-Archives of Medical Research Volume 29, No. 3, pp. 231-234,1998

Printed in Mexico

Original Article

Intestinal Colonization of the Infant Mouse Model

By Attenuated and Virulent Vibrio cholerae Strains

BARBARA CEDRE MARRERO, LUIS GUILLERMO GARCIA IMIA,HILDA MARIA GARCIA SANCHEZ, MILDREY FARIÑAS MEDINA,ARTURO TALAVERA CORONEL and JUAN FRANCISCO INFANTE BOURZAC

Fin/ay lnstitute, Center of Research and Production of Sera and Vaccine, Havana, Cuba

Received for publication March 5, 1997;acepted June 2, 1998 (97/032).

Abstract

Background: Intestinal colonization ofhumans withvirulent Vibrio cholerae stimulates substantial,lasting immunity against reinfection. The purposeofthis study was to evaluate thecolonizing capabilityofvarious Vibriocholeraestrains which are promisingcandidates to oral vaccine.Methods: Infant mouse model modification wasused. In order to standardize the method, severalparameters were tested, such as culture mediumand optimal time of incubation and appropriatenumber of cells to be inoculated. The following weretested: Vibrio cholerae strain 81, 413, and 251A,which were obtained at the Molecular BiologyDepartment of the National Center for Scientific

Research, Ravana, Cuba. Their virulence cassetteswhich code for the main virulence factors weredeleted.Results: Good variancecoefficient (VC)wasobtainedin repeated experiments. The colonizing propertiesof attenuated Vibrio cholerae strains evaluated bythis method correlated well with those observed forparental strains.Conclusions: Genetically attenuated Vibrio cholerastrains have the same intestinal colonization level astheir parental strains in the infant mouse model;thus, genetic manipulation does not affect genes thatencode for the synthesis of colonization factors.(Arch Med Res 1998; 29:231)

KEY WORDS: Colonization; Vibrio cholerae; Infant mouse modelo

~

~ Introduction

Sma11intestine colonization by Vibriocholeraeisoneof the most important steps in the pathogenic mecha-nisms of the cholera disease (1-3). In order to obtain asafe and effective vaccine against cholera, it is essentialto achieve the optimal expression of the factors thatensure adherence and multiplication of bacteria in the

Correspondence to:Bárbara Cedré Marrero, Finlay Institute, Center for Sera and Vac-cines, Ave.27No. 19805,LaLisa, Apartado Postal 16017, 11600LaHabana, Cuba. Tel: (53)7218176; FAX: (53)7 33 6075; E-mail:finlayci@ceniai.cu

small intestine, thus stimulating a protective and long-lasting irnmune response.

Vibrio choleraeEl Tor and Classicbiotypeproducetwo colonization factors, MSHA and TCP (4). Theinactivation of these protein genes reduces the cellcolonization capacity in the "infant mouse model."

Different studies have demonstrated that bacterialcolonization is the most important aspect in the irnmu-nogenicity of the microorganism (5,6). On the otherhand, bacterial virulence isc10selyrelatedwith its abilityto adhere and multiply in the small intestine (4).

The infant mouse model was studied in differentconditions in order to evaluate the colonization capabil-

231

232

ity of some genetically attenuated Vibrio cholerae stra-lOS.

The purpose of this study was to evaluate the coloniz-ing capability of various Vibrio cholerae strains whichare strong candidates to oral vaccine. This paper de-scribes the methodology used and the results of studieson colonization of genetically attenuated strains of Vibriocholerae with reference to their parental virulento Theresults of this study have shown that these strains main-tain the colonization capacity and the possibility to elicitan antibacterial immune response despite genetic ma-nipulations.

Materials and Methods

Animals. Intestinal colonization assay was done ininbred Balb/c 2- to 4-day-old mice, with weights rang-ing from 1.5 - 2 g.

Bacteria. The sources and characteristics of the Vibrio

cholerae strains used in this study are shown in Table l.Organisms were maintained at -70°C in 10% skim milkcontaining 20% glycerol.

Media. In the first experiment, C6706 and 413 strainswere grownin Tryptone Soya Broth (TSB)(OXOID) andin Colonization Factors Broth (CFB) in order to selectthe best media for the subsequent assays. Cultures wereincubated at 37°C, 200 rpm, for 6 h.

