Clinical Chemistry

Sweetman,, Lawrence

doi: 10.1093/clinchem/47.11.1937pmid: N/A

Major expansion of newborn screening for inherited metabolic disorders is taking place across the US and around the world as newer analytical technology is applied. Historically, each disorder to be screened required a separate test with associated costs and requirement for a portion of the dried-blood-spot specimen from a heel stick. This limitation of the existing tests was partially responsible for the limitation of mandated newborn screening in the US to a small number of disorders (usually three to seven, depending on the state). The technique of tandem mass spectrometric (MS) analysis of dried-blood spots was first proposed for newborn screening in 1990 by Millington et al. (1). Using ionization techniques of fast atom bombardment or liquid secondary ionization with tandem MS, they simultaneously determined a large number of acylcarnitines as an acylcarnitine profile. This allowed newborn screening for numerous inherited fatty acid oxidation and organic acid disorders by a single procedure. Tandem MS was extended to amino acids, including phenylalanine, the screening target for detection of the phenylketonuria (PKU) test (2)(3)(4), and to other disorders, with several such tests described in these pages during the last 8 years (3)(5)(6)(7)(8)(9). The development and application of electrospray ionization (ESI) tandem MS with its ability to be automated made high-volume tandem MS screening for amino acid, organic acid, and fatty acid metabolic disorders practical by the mid-1990s (10)(11)(12). Automated ESI tandem MS newborn screening of amino acids and acylcarnitines extracted from a single punch of a dried-blood spot, with stable-isotope-labeled internal standards and derivatization to butyl esters, can be performed with 2–4 min of instrument time. This single test is capable of detecting ∼20–40 inherited metabolic disorders, depending on which analytes are measured and how different disorders are defined (10)(11)(12)(13). The number of disorders is based on experience with detection of increased metabolites in blood from children affected with these disorders, although not all have been demonstrated to be detectable in the neonate. The large increase in the number of inherited metabolic disorders detectable in the newborn period because of tandem MS screening greatly extends the possibilities of early, generally presymptomatic, diagnosis and treatment to minimize morbidity and mortality for many affected children. The aggregate incidence of disorders detectable by tandem MS (including PKU and other amino acid disorders that are currently screened for in some states) was reportedly ∼1 in 3400 newborns for 168 000 newborns screened in Pennsylvania and North Carolina [summarized in (14)], 1 in 4700 for 137 000 babies screened in Australia (13), and 1 in 3800 in >160 000 babies screened in New England by Zytkovicz et al. (15) as described in this issue of Clinical Chemistry. These initial publications of the experiences of large-scale newborn screening for metabolic disorders by tandem MS also emphasize some of the complications of this expansion: (a) establishment of appropriate cutoffs for the large number of analytes to minimize false negatives without an excessive number of false positives; and (b) determination of a definitive diagnosis for an infant with an abnormal increase detected by tandem MS screening. False positives cause parental anxiety and are expensive in terms of professional time and effort to obtain repeat specimens for retesting and follow-up (16). The first publication of a computer algorithm for tandem MS newborn screening set the upper limit of the reference interval at the 99.5 percentile, which would give 0.5% positives for each analyte, which is reasonable for a single analyte test and disorder (12). When screening for 20–40 disorders (analytes) in a single test (assuming that each is independent of the others), the aggregate incidence of false positives would be as high as 18%, whereas the true positives would be ∼0.03%. To minimize false positives, Zytkovicz et al. (15) generally set the abnormal flags at the 99.98 percentile, so that only 0.02% would be flagged positive for each analyte, making the false-positive rate (for ∼20 independent tests) ∼0.4% with a true-positive rate of 0.03%. This minimizes false positives but raises some concern about possible false negatives because these flagging values were 5–13 SD above the normal mean for the analytes. Zytkovicz et al. (15) point out that different cutpoints need to be established for different subpopulations of newborns. Only 5% of their newborns had very low birth weights or were in neonatal intensive care units, but they accounted for 50% of the false positives. Optimal discrimination may require determination of cutpoints for a larger number of subpopulations. These might include full-term, premature or low birth weight, very low birth weight, very low birth weight on total parenteral nutrition, specimen collection at <24 h of age (early discharge), and collection at 1–3 days of age. Additionally, blood concentrations of some analytes change with age, requiring different decision limits for repeat specimens obtained many days or weeks after birth to investigate a positive initial result in the neonate. Another challenge illustrated in the report by Zytkovicz et al. (15) is confirmation of a diagnosis suggested by an abnormal screening result. Tandem MS is sometimes considered to provide definitive diagnoses, as is true for some disorders. In medium-chain acyl-CoA dehydrogenase deficiency (MCAD), the pattern of increased acylcarnitines is diagnostic. Even in this case, molecular analyses of mutations in the MCAD gene suggest that different mutations can produce different degrees of metabolite increases in affected babies (17). Many of the abnormal analyte increases found in tandem MS screening are not pathognomonic of a single disorder and can be produced by several different genetic disorders. In these cases, additional, more-specific diagnostic tests are required. Some of the lack of specificity of the tandem MS results reflects the inability of the technique to distinguish isobaric compounds. For example, isobutyrylcarnitine and butyrylcarnitine have the same mass and are detected as one analyte (C4-acylcarnitine). Increased C4-acylcarnitine may reflect either of two very different disorders: isobutyryl-CoA dehydrogenase deficiency in the catabolic pathway of valine (increased isobutyrylcarnitine) or short-chain acyl-CoA dehydrogenase deficiency in fatty acid β-oxidation (increased butyrylcarnitine). Differential diagnosis requires additional tests, such as urinary organic acids, DNA mutation analysis, determination of the integrity of different metabolic pathways in intact cells, or assay of individual enzyme activities in cells. In other cases, different inherited disorders produce increases in the same abnormal compound. An example is increased propionylcarnitine (C3-acylcarnitine), a finding that requires differentiation of several disorders: propionyl-CoA carboxylase deficiency, methylmalonyl-CoA mutase deficiency, several cobalamin disorders, and even dietary deficiency of vitamin B12. This requires the expertise of clinical and biochemical geneticists and a variety of diagnostic tests. The experience presented by Zytkovicz et al. (15) in this issue of Clinical Chemistry, as well as the experience of others (12)(13), is valuable for pointing out some of the complications that need to be addressed as tandem MS newborn screening is implemented in more and more states for an increasing number of metabolic disorders. Resolving the issues of establishing appropriate reference intervals for subpopulations of newborns, determining limits to minimize false positives and false negatives, developing protocols for the differential diagnosis of suspected disorders, compilation of incidences of disorders, and documentation of outcomes for validation of the programs will require the sharing of information among screening laboratories and their genetic consultants. This goal is being encouraged and greatly assisted by the collaborative efforts of the Genetic Services Branch of the Maternal and Child Health Bureau in the Health Resources and Services Administration (HRSA) in Washington, the CDC in Atlanta, and the National Newborn Screening and Genetic Resource Center (NNSGRC) in Austin (18). References 1 Millington DS, Kodo N, Norwood DL, Roe CR. Tandem mass spectrometry: a new method for acylcarnitine profiling with potential for neonatal screening for inborn errors of metabolism. J Inherit Metab Dis 1990 ; 13 : 321 -324. Crossref Search ADS PubMed 2 Millington DS, Kodo N, Terada N, Roe D, Chase DH. The analysis of diagnostic markers of genetic disorders to human blood and urine using tandem mass spectrometry with liquid secondary ion mass spectrometry. Int J Mass Spectrom Ion Proc 1991 ; 111 : 211 -228. Crossref Search ADS 3 Chace DH, Millington DS, Terada N, Kahler SG, Roe CR, Hofman LF. Rapid diagnosis of phenylketonuria by quantitative analysis for phenylalanine and tyrosine in neonatal blood spots by tandem mass spectrometry. Clin Chem 1993 ; 39 : 66 -71. Crossref Search ADS PubMed 4 Chace DH, Sherwin JH, Hillman SL, Lorey F, Cunningham GC. Use of phenylalanine-to-tyrosine ratio determined by tandem mass spectrometry to improve newborn screening for phenylketonuria of early discharge specimens collected in the first 24 h. Clin Chem 1998 ; 44 : 2405 -2409. Crossref Search ADS PubMed 5 Chace DH, Hillman SL, Millington DS, Kahler SG, Roe CR, Naylor EW. Rapid diagnosis of maple syrup urine disease in blood spots from newborns by tandem mass spectrometry. Clin Chem 1995 ; 41 : 62 -68. Crossref Search ADS PubMed 6 Chace DH, Hillman SL, Millington DS, Kahler SG, Adam BW, Levy HL. Rapid diagnosis of homocystinuria and other hypermethioninemias from newborns’ blood spots by tandem mass spectrometry. Clin Chem 1996 ; 42 : 349 -355. Crossref Search ADS PubMed 7 Chace DH, Hillman SL, Van Hove JLK, Naylor EW. Rapid diagnosis of MCAD deficiency: quantitative analysis of octanoylcarnitine and other acylcarnitines in newborn blood spots by tandem mass spectrometry. Clin Chem 1997 ; 43 : 2106 -2113. Crossref Search ADS PubMed 8 Jones PM, Quinn R, Fennessey PV, Tjoa S, Goodman SI, Fiore S, et al. Improved stable isotope dilution-gas chromatography-mass spectrometry method for serum or plasma free 3-hydroxy-fatty acids and its utility for the study of disorders of mitochondrial fatty acid β-oxidation. Clin Chem 2000 ; 46 : 149 -155. Crossref Search ADS PubMed 9 Ito T, van Kuilenburg ABP, Bootsma AH, Haasnoot AJ, van Cruchten A, Wada Y, van Gennip AH. Rapid screening of high-risk patients for disorders of purine and pyrimidine metabolism using HPLC-electrospray tandem mass spectrometry of liquid urine or urine-soaked filter paper strips. Clin Chem 2000 ; 46 : 445 -452. Crossref Search ADS PubMed 10 Rashed MS, Ozand PT, Bucknall MP, Little D. Diagnosis of inborn errors of metabolism from blood spots by acylcarnitines and amino acids profiling using automated electrospray tandem mass spectrometry. Pediatr Res 1995 ; 38 : 324 -331. Crossref Search ADS PubMed 11 Johnson AW, Mills K, Clayton PT. The use of automated electrospray ionization tandem MS for the diagnosis of inborn errors of metabolism from dried blood spots. Biochem Soc Trans 1996 ; 24 : 932 -938. Crossref Search ADS PubMed 12 Rashed MS, Bucknall MP, Little D, Awad A, Jacob M, Alamoudi M, et al. Screening blood spots for inborn errors of metabolism by electrospray tandem mass spectrometry with a microplate batch process and a computer algorithm for automated flagging of abnormal profiles. Clin Chem 1997 ; 43 : 1129 -1141. Crossref Search ADS PubMed 13 Wiley V, Carpenter K, Wilcken B. Newborn screening with tandem mass spectrometry 12 months’ experience in NSW Australia. Acta Paediatr Suppl 1999 ; 88 : 48 -51. 14 Sweetman L. Newborn screening by tandem mass spectrometry (MS-MS). Clin Chem 1996 ; 42 : 345 -346. Crossref Search ADS PubMed 15 Zytkovicz TH, Fitzgerald EF, Marsden D, Larson CA, Shih VE, Johnson DM, et al. Tandem mass spectrometric analysis for amino, organic, and fatty acid disorders in newborn dried blood spots: a two-year summary from the New England Newborn Screening Program. Clin Chem 2001 ; 47 : 1945 -1955. Crossref Search ADS PubMed 16 Levy HL. Newborn screening by tandem mass spectrometry: a new era. Clin Chem 1998 ; 44 : 2401 -2402. Crossref Search ADS PubMed 17 Andresen BS, Dobrowolski SF, O’Reilly L, Muenzer J, McCandless SE, Frazier DM, et al. Medium-chain acyl-CoA dehydrogenase (MCAD) mutations identified by MS/MS based prospective screening of newborns differ from those observed in patients with clinical symptoms: identification and characterization of a new, prevalent mutation that results in mild MCAD deficiency. Am J Hum Genet 2001 ; 68 : 1408 -1418. Crossref Search ADS PubMed 18 National Newborn Screening and Genetics Resource Center. http://genes-r-us.uthscsa.edu (Accessed September 22, 2001).. © 2001 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)

Ruiz,, Agustín;Antiñolo,, Guillermo;Marcos,, Irene;Borrego,, Salud

doi: 10.1093/clinchem/47.11.1939pmid: N/A

Abstract Background: The multiple endocrine neoplasia 2 (MEN 2) syndromes [MEN 2A, MEN 2B, and familial medullary thyroid carcinoma (FMTC)] are caused by germline mutations of the RET protooncogene. Because 85% of MEN 2A patients and 30% of FMTC patients have mutations at codon 634, the recommended molecular analyses begin at exon 11, where codon 634 is located. Methods: We scanned codon 634 of the RET protooncogene with real-time PCR and fluorescence resonance energy transfer (FRET), using a unique pair of internal probes to detect mutations localized at codon 634. We compared results with sequencing results in 66 patients. Results: The method detected all codon 634 mutations available in our laboratory (Cys634Tyr, Cys634Arg, Cys634Phe, Cys634Trp). Comparing this method with the direct sequencing of exon 11 in a cohort of 66 patients with MTC, the system identified all 14 MTC patients carrying germline mutations at codon 634. One apparent false-positive result occurred among 52 patients. Conclusions: The simultaneous scanning of multiple mutations is possible with the FRET system. The method allows rapid characterization of germline mutations at codon 634 in MTC patients. Medullary thyroid carcinoma (MTC) 1 represents ∼5–10% of malignant thyroid tumors. Approximately 75% of all MTC cases are sporadic, and 25% are grouped in a hereditary cancer syndrome known as multiple endocrine neoplasia 2 (MEN 2). MEN 2 is an autosomal dominant that is associated with mutations in the RET protooncogene. MEN 2 includes MEN 2A and 2B, and familial MTC (FMTC). Inter- and intrafamilial variation is seen in the age of presentation and the incidence of pheochromocytoma and hyperparathyroidism. Classically, it was thought that the penetrance was 70% at the age of 70 (1). The RET protooncogene, localized in 10q11.2, encodes a receptor with tyrosine-kinase activity (2)(3)(4). The genomic structure of RET contains 21 exons distributed throughout 60 kilobases (5). Data collected by the International RET Mutation Consortium (6) indicate that in 92% of MEN 2 cases, germline mutations are found in six exons of the RET gene: 10, 11, 13, 14, 15, and 16. In MEN 2A, the most common form of MEN 2, 98% of patients have germline mutations involving five cysteine codons partially encoding the extracellular domain of the protein (codons 609, 611, 618, 620, and 634). Furthermore, in MEN 2A, 85% of the families have a mutation at codon 634 (6)(7)(8). In classic FMTC, 88% of the cases carry a germline mutation in RET, 30% of the cases of FMTC having a mutation at codon 634 (6). Therefore, the highest frequency of mutations associated with familial cases of MTC appears at codon 634. The rest of the mutations associated with MEN 2A and FMTC are found distributed at the other four cysteine codons mentioned above (609, 611, 618, and 620, exon 10). Subsequently, other mutations localized at other cysteine codons (630 and 631) and noncysteinic codons of the intracellular tyrosine-kinase domain of the RET protein have been reported (codons 768, 790, 791, 804, and 891) (1). Increasing knowledge in recent years about the molecular basis of MEN 2 syndromes has been translated into clinical practice (9). The results of genetic analyses produce dramatic alterations of the course of the disease because at-risk individuals positive for a mutation associated with MEN 2 can undergo prophylactic thyroidectomy, which is believed to be life-saving. At present, no single-step method exists to carry out a molecular analysis of the hotspots related to MEN 2A and FMTC that can be applied to routine diagnosis. In most clinical laboratories, the direct sequencing of exons containing hotspots of the RET protooncogene is routinely performed in all MTC patients (except for cases in which MEN 2B is suspected, where a direct analysis of the M918T mutation is the first step of the molecular evaluation). This approach, although highly sensitive, is tedious, expensive, and awkward when dealing with large numbers of MTC cases. In this report, we present the use of real-time PCR and fluorescence resonance energy transfer (FRET) (10) to simultaneously screen the alterations of codon 634 of the RET protooncogene in a cohort of 66 MTC patients. Materials and Methods patients Patients (n = 66) diagnosed with and treated for MTC between 1994 and 2000 in Andalucia (southern Spain) were included in this study. Informed consent was obtained from all. In all cases, we performed a systematic search for mutations using semiautomatic sequencing of exons 10, 11, 13, 14, 15, and 16 of the RET protooncogene as routine scanning of molecular alterations related to the familial forms of MTC, following the protocols reported previously (7)(8). dna extraction We obtained 15 mL of peripheral blood from all patients to isolate germline DNA from leukocytes. DNA extraction was performed according to standard protocols, as described elsewhere (11). For the DNA, we measured absorbances at 260 and 280 nm to estimate the concentration. We prepared aliquots of DNA at a concentration of 25 ng/μL. The rest of the stock was cryopreserved at −80 °C. primers We selected amplification primers for the PCR of exon 11 of the RET protooncogene using the PRIME command of the Wisconsin GCG package (Genetics Computer Group, Inc.), following the manufacturer’s instructions. The DNA sequence used to carry out this study corresponds to the partial genomic sequence of the 3′ end of the RET protooncogene (GenBank no. AJ243297). The selected pair of primers amplified a segment of 448 bp, corresponding to positions 14782–14800 (forward primer, taggagggggcagtaaatg) and 15210–15227 (reverse primer, ggatcttgaaggcatccac) according to GenBank file AJ243297 (Fig. 1 ). The RETex11 anchor probe (aagcctcacaccacccccacccacagat) was 3′ labeled with fluorescein, and RETex11 sensor (actgtgcgacgagctgtgccgc) was 5′ labeled with LC-Red640 dye and 3′ phosphorylated as indicated (Fig. 1 ). Figure 1. Open in new tabDownload slide Schematic representation of the FRET system designed to test codon 634 of the RET protooncogene. The scanned region is covered by the sensor probe (RETex11 Sensor). Underlined sequences represent the amplification primers. Figure 1. Open in new tabDownload slide Schematic representation of the FRET system designed to test codon 634 of the RET protooncogene. The scanned region is covered by the sensor probe (RETex11 Sensor). Underlined sequences represent the amplification primers. pcr conditions PCR was performed in the LightCycler® system (Roche). PCR was performed to amplify the segment of the RET protooncogene that flanks codon 634 (Fig. 1 ) at a final volume of 10 μL using 25 ng of genomic DNA, 1 μM each amplification primer, 4.4 mM MgCl2, 0.2 μM each detection probe (RETex11 anchor and sensor probes; Fig. 1 ), and 1 μL of LC Master Hybridization Probes Mix (Roche). We used an initial denaturation step of 95 °C for 2 min followed by 55 cycles of 95 °C for 0 s, 66 °C for 15 s, and 72 °C for 15 s. To check the specificity of the PCR products obtained, a conventional electrophoresis of amplified samples was routinely performed (Fig. 2 ). Figure 2. Open in new tabDownload slide Melting experiment of amplicons (A), monitoring of DNA amplification in real-time using FRET probes (B), and conventional electrophoresis (C). (A), melting experiment of amplicons obtained with primers designed to amplify the genomic DNA region containing exon 11 of the RET protooncogene and SYBR Green I fluorochrome. A unique DNA species with a Tm of 93 °C was observed. (B), homogeneous and exponential increases of fluorescence can be observed in cycle 32 in DNA samples. (C), conventional electrophoresis of the PCR products obtained during the amplification in the LightCycler. A unique 442-bp band can be observed in each PCR reaction. Figure 2. Open in new tabDownload slide Melting experiment of amplicons (A), monitoring of DNA amplification in real-time using FRET probes (B), and conventional electrophoresis (C). (A), melting experiment of amplicons obtained with primers designed to amplify the genomic DNA region containing exon 11 of the RET protooncogene and SYBR Green I fluorochrome. A unique DNA species with a Tm of 93 °C was observed. (B), homogeneous and exponential increases of fluorescence can be observed in cycle 32 in DNA samples. (C), conventional electrophoresis of the PCR products obtained during the amplification in the LightCycler. A unique 442-bp band can be observed in each PCR reaction. positive controls We molecularly characterized a group of four independent controls by sequencing and rechecking with PCR restriction fragment-length polymorphism analysis. The positive heterozygote controls corresponded to four mutations that affect codon 634 of the RET protooncogene in MTC patients (Cys634Tyr, Cys634Arg, Cys634Phe, Cys634Trp). fret melting curves The probes (Fig. 1 ) were purchased from TIB MOLBIOL (Berlin). The final conditions to obtain the specific melting curve were 95 °C for 20 s, 72 °C for 25 s, 70 °C for 25 s, 50 °C for 0 s, and 95 °C for 0 s [with a temperature-transfer speed of 0.5 °C/s in each step, except the first step (20 °C/s) and the last step, in which the speed of temperature transfer was 0.4 °C/s]. In the last step, a continuous fluorometric register was performed (F2/F1), fixing the gains of the system at 1, 30, and 30 on channels F1, F2, and F3, respectively. Results Specific amplification of the DNA sequences corresponding to the sequences of the internal probes was indicated by an increase of fluorescence in channel F2 (corresponding to the emission of fluorochrome RED640, which is activated during the FRET process; Fig. 2 ) in the LightCycler system. Specific amplification was monitored in this manner for all samples tested. In addition, when we checked the PCR product using SYBR® Green I fluorochrome and melting analysis in the LightCycler system, a unique DNA species melting at 93 °C could be observed (Fig. 2 ). We performed mutation scanning for specific mutations at codon 634 of the RET protooncogene using fluorescence monitoring during the melting experiment. This process consisted of a temperature ramp coupled to a continuous fluorometric register. With the short-cooling melting procedure (95 °C for 0 s, 50 °C for 0 s, 95 °C for 0 s with a temperature-transfer speed of 0.4 °C/s in the last step), the results of the FRET experiment suggested that the probes do not hybridize well under these conditions. Therefore, we designed a new melting experiment in which several conditions were modified and new incubation steps were included for the anchor and sensor probes at the theoretical annealing temperature (see Materials and Methods for details). Fluorescence melting peak analysis revealed a high melting transition with a melting temperature (Tm) at 75 °C, corresponding to the wild-type sequence. To determine whether FRET can detect codon 634 mutations, we genotyped individuals known to carry codon 634 mutations (Cys→Arg, Cys→Phe, Cys→Tyr, and Cys→Trp). Our results show that the system designed detects all of the mutations in the sequence mentioned above, with specific Tms for each variant (Fig. 3 ). These results seem to show that the FRET system could be used to scan the Cys634 mutation hotspot without the need to design a specific probe for each mutation. Figure 3. Open in new tabDownload slide Tms of four different mutations at codon 634 of the RET protooncogene. Each DNA mutation has an specific melting point using the probe design reported in this work: Cys634Arg, 72 °C; Cys634Tyr, 70.3 °C; Cys634Phe, 68.8 °C; and Cys634Trp, 67.8 °C. The minimum shift of Tm between the wild-type allele and any mutation (Cys634Arg) is 2.5 °C. Figure 3. Open in new tabDownload slide Tms of four different mutations at codon 634 of the RET protooncogene. Each DNA mutation has an specific melting point using the probe design reported in this work: Cys634Arg, 72 °C; Cys634Tyr, 70.3 °C; Cys634Phe, 68.8 °C; and Cys634Trp, 67.8 °C. The minimum shift of Tm between the wild-type allele and any mutation (Cys634Arg) is 2.5 °C. To evaluate our method against the gold-standard mutation-detection technique of direct sequencing, we blindly tested a cohort of 66 MTC patients who had been studied previously in our laboratory with direct sequencing of PCR products of exon 11 of the RET protooncogene. Our method took ∼120 min to reliably analyze codon 634 of the RET protooncogene in all 66 samples. Among these 66, 14 (21%) were known to be mutation positive at codon 634, and FRET detected all 14 mutations. Only 1 false positive (1 of 52 known mutation negative cases) was detected during the study. We believe that the false positive was caused by low yields of PCR product during amplification, a problem that can be avoided by double checking all samples in each study. FRET analysis has only two steps, DNA extraction and real-time PCR, whereas sequencing protocols involve a multistep and time-consuming processing of the samples. Discussion With the increasing demand for genetic testing (12), real-time PCR in combination with FRET has been used to genotype various gene mutations (13)(14)(15)(16)(17)(18)(19). The technology can simultaneously scan the spectrum of point mutations occurring in codons 12, 13, and 61 of Nras, using a combination of fluorescence melting curve analyis and multicolor fluorometry (20). It has not, however, been used to simultaneously scan multiple mutations at a mutational hotspot, despite its versatility (13)(21). We designed a multistep melting experiment to resolve the codon 634 mutations of the protooncogene RET. To favor the hybridization of the probes, we included new incubation steps for the anchor and sensor probes at the theoretical annealing temperatures (72 °C and 70 °C, respectively). A slow ramp speed during the annealing phase and new annealing steps during the melting process can enhance the signal intensity compared with one-step annealing (20)(22). Our system detects the four codon 634 mutations available in our laboratory. The differences in Tm between the wild-type allele and mutant alleles correlate with the mismatch stability (G:T Arg and Tyr, G:A Phe, and C:C Trp). A shorter detection probe may facilitate discrimination between the Arg634 and Cys634 alleles (less destabilizing mismatch) (23). Our system differs greatly from the template-directed dye-terminator incorporation method that has been successfully applied for the detection of RET hotspot mutations (24). In template-directed dye-terminator incorporation, each mutation is detected by a specific primer and the PCR products must be processed (24)(25). In contrast, the method described here uses only one pair of probes that detects all possible changes in the positions in which the sensor probe hybridizes (Fig. 1 ), and no post-PCR processing is needed because the melting is coupled to the PCR program. Because RET testing is part of the standard of care in MEN 2, many clinical laboratories use the perceived gold standard of sequencing (1). In general, sequencing is labor-intensive and expensive, even in the semiautomated setting (26). We believe that our method could be used as an initial screen of the RET codon 634 hotspot. However, given that this technique is relatively new, we would still recommend that when a variant is detected, reanalysis of the sample, either with specific digestion (PCR-restriction fragment length polymorphism analysis) or with direct sequencing, should be performed before a clinical result is issued. We suspect that given further experience with this technique, the secondary confirmation may no longer be necessary. If no alteration is detected during the melting experiment, it may be possible to issue a negative result for this hotspot without additional molecular tests. Because 75% of MTC cases are sporadic, this would reduce both response time and labor. We cannot demonstrate that FRET detects every mutation at codon 634 because we did not test all possible mutations. However, taking into account the performance on our known mutation set, as well as our study of 66 consecutive MTC cases, we feel that FRET will detect all variants. In conclusion, our study suggests that the combined use of FRET and real-time PCR is feasible as a scanning method for small regions of DNA. Our pilot study has shown the successful application of FRET to a hotspot for mutations in RET. The approach can likely be adopted for other disease hotspot sites in other genes. We are deeply grateful to the MTC patients and their families for participation in this study. We are very grateful to Charis Eng, who provided invaluable help and comments on the manuscript. We also thank TIB MOLBIOL for their help. Matilde Romero provided technical assistance. This study was partially funded by the Fondo de Investigación Sanitaria (FIS 01/0551) and Consejería de Salud/Comunidad Autónoma de Andalucía (CAA 47/99 and CAA 116/00), Spain. I.M. is the recipient of a fellowship from the Instituto de Salud Carlos III (Grant 99/4250; Ministerio de Sanidad y Consumo, Spain). 1 " Nonstandard abbreviations: MEN 2, multiple endocrine neoplasia 2; MTC, medullary thyroid carcinoma; FMTC, familial MTC; FRET, fluorescence resonance energy transfer; and Tm, melting temperature. References 1 Eng C. RET proto-oncogene in the development of human cancer [Review]. J Clin Oncol 1999 ; 17 : 380 -393. Crossref Search ADS PubMed 2 Gardner E, Papi L, Easton DF, Cummings T, Jackson CE, Kaplan M, et al. Genetic linkage studies map the multiple endocrine neoplasia type 2 loci to a small interval on chromosome 10q11.2. Hum Mol Genet 1993 ; 2 : 241 -246. Crossref Search ADS PubMed 3 Mulligan LM, Kwok JB, Healey CS, Elsdon MJ, Eng C, Gardner E, et al. Germ-line mutations of the RET proto-oncogene in multiple endocrine neoplasia type 2A. Nature 1993 ; 363 : 458 -460. Crossref Search ADS PubMed 4 Donis-Keller H, Dou S, Chi D, Carlson KM, Toshima K, Lairmore TC, et al. Mutations in the RET proto-oncogene are associated with MEN 2A and FMTC. Hum Mol Genet 1993 ; 2 : 851 -856. Crossref Search ADS PubMed 5 Myers SM, Eng C, Ponder BA, Mulligan LM. Characterization of RET proto-oncogene 3′ splicing variants and polyadenylation sites: a novel C-terminus for RET. Oncogene 1995 ; 11 : 2039 -2045. PubMed 6 Eng C, Clayton D, Schuffenecker I, Lenoir G, Cote G, Gagel RF, et al. The relationship between specific RET proto-oncogene mutations and disease phenotype in multiple endocrine neoplasia type 2. International RET mutation consortium analysis. JAMA 1996 ; 276 : 1575 -1579. Crossref Search ADS PubMed 7 Sanchez B, Antiñolo G, Navarro E, Japon MA, Conde AF, Astorga R, Borrego S. Cys634 mutations in the RET proto-oncogene in Spanish families affected by MEN 2A. Hum Mutat 1998 ; S1 : S72 -S73. 8 Sanchez B, Robledo M, Biarnes J, Saez ME, Volpini V, Benitez J, et al. High prevalence of the C634Y mutation in the RET proto-oncogene in MEN 2A families in Spain. J Med Genet 1999 ; 36 : 68 -70. PubMed 9 Eng C. Seminars in medicine of the Beth Israel Hospital, Boston. The RET proto-oncogene in multiple endocrine neoplasia type 2 and Hirschsprung’s disease [Review]. N Engl J Med 1996 ; 335 : 943 -951. Crossref Search ADS PubMed 10 Wittwer CT, Herrmann MG, Moss AA, Rasmussen RP. Continuous fluorescence monitoring of rapid cycle DNA amplification. Biotechniques 1997 ; 22 : 130 -131. Crossref Search ADS PubMed 11 Dracapoly NH Haines JL Korf BR Moir CG Seidman LE Seidman JGet al eds. Current protocols in human genetics 1994 : A.3B1-2 John Wiley & Sons New York. . 12 Ravine D. Automated mutation analysis [Review]. J Inherit Metab Dis 1999 ; 22 : 503 -518. Crossref Search ADS PubMed 13 Bernard PS, Ajioka RS, Kushner JP, Wittwer CT. Homogeneous multiplex genotyping of hemochromatosis mutations with fluorescent hybridization probes. Am J Pathol 1998 ; 153 : 1055 -1061. Crossref Search ADS PubMed 14 von Ahsen N, Schutz E, Armstrong VW, Oellerich M. Rapid detection of prothrombotic mutations of prothrombin (G20210A), factor V (G1691A), and methylenetetrahydrofolate reductase (C677T) by real-time fluorescence PCR with the LightCycler. Clin Chem 1999 ; 45 : 694 -696. Crossref Search ADS PubMed 15 von Ahsen N, Oellerich M, Schutz E. Use of two reporter dyes without interference in a single-tube rapid-cycle PCR: α1-antitrypsin genotyping by multiplex real-time fluorescence PCR with the LightCycler. Clin Chem 2000 ; 46 : 156 -161. Crossref Search ADS PubMed 16 Nauck M, Wieland H, Marz W. Rapid, homogeneous genotyping of the 4G/5G polymorphism in the promoter region of the PAII gene by fluorescence resonance energy transfer and probe melting curves. Clin Chem 1999 ; 45 : 1141 -1147. Crossref Search ADS PubMed 17 Aslanidis C, Nauck M, Schmitz G. High-speed detection of the two common α1-antitrypsin deficiency alleles Pi*Z and Pi*S by real-time fluorescence PCR and melting curves. Clin Chem 1999 ; 45 : 1872 -1875. Crossref Search ADS PubMed 18 Aslanidis C, Nauck M, Schmitz G. High-speed prothrombin G→A 20210 and methylenetetrahydrofolate reductase C→T 677 mutation detection using real-time fluorescence PCR and melting curves. Biotechniques 1999 ; 27 : 234 -238. Crossref Search ADS PubMed 19 Ruiz A, Royo JL, Rubio A, Borrego S, Leal M, Sanchez B, et al. Spectrofluormetric analysis of CCR5-Δ32 allele using real-time PCR: prevalence in Southern Spain HIV+ patients and non-infected population [Letter]. AIDS Res Hum Retroviruses 2001 ; 17 : 191 -193. Crossref Search ADS PubMed 20 Elenitoba-Johnson KSJ, Bohling SD, Wittwer CT, King TC. Multiplex PCR by multicolor fluorimetry and fluorescence melting curve analysis. Nat Med 2001 ; 7 : 249 -253. Crossref Search ADS PubMed 21 Gundry CN, Bernard PS, Herrmann MG, Reed GH, Wittwer CT. Rapid F508del, F508C assay using fluorescent hybridization probes. Genet Test 1999 ; 3 : 365 -370. Crossref Search ADS PubMed 22 Bernard PS, Pritham GH, Wittwer CT. Color multiplexing hybridization probes using the apolipoprotein E locus as a model system for genotyping. Anal Biochem 1999 ; 273 : 221 -228. Crossref Search ADS PubMed 23 Meuer S Wittwer C Nakagawa K eds. Rapid cycle real-time PCR. Methods and applications 2001 : 11 -19 Springer Berlin. . 24 Chen X, Zehnbauer B, Gnirke A, Kwok PY. Fluorescence energy transfer detection as a homogeneous DNA diagnostic method. Proc Natl Acad Sci U S A 1997 ; 94 : 10756 -10761. Crossref Search ADS PubMed 25 Chen X, Kwok PY. Template-directed dye-terminator incorporation (TDI) assay: a homogeneous DNA diagnostic method based on fluorescence resonance energy transfer. Nucleic Acids Res 1997 ; 25 : 347 -353. Crossref Search ADS PubMed 26 Eng C, Vijg J. Genetic testing: the problems and the promise [Review]. Nat Biotechnol 1997 ; 15 : 422 -426. Crossref Search ADS PubMed © 2001 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)

