CLIN. CHEM. 20/1,30-35(1974)

30 CLINICAL CHEMISTRY, Vol. 20, No. 1, 1974

Rapid Two-Dimensional lmmunoelectrophoresisof

HumanSerum Proteins

G. L Wright, Jr.,1 Louis POliack, and D. B. Roberts

We describe two versions of a two-dimensionalimmunoelectrophoretic system and illustrate theiruse for separating and estimating human serum pro-teins. Each is sensitive, rapid, and simple to per-form, and requires no special apparatus and only0.1-0.6 ml of antiserum. Both procedures appear tobe potentially useful in the clinical laboratory. Timerequired to obtain a stained pattern varies from 4.5to 7 h, depending on the intended purpose and sen-sitivity (number of possible precipitin peaks) desired.With the faster method (“rapid screening”) 25 to 35precipitin peaks are detected. With the slower butmore sensitive method 40 to >50 peaks are detect-ed. The sensitivity of either version is a function citthe titer of the polyvalent antiserum to whole humanserum. Of the precipitin peaks thus far detected, 35have been identified.

IEP2 is a highly sensitive technique for study ofproteins in both normal and pathologic sera andother biological fluids and has been successfully ap-plied to monitoring serum profiles of patients (1).Yet, in spite of its prognostic value, IEP has fallenshort of expectations as a practical diagnostic tooi,primarily owing to difficulties in interpreting IEPpatterns and quantitating the arcs, as well as to theinconsistent resolution of abnormal proteins charac-teristic of disease processes.

Development of two-dimensional or “cross” elec-trophoresis by Laurell (2) has increased the potentialuse of IEP in the clinical laboratory. This methodprovides greater sensitivity and facilitates interpre-tation and quantitation of the precipitin peaks ob-tained. Within the last four years, numerous modifi-cations have been proposed in an attempt to improvethe clarity of such electrophoretic patterns, to identi-fy the precipitin peaks, and to decrease the time re-quired to perform the test.

Department of Microbiology, George Washington UniversityMedical Center, Washington, D. C. 20037.

‘Request for reprints and other inquiries should be addressedto G. L. W. Present address: Department of Microbiology andCell Biology, Eastern Virginia Medical School, Smith RogersHall, 358 Mowbray Arch, Norfolk, Va. 23507.

2 Nonstandard abbreviations used: IEP, immunoelectrophor-esis; 2D-IEP, two-dimensional immunoelectrophoresis; and NHS,normal human serum.

Received Aug. 20, 1973; accepted Oct. 20, 1973.

Clarke and Freeman (3), Raisys and Aryan (4),

Kr#{246}ll(5, 6), and Weeke (7) have used “macro” tech-niques, which necessitate large amounts of antiserum(2-5 ml) and a 17-24 h electrophoretic separation inthe second direction. They could quantitate many ofthe serum proteins by measuring either the area orheight of the precipitin peaks, which are proportion-al to their antigen concentration. However, suchmethods were not economically practical, and fur-ther modifications were needed to facilitate clinicalapplication.

Accordingly, Firestone and Aronson (8), Davies etal. (9), and Crowle (10) have developed “micro”techniques that require much smaller amounts of an-tiserum (125-300 id) and much shorter runningtimes (3-3.5 h). Individual serum proteins could beestimated, and some 30-40 precipitin peaks wereconsistently observed. Yet the performance of theseprocedures is somewhat awkward and requires spe-cially constructed apparatus.

This report describes two further modifications: a“micro” 2D-IEP procedure and a “rapid screening”2D-IEP procedure. These techniques were developedfor clinical applications: first, only economical butpracticable amounts of serum and antiserum are re-quired; second, less than 8 h is required to produce acompletely stained electrophoretic pattern; third, theserum protein pattern is easily identified and quanti-tated; and fourth, the procedure is uncomplicated,readily mastered, and requires no specially-con-structed apparatus.

