ERAP2 is often seen in the shadow of its more famous and clinically progressed cousin (ERAP1). But could this novel target rival its counterpart’s clinical success?
T cells are wired to recognize non-self-peptides, but their specificity is imperfect. They scan short protein fragments on cell surfaces and respond to anything that deviates from ‘self’ to trigger their receptors. If the match crosses that threshold, the cell displaying it becomes the target for killing.
Endoplasmic reticulum aminopeptidase 2 (ERAP2) determines which peptides are displayed in that first encounter. It trims antigenic peptides before they are loaded onto MHC-I molecules to shape around 30% of the immunopeptidome available to CD8⁺ T cells. In effect, ERAP2 holds influence over what the immune system gets to see.
That power cuts both ways. In autoimmune disease, ERAP2 generates peptides that drive CD8⁺ T cells to attack healthy tissue. In cancer, it can edit away tumor peptides that might otherwise flag malignant cells for destruction.
Greywolf Therapeutics is building a pioneering program focused on selective ERAP2 inhibition. The goal is two-fold: block autoimmune T cell attacks by preventing presentation of disease-associated self-peptides and enhance anti-tumor immunity by unmasking neoantigens normally edited out by ERAP2’s trimming preferences.
“ERAP2 is one of the last untouched levers in antigen modulation” says Kris Clark, PhD, Greywolf’s Head of Biology. “We now have the structural tools, the biochemical understanding, and the genetic validation to go after it in a way no one could before.”
ERAP2: the missing half of antigen modulation
“If ERAP1 was the first antigen-processing enzyme to gain drug-development traction, ERAP2 is the non-redundant brother. The two share close to 50% sequence identity but cut very different peptide substrates,” says Kris.
ERAP1 trims peptides down to 9 amino acids in length, preferring longer peptides with hydrophobic residues (leucine). ERAP2, by contrast, prefers peptides with basic residues (lysine or arginine) at the N-terminus. This means that ERAP1 and ERAP2 edit different slices of the antigen repertoire.
Genetic studies make the case that this slice matters. Homozygous loss-of-function ERAP2 alleles are found in about 25% of Europeans. They are under balancing selection, with functional and non-functional alleles persisting because each offers survival advantages in different contexts.
“For years, ERAP2 has proven a difficult nut to crack.” Kris continued.
ERAP2's active site is dynamic, shifting shape during catalysis, meaning early screening attempts returned either non-specific hits or compounds too weak to matter in cells (3). Light appeared at the end of the tunnel after Greywolf's hit finding campaign identified a new starting point for which the binding could be rationalized at the structural level.
Medicinal chemistry followed. Greywolf’s drug candidates have high potency, high selectivity over ERAP1, and clean off-target panels.
Immunopeptidomics revealed ERAP2’s distinctive antigen signature– the over-representation of basic-residue peptides in its absence (4–6) – and under-representation of basic-residue epitopes when human ERAP2 was expressed in our knock-in mouse models. In cell assays, inhibiting ERAP2 produces the expected enrichment of lysine/arginine N-terminal peptides.
Repelling autoimmune T cell attacks
In many MHC-I–linked autoimmune diseases, the problem is less about a failure to detect threats, but rather the overexuberance of the immune system to detect self-peptides closely related to non-self (malignancy or infection). ERAP2 contributes to this error by generating self-peptides that mimic foreign antigens closely enough to trigger CD8⁺ T cells.
“Blocking ERAP2 alters the first step in CD8⁺ T-cell activation (Signal 1) by reshaping the HLA class I antigenic repertoire via peptide trimming. Without ERAP2 many low-abundance self-antigens that are detected by T cells by cross-reactive TCRs after an infectious trigger are no longer generated and therefore do not reach the cell surface” says Samantha Bucktrout PhD, Greywolf’s Head of Research.
HLA-I presentation continues, but the autoimmune causing peptide–HLA-I complexes are replaced by alternative self-peptides that have been encountered during central tolerance and not associated with the initial trigger event. In effect, ERAP2 inhibition removes the antigenic stimulus that sustains autoimmune responses while preserving recognition of infection- and tumor-derived peptides.
