The role of YKL-40 in the pathogenesis of autoimmune diseases: a comprehensive review

YKL-40, a chitinase-3-like protein 1 (CHI3L1) or human cartilage glycoprotein 39 (HC gp-39), is expressed and secreted by various cell-types including macrophages, chondrocytes, fibroblast-like synovial cells and vascular smooth muscle cells. Its biological function is not well elucidated, but it is speculated to have some connection with inflammatory reactions and autoimmune diseases. Although having important biological roles in autoimmunity, there were only attempts to elucidate relationships of YKL-40 with a single or couple of diseases in the literature. Therefore, in order to analyze the relationship between YKL-40 and the overall diseases, we reviewed 51 articles that discussed the association of YKL-40 with rheumatoid arthritis, psoriasis, systemic lupus erythematosus, Behçet disease and inflammatory bowel disease. Several studies showed that YKL-40 could be assumed as a marker for disease diagnosis, prognosis, disease activity and severity. It is also shown to be involved in response to disease treatment. However, other studies showed controversial results particularly in the case of Behçet disease activity. Therefore, further studies are needed to elucidate the exact role of YKL-40 in autoimmunity and to investigate its potential in therapeutics.

It is expressed and secreted by various kinds of

Ivyspring
International Publisher cells such as chondrocytes, fibroblast-like synovial cells, vascular smooth muscle cells, and macrophages. Its exact pathophysiological function and mechanism of action is currently unknown, but it is thought to play a role in inflammation, tissue remodeling, and angiogenesis [11]. It is involved in the activation of the innate immune system, extracellular matrix remodelling, [8,12] and the differentiation of CD14+ monocytes to CD14-and CD16+ macrophages [13]. YKL-40 stimulates the proliferation of human connective tissue cells (fibroblasts, chondrocytes, synovial cells) in a dose-dependent manner [14,15] and also plays important roles in the antigen-induced T-helper 2-response, antigen sensitization and IgE induction as well as activation of innate immune cells [16]. Recent studies have suggested important roles of YKL-40 in various autoimmune diseases. However, these findings have not been reviewed comprehensively. This review will discuss the potential of YKL-40 in the pathogenesis of autoimmune and rheumatic diseases, such as rheumatoid arthritis, psoriasis, systemic lupus erythematosus, Behçet disease, and inflammatory bowel disease. Possible therapeutic strategies by targeting YKL-40 will also be discussed.

Rheumatoid arthritis (RA) and YKL-40
Rheumatoid arthritis (RA) is a chronic autoimmune disease principally affecting the synovial joints, causing pain and limitation of motion by the destruction of arthritic cartilage and joint ankylosis. The main aim of RA treatment is to achieve early remission with minimal disease activity. Disease modifying anti-rheumatic drugs (DMARDs) (i.e. sulfasalazine, methotrexate and hydroxychloroquine), glucocorticoids (GC) and biologics such as anti-tumour necrosis factor (TNF)-α inhibitors have been used in the treatment of RA. Disease activity is usually assessed by some indices, such as the 28-joint disease activity score (DAS28) [17] which evaluates the count of tender and swollen joints, inflammation and patient´s assessment of the disease activity.
Although the exact mechanism of joint destruction is still elusive, the presentation of exogenous and autologous antigens by antigen presenting cells (APCs) to T-cells are considered to be an early process of disease, and YKL-40 has been recognized as a potential candidate autoantigen [2,3,4,5,7]. Recently, there have been publications on important roles of YKL-40 in the pathogenesis of RA such as correlations with disease activity of RA and diagnostic or prognostic biomarkers. We will discuss the role of YKL-40 in RA pathogenesis as well as its potential therapeutic target.
A total of 47 articles discussing the association between RA and YKL-40 were identified from PubMed search. There were two review articles: one about the association between YKL-40 and RA [18], and the other about the association between YKL-40, inflammation, and cancer [19]. Twenty-three articles were reviewed in these two review articles [14,15,16,. From 21 articles newly published after the review, one article was excluded due to multiple confounding variables [41], and another was excluded because it was a phase 1 clinical trial which did not discuss the effect of YKL-40 [42]. We have summarized the key findings of 43 eligible articles ( Table 1), 24 of which were included in reviews by Johansen et al. [18] and Kzhyshkowska et al. [19] which have been published thereafter. Table 1. List of studies on the relationship between RA and YKL-40 (chitinase-3-like protein-1, human cartilage glycoprotein-39).

