Journal of Autoimmunity (1992) $691-702

T-Lymphocyte Reactivity to the Recombinant Mycobacterial6G and 709kDa Heat Shock Proteins in Multiple Sclerosis

Marco Salvetti, Carla Buttinelli, Giovanni Ristori, Maurizio Carbonari,* Michela Cherchi,* Marco Fiorelli, Maria Grazia Grasso, Luigi Toma and Carlo Pozzilli Third Chair of Neurology,

Department

Clinical Immunology,

of Neurological

Universitci ‘La Sapienza’,

Sciences and *Department 00185Rome,

of

Italy

(Received 11 May 1992 and accepted 17 August 1992)

Owing to their conservation and immunogenicity, heat shock proteins (hsps) represent a class of potential autoantigens. Moreover, they could be targets for y6 T lymphocytes, which are prominent in various immune disorders. We studied the T cell proliferative primary responses to recombinant M. bovis 65 kDa hsp (hsp65) and M. tuberculosis 70 kDa hsp (hsp70) in 31 patients with multiple sclerosis (MS), 19 patients with other neurological diseases (OND) and 19 healthy individuals. Positive responses to hsp70, but not to hsp65 were significantly more frequent in patients with MS than in patients with OND or in healthy individuals. In order to verify and refine these results and to characterize the hsp reactive T lymphocytes, we screened 147 PPD-specific long-term T cell lines (76 from 10 patients with MS and 71 from 12 healthy donors) for their proliferative response to hsp65 and hsp70. hsp70-reactive T lines were significantly more common in patients with MS than in healthy controls. The number of T lines responding to hsp65 increased in the MS group only slightly. In 19 T lymphocyte lines from patients with MS and healthy donors, a cytofluorometric analysis was performed with special attention paid to distinct T cell receptor 76 determinants. With one exception, in each line the population of 78 T cells remained a minority. We conclude that an increased T cell response to mycobacterial hsp70 may be present in patients with multiple sclerosis.

Correspondence to: Dr Marco Salvetti, Dipartmento di Scienze Neurologiche, v.le dell’Universit8 30, 00185Rome, Italy. Tel: (39) 6-49914441. Fax: (39) 6-4440790. 691 0896-841 l/92/060691

+ 12 $08.00/O

0 1992 Academic Press Limited

692 Marco Salvetti et al.

Introduction An immune response to heat shock proteins (hsps) is thought to be part of a general, initial defense against infection: hsps are conserved and immunodominant determinants, expressed by all organisms in which they have been sought, following various forms of cellular stress. During infection, microbial as well as host cell hsps are induced. This simultaneous presence of homologous antigens would allow hsp specific T lymphocytes to cope with both heterogeneous invading micro-organisms and the cellular changes they might have produced [ 11. Although advantageous, such a response is intrinsically at risk of autoimmunity and examples of this possibility have been reported in human and experimental immune disorders [2,3]. Mycobacterial preparations seem to harbor a ligand for y6 T cell receptor-bearing lymphocytes, capable of triggering a preferential activation of these cells. This ligand has not yet been identified, although hsps are considered as suitable candidates. y6 T lymphocytes have been isolated at inflammatory lesion sites in rheumatoid arthritis and in a polymyositis patient [4-6]. These cells are increased in number in other immunological diseases [7]. In rheumatoid arthritis synovial fluid, y6 T cells specific for Mycobacterium bovis 65 kDa hsp (hsp65) have been described [4]. However, hsp specific y6 clones are part of the immune repertoire of healthy individuals [8]. With respect to multiple sclerosis, hsps are constitutively expressed but also stress-induced in the mammalian brain [9]. Here, preferential induction of members of the hsp70 family seems to occur in glial and vascular elements [l&12] and in those mammalian brain fiber tracts where oligodendroglia predominate [ 131. The number of y6 T cells is higher in the cerebrospinal fluid of patients with inflammatory diseases of the central nervous system [ 141. These cells are able to lyse oligodendrocytes in vitro [ 151. Finally, y6 lymphocytes, and oligodendrocytes that express hsp65, are associated with chronic lesions in MS [ 161. Our study was aimed at determining whether preferential T cell recognition of mycobacterial hsp65 and/or hsp70 occurs in MS and which T cell subsets may be involved in the response to these antigens. Methods Donors

