The endoplasmic reticulum aminopeptidase associated with Ag processing, ERAAP, plays an important role in the trimming of antigenic peptides for presentation at the cell surface complexed with MHC class I molecules. Tumors express varying levels of ERAAP, highlighting a possible mechanism of immune-evasion through alteration of the peptide repertoire. Using the CT26 tumor model, we investigated the effects of ERAAP modulation on peptide presentation and the use of ERAAP inhibition as an antitumor therapy. We show that generation of the cross-protective tumor Ag GSW11 in the colorectal carcinoma CT26 is increased when ERAAP expression is reduced. BALB/c mice with reduced ERAAP expression challenged with CT26 induced protective immunity that was mediated by CD8+ T cells. This antitumor immunity also protected mice when rechallenged with wild-type CT26 tumor; strong CD8+ T cell responses to GSW11 were observed, despite its presentation being considerably lower. Furthermore, boosting the tumor immunogenicity through inhibition of ERAAP function with the small molecule inhibitor leucinethiol in vitro, or in established tumors in vivo, abrogated tumor growth and prolonged survival. Thus, our results highlight the promising possibility of using modulation of ERAAP to generate protective antitumor responses as a strategy for cancer immunotherapy.

The repertoire of antigenic peptides derived from intracellular proteins consists of thousands of different 8–10-aa peptides, many of which are generated as N-terminally extended precursors following partial hydrolysis of substrates by cytosolic proteases, including the proteasome (1, 2). These are then transported into the endoplasmic reticulum by the TAP. A key final step of the pathway involves aminopeptidases, which trim the extra N-terminal residues to generate peptides (1) that are an optimal length for loading onto MHC class I (MHC I) molecules. This completes MHC I biogenesis and permits exit from the endoplasmic reticulum to the cell surface (36), where they present the processed peptides to CTL. In humans, two aminopeptidases perform this function, ERAP1 and ERAP2 (79); in mice, only one aminopeptidase exists, ERAAP (10). ERAAP is critical for the generation of many antigenic epitopes in vivo and can influence the generation of the antigenic peptide repertoire (1114). ERAAP can also destroy antigenic epitopes by trimming them to lengths too small for MHC binding, thus its characterization as an antigenic peptide editor (12).

In tumors, expression of ERAP1 and ERAP2 was shown to vary considerably, with high/moderate levels in skin cancer and melanoma and low levels in breast and lung cancer (1517). Varying expression of ERAP1/2 in tumors could indicate a mechanism by which the tumors evade detection by CTL through alteration of the peptide repertoire presented by MHC I, perhaps reducing the generation of dominant tumor-derived peptide epitopes either by reduced trimming activity or by destroying them by overtrimming.

To investigate the contribution of ERAAP to immunogenicity and immunological control of tumor growth, we used the transplantable murine colon carcinoma model CT26 and modulated ERAAP levels through small interfering RNA (siRNA). This model allowed us to investigate the correlation among ERAAP expression, the generation of well-characterized tumor epitopes, and protective and durable antitumor immunity. We found that reducing ERAAP levels enhanced tumor immunogenicity, thereby opening a window of opportunity for the potential use of small molecule ERAAP inhibitors to enhance antitumor immunity. Accordingly, treatment of CT26 with the ERAAP inhibitor leucinethiol (leuSH) prior to injection or following tumor establishment in vivo led to prolonged survival.

All CT26 and P815 tumor cells, COS-7 cells, and the GSW11-specific lacZ-inducible CD8+ T cell hybridoma CCD2Z were maintained as previously described (18). In all experiments, mice were injected s.c. with 105 CT26 tumor cells in PBS. All cell lines were tested and were mycoplasma free.

BALB/c mice were bred locally under specific pathogen–free conditions and were used at 6–8 wk of age. Mice were housed in conventional facilities during experimental procedures. All experiments were carried out under license according to local ethical committee approval under United Kingdom Home Office license guidelines. Hybridomas secreting CD8 (YTS 169.4.2.1, YTS 156.7.7, both rat IgG2b) and CD4 (YTS 191.1.2, YTA 3.1.2, both rat IgG2b)-specific mAb were described previously (19) and were grown in culture and purified by precipitation in saturated ammonium sulfate. For depletion, mice were injected with 100 μg each mAb twice with a 1-d interval. For intratumoral administration of leuSH, mice were challenged with CT26 tumor, and 100 μM leuSH in 0.1 M DTT in 50 μl was injected into the tumor at days 7 and 10. Controls received DTT-only injections.

The GSW11-specific lacZ-inducible T cell hybridoma CCD2Z was cocultured with CT26 (wild-type, ERAAP knockdown or inhibitor treated) or APCs that were transfected with Ag cDNAs. The lacZ activity was measured as described previously (20). For inhibitor treatment of CT26, cells were incubated with 10 μM MG132 (Calbiochem), 100 μM AAF-CMK (Calbiochem), or 30 μM leuSH (Sigma) for 16 h. Following incubation, the treated CT26 were assayed for stimulation of CCD2Z.

For stable ERAAP knockdown, ERAAP-specific siRNA (CT26-ERAAPmixed) or scrambled siRNA (ERAAPhi CT26) was cloned into the short hairpin RNA (shRNA) expression vector pSilencer 4.1 CMV/hygro (Life Technologies) and transfected into CT26 using FuGENE6 (Roche), according to the manufacturer’s instructions. After 24 h, transfected cells were selected by the addition of 0.5 μg hygromycin. CT26-ERAAPmixed cells were subcloned and screened for ERAAP protein knockdown using Western blot of CT26 cell lysates.

