- Email: [email protected]
T-cell subsets in human lymphocytes maintained in IL-2 medium after PHA or mixed lymphocyte reaction activation
CLINICAL
IMMUNOLOGY
T-Cell Medium JAMES
AND
IMMUNOPATHOLOGY
27, 444-451 (1983)
Subsets in Human Lymphocytes after PHA or Mixed Lymphocyte T. KLJRNICK,?
ANTHONY R. WARRENS, CAROL P. LEARY
Maintained in IL-2 Reaction Activation’
RICHARD
A. MOSCICKI,
ANI>
The culture of human T lymphocytes in interleukin-2 (IL-2) containmg growth factor medium results in a significant shift in the T-lymphocytes subsets isolated from such cultures at weekly intervals. If normal peripheral blood mononuclear cells are stimulated with phytohemagglutinin (PHA) or in a mixed lymphocyte reaction (MLRI. the resulting T lymphoblasts can be propagated in growth factor medium. Staining of the cultured cells with monoclonal antibodies was evaluated by indirect immunofluorescence on a laser-activated flow cytometer (Ortho Spectrum 111). The antibodies used were: OKT3 (mature T lymphocytes). OKT4 (helper/inducer 7‘ lymphocytes). OKT8 (cy,totoxic/suppressor T lymphocytes. OKTIO (immature and “activated” lymphocytes). OKTI la (cells which rosette with sheep erythrocytesl. and OKIa-I (HLA-DR constant region). Both PHA and MLR activation resulted in initial preservation of the OKT4 + subset predominance over OKT8 + T lymphocytes noted on normal circulating blood lymphocytes. However. during culture in T-cell growth factor medium. there was a progressive increase in the percentage of OKT8t cells. and a concomitant decrease in OKT4 + lymphoblasts. The increase in OKT8 + cells in the MLR-stimulated culture\ was paralleled by an increase in specific cell-mediated cytotoxicity against the stimulating lymphocyte population. In addition to the shift in T-lymphocyte subset. there wa\ virtual 100% staining with OKT3 and OKTl la, indicating the T-cell nature of the proliferating cells. OKTIO which was present on a small subset of fresh blood lymphocyte\ appeared rapidly in stimulated cultures, and was retained on virtually all lymphoblasts of either OKT4 + or OKT8 + subset. OKIa-I cells increased slowly in PHA-stimulated cultures. HLA-DR+ T cells were detected earlier in MLR cultures. The activation of T lymphocytes results in a significant increase in the number of molecules of OKTI la bound per cell. in concert with the increased avidity of T lymphoblasts for sheep erythrocytes. The significant change in the phenotype and function of lymphoblasts isolated from long-term cultures demonstrates the importance of monitoring cultures. and the potential hazards in equating a cultured cell population with a freshly isolated one.
INTRODUCTION Evaluation of human lymphocytes in the past decade has relied heavily on mitogen activation of peripheral blood lymphocytes. Many functional characteristics have been demonstrable from the lectin-stimulated peripheral blood cells (I-5). More recently, it has been shown that these activated lymphocytes can be maintained and propagated in culture by the addition of soluble factors present I This work was supported in part by Grant AI-1741 1 and CA29601 from the NIH of Health and in part was aided by Basil O’Connor Stacter Research Grant 5-324 from March of Dimes Birth Defects Foundation. Dr. Moscicki was supported by PHS Training Grant T32-A107040. z To whom correspondence should be addressed: Cox 5. Massachusetts General Hospital, I‘ruit Street, Boston, Mass. 02114. 444 0090-1229/83 $1.50 Copyright 0 1983 by Academic Press. Inc. All rights of reproduction in any form rewved
HUMAN
T-CELL
SUBSETS
GROWN
IN IL-2
445
in the supernatants of PHA-stimulated cultures (P-SUP) (6). This activity has been designated as interleukin-2 (IL-2) (7). In the early functional reports on IL2-maintained T-cell blasts, there was an indication that mixed lymphocyte reaction T-cell cultures from mice showed increasing cytotoxic activity at sequentially later time points in culture (8). While T cells which have been cultured and even cloned from IL-2-containing medium have been capable of demonstrating multiple functional capacities (9-12), these early mouse studies suggested that in a bulk culture there might be selection for a particular T-cell subset in IL-2 medium. Monoclonal antibodies which distinguish human T-cell subsets shed new light on T-cell interactions in vivo and in vitro (13-18). By immunofluorescence on a laser-activated flow cytometer (19), it has been possible to assess large numbers of cells stained with these monoclonal reagents and establish changes in the Tcell subsets over time in culture. We have assessed phenotypic changes in human T cells stimulated with PHA, or in a mixed lymphocyte reaction (MLR) after continued culture in IL-2 containing medium (P-SUP). MLR-stimulated blasts were assessed for specific cytotoxic activity after culture in P-SUP. MATERIALS
AND METHODS
PHA stimulation. Normal human peripheral blood mononuclear cells were isolated from Ficoll-Hypaque gradients (20). Lymphocytes at lOVm1 were stimulated in t&o with optimal doses of PHA ( 1 ug/ml of leucoagglutinin, Pharmacia Fine Chemicals, Uppsala, Sweden) for 3 days in RPM1 1640 containing 5% normal human serum, antibiotics, and 5 x 10.’ M 2-mercaptoethanol in 5% CO,humidified 37°C atmosphere. Cells were then resuspended at 10’ blasts/ml in Tcell growth factor medium (interleukin-2-containing medium). Cells were diluted back to lOVm1 every third day. Growth factor medium. The growth factor medium (P-SUP) was prepared from human tonsil lymphocytes stimulated with optimal doses of PHA. Growth factor was produced with and without residual free PHA. Medium was prepared as previously described, by collecting supernatants from stimulated tonsils after 48 hr (2 I). A lectin-free supernatant was also prepared by washing tonsil leukocytes after only 2 hr in the presence of PHA, and reculturing them in a lectin-free RPMI-2% human serum medium, and collecting the final medium after 48 hr. Such washed growth factor medium is free of PHA as tested in a mitogenesis assay. Both the PHA-containing and lectin-free medium were used at 25% dilutions in RPM1 1640 containing 5% pooled human serum. Mixed lymphocyte reaction cultures. Mixed lymphocyte reactions T-cell cultures were established by coculturing Ficoll-Hypaque-separated mononuclear cells at 10Vml with a similar number of X-irradiated (2000 R) allogeneic PBLs. After 6 days, the blast cells were collected on a Percoll density gradient as previously described (22) and cultured in lectin-free growth factor medium. At weekly intervals, a Vr-release assay (21) was performed on both autologous and appropriate stimulator allogeneic PHA blasts to determine specific cytotoxic activity of the MLR T cells. The targets were PHA blasts grown in T-cell growth factor medium for 7-14 days. Percentage specific lysis was calculated using the formula: % specific lysis = (1 - [(M - S) - (E - S)I(M - S)]} x 100, where M =
446
KURNICK
ET AL.
