IMMUNOPATHOGENESIS OF BEHÇET’S DISEASE

Abstract

Behçet’s disease (BD) is a multi-system inflammatory disorder with vasculitic features. It does not suit any of the current pathogenesis-driven disease classifications well, a unifying concept of its pathogenesis is not unanimously conceivable at present, and its etiology is obscure.

Still, evidence from immunogenetic and other studies supports the notion of a complex-polygenic disease with robust innate effector responses, reconstitution of regulatory T cells upon successful treatment, and first clues to the role of an, as of yet, underexplored adaptive immune system and its antigen recognition receptors.

Without an attempt to be comprehensive, this review aims to collect and organize impactful parts of this evidence in a way that allows the reader to appreciate the work done and define the efforts needed now. The focus is on literature and notions that drove the field into new directions, whether recent or more remote.

1. Introduction to Behçet’s disease (BD)

Behçet’s disease (BD) is a multisystem immune-mediated disorder with vasculitic and peri-vasculitic components. It is clinically characterized by relapsing aphthous ulcers in the oral mucosa in nearly all cases and recurrent ulcers in the genitalia as well as inflammation in the interior of the eye (uveitis) in most patients [1]. Skin involvement with nodular or pustular rashes in a characteristic distribution pattern occurs frequently, and the disease can affect and seriously damage most major organ systems, typically except for the kidneys. Overall morbidity of untreated BD is high as there is a significant risk of blindness through the involvement of the posterior pole of the eye – its most common major organ manifestation. Central Nervous System (CNS) and blood vessel involvement can be life-threatening but are rarer. Gastrointestinal (GI) disease, characterized by deep ulcers in the intestine, is exceedingly uncommon, at least in non-East Asian populations, but can be fatal given a high risk of bowel perforation.

BD is relatively rare globally but not regionally and locally in high-incidence areas and immigrant communities, primarily affecting people of South European, North African, Middle Eastern, and East Asian descent. Disease prevalence is best known in Turkey, where it may reach 400 /100,000, roughly between the estimated US prevalence of systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA). BD prevalence strongly correlates with the minor allele frequency of HLA-B*51 in the general population, which is highest amongst populations inhabiting the interconnected landmasses of Southern Europe, the Middle East, the Maghreb region in Northern Africa, and the Far East.

Although, collectively, studies report a roughly equal male-female gender ratio, most BD patients with severe disease are male. The disease typically manifests in the third decade of life and usually abates with progressing age. BD does occur in children, but not typically before the age of five and infrequently before the age of ten. Young age at disease onset and male sex are strong predictors of severe disease. BD is complex-polygenic with cases of familial clustering and at least one well-established HLA class I association (HLA-B*51). While robust and significant, the strength of this association is much weaker than that of HLA-B*27 with ankylosing spondylitis (AS) and acute anterior uveitis (AAU) or of HLA-A*29 with Birdshot chorioretinopathy – the strongest one known for human disease.

Directly disease-related morbidity is highest in the third and fourth decades of life. Although we know little about treatment-related morbidity from formally conducted studies, it is likely significant, given the still widespread use of highly toxic medications for its management in large parts of the world, such as cyclosporine for eye disease or cyclophosphamide for some forms of vascular or CNS involvement.

Diagnosis of BD is clinical. Sets of diagnostic criteria exist to aid in diagnosis and enhance research [2]. Despite those efforts, many conditions that can mimic BD often create challenges for clinicians with limited exposure to BD and those practicing in clinical contexts where BD is rare, but mimics are not. Typical examples in the US include nutritional deficiencies (e.g., B vitamins, iron, zinc, folate) causing oral and genital ulcerations, autoimmune-blistering or Stevens-Johnson spectrum disorders affecting the mucous membranes, and, perhaps most impactful, inflammatory bowel diseases, especially Crohn’s disease (CD), which can mimic most aspects of BD, sometimes preceding intra-luminal GI disease or even existing independently of it, such as in some cases of vulvar Crohn’s disease. These difficulties very likely have immensely biased numerous research studies.

Most clinicians tailor the treatment of BD to disease severity assessed as the risk of permanent organ damage and quality of life considerations. Skin, mucosal, and joint involvement is often less severe given the low to moderate risk of permanent organ dysfunction, but ocular, major vascular, CNS and gastrointestinal involvement potentially confer life-long disability, including blindness, or can even be lethal. [3]. Corticosteroids are helpful for acute disease, with colchicine, azathioprine (AZA), cyclosporin-A, cyclophosphamide, interferon alfa (IFNα), and tumor necrosis factor-alpha (TNFα) inhibitors, among other agents used as induction or maintenance therapy [4].

