APPLIED MICROBIOLOGY, Jan. 1973, p. 26-36Copyright 0 1973 American Society for Microbiology

Vol. 25, No.1Printed in U.S.A.

Determination of the Optimal AmmoniumSulfate Concentration for the Fractionation of

Rabbit, Sheep, Horse, and Goat AntiseraG. ANN HEBERT, PATRICIA L. PELHAM, AND BERTIE PITTMAN

U.S. Department of Health, Education, and Welfare, Health Services and Mental Health Administration,Center for Disease Control, Atlanta, Georgia 30333

Received for publication 6 October 1972

Various ammonium sulfate concentrations and reaction conditions were

employed in the fractionation of sera from rabbits, sheep, horses, and goats.Precipitates and supernatant fluids were analyzed by electrophoresis to studythe effects of the controlled variables. At room temperature, the thirdprecipitate in 35% saturated (NH)2SO4 was the best fraction from both rabbitand sheep sera; 80 to 90% of the gamma globulins were recovered. The secondand third precipitates of horse sera proteins in 30% saturated (NH),2SO0 wereboth satisfactory, but only 44% of the gamma globulin was recovered after threeprecipitations. Goat sera yielded a very satisfactory fraction; 80% of the gammaglobulin was recovered after two precipitations-the first in 30% and the secondin 45% saturated (NH,)2SO. The composition of these fractions was notinfluenced by the pH of the sulfate solutions (pH 5.8 and 7.2), by a range ofnormal room temperatures (20 to 30 C), or by diluting the sera beforefractionation. Crude globulins and fluorescein isothiocyanate-labeled globulinswere successfully refractionated by one precipitation in the optimal sulfateconcentration for the appropriate animal species. The refractionated productscontained considerably less beta and alpha globulins than did the original crudefractions and little or no albumin.

The most frequently used method for frac-tionating antisera is precipitation in the coldwith half-saturated ammonium sulfate. Theprocedure is simple and inexpensive, and theresulting crude antibody fractions from theantisera of a number of animal species havebeen successfully used in many areas of im-munology. In this laboratory these crude frac-tions have been conjugated with fluoresceinisothiocyanate (FITC) for use as fluorescent-antibody (FA) reagents. During a series ofstudies designed to improve FA methodology,these crude fractions were analyzed by cel-lulose acetate strip electrophoresis. The resultsrevealed that most of the fractions containedless than 50% gamma globulin and that manyof them contained undesirable amounts ofalbumin. A survey of commercial antibacterialFA reagents also had shown a need for im-proved fractionation procedures (5). This needfor improved fractionation coupled with theneed to retain the simple and inexpensiveaspects of the (NH,)2SO4 procedure led to this

26

systematic study of various concentrations ofthe salt for serum fractionation. (Presented inpart at the Annual Meeting of the AmericanSociety for Microbiology, Minneapolis, Minn.,3 May 1971.).

MATERIALS AND METHODSAntisera. The antisera used in these studies were

produced in rabbits, sheep, horses, and goats againsta variety of bacterial antigens including Bacillusanthracis, Bordetella bronchiseptica, Bordetella per-tussis, Escherichia coli, Pseudomonas pseudomallei,Salmonella, Shigella dysenteriae, Shigella sonnei,and Yersinia pestis. Some normal sera were alsoused.Ammonium sulfate. A stock solution of satu-

rated ammonium sulfate (SAS) was prepared andstored at room temperature (approximately 25 C).Working solutions of 50, 60, 70, 80, and 90% SAS wereprepared (v/v) fresh as needed from the stock satu-rated solution. Equal volumes of these solutions andvarious antisera resulted in reaction mixtures of 25,30, 35, 40, 45, and 50% SAS (6).

