INFECTION AND IMMUNITY, June 1976, p. 1692-1698Copyright ©D 1976 American Society for Microbiology

Vol. 13, No. 6Printed in U.SA.

Preparation of a Purified Antigenic Cholera ToxoidR. GERMANIER,* E. FORER, S. VARALLYAY, AND T. M. INDERBITZIN

Swiss Serum and Vaccine Institute, Berne, Switzerland

Received for publication 25 November 1975

Purified cholera enterotoxin was prepared by methods described by Finkel-stein and Lo Spalluto (1970). This toxin was detoxified by treatment with heatand formaldehyde. Heating cholera toxin at 60 C for 25 min resulted in theformation ofa polymer named procholeragenoid by Finkelstein et al. (1971). Theweak toxic activity of this product was removed by treatment with formalin. Noresidual toxicity could be demonstrated in formalinized procholeragenoid by therabbit ileal loop assay and the highly sensitive rabbit skin test. This toxoid wasnevertheless at least as antigenic in the rabbit as was the toxin. No reversion totoxicity was observed in vivo and in vitro at 4 C. The toxicity of formalinizedprocholeragenoid never exceeded 1/5,000 to 1/10,000 that of the toxin.

Studies in human volunteers (1) have shownthat infection with Vibrio cholerae is followedby complete protection against a second chal-lenge, but that vaccination with presentlyavailable cholera vaccines leads only to partialprotection. During their illness, cholera pa-tients develop antibodies against both the vi-brio and the enterotoxin. Vaccination withwhole-cell cholera vaccines induces only theformation of vibriocidal antibiodies (15).The protective value oftoxin-neutralizing an-

tibodies has been shown in various animalmodels (5, 11, 18).The purpose of the present study was to de-

velop a purified, antigenic cholera toxoid suffi-ciently detoxified and stable to be added to thepresently available whole-cell vaccines. Meth-ods for the in vitro production of cholera toxin (9)and for its purification (2, 9) and testing (3, 12)have been described. The usual method of de-toxification with formaldehyde leads to a prod-uct that tends to revert in vitro and in vivo totoxicity (16). Reacting cholera toxin with glu-taraldehyde (19) yields a highly detoxified andstable product, but one that is only weaklyantigenic without adjuvant (20).

Finkelstein et al. (7) have described a "natu-ral" toxoid, choleragenoid, and also a high-mo-lecular-weight polymer, designated as prochol-eragenoid, that is an intermediate in theheat-mediated conversion of cholera toxin tocholeragenoid. These two toxoids have beenshown by Fujita and Finkelstein (10) to behighly immunogenic in animal models. In ourpreliminary studies, choleragenoid could be ob-tained only in low yields, whereas procholera-genoid was transformed quantitatively frompurified choleragen. We therefore decided todevelop a toxoid based on procholeragenoid.

MATERIALS AND METHODSProduction of purified toxin. V. cholerae strain

Inaba 569B, obtained from N. K. Dutta via W. Bur-rows, was used in all these experiments. The toxinwas produced in experimental fermenters of 10 to200 liters. The syncase medium of Finkelstein andLo Spalluto (9) was inoculated with 107 viable cells/liter of an overnight culture and incubated undercontinuous stirring (550 rpm) at 30 C and aeration (1liter/liter per min). After 17 h, the temperature wasraised to 37 C and the culture was inactivated by theaddition of 1 ml of a 10% thimerosal solution/liter.The cells were removed by centrifugation at 6,000 xg for 15 min in a Sorvall RC-3 centrifuge with theHG-4 rotor (4 x 1 liter) with a radius of 9.83 inches(about 24.6 cm). The toxin was concentrated andpurified as described by Finkelstein and Lo Spalluto(9) by Al(OH)3 adsorption and elution.

Preparation of monospecific antiserum. Fiveadult male sheep and three goats were injected oncein the hind legs with 1 mg of purified cholera toxinemulsified in complete Freund adjuvant. After 31days blood samples were analyzed, and the animal(sheep no. 5911) with the lowest vibriocidal antibodytiter was bled 42 days after receiving the immuniz-ing toxin injection.Immunochemical and chemical methods. The ra-

dial immunodiffusion technique of Mancini et al.(14) was used to make quantitative cholera toxindeterminations.