Materials and MethodsAntisera. Serum (NHS) was pooled from 102 nor-

mal, healthy individuals, and stored at -20 #{176}Cinsmall aliquots until required. Antiserum was pre-pared against the NHS pool in rabbits according tothe method described by Hirschfield (11). Booster,regimens were used when the initial test-bleeding re-vealed a low-titered antiserum (i.e., few precipitincurves). Commercial polyvalent and monovalent an-tihuman antisera were obtained from Behring Diag-nostic, Somerville, N. J. 08876; Kallestad Laborato-ries, Chaska, Minn. 55318; Hyland, Division Traven-ol Laboratories, Inc., Costa Mesa, Calif. 92626; Pen-tex,’ Miles Laboratories, Inc., Kankakee, Ill. 60901;

AntIserumMatrIxSerum

StrIp

Removed

CLINICAL CHEMISTRY, Vol. 20, No. 1, 1974 31

and Meloy Laboratories Inc., Springfield, Va. 22151.Two-dimensional immunoelectrophoresis: Method

I. The “micro” 2D-IEP procedure is a modificationof a “macro” technique described by Roberts et al.(12). A 5 X 7.5 cm slide was placed on a level surfaceand coated with 6 ml of hot (90 #{176}C)agarose (“Sea-Kem”; Bausch & Lomb, Rochester, N. Y. 14602),dissolved in the barbital gel buffer (well buffer) de-scribed below, diluted with an equal volume of dis-tilled water containing 0.1 g of merthiolate per liter(final concentration of agarose, 10 g/liter). An anti-gen well (2 mm diameter) was punched with its cen-ter 1 cm from both the bottom and the cathodicsides of the slide, and filled with 4 Ml of an equalmixture of NHS and a saturated aqueous solution ofEvans Blue dye. Four hundred-fifty milliliters of 0.06ionic strength barbital well buffer, pH 8.2, was placedin each side of a Colab (Chicago Heights, 111. 60411)electrophoresis chamber. Cellulose sponges, whichserve as supports and as buffer wicks, were placed inthe buffer and covered with filter paper to preventscarring the agarose surface. The agarose-coatedslides were placed, agarose-surface down, across thesponge wicks so that the direction of the first electro-phoresis was carried out at 10 mA per slide (9 V/cm)at 4 #{176}Cuntil the blue albumin front had reached apoint 0.8 cm from the anodic edge of the slide (1.5-2h). After this first electrophoretic separation, theagarose above a line parallel to and 1.5 cm from thebottom edge of the slide was removed and replacedwith 4.2 ml of a warm (55 #{176}C)mixture consisting of2.1 ml of a 20 g/liter solution of agarose in gel bufferand 2.1 ml of an appropriate mixture of antiserumand gel buffer. The amount of antiserum used perslide varied between 0.2 ml and 0.8 ml, depending onits titer, After the agarose-antiserum mixture solidi-fied, the slides were inverted and returned to theelectrophoresis chamber so that the direction of elec-trophoresis was parallel to the 5-cm edge of the slide.Electrophoresis in the second direction was carriedout at 9 V/cm (4 #{176}C)until the dye front reached apoint 0.5 cm from the anodic edge of the slide (1-1.5h). After this second electrophoresis, the slides werewashed in agitating saline (9 g/liter) that was re-placed at 30-mm intervals for 2 h. The slides werethen dried under bibulous paper,’ stained with a 5g/liter solution of Coomassie Brilliant Blue in ethan-ol:glacial acetic acid:distilled water (4.5:1:4.5, byvol) (13) for 7 mm, destained in solvent, allowed todry, and photographed. The total time required toperform the test was 6.5-7 h. Diagrammatic repre-sentations of the precipitin patterns were producedby placing the stained slide in a photographic en-larger and tracing the projected image on graphpaper.

Two-dimensional immwioelectrophoresis: Method

II. The “rapid screening” 2D-IEP procedure was afurther modification of the “micro” 2D-IEP proce-dure described above. Slides 5 X 7.5 cm were coatedwith 6 ml of a 10 g/liter hot agarose solution in gelbuffer. The agarose was then cut and the designated

Template For “RapId ScreenIng” Method

7.5cm

Fig. 1. Diagrammatic illustration of the template used forpreparing a “rapid screening” two-dimensional immu-noelectrophoretic slide

portions removed as illustrated in Figure 1. An equalvolume of agarose (20 g/liter) and antiserum3 in gelbuffer was mixed at 55 #{176}Cand 1.2 ml of the mixturewas poured into each of the three antiserum-agarosematrix troughs and allowed to solidify. Antigen wells(diameter, 2 mm) were punched with their centers 1cm from the cathodic edge of the slide and 0.5 cmfrom each edge of the antigen strip and filled with 4Mlof serum:Evans blue dye mixture (1:3, by vol).