Zytkovicz, Thomas, H;Fitzgerald, Eileen, F;Marsden,, Deborah;Larson, Cecilia, A;Shih, Vivian, E;Johnson, Donna, M;Strauss, Arnold, W;Comeau, Anne, Marie;Eaton, Roger, B;Grady, George, F

doi: 10.1093/clinchem/47.11.1945

, von Ahsen, Nicolas;Wittwer, Carl, T;Schütz,, Ekkehard

doi: 10.1093/clinchem/47.11.1956pmid: N/A

Abstract Background: Many techniques in molecular biology depend on the oligonucleotide melting temperature (Tm), and several formulas have been developed to estimate Tm. Nearest-neighbor (N-N) models provide the highest accuracy for Tm prediction, but it is not clear how to adjust these models for the effects of reagents commonly used in PCR, such as Mg2+, deoxynucleotide triphosphates (dNTPs), and dimethyl sulfoxide (DMSO). Methods: The experimental Tms of 475 matched or mismatched target/probe duplexes were obtained in our laboratories or were compiled from the literature based on studies using the same real-time PCR platform. This data set was used to evaluate the contributions of [Mg2+], [dNTPs], and [DMSO] in N-N calculations. In addition, best-fit coefficients for common empirical formulas based on GC content, length, and the equivalent sodium ion concentration of cations [Na+eq] were obtained by multiple regression. Results: When we used [Na+eq] = [Monovalent cations] + 120(⁠ \(\sqrt{{[}Mg^{2{+}}{]}\ {-}\ {[}dNTPs{]}}\) ⁠) (the concentrations in this formula are mmol/L) to correct ΔS0 and a DMSO term of 0.75 °C (%DMSO), the SE of the N-N Tm estimate was 1.76 °C for perfectly matched duplexes (n = 217). Alternatively, the empirical formula Tm (°C) = 77.1 °C + 11.7 × log[Na+eq] + 0.41(%GC) − 528/bp − 0.75 °C(%DMSO) gave a slightly higher SE of 1.87 °C. When all duplexes (matched and mismatched; n = 475) were included in N-N calculations, the SE was 2.06 °C. Conclusions: This robust model, accounting for the effects of Mg2+, DMSO, and dNTPs on oligonucleotide Tm in PCR, gives reliable Tm predictions using thermodynamic N-N calculations or empirical formulas. DNA amplification and detection techniques often depend on oligonucleotide melting temperature (Tm). 1 The Tm of a DNA duplex, defined as the temperature where one-half of the nucleotides are paired and one-half are unpaired (1), corresponds to the midpoint of the spectroscopic hyperchromic absorbance shift during DNA melting (2). The Tm indicates the transition from double helical to random coil formation and is related to the DNA GC base content (2). Other important factors for DNA stability are the cation concentration of the surrounding buffer (3)(4) and the DNA double strand length (5). The Tm of polymer DNA can be predicted by formulas that account for differing GC content in various buffers (6). A term for the length of the duplex can also be included (1)(7)(8). The most accurate prediction of Tm for oligonucleotide DNA uses the thermodynamic nearest-neighbor (N-N) model [see Ref. (9) for review and parameters]. N-N calculations for Tm prediction are useful on microarrays (10) and for the selection of PCR primers and hybridization probes (11). Empirical data from probe melting curve analysis during real-time PCR correlate well with theoretical predictions (12)(13). The N-N model is based on the assumption that probe hybridization energy can be calculated from the enthalpy and entropy of all N-N pairs, including a contribution from each dangling end. Thermodynamic values for entropy and enthalpy of each possible matched N-N and dangling ends have been determined (9)(14). Dangling-end effects account for the stacking energy of a shorter probe on a more lengthy target (15)(16). The entropy (ΔS) is salt dependent, and ΔS0 must be corrected if the ionic environment is different from 1 mol/L NaCl, the salt concentration at which most thermodynamic values have been derived. However, Mg2+ is present in PCR as an important cofactor for Taq DNA polymerase and strongly influences ΔS (17). Deoxynucleotide triphosphates (dNTPs) are also essential and chelate some of the available Mg2+ (18). In addition, dimethyl sulfoxide (DMSO) is commonly used as a cosolvent (19) to facilitate amplification from difficult templates. Addition of DMSO decreases the Tm (20)(21)(22), which must be taken into account when primer Tm is calculated (23). To get a deeper insight into probe Tm under common PCR buffer conditions, we have compiled Tm data from 475 different DNA duplexes. These assays were performed with different concentrations of Mg2+ and DMSO, reflecting current PCR laboratory practice. We provide an empirical ΔS compensation for the Mg2+ and dNTP influence on ionic strength. In addition, best-fit coefficients for simpler formulas based on GC content, length, and the equivalent sodium ion concentration are determined for convenient bench-side use that offer accurate Tm predictions. Materials and Methods basic principle The LightCycler real-time PCR machine (Roche Molecular Biochemicals) is capable of detecting the hybridization of adjacent fluorescent dye-labeled probes by fluorescence resonance energy transfer (24). Assay design for the detection of single nucleotide polymorphisms requires that the sensor probe (the probe covering the mutation) has a lower Tm than the detection probe, which stays hybridized during the melting cycle. The observed Tms from matched and mismatched hybridizations can be predicted using the N-N model (12). Probes have a Tm similar to PCR primers (50–70 °C), and the results from probe Tm prediction can also be used to predict primer Tms, a critical parameter for PCR performance (23)(25)(26)(27). data collection All assays were performed using the LightCycler real-time PCR instrument. Most Tms were measured in our laboratories (n = 388) during the course of genotyping experiments [for examples, see Refs (12)(28)]. Some additional Tms were extracted from the literature when complete experimental conditions were published (n = 87). In total, 162 different probes were used with various templates and conditions, including 221 completely matched hybridizations, 237 single mismatches, and 17 two-point mismatches. Forty assays were based on melting curves detected with SYBR Green I, whereas the remainder were based on the melting of fluorescent oligonucleotide probes. DMSO was used in 206 assays in concentrations ranging from 2.5% to 10%. These data are available as an online supplement at Clinical Chemistry Online (http://www.clinchem.org/content/vol47/issue11). statistical analysis N-N calculations. The entropy and enthalpy were calculated from probe sequences at standard conditions (1 mol/L NaCl) as described in more detail elsewhere (9)(29)(30). In addition we considered the published thermodynamic data for dangling-end contributions (14). Mismatches were accounted for by the thermodynamic data reported by Allawi and SantaLucia (31)(31)(32)(33)(34) and Peyret et al. (35). The PCR DNA target concentration was set to 50 nM. SYBR Green I, if present, was assumed to increase the Tm by 1 °C at 1:20 000 dilution, based on own preliminary data. Calculations were performed with ExcelTM for Windows (Microsoft), using the built-in statistical functions. The Pearson r2 was used for correlations, and standard linear regression was used for relating observed to measured Tm. Thermodynamic N-N stability calculations were performed using MeltCalc, a spreadsheet software for Excel (36). The observed Tm was used as the dependent variable in a multiple variable fit to determine the DMSO coefficient and the best formula for Mg2+ influence. Prior evidence suggested that the influence of Mg2+ on Tm is stoichiometrically reduced by dNTPs. Initial calculations based on r2 indicated that the relationship of DMSO to Tm was linear, whereas [Mg2+] was nonlinear. Therefore, our model was: Tm (observed) = Tm (predicted) − a × DMSO (%), with [Na+eq] = [monovalent cations] + b × ([Mg2+] − [dNTP])c. The parameters a, b, and c were optimized to minimize the prediction error by stepwise incremental iterations. With a = 0.75, b = 120, and c = 0.5, only 18% of the predicted values fell outside a 5% error limit. The nonlinear effect of [Mg2+] on [Na+eq] was best approximated by the square-root function, which is in agreement with a previous report (37). Alternative formulas. Using our empirical data set, we evaluated several simpler formulas for their ability to predict Tm. These formulas cannot properly account for the presence of single mismatches. Therefore, only data for matched probe/template duplexes were used. The Wallace–Ikatura rule is often used as a rule of thumb when primer Tm is to be estimated at the bench (1)(38). However, the formula was originally applied to the hybridization of probes in 1 mol/L NaCl (1) and is an estimate of the denaturation temperature (Td): \[T_{\mathrm{d}}({^\circ}\mathrm{C}){=}2(\mathrm{A}{+}\mathrm{T}){+}4(\mathrm{G}{+}\mathrm{C})\] Another equation for the effective priming temperature (Tp) was suggested by Wu et al. (25): \[T_{\mathrm{p}}({^\circ}\mathrm{C}){=}22{+}1.46\mathrm{L}_{\mathrm{n}}\] where Ln = 2(G + C) + (A + T). Marmur and Doty (2) originally established a formula to correlate GC content (%GC) to the Tm of long duplexes at a given ionic strength. Chester and Marshak (23) added a term to account for DNA strand length (n in base pairs) to estimate primer Tm: \[T_{\mathrm{m}}{=}69.3{+}0.41(\%\mathrm{GC}){-}650/\mathrm{n}\] The Marmur–Schildkraut–Doty equation also accounts for ionic strength with a term for the Na+ concentration (1)(2)(6)(7)(8). \[T_{\mathrm{m}}({^\circ}\mathrm{C}){=}81.5\ {^\circ}\mathrm{C}{+}16.6(\mathrm{log}{[}\mathrm{Na}^{{+}}{]}){+}0.41(\%\mathrm{GC}){-}b/\mathrm{n}\] Values between 500 and 750 have been used for b (5)(23), a value that may increase with the ionic strength (8). Another modification is that of Wetmur (1): \[T_{\mathrm{m}}{=}81.5{+}16.6{\times}\mathrm{log}\left(\frac{{[}\mathrm{Na}^{{+}}{]}}{1.0{+}0.7{\times}{[}\mathrm{Na}^{{+}}{]}}\right){+}0.41(\%\mathrm{GC}){-}500/\mathrm{n}\] Eqs. 4 and 5 assume that the stabilizing effects of cations are the same on all base pairs. However, Owen et al. (39) observed that the slopes of Tm vs log[Na+] decrease with increasing GC content, leading to the following final formula for the estimation of Tm in polymer DNA (40)(41)(42): \[T_{\mathrm{m}}({^\circ}\mathrm{C}){=}87.16{+}0.345(\%\mathrm{GC}){+}\mathrm{log}{[}\mathrm{Na}^{{+}}{]}{\times}{[}20.17{-}0.066(\%\mathrm{GC}){]}\] Results descriptive statistics Empirical oligonucleotide Tms were obtained under 475 unique conditions typical of current PCR practice. The Mg2+ concentration was 1–6 mM (mean, 2.9 ± 1.17 mM), the length of the probes was 11–45 bp (median, 19 bp), and the GC content was 17–88% (mean, 52.5% ± 9.6%). DMSO was used in 206 assays in concentrations ranging from 2.5% to 10%. In 57 assays, synthesized oligonucleotides were used without PCR. Probe concentrations were 0.025–0.5 μM (mean, 0.15 ± 0.087 μM). The monovalent cation concentration was 20–54 mM, depending on the PCR buffer system. The measured probe Tms varied from 43.2 to 74.7 °C (mean, 59.1 ± 6.42 °C). statistical analysis N-N calculations. The best fit values for [Mg2+] and [dNTP] coefficients revealed that the Na+ equivalents (Na+eq) were approximated by: \[{[}\mathrm{Na}^{{+}}_{\mathrm{eq}}{]}{=}{[}\mathrm{Monovalent\ cations}{]}{+}120(\sqrt{{[}\mathrm{Mg}^{2{+}}{]}{-}{[}\mathrm{dNTPs}{]}})\] Note that the concentrations in Eq. 7 are in mmol/L. In all other equations, concentrations are in mol/L. Monovalent cations are typically present as K+ and Tris+ in PCR buffer (18)(26). K+ is similar to Na+ in regard to duplex stabilization (17). ΔS0 was corrected for the salt concentration as follows (9): \[S^{0}{[}\mathrm{Na}^{{+}}{]}{=}S^{0}{[}1\mathrm{M\ Na}^{{+}}{]}{+}0.847{\times}\mathrm{n}{\times}\mathrm{log}{[}\mathrm{Na}^{{+}}{]}\] where n is the total number of phosphates in the duplex divided by 2. This is equal to the oligonucleotide length minus 1. Each percentage of DMSO (by volume) decreased the Tm by 0.75 °C. GC content had no obvious influence on the DMSO factor. Using these equations, we found a good regression of predicted vs observed Tm (Fig. 1 ). The mean prediction error was 0.2 ± 2.18 °C, which is within the error range for N-N calculations. Figure 1. Open in new tabDownload slide Observed vs predicted Tm for 475 melting temperature assays. (- - - - -), line of identity; (——–), regression line: y = 1.00x − 0.29. Figure 1. Open in new tabDownload slide Observed vs predicted Tm for 475 melting temperature assays. (- - - - -), line of identity; (——–), regression line: y = 1.00x − 0.29. Alternative formulas. The Wallace–Ikatura rule (Eq. 1 ) overestimates the Tm of long duplexes and gives reasonable results only in the range of 14–20 bp (1)(38). Therefore, only duplexes shorter than 21 bp were included in the analysis (Table 1 ). When the analysis was extended to include duplexes of up to 24 bp, r2 decreased to 0.64. Table 1. Comparison of the predictive performance of different equations for the calculation of oligonucleotide Tm.1 . N-N calculations . . Equation . . . . . . . Including mismatches . Without mismatches . 12 . 22 . 9 . 10 . 11 . 12 . n 475 217 131 200 217 217 217 217 r2 0.90 0.89 0.78 0.60 0.86 0.87 0.88 0.88 Slope 1.00 1.05 1.20 0.95 1.01 0.96 0.97 1.01 Intercept, ΔTm −0.29 −3.17 −18.44 5.83 −0.60 2.65 2.32 −0.54 Sy|x 2.06 1.76 2.21 3.13 1.99 1.88 1.87 1.86 Mean ± SD of differences (observed minus predicted) 0.1 ± 2.18 0.2 ± 1.96 6.1 ± 3.13 −2.5 ± 3.84 −0.1 ± 1.93 0.0 ± 2.16 −0.1 ± 2.00 0.1 ± 2.18 . N-N calculations . . Equation . . . . . . . Including mismatches . Without mismatches . 12 . 22 . 9 . 10 . 11 . 12 . n 475 217 131 200 217 217 217 217 r2 0.90 0.89 0.78 0.60 0.86 0.87 0.88 0.88 Slope 1.00 1.05 1.20 0.95 1.01 0.96 0.97 1.01 Intercept, ΔTm −0.29 −3.17 −18.44 5.83 −0.60 2.65 2.32 −0.54 Sy|x 2.06 1.76 2.21 3.13 1.99 1.88 1.87 1.86 Mean ± SD of differences (observed minus predicted) 0.1 ± 2.18 0.2 ± 1.96 6.1 ± 3.13 −2.5 ± 3.84 −0.1 ± 1.93 0.0 ± 2.16 −0.1 ± 2.00 0.1 ± 2.18 1 See Materials and Methods for a detailed description of the formulas. 2 Only oligonucleotides <21 bp were included in the analysis. 3 Only oligonucleotides with Ln <39 (see Eq. 2 ) were included in the analysis. Open in new tab Table 1. Comparison of the predictive performance of different equations for the calculation of oligonucleotide Tm.1 . N-N calculations . . Equation . . . . . . . Including mismatches . Without mismatches . 12 . 22 . 9 . 10 . 11 . 12 . n 475 217 131 200 217 217 217 217 r2 0.90 0.89 0.78 0.60 0.86 0.87 0.88 0.88 Slope 1.00 1.05 1.20 0.95 1.01 0.96 0.97 1.01 Intercept, ΔTm −0.29 −3.17 −18.44 5.83 −0.60 2.65 2.32 −0.54 Sy|x 2.06 1.76 2.21 3.13 1.99 1.88 1.87 1.86 Mean ± SD of differences (observed minus predicted) 0.1 ± 2.18 0.2 ± 1.96 6.1 ± 3.13 −2.5 ± 3.84 −0.1 ± 1.93 0.0 ± 2.16 −0.1 ± 2.00 0.1 ± 2.18 . N-N calculations . . Equation . . . . . . . Including mismatches . Without mismatches . 12 . 22 . 9 . 10 . 11 . 12 . n 475 217 131 200 217 217 217 217 r2 0.90 0.89 0.78 0.60 0.86 0.87 0.88 0.88 Slope 1.00 1.05 1.20 0.95 1.01 0.96 0.97 1.01 Intercept, ΔTm −0.29 −3.17 −18.44 5.83 −0.60 2.65 2.32 −0.54 Sy|x 2.06 1.76 2.21 3.13 1.99 1.88 1.87 1.86 Mean ± SD of differences (observed minus predicted) 0.1 ± 2.18 0.2 ± 1.96 6.1 ± 3.13 −2.5 ± 3.84 −0.1 ± 1.93 0.0 ± 2.16 −0.1 ± 2.00 0.1 ± 2.18 1 See Materials and Methods for a detailed description of the formulas. 2 Only oligonucleotides <21 bp were included in the analysis. 3 Only oligonucleotides with Ln <39 (see Eq. 2 ) were included in the analysis. Open in new tab The equation for the effective priming temperature by Wu et al. (25) (Eq. 2 ) is similar to the Wallace–Ikatura rule. Only oligonucleotides with Ln <39 (see Eq. 2 ) were included in the analysis, as suggested by the authors. The oligonucleotide Tm is overestimated by a mean of 2.5 °C by this formula (Table 1 ). Including the length dependence of Chester and Marshak [Eq. 3 ; (23)] improved r2, but the intercept and difference between observed and predicted Tms were poor (data not shown). However, a better fit was obtained with new constants obtained from stepwise iterations (fit variable: length dependency term): \[T_{\mathrm{m}}{=}69.3{+}0.41(\%\mathrm{GC}){-}535/\mathrm{n}\] To improve Eq. 4 , a dTm/dlog[Na+] of 11.7 °C was used as suggested for oligonucleotide DNA (9) instead of the commonly used 16.6 °C (6). Best-fit values for the temperature offset and length dependence were then obtained by stepwise iterations (fit variables: Tm offset, length dependency term): \[T_{\mathrm{m}}({^\circ}\mathrm{C}){=}77.1\ {^\circ}C{+}11.7{\times}\mathrm{log}{[}\mathrm{Na}^{{+}}{]}{+}0.41(\%\mathrm{GC}){-}528/\mathrm{n}\] In the same manner, the modification for the formula of Wetmur (1) is (fit variables: Tm offset, length dependency term): \[T_{\mathrm{m}}{=}77.8{+}11.7{\times}\mathrm{log}\left(\frac{{[}\mathrm{Na}^{{+}}{]}}{1.0{+}0.7{\times}{[}\mathrm{Na}^{{+}}{]}}\right){+}0.41(\%\mathrm{GC}){-}528/\mathrm{n}\] The fit of this equation could not be improved (data not shown) by the introduction of a salt dependence term for b (see Eq. 4 ) as suggested previously (8). Eq. 6 has been suggested as the best predictor of polymer DNA Tm (41)(42). The best fit of constants in this formula based on our oligonucleotide data produced (fit variables: Tm offset, oligomer [Na+] dependency term, length dependency term added): \[T_{\mathrm{m}}({^\circ}\mathrm{C}){=}80.4{+}0.345(\%\mathrm{GC}){+}\mathrm{log}{[}\mathrm{Na}^{{+}}{]}{\times}{[}17.0{-}0.135(\%\mathrm{GC}){]}{-}550/\mathrm{n}\] Discussion Many techniques in modern molecular biology depend on oligonucleotide duplex formation. Perhaps the most universal method in use today is the PCR, with applications including amplification, cloning, mutation detection, and mutagenesis. A precise knowledge of oligonucleotide Tm is useful for rapid optimization of assays. The most accurate predictions of oligonucleotide Tms use the N-N model. N-N parameters have recently been improved (9), and their ability to predict the Tm of unknown oligonucleotide duplexes demonstrated (30). However, these parameters usually are determined in 1 mol/L NaCl and need to be corrected for the conditions in the PCR. Although correction factors for different NaCl concentrations have been published (9), we are not aware of an investigation into nucleic acid stability under typical PCR buffer conditions, even given its high practical value. It is well appreciated that Mg2+ stabilizes duplex DNA 80- to 100-fold (3) to as much as 140-fold (17) more than Na+. Our findings indicate a nonlinear effect of [Mg2+] on [Na+eq] in the order of 120(⁠ \(\sqrt{{[}Mg^{2{+}}{]}}\) ⁠). The N-N model accurately predicts probe Tm in the LightCycler analysis system (12). We compiled and analyzed measured Tm data for matched and mismatched hybridization probes from different laboratories. Therefore, some interlaboratory variation is expected from different reagents and protocols. The temperature transition rates during melting curve acquisition (usually 0.1–0.2 °C/s) are too fast to achieve equilibrium conditions and cause a slight overestimation of the true Tm (43). Probe sequence choices reflect the demands of mutation detection and are not designed particularly for two-state behavior. Even with these limitations, the predictive accuracy we have achieved underscores the robustness of the parameterization. Systematic errors introduced by fluorescent dyes were negligible in another study (10), and fluorescence resonance energy transfer probes themselves have successfully been used to derive thermodynamic parameters (44). Our findings should be equally applicable to probe and primer oligonucleotides, thereby allowing in silico optimization of primers and probes, saving both on time required for optimization and costs for probe resynthesis (12)(29)(45)(46)(47). An extensive survey of alternative formulas used for the prediction of perfectly matched oligonucleotide DNA was also performed. The Wallace–Ikatura rule (Eq. 1 ) is often used as a rough predictor of primer Tm but has limited accuracy, especially for longer oligonucleotides. This rule assumes a salt concentration of 1 mol/L NaCl, which is typical for dot blots and other hybridizations but not PCR. The fact that it is used for PCR applications is more a testament of the robustness of PCR toward different annealing temperatures than evidence for accurate Tm estimates. The same is true for the formula of Wu et al. (25). Many formulas were originally designed to relate measured Tm and GC content of polymer DNA (Eqs. 3–6). The inclusion of additional terms for ionic strength (6), length dependency (5), and GC dependency of ionic strength (39) has led to more accurate estimates for polymer DNA. Eqs. 9–12 have been specifically optimized for oligonucleotide Tm estimation by best-fit estimates of our data set. Eq. 10 is recommended as a tradeoff between accuracy and ease of use. Table 2 gives primer Tms for common primer compositions and PCR conditions. These estimates may not be accurate for certain sequences with a biased N-N composition (41). Furthermore, mismatches are strongly dependent on their N-N bases and require more laborious N-N calculations. Table 2. Standard primer Tms (°C) calculated for different lengths, Mg2+ concentration, and GC content at 0.8 mM dNTPs and 50 mM monovalent ion concentration without mismatches and with no addition of cosolvents.1 %GC . Mg2+, mM . Tm, °C . . . . . . 18 bp . 20 bp . 22 bp . 24 bp . 40 1.0 52.6 55.6 58.0 60.0 1.5 54.5 57.5 59.9 61.9 2.0 55.5 58.4 60.8 62.8 2.5 56.2 59.1 61.5 63.5 3.0 56.7 59.6 62.0 64.0 50 1.0 56.7 59.7 62.1 64.1 1.5 58.6 61.6 64.0 66.0 2.0 59.6 62.5 64.9 66.9 2.5 60.3 63.2 65.6 67.6 3.0 60.8 63.7 66.1 68.1 60 1.0 60.8 63.8 66.2 68.2 1.5 62.7 65.7 68.1 70.1 2.0 63.7 66.6 69.0 71.0 2.5 64.4 67.3 69.7 71.7 3.0 64.9 67.8 70.2 72.2 %GC . Mg2+, mM . Tm, °C . . . . . . 18 bp . 20 bp . 22 bp . 24 bp . 40 1.0 52.6 55.6 58.0 60.0 1.5 54.5 57.5 59.9 61.9 2.0 55.5 58.4 60.8 62.8 2.5 56.2 59.1 61.5 63.5 3.0 56.7 59.6 62.0 64.0 50 1.0 56.7 59.7 62.1 64.1 1.5 58.6 61.6 64.0 66.0 2.0 59.6 62.5 64.9 66.9 2.5 60.3 63.2 65.6 67.6 3.0 60.8 63.7 66.1 68.1 60 1.0 60.8 63.8 66.2 68.2 1.5 62.7 65.7 68.1 70.1 2.0 63.7 66.6 69.0 71.0 2.5 64.4 67.3 69.7 71.7 3.0 64.9 67.8 70.2 72.2 1 Eqs. 7 and 10 (see Material and Methods) were used for the calculations. Open in new tab Table 2. Standard primer Tms (°C) calculated for different lengths, Mg2+ concentration, and GC content at 0.8 mM dNTPs and 50 mM monovalent ion concentration without mismatches and with no addition of cosolvents.1 %GC . Mg2+, mM . Tm, °C . . . . . . 18 bp . 20 bp . 22 bp . 24 bp . 40 1.0 52.6 55.6 58.0 60.0 1.5 54.5 57.5 59.9 61.9 2.0 55.5 58.4 60.8 62.8 2.5 56.2 59.1 61.5 63.5 3.0 56.7 59.6 62.0 64.0 50 1.0 56.7 59.7 62.1 64.1 1.5 58.6 61.6 64.0 66.0 2.0 59.6 62.5 64.9 66.9 2.5 60.3 63.2 65.6 67.6 3.0 60.8 63.7 66.1 68.1 60 1.0 60.8 63.8 66.2 68.2 1.5 62.7 65.7 68.1 70.1 2.0 63.7 66.6 69.0 71.0 2.5 64.4 67.3 69.7 71.7 3.0 64.9 67.8 70.2 72.2 %GC . Mg2+, mM . Tm, °C . . . . . . 18 bp . 20 bp . 22 bp . 24 bp . 40 1.0 52.6 55.6 58.0 60.0 1.5 54.5 57.5 59.9 61.9 2.0 55.5 58.4 60.8 62.8 2.5 56.2 59.1 61.5 63.5 3.0 56.7 59.6 62.0 64.0 50 1.0 56.7 59.7 62.1 64.1 1.5 58.6 61.6 64.0 66.0 2.0 59.6 62.5 64.9 66.9 2.5 60.3 63.2 65.6 67.6 3.0 60.8 63.7 66.1 68.1 60 1.0 60.8 63.8 66.2 68.2 1.5 62.7 65.7 68.1 70.1 2.0 63.7 66.6 69.0 71.0 2.5 64.4 67.3 69.7 71.7 3.0 64.9 67.8 70.2 72.2 1 Eqs. 7 and 10 (see Material and Methods) were used for the calculations. Open in new tab The effect of DMSO on thermal stability of DNA has been investigated before. Our factor of 0.75 °C decrease in Tm per 1% DMSO is similar to previous findings of 0.6 °C per 1% DMSO (20), 0.675 °C per 1% DMSO (22), and 0.5 °C per 1% DMSO (21). These prior studies were performed on polymer DNA, suggesting that DMSO may have a slightly greater effect on oligomer DNA. Because template priming during PCR is a kinetic process, efficient, specific priming should occur at the primer Tm (25), suggesting that the Tm can be used as the annealing temperature in PCR. Because efficient amplification is dependent on hybridization of both primers, it is rational to use the Tm of the least stable primer. The temperature used for annealing in PCR also depends on the annealing time. For example, allele-specific amplification can be achieved with rapid cycling (“0” s annealing) at a lower annealing temperature than conventional cycling with a longer annealing time (48). Finally, Tm is not just a property of an oligonucleotide, but a property of an oligonucleotide under specific conditions and at a given concentration. In conclusion, we have developed a robust model for the effects of Mg2+, DMSO, and dNTPs on oligonucleotide Tm under common PCR buffer conditions. This enables reliable Tm predictions and in silico primer and probe optimization using thermodynamic N-N calculations. We performed an extensive evaluation of different equations advocated for PCR primer Tm prediction. These formulas have been parameterized to accommodate for standard PCR conditions and are now useful for rapid calculation of the Tm of perfectly matched oligonucleotides. It is rational to use the Tm of the least stable PCR primer as the annealing temperature in PCR. The MeltCalc software is copyrighted by Ekkehard Schütz and Nicolas von Ahsen. 1 " Nonstandard abbreviations: Tm, melting temperature; N-N, nearest-neighbor; dNTP, deoxynucleotide triphosphate; and DMSO, dimethyl sulfoxide. References 1 Wetmur JG. DNA probes: applications of the principles of nucleic acid hybridization. Crit Rev Biochem Mol Biol 1991 ; 26 : 227 -259. Crossref Search ADS PubMed 2 Marmur J, Doty P. Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature. J Mol Biol 1962 ; 5 : 109 -118. Crossref Search ADS PubMed 3 Thomas R. [Investigations into denaturation of DNA]. Biochim Biophys Acta 1954 ; 14 : 231 -240. Crossref Search ADS PubMed 4 Dove WF, Davidson N. Cation effects on the denaturation of DNA. J Mol Biol 1962 ; 5 : 467 -478. Crossref Search ADS 5 Hall TJ, Grula JW, Davidson EH, Britten RJ. Evolution of sea urchin non-repetitive DNA. J Mol Evol 1980 ; 16 : 95 -110. Crossref Search ADS PubMed 6 Schildkraut C, Lifson S. Dependence of the melting temperature of DNA on salt concentration. Biopolymers 1965 ; 3 : 195 -208. Crossref Search ADS PubMed 7 Wahl GM, Berger SL, Kimmel AR. Molecular hybridization of immobilized nucleic acids: theoretical concepts and practical considerations. Methods Enzymol 1987 ; 152 : 399 -407. Crossref Search ADS PubMed 8 Britten RJ, Graham DE, Neufeld BR. Analysis of repeating DNA sequences by reassociation. Methods Enzymol 1974 ; 29 : 363 -418. Crossref Search ADS PubMed 9 SantaLucia J, Jr. A unified view of polymer, dumbbell, and oligonucleotide DNA nearest-neighbor thermodynamics. Proc Natl Acad Sci U S A 1998 ; 95 : 1460 -1465. Crossref Search ADS PubMed 10 Fotin AV, Drobyshev AL, Proudnikov DY, Perov AN, Mirzabekov AD. Parallel thermodynamic analysis of duplexes on oligodeoxyribonucleotide microchips. Nucleic Acids Res 1998 ; 26 : 1515 -1521. Crossref Search ADS PubMed 11 Rychlik W, Rhoads RE. A computer program for choosing optimal oligonucleotides for filter hybridization, sequencing and in vitro amplification of DNA. Nucleic Acids Res 1989 ; 17 : 8543 -8551. Crossref Search ADS PubMed 12 von Ahsen N, Oellerich M, Armstrong VW, Schütz E. Application of a thermodynamic nearest-neighbor model to estimate nucleic acid stability and optimize probe design: prediction of melting points of different mutations of apolipoprotein B 3500 and factor V Leiden with a hybridization probe genotyping assay on the LightCycler. Clin Chem 1999 ; 45 : 2094 -2101. Crossref Search ADS PubMed 13 von Ahsen N, Oellerich M, Schütz E. DNA base bulge vs unmatched end formation in probe-based diagnostic insertion/deletion genotyping: genotyping the UGT1A1 (TA)n polymorphism by real-time fluorescence PCR. Clin Chem 2000 ; 46 : 1939 -1945. Crossref Search ADS PubMed 14 Bommarito S, Peyret N, SantaLucia J, Jr. Thermodynamic parameters for DNA sequences with dangling ends. Nucleic Acids Res 2000 ; 28 : 1929 -1934. Crossref Search ADS PubMed 15 Hayes FN, Lilly EH, Ratliff RL, Smith DA, Williams DL. Thermal transitions in mixtures of polydeoxyribodinucleotides. Biopolymers 1970 ; 9 : 1105 -1117. Crossref Search ADS PubMed 16 Lane MJ, Paner T, Kashin I, Faldasz BD, Li B, Gallo FJ, et al. The thermodynamic advantage of DNA oligonucleotide ‘stacking hybridization’ reactions: energetics of a DNA nick. Nucleic Acids Res 1997 ; 25 : 611 -617. Crossref Search ADS PubMed 17 Nakano S, Fujimoto M, Hara H, Sugimoto N. Nucleic acid duplex stability: influence of base composition on cation effects. Nucleic Acids Res 1999 ; 27 : 2957 -2965. Crossref Search ADS PubMed 18 Baumforth KR, Nelson PN, Digby JE, O’Neil JD, Murray PG. Demystified … the polymerase chain reaction. Mol Pathol 1999 ; 52 : 1 -10. Crossref Search ADS PubMed 19 Pomp D, Medrano JF. Organic solvents as facilitators of polymerase chain reaction. Biotechniques 1991 ; 10 : 58 -59. PubMed 20 Musielski H, Mann W, Laue R, Michel S. Influence of dimethylsulfoxide on transcription by bacteriophage T3- induced RNA polymerase. Z Allg Mikrobiol 1981 ; 21 : 447 -456. Crossref Search ADS PubMed 21 Cullen BR, Bick MD. Thermal denaturation of DNA from bromodeoxyuridine substituted cells. Nucleic Acids Res 1976 ; 3 : 49 -62. Crossref Search ADS PubMed 22 Escara JF, Hutton JR. Thermal stability and renaturation of DNA in dimethyl sulfoxide solutions: acceleration of the renaturation rate. Biopolymers 1980 ; 19 : 1315 -1327. Crossref Search ADS PubMed 23 Chester N, Marshak DR. Dimethyl sulfoxide-mediated primer Tm reduction: a method for analyzing the role of renaturation temperature in the polymerase chain reaction. Anal Biochem 1993 ; 209 : 284 -290. Crossref Search ADS PubMed 24 de Silva D, Wittwer CT. Monitoring hybridization during polymerase chain reaction. J Chromatogr B 2000 ; 741 : 3 -13. Crossref Search ADS 25 Wu DY, Ugozzoli L, Pal BK, Qian J, Wallace RB. The effect of temperature and oligonucleotide primer length on the specificity and efficiency of amplification by the polymerase chain reaction. DNA Cell Biol 1991 ; 10 : 233 -238. Crossref Search ADS PubMed 26 Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn GT, et al. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 1988 ; 239 : 487 -491. Crossref Search ADS PubMed 27 Rychlik W, Spencer WJ, Rhoads RE. Optimization of the annealing temperature for DNA amplification in vitro [published erratum appears in Nucleic Acids Res 1991;19:698]. Nucleic Acids Res 1990 ; 18 : 6409 -6412. Crossref Search ADS PubMed 28 Crockett AO, Wittwer CT. Fluorescein-labeled oligonucleotides for real-time PCR: using the inherent quenching of deoxyguanosine nucleotides. Anal Biochem 2001 ; 290 : 89 -97. Crossref Search ADS PubMed 29 von Ahsen N, Schütz E. Using the nearest neighbor model for the estimation of matched and mismatched hybridization probe melting points and selection of optimal probes on the LightCycler. Meuer S Wittwer C Nakagawara K-I eds. Rapid cycle real time PCR—methods and applications 2001 : 43 -56 Springer Verlag Berlin. . 30 Owczarzy R, Vallone PM, Gallo FJ, Paner TM, Lane MJ, Benight AS. Predicting sequence-dependent melting stability of short duplex DNA oligomers. Biopolymers 1997 ; 44 : 217 -239. Crossref Search ADS PubMed 31 Allawi HT, SantaLucia J, Jr. Thermodynamics and NMR of internal G*T mismatches in DNA. Biochemistry 1997 ; 36 : 10581 -10594. Crossref Search ADS PubMed 32 Allawi HT, SantaLucia J, Jr. Nearest-neighbor thermodynamics of internal A*C mismatches in DNA: sequence dependence and pH effects. Biochemistry 1998 ; 37 : 9435 -9444. Crossref Search ADS PubMed 33 Allawi HT, SantaLucia J, Jr. Thermodynamics of internal C*T mismatches in DNA. Nucleic Acids Res 1998 ; 26 : 2694 -2701. Crossref Search ADS PubMed 34 Allawi HT, SantaLucia J, Jr. Nearest neighbor thermodynamic parameters for internal G*A mismatches in DNA. Biochemistry 1998 ; 37 : 2170 -2179. Crossref Search ADS PubMed 35 Peyret N, Seneviratne PA, Allawi HT, SantaLucia J, Jr. Nearest-neighbor thermodynamics and NMR of DNA sequences with internal A*A, C*C, G*G, and T*T mismatches. Biochemistry 1999 ; 38 : 3468 -3477. Crossref Search ADS PubMed 36 Schütz E, von Ahsen N. Spreadsheet software for thermodynamic melting point prediction of oligonucleotide hybridization with and without mismatches. Biotechniques 1999 ; 27 : 1218 -1224. Crossref Search ADS PubMed 37 Wetmur JG. Nucleic acid hybrids, formation and structure of. Meyers RA eds. Encyclopedia of molecular biology and molecular medicine 1996 ; Vol. 4 : 235 -243 VCH Verlagsgesellschaft Weinheim, Germany. . 38 Meinkoth J, Wahl G. Hybridization of nucleic acids immobilized on solid supports. Anal Biochem 1984 ; 138 : 267 -284. Crossref Search ADS PubMed 39 Owen RJ, Hill LR, Lapage SP. Determination of DNA base compositions from melting profiles in dilute buffers. Biopolymers 1969 ; 7 : 503 -516. Crossref Search ADS PubMed 40 Frank-Kamenetskii MD. Simplification of the empirical relationship between melting temperature of DNA, its GC content and concentration of sodium ions in solution. Biopolymers 1971 ; 10 : 2623 -2624. Crossref Search ADS PubMed 41 Blake RD. Denaturation of DNA. Meyers RA eds. Encyclopedia of molecular biology and molecular medicine 1996 ; Vol. 2 : 1 -19 VCH Verlagsgesellschaft Weinheim, Germany. . 42 Blake RD, Delcourt SG. Thermal stability of DNA. Nucleic Acids Res 1998 ; 26 : 3323 -3332. Crossref Search ADS PubMed 43 Gundry CN, Bernard PS, Herrmann MG, Reed GH, Wittwer CT. Rapid F508del and F508C assay using fluorescent hybridization probes. Genet Test 1999 ; 3 : 365 -370. Crossref Search ADS PubMed 44 Gelfand CA, Plum GE, Mielewczyk S, Remeta DP, Breslauer KJ. A quantitative method for evaluating the stabilities of nucleic acids. Proc Natl Acad Sci U S A 1999 ; 96 : 6113 -6118. Crossref Search ADS PubMed 45 Schütz E, von Ahsen N, Oellerich M. Genotyping of eight thiopurine methyltransferase mutations: three-color multiplexing, “Two-Color/Shared” anchor, and fluorescence-quenching hybridization probe assays based on thermodynamic nearest-neighbor probe design. Clin Chem 2000 ; 46 : 1728 -1737. Crossref Search ADS PubMed 46 Schütz E, von Ahsen N. Genotyping the most common thiopurine methyltransferase mutations with the LightCycler. Meuer S Wittwer C Nakagawara K-I eds. Rapid cycle real time PCR—methods and applications 2001 : 143 -152 Springer Verlag Berlin. . 47 von Ahsen N, Oellerich M, Schütz E. A method for homogeneous color-compensated genotyping of factor V (G1691A) and methylenetetrahydrofolate reductase (C677T) mutations using real-time multiplex fluorescence PCR. Clin Biochem 2000 ; 33 : 535 -539. Crossref Search ADS PubMed 48 Wittwer CT, Marshall BC, Reed GH, Cherry JL. Rapid cycle allele-specific amplification: studies with the cystic fibrosis ΔF508 locus. Clin Chem 1993 ; 39 : 804 -809. Crossref Search ADS PubMed © 2001 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)

