Delayed rejection of MHC class II-disparate skin allografts in mice treated with farnesyltransferase inhibitors

Abstract

Farnesyltransferase inhibitors (FTIs), developed as anti-cancer drugs, have the potential to modulate immune responses without causing nonspecific immune suppression. We have investigated the possibility that FTIs, by affecting T cell cytokine secretion, can attenuate alloreactive immune responses. The effects of FTIs on murine alloreactive T cells were determined both in vitro, by measuring cytokine secretion or cell proliferation in mixed lymphocyte cultures, and in vivo, by performing skin allografts from H-2^bm12 mice to MHC class II-disparate B6 mice. We found that two different FTIs, ABT-100 and L-744,832, blocked secretion of IFN-γ, IL-2, IL-4, and TNF-α from naïve T cells in vitro. ABT-100 and L-744,832 blocked cytokine production from both CD4⁺ and CD8⁺ naïve T cells stimulated with CD3 and CD28 antibodies, but only if the cells were pretreated with the FTIs for 48 h. Proliferation of alloreactive T cells in mixed lymphocyte cultures was blocked by either FTI. We also found that the proliferation of enriched T cells stimulated with IL-2 was blocked by ABT-100 treatment. In mice with an MHC class II-disparate skin graft, rejection of primary allografts was significantly delayed by treatment with either ABT-100 or L-744,832. Secondary rejection in mice previously primed to the alloantigen was found to be unaffected by L-744,832 treatment. We have shown that FTIs can block T cell cytokine secretion and attenuate alloreactive immune responses. The ability of FTIs to block secretion of cytokines, including IFN-γ and IL-4, from naïve T cells provides a likely biological mechanism for the specific suppression of class II MHC-mediated allorejection. These results demonstrate that FTIs may have useful immunomodulatory activity due to their ability to delay priming to alloantigens.

1. Introduction

Farnesyltransferase inhibitors (FTIs) were developed to block the function of proteins involved in oncogenic signaling by preventing their proper localization. The enzyme farnesyltransferase attaches a prenyl group to at least 17 cellular proteins, including H-Ras, K-Ras, and N-Ras, and this post-translational modification is necessary for the proper subcellular targeting and function of these substrate proteins [1]. The anti-neoplastic activity of FTIs does not depend on the presence of activated Ras oncogenic proteins and the activity of these drugs likely depends on their combined inhibitory effects on several farnesylated proteins. We have recently shown that FTIs are effective in a mouse model of mature B cell lymphoma [2], a tumor that depends on antigen receptor signaling for neoplastic growth [3]. Early clinical trials have shown that FTIs may be effective in treating hematologic malignancies and Phase III trials are ongoing for these diseases [4].

This study utilizes two structurally unrelated FTIs, L-744,832 and ABT-100. L-744,832 is a peptidomimetic inhibitor that has been shown to affect the growth and survival of tumor cells both in vitro [5–7] and in mouse models [8–14]. ABT-100 has been shown to have anti-neoplastic activity in xenografts of human cancer cell lines and in animal models [15–17]. This activity may be due, in part, to anti-angiogenic effects of ABT-100 [16] related to its ability to block VEGF expression [17]. Because ABT-100 and L-744,832 are not structurally related to each other, they are unlikely to share similar nonspecific effects.

Although FTIs are not known to generally suppress immune responses, several studies have suggested that they may have specific immunomodulatory activity [18–24]. FTIs do not appear to block lymphocyte activation by antigen. For example, we have found that the proliferation of murine primary B cells stimulated with anti-IgM and anti-CD40 is unaffected by concentrations of L-744,832 that block the proliferation of a B cell lymphoma [2]. The ability of HMG-CoA reductase inhibitors to block cardiac rejection and allo-antibody production in humans [25,26], first suggested that prenylated proteins may be important for allorejection (reviewed in [27]). Several farnesylated proteins are known to be involved in lymphocyte signaling, such as Ras, RhoB, and Rheb [1]. Inhibition of Ras can block IL-2 production in activated lymphocytes [28], but a Ras-independent pathway for IL-2 production has also been described [29]. A prenylation inhibitor that blocks both farnesylation and geranylgeranylation, L-778,123, can prevent IL-2-mediated proliferation of lymphocytes, but has no effect on IL-2 secretion [23]. FTI treatment can block secretion of inflammatory cytokines from monocytes or macrophages [18,20] as well as Th1 and Th2 cytokines from murine T cell clones [21]. The inhibition of cytokine secretion from Th1 and Th2 cell clones does not depend on Ras/MAP kinase inhibition but, rather, is due to a post-transcriptional block of cytokine synthesis [21]. In a human cancer cell line, the SCH66336 FTI has been shown to block mTOR signaling by blocking the farnesylation of Rheb [30]. It appears that ABT-100 may act through a similar mechanism to block the activation of p70S6 kinase, thereby affecting the translation of certain mRNAs, including those for certain cytokines. If FTIs can block secretion of certain cytokines under specific conditions, then this could explain their ability to affect particular immune responses without general immune suppression.