Immunodiffusion assays by the Ouchterlonymethod were performed with the horse anticholera-gen SSVI/EC-3 and the sheep anticholeragen 5911.

Immunoelectrophoresis was performed in aSpinco paper electrophoresis cell on glass plates (8.5by 10 cm) with Noble agar (Difco), using a 0.01 Mbarbital buffer of pH 8.6 and 30 mA for 1 h. Antise-rum EC-3 was added to the trough.Acrylamide gel electrophoresis was performed by

the method of Ornstein and Davis (17).Protein determinations were performed with the

Lowry modification (13) of the Folin method; bovineserum albumin was used as a standard.

1692

on Novem

ber 20, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

PURIFIED ANTIGENIC CHOLERA TOXOID

Determination of toxin activities. The choleratoxin potency of various toxin preparations was de-termined by the rabbit ileal loop (12) and skin as-

says (3). Ileal loop activity was determined by themethod of Kasai and Burrows (12).The skin vascular permeability potency was esti-

mated by the limit-of-bluing titration method de-scribed by Craig et al. (3).

Detoxification of cholera toxin. Cholera toxin intris(hydroxymethyl)aminomethane-ethylenediaminetetraacetic acid (Tris-EDTA) buffer (6 mg/ml), pH7.5, was heated for 25 min at 60 C. The Tris-EDTAbuffer was then replaced with 0.1 M sodium boratebuffer, pH 8.5. Formaldehyde was added to thechilled solution to achieve a final concentration of0.2%. After 60 h of incubation at 30 C, the remain-ing formaldehyde was removed by dialysis, thebuffer was replaced by phosphate-buffered saline,and the toxoid concentration was adjusted to 1 mg ofprotein per ml.

Determination of antigenicity. Antigenicity ofvarious toxin and toxoid preparations was deter-mined by immunizing rabbits intramuscularly with1 to 100 ,g of protein per kg of body weight. Fourweeks later the rabbits received the same dose as abooster. The rabbits were then bled by heart punc-ture 4 and 6 weeks after the first injection. Theantitoxin level in sera was determined by the intra-dermal neutralization test (3). Serial dilutions ofpreviously heated sera (30 min, 56 C) were incu-bated with an equal volume of a dilution ofthe crudetoxin NIH-002, supplied by J. C. Feeley and contain-ing 1 limit-of-bluing unit per ml. In each test theprovisional standard antiserum EC-3, containing4,470 antitoxin units per ml, was titrated.

Vibriocidal activity of sera was determined by themethod of Finkelstein (4).

Determination of immunogenicity. Immunogen-icity was estimated according to Craig (2) and Rap-paport et al. (19) on the basis of the suppression ofbluing, which occurred when intramuscularly im-munized guinea pigs were challenged by intrader-mal injection of serial dilutions of toxin.

Reversion tests. In vitro reversion was estimatedby measuring the rabbit skin and ileal activity of

toxoid samples kept for 4 and 16 weeks at 4 and 37 C.In vivo reversion was estimated according to Craiget al. (3) and Rappaport et al. (19) by injecting serialdilutions of toxoid preparations intradermally intorabbit skin and then measuring and recording thediameters of induration daily during a 15-day pe-

nod.Stability. Stability of the toxins and toxoids was

estimated by measuring the antigenicity in rabbitsofsamples freshly prepared and stored for 4 weeks at4 and 37 C.

RESULTSPurification of toxin. Production of cholera

toxin in 10-liter fermenters reproduciblyyielded 35 to 40 mg of toxin per liter. The suc-

cessive steps in the purification of this crudetoxin were checked by radial immunodiffusion,protein determinations, and biological assays

(Table 1).In the first purification step, toxin was ad-

sorbed from a cell-free culture on aluminumhydroxide and eluted with Tris-EDTA buffer,pH 8.2. The eluant was concentrated on a UM20E ultrafilter and chromatographed on a col-umn of Sephadex G-75. Three distinct peaksdesignated S-1, S-2, and S-3 were obtained.Fraction S-1 consisted mainly of lipopolysac-charides and nucleic acids; proteins were dem-onstrated only in minute amounts. This frac-tion showed no enterotoxic activity but wasable to induce the production of vibriocidal an-

tibodies in rabbits (Table 2). Fraction S-2 con-tained the cholera toxin, and fraction S-3 in-cluded the almost atoxic choleragenoid (Table1).

