lymphocytes in thymomas are tolerant to self-mhc

CELLULAR IMMUNOLOGY 137,438-447 (1991)

Lymphocytes in Thymomas Are Tolerant to Self-MHC

YOSHITAKA FUJII, MEINOSHIN OKUMURA,KEIJI INADA, ANDKAZUYANAKAHARA

Immunology Laboratory, First Department qf‘surgery, Osaka University Medical School, l-l-50 Fukushima, Osaka 553, Japan

Received May I, 1991; accepted June IO, 1991

Neoplastic epithelial cells of thymoma apparently retain the function of the cortical epithelial cells of normal thymus because a large number of nonneoplastic T cells in thymomas are often CD4+8+. However, the lack of medullary structure suggests that thymomas may lack some of the function of the normal thymus, especially the function of the medullary interdigitating cells to induce tolerance to self-antigens on T cells. Thymoma is often associated with autoimmune diseases, most frequently, myasthenia gravis. This suggests that the microenvironment of a thymoma may not be able to induce T cell tolerance to self-antigens. We addressed this question by testing the lymphocytes in thymomas for proliferative responses to mitogens and allogeneic or autologous stimulator cells. The lymphocytes in thymomas proliferated consistently in response to PHA and to allogeneic cells even when the response to OKT3 was undetectable. However, neither the thymoma lymphocytes nor the peripheral blood lymphocytes in these patients proliferated in response to autotogous cells. Q 1991 Academic Press, 1~.

INTRODUCTION

Human thymoma is a neoplasm of thymic epithelial cells (1). The unique character of thymoma is its association with myasthenia gravis (MG)’ (2) and the presence of various autoantibodies in the serum of patients bearing a thymoma (3). Since the thymus is the organ that confers self-tolerance on developing T cells (4-6), it is imag- inable that the breakdown of tolerance in thymoma patients is related to the tolerance- inducing function of the thymus. However, there has been no evidence to support this hypothesis.

,4 change that makes the thymus look like a thymoma occurs when cyclosporin A is administered (7, 8). Severe medullary atrophy and depletion of mature single positive cells in the thymus in cyclosporin A-treated mice resemble typical thymomas with abundant double positive cells and no medullary structure. Immunohistochemical stuldies using antibodies that distinguish cortical from medullary epithelial cells show that many thymomas derive from cortical epithelial cells (9-l 1). In both cyclosporin A-treated mice (12) and thymoma patients, autoimmune diseases occur frequently. In a recent experiment, T cells with receptors for self-I-E molecule (which are deleted in untreated mice) were shown to appear after treatment with cyclosporin A (13). The lack of medulla in both cyclosporin A-treated thymus and thymoma may be related to the incomplete delection of autoreactive lymphocytes. This agrees with studies

I Abbreviations used: AChR, acetylcholine receptor; MC, myasthenia gravis.

438

000%8749/91 $3.00 Copyright 0 1991 by Academic Press. Inc. All rights of reproduction in any form reserved

THYMOMA LYMPHOCYTES ARE TOLERANT TO SELF-MHC 439

suggesting the medullary dendritic cells as responsible for deletion of autoreactive T cells (4-6).

Because most thymomas have a large number of nonmalignant T cells which apparently have developed under the influence of the epithelial cells of the thymoma, study of T cells in the thymoma may reveal the abnormal function of epithelial cells. In addition, thymomas offer unique experimental material because it is possible to evaluate the effect of cortical epithelial cells without significant contribution from the medullary components, i.e., medullary dendritic cells and epithelial cells.

We report here the responses of lymphocytes in the thymoma to autologous and allogeneic stimulator cells in mixed lymphocyte reaction as well as their responses to mitogens. The results indicate that lymphocytes in the thymoma as well as the peripheral blood cells in these patients in fact distinguish allo- from self-MHC antigens.

