- Email: [email protected]
T-cell subset analysis of cryopreserved human peripheral blood mononuclear cells
mmunoh~gv Letters, 7 (1983) 119 122 :lsevier ,tier 423
T-CELL SUBSET
ANALYSIS OF CRYOPRESERVED HUMAN BLOOD MONONUCLEAR CELLS
PERIPHERAL
Marian E. LUDGATE, Patricia R. DRYDEN, Anthony P. WEETMAN and Alan M. McGREGOR* Department ~f Medicine, Welsh National School of Medicine, Heath Park, CardiJf CF4 4XN, Wales (Received I July 1983) (Modified version received 8 August 1983) (Accepted 9 August 1983)
1. Summary A criticism of current techniques for monitoring changes in T-cell subset numbers over extended periods in individuals with disease states in which such changes might provide insight is the fact that serial samples taken are usually analysed fresh and therefore not in the same assay. To try to overcome this problem we have stored peripheral blood mononuclear cells frozen in liquid nitrogen and thence examined their ability to form sheep red blood cell (E) rosettes and to label with OK monoclonal antibodies. Results obtained show that cell viabilities following freezing and T-cell subset analysis of Erosette positive cells are no different when fresh or frozen and subsequently thawed peripheral blood mononuclear cells are used. 2. Introduction Day-to-day variation in test procedures for the investigation of immune function in patients may mask real changes or suggest alterations which are in fact a manifestation of the variation in the test procedure rather than the immune system. Since in a number of immunologically mediated diseases alterations in immune function over the course of the' disease can provide important information on the natural history and responsiveness of the disease to Key words: T-lymphocytes - freezing - flow cytometry * To whom correspondence should be addressed. 0165 2478/83/$3.00 © Elsevier Science Publishers B.V.
therapy, a reliable means for assessing change in such parameters over time is essential. With the development and availability of the OKT series of monoclonal antibodies raised against human lymphocyte subpopulations [1] a number of serial studies of alterations in T-cell subset level~ with time in a variety of autoimmune diseases have been reported [2-4]. Such studies have depended on serial analysis of freshly obtained lymphocytes, taken not only on different days but often without regard for the time of day and its possible influence on subset numbers [5], To try to overcome the problems inherent in day-today variation in assay procedures of this type it is necessary to establish a means for analysing a number of samples simply and reproducibly, collected over an extended period, from a single individual within the same assay. In an elegant series of experiments Birkeland [6] demonstrated the feasability of long term cryopreservation of human lymphocytes for sequential testing of immune competence using a cryobiological freezing apparatus. In his study he was able to show excellent reproducibility of sheep red blood cell (E) rosette positive T-cell numbers in lymphocyte samples taken at different times and subsequently tested as a batch. Using E-rosette positive T-cell preparations prepared from human peripheral blood we have examined the ability of such cells to label with OK monoclonal antibodies and compared the results of OKT+E + cells obtained fresh with those obtained after freezing (without the aid of cryobiological freezing apparatus) and storage in liquid nitrogen. l 19
3. Methods Peripheral blood mononuclear cells (PBM) isolated by Ficoll-Hypaque density centrifugation from 20 ml of heparinised blood were obtained on two separate occasions (1 wk apart but at the same time of day) from four healthy donors. The cells were washed three times in RPMI 1640 culture medium and thence resuspended, after counting and determination of viability with acridine orange and ethidium bromide [7] at l0 7 viable cells/ml. The cells obtained on the first occasion were then frozen to await parallel analysis with the fresh PBM taken 1 wk later. Freezing of cells was performed in initial studies by adding 500 #1 of fetal calf serum (FCS, 9 parts) plus dimethylsulfoxide (DMSO, 1 part) dropwise to a cell pellet containing 10 to 20× 106 PBM, with resuspension of the pellet. The PBM resuspended in 500 #1 of F C S - D M S O were transferred to a single freezing vial (round bottomed, 2 ml total volume, polypropylene, Nunc). Viabilities of cells recovered after this procedure were low (range 50 85%) and unpredictable so that for the present study 250 #1 of neat FCS was added to the pellet at 4 ° C and thereafter a 250 #1 aliquot of DMSO:FCS (1:4) added dropwise. The resuspended PBM kept on ice in freezing vials, were transferred rapidly to a well-insulated polystyrene container (of 10 X 10 X 20 cm external dimension and with 3 cm thick walls) and kept in this at 70° C overnight (at least 16 h) before transfer to liquid nitrogen storage tank. Prior to their use as a means of comparison with fresh PBM, the frozen PBM were thawed rapidly in a pre-warmed water bath (37° C) and thence transferred as soon as the last crystal of ice had melted to RPMI 1640 at 20°C in which they were left for 20 rain. To ensure that the thawing process was as rapid as possible only round bottomed freezing vials (described above) were used as air collecting in the space below free-standing vials delayed the process of cell thawing. The fresh and frozen PBM from the 4 individuals were then washed twice in RPMI 1640, assessed for cell viability and separated into sheep red blood cells (E) rosette-enriched (E+) and -depleted ( E ) populations [8]. Briefly this was performed by mixing 2 ml FCS, 5 ml of E (1 X 108/ml and 4 ml of PBM (3 to 5 X 106/ml) in RPMI; the mixture was centrifuged at 300 X g for 5 rain thence incubat120
ed on ice for an hour. After gentle resuspension the mixture was resuspended with Ficoll Hypaque and by density centrilugation (400 X g for 35 min) separation of an E~rosette preparation (pellet) from an E population (gradient interface) was achieved. Following hypotonic lysis to remove the sheep red cells from the E k population the cells were filtered through a 320 stainless steel mesh to remove cell debris and resuspended in phosphate buffered saline (PBS) supplemented with albumin (A, 0.2% w/v) and sodium azide (A, 0.2% w/v) at 20× 106 E+ cells/ml and maintained at room temperature. A sample of cells from both the fresh and frozen E+ populations of the 4 individuals was taken for estimation of B-cell contamination of the T-enriched population. B-cell numbers were assessed by direct immunofluorescence using a polyvalent commercially available fluorescein-labelled anti-human immunoglobulin preparation (Burroughs Wellcome) to detect surface immunoglobulin staining. By phagocyte-latex bead ingestion it was possible to differentiate these cells from true B cells [9]. The remainder of the E+ fresh and frozen populations were used for T-cell sub-set analysis using a fluorescence-activated cell sorter (FACS llI, Becton-Dickinson) to distinguish and enumerate the cell populations recognised by the commercially available mouse monoclonal antibodies, OKT3, OKT4, OKT8, OKIa and OKMI (Ortho Diagnostics) [10]. Briefly 50 ~1 aliquots of PBM were treated with 50 #1 of 1 in 10 dilution (in PBS AA) of each of the OK monoclonals and incubated for 5 min at 20° C. Cells were then washed twice in PBS AA treated with an in-house fluorescein-labelled sheep anti-mouse immunoglobulin (50/al of a I in 10 dilution in PBS-AA) for 5 min at 20° C, washed as bef6re, resuspended in I ml PBS AA and kept on ice until counting. Cells counted by flow cytometry are expressed as the percentage of the total cells counted, which fluoresce with the particular OK monoclonal. A minimum of 5000 fluorescing cells were counted. Peripheral blood mononuclear cells from a single subject were frozen as described above and 6 separate vials of cells were analysed separately following E-rosette enriching by estimation of T-cell subset numbers. Three vials were used on one occasion and the other 3 on three separate occasions to assess the
variability in the procedure. Data on the influence of the methodology described on OKT8 cell numbers within and between assays is shown. Analysis of the data was made using the paired Student's t-test.