Fractionation. The following procedure for frac-

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

VoL 25, 1972 (NH4)2SO4 CONCENTRATION IN ANTISERA FRACTIONATION

tionation was used. A volume of serum was gentlystirred while an equal volume of an ammoniumsulfate solution was slowly added and mixed well.The reaction mixture was set aside at room tempera-ture for 4 hr and then centrifuged to pack theprecipitated protein. The supernatant fluid was re-moved and stored for later analysis. The precipitatewas resuspended and dissolved in distilled water to afinal volume equal to the original volume of serum.For a second precipitation, the dissolved protein wasgently stirred while an equal volume of an ammo-nium sulfate solution was slowly added. The mixturewas immediately centrifuged to pack the formedprecipitate, and the supernatant fluid was discarded.The precipitate was dissolved and brought to volumeas before. A third precipitation was handled in thesame manner.

All fractions were dialyzed against frequentchanges (3) ofpH 8, 0.85% NaCl solution until sulfatewas no longer detected in the dialysate (7). Thesaline was brought to pH 8 with a few drops of 10%NaOH. A small volume of saturated barium chloridesolution was added to an equal volume of well-mixedsaline dialysate to check for the presence of sulfate. Ifno cloudiness resulted, the dialyzed fraction wasconsidered substantially free of sulfate.

Protein. Protein concentrations were measured bythe Biuret method (2) with a Beckman DB spectro-photometer. Protein compositions were determinedby cellulose acetate strip electrophoresis (CASE)with the Beckman Microzone (1) equipment andprocedure with a slight modification. After the mem-branes were stained with Ponceau S, they were rinsedin 5% acetic acid until the background was white;then they were gently blotted to remove excessmoisture and placed between dry blotters on a flatsurface beneath a glass thin-layer chromatographyplate to dry. After drying, the uncleared membraneswere read on a Beckman Densitometer, model R-110,at the recommended settings.

Electrophoreuis interpretation. The integrateddensitometer tracings of the proteins subjected toelectrophoresis were interpreted by the followingmethod (Fig. 1). (i) Each of these proteins, whenisolated and studied in the pure state, has a peak ofGaussian configuration. Use Gaussian projections toextrapolate to the base line the right side of thegamma peak and the left side of the albumin peak.(ii) Drop perpendiculars through these projections sothat an area "A" between the original tracing and theprojection is equal to an area "B" inside the projec-tion and on the opposite side of the perpendicular.Extend these perpendiculars down through the inte-grator trace. (iii) The integrated tracing of theproteins subjected to electrophoresis has been parti-tioned into three regions-gamma globulin, beta andalpha globulins, and albumin. To calculate the areain each region, add the individual counts registeredfor each region and obtain the relative proportion ofeach.Fluoresein and fluorescein to protein ratios.

Fluorescein isothiocyanate was determined as pro-tein-bound FITC by absorbance at A max (near 495nm) in 0.1 N NaOH, and related to a pure fluoresceindiacetate reference standard (9). The fluorescein toprotein (F/P) ratio was calculated from the FITC andprotein measurements and expressed as microgramsof protein-bound FITC per milligram of protein.

Specific antibody titration. The specific stain-ing titers of all conjugates were determined byroutine FA staining procedures (10, 11). The highestdilution, which gave a 4+ staining of the bacterialcells, was recorded as the specific titer.

RESULTSPreliminary studies. Many of these studies

were done with pools of animal sera rather thanindividual sera. Once a procedure was workedout it was applied to other pooled sera as well

FIG. 1. Method used to interpret electrophoresis tracings.

27

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

HEBERT, PELHAM, AND PITTMAN

as individual antisera. The electrophoresis pro-files of rabbit, sheep, horse, and goat sera poolsare shown in Fig. 2. They are easily distin-guished from each other, and each is verycharacteristic of its species. The rabbit profileis the least distinctive and is quite similar tothe familiar human serum profile. The sheepgamma globulin resolves as two separate peaks.Horse serum profiles exhibit two strong peaksin the alpha globulin region. In goat serum,gamma is the dominant globulin. These fea-tures remain constant within the species, andthe differences in individual animals are minorexcept for fluctuations in the gamma globulinconcentrations, which reflect immunologicalstatus.

