Scand. J. Immunol, 31, 361-366, 1990

Low T- and B-Cell Reactivity is an ApparentlyParadoxical Request for MurineImmunoprotection Against Streptococcus mutansMurine Protection can be Achieved by Immunization Against aB-Cell Mitogen Produced by these Bacteria

R, SOARES. p. FERREIRA. M. M G. SANTAREM,M. TEIXEIRA DA SILVA & M. ARALA-CHAVES

Departmetit of Immunology, Institute for Biomedica! Sciences Abel Salazar. andLaboralory of Microbiology, Centre for Experimental Cytology (INIC). Porto. Portugal

Soares. R.. Ferreira, P., Santarem, M.M.O.. Teixeira da Silva, M. & Arala-Chaves, M. Low T-and B-Cell Reaetivity is an Apparently Paradoxical Request for Murine ImmunoprolectionAgainst Streptococcus mutans. I. Murine Proteetionean be Aehieved by Immtinization Against aB-Cell Mitogen Produeed by these Bacteria. Scund. J. Immunol. 31, 361-366, 1990

C57BL/6 mice thymectomi7ed as adults or depleted of CD4 ' cells were much less susceptible thanintact conventional mice to the B-cell mitogenic and specific immunosuppressive effects of aprotein designated as F5'EP-Sm secreted by Streptococcus mutans. These mice were alsoeonsiderably more resistant to infection by these baeteria than intaet individuals. Theimmunosuppressor effect of F5'EP-Stn was also abrogated, however, m conventional intact tnicewhen immunized intraperitoneally against heat-inactivated F5'EP-Sm. On the other hand.resistance to bacterial infection could be achieved by immunization of conventional intactC57BL/6 mice against heai-inactivated F5'EP-Sm by intraperitoneal or intradermal routes evenwhen the animals were infected 3 tnonths after immunization and even when the immunizationprocedure did not include Freund"s adjuvant, which was the case with the intradermal route.Interestingly, the protection against the bacterial infeetion was aeeompanied by only a minorincrease in specific serum antibodies against F5'EP-Sm.

These results are discussed in the context of adequate strategies for immunoprotection againstStreptococcus mutans and other micro-organisms whieh are secretors of substances ihat shareboth B-cell mitogenic and immunosuppressive properties and which are thus able to suppress theimmune response by overstimulation of the immune system of the hosl.

Mario Arala-Chaves, Department of Immunology. Instituto de Ciencieas Biomedicas Abel Salazar.Largo da Escola Medica. 4000, Porto, Portugal

Non-specific B-ccl! mitogenicity has beenascribed to bacteria such as Brucella melitensis.Streptococcus intermedius and Streptococcusmutans [2.7.22]. lo viruses such as dchydrogcnaseand African swine fever virus (ASFV) [3. 5], andto parasites such as Trypanosoma cruzi andLcishmania [10, 12, 17].

We have previously shown that in the cases ofStreptococcus Intermedius, Streptococcus mtitans,and ASFV the B-cell mitogenicity is due toproteins sharing immunosuppressive propertiesthat are secreted by these micro-organisms into

the culture medium [2. 3, 7]. At least in the case ofthe protein secreted by the ASFV. both B-cellmitogenicity and specific immunosuppressiveproperties are abrogated by T-cell depletion [16].

We have also shown that the immunosuppres-sive/B-cell mitogenic protein secreted by S,»»//«n.v (designated as F5'EP-Sm) plays an impor-tant role in the survival of the baeterium in thehost [18]. Aitning at protecting the host againstthis bacterium we decided therefore to study theconditions in which the biological effects of F5'-EP-Sm were neutralized in vivo.

361

362 R, Soares et al.

M A T E R I A L S A N D M E T H O D S

Mice. C57BL/6 male mice 6 8-weeks old, bred at theGulbenkian Institute of Science were used.

In vivo treatments for lymphocyte depletion. Groups ofmice were thymectomized al 4 6 weeks of age bysuction of the Ihymic lobes under ether anaesthesia.Depletion of T-helper cells was accomplished by usingrat monoclonal antibody anti-CD4(GK 1.5) [6| (a kindgift from Dr P. Minoprio, Pasteur Institute, Paris).Anti-CD4 antibody was administered inlraperitoneallyin doses of 10 fig according to difTereni schemes (seebelow). Each dose has previously been shown [ 15], andconfirmed by us. to result in functional inactivalion andremoval of more than 9O'Xi of CD4^ cells.

