615
CD7- T CELLS IN RHEUMATOID ARTHRITIS ANDREW I. LAZAROVITS, MARTIN J. WHITE, and JACOB KARSH Objective. Rheumatoid arthritis (RA) is characterized by decreased expression of CD7 in the peripheral blood and in the synovium. The present study was designed to identify the basis for and functional consequences of this decreased expression. Methods. Peripheral blood lymphocytes from normal controls and from patients with RA or systemic lupus erythematosus (SLE), and T cell lines derived from rheumatoid synovium, were evaluated using 3-color fluorescence-activated cell sorter analysis. Results. Normal subjects and most SLE patients expressed homogeneous, bright CD7 on CD4+, CD45RA+ cells, whereas RA patients demonstrated a significantly increased proportion of CD7- cells. T cell lines derived from rheumatoid synovium demonstrated a striking deficiency of CD7 on CD4+, CD45RA- cells. CD4+, CD45RA+ cells from RA patients changed phenotype after in vitro activation to CD45RA negativCD7-, CD4+, ity, with up-regulation of CD7. Presented in part at the 75th Annual Meeting of the Federation of American Societies for Experimental Biology, Atlanta, GA, April 1991. From the John P. Robarts Research Institute, University Hospital, and the University of Western Ontario, London, and the Ottawa General Hospital and the University of Ottawa, Ontario, Canada. Supported by grants from the Medical Research Council of Canada, The Kidney Foundation of Canada, and the Canadian Arthritis Society. Andrew I. Lazarovits, MD: John P. Robarts Research Institute, University Hospital, and University of Western Ontario; Martin J. White, BSc: John P. Robarts Research Institute, University Hospital, and University of Western Ontario; Jacob Karsh, MD: Ottawa General Hospital and University of Ottawa. Address reprint requests to Andrew I. Lazarovits, MD, University Hospital, Room 4TU46, 339 Windermere Road, Box 5339, London, Ontario, N6A 5A5 Canada. Submitted for publication December 26, 1990; accepted in revised form January 14, 1992. Arthritis and Rheumatism, Vol. 35, No. 6 (June 1992)
CD45RA- cells were assessed for their ability to induce pokeweed mitogenariven IgM and IgM-rheumatoid factor synthesis, and they were found to be potent helperhducer cells. An increased population of CD7-, CD4+ cells in peripheral blood was found to predict a low response to recall antigens. Conclusion. The low expression of CD7 in RA may explain some of the immune abnormalities which may contribute to the pathogenesis of this disease.
The CD7 cluster of monoclonal antibodies (MAb) identifies a 40-kd structure present on a major subset of human T cells ( 1 4 ) . CD7+ T cells perform helper, suppressor, and cytotoxic functions (1,3,4). CD7 provides an alternate (non-antigen-specific) pathway of T cell activation (5,6) and is a member of the immunoglobulin gene superfamily (7). We have reported previously that the expression of the CD7 antigen in patients with rheumatoid arthritis (RA) is markedly decreased both on peripheral blood lymphocytes (PBL) (as measured by quantitative immunofluorescence) and on synovial T cells (8). Morishima et a1 have noted that a significant proportion of CD4-t T cells lack CD7 (4). Since CD4+, CD29high,CD45RA- T cells have been reported to be increased in patients with RA (9-13), we evaluated the expression of CD7 on CD4+, CD29high, CD45RA- and CD4+, CD29'"", CD45RA+ T cells in normal subjects and in patients with RA and systemic lupus erythematosus (SLE). Our experiments indicated that normal, RA, and SLE peripheral blood CD4+, CD29high,CD45RA- cells are heterogeneous in their expression of CD7. However, whereas CD4+, CD29'"", CD45RA-t PBL from normal subjects and from most patients with SLE are homogeneously CD7
LAZAROVITS ET AL
616
bright, CD4+, CD29'OW, CD45RA+ PBL from R A patients are divided into CD7+ and CD7- subsets. Furthermore, T cell lines derived from rheumatoid synovium demonstrate a striking increase in t h e num-
ber of CD7-, CD4+, CD45RA- cells.