The CASE profiles of serum fractionsobtained after three precipitations of these foursera in 50% SAS are shown in Fig. 3. Thenumerical data extracted from Fig. 2 and 3 arepresented in Table 1. These crude. fractionscontained only 23 to 48% gamma globulin andas much as 9 to 21% albumin. All of the gammaglobulins present in the original sera wererecovered, but many other proteins were alsoprecipitated.Rabbit studies. The CASE profiles of the

first, second, and third precipitations of rabbitE. coli antisera in 50, 45, 40, and 35% SAS areshown in Fig. 4. These tracings show that, asthe concentration of ammonium sulfate was de-creased, the gamma globulin percentage in the

FIG. 2. Electrophoretic profiles of pooled animal sera.

28 APPL. MICROBIOL.

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

VOL. 25, 1972 (NH4)2SO4 CONCENTRATION IN ANTISERA FRACTIONATION

9mprofiles of animal serum fractions after three precipitations in 50% saturated

precipitates was increased. Consecutive pre-cipitations with a given concentration of am-

monium sulfate did not change the generalratio of the amount of gamma globulin to theamount of beta and alpha globulins, but theydid reduce the percentage of albumin in theprecipitate. The third precipitation in 35%SAS gave a fraction which was 65% gammaglobulin and only 1% albumin. Very littleprecipitate formed in 30% SAS, and its compo-sition was not desirable. No precipitation oc-

curred in 25% SAS. Seven other bacterialantisera of rabbit origin were fractionated by

three consecutive precipitations in 35% SAS.The data from CASE of all eight antisera andtheir sulfate fractions are given in Table 2.Gamma globulin in the original antisera rangedfrom 5 to 21%, and the fractions contained from54 to 68% gamma globulin with a maximum of2% albumin.

Sheep studies. The CASE profiles of thefirst supernatant fractions and the third con-

secutive precipitates of sheep serum in 45, 40,35, and 30% SAS are shown in Fig. 5. Withsheep serum as with rabbit serum, decreasingthe sulfate concentration resulted in an in-

FIG. 3. Electrophoretic(NH4)2SO4

29

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

HEBERT, PELHAM, AND PITTMAN APPL. MIcROBIoL.

creased percentage of gamma globulin in the supernatant fractions and third precipitates ofprecipitate, and repeated precipitations in a horse serum in 45, 40, 35, and 30% SAS aregiven SAS concentration yielded reduced per- shown in Fig. 6. Decreasing the sulfate concen-centages of albumin. The third precipitation in35% SAS gave a fraction containing 68% TABLE 2. Electrophoretic results from eight rabbitgamma globulin and less than 1% albumin, bacterial antisera and their fractions obtained withWith this amount of sulfate, a small amount of optimala (NH4),SO, concentrationsgamma globulin was lost in the first superna-tant fraction, but considerably more was lost in Composition by CASE"the first supernatant fraction obtained with Antisera %30% SAS without improvement in the composi- % beta- %

tion of the third precipitate. gamma alpha albumin

Horse studies. CASE profiles of the first P. pseudomallei: Original 5 26 69

their0%satratedNH,),S, fi-ations . csuomai: i Original 7 27 66Fractiona 75 27

CASE"Y.proteinsnfouranmalserumpoolsadE Original 8 37 5l Composition by CSE Y.psi: Fraction |61 |38 1

Serum pool % % % S. dysenteriae: Original 10 34 56gamma beta-alpha albumin Fraction 68 32 0

B. bronchiseptica: Original 12 33 55Rabbit: Original 17 25 58 Fraction 55 45 0

Fraction" 33 58 9 B. pertussis: Original 14 34 52Sheep: Original 26 32 42 Fraction 64 35 1

Fraction 44 47 9 B. anthracis: Original 17 38 45Horse: Original 20 46 34 Fraction 66 34 0

Fraction 23 65 12 Salmonella: Original 21 36 43Goat: Original 30 36 34 Fraction 59 40 1

Fraction 48 31 21a Determined in the present study.

a Cellulose acetate strip electrophoresis. "Cellulose acetate strip electrophoresis.° Crude fraction after three precipitations in 50% c Antiserum fraction after three precipitations in

saturated (NH4),S04. 35% saturated (NH4)ISO4.