Total acid-soluble peptide pools from tumor cells CT26, CT26-ERAAPmixed, ERAAPmed CT26, and ERAAPlo CT26 were extracted as described (12). Briefly, 108 cells were washed with PBS and extracted with 400 μl 10% boiling formic acid in the presence of a 10 μM “martyr” peptide ASNENMETM for 10 min. Cellular debris was removed by centrifugation and fractionated by HPLC after filtration through a 10-kDa filter (Millipore, Billerica, MA). Reverse-phase C18 column (Grace Vydac, Hesperia, CA; 2.1 × 250 mm, 5 μM, 300 Å) was run in 0.1% trifluoroacetic acid in water (solvent A) and 0.1% trifluoroacetic acid in acetonitrile (solvent B). The gradient was from 17 to 45% (solvent B) at a rate of 1%/min, with fractions collected and dried in a vacuum centrifuge. Mock injections with sample buffer alone were performed prior to each extracted sample, using the same column and identical run conditions. The collected fractions were assayed in the same experiment, using the same APCs and T cells, in parallel with fractions from the cell extracts and synthetic peptide standards.

Spleen and draining lymph node cells from CT26-challenged mice (wild-type or ERAAP knockdown) were harvested between days 10 and 12. The CD8+ T cell responses to GSW11 and AH1 Ags were measured by production of IFN-γ by those cells in response to incubation with peptide. CD8+ T cells, APCs, and peptides were cultured together for 4 h in the presence of brefeldin A before being stained for surface CD8 and intracellular IFN-γ (BD Biosciences). For assessment of cell surface MHC I expression, H-2Dd–FITC (34-2-12) and H-2Ld-FITC (B22) Abs were incubated with CT26 for 30 min on ice and washed. Cells were then analyzed on a FACSCanto (BD Biosciences), and data were analyzed with FlowJo software (TreeStar). CD8+ lymphocytes were gated, and the percentage of IFN-γ–producing cells was calculated from total CD8+ T cells.

Statistical significance was assessed by the two-tailed paired Student t test or one-way ANOVA, using a Bonferroni multiple-comparison posttest. Survival data are presented as KaplanMeier plots.

In the absence of vaccination, protective immunity against the transplanted CT26 colon carcinoma requires the induction of a strong CTL response against the GSW11 epitope (18). Anti-GSW11 CTLs are preferentially suppressed by regulatory T cells (Tregs) and are only allowed to expand when Tregs are depleted prior to tumor challenge (18). Because GSW11-specific CD8+ T cell responses are important for tumor rejection, we were interested to know how the GSW11 epitope (GGPESFYCASW) was generated by proteases. Because GSW11 is relatively “long” for a CTL epitope, does not conform to the canonical H-2Dd binding motif (XGPXXXXX[I,L,F]) (21), and has a very low affinity for MHC I (t1/2 = 20 min) (18), we wondered whether it would be unusually susceptible to overtrimming by ERAAP, as observed for other CTL epitopes (12).

We incubated CT26 tumor cells with MG132 (a proteasome inhibitor), AAF-CMK (an inhibitor of tripeptidyl peptidase II [TPPII]), and leuSH (an ERAAP inhibitor). We also transfected COS-7 cells with the gp70 cDNA construct C8 (which contains the GSW11 epitope) and/or a TAP inhibitor derived from human herpesvirus 1, ICP47 (22, 23). The effect of the inhibitors on the generation of GSW11 and its cell surface presentation by H-2Dd were assessed using the GSW11-specific T cell hybridoma CCD2Z, which is able to detect nanomolar concentrations of presented peptide (Supplemental Fig. 1) (18). CCD2Z is a fusion between the GSW11-specific CTL, CCD2, and BWZ.36 CD8α fusion partner that contains an inducible reporter gene encoding β-galactosidase (lacZ) (24). Activation of the CCD2Z T cell hybridoma is assayed by measuring lacZ activity. MG132-treated CT26 and C8 and ICP47 cotransfection resulted in complete abrogation of CCD2Z stimulation (Fig. 1A, 1B). AAF-CMK treatment had no effect on CCD2Z stimulation, indicating that TPPII has no role in the generation of GSW11 (Fig. 1A). In contrast, incubation of CT26 with leuSH resulted in an increase in CCD2Z stimulation (Fig. 1A). Therefore, we concluded that, although generation of GSW11 requires both the proteasome and TAP for presentation, it is overtrimmed and destroyed by ERAAP, resulting in a steady-state level of presentation that is detectable but low.

FIGURE 1.

GSW11 is overtrimmed by ERAAP. (A) CT26 tumor cells were incubated with leuSH (ERAAP inhibitor), AAF-CMK (TPPII inhibitor), or MG-132 (proteasome inhibitor). Generation of GSW11 was assessed using the lacZ-inducible GSW11-specific T cell hybridoma CCD2Z. (B) COS-7 cells were transfected with H-2Dd and the GSW11-encoding C8 or vector only cDNA and/or ICP47 (TAP inhibitor). GSW11 generation was assessed by activation of CCD2Z. (C) CT26 cells were transfected with ERAAP or laminB1-specific siRNA or were mock transfected, and GSW11 generation was assessed by activation of CCD2Z. (D) The effect of ERAAP siRNA knockdown on cell surface MHC class I expression was assessed by flow cytometry. Data shown are representative of four experiments.

FIGURE 1.