Table 1 Assessment of cell viability in fresh and frozen peripheral blood mononuclear cells (PBM) from four normal individuals PBM
Fresh Frozen
4. Results
Subject 1
2
3
4
94.9 ± 2.2 a 88.3 ± 2.9
95.5 ± 2.1 90.6 ± 2.1
95.3 ± 1.7 85.0 ± 4.4
95.2 ± 2.6 90.6 ± 4.5
a Percentage viable cells (mean ± SD).
Whereas the mean levels of viable cells obtained in the PBM which had undergone prior freezing and thence thawing were lower than those of the fresh cells (Table 1), the differences between fresh and frozen cells did not achieve significance either for the individuals alone or for the pooled 4 fresh (95.2 + 2.2% mean + SD) compared with the 4 frozen (88.6 + 4.3) samples (P always > 0.1). Analysis of B-cell numbers in fresh and frozen PBM (Table 2) were not significantly different ( P > 0.1). Following T-cell enrichment, B-cell contamination of the E + population was trivial and no different in the fresh or frozen populations. T-cell subset analysis (Table 3) of the fresh and frozen cells shows that as expected following T-cell enrichment, T-cell numbers are increased in the frozen E + population as compared with the frozen PBM but that no differences are apparent in the fresh and frozen E + T-cell subset populations. Since the B-cell contamination is so small (approximately 1%) we have subtracted O K M I from OKIa and the resulting number of cells are referred to as "activated T cells". Comparison of the data on fresh and frozen samples for all the markers in all the patients showed no significant differences.
Analysis of OKT8 positive cells from a single individual on three separate samples on one occasion and thence on a further three samples on three separate occasions show levels of 29, 33 and 26% and 31, 28 and 28%, respectively.
Table 2 B-cell numbers in fresh and frozen peripheral blood mononuclear and T-cell enriched cell preparations Subject
Percentage of positive cells PBM
1 2 3 4
E+ cells
Fresh (A)
Frozen (B)
Fresh (C)
Frozen (D)
8.6 11.4 14.2 6.1
7.8 11.2 8.4 5.1
1.0 0.8 2.0 0.6
1.9 1.0 2.0 0.6
10.15:3.0 a
8.1±2.2
1.15:0.3
1.3±0.3
a Mean + SEM. There was no significant difference (P > 0.1) between A and B and between C and D.
Table 3 Influence of cell preparation on T-cell subset analysis Cell type present
OKT3+ OKT4+ OKT8+ Activated T cells (OKIa-OKMI)
Cell preparationa Fresh E-rosette positive cells (%)
Frozen E-rosette positive cells (%)
Frozen peripheral blood mononuclear cells (%)
84.5+3.0 51.0± 3.6 31.5 ± 2.8
81.1 + 1.9 47.8 + 2.1 32.3 5:3.3
62.8+ 1.4 36.3 -4- 2.6 23.5 + 1.6
1.5±0.3
3.05:1.6
6.0± I.l
a Percentage of cells positive
mean ± SEM (n = 4).
121
5. Discussion
Acknowledgements
Freezing of P B M in D M S O overnight at 70 ° C prior to storage in liquid n i t r o g e n did not significantly alter the viabilities as c o m p a r e d with fresh cells obtained subsequently f r o m the same individuals. Using these P B M it was possible to obtain T-cell enriched p o p u l a t i o n s of frozen P B M as effectively as for fresh P B M by E-rosette f o r m a t i o n and to enumerate the T-cell subsets within the E + populations using F A C S analysis of the positive fluorescence associated with binding of O K m o n o c l o n a l s to the specific T cells recognising the antibodies. T h e r e was no significant difference in the T-cell subset analysis obtained on fresh and frozen P B M f r o m the same individual. Using this technique it will be possible to store serial samples f r o m individuals t h r o u g h the course of their disease and thence, by subsequent thawing, analyse all the samples u n d e r the s a m e conditions within the s a m e assay. In so d o i n g a n u m b e r of criticisms, currently applicable to such analyses using fresh cells on different occasions, will be r e m o v e d and alterations observed in T-cell subset n u m b e r s can then be taken as being m o r e closely representative of the disease investigated rather than the m e t h o d o l o g y used.
