- Email: [email protected]
Defective signal transduction in CD4−CD8− T cells of lpr mice
CELLULAR
IMMUNOLOGY
123,396-404 (1989)
Defective Signal Transduction
in CD4XD8-
T Cells of Ipr Mice
FRANCES STAFFORDBRADY,* EUI SUGIYAMA,* DWIGHT R. ROBINSON,* MAN-SUNSY,~ JOSEPH V. BONVENTRE,~ AND EDWARD T. H, YEH* *Arthritis and$Renal Units of the Department ofMedicine, and iDepartment ofPathology, Massachusetts Generai Hospital, Harvard Medical School, Boston, Massachusetts 021I4 Received April 15, 1989; accepted J&y 17, 1989 Mice homozygous for the Ipr gene develop a lymphoproliferative disorder due to expansion of a subset of CXWCDK T cells. Triggering of the T-cell receptor in these /pr T cells does not lead to tramlocation of protein kinase C or phosphorylation of CR3, interleukin-2 production, or proliferation, whereas a combination of phorbol ester and calcium ionophore does. Stimulation with concanavalin A or anti-CD3 induces phosphoinositide hydrolysis. The rise in inositol bisphosphate, inositol triphosphate, and inositol tetralcisphosphate, identified by HPLC, is similar in +/+ and Ipr T cells. The concentration of cytoplasmic free calcium ([Caz’li), however, under basal and stimulated conditions is significantly lower in lpr T cells. The lower basal [Ca2+li may explain why induction of proliferation with phorbol ester and calcium ionophore requires a higher concentration of ionophore in these cells than in normal T cells. The lower [Ca2’li obtained on stimulation may contribute to the activation defect of CD4CD8- lpr T cells. Q 19%9Academic Press,Inc.
INTRODUCTION Mice homozygous for the Ipr gene develop generalizedlymphadenopathy due to accumulation of a subsetof abnormal T cells (1). The phenotype of lpr abnormal T cellsis similar to that of a small subsetof CD4-CDS- thymocytes (2-4). T-cell recep tor (TCR) is expressedon their surface(4, S),albeit at a lower density than on normal T cells(6). Stimulation of lpr T cellsby mitogens,however,doesnot result in proliferative responses( 1,4,7). Triggering of the TCR in normal T cells activatesphospholipaseC, which hydrolyzes phosphatidylinositol 4,5-bisphosphate to inositol 1,4,5-trisphosphate( 1,4,51P&and diacylglycerol(8). 1,4,5-IP3raisesthe concentration of cytoplasmic free calcium ([Ca2’li), by intracellular mobilization (8), and probably also by influx through a calcium channel (9). A combination of calcium ionophore, which raises [Ca’+]i, and phorbol my&ate acetate(PMA), which mimics diacylglycerol in activation of protein kinaseC, induces proliferation ofnormal T cells,highlighting the importance of thesesecondmessengersin T-cell activation (8,10). It hasbeen reported that mitogen-induced phosphoinositide hydrolysis is deficient in lpr T cells (6). Concomitant changesin [Ca”]i have not been studied. Protein kinase C cannot be activated by lectin ($6) but can be activated by PMA (4-6), asjudged by translocation of enzyme activity from cytosol to membrane (4,6), and phosphorylation of CD3 y and c (5). 394 ooo8-8744/89 $3.00 Copyright 0 1989 by Academic Refs, Inc. All rights ofreproduction in any fwm rescrvgd.
SIGNAL
TRANSDUCTION
IN Ipr T CELLS
397
These observations suggest that lpr T cells have a defect in the signaling pathway between triggering of the TCR and activation of protein kinase C. To further analyze the defect in signal transduction in lpr T cells, we examined phosphoinositide hydrolysis and correlated this with changes in [Ca2’]i. We found that mitogen-induced generation of inositol phosphates was similar in these cells and congenic +/+ T cells. The concomitant rise in [Ca2’]i, however, was consistently lower than that in normal T cells. This subnormal calcium response may partly account for the activation defect in lpr T cells. MATERIALS AND METHODS
Animals and Reagents MRL/MpJ (MRL), C3H/HeJ (C3H), and C57BL/6 (B6) mice homozygous for the
Ipr gene and their +/+ counterparts (Jackson Laboratory, Bar Harbor, ME) were used between the agesof 4 and 6 months. Low-Tox rabbit complement was obtained from Cedarlane Laboratories (Westbury, NY). All other reagents were from Sigma. Anti-CD3 (145-2Cll) and other monoclonal antibodies were used as cell culture supernatants.
T-Cell Purification Abnormal lpr T cells were purified from lymph nodes of mice homozygous for the lpr gene by treatment with monoclonal antibodies to CD4 (GK 1.5; ATCC, Rockville, MD), CD8 (53.6.72; ATCC), and T-cell-activating protein (TAP; 3E7.1 or lE7) for 30 min on ice, followed by complement-mediated lysis for 30 min at 37°C (4). The term lpr T cells throughout the paper refers to this population of purified abnormal lymph node T cells. Normal T cells were obtained by passageof spleen cells from +/+ mice over nylon wool columns. In vitro proliferation assayswith concanavalin A (Con A) and anti-CD3 were performed to check the purity of the +/+ and lpr T cells.
In Vitro Proliferation Assay Cells were resuspended in RPM1 1640 medium supplemented with 10%heat-inactivated fetal calf serum, 2 mM L-glutamine, 100 U/ml penicillin, 100 &ml streptomycin, 0.25 pg/ml fungizone, 0.1 m&f nonessential amino acids, 20 mM Hepes, and 5 X lOA M 2-mercaptoethanol; and 5 X lo5 cells in a final volume of 200 &well were cultured for 48 hr in 96-well llat-bottom tissue culture plates. [3H]Thymidine ( 1 &i/well) was added for the last 16 hr of culture. Cells were harvested with a PHD automatic cell harvester (Cambridge, MA) and were counted in a scintillation counter. Data are represented as the mean and SEM of triplicate cultures.