Authors, year
Study design Main study findings van Bilsen JH et al, 2004 [20] -PBMCs obtained from RA patients were stimulated in vitro with HC gp-39.
-Elispot analyses were used to analyze the production of cytokines.
• HC gp-39 may have anti-inflammatory phenotype in healthy individuals. • gp-39 immune response in RA patients has pro-inflammatory phenotype. . [21] -HC gp-39 was isolated using PCR and western blotting.
-Northern blotting and RT-PCR was used to detect HC gp-39 mRNA both in synovial membranes and in cartilage acquired from patients with RA.
• Purified and sequenced HC gp-39 has regions similar to three other mammalian secretory proteins.
• HC gp-39 mRNA was present in articular chondrocytes.
Baeten D et al, 2000 [14] -Immunohistochemistry and flow cytometry were used to study expression of HC gp-39 in synovium and PBMCs.
-Synthesis and secretion of HC gp-39 were evaluated by RT-PCR and ELISA.
• HC gp-39 expression was more significant in RA patients compared to spondyloarthropathy patients and healthy controls.
• gp-39 level is correlated with the degree of joint destruction in RA.
Kirkpatrick RB et al, 1997 [15] -In situ hybridization was used to detect HC gp-39 mRNA in macrophages obtained from ST of 5 RA patients.
• HC gp-39 is expressed in primary human macrophages. • Its expression is closely related to differentiation of monocytes into macrophages in RA patients. Harvey S et al, 1998 [16] -Sandwich-type ELISA was used to compare serum Chondrex (HC gp-39) in healthy controls and in arthritis groups. -Assessment of sYKL-40, sIGF-I, and sIL-6 levels in 72 RA patients and 40 healthy subjects measured by ELISA.
• RA patients had significantly higher sYKL-40 levels than the healthy controls. • sYKL-40 levels were positively correlated with sIL-6 and CRP levels but negatively correlated with sIGF-I.
• sYKL-40 levels correlated with radiological score and severity of functional disability but not with joint pain. Johansen et al, 2001 [33] -Analysis of sYKL-40 samples from RA patients treated with DMARDs for 36 months measured by ELISA.
• sYKL-40 levels in ERA patients were significantly correlated with radiological progression determined by the Larsen score • sYKL-40 may provide information independent of CRP on disease activity. Johansen et al., 1999 [34] -1-year longitudinal study of sYKL-40 obtained from 156 RA patients by RIA.

•
In RA, sYKL-40 levels correspond to disease activity: levels decreased significantly in active patients who became clinically inactive, while levels increased in inactive patients who developed active RA. Volck et al., 2001 [35] -Analysis of serum and synovial fluid YKL-40and other biochemical markers determined by RIA and ELISA.
• YKL-40 was detected in the synovial membrane of RA and OA patients.

•
The number of YKL-40 positive cells in the inflamed synovial membrane is positively correlated with the severity of inflammation.
• Intra-articular glucocorticoid injection was followed by a decline in sYKL-40 levels. Vos K et al, 2000 [36] -Growth with 5 different HC gp-39 derived peptides followed by measuring multiplication of PBMC as well as recording disease activity score in RA, SLE, IBD, OA and healthy controls.

•
In autoimmune diseases including RA, HC gp-39 derived peptides are the objects of T cell immunity.

•
Especially HC gp-39 derived peptide 259-271 are known to be correlated with disease severity in RA. den Broeder A et al, 2002 [37] -Univariate and multivariate analyzes were used to assess the association between serum markers of radiological progression and cartilage and synovial membrane turnover (HC gp-39).
-Integrated measures of disease activity.

•
Long term TNF-alpha neutralization decreased the levels of HC gp-39, in RA patients. Peltomaa et al., 2001 [38] -Analysis of sYKL-40 in 52 early onset RA patients by ELISA during a 2-year prospective follow-up

•
Baseline sYKL-40 levels prior to anti-rheumatic therapy were significantly higher in ERA patients compared to healthy controls. • sYKL-40 is an inflammatory marker that correlates with disease activity, but does not have predictive value concerning radiographic progression or clinical course. Combe B et al, 2001 [39] -A cohort study of 191 RA patients conducted for three years.

•
The average odds ratio of YKL-40was 1.3(p=0.5) in predicting radiographic progression. Fusetti F et al, 2003 [40] -Analysis of crystallized structures of HC gp-39 by using the hanging drop vapor-diffusion method with recombinant HC gp-39.

•
HC gp-39 is apparently eligible autoantigen of RA and though its exact physiological function is unexplained. -Repeated measurements of pIL-6, pVEGF, and sYKL-40 for 25 Danish RA patients during treatment.

•
In treatment of RA, 18 of 25 patients had significantly decreased pIL-6, pVEGF, and sYKL-40 which were significantly elevated before treatment. Tsark EC etval., 2002 [44] -Blood monocyte-derived dendritic cell and macrophage were incubated with native human CII, HC gp39, and synovial fluid from RA patients. and the comparison with T cell hybrids produced by immunizing DR4-transgenic mice with CII 259-263 and HC gp39 263-275 peptides.

•
Human ex vivo differentiated dendritic cell and macrophage that were incubated with synovial fluid from RA patients differentially presented CII and HC gp39 MHC II epitopes.
• CII and HC gp39 might be important autoantigens in RA immunopathology and show different mechanisms of their presentation by dendritic cell and macrophage. Boots AM et al., 2007 [45] -Competition binding assay was used to map epitopes that strongly bind to T cell hybridomas obtained from HC gp-39-immunized HLA-DR4 transgenic mice.

•
Anchor variant peptide can modify antigen-specific, pro-inflammatory response in HLA-DR4 transgenic mice in a dampening fashion. -MIA and HC gp-39 levels in SF and ST were determined by enzyme-linked immunosorbent assay and immunohistochemistry.
• HC gp-39 and MIA were found in the SF and ST of patients with arthritis, and they were mainly presented by HLA-DR myeloid dendritic cells and B cells.