PATIENTS AND METHODS Antibodies. MAb clustering with CD4 (Leu-3), CD8 (Leu-2), CD16 (Leu-1 l), and CD57 (Leu-7) were purchased from Becton-Dickinson (Mountain View, CA). MAb clustering with CD29 (4B4) and CD45RA (2H4) were purchased from Coulter (Hialeah, FL). CD74 (OKIal) was purchased from Ortho (Don Mills, Ontario, Canada). CD37 was purchased from Immunotech (Westbrook, ME). The following MAb were prepared in our laboratory: CD3 (12F6) (14), CD5 (24G4) (this MAb immunoprecipitates Tp67 and prevents directly fluoresceinated Leu-1 from binding to T cells), CD7 (7G5) (1416), CD71 (Act 11, transfemn receptor) (14), and Act I (a nonclustered late lymphocyte activation antigen) (17). MEM93, which clusters with CD45RA (18). was a generous gift from Dr. V. Horejsi. Fluorescein isothiocyanate-conjugated goat anti-mouse Ig (GAM-FITC) was purchased from Tag0 (Burlingame, CA). Avidin-allophycocyanin was purchased from Becton-Dickinson. Study subjects. Venous blood samples were obtained from 6 healthy volunteers (4 women, 2 men) and from I I patients with RA (7 women, 4 men) and 5 patients with SLE (4 women, 1 man), who met American College of Rheumatology (formerly, the American Rheumatism Association) criteria for these diseases (19,20). The average age of the healthy subjects, RA patients, and SLE patients was 39, 54, and 32 years, respectively (range 33-45, 40-79, and 2 6 3 6 , respectively). All of the RA patients had uncomplicated disease. All were taking nonsteroidal antiinflammatory drugs (NSAIDs); 2 were taking hydroxychloroquine. The SLE patients were unselected, consecutive outpatients, all of whom had chronic, stable disease without recent flares in disease activity. All were taking prednisone (20,000 lymphocytes per test were stored in "list mode" on a Digital Microvax computer, and data were analyzed using Consort 40 and "paint a gate" software. The percentages of positively stained cells were determined based on markers formed around irrelevant control antibodies. Quantitative fluorescence measurements for the CD7 antigen were performed using the mean channel fluorescence (MCF) (8,lS). In vitro changes in lymphocyte phenotype. To assess the ability of rheumatoid CD4+, CD45RA+ lymphocytes to change to a CD29high, CD45RA- phenotype, FicollHypaque-separated PBL from 2 patients were incubated with 4B4 and Leu-2 and were then sterile-sorted using the FACS, to yield a population of CD8-, CD29"'" lymphocytes, using methods previously described (21). Negative selection was used to obtain the CD4+, CD45RA+ cells, so as to avoid any possible functional effects these MAb might have on phytohemagglutinin (PHAkinduced activation and CD7 expression. The sorted cells were placed in complete medium at a concentration of 106/ml. Monocytes (obtained from the FACS) were added to a concentration of 5%, and the cells were cultured in the presence or absence of I &ml PHA at 37°C in a 5% CO, humidified atmosphere. Cells were removed at various time points over 1 week and were analyzed for expression of CD7, decrease in CD45RA, and increase in CD29. Modulation of CD7 by RA plasma. To determine the possible presence of a factor in RA plasma that could induce modulation of CD7, normal mononuclear cells were cultured in complete medium supplemented to 10% with RA plasma or were cultured with 7G5 (1 pg/ml), which is known to rapidly induce modulation of CD7 (8,1416). The cells were removed at various time points over 24 hours and assessed by flow cytometry, for CD7 antigen expression. CD7- T cells and induction of pokeweed mitogendriven IgM and IgM-rheumatoid factor synthesis. Mononuclear cells were obtained from the heparinized peripheral blood of 4 normal donors, using Ficoll-Hypaque density centrifugation. Monocytes were depleted by adhering the mononuclear cells to plastic petri dishes (Falcon, Oxnard, CA). T cells were obtained by overnight incubation with sheep red blood cells (SRBC), followed by density gradient centrifugation and ammonium chloride lysis of the adhering SRBC. The resulting E rosette-positive cells were >90% pure T cells as judged by reactivity with a CD3 MAb. CD4+ T cells were obtained by negative selection, using MAb CD8 and immunomagnetic beads (Dynal, Great Neck, NY). The resulting cells were >85% pure as judged by CD4 reactivity, and were contaminated 90% pure as judged by staining with CAM-FITC. The CD4+, CD45RA- cells were further fractionated into CD7+ and CD7- subpopulations on the FACS, using MAb 7G5 according to the technique described above. The resulting population was >98% pure as judged by reanalysis of the sorted cells.