FTIs could be clinically useful as immunomodulators by affecting cytokine secretion from alloreactive T cells. The ABT-100 FTI has been shown to delay rejection of rat cardiac allografts mismatched at both class I and class II MHC [24], demonstrating that FTIs can affect alloreactive immune responses in a non-stringent rejection model. Cytokines are responsible for regulating the tolerance or rejection of allografted organs at several steps of the alloreactive immune response [31]. Manipulation of the balance of cytokines may make it possible to promote the tolerance of allografts without nonspecific immune suppression.

2. Objective

The possibility that FTIs can modulate immune responses, such as allorejection, requires further investigation. The goals of this study were to determine if FTIs could block cytokine secretion from alloreactive T cells and delay an alloreactive immune response.

3. Materials and methods

3.1. Mice and reagents

C57BL/6 mice (B6) were purchased from Charles River Laboratories. BALB/c and B6-H2-Ab1^bm12 mice (bm12) were purchased from The Jackson Laboratory. All procedures involving animals were reviewed and approved by the Bucknell University IACUC. L-744,832 (BIOMOL) was dissolved at 2.5 mg/ml in DMSO and stored for no more than 1 mo at −80 °C prior to use in tissue culture experiments. For treatment of mice, L-744,832 was dissolved at 2.5 mg/ml in HBSS (Sigma), sterilized by filtration, and stored at −80 °C for no more than 1 wk prior to i.v. injection b.i.d. at 20 mg/kg. ABT-100, generously provided by Abbott Laboratories, was dissolved at 10 mg/ml in DMSO and stored at −20 °C prior to use in tissue culture experiments. For treatment of mice, ABT-100 was partially dissolved in ethanol at a concentration of 50 mg/ml, then diluted to 0.625 mg/ml in 0.5% hydroxypropylmethylcellulose (HPMC). The solution was adjusted to a pH of 3.5 and brought to a final concentration of 1% ethanol, 0.4% HPMC and 0.5 mg/ml ABT-100. The solution was stored at room temperature for no more than 2 weeks and provided to mice ad libitum as drinking water. This concentration of ABT-100 in drinking water was determined to administer an average of 100 mg/kg/day of the drug in separate experiments (data not shown).

3.2. ELISPOT assays

Splenocytes were isolated from B6 mice by macerating spleens, treating with RBC Lysis Buffer (Biolegend), and resuspending in complete RPMI media (RPMI 1640 (Mediatech) with 10% heat-inactivated FBS (PAA North America)). Cells at either 1 × 10⁶ or 2 × 10⁷ cells/ml were treated with ABT-100 or L-744,832 beginning 24 h prior to stimulation either with 5 μg/ml anti-CD3ε and 10 μg/ml anti-CD28 (Biolegend) or with allogeneic cells. BALB/c or bm12 stimulator cells were prepared by treating splenocytes for 20 min with 50 μg/ml mitomycin C (Sigma). Stimulator cells at 4 × 10⁶ cells/ml were either used fresh or cryopreserved prior to use. ELISPOT kits from eBioscience were used according to the manufacturer’s instructions. IFN-γ was measured for 24 h and IL-2, IL-4, and TNF-α were measured for 48 h. Data was blinded prior to manual spot counting.