“We’re not focused on broadly switching off immunity” says Sam. “Instead, we want to remove the specific antigens that drive harmful responses but leave the rest of the immune system operational.”
Multiple genome-wide association studies link ERAP2 polymorphisms to autoimmune risk, and these variants affect the presence and functionality of ERAP2. Individuals with non-functional or reduced expression ERAP2 variants are protected from developing certain autoimmune diseases, even if they have all the other risk factors: genetic associations other than ERAP2, and infectious triggers associated with autoimmunity. This is the best evidence for blocking an active ERAP2 that nature can provide.
Primary biliary cholangitis (PBC): High ERAP2 and HLA-C*04:01 combine to produce a liver-targeting peptide repertoire; inhibition could reduce pathogenic CD8⁺ T cell responses. The underlying, immune-mediated driving component of PBC has yet to be addressed therapeutically, thus remains a large unmet need.
Crohn’s disease: GWAS point to HLA-B*27 and HLA-B*35 interactions, with high ERAP2 expression linked to disease risk.
Juvenile idiopathic arthritis (JIA): eQTL variants affecting ERAP2 expression alter disease susceptibility, with some alleles protective and others increasing risk.
Ankylosing spondylitis (AS): ERAP2 risk alleles are associated with disease in both B27⁺ and B27– patients. Mechanistic studies performed at GWT (our internal, confidential) support the genetics association data that ERAP-generated self-peptides activate pathogenic TCR clones.
Birdshot chorioretinopathy (BSCR): In HLA-A*29 carriers, high ERAP2 expression increases the abundance of a specific peptide motif that appears to drive ocular CD8⁺ infiltration.
Sam continues, “In each case, the mechanism is consistent: ERAP2 has been a determinant in presenting antigens that are causing aberrant T cell responses, sustaining healthy tissue destruction and a cycle of inflammation.”
An ERAP2 inhibitor could , in principle, do what no current autoimmune therapy can – remove the disease-driving antigens from the display case altogether. That is a fundamentally different approach from the systemic immunosuppression used today, which blunts protective immunity and raises infection risk.
“Because ERAP2 shapes only a subset of peptides, inhibition would leave much of antigen presentation untouched. Protective CD8⁺ responses to viruses and tumors should remain functional – a critical distinction for long-term safety.”
Greywolf’s development plan treats ERAP2 inhibition not as a stand-alone replacement for ERAP1 programs, but as the second arm of a complementary strategy. The first approach would be ERAP2 monotherapy. Here, Greywolf would target diseases where ERAP2 is the primary driver, such as Crohn’s, PBC, JIA and BSCR, and HLA-B*27 negative axSpA.
Turning tumors into visible targets
Where autoimmunity is a case of too much recognition, cancer is often the opposite – it’s not enough. Tumors can present so few peptides that trigger T cell responses on their surface that CD8⁺ T cells simply do not detect them or have become unresponsive to them (8).
“The activity of ERAP2 creates a collection of peptides that effectively compete for MHC-I binding: by favoring presentation of ERAP2-dependent 9mer peptides – most of which have been seen during T cell central tolerance – it displaces tumor-derived peptides with altered-self motifs that could otherwise trigger a response,” says Kris.
“Inhibiting ERAP2 reverses that edit. Peptides normally destroyed by its trimming preferences remain intact, bind MHC-I, and appear as new neoantigens for immune inspection.”
Genetic variability between patients will influence the impact of ERAP2 inhibition. Some HLA haplotypes prefer to present peptides with basic residues at position +1 – the same peptides ERAP2 removes. In patients with those HLA types, ERAP2 inhibition can restore the display of tumor-associated peptides that would otherwise be lost. But in patients whose HLA molecules do not bind those peptides, the antigen repertoire may remain largely unchanged.
Across common cancers – including bladder, breast, lung, melanoma, and colorectal – patients with low ERAP2 expression and ‘optimal’ HLA types have significantly better outcomes. The implication is that in these genetic contexts, losing ERAP2 exposes immunogenic antigens that the immune system can act on.