PhosphoinositideHydrolysis Cells were labeled overnight with 25 &i/ml myo-[2-3H]inositol (Amersham Corp., Amersham, UK) in inositol-free RPM1 1640 supplemented with 1% BSA, ~-@amine, and antibiotics. For each experiment, 10’ cells were washed and incubated in 0.6 ml inositol-free RPM1 1640 for 40 min in a 5% CO* incubator. LiC12 5 mM was added for the last 10 min of incubation in half of the experiments on both normal
398
STAFFORD-BRADY
ET AL.
and abnormal T cells. The increase in inositol polyphosphates on stimulation was similar in the experiments with and without LiC12. Following stimulation, ice-cold 10% TCA was added to stop the reaction and the reaction mixture was centrifuged. The supernatant was treated with 1 ml ice-cold Freon amine and centrifuged again. The aqueous phase was applied to a Partisil SAX 10 HPLC column (Whatman, Maidstone, England) and eluted with a gradient of ammonium phosphate, as previously described (11). The eluant was monitored using a Berthold LB506C in-line HPLC scintillation counter. Identification of different inositol phosphates was based on the retention time of standard 3H-labeled inositol 1-phosphate, inositol 1,Cbisphosphate, inositol 1,4,5-trisphosphate, and inositol 1,3,4,Qetrakisphosphate (Amersham) (Fig. la). The inositol phosphate peaks were wider in the experimental tracings, presumably due to the presence of other isomers, e.g., inositol 1,3-bisphosphate, inositol 3,4-bisphosphate and inositol 1,3,4-trisphosphate ( 11). All isomers were included in the calculations. An unpaired Student’s t test was used to compare generation of inositol phosphates in +/+ and Ipr T cells. Intracellular Free Calcium Mean [Ca2+]i was measured in populations of cells before and after stimulation. A Deltascan spectrophotometer (Photon Technology International, Princeton, NJ) was used with dual excitation at 340 and 380 nm and an emission wavelength of 5 10 nm. Cells at a density of 5 X 106/ml in RPM1 1640 supplemented with 1.5%gelatin were loaded with 2.5 &ml of 1 mM membrane-permeant fura- AM (Molecular Probes, Inc., Eugene, OR) in dimethyl sulfoxide (DMSO) at 37°C for 1 hr. For each measurement, 10’ cells were resuspended in 2 ml of buffer, consisting of 145 mM NaCI, 5 mM KCl, 0.5 mM MgS04, 1 mM Na2HP04, 5 mA4 glucose, 20 mM Hepes, 0.5% BSA, 0.5% gelatin, and 2 nut4 CaC12in water (pH 7.4, 37°C). Autofluorescence of unloaded cells incubated with DMSO and treated in a similar fashion was subtracted. Calibration of [Ca2+li was performed at the end of each experiment using 10 p&Y ionomycin, followed by 0.1 nut4 digitonin (R,ax) and 15 mM EGTA + 24 r&l4 Tris (R,i”), The [Ca2+]i was calculated according to Grynkiewicz et al. (12). Statistical significance of the difference between [Ca2+]i in +/+ and lpr T cells was determined by an unpaired Student’s t test. RESULTS Purified CD4CD8- T cells from the lymph nodes of MRL-lpr/lpr mice did not proliferate in response to Con A or anti-CD3 plus PMA, whereas unselected T cells from lymph nodes of the same mice did (Table 1). The total population contained a minority of normal T cells, which accounted for the proliferative responsesseen. We previously reported that proliferative responses,seenin normal T cells, were not seen in CD4-CD8- lpr T cells on treatment with PMA and 0.06 ~~ calcium ionophore A23 187 (4). Calcium ionophore (A23 187 or ionomycin), at a concentration 4 to 20 times greater than that required to activate normal T cells was, however, able to synergize with PMA to induce proliferation of Ipr T cells (Table l), as has been previously reported (2, 13). Similar results were obtained with T cells from B6 and C3H mice (data not shown). The proliferation data suggestedthat lpr T cells have a defect in the signaling pathway between triggering of the TCR and activation of protein kinase C. It was, there-
SIGNAL TRANSDUCTION
399
IN Ipr T CELLS
TABLE 1 Proliferative Responsesof Lymph Node T Cells of MRL-lpr/lpr
Mice
[ ‘H]Thymidine incorporation (cpm X 10e3) Stimulation Experiment 1 Medium Con A 2.5 &ml Anti-CD3 (1:lO) PMA 2 rig/ml Anti-CD3 (1: 10) + PMA 2 rig/ml Experiment 2 Medium PMA 5 rig/ml Ionomycin 250 rig/ml Ionomycin 1 pg/ml Ionomycin 250 rig/ml + PMA 5 rig/ml Ionomycin 1 &ml + PMA 5 rig/ml
Total lymph node T cells
CD4- CD8- lpr T cells
0.5+ 44.9 * 3.4 f 0.9 f 96.3 -I
0.1 1.5 0.5 0.4 18
0.3fO.l 0.4kO.l 0.5 Ii 0.3 0.2 kO.1 0.6~0.1
0.5 f 2.4+ 1.3* 0.8% 93.2 f 103.4+
0.1 0.2 0.1 0.1 4.7 2.5
2.0 k 0.2 1.3 + 0.3 1.6 kO.5 0.8 +O.l 7.8 f 1.0 84.0 -e 2.8
Note. 5 X lo5 untreated lymph node cells or purified CD4-CDS-TAP- T cells from MRL-Ipr/lpr mice were cultured in the presence of the indicated reagents. Cultures were incubated for 48 hr, pulsed with 1 &/well [ 3Hlthymidine for the last 16 hr, harvested, and counted.