CD7- CELLS IN RA
617
Table 1. Total CD7 negativity and CD7 negativity among CD4+ peripheral blood lymphocytes (PBL) and CD8+ PBL*
B cells were obtained from the mononuclear cell preparation by incubating the cells with CDlPconjugated immunomagnetic beads (Dynal) followed by overnight incubation, which allowed the beads to come off the cells. The beads were then removed with the magnet. Purity was >98% as judged by reactivity with MAb CD37. The highly purified T cell subsets were added in various numbers to 20,000 B cells and 5,000 monocytes per well in replicate cultures in 96-well round-bottomed microtiter plates (Nunc, Naperville, IL), in the presence or absence of 1 pglml pokeweed mitogen (PWM; Sigma. St. Louis, MO) in medium consisting of RPMI 1640 supplemented with 10% fetal calf serum, L-glutamine, 2 mercaptoethanol, and penicillin-streptomycin, as described (23). The plates were incubated for 8 days at 37°C in a 5% CO, humidified atmosphere. The culture supernatants were assayed for total IgM-rheumatoid factor (IgM-RF) by enzymelinked immunosorbent assay, using previously described methods (23). Studies of T cell proliferative response to recall antigens. The functional consequences of an increased subpopulation of CD7- T cells were assessed by examining the relationship between T cell proliferation in response to recall antigens and percentage of CD7- T cells in the peripheral blood. Five normal controls and 11 RA patients were evaluated. Mononuclear cells were separated from heparinized peripheral blood using Ficoll-Hypaque. The cells were cultured in 200-pi volumes of complete medium in 96-well flat-bottomed plates (Nunc) at a concentration of lO"lm1, in replicates of 3-6. The cultures were incubated for 6 days in the presence or absence of PWM (2 pg/ml), tetanus toxoid (1 pglrnl; Connaught Laboratories, Willowdale, Ontario, Canada), streptokinase (Streptase, 50 units/ml; Hoechst Canada, Montreal, Quebec, Canada), Cundidu ulbicuns (1 :10,OOO dilution; Bencard, Mississauga, Ontario, Canada), and bacillus Calmette-Guerin (10 pg/ml; Connaught). Tritiated thymidine (1 pCi; ICN, Montreal, Quebec, Canada) was added to each well during the final 4 hours of the culture. The cells were harvested onto glass-fiber filters using an automated apparatus (Skatron, Leir, Norway) and were assessed by beta-scintillation counting. Statistical analysis. Flow cytometry findings in the healthy subjects, the RA patients, and the SLE patients were compared using the Wilcoxon rank sum test.
Subjects
CD7-
Normal controls RA patients
19.1 t 2.0 25.7 f 6.0t 20.4 2 3.9
SLE patients
CD7-, CD4+ 6.1 8.6 2.5
2
0.9
?
3.2
2
1.6
CD7-, CD8+ 2.1 2 1 . 1 5.2 t 4.7 2.2 2 1.4
* PBL were studied by. 3-color fluorescence-activated cell sorter analysis. There was a substantial percentage of CD7-, CD4-. CD8- cells, not shown in the table. Values are the mean 2 SD percentage. RA = rheumatoid arthritis; SLE = systemic lupus erythematosus. t P < 0.05 versus normal controls.
RESULTS Flow cytometry. CD7 expression on normal, RA, and SLE PBL. RA PBL demonstrated decreased expression of CD7 compared with normal controls (Table I ) , confirming our previous observations in 26 patients with RA and 15 controls (8). The mean percentage of CD7- cells a m o n g the controls was 19.1% (MCF 348), compared with 25.7% (MCF 326) among the RA patients ( P < 0.05 for percentage CD7cells and for MCF). To further evaluate CD7 antigen expression, 3-color multiparameter flow c y t o m e t r y was performed. CD4+ and CD8+ cells were selected ("gated") for analysis. CD7 expression w a s heterogeneous on CD4+, CD45RA- cells both in normal subjects and in patients with RA (mean 14.9%, MCF 314 a n d mean 18.9%, MCF 309, respectively: P not significant) (Table 2). Normal CD4+, CD45RA+ cells were more homogeneously CD7 bright, whereas the RA patients had a sizable population of CD7-, CD4+, CD45RA+ cells, seen as a clearly negative peak (mean 8.3%, MCF 41 1 and mean 15.0%, MCF 389, respectively; P < 0.01 for percentage; P < 0.05 for MCF). CD7 expression was somewhat dim in CD8+, CD45RA+ PBL from RA patients compared with healthy con-
Table 2. CD7 negativity among CD4+, CD45RA- peripheral blood lymphocytes (PBL), CD4+, CD45RA+ PBL, CD8+, CD45RA- PBL, and CD8+, CD45RA+ PBL* ~
Subjects
CD7-, CD4+, CD45RA-
CD7-, CD4+, CD45RA+
CD7-, CD8+, CD45RA-
CD7-, CD8+, CD45RA+
Normal controls RA patients SLE patients
14.9 5 3.0 18.9 f 7.8 20.2 t 6.8
8.3 2 3.0
9.3 ? 4.2 15.2 2 7.4
4.4 ? 0.9 12.0 t 8.0 11.1 t 8.7
15.0 f 6.2t
8.5
2
14.0
31.9
2 13.0
* PBL were studied by 3-color fluorescence-activatedcell sorter analysis; CD7 expression was determined on cells "gated" for the given phenotype. There was a substantial percentage of CD7-, CD4-, CD8- cells, not shown in the table. Values are the mean 2 SD percentage. RA = rheumatoid arthritis; SLE = systemic lupus erythematosus. t P < 0.01 versus normal controls.