MATERIALS AND METHODS

Patients

Patients are listed in Table 1. Patients 1 and 2 had moderate MG symptoms and positive serum anti-AChR antibodies as assayed using an radioimmunoassay (14). All the patients underwent extended thymectomy together with the thymoma. Histolog- ically, thymomas in Patients l-4 were mixed lympho-epithelial type. We included Patient 5 as an example of epithelial type thymoma. We studied the thymus accompanying the thymoma in Patients l-3. The thymus accompanying the thymoma in Patients 4 and 5 were involuted and we could not recover enough cell for the study. Heparinized venous blood was collected at the time of operation. Hematoxylin-Eosin- stained sections of the thymoma and accompanying thymus were examined and judged either as mixed or epithelial according to the relative abundance of lymphocytes in relation to the epithelial cell component.

Preparation of Lymphocytes

The lymphocytes were isolated by mechanically teasing and pressing the thymoma or thymus against a stainless mesh. The cells were counted and the viability was >90% as assessed by Trypan blue dye exclusion test. Since it was not possible to recover all the lymphocytes from thymoma or thymus tissue, the total cell count reported in this paper was an estimate of actual cell number.

The peripheral blood lymphocytes (PBL) were separated on Ficoll-Hypaque (Lym- phoprep, Nycomed, Oslo, Norway) density gradient and washed three times before culture.

TABLE I

Patient Characteristics

Age/sex Thymoma histology Associated disease

1 39/M 2 45/F 3 34/M 4 55/F 5 70/M

Mixed Mixed Mixed Mixed Epithelial

MG MG - - -

440 FUJI1 ET AL.

Monoclonal Antibodies

FITC-conjugated anti-CD8 (Leu2), anti-TCR (WT3 l), anti-CD45RA (Leul8), anti- HLA DR, anti-IL2R, and phycoerythrin (PE)-conjugated anti-CD4 (Leu3a), anti-CD3 (Leu4) were purchased from Beckton Dickinson (Mountain View, CA). FITC-conjugated anti-CD45RO (UCHLI) was purchased from DAK0 (Glostrup, Denmark). PE-conjugated anti-CD45RA (2H4) was purchased from Coulter Immunology (Hi- adeah, FL).

FAG’S Analysis

Lymphocytes (0.5- 1 X 106) in 100 yl of PBS with 0.1% NaN3 were mixed with 10 ~1 each of a combination of FITC- or PE-conjugated monoclonal antibodies. After 30 min on ice (for the TCR/CD3 combination, PE-anti-CD3 was added 15 min after a’dding FITC-anti-TCR as suggested by the manufacturer), the cells were washed twice and propidium iodide was added at a final concentration of 5 pg/ml before the analysis using FACS (Beckton Dickinson). Data from 10,000 cells were collected and analyzed after gating according to forward and side scatter pattern using LYSIS program.

Cell Culture

Lymphocytes (3 X 105) from the thymoma, thymus, or peripheral blood were cultured in triplicate in 200 ~1 of complete media (RPM1 1640 in 10% fetal calf serum, 2 mA4 glutamine, 5 X lo-’ M 2-ME) with PHA (Sigma, 1 pg/ml), OKT3 (Ortho Diagnostics, Raritan, NJ, 10 rig/ml), or 3 X lo5 stimulator cells which had been treated with mitomycin C (MMC, 25 pg/ml) for 30 min and washed three times. These con- centrations had been determined by preliminary experiments. The allogeneic stimulator cells were from a healthy volunteer. HLA typing of the thymoma patients was not done but assumed to be sufficiently different because the peripheral blood cells of these patients responded to MMC-treated cells from the volunteer. The cells were pulsed with 0.5 &i of [3H]thymidine on Day 3 and harvested 6 hr later. The means and standard errors of incorporated thymidine are shown in the figures.

RESULTS

Lymphocytes Recovery from the Thymoma

The number of cells recovered from a thymoma is an estimate of lymphocyte gen- eration in the thymoma and varied from 1.1 X lo6 in Case 5 to as many as 1.5 X lo9 pelr gram wet tissue in Case 3 (second column of Fig. 1). The latter compares to the cell recovery from a neonatal thymus. An epithelial thymoma (Case 5) contained very few cells.