S, mutans and preparation of bacterial products, S.mutans E49 was obtained from the American TypeCullure Collection (ATCC) and was cultured on choco-late agar medium for 24 h under anaerobic conditionsat37 C. For the purposesof the preparation ofbacterialproducts, 7 X 10** micro-organisms/I of 5. mutans weresubsequently inoculated into Shacdier broth medium(Biomerieux. Marcy-FEtoile. Charbonieres les Bains,Franue) and cullured for 2 days also under anaerobicconditions and at the same temperature as above.F5'EP-Sm was obtained from the supernatants of thecultures after successive fractionation by ion-exchangechromatography and double-preparative isoelectro-focusing, as described in detail previously [18].

5 50

10

FIG. I. Primary immune response againsl SRBC ofF5'EP-Sm-treated C57BL,.6 mice non-immunized (•)or previously immunized intraperitoneullyand boosted.15 days later, with heat-inactivated F5'EP-Sm (H). Theresults are expressed as percentage of the control(F5'EP-Sni-untreated mice primed with SRBC) and aremeans and one standard deviation (SD). !n these and inthe foliowitig figures the results were calculated on thebasis of quadruplicate samples for both experimentaland control animals and represent typical experiments.The mean-(-l SD of the numbers of IgM-secretingdirect spleen PFC of ihe control mice, expressing theprimary immune response to SRBC, were 11003 + 995per 10̂ spleen cells.

Inactivation of F5'EP-Sm. Both immunosuppressiveand mitogenic properties of F5'EP-Sm were inactivatedby heating the protein at 70 C for 30 min.

Immunization protocol against E5'EP-Sm. Conven-tional C57BL/'6 mice were immunized wilh F5'EP-Smeither inlraperitoneally (i.p.) or intradermally. Miceimmunized inlraperitoneally received 0.5 /ig of heat-inactivated F5'EP-Stn dissolved in phosphate-bufleredsaline (PBS) three times per week for 3 weeks. On firstadministration the product was mixed with completeFreund's adjuvant and on the fourth and seventhadministrations with incomplete Freund's adjuvant.Mice immunized intradermally received 1 ̂ g of heat-inactivated F5'EP-Sm three times per week, for 4 weeks.Some of the mice immunized intraperitoneally wereboosted 15 days after immunization. The remainingmice were boosted 90 days later. All boosts consisted ofone single administration of the same amount of F ^ EP-Sm used for each immunization protocol. All immu-nized mice were challenged with .V. mutans or withF5'EP-Sm ! week after boost. Controls of mice immu-nized intraperitoneally and intradermally consisted olC57BL/6 mice, of the same age, treated with PBS andcomplete Freund's adjuvant or PBS alone respectively.

Infection with S. mutans. Mice were infected intraper-itoneally with 2.5 X 10'* streptococci. Three days laterthe mice were killed and the peritoneal exudate wascollected with 4 ml of PBS. The number of streptococciin the peritoneal cavity was determined in Gram-stainedcytospin preparations of the peritoneal exudate bycounting the bacteria and the leucocytes; the actualnumber of streptococci per peritoneal cavity wasobtained by extrapolation from the ratio of streptococcito leucocytes, taking into account the total number ofleucocytes determined with an electronic cell counter.

50

25

15 -

5 -

FIG. 2. Comparison of the suppressor effects on theprimary immune response against SRBC induced byF5'EP-Sm in conventional C57BL/6 mice (•) orC57BL/6 mice depleted of CD4 * eells (H) or thymecto-mized (•). The results are expressed as percentage of thecontrols (F5'EP-Sm-untre:ited C57BL;'6 mice). Controlvalues of the primary immune response were the same asthose in Fig. 1.

Immunoprotection against Streptococcus mutans 363

As.say of mice anti-E5'EP-Sm antihody. We measuredthe titre of anti-F5'EP-Sm antibodies in the sera ofC57BL/6 mice using enzyme-linked immunosorbentassay (ELIS.A) described elsewhere [14]. Briefiy 5-mmpolystyrene niicrotitre plates (Nunc. DK-4000, Ros-kilde, Denmark) were coated with 10 //g of F5'EP-Smovernight at 4 Cand then saturated for 30 min with \"A.gelatine in PBS at room temperature. Appropriate serialdilutions of the serum samples were then ineubated inthe plates for 2 h at room temperature. After washing,the bound antibodies were revealed by the addition ofperoxidase-labelled goat anti-mouse total Ig antibodies(Southern Biotechnology Associates, I60A OxmoorBoulevard, Birmingham. Ala., USA). Orthophenylcne-diamine (Sigma Chemical Company, St Louis Miss.,USA; 6.1178) and HjO: were used to develop thereaction, and the colorimetrie change was measuredwith a Biotek Chromoscan at 450 nm.

