Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

HLA class II molecules are polymorphic proteins expressed on a subset of white blood cells that includes macrophages, B cells, dendritic cells and activated T cells (Barber and Parham, 1993). Their normal function is to bind peptides generated by intracellular (for review, see Brodsky and Guagliardi, 1991) and possibly by extracellular (Adorini et al., 1991) degradation of foreign antigens. The subsequent recognition of complexes of class II molecules and peptides by TCR on CD4-positive T cells is an essential step in the generation of both antibody and cellular immune responses. The diverse nature of the HLA class II molecules is a result of both the different class II isotypes (DP, DQ and DR) and the many different allelic forms of these isotypes that exist (Marsh and Bodmer, 1992). This diversity provides the species with the capacity to respond to a vast variety of antigens and restricts the ability of an individual to respond to a given antigen (Abbas et al., 1991).

Certain class II polymorphisms are associated with increased relative risks of developing autoimmune diseases. For example, the DR alleles DR4 Dw4 (DRB1*0401), DR4 Dw14 and DR1 are associated with susceptibility to RA and with the severity of this disease (for review, see Winchester, 1994). These three DR molecules share a highly conserved sequence in the third hypervariable region of the -chain, suggesting a molecular basis for the disease pathogenesis. Analyses of the repertoires of peptides that bind to DR4 and DR1 molecules have identified both promiscuous and specific binding peptides (Chicz et al., 1993; Hammer et al., 1993; O'Sullivan et al., 1990; Sette et al., 1993). These studies suggest that it should be possible to exploit the differences between the allelic forms to develop inhibitors that selectively block specific DR molecules.

Inhibition of the interactions between any of the components of the trimolecular complex of HLA class II, peptide and TCR, using monoclonal antibodies, is sufficient to block CD4-positive T cell-mediated immune responses in multiple experimental models of autoimmune disease (for review, see Smilek et al., 1990). The ability of antagonist peptides to compete for antigen presentation has been well characterized in vitro in humans and both in vitro and in vivo in animals (for reviews, see Smilek et al., 1990; Adorini and Nagy, 1990). In addition, competitor peptides have been used to treat disease in animal models of autoimmunity, including experimental autoimmune encephalitis, adjuvant-induced arthritis (Wauben et al., 1994), autoimmune carditis (Smith and Allen, 1991) and spontaneous diabetes in nonobese diabetic mice (Hurtenbach et al., 1993). Thus, even peptides can be utilized as small molecule antagonists for the treatment of class II-associated autoimmune diseases. The association of DRB1*0401 with RA makes it an attractive molecular target for disease intervention. This report describes the characterization of a peptide-based inhibitor of peptide binding to DRB1*0401 and its potential as a therapeutic agent.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Synthesis of inhibitors. The synthesis of the protease-resistant peptidomimetic SC-67655 is described separately (Hanson et al., 1996). HA307-319, SC-64939, SC-69050 and SC-69054 were synthesized, with fluorenylmethoxycarbonyl chemistry, on an Applied Biosystems model 433A synthesizer and purified on a C18 column by reverse-phase high-pressure liquid chromatography. The compositions of the inhibitors were verified by mass spectroscopy. The structures of the peptidomimetics are shown in table 1.

Cell lines. The DR-homozygous B lymphoblastoid cell lines GM03104 (DR1), GM03161 (DR2), GM03098A (DR3), GM03164 (DRB1*0401, formerly DR4 Dw4), GM03153B (DR6), GM03163 (DR7) and GM08598 (DR8) were obtained from the National Institutes of General Medical Sciences Human Genetic Mutant Cell Repository (Camden, NJ). The DR5 family-homozygous, B lymphoblastoid Swei cell line (DRB1*1101) was originally obtained from Dr. John Hansen, Fred Hutchinson Cancer Research Center (Seattle, WA). The MT (DR4 Dw14) and YAR (DR4 Dw10) cell lines were gifts from Dr. Gerald Nepom (Virginia Mason Research Center, Seattle, WA). The B lymphoblastoid cell lines were maintained in RPMI 1640 medium (Mediatech, Washington, DC) supplemented with 15% fetal calf serum (Hyclone Laboratories, Logan, UT) and 2 mM glutamine (Mediatech). The mouse macrophage J774A.1 cell line was purchased from the American Type Culture Collection (Rockville, MD) and was maintained in Dulbecco's modified Eagle's medium (Mediatech) with 10% fetal calf serum (Hyclone).