3.3. Flow cytometry

Splenocytes were incubated in complete RPMI media at 1 × 10⁶ cells/ml in the presence or absence of ABT-100 or L-744,832. Cells were stimulated by the addition of 5 μg/ml anti-CD3ε and 10 μg/ml anti-CD28. For intracellular cytokine labeling, cells were treated with 3 μg/ml brefeldin A (eBiosciences) for the final 4 h of stimulation. Cells were then blocked with anti-CD16/32 and labeled with fluorescent anti-CD4 and anti-CD8. Cells were fixed and permeabilized with Cytofix/cytoperm buffer (BD Biosciences) and then labeled with fluorescent anti-TNF-α or anti-IFN-γ antibodies (Biolegend). Flow cytometry was performed on a BD FACScan using Cellquest software for data acquisition and analysis.

3.4. Proliferation assays

Total splenocytes or splenocytes enriched for T cells by negative selection paramagnetic beads (Invitrogen) from B6 mice were suspended at 2 × 10⁶ cells/ml and treated with ABT-100 beginning 24 h prior to stimulation. Cells were then stimulated either with 5 μg/ml anti-CD3ε and 10 μg/ml anti-CD28 or with 2 ng/ml (~10 U/ml) murine IL-2 (Peprotech). For mixed lymphocyte cultures (MLCs), B6 responder cells were stimulated with mitomycin C-treated BALB/c cells. For ³H-thymidine measurements, samples were stimulated for 72 h and then 5 μCi ³H-methyl thymidine (Sigma) was added and thymidine incorporation was measured 18 h later by filter binding. For BrdU measurements, after 36 h of stimulation, BrdU was added for 24 h and incorporation was quantified using an ELISA kit (Roche Diagnostics).

3.5. Skin allografts

Skin grafts were performed using full thickness tail skin (~0.5 × 1.0 cm) from bm12 mice or B6 mice onto the dorsal trunk of B6 recipients. Seven days after surgery, bandages were removed and grafts were scored daily for rejection, which was considered complete when > 90% of the donor tissue was necrotic.

4. Results

To determine whether FTI treatment could block the secretion of cytokines from naïve T cells when activated, we performed ELISPOT assays on murine splenocytes in culture. Splenocytes were harvested from B6 mice, treated with 0 to 1 μM ABT-100 or 1 to 10 μM L-744,832 and transferred to ELISPOT plates after 24 h. Antibodies to CD3 and CD28 were then used to stimulate T cells and cytokine production was measured (Fig. 1). Unstimulated cells did not secrete cytokines and CD3/CD28 stimulation promoted secretion of IFN-γ, IL-2, IL-4, and TNF-α. FTI treatment decreased the number of cytokine secreting cells for all four cytokines measured. The largest effect on cytokine secretion was seen with 1 μM ABT-100, where the treated cells were detected secreting cytokines at levels between 14% and 29% of the untreated cells. Either ABT-100 or L-744,832 treatment significantly blocked the secretion of each of these cytokines at both concentrations tested.

Fig. 1.

Effects of FTIs on cytokine secretion from antibody-stimulated T cells. Cytokine secreting cells were detected using ELISPOT assays for IFN-γ (a), IL-2 (b), IL-4 (c), or TNF-α (d). B6 splenocytes (1 × 10⁶ cells/ml) were treated with the indicated concentration of ABT-100 or L-744,832 beginning 24 h before stimulation. ELISPOT assays were then performed on 1 × 10⁵ cells/well stimulated with 5 μg/ml anti-CD3 and 10 μg/ml anti-CD28 for 24 h (a) or 48 h (b-d). For each condition, the mean and standard deviation are shown for triplicate samples. An asterisk indicates a statistically significant difference between the sample and untreated cells as determined by Student’s t test (p<.05).

The results shown in Fig. 1 would be expected if the FTIs were causing cell death at the concentrations tested. To determine if these concentrations of ABT-100 or L-744,832 were cytotoxic, we measured cell death and T cell activation under the same conditions used for the ELISPOT measurements. We found that viability decreased between 5 and 15% with ABT-100 or L-744,832 treatment (unpublished results). T cell activation, as measured by CD25 expression was also not substantially affected by treatment with ABT-100 or L-744,832 (unpublished results). Together, these two results demonstrate that the concentrations of either ABT-100 or L-744,832 that were able to block the secretion of Th1 and Th2 cytokines did not block activation of T cell signaling or cause cell death.

