Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

The Possibility of Detecting Individual Proteins in Blood Serum by Differentiation of Solubility Curves in Concentrated Sodium Sulfate Solutions *

The Possibility of Detecting Individual Proteins in Blood Serum by Differentiation of Solubility... THE POSSIBILITY OF DETECTING INDIVIDUAL PRO- TEINS IN BLOOD SERUM BY DIFFERENTIATION OF SOLUBILITY CURVES IN CONCENTRATED SODIUM SULFATE SOLUTIONS* C. L. H. MAJOOR During the latter part of the 19th century and in the early 20th much has been devoted to the behavior of the serum pentury study proteins. As a result, it has been found possible to separate the alibumin and some of the globulins of serum by means of a variety of concentrated salt solutions, as well as by dialysis and electrodial- ysis. this Yet, despite success, there remain sharp differences of opinion as to the number of individual be found in proteins to serum and no agreement has been reached on the question as whether to or not the various protein fractions should be considered as chemical entities. There has been, however, a general acceptance of the view of Haslaml' that only by fractional precipitation can the serum pro- teins be separated relatively completely. Since this method proved unsuited to clinical use, it was essential to search for a simpler pro- cedure and one which would but a involve single precipitation. At first the separation of albumin and globulins was generally effected by the use of a half-saturated solution of (NH4)2SO4, but in 1921 Howe12 published his method using a solution containing 215 gm. of anhydrous sodium sulfate per liter and this technique has been quite A universally adopted. great advantage of the method lies in the fact that it a permits direct determination of the nitrogen content of the filtrate by the Kjeldahl method. The concentration of 215 gm. per liter was chosen by Howe upon the basis of 6 solubility curves worked out with calf serum and plasma. In his graphs the values on the abscissa represent salt con- centrations and those on the ordinate indicate the quantity of protein N remaining in solution at the concentrations given. Howe believed From the "Binnen-Gasthuis" of of Amster- Department Medicine, University dam. The data here are derived from a presented Thesis offered to the Faculty of Medicine of the of Amsterdam University in 1942 in partial fulfilment of the the requirements for of Doctor of Medicine. is made degree Acknowledgment of the technical assistance of Miss W. J. Martens. YALE JOURNAL OF BIOLOGY AND MEDICINE that he noted that the quantity of protein remaining in solution with a concentration of 210 gm. of Na2SO4 per liter was not diminished when the concentration was raised to 220 gm. per liter. Therefore, he concluded that his curves had at one point flattened out, and that at this point there was a so-called "critical zone" (Fig. 1, I). But these critical zones have been found only in the 6 calf sera, and, ffi | . ,$q ,,, 1 L. PLA3S"A PMMIATE SOL| ^ n PkMff 1. of curves Howe and Butler and Montgomery (II). PIG. Types solubility by (I) by are not evident. The indeed, .convincingly acceptance they always that these zones exist-the of the method- of the idea quintessence for H~owe's me-thod has been almost uni- is although surprising, examination of human sera there has never -been used in the formly conclusive that critical zones exist. proo'f demonstrate them in curves of the same for Wu"8 failed to type abut Streef and showed experimentally serum, Streef-Spaan"3 sheep that the of -human oxalated ha-rdly changed solutbility plasma between 200 and 220 of liter. Their studies Na2SO4 gm. per gm. to constitute an affir-mation of the existence of the are too incomplete of Howe. Further of Howe's critical zones que-stion interpretation work of Butler and means is the Montgomery,3 who, by by provided SOLUBILITY CURVES OF SERUM PROTEINS 421 of a method similar to that of Howe, derived a different wholly type of solubility curve, lacking evidence of a true critical zone. As salt solution Butler and Montgomery employed a in mixture, equal parts, of KH2PO4 and K2HP04. Curves based their upon experi- mental findings show (Fig. 1, II) a well-defined and a angle steep of fall the curve at a concentration of molar 2.3 It is assumed PO4-. that the greater part of the gldbulin has been precipitated at this point and that a precipitation of albumin is to begin. Butler and Montgomery determined the albumin content of two sera by means of a single precipitation, employing their own method and also that of Howe. With their own technique values of 31.0 and 36.2 were found; with the Howe method corresponding values were 40.5 and 43.6 gm. of albumin per liter. The critical zone reported by Howe and the sharp angle found by Butler and must Montgomery therefore lie at different points on the solubility curve. In any event, the apparent uncertainty con- cerning the shape of the solubility curve of serum proteins in con- centrated salt solutions and the question as to what salt concentra- tion will most completely effect a separation of albumin and globu- lins indicate that further study is to be desired. Methods The sera examined were a those of number of healthy young and people those of a few patients. In order to eliminate individual differences in sera as far as possible, solubility curves were also de.termined upon the pooled sera of normal individuals. Various quantities of distilled water 1 were added to cc. of serum, after which a concentrated solution* of Na2SO4 was added up to a total volume of 30 cc. In this way Na2SO4 concentrations were obtained that varied from about 170 to 290 gm. per liter, and these were used throughout the greater part of the experiments (Figs. 2 and 3). In the last experiments (Figs. 4 and 5) the concentrations of Na2SO4 ranged from 100 to 400 gm. per liter. After standing in the incubator (310 to 370 C.) for about 12 hours the liquid above the precipitated globulin was almost clear and this was filtered -without shaking. During filtration the funnels were tightly covered with watchglasses. The results were not modified if shaking up the globulin precipi- tate were omitted, and a great advantage is gained in that the time of filtration * This had been with solution prepared It was freed chemically pure Na2SO4. was at of CO2 by means of and 7. The concentration deter- boiling, kept pH was mined and was controlled of Na+ gravimetrically by determination the content. The stock solution was the incubator in a kept in container. Pyrex glass 422 YALE JOURNAL OF BIOLOGY AND MEDICINE is considerably shortened and evaporation in the incubator is diminished.*- The 30 cc. dilutions of sera were prepared in duplicate, and filtration was done in duplicate. For each filtrate two nitrogen determinations were made, and the average of these values served for the calculation of the protein con- tent of the filtrate. In the graphs will be found two protein values given for each concentration of Na2SO4, each of the values having been obtained from the average of two nitrogen determinations of separate filtrates. The of the 4 nitrogen content filtrates, as expressed in graphs and 5, was ascertained by use of the unmodified iodometric method of Teorell,25 but for the greater of the work a of method portion simplification this was adopted.20 Both methods permit accurate work. The standard deviation calculated from 67 determinations of the nitrogen content, expressed as pro- tein, of a standard solution was 0.265 gm. per liter. The (NH4)2SO4 average of 2 determinations of the nitrogen content of a protein filtrate, calculated by the formula S.D.ii = ,/5 was 0.187 gm. per liter. These values were obtained with the original method of Teorell, while with our modified technique corresponding values of 0.298 and 0.211 gm. per liter were found (140 determinations). Results In Fig. 2 are charted the results obtained with three different sera. The first serum was from C., a 'healthy man, 25 years of age. The second was from A. M., a woman 52 years old, who suffered from mitral and aortic insufficiency, diabetes, and purulent otitis media. She had never had any considerable degree of fever and on the the serum was day obtained there was no marked cardiac diffi- had culty. She some albuminuria; this had been pronounced when she entered the hospital but had diminished to 1.5 gm. of protein per liter The (Esbach). albuminuria was due, in all probability, to an intercapillary glomerulosclerosis. The third subject was patient S., a male 50 years old, having a chronic nonspecific pneumonia with a right-sided empyema, bilateral pulmonary tuberculosis, and Laen- nec's cirrhosis. This patient died four days after the -blood was col- lected, and post-mortem examination confirmed the diagnosis. In work done in 1942 it was found20 that too values be secured if high may is not and and filtration quickly accomplished. Also, Robinson, Price, Hogden2l that when filter are used later Streef and Streef-Spaan23 and Harris9 noted soft papers adsorption of albumin is to be expected. In this work only hard paper (Schleicher and was as no takes Since with Schuill No. 575) used, practically adsorption place. hastened not such a paper filtration is slow, the process is materially by shaking up the globulin. SOLUBILITY CURVES OF SERUM PROTEINS 423 In Fig. 2, it will be noted, the curve has not been constructed through the points plotted. This was intentional, since the presenta- tion of a curve always tends to interfere with an unprejudiced evalua- tion of the results.* Our graphs show, in the first place, the great difference between the normal serum of C. and the two pathological sera. If these sera had been treated according to Howe with only 215 gm. of Na2SO4 per liter we would have found, as can easily be seen, albumin values of 56.1, 35.7, and 21.9 gm. per liter. It is generally agreed that a value of 21.9 indicates a markedly pathological condition. That an liter albumin content of 35.7 gm. per is also far below normal is per- haps not quite so evident, but the situation becomes very clear when we compare ithe curves for A. M., and C., for in serum C. we find that treatment with 277 gm. per liter leaves as much protein in solu- tion as is to be found in serum A. M. after treatment with 180 gm. of Na2SO4 per liter. In second the the place, regular configuration of the curves shows that method the salting out applied to protein determinations, can, if carefully done, give reliable results with but a small range of error. These results can be duplicated by others making use of the same technique and observing the same precautions. Moreover, Fig. 2 clearly shows that a small inaccuracy in the solutions causes Na2SO4 a in found. but very small change the protein value The line drawn in Fig. 3 shows that this difference amounts to about 1.8 gm. of protein per liter when the salt concentration is increased or decreased 10 liter. by gm. per In general, the three curves approximate a straight line and have about the same slope. But the critical zone of Howe at between 210 and 220 of liter cannot be As a gm. Na2SO4 per identified. matter of fact, the curves occasionally show horizontal portions, but they occur at various concentrations. Furthermore, curves derived from duplicate determinations made after an interval of a few days (published elsewhere) show these horizontal parts to have vanished or to have occurred in some other of the curve segment (Fig. 4). * made the of Butler In this connection, comment may be of work and Mont- mentioned above. of the values found them are gomery3 Many by experimentally a bit far from their drawn curves. This is located too arbitrarily solubility espe- true the which is constructed at 2.3 molar - -. It would cially of sharp angle, P04- which be the be no to draw other would in possible, doubt, curves, steeper neighbor- the said concentration. hood of 424 YALE JOURNAL OF BIOLOGY AND MEDICINE p-0ten $1UY:. 