Since preparation S-2 induced the productionof vibriocidal antibodies (Table 2), it was puri-fied further by gel filtration on a column ofagarose A-5m (8). The toxin fraction (A-2) ob-tained by this chromatography consistedmainly of cholera toxin with trace amounts of

TABLE 1. Characteristics ofpurification productsDegree of purity Skin ac-

tivity Loop ac-

Product Vol (ml) (mg/mP ) tal (mg) toxin/mgOf (Lb2d/Mg (U/mg of(mg/mi tel mgY~ oxin/mg OD2../I20 of pro-enofepro teinf rten

Culture filtrate 55,000 0.98 1,850 0.037 0.47 560 300Eluant from aluminum hydroxide con- 64 25.7 1,380 0.92 1.2 15,000 5,100

centrated on UM 20EFraction S-1 from G-75 240 0.13 0 0 0.9Fraction S-2 from G-75 concentrated on 41 29.8 1,240 1.02 1.8 49,000 6,900PM 30

Fraction S-3 from G-75 concentrated on 14 7.9 163 1.46 1.5 12 1UM 10a Measured by radial immunodiffusion.b Optical density at 280 nm/optical density at 260 nm." Lb2,,, Amount of toxin giving 4-mm-diameter increase in permeability in the rabbit skin in presence of

1 unit of antitoxin.

VOL. 13, 1976 1693

on Novem

ber 20, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

1694 GERMANIER ET AL.

TABLE 2. Level of vibriocidal antibodies in rabbitsera 4 weeks after one intramuscular immunizationwith various toxin preparations in complete Freund

adjuvant (mean values from four rabbits)Immunizing Vibriocidal

Product (toxin fractions) dosage (ugI antibody ti-kg of body ter (vU/mlrwt)

Sephadex G-75, fraction 10 1 x 107S-i 3 1 x 106

1 70,0000.3 400

Sephadex G-75, fraction S-2 100 24030 5010 203 10

Agarose, fraction A-2 100 10Formalinized procholera- 100 600genoid 10 50

" VU, Vibriocidal units; serum dilution killing50% of the vibrios.

choleragenoid. Immunization of rabbits withthis toxin (A-2) did not give rise to vibriocidalantibodies after a single application of 0.1 mg.This preparation was used as the antigen in theproduction of a monospecific antiserum andalso as the standard in quantitative toxin deter-minations by radial immunodiffusion.

In acrylamide gel electrophoresis (Fig. 1),choleragenoid (S-3) was separated in three dis-tinct bands, none of which was identical to thecholera toxin band. This phenomenon also hasbeen described by Finkelstein and Lo Spalluto(8). In our experiments, both the purified toxinpreparations S-2 and A-2 and the referencecholera toxin of Finkelstein (lot 1071, preparedunder contract for the National Institute of Al-lergy and Infectious Diseases and provided byR. Northrup) were separated in similar fashionin three bands.Production of a monospecific antiserum.

The SSVI antisera EC-3 and EC-8 were ob-tained in previous studies by repeated immuni-zation of a horse with semipurified toxin prepa-rations. Antiserum EC-3 was thoroughly ti-trated by J. P. Craig and has been proposed as aprovisional standard antitoxin by the NationalInstitutes of Health. In keeping with the immu-nization scheme, this serum contains antibod-ies against the toxin and also against varioussomatic antigens. It has been used in immuno-chemical tests to check the purity of toxin prep-arations.The monospecific antiserum SSVI no. 5911

was obtained by immunizing a sheep with 1 mgof the purified toxin preparation (A-2). Thelevel of toxin antibodies in antiserum no. 5911was two to four times higher than that in anti-serum EC-3 (Table 3). According to the immu-

nization scheme (one single injection), it had amuch lower avidity than EC-3. Vibriocidal ac-tivity could be demonstrated best up to a 1:10dilution. Whereas antisera EC-3 and EC-8 gavefour precipitin bands in Ouchterlony immuno-diffusion against a semipurified toxin, antise-rum no. 5911 exhibited only one precipitin linein this test (Fig. 2). Similar results were ob-tained in immunoelectrophoresis tests.