This w o r k was s u p p o r t e d by grants f r o m the Wellc o m e Trust and A c t i o n Research for the Crippled Child. We are grateful to Miss Annette Berry for expert secretarial assistance.
122
References [1] Kung, P. C., Goldstein, G., Reinherz, E. L. and Schlossman, S. F. (1979) Science 206, 347 349. [2] Reinherz, E. L., Weiner, H. L., Hauser, S. L., Cohen, J. A., Distaso, J. A. and Schlossman, S. F. (1980) N. Engl. J. Med. 303, 124 129. [3] Veys, E. M., Hermanns, P., Goldstein, G., Kung, P., Schindler, J. and Van Wauwe, J. (1981) Int. J. Immunopharmacol. 3, 313 319. [4] Jackson, R. A., Morris, M. A., Haynes, B. F. and Eisenbarth, G. S. (1982) N. Engl. J. Med. 306, 785 788. [5] Ritchie, A. W. S., Oswald, I., Micklem, H. S., Boyd, J. E., Elton, R. E., Jazwinska, E. and James, K. (1983) Br. Med. J.286, 1773 1775. [6] Birkeland, S. A. (1980) J. lmmunol. Methods 35, 57 67. [7] Parks, D. R., Bryan, V. M., Oi, V. T. and Herzenberg, 1.. A. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 1962 1966. [8] Dean, J. H., Silva, J. S., McCoy, J. L., Leonard, C. M., Cannon, G. B. and Herberman, R. B. (1975),I. lmmunol. 115, 1449 1454. [9] Winchester, R. J. and Fu, S. M. (1976) Scand. J. lmmunol. 5 (Suppl. 5) 77 82. [10] Hoffman, R. A., Kung, P. C., Hansen, W. P. and Goldstein, G. (1980) Proc. Natl. Acad. Sci. U.S.A. 77, 4914 4917.
T-CELL SUBSET
ANALYSIS OF CRYOPRESERVED HUMAN BLOOD MONONUCLEAR CELLS
PERIPHERAL
Marian E. LUDGATE, Patricia R. DRYDEN, Anthony P. WEETMAN and Alan M. McGREGOR* Department ~f Medicine, Welsh National School of Medicine, Heath Park, CardiJf CF4 4XN, Wales (Received I July 1983) (Modified version received 8 August 1983) (Accepted 9 August 1983)
1. Summary A criticism of current techniques for monitoring changes in T-cell subset numbers over extended periods in individuals with disease states in which such changes might provide insight is the fact that serial samples taken are usually analysed fresh and therefore not in the same assay. To try to overcome this problem we have stored peripheral blood mononuclear cells frozen in liquid nitrogen and thence examined their ability to form sheep red blood cell (E) rosettes and to label with OK monoclonal antibodies. Results obtained show that cell viabilities following freezing and T-cell subset analysis of Erosette positive cells are no different when fresh or frozen and subsequently thawed peripheral blood mononuclear cells are used. 2. Introduction Day-to-day variation in test procedures for the investigation of immune function in patients may mask real changes or suggest alterations which are in fact a manifestation of the variation in the test procedure rather than the immune system. Since in a number of immunologically mediated diseases alterations in immune function over the course of the' disease can provide important information on the natural history and responsiveness of the disease to Key words: T-lymphocytes - freezing - flow cytometry * To whom correspondence should be addressed. 0165 2478/83/$3.00 © Elsevier Science Publishers B.V.
