Glycosylphosphatidylinositol: a candidate system for interleukin-2 signal transduction.

The mechanism of interleukin-2 (IL-2) signal transduction was analyzed by use of an inducible B lymphoma. Like normal antigen-activated B lymphocytes, the lymphoma cells respond to IL-2 by proliferating and differentiating into antibody-secreting cells; both responses are blocked by a second interleukin, IL-4. Analyses of the signaling pathway showed that IL-2 stimulated the rapid hydrolysis of an inositol-containing glycolipid to yield two possible second messengers, a myristylated diacylglycerol and an inositol phosphate-glycan. The myristylated diacylglycerol response exhibited the same IL-2 dose dependence as the growth and differentiative responses, and the generation of both hydrolysis products was inhibited by IL-4. These correlations implicate the glycosyl-phosphatidylinositol system in the intracellular relay of the IL-2 signal.

Suppressive effects of insulin and insulin-like growth factor-1 (IGF1) on immune responses.

This paper describes conditions wherein the serum peptides insulin and IGF1, which are typically associated with growth-promoting functions, can suppress in vitro immune responses. IL 2-induced proliferation of lymphocytes as well as in vitro antibody-producing cultures are suppressed by physiologic concentrations of IGF1 or by superphysiologic concentrations of insulin. Suppression of IL 2-induced proliferation is not overcome by increasing the IL 2 concentration and is mediated only during the first 24 to 48 hr of the 110-hr incubation period required to measure the proliferative response to IL 2. By analogy to other biologic systems, these effects of insulin and of IGF1 are probably mediated by occupancy of the IGF1-receptor, which is cross-occupied by insulin at superphysiologic concentrations. These data support the possibility of a novel function for these endocrine and endocrine-like peptides and also expands their range of biologic activities to within the immune system.

Interleukin 2-dependent phosphorylation of interleukin 2 receptors and other T cell membrane proteins.

The addition of IL 2 to Con A-activated splenic T cells induced the rapid and time-dependent phosphorylation of membrane proteins with m.w. of 115,000 to 105,000, 90,000, and 66,000, and to a lesser extent 55,000 to 58,000, 40,000, and 34,000. Immunoprecipitations conducted with an anti-IL 2 receptor antibody indicated that the murine IL 2 receptor (55,000 to 58,000) was included in the set of IL 2-dependent phosphoproteins. Phosphorylation of these same proteins was also seen after IL 2 treatment of PHA-activated T cells and of the IL 2-dependent line CTLL-2. Membrane phosphorylation was dependent on physiologically relevant IL 2 concentrations (0.2 to 1 ng/ml), and was detected as early as 1 min after IL 2 addition, with maximal levels of phosphorylation achieved by 15 min. In contrast to these observations, the pattern of cytoplasmic protein phosphorylation remained unchanged after IL 2 addition, although IL 2 did augment the level of preexisting cytoplasmic phosphorylation induced by lectin. The pattern of membrane protein phosphorylation induced by IL 2 also overlapped in part with that induced after stimulation of Con A-activated T cells with the phorbol ester PMA. IL 2-stimulated phosphorylation was inhibited by the addition of agents that both stimulate cyclic AMP-dependent protein kinases and block lymphocyte mitogenesis. No effect was seen upon addition of agents that enhance cyclic GMP-dependent protein kinases. These observations support a role for specific membrane as opposed to cytoplasmic protein phosphorylation in the regulation of lymphocyte growth by IL 2, and also suggest that protein kinase A, and perhaps protein kinase C, participate as regulators of the IL 2 signaling mechanism.

Assessment of the diabetogenic drugs alloxan and streptozotocin as models for the study of immune defects in diabetic mice.

