Clinical Chemistry

Valdes,, Roland;Jortani, Saeed, A

doi: 10.1093/clinchem/44.9.1883pmid: N/A

Digoxin is one of the few therapeutic drugs for which antidotal therapy is available (1) . Administration of Fab fragments has successfully been used to reverse the effects of life-threatening digoxin overdoses for over 20 years. Digoxin-specific Fab fragments are produced by cleaving sheep digoxin-specific IgG with papain to form two Fab fragments (50 000 Da each) and one 50 000-Da Fc fragment (2) . As determined by equilibrium dialysis with one such preparation of Fab fragments, the intrinsic affinity constant for binding of digoxin was 1010 L/mol (3) . Depending on the experimental conditions used (e.g., K+ concentration and type of membrane isolated), the affinity constant for binding of digoxin to its receptor (the sodium pump) has been reported to be in the range 106-108 L/mol (4)(5) . The rationale for using Fab fragments to reduce toxic effects of digoxin is their greater affinity for digoxin when compared with the sodium pump. As a consequence, the extracellular unbound digoxin concentration is lowered, and further equilibration of receptor-bound digoxin with the fluid in the extracellular space leads to the eventual release of digoxin from its receptor sites. The affinity of digitoxin for binding to Fab fragments is 10-fold lower than that of digoxin; however, it is high enough to allow clinical utility of Fab fragments in cases of digitoxin intoxication as well (3) . The dose of Fab fragments should be approximately equimolar to the total body digoxin load, which is determined according to the serum digoxin concentration and/or patient’s medical history (6) . Digibind® (Glaxo Wellcome) is the most common brand of anti-digoxin Fab fragments used in the United States and worldwide. Boehringer-Mannheim also produces anti-digoxin Fab fragments intended for use as an antidote. Anti-digoxin Fab fragments have been suggested and used as antidotes in intoxications caused by other digitalis-like molecules such as oleandrin (7)(8)(9) , bufadienolide-containing aphrodisiacs (10) , digitoxin ((11)), and foxglove extract (12) ; they have also been successfully used to reverse hypertension believed linked to increased digoxin-like immunoreactive factors in blood (13) . In 90% of digitalis-induced intoxicated patients, a median initial favorable response time of 19 min was achieved after administration of Digibind (14) . Hyperkalemia associated with digitalis poisoning is also reversed after administration of Fab fragments (2) . Fab fragments have a volume of distribution of 0.4 L/kg, an elimination half-life of 16–20 h, and a systemic clearance of 0.324 mL · min−1 · kg−1 (15) . Both renal and nonrenal (i.e., metabolism and/or removal by the reticuloendothelial system) routes are responsible for the elimination of Fab fragments with ∼65% of both Fab as well as Fab-digoxin complexes eliminated by the kidneys (15) . Such complexes are not removed by hemodialysis or even by continuous arteriovenous hemofiltration (16) . In patients with reduced or no renal function, Fab fragments can remain in plasma for 2–3 weeks after administration because of decreased renal elimination. After administration of Fab fragments, there is approximately a 10- to 30-fold increase in total digoxin concentration in the serum, whereas the unbound fraction of the drug decreases rapidly (17) . Because the unbound fraction of digoxin is the pharmacologically active form, its accurate and reliable measurement in serum collected at various times after administration of Fab fragments may be clinically important (18) . The rationale here is that rapid removal of all bioactive digoxin from a patient in need of the drug is not clinically efficacious, and an ability to titer the unbound digoxin would seem optimal. Although no definitive evidence has yet been presented, the concept seems rational. Furthermore, the pharmacokinetics of digoxin after administration of Fab fragments becomes dependent on the disposition of Fab (19) . Monitoring of unbound digoxin concentrations after administration of Fab fragments may be appropriate to establish recrudescent toxicity in renal failure patients, to assess the need for more Fab to be administered, and to establish the need for reintroduction of digoxin (15) . However, specimens collected from patients treated with digoxin Fab fragments usually give grossly erroneous and misleading values for digoxin concentrations by most immunoassays (20) . Assay antibody and Fab fragments have similar affinities for digoxin and the tracer. Therefore, the most likely mechanism of interference of Fab fragments in digoxin assays is by binding of the assay tracer components. In fact, over the last few years, over a dozen publications have addressed the issue of discrepant digoxin values caused by the presence of this antidote in blood. One study in particular noted discrepancies between several immunoassays for digoxin as long as 14 days after a patient had been treated with Fab (21) . Generally, immunoassays involving wash steps are less prone to interference by Fab fragments because the unbound Fab fragments are removed before the addition of tracer during the wash step (17) . In some digoxin immunoassays that require acid precipitation of proteins, pretreatment of sample with acid releases the Fab-sequestered digoxin; subsequently, total digoxin present in the serum including the bound and the unbound fractions is measured. This usually gives very large measured values. There appears to be no clinical value in measuring total serum digoxin concentrations in patients treated with the antidote. One method for measuring the unbound digoxin is the use of ultrafiltration before the immunoassay. Although this procedure is cumbersome, it has been used successfully (18)(22) . In the study reported in this issue, Ocal and Green (23) have used the Stratus and AxSYM digoxin immunoassays to analyze three samples from a patient treated with Digibind. In addition, they also analyzed several samples to which Digibind was added at different amounts to a sample containing digoxin. These authors concluded that there was good correlation between these assays in measuring unbound digoxin in the presence of Digibind. We believe that ultrafiltration of a portion of these samples followed by measurement of the unbound digoxin by an immunoassay (previously shown to be free of matrix-dependent error) might have been used to monitor the accuracy of these immunoassays in measuring unbound digoxin in the presence of Fab fragments. The problem of bias caused by matrix in the immunoassays used for measuring the concentration of digoxin in the ultrafiltrate has been previously recognized (24) . Ujhelyi et al. ((17)) have also reported a mean prediction error of 0.62 μg/L (0.54–0.70 μg/L) for the Baxter Dade Stratus digoxin immunoassay as a result of bias between ultrafiltrate and serum matrices. Because the Abbott TDx fluorescence polarization immunoassay digoxin assay measures the sample after protein precipitation, the measured matrix is protein-free. Therefore, we agree with the recommendation by Ujhelyi et al. of using that assay to monitor the unbound digoxin concentration in the ultrafiltrates as reference in their study. However, just because the Stratus and the AxSYM results for measuring unbound digoxin in Fab-treated patients correlate, one cannot conclude that the results are accurate as well. Therefore, as recommended in the recently established Guidelines for Therapeutic Drug Monitoring developed by the National Academy of Clinical Biochemistry (24) , the authors might have prepared a pool of digoxin-free serum ultrafiltrate. Then varying concentrations of digoxin could have been added to the ultrafiltrate and the serum followed by analysis for digoxin. If results for the serum and the ultrafiltrate were statistically equivalent, the assay would have been better documented as suitable to monitor digoxin in the ultrafiltrates of the correlation samples. Serum samples to which various concentrations of digoxin and Fab fragments have been added may not accurately represent samples collected from antidote-treated patients. Ocal and Green (23) have used three samples from a Digibind-treated patient in addition to several samples to which Fab was added by titrating their stock digoxin sample. We commend them for using actual patient samples for the correlation studies; however, use of an appropriate reference to monitor accuracy of their results would have buttressed their conclusions. Despite several reports in the literature, including the current paper by Ocal and Green (23) on the methodology for measuring unbound digoxin in Fab-treated patients, the issue of therapeutic range has not been adequately addressed. It is not clear if the fraction of digoxin bound to albumin (∼20% of that in serum) is also measured as “free” by the immunoassays attempting to directly measure the pharmacologically active fraction of digoxin in the serum. Until this question is addressed, we believe that the current therapeutic ranges for digoxin should be used with caution in patients treated with the antidote; this point should be considered before redigitalizing Fab-treated digoxin-overdosed patients. An important question is whether the concept of treating poisoned patients with Fab raised against other drugs or poisons can be more generally applied beyond digitalis intoxication. Thorough discussion of this topic is beyond the scope of this editorial. Nevertheless, the following fundamental questions need to be addressed as part of assessing if the unbound concentration of digoxin or any other drug can be clinically useful after treatment of an overdose with a binding agent. First, should such monitoring be performed in all patients treated with Fab fragments or should it be reserved for selected patients such as those with renal failure? Second, what reference interval should be used for the unbound digoxin in Fab-treated patients? Third, what controls should be used (i.e., would regular controls containing only digoxin suffice in monitoring the performance of the assay in presence of Fab)? Fourth, how often and how long after administration of Fab fragments should unbound digoxin concentrations be monitored? Fifth, because Fab fragments are also used to treat other cardiac glycoside poisonings, such as from ingestion of oleander plants, can monitoring serum in terms of digoxin equivalents be useful in assessing the antidotal therapy? We believe that both the analytical and clinical issues related to monitoring unbound digoxin in the presence of Fab fragments be done with these questions in mind and thank Ocal and Green for bringing this important issue once again to the forefront of laboratory medicine. References 1 Smith TW, Haber E, Yeatman L, Butler VP. Reversal of advanced digoxin intoxication with FAB fragments of digoxin-specific antibodies. New Engl J Med 1976 ; 294 : 797 -800. Crossref Search ADS PubMed 2 Wegner TL, Butler VP, Haber E, Smith TW. Treatment of 63 severely digitalis-toxic patients with digoxin-specific antibody fragments. J Am Coll Cardiol 1985 ; 5 : 118A -123A. Crossref Search ADS PubMed 3 Smith TW, Butler VP, Haber E, Fozzard H, Marcus FI, Bremner F, et al. Treatment of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragments, experience in 26 cases. New Engl J Med 1982 ; 307 : 1357 -1362. Crossref Search ADS PubMed 4 Noel F, Fagoo M, Godfraind T. A comparison of the affinities of rat (Na+-K+)-ATPase isoenzymes for cardioactive steroids, role of lactone ring, sugar moiety and KCl concentration. Biochem Pharmacol 1990 ; 40 : 2611 -2616. Crossref Search ADS PubMed 5 Schmidt TA, Svendsen JH, Haunso S, Kjeldsen K. Quantification of the total Na,K-ATPase concentration in atria and ventricles from mammalian species by measuring 3H-ouabain binding to intact myocardial samples. Stability to short term ischemia reperfusion. Basic Res Cardiol 1990 ; 85 : 411 -427. Crossref Search ADS PubMed 6 Hursting MJ, Raisys VA, Ophelm KE, Bell JL, Trobaugh GB, Smith TW. Determination of free digoxin concentrations in serum for monitoring Fab treatment of digoxin overdose. Clin Chem 1987 ; 33 : 1652 -1655. Crossref Search ADS PubMed 7 Clark RF, Selden BS, Curry SC. Digoxin-specific Fab fragments in the treatment of oleander toxicity in a canine model. Ann Emerg Med 1991 ; 20 : 1073 -1077. Crossref Search ADS PubMed 8 Shumaik GM, Wu AW, Ping AC. Oleander poisoning: treatment with digoxin-specific Fab antibody fragments. Ann Emerg Med 1988 ; 17 : 732 -735. Crossref Search ADS PubMed 9 Safadi R, Levy I, Amitai Y, Caraco Y. Beneficial effect of digoxin-specific Fab antibody fragments in oleander intoxication. Arch Intern Med 1995 ; 155 : 2121 -2125. Crossref Search ADS PubMed 10 Pierach CA. Digoxin-like toxicity and death from a purported aphrodisiac. JAMA 1996 ; 275 : 988 . PubMed 11 Kurowski V, Iven H, Djonlagic H. Treatment of a patient with severe intoxication by Fab fragments of anti-digitalis antibodies. Intensive Care Med 1992 ; 18 : 439 -442. Crossref Search ADS PubMed 12 Rich SA, Libera JM, Locke RJ. Treatment of foxglove extract poisoning with digoxin-specific fab fragments. Ann Emerg Med 1993 ; 22 : 1904 -1907. Crossref Search ADS PubMed 13 Kreps HH, Graves SW, Price DA, Lazarus M, Ensign A, Soszynski PA, Hollenberg NK. Reversal of sodium pump inhibitor induced vascular smooth muscle contraction with digibind. Stoichiometry and its implications. Am J Hypertension 1996 ; 9 : 39 -46. Crossref Search ADS 14 Antman EM, Wenger TL, Butler VP, Jr, Haber E, Smith TW. Treatment of 150 cases of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragments. Circulation 1990 ; 81 : 1744 -1752. Crossref Search ADS PubMed 15 Ujhelyi MR, Robert S. Pharmacokinetic aspect of digoxin-specific Fab therapy in the management of digitalis toxicity. Clin Pharmacokinet 1995 ; 28 : 483 -493. Crossref Search ADS PubMed 16 Berkovitch M, Akilesh MR, Gerace R. Acute digoxin overdose in a newborn with renal failure: use of digoxin immune Fab and peritoneal dialysis. Ther Drug Monit 1994 ; 16 : 531 -533. Crossref Search ADS PubMed 17 Ujhelyi MR, Green PJ, Cummings DM, Robert S, Vlasses PH, Zarowitz BJ. Determination of free serum digoxin concentrations in digoxin toxic patients after administration of digoxin Fab antibodies. Ther Drug Monit 1992 ; 14 : 147 -154. Crossref Search ADS PubMed 18 George S, Braithwaite RA, Hughes EA. Digoxin measurements following plasma ultrafiltration in two patients with digoxin toxicity treated with specific Fab fragments. Ann Clin Biochem 1994 ; 31 : 380 -382. Crossref Search ADS PubMed 19 Sinclair AJ, Hewick DS, Johnston PC, Stevenson IH, Lemon M. Kinetics of digoxin and anti-digoxin antibody fragments during treatment of digoxin toxicity. Br J Clin Pharmacol 1989 ; 28 : 352 -356. Crossref Search ADS PubMed 20 Rainey PM. Effects of digoxin immune Fab (ovine) on digoxin immunoassays. Am J Clin Pathol 1989 ; 92 : 779 -786. Crossref Search ADS PubMed 21 Miller JJ, Straub RW, Valdes R, Jr. Analytical performance of a monoclonal digoxin assay with increased specificity on the ACS:180. Ther Drug Monit 1996 ; 18 : 65 -72. Crossref Search ADS PubMed 22 Lemon M, Andrews DJ, Binks AM, Georgiou GA. Concentrations of free serum digoxin after treatment with antibody fragments. Br Med J 1987 ; 295 : 1520 -1521. Crossref Search ADS 23 Ocal IT, Green TR. Serum digoxin in the presence of Digibind: determination of digoxin by the Abbott AxSYM and Baxter Stratus II immunoassays by direct analysis without pretreatment of serum samples. Clin Chem 1998 ; 44 : 1947 -1950. Crossref Search ADS PubMed 24 Valdes R, Jr, Jortani SA, Gheorghiade M. Standards of laboratory practice: cardiac drug monitoring. Clin Chem 1998 ; 44 : 1096 -1109. Crossref Search ADS PubMed © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Wright, Anthony J, A;Finglas, Paul, M;Southon,, Susan

doi: 10.1093/clinchem/44.9.1886pmid: N/A

Abstract Neural tube defects can be prevented by adequate intake of periconceptional folate, and inverse associations between folate status and cardiovascular disease and various cancers have been noted. Thus, there is renewed interest in the analysis of red cell folate (RCF) as an indicator of folate deficiency risk. Assessment of the assumptions that underpin RCF assays indicates that many are false. Published literature suggests that increased deoxy-hemoglobin (which can bind RCF electrostatically) yields more assayable folate, and increased oxy-hemoglobin (which cannot bind RCF) yields less assayable folate. It is argued that as deoxy-hemoglobin picks up oxygen and switches quaternary structure, any bound folate must, on purely theoretical grounds, become physically “trapped”. Venous blood taken for analysis is 65% to 75% saturated with oxygen, and pro-rata “trapping” will lead to serious underestimation of RCF. Hence, doubt is cast over the validity of all previous RCF values. Some strategies for accurately assessing RCF are suggested. RCF, red cell folate, Hb, hemoglobin, MA, microbiological assay, RFPB, radio-folate-binding protein, AA, ascorbic acid, PteGlu, pteroylmonoglutamic acid (folic acid), 5CH3-H4-PteGlun, 5-methyltetrahydrofolic acid polyglutamates, 5CH3-H4-PteGlu, 5-methyltetrahydrofolic acid monoglutamate, DPG, 2,3-diphosphoglycerate, T-, tense, R-, relaxed. Spina bifida and related neural tube defects of anencephaly and encephalocele are major causes of perinatal, infant, and childhood mortality and morbidity. Although screening (maternal serum α-fetoprotein estimation or fetal ultrasound examination) has led to a fall in the number of children born with neural tube defects through termination of pregnancy (1)(2) , it cannot prevent the development of these defects or the associated parental distress. There is now firm evidence that the majority of neural tube defects can be prevented by an adequate intake of periconceptional folate, a B-group vitamin found especially in leafy green vegetables, some other vegetables and fruits, and whole-grain cereals. All studies (case-control and cohort studies; nonrandomized and randomized controlled trials) (3) showed a reduction in risk with increased intake of either dietary folates or supplemental folic acid. These studies have led to renewed interest in the analysis of red blood cell folate (RCF)1 as an indicator of folate deficiency risk in women of child-bearing age. Additional interest in the analysis of RCF as an indicator of folate deficiency risk in the general population is generated by the inverse associations between folate status and the risk of cardiovascular disease (4) and various cancers (5)(6) . Because of the increasing importance of folate nutrition to public health, a “round robin” interlaboratory comparison study was conducted to assess differences among methods (7) . This exercise, as others before it (8)(9) , demonstrated the large intraassay and interassay variation in whole-blood folate analysis. There are discussions in the literature about method comparisons (8)(10)(11) and possible sources of variability (9)(12)(13)(14)(15)(16) . However, to date, the influence of the degree of hemoglobin (Hb) oxygenation on the estimation of RCF concentration has not been assessed critically. The degree of Hb oxygenation, however, could be of prime importance with respect to the accuracy and stability of both the microbiological (MA) and radio-folate-binding protein (RFBP) assays. This report questions currently held assumptions relating to RCF analysis and then evaluates critically the potential for gross inaccuracy in RCF analysis resulting from differences in the oxygenation state of Hb as a sample progresses through the analytical procedure. current method of rcf analysis Venous whole-blood samples are transferred immediately to either EDTA or lithium-heparin coated tubes. A subsample is then hemolyzed by 10-fold dilution in a hypotonic aqueous solution [usually a freshly prepared solution of 1% ascorbic acid (AA) (7)(9)(10) ]. The remaining blood is sent for general hematological analysis, which includes a hematocrit or packed cell volume. Subsequent to deconjugating the hemolysate folate (generally agreed to be polyglutamates of 5-methyltetrahydrofolic acid; 5CH3-H4-PteGlun) with plasma folate conjugase (γ-glu-x carboxypeptidase; EC 3.4.19.9) associated with the “whole-blood” sample, the folate monoglutamate (5CH3-H4-PteGlu) concentration is assessed by either MA or RFBP assays, which generally use folic acid (pteroyl monoglutamic acid; PteGlu), the parent and fully oxidized folate form, as the assay calibrant. The final answer is expressed (using the hematocrit or packed cell volume value of the original sample) as a folate concentration (μg/L or nmol/L) per liter of red blood cells. assumptions We make the following assumptions: 1. That 10-fold dilution of a whole-blood sample is sufficient to ensure complete lysis of all red blood cells; 2. That total lysis of all red blood cells will liberate erythrocyte folate completely to unfettered access by plasma folate conjugase and consequently ensure complete RCF deconjugation to the monoglutamate form before analysis by either MA or RFBP assay; 3. That allowing deconjugation to proceed for an excess of time, which would allow assayable folate concentration to plateau, is synonymous with ensuring complete deconjugation; 4. That AA is used in the hemolysate diluent as an “anti-oxidant”, which will protect the assay from folate loss, because RCF exists in chemically reduced forms (5CH3-H4-PteGlun) that can potentially undergo oxidation; 5. That folate analysis is not affected by an acidic hemolysate pH causing perturbation in the pH of the terminal MA or RFBP assay, which would lead to an inappropriately increased/decreased response to deconjugated RCF (5CH3-H4-PteGlu) in comparison with the assay calibrant PteGlu; 6. That the MA or RFBP assay of folate is not affected by RCF binding to any erythrocyte or plasma component; and 7. That RCF assays are not affected by the degree of oxygenation of Hb in the hemolysates. evaluation of these assumptions 1. Red cell membranes enclose high concentrations of impermeant anions, principally Hb (5 mmol/L), 2,3-diphosphoglycerate (DPG; 5 mmol/L), and glutathione (2–3 mmol/L), and any osmotic imbalance creates an intolerable osmotic gradient, allowing the instant entry of water. Erythrocytes, when placed in a hypotonic medium, can swell by 60% to 70% before the critical hemolytic volume is reached, and the stretched membrane acquires transient holes that cause leakage of cellular contents (17) . In our experience, visual inspection under a microscope confirms that all erythrocytes are hemolyzed when whole blood is diluted 10-fold with an aqueous hypotonic solution (e.g., 1% AA). 2. Knowing that 10-fold dilution of whole blood will produce total lysis is not synonymous with the assumption that there is complete liberation of RCF to unfettered access by plasma folate conjugase. Such an assumption would be false if there were any evidence for folate being bound to any erythrocyte or plasma component (see evaluation of assumption 6). 3. The optimal deconjugation of folyl-polyglutamates with folate conjugase enzyme proceeds as a function of buffer, pH, amount of enzyme, temperature, and reaction time (18) . Subsequent to erythrocyte hemolysis in 1% AA, RCF is deconjugated with naturally occurring plasma folate conjugase enzyme, with reaction time being the only variable utilized, at any given temperature, to optimize assayable folate. However, even if assayable folate is optimized, this does not infer that all RCF has been deconjugated. It only infers that all RCF made available to the plasma conjugase enzyme has been deconjugated. 4. The use of AA in hemolysates goes back to the 1950s. After it was observed that protein-free dialysates of whole blood had up to 100 times the microbiological folate activity of the same blood assayed after diluting with water and heating (19) , it was further demonstrated that if bloods were diluted with a buffered solution containing AA and then autoclaved to precipitate proteins, folate activity comparable with that after dialysis was obtained in the clear protein-free supernatant solution (20) . The addition of AA to the MA does not produce a higher growth response to the assay calibrant (PteGlu) (21) , and hence the increase in RCF MA response is not an artifact of an underlying additional need for ascorbic acid to be incorporated into the assay growth medium. Because RCF exists in reduced forms that can potentially undergo oxidation, it has always been assumed that the only role of AA in RCF assays is as an antioxidant. The ubiquitous presentation of this theory has, inevitably, closed minds to the possibility of alternative theories by which the addition of AA could produce an increased MA (or RFBP) response. At this point, it may be pertinent to note that although AA technically has the ability to act as a prooxidant (especially in the presence of metal ions), the addition of AA to hemolysates has always resulted in an increase of assayable folate, thus ruling out the probability of it acting as a prooxidant in the context of erythrocyte folate analysis. Data from a recent report (22) raise questions about the validity of the antioxidant role, at least over short-term hemolysate storage intervals, and hence compel examination of the merits of alternative theories. MA RCF values have been compared after various periods of hemolysate storage, in whole-blood diluted conventionally (×10) with fresh 1% AA (pH 2.8), and the same blood diluted ×10 with 1% AA adjusted to pH 6.0 (AA-pH 6.0). After 1 day of storage, AA (pH 2.8) hemolysates had an assayable folate concentration 65% higher than AA-pH 6.0 hemolysates (22) . For such a short period of storage, it can be argued that ascorbate ions would be present in both lysates at sufficient concentrations to exert an antioxidant influence, and therefore, the higher RCF values cannot be attributed to the antioxidant role of AA. Thus, any theory that attributes an antioxidant role as the only effect of AA (pH 2.8) addition to hemolysates becomes untenable. Dilution of whole blood with 1% AA (pH 2.8) lowers hemolysate pH to ∼3.6. Human plasma folate conjugase (γ-glu-x carboxypeptidase; EC 3.4.19.9) exhibits maximum activity at pH 4.5 but functions across a broad spectrum from about pH 3.5–7.5 (23) . Therefore, the addition of AA does not produce a hemolysate pH optimally appropriate for the action of plasma folate conjugase, any more than does dilution of neutral whole blood with water. 5. Both MA and RFBP assay response to deconjugated RCF (5CH3-H4-PteGlu) is pH dependent (12)(15) . If the assay pH is not well chosen, these assays can display a disparate response toward 5CH3-H4-PteGlu, as opposed to their PteGlu calibrant. In our experience, the assay medium of the MA (at pH 6.2) is not perturbed by the appropriate addition of hemolysates in 1% AA. Commercial RFBP assays used currently are usually well buffered at their optimum pH, and it is unlikely that the addition of a small amount of an acidic hemolysate would compromise these assays. However, slight shifts in pH have been observed during sample processing using some RFBP assays (9) , and therefore the onus is on users to verify the pH stability of their chosen kit before use. 6. A major clue as to why RCF values are higher when 1% AA (pH 2.8) is used as the diluent for whole blood is provided by the fact that assayable folate concentration in 1% AA-pH 6.0 lysates, although initially lower, increased during storage (22) . This prompted the conclusion that the initial discrepancy was because of folate being bound rather than being destroyed. In reverse, the explanation for obtaining higher RCF values when using 1% AA (pH 2.8) as whole-blood diluent must, therefore, be because it releases at least some folate that would otherwise be bound. Therefore, the assumption that the availability of hemolysate folate for terminal MA or RFBP assay is not affected by folate binding to any erythrocyte or plasma component is false. Furthermore, it is clear from research published previously (23)(24)(25)(26)(27) that folate can be bound by Hb; indeed, it has been argued that most, if not all, RCF in venous blood samples is bound to Hb. 7. A change in hemolysate pH could initiate the “Bohr effect” (28)(29) , where the capacity of Hb to carry oxygen is decreased remarkably by acidity; i.e., Hb is deoxygenated. One report in the literature (30) (the implications of which appear to have been overlooked) reported that the degree of Hb oxygenation altered RCF amounts such that deoxygenation increased, and oxygenation decreased, assayable folate concentration in both MA and RFBP assays. This phenomenon was shown to be totally reversible. The direct binding of folate to Hb was suggested both by the fact that the degree of oxygenation did not affect plasma folate analysis and by experiments in which the rise with deoxygenation was not seen in red cells exposed previously to carbon monoxide or treated with cyanate. The same study showed that assayable RCF concentration was generally higher in venous blood samples, where blood would have been less saturated with oxygen, when compared with arterial blood samples. It is clear from the evaluation of the assumptions that underpin current RCF assays that many are false. However, before considering the implications of RCF binding to Hb and the effect of Hb oxygenation, it is first necessary to have an appreciation of the general structure and function of Hb. the structure, function, and allostery of Hb The Hb of most human adults consists (28)(29) of four folded (tertiary structure) myoglobin-like polypeptide chains: two α (141 amino acid residues) and two β (146 amino acid residues). Each polypeptide subunit contains a heme group, in which the iron atom is in the ferrous [Fe(II)] state, and a single oxygen-binding site. The subunits affiliate in pairs as αβ heterodimers. The Hb molecule is nearly spherical, with a diameter of 5.5 nm, and has a quaternary tetramer structure composed of two heterodimers (α1β1 and α2β2) that associate with twofold symmetry in an approximate tetrahedral array such that each α chain is in contact with both β chains. In contrast, there are few interactions between the two α chains or between the two β chains. The α1β1 and α2β2 pairs, being made up of irregularly shaped polypeptide chains, do not fit each other precisely; through the center of the Hb molecule, along the axis of heterodimer symmetry (the molecular dyad axis) runs a central cavity that is populated by a variety of polar side chains. The cavity extends for a depth of 5 nm, with entry points at both the top (α1α2 entrance) and the bottom (β1β2 entrance) of the molecule. Its shape in horse oxy-Hb (which is very similar to human oxy-Hb) has been described (31) as most easily represented by two boxes, each ∼2 nm long, 0.8 nm wide, and 2.5 nm deep (i.e., one-half the depth of the Hb molecule). One box separates the α chains from each other and the other the β chains. The two boxes are set one above the other with their 2-nm-long axes at right angles to the dyad axis and to each other, each box being open at the top and bottom. Hb is an allosteric protein that undergoes conformational change in quaternary structure from the T (tense) state (deoxy-Hb) to the R (relaxed) state (oxy-Hb) upon binding its first oxygen molecule such that the binding of additional oxygen molecules is enhanced. In other words, oxygen binds cooperatively to Hb. In contrast to myoglobin, which has a hyperbolic oxygen dissociation curve and 50% saturation at 1 torr oxygen pressure, Hb has a sigmoidal dissociation curve and 50% saturation at 26 torr. This lower affinity of Hb for oxygen gives it the latitude to be modified. Increasing concentrations of hydrogen ion and CO2 promote the release of oxygen (the Bohr effect). An important ligand of human Hb is DPG (32) . This highly anionic organic phosphate is present in human erythrocytes at approximately the same molar concentration as Hb and lowers profoundly the oxygen affinity of Hb by cross-linking deoxy-Hb. Hb stripped of DPG loses its sigmoidal oxygen-binding relationship and exhibits an affinity for oxygen similar to myoglobin (29) . DPG, with which folyl-polyglutamate can compete (27) , binds to deoxy-Hb just inside the entrance to the β1β2 cavity (with a stoichiometry of 1-DPG/Hb tetramer) because it is stereochemically complementary to a constellation of six positively charged groups facing the central cavity of the Hb molecule. DPG cannot bind to oxy-Hb because the entrance to the cavity is reduced greatly as Hb switches from the deoxy- to the oxy-molecular state. In humans, it is DPG that promotes the sigmoidal oxygen-binding relationship with Hb, and its presence is essential for the release of oxygen in the tissues, allowing the discrepancy in oxygen saturation between arterial (about 96% saturated) and venous blood (about 64% saturated) (28) . In the transition from deoxy- to oxy-Hb, large structural changes take place at the α1β2 contact but only small ones at the α1β1 contact. The α1β1 heterodimer rotates, relative to the α2β2 heterodimer, by 15° and, at the same time, the two heterodimers move closer together. As a result, the oxy-Hb molecule has a more compact structure than deoxy-Hb, and the central cavity becomes smaller with the “box” separating the β chains, reducing from an open-ended slot about 0.8 nm wide to a box about 1.8 nm long and 0.5 nm wide (33) . In combination with a slight alteration in tertiary structure, where the A and H helices of the two β chains move closer together, the β1β2 cavity in oxy-Hb becomes too small to accommodate DPG and, thus, expels it. the binding of folate to Hb It is clear from published research (24)(25)(26)(27) that folate can be bound by Hb and that most, if not all, RCF in venous blood samples could be associated electrostatically with deoxy-Hb. Folyl-polyglutamate binds within the β1β2 cavity, with an affinity similar to DPG (25) , to deoxy-Hb tetramers with all three structural elements (pteridine moiety, p-aminobenzoyl portion, and glutamyl groups) contributing to the binding energy (27) . The affinity increases with the number of glutamyl residues. Because the glutamate residues run through the central cavity, only one molecule can be bound per deoxy-Hb tetramer. Unlike DPG, which electrostatically binds at the entrance to the cavity, the pteroyl and p-aminobenzoyl groups of the folyl-polyglutamate are buried deep within the central cavity, nestled against an interior edge of the α1β1 interface, with the glutamate residues remaining at the entrance to the cavity where the first two or three glutamates interact with the DPG binding site. It is suggested that at least the fourth and higher glutamate residues exit the central cavity through the DPG binding site and either extend beyond the deoxy-Hb tetramer into the bulk solvent or, more likely, bind in a disordered way to the basic residues on the tetramer surface (27) . Folyl-polyglutamates do not bind to oxy-Hb tetramers, although removal of the pteridine moiety produces binding of the resulting p-aminobenzoylpolyglutamate residues (26) . The (obvious) conclusion drawn is that the bulky pteridine residue is too large to penetrate the smaller β1β2 entrance to the central cavity, which arises as Hb switches from the T (deoxy-) to the R (oxy-) state, thus preventing the access of folate to the Hb-folate binding sites. Hb-folate binding and rcf analysis: a paradox? Removal of oxygen from oxy-Hb either directly [with N2 (30) ] or indirectly [by increasing hemolysate acidity with AA (22) ] produces more deoxy-Hb. However, in spite of the fact that it is deoxy-Hb that clearly binds folate (26) , assayable RCF concentration increases. Conversely, direct oxygenation of Hb (30) produces a remarkable fall in assayable RCF, even though it is clear that the oxy-Hb tetramer does not bind folate (26) . It is easy to assume that an increase in the noncovalent electrostatic binding of RCF to deoxy-Hb should yield a decrease in assayable RCF and that, conversely, an increase in oxygenated Hb should, automatically, produce folate release and an increase in assayable RCF. However, although RCF binds to deoxy-Hb, this is not an impediment to analysis (30) . The binding affinity of deoxy-Hb for polyglutamate folate is low, and it is only the high binding capacity of Hb (there are between 5 000 and 10 000 Hb molecules for every folate molecule) that leads to the conclusion that most, if not all, RCF is bound to deoxy-Hb (24)(25)(26) at the oxygen saturation (64% to 75%) (26)(28) associated with venous blood samples. Consequently, it is quite easy to envisage the scenario by which folate bound to deoxy-Hb is assayable. Once red blood cells have been lysed, plasma conjugase will compete for the pteroyl-polyglutamate (5CH3-H4-PteGlun) that is bound (with low affinity) to deoxy-Hb. Because most of the polyglutamate residues of the red cell folate reside outside of the entrance to the β1β2 central activity of deoxy-Hb, they are always accessible to plasma conjugase and the sequential removal of glutamyl residues, which will produce decreased folate binding affinity. Consequently, even if not initially successful in wrestling folate away from deoxy-Hb, the removal of glutamate residues from RCF by plasma conjugase will accelerate the dis-association of folate from deoxy-Hb and help to ensure complete deconjugation of RCF to the monoglutamate form (5CH3-H4-PteGlu). The monoglutamate form of the parent compound (PteGlu) has little affinity for binding to deoxy-Hb when compared with its polyglutamate counterparts (27) , and because the addition of a methyl group has been shown to reduce binding affinity still further (27) , it could be argued that both MA or RFBP should be able to assay fully deconjugated RCF in the presence of “folate-stripped” deoxy-Hb without the problem of substantial folate rebinding. Why oxygenation of Hb should produce a substantial decrease in assayable RCF (30) , even though it is clear that the oxy-Hb tetramer does not bind folate (26) , is puzzling. Although experimental evidence shows clearly that Hb, preexisting in the oxy-Hb state, does not bind folyl-polyglutamate (26) , this does not mean that oxy-Hb cannot trap folate. Hb, although permanently in the presence of RCF, circulates in the body continuously, picking up oxygen in the lungs and depositing it in tissue. Returning, partially deoxygenated venous blood (conditions where RCF would be bound to deoxy-Hb) is pumped from the right ventricle of the heart via the pulmonary artery to the lungs. As soon as a deoxy-Hb molecule picks up its first molecule of oxygen, its quaternary structure abruptly snaps from the T-state to the R-state. If a deoxy-Hb molecule is carrying a molecule of bound folate when it converts to the R-state, then the folate must be trapped in the tetramer cavity for the very same reason that folate cannot enter the cavity of a preexisting oxy-Hb molecule to be bound, i.e., the bulky pteridine moiety will prevent passage through the narrowed entrance of the β1β2 cavity present in the new quaternary structure. It is therefore irrelevant whether the pteridine, p-aminobenzoyl, and glutamate electrostatic bonding of the folate molecule with Hb is disrupted instantaneously, such that the folate molecule becomes dissociated from the Hb tetramer, because the folate molecule must remain trapped until the Hb molecule becomes fully deoxygenated once again. If true, previous conclusions that most, if not all, RCF in venous blood would be associated with deoxy-Hb, and not with oxy-Hb, should be re-assessed. implications The implications for RCF assay are important, because even venous blood has substantial partial oxygen saturation (64% to 75%) (26)(28) . Assuming, in the best-case scenario, that the percentage of oxy-Hb is no higher than the percentage of partial oxygen saturation, if RCF is “trapped” pro-rata to the proportion of oxy-Hb in venous blood, then RCF concentration could be underassayed by a factor of at least three- or fourfold. This casts serious doubt over all previously published RCF values. One report (30) , in which whole blood was flushed with nitrogen, indicated that assayable RCF may only be increased up to twofold. However, deoxygenated red blood cells were subsequently lysed with an AA solution within which air (oxygen) was dissolved, and hence substantial amounts of RCF could have been retrapped before being fully deconjugated. Importantly, deoxy-Hb (which binds RCF) may keep folate in a state that can be successfully deconjugated and analyzed. The novel interpretation of folate-trapping by oxy-Hb, presented in this report, questions the effectiveness of the current method for RCF analysis. the way forward Although, superficially at least, it would seem easy to collect additional, practical evidence in support of our new theory, such thoughts would be naive. Even the simple step of producing a hemolysate creates a serious problem in that the concentration of naturally occurring allosteric effectors fall 10-fold, shifting the equilibrium of Hb to the R-state, which results in Hb tetramers having a much greater avidity for oxygen when in free solution than when originally contained within erythrocytes. This physical phenomenon must not only be neutralized, but the original affinity of Hb for oxygen must be reversed effectively; but how? Strategies need to be developed that allow RCF to be deconjugated while all Hb is in the T-state quaternary structure; this will usually, but not exclusively (34)(35) , mean that all Hb will need to be in the deoxy-Hb state. Venous whole-blood samples could be equilibrated with a gas (e.g., 100% N2, 100% CO2) until the partial pressure of O2 is reduced to a minimum. However, although it may be efficient at preventing Hb from carrying oxygen, CO should not be considered because it may mimic the allosteric effect of oxygen and cause Hb to snap to the R-state. This could itself provoke the trapping of RCF. Because AA could protect hemolysates that are destined to be stored before analysis from oxidation, 10-fold dilution of deoxygenated blood (to engender complete lysis) should probably still be carried out with a solution of 1% AA, provided that the diluent is deoxygenated immediately before use. To avoid reoxygenation during the subsequent deconjugation step, hemolysates could be placed in containers, flushed with an appropriate gas, and sealed with a gas-tight closure. The addition of chloride ions (a newly discovered allosteric effector) (36) or competitors for deoxy-Hb/folate binding such as DPG or inositol hexaphosphate [which has an even greater binding affinity and, thus, is even more potent than DPG in reducing the oxygen affinity of Hb (28)(32) ] to the lysate diluent (1% AA) should be considered. A search should be made for compounds that not only have a high binding affinity for Hb but can also cross-link the tetramer to permanently retain Hb in the T-state quaternary structure. In the interim, the usefulness of compounds that are already known to be synergistic with DPG (37)(38) should be examined. Institute of Food Research, Norwich Research Park, Colney, Norwich NR4 7UA, UK. Statement of interest: The authors have no commercial interests. The Institute of Food Research is a company limited by guarantee, with charitable status (Reg. Charity No. 1058499), grant aided by the Biotechnology and Biological Sciences Research Council (BBSRC). References 1 Office of Population Census and Surveys. Congenital Malformation Statistics, Series MB3, No. 7. London: Her Majesty’s Stationery Office, 1991.. 2 Chan A, Robertson EF, Haan EA, Keane RJ, Ranieri E, Carney A. Prevalence of neural tube defects in South Australia, 1966–1991: effectiveness and impact of prenatal diagnosis. Br Med J 1991 ; 307 : 703 -706. 3 Bower C, Stanley FJ. Issues in the prevention of spina bifida. J R Soc Med 1996 ; 89 : 436 -442. Crossref Search ADS PubMed 4 Wilcken DEL. Homocystenaemia and cardiovascular risk: the relevance of folic acid supplementation. Aust J Nutr Diet 1996 ; 53 (Suppl): S11 -S12. 5 Mason JB. Folate and colonic carcinogenesis–searching for a mechanistic understanding. J Nutr Biochem 1994 ; 5 : 170 -175. Crossref Search ADS 6 Fenech M. Folate and cancer initiation. Will folate fortification help to prevent genetic events that could initiate cancer?. Aust J Nutr Diet 1996 ; 53 (Suppl): S13 -S17. 7 Gunter EW, Bowman BA, Caudill SP, Twite DB, Adams MJ, Sampson EJ. Results of an international round robin for serum and whole-blood folate. Clin Chem 1996 ; 42 : 1689 -1694. Crossref Search ADS PubMed 8 Dawson DW, Fish DI, Frew IDO, Roome T, Tilston I. Laboratory diagnosis of megaloblastic anaemia: current methods assessed by external quality assurance trials. J Clin Pathol 1987 ; 40 : 393 -397. Crossref Search ADS PubMed 9 Van den Berg H, Finglas PM, Bates CJ. FLAIR intercomparisons on serum and red cell folate. Int J Vitam Nutr Res 1994 ; 64 : 288 -293. PubMed 10 McGown EL, Lewis CM, Dong MH, Sauberlich HE. Results with commercial radioassay kits compared with microbiological assay of folate in serum and whole-blood. Clin Chem 1978 ; 24 : 2186 -2191. Crossref Search ADS PubMed 11 Gilois CR, Dunbar DR. Measurement of low serum and red cell folate levels: a comparison of analytical methods. Med Lab Sci 1987 ; 44 : 33 -40. PubMed 12 Phillips DR, Wright AJA. Studies on the response of Lactobacillus casei to different folate monoglutamates. Br J Nutr 1982 ; 47 : 183 -189. Crossref Search ADS PubMed 13 Wright AJA, Phillips DR. The threshold growth response of Lactobacillus casei to 5-methyltetrahydrofolic acid: implications for folate assays. Br J Nutr 1985 ; 53 : 569 -573. Crossref Search ADS PubMed 14 Brown RD, Jun R, Hughes W, Watman R, Arnold B, Kronenberg H. Red cell folate assays: some answers to current problems with radioassay variability. Pathology 1990 ; 22 : 82 -87. PubMed 15 Van den Berg H, Floor B. Accuracy of the serum folate radioassay. Colloq INSERM 1990 ; 197 : 441 -447. 16 Levine S. Analytical inaccuracy for folic acid with a popular commercial vitamin B12/folate kit. Clin Chem 1993 ; 39 : 2209 -2210. Crossref Search ADS PubMed 17 Jandle JH. Physiology of red cells. Blood: textbook of haematology 1987 : 49 -109 Little, Brown and Company Boston. . 18 Gregory JF. Chemical and nutritional aspects of folate research: analytical procedures, methods of folate synthesis, stability and bioavailability of dietary folates. In: Kinsella J, ed. Advances in food and nutrition research. London: Academic Press, 1989;33:1–101.. 19 Toennies G, Frank HG, Gallant DL. On the folic acid activity of human blood. J Biol Chem 1953 ; 200 : 23 -30. PubMed 20 Toennies G, Usdin E, Phillips PM. Precursors of the folic acid-active factors of blood. J Biol Chem 1956 ; 221 : 855 -863. PubMed 21 Harper TA. A modified “aseptic addition” assay procedure for the measurement of serum “folic-acid” activity. Nature 1965 ; 207 : 947 -949. Crossref Search ADS PubMed 22 Wright AJA, Romera JM, Finglas PM, Southon S. Erythrocyte folate analysis: effect of pH of ascorbic acid diluent, point of deconjugation, and length and temperature of haemolysate storage [Abstract]. Proc Nutr Soc 1997 ; 56 : 145A . 23 Lakshmaiah N, Ramasatri BV. Folic acid conjugase from plasma. I. Partial purification and properties. Int J Vitam Nutr Res 1975 ; 45 : 183 -193. PubMed 24 Benesch R, Waxman S, Benesch R, Baugh C. The binding of folyl polyglutamates by hemoglobin. Biochem Biophys Res Commun 1982 ; 106 : 1359 -1363. Crossref Search ADS PubMed 25 Benesch RE, Benesch R, Kwong S, Baugh CM. A pteroylpolyglutamate binds to tetramers in deoxyhemoglobin but to dimers in oxyhemoglobin. Proc Natl Acad Sci USA 1983 ; 80 : 6202 -6205. Crossref Search ADS PubMed 26 Benesch RE, Kwong S, Benesch R, Baugh CM. The binding of folyl- and antifolylpolyglutamates to hemoglobin. J Biol Chem 1985 ; 260 : 14653 -14658. PubMed 27 Arnone A, Rogers PH, Benesch RE, Benesch R, Kwong S. The interaction of folylpolyglutamates with deoxyhemoglobin: identification of the binding site. J Biol Chem 1986 ; 261 : 5853 -5857. PubMed 28 Lehninger AL. Globular proteins: the structure and function of hemoglobin. Anderson S Fox J eds. Principles of biochemistry 1982 : 169 -206 Worth Publishers, Inc. New York. . 29 Stryker L. Oxygen-transporting proteins: myoglobin and hemoglobin. Biochemistry 3rd ed. 1988 : 143 -176 W. H. Freeman and Company New York. . 30 Nelson NL, Klausner JS, Branda RF. Oxygenation alters red cell folate levels. Br J Haematol 1983 ; 55 : 235 -242. Crossref Search ADS PubMed 31 Perutz MF, Muirhead H, Cox JM, Goaman LCG. Three-dimensional fourier synthesis of horse oxyhaemoglobin at 2.8 angstrom resolution: the atomic model. Nature 1968 ; 219 : 131 -139. Crossref Search ADS PubMed 32 Benesch R, Benesch RE. Intracellular organic phosphates as regulators of oxygen release by haemoglobin. Nature 1969 ; 221 : 618 -622. Crossref Search ADS PubMed 33 Fermi G, Perutz MF. In: Phillips D, Richards F, eds. Atlas of molecular structures in biology. 2. Haemoglobin and myoglobin. Oxford: Clarendon Press, 1981.. 34 Paoli M, Liddington R, Tame J, Wilkinson A, Dodson G. Crystal structure of T state haemoglobin with oxygen bound at all four haems. J Mol Biol 1996 ; 256 : 775 -792. Crossref Search ADS PubMed 35 Bettati S, Mozzarelli A. T state hemoglobin binds oxygen noncooperatively with allosteric effects of protons, inositol hexaphosphate, and chloride. J Biol Chem 1997 ; 272 : 32050 -32055. Crossref Search ADS PubMed 36 Perutz MF, Shih DT-b, Williamson D. The chloride effect in human haemoglobin: a new kind of allosteric mechanism. J Mol Biol 1994 ; 239 : 555 -560. Crossref Search ADS PubMed 37 Lalezari I, Rahbar S, Lalezari P, Fermi G, Perutz MF. LR16, a compound with potent effects on the oxygen affinity of hemoglobin, on blood cholesterol, and on low density lipoprotein. Proc Natl Acad Sci U S A 1988 ; 85 : 6117 -6121. Crossref Search ADS PubMed 38 Lalezari I, Lalezari P, Poyart C, Marden M, Kister J, Bohn B, Fermi G, Perutz MF. New effectors of human hemoglobin: structure and function. Biochemistry 1990 ; 29 : 1515 -1523. Crossref Search ADS PubMed © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Gross,, Ulrich;Sassa,, Shigeru;Jacob,, Karl;Deybach,, Jean-Charles;Nordmann,, Yves;Frank,, Margareta;Doss, Manfred, O