Inoculum Size. The inoculum was administered byoral route in cellular suspensions with 103, 104 and 105bacteria in 50 /lL of phosphate buffered saline (PBS) pH7.4 containing 0.01% Evans Blue dye using an eyeneedle and syringe. Six animals per dose were used.

Optimal Incubation Time. After inoculation with 105cells, the animals were divided into groups and weresacrificed at 6, 12, 18 and 24 h.

Table1SourcesandDescriptionsof the VibriocholeraeStrainsUsedin

ThisStudy

StrainC7258*C6706*81413SG251*251A569B*

Description

Wild type, El Tor, Ogawa, Pero 1991

Wild type, El Tor, Inaba, Pero 1991

ctxA- ctxB- zor ace' orfU- cep-

ctxA" ctxB- zor ace- orfU- cepo

Wild type, 0139 serogroup.

ctxA" ctxB- zor ace- orfU. cep-

Wild type al, Classical, Inaba

Source

7

7

8

8

9

8

10

* Provided by Dr. Richard A. Finkelstein, Department ofMolecular Microbiology and Immunology, School of Medicine,University of Missouri, Columbia, MO, USA.

CEDRE, GARCIA, GARCIA, ETAL.

In all cases, groups of six mice were used to test eachone of the parameters, and the experiments were re-peated three times.

Colonization Assay. The method is a modificationreported by others (11,12). This assay was made usingthe best parameters as determined in the experimentsdescribed above. Mice were separatedfromtheir parentsthe day before, in order for them to have their stomachempty at the time of challenge. Fifty microliters of ce-llular suspensions were given by oral route with an eyeneedle and a syringe. After the incubation time, theanimals were sacrificed, the small intestines removedand weighed in groups of ten, and then washed. Theintestines were homogenized in PBS at 13,500rpm for15 sec in an Ultraturrax T 25 lKA (Labortechnik,Beme,Switzerland). R:

Determination of Bacteria Counts. Thecolonizingproperties of Vibrio cholerae were assayed by countingviable vibrios in the intestinal wall after it had been c---..washed.Numbersof bacteriarecoverablefrom mice --------

were determined by homogenization in PBS of pooledintestines from ten mice in each group. Suitable dilutionsin PBS of the homogenates were plated on Thiosulphate-Citrate-Bile salt-Sucrose (TCBS) (OXOID) agar forviable counting, and were cultivated for 24 h. The resultswere expressed as CFU/g of intestine.

Analysis of variance was used for comparing the indexof colonizing of attenuated and virulent strains.

Results

Vibrio cholerae C6706 and 413 strains were chosen inorder to select the best culture medium, the formerbecause it is a wild-type strain, and the latter because itis a manipulated strain, comparing their behavior in bothmedia. Figure 1shows the comparison between TSB and ~

IU CFB Medium

CFU /g1.00E+10

.TSB Medium

1.00E+08

1.00E+06

1.00E+04

1.00E+02

1.00E+00

6706413 Strains

Figure 1. Colonization index of two V. cho/erae strains in theinfant mouse model cultured in TSB and CFB media.

,VIBRIO CHOLERAE COLONIZA TION IN INFANT MOUSE

.. .1.00E+03IJ 1.00E+04

D1.00E+05CFU/g1.00E+09

1.00E+06

1.00E+03

1.00E+O

5698 C7258 Strains

Figure 2. Evaluation of different inoculumsizes inthe infant

~mouse mode!. CeUswere grown in TSB medium.

CFB media. There is no difference in CFU/g of intestine,regardless of the media used to culture the ce11s.On theother hand, the results were very similar for both strains.

The experiment to select the inoculum size is summa-~;zed in Figure 2. The experiment was repeated three

times and, in a11cases, the highest colonization indexwas obtained with 105CFU for the strains used. In thiscase, two strains were chosen, a highly virulent, C7258,and a strain with a low colonization index, 569B strain.Higher doses were not assayed because they constitutethe 50% lethal dose (LD 50)for both of these strains (datanot shown). Figure 3 presents the results for three differ-ent strains and four incubation times. For all strains, the24-h period from inoculation to animal sacrifice showedthe highest values of CFU/g of intestine. Other periodsof time were not proved, because at that time, a11strainswere capable of colonizing small intestine.