S&ace Antigen Phenotype of Lymphocytes in the Thymoma

The results of two-color FACS analysis of the thymoma lymphocytes are shown in Table 2. The contour plots of two-color FACS analysis in Patient 1 are shown in Fig. 1 as an example. The frequency of double positive (CD4+CD8+) cells in thymomas varied from 1.3% in Case 5 to 83.1% in Case 1. In an earlier report, we noted that the mixed type thymomas without MG tended to have high frequencies of double positive cells (15). Case 4 thymoma is an exception according to this notion. The cell recovery from this thymoma ( 179 million/g) was less than the other mixed type thymomas and


a.

CD4

b.

cm

d.

co3

e.

T cd I antigen recmtor

CD45RO

CD45RO

DR

FIG. 1. Surface antigen phenotype of lymphocytes from the thymoma, thymus, or peripheral blood in Patient 1. Staining with FITC/PE-labeled antibodies to (a) CD8/CD4, (b) TCR/CD3, (c) CD45RA/CD3, (d) CD45RO/CD3, (e) CD45RO/CD45RA, (f) HLA-DR/CD3. Numbers indicate the percentage of cells belonging to each quadrant.

this thymoma may be an intermediate between the mixed and epithelial type. The small number of cells recovered from Case 5 thymoma and the single positive phenotype of recovered cells suggest that most of the cells in this thymoma were contam-

442 FUJI1 ET AL.

TABLE 2

Phenotype of Lymphocytes in the Thymoma or Thymus

Thymoma patient CD4+8+ CD4+8- CD4-8+ WT3 1 +CD3+

1 83.1 8.5 5.9 54.6 2 45.0 26.3 26.2 63.7 3 78.5 4.9 14.7 52.3 4 30.7 43.4 16.5 31.4 5 1.3 49.2 11.3 nt

CD45RA+CD3+ CD45RA+RO- CD45RA-RO+

1 4.0 4.0 86.4 2 5.9 nt nt 3 8.4 1.2 81.0 4 10.5 4.2 65.9

DR+CD3+ DR+CD3- IL2R+CD3+

1 2 3 4

Thymus patient0

0.6 0.6 0 0.4 0.4 0.4 0.3 1.3 0.2 8.0 1.5 0.6

CD4+8+ CD4+8- CD4-8+ WT31+CD3+

1 37.9 18.6 13.3 45.4 2 35.7 27.2 19.8 64.8 3 71.6 7.8 4.6 48.5

CD45RA+CD3+ CD45RA+RO- CD45RA-RO+

1 12.1 22.2 55.6 2 26.2 nt nt 3 3.2 1.1 72.5

DR+CD3+ DR+CD3- IL2R+CD3+

1 2.0 6.9 0.4 2 1.7 8.2 4.1 3 0.2 1.9 0.7

Note. Percentage of cells with indicated phenotype as determined by two-color FACS analysis. Gate was set to include cells with the size of lymphocytes and without propidium iodide staining.

a The uninvolved thymus adjacent to the thymoma.

inating peripheral blood cells. The responses to mitogens also indicated this possibility (see Responses to Mitogen). Thus, the thymomas we report here were heterogeneous with1 regard to the frequencies of double positive cells.

Staining of the thymoma lymphocytes with anti-T cell receptor (WT31) and anti- CD3 showed variable levels of expression of TCR/CD3 molecules (Fig. 1). The percentage of WT3 1 + cells shown in Table 2 includes both WT3 l’Ow and WT3 1 high cells as shown in Fig. 1.

Staining with antibodies to CD45 isotypes showed mutually exclusive expression of CD45RA and CD45RO on the thymoma lymphocytes as in the PBL (Fig. 1).


CD45RA+CD3+ cells were rare in thymomas even in Cases 2 and 4 with relatively large proportions of single positive cells (52.5 and 59.9%, Table 2) indicating that there are significant number of single positive CD45RO+CD45RA- cells. This was also evident in the staining with anti-CD45RA and anti-CD45RO (UCHLl).

The phenotype of the PBL in the patients with thymoma was unremarkable suggesting that the large number of double positive cells generated in the thymoma are not allowed to emigrate from the thymoma or do not survive in the periphery.