Authors, year
Study design Main study findings -IL-2 production of T cell hybridomas generated by immunizing DR4-transgenic mice with HC gp-39 or MIA were used to study presentation of MIA and HC gp-39 by SF cells. Li et al., 2017 [47] -In vivo study using recombinant human YKL-40.
• IL-1β is the most important factor involved in the inflammatory process in RA • TNF-α might induce secretion of YKL-40 from chondrocytes.

•
Serum and synovial YKL-40, IL-1β, and TNF-α levels are strongly correlated, suggesting that these molecules together play an important role in RA pathogenesis and disease activity.

•
No correlation exists between YKL-40 and IL-6 levels. Nielsen et al., 2011 [49] -Analysis of serum and whole blood samples from 308 RA patients and 605 healthy blood donors.

•
The g.-131 (C > G) promoter SNP in the CHI3L1 gene is strongly associated with the serum concentration of YKL-40 in both RA patients and in healthy controls.

•
The g.-131 (C > G) promoter SNP does not appear to be a direct risk factor of RA itself. Baeten D et al., 2004 [50] -Synovial biopsy samples were gathered for diagnostic evaluation from 154 patients. -Two investigator-initiated RCTs performed on treatment-naïve ERA patients.
• They did not appear to be predictive of either clinical remission or radiographic progression. Väänänen et al., 2017 [54] -RCT (NEO-RACo study) on 99 ERA patients undergoing DMARD therapy for 26 weeks.

•
Baseline pYKL-40 levels in ERA patients showed a positive correlation with disease activity as well as with IL-6 and MMP-3 levels. • During DMARD treatment, YKL-40 levels decreased significantly, which appears to be directly related to the anti-rheumatic effect of DMARDs.
• YKL-40 could potentially be used as a biomarker of disease activity in both ERA and RA patients undergoing active DMARD treatment. Turkyilmaz et al., 2013 [55] -Cross-sectional study of 42 ERA patients and 35 healthy subjects.
-Evaluation of arterial stiffness by cf-PWV and IMT-C by carotid ultrasonography.

•
The patients' sYKL-40 levels were strongly correlated with both cf-PWV and DAS28. • sYKL-40 levels could reflect early atherosclerosis, a major factor that contributes to mortality, as well as disease activity in ERA patients.

Kazakova et al., 2012 [56]
-Comparison of YKL-40 levels in serum and synovial fluid and ultrasonographic findings obtained from 25 RA patients and 40 healthy subjects.

•
The Bulgarian RA patients were found to have significantly higher sYKL-40 levels than the healthy subjects.

•
There is a positive correlation between YKL-40 in serum and synovial fluid and sonographic parameters. Knudsen et al., 2008 [56] -Analysis of biochemical measurements, radiographs, and MRI images obtained from 51 ERA patients and 21 PA patients.
• pIL-6, sYKL-40, pVEGF, CRP, and ESR were elevated in RA patients but not in PA patients. • Only pIL-6 was related to treatment response and progressive erosive disease in ERA; sYKL-40 showed no association. Bakker M et al., 2012 [57] -Twelve biomarkers, including YKL-40, were used to calculate the MBDA score of RA patients.

•
YKL-40 was one of the markers of the MBDA score designed to measure RA disease activity.

•
The MBDA score showed a significant correlation in measuring disease activity of RA. Tanaka Y et al., 2014 [51] -Analysis of HC gp-39 serum levels and mRNA expressions using ELISA and RT-PCR in GPI-induced arthritis.
• Addition of recombinant HC gp-39 blocked GPI-specific T cell proliferation and cytokine production, which means HC gp-39 in CD4+ T cells might play a regulatory role in early RA. Knudsen et al., 2006 [58] -1-year pilot study of sYKL-40 in 20 RA patients undergoing 52 weeks of infliximab and concomitant methotrexate therapy

•
High baseline pIL-6 is significantly related to radiographic joint destruction progression.
• It is unknown whether sYKL-40, independent of ESR, can provide information on joint destruction and disease activity in RA patients being treated with infliximab. Syversen et al., 2009 [59] -Cohort study of 238 RA patients followed for 10 years.
-Analysis of serum samples and radiograph data.

•
In contrast to sCTX-1 levels, sYKL-40 levels are weakly associated with radiographic progression and inflammation in RA.
• sYKL-40 will likely be much less useful as a prognostic marker of RA. Landewé RB et al., 2010 [60] -A phase II, 13-week multicenter, double-blinded RCT on patients with RA to receive either intranasal applications of placebo. or fully human, recombinant HC gp-39 (Org39141) in differential doses once a week.

•
Intranasal administration of Org39141 was proven to be safe but showed no efficacy over placebo. Wolvers DA et al., 1999 [61] -Mouse model was intra-nasally administered with OVA and HC gp-39 for tolerance induction.
-Lymph nodes were surgically removed and re-transplanted to see importance of lymph nodes in tolerance induction.

•
Certain lymph nodes that drain the nasal mucosa are essential in intranasal tolerance induction of model Ag (OVA) and HC gp-39.