LAZAROVITS ET AL
618
Figure 1. CD7 expression on CD4+ peripheral blood lymphocytes (PBL) from a normal subject and a patient with rheumatoid arthritis (RA). PBL from a normal donor (A) or from a patient with RA (8) were subjected to 3-color fluorescence-activated cell sorter analysis. Three-dimensional bivariate histograms displaying dual-parameter correlated data, obtained after selecting for CD4+ cells, are shown. The histograms have been rotated: 0 marks the origin, and arrows indicate increasing fluorescence intensity. Note the presence of increased numbers of CD4+, CD7-, CD45RA+ and CD4+, CD4SRA- cells in the RA patient compared with the normal donor.
trols. The percentage of CD7- cells among these PBL did not differ significantly between RA patients (4.4%) and controls (12.0%), but the MCF was lower in RA patients (349 versus 404; P < 0.05). Figure 1 shows the populations of CD7-, CD4+ and CD7+, CD4+ cells and the population of CD4+, CD7-, CD45RA+ cells in a patient with RA. As can also be seen, there was an expanded population of CD4+, CD45RA- cells in the RA patient compared with a normal control. The mean ? SD percentage of these cells was 18 +- 6% among controls and 27.5 ? 6% among RA patients (P < 0.01), confirming the observations of others (10). Similar alterations in CD7 expression were seen when CD4+ cells were divided into CD29highand CD29"" subgroups in the 3-color FACS analyses (results not shown). The phenotype of the PBL of SLE patients was different from those seen in the normal subjects or the RA patients, and showed considerable heterogeneity among the patients studied. Unlike the finding in patients with RA, the mean percentage of CD7- cells among SLE patients did not differ from that in normal subjects (20.4% and 19.I%, respectively) (Table I). There was no increase in the CD7-, CD4-t subpopulation. There was heterogeneity of CD7 expression on both CD4+, CD45RA- and CD8+, CD45RAcells (20.2% and 31.9% respectively) (Table 2). The percentage of CD7-, CD4+, CD45RA+ cells varied widely among the 5 SLE patients (1.2%, 1.6%. 0.7%, 5.3%, and 33.9%). The patient with 33.9% CD7-, CD4+, CD45RA+ cells had a large negative peak on the FACS histogram (not shown), while these cells in the other 4 SLE patients were uniformly CD7 bright.