fore, of interest to study production of second messengersin response to Con A or anti-CD3. Stimulation of abnormal B64pr/lpr T cells with 5 pg/ml Con A for 20 min resulted in a rise in inositol bisphosphate (IP,), inositol trisphosphate (IP& and inositol tetrakisphosphate (1P.J (Fig. 1). Similar results were obtained using MRLlpr/lpr and C3H-Zpr/lpr T cells (Figs. 2 and 3). Lack of a proliferative response to Con A or anti-CD3 confirmed that the ipr T cells were adequately purified (data not shown). Generation of inositol phosphates in response to varying dosesof Con A was similar in +/+ (Fig. 2a) and lpr (Fig. 2b) T cells. The kinetics of phosphoinositide hydrolysis in response to Con A and anti-CD3 were examined (Fig. 3). On stimulation with 10 pg/ml Con A for 20 min, the mean percentage increase in IP2, IP3, and IP, in +/+ T cells was 186, 159, and 112%, respectively (Fig. 3a), and in lpr T cells was 224, 111, and 3 12%(Fig. 3b). These increasesrepresented maximal inositol phosphate production and were comparable in the two cell types. Generation of inositol phosphates in response to anti-CD3 was maximal at 10 min and was similar in +/+ (Fig. 3c) and lpr (Fig. 3d) T cells. In view of the normal production of IP3 on activation of lpr T cells, it was of interest to determine concomitant changesin [Ca2’]i. A rise in [Ca2+]iwas seenin the normal and abnormal cells following stimulation (Fig. 4). This calcium response provided further evidence for phosphoinositide hydrolysis in Zpr T cells. Although there was considerable overlap, basal [Ca2’]i tended to be lower in the abnormal T cells, with a mean value of 52 nM compared with 65 nM in the control T cells (P < 0.0 1). At all time points following stimulation with Con A at 10 pg/ml (Fig. 4a) and 2.5 pg/ml (Fig. 4b), the [Ca2+]i was significantly lower in lpr than in +/+ T cells (P < 0.05). Ionomycin at 10 PMinduced a similar increase in [Ca*+]i in the normal and abnormal T cells. There was no rise in [Ca2+]i in response to 10 pug/mlCon A in both +/+ and
400
STAFFORD-BRADY
127
ET AL.
I .I-IPZ
a
cps I.3.4.S.IP4
0
_Ik C
Time (min) FIG. 1. Generation of inositol phosphates by B6-lpr/lpr abnormal T cells. The elution profile of ‘Hlabeled standard inositol phosphates is shown in (a). Myo-[2-3H]inositol-labeled cells were unstimulated (b) or stimulated with 5 &ml Con A for 20 min (c). Water-soluble inositol phosphates were extracted and separated by HPLC. Two isomers of IPr, presumably 1,3-IPr and 1,4-IP2, were separated in (c). Labeled substanceseluted prior to IP, included inositol and glycerophosphorylinositol.
lpr T cells when incubated for 5-10 min in calcium-free buffer to which 2 mMEGTA was added 2-4 min prior to stimulation (data not shown). The spectral characteristics of intracellular fura- were similar in +/+ and Ipr T cells. The difference in [Ca”]i, therefore, did not appear to be due to different handling of the fluorescent dye by the two cell types.
SIGNAL TRANSDUCTION
IN lpr T CELLS
401
ConA (&ml) FIG. 2. Generation of inositol phosphates in +/+ (a) and Ipr (b) T cells in response to stimulation with varying dosesof Con A for 20 min. Data represent mean percentage increases in inositol phosphates over unstimulated controls, for three experiments on normal T cells from MRL-+/+ and B6-+/+ mice (a) and five experiments on abnormal T cells from MRL-lpr/lpr, B6-Zpr/lpr, and C3H-Ipr/lpr mice (b). The mean counts per second in unstimulated cells for IP2, IPX, and IP, in +/+ T cells were 88 + 23,39 k 6, and 39 + 2, respectively; and in lpr T cells these were I 18 f 18,6 1 -t 6, and 34 +-3, respectively. For all comparisons 0.1 < P < 0.4, except for 1P4production in responseto 10 &ml Con A, for which 0.05 < P < 0.1.
DISCUSSION Scholz et al. reported an increasein total inositol phosphatesof 57-66% in +/+ T cells and only 13-l 8% in lpr T cells in responseto PHA (6, lo), and 44% in +/+ and 17% in Zpr T cells in responseto Con A (10). We have demonstrated substantial increasesin IP,, IP,, and IP, in both +/+ and lpr T cells. In fact, the rise in inositol phosphatestended to be higher in lpr T cells.The discrepanciesmay be accountedfor by differencesin labeling time and inositol phosphateseparationtechnique. Scholzet al. labeled the cells for 2 hr, in contrast to our 16 hr. Furthermore, they separatedthe inositol phosphateson a Dowex column and reported only changesin total inositol phosphates.Changesin IP2, IP3, and IP, may have been masked when these were not measured separately from more abundant labeled molecules, such as IP, and glycerophosphorylinositol (Fig. 1). We used a sensitiveHPLC technique, which allowed us to clearly separateand quantitate increasesin IPz, IP3, and IP, (Fig. 1). Although the separation was not adequateto measureIP3 isomersindividually in all experiments,the concentration of 1,4,5-IPj ,the isomer which mediates calcium flux (8,9), identified with a radiolabeled standard, was noted to increasein both +/+ and lpr T cells.
402
STAFFORD-BRADY
320-
a
ET AL. C
IPZ+ IP,--oIP4--t-
160
Time (min)
FIG.3. Kinetics of phosphoinositide hydrolysis in +/+ (a, c) and Ipr (b, d) T cells in response to 10 rgl ml Con A (a, b) or anti-CD3 hybridoma culture supematant, 15 dilution (c, d). Data represent mean percentage increasesin inositol phosphates over unstimulated controls for the following number of experiments: three in (a) and two in (c) on MRL-+/+ and B6-+/+ T cells; five in (b) and three in (d) on MRLIprflpr, B64prflpr, and C3H-lpr/lpr T cells. The P values for the maximal inositol phosphate production in response to either Con A or anti-CD3 were 0.1 < P < 0.375 for IP2 and IP3 and 0.05 < P < 0.1 for IP,.