doi: 10.1093/clinchem/44.9.1892pmid: N/A

Abstract 5-Aminolevulinic acid dehydratase (ALAD) activity in two patients with compound heterozygous 5-aminolevulinic acid dehydratase deficiency porphyria was studied over the last 20 years. The patients’ enzyme activity was <10% from 1977 to 1997. An acute crisis in each patient was successfully treated by infusion of glucose and heme arginate. After this therapy both urinary 5-aminolevulinic acid (ALA) and total porphyrins were diminished to 65% in patient B. In patient H, ALA was decreased to 80%, and total porphyrins were reduced to 15% after treatment with heme arginate and glucose. The patients remained free of symptoms after this therapy. Family studies of patient B showed cross-reactive immunological material (CRIM), in which the maternal mutation is CRIM(+), whereas the paternal mutation is CRIM(−). Incubation of erythrocyte lysates with ALA decreased porphyrin formation, whereas incubation with porphobilinogen produced porphyrin concentrations within reference values in both patients, confirming that ALAD activity is rate-limiting in these cells. ALAD, 5-aminolevulinic acid dehydratase, ALA, 5-aminolevulinic acid, PBG, porphobilinogen, ADP, 5-aminolevulinic acid dehydratase deficiency porphyria, PBGD, porphobilinogen deaminase, CRIM, cross-reactive immunological material. 5-Aminolevulinic acid dehydratase (ALAD)5 is the second enzyme in the heme biosynthetic pathway, which is cytosolic and nonlimiting in heme synthesis in healthy cells. The enzyme catalyzes the condensation of two molecules of 5-aminolevulinic acid (ALA) to form one molecule of the monopyrrole porphobilinogen (PBG). Activity of this enzyme is markedly inhibited by environmental toxins such as lead (1) or markedly decreased in an inherited enzyme deficiency. ALAD activity is present in great excess in the healthy liver, and a partial enzyme deficiency such as in heterozygous ALAD deficiency is not accompanied by any clinical consequence (2) . This study deals with the two currently surviving patients (patients H and B) with ALAD deficiency porphyria (ADP). ADP was first reported in these two young men, not related to each other, who have an intermittent severe acute hepatic porphyria syndrome and ALAD activity ∼1% of controls (3) . Since 15 years of age, both patients suffered from repeated abdominal-neurologic crises with cardiovascular symptoms, persistent paresis, and transient respiratory paralysis. Despite marked ALAD deficiency, there was no anemia in either patient. Family studies demonstrated that ALAD deficiency was inherited as an autosomal recessive trait (4) . Cloning and expression of the defective genes of one patient demonstrated that the patient was heteroallelic for ALAD deficiency with two separate point mutations, one producing an inactive enzyme and the other producing an unstable enzyme (5) . Another point mutation has been detected in the second patient. Because he has a wild-type residue at this site in the other allele, this patient is also compound heterozygous for ALAD deficiency, although the second mutation has not yet been defined (6) . Because there have been few studies that followed enzymatic and biochemical courses of acute hepatic porphyrias, we thought it important to report complete immunological and enzymatic data, as well as biochemical changes over a period of 20 years. Materials and Methods Determination of ALAD activity was performed with 100 μL of packed erythrocytes in 80 mmol/L sodium phosphate buffer, pH 6.4, containing 8 mmol/L ALA according to the European standard method. PBG formed in the assay was determined spectrophotometrically with Ehrlich’s reagent (7) . ALA and PBG were determined spectrophotometrically after isolation by ion-exchange chromatography. Porphyrins were analyzed spectrophotometrically as methyl esters after separation by high-performance thin-layer chromatography (8) . Total porphyrins were calculated from the sum of individual porphyrins. Urinary coproporphyrin isomers I-IV were quantitated by isocratic ion-pair HPLC (9) . Zinc protoporphyrin and protoporphyrin in erythrocytes were analyzed using reversed-phase ion-pair HPLC (10) , with simultaneous fluorometric detection of both substances. Porphobilinogen deaminase (PBGD) and uroporphyrinogen decarboxylase activities in erythrocytes were determined as described previously (11) . Uroporphyrinogen-III synthase activity was measured using erythrocyte lysates by a coupled enzymatic assay (12) . Coproporphyrinogen oxidase, protoporphyrinogen oxidase, and ferrochelatase activities were assayed using lymphocytes, according to the methods reported previously (13)(14)(15) . Incubation of erythrocyte lysates with exogenous ALA and PBG was carried out in the dark at 37 °C for 2 h, as described previously (16) . ALAD and PBGD activities in lymphocytes were determined, and lymphocytes of patient B and his family members were isolated and transformed by infection with Epstein-Barr virus according to Sassa et al. (17) . ALAD concentrations in erythrocytes were determined by rocket immunoelectrophoresis with an antibody against purified human ALAD. ALAD from erythrocytes from subjects of family B was partially purified using anion-exchange chromatography. Three microliters of enzyme fractions was applied to each well. A calibration curve of homogeneously purified ALAD from nondiseased erythrocytes ranged from 10 to 50 mg/L (17) . Total soluble protein was measured according to the method of Bradford (18) . These investigations were in accordance with the current revision of the Helsinki Declaration of 1975. Results clinical and laboratory findings ALAD activity and urinary excretion of heme precursors in both patients with ADP were determined over a period of 20 years. The erythrocyte ALAD activity of both patients was <10% of that in controls [reference range, 283 ± 41 nkat/L, (x ± SD, n = 50), CV = 5.2%] during a period of 20 years. In patient B, erythrocyte PBGD, uroporphyrinogen-III synthase, and uroporphyrinogen decarboxylase as well as lymphocyte coproporphyrinogen oxidase, protoporphyrinogen oxidase, and ferrochelatase activities were additionally examined; all were within reference values. Urinary ALA excretion was 44-fold and 50-fold in 1979 and 1983, respectively, in patient H. Urinary ALA excretion was increased 44-fold in 1979 and 80-fold in 1985 in patient B (Fig. 1). Urinary PBG excretion was increased fourfold in 1983 in patient H. It increased fourfold in 1979 and fivefold in 1986 in patient B (Fig. 2). Urinary total porphyrins of both patients (Fig. 3) followed the course of ALA and PBG, with 90% coproporphyrin and 5% pentacarboxyporphyrin. The ratio of the urinary coproporphyrin isomers I:II:III:IV was 3.1% ± 0.5%: 4.2% ± 1.7%: 84.0% ± 1.7%: 8.7% ± 2.5% (x ± SD). Fecal porphyrins of both patients were within reference values (data not shown). Figure 3. Open in new tabDownload slide Urinary total porphyrin excretion of patients B and H. −, patient H; ▪, patient B. Figure 3. Open in new tabDownload slide Urinary total porphyrin excretion of patients B and H. −, patient H; ▪, patient B. Figure 2. Open in new tabDownload slide Urinary PBG excretion of patients B and H. −, patient H; ▪, patient B. Figure 2. Open in new tabDownload slide Urinary PBG excretion of patients B and H. −, patient H; ▪, patient B. Figure 1. Open in new tabDownload slide Urinary ALA excretion of patients B and H. −, patient H; ▪, patient B. Figure 1. Open in new tabDownload slide Urinary ALA excretion of patients B and H. −, patient H; ▪, patient B. In 1983 patient H suffered from an acute porphyric crisis that was associated with an excessive intake of alcohol (300 g in one day). Urinary excretion of ALA, PBG, and total porphyrins were 2.6 mmol, and 29 and 10.7 μmol per day, respectively. After intensive treatment with glucose, heme arginate, diet, and physiotherapy, the clinical status of both patients improved, despite persistently high levels of ALA (between 0.45 and 2.1 mmol), PBG (between 4 and 17 μmol), and total porphyrin (between 2 and 9 μmol) excretion. In 1994, ALA excretion of ∼1.6 mmol was found; however, it was not associated with an acute attack. In the first half of 1997, patient H was in good health. treatment of acute crisis In the summer of 1997, patient H worked strenuously for ∼9 h without eating. The next morning, he suffered from abdominal colic, weakness in the legs and arms, paresthesia and partial paresis, hypertension, and tachycardia, and he lost appetite. His urinary ALA was 2.2 mmol/24 h. Heme arginate infusion was initiated immediately, which produced good clinical and biochemical responses, and the symptoms disappeared after 3 days. After he was treated with heme arginate, his coproporphyrin excretion was found increased, probably because of improved metabolism of ALA into porphyrins. His ALA excretion increased after the cessation of heme arginate treatment but remained at lower levels than during the acute period (Fig. 4). Figure 4. Open in new tabDownload slide Time course of ALA and total porphyrins of patient H under therapeutic treatment of heme arginate and glucose. ▪, ALA; −, porphyrin. Figure 4. Open in new tabDownload slide Time course of ALA and total porphyrins of patient H under therapeutic treatment of heme arginate and glucose. ▪, ALA; −, porphyrin. The response to heme arginate in patient B is shown in Fig. 5 . This patient was more resistant to heme arginate treatment than the first patient. He had no response to the first heme arginate treatment. Probably his condition was compromised because of vomiting and refusal of food at that time. A much better response, however, was achieved (days 15–20) when heme arginate treatment was combined with glucose infusion. In the spring of 1997, he suffered again from a severe acute abdominal neurologic manifestation and responded well to heme arginate. Figure 5. Open in new tabDownload slide Time course of ALA and total porphyrins of patient B under therapeutic treatment of heme arginate and glucose. ▪, ALA; −, porphyrin. Figure 5. Open in new tabDownload slide Time course of ALA and total porphyrins of patient B under therapeutic treatment of heme arginate and glucose. ▪, ALA; −, porphyrin. porphyrins in erythrocytes In the erythrocytes of patients H and B, protoporphyrin was increased 12- and 3-fold, respectively, compared with healthy controls. Zinc protoporphyrin was enhanced 15- and 9-fold in patients H and B, respectively, compared with healthy controls. In the erythrocytes of the parents of patient H, these porphyrins were within reference values (Table 1). Table 1. Protoporphyrin and zinc protoporphyrin in erythrocytes of patient H, his parents, and patient B compared with the reference range (¯x ± SD, n = 20). Subject . Zinc protoporphyrin, nmol/L . Protoporphyrin, nmol/L . Father of patient H 370 70 Mother of patient H 440 50 Patient H 4300 830 Patient B 2400 210 Reference range 280± 110 70± 30 Subject . Zinc protoporphyrin, nmol/L . Protoporphyrin, nmol/L . Father of patient H 370 70 Mother of patient H 440 50 Patient H 4300 830 Patient B 2400 210 Reference range 280± 110 70± 30 Open in new tab Table 1. Protoporphyrin and zinc protoporphyrin in erythrocytes of patient H, his parents, and patient B compared with the reference range (¯x ± SD, n = 20). Subject . Zinc protoporphyrin, nmol/L . Protoporphyrin, nmol/L . Father of patient H 370 70 Mother of patient H 440 50 Patient H 4300 830 Patient B 2400 210 Reference range 280± 110 70± 30 Subject . Zinc protoporphyrin, nmol/L . Protoporphyrin, nmol/L . Father of patient H 370 70 Mother of patient H 440 50 Patient H 4300 830 Patient B 2400 210 Reference range 280± 110 70± 30 Open in new tab in vitro studies Porphyrin formation from ALA and PBG was evaluated in erythrocyte lysates in both patients (Table 2). With 10−5, 10−4, and 10−3 mol/L ALA, the amount of total porphyrins formed by erythrocyte lysates of patients B and H was 69%, 41%, and 19% compared with the amount formed by healthy controls. With 0.5 × 10−3 mol/L PBG, total porphyrins formed were not different from healthy controls in both patients. Table 2. Total porphyrins in nmol · g−1 total soluble protein · h−1 from lysates of erythrocytes from patients B and H after incubation with various concentrations of ALA and PBG. Subject . ALA . . . PBG0.5 × 10−3 mol/L . . 10−5 mol/L . 10−4 mol/L . 10−3 mol/L . . Patient B 3.8 8.6 12.8 68.8 Patient H 3.1 7.4 13.2 62.1 Reference range (n = 4) 5.0 19.6 66.7 69.3 Subject . ALA . . . PBG0.5 × 10−3 mol/L . . 10−5 mol/L . 10−4 mol/L . 10−3 mol/L . . Patient B 3.8 8.6 12.8 68.8 Patient H 3.1 7.4 13.2 62.1 Reference range (n = 4) 5.0 19.6 66.7 69.3 Open in new tab Table 2. Total porphyrins in nmol · g−1 total soluble protein · h−1 from lysates of erythrocytes from patients B and H after incubation with various concentrations of ALA and PBG. Subject . ALA . . . PBG0.5 × 10−3 mol/L . . 10−5 mol/L . 10−4 mol/L . 10−3 mol/L . . Patient B 3.8 8.6 12.8 68.8 Patient H 3.1 7.4 13.2 62.1 Reference range (n = 4) 5.0 19.6 66.7 69.3 Subject . ALA . . . PBG0.5 × 10−3 mol/L . . 10−5 mol/L . 10−4 mol/L . 10−3 mol/L . . Patient B 3.8 8.6 12.8 68.8 Patient H 3.1 7.4 13.2 62.1 Reference range (n = 4) 5.0 19.6 66.7 69.3 Open in new tab In Epstein-Barr virus-transformed lymphoblastoid cells, ALAD and PBGD activities were examined in patient B and in his family members. ALAD activity was 2.5%, 33%, 46%, and 30% from the patient, the mother, the sister, and the brother, respectively (Table 3). PBGD activity was within the reference range for all these subjects (Table 3). Table 3. ALAD and PBGD activities in Epstein-Barr virus-transformed lymphocytes from family B. Subject . ALAD, fkat/g . PBGD, pkat/g . Patient B 2.0 3.7 Mother 26 3.4 Sister 36 3.5 Brother 24 3.3 Reference range (¯x± SD, n = 5) 79± 19 4.4± 1 Subject . ALAD, fkat/g . PBGD, pkat/g . Patient B 2.0 3.7 Mother 26 3.4 Sister 36 3.5 Brother 24 3.3 Reference range (¯x± SD, n = 5) 79± 19 4.4± 1 Open in new tab Table 3. ALAD and PBGD activities in Epstein-Barr virus-transformed lymphocytes from family B. Subject . ALAD, fkat/g . PBGD, pkat/g . Patient B 2.0 3.7 Mother 26 3.4 Sister 36 3.5 Brother 24 3.3 Reference range (¯x± SD, n = 5) 79± 19 4.4± 1 Subject . ALAD, fkat/g . PBGD, pkat/g . Patient B 2.0 3.7 Mother 26 3.4 Sister 36 3.5 Brother 24 3.3 Reference range (¯x± SD, n = 5) 79± 19 4.4± 1 Open in new tab immunologic studies The ALAD concentrations of patient B, his father, and his mother were 27%, 35%, and 77%, respectively, when compared with the concentrations in healthy controls. The erythrocyte concentrations of ALAD were 62% and 45% in his sister and brother compared with reference values, respectively. Erythrocyte ALAD activities were 1%, 43%, and 25% in the patient, his father, and his mother, respectively. The sister and the brother showed 39% and 37% erythrocyte ALAD activity. Thus in this patient, his mother, his sister, and his brother, ALAD activity was lower than its concentration, indicating that the ALAD mutation in these subjects is cross-reactive immunological material-positive [CRIM(+)]. In contrast, the patient’s father had CRIM(−) mutation. Discussion alad activity ADP was characterized by deficiency of erythrocyte ALAD activity in two patients. They are both alive and generally in good health at the age of 40, indicating that markedly diminished ALAD activity is still sufficient for heme synthesis. An acute attack was associated with excessive alcohol ingestion in one of our patients. Alcohol was shown experimentally to increase 5-aminolevulinic acid synthase activity in the liver (19) . regulatory aspects Our results demonstrated that acute crises in both patients can be managed by therapy with heme arginate in combination with glucose. Protoporphyrin and zinc protoporphyrin concentrations in the erythrocytes of these patients were markedly increased. Despite the decreased ALAD activity in these patients, an overproduction of PBG and porphyrins occurs. The excessive urinary coproporphyrin excretion cannot be caused by an enzymatic deficiency, because coproporphyrinogen oxidase activity was within reference values. Oral ALA loading tests in healthy persons show an excretion of ∼3.3 μmol/L urinary total porphyrins with a portion of 62% coproporphyrin during the first 24 h (20) . The production of coproporphyrin from ALA in patients with ADP may be the consequence of an alteration in the regulation of heme biosynthesis, which is mimicked by the dominance of urinary coproporphyrin in healthy persons after loading with ALA. porphyrin formation from ala and pbg Urinary ALA excretion was increased ∼50-fold, whereas urinary PBG excretion was ∼5-fold in both patients. Although this finding is compatible with the relative rate-limiting nature of ALAD deficiency in these patients (21) , it is necessary to show that in fact ALAD in cells functions as a rate-limiting enzyme in porphyrin formation. Our findings in Table 2 demonstrate that porphyrin formation from ALA is diminished, whereas its formation from PBG is within reference values. This clearly indicates that ALAD functions as a rate-limiting enzyme for porphyrin synthesis. This finding is consistent with our finding on deficient ALAD activity, decreased ALAD protein, and the aberrant phenotype of the mutant ALAD expressed by the patient’s cDNA (5)(6) . immunological findings Our results indicate that ADP in patient B is caused by two separate point mutations (5) . These data agree with the observation of Table 3 , where ALAD activity in lymphocytes of the compound heterozygous patient B (5) is <10% and is 32% in his heterozygous mother, sister, and brother, all of whom have the maternal mutation (5) . The latter molecular genetic analysis is also confirmed by the data from erythrocytes, which show that all these persons have a CRIM(+) mutation. The father’s mutation is CRIM(−). The different CRIM types of the patient’s mother and father show that one must deal with two different mutations in the patient and thus with compound heterozygosity. The maternal CRIM(+) and the paternal CRIM(−) mutations produce a CRIM(+) mutation in the compound heterozygous patient. adp patients ADP is an extremely rare disease. Only four patients have been reported thus far, which include the two young men from Germany (3) , one child from Sweden ((22)), and one elderly patient from Belgium (23) . The Swedish child was also a compound heterozygous subject for ALAD deficiency (24) . The Swedish child underwent liver transplantation (25) ; however, he died 2 years and 9 months after the transplantation at the age of 9 years (S. Thunell, personal communication). The Belgian patient also died at the age of 63, 2 years after the onset of the disease. Thus the two subjects studied in this report are the only ones who are alive. We are grateful to A. G. Freesemann for measuring uroporphyrinogen-III synthase activity. We also acknowledge the skillful technical assistance of Martina Wenz, Sabine Preis, and Heidrun Schudarek. The study was supported by the German Research Association (Grant Gr 1363) and US Public Health Service (DK-32890). The investigation of the excretory variables of both patients was supported by the Hans-Fischer-Gesellschaft, Munich. References 1 Conner EA, Fowler BA. Biochemical and immunological properties of hepatic δ-aminolevulinic acid dehydratase in channel catfish (Ictalurus punctatus). Aquat Toxicol 1994 ; 28 : 37 -52. Crossref Search ADS 2 Bird TD, Hammernyik P, Nutter JY, Labbe RF. Inherited deficiency of δ-aminolevulinic acid dehydratase. Am J Hum Genet 1979 ; 31 : 662 -668. PubMed 3 Doss M, v Tiepermann R, Schneider J, Schmid H. New type of hepatic porphyria with porphobilinogen synthase defect and intermittent acute clinical manifestation. Klin Wochenschr 1979 ; 57 : 1123 -1127. Crossref Search ADS PubMed 4 Doss M, Benkmann HG, Goedde HW. δ-aminolevulinic acid dehydrase (porphobilinogen synthase) in two families with inherited enzyme deficiency. Clin Genet 1986 ; 30 : 191 -198. PubMed 5 Ishida N, Fujita H, Fukuda Y, Noguchi T, Doss M, Kappas A, Sassa S. Cloning and expression of the defective genes from a patient with δ-aminolevulinate dehydratase porphyria. J Clin Investig 1992 ; 89 : 1431 -1437. Crossref Search ADS PubMed 6 Akagi R, Meguro K, Doss M, Sassa S. Molecular studies of the gene defect of ALA dehydratase deficiency porphyria: a new point mutation identified in a second German patient. Porphyrins 1992 ; 1 : 267 -272. 7 Geisse S, Brüller HJ, Doss M. Porphobilinogen synthase (δ-aminolevulinic acid dehydratase) activity in human erythrocytes: reactivation by zinc and dithiothreitol depending on influence of storage. Clin Chim Acta 1983 ; 135 : 239 -245. Crossref Search ADS PubMed 8 Doss MO. Porphyrins and porphyrin precursors. In: Curtius HC, Roth M, eds. Clinical biochemistry–principles and methods. New York: W. de Gruyter, 1974:2:1323–71.. 9 Jacob K, Doss MO. Composition of urinary coproporphyrin isomers I-IV in human porphyrias. Eur J Clin Chem Clin Biochem 1993 ; 31 : 617 -624. PubMed 10 Meyer HD, Jacob K, Vogt W. Ion-pair reversed-phase high-performance liquid chromatographic determination of porphyrins from red blood cells. Chromatographia 1982 ; 16 : 190 -191. Crossref Search ADS 11 Doss MO. Dual porphyria in double heterozygotes with porphobilinogen deaminase and uroporphyrinogen decarboxylase deficiencies. Clin Genet 1989 ; 35 : 146 -151. PubMed 12 Freesemann AG, Hofweber K, Doss MO. Coexistence of deficiencies of uroporphyrinogen III synthase and decarboxylase in a patient with congenital erythropoietic porphyria and in his family. Eur J Clin Chem Clin Biochem 1997 ; 35 : 35 -39. PubMed 13 Nordmann Y, Grandchamp B. Hereditary coproporphyria. Demonstration of a genetic defect in coproporphyrinogen metabolism. Doss M eds. Diagnosis and therapy of porphyrias and lead intoxication 1978 : 77 -81 Springer-Verlag Berlin. . 14 Deybach JC, de Verneuil H, Nordmann Y. The inherited enzymatic defect in porphyria variegata. Hum Genet 1981 ; 58 : 425 -428. Crossref Search ADS PubMed 15 Deybach JC, da Silva V, Pasqier Y, Nordmann Y. Ferrochelatase in human erythropoietic protoporphyria: the first case of a homozygous form of the enzyme deficiency. In: Nordmann Y, ed. Porphyrins and porphyrias 1986;134:163–73.. 16 Doss M. Metabolism of δ-aminolaevulinic acid and porphobilinogen in human erythrocytes in acute intermittent porphyria. Enzyme (Basel) 1973 ; 16 : 343 -353. 17 Sassa S, Fujita H, Doss M, Hassoun A, Verstraeten L, Mercelis R, Kappas A. Hereditary hepatic porphyria due to homozygous δ-aminolevulinic acid dehydratase deficiency: studies in lymphocytes and erythrocytes. Eur J Clin Investig 1991 ; 21 : 244 -248. Crossref Search ADS 18 Bradford MM. A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976 ; 72 : 248 -254. Crossref Search ADS PubMed 19 Sieg I, Doss MO, Kandels H, Schneider J. Effect of alcohol on δ-aminolevulinic acid dehydratase and porphyrin metabolism in man. Clin Chim Acta 1991 ; 202 : 211 -218. Crossref Search ADS PubMed 20 Doss MO. Porphyrinurias and occupational disease. Ann N Y Acad Sci 1987 ; 514 : 204 -218. Crossref Search ADS PubMed 21 de Verneuil H, Doss M, Brusco N, Beaumont C, Nordmann Y. Hereditary hepatic porphyria with delta aminolevulinate dehydrase deficiency: immunologic characterization of the non-catalytic enzyme. Hum Genet 1985 ; 69 : 174 -177. Crossref Search ADS PubMed 22 Thunell S, Holmberg L, Lundgren J. Aminolaevulinate dehydratase porphyria in infancy: a clinical and biochemical study. J Clin Chem Clin Biochem 1987 ; 25 : 5 -14. PubMed 23 Hassoun A, Verstraeten L, Mercelis R, Martin JJ. Biochemical diagnosis of an hereditary aminolaevulinate dehydratase deficiency in a 63-year old man. J Clin Chem Clin Biochem 1989 ; 27 : 781 -786. PubMed 24 Plewinska M, Thunell S, Holmberg L, Wetmur JG, Desnick RJ. δ-Aminolevulinate dehydratase deficient porphyria: identification of the molecular lesions in a severely affected homozygote. Am J Hum Genet 1991 ; 49 : 167 -174. PubMed 25 Thunell S, Henrichson A, Floderus Y, Groth CG, Eriksson BG, Barkholt L, et al. Liver transplantation in a boy with acute porphyria due to aminolaevulinate dehydratase deficiency. Eur J Clin Chem Clin Biochem 1992 ; 30 : 599 -606. PubMed © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Bräutigam,, Christa;Wevers, Ron, A;Jansen, Riet J, T;Smeitink, Jan A, M;Andel, Johanneke F, de Rijk-van;Gabreëls, Fons J, M;Hoffmann, Georg, F