The efficacy of the diabetogenic drugs streptozotocin and alloxan were evaluated as models for the study of immune defects associated with diabetes. Streptozotocin- or alloxan-treated mice, with a stable hyperglycaemia of 25-33 mmol/l plasma glucose, were severely impaired in their ability to mount antibody forming, mitogenic, or delayed-type hypersensitivity responses in vivo. Treatment of alloxan-diabetic mice with insulin in vivo completely reversed all immune defects, while insulin treatment of streptozotocin-diabetic mice restored immune function to only 70-80% of normal levels. Results obtained by viability measurements and in vitro biological assays of lymphoid function, including proliferation in response to T- and B-cell mitogens, the production of interleukin-2 by T cells, and the production of interleukin-1 by macrophages indicated that direct exposure to alloxan for 48 h (at concentrations less than or equal to 14 mmol/l) had no adverse effects on lymphoid activity, while exposure to streptozotocin was routinely toxic at concentrations greater than or equal to 1 mmol/l. Both alloxan and streptozotocin exhibited strong toxicity in vitro for isolated pancreatic islet cells. Finally, lymphocytes from streptozotocin-diabetic mice, or cells incubated in vitro with streptozotocin, contained numerous chromosomal abnormalities indicative of DNA strand breakage. Such abnormalities were absent in alloxan-diabetic mice and in cells incubated with alloxan in vitro. These results indicate that immune dysfunction associated with streptozotocin is attributable to direct and irreversible impairment of lymphoid cell function and viability. In contrast, immune dysfunction associated with alloxan-diabetes appears to be a consequence of the diabetic state.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of a monoclonal rat anti-mouse interleukin 2 (IL-2) receptor antibody and its use in the biochemical characterization of the murine IL-2 receptor.

Anti-murine interleukin 2 (IL-2) receptor monoclonal antibodies (mAb) were made from rats immunized with murine cytotoxic lymphocytes. One mAb, designated M7/20, strongly inhibited the proliferation of both IL-2 dependent CTLL-2 cells and concanavalin A (Con A)-induced T-cell blasts. Inhibition was linearly dependent on the concentrations of both M7/20 and IL-2. Utilizing FACS analysis, M7/20 was shown to bind selectively to mitogen-activated T lymphocytes and, to a lesser degree, to activated B lymphocytes. 125I-Labeled M7/20 binding assays indicated that 48-hr Con A-induced T-cell blasts possessed 89,000 binding sites/cell with a Kd of 1.2 X 10(-9) M. Competitive binding analyses indicated that M7/20 and IL-2 occupy the same or overlapping cell surface sites. Preliminary biochemical characterization of M7/20 immunoprecipitates of detergent extracts from both surface-iodinated and internally D-[3H]glucosamine-labeled T lymphoblasts indicated that the murine IL-2 receptor is an N-glycosylated 58,000-Da glycoprotein. Together these results suggest that mAb M7/20 binds at or near the IL-2-binding epitope on the murine IL-2 receptor and, thus, upon manipulation may act as an IL-2 agonist.

Polyclonal B cell stimulation and interleukin 1 induction by the mucoid exopolysaccharide of Pseudomonas aeruginosa associated with cystic fibrosis.

Mucoid exopolysaccharide isolated from Pseudomonas aeruginosa obtained from colonized cystic fibrosis patients was found to be a potent mitogen for mouse lymphocytes. The responding lymphocyte was a B cell, and we found no evidence that T cell could proliferate or synergize with B cells in response to the mucoid exopolysaccharide. Proliferation was not inhibitable by polymyxin B, which blocks lipopolysaccharide (LPS)-induced proliferation, indicating that a minor LPS contaminant in the purified exopolysaccharide was not the mitogenic component. Mucoid exopolysaccharide induced secretion of IgG, suggesting that it is polyclonal mitogen. It also induced splenic adherent cells (macrophages) to produce interleukin 1. We propose that mucoid exopolysaccharide produced by P. aeruginosa present in the lungs of cystic fibrosis patients may have potent in vivo consequences resulting in aberrant immunoregulation and inhibition of effective immune elimination of P. aeruginosa.

Induction of human interleukin-1 by toxic-shock-syndrome toxin-1.

Strains of Staphylococcus aureus isolated from patients with toxic shock syndrome (TSS) make a characteristic protein known as toxic-shock-syndrome toxin-1 (TSST-1), but the role of this protein in the pathogenesis of TSS is not certain. We have purified TSST-1 by using a combination of alcohol precipitation, isoelectric focusing, and gel chromatography. TSST-1 has an isoelectric point of 7.2 and a molecular weight of 23,100, in accordance with previously published determinations for this protein, and is serologically identical to pyrogenic exotoxin C and staphylococcal enterotoxin F. In highly purified form, TSST-1 is a potent inducer of interleukin-1 production by human monocytes, as quantitated in a thymocyte-proliferation assay. This capability is not attributable to contamination by other staphylococcal products or gram-negative endotoxin and can be blocked by hydrocortisone. Many features of TSS suggest that induction of interleukin-1 by TSST-1 in vivo may play a central role in the elaboration of this disease.