Correlation between YKL-40 and pathogenesis of rheumatoid arthritis
YKL-40 is an important autoantigen in RA immunopathology. Several researches showed locally elevated concentration of YKL-40 in serum and synovial fluid of RA patients, suggesting that YKL-40 plays an important role in RA pathophysiology [43][44][45]. The immune reaction induced by YKL-40 differs, showing anti-inflammatory phenotypes in healthy people versus proinflammatory phenotypes in half of the RA patients [20]. mRNA for YKL-40 was identified specifically in human articular chondrocytes and liver [21] and overexpressed in the synovium and peripheral blood mononuclear cells (PBMCs) from RA patients [14]. YKL-40 expression is closely related to differentiation of monocytes into macrophages which produce higher levels of YKL-40 in rheumatoid synovium [15]. YKL-40 binds to the HLA-DR4 peptide-binding motif which leads to mononuclear cell proliferation [16,22]. Intracellular HLA-DM also has a crucial role in presenting YKL-40 to CD4+ T cell through "epitope editing" [23]. Moreover, the APCs presenting the immunodominant epitope of YKL-40 mainly appeared at the primary disease sites [24]. One phase 1 study reported that HLA-DR4B1-Org3660l, a peptide derived from YKL-40, inactivated T cell response and induced immune tolerance in RA patients [25]. It also indicates that YKL-40 has an association with RA pathogenesis.
Endogenous processing and major histocompatibility complex (MHC) II-mediated presentation of YKL-40, play major roles in the pathophysiology of RA [46]. Human ex-vivo differentiated DR4+ dendritic cells and macrophages, which were similar to synovial joint APC, were also capable of MHC II-presentation of YKL-40 epitopes [44]. A study using binding assay visualized the constitution of HLA-DR4 (B1*0410)/YKL-40 complex and showed that this can modify antigen-specific, pro-inflammatory responses in HLA-DR4 transgenic mice [45]. Not only inducing T cell proliferation, YKL-40 also activates the FAK/PI3K/Akt pathway, thereby inducing the production of IL-18 in osteoblasts and the inhibition of miR-590-3p. This, in turn, stimulates endothelial progenitor cell (EPC) angiogenesis, subsequently promoting inflammation, a process critical to the pathogenesis of RA [47]. Serum and synovial YKL-40 levels were strongly correlated with the levels of other inflammatory cytokines like IL-1β and TNF-α, suggesting that these molecules together play an important role in RA pathogenesis and disease activity [48]. Moreover, Petersson et al. (2006) also reported that arginine-vasopressin (AVP) and parathyroid hormone-related peptide (PTHrP) stimulated YKL-40 secretion particularly in RA chondrocytes [26,32], while Nielsen et al. suggested the association between the g.-131 (C > G) promoter SNP in the CHI3L1 gene and the serum concentration of YKL-40 [49]. Proposed pathogenic mechanisms of YKL-40 in RA are demonstrated in Figure 1.

Diagnostic marker
Since the serum and synovial fluid YKL-40 levels in RA patients were higher than those in the healthy controls, YKL-40 seemed to be a useful diagnostic marker for RA [16,27]. In one study which synovial biopsy samples are obtained from 154 patients, presentation of specific YKL-40 peptides in the context of class Ⅱ MHC was a highly specific histopathologic marker with a PPV of >90% for atypical RA patients [50].
However, mere serum or synovial YKL-40 level has a significant limitation to be an effective diagnostic tool since it cannot differentiate RA from other joint inflammatory diseases [28]. Production of YKL-40 might be a just response to an altered tissue environment, since YKL-40 plays a role in cartilage-remodeling process [16]. Moreover, some argued that YKL-40 values have a limited efficacy as a diagnostic marker. Harvey et al. (2000) suggested that serum YKL-40 level does not provide any additional information that would not be attainable by means of conventional biochemical measurements of disease activity like ESR or CRP levels [29]. Sekine et al. (2001) also suggested that it is not a sensitive marker for RA diagnosis as YKL-40 was detected in only 1 of 87 RA patients [30].
In contrast to mere expression of YKL-40, monoclonal antibody 12A staining in the synovial membrane was highly specific for RA. Monoclonal antibody 12A, which acts directly against HLA-DR4/HC gp-39 263-275 complex, can be a novel immune-pathologic diagnostic tool for RA [24]. Steenbakkers et al. showed that monoclonal antibody 12A has the highest specificity for the complex among 5 monoclonal antibodies that bind to HLA-DR4/HC gp-39 263-275 complex, and therefore can be used to effectively detect MHC/peptide/TCR complexes in the synovium of RA patients [31]. Also, in review paper by Johansen et al., the author suggested YKL-40 as a potential diagnostic marker of RA because of higher YKL-40 elevation in RA rather than OA compared to healthy people [18]. However, there are few papers about YKL-40 as a diagnosis marker. Therefore, whether YKL-40 can be used to diagnosis in RA should be further studied.