doi: 10.1093/clinchem/44.9.1897pmid: N/A

Abstract We report the biochemical hallmarks of tyrosine hydroxylase deficiency with emphasis on reliable diagnostic strategies of four new cases of an inborn error of tyrosine hydroxylase (TH). Three of our patients from different parts of the Netherlands were found homozygous for a mutation in exon 6 (G698A) of the TH gene, and one patient was found compound heterozygous for the same mutation and an additional mutation in exon 3. The first clinical symptoms of hypokinesia, rigidity of arms and legs and axial hypotonia, developed between 3 and 7 months of age. Cerebrospinal fluid investigations revealed a characteristic metabolite constellation in every case: low homovanillic acid (HVA) and 3-methoxy-4-hydroxy-phenylethyleneglycol concentrations in the presence of normal reference range 5-hydroxyindolacetic acid concentrations. Strict adherence to a standardized lumbar puncture protocol and adequate age-related reference values are essential for diagnosis of this “new” treatable neurometabolic disorder. Urinary measurements of HVA, vanillylmandelic acid, and catecholamines can lead to false-negative conclusions. All patients showed a remarkable clinical improvement on a low dose of l-dihydroxyphenylalanine/(S)-2-(3,4-dihydroxybenzyl)-2-hydrazinpropionic acid. During treatment, cerebrospinal fluid HVA, and 3-methoxy-4-hydroxy-phenylethyleneglycol increased substantially. l-dopa, l-dihydroxyphenylalanine, TH, tyrosine hydroxylase, HVA, homovanillic acid, CNS, central nervous system, VMA, vanillylmandelic acid, CSF, cerebrospinal fluid, MHPG, 3-methoxy-4-hydroxyphenylethyleneglycol, 5-HIAA, 5-hydroxyindolacetic acid, DRD, dopa-responsive dystonia, 3-OMD, 3-o-methyldopa, carbidopa, (S)-2-(3,4-dihydroxybenzyl)-2-hydrazinpropionic acid. The conversion of l-tyrosine to l-dihydroxyphenylalanine (l-dopa),4 catalyzed by the enzyme tyrosine hydroxylase (TH; EC 1.14.16.2), is the rate-limiting step in the biosynthesis of the catecholamines dopamine, norepinephrine, and epinephrine. The stereospecific enzyme, an iron-containing mixed function oxidase, requires molecular oxygen and a tetrahydropteridine cofactor and is present in specific brain areas, in all sympathetically innervated tissues, and in the adrenal medulla (1) . The major catabolic product of dopamine is homovanillic acid (HVA; Fig. 1). The major catabolic product in the central nervous system (CNS) from norepinephrine is 3-methoxy-4-hydroxy-phenylethyleneglycol (MHPG) (2) , whereas vanillylmandelic acid (VMA) is the major norepinephrine catabolite outside the CNS. Figure 1. Open in new tabDownload slide Metabolism of serotonin and the catecholamines, showing the role of TH and the metabolites measured. Neo, neopterin; GTP, guanosine triphosphate; NH2TP, dihydroneopterin triphosphate; BH4, tetrahydrobiopterin; qBH2, quinonoid dihydrobiopterin; TR, tryptophan hydroxylase; COMT, catechol-ortho-methyltransferase; AADC, aromatic l-amino acid decarboxylase; MAO, monoamine oxidase; DOPAC, 3,4-dihydrioxyphenylacetic acid; 3MT, 3-methoxytyramine; DβH, dopamine β-hydroxylase; PNM, phenylethanolamine-N-methyltransferase; NM, normetanephrine; M, metanephrine; ALD, intermediate aldehyde (3-methoxy-4-hydroxyphenylhydroxyacetaldehyde); aldd., aldehyde dehydrogenase; alcd., alcohol dehydrogenase; dotted arrow, several steps are involved. Figure 1. Open in new tabDownload slide Metabolism of serotonin and the catecholamines, showing the role of TH and the metabolites measured. Neo, neopterin; GTP, guanosine triphosphate; NH2TP, dihydroneopterin triphosphate; BH4, tetrahydrobiopterin; qBH2, quinonoid dihydrobiopterin; TR, tryptophan hydroxylase; COMT, catechol-ortho-methyltransferase; AADC, aromatic l-amino acid decarboxylase; MAO, monoamine oxidase; DOPAC, 3,4-dihydrioxyphenylacetic acid; 3MT, 3-methoxytyramine; DβH, dopamine β-hydroxylase; PNM, phenylethanolamine-N-methyltransferase; NM, normetanephrine; M, metanephrine; ALD, intermediate aldehyde (3-methoxy-4-hydroxyphenylhydroxyacetaldehyde); aldd., aldehyde dehydrogenase; alcd., alcohol dehydrogenase; dotted arrow, several steps are involved. TH deficiency can cause the autosomal recessive form of dopa responsive dystonia (DRD); in its clinical form, it is also known as Segawa’s disease. The autosomal dominant form of DRD results from a mutation in the GTP I cyclohydrolase gene (3) . Clayton et al. (4) were the first to suggest TH deficiency for the recessive form of DRD. The TH gene on chromosome 11p15.5 was sequenced by Lüdecke et al. (5) . A point mutation (Q381K) in exon 11 (5)(6) and a point mutation (L205P) in exon 5 (7) were found as disease-causing mutations in the two patients with TH deficiency described thus far. In this study, we report the diagnostic methodology and the biochemical hallmarks of TH deficiency in the CNS. Four unrelated families with the characteristic clinical signs and symptoms of recessive Segawa’s disease are described. The patients share a point mutation (R233H) in exon 6 in the TH gene recently identified by van den Heuvel et al. (8) . Three of our patients are homozygous for the R233H mutation, and one is compound heterozygous for this mutation and for another mutation in exon 3 of this gene. Specific biochemical analyses are compatible with a markedly reduced biosynthesis of dopamine in the CNS because of a deficient function of the TH enzyme system and may enable the diagnosis of additional patients worldwide. Materials and Methods materials Standard substances, all of high analytical grade, and Sephadex G-10 were obtained from Sigma Chemical Co.; HPLC-grade methanol was obtained from Fisons; VMA urine standard, Iso-VMA internal standard, Lyphochek (quantitative urine control), and extraction columns for VMA (cat. no. 195-5005) and catecholamines (no. 189-2202) were obtained from Bio-Rad. All other chemicals were of analytical grade or higher. csf and urine By using a standardized protocol, lumbar CSF was collected between 0830 and 1200 and stored at −70 °C until analyzed. At the age of 0–1 year, the first 2 mL were used for HVA, 5-HIAA, and MHPG determination. At the age of 2–12 years, the first 5 mL of CSF were used for routine investigations; subsequently, the 3 mL were used for HVA, 5-HIAA, and MHPG determination. At the age >12 years, the first 8 mL were for routine CSF investigations, and the subsequent 3 mL were used for measurement of neurotransmitter metabolites. By using the same standardized protocol, reference values were established on CSF left over from routine investigations. Retrospectively, these patients were classified as suitable reference subjects because of lack of evidence for endocrine or metabolic abnormalities. Patients with epilepsy or extrapyramidal signs and symptoms were excluded. Appropriate studies ruled out immunological or chronic infectious diseases, deficiencies, and disorders caused by toxic agents. Samples containing red blood cells were eliminated. The study was approved by the ethical committee of the University of Marburg. CSF samples for the investigations of the gradient in the CSF column were obtained with informed consent of the patients involved. Twenty-four-hour urine was collected into 5 mL of 3 mol/L hydrochloric acid and stored at −20 °C. Reference intervals for urine were obtained from children without metabolic disease, neuroblastoma, or phaochromocytoma. procedures (a) Measurement of HVA and 5-HIAA in CSF and urine We adapted the method of Westerink and Mulder (9) for determination of HVA and 5-HIAA in CSF and urine with the following modifications: to 1 mL of CSF samples, 20 μL of 12.6 mol/L formic acid was added, and the pH ± 2.5 was controlled. The urine samples were centrifuged (10 min, 3000 rpm), and to 50 μL of supernatant, 5 mL Millipore water and 20 μL of 12.6 mol/L formic acid were added, and the pH ± 2.5 was controlled. Five hundred microliters of the CSF sample or urine sample were applied to the Sephadex G-10 column. After washing the Sephadex G-10 column successively with 3.5 mL of 25 mmol/L formic acid and 1.5 mL of 0.2 mol/L phosphate solution, the metabolites were eluted with 2 mL of 34 mmol/L ammonia in a tube containing 50 μL of 12.6 mol/L formic acid and 50 μL of 2.3 mmol/L ascorbic acid. HPLC was performed using a mobile phase of 0.2 mol/L phosphate solution and 0.1 mol/L citric acid (30:70, by volume), pH 3.5, and 350 mL of methanol; a Spectra-Physics SP 8800 gradient pump; a Spectra-Physics SP 8880 autosampler and a SP 8760 autosampler cooler (15 °C); and a 15-cm × 4.6-mm (i.d.) Nucleosil 5-μm RP-18 column. The flow rate was 1.0 mL/min. One hundred microliters of the eluate were injected into the system, and detection was by a Spark Holland amperometric detector with the analytical electrode set at +0.64 V, range × 1, offset × 0.1, and 50 nA full scale. Detector output was integrated by using the PC 1000 software system, Ver. 3.01 (Thermo Separations). (b) Measurement of MHPG in CSF We used the same method as described above for HVA and 5-HIAA with the following modifications. After the washing step of the Sephadex G-10 column, MHPG was eluted with 2.0 mL of 25 mmol/L formic acid and 0.5 mL of 0.2 mol/L phosphate solution. Chromatography was performed with a mobile phase of 6.25 mmol/L phosphate buffer, pH 4.0, containing, per liter, 6 mmol of citric acid, 6 mmol of sodium chloride, 7 mmol of sodium perchlorate, 1 mmol of octyl sodium sulfate, 0.2 mmol of sodium EDTA, and 5 mL of methanol; a Spectra-Physics SP8810 isocratic pump; a Rheodyne 7125 injector; and a 25-cm × 4.6-mm (i.d.) Nucleosil 5-μm C18 column. The flow rate was 0.6 mL/min. Fifty microliters of the eluate were injected into the system, and detection was by an electrochemical detector Decade (Antec) with the analytical cell set at +0.70 V, range × 2, and damping 5. Detector output was integrated using the PC 1000 software system, Ver. 3.01 (Thermo Separations). (c) Measurement of L-dopa and 3-OMD in CSF l-dopa and 3-OMD were measured by HPLC on a 25-cm × 4.6-mm (i.d.) Progidy 5-m 5 ODS-2 column in combination with fluorescence detection (excitation, 278 nm; emission, 325 nm) as described by Hyland (10). (d) Measurement of VMA in urine VMA in urine was determined by HPLC on a Bio-Rad RP-ODS 5 column in combination with electrochemical detection after sample preparation using a commercially available kit (Bio-Rad, no. 1955001). (e) Measurement of dopamine, norepinephrine, and epinephrine in urine After extraction of the catecholamines with cation exchange columns (Bio-Rad, no. 189-2202), we included an additional cleaning step on a Sephadex G-10 column. Forty microliters of 86.1 mol/L perchloric acid were added to 2 mL of the boric acid eluate, and after mixing 1 mL of this solution, it was applied to a Sephadex G-10 column. The column was washed with 2.5 mL of 25 mmol/L formic acid, and the catecholamines were eluted with 2.5 mL of 25 mmol/L formic acid. The catecholamine concentrations were determined with HPLC on a 15-cm × 4.6-mm (i.d.) Supelcosil 5-μm C18 column with electrochemical detection. The system was calibrated by injecting 50 μL of eluate after cation exchange and by cleaning on Sephadex G-10 with a solution containing, per liter, 200 nmol of dopamine, norepinephrine, and epinephrine and 200 nmol of dihydroxybenzylamine (internal standard) in 4 mmol/L formic acid. cases Our patients came from four unrelated families from different parts of the Netherlands. The parents were all healthy. All patients presented with signs and symptoms characteristic of the recessive form of DRD. A detailed clinical description will be published elsewhere. In short, after normal pregnancy and delivery, progressive severe motor retardation with predominant extrapyramidal symptoms first became obvious at ages between 3 and 7 months, whereas psychosocial development appeared relatively healthy. The children appeared hypokinetic with mask face, rigidity of arms and legs, and axial hypotonia. No diurnal fluctuation in the symptoms was observed. Routine clinical chemistry, electroencephalogram, and magnetic resonance imaging and computed tomography neuroimaging were within normal reference intervals in all cases. After establishing the diagnosis of TH deficiency and starting therapy with l-dopa together with (S)-2-(3,4-dihydroxybenzyl)-2-hydrazinpropionic acid (carbidopa) there was a clear improvement of symptoms. In all cases, a G698A transition was found by direct sequencing of exon 6 of the TH gene. This transition produces an amino acid change from arginine to histidine (R233H) (8) . Patient I was heterozygous for the G698A mutation and also for a one-base deletion in exon 3 of the TH gene. The other three were homozygous for the G698A mutation. CSF and urine of the children were investigated before any medication was initiated and also during therapy with l-dopa. Results the reference values Table 1 shows our age-related reference values of the CSF metabolites HVA, 5-HIAA, HVA/5-HIAA ratio, and MHPG. All CSF samples included in this reference range study were obtained with the standardized protocol described in Materials and Methods. For the first 2 years of life, there were only relatively few data points available that could be included in the statistical calculation. There was a marked effect of age on CSF metabolite concentrations. The median concentration of free MHPG decreases rapidly in the first year of life. In later life, the reference range for MHPG does not change substantially any more. The metabolites HVA and 5-HIAA decrease rapidly in early neonatal life and decrease steadily until they reach adult concentrations at the age of 15 years. The HVA/5-HIAA ratio increases from birth up to the age of 12 years. Lower values for this ratio are found in adults. Table 1. Age-related reference values for CSF concentrations of HVA, 5-HIAA, HVA/5-HIAA ratio, and MHPG (all concentrations in nmol/L). . P1 . ≤0.3 . 0.3 to ≤0.7 . 0.7 to ≤1 . 1 to ≤2 . 2 to ≤5 . 5 to ≤8 . 8 to ≤12 . 12 to ≤15 . >15 . n (HVA, 5-HIAA) 7 11 10 16 49 25 22 16 81 HVA P 2.5 543 478 488 429 384 346 339 211 87 P 50 780 566 557 605 510 505 449 397 213 P 97.5 1142 895 664 789 769 716 668 540 372 5-HIAA P 2.5 383 231 190 156 110 100 109 95 58 P 50 531 322 229 207 183 166 152 148 105 P 97.5 1028 618 301 275 265 245 214 173 190 ratio HVA/5-HIAA P 2.5 0.8 1.3 2.1 1.6 1.8 2.3 1.9 1.8 1.2 P 50 1.4 1.7 2.5 2.8 2.9 3.1 3.1 2.7 2.0 P 97.5 1.9 3.1 2.9 3.3 4.4 4.0 3.8 4.1 3.1 n (MHPG) 7 11 8 14 42 18 10 7 74 MHPG P 2.5 84 49 40 33 35 37 32 41 29 P 50 116 67 49 54 50 49 54 53 44 P 97.5 277 116 71 71 64 75 68 82 64 . P1 . ≤0.3 . 0.3 to ≤0.7 . 0.7 to ≤1 . 1 to ≤2 . 2 to ≤5 . 5 to ≤8 . 8 to ≤12 . 12 to ≤15 . >15 . n (HVA, 5-HIAA) 7 11 10 16 49 25 22 16 81 HVA P 2.5 543 478 488 429 384 346 339 211 87 P 50 780 566 557 605 510 505 449 397 213 P 97.5 1142 895 664 789 769 716 668 540 372 5-HIAA P 2.5 383 231 190 156 110 100 109 95 58 P 50 531 322 229 207 183 166 152 148 105 P 97.5 1028 618 301 275 265 245 214 173 190 ratio HVA/5-HIAA P 2.5 0.8 1.3 2.1 1.6 1.8 2.3 1.9 1.8 1.2 P 50 1.4 1.7 2.5 2.8 2.9 3.1 3.1 2.7 2.0 P 97.5 1.9 3.1 2.9 3.3 4.4 4.0 3.8 4.1 3.1 n (MHPG) 7 11 8 14 42 18 10 7 74 MHPG P 2.5 84 49 40 33 35 37 32 41 29 P 50 116 67 49 54 50 49 54 53 44 P 97.5 277 116 71 71 64 75 68 82 64 1 P, percentile. Open in new tab Table 1. Age-related reference values for CSF concentrations of HVA, 5-HIAA, HVA/5-HIAA ratio, and MHPG (all concentrations in nmol/L). . P1 . ≤0.3 . 0.3 to ≤0.7 . 0.7 to ≤1 . 1 to ≤2 . 2 to ≤5 . 5 to ≤8 . 8 to ≤12 . 12 to ≤15 . >15 . n (HVA, 5-HIAA) 7 11 10 16 49 25 22 16 81 HVA P 2.5 543 478 488 429 384 346 339 211 87 P 50 780 566 557 605 510 505 449 397 213 P 97.5 1142 895 664 789 769 716 668 540 372 5-HIAA P 2.5 383 231 190 156 110 100 109 95 58 P 50 531 322 229 207 183 166 152 148 105 P 97.5 1028 618 301 275 265 245 214 173 190 ratio HVA/5-HIAA P 2.5 0.8 1.3 2.1 1.6 1.8 2.3 1.9 1.8 1.2 P 50 1.4 1.7 2.5 2.8 2.9 3.1 3.1 2.7 2.0 P 97.5 1.9 3.1 2.9 3.3 4.4 4.0 3.8 4.1 3.1 n (MHPG) 7 11 8 14 42 18 10 7 74 MHPG P 2.5 84 49 40 33 35 37 32 41 29 P 50 116 67 49 54 50 49 54 53 44 P 97.5 277 116 71 71 64 75 68 82 64 . P1 . ≤0.3 . 0.3 to ≤0.7 . 0.7 to ≤1 . 1 to ≤2 . 2 to ≤5 . 5 to ≤8 . 8 to ≤12 . 12 to ≤15 . >15 . n (HVA, 5-HIAA) 7 11 10 16 49 25 22 16 81 HVA P 2.5 543 478 488 429 384 346 339 211 87 P 50 780 566 557 605 510 505 449 397 213 P 97.5 1142 895 664 789 769 716 668 540 372 5-HIAA P 2.5 383 231 190 156 110 100 109 95 58 P 50 531 322 229 207 183 166 152 148 105 P 97.5 1028 618 301 275 265 245 214 173 190 ratio HVA/5-HIAA P 2.5 0.8 1.3 2.1 1.6 1.8 2.3 1.9 1.8 1.2 P 50 1.4 1.7 2.5 2.8 2.9 3.1 3.1 2.7 2.0 P 97.5 1.9 3.1 2.9 3.3 4.4 4.0 3.8 4.1 3.1 n (MHPG) 7 11 8 14 42 18 10 7 74 MHPG P 2.5 84 49 40 33 35 37 32 41 29 P 50 116 67 49 54 50 49 54 53 44 P 97.5 277 116 71 71 64 75 68 82 64 1 P, percentile. Open in new tab the rostro caudal gradient To investigate a concentration gradient in CSF for the metabolites HVA, 5-HIAA, and MHPG in CSF from different levels of the spinocisternal system, consecutive CSF fractions from adults without deficiency in the neurotransmitter metabolism were taken and analyzed. There was a steep gradient for HVA and 5-HIAA in spinal CSF of two adult patients, with an increase of 60% in HVA concentrations and of 95% in 5-HIAA concentrations in the first two CSF fractions. Fig. 2 gives a representative example. From the first fraction to the last CSF fraction (35 mL), there was an increase of 124% for HVA and 173% for 5-HIAA. The HVA/5-HIAA ratio varied between 1.03 and 1.48 in the various fractions. The metabolite MHPG was unaffected by the volume fraction (Fig. 2). Figure 2. Open in new tabDownload slide Rostro caudal gradient in different CSF volume fractions (lumbal to ventricular) of an adult patient. □, HVA concentrations; •, 5-HIAA concentrations; ♦, MHPG concentrations; ×, HVA/5-HIAA ratio × 10. Figure 2. Open in new tabDownload slide Rostro caudal gradient in different CSF volume fractions (lumbal to ventricular) of an adult patient. □, HVA concentrations; •, 5-HIAA concentrations; ♦, MHPG concentrations; ×, HVA/5-HIAA ratio × 10. the stability of the metabolites in CSF To investigate the stability of the monoamine metabolites in CSF, we took fresh CSF samples from two adult patients without neurometabolic disease. One portion of the CSF was immediately stored at −70 °C. Separate volumes of the sample of the first patient were left at room temperature (25 °C) for 1.5, 4, 20, 27, and 51 h and of the second patient for 1.5, 3, 4, 6, 23, and 27 h and subsequently stored at −70 °C. The CSF HVA, 5-HIAA, and MHPG concentrations of the first patient were stable at room temperature (25 °C) for at least 27 h (HVA: mean 303 nmol/L, SD 10; 5-HIAA: mean 129 nmol/L, SD 6; MHPG: mean 47 nmol/L, SD 2; HVA/5-HIAA ratio: mean 2.4, SD 0.1) after CSF collection without any addition of antioxidants or buffers in the analyzed samples. After 51 h, the HVA concentration increased >2 SDs. In the second CSF sample, the metabolites were also found to be stable during 27 h at room temperature. pretreatment metabolite concentrations in body fluids Clinical chemical routine investigations were within normal reference intervals in all patients, as were metabolic investigations including organic acids in urine and amino acids in urine, blood, and CSF. Phenylalanine was within normal reference intervals in all body fluids investigated. Tyrosine as substrate of tyrosine hydroxylase was in the normal reference interval in urine, plasma, and CSF [for CSF, patient I: 14 μmol/L (reference, 6–19 μmol/L); patient II, 13 μmol/L (reference, 6–19 μmol/L); patient III, 10 μmol/L (reference, 5–13 μmol/L); and patient IV, 12 μmol/L (reference, 5–13 μmol/L] per method according to Gerrits et al. (11) . The pterin concentrations in CSF and urine (biopterin, neopterin, and their ratio) were all completely within the normal reference interval. Determinations of the dihydropteridine reductase activity in blood showed results within the normal reference intervals in all four patients (data not given). These findings exclude a defect of tetrahydrobiopterin biosynthesis or recycling. Fig. 3 shows a characteristic chromatogram of the CSF metabolites HVA and 5-HIAA of patient III and of a control subject. In four patients with TH deficiency, analyses of the CSF revealed low concentrations of the dopamine metabolite HVA and the norepinephrine metabolite MHPG, whereas the serotonin metabolite 5-HIAA was always in the normal reference interval (Table 2). The HVA/5-HIAA ratio in CSF was abnormally low in all patients. These results were confirmed in repeat CSF samples from all four patients (data not shown). Table 2. CSF metabolite concentrations in four patients with TH deficiency before treatment (all concentrations in nmol/L). Patient . Age . HVA(P1 2.5:384) . 5-HIAA(P 2.5:110) . HVA/5-HIAA ratio (P 2.5:1.8) . MHPG(P 2.5:35) . I, male 2 y 1 mo 117 151 0.77 13 II, female 2 y 5 mo 111 233 0.48 6 III, male 3 y 3 mo 76 268 0.28 12 IV, female 3 y 1 mo 31 234 0.13 2 Patient . Age . HVA(P1 2.5:384) . 5-HIAA(P 2.5:110) . HVA/5-HIAA ratio (P 2.5:1.8) . MHPG(P 2.5:35) . I, male 2 y 1 mo 117 151 0.77 13 II, female 2 y 5 mo 111 233 0.48 6 III, male 3 y 3 mo 76 268 0.28 12 IV, female 3 y 1 mo 31 234 0.13 2 1 P, percentile; reference ranges of the metabolites are given in the top bar. Open in new tab Table 2. CSF metabolite concentrations in four patients with TH deficiency before treatment (all concentrations in nmol/L). Patient . Age . HVA(P1 2.5:384) . 5-HIAA(P 2.5:110) . HVA/5-HIAA ratio (P 2.5:1.8) . MHPG(P 2.5:35) . I, male 2 y 1 mo 117 151 0.77 13 II, female 2 y 5 mo 111 233 0.48 6 III, male 3 y 3 mo 76 268 0.28 12 IV, female 3 y 1 mo 31 234 0.13 2 Patient . Age . HVA(P1 2.5:384) . 5-HIAA(P 2.5:110) . HVA/5-HIAA ratio (P 2.5:1.8) . MHPG(P 2.5:35) . I, male 2 y 1 mo 117 151 0.77 13 II, female 2 y 5 mo 111 233 0.48 6 III, male 3 y 3 mo 76 268 0.28 12 IV, female 3 y 1 mo 31 234 0.13 2 1 P, percentile; reference ranges of the metabolites are given in the top bar. Open in new tab Figure 3. Open in new tabDownload slide Chromatograms of the biogenic amines in CSF with HPLC in combination with electrochemical detection. (A) CSF of a control person (age, 5 years; 5-HIAA, 231 nmol/L; HVA, 674 nmol/L). (B) CSF of TH-deficient patient IV. Figure 3. Open in new tabDownload slide Chromatograms of the biogenic amines in CSF with HPLC in combination with electrochemical detection. (A) CSF of a control person (age, 5 years; 5-HIAA, 231 nmol/L; HVA, 674 nmol/L). (B) CSF of TH-deficient patient IV. The CSF concentration of HVA ranged from 8% of the age-related lower reference range limit (percentile, 2.5) in patient IV to 30% in patient I. The CSF concentration of MHPG ranged from 6% to 37% of the lower limit of the age-related reference range (percentile, 2.5). Concentrations of CSF l-dopa were not detectable in all patients (reference, 0–10 nmol/L). Similarly, 3-OMD, a metabolite deriving from l-dopa (Fig. 1), was not detectable in all patients (reference, 0–50 nmol/L), and the concentration of vanillactic acid in urine was not increased, thus excluding aromatic l-amino acid decarboxylase deficiency. Pretreatment urinary HVA (Table 3) was found decreased in three patients and within normal reference intervals in patient IV. The VMA concentration was decreased in all four cases. 5-HIAA concentrations in all samples were within the normal reference interval. Remarkable were the values within the normal reference intervals of the catecholamines norepinephrine, dopamine, and epinephrine in most of the urine samples, except the urine of patient III, which had decreased values of norepinephrine and dopamine. In all four cases, the epinephrine/norepinephrine ratio was increased (1.0 to 5.4; reference, <1). Unexpectedly, epinephrine was repeatedly found increased in the urine of patient I (about 255% above the upper limit of the age-related reference range). The concentrations of normetanephrine and metanephrine in two different urine samples of patient I were also measured: the normetanephrine concentrations were within the normal reference intervals, and the metanephrine concentrations were increased (about 38% to 257% above the upper limit of the reference range in the two samples). Table 3. Urine metabolite concentrations in four patients with TH deficiency before treatment: HVA, 5-HIAA, and VMA concentrations in μmol/mmol creatinine and norepinephrine (NE), epinephrine (E), and dopamine (DA) concentrations in nmol/mmol creatinine. Patient . Age . HVA . Reference HVA . 5-HIAA . Reference 5-HIAA . VMA . Reference VMA . NE . Reference NE . E . Reference E . E/NE ratio (reference, <1.0) . DA . Reference DA . I, male 2 y 1 mo 2.5 5–15 4.1 3–12 1.1 2–15 9.5 8–70 50.9 1–20 5.4 70 50–700 II, female 2 y 7 mo 4.1 5–15 7.6 3–12 1.2 2–15 13.5 10–100 21.2 1.5–30 1.6 195 70–975 III, male 3 y 3 mo 1.4 2–10 3.4 1–10 1.0 2–10 3.8 7–85 4.5 1.5–30 1.2 14 70–825 IV, female 3 y 1 mo 5.3 2–10 10.0 1–10 1.2 2–10 10.6 7–85 10.6 1.5–30 1.0 293 70–825 Patient . Age . HVA . Reference HVA . 5-HIAA . Reference 5-HIAA . VMA . Reference VMA . NE . Reference NE . E . Reference E . E/NE ratio (reference, <1.0) . DA . Reference DA . I, male 2 y 1 mo 2.5 5–15 4.1 3–12 1.1 2–15 9.5 8–70 50.9 1–20 5.4 70 50–700 II, female 2 y 7 mo 4.1 5–15 7.6 3–12 1.2 2–15 13.5 10–100 21.2 1.5–30 1.6 195 70–975 III, male 3 y 3 mo 1.4 2–10 3.4 1–10 1.0 2–10 3.8 7–85 4.5 1.5–30 1.2 14 70–825 IV, female 3 y 1 mo 5.3 2–10 10.0 1–10 1.2 2–10 10.6 7–85 10.6 1.5–30 1.0 293 70–825 Open in new tab Table 3. Urine metabolite concentrations in four patients with TH deficiency before treatment: HVA, 5-HIAA, and VMA concentrations in μmol/mmol creatinine and norepinephrine (NE), epinephrine (E), and dopamine (DA) concentrations in nmol/mmol creatinine. Patient . Age . HVA . Reference HVA . 5-HIAA . Reference 5-HIAA . VMA . Reference VMA . NE . Reference NE . E . Reference E . E/NE ratio (reference, <1.0) . DA . Reference DA . I, male 2 y 1 mo 2.5 5–15 4.1 3–12 1.1 2–15 9.5 8–70 50.9 1–20 5.4 70 50–700 II, female 2 y 7 mo 4.1 5–15 7.6 3–12 1.2 2–15 13.5 10–100 21.2 1.5–30 1.6 195 70–975 III, male 3 y 3 mo 1.4 2–10 3.4 1–10 1.0 2–10 3.8 7–85 4.5 1.5–30 1.2 14 70–825 IV, female 3 y 1 mo 5.3 2–10 10.0 1–10 1.2 2–10 10.6 7–85 10.6 1.5–30 1.0 293 70–825 Patient . Age . HVA . Reference HVA . 5-HIAA . Reference 5-HIAA . VMA . Reference VMA . NE . Reference NE . E . Reference E . E/NE ratio (reference, <1.0) . DA . Reference DA . I, male 2 y 1 mo 2.5 5–15 4.1 3–12 1.1 2–15 9.5 8–70 50.9 1–20 5.4 70 50–700 II, female 2 y 7 mo 4.1 5–15 7.6 3–12 1.2 2–15 13.5 10–100 21.2 1.5–30 1.6 195 70–975 III, male 3 y 3 mo 1.4 2–10 3.4 1–10 1.0 2–10 3.8 7–85 4.5 1.5–30 1.2 14 70–825 IV, female 3 y 1 mo 5.3 2–10 10.0 1–10 1.2 2–10 10.6 7–85 10.6 1.5–30 1.0 293 70–825 Open in new tab metabolite concentrations during treatment in body fluids In our patients, we initially used low dose l-dopa of about 3 mg/kg/day, together with the decarboxylase inhibitor carbidopa, of 0.75 mg/kg/day to block peripheral conversion of l-dopa into dopamine. The medication was given in three divided doses per day. This resulted in an immediate clinical improvement. In some patients, additional clinical improvements could be achieved by further increasing the daily l-dopa dose after some time. Table 4 gives the neurotransmitter metabolites in CSF under treatment. Both HVA, MHPG, and the HVA/5HIAA ratio in CSF increased on treatment with 2.9–4.4 mg/kg/day of l-dopa but remained below the age-related reference range (compared with the lower reference range limit of percentile 2.5: HVA, 35–51%; MHPG, 42–51%; and HVA/5-HIAA, 28–61%; Table 4). When higher doses of l-dopa (5–6.8 mg/kg/day) together with carbidopa were given, a further increase of CSF neurotransmitter metabolites could be observed (Table 4). However, their concentrations did not normalize (compared with the lower reference range limit of percentile 2.5: HVA, 48–58%; MHPG, 51–94%; and HVA/5-HIAA, 44–84%). The CSF concentrations of l-dopa and 3-OMD were 1 to 55 times higher than the upper limit of the reference range in different samples (Table 4). Table 4. Effect of l-dopa therapy of the TH-deficient patients on metabolite concentrations in CSF (all concentrations in nmol/L). Patient . Age . l-dopa1 doses,mg/kg/day . HVA(P2 2.5:384) . 5-HIAA(P 2.5:110) . HVA/5-HIAA ratio (P 2.5:1.8) . MHPG(P 2.5:35) . l-dopa(reference, 0–10) . 3-OMD(reference, 0–50) . I, male 3 y 7 mo 2.9 195 178 1.09 16 10 170 4 y 7 mo 5.0 223 148 1.51 33 22 741 II, female 3 y 2 mo 3.0 187 263 0.70 16 25 407 4 y 1 mo 4.4 193 206 0.94 17 18 506 III, male 3 y 11 mo 5.7 220 263 0.80 21 41 365 4 y 10 mo 6.8 185 199 0.93 18 55 577 IV, female 3 y 7 mo 3.1 136 268 0.50 15 13 179 4 y 6 mo 4.2 147 191 0.77 18 15 436 Patient . Age . l-dopa1 doses,mg/kg/day . HVA(P2 2.5:384) . 5-HIAA(P 2.5:110) . HVA/5-HIAA ratio (P 2.5:1.8) . MHPG(P 2.5:35) . l-dopa(reference, 0–10) . 3-OMD(reference, 0–50) . I, male 3 y 7 mo 2.9 195 178 1.09 16 10 170 4 y 7 mo 5.0 223 148 1.51 33 22 741 II, female 3 y 2 mo 3.0 187 263 0.70 16 25 407 4 y 1 mo 4.4 193 206 0.94 17 18 506 III, male 3 y 11 mo 5.7 220 263 0.80 21 41 365 4 y 10 mo 6.8 185 199 0.93 18 55 577 IV, female 3 y 7 mo 3.1 136 268 0.50 15 13 179 4 y 6 mo 4.2 147 191 0.77 18 15 436 1 Carbidopa dose is 25% of l-dopa dose. 2 P, percentile; reference ranges of the metabolites are given in the top bar. Open in new tab Table 4. Effect of l-dopa therapy of the TH-deficient patients on metabolite concentrations in CSF (all concentrations in nmol/L). Patient . Age . l-dopa1 doses,mg/kg/day . HVA(P2 2.5:384) . 5-HIAA(P 2.5:110) . HVA/5-HIAA ratio (P 2.5:1.8) . MHPG(P 2.5:35) . l-dopa(reference, 0–10) . 3-OMD(reference, 0–50) . I, male 3 y 7 mo 2.9 195 178 1.09 16 10 170 4 y 7 mo 5.0 223 148 1.51 33 22 741 II, female 3 y 2 mo 3.0 187 263 0.70 16 25 407 4 y 1 mo 4.4 193 206 0.94 17 18 506 III, male 3 y 11 mo 5.7 220 263 0.80 21 41 365 4 y 10 mo 6.8 185 199 0.93 18 55 577 IV, female 3 y 7 mo 3.1 136 268 0.50 15 13 179 4 y 6 mo 4.2 147 191 0.77 18 15 436 Patient . Age . l-dopa1 doses,mg/kg/day . HVA(P2 2.5:384) . 5-HIAA(P 2.5:110) . HVA/5-HIAA ratio (P 2.5:1.8) . MHPG(P 2.5:35) . l-dopa(reference, 0–10) . 3-OMD(reference, 0–50) . I, male 3 y 7 mo 2.9 195 178 1.09 16 10 170 4 y 7 mo 5.0 223 148 1.51 33 22 741 II, female 3 y 2 mo 3.0 187 263 0.70 16 25 407 4 y 1 mo 4.4 193 206 0.94 17 18 506 III, male 3 y 11 mo 5.7 220 263 0.80 21 41 365 4 y 10 mo 6.8 185 199 0.93 18 55 577 IV, female 3 y 7 mo 3.1 136 268 0.50 15 13 179 4 y 6 mo 4.2 147 191 0.77 18 15 436 1 Carbidopa dose is 25% of l-dopa dose. 2 P, percentile; reference ranges of the metabolites are given in the top bar. Open in new tab Using the treatment with the low dose of l-dopa, we also analyzed the catecholamines and metabolites in the urine of the patients and found a similar picture in all samples. As expected, we found high concentrations of dopamine (6 to 25 times higher than the upper limit) and HVA (within normal reference intervals to 4 times higher than the upper limit) and values within normal reference intervals for 5-HIAA, VMA, norepinephrine, and epinephrine. Without exception, the ratio of epinephrine/norepinephrine reached values within the normal reference intervals. In patient I, the increased concentration of epinephrine that was observed in two samples before treatment also fully reached values within normal reference intervals. Discussion analytical aspects with diagnostic relevance There are many pitfalls in accurate measurements and interpretation of neurotransmitter metabolites in CSF from sampling to analytical procedures. In agreement with previous reports (12)(13)(14)(15) , HVA and 5-HIAA concentrations were significantly correlated (r = −0.74 for HVA and −0.45 for 5-HIAA) with age. The metabolites HVA and 5-HIAA decrease with age over the first years of life and for MHPG over the first months of life. Therefore, adequate age-related reference values are of utmost importance. In accordance with previous studies (16)(17) , a steep gradient from lumbar to ventricular CSF was observed for the CSF metabolites HVA and 5-HIAA. No gradient was found for the metabolite MHPG and for the HVA/5-HIAA ratio. For correct interpretation of neurotransmitter metabolite analysis, a standardized protocol including standardization of the time of lumbar puncture and of the CSF volume fraction used for analyses of neurotransmitter metabolite is essential. We have presented reference values for a specific volume fraction (see Materials and Methods). To our knowledge, this is the first study on neurotransmitter metabolites in CSF that uses defined CSF volume fractions and gives appropriate age-related reference values based on a standardized protocol for the lumbar puncture. Because of the interplay between neurotransmitters and their receptors, small deviations of metabolite concentrations or of ratios of metabolites can be of diagnostic importance. Over the last years, we were involved in the workup of several patients in whom a diagnosis could not be established initially because of sampling errors or interpretation difficulties because of inadequate reference ranges. In CSF samples of two adult patients, we found the neurotransmitter metabolites HVA, 5-HIAA, and MHPG stable for at least 27 h at room temperature, in agreement with an earlier publication (18) . In contrast, pterin concentrations in CSF are unstable at room temperature and are also light sensitive (13) . Because pterin analysis is important in the differential diagnosis of genetic defects of biogenic monoamine metabolism, CSF samples for all of these investigations have to be kept in the dark and stored immediately at −70 °C. When interpreting concentrations of conjugated catecholamines in urine, it is important to consider dietary influences. Particular caution is necessary with food known to contain biogenic amines, i.e., bananas (19)(20) . An influence of diet on CSF catecholamine metabolites has not yet been fully investigated. TH is mainly expressed in brain and adrenal medulla. Therefore, direct measurement of TH enzymatic activity in tissue samples is not a diagnostic option, and final proof can only be obtained by molecular genetic analysis. Until now, only two patients have been described in the literature with DNA-confirmed defects in TH (6)(7) . Our experience suggests that TH deficiency is a rare but widely underdiagnosed inborn error of metabolism. diagnostic hallmarks of th deficiency TH deficiency leads to depressed concentrations of l-dopa and consecutively to low concentrations of the catecholamines dopamine, norepinephrine, and epinephrine. The measurement of phenylalanine and tyrosine in body fluids does not provide any clues to the diagnosis of TH deficiency. Furthermore, urinary measurements of VMA, HVA, and catecholamines can give values within normal reference intervals and lead to erroneous interpretations. The ratio of urinary epinephrine/norepinephrine may prove to be a useful indicator. As yet we conclude that the biochemical diagnosis of TH deficiency can only be made reliably by measuring neurotransmitter metabolites in CSF. Reduced dopamine synthesis leads to decreased CSF concentrations of HVA and MHPG. Together with unaffected pterin and CSF tyrosine and 5-HIAA concentrations, these findings are the diagnostic hallmarks of isolated TH deficiency. relation between mutation and csf neurotransmitter metabolites For all four patients, the CSF HVA concentration ranged between 8% and 30% of the lower reference range limit, whereas MHPG ranged from 6% to 37%. The patient (I) with the highest CSF HVA concentration is heterozygous for the G698A mutation in exon 6 of the TH gene and for another mutation in exon 3 of the TH gene. The other three patients are homozygous for the same mutation. Different mutations in the gene may produce a variable residual enzyme activity and therefore various CSF HVA and MHPG concentrations between patients. The different mutations may give rise to a wider clinical and biochemical spectrum. therapy There was a marked improvement of all clinical signs and symptoms under treatment with low doses of l-dopa/carbidopa. HVA concentrations in CSF increased substantially but still were below the lower reference range limit. After increasing the dose (Table 4), an additional clinical improvement could be achieved corresponding to an additional increase in HVA concentrations in CSF but still below the lower limit of the reference range. In earlier publications (4)(7) , higher doses of l-dopa (up to 10 mg l-dopa/kg/day) were given, and a healthy HVA concentration in CSF could be achieved. However, some of our patients responded with hyperkinesia to increasing doses. Also, it should be considered that adverse effects may occur in the long run because the l-dopa therapy has to be continued life-long. Therefore, the therapy strategy in each patient has to find a balance between the short-term beneficial aspects and the longer term side effects of the therapy. urine analysis The biochemical picture in urine warrants additional comments. Low to healthy HVA and VMA concentrations, together with healthy 5-HIAA concentrations, occur in all samples. The catecholamines are in the normal reference interval in many urine samples of our patients. Because TH in the adrenals and in the CNS derives from the same gene, TH in the adrenals is expected to be equally deficient in the patients. Therefore, the often healthy urinary catecholamine concentrations and the lack of clinical evidence for adrenal malfunction are ill-understood. Dietary influences may play a role in the healthy urinary concentrations of the catecholamines (19)(20) . We have no explanation for the increased concentration of epinephrine in patient I in different urine samples and the disturbed epinephrine/norepinephrine ratio in all four cases. The high epinephrine concentration in the urine of patient I was also found in another laboratory using an independent technique. Because we also found high concentrations of metanephrine in the urine of the patient, these findings are unlikely to be artifacts. There has been speculation in the literature about alternative pathways in catecholamine metabolism (1)(21) . Attempts have been made to find alternative pathways in vivo and to determine whether they are functionally important. Biosynthetic pathways from l-tyrosine to norepinephrine and epinephrine are possible via p-tyramine, octopamine, and synephrine (1)(21) . Such alternative pathways of epinephrine and norepinephrine synthesis may become important in TH-deficient patients. We thank N. Blau for measurements of pterins as well as of the activity of dihydropteridine reductase, N. Abeling for measurement of metanephrines, and the statistical department of the University Nijmegen for deliberation. References 1 Cooper JR, Bloom FE, Roth RH. The biochemical basis of neuropharmacology. New York: Oxford University Press, 1996:230pp.. 2 Kopin IJ, Gordon EK, Jimerson DC, Polinsky RJ. Relation between plasma and cerebrospinal fluid levels of 3-methoxy-4-hydroxyphenylglycol. Science 1983 ; 219 : 73 -75. Crossref Search ADS PubMed 3 Ichinose H, Ohye T, Takahshi E, Seki N, Hori T, Segawa M, et al. Hereditary progressive dystonia with marked diurnal fluctuation caused by mutations in the GTP cyclohydrolase I gene. Nat Genet 1994 ; 8 : 236 -242. Crossref Search ADS PubMed 4 Clayton PT, Heales SJR, Brand M, Clelland J, Surtees RJ. An infant with clinical features, a metabolic profile to treatment suggestive of tyrosine hydroxylase deficiency. Poster presentation 49, 32nd Annu. Symp. Society for the Study of Inborn Errors of Metablism, Edinburgh, 1994.. 5 Lüdecke B, Dworniczak B, Bartholomé K. A point mutation in the tyrosine hydroxylase gene associated with Segawa’s syndrome. Hum Genet 1995 ; 95 : 123 -125. PubMed 6 Knappskog PM, Flatmark T, Mallet J, Lüdecke B, Bartholomé K. Recessively inherited l-dopa-responsive dystonia caused by a point mutation (Q381K) in the tyrosine hydroxylase gene. Hum Mol Genet 1995 ; 4 : 1209 -1212. Crossref Search ADS PubMed 7 Lüdecke B, Knappskog PM, Clayton PT, Surtees RAH, Clelland JD, Heales SJR, et al. Recessively inherited l-dopa-responsive parkinsonism in infancy caused by a point mutation (L205P) in the tyrosine hydroxylase gene. Hum Mol Genet 1996 ; 5 : 1023 -1028. Crossref Search ADS PubMed 8 van den Heuvel LPWJ, Luiten B, Smeitink JAM, de Rijk-van Andel JF, Steenbergen-Spanjers GCH, Jansen RJT, et al. A common point mutation in the tyrosine hydroxylase gene in autosomal recessive l-dopa responsive dystonia (DRD) in the Dutch population. Hum Genet (in press).. 9 Westerink BHC, Mulder TBA. Determination of picomole amounts of dopamine, noradrenaline, 3,4-dihydroxyphenylalanine, 3,4-dihydroxphenylacetic acid, homovanillic acid, and 5-hydroxyindolacetic acid in nervous tissue after one-step purification on Sephadex G-10, using high-performance liquid chromatography with a novel type of electrochemical detection. J Neurochem 1981 ; 36 : 1449 -1462. Crossref Search ADS PubMed 10 Hyland K, Surtees RAH, Rodeck C, Clayton PT. Aromatic l-amino acid decarboxylase deficiency: clinical features, diagnosis, and treatment of a new inborn error of neurotransmitter amine synthesis. Neurology 1992 ; 42 : 1980 -1988. Crossref Search ADS PubMed 11 Gerrits GP, Trijbels FJ, Monnens LA, Gabreëls FJ, de Abreu RA, Theeuwes AdG, et al. Reference values for amino acids in cerebrospinal fluid of children determined using ion-exchange chromatography with fluorimetric detection. Clin Chim Acta 1989 ; 182 : 271 -280. Crossref Search ADS PubMed 12 Anderson GM, Hoder EL, Shaywitz BA, Cohen DJ. Neurotransmitter precursors and metabolites in CSF of human neonates. Dev Med Child Neurol 1985 ; 27 : 207 -214. PubMed 13 Hyland K, Surtees RAH, Heales SJR, Bowron A, Howells DW, Smith I. Cerebrospinal fluid concentrations of pterins and metabolites of serotonin and dopamine in a pediatric reference population. Pediatr Res 1993 ; 34 : 10 -14. Crossref Search ADS PubMed 14 Langlais PJ, Walsh FX, Bird ED, Levy HL. Cerebrospinal fluid neurotransmitter metabolites in neurologically normal infants and children. Pediatrics 1985 ; 75 : 580 -586. PubMed 15 Seifert WE, Foxx JL, Butler IJ. Age effect on dopamine and serotonin metabolite levels in cerebrospinal fluid. Ann Neurol 1979 ; 8 : 38 -42. 16 Sjöström R, Ekstedt J, Änggård E. Concentration gradients of monoamine metabolites in human cerebrospinal fluid. J Neurol Neurosurg Psychiatry 1975 ; 38 : 666 -668. Crossref Search ADS PubMed 17 Kruesi MJP, Swedo SE, Hamburger SD, Potter WZ, Rapoport JL. Concentration gradient of CSF monoamine metabolites in children and adolescents. Biol Psychiatry 1988 ; 24 : 507 -514. Crossref Search ADS PubMed 18 Langlais PJ, Bird ED, McEntee WJ. Stability of monoamine metabolites in human cerebrospinal fluid. Ann Neurol 1981 ; 12 : 48 -51. 19 Davidson L, Van dongen R, Beilin LJ. Effect of eating bananas on plasma free and sulfate-conjugated catecholamines. Life Sci 1981 ; 29 : 1773 -1778. Crossref Search ADS PubMed 20 Mashige F, Matsushima Y, Kanazawa H, Sakuma I, Takai N, Bessho F, et al. Acidic catecholamine metabolites and 5-hydroxyindolacetic acid in urine: the influence of diet. Ann Clin Biochem 1996 ; 33 : 43 -49. Crossref Search ADS PubMed 21 Sandler M, Uthven CRJ. The biosynthesis and metabolism of the catecholamines. Progr Med Chem 1969 ; : 200 -246. © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