Acquisition of paralytic activity by inducer cells.

Clones of inducer cells activated by antigen and class II major histocompatibility complex (MHC) gene products synthesize large amounts of several distinct mRNA species not detected in other activated cell types, including antigen-activated suppressor or killer cell clones. Although inducer cells continued to synthesize and secrete peptides at about the same rate for 5 days after activation by adherent cells and antigen, they expressed a new set of mRNAs approximately 3 days after activation. We therefore tested the functional activity of purified Ly1 cells at different days after activation by adherent cells and antigen. We wished to find out (a) whether there was a qualitative or quantitative change in the level of inducer activity, and if so, (b) whether this functional change was an intrinsic property of Ly1 cells or, rather, was dependent on signals from the cellular environment, in particular from adherent antigen-presenting cells. We incubated purified Ly1 cells and splenic adherent cells for 1-7 days in vitro, and tested inducer activity in cultures containing highly purified B cells and sheep erythrocytes (SRBC). Anti-SRBC plaqueforming cells were counted 5 days later. We found that inducer cells that have been incubated with adherent cells in culture for more than 72 h (a) did not induce B cells to produce antibody, and (b) prevented virgin but not immune B cells from receiving T-helper signals. We term this latter phenomenon "paralysis." Acquisition of the ability to paralyze virgin B cells required an I-E gene-regulated interaction between inducer cells and in vitro-activated adherent cells. This interaction did not require antigen and was associated with transition from Ly1:Qa1- to Ly1:Qa1+ cell-surface phenotype. Taken together these findings indicate that (a) interactions between inducer cells and adherent antigen-presenting cells result first in classical inducer ("helper") activity and later in expression of paralytic activity and (b) sequential expression of these inducer activities depends on two distinct signals, supplied by resting and activated adherent cells, respectively. The signal supplied by autologous activated adherent cells is regulated by I-E gene products and is independent of corecognition of foreign proteins. The physiologic significance of this paralytic inducer activity in the prevention of potentially harmful immune reactions to foreign microbial agents is discussed.

Role of Ly-1:Qa1- and Ly-1:Qa1+ inducer T cells in activation of Ly-23 effectors of suppression of antibody production in mice.

Ly-2+ effectors of T cell-mediated suppression require inducing signals from antigen and a helper cell bearing the Ly-1+:Qa1+ surface phenotype. In this report, we have further examined the helper cell requirements for suppressor cell induction of antibody production in mice. By using the T cell subset education procedure in vitro, we have activated T cells to sheep red blood cells (SRBC) antigens and then purified Ly-2 cells before testing for suppressor activity in assay cultures of defined T and B cell subsets. We have confirmed our previous observations that Ly-1+:Qa1+ cells are required for activation of T suppressors, but have found that under the appropriate conditions, there is not a strict requirement for the Ly-123 subset of T cells. Furthermore, if Ly-23 cells are stimulated in the presence of Ly-1+:Qa1- T cells, effective suppressors can be obtained only if a source of Ly-1:Qa1+ inducers is added to the assay culture. If Ly-23 cells are activated by antigen in the absence of Ly-1 cells, subsequent exposure to the Ly-1+:Qa1+ subset under the conditions tested here is not sufficient to activate suppressors. These results show that effectors of suppression, like B cells and cytotoxic T lymphocytes, may respond to two helper cells.

"Nonspecific" immunoenhancing T cells in tumor-bearing mice include anti-idiotypic subsets.