The {chain of CD3 in Ipr T cells has been shown to be constitutively phosphorylated on tyrosine residues, whereas such phosphorylation occurred in normal T cells only on activation (5). The role of !: chain phosphorylation is not known. We have shown that such phosphorylation did not interfere with the ability of CD3 to transduce signals, as demonstrated by Con A- and anti-CD3-induced phosphoinositide hydrolysis. In contrast to the comparable inositol phosphate generation in +/+ and Zpr T cells, [Ca2”]i, under basal and stimulated conditions, was significantly lower in /pr T cells. Inositol phosphates, in particular 1,4,5-IP3, have been shown to mediate a rise in [Ca2’]i by intracellular mobilization (8), and probably also by influx through calcium channels (9). We found that prior addition of 2 mM EGTA prevented the Con Ainduced [Ca”]i increase in both +/+ and lpr T cells. This suggeststhat calcium influx plays an important role in the increase in [Ca”]i in both cell types. EGTA, however, not only may chelate extracellular calcium and prevent calcium influx, but also may deplete intracellular stores (14). These data, therefore, do not rule out a contribution to the rise in [Ca2’]i from internal stores. Stimulation of normal T cells generally first results in mobilization of intracellular calcium (8, 14). Kf efflux then serves to hyperpolarize the membrane, which would potentially augment calcium influx across a voltage-gated calcium channel. Chandy et al. have reported that lpr T cells
SIGNAL TRANSDUCTION
IN Ipr T CELLS
403
a
0
10
20
Time (min) FIG. 4. Calcium response to Con A in +/+ and lpr T ceIIs. [Ca”]i was measured in fura-2-loaded cells stimulated with 10 &ml Con A (a) or 2.5 &ml Con A (b). Data represent mean [Ca”]i and SEM for eight experiments each on -t/+ and ZprT cells in (a) and four experiments each in (b). Results include at least one experiment each on +/+ and ZprT cells of the three strains of mice studied, MRL, C3H, and B6.
have I-type voltage-gated K+ channels, which open at 0 mV, in contrast to n and n’ K+ channels on normal T cells, which open at -30 mV ( 15, and G. Chandy, personal communication). The higher membrane potential required to open Kt channels in lpr T cells could result in a reduced calcium response to mitogen. The lower [Ca2’]i obtained on stimulation may contribute to the defect in signal transduction in lpr T cells. We do not know whether the reduced [Ca2’]i is solely responsible for the activation defect in Ipr T cells. Many other abnormalities have been noted in these cells, such asthe absenceof TAP, a glycosylphosphatidylinositol-linked protein, which may play an important role in activation, as suggestedby the inability of TAP-negative hybridomas to be activated through the TCR (16). Double positive thymocytes share with Ipr T cells an absence of TAP, as well as an inability to secrete IL-2, express IL-2 receptor, or proliferate in responseto T-cell mitogens. It is of interest that these immature cells also demonstrate normal phosphoinositide hydrolysis (Pechet and Yeh, unpublished observations) and a lower calcium response to anti-CD3 than mature T cells (17). Further analysis of these naturally occurring signaling variants may allow us to arrive at a deeper understanding of signal transduction in normal T cells. ACKNOWLEDGMENTS We gratefully acknowledge Dr. Kenneth Rock and Dr. Jeffrey Bluestone for their contribution of hybridomas. We thank Dr. Stephen Krane for his support and encouragement. This work was supported by NIH
404
STAFFORD-BRADY
ET AL.
Grants AR-03564, AR-19427, AR-38018, D&39773, DK-38452, and DK-38165. E. T. H. Yeh is the recipient of an Arthritis Investigator Award and an Upjohn Scholar Award. J. V. Bonventre is an Established Investigator of the American Heart Association.
REFERENCES Theofilopoulos, A. N., and Dixon, F. J., Adv. Irnmunol. 37,269, 1985. Katagiri, K., Katagiri, T., Eisenberg, R. A., Ting, J., and Cohen, P. L., J. Immunol. 138,149,1987. Gause, W. C., Mountz, J. D., and Steinberg, A. D., .I. Immunol. 140,1,1988. Sy, M.-S., Wang, P. T. H., Ju, S.-T., Weston, K. M., Alarcon, B., Terhorst, C., and Yeh, E. T. H., Cell. Zmmunol. 113,82, 1988. 5. Samelson, L. E., Davidson, W. F., Morse, H. C., III, and Klausner, R. D., Nufure (London)324,674, 1986, 6. Scholl, W., Isakov, N., Mally, M. I., Theofdopoulos, A. N., and Altman, A., J. Biol. Chem.263,3626,
1. 2. 3. 4.
1988. 7. Wofsy, D., Murphy, E. D., Roths, J. B., Dauphinee, M. J., Kipper, S. B., and Talal, N., J. Exp. Med. 154,1671,1981. 8. Weiss, A., Imboden, J., Hardy, K., Manger, B., Terhorst, C., and Stobo, J., Annu. Rev.Zmmunol.4, 593,1986. 9. Kuno, M., and Gardner, P., Nature (London)326,301, 1987. 10. Isakov, N., Mally, M. I., Scholz, W., and Altman, A., Immunol. Rev.95,89, 1987. 11. Cunha-Melo, J. R., Dean, N. M., Moyer, J. D., Maeyama, K., and Beaven, M. A., J. Eiol. Chem.262, 11455,1987. 12. Gtynkiewicz, G., Poenie, M., and Tsien, R. Y., J. Biol. Chem.260,3440,1985. 13. Dumont, F. J.,J. Zmmunol.U&4106, 1987. 14. Gelfand, E. W., Mills, G. B., Cheung, R. K., Lee, J. W. W., and Grinstein, S.,Immunol. Rev.95,59, 1987. 15. Chandy, K. J., DeCoumey, T. E., Fischbach, M., Talal, N., Cahalan, M. D,, and Gupta, S., Science 233,1197, 1986. 16. Yeh, E. T. H., Riser, H., Bamezai, A., and Rock, K. L., Cell52,665, 1988. 17. Havran, W. L., Poenie, M., Kimura, J., Tsien, R., Weiss, A., and Allison, J. P., Nature(London)330, 170,1987.
