Rheumatoid arthritis (RA) is a long-term autoimmune inflammatory disease affecting the joints, connective tissue and organs, manifesting in joint pain, swelling, stiffness, and loss of joint function (1, 2). Multiple factors influence RA development, including genetic predisposition and various environmental and lifestyle elements. It has been shown that women are more likely to suffer from RA, and the ratio of women to men is approximately 3:1 (3). The heredity of RA is around 65%, which emphasizes the importance of genetics (4). Among genetic factors, various genes have been associated with RA predisposition. Genetic association studies on several different populations have identified 150 candidate loci with polymorphisms connected to RA, primarily related to seropositive disease and HLA associations (5, 6).

During the autoimmune process, immune responses are triggered against peptides presented in the context of Major Histocompatibility Complex (MHC) class I or class II molecules, referred to as HLA (human leukocyte antigen) in humans. Although rheumatoid arthritis is primarily associated with HLA class II (7, 8), nevertheless some HLA class I associations have also been observed (9, 10), as supported by the association of some KIR genes (11–13) whose products, killer cell immunoglobulin-like receptors, recognize an absence or decrease in the expression of MHC class I molecules on the surface of target cells. MHC-I-bound peptides influence this interaction. KIR molecules are expressed on NK cells and subpopulations of T lymphocytes (14) which could be engaged in joint inflammation and destruction.

Antigenic peptides are formed and loaded on MHC class I molecules in the endoplasmic reticulum (ER) during a multistep process involving many proteins collectively named antigen processing and presentation machinery (APM) (15). Important elements of the APM are the endoplasmic reticulum aminopeptidases ERAP1 and ERAP2. ERAP1 and the closely related ERAP2 belong to the M1 family of zinc-metallopeptidase enzymes. Both aminopeptidases are IFN-γ and TNF-α inducible and ubiquitously expressed. They share ~ 49% sequence similarity and can form functional heterodimers (16–18). They are assigned to produce optimal-sized peptides for HLA class I binding by deleting amino acids from the unblocked N-terminus of peptides. Trimmed peptides with 8-10 amino acids can be directly presented by MHC class I molecules on the cell surface (19).

ERAP1 and ERAP2 single nucleotide polymorphisms (SNPs) affect both activity and specificity of the enzymes. Therefore, ERAP1 SNPs have been correlated with HLA-associated diseases like ankylosing spondylitis (AS) (20), psoriasis (21, 22), type 1 diabetes (T1D) (23), inflammatory bowel disease (24), Behçet’s disease (25), cervical cancer (26), and hypertension (27), suggesting that ERAP1 polymorphisms can exert diverse effects. Numerous literature data indicate that the ERAP1 and ERAP2 SNPs may not only affect the correct trimming and presentation of antigenic peptides to the respective receptors in the context of HLA class I, but also actively modulate peptide diversity (28, 29).

The aim of this work was to examine the role of ERAP1/ERAP2 polymorphisms in the susceptibility to RA and the severity of this disease. Moreover, we wanted to check the impact of the ERAP2 protein tested in the serum of patients with clinical manifestations. This work is the first report presenting the association of ERAP2-secreted protein with RA.

It’s known that RA is an inflammatory multi-factorial autoimmune disease that mainly harms small joints and progresses into erosive-destructive polyarthritis. The disease evolves in middle and older ages, primarily in women, and is characterized by painful and autoinflammatory status with prolonged disorders. Several genes are known, and some polymorphisms are associated with progression of RA (5, 36, 37). The ERAP1 and ERAP2 genes encode endoplasmic reticulum aminopeptidases, forming peptides that create antigens presented by HLA class I molecules. The polymorphism of these genes influences the diverse functions of aminopeptidases and affects the formation of HLA-antigen complexes (28, 33).