The first electrophoretic separation was run paral-lel to the 5 cm edge of the slides at a current of 10mA per slide (9 V/cm) at 4 #{176}Cuntil the blue albu-min front had reached a point 0.8 cm from the anod-ic edge of the slides (1 h). The slides were thenturned 90#{176},and the second electrophoresis was car-ried out at 20 mA per slide (7 V.cm) at 4 #{176}Cuntil theblue albumin front had reached the bottom edge ofthe next anodic serum strip (about 45-60 mm). Theslides were then washed, stained, destained, anddried as previously described. The total time re-quired to perform the test was 4.5-5 h.

ResultsRepresentative two-dimensional immunoelectro-

phoregrams of normal human serum obtained byMethod I are presented in Figure 2. The patterns ob-tained when the same serum sample was analyzed onseveral slides run at the same time or on differentdays were very reproducible if the same polyvalentantiserum was used. The patterns obtained in Figure2 were obtained with 400 Ml of the same antiserum.Although quantitative differences were observedwhen different normal serum samples were runagainst the same antiserum, the number and posi-tion of the precipitin curves were consistent. It is,however, important to point out that not all individ-ual or pooled polyvalent antisera prepared in rabbitsagainst the same NHS pool were consistently able toprecipitate the same serum protein antigens. For ex-ample, in Figure 2 the thyroxine-binding prealbuminis readily observed in the two-dimensional patternsrepresenting three different serum samples. At firstglance, this antiserum appears to be of the quality

amount of antiserum required depended upon its titer.

B

CV C

I

32 CLINICAL CHEMISTRY, Vol. 20, No. 1, 1974

4-

FIg. 2. Two-dimensiona! immunoelectrophoretic patternsobtained when different normal human serum samplesare reacted with the same polyvalent antiserum2 MI of serum, 400 MI of antiserum. (A) IndIvIdual serum sample 1, (B)NHS pool, (C) IndIvIdual serum sample 2. Arrow Indicates posItion ofprealbumin peak

and high titer necessary to achieve maximum sensi-tivity and economy (i.e., lesser quantities of antise-rum) for this IEP technique. However, identificationof some of the precipitin cUrves and analysis of dif-ferent batches of antisera revealed that this antise-rum (Figure 2) failed to precipitate the aj-lipopro-tein.

A second polyvalent antiserum prepared in rabbitsagainst the same NHS pool (Figure 2B) not onlypermitted the detection of the ai-lipoprotein, but re-vealed the genetic polymorphism (13) of this serumprotein as indicated by the triphasic shape of theprecipitin curve in the NHS pool pattern (Figure 3A)and the diphasic pattern of the ai-lipoprotein as ob-served in an individual serum pattern (Figure 3B).The location, identity, and polymorphism of the al-

lipoprotein were confirmed by the use of a monospe-cific antiserum (Figure 3C). Although the polyvalentantiserum used to obtain the patterns in Figure 3Aand B contained a wide spectrum of high’ titer anti-bodies for the a and globulins, it failed to precipi-tate the prealbumin.

It is most important to have a polyvalent antise-rum containing a high titer of antibodies directedagainst a diverse spectrum of serum protein antigen-ic determinants, to achieve maximum sensitivity ofthe “micro” 2D-IEP methods described herein. Be-cause of the potential application of the rapid 2D-IEP systems in the clinical laboratory, several com-mercial rabbit polyvalent antisera to whole humanserum were evaluated by Method I.

The difference in 2D-IEP patterns observed afteranalysis of the same serum sample with four com-

Fig. 3. Two-dimensional immunoelectrophoretic patternsof normal human serum, illustrating the genetic polymor-phism of the a1-lipoprotein (a1LP) antigen (arrows)The patterns in A and B were obtaIned with two different serum samplesreacted with the same polyvalent antiserum. The antiserum was differentfrom that used to obtain the patterns in Figure 2. 2 MI of serum, 600 MI ofantiserum. Pattern C was obtained with monospecific antisera againstthe a1-LP. 2 M1of serum. 200 MI of antiserum

Fig. 4. Two-dimensional immunoelectrophoretic patternsfor a pool of normal human serum, as obtained with fourdifferent commercial rabbit polyvalent antisera to wholehuman serum(A) Kallestad, (B) Behring, (C) Hyland, (0) Pentex. 2 MI of serum, 600MI of antiserum. Note difference in the four patterns with respect to num-ber of precipitin peaks.