IMMUNOLOGY
123,396-404 (1989)
Defective Signal Transduction
in CD4XD8-
T Cells of Ipr Mice
FRANCES STAFFORDBRADY,* EUI SUGIYAMA,* DWIGHT R. ROBINSON,* MAN-SUNSY,~ JOSEPH V. BONVENTRE,~ AND EDWARD T. H, YEH* *Arthritis and$Renal Units of the Department ofMedicine, and iDepartment ofPathology, Massachusetts Generai Hospital, Harvard Medical School, Boston, Massachusetts 021I4 Received April 15, 1989; accepted J&y 17, 1989 Mice homozygous for the Ipr gene develop a lymphoproliferative disorder due to expansion of a subset of CXWCDK T cells. Triggering of the T-cell receptor in these /pr T cells does not lead to tramlocation of protein kinase C or phosphorylation of CR3, interleukin-2 production, or proliferation, whereas a combination of phorbol ester and calcium ionophore does. Stimulation with concanavalin A or anti-CD3 induces phosphoinositide hydrolysis. The rise in inositol bisphosphate, inositol triphosphate, and inositol tetralcisphosphate, identified by HPLC, is similar in +/+ and Ipr T cells. The concentration of cytoplasmic free calcium ([Caz’li), however, under basal and stimulated conditions is significantly lower in lpr T cells. The lower basal [Ca2+li may explain why induction of proliferation with phorbol ester and calcium ionophore requires a higher concentration of ionophore in these cells than in normal T cells. The lower [Ca2’li obtained on stimulation may contribute to the activation defect of CD4CD8- lpr T cells. Q 19%9Academic Press,Inc.
INTRODUCTION Mice homozygous for the Ipr gene develop generalizedlymphadenopathy due to accumulation of a subsetof abnormal T cells (1). The phenotype of lpr abnormal T cellsis similar to that of a small subsetof CD4-CDS- thymocytes (2-4). T-cell recep tor (TCR) is expressedon their surface(4, S),albeit at a lower density than on normal T cells(6). Stimulation of lpr T cellsby mitogens,however,doesnot result in proliferative responses( 1,4,7). Triggering of the TCR in normal T cells activatesphospholipaseC, which hydrolyzes phosphatidylinositol 4,5-bisphosphate to inositol 1,4,5-trisphosphate( 1,4,51P&and diacylglycerol(8). 1,4,5-IP3raisesthe concentration of cytoplasmic free calcium ([Ca2’li), by intracellular mobilization (8), and probably also by influx through a calcium channel (9). A combination of calcium ionophore, which raises [Ca’+]i, and phorbol my&ate acetate(PMA), which mimics diacylglycerol in activation of protein kinaseC, induces proliferation ofnormal T cells,highlighting the importance of thesesecondmessengersin T-cell activation (8,10). It hasbeen reported that mitogen-induced phosphoinositide hydrolysis is deficient in lpr T cells (6). Concomitant changesin [Ca”]i have not been studied. Protein kinase C cannot be activated by lectin ($6) but can be activated by PMA (4-6), asjudged by translocation of enzyme activity from cytosol to membrane (4,6), and phosphorylation of CD3 y and c (5). 394 ooo8-8744/89 $3.00 Copyright 0 1989 by Academic Refs, Inc. All rights ofreproduction in any fwm rescrvgd.
SIGNAL
TRANSDUCTION
IN Ipr T CELLS
397
These observations suggest that lpr T cells have a defect in the signaling pathway between triggering of the TCR and activation of protein kinase C. To further analyze the defect in signal transduction in lpr T cells, we examined phosphoinositide hydrolysis and correlated this with changes in [Ca2’]i. We found that mitogen-induced generation of inositol phosphates was similar in these cells and congenic +/+ T cells. The concomitant rise in [Ca2’]i, however, was consistently lower than that in normal T cells. This subnormal calcium response may partly account for the activation defect in lpr T cells. MATERIALS AND METHODS
Animals and Reagents MRL/MpJ (MRL), C3H/HeJ (C3H), and C57BL/6 (B6) mice homozygous for the
Ipr gene and their +/+ counterparts (Jackson Laboratory, Bar Harbor, ME) were used between the agesof 4 and 6 months. Low-Tox rabbit complement was obtained from Cedarlane Laboratories (Westbury, NY). All other reagents were from Sigma. Anti-CD3 (145-2Cll) and other monoclonal antibodies were used as cell culture supernatants.
T-Cell Purification Abnormal lpr T cells were purified from lymph nodes of mice homozygous for the lpr gene by treatment with monoclonal antibodies to CD4 (GK 1.5; ATCC, Rockville, MD), CD8 (53.6.72; ATCC), and T-cell-activating protein (TAP; 3E7.1 or lE7) for 30 min on ice, followed by complement-mediated lysis for 30 min at 37°C (4). The term lpr T cells throughout the paper refers to this population of purified abnormal lymph node T cells. Normal T cells were obtained by passageof spleen cells from +/+ mice over nylon wool columns. In vitro proliferation assayswith concanavalin A (Con A) and anti-CD3 were performed to check the purity of the +/+ and lpr T cells.
In Vitro Proliferation Assay Cells were resuspended in RPM1 1640 medium supplemented with 10%heat-inactivated fetal calf serum, 2 mM L-glutamine, 100 U/ml penicillin, 100 &ml streptomycin, 0.25 pg/ml fungizone, 0.1 m&f nonessential amino acids, 20 mM Hepes, and 5 X lOA M 2-mercaptoethanol; and 5 X lo5 cells in a final volume of 200 &well were cultured for 48 hr in 96-well llat-bottom tissue culture plates. [3H]Thymidine ( 1 &i/well) was added for the last 16 hr of culture. Cells were harvested with a PHD automatic cell harvester (Cambridge, MA) and were counted in a scintillation counter. Data are represented as the mean and SEM of triplicate cultures.