Percent Saturated (NH4)2 S04 in the Serum Reaction Mixtureppta 50% 45% 40% 35%

2nd

7y A 7A YA 7Y AFIG. 4. Electrophoretic profiles of (NH4),SO4 fractions of rabbit antiserum for E. coli. a Precipitation.

30

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

VOL. 25, 1972 (NH4)2SO4 CONCENTRATION IN ANTISERA FRACTIONATION

FIG. 5. Electrophoretic profiles of sheep serum fractions obtained with (NH4)2S04. a Saturated ammoniumsulfate.

% SASa

45

____ _ _ _ _ _ _ _ ~~~~~~~35_

30

b 0e"M albumin gamma albumin

FIG. 6. Electrophoretic profiles of horse serum fractions obtained with (NH4)2SO4. a Saturated ammoniumsulfate.

tration had the same general effect on horseserum as that described for rabbit and sheepsera. At 35% SAS some gamma globulin re-mained in the supernatant fraction, but thethird precipitate still contained a high percent-

age of beta and alpha globulins. Lowering theconcentration of SAS to 30% left a largeamount of gamma globulin in the first superna-tant fraction but gave a third precipitate whichwas 83% gamma globulin with no albumin.

31

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

HEBERT, PELHAM, AND PITTMAN

Goat studies. CASE profiles of goat E. coliantiserum fractions are shown in Fig. 7. De-creasing the sulfate concentration caused agreater change in the composition of the firstprecipitate of goat antiserum than it had withany of the other three animal sera. The firstprecipitate in 30% SAS contained 71% gammaglobulin and only 6% albumin. No precipitateformed when a second sample of 60% SAS wasused. When a 45% concentration of SAS wasused to refractionate the 30% SAS precipitate,a fraction containing 81% gamma globulin andonly 1% albumin was obtained.Recommended optimals. The CASE pro-

files of the improved fractions from rabbit,sheep, horse, and goat sera, obtained by usingthe optimal ammonium sulfate concentrationsfor precipitation, are shown in Fig. 8. Rabbitand sheep sera were handled in the same waywith three consecutive precipitations in 35%SAS. Horse serum required three precipita-tions in 30% SAS to effectively remove its highpercentage of beta and alpha globulins. Goatserum required precipitation in 30% SAS fol-lowed by precipitation in 45% SAS. Underthese conditions the percentage of gammaglobulin recovered from the sera of rabbits,sheep, and goats was above 80%, and from thatof horses, slightly less than 50% (Table 3).Refractionation. Crude globulin fractions

of rabbit antisera which had been preparedseveral years earlier by three precipitations in

50% SAS were subjected to a single precipita-tion in 35% SAS. The improvement was dem-onstrated by CASE profiles (Fig. 9). Several FAconjugates for Salmonella and Shigella detec-tion have also been improved by a singleprecipitation in the optimal ammonium sulfateconcentration appropriate for the animal spe-cies involved. The data for one of the Shigellaconjugates are shown in Fig. 10. Refractionatedconjugates were dialyzed in pH 7.6 phosphate-buffered saline (PBS) for sulfate removal.Since PBS reacts with saturated BaCl2 to forma precipitate, BaCl2 could not be used to checkfor the presence of sulfate; therefore, the dial-ysis time was extended to insure completeremoval of the sulfate. Both globulins andconjugates were composed of considerably lessalbumin and beta-alpha globulins after oneoptimal precipitation. The Shigella conjugatelost 30% of its total protein, but only 7% of itsgamma globulin. Although its FITC concentra-tion was reduced 55% and the F/P ratio 36%,the specific antibody titer did not change.Effect of PH. The unadjusted pH of all of

the ammonium sulfate solutions as measuredon the pH meter was 5.8. The pH of allserum-sulfate reaction mixtures ranged from7.2 to 7.6, depending upon the pH of theoriginal serum. Solutions of 90, 80, and 70%SAS were raised to pH 7.2 with 10% NaOH.Rabbit B. bronchiseptica antiserum was thenfractionated in 45, 40, and 35% SAS with both

% SAS8

-_un. Mbunin 9"Im. Aminin

FIG. 7. Electrophoretic profiles of goat E. coli antiserum precipitates obtained after one fractionation with.various concentrations of (NH4)S04. a Saturated ammonium sulfate.