Yamada,, Magohei;Marui,, Yoji;Hayashi,, Chozo;Miki,, Yasuyoshi;Takemura,, Shoji

doi: 10.1093/clinchem/47.11.1962pmid: N/A

Abstract Background: Several thiocholine alkanoyl esters were newly synthesized and explored as substrates for the assay of human serum cholinesterase after being subjected to the Ellman reaction (Arch Biochem Biophys 1958;74:443–50 and Arch Biochem Biophys 1959;82:70–7). Methods: We synthesized thiocholine ester iodides by the method of Renshow et al. (J Am Chem Soc 1938;60:1765–70). We examined solubility in H2O, substrate specificity serum for cholinesterase, (spontaneous) self-hydrolysis, storage stability, and reaction conditions for measurement of the activity of the enzyme. Results: Isobutyryl and cyclohexane-carboxyl esters showed the best efficiency for the specific and stable assay of human serum cholinesterase. Aqueous solubility of each was >10 mmol/L, and the reactivity with acetylcholinesterase was negligible. For isobutyryl and cyclohexane-carboxyl esters, respectively, spontaneous hydrolysis in the aqueous phase was ∼1/25 and ∼1/175 slower than the enzymatic hydrolysis, and assays with these substrates were linear to 1800 and 3000 U/L, respectively. The Km values of these acylthiocholines with human cholinesterase were almost equivalent (6.9 × 10−3 mmol/L). The substrates were stable in aqueous solution and in the solid state as the iodides for at least 5 years at 5 °C. Conclusions: The isobutyrate and cyclohexane-carboxylate of thiocholine are suitable for the specific assay of human serum cholinesterase. Thiocholine esters such as acetate, propionate, and n-butyrate have been used conventionally as substrates in the assay of human serum cholinesterase (1)(2) with the Ellman reaction (3)(4). The Ellman reaction involves the reaction of 5,5-dithio-bis(2-nitrobenzoic acid) (DTNB) with thiocholine liberated from its esters by enzymatic hydrolysis. The yellow 5-thio-2-nitro-benzoate (TNB) thus formed is detected by colorimetry. Disadvantages, especially for application to automatic measurements, lie in the use of these conventional assay reagents. (a) The assay procedure requires a high dilution of the serum samples (∼100-fold) (1), whereas the desirable dilution for accurate manual operations would likely be <30-fold, which is also readily applicable to automatic systems. (b) The optimum pH for the enzymatic reaction is ∼8–8.5, whereas the conventional measurements are usually performed at pH 7.6 to minimize the undesirable self-hydrolysis of the thiocholine esters (1). Using the synthetic substrate, 2,3-dimethoxybenzoylthiocholine, we previously simplified the assay procedure without manipulating the high dilution and reduced self-hydrolysis at the optimum pH of the reaction (5). In the current study, a series of esters were synthesized and their applications as assay reagents for human serum cholinesterase were explored. Two additional esters, thiocholine isobutyrate and cyclohexane-carboxylate, were also explored as suitable substrates for the assay. Materials and Methods synthesis of thiocholine esters Thirteen thiocholine ester iodides (1–13 in Table 1 ) were synthesized by a modified procedure of Renshow et al. (6). Table 1. Synthesized aliphatic acylthiocholines. RCO-S-CH2CH2-N+(CH3)3I− . . . . . Compound no. . R . Thiocholine . Melting point, °C . Yield,%1 . 1 Me3C- pivaloyl- 162–163 92 2 Me2CH- isobutyryl- 157–158 95 3 HOCO(CH2)2- succinyl- 211–213 42 4 Et2CH- 2-ethylbutyryl- 225–226 34 5 EtMeCH- 2-methylbutyryl- 215–217 22 6 c-C3H5-1 c-propanecarboxyl- 188–189 50 7 c-C4H7- c-butanecarboxyl- 164–165 46 8 c-C5H9- c-pentanecarboxyl- 221–222 40 9 c-C6H11- c-hexanecarboxyl- 159–161 59 10 c-C7H13- c-heptanecarboxyl- 219–220 40 11 adamantyl- adamantanecarboxyl- 227–228 91 12 PhCH2- phenylacetyl- 182–184 58 13 Ph(CH2)2- phenylpropionyl- 200–201 65 RCO-S-CH2CH2-N+(CH3)3I− . . . . . Compound no. . R . Thiocholine . Melting point, °C . Yield,%1 . 1 Me3C- pivaloyl- 162–163 92 2 Me2CH- isobutyryl- 157–158 95 3 HOCO(CH2)2- succinyl- 211–213 42 4 Et2CH- 2-ethylbutyryl- 225–226 34 5 EtMeCH- 2-methylbutyryl- 215–217 22 6 c-C3H5-1 c-propanecarboxyl- 188–189 50 7 c-C4H7- c-butanecarboxyl- 164–165 46 8 c-C5H9- c-pentanecarboxyl- 221–222 40 9 c-C6H11- c-hexanecarboxyl- 159–161 59 10 c-C7H13- c-heptanecarboxyl- 219–220 40 11 adamantyl- adamantanecarboxyl- 227–228 91 12 PhCH2- phenylacetyl- 182–184 58 13 Ph(CH2)2- phenylpropionyl- 200–201 65 1 c, cyclo-. 2 Overall yield from the starting acid chlorides. Open in new tab Table 1. Synthesized aliphatic acylthiocholines. RCO-S-CH2CH2-N+(CH3)3I− . . . . . Compound no. . R . Thiocholine . Melting point, °C . Yield,%1 . 1 Me3C- pivaloyl- 162–163 92 2 Me2CH- isobutyryl- 157–158 95 3 HOCO(CH2)2- succinyl- 211–213 42 4 Et2CH- 2-ethylbutyryl- 225–226 34 5 EtMeCH- 2-methylbutyryl- 215–217 22 6 c-C3H5-1 c-propanecarboxyl- 188–189 50 7 c-C4H7- c-butanecarboxyl- 164–165 46 8 c-C5H9- c-pentanecarboxyl- 221–222 40 9 c-C6H11- c-hexanecarboxyl- 159–161 59 10 c-C7H13- c-heptanecarboxyl- 219–220 40 11 adamantyl- adamantanecarboxyl- 227–228 91 12 PhCH2- phenylacetyl- 182–184 58 13 Ph(CH2)2- phenylpropionyl- 200–201 65 RCO-S-CH2CH2-N+(CH3)3I− . . . . . Compound no. . R . Thiocholine . Melting point, °C . Yield,%1 . 1 Me3C- pivaloyl- 162–163 92 2 Me2CH- isobutyryl- 157–158 95 3 HOCO(CH2)2- succinyl- 211–213 42 4 Et2CH- 2-ethylbutyryl- 225–226 34 5 EtMeCH- 2-methylbutyryl- 215–217 22 6 c-C3H5-1 c-propanecarboxyl- 188–189 50 7 c-C4H7- c-butanecarboxyl- 164–165 46 8 c-C5H9- c-pentanecarboxyl- 221–222 40 9 c-C6H11- c-hexanecarboxyl- 159–161 59 10 c-C7H13- c-heptanecarboxyl- 219–220 40 11 adamantyl- adamantanecarboxyl- 227–228 91 12 PhCH2- phenylacetyl- 182–184 58 13 Ph(CH2)2- phenylpropionyl- 200–201 65 1 c, cyclo-. 2 Overall yield from the starting acid chlorides. Open in new tab N,N-Dimethylamino-ethanethiol hydrochloride (1 mol) was suspended in ether (10-fold the amount of the hydrochloride), and the mixture was stirred with aqueous sodium carbonate (1.2 mol) at room temperature. The ether layer was separated and dried over anhydrous sodium sulfate. The corresponding acid chloride (1.1 mol) was added dropwise under ice-cooling conditions into the stirred ether solution. The separated crystalline ester hydrochloride was filtered and dissolved in water (12-fold the weight of the ester hydrochloride). The solution was alkalized with 100 mmol/L aqueous sodium hydroxide and extracted with ether. The extract was dried over anhydrous magnesium sulfate and stirred with methyl iodide (1.5 mol equivalent to the ester hydrochloride) until no more crystal was precipitated. The product was filtered, dried, and recrystallized from hot water. The melting points and yields of the products are listed in Table 1 . The elemental analysis for carbon, hydrogen, sulfur, and nitrogen were satisfactory (error, within 0.3%). Propionyl- and n-butyryl-thiocholine iodides and starting materials for syntheses were purchased from Aldrich Chemical Co. reagents for the assay Reagent materials were obtained from Aldrich Chemical Co. The human serum cholinesterase (EC 3.1.1.8) and acetylcholinesterase (EC 3.1.1.7) preparations were obtained from Sigma Chemical Japan, Ltd. Reagents were prepared as follows. Reagent 1 was a solution of 0.3 mmol of DTNB and 240 mmol of tris(hydroxymethyl)aminomethane in 1 L of water, which was adjusted to pH 8.0 by the addition of maleic acid. Reagent 2 was 6.0 mmol/L thiocholine ester iodide. general procedure for the assay A mixture of 2.4 mL of reagent 1 and 0.1 mL of the aqueous solution, containing 13 000 U/L of the enzyme, was preincubated for 5 min at 30 °C; 0.5 mL of reagent 2 was then added to the mixture. The change of the absorbance (ΔE/min) was measured every minute at 405 nm in a cell with a 1-cm pathlength. The susceptibility of the substrates, expressed in terms of Δμmol (substrate) · min−1 · L−1 (U/L), was calculated by the equation U/L = (A − B)/ε × V/v, where A represents the overall reaction rate (ΔE overall/min), B the reagent blank (ΔE blank/min) attributable to the self-hydrolytic reaction, and ε the molar absorptivity of TNB at 405 nm (13 300 L/mol = 0.0133 L/μmol). V/v is the ratio between final and sample volumes (3/0.1 = 30). The enzymatic reaction took place at 37 °C. examination of the stability The storage stability as the aqueous solution was examined with the use of the Arrhenius relationship between temperature and degradation rate according to Garrett and Carper (7). Aqueous solutions of thiocholine ester iodide samples (5 mmol/L) were kept in appropriate vials at 25 °C, 37 °C, and 50 °C in constant temperature baths. The DTNB solution was prepared separately, and the concentration was 0.1 mmol/L in the 100 mmol/L Tris buffer solution (pH 8.0). Samples of the thiocholine ester solutions were withdrawn from vials at intervals. After acclimatization at room temperature, 0.02 mL of the solution was mixed with 2.5 mL of the DTNB solution and 0.2 mL of serum cholinesterase preparation. The concentration of cholinesterase was chosen to ensure that the absorbance of the total TNB after the enzyme reaction would be as close to 0.5 as possible at time zero. The mixed solution was kept at 37 °C for 1 h, and the absorbance of TNB was measured at 405 nm to estimate the amount of the total thiocholine (At) liberated by the enzyme reaction from the remaining thiocholine ester plus that formed by the thermal degradation (hydrolysis). As the reference, the absorbance of the solution with the DTNB solution added, but not the enzyme preparation, was measured to estimate the amount of thiocholine formed nonenzymatically (As). The rate of degradation was followed for a period of 31 days. The first-order rate constant, k, at each of the temperatures was estimated as the slope of the plot of log (At − As) for the remaining thiocholine esters vs time (h). The values of log k were then subjected to the Arrhenius plot and plotted against the corresponding 1/T (T is the absolute temperature). By extrapolating the linear Arrhenius relationship derived by the regression analysis, the rate constant at 5 °C was estimated, from which the stability of the aqueous substrate solution in terms of t0.8, or the 20% loss time, was calculated. Results Reaction and self-hydrolysis rates of the synthesized thiocholine esters were examined by the general procedure described above at pH 8.0. As shown in Table 2 , isobutyrate (no. 2) exhibited a relatively high reactivity and a low self-hydrolysis. The presence of the secondary α-carbon atom in the aliphatic acyl moiety seems appropriate for the enzymatic reaction with lower self-hydrolysis. This also applies to the 2-ethylbutyryl (no. 4) and 2-methylbutyryl (no. 5) derivatives. Electron-donating and medium-sized steric effects of secondary alkyl groups may contribute to the low self-hydrolysis. Table 2. Properties of aliphatic acylthiocholines as substrates of human serum cholinesterase. . RCO-S-CH2CH2-N+(CH3)3I− . . . . . Compound no. . R . Reactivity, A (ΔE/min)1 . Rate of self-hydrolysis,1B (ΔE/min) . Ratio, A/B . Solubility in H2O, 10 mmol/L . 1 Me3C- 0.004 (0.008) 0.001 (0.004) 4.0 (2.0) +2 2 Me2CH- 0.155 (0.775) 0.006 (0.012) 25.8 (64.6) ++3 3 HOCO(CH2)2- 0.018 (0.087) 0.036 (0.068) 0.5 (1.3) + 4 Et2CH- 0.15 (0.725) 0.005 (0.012) 30.0 (60.4) + 5 EtMeCH- 0.14 (0.698) 0.006 (0.015) 23.3 (46.5) + 6 c-C3H5- 0.103 (0.504) 0.009 (0.016) 11.4 (31.5) ++ 7 c-C4H7- 0.098 (0.492) 0.010 (0.021) 9.8 (23.4) ++ 8 c-C5H9- 0.17 (0.862) 0.008 (0.018) 21.3 (47.4) ++ 9 c-C6H11- 0.35 (1.850) 0.002 (0.010) 175.0 (185.0) ++ 10 c-C7H13- 0.004 (0.007) 0.005 (0.007) 0.8 (1.0) ++ 11 adamantyl- 0.002 (0.008) 0.008 (0.009) 0.2 (0.9) + 12 PhCH2- 0.085 (0.435) 0.006 (0.012) 14.1 (36.3) ++ 13 Ph(CH2)2- 0.098 (0.499) 0.006 (0.015) 16.3 (33.2) ++ 14 C2H5-4 0.70 (1.36) 0.012 (0.015) 58.3 (90.7) ++ 15 n-C3H7-45 1.02 (2.06) 0.004 (0.006) 255.0 (343.3) ++ 16 2,3-(MeO)2C6H3-6 0.07 (0.361) 0.003 (0.012) 23.3 (30.1) ++ . RCO-S-CH2CH2-N+(CH3)3I− . . . . . Compound no. . R . Reactivity, A (ΔE/min)1 . Rate of self-hydrolysis,1B (ΔE/min) . Ratio, A/B . Solubility in H2O, 10 mmol/L . 1 Me3C- 0.004 (0.008) 0.001 (0.004) 4.0 (2.0) +2 2 Me2CH- 0.155 (0.775) 0.006 (0.012) 25.8 (64.6) ++3 3 HOCO(CH2)2- 0.018 (0.087) 0.036 (0.068) 0.5 (1.3) + 4 Et2CH- 0.15 (0.725) 0.005 (0.012) 30.0 (60.4) + 5 EtMeCH- 0.14 (0.698) 0.006 (0.015) 23.3 (46.5) + 6 c-C3H5- 0.103 (0.504) 0.009 (0.016) 11.4 (31.5) ++ 7 c-C4H7- 0.098 (0.492) 0.010 (0.021) 9.8 (23.4) ++ 8 c-C5H9- 0.17 (0.862) 0.008 (0.018) 21.3 (47.4) ++ 9 c-C6H11- 0.35 (1.850) 0.002 (0.010) 175.0 (185.0) ++ 10 c-C7H13- 0.004 (0.007) 0.005 (0.007) 0.8 (1.0) ++ 11 adamantyl- 0.002 (0.008) 0.008 (0.009) 0.2 (0.9) + 12 PhCH2- 0.085 (0.435) 0.006 (0.012) 14.1 (36.3) ++ 13 Ph(CH2)2- 0.098 (0.499) 0.006 (0.015) 16.3 (33.2) ++ 14 C2H5-4 0.70 (1.36) 0.012 (0.015) 58.3 (90.7) ++ 15 n-C3H7-45 1.02 (2.06) 0.004 (0.006) 255.0 (343.3) ++ 16 2,3-(MeO)2C6H3-6 0.07 (0.361) 0.003 (0.012) 23.3 (30.1) ++ 1 Concentration of serum cholinesterase was 13 000 U/L. Measured at pH 8.0 and 30°C unless noted. Values in parentheses are the data obtained at pH 8.0 and 37°C. 2 +, the solubility of 10 mmol/L was not attained at <25°C. 3 ++, the solubility of 10 mmol/L was attained at >5°C. 4 Purchased from Aldrich Chemical. Measured with 130 U/L of the enzyme solution, and the value was multiplied by 100 to give the “A” value comparable to those of the other compounds. 5 Measured at pH 7.0 and 30°C (37°C). 6 Synthesized in our previous study (5). Open in new tab Table 2. Properties of aliphatic acylthiocholines as substrates of human serum cholinesterase. . RCO-S-CH2CH2-N+(CH3)3I− . . . . . Compound no. . R . Reactivity, A (ΔE/min)1 . Rate of self-hydrolysis,1B (ΔE/min) . Ratio, A/B . Solubility in H2O, 10 mmol/L . 1 Me3C- 0.004 (0.008) 0.001 (0.004) 4.0 (2.0) +2 2 Me2CH- 0.155 (0.775) 0.006 (0.012) 25.8 (64.6) ++3 3 HOCO(CH2)2- 0.018 (0.087) 0.036 (0.068) 0.5 (1.3) + 4 Et2CH- 0.15 (0.725) 0.005 (0.012) 30.0 (60.4) + 5 EtMeCH- 0.14 (0.698) 0.006 (0.015) 23.3 (46.5) + 6 c-C3H5- 0.103 (0.504) 0.009 (0.016) 11.4 (31.5) ++ 7 c-C4H7- 0.098 (0.492) 0.010 (0.021) 9.8 (23.4) ++ 8 c-C5H9- 0.17 (0.862) 0.008 (0.018) 21.3 (47.4) ++ 9 c-C6H11- 0.35 (1.850) 0.002 (0.010) 175.0 (185.0) ++ 10 c-C7H13- 0.004 (0.007) 0.005 (0.007) 0.8 (1.0) ++ 11 adamantyl- 0.002 (0.008) 0.008 (0.009) 0.2 (0.9) + 12 PhCH2- 0.085 (0.435) 0.006 (0.012) 14.1 (36.3) ++ 13 Ph(CH2)2- 0.098 (0.499) 0.006 (0.015) 16.3 (33.2) ++ 14 C2H5-4 0.70 (1.36) 0.012 (0.015) 58.3 (90.7) ++ 15 n-C3H7-45 1.02 (2.06) 0.004 (0.006) 255.0 (343.3) ++ 16 2,3-(MeO)2C6H3-6 0.07 (0.361) 0.003 (0.012) 23.3 (30.1) ++ . RCO-S-CH2CH2-N+(CH3)3I− . . . . . Compound no. . R . Reactivity, A (ΔE/min)1 . Rate of self-hydrolysis,1B (ΔE/min) . Ratio, A/B . Solubility in H2O, 10 mmol/L . 1 Me3C- 0.004 (0.008) 0.001 (0.004) 4.0 (2.0) +2 2 Me2CH- 0.155 (0.775) 0.006 (0.012) 25.8 (64.6) ++3 3 HOCO(CH2)2- 0.018 (0.087) 0.036 (0.068) 0.5 (1.3) + 4 Et2CH- 0.15 (0.725) 0.005 (0.012) 30.0 (60.4) + 5 EtMeCH- 0.14 (0.698) 0.006 (0.015) 23.3 (46.5) + 6 c-C3H5- 0.103 (0.504) 0.009 (0.016) 11.4 (31.5) ++ 7 c-C4H7- 0.098 (0.492) 0.010 (0.021) 9.8 (23.4) ++ 8 c-C5H9- 0.17 (0.862) 0.008 (0.018) 21.3 (47.4) ++ 9 c-C6H11- 0.35 (1.850) 0.002 (0.010) 175.0 (185.0) ++ 10 c-C7H13- 0.004 (0.007) 0.005 (0.007) 0.8 (1.0) ++ 11 adamantyl- 0.002 (0.008) 0.008 (0.009) 0.2 (0.9) + 12 PhCH2- 0.085 (0.435) 0.006 (0.012) 14.1 (36.3) ++ 13 Ph(CH2)2- 0.098 (0.499) 0.006 (0.015) 16.3 (33.2) ++ 14 C2H5-4 0.70 (1.36) 0.012 (0.015) 58.3 (90.7) ++ 15 n-C3H7-45 1.02 (2.06) 0.004 (0.006) 255.0 (343.3) ++ 16 2,3-(MeO)2C6H3-6 0.07 (0.361) 0.003 (0.012) 23.3 (30.1) ++ 1 Concentration of serum cholinesterase was 13 000 U/L. Measured at pH 8.0 and 30°C unless noted. Values in parentheses are the data obtained at pH 8.0 and 37°C. 2 +, the solubility of 10 mmol/L was not attained at <25°C. 3 ++, the solubility of 10 mmol/L was attained at >5°C. 4 Purchased from Aldrich Chemical. Measured with 130 U/L of the enzyme solution, and the value was multiplied by 100 to give the “A” value comparable to those of the other compounds. 5 Measured at pH 7.0 and 30°C (37°C). 6 Synthesized in our previous study (5). Open in new tab Pivaloyl (no. 1) and adamantanecarboxyl (no. 11) derivatives showed a very low reactivity. This may be attributable to a severe hindrance at the tertiary alkyl groups that interferes with the approach of the enzyme molecule to the thioester linkage, as well as the linkage of water to the acylated enzyme. With the succinyl derivative (no. 3), the enzymatic reactivity decreased and the self-hydrolysis increased. The negatively charged carboxylate group in the neighborhood of the thioester bond may accelerate the self-hydrolysis. In the series of cycloalkanethiocarboxylates (nos. 6–10) in Table 2 , the reactivity increased with the number of the ring carbons up to six. The cyclohexanethiocarboxyl (no. 9) derivative exhibited the highest reactivity and the lowest self-hydrolysis rate in this series. Cycloheptanethiocarboxylate (no. 10) showed a very low reactivity. The dimensions of the cycloheptane ring appear to be intolerable for the enzyme reaction. Although the reactivities of the 2-ethylbutyryl (no. 4) and 2-methylbutyryl (no. 5) derivatives did not differ significantly from that of isobutyrylthiocholine (no. 2), their aqueous solubility was lower than that of isobutyryl thiocholine as shown in Table 2 . Thus, the isobutyrate (no. 2) and cyclohexanecarboxylate (no. 9) derivatives of thiocholine were selected for detailed examinations. Table 3 shows the results for reactivities (ΔE) of isobutyryl- (no. 2) and cyclohexanecarboxyl- (no. 9) thiocholines within solutions of human serum cholinesterase (EC 3.1.1.8; 13 000 U/L) and acetylcholinesterase (EC 3.1.1.7; 13 000 U/L), with the propionyl (no. 14) and n-butyryl (no. 15) compounds as references. The reactivity of isobutyryl and cyclohexanecarboxyl compounds with acetylcholinesterase was negligible (almost equal to their rate of self-hydrolysis), demonstrating that these substrates are specific to serum cholinesterase. However, the propionyl- and n-butyryl-thiocholines showed some reactivity with acetylcholinesterase. Table 3. Specificity of selected compounds as substrates of human serum cholinesterase. Compound no. . Substrate reactivity, U/L (ΔE/min)1 . . . . HSChE2 . AChE3 . Self-hydrolysis . 2 349 (0.155) 13.5 (0.006) 13.5 (0.006) 9 789 (0.350) 4.5 (0.002) 4.5 (0.002) 144 1579 (0.70) 40.2 (0.018) 28.0 (0.012) 1545 2301 (1.02) 56.4 (0.025) 9.5 (0.004) Compound no. . Substrate reactivity, U/L (ΔE/min)1 . . . . HSChE2 . AChE3 . Self-hydrolysis . 2 349 (0.155) 13.5 (0.006) 13.5 (0.006) 9 789 (0.350) 4.5 (0.002) 4.5 (0.002) 144 1579 (0.70) 40.2 (0.018) 28.0 (0.012) 1545 2301 (1.02) 56.4 (0.025) 9.5 (0.004) 1 The concentration of the enzyme was 13 000 U/L. The reaction was carried out at 30°C and pH 8.0, unless otherwise noted. 2 Human serum cholinesterase. 3 Acetylcholinesterase. 4 The 130 U/L enzyme was solution was used for the measurement, and the reactivity values presented here are those multiplied by 100. 5 The reaction was at 30°C and pH 7.0. Open in new tab Table 3. Specificity of selected compounds as substrates of human serum cholinesterase. Compound no. . Substrate reactivity, U/L (ΔE/min)1 . . . . HSChE2 . AChE3 . Self-hydrolysis . 2 349 (0.155) 13.5 (0.006) 13.5 (0.006) 9 789 (0.350) 4.5 (0.002) 4.5 (0.002) 144 1579 (0.70) 40.2 (0.018) 28.0 (0.012) 1545 2301 (1.02) 56.4 (0.025) 9.5 (0.004) Compound no. . Substrate reactivity, U/L (ΔE/min)1 . . . . HSChE2 . AChE3 . Self-hydrolysis . 2 349 (0.155) 13.5 (0.006) 13.5 (0.006) 9 789 (0.350) 4.5 (0.002) 4.5 (0.002) 144 1579 (0.70) 40.2 (0.018) 28.0 (0.012) 1545 2301 (1.02) 56.4 (0.025) 9.5 (0.004) 1 The concentration of the enzyme was 13 000 U/L. The reaction was carried out at 30°C and pH 8.0, unless otherwise noted. 2 Human serum cholinesterase. 3 Acetylcholinesterase. 4 The 130 U/L enzyme was solution was used for the measurement, and the reactivity values presented here are those multiplied by 100. 5 The reaction was at 30°C and pH 7.0. Open in new tab The effects of pH variations between 7.0 and 9.0 on the reactivity and the rate of self-hydrolysis were examined for the isobutyrate (no. 2) and cyclohexane-carboxylate (no. 9) derivatives. The maximum reactivity was observed at pH 8–8.25 (100%), as shown in Fig. 1 . The reactivity of the blank reaction (self-hydrolysis) was <5% in the same pH range. Thus, the selection of pH 8.0 in the standard procedure is justified. The effects of pH variations between 7.0 and 8.5 for the propionate (no. 14) are shown in Table 4 . Whereas the enzymatic and the self-hydrolytic reactivities were much higher in this pH range, the pH dependence was somewhat similar to those of the isobutyrate (no. 2) and cyclohexanecarboxylate (no. 9). Figure 1. Open in new tabDownload slide Influence of the pH on the enzymatic and self-hydrolytic reactions of selected substrate. Compound no. 2 (isobutyrate): ○, enzymatic reaction; □, self-hydrolysis. Compound no. 9 (cyclohexane-carboxylate): •, enzymatic reaction; ▪, self-hydrolysis. Figure 1. Open in new tabDownload slide Influence of the pH on the enzymatic and self-hydrolytic reactions of selected substrate. Compound no. 2 (isobutyrate): ○, enzymatic reaction; □, self-hydrolysis. Compound no. 9 (cyclohexane-carboxylate): •, enzymatic reaction; ▪, self-hydrolysis. Table 4. pH dependence of reactivity of propionylthiocholine.1 pH . Reactivity U/L (ΔE/min) . . Self-hydrolysis, U/L (ΔE/min) . . . 30°C . 37°C . 30°C . 37°C . 7.0 1421 (0.63) 2761 (1.22) 9.3 (0.004) 12.2 (0.005) 7.6 1502 (0.67) 2914 (1.29) 12.6 (0.006) 14.4 (0.006) 8.0 1579 (0.70) 3068 (1.36) 28.0 (0.012) 34.3 (0.015) 8.5 1574 (0.70) 3063 (1.36) 50.5 (0.022) 60.0 (0.027) pH . Reactivity U/L (ΔE/min) . . Self-hydrolysis, U/L (ΔE/min) . . . 30°C . 37°C . 30°C . 37°C . 7.0 1421 (0.63) 2761 (1.22) 9.3 (0.004) 12.2 (0.005) 7.6 1502 (0.67) 2914 (1.29) 12.6 (0.006) 14.4 (0.006) 8.0 1579 (0.70) 3068 (1.36) 28.0 (0.012) 34.3 (0.015) 8.5 1574 (0.70) 3063 (1.36) 50.5 (0.022) 60.0 (0.027) 1 The concentration of the enzyme used for the measurement was 130 U/L. The reactivity values were those multiplied by 100 to make the standard identical with those shown in Tables 2 and 3 . Open in new tab Table 4. pH dependence of reactivity of propionylthiocholine.1 pH . Reactivity U/L (ΔE/min) . . Self-hydrolysis, U/L (ΔE/min) . . . 30°C . 37°C . 30°C . 37°C . 7.0 1421 (0.63) 2761 (1.22) 9.3 (0.004) 12.2 (0.005) 7.6 1502 (0.67) 2914 (1.29) 12.6 (0.006) 14.4 (0.006) 8.0 1579 (0.70) 3068 (1.36) 28.0 (0.012) 34.3 (0.015) 8.5 1574 (0.70) 3063 (1.36) 50.5 (0.022) 60.0 (0.027) pH . Reactivity U/L (ΔE/min) . . Self-hydrolysis, U/L (ΔE/min) . . . 30°C . 37°C . 30°C . 37°C . 7.0 1421 (0.63) 2761 (1.22) 9.3 (0.004) 12.2 (0.005) 7.6 1502 (0.67) 2914 (1.29) 12.6 (0.006) 14.4 (0.006) 8.0 1579 (0.70) 3068 (1.36) 28.0 (0.012) 34.3 (0.015) 8.5 1574 (0.70) 3063 (1.36) 50.5 (0.022) 60.0 (0.027) 1 The concentration of the enzyme used for the measurement was 130 U/L. The reactivity values were those multiplied by 100 to make the standard identical with those shown in Tables 2 and 3 . Open in new tab The influence of the buffer concentration was examined for the reactivity of isobutyrate (no. 2) and cyclohexanecarboxylate (no. 9) in the range between 10 and 1000 mmol/L. The maximum reactivity was 100 mmol/L at pH 8.0 and lower at increased concentrations (∼70% in 1000 mmol/L). The Km value was calculated from the plot of ΔE values against the reciprocals of the concentrations for the substrate. The Km for both the isobutyrate and cyclohexane-carboxylate derivatives was ∼6.9 × 10−3 mmol/L. Various concentrations of the human serum cholinesterase were assayed by the general procedure. Very good linearities (r >0.999) were obtained for isobutyrate (no. 2) and cyclohexane-carboxylate (no. 9) substrates up to 1800 and 3000 U/L, respectively. The pseudo first-order degradation (hydrolysis) rate constants in the aqueous buffer solution (pH 8) for the thiocholine ester iodides at various temperatures are listed in Table 5 . The log k value and t0.8 (20% loss time) at 5 °C were estimated from the log k values at higher temperatures. With storage at 5 °C, the 20% loss time of the isobutyryl (no. 2) and cyclohexanecarboxyl (no. 9) analogs was ∼5–6 years. Because the amount of substrate reagents used for the clinical assay was selected to greatly exceed that of the enzyme, the 20% loss would be of no significance for practical measurements. We also confirmed that there is no difference in the melting point or the substrate efficiency of the aqueous solution of solid samples for isobutyryl and cyclohexanecarboxyl thiocholines between those stored for 10 years at 5 °C and those that were freshly prepared. Table 5. Degradation rate of selected acylthiocholines. Compound no. . log k, 1st order, h−1 . . . . t0.8 (5°C) . . . 25°C . 37°C . 50°C . 5°C (calc.)a . ×104 h . Years . 2 −4.64 −4.34 −3.94 −5.30 4.45 5.1 9 −4.86 −4.64 −4.34 −5.36 5.06 5.7 14 −4.16 −3.68 −3.34 −4.90 1.78 2.0 Compound no. . log k, 1st order, h−1 . . . . t0.8 (5°C) . . . 25°C . 37°C . 50°C . 5°C (calc.)a . ×104 h . Years . 2 −4.64 −4.34 −3.94 −5.30 4.45 5.1 9 −4.86 −4.64 −4.34 −5.36 5.06 5.7 14 −4.16 −3.68 −3.34 −4.90 1.78 2.0 1 Calc., estimated based on Arrhenius plots using values at 25, 37, and 50°C. Open in new tab Table 5. Degradation rate of selected acylthiocholines. Compound no. . log k, 1st order, h−1 . . . . t0.8 (5°C) . . . 25°C . 37°C . 50°C . 5°C (calc.)a . ×104 h . Years . 2 −4.64 −4.34 −3.94 −5.30 4.45 5.1 9 −4.86 −4.64 −4.34 −5.36 5.06 5.7 14 −4.16 −3.68 −3.34 −4.90 1.78 2.0 Compound no. . log k, 1st order, h−1 . . . . t0.8 (5°C) . . . 25°C . 37°C . 50°C . 5°C (calc.)a . ×104 h . Years . 2 −4.64 −4.34 −3.94 −5.30 4.45 5.1 9 −4.86 −4.64 −4.34 −5.36 5.06 5.7 14 −4.16 −3.68 −3.34 −4.90 1.78 2.0 1 Calc., estimated based on Arrhenius plots using values at 25, 37, and 50°C. Open in new tab Discussion Isobutyryl- and cyclohexanecarboxyl-thiocholines exhibit a moderate reactivity with serum cholinesterase preparations at pH 8.0 with a high selectivity against acetylcholinesterase and a sufficient ratio over spontaneous hydrolysis. Propionyl- and n-butyryl-thiocholines are much more susceptible to enzymatic hydrolysis than the branched alkanoyl thiocholines, as shown in Table 2 . The reactivity of nonbranched alkanoyl thiocholines is too high for accurate measurement unless the serum sample is highly diluted. In the assay method of Dietz et al. (1), the concentration of the human serum cholinesterase is measured at pH 7.6 with propionylthiocholine as the substrate. As shown in Table 4 , the reactivity at pH 7.6 was ∼5% lower. At pH 7.0, it was ∼10% lower than that at pH 8.0. The self-hydrolysis at pH 7.6 was >50% lower than that at pH 8.0. The Dietz method carried out at the lower-than-optimum pH may be compromised by the decrease in self-hydrolysis. The ratios of the enzymatic reactivity vs self-hydrolysis of propionyl- and n-butyryl-thiocholines were among the highest (even under nonoptimum conditions). Those of isobutyrate and cyclohexane-carboxylate at pH 8 came next, as shown in Table 2 . Isobutyryl- and cyclohexanecarboxyl-thiocholines cover concentration ranges of serum cholinesterase approximately two- to threefold broader than propionyl-thiocholine (1). In addition, they are very stable in the solid state as iodide esters and in aqueous solution, making them directly utilizable as assay reagents for at least 5 years at 5 °C. In contrast, the t0.8 time for the propionyl ester iodide solution is estimated at ∼2 years. Because of the moderate reactivity that does not require high dilution of the serum samples, the isobutyrate and cyclohexanecarboxylate derivatives are favored for the serum cholinesterase assay, especially under automated conditions. References 1 Dietz AA, Rubinstein HN, Lubrano T. Colorimetric determination of serum cholinesterase and its generic variants by the propionylthiocholine-dithio-bis-(nitrobenzoic acid) procedure. Clin Chem 1973 ; 19 : 1309 -1313. Crossref Search ADS PubMed 2 Dietz AA, Rubinstein HN. Standardization of the Ellman reaction. Clin Biochem 1972 ; 5 : 136 -138. Crossref Search ADS PubMed 3 Ellman GL. A colorimetric method for determining low concentrations of mercaptans. Arch Biochem Biophys 1958 ; 74 : 443 -450. Crossref Search ADS PubMed 4 Ellman GL. Tissue sulfhydryl groups. Arch Biochem Biophys 1959 ; 82 : 70 -77. Crossref Search ADS PubMed 5 Yamada M, Marui Y, Hayashi C, Takemura S. Benzoylthiocholine derivatives as substrates for pseudocholinesterase: synthesis and application. Chem Pharm Bull (Tokyo) 1987 ; 35 : 1491 -1496. Crossref Search ADS PubMed 6 Renshow RR, Dreisbach PF, Ziff M, Gerrn D. Thioesters of choline and α-methylcholine and their physiological activity: onium compounds. J Am Chem Soc 1938 ; 60 : 1765 -1770. Crossref Search ADS 7 Garrett ER, Carper RF. Prediction of stability in pharmaceutical preparations. I. Color stability in a liquid multisulfa preparation. J Am Pharm Assoc 1955 ; 44 : 515 -519. Crossref Search ADS © 2001 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)