6C $ + 4. C. Serum 4: *4. + 5C Serum A.M. Mej. rPe -e W4 I. a )k Serum S. ,B * * s 22 23 24 25 26 17 W 19 20 21 h@CC sCfUIL KNSO0 SOLUBILITY CURVES OF SERUM PROTEINS 425 It is to be assumed, therefore, that the flat portions are artificial and are due proibably to small differences in evaporation during the filtration process in the incubator. In our opinion, it seems unlikely that the small, insignificant, horizontal parts in our curves, which do not always assume the same position, indicate the concentration at which precipitation of the albumin begins. Also, if in a mixture of is two substances the point approached at which one of them is beginning to precipitate, one would in a not, fact, expect minute critical zone such as that found but rather a sudden fall in by Howe, the curve which solubility pre- viously showed a horizontal portion or an evident inclination. Facts such as these, as well as our own experiments, lead us to reject wholly Howe's coinclusion. Though our solutbility curves resemble straight lines, it is theo- retically unlikely that the quantity of protein s remaining in solution is proportional to the salt concentration m. At least, Cohn4 has pointed out that not s, but log s is proportional to m, and he derived the following formula log s=cam+~ where a is a constant the determining slope of the line, and is also a constant which fixes the location of the line on the graph. This becomes dear if it is assumed that m=0, when s. P=log has pointed out that this formula also applies to solutions Sorensen22 of very pure serum In albumin. order to test my data by this for- mula the 11 values of solubility curves have been collected as in Fig. 3* It is possible to construct a regression equation, deriving a straight line corresponding as closely as possi-ble to the location of all points, by use of the simple formula derived from Hill."' The equation is s=90.5-1.78 m and the straight line is to be found in Fig. 3. Whereas the points are spread in the regularly neighborhood of the line at concentrations ranging from 170 to 220 gm. of Na2SO4 per liter, they are at con- * for the two curves of 2 all curves are derived from Except Fig. serum or mixtures from have serum healthy young people (details published elsewhere). bqen In all curves to the same order to bring was effected as level, adjustment follows: The salt concentration value of 230 liter is located at gm. per about the center of the The value of the content found at figure. protein this concentration of increased decreased Na2SO4 was or so units that it reached by many 50 to 51 gm. and all other of the same curve were corrected in per liter, points the same way. 426 YALE JOURNAL OF BIOLOGY AND MEDICINE centrations from 220 to 250 gm. per liter above the line without exception, and at a concentration of 260 gm. per liter and higher more and more points are to be found below the line, even far below. Powtg;A 6QC f"*A. 55.c $4, 0 0 , - 0 ? 40 " :0 0 to 0 *0 9 to Id# :, * . : 'Q * ; 3\O PoAM V. rug& 1- '( I' All .LaSM I,5OM4t 2.00 nWo Vseru. KIM . titm SC SOllJWce Fixu 3. a straight line does not provide an ideal expression of the Thus, of the Probably the ideal representation of the solu- location points. line is a curved one, which is slightly flattened out between bility and 250 of Na2SO4 per liter and then falls more rapidly. 220 gm. the line is not straight, there is no justification for Yet, although a to be present; in fact, as is evident from the assuming sharp angle curves of 2 and the in direction produce individual Figs. 3, changes curve. rather smooth a SOLUBILITY CURVES OF SERUM PROTEINS 427 Howe, himself noticed the steeper fall in the solubility curves at higher in concentrations of Na2SO4. He"3 drew attention, 1923, to the fact that the largest quantity of albumin precipitates between 1.75 molar and 2.25 molar Na2SO4 (=248.5 and 319.5 gm. per liter), but he seems to have been so impressed by the critical zone he had disclosed that he paid but to inclina- scant attention the changed tion of his lines. Working with sheep serum, Wu28 constructed curves through points obtained in observations similar to ours, and noted that the slope of the curves of became steeper at concentrations Na2SO4 higher than 1.75 molar. Between 2.00 and 2.25 molar this inclina- tion became very steep indeed. This change in direction was much less evident in his work with horse serum. Wu did not work with human serum. However, the value of his curve is limited since he studied only a few different salt concentrations. In the of graphs Geill7 also, the solubility curve is at distinctly steeper, not the classi- cal 50 per cent, but at a 67 per cent saturation of (NH4)2SO4. In order to permit a more complete understanding of the solu- of bility properties albumin rather large amounts of serum were obtained from 6 healthy young persons. These were examined at con- centrations of from of 106.6 to 386.6 gm. Na2SO4 per liter (0.75 2.75 to With the to molar). higher concentration, according Howe,13 the serum proteins should precipitate completely. With this material the solubility curve was determined twice with about a week's interval. In Fig. 4 the first curve is indicated by xx, with each x representing the protein content of an individual filtrate. Protein values were calculated from the average of two nitrogen determinations. The second in curve the figure is indicated by 4+. Examining the two curves in their entirety, in we see, the first place, that practically all of the serum proteins are involved. At about 400 gm. of Na2SO4 per liter the protein has been precipitated practically completely, while at 106.6 per liter gm. the protein con- tent of the filtrate agrees with the total protein content (74.1 gm. per liter). In the second place, it is clear that the points form two real curves and not straight lines. The second steep part of the curve begins at exactly the same point which, as was indicated in Fig. 3, formed the beginning of a curve. The steepest fall in the curves is found at 305 gm. of Na2SO4 per liter, that is, at a concentration of 2.1 moles of salt liter. This is per precisely the same concentration at which Howe observed a maximum precipitation of albumin 428 YALE JOURNAL OF BIOLOGY AND MEDICINE (between 1.75 and 2.25 moles). In our graphs no sharp angles can Je noted. Both curves form smooth lines, particularly the portions CT Lt. scram AgP sC : . To so atL so I* $w so ao go 2, N6Aoo.cc. Fma. . between 1.75 and 2.1 molar where the inclination becomes Na2SO4 this that it seems to much steeper. It is for reason necessary SOLUBILITY CURVES OF SERUM PROTEINS 429 reject curves showing a sharp angle as constructed by Butler and Montgomery.3 In accord with the method of these American investigators, the logarithms of the values found were and determined, these indi- cate the of in quantity protein remaining solution in the individual of salt. In concentrations the graphs these values are expressed on ordinate at the the right-hand side of the figure. To avoid confusion this has been done only in the case of the second curve which is repre- sented by * -. Only in the zone which is pertinent, that is, the region of the large drop, are some logarithm values of the first curve noted and these are indicated by Obviously the double S curve, as seen in the original curves, has become in flatter the logarithmic curves, but it can still be easily identified. In their work, Butler and Montgomery constructed straight lines through these logarithmic points, in accord with the formula of Cohn, and these lines, one for albumin and one for globulin, inter- sect forming an obtuse angle at about 2.3 moles of phosphate. These log lines, like the original solubility curves of these authors, are open to some criticism, for the points experimentally fixed deviate, even on the small scale used for the graphs, sometimes rather far from th-e st.raight lines. Our own work makes it evident that the log lines are approximately straight, but we must note that the word must approximate be used. The formula of Cohn does not mean more than this. Hence, we really are dealing with a slightly curved line, as is demonstrated at about 300 gm. of per liter 2.1 Na2SO4 (= moles), where straight lines ought to intersect but where a curve can be clearly observed instead. There is, however, no doulbt of the fact that our curves clearly show two steep portions and it is not too bold to assume that these steep portions are caused by the precipitation of a substance (or, at least the greater part of one) with definite solubility properties at pre- cisely these points on the curve. The best way to judge the correct- ness of this interpretation is to differentiate both solubility curves as to the of with quantity protein precipitated each change of concentra- tion (Ac= 13.3 gm. of Na2SO4 per liter). TIherefore, the differ- ences between the protein concentrations found at each successive salt concentration have been calculated. For example, in the first curve the difference in protein concentration between the filtrate at and that at 306.6 320 gm. of Na2SO4 per liter amounts to 12.8 gm. of protein per liter. These differences have been recorded on the 430 YALE JOURNAL OF BIOLOGY AND MEDICINE ordinate at a somewhat enlarged scale (AS). The result can be seen in the broken line at bottom of 4. The first the Fig. experiment is indicated by the dotted lines, the second by the solid lines. Now, if our concept is correct and if the solubility curves are to be con- sidered as double probability its differentiation to curves, ought approximate the double curve of From 4 we see that Quetelet. Fig. this distinct is almost the case, the broken line shows two peaks. The concentration high peak on the right (point A) occurs at a salt of about 305 gm. of Na2SO4 per liter. The lower peak occurs at 145 gm. per liter for the second curve for and at 152 gm. per liter the first one (point G). Hence it may be concluded upon the basis of solubility properties of serum proteins in concentrated solutions of Na2SO4 that the existence of at least two individual fractions is to be assumed. The first fraction, which ought to retain the name of euglobulin, has its precipitation maximum at 145 gm. of Na2SO4 per liter (=about 1 mole). Its precipitation begins at about 115 gm. per liter. As this portion of the differentiation curve seems to be rather symmetrical, it may be inferred that euglobulin is almost completely precipitated at 190 liter gm. per (point P1). The Na2SO4 concentration at which the second peak occurs in the curve indicates that we are dealing with the classic serum albu- min. Though the figure does not have sufficient accuracy to permit a final conclusion, it appears that this peak is not quite symmetrical. Such asymmetry could have been due to the precipitation of a third protein fraction along with the albumin at salt concentrations of from 250 to 305 gm. per liter. At all events one can assume that albumin precipitation does not begin at a concentration of less than 250 gm. per liter (point P2 on the figure). If there existed in serum only the two protein fractions, albumin and the euglobulin, solubility curve should run horizontally between 190 and 250 gm. of Na2SO4 per liter. That this is not the case is shown clearly in Fig. 4 and can also be seen in Figs. 2 and 3. Between these concentrations protein is precipitated also. Again, this becomes clear from the broken lines at the bottom of Fig. 4, and if at these concentrations no protein precipitation were taking the place lines ought to coincide with the zero line. Actually this is not the case. However, distinct and constantly occurring peaks are not to be seen. The peak between 240 and 253.3 gm. per liter of SOLUBILITY CURVES OF SERUM PROTEINS 431 