Detoxification of cholera toxin. The heat-mediated conversion of cholera toxin to cholera-genoid proceeds through the high-molecular-weight intermediate procholeragenoid (7). Fig-ure 3 shows the course of such a conversion.After only 2 min of exposure at 60 C, the threetoxin bands had disappeared completely. Nocholeragenoid was formed during this period.The toxin band was replaced by a diffuse bandthat penetrated the soft 3.5% gel only slowly.At 20 min this band had concentrated into adense band that failed to penetrate the separat-ing gel. After this heating period, three distinctbands appeared in the choleragenoid region. Ofthe three bands, the upper one clearly domi-nated the other two. Further heating for 45 minresulted in the complete disappearance of thehigh-molecular-weight procholeragenoid band.Figure 4 shows the results of the conversion

...; .- §

{. ..; g., ,, l

i,:12;he

--...^-. | ..-i; iX

3 l41 5..

FIG. 1. Acrylamide gel electrophoresis of alumi-num hydroxide eluant (1 and 4); fraction S-2 (toxin)(2); fraction S-3 (choleragenoid) (3); Finkelstein'sreference cholera toxin, lot 1071 (5); and referencecholeragenoid (6).

INFECT. IMMUN.

I

on Novem

ber 20, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

PURIFIED ANTIGENIC CHOLERA TOXOID

TABLE 3. Characteristics of the two toxin antisera, EC-3 and no. 5911

Antiserum Skin antitoxic activity Ileal loop antitoxic Avd Vibriocidal anti-(U/mi) activity (U/ml) vidity body titer (VU/ml)P

SSVI/EC-3 4,470" 24,000 High 100,000SSVI no. 5911 16,760 41,000 Low 10

" Standardized by J. P. Craig.b VU, Vibriocidal units; serum dilution killing 50% of the vibrios.

FIG. 2. Agar gel immunodiffusion ofsemipurifiedcholera toxin (aluminum hydroxide eluant) againstthe polyvalent SSVI antisera EC-3 (left) and EC-8(right) and against the monovalent antiserum SSVIno. 5911 (above) and the monovalent choleragenoidantiserum of Finkelstein (6) (below).

of choleragen to procholeragenoid, as demon-strated by immunoelectrophoresis. After 5 minat 60 C, the cholera toxin was converted toprocholeragenoid, which appeared as a denseprecipitin band at the top of the wells. At 15min, a new band appeared on the left; thisconsisted of choleragenoid. During the intervalfrom 5 to 25 min, a precipitin band was alsoformed in the region of the choleragen. On thebasis of the disc electrophoresis pattern (Fig. 3)as well as measurements ofthe toxin potency ofthese products, it can be concluded that thisline could not be formed by incompletely trans-formed choleragen. Formaldehyde-treated pro-choleragenoid exhibited only one precipitin linein this assay.

Procholeragenoid produced in this way (25min at 60 C) was only weakly active in therabbit skin and the ileal loop assays (Table 4).Various attempts to eliminate the remainingtoxic activity by purification failed. It must beconcluded that this remaining weak toxicitywas not due to choleragen impurities but in-stead was inherent in the procholeragenoid.The same conclusion could be drawn from thedisc electrophoretic patterns (Fig. 3). This weakresidual toxicity of procholeragenoid was abol-

1 2. 3:14 5 6

FIG. 3. Acrylamide gel (3.5%) electrophoresis ofcholera toxin (1) and oftoxin after 2 (2), 5 (3), 10 (4),20 (5), 25 (6), and 45 min (7) at 60 C.

ished by treatment with formaldehyde (Table4).