Responses to PHA and OKT3

The lymphocytes from all five thymomas responded to PHA. Except in Case 5, the magnitude of the response to PHA was lower than that of PBL that were cultured in the same experiment. The lymphocytes from Case 5 thymoma responded to PHA as vigorously as the peripheral blood cells.

The responses of thymoma lymphocytes to OKT3 were undetectable in Cases l- 3. The PBL in these patients showed similar responses to PHA and OKT3. The response of thymoma lymphocytes to OKT3 was low but detectable in Case 4 and was very good in Case 5. Apparently, the cells with higher frequencies of single positive cells responded better to OKT3 (Fig. 2).

Responses to Autologous and Allogeneic Stimulator Cells

The lymphocytes from thymomas consistently proliferated in response to MMC- treated allogeneic peripheral blood cells. The magnitude of the mixed lymphocyte reaction in Cases 4 and 5 even exceeded those of peripheral blood cells. In contrast, the response to autologous blood cells was very low in all thymomas. The responses to autologous cells were l/4 (Case 3) to l/204 (Case 5) of those to allogeneic cells. It was below the background in Case 2. The low background proliferations of thymoma cells (see legend to Fig. 3) also suggest the absence of autoreactive T cells. Overall, the mixed lymphocyte reaction in the lymphocytes in the thymomas showed almost complete unresponsiveness to self-MHC.

We evaluated the responses of the lymphocytes from the apparently normal thymus accompanying the thymoma in Cases l-3. The cells from the thymus also showed clear preference to allogeneic than autologous stimulator cells. The involuted thymus in other cases did not yield enough cells for the experiment.

DISCUSSION

Using mixed lymphocyte reaction, we showed that lymphocytes from five thymomas, with or without myasthenia gravis, did not detectably react to MMC-treated autologous blood cells while consistently responding to allogeneic cells. The identity of cells in the thymoma that are proliferating in these cultures is not definitely clear and there are three possibilities. First, it may be the malignant epithelial cells or other nonlym- phoid cells that are proliferating. This is not likely because epithelial cells in the thymoma are not readily released in the cell suspension unless enzyme digestion is used (less than 1% keratin positive cells in the thymoma cell suspensions) and it is not likely that they discriminate between allo- and self-MHC (Fig. 3).

Second, it could be argued that the contaminating peripheral blood cells were responsible for most of the proliferative responses; the thymoma lymphocytes being premature to respond to alloantigens. This possibility cannot be ruled out in Case 5

444 FUJI1 ET AL.

Cell Double

Positive Proliferative Response to

Mitogen

728

1514

179

1.1

785

30.7

1.3

PHA oKr3 f+IAoKn

1%. 2. Responses to mitogens of lymphocytes from the thymoma and peripheral blood. Numbers on the left of the figure are the number of cells recovered from each thymoma (left column) and the frequencies of double positive cells in the cell suspension of thymoma (right column). Values are the mean and standard error of tritiated thymidine uptake (cpm) stimulated with PHA (closed bars) or OKT3 (open bars).

where the response of the thymoma cells was similar to that of PBL. However, in other four cases, the response to anti-OKT3 was much lower in the thymoma cell cultures than in PBL cultures even when the response to PHA was comparable. This indicates that the level of PBL contamination was not large in these thymomas. Also,

THYMOMA LYMPHOCYTES ARE TOLERANT TO SELF-MHC

Thymus 9303 t PBL 8300 t

Thymoma

12co3, I _ Thvmoma

Thymus i

Stimulator cells FIG. 3. Responses to allogeneic (closed bars) or syngeneic (open bars) stimulator cells of lymphocytes

from the thymoma, thymus, and peripheral blood. Values are the mean and standard error of tritiated thymidine uptake (cpm). Counts of thymoma cells only were (Cases 1-5) 84, 349, 346, 1095, and 1389 cpm. Counts of thymus cells only were (Cases l-3) 452, 516, and 231 cpm. Counts for PBL only were (Cases l-5) 442, 642, 412, 800, and 95 cpm. These, as well as the counts of MMC-treated stimulator cells (< 100 cpm), have been subtracted.