Plaiiue-forming cell (PTC) assays. Splenic produc-tion of specific anti-sheep red blood eells (SRBC) IgMantibody production was evaluated by direct haemoly-tic PFC assay [4]. Non-specific Ig production wasassessed by the protein A PFC assay [9]. Rabbit antiserafor total Ig or specific for each isotype of mouse Ig wereused as developing antibodies and were previously

characterized in detail [8]. Each experiment was per-formed in quadruplicate samples for both F5'EP-Sm-treated mice and corresponding controls.

Immunization wilh SRBC and imniuno.suppre.s.Kionprotocol. Mice were immunized with an intraperitonealinjection of 5 x 10' SRBC, and the splenic direct andindirect PFC responses assayed 5 days later. Treatmentswith F5'EP-Sm were carried out by intraperitonealinjection of 4 units of biological activity (UBA) 2 daysbefore priming with SRBC. A UBA was defined as theamouniofF5 EP-Sm that reduced by 50",, the numbersof direct haemolytic splenic PFC of treated animals.

RESULTS

Immunosuppression and B-cell polyclonal acti-vation by F5'EP-Sm

The suppressor effect of F5'EP-Sm on theprimary immune response of C57BL/6 miceagainst SRBC was reduced when this protein wasassayed in mice immunized intraperitoneally andboosled 15 days later (Fig. 1). Surprisingly,

•5 25

10

T

1 igG 2a IqG 2b 3

FIG. 3. Spleen non-specific Ig production of the indicated isotypes 5 days afterF5'EP-Sm treatment of SRBC-unprimed C57BL/6 mice either depleted ofCD4*^ cells (0) or thymectomized (•). Results are means and one SD of non-specific Ig-secreting PFC and are expressed as percentage of the values ofspleen non-specific Ig-secreting PFC found in conventional F5'EP-Sm-treated mice. The mean-l-1 SD of these values, per 10'' spleen cells, were13898-1-1679, 19963-1-1921, 17675-F 1921. 2333-1-2400, and 11747-H984 fornon-specific IgGI, lgG2a, IgG2b, lgG3 and IgM, and specific IgM respect-ively. The mean + 1 SDofthesePFC.per 10̂ spleen eells, found in E5'EP-Sm-untreated mice were 191-1-25,551-)- 56,493 + 53, 212+42, and 2099-i- 287 fornon-specific IgGI, IgG2a. IgG2b, IgG3, and IgM respectively.

364 R. Soare.s ct al.

however, only low ELISA titres (1/252 to 1/500)of serum ;tntibodies against F5'EP-Sm werefound in these mice, atid in all Ihe remaininganimals immunized against this protein, com-pared to an average of I/I 10 ELISA litresobserved in non-immunized mice. A similarreduction of the immunosupressor effect was alsoobserved in miee depleted of CD4' cells orthymectomized {Fig, 2).In these two groups ofmice, the loss of immunosuppressor effect wasassociated wilh a decrease in ihe B-cell polyclonaiactivation also induced by F5'EP-Sm (Fig. 3).

Survival of S, mutans in the peritoneal cavity ofC57BLI6 mice

The number of 5. mutans found, 3 days afterthe infection with this bacterium, in the peritonealcavity of mice depleted of CD4' cells or thymec-tomized was considerably smaller than thatfound in conventional mice (Fig. 4). Three daysafter infection, no micro-organisms were found inthe peritoneal cavity of mice previously immu-nized intraperitoneally and boosted 15 days laterwith F5'EP-Sm (Fig. 5). On the contrary, severaltens ofmillionsofmicro-organisms were found inthe peritonea! cavity of non-immunized animals 3days after being equally infected with .S'. tnutaus.

FIG. 4. Means-)-1 SD of the number of .S./HW/UH.V foundin the peritoneal cavity of conventional C57BL,/6 mice(•) and C57BL/6 mice depleted of CD4 ' cells (@) orthymectomized (O). 3 days after being inoculatedintraperitoneally with 2.5x10'^ of these micro-orga-nisms.

10

FIG. 5. Percentage of the control values of the number ofS, mutans found in the peritoneal cavity of C57BL/6mice immunized intraperitoneally and boosted 15 dayslater ( • ) , immunized intraperitoneally and boosted 90days later (•) and immunized intradermally andboosted 90 days later (B), 3 days after being inoculatedwith 2.5 X 10'' micro-organistns. In every instance themice were infected 1 week after boost. The mean + 1 SDof the number of micro-organisms found in the perito-neal cavity were 5,1 x 10''-(-7x 10'' for the controls(non-immunized) of the mice immunized intraperito-neally and boosted 15 days later and 6.2 x I0'̂ + 6x 10̂for the controls of the mice boosted 90 days afterimmunization intraperitoneally or intradermally.