Blood samples were drawn from 31(18 males and 13 females, aged 16-52 years, mean 30.0 rf: 8.6) patients with a clinically definite form [ 171 of relapsing-remitting (n = 27) or chronic-progressive (n = 4) MS [ 181. The mean duration of disease was 6.6 + 3.3 years. Nineteen patients with other neurological diseases (9 cerebrovascular, 6 inflammatory, 2 degenerative, 1 peripheral neuropathy and 1 trauma; 10 males and 9 females aged 16-79 years, mean 49.7 f 20.30) and 19 healthy individuals (8 males and 11 females, aged 25-69 years, mean 35.7 f 11.1) served as controls. Purified protein derivative (PPD)-specific T cell lines were generated from 10 patients with definite MS (6 males and 4 females, aged 16-52 years, mean 32.8 f 10.9) and 12 healthy individuals (7 males and 5 females, aged 24-55 years, mean 33.2 &-11.5). Most of the donors were part of the primary response study. None of the patients had received immunosuppressants for at least 3 months before entering the study.

hsp-specific T lymphocytes

693

Antigens PPD (Statens Serum Institut, Copenhagen, Denmark) concentration in culture and in proliferation assays was 10 ug/ml. Purified recombinant M. bovis 65 kDa hsp [ 191 and M. tuberculosis 70 kDa hsp [20] (made available by Dr Jan van Embden, Bilthoven, The Netherlands, through the UNDP/World Bank/WHO) were used at 10 pg/ml. Human myelin basic protein (MBP; kindly provided by Prof. H. Wekerle) was used at 30 pg/ml. Generation of T-lymphocyte lines Antigen specific T-cell lines were selected and expanded as described [21]. Briefly, peripheral blood mononuclear cells (PBMC) were separated by Ficoll-Hypaque density gradient centrifugation. After washing, the cells were resuspended in complete culture medium (RPM1 1640 supplemented with 2 mM glutamine, 100 units/ml penicillin, 100 pg/ml streptomycin along with 5% heat-inactivated autologous or pooled AB serum) to a final concentration of 2 x lo6 cells/ml and seeded in 100 ul volumes into round-bottom microtiter wells in the presence of antigens. Twenty wells were seeded for each donor. Cells were expanded by adding to each well 100 ~1 of complete medium containing 30 units/ml of recombinant human interleukin-2 (IL-2) (Cellular Products Inc., Buffalo, NY, USA) every 3 to 4 days for 15 days. Cultures were then restimulated with the relevant antigen and 1.5 x lo5 irradiated (4,000 rad) PBMC/well as antigen-presenting cells (APCs). Specifically growing wells were selected for further expansion. At the time of the third restimulation with antigen ( N 45th day of culture), PPDspecific T lines were tested for their proliferative response to hsp65 and hsp70. Proliferation assays For primary proliferation tests, 2 x lo5 PBMC were cultured in triplicate with or without antigens or mitogen for 5 days. Sixteen hours before harvesting on glass fiber filters, 0.5 uCi/well of [3H]thymidine (Amersham International, UK) was added to each well. A Stimulation Index [SI; (counts per minute of cells + antigen) t counts per minute of cells alone] > 2.5 was regarded as a positive response. The specificity of T-cell lines was assayed as follows. Samples of 2 x lo4 cells from individual antigen specific T-cell lines were placed in culture in duplicates or triplicates with 1.5 x lo5 irradiated (4,000 rad) PBMC as antigen presenting cells and in the presence or absence of antigens or mitogen (2.5 ug/ml phytohemagglutinin). After 72 h, the cultures were labelled with [3H]thymidine (0.5 @/well) and harvested 16 h later. Antigen specific lines were defined as showing both an SI higher than 3 and a delta count per minute higher than 500. In each donor, the response profile to hsps was tested on T cell lines derived from the same blood sample and cultured in parallel. CytoJEuorographic analysis Cells were stained with the following fluorochrome-conjugated monoclonal antiTCR antibodies: I$ chains (WT31, Becton Dickinson), delta chain (TCRGl, T Cell

694

Marco Salvetti et al. Table

1. Percentages

of positive primary responses to antigens

Antigen PPD MBP hsp65 hsp70

OND (n= 19)

Healthy (?a= 19)