Correlation with disease activity
There may be various criteria for determining disease activity, such as DAS28, inflammation, radiological score (Larsen score), functional disability, erosion which means joint destruction, and clinical course. In order to explain the relationship between YKL-40 and disease activity, it is necessary to explain whether or not there is a significant relationship with YKL-40 level in each criterion. In addition, whether or not YKL-40 is a marker of disease activity also depends on whether it is a baseline value or a value that reflects treatment response. In fact, Table 1 summarizes many studies explaining the relationship between YKL-40 by distinguishing the baseline state from the treatment response state.
The association between YKL-40 and RA disease activity have been studied primarily via the measurement of YKL-40 levels in the serum (sYKL-40). sYKL-40 levels are increased significantly in RA patients compared to healthy controls [16,[32][33], inactive RA groups [16], or in inactive patients who developed active RA [34]. YKL-40 levels have also been studied in the synovium, which have shown an increase of YKL-40 levels in the synovial fluid [35] or an increase in the number of YKL-40 positive cells in the synovial membrane [27]. Correlations have been found between disease activity and sYKL-40 levels in RA patients [27,36], and even between disease activity and the immune response to YKL-40 levels in vitro [37].
In one study, sYKL-40 level was positively correlated with radiological scores used to assess joint destruction, but not with joint pain or swelling [32]. Similarly, YKL-40 expression in the synovial tissue was reported to have correlation with joint destruction [14].
Only one paper was an in vitro experiment [51]. The other papers were in vivo experiments [43,50,[52][53][54][55][56][57]. One of the seven papers evaluated the correlation between sYKL-40, cf-PWV and IMT-C as well as DAS28 [54]. Also, one of the seven papers evaluated sYKL-40 and synovial thickening and vascularization scores [55]. In the early phase of GPI-induced arthritic mice study, high sYKL-40 was detected [51]. In two randomized controlled trial (RCT) studies, sYKL-40 level had a positive correlation with early rheumatoid arthritis (ERA) disease activity [52,53]. In a cross-sectional study including 42 ERA patients and 35 healthy patients, sYKL-40 levels were highly correlated with DAS28 [54]. Another study including 51 ERA patients and 21 polyarthritis (PA) patients, increased YKL-40 in serum was observed only in RA patients [56]. In one study where 25 Danish RA patients got repeated measurement of various serum markers during treatment, YKL-40 serum was significantly elevated in RA patients [43]. In one study, the MBDA score where twelve markers including serum YKL-40 are combined was significantly correlated with RA disease activity [50]. Interestingly, a study evaluating arterial stiffness by cf-PWV and IMT-C by carotid ultrasonography [54] showed that sYKL-40 was highly correlated with cf-PWC and IMT-C as well as DAS28. These results suggest the possibility of early detection of atherosclerosis using sYKL-40 [54]. Correlation between YKL-40 level and synovial thickening and vascularization score was observed in one study [55]. Therefore, YKL-40 serum could be used as a disease activity marker of RA.

Correlation with long-term prognosis
One study showed that sYKL-40 levels have a correlation with RA progression (assessed by the Larsen score and the extent of bone erosions) [32], while other studies suggested no association between sYKL-40 and RA progression [36,39]. Four papers using YKL-40 as a marker are available for long-term prognosis of RA. The others evaluated YKL-40 as a long-term prognosis marker in RA patients [52,58,59]. Four papers had a consensus that sYKL-40 seemed to be not a long-term prognosis marker and to be not associated with radiographic progression in RA [18,52,58,59]. In one cohort study where 238 RA patients are followed for 10 years, sYKL-40 level didn't show any correlation in 5-and 10-years radiographic progression [59]. In 1-year pilot study where 20 RA patients got infliximab and concomitant methotrexate therapy for 52 weeks, YKL-40 did not show correlation with radiographic joint destruction [58]. Also, in two investigatory-initiated RCTs on naïve ERA patients, YKL-40 was not useful in predicting patient's prognosis in radiographic or clinical aspects [52]. Therefore, YKL-40 did not seem to be a long-term prognosis marker, nor was it associated with radiographic progression in RA.

Effect of YKL-40 suppression in rheumatoid arthritis
Three papers evaluating YKL-40 suppression are available for RA treatment [51,60,61]. Injecting YKL-40 in an early phase of GPI-induced arthritis mice caused decreased antigen-specific T cell proliferation and cytokine production [51]. By using surgical procedures, a study investigated the relationship between and intra-nasal YKL40 injection in lymph nodes and tolerance induction in mice [61]. They concluded that lymph nodes that drain the nasal mucosa is crucial in tolerance induction in mouse model AG (OVA) [61,67]. This showed the possibility of injecting YKL-40 as RA treatment [51,61]. However, in a phase II, double-blinded RCT study where RA patients received internal YKL-40 therapy, there was no significant change in DAS28 compared to placebo group [60]. Therefore, other strategies including changes in way to convey YKL-40 or combination of other immune regulators should be studied.