PhosphoinositideHydrolysis Cells were labeled overnight with 25 &i/ml myo-[2-3H]inositol (Amersham Corp., Amersham, UK) in inositol-free RPM1 1640 supplemented with 1% BSA, ~-@amine, and antibiotics. For each experiment, 10’ cells were washed and incubated in 0.6 ml inositol-free RPM1 1640 for 40 min in a 5% CO* incubator. LiC12 5 mM was added for the last 10 min of incubation in half of the experiments on both normal
398
STAFFORD-BRADY
ET AL.
and abnormal T cells. The increase in inositol polyphosphates on stimulation was similar in the experiments with and without LiC12. Following stimulation, ice-cold 10% TCA was added to stop the reaction and the reaction mixture was centrifuged. The supernatant was treated with 1 ml ice-cold Freon amine and centrifuged again. The aqueous phase was applied to a Partisil SAX 10 HPLC column (Whatman, Maidstone, England) and eluted with a gradient of ammonium phosphate, as previously described (11). The eluant was monitored using a Berthold LB506C in-line HPLC scintillation counter. Identification of different inositol phosphates was based on the retention time of standard 3H-labeled inositol 1-phosphate, inositol 1,Cbisphosphate, inositol 1,4,5-trisphosphate, and inositol 1,3,4,Qetrakisphosphate (Amersham) (Fig. la). The inositol phosphate peaks were wider in the experimental tracings, presumably due to the presence of other isomers, e.g., inositol 1,3-bisphosphate, inositol 3,4-bisphosphate and inositol 1,3,4-trisphosphate ( 11). All isomers were included in the calculations. An unpaired Student’s t test was used to compare generation of inositol phosphates in +/+ and Ipr T cells. Intracellular Free Calcium Mean [Ca2+]i was measured in populations of cells before and after stimulation. A Deltascan spectrophotometer (Photon Technology International, Princeton, NJ) was used with dual excitation at 340 and 380 nm and an emission wavelength of 5 10 nm. Cells at a density of 5 X 106/ml in RPM1 1640 supplemented with 1.5%gelatin were loaded with 2.5 &ml of 1 mM membrane-permeant fura- AM (Molecular Probes, Inc., Eugene, OR) in dimethyl sulfoxide (DMSO) at 37°C for 1 hr. For each measurement, 10’ cells were resuspended in 2 ml of buffer, consisting of 145 mM NaCI, 5 mM KCl, 0.5 mM MgS04, 1 mM Na2HP04, 5 mA4 glucose, 20 mM Hepes, 0.5% BSA, 0.5% gelatin, and 2 nut4 CaC12in water (pH 7.4, 37°C). Autofluorescence of unloaded cells incubated with DMSO and treated in a similar fashion was subtracted. Calibration of [Ca2+li was performed at the end of each experiment using 10 p&Y ionomycin, followed by 0.1 nut4 digitonin (R,ax) and 15 mM EGTA + 24 r&l4 Tris (R,i”), The [Ca2+]i was calculated according to Grynkiewicz et al. (12). Statistical significance of the difference between [Ca2+]i in +/+ and lpr T cells was determined by an unpaired Student’s t test. RESULTS Purified CD4CD8- T cells from the lymph nodes of MRL-lpr/lpr mice did not proliferate in response to Con A or anti-CD3 plus PMA, whereas unselected T cells from lymph nodes of the same mice did (Table 1). The total population contained a minority of normal T cells, which accounted for the proliferative responsesseen. We previously reported that proliferative responses,seenin normal T cells, were not seen in CD4-CD8- lpr T cells on treatment with PMA and 0.06 ~~ calcium ionophore A23 187 (4). Calcium ionophore (A23 187 or ionomycin), at a concentration 4 to 20 times greater than that required to activate normal T cells was, however, able to synergize with PMA to induce proliferation of Ipr T cells (Table l), as has been previously reported (2, 13). Similar results were obtained with T cells from B6 and C3H mice (data not shown). The proliferation data suggestedthat lpr T cells have a defect in the signaling pathway between triggering of the TCR and activation of protein kinase C. It was, there-
SIGNAL TRANSDUCTION
399
IN Ipr T CELLS
TABLE 1 Proliferative Responsesof Lymph Node T Cells of MRL-lpr/lpr
Mice
[ ‘H]Thymidine incorporation (cpm X 10e3) Stimulation Experiment 1 Medium Con A 2.5 &ml Anti-CD3 (1:lO) PMA 2 rig/ml Anti-CD3 (1: 10) + PMA 2 rig/ml Experiment 2 Medium PMA 5 rig/ml Ionomycin 250 rig/ml Ionomycin 1 pg/ml Ionomycin 250 rig/ml + PMA 5 rig/ml Ionomycin 1 &ml + PMA 5 rig/ml
Total lymph node T cells
CD4- CD8- lpr T cells
0.5+ 44.9 * 3.4 f 0.9 f 96.3 -I
0.1 1.5 0.5 0.4 18
0.3fO.l 0.4kO.l 0.5 Ii 0.3 0.2 kO.1 0.6~0.1
0.5 f 2.4+ 1.3* 0.8% 93.2 f 103.4+
0.1 0.2 0.1 0.1 4.7 2.5
2.0 k 0.2 1.3 + 0.3 1.6 kO.5 0.8 +O.l 7.8 f 1.0 84.0 -e 2.8
Note. 5 X lo5 untreated lymph node cells or purified CD4-CDS-TAP- T cells from MRL-Ipr/lpr mice were cultured in the presence of the indicated reagents. Cultures were incubated for 48 hr, pulsed with 1 &/well [ 3Hlthymidine for the last 16 hr, harvested, and counted.