Yamamichi,, Hiroshi;Kasakura,, Shinpei;Yamamori,, Shunzi;Iwasaki,, Ryu;Jikimoto,, Takumi;Kanagawa,, Sugayo;Ohkawa,, Jiro;Kumagai,, Shunichi;Koshiba,, Masahiro

doi: 10.1093/clinchem/47.11.1967pmid: N/A

Abstract Background: We describe a 56-year-old woman admitted to the hospital with a diagnosis of acute myocardial infarction without an increase of serum creatine kinase (CK) activity during her clinical course. She died on the 11th hospital day, and the diagnosis was confirmed by autopsy. The patient had had no previous muscular symptoms. Methods: Expression of the CK-muscle (CK-M) protein in cardiac tissue was examined by immunoblotting and immunochemical staining. CK-M mRNA expression was estimated by semiquantitative reverse transcription–PCR. Gene structure of CK-M was determined by Southern blotting and direct sequencing of 2251 bp. Existence of a point mutation in the CK-M gene was examined by restriction fragment length polymorphism analysis of PCR products (PCR-RFLP) in the patient and in 108 controls. Results: CK-M protein in the myocardial tissue of the patient was substantially lower (103 ± 7 ng/mg protein) than in control myocardial tissue (35 800 ± 2860 ng/mg protein). Immunoreactive CK-M in the patient tissue sample was 0.3% of the value for the control sample. CK-M mRNA was 53-fold less in the patient sample compared with the control. This very low expression of CK-M mRNA was considered to be the primary reason for CK-M deficiency. Direct sequencing revealed a point mutation at residue 54 in exon 2, which was specific for the patient. No other abnormalities were found in the CK-M gene of the patient. Conclusions: This report identifies a molecular abnormality in human CK deficiency and discusses the physiologic relevance of CK-M. Creatine kinase [(CK) 1 ; EC2.7.3.2] is an 82-kDa protein consisting of two subunits, CK-muscle (CK-M) and -brain (CK-B). Three isoenzymes, CK-MM, CK-MB, and CK-BB, are formed by hybridization of CK-M and -B subunits (1). CK plays a major role in fast muscle contraction by supplying creatine phosphate, which is used for ATP production, especially under anaerobic conditions. Increased CK activity accompanied by the appearance of CK-MB isoenzyme is considered to be a most useful tool in the diagnosis of acute myocardial infarction (AMI) (2). In the only case report of human CK deficiency, in which the patient did not have a definitive muscular involvement, the mechanism(s) of CK deficiency was not investigated (3). In this report, however, we describe 56-year-old female who complained of chest pain with typical clinical findings for AMI, except for the absence of CK-MB isoenzyme activity and the absence of increased total CK activity. An autopsy confirmed that the patient suffered from AMI. Thus the experiments in this study were designed to clarify the cause of the lack of total CK activity in this patient. We identify, for the first time, the molecular mechanism that may have accounted for the major reduction in CK mRNA, which led to the deficiency in CK protein expression. case report A 56-year-old female was admitted in a state of shock to the emergency room with chest pain of 1-day duration. The patient had been well until a day before admission, except for mild diabetes, which had not required any medication. Clinical symptoms and the electrocardiogram findings (increased S-T and terminal T inversion on II, III, and aVF leads; S-T depression on V1-V4 leads) indicated that the patient suffered a severe AMI of the posterior wall. Laboratory data demonstrated inconsistent findings with this diagnosis. Total activities of the non-cardiac-specific enzymes lactate dehydrogenase (LDH) and aspartate aminotransferase were substantially increased (6900 U/L and 2550 U/L, respectively; the reference intervals for LDH and aspartate aminotransferase were 200–400 U/L and 8–40 U/L, respectively). However, the flipped pattern of LDH-1 and LDH-2 was present, with a ratio of 1.72. No increase of serum CK activity (37 U/L; reference interval, 15–130 U/L) nor any CK-MB activity was detected. Soon after admission, an intraaortic balloon pump was placed to treat the patient for cardiogenic shock. The percutaneous transluminal coronary angioplasty was unsuccessful. On the 5th hospital day, serum CK activity had dropped to 11 U/L, whereas LDH and aspartate aminotransferase activities were 2055 and 70 U/L, respectively. The patient died on the 11th hospital day, and an autopsy was performed. Pathologic studies confirmed the infero-postlateral wall infarction of the heart. The patient had two daughters by normal delivery, neither of whom had experienced any muscular or cardiac symptoms. Materials and Methods The studies presented here were performed in muscle samples and sera after family members of the patient and a nonrelated control had given their approval on behalf of each. Informed consent was also obtained from healthy volunteers, from whom DNA samples were obtained and analyzed for the point mutation of CK. immunoblotting for the detection of ck-m subunit Right ventricular cardiac tissue, skeletal muscle (pectoralis major muscle), and smooth muscle (muscle layer of the rectum) from the patient were obtained 4 h after death and stored at −80 °C for 1 week before examination. The control tissues were obtained from a nonrelated patient who had died of acute myelogenous leukemia. To identify the CK-M protein expression, 15 μg of homogenates from cardiac and skeletal muscles of the patient and from the control, along with purified human CK-MM (BiosPacific) were immunoblotted with polyclonal goat anti-CK-M (BiosPacific). A gel prepared in parallel was stained with Coomassie Brilliant Blue G-250 (Bio-Rad) to ensure that equivalent amounts of proteins were electrophoresed on each lane. determination of total ck, ck isoenzyme, and ck subunit activities Total CK activity was determined by the reverse reaction with NADH production (CK-NAC; Boehringer). CK isoenzymes were separated on agarose films (CK isoenzyme gel 8; Corning) and the enzyme activity of each isoenzyme was determined with Cardiotrack CK (Corning) according to the manufacturer’s instructions. Total CK activities were measured on 10 different parts each of right ventricular cardiac tissue, skeletal muscle (pectoralis major muscle), and smooth muscle (muscle layer of the rectum) from the patient and control. Immunoreactive CK-M and CK-B subunits were measured on 5 different parts each of right ventricular cardiac tissue, skeletal muscle, and smooth muscle from the patient and control by the sensitive sandwich enzyme immunoassay method previously described (4)(5). immunohistochemical detection of ck-m subunit Patient and control pectoralis major muscle and cardiac muscle (10 g each) were fixed in 200 mL/L formalin for 15 h at room temperature. Sections of 4-μm thickness were stained with rabbit anti-human CK-M antibody (Ab), which was a kind gift from Dr. Kato (Department of Biochemistry, Institute for Developmental Research, Aichi Prefectural Colony, Kamily, Kasugai, Aichi, Japan) (4), followed by visualization with the aid of alkaline phosphatase-conjugated goat anti-rabbit IgG (Pierce Chemical Co.) and p-nitrophenyl phosphate (Pierce) as a substrate. extraction of genomic dna and southern blot analysis of ck-m gene After the proteinase K digestion, high-molecular weight genomic DNA was extracted from the pectoris major muscles of both the patient and the control. DNA from each sample (5 μg each) was digested with the restriction enzymes EcoRI, BamHI, HindIII, or TaqI (Boehringer Mannheim), and Southern hybridization was performed with a 32P-labeled CK-M cDNA probe obtained from the ATCC (6). determination of ck-m and β-actin mrna expression by semiquantitative reverse transcription-pcr Total RNA was prepared from the cardiac muscle of the patient and the control by the single-step method described by Chomczynski and Sacchi (7). The first-strand cDNA was synthesized with the cDNA Synthesis Kit (Boehringer Mannheim) according to the manufacturer’s instructions. The quality of the cDNA was checked by PCR amplification of a so-called “housekeeping” gene, β-actin. The expected size of the PCR product for β-actin was 305 bp. PCR primers for the CK-M gene amplification were designed by Bailly et al. (8). A set of primers, nos. 1 and 16, was used for the amplification of the entire open reading frame of CK-M mRNA (9), with an expected size of the PCR product of 1378 bp. The amplification reaction was terminated during the exponential phase, and one-tenth of the PCR mixture was electrophoresed on a 3% (w/v) NuSieve 3:1 agarose gel (FMC BioProducts) followed by ethidium bromide staining. The amount of each PCR product was compared by densitometric measurement of the band (Atto Densitograph; Atto). PCR quantification of β-actin was used to normalize the differences in cDNA quantity between samples. Previous examination revealed that up to a twofold difference between two independent quantifications of the same sample was not significant (10) (data not shown). sequencing of the ck-m gene Three sets of primers (nos. 1 and 5, 4 and 9, and 10 and 16) were used to amplify exons 1–3, 4–5, and 6–8, respectively, of the CK-M cDNA from the cardiac muscle of the patient (8). We purified each PCR product and determined the DNA sequence directly using the Taq DyeDeoxy Terminator Cycle Sequence reagent set (Applied Biosystems) according to the manufacturer’s instructions. Sequencing gel electrophoresis and the data analyses were performed on a DNA sequencer (Model 373 A; Applied Biosystems). restriction fragment length polymorphism analysis of pcr products for genomic dna from healthy controls As shown in the Results section, an A→G transition was found in exon 2 of the CK-M gene of the patient, which eliminated the restriction-recognition sequence for the restriction enzyme Tth111I. To investigate the occurrence and frequency of this mutation in the healthy population, genomic DNA was extracted from the peripheral blood mononuclear cells of nonrelated healthy volunteers (58 males and 50 females). Exon 2 of CK-M was amplified with primer nos. 2 and 3 (8), and one-half of the PCR products was digested by Tth111I (Boehringer Mannheim) at 65 °C for 1 h. Undigested and digested PCR products (10 μL each) were electrophoresed on a 3% (w/v) NuSieve 3:1 agarose gel and examined by ethidium bromide staining. Results The patient’s serum CK activity was low and did not increase during the entire disease course, whereas no CK-MB isoenzyme activity was detected. As shown in Table 1 , total CK activity of the patient’s tissues was 1/10 for skeletal muscle, 1/35 for cardiac muscle, and 1/24 for smooth muscle compared with those of the control tissues. The patient’s CK-B was decreased in cardiac and smooth muscles compared with the control. Although the patient’s skeletal muscle CK-B was higher than that of the control, some of skeletal muscle CK-M and CK-B of the patient was only ∼2% of the value for the control. Table 1. Total CK activities and immunoreactive CK-M and CK-B in patient and control muscle samples (mean ± SD). . Total CK, mU/g (wet weight) . CK-M, ng/mg of protein . CK-B, ng/mg of protein . Control Skeletal muscle 951 000 ± 76 080 98 000 ± 7840 90 ± 7 Cardiac muscle 516 000 ± 40 250 35 800 ± 2860 360 ± 29 Smooth muscle 239 000 ± 17 690 1460 ± 100 6500 ± 455 Patient Skeletal muscle 94 700 ± 7390 550 ± 40 1830 ± 130 Cardiac muscle 14 500 ± 1100 103 ± 7 271 ± 6 Smooth muscle 10 300 ± 710 5 ± 0.2 1151 ± 80 . Total CK, mU/g (wet weight) . CK-M, ng/mg of protein . CK-B, ng/mg of protein . Control Skeletal muscle 951 000 ± 76 080 98 000 ± 7840 90 ± 7 Cardiac muscle 516 000 ± 40 250 35 800 ± 2860 360 ± 29 Smooth muscle 239 000 ± 17 690 1460 ± 100 6500 ± 455 Patient Skeletal muscle 94 700 ± 7390 550 ± 40 1830 ± 130 Cardiac muscle 14 500 ± 1100 103 ± 7 271 ± 6 Smooth muscle 10 300 ± 710 5 ± 0.2 1151 ± 80 Open in new tab Table 1. Total CK activities and immunoreactive CK-M and CK-B in patient and control muscle samples (mean ± SD). . Total CK, mU/g (wet weight) . CK-M, ng/mg of protein . CK-B, ng/mg of protein . Control Skeletal muscle 951 000 ± 76 080 98 000 ± 7840 90 ± 7 Cardiac muscle 516 000 ± 40 250 35 800 ± 2860 360 ± 29 Smooth muscle 239 000 ± 17 690 1460 ± 100 6500 ± 455 Patient Skeletal muscle 94 700 ± 7390 550 ± 40 1830 ± 130 Cardiac muscle 14 500 ± 1100 103 ± 7 271 ± 6 Smooth muscle 10 300 ± 710 5 ± 0.2 1151 ± 80 . Total CK, mU/g (wet weight) . CK-M, ng/mg of protein . CK-B, ng/mg of protein . Control Skeletal muscle 951 000 ± 76 080 98 000 ± 7840 90 ± 7 Cardiac muscle 516 000 ± 40 250 35 800 ± 2860 360 ± 29 Smooth muscle 239 000 ± 17 690 1460 ± 100 6500 ± 455 Patient Skeletal muscle 94 700 ± 7390 550 ± 40 1830 ± 130 Cardiac muscle 14 500 ± 1100 103 ± 7 271 ± 6 Smooth muscle 10 300 ± 710 5 ± 0.2 1151 ± 80 Open in new tab On the immunoblot with anti-human CK-M shown in Fig. 1A , the control skeletal and cardiac muscles demonstrated single bands of 43 kDa (lanes 3 and 5, respectively), which corresponded to the band from the purified CK-M protein (lane 1). No bands appeared in the skeletal and cardiac muscle samples from the patient (lanes 2 and 4, respectively). Coomassie Brilliant Blue staining of the gel with sodium dodecyl sulfate–polyacrylamide gel electrophoresis prepared in parallel confirmed that equivalent amounts of total protein were loaded on each lane (data not shown). Figure 1. Open in new tabDownload slide Immunoblotting for CK-M of the muscle tissue extracts and CK isoenzyme study of sera on agarose film. (A), immunoblotting for CK-M. Lane 1, purified human CK-M; lane 2, cardiac muscle extract from the patient; lane 3, cardiac muscle extract from the control; lane 4, skeletal muscle from the patient; lane 5, skeletal muscle extract from the control. (B), CK isoenzyme study on agarose film. Top panel, control sera; bottom panel, patient sera. Saline, control absorption with saline; anti-CK-M, absorption with anti-human CK-M Ab. Figure 1. Open in new tabDownload slide Immunoblotting for CK-M of the muscle tissue extracts and CK isoenzyme study of sera on agarose film. (A), immunoblotting for CK-M. Lane 1, purified human CK-M; lane 2, cardiac muscle extract from the patient; lane 3, cardiac muscle extract from the control; lane 4, skeletal muscle from the patient; lane 5, skeletal muscle extract from the control. (B), CK isoenzyme study on agarose film. Top panel, control sera; bottom panel, patient sera. Saline, control absorption with saline; anti-CK-M, absorption with anti-human CK-M Ab. Serum and myocardial CK isoenzymes from the control and patient were separated on an agarose film (Fig. 1B ). Serum CK-MM activity was substantially reduced, and no serum CK-MB activity was detected in the patient sample. Because of the great reduction of CK activity, it was necessary to apply ∼30-fold more of the protein from the patient sample on the gel to visualize the bands. Consistent with the CK activities (Table 1 ) and immunoblotting results (Fig. 1A ), CK-MB in cardiac muscles was detected in the patient sample, but its activity was substantially reduced. There was essentially no detectable CK-MM band in the cardiac muscle of the patient. CK-MM was the major component of the CK isoenzymes in the control cardiac muscle (data not shown). CK-BB was not visible in the control and patient samples as expected. As shown in Fig. 1B , the addition of anti-human CK-M Ab completely eliminated the control serum CK-MM. The Ab also eliminated the CK-MB band of the control serum, presumably because of the reactivity of the Ab to the CK-M component of CK-MB isoenzyme. In the patient samples, however, the faint CK-MM band in serum was not completely adsorbed by the Ab, suggesting the existence of the dimer form, mitochondrial CK (CK-mit). In the immunohistochemical examination shown in Fig. 2 , the control myocardium yielded positive staining (Fig. 2A ), whereas the anti-human CK-M Ab did not detect the CK-M protein in the heart muscle of the patient (Fig. 2B ). Figure 2. Open in new tabDownload slide Immunohistochemical detection of CK-M subunit (original magnification, ×400). Sections of myocardium of the control (A) and the patient (B) were stained with anti-CK-M, followed by visualization by alkaline phosphatase and p-nitrophenyl phosphate. Figure 2. Open in new tabDownload slide Immunohistochemical detection of CK-M subunit (original magnification, ×400). Sections of myocardium of the control (A) and the patient (B) were stained with anti-CK-M, followed by visualization by alkaline phosphatase and p-nitrophenyl phosphate. To identify the underlying molecular mechanism of this deficiency, we first performed Southern blot analysis for the CK-M gene using the genomic DNA of muscles extracted from the patient and control. The blot did not show any major abnormalities on the genomic DNA for CK-M of the patient (Fig. 3 ). We next examined the CK-M mRNA expression to determine whether the observed defect in CK-M protein expression was attributable to a transcriptional abnormality. During exponential PCR amplification, cDNA from the myocardium of the patient did yield a faint band for the CK-M gene (Fig. 4 , top panel), whereas the β-actin expression in the patient sample was comparable to that of the control (Fig. 4 , lower panel). Two independent quantifications with the applied mRNA amount normalized by means of β-actin expression revealed that the patient’s CK-M mRNA was 68.2- and 37.7-fold (mean, 53-fold) lower than that of the control. Figure 3. Open in new tabDownload slide Southern blot analysis of CK-M gene. Genomic DNAs from pectoral muscles of both the control and the patient were digested with the restriction enzymes shown, and the Southern blot hybridization was performed with 32P-labeled CK-M cDNA as a probe. E, EcoRI; B, BamHI; H, HindIII; T, TaqI. Figure 3. Open in new tabDownload slide Southern blot analysis of CK-M gene. Genomic DNAs from pectoral muscles of both the control and the patient were digested with the restriction enzymes shown, and the Southern blot hybridization was performed with 32P-labeled CK-M cDNA as a probe. E, EcoRI; B, BamHI; H, HindIII; T, TaqI. Figure 4. Open in new tabDownload slide Semiquantitative reverse transcription–PCR analysis of CK-M gene expression in cardiac muscle. Top panel, CK-M; bottom panel, β-actin. Figure 4. Open in new tabDownload slide Semiquantitative reverse transcription–PCR analysis of CK-M gene expression in cardiac muscle. Top panel, CK-M; bottom panel, β-actin. Direct sequencing of the whole CK-M cDNA, as shown in Fig. 5 , identified a single nucleotide transition of A→G in exon 2, which led to the amino acid substitution of glycine (GGC) for aspartate (GAC; wild type) at residue 54. To determine whether the single nucleotide transition found in the patient mRNA was a polymorphism of the CK-M gene, the genomic DNA sequences from 108 healthy volunteers (58 males and 50 females) were examined. PCR-amplified fragments of exon 2 of CK-M DNA were digested by Tth111I because the mutation in the corresponding sequence should cause the destruction of the restriction site of the enzyme. All of the DNA sequences tested were successfully digested by the restriction enzyme (data not shown), which confirmed that the nucleotide transition observed was a point mutation and specific for the patient. Figure 5. Open in new tabDownload slide cDNA sequencing of the CK-M gene. A single nucleotide transition of A→G found at residue 54 in presumably the fourth α helix led to the amino acid change of wild-type aspartate (GAC; panel A) to glycine (GGC; panel B). The triple nucleotides at residue 54 are underlined. Figure 5. Open in new tabDownload slide cDNA sequencing of the CK-M gene. A single nucleotide transition of A→G found at residue 54 in presumably the fourth α helix led to the amino acid change of wild-type aspartate (GAC; panel A) to glycine (GGC; panel B). The triple nucleotides at residue 54 are underlined. Discussion We identified the transition in exon 2 of the CK gene of the patient where adenine was replaced by guanine (GAC→GGC), leading to the amino acid substitution with nonpolar neutral glycine of acidic aspartate (wild type). This mutation was specific for our patient because none of the DNAs derived from 108 nonrelated healthy volunteers showed any abnormality at this nucleotide position, as evidenced by the successful restriction digestion of PCR products (data not shown). The gene for CK-M on human chromosome 19q13.2–19q13.3 is one of the most tightly linked markers of myotonic dystrophy (DM) (11)(12). Bailly et al. (8) reported that sequencing of the CK-M cDNA from the skeletal muscle of an individual with DM revealed two novel polymorphisms but no translationally significant mutation. They concluded that, in light of genetic homogeneity shown to date for DM, a defect in the coding segment of the CK-M gene is probably not a cause in all cases of DM. Thus, evidence is lacking that mutations in the CK gene may be the cause of the muscle diseases (13). In this regard, the generation and analysis of CK-M-deficient animals are of great importance for an understanding of the true physiologic roles of CK. CK-M-deficient mice were first described by van Deursen et al. (14). Similar to our case, the mutant mice were fertile and appeared to have no abnormalities. Although a detailed examination of the muscular system of the mutant mice revealed the inability to perform the burst activity; phosphocreatine and ATP concentrations appeared normal in CK-M-deficient skeletal muscles. In addition, the recovery rates of phosphocreatine and inorganic phosphoric acid concentrations after muscular stimulation had stopped were also comparable, suggesting that the hydrolyzed phosphocreatine was replenished via a CK-mit reaction. CK-mit is a distinct isoenzyme of CK that is associated with the outer face of the mitochondrial inner membrane and is functionally linked to oxidative phosphorylation (15). CK-mit has been detected in normal human heart and liver (16), as well as sera from patients with AMI (17). More than 90% of tissue CK-mit is octameric (18) and exhibits cathodal migration on electrophoresis because of its positive charge, whereas CK-BB and CK-MB are negatively charged and CK-MM is neutral (19)(20). However, Kanemitsu et al. (16) showed that serum octameric CK-mit displayed a tendency to dissociate into stable dimers, which stayed in the same position as CK-MM during electrophoresis. On the agarose film electrophoresis (Fig. 1B ), a faint band remained at the position of CK-MM in the serum of the patient treated with the CK-M Ab. The finding that the CK-MM of the control serum, which was much more abundant than that of the patient, was completely adsorbed by the anti-CK-M Ab suggests that the residual CK-MM of the patient was also adsorbed. Therefore, this faint band at the CK-MM position in the serum of the patient may represent the CK-mit of dimeric form. We did not reliably determine the CK-mit activity in the cardiac tissue of the patient, mainly because of the limited availability of samples. It is reported, however, that in normal cardiac cells, the mitochondrial component contains 30–40% of the cardiac cell CK activity (21). We believe that CK-mit compensated for most of the CK-M functions in our patient. Alternatively, CK function may be compensated by the different kinase(s). Dzeja et al. (22) reported an apparent compensatory shift in phosphotransfer catalysis from the CK to the adenylate kinase system with increasing muscle contraction or graded chemical inhibition of CK activity. Thus it is possible that the adenylate kinase system may, in part, have taken over the compromised CK function in our patient. The degree of expression of a gene is affected not only by the efficiency of transcription and/or translation, but also by the stability of mRNA (23). Because a poly(A) tail and 3′-untranslated region (UTR) are considered important for determining the fate of mRNA, we determined the 5′-UTR sequence up to 975 bp upstream of the initiation codon of CK-M mRNA and the 3′-UTR sequence 130 bp downstream of open reading frame. We found no differences in the 3′-UTR, but in the 5′-UTR, we identified 14 nucleotide differences from the sequence reported by Trask et al. (9). None of the differences in the 5′-UTR, however, was in the promoter or enhancer regions (data not shown). We also searched for E2A gene abnormalities because E2A is considered one of the essential factors for the transcription of many genes, including CK-M (24), but Southern blot analysis revealed no evident abnormality on E2A DNA (data not shown). Although the precise molecular mechanism(s) leading to the extremely low expression of CK-M mRNA in our case are unclear, the apparent absence of any abnormalities in the promoter and enhancer regions and the 3′-UTR, as well as in the transcription factor E2A gene, suggests that the observed point mutation in exon 2 of CK-M mRNA is responsible for the deficient CK-M protein expression. It will be of interest to see whether the site-directed mutagenesis of CK-M gene reproduces the CK-M deficiency similar to our patient, or whether the CK-M expression can be reconstituted when the mutated CK-M gene is introduced into the CK-M deficient mice to mimic our case. Such experiments may lead to a better understanding of the true physiologic and pathologic roles of CK in the human muscular system. The diagnosis of AMI in our patient was determined by the combination of clinical symptoms, electrocardiogram, and non-cardiac-specific biochemical markers. Shibuya et al. (3) reported a case of 46-year-old male in whom serum CK was extremely low and CK-M was absent. The patient was initially diagnosed as having a myocardial ischemia from his clinical symptoms, but further medical tests showed no evidence of ischemia. Recent studies have revealed that cardiac troponin I is uniquely specific for the heart and that cardiac troponin T if measured by the Roche cardiac troponin T assay is not interfered by the isoforms expressed in skeletal muscle (25). Furthermore, McLaurin et al. (26) reported that cardiac troponin I and cardiac troponin T concentrations were within normal limits in patients without the recent myocardial injury, in whom CK-MB activities were increased. Thus, cardiac troponin I and T should be superior to measurement of CK-MB for detecting cardiac damage. The measurement of cardiac troponin I and T would have been useful in our patient and the patient of Shibuya et al. (3) to make accurate clinical evaluation of recent myocardial injury. This work is supported in part by the Kobe City Fund for the Promotion of Medical Research. We thank Dr. Tomoko Nakamura-Wada and Mari Hyakuta (Kobe University) for their helpful discussion, and Nobuhide Hayashi (Kobe University) and Jan K. Visscher (English Language Consultant) for editorial assistance and preparation of the manuscript. 1 " Nonstandard abbreviations: CK, creatine kinase; CK-M and -B, CK-muscle and -brain, respectively; AMI, acute myocardial infarction; Ab, antibody; UTR, untranslated region; CK-mit, mitochondrial CK; and DM, myotonic dystrophy. References 1 Dawson DM, Eppenberger HM, Kaplan NO. Creatine kinase: evidence for a dimeric structure. Biochem Biophys Res Commun 1965 ; 21 : 346 -353. Crossref Search ADS PubMed 2 Katz IA, Irwig L, Vinen JD, March L, Wyndham LE, Luu T, et al. Biochemical markers of acute myocardial infarction: strategies for improving their clinical usefulness. Ann Clin Biochem 1998 ; 35 : 393 -399. Crossref Search ADS PubMed 3 Shibuya J, Matsumoto T, Takahashi K, Sugisawa K, Yasutomi N, Kawashima S, et al. The first report of a case with acute myocardial infarction showing familial deficiency of creatine kinase. Intern Med 1992 ; 31 : 611 -616. Crossref Search ADS PubMed 4 Kato K, Shimizu A. Highly sensitive enzyme immunoassay for human creatine kinase MM and MB isozymes. Clin Chim Acta 1986 ; 158 : 99 -108. Crossref Search ADS PubMed 5 Kato K, Shimizu A, Ishiguro Y, Mokuno K, Ariyoshi Y, Nakajima T. Highly sensitive enzyme immunoassay for human creatine kinase BB isozyme. Clin Chim Acta 1985 ; 150 : 31 -40. Crossref Search ADS PubMed 6 Fukutani H, Naoe T, Yoshida H, Yamamori S, Ohno R. Molecular heterogeneity of the PML gene rearrangement in acute promyelocytic leukemia: prevalence and clinical significance. Jpn J Cancer Res 1993 ; 84 : 257 -264. Crossref Search ADS PubMed 7 Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987 ; 162 : 156 -159. PubMed 8 Bailly J, MacKenzie AE, Leblond S, Korneluk RG. Assessment of a creatine kinase isoform M defect as a cause of myotonic dystrophy and the characterization of two novel CKMM polymorphisms. Hum Genet 1991 ; 86 : 457 -462. PubMed 9 Trask RV, Strauss AW, Billadello JJ. Developmental regulation and tissue-specific expression of the human muscle creatine kinase gene. J Biol Chem 1988 ; 263 : 17142 -17149. PubMed 10 Koshiba M, Apasov S, Sverdlov V, Chen P, Erb L, Weisman GA, et al. Transient up-regulation of P2Y2 nucleotide receptor mRNA expression is an immediate early gene response in activated thymocytes. Proc Natl Acad Sci U S A 1997 ; 94 : 831 -836. Crossref Search ADS PubMed 11 Brunner HG, Korneluk RG, Coerwinkel-Driessen M, MacKenzie A, Smeets H, Lambermon HM, et al. Myotonic dystrophy is closely linked to the gene for muscle-type creatine kinase (CKMM). Hum Genet 1989 ; 81 : 308 -310. Crossref Search ADS PubMed 12 Korneluk RG, MacKenzie AE, Nakamura Y, Dube I, Jacob P, Hunter AG. A reordering of human chromosome 19 long-arm DNA markers and identification of markers flanking the myotonic dystrophy locus. Genomics 1989 ; 5 : 596 -604. Crossref Search ADS PubMed 13 Wyss M, Kaddurah-Daouk R. Creatine and creatinine metabolism. Physiol Rev 2000 ; 80 : 1107 -1213. Crossref Search ADS PubMed 14 van Deursen J, Heerschap A, Oerlemans F, Ruitenbeek W, Jap P, ter Laak H, et al. Skeletal muscles of mice deficient in muscle creatine kinase lack burst activity. Cell 1993 ; 74 : 621 -631. Crossref Search ADS PubMed 15 Wyss M, Smeitink J, Wevers RA, Wallimann T. Mitochondrial creatine kinase: a key enzyme of aerobic energy metabolism. Biochim Biophys Acta 1992 ; 1102 : 119 -166. Crossref Search ADS PubMed 16 Kanemitsu F, Kawanishi I, Mizushima J. Characteristics of mitochondrial creatine kinases from normal human heart and liver tissues. Clin Chim Acta 1982 ; 119 : 307 -317. Crossref Search ADS PubMed 17 James GP, Harrison RL. Creatine kinase isoenzymes of mitochondrial origin in human serum. Clin Chem 1979 ; 25 : 943 -947. Crossref Search ADS PubMed 18 Schlegel J, Wyss M, Eppenberger HM, Wallimann T. Functional studies with the octameric and dimeric form of mitochondrial creatine kinase. Differential pH-dependent association of the two oligomeric forms with the inner mitochondrial membrane. J Biol Chem 1990 ; 265 : 9221 -9227. PubMed 19 Roberts R, Grace AM. Purification of mitochondrial creatine kinase. Biochemical and immunological characterization. J Biol Chem 1980 ; 255 : 2870 -2877. PubMed 20 Lee KN, Csako G, Bernhardt P, Elin RJ. Relevance of macro creatine kinase type 1 and type 2 isoenzymes to laboratory and clinical data. Clin Chem 1994 ; 40 : 1278 -1283. Crossref Search ADS PubMed 21 Saks VA, Kupriyanov VV, Elizarova GV, Jacobus WE. Studies of energy transport in heart cells. The importance of creatine kinase localization for the coupling of mitochondrial phosphorylcreatine production to oxidative phosphorylation. J Biol Chem 1980 ; 255 : 755 -763. PubMed 22 Dzeja PP, Zeleznikar RJ, Goldberg ND. Adenylate kinase: kinetic behavior in intact cells indicates it is integral to multiple cellular processes. Mol Cell Biochem 1998 ; 184 : 169 -182. Crossref Search ADS PubMed 23 Grunberg-Manago M. Messenger RNA stability and its role in control of gene expression in bacteria and phages. Annu Rev Genet 1999 ; 33 : 193 -227. Crossref Search ADS PubMed 24 Edmondson DG, Olson EN. Helix-loop-helix proteins as regulators of muscle-specific transcription. J Biol Chem 1993 ; 268 : 755 -758. PubMed 25 Apple FS. The specificity of biochemical markers of cardiac damage: a problem solved. Clin Chem Lab Med 1999 ; 37 : 1085 -1089. PubMed 26 McLaurin M, Apple FS, Henry TD, Sharkey SW. Cardiac troponin I and T concentrations in patients with cocaine-associated chest pain. Ann Clin Biochem 1996 ; 33 : 183 -186. Crossref Search ADS PubMed © 2001 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)