Broadening the antigen space for immunotherapy
Immune checkpoint inhibitors (ICIs) depend on having antigens for T cells to recognize. In tumors with a low mutational burden or limited antigen presentation, ICIs underperform because there is nothing for the revitalized T cells to target (9).
ERAP2 inhibition can change that baseline. By revealing a set of antigens that would otherwise be trimmed away, it could increase the diversity and novelty of targets. In preclinical immunopeptidomics, ERAP2 knockdown or inhibition enriched the MHC-I pool for peptides with N-terminal basic residues, consistent across multiple tumor cell lines.
The expected clinical pattern mirrors ERAP1 inhibition in our ongoing oncology studies:
Immediate pharmacodynamic signature: shift in antigen composition.
Novel T cell responses: expansion of TCR clonotypes recognizing newly presented neoantigens.
Potential synergy with ICIs: more antigenic ‘fuel’ for checkpoint-released T cells to act on.
From concept to clinic
“The translational plan is staged. Autoimmune indications such as PBC and Crohn’s or PBC, with well-defined genetic drivers and no curative therapies, are the most direct route to proof of concept. Oncology trials will start with biomarker-enriched cohorts to maximize early efficacy signals,” says Kris.
ERAP2 has moved from an academic curiosity to a translational target because the assets now exist to inhibit it selectively and measure the effects. Greywolf’s program is built on three core technical assets.
1. Structure-guided chemistry has delivered ERAP2 inhibitors with high potency and selectivity.
2. ERAP2 inhibition alters the composition of peptides presented on MHC-I, and this peptidome shift provides evidence of target engagement. Similar changes have been used as proof-of-mechanism in the EMITT-1 trial with ERAP1 (10, 11).
3. The development plan prioritizes autoimmune indications with well-defined genetic drivers and limited treatment options, followed by oncology trials in biomarker-selected cohorts.
Where could ERAP2 inhibition take antigen modulation next?
If ERAP2 inhibition delivers in the clinic, it will open a complimentary dimension in controlling immune visibility. The field has spent the last decade refining therapies that boost or block T cell activation (Signals 2 and 3). ERAP2, like ERAP1, sits upstream at the initial detection event (Signal 1) – deciding which antigens T cells get the chance to see.
In autoimmunity, that means the possibility of deleting disease-causing peptides from the antigen pool entirely, shutting down pathogenic CD8⁺ responses while leaving antiviral and anticancer immunity untouched. In oncology, it means unveiling a previously hidden set of neoantigens to which the immune system can mount a fresh attack.
“We see ERAP2 as a precision strike” says Samantha Bucktrout. “In the right genetic and disease context, it allows us to rewrite the set of antigens in play. This is a fundamentally different way to treat immune-driven disease.”
The combination of ERAP1 and ERAP2 inhibitors offers a modular way to reshape the MHC-I repertoire. Broader sets of peptides are removed and revealed than with singular treatments. Together, they could address immune misdirection in autoimmunity and immune evasion in cancer through non-overlapping mechanisms.
ERAP2 inhibition is supported by strong genetic work: knockout variants and reduced expression variants are highly protective against autoimmune disease and improve cancer outcomes in patients with permissive HLA types. Structure-guided chemistry has now made this target tractable, and the next phase of work will focus on advancing selective inhibitors with defined antigen modulation effects into the clinic.
References
1. J. Klunk, T. P. Vilgalys, C. E. Demeure, X. Cheng, M. Shiratori, J. Madej, R. Beau, D. Elli, M. I. Patino, R. Redfern, S. N. DeWitte, J. A. Gamble, J. L. Boldsen, A. Carmichael, N. Varlik, K. Eaton, J.-C. Grenier, G. B. Golding, A. Devault, J.-M. Rouillard, V. Yotova, R. Sindeaux, C. J. Ye, M. Bikaran, A. Dumaine, J. F. Brinkworth, D. Missiakas, G. A. Rouleau, M. Steinrücken, J. Pizarro-Cerdá, H. N. Poinar, L. B. Barreiro, Evolution of immune genes is associated with the Black Death. Nature 611, 312–319 (2022).