Although anti-TNF agents have dramatically improved treatment success for most major organ manifestations of BD, there are still many unmet needs. These include the induction of drug-free remission and the identification of distinct disease endotypes allowing the application of personalized medicine approaches. The increased risk of tuberculosis (TB) and its high prevalence in many BD-endemic regions and affected immigrant populations, together with the high cost of these agents and the risk of severe rebound disease upon withdrawal, poses additional challenges when using an anti-TNF agent in BD. Given the lack of mechanistic understanding, targeted therapies do not exist for BD, and chronic systemic immunosuppression generally has significant adverse effects. By better understanding the mechanisms that drive this disease, more effective targeted therapies can be developed, thereby improving the lives of BD patients.

2. Etiology

The etiology of BD is unknown. While there is evidence for dysregulation in multiple immune cell compartments and pathways, causal stimuli inciting these events remain obscure but are suspected to include environmental factors or modified self. The latter likely involves altered antigen presentation and innate immune activation [5, 6]. There is also compelling evidence of significant genetic and epigenetic risk, classifying BD as a complex polygenic, immune-mediated disease of unknown etiology.

3. Immunogenetics

To date, immunogenetic studies still provide the most potent and comprehensive clues to targets of interest in BD pathogenesis [Figure 1]. The first GWAS study in BD found associations with genes thought to be involved in innate immunity and T-cell receptor signaling [7]. Additional dense genotyping studies demonstrated that the HLA-B/MICA locus was significantly associated with genetic risk for BD [8]. A subsequent major genome-wide association study in 2013 in a Turkish population identified associations at CCR1, STAT4, and KLRC4 loci and ERAP1. Epistatic interaction between HLA-B*51 and ERAP1 conferring increased BD risk was detected [9]. A smaller GWAS study in 2013 in a Korean population found an association between BD and the GIMAP locus, a protein involved in T-cell survival. Experiments demonstrated that CD4 T cells from BD patients showed a lower level of GIMAP4 mRNA, and GIMAP4 knockdown was protective against Fas-mediated apoptosis [10]. Overall, these results are examples that suggest that T-cell aberration may contribute to the development of BD.

Figure 1:

Most polymorphisms linked to BD risk are shared with other diseases, especially Recurrent Aphthous Stomatitis (bold genes at the bottom of the pyramid), or can be mimics of BD (e.g., SJS involving the mucosal membranes; HLA-B*27-associated diseases; IBD, PS). However, epistasis of HLA-B*51 with ERAP1-Hap10 recessively confers the highest BD risk known to date and is BD ‘specific.’ The haplotype ERAP1-Hap10, associated with BD risk in the presence of HLA-B*51, protects from AS/AAU and PS in the presence of their HLA class I risk alleles, i.e., HLA-B*24 and HLA-C*06, respectively. (AS = ankylosing spondylitis; AAU = acute anterior uveitis; IBD = inflammatory bowel diseases; PBC = primary biliary cirrhosis; PS = psoriasis; SJS = Stevens-Johnson syndrome) Figure created with BioRender.com

More recently, these GWAS studies led to a more comprehensive study of Turkish patients demonstrating that the combination (epistasis) of HLA-B*51 and the ERAP1 coding variant Hap10 conferred an 11-fold increased risk of developing BD vs. the 4-fold risk of HLA-B*51 alone [11]. HLA-B51 is an MHC class I protein responsible for antigen presentation to effector CD8⁺ T cells. Endoplasmic reticulum aminopeptidase 1 (ERAP1) is a protease that resides in the endoplasmic reticulum and is responsible for trimming peptides that are ultimately presented on HLA class I molecules, such as HLA-B51, to CD8⁺ T cells. Thus, altered trimming of peptides by ERAP1 that HLA-B51 then presents is likely involved in the BD pathogenesis of subjects carrying the risk variant. The aberrant HLA-B51-bound peptidome in BD presents to effector CD8+ T cells, an abundant cell type in the anterior chamber fluid of BED patients during active ocular inflammation [12]. A 2016 whole exome sequencing study in patients of European descent identified rare variants associated with BD affecting LIMK2 and NEIL1, which encode proteins involved in cytoskeleton reorganization or cell motility and base excision repair, respectively. These functional domains may relate to the aberrant leukocyte infiltration and oxidative stress seen in BD [13]. An epigenome-wide study in monocytes and CD4 T cells from BD patients demonstrated differential DNA methylation in genes thought to regulate cytoskeletal dynamics. This pattern was affected by the treatment of BD, with the reversal of altered DNA methylation noted in treated patients whose disease was in remission [14]. These studies strongly indicate a central role in adaptive immunity, antigen presentation, and T-cell activity.