García,, Antonia;Barbas,, Coral;Aguilar,, Rosa;Castro,, Mario

doi: 10.1093/clinchem/44.9.1905pmid: N/A

Abstract Organic acids analysis is a powerful technique in the diagnosis of inborn errors of metabolism. Clinically, patients present with severe symptoms, and early detection and appropriate treatment are often lifesaving. Most of the existing methods are based on gas chromatography in combination with mass spectrometry and require sophisticated equipment and complex sample pretreatment and derivatization. We propose a rapid, simple, and automated capillary electrophoretic method for routine analysis of urine to detect 27 organic acids related to metabolic diseases. With this method, direct measurements are performed on samples after initial centrifugation and dilution, if needed. Separation is performed in pH 6.0 phosphate buffer with methanol added as an organic modifier, −10 kV applied potential, and ultraviolet detection at 200 nm. The assay is completed in <15 min, and alternative separation conditions are proposed in case of overlapping peaks. The developed method allows the identification and quantitation of methylmalonic, pyroglutamic, and glutaric acids in samples of patients with diseases related to these acids. The analysis of organic acids in urine is a well-established procedure for the diagnosis of inherited disorders of amino acid and organic acid metabolism. The large number of organic acids in urine and the complexity of the mixture makes its separation and quantitation very difficult; gas chromatography–mass spectrometry is the most reliable technique for this purpose (1) . However, despite the number of methodologies described, they all require laborious sample pretreatment, expensive and sophisticated equipment, and highly qualified personnel; in fact, most laboratories spend a long time extracting, purifying, and derivatizing organic acids from urine before a patient sample is ready to be analyzed by gas chromatography–mass spectrometry. A rapid diagnosis of critically ill newborns who present with coma and metabolic acidosis is crucial to instituting the adequate therapy and avoiding fatal consequences. Capillary electrophoresis (CE) is suitable for detecting important changes in the metabolic profiles of body fluids and provides a rapid and simple alternative to other techniques in routine analysis (2)(3)(4)(5) . A previous study separated and quantified nine short-chain organic acids (6) . The aim of the present study was to separate and identify more organic acids, including new groups of aromatic acids, amino acids, and keto acids with relevance for diagnostic purposes. Three different internal standards (ISs) have been proposed and evaluated, and the intra- and interassay precision for migration time have been studied. Therefore, a rapid and simple screening method for acidurias has been developed, and some pathologic samples have been tested. Methylmalonic, pyroglutamic, and glutaric aciduria were detected in pathological samples in which these metabolites showed increased concentrations, and the peaks were easily identified. Materials and Methods apparatus Capillary zonal electrophoresis was performed on a Beckman System 5500 (P/ACE) equipped with an ultraviolet detector set at 200 nm, an automatic injector, and a 37-cm total length (75-μm i.d.), polyacrylamide gel-pretreated column cartridge. All experiments were carried out at 25 °C. Sample injections were made by pressure for 5 s with an applied reversed voltage of 10 kV for buffer S and 15 kV for buffer A. chemicals Calibrators Calibrators and ISs were obtained from Sigma Chemical Co. Aminoadipic acid, ketoglutaric acid disodium salt, 3-hydroxybutyric acid sodium salt, p-hydroxyphenylpyruvic, oxalacetic, and glycolic acids were 98% minimum purity; lactic acid lithium salt was 97% minimum purity. The rest of chemicals were analytical-reagent grade (>99% purity). Buffers Phosphoric acid (85%) was from Merck. Acetic acid and sodium hydroxide were from Panreac, and methanol was from Scharlau. Buffer solutions and all dilutions were prepared with water purified by a MilliQ-System (Millipore). The electrophoretic buffer S, pH 6.0, contained 0.2 mol/L phosphate and 100 mL/L methanol (6) . The second buffer used in the study, buffer A, was 0.2 mol/L phosphoric acid and 0.01 mol/L acetic acid adjusted to pH 4.0 with sodium hydroxide and did not contain methanol. samples Fresh urine samples were collected from healthy and ill babies under 4 months and refrigerated at −20 °C. Pathological samples corresponded to methylmalonic, glutaric, and pyroglutamic acidurias. Before analysis, samples were diluted with water (1 volume of sample plus 2 volumes of water) and centrifuged for 3 min at 2000g. Results and Discussion As shown in electropherogram A in Fig. 1 , 22 calibrators of different organic acids listed in Table 1 A were separated with buffer S as electrolyte at −10 kV in <12 min. Table 1 shows working, health-related, and pathological concentrations of these acids in urine; working concentrations are either in the pathological range or in a few cases slightly below it, except for oxalic acid, which is above the pathological range. However, the linear response for this acid in the present conditions has been proven in the range of 250-2000 mmol/mol creatinine, and therefore, lower concentrations can be detected. Within-run (n = 6) and between-run migration times (n = 6) with phthalic acid as IS and without an IS on a single day (within-run) and on 6 different days, with six different electrolyte batches and made by three different operators (between-run) were studied. The within-run CV ranged between 0.17% and 0.65% without an IS and between 0.04% and 0.39% with an IS. Because the between-run CV for migration times is higher without an IS (1.49–2.27%) than it is with one (0.06–1.40%), the use of the IS is especially recommended if the calibrator mixture is not run with the batch. Table 1. Organic acid concentrations in healthy and pathological urines. Peak number . Organic acid . Migration time, min . Working conc, mmol/L . Working conc1 . Health-related values1 . Pathological values1 . A. Buffer S Fig. 1 . Electropherogram A 1 Oxalic 4.84 0.80 1000 <100 100–350 2 Oxalacetic 5.48 0.20 250 3 Fumaric 5.61 0.05 63 <10 3000–4000 4 2-Ketoglutaric 6.00 0.40 500 <150 150–1100 5 Malic 6.19 2.00 2500 <100 6 Methylmalonic 6.31 2.00 2500 <2 150–15 500 7 Glutaric 6.63 4.00 5000 <10 500–22 000 8 Citric 6.74 0.80 1000 <300 9 Phthalic (IS) 7.07 0.05 63 10 N-Acetyl-l-aspartic 7.21 0.40 500 <2 1000–7000 11 Glycolic 7.32 4.00 5000 <100 >100 12 Acetoacetic 8.36 0.80 1000 <2 50–20 000 13 Propionic 8.49 1.00 1250 14 Lactic 8.64 1.00 1250 <100 1000–30 000 15 2-Ketoisovaleric 8.92 0.40 500 <2 300–800 16 3-Hydroxybutyric 10.05 4.00 5000 <100 100–50 000 17 2-Hydroxyisovaleric 10.53 2.00 2500 <2 850–3600 18 4-Methyl-n-valeric 11.13 4.00 5000 19 Phenyllactic 11.82 0.05 63 <2 200–1000 20 Homogentisic 12.24 0.05 63 <2 1000–5000 21 Hippuric 12.53 0.05 63 <300 22 Aminoadipic 13.81 2.00 2500 Fig. 1 . Electropherogram B 9 Phthalic (IS) 7.15 0.05 63 23 Pyroglutamic 10.19 0.50 625 42–115 4000–30 000 24 Orotic 10.57 0.05 63 0–11 30–5600 25 Xanthurenic 12.20 0.05 63 Fig. 1 . Electropherogram C 26 Pyruvic 7.10 0.80 1000 0–12 50–10 000 9 Phthalic (IS) 7.15 0.05 63 27 Phenylpyruvic 10.46 0.20 250 0–4 300–1000 28 4-Hydroxyphenylpyruvic 11.90 0.20 250 <2 140–2000 B. Buffer A Fig. 2 . Electropherogram A 1 Pyruvic 3.49 0.80 2 2-Ketocaproic (IS) 4.65 1.00 3 Orotic 4.83 0.05 4 Phenylpyruvic 4.93 0.20 5 Pyroglutamic 5.18 0.50 6 N-Acetyl-l-aspartic 5.26 0.40 7 Hydroxyphenylpyruvic 5.47 0.20 8 Xanthurenic 5.98 0.05 9 Phenyllactic 7.09 0.05 10 Tropic (IS)2 9.58 0.10 11 Homogentisic 11.88 0.05 Peak number . Organic acid . Migration time, min . Working conc, mmol/L . Working conc1 . Health-related values1 . Pathological values1 . A. Buffer S Fig. 1 . Electropherogram A 1 Oxalic 4.84 0.80 1000 <100 100–350 2 Oxalacetic 5.48 0.20 250 3 Fumaric 5.61 0.05 63 <10 3000–4000 4 2-Ketoglutaric 6.00 0.40 500 <150 150–1100 5 Malic 6.19 2.00 2500 <100 6 Methylmalonic 6.31 2.00 2500 <2 150–15 500 7 Glutaric 6.63 4.00 5000 <10 500–22 000 8 Citric 6.74 0.80 1000 <300 9 Phthalic (IS) 7.07 0.05 63 10 N-Acetyl-l-aspartic 7.21 0.40 500 <2 1000–7000 11 Glycolic 7.32 4.00 5000 <100 >100 12 Acetoacetic 8.36 0.80 1000 <2 50–20 000 13 Propionic 8.49 1.00 1250 14 Lactic 8.64 1.00 1250 <100 1000–30 000 15 2-Ketoisovaleric 8.92 0.40 500 <2 300–800 16 3-Hydroxybutyric 10.05 4.00 5000 <100 100–50 000 17 2-Hydroxyisovaleric 10.53 2.00 2500 <2 850–3600 18 4-Methyl-n-valeric 11.13 4.00 5000 19 Phenyllactic 11.82 0.05 63 <2 200–1000 20 Homogentisic 12.24 0.05 63 <2 1000–5000 21 Hippuric 12.53 0.05 63 <300 22 Aminoadipic 13.81 2.00 2500 Fig. 1 . Electropherogram B 9 Phthalic (IS) 7.15 0.05 63 23 Pyroglutamic 10.19 0.50 625 42–115 4000–30 000 24 Orotic 10.57 0.05 63 0–11 30–5600 25 Xanthurenic 12.20 0.05 63 Fig. 1 . Electropherogram C 26 Pyruvic 7.10 0.80 1000 0–12 50–10 000 9 Phthalic (IS) 7.15 0.05 63 27 Phenylpyruvic 10.46 0.20 250 0–4 300–1000 28 4-Hydroxyphenylpyruvic 11.90 0.20 250 <2 140–2000 B. Buffer A Fig. 2 . Electropherogram A 1 Pyruvic 3.49 0.80 2 2-Ketocaproic (IS) 4.65 1.00 3 Orotic 4.83 0.05 4 Phenylpyruvic 4.93 0.20 5 Pyroglutamic 5.18 0.50 6 N-Acetyl-l-aspartic 5.26 0.40 7 Hydroxyphenylpyruvic 5.47 0.20 8 Xanthurenic 5.98 0.05 9 Phenyllactic 7.09 0.05 10 Tropic (IS)2 9.58 0.10 11 Homogentisic 11.88 0.05 1 mmol/mol creatinine. 2 3-Hydroxy-2-phenylpropanoic acid. Open in new tab Table 1. Organic acid concentrations in healthy and pathological urines. Peak number . Organic acid . Migration time, min . Working conc, mmol/L . Working conc1 . Health-related values1 . Pathological values1 . A. Buffer S Fig. 1 . Electropherogram A 1 Oxalic 4.84 0.80 1000 <100 100–350 2 Oxalacetic 5.48 0.20 250 3 Fumaric 5.61 0.05 63 <10 3000–4000 4 2-Ketoglutaric 6.00 0.40 500 <150 150–1100 5 Malic 6.19 2.00 2500 <100 6 Methylmalonic 6.31 2.00 2500 <2 150–15 500 7 Glutaric 6.63 4.00 5000 <10 500–22 000 8 Citric 6.74 0.80 1000 <300 9 Phthalic (IS) 7.07 0.05 63 10 N-Acetyl-l-aspartic 7.21 0.40 500 <2 1000–7000 11 Glycolic 7.32 4.00 5000 <100 >100 12 Acetoacetic 8.36 0.80 1000 <2 50–20 000 13 Propionic 8.49 1.00 1250 14 Lactic 8.64 1.00 1250 <100 1000–30 000 15 2-Ketoisovaleric 8.92 0.40 500 <2 300–800 16 3-Hydroxybutyric 10.05 4.00 5000 <100 100–50 000 17 2-Hydroxyisovaleric 10.53 2.00 2500 <2 850–3600 18 4-Methyl-n-valeric 11.13 4.00 5000 19 Phenyllactic 11.82 0.05 63 <2 200–1000 20 Homogentisic 12.24 0.05 63 <2 1000–5000 21 Hippuric 12.53 0.05 63 <300 22 Aminoadipic 13.81 2.00 2500 Fig. 1 . Electropherogram B 9 Phthalic (IS) 7.15 0.05 63 23 Pyroglutamic 10.19 0.50 625 42–115 4000–30 000 24 Orotic 10.57 0.05 63 0–11 30–5600 25 Xanthurenic 12.20 0.05 63 Fig. 1 . Electropherogram C 26 Pyruvic 7.10 0.80 1000 0–12 50–10 000 9 Phthalic (IS) 7.15 0.05 63 27 Phenylpyruvic 10.46 0.20 250 0–4 300–1000 28 4-Hydroxyphenylpyruvic 11.90 0.20 250 <2 140–2000 B. Buffer A Fig. 2 . Electropherogram A 1 Pyruvic 3.49 0.80 2 2-Ketocaproic (IS) 4.65 1.00 3 Orotic 4.83 0.05 4 Phenylpyruvic 4.93 0.20 5 Pyroglutamic 5.18 0.50 6 N-Acetyl-l-aspartic 5.26 0.40 7 Hydroxyphenylpyruvic 5.47 0.20 8 Xanthurenic 5.98 0.05 9 Phenyllactic 7.09 0.05 10 Tropic (IS)2 9.58 0.10 11 Homogentisic 11.88 0.05 Peak number . Organic acid . Migration time, min . Working conc, mmol/L . Working conc1 . Health-related values1 . Pathological values1 . A. Buffer S Fig. 1 . Electropherogram A 1 Oxalic 4.84 0.80 1000 <100 100–350 2 Oxalacetic 5.48 0.20 250 3 Fumaric 5.61 0.05 63 <10 3000–4000 4 2-Ketoglutaric 6.00 0.40 500 <150 150–1100 5 Malic 6.19 2.00 2500 <100 6 Methylmalonic 6.31 2.00 2500 <2 150–15 500 7 Glutaric 6.63 4.00 5000 <10 500–22 000 8 Citric 6.74 0.80 1000 <300 9 Phthalic (IS) 7.07 0.05 63 10 N-Acetyl-l-aspartic 7.21 0.40 500 <2 1000–7000 11 Glycolic 7.32 4.00 5000 <100 >100 12 Acetoacetic 8.36 0.80 1000 <2 50–20 000 13 Propionic 8.49 1.00 1250 14 Lactic 8.64 1.00 1250 <100 1000–30 000 15 2-Ketoisovaleric 8.92 0.40 500 <2 300–800 16 3-Hydroxybutyric 10.05 4.00 5000 <100 100–50 000 17 2-Hydroxyisovaleric 10.53 2.00 2500 <2 850–3600 18 4-Methyl-n-valeric 11.13 4.00 5000 19 Phenyllactic 11.82 0.05 63 <2 200–1000 20 Homogentisic 12.24 0.05 63 <2 1000–5000 21 Hippuric 12.53 0.05 63 <300 22 Aminoadipic 13.81 2.00 2500 Fig. 1 . Electropherogram B 9 Phthalic (IS) 7.15 0.05 63 23 Pyroglutamic 10.19 0.50 625 42–115 4000–30 000 24 Orotic 10.57 0.05 63 0–11 30–5600 25 Xanthurenic 12.20 0.05 63 Fig. 1 . Electropherogram C 26 Pyruvic 7.10 0.80 1000 0–12 50–10 000 9 Phthalic (IS) 7.15 0.05 63 27 Phenylpyruvic 10.46 0.20 250 0–4 300–1000 28 4-Hydroxyphenylpyruvic 11.90 0.20 250 <2 140–2000 B. Buffer A Fig. 2 . Electropherogram A 1 Pyruvic 3.49 0.80 2 2-Ketocaproic (IS) 4.65 1.00 3 Orotic 4.83 0.05 4 Phenylpyruvic 4.93 0.20 5 Pyroglutamic 5.18 0.50 6 N-Acetyl-l-aspartic 5.26 0.40 7 Hydroxyphenylpyruvic 5.47 0.20 8 Xanthurenic 5.98 0.05 9 Phenyllactic 7.09 0.05 10 Tropic (IS)2 9.58 0.10 11 Homogentisic 11.88 0.05 1 mmol/mol creatinine. 2 3-Hydroxy-2-phenylpropanoic acid. Open in new tab Figure 1. Open in new tabDownload slide Electropherograms of calibrators and samples analyzed using buffer S. Electropherogram A, separation of 22 organic acids by capillary zonal electrophoresis; electropherograms Band C, new organic acids added. For peak identification, see Table 1A. Electropherogram D, healthy diluted (1:3) urine. Applied voltage, −10 kV (see Materials and Methods). Figure 1. Open in new tabDownload slide Electropherograms of calibrators and samples analyzed using buffer S. Electropherogram A, separation of 22 organic acids by capillary zonal electrophoresis; electropherograms Band C, new organic acids added. For peak identification, see Table 1A. Electropherogram D, healthy diluted (1:3) urine. Applied voltage, −10 kV (see Materials and Methods). Electropherogram D in Fig. 1 is a diluted (1:3) urine from a healthy volunteer. No interfering peaks appear in healthy urine under the present conditions (buffer S, −10 kV). Another six related organic acids—pyroglutamic, orotic, xanthurenic, pyruvic, phenylpyruvic and p-hydroxyphenylpyruvic, also included in Table 1 —were added to those previously separated. Electropherograms B and C in Fig. 1 show overlapping with some of the other 22 compounds. Pyruvic, phenyl pyruvic, and p-hydroxyphenyl pyruvic acids with these electrophoretic conditions gave wide peaks, probably because of a tautomeric ketoenolic equilibrium. Pyruvic acid overlaps with phthalic, N-acetylaspartic, and glycolic acids; pyroglutamic acid overlaps with 3-hydroxybutyric acid; phenylpyruvic acid overlaps with orotic and hydroxyisovaleric acids; and p-hydroxyphenyl pyruvic and phenyllactic acids overlap with xanthurenic and homogentisic acids. In view of this, if a peak was increased in a suspicious pathological sample in the corresponding migration time, a new buffer system was developed to clearly confirm the compound. Electropherogram A in Fig. 2 shows the complete separation of the calibrators in Table 1B in a system using buffer A run at −15 kV. Pyruvic acid appeared at 3.49 min in the new buffer; N-acetylaspartic appeared at 5.26 min. Glycolic acid did not appear, and phthalic acid interference was easy to eliminate because it was the IS. Pyroglutamic acid appeared at 5.18 min, whereas 3-hydroxybutyric acid did not. Phenylpyruvic acid appeared at 4.93 min and orotic acid at 4.83 min; however, their appearances are very different. Finally, hydroxyisovaleric did not appear; p-hydroxyphenyl pyruvic acid appeared at 5.47 min, xanthurenic acid at 5.98 min, phenyllactic acid at 7.09 min, and homogentisic at 11.88 min. To improve precision, two ISs were added: 2-ketocaproic acid (4.65 min) and tropic acid (9.58 min). Within-run (n = 6) imprecision and total between-run imprecision with and without the ISs were compared. Within-run imprecision studies without the IS gave CV values ranging between 0.33% and 1.97%, whereas with relative migration times referred to the closest IS, the CV ranged between 0.10% and 0.75%. The total between-run CV ranged between 1.25% and 4.93% without the IS and between 0.22% and 1.49% with relative migration times referred to the closest IS. Electropherogram B in Fig. 2 belongs to a diluted (1:3) healthy urine analyzed under the same conditions with buffer A and shows no interferences. Figure 2. Open in new tabDownload slide Electropherograms of peaks that overlapped in buffer S, analyzed using buffer A. Electropherogram A, separation of the overlapped peaks. For peak identification, see Table 1B. Electropherogram B, healthy diluted (1:3) urine. Applied voltage, −15 kV (see Materials and Methods). Figure 2. Open in new tabDownload slide Electropherograms of peaks that overlapped in buffer S, analyzed using buffer A. Electropherogram A, separation of the overlapped peaks. For peak identification, see Table 1B. Electropherogram B, healthy diluted (1:3) urine. Applied voltage, −15 kV (see Materials and Methods). suggested complete method First, a daily run of the calibrator mixture is recommended, although when working with an IS, it could be run alone and relative retention times taken as an index. Second, diluted (1:3) and centrifuged samples will be analyzed using buffer S. If any peak with a migration time corresponding to a calibrator increases, a pathology may be suspected; it should be confirmed by co-injecting the presumed compound with the sample or adding the compound to the sample. Third, if one of the compounds corresponding to the retention time of the calibrators listed in Table 1, A and B, increases, a second analysis using buffer A is recommended to confirm the assignment. analysis of urine samples Organic acidurias, although clinically important, are not common, and samples are difficult to obtain. It may be that the diagnosis is missed in some cases because the disorders are not screened at birth. Samples were provided by Hospital Virgen del Rocío, Hospital la Macarena, and Hospital La Paz and were obtained in agreement with the ethics committees of the respective centers. Pathological samples corresponding to methylmalonic and glutaric acidurias were tested following the method described above; the electropherograms are shown in Fig. 3 , where healthy urine, pathological samples, and calibrators are included. Pyroglutamic urine was diluted with water using 1 volume of sample and 19 volumes of water (final dilution, 1:20) because of the high concentration of pyroglutamic acid in it. The electropherogram including a healthy urine and the corresponding calibrators is shown in Fig. 4 . When a peak related to these diseases appears, it is clearly differentiated from a healthy sample. Figure 4. Open in new tabDownload slide Electropherograms of urine from a patient with pyroglutamic aciduria compared with healthy urine and calibrators. Conditions: applied voltage, −10 kV; buffer S (see Materials and Methods). Figure 4. Open in new tabDownload slide Electropherograms of urine from a patient with pyroglutamic aciduria compared with healthy urine and calibrators. Conditions: applied voltage, −10 kV; buffer S (see Materials and Methods). Figure 3. Open in new tabDownload slide Electropherograms of urine from patients with methylmalonic or glutaric aciduria compared with healthy urine and calibrators. Conditions: applied voltage, −10 kV; buffer S (see Materials and Methods). Figure 3. Open in new tabDownload slide Electropherograms of urine from patients with methylmalonic or glutaric aciduria compared with healthy urine and calibrators. Conditions: applied voltage, −10 kV; buffer S (see Materials and Methods). Before quantitation, the calibrators and sample linearity were tested; correlation coefficients were found to be >0.99. The concentrations obtained were as follows: methylmalonic acid, 0.011 mol/L (3089 mmol/mol creatinine); pyroglutamic acid, 0.017 mol/L (25 682 mmol/mol creatinine); and glutaric acid, 0.055 mol/L (82 879 mmol/mol creatinine). The interpretation of organic acid concentrations for diagnostic purposes depends heavily on a pattern of abnormalities because the increase of a single compound may not be diagnostic. Relative amounts of compounds can also be informative. Some disorders, such as propionic acidemia, methylmalonic acidemia, pyroglutamic acidemia, and glutaric acidemia, can be reliably diagnosed from organic acid excretions because of the consistently high increases of characteristic acids. Methylmalonic acidemia results from the deficiency of the cobalamin-dependent enzyme methylmalonyl-CoA mutase (1) . It is one of the most frequently diagnosed organic acidurias (6 cases in 3 years from 1000 children) (6) ; chemically the urine of a patient with this disorder is characterized by large amounts of methylmalonic acid, which is almost undetectable in the urine of healthy subjects (7) . During a ketotic crisis, the increase of ketone bodies such as 3-hydroxybutyrate is higher than the methylmalonic peak (8) . In Fig. 3 , the methylmalonic acid peak is clearly increased. Prenatal detection has been accomplished by measurement of methylmalonate in amniotic fluid and maternal urine at midtrimester (9) ; as in some other inherited metabolic disorders, treatment in the early weeks or months of life is most important (7) . Pyroglutamic acidemia, or 5-oxoprolinuria, is caused by a glutathione synthetase deficiency, an inherited metabolic condition that may show in early infancy as persistent or acute metabolic acidosis associated with chronic hemolytic anemia (1) . The presence or absence of ketonuria associated with metabolic acidosis is the major clinical key to the diagnosis; when metabolic acidosis occurs with an anion gap within reference values and without hyperlacticacidemia or hypoglycemia, pyroglutamic aciduria is rarely diagnosed (1 case in 3 years from 1000 children) (6) , and it may show early in life with constant, isolated metabolic acidosis. If an increased pyroglutamic peak is detected in patient urine, pyroglutamic aciduria may be diagnosed instead of renal tubular acidosis type II (10) . Glutaric acidemia is caused by an isolated deficiency of mitochondrial glutaryl-CoA dehydrogenase (glutaric aciduria type I) or by the deficiency of mitochondrial electron transport flavoprotein or electron transport flavoprotein dehydrogenase (1) . Diagnosis is made on the basis of increased glutaric and 3-hydroxyglutaric acids in urine (10) . Most infants whose treatment began before the onset of symptoms have developed normally (11) , even those with a prenatal diagnosis of glutaric aciduria type I after the discovery of previous cases in the same family (12) . Under those circumstances, the interest of early diagnosis is enormous. Fig. 3 shows the marked glutaric peak in a pathological sample, which is >1000-fold higher than in healthy urine (10) and is usually out of detection limits. In reported results on screening of organic acidemias, the incidence ranges between 4% (7) and 6.3% ((13)). However, these reports depend mostly on the selection of patients because, except for classical phenylketonuria, the disorders are not screened at birth. Furthermore, diagnoses could have been missed in some cases because of the severe vital prognosis, unless a specific treatment was immediately instituted, at least for some of them. Moreover, the number of diagnosed inborn errors of metabolism is growing constantly because of the improvement and widespread availability of analytical techniques (14) . Because of the importance of genetic counseling in such diseases, the diagnosis is valuable even in less urgent cases or in postmortem samples (7) . Conclusion The CE method described as applied to urine is rapid, automated, simple, and inexpensive. It requires only a small volume of urine (50 μL) and no sample preparation. It permits separation, detection, and even identification in <15 min of a wide range of organic acids related to metabolic disorders. The urgent evaluation of a critically ill newborn is a frequent incident in neonatal intensive care units. Lethargy, coma, vomiting, seizures, and death occur in the first few days of life; thus, the time spent in diagnosis is crucial. Finally, the proposed method allows the operator to become familiar with the patterns of nonpathological samples and to recognize immediately “true abnormal profiles” in urine of children. It could be a valuable tool in the routine diagnostic system, mainly in newborns, applied to severe diseases that need a specific and early treatment. Facultad de CC Experimentales y Técnicas. Universidad S. Pablo-CEU, Urbanización Montepríncipe Ctra. Boadilla del Monte, km 5,3–28668 Madrid, Spain. The present study was supported by Universidad S. Pablo-CEU project No. 12/97. We are also grateful to all the persons who have provided urine samples from newborns, to Fidel Gayoso for kind clinical support, and to Enrique Torija for technical support. References 1 Lehotay DC, Clarke JT. Organic acidurias and related abnormalities. Crit Rev Clin Lab Sci 1995 ; 32 : 377 -429. Crossref Search ADS PubMed 2 Jellum E, Dollekamp H, Brunsvig A, Gislefoss R. Diagnostic applications of chromatography and capillary electrophoresis. J Chromatogr B 1997 ; 689 : 155 -164. Crossref Search ADS 3 Jellum E, Thorsrud AK, Time E. Capillary electrophoresis for diagnosis and studies of human disease, particularly metabolic disorders. J Chromatogr 1991 ; 559 : 455 -465. Crossref Search ADS PubMed 4 Jellum E, Dollekamp H, Blessum C. Capillary electrophoresis for clinical problem solving: analysis of urinary organic diagnostic metabolites and serum proteins. J Chromatogr B 1996 ; 683 : 55 -65. Crossref Search ADS 5 Hong J, Baldwin RP. Profiling clinically important metabolites in human urine by capillary electrophoresis and electrochemical detection. J Capillary Electrophor 1997 ; 4 : 65 -71. PubMed 6 Barbas C, Adeva A, Aguilar R, Rosillo M, Rubio T, Castro M. Quantitative determination of short chain organic acids in urine by capillary electrophoresis. Clin Chem 1998 ; 44 : 1340 -1342. Crossref Search ADS PubMed 7 Divry P, Vianey Liaud C, Cotte J. Routine gas chromatographic/mass spectrometric analysis of urinary organic acids. Results over a three-year period. Biomed Environ Mass Spectrom 1987 ; 14 : 663 -668. Crossref Search ADS PubMed 8 Fenton WA, Rosenberg LE. Disorders of propionate and methylmalonate metabolism. Scriver CR Beaudet AL Sly WS Valle D eds. The metabolic and molecular basis of inherited disease 7th ed. 1995 : 1423 McGraw-Hill New York. . 9 Tavares de Almeida I, Duran M, Silva MFB, Portela R, Cabral A, Tasso T, et al. Mild form of methylmalonic aciduria misdiagnosed as propionic acidemia during a ketotic crisis. J Inherit Metab Dis 1991 ; 14 : 259 -262. Crossref Search ADS PubMed 10 Mahoney MJ, Rosenberg LE, Lindbland B, Waldenstrom J, Zetterstrom R. Prenatal diagnosis of methylmalonic acidemia. Acta Paediatr Scand 1975 ; 64 : 44 -51. PubMed 11 Stephen. Clinical phenotypes: diagnosis/algorithms. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular basis of inherited disease, 7th ed. New York: McGraw-Hill, 1995:339.. 12 Iafolla AK, Kahler SG. Megaloencephaly in the neonatal period as the initial manifestation of glutaricaciduria type I. J Pediatr 1989 ; 114 : 1004 -1008. Crossref Search ADS PubMed 13 Jellum E, Thorensen O, Horn L, Seip R, Nilsen E, Kvittingen EA, Syokke O. Advances in the use of computerized gas chromatography-mass spectrometry and high-performance liquid chromatography with rapid scanning detection for clinical diagnosis. J Chromatogr 1989 ; 468 : 43 -53. Crossref Search ADS PubMed 14 Coelho JC, Wajner M, Burin MG, Vargas CR, Giugliani R. Selective screening of 10,000 high-risk Brazilian patients for the detection of inborn errors of metabolism. Eur J Pediatr 1997 ; 156 : 650 -654. Crossref Search ADS PubMed © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Bleier,, Jürgen;Vorderwinkler,, Karl-Paul;Falkensammer,, Jürgen;Mair,, Peter;Dapunt,, Otto;Puschendorf,, Bernd;Mair,, Johannes