32 APPL. MICROBIOL.

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

pmn albumin

FIG. 8. Electrophoretic profiles of antisera fractions obtained with optimal percent saturated solutions of(NH4)S04. ppt, Precipitate.

TAHE 3. Recovery of gamma globulin from antisera after fractionation with optimala (NH,)jSO4concentrations

No. Protein % gamma Gamma % gammaAntisera SAS" pptsc (mg/ml) CASEd (mg/ml) recovery

Rabbit: Original NA' NA 67 17 11 NAFraction 35 3 15 66 10 91

Sheep: Original NA NA 74 26 19 NAFraction 35 3 24 68 16 84

Horse: Original NA NA 81 20 16 NAFraction 30 3 8.4 83 7 44

Goat: Original NA NA 60 30 18 NAFraction 30,45 1,1 18 81 15 83

a Determined in the present study.,'Saturated (NH4)2S04.c Precipitations.d Cellulose acetate strip electrophoresis.'Not applicable.

33

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

HEBERT, PELHAM, AND PITTMAN

improved globulinb

Shigella dysenteriae A2

gamma albumin

FIG. 9. Improvement in crude globulins from Shigella rabbit antisera after one precipitation with theoptimal (NH4)2SO4 concentration. The crude globulin (a) was obtained after three precipitations in 50%osaturated (NH4)2SO4. The improved globulin (b) was obtained after one additional precipitation of the crudeglobulin, using 35% saturated (NH4)2SO4.

.on sprotein % am+ pam F/Pc FITCd

Mbjuats/ ASmg/.I ratio gtmdmg/ml CAS6 aii a g/I

origina 10.8 49 5.3 14 149

improvd 7.6 65 4.9 9 67.5% change -30 +16 -7 -36 -55

FIG. 10. Improvement in rabbit Shigella dysen-teriae (Al) conjugate after one precipitation with theoptimal (NH4)2SO4 concentration. a Original conju-gate was fractionated by one precipitation in 35%saturated (NH4)2SO4. b Cellulose acetate strip elec-torphoresis. c Fluorescein to protein ratio. d Fluo-rescein isothiocyanate.

the adjusted pH 7.2 reagents and unadjustedpH 5.8 reagents. After three consecutive pre-cipitations under the six different conditions,the three fractions from the pH 7.2 sulfatesolutions were practically identical to thoseobtained with the pH 5.8 reagents (Table 4).Effect of dilution. Some rabbit B. bron:

chiseptica antiserum was also used to study theeffect of dilution of serum prior to fractiona-tion. Equal volumes of the original antiserumand of a 1:2 and 1:10 dilutions were eachprecipitated three times in 40% SAS. The finalfractions had practically identical CASE pro-files (Table 5).

Effect of temperature. To examine thepossible influence of room temperature varia-tions, we prepared and stored saturated solu-tions of ammonium sulfate at 20, 25, and 30 C.At each temperature, 70% SAS was preparedand used to fractionate rabbit B. bronchisep-tica antiserum. The fractions obtained afterthree precipitations in 35% SAS at each tem-

34 APPL. MICROBIOL.

_ -, _,

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

VOLT 25, 1972 (NH4)2SO4 CONCENTRATION IN ANTISERA FRACTIONATION

perature were practically identical (Table 6).

