Deficient T-cell responses in non-responders to hepatitis B vaccination: absence of TH1 cytokine production

lmmum~h~g r Letter,~. 39 ( 1994'1 163-168 0165 - 2478

94

$ ".00 ,i' 1994 Elsevier Science B V. All rights reserved

I M LET 021)73

Deficient T-cell responses in non-responders to hepatitis B vaccination: absence of TH1 cytokine production J. V i n g e r h o e t s ~''*, G . V a n h a m a, k. K e s t e n s a, G . P e n n e a, G . k e r o u x - R o e l s b a n d P. G i g a s e a JLabt~ratort of Pathologr and hnmunolog.~. Institute ol 1-roptcal Medicine, ,Vationalt,~lraat 155. B-2000 .4ntwerpen I, Belgtum." "bDepartment ot Clinical Chemt~trr. ( oiver~it.r Ho~pttal Ghent. Dr' Phuelaan I,~5. B-9000 Ghent. Belgium IRecei~ed 15 Notember 1993: accepted 23 No~ember 1993~

1.

Summary

The inability to mount a protective level (/> 10 IU 1) of hepatitis B surface antigen (HBsAg)-specific antibodies after vaccination is presumabl3 the consequence of a defect in the cellular immune regulation. We compared the in vitro immune responses of peripheral blood monont, clear cells (PBMC) from high. intermediate and non-responders, after stimulation with recombinant HBsAg. The absence of a proliferative response in non-responders was not reversed by removal of C D 8 - T cells, indicating that HBsAg-specific CD8 -~ T-cell-induced suppression was not the tmderlying cause of non-responsiveness. Non-responders did not produce cytokines after HBsAg stimulation. High responders displayed a typical Thl-like profile since their PBMC produced interleukin-2 (Ik-2) and ?,-interferon (IFN2') and no detectable Ik-4 or Ik-5 upon stimulation with HBsAg. 2.

Introduction Vaccination ~ith HBsAg offers an effective

protection against infection with the hepatitis B virus in the majority of vaccine recipients. Two to 10% of healthy adult vaccinees fail to generate a protective immune response to the antigen ( 1> 10 l U l l [I,2]. This non-responsiveness in immtmocompetent individuals may emerge from a dysfunctional cellular immune system since most non-responders can be induced to produce antiHBs with higher or additional doses of HBsAg or with increased non-specific T-cell help [3]. Although Milich and co-workers [3-5] clearly demonstrated the presence of an MHC-controlled immune response in the mouse s~stem, the presence of an MHC-linked control mechanism in hunaans is still controxersial and the precise mechanisms causing these variable immune responses remain uncertain. To dex elop a better understanding of the cellular basis of non-responsiveness, we investigated in ~itro activation characteristics of peripheral blood lymphocytes from ~accinated donors after stimulation with recombinant HBsAg IrHBsAg). The possibility of CD8 ÷ T-cell-mediated suppression was first e~aluated. Next, we compared proliferatixe responses and c~,tokine profiles in PBMC from high. intermediate and non-responders to HBsAg.

Kt.l ~tord~ Hepatitis B ~,acclnatloll; Immune response: Hepatins B ~urface antigen: T-cell acti~auon: C.~tokine: Suppressmn

"Corresponding author: Johan Vingerhoets. B.Sc.. Institute of Tropical Medicine. Laborator.~ of Patholog3 and Immunolog.v. NatJonalestraat 155, B-2000 Antx~erpen I. Belgium. Tel.: 132-3) 247-62-28: Fax: q'32-31 247-62-31.

SSDI () 1 6 5 - 2 4 7 8 ( 9 3

) El) 1 6 4 - 8

4hhrevtatton~,." HBsAg, hepatitis B surface anngen: HBV, hepatitis B titus: HIV. h u m a n immunodeficienc.~ titus: HLA, human leucoc.~te antigen, IFN. interferon: Ik. mterleukin: ILl, international units: mAb, monoclonal antibod.~: MHC. major histocompaubdn.~ comple,~: Th. T helper.