therapy, a reliable means for assessing change in such parameters over time is essential. With the development and availability of the OKT series of monoclonal antibodies raised against human lymphocyte subpopulations [1] a number of serial studies of alterations in T-cell subset level~ with time in a variety of autoimmune diseases have been reported [2-4]. Such studies have depended on serial analysis of freshly obtained lymphocytes, taken not only on different days but often without regard for the time of day and its possible influence on subset numbers [5], To try to overcome the problems inherent in day-today variation in assay procedures of this type it is necessary to establish a means for analysing a number of samples simply and reproducibly, collected over an extended period, from a single individual within the same assay. In an elegant series of experiments Birkeland [6] demonstrated the feasability of long term cryopreservation of human lymphocytes for sequential testing of immune competence using a cryobiological freezing apparatus. In his study he was able to show excellent reproducibility of sheep red blood cell (E) rosette positive T-cell numbers in lymphocyte samples taken at different times and subsequently tested as a batch. Using E-rosette positive T-cell preparations prepared from human peripheral blood we have examined the ability of such cells to label with OK monoclonal antibodies and compared the results of OKT+E + cells obtained fresh with those obtained after freezing (without the aid of cryobiological freezing apparatus) and storage in liquid nitrogen. l 19
3. Methods Peripheral blood mononuclear cells (PBM) isolated by Ficoll-Hypaque density centrifugation from 20 ml of heparinised blood were obtained on two separate occasions (1 wk apart but at the same time of day) from four healthy donors. The cells were washed three times in RPMI 1640 culture medium and thence resuspended, after counting and determination of viability with acridine orange and ethidium bromide [7] at l0 7 viable cells/ml. The cells obtained on the first occasion were then frozen to await parallel analysis with the fresh PBM taken 1 wk later. Freezing of cells was performed in initial studies by adding 500 #1 of fetal calf serum (FCS, 9 parts) plus dimethylsulfoxide (DMSO, 1 part) dropwise to a cell pellet containing 10 to 20× 106 PBM, with resuspension of the pellet. The PBM resuspended in 500 #1 of F C S - D M S O were transferred to a single freezing vial (round bottomed, 2 ml total volume, polypropylene, Nunc). Viabilities of cells recovered after this procedure were low (range 50 85%) and unpredictable so that for the present study 250 #1 of neat FCS was added to the pellet at 4 ° C and thereafter a 250 #1 aliquot of DMSO:FCS (1:4) added dropwise. The resuspended PBM kept on ice in freezing vials, were transferred rapidly to a well-insulated polystyrene container (of 10 X 10 X 20 cm external dimension and with 3 cm thick walls) and kept in this at 70° C overnight (at least 16 h) before transfer to liquid nitrogen storage tank. Prior to their use as a means of comparison with fresh PBM, the frozen PBM were thawed rapidly in a pre-warmed water bath (37° C) and thence transferred as soon as the last crystal of ice had melted to RPMI 1640 at 20°C in which they were left for 20 rain. To ensure that the thawing process was as rapid as possible only round bottomed freezing vials (described above) were used as air collecting in the space below free-standing vials delayed the process of cell thawing. The fresh and frozen PBM from the 4 individuals were then washed twice in RPMI 1640, assessed for cell viability and separated into sheep red blood cells (E) rosette-enriched (E+) and -depleted ( E ) populations [8]. Briefly this was performed by mixing 2 ml FCS, 5 ml of E (1 X 108/ml and 4 ml of PBM (3 to 5 X 106/ml) in RPMI; the mixture was centrifuged at 300 X g for 5 rain thence incubat120