4. Tissue and Cellular Immunobiology

Many BD manifestations exhibit features of systemic perivasculitis, in which early neutrophil infiltration, endothelial damage, and – to varying degrees – fibrinoid necrosis mediated by macrophages and neutrophils have been reported [15]. Neutrophil infiltration occurs in mucocutaneous aphthae and nodular cutaneous lesions, ocular lesions, and the skin pathergy reaction [16–18]. In addition, serum myeloperoxidase levels, generated by active neutrophils, are also elevated [19]. Although genetic factors influence BD pathogenesis, most carriers of HLA-B*51 do not develop BD, suggesting that additional factors, such as environmental stimuli, must drive the disease. These may include microbes, molecular mimics such as Hsp60, and others [20]. In vitro studies found that antigens from Streptococcus, E. coli, and Staphylococcus aureus, as well as non-peptide antigens, strongly activated γδ T cells in BD patients, suggesting that T lymphocytes of BD patients may be hyperactive against bacterial antigens generally and not simply attuned to a specific antigenic signal [21, 22]. Due to high sequence homology between human 60-kDa HSP (HSP60) and bacterial 65-kDa HSP (HSP65), peptides from the human HSP60 both stimulate the proliferation of γδT cells in BD patients, and autoantibodies against both the human and bacterial proteins have been reported in BD patients [23, 24]. Thus, aberrant antigen presentation in the setting of molecular mimicry may also be contributing to BD pathogenesis. Finally, better oral hygiene is associated with decreased BD activity, suggesting that genetics alone do not determine BD pathogenesis [25].

Knowledge about the cellular machinery initiating, maintaining, and controlling target organ inflammation in BD is rudimentary. Nevertheless, published data regarding most major immune cell types in some contexts exist. The rapid progress of scRNA sequencing technology that now allows combining phenotypic information with antigen recognition receptors will likely enable a more granular understanding of the players involved in the immunopathogenesis of BD over the coming years.

4.1. B cells

B cells are the least investigated common immune cell type in BD. This lack of attention stems most likely from the frequently cited "lack of autoantibodies" in BD, the observation that B cell infiltrates in tissue are rare or absent in BD target organ lesions, and the notion of a vital "auto-inflammatory" component to the disease. Our advances in technical capability, in particular as pertains to V(D)J sequencing analyses, have already begun to shed new light on this compartment (see the section “antigen recognition receptors” below). One trial with rituximab, an antibody targeted against CD20 expressed only by pre-B and mature B cells, demonstrated reduced disease activity [26]. This result suggested a role for B cells in BD pathogenesis and led to further studies of the effect of anti-TNF therapy. Using samples from treatment-naive and adalimumab-treated patients (an anti-TNF agent), reduced numbers of circulating memory B cells and proliferation defects were noted. T and B cells infiltrating ulcerated tissues occurred in patients with active BD. Treatment with TNF blockers appeared to reverse these B cell aberrations, implicating B cells in the pathogenesis of BD, at least in a supplementary role [27].

4.2. T cells

Studies of BD have consistently demonstrated aberrant, excessive activation of innate and adaptive immunity, suggestive of autoinflammatory and autoimmune disease, respectively. Given the prominence of mucosal lesions in the disease, a complex interplay between the mucosal surface microbiome and neutrophil-initiated immune responses that subsequently activate a dysregulated adaptive response is conceivable [28]. T cells mediate neutrophil activation by producing pro-inflammatory cytokines, particularly IL-17 secretion by T_H17 cells. Given the T_H17 imbalance seen in BD, T cells may unleash the neutrophilic response that characterizes many BD lesions [29]. There is ample evidence that TNFα plays a central role [30, 31]. This inflammatory cytokine can be produced by and act upon T cells to promote activation and proliferation. TNFα can also induce apoptosis of highly activated effector T cells and act dichotomously upon Tregs to downregulate their suppressive capacity and promote their proliferation [32]. The latter seems to be underpinned by the finding that significant Treg reconstitution occurs in BD patients after TNFα inhibitors achieve control of their disease [33, 34]. Accordingly, increased levels of TNFα predominate in both serum and aqueous humor of BD patients and increased levels of soluble TNFα receptors in peripheral blood [31, 35]. The striking clinical success of TNF-inhibiting agents in ameliorating many disease manifestations of BD further underpins the high likelihood of a central role for this cytokine and its cells of origin in BD pathogenesis.