GSW11 is overtrimmed by ERAAP. (A) CT26 tumor cells were incubated with leuSH (ERAAP inhibitor), AAF-CMK (TPPII inhibitor), or MG-132 (proteasome inhibitor). Generation of GSW11 was assessed using the lacZ-inducible GSW11-specific T cell hybridoma CCD2Z. (B) COS-7 cells were transfected with H-2Dd and the GSW11-encoding C8 or vector only cDNA and/or ICP47 (TAP inhibitor). GSW11 generation was assessed by activation of CCD2Z. (C) CT26 cells were transfected with ERAAP or laminB1-specific siRNA or were mock transfected, and GSW11 generation was assessed by activation of CCD2Z. (D) The effect of ERAAP siRNA knockdown on cell surface MHC class I expression was assessed by flow cytometry. Data shown are representative of four experiments.

Close modal

To confirm the role of ERAAP in destroying GSW11, CT26 tumor cells were transfected with ERAAP-specific siRNA. LaminB1-specific siRNA knockdown was used as a control. ERAAP siRNA knockdown resulted in an increase in CCD2Z stimulation (Fig. 1C), consistent with the results following leuSH treatment (Fig. 1A). As expected, laminB1 knockdown of CT26 did not affect CCD2Z stimulation (Fig. 1C). The increase in GSW11 generation occurred despite an overall decrease in MHC I levels; MHC I downregulation is typically observed in ERAAP-knockout cells (Fig. 1D) (1114). This confirmed that the generation of GSW11 is ERAAP sensitive.

To investigate the effect of GSW11 presentation on antitumor immunity, we created stable ERAAP knockdown lines of CT26 using an ERAAP-specific shRNA vector. A polyclonal line was established (CT26-ERAAPmixed) from which two clones were generated that showed greater stimulation of CCD2Z than control-transfected CT26 (Fig. 2A). Analysis of ERAAP expression revealed that these showed an ∼80% (ERAAPlo CT26) and 65% (ERAAPmed CT26) reduction in protein levels (Fig. 2B), thus establishing a correlation between the level of ERAAP expression and the level of CCD2Z stimulation.

FIGURE 2.

Generation of stable ERAAP knockdown CT26. (A) CT26 were stably transfected with ERAAP (CT26-ERAAPmixed) or scrambled shRNA (ERAAPhi CT26) and assessed for stimulation of CCD2Z. Two clones established from the polyclonal CT26-ERAAPmixed cell line, ERAAPmed CT26 and ERAAPlo CT26, were also assessed for CCD2Z stimulation. (B) ERAAP expression in all CT26 cell lines was assessed by Western blot. Data shown are representative of three experiments.

FIGURE 2.

Generation of stable ERAAP knockdown CT26. (A) CT26 were stably transfected with ERAAP (CT26-ERAAPmixed) or scrambled shRNA (ERAAPhi CT26) and assessed for stimulation of CCD2Z. Two clones established from the polyclonal CT26-ERAAPmixed cell line, ERAAPmed CT26 and ERAAPlo CT26, were also assessed for CCD2Z stimulation. (B) ERAAP expression in all CT26 cell lines was assessed by Western blot. Data shown are representative of three experiments.

Close modal

To confirm that the increase in CCD2Z stimulation observed with reduced ERAAP protein was due to an increase in the generation of GSW11, we generated peptide extracts from tumor cells with different ERAAP expression levels. These were fractionated by reverse-phase HPLC, and the amount of GSW11 was quantified in the fractionated mixture using CCD2Z as a biological detector. The profile obtained from CT26 peptide extracts revealed one significant peak that coeluted with synthetic GSW11 (Fig. 3A). Analysis of the polyclonal line CT26-ERAAPmixed revealed a 3-fold increase in GSW11 compared with CT26 (Fig. 3B). Greater increases in GSW11 were observed for ERAAPmed and ERAAPlo CT26 (8- and 30-fold increases, respectively; Fig. 3C, 3D). Importantly, the amount of GSW11 generated correlated with the amount of ERAAP protein present within the cells. The presence of an additional peak in ERAAP-knockdown CT26 extracts eluting ahead of GSW11 is also seen in CT26 cells when developed for a longer period of time; this was shown to be an oxidized adduct (or dimer) that probably forms during the extraction process (18). Therefore, we concluded that the increased response of CCD2Z observed after ERAAP knockdown is due to an increase in the amount of GSW11 generated in these cells.

FIGURE 3.

Knockdown of ERAAP expression increases the amount of GSW11 generated. Peptides were extracted from CT26 (A), CT26-ERAAPmixed (B), ERAAPmed CT26 (C), or ERAAPlo CT26 (D) cells and fractionated on a reverse-phase C18 column by HPLC. Each fraction was dried, resuspended in PBS, and tested for its ability to stimulate lacZ activity in CCD2Z T hybridoma with P815 cells as APCs. The arrow indicates the peak activity of synthetic GSW11 peptide run under identical conditions. Data shown are representative of four experiments.

FIGURE 3.

Knockdown of ERAAP expression increases the amount of GSW11 generated. Peptides were extracted from CT26 (A), CT26-ERAAPmixed (B), ERAAPmed CT26 (C), or ERAAPlo CT26 (D) cells and fractionated on a reverse-phase C18 column by HPLC. Each fraction was dried, resuspended in PBS, and tested for its ability to stimulate lacZ activity in CCD2Z T hybridoma with P815 cells as APCs. The arrow indicates the peak activity of synthetic GSW11 peptide run under identical conditions. Data shown are representative of four experiments.