ed on ice for an hour. After gentle resuspension the mixture was resuspended with Ficoll Hypaque and by density centrilugation (400 X g for 35 min) separation of an E~rosette preparation (pellet) from an E population (gradient interface) was achieved. Following hypotonic lysis to remove the sheep red cells from the E k population the cells were filtered through a 320 stainless steel mesh to remove cell debris and resuspended in phosphate buffered saline (PBS) supplemented with albumin (A, 0.2% w/v) and sodium azide (A, 0.2% w/v) at 20× 106 E+ cells/ml and maintained at room temperature. A sample of cells from both the fresh and frozen E+ populations of the 4 individuals was taken for estimation of B-cell contamination of the T-enriched population. B-cell numbers were assessed by direct immunofluorescence using a polyvalent commercially available fluorescein-labelled anti-human immunoglobulin preparation (Burroughs Wellcome) to detect surface immunoglobulin staining. By phagocyte-latex bead ingestion it was possible to differentiate these cells from true B cells [9]. The remainder of the E+ fresh and frozen populations were used for T-cell sub-set analysis using a fluorescence-activated cell sorter (FACS llI, Becton-Dickinson) to distinguish and enumerate the cell populations recognised by the commercially available mouse monoclonal antibodies, OKT3, OKT4, OKT8, OKIa and OKMI (Ortho Diagnostics) [10]. Briefly 50 ~1 aliquots of PBM were treated with 50 #1 of 1 in 10 dilution (in PBS AA) of each of the OK monoclonals and incubated for 5 min at 20° C. Cells were then washed twice in PBS AA treated with an in-house fluorescein-labelled sheep anti-mouse immunoglobulin (50/al of a I in 10 dilution in PBS-AA) for 5 min at 20° C, washed as bef6re, resuspended in I ml PBS AA and kept on ice until counting. Cells counted by flow cytometry are expressed as the percentage of the total cells counted, which fluoresce with the particular OK monoclonal. A minimum of 5000 fluorescing cells were counted. Peripheral blood mononuclear cells from a single subject were frozen as described above and 6 separate vials of cells were analysed separately following E-rosette enriching by estimation of T-cell subset numbers. Three vials were used on one occasion and the other 3 on three separate occasions to assess the
variability in the procedure. Data on the influence of the methodology described on OKT8 cell numbers within and between assays is shown. Analysis of the data was made using the paired Student's t-test.
Table 1 Assessment of cell viability in fresh and frozen peripheral blood mononuclear cells (PBM) from four normal individuals PBM
Fresh Frozen
4. Results
Subject 1
2
3
4
94.9 ± 2.2 a 88.3 ± 2.9
95.5 ± 2.1 90.6 ± 2.1
95.3 ± 1.7 85.0 ± 4.4
95.2 ± 2.6 90.6 ± 4.5
a Percentage viable cells (mean ± SD).
Whereas the mean levels of viable cells obtained in the PBM which had undergone prior freezing and thence thawing were lower than those of the fresh cells (Table 1), the differences between fresh and frozen cells did not achieve significance either for the individuals alone or for the pooled 4 fresh (95.2 + 2.2% mean + SD) compared with the 4 frozen (88.6 + 4.3) samples (P always > 0.1). Analysis of B-cell numbers in fresh and frozen PBM (Table 2) were not significantly different ( P > 0.1). Following T-cell enrichment, B-cell contamination of the E + population was trivial and no different in the fresh or frozen populations. T-cell subset analysis (Table 3) of the fresh and frozen cells shows that as expected following T-cell enrichment, T-cell numbers are increased in the frozen E + population as compared with the frozen PBM but that no differences are apparent in the fresh and frozen E + T-cell subset populations. Since the B-cell contamination is so small (approximately 1%) we have subtracted O K M I from OKIa and the resulting number of cells are referred to as "activated T cells". Comparison of the data on fresh and frozen samples for all the markers in all the patients showed no significant differences.
Analysis of OKT8 positive cells from a single individual on three separate samples on one occasion and thence on a further three samples on three separate occasions show levels of 29, 33 and 26% and 31, 28 and 28%, respectively.
Table 2 B-cell numbers in fresh and frozen peripheral blood mononuclear and T-cell enriched cell preparations Subject
Percentage of positive cells PBM
1 2 3 4
E+ cells
Fresh (A)
Frozen (B)
Fresh (C)
Frozen (D)
8.6 11.4 14.2 6.1
7.8 11.2 8.4 5.1
1.0 0.8 2.0 0.6
1.9 1.0 2.0 0.6
10.15:3.0 a
8.1±2.2
1.15:0.3
1.3±0.3
a Mean + SEM. There was no significant difference (P > 0.1) between A and B and between C and D.
