Inhibition of Cell Cycle Progression by a Synthetic Peptide Corresponding to Residues 65–79 of an HLA Class II Sequence: Functional Similarities but Mechanistic Differences with the Immunosuppressive Drug Rapamycin¹

Tcells are crucial for the defense of the host against infection, and the key to T cell activation is recognition of nonself. Processed peptides from both internal and external sources are bound to MHC molecules and presented to T cells by APC (1). Other regions of MHC molecules interact with T cell coreceptors CD4 or CD8 (2, 3, 4, 5, 6). Not only are MHC molecules critical for inducing immune responses, but they are also important in controlling them. Thymocytes are deleted if they cannot recognize self MHC molecules or if they strongly recognize self peptides in the context of self MHC molecules. In the periphery, recognition of self peptides presented by self MHC molecules can lead to deletion or anergy of mature T cells (7, 8). Recently, it has been demonstrated that lysis of target cells by NK cells can be inhibited by the specific recognition of MHC class I molecules on the target cell by killer cell-inhibitory receptors (KIR)³ on the NK cell (9, 10). Thus, MHC molecules are important for both initiation and regulation of normal immune responses.

In addition to their natural role in host defense and T cell education, MHC molecules play a pivotal role in the process of graft rejection. Both direct recognition of allogeneic MHC and indirect recognition through presentation of allogeneic MHC molecules as peptides in host MHC molecules can stimulate rejection (11, 12). Conversely, MHC molecules have also been implicated in the prevention of graft rejection. Circumstantial evidence comes from studies in which concurrent donor-specific transfusions prolong graft survival in humans, and studies in rodents in which injection of spleen cells or lymphocytes from the donor strain induce tolerance to the graft (13, 14). MHC molecules themselves are shown to induce unresponsiveness by injection of cells transfected with class I or class II donor MHC molecules into graft recipients (14). High doses of purified class I molecules induce tolerance in the rat allograft model; however, this is not the case for purified class II molecules (15, 16). In contrast, peptides corresponding to the β-chain of donor rat class II MHC could initiate unresponsiveness to a transplant via oral or intrathymic administration in the recipient (17, 18). Additionally, synthetic peptides derived from nonpolymorphic regions of MHC class II have inhibitory effects (19).

During the past decade, we have characterized the effects of synthetic peptides corresponding to HLA class I sequences on immune responses. A peptide corresponding to residues 222–235 of the α₃ domain, the CD8 binding domain of the HLA class I molecule, blocks the differentiation of pre-CTL into mature effector cells in an allele-unrestricted manner (5). We reported that peptides corresponding to residues 98–113 or 56–69 of HLA-A2 block CTL responses in an allele-specific manner (20, 21). Finally, a peptide derived from residues 75–84 of the α₁ α-helix of HLA-B2702 inhibits lysis of target cells by CTL in an allele-nonspecific manner (22). This peptide prevents graft rejection in rodent models (23, 24, 25).

We sought to determine whether synthetic peptides corresponding to conserved regions of class II molecules could have similar immunoregulatory effects. After testing a panel of peptides, we found that a peptide corresponding to residues 56–80 of DQA03011 inhibits a variety of immune responses, and this activity is localized to residues 65–79. In this study, we characterize the inhibitory effects of this peptide and describe its mechanism of action.

Peptides were synthesized using Fmoc chemistry on an Applied Biosystems (Foster City, CA) peptide synthesizer, and were purified to greater than 95% purity by HPLC (26), and the composition was confirmed by mass spectrometry. Stock solutions of peptide were made by dissolving peptide in DMSO (25 mg/ml). Cyclosporin A was a gift from Merck Research Laboratories (Rahway, NJ), and rapamycin was a gift from Wyeth Ayerst (Philadelphia, PA). rIL-2 was obtained from Sigma Chemical Co. (St. Louis, MO). Abs specific for cyclin E, p21, p27, and cdk2 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Goat anti-mouse horseradish peroxidase and goat anti-rabbit horseradish peroxidase were obtained from Caltag (South San Francisco, CA). Abs to CD2, CD3, CD4, CD8, CD28, CD45RO, CD69, IL-2Rα,β,γ, class I, and class II were purchased from PharMingen (San Diego, CA). Unless specified, all other reagents were obtained from Sigma Chemical Co.