Stability of the formalinized procholerage-noid. Because of the known tendency of formal-dehyde-detoxified choleragen to revert to par-tial toxicity, the formaldehyde-treated prochol-eragenoid also was checked for stability. Solu-tions of formalinized procholeragenoid (1 mg/ml) were kept at 4 and 37 C and were assayedfor toxin potency at 1 and 4 months (Table 4).

In no case could any toxic activity be demon-strated in the ileal loop assay. Also, no rever-sion was demonstrated with the more sensitiveskin test in samples kept for as long as 4months at 4 C. Weak activity was demon-strated, however, with this assay in sampleskept for 1 month at 37 C. Further incubationfor another 3 months did not increase this activ-ity. Toxic activity offormalinized procholerage-noid was never higher than 1/5,000 to 1/10,000that of the toxin.The possibility of in vivo reversion to toxicity

VOL. 13, 1976 1695

on Novem

ber 20, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

1696 GERMANIER ET AL.

FIG. 4. Immunoelectrophoresis of cholera toxin(CT) and oftoxin after 5,15, and 25 min at 60 C. FP,Formalin-treated procholeragenoid.

was studied by making a daily record for 2weeks of induration diameters produced by an

intracutaneous injection of 3 ,ug of formalinizedprocholeragenoid. Whereas a control injectionof 3 ,ug of cholera toxin induced an induration of15- to 17-mm diameter in the 2 to 3 days afterapplication, injection of the same dose of for-malin-treated procholeragenoid produced a

small induration zone that never reached the 7-mm limit.

Antigenicity of formalinized procholerage-noid. By intramuscular immunization of rab-bits with toxin and various toxoids in doses of 1to 100 ,ug per kg of body weight, the potency ofthese preparations to elicit the formation oftoxin-neutralizing antibodies was estimated(Table 5). A good immune response could beobtained with as little as 1 ,ug/kg. With dosesfrom 1 to 100 ,ug/kg, a good correlation existedbetween immunization dose and antibody titer.The toxoids choleragenoid and procholerage-noid were not significantly less antigenic, andthe formaldehyde treatment did not reduce theantigenic potency of the procholeragenoid. In-jection of a booster dose 4 weeks after the pri-mary immunization resulted in a 10- to 20-foldincrease in serum antitoxin level.

Storing the formalinized procholeragenoid at4 and 37 C for 4 weeks in solutions of 1 mg per

ml of phosphate-buffered saline (0.15 M, pH7.4) did not significantly weaken the antigenicpotency.Similar results were observed when the im-

munogenicity of formalinized procholeragenoidwas estimated in guinea pigs. Two groups ofanimals that were immunized with 1 and 10 ,tgof toxoid followed by a booster with the same

dose after 4 weeks needed a 100- to 300-foldlarger dose of toxin to produce an 8-mm bluinglesion (Table 6).

After immunizing rabbits with 100 ,.g of for-malinized procholeragenoid, a weak but dis-tinct rise in vibriocidal antibodies was observed(Table 2). Since the procholeragenoid used hadbeen prepared from the toxin preparation S-2,this was not surprising. Obviously, the somaticantigens contained in this material were inacti-vated neither by heat (25 min at 60 C) nor byformaldehyde treatment.

DISCUSSIONFinkelstein (6) has reported the isolation of

the cholera enterotoxin and of its "natural" tox-oid, choleragenoid. Our semipurified choleratoxin preparations also contained importantamounts of choleragenoid. This toxoid, whichcould not be demonstrated in culture filtrates,was formed in the course of toxin purificationunder certain circumstances such as aluminumhydroxide adsorption and elution. The conver-sion of choleratoxin to choleragenoid also couldbe produced under well-controlled conditions byheating at 60 C. In this process, the intermedi-ary product procholeragenoid was also formed.By limiting the heating to between 20 and 25min, an almost quantitative conversion of chol-eragen to procholeragenoid was attained. Discelectrophoretic patterns indicate that the for-mation of the procholeragenoid proceedsthrough stages of intermediate aggregation.Heating for 60 min resulted in complete conver-sion of procholeragenoid to choleragenoid.Reversion of choleragenoid or procholerage-

noid to higher toxicity was not observed. Theresidual toxicity of procholeragenoid was, likethat of choleragenoid, reduced to at least 1/1,000 to 1/5,000 that of toxin. In the develop-ment of a toxoid vaccine, such a detoxificationwould not be sufficient. On the basis of resultsobtained by Craig et al. (3) in human volun-teers, it can be estimated that parenteral appli-