No clinical difference was evident between this patient and the other patients with SLE. Low CD7 expression on T cells derived from rheumatoid synovium. Since many studies have shown that the phenotype of the T cells in RA peripheral blood is quite different from the phenotype of the T cells in the synovium of RA patients (9-12,24), we evaluated the expression of CD7 on T cell lines derived from the synovium. As shown in Figure 2, CD7 expression was diminished on the synovium-derived T cell lines, compared with a T cell line derived from the peripheral blood of a normal donor, which was propagated in the same manner. It was notable that all of the cell lines had lost expression of CD45RA, and were uniformly CD29high(results not shown). Figure 2 also demonstrates that, whereas the synovium-derived T cell lines were somewhat low in CD7 antigen expression on CD8+ cells compared with the peripheral blood4erived T cell line, there was a substantial decrease in CD7 expression on the CD4+ RA cells compared with the normal cells. The cell lines derived from the synovium also had the following phenotypic markers: CD3 >98%, CD5 93%, CD16
CD7- T CELLS IN RHEUMATOID ARTHRITIS ANDREW I. LAZAROVITS, MARTIN J. WHITE, and JACOB KARSH Objective. Rheumatoid arthritis (RA) is characterized by decreased expression of CD7 in the peripheral blood and in the synovium. The present study was designed to identify the basis for and functional consequences of this decreased expression. Methods. Peripheral blood lymphocytes from normal controls and from patients with RA or systemic lupus erythematosus (SLE), and T cell lines derived from rheumatoid synovium, were evaluated using 3-color fluorescence-activated cell sorter analysis. Results. Normal subjects and most SLE patients expressed homogeneous, bright CD7 on CD4+, CD45RA+ cells, whereas RA patients demonstrated a significantly increased proportion of CD7- cells. T cell lines derived from rheumatoid synovium demonstrated a striking deficiency of CD7 on CD4+, CD45RA- cells. CD4+, CD45RA+ cells from RA patients changed phenotype after in vitro activation to CD45RA negativCD7-, CD4+, ity, with up-regulation of CD7. Presented in part at the 75th Annual Meeting of the Federation of American Societies for Experimental Biology, Atlanta, GA, April 1991. From the John P. Robarts Research Institute, University Hospital, and the University of Western Ontario, London, and the Ottawa General Hospital and the University of Ottawa, Ontario, Canada. Supported by grants from the Medical Research Council of Canada, The Kidney Foundation of Canada, and the Canadian Arthritis Society. Andrew I. Lazarovits, MD: John P. Robarts Research Institute, University Hospital, and University of Western Ontario; Martin J. White, BSc: John P. Robarts Research Institute, University Hospital, and University of Western Ontario; Jacob Karsh, MD: Ottawa General Hospital and University of Ottawa. Address reprint requests to Andrew I. Lazarovits, MD, University Hospital, Room 4TU46, 339 Windermere Road, Box 5339, London, Ontario, N6A 5A5 Canada. Submitted for publication December 26, 1990; accepted in revised form January 14, 1992. Arthritis and Rheumatism, Vol. 35, No. 6 (June 1992)
CD45RA- cells were assessed for their ability to induce pokeweed mitogenariven IgM and IgM-rheumatoid factor synthesis, and they were found to be potent helperhducer cells. An increased population of CD7-, CD4+ cells in peripheral blood was found to predict a low response to recall antigens. Conclusion. The low expression of CD7 in RA may explain some of the immune abnormalities which may contribute to the pathogenesis of this disease.
The CD7 cluster of monoclonal antibodies (MAb) identifies a 40-kd structure present on a major subset of human T cells ( 1 4 ) . CD7+ T cells perform helper, suppressor, and cytotoxic functions (1,3,4). CD7 provides an alternate (non-antigen-specific) pathway of T cell activation (5,6) and is a member of the immunoglobulin gene superfamily (7). We have reported previously that the expression of the CD7 antigen in patients with rheumatoid arthritis (RA) is markedly decreased both on peripheral blood lymphocytes (PBL) (as measured by quantitative immunofluorescence) and on synovial T cells (8). Morishima et a1 have noted that a significant proportion of CD4-t T cells lack CD7 (4). Since CD4+, CD29high,CD45RA- T cells have been reported to be increased in patients with RA (9-13), we evaluated the expression of CD7 on CD4+, CD29high, CD45RA- and CD4+, CD29'"", CD45RA+ T cells in normal subjects and in patients with RA and systemic lupus erythematosus (SLE). Our experiments indicated that normal, RA, and SLE peripheral blood CD4+, CD29high,CD45RA- cells are heterogeneous in their expression of CD7. However, whereas CD4+, CD29'"", CD45RA-t PBL from normal subjects and from most patients with SLE are homogeneously CD7
LAZAROVITS ET AL
616
bright, CD4+, CD29'OW, CD45RA+ PBL from R A patients are divided into CD7+ and CD7- subsets. Furthermore, T cell lines derived from rheumatoid synovium demonstrate a striking increase in t h e num-
ber of CD7-, CD4+, CD45RA- cells.