maximum release, S = spontaneous release, and E = experimental value of effector combination being tested. Assessment of T-cell subsets. At the initiation of culture and thereafter at weekly intervals, the cultured cells were stained with monoclonal antibodies as outlined below. Lymphocytes, IOh per sample, were stained for 30 min with 5 ~1 of monoclonal antibodies, followed by a phosphate-buffered saline (PBS) wash and a second incubation on ice with fluoresceinated goat anti-mouse immunoglobulin at a l/40 dilution (Cappel Labs, Cochranville. Pa.). The monoclonal antibodies (“Orthoclone,” Ortho Pharmaceuticals, Raritan, N.J.) are designated as OKT3. OKT4, OKT8, OKTlO, OKTlla, and OKla-1. After washing, the stained cells were fixed for IO min in buffered 1% paraformaldehyde and finally washed and stored in PBS until they were evaluated. Fixed cells could be observed up to 1 week following staining. Evaluation of fluorescence staining was carried out on a prototype cytofluorograph (Ortho Instruments, Westwood, Mass.). This laser-activated flow cytometer allows rapid analysis of cells based on size and surface morphology and fluorescence intensity (20). Data generated were stored on an Apple II computer interlinked with the flow cytometer, permitting standardization of the data from one day to the next. The threshhold for positive fluorescence was selected so that less than 2% of cells incubated with control mouse ascites and fluoresceinated goat anti-mouse reagent were scored as positive. RESULTS PHA Cultures. Figure 1 demonstrates the change in the percentage of staining with various monoclonal anti-T-cell reagents over a 4-week period in culture in P-SUP. At the initiation of culture, the T4 subset is the dominant T-cell population and the percentage of T&bearing cells comprises approximately half the percentage of T4 cells. T3 and Tlla, which are present on virtually all peripheral blood T lymphocytes, constitute the majority of the peripheral blood cells isolated from the Ficoll-Hypaque gradient. The antigen recognized by OKIa-I (which is a constant region of the HLA-DR molecule) (16, 23) is virtually lacking from the resting T lymphocyte, but appears here on a small percentage of cells as would be expected in a sample which contains some B cells and macrophages. TIO antigen is present on less than 15% of the freshly isolated lymphocytes. After the first week of activation, TlO appears on virtually all the cells and remains elevated thereafter. After 1 week the percentage of T4 cells begins to decline until at 4 weeks this population constitutes less than 10% of the T-cell blasts which are proliferating in culture. Concomitantly, the percentage of T8 lymphoblasts rise dramatically and constitutes the vast majority of the T-cell blasts at the end of this l-month culture period. HLA-DR-antigen-bearing lymphocytes rise and this antigen is present on virtually all of the lymphoblasts at the end of the culture. The selection for T cells is demonstrated by the virtual uniform staining with anti-T3 and Tl la on all of the cells. The selection toward T8-bearing cells occurs in IL-2-containing medium either with or without residual PHA. MLR Cultures. Figure 2 shows the specific cytotoxic activity of MLR blasts tested after 7, 14, and 21 days in culture. This example is representative of what
HUMAN
0’
’ 0
T-CELL
SUBSETS
7
GROWN
14 DAYS
IN
447
IL-7
21
L 1 28
I 21
4 28
tN (-XJl Tt JRF
IOO-
20.
I O
b
I 0
7
14
DAYS IN CULTURE FIG. 1. (a) Change in OKT3, OKT4, and OKTS staining of PHA blasts grown in IL-2 medium. The percentage positive (with standard deviations shown by vertical bars) for human peripheral blood lymphocytes assayed in culture at weekly intervals. (Each point represents the average of 6-16 samples). FIG. I. (b) Change in OKTll, OKTIO, and OKla-I (D) staining of PHA blasts grown in IL-2 medium. The percentage positive (with standard deviations shown by vertical bars) for human peripheral blood lymphocytes in culture at weekly intervals. (Each point represents the average of 616 samples).