TABLE 4. Toxic activity of cholera toxin,procholeragenoid, and formalinized

procholeragenoid

Skin activity Ileal loop ac-Product (Lb tivity (U/(Lb2omg)a mg)

Cholera toxin (S-2) 49,000 6,900Procholeragenoid 50 15Formalinized procholera- <1 <1genoid4 weeks at 4 C <1 <14 weeks at 37 C 9 <116 weeks at 4 C <116 weeks at 37 C 3

a Lb20 expressed as in Table 1.

INFECT. IMMUN.

on Novem

ber 20, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

PURIFIED ANTIGENIC CHOLERA TOXOID

TABLE 5. Antigenicity of cholera toxin and toxoids in rabbits

Antitoxic units/ml at:a

Antigen Immunizationdose (,ug/kg) 3 weeks 4 weeks 2 weeks after boos-

ter-b

Choleragen (S-2) 1 2 710 26 77

100 203 350Choleragenoid (S-3) 100 90 205Procholeragenoid 1 3 20

10 20 75100 110 210

Formalinized procholeragenoid 1 2 13 265c10 18 82 950

100 250 480 7,2504 weeks at 4 C 1 6

10 23100 500

4 weeks at 37 C 1 510 50

100 184

a Average from two or three experiments.b Some animals were boostered 4 weeks after the first immunization with the same doses.c Booster with solutions freshly prepared from lyophilized toxoid.

TABLE 6. Inhibition of skin bluing reaction inguinea pigs after immunization with formalinized

procholeragenoid

Fold in-

Immrnu- crease oftoxin re-

Antigen gdose quired to

(ILg)a produce 8-doe

mm bluing

lesionb

Formalinized procholerage- 1 75noid 10 375

Formalinized procholerage- 1 104noid + 0.2% aluminum hy- 10 235droxide, pH 7.0a Two injections at 4-week interval.h Toxin injected intradermally 2 weeks after sec-

ond immunization.

cation of procholeragenoid in doses required togive a proper immune response would elicitsome local reactions. Various attempts to fur-ther purify the procholeragenoid, which can beaccomplished easily because of its high molecu-lar weight, showed that the residual toxicity ofprocholeragenoid is not due to either toxin im-purities or partial reversion to the toxin. Pro-choleragenoid thus had to be detoxified furtherin a second step by treatment with formalde-hyde. Formalinized procholeragenoid was sostrongly detoxified that activity could not bedemonstrated even in the highly sensitive rab-bit skin test. Storage for 4 weeks at 4 C did notproduce reversion to higher toxicity. After stor-age for 4 weeks at 37 C, a weak toxic activity

was demonstrated in the skin test; however,prolonged storage at 37 C did not result in anincrease of this toxicity.The antigenicity of the cholera toxin was not

reduced by the detoxification procedures. Intra-muscular injection of 1 to 100 ,g offormalinizedprocholeragenoid induced a dose-dependentproduction of toxin-neutralizing antibodiescomparable to that induced by toxin. Adminis-tration of a booster dose resulted in an impor-tant rise in serum antibody level. Furthermore,it was shown that immunization with fonnalin-ized procholeragenoid can protect guinea pigsagainst a subsequent toxin-induced skin reac-tion.

Besides a high titer of toxin-neutralizing an-tibodies, the toxoid described in this paper alsoinduces, on a much lower level, the productionof vibriocidal antibodies. We would like to em-phasize that the purer toxin preparation A-2can be used at any time to prepare, by toxoida-tion or by gel filtration of the procholeragenoid,a toxoid free of somatic antigens.

