Inflammatory bowel diseases (IBD) are a group of chronic inflammatory conditions of the gastrointestinal tract, including ulcerative colitis (UC) and Crohn’s disease (CD). The precise etiology of IBDs remains unknown, though their relation to multiple and diverse genetic, immunological and environmental factors is accepted. Genome-wide association studies (GWAS) have identified immune-related genes associated to susceptibility and/or clinical manifestations that point to an inappropriate regulation of innate and/or adaptive immune responses in IBD (1). However, these polymorphisms alone do not account for IBD heritability, suggesting that other environmental, epigenetic and genetic factors, including rare variants, must be in place (1).

CD5 and CD6 are paralogs sharing homology in tissue expression patterns, structure and function (2–4). They encode signal-transducing surface co-receptors expressed on all T and B1a cells and involved in the fine tuning of intracellular activation signals delivered upon specific antigen recognition by lymphocyte’s clonotypic receptors (5). Both CD5 and CD6 receptors are composed of an extracellular region encompassing three tandem scavenger receptor cysteine-rich (SRCR) domain repeats, a transmembrane region, and a cytoplasmic region devoid of catalytic activity but well adapted for phosphorylation and association with downstream signaling effectors. Importantly, CD5 and CD6 are physically associated with the T cell receptor complex (TCR) with which co-localize at the center of the immunological synapse (6), providing inhibitory (CD5) and activating/inhibitory (CD6) signals (7). This is likely achieved through interaction with endogenous counter-receptors such as CD166/activated leukocyte cell adhesion molecule (ALCAM) (8), Galectins 1 and 3 (9), and CD318/CUB domain-containing protein 1 (CDCP-1) (10) for CD6, and still ill-defined ligands (CD72, IgVh framework, gp200, gp40-80, gp150, IL-6 and CD5 itself) for CD5 (11–18). Both molecules also act as pattern-recognition receptors (PRRs) for microbial-associated molecular patterns (MAMPs), where CD5 interacts with fungal (β-glucan) (19), viral (hepatitis C virus) (20), and parasitic (E. granulosus teguments) structures (21), while CD6 does it with bacterial (lipopolysaccharide, lipoteichoic acid and peptidoglycan) (22), viral (gp120 HIV-1) (23), and parasitic (E. granulosus tegument) structures (21). This dual role of CD5 and CD6 as both immunomodulatory and microbial PRR receptors is supported by pre-clinical models of infection, autoimmunity and cancer involving Cd5- and Cd6-deficient mouse lines, as well as infusion of wild-type mice with soluble CD5 and CD6 proteins (24–26).

To date, no CD5 or CD6 deficiencies have been reported in humans. However, functionally relevant single nucleotide polymorphisms (SNPs) of CD5 and CD6 have been identified, which act as susceptibility or disease modifier markers in autoimmune and neoplastic processes. Allelic combinations of the CD5 rs2241002 and rs2229177 SNPs resulting in hyper-reactivity to TCR stimulation are associated to more severe systemic lupus erythematosus (SLE) forms, but predict better prognosis in chronic lymphocytic leukemia (CLL) and melanoma (27–29). Moreover, GWAS have involved CD5 (rs2229177) in rheumatoid arthritis susceptibility (30). Regarding CD6, the rs12360861, rs17824933 and rs11230563 SNPs are revealed as disease modifiers in psoriasis, and as susceptibility markers in multiple sclerosis (MS) and Behçet’s disease (31–34). Also, GWAS and meta-analyses have associated the CD6 rs11230563 SNP to IBD susceptibility (35, 36). However, its role as a disease modifier in IBD, and the involvement of other neighboring SNPs from the CD6 and CD5 genes, and from the functionally related CD166/ALCAM gene, are still unknown.

The present work explores the consequence of CD5 and CD6 expression and variation in experimental and clinical IBD. To this end, we first analyzed the impact of Cd5 and Cd6 deficiency on dextran sulphate sodium (DSS)-induced colitis, an experimental model of human IBD (37). Subsequent clinical association studies assessed the impact of CD5 and CD6 variations on different clinically relevant manifestations and therapeutic requirements of CD and UC.

We provide experimental and clinical evidence for the involvement of CD5 and CD6 expression and variation in IBD. Previous GWAS and meta-analysis studies have identified the CD6 locus (SNP rs11230563) as a susceptibility marker in CD and UC, thus supporting its contribution to IBD etiopathogenesis (35, 36). Here, we used genetically modified mice and candidate gene-driven association analyses to clinical traits and prognosis with functionally relevant SNPs from the CD5 and CD6 paralogs, as well as from CD166/ALCAM.