2. M. Al-kaabi, P. Deshpande, M. Firth, R. Pavlos, A. Chopra, H. Basiri, J. Currenti, E. Alves, S. Kalams, J. Fellay, E. Phillips, S. Mallal, M. John, S. Gaudieri, Epistatic interaction between ERAP2 and HLA modulates HIV-1 adaptation and disease outcome in an Australian population. PLOS Pathog. 20, e1012359 (2024).
3. M. Laura, G. Ronan, L. B. Vy, G. Valentin, C.-A. Omar, C. Virgyl, C. Piveteau, R. Melissa, F. Charlotte, W. Sandrine, C. Julie, D.-R. Julie, B. Damien, L. Florence, van E. Peter, D. Benoit, D.-P. Rebecca, Modulators of hERAP2 discovered by high-throughput screening. Eur. J. Med. Chem. 211, 113053 (2021).
4. J. A. L. de Castro, How ERAP1 and ERAP2 Shape the Peptidomes of Disease-Associated MHC-I Proteins. Front. Immunol. 9, 2463 (2018).
5. A. M. Andrés, M. Y. Dennis, W. W. Kretzschmar, J. L. Cannons, S.-Q. Lee-Lin, B. Hurle, N. C. S. Program, P. L. Schwartzberg, S. H. Williamson, C. D. Bustamante, R. Nielsen, A. G. Clark, E. D. Green, Balancing Selection Maintains a Form of ERAP2 that Undergoes Nonsense-Mediated Decay and Affects Antigen Presentation. PLoS Genet. 6, e1001157 (2010).
6. L. Saveanu, O. Carroll, V. Lindo, M. D. Val, D. Lopez, Y. Lepelletier, F. Greer, L. Schomburg, D. Fruci, G. Niedermann, P. M. van Endert, Concerted peptide trimming by human ERAP1 and ERAP2 aminopeptidase complexes in the endoplasmic reticulum. Nat. Immunol. 6, 689–697 (2005).
7. E. Lorente, M. G. Fontela, E. Barnea, A. J. Martín-Galiano, C. Mir, B. Galocha, A. Admon, P. Lauzurica, D. López, Modulation of Natural HLA-B*27:05 Ligandome by Ankylosing Spondylitis-associated Endoplasmic Reticulum Aminopeptidase 2 (ERAP2)*. Mol. Cell. Proteom. 19, 994–1004 (2020).
8. I. Temponeras, G. Stamatakis, M. Samiotaki, D. Georgiadis, H. Pratsinis, G. Panayotou, E. Stratikos, ERAP2 Inhibition Induces Cell-Surface Presentation by MOLT-4 Leukemia Cancer Cells of Many Novel and Potentially Antigenic Peptides. Int. J. Mol. Sci. 23, 1913 (2022).
9. M. Lee, R. M. Samstein, C. Valero, T. A. Chan, L. G. T. Morris, Tumor mutational burden as a predictive biomarker for checkpoint inhibitor immunotherapy. Hum. Vaccines Immunother. 16, 112–115 (2020).
10. T. Lillie, G. Kichenadasse, J. (Jenny) Liu, T. H. Guerrero, E. Calvo, H. Jakobsson, V. Moreno, H. K. Gan, P. Joyce, N. Hyland, M. Giovannetti, T. Palmer, C. McAlpine, D. Green, S. N. Symeonides, EMITT-1: Proof-of-mechanism immunopeptidome (ImPD) effects at target PK exposure, in a phase 1 study of GRWD5769 (a first-in-class inhibitor of Endoplasmic Reticulum Aminopeptidase 1 [ERAP1]) in patients with solid malignancies. JCO 42, 2589–2589 (2024).
11. NCT06923761 | EMITT-1 (ERAP Mediated Immunopeptidome Targeting Trial - 1) | ClinicalTrials.gov. https://clinicaltrials.gov/study/NCT06923761.