PATIENTS AND METHODS Antibodies. MAb clustering with CD4 (Leu-3), CD8 (Leu-2), CD16 (Leu-1 l), and CD57 (Leu-7) were purchased from Becton-Dickinson (Mountain View, CA). MAb clustering with CD29 (4B4) and CD45RA (2H4) were purchased from Coulter (Hialeah, FL). CD74 (OKIal) was purchased from Ortho (Don Mills, Ontario, Canada). CD37 was purchased from Immunotech (Westbrook, ME). The following MAb were prepared in our laboratory: CD3 (12F6) (14), CD5 (24G4) (this MAb immunoprecipitates Tp67 and prevents directly fluoresceinated Leu-1 from binding to T cells), CD7 (7G5) (1416), CD71 (Act 11, transfemn receptor) (14), and Act I (a nonclustered late lymphocyte activation antigen) (17). MEM93, which clusters with CD45RA (18). was a generous gift from Dr. V. Horejsi. Fluorescein isothiocyanate-conjugated goat anti-mouse Ig (GAM-FITC) was purchased from Tag0 (Burlingame, CA). Avidin-allophycocyanin was purchased from Becton-Dickinson. Study subjects. Venous blood samples were obtained from 6 healthy volunteers (4 women, 2 men) and from I I patients with RA (7 women, 4 men) and 5 patients with SLE (4 women, 1 man), who met American College of Rheumatology (formerly, the American Rheumatism Association) criteria for these diseases (19,20). The average age of the healthy subjects, RA patients, and SLE patients was 39, 54, and 32 years, respectively (range 33-45, 40-79, and 2 6 3 6 , respectively). All of the RA patients had uncomplicated disease. All were taking nonsteroidal antiinflammatory drugs (NSAIDs); 2 were taking hydroxychloroquine. The SLE patients were unselected, consecutive outpatients, all of whom had chronic, stable disease without recent flares in disease activity. All were taking prednisone (20,000 lymphocytes per test were stored in "list mode" on a Digital Microvax computer, and data were analyzed using Consort 40 and "paint a gate" software. The percentages of positively stained cells were determined based on markers formed around irrelevant control antibodies. Quantitative fluorescence measurements for the CD7 antigen were performed using the mean channel fluorescence (MCF) (8,lS). In vitro changes in lymphocyte phenotype. To assess the ability of rheumatoid CD4+, CD45RA+ lymphocytes to change to a CD29high, CD45RA- phenotype, FicollHypaque-separated PBL from 2 patients were incubated with 4B4 and Leu-2 and were then sterile-sorted using the FACS, to yield a population of CD8-, CD29"'" lymphocytes, using methods previously described (21). Negative selection was used to obtain the CD4+, CD45RA+ cells, so as to avoid any possible functional effects these MAb might have on phytohemagglutinin (PHAkinduced activation and CD7 expression. The sorted cells were placed in complete medium at a concentration of 106/ml. Monocytes (obtained from the FACS) were added to a concentration of 5%, and the cells were cultured in the presence or absence of I &ml PHA at 37°C in a 5% CO, humidified atmosphere. Cells were removed at various time points over 1 week and were analyzed for expression of CD7, decrease in CD45RA, and increase in CD29. Modulation of CD7 by RA plasma. To determine the possible presence of a factor in RA plasma that could induce modulation of CD7, normal mononuclear cells were cultured in complete medium supplemented to 10% with RA plasma or were cultured with 7G5 (1 pg/ml), which is known to rapidly induce modulation of CD7 (8,1416). The cells were removed at various time points over 24 hours and assessed by flow cytometry, for CD7 antigen expression. CD7- T cells and induction of pokeweed mitogendriven IgM and IgM-rheumatoid factor synthesis. Mononuclear cells were obtained from the heparinized peripheral blood of 4 normal donors, using Ficoll-Hypaque density centrifugation. Monocytes were depleted by adhering the mononuclear cells to plastic petri dishes (Falcon, Oxnard, CA). T cells were obtained by overnight incubation with sheep red blood cells (SRBC), followed by density gradient centrifugation and ammonium chloride lysis of the adhering SRBC. The resulting E rosette-positive cells were >90% pure T cells as judged by reactivity with a CD3 MAb. CD4+ T cells were obtained by negative selection, using MAb CD8 and immunomagnetic beads (Dynal, Great Neck, NY). The resulting cells were >85% pure as judged by CD4 reactivity, and were contaminated 90% pure as judged by staining with CAM-FITC. The CD4+, CD45RA- cells were further fractionated into CD7+ and CD7- subpopulations on the FACS, using MAb 7G5 according to the technique described above. The resulting population was >98% pure as judged by reanalysis of the sorted cells.