Close modal

Having shown that ERAAP-knockdown CT26 generated and presented significantly more GSW11 compared with CT26, we wanted to determine whether the increase in GSW11 presentation was able to overcome Treg suppression and induce GSW11-specific T cells in vivo. We inoculated BALB/c mice with ERAAPlo, ERAAPmed, or untreated CT26 and assessed tumor growth and survival. Mice inoculated with CT26 had detectable tumors from day 8 that continued to grow until reaching the humane end point, with a median survival time of 20 d (Fig. 4A, 4C). All mice challenged with ERAAPlo and ERAAPmed CT26 had tumor growth at days 8–12 that was subsequently cleared in the majority of cases (8/12), indicating a role for an effective adaptive antitumor response in these mice. Mice challenged with ERAAPlo CT26 showed a significant increase in survival time compared with CT26, with >90% of mice rejecting the tumor (p < 0.001; Fig. 4A, 4C). Inoculation with ERAAPmed CT26 induced tumor protection in 50% of mice (Fig. 4A), with a median survival time of 64.5 d compared with 20 d for CT26 (p < 0.005; Fig. 4A, 4C). These results showed that the increased generation of GSW11 resulting from ERAAP knockdown produced a protective antitumor response whose intensity correlated negatively with the level of ERAAP expression. To investigate the cells responsible for this tumor protection, we depleted mice of either CD4 or CD8 T cells prior to challenge with ERAAPlo CT26. Depletion of CD8 T cells completely abrogated the ability to clear tumor (Fig. 4B). Depletion of CD4 T cells also resulted in a small decrease in tumor rejection, although the majority of mice survived tumor challenge (Fig. 4B), confirming that the rejection of ERAAPlo CT26 tumor was primarily mediated by CD8+ T cells.

FIGURE 4.

ERAAPlo CT26 induces CD8+ T cells capable of mediating tumor rejection. BALB/c mice were challenged with wild-type (n = 6), ERAAPmed (n = 6), or ERAAPlo (n = 7) CT26, and tumor survival (A) and tumor growth (C) were assessed. **p < 0.01, ***p < 0.001, versus CT26. (B) Untreated CD4+ or CD8+ T cell–depleted ERAAPlo CT26–challenged mice were assessed for tumor survival. Control mice were challenged with wild-type CT26 (n = 5/group). *p < 0.05, versus ERAAPlo CT26. Data represent two independent experiments with at least five mice/group.

FIGURE 4.

ERAAPlo CT26 induces CD8+ T cells capable of mediating tumor rejection. BALB/c mice were challenged with wild-type (n = 6), ERAAPmed (n = 6), or ERAAPlo (n = 7) CT26, and tumor survival (A) and tumor growth (C) were assessed. **p < 0.01, ***p < 0.001, versus CT26. (B) Untreated CD4+ or CD8+ T cell–depleted ERAAPlo CT26–challenged mice were assessed for tumor survival. Control mice were challenged with wild-type CT26 (n = 5/group). *p < 0.05, versus ERAAPlo CT26. Data represent two independent experiments with at least five mice/group.

Close modal

To investigate the specificity of the anti-ERAAPlo CT26 response in more detail, we inoculated mice with either ERAAPlo CT26 or CT26 and assessed the anti-AH1 and anti-GSW11 CD8+ T cell responses in spleens and draining lymph nodes using intracellular cytokine staining. As shown previously (18), anti-AH1 CD8+ T cell responses were observed following inoculation with CT26 (∼0.4%), with no GSW11-specific responses detected above that seen in unstimulated controls (Fig. 5A). Interestingly, despite detectable anti-AH1 responses in tumor and draining lymph nodes, these mice succumb to tumor by day 25. In contrast, all mice challenged with ERAAPlo CT26 had GSW11-specific CD8+ T cell responses that were much greater than those observed in CT26-challenged mice (∼0.75%, p < 0.001; Fig. 5A), similar to those observed in mice depleted of Tregs (18). The anti-GSW11 response dominated over AH1-specific responses (∼0.75% versus ∼0.6%; p < 0.05) in these mice, even though the latter were also increased compared with CT26 challenge (∼0.6% versus ∼0.4%; Fig. 5A). Because >90% of mice challenged with ERAAPlo CT26 reject the tumor, the correlation of protection and the appearance of GSW11-specific CD8+ T cells suggests the importance of these cells in clearance of the tumor. Although AH1-specific CD8+ T cells were also increased, the modest difference suggests that these responses are unlikely to play a major role in tumor clearance.

FIGURE 5.

ERAAPlo CT26 challenge induces responses to GSW11-specific T cells. (A) Responses to AH1 and GSW11 Ags were assessed in mice challenged with either CT26 or ERAAPlo CT26. (B) Representative dot plots of AH1- and GSW11-specific CD8+ T cell responses in CT26- or ERAAPlo CT26–challenged mice. Percentages adjacent to outlined areas indicate CD8+ IFN-γ–producing cells. (C) ERAAPlo CT26–immune mice were rechallenged 50 d later with either ERAAPlo CT26 or CT26, and tumor survival was assessed. Control naive mice were challenged with CT26 (n = 5/group). Data are representative of two independent experiments. Anti-AH1 and anti-GSW11 CD8+ T cell responses from ERAAPlo CT26–rechallenged (D) or CT26-rechallenged (E) ERAAPlo CT26–immune mice were assessed. *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 5.

ERAAPlo CT26 challenge induces responses to GSW11-specific T cells. (A) Responses to AH1 and GSW11 Ags were assessed in mice challenged with either CT26 or ERAAPlo CT26. (B) Representative dot plots of AH1- and GSW11-specific CD8+ T cell responses in CT26- or ERAAPlo CT26–challenged mice. Percentages adjacent to outlined areas indicate CD8+ IFN-γ–producing cells. (C) ERAAPlo CT26–immune mice were rechallenged 50 d later with either ERAAPlo CT26 or CT26, and tumor survival was assessed. Control naive mice were challenged with CT26 (n = 5/group). Data are representative of two independent experiments. Anti-AH1 and anti-GSW11 CD8+ T cell responses from ERAAPlo CT26–rechallenged (D) or CT26-rechallenged (E) ERAAPlo CT26–immune mice were assessed. *p < 0.05, **p < 0.01, ***p < 0.001.