To determine if the production of cytokines in both CD4⁺ and CD8⁺ T cells were affected by FTI treatment, we used intracellular cytokine staining to measure TNF-α and IFN-γ (Fig. 2). Splenocytes were treated with 1.0 μM ABT-100 or 10 μM L-744,832 or left untreated. For TNF-α measurements, after 48 h of FTI treatment, cells were stimulated with anti-CD3/anti-CD28 for 4 h. For IFN-γ measurements, after 24 h of FTI treatment, cells were stimulated with anti-CD3/anti-CD28 for 24 h. Brefeldin A was used to block cytokine secretion for the final 4 h of stimulation. Following brefeldin A treatment, cells were labeled with CD4 and CD8 fluorescent antibodies, fixed, permeabilized, and then stained with a fluorescent antibody to the appropriate cytokine. We found that TNF-α was produced by both CD4⁺ and CD8⁺ T cells (Fig. 2a and d) and that IFN-γ was produced predominantly by CD8⁺ T cells (Fig. 2g). When the cells were treated with ABT-100 or L-744,832, we found that the amount of cytokine produced decreased substantially in both CD4⁺ and CD8⁺ T cells (Fig. 2). These results demonstrate that production of cytokines in both CD4⁺ and CD8⁺ naïve Tcells is sensitive to FTI treatment.

Fig. 2.

Effects of FTIs on production of TNF-α and IFN-γ in CD4⁺ and CD8⁺ T cells. Intracellular cytokine staining was used to measure TNF-α and IFN-γ production in brefeldin A-treated cells. Splenocytes at 1 × 10⁶ cells/ml were treated with 1.0 μM ABT-100 (b, e, and h), 10 μM L-744,832 (c, f, and i), or left untreated (a, d, and g). Cells were stimulated with 5 μg/ml anti-CD3 and 10 μg/ml anti-CD28 either 44 h later (a-f) or 24 h later (g-i). Brefeldin A was added 44 h after FTI treatment and 4 h later cells were stained with fluorescent antibodies and analyzed by flow cytometry. The percentage of viable cells in selected quadrants is shown.

We confirmed that the effects of ABT-100 or L-744,832 on cytokine secretion required preincubation of the cells with FTI. When we measured TNF-α in T cells that were stimulated immediately after FTI addition (Fig. 3), we found that neither ABT-100 nor L-744,832 blocked production of the cytokine. The inhibitory effects of FTI treatment required incubation of the cells for at least 48 h (unpublished results), presumably because protein turnover of farnesylated proteins is necessary.

Fig. 3.

Effects of FTIs on production of TNF-α in T cells without pretreatment. Intracellular cytokine staining was used to measure TNF-α production in brefeldin A-treated cells. Splenocytes at 1 × 10⁶ cells/ml were left untreated (a), treated with 1.0 μM ABT-100 (b), or treated with 10 μM L-744,832 (c). At the same time, cells were stimulated with 5 μg/ml anti-CD3 and 10 μg/ml anti-CD28 and Brefeldin A was added. After 4 h, cells were stained with fluorescent antibodies and analyzed by flow cytometry. The percentage of viable cells in selected quadrants is shown.

ELISPOTs on MLCs were used to determine if the ABT-100 FTI could block cytokine secretion from naïve, alloreactive T cells. B6 splenocytes, treated with 0 to 1 μM ABT-100, were stimulated with mitomycin C-treated cells from either BALB/c mice, which are fully MHC mismatched, or bm12 mice, which have a single class II MHC mismatch (Fig. 4). We measured secretion of the Th1 cytokine, IFN-γ, after 24 h and the Th2 cytokine, IL-4, after 48 h. The allogeneic cells caused secretion of IFN-γ and IL-4 from untreated responder splenocytes. ABT-100 at 1.0 μM significantly blocked secretion of both of these cytokines from class II-disparate alloreactive T cells and secretion of IFN-γ from class I/II-disparate alloreactive T cells.

Fig. 4.