The human ERAP1 and ERAP2 genes are encoded in the short arm chromosome 5q15 in the opposite orientation, and they likely have two shared regulatory components. ERAP1 is highly polymorphic, with strong linkage disequilibrium apparent across the gene. Some of the single nucleotide polymorphisms (SNPs) present in ERAP1 may cause changes in the gene expression and alter the enzymatic activity and substrate binding (29, 38). The potential impact of some of these variants on ERAP1 biological functions can be inferred from their location in the protein structure. For example, rs27044 (Q730E) is exposed on the inner area of the C-terminal cavity, which could affect substrate sequence or length specificity. Other polymorphisms, such as rs26653 (R127P) and rs30187 (K528R) (alteration in the ligand’s entry pocket), could indirectly affect either substrate specificity or enzymatic activity by changing the conformational motives between the open and closed form of the enzyme (17). Rs7063 which is present in the middle of a conserved transcription termination sequence (AATAAA) in the 3`-UTR of ERAP1 19E isoform, may be involved in the transcription termination process (33). Finally, the rs26618 (Ile276Met) may affect ERAP1 enzymatic activity. Combinations of SNPs in the ERAP1 gene form 10 haplotypes of different activity (39, 40).

A significant functional difference in the rs30187 C/T (R528K) polymorphism of the ERAP1 gene diminishes the efficacy of peptide trimming by changing ERAP1 from an active to an inactive form (18). Here, the TT genotype was less common in Ukrainian patients with RA than in the control group, indicating this genotype’s protective role against RA. On the contrary, in Akbulut et al. (2022), the TT genotype was associated with RA disease risk in the Turkish population (32). The association of the rs30187 polymorphism with susceptibility to different diseases other than RA has been shown in many studies. An elevated disease risk for TT genotype-positive patients was observed in ankylosing spondylitis (AS) in the Polish population by Wisniewski et al. (41) and also in several other autoimmune diseases in different populations, Caucasian or Han Chinese (42).

In the case of ERAP1 rs27044 C > G (Q730E), we found the same pattern as in the case of rs30187 – the GG genotype was seen less frequently in patients with RA than in the control group, indicating the protective role of this genotype against RA. Rs27044 polymorphism is also significantly associated with an increased risk of AS, especially in some Caucasians, Han Chinese, and Koreans (43, 44), and spondyloarthritis in the French population, but not in the Belgian population (42).

We also found some other significant results concerning the ERAP1 gene: RA patients homozygotic for rs26618T were more often observed than in the control group, indicating an association with predisposition to the disease. The opposite was observed with the CT genotype demonstrating a protective effect. For cervical cancer, the CC genotype of rs26618 was a risk factor compared with the TT or CT genotype in the Chinese population (45) but not in the Dutch (46). We previously found an association of the rs26618*C allele and the rs26618*CC genotype with atopic dermatitis in the Polish population. In addition, as rs26618*C encodes the 276Met variant, we found that the enzyme with this substitution produces a peptide for the HLA-C*15 molecule, also associated with atopic dermatitis in the Polish population, with two times lower activity than the major variant, 276Ile (31).

It should be also mentioned that some ERAP1 polymorphisms tested in our study were associated with disease activity and other clinical manifestations such as the presence of anti-CCP and anti-MCV antibodies. We found the rs26653 GG genotype more frequent in patients with a mild stage of disease (DAS28 < 3.2) in comparison to more severe stages of disease. In addition, this genotype was more frequent in anti-CCP negative and anti-MCV negative patients. This indicates protection against disease progression.

We must underline that in our study, the distribution of all tested SNPs in controls was consistent with the HWE, indicating correct genotyping. According to the National Library of Medicine, the frequency of the T rs30187 minor allele in the Ukrainian population (36.9%) was only slightly different from the European population (35.3%) [https://www.ncbi.nlm.nih.gov/snp/rs30187]. The frequency of the G allele in the rs27044 polymorphism in the Ukrainian population was 31%, while it was 27% in other Caucasians.

How can we explain the differences observed in our research and others on the role of ERAP1 gene in rheumatic diseases, especially RA and AS? HLA class II alleles are generally connected to seropositive diseases: several autoimmune conditions (including coeliac disease, T1D, autoimmune thyroid disease, systemic lupus erythematosus (SLE), and others) are related to the HLA-DR3-DQ2 haplotype. RA is associated with different alleles of HLA-DRB1. By contrast, seronegative diseases are generally correlated with HLA class I alleles, which are disease-specific (i.e., AS is associated with HLA-B27 (47, 48) and psoriasis with HLA-Cw6 (49), etc.). In our study, the percentage of seropositive and negative patients was almost equal. It suggests that both pathomechanisms might be possible.