the first curve (dotted line) is not to be seen distinctly in the second curve. The great variability at these concentrations has impressed us when dealing with other sera which have been examined twice Such variability can be explained in the first instance (Majoor20). by assuming that evaporation during the process of filtration in the incubator was not uniform on the two days when the serum was examined, and in the second place one must consider the possibility that the precipitation maximum of those proteins which precipitate between 190 and 250 gm. of Na2SO4 per liter may change slightly from one day to another. It is possible that a protein fraction which has a precipitation maximum of asbout 253 gm. per liter may, on one day, precipitate completely between 240 and 253.3 gm. per liter, whereas on another day its precipitation may spread over these and (Fig. 4). higher concentrations know that between 190 and 250 gm. of Na2SO4 At any rate, we per liter a third protein fraction is precipitated which, because of the absence of a distinct precipitation maximum in the curves, we shall continue to call pseudoglobulin. This consists perhaps of some smaller homogeneous fractions, possibly of an uninterrupted series of proteins each with somewhat greater and perhaps not quite con- solubility properties. stant area which a differentiation curve exhibits between two The special concentrations is a measure of the quantity of protein precipi- these concentrations. Thus, if the total protein con- tated between stant is known it is possible to calculate the albumin, the pseudoglob- ulin, and the euglobulin content. Such a calculation can be only correct because of the Ac's being chosen relatively approximately (= 13.3 gm. of Na2SO4 per liter) and the necessity of roughly estimating the shape of the right part of the euglobulin curves and the left portion of the albumin curve. The pseudoglobulin curve is subtracting the values for albumin and globulin from that derived by t!he total content. of protein in this fashion we find: Calculated - -) Second curve (44.oo. ) First curve (XX.00. 39.3 liter 39.1 gm. per liter gm. per Albumin 12.3 " " " 12.3 " " " Pseudoglobulin 20.8 " " " 21.9 " " " Euglobulin liter liter Total 72.4 73.3 gm. gm. per per 432 YALE JOURNAL OF BIOLOGY AND MEDICINE According to Howe one would have found, as can be deduced by interpolation: First curve Second curve Albumin 48.8 per liter 48.6 gm. gm. per liter Pseudoglobulin II (215 gm. of Na2SO4 per liter) 7.6 " " C 6.6 " c Pseudoglobulin I " " (170 gm. of " Na2SO4 per liter) 12.6 11.5 " Euglobulin (140 gm. of " " cc cc Na2SO4 per liter) 4.8 C 7.9 "c Total 73.8 gm. per liter liter 74.6 gm. per We have already concluded that the flat part of the curves which Howe believed he had found between of 210 and 220 gm. Na2SO4 per liter has no reality and that the certainly it does not indicate at which albumin point precipitation begins. From the data pre- sented above it becomes even more clear that the flat part Howe thought he had found at 140 gm. pier liter does not exist at all (at least not in the case of human serum). At exactly this concen-tration of Na2SO4 we find the maximum precipitation of euglobulin and if one should like to approximate' by means of a single precipitation the euglobulin as exactly as possible one should choose a concentra- tion of Na2SO4 of about 190 gm. per liter. In the case of our first curve the euglobulin content would then have been 20.2 gm. per liter, in our second curve the value would have been 21.5 gm. per liter. These values agree rather well with those calculated from the area of the in euglobulin peak Fig. 4. One may also note that when the and euglobulin pseudoglobulin I fractions, according to Howe, are counted the together true euglobulin content is also approximated. In and 1940 Balint Balintl stated that Howe's pseudoglobulin I was identical to his euglobulin in a conclusion based upon determina- tions of tyrosin, tryptophane, cystine, arginine, and histidine in serum protein fractions. Howe's two fractions are thus to be regarded as a single protein. The albumin content as calculated from the area of the albumin peak proves to be about 10 gm. per liter lower than is the value as determined to Howe. according Evidently this is true because at 215 of liter gm. Na2SO4 per there still remains a fairly large amount of in pseudoglobulin solution. If, however, the albumin content had been determined by a single salting out by means of a concentra- tion of Na2SO4 at which albumin is about to precipitate, for example, SOLUBILITY CURVES OF SERUM PROTEINS 433 about 260 gm. per liter (= 1.8 molar), the results would then have been 41.9 and 42.7 of albumin liter of serum and these gm. per values agree much better with those calculated from the area of the peak. to this It was obviously of importance subject interpretation of the protein solubility curve by determination of the protein values in a clinical case with a highly pathological serum. Consequently determinations were made on the blood of the following patient: L. O., male, 19 years of age. The patient had complained of weakness and emaciation for some months. There was no pruritus. The general examination was negative. Roentgen examination of the chest revealed a marked broadening of the hilar region bilater- ally. Urine examination was Von and Mantoux negative. Pirquet tests and I of The blood (0.1 mg. tuberculin) were negative. counts showed no abnormalities. The blood sedimentation test 1 10 mm. after 1 In (Westergren) was hour. the formol-gel reac- tion 1 there was positive flocculation after hour. The Mancke- Sommer in reaction showed positive flocculation 8 tubes (see de Vries27). Culture after gastric lavage showed no tubercle bacilli. probable diagnosis of lymphogranulomatosis benigna (Besnier- Boeck's disease) was made. The result of the examination of the serum proteins from this patient is presented in Fig. 5. The curve was differentiated in the same way as was the one offered in Fig. 4 and calculations of the surface area of the peaks showed: Albumin 23.3 gm. per liter Pseudoglobulin 21.4 gm. " " Euglobulin " " 43.3 gm. Total 88.1 gm. per liter By interpolation, the results according to Howe could be derived from this curve: Albumin 42.0 gm. per liter Pseudoglobulin II (215 gm. of Na2SO4 per liter) " 8.5 gm. Pseudoglobulin (170 gm. of "c Na2SO4 per liter) 18.2 gm. Euglobulin of "c (140 gm. per liter) 23.3 gm. Na2SO4 Total 92.0 gm. per liter 434 YALE JOURNAL OF BIOLOGY AND MEDICINE If salt concentrations had been as used recommended in the text in the discussion of the Fig. 4, results would have been: Albumin 31.3 liter gm. per Pseudoglobulin 16.4 gm. " Euglobulin 44.3 gm. " " Total 92.0 gm. per liter The difference in the results when the total protein content is calculated from the total area of individual peaks and when is it determined in the ordinary way, that is, from the difference between total N and residual N, is due to the fact that this particular serum did not show a precipitation of protein between and 366 340 gm. of Na2SO4 per liter, whereas between 366 and 380 liter gm. per a rather significant quantity of protein was precipitated. For this rea- son there appears a at separate peak the extreme right of the figure. In the calculations based on area of the the peaks this small fraction is disregarded. In the other calculations it is added to the albumin and the total protein values, which thus become a little too high. Upon the basis of this single observation it is not to possible conclude that substances of this type are frequently to be found in pathological sera. One would be inclined to that suppose precipitation of a poly- peptide-like substance would occur in this region. It is evident in this case also that the three fractions as calculated from the area of the peaks agree rather well with the results obtained from a single precipitation by solutions containing 190 and 260 gm. of Na2SO4 per liter. Of more importance is the fact that the albumin in peak this highly pathological serum is situated at exactly the same place as was the case in Fig. 4. The eugldbulin peak appears somewhat to the left, but the removal is so small that it is questionable if it is of any importance. Discussion Thus by means of differentiation curves of the serum proteins in concentrated Na2SO4 solutions, it is possible to show the existence of three individual protein fractions in serum, and the graphs permit us to determine approximately the concentrations of these fractions. A comparison of the above results with those of the two modern methods of protein study leads to surprising conclusions. Thanks to the work of Svedberg and his collaborators with the ultracentrifuge, 435 CURVES OF SERUM PROTEINS SOLUBILITY s Ls x 70 _ s0- 15 X ~~~~~+ 30 Xx X.0X KK S I; ZO 22 24 Z 2A 30 32 3A3A 36 40 i1 14 NLl3O, cc. gr/oo Fia. 5. 436 YALE OF JOURNAL BIOLOGY AND MEDICINE many new data about proteins have come to For clinical light. pur- poses this method is obviously too complicated. But few communi- cations have appeared in the literature dealing with the study of human sera pathological by means of the ultracentrifuge (MacFar- lane"9), yet it remarkable is to find the close resemblance of the graphs in our 4 and with the Figs. 5 ultracentrifugal photograms worked out by means of the scale method of Lamm (MacFarlane,"9 Important, too, is the resemblance of our curves to Svedberge4). the figures obtaiined by photographic registration of the electropho- retic examination in the Tiselius protein apparatus (Tiselius,2e 16 means of this Longsworth,"5 Luetscher,17 Gutman8). By method many pathological sera have been examined. As is shown by Fig. 6, derived from the work of Longsworth et there can be al.," dis- in normal serum the three tinguished glob- and from the ulins-a, apart A y-quite 3, high albumin peak A. The peak 4 is due to fibrinogen. From what has been stated above it may be asked if it is not permissible to assume that the fractions obtained by means of the three methods are really identical.* It is already known, for example, that the a and a globulins of Tiselius for the greater part possess a molecular weight as high as globulin, which in the ultracentrifuge 7 proves to be so In the very homogeneous. 