Devaraj,, Sridevi;Hirany, Shaina, V;Burk, Raymond, F;Jialal,, Ishwarlal

doi: 10.1093/clinchem/47.11.1974pmid: N/A

Abstract Background: Oxidative stress is pivotal in atherogenesis. Although the most widely used indirect assay to quantify oxidative stress is LDL oxidative susceptibility, direct assays such as urinary F2-isoprostanes have shown great promise. Methods: We evaluated the utility of both a direct measure of oxidative stress (urinary F2-isoprostanes) and an indirect measure of copper-catalyzed, LDL oxidation in a model of increased oxidative stress (diabetes). We also evaluated an enzyme immunoassay (EIA) method for urinary F2-isoprostanes with a gas chromatography–mass spectrometry method. Results: Excellent intraassay and interassay CVs of <4% were obtained with our EIA method. A good correlation was obtained between the two methods (r = 0.80; n = 68) of F2-isoprostane measurement. An excellent correlation for F2-isoprostane concentrations was obtained between a timed collection vs 24-h urine (r = 0.96; n = 46). Baseline F2-isoprostane concentrations by EIA were significantly higher in both type 2 diabetics with and without macrovascular complications compared with controls (P <0.001). Supplementation with α-tocopherol led to a significant reduction in F2-isoprostane concentrations in all diabetic patients compared with baseline values (2.51 ± 1.76 compared with 1.69 ± 1.32 ng/mg creatinine; P <0.001). There were no significant differences in LDL oxidation in both diabetic groups compared with controls. α-Tocopherol supplementation led to significant increases in the lag phase of oxidation as measured by 3 indices in all groups. Conclusions: The measurement of urinary F2-isoprostanes provides a direct measure of in vivo lipid peroxidation and oxidative stress and appears to be superior to an indirect measure, e.g., LDL oxidative susceptibility, in type 2 diabetes. Atherosclerotic vascular disease is the major cause of mortality and morbidity in the US. Data continue to accrue supporting the hypothesis that oxidative stress is pivotal in the genesis of the atherosclerotic lesion (1)(2). There are several direct as well as indirect measures for assaying oxidative stress. Although the most widely used indirect method for measuring oxidative stress is the measurement of LDL oxidative susceptibility (3), direct assays such as measurement of urinary F2-isoprostanes have shown great promise. Much evidence implicates oxidative modification of LDL in the pathogenesis of atherosclerosis (4)(5). The diabetic state confers an increased propensity to accelerated atherosclerosis. Factors that may contribute to increased oxidative stress in diabetic patients include antioxidant deficiencies (decreased ascorbate, glutathione, and superoxide dismutase), protein glycation (glucooxidation), and increased production of reactive oxygen species (superoxide, hydrogen peroxide) (6)(7). Other evidence for increased oxidative stress in diabetes includes increased oxidative DNA damage as well as increased titers of autoantibodies to oxidized LDL (8)(9)(10). However, data on the oxidizability of LDL in diabetic patients are conflicting (8)(11)(12)(13)(14). Recently, the discovery of F2-isoprostanes, which are prostaglandin-like compounds produced in vivo by free radical peroxidation of arachidonic acid, has allowed for the direct assessment of in vivo lipid peroxidation in plasma (15)(16)(17). Thus, quantification of F2-isoprostanes may provide a reliable direct measure of oxidative stress in vivo. Increased concentrations of F2-isoprostanes have been reported in type 2 diabetic patients (18)(19), further underscoring the increased oxidative stress present the diabetic state. The major drawback in measurement of F2-isoprostanes is that the methodology involves gas chromatography–mass spectrometry (GC-MS), 1 which, although sensitive and accurate, is laborious and time- consuming and may not be available in many laboratories. This has prompted development of commercial immunoassay methods for the measurement of F2-isoprostanes. To date, no studies have compared a direct and indirect measure of increased oxidative stress; the present study was undertaken to evaluate the utility of these indices of oxidative stress, i.e., F2-isoprostanes and LDL oxidative susceptibility in a model of increased oxidative stress, type 2 diabetes. We also evaluated an enzyme immunoassay (EIA) method for measurement of urinary F2-isoprostanes and validated it against the “gold standard” GC-MS method. Furthermore, we assessed the effect of α-tocopherol supplementation on urinary F2-isoprostanes in these diabetic patients. Materials and Methods patients This study was approved by the Institutional Review Board. Volunteers (n = 75) were recruited after giving informed consent. They were divided into three groups: type 2 diabetic patients without macrovascular complications [(DM2); n = 25]; type 2 diabetic patients with macrovascular complications (DM2-MV; n = 25); and age- and sex-matched healthy controls (n = 25). Selection criteria for the participants have been described previously (20). All volunteers gave informed consent. Fasting blood was obtained from all the participants at baseline and after 3 months of supplementation with all-rac α-tocopherol (1200 IU/day) and after a 2-month washout phase. Urine samples collected after 24 h were stored at −70 °C for F2-isoprostane analysis by EIA and GC-MS. Plasma samples were stored at −70 °C and analyzed for α-tocopherol by HPLC and for LDL oxidation as described previously (20). For comparison of F2-isoprostane concentrations in timed vs 24-h urine collection, 46 volunteers were requested to collect a first morning urine sample and a 24-h collection sample on different days. Samples were frozen at −70 °C until analysis of F2-isoprostane concentrations by EIA. ldl isolation and oxidation Fasting blood (60 mL) anticoagulated with EDTA was obtained for studies of LDL oxidation. LDL (density, 1.019–1.063 kg/L) was isolated by preparative ultracentrifugation from plasma collected in EDTA (1 g/L) as described previously (21). After dialysis against 150 mmol/L NaCl and 1 mmol/L EDTA (pH 7.4), LDL was filtered and protein content was measured as described previously (21). After overnight dialysis against phosphate-buffered saline, pH 7.4, LDL (100 ng/L protein) was oxidized with 5 μmol/L copper at 37 °C for 8 h. The time points were 0, 0.5, 1, 1.5, 2, 2.5, 3, 4, 5, and 8 h, respectively. At the respective time points, LDL oxidation was arrested with 200 μmol/L EDTA and 40 μmol/L butylated hydroxytoluene followed by refrigeration. We used three assays to quantify LDL oxidation to better appreciate lipid peroxidation and aldehydic modification of apolipoprotein B-100 (apo B-100) (4)(22). We determined the amount of conjugated dienes formed by monitoring the absorbance of the LDL sample at 234 nm at various time points using a phosphate-buffered saline blank (21). We measured the lipid peroxide content of LDL by the ferrous oxide-xylenol orange method (23), and we measured apo B fluorescence of LDL samples after dilution in a spectrofluorometer with excitation set at 360 nm and emission at 430 nm using a 5-nm slit width (21). We computed the lag phase of oxidation using the time course curve (21) from the three indices of oxidation. eia analysis of f2-isoprostanes Purification and extraction of urine samples was performed before EIA analysis as reported previously (24). The pH of the urine samples was adjusted to <4.0 with 1.0 mol/L HCl, and a 1-mL aliquot of urine was extracted on a Bakerbond SPE C18 column (JT Baker) that had previously been rinsed with 5 mL of methanol followed by 5 mL of ultrapure water (Cayman Chemical; catalog no. 400000). The columns were then washed with ultrapure water, allowed to dry, and then equilibrated with 5 mL of hexane. After eluting with ethyl acetate and 10 ml/L methanol, sodium acetate was added to the eluate, which was then vortex-mixed and subjected to silica gel chromatography. Elution was carried out with ethyl acetate and methanol (1:1 by volume); eluates were then dried under nitrogen and reconstituted with EIA buffer (1 mL). The extracted samples were diluted 1:5 to 1:8 with EIA buffer and assayed according to manufacturer’s protocol for the 8-Isoprostane Enzyme Immunoassay method (Cayman Chemical). The EIA exhibits >1% cross-reactivity with 8-isoprostaglandin F3 α (21%), 8-isoprostaglandin E2 (1.8%), 2,3,dinor 8-isoprostaglandin F2 α (1.7%), and 8-isoprostaglandin E1 (1.6%). We measured urine creatinine on the Olympus (Redondo, CA) by the Jaffe reaction using standard techniques. F2-Isoprostane concentrations are expressed as ng/mg of creatinine. gc-ms analysis of f2-isoprostane We measured F2-isoprostanes in urine on frozen samples after thin-layer chromatographic purification by GC negative-ion chemical ionization and electron-ionization MS using a deuterated internal standard in the laboratory of Dr. Raymond F. Burk (Vanderbilt University, Nashville, TN) (25). Catalytic hydrogenation and formation and analyses of boronate derivatives were carried out as described previously (25). statistics Linear regression analysis was performed using the Microsoft Excel software. Comparison of differences between groups was determined by Wilcoxon signed-rank test with the use of Sigma Stat software. The degree of significance was set at <0.05. Results precision Intraassay (n = 5) and interassay (n = 5) CVs of <4% for three concentrations of F2-isoprostanes were obtained for the EIA method. The precision and accuracy for the GC-MS method were ± 6% and 96%, respectively (25). The lowest limits of detection with the EIA and GC-MS methods were 4 ng/L and 5 ng/L, respectively. validation of eia vs gc-ms The EIA assay was validated for F2-isoprostanes determination in urine by comparison with the GC-MS method. A good correlation was obtained between the two methods (r = 0.80; n = 68; Fig. 1 ). The median F2-isoprostane concentrations for the EIA and GC-MS methods were 1.74 and 1.70 ng/mg of creatinine, respectively. Figure 1. Open in new tabDownload slide Validation of enzyme immunoassay for F2-isoprostanes with GC-MS. F2-Isoprostane concentrations were analyzed in urine samples (n = 68) by both the EIA method and by the GC-MS as described in Materials and Methods. Figure 1. Open in new tabDownload slide Validation of enzyme immunoassay for F2-isoprostanes with GC-MS. F2-Isoprostane concentrations were analyzed in urine samples (n = 68) by both the EIA method and by the GC-MS as described in Materials and Methods. comparison of timed vs 24-h urine collection We assessed the variation in F2-isoprostane excretion between the first morning urine collection and 24-h urine collection in 46 volunteers. F2-Isoprostane concentrations were 0.41–10.6 ng/mg of creatinine. An excellent correlation for F2-isoprostane concentrations was obtained between the two collection methods (r = 0.96; n = 46, Fig. 2 ). The mean F2-isoprostane concentrations in timed vs 24-h collection were 2.26 ± 1.79 and 2.38 ± 2.11 ng/mg of creatinine, respectively. Figure 2. Open in new tabDownload slide Comparison of early morning vs 24-h urine for measurement of isoprostanes. Volunteers (n = 46) were asked to collect an early morning urine sample as well as a 24-h urine collection on different days. Samples were frozen at −70 °C, and F2-isoprostanes were analyzed in both samples by EIA as described in Materials and Methods. Figure 2. Open in new tabDownload slide Comparison of early morning vs 24-h urine for measurement of isoprostanes. Volunteers (n = 46) were asked to collect an early morning urine sample as well as a 24-h urine collection on different days. Samples were frozen at −70 °C, and F2-isoprostanes were analyzed in both samples by EIA as described in Materials and Methods. f2-isoprostane concentrations in diabetic patients F2-Isoprostane concentrations by EIA were significantly higher in DM2 and DM2-MV patients compared with age- and sex-matched controls (2.03 ± 1.17 and 2.61 ± 1.53 ng/mg creatinine, respectively, compared with 0.71 ± 0.35 ng/mg creatinine; P <0.001, Fig. 3 ); however, no significant differences existed at baseline between DM2 and DM2-MV for F2-isoprostane concentrations (P = 0.37). We examined the effect of α-tocopherol supplementation (1200 IU/day) on urinary excretion of F2-isoprostane in diabetic patients with and without macrovascular complications, using the EIA and GC-MS methods. Figure 3. Open in new tabDownload slide F2-Isoprostane concentrations in type 2 diabetic patients. Twenty-four-hour urine samples were obtained from controls and DM2-MV and DM2 patients and stored at −70 °C. Data represent 25th percentile, median, and 75th percentile of F2-isoprostane concentrations by EIA in the three groups. ∗, P <0.001 by Wilcoxon signed-rank test. Figure 3. Open in new tabDownload slide F2-Isoprostane concentrations in type 2 diabetic patients. Twenty-four-hour urine samples were obtained from controls and DM2-MV and DM2 patients and stored at −70 °C. Data represent 25th percentile, median, and 75th percentile of F2-isoprostane concentrations by EIA in the three groups. ∗, P <0.001 by Wilcoxon signed-rank test. α-Tocopherol supplementation was associated with an increase in plasma α-tocopherol concentrations as reported previously (20). Supplementation with α-tocopherol led to a significant reduction in F2-isoprostane concentrations in the combined diabetic group compared with baseline values (2.51 ± 1.76 ng/mg of creatinine compared with 1.69 ± 1.32 ng/mg of creatinine; P <0.001). However, when the diabetic patients were subgrouped into DM2 and DM2-MV, only the reduction in DM2-MV was significant after supplementation (P <0.001; Table 1 ) in both the EIA and GC-MS methods. In addition, F2-isoprostane concentrations returned to baseline values after the 6-week washout phase in all groups (P >0.6). Table 1. Effect of α-tocopherol on F2-isoprostane concentrations in diabetic patients.1 . GS-MS . . EIA . . . DM2 . DM2-MV . DM2 . DM2-MV . Baseline 1.7 (1.5, 2.6) 2.5 (1.7, 3.2) 1.7 (1.4, 2.4) 2.4 (1.6, 3.3) α-Tocopherol phase 1.4 (1.2, 1.6) 1.42 (1, 2.1) 1.4 (1.1, 1.6) 1.52 (1, 1.7) Washout phase 1.93 (1.6, 2.4) 2.63 (1, 3.1) 2.03 (1.5, 2.3) 2.63 (1.3, 2.9) . GS-MS . . EIA . . . DM2 . DM2-MV . DM2 . DM2-MV . Baseline 1.7 (1.5, 2.6) 2.5 (1.7, 3.2) 1.7 (1.4, 2.4) 2.4 (1.6, 3.3) α-Tocopherol phase 1.4 (1.2, 1.6) 1.42 (1, 2.1) 1.4 (1.1, 1.6) 1.52 (1, 1.7) Washout phase 1.93 (1.6, 2.4) 2.63 (1, 3.1) 2.03 (1.5, 2.3) 2.63 (1.3, 2.9) 1 Data are presented as median (25th and 75th percentiles). 2 P <0.001 compared with baseline by Wilcoxon signed-rank test. 3 P >0.6 compared with baseline by Wilcoxon signed-rank test. Open in new tab Table 1. Effect of α-tocopherol on F2-isoprostane concentrations in diabetic patients.1 . GS-MS . . EIA . . . DM2 . DM2-MV . DM2 . DM2-MV . Baseline 1.7 (1.5, 2.6) 2.5 (1.7, 3.2) 1.7 (1.4, 2.4) 2.4 (1.6, 3.3) α-Tocopherol phase 1.4 (1.2, 1.6) 1.42 (1, 2.1) 1.4 (1.1, 1.6) 1.52 (1, 1.7) Washout phase 1.93 (1.6, 2.4) 2.63 (1, 3.1) 2.03 (1.5, 2.3) 2.63 (1.3, 2.9) . GS-MS . . EIA . . . DM2 . DM2-MV . DM2 . DM2-MV . Baseline 1.7 (1.5, 2.6) 2.5 (1.7, 3.2) 1.7 (1.4, 2.4) 2.4 (1.6, 3.3) α-Tocopherol phase 1.4 (1.2, 1.6) 1.42 (1, 2.1) 1.4 (1.1, 1.6) 1.52 (1, 1.7) Washout phase 1.93 (1.6, 2.4) 2.63 (1, 3.1) 2.03 (1.5, 2.3) 2.63 (1.3, 2.9) 1 Data are presented as median (25th and 75th percentiles). 2 P <0.001 compared with baseline by Wilcoxon signed-rank test. 3 P >0.6 compared with baseline by Wilcoxon signed-rank test. Open in new tab ldl oxidative susceptibility in diabetic patients LDL oxidative susceptibility was monitored by three indices of oxidation: (a) conjugated dienes, (b) lipid peroxides, and (c) apo B fluorescence. We found no significant differences in lag phase, using all three indices of LDL oxidative susceptibility, in both diabetic groups compared with controls. Furthermore, there was no significant difference between the DM2 and DM2-MV groups (Table 2 ). α-Tocopherol supplementation led to significant increases in lag phase of oxidation as measured by all three indices in the control, DM2, and DM2-MV groups, respectively, as reported previously for conjugated dienes (20). Table 2. LDL oxidative susceptibility in type 2 diabetic patients and matched controls.1 . Controls . DM2 . DM2-MV . Conjugated dienes lag phase, min 65.6 ± 11.6 58.6 ± 11.2 56.9 ± 14.9 Lipid peroxides lag phase, min 66.2 ± 8.4 64.5 ± 11.0 63.8 ± 13.9 apo B fluorescence lag phase, min 38.1 ± 18.2 30.1 ± 16.2 29.1 ± 19.2 . Controls . DM2 . DM2-MV . Conjugated dienes lag phase, min 65.6 ± 11.6 58.6 ± 11.2 56.9 ± 14.9 Lipid peroxides lag phase, min 66.2 ± 8.4 64.5 ± 11.0 63.8 ± 13.9 apo B fluorescence lag phase, min 38.1 ± 18.2 30.1 ± 16.2 29.1 ± 19.2 1 Data are expressed as mean ± SD. Open in new tab Table 2. LDL oxidative susceptibility in type 2 diabetic patients and matched controls.1 . Controls . DM2 . DM2-MV . Conjugated dienes lag phase, min 65.6 ± 11.6 58.6 ± 11.2 56.9 ± 14.9 Lipid peroxides lag phase, min 66.2 ± 8.4 64.5 ± 11.0 63.8 ± 13.9 apo B fluorescence lag phase, min 38.1 ± 18.2 30.1 ± 16.2 29.1 ± 19.2 . Controls . DM2 . DM2-MV . Conjugated dienes lag phase, min 65.6 ± 11.6 58.6 ± 11.2 56.9 ± 14.9 Lipid peroxides lag phase, min 66.2 ± 8.4 64.5 ± 11.0 63.8 ± 13.9 apo B fluorescence lag phase, min 38.1 ± 18.2 30.1 ± 16.2 29.1 ± 19.2 1 Data are expressed as mean ± SD. Open in new tab Discussion Cardiovascular disease is the major cause of mortality and morbidity in the US. Oxidative stress plays a crucial role in the genesis and progression of the atherosclerotic lesion. There are several direct as well as indirect measures for assaying oxidative stress. Although the most widely used indirect method for measuring oxidative stress is measurement of LDL oxidative susceptibility, direct assays such as measurement of urinary F2-isoprostanes have shown great promise. In this report, we have validated an EIA method for quantification of urinary F2-isoprostanes, sensitive markers of in vivo oxidative stress, and compared the method with the gold standard GC-MS method. In addition, using a model of oxidative stress, type 2 diabetes, we highlighted the divergence between LDL oxidative susceptibility as assessed by three techniques and urinary F2-isoprostanes. F2-Isoprostanes are prostaglandin-like compounds formed in vivo from free radical-catalyzed peroxidation of arachidonic acid, mainly via a noncycloxygenase-dependent mechanism. F2-Isoprostanes are found in body tissues in the esterified form and in biologic fluids, such as plasma and urine, in the free form (15)(16)(17). The relevance of measurement of urinary F2-isoprostanes with regard to atherosclerosis has been brought forth in many studies (15)(16)(17)(18)(19). F2-Isoprostanes are increased after LDL oxidation by macrophages, endothelial cells, or copper. Increased concentrations have been detected in oxidized LDL and also in patients with established risk factors for premature atherosclerosis, such as diabetes, hypercholesterolemia, and smoking (15)(16)(17)(18)(19). F2-Isoprostanes have been found to localize in foam cells in human atherosclerotic lesions. Furthermore, α-tocopherol supplementation has been found to suppress F2-isoprostanes and atherogenesis in apo-E-deficient mice. In humans, α-tocopherol supplementation has been shown to lower urinary F2-isoprostanes in patients with hypercholesterolemia or diabetes (15)(16)(17)(18)(19). Also, in a recent report, we showed that α-tocopherol supplementation (400 IU/day) can decrease urinary F2-isoprostanes in healthy volunteers (4), but this was not confirmed by another recent study (26), probably because of the study’s small sample size (n = 5). F2-Isoprostanes can be measured accurately and sensitively by a solid-phase extraction procedure, followed by selective-ion monitoring GC-MS, using tritiated prostaglandin F2-α as internal standard (25). However, although GC-MS methods are the method of choice, they are technique dependent and involve sophisticated instrumentation that is not available in most laboratories. We have shown excellent intra- and interassay precision for the EIA method, as is seen for the GC-MS method. We also demonstrate an excellent correlation with the GC-MS method, and we show that the EIA method can measure accurately on a timed vs 24-h specimen. We show substantial increases in a model of oxidative stress, type 2 diabetes, using this assay and also show that it can be modulated with antioxidant therapy. Proudfoot et al. (27) previously compared the measurement of F2-isoprostanes in urine by EIA and GC-MS and showed a poor correlation; however, only 14 samples were assayed from healthy volunteers, and it is not clear whether both methods were conducted at the same time. Wang et al. (28), however, measured F2-isoprostanes by both GC-MS and EIA in 9 healthy volunteers and showed a good correlation between the two methods (r = 0.99). Thus, the measurement of F2-isoprostanes by EIA may provide a sensitive, specific, and noninvasive method for the assessment of in vivo lipid peroxidation in humans, a method that is simpler and less expensive. Furthermore, large numbers of samples can be quantified at the same time with this method. Whereas the measurement of urinary F2-isoprostanes is a direct measure of oxidative stress, LDL oxidative susceptibility is an indirect measure of oxidative stress. LDL is isolated from plasma and then subjected to oxidative stress, whereas F2-isoprostanes are directly measured in urine (29). Furthermore, in disease states, the patients are on drugs that could potentially partition in LDL and alter its oxidative susceptibility. Thus, isoprostanes appear to be superior in this case. Also, it is possible that we may have seen significant differences in LDL oxidizability with larger sample sizes. Because urinary F2-isoprostanes could also derive from local production in the kidney, caution should be exercised in using this as a measure of oxidative stress in patients with chronic renal failure (25). It should also be pointed out that the EIA, although largely specific for 8-isoprostane, also exhibits cross-reactivity with certain other prostaglandins. Furthermore, as pointed out by Lawson et al. (30), all of the 64 possible isomers of F2-isoprostanes share the same ring structure, and it is believed that prostaglandin antigenicity is largely directed toward the ring. Thus, it is possible that the antibody in this EIA also recognizes other isoprostanes, accounting for the similar concentrations obtained with GC-MS. Although the diabetic state has been shown to have increased oxidant stress, as evidenced by increased concentrations of superoxide release and antioxidant deficiencies as well as increased F2-isoprostane concentrations by GC-MS, there is conflict in the literature with regard to in vitro susceptibility of LDL to oxidation, as assessed by the lag phase in type 2 diabetic patients (8)(11)(12)(13)(20)(31). In our study, LDL oxidative susceptibility, as measured by lag phase using three indices, conjugated dienes, lipid peroxides, and apo B fluorescence, was not significantly increased in type 2 diabetic patients (20). In this report, we also show that α-tocopherol supplementation leads to a decrease in LDL oxidative susceptibility (20). This finding confirms the reports of previous investigators (32)(33). Thus, measurement of urinary F2-isoprostanes provides direct measure of lipid peroxidation and whole body oxidative stress and appears to be superior to indirect measures, such as LDL oxidative susceptibility, in certain conditions of increased oxidative stress, such as type 2 diabetes. This work was supported by grants from the National Institutes of Health (1 RO1 AT00005, K24 AT 00596, DK 26657) and the American Diabetes Association. We thank Amy Motley for assistance with the GC-MS assays. 1 " Nonstandard abbreviations: GC-MS, gas chromatography–mass spectrometry; EIA, enzyme immunoassay; DM2, type 2 diabetic patients without macrovascular complications; DM2-MV, type 2 diabetic patients with macrovascular complications; and apo, apolipoprotein. References 1 Kunsch C, Medford RM. Oxidative stress as a regulator of gene expression in the vasculature. Circ Res 1999 ; 85 : 753 -766. Crossref Search ADS PubMed 2 Witztum JL, Berliner JA. Oxidized phospholipids and isoprostanes in atherosclerosis. Curr Opin Lipidol 1998 ; 9 : 441 -448. Crossref Search ADS PubMed 3 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 4 Devaraj S, Jialal I. Oxidized low density lipoprotein and atherosclerosis. Int J Clin Lab Res 1996 ; 26 : 178 -184. Crossref Search ADS PubMed 5 Berliner JA, Heinecke JW. The role of oxidized lipoproteins in atherogenesis. Free Radic Biol Med 1992 ; 92 : 127 -143. 6 Bierman E. Atherogenesis in diabetes. Arterioscler Thromb 1992 ; 12 : 647 -656. Crossref Search ADS PubMed 7 Banga JD, Sixma JJ. Diabetes mellitus, vascular disease and thrombosis. Clin Hematol 1986 ; 15 : 465 -492. 8 Babiy A, Gebicki J, Sullivan DR, Willey K. Increased oxidizability of plasma lipoproteins in diabetic patients can be decreased by probucol therapy and is not due to glycation. Biochem Pharmacol 1992 ; 43 : 995 -1000. Crossref Search ADS PubMed 9 Dandona P, Thusu K, Cook S, Snyder B, Makowski J, Armstrong D, Nicotera T. Oxidative damage to DNA in diabetes mellitus. Lancet 1996 ; 347 : 444 -445. Crossref Search ADS PubMed 10 Bellomo G, Maggi E, Poli M, Agosta FG, Bollati P, Finardi G. Autoantibodies against Ox-LDL in NIDDM. Diabetes 1995 ; 44 : 60 -66. Crossref Search ADS PubMed 11 Yoshida H, Ishikawa T, Nakamura H. Vitamin E/Lipid peroxide ratio and susceptibility of LDL to oxidative modification in NIDDM. Arterioscler Thromb Vasc Biol 1997 ; 17 : 1438 -1446. Crossref Search ADS PubMed 12 Beaudeux JL, Guillausseau PJ, Peynet J, Flourie F, Assayag M, Tielmans D, et al. Enhanced susceptibility of LDL to in vitro oxidation in type 1 and 2 diabetic patients. Clin Chim Acta 1995 ; 239 : 131 -141. Crossref Search ADS PubMed 13 Cominacini L, Garbin U, Pastorino AM, Fratta Pasini A, Campagnola M, De Santis A, et al. Increased susceptibility of LDL to in vitro oxidation in patients with IDDM and NIDDM. Diabetes Res 1994 ; 26 : 173 -184. PubMed 14 Mironova MA, Klein RL, Virella GT, Lopes-Virella MF. Anti-modified LDL antibodies, LDL-containing immune complexes, and susceptibility of LDL to in vitro oxidation in patients with type 2 diabetes. Diabetes 2000 ; 49 : 1033 -1041. Crossref Search ADS PubMed 15 Pratico D. F2-Isoprostanes: sensitive and specific non-invasive indices of lipid peroxidation in vivo. Atherosclerosis 1999 ; 147 : 1 -10. Crossref Search ADS PubMed 16 Roberts LJ, II, Morrow J. The generation and actions of isoprostanes. Biochim Biophys Acta 1997 ; 1345 : 121 -135. Crossref Search ADS PubMed 17 Patrono C, Fitzgerald G. Isoprostanes—potential markers of oxidant stress in atherothrombotic disease. Arterioscler Thromb Vasc Biol 1997 ; 17 : 2309 -2315. Crossref Search ADS PubMed 18 Davi G, Ciabattoni G, Consoli A, Mezzetti A, Falco A, Santarone G, et al. In-vivo formation of 8-ep PGF2-α and platelet activation in diabetes mellitus-effect of improved metabolic control and vitamin E supplementation. Circulation 1999 ; 99 : 224 -229. Crossref Search ADS PubMed 19 Gopaul NK, Anggard EE, Mallet AI, Betteridge DJ, Wolff SP, Nourooz-zadeh J. Plasma 8-epi PGF2 α levels are increased in individuals with NIDDM. FEBS Lett 1995 ; 368 : 225 -229. Crossref Search ADS PubMed 20 Devaraj S, Jialal I. LDL. Post-secretory modification, monocyte function and circulating adhesion molecules in type 2 diabetic patients with and without macrovascular complications: the effect of α tocopherol supplementation. Circulation 2000 ; 102 : 191 -196. Crossref Search ADS PubMed 21 Jialal I, Fuller CJ, Huet BA. The effect of a-tocopherol supplementation on LDL oxidation: a dose-response study. Arterioscler Thromb Vasc Biol 1995 ; 15 : 90 -97. 22 Devaraj S, Jialal I. Assessment of lipoprotein peroxidation. Rifai N Warnick GR Dominiczak MH eds. Handbook of lipoprotein testing 2001 : 465 -479 AACC Press Washington. . 23 Marangon K, Devaraj S, Jialal I. Measurement of protein carbonyls by a sensitive ELISA. Clin Chem 1999 ; 45 : 577 -578. Crossref Search ADS PubMed 24 Marangon K, Devaraj S, Tirosh O, Packer L, Jialal I. Comparison of the effect of α-lipoic acid and α-tocopherol supplementation on measures of oxidative stress. Free Radic Biol Med 1999 ; 27 : 1114 -1121. Crossref Search ADS PubMed 25 Morrow JD, Jackson Roberts L. Mass spectrometric quantitation of F2-isoprostanes in biological fluids and tissues as a measure of oxidant stress. Methods Enzymol 1998 ; 300 : 3 -12. 26 Meagher EA, Barry OP, Lawson JA, Rokach J, FitzGerald GA. Effects of vitamin e on lipid peroxidation in healthy persons. JAMA 2001 ; 285 : 1178 -1182. Crossref Search ADS PubMed 27 Proudfoot J, Barden A, Mori TA, Burke V, Croft KD, Beilin LJ, Puddey IB. Measurement of urinary F2-isoprostanes as markers of in vivo lipid peroxidation–a comparison of EIA with GC/MS. Anal Biochem 1999 ; 272 : 209 -215. Crossref Search ADS PubMed 28 Wang Z, Ciabattoni G, Creminon C, Lawson J, Fitzgerald GA, Patrono C, Maclouf J. Immunological characterization of urinary 8-epi PGF2 α excretion in man. J Pharmacol Exp Ther 1995 ; 275 : 94 -100. PubMed 29 Jialal I. Evolving lipoprotein risk factors: lipoprotein 9a and oxidized LDL. Clin Chem 1998 ; 44 : 1827 -1832. Crossref Search ADS PubMed 30 Lawson JA, Rokach J, Fitzgerald GA. Isoprostanes: formation, analysis and use as indices of lipid peroxidation in vivo. J Biol Chem 1999 ; 274 : 24441 -24444. Crossref Search ADS PubMed 31 Oranje WA, Wolffenbuttel HR. Lipid peroxidation and atherosclerosis in type 2 diabetes. J Lab Clin Med 1999 ; 134 : 19 -32. Crossref Search ADS PubMed 32 Reaven PD, Herold DA, Barnett J, Edelman S. Effects of vitamin E on susceptibility of LDL and LDL subfractions to oxidation and on protein glycation in NIDDM. Diabetes Care 1995 ; 18 : 807 -816. Crossref Search ADS PubMed 33 Fuller CJ, Chandalia M, Garg A, Grundy SM, Jialal I. RRR-AT supplementation at pharmacological doses decreases LDL oxidative susceptibility but not protein glycation in patients with diabetes mellitus. Am J Clin Nutr 1996 ; 63 : 753 -759. Crossref Search ADS PubMed © 2001 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)