73.7 21.0 31.5 47.4

100.0 10.5 50.0 50.0

90.3 4a.4* 46.4 85.7t$

*MS ershealthy: x2=5.932; P=O.O15. tMS vshealthy: x2= 5.240; P=O.O22. $MS wsOND: x2= 6.230; Z’=0.013. Science) and anti-V61 (Delta TCS-1, T Cell Science). The following antibodies (Becton Dickinson) were also used: anti-CD4 (Leu3a), anti-CD8 (Leu2a), anti-CD3 (Leu4), anti-CD16 (Leull) and anti-CD56 (Leu19). Staining was performed according to manufacturers’ instructions. For fresh blood samples, erythrocytes were lysed before staining. Cytofluorographic analyses were performed using a Cytoron (Ortho) flow cytometer, on at least 1 .OOscatter-gated lymphoid cells. Results

Primary response to antigens The results of the proliferation assays are summarized in Table 1. The three groups were comparable with respect to their reactivity to PPD. MS patients responded to hsp70 (85.7%) significantly more frequently than patients with OND (47.4%; Chi-square (x2) =6.23, P 3) proliferative response to hsp65 (x2 = 1.2; not significant) while 19 (25.0%) and 5 (7.0%), respectively, were positive against hsp70 (x2 = 9.77; P < 0.005) (Tables 3a and b). .

hsp-specific T lymphocytes 695 Table 2a. HLA haplotypes andprimary response to antigens in MSpatientsfrom PPD specific T cell lines were assayed Primary

HLA Patient S.P. T.A. C.T. G.A. G.U. C.A. A.M. C.P. S.C. P.F.

DR

A

B

Cw

DQw

2,3 3,30 3,30 11,30 2,28 1,2 1,31 1,32

7 13,51 7 18,41

7 ::

193 23

I 35,40 8,37 27,44

2,3 1 1,3 3 1,3

3,26

44,52

2,5 4,l 4,6 I 1 nd 7

2,5 4,5 3 3,5 2 5,lO 2,5 2,4

1,3

4,lO (w53)

(~52) (w52,53) (~52) (~52) (nd) (~52) (~52) (w52,53)

response

whom

(SI)

PPD

MBP

hsp65

hsp70

15.5 7.0 5.3 9.9 13.7 3.7 2.8 6.1 14.0 52.0

0.5 0.2 1.5 2.6 5.6 2.7 0.7 2.1 5.6 3.3

nd 0.4 0.4 7.4 3.9 0.9 0.4 nd 1.2 1.6

nd 2.5 0.7 5.6 3.4 0.7 5.8 nd 12.3 16.1

Results are given as SI (stimulation index). nd = not determined.

Table 2b. HLA haplotypes and primary response to antigens in healthy donors from whom PPD specijic T cell lines were assayed Primary

HLA Healthy donor F.P. T.L. C.V. M.L. M.C. M.A. M.M. F.A. O.A. V.C. A.N. B.U.

A

B

Cw

DQw

1,30

13,17

2

1,28 11,28 1,2 3,30 3,ll 11,30 30 24 2,ll

8,60 22,35 8,51 18,40 40,41 13,41 13,18 18,49 35,51

6 nd 3,7 3,4 7 5 2 2,6 5,6 7 4 nd

2 1 1,2 2 3 3,2 2 1 1

response

(S I)

DR

PPD

MBP

hsp65

hsp70

7 (w53)

7.4 2.6 4.8 4.4 21.6 5.6 5.9 7.5 6.6 4.8 13.8 7.6

0.8 1.2 0.4 1.1 2.1 2.0 2.0 2.7 1.4 0.9 0.8 2.4

0.8 1.6 0.6 0.6 1.1 3.3 2.5 8.0 4.3 nd 0.7 0.8

2.1 1.5 0.5 4.8 6.2 5.1 3.0 5.9 6.3 nd 2.2 9.5

3 2,6 2,3 3 5 5,7 3,7 5,6 1,6

(~52) (~52) (~52) (~52) (w52) (w52,53) (w52,53) (~52) (~52)

Results are given as SI (stimulation index). nd = not determined.

y6 T cell receptor expression The percentage of y6 T cells was assayed in PPD-specific T lines from six patients and three healthy individuals. From each subject, multiple lines which differed in their pattern of reactivity to hsp70 were chosen for cytofluorometric analysis. A variable association between the proliferative response to hsp70 and the proportion of TCRG 1 + cells was observed in T cell lines from five of six patients and one of three

696 Marco Salvetti et al. Table 3a. Frequencies of hsp65 and hsp70 specific Tcell linesfrom patients with MS. Results are expressed as number of hsp reactive T cell lines/ number of PPD specific T cell lines tested

Table 3b. Frequencies of hsp65 and hsp70 speciJic T cell lines from healthy donors. Results are expressed as number of hsp reactive T cell lines/ number of PPD specific T cell lines tested

Patient

hsp65

Healthy

S.P.