Close modal

Having shown that challenge with ERAAPlo CT26 induces robust protective antitumor immunity that is CD8 dependent, we went on to show that tumor challenge also induced immunological memory. We rechallenged ERAAPlo CT26 immune mice with ERAAPlo CT26 50 d after the primary challenge. All mice that rejected the first tumor challenge also rejected the second (Fig. 5C). Examination of CD8+ T cell responses in mice 10 d after rechallenge revealed increased anti-GSW11 and anti-AH1 T cell responses compared with those observed in the primary challenge with dominant anti-GSW11 responses (∼2.25% versus ∼1%, p < 0.01; Fig. 5D).

To consider ERAAP silencing as a potential antitumor therapy, the induced immunity following primary challenge with ERAAPlo CT26 would also need to be effective against tumor cells with normal levels of ERAAP (ERAAP silencing may not be effective in all tumor cells or those at additional sites in the body). The presence of anti-AH1 and anti-GSW11 CD8+ T cell responses in ERAAPlo CT26–challenged mice suggests that ERAAPlo CT26 immune mice would be able to clear wild-type CT26, despite a lower level of epitope production. To assess this, ERAAPlo CT26 immune mice were immunized with CT26 50 d after primary tumor challenge. All ERAAPlo CT26–immune mice were protected from CT26 tumor challenge, whereas control naive mice succumbed to the tumor by day 25 (Fig. 5C). Assessment of CD8+ T cell responses 10 d following CT26 challenge revealed strong anti-GSW11 and anti-AH1 responses greater than those observed after primary ERAAPlo CT26 challenge (2 and 1.25% versus 0.8 and 0.6%; Fig. 5A, 5E). This suggests that the dominant protective CD8+ T cell responses in these mice were directed toward GSW11 and AH1 peptide epitopes. Therefore, these memory CTL responses to epitopes shared between ERAAPlo CT26 and CT26 (GSW11 and AH1) allow rejection of CT26 tumor following rechallenge, even though the presentation of these epitopes is considerably lower in the latter. The results also highlight the validity of attenuating ERAAP function as a strategy for the induction of protective antitumor immunity.

We investigated whether attenuation of ERAAP activity in CT26 with leuSH treatment either prior to injection or in situ would trigger tumor rejection. Treatment of CT26 with LeuSH in vitro results in a transient inhibition of ERAAP that persists for 24–48 h (Supplemental Fig. 2). Mice challenged with CT26 treated in vitro with leuSH showed prolonged survival in four of seven recipients (Fig. 6A; p < 0.05), with two mice surviving beyond day 50. Tumor was present in these two mice but was stable (Fig. 6C), indicating that a protective anti-CT26 response was induced that favored tumoristasis over rejection.

FIGURE 6.

ERAAP inhibition attenuates tumor growth in vivo. (A) Survival of BALB/c mice challenged with in vitro leuSH-treated CT26 or CT26 (n = 7/group). (B) CT26-challenged BALB/c mice were injected with leuSH and DTT or DTT only intratumorally on days 7 and 10 (n = 7/group). Tumor-growth kinetics of BALB/c mice challenged with in vitro (C) or in vivo (D) leuSH-treated CT26. Arrows indicate times of intratumor injections. *p < 0.05.

FIGURE 6.

ERAAP inhibition attenuates tumor growth in vivo. (A) Survival of BALB/c mice challenged with in vitro leuSH-treated CT26 or CT26 (n = 7/group). (B) CT26-challenged BALB/c mice were injected with leuSH and DTT or DTT only intratumorally on days 7 and 10 (n = 7/group). Tumor-growth kinetics of BALB/c mice challenged with in vitro (C) or in vivo (D) leuSH-treated CT26. Arrows indicate times of intratumor injections. *p < 0.05.

Close modal

We next investigated whether inhibition of ERAAP by topical delivery of leuSH could prolong survival in mice where tumor was established. CT26-challenged mice with a palpable tumor (∼0.5 cm2) were injected intratumorally with leuSH at day 7, with a repeat dose given at day 10. Intratumoral administration of leuSH resulted in the attenuation of tumor growth, leading to prolonged survival in five of seven mice (p < 0.05; Fig. 6B, 6D). Tumor growth was arrested in three mice, with two showing a dramatic reduction in tumor size (Fig. 6D). In all instances of tumor growth arrest, the tumor size remained stable throughout the remainder of the experiment, again indicating the presence of a tumoristatic T cell response. Control mice injected with DTT alone showed no retardation of tumor growth (Fig. 6D). Attenuation of tumor growth in most mice after inhibition of ERAAP function provides proof of principle that small molecule inhibitors of ERAAP might have therapeutic value, particularly for tumors in which high expression correlates with poorer prognosis.

We used ERAAP silencing to induce a robust protective antitumor immunity to CT26 and show that this immunity is CD8+ T cell mediated, which also protects against challenge with wild-type CT26. Although it is not known whether modulation of ERAAP expression or activity leads to immune-evasion by human tumors, it is interesting to note that ERAP1/2 expression in human cancers is variable; one study (15) showed that, unlike ovarian, breast, and lung cancers, which tended to downregulate ERAAP expression, colorectal adenocarcinoma showed enhanced expression relative to normal gut tissue. This observation is consistent with the model that we investigated in which a potent tumor epitope is overprocessed by ERAAP, rendering it cryptic. Downregulation of ERAAP in CT26 led to the increased generation of protective epitopes and, in particular, the cryptic epitope GSW11, leading to strong antitumor CTL and tumor rejection. Importantly, ERAAPlo CT26–immune mice were also able to reject CT26 when challenged 50 d after rejection of the primary tumor. We also showed that transient ERAAP inhibition with leuSH attenuated tumor growth in established tumors, demonstrating the potential of this strategy as a therapy.