Competition ELISA. Inhibitors (0.01-100 µM) were analyzed for their ability to block the binding of 30 nM B-HA peptide to purified DR molecules as previously described (Kirschmann et al., 1995).

Cellular peptide-binding assay. The binding of B-HA peptide to cells was previously described (Woulfe et al., 1995). The ability of inhibitors to block the binding of 10 µM B-HA to cells was determined by incubating cells with 1 to 300 µM competitor.

T cell proliferation assays. The ability of inhibitors to block antigen-specific T cell proliferation was evaluated by using a modification of the assay previously described (Fu et al., 1995). Peripheral blood lymphocytes from HLA DR-typed donors were isolated, using Ficoll-Hypaque (Pharmacia, Piscataway, NJ), from freshly drawn, heparinized, venous blood. The cells were irradiated (3000 rad) by using a ¹³⁷Cs source and were used as APC. The APC were incubated under serum-free conditions in Excell 300 medium (JRH Biosciences, Lenexa, KS), with varying concentrations (0.01-500 µM) of inhibitors, for 2 to 4 hr at 37°C in 5% CO₂. Suboptimal concentrations of HA307-319 peptide (3-10 nM) or HA protein (2-7 nM) were added to the wells for 2 hr after 2 hr of incubation with the inhibitors. The APC were washed with serum-free medium three times, to remove unbound inhibitors and peptides. Clonal human DRB1*0401-restricted, HA307-319-specific T cells (Fu et al., 1995) or clonal human DRB1*1101-restricted, HA307-319-specific T cells (McKinney et al., 1994) were added to the wells and incubated for 3 days at 37°C in 5% CO₂. Phytohemagglutinin (5 µg/ml) was added to some of the wells in the presence or absence of inhibitor peptides to determine cytotoxicity. Proliferation was determined as previously described (Fu et al., 1995).

Cell association assays. The ability of inhibitors to associate with human RBC or the mouse macrophage cell line (J774A.1) was determined with a modification of the previously described assay (Ohsako et al., 1993). RBC were obtained from freshly drawn blood in heparinized tubes. The RBC were washed with 0.15 M saline and resuspended as a 62.5% (v/v) solution in saline. Inhibitors (1 mM in PBS) were added at a final concentration of 200 µM to generate a 50% (v/v) solution of RBC or added to a confluent (~1.7 × 10⁷ cells) 150-cm² flask of J774A.1 cells and incubated for 1 hr at 37°C. The cells were briefly pelleted in an Eppendorf microfuge (3000 × g for 30 sec), and the extracellular supernatants were collected and saved for further analysis. The RBC and macrophage pellets were rapidly washed three times with saline or PBS, respectively. The cells were hypotonically lysed in an equal volume of water, and octylglucoside was added to a final concentration of 1%, to generate cell extracts. Cell membranes were removed from the extracts by centrifugation at 20,000 × g for 30 min for RBC or 130,000 × g (for 30 min) for J774A.1 cells. The cell extracts and the extracellular supernatants were centrifuged through molecular weight 10,000 filters (Amicon, Beverly, MA) and evaluated for the presence of inhibitory activity with the competition ELISA.

Statistical analyses. Data were analyzed for significant differences by the one-tailed Student's t test.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Design of small peptidomimetics. The compositions of HA307-319, the peptide-based inhibitor SC-67655 and the control molecule SC-64939 are shown in table 1. SC-67655 and SC-64939 can be considered as a pentamer and a heptamer, respectively, based upon the number of nitrogens capable of forming peptide bonds within each compound. SC-64939 is an acetylated, aminated and truncated variant of HA309E, a previously described (Woulfe et al., 1995) mutant HA307-319 peptide that does not bind to DRB1*0401. SC-69050 and SC-69054 contain identical hydrophobic membrane-translocating and spacer sequences (Lin et al., 1995), which are conjugated to the respective amino-termini of the inhibitors.