O/4 O/2 2/10 217 l/7 5/10 O/l O/4 l/l1 13120

l/4 O/2 3/10 217 l/7 4/10 O/l O/4 l/l1 7120

24176

19176

T.A. CT. G.A. G.U. C.A. A.M. C.P. S.C. P.F.

hsp70

donor

hsp65

hsp70

F.P. T.L. C.V. M.L. M.C. M.A. M.M. F.A. O.A. V.C. A.N. B.U.

l/2

112

217 213 213 O/2 O/5 O/5 O/3 l/5 O/6 l/11 7119

l/7 l/3 O/3 O/2 O/5 O/5 O/3 O/5 l/6 o/11 l/19

16/71

5171

healthy individuals (Tables 4a and b). The number of TCR y6 lymphocytes of the lines from patients and controls was not related to the percentage of peripheral blood TCR y6 bearing cells or to the V61+ subset. Line I’2 from patient S.P. displayed a high percentage of y6 T cells (63% TCRG 1’). A more detailed cytofluorographic analysis was performed on this T line: almost all the y6+ cells appeared to use the V61 encoded 6 chain (55% GTCSl ‘). Double staining for CD4-TCRGl, CD8-TCRGl and CD4-CD8 co-expression gave 60/6, 5696 and 5%, respectively. Furthermore, the cells were uniformly negative for the natural killer cell (NK) markers CD16 and CD56. These results were confirmed after the fourth and the fifth re-stimulation with PPD (60 and 75 days in culture). However, a slight reduction in the fraction of TCRGl+ and GTCSl+ cells, which paralleled a decrease in the proliferative response to hsp70, was evident (data not shown). This phenotype was peculiar to the line since one PPD but not hsp70specific, one MBP-specific and five hsp65-specific T cell lines (these five lines are not included in the frequency analysis since they were selected directly with hsp65) from this patient were > 9994 c@TCR+/CD4+ (Table 5).

Discussion In our study, two different approaches provide evidence of an increased T cell reactivity to mycobacterial hsp70 in MS. The first, namely the primary proliferative response of unselected T cells, shows that the positive responses to this antigen were

hsp-specific T lymphocytes

697

Table 4a. Proliferative responses to antigens and percentage of y6 lymphocytes ( TCRGl +) of PPD specij% T cell lines from six MS patients. The percentages of TCR ~6 ( TCRGl’) and V&l (GTCSl’) bearing subsets in the peripheral blood of the patients are also shown

Proliferative response (SI)

‘)(,of y6 in the periphery TCRGl+

GTCSl+

S.P.

7.0

0.3

T.A.

15.0

1.4

C.T.

2.0

1.9

G.A.

5.9

5.9

G.U.

4.0

2.6

C.A.

4.0

0.2

Patient

‘lo y6 in each T line

T line

PPD

hsp65

hsp70

TCRGl+

SPP2 SPP8 TAP1 TAP2 CTP16 CTPl GAP1 1 GAP3 GAP9 GUP9 GUP16 CAP9 CAP1 1

6.3 25.0 11.2 5.1 19.2 44.5 12.0 9.4 19.6 31.1 29.4 15.7 8.3

0.7 0.6 0.8 0.7 10.5 0.8 5.6 0.4 0.6 1.3 1.2 4.6 0.8

5.8 0.8 1.5 0.7 17.7 5.9 7.0 8.0 0.9 4.6 0.5 5.5 0.8

63 0 3 1 12 6 5 2 0 3 0 0 0

Table 4b. Proliferative responses to antigens and percentage of y6 lymphocytes f TCRGI ’ ) of PPD specific T cell lines from three healthy individuals. The percentages of TCR y6 ( TCRGI ’ 1 and V61 (6 TCSl i ) bearing subsets in the peripheral blood are also shown

‘lo of y6 in the periphery

‘” 76 in each T line

Proliferative response (SI)

TCRGl+

GTCSl+

T line

PPD

hsp65

hsp70

TCRG 1’

F.P.

3.1

0.6

T.L.

3.3

0.6

C.V.