mercial antisera is shown in Figure 4. The concen-tration of antiserum was the same (600 Ml) in allelectrophoretic slides. Based on the largest numberof precipitin peaks, the Behring antiserum with 38precipitin peaks was by far the best quality antise-rum tested and compared well with the best polyva-lent antiserum produced in this laboratory (41peaks). Next best was Kallestad (29 peaks), followedby Hyland (25 peaks), and Pentex (18 peaks). Anti-serum obtained by Meloy (not shown in Figure 4)gave a similar pattern to that of Hyland. The num-ber of precipitin peaks depended on the titer of the

CLINICAL CHEMISTRY, Vol. 20, No. 1, 1974 33

Fig. 5. Composite diagrammatic representation of the two-dimensional immunoelectrophoretic pattern of normal humanseraThe proteins identified are indicated by number: (1) prealbumin, (2) albumin, (3) a,-T glycoprotein. (4) a1-acid giycoprotein, (5) a1-chymotrypsin, (6)a,-antitrypsin, (8) inter a,-trypsln inhibitor, (9) a1-iipoprotein, (10) a1-cerutoplasmin, (11) a2-ceruloplasmin, (12) a2,macroglobulin, (13) haptoglobin,(14) a2 HS-glycoprotein, (15) haptoglobin, (16) fibrinogen, (17) hemopexin, (18) transferrin, (19) $1-lipoprotein, (20) $1-lipoprotein, (21) $1-A globu-ferrln, (22) $1-A globulin, (23) $c globulin, (24) $1-E globulin, (25) $2-glycoprotein Il, (26) Ig 0, (27) lg M, (28) Ig A, (29) lg 6, (30) $2-glycoproteinI, (31) Zn-a2-glycoprotein, (33) a1-easiiy precipitable glycoprotein, (34) group-specific components (Gc), (35) group-specific components (Gc), (40)cholinesterase, (41) plasminogen. Precipitin curves not identified: 7, 32, 36, 37, 38, 39, and 42

antiserum. The number of precipitin curves thatcould be detected by a poor or low-titered antiserumcould be increased by increasing the concentration ofantiserum. But then the “micro” 2D-IEP system wasno longer economically performed.

The number of serum proteins detected in NHS byMethod I with antisera produced in this laboratorywas 42. The 35 serum proteins identified to date areindicated in Figure 5. These proteins were identifiedby monospecific antisera and specific staining reac-tions (7).

The patterns obtained by Method II or the “rapidscreening” 2D-IEP modification are presented inFigure 6. The patterns were similar to the larger ver-sion (Method I), although fewer precipitin peaks(25-35 peaks) were observed. The patterns were re-producible within a given run and from day to day aslong as the same concentration of the same antise-rum was used. Again, it was important to have ahigh-titered polyvalent antiserum to achieve maxi-mum sensitivity (25-35 peaks) while keeping the re-quired antiserum volume to a minimum (100 Ml).Although the number of precipitin peaks could beincreased by increasing the concentration of the an-tiserum, it was necessary to keep the quantity of an-tiserum low, otherwise the background would be sointensely stained as to mask many of the precipitinpeaks.

Figure 7 shows some electrophoretic (Method I)patterns of sera from patients with pathologic disor-ders, analyzed by the use of antisera against NHS.As expected, a marked elevation in the immunoglob-iilins (IgG, 1gM, IgA) was observed in the alcoholiccirrhosis serum pattern (Figure 6B). Aside from theexpected elevations of the a- and fl-globulins in theimmunoelectrophoregrams of sera from patients withrenal disorders (Figure 6A and D), the total numberof precipitin peaks increased from the 42 detected inNHS to more than 50 peaks, with the same polyva-lent antiserum that was used to obtain the normalserum patterns in Figure 3A and B. Such observa-tions were consistently found, yet their significanceremains to be determined.

The pattern obtained with serum from a patientwith macroglobulinemia (Waldenstr#{246}m)showed amarked decrease in albumin, and in the a- and fi-globulins. Characteristic of this disorder was thefinding of an elevated diphasic 1gM precipitin curve.The shape of the 1gM curve agrees with the usual“seagull-winged” 1gM precipitin line observed byconventional IEP, which is indicative of an 1gMmonoclonal gammapathy.

DiscussionThe 2D-IEP modifications described in this report

are potentially applicable in the clinical laboratory,

34 CLINICAL CHEMISTRY, Vol. 20, No. 1, 1974

A

B

C

Fig. 6. Two-dimensional immunoelectrophoretic patternsobtained by the “rapid screening” methodThe same serum sample was used in patterns A and B; a different NHSwas used in pattern C. The same antiserum was used to obtain the threepatterns. 1 M’ of serum. 100 MI of antiserum

Fig. 7. Two-dimensional immunoelectrophoretic patternsof sera from patients with disease(A) Membranous glomerulonephritis, (B) alcoholic cirrhosis, (C) macro-giobulinemia. (0) nephrotic syndrome. The same polyvalent anti-NHSwas used to obtain the four patterns. 2 M’ of serum. 600 MI of antiserum.Immunoglobulins IgA (A), lgo (0), and 1gM (M)

in that the procedures are simple, rapid, economical,and require no special-purpose apparatus.