doi: 10.1093/clinchem/44.9.1912pmid: N/A

Abstract We investigated the net myocardial release of creatine kinase isoenzyme MB (CKMB), myoglobin, cardiac troponin T (cTnT), cardiac troponin I (cTnI), and cardiac β-type myosin heavy chain (β-MHC) into the coronary circulation after cardioplegic cardiac arrest in humans. Cardiac markers were measured in paired arterial, central venous, and coronary sinus blood in 19 patients undergoing elective coronary artery bypass grafting (CABG) before aortic cross-clamping and 1, 5, 10, and 20 min after aortic declamping. cTnT and cTnI were released into the coronary sinus in parallel to each other and almost simultaneously to myoglobin and CKMB within 20 min of reperfusion. In contrast, no β-MHC was released in the same patients during the study period. The average soluble cTnT and cTnI pools in right atrial appendages of 11 patients with right atrial and right ventricular pressures within reference values were comparable and were ∼8% of total myocardial troponin content. The soluble β-MHC pool was <0.1% in all patients. Our results demonstrate the impact of the different intracellular compartmention of regulatory and contractile proteins on their early release from damaged myocardium. cTnT, cardiac troponin T, cTnI, cardiac troponin I, CK, creatine kinase, CKMB, creatine kinase MB isoenzyme, MHC, myosin heavy chain, β-MHC, cardiac β-type heavy chain myosin, LD, lactate dehydrogenase, CABG, coronary artery by-pass grafting. Laboratory measurement of cardiac regulatory and contractile proteins has considerably improved the diagnosis of myocardial damage caused by several diseases, such as acute and perioperative myocardial infarction, myocarditis, or heart contusion (1)(2) . Troponin T and troponin I are regulatory proteins of the muscular thin filaments and part of the troponin-tropomyosin complex in striated muscle. They exist as specific isoforms, cardiac troponin T (cTnT)4 and cardiac troponin I (cTnI), in the myocardium (3)(4) . The conventional markers creatine kinase (CK) and its MB isoenzyme (CKMB) as well as myoglobin are predominantly cytosolic proteins and not heart-specific. Cardiac β-type myosin heavy chain (β-MHC) is the predominant MHC type in human adult healthy and diseased myocardium (5) . Together with the myosin light chains, β-MHC forms cardiac myosin, the major structural protein of the myocardium. Cardiac β-MHC is co-expressed in slow twitch skeletal muscle fibers and is therefore not cardiac-specific (5) . Previous clinical (1)(6)(7)(8) and experimental (9)(10)(11) studies investigated the release kinetics of myocardial macromolecules after myocardial damage. Our recent experimental observations showed that cTnT, cTnI, CK, and lactate dehydrogenase (LD) increase in parallel in effluents from isolated perfused Langendorff rat hearts after 60 min of hypoxia-induced myocardial damage (11) . In open heart surgery the aortic cross-clamping with cardioplegic cardiac arrest induces a global myocardial ischemia, and hypothermia is induced to protect the myocardium. This human model of controlled ischemia facilitates the investigation of the pathophysiological events occurring during myocardial ischemia and reperfusion in humans. Unique to our study were the direct determination of the net myocardial release of CKMB, myoglobin, cTnT, cTnI, and β-MHC into the coronary circulation during reperfusion after cardioplegic cardiac arrest in patients undergoing elective coronary artery by-pass grafting (CABG) and the quantification of soluble cTnI, TnT, and β-MHC pools in human fresh myocardium. Patients and Methods patients All procedures were in accordance with the Helsinki Declaration of 1975, as revised in 1983. After institutional approval and informed consent, 19 patients undergoing elective CABG were studied (16 men, 3 women; median age, 60 years; range, 37–73 years). None of the 19 patients had clinical signs of congestive heart failure. The median preoperative left ventricular ejection fraction was 56% (range, 18–86%, <40% in three patients). A three-vessel disease was found in 14 of the 19 patients, with left main artery stenosis in 3 of them. Patients received a median of three grafts (range, 2–4). The internal mammary artery was used as a by-pass vessel in 15 patients. Perioperative myocardial infarction was diagnosed when CKMB exceeded 50 μg/L in the morning of the first postoperative day, and new Q waves developed perioperatively in at least two contiguous leads of the electrocardiogram (12) . anesthesia, cardiopulmonary by-pass technique, and cardioplegia Anesthesia was induced with 0.3–0.4 mg/kg midazolam and 5–10 μg/kg fentanyl in all patients. Endotracheal intubation was facilitated with 0.1 mg/kg vecuronium. Anesthesia was maintained using a continuous infusion of fentanyl (20 μg · kg−1 · h−1) and midazolam (0.15 mg · kg−1 · h−1). Additional bolus doses of fentanyl and isoflurane (<1.5%) were administered according to clinical requirements. All patients received an initial bolus dose of aprotinin (280 mg) over a period of 15 min followed by a continuous infusion (70 mg/h). A standard cardiopulmonary by-pass technique (roller pump, membrane oxygenator, and cardiotomy reservoir) with moderate systemic hypothermia (core temperature, 30–32 °C) was used in all patients. The extracorporal circuit was primed with 100 mL of 200 g/L human albumin, 250 mL of 200 g/L mannitol, and Ringer’s lactate. The pump prime also contained 280 mg of aprotinin. Blood was added to the prime only when preoperative hemoglobin was <100 g/L. After aortic and right atrial cannulation, a 14F coronary sinus retrograde perfusion catheter was inserted. Cardiopulmonary by-pass was started, and the body core temperature was reduced to 30–32 °C. Additional topical cooling of the heart was applied. The heart started to fibrillate with topical cooling in all patients, and the coronary arteries were inspected. Thereafter the aorta was cross-clamped. Myocardial protection during aortic cross-clamping was achieved by combined antegrade and retrograde, cold, multiple dose hyperkalemic cardioplegia: 1000–1200 mL of cold (6–8 °C) St. Thomas Hospital II solution was infused into the aortic root, followed by 300 mL infused via the coronary sinus perfusion catheter. After completion of each peripheral anastomosis, an additional dose of 300–500 mL of cardioplegic solution was infused through the aortic root, the venous grafts, and the coronary sinus perfusion catheter. Oxygenated blood from the by-pass circuit was added to these additional doses of cardioplegia (ratio of blood to St. Thomas solution, 1:3). Systemic rewarming was started while the final peripheral anastomosis was performed. After aortic declamping, high by-pass flows were maintained to completely unload the beating heart and prevent it from ejecting blood during reperfusion. blood sampling and laboratory analysis A single central venous blood sample (baseline sample) was obtained immediately after induction of anesthesia. Additional paired arterial, central venous, and coronary sinus blood samples were obtained after atrial cannulation before aortic cross-clamping (preischemic sample) and 1, 5, 10, and 20 min after aortic declamping (reperfusion samples). Blood was collected from the arterial line of the by-pass circuit (arterial sample), from the central venous catheter (central venous sample; correct position of the catheter in the central venous circulation was assured by chest x-ray), and from the pressure monitoring line of the coronary sinus perfusion catheter (coronary sinus sample; correct position of the perfusion catheter was assured by the surgeon each time before blood was sampled). All blood samples were assayed for lactate, cTnT, cTnI, CKMB mass, myoglobin, and cardiac β-MHC. Lactate was determined without delay. Blood samples for cTnT, cTnI, CKMB mass, myoglobin, and β-MHC measurements were centrifuged immediately (2000g for 15 min), and the plasma was frozen and stored at −20 °C until analysis. To adjust for hemodilution during by-pass, the results of cTnT, cTnI, CKMB mass, myoglobin, and β-MHC were expressed per gram of total serum protein. Total protein concentrations were measured by the Biuret method (Merck). The reference interval is 67–87 g/L. LD and lactate LD activity and lactate concentrations were determined enzymatically (Boehringer Mannheim). Myocardial lactate production was calculated as the coronary sinus lactate concentration minus the arterial lactate concentration. The cumulative myocardial lactate production during reperfusion was calculated as the mean of all four measurements during reperfusion. cTnT and cTnI Commercially available enzyme immunoassays developed by Katus et al. (13) (Boehringer Mannheim) and by Larue et al. (14) (ERIA Diagnostics Pasteur) were used for cTnT and cTnI determination, respectively. The upper limit of the reference interval for the cTnT assay used was 0.2 μg/L (13) . The enzyme immunometric assay used for cTnI measurements showed no cross-reactivity with skeletal muscle troponin I or other cardiac proteins (14) . The upper limit of the reference interval for cTnI is 0.1 μg/L. CKMB CKMB mass concentration was measured by a microparticle enzyme immunoassay (Abbott) for use with the Abbott IMx automated analyzer (15) . The upper limit of the reference interval is 5 μg/L. Myoglobin Myoglobin was determined by a commercially available immunoassay (OPUS, Behringwerke AG). The upper limit of the reference interval is 70 μg/L (16) . β-MHC Cardiac β-MHC plasma concentrations were measured by an immunoradiometric assay (ERIA Diagnostics Pasteur) developed by Larue et al. (17) . Because of the strong structural similarity of cardiac β-MHC and MHC of slow-twitch skeletal muscle fibers, this assay cross-reacts strongly with human slow-twitch skeletal MHC. The upper limit of the reference interval is 400 microunits/L; one microunit/L corresponds to 1 μg/L (17) . The myocardial release of cardiac markers (CKMB, myoglobin, cTnT, cTnI, and β-MHC) was calculated as the coronary sinus concentration minus the corresponding arterial concentration at the same measuring time point. Cumulative cardiac marker release was calculated as the mean of net release at all four measuring time points during reperfusion. determination of the soluble cTnT, cTnI, and β-MHC pool in atrial tissue specimens At our institution, right atrial appendages are excised as part of the routine surgical procedure during cannulation in CABG patients when a standard cardiopulmonary by-pass technique is used during heart surgery. To avoid a possible bias from hemodynamic overload of the right heart, only right atrial appendages from 11 patients with right ventricular and right atrial pressures within reference values were used to determine the soluble cTnT, cTnI, and β-MHC pools in fresh myocardium. After excision, the tissue was immediately rinsed in ice-cold cardioplegic solution, shock-frozen in liquid nitrogen, and stored at −80 °C until further analysis. The sarcoplasmatic and structurally bound fractions of cardiac regulatory and contractile proteins were determined according to a previously published protocol (6)(18) . Briefly, the tissue was homogenized in a buffer (0.05 mol/L Tris hydrochloride, 2 mmol/L EDTA, and 0.5 mmol/L dithiothreitol, pH 7.0) and stirred at 4 °C for 1 h. Thereafter the insoluble molecules were sedimented by ultracentrifugation (1 h, 100 000g, 4 °C). The pellet was washed, and centrifugation was repeated. Troponins were extracted from insoluble myofibrils by homogenization of the precipitate in buffer containing 0.4 mol/L potassium chloride, 0.1 mol/L potassium dihydrogen phosphate, 0.05 mol/L dipotassium hydrogen phosphate, 0.04 mol/L sodium pyrophosphate, and 0.01 mol/L magnesium chloride, pH 7.0, and stirred for 1 h at 4 °C. The solubilized troponin complex was separated from insoluble actomyosin and cellular debrids by centrifugation (1 h, 20000g, 4 °C). This extraction step was repeated once. The remaining pellet with actin-myosin filaments was resuspended in 50 mmol/L sodium pyrophosphate (pH 7.4) for β-MHC measurements. The LD, CKMB, myoglobin, troponin, and β-MHC concentrations were measured in the soluble and bound fractions. LD, a soluble cytosolic enzyme, was used to check the quality of fraction separation by the centrifugation protocol used. The efficacy of this protocol for troponin solubilization has been documented previously (6) . The cytosolic fraction was divided by the total fraction (cytosolic plus myofibrillar) to obtain the percentage of the measured markers in the cytosolic fractions. statistics Cumulative myocardial lactate production and cardiac marker release were calculated as the mean of all four measurements during reperfusion according to Matthews et al. (19) . The medians, ranges, and interquartile ranges were calculated to describe continuous variables. Spearman rank correlation coefficients were calculated to describe the association between variables. Nonparametric analysis of variance (Friedman two-way ANOVA) and the Wilcoxon matched pairs signed-ranks test were used for statistical analysis. P values <0.05 were considered significant. Results None of the 19 patients fulfilled the criteria for diagnosing perioperative myocardial infarction, and none of the patients suffered postby-pass cardiac failure. All were easily weaned from extracorporal circulation with a dose of dopamine of <5 μg · kg−1 · min−1. The median by-pass time was 104 min (range, 56–181 min). The median aortic cross-clamping time was 54 min (range, 24–84 min). Moderate myocardial ischemia during aortic cross-clamping was indicated by a myocardial production of lactate immediately after aortic declamping (median, 34 mg/L; range, 2.0–77.0 mg/L 1 min after aortic declamping). There were no significant correlations between aortic cross-clamping times and cumulative cardiac marker release. However, we found moderately significant correlations between cumulative lactate production and cumulative cardiac marker net release (myoglobin: r = 0.49, P = 0.037; CKMB: r = 0.71, P = 0.002; cTnI: r = 0.54, P = 0.017; cTnT: r = 0.76, P = 0.0001). myoglobin, ckmb mass, cTnT, and cTnI Intraoperatively, the concentrations of myoglobin, CKMB mass, cTnT, and cTnI rose significantly (P <0.0004) within 20 min of reperfusion compared with the central venous baseline values obtained immediately after induction of anesthesia. Myoglobin, CKMB mass, cTnT, and cTnI markedly exceeded their upper reference limits. The highest concentrations of all four analytes were measured in the coronary sinus (from 18-fold to 62-fold above the upper reference limit) after aortic declamping (Figs. 1–4). Coronary sinus myoglobin and CKMB mass concentrations were significantly (P <0.039) higher than corresponding arterial concentrations in all four paired blood samples obtained after aortic declamping, indicating myoglobin and CKMB mass release from the human heart. There was no release of myoglobin and CKMB before aortic cross-clamping, and in contrast, there was considerable release of both markers within 20 min after aortic declamping (Figs. 1 and 2). These arterial coronary sinus myoglobin and CKMB mass concentration differences showed significant (P <0.014) increases immediately (with the first minute) after aortic declamping compared with baseline values before aortic cross-clamping, which indicated a very rapid release from the myocardium with the onset of reperfusion. Figure 1. Open in new tabDownload slide Myoglobin concentrations in arterial, central venous, and coronary sinus blood during reperfusion after cardioplegic cardiac arrest in 19 uneventful patients. Concentrations are given per gram of total serum protein. Data given as median (bars) and interquartile range (IQR, error bars). HLM, heart-lung machine; ACC, aortic cross-clamping; AUC, aortic unclamping. *, significantly higher than corresponding arterial HLM myoglobin concentration. Figure 1. Open in new tabDownload slide Myoglobin concentrations in arterial, central venous, and coronary sinus blood during reperfusion after cardioplegic cardiac arrest in 19 uneventful patients. Concentrations are given per gram of total serum protein. Data given as median (bars) and interquartile range (IQR, error bars). HLM, heart-lung machine; ACC, aortic cross-clamping; AUC, aortic unclamping. *, significantly higher than corresponding arterial HLM myoglobin concentration. Figure 2. Open in new tabDownload slide CKMB mass concentrations in arterial, central venous, and coronary sinus blood during reperfusion after cardioplegic cardiac arrest in 19 uneventful patients. Concentrations are given per gram of total serum protein. Data given as median (bars) and interquartile range (IQR, error bars). HLM, heart-lung machine; ACC, aortic cross-clamping; AUC, aortic unclamping. *, significantly higher than corresponding arterial HLM CKMB concentration. Figure 2. Open in new tabDownload slide CKMB mass concentrations in arterial, central venous, and coronary sinus blood during reperfusion after cardioplegic cardiac arrest in 19 uneventful patients. Concentrations are given per gram of total serum protein. Data given as median (bars) and interquartile range (IQR, error bars). HLM, heart-lung machine; ACC, aortic cross-clamping; AUC, aortic unclamping. *, significantly higher than corresponding arterial HLM CKMB concentration. Figure 3. Open in new tabDownload slide cTnT concentrations in arterial, central venous, and coronary sinus blood during reperfusion after cardioplegic cardiac arrest in 19 uneventful patients. Concentrations are given per gram of total serum protein. Data given as median (bars) and interquartile range (IQR, error bars). HLM, heart-lung machine; ACC, aortic cross-clamping; AUC, aortic unclamping. *, significantly higher than corresponding arterial HLM cTnT concentration. Figure 3. Open in new tabDownload slide cTnT concentrations in arterial, central venous, and coronary sinus blood during reperfusion after cardioplegic cardiac arrest in 19 uneventful patients. Concentrations are given per gram of total serum protein. Data given as median (bars) and interquartile range (IQR, error bars). HLM, heart-lung machine; ACC, aortic cross-clamping; AUC, aortic unclamping. *, significantly higher than corresponding arterial HLM cTnT concentration. Figure 4. Open in new tabDownload slide cTnI concentrations in arterial, central venous, and coronary sinus blood during reperfusion after cardioplegic cardiac arrest in 19 uneventful patients. Concentrations are given per gram of total serum protein. Data given as median (bars) and interquartile range (IQR, error bars). HLM, heart-lung machine; ACC, aortic cross-clamping; AUC, aortic unclamping. *, significantly higher than corresponding arterial HLM cTnI concentration. Figure 4. Open in new tabDownload slide cTnI concentrations in arterial, central venous, and coronary sinus blood during reperfusion after cardioplegic cardiac arrest in 19 uneventful patients. Concentrations are given per gram of total serum protein. Data given as median (bars) and interquartile range (IQR, error bars). HLM, heart-lung machine; ACC, aortic cross-clamping; AUC, aortic unclamping. *, significantly higher than corresponding arterial HLM cTnI concentration. The concentrations of coronary sinus cTnT and cTnI after aortic declamping were also higher than the corresponding arterial concentrations, indicating considerable myocardial net release of troponins within 20 min after reperfusion (Figs. 3 and 4). Five, 10, and 20 min after aortic declamping, myocardial cTnT and cTnI release was significantly (P <0.033) higher compared with baseline values before aortic cross-clamping. In contrast to myoglobin and CKMB, the difference among baseline and release values for cTnT and cTnI at 1 min after aortic declamping were not significant. cardiac β-mhc We found no significant increase in β-MHC compared with baseline values within the first 20 min of reperfusion after hypothermic cardioplegic cardiac arrest during CABG. In all 19 patients, β-MHC concentrations did not increase over the upper limit of the reference interval during our observation period. The highest value measured (157 microunits/L) was found in the coronary sinus blood of one patient 20 min after reperfusion, which was still clearly within the reference interval. quantification of soluble troponin and myosin pools We found a considerable amount of unbound troponins in the cytoplasma of myocardium, whereas the soluble β-MHC pool was negligible when compared with the total MHC content (Table 1). Table 1. Intracellular compartmentation of cardiac markers in fresh human myocardium.1 . Cytosolic pool . . . Per g of wet weight . % of total content . cTnI 1027 μg (417–1265)2 8.3 (4.5–10.9)2 cTnT 278 μg (194–396) 7.5 (4.3–10.6) Myoglobin 973 μg (623–1606) 98 (96–99) CKMB 111 μg (62–228) 91 (78–94) β-MHC 670 milliunits (410–930) <0.1 in all . Cytosolic pool . . . Per g of wet weight . % of total content . cTnI 1027 μg (417–1265)2 8.3 (4.5–10.9)2 cTnT 278 μg (194–396) 7.5 (4.3–10.6) Myoglobin 973 μg (623–1606) 98 (96–99) CKMB 111 μg (62–228) 91 (78–94) β-MHC 670 milliunits (410–930) <0.1 in all 1 Data of 11 right atrial appendages from hearts with right ventricular and right atrial pressures within reference values. Appendages were excised during surgery as part of the routine procedure. 2 Data given as median and interquartile range (in parentheses). Open in new tab Table 1. Intracellular compartmentation of cardiac markers in fresh human myocardium.1 . Cytosolic pool . . . Per g of wet weight . % of total content . cTnI 1027 μg (417–1265)2 8.3 (4.5–10.9)2 cTnT 278 μg (194–396) 7.5 (4.3–10.6) Myoglobin 973 μg (623–1606) 98 (96–99) CKMB 111 μg (62–228) 91 (78–94) β-MHC 670 milliunits (410–930) <0.1 in all . Cytosolic pool . . . Per g of wet weight . % of total content . cTnI 1027 μg (417–1265)2 8.3 (4.5–10.9)2 cTnT 278 μg (194–396) 7.5 (4.3–10.6) Myoglobin 973 μg (623–1606) 98 (96–99) CKMB 111 μg (62–228) 91 (78–94) β-MHC 670 milliunits (410–930) <0.1 in all 1 Data of 11 right atrial appendages from hearts with right ventricular and right atrial pressures within reference values. Appendages were excised during surgery as part of the routine procedure. 2 Data given as median and interquartile range (in parentheses). Open in new tab Discussion The present study investigated the net myocardial release kinetics of cardiac markers during reperfusion after hypothermic cardioplegic cardiac arrest in CABG patients. The novel and unique approach of this study was to measure these cardiac markers in parallel in central venous, coronary sinus, and arterial blood samples. In our study population, we found significant myocardial release of myoglobin, CKMB mass, cTnT, and cTnI into the coronary circulation after aortic declamping within 20 min of reperfusion. The myocardial net release of CKMB, myoglobin, and cardiac troponins indicates that myocardial damage occurred during aortic cross-clamping and cardioplegic cardiac arrest. However, we found no increase in β-MHC within the first 20 min of reperfusion in the same patients, and the plasma concentrations of β-MHC clearly stayed within reference values in all patients during the study period. The time courses of myocardial myoglobin, CKMB mass, cTnT, and cTnI release were similar but not completely identical. In contrast to CKMB and myoglobin, which both showed a significant myocardial release as early as 1 min after aortic declamping, the release of troponins was significantly higher compared with baseline values starting at 5 min after aortic declamping, although troponins also started to be released from myocardium with aortic declamping. These data obtained during CABG are in accordance with our recent clinical observation of roughly equivalent early sensitives of cTnT, cTnI, myoglobin, and CKMB mass after myocardial infarction (8) and the clinical observation by Apple et al. (20) of similar early release kinetics of cTnI, cTnT, myoglobin, and CKMB in the systemic circulation in nine patients with acute myocardial infarction and complete reperfusion of the infarct-related coronary artery. However, the 30-min increase rates for CKMB and myoglobin in this study were significantly higher than for either cTnI or cTnT, which is explained very well by our findings. Also unique to our study is the quantification of the soluble troponins and β-MHC pools in fresh human myocardium, which revealed no significant amount of β-MHC in the cytoplasma but did reveal small cytosolic troponin pools in right atrial appendages that were removed for cannulation as a part of the routine surgical procedure. The amounts of soluble TnT and TnI (in percentage of total troponin content) were comparable. Although it can be expected that cTnT and cTnI are present in equimolar concentrations in myocardial tissue, we found marked differences when cytosolic pools were calculated as μg/g wet weight of myocardium. These differences cannot be explained by differences in troponin molecular masses. They are because of differences in troponin assay standardization; consequently the numeric results of different troponin assays may differ substantially. This study is the first to exclude the presence of a significant soluble β-MHC fraction in the cytoplasma of human myocardium. A possible limitation is that myocardial tissue from right atrial appendages may not be representative of ventricular myocardium; however, we could test only atrial appendages for obvious ethical reasons. Our results confirm earlier reports on the presence of significant soluble troponin pools in the cytoplasma of human ventricular myocardium obtained at autopsy (6)(18) . The percentages of soluble troponin pools in atrial tissue were in the range of the values found in these previous studies; in addition, the values of troponin and MHC pools were in agreement with the results from fresh left ventricular myocardium of a heart that was explanted but finally not used for transplantation (unpublished results). Our study obviously demonstrates the impact of the intracellular compartmentation of a molecule on the rate of its increases after myocardial damage. Myoglobin and CKMB are cytosolic or predominantly cytosolic molecules, which was confirmed by our results. They are released somewhat more rapidly than regulatory proteins after myocardial damage. cTnT and cTnI are released in parallel to each other. Both troponins are mainly structurally bound and have comparably large soluble pools in myocardium. The differences in myocardial release of cardiac troponins and MHC after cardioplegic cardiac arrest in association with aortic cross-clamping is striking and can be explained very well by their different intracellular compartmentations. It is very likely that the cytosolic pools of cTnT and cTnI account for the rapid early myocardial release in parallel to myoglobin and CKMB during reperfusion of the myocardium. Our study period was too short to detect β-MHC release from the human heart. Blood sampling was limited to 20 min after aortic declamping because the coronary sinus perfusion catheter had to be removed after this time point. We therefore cannot comment on the release of β-MHC into the coronary circulation thereafter. In previous studies, β-MHC also increased after CABG in uncomplicated CABG patients but not before the first day after surgery (1)(21) . Our study results in humans confirm previous experimental results (9)(10)(11) that suggested that the intracellular compartmentation of a molecule has a great impact on the rate at which it is released after myocardial damage. In isolated perfused rat hearts, cTnT increased in parallel to CK and LD after global ischemia, reoxygenation, or the calcium paradox (9)(10) . A simultaneous release of cTnT, cTnI, LD, and CK was also found in isolated perfused Langendorff rat hearts during reoxygenation after 60 min of hypoxia (11) . Compared with these experimental studies, the extent of myocardial damage in the patients we studied was small because several cardioprotective measures are routinely used to minimize myocardial damage during cardiac arrest in heart surgery, and no patient sustained a perioperative myocardial infarction. However, myocardial ischemia occurred during CABG, which was indicated by myocardial lactate production, although none of the 19 patients fulfilled the criteria for diagnosing perioperative myocardial infarction. There were moderately significant correlations between cumulative release of myoglobin, CKMB, and cardiac troponins and cumulative myocardial lactate production, which is an excellent measure for myocardial ischemia. In contrast to experimental studies, global myocardial ischemia might not be the only cause of myocardial protein release in our patients. The mechanical trauma to the heart during CABG is very small; however, right atriotomy was performed in all of our patients for cannulation of the great vessels. Despite the small myocardial mass of the right atrium, this may also contribute to a release of myocardial macromolecules and may have negatively impaired correlations among myocardial lactate production and cardiac marker release. Jürgen Bleier was supported by the “Hans und Blanca Moser Stiftung” (Vienna, Austria). References 1 Mair J, Puschendorf B, Michel G. Clinical significance of cardiac contractile proteins for the diagnosis of myocardial injury. Adv Clin Chem 1994 ; 31 : 63 -98. Crossref Search ADS PubMed 2 Mair P, Mair J, Seibt I, Wieser C, Furtwaengler W, Waldenberger F, et al. Cardiac troponin T: a new marker of myocardial tissue damage in bypass surgery. J Cardiothorac Vasc Anesth 1993 ; 7 : 674 -678. Crossref Search ADS PubMed 3 Graeser ML, Gergeley J. Purification and properties of the components from troponin. J Biol Chem 1973 ; 248 : 2125 -2133. PubMed 4 Wilkinson JM, Grand JRA. Comparison of amino acid sequence of troponin I from different striated muscles. Nature 1978 ; 271 : 31 -35. Crossref Search ADS PubMed 5 Harrington WF, Rodgers ME. Myosin. Annu Rev Biochem 1984 ; 53 : 35 -73. Crossref Search ADS PubMed 6 Katus HA, Remppis A, Scheffold T, Diederich KW, Kuebler W. Intracellular compartmentation of cardiac troponin T and its release kinetics in patients with reperfused and nonreperfused myocardial infarction. Am J Cardiol 1991 ; 67 : 1360 -1367. Crossref Search ADS PubMed 7 Mair J, Thome-Kromer B, Wagner I, Lechleitner P, Dienstl F, Puschendorf B, et al. Concentration time courses of troponin and myosin subunits after acute myocardial infarction. Coron Artery Dis 1994 ; 5 : 865 -872. PubMed 8 Mair J, Morandell D, Genser N, Lechleitner P, Dienstl F, Puschendorf B. Equivalent early sensitivities of myoglobin, creatine kinase MB mass, creatine kinase isoform ratios, and cardiac troponins I and T for acute myocardial infarction. Clin Chem 1995 ; 41 : 1266 -1272. Crossref Search ADS PubMed 9 Remppis A, Scheffold T, Greten J, Haass M, Greten T, Kübler W, et al. Intracellular compartmentation of troponin T: release kinetics after global ischemia and calcium paradox in the isolated perfused rat heart. J Mol Cell Cardiol 1995 ; 27 : 793 -803. Crossref Search ADS PubMed 10 Asayama J, Yamahara Y, Otha B, Miyazaki H, Tatsumi T, Matsumoto T, et al. Release kinetics of cardiac troponin T in coronary effluent from isolated rat hearts during hypoxia and reoxygenation. Basic Res Cardiol 1992 ; 87 : 428 -436. PubMed 11 Vorderwinkler KP, Mair J, Puschendorf B, Hempel A, Schlüter KD, Piper HM. Cardiac troponin I increases in parallel to cardiac troponin T, creatine kinase and lactate dehydrogenase in effluents from isolated perfused rat hearts after hypoxia-reoxygenation-induced myocardial injury. Clin Chim Acta 1996 ; 251 : 113 -117. Crossref Search ADS PubMed 12 Mair P, Mair J, Seibt I, Balogh D, Puschendorf B. Creatine kinase MB mass concentration in the diagnosis of perioperative myocardial infarction after coronary artery bypass grafting. Clin Chim Acta 1994 ; 224 : 203 -207. Crossref Search ADS PubMed 13 Katus HA, Looser S, Hallermayer K, Remppis A, Scheffold T, Borgya A, et al. Development and in vitro characterization of a new immunoassay for cardiac troponin T. Clin Chem 1992 ; 38 : 386 -393. Crossref Search ADS PubMed 14 Larue C, Calzolari C, Bertinchant JP, Leclercq F, Grolleau R, Pau B. Cardiac-specific immunoenzymometric assay of troponin I in the early phase of acute myocardial infarction. Clin Chem 1993 ; 39 : 972 -979. Crossref Search ADS PubMed 15 Brandt DR, Gates RC, Eng KE, Forsthy CM, Korom GK, Nitro AS, et al. Quantifying the isoenzyme of creatine kinase with the Abbott “IMx” immunoassay analyzer. Clin Chem 1990 ; 36 : 375 -378. Crossref Search ADS PubMed 16 Mair J, Artner-Dworzak E, Lechleitner P, Morass B, Smidt J, Wagner I, et al. Early diagnosis of acute myocardial infarction by a newly developed rapid immunoturbidimetric assay for myoglobin. Br Heart J 1992 ; 68 : 462 -468. Crossref Search ADS PubMed 17 Larue C, Calzolari C, Leger JJ, Leger J, Pau B. Immunoradiometric assay of myosin heavy chain fragments in plasma for investigation of myocardial infarction. Clin Chem 1992 ; 37 : 78 -82. 18 Mair J, Genser N, Morandell D, Maier J, Mair P, Lechleitner P, et al. Cardiac troponin I in the diagnosis of myocardial injury and infarction. Clin Chim Acta 1996 ; 245 : 19 -38. Crossref Search ADS PubMed 19 Matthews JNS, Altman DG, Cambell MJ, Royston P. Analysis of serial measurements in medical research. Br Med J 1990 ; 300 : 230 -235. Crossref Search ADS 20 Apple FS, Sharkey SW, Henry TD. Early serum cardiac troponin I and T concentrations after successful thrombolysis for acute myocardial infarction. Clin Chem 1995 ; 41 : 1197 -1198. Crossref Search ADS PubMed 21 Seguin JR, Saussine M, Ferriere M, Leger JJ, Leger J, Larue C, et al. Myosin: a highly sensitive indicator of myocardial necrosis after cardiac operations. J Thorac Cardiovasc Surg 1989 ; 98 : 397 -401. Crossref Search ADS PubMed © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Ricchiuti,, Vincent;Voss, Ellen, M;Ney,, Arthur;Odland,, Mark;Anderson, Page A, W;Apple, Fred, S