Ceder,, Gunnel;Jones, Alan, Wayne

doi: 10.1093/clinchem/47.11.1980pmid: N/A

Abstract Background: Both the illicit drug heroin and the prescription drug codeine are metabolized to morphine, which tends to complicate interpretation of opiate-positive samples. We report here the concentrations of morphine and codeine, the morphine/codeine ratios, and 6-acetylmorphine (6-AM) in blood specimens from individuals arrested for driving under the influence of drugs (DUID) in Sweden. The results were compared with positive findings of 6-AM in urine as evidence of heroin intake. Methods: In 339 DUID suspects, both blood and urine specimens were available for toxicologic analysis. In another 882 cases, only blood was available. All specimens were initially analyzed by immunoassay, and the positive results were verified by isotope-dilution gas chromatography–mass spectrometry. In routine casework, the limits of quantification (LOQs) for unconjugated opiates were 5 ng/g for blood and 20 μg/L for urine. Results: The median concentration of morphine in blood was 30 ng/g with 2.5 and 97.5 percentiles of 5 and 230 ng/g, respectively (n = 979). This compares with a median codeine concentration of 20 ng/g and 2.5 and 97.5 percentiles of 5 and 592 ng/g, respectively (n = 784). The specific metabolite of heroin, 6-AM, was identified in only 16 of 675 blood specimens (2.3%). This compares with positive findings of 6-AM in 212 of 339 urine samples (62%) from the same population of DUID suspects. When 6-AM was identified in urine, the morphine/codeine ratio in blood was always greater than unity (median, 6.0; range, 1–66). In 18 instances, 6-AM was present in urine, although morphine and codeine were below the LOQ in blood. The morphine/codeine ratio in blood was greater than unity in 85% of DUID cases when urine was not available (n = 506), and the median morphine and codeine concentrations were 70 ng/g and 10 ng/g, respectively. When morphine/codeine ratios in blood were less than unity (n = 76), the median morphine and codeine concentrations were 10 ng/g and 180 ng/g, respectively. Conclusions: Only 2.3% of opiate-positive DUID suspects were verified as heroin users on the basis of positive findings of 6-AM in blood. A much higher proportion (62%) were verified heroin users from 6-AM identified in urine. When urine was not available for analysis, finding a morphine/codeine concentration ratio in blood above unity suggests heroin use and not medication with codeine. This biomarker indicated that 85% of opiate-positive DUID blood samples were from heroin users. Determining whether a person has taken heroin as opposed to a prescription drug containing codeine is not always easy because both opiates undergo metabolism into morphine (1)(2)(3)(4). Moreover, clandestine heroin preparations contain acetylcodeine as an impurity, and this opiate is readily deacetylated to produce codeine (5)(6). Accordingly, forensic toxicology reports showing morphine and codeine in blood or urine could have several possible explanations. Heroin has a specific metabolite, 6-acetylmorphine (6-AM), 1 but its very short half-life of 10–20 min means that the window of detection in blood is only 1–2 h after administration (3)(4). For this reason, 6-AM is seldom identified in forensic blood samples above the limits of quantification (LOQs) by current methods of gas chromatography–mass spectrometry (GC-MS) (7)(8). However, evidence of heroin intake is easier to obtain by analysis of urine specimens because 6-AM remains in urine for much longer than in blood (8)(9)(10)(11). The traffic police in Sweden try to obtain specimens of both blood and urine for toxicologic analysis, especially from individuals suspected of driving under the influence of drugs (DUID). This gave us the opportunity to compare the concentrations of morphine, codeine, and 6-AM in blood with the frequency of finding 6-AM in urine as unequivocal evidence of heroin intake. Materials and Methods All blood and urine specimens from impaired drivers apprehended in Sweden are submitted for analysis to one central laboratory, the National Laboratory of Forensic Chemistry located at the University Hospital in Linköping. Drivers are apprehended for committing a moving traffic violation, during routine sobriety controls, or after traffic crashes. The material for this study was taken from the routine case files and toxicologic reports of blood or urine or both. Table 1 summarizes the numbers of DUID suspects involved in the study and whether blood or urine was available for analysis. Table 1. Distribution of DUID cases in Sweden before and after introduction of zero-tolerance legislation (July 1999). Specimens available for analysis . January 1992 to June 1999 . July 1999 to December 2000 . Blood only 4569 (77%) 2922 (56%) Both blood and urine 1348 (23%) 2248 (44%) Total 5917 (100%) 5176 (100%) Specimens available for analysis . January 1992 to June 1999 . July 1999 to December 2000 . Blood only 4569 (77%) 2922 (56%) Both blood and urine 1348 (23%) 2248 (44%) Total 5917 (100%) 5176 (100%) Open in new tab Table 1. Distribution of DUID cases in Sweden before and after introduction of zero-tolerance legislation (July 1999). Specimens available for analysis . January 1992 to June 1999 . July 1999 to December 2000 . Blood only 4569 (77%) 2922 (56%) Both blood and urine 1348 (23%) 2248 (44%) Total 5917 (100%) 5176 (100%) Specimens available for analysis . January 1992 to June 1999 . July 1999 to December 2000 . Blood only 4569 (77%) 2922 (56%) Both blood and urine 1348 (23%) 2248 (44%) Total 5917 (100%) 5176 (100%) Open in new tab The police typically conduct a breath-alcohol screening test at the roadside with the aid of a hand-held instrument (12), either an Alcolmeter S-D2 or Alcolmeter 400 (Lion Laboratories). Performing various field-sobriety tests is unnecessary before making the preliminary breath-alcohol analysis. If the breath-alcohol test is negative and the police suspect impairment by drugs other than alcohol, specimens of blood and urine are taken for toxicologic analysis. The police make this decision on the basis of the general appearance and behavior of the suspect. More recently, the police also examine pupil size and eye movements, as well as other signs and symptoms of drug impairment. Examination of DUID suspects by drug recognition experts is not an option in Sweden. However, in some situations, a physician makes a clinical examination in conjunction with obtaining blood and urine samples for forensic toxicology. Venous whole blood was submitted in two 10-mL gray-stoppered Vacutainer tubes (Becton Dickinson) containing 100 mg of sodium fluoride and 25 mg of sodium oxalate as preservatives. Urine specimens were sent for analysis in two 10-mL, plastic screw-capped tubes, each containing 100 mg of sodium fluoride as preservative. A qualitative screening analysis of urine and/or blood was made for five classes of abused drugs (opiates, cannabinoids, amphetamine analogs, cocaine metabolites, and benzodiazepines) by immunoassay techniques such as Emit or CEDIA with the Hitachi 717. The most commonly prescribed sedative-hypnotics and analgesics were analyzed in blood by capillary column gas chromatography with nitrogen-phosphorous detection (13). All positive immunoassay results were verified by more specific methods (liquid chromatography–mass spectrometry or GC-MS). After solid-phase extraction, opiates were determined by GC-MS with the use of deuterium-labeled internal standards and selected-ion monitoring (14). The blood aliquots were measured by weight and concentrations of drugs reported in mass/mass units (ng/g) because the fluidity of specimens differs widely, especially in postmortem bloods where it is not always practical to dispense aliquots by volume. After extraction and before gas chromatography, the opiates were converted to their pentafluoropropionic acid anhydride derivatives. For analysis of urine, the aliquots were measured by volume and results were thus reported mass/volume (μg/L). Before GC-MS, the trimethylsilylester derivatives were prepared by treatment with N,O-bis(trimethylsilyl)trifluoroacetamide. The concentrations of opiates in blood and urine (morphine, codeine, and 6-AM) were quantitatively determined in a single GC-MS analytical run. The specimens were not hydrolyzed; therefore, the concentrations of unconjugated opiates could be reported. The LOQs for codeine, morphine, and 6-AM in blood were 5 ng/g and 20 μg/L for urine. Calibration curves were constructed up to a maximum concentration of 1000 μg/L in urine and values above this concentration were reported as >1000 μg/L. The frequency distributions of the concentrations of morphine and codeine in blood were markedly skewed, so the median value and 2.5 and 97.5 percentiles were used to characterize the data. Results Table 2 presents the drugs verified positive in blood from individuals apprehended for DUID in Sweden during 2000. Many individuals had several different drugs present in their blood, including scheduled narcotic drugs, prescription medications (mainly benzodiazepines), as well as various alcohol and drug combinations. Amphetamine and cannabis are the most commonly encountered drugs in DUID suspects in Sweden, but morphine and codeine are also high on the list of substances confirmed present in blood specimens. Table 2. Analytical results for the most commonly encountered drugs in blood samples from DUID suspects in Sweden during 2000.1 Drug . Year 2000 (n = 3808) . Amphetamine 2395 (63%) Tetrahydrocannabinol 1119 (29%) Diazepam 743 (20%) Morphine 382 (10%) Methamphetamine 364 (10%) Flunitrazepam 347 (9%) Codeine 245 (6%) Paracetamol (acetaminophen) 129 (3%) MDMA 116 (3%) Phenazone 109 (3%) Benzoylecgonine 89 (2%) Nitrazepam 81 (2%) Alprazolam 80 (2%) d-Proxyphene 67 (2%) Drug . Year 2000 (n = 3808) . Amphetamine 2395 (63%) Tetrahydrocannabinol 1119 (29%) Diazepam 743 (20%) Morphine 382 (10%) Methamphetamine 364 (10%) Flunitrazepam 347 (9%) Codeine 245 (6%) Paracetamol (acetaminophen) 129 (3%) MDMA 116 (3%) Phenazone 109 (3%) Benzoylecgonine 89 (2%) Nitrazepam 81 (2%) Alprazolam 80 (2%) d-Proxyphene 67 (2%) 1 Note that many specimens contained several different drug classes. Open in new tab Table 2. Analytical results for the most commonly encountered drugs in blood samples from DUID suspects in Sweden during 2000.1 Drug . Year 2000 (n = 3808) . Amphetamine 2395 (63%) Tetrahydrocannabinol 1119 (29%) Diazepam 743 (20%) Morphine 382 (10%) Methamphetamine 364 (10%) Flunitrazepam 347 (9%) Codeine 245 (6%) Paracetamol (acetaminophen) 129 (3%) MDMA 116 (3%) Phenazone 109 (3%) Benzoylecgonine 89 (2%) Nitrazepam 81 (2%) Alprazolam 80 (2%) d-Proxyphene 67 (2%) Drug . Year 2000 (n = 3808) . Amphetamine 2395 (63%) Tetrahydrocannabinol 1119 (29%) Diazepam 743 (20%) Morphine 382 (10%) Methamphetamine 364 (10%) Flunitrazepam 347 (9%) Codeine 245 (6%) Paracetamol (acetaminophen) 129 (3%) MDMA 116 (3%) Phenazone 109 (3%) Benzoylecgonine 89 (2%) Nitrazepam 81 (2%) Alprazolam 80 (2%) d-Proxyphene 67 (2%) 1 Note that many specimens contained several different drug classes. Open in new tab Table 3 gives the concentrations of morphine, codeine, and 6-AM in blood samples from DUID suspects apprehended between 1992 and 2000. As expected, the concentrations span a wide range in this population, which comprises people taking therapeutic doses of opiates, as well as many drug addicts. The heroin metabolite 6-AM was determined 43 times at a median concentration of 8 ng/g (2.5 and 97.5 percentiles, 5 and 104 ng/g). The highly skewed distributions of morphine and codeine concentrations in blood are shown in Figs. 1 and 2 with medians of 30 ng/g for morphine and 20 ng/g for codeine. Figure 1. Open in new tabDownload slide Frequency distribution of the concentrations of unconjugated morphine in whole blood form DUID suspects in Sweden (n = 979). Figure 1. Open in new tabDownload slide Frequency distribution of the concentrations of unconjugated morphine in whole blood form DUID suspects in Sweden (n = 979). Figure 2. Open in new tabDownload slide Frequency distribution of the concentrations of unconjugated codeine in whole blood from DUID suspects apprehended in Sweden (n = 784). Figure 2. Open in new tabDownload slide Frequency distribution of the concentrations of unconjugated codeine in whole blood from DUID suspects apprehended in Sweden (n = 784). Table 3. Concentrations of morphine, codeine, and 6-AM in opiate-positive blood specimens from motorists apprehended for DUID in Sweden between 1992 and 2000. Opiate . Number . Concentration of opiates in blood, median (2.5 and 97.5 percentiles) . Morphine 979 30 ng/g (5 and 230) Codeine 784 20 ng/g (5 and 592) 6-AM 43 8 ng/g (5 and 104) Opiate . Number . Concentration of opiates in blood, median (2.5 and 97.5 percentiles) . Morphine 979 30 ng/g (5 and 230) Codeine 784 20 ng/g (5 and 592) 6-AM 43 8 ng/g (5 and 104) Open in new tab Table 3. Concentrations of morphine, codeine, and 6-AM in opiate-positive blood specimens from motorists apprehended for DUID in Sweden between 1992 and 2000. Opiate . Number . Concentration of opiates in blood, median (2.5 and 97.5 percentiles) . Morphine 979 30 ng/g (5 and 230) Codeine 784 20 ng/g (5 and 592) 6-AM 43 8 ng/g (5 and 104) Opiate . Number . Concentration of opiates in blood, median (2.5 and 97.5 percentiles) . Morphine 979 30 ng/g (5 and 230) Codeine 784 20 ng/g (5 and 592) 6-AM 43 8 ng/g (5 and 104) Open in new tab In 675 opiate-positive blood samples, 6-AM was verified above the LOQ only 16 times (2.3%). By contrast, 6-AM was verified in 212 of 339 urine samples analyzed (62%). In 89 urine specimens containing 6-AM, the median morphine/codeine ratio in the corresponding blood sample was 6.0 (range, 1–66). In 18 instances, 6-AM was present in urine, although morphine and codeine were below the LOQ in blood. Another 105 urine samples contained 6-AM when morphine was above the LOQ in blood (median concentration, 10 ng/g) but the concentration of codeine in blood was below the LOQ. The frequency distribution of morphine/codeine ratios in blood when urine was not available for analysis, and when morphine and codeine were both above the LOQ (n = 506), is shown in Fig. 3 . In 76 instances (15%), the morphine/codeine ratio was less than unity (median, 0.057), and the corresponding median blood-morphine concentration was 10 ng/g compared with 180 ng/g for codeine, which gives a median codeine/morphine ratio of 18:1. This high codeine/morphine ratio most likely reflects intake of the prescription drug codeine. In 430 cases, the morphine/codeine ratio was greater than unity (median, 5.0; range, 1.1–67). For these cases, the median concentration of morphine was 70 ng/g (range, 7–620 ng/g) compared with 10 ng/g (range, 1–100 ng/g) for codeine, indicating heroin use. These results are summarized in Table 4 . Figure 3. Open in new tabDownload slide Frequency distribution of the concentration ratios of unconjugated morphine to codeine in specimens of whole blood from DUID suspects apprehended in Sweden (n = 506). Figure 3. Open in new tabDownload slide Frequency distribution of the concentration ratios of unconjugated morphine to codeine in specimens of whole blood from DUID suspects apprehended in Sweden (n = 506). Table 4. Concentrations of morphine and codeine in blood samples from motorists apprehended for DUID in Sweden when both opiates were above the LOQs (5 ng/g) by GC-MS. Opiate in blood . Number . Median (2.5 and 97.5 percentiles) . Morphine 506 50 ng/g (6 and 268) Codeine 506 10 ng/g (5 and 451) Morphine/Codeine ratio >1 430 Morphine, 70 ng/g; codeine, 10 ng/g Morphine/Codeine ratio <1 76 Morphine, 10 ng/g; codeine, 180 ng/g Opiate in blood . Number . Median (2.5 and 97.5 percentiles) . Morphine 506 50 ng/g (6 and 268) Codeine 506 10 ng/g (5 and 451) Morphine/Codeine ratio >1 430 Morphine, 70 ng/g; codeine, 10 ng/g Morphine/Codeine ratio <1 76 Morphine, 10 ng/g; codeine, 180 ng/g Open in new tab Table 4. Concentrations of morphine and codeine in blood samples from motorists apprehended for DUID in Sweden when both opiates were above the LOQs (5 ng/g) by GC-MS. Opiate in blood . Number . Median (2.5 and 97.5 percentiles) . Morphine 506 50 ng/g (6 and 268) Codeine 506 10 ng/g (5 and 451) Morphine/Codeine ratio >1 430 Morphine, 70 ng/g; codeine, 10 ng/g Morphine/Codeine ratio <1 76 Morphine, 10 ng/g; codeine, 180 ng/g Opiate in blood . Number . Median (2.5 and 97.5 percentiles) . Morphine 506 50 ng/g (6 and 268) Codeine 506 10 ng/g (5 and 451) Morphine/Codeine ratio >1 430 Morphine, 70 ng/g; codeine, 10 ng/g Morphine/Codeine ratio <1 76 Morphine, 10 ng/g; codeine, 180 ng/g Open in new tab Discussion Since July 1999, Sweden has enforced a zero-tolerance law for DUID offenses, which means that the presence of a scheduled narcotic drug in blood is sufficient for prosecution regardless of whether the person exhibits any signs and symptoms of impairment. For the use of prescription drugs (e.g., benzodiazepines) without clear-cut evidence of impairment, the prosecution is required to prove that the recommended dose was exceeded. This requires an expert opinion to relate the blood drug concentration to the expected therapeutic concentration for the particular medication, which is not always easy. If scheduled drugs like amphetamine, methamphetamine, cocaine, and tetrahydrocannabinol are present in blood above the LOQs by GC-MS, this is sufficient for prosecution, regardless of whether the person’s driving ability was impaired. Zero-tolerance laws for narcotic drugs are similar in principle to the widely used per se laws for driving under the influence of alcohol. Because all specimens are sent to one forensic toxicology laboratory in Sweden (population 8.8 million), the LOQs for scheduled drugs in blood determined by GC-MS are effectively the threshold concentration limits for prosecution. Since the introduction of this zero-tolerance law for narcotics, the number of DUID cases submitted by the police has increased approximately fivefold. This sharp increase after July 1999 stems from an increased activity and enthusiasm by the police and prosecutors, who are now more likely to win a conviction. Moreover, the police are trained to recognize signs and symptoms of drug abuse and are allowed to monitor pupil size and look for nystagmus. This dramatic increase in the number of DUID cases can also be explained, at least in part, by prior knowledge about criminal elements and drug abusers in the community. The ultimate evidence of heroin use is finding 6-AM in blood or urine. However, only 2.3% of opiate-positive DUID suspects were verified as heroin users based on the analysis of the 6-AM metabolite in blood. This compares with 62% confirmed users of heroin after the analysis of 6-AM in urine. The window of detection for 6-AM in blood and urine depends on the dose of heroin taken, the route of administration, and the frequency of heroin intake (10). The concentrations of opiates excreted in urine, including 6-AM, vastly exceeds the concentration in blood (15)(16). Interpreting the concentrations of morphine and codeine measured in urine is more complicated, owing to the irregular frequency of urination and pooling of urine in the bladder, as well as variable diuresis and fluctuations in pH, which will affect the results of toxicologic analysis (16)(17)(18)(19). Moreover, cleavage of glucuronides is often a tricky problem depending on whether acid hydrolysis or enzymatic methods are used and the source of β-glucuronidase (20). The unconjugated drug concentrations of opiates in blood, more so than urine, reflect fairly recent intake (17). Analysis of drugs of abuse in urine provides evidence of intake but does not permit conclusions to be drawn about coexisting concentrations in blood or impairment of the individual at the time of voiding. Even the concentrations of unconjugated opiates in blood or serum are difficult to relate to impairment, because of the development of tolerance. The introduction of so-called zero-tolerance laws for scheduled drugs in traffic cases sidesteps these problems and offers an effective means of combating the problem of DUID. The forensic blood samples that are sent for analysis to our laboratory contain 1% sodium fluoride as preservative, which tends to cause hemolysis. Accordingly, the concentrations of drugs and poisons are determined in whole blood and not plasma or serum, which are the specimens more commonly used for therapeutic drug monitoring. Because the distribution ratio of morphine between plasma and whole blood is close to unity (21), finding a morphine/codeine ratio >1.0 in plasma or serum also furnishes evidence of heroin use. Studies on the pharmacokinetics of codeine have shown that morphine/codeine ratios in plasma remain less than unity at all times after administration (1)(2)(18). When 100 mg of codeine phosphate was given to healthy volunteers, the peak plasma morphine concentration was only 3.2% of the peak codeine concentration (19). Furthermore, the morphine/codeine ratios in plasma remained less than unity for up to 23 h postdosing (19). Similar results have been reported after a single oral dose of 60 mg of codeine and also after repeated intake (1)(18). Morphine is also a metabolite of ethyl morphine, the active ingredient in various antitussive medications (22)(23). However only a small fraction of the ethyl morphine is seemingly converted to morphine, and both parent drug and metabolite are easily detected in urine for up to 24 h after intake (22). Since the first report in Sweden in 1982 (24), the presence of various opiates in poppy seeds and other food products is well recognized as a confounding factor in urine drug testing programs. Finding 6-AM in urine can be used to dismiss the allegation that the source of morphine and codeine in urine was from foods laced with poppy seeds (25). The concentrations of unconjugated morphine and codeine in blood samples after eating poppy seed cakes will hardly be expected to exceed the LOQ of the GC-MS method of analysis (26). Because of polymorphisms of the CYP2D6 enzyme, large interethnic differences in the rate and extent of demethylation of codeine exist (27). This leads to slow and rapid metabolizers of codeine depending on the particular genotype inherited (27)(28). Most Caucasians seem to be rapid metabolizers of codeine, and morphine/codeine ratios within this population are always much less than unity (27). In a group of Chinese volunteers with low capability of metabolizing codeine, the morphine/codeine concentration ratios after a dose of 50 mg of codeine were even less than in Caucasians. Accordingly, finding a high morphine/codeine ratio in forensic blood specimens cannot be attributed to a person’s inherent ability to convert codeine into morphine. Many DUID suspects with morphine and codeine identified in blood claim they took the prescription drug codeine or received morphine for relief of pain (e.g., from injuries resulting from a traffic crash). However, the results reported here, based on finding 6-AM in urine, suggests that when morphine/codeine ratios in blood are greater than unity, the person has more likely used heroin than taken the legal drug codeine. This allowed us to conclude that in 506 DUID suspects with morphine and codeine verified in blood above the LOQ, 85% had used heroin. Intake of mixtures of codeine and morphine or codeine and heroin obviously complicates the interpretation of morphine/codeine ratios in blood. The rapid increase of immigration into Sweden has seen new drug cultures emerging. An example is the abuse of raw opium, which is likely to give a different pattern of morphine and codeine concentrations in body fluids. However, experience with two documented cases of raw opium abuse gave morphine/codeine ratios in blood of 1.5 and 2.0. Accordingly, such cases would not be confused with the intake of codeine, for which morphine/codeine ratios are always much less than unity (19). From the results of this study, we recommend that if urine is not available for toxicologic analysis, the finding of a morphine/codeine concentration ratio in blood exceeding unity is strong evidence that the person had used heroin, as opposed to having taken a prescription drug containing codeine. 1 " Nonstandard abbreviations: 6-AM, 6-acetylmorphine; LOQ, limit of quantification; GC-MS, gas chromatography–mass spectrometry; and DUID, driving under the influence of drugs. References 1 Findlay JWA, Jones EC, Butz RF, Welch RM. Plasma codeine and morphine concentrations after therapeutic oral doses of codeine-containing analgesics. Clin Pharmacol Ther 1978 ; 24 : 60 -68. Crossref Search ADS PubMed 2 Quiding H, Anderson P, Bondesson U, Boréus L-O, Hynning P-Å. Plasma concentrations of codeine and its metabolite morphine after single and repeated oral administration. Eur J Clin Pharmacol 1986 ; 30 : 673 -677. Crossref Search ADS PubMed 3 Goldberger BA, Darwin WD, Grant TM, Allen AC, Caplan YH, Cone EJ. Measurement of heroin and its metabolites by isotope dilution electron-impact mass spectrometry. Clin Chem 1993 ; 39 : 670 -675. Crossref Search ADS PubMed 4 Inturrisi CE, Max MB, Foley KM, Schultz M, Shin S-U, Houde RW. The pharmacokinetics of heroin in patients with chronic pain. N Engl J Med 1984 ; 310 : 1213 -1217. Crossref Search ADS PubMed 5 O’ Neal CL, Poklis A. The detection of acetylcodeine and 6-acetylmorphine in opiate positive urines. Forensic Sci Int 1998 ; 95 : 1 -10. Crossref Search ADS PubMed 6 Staub C, Marset M, Mino A, Mangin P. Detection of acetylcodeine in urine as an indicator of illicit heroin use: method validation and results of a pilot study. Clin Chem 2001 ; 47 : 301 -307. Crossref Search ADS PubMed 7 Christophersen AS, Ceder G, Kristinsson J, Lillsunde P, Stenntoft A. Drugged driving in the Nordic countries–a comparative study between five countries. Forensic Sci Int 1999 ; 106 : 173 -190. Crossref Search ADS PubMed 8 Jones AW. Heroin use by motorists in Sweden confirmed by analysis of 6-acetylmorphine in urine [Letter]. J Anal Toxicol 2001 ; 25 : 353 -355. Crossref Search ADS PubMed 9 Fehn J, Megges G. Detection of O6-Monoacetylmorphine in urine samples by GC/MS as evidence for heroin use. J Anal Toxicol 1985 ; 9 : 134 -138. Crossref Search ADS PubMed 10 Cone EJ, Welch P, Mitchell JM, Paul BD. Forensic drug testing for opiates: I. Detection of 6-acetylmorphine in urine as an indicator of recent heroin exposure; drug and assay considerations and detection times. J Anal Toxicol 1991 ; 15 : 1 -7. Crossref Search ADS PubMed 11 Paul BD, Shimomura ET, Smith ML. A practical approach to determine cutoff concentrations for opiate testing with simultaneous detection of codeine, morphine, and 6-acetylmorphine in urine. Clin Chem 1999 ; 45 : 510 -519. Crossref Search ADS PubMed 12 Jones AW. Evaluation of breath-alcohol instruments. IV. Roadside tests with Alcolmeter pocket model. Forensic Sci Int 1985 ; 28 : 157 -165. Crossref Search ADS PubMed 13 Eklund A, Jonsson J, Schuberth J. A procedure for simultaneous screening and quantification of basic drugs in liver, utilizing capillary gas chromatography and nitrogen sensitive detection. J Anal Toxicol 1983 ; 7 : 24 -28. Crossref Search ADS PubMed 14 Schuberth J, Schuberth JS. Gas chromatographic-mass spectrometric determination of morphine, codeine and 6-monoacetyl morphine in blood extracted by solid phase. J Chromatogr 1989 ; 490 : 444 -449. Crossref Search ADS PubMed 15 Cone EJ, Welch P, Paul BD, Mitchell JM. Forensic drug testing for opiates, III. Urinary excretion rates of morphine and codeine following codeine administration. J Anal Toxicol 1991 ; 15 : 161 -166. Crossref Search ADS PubMed 16 Lafolie P, Beck O, Lin Z, Albertioni F, Boréus L. Urine and plasma pharmacokinetics of codeine in healthy volunteers: implications for drugs-of-abuse testing. J Anal Toxicol 1996 ; 20 : 541 -546. Crossref Search ADS PubMed 17 Boerner U, Abbott S, Roe RL. The metabolism of morphine and heroin in man. Drug Metab Rev 1975 ; 4 : 39 -73. Crossref Search ADS PubMed 18 Shah JC, Mason WD. Plasma codeine and morphine concentrations after a single oral dose of codeine phosphate. J Clin Pharmacol 1990 ; 30 : 764 -766. Crossref Search ADS PubMed 19 Kronstrand R, Jones AW. Concentration ratios of codeine-to-morphine in plasma after a single oral dose (100 mg) of codeine phosphate [Letter]. J Anal Toxicol 2001 ; 25 : 486 -487. Crossref Search ADS PubMed 20 Romberg RW, Lee L. Comparison of the hydrolysis rates of morphine-3-glucuronide and morphine-6-glucuronide with acid and β-glucuronidase. J Anal Toxicol 1995 ; 19 : 157 -162. Crossref Search ADS PubMed 21 Hand CW, Moore RA, Sear JW. Comparison of whole blood and plasma morphine. J Anal Toxicol 1988 ; 12 : 234 -235. Crossref Search ADS PubMed 22 Popa C, Beck O, Brodin K. Morphine formation from ethylmorphine: implications for drugs-of-abuse testing in urine. J Anal Toxicol 1998 ; 22 : 142 -147. Crossref Search ADS PubMed 23 Aasmundstad TA, Xu BQ, Johansson I, Ripel Å, Bjorneboe A, Christophersen AS, et al. Biotransformation and pharmacokinetics of ethylmorphine after a single oral dose. Br J Clin Pharmacol 1995 ; 39 : 611 -620. Crossref Search ADS PubMed 24 Bjerver K, Jonsson J, Nilsson A, Schuberth J, Schuberth JS. Morphine intake from poppy seed food. J Pharm Pharmacol 1982 ; 34 : 798 -801. Crossref Search ADS 25 Mule SJ, Casella GA. Rendering the “poppy-seed defense” defenseless: identification of 6-monoacetylmorphine in urine by gas chromatography/mass spectrometry. Clin Chem 1988 ; 34 : 1427 -1430. Crossref Search ADS PubMed 26 Hayes LW, Krasselt WG, Mueggler PA. Concentrations of morphine and codeine in serum and urine after ingestion of poppy seeds. Clin Chem 1987 ; 33 : 806 -808. Crossref Search ADS PubMed 27 Yue QY, Hasselström J, Svensson JO, Säwe J. Pharmacokinetics of codeine and its metabolites in Caucasian healthy volunteers: comparisons between extensive and poor hydroxylators of debrisoquine. Br J Clin Pharmacol 1991 ; 31 : 635 -642. Crossref Search ADS PubMed 28 Yue QY, Svensson JO, Sjöqvist F, Säwe J. A comparison of the pharmacokinetics of codeine and its metabolites in healthy Chinese and Caucasian extensive hydroxylators of debrisiquine. Br J Clin Pharmacol 1991 ; 31 : 643 -647. Crossref Search ADS PubMed © 2001 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)

Kushnir, Mark, M;Komaromy-Hiller,, Gabor;Shushan,, Bori;Urry, Francis, M;Roberts, William, L