PBL were obtained from volunteer donors or buffy coats from Stanford Blood Center and were isolated via Ficoll-Hypaque density centrifugation. T cells were isolated following adherence of macrophages on plastic petri dishes and removal of B cells on nylon wool columns (27). T cell purities were greater than 90% by staining with anti-CD3 mAb. Cells were cultured in RPMI 1640 supplemented with 10% FBS, 100 U/ml penicillin/streptomycin, 2 mM l-glutamine, and 1 mM HEPES. For preactivated cells, purified T cells were cultured with 0.8 μg/ml PHA-P for 72 h before stimulation with rIL-2 (100 U/ml), as described (28).

Proliferation assays were performed essentially as described (27). For each stimulation protocol/cell type, the time for the optimal proliferative response was determined empirically. For human PBL, 96-well plates were coated with 50 μl of anti-CD3 mAb (5 μg/ml) in PBS for 1 h at 37°C. PBL were added at 4 × 10⁵ cells/well and peptide at a concentration of 40 μM. The cells were pulsed with [³H]TdR (1 μCi/well) (DuPont NEN, Boston, MA) and were harvested with a PhD cell harvester at 72 h. [³H]TdR incorporation was determined by liquid scintillation counting. Purified T cells were cultured at 2 × 10⁵ cells/well, pulsed at 72 h, and harvested at 96 h. Where indicated, rIL-2 (100 U/ml) or anti-CD28 mAb (1 μg/ml) was added at the start of culture. In indicated assays, PHA-P (5 μg/ml) or PMA and ionomycin (10 ng/ml and 250 ng/ml, respectively) were used to stimulate cells, and with these treatments, cells were pulsed at 24 h and harvested at 48 h. Preactivated cells were stimulated with rIL-2 (100 U/ml), pulsed at 24 h, and harvested at 48 h. Cyclosporin A (100 nM) and rapamycin (100 nM) were used in indicated experiments.

PBL (4 × 10⁶/ml) were incubated in the presence of peptide (40 μM) or rapamycin (100 nM) with or without immobilized anti-CD3 mAb (5 μg/ml) for 24 h at 37°C in 2-ml wells. The cells were then washed three times with PBS and were replated at the same density. After a 7-day incubation, the cells were washed, plated at equivalent densities in 96-well microtiter plates, and restimulated with anti-CD3 mAb (5 μg/ml). After 48 h, the cells were pulsed with [³H]TdR and were harvested at 72 h. In some experiments, 100 U/ml of rIL-2 was added during the restimulation.

PBL (2 × 10⁵/well) were cultured with PHA-P (5 μg/ml) in the presence of peptide (40 μM), cyclosporin A (100 nM), rapamycin (100 nM), or medium. At 48 h, the cells were washed in PBS and resuspended in 250 μl of a solution containing 4 mM sodium citrate, 0.05% Nonidet P-40, 0.45 mg/ml RNase, and 50 μg/ml propidium iodide. The cells were incubated on ice for 10 min, at which time 25 μl of 1.5 M sodium chloride was added. The cells were analyzed on a Becton Dickinson (San Jose, CA) FACScan at the Stanford FACS facility. Propidium iodide fluorescence was plotted on a linear scale vs orthogonal scatter.

PBL (2 × 10⁵ cells) were stained as previously described (29) with the indicated FITC-conjugated Abs (1 μg/10⁶ cells). Cells were also stained with propidium iodide immediately before analysis to exclude dead cells.