In a recent study using a transcriptomic database of sorted immune cell subsets from BD and healthy control patients, Okubo et al. found significant increases of Th17 cells in the peripheral blood of BD patients. The authors also showed upregulation of the NFkB pathway in the BD Th17 cells. Looking specifically at HLA-B*51⁺ BD patients, they found upregulation of IL-17-associated genes in CD8⁺ memory cells, suggesting enrichment and activation of Tc17 cells in BD [36].

While most of the published literature suggests a predominant role for CD4 T cells in the disease, several considerations call for a partial revision and refinement of the current concept. The well-established association with HLA class I alleles requires antigen presentation to CD8 effector cells as a part of the conceptual framework. Evidence for CD8 T cells at effector sites does exist, including in the eye [37–39], the skin [40, 41], mucosal lesions [42], cutaneous pathergy reactions [43, 44].

More recently, we showed modulation of the CD8 T cell phenotype in an HLA-B*51 background through a peptide-trimming aminopeptidase (ERAP1) towards antigen-experienced CD8 cells, suggesting the presence of an HLA class I restricted process activating these cells [6]. Another aspect worth considering is that the scientific community’s ability to determine and investigate antigen-specificity in human systems at a large scale used to be extremely limited. This notion has dramatically changed through the advent of single-cell sequencing with analysis of the V(D)J. Total numbers of T cells do not necessarily reflect the biological importance of a cell type whose major functional determinant lies in its unique ability to recognize cognate antigens in HLA restriction, and minimal numbers of antigenspecific cells can initiate robust immune responses.

Lastly, a large proportion of CD8 T cells, granzyme K⁺ CD8 T cells, are capable of secreting substantial amounts of TNF [45] and may contribute to the TNF-centric effector cytokine patterns observed in BD.

4.3. NK cells

Natural killer (NK) cells arise from a common NK/T-cell/B-cell progenitor. As frontline effector cells of the innate immune system, NK cells participate in early cytotoxic responses against pathogens and perform tumor immunosurveillance. They also have protective and pathogenic roles in autoimmune diseases, such as systemic lupus erythematosus [46]. These features stem from their ability to regulate other immune cells; for instance, NK cells can mediate the release of cytokines (such as TNFα, IFNγ, and IL-2) that help CD4⁺ T cells differentiate into Th1 cells. Indeed, these Th1-associated cytokines are known to correlate with BD disease activity [47]. NK cells include CD56^dimCD16⁺ NK cells, which are potently cytotoxic but have limited production of cytokines and chemokines and represent the majority of NK cells in peripheral blood, and regulatory CD56^brightCD16⁻ NK cells which strongly produce cytokines and chemokines when stimulated but have limited cytotoxicity [48].

There are many studies on the role of NK cells in BD, although conflicting results have hampered their interpretation. One early study demonstrated increased NK cells in peripheral blood in patients with active BD compared to inactive BD and healthy controls but decreased NK cell activity (based on in vitro cytolysis) in these patients [49]. A subsequent study also reported increased CD69⁺ NK cell quantity in active BD patients but similar cytotoxic function. It also found that NK cells from inactive BD patients (but not active BD) had downregulated IL-12 receptor b2 expression and could suppress IFNγ production by CD4⁺ T cells derived from active BD patients. The authors suggested that NK cells may thus impact BD flare/remission via relative NK1/NK2 bias, with NK2 cells suppressing IFNy and the Th1 response [50]. Hamzaoui et al. investigated NK cells specifically from bronchoalveolar lavage of BD patients with pulmonary vascular involvement. They found that NK cell quantity was reduced in BAL from BD patients, with reduced cytotoxicity and lymphokine-activated killer activity [51]. In contrast, a study of BD patients with only mucocutaneous involvement found no difference in total NK cell frequency nor NK cytotoxic activity in BD NK cells compared to healthy controls; an increase in the IFNγ-secreting NK1 subset was noted [52]. Similarly, a study of uveitis-affected BD patients demonstrated NK cell cytokine production patterns corresponding to a Th1/NK1 profile (TNFα, IFNγ, and IL-2) during periods of BD activity and a switch to Th2/NK2 (IL-4 and IL-10) predominant cytokine pattern during remission [53].