04 globulin of the dia- peak ultracentrifugal 4XL therefore the dis- gram electroplhoretically tinguishable fractions a, ,3, and y are all present. Fia. 6. Electrophoretic pat- It is my opinion, however, that the tern of normal human plasma. (From: Longsworth, L. G., Th. agreement between precipitation diagrams Shedlovsky, and D. A. MeIn- nes: J. and those means of Exper. Med., obtained the electro 1939, 70, by 399.) phoretic technique is even greater, though I must admit that by means of electrophoresis the pseudoglobulin has been divided into three individual fractions, az, ,, and which in 2, * Referee's note: Recent data, undoubtedly unavailable to the author, indicate that this may be an over-simplification. Svensson, H.: J. Biol. Chem., 1941, 139, 805; Pedersen, K. O.: Ultracentrifugal Studies on Serum and Serum Fractions, Almqvist and Weksells, Uppsala, 1945. SOLUBILITY OF PROTEINS 437 CURVES SERUM my work first could not well be differentiated. Thus, the method is certainly more efficient but that both methods lead to almost identi- cal results can be assumed theoretically, I the believe, upon following basis, namely, that precipitation by means of concentrated salt solu- tions as well as by electrophoresis are methods in which the electrical charge of protein molecules plays a part. Quite aside from the resemblance of the two of types diagram, there is also the following argument which used to may be support the view stated above. In the electrophoretic diagrams globulin shows the highest and the most symmetrical globulin peak. Recently Gutman8 has reported that he has this be proved y globulin to identi- cal with the a globulin of Kendall.'4 The latter author succeeded in from human isolating serum globulin an distin- immunologically guishable protein which normally is present to a concentration of from 4 to 10 gm. per liter. The rest of the globulin behaves immu- nologically as though it were a mixture and has been In called x. the ultracentrifuge y globulin behaves like a rather homogeneous su.bstance. Thus, this protein can be characterized rather exactly in differ- ent ways. It is strange, as Gutman after points out, that precipita- tion by the method of Howe, it appeared in the euglobulin, then in the pseudoglobulin I fraction and that it sometimes partially precipi- tated with pseudoglobulin II. Since I have shown that euglobulin has its precipitation maximum at the concentration recommended by Howe for total precipitation and that total precipitation is not attained until about 190 gm. of Na2SO4 per liter is reached, it seems permissible to suggest that the euglobulin, as I use the term, the 'y globulin of Tiselius, and the a globulin of Kendall are identical. Pseudoglobulin, the protein which precipitates at from about 190 to 255 gm. of Na2SO4 per liter, is probably identical with the a, P, and 2 globulins of Tiselius and with the immunologically non-homo- geneous x globulin of Kendall. We may recall the conclusions of de Vries.27 He pointed out that positive reactions with the Takata Ara or Mancke-Sommer technique aside depend, from the necessarily low albumin content, upon an of increase euglobulin, but sometimes they depend exclu- sively upon an increase in pseudoglobulin I. This conclusion may not be final since is it based only upon the separation of fractions according to the method of Howe, and the statement should read that the reaction of Takata Ara or Mancke-Sommer is positive when 438 YALE JOURNAL OF BIOLOGY AND MEDICINE the actual euglobulin, that the is, y globulin of Tiselius and a globu- lin of Kendall is increased. This thesis is supported by the fact that it has already been shown that in cases of cirrhosis of the liver in which the Takata Ara reaction is generally positive there is an increase in both the y globulin of Tiselius and the a globulin of Kendall. My own observations on patient 0. afford an (Fig. 5) even for this stronger argument conclusion, for this serum which gave very strongly positive Mancke-Sommer reaction had greatly increased euglobulin with hardly any increase in pseudoglobulin. It is of some importance to note that in this highly pathological serum there was no indication a of quantitative in the change serum proteins. However, the possibility of such a in change some dis- eases should be given consideration, as has been shown recently by Wuhrmann and Leuthardt.29 These workers report solubility curves from human sera in phosphate solutions, applying the method of Butler and Montgomery.3 Though the Swiss autho;rs do not differ- entiate their curves, the precipitation maximum for albumin is apparent. For one normal and two pathological sera the maximum is located - - -. at about 2.1 molar P04 In the case of a patient with chronic nephrosis this is shifted considerably to the left. Since with the - - serum at the concentration of P04- the serum protein has been completely precipitated, the authors conclude that no albumin at all was present. A more logical conclusion, it seems to me, would be that albumin has been quantitatively changed. It has become less soluble in concentrated salt solutions and, consequently, the pre- cipitation maximum has shifted to the left. I have had as yet no to confirm opportunity this observation of Wuhrmann and Leuthardt. By means of other technical methods also, observations have been made which point out quantitative changes in the serum proteins in patients exhibiting a nephrotic syndrome. MacFarlane"9 noted that the urinary protein of such a patient shows in the ultracentrifuge a much more polydisperse pattern than does the serum albumin of healthy people. Bourdillon2 in 1939 measured the osmotic pressure of serum and urine albumin in normal persons and in patients with the nephrotic and syndrome, from his results he calculated the molec- ular weight. In the patients he arrived at figures considerably different from normal values. Leutscher"7 noted by means of the Tiselius technique that the serum alibumin of nephrotic subjects is of SOLUBILITY CURVES OF SERUM PROTEINS 439 abnormal composition. These observations make it that the urgent work of the Swiss authors be repeated. Attention has been called to the work of who already Haslam,10 thought that the various serum could be isolated means proteins by of fractional and others have also precipitation. S0rensen22 used this technique. because of seems to However, my results, it necessary question the of this method unless or utility special selected concen- trations of salt are used. that for Let us assume fractional precipi- tation the concentration of 215 gm. of per liter, which is Na2SO4 that of should used. Howe, be (This is about a half-saturated solu- tion and C. about at 32° 470 gm. of anhydrous can be dis- Na2SO4 solved in 1 if liter.) Even precipitation is repeated only a few of times a division the pseudogldbulin into two cannot be parts avoided. Thus the albumin filtrate will contain a always protein with globulin properties and will the globulin precipitate always be partly soluble in water. It is that probable the classic separation method by.means of (NH4)2SO4 will suffer from the same fault, inasmuch as Howe obtained approximately the same results use by of a salt solution containing 215 gm. of Na2SO4 per liter as when he employed half-saturated (NH4)2SO4 solution. Our study makes it appear most likely that serum proteins can be the more completely isolated by a single precipitation when the salt concentrations used are correctly chosen than can be the case by use of fractional precipitation by means of half-saturated (NH4)2SO4 or Na2SO4 solutions. Solutions containing 190 gm. per liter and 260 gm. per liter of serve best for Na2SO4 this purpose. Cohn and his collaborators5' arrived at the 6 same results when serum proteins were treated with (NH4)2SO4 solutions of exactly known concentra- tion and or with pH ethanol-water mixtures. When they chose a suitable concentration these workers succeeded in obtaining almost pure solutions of y, a, and d globulins and of albumin. The purity of the fractions obtained by precipitation was controlled by the Tiselius electrophoresis technique. Attention should be given to one more point'by Luetscher"7 who states, upon the basis of the of results his electrophoretic studies, that the results obtained by of use the new method approximate those of Howe if one is with almost dealing normal sera. In pathological sera the decrease of albumin the content is often much greater than is to be expected from tests made according to Howe. This fact is also shown in our Figs. 4 and 5. To illustrate this point I present values 440 YALE JOURNAL OF BIOLOGY AND MEDICINE albumin content of and of normal below expressing the pathological sera: Albumin content By preciritation According to Howe Calculated from with 260 gm. of (210 gm. of Na2SO4 surface area per liter per liter) Na2SO4 Nornal 39.3 39.1 41.9 42.7 48.8 48.6 Pathological 23.3 31.3 42.0 Howe was about 9.5 This table shows that the result according to serum it gm. per liter (24 per cent) too high, while in pathological was 18.7 per liter (80 per cent). Summwry 1. A number of sera were examined with Na2SO4 solutions of From the solubility curves based on widely varying concentrations. it these studies appeared: of from 210 to 220 gm. (a) That in the case of a concentration of Na2SO4 per liter there does not appear a constant "critical zone" in believed. the curve as Howe on That the concentration of 215 gm. per liter, which the (b) albumin and strength of Howe's work is generally used to separate globulin, lacks theoretical justification. (c) That there is no acute angle in the solubility curves of the such as was accepted by Butler and Montgomery. serum proteins That differentiating the solubility curves three distinctly (d) by the distinguishable serum proteins can be recognized, namely (i) albumin which about 305 of per largely precipitates at gm. Na2SO4 liter; about 145 (ii) the euglobulin which largely precipitates at of liter; (iii) the pseudoglobulin which does not gm. Na2SO4 per for for this have a clearly defined salt concentration precipitation, protein does not behave as a homogeneous substance. It may consist of some small fractions. (e) That it is possible to separate the three serum proteins fairly completely by using concentrations of Na2SO4 equal to 190 and liter. 260 gm. per in in is here is 2. Euglobulin, the sense which the word used, the most probably identical with electrophoretically distinguishable y globulin of Tiselius and with the immunologically characterized ac of globulin of Kendall. It is largely an augmentation in the amount this protein which renders positive the reactions of Mancke-Sommer and Takata Ara. SOLUBILITY CURVES OF SERUM PROTEINS 3. Serum which yields a strongly positive Mancke-Sommer reaction showed only marked quantitative changes in the pattern of the serum proteins. Qualitative changes could not be demonstrated. REFERENCES I Balint, P., and M. Balint: Biochem. Ztschr., 1940, 306, 296. 2 Bourdillon, J.: J. Exper. Med., 69, 819. 1939, 3 Butler, A. M., and H. Montgomery: J. Biol. Chem., 1932, 99, 173. Cohn, E. J.: Physiol. Rev., 1925, 5, 349. 5 Cohn, E. J., J. A. Luetscher, J. L. Oncley, S. H. Armstrong, and B. D. Davis: J. Am. Chem. Soc., 1940, 62, 3396. 6 Cohn, E. J., T. L. McMeekin, J. L. Oncley, J. M. Newell, and W. L. Hughes: J. Am. Chem. Soc., 1940, 62, 3386. 7 Geill, T.: Ztschr. f. klin. 334. Med., 1929, 110, 8 D. H. E. B. and Kabat: Gutman, A. B., Moore, Gutman, V. McClellan, E. A. J. Clin. Invest., 1941, 20, 765. 9 R. C.: Harris, J. Biol. Chem., 1939, 127, 751. Haslam, H. C.: Biochem. J., 1913, 7, 492. 11 Hill, A. Bradford: Principles of medical statistics. 1937. London, 12 Howe, P. E.: J. Biol. Chem., 1921, 49, 93. 13 Howe, P. E.: J. Biol. Chem., 1923, 57, 241. 14 Kendall, F. E.: J. Clin. Invest., 1937, 16, 921. 15 Longsworth, L. G., and D. A. McInnes: J. Exper. Med., 1940, 71, 77. 16 Longsworth, L. G., Th. Shedlovsky, and D. A. McInnes: J. Exper. Med., 1939, 70, 399. 17 A. Luetscher, J. Jr.: J. Clin. Invest., 1940, 19, 313. 18 Luetscher, J. A. Jr.: J. Clin. Invest., 1941, 20, 99. 19 MacFarlane, A. S.: Biochem. J., 1935, 29, 407, 660, 1175, 1202, 1209. 20 Over de Beteekenis van Majoor, C. L. H.: het Serumalbumine voor de beoordeeling en Behandeling van Inwendige en Chirurgische ziekten. Disseration, Amsterdam, 21 Robinson, H. W., J. W. Price, and C. G. Hogden: J. Biol. Chem., 1937, 120, 22 S. P. L.: Compt. du rend. d. trav. lab. Carlsberg, serie physiol., Sorensen, 1930-31, 18, No. 5. 23 Streef, G. H., and A. H. Streef-Spaan: Geneesk. tijdschr. v. Nederl.-Indie, 1938, 77, 2816. 24 Svedberg, Th.: Kolloid Ztschr., 1938, 85, 119. 25 T.: Acta med. 305. Teorell, Scandinav., 1928, 68, Tiselius, A.: Biochem. J., 1937, 31, 1464. 27 de Vries, A.: Acta med. Scandinav., 1938-39, 98, 95; 1939, 99, 425. 28 Wu, H.: Chinese J. Physiol., 125. 1933, 7, 29 Wuhrmann, F., and F. Leuthardt: Helvet. med. acta, 1941, 7, 565. The writer wishes to acknowledge his indebtedness to Prof. Dr. J. G. G. Borst, who directed this work, to Dr. J. Spaander for valuable suggestions, and to Capt.. E. R. Harvey, U.S.A., for indispensable help with translation and corrections. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Yale Journal of Biology and Medicine Pubmed Central

The Possibility of Detecting Individual Proteins in Blood Serum by Differentiation of Solubility Curves in Concentrated Sodium Sulfate Solutions *

Majoor, C. L. H.