163

3.

3.1.

Materials and Methods

Subjects

Healthy volunteers, aged 23~-0 years, previously vaccinated with a standard administration of 3 vaccine doses at months 0. 1 and 6, were studied. Anti-HBs levels were determined by ELISA (Sorin Biomedica, Italy) prior to the collection of blood samples. Non-responders, intermediate responders and high responders were defined as vaccinees having anti-HBs levels of ~< 10 IU,,I, between 10 and 10000 IU/I and >10000 IUI, respectively.

3.2.

Cell preparations

PBMC were obtained from EDTA-anticoagulated venous blood b.x density gradient centrifugation on Ficoll-Paque (Pharmacia LKB, Uppsala, Sweden). After washing. PBMC were resuspended in complete medium consisting of RPMI1640 supplemented with L-glutamine (2 raM), sodium-pyruvate (1 mM), penicillin (100 Uqnl), streptomycin (100 l~g,'ml) (all reagents from Gibco, Paisley, Scotland), 2-mercapto-ethanol (5 × 10 -~ M, Merck-Belgolabo, Overijse, Belgium) and 10% heat-inactivated pooled human serum (negative for all serological markers of HBV infection ).

3.3.

Prol(/eration a~sav

PBMC were cultured at 2 x 10~',ml (4 × 10~ well) in complete medium for 6 days with various concentrations of rHBsAg (DVP23, provided by Smith Kline Beecham Biologicals, Rixensart, Belgium). Varicella zoster antigen (VZAg, BehringHoechst, Brussels, Belgium: lot nr. 412858A), tetanus toxoid (TT, Wyeth Laboratories, PA: lot nr. 4878222) or poke weed mitogen (PWM, Gibco, Paisley, Scotland: lot hr. 10N5003) were used as control stimuli (I 500 final dilution). Proliferation of PBMC x~as determined by measuring incorporation of [3H]thymidine (0.4 /,lCi well: Amersham, UKI during the final 20 h of the culture period. Results were expressed as mean cpm +standard deviation or as a stimulation index (SI =mean cpm of stimulated cultures mean 164

cpm of unstimulated cultures). The proliferative response was considered positive when SI was greater than 2.

3.4.

Separation of lymphocyte subsets

CD8 + T cells were purified with an immunomagnetic separation technique (Dynabeads M450 CD8 and D E T A C H a B E A D from D.~nal, Oslo, Norway) according to manufacturer's instructions. The positive fraction consisted of more than 99°,'o CD8+ ceils. The negative fraction contained less than 5°o CD8 ÷ cells. CD8 ~T-cell-depleted PBMC, purified CD8 + T cells, and reconstituted PBMC (200,0 CD8 ÷ cells plus 80°0 CD8 PBMC) were stimulated as described for unseparated PBMC.

3.5.

Detection of ~:vtokhws

IL-2 s~as determined using IL-2-dependent proliferation of the CTLL-2 line. To increase the sensitivity of the IL-2 measurement, an IL-2R~-specific mAb (anti-TAC 2 ~g, ml, provided by Dr. T. Waldman) was added during the entire culture period. This blocked interaction between IL-2 and the high-affinity IL-2 receptor and thus prevented IL-2 consumption [6]. IFN). production x~as deten'nined using a bioassay based on induction of MHC class II antigens on COLO-205 cells (ATCC CCL 2221. Class I1 expression was measured b) flow cytometry using an FITC-conjugated anti-HLA-DR mAb (Becton Dickinson, Erembodegem, Belgium) [7]. IL-4 was measured ~ith a commercially available immunoassay (Biotrak ELISA, Amersham, UK: sensitivity 3 pg, ml). IL-5 ~as measured using a bioassa 3 based on the proliferation of BaF:~-cells, a mouse cell line (BaF3) transfected with cDNA coding for the human IL-5 receptor :~-chain (BaF,:~-cells, kindly donated b 3 Dr. R. Devos. Roche Research Ghent, Belgium.). Standardization of the cytokine assays x~as done b) adding serial dilutions of recombinant (r) IL-2, rIFN),, rlL-4 and rIL-5, respectively. All bioassays ~ere specific and had a sensitivity comparable x~ith commercially available immunoas-