Splenocytes from DBA/2 mice inoculated 3 wk earlier with syngeneic P815 mastocytoma tumor cells produce increased numbers of antibody plaque-forming cells (PFC) when stimulated with either sheep red blood cells (SRBC) or phosphorylcholine (PC) on Streptococcus pneumoniae R36a in vitro. The nature of this nonspecific hyperreactivity was investigated in mixed cultures of purified splenic T and B cells. The addition of T cells from P815 tumor-bearing mice (TP815) into the cultures of normal B cells produced a significant enhancement of the PFC response to both SRBC and PC, when compared with the effect of normal T cells added to control cultures. The idiotypic profile of the enhanced anti-PC response was studied by a PFC-inhibition assay with monoclonal antibodies against two distinct idiotopic determinants (Id) of the T15 family. Normal B cells produced greater than 90% of T15 Id-positive (Id+) PFC. Addition of normal T cells diminished the proportion of T15 Id+ PFC to approximately 60%, whereas the rest of PFC were Id-. Addition of the immunoenhancing TP815 cells into the normal B cells cultures elevated the number of both T15 Id+ and Id- PFC responses, proportionally. However, when TP815 cells were first incubated on T15 protein-coated dishes and the non-adherent fraction was added to B cell cultures, the anti-PC PFC response remained enhanced but consisted of predominently T15 Id- PFC. These observations suggest that the early stage of P815 tumor growth activates various populations of specific helper/amplifier T cells including subsets with anti-idiotypic activity and that the generalized increase of antibody response to various antigens in tumor-bearing mice may be regarded as a polyclonal activation of specific T cells.

Augmentation of B-cell responsiveness by a tumor-activated T-cell factor.

The growth of the P815 mastocytoma in syngeneic DBA/2 mice led to an activation of Ly1+2- T cells. These T cells produced a soluble factor or factors in culture which, when added to normal spleen cells or B cells in the presence of syngeneic Ly1 cells, caused a genetically unrestricted augmentation of the plaque-forming response toward sheep red blood cells (SRBC). The culture supernatant of the activated T cells did not support the proliferation of an interleukin-2 (IL-2)-dependent cell, nor exhibit properties of late-acting TRF. Active supernatants appeared to affect directly B cells during the first 48 hr of culture with SRBC in such a way as to make them more responsive to antigen-specific Ly1-cell help.

Activation of T cells in tumor-bearing mice.

Subcutaneous transplantation of the syngeneic P815 mastocytoma in DBA/2J mice induced an activation of splenic T cells which resulted in a hyperresponsiveness of the tumor-bearing animal to the unrelated antigens pneumococcal polysaccharide (Pn) and sheep red blood cells (SRBC). These tumor-activated T cells appeared to increase the plaque-forming cell (PFC) potential of suboptimal numbers of spleen cells, caused normal spleen cells to express increased numbers of PFC, and produced lymphokine(s) which also increased PFC responses of normal splenocytes. The tumor-activated T cells responsible for stimulating normal splenocytes in an in vitro antibody response were shown to be Ly+2- cells. The activity of the tumor-activated T-cell supernatants was not genetically restricted and required additional Ly1 T cells in order to induce rigorously clean B cells to produce antibody. The T cells capable of stimulating non-specific antibody responses were also capable of slowing tumor growth when injected with tumor cells in normal recipient mice. These results suggest that T cells activated by tumor antigens release immunostimulatory lymphokines and, at the same time, are capable of leading to inhibition of tumor growth.

Definition of two pathways for generation of suppressor T-cell activity.

Antigen-stimulated Ly1 cells induce T cells from nonimmune donors to develop potent feedback suppressive activity. Suppression is mediated by Ly23 suppressor T (Ts) cells, which are generated from either Ly23 or Ly123 precursors. Ts activity generated from Ly23 precursors requires a strong inducer signal and is rapidly expressed but short lived. In contrast, Ts activity from Ly123 precursors is relatively long lived and is efficiently generated by relatively low levels of inducer signals. Induction of both Ly123 and Ly23 precursors to become Ts cells requires that both cells share genes linked to the Ig-H locus.

Cloned T cells with "co-helper" activity. I. Biologic activities of the clone.

Clones of sheep erythrocyte-(SRBC) specific helper T cells with the surface phenotype Thy-1+, Ly-1+, Ly-2- have been derived that grow in vitro in the absence of exogenous antigen or added growth factors. The IL 2-independent clone, 101.6 has been shown to produce a supernatant factor that augments the primary anti-SRBC but not anti-burro RBC responses of whole spleen cells or Ly-1 T plus B cell cultures. The supernatant does not help B cells directly. This augmenting activity is terminated "co-helper" because the enhancement requires the presence of normal Ly-1 T cells. The supernatant of 101.6 was not shown to contain IL 2; co-helper activity was distinguishable from IL 2 activity by absorption with SRBC but not with Con A blasts, and we observed that co-helper activity does not act on spleen cells that differ at the major histocompatibility complex.

Use of anti-idiotype immunosorbents to isolate circulating antigen-specific T cell-derived molecules from hyperimmune sera.