The Yale Journal of Biology and Medicine , Volume 18 (5) – May 1, 1946
Download PDF
23 pages

Loading next page...
 
/lp/pubmed-central/the-possibility-of-detecting-individual-proteins-in-blood-serum-by-JhEcMKqXT2

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

ISSN
0044-0086
eISSN
1551-4056
Publisher site
See Article on Publisher Site

Abstract

THE POSSIBILITY OF DETECTING INDIVIDUAL PRO- TEINS IN BLOOD SERUM BY DIFFERENTIATION OF SOLUBILITY CURVES IN CONCENTRATED SODIUM SULFATE SOLUTIONS* C. L. H. MAJOOR During the latter part of the 19th century and in the early 20th much has been devoted to the behavior of the serum pentury study proteins. As a result, it has been found possible to separate the alibumin and some of the globulins of serum by means of a variety of concentrated salt solutions, as well as by dialysis and electrodial- ysis. this Yet, despite success, there remain sharp differences of opinion as to the number of individual be found in proteins to serum and no agreement has been reached on the question as whether to or not the various protein fractions should be considered as chemical entities. There has been, however, a general acceptance of the view of Haslaml' that only by fractional precipitation can the serum pro- teins be separated relatively completely. Since this method proved unsuited to clinical use, it was essential to search for a simpler pro- cedure and one which would but a involve single precipitation. At first the separation of albumin and globulins was generally effected by the use of a half-saturated solution of (NH4)2SO4, but in 1921 Howe12 published his method using a solution containing 215 gm. of anhydrous sodium sulfate per liter and this technique has been quite A universally adopted. great advantage of the method lies in the fact that it a permits direct determination of the nitrogen content of the filtrate by the Kjeldahl method. The concentration of 215 gm. per liter was chosen by Howe upon the basis of 6 solubility curves worked out with calf serum and plasma. In his graphs the values on the abscissa represent salt con- centrations and those on the ordinate indicate the quantity of protein N remaining in solution at the concentrations given. Howe believed From the "Binnen-Gasthuis" of of Amster- Department Medicine, University dam. The data here are derived from a presented Thesis offered to the Faculty of Medicine of the of Amsterdam University in 1942 in partial fulfilment of the the requirements for of Doctor of Medicine. is made degree Acknowledgment of the technical assistance of Miss W. J. Martens. YALE JOURNAL OF BIOLOGY AND MEDICINE that he noted that the quantity of protein remaining in solution with a concentration of 210 gm. of Na2SO4 per liter was not diminished when the concentration was raised to 220 gm. per liter. Therefore, he concluded that his curves had at one point flattened out, and that at this point there was a so-called "critical zone" (Fig. 1, I). But these critical zones have been found only in the 6 calf sera, and, ffi | . ,$q ,,, 1 L. PLA3S"A PMMIATE SOL| ^ n PkMff 1. of curves Howe and Butler and Montgomery (II). PIG. Types solubility by (I) by are not evident. The indeed, .convincingly acceptance they always that these zones exist-the of the method- of the idea quintessence for H~owe's me-thod has been almost uni- is although surprising, examination of human sera there has never -been used in the formly conclusive that critical zones exist. proo'f demonstrate them in curves of the same for Wu"8 failed to type abut Streef and showed experimentally serum, Streef-Spaan"3 sheep that the of -human oxalated ha-rdly changed solutbility plasma between 200 and 220 of liter. Their studies Na2SO4 gm. per gm. to constitute an affir-mation of the existence of the are too incomplete of Howe. Further of Howe's critical zones que-stion interpretation work of Butler and means is the Montgomery,3 who, by by provided SOLUBILITY CURVES OF SERUM PROTEINS 421 of a method similar to that of Howe, derived a different wholly type of solubility curve, lacking evidence of a true critical zone. As salt solution Butler and Montgomery employed a in mixture, equal parts, of KH2PO4 and K2HP04. Curves based their upon experi- mental findings show (Fig. 1, II) a well-defined and a angle steep of fall the curve at a concentration of molar 2.3 It is assumed PO4-. that the greater part of the gldbulin has been precipitated at this point and that a precipitation of albumin is to begin. Butler and Montgomery determined the albumin content of two sera by means of a single precipitation, employing their own method and also that of Howe. With their own technique values of 31.0 and 36.2 were found; with the Howe method corresponding values were 40.5 and 43.6 gm. of albumin per liter. The critical zone reported by Howe and the sharp angle found by Butler and must Montgomery therefore lie at different points on the solubility curve. In any event, the apparent uncertainty con- cerning the shape of the solubility curve of serum proteins in con- centrated salt solutions and the question as to what salt concentra- tion will most completely effect a separation of albumin and globu- lins indicate that further study is to be desired. Methods The sera examined were a those of number of healthy young and people those of a few patients. In order to eliminate individual differences in sera as far as possible, solubility curves were also de.termined upon the pooled sera of normal individuals. Various quantities of distilled water 1 were added to cc. of serum, after which a concentrated solution* of Na2SO4 was added up to a total volume of 30 cc. In this way Na2SO4 concentrations were obtained that varied from about 170 to 290 gm. per liter, and these were used throughout the greater part of the experiments (Figs. 2 and 3). In the last experiments (Figs. 4 and 5) the concentrations of Na2SO4 ranged from 100 to 400 gm. per liter. After standing in the incubator (310 to 370 C.) for about 12 hours the liquid above the precipitated globulin was almost clear and this was filtered -without shaking. During filtration the funnels were tightly covered with watchglasses. The results were not modified if shaking up the globulin precipi- tate were omitted, and a great advantage is gained in that the time of filtration * This had been with solution prepared It was freed chemically pure Na2SO4. was at of CO2 by means of and 7. The concentration deter- boiling, kept pH was mined and was controlled of Na+ gravimetrically by determination the content. The stock solution was the incubator in a kept in container. Pyrex glass 422 YALE JOURNAL OF BIOLOGY AND MEDICINE is considerably shortened and evaporation in the incubator is diminished.*- The 30 cc. dilutions of sera were prepared in duplicate, and filtration was done in duplicate. For each filtrate two nitrogen determinations were made, and the average of these values served for the calculation of the protein con- tent of the filtrate. In the graphs will be found two protein values given for each concentration of Na2SO4, each of the values having been obtained from the average of two nitrogen determinations of separate filtrates. The of the 4 nitrogen content filtrates, as expressed in graphs and 5, was ascertained by use of the unmodified iodometric method of Teorell,25 but for the greater of the work a of method portion simplification this was adopted.20 Both methods permit accurate work. The standard deviation calculated from 67 determinations of the nitrogen content, expressed as pro- tein, of a standard solution was 0.265 gm. per liter. The (NH4)2SO4 average of 2 determinations of the nitrogen content of a protein filtrate, calculated by the formula S.D.ii = ,/5 was 0.187 gm. per liter. These values were obtained with the original method of Teorell, while with our modified technique corresponding values of 0.298 and 0.211 gm. per liter were found (140 determinations). Results In Fig. 2 are charted the results obtained with three different sera. The first serum was from C., a 'healthy man, 25 years of age. The second was from A. M., a woman 52 years old, who suffered from mitral and aortic insufficiency, diabetes, and purulent otitis media. She had never had any considerable degree of fever and on the the serum was day obtained there was no marked cardiac diffi- had culty. She some albuminuria; this had been pronounced when she entered the hospital but had diminished to 1.5 gm. of protein per liter The (Esbach). albuminuria was due, in all probability, to an intercapillary glomerulosclerosis. The third subject was patient S., a male 50 years old, having a chronic nonspecific pneumonia with a right-sided empyema, bilateral pulmonary tuberculosis, and Laen- nec's cirrhosis. This patient died four days after the -blood was col- lected, and post-mortem examination confirmed the diagnosis. In work done in 1942 it was found20 that too values be secured if high may is not and and filtration quickly accomplished. Also, Robinson, Price, Hogden2l that when filter are used later Streef and Streef-Spaan23 and Harris9 noted soft papers adsorption of albumin is to be expected. In this work only hard paper (Schleicher and was as no takes Since with Schuill No. 575) used, practically adsorption place. hastened not such a paper filtration is slow, the process is materially by shaking up the globulin. SOLUBILITY CURVES OF SERUM PROTEINS 423 In Fig. 2, it will be noted, the curve has not been constructed through the points plotted. This was intentional, since the presenta- tion of a curve always tends to interfere with an unprejudiced evalua- tion of the results.* Our graphs show, in the first place, the great difference between the normal serum of C. and the two pathological sera. If these sera had been treated according to Howe with only 215 gm. of Na2SO4 per liter we would have found, as can easily be seen, albumin values of 56.1, 35.7, and 21.9 gm. per liter. It is generally agreed that a value of 21.9 indicates a markedly pathological condition. That an liter albumin content of 35.7 gm. per is also far below normal is per- haps not quite so evident, but the situation becomes very clear when we compare ithe curves for A. M., and C., for in serum C. we find that treatment with 277 gm. per liter leaves as much protein in solu- tion as is to be found in serum A. M. after treatment with 180 gm. of Na2SO4 per liter. In second the the place, regular configuration of the curves shows that method the salting out applied to protein determinations, can, if carefully done, give reliable results with but a small range of error. These results can be duplicated by others making use of the same technique and observing the same precautions. Moreover, Fig. 2 clearly shows that a small inaccuracy in the solutions causes Na2SO4 a in found. but very small change the protein value The line drawn in Fig. 3 shows that this difference amounts to about 1.8 gm. of protein per liter when the salt concentration is increased or decreased 10 liter. by gm. per In general, the three curves approximate a straight line and have about the same slope. But the critical zone of Howe at between 210 and 220 of liter cannot be As a gm. Na2SO4 per identified. matter of fact, the curves occasionally show horizontal portions, but they occur at various concentrations. Furthermore, curves derived from duplicate determinations made after an interval of a few days (published elsewhere) show these horizontal parts to have vanished or to have occurred in some other of the curve segment (Fig. 4). * made the of Butler In this connection, comment may be of work and Mont- mentioned above. of the values found them are gomery3 Many by experimentally a bit far from their drawn curves. This is located too arbitrarily solubility espe- true the which is constructed at 2.3 molar - -. It would cially of sharp angle, P04- which be the be no to draw other would in possible, doubt, curves, steeper neighbor- the said concentration. hood of 424 YALE JOURNAL OF BIOLOGY AND MEDICINE p-0ten $1UY:. 