says (lower detection limit: 0.05 U:'ml, 0.1 IU/ml and 5 pg.ml for the IL-2, IFN7 and IL-5 assay, respectively). Antigen-induced cytokine production in culture supernatants was measured after 24, 48 or 72 h of stimulation, depending on the experimental conditions.

40

35 30 "o

c:

I

25

f.n

i

~ 2o ,1=

Results

4.

E

4.1.

I

"

TT

PWM

15

e~

Proliferative response to r H B s A g

U

The proliferative response of PBMC from a high, intermediate and non-responder to various concentrations of rHBsAg is shown in Fig. I. A significant proliferative response (SI >2) was detected in high responders only and ~as clearly dose-dependent. Intermediate responders showed a marginal response whereas no proliferation could be induced in PBMC from non-responders. The proliferative responses to control antigen or mitogen stimulation were not different in the 4 high responders as compared to 2 intermediate and 3 non-responders tested (Table I and Fig. I). The lack of proliferative response to HBsAg in non-responders was certainly not a sign of a generalized immune defect but ~.as restricted to HBsAg alone.

10

i

o

0.003

0.03

0.3

3 pg/ml

rHBsAg F~g. I. Proliferative responses o f P B M C from a representative high responder I.II I, intermedmte responder I A ; and non-responder ( O ) to m e d i u m control, rHBsAg, T T and P W M . Results are expressed as mean cpm +_ S D of triplicate cultures.

4.2.

Elimitzation o f CD8 + T cells

PBMC were depleted of CD8 ÷ T cells by an immunomagnetic separation technique before the onset of culture. Removal o f C D 8 ÷ T cells from a non-responder did not restore the proliferation to rHBsAg but slightly increased the response to

TABLE I P R O L I F E R A T I V E RESPONSES IN HIGH . A N T I G E N DOSE FOR EACH SUBJECT

INTERMEDIATE

AND

NON-RESPONDERS O B T A I N E D

AT O P T I M A L

Results are expressed as mean cprn + s t a n d a r d de~mtion. N D = not done. Responder

M e d i u m alone

HBsAg

High I 2 3 4

1883 3803 6256 301

31370 10269 44341 5235

Intermedmte I 2

4473 + 8738 +

+ 455 +_ 1821 + 1850 + 56

885 617

+ 7717 +_ 191)6 + 14721 + 3455

7246 + 2599 7431 _+ 786

T e t a n u s toxotd

VZAg

23758 + 1062 ND ND ND

23281 10659 98665 21363

14791) + 2655 ND

ND 50918 ± 12266

ND ND ND

51381 _.+ 81)10 20263 + 2607 39029 +. 3146

+ 3145 -2_ 696 _.+ 5448 _+ 8174

Non

I 2 3

3431 + 257 3644 _-L- 3119 3726 + 1203

,9,0 + 3029 + 3519 +

851 342 851

165

A)

B) HIGH R E S P O N D E R

NON-RESPONDER cDm

cpm (Thousands)

(Thousands) "O 60 50

kx.

40

i'i

302010-

X

0 PBMC

TS(-)

T8(-) +T~('~)

TS(+)

T8(.)

PBMC

I

['70 ~Jg,'mt t-]D2 ~g,HBsAg rnl [~]1 #grnl E~5 ~zg'm~

A

"re(.)