CD7- CELLS IN RA
617
Table 1. Total CD7 negativity and CD7 negativity among CD4+ peripheral blood lymphocytes (PBL) and CD8+ PBL*
B cells were obtained from the mononuclear cell preparation by incubating the cells with CDlPconjugated immunomagnetic beads (Dynal) followed by overnight incubation, which allowed the beads to come off the cells. The beads were then removed with the magnet. Purity was >98% as judged by reactivity with MAb CD37. The highly purified T cell subsets were added in various numbers to 20,000 B cells and 5,000 monocytes per well in replicate cultures in 96-well round-bottomed microtiter plates (Nunc, Naperville, IL), in the presence or absence of 1 pglml pokeweed mitogen (PWM; Sigma. St. Louis, MO) in medium consisting of RPMI 1640 supplemented with 10% fetal calf serum, L-glutamine, 2 mercaptoethanol, and penicillin-streptomycin, as described (23). The plates were incubated for 8 days at 37°C in a 5% CO, humidified atmosphere. The culture supernatants were assayed for total IgM-rheumatoid factor (IgM-RF) by enzymelinked immunosorbent assay, using previously described methods (23). Studies of T cell proliferative response to recall antigens. The functional consequences of an increased subpopulation of CD7- T cells were assessed by examining the relationship between T cell proliferation in response to recall antigens and percentage of CD7- T cells in the peripheral blood. Five normal controls and 11 RA patients were evaluated. Mononuclear cells were separated from heparinized peripheral blood using Ficoll-Hypaque. The cells were cultured in 200-pi volumes of complete medium in 96-well flat-bottomed plates (Nunc) at a concentration of lO"lm1, in replicates of 3-6. The cultures were incubated for 6 days in the presence or absence of PWM (2 pg/ml), tetanus toxoid (1 pglrnl; Connaught Laboratories, Willowdale, Ontario, Canada), streptokinase (Streptase, 50 units/ml; Hoechst Canada, Montreal, Quebec, Canada), Cundidu ulbicuns (1 :10,OOO dilution; Bencard, Mississauga, Ontario, Canada), and bacillus Calmette-Guerin (10 pg/ml; Connaught). Tritiated thymidine (1 pCi; ICN, Montreal, Quebec, Canada) was added to each well during the final 4 hours of the culture. The cells were harvested onto glass-fiber filters using an automated apparatus (Skatron, Leir, Norway) and were assessed by beta-scintillation counting. Statistical analysis. Flow cytometry findings in the healthy subjects, the RA patients, and the SLE patients were compared using the Wilcoxon rank sum test.
Subjects
CD7-
Normal controls RA patients
19.1 t 2.0 25.7 f 6.0t 20.4 2 3.9
SLE patients
CD7-, CD4+ 6.1 8.6 2.5
2
0.9
?
3.2
2
1.6
CD7-, CD8+ 2.1 2 1 . 1 5.2 t 4.7 2.2 2 1.4
* PBL were studied by. 3-color fluorescence-activated cell sorter analysis. There was a substantial percentage of CD7-, CD4-. CD8- cells, not shown in the table. Values are the mean 2 SD percentage. RA = rheumatoid arthritis; SLE = systemic lupus erythematosus. t P < 0.05 versus normal controls.
RESULTS Flow cytometry. CD7 expression on normal, RA, and SLE PBL. RA PBL demonstrated decreased expression of CD7 compared with normal controls (Table I ) , confirming our previous observations in 26 patients with RA and 15 controls (8). The mean percentage of CD7- cells a m o n g the controls was 19.1% (MCF 348), compared with 25.7% (MCF 326) among the RA patients ( P < 0.05 for percentage CD7cells and for MCF). To further evaluate CD7 antigen expression, 3-color multiparameter flow c y t o m e t r y was performed. CD4+ and CD8+ cells were selected ("gated") for analysis. CD7 expression w a s heterogeneous on CD4+, CD45RA- cells both in normal subjects and in patients with RA (mean 14.9%, MCF 314 a n d mean 18.9%, MCF 309, respectively: P not significant) (Table 2). Normal CD4+, CD45RA+ cells were more homogeneously CD7 bright, whereas the RA patients had a sizable population of CD7-, CD4+, CD45RA+ cells, seen as a clearly negative peak (mean 8.3%, MCF 41 1 and mean 15.0%, MCF 389, respectively; P < 0.01 for percentage; P < 0.05 for MCF). CD7 expression was somewhat dim in CD8+, CD45RA+ PBL from RA patients compared with healthy con-
Table 2. CD7 negativity among CD4+, CD45RA- peripheral blood lymphocytes (PBL), CD4+, CD45RA+ PBL, CD8+, CD45RA- PBL, and CD8+, CD45RA+ PBL* ~
Subjects
CD7-, CD4+, CD45RA-
CD7-, CD4+, CD45RA+
CD7-, CD8+, CD45RA-
CD7-, CD8+, CD45RA+
Normal controls RA patients SLE patients
14.9 5 3.0 18.9 f 7.8 20.2 t 6.8
8.3 2 3.0
9.3 ? 4.2 15.2 2 7.4
4.4 ? 0.9 12.0 t 8.0 11.1 t 8.7
15.0 f 6.2t
8.5
2
14.0
31.9
2 13.0
* PBL were studied by 3-color fluorescence-activatedcell sorter analysis; CD7 expression was determined on cells "gated" for the given phenotype. There was a substantial percentage of CD7-, CD4-, CD8- cells, not shown in the table. Values are the mean 2 SD percentage. RA = rheumatoid arthritis; SLE = systemic lupus erythematosus. t P < 0.01 versus normal controls.