Effects of FTIs on cytokine secretion from alloreactive T cells. B6 splenocytes at 2 × 10⁷ cells/ml were treated with ABT-100 beginning 24 h before stimulation and then 2 × 10⁶ cells/well were stimulated with 4 × 10⁵ mitomycin C-treated BALB/c or bm12 splenocytes for 24 h (a) or 48 h (b). Cytokine secreting cellswere detected using ELISPOT assays for IFN-γ (a) or IL-4 (b). For each condition, the mean and standard deviation are shown for triplicate samples. An asterisk indicates a statistically significant difference between the sample and untreated cells as determined by Student’s t test (p<.05).

We next measured proliferation in MLCs to determine if the block in cytokine secretion by ABT-100 could affect the cytokine-dependent proliferation of alloreactive T cells. B6 splenocytes were stimulated with mitomycin C-treated BALB/c cells and proliferation was measured by ³H-thymidine incorporation (Fig. 5a). Treatment of the responder cells with 0.1 μM orgreater ABT-100 significantly blocked the proliferation of the alloreactive T cells stimulated by BALB/c cells. This block in proliferation was accompanied by a smaller decrease in cell viability at the highest concentration of ABT-100 tested (unpublished results), but the block in proliferation was much greater than the decrease in viability.

Fig. 5.

Effects of ABT-100 on proliferation of T cells in vitro. (a) Proliferation of alloreactive T cells in MLCs was measured using ³H-thymidine incorporation. B6 splenocytes at 2 × 10⁶ cells/ml were treated with the indicated concentration of ABT-100 and then stimulated with 4 × 10⁶ cells/ml mitomycin C-treated BALB/c stimulator cells. After 72 h of stimulation ³H-thymidine incorporation was measured for 18 h by filter binding. Proliferation is shown relative to cells stimulated with allogeneic cells but not treated with ABT-100. (b) Proliferation of purified T cells stimulated with antibodies or with IL-2 was measured using BrdU incorporation. T cells were purified from B6 splenocytes by negative selection. The enriched T cells at 2 × 10⁶ cells/ml were treated with the indicated concentration of ABT-100 for 24 h and then 2 × 10⁵ cells/well were stimulated for 60 h with 5 μg/ml anti-CD3 and 10 μg/ml anti-CD28 or with 2 ng/ml murine IL-2. During the last 24 h of stimulation BrdU incorporationwas measured with a colorimetric assay. Proliferation is shown relative to cells stimulated with antibodies or with IL-2 but not treated with ABT-100. For each condition, the mean and standard deviation are shown for triplicate samples. An asterisk indicates a statistically significant difference between the sample and untreated cells as determined by Student’s t test (p<.05).

The decreased proliferation of FTI-treated alloreactive T cells could be caused by blocking the secretion of cytokines or by inhibiting the mitogenic effects of cytokines on lymphocytes. To determine if cytokine-stimulated cell proliferation was blocked by ABT-100 treatment, we measured T cell proliferation induced by IL-2 (Fig. 5b). B6 splenocytes were enriched for T cells by negative selection and then treated with 0 to 10 μM ABT-100. After 24 h, the T cells were then stimulated with 2 ng/ml murine IL-2 and proliferation was measured by BrdU incorporation. For comparison, we separately stimulated cells with CD3/CD28 antibodies. Both IL-2 and antibody stimulation caused proliferation of the purified T cells, and this proliferation was significantly blocked by treatment with ABT-100. Higher doses of ABT-100 were required to maximally block proliferation induced by IL-2, but both stimuli were inhibited similarly by 1 μM ABT-100 (Fig. 5b). Together with the previous data, this demonstrates that FTI treatment can both block the secretion of cytokines and block the proliferative response of T cells to mitogenic cytokines.