T8(-) +TS(+)

HB~Ag 1 ['70 ~g ml [--~0 2 ug,ml 1'7 1 tagml [~ 5 ~g,ml

Fi£.. 2. Prolil'eratl~e responses o f total P B M C , CD8( ÷ ) T-cell-depleted P B M C (TS( - )), reconstituted P B M C I ]-8( - I + TS( + i~ and purified CD8( ÷ ) T cells (T8( + j~ upon st=mulation ~=th tarious concentrations o f HBsAg in a N R q.-~) and in a HR (B~. Results are expressed as mean cpm o f quadruphcate cultures

VZAg, as sho~n in Fig. 2. Proliferation of CD8depleted PBMC from the high responder to both rHBsAg and VZAg was slightly increased, probably due to a relative enrichment of CD4 ÷ T cells in the C D 8 - fraction. Highly enriched C D 8 - T cells from non- or high responders did not proliferate to rHBsAg or VZAg (Fig. 2).

4.3. Pro~h.'tion of cytokines Tx~o subjects of each responder group ~ere tested for cytokine production in ~itro. The production of IL-2 was shoran in cultures of PBMC from high and intermediate responders using an optimal concentration of rHBsAg (bet~een 0.3 and 3 llg ml, dependent on the individual) (Fig.

A)

B) PRODUCTION OF INTERLEUKIN 2

PRODUCTION OF INTERFERON GAMMA ~/ml

u/ml 18

30

14-

25

12-

2O i-

o8-

00-

10

04-

02

0

tlL'J

,iq[

vZAg Mesa 12n

i=[i.lll=

VZ~g *esA0 24h

t

.=r,rl

VZ.'0 ~0~g 48n

I--]NR I-']lR E ] H R ]

ill

i

['It

VZAg .St,Lg 72h

0

VZAg

H0sAg 24n

4ZAg

HBsAg 72h

["-]NR ['--]IR ['-]HR

Fig. 3 Production of IL-2 (A) and o f IFN,'q B) b} P B M C upon stimulation ~ith V Z A g or r H B s A ~ after 12, 24, 48 or 72 h o f cuhure. The Io~er detection limit for IL-2 is indicated b> the horizontal line (AI. One representative ~accmee of the 3 ~roups INR, IR and HR) is sho~n. Culture of PBMC m the absence o f a n u e e n did not reduce IL-2 or IFN7 secretion.

166

3A). IL-2 secretion by PBMC from high responders was detectable after 12 h and increased until 72 h after the start of the culture. PBMC from intermediate responders started to produce IL-2 after 24 h whereas no IL-2 was produced by PBMC from the non-responders (Fig. 3A). Control antigen (VZAg) induced significant amounts of IL-2 in the culture supernatants from high, intenaaediate and also from non-responders (Fig. 3A). IFN'; production induced by rHBsAg could only be detected in culture supernatants from high-responder PBMC (Fig. 3B) and w,as completel3 inhibited by the addition of anti-TAC during the culture period (results not shox~n). Stimulation with VZAg induced similar secretion of IFN'; b3, PBMC from high, intermediate and non-responders (Fig. 3B). Despite the use of sensitive and specific assays, no secretion of Ik-4 and Ik-5 could be measured after rHBsAg stimulation of PBMC from ant of the vaccinees. 5.

Discussion

The failure to mount antibodies to HBsAg is clearly related to MHC class ll-linked immune response mechanisms in mice [3-5]. In humans, similar genetic and immunological mechanisms could be operative. Immune response genes linked to MHC genes are thought to control nonresponsiveness either through (I) failure in antigen presentation and CD4 ~- T-cell activation [8], or 121 through CD8 + HBsAg-specific T-cellmediated suppression [9]. The discrepant findings in these studies could be attributed to ethnic dift'erences in the study population or to different immunization routes. Since anti-HBs production is a T-cell-dependent process, several authors have studied T-cell proliferation in vaccine recipients. However, reports on detection of HBsAg-specific T-cell responses in fresh PBMC are not always consistent and indicate the need for an optimal assay system [10-15]. We used a reliable proliferation assay to measure cellular responses to HBsAg in responders and non-responders [16]. Our results sho~ed that the data from the proliferation assays in vitro paralleled the serological division in high, inter-