We experienced none of the procedural difficultieswith the methods of Clarke and Freeman (3) andDavies et al. (9), as evaluated by Fuller and Keyser(14). These included difficulties in maintaining theproper gel pH and electrolyte conditions during theelectrophoretic run, the appearance of distortions inthe gel caused by current abnormalities, the necessi-ty for excessive manipulation of apparatus, and diffi-culty in removing the unprecipitated proteins fromthe gel.

The optimal electrophoretic conditions were main-tained in the present procedures by reversing theelectrodes after every second electrophoresis and bychanging the buffer in the electrophoretic chamber

after four complete (both directions) electrophoreticruns. The only problem encountered in either of the2D-IEP modifications was the presence of a stainingartifact as seen in the upper left corner of the pat-terns in Figures 2 and 7. This was due to proteincontamination remaining in the sponge wicks from aprevious run. This could easily be eliminated byrinsing the sponge wicks with water after each run orprotecting the sponge wicks with a sheet of cellulosemembrane cut from dialysis tubing.

The first 2D-IEP procedure (Method I) providesoptimal conditions where maximum sensitivity (i.e.,greatest number of precipitin peaks) is desired. Afterthe second electrophoresis the slides do not requirean independent incubation step for development ofthe precipitin arcs since the slides are washed andincubated simultaneously in agitating isotonic sa-line. Forty-two precipitin peaks can be obtained withNHS in reaction with as little as 400 Mi of high-ti-tered polyvalent antiserum. Over 50 precipitin peakswere observed with sera from patients with renal ab-normalities, analyzed with 400 Ml of the same poly-valent antiserum to whole NHS. Thirty-five of theserum antigens have been identified and many con-firm the identifications of Weeke (7). The entire pro-cedure, from preparing the agarose slide to obtaininga dried and stained slide, requires about 6.5-7 h.

If speed is desired rather than maximum sensitivi-ty, then the second modification (Method II) may bemore appropriate as a quick screening procedure fordetermining qualitative alterations in the serum pro-teins. All of the procedural manipulations, includingcoating the slide with agarose, punching out the an-tigen well, and applying the agarose-antiserum mix-ture, are performed at the start and require onlyabout 30 mm. Once the slide is placed in the electro-phoresis chamber, the only additional step is to turnthe slide 90#{176},after the first electrophoretic separa-tion is completed, to allow electrophoresis of the an-tigen into the agarose-antiserum matrix. The entireprocedure takes about 5 h, so that a trained labora-tory technician, with two electrophoresis chambers,could run 48 individual serum samples in an 8-hworking day. The vertical alignment of the devel-oped patterns also allows individual serum compo-nents from a number of samples to be compared.

The most important parameter for success in ob-taining maximum sensitivity and reproducibility ishaving a polyvalent antiserum containing high titerof antibodies to a great diversity of serum proteinantigenic determinants. Since the quality (i.e., num-ber of antibody specificities) and titer of antiseravaries from batch to batch, the optimal amount tobe used cannot be assigned to either of the methodsdescribed in this report. The various commerciallyprepared antisera gave no more consistent immuno-electrophoregrams than did antisera prepared in ourlaboratory. Each antiserum, therefore, must be test-ed against a standard serum pool and the antiserumconcentration adjusted to give optimal 2D-IEP pat-terns. Although increasing the concentration of a

CLINICAL CHEMISTRY, Vol.20, No. 1, 1974 35

poor titer antiserum increases the number of precipi-tin peaks, the test is no longer economical, Withgood quality antiserum, as little as 0.1 ml can resultin as many as 35 precipitin curves if Method II isused. This procedure is then not only fast but veryeconomical.

Most of the emphasis in developing or modifyingexisting 2D-IEP systems has been on obtaining arapid procedure (8, 9, 10, 15), designing special ap-paratus (4, 8, 9, 15), identifying the precipitin curves(3, 5-7, 16), and quantitating the serum protein an-tigens (1, 3, 17-19). Our major emphasis in thisstudy was on the development of a rapid 2D-IEPprocedure by using existing and simple apparatus,and on the identification of the protein antigens forthe purpose of being able to determine the reproduc-ibility and quality of the antisera.