Furtherrnore. even in the peritoneal cavity of themice boosted 3 months after being immunizedintraperitoneally or intradermally. the numbersof bacteria found 3 days after infection wereconsiderably reduced comparatively to non-immunized mice, although less so in the formerthan in ihe latter group o^ mice (Fig. 5).

DISCUSSION

We have previously observed that both immuno-suppressor and B-cell mitogenic properties ofF5'EP-Sm were markedly reduced in thymecto-mized mice [7]. In the present report we not onlyconfirm these observations but we also deseribe asimilar occurrence, although to a lesser extent, inmice depicted of CD4" cells. Moreover, we shownow thai this reduction of the immunosuppres-sive and B-ccll mitogenic properties of F5'EP-Sm,in miee depleted of T cells (new immigrants in thecase of adult thymeetomized mice and helper Teells in the case of individuals depleted of CD4 'eells). is correlated with a lower susceptibility toinfection by S. mutans. Although apparentlypradoxical, this lesser susceptibility to infection

Immunoprotection against Streptococcus mutans 365

by S. mutans of mice depleted of T ceils mayexplain the also apparent contradictory finding ofF5'EP-Sm sharing stimulatory and inhibitoryproperties, such as B-cell mitogenicity and immu-nosuppression respectively. Indeed, the genera-tion of T-suppressor eells has been repeatedlyascribed to B lymphocytes [1. II, 13, 19-21]. andwe have recently observed [16] that, in the case ofa similar B-cell mitogenic/immunosuppressiveprotein released by ASFV, the generation of T-suppressor cells induced by activated B cells wasmediated by T-helper cells which have feedbackstimulatory effects on B cells. It is thus temptingtoassume that a similar event occurs in the case ofF5'EP-Sm which ultimately explains why micedepleted of T cells are less susceptible to bacterialinfection and to the B-cell mitogenic and immu-nosuppressor effects of this protein.

In other previously reported data [18] we haveshown that mice treated with F5'EP-Sm weremore susceptible to infection by S. mutctns, thussuggesting that the B-cell mitogenic protein se-creted by this micro-organism plays an importantrole in the protection ofthc bacteria in the host.Confirming this assumption are the observationsof this paper that mice immunized against thisprotein are mueh less susceptible to the bacterialinfection than non-immunized individuals, evenwith relatively low serum specific antibody titres.

All these observations suggest that classicalstrategies of immunoprotection aiming at apotentiation of the immune system against infec-tious micro-organisms may, in certain eases,produce the opposite result, as micro-organismssecreting B-cetl mitogens seem to survive in thehost through an overstimulation of the immunesystem. Thus the proper strategy in these casesseems to be immunostimulation of host againstthe B-cell mitogenic protein, instead of the immu-nostimulation against the micro-organism itself.

ACKNOWLEDGMENTS

This work was supported by "Junla Nacional deInvestigagao Cientilica e Tecnoiogica' (JNICT)and by 'Institulo Nacional de Investigaijao Cien-tifica" (INIC).

REFERENCES

I L"Age-Stehr, J., Teichmann, J., Gershon, R.K. &

Cantor, H. Stimulation of regulatory T eel! circuitsby lmmunoglobulin-dependent structures on acti-vated B cells. J. Immunol. 10, 2!, 19S0.

2 Arala-Chaves, M.P., Ribeiro, A.S., Santarem,M.M.G. & Coutinho, A. Strong mitogenic effect formurine B tymphoeytes of an immunosuppressorsubstance released by Streptococcus intermedius.Infect. Immunol. 54, 543, 1986.

3 Arala-Chaves, M.P., Ribeiro, A.S., Vilanova, M.,Porto, M.T., Santarem, M.M.G. & Lima, M.Correlation between B-cell mitogenicity and immu-nosuppressor effects of a protein released by poreinemonocytes infected with African swine fever virus.Am.J. Vet. Res. 4% 1955, 1988.

4 Benner, R.. Meina, F., Van der Meulen, G.M. &Van Muiswinkel, W.B. Antibody formation inmouse bone marrow. I. Evidence for the develop-ment of plaque forming cells in situ. Immunology 26,247, 1974.