448
KURNICK
ET
AL
70-
60 ul 3j
50-
> J "
4ci-
c 2
30-
% 8
*O IO-
1
I
I
I
I
I
I
1.5
3
6
12
25
50
TARGET
: EFFECTOR
RATIO
FIG. 2. Change in specific cytotoxicity in MLR blasts maintained in IL-2 medium. Specific i’C~release at various effector:target ratios are shown for cells cultured for 7 (+ ). 14 (0). and 71 dayz f-t+).
was noted in 15 different MLR cultures in which all showed a progressive increase in Wr release during culture in P-SUP. The changes in T-cell surface markers as detected with the monoclonal antibodies (Fig. 3) was similar to that seen with the PHA-stimulated cells, except that at 7 days, the T4+ cells were more prevalent and T8+ cells were accordingly fewer in the MLR. However, as with the PHA blasts, the T4 predominance quickly was reversed so that by 3 weeks in P-SUP. the percentage of T8+ cells was greater than 70% and T4+ were in the low 20% range. Also, OKIa-I staining appeared on significant numbers of cells at Day 7 in the MLR cultures. Assessment of individual cell jluorescence. The flow cytometer distinguishes positive- from negative-staining cells and permits evaluation of the intensity of fluorescence. The fluorescence intensity is indicated by fluorescence channel. with brighter cells falling in higher numbered channels. By this criterion. the amount of Tlla antigen increases dramatically after T-cell activation. Freshly isolated peripheral blood T cells stain rather weakly with OKTlla (mean fluorescence channel = 40), while activated cells stained more brightly with OKTlla than with any of the other monoclonal antibodies used in this study (mean fluorescence channel = 129) at Day 7. The intensity of staining with OKT4 likewise increases after activation, but staining with OKT3 remains relatively unchanged.
HUMAN
0
T-CELL
SUBSETS
;
1’4
DAYS
449
GROWN IN IL-?
2;
IN CULTURE
FIG. 3. Change in phenotype of MLR blasts during growth factor medium culture. Surface fluorescence was determined on a flow cytometer using monoclonal antibodies OKT3 (3), OKT4 (4). OKT8 (8). and OKIa-I (D). MLR blasts were harvested after 6 days and placed in growth factor medium, and aliquots stained on Days 7, 14. and 21. Standard deviation shown in vertical bars (average of 6-16 samples).
DISCUSSION
The initial reports on the use of soluble growth factors to maintain growth in human T lymphocytes demonstrated the extreme selective nature of PHA-induced T-cell growth factor for T cells. Indeed, the first observations were made on bone marrow cells, where a minor mature T-cell population became the predominant cell type during culture (6). With the recent availability of antibodies which can distinguish human T-cell subsets (13-19), the possibility was presented of observing possible T-cell subset selection during in vitro culture in supernatants from PHA-stimulated lymphocytes. The overall T-cell selective nature of the PHA-induced growth factor medium is confirmed by the increase in the percentage of T3- and Tlla-bearing cells (although these cells also predominate early in culture), but our results show that while both T4- and T&bearing lymphocytes can be propagated in culture, there is a progressive selection for cells bearing the T8 marker, and lacking the T4 antigen. These findings emphasize the potential perils in evaluating a cultured cell population as a model for in vivo lymphocyte function. The anti-HLA-DR constant region-specific antibody (OKIa) confirms the progressive rise in the cell population bearing DR antigen which has been noted by others (16, 23). It is of interest to note here, however, that the DR-bearing cells must also bear T8, and that a previous communication using soluble antigenstimulated T cells demonstrated that T4-bearing lymphoblasts can likewise be DR positive (9). Thus, the presence of DR molecules on the surface of T lymphoblasts
450
KURNICK
ET
AL.
cannot be used as an indicator of T-cell subset. The TIO marker rose in percent after T-cell stimulation, and remained elevated throughout the culture period. While this marker was initially noted primarily on immature thymocytes (24). it is clear that it is also present on activated normal T cells, and its presence with other mature T-cell markers should not be equated with abnormality in maturation. Also, the increased intensity of staining with anti-T1 la antibody may be correlated with the more vigorous sheep red blood cell-rosetting capacity of activated T cells (25). While intensity of OKT4 staining also increases. OKT3 intensity per cell does not change significantly after activation, indicating that the increase in antigen per cell is not merely an indication of increased cell size. These data demonstrate that T-cell subset selection does occur during prrtlonged culture in IL-2-containing medium when both T4- and T8+ cells are present. In our previous study where only T4’ cells were activated with soluble antigen, these cells could still be maintained for many months in culture (9) so the selective capacity of the IL-2 medium does not exclude prolonged T4’ subset culture, but by some still unknown mechanism, selects for TX- cells when both subsets are present. We cannot state whether the eventual T8+ predominencc results from more rapid proliferation of these cells, or some selective pressure exerted by the T8+ cells against the OKT4+ cells (such as suppression