doi: 10.1093/clinchem/44.9.1919pmid: N/A

Abstract The purpose of this study was to determine whether the two monoclonal anti-cardiac troponin T (cTnT) antibodies (MAbs) used in the second generation cTnT assay by Boehringer Mannheim (BM, capture Ab, M11.7; detection Ab, M7) would detect cTnT isoforms expressed in human skeletal muscle in response to chronic renal disease (CRD). cTnT expression was examined in skeletal muscle biopsies obtained from 45 CRD patients, as well as nondiseased human heart (n = 3) and skeletal muscle (n = 3). cTnT proteins were resolved by modified 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and probed with the following anti-cTnT MAbs: M11.7; M7; JS-2, Lakeland Biomedical; and 13–11, Duke University. All four antibodies detected the cTnT isoforms (Ta, Te) expressed in human myocardium. In 20 of 45 skeletal muscle biopsies, MAb M11.7 recognized its epitope in one to three proteins, molecular mass 34–36 kDa, designated Te, Td, and Tc; the strongest signal was that of Te. The same proteins were recognized by MAbs JS-2 and 13–11. The BM M7 antibody did not detect the cTnT isoforms in the molecular mass range of 34–36 kDa. However, MAb M7 did detect a cTnT isoform, molecular mass 39 kDa, in 2 of 45 biopsies. This isoform had an electrophoretic mobility similar to the predominant heart cTnT isoform, Ta. We conclude that cTnT isoforms are expressed in the skeletal muscle of CRD patients. However, given the epitopes recognized by the BM MAbs M7 and M11.7 and the variable presence of these cTnT isoforms in skeletal muscle, the second generation BM cTnT assay will not detect these isoforms if they are released from skeletal muscle into the circulation. cTnT, cardiac troponin T, cTnI, cardiac troponin I, MAb, monoclonal antibody, CRD, chronic renal disease, BM, Boehringer Mannheim, TTBS, Tween-Tris-buffered saline. Cardiac troponin T (cTnT)4 and cardiac troponin I (cTnI) have unique amino-terminal sequences that differentiate them from their respective skeletal muscle isoforms; these differences have allowed for the development of isoform-specific monoclonal antibodies (MAbs) (1)(2)(3) . These antibodies have been used to examine the changes in cTnT and cTnI isoform expression that occur during muscle development. Four isoforms of cTnT are expressed in developing cardiac muscle through combinatorial alternative splicing of two 5′ exons in a developmentally regulated manner (4) . cTnT isoforms have also been described in fetal human skeletal muscle (5) , where there is a developmental down-regulation of cTnT and up-regulation of skeletal isoforms of TnT, which leads to the absence of cTnT in nondiseased adult skeletal muscle (5)(6) . In contrast to TnT, slow skeletal muscle TnI is the predominant isoform in developing human heart tissue (7)(8) . In cardiac tissue during the postnatal period, slow skeletal muscle TnI is gradually down-regulated in proportion to the up-regulation of cTnI. Nondiseased human cardiac muscle contains a single cTnI, which is not detected in healthy adult skeletal muscle. Furthermore, skeletal muscle does not express cTnI at any point during development. Several cTnT isoforms have been shown to be re-expressed during regeneration in adult rat skeletal muscle after injury or denervation (9) and in human skeletal muscle from patients with Duchenne muscular dystrophy (10) and polymyositis ((10)) and from chronic renal disease (CRD) patients (11) . These preliminary human studies described cTnT isoforms in skeletal muscle using immunochemical staining analysis (8) and Western blot analysis (10)(11) . A better characterization of cTnT isoform expression using more fully characterized antibodies was the goal of the current study. The mechanism for expression of cTnT isoforms in skeletal muscle from renal failure patients is likely associated with peripheral myopathy associated with renal disease (12) . Expression of cTnT isoforms in diseased or regenerating skeletal muscle appears to represent re-expression of the fetal gene, given that fetal skeletal muscle expresses cTnT (9) . To date, there has not been any evidence for a change in cTnI isoform expression even under adverse cardiac conditions such as hypertrophy and end-stage heart disease (13) or skeletal muscle disease associated with Duchenne muscular dystrophy (14) or CRD ((11)). The presence of measurable amounts of circulating cTnT and cTnI in blood is a specific indicator of heart muscle damage. This observation serves as the basis for using cTnT and cTnI measurement to diagnose acute myocardial infarction (15)(16) . However, studies have also established that cTnT and cTnI can be released during unstable angina (17)(18) . This observation led to the recognition of minor myocardial injury (19) . The incidence and prognostic value of increased cTnT and cTnI concentrations in chronic hemodialysis patients, independent of their history of coronary artery disease, has not been fully characterized. Numerous brief reports have described increases in cTnT in chronic hemodialysis patients without supportive clinical evidence of myocardial disease (11)(20)(21) . However, no mechanisms are available at the present to explain discordances between increased cTnT and cTnI in serum or plasma of CRD patients. The new Boehringer Mannheim (BM) second generation assay for cTnT has also demonstrated unexplained increases of serum cTnT in ∼10–20% of selected renal disease patients who have no clinically documented ischemic heart disease (11)(22) . These unexplained increases of cTnT in chronic hemodialysis patients have raised questions about the cardiac specificity of the BM cTnT immunoassay in detecting myocardial injury and the possibility of false-positive results stemming from renal disease-induced cTnT isoform expression in skeletal muscle (11) . To resolve whether the BM second generation cTnT immunoassay could yield false-positive results secondary to skeletal muscle expression of cTnT isoforms or fetal isoforms cross-reacting in the earlier studies, we have carefully characterized cTnT isoform expression in skeletal muscle from patients with CRD. We have assessed whether the two MAbs used in the second generation BM cTnT immunoassay would recognize these isoforms. To this end we probed the proteins expressed in the skeletal muscle of CRD patients with the capture and detection MAbs used in the second generation BM cTnT assay and two other cTnT-specific MAbs. Materials and Methods subjects Skeletal muscle biopsies were obtained from 45 CRD patients (21 women and 24 men; ages 29–81 years) on hemodialysis (range, 1 month to 10 years) or at the time of renal transplantation. Human heart and skeletal muscle specimens were obtained at autopsy within 24 h of death from three subjects who expired after non-cardiac-related illnesses. Informed consent was obtained from all subjects according to the Institutional Human Subjects Research Review Board guidelines. The biopsied tissues were frozen in liquid nitrogen and stored at −70 °C until analysis. protein extraction Samples (∼50 mg) of frozen nondiseased human heart muscle (n = 3), nondiseased human skeletal muscle (n = 3), and diseased skeletal muscle from patients with CRD (n = 45) were coarsely ground in a liquid nitrogen-cooled mortar and then added to 1 mL of ice-cold buffer (200 mmol/L potassium phosphate, pH 7.4, 5.0 mmol/L EGTA, 5.0 mmol/L β-mercaptoethanol, and 100 mL/L glycerol) to release both mitochondrial and cytoplasmic proteins (23) . The samples were homogenized at 4 °C. The procedures yielded >98% recovery of both cytosolic and myofibril proteins (24) . The supernatants were used immediately for protein analysis and Western blotting. Western blot analysis of the samples over the 2-month experimental period did not show any degradation of the cTnT or cTnI proteins (data not shown). Protein concentrations were determined using a modified Lowry method (25) with bovine serum albumin as a calibrator. antibodies Five different primary MAbs were selected for use in Western blotting on the basis of preliminary tests that characterized antibody specificity using purified human cTnI and cTnT proteins. A mouse MAb specific for cTnI (JS-1, residues recognized on cTnI protein sequence not available by manufacturer) was a gift from Lakeland Biomedical, Minneapolis, MN, and was used at 2 mg/L (10) . Four MAbs specific for cTnT were used (all at a 2 mg/L concentration): JS-2 (residues recognized on cTnT protein sequence not available by manufacturer) was a gift from Lakeland Biomedical (10) ; 13–11 (recognized residues 68–79 on cTnT protein sequence) from Duke University (6) ; M7 and M11.7 (recognized residues 125–131 and 136–147, respectively, on cTnT protein sequence), were provided by Boehringer Mannheim, GmbH (a gift from Dr. Klaus Hallermayer, Boehringer Mannheim GmbH, Germany) (22) . western blot analysis Protein extracts, 50 μg, were size-fractionated on sodium dodecyl sulfate-polyacrylamide gels using the method of Laemmli (26) with the following modifications: 30% acrylamide and 1.1% bis-acrylamide stock solutions were used in 7.5% running gels and 3.3% stacking gels (4) . Proteins were subsequently transferred to Hybond nitrocellulose membranes (27) (Amersham). Nonspecific binding sites were blocked by incubating the membranes in a blocking buffer containing 50 g/L nonfat dry milk in Tween-Tris-buffered saline (TTBS; 20 mmol/L Tris-HCl, pH 7.6, 137 mmol/L NaCl, 0.5 mL/L Tween-20) for 1 h. The primary antibody was diluted in antibody buffer (10 g/L nonfat dry milk in TTBS) and incubated with the membranes for 2 h on a rotating cylinder. The membranes were washed three times in changes of TTBS buffer for 30 min. Appropriate horseradish peroxidase-labeled secondary antibodies (sheep anti-mouse) were then incubated with the membranes for 1 h. The membranes were again washed three times in TTBS buffer before a 1-min incubation with ECLTM chemiluminescent substrate (Amersham). Light emission was detected by exposure to Fuji RX autoradiography film in the presence of Cronex intensifying screens (Fisher Scientific). Signal intensities within the linear range were quantitated using laser densitometry (Molecular Dynamics, Inc.). Linearity was established by analysis of a calibration curve generated with known amounts of total protein by Western blot (data not shown). Results m11.7 epitope recognition in cardiac and skeletal muscle A representative Western blot of nondiseased human heart muscle (lanes 1–3), nondiseased human skeletal muscle (lanes 4–6), and skeletal muscle from patients with CRD (lanes 7–13) probed with MAb M11.7 is illustrated in Fig. 1 A. One major cTnT isoform, molecular mass 39 kDa, and one minor cTnT isoform, molecular mass 34 kDa, were recognized in nondiseased myocardium. As anticipated, M11.7 did not recognize its epitope in nondiseased skeletal muscle. However, one to three proteins, with approximate molecular masses of 34, 35, and 36 kDa, were recognized by M11.7 in skeletal muscle from 20 of 45 patients with CRD. No other proteins containing the cardiac-specific epitope of M11.7, including the major cTnT isoform band, molecular mass 39 kDa, found in nondiseased heart samples, were identified in these 45 preparations. Figure 1. Open in new tabDownload slide Western immunoblots of nondiseased human heart muscle (NHHM; lanes 1–3), nondiseased human skeletal muscle (NHSM; lanes 4–6), and skeletal muscle from CRD patients (lanes 7–13) probed with cardiac-specific TnT MAbs M11.7 (A) and M7 (B). Both MAbs were supplied by BM. Positions of the molecular mass standards (MW) are shown on the left. Figure 1. Open in new tabDownload slide Western immunoblots of nondiseased human heart muscle (NHHM; lanes 1–3), nondiseased human skeletal muscle (NHSM; lanes 4–6), and skeletal muscle from CRD patients (lanes 7–13) probed with cardiac-specific TnT MAbs M11.7 (A) and M7 (B). Both MAbs were supplied by BM. Positions of the molecular mass standards (MW) are shown on the left. m7 epitope recognition in cardiac and skeletal muscle A representative Western blot of human heart, nondiseased skeletal muscle, and muscle biopsies from patients with CRD probed with M7 is illustrated in Fig. 1B . Similar to M11.7, two cTnT isoforms were recognized by M7 in nondiseased myocardium. The recognition of these two cTnT isoforms seemed to differ between M11.7 and M7. The signal intensity for the cTnT isoform, molecular mass 34 kDa, was less with M7. Similar to M11.7, M7 did not recognize its epitope in nondiseased skeletal muscle. However, in 2 of 45 biopsies obtained from patients with CRD, M7 recognized a protein with an approximate molecular mass of 39 kDa that comigrates with the major cTnT isoform expressed in the heart. Comparison of four anti-cTnT MAbs Western blot analysis demonstrates important differences in protein recognition among the four cTnT MAbs (Fig. 2). MAb 11.7, 13–11, and JS-2 recognized the identical pattern of cTnT isoforms in myocardium and in skeletal muscle from patients with CRD. All three MAbs recognized the cTnT isoforms with the lower molecular mass of ∼34–36 kDa expressed in some of the skeletal muscle biopsies from renal patients (20 of 45); however, they did not recognize the large cTnT isoform, molecular mass 39 kDa, detected by M7. None of the MAbs recognized their cardiac-specific epitopes in nondiseased human skeletal muscle. Western blots of TnT isolated from fast and slow skeletal muscle (using high loads, 1 μg per lane) probed with M11.7 (Fig. 3 A) and MAbs 13–11 and JS-2 (data not shown) demonstrated a minor cross-reactivity with a 33-kDa protein. In contrast, M7 did not recognize its epitope in these fast and slow skeletal muscle preparations (Fig. 3B). The Western blot illustrated in Fig. 3 further demonstrates the absence of the epitopes of M11.7 and M7 in nondiseased skeletal muscle. Figure 3. Open in new tabDownload slide Western immunoblots of nondiseased human heart muscle (lane 1), purified cTnT (lane 2), nondiseased human skeletal muscle (lane 3), purified human slow (lane 4) and fast (lane 5) twitch skeletal TnT protein, and skeletal muscle from CRD patients (lanes 6and 7) probed with cardiac-specific TnT MAbs M11.7 (A) and M7(B). Positions of the molecular mass standards (MW) are shown on the left. Figure 3. Open in new tabDownload slide Western immunoblots of nondiseased human heart muscle (lane 1), purified cTnT (lane 2), nondiseased human skeletal muscle (lane 3), purified human slow (lane 4) and fast (lane 5) twitch skeletal TnT protein, and skeletal muscle from CRD patients (lanes 6and 7) probed with cardiac-specific TnT MAbs M11.7 (A) and M7(B). Positions of the molecular mass standards (MW) are shown on the left. Figure 2. Open in new tabDownload slide Western immunoblots of nondiseased human heart muscle (lane 1), nondiseased human skeletal muscle (lanes 2and 3), an equal mixture of human heart and skeletal muscle from CRD patients (lanes 4 and 6), and skeletal muscle from CRD patients (lanes 5 and 7) probed with cardiac-specific TnT MAbs: (A) M11.7 (BM); (B) M7 (BM); (C) 13–11 (Duke University); and (D) JS-2 (Lakeland Biomedical). Positions of the molecular mass standards (MW) are shown on the left. Figure 2. Open in new tabDownload slide Western immunoblots of nondiseased human heart muscle (lane 1), nondiseased human skeletal muscle (lanes 2and 3), an equal mixture of human heart and skeletal muscle from CRD patients (lanes 4 and 6), and skeletal muscle from CRD patients (lanes 5 and 7) probed with cardiac-specific TnT MAbs: (A) M11.7 (BM); (B) M7 (BM); (C) 13–11 (Duke University); and (D) JS-2 (Lakeland Biomedical). Positions of the molecular mass standards (MW) are shown on the left. cTnI expression in cardiac and skeletal muscle A single cTnI isoform (25 kDa) was detected by MAb JS-1 in nondiseased adult myocardium (data not shown). MAb JS-1 did not recognize its cTnI-specific epitope in any of the skeletal muscle preparations, including those from patients with CRD (data not shown). Discussion The findings of this study make two important contributions to the fields of human cTnT isoform expression and skeletal muscle response to CRD. First, we demonstrate cTnT isoform expression in adult human skeletal muscle obtained from patients with CRD, using three well-characterized anti-cTnT MAbs (M11.7 and M7 (22) and 13–11 (4) ), confirming an earlier report from our laboratory (11) . The expression of cTnT isoforms in skeletal muscle has been described previously in differentiating C2C12 myoblasts, a mouse skeletal muscle cell line (28) , regenerating rat muscle fibers after cold injury (9) , mature rat muscle fibers after denervation (9) , and diseased human skeletal muscle from Duchenne muscular dystrophy patients (10) . Our findings contrast with the recent report of Haller et al. (29) , which showed that no evidence of cTnT expression at the mRNA or protein level was demonstrated in truncal skeletal muscle biopsies from five patients with end-stage renal disease. Because the M7 and M11.7 antibodies were not used in their tissue studies, it is difficult to correlate with findings in serum measured by the second generation BM cTnT assay. Characterization of cTnT isoforms in total RNA from our 45 biopsies is currently in progress. Second, because of the differential detection by the two BM MAbs M11.7 and M7 of the cTnT isoforms expressed in skeletal muscle in the presence of CRD, the second generation serum cTnT assay by BM will detect only cTnT isoforms expressed in the adult human heart. Our detection of a major cardiac isoform of TnT (Ta) detected by both MAb M11.7 and MAb M7 confirmed the observations of Muller-Bardorff et al. (22) . The 39-kDa isoform detected only by M7 (band Ta′) may correspond to an isoform lacking the M11.7 epitope (residues 136–147) present by definition in the adult cTnT isoform. If the band Ta′ was adult cardiac isoform it would be positive for both M11.7 and M7. That isoform Ta′ co-migrates with adult isoform Ta is not unsurprising because small molecular weight differences will not be resolved by the gels. The two to three isoforms recognized by M11.7 would not be detected by M7, whereas the 39-kDa isoform will not be captured by M11.7. Clinically therefore, on the basis of our study, increased concentrations of circulating cTnT in serum or plasma of CRD patients cannot be considered false-positive results. Ischemic heart disease continues to be the major cause of death in renal dialysis patients. Approximately 40% of overall mortality in chronic dialysis patients is caused by ischemic heart disease, with ∼25% of these ischemic deaths attributed to acute myocardial infarction (30) . The risk of cardiac death is higher in older, diabetic patients. Approximately 250 000 patients were treated for end-stage renal disease in the United States in 1993. Despite the high incidence of cardiac disease in dialysis patients, there are no data on outcomes of dialysis patients with acute myocardial infarction. No substantial studies have focused on chronic dialysis patients for identifying biochemical markers that would provide useful prognostic information and permit the early identification of patients with increased risk of death. Therefore, it is likely that examining serum cTnT concentrations in CDR patients will provide new and helpful information to risk stratify this patient population (31) , as shown for unstable angina patients (17)(18) . Large-scale outcome studies using both cTnT and cTnI may prove useful in the care of these patients. The re-expression of multiple isoforms of cTnT in diseased human skeletal muscle parallels, probably, results from the expression of these isoforms in differentiating myotubes (28) and is consistent with the expression of developmentally expressed fetal isoforms, as previously described for both cTnT (5)(32) and creatine kinase isoenzymes (33) . In human fetal skeletal muscle, fetal cTnT isoforms are transiently expressed. Lack of cTnI expression in skeletal muscle agrees with previous studies (11)(14) . The existence of multiple muscle-specific isoforms is common among contractile proteins. Different genes encode for the different TnT isoforms in different striated muscles. Further diversity of these isoforms arises from combinatorial alternative splicing of the primary transcripts of the three striated muscle TnT genes (1)(2)(3)(4)(6) . In humans, four cTnT isoforms have been described at the product level in fetal cardiac muscle (5) . Fig. 4 represents a schematic of our findings that describe cTnT isoform expression in nondiseased human heart and skeletal muscle and skeletal muscle obtained from CRD patients, based on the proteins recognized by the two MAbs used in the second generation BM cTnT immunoassay. Both the capture MAb M11.7 and the detection MAb M7 detect several cTnT isoforms in heart, with cTnTa being the predominantly expressed isoform. No cTnT isoforms were recognized in nondiseased skeletal muscle. In the skeletal muscle of renal disease patients, the capture M11.7 MAb detected one to three minor cTnT isoforms, at 34, 35, and 36 kDa; however, it never detected the major cTnT isoform expressed at 39 kDa found in heart muscle. Conversely, the detection MAb M7 rarely (2 of 45 samples) detected a cTnT isoform band at 39 kDa (Ta′) but never detected any cTnT isoforms at lower molecular weights. Figure 4. Open in new tabDownload slide Schematic representation of Western immunoblot findings for nondiseased human heart muscle (NHHM), nondiseased human skeletal muscle (NHSM), and skeletal muscle from CRD patients probed with the two cardiac-specific TnT MAbs from BM (A) M11.7 (capture Ab) and (B) M7 (detection Ab). cTnT isoforms are designated a, b, c, d, and e based on decreasing molecular weights; 39, 38, 36, 35, and 34 kDa, respectively. a′ represents the isoform with an electrophoretic mobility similar to the predominant heart cTnT isoform Ta. Positions of the molecular mass standards (MW) are shown on the left. Figure 4. Open in new tabDownload slide Schematic representation of Western immunoblot findings for nondiseased human heart muscle (NHHM), nondiseased human skeletal muscle (NHSM), and skeletal muscle from CRD patients probed with the two cardiac-specific TnT MAbs from BM (A) M11.7 (capture Ab) and (B) M7 (detection Ab). cTnT isoforms are designated a, b, c, d, and e based on decreasing molecular weights; 39, 38, 36, 35, and 34 kDa, respectively. a′ represents the isoform with an electrophoretic mobility similar to the predominant heart cTnT isoform Ta. Positions of the molecular mass standards (MW) are shown on the left. The differential recognition of the BM antibodies of cTnT isoforms expressed in skeletal muscle from patients with CRD can be explained, at least in part, by combinatorial splicing of the cTnT primary transcript, causing the loss of the M7 and M11.7 epitopes. Although the commonly expressed cTnT isoforms in the heart are the result of splicing of two 5′ exons, rare combinatorial alternative splicing of exons encoding the central region of cTnT has been described in cDNA and PCR amplicons (4) . MAb 13–11, whose epitope is made up of residue 68–79, would be expected to recognize all of the cTnT isoforms in that the exon that encodes for its epitope does not undergo splicing. However, the epitopes of M7, residues 125–131, and M11.7, residues 136–147, appear to be encoded by two of three adjoining exons that undergo combinatorial alternative splicing (4) . The exclusion of two exons, including the one that encodes for the epitope of M7, would lead to a loss of 39 residues and, therefore, the smaller cTnT isoforms recognized by M11.7 in skeletal muscle, and the failure of M7 to recognize them. Given the relatively larger size of the cTnT isoform expressed in skeletal muscle by M7 and not by M11.7, the exon encoding the sequence containing the M11.7 epitope must be shorter than proposed (4) to yield a cTnT isoform that co-migrates with the larger cTnT isoform expressed in the adult human heart. In conclusion, taking into account the structure of the epitopes of the BM antibodies and the presence of these epitopes in the cTnT isoforms expressed in the patients with CRD, if these cTnT isoforms are released from skeletal muscle into the circulation, they would not be measured by the BM second generation cTnT assay. Therefore, we conclude that in patients with CRD an increased serum cTnT concentration, as measured by the second generation BM cTnT immunoassay, originates from the heart and is not a false positive that results from skeletal muscle expression of cTnT. This work was supported in part by NIH HL20749 and HL42250 (PAWA). References 1 Potter JD, Gergeley J. Troponin, tropomyosin, and actin interactions in the calcium regulation of muscle contraction. Biochemistry 1974 ; 13 : 2697 -2703. Crossref Search ADS PubMed 2 Dhoot GK, Grearson N, Perry SV. Polymorphic forms of troponin T and troponin C and their localization in striated muscle cell types. Exp Cell Res 1979 ; 122 : 339 -350. Crossref Search ADS PubMed 3 Wilkinson JM, Grand JA. Comparison of amino acid sequence of troponin I from different striated muscles. Nature 1978 ; 271 : 31 -35. Crossref Search ADS PubMed 4 Townsend PJ, Barton PJR, Yacoub HH, Farga H. Molecular cloning of human cardiac troponin isoforms: expression in developing and failing hearts. J Mol Cell Cardiol 1995 ; 27 : 2223 -2236. Crossref Search ADS PubMed 5 Anderson PAW, Malouf NN, Oakley AE, Pagani ED, Allen PD. Troponin T isoform expression in humans: a comparison among normal and failing adult heart, fetal heart, and adult and fetal skeletal muscle. Circ Res 1991 ; 69 : 1226 -1233. Crossref Search ADS PubMed 6 Anderson PAW, Oakley AE. Immunological identification of five troponin T isoforms reveals an elaborate maturational troponin T profile in rabbit myocardium. Circ Res 1989 ; 65 : 1087 -1093. Crossref Search ADS PubMed 7 Saggin L, Gorza L, Ausoni S, Schiaffino S. Troponin switching in the developing heart. J Biol Chem 1989 ; 264 : 16299 -16302. PubMed 8 Hunkeler NM, Kullman J, Murphy AM. Troponin I isoform expression in human heart. Circ Res 1991 ; 69 : 1409 -1414. Crossref Search ADS PubMed 9 Saggin L, Gorza L, Ausoni S, Schiaffino S. Cardiac troponin T in developing, regenerating and denervated rat skeletal muscle. Development 1990 ; 110 : 547 -554. PubMed 10 Bodor GS, Survant L, Voss EM, Smith S, Porterfield D, Apple FS. Cardiac troponin T composition in normal and regenerating skeletal muscle. Clin Chem 1997 ; 43 : 476 -484. Crossref Search ADS PubMed 11 McLaurin MD, Apple FS, Voss EM, Herzog CA, Sharkey SW. Cardiac troponin I, cardiac troponin T, and creatine kinase MB in dialysis patients without ischemic heart disease: evidence of cardiac troponin T expression in skeletal muscle. Clin Chem 1997 ; 43 : 976 -982. Crossref Search ADS PubMed 12 Diesel W, Emms M, Knight B, Noakes TD, Swanepoel CR, Smit R. Morphology features on the myopathy associated with chronic renal failure. Am J Kidney Dis 1993 ; 22 : 677 -684. Crossref Search ADS PubMed 13 Sasse S, Brand NJ, Kyprianou P, Dhoot GK, Wade R, Arai M, et al. Troponin I gene expression during human cardiac development and in end-stage heart failure. Circ Res 1993 ; 72 : 932 -938. Crossref Search ADS PubMed 14 Bodor GS, Porterfield D, Voss EM, Smith S, Apple FS. Cardiac troponin I is not expressed in fetal and healthy or diseased adult human skeletal tissue. Clin Chem 1995 ; 41 : 1710 -1715. Crossref Search ADS PubMed 15 Katus HA, Scheffold T, Remppis A, Zehelein J. Proteins of the troponin complex. Lab Med 1992 ; 23 : 311 -317. Crossref Search ADS 16 Adams J, Bodor G, Davila-Roman V, Delmez JA, Apple FS, Ladenson JH, Jaffe AS. Cardiac troponin I: a marker with high specificity for cardiac injury. Circulation 1993 ; 88 : 101 -106. Crossref Search ADS PubMed 17 Ohman EM, Armstrong PW, Christenson RH, Granger CB, Katus HA, Hamm CW, et al. Cardiac troponin T levels for risk stratification in acute myocardial ischemia. N Engl J Med 1996 ; 335 : 1333 -1341. Crossref Search ADS PubMed 18 Galvani M, Ottani F, Ferrini D, Ladenson JH, Destro A, Baccos D, et al. Prognostic influence of elevated values of cardiac troponin I in patients with unstable angina. Circulation 1997 ; 95 : 2053 -2059. Crossref Search ADS PubMed 19 Gerhardt W, Katus H, Ravkilde J. S-Troponin T in suspected myocardial injury compared with mass and catalytic concentrations of S-creatine kinase isoenzyme MB. Clin Chem 1991 ; 37 : 1405 -1411. Crossref Search ADS PubMed 20 Collinson PO, Stubbs PJ, Rosalki SB. Cardiac troponin T in renal disease. Clin Chem 1995 ; 41 : 1671 -1673. Crossref Search ADS PubMed 21 Bhayana V, Gougoulias T, Cohee S. Discordance between results for serum troponin T and I in renal disease. Clin Chem 1995 ; 41 : 312 -317. Crossref Search ADS PubMed 22 Muller-Bardorff M, Hallermayer K, Schroder A, Ebert C, Borgya A, Gerhardt W, et al. Improved troponin T ELISA specific for cardiac troponin T isoform: assay development and analytical and clinical validation. Clin Chem 1997 ; 43 : 458 -466. Crossref Search ADS PubMed 23 Smeitink J, Wevers R, Hulshof Ruitenbeek W, Lith TV, Sengers R, Trjbels F. A method for quantitative measurement of mitochondrial creatine kinase in human skeletal muscle. Ann Clin Biochem 1992 ; 29 : 196 -201. Crossref Search ADS PubMed 24 Voss EM, Sharkey SW, Gernert A, Murakami MA, Johnson RB, Hsieh CC, Apple FS. Human and canine cardiac troponin T and CK MB distribution in normal and diseased myocardium. Arch Path Lab Med 1995 ; 119 : 799 -806. PubMed 25 Lowry OH, Rosebrough NJ, Farr AL, Rangall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 1951 ; 193 : 265 -275. PubMed 26 Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970 ; 227 : 680 -685. Crossref Search ADS PubMed 27 Towbin H, Staehelin T, Gordin J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedures and some applications. Proc Natl Acad Sci U S A 1979 ; 76 : 4350 -4354. Crossref Search ADS PubMed 28 McMahon DK, Anderson PAW, Wassar R, Gunting JB, Saba Z, Oakley AE, Malouf NN. C2 C12 cells: biophysical, biochemical, and immunocytochemical properties. Am J Physiol 1994 ; 266 : C1795 -C1802. Crossref Search ADS PubMed 29 Haller C, Zehelein J, Remppis A, Muller-Bardorff M, Katus HA. Cardiac troponin T in patients with end-stage renal disease: absence of expression in truncal skeletal muscle. Clin Chem 1998 ; 44 : 930 -938. Crossref Search ADS PubMed 30 Lindner A, Charra B, Sheppard DJ. Accelerated atherosclerosis in prolonged maintenance hemodialysis. N Engl J Med 1994 ; 290 : 697 -701. 31 Apple FS, Sharkey SW, Hoeft P, Skeate R, Voss EM, Dahlmeier BA, Preese L. Prognostic value of serum cardiac troponin I and T in chronic dialysis patients: a one year outcomes analysis. Am J Kidney Dis 1997 ; 29 : 399 -403. Crossref Search ADS PubMed 32 Anderson PAW, Greig A, Mark TA, Malouf NN, Oakley AE, Ungerleider RM, et al. Molecular basis of human cardiac troponin T isoforms expressed in developing, adult and failing heart. Circ Res 1995 ; 76 : 681 -686. Crossref Search ADS PubMed 33 Apple FS, Billadello JJ. Expression of creatine kinase M and B mRNAs in treadmill trained rat skeletal muscle. Life Sci 1994 ; 55 : 585 -592. Crossref Search ADS PubMed © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Heeschen,, Christopher;Goldmann, Britta, U;Moeller, Robert, H;Hamm, Christian, W