6C $ + 4. C. Serum 4: *4. + 5C Serum A.M. Mej. rPe -e W4 I. a )k Serum S. ,B * * s 22 23 24 25 26 17 W 19 20 21 h@CC sCfUIL KNSO0 SOLUBILITY CURVES OF SERUM PROTEINS 425 It is to be assumed, therefore, that the flat portions are artificial and are due proibably to small differences in evaporation during the filtration process in the incubator. In our opinion, it seems unlikely that the small, insignificant, horizontal parts in our curves, which do not always assume the same position, indicate the concentration at which precipitation of the albumin begins. Also, if in a mixture of is two substances the point approached at which one of them is beginning to precipitate, one would in a not, fact, expect minute critical zone such as that found but rather a sudden fall in by Howe, the curve which solubility pre- viously showed a horizontal portion or an evident inclination. Facts such as these, as well as our own experiments, lead us to reject wholly Howe's coinclusion. Though our solutbility curves resemble straight lines, it is theo- retically unlikely that the quantity of protein s remaining in solution is proportional to the salt concentration m. At least, Cohn4 has pointed out that not s, but log s is proportional to m, and he derived the following formula log s=cam+~ where a is a constant the determining slope of the line, and is also a constant which fixes the location of the line on the graph. This becomes dear if it is assumed that m=0, when s. P=log has pointed out that this formula also applies to solutions Sorensen22 of very pure serum In albumin. order to test my data by this for- mula the 11 values of solubility curves have been collected as in Fig. 3* It is possible to construct a regression equation, deriving a straight line corresponding as closely as possi-ble to the location of all points, by use of the simple formula derived from Hill."' The equation is s=90.5-1.78 m and the straight line is to be found in Fig. 3. Whereas the points are spread in the regularly neighborhood of the line at concentrations ranging from 170 to 220 gm. of Na2SO4 per liter, they are at con- * for the two curves of 2 all curves are derived from Except Fig. serum or mixtures from have serum healthy young people (details published elsewhere). bqen In all curves to the same order to bring was effected as level, adjustment follows: The salt concentration value of 230 liter is located at gm. per about the center of the The value of the content found at figure. protein this concentration of increased decreased Na2SO4 was or so units that it reached by many 50 to 51 gm. and all other of the same curve were corrected in per liter, points the same way. 426 YALE JOURNAL OF BIOLOGY AND MEDICINE centrations from 220 to 250 gm. per liter above the line without exception, and at a concentration of 260 gm. per liter and higher more and more points are to be found below the line, even far below. Powtg;A 6QC f"*A. 55.c $4, 0 0 , - 0 ? 40 " :0 0 to 0 *0 9 to Id# :, * . : 'Q * ; 3\O PoAM V. rug& 1- '( I' All .LaSM I,5OM4t 2.00 nWo Vseru. KIM . titm SC SOllJWce Fixu 3. a straight line does not provide an ideal expression of the Thus, of the Probably the ideal representation of the solu- location points. line is a curved one, which is slightly flattened out between bility and 250 of Na2SO4 per liter and then falls more rapidly. 220 gm. the line is not straight, there is no justification for Yet, although a to be present; in fact, as is evident from the assuming sharp angle curves of 2 and the in direction produce individual Figs. 3, changes curve. rather smooth a SOLUBILITY CURVES OF SERUM PROTEINS 427 Howe, himself noticed the steeper fall in the solubility curves at higher in concentrations of Na2SO4. He"3 drew attention, 1923, to the fact that the largest quantity of albumin precipitates between 1.75 molar and 2.25 molar Na2SO4 (=248.5 and 319.5 gm. per liter), but he seems to have been so impressed by the critical zone he had disclosed that he paid but to inclina- scant attention the changed tion of his lines. Working with sheep serum, Wu28 constructed curves through points obtained in observations similar to ours, and noted that the slope of the curves of became steeper at concentrations Na2SO4 higher than 1.75 molar. Between 2.00 and 2.25 molar this inclina- tion became very steep indeed. This change in direction was much less evident in his work with horse serum. Wu did not work with human serum. However, the value of his curve is limited since he studied only a few different salt concentrations. In the of graphs Geill7 also, the solubility curve is at distinctly steeper, not the classi- cal 50 per cent, but at a 67 per cent saturation of (NH4)2SO4. In order to permit a more complete understanding of the solu- of bility properties albumin rather large amounts of serum were obtained from 6 healthy young persons. These were examined at con- centrations of from of 106.6 to 386.6 gm. Na2SO4 per liter (0.75 2.75 to With the to molar). higher concentration, according Howe,13 the serum proteins should precipitate completely. With this material the solubility curve was determined twice with about a week's interval. In Fig. 4 the first curve is indicated by xx, with each x representing the protein content of an individual filtrate. Protein values were calculated from the average of two nitrogen determinations. The second in curve the figure is indicated by 4+. Examining the two curves in their entirety, in we see, the first place, that practically all of the serum proteins are involved. At about 400 gm. of Na2SO4 per liter the protein has been precipitated practically completely, while at 106.6 per liter gm. the protein con- tent of the filtrate agrees with the total protein content (74.1 gm. per liter). In the second place, it is clear that the points form two real curves and not straight lines. The second steep part of the curve begins at exactly the same point which, as was indicated in Fig. 3, formed the beginning of a curve. The steepest fall in the curves is found at 305 gm. of Na2SO4 per liter, that is, at a concentration of 2.1 moles of salt liter. This is per precisely the same concentration at which Howe observed a maximum precipitation of albumin 428 YALE JOURNAL OF BIOLOGY AND MEDICINE (between 1.75 and 2.25 moles). In our graphs no sharp angles can Je noted. Both curves form smooth lines, particularly the portions CT Lt. scram AgP sC : . To so atL so I* $w so ao go 2, N6Aoo.cc. Fma. . between 1.75 and 2.1 molar where the inclination becomes Na2SO4 this that it seems to much steeper. It is for reason necessary SOLUBILITY CURVES OF SERUM PROTEINS 429 reject curves showing a sharp angle as constructed by Butler and Montgomery.3 In accord with the method of these American investigators, the logarithms of the values found were and determined, these indi- cate the of in quantity protein remaining solution in the individual of salt. In concentrations the graphs these values are expressed on ordinate at the the right-hand side of the figure. To avoid confusion this has been done only in the case of the second curve which is repre- sented by * -. Only in the zone which is pertinent, that is, the region of the large drop, are some logarithm values of the first curve noted and these are indicated by Obviously the double S curve, as seen in the original curves, has become in flatter the logarithmic curves, but it can still be easily identified. In their work, Butler and Montgomery constructed straight lines through these logarithmic points, in accord with the formula of Cohn, and these lines, one for albumin and one for globulin, inter- sect forming an obtuse angle at about 2.3 moles of phosphate. These log lines, like the original solubility curves of these authors, are open to some criticism, for the points experimentally fixed deviate, even on the small scale used for the graphs, sometimes rather far from th-e st.raight lines. Our own work makes it evident that the log lines are approximately straight, but we must note that the word must approximate be used. The formula of Cohn does not mean more than this. Hence, we really are dealing with a slightly curved line, as is demonstrated at about 300 gm. of per liter 2.1 Na2SO4 (= moles), where straight lines ought to intersect but where a curve can be clearly observed instead. There is, however, no doulbt of the fact that our curves clearly show two steep portions and it is not too bold to assume that these steep portions are caused by the precipitation of a substance (or, at least the greater part of one) with definite solubility properties at pre- cisely these points on the curve. The best way to judge the correct- ness of this interpretation is to differentiate both solubility curves as to the of with quantity protein precipitated each change of concentra- tion (Ac= 13.3 gm. of Na2SO4 per liter). TIherefore, the differ- ences between the protein concentrations found at each successive salt concentration have been calculated. For example, in the first curve the difference in protein concentration between the filtrate at and that at 306.6 320 gm. of Na2SO4 per liter amounts to 12.8 gm. of protein per liter. These differences have been recorded on the 430 YALE JOURNAL OF BIOLOGY AND MEDICINE ordinate at a somewhat enlarged scale (AS). The result can be seen in the broken line at bottom of 4. The first the Fig. experiment is indicated by the dotted lines, the second by the solid lines. Now, if our concept is correct and if the solubility curves are to be con- sidered as double probability its differentiation to curves, ought approximate the double curve of From 4 we see that Quetelet. Fig. this distinct is almost the case, the broken line shows two peaks. The concentration high peak on the right (point A) occurs at a salt of about 305 gm. of Na2SO4 per liter. The lower peak occurs at 145 gm. per liter for the second curve for and at 152 gm. per liter the first one (point G). Hence it may be concluded upon the basis of solubility properties of serum proteins in concentrated solutions of Na2SO4 that the existence of at least two individual fractions is to be assumed. The first fraction, which ought to retain the name of euglobulin, has its precipitation maximum at 145 gm. of Na2SO4 per liter (=about 1 mole). Its precipitation begins at about 115 gm. per liter. As this portion of the differentiation curve seems to be rather symmetrical, it may be inferred that euglobulin is almost completely precipitated at 190 liter gm. per (point P1). The Na2SO4 concentration at which the second peak occurs in the curve indicates that we are dealing with the classic serum albu- min. Though the figure does not have sufficient accuracy to permit a final conclusion, it appears that this peak is not quite symmetrical. Such asymmetry could have been due to the precipitation of a third protein fraction along with the albumin at salt concentrations of from 250 to 305 gm. per liter. At all events one can assume that albumin precipitation does not begin at a concentration of less than 250 gm. per liter (point P2 on the figure). If there existed in serum only the two protein fractions, albumin and the euglobulin, solubility curve should run horizontally between 190 and 250 gm. of Na2SO4 per liter. That this is not the case is shown clearly in Fig. 4 and can also be seen in Figs. 2 and 3. Between these concentrations protein is precipitated also. Again, this becomes clear from the broken lines at the bottom of Fig. 4, and if at these concentrations no protein precipitation were taking the place lines ought to coincide with the zero line. Actually this is not the case. However, distinct and constantly occurring peaks are not to be