doi: 10.1093/clinchem/47.11.1993pmid: N/A

Abstract Background: Methylmalonic acid (MMA) is a dicarboxylic acid whose concentration can be increased in blood and urine in patients with an inborn error of metabolism or vitamin B12 deficiency. We developed a method for the selective analysis of dicarboxylic acids that exploits the high specificity of tandem mass spectrometry (MS/MS) and the substantial difference in fragmentation patterns of the isomers methylmalonic (MMA) and succinic acid (SA). Methods: Dicarboxylic acids were extracted from samples with methyl-tert-butyl ether and derivatized with butanolic HCl to form dibutyl esters. The derivative was injected into the liquid chromatography (LC)-MS/MS system using TurboIonSprayTM (nebulizer-assisted electrospray) ionization and quantified by the multiple reaction monitoring mode of MS/MS. Results: The assay for MMA was linear up to 150 μmol/L. The total imprecision was ≤7.5% at both low and high concentrations. The limits of quantification and detection were 0.1 and 0.05 μmol/L, respectively. The degree of interference from SA could be predicted from the branching ratios of the major product ions. Conclusions: The method is specific for dicarboxylic acids. The LC-MS/MS analysis for MMA requires minimal chromatographic separation and takes <60 s per sample. The entire analysis, including sample preparation, for a batch of 100 specimens can be performed in <4 h. Methylmalonic acid (MMA) 2 is a metabolic intermediate in the conversion of propionic acid to succinic acid (SA) (1). Measurement of MMA has become an important diagnostic procedure in clinical chemistry because of the accumulated evidence that increased MMA is a marker of vitamin B12 deficiency (2)(3)(4)(5). Vitamin B12 is an essential cofactor for the enzymatic carbon rearrangement of MMA to SA, and the lack of vitamin B12 leads to increased MMA. A deficiency of vitamin B12 causes a serious and often irreversible neurologic disorder known as subacute combined degeneration of the spinal cord (2). Moderately increased MMA (>0.4 μmol/L in serum or plasma and >3.6 mmol/mol creatinine in urine) is an early indicator of acquired vitamin B12 deficiency. A massive increase of MMA in serum, plasma, or urine (100- to 1000-fold above the concentrations characteristic for the vitamin B12 deficiency) is indicative of methylmalonic acidemia, an inborn metabolic disorder (3). Although either serum MMA or serum vitamin B12 can be used for detecting vitamin B12 deficiency, there are advantages to measuring MMA instead of vitamin B12. This is related to the facts that (a) serum or plasma vitamin B12 concentrations may not adequately reflect tissue vitamin B12 status, (b) MMA is more stable than vitamin B12, (c) serum MMA concentrations are 1000-fold greater than serum vitamin B12 and therefore easier to measure accurately, and (d) an increase rather than a decrease in concentration is measured in vitamin B12 deficiency. Since the 1960s, efforts have been directed toward developing a rapid, simple, sensitive, and specific method for MMA analysis in biological fluids. Sample preparation for the analysis of MMA usually consists of extraction from the sample matrix and, frequently, subsequent derivatization. To be able to detect vitamin B12 deficiency, the method should be able to accurately measure low concentrations of MMA (∼1 μmol/L in urine and 0.1 μmol/L in serum). Derivatization of MMA is needed to improve MMA detection with an ultraviolet or a fluorescence detector in HPLC procedures (6)(7)(8)(9) or to convert it to a derivative amenable to gas chromatography (GC) separation and detection (10)(11)(12)(13)(14)(15)(16). The major obstacle for MMA analysis in biological fluids is the potential interference from other low-molecular weight organic acids and especially from the naturally occurring structurally related isomer SA, which is a product of MMA degradation and is usually present in samples at a concentration greater than that of MMA. SA interference is difficult to overcome because the chromatographic characteristics and mass spectra of SA are almost identical to those of MMA. MMA is usually analyzed by GC–mass spectrometry (MS) in the selected-ion monitoring mode. All derivatizing reagents used for GC-MS analysis of MMA produce nearly identical mass spectra for MMA and SA derivatives; thus, for the quantification of MMA, adequate chromatographic separation between the peaks is required. The major disadvantage associated with GC-MS methods in general is relatively low instrument throughput (usually three to six samples per hour). There is, therefore, a need for a high-throughput method of determining MMA in biological samples to effectively detect vitamin B12 deficiency and inherited disorders associated with increased MMA. Different derivatives can be used to enhance selectivity and specificity to organic acids for liquid chromatography (LC)-MS analysis. Recent work by Johnson (17) involved the determination of long- and very long-chain fatty acids by nebulizer-assisted electrospray (ES). The fatty acids were derivatized to form dimethylaminoethyl esters via a two-step condensation reaction using oxalyl chloride and dimethylaminoethanol as reagents. The reagent produces a strongly basic derivative, which enhances the response of all organic acids in the positive-ion mode without any differences in fragmentation between the structural isomers MMA and SA. One of the clinically important groups of organic acids (of which MMA and SA are representatives) is dicarboxylic acids. The dicarboxylic acids usually are analyzed by chromatographic methods along with other organic acids. The difficulty in the target analysis for these and other low-molecular weight organic acids relates to potential interference from other acids caused by similar properties that do not allow selective extraction and detection. The analysis usually requires extended instrument run time because of the need to chromatographically separate the acids from each other and to condition the chromatographic column before the following injection to elute other coextracted compounds. To date, there is no method described in the literature that is selective only for dicarboxylic acids. Recently, Magera et al. (18) published a method for the determination of MMA as the n-butyl ester derivative in plasma and urine by LC–tandem MS (MS/MS). The method is based on chromatographic separation of MMA from SA and selective fragmentation of MMA butyl ester. The analysis time per sample was 3 min, which is significantly faster than conventional GC-MS methods. Among the disadvantages of the method are time-consuming sample preparation involving solid-phase extraction, which includes a two-step separation of the extract from a residue produced during the extraction, and the relatively narrow linear range of the method. The intent of this work was to develop a rapid method for the selective analysis of dicarboxylic acids (and MMA in particular) by LC-MS/MS, based on the power of the MS/MS analyzer to differentiate between MMA and SA, that does not require chromatographic separation of these isomeric analytes. Materials and Methods reagents MMA was purchased from Sigma Chemical Co., and d3-MMA was purchased from Cambridge Isotope Laboratories. Methanol, acetonitrile, methyl-tert-butyl ether (MTBE), and phosphoric acid were all HPLC grade from Fisher Scientific. Hydrochloric acid (3 mol/L) in n-butanol was purchased from Regis Technologies, Inc. All other chemicals were of the highest purity available commercially. apparatus A PE Series 200 HPLC system (Perkin-Elmer Analytical Instruments) was equipped with a Luna C18 column (30 mm × 3.0 mm; 3-μm particles; Phenomenex). The mobile phase consisted of 850 mL/L methanol and 150 mL/L ammonium formate buffer (0.005 mol/L), pH 6.5. The mobile phase flow rate was 750 μL/min, and the LC column effluent split flow was 500–600 μL/min. The column temperature was 40 °C, the injection volume was 3 μL, and the injection interval was 60 s. An API 2000 (Applied Biosystems/MDS SCIEX, Foster City, CA) tandem mass spectrometer was used in the positive-ion mode with TurboIonSprayTM (TIS) interface. Quantitative analysis was performed in the multiple reaction monitoring (MRM) mode. The collision gas was nitrogen with a cell pressure of 1.1 Pa. The TIS capillary voltage was 6.0 kV, the orifice voltage was 21V, and the collision energy was 15 V. The MRM transitions monitored were m/z 231→119 and 231→175 for MMA, and 234→122 and 234→178 for d3-MMA. The product ions m/z 119 and 122 were quantitative, whereas the product ions m/z 175 and 178 were qualitative. Quantitative data analysis was performed with TurboQuanTM (Applied Biosystems/MDS SCIEX) software. assay procedures Preparation of calibrators and controls. Dialyzed plasma was prepared from a pool of human plasma and showed no detectable MMA and SA. Samples for method precision, sensitivity, and linearity studies were prepared in dialyzed human plasma and stored at 2–8 °C. Dialyzed plasma used for the preparation of calibrators and quality-control samples was supplemented with SA to a concentration of 6 μmol/L. Calibrators were at concentrations of 0.2, 0.4, 0.75, 1.0, 1.5, and 2.0 μmol/L. The calibrator containing 0.4 μmol/L MMA was used to establish the qualitative ion mass ratio of the intensities of the product ion fragments m/z 231→175/231→119 of MMA and m/z 234→178/234→122 of d3-MMA. The qualitative ion mass ratio acceptability limits for the controls and test samples were established as ± 40% of the values observed in the calibrator. Organic acid ester derivatives. The n-butyl esters of organic acids for qualitative mass spectral measurements were prepared by transferring 400 nmol of each acid into glass tubes. The solvent was evaporated and the residue reconstituted with 40 μL of n-butanol containing 3 mol/L HCl. The tubes were incubated at 60 °C for 15 min. Excess derivatizing reagent was evaporated, and the remaining residue was reconstituted with 4 mL of methanol containing 50 mL/L ammonium formate (0.005 mol/L), pH 6.5. For both the MS and MS/MS experiments, the ion source and analyzer conditions were the same as those used for the MMA n-butyl ester derivative. The samples were infused by syringe at a flow rate of 5 μL/min into the TIS ion source. Relative ionization efficiencies and fragmentation patterns were evaluated for the methyl, propyl, isopropyl, and amyl esters of MMA and SA. The various esters of MMA and SA were prepared by transferring 400 nmol of each acid into glass tubes. The solvent was evaporated, and the residues were reconstituted with 100 μL of the corresponding alcohols and 50 μL of concentrated sulfuric acid. The tubes were incubated at 60 °C for 15 min. The esters were extracted from the mixture with hexane, the tubes were centrifuged, and the organic layers were transferred into new tubes. The solvent was evaporated, and the residues were reconstituted with 4 mL of methanol containing 50 mL/L ammonium formate (0.005 mol/L). sample preparation Samples were aliquoted into disposable glass tubes (1 mL for serum/plasma analysis, and 0.1 mL of sample and 0.9 mL of water for urine analysis). To this, 100 μL of the working internal standard solution and 3 mL of MTBE containing 30 mL/L phosphoric acid were added. The tubes were then vortex-mixed for 5 min and centrifuged at 3000g for 10 min. The supernatant was transferred to a second set of tubes, the solvent was evaporated, and 40 μL of n-butanol containing 3 mol/L HCl was added. The mixture was incubated at 50 °C for 5 min. The excess derivatizing reagent was evaporated, and the residues were reconstituted with 75 μL of a mixture of methanol–0.005 mol/L ammonium formate (1:1 by volume) and transferred to labeled autosampler vials. recovery studies The experiments to evaluate an absolute extraction recovery were performed with human plasma samples containing MMA at concentrations of 0.4 and 4 μmol/L. The internal standard was added to the first group of samples before extraction, whereas for the second group, the internal standard was added into the extract. The dried extracts were derivatized according to the same procedure and analyzed at the same time. The percentage of recovery was determined by comparing the MMA concentration in the samples to which the internal standard was added after extraction to the results obtained for the samples to which the internal standard was added before extraction. precision, linearity, and sensitivity studies Method precision was determined by analyzing three replicates per day of plasma samples containing MMA at concentrations of 0.3, 0.75, and 1.2 μmol/L over a 5-day period. In addition, method precision was determined from the control values observed during routine use of the method (concentrations of 0.3 and 1.0 μmol/L in duplicate within 15 days). Instrument imprecision was determined by repetitive injections of an extracted sample containing 0.4 μmol/L MMA from the same vial. Linearity was evaluated by analyzing supplemented samples prepared at 1, 50, 100, 125, 150, 200, and 250 μmol/L. Method sensitivity was determined by analyzing supplemented samples containing progressively lower concentrations of MMA. We used a criterion of maintaining accuracy within ± 15%, imprecision (CV) <10%, and a qualitative ion ratio within ± 40% of the value set by the calibration to determine the upper limit of linearity and limit of quantification for the assay. Each sample was analyzed in duplicate over a 2-day period. patient sample comparison studies A total of 591 samples were included in the correlation study. A group of 211 samples was analyzed with the in-house GC-MS assay (15): 182 serum and plasma samples from patients, 13 urine samples from patients, and 48 MMA-supplemented samples in dialyzed plasma. A group of 380 serum and plasma samples from patients, with concentrations ranging from 0.1 to 3.5 μmol/L, were correlated to the in-house LC-MS/MS assay similar to one described by Magera et al. (18) that used chromatographic separation of MMA and SA. Sample preparation for the LC-MS/MS procedure was the same as for the evaluated method. The analysis was performed on the same LC column with a mobile phase consisting of 700 mL/L methanol and 300 mL/L ammonium formate buffer (0.005 mol/L), pH 6.5. The MMA-supplemented samples included in the study were used because the number of patient samples with MMA concentrations >10 μmol/L was not sufficient for the correlation. To account for bias in both the reference and the evaluated methods, the results were analyzed by Deming regression (19). interference studies Interference was evaluated by analyzing n-butyl esters of clinically important organic acids. The total number of acids included in the study was 77, with a concentration of 10 000 μmol/L for each individual acid. The esterified acids were analyzed by the evaluated method in MRM and product-ion scan (with precursor ion m/z 231) modes. Results ionization and fragmentation The ionization efficiency of n-butyl esters of different classes of organic acids was studied by positive-mode ES-MS. The molecular ions [M + H]+ and [M + NH4]+ were observed for all dicarboxylic acids that were evaluated, including MMA, SA, malonic, 2-hydroxyglutaric, adipic, ethylmalonic (EMA), dimethylmalonic (DiMMA), suberic, sebacic, and dodecanedioic acids. The absolute abundances of the [M + H]+ and [M + NH4]+ molecular ions of the studied dicarboxylic acids were found to be comparable to one another with an increased intensity of the [M + NH4]+ ion for the acids with higher molecular weights. A representative group of monocarboxylic acid n-butyl esters included 2-hydroxy-3-methylvaleric, 2-hydroxyisovaleric, 2-oxoisovaleric, and 2-hydroxybutyric acids. These acids were subjected to the same derivatization and ionization conditions. Under these conditions, monocarboxylic acid n-butyl ester molecular ions were not detectable, in contrast to the molecular ions of the dicarboxylic acids (Fig. 1 ). Figure 1. Open in new tabDownload slide Abundances of the M+1 (□) and M+17 (▪) mass ion fragments of butyl esters of selected dicarboxylic (columns A–I) and monocarboxylic (columns J–M) acids, measured at quadrupole 1. Dicarboxylic acids: A, MMA; B, malonic acid; C, 2-OH-glutaric acid; D, adipic acid; E, suberic acid; F, sebacic acid; G, dodecanedioic acid; H, EMA; I, DiMMA. Monocarboxylic acids: J, 2-OH-3-methylvaleric acid; K, 2-OH-isovaleric acid; L, 2-oxyisovaleric acid; M, 2-OH-butyric acid. Figure 1. Open in new tabDownload slide Abundances of the M+1 (□) and M+17 (▪) mass ion fragments of butyl esters of selected dicarboxylic (columns A–I) and monocarboxylic (columns J–M) acids, measured at quadrupole 1. Dicarboxylic acids: A, MMA; B, malonic acid; C, 2-OH-glutaric acid; D, adipic acid; E, suberic acid; F, sebacic acid; G, dodecanedioic acid; H, EMA; I, DiMMA. Monocarboxylic acids: J, 2-OH-3-methylvaleric acid; K, 2-OH-isovaleric acid; L, 2-oxyisovaleric acid; M, 2-OH-butyric acid. The fragmentation of various dicarboxylic acids was investigated to determine whether differences exist among the dicarboxylic acids and between the isomers, such as MMA and SA. The MS/MS product ion spectra for MMA and SA n-butyl ester [M + H]+ molecular ions are presented in Fig. 2, C and D. In addition to n-butyl esters, several other alkyl esters were evaluated for relative ionization efficiencies, fragmentation patterns, and the uniqueness of the product ion m/z 119 of MMA relative to SA (Fig. 3 ). Figure 2. Open in new tabDownload slide Negative-ion mode mass spectra of underivatized MMA (A) and SA (B), positive-ion mode mass spectra of the [M + 1]+ product ions of the n-dibutyl esters of MMA (C) and SA (D), and proposed structures of MMA ionization (structures in A and C). Figure 2. Open in new tabDownload slide Negative-ion mode mass spectra of underivatized MMA (A) and SA (B), positive-ion mode mass spectra of the [M + 1]+ product ions of the n-dibutyl esters of MMA (C) and SA (D), and proposed structures of MMA ionization (structures in A and C). Figure 3. Open in new tabDownload slide Intensities of the M+1 and M+17 mass ion fragments of selected diesters of MMA and SA (A) and intensity of the product ion m/z 119 originating from M+1 molecular ions of the diesters of MMA and SA (B), as measured at quadrupole 3 (Q3). Figure 3. Open in new tabDownload slide Intensities of the M+1 and M+17 mass ion fragments of selected diesters of MMA and SA (A) and intensity of the product ion m/z 119 originating from M+1 molecular ions of the diesters of MMA and SA (B), as measured at quadrupole 3 (Q3). Because the detection is specific to dicarboxylic acids, LC separation plays a lesser role in the method. A short LC column was used to separate compounds from the solvent front to eliminate possible signal suppression. The chromatograms for the monitored mass ion transitions for the extracted plasma control containing 0.4 μmol/L MMA and 1.5 μmol/L d3-MMA internal standard are presented in Fig. 4 . The calibrator containing MMA at 0.4 μmol/L and SA at 6 μmol/L was used to set a threshold for the branching ratio for the positive samples. The acceptance limit for the branching ratio was established in every run as ± 40% of the value observed in the 0.4 μmol/L calibrator. Figure 4. Open in new tabDownload slide Typical MRM profiles for an extracted serum sample containing 0.4 μmol/L MMA and 1.5 μmol/L d3-MMA internal standard. PanelsA and B show MMA transitions m/z 231→119 and m/z 231→175, respectively; panelsC and D show d3-MMA transitions m/z 234→122 and m/z 234→178, respectively. Figure 4. Open in new tabDownload slide Typical MRM profiles for an extracted serum sample containing 0.4 μmol/L MMA and 1.5 μmol/L d3-MMA internal standard. PanelsA and B show MMA transitions m/z 231→119 and m/z 231→175, respectively; panelsC and D show d3-MMA transitions m/z 234→122 and m/z 234→178, respectively. performance characteristics The absolute extraction recovery of MMA for the method was 77% ± 11%. The values obtained for within-run, between-run, and total imprecision for the results in the experiments for the precision study along with between-run precision determined from the control values observed over 3 weeks for the method in routine testing are presented in Table 1 . The CV for instrument imprecision was determined by re-injecting a sample from the same vial (n = 10) and was 0.9%. The assay was linear up to 150 μmol/L (r2 = 0.998). The linear regression equation at concentrations of 0.1–1.2 μmol/L was: y = 0.997x + 0.012 (r2 = 1.000). The limit of quantification for the method was 0.1 μmol/L with an observed accuracy of 98.5% and imprecision of 7%. The limit of detection was 0.05 μmol/L. Table 1. Method accuracy and imprecision. MMA, μmol/L . Within-run CV, % . Between-run CV, % . Total CV, % . Accuracy, % . 0.301 4.5 0.4 4.5 100 0.751 2.3 0.8 2.4 98.7 1.201 1.7 2.7 3.2 97.5 0.32 4.8 5.8 7.5 93.9 1.02 2.7 4.5 5.2 99.1 MMA, μmol/L . Within-run CV, % . Between-run CV, % . Total CV, % . Accuracy, % . 0.301 4.5 0.4 4.5 100 0.751 2.3 0.8 2.4 98.7 1.201 1.7 2.7 3.2 97.5 0.32 4.8 5.8 7.5 93.9 1.02 2.7 4.5 5.2 99.1 1 Three replicates per run over 5 days. 2 Two replicates per run over 15 days. Open in new tab Table 1. Method accuracy and imprecision. MMA, μmol/L . Within-run CV, % . Between-run CV, % . Total CV, % . Accuracy, % . 0.301 4.5 0.4 4.5 100 0.751 2.3 0.8 2.4 98.7 1.201 1.7 2.7 3.2 97.5 0.32 4.8 5.8 7.5 93.9 1.02 2.7 4.5 5.2 99.1 MMA, μmol/L . Within-run CV, % . Between-run CV, % . Total CV, % . Accuracy, % . 0.301 4.5 0.4 4.5 100 0.751 2.3 0.8 2.4 98.7 1.201 1.7 2.7 3.2 97.5 0.32 4.8 5.8 7.5 93.9 1.02 2.7 4.5 5.2 99.1 1 Three replicates per run over 5 days. 2 Two replicates per run over 15 days. Open in new tab method comparison Comparison with the in-house GC-MS method (15) for serum, plasma, and urine samples (Fig. 5 ) showed close concordance over the entire range of evaluated concentrations. In addition to the correlation with the GC-MS test, method performance was evaluated relative to a LC-MS/MS method that chromatographically resolved SA and MMA in a way similar to a previously described method (18). Samples included in the study contained MMA at concentrations corresponding to the general population and slightly increased concentrations, characteristic of vitamin B12 deficiency (Fig. 6 ). No significant bias was seen between the methods for the serum/plasma and urine samples. Of the 380 samples included in the correlation with the LC-MS/MS method, the m/z 175/119 ion ratio was outside of the acceptable range in 41 samples (10.8%) with MMA concentrations ≥0.4 μmol/L. Figure 5. Open in new tabDownload slide Comparison of MS/MS values with GC-MS for MMA analysis in serum, plasma, and urine (n = 211; A) and Bland–Altman plot of differences (B). (A), linear regression: y = 0.95x + 0.51 (r2 = 0.998; Sy|x = 2.6). (B), results <2.5 μmol/L. The dashed lines represent 2 SD (0.12 μmol/L). Mean difference, −0.03 μmol/L. Figure 5. Open in new tabDownload slide Comparison of MS/MS values with GC-MS for MMA analysis in serum, plasma, and urine (n = 211; A) and Bland–Altman plot of differences (B). (A), linear regression: y = 0.95x + 0.51 (r2 = 0.998; Sy|x = 2.6). (B), results <2.5 μmol/L. The dashed lines represent 2 SD (0.12 μmol/L). Mean difference, −0.03 μmol/L. Figure 6. Open in new tabDownload slide Comparison of MS/MS values with LC-MS/MS for MMA analysis in serum and plasma (n = 380; A) and Bland–Altman plot of differences (B). (A), linear regression: y = 0.996x + 0.009 (r2 = 0.999; Sy|x = 0.07). (B), results <2 μmol/L. The dashed lines represent 2 SD (0.12 μmol/L). Mean difference, −0.01 μmol/L. Figure 6. Open in new tabDownload slide Comparison of MS/MS values with LC-MS/MS for MMA analysis in serum and plasma (n = 380; A) and Bland–Altman plot of differences (B). (A), linear regression: y = 0.996x + 0.009 (r2 = 0.999; Sy|x = 0.07). (B), results <2 μmol/L. The dashed lines represent 2 SD (0.12 μmol/L). Mean difference, −0.01 μmol/L. An interference study showed that the only acid potentially causing interference with MMA analysis was SA. SA, when present at a concentration of 20 μmol/L, produced a signal equivalent to 0.4 μmol/L MMA with a significantly increased qualitative ion mass ratio (>2, when <0.75 is expected for MMA). Hemolysis did not interfere with the LC-MS/MS assay, whereas lipemic samples produced a substantial amount of precipitate in the final solution. Commonly used serum collection tubes (EDTA, heparin, oxalate, and citrate) were evaluated for compatibility with the method. Interference was observed only with the citrate collection tube. Citric acid at physiologic concentrations did not interfere with MMA, but when present in the collection tube, it interfered with the m/z 175 qualitative ion of MMA and caused a significant increase in the qualitative ion mass ratio m/z 175/119 (>20, when 0.75 was expected). No carryover was observed at up to 1000 μmol/L MMA. A set of 335 random patient serum and plasma samples submitted for MMA analysis was analyzed for SA to determine the distribution of concentrations usually present in the samples. A histogram of the observed SA concentrations is shown in Fig. 7 . Figure 7. Open in new tabDownload slide SA distribution in random patient serum and plasma samples (n = 334). Figure 7. Open in new tabDownload slide SA distribution in random patient serum and plasma samples (n = 334). Discussion At present, GC-MS is considered the methodology of choice for MMA analysis in body fluids. The main disadvantage of the methodology is the low sample throughput. LC-MS/MS has a potential for higher throughput for MMA quantification. Organic acids are readily detected as anions in the negative-ion mode (20)(21)(22)(23) when atmospheric pressure ionization is used; however, the lack of specificity in the negative-ion mode far outweighs any sensitivity advantage that may be realized. In the negative-ion mode, the structural isomers MMA and SA are indistinguishable from each other (Fig. 2, A and B). Therefore, negative-ion mode detection requires adequate chromatographic separation of MMA from SA and other low-molecular weight organic acids potentially present in biological samples (20)(21)(22)(23). To speed up the analysis, we attempted to identify conditions that would produce unique mass ion fragments for MMA. One of the possibilities for obtaining selective fragmentation for MMA and SA is to analyze their derivatives rather than the parent compounds. For LC-MS applications using atmospheric pressure ionization techniques such as TIS, organic acid esterification is counterintuitive because the nonpolar derivative in solution is less likely to be ionized than its more polar underivatized form. Despite this, our experiments showed that dicarboxylic acid diesters produced positively charged molecular ions. The ionization study for the dicarboxylic acids (Fig. 1 ) demonstrated that all of the evaluated n-butyl esters of dicarboxylic acids produced [M + H]+ molecular ions and [M + NH4]+ ammonium adducts, if ammonium ions were present in solution. It was observed that neither the number of methylene groups between the carboxyl groups nor additional functional groups present within the structure (e.g., hydroxy or keto groups) affected the ability of the n-butyl diesters to produce stable positively charged molecular ions. This positive-ion formation for dicarboxylic acid diesters can be explained through resonance of the tautomer structures promoted by the delocalization of charge within the molecule (Fig. 2C ) and mobility of the hydrogen atoms attached to active α carbons (24). The molecules may exist in solution as protonated or ammoniated adducts, both of which were observed in the experiments (Figs. 1 and 3 ). To support the hypothesis that positive-ion formation takes place through a resonance of keto-enol equilibrium within the part of the MMA molecule between the two carboxyl groups, we evaluated, along with other acids, fragmentation of n-butyl diesters of two other structural isomers dicarboxylic acids, DiMMA and EMA acids (Fig. 1 ). The results of the experiment demonstrated that in single-MS mode, the [M + H]+ ion at m/z 245 was the major ion of the n-butyl-EMA and the intensity of the [M + H]+ ion of n-butyl-DiMMA at m/z 245 was not higher than the background noise. A similar result was observed for the ammonium adduct ion m/z 262, which was abundant for n-butyl-EMA and completely absent for n-butyl-DiMMA. The difference between the isomers is that the α carbon in EMA is tertiary, whereas in DiMMA it is quaternary. The phenomenon that the [M + H]+ ion does not exist for DiMMA can be explained by the fact that the α carbon in the DiMMA molecule does not have an active hydrogen that would allow tautomer formation. The absolute abundances of the [M + H]+ and [M + NH4]+ molecular ions of the evaluated esters of MMA and SA were substantially different from one ester to another, with increased intensities for the fragments from the esters with higher molecular weight (Fig. 3 ). Comparison of the results for fragmentation of different diesters of MMA demonstrated that the n-butyl derivative provided the best specificity for MMA vs SA, whereas the n-amyl derivative showed better sensitivity but less specificity than the n-butyl diester (Fig. 3 ). All other diesters created less specific and less sensitive analytical conditions for MMA. Because the sensitivity of the n-butyl derivative was adequate, we chose to use this derivatizing reagent. A study of the collisionally induced dissociation fragmentation mechanism for different dicarboxylic acids revealed that MS/MS fragment ions at m/z 175 and 119 (loss of 56 and 112 Da) in n-butyl-MMA are unique compared with the other evaluated dicarboxylic acids derivatives. For n-butyl-SA the major product ions from the M+1 molecular ion were obtained by the loss of 74 and 130 Da. Fragmentation of the molecular ions of a series of dicarboxylic acid n-butyl diesters (2-hydroxyglutaric, adipic, suberic, sebacic and dodecanedioic acids) was similar to that for SA. Of the evaluated dicarboxylic acids, only MMA and malonic acid showed the neutral losses of 56 and 112 Da in the MS/MS mode; the rest of the acid diesters showed a preferred neutral loss of 74 and 130 Da, similar to SA (Fig. 2 ). At the same time we did not observe [M + H]+ and [M + NH4 ]+ molecular ions for any of the evaluated monocarboxylic acid esters (Fig. 1 ). The major product ions observed by ES-MS/MS of dicarboxylic acid dibutyl ester [M + H]+ as the precursor ion are presented in Scheme 1. Scheme 1. Open in new tabDownload slide Major product ions observed by ES-MS/MS of dicarboxylic acid dibutyl esters. Scheme 1. Open in new tabDownload slide Major product ions observed by ES-MS/MS of dicarboxylic acid dibutyl esters. The fragment ions B and D in Scheme 1 are common to all dicarboxylic acids, but the fragment ions A and C are almost completely missing in SA. The observed difference in fragmentation provides the necessary specificity for the method. The sensitivity of the evaluated method and the fragmentation pattern that were observed for dibutyl MMA differed from the sensitivity and fragmentation pattern described by Magera et al. (18) for their method. It appeared that, under the conditions used in the present method, the product ion spectrum exhibited extra mass ion fragments at m/z 175, 157, and 101 for MMA and SA. Although additional product ions detract from the ultimate achievable sensitivity, they enhance the specificity of the assay by permitting monitoring of the product ion-mass branching ratios. Sample preparation for the procedure incorporated liquid/liquid extraction with MTBE of a sample acidified with phosphoric acid. As determined previously (15), MTBE is a selective solvent for separating MMA from a sample matrix, leading to a cleaner extract compared with ethyl acetate and other organic solvents typically used for the extraction of organic acids. The selectivity for dicarboxylic acids and the unique fragmentation of the MMA diester form the basis for this method of MMA analysis. Total instrumental analysis time was 60 s from injection to injection, with an MMA retention time of ∼40 s, one-third of the time required for the recently published LC-MS/MS method (18). There is very little retention of the analyte on the HPLC column, which serves only to separate MMA from the solvent front and to provide an improvement of the chromatographic peak shape. The ratio of the integrated peak response for MMA transition m/z 231→119 vs the d3-MMA transition m/z 234→122 was used to calculate the concentration of MMA. The MS/MS transitions 231→175 and 234→178 were used for qualitative confirmation purposes to assure correct identification of MMA and the absence of interference from coextracted sample constituents. The key features of the present LC-MS/MS method are its selectivity for dicarboxylic acids and the ability to differentiate MMA from SA, a potential endogenous interferent. SA is the final product of the metabolic conversion of propionic acid and is present at concentrations greater than those of MMA. Fortunately, the m/z 231→119 and m/z 231→175 fragmentation pathways are ∼100 and 30 times more abundant for MMA than for SA, respectively. The SA n-butyl ester produced a minor amount of the same product fragment ions at m/z 119 and 175 as MMA. The relative abundances of the ions are substantially different between the compounds. The branching ion-mass ratio m/z 175/119 for MMA is 0.3 ± 0.05, whereas the ratio for the same fragments of SA is 2.0 ± 0.1. When both MMA and SA are present in a sample, the branching ratio of the two product ions enables the estimation of the SA contribution to MMA quantification. To determine the potential extent of SA interference with the method and to establish a baseline for the expected SA concentration, we evaluated SA concentration distribution in patient samples. The mean, mode, and SD for SA distribution in the patient samples were 5.7, 5.1, and 4.9, respectively. Because SA is a typical constituent of samples and the distribution of the concentrations is relatively narrow, to compensate for SA contribution to MMA quantification and to establish acceptability limits for qualitative ion ratios we included SA in the calibrators and controls at a concentration corresponding to the mean value observed in the studied population. From the results of the method evaluation, the branching ion-mass ratios appeared to be highly reproducible and can serve as an accurate predictor of SA interference with MMA quantification. Potential interference from SA was evaluated from the branching ratios only for samples with MMA concentrations >0.4 μmol/L. For samples with MMA concentrations below the cutoff, the branching ion ratio was not evaluated. If the branching ion ratio did not exceed the value set by the calibration, the quantitative results were accepted. Samples with ion ratios outside of the limits should be reanalyzed by a method that chromatographically separates MMA and SA. The branching ion ratio of m/z 178/122 was used to assess potential interference with d3-MMA in all samples (negative and positive). The validity of this approach was evaluated through correlation with a LC-MS/MS method that chromatographically resolved MMA and SA (Fig. 6 ). The results showed good agreement between the methods. Evaluation of the results of the correlation showed that 89.2% of the analyzed samples produced acceptable results, whereas 10.8% of the samples needed to be reanalyzed by a method that chromatographically resolves MMA and SA because of unacceptable qualitative ion ratios. All samples with a MMA concentration >1.5 μmol/L showed m/z 175/119 ion ratios within acceptance limits and a narrow distribution of 0.44–0.56. The qualitative ion mass ratio for the internal standard was outside of the established limits in only two of the analyzed patient samples. In conclusion, we present a method for the selective detection of dicarboxylic acids based on the formation of positively charged molecular ions and subsequent unique fragmentation. The advantages of the method are its sensitivity to dicarboxylic acids and the ability to selectively quantify the structural isomers MMA and SA. The selectivity of the method is based on the facts that nonderivatized acids are transparent to the detector and that the esters of monocarboxylic acids do not produce stable [M + H]+ and [M + NH4]+ molecular ions. Such selective detection eliminates the necessity for extensive chromatographic separation and shortens the instrument analysis time for MMA to less than one-tenth of the analysis time of a conventional GC-MS method. The instrumental analysis portion for our method is ∼3 times faster than a recently published method (18) that uses MS/MS. The entire analysis, including sample preparation for a batch of 100 specimens, can be performed in <4 h. We thank MDS SCIEX for providing a LC-MS/MS instrument, and the ARUP Institute for Clinical and Experimental Pathology for financial support. 1 " Present address: Specialty Laboratories, Santa Monica, CA 90404-3900. 2 " Nonstandard abbreviations: MMA, methylmalonic acid; SA, succinic acid; GC, gas chromatography; MS, mass spectrometry; LC, liquid chromatography; ES, electrospray; MS/MS, tandem MS; MTBE, methyl-tert-butyl ether; TIS, TurboIonSpray; MRM, multiple reaction monitoring; EMA, ethylmalonic acid; and DiMMA, dimethylmalonic acid. References 1 Cox EV, White AM. Methylmalonic acid excretion: an index of vitamin B12 deficiency. Lancet 1962 ; 2 : 853 -856. PubMed 2 Fairbanks VF, Klee GG. Biochemical aspects of hematology. Burtis CA Ashwood ER eds. Tietz textbook of clinical chemistry 3rd ed. 1999 : 1692 -1693 WB Saunders Philadelphia. . 3 Fenton WA, Rosenberg LE. Disorders of propionate and methylmalonate metabolism. Scriver CR Beaudet AL Sly WS Valle D eds. The metabolic and molecular bases of inherited disease 8th ed. 2001 : 2177 -2193 McGraw-Hill New York. . 4 Barness LA, Young D, Mellman WJ, Kahn SB, Williams WJ. Methylmalonate excretion in a patient with pernicious anemia. N Engl J Med 1963 ; 268 : 144 -146. Crossref Search ADS PubMed 5 Bashir HV, Hinterberger H, Jones BP. Methylmalonic acid excretion in vitamin B12 deficiency. Br J Haematol 1966 ; 12 : 704 -711. Crossref Search ADS PubMed 6 Schneede J, Ueland PM. Automated assay of methylmalonic acid in serum and urine by derivatization with 1-pyrenyldiazomethane, liquid chromatography, and fluorescence detection. Clin Chem 1993 ; 39 : 392 -399. Crossref Search ADS PubMed 7 Babidge PJ, Babidge WJ. Determination of methylmalonic acid by high-performance liquid chromatography. Anal Biochem 1994 ; 216 : 424 -426. Crossref Search ADS PubMed 8 Gibbs BF, Itiaba K, Mamer OA, Crawhall JC, Cooper BA. A rapid method for the analysis of urinary methylmalonic acid. Clin Chim Acta 1972 ; 38 : 447 -453. Crossref Search ADS PubMed 9 Norman EJ, Martelo OJ, Denton MD. Cobalamin (vitamin B12) deficiency detection by urinary methylmalonic acid quantitation. Blood 1982 ; 59 : 1128 -1131. Crossref Search ADS PubMed 10 Nakamura E, Rosenberg LE, Tanaka K. Microdetermination of methylmalonic acid and other short chain dicarboxylic acids by gas chromatography: use in prenatal diagnosis of methylmalonic acidemia and in studies of isovaleric acidemia. Clin Chim Acta 1976 ; 68 : 127 -140. Crossref Search ADS PubMed 11 Marcell PD, Stabler SP, Podell ER, Allen RH. Quantitation of methylmalonic acid and other dicarboxylic acids in normal serum and urine using capillary gas chromatography-mass spectrometry. Anal Biochem 1985 ; 150 : 58 -66. Crossref Search ADS PubMed 12 Stabler SP, Marcell PD, Podell ER, Allen RH, Lindenbaum J. Assay of methylmalonic acid in the serum of patients with cobalamin deficiency using capillary gas chromatography-mass spectrometry. J Clin Invest 1986 ; 77 : 1606 -1612. Crossref Search ADS PubMed 13 Rasmussen K. Solid-phase sample extraction for rapid determination of methylmalonic acid in serum and urine by a stable-isotope-dilution method. Clin Chem 1989 ; 35 : 260 -264. Crossref Search ADS PubMed 14 McCann MT, Thompson MM, Gueron IC, Lemieux B, Giguere R, Tuchman M. Methylmalonic acid quantification by stable isotope dilution gas chromatography-mass spectrometry from filter paper urine samples. Clin Chem 1996 ; 42 : 910 -914. Crossref Search ADS PubMed 15 Kushnir MM, Komaromy-Hiller G. Optimization and performance of a rapid gas chromatography-mass spectrometry analysis for methylmalonic acid. J Chromatogr B 2000 ; 741 : 231 -241. Crossref Search ADS 16 Allen RH, Stabler SP, Lindenbaum J, inventors. US patent 4,940,658, July 10, 1990.. 17 Johnson DW. Dimethylaminoethyl esters for trace, rapid analysis of fatty acids by electrospray tandem mass spectrometry. Rapid Commun Mass Spectrom 1999 ; 13 : 1 -6. Crossref Search ADS PubMed 18 Magera MJ, Helgeson JK, Matern D, Rinaldo P. Methylmalonic acid measurement in plasma and urine by stable-isotope dilution and electrospray tandem mass spectrometry. Clin Chem 2000 ; 46 : 1804 -1810. Crossref Search ADS PubMed 19 Cornbleet PJ, Gochman N. Incorrect least-squares regression coefficients in method-comparison analysis. Clin Chem 1979 ; 25 : 432 -438. Crossref Search ADS PubMed 20 Hunt DF, Giordani AB, Rhodes G, Herold DA. Mixture analysis by triple-quadrupole mass spectrometry: metabolic profiling of urinary carboxylic acids. Clin Chem 1982 ; 28 : 2387 -2392. Crossref Search ADS PubMed 21 Mills GA, Walker V, Cleanch MR, Parr VC. Analysis of urinary organic acids by Plasmaspray liquid chromatography/mass spectrometry. Biomed Environ Mass Spectrom 1988 ; 16 : 259 -261. Crossref Search ADS PubMed 22 Buchanan NB, Munzer J, Thoene JG. Positive-ion thermospray liquid chromatography- mass spectrometry: detection of organic acidurias. J Chromatogr B 1990 ; 534 : 1 -11. Crossref Search ADS 23 Kajia M, Niwa T, Watanabe K. Analysis of urinary organic acids by liquid chromatography-atmospheric pressure chemical ionization mass spectrometry. J Chromatogr B 1993 ; 622 : 263 -268. Crossref Search ADS 24 Ege SN. Organic chemistry. Structure and reactivity 1999 : 651 -662 Houghton Mifflin Boston. . © 2001 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)