or cytotoxicity). In MLR-induced T-cell proliferation. we have been able to confirm a functional change which accompanies the phenotypic shift toward T8 r cells. The increased cytotoxic activity is consistent with the T8’- phenotype (17). We have previously reported that PHA blasts which contain T4+ and T8’- cells can both help and suppress in a B-cell maturation assay with pokeweed mitogen stimulation (9). Help by PHA blasts can be demonstrated on purified B even when T8’ cells begin to predominate. The suppression which can also be observed is difficult to interpret since even pure T4+ cells can induce suppression of B-cell maturation when an inducible suppressor T cell is present (9). Since even the use of cloned T-cell blasts does not clarify the precise mechanism of suppression detected in multiple cell mix experiments, we have elected not to elaborate on these earlier findings when evaluating a heterogenous T-cell population. The absorbtion of IL2 by activated T cells may play a role in in \*itrc/ suppression (26). However, the cytotoxic T-cell function is straightforward and the increase during P-SUP culture parallels the shift toward T8+, T4m cells. A final note on the culture is the frequent slowing of proliferation during the long-term culture. The Cweek time period observed in this study was necessitated by the cessation of growth of the cultures at approximately this time point. Further culture required the addition of irradiated fresh peripheral blood cells and lectin (22) to reinstate soluble IL-2 responsiveness. We have elected not to report in detail on subsets of lymphoblasts restimulated with irradiated cells and PHA. but note here that there is no further alteration in T-cell subsets after such stimulation, and after several months in culture with combinations of irradiated cells and soluble factor, a predominant T8+ population is retained with a small residual T4+ population. Specificity of MLR cytotoxicity is similary maintained. While the cessation in growth at the 4-week time point could indicate the need
HUMAN
T-CELL
SUBSETS
GROWN
IN IL-2
451
for a T4+ population to retain proliferative capacity of the T8+ cells, a similar slowing in growth has also been noted in soluble antigen-stimulated T-cell cultures where only T4+ cells are present. IL-2 responsiveness thus relies on more complex culture conditions. ACKNOWLEDGMENTS The authors gratefully acknowledge the generous gift of monoclonal antibodies from Drs. Gideon Goldstein and Patrick Kung of Ortho Pharmaceuticals, and the provision and technical support for the Spectrum III prototype flow cytometer from Ortho Diagnostics. We also appreciate the advice and support of Dr. Robert B. Colvin in carrying out these experiments. The technical support of Mr. Paul Steinwachs is gratefully acknowledged. The expert secretarial assistance of Ms. Susan Adams and Ms. Michele Angelo was much appreciated.
REFERENCES I. @eaves, M. F., Janossy, G., and Doenhoff, M., J. Exp. Med. 140, 1, 1974. 2. Waldman, T. A., Durm, M., Broder. S., Blackman, M., Blaese, R. M., and Strober, W.. Lancet 2, 609, 1974. 3. Fauci, A. S., and Pratt, K. R., J. Exp. Med. 144 674. 1976. 4. Keightley, R. G., Cooper, M. D., and Lawton. A. R., J. Immunol. 117, 1538, 1976. 5. Kumick. J. T.. Bell, C., and Grey, H. M., Stand. J. tmmunol. 5, 771, 1978. 6. Morgan, D. A., Ruscetti, F. W., and Gallo, R., Science 193, 1007. 1976. 7. Letter to the Editor, J. Immunol. 123, 2928, 1979. 8. Gillis, S., and Smith, K. A., Nature (London) 268, 154, 1977. 9. Kurnick, J. T., Hayward, A. R., and Altevogt, P.. J. Immunol. 126, 1307. 1981. IO. Lotze, M. T., Strausser, J. L., and Rosenberg, S. A., J. tmmunol. 124, 2972, 1980. II. Bach, F. H.. Inouye, H.. Hank, J. A., and Alter, B. M., Nature (London) 281, 307, 1979. 12. Fleisher, T. A.. Greene, W. C., Uchiyama, T., Goldman, C. K., Nelson, D. L., Blaese. M., and Waldmann, T. A., J. Exp. Med. 154, 156, 1981. 13. Reinherz, E. L.. Kung, P. C., Goldstein, G.. and Schlossman, S. F., Proc. Nar. Acad. Sci. USA 76, 4061, 1979. 14. Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman. S. F., J. Immunol. 123, 2894, 1979. 15. Reinherz, E. L., Kung, P. C.. Goldstein, G.. and Schlossman, S. F., J. tmmunol. 123, 1312, 1979. 16. Reinherz, E. L., Kung, P. C., Pesando, J. M.. Ritz, J.. Goldstein, G.. and Schlossman, S. F., J. Exp. Med. 150, 1472, 1979. 17. Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman, S. F.. J. tmmunol. 124, 1301. 1980. 18. Kung, P. C.. Tulle, M. A., DeMaria. M. A., Butler, M. S., Lifte. J., and Goldstein, G., Transplant. hoc. 12, (Suppl.), 141, 1980. 19. Hoffman, R. A.. Kung, P. C., Hansen, W. P.. and Goldstein, G., Pror. Nat. Acad. .Sci. USA 77, 4914. 1980. 20. Boyum, A., Stand. J. tmmunol. 5 (Suppl.), 9, 1976. 21. Kurnick, J. T.. Gronvik, K.. Kimura, A. K.. Lindblom, V. B., Skoog, V. T., Sjoberg. 0.. and Wigzell, H.. J. tmnzunol. 122, 1255. 1979. 22. Kurnick, J. T., Ostberg, L., Stegagno, M.. Kimura. A. K., Om, A., and Sjuberg, 0.. Stand. J. tmmunol.
10, 563, 1979.
23. Fu, S. M.. Chiorazzi, N.. Wang, C. A. B., .I. Exp. Med. 148, 1423, 1978. 24. Reinherz. E. L., and Schlossman, S. 25. Yu, D. T. Y.. J. tmmunol. 116, 1719, 26. Palacios, R.. and Moller. G., J. Exp.
Y., Montageri, G., Kunkel, H., Ho, H. A., and Gottlieb. F., Cells 19, 821, 1980. 1976. Med. 153,
1360. 1980.