seen. The peak between 240 and 253.3 gm. per liter of SOLUBILITY CURVES OF SERUM PROTEINS 431 the first curve (dotted line) is not to be seen distinctly in the second curve. The great variability at these concentrations has impressed us when dealing with other sera which have been examined twice Such variability can be explained in the first instance (Majoor20). by assuming that evaporation during the process of filtration in the incubator was not uniform on the two days when the serum was examined, and in the second place one must consider the possibility that the precipitation maximum of those proteins which precipitate between 190 and 250 gm. of Na2SO4 per liter may change slightly from one day to another. It is possible that a protein fraction which has a precipitation maximum of asbout 253 gm. per liter may, on one day, precipitate completely between 240 and 253.3 gm. per liter, whereas on another day its precipitation may spread over these and (Fig. 4). higher concentrations know that between 190 and 250 gm. of Na2SO4 At any rate, we per liter a third protein fraction is precipitated which, because of the absence of a distinct precipitation maximum in the curves, we shall continue to call pseudoglobulin. This consists perhaps of some smaller homogeneous fractions, possibly of an uninterrupted series of proteins each with somewhat greater and perhaps not quite con- solubility properties. stant area which a differentiation curve exhibits between two The special concentrations is a measure of the quantity of protein precipi- these concentrations. Thus, if the total protein con- tated between stant is known it is possible to calculate the albumin, the pseudoglob- ulin, and the euglobulin content. Such a calculation can be only correct because of the Ac's being chosen relatively approximately (= 13.3 gm. of Na2SO4 per liter) and the necessity of roughly estimating the shape of the right part of the euglobulin curves and the left portion of the albumin curve. The pseudoglobulin curve is subtracting the values for albumin and globulin from that derived by t!he total content. of protein in this fashion we find: Calculated - -) Second curve (44.oo. ) First curve (XX.00. 39.3 liter 39.1 gm. per liter gm. per Albumin 12.3 " " " 12.3 " " " Pseudoglobulin 20.8 " " " 21.9 " " " Euglobulin liter liter Total 72.4 73.3 gm. gm. per per 432 YALE JOURNAL OF BIOLOGY AND MEDICINE According to Howe one would have found, as can be deduced by interpolation: First curve Second curve Albumin 48.8 per liter 48.6 gm. gm. per liter Pseudoglobulin II (215 gm. of Na2SO4 per liter) 7.6 " " C 6.6 " c Pseudoglobulin I " " (170 gm. of " Na2SO4 per liter) 12.6 11.5 " Euglobulin (140 gm. of " " cc cc Na2SO4 per liter) 4.8 C 7.9 "c Total 73.8 gm. per liter liter 74.6 gm. per We have already concluded that the flat part of the curves which Howe believed he had found between of 210 and 220 gm. Na2SO4 per liter has no reality and that the certainly it does not indicate at which albumin point precipitation begins. From the data pre- sented above it becomes even more clear that the flat part Howe thought he had found at 140 gm. pier liter does not exist at all (at least not in the case of human serum). At exactly this concen-tration of Na2SO4 we find the maximum precipitation of euglobulin and if one should like to approximate' by means of a single precipitation the euglobulin as exactly as possible one should choose a concentra- tion of Na2SO4 of about 190 gm. per liter. In the case of our first curve the euglobulin content would then have been 20.2 gm. per liter, in our second curve the value would have been 21.5 gm. per liter. These values agree rather well with those calculated from the area of the in euglobulin peak Fig. 4. One may also note that when the and euglobulin pseudoglobulin I fractions, according to Howe, are counted the together true euglobulin content is also approximated. In and 1940 Balint Balintl stated that Howe's pseudoglobulin I was identical to his euglobulin in a conclusion based upon determina- tions of tyrosin, tryptophane, cystine, arginine, and histidine in serum protein fractions. Howe's two fractions are thus to be regarded as a single protein. The albumin content as calculated from the area of the albumin peak proves to be about 10 gm. per liter lower than is the value as determined to Howe. according Evidently this is true because at 215 of liter gm. Na2SO4 per there still remains a fairly large amount of in pseudoglobulin solution. If, however, the albumin content had been determined by a single salting out by means of a concentra- tion of Na2SO4 at which albumin is about to precipitate, for example, SOLUBILITY CURVES OF SERUM PROTEINS 433 about 260 gm. per liter (= 1.8 molar), the results would then have been 41.9 and 42.7 of albumin liter of serum and these gm. per values agree much better with those calculated from the area of the peak. to this It was obviously of importance subject interpretation of the protein solubility curve by determination of the protein values in a clinical case with a highly pathological serum. Consequently determinations were made on the blood of the following patient: L. O., male, 19 years of age. The patient had complained of weakness and emaciation for some months. There was no pruritus. The general examination was negative. Roentgen examination of the chest revealed a marked broadening of the hilar region bilater- ally. Urine examination was Von and Mantoux negative. Pirquet tests and I of The blood (0.1 mg. tuberculin) were negative. counts showed no abnormalities. The blood sedimentation test 1 10 mm. after 1 In (Westergren) was hour. the formol-gel reac- tion 1 there was positive flocculation after hour. The Mancke- Sommer in reaction showed positive flocculation 8 tubes (see de Vries27). Culture after gastric lavage showed no tubercle bacilli. probable diagnosis of lymphogranulomatosis benigna (Besnier- Boeck's disease) was made. The result of the examination of the serum proteins from this patient is presented in Fig. 5. The curve was differentiated in the same way as was the one offered in Fig. 4 and calculations of the surface area of the peaks showed: Albumin 23.3 gm. per liter Pseudoglobulin 21.4 gm. " " Euglobulin " " 43.3 gm. Total 88.1 gm. per liter By interpolation, the results according to Howe could be derived from this curve: Albumin 42.0 gm. per liter Pseudoglobulin II (215 gm. of Na2SO4 per liter) " 8.5 gm. Pseudoglobulin (170 gm. of "c Na2SO4 per liter) 18.2 gm. Euglobulin of "c (140 gm. per liter) 23.3 gm. Na2SO4 Total 92.0 gm. per liter 434 YALE JOURNAL OF BIOLOGY AND MEDICINE If salt concentrations had been as used recommended in the text in the discussion of the Fig. 4, results would have been: Albumin 31.3 liter gm. per Pseudoglobulin 16.4 gm. " Euglobulin 44.3 gm. " " Total 92.0 gm. per liter The difference in the results when the total protein content is calculated from the total area of individual peaks and when is it determined in the ordinary way, that is, from the difference between total N and residual N, is due to the fact that this particular serum did not show a precipitation of protein between and 366 340 gm. of Na2SO4 per liter, whereas between 366 and 380 liter gm. per a rather significant quantity of protein was precipitated. For this rea- son there appears a at separate peak the extreme right of the figure. In the calculations based on area of the the peaks this small fraction is disregarded. In the other calculations it is added to the albumin and the total protein values, which thus become a little too high. Upon the basis of this single observation it is not to possible conclude that substances of this type are frequently to be found in pathological sera. One would be inclined to that suppose precipitation of a poly- peptide-like substance would occur in this region. It is evident in this case also that the three fractions as calculated from the area of the peaks agree rather well with the results obtained from a single precipitation by solutions containing 190 and 260 gm. of Na2SO4 per liter. Of more importance is the fact that the albumin in peak this highly pathological serum is situated at exactly the same place as was the case in Fig. 4. The eugldbulin peak appears somewhat to the left, but the removal is so small that it is questionable if it is of any importance. Discussion Thus by means of differentiation curves of the serum proteins in concentrated Na2SO4 solutions, it is possible to show the existence of three individual protein fractions in serum, and the graphs permit us to determine approximately the concentrations of these fractions. A comparison of the above results with those of the two modern methods of protein study leads to surprising conclusions. Thanks to the work of Svedberg and his collaborators with the ultracentrifuge, 435 CURVES OF SERUM PROTEINS SOLUBILITY s Ls x 70 _ s0- 15 X ~~~~~+ 30 Xx X.0X KK S I; ZO 22 24 Z 2A 30 32 3A3A 36 40 i1 14 NLl3O, cc. gr/oo Fia. 5. 436 YALE OF JOURNAL BIOLOGY AND MEDICINE many new data about proteins have come to For clinical light. pur- poses this method is obviously too complicated. But few communi- cations have appeared in the literature dealing with the study of human sera pathological by means of the ultracentrifuge (MacFar- lane"9), yet it remarkable is to find the close resemblance of the graphs in our 4 and with the Figs. 5 ultracentrifugal photograms worked out by means of the scale method of Lamm (MacFarlane,"9 Important, too, is the resemblance of our curves to Svedberge4). the figures obtaiined by photographic registration of the electropho- retic examination in the Tiselius protein apparatus (Tiselius,2e 16 means of this Longsworth,"5 Luetscher,17 Gutman8). By method many pathological sera have been examined. As is shown by Fig. 6, derived from the work of Longsworth et there can be al.," dis- in normal serum the three tinguished glob- and from the ulins-a, apart A y-quite 3, high albumin peak A. The peak 4 is due to fibrinogen. From what has been stated above it may be asked if it is not permissible to assume that the fractions obtained by means of the three methods are really identical.* It is already known, for example, that the a and a globulins of Tiselius for the greater part possess a molecular weight as high as globulin, which in the ultracentrifuge 7 proves to be so In the very homogeneous. 