To resolve contradictory findings in prior literature, Hasan et al. performed detailed studies of NK cells in BD patients with and without active disease compared to healthy controls [54]. They found that CD56^dim and CD56^bright NK cells were depleted within the peripheral blood compartment of BD patients compared to healthy controls, with a more pronounced depletion in patients with current disease activity than in those without active disease. They also noted a proportionately more significant loss of the CD56^dim subset compared to the reduction in CD56^bright cells. Separately, they also noted that azathioprine independently induced NK cell depletion in BD patients, similar to SLE, IBD, and RA observations. The authors hypothesized that NK cell depletion from peripheral blood in BD occurs due to net trafficking into disease-active tissue sites, which is why BD activity correlates with depletion. The percentage of CD56^bright NK cells, with IFNγ production ability, was increased in BD patients corresponding with earlier studies suggesting NK cells promote Th1 cytokine production in BD; at the single cell level, IFNγ production did not increase significantly in this subset, however [54].

Ultimately, it is possible to conceptualize these results, considering that BD in different anatomical sites may reflect different disease states with corresponding differences in NK cell homing and proliferation. There is supporting evidence for this idea in different sites of disease activity, with NK tissue trafficking in active BD also demonstrated in histological samples in BD uveitis [55]. Furthermore, as BD evolves from active to inactive states, NK1/NK2 bias likely shifts accordingly. Given NK cells’ ability to guide adaptive immune responses, it is plausible that NK1 cells may be driving the disease instead of merely following it.

4.4. Neutrophils

Neutrophils play an essential role in many tissue lesions and pathological states of BD. As discussed above, neutrophils infiltrate many of the affected tissues in BD – particularly in the early stages of an inflammatory episode when they are hyperreactive, based on cytokine findings and functional assays. Neutrophils are central innate immune effectors, serving as a primary defense again infection. In BD, neutrophilic vasculitis is a common histological finding [56]. These highly active perivascular neutrophils produce ROS, which drives fibrinogen oxidation and thrombus formation [57], in addition to endothelial injury and dysfunction. Surface and functional changes in neutrophils in BD patients were reported early on, including CD10/CD14 upregulation, suggestive of priming and activation, and CD16 downregulation, suggestive of activation and subsequent apoptosis [58]. Recent work has also shown that neutrophil extracellular traps (NETs) are present in some BD patients and contribute to a procoagulant state, thus possibly contributing to BD-associated thrombosis [59]. Furthermore, these BD NETs may be driving macrophage hyperactivation and encouraging IFNγ⁺ CD4⁺ (e.g., TBET-expressing Th1) T cell differentiation. [60] However, whether genetic or pathogen-derived factors precipitate neutrophil hyperreactivity in BD remains unclear. Familial Mediterranean Fever (FMF), an autoinflammatory disease with a similar geographical distribution, is caused by pyrin-encoding MEFV gene mutations. Although clinically distinct from BD, FMF also entails neutrophil-mediated intense inflammation [61]. Many Middle Eastern BD patients carry FMF mutations, and in those, it is conceivable that the latter may contribute to and convolute the picture regarding the observed pathophenoptype at the tissue level.

In an attempt to clarify whether the increased activity of BD neutrophils was characteristic of the full spectrum of BD patients, Perazzio et al. 2015 studied phagocyte function and activation in cells from control, septic, inactive BD, and active BD patients to determine if cell-extrinsic factors could be necessary [62]. They found that neutrophils and monocytes from highly active BD patients had some evidence of increased activity, although they were not inherently more phagocytic than healthy controls. They also found that this phagocytic hyperactivity correlated with Behçet’s Disease Current Activity Form (BDCAF) scores. This study thus positioned phagocytes (namely neutrophils and monocytes) as essential players in BD pathogenesis that drive severe BD manifestations, whose hyperactivity may not be intrinsic to these cell populations but rather a form of constitutive activation secondary to a possible unknown soluble factor.

While the importance of neutrophil effector function seems indisputable, many authors have found it difficult to reconcile this aspect of BD pathogenesis with the notion of a potentially HLA class I-restricted disease. Similar considerations apply to the understanding of other diseases with HLA class I associations and partially overlapping clinical phenotypes where evidence of neutrophil responses at effector sites exists, e.g., neutrophilic cell predominance in the aqueous humor during acute anterior uveitis and neutrophilic dermatoses in HLA-B27 associated disease. In their review of inflammatory non-autoantibody-associated spondyloarthropathies and related disorders, including Behçet’s, Macleod et al. note several important shared features [63]. Namely, they share prominent neutrophilic inflammation and common genetic polymorphisms in the IL-23–IL-17 axis and CD8⁺ T cell-mediated immunity (particularly class I MHC, RUNX3, TBX21, and ERAP1). While the role of the IL-23–IL-17 axis in driving neutrophil recruitment and activation is well appreciated, the authors propose that the neutrophils recruited and activated at disease sites by T cells and IL-17, in turn, can promote the development of Th17 and Th17-type CD8⁺ (Tc17) T cells. Neutrophils can achieve this by producing IL-23, inflammatory factors like macrophage migration inhibitory factor, NETs, and proteases. Thus, rather than merely terminal effectors, neutrophils may be first-line innate immune responders that drive complex positive feedback loops to regulate the adaptive immune response, thereby offering a potential mechanistic explanation for the sterile neutrophilic inflammation observed in these diseases with strong MHC class I associations including BD.