doi: 10.1093/clinchem/44.9.1925pmid: N/A

Abstract Detection of cardiac troponin I (cTnI) in patients suspected of having an acute coronary syndrome is highly predictive for an adverse outcome. We evaluated a bedside test for cTnI that uses a polyclonal capture antibody and two monoclonal indicator antibodies. Clinical studies were performed in patients with acute coronary syndrome and patients with chest pain but no evidence of acute myocardial injury. The whole-blood, 15-minute assay had a concordance of 98.9% with an ELISA for cTnI and a detection limit of 0.14 μg/L, and the device tolerated temperatures between 4 °C and 37 °C. Diagnostic sensitivity for myocardial infarction at arrival (3.5 ± 2.7 h after onset of symptoms) was 60% [creatine kinase isoenzyme MB (CK-MB) mass, 48%; CK activity, 36%; P < 0.01], and 4 h later, diagnostic sensitivity was 98% (CK-MB mass, 91%; CK activity, 61%; P < 0.01). In 38% of the patients with unstable angina, at least one positive cTnI test was found (CK-MB mass, 4%; CK activity, 2%). No false-positive test results were found in renal failure or injury of skeletal muscle. We conclude that the diagnostic efficacy of the cTnI rapid test was comparable with the cTnI ELISA and superior to CK-MB determination. Therefore, this device could facilitate decision-making in patients with chest pain at the point of care. Acute myocardial infarction is highly probable in patients with typical anginal pain accompanied by marked ST-segment elevation. In these patients, thrombolytic therapy is usually administered independently of confirmation by biochemical markers (1) . However, in up to 50% of the patients with acute myocardial infarction, the electrocardiogram is nondiagnostic, and the diagnosis depends on biochemical markers (2)(3)(4)(5) . About 4% of patients with developing myocardial infarction are inappropriately sent home (4)(5)(6)(7) , and electrocardiographic changes in high-risk patients with unstable angina are even less specific (8) . Measurements of creatine kinase isoenzyme MB (CK-MB) activity are rapid and inexpensive but suffer from a low cardiac specificity and poor sensitivity to detect minor myocardial injury (2)(4)(9) . CK-MB measurements, even when determined immunologically, appear to be inadequate for clinical decision-making in the emergency room (9)(10)(11)(12) . Cardiac troponins have recently been introduced as cardiac markers (12)(13)(14)(15)(16)(17)(18)(19) . Only cardiac troponin I (cTnI) has not yet been detected in skeletal muscle cells at any stage of development or pathological condition (15)(18)(19) . cTnI can be distinguished by monoclonal antibodies recognizing the amino acid sequence distinct for cardiac muscle cells (17)(18)(19)(20) . cTnI determination promises higher diagnostic efficacy and potency for new diagnostic options (12)(13)(14) : Wider diagnostic window of time with early appearance (attributed to release from cytosolic pool) and prolonged appearance (attributed to proteolysis of the contractile apparatus). High proportional rise of troponins (reflecting the low plasma cTnI concentrations in healthy persons), allowing detection of minor myocardial damage. No cross-reactivity of skeletal isoforms. Lack of cTnI in skeletal muscle tissue. A recently developed rapid assay for cTnI can be performed at the point of care within 15 min by medical or paramedical staff and might facilitate risk stratification of patients in emergency rooms (12)(21) . Therefore, we compared its analytical performance with two established ELISA systems. Materials and Methods rapid assay The rapid assay system uses chromatographic solid-phase technology (Spectral Diagnostics) with a cocktail of two gold-labeled monoclonal mouse indicator antibodies (8I-7 and 2I-14) and a biotinylated polyclonal goat capture antibody. The test system contains at least 0.3 μg of each antibody (17)(21) . The selected antibodies bind both free and complexed cTnI with high affinity (17)(22) . We added 200 μL of heparinized whole blood or centrifuged plasma to the device. After separation of cellular blood components from the plasma fraction by a glass fiber fleece, the migrating plasma dissolves a buffer and solubilizes the adsorbed antibodies. The antibodies and the patient’s cTnI molecule form sandwich complexes, which migrate to the signal zone and accumulate in the reading window by means of biotin-streptavidin interaction. Positive results (cTnI ≥0.1 μg/L; manufacturer-defined discriminator) are indicated by a color line that appears within 15 min. The unreacted indicator antibodies pass on and bind to the control line consisting of solid phase anti-mouse IgG antibodies (≥0.2 μg). The appearance of this control line distal to the signal line confirms impeccable test function, including unimpeded plasma flow. quantitative cardiac marker determination The two identical indicator antibodies, but as Fab fragments, were used on a semiautomated one-step cTnI ELISA system on the basis of “big bead” technology performed on the Cobas EIA® (Roche Diagnostic Systems) (17)(23)(24) . After incubation of the serum samples with 100 μL of conjugate and the coated bead, the exceeding conjugate was removed by radial elution. The second incubation period was performed with 250 μL of solution of 3,3′,5,5′-tetramethyl-benzidin and hydrogen peroxide converted into a colored product by horseradish peroxidase-labeled antibodies. The reaction was terminated by adding 1000 μL of stopping solution (50 mL/L sulfuric acid) to allow cTnI quantification photometrically. The lower detection limit of this assay system is 0.03 μg/L (17) , and a discriminator value of 0.1 μg/L was used. For calibration, six calibrators containing purified recombinant cTnI (Spectral Diagnostics), ranging from 0 to 6.0 μg/L, were used. Interassay precision was 6.2% at 0.2 μg/L and 5.5% at 2.4 μg/L. In parallel, cTnI and CK-MB mass were quantified by means of a commercially available analyzer system (Access® analyzer, Beckmann Diagnostics). It is based on chemiluminescence technology, and the capture antibody is fixed to magnetic microparticles. The turnaround time is 18 min. The detection limit for cTnI is 0.03 μg/L. Values ≥0.1 μg/L were considered to be positive (25) , and day-to-day precision (CV) was 8.1% at 0.2 μg/L and 4.6% at 2.4 μg/L. The detection limit for CK-MB mass is 0.15 μg/L, and the upper reference is 5.0 μg/L (26) . The interassay CVs were 8.4% at 8.2 μg/L and 7.2% at 14.7 μg/L. Catalytic activity of total CK was measured colorimetrically in the emergency laboratory at room temperature (Hitachi 717, Boehringer Mannheim) and with different cutoffs for women (70 U/L) and men (80 U/L) according to the manufacturer’s specifications. All biochemical analyses were performed by technicians unaware of the patients’ histories. evaluation of test performance Blood samples were collected from 10 patients with acute myocardial infarction and 5 patients with minor myocardial injury (high frequency count ablation, unstable angina, and myocarditis). After quantitative determination of cTnI two times by Cobas EIA, the samples were diluted stepwise with human serum from healthy volunteers (cTnI concentration ≤0.03 μg/L) down to a cTnI concentration below the analytical sensitivity of the Cobas EIA system. Testing of each dilution step was performed five times separately by blinded investigators (n = 6) who were trained in the assessment of the rapid test. The cutoff for the rapid assay was calculated as the cTnI concentration determined with the Cobas EIA (identical antibodies) detected in all trials by each of the blinded observers. To evaluate the effects of temperature, rapid tests were duplicated on devices heated to 37 °C, cooled to 4 °C, and frozen to −20 °C for 7 days. clinical evaluation The study population was recruited from the emergency room of the University Hospital Hamburg between September 1, 1995 and October 31, 1997. All of these patients had chest pain lasting 12 h or less and were stratified into four groups of consecutive patients: group 1, 159 patients with acute myocardial infarction, which was defined in the absence of ST-segment elevations as an increase of total CK activity within 24 h after onset of chest pain higher than twice the upper limit of normal associated with increased CK-MB; group 2, 321 patients with angina at rest associated with ST-T-segment aberrations. Patients had unstable angina according to the Braunwald classification III B (53%) + C (47%) (27) ; group 3, 37 patients with cardiac chest pain of nonischemic origin related to acute congestive heart failure, perimyocarditis, ventricular arrhythmia, or rejection after heart transplantation who had no marked coronary stenosis (≥70% diameter stenosis); and group 4, 120 patients with acute chest pain but no evidence of acute myocardial injury. These patients had no ST-segment changes and underwent treadmill testing, stress echocardiography, or angiogram, providing no abnormalities. During 30 days of follow-up, no cardiac event was documented in this group. Furthermore, we investigated 27 asymptomatic patients (ages, 34 ± 12 years) with end-stage renal failure but no history of coronary heart disease recruited from the department of nephrology at the time of periodic hemodialysis. statistical Analysis All results for continuous variables are expressed as means ± SDs (range), and for comparison between two groups, the Mann–Whitney U-test was used. Comparison of categorical variables (positive/negative) were generated by the McNemar test (two-sided) with appropriate degrees of freedom. The calculations were done with SPSS 6.1 (SPSS, Inc.), and P values <0.05 were considered statistically significant. Analytical results were considered concordant if the rapid assay was interpreted as positive and the quantitative determination by ELISA was above cutoff or if rapid assay was interpreted as negative and the ELISA result was below cutoff. Otherwise, the result was categorized as discordant (false-negative/false-positive). The diagnostic sensitivity was calculated as the number of positive test results among all patients with acute myocardial infarction observed, and the diagnostic specificity was defined as the number of negative test results among all patients with excluded acute myocardial infarction. Results detection limit of the rapid device Under blinded conditions, 1198 blood samples were analyzed with the rapid device by two trained observers and with both ELISA systems, the Cobas EIA and the Access analyzer. The concordance of the rapid device (29% positive test results) with the Cobas EIA (26% positive results) was 98.9% (Fig. 1), and with the Access analyzer (23% positive results) was 97.2% at predefined cutoff values of 0.1 μg/L. Figure 1. Open in new tabDownload slide Concordance between the quantitative troponin I determination (Cobas EIA) and the rapid test result in 1188 blood samples. Troponin I concentrations are listed according to a positive or negative rapid test result for each sample. Figure 1. Open in new tabDownload slide Concordance between the quantitative troponin I determination (Cobas EIA) and the rapid test result in 1188 blood samples. Troponin I concentrations are listed according to a positive or negative rapid test result for each sample. All cTnI values ≥0.14 μg/L (Cobas EIA; n = 302) were categorized as positive rapid test results. For cTnI values between 0.06 and 0.14 μg/L, an increasing rate of positive test results was found. Only two samples with cTnI values <0.06 μg/L showed a positive rapid test result (0.03 and 0.04 μg/L) and have to be considered as false-positive. One test kit indicated malfunction by absence of a control signal, despite obviously adequate plasma migration. Intensity and speed of color development were apparently proportional to the cTnI concentration. In samples with cTnI amounts ≥4 μg/L, a positive test signal developed within 5 min, and samples with cTnI concentrations ≤0.4 μg/L could not be rated earlier than 15 min after serum application (Fig. 2). To prevent false-positive test results, the time interval from blood application to test assessment should not exceed 15 min. Figure 2. Open in new tabDownload slide Rapid device: intensity and speed of signal line development according to the troponin I concentration in the blood sample (Cobas EIA). Figure 2. Open in new tabDownload slide Rapid device: intensity and speed of signal line development according to the troponin I concentration in the blood sample (Cobas EIA). interobserver variability The rapid test results of a stepwise dilution series revealed from patients with myocardial injury were categorized by six blinded observers who were trained but varied in experience (50–1000 test assessments). In the range from 0.06 to 0.14 μg/L, significant, (P = 0.02) interobserver variability was observed (Fig. 3). At a cTnI concentration of 0.1 μg/L, four observers rated the test as positive and two as negative. The discordance correlated with the observer’s experience in assessment of rapid assays. Experienced observers rated all tests at cTnI concentrations of 0.08 μg/L as positive, whereas trained but unexperienced observers rated all of these tests as negative (P < 0.01). Additionally, insufficient artificial illumination and hemolytic or lipemic samples increased the incidence of discrepant classifications. All test results with cTnI concentrations above 0.14 μg/L were considered positive by all observers, independent of experience. Figure 3. Open in new tabDownload slide Stepwise dilution series in five patients with acute myocardial infarction. Quantitative troponin I determination (Cobas E1A) and rapid test result are plotted for each patient. (▴), positive rapid test result; (○), negative rapid test result. Figure 3. Open in new tabDownload slide Stepwise dilution series in five patients with acute myocardial infarction. Quantitative troponin I determination (Cobas E1A) and rapid test result are plotted for each patient. (▴), positive rapid test result; (○), negative rapid test result. temperature stability Incubation of 15 test devices at 37 °C for 7 days did not influence the analytical sensitivity of the test system (0.10 vs 0.11 μg/L for incubated devices). Similarly, lowering the temperature of 20 tests devices down to 4 °C was not associated with a decrease of analytical sensitivity (0.10 vs 0.10 μg/L for cooled devices). However, the freezing of 20 test devices at −20 °C for 7 days did affect the sensitivity (0.10 vs 0.45 μg/L for frozen devices; P < 0.01). clinical evaluation Patients with acute myocardial infarction (n = 159) In 159 patients with acute myocardial infarction, the first determination of the cardiac markers was performed 3.5 ± 2.7 (0.5–12) h after onset of symptoms. A positive cTnI bedside test was found in 60% of the patients, with increased cTnI concentrations with the Cobas EIA in 58% and with the Access analyzer in 54% of the patients (not significant for both). Four hours later, at least one positive bedside test was obtained in 98%. Increased quantitative cTnI concentrations with the Cobas EIA were found in 94% and in 92% of the patients with the Access analyzer (not significant for both). Immunoassay determination of CK-MB revealed increased values at arrival in 48% (P = 0.004 compared with cTnI rapid assay) and for conventional CK activity in 36% of the patients (P <0.001). Four hours later, in 91% and 61% of the patients, CK-MB mass (P = 0.03) and CK enzyme activity (P < 0.001), respectively, were increased, which was equal to the diagnostic sensitivity of the cTnI rapid device at the time of arrival (60%) (Fig. 4). Figure 4. Open in new tabDownload slide Diagnostic sensitivity of CK enzyme activity, CK-MB mass, and troponin I (rapid device) for the diagnosis of acute myocardial infarction at arrival in the emergency room and 4 h later. ∗, P = 0.004 vs CK-MB mass/P <0.001 vs CK activity; ∗∗, P = 0.03 vs CK-MB mass/P <0.001 vs CK activity. Figure 4. Open in new tabDownload slide Diagnostic sensitivity of CK enzyme activity, CK-MB mass, and troponin I (rapid device) for the diagnosis of acute myocardial infarction at arrival in the emergency room and 4 h later. ∗, P = 0.004 vs CK-MB mass/P <0.001 vs CK activity; ∗∗, P = 0.03 vs CK-MB mass/P <0.001 vs CK activity. Patients with unstable angina (n = 321) Among 321 patients with unstable angina, 122 patients (38%) had at least one positive cTnI test result within 4 h after arrival and 7.4 ± 2.9 h after onset of symptoms, respectively (Fig. 5). Positive results were confirmed by quantitative determination in 92% of the patients for Cobas EIA. Only one patient with a negative cTnI rapid assay had a cTnI value of 0.12 μg/L (Cobas EIA) and 0.09 μg/L (Access). In 98% of the cTnI-positive patients, a substantial coronary stenosis could be detected by angiogram. Increased CK-MB mass was found in 13 patients (4%), and in 7 patients (2%), CK enzyme activity was increased (81–134 U/L). Figure 5. Open in new tabDownload slide Proportion of positive test results in different categories of patients for CK-MB mass (Access) and troponin I determination (rapid device). Figure 5. Open in new tabDownload slide Proportion of positive test results in different categories of patients for CK-MB mass (Access) and troponin I determination (rapid device). Patients with cardiac chest pain of nonischemic origin (n = 37) Patients had acute congestive heart failure (n = 15), perimyocarditis (n = 11), ventricular arrhythmia (n = 8), or rejection after heart transplantation (n = 3). In patients with congestive heart failure (73% ischemic), eight positive rapid test results (53%) were documented, but none of them had marked coronary stenosis. In seven patients with perimyocarditis (58%), a positive test result was documented, and all patients with ventricular arrhythmia and rejection of transplanted heart had increased cTnI concentrations (Fig. 5). All positive test results were confirmed by Cobas EIA and in 95% by the Access analyzer. Patients with chest pain of noncardiac origin (n = 120) Patients had acute pulmonary embolism (n = 5), acute pleuritis (n = 7), pneumonia (n = 12), musculoskeletal syndrome (n = 51), or traumatic chest pain (n = 45). Positive cTnI tests were found in two patients with fulminant pulmonary embolism and echocardiographic signs of right ventricular overloading (Fig. 5). Both were confirmed by Cobas EIA and Access analyzer. No cTnI was detected in the other patients, both by rapid device and ELISA. CK-MB mass was increased in 32 (71%) patients with traumatic chest pain, in 3 (6%) patients with musculoskeletal syndrome, and in 4 (33%) patients with pneumonia. Patients with end-stage renal failure (n = 27) In 27 patients with end-stage renal failure but no history of coronary heart disease, 53 blood samples were collected before and after planned hemodialysis. No positive bedside tests were detected, and likewise no increased cTnI concentrations were determined by ELISA (Fig. 5). CK-MB mass (9.5 ± 7.2 μg/L) was increased in 8 patients (30%), and CK activity (128 ± 56 U/L) was increased in 12 patients (44%). Discussion Our results indicate that the new qualitative bedside test is a reliable, convenient, and rapid method for the qualitative determination of cTnI. The device was easy to handle and not affected by temperature fluctuations between 4 and 37 °C. However, increasing analytical sensitivity was documented after storage at −20 °C for 7 days. The rapid test system provided a high test accuracy compared with both ELISA systems. The major advantage of a rapid test system is the applicability of whole blood anticoagulated by heparin so that additional time-consuming preparations such as clotting and centrifugation can be omitted. The performance of the rapid test system at the point of care by medical or paramedical personnel avoids additional time delay and possible sample damage or confusion. But the absence of a quantitative test result represents a disadvantage because positive correlation between the amount of troponin release and cardiac risk has been shown (8)(28) . Although the speed of color development and the intensity of the test signal apparently depends on the cTnI concentration in the blood sample, the rapid device provides at best a semiquantitative test result. However, no different therapeutic strategies according to the troponin concentration have been established thus far (12)(28) , favoring cTnI rapid testing for rapid decision-making and in facilities without access to quantitative troponin measurements. The analytical sensitivity of the rapid assay is influenced by factors other than the analytical characteristics of the test device, as well. The variability of visual assessment as to the presence or absence of the signal line at low troponin concentrations represents a critical factor in clinical routine. The cutoff defined in our analysis as the cTnI concentration leading to positive test results by all observers in all repetitive samples was settled at 0.14 μg/L. However, we could demonstrate that between 0.10 and 0.14 μg/L, the sensitivity depends on the users’ training status. Therefore, a careful acquaintance with training samples is recommended. The highly experienced personnel achieved a detection limit of 0.08 μg/L cTnI without an increase in false-positive test results. The rapid test device revealed a distinct improvement to CK-MB mass determination with higher diagnostic sensitivity for the detection of acute myocardial infarction (60% vs 48%; 3.5 ± 2.7 h after onset of pain; P <0.01). The cTnI determination is expected to reveal absolute myocardial specificity because cTnI is not expressed in fetal and healthy or diseased adult human skeletal muscle tissue, promising no false-positive test results in patients with substantial skeletal muscle damage or renal failure (29)(30)(31)(32)(33) . The investigated cTnI test systems produced no positive results in patients with end-stage renal failure and acute or chronic skeletal muscle injury, whereas 30% and 71% of the patients, respectively, had increased CK-MB mass concentrations. Diagnostic specificity, however, for the detection of acute myocardial infarction according to WHO criteria was lower than that of CK-MB because of an marked rate of positive cTnI test results in patients with unstable angina (38% vs 4% for CK-MB mass; P <0.001). The low incidence of patients with increased CK-MB concentrations might be related to strict exclusion of all patients with non-Q-wave acute myocardial infarction. Additional positive cTnI test results were documented in patients with acute congestive heart failure, myocarditis, ventricular arrhythmia, and rejection of transplanted heart. Recent studies indicate that, independent of the diagnosis of myocardial infarction, the detection of troponins is associated with increased subsequent cardiac risk; therefore, these positive test results are not to be rated as false-positive (12)(13)(14) . In conclusion, the use of this cTnI rapid device could improve efficacy and safety of decision-making in patients with chest pain that might produce more cost-effective use of intensive care facilities. Future prospective studies must define the role of the new diagnostic marker in patient management to improve their adverse prognosis (12)(34) . References 1 . Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group. Indications for fibrinolytic therapy in suspected acute myocardial infarction. Collaborative overview of early mortality and major morbidity results from all randomised trials of more than 1000 patients. Lancet 1994 ; 343 : 311 -322. Crossref Search ADS PubMed 2 Gibler WB, Young GP, Hedges JR, Lewis LM, Smith MS, Carleton SC, et al. Acute myocardial infarction in chest pain patients with nondiagnostic ECG’s: serial CK-MB sampling in the emergency department. Ann Emerg Med 1992 ; 21 : 504 -512. Crossref Search ADS PubMed 3 Brush JE, Brand D, Acampora D, Chalmer B, Wackers FJ. Use of the initial electrocardiogram to predict in-hospital complications of acute myocardial infarction. N Engl J Med 1985 ; 312 : 1137 -1141. Crossref Search ADS PubMed 4 Lee TH, Rouan GW, Weisberg MC, Brand DA, Cook EF, Acampora D, Goldman L. Sensitivity of routine clinical criteria for diagnosing myocardial infarction within 24 h of hospitalization. Ann Intern Med 1987 ; 106 : 81 -186. 5 Rude RE, Poole KW, Muller JE, Rude RE, Poole WK, Muller JE, et al. Electrocardiographic and clinical criteria for recognition of acute myocardial infarction based on analysis of 3697 patients. Am J Cardiol 1983 ; 52 : 936 -942. Crossref Search ADS PubMed 6 Rusnak RA, Stair TO, Hansen K, Fastow JS. Litigation against the emergency physician: common features in cases of missed myocardial infarction. Ann Emerg Med 1989 ; 18 : 1029 -1034. Crossref Search ADS PubMed 7 McCarthy B, Beshansky JR, D’Agostino RB, Selker HP. Missed diagnoses of acute myocardial infarction in the emergency department: results from a multicenter study. Ann Emerg Med 1993 ; 22 : 579 -582. Crossref Search ADS PubMed 8 . The TIMI IIIB Investigators. Effects of tissue plasminogen activator and a comparison of early invasive and conservative strategies in unstable angina and non-Q-wave myocardial infarction. Results of the TIMI IIIB Trial. Circulation 1994 ; 89 : 1545 -1556. Crossref Search ADS PubMed 9 Bakker AJ, Koelemay MJ, Gorgels JP, van Vlies B, Smits R, Tijssen JG, Haagen FD. Failure of new biochemical markers to exclude acute myocardial infarction at admission. Lancet 1993 ; 342 : 1220 -1222. Crossref Search ADS PubMed 10 Hamm CW, Ravkilde J, Gerhardt W, Jorgensen P, Peheim E, Ljungdahl L, et al. The prognostic value of serum troponin T in unstable angina. N Engl J Med 1992 ; 327 : 146 -150. Crossref Search ADS PubMed 11 Ravkilde J, Nissen H, Horder M, Thygesen K. Independent prognostic value of serum creatine kinase isoenzyme MB mass, cardiac troponin T and myosin light chain levels in suspected acute myocardial infarction. J Am Coll Cardiol 1995 ; 25 : 574 -581. Crossref Search ADS PubMed 12 Hamm CW, Goldmann BU, Heeschen C, Kreymann G, Berger J, Meinertz T. Emergency room triage of patients with acute chest based on rapid testing for troponin T or troponin I. N Engl J Med 1997 ; 337 : 1648 -1653. Crossref Search ADS PubMed 13 Adams JE, III, Bodor GS, Dávila-Román VG, Delmez JA, Apple FS, Ladenson JH, Jaffe AS. Cardiac troponin I is a marker with high specificity for cardiac injury. Circulation 1993 ; 88 : 101 -106. Crossref Search ADS PubMed 14 Antman EM, Tanasijevic MJ, Thompson B, Schactman M, McCabe CH, Cannon CP, et al. Cardiac-specific troponin I levels to predict the risk of mortality in patients with acute coronary syndromes. N Engl J Med 1996 ; 335 : 1342 -1349. Crossref Search ADS PubMed 15 Wade R, Eddy R, Shows TB, Kedes L. cDNA sequences, tissue specific expression, and chromosomal mapping of the human slow-twitch muscle isoform of troponin I. Genomics 1990 ; 7 : 346 -357. Crossref Search ADS PubMed 16 MacGeoch C, Garton PJ, Vallins WJ, Bhavsar P, Spurr NK. The human cardiac troponin I locus: assignment to chromosome 19p13.2–19q13.2. Hum Genet 1991 ; 88 : 101 -104. Crossref Search ADS PubMed 17 Davies E, Gawad Y, Takahashi M, Shi Q, Lam P, Styba G, et al. Analytical performance and clinical utility of a sensitive immunoassay for determination of human cardiac troponin I. Clin Biochem 1997 ; 30 : 479 -490. Crossref Search ADS PubMed 18 Cummins P, Perry SV. Troponin I from human skeletal and cardiac muscles. Biochem J 1978 ; 171 : 251 . Crossref Search ADS PubMed 19 Humphreys JE, Cummins P. Regulatory proteins of the myocardium: atrial and ventricular tropomyosin and troponin I in the developing and adult bovine and human heart. J Mol Cell Cardiol 1984 ; 16 : 643 -657. Crossref Search ADS PubMed 20 Wilkinson JM, Grand RJA. Comparison of amino acid sequence of troponin I. Nature 1978 ; 271 : 31 -35. Crossref Search ADS PubMed 21 Heeschen C, Goldmann BU, Möller RH, Hamm CW. Evaluation of a new troponin I beside test in acute myocardial infarction [Abstract]. Circulation 1996 ; 94 : 322 . 22 Katrukha AG, Bereznikova AV, Esakova TV, Pettersson K, Lovgren T, Severina ME, et al. Troponin I is released in bloodstream of patients with acute myocardial infarction not in free form but as complex. Clin Chem 1997 ; 43 : 1379 -1385. Crossref Search ADS PubMed 23 Takahashi M, Lee L, Jackowski G. Development and characterization of monoclonal antibodies specific for human cardiac troponin I [Abstract]. Clin Chem 1993 ; 39 : 1133 . 24 Coulter KM, Takahashi M, Lee L. Development of a novel cardiac troponin I mono-poly ELISA that uses chicken IgY [Abstract]. Clin Chem 1994 ; 40 : 1030 . 25 Larue C, Calzolari C, Bertinchant JP, Leclerq F, Grolleau R, Pau B. Cardiac-specific immunoenzymometric assay of troponin I in the early phase of acute myocardial infarction. Clin Chem 1993 ; 39 : 972 -979. Crossref Search ADS PubMed 26 Chapelle JP, El Allaf M. Automated quantification of creatine kinase MB isoenzyme in serum by radial partition immunoassay, with use of the stratus analyzer. Clin Chem 1990 ; 36 : 99 -101. Crossref Search ADS PubMed 27 Braunwald E. Unstable angina: a classification. Circulation 1989 ; 80 : 410 -414. Crossref Search ADS PubMed 28 . for the FRISC Study GroupLindahl B, Venge P, Wallentin L. Relation between troponin T and the risk of subsequent cardiac events in unstable coronary artery disease. Circulation 1996 ; 93 : 1651 -1657. Crossref Search ADS PubMed 29 Anderson PA, Greig A, Mark TM. Molecular basis of human cardiac troponin T isoforms expressed in the developing, adult, and failing heart. Circ Res 1995 ; 76 : 681 -686. Crossref Search ADS PubMed 30 Bhayana V, Gougoulias T, Cohoe S. Discordance between results for serum troponin T and Troponin I in renal disease. Clin Chem 1995 ; 41 : 312 -317. Crossref Search ADS PubMed 31 Haller C, Stevanovich A, Katus HA. Are cardiac troponins reliable serodiagnostic markers of cardiac ischaemia in end-stage renal disease? [Editorial]. Nephrol Dial Transplant 1996 ; 11 : 941 -944. Crossref Search ADS PubMed 32 McLaurin MD, Apple FS, Voss EM, Herzog CA, Sharkey SW. Cardiac troponin I, cardiac troponin T, and creatine kinase MB in dialysis patients without ischemic heart disease: evidence of cardiac troponin T expression in skeletal muscle. Clin Chem 1997 ; 43 : 976 -982. Crossref Search ADS PubMed 33 Apple FS, Sharkey SW, Hoeft P, Skeate R, Voss E, Dahlmeier BA, Preese LM. Prognostic value of serum cardiac troponin I and T in chronic dialysis patients: a 1-year outcomes analysis. Am J Kidney Dis 1997 ; 29 : 399 -403. Crossref Search ADS PubMed 34 Keffer JH. The cardiac profile and proposed practice guideline for acute ischemic heart disease. Am J Clin Pathol 1997 ; 107 : 398 -409. Crossref Search ADS PubMed © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Duell, P, Barton;Wheaton, David, L;Shultz,, Amy;Nguyen,, Hien

doi: 10.1093/clinchem/44.9.1931pmid: N/A

Abstract Melatonin has been suggested as a potent antioxidant that may protect against development of atherosclerosis and cancer; however, these effects are unproven and controversial. The antioxidant capacity of melatonin was tested in comparison with α-tocopherol, ascorbic acid, and the melatonin precursors tryptophan and serotonin, by measuring inhibition of metal ion-mediated and human macrophage-mediated oxidation of LDL. Melatonin had weak antioxidant activity that was detectable only at concentrations 10 000- to 100 000-fold higher than physiologic concentrations. These results were comparable with published data showing that the radical scavenging activity of melatonin requires markedly supraphysiologic concentrations. In contrast, α-tocopherol was 50- to 100-fold more potent and was efficacious at physiologic concentrations. Ascorbic acid and tryptophan also were active at physiologic concentrations and were significantly more potent than melatonin. In summary, extremely supraphysiologic concentrations of melatonin had only weak antioxidant activity, which was surpassed by α-tocopherol, ascorbic acid, and tryptophan. Melatonin is a lipophilic indoleamine hormone, derived from tryptophan, that is secreted by the pineal gland primarily during periods of darkness (1)(2) . It is believed to play a major role in the regulation of diurnal rhythms in vertebrate animals, including humans (2)(3)(4) . It also has been suggested as a powerful antioxidant that scavenges superoxide, hydroxyl, and peroxyl radicals (5)(6)(7)(8)(9) ; however, these effects have been observed primarily at markedly supraphysiologic concentrations. Some investigators have proposed that the antioxidant properties of melatonin may protect against development of cancer, atherosclerosis, and other consequences of aging (10)(11)(12)(13) ; however, these effects remain unproven and controversial (14)(15) . A large body of experimental evidence supports the hypothesis that oxidation of LDL contributes to the development of atherosclerosis (16)(17)(18)(19)(20)(21) . Moreover, it is postulated that inhibition of LDL oxidation by antioxidants might protect against the development of atherosclerosis (22)(23)(24)(25)(26)(27) . In both human and animal studies, resistance of LDL to oxidation ex vivo often has been associated with decreased atherosclerosis (19)(25)(26)(28) . Moreover, antioxidant administration inhibits oxidation of LDL and typically, but not always, has been associated with decreased progression of atherosclerosis (19)(21)(22)(25)(26) . The strongest data from human subjects come from the Cambridge Heart Antioxidant Study, in which the risk of cardiovascular death and nonfatal myocardial infarction was reduced 47% in patients with coronary disease who received 400–800 IU of vitamin E daily (27) . Because melatonin has been reported to be a powerful antioxidant with possible antiatherogenic properties, we endeavored to test the relative capacity of melatonin, its structurally related precursors, tryptophan and serotonin, and the antioxidant vitamins, α-tocopherol and ascorbic acid, to inhibit oxidative modification of LDL in vitro. Materials and Methods lipoprotein isolation LDL was isolated from pooled fresh human plasma by sequential ultracentrifugation in the density range 1.019–1.063 kg/L using standard methods (29)(30) . LDL was stored in 1 mmol/L EDTA in phosphate-buffered saline (9 g/L NaCl) under nitrogen at 4 °C in the dark and was used within 2 weeks after isolation. EDTA was removed from LDL samples before oxidation assays by extensive dialysis in degassed phosphate-buffered saline (9 g/L NaCl) or desalting with a Sephadex G-25 (PD-10) column (Sigma Chemical Co.) equilibrated with degassed phosphate-buffered saline (9 g/L NaCl). cell culture Human monocyte-derived macrophages were obtained by Ficoll/Hypaque density gradient centrifugation of blood from healthy donors and grown in primary culture in RPMI-1640 culture medium (Gibco BRL) with 200 mL/L autologous serum at 37 °C in humidified incubators containing 5% CO2/95% air as previously described (30)(31)(32) . Cells were plated at a density of 1–2 × 106 cells per 22-mm plastic well and used after 7–10 days. cell-free ldl oxidation LDL was oxidized in vitro by incubating 0.52 mmol/L (200 μg/mL) LDL-cholesterol in phosphate-buffered saline (9 g/L NaCl) in the presence of 5 μmol/L copper sulfate at 20 °C for 8 h in a temperature-controlled, multicuvette Shimadzu spectrophotometer. In other experiments, LDL was oxidized by incubation of 1.82 × 10−7 mol/L (100 μg/mL) LDL protein in Ham’s F-10 medium at 37 °C (Gibco BRL) for 18–24 h. Oxidative modification of LDL was monitored by determining the formation of conjugated dienes by semicontinuous measurements of the absorbance at 234 nm (33)(34) or sequential measurements of thiobarbituric acid-reactive substances (TBARSs) (35)(36) . These standard methods for determining the susceptibility of LDL to oxidative modification have been used extensively to evaluate the antioxidant properties of various compounds (33)(34) . The lag time for LDL oxidation was measured as the intercept of tangent lines for the initiation and propagation phases of the curve showing the time course for formation of conjugated dienes (33)(34) . The propagation rate was measured as the slope of the propagation phase during formation of conjugated dienes (33)(34) . cell-mediated ldl oxidation Macrophage-mediated oxidation of LDL was measured by incubating 1.82 × 10−7 mol/L (100 μg/mL) LDL protein in triplicate 22-mm wells with 1–2 × 106 cells in Ham’s F-10 medium at 37 °C for 18–24 h. Cell-free control wells were used for all conditions. At the end of incubation, oxidation of LDL was arrested by chilling the medium and adding 200 μmol/L EDTA and 40 μmol/L butylated hydroxytoluene. Aliquots were assayed for TBARS content (35)(36) and normalized for the amount of cell protein determined by a modified method of Hartree (37) , using a bicinchoninic acid microtiter plate assay (Pierce Chemical Co.). Cell-mediated oxidation was calculated as the difference between TBARS content in cell-containing and cell-free conditions. ldl electrophoresis Electrophoretic mobility of control and modified LDL was assessed by 0.8% agarose gel electrophoresis at pH 8.6 in barbitol buffer (38) . LDL was visualized with Sudan black staining (38) . Relative electrophoretic mobility was calculated as the ratio of migration of modified LDL compared with control LDL. preparation of antioxidants Melatonin and α-tocopherol were dissolved in ethanol. Tryptophan was solubilized in 0.5 mol/L (0.5 N) hydrochloric acid, and serotonin and ascorbic acid were dissolved in distilled water or ethanol. Antioxidant solutions were prepared fresh, protected from light, and added to the incubation medium in concentrations ranging from 0–50 μmol/L (Sigma). Control samples of LDL with equal volumes of diluent were used as the reference measurement in every experiment. The final concentration of ethanol in the incubation medium was ≤20 mL/L. The final concentration of HCl in experiments with tryptophan was ≤0.25 mmol/L. Neither ethanol nor dilute HCl at these concentrations affected the lag time or propagation rate for LDL oxidation. statistical analysis Statistical analyses were done with Mann–Whitney rank sum and t-testing using Sigmastat statistical software (Jandel Scientific). P values <0.05 were considered statistically significant. Results In a cell-free system, melatonin at concentrations up to 5 μmol/L had no appreciable effect on LDL oxidation measured as copper-mediated conjugated diene formation (Fig. 1). At a maximal concentration of 50 μmol/L melatonin, there was moderate inhibition of LDL oxidation manifested as a 19% ± 9% increase in the lag time (P = 0.036) and 48% ± 16% decrease in propagation rate (P = 0.016). This concentration of melatonin is 10 000- to 100 000-fold greater than peak physiologic plasma concentrations (∼45–900 pmol/L) (39) and 50- to 100-fold higher than maximal serum concentrations achieved after large pharmacologic doses of melatonin (39)(40) . Figure 1. Open in new tabDownload slide Dose–response inhibition of LDL oxidation by melatonin and α-tocopherol. LDL at a concentration of 200 g/L cholesterol in degassed phosphate-buffered saline was incubated with 5 μmol/L copper sulfate and variable concentrations of melatonin or α-tocopherol for 8 h at 20 °C. The lag time was measured as described in Materials and Methods. The results are representative of 13 experiments using different preparations of LDL. Values are means ± SD. Figure 1. Open in new tabDownload slide Dose–response inhibition of LDL oxidation by melatonin and α-tocopherol. LDL at a concentration of 200 g/L cholesterol in degassed phosphate-buffered saline was incubated with 5 μmol/L copper sulfate and variable concentrations of melatonin or α-tocopherol for 8 h at 20 °C. The lag time was measured as described in Materials and Methods. The results are representative of 13 experiments using different preparations of LDL. Values are means ± SD. In contrast, α-tocopherol, another lipid-soluble antioxidant (34)(41) , was about 50-fold more potent compared with melatonin (P <0.001; Fig. 1). At a concentration of 5 μmol/L, α-tocopherol increased the lag time for conjugated diene formation by 47% ± 17% (P = 0.036 vs control) without significantly affecting the propagation rate. LDL oxidation was essentially abolished in the presence of 50 μmol/L α-tocopherol (data not shown). These concentrations of α-tocopherol are comparable to physiologic serum concentrations of 12–46 μmol/L. Ascorbic acid, an effective water-soluble antioxidant, also was significantly more potent as an antioxidant compared with melatonin (P <0.001; Table 1). Table 1. Effect of melatonin, melatonin precursors, and other antioxidants on the lag time and propagation rate for conjugated diene formation during copper ion-mediated LDL oxidation.1 . Change in lag time, % of control . . Change in propagation rate . . Physiologic serum concentration . . 5 μmol/L . 50 μmol/L . 5 μmol/L . 50 μmol/L . . Melatonin −2 ± 9 19 ± 92 0 ± 8 −48 ± 162 Peak, 45–900 pmol/L Tryptophan 12 ± 92 117 ± 222 −11 ± 92 −44 ± 352 25–125 μmol/L Serotonin −40 ± 52 >300 −26 ± 122 NA3 0.45–1.20 μmol/L Ascorbic acid 37 ± 72 94 ± 142 8 ± 22 62 ± 282 30–110 μmol/L α-Tocopherol 47 ± 172 >300 2 ± 16 NA3 12–46 μmol/L . Change in lag time, % of control . . Change in propagation rate . . Physiologic serum concentration . . 5 μmol/L . 50 μmol/L . 5 μmol/L . 50 μmol/L . . Melatonin −2 ± 9 19 ± 92 0 ± 8 −48 ± 162 Peak, 45–900 pmol/L Tryptophan 12 ± 92 117 ± 222 −11 ± 92 −44 ± 352 25–125 μmol/L Serotonin −40 ± 52 >300 −26 ± 122 NA3 0.45–1.20 μmol/L Ascorbic acid 37 ± 72 94 ± 142 8 ± 22 62 ± 282 30–110 μmol/L α-Tocopherol 47 ± 172 >300 2 ± 16 NA3 12–46 μmol/L 1 Values indicate percentage of change in lag time or propagation rate ± SD and represent data from three to six separate experiments. 2 P <0.05 compared with control. 3 NA indicates that values could not be calculated because oxidation was completely inhibited during 8 h of incubation. Open in new tab Table 1. Effect of melatonin, melatonin precursors, and other antioxidants on the lag time and propagation rate for conjugated diene formation during copper ion-mediated LDL oxidation.1 . Change in lag time, % of control . . Change in propagation rate . . Physiologic serum concentration . . 5 μmol/L . 50 μmol/L . 5 μmol/L . 50 μmol/L . . Melatonin −2 ± 9 19 ± 92 0 ± 8 −48 ± 162 Peak, 45–900 pmol/L Tryptophan 12 ± 92 117 ± 222 −11 ± 92 −44 ± 352 25–125 μmol/L Serotonin −40 ± 52 >300 −26 ± 122 NA3 0.45–1.20 μmol/L Ascorbic acid 37 ± 72 94 ± 142 8 ± 22 62 ± 282 30–110 μmol/L α-Tocopherol 47 ± 172 >300 2 ± 16 NA3 12–46 μmol/L . Change in lag time, % of control . . Change in propagation rate . . Physiologic serum concentration . . 5 μmol/L . 50 μmol/L . 5 μmol/L . 50 μmol/L . . Melatonin −2 ± 9 19 ± 92 0 ± 8 −48 ± 162 Peak, 45–900 pmol/L Tryptophan 12 ± 92 117 ± 222 −11 ± 92 −44 ± 352 25–125 μmol/L Serotonin −40 ± 52 >300 −26 ± 122 NA3 0.45–1.20 μmol/L Ascorbic acid 37 ± 72 94 ± 142 8 ± 22 62 ± 282 30–110 μmol/L α-Tocopherol 47 ± 172 >300 2 ± 16 NA3 12–46 μmol/L 1 Values indicate percentage of change in lag time or propagation rate ± SD and represent data from three to six separate experiments. 2 P <0.05 compared with control. 3 NA indicates that values could not be calculated because oxidation was completely inhibited during 8 h of incubation. Open in new tab Comparable differences between the antioxidant capacity of melatonin and α-tocopherol were observed when LDL oxidation was quantified by measuring TBARSs after incubating LDL in Ham’s F-10 medium [containing 3 mmol/L (0.834 mg/mL) FeSO4 · 7 H2O and 10 μmol/L (0.0025 mg/mL) CuSO4 · 5 H2O] for 18 h at 37 °C. α-Tocopherol at a concentration of 50 μmol/L reduced LDL oxidation by 61% ± 4% (P <0.001 vs control) whereas equimolar concentrations of melatonin reduced LDL oxidation by only 22% ± 3% (P <0.001 vs control and α-tocopherol; Fig. 2). Figure 2. Open in new tabDownload slide Inhibition of cell-free oxidation of LDL in Ham’s F-10 medium by 50 μmol/L melatonin and α-tocopherol (vitamin E). LDL at a concentration of 100 mg/L protein was incubated for 18 h at 37 °C and assayed for TBARSs. Values are means ± SD for triplicate measurements. Results are representative of three experiments using different preparations of LDL. (*), P <0.001 vs control or melatonin; (#), P <0.001 vs control. Figure 2. Open in new tabDownload slide Inhibition of cell-free oxidation of LDL in Ham’s F-10 medium by 50 μmol/L melatonin and α-tocopherol (vitamin E). LDL at a concentration of 100 mg/L protein was incubated for 18 h at 37 °C and assayed for TBARSs. Values are means ± SD for triplicate measurements. Results are representative of three experiments using different preparations of LDL. (*), P <0.001 vs control or melatonin; (#), P <0.001 vs control. Because the indole moiety is presumed to be responsible for radical scavenging activity of melatonin, the relative antioxidant potency of tryptophan and serotonin, indole precursors of melatonin, was tested (Table 1). At concentrations <5 μmol/L, neither tryptophan nor serotonin significantly influenced the lag time for LDL oxidation (data not shown). However, at a concentration of 5 μmol/L, the antioxidant activity of tryptophan was much greater than equimolar amounts of melatonin, producing a 12% ± 9% increase in the lag time compared with −2% ± 9% for melatonin (P = 0.045; Table 1). In contrast, 5 μmol/L serotonin appeared to enhance LDL oxidation, producing a consistent 40% ± 5% decrease in the lag time (P = 0.016). At a concentration of 50 μmol/L, tryptophan increased the lag time by 117% ± 22% (P = 0.016), whereas serotonin completely inhibited LDL oxidation during the 8-h incubation. To test the effects of melatonin on cell-mediated oxidation, primary cultures of human monocyte-derived macrophages and cell-free control wells were incubated with 1.82 × 10−7 mol/L (100 μg/mL) LDL protein for 18 h at 37 °C in Ham’s F-10 medium. α-Tocopherol at a concentration of 50 μmol/L reduced oxidation of LDL measured by TBARSs by 87% ± 10% (P <0.001 vs control or melatonin), whereas equimolar concentrations of melatonin reduced LDL oxidation by only 21% ± 10% (P = 0.039 vs control; Fig. 3). Figure 3. Open in new tabDownload slide Inhibition of cell-mediated oxidation of LDL by 50 μmol/L melatonin and α-tocopherol. LDL at a concentration of 100 mg/L protein was incubated with human monocyte-derived macrophages in Ham’s F-10 medium for 18 h at 37 °C. Cell-mediated oxidation was calculated as the difference between total TBARSs in the extracellular medium minus TBARSs in cell-free conditions. Values are means ± SD for triplicate measurements. Results are representative of three experiments using different preparations of LDL. (*), P <0.001 vs control or melatonin; (#), P = 0.039 vs control. Figure 3. Open in new tabDownload slide Inhibition of cell-mediated oxidation of LDL by 50 μmol/L melatonin and α-tocopherol. LDL at a concentration of 100 mg/L protein was incubated with human monocyte-derived macrophages in Ham’s F-10 medium for 18 h at 37 °C. Cell-mediated oxidation was calculated as the difference between total TBARSs in the extracellular medium minus TBARSs in cell-free conditions. Values are means ± SD for triplicate measurements. Results are representative of three experiments using different preparations of LDL. (*), P <0.001 vs control or melatonin; (#), P = 0.039 vs control. To assess the effects of melatonin on apolipoprotein B modification, the relative electrophoretic mobility of LDL was determined after cell-mediated and cell-free oxidation of 1.82 × 10−7 mol/L (100 μg/mL) LDL protein in Ham’s F-10 medium for 18 h at 37 °C (Fig. 4). Mobility of LDL incubated with human monocyte-derived macrophages was unaffected by 50 μmol/L melatonin, whereas equimolar α-tocopherol prevented 38% of increased mobility of modified LDL. Under cell-free conditions, 50 μmol/L melatonin blocked 16% of increased mobility of modified LDL; however, α-tocopherol blocked 56% of the increase. Figure 4. Open in new tabDownload slide Alteration of electrophoretic mobility of LDL by melatonin and α-tocopherol. LDL was incubated at a concentration of 1.82 × 10−7 mol/L (100 μg protein/mL) in Ham’s F-10 medium for 18 h at 37 °C in the presence of human monocyte-derived macrophages (lanes 1–4) or cell-free conditions (lanes 5–8). Electrophoretic mobility of unmodified control LDL (lanes 1 and 5) was compared with LDL incubated in Ham’s F-10 medium (lanes 2and 6) or Ham’s F-10 medium with 50 μmol/L melatonin (lanes 3 and 7) or equimolar α-tocopherol (lanes 4 and 8). Results are representative of three experiments using different preparations of LDL. Figure 4. Open in new tabDownload slide Alteration of electrophoretic mobility of LDL by melatonin and α-tocopherol. LDL was incubated at a concentration of 1.82 × 10−7 mol/L (100 μg protein/mL) in Ham’s F-10 medium for 18 h at 37 °C in the presence of human monocyte-derived macrophages (lanes 1–4) or cell-free conditions (lanes 5–8). Electrophoretic mobility of unmodified control LDL (lanes 1 and 5) was compared with LDL incubated in Ham’s F-10 medium (lanes 2and 6) or Ham’s F-10 medium with 50 μmol/L melatonin (lanes 3 and 7) or equimolar α-tocopherol (lanes 4 and 8). Results are representative of three experiments using different preparations of LDL. Discussion Melatonin has been suggested to have potent antioxidant properties that may prevent the development of cancer, atherosclerosis, and other consequences of aging (5)(10)(11)(12)(13) ; however, these hypothetical effects are unproven (14)(15) . In some animal studies, melatonin has been shown to have antioxidant properties in vivo, but often only at very high parenteral doses, e.g., 10 to 450 mg/kg body weight (5)(9)(10) . In one small human study, nocturnal secretion of melatonin was decreased in 15 patients with coronary atherosclerosis (42) ; however, these data are insufficient to allow conclusions about the relationship between melatonin, antioxidant activity, and vascular disease (43) . Thus, conclusive studies regarding the relevance of antioxidant properties of melatonin in prevention of disease are not available. Because oxidation of LDL is believed to play an important etiologic role in the development of atherosclerosis, the capacity of melatonin to inhibit oxidation of LDL was tested in a standardized in vitro system. The susceptibility of LDL to undergo oxidation in this assay has been correlated with the severity of atherosclerosis in men with myocardial infarction (28) . Although the results of other studies have suggested that high concentrations of melatonin may inhibit LDL oxidation (44)(45)(46)(47) , dose–response data comparing the capacity of melatonin to inhibit LDL oxidation with those of α-tocopherol and other antioxidants have been limited. Melatonin had no antioxidant activity at physiologic concentrations and only moderate antioxidant activity at concentrations that were 4–6 orders of magnitude greater than peak physiologic concentrations and 50- to 100-fold higher than maximal serum concentrations achievable after large oral doses of melatonin up to 240 mg (39)(40) . Very large doses of melatonin >1000 mg might achieve transient serum concentrations >5 μmol/L; however, the safety and clinical relevance of such doses are unclear. Although the indole moiety in melatonin has been suggested to be responsible for antioxidant activity (5) , serotonin and tryptophan differed substantially from melatonin in their capacity to inhibit LDL oxidation. Physiologic concentrations of tryptophan (25–125 μmol/L) significantly inhibited LDL oxidation, whereas physiologic concentrations of serotonin (0.45–1.20 μmol/L) were inactive. Moreover, at a concentration of 5 μmol/L, serotonin appeared to have prooxidant activity, producing accelerated oxidation of LDL. Halliwell and co-workers (48) also showed that serotonin was strongly prooxidant in an Fe(3+)-EDTA H2O-deoxyribose system. In other studies, 25 μmol/L serotonin stimulated uptake of oxidized LDL by macrophages (49) . The biological importance of these findings is uncertain; however, these data suggest that supraphysiologic concentrations of serotonin may have the potential to enhance oxidation under some conditions. Both tryptophan and serotonin were more potent than melatonin at a concentration of 50 μmol/L. α-Tocopherol has previously been demonstrated to be a potent inhibitor of LDL oxidation (34)(41)(50) and is hypothesized to protect against the development of atherosclerosis (19)(21)(22)(25)(26) . In one recent double-blind placebo-controlled clinical trial, supplementation with α-tocopherol reduced the risk of cardiovascular death and nonfatal myocardial infarction by 47% (27) . In the present study, α-tocopherol clearly was the most potent antioxidant and was ∼50-fold more efficacious than melatonin. Moreover, α-tocopherol had significant antioxidant activity at concentrations that were comparable to physiologic serum concentrations. Ascorbic acid, a water-soluble chain-breaking antioxidant, also was more potent than melatonin at all concentrations, and had significant antioxidant activity at physiologic concentrations. The results of recent studies also demonstrated that melatonin did not substantially inhibit oxidation of LDL at concentrations <10–20 μmol/L (44)(45)(46)(47) . Moreover, a comparable dose–response relationship was demonstrated when the antioxidant activity of melatonin was tested in a specific radical-scavenging system using the radical-trapping reagent 2,2′-azinobis(3-ethylbenzothiazoline-6-sulfonic acid) (8) . In those studies, melatonin appeared to have greater radical-scavenging activity than tryptophan or ascorbic acid; however, it nonetheless had minimal scavenging activity at concentrations <50 μmol/L (8) . Maximal radical-scavenging activity of melatonin in those studies occurred at concentrations of 150–200 μmol/L (8) . Thus, melatonin appears to have substantial antioxidant activity only at markedly supraphysiologic concentrations. In summary, extremely supraphysiologic concentrations of melatonin had weak antioxidant properties in this study; however, physiologic concentrations of α-tocopherol and ascorbic acid were significantly more efficacious at equimolar concentrations. Similarly, tryptophan and serotonin were significantly more potent than melatonin at the highest concentrations. Although high doses of melatonin (e.g., 10–450 mg/kg body weight parenterally) have been shown to have antioxidant properties in experimental animals (5)(6)(7)(8)(9) , there currently are no data in animals or humans that conclusively demonstrate that melatonin plays a role in prevention of atherosclerosis (14)(15) . These results suggest that the potential biological relevance of antioxidant properties of melatonin is uncertain and needs to be interpreted with caution until definitive studies are completed. Division of Endocrinology, Diabetes and Clinical Nutrition, L465, Oregon Health Sciences University, Portland, OR 97201-3098. This work was supported in part by the National Institutes of Health (grant RR00334), the Collins Medical Trust, and the Medical Research Foundation of Oregon. References 1 Lerner AB, Case JD, Heinzelman RV. Structure of melatonin. J Am Chem Soc 1959 ; 81 : 6084 -6085. 2 Reiter RJ. Melatonin: the chemical expression of darkness. Mol Cell Endocrinol 1991 ; 79 : C153 -C158. Crossref Search ADS PubMed 3 Lewy AJ, Sack RL, Blood ML, Bauer VK, Cutler NL, Thomas KH. Melatonin marks circadian phase position and resets the endogenous circadian pacemaker in humans. Ciba Found Symp 1995 ; 183 : 303 -307. PubMed 4 Dijk DJ, Boulos Z, Eastman CI, Lewy AJ, Campbell SS, Terman M. Light treatment for sleep disorders: consensus report. II. Basic properties of circadian physiology and sleep regulation. J Biol Rhythms 1995 ; 10 : 113 -125. Crossref Search ADS PubMed 5 Reiter RJ, Tan D-X, Poeggeler BR, Menendez-Pelaez A, Chen L-D, Saarela S. Melatonin as a free radical scavenger: implications for aging and age related diseases. Ann N Y Acad Sci 1994 ; 719 : 1 -12. Crossref Search ADS PubMed 6 Tan D-X, Chen L-D, Poeggeler B, Manchester LC, Reiter RJ. Melatonin: a potent, endogenous hydroxyl radical scavenger. Endocr J 1993 ; 1 : 57 -60. 7 Pieri C, Maurizio M, Moroni F, Recchioni R, Marcheselli F. Melatonin: a peroxyl radical scavenger more effective than vitamin E. Life Sci 1994 ; 55 : 271 -276. 8 Poeggeler B, Reiter RJ, Hardeland R, Tan D-X, Barlow-Walden LR. Melatonin and structurally-related, endogenous indoles act as potent electron donors and radical scavengers in vitro. Redox Report 1996 ; 2 : 179 -184. Crossref Search ADS PubMed 9 Sewerynek E, Poeggeler B, Melchiorri C, Reiter RJ. H2O2-induced lipid peroxidation in rat brain homogenates is greatly reduced by melatonin. Neurosci Lett 1995 ; 195 : 203 -205. Crossref Search ADS PubMed 10 Poeggeler B, Reiter RJ, Tan D-X, Chen L-D, Manchester LC. Melatonin, hydroxyl radical-mediated oxidative damage, and aging: a hypothesis. J Pineal Res 1993 ; 14 : 151 -168. Crossref Search ADS PubMed 11 Pierpaoli W, Yi CX, Dall’Ara A. Aging-postponing effects of circadian melatonin: experimental evidence, significance and possible mechanisms. Int J Neurosci 1990 ; 51 : 339 -340. Crossref Search ADS PubMed 12 Armstrong SM, Redman JR. Melatonin: a chronobiotic with anti-aging properties?. Med Hypotheses 1991 ; 34 : 300 -309. Crossref Search ADS PubMed 13 Grad BR, Rozencwaig R. The role of melatonin and serotonin in aging: update. Psychoneuroendocrinology 1993 ; 18 : 283 -295. Crossref Search ADS PubMed 14 Bonn D. Melatonin’s multifarious marvels: miracle or myth?. Lancet 1996 ; 347 : 184 . Crossref Search ADS 15 Brzezinski A. Melatonin in humans. N Engl J Med 1997 ; 336 : 186 -195. Crossref Search ADS PubMed 16 Steinberg D. Lewis A. Conner Memorial Lecture–oxidative modification of LDL and atherogenesis. Circulation 1997 ; 95 : 1062 -1071. Crossref Search ADS PubMed 17 Witztum JL, Steinberg D. Role of oxidized low density lipoprotein in atherogenesis. J Clin Investig 1991 ; 88 : 1785 -1792. Crossref Search ADS PubMed 18 Hajjar DP, Haberland ME. Lipoprotein trafficking in vascular cells. Molecular Trojan horses and cellular saboteurs. J Biol Chem 1997 ; 272 : 22975 -22978. Crossref Search ADS PubMed 19 Jialal I, Devaraj S. Low-density lipoprotein oxidation, antioxidants, and atherosclerosis: a clinical biochemistry perspective. Clin Chem 1996 ; 42 : 498 -506. Crossref Search ADS PubMed 20 Navab M, Berliner JA, Watson AD, Hama SY, Territo MC, Lusis AJ, et al. The Yin and Yang of oxidation in the development of the fatty streak. A review based on the 1994 George Lyman Duff Memorial Lecture. Arterioscler Thromb Vasc Biol 1996 ; 16 : 831 -842. Crossref Search ADS PubMed 21 Reaven PD, Witztum JL. Oxidized low density lipoproteins in atherogenesis: role of dietary modification. Annu Rev Nutr 1996 ; 16 : 51 -71. Crossref Search ADS PubMed 22 Steinberg D. Clinical trials of antioxidants in atherosclerosis: are we doing the right thing?. Lancet 1995 ; 346 : 36 -38. Crossref Search ADS PubMed 23 Kwiterovich PO, Jr. The effect of dietary fat, antioxidants, and pro-oxidants on blood lipids, lipoproteins, and atherosclerosis. J Am Diet Assoc 1997 ; 97 (Suppl 7): S31 -S41. Crossref Search ADS PubMed 24 Keaney JF, Jr, Vita JA. Atherosclerosis, oxidative stress, and antioxidant protection in endothelium-derived relaxing factor action. Prog Cardiovasc Dis 1995 ; 38 : 129 -154. Crossref Search ADS PubMed 25 Duell PB. The role of dietary antioxidants in prevention of atherosclerosis. Endocrinologist 1995 ; 5 : 347 -356. Crossref Search ADS 26 Jha P, Flather M, Lonn E, Farkouh M, Yusuf S. The antioxidant vitamins and cardiovascular disease. Ann Intern Med 1995 ; 123 : 860 -872. Crossref Search ADS PubMed 27 Stephens NG, Parsons A, Schofield PM, Kelly F, Cheeseman K, Mitchinson MJ. Randomised controlled trial of vitamin E in patients with coronary disease: Cambridge Heart Antioxidant Study (CHAOS). Lancet 1996 ; 347 : 781 -786. Crossref Search ADS PubMed 28 Regnström J, Nilsson J, Tornvall P, Landou C, Hamsten A. Susceptibility to low-density lipoprotein oxidation and coronary atherosclerosis in man. Lancet 1992 ; 339 : 1183 -1186. Crossref Search ADS PubMed 29 Lipid Research Clinics Program. Manual of Laboratory Operations: Lipid and Lipoprotein Analysis. NIH 75-628. Washington, DC: US Department of Health, Education, and Welfare, 1974:1.. 30 Duell PB, Oram JF, Bierman EL. Nonenzymatic glycosylation of HDL resulting in inhibition of high affinity binding to cultured human fibroblasts. Diabetes 1990 ; 39 : 1257 -1263. Crossref Search ADS PubMed 31 Duell PB, Oram JF, Bierman EL. Nonenzymatic glycosylation of high density lipoprotein impairs HDL receptor-mediated cholesterol efflux. Diabetes 1991 ; 40 : 377 -384. Crossref Search ADS PubMed 32 Chait A, Mazzone T. A secretory product of human monocyte-derived macrophages stimulates low density lipoprotein receptor activity in arterial smooth muscle cells and skin fibroblasts. Arteriosclerosis 1982 ; 2 : 134 -141. Crossref Search ADS PubMed 33 Esterbauer H, Striegl G, Puhl H, Rotheneder M. Continuous monitoring of in vitro oxidation of human low density lipoprotein. Free Radic Res Commun 1989 ; 6 : 67 -75. Crossref Search ADS PubMed 34 Esterbauer H, Dieber-Rotheneder M, Waeg G, Rabl H. Effect of antioxidants on oxidative modification of LDL. Ann Med 1991 ; 23 : 573 -581. Crossref Search ADS PubMed 35 Buege JA, Aust SD. Microsomal lipid peroxidation. Methods Enzymol 1978 ; 52 : 302 -310. Crossref Search ADS PubMed 36 Heinecke JW, Baker L, Rosen H, Chait A. Superoxide-mediated modification of low density lipoprotein by arterial smooth muscle cells. J Clin Investig 1986 ; 77 : 757 -761. Crossref Search ADS PubMed 37 Hartree EF. Determination of protein: a modification of the Lowry method that gives a linear photometric response. Anal Biochem 1972 ; 48 : 422 -427. Crossref Search ADS PubMed 38 Noble RP. Electrophoretic separation of plasma lipoproteins in agarose gel. J Lipid Res 1968 ; 9 : 693 -700. PubMed 39 Cagnacci A, Elliott JA, Yen SS. Amplification of pulsatile LH secretion by exogenous melatonin in women. J Clin Endocrinol Metab 1991 ; 73 : 210 -212. Crossref Search ADS PubMed 40 Waldhauser F, Lieberman HR, Lynch HJ, Waldhauser M, Herkner K, Frisch H, et al. A pharmacological dose of melatonin increases PRL levels in males without altering those of GH, LH, FSH, TSH, and testosterone or cortisol. Neuroendocrinology 1987 ; 46 : 125 -130. Crossref Search ADS PubMed 41 Jialal I, Grundy SM. Effect of dietary supplementation with α-tocopherol on the oxidative modification of low density lipoprotein. J Lipid Res 1992 ; 33 : 899 -906. PubMed 42 Brugger P, Marktl W, Herold M. Impaired nocturnal secretion of melatonin in coronary heart disease. Lancet 1995 ; 345 : 1408 . Crossref Search ADS PubMed 43 Nagtegaal E, Smits M, Swart W, van der Meer G, Kerkhof G. Melatonin secretion and coronary heart disease. Lancet 1995 ; 346 : 1299 . Crossref Search ADS PubMed 44 Pieri C, Marra M, Gáspár R, Damjanovich S. Melatonin protects LDL from oxidation but does not prevent the apolipoprotein derivatization. Biochem Biophys Res Commun 1996 ; 222 : 256 -260. Crossref Search ADS PubMed 45 Kelly MR, Loo G. Melatonin inhibits oxidative modification of human low-density lipoprotein. J Pineal Res 1997 ; 22 : 203 -209. Crossref Search ADS PubMed 46 Abuja PM, Liebmann P, Hayn M, Schauenstein K, Esterbauer H. Antioxidant role of melatonin in lipid peroxidation of human LDL. FEBS Lett 1997 ; 413 : 289 -293. Crossref Search ADS PubMed 47 Seegar H, Mueck AO, Lippert TH. Effect of melatonin and metabolites on copper-mediated oxidation of low density lipoprotein. Br J Clin Pharmacol 1997 ; 44 : 283 -284. PubMed 48 Marshall KA, Reiter RJ, Poeggeler B, Aruoma OI, Halliwell B. Evaluation of the antioxidant activity of melatonin in vitro. Free Radical Biol Med 1996 ; 21 : 307 -315. Crossref Search ADS 49 Aviram M, Fuhrman B, Maor I, Brook GJ. Serotonin increases macrophage uptake of oxidized low density lipoprotein. Eur J Clin Chem Clin Biochem 1992 ; 30 : 55 -61. PubMed 50 Jialal I, Fuller CJ, Huet BA. The effect of α-tocopherol supplementation on LDL oxidation. A dose response study. Arterioscler Thromb 1995 ; 15 : 190 -198. Crossref Search ADS © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Laryea, Maurice, D;Steinhagen,, Folkert;Pawliczek,, Sandra;Wendel,, Udo