We immunized four different sheep with antigen-binding material found in the serum of BALB/c mice 4 days after primary immunization with sheep erythrocytes (SRBC). The resultant antibodies made by the sheep contained a specificity(ies) that appeared to react with a dominant idiotype present on SRBC-specific Lyt-2+ T cells. The antiserum made by the sheep markedly inhibited the formation of antigen-specific rosettes by SRBC educated T cells but did not inhibit T cells educated to other heterologous erythrocytes from forming crossreacting rosettes with SRBC or specific rosettes with the homologous erythrocytes. The "anti-Id serum" was depleted of all activity against known immunoglobulin isotypes and light chains and then was used to isolate antigen-binding molecules from mice that were hyperimmunized with SRBC. The ShId+ material so isolated could be divided into two main groups--one that expressed immunoglobulin determinants, and one that did not. The former represented 15-25% of the ShId+ protein isolated and comprised a minority of the anti-SRBC antibody in the anti-SRBC serum; the latter group of proteins bound sheep glycophorin specifically and expressed constant region determinants found on a number of other antigen-specific T cell factors. These experiments suggest that antigen-binding molecules made by T cells display much less heterogeneity than do antibodies and also show that the serum of hyperimmune mice contains significant amounts of T cell-derived antigen-specific immunoregulatory molecules.

Immunochemical characterization of the mucoid exopolysaccharide of Pseudomonas aeruginosa.

Alginic acid-like mucoid exopolysaccharide was isolated from three strains of Pseudomonas aeruginosa obtained from the sputa of patients with cystic fibrosis. Purified mucoid antigens were greater than 99% uronic acid. With a hemagglutination assay, antibody responses to the mucoid exopolysaccharide were documented after immunization of rabbits with either whole mucoid organisms or purified mucoid exopolysaccharide. The mucoid antigen from one strain (no. 2192) was composed predominantly of a single serologic epitope shared among 40 alginate exopolysaccharides from different clinical isolates. The mucoid exopolysaccharide from the other two strains (nos. 1 and 258) had a serotype-specific determinant in addition to the common epitope. Analyses of antibody in sera from normal adults, children, and patients with cystic fibrosis culture-positive and culture-negative for mucoid P. aeruginosa showed a highly significant (P less than 0.001) association between increased hemagglutination titers and positive cultures for mucoid P. aeruginosa.

Immunoglobulin idiotopes expressed by T cells. I. Expression of distinct idiotopes detected by monoclonal antibodies on antigen-specific suppressor T cells.

The idiotopic repertoire expressed by antigen-specific suppressor T cells (Ts) generated by Streptococcus pneumoniae strain R36a (Pn) in BALB/c strain mice was investigated using a panel of five monoclonal anti-idiotopic antibodies against TEPC-15/HOPC-8 myeloma proteins. Previous studies suggested that the anti-idiotopic antibodies recognize distinct idiotopic determinants within the T15 idiotype, and that Pn-reactive B cells express all of those idiotopes as shown by a specific inhibitory effect of the anti-idiotopic antibodies on induction of anti-Pn response in vitro as well as on the mature antibody plaque-forming cells. In this study we asked the question of whether anti-idiotopic (Id) can block the inductive and/or effector phases of generation of Ts which act on the Pn-reactive B cells. The presence of anti-Id during the activation of T cells with Pn did not prevent the generation of Ts. However, suppression mediated by Ts on responder lymphocytes (cultures of spleen cells or B cels) was inhibited (reversed) by four out of five anti-Id. Some of the antibodies recognize hapten (phosphorylcholine)-inhibitable Id in the paratope of Ig whereas others are directed against nonparatopic Id. These data indicate that the antigen receptor on Ts includes VH sequences both within and without the immunoglobulin in paratope, and that the Id repertoir of Ts overlaps with that of B cells.

Immunoregulatory circuits which modulate responsiveness to suppressor cell signals: characterization of an effector cell in the contrasuppressor circuit.