Received December 3, 1982: accepted with revision December 9, 1982.
IMMUNOLOGY
T-Cell Medium JAMES
AND
IMMUNOPATHOLOGY
27, 444-451 (1983)
Subsets in Human Lymphocytes after PHA or Mixed Lymphocyte T. KLJRNICK,?
ANTHONY R. WARRENS, CAROL P. LEARY
Maintained in IL-2 Reaction Activation’
RICHARD
A. MOSCICKI,
ANI>
The culture of human T lymphocytes in interleukin-2 (IL-2) containmg growth factor medium results in a significant shift in the T-lymphocytes subsets isolated from such cultures at weekly intervals. If normal peripheral blood mononuclear cells are stimulated with phytohemagglutinin (PHA) or in a mixed lymphocyte reaction (MLRI. the resulting T lymphoblasts can be propagated in growth factor medium. Staining of the cultured cells with monoclonal antibodies was evaluated by indirect immunofluorescence on a laser-activated flow cytometer (Ortho Spectrum 111). The antibodies used were: OKT3 (mature T lymphocytes). OKT4 (helper/inducer 7‘ lymphocytes). OKT8 (cy,totoxic/suppressor T lymphocytes. OKTIO (immature and “activated” lymphocytes). OKTI la (cells which rosette with sheep erythrocytesl. and OKIa-I (HLA-DR constant region). Both PHA and MLR activation resulted in initial preservation of the OKT4 + subset predominance over OKT8 + T lymphocytes noted on normal circulating blood lymphocytes. However. during culture in T-cell growth factor medium. there was a progressive increase in the percentage of OKT8t cells. and a concomitant decrease in OKT4 + lymphoblasts. The increase in OKT8 + cells in the MLR-stimulated culture\ was paralleled by an increase in specific cell-mediated cytotoxicity against the stimulating lymphocyte population. In addition to the shift in T-lymphocyte subset. there wa\ virtual 100% staining with OKT3 and OKTl la, indicating the T-cell nature of the proliferating cells. OKTIO which was present on a small subset of fresh blood lymphocyte\ appeared rapidly in stimulated cultures, and was retained on virtually all lymphoblasts of either OKT4 + or OKT8 + subset. OKIa-I cells increased slowly in PHA-stimulated cultures. HLA-DR+ T cells were detected earlier in MLR cultures. The activation of T lymphocytes results in a significant increase in the number of molecules of OKTI la bound per cell. in concert with the increased avidity of T lymphoblasts for sheep erythrocytes. The significant change in the phenotype and function of lymphoblasts isolated from long-term cultures demonstrates the importance of monitoring cultures. and the potential hazards in equating a cultured cell population with a freshly isolated one.
INTRODUCTION Evaluation of human lymphocytes in the past decade has relied heavily on mitogen activation of peripheral blood lymphocytes. Many functional characteristics have been demonstrable from the lectin-stimulated peripheral blood cells (I-5). More recently, it has been shown that these activated lymphocytes can be maintained and propagated in culture by the addition of soluble factors present I This work was supported in part by Grant AI-1741 1 and CA29601 from the NIH of Health and in part was aided by Basil O’Connor Stacter Research Grant 5-324 from March of Dimes Birth Defects Foundation. Dr. Moscicki was supported by PHS Training Grant T32-A107040. z To whom correspondence should be addressed: Cox 5. Massachusetts General Hospital, I‘ruit Street, Boston, Mass. 02114. 444 0090-1229/83 $1.50 Copyright 0 1983 by Academic Press. Inc. All rights of reproduction in any form rewved
HUMAN
T-CELL
SUBSETS
GROWN
IN IL-2
445
in the supernatants of PHA-stimulated cultures (P-SUP) (6). This activity has been designated as interleukin-2 (IL-2) (7). In the early functional reports on IL2-maintained T-cell blasts, there was an indication that mixed lymphocyte reaction T-cell cultures from mice showed increasing cytotoxic activity at sequentially later time points in culture (8). While T cells which have been cultured and even cloned from IL-2-containing medium have been capable of demonstrating multiple functional capacities (9-12), these early mouse studies suggested that in a bulk culture there might be selection for a particular T-cell subset in IL-2 medium. Monoclonal antibodies which distinguish human T-cell subsets shed new light on T-cell interactions in vivo and in vitro (13-18). By immunofluorescence on a laser-activated flow cytometer (19), it has been possible to assess large numbers of cells stained with these monoclonal reagents and establish changes in the Tcell subsets over time in culture. We have assessed phenotypic changes in human T cells stimulated with PHA, or in a mixed lymphocyte reaction (MLR) after continued culture in IL-2 containing medium (P-SUP). MLR-stimulated blasts were assessed for specific cytotoxic activity after culture in P-SUP. MATERIALS
AND METHODS
PHA stimulation. Normal human peripheral blood mononuclear cells were isolated from Ficoll-Hypaque gradients (20). Lymphocytes at lOVm1 were stimulated in t&o with optimal doses of PHA ( 1 ug/ml of leucoagglutinin, Pharmacia Fine Chemicals, Uppsala, Sweden) for 3 days in RPM1 1640 containing 5% normal human serum, antibiotics, and 5 x 10.’ M 2-mercaptoethanol in 5% CO,humidified 37°C atmosphere. Cells were then resuspended at 10’ blasts/ml in Tcell growth factor medium (interleukin-2-containing medium). Cells were diluted back to lOVm1 every third day. Growth factor medium. The growth factor medium (P-SUP) was prepared from human tonsil lymphocytes stimulated with optimal doses of PHA. Growth factor was produced with and without residual free PHA. Medium was prepared as previously described, by collecting supernatants from stimulated tonsils after 48 hr (2 I). A lectin-free supernatant was also prepared by washing tonsil leukocytes after only 2 hr in the presence of PHA, and reculturing them in a lectin-free RPMI-2% human serum medium, and collecting the final medium after 48 hr. Such washed growth factor medium is free of PHA as tested in a mitogenesis assay. Both the PHA-containing and lectin-free medium were used at 25% dilutions in RPM1 1640 containing 5% pooled human serum. Mixed lymphocyte reaction cultures. Mixed lymphocyte reactions T-cell cultures were established by coculturing Ficoll-Hypaque-separated mononuclear cells at 10Vml with a similar number of X-irradiated (2000 R) allogeneic PBLs. After 6 days, the blast cells were collected on a Percoll density gradient as previously described (22) and cultured in lectin-free growth factor medium. At weekly intervals, a Vr-release assay (21) was performed on both autologous and appropriate stimulator allogeneic PHA blasts to determine specific cytotoxic activity of the MLR T cells. The targets were PHA blasts grown in T-cell growth factor medium for 7-14 days. Percentage specific lysis was calculated using the formula: % specific lysis = (1 - [(M - S) - (E - S)I(M - S)]} x 100, where M =
446
KURNICK
ET AL.