doi: 10.1093/clinchem/44.9.1937pmid: N/A

Abstract A simple and convenient method using commercially available derivatization reagents is described for the measurement of betaine and N,N-dimethylglycine (DMG) in blood and urine. Precolumn derivatization of plasma or urine is performed directly in acetonitrile without extraction with p-bromophenacyl bromide and crown ether as catalyst. The p-bromophenacyl ester derivatives are then separated by high-performance liquid chromatography, using an isocratic system of acetonitrile and water containing choline. Effluent was monitored at 254 nm. The limit of detection was 5 μmol/L for betaine and 2 μmol/L for DMG. Analytical recovery was >97% for both analytes. Total and within-day CVs were 2.0–4.4% and 0.9–2.2% for DMG. For betaine, the total and within-day CVs were 1.3–5.3% and 0.4–3.8%, respectively. The method is precise and cost-effective and has been used successfully to determine the concentrations of DMG and betaine in human plasma and urine. Interest in the medical use of betaine [2(N,N,N-trimethylammonium) acetate] has been generated by the knowledge of its importance as an alternative homocysteine methylating agent. It functions as a substrate for the enzyme betaine-homocysteine methyltransferase (EC 2.1.1.5), which catalyzes the remethylation of homocysteine to methionine. Treatment with betaine was effective in the different forms of homocystinuria caused by cystathionine β-synthase (EC 4.2.1.22) and 5,10-methylenetetrahydrofolate reductase (MTHFR; EC 1.1.99.15) deficiencies, inborn errors of the transsulfuration, and one-carbon pathways. Betaine previously has been shown to reduce plasma concentrations of homocysteine and to increase methionine (1)(2)(3)(4) . In MTHFR deficiency, treatment with betaine is most effective in reversing demyelination of the brain and spinal cord (5) . Although large amounts of betaine are often given orally to patients with these metabolic disorders, little is known about its absorption in the gut and its metabolism. The extent of inhibition on betaine homocysteine methyltransferase by the formed N,N-dimethylglycine (DMG) is uncertain. To optimize therapies in these patients, a sensitive, specific, and reliable method appears necessary to monitor the concentrations of betaine and DMG in blood and urine. Problems associated with the isolation, detection, and measurement of quaternary ammonium compounds, including betaines in biological materials, have been reviewed by Gorham (6) . Since then, several methods have been described in the literature for the separation of these naturally occurring compounds. We recently described a method for the determination of betaine and DMG in urine, using ultraviolet absorbance (7) . This method lacks sensitivity, which leads to the need to label the substances with absorbing or fluorescing reagents to improve detection limits. Methods using proton nuclear magnetic resonance (8)(9)(10) require large capital outlay and a high degree of technical expertise. Recently, considerable improvement in the analysis of these compounds was reported by Allen et al. (11) . Their assays are based on an isotope dilution method using gas chromatography–mass spectrometry. However, the method is very laborious because it requires betaine to be converted to DMG by partially purified rat liver betaine homocysteine methyltransferase, which is not commercially available. Here, we report a simple and sensitive isocratic HPLC-UV method for the determination of DMG and betaine in plasma and urine. Materials and Methods chemicals Anhydrous betaine and 18-crown-6 were obtained from Sigma Chemical Co. DMG was purchased from Fluka. p-Bromophenacyl bromide was a product of Pierce. All other reagents and solutions were of analytical grade and purchased from Merck. standards DMG and betaine were dissolved in water at a concentration of 1 mmol/L and diluted with water to the final concentrations used during the analysis. subjects and sample handling Venous blood from five children with homocystinuria (ages, 1–14 years) and 12 healthy volunteers (ages, 30–50 years) was collected into a Vacutainer Tube containing EDTA. Four patients suffered from MTHFR deficiency and were treated with up to 600 mg/kg betaine monohydrate daily; one patient had cystathionine-β-synthase deficiency and received 200 mg/kg betaine. Because the blood samples were also used for the determination of total homocysteine, they were placed on ice after collection, and plasma was obtained without delay by centrifuging the blood samples within 30 min after collection at 2000g for 10 min at room temperature. Spontaneous urine samples were collected into plastic tubes. Plasma and urine samples were stored at −20 °C until analysis, usually within 14 days. Hemolytic and lipemic plasma were also used in the preliminary study for the assay. reagent preparation The derivatizing solution was made by dissolving 66 mg (2.5 mmol) of 18-crown-6 and 1390 mg (50 mmol) of 4-bromophenacyl bromide in 100 mL of acetonitrile. derivatization procedure Urine samples were diluted up to 10-fold with distilled water before assay. Plasma was used without dilution for the assay. To 50 μL of sample or calibrator solution was added 50 μL of 100 mmol/L KH2PO4. After the solution was vortex-mixed, 900 μL of derivatizing solution was added, and the mixing continued. The tubes were capped, vortex-mixed, and heated to 80 °C for 60 min. After the mixture was cooled to room temperature, it was again vortex-mixed and centrifuged at 1000g. Fifteen microliters of the supernatant containing the phenacyl esters of DMG and betaine was injected directly into the HPLC. equipment The HPLC consisted of a Model 501 pump coupled to a Wisp Model 712 autosampler. Detection was with a 490 Programmable Multiwavelength Detector connected to dual channel monitor (Waters Associates) and a Shimadzu integrator CR3A. The column was a SupelcosilTM LC-SCX, 5 μm, 25 cm × 4.6 cm (Supelco Inc.). chromatographic conditions Sample elution was isocratic over 20 min, using a mobile phase containing 22 mmol/L choline in 900 mL/L acetonitrile and 100 mL/L water. The mobile phase was degassed for 30 min in an ultrasonic bath before use. The flow rate was 1.5 mL/min. The detector was set to monitor the analytes at 254 nm. All chromatography was performed at room temperature. Results optimization of the assay The derivatization kinetics were determined at different temperatures, i.e., 20, 40, 60, 80, 90, and 120 °C, for both DMG and betaine. Samples were taken at 10-min intervals for 90 min. The optimum conditions proved to be 60 min at 80 °C. Samples could be derivatized directly without addition of KH2PO4; however, DMG gave a double peak, probably from alkylation of its amino group under the reaction conditions. hplc elution profiles Careful selection of the ionic strength of choline and the amounts of acetonitrile and water in the mobile phase was found to be effective for the separation of DMG and betaine. We determined that 22 mmol/L choline in 900 mL/L acetonitrile allowed resolution of the phenacyl bromide ester derivatives of DMG and betaine from other compounds present in the mixture. Under the conditions chosen, the retention times of the phenacyl bromide esters of DMG and betaine were 12.7 and 14.8 min, respectively (Fig. 1 A). Figure 1. Open in new tabDownload slide Elution profiles of DMG and betaine calibrators (A) and plasma (B) and urine (C) from a patient with homocystinuria, caused by 5,10-methylenetetrahydrofolate deficiency, undergoing therapy with betaine. (B) DMG = 7.8 μmol/L; betaine = 101.7 μmol/L; full scale = 0.01 absorbance units. (C) Full scale = 0.02 absorbance units. Figure 1. Open in new tabDownload slide Elution profiles of DMG and betaine calibrators (A) and plasma (B) and urine (C) from a patient with homocystinuria, caused by 5,10-methylenetetrahydrofolate deficiency, undergoing therapy with betaine. (B) DMG = 7.8 μmol/L; betaine = 101.7 μmol/L; full scale = 0.01 absorbance units. (C) Full scale = 0.02 absorbance units. Chromatograms of plasma and urine of a patient with homocystinuria caused by MTHFR deficiency are shown in Fig. 1 , B and C; Fig. 2 shows plasma of an unaffected subject with and without added DMG. Betaine and DMG peaks in the plasma samples correspond to the retention of the analytes in the calibrator solution. The peaks are well-resolved, with no extraneous substance interfering with the assay. All of the UV-absorbing compounds eluted within 25 min of injection. Peak 1 in the chromatograms of the urine samples was identified as creatinine by retention time and co-chromatography with an authentic standard. Therefore, it is possible to estimate DMG, betaine, and creatinine in the present system. Figure 2. Open in new tabDownload slide Elution profiles of the plasma of a healthy subject (A) and plasma from that same subject (B) with 8.0 μmol/L DMG added. Full scale = 0.01 absorbance units. (A) Betaine = 21.6 μmol/L; DMG = 5.7 μmol/L. Figure 2. Open in new tabDownload slide Elution profiles of the plasma of a healthy subject (A) and plasma from that same subject (B) with 8.0 μmol/L DMG added. Full scale = 0.01 absorbance units. (A) Betaine = 21.6 μmol/L; DMG = 5.7 μmol/L. limit of detection and sensitivity The limits of detection for the assay, defined as four times the signal-to-noise ratio, were determined to be 2 and 5 μmol/L for DMG and betaine, respectively. These concentrations are lower than the basal concentrations of DMG and betaine in human plasma and urine. Sensitivity can be increased by using larger quantities of matrix in the case of plasma or by changing the dilution factor of the urine. Sensitivity could also be enhanced by injecting more sample into the HPLC rather than the 15 μL used here. linearity The linearity of the method was assessed by analyzing DMG and betaine calibrators ranging in concentration from 2 to 200 μmol/L, using 15 μL samples. DMG and betaine were linearly related to peak height, and this relationship was maintained over the range tested. The regression equations (± SD) were: y = 220 (± 2)x − 348 (± 200) μV (r2 = 0.99) for DMG, and y = 121(± 1)x − 185 (± 111) μV (r2 = 0.99) for betaine. recovery Recoveries of DMG and betaine added to urine and plasma samples were 97–101% (Table 1). Table 1. Recoveries of DMG and betaine calibrators from urine and plasma. . Plasma . . . Urine . . . . Amount added, μmol/L . Amount recovered, μmol/L (n = 5) . Recovery, % . Amount added, μmol/L . Amount recovered, μmol/L (n = 3) . Recovery, % . DMG 24.2 24.6 101.5 24.2 23.6 97.5 48.5 48.4 99.9 48.5 49.8 102.7 64.7 64.9 100.4 64.7 64.1 99.1 96.9 98.0 101.1 96.9 95.7 98.8 193.9 194.9 100.5 193.9 193.7 99.9 Betaine 21.3 21.2 99.5 21.3 20.8 97.5 42.7 42.6 99.8 42.7 41.9 98.2 56.9 55.4 97.3 56.9 57.3 100.7 85.5 84.6 98.9 85.5 84.4 98.7 170.7 170.3 99.7 170.7 171.7 100.6 . Plasma . . . Urine . . . . Amount added, μmol/L . Amount recovered, μmol/L (n = 5) . Recovery, % . Amount added, μmol/L . Amount recovered, μmol/L (n = 3) . Recovery, % . DMG 24.2 24.6 101.5 24.2 23.6 97.5 48.5 48.4 99.9 48.5 49.8 102.7 64.7 64.9 100.4 64.7 64.1 99.1 96.9 98.0 101.1 96.9 95.7 98.8 193.9 194.9 100.5 193.9 193.7 99.9 Betaine 21.3 21.2 99.5 21.3 20.8 97.5 42.7 42.6 99.8 42.7 41.9 98.2 56.9 55.4 97.3 56.9 57.3 100.7 85.5 84.6 98.9 85.5 84.4 98.7 170.7 170.3 99.7 170.7 171.7 100.6 Open in new tab Table 1. Recoveries of DMG and betaine calibrators from urine and plasma. . Plasma . . . Urine . . . . Amount added, μmol/L . Amount recovered, μmol/L (n = 5) . Recovery, % . Amount added, μmol/L . Amount recovered, μmol/L (n = 3) . Recovery, % . DMG 24.2 24.6 101.5 24.2 23.6 97.5 48.5 48.4 99.9 48.5 49.8 102.7 64.7 64.9 100.4 64.7 64.1 99.1 96.9 98.0 101.1 96.9 95.7 98.8 193.9 194.9 100.5 193.9 193.7 99.9 Betaine 21.3 21.2 99.5 21.3 20.8 97.5 42.7 42.6 99.8 42.7 41.9 98.2 56.9 55.4 97.3 56.9 57.3 100.7 85.5 84.6 98.9 85.5 84.4 98.7 170.7 170.3 99.7 170.7 171.7 100.6 . Plasma . . . Urine . . . . Amount added, μmol/L . Amount recovered, μmol/L (n = 5) . Recovery, % . Amount added, μmol/L . Amount recovered, μmol/L (n = 3) . Recovery, % . DMG 24.2 24.6 101.5 24.2 23.6 97.5 48.5 48.4 99.9 48.5 49.8 102.7 64.7 64.9 100.4 64.7 64.1 99.1 96.9 98.0 101.1 96.9 95.7 98.8 193.9 194.9 100.5 193.9 193.7 99.9 Betaine 21.3 21.2 99.5 21.3 20.8 97.5 42.7 42.6 99.8 42.7 41.9 98.2 56.9 55.4 97.3 56.9 57.3 100.7 85.5 84.6 98.9 85.5 84.4 98.7 170.7 170.3 99.7 170.7 171.7 100.6 Open in new tab imprecision Total and within-run imprecision (CV) measured on plasma and urine at three concentrations was assessed by analyzing the samples 20 times within 1 day and over 30 separate days (12) . The amounts added to the specimens were chosen to cover the ranges of the calibration curves and to include a specimen of a high value, as encountered in patients being treated with betaine (Table 2). Table 2. Precision of DMG and betaine measurements. Specimen . Mean, μmol/L . Total imprecision (30 days) . . Within-day imprecision (20 runs within 1 day) . . . . SD . CV, % . SD . CV, % . DMG Plasma pool 6.9 0.3 4.2 0.3 2.1 64.71 70.7 1.3 2.0 1.0 2.0 194.01 207.1 4.7 2.3 1.8 0.9 Urine pool 13.6 0.6 4.4 0.3 2.2 64.71 80.2 2.1 2.6 1.8 2.2 194.01 212.5 6.3 3.0 3.2 1.5 Betaine Plasma pool 26.5 1.7 5.1 0.7 2.8 56.92 83.2 1.8 2.2 0.9 1.1 170.72 196.2 2.3 2.3 1.3 0.7 Urine pool 57.7 3.1 5.3 2.2 3.8 56.92 114.9 2.4 2.4 1.6 1.4 170.72 228.7 3.3 1.3 0.8 0.4 Specimen . Mean, μmol/L . Total imprecision (30 days) . . Within-day imprecision (20 runs within 1 day) . . . . SD . CV, % . SD . CV, % . DMG Plasma pool 6.9 0.3 4.2 0.3 2.1 64.71 70.7 1.3 2.0 1.0 2.0 194.01 207.1 4.7 2.3 1.8 0.9 Urine pool 13.6 0.6 4.4 0.3 2.2 64.71 80.2 2.1 2.6 1.8 2.2 194.01 212.5 6.3 3.0 3.2 1.5 Betaine Plasma pool 26.5 1.7 5.1 0.7 2.8 56.92 83.2 1.8 2.2 0.9 1.1 170.72 196.2 2.3 2.3 1.3 0.7 Urine pool 57.7 3.1 5.3 2.2 3.8 56.92 114.9 2.4 2.4 1.6 1.4 170.72 228.7 3.3 1.3 0.8 0.4 1 Added DMG, μmol/L. 2 Added betaine, μmol/L. Open in new tab Table 2. Precision of DMG and betaine measurements. Specimen . Mean, μmol/L . Total imprecision (30 days) . . Within-day imprecision (20 runs within 1 day) . . . . SD . CV, % . SD . CV, % . DMG Plasma pool 6.9 0.3 4.2 0.3 2.1 64.71 70.7 1.3 2.0 1.0 2.0 194.01 207.1 4.7 2.3 1.8 0.9 Urine pool 13.6 0.6 4.4 0.3 2.2 64.71 80.2 2.1 2.6 1.8 2.2 194.01 212.5 6.3 3.0 3.2 1.5 Betaine Plasma pool 26.5 1.7 5.1 0.7 2.8 56.92 83.2 1.8 2.2 0.9 1.1 170.72 196.2 2.3 2.3 1.3 0.7 Urine pool 57.7 3.1 5.3 2.2 3.8 56.92 114.9 2.4 2.4 1.6 1.4 170.72 228.7 3.3 1.3 0.8 0.4 Specimen . Mean, μmol/L . Total imprecision (30 days) . . Within-day imprecision (20 runs within 1 day) . . . . SD . CV, % . SD . CV, % . DMG Plasma pool 6.9 0.3 4.2 0.3 2.1 64.71 70.7 1.3 2.0 1.0 2.0 194.01 207.1 4.7 2.3 1.8 0.9 Urine pool 13.6 0.6 4.4 0.3 2.2 64.71 80.2 2.1 2.6 1.8 2.2 194.01 212.5 6.3 3.0 3.2 1.5 Betaine Plasma pool 26.5 1.7 5.1 0.7 2.8 56.92 83.2 1.8 2.2 0.9 1.1 170.72 196.2 2.3 2.3 1.3 0.7 Urine pool 57.7 3.1 5.3 2.2 3.8 56.92 114.9 2.4 2.4 1.6 1.4 170.72 228.7 3.3 1.3 0.8 0.4 1 Added DMG, μmol/L. 2 Added betaine, μmol/L. Open in new tab internal standards No internal standard was used in this assay. Commercially available substances such as sulfonobetaine, trigonelline, and others with similar structures and retention times appeared from proton nuclear magnetic resonance spectrometry to be present in urine and probably plasma samples. Direct derivatization without sample preparation, however, allowed the peak sizes to be monitored efficiently with external standards. measurement of reference and high values Plasma The concentration ranges of DMG and betaine in the 12 healthy subjects were 4–13 μmol/L and 20–144 μmol/L, respectively. In patients being treated with betaine, the ranges were from 8 to 228 μmol/L for DMG and from 20 to 2680 μmol/L for betaine. Urine The range in the healthy subjects was from 0.8 to 11.6 mmol/mol creatinine for DMG and from 6.4 to 92.7 mmol/mol creatinine for betaine. In urine of betaine-treated patients, the concentrations were increased up to 3.6 mol/mol creatinine for DMG and up to 20.8 mol/mol creatinine for betaine. Discussion The ability of phenacyl bromide to react with carboxyl groups has been known for some time (13) . Recently, van Kempen et al. (14) and Gorham et al. ((15)) demonstrated that this reagent reacts with quaternary ammonium compounds and can be used for their quantification. We have adapted and modified these methods for the determination of DMG and betaine in plasma and urine. Our goal was to develop a method for clinical use that would be simple and rapid, would use commercially available reagents, would have a one-step sample preparation, would involve isocratic elution with a single column, would allow simultaneous analyses of DMG and betaine in plasma and urine, and would require only a small sample. The present method appears to have certain advantages over previously reported HPLC methods, principally because most of the interfering substances do not have to be removed before derivatization (7)(14)(15) . This reduces losses of DMG and betaine and the time required for analysis. The absence of this critical step leads to satisfactory criteria of reproducibility and repeatability. The limits of detection were 2 μmol/L for DMG and 5 μmol/L for betaine. Recovery was >97%. Moreover, it is also possible to estimate creatinine in urine samples in the same assay using our system, an advantage for the calculation of urine values related to creatinine. The method also avoids gradient elution, which requires a sophisticated HPLC apparatus. Samples are usually analyzed on the day of derivatization. No detectable losses were found during the course of a working day or in derivatized samples stored at 4 °C for a day or at −20 °C for a week. This procedure was used to determine betaine and DMG concentrations in plasma and urine samples from a limited number of healthy individuals and from patients receiving betaine monohydrate in doses of 200–600 mg/kg body weight per day. Interestingly, in healthy subjects we found somewhat wider ranges for betaine in plasma and urine than Allen et al. (11) , with 20–144 vs 17.6–73.3 μmol/L and 6.4–92.7 vs 2.3–55.9 mmol/mol creatinine. These differences may reflect methodological differences or different dietary intakes. Ease of derivative formation, coupled with a simple chromatographic separation, increases the potential of this method for routine analysis of DMG and betaine in plasma and urine. Heinrich-Heine University, Children’s Hospital, Metabolic Unit, Moorenstrasse 5, D-40225 Düsseldorf, Germany. References 1 Wilcken DE, Dudman NP, Tyrrell PA. Homocystinuria due to cystathionine beta-synthase deficiency–the effects of betaine treatment in pyridoxine-responsive patients. Metabolism 1985 ; 34 : 1115 -1121. Crossref Search ADS PubMed 2 Wendel U, Bremer HJ. Betaine in the treatment of homocystinuria due to 5,10-methylenetetrahydrofolate reductase deficiency. Eur J Pediatr 1984 ; 142 : 147 -150. Crossref Search ADS PubMed 3 Holme E, Kjellman B, Ronge E. Betaine for treatment of homocystinuria caused by methylenetetrahydrofolate reductase deficiency. Arch Dis Child 1989 ; 64 : 1061 -1064. Crossref Search ADS PubMed 4 Smolin LA, Benevenga NJ, Berlow S. The use of betaine for the treatment of homocystinuria. J Pediatr 1981 ; 99 : 467 -472. Crossref Search ADS PubMed 5 Surtees R, Leonard J, Austin S. Association of demyelination with deficiency of cerebrospinal-fluid S-adenosylmethionine in inborn errors of methyl-transfer pathway. Lancet 1991 ; 338 : 1550 -1554. Crossref Search ADS PubMed 6 Gorham J. Quantitative analysis of quaternary ammonium compounds by ion exchange and ion pair high performance liquid chromatography. Williams PA Hudson MJ eds. Recent developments in ion exchange 1987 : 79 -86 Elsevier Applied Science London. . 7 Laryea MD, Zass R, Ritgen J, Wendel U. Simultaneous determination of betaine and N,N-dimethylglycine in urine. Clin Chim Acta 1994 ; 230 : 169 -175. Crossref Search ADS PubMed 8 Wevers RA, Engelke U, Heerschap A. High-resolution 1H-NMR spectroscopy of blood plasma for metabolic studies. Clin Chem 1994 ; 40 : 1245 -1250. Crossref Search ADS PubMed 9 Davies SE, Chalmers RA, Randall EW, Iles RA. Betaine metabolism in human neonates and developing rats. Clin Chim Acta 1988 ; 178 : 241 -249. Crossref Search ADS PubMed 10 Dudman NP, Wilcken DE, Wang J, Lynch JF, Macey D, Lundberg P. Disordered methionine/homocysteine metabolism in premature vascular disease. Its occurrence, cofactor therapy, and enzymology. Arterioscler Thromb 1993 ; 13 : 1253 -1260. Crossref Search ADS PubMed 11 Allen RH, Stabler SP, Lindenbaum J. Serum betaine, N,N-dimethylglycine and N-methylglycine levels in patients with cobalamin and folate deficiency and related inborn errors of metabolism. Metabolism 1993 ; 42 : 1448 -1460. Crossref Search ADS PubMed 12 Koch DO, Peters T, Jr. Selection and evaluation of methods. Kaplan L Pesce A eds. Tietz textbook of clinical chemistry 2nd ed. 1994 : 508 -525 WB Saunders Philadelphia. . 13 Durst HD, Milano M, Kitka EJ, Connely SA, Grushka E. Phenacyl esters of fatty acids via crown ether catalysts for enhanced ultraviolet detection in liquid chromatography. Anal Chem 1975 ; 47 : 1797 -1801. Crossref Search ADS PubMed 14 van Kempen TA, Odle J. Quantification of carnitine esters by high-performance liquid chromatography. Effect of feeding medium-chain triglycerides on the plasma carnitine ester profile. J Chromatogr 1992 ; 584 : 157 -165. Crossref Search ADS PubMed 15 Gorham J, Mcdonnell E, Wyn Jones RG. Determination of betaines as ultraviolet-absorbing esters. Anal Chim Acta 1982 ; 138 : 277 -283. Crossref Search ADS © 1998 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)