Assays were performed as described (30) with the following modifications. Preactivated T cells (5 × 10⁶) were stimulated with 100 U/ml rIL-2. At 24 h, the cells were washed in PBS and lysed in lysis buffer (0.5% Triton X-100, 1 mM DTT, 10 mM β-glycerophosphate, 20 mM sodium fluoride, 0.2 mM sodium orthovanadate, 10 μg/ml leupeptin, and 5 μg/ml aprotinin). Lysates were incubated on ice for 10 min, then spun down for 3 min in a microcentrifuge. The amount of protein in each supernatant was measured using the Bio-Rad protein assay (Bio-Rad, Hercules, CA). Anti-cyclin E Ab (2 μg) was added to supernatants containing equivalent amounts of protein, and the samples were rocked at 4°C for 1 h. Protein A/G beads (15 μl) (Santa Cruz Biotechnology) were added, and the samples were rocked for 45 min at 4°C. The beads were washed twice in lysis buffer and twice in kinase buffer (20 mM HEPES, pH 7, 10 mM β-glycerophosphate, 5 mM magnesium chloride, 10 μg/ml leupeptin, and 1 mM DTT). Kinase activity was assayed by resuspending the beads in 40 μl of kinase buffer supplemented with 8 μCi [γ-³²P]ATP, 4 μM ATP, and 20.6 mg/ml histone H1 type III-S. Samples were incubated at 37°C for 15 min, and the reaction was stopped by adding 40 μl of ice-cold 20 mM EDTA, pH 8. Triplicate aliquots were spotted onto phosphocellulose filters (Life Technologies, Grand Island, NY). The filters were soaked briefly in 1% H₃PO₄/10 mM Na₄P₂O₇ and washed three times for 15 min in 1% H₃PO₄. ³²P incorporation was determined by liquid scintillation counting.

Preactivated cells (1–2 × 10⁷) were treated with peptide and 100 U/ml IL-2 for 24 h, and immunoprecipitations were performed as described (28). For Western blots, 2.5 × 10⁶ preactivated T cells were incubated with peptide (40 μM) and 100 U/ml rIL-2 and prepared as described (31). Proteins were separated on either 7.5% or 12% SDS-PAGE gels. The proteins were transferred electrophoretically to supported nitrocellulose (Amersham, Arlington Heights, IL), and the membrane was blocked with a solution of 5% nonfat dry milk in 5 mM Tris-buffered saline, pH 7.5, with 0.05% Tween-20. Blots were probed with the indicated Abs at 1:1000. This step was followed by a secondary Ab conjugated to horseradish peroxidase at 1:3000, and proteins were visualized by chemiluminescence.

HLA class II molecules are heterodimers composed of an α- and a β-chain. While the β-chain is highly polymorphic, the α-chain is relatively conserved (32). From the hypothetical model of class II, later confirmed by the crystal structure of HLA-DR (33, 34), we selected the α₁ α-helix region for study based on our previous studies with MHC class I (20, 21, 22). Peptides corresponding to the α-helix of DRA, DQA, and DPA (32) were synthesized.

The peptide corresponding to residues 56–80 of DQA03011 blocked proliferation of PBL to anti-CD3 mAb (data not shown). Overlapping 15-amino acid peptides were synthesized from this sequence and were tested for function. A peptide corresponding to residues 65–79 (Table I), designated DQ 65–79, recapitulated the effects of the full-length peptide. The DQ 65–79 peptide inhibited [³H]thymidine incorporation by PBL or purified T cells stimulated with anti-CD3 mAb (Fig. 1, A and B). Furthermore, the inhibitory effect of the peptide was not reversed by the addition of anti-CD28 mAb to purified T cells stimulated by anti-CD3 mAb (Fig. 1,C). Inhibition was not limited to stimulation via anti-CD3 mAb, as PHA-P-stimulated (Fig. 2) or alloantigen-stimulated cells (data not shown) were also blocked by peptide; however, DQ 65–79 did not block PMA- and ionomycin-stimulated proliferation (Fig. 1,D). Inhibition of proliferation by peptide was observed with cells obtained from 35 donors, indicating the effects were allele unrestricted. The inhibition of proliferation was not due to cell death because no significant increase in staining by either trypan blue or propidium iodide was observed over the course of the proliferation assay (data not shown). A second peptide, in which aspartic acid was substituted for asparagine at amino acid 72, did not inhibit proliferation like the DQ 65–79 peptide (Table I and Figs. 1 and 2). The DQ 65–79 peptide and its derivatives were used to characterize the immunoregulatory effects and mechanism of action.