maximum release, S = spontaneous release, and E = experimental value of effector combination being tested. Assessment of T-cell subsets. At the initiation of culture and thereafter at weekly intervals, the cultured cells were stained with monoclonal antibodies as outlined below. Lymphocytes, IOh per sample, were stained for 30 min with 5 ~1 of monoclonal antibodies, followed by a phosphate-buffered saline (PBS) wash and a second incubation on ice with fluoresceinated goat anti-mouse immunoglobulin at a l/40 dilution (Cappel Labs, Cochranville. Pa.). The monoclonal antibodies (“Orthoclone,” Ortho Pharmaceuticals, Raritan, N.J.) are designated as OKT3. OKT4, OKT8, OKTlO, OKTlla, and OKla-1. After washing, the stained cells were fixed for IO min in buffered 1% paraformaldehyde and finally washed and stored in PBS until they were evaluated. Fixed cells could be observed up to 1 week following staining. Evaluation of fluorescence staining was carried out on a prototype cytofluorograph (Ortho Instruments, Westwood, Mass.). This laser-activated flow cytometer allows rapid analysis of cells based on size and surface morphology and fluorescence intensity (20). Data generated were stored on an Apple II computer interlinked with the flow cytometer, permitting standardization of the data from one day to the next. The threshhold for positive fluorescence was selected so that less than 2% of cells incubated with control mouse ascites and fluoresceinated goat anti-mouse reagent were scored as positive. RESULTS PHA Cultures. Figure 1 demonstrates the change in the percentage of staining with various monoclonal anti-T-cell reagents over a 4-week period in culture in P-SUP. At the initiation of culture, the T4 subset is the dominant T-cell population and the percentage of T&bearing cells comprises approximately half the percentage of T4 cells. T3 and Tlla, which are present on virtually all peripheral blood T lymphocytes, constitute the majority of the peripheral blood cells isolated from the Ficoll-Hypaque gradient. The antigen recognized by OKIa-I (which is a constant region of the HLA-DR molecule) (16, 23) is virtually lacking from the resting T lymphocyte, but appears here on a small percentage of cells as would be expected in a sample which contains some B cells and macrophages. TIO antigen is present on less than 15% of the freshly isolated lymphocytes. After the first week of activation, TlO appears on virtually all the cells and remains elevated thereafter. After 1 week the percentage of T4 cells begins to decline until at 4 weeks this population constitutes less than 10% of the T-cell blasts which are proliferating in culture. Concomitantly, the percentage of T8 lymphoblasts rise dramatically and constitutes the vast majority of the T-cell blasts at the end of this l-month culture period. HLA-DR-antigen-bearing lymphocytes rise and this antigen is present on virtually all of the lymphoblasts at the end of the culture. The selection for T cells is demonstrated by the virtual uniform staining with anti-T3 and Tl la on all of the cells. The selection toward T8-bearing cells occurs in IL-2-containing medium either with or without residual PHA. MLR Cultures. Figure 2 shows the specific cytotoxic activity of MLR blasts tested after 7, 14, and 21 days in culture. This example is representative of what
HUMAN
0’
’ 0
T-CELL
SUBSETS
7
GROWN
14 DAYS
IN
447
IL-7
21
L 1 28
I 21
4 28
tN (-XJl Tt JRF
IOO-
20.
I O
b
I 0
7
14
DAYS IN CULTURE FIG. 1. (a) Change in OKT3, OKT4, and OKTS staining of PHA blasts grown in IL-2 medium. The percentage positive (with standard deviations shown by vertical bars) for human peripheral blood lymphocytes assayed in culture at weekly intervals. (Each point represents the average of 6-16 samples). FIG. I. (b) Change in OKTll, OKTIO, and OKla-I (D) staining of PHA blasts grown in IL-2 medium. The percentage positive (with standard deviations shown by vertical bars) for human peripheral blood lymphocytes in culture at weekly intervals. (Each point represents the average of 616 samples).