mediate and non-responders in vivo. The lack of proliferation after in vitro stimulation of non-responder PBMC, was shown to be HBsAg-specific since the responses to control antigens or mitogen were not affected. The present study also showed that CD8 ÷ Tcell-mediated suppressor activity was not the mechanism causing the lack of HBsAg-specific response in non-responding vaccinees. This observation confirmed the data reported b3 Egea et al. who did not find evidence of suppression in Caucasian non-responders. Both our and Egea's results are therefore different from the Japanese data where CD8 ~- T-cell-mediated suppression was clearly demonstrated on the cellular basis of non-responsiveness in Caucasian vaccinees [8,9]. Several studies have looked at the precise role of T h l - and Th2-derived cytokines in the immune response to vaccination against different organisms in mice and man. Single vaccination of mice using attenuated Schistosoma or Leishmania organisms, evoked strong and protective Thl responses with in vitro production of II_-2 and IFN'; [17,18]. Protection could be do~n-regulated by anti-IFN'; or IL-10, suggesting the in vivo importance of a Thl activity dominating the Th2 response [19,20]. However, multiple vaccination with irradiated S. mansoni stimulated Th2 responses in vitro and those responses reinforced protective immunit.~ [17]. In ~itro production of Thl-like c3tokines by human PBMC together with lymphocyte proliferation have been demonstrated after intradermal immunization with a bacterial vaccine strain [21]. A predominant Th2 response with increased IL-4 production and decreased IFN'; production in vitro. ~as shown after measles virus ~,accination [22]. Evidence of in vitro Th2-1ike cytokine production was also given in hepatitis B vaccinees, where CD4 + T-cell clones from responders could produce IL-4 [23]. HBsAg-specific IL-2 production in vitro has not been thoroughly investigated until no~. One report described a biphasic IL-2 secretion in hepatitis immune individuals [24]. Our findings differ from the observations of S31van and Hellstr6m [24]. possibl3 due to differences in the experimental conditions used. We showed that the production of IL-2 upon rHBsAg stimulation in vitro paralleled high. intermediate and low responsive167

ness in vivo. These data theretbre suggested that the rHBsAg-induced production of IL-2 in vitro was more sensitive than the proliferation assay, since PBMC of intermediate responders did produce IL-2 but showed no proliferative response upon stimulation with rHBsAg. Similarly, peptide-induced IL-2 production in humans immunized against HIV has indeed been shown to be more sensitive than lymphoproiiferation [25]. The HBsAg-specific production of IFN~, was restricted to high-responder PBMC and was dependent on IL-2-IL-2 receptor interaction, since anti-TAC blocked all IFN'; production. We did not observe any production of the Th2like cytokines IL-4 and IL-5 upon HBsAg stimulation, irrespective of the in vivo responder status. To what extent our in vitro system alloCved the generation of Th2-1ike responses is still not clear. Since our data show that PBMC of responders produce IFN 7, known to have an antagonistic effect on Th2 function, possible Th2-1ike activit3 could be inhibited. In summary, the data presented here shov, that, in our Caucasian population, CD8 ÷ T-cellmediated suppression does not induce non-responsiveness. Moreover, Thl but not Th2-1ike activity could be clearly demonstrated in high responders, to some extent in intermediate responders, but not in non-responders. Acknowledgements We thank Dr. T. Waldman for the kind gift of anti-TAC mAb. Dr. R. Devos for kindly providing BaF.:¢ cells and Smith Kline Beecham Biologicals for supplying rHBsAg. The authors are grateful to Ms. Greet Verhulst for excellent assistance in processing this manuscript. We also thank Dr. Chris Locher for critical reading of the manuscript. This work was supported by the Belgian Fund tbr Scientific Research (NFWO). Grant 3.0[)95.90.