DISCUSSIONFor most immunological applications, the

purpose of serum fractionation is to separateand recover the gamma- or immunoglobulins.Except for the column fractionation methodsand even as a preparative step for these tech-niques, the most widely used method of serumfractionation is precipitation in 50% SAS.Three precipitations are usually recommendedto free the material of hemoglobin, but fre-quently the adequacy of the fractionation interms of the protein composition of the finalproduct is not determined. When rabbit, sheep,horse, and goat sera were fractionated by threeprecipitations in 50% SAS, all of the finalfractions (Fig. 3) contained more beta andalpha globulins and more albumin thangamma globulin. By using the optimal concen-tration of (NH4)3SO4 for a given animal speciesof serum, all of the final fractions (Fig. 8)contained predominantly more gamma globu-lin than any other protein with very little or noalbumin.More than 80% of the gamma globulins

present in the original rabbit, sheep, and goatsera were recovered by the improved proce-dures. The results with horse serum were not assatisfactory. After the serum was precipitated

TABLE 4. Composition (%) by CASEa of rabbitserum fractions obtained with three concentrations of

(NH4)2S04 at two different pH levels

Proteins 45% SASc 40% SAS 35%SASin third pH pH pH pH pH pHpptb 5.8 7.2 5.8 7.2 5.8 7.2

Gamma 35 35 51 46 55 56Beta-alpha 63 63 48 52 45 44Albumin 2 2 1 2 0 0

a Cellulose acetate strip electrophoresis.b Precipitate.c Saturated (NH4)2804-

three times in 30% SAS, the high percentage ofbeta and alpha globulins was reduced, but only44% of the gamma globulin was recovered. Thisfraction was 83% gamma globulin and con-tained no albumin. Approximately 60% of thehorse gamma globulin was recovered after onlytwo precipitations in 30% SAS, and, after thatstep, the fraction contained 74% gamma globu-lin with only 1% albumin. The recovery of ahigher percentage of gamma globulin after two30% SAS precipitations makes the second pre-cipitate as desirable as the third in 30% SAS,unless the elimination of all albumin is a majorobjective.Using 35% SAS, we selectively precipitated

rabbit gamma globulin and recovered a highpercentage of it regardless of the percentage inthe original antiserum. The gamma globulinconcentrations (5 to 21%) in the eight rabbitantisera reported in Table 2 were not correlatedwith the percentage of gamma globulin ob-tained in the product after fractionation. Thegamma globulin recovered from these antiseraranged from 84 to 91% of the original concen-tration.These serum fractionation procedures were

reliable under variable reaction conditions. Atthe values tested, the composition of the frac-tions was not altered by the pH of the ammo-nium sulfate solutions, probably because thebuffering capacity of serum kept the pH of themixture within a very narrow range. However,

TABLE 6. Composition (%) by CASEa of the third(NH4),S04 precipitate of rabbit serum obtained at

three temperature levels

35% SAS"Proteins

20C 25C 30C

Gamma 59 56 57Beta-alpha 41 44 43Albumin 0 0 0

a Cellulose acetate strip electrophoresis.Saturated (NH4),S04.

TABLE 5. Composition (%) by CASEa of rabbit serum fractions obtained from diluted and undiluted serum

40% SAS", 1st pptc 40%, 2nd ppt 40%, 3rd pptProteins

Und | 1:2 1:10 Und 1:2 1:10 Und 1:2 1:10

Gamma 34 36 43 44 48 50 51 49.8 54Beta-alpha 45 44 43 52 50 48 48 49.8 44.5Albumin 21 20 14 4 2 2 1 0.4 1.5a Cellulose acetate strip electrophoresis.I Saturated (NH4)ISO4c Precipitate.d Undiluted.

35

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from

HEBERT, PELHAM, AND PITTMAN

diluted serum has less buffering ability, and, ifit is fractionated with ammonium sulfate thathas not been neutralized, careful attentionshould be given to pH to avoid acid reactionconditions. No differences were noted in thecomposition of fractions obtained by precipita-tion of rabbit antiserum with 35% SAS at 20,25, and 30 C; this finding indicates that varia-tions in temperature within these limits had noeffect. Because the molarity of SAS varies withtemperature, no attempt has been made toexpress concentration in molar equivalents,but the SAS used at approximately 25 C wasapproximately 4.09 M. The data indicated thatthe composition of fractions prepared fromundiluted serum was approximately the sameas the composition of fractions prepared fromserum diluted either 1:2 or 1: 10. The reliabil-ity of these procedures depends upon propertechnique. Ammonium sulfate solutions mustbe slowly added to serum while it is gentlystirred. When sulfate solutions are poured intosera and the reaction mixtures are shaken orstirred vigorously, the final fractions containconsiderably more albumin and beta and alphaglobulins than do properly prepared fractions.