448
KURNICK
ET
AL
70-
60 ul 3j
50-
> J "
4ci-
c 2
30-
% 8
*O IO-
1
I
I
I
I
I
I
1.5
3
6
12
25
50
TARGET
: EFFECTOR
RATIO
FIG. 2. Change in specific cytotoxicity in MLR blasts maintained in IL-2 medium. Specific i’C~release at various effector:target ratios are shown for cells cultured for 7 (+ ). 14 (0). and 71 dayz f-t+).
was noted in 15 different MLR cultures in which all showed a progressive increase in Wr release during culture in P-SUP. The changes in T-cell surface markers as detected with the monoclonal antibodies (Fig. 3) was similar to that seen with the PHA-stimulated cells, except that at 7 days, the T4+ cells were more prevalent and T8+ cells were accordingly fewer in the MLR. However, as with the PHA blasts, the T4 predominance quickly was reversed so that by 3 weeks in P-SUP. the percentage of T8+ cells was greater than 70% and T4+ were in the low 20% range. Also, OKIa-I staining appeared on significant numbers of cells at Day 7 in the MLR cultures. Assessment of individual cell jluorescence. The flow cytometer distinguishes positive- from negative-staining cells and permits evaluation of the intensity of fluorescence. The fluorescence intensity is indicated by fluorescence channel. with brighter cells falling in higher numbered channels. By this criterion. the amount of Tlla antigen increases dramatically after T-cell activation. Freshly isolated peripheral blood T cells stain rather weakly with OKTlla (mean fluorescence channel = 40), while activated cells stained more brightly with OKTlla than with any of the other monoclonal antibodies used in this study (mean fluorescence channel = 129) at Day 7. The intensity of staining with OKT4 likewise increases after activation, but staining with OKT3 remains relatively unchanged.
HUMAN
0
T-CELL
SUBSETS
;
1’4
DAYS
449
GROWN IN IL-?
2;
IN CULTURE
FIG. 3. Change in phenotype of MLR blasts during growth factor medium culture. Surface fluorescence was determined on a flow cytometer using monoclonal antibodies OKT3 (3), OKT4 (4). OKT8 (8). and OKIa-I (D). MLR blasts were harvested after 6 days and placed in growth factor medium, and aliquots stained on Days 7, 14. and 21. Standard deviation shown in vertical bars (average of 6-16 samples).
DISCUSSION
The initial reports on the use of soluble growth factors to maintain growth in human T lymphocytes demonstrated the extreme selective nature of PHA-induced T-cell growth factor for T cells. Indeed, the first observations were made on bone marrow cells, where a minor mature T-cell population became the predominant cell type during culture (6). With the recent availability of antibodies which can distinguish human T-cell subsets (13-19), the possibility was presented of observing possible T-cell subset selection during in vitro culture in supernatants from PHA-stimulated lymphocytes. The overall T-cell selective nature of the PHA-induced growth factor medium is confirmed by the increase in the percentage of T3- and Tlla-bearing cells (although these cells also predominate early in culture), but our results show that while both T4- and T&bearing lymphocytes can be propagated in culture, there is a progressive selection for cells bearing the T8 marker, and lacking the T4 antigen. These findings emphasize the potential perils in evaluating a cultured cell population as a model for in vivo lymphocyte function. The anti-HLA-DR constant region-specific antibody (OKIa) confirms the progressive rise in the cell population bearing DR antigen which has been noted by others (16, 23). It is of interest to note here, however, that the DR-bearing cells must also bear T8, and that a previous communication using soluble antigenstimulated T cells demonstrated that T4-bearing lymphoblasts can likewise be DR positive (9). Thus, the presence of DR molecules on the surface of T lymphoblasts
450
KURNICK
ET
AL.
cannot be used as an indicator of T-cell subset. The TIO marker rose in percent after T-cell stimulation, and remained elevated throughout the culture period. While this marker was initially noted primarily on immature thymocytes (24). it is clear that it is also present on activated normal T cells, and its presence with other mature T-cell markers should not be equated with abnormality in maturation. Also, the increased intensity of staining with anti-T1 la antibody may be correlated with the more vigorous sheep red blood cell-rosetting capacity of activated T cells (25). While intensity of OKT4 staining also increases. OKT3 intensity per cell does not change significantly after activation, indicating that the increase in antigen per cell is not merely an indication of increased cell size. These data demonstrate that T-cell subset selection does occur during prrtlonged culture in IL-2-containing medium when both T4- and T8+ cells are present. In our previous study where only T4’ cells were activated with soluble antigen, these cells could still be maintained for many months in culture (9) so the selective capacity of the IL-2 medium does not exclude prolonged T4’ subset culture, but by some still unknown mechanism, selects for TX- cells when both subsets are present. We cannot state whether the eventual T8+ predominencc results from more rapid proliferation of these cells, or some selective pressure exerted by the T8+ cells against the OKT4+ cells (such as suppression or cytotoxicity). In MLR-induced