446 FUJI1 ET AL

the frequency of red blood cells in the thymoma cell suspension (before Ficoll sepa- ration) was below 2% except in Case 5 indicating that the contaminating PBL would be extremely low.

Third, the result may suggest that the microenvironment of the thymoma (most of which consists of epithelial cells of cortical origin) can delete or inactivate T cells that are reactive to self-MHC antigens. This agrees with reports that thymic epithelial cells can induce T cell tolerance (16-20) although it may be less effective than dendritic cells in doing so (5,2 1,22).

Clinically, this agrees with our experience that general graft-versus-host type disease is not frequent in thymoma patients. Also, the PBL of the thymoma patients did not proliferate in response to autologous stimulator cells in the mixed lymphocyte cultures (Fig. 2). Thus, myasthenia gravis in thymoma patients is probably not a consequence of a graft-versus-host disease as in the reported cases of myasthenia gravis after bone marrow transplantation (23). However, it should be pointed out that the maturation of T cells in the thymoma may not be complete. It is possible that even if the thymomas have allowed autoreactive T cells to mature, the response of these T cells to autoantigen may not be apparent until the later stage of maturation which may not necessarily take place in the thymoma.

The MHC antigens are expressed in thymomas as in the normal thymic cortex (11) and these MHC antigens could have been the antigen that deleted self-MHC reactive T ‘cells. We have not, however, tested the response of thymoma lymphocytes to autoantigens other than self-MHC. In the normal thymus, cortex is believed to be relatively “sterile” (24) and do not usually permit penetration of macromolecules in the extracellular fluid. In contrast, the medulla is bathed with circulating antigens and dendritic cells can present these antigens (25). Whether a thymoma allows penetration of macromolecules is not known and whether thymic epithelial cells can present nom- inall antigens is controversial (25-27). Thymomas lack the medulla and DR+ non-T cells with dendritic morphology are rare in the cell suspension from thymomas (Fig. 1 and Table 2) and it is possible that circulating antigens are not effectively presented in the thymoma. In this case, the thymoma lymphocytes have little chance to encounter circulating autoantigens during their development and could react to autoantigen if they can mature and leave the thymoma.

It is difficult to imagine that all the insoluble membrane proteins in the body are expressed in the thymus. The T cell tolerance to membrane proteins like AChR may be maintained by a mechanism different from the one that involves the thymus (28- 34). As for the AChR, there have been conflicting reports on the presence of AChR in the thymus. Whether AChR protein is expressed in the thymus at a sufficient level to delete every AChR-reactive T cell is doubtful (35-38). Therefore, tolerance to AChR may be maintained in the periphery. From the results of our present study, a thymoma does not seem to break this tolerance centrally by allowing self-MHC reactive T cells to mature and emigrate from the thymoma.

REFERENCES

1, Rosai, J., and Levine, G. D., “Atlas of Tumor Pathology,” 2nd Series Fascicle 13. U.S. Armed Forces Institute of Pathology, 1976.

2. Monden, Y., Nakahara, K., Kagotani, K., Fujii, Y., Masaoka, A., and Kawashima, Y., Ann. Thoruc. Surg. 38, 46, 1984.

3. Aarli, J. A., Lefvert, A. K., and Tonder, O., J. Neuroimmunol. 1, 42 I, 198 1. 4. Kappler, J. W., Roehm, N., and Marrack, P., CeN 49, 273, 1987.


5. Sprent, J., Lo, D., Gao, E-K., and Ron, Y., Immunol. Rev. 101, 173, 1988. 6. Schwartz, R. H., Cell 57, 1073, 1989. 7. Demetris, A. J., Nalesnik, A., Kunz, H. W., Gill, T. J., III, and Shinozuka, H., Transplantation 38,239,

1984. 8. Hattori, A., Kunz, H. W., Gill, T. J., III, and Shinozuka, H., Am. J. Pathol. 128, 111, 1987. 9. Thomas, J. A., Willcox, H. N. A., and Newsom-Davis, J., J. Neuroimmunol. 3, 319, 1982.