LITERATURE CITED1. Cash, R. A., S. I. Music, J. P. Libonati, J. P. Craig, N.

F. Pierce, and R. B. Hornick. 1974. Response of manto infection with Vibrio cholerae. II. Protection fromillness afforded by previous disease and vaccine. J.Infect. Dis. 130:325-333.

2. Craig, J. P. 1971. Cholera toxins, p. 189-254. In S.Kadis, T. C. Montie, and S. J. Ajl (ed.), Microbialtoxins, vol. IIA. Academic Press Inc., New York.

3. Craig, J. P., E. R. Eichner, and R. B. Hornick. 1972.Cutaneous responses to cholera skin toxin in man. I.Responses in unimmunized American males. J. In-fect. Dis. 125:203-215.

VOL. 13, 1976 1697

on Novem

ber 20, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

1698 GERMANIER ET AL.

4. Finkelstein, R. A. 1962. Vibriocidal antibody inhibition(VAI) analysis. J. Immunol. 89:264-271.

5. Finkelstein, R. A. 1970. Antitoxic immunity in experi-mental cholera. Infect. Immun. 1:464-467.

6. Finkelstein, R. A. 1970. Monospecific equine antiserumagainst cholera exo-enterotoxin. Infect. Immun.2:691-697.

7. Finkelstein, R. A., K. Fujita, and J. J. Lo Spalluto.1971. Procholeragenoid: an aggregated intermediatein the formation of choleragenoid. J. Immunol.107:1043-1051.

8. Finkelstein, R. A., and J. J. Lo Spalluto. 1969. Patho-genesis of experimental cholera. J. Exp. Med.130:185-202.

9. Finkelstein, R. A., and J. J. Lo Spalluto. 1970. Produc-tion of highly purified choleragen and choleragenoid.J. Infect. Dis. 121(Suppl.):S63-S72.

10. Fujita, K., and R. A. Finkelstein. 1972. Antitoxic im-munity in experimental cholera. J. Infect. Dis.125:647-655.

11. Holmgren, J., A. Anderson, G. Wallerstrom, and 0.

Ouchterlony. 1972. Experimental studies on choleraimmunization. II. Evidence for protective antitoxicimmunity mediated by serum antibodies as well as

local antibodies. Infect. Immun. 5:662-667.12. Kasai, G. J., and W. Burrows. 1956. The titration of

cholera toxin and antitoxin in the rabbit ileal loop. J.Infect. Dis. 116:606-614.

13. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.

INFECT. IMMUN.

Randall. 1951. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

14. Mancini, G., A. 0. Carbonara, and J. F. Heremans.1965. Immunochemical quantitation of antigens bysingle radial immunodiffusion. Immunochemistry2:235-254.

15. Mosley, W. H. 1969. The role of immunity in cholera.Tex. Rep. Biol. Med. 27:227-241.

16. Northrup, R. S., and F. V. Chisari. 1972. Response ofmonkeys to immunization with cholera toxoid, toxinand vaccine. J. Infect. Dis. 125:471-479.

17. Ornstein, L., and B. J. Davis. 1962. Disc electrophore-sis. Distillation Products Industries, Eastman KodakCo., Rochester, N. Y.

18. Pierce, N. F., E. A. Kaniecki, and R. S. Northrup.1972. Protection against experimental cholera by an-titoxin. J. Infect. Dis. 126:606-616.

19. Rappaport, R. S., G. Bonde, T. McCann, B. Rubin, andH. Tint. 1974. Development of a purified cholera tox-oid. II. Preparation of a stable, antigenic toxoid byreaction of purified toxin with glutaraldehyde. Infect.Immun. 9:304-317.

20. Saletti, M., and A. Ricci. 1974. Experiments with chol-era toxin detoxified with glutaraldehyde. Bull.W.H.O. 51:633-639.

21. Spyrides, G. J., and J. C. Feeley. 1970. Concentrationand purification of cholera exotoxin by adsorption onaluminum compound gels. J. Infect. Dis.121(Suppl.):S96-S99.

on Novem

ber 20, 2020 by guesthttp://iai.asm

.org/D

ownloaded from