fore, of interest to study production of second messengersin response to Con A or anti-CD3. Stimulation of abnormal B64pr/lpr T cells with 5 pg/ml Con A for 20 min resulted in a rise in inositol bisphosphate (IP,), inositol trisphosphate (IP& and inositol tetrakisphosphate (1P.J (Fig. 1). Similar results were obtained using MRLlpr/lpr and C3H-Zpr/lpr T cells (Figs. 2 and 3). Lack of a proliferative response to Con A or anti-CD3 confirmed that the ipr T cells were adequately purified (data not shown). Generation of inositol phosphates in response to varying dosesof Con A was similar in +/+ (Fig. 2a) and lpr (Fig. 2b) T cells. The kinetics of phosphoinositide hydrolysis in response to Con A and anti-CD3 were examined (Fig. 3). On stimulation with 10 pg/ml Con A for 20 min, the mean percentage increase in IP2, IP3, and IP, in +/+ T cells was 186, 159, and 112%, respectively (Fig. 3a), and in lpr T cells was 224, 111, and 3 12%(Fig. 3b). These increasesrepresented maximal inositol phosphate production and were comparable in the two cell types. Generation of inositol phosphates in response to anti-CD3 was maximal at 10 min and was similar in +/+ (Fig. 3c) and lpr (Fig. 3d) T cells. In view of the normal production of IP3 on activation of lpr T cells, it was of interest to determine concomitant changesin [Ca2’]i. A rise in [Ca2+]iwas seenin the normal and abnormal cells following stimulation (Fig. 4). This calcium response provided further evidence for phosphoinositide hydrolysis in Zpr T cells. Although there was considerable overlap, basal [Ca2’]i tended to be lower in the abnormal T cells, with a mean value of 52 nM compared with 65 nM in the control T cells (P < 0.0 1). At all time points following stimulation with Con A at 10 pg/ml (Fig. 4a) and 2.5 pg/ml (Fig. 4b), the [Ca2+]i was significantly lower in lpr than in +/+ T cells (P < 0.05). Ionomycin at 10 PMinduced a similar increase in [Ca*+]i in the normal and abnormal T cells. There was no rise in [Ca2+]i in response to 10 pug/mlCon A in both +/+ and
400
STAFFORD-BRADY
127
ET AL.
I .I-IPZ
a
cps I.3.4.S.IP4
0
_Ik C
Time (min) FIG. 1. Generation of inositol phosphates by B6-lpr/lpr abnormal T cells. The elution profile of ‘Hlabeled standard inositol phosphates is shown in (a). Myo-[2-3H]inositol-labeled cells were unstimulated (b) or stimulated with 5 &ml Con A for 20 min (c). Water-soluble inositol phosphates were extracted and separated by HPLC. Two isomers of IPr, presumably 1,3-IPr and 1,4-IP2, were separated in (c). Labeled substanceseluted prior to IP, included inositol and glycerophosphorylinositol.
lpr T cells when incubated for 5-10 min in calcium-free buffer to which 2 mMEGTA was added 2-4 min prior to stimulation (data not shown). The spectral characteristics of intracellular fura- were similar in +/+ and Ipr T cells. The difference in [Ca”]i, therefore, did not appear to be due to different handling of the fluorescent dye by the two cell types.
SIGNAL TRANSDUCTION
IN lpr T CELLS
401
ConA (&ml) FIG. 2. Generation of inositol phosphates in +/+ (a) and Ipr (b) T cells in response to stimulation with varying dosesof Con A for 20 min. Data represent mean percentage increases in inositol phosphates over unstimulated controls, for three experiments on normal T cells from MRL-+/+ and B6-+/+ mice (a) and five experiments on abnormal T cells from MRL-lpr/lpr, B6-Zpr/lpr, and C3H-Ipr/lpr mice (b). The mean counts per second in unstimulated cells for IP2, IPX, and IP, in +/+ T cells were 88 + 23,39 k 6, and 39 + 2, respectively; and in lpr T cells these were I 18 f 18,6 1 -t 6, and 34 +-3, respectively. For all comparisons 0.1 < P < 0.4, except for 1P4production in responseto 10 &ml Con A, for which 0.05 < P < 0.1.
DISCUSSION Scholz et al. reported an increasein total inositol phosphatesof 57-66% in +/+ T cells and only 13-l 8% in lpr T cells in responseto PHA (6, lo), and 44% in +/+ and 17% in Zpr T cells in responseto Con A (10). We have demonstrated substantial increasesin IP,, IP,, and IP, in both +/+ and lpr T cells. In fact, the rise in inositol phosphatestended to be higher in lpr T cells.The discrepanciesmay be accountedfor by differencesin labeling time and inositol phosphateseparationtechnique. Scholzet al. labeled the cells for 2 hr, in contrast to our 16 hr. Furthermore, they separatedthe inositol phosphateson a Dowex column and reported only changesin total inositol phosphates.Changesin IP2, IP3, and IP, may have been masked when these were not measured separately from more abundant labeled molecules, such as IP, and glycerophosphorylinositol (Fig. 1). We used a sensitiveHPLC technique, which allowed us to clearly separateand quantitate increasesin IPz, IP3, and IP, (Fig. 1). Although the separation was not adequateto measureIP3 isomersindividually in all experiments,the concentration of 1,4,5-IPj ,the isomer which mediates calcium flux (8,9), identified with a radiolabeled standard, was noted to increasein both +/+ and lpr T cells.
402
STAFFORD-BRADY
320-
a
ET AL. C
IPZ+ IP,--oIP4--t-
160
Time (min)
FIG.3. Kinetics of phosphoinositide hydrolysis in +/+ (a, c) and Ipr (b, d) T cells in response to 10 rgl ml Con A (a, b) or anti-CD3 hybridoma culture supematant, 15 dilution (c, d). Data represent mean percentage increasesin inositol phosphates over unstimulated controls for the following number of experiments: three in (a) and two in (c) on MRL-+/+ and B6-+/+ T cells; five in (b) and three in (d) on MRLIprflpr, B64prflpr, and C3H-lpr/lpr T cells. The P values for the maximal inositol phosphate production in response to either Con A or anti-CD3 were 0.1 < P < 0.375 for IP2 and IP3 and 0.05 < P < 0.1 for IP,.