To test if FTI treatment could affect an alloreactive immune response in vivo, we measured the effects of ABT-100 or L-744,832 treatment on rejection of class II MHC mismatched skin allografts in mice. Shown in Fig. 6a, 5 B6 mice were treated continuously with 0.5 mg/ml ABT-100 in their drinking water (~100 mg/kg/day) beginning 2 days before surgery. Full thickness tail skin from a bm12 mouse was engrafted onto the trunks of these mice and 4 untreated B6 mice. The mean graft survival time for the untreated mice was 13.8 ± 1.5 days and was 38.4 ± 23.0 days for the mice treated with ABT-100 (Fig. 6a). We also determined if the L-744,832 FTI could delay skin allograft rejection (Fig. 6b). Intravenous treatment twice daily with ~40 mg/kg/day L-744,832 began the day after surgery and continued for 7 days. The mean graft survival time for the 7 untreated mice was 12.8±3.8 days and was 26.4±14.2 days for the 8 mice treated with L-744,832. Both ABT-100 and L-744,832 treatments cause significant delays (p<.05) in the rejection of class II MHC mismatched skin allografts on naïve mice.

Fig. 6.

Effects of FTIs on skin allograft rejection in primary responses. Skin allografts were performed using full thickness tail skin from bm12 donor mice onto the trunks of B6 mice. (a) Four transplant recipients were left untreated and 5 recipients were treated orally with 100 mg/kg/day ABT-100 beginning 2 days prior to surgery. (b) Eight transplant recipient mice were treated with 40 mg/kg/day L-744,832 i.v. for 7 days and 7 recipients were left untreated. Control B6 mice that received syngeneic skin grafts from B6 donor mice are shown for comparison.

We next performed second set grafts to determine if the effects of FTI treatment on allograft rejection could also delay rejection mediated by T cells primed to the class II alloantigen (Fig. 7). Skin allograft surgeries using bm12 tail skin were performed on a cohort of B6 mice and rejection was allowed to occur without FTI treatment. Approximately 4 weeks later, a second bm12 skin allograft was performed on the opposite side of the trunk of each mouse. A group of 6 mice were treated with L-744,832 as above for 7 days beginning the day after the second surgery and 7 mice were left untreated. The mean graft survival time for the untreated mice was 9.7±1.6 days and was 10.5±1.8 days for the treated mice. The difference between these rejection times was not significant. Under the same conditions that L-744,832 delayed primary allograft rejection, this FTI did not affect allograft rejection in mice previously primed with the mismatched class II MHC alloantigen.

Fig. 7.

Effects of L-744,832 on skin allograft rejection in a secondary alloreactive response. Skin allografts were performed on B6 mice that had previously rejected a bm12 graft approximately 4 weeks earlier. Seven of the recipient mice were left untreated and 6 of the mice were treated with 40 mg/kg/day L-744,832 i.v. for 7 days following the transplant.

5. Discussion

5.1. Specificity of FTIs

By demonstrating that FTIs could affect alloreactive immune responses in vitro and in vivo, we have provided evidence that FTIs could be useful immune modulators. We used two different, structurally unrelated, FTIs for several of the experiments (Figs. 1–3 and 6) to establish that the effects we observed were specific to the inhibition of farnesyltransferase. FTIs can have off-target effects, including proteasome inhibition [32], but the observation that multiple FTIs exhibit the same effect increase the likelihood that the effects are due to the inhibition of prenylation. A previous study on cytokine secretion by murine Th1 and Th2 clones, showed that five different FTIs, including L-744,832, could block cytokine secretion [21].

The delay between FTI treatment and the effects that we have observed is consistent with the hypothesis that protein prenylation is the target of FTI treatment. Because farnesylation is an irreversible modification within the cell, protein turnover is required before the function of modified proteins is blocked. In this study, when we examined the effects of pretreating cells with FTIs for less than 24 h, we did not observe inhibition of cytokine production (Fig. 3).

At high doses, FTIs are known to inhibit geranylgeranyltransferase I [1], a prenylation enzyme closely related to farnesyltransferase. For our in vitro experiments, significant effects were observed at FTI concentrations that are not known to block geranylgeranyltransferase. For example, in tissue culture experiments the ABT-100 FTI has been shown to have an EC₅₀ for farnesyltransferase of 0.7 nM and for geranylgeranyltransferase of >30 μM [17]. For our experiments in mice (Figs. 6 and 7), we used the maximal dose possible in order to increase the likelihood of observing an effect. For L-744,832 we used the highest dose that the mice could tolerate (unpublished results) and for ABT-100 we used the highest dose that could be reliably dissolved. Future experiments will be necessary to determine if lower doses of the drugs are effective. However, given that both FTIs showed similar efficacy and that ABT-100 has demonstrated >100,000-fold selectivity for farnesyltransferase inhibition over geranylgeranyltransferase [17], it is most likely that the effects we have observed are due to farnesyltransferase inhibition.