5 Couteiier, J.P., & Van Snick, J. Isotypically res-tricted activation of B lymphocytes by lactic dehyd-rogenase virus. Eur. J. Immunol. 15, 250, 1985.

6 Dialynas, DP. . Quan, Z.S.. Wall, K.A., Pierres. A.,Quintans. J.. Loken. M.R., Pierres, M. & Fitch.F.W. Characterization of Ihe murine T cell surfacemolecule designated L2T4 identified by monoclonalantibody GK.1.5: similarity of L.1T4 to the humanLeu-3/T4 molecule. J. Immunol. 13, 2445, 1983.

7 Ferreira, P., Soares, R., Ribeiro, A. & Aritln-Chaves, M.P. Correlation between specific immuno-suppression and polyclonal B cell activationinduced by a protein secreted by the Streptoccusmutans. Scand. J. Immunol. 27, 549, 1988.

S Forni, L. Reagents for immunofluorescence andtheir use for studying lymphoid cell products. P. 151in Lefkovitz, I. & Pernis, B. (eds) ImmunologicalMethods. Academic Press, New York, 1979.

9 Gronowicz, E.. Coutinho, A. & Melchers, F.A.plaque assay foral! cells secreting Ig of a given typeor class. Eur. J. Immunol, 6, 588, 1976.

10 Grosskinsky, C M . & Askonas. B.A. Macrophagesas primary target cells and mediators of immunedisfunction in african trypanosomiasis. Infect.Immunol. 33, 149. 1981.

11 Hausman, P.B.. Sherr, D.H. & Dorf, M.E. An invitro system for the generation of suppressor cellsand the requirements for B cells in their induction. 7.Immunol. 134. 13H8, 1985.

12 lnipero-Lima, M.R., Joskowiez, M., Coutinho, A.,Kipnis, H.L. & Eisen H. Very large and isotypicaltyatypical polyclonal PFC responses in mice infectedwith Trvpanosoma cruzi. Eur. J. Immunol. IS, 201,1985.

1.1 Kennedy, M.W. & Thomas, D.B. A regulatory rolefor the memory B cell as suppressor-induced offeedback control. J. Exp. Med. 157, 547, 1983.

14 Lundkvisl, 1., Hoimberg, D., Ivars, F. & Coutinho,A. The immune response to bacterial dextrans. III.Ontogenic development and strain distribution ofspecific clonal precursors. Eur. J. Imnntnoi 16, 357,1986.

15 Minoprio, P., Eisen, H., Joskowiez, M., Pereira, P.& Coutinho, A. Suppression of polyelonal antibodyproduction in Trypanosoma crwri-infected mice by

366 R. Soares et al.

treatment with anti-L3T4 antibodies. J. Immunol.139,545. 1987.

16 Ribeiro, A.S., Arala-Chaves, M.P.. Vilanova, M.Porto, M.P. & Coutinho, A. Role of B and Tlymphocytes in the specific immunosuppressioninduced by a protein released by porcine monocytesinfected with African swine fever virus. Submittedto Eur. J. Immunol.

17 Sacks. D.L., Scott, PA., Asofsky, R. & Sher, FA.Cutaneous leishmaniasis in anti-lgM-treated mice:enhanced resistance due to functional depletion of aB cell-dependent T eell involved in the suppressorpathway. J. Imrmtnot. 132. 2077, 1984.

18 Santarem, M.M.G., Porto, M.T., Ferreira, P.,Soares, R. & Arala-Chaves, M.P. Semi-purificationof an immunosuppressor substance secreted byStreptococcus mutans which plays a role in theprotection of the bacteria in the host. Scand. J.Immunol, (in press).

19 Shimamura, T. & Yoshida, T. Involvement of

antigen and 1-J determinants in the induction ofeffector T suppressor cells by immune B cells. Cell.Immunol. 112,214, 1988.

20 Shimamura, T., Hashimoto, K. & Sasaki, S. Feed-back suppression of the immune response in vivo.Cell. Immunol. 68, 104, 1982.

21 Taylor, C.E., Stashak, P.W., Caldes, G., Preseott,B., Chused, T.E., Brooks, A. & Baker, P.J. Acti-vation of antigen specific suppressor T cell by B cellsfrom mice immunized with type III pneumococcalpoiysacharide. J. E.xp. Med. 158, 703, 1983.

22 Vendrell, J.P., Rabesandratana. H., Huguet, M.F.,Cannat, A. & Serre, A. Brucella fractions behave asnonspecific mitogens and polyclonal B-eell acti-vators for human lymphocytes. Infect. Immunol. 50,310,1985.

Received 24 July 1989Accepted in revised form 7 November 1989