10. Takacs, L., Savino, W., Monostori, E., Ando, I., Bach, J.-F., and Dardenne, M., J. Immunol. 138,687, 1987.

11. Willcox, N., Schluep, M., Ritter, M. A., Schuurman, H. J., Newsom-Davis, J., and Christensson, B., Am. J. Pathol. 127, 447, 1987.

12. Sakaguchi, S., and Sakaguchi, N., J. Immunol. 142, 471, 1989. 13. Gao, E.-K., Lo, D., Cheney, R., Kanagawa, O., and Sprent, J., Nature 336, 176, 1988. 14. Lindstrom, J., Einarson, B., and Tzartos, S., In “Methods in Enzymology,” Vol. 74, pp. 432. Academic

Press, New York, 1981. 15. Fujii, Y., Hayakawa, M., Inada, K., and Nakahara, K., Eur. J. Immunol. 20, 2355, 1990. 16. Flajnik, M. F., Du Pasquier, L., and Cohen, N., Eur. J. Immunol. 15, 540, 1985. 17. Jordan, R. K., Robinson, J. H., Hopkinson, N. A., House, K. C., and Bentley, A. L., Nature 314,454,

1985. 18. Jenkinson, E. J., Jhittay, P., Kingston, R., and Owen, J. T., Transplantation 39, 331, 1985. 19. Ohki, H., Martin, C., Corbel, C., Coltey, M., and le Douarin, N. M., Science 237, 1032, 1987. 20. Salaun, J., Bandeira, A., Khazaal, I., Calman, F., Coltey, M., Coutinho, A., and Le Douarin, N. M.,

Science 247, 147 1, 1990. 2 1. Marrack, P., Lo, D., Brinster, R., Palmiter, R., Birkly, L., FIavell, R. H., and Kappler, J., Cell 53, 627,

1988. 22. Sprent, J., Gao, E.-K., and Webb, S. R., Science 248, 1357, 1990. 23. Lefvert, A. K., and Bjorkholm, M., N. Engl. J. Med. 318, 170, 1987. 24. Van Ewijk, W., Am. J. Anat. 170, 31 I, 1984. 25. Kyewski, B. A., Fathman, C. G., and Kaplan, H. S., Nature 308, 196, 1984. 26. Marrack, P., McCormack, J., and Kappler, J., Nature 338, 503, 1989. 27. Lorenz, R. G., and Allen, P. M., Nature 340, 557, 1989. 28. Adams, T. E., Alpert, S., and Hanahan, D., Nature 325, 223, 1987. 29. Lo, D., Burkly, L. C., Widera, G., Cowing, C., Flavell, R. A., Palmiter, R. D., and Brinster, R. L., Cell

53, 159, 1988. 30. Burkly, L. C., Lo, D., Kanagawa, O., Brinster, R. L., and Flavell, R. A., Nature 342, 564, 1989. 3 1. Morahan, G., Allison, J., and Miller, J. F. A. P., Nature 339, 622, 1989. 32. Rammensee, H.-G., Kroschewski, R., and Frangoulis, B., Nature 339, 541, 1989. 33. Corbel, C., Martin, C., Ohki, H., Coltey, M., Hlozanek, I., and Le Douarin, N. M., Int. Immunol. 2,

33, 1990. 34. Burkly, L. C., Lo, D., and Flavell, R. A., Science 248, 1364, 1990. 35. Kao, I., and Drachman, D., Science 195, 74, 1977. 36. Kirchner, T., Hoppe, F., Muller-Hermelink, H. K., SchaIke, B., and Tzartos, S., Lancet 1, 218, 1987. 37. Nicholson, G. A., and Appel, S. H., J. Neurol. Sci. 34, 101, 1977. 38. Murphy, A., Drachman, D., Satya-Murti, S., Pestronk, A., and Eggleston, J., Muscle Nerve3,293, 1980.

lymphocytes in thymomas are tolerant to self-mhc

Documents