The farnesylated protein or proteins that are the functional target of FTI treatment are not known. Up to 17 proteins are farnesylated within the cell and several of them are known to be involved in lymphocyte signaling [1]. Interestingly, transgenic mice that express dominant negative H-Ras show delayed rejection of skin allografts that are fully mismatched at MHC class I and II [33]. This observation suggests that H-Ras inhibition may be involved in the action of FTIs during allograft rejection. However, the inhibition of cytokine secretion by FTI treatment does not appear to involve Ras inhibition [21]. Therefore, it appears that some combination of farnesylated proteins are the likely targets of FTI inhibition, as has been shown for the anti-neoplastic activity of these drugs [1].

5.2. FTIs block cytokine production and cytokine-induced proliferation in T cells

The ability of ABT-100 and L-744,832 to block secretion of cytokines provides a possible mechanism for their specific immunosuppression activity. ABT-100 prevented secretion of IFN-γ, IL-2, IL-4, and TNF-α from naïve T cells, extending the previous observations that this FTI blocks secretion of IFN-γ and IL-2 from murine Th1 clones and secretion of IL-4 and IL-5 from murine Th2 clones [21]. We confirmed that FTIs were blocking cytokine production, and not simply killing the cytokine-producing cells, by measuring cell viability and by directly measuring cytokine synthesis within viable T cells (Fig. 2). In addition to affecting CD4⁺ T cells, we have demonstrated that ABT-100 can block secretion of TNF-α and IFN-γ from naïve CD8⁺ T cells (Figs. 1 and 2). This result is consistent with other observations that TNF-α secretion can be blocked by FTIs both in vitro and in vivo [18,20]. A less robust, but still significant, inhibition of cytokine secretion was seen when allogeneic stimulator cells were used instead of anti-CD3 and anti-CD28 antibodies (Fig. 4). Our further observation that the mitogenic response of T cells to IL-2 is blocked by ABT-100 (Fig. 5b) indicates that the effects of cytokines on target cells may also be blocked by FTIs. The block in proliferation is not mediated by preventing the expression of high affinity IL-2 receptors on activated T cells, which we have found is unaffected by FTI treatment (unpublished results). This observation is consistent with previous results showing that ABT-100 does not effect expression of CD25 or other activation markers on phytohemagglutinin-activated human peripheral blood mononuclear cells [24]. Together, our results show that FTIs may have several effects on immune responses by modulating the production of cytokines from a variety of cell types and the mitogenic responses of cells to cytokines.

5.3. FTIs block proliferation of alloreactive T cells in MLCs

FTI treatment was found to block alloreactive cell proliferation in both CD4⁺ and CD8⁺ T cells. Using BALB/c stimulator cells mismatched at both class I and II MHC, ABT-100 was found to inhibit proliferation by up to 85% (Fig. 5a). This provides clear evidence that the effects of FTI treatment on cytokine secretion can delay an alloreactive immune response and that the target of FTI action is the alloreactive T cell. While the in vitro results do not preclude additional effects on the recognition or effector phases of allorejection, a principal target of FTI action appears to be the expansion phase. Proliferation of Tcells under these conditions is dependent on the mitogenic activity of cytokines; we also found that T cell proliferation induced by IL-2 was blocked by ABT-100 (Fig. 5b). We conclude that FTIs can affect both the secretion of cytokines and their mitogenic stimulation of lymphocytes.

5.4. FTI treatment can delay alloreactive rejection in vivo

Treatment with either ABT-100 or L-744,832 delayed rejection of class II MHC-disparate primary allografts (Fig. 6), directly demonstrating that FTIs can affect immune responses in vivo. Our results indicate that FTIs may delay the expansion of CD4⁺ T cells once primed to alloantigens by affecting cytokine secretion. By delaying the expansion of the alloreactive T cells (Fig. 5), FTIs can slow the rejection of primary allografts but do not affect secondary rejection (Fig. 7), because the expansion of alloreactive clones has already occurred.