The results of the above studies have a110wed us to

determine the best parameters and design the experimentto compare the colonization index of different Vibrio

~holerae strains (Figure 4). The strains tested showed,imilar colonization indexes, except Vibrio cholerae

CFU/g

~. 1E+11

/~

!:::!6h 012h

1E+07

1000

0.1

C 7258 81569B Strains

Figure 3. Kinetics of colonization in the infant mouse model of11.cho/erae strains. Animals were sacrificed at different timesafter inoculation of 105ceUs.

233

CFU/g1.00E+11

1.00E+10

1.00E+09

1.00E+08

1.00E+07

1.00E+0600 _ ~ M m ~ m~ m R ; m ~ ~~ ~ ~ ~ N

o o rJJ Strains

Figure 4. Colonization index of different 11.cho/erae strains ininfant mouse mode!. Groups of 10 animals were inoculatedby oral route with viable 11.cholerae strains and sacrificed 24h later.

Classical Inaba 569B strain, with low variation coeffi-cient. Variation coefficients (VC) expressed in percent(%) of the studied strains in this animal model in thisstudy were from 1.43 - 6.63 for transformed strains and2.42 - 8.04 in the epidemic strains.

Discussion

Vibrio cholerae is a non-invasive pathogen, and con-sequently, adherence ano colonization are critical fac-tors in the disease process. The infant mouse model hasbeen widely used in the evaluation of the virulence andcolonization capacity of Vibrio cholerae (11-13).

MSHA is one of the colonization factors in Vibriocholerae El Tor strains, and other authors have obtainedthe best expression levels of this protein using TSB as aculture medium (14). In addition to MSHA, other anti-gens and colonization factors are expressed in CFB (15).TSB and CFB were used to grow the vibrios. Vibrioscultured in both media showed the sarne adhesive prop-erties. The TSB medium was selected because it is easier

to prepare and has a high yield in biomass production.To determine the optimal inoculum size, 103, 104and

105 CFU of the we11-known Vibrio cholerae strains

C7258 and 569B were inoculated. The former is a veryvirulent strain isolated from a cholera patient in Peru,and several reports have been done on the poor coloni-zation capability of the 569B strain. The latter showsdiminished intestinal colonization in humans (16), mice(17) and rabbits (18). In 24 h, both strains had colonizedthe sma11 intestine of infant mice but with significantdifferences (p <0.025).

Vibrio cholerae 81 and 569B strains were unable tocolonize the intestine both at 6 and 12h after inoculation,the last one even at 18 h. Only 24 h after the inoculumadministration, a11strains achieved measurable levels ofintestinal colonization.

234

Vibrio cholerae C7258 and C6706 strains, isolatedfrom cholera patients in Pern, as well as SG 251, in-creased their cellular concentration in the intestine, 1Q4and lOstimes with respect to the initial inoculum. Simi-lar results were obtained with attenuated strains 81,413,251A, produced in the Molecular Biology Departmentof the National Center for Scientific Research (8).Thesestrains have proven to be avirnlent in the Rabbit IlealLoop model, and have LDsovalues between 102and 103higher than the virulent parental strains (8). On the otherhand, the 569B strain showed a very low colonizationlevel, with an increase of only 100times the cell numberper gram of intestine.

Analysis of variance indicates the non-existence ofsignificantdifferencesbetweenparentalepidemicstrainsand their transformer strains.

Even though references conceming the precision ofthe model used do not appear in the literature in general,calculated variation coefficients are lower than thoseaccepted, even for irnmunoenzymatic assays, which arereported to be lower than 10 - 11% (19,20).

The fact that the parental strains and the attenuatedstrains present similar colonization indexes is in agree-ment with the value expected, because in the genetictransformation, the genes that regulate the adherencecapability were not altered in the study.

The results obtained conceming the VC indicate thatthis model presents an acceptable precision, and is ableto ',distinguish among strains of different colonizingcapabilities.

The principal conclusion of this work is that under theexperimental conditions used here, genetically attenu-ated Vibrio cholerae strains have the same intestinalcolonization level as their parental strains in the infantmouse model. It seems that genetic manipulation doesnot affect genes that encode for the synthesis of coloni-zation factors.

Acknowledgments

The authors wish to express appreciation to JorgeBeDÍtezfor providing attenuated strains, and to Dr. Gus-tavo Sierra for critical review of the results of this work.

References

1. Beachey EH. Bactererial adherence: adhesin-receptor interac-tions mediating the attachment of bacteria to mucosal surfaces.J Infect Dis 1981; 143:325.