ERAAP-deficient cells were shown to present peptides with N-terminal extensions, as well as novel peptides uncovered in the absence of ERAAP (21, 25, 26). Interestingly, some peptide epitopes were overtrimmed by ERAAP, highlighting its role as a peptide editor (12). The increased T cell responses, together with the 30-fold increase in peptide generation when ERAAP is downregulated/inhibited, strongly indicate that the potent antitumor epitope GSW11 is overtrimmed. The precise mechanism for peptide trimming by ERAAP is not known; however, from other examples of aminopeptidases, it is likely that it trims a single amino acid at a time in a bind, trim, release mechanism. This suggests that ERAAP must compete for peptide binding with other molecules, such as MHC I, and this competition would ultimately determine whether a peptide is overtrimmed; GSW11 has a poor affinity for H-2Dd (its half-life for dissociation is only 20 min), which might bias competition toward ERAAP binding over MHC I, thereby providing more opportunity for ERAAP to bind and overtrim GSW11.

The induction of anti-GSW11 CD8+ T cell responses is fundamental to the clearance of CT26 tumor. In naive mice challenged with CT26, responses to GSW11 are absent, but we showed that GSW11-specific T cell responses are observed following Treg depletion prior to challenge, revealing that Tregs suppress these T cells (18). Downregulation of ERAAP in CT26 (ERAAPlo CT26) led to an increased abundance of GSW11 and the subsequent induction of GSW11-specific T cell responses similar to those observed in CT26-challenged Treg-depleted mice. This suggests that increasing the peptide abundance in CT26 tumors facilitates effective priming and activation of GSW11-specific T cells, thus overcoming Treg suppression. There are many mechanisms for Treg suppression, with strategies to overcome the effects of Tregs on protective T cell responses based on the depletion/blocking of important cells or interactions, such as Treg depletion (19), CTLA-4 (27), PD-L1 (28). These approaches, although showing promise, rely on the modulation of all immune responses, not just those restricted to the tumor, and were shown to cause undesirable autoimmune responses. Therefore, the use of ERAAP downregulation may provide a new approach by which Treg suppression of CTL responses can be overcome.

Alteration of ERAAP activity was shown to induce immune responses to tumors (mediated by NK cells) (29) and ERAAP-deficient cells (mediated by CD8+ T cells restricted by MHC Ia and Ib) (25, 26), but a link between ERAAP expression/function and immune control of human tumors has yet to be made. There is growing evidence for the importance of tumor-infiltrating lymphocytes (including CTL) in controlling the progression of tumors, including colorectal cancer (30, 31). Furthermore, in human colon cancer there is a negative correlation between tumor survival and tumor-infiltrating Tregs, which adds further significance to our observation that ERAAP downregulation allows anti-GSW11 CTLs to overcome epitope-specific immunosuppression by Tregs (32). The same is likely to be true for other cryptic epitopes that arise as a result of ERAAP silencing, which, for other Ags, were shown to be highly immunogenic (21, 25). Therefore, it is probable that these Ags play a role in the rejection of ERAAP-silenced tumor cells (gene knockdown or leuSH treated). However, these Ags, like GSW11, would also have to be presented, albeit at a lower level, on ERAAP-competent tumor cells to account for the robust memory response against unmanipulated CT26 cells in rechallenge experiments. CD8+ T cell responses directed against GSW11 are important for CT26 clearance in ERAAPlo CT26–immune mice; however, this may be indicative of other similar epitopes because anti-CT26 CD8+ T cells of other specificities (albeit subdominant) were identified (18).

We found that molecular inhibition of ERAAP in CT26 by preincubation with leuSH, which induces a transient loss of ERAAP activity lasting <48 h (Supplemental Fig. 2), also induced antitumor immunity. This indicated the possibility of using topical ERAAP inhibitors in vivo in a therapeutic setting, and we found that intratumoral injection of established and palpable tumors with leuSH attenuated their growth in most instances. LeuSH is not an attractive therapeutic candidate for many reasons, including its lack of specificity, redox number (requiring coadministration of a reducing agent), and high Kd (∼50 μM), which do not justify the further pharmacokinetic studies and dosing trials that would be needed to optimize these therapy experiments. Nevertheless, its clear effect on immune control of tumor growth in the majority of mice and a corresponding increase in survival indicate that more specific, potent inhibitors of ERAAP may have significant therapeutic potential. Furthermore, even in cases in which in vivo leuSH treatment did not lead to long-term survival (5/7 cases), such treatment may prolong the “equilibrium phase” of tumor growth (33) during which progression of an established tumor is kept in check via natural immunity. In turn, this may widen a window of opportunity for adjunct CTL-targeted therapies (e.g., vaccination) or immunomodulation (e.g., anti–CTLA-4, Treg ablation). In this way, small molecule inhibitors of ERAAP might be delivered with biological immunotherapeutics in the future, taking their place among a growing armory of anticancer combination therapies.

We thank Nasia Kontouli for technical expertise and Denise Boulanger for helpful discussions.

This work was supported by Cancer Research UK Project Grant C25722/A8410 (to E.J. and T.E.) and Cancer Research UK Programme Grant C7056/A11946 (to T.E.). G.S. was awarded a Cancer Research UK Ph.D. studentship.

The online version of this article contains supplemental material.

Abbreviations used in this article:

leuSH

leucinethiol

MHC I

MHC class I

shRNA

short hairpin RNA

siRNA

small interfering RNA

TPPII

tripeptidyl peptidase II

Treg

regulatory T cell.