FTIs may also affect priming of the alloreactive immune response. Priming of CD4⁺ T cells to MHC class II alloantigens involves several steps, including the direct pathway of antigen presentation, recognition of MHC by CD4⁺ T cells, and activation of alloreactive T cells. Our results from in vitro assays indicate that FTIs can block the cytokine secretion that is involved in both the proliferation and differentiation of alloreactive T cells after they have been primed. While we cannot exclude possible effects of FTIs on antigen presentation or recognition of alloantigens, the effects of FTIs on cytokine secretion are sufficient to explain the delay in allograft rejection.

5.5. Possible mechanisms of FTI action in allograft rejection

We chose a stringent model of MHC class II-mediated skin allograft rejection because it had previously been shown to be dependent on the secretion of a single cytokine, IFN-γ B6 mice deficient in IFN-γ or treated with anti-IFN-γ antibodies are not capable of rejecting bm12 allografts, although they exhibit normal allorejection of class I mismatched and fully mismatched grafts [34,35]. We demonstrated that treatment with either ABT-100 or L-744,832 can delay bm12 graft rejection in mice (Fig. 5). Therefore, we can conclude that FTI treatment affects MHC II allorecognition and the activation of naïve CD4⁺ T cells. We provide two possible mechanisms for this activity of xFTIs involving IFN-γ By blocking the secretion of IFN-γ from activated naïve CD4⁺ T cells, FTIs may prevent the generation of a proinflammatory environment and the IFN-γ-dependent secretion of the Mig chemokine, which is also required for bm12 graft rejection in B6 mice [36]. Second, FTIs may block the effects of IFN-γ and other cytokines on the proliferation and/or survival of effector cells, as we found for IL-2 (Fig. 5b).

It is likely that FTI treatment can affect other cytokines in addition to those measured in this study. We demonstrated that FTI treatment affected the Th2 cytokine, IL-4 (Fig. 1). Class II mismatched graft rejection involves secretion of both IL-4 andIL-5 by Th2 cells [37,38].A third possible mechanism of action of FTIs on graft rejection involves these Th2 cytokines. FTI treatment has the potential to affect the initial coordination of the alloreactive response by stimulated naïve T cells and alter their differentiation into Th1, Th2, and Treg cells. Inhibition of Ras has been shown to increase FoxP3 expression in a human T cell line [39], suggesting that FTI treatment could increase the number of Treg cells induced during an immune response, presumably at the expense of Th1 or Th2 cells. To determine if FTI treatment affects the balance of Th1, Th2, and Treg cells generated, future studies will be necessary to measure the effects of FTIs on cytokine production in vivo and the generation of allospecific effector cells in graft-draining lymph nodes.

Although several mechanisms may be involved in delaying allograft rejection, our results suggest that the block in IFN-7 secretion may be the most critical. The observation that second set rejection was not affected by L-744,832 treatment (Fig. 7) demonstrated that, after CD4⁺ T cells had been primed to the MHC II alloantigen, the FTI no longer significantly delayed the rejection process. This observation is similar to that seen in IFN-γ knockout mice that readily reject bm12 skin allografts only after priming to the alloantigen [34]. The ability of FTI-treated mice to rapidly reject second set grafts demonstrates that FTI treatment does not block the effector mechanisms responsible for tissue cytotoxicity, including lymphocyte migration and the production of proinflammatory cytokines within the skin graft. Our results suggest that once an alloreactive response had taken place during the primary rejection, treatment with FTIs was unable to delay the strong allorejection response of the primed effector T cells. This observation may suggest that FTIs would be less useful to treat ongoing alloreactive immune responses than as preventative therapy.

5.6. Implications

Our results have particular implications for the treatment of malignancies by hematopoietic stem cell (HSC) transplantation and the induction of tolerance following solid organ transplantation. By including an FTI in the preparative or maintenance phases of HSC therapy, we suggest that it might be possible to utilize both the direct anti-neoplastic and the potential graft versus host disease inhibitory activities of an FTI to improve outcomes. Through similar immunomodulatory mechanisms, it may be possible to use FTIs to delay allorejection and induce tolerance following solid organ transplantation.