Refractionation of very crude 50% SAS glob-ulins with a single precipitation in the opti-mal SAS concentration for the particular ani-mal species greatly improved the compositionof the fractions. Most of the albumin and alarge portion of the beta and alpha globulinswere effectively removed.Crude FITC conjugates were also success-

fully refractionated with optimal SAS concen-trations. Only very small amounts of the la-beled gamma globulins were lost, and thespecific titers did not change. Approximatelyone-third of the total labeled proteins wereremoved, and the FITC concentrations werereduced more than 50%. Because of the directrelationship between FITC concentration andnonspecific staining (4), the improved conju-gates should exhibit less nonspecific stainingthan before. The large amount of FITC re-moved by reducing the percentage of labeledbeta and alpha globulins and eliminating the

labeled albumin reflects the high avidity ofthese proteins for FITC (8); it also illustratesthe need for employing optimum procedures forfractionation before the antibody is labeledwith a fluor.

ACKNOWLEDGMENTWe thank Berenice M. Thomason for specific analysis of

the enteric reagents, consultation, and assistance in thepreparation of this manuscript.

LITERATURE CITED1. Beckman Instruments, Inc. 1965. Instruction manual for

model R-101 Microzone electrophoresis cell. Fullerton,Calif.

2. Gornall, A. G., C. J. Bardawill, and M. M. David. 1949.Determination of serum proteins by means of theBiuret reaction. J. Biol. Chem. 177:751-766.

3. Hebert, G. Ann, and Bertie Pittman. 1965. -Factorsaffecting removal of (NH4)2S04 from salt fractionatedserum globulins employing a spectrophotometricprocedure for determination of sulfate. Health Lab.Sci. 2:48-53.

4. Hebert, G. Ann, Bertie Pittman, and William B. Cherry.1967. Factors affecting the degree of nonspecific stain-ing given by fluorescein isothiocyanate labeled globu-lins. J. Immunol. 98:1204-1212.

5. Hebert, G. Ann, Bertie Pittman, and William B. Cherry.1971. The definition and application of evaluationtechniques as a guide for the improvement of fluores-cent antibody reagents. Anal. N. Y. Acad. Sci.177:54-69.

6. Hebert, G. Ann, Bertie Pittman, Roger M. McKinney,and William B. Cherry. 1972. The preparation andphysicochemical characterization of fluorescent anti-body reagents. U.S. Dept. of Health, Education, andWelfare, Center for Disease Control, Atlanta, Ga.

7. Kaufman, L., and W. B. Cherry. 1961. Technical factorsaffecting the preparation of fluorescent antibody rea-gents. J. Immunol. 87:72-79.

8. McKinney, R. M., J. T. Spillane, and G. W. Pearce.1964. Factors affecting the rate of reaction of fluores-cein isothiocyanate with serum proteins. J. Immunol.93:232-242.

9. McKinney, R. M., J. T. Spillane, and G. W. Pearce.1964. Fluorescein diacetate as a reference color stand-ard in fluorescent antibody studies. Anal. Biochem.9:474-476.

10. Thomason, Berenice M., Glenda S. Cowart, and WilliamB. Cherry. 1965. Current status of immunofluores-cence techniques for rapid detection of shigellae infecal specimens. Appl. Microbiol. 13:605-613.

11. Thomason, Berenice M., and Joy G. Wells. 1971. Prepar-ation and testing of polyvalent conjugates for fluores-cent-antibody detection of salmonellae. Appl. Mi-crobiol. 22:876-884.

36 APPL. MICROBIOL.

on April 13, 2020 by guest

http://aem.asm

.org/D

ownloaded from