1
Kloetzel
P. M.
2004
.
Generation of major histocompatibility complex class I antigens: functional interplay between proteasomes and TPPII.
Nat. Immunol.
5
:
661
669
.
2
Shastri
N.
,
Schwab
S.
,
Serwold
T.
.
2002
.
Producing nature’s gene-chips: the generation of peptides for display by MHC class I molecules.
Annu. Rev. Immunol.
20
:
463
493
.
3
Cresswell
P.
,
Ackerman
A. L.
,
Giodini
A.
,
Peaper
D. R.
,
Wearsch
P. A.
.
2005
.
Mechanisms of MHC class I-restricted antigen processing and cross-presentation.
Immunol. Rev.
207
:
145
157
.
4
Garbi
N.
,
Hämmerling
G.
,
Tanaka
S.
.
2007
.
Interaction of ERp57 and tapasin in the generation of MHC class I-peptide complexes.
Curr. Opin. Immunol.
19
:
99
105
.
5
Rock
K. L.
,
York
I. A.
,
Goldberg
A. L.
.
2004
.
Post-proteasomal antigen processing for major histocompatibility complex class I presentation.
Nat. Immunol.
5
:
670
677
.
6
Serwold
T.
,
Gaw
S.
,
Shastri
N.
.
2001
.
ER aminopeptidases generate a unique pool of peptides for MHC class I molecules.
Nat. Immunol.
2
:
644
651
.
7
Saric
T.
,
Chang
S. C.
,
Hattori
A.
,
York
I. A.
,
Markant
S.
,
Rock
K. L.
,
Tsujimoto
M.
,
Goldberg
A. L.
.
2002
.
An IFN-gamma-induced aminopeptidase in the ER, ERAP1, trims precursors to MHC class I-presented peptides.
Nat. Immunol.
3
:
1169
1176
.
8
Saveanu
L.
,
Carroll
O.
,
Lindo
V.
,
Del Val
M.
,
Lopez
D.
,
Lepelletier
Y.
,
Greer
F.
,
Schomburg
L.
,
Fruci
D.
,
Niedermann
G.
,
van Endert
P. M.
.
2005
.
Concerted peptide trimming by human ERAP1 and ERAP2 aminopeptidase complexes in the endoplasmic reticulum.
Nat. Immunol.
6
:
689
697
.
9
York
I. A.
,
Chang
S. C.
,
Saric
T.
,
Keys
J. A.
,
Favreau
J. M.
,
Goldberg
A. L.
,
Rock
K. L.
.
2002
.
The ER aminopeptidase ERAP1 enhances or limits antigen presentation by trimming epitopes to 8-9 residues.
Nat. Immunol.
3
:
1177
1184
.
10
Serwold
T.
,
Gonzalez
F.
,
Kim
J.
,
Jacob
R.
,
Shastri
N.
.
2002
.
ERAAP customizes peptides for MHC class I molecules in the endoplasmic reticulum.
Nature
419
:
480
483
.
11
Firat
E.
,
Saveanu
L.
,
Aichele
P.
,
Staeheli
P.
,
Huai
J.
,
Gaedicke
S.
,
Nil
A.
,
Besin
G.
,
Kanzler
B.
,
van Endert
P.
,
Niedermann
G.
.
2007
.
The role of endoplasmic reticulum-associated aminopeptidase 1 in immunity to infection and in cross-presentation.
J. Immunol.
178
:
2241
2248
.
12
Hammer
G. E.
,
Gonzalez
F.
,
Champsaur
M.
,
Cado
D.
,
Shastri
N.
.
2006
.
The aminopeptidase ERAAP shapes the peptide repertoire displayed by major histocompatibility complex class I molecules.
Nat. Immunol.
7
:
103
112
.
13
Yan
J.
,
Parekh
V. V.
,
Mendez-Fernandez
Y.
,
Olivares-Villagómez
D.
,
Dragovic
S.
,
Hill
T.
,
Roopenian
D. C.
,
Joyce
S.
,
Van Kaer
L.
.
2006
.
In vivo role of ER-associated peptidase activity in tailoring peptides for presentation by MHC class Ia and class Ib molecules.
J. Exp. Med.
203
:
647
659
.
14
York
I. A.
,
Brehm
M. A.
,
Zendzian
S.
,
Towne
C. F.
,
Rock
K. L.
.
2006
.
Endoplasmic reticulum aminopeptidase 1 (ERAP1) trims MHC class I-presented peptides in vivo and plays an important role in immunodominance.
Proc. Natl. Acad. Sci. USA
103
:
9202
9207
.
15
Fruci
D.
,
Ferracuti
S.
,
Limongi
M. Z.
,
Cunsolo
V.
,
Giorda
E.
,
Fraioli
R.
,
Sibilio
L.
,
Carroll
O.
,
Hattori
A.
,
van Endert
P. M.
,
Giacomini
P.
.
2006
.
Expression of endoplasmic reticulum aminopeptidases in EBV-B cell lines from healthy donors and in leukemia/lymphoma, carcinoma, and melanoma cell lines.
J. Immunol.
176
:
4869
4879
.
16
Fruci
D.
,
Giacomini
P.
,
Nicotra
M. R.
,
Forloni
M.
,
Fraioli
R.
,
Saveanu
L.
,
van Endert
P.
,
Natali
P. G.
.
2008
.
Altered expression of endoplasmic reticulum aminopeptidases ERAP1 and ERAP2 in transformed non-lymphoid human tissues.
J. Cell. Physiol.
216
:
742
749
.
17
Kamphausen
E.
,
Kellert
C.
,
Abbas
T.
,
Akkad
N.
,
Tenzer
S.
,
Pawelec
G.
,
Schild
H.
,
van Endert
P.
,
Seliger
B.
.
2010
.