To determine which amino acids were necessary for peptide inhibition, peptides were synthesized in which each amino acid of the sequence was substituted individually with serine (Fig. 2). Peptides with serine substitutions at residues 65 and 67 maintained activity, while serine substitutions at 70 or 74 increased activity. All remaining substituted peptides lost functional activity.

We next examined whether peptide-induced unresponsiveness was transitory or long-lived. PBL treated with peptide and/or anti-CD3 mAb were washed, rested 7 days, and restimulated with anti-CD3 mAb alone. Cells initially treated with both the DQ 65–79 peptide and anti-CD3 mAb were unresponsive to restimulation when compared with cells initially treated with anti-CD3 mAb with either DMSO or the DQ 65–79 72D peptide (Fig. 3,C). Addition of exogenous IL-2 during the restimulation did not reverse the unresponsiveness of the DQ 65–79 peptide-treated cells (Fig. 3,D). Interestingly, cells cultured with the DQ 65–79 peptide alone were not unresponsive upon restimulation (Fig. 3 B). These cells, after washing and resting, were indistinguishable from control cells in their response to anti-CD3 mAb. Thus, simultaneous signals provided by peptide and anti-CD3 mAb are required to induce long term unresponsiveness.

We examined the effect of the DQ 65–79 peptide on two key events involved in T cell activation: early gene expression and cell surface receptor expression. In some experiments, the effects of the DQ 65–79 peptide were compared with the immunosuppressive drugs cyclosporin A or rapamycin, agents that have divergent mechanisms of action (35).

The hallmark of cyclosporin A activity is its block of IL-2 gene expression, a critical early event in T cell proliferation (35). Cyclosporin A treatment also decreases expression of other early genes such as IFN-γ and IL-2Rα (36). In contrast, rapamycin does not affect the expression of these early genes, but instead blocks signaling events downstream of the IL-2R (29). No decrease in IL-2, IL-2Rα, or IFN-γ gene expression was detected by Northern blot analysis of PHA-P-activated PBL treated with the DQ 65–79 peptide as compared with DMSO-treated cells (data not shown). In agreement with these data, both the DQ 65–79 peptide and rapamycin blocked anti-CD3- or rIL-2-induced proliferation of preactivated PBL, while cyclosporin A blocked only anti-CD3-induced proliferation (Figs. 4, A and B).

Because the DQ 65–79 peptide inhibited IL-2-mediated proliferation, cell surface expression of proximal signaling molecules was examined. The levels of CD2, CD4, CD8, MHC class I, and MHC class II were identical in anti-CD3 mAb-stimulated PBL in the presence or absence of the DQ 65–79 peptide (data not shown). Additionally, there was no difference between the expression of the activation markers IL-2Rα, CD69, and CD45RO in peptide-treated or untreated cells (data not shown). Because the peptide inhibited IL-2-mediated proliferation (Fig. 4 B), the expression of the IL-2Rβ and IL-2Rγ chains also was examined; however, DQ 65–79 had no effect on either of these molecules (data not shown). These findings indicate that, like rapamycin, the functional effects of the DQ 65–79 peptide are most likely downstream of the IL-2R.

To further examine the effect of the DQ 65–79 peptide on proliferation, cell cycle and kinetics studies were performed. Analysis of DNA content showed that the peptide, along with cyclosporin A and rapamycin, prevented DNA replication by blocking the G₁ to S transition (Table II). We examined the kinetics of peptide activity by adding it at various time points after activation. [³H]Thymidine incorporation of PBL stimulated with anti-CD3 mAb was inhibited when the DQ 65–79 peptide was added up to 24 h after stimulation (Fig. 5). The kinetics data concur with the block at G₁/S of the cell cycle. The ability of the DQ 65–79 peptide to block T cell proliferation at a relatively late time was similar to that of rapamycin, but contrasted with cyclosporin A, which lost its effect by 24 h.