4.5. Monocytes and other myeloid cells

Monocytes bind antibody-bound or complement-opsonized microbes through either phagocytosis or using surface receptors such as cC1qR and gC1qR that primarily recognize C1q and PAMPs. In the traditional perception, monocytes consist of three main subtypes: 'classical' CD14⁺⁺ CD16⁻ monocytes which express high levels of CD14 (LPS receptor) but lack CD16 (Fc receptor); ‘non-classical’ CD14⁺ CD16⁺⁺ monocytes with low CD14 and high CD16 expression; and intermediate CD14⁺⁺CD16⁺ monocytes with high CD14 and low CD16 expression. Activated by a PAMP or microbial antigen, “non-classical” monocytes produce high amounts of proinflammatory cytokines, including TNF and IL-12, and participate in antigen presentation [64]. This classification is under revision, given new insights from single-cell RNA sequencing, which dissects cell subsets at much more granularity.

In a recent study, Zheng and colleagues combined single-cell RNA sequencing with bulk sequencing to study PBMCs and monocytes, particularly at high resolution in BD patients [65]. BD patients had approximately twice the proportion of circulating monocytes, with the notable expansion of a C1q-high (C1q^hi) monocyte population in BD. The authors identified differentially expressed genes in BD versus healthy controls and found these were notably upregulated in monocytes and dendritic cells and downregulated in B cells. In concordance with previous studies, they also noted reduced memory B cell populations in BD patients. The differentially expressed genes across multiple lineages map primarily to bacterial and viral infection pathways, consistent with the hypothesis that infectious agents play a role in BD etiology (see also discussion of B cell repertoire IGHV below). In the T cell population, Th1 cell differentiation predominated. In the total monocyte population, IFNg response, exogenous antigen, and neutrophil activation were all increased, and the increases noted for these genes in the bulk sequencing were attributable primarily to the monocyte fraction. The C1q^hi monocytes in BD patients highly express complement C1q and macrophage markers like CD68. Apart from increased C1q^hi monocytes, BD patients had reduced numbers of monocyte-derived dendritic cells (moDC). Phenotypically, moDC had reduced CD14 expression, high expression of DC markers (CLEC10A, FCER1A, CST3, and CD74), enrichment of antigen presentation pathways, and a DC-specific gene signature. Differentiation trajectories suggested that MHC-II monocytes (highly expressing HLA-DRA, HLA-DQA1, HLA-DPB1, and CD74) terminally differentiated into either C1q monocytes or moDCs, with BD monocytes preferentially differentiating into C1q monocytes and healthy control monocytes preferentially differentiating into moDCs, further implying that the enriched C1q monocyte population in BD was driving the increased inflammation seen in BD. C1q^hi BD monocytes overexpressed proinflammatory cytokines (TNF and IL6) at baseline, FcR activators, recruiters of cytoskeleton remodelers, and MHC-II components. The C1q^hi monocytes demonstrated enhanced phagocytic ability and increased TNF/IL6 production in response to LPS compared to CD16⁺ and CD16⁻ monocytes. The authors sorted CD14+ cells into C1q^hi, CD16⁺, and CD16⁻ monocytes.

C1q^hi monocytes displayed global overexpression of IFNg-induced genes mediated by STAT1 and IRF1. Correspondingly, they found elevated serum IFNg in BD patients, with increased Th1 (IFNg⁺ CD4) cells suggesting crosstalk between CD4⁺ T cells and C1q^hi monocytes and IFNg promoting the generation of C1q^hi monocytes. Expanding to look at additional previously published BD cohorts, they were able to determine upregulated expression levels of C1q genes in bulk RNASeq data, accompanied by elevated C1q protein levels in serum. The numbers of C1q^hi monocytes positively correlated with ESR and BDCAF scores. Finally, IFNg-induced expression in C1q^hi monocytes and the proportion of C1q^hi monocytes decreased with immunosuppressant treatment in BD patients. Assessment after the use of the JAK-STAT inhibitor tofacitinib replicated this finding. Thus, the authors concluded that C1q^hi monocytes appeared to be impactful actors in BD pathogenesis, potential clinical targets, and valuable markers of BD disease activity [65]. Another recent study supports these findings by applying RNA-seq and IF/IHC to BD patient skin samples: A higher proportion of monocytes, particularly ISG15⁺ Mono and C1q⁺ Mono subsets, was found in the skin of BD patients compared to controls. Expression analysis demonstrated that the IFNγ pathway was activated in the skin of BD [66].