zur Stadt,, Udo;Rischewski,, Johannes;Schneppenheim,, Reinhard;Kabisch,, Hartmut

doi: 10.1093/clinchem/47.11.2003pmid: N/A

Abstract Background: Denaturing HPLC (DHPLC) can be used to screen DNA for known and unknown mutations. We describe a novel, HPLC-based method for discrimination among polyclonal, oligoclonal, and/or clonal T-cell receptor γ (TCR-γ) rearrangements in samples from children with newly diagnosed acute lymphoblastic leukemia. Methods: TCR rearrangements were PCR amplified from initial leukemic samples and, after heteroduplex-induction, the clonality status of each product was evaluated. To attain this, we used DHPLC on a high-resolution micropellicular matrix. Running conditions were established by melting-curve analysis of known clonal and polyclonal products and melting-point prediction software. Elution profiles were studied at 50 °C (native) and, to achieve optimal separation, at different column temperatures between 56 and 64 °C. Results: For VγI-Jγ1.3/2.3 rearrangements, an analysis temperature of 60 °C with a linear triethylammoniumacetate—acetonitrile gradient separated clonal bands from the polyclonal background amplification. Less than 15% clonal PCR product was detectable in mixtures of initial leukemic cell DNA and polyclonal DNA. Biallelic rearrangements produced two sharp peaks. Clonality of PCR products from 100 initial leukemic samples was completely identified in all investigated cases. Conclusions: Heteroduplex analysis with standardized DHPLC conditions simplifies the detection of unknown clonal or polyclonal TCR rearrangements in newly diagnosed leukemias. Clonal targets for detection of minimal residual disease are available after a short, automated analysis of PCR amplified rearrangements. Initial diagnosis of acute lymphoblastic leukemia (ALL) 2 is established by routine cytomorphologic and immunphenotypical analysis. Important improvements to achieve better outcomes are based on the identification of new clinical and biologic markers. The initial response to remission-induction therapy is one of the most important prognostic factors in childhood ALL. For example, a slow response to induction therapy is associated with a high risk of relapse (1)(2). Complete remission is defined as a blast cell fraction of <5% in the bone marrow examined by light microscopy. Detection of minimal residual disease (MRD) is a new practical tool for a more exact measurement of remission induction during therapy because leukemic blasts can be detected down to 10−4–10−6. In the future, this may be the way for a new definition of remission in childhood ALL (3). PCR-based MRD analysis uses clonal antigen receptor rearrangements detectable in ∼90–95% of the investigated patient samples. Amplification of polyclonal products often leads to false-positive PCR amplicons not suitable for MRD analysis. Therefore, PCR analysis alone will only be sufficient if a more detailed analysis follows. Single-strand conformation analysis, heteroduplex analysis, temperature gradient gel electrophoresis, or gene scanning analysis have been described to circumvent this background problem (4)(5)(6)(7)(8)(9)(10). Previous reports indicated high sensitivity and specificity of denaturing HPLC (DHPLC) for mutation detection (11)(12)(13)(14)(15) and for detection of single-nucleotide polymorphisms (16). The detection principle is based on the instability of a PCR product’s helical structure at or close to the melting temperature (partially denatured PCR product). Sequence variations will increase or decrease the number and spatial positions of hydrogen bonds within a double-stranded (full-helical) PCR product. This causes an altered percentage of nonhelical, or partially single-stranded segments within the PCR product. The conformational changes become detectable because of the altered physico-chemical interaction of the negatively loaded surface of the PCR product and the reversed-phase column. The higher the single-stranded fraction, the earlier the PCR product will elute from the column as the negative loads on the DNA surface are further spaced. The interaction of negative surface loads of the PCR product and the lipophilic surface of the column is mediated by a slow linear decrease of a triethylammonium acetate (TEAA) buffer. At partial-denaturing conditions, this buffer decrease leads to different retention times of hetero- and homoduplex products [sequence-dependent separation; for a review see Ref. (17)]. In general, PCR products contain two alleles, A and B, and each allele has a sense (A, B) and an antisense (A*, B*) orientation. After heat denaturation and during heteroduplex induction, single strands are allowed to slowly reanneal, and homoduplexes (AA*, BB*) and heteroduplexes (AB*, BA*) will be formed. At partial-denaturing conditions, the percentage of nonhelical PCR product will be higher in the heteroduplex fraction, and this leads to an earlier elution of this fraction. The method uses an automated instrumentation setup, and each sample can be analyzed within 5–10 min. The aim of our investigation was to establish a routine method for identification of clonal rearrangements in newly diagnosed leukemias with this highly automated system. We applied the DHPLC principle to the detection of a present clonal cell fraction within a polyclonal background. The polyclonal product contains a multitude of sequence variations regarding length and base composition in the junctional region. With partial-denaturing HPLC conditions, no detectable single clonal peak should be present because the given multitude of sequence variations should lead to an even higher amount of heteroduplexes. In contrast, a clonal PCR product should contain one highly predominant sequence and should therefore elute as one distinct homoduplex peak. To test this, we evaluated the sensitivity and reliability of this method for the most common VγI-Jγ1.3/2.3 rearrangement, using patient samples with a known or unknown clonality status. Here we describe the standard conditions for these targets and our evaluation of the method on 100 initial leukemic samples. Materials and Methods Bone marrow samples from the time of diagnosis were collected during the 1997 Co-operative Study Group for Childhood Acute Leukemia (COALL) trial. Ficoll-separated cells were frozen in 100 mL/L dimethyl sulfoxide–500 mL/L fetal bovine serum–RPMI 1640 and stored at −80 °C until use. DNA from 106–107 cells was isolated using the High Pure PCR Template Preparation Kit (Roche Molecular Diagnostics). Three different leukemic samples and a mixture of blood mononuclear cells from five different healthy donors (as a polyclonal control) were investigated. pcr and heteroduplex induction We performed PCR in 50-μL reactions containing 100–300 ng of DNA for each reaction. We identified PCR targets for detection of MRD from initial leukemic DNA using a standard set of primers (18)(19). The PCR product was subjected to heteroduplex induction as follows: preheating at 95 °C; denaturing for 5 min at 95 °C; gradually reannealing from 95 °C to 10 °C with 1 °C/min; and storage at 4 °C before analysis. Samples (5–10 μL) were injected in a preheated C18 reversed-phase column with nonporous poly-(styrene-divinybenzene) particles (DNASep®; Transgenomic). The injected sample was eluted within a linear acetonitrile gradient consisting of buffer A (0.1 mol/L TEAA) and buffer B (0.1 mol/L TEAA, 250 mL/L acetonitrile) with a 2% increase of buffer B per minute. PCR products were separated with a flow rate of 0.9 mL/min. Retention time was measured on-line via ultraviolet absorbance at 254 nm in the eluate. Result diagrams showed absorbance intensity in millivolts over the retention time in minutes (mV/min) after injection into the column. analysis temperature and gradient estimation We analyzed PCR products on the HPLC system, using the following analytical conditions for optimization strategies. (a) We predicted the melting-temperature area of a 532-bp VγI-Jγ1.3/2.3 product (resulting from a theoretically rearranged germline product without deleted or added nucleotides). This sequence serves as the basis for the melting-point prediction software analysis (WAVEMAKER; Transgenomic). (b) We estimated the melting profile by serial injections of PCR products at different column temperatures. The latter always included a run at 50 °C (complete double-stranded product), at 70 °C (complete single-stranded product), and at several temperatures (56–64 °C) close to the predicted melting temperature. Several PCR products were used to establish the method: a polyclonal, a clonal, a biclonal, and two mixed-clonal products. Results software-predicted melting profile The percentage of helical (double-stranded) PCR product within a temperature range of 50–70 °C is shown in Fig. 1B . The melting point was defined as the temperature at which 50% of the PCR product had a helical structure. To achieve optimal separation of homo- and heteroduplexes, DHPLC analysis had to be performed close to the melting point of the PCR product. The theoretical percentage of helical, i.e., fully double-stranded, product vs the expected analysis temperature is demonstrated in Fig. 1 . The software demonstrated that the PCR product stayed fully double stranded up to 55 °C. Thereafter, a two-step melting profile was observed, giving 56 and 59 °C as the points with the steepest decline from double stranded to partially single stranded. At 65 °C, all product was predicted to be melted (fully single stranded). Different junctional regions led to a different temperature stability of the helical configuration. These differences had the most influence on the retention patterns 1 or 2 °C below the points of steepest decline (56 and 59 °C). Figure 1. Open in new tabDownload slide Melting-curve prediction software. (A), sequence of the analyzed TCR γ rearrangement. The junctional region is shown as the product of the two rearranged germline segments without any diversity. Primer sequences are underlined and comparable to the VγI-Jγ1.3/2.3 primer pair (18). (B), melting profile of a TCR γ product. The percentage of helical fraction vs temperature is shown. The PCR product has two melting domains (↑). This curve is used to select the temperature for the optimization runs. (C), local melting behavior of the TCR γ PCR product. The base composition vs the melting temperature (Tm) of distinct regions is plotted. This profile demonstrates that differences in the junctional region are more likely detectable in the second melting domain (2↑). Figure 1. Open in new tabDownload slide Melting-curve prediction software. (A), sequence of the analyzed TCR γ rearrangement. The junctional region is shown as the product of the two rearranged germline segments without any diversity. Primer sequences are underlined and comparable to the VγI-Jγ1.3/2.3 primer pair (18). (B), melting profile of a TCR γ product. The percentage of helical fraction vs temperature is shown. The PCR product has two melting domains (↑). This curve is used to select the temperature for the optimization runs. (C), local melting behavior of the TCR γ PCR product. The base composition vs the melting temperature (Tm) of distinct regions is plotted. This profile demonstrates that differences in the junctional region are more likely detectable in the second melting domain (2↑). The melting behavior of the PCR products depends on the base-pair composition of a given sequence. The different interval domains of the PCR product with respect to their local melting point are demonstrated in Fig. 1C . An AT-rich region in the first part of the product leads to an earlier melting domain at 56 °C. Our region of interest (the junctional region of rearranged TCR genes; Fig. 1A ) has a relatively higher GC content, leading to a higher temperature stability, and is expected to melt at ∼59 °C. Typically, melting-curve prediction software visualizes the melting behavior of each target; it does not replace optimization runs to define analysis conditions. In the following experiments we compared these theoretical predictions to the practical approaches. dhplc analysis conditions All optimization runs were performed with gradients according to Tables 1 and 2 . After heteroduplex induction, 8–10-μL samples were injected into the preheated column. A polyclonal VγI-Jγ1.3/2.3 PCR product was analyzed at 50, 56, 57, 58, 59, 60, 61, 62, 63, 64, and 70 °C (Fig. 2 ). At 50 °C, the polyclonal products elute as a single peak. This is in accordance with the literature (17), which describes a length-discrimination performance of 1–3% at nondenaturing conditions. A 530-bp product with a length variation (gaussian distribution) of ± 20 nucleotides should not be discernable. With higher temperatures, this peak “melts” and a broader peak elutes earlier from the column. At 60 °C, a plateau-like peak elutes between 10 and 12 min. In contrast, the resulting clonal VγI-Jγ 1.3 product from a T-cell ALL (T-ALL) sample at the time of diagnosis showed a distinct (clonal) peak at this temperature (Fig. 2A ). This peak corresponds to a sequence with seven deleted and six randomly added nucleotides in the junctional region (Table 3 ). Table 1. DHPLC conditions for separation of VγI-Jγ1.3/2.3 targets. Time, min . Buffer, % (50°C) . . Time, min . Buffer, % (60°C) . . . A . B . . A . B . 0 65 35 0 57 43 0.1 601 40 0.1 521 48 16.1 28 72 7.1 38 62 16.2 0 100 7.2 0 100 16.7 0 100 7.7 0 100 16.8 65 35 7.8 57 43 18.0 65 35 9.0 57 43 Time, min . Buffer, % (50°C) . . Time, min . Buffer, % (60°C) . . . A . B . . A . B . 0 65 35 0 57 43 0.1 601 40 0.1 521 48 16.1 28 72 7.1 38 62 16.2 0 100 7.2 0 100 16.7 0 100 7.7 0 100 16.8 65 35 7.8 57 43 18.0 65 35 9.0 57 43 1 2% decrease/min. Open in new tab Table 1. DHPLC conditions for separation of VγI-Jγ1.3/2.3 targets. Time, min . Buffer, % (50°C) . . Time, min . Buffer, % (60°C) . . . A . B . . A . B . 0 65 35 0 57 43 0.1 601 40 0.1 521 48 16.1 28 72 7.1 38 62 16.2 0 100 7.2 0 100 16.7 0 100 7.7 0 100 16.8 65 35 7.8 57 43 18.0 65 35 9.0 57 43 Time, min . Buffer, % (50°C) . . Time, min . Buffer, % (60°C) . . . A . B . . A . B . 0 65 35 0 57 43 0.1 601 40 0.1 521 48 16.1 28 72 7.1 38 62 16.2 0 100 7.2 0 100 16.7 0 100 7.7 0 100 16.8 65 35 7.8 57 43 18.0 65 35 9.0 57 43 1 2% decrease/min. Open in new tab Table 2. DHPLC conditions for optimization run combinations at 50 °C (native) and at 60 °C (partial denaturing). Time, min . Buffer, % . . . A . B . 0 49 51 0.1 441 56 8.1 28 72 8.2 0 100 8.7 0 100 8.8 49 51 9.7 49 51 Time, min . Buffer, % . . . A . B . 0 49 51 0.1 441 56 8.1 28 72 8.2 0 100 8.7 0 100 8.8 49 51 9.7 49 51 1 2% decrease/min. Open in new tab Table 2. DHPLC conditions for optimization run combinations at 50 °C (native) and at 60 °C (partial denaturing). Time, min . Buffer, % . . . A . B . 0 49 51 0.1 441 56 8.1 28 72 8.2 0 100 8.7 0 100 8.8 49 51 9.7 49 51 Time, min . Buffer, % . . . A . B . 0 49 51 0.1 441 56 8.1 28 72 8.2 0 100 8.7 0 100 8.8 49 51 9.7 49 51 1 2% decrease/min. Open in new tab Figure 2. Open in new tabDownload slide Optimal temperature selection. DHPLC melting profiles of four VγI-Jγ1.3/2.3 products. Melting curves are drawn as intensity (mV) over retention time (min) at 11 different column oven temperatures. At 50 °C, all products show a single elution peak between 14 and 15 min (peak1). Denatured products (at 70 °C) are eluted after 6–7 min (peak3), whereas partially denatured products (56–64 °C) with different helical fractions are eluted later (peak 2). PanelA shows a polyclonal (left) and a clonal (right) product, and panelB, a biclonal (left) product from one patient sample and two mixed-clonal (right) products from two patients, respectively. The most discriminating temperature for this target is represented at 60 °C. The elution profiles of clonal and biclonal products are mainly influenced by the base composition of the junctional region and are nearly independent of product length. Figure 2. Open in new tabDownload slide Optimal temperature selection. DHPLC melting profiles of four VγI-Jγ1.3/2.3 products. Melting curves are drawn as intensity (mV) over retention time (min) at 11 different column oven temperatures. At 50 °C, all products show a single elution peak between 14 and 15 min (peak1). Denatured products (at 70 °C) are eluted after 6–7 min (peak3), whereas partially denatured products (56–64 °C) with different helical fractions are eluted later (peak 2). PanelA shows a polyclonal (left) and a clonal (right) product, and panelB, a biclonal (left) product from one patient sample and two mixed-clonal (right) products from two patients, respectively. The most discriminating temperature for this target is represented at 60 °C. The elution profiles of clonal and biclonal products are mainly influenced by the base composition of the junctional region and are nearly independent of product length. Table 3. Sequences of the analyzed clonal/biclonal TCR γ targets. Germline . ANNNN . GAATTATTATAAGAAAC . Length . 1a ACGGG CCAGGA TATAAGAAAC −7/+6/0 1b ACAGG CGCCCCATGTCCCATAA ATAAGAAAC −8/17/0 2 ATGCG GAGGTC TATAAGAAAC −7/+6/0 3 ACG TT AATTATTATAAGAAAC −1/+2/−2 Germline . ANNNN . GAATTATTATAAGAAAC . Length . 1a ACGGG CCAGGA TATAAGAAAC −7/+6/0 1b ACAGG CGCCCCATGTCCCATAA ATAAGAAAC −8/17/0 2 ATGCG GAGGTC TATAAGAAAC −7/+6/0 3 ACG TT AATTATTATAAGAAAC −1/+2/−2 Open in new tab Table 3. Sequences of the analyzed clonal/biclonal TCR γ targets. Germline . ANNNN . GAATTATTATAAGAAAC . Length . 1a ACGGG CCAGGA TATAAGAAAC −7/+6/0 1b ACAGG CGCCCCATGTCCCATAA ATAAGAAAC −8/17/0 2 ATGCG GAGGTC TATAAGAAAC −7/+6/0 3 ACG TT AATTATTATAAGAAAC −1/+2/−2 Germline . ANNNN . GAATTATTATAAGAAAC . Length . 1a ACGGG CCAGGA TATAAGAAAC −7/+6/0 1b ACAGG CGCCCCATGTCCCATAA ATAAGAAAC −8/17/0 2 ATGCG GAGGTC TATAAGAAAC −7/+6/0 3 ACG TT AATTATTATAAGAAAC −1/+2/−2 Open in new tab In two other experiments, a biclonal patient sample and two mixed-clonal products were analyzed under the same conditions (Fig. 2B ). The aim was to test the effect of sequence variations only and in combination with length differences. The biclonal product (amplified from one patient sample) was cloned and sequenced and showed a length difference of 10 nucleotides (germlines 1a and 1b in Table 3 ). The clonal products from two known samples that were mixed at equal parts after PCR amplification are shown in Fig. 2 . The two products showed no length difference but a different junctional region (germlines 2 and 3 in Table 3 ). The elution profiles from the DHPLC analysis are highly sequence specific and have a distinct retention time. Length separation plays only a minor role at partial-denaturing conditions. The characteristic profiles of four clonal and biclonal products together with a polyclonal control are summarized in Fig. 3 . Sequence variations in the junctional region have a dramatic effect on the separation power of these fragments. Elution times differed by >0.6 min (last three elution profiles in Fig. 3 ). Therefore, separation was observed only with different helical fragments, i.e., induction of heteroduplexes via different junctional regions. The sequences of the analyzed products are summarized in Table 3 . We cloned and sequenced the biclonal rearrangement. Two different clones corresponding to the biclonal target were analyzed at 50 and 60 °C. To reduce analysis time, the TEAA gradient was shortened to the discriminating area according to Table 2 . New conditions corresponded to the gradient between 3 min and 8 min of the previous runs. DHPLC analysis with this shortened, optimized gradient indicated that the broad biclonal peak was a consequence of two overlapping single peaks (Fig. 3 , lower right panel at 60 °C). No genomic artifacts or heteroduplex peaks were observed. Figure 3. Open in new tabDownload slide Partial-denaturing HPLC analysis (DHPLC). Summary of five optimized DHPLC runs at a column oven temperature of 60 °C. A polyclonal PCR product was compared with biclonal and clonal amplicons. Retention profiles are highly specific for a distinct product as demonstrated by the artificially mixed biclonal product originating from two clonal patient samples. Each peak from the mixed elution profile is identical to the peaks from single amplified patient samples a and b. The lower panels show the two cloned fragments from the biclonal patient sample analyzed at 50 °C (left) and at 60 °C (right). Figure 3. Open in new tabDownload slide Partial-denaturing HPLC analysis (DHPLC). Summary of five optimized DHPLC runs at a column oven temperature of 60 °C. A polyclonal PCR product was compared with biclonal and clonal amplicons. Retention profiles are highly specific for a distinct product as demonstrated by the artificially mixed biclonal product originating from two clonal patient samples. Each peak from the mixed elution profile is identical to the peaks from single amplified patient samples a and b. The lower panels show the two cloned fragments from the biclonal patient sample analyzed at 50 °C (left) and at 60 °C (right). detection limit for identification of clonal targets at diagnosis Because most of the initial leukemic cell samples contained variable numbers of nonpathologic hematopoietic cells, sensitivity testing for identification of clonal targets was performed. Because bone marrow or peripheral blood usually contains at least >20% leukemic blasts at diagnosis, a detection limit of 15% in the mononuclear fraction should be sufficient to identify clonal targets. With the buffer gradient described above at 60 °C, several dilution steps were analyzed for evaluation of the minimal amount of leukemic cells necessary for detection of clonality. We serially diluted leukemic cell DNA in 12.5% steps into unrelated polyclonal DNA, PCR amplified the VγI-Jγ1.3/2.3 fragment, and subsequently induced heteroduplex formation. Down to a dilution of 12.5%, a clonal-leukemia-specific elution peak was detectable (Fig. 4 ). Figure 4. Open in new tabDownload slide Detection-limit testing. Leukemic DNA from the time of diagnosis was serially diluted in unrelated polyclonal peripheral blood lymphocyte DNA from five independent healthy donors. After heteroduplex induction, PCR products were subjected to the DHPLC analysis. DHPLC was performed at the optimized column oven temperature at 60 °C. The detection limit of the clonal product is 12.5% in a polyclonal background. Figure 4. Open in new tabDownload slide Detection-limit testing. Leukemic DNA from the time of diagnosis was serially diluted in unrelated polyclonal peripheral blood lymphocyte DNA from five independent healthy donors. After heteroduplex induction, PCR products were subjected to the DHPLC analysis. DHPLC was performed at the optimized column oven temperature at 60 °C. The detection limit of the clonal product is 12.5% in a polyclonal background. mixing experiments We tested the discrimination power of the method, using two targets with known sequences. We mixed different percentages of genomic DNA from two patient samples, amplified them by PCR, and analyzed the products at 50 and 60 °C. At 50 °C, all sample mixtures showed weak heteroduplex peaks during elution from the column. Nevertheless, only one main elution peak was observed with this condition. On the other hand, the biclonal products were separated at 60 °C with different admixtures of genomic DNA from each patient before PCR (Fig. 5 ). Figure 5. Open in new tabDownload slide Mixing experiments. DNA from two patient samples was mixed in different proportions, and PCR was performed using the VγI-Jγ1.3/2.3 primer pair. After heteroduplex induction, products were analyzed at 50 °C (native) and at 60 °C. At 50 °C, small heteroduplexes were induced in all mixed samples (upper panel, →). In DHPLC at optimized temperature and buffer gradient, all clonal and biclonal products were identified with the specific retention profile and unique retention times (lower panel, →). Figure 5. Open in new tabDownload slide Mixing experiments. DNA from two patient samples was mixed in different proportions, and PCR was performed using the VγI-Jγ1.3/2.3 primer pair. After heteroduplex induction, products were analyzed at 50 °C (native) and at 60 °C. At 50 °C, small heteroduplexes were induced in all mixed samples (upper panel, →). In DHPLC at optimized temperature and buffer gradient, all clonal and biclonal products were identified with the specific retention profile and unique retention times (lower panel, →). screening of initial patient samples DNA samples from 100 ALL samples at diagnosis [56 common ALL (C-ALL), 16 PRE-B-cell ALL (-B-ALL), 2 PRO-B-ALL, and 26 T-ALL] were analyzed for the presence of clonal rearrangements at the VγI-Jγ1.3/2.3 locus. In T-ALL, 24 of 26 samples showed clonal rearrangements, 7 of the 24 positive samples were rearranged on both alleles (29%). In 74 B-cell precursor ALL 36 clonal VγI-Jγ1.3/2.3 rearrangements were identified, 6 of these 36 patients samples were biclonal at this locus (17%). A more detailed description of the rearrangement pattern according to the immunologic subgroups is listed in Table 4 . Table 4. Clonality analysis of VγI-Jγ1.3/2.3 rearrangements in childhood precursor B-ALL and T-ALL. Subgroup . No. . Clonal . Biclonal . Polyclonal . All samples 100 47 13 40 C-ALL 56 23 4 29 PRE-B-ALL 16 7 2 7 PRO-B-ALL 2 2 T-ALL 26 17 7 2 Subgroup . No. . Clonal . Biclonal . Polyclonal . All samples 100 47 13 40 C-ALL 56 23 4 29 PRE-B-ALL 16 7 2 7 PRO-B-ALL 2 2 T-ALL 26 17 7 2 Open in new tab Table 4. Clonality analysis of VγI-Jγ1.3/2.3 rearrangements in childhood precursor B-ALL and T-ALL. Subgroup . No. . Clonal . Biclonal . Polyclonal . All samples 100 47 13 40 C-ALL 56 23 4 29 PRE-B-ALL 16 7 2 7 PRO-B-ALL 2 2 T-ALL 26 17 7 2 Subgroup . No. . Clonal . Biclonal . Polyclonal . All samples 100 47 13 40 C-ALL 56 23 4 29 PRE-B-ALL 16 7 2 7 PRO-B-ALL 2 2 T-ALL 26 17 7 2 Open in new tab Discussion The aim of this study was to establish a simplified method for discrimination between clonal and polyclonal TCR-γ rearrangements in newly diagnosed ALL. To achieve this, we examined the specificity and sensitivity of DHPLC for the detection of clonality in newly diagnosed ALLs. PCR amplification of antigen receptor rearrangements followed by agarose gel electrophoresis does not discriminate exactly between clonal and polyclonal cases. For an exact clonality analysis, further labor-intensive methods are often necessary. In addition, both alleles are often rearranged and not distinguishable by conventional length separation. Heteroduplex analysis followed by acrylamide gel electrophoresis separate monoclonal bands from polyclonal background smears (4). This method is time-consuming and has to be established for each laboratory. Here we demonstrate the potential of an automated HPLC-based system for clonality assessment in newly diagnosed leukemic samples with unknown rearrangements. Polyclonal samples together with known and unknown Vγ-Jγ rearrangements were tested for eligibility in routine diagnosis. For prediction of the optimal column oven temperature, we used a prediction software program (WAVEMAKER; Transgenomic) and tested it empirically. Fragments were run at multiple temperatures to determine the temperature at which the clonal or polyclonal targets were optimally separated. Both methods demonstrated similar results regarding the best column oven temperature. Whereas the software predicted 59.4 °C, the best practical temperature was 60 °C (20). After identification of optimal separation conditions (buffer gradient and column oven temperature), distinct mixed-clonal products of known length and known junctional regions were separated. We showed that this separation greatly depends on the heteroduplex conformation of the analyzed product and the applied column oven temperature conditions. Length independence is one of the major advantages of this type of separation. Targets with minor length differences but large junctional differences in homogeneity were separated with a high resolution (Fig. 1B ). Because a gaussian length distribution is often observed in samples with a high percentage of polyclonal lymphocytes, clonal sequences may be undetectable by length separation only. DHPLC analysis distinguishes mainly on the basis of the sequence pattern around the junctional region of each target. Therefore, even alleles with identical length can be discriminated from each other. In contrast, we always found a substantially reduced retention time of polyclonal sequences, ensuring a highly specific and reproducible separation of clonal and polyclonal targets. The found sensitivity is high enough for the identification of clonal targets in samples from the time of diagnosis. The evaluation of 100 initial leukemic samples gave the expected results compared with published frequencies of clonal TCR-γ rearrangements in B-cell precursor and T-ALL (21)(22). Because of a limited junctional region diversity in TCR-γ rearrangements, a high-resolution separation technique is needed to clearly identify clonality. Because nonpathologic complete VJ-rearrangements are often present in the peripheral blood, background amplification is a major pitfall during MRD target identification. These cases require cloning with subsequent sequencing, which makes the analysis extremely laborious and expensive, but often not successful for target identification. DHPLC is a practical approach for identification of clonality that uses the native PCR product and can be automated. After initial PCR, the time required for DHPLC analysis and target identification is ∼12 min. The method needs no further pre- or postanalytical hands-on time (e.g., gel preparation or gel staining). The results achieved are highly reproducible and interpretation is easy. The versatility of DHPLC analysis (14), the minimal preanalytical preparation time, and the minimal cost of approximately 1 Euro per sample makes the described principle applicable for routine molecular laboratories. Analysis conditions for these targets are now available for all users and can easily be adapted without major changes. In addition, other MRD targets can be established by applying the described optimization strategy. Because of the automated setup, results are highly reproducible even in different laboratories. This work was supported by the Fördergemeinschaft Kinderkrebszentrum Hamburg. 1 " These authors contributed equally to the work. 2 " Nonstandard abbreviations: ALL, acute lymphoblastic leukemia; MRD, minimal residual disease; DHPLC, denaturing HPLC; TEAA, triethylammonium acetate; T-ALL, T-cell ALL; and B-ALL, B-cell ALL. References 1 van Dongen JJ, Seriu T, Panzer-Grumayer ER, Biondi A, Pongers-Willemse MJ, Corral L, et al. Prognostic value of minimal residual disease in acute lymphoblastic leukaemia in childhood [see comments]. Lancet 1998 ; 352 : 1731 -1738. Crossref Search ADS PubMed 2 Cave H, van der Werff ten Bosch J, Suciu S, Guidal C, Waterkeyn C, Otten J, et al. Clinical significance of minimal residual disease in childhood acute lymphoblastic leukemia. European Organization for Research and Treatment of Cancer–Childhood Leukemia Cooperative Group [see comments]. N Engl J Med 1998 ; 339 : 591 -598. Crossref Search ADS PubMed 3 Pui CH, Campana D. New definition of remission in childhood acute lymphoblastic leukemia. Leukemia 2000 ; 14 : 783 -785. Crossref Search ADS PubMed 4 Langerak AW, Szczepanski T, van der Burg M, Wolvers-Tettero IL, van Dongen JJ. Heteroduplex PCR analysis of rearranged T cell receptor genes for clonality assessment in suspect T cell proliferations. Leukemia 1997 ; 11 : 2192 -2199. Crossref Search ADS PubMed 5 Delabesse E, Burtin ML, Millien C, Madonik A, Arnulf B, Beldjord K, et al. Rapid, multifluorescent TCRG Vγ and Jγ typing: application to T cell acute lymphoblastic leukemia and to the detection of minor clonal populations. Leukemia 2000 ; 14 : 1143 -1152. Crossref Search ADS PubMed 6 Kneba M, Bolz I, Linke B, Hiddemann W. Analysis of rearranged T-cell receptor β-chain genes by polymerase chain reaction (PCR) DNA sequencing and automated high resolution PCR fragment analysis. Blood 1995 ; 86 : 3930 -3937. Crossref Search ADS PubMed 7 Linke B, Pyttlich J, Tiemann M, Suttorp M, Parwaresch R, Hiddemann W, et al. Identification and structural analysis of rearranged immunoglobulin heavy chain genes in lymphomas and leukemias. Leukemia 1995 ; 9 : 840 -847. PubMed 8 Bottaro M, Berti E, Biondi A, Migone N, Crosti L. Heteroduplex analysis of T-cell receptor γ gene rearrangements for diagnosis and monitoring of cutaneous T-cell lymphomas. Blood 1994 ; 83 : 3271 -3278. Crossref Search ADS PubMed 9 Bourguin A, Tung R, Galili N, Sklar J. Rapid, nonradioactive detection of clonal T-cell receptor gene rearrangements in lymphoid neoplasms. Proc Natl Acad Sci U S A 1990 ; 87 : 8536 -8540. Crossref Search ADS PubMed 10 Davis TH, Yockey CE, Balk SP. Detection of clonal immunoglobulin gene rearrangements by polymerase chain reaction amplification and single-strand conformational polymorphism analysis [see comments]. Am J Pathol 1993 ; 142 : 1841 -1847. PubMed 11 Liu W, Smith DI, Rechtzigel KJ, Thibodeau SN, James CD. Denaturing high performance liquid chromatography (DHPLC) used in the detection of germline and somatic mutations. Nucleic Acids Res 1998 ; 26 : 1396 -1400. Crossref Search ADS PubMed 12 Liu WO, Oefner PJ, Qian C, Odom RS, Francke U. Denaturing HPLC-identified novel FBN1 mutations, polymorphisms, sequence variants in Marfan syndrome and related connective tissue disorders. Genet Test 1997 ; 1 : 237 -242. Crossref Search ADS PubMed 13 Arnold N, Gross E, Schwarz-Boeger U, Pfisterer J, Jonat W, Kiechle M. A highly sensitive, fast, and economical technique for mutation analysis in hereditary breast and ovarian cancers. Hum Mutat 1999 ; 14 : 333 -339. Crossref Search ADS PubMed 14 Xiao W, Oefner PJ. Denaturing high-performance liquid chromatography: a review. Hum Mutat 2001 ; 17 : 439 -474. Crossref Search ADS PubMed 15 Rischewski J, Schneppenheim R. Screening strategies for a highly polymorphic gene: DHPLC analysis of the Fanconi anemia group A gene. J Biochem Biophys Methods 2001 ; 47 : 53 -64. Crossref Search ADS PubMed 16 Cargill M, Altshuler D, Ireland J, Sklar P, Ardlie K, Patil N, et al. Characterization of single-nucleotide polymorphisms in coding regions of human genes. [published erratum appears in Nat Genet 1999;23:373] Nat Genet 1999 ; 22 : 231 -238. Crossref Search ADS PubMed 17 Huber CG. Micropellicular stationary phases for high-performance liquid chromatography of double-stranded DNA. J Chromatogr A 1998 ; 806 : 3 -30. Crossref Search ADS PubMed 18 Pongers-Willemse MJ, Seriu T, Stolz F, d’Aniello E, Gameiro P, Pisa P, et al. Primers and protocols for standardized detection of minimal residual disease in acute lymphoblastic leukemia using immunoglobulin and T cell receptor gene rearrangements and TAL1 deletions as PCR targets: report of the BIOMED-1 CONCERTED ACTION: investigation of minimal residual disease in acute leukemia. Leukemia 1999 ; 13 : 110 -118. PubMed 19 Deane M, Norton JD. Immunoglobulin heavy chain variable region family usage is independent of tumor cell phenotype in human B lineage leukemias. Eur J Immunol 1990 ; 20 : 2209 -2217. Crossref Search ADS PubMed 20 Jones AC, Austin J, Hansen N, Hoogendoorn B, Oefner PJ, Cheadle JP, et al. Optimal temperature selection for mutation detection by denaturing HPLC and comparison to single-stranded conformation polymorphism and heteroduplex analysis. Clin Chem 1999 ; 45 : 1133 -1140. Crossref Search ADS PubMed 21 Szczepanski T, Beishuizen A, Pongers-Willemse MJ, Hahlen K, Van Wering ER, Wijkhuijs AJ, et al. Cross-lineage T cell receptor gene rearrangements occur in more than ninety percent of childhood precursor-B acute lymphoblastic leukemias: alternative PCR targets for detection of minimal residual disease. Leukemia 1999 ; 13 : 196 -205. Crossref Search ADS PubMed 22 Szczepanski T, Langerak AW, Willemse MJ, Wolvers-Tettero IL, van Wering ER, van Dongen JJ. T cell receptor γ (TCRG) gene rearrangements in T cell acute lymphoblastic leukemia reflect ’end-stage’ recombinations: implications for minimal residual disease monitoring. Leukemia 2000 ; 14 : 1208 -1214. Crossref Search ADS PubMed © 2001 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)