LAZAROVITS ET AL
618
Figure 1. CD7 expression on CD4+ peripheral blood lymphocytes (PBL) from a normal subject and a patient with rheumatoid arthritis (RA). PBL from a normal donor (A) or from a patient with RA (8) were subjected to 3-color fluorescence-activated cell sorter analysis. Three-dimensional bivariate histograms displaying dual-parameter correlated data, obtained after selecting for CD4+ cells, are shown. The histograms have been rotated: 0 marks the origin, and arrows indicate increasing fluorescence intensity. Note the presence of increased numbers of CD4+, CD7-, CD45RA+ and CD4+, CD4SRA- cells in the RA patient compared with the normal donor.
trols. The percentage of CD7- cells among these PBL did not differ significantly between RA patients (4.4%) and controls (12.0%), but the MCF was lower in RA patients (349 versus 404; P < 0.05). Figure 1 shows the populations of CD7-, CD4+ and CD7+, CD4+ cells and the population of CD4+, CD7-, CD45RA+ cells in a patient with RA. As can also be seen, there was an expanded population of CD4+, CD45RA- cells in the RA patient compared with a normal control. The mean ? SD percentage of these cells was 18 +- 6% among controls and 27.5 ? 6% among RA patients (P < 0.01), confirming the observations of others (10). Similar alterations in CD7 expression were seen when CD4+ cells were divided into CD29highand CD29"" subgroups in the 3-color FACS analyses (results not shown). The phenotype of the PBL of SLE patients was different from those seen in the normal subjects or the RA patients, and showed considerable heterogeneity among the patients studied. Unlike the finding in patients with RA, the mean percentage of CD7- cells among SLE patients did not differ from that in normal subjects (20.4% and 19.I%, respectively) (Table I). There was no increase in the CD7-, CD4-t subpopulation. There was heterogeneity of CD7 expression on both CD4+, CD45RA- and CD8+, CD45RAcells (20.2% and 31.9% respectively) (Table 2). The percentage of CD7-, CD4+, CD45RA+ cells varied widely among the 5 SLE patients (1.2%, 1.6%. 0.7%, 5.3%, and 33.9%). The patient with 33.9% CD7-, CD4+, CD45RA+ cells had a large negative peak on the FACS histogram (not shown), while these cells in the other 4 SLE patients were uniformly CD7 bright.
No clinical difference was evident between this patient and the other patients with SLE. Low CD7 expression on T cells derived from rheumatoid synovium. Since many studies have shown that the phenotype of the T cells in RA peripheral blood is quite different from the phenotype of the T cells in the synovium of RA patients (9-12,24), we evaluated the expression of CD7 on T cell lines derived from the synovium. As shown in Figure 2, CD7 expression was diminished on the synovium-derived T cell lines, compared with a T cell line derived from the peripheral blood of a normal donor, which was propagated in the same manner. It was notable that all of the cell lines had lost expression of CD45RA, and were uniformly CD29high(results not shown). Figure 2 also demonstrates that, whereas the synovium-derived T cell lines were somewhat low in CD7 antigen expression on CD8+ cells compared with the peripheral blood4erived T cell line, there was a substantial decrease in CD7 expression on the CD4+ RA cells compared with the normal cells. The cell lines derived from the synovium also had the following phenotypic markers: CD3 >98%, CD5 93%, CD16