Binding of peptidomimetics to DRB1*0401. The abilities of the peptidomimetics to compete with B-HA peptide for binding to purified DRB1*0401 were determined by ELISA, and a comparison of the activities is shown in figure 1. SC-67655 and HA307-319 inhibited the binding of B-HA in a dose-dependent manner, whereas SC-64939 had no effect. The average concentrations of HA307-319, SC-67655 and SC-64939 required to yield 50% inhibition (IC₅₀) in five experiments were 62, 50 and >200,000 nM, respectively. Therefore, the small peptide-based inhibitor bound to DRB1*0401 as well as or better than the 13-residue HA307-319 peptide. SC-69050 also inhibited biotinylated peptide binding, but to a lesser extent than did SC-67655. The average concentration of SC-69050 needed to inhibit binding by 50% was 26.7 µM (n = 2). In contrast, SC-69054 competed poorly with B-HA for binding to DRB1*0401 (IC₅₀ > 100 µM, n = 2).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   A peptidomimetic has high affinity for DRB1*0401. The ability of inhibitors to compete for B-HA binding to purified DRB1*0401 molecules was determined by competition ELISA. Inhibitors were evaluated in at least two experiments. Results of representative experiment are shown.

Specificity of inhibitors for DRB1*0401. The specificity of binding of the peptide-based inhibitors was evaluated using a panel of homozygous typing cells, each of which expressed a single DRB1 gene product (fig. 2). As previously reported, HA307-319 dose-dependently inhibited the binding of B-HA on each of the cell lines tested. In contrast, SC-64939 did not compete in a dose-dependent manner for biotinylated peptide binding to any of the cell lines. SC-67655 inhibited binding of B-HA to the cell line that expressed DRB1*0401 but did not block binding of B-HA to any of the other cell lines tested. These results demonstrate that, in contrast to the HA307-319 peptide, SC-67655 is selective in its ability to bind different DR molecules. The specificity of SC-67655 was further analyzed in an ELISA, using purified DR4 Dw14, DR4 Dw10 and DR1 proteins. SC-67655 did not compete for HA307-319 binding to these alleles (data not shown).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   The peptidomimetic is specific for DRB1*0401 (DR4). Inhibition of B-HA binding to B lymphoblastoid cell lines was evaluated using the cellular peptide binding assay. Results shown are averages of two experiments.

Inhibition of a functional T cell response to peptide antigen. Human peripheral blood lymphocytes that expressed DRB1*0401 were incubated with HA307-319 peptide and increasing concentrations of SC-67655 or SC-64939 and evaluated for their ability to block DRB1*0401-restricted, antigen-specific, T cell proliferation (fig. 3A). SC-67655 inhibited T cell proliferation with an IC₅₀ of <7.5 µM (n = 2) but did not inhibit T cell proliferation in response to phytohemagglutinin (data not shown), indicating that the compound was not cytotoxic. No inhibition of T cell proliferation was observed when SC-64939 was added to the cultures (IC₅₀ > 200 µM). The ability of SC-67655 to block proliferation was dependent upon the presence of DRB1*0401. Incubation of SC-67655 with human peripheral blood lymphocytes derived from DRB1*1101 donors pulsed with HA 307-319 peptide had no effect on the ability of a HA307-319-specific, DRB1*1101-restricted T cell clone to proliferate (IC₅₀ > 200 µM, n = 2) (fig. 3B).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   The peptidomimetic specifically inhibits DRB1*0401-restricted T cell proliferation in response to peptide but not protein antigen. Effects of inhibitors on the ability of a T cell clone to proliferate in response to APC pulsed with HA307-319 peptide and DRB1*0401-positive APC (A), HA307-319 peptide and DRB1*1101-positive APC (B) and HA protein and DRB1*0401-positive APC (C) were determined. Inhibitors were evaluated in at least two experiments. Results of representative experiments are shown.

Blockade of functional T cell response to protein antigen. Incubation of human peripheral blood lymphocytes from DRB1*0401 donors with intact HA protein antigen also stimulates the proliferation of peptide-specific, DRB1*0401-restricted T cells. Human peripheral blood lymphocytes that expressed DRB1*0401 were incubated with HA protein and increasing concentrations of SC-67655 or SC-64939 and evaluated for their ability to trigger T cell proliferation (fig. 3C). No inhibition of T cell proliferation was observed under any of the conditions tested (IC₅₀ > 200 µM, n = 2).