Because the DQ 65–79 peptide functioned until relatively late after activation and blocked the G₁/S cell cycle transition, we focused on events that occur during this time period. The cyclin-dependent kinases (cdks) are key regulators of cell cycle progression. Activation of the cdks is a complex process, requiring both the recruitment of positive and the release of negative regulators. These kinases are regulated positively by association with cyclin proteins and by threonine phosphorylation, and they are regulated negatively by tyrosine phosphorylation and the association with either the Ink, Cip, or Kip families of inhibitory proteins. The negative signals from these inhibitory proteins overwhelm positive signals. Thus, an inhibitory protein must be removed from the cyclin-cdk complex before the kinase can become active (37, 38).

Cdk2 is critical for the transition from G₁ into S phase of the cell cycle. It must associate with cyclin E and be phosphorylated on threonine 160 for activation. Cdk2 can be regulated negatively by phosphorylation of tyrosine 15 and by the association of either p21^Cip1 or p27^Kip1 (28). Because the DQ 65–79 peptide inhibited cell cycle progression at G₁/S, we examined its effects on cdk2 kinase activity. T cells stimulated with IL-2 in the presence or absence of the DQ 65–79 peptide were assayed for cdk2 kinase activity. Phosphorylation of the cdk2 kinase substrate histone H1 was decreased with DQ 65–79 peptide treatment (Fig. 6,A). This effect was not due to a change in the expression of cdk2 or cyclin E because the amount of these proteins was essentially equivalent between cells treated with the DQ 65–79 peptide and controls over a 2-day time course (Fig. 6,C). Nor was inhibition due to a detectable change in phosphorylation, as the phosphorylation pattern was equivalent between treatment conditions, as shown by the doublet bands (Fig. 6,B). Cdk2 in DQ 65–79 peptide-treated cells was able to form a complex with cyclin E, as shown by the coimmunoprecipitation of cdk2 with Abs specific for cyclin E (Fig. 6,B). There was no significant change in complex formation in cells treated with the DQ 65–79 peptide as compared with controls (Fig. 6, B and C). However, the steady state level of the inhibitory protein p27 was higher, while the level of p21 was lower, in the cells treated with the DQ 65–79 peptide as compared with cells treated with the control 72D peptide (Fig. 6,C). Rapamycin retards G₁/S cell cycle progression by the inhibition of the cdk2 kinase via a block in p27 degradation, and also blocks the up-regulation of p21 (Fig. 6 C) (30, 39, 40). Thus, DQ 65–79 affected events downstream of the IL-2R in a manner similar to rapamycin, resulting in the inhibition of cell cycle progression and subsequent proliferation.

Because of the multiple functional similarities between DQ 65–79 and rapamycin, we asked whether the DQ 65–79 peptide bound to the immunophilin FK-506-binding protein (FKBP). Rapamycin and FK-506 bind to FKBP, and this complex formation is critical for their function (41). Rapamycin and FK-506 have different effects and are mutually antagonistic. Excess FK-506 prevents rapamycin-mediated inhibition of IL-2-mediated proliferation, and likewise, an excess of rapamycin prevents FK-506-mediated inhibition of PMA- and ionomycin-stimulated proliferation (41). Since the DQ 65–79 peptide blocked IL-2-mediated proliferation, but not PMA- and ionomycin-mediated proliferation, we conducted similar experiments in which the DQ 65–79 peptide was substituted for rapamycin. The DQ 65–79 peptide had no effect on FK-506-mediated inhibition of PBL proliferation stimulated with PMA and ionomycin in a range of 10 to 10,000 molar excess (Fig. 7,A). Likewise, no concentration of FK-506 tested blocked the effect of the DQ 65–79 peptide on IL-2-mediated proliferation, although FK-506 did reverse the inhibition of proliferation by rapamycin (Fig. 7 B). Therefore, the DQ 65–79 peptide does not compete for binding to the FK-506 binding site on FKBP, and probably does not bind to FKBP at all. Thus, while DQ 65–79 is functionally similar to rapamycin, it acts via a distinct mechanism.