Interestingly, deficiencies in the early components of the classical complement pathway (such as C1q, C4, and C2) are associated with SLE, with homozygous C1q deficiency being the most substantial genetic susceptibility [67]. C1q plays a vital role in the clearance of apoptotic cells and also helps determine activation thresholds of B and T cells, such that C1q deficiency may cause loss of peripheral tolerance [68]. The role of C1q is likely complex, as both its presence and absence appear to drive dysregulated inflammatory states. C1q may thus function as both a sensor of danger but also as a guarantor of a steady state; at rest, C1q may maintain cells in a monocyte-like innate immune state and then, upon recognition of a PAMP "danger" signal, shift monocytes toward a DC lineage to engage adaptive immunity [64]. Therefore, deficiency or excess in C1q leads to dysregulated inflammatory states. Megakaryocytes and platelets may also be significant participants in BD pathology. Once the thrombosis cascade activates in BD due to neutrophil-derived ROS and NETs, platelets may be recruited to areas of endothelial damage, degranulating and aggregating, thereby facilitating further inflammation and thrombosis [69].

5. Antigen Recognition Receptors

The technical advances in our ability to link sequencing of antigen recognition receptors to gene expression profiles of immune cells at the single cell level allow for a profound reassessment of immune cells orchestrating adaptive immune responses in immune-mediated human diseases. While, in BD, this work is just beginning, several studies illustrate the potential strength and insights attainable.

5.1. BCR

One major study analyzed the B-cell repertoire (BCR) in six immune-mediated diseases, including BD in patients with active disease prior to treatment, demonstrating significant differences in BCR. In particular, they found that IgA was the dominant isotype in BD patients compared to healthy controls. Genes of the V1 family were over-represented in all IMDs studied; in BD, IGHV1–46, IGHV1–3, and IGHV1–69 were noted to be highly expressed. These genes are known to increase expression in infection, and their encoded proteins bind to microbial antigens and autoantigens. Unexpectedly and uniquely among IMD examined, there was reduced BCR clonality in BD in both BCR clonal expansion and clonal diversification. Additionally, class switching recombination was also reduced in BD [70]. Taken together, these data regarding BCR in BD combined with the changes in B cells reported by van der Houwen point to significant changes in BCR and B cells in BD and suggest a role for autoantigens in the disease [27].

5.2. TCR

Two T cell populations, namely αβ T and γδ T cells, are defined by the expression of either an αβ TCR or a γδ TCR, respectively. Complete αβ or γδ TCRs consist of TCRα/TCRβ (TCRαβ) or TCRγ/TCRδ (TCRγδ) Ag binding heterodimers combined with a CD3 complex. The two receptors have significant differences; they are structurally distinct regarding physical geometry, plasma membrane organization, and the accessibility of signaling motifs in the CD3 intracellular tails. They also have different glycosylation patterns. Overall, there are fewer VDJ combinations for γδ T cells compared to αβ T cells.

Additionally, while activating αβ TCR ligands causes exposure of the proline-rich sequence (PRS) in CD3ε to induce αβ TCR signaling, CD3ε PRS exposure does not occur by binding of ligands to the γδ TCR, and γδ TCR signaling occurs independently of CD3ε PRS exposure [71]. Both types can exist on non-MHC-restricted ‘unconventional’ T cells, which can recognize non-polymorphic antigen-presenting molecules. These molecules are encoded by genes outside the MHC locus (e.g., CD1 family and MR1) or within the MHC locus (HLA-E, HFE, H2-M3, and H2-Q9). These T cells are deemed 'unconventional' because they do not recognize classical peptide antigens, are not donor restricted, and circulate abundantly during early infection and reinfection. Unconventional T cells have more limited TCR diversity and are often found in non-lymphoid tissues [72]. Fortune et al. found higher levels of γδ T cells in BD patient-derived peripheral blood compared to healthy controls, similar to Systemic Lupus Erythematosus (SLE) patients [73]. This increase appeared specific to BD patients with inflammatory arthritis and did not occur in BD patients with ocular and mucocutaneous disease manifestations, perhaps reflective of the distinct localization of γδ T cells. Likewise, BD patients with active neurological disease have increased Vδ1 T cells in their cerebrospinal fluid [74]. Functionally, γδ T cells are known to be strong Th1 and Th17 inducers, and accordingly, the percentages of Th17 cells and IL17 are increased in BD [75].