References [I] Eddleston. A. (19901 Lancet 335. 1142.

168

[2] Craven, D.E., A~deh. Z.L., Kunches, L.M.. Yunis. E.J.. D~enstag, J.L.. Werner, B.G., Polk, F., Sn>dman, D.R., Plait, R. and Crumpacker, C.S. (1986) Ann. Intern. Med. 105. 356. [3] Mihch. D.R., Alexander, H. and Chisari, F.V. {1983) J. Immunol. 130. 1401. [4] Milich. D.R. and Chlsari. F.V. (1982)J. Immunol. 129, 320. [5] Millch, D.R., Leroux-Roels. G.G. and Chisari, F V (1983) J. Immunol. 130. 1395. [6] Baroja, M.L. and Ceuppens. J i . (1987) J. Immunol. Methods 98. 267. [7] Current Protocols m Immunolog.,, (1991~ IColigan. J.E.. Kruisbeek, A.M., Margulies, A.M., She~ach, E.M. and Strober, W., Eds.), Greene Publ. and Wile)-Interscience, New York. [8] Egea. E., Iglesms, A., Salazar, M., Morimoto. C.. Kruskall, M.S.. A~deh, Z., Schlossman. S.F.. Alper, C A . and Yums, EJ. fl991) J. Enp. Med. 173, 531 [9] Watanabe. H., Okumura, M.. Htra>ama, K. and Sasazukl. T. (Iq901 Tissue Antigens 36, 69. [10] Cehs. E.. Kung. P.C. and Chang, T.W. ( 19841J. Immunol. 132, 1511. [I I1 Cupp~, T.R.. Gerin. J.L.. Purcell. R.H., Goldsmith. P.K. and Faucl, A.S. (1984)J. Chn. In~est. 74, 1204. [12] Filion, L.G., Saginur, R. and Szczerbak. N. (1988) Chn. Exp. Immunol. 71. 405. [131 Ste~ard. M W., Sisle), B.M., Stanle). S.E. and Brown, C.R. 11988) Clin. E',,p. Immunol. 71, 19. [141 Fernan, A.. Career, C.J.R, and Cooksle~, ~,V.G.E. (1989) Clin. Exp. Immunol. 76. 222. [15] Degrassl. A., Mariani, E., Honorati, M.C., Roda, P., Miniero. R.. Capelh. M and Facchim, A. (1992) Vaccine 10. 617. [161 Leroux-Roels, G., Van Hecke, E., Michlelsen. W.. Voet, P., Hauser. P and P~tre, J, (1993) Vaccine. [I 7] Caulada-Benedetti. Z., AI-Zamel, F.. Sher. A. and James, S. 1'1991) J. Immunol. 146, 1655. [18] Kaye. P.M.. Curr3. A.J. and Black~.~.ell. J.M. 119911 J Immunol. 146, 2763. [lq] Sm)thles, L.E.. Coulson, PS. and Wdson, R.A. 11992.1J. Immunol. 149. 3654. [20] Sher. A., Fiorentino. D , Caspar. P., Pearce. E. and Mosmann, T.l'1991)J. Immunol 147, 2713. [211 Karttunen. R., Surcel. H.-M . Andersson, G.. Ekre, H.-P. and Her~a, E. (1991) J. Clin. Microbiol. 29, '7,53. [22] Ward. B.J. and Griffin. D.E. 11993) Chn Immunol. Immunopathol. 67. 171. [23] Tsutsui. H.. Mizoguchi, Y. and Monsa~a. S. 11991) Scand J. Immunol. 34. 433. [24] S>l,.an. S.P.E. and HellstrL~m, LiB 11989) Clin. Exp. Immunol. 78. 207. [251 Clenct. M.. Stocks. N.I.. Zajac. R.A. Bos~.,.ell. RN.. Bernsteln. D.C., Mann, D.L., Shearer. G.M. and Berzofsk}. J.A. ( 19891 Nature 339, 383.