A synthetic peptide corresponding to residues 65–79 of the α-chain of DQA03011 inhibits the proliferation of PBL and purified T cells in an allele-nonspecific manner. This peptide induces long-term unresponsiveness that is not reversible with IL-2 and is dependent upon concomitant activation through the TCR or IL-2R. The DQ 65–79 peptide does not inhibit early mRNA accumulation or cell surface receptor expression, but instead blocks the G₁ to S transition of the cell cycle. Cells treated with the DQ 65–79 peptide exhibit repressed cdk2 kinase activity and prolonged expression of the p27 inhibitor.

We and others have demonstrated that either intact MHC molecules or synthetic peptides corresponding to polymorphic regions of MHC molecules inhibit target cell lysis of CTL specific for those MHC Ags (17, 21). We later demonstrated that synthetic peptides corresponding to residues 75–84 of the class I molecule HLA-B2702 block lysis of target cells by CTL in an allele-nonrestricted manner (22). Peptides derived from this region of class I, either alone or in combination with a subtherapeutic dose of cyclosporin A, prevent rejection of heterotopic heart allografts in rodents (23, 24, 25, 42). Lymphocytes isolated from tolerant animals differentiate into donor-specific CTL in vitro, consistent with the notion that the cells are functionally inactivated in vivo (42).

The inhibitory effect of the DQ 65–79 peptide is sequence specific, but not MHC restricted. Peptides synthesized with single amino acid substitutions at individual residues 66, 68, 69, 71–73, and 75–79 lost function. These residues are located throughout the length of the peptide, indicating that the overall sequence of the peptide is important for its function. This contrasts with our previous findings with the synthetic peptide corresponding to residues 75–84 of the class I HLA-B2702 sequence. This peptide blocks lysis of target cells by CTL, and this effect is completely dependent on the presence of an isoleucine at position 80 (22). Substitution of the other residues had little, if any, effect on the function of the class I peptide.

A potential mechanism for DQ 65–79 peptide function is that it interferes with Ag presentation by binding to the peptide groove of MHC molecules. Peptides derived from this conserved region of HLA DQ do bind to HLA class II molecules (43, 44). However, the DQ 65–79 peptide blocks proliferation of all T cells tested, indicating an allele-unrestricted effect. Furthermore, the DQ 65–79 peptide inhibits proliferation of purified T cells stimulated by anti-CD3 mAb or PHA-P, both of which bypass APC. These findings suggest that the peptide-mediated inhibition is independent of Ag presentation.

It has been postulated that soluble whole or fragmented MHC molecules modulate immune responses through interaction with the TCR, potentially by steric inhibition of the TCR interaction with membrane-bound MHC/Ag complexes (3, 5). It is possible that the DQ 65–79 peptide bypasses allele specificity by binding to conserved regions of the TCR. It is also possible that the DQ 65–79 peptide could interfere with costimulation, causing cells to become anergic (45). This type of unresponsiveness can often be overcome by the addition of exogenous IL-2. However, neither exogenous IL-2 nor anti-CD28 mAb reversed the inhibitory effect of the peptide. Therefore, the DQ 65–79 peptide functions through a mechanism not involving TCR/MHC interactions or the induction of classical anergy.

The DQ 65–79 peptide described in this work and the class I-derived B2702 75–84 peptide identified previously are derived from analogous regions of MHC class II and MHC class I, respectively. This region of native MHC class I is important in binding the p58 and p70 KIR of NK cells (9). There is evidence to suggest that functionally similar inhibitory receptors for MHC class II exist, although they have not yet been identified (46, 47). Although the B2702.75–84 peptide blocks NK-mediated killing, peptide function is not dependent on any defined KIR (J. E. Goldberg and C. Clayberger, unpublished observations). Various KIRs are expressed on subsets of NK cells and a small percentage of T cells (9). Unlike KIR, the allele-nonspecific effects of the DQ 65–79 peptide suggest that a potential inhibitory ligand would have to be broadly expressed.