Table 3 Influence of cell preparation on T-cell subset analysis Cell type present
OKT3+ OKT4+ OKT8+ Activated T cells (OKIa-OKMI)
Cell preparationa Fresh E-rosette positive cells (%)
Frozen E-rosette positive cells (%)
Frozen peripheral blood mononuclear cells (%)
84.5+3.0 51.0± 3.6 31.5 ± 2.8
81.1 + 1.9 47.8 + 2.1 32.3 5:3.3
62.8+ 1.4 36.3 -4- 2.6 23.5 + 1.6
1.5±0.3
3.05:1.6
6.0± I.l
a Percentage of cells positive
mean ± SEM (n = 4).
121
5. Discussion
Acknowledgements
Freezing of P B M in D M S O overnight at 70 ° C prior to storage in liquid n i t r o g e n did not significantly alter the viabilities as c o m p a r e d with fresh cells obtained subsequently f r o m the same individuals. Using these P B M it was possible to obtain T-cell enriched p o p u l a t i o n s of frozen P B M as effectively as for fresh P B M by E-rosette f o r m a t i o n and to enumerate the T-cell subsets within the E + populations using F A C S analysis of the positive fluorescence associated with binding of O K m o n o c l o n a l s to the specific T cells recognising the antibodies. T h e r e was no significant difference in the T-cell subset analysis obtained on fresh and frozen P B M f r o m the same individual. Using this technique it will be possible to store serial samples f r o m individuals t h r o u g h the course of their disease and thence, by subsequent thawing, analyse all the samples u n d e r the s a m e conditions within the s a m e assay. In so d o i n g a n u m b e r of criticisms, currently applicable to such analyses using fresh cells on different occasions, will be r e m o v e d and alterations observed in T-cell subset n u m b e r s can then be taken as being m o r e closely representative of the disease investigated rather than the m e t h o d o l o g y used.
This w o r k was s u p p o r t e d by grants f r o m the Wellc o m e Trust and A c t i o n Research for the Crippled Child. We are grateful to Miss Annette Berry for expert secretarial assistance.
122
References [1] Kung, P. C., Goldstein, G., Reinherz, E. L. and Schlossman, S. F. (1979) Science 206, 347 349. [2] Reinherz, E. L., Weiner, H. L., Hauser, S. L., Cohen, J. A., Distaso, J. A. and Schlossman, S. F. (1980) N. Engl. J. Med. 303, 124 129. [3] Veys, E. M., Hermanns, P., Goldstein, G., Kung, P., Schindler, J. and Van Wauwe, J. (1981) Int. J. Immunopharmacol. 3, 313 319. [4] Jackson, R. A., Morris, M. A., Haynes, B. F. and Eisenbarth, G. S. (1982) N. Engl. J. Med. 306, 785 788. [5] Ritchie, A. W. S., Oswald, I., Micklem, H. S., Boyd, J. E., Elton, R. E., Jazwinska, E. and James, K. (1983) Br. Med. J.286, 1773 1775. [6] Birkeland, S. A. (1980) J. lmmunol. Methods 35, 57 67. [7] Parks, D. R., Bryan, V. M., Oi, V. T. and Herzenberg, 1.. A. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 1962 1966. [8] Dean, J. H., Silva, J. S., McCoy, J. L., Leonard, C. M., Cannon, G. B. and Herberman, R. B. (1975),I. lmmunol. 115, 1449 1454. [9] Winchester, R. J. and Fu, S. M. (1976) Scand. J. lmmunol. 5 (Suppl. 5) 77 82. [10] Hoffman, R. A., Kung, P. C., Hansen, W. P. and Goldstein, G. (1980) Proc. Natl. Acad. Sci. U.S.A. 77, 4914 4917.