2. Srivastava R, Sinha VB, Srivasta BS. Events in the pathogenesisof experimental cholera: role ofbacteria adherence and multipli-cation. J Med Microbiol1980; 13:l.

3. Finkelstein RA. Cholera. In: Germanier R, editor. BacterialVaccine. Orlando, FL: Academic Press; 1984. p. 107.

CEDRE, GARCIA, GARCIA, ET AL.

4. Osek J, Jonson G, Svennerholm A-M, Holmgren J. Role ofantibodies againstbiotype-specitic Vibriocholeraepili in protec-tion against experimental classical and El Tor cholera. InfectImmun 1994; 62:2901.

5. Neoh SH, Rowley D. The antigens of Vibriocholerae involvedin thevibriocidalactionof antibodyand complementoJ InfectDis1970; 121:503.

6. Jonson G, Holmgren J, Svennerholm A-M. Analysis of expres-sion of toxin-coregulated pili in Classical and El Tor Vibriocholerae 01 in vitro and in vivo. Infect Immun 1992; 60:4278.

7. Hase CC, Thai LS, Boesman-Finkelstein M, Mar VL, BurnetteWN, Kaslow HR, Stevens LA, Moss J, Finkelstein RA. Con-struction and characterization of recombinant Vibrio cholerae

strains producing inactive cholera toxin analogs. Infect Immun1994; 62:3051.

8. Benítez J, SilvaA, Rodríguez BL, Fando R, Campos J, RobertA,García HM, García L, Pérez JL, Oliva R,. Torres CA, Ledón T.

Geneticmanipulationof Vibriocholeraefor vaccinedevelop:, ,ment: construction of live attenuated El Tor candidate vaccinlstrains. Arch Med Res 1996; 27:275.

9. Bhadra RK, Roychomdhury S, Eanerjee RK, Kar S, MajumdarR, Sengupta S, Chatterjee S, Khetawal G, Des J. Cholera toxin(ctx) genetic element in Vibrio cholerae 0139. Microbiology1995; 14:1977. ~

10. Finkelstein RA, LoSpalluto JJ. Pathogenesis of experimentalcholera: preparation and isolation of choleragen andcholeragenoid. J Exp Med 1969; 130:185.

11. Dutta NK, Habbu MK. Experimental cholera in infant rabbits: amethod for hemotherapeutic investigation. BrJ Pharmacol1955;10:153.

12. Ujiiye A, Nakatomi M, Utsunomiya A, Mitsui K, Sogame S,Iwanaga M, Kobari K. Experimental cholera in mice. 1. Firstreport on the oral infection. Trop Med 1968; 10:65.

13. Baselski V, Briggs R, Parker C. Intestinal fluid accumulationinduced by oral challenge with Vibriocholerae or cholera toxinin infant mice. Infect Immun 1977; 15:704.

14. Jonson G, SanchezJ, SvennerholmA-M. Expression and differ-ent biotype-associated cell-bond haemagglutinins of Vibriocholerae 01. J Gen Microbiol1989; 135:111.

15. Evans DG, Evans DJ Jr, Tjoa W. Haemagglutination ofhumangroup A erythrocytes by enterotoxigenic Escherichia coli iso-lated from adults with diarrhea: correlation with colonizati01----factors. Infect Immun 1977; 18:330.

16. Ketley 1M, Kaper 18, Herrington D, Losonsky G, Levine MM.Diminished irnmunogenicity of a recombination-deticient de-rivative of Vibrio cholerae vaccine strain CVD 103. InfectIrnmun 1990;58:1481. ~

17. Levine MM, Kaper 18, Herrington D, Ketley J, Losonsky G;Tacket CO,Tall B, Cryz S. Safety, irnmunogenicityand efficacyof recombinant live oral cholera vaccines, CVD 1()3and CVD1()3HgR. Lancet 1988; 2:467.

18. Pierce NF, Kaper 18, Mekalanos JJ, Cray WC. Role of choleratoxin in enteric colonization by Vibrio cholerae 01 in rabbits.Infect Immun 1985; 50:813.

19. Thielman K. Principios metodológicos en Bioquímica Clínica.Havana: Editorial Instituto del Libro; 1973.

20. Tijsjen P. Practice and theory of enzyme immunoassays.Amsterdam: Elsevier; 1985, ch. 15.