Now that rapid, reproducible, and economical pro-cedures have been developed in our laboratory andmost of the serum proteins identified, we are cur-rently attempting to establish a quantitative normalserum profile with respect to sex and age for each ofthe serum antigens. After a normal serum profile isobtained with reference antiserum and quantitated,normal serum standards are established, it may bepossible to quantitate 20 or more of the biologicallyimportant serum proteins simultaneously on thesame slide and thereby eliminate the present needfor different radial immunodiffusion plates to quan-titate each serum protein. This would not only savetime but would be considerably less expensive, mak-ing the rapid 2D-IEP systems increasingly valuableto the clinical laboratory for detecting both qualita-tive and quantitative differences in patients’ sera.

This study was supported in part by a grant from the John A.

Hartford Foundation, Inc., New York, N. Y.

References

1. Clarke, J. G., Freeman, T., and Pryse-Phihips, W. E. M.,Serum proteins changes in Still’s disease, rheumatoid arthritisand gout. Brit. J. Exp. Pat hol. 5, 441 (1970).

2. Laurell, C-B., Antigen-antibody cross-electrophoresis. Anal.Biochem. 10, 358(1965).

3. Clarke, H. G. M., and Freeman, T., Quantitative immunoelec-trophoresis of normal serum proteins. Clin. Sci. 35,403(1968).4. Raisys, V. A., and Aryan, D. A., Determination of proteins inbiological fluids by electroimmunodiffusion and two-dimensionalimmunoelectrophoresis. Clin. Chem. 17,745(1971).5. KrOll, J., On the immunoelectrophoretical identification andquantitation of serum proteins. Scand. J. Clin. Lab. Invest. 22,79(1968).6. Kroll, J., Immunochemical identification of specific precipitinlines in quantitative immunoelectrophoresis patterns. Scand. J.Clin. Lab. Invest. 24, 55 (1969).

7. Weeke, B., Serum protein identification by means of the Lau-rell cross-electrophoresis. Scand. J. Clin. Lab. Invest. 25, 269(1970).8. Firestone, H. J., and Aronson, S. B., Thin gel quantitationtwo-dimensional immunoelectrophoretic method. Amer. J. Clin.Pat hol. 52, 615 (1969).

9. Davies, D. R., Spurr, E. D., and Versey, J. B., Modification tothe technique of two-dimensional immunoelectrophoresis. Clin.Sci. 40,411 (1971).10. Crowle, A. J., Preparatory electroimmunodiffusion for makingprecipitins to selected native antigens. Immunol. Commun. 1, 325(1972).11. Hirschfield, J., Immunoelectrophoretic procedure and appli-cation of the study of group specific variations in sara. Sci. Tools7,18(1960).12. Roberts, D. B., Wright, G. L., Affronti, L. F., and Reich, M.,Characterization and comparison of mycobacterial antigens bytwo-dimensional immunoelectrophoresis. Infec. Immunity 6, 564(1972).13. Schultze, H. E., and Hermanns, J. F., Molecular Biology ofHuman Proteins, 1, Elsevier, New York, N.Y., 1966, p 186.

14. Fuller, J. M., and Keyser, J. W., Some technical aspects ofquantitative immunoelectrophoresis of human serum and cere-brospinal fluid. Clin. Chem. 18,625(1972).

15. Pizzolato, M. A., Two-dimensional immunoelectrophoresis oncellulose acetate: improved method for routine protein estima-tion. Clin. Chirn. Acta 45, 207(1973).16. Weeke, B., Carbamylated transferrin used as the reference inthe Laurell cross-electrophoresis. Scand. J. Clin. Lab. Invest. 25,161 (1970).17. Kroll, J., Changes in the $ic- $1A-globulin during the coagu-lation process demonstrated by means of a quantitative immu-noelectrophoretic method. In Protides of Biological Fluids, 17thColloquium. H. Peters, Ed. Pergamon Press, Elmsford, N. Y. 10523,1969, p 529.18. Lyngbye, J., and Kroll, J., Quantitative immunoelectrophor-esis of proteins in serum from a normal population: Season-, age-,and sex-related variations. Clin. Chem. 17, 495(1971).

19. Stephan, V. W., and Frahm, U., Quantitative simultan im-munoelectrophorese. Z. Kim. Chem. Kim. Biochem. 9, 224(1971).