Positive correlation between psoriasis and YKL-40
Multiple biomarkers have been used for the diagnosis of psoriatic diseases. Increase in white blood cell count, C-reactive protein (CRP), and several cytokines such as IFN-γ were found in patients with psoriasis; however, these markers alone lack sensitivity and specificity for evaluating the severity of disease [75]. Several studies recently showed the possibility of YKL-40 as a new diagnostic marker for psoriasis and its possible correlation with disease severity. Imai Y et al. first observed the elevation of YKL-40 in patients with psoriasis vulgaris compared with control group, and the difference was more obvious in patients with generalized pustular psoriasis, a more severe inflammatory form of psoriasis [64]. sYKL-40 level was also suggested to show positive relationship with the severity of skin lesions in patients with psoriatic arthritis, a type of arthritis that occurs in psoriasis patients [65]. Ahmed S et al. stated that sYKL-40 level might be used for evaluating disease activity and angiogenesis in patients with psoriasis [66]. YKL-40 could also be helpful for diagnosing and monitoring patients with psoriatic arthritis as an inflammatory marker [67,68]. Salomon et al. stated that an increased YKL-40 level reflects the presence of systemic inflammation rather than cutaneous lesions [67] and found its correlation with the severity of psoriatic arthritis, which was scored with 28-joint Disease Activity Score (DAS-28) in the study [68]. There were conflicting findings regarding YKL-40's legitimacy in evaluating treatment response. Baran A et al. stated that YKL-40 might be interpreted as a marker of psoriasis, but not for evaluating metabolic condition or efficacy of treatment [69]. However, decreased sYKL-40 level after narrow-band ultraviolet B phototherapy was observed in psoriasis vulgaris patients [70]. -Diagnosis YKL-40 in 18 PsA patients and 29 controls based on CASPAR criteria using enzyme-linked immunoassay kit. -60 psoriasis patients and 30 healthy controls were evaluated for endothelial function and serum YKL-40.

•
The increased level of YKL-40 in psoriasis might be related to ED.   Although HC gp-39 thought to be an index of cartilage damage initially, it should be considered as an index of chondrocyte activation due to subsequent studies. Vos K et al., 2000 [80] -Growth with 5 different HC gp-39 derived peptides followed by measuring multiplication of PBMC as well as recording disease activity score in RA, SLE, IBD, OA and healthy controls.
• In autoimmune diseases including SLE, HC gp-39 derived peptides are the objects of T cell immunity. • The T-cell response to the various HC gp-39 derived peptide was low in the SLE patients. Furthermore, sYKL-40 level is also related to vascular defects in psoriasis patients. YKL-40 level is positively correlated with carotid intima-media thickness and defective aortic elasticity in patients with psoriasis [71]. According to Erfan G et al, YKL-40 might be associated with endothelial dysfunction in psoriasis and could be applied for managing cardiovascular diseases in high-risk psoriasis patients [72].

Negative or no correlation between psoriasis and YKL-40
There are two studies that show negative or none correlation between psoriasis and YKL-40. Ataseven et al. reported that the YKL-40 levels of psoriasis group are not significantly different from the levels of healthy controls [73]. Jensen et al. revealed that there is no correlation between YKL-40 and psoriasis severity [74]. However in psoriatic arthritis group, which is a type of arthritis that occurs in psoriasis patients, YKL-40 level had association with disease severity and treatment response [74].

Systemic lupus erythematosus (SLE) and YKL-40
SLE is an autoimmune disease that involves multiple organs including the skin, kidney, brain, and joints [76]. Majority of patients are women of childbearing age, which accounts for 90% of total SLE patients [77]. There are several classic markers that have been used to estimate disease activity in SLE patients, such as anti-dsDNA titer and measurements of C3, C4, CH50 [78]. A total of 3 associations between SLE and YKL-40 were identified from a PubMed search (Table 3) [36,79,80].
Two studies which compared YKL-40 levels between SLE patients and healthy controls showed concordant findings. Vos et al. first revealed that SLE patients have higher YKL-40 levels than healthy controls [36]. However, YKL-40 levels in SLE patients were lower compared to RA patients and did not correlate with disease activity [36]. Wcisło et al. also showed that average plasma levels of YKL-40 is about twice higher in the SLE group than in controls [79]. Likewise, YKL-40 had no correlation with disease activity or the severity of joint involvement [80]. There was an in vitro study that showed reduced reactivity of T cells from SLE patients in response to YKL-40 [80]. Vos et al. incubated T-cells in the settings of five types of YKL-40 and compared its growth rate between different disease entities including SLE [80]. They studied the responses of T-cells to YKL-40 in patients with various inflammatory conditions [80] and found that T-cells from RA patients showed a proliferative response to YKL-40 [80]. In contrast, T-cells from SLE patients showed low response to YKL-40 [80].

Behcet's disease (BD) and YKL-40
BD is a systemic inflammatory disorder with recurrent oral ulcers, genital ulcers, eye lesions, and skin lesions [81]. It can also involve joints, central nervous system, gastrointestinal tract, and large vessels [81]. Although the pathogenesis of BD is not fully known, neutrophilic hyperactivity and overproduction of pro-inflammatory cytokines and reactive oxygen species (ROS) are considered to be major mechanism of the disease [81,82,83]. PubMed search identified two studies describing the association between BD and YKL-40 (Table 4) [84,85].
Seo et al. first reported the correlation between BD and YKL-40 [84]. By comparing sYKL-40 levels of control group and inactive/active BD patients, they revealed that YKL-40 was significantly higher in BD patients [84]. In addition, patients in the active BD group showed an elevation in sYKL-40 levels compared with inactive BD patients [84]. Since YKL-40 levels showed a positive correlation with disease activity, they proposed YKL-40 as an alternative marker to monitor disease activity in BD patients [84]. Bilen et al. also reported YKL-40 elevation in BD patients [85]. However, this study failed to show the association between YKL-40 levels and disease activity, which is inconsistent with the previous study [85]. Therefore, further research is required to clarify the correlation between YKL-40 and BD activity.