2.0

0.1

FPP2 FPP8 TLP9 TLPlO CVP6 CVPl

11.8 5.0 4.2 4.0 97.6 69.4

0.5 0.9 1.5 1.6 27.6 3.6

3.0 0.7 3.2 0.6 24.4 1.9

12 8 8 10 2 2

Patient

significantly more frequent in patients with MS compared with OND patients and healthy blood donors. The reactivity to MBP parallels this pattern although the difference between MS and OND patients is not statistically significant. This pattern of response to MBP reproduces previous findings [22] indicating that no bias was introduced by the selection of patients or by the culturing system we used. The second strategy was devised in order to reduce the risks of eliciting T cell responses against contaminant antigens (derived from the Escherichia coli used to produce the recombinant proteins) [23]. For this reason we decided to select and

698

Marco Salvetti et al.

Table 5. Proliferative responses and percentages of y6 T lymphocytes ( TCRGl’) in T cell lines from patient S.P. Proliferative response (SI) hsp70

MBP

(TCRGl+)

PPD-specific T lines P2 6.3 0.7 P8 25.0 0.6

5.8 0.8

0.5 0.6

63 0

MBP-specific T line Ml3 0.9 0.7

0.9

26.8

0

PPD

hsp65

O’ ‘0 Y6

hsp65-specific T lines H601 4.3 nd H605 2.0 1.1 H611 2.7 nd H613 5.8 nd H614 6.2 nd

11.8

6.1 4.4 10.1 17.5

nd

0

nd nd nd nd

0 0 0 0

T cell lines were expanded in the presence of PPD, MBP and hsp65 (‘PPD, MBP and hsp65 specific T lines’). P2, P8 and Ml3 T lines were established from the same sample of peripheral blood mononuclear cells while all the hsp65-specific TCL were started 2 months later from a new sample. nd = not determined.

expand the T cell lines with PPD and then assess the reactivity to the hsps rather than measure the yield of hsp-specific lines generated by direct stimulation with the recombinant proteins. The majority of the lines reacted with only one hsp, supporting the validity of this strategy. Furthermore, preliminary experiments with epitope mapping show that these T lines can recognize synthetic peptides that represent sequences of the mycobacterial hsps. The results indicate that, following repeated exposures to PPD, the yield of hsp-reactive T lines is higher in patients with MS than in healthy subjects. This difference is significant when the response to hsp70 is considered. However, this observation seems to be due to a relatively low number of lines proliferating to hsp70 among controls rather than to an increase in the number of such lines in the MS group: in patients with MS, the percentage of hsp70 responsive lines is slightly lower than that of the hsp65 reactive ones (25 us 31.6%); among controls, this difference is more prominent (7.0 us 22.5%; P=O.O2, McNemar’s test). The low responsiveness to hsp70 among controls was not evident at the primary response level. Non-specific reactivities against contaminants of the hsp70 preparation may account for this observation although other explanations cannot be ruled out at the present stage: the studies performed so far on healthy donors have not conclusively addressed the issue of a possible immunodominance of either of these two proteins [ 19,24-301. Further investigations, including limiting dilution analyses, are needed in order to confirm the prevailing reactivity to hsp65 in healthy individuals and to clarify whether it

hsp-specific T lymphocytes

699

derives from diversities either in precursor frequencies or in the regulation of activation and/or suppression. These diversities may be absent in MS: members of the hsp70 family may be either primary or secondary targets of the autoaggressive immune response in this disease, depending on their immunogenicity [3 1,261 associated with their stress-induced expression in myelinated fiber tracts and in glial and brain vascular elements [lO-131. The slight increase in reactivity to hsp65 in the MS group may also be of relevance since immature oligodendrocytes expressing hsp65 have recently been described in chronic MS lesions [ 161. Epitope mapping studies, which include sequences from the human homologs of hsp65 and hsp70, are required in MS in order to ascertain the possible recognition of conserved epitopes. Following our results, we decided to restrict the analysis of the patterns of membrane TCR y6 expression to those PPD-specific T cell lines that proliferated in response to hsp70. This study did not reveal clear cut differences between lines from patients and those from controls. The weak association between hsp70 reactivity and percentage of y6 T lymphocytes, evident in some T cell lines, may suggest a potential ability of hsp70 to stimulate antigen-specific y6 T cell expansion. Nevertheless, with the exception of line SPP2, the y6 population in each line remained a minority. This may be because of the competition with afl T cells having the same antigen specificity and capable of better survival in vitro [32]. Moreover, hsp70 is not among the major mycobacterial ligands for y6 cells since these have been shown to have an apparent molecular mass of