04 globulin of the dia- peak ultracentrifugal 4XL therefore the dis- gram electroplhoretically tinguishable fractions a, ,3, and y are all present. Fia. 6. Electrophoretic pat- It is my opinion, however, that the tern of normal human plasma. (From: Longsworth, L. G., Th. agreement between precipitation diagrams Shedlovsky, and D. A. MeIn- nes: J. and those means of Exper. Med., obtained the electro 1939, 70, by 399.) phoretic technique is even greater, though I must admit that by means of electrophoresis the pseudoglobulin has been divided into three individual fractions, az, ,, and which in 2, * Referee's note: Recent data, undoubtedly unavailable to the author, indicate that this may be an over-simplification. Svensson, H.: J. Biol. Chem., 1941, 139, 805; Pedersen, K. O.: Ultracentrifugal Studies on Serum and Serum Fractions, Almqvist and Weksells, Uppsala, 1945. SOLUBILITY OF PROTEINS 437 CURVES SERUM my work first could not well be differentiated. Thus, the method is certainly more efficient but that both methods lead to almost identi- cal results can be assumed theoretically, I the believe, upon following basis, namely, that precipitation by means of concentrated salt solu- tions as well as by electrophoresis are methods in which the electrical charge of protein molecules plays a part. Quite aside from the resemblance of the two of types diagram, there is also the following argument which used to may be support the view stated above. In the electrophoretic diagrams globulin shows the highest and the most symmetrical globulin peak. Recently Gutman8 has reported that he has this be proved y globulin to identi- cal with the a globulin of Kendall.'4 The latter author succeeded in from human isolating serum globulin an distin- immunologically guishable protein which normally is present to a concentration of from 4 to 10 gm. per liter. The rest of the globulin behaves immu- nologically as though it were a mixture and has been In called x. the ultracentrifuge y globulin behaves like a rather homogeneous su.bstance. Thus, this protein can be characterized rather exactly in differ- ent ways. It is strange, as Gutman after points out, that precipita- tion by the method of Howe, it appeared in the euglobulin, then in the pseudoglobulin I fraction and that it sometimes partially precipi- tated with pseudoglobulin II. Since I have shown that euglobulin has its precipitation maximum at the concentration recommended by Howe for total precipitation and that total precipitation is not attained until about 190 gm. of Na2SO4 per liter is reached, it seems permissible to suggest that the euglobulin, as I use the term, the 'y globulin of Tiselius, and the a globulin of Kendall are identical. Pseudoglobulin, the protein which precipitates at from about 190 to 255 gm. of Na2SO4 per liter, is probably identical with the a, P, and 2 globulins of Tiselius and with the immunologically non-homo- geneous x globulin of Kendall. We may recall the conclusions of de Vries.27 He pointed out that positive reactions with the Takata Ara or Mancke-Sommer technique aside depend, from the necessarily low albumin content, upon an of increase euglobulin, but sometimes they depend exclu- sively upon an increase in pseudoglobulin I. This conclusion may not be final since is it based only upon the separation of fractions according to the method of Howe, and the statement should read that the reaction of Takata Ara or Mancke-Sommer is positive when 438 YALE JOURNAL OF BIOLOGY AND MEDICINE the actual euglobulin, that the is, y globulin of Tiselius and a globu- lin of Kendall is increased. This thesis is supported by the fact that it has already been shown that in cases of cirrhosis of the liver in which the Takata Ara reaction is generally positive there is an increase in both the y globulin of Tiselius and the a globulin of Kendall. My own observations on patient 0. afford an (Fig. 5) even for this stronger argument conclusion, for this serum which gave very strongly positive Mancke-Sommer reaction had greatly increased euglobulin with hardly any increase in pseudoglobulin. It is of some importance to note that in this highly pathological serum there was no indication a of quantitative in the change serum proteins. However, the possibility of such a in change some dis- eases should be given consideration, as has been shown recently by Wuhrmann and Leuthardt.29 These workers report solubility curves from human sera in phosphate solutions, applying the method of Butler and Montgomery.3 Though the Swiss autho;rs do not differ- entiate their curves, the precipitation maximum for albumin is apparent. For one normal and two pathological sera the maximum is located - - -. at about 2.1 molar P04 In the case of a patient with chronic nephrosis this is shifted considerably to the left. Since with the - - serum at the concentration of P04- the serum protein has been completely precipitated, the authors conclude that no albumin at all was present. A more logical conclusion, it seems to me, would be that albumin has been quantitatively changed. It has become less soluble in concentrated salt solutions and, consequently, the pre- cipitation maximum has shifted to the left. I have had as yet no to confirm opportunity this observation of Wuhrmann and Leuthardt. By means of other technical methods also, observations have been made which point out quantitative changes in the serum proteins in patients exhibiting a nephrotic syndrome. MacFarlane"9 noted that the urinary protein of such a patient shows in the ultracentrifuge a much more polydisperse pattern than does the serum albumin of healthy people. Bourdillon2 in 1939 measured the osmotic pressure of serum and urine albumin in normal persons and in patients with the nephrotic and syndrome, from his results he calculated the molec- ular weight. In the patients he arrived at figures considerably different from normal values. Leutscher"7 noted by means of the Tiselius technique that the serum alibumin of nephrotic subjects is of SOLUBILITY CURVES OF SERUM PROTEINS 439 abnormal composition. These observations make it that the urgent work of the Swiss authors be repeated. Attention has been called to the work of who already Haslam,10 thought that the various serum could be isolated means proteins by of fractional and others have also precipitation. S0rensen22 used this technique. because of seems to However, my results, it necessary question the of this method unless or utility special selected concen- trations of salt are used. that for Let us assume fractional precipi- tation the concentration of 215 gm. of per liter, which is Na2SO4 that of should used. Howe, be (This is about a half-saturated solu- tion and C. about at 32° 470 gm. of anhydrous can be dis- Na2SO4 solved in 1 if liter.) Even precipitation is repeated only a few of times a division the pseudogldbulin into two cannot be parts avoided. Thus the albumin filtrate will contain a always protein with globulin properties and will the globulin precipitate always be partly soluble in water. It is that probable the classic separation method by.means of (NH4)2SO4 will suffer from the same fault, inasmuch as Howe obtained approximately the same results use by of a salt solution containing 215 gm. of Na2SO4 per liter as when he employed half-saturated (NH4)2SO4 solution. Our study makes it appear most likely that serum proteins can be the more completely isolated by a single precipitation when the salt concentrations used are correctly chosen than can be the case by use of fractional precipitation by means of half-saturated (NH4)2SO4 or Na2SO4 solutions. Solutions containing 190 gm. per liter and 260 gm. per liter of serve best for Na2SO4 this purpose. Cohn and his collaborators5' arrived at the 6 same results when serum proteins were treated with (NH4)2SO4 solutions of exactly known concentra- tion and or with pH ethanol-water mixtures. When they chose a suitable concentration these workers succeeded in obtaining almost pure solutions of y, a, and d globulins and of albumin. The purity of the fractions obtained by precipitation was controlled by the Tiselius electrophoresis technique. Attention should be given to one more point'by Luetscher"7 who states, upon the basis of the of results his electrophoretic studies, that the results obtained by of use the new method approximate those of Howe if one is with almost dealing normal sera. In pathological sera the decrease of albumin the content is often much greater than is to be expected from tests made according to Howe. This fact is also shown in our Figs. 4 and 5. To illustrate this point I present values 440 YALE JOURNAL OF BIOLOGY AND MEDICINE albumin content of and of normal below expressing the pathological sera: Albumin content By preciritation According to Howe Calculated from with 260 gm. of (210 gm. of Na2SO4 surface area per liter per liter) Na2SO4 Nornal 39.3 39.1 41.9 42.7 48.8 48.6 Pathological 23.3 31.3 42.0 Howe was about 9.5 This table shows that the result according to serum it gm. per liter (24 per cent) too high, while in pathological was 18.7 per liter (80 per cent). Summwry 1. A number of sera were examined with Na2SO4 solutions of From the solubility curves based on widely varying concentrations. it these studies appeared: of from 210 to 220 gm. (a) That in the case of a concentration of Na2SO4 per liter there does not appear a constant "critical zone" in believed. the curve as Howe on That the concentration of 215 gm. per liter, which the (b) albumin and strength of Howe's work is generally used to separate globulin, lacks theoretical justification. (c) That there is no acute angle in the solubility curves of the such as was accepted by Butler and Montgomery. serum proteins That differentiating the solubility curves three distinctly (d) by the distinguishable serum proteins can be recognized, namely (i) albumin which about 305 of per largely precipitates at gm. Na2SO4 liter; about 145 (ii) the euglobulin which largely precipitates at of liter; (iii) the pseudoglobulin which does not gm. Na2SO4 per for for this have a clearly defined salt concentration precipitation, protein does not behave as a homogeneous substance. It may consist of some small fractions. (e) That it is possible to separate the three serum proteins fairly completely by using concentrations of Na2SO4 equal to 190 and liter. 260 gm. per in in is here is 2. Euglobulin, the sense which the word used, the most probably identical with electrophoretically distinguishable y globulin of Tiselius and with the immunologically characterized ac of globulin of Kendall. It is largely an augmentation in the amount this protein which renders positive the reactions of Mancke-Sommer and Takata Ara. SOLUBILITY CURVES OF SERUM PROTEINS 3. Serum which yields a strongly positive Mancke-Sommer reaction showed only marked quantitative changes in the pattern of the serum proteins. Qualitative changes could not be demonstrated. REFERENCES I Balint, P., and M. Balint: Biochem. Ztschr., 1940, 306, 296. 2 Bourdillon, J.: J. Exper. Med., 69, 819. 1939, 3 Butler, A. M., and H. Montgomery: J. Biol. Chem., 1932, 99, 173. Cohn, E. J.: Physiol. Rev., 1925, 5, 349. 5 Cohn, E. J., J. A. Luetscher, J. L. Oncley, S. H. Armstrong, and B. D. Davis: J. Am. Chem. Soc., 1940, 62, 3396. 6 Cohn, E. J., T. L. McMeekin, J. L. Oncley, J. M. Newell, and W. L. Hughes: J. Am. Chem. Soc., 1940, 62, 3386. 7 Geill, T.: Ztschr. f. klin. 334. Med., 1929, 110, 8 D. H. E. B. and Kabat: Gutman, A. B., Moore, Gutman, V. McClellan, E. A. J. Clin. Invest., 1941, 20, 765. 9 R. C.: Harris, J. Biol. Chem., 1939, 127, 751. Haslam, H. C.: Biochem. J., 1913, 7, 492. 11 Hill, A. Bradford: Principles of medical statistics. 1937. London, 12 Howe, P. E.: J. Biol. Chem., 1921, 49, 93. 13 Howe, P. E.: J. Biol. Chem., 1923, 57, 241. 14 Kendall, F. E.: J. Clin. Invest., 1937, 16, 921. 15 Longsworth, L. G., and D. A. McInnes: J. Exper. Med., 1940, 71, 77. 16 Longsworth, L. G., Th. Shedlovsky, and D. A. McInnes: J. Exper. Med., 1939, 70, 399. 17 A. Luetscher, J. Jr.: J. Clin. Invest., 1940, 19, 313. 18 Luetscher, J. A. Jr.: J. Clin. Invest., 1941, 20, 99. 19 MacFarlane, A. S.: Biochem. J., 1935, 29, 407, 660, 1175, 1202, 1209. 20 Over de Beteekenis van Majoor, C. L. H.: het Serumalbumine voor de beoordeeling en Behandeling van Inwendige en Chirurgische ziekten. Disseration, Amsterdam, 21 Robinson, H. W., J. W. Price, and C. G. Hogden: J. Biol. Chem., 1937, 120, 22 S. P. L.: Compt. du rend. d. trav. lab. Carlsberg, serie physiol., Sorensen, 1930-31, 18, No. 5. 23 Streef, G. H., and A. H. Streef-Spaan: Geneesk. tijdschr. v. Nederl.-Indie, 1938, 77, 2816. 24 Svedberg, Th.: Kolloid Ztschr., 1938, 85, 119. 25 T.: Acta med. 305. Teorell, Scandinav., 1928, 68, Tiselius, A.: Biochem. J., 1937, 31, 1464. 27 de Vries, A.: Acta med. Scandinav., 1938-39, 98, 95; 1939, 99, 425. 28 Wu, H.: Chinese J. Physiol., 125. 1933, 7, 29 Wuhrmann, F., and F. Leuthardt: Helvet. med. acta, 1941, 7, 565. The writer wishes to acknowledge his indebtedness to Prof. Dr. J. G. G. Borst, who directed this work, to Dr. J. Spaander for valuable suggestions, and to Capt.. E. R. Harvey, U.S.A., for indispensable help with translation and corrections.

Journal

The Yale Journal of Biology and Medicine – Pubmed Central

Published: May 1, 1946

There are no references for this article.