In contrast to ERAP1, ERAP2 shows limited polymorphism. The two major ERAP2 isoforms result from an alternative splicing in correspondence to the pivotal SNP rs2248374 (A/G) present within the 5′-splice site of exon 10. The ERAP2-A isoform (A allele) encodes a full-length and functional protein, whereas the ERAP2-B isoform (G allele) is degraded by nonsense-mediated decay (50, 51). Since both alleles of rs2248374 have similar frequency in most populations, only three-fourths of individuals [AA homozygotes (25%) and AG heterozygotes (50%)] express a functional ERAP2 isoform, while the remaining individuals (with GG genotype – occurring in ~25% of the population) express two undetectable isoforms of ERAP2-B, resulting in reduced MHC class I surface expression in lymphoblastoid cell lines (50). In addition, viral and also perhaps bacterial infections induce, in rs2248374G individuals, the expression of a third truncated transcript giving a shortened protein devoid of enzymatic activity but nevertheless possibly capable of interfering with ERAP1 and ERAP2 function (14, 34, 35). The polymorphism of the ERAP2 gene was associated with several immune-mediated diseases including AS (41, 52), psoriasis (21, 22) and preeclampsia (53). Interestingly, as ERAP1 and ERAP2 genes are closely linked on chromosome 5q15, the defective ERAP2 allele is frequently correlated to a highly active ERAP1 haplotype, and vice versa (40).

During our study, we observed a protective role of the ERAP2 rs2248374 AA genotype in RA susceptibility, but it was only seen in women. However, when analyzing the distribution of ERAP2 rs2248374 genotypes depending on disease activity, we found that the frequency of the AA genotype was highest in patients with severe disease compared to patients with mild disease. This rather points to the ERAP2 A haplotype and the production of full-length ERAP2 proteins as a risk for RA progression. This observation is consistent with other studies showing that the ERAP2 A haplotype increases the risk of autoimmune diseases such as Crohn’s disease, juvenile idiopathic arthritis (JIA), and BSCR (54). Additionally, in our study, RF-positive patients exhibited a significantly higher frequency of the GG genotype (no active ERAP2) than RF-negative which suggests that the lack of the functional enzyme may favor the production of RF. Thus, we may speculate that ERAP2 (encoded by rs2248374A allele) eliminates rheumatogenic peptide(s) including those stimulating RF production, but once the patient gets RA, then ERAP2 may produce peptides aggravating the disease or destroy protective peptides.

Moreover, we found that RA patients secreted higher concentrations of ERAP2 than controls. We also observed a linear increase in ERAP2 levels in patients according to DAS28, suggesting a role of the ERAP2 protein in the severity of RA. Additionally, ROC analysis indicated 5.85 ng/mL of ERAP2 as a point of severe RA development. It is known that pro-inflammatory cytokines such as IFNs and TNF-α significantly increase the expression of ERAP2 (55, 56). Chronic inflammation in RA supports increased expression and secretion of ERAP2, as observed in our study. Furthermore, additional SNPs influence ERAP2 expression, altering enhancer-promoter interactions and affecting expression (57). These SNPs deserve further study.

However, it should be emphasized that our study possesses some limitations. First, the number of patients and controls is too small for SNP analysis. Power calculations showed that to achieve 80% power, it would be necessary to test 2357 patients and 1892 controls for rs30187, 634 patients and 509 controls for rs27044, 1506 patients and 1209 controls for rs26653, 845 patients and 678 controls for rs26618, 7498 patients and 6020 controls for rs7063, and 3883 patients and 3118 controls for rs2248374. In turn, power test analysis for ERAP2 concentration showed that to achieve a result of 80% we needed 75 subjects in each group, which was met in our study. Moreover, a potential confounding factor in the study may be the fact that the control group was not fully matched in terms of age (p < 0.0002 in comparison to the patients) and sex (p < 0.0045 in comparison to the patients). However, age does not influence SNP genotype. Also, results in ERAP2 secretion were independent of sex. RA females and males possessed similar concentrations of ERAP2 ( Figure 1A ). The factors determining the secretion of ERAP2 were DAS28 ( Figure 1B ) and RF ( Figure 1C ).

In conclusion, we found genetic differences in ERAP1 and ERAP2 genes between RA patients and healthy controls. Additionally, the ERAP2 protein tested in serum could be a valuable biomarker of RA severity.