The expression of cytokines and signaling receptors is critical for the initiation of proliferation. Unlike cyclosporin A (36), no inhibition of IL-2 or IFN-γ expression was detected in DQ 65–79 peptide-treated T cells stimulated with PHA-P. Expression of all cell surface receptors tested, including all three chains of the IL-2R, was normal in peptide-treated cells stimulated with anti-CD3 mAb or PHA-P. Instead, the peptide interfered with signals downstream of the IL-2R. Although cdk2 activity was reduced, we could not detect any differences in the phosphorylation pattern of cdk2 or its association with cyclin E in lysates from cells activated in the presence of the DQ 65–79 peptide. Instead, the peptide prolonged the presence of the inhibitor molecule p27 (39). Although it is not known how rapamycin or the DQ 65–79 peptide modulates p27 expression, it may involve the ubiquitin degradation pathway (48).

The DQ 65–79 peptide and rapamycin share functional and mechanistic activities, but they differ in several important ways. First, rapamycin, in a complex with FKBP, blocks Saccharomyces cerevisiae yeast cell division by binding to and inhibiting the activity of TOR (target of rapamycin), a phosphoinositol kinase family member required for growth (49). In contrast, the DQ 65–79 peptide did not affect division of S. cerevisiae yeast cells at any concentration tested (S.-C. Lyu and C. Clayberger, unpublished observations). Furthermore, our experiments show that DQ 65–79 peptide and rapamycin do not bind to the same site on FKBP, and it is likely that the DQ 65–79 peptide does not bind FKBP at all.

Rapamycin has pleiotropic effects on cell proliferation downstream of the IL-2R. As a complex with FKBP, rapamycin blocks the activity of mTOR, the mammalian homologue of TOR, which is critical for cap-dependent translation. and for the activation of p70 S6 kinase (50, 51). The p70 S6 kinase has also been implicated in the regulation of CREB/ATF transcription factors, which are involved in the expression of proliferating cell nuclear Ag, a protein that is required for entry into the S phase of the cell cycle (52). We are currently investigating whether the DQ 65–79 peptide affects these and other signaling events downstream of the IL-2R.

Do analogous peptides corresponding to regions of MHC molecules exist in vivo, and do they mediate similar functions? Although there are no definitive answers to these questions, several lines of evidence suggest that soluble peptides of MHC molecules are present and can mediate immunoregulatory functions in vivo. It has been recognized for some time that class I MHC molecules exist in both membrane-bound and soluble forms, the latter resulting from alternative splicing (53). Soluble class II molecules also exist, and activated lymphocytes produce large quantities of soluble MHC molecules (54, 55, 56). Soluble MHC molecules inhibit immune responses in both allele-restricted and unrestricted ways, and they are likely to be responsible for part or all of the nonspecific immunosuppression observed following blood transfusion in a number of clinical settings (57, 58).

MHC-derived peptides may also mediate some of the phenomena attributed to suppressor cells. There are many studies that document a potent suppressor effect of T lymphocytes or their products, and many of these effects map to the MHC. However, when a thorough analysis of the MHC was conducted, no suppressor genes were identified (59). Our findings suggest that peptides, derived from intact MHC molecules by either pre- or post-translational events, can mediate immunoregulatory effects. The generation of peptides from MHC molecules may occur only in restricted sites or in response to particular stimuli, and secretion and/or processing of these molecules by activated lymphocytes may be important in turning off immune responses. Characterization of such peptides in vitro should prove useful in the design of novel therapies to dampen the in vivo immune response in transplant rejection and autoimmune disease.

We thank Elizabeth Mellins and Jodi Goldberg for helpful discussions and critical reading of the manuscript.