Etiopathogenic factors for IBD include host genetic susceptibility, dysregulated immune response, intestinal dysbiosis, and impairment of intestinal epithelial barrier function. Under normal circumstances, there is continuous crosstalk between gut microbiota and the immune system, where gut microbiota modulates the host’s innate and adaptive immunity and vice versa (47). Gut microbiota is in close contact with the intestinal barrier, consisting of an epithelial cell layer and a variety of immune cells of hematopoietic origin. Cells from the intestinal barrier (both epithelial and hematopoietic) sense and signal the presence of microbial components via PRRs, which belong to different structural families such as lectin C-type, leucine-rich repeats (LRR), immunoglobulin (Ig), or scavenger receptor cysteine-rich (SRCR) domains (48). Both CD5 and CD6 are lymphocytic members of the SRCR superfamily, expressed by all T cells and the B1a subset responsible for production of polyreactive natural IgM antibodies (4, 49). CD5 and CD6 are also represented in certain immune cells subsets present in mucosal barriers such as regulatory T (Treg) and B (B1a, Breg) cells, certain macrophage and dendritic cells and innate lymphoid cells (NK, iNKT, ILCs) (50–52). CD166/ALCAM, the best characterized CD6 ligand, is also found in the gastrointestinal epithelial tract (53). Increased expression of both CD6 and CD166/ALCAM has been reported in inflamed mucosa from IBD patients, a fact that is attributed to higher CD6-expressing T cell infiltration rather than surface CD6 expression levels (54). This may relate to quantitative trait loci studies in which the rs11230584 SNP in the intergenic region between CD5 and CD6 modulates expression of both genes in IBD patients but not in healthy controls (55). Taken together, their tissue and cell expression pattern, microbial recognition properties and ability to modulate lymphocyte activation/differentiation and cell adhesion provide the basis for considering both CD5 and CD6 as contributors to IBD pathogenesis.

The observation that both CD5- and CD6-deficient mice differ in their response to DSS-induced colitis further supports their involvement in IBD. In Cd5 ^-/- mice, attenuated DSS-induced colitis was observed in agreement with a previous report (43). The mechanism underlying such attenuated colitis has already been explored and attributed to increased suppressive function of Treg cells from Cd5 ^-/- mice (43), a fact that was not confirmed by others (56). An alternative mechanism could be the increased activation-induced cell death (AICD) in Cd5 ^-/- effector T cells as a result of inhibitory role assigned to the CD5 receptor (7, 57). Further evidence for CD5 expression involvement in IBD comes from recent report showing that inducible Cd5-deficient mice in the autoimmune-prone non-obese diabetic (NOD) background undergo exacerbated DSS-induced colitis by modifying T cell effector function (58).

Regarding Cd6 ^-/- mice, no analysis of DSS-induced colitis has been brought forward, in spite of reports of Cd6 ^-/- mice behavior in several other immune-related inflammatory disease models (i.e., intestinal ischemia-reperfusion, bovine or avian type II collagen-induced arthritis, chronic graft-versus-host disease-induced lupus-like, imiquimod-induced psoriasis-like skin inflammation, experimental autoimmune encephalitis, and autoimmune uveitis) (10, 34, 39, 59–62). CD6 deficiency results in attenuated or exacerbated phenotypes according to mouse background and experimental models responsive to different underlying mechanisms (e.g., increased AICD or defective Treg function). This puzzling situation has been unveiled by CD6 receptor’s multitask signalosome with opposite functions in T cell activation (7). CD6 multifaceted role accounts for past difficulties in classifying it as a co-inhibitory or -stimulatory receptor.

Here we observed that Cd6 ^-/- mice exhibit an increased body weight loss and DAI upon DSS-colitis induction, in conjunction with differential expression of certain mRNA transcripts. This included decreased expression of Ifng —the prototypical Th1 cytokine— and Gata3 —the master regulator of Th2 differentiation—, no differences in Il17a and Il10 expression, and increased expression of Il6 and Cxcl1 —a cytokine and a chemokine involved in the Th17 function, all this pointing to a somehow misbalanced Th1/Th2/Th17 response. Additionally, reduced mRNA expression of Ncr1 —coding for NKp46, one of the NK triggering receptors— and a trend towards increased Lcn2 —coding for lipocalin-2, also named NGAL, a neutrophil secondary granule marker— in colons of Cd6 ^-/- mice undergoing DSS-induced colitis was observed. These findings would fit with the observation that i) decreased NK cell activity would lead to increased granulocyte infiltrate in DSS-induced colitis (63), and ii) increased lipocalin-2 expression would act as marker and a counter reactor of colonic inflammation (64, 65).