Association of inhibitor with cells. The ability of SC-67655 to block peptide-induced but not protein-induced proliferation suggested the possibility that SC-67655 did not enter APC. To evaluate the permeability of the inhibitor under conditions where little or no proteolysis was expected, RBC were used as a model system instead of APC. The ability of SC-67655 to penetrate cells was determined by incubating the inhibitor with RBC, and the intracellular lysates and extracellular supernatants were analyzed for the presence of inhibitory activity using the ELISA (fig. 4A). Lysates generated from RBC incubated with SC-67655 contained only 0.38% of the inhibitory activity of SC-67655 found in the extracellular supernatant. The low level of activity detected in RBC lysates incubated with SC-67655 might be due to either a low level of uptake or nonspecific association of the compound with the cells. To distinguish between these possibilities, SC-67655 and sucrose were each loaded into RBC ghosts, and the release of inhibitory activity into the extracellular supernatants was measured as a function of time (fig. 4B). No inhibitory activities were detected in the extracellular supernatants. To demonstrate that the ghosts were indeed loaded, the ghosts were lysed at the final time point and evaluated for inhibitory activities (fig. 4B). As expected, lysates prepared from ghosts loaded with SC-67655 completely abrogated the binding of B-HA to DRB1*0401, whereas lysates from ghosts containing sucrose had no inhibitory activity. These results demonstrated that SC-67655 does not penetrate RBC cell membranes. Consistent with this finding, no inhibitory activity was detected in lysates generated from a murine macrophage cell line (J774A.1) incubated with SC-67655 (data not shown).

View larger version (26K): [in this window] [in a new window]   Fig. 4.   The peptidomimetic does not accumulate within RBC. The ability of inhibitors to permeate RBC was determined by competition ELISA. The inhibitory activities present within intracellular and extracellular fractions generated from RBC incubated in the presence or absence of inhibitor were determined in two experiments. A, representative experiment; B, release of inhibitor from RBC-loaded ghosts over time.

Signal peptide-inhibitor complex inhibition of functional T cell responses to protein antigens. The inability of SC-67655 to enter cells supports the hypothesis that molecules that function by blocking the peptide binding groove must penetrate APC to block the presentation of protein antigens. To test this hypothesis, the ability of the signal-peptide conjugates to block antigen-specific DRB1*0401-restricted T cell proliferation induced by a protein antigen was determined (fig. 5A). SC-69050 conjugate consistently inhibited T cell proliferation in response to protein antigen, with an average IC₅₀ of 80 µM (n = 9). SC-69054 also inhibited proliferation (IC₅₀ = 150 µM) in four experiments, but to a lesser extent than SC-69050, and did not cause inhibition in five experiments (IC₅₀ > 200 µM, n = 5). The inhibition observed with SC-69050 (n = 9) was statistically different from the inhibition observed with SC-69054 (n = 4) (P < .05). Furthermore, SC-69050 did not specifically inhibit the proliferation of human peripheral blood lymphocytes derived from a DRB1*1101 donor that had been pulsed with HA protein (fig. 5B). This finding confirms that a portion of the inhibition observed in the previous experiments was mediated via the DRB1*0401 molecule.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   A signal peptide-peptidomimetic conjugate specifically inhibits DRB1*0401-restricted T cell proliferation in response to a protein antigen. Effects of inhibitors on the ability of a T cell clone to proliferate in response to APC pulsed with HA protein and DRB1*0401-positive APC (A) and DRB1*1101-positive APC (B) were determined. Inhibitors were evaluated in nine experiments using APC from four DRB1*0401-positive donors and four experiments using APC from one DRB1*1101-positive donor. Results from a representative experiment are shown.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The associations of HLA DRB1*0401 with increased susceptibility to RA and with increased disease severity suggest a potential role for the molecule in pathogenesis of this disease. The efficacy of antirheumatic agents aimed at inhibiting antigen-specific T cell responses supports the involvement of DRB1*0401. For example, hydroxychloroquine has been shown in vitro to inhibit endosomal and lysosomal activities such as the processing and presentation of antigens (Fox and Kang, 1993; Ziegler and Unanue, 1982). The drug disodium aurothiomalate was recently reported to bind to certain DR-peptide complexes and prevent subsequent T cell recognition (Griem et al., 1995). Finally, cyclosporine A inhibits CD4-positive T cell activation, which can be stimulated via the TCR, by DR-peptide complexes (Wells and Tugwell, 1993). Our interest in characterizing an inhibitor that blocks the DRB1*0401 peptide-binding groove was based on the hypothesis that the interactions between the TCR on CD4-positive T cells and DRB1*0401-peptide complexes present on APC play a critical role in the pathogenesis of RA.