The {chain of CD3 in Ipr T cells has been shown to be constitutively phosphorylated on tyrosine residues, whereas such phosphorylation occurred in normal T cells only on activation (5). The role of !: chain phosphorylation is not known. We have shown that such phosphorylation did not interfere with the ability of CD3 to transduce signals, as demonstrated by Con A- and anti-CD3-induced phosphoinositide hydrolysis. In contrast to the comparable inositol phosphate generation in +/+ and Zpr T cells, [Ca2”]i, under basal and stimulated conditions, was significantly lower in /pr T cells. Inositol phosphates, in particular 1,4,5-IP3, have been shown to mediate a rise in [Ca2’]i by intracellular mobilization (8), and probably also by influx through calcium channels (9). We found that prior addition of 2 mM EGTA prevented the Con Ainduced [Ca”]i increase in both +/+ and lpr T cells. This suggeststhat calcium influx plays an important role in the increase in [Ca”]i in both cell types. EGTA, however, not only may chelate extracellular calcium and prevent calcium influx, but also may deplete intracellular stores (14). These data, therefore, do not rule out a contribution to the rise in [Ca2’]i from internal stores. Stimulation of normal T cells generally first results in mobilization of intracellular calcium (8, 14). Kf efflux then serves to hyperpolarize the membrane, which would potentially augment calcium influx across a voltage-gated calcium channel. Chandy et al. have reported that lpr T cells
SIGNAL TRANSDUCTION
IN Ipr T CELLS
403
a
0
10
20
Time (min) FIG. 4. Calcium response to Con A in +/+ and lpr T ceIIs. [Ca”]i was measured in fura-2-loaded cells stimulated with 10 &ml Con A (a) or 2.5 &ml Con A (b). Data represent mean [Ca”]i and SEM for eight experiments each on -t/+ and ZprT cells in (a) and four experiments each in (b). Results include at least one experiment each on +/+ and ZprT cells of the three strains of mice studied, MRL, C3H, and B6.
have I-type voltage-gated K+ channels, which open at 0 mV, in contrast to n and n’ K+ channels on normal T cells, which open at -30 mV ( 15, and G. Chandy, personal communication). The higher membrane potential required to open Kt channels in lpr T cells could result in a reduced calcium response to mitogen. The lower [Ca2’]i obtained on stimulation may contribute to the defect in signal transduction in lpr T cells. We do not know whether the reduced [Ca2’]i is solely responsible for the activation defect in Ipr T cells. Many other abnormalities have been noted in these cells, such asthe absenceof TAP, a glycosylphosphatidylinositol-linked protein, which may play an important role in activation, as suggestedby the inability of TAP-negative hybridomas to be activated through the TCR (16). Double positive thymocytes share with Ipr T cells an absence of TAP, as well as an inability to secrete IL-2, express IL-2 receptor, or proliferate in responseto T-cell mitogens. It is of interest that these immature cells also demonstrate normal phosphoinositide hydrolysis (Pechet and Yeh, unpublished observations) and a lower calcium response to anti-CD3 than mature T cells (17). Further analysis of these naturally occurring signaling variants may allow us to arrive at a deeper understanding of signal transduction in normal T cells. ACKNOWLEDGMENTS We gratefully acknowledge Dr. Kenneth Rock and Dr. Jeffrey Bluestone for their contribution of hybridomas. We thank Dr. Stephen Krane for his support and encouragement. This work was supported by NIH
404
STAFFORD-BRADY
ET AL.
Grants AR-03564, AR-19427, AR-38018, D&39773, DK-38452, and DK-38165. E. T. H. Yeh is the recipient of an Arthritis Investigator Award and an Upjohn Scholar Award. J. V. Bonventre is an Established Investigator of the American Heart Association.
REFERENCES Theofilopoulos, A. N., and Dixon, F. J., Adv. Irnmunol. 37,269, 1985. Katagiri, K., Katagiri, T., Eisenberg, R. A., Ting, J., and Cohen, P. L., J. Immunol. 138,149,1987. Gause, W. C., Mountz, J. D., and Steinberg, A. D., .I. Immunol. 140,1,1988. Sy, M.-S., Wang, P. T. H., Ju, S.-T., Weston, K. M., Alarcon, B., Terhorst, C., and Yeh, E. T. H., Cell. Zmmunol. 113,82, 1988. 5. Samelson, L. E., Davidson, W. F., Morse, H. C., III, and Klausner, R. D., Nufure (London)324,674, 1986, 6. Scholl, W., Isakov, N., Mally, M. I., Theofdopoulos, A. N., and Altman, A., J. Biol. Chem.263,3626,
1. 2. 3. 4.
1988. 7. Wofsy, D., Murphy, E. D., Roths, J. B., Dauphinee, M. J., Kipper, S. B., and Talal, N., J. Exp. Med. 154,1671,1981. 8. Weiss, A., Imboden, J., Hardy, K., Manger, B., Terhorst, C., and Stobo, J., Annu. Rev.Zmmunol.4, 593,1986. 9. Kuno, M., and Gardner, P., Nature (London)326,301, 1987. 10. Isakov, N., Mally, M. I., Scholz, W., and Altman, A., Immunol. Rev.95,89, 1987. 11. Cunha-Melo, J. R., Dean, N. M., Moyer, J. D., Maeyama, K., and Beaven, M. A., J. Eiol. Chem.262, 11455,1987. 12. Gtynkiewicz, G., Poenie, M., and Tsien, R. Y., J. Biol. Chem.260,3440,1985. 13. Dumont, F. J.,J. Zmmunol.U&4106, 1987. 14. Gelfand, E. W., Mills, G. B., Cheung, R. K., Lee, J. W. W., and Grinstein, S.,Immunol. Rev.95,59, 1987. 15. Chandy, K. J., DeCoumey, T. E., Fischbach, M., Talal, N., Cahalan, M. D,, and Gupta, S., Science 233,1197, 1986. 16. Yeh, E. T. H., Riser, H., Bamezai, A., and Rock, K. L., Cell52,665, 1988. 17. Havran, W. L., Poenie, M., Kimura, J., Tsien, R., Weiss, A., and Allison, J. P., Nature(London)330, 170,1987.