Spleen cells from neonatal animals, placed in culture for 6 days spontaneously develop the ability to block the activity of suppressor T cells, a phenomenon that is referred to as contrasuppression. The effector cell which is derived from the interactions among the cells which comprise a contrasuppressor "circuit" is an Ly-1 T cell. It can be separated from Ly-1 helper cells by three criteria other than function: its generation is dependent on Ly-2+ cells, it is I-J+, and it sticks to the Vicia villosa lectin. Those cells which deliver help to B cells under the experimental conditions studied are not dependent on Ly-2+ cells for generation and neither express determinants that our anti-I-J antisera recognize nor stick to V. villosa. The mechanism by which these Ly-1 contrasuppressor cells function was elucidated by adding them to "'intermediate cultures" containing activated Ly-2 suppressor cells and in vivo immunized Ly-1.1-congenic helper cells. After 48 h in these intermediate cultures, the neonatal Ly-1.2 contrasuppressor cells and the Ly-2 suppressor cells were removed by treatment with the appropriate antiserum plus complement. The remaining activity of the in vivo generated Ly-1.1 helper cells was assayed in fresh cultures of B cells. The contrasuppressor cells not only diminished suppression of the Ly-1 helper cells by the Ly-2 suppressor cells in the intermediate culture, but actually conferred a state of relative resistance to suppression upon the helper cells. This state persisted after the contrasuppressor cells were removed. Why such a cellular circuit, which confers resistance to suppression, might be beneficial to neonatal mice and how considering its attributes might help explain some immunological paradoxes is the subject of discussion.

Contrasuppression. A novel immunoregulatory activity.

We have described an interaction between two T cells subsets that results in interference with the expression of Ly-1-, 2+ (Ly-2) T cell-mediated suppression. We refer to this novel immunoregulatory activity as contrasuppression. The T cell responsible for the induction of contrasuppression (inducer cell) expresses the phenotype Ly-1-, 2+;I-J+;Qa-1+. This phenotype distinguishes it from the suppressor effector cells which we find to be I-J-2.3. An I-J+ soluble mediator from the contrasuppressor inducer cell acts on another cell (acceptor cell) that expresses the phenotype Ly-1+, 2+; I-J+; Qa-1+. This phenotype distinguishes it from T helper cells. Both the inducer cell (or its biologically active mediator) and its acceptor cell are required for the expression of contrasuppression. Because contrasuppressor cells can block the suppressive activity of cell-free mediators released by Ly-2 suppressor T cells, the mechanism of contrasuppression is either separated from or in addition to the inactivation of suppressor cells themselves. The potential importance of contrasuppressor activity in the regulation of suppressor T cell activity in allowing immunologic memory to be expressed and in permitting microenvironmental immune regulation is discussed.

Immunoregulatory circuits that modulate responsiveness to suppressor cell signal. Failure of B10 mice to respond to suppressor factors can be overcome by quenching the contrasuppressor circuit.

The in vitro antibody response of spleen cells from B10 strain mice is not suppressed by factor preparations made by primed Ly-2 T cells, although these preparations can suppress the in vitro antibody response of spleen cells from other mouse strains (1-3)2. The factor preparations from Ly-2 cells contain at least two separable activities: one that acts as a suppressor moiety (Ly-2 T cell suppressor factor [Ly-2 TsF]) and a second factor that acts as an inducer of contrasuppression (Ly-2 TcsiF); the latter initiates a series of cellular interactions that leads to the inhibition of suppression that we refer to as contrasuppression. Removal of components (either cellular or humoral) of the contrasuppressor circuit makes spleen cells from B10 strain mice as easily suppressible as are those of other mouse strains. Thus, removal of the contrasuppressor inducer cell and/or its biologically active product with the use of an anit-J serum, or removal of the functional acceptor of the inducer cell with the same or other (Ly-2; Qa-1) antisera breaks the B10 suppressor barrier. Contrasuppressive activity. but not helper activity can be eluted from anit-I-J immunoabsorbents. The addition of B10 T cells to either B6 or B10 spleen cell culture deprived of acceptor cells for the TcsiF reconstitutes contrasuppression more efficiently than does the addition of C57BL/6 T cells. Ly-2 TcsiF is more cross-reactive than is Ly-2 TsF so that absorption of factor preparations from sheep erythrocyte-primed Ly-2 cells with horse erythrocytes also breaks the B10 suppressor barrier. The hyperresponsiveness of splenic T cells from B10 strains to Ly-2 TcsiF may be an in vitro exaggeration of a normal in vivo process. Thus it is possible that one can take advantage of this unusual situation to help dissect out the cellular and subcellular components of T cell circuits that moldulate sensitivity to immunoregulatory signals.