Given the multifaceted nature of CD6, the above-mentioned mechanistic findings for exacerbated symptoms during DSS colitis in Cd6 ^-/- mice do not exclude other possibilities such as decreased Treg functionality in Cd6 ^-/- mice (59), a cell subset known for its role in mucosal protection during DSS-induced colitis (66). Another possibility could be the deficient production of natural antibodies, as found in Cd6 ^-/- mice from DBA-1 background (60), and confirmed by us in the C57BL/6 background Cd6 ^-/- mice used here (67). Natural antibodies are an innate component of humoral immunity and have a protective role in IBD (68, 69).

In humans, no CD5 or CD6 deficiencies have been reported. Functionally relevant CD5 or CD6 SNPs previously described act as susceptibility or disease modifier markers for immune-related disorders. Regarding CD5, the rs2241002 and rs2229177 SNPs cause nonsynonymous substitutions at the extracellular (Pro224>Leu) and cytoplasmic (Ala471>Val) regions, respectively, which are relevant to CD5-mediated signal transduction (70). Thus, homozygous carriers of the ancestral Pro224-Ala471 (rs2241002^C and rs2229177^C) CD5 haplotype are hyper-reactive to TCR/CD3 cross-linking, and present more severe clinical forms of SLE (27) but better CLL and melanoma prognosis (28, 29). Regarding the functionality of CD6 SNPs, the intronic rs17824933^G allele identified as a susceptibility marker for MS causes over-expression of a CD6 isoform devoid of the CD166/ALCAM-binding domain (CD6Δd3) concomitant with diminished proliferation and long-term activation of CD4⁺ T cells (71, 72). Also, the MS-protective haplotype involving the CD6 rs11230563^C and rs2074225^C SNPs results in higher surface CD6 expression on several lymphocyte subsets (CD4⁺ and CD8⁺ naïve T, and NKT cells) (31).

In our genetic analysis the CD5 rs2241002 SNP, which causes a nonsynonymous substitution in the extracellular SRCR domain 2 of CD5 (Pro224>Leu), showed association with CD location. Further analyses showed association of CD5 haplotypes containing the cytoplasmic rs2229177^T variant with severity parameters in CD (requirement of biological treatments) and UC (poor prognosis) patients. The rs2229177^T variant involves the substitution of ancestral Ala471 for Val, which results in increased CD5 inhibitory capacity (27, 70). This can turn activated lymphocytes less sensitive to AICD and more damaging, thus making more intensive therapies necessary.

Analysis of CD6 SNPs showed association of the rs17824933^G allele with preferred ileal CD location and increased UC extent. These results consolidate the damaging effect of the rs17824933^G allele in inflammatory diseases, as suggested from its association with more aggressive forms of psoriasis and with MS susceptibility (34, 46). Patients with left-sided or extensive UC also tend to need more aggressive therapies and are at higher risk of developing colorectal cancer (73). A relatively short follow-up (median 12.37 years; Q1 7.43 years; Q3 19.21 years) may underlie the lack of significant differences observed for this SNP regarding prognosis ( Supplementary Table 1 ).

The CD6 rs17824933^G allele was further associated with lower risk of ankylosing spondylitis in the whole IBD cohort. This result appears to contradict the above-mentioned deleterious contribution of this variant in UC, as well as in psoriasis and MS. However, this variant also showed association with a more ileal location of CD. Articular extra-intestinal manifestations of IBD are more common in patients with colonic disease than in those with small-bowel disease (74). Thus, preferential ileal location in CD patients may account for the association of rs17824933^G with lower ankylosing spondylitis risk.

The study also showed association of the CD6 rs12360861 SNP with prognosis in CD patients but not susceptibility, in agreement with a major genetic contribution to prognosis from loci distinct from those driving disease susceptibility, applicable in this case (42).

As stated above, genetic susceptibility is only one of the known factors in IBD etiopathogenesis. The importance played by other environmental factors is illustrated by seasonal onset and exacerbation patterns in IBD patients (75). We have observed seasonal variations also for Cd6 ^-/- mice regarding susceptibility to DSS-induced colitis. More precisely, the exacerbated DSS-induced colitis phenotype of Cd6 ^-/- mice manifested during the spring/summer but not the autumn/winter season ( Figure 1C ), reminiscent of other mouse models of human diseases (i.e., EAE) (76). Though incompletely understood, it has been proposed that seasonal variations might be regulated by endogenous circannual rhythms, since they are found even when the animals are subjected to a constant, controlled environment, and genetically regulated (77). Seasonality in Cd5 ^-/- mice could be neither confirmed nor denied, since the two experiments performed were carried out in the summer season (July and September). However, recent unpublished results from a collaborative study show seasonality phenomena in Cd5^-/- mice upon mannan-induced psoriatic arthritis induction (Merino R and Merino J, University of Cantabria, Spain).