Another recent study shed light on TCRβ repertoires and showed less clonal diversity in BD than in HD. The clonal expansion also occurred at the tissue level in skin lesions, and it was a male-predominant finding [76]. The authors did not identify TCRα chains paired with TCRβ, which would be necessary for reconstructing and experimentally expressing TCRs for functional testing and antigen discovery. Nonetheless, it importantly points to the presence of clonal expansion, confirming results from earlier work [77, 78]. Current and future work will identify α/βTCR, allowing us to test hypotheses that imply HLA-class I restricted antigen recognition by CD8 T cells and their target antigens.

5.3. NK/ NKT cell receptors

Natural killer (NK) cells are essential first-line immune effectors that can recognize and kill cells expressing low levels or no self-MHC, thereby preventing viral infection, tumor growth, and metastatic spread. They possess killer Ig-like receptors (KIRs), which are part of the immunoglobulin-like superfamily, and have specific ligands in the HLA-class I family [79]. In a study of 40 patients with BD, KIR expression was abnormal in BD patients with severe eye disease, suggesting that NK cells in BD patients may lack appropriate inhibitory KIRs, thus failing to recognize self-MHC and causing autologous tissue damage [80]. At the genetic level, KIRs have been studied extensively in BD, but these studies have failed to establish a clear link to disease risk. The finding that HLA-B51 can engage KIR3DL, an inhibitory KIR, through its Bw4 epitope has made NK cells and KIR a candidate for a mechanistic explanation of the HLA-B*51 disease association, but conclusive experimental evidence is lacking to date. The complex and incompletely understood biology of most human NK cell receptors makes their future study worthwhile but technically challenging.

6. Conceptual Approaches to Understanding BD Biology Today

The multiorgan involvement, aberrations in multiple cellular populations, and the polygenic nature of BD pose significant challenges to understanding its pathogenesis. Consistently observed alterations in innate and adaptive immune responses point to the involvement of multiple cell types, pathways, and immune-pathophysiological states without an easily discernable 'common theme.' While results from twin studies and immunogenetic population-based work demonstrate a substantial genetic component whose impact exceeds that observed in classical autoimmune diseases such as RA and type-I-diabetes mellitus, mechanistic evidence for conceptually compelling players in affected carriers, such as ERAP1 and HLA-B*51, has just begun to evolve [Figures 1, 2].

Figure 2: Timeline of selected mile stones in BD discovery.

Refers to results from [6, 9, 12, 22, 24, 30, 31, 81–84] Figure created with BioRender.com

Important current areas of interest are the biology of neutrophil and monocyte responses in the disease, understanding the effects of aberrant antigen processing and presentation in terms of their downstream cellular effectors, and the identification of pathogenic antigens, if indeed present. Single-cell immune phenotyping and gene expression analyses have been powerful techniques for sampling immunological states in health and disease. They have emerged to show clear value in driving recent advances in BD, and these efforts must continue at a large scale. In the same vein, much work remains to understand antigen receptor repertoires and function, which are essential to discovering disease-specific motives, allowing the identification of their binding partners – HLA restricted or not. Although conceptually appealing, the postulation of BD being an "MHC-I-opathy" is far from proven and may not be universally applicable throughout all patient subsets. BD also escapes the prevailing classification schemes of autoimmunity and autoinflammation. It is, therefore, paramount to keep an open, data-driven mind and pursue an integrative, comprehensive view of its pathobiology while resisting the urge to classify the unknown.

The immune biologic “picture” of BD today is a complex-polygenic disease driven by an episodically perturbed balance of pathogenic effector and tolerogenic processes involving molecules produced by varying effector cells in different tissues. As is likely the case in most human immune-mediated disorders, it is conceivable that BD states manifest through molecular "effector” and “control” signatures that are generated and executed by a variety of cellular players and pathways with significant functional plasticity and versatility, entailing the ability to assume profoundly different roles at varying time points, tissues, and physiological contexts.

The current cell-centric focus of the field, with its enthusiasm for single-cell approaches, marks an essential early stage in discovering these immune signatures. Ultimately, understanding the mechanisms through which they act in disease causation, maintenance, and resolution through hypothesis-driven research will be necessary to develop and meaningfully apply targeted treatments to benefit patients afflicted with BD.