Distinct molecular mechanisms leading to deficient expression of ER-resident aminopeptidases in melanoma.
Cancer Immunol. Immunother.
59
:
1273
1284
.
18
James
E.
,
Yeh
A.
,
King
C.
,
Korangy
F.
,
Bailey
I.
,
Boulanger
D. S.
,
Van den Eynde
B. J.
,
Murray
N.
,
Elliott
T. J.
.
2010
.
Differential suppression of tumor-specific CD8+ T cells by regulatory T cells.
J. Immunol.
185
:
5048
5055
.
19
Golgher
D.
,
Jones
E.
,
Powrie
F.
,
Elliott
T.
,
Gallimore
A.
.
2002
.
Depletion of CD25+ regulatory cells uncovers immune responses to shared murine tumor rejection antigens.
Eur. J. Immunol.
32
:
3267
3275
.
20
Kanaseki
T.
,
Shastri
N.
.
2008
.
Endoplasmic reticulum aminopeptidase associated with antigen processing regulates quality of processed peptides presented by MHC class I molecules.
J. Immunol.
181
:
6275
6282
.
21
Blanchard
N.
,
Kanaseki
T.
,
Escobar
H.
,
Delebecque
F.
,
Nagarajan
N. A.
,
Reyes-Vargas
E.
,
Crockett
D. K.
,
Raulet
D. H.
,
Delgado
J. C.
,
Shastri
N.
.
2010
.
Endoplasmic reticulum aminopeptidase associated with antigen processing defines the composition and structure of MHC class I peptide repertoire in normal and virus-infected cells.
J. Immunol.
184
:
3033
3042
.
22
Früh
K.
,
Ahn
K.
,
Djaballah
H.
,
Sempé
P.
,
van Endert
P. M.
,
Tampé
R.
,
Peterson
P. A.
,
Yang
Y.
.
1995
.
A viral inhibitor of peptide transporters for antigen presentation.
Nature
375
:
415
418
.
23
Hill
A.
,
Jugovic
P.
,
York
I.
,
Russ
G.
,
Bennink
J.
,
Yewdell
J.
,
Ploegh
H.
,
Johnson
D.
.
1995
.
Herpes simplex virus turns off the TAP to evade host immunity.
Nature
375
:
411
415
.
24
Sanderson
S.
,
Shastri
N.
.
1994
.
LacZ inducible, antigen/MHC-specific T cell hybrids.
Int. Immunol.
6
:
369
376
.
25
Hammer
G. E.
,
Gonzalez
F.
,
James
E.
,
Nolla
H.
,
Shastri
N.
.
2007
.
In the absence of aminopeptidase ERAAP, MHC class I molecules present many unstable and highly immunogenic peptides.
Nat. Immunol.
8
:
101
108
.
26
Nagarajan
N. A.
,
Gonzalez
F.
,
Shastri
N.
.
2012
.
Nonclassical MHC class Ib-restricted cytotoxic T cells monitor antigen processing in the endoplasmic reticulum.
Nat. Immunol.
13
:
579
586
.
27
Read
S.
,
Malmström
V.
,
Powrie
F.
.
2000
.
Cytotoxic T lymphocyte-associated antigen 4 plays an essential role in the function of CD25(+)CD4(+) regulatory cells that control intestinal inflammation.
J. Exp. Med.
192
:
295
302
.
28
Curran
M. A.
,
Montalvo
W.
,
Yagita
H.
,
Allison
J. P.
.
2010
.
PD-1 and CTLA-4 combination blockade expands infiltrating T cells and reduces regulatory T and myeloid cells within B16 melanoma tumors.
Proc. Natl. Acad. Sci. USA
107
:
4275
4280
.
29
Cifaldi
L.
,
Lo Monaco
E.
,
Forloni
M.
,
Giorda
E.
,
Lorenzi
S.
,
Petrini
S.
,
Tremante
E.
,
Pende
D.
,
Locatelli
F.
,
Giacomini
P.
,
Fruci
D.
.
2011
.
Natural killer cells efficiently reject lymphoma silenced for the endoplasmic reticulum aminopeptidase associated with antigen processing.
Cancer Res.
71
:
1597
1606
.
30
Pagès
F.
,
Berger
A.
,
Camus
M.
,
Sanchez-Cabo
F.
,
Costes
A.
,
Molidor
R.
,
Mlecnik
B.
,
Kirilovsky
A.
,
Nilsson
M.
,
Damotte
D.
, et al
.
2005
.
Effector memory T cells, early metastasis, and survival in colorectal cancer.
N. Engl. J. Med.
353
:
2654
2666
.
31
Ropponen
K. M.
,
Eskelinen
M. J.
,
Lipponen
P. K.
,
Alhava
E.
,
Kosma
V. M.
.
1997
.
Prognostic value of tumour-infiltrating lymphocytes (TILs) in colorectal cancer.
J. Pathol.
182
:
318
324
.
32
Salama, P., M. Phillips, F. Grieu, M. Morris, N. Zeps, D. Joseph, C. Platell, and B. Iacopetta. 2009. Tumor-infiltrating FOXP3+ T regulatory cells show strong prognostic significance in colorectal cancer. J. Clin. Oncol. 27: 186–192
.
33
Koebel
C. M.
,
Vermi
W.
,
Swann
J. B.
,
Zerafa
N.
,
Rodig
S. J.
,
Old
L. J.
,
Smyth
M. J.
,
Schreiber
R. D.
.
2007
.
Adaptive immunity maintains occult cancer in an equilibrium state.
Nature
450
:
903
907
.

The authors have no financial conflicts of interest.