The main strengths of our study are the use of a large patient cohort, which together with the experimental mouse model highlights a role for CD5 and CD6 in IBD. We also acknowledge some caveats in our study. Separate breeding of Cd5 ^-/- or Cd6 ^-/- mice and their wild-type counterparts can be a source of confusion, which we minimized by periodic colony refresh, the use of high sample sizes, and a large number of repetitions. We measured mRNA expression as a proxy for protein expression, but correlation between mRNA and protein levels is limited. Therefore, further protein expression assays (e.g.: flow cytometry) will be needed to ascertain the molecular mechanisms underlying differences between Cd6 ^-/- and their wild-type counterparts. Similarly, further molecular mechanisms driving clinical SNP associations in IBD patients are to be identified.

In conclusion, our findings support a role for the CD5 and CD6 lymphocyte receptors in the pathophysiology of IBD and hint at their potential in patient stratification and as therapeutic targets. The latter is particularly valid for CD6, where the humanized anti-CD6 mAb Itolizumab currently represents a therapeutic option in several immune-mediated disorders (78), since could modulate the activity of the T cell subsets (i.e., Th1 and Th17) involved in their pathogeny (79).

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author/s

The studies involving human participants were reviewed and approved by Ethical Committee of Clinical Research of the Hospital Clínic de Barcelona. The patients/participants provided their written informed consent to participate in this study. The animal study was reviewed and approved by Comitè Ètic d’Experimentació Animal, Universitat de Barcelona, Spain.

Alfredo J. Lucendo, Gastroenterology Department, Hospital General de Tomelloso, IIS-IP, Ciudad Real, Spain, and Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd). Jordi Guardiola, Hospital Universitari de Bellvitge (IDIBELL), l’Hospitalet de Llobregat, Spain. Xavier Calvet, Servei d’Aparell Digestiu, Hospital Universitari Parc Taulí, Departament de Medicina, Universitat Autònoma de Barcelona, Sabadell, Spain, and CIBERehd. Lorenzo Oliván, Hospital San Jorge, Huesca, Spain. Marta Piqueras, Gastroenterology Department, Consorci Sanitari de Terrassa, Barcelona, Spain.

Conceptualization: FL, AS, SC-L, and JP. Mouse studies: SC-L, MV-dA, CC, NA, AL-P, RG-C, EC, MiE, JL, KA, JHN, PE, and AS. Genetic studies: SC-L, MV-dA, CC, NA, AL-P, BS, EC, and JL. Sample and clinical information collection: ER, IO, JPG, MaE, LM, DB, EI, EG-P, MDM-A, AJL, JG, XC, LO, MP, and JP. Statistical analyses and figures: SC-L, EC, and JL. Writing original draft: SC-L and FL. All authors read, critically revised and approved the final version of the manuscript.

This work was supported by Spanish Ministerio de Economía y Competitividad (MINECO, SAF2016-80535-R) and Ministerio de Ciencia e Innovación (MCIN/AEI/10.13039/501100011033, PID2019-106658RB-I00), co-financed by European Development Regional Fund “A way to achieve Europe” ERDF, and Agència de Gestió d’Ajuts Universitaris i de Recerca from Generalitat de Catalunya (2017/SGR/1582). SC-L, MV-dA, CC, AL-P, and EC are recipients of fellowships from Spanish Ministerio de Educación, Cultura y Deporte (FPU15/02897), Spanish MINECO (BES-2014-069237 and BES-2017-082107), Chilean Agencia Nacional de Investigación y Desarrollo (2018-72190154), and European Community Seventh Framework Program (FP7/2007/2013; 229673), respectively. SC-L and JL are recipients of short-term fellowships from European Federation of Immunological Societies-Immunology Letters (EFIS-IL) and Erasmus+ from the European Union, respectively. The ENEIDA registry of GETECCU is supported by Biogen, Janssen, Takeda and Pfizer. The funders were not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication.

We thank Belchin Kostov for statistical analysis support, Marcos Isamat for manuscript editing and critical comments, and Silvia Ariño for technical help with RT-PCR and immunohistochemistry quantification and interpretation. We are indebted to GETECCU and the IDIBAPS Biobank for clinical data and sample procurement.

FL is founder and ad-honorem scientific advisor of Sepsia Therapeutics.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.