Correlation between YKL-40 and disease activity in inflammatory bowel disease
A link between YKL-40 and the activity of IBD lies on the process of fibrosis and inflammation during the natural course of IBD [86]. Ten papers reported increased levels of YKL-40 associated with the severity of the disease [86][87][88][89][90][91][92][93][94][95]. In both Crohn's disease and ulcerative colitis, the sYKL-40 levels were significantly correlated with C-reactive protein level and disease activity [86]. In Crohn's disease, sYKL-40 level was shown higher in patients with intestinal strictures than in those without intestinal strictures [87]. Vind et al. found that sYKL-40 level is increased in 40-50% of ulcerative colitis and Crohn's disease patients with active disease, and the level was also increased in 30% of patients with Crohn's disease which is clinically inactive [88]. Ytting H et al. reported a case where s YKL-40 level is increased as an appearance of Sweet's syndrome in a patient previously diagnosed with ulcerative colitis, sYKL-40 level acted as a marker of disease activity of Sweet's syndrome [89]. Increased sYKL-40 level was also discovered as the marker for demonstrating articular involvement in inflammatory bowel disease [90]. Punzi L et al. discovered a similar result showing sYKL-40 level was only increased in IBD patient with arthritis, IBD patients without arthritis showed no difference from healthy controls [94]. Among the peptides derived from YKL-40, the human cartilage glycoprotein-39 263-275 level was increased in IBD [91]. The fecal YKL-40 level was found to be a marker to assess endoscopic ulceration level in IBD [89]. The fecal sYKL-40 level was also correlated with the severity and disease activity in IBD [93]. In Crohn's disease, sYKL-40 is an autoantigenic target, IgA and secretory IgA antibodies against sYKL-40 level can serve as a marker facilitating the serological diagnosis [95].

Correlation between YKL-40 and development of cancer in inflammatory bowel disease
The YKL-40 expression in colonic epithelial cells is a biomarker for neoplastic changes in IBD [95]. In a mouse model, the YKL-40 induced cell proliferation and survival while involving down-regulation of the pro-apoptotic S100A9 protein, which promoted tumorigenic changes in the colon and led tumor cells survive and proliferate [97]. In another mouse model, colonic epithelial cells with high levels of YKL-40 expression exhibited malignant transformation in vivo when exposed to azoxymethane, a well-known colonic carcinogen [98]. Higher expression of the YKL-40 was found in the distal colon compared to the proximal colon in mice, resulting in higher chances of tumorigenesis in the distal colon [99].

YKL-40 as biomarker of other human diseases with autoimmune mechanism.
The serum YKL-40 levels were higher in the multiple sclerosis (MS) group than in control, and which levels in the patients with relapsing remitting multiple sclerosis (RRMS) were correlated with the patients' expanded disability status scale (EDSS) scores and ages. No relationships were determined between the serum YKL-40 levels and the other variables. The findings from this study suggested that YKL-40 may be a useful marker for the inflammatory process of MS [100]. Serum levels of YKL-40 were significantly higher in 40 female patients with systemic sclerosis compared to 14 healthy female controls. In contrast, miR-214 expression in plasma from SSc patients was significantly downregulated compared to controls which shows that the binding of YKL-40 and miR-214 is involved in the mechanism of inflammation and fibrosis [101]. YKL-40 was higher in the poorly controlled symptom and exacerbation group and in patients with non-atopic asthma compared with stable asthma. YKL-40 appears to increase in proportion to the degree of inflammation in diseases with an immune mechanism [102]. Compared to neurological controls, increased YKL-40 levels were detected in sCJD and Alzheimer's disease (AD) but not in vascular dementia (VaD) or in dementia with Lewy bodies (DLB)/Parkinson's disease dementia (PDD). Further, two independent patient cohorts were used to validate the increased CSF YKL-40 levels in Creutzfeldt-Jakob disease (sCJD). YKL-40 is a disease-specific marker of neuroinflammation showing its highest levels in prion diseases [103]. -Comparison of plasma levels of chitinase-3-like 1 protein and its association with malondialdehyde in 51 Behcet's disease patients and 28 healthy controls by SPSS 20.

•
Chitinase-3-like 1 protein might be associated with Behcet's disease. • The levels of malondialdehyde had no significant correlation with Chitinase-3-like 1 protein.

Authors, year
Study design Main study findings Strong Correlation between YKL-40 and disease activity Koutroubakis I. E et al., 2003 [86] -Detection of sYKL-40 values in 94 UC, 85 CD, 23 non-IBD intestinal inflammation patients and 70 healthy controls by using ELISA.