DRB1*0401 is an attractive molecular target for therapeutic intervention in RA, for a number of reasons. By preventing binding of antigenic peptide, DRB1*0401 groove blockers should block the interactions between APC and putative pathogenic T cells and thus should have disease-modifying activity. We have shown that a small molecule inhibitor can disrupt the trimolecular complex of DR, peptide and TCR. Such a small molecule could be chronically administered to prevent the initiation or exacerbation of disease. Moreover, the polymorphic differences between class II molecules can be exploited to develop highly specific inhibitors. Although considerably smaller than the majority of naturally processed peptides eluted from DRB1*0401 molecules (Kirschmann et al., 1995), SC-67655 contained sufficient information to impart selective binding to DRB1*0401. Because an individual may express three to eight class II molecules, development of an inhibitor highly specific for DRB1*0401 would leave the remainder of class II molecules free to bind peptides and thus would reduce the possibility of causing general immunosuppression.

Several factors must be considered in the development of an inhibitor that blocks the DRB1*0401 peptide groove as a therapeutic agent. First, an inhibitor that acts by blocking the groove must be potent, because a few hundred class II-peptide complexes have been shown to trigger antigen-specific T cell proliferation (Demotz et al., 1990; Harding and Unanue, 1990). Consistent with this observation, in our hands a 1000-fold molar excess of SC-67655 was needed to inhibit proliferation to peptide.

Second, if protein antigens are involved in the autoimmune response that results in RA, then either the inhibitor must be capable of penetrating APC and competing intracellularly with the agonist peptide for binding to DRB1*0401 or it must be capable of displacing bound peptide at the cell surface (Adorini et al., 1989; deKroon and McConnell, 1994; Frumento et al., 1994; Fridkis-Hareli et al., 1994). Surprisingly, although peptides that specifically bind to class II molecules have been used to inhibit T cell responses to protein antigens in vitro (Adorini et al., 1991) and in vivo (Adorini et al., 1991; Guery et al., 1992, 1993; Smith and Allen, 1991), it is not known whether the inhibiting peptides function via intracellular or extracellular mechanisms. SC-67655 was able to inhibit T cell proliferation in response to peptide antigen but not protein antigen. These results suggest that the peptidomimetic could compete for peptide binding to cell surface DRB1*0401 but could not compete for intracellular peptide binding.

Inhibitor molecules that could penetrate APC could function by blocking the intracellular peptide binding groove and thus prevent the presentation of processed protein antigens. We demonstrated that a DRB1*0401 groove-blocker conjugated to a signal peptide sequence was capable of inhibiting T cell proliferation in response to protein antigen. The data suggest that a portion of the inhibition observed in some donors may be attributable to the signal peptide itself. Inhibition by the signal peptide moiety may be mediated by interference with the uptake of protein, protein proteolysis or trafficking of class II molecules in some donors. However, significantly greater and consistent inhibition of protein-induced T cell proliferation was observed using the SC-69050 conjugate, suggesting that inhibition was mediated through blockade of the DRB1*0401 molecule. In addition, SC-69050 did not specifically inhibit antigen-specific, DRB1*1101-restricted proliferation. This result further supports a mechanism of action that involves DRB1*0401 blockade, and it excludes the possibility that the SC-67655 moiety merely enhanced the nonspecific inhibitory activity of the signal sequence through a non-DRB1*0401-mediated mechanism.

Our experiments demonstrate that inhibitors that prevent peptide binding to the groove of DRB1*0401 are capable of blocking functional immune responses. Although the signal peptide-groove blocker conjugates are unlikely to be of therapeutic value due to their large size, high cost and hydrophobicity, they were useful as tools to demonstrate the feasibility of intracellular blockade. The development of inhibitors that prevent peptide binding to class II molecules and that could be selectively delivered and concentrated in the compartments of peptide loading within APC may be suitable for further development as drug candidates. Such inhibitors with different specificities for class II molecules could be useful to treat a variety of autoimmune diseases.