T-cell proliferation. we have been able to confirm a functional change which accompanies the phenotypic shift toward T8 r cells. The increased cytotoxic activity is consistent with the T8’- phenotype (17). We have previously reported that PHA blasts which contain T4+ and T8’- cells can both help and suppress in a B-cell maturation assay with pokeweed mitogen stimulation (9). Help by PHA blasts can be demonstrated on purified B even when T8’ cells begin to predominate. The suppression which can also be observed is difficult to interpret since even pure T4+ cells can induce suppression of B-cell maturation when an inducible suppressor T cell is present (9). Since even the use of cloned T-cell blasts does not clarify the precise mechanism of suppression detected in multiple cell mix experiments, we have elected not to elaborate on these earlier findings when evaluating a heterogenous T-cell population. The absorbtion of IL2 by activated T cells may play a role in in \*itrc/ suppression (26). However, the cytotoxic T-cell function is straightforward and the increase during P-SUP culture parallels the shift toward T8+, T4m cells. A final note on the culture is the frequent slowing of proliferation during the long-term culture. The Cweek time period observed in this study was necessitated by the cessation of growth of the cultures at approximately this time point. Further culture required the addition of irradiated fresh peripheral blood cells and lectin (22) to reinstate soluble IL-2 responsiveness. We have elected not to report in detail on subsets of lymphoblasts restimulated with irradiated cells and PHA. but note here that there is no further alteration in T-cell subsets after such stimulation, and after several months in culture with combinations of irradiated cells and soluble factor, a predominant T8+ population is retained with a small residual T4+ population. Specificity of MLR cytotoxicity is similary maintained. While the cessation in growth at the 4-week time point could indicate the need
HUMAN
T-CELL
SUBSETS
GROWN
IN IL-2
451
for a T4+ population to retain proliferative capacity of the T8+ cells, a similar slowing in growth has also been noted in soluble antigen-stimulated T-cell cultures where only T4+ cells are present. IL-2 responsiveness thus relies on more complex culture conditions. ACKNOWLEDGMENTS The authors gratefully acknowledge the generous gift of monoclonal antibodies from Drs. Gideon Goldstein and Patrick Kung of Ortho Pharmaceuticals, and the provision and technical support for the Spectrum III prototype flow cytometer from Ortho Diagnostics. We also appreciate the advice and support of Dr. Robert B. Colvin in carrying out these experiments. The technical support of Mr. Paul Steinwachs is gratefully acknowledged. The expert secretarial assistance of Ms. Susan Adams and Ms. Michele Angelo was much appreciated.
REFERENCES I. @eaves, M. F., Janossy, G., and Doenhoff, M., J. Exp. Med. 140, 1, 1974. 2. Waldman, T. A., Durm, M., Broder. S., Blackman, M., Blaese, R. M., and Strober, W.. Lancet 2, 609, 1974. 3. Fauci, A. S., and Pratt, K. R., J. Exp. Med. 144 674. 1976. 4. Keightley, R. G., Cooper, M. D., and Lawton. A. R., J. Immunol. 117, 1538, 1976. 5. Kumick. J. T.. Bell, C., and Grey, H. M., Stand. J. tmmunol. 5, 771, 1978. 6. Morgan, D. A., Ruscetti, F. W., and Gallo, R., Science 193, 1007. 1976. 7. Letter to the Editor, J. Immunol. 123, 2928, 1979. 8. Gillis, S., and Smith, K. A., Nature (London) 268, 154, 1977. 9. Kurnick, J. T., Hayward, A. R., and Altevogt, P.. J. Immunol. 126, 1307. 1981. IO. Lotze, M. T., Strausser, J. L., and Rosenberg, S. A., J. tmmunol. 124, 2972, 1980. II. Bach, F. H.. Inouye, H.. Hank, J. A., and Alter, B. M., Nature (London) 281, 307, 1979. 12. Fleisher, T. A.. Greene, W. C., Uchiyama, T., Goldman, C. K., Nelson, D. L., Blaese. M., and Waldmann, T. A., J. Exp. Med. 154, 156, 1981. 13. Reinherz, E. L.. Kung, P. C., Goldstein, G.. and Schlossman, S. F., Proc. Nar. Acad. Sci. USA 76, 4061, 1979. 14. Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman. S. F., J. Immunol. 123, 2894, 1979. 15. Reinherz, E. L., Kung, P. C.. Goldstein, G.. and Schlossman, S. F., J. tmmunol. 123, 1312, 1979. 16. Reinherz, E. L., Kung, P. C., Pesando, J. M.. Ritz, J.. Goldstein, G.. and Schlossman, S. F., J. Exp. Med. 150, 1472, 1979. 17. Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman, S. F.. J. tmmunol. 124, 1301. 1980. 18. Kung, P. C.. Tulle, M. A., DeMaria. M. A., Butler, M. S., Lifte. J., and Goldstein, G., Transplant. hoc. 12, (Suppl.), 141, 1980. 19. Hoffman, R. A.. Kung, P. C., Hansen, W. P.. and Goldstein, G., Pror. Nat. Acad. .Sci. USA 77, 4914. 1980. 20. Boyum, A., Stand. J. tmmunol. 5 (Suppl.), 9, 1976. 21. Kurnick, J. T.. Gronvik, K.. Kimura, A. K.. Lindblom, V. B., Skoog, V. T., Sjoberg. 0.. and Wigzell, H.. J. tmnzunol. 122, 1255. 1979. 22. Kurnick, J. T., Ostberg, L., Stegagno, M.. Kimura. A. K., Om, A., and Sjuberg, 0.. Stand. J. tmmunol.
10, 563, 1979.
23. Fu, S. M.. Chiorazzi, N.. Wang, C. A. B., .I. Exp. Med. 148, 1423, 1978. 24. Reinherz. E. L., and Schlossman, S. 25. Yu, D. T. Y.. J. tmmunol. 116, 1719, 26. Palacios, R.. and Moller. G., J. Exp.
Y., Montageri, G., Kunkel, H., Ho, H. A., and Gottlieb. F., Cells 19, 821, 1980. 1976. Med. 153,
1360. 1980.
Received December 3, 1982: accepted with revision December 9, 1982.