•
Mean sYKL-40 concentrations increased in UC and CD patients than in healthy controls (P<0.0001, respectively), but not significantly higher than in non-IBD intestinal inflammation patients.

•
Localization also attributed to higher YKL-40 levels as CD in ileum showed higher levels of YKL-40 than in ileocolon or colon.
• YKL-40 values in CD were highly detected in the healthy controls (P=0.000) and increased as disease activity increased (r=0.681, p=0.000). • YKL-40 concentrations in severe active UC increased than in inactive UC and healthy controls(p<0.001). • Median YKL-40 concentrations in UC patients and clinical activity were correlated. • YKL-40 concentrations in severe active CD were also highly detected than in healthy controls(p<0.001), but not significantly higher than in inactive CD patients. • YKL-40 levels were highly detected in JIBD patients (P =0.000003) than in IBD patients, but there were no differences for CRP and SAA values. -Fecal calprotectin and CHI3L1 was measured in 86 adult IBD patients and compared to CDEIS score in CD patients or MES in UC patients, evaluating both level as biomarker of activity in IBD.

•
Both fecal CHI3L1 and calprotectin levels correlated with endoscopic activity score in CD and UC.
• Fecal CHI3L1 seem to be reliable biomarker in assessing endoscopic activity of IBD.
Aomatsu T et al., 2011 [93] -Analyzed the level of CHI3L1 and calprotectin in fecal samples from pediatric UC, CD patients and healthy control by ELISA.
• Active IBD patients' fecal CHI3L1 levels were higher compared with healthy control, and were correlated with disease activity. Punzi L et al., 2003. [94] -Comparison of serum HC gp39 levels and CRP in 58 IBD-nonA, 63 IBD-A, and in 20 healthy controls. IBD patients were also divided into aIBD and naIBD.

•
The serum HC gp39 values in IBD-A patients were highly detected than in controls (p<0.01) and IBD-nonA (p<0.001) patients.

•
Otherwise, the values between controls and IBD-nonA were not different.

•
The serum CRP levels were high in IBD-A(p<0.01) and IBD-nonA(p<0.05) than in the controls.
• CRP levels were highly detected in aIBD-nonA than in naIBD-nonA (p<0.05), the levels between aIBD-A and naIBD-A were not different. Deutschmann   • YKL-40 serum levels were significantly higher in patients compared to controls (p=0.0042). In contrary, miR-214 expression in plasma of SSc patients was significantly downregulated compared to controls (p=0.0058).

Krzysztof et al., 2015 [102]
The study group comprised 167 patients, including 116 women and 51 men aged 18-88 years with chronic asthma to investigate the role of YKL-40 as a possible marker of asthma.

Concluding remarks and future perspectives
The biological function of YKL-40 glycoprotein, also known as chitinase-3-like protein 1 (CHI3L1) [1] or human cartilage glycoprotein 39 (HC gp-39) [3], is not that clear, but it is speculated to have some connection with inflammatory reactions and autoimmune diseases [9,11]. We reviewed 51 articles that discussed the association of YKL-40 with RA, Psoriasis, SLE, BD and IBD. Results highlight the value of YKL-40 as biomarker of autoimmunity and rheumatic diseases.
In the first place, we found 21 articles discussing associations between RA and YKL-40. Two pathogenesis of RA related to YKL-40 are unknown; acting as an autoantigen or an inducing factor of IL-18 expression [44][45][46][47]. There were two articles including a review article on using YKL-40 as a diagnostic marker of RA [18,50]. Both showed the possibility of YKL-40 as diagnostic marker of RA, and one of them demonstrated the presentation of specific YKL-40 peptides in the context of MHC-II [50]. As a result, YKL-40 could be assumed as a diagnostic marker of RA, but further studies are needed. There were nine articles investigating the correlation of RA disease activity with YKL-40 [18,43,[52][53][54][55][56][57], all of which demonstrated a correlation between YKL-40 level and disease activity of RA. Furthermore, one of the papers that used cf-PWV and IMT-C as disease activity indices presented the potentiality of YKL-40 as an early detection marker of atherosclerosis in RA patients [54]. Including a review article, there were five articles on evaluating s YKL-40 level during RA treatment [18,43,53,56,58]. They revealed YKL-40 levels decreased after therapy by DMARD, infliximab or anti-TNF-alpha agents. Notably, in one of those studies, YKL-40 level was decreased only in patients that achieved remission, but not in all RA patients [56]. As follows, YKL-40 could be used to evaluate treatment response. We could find four articles examining the availability of YKL-40 as long-term prognosis of RA, including a review article [18,52,68,59]. But there was no evidence at all that sYKL-40 level is associated with radiographic progression and long-term prognosis. Lastly, three papers are available for RA treatment using YKL-40 suppression [51,60,61]. This protocol had some controversy, so further studies are warranted.
There were 11 studies on relation between psoriasis and YKL-40. Five of them imply YKL-40 could be used for evaluating disease activity because YKL-40 level is elevated in a more severe form of inflammation than in a less severe form or in controls and in systemic inflammation rather than in cutaneous lesions [64][65][66][67][68], though Jensen et al. argued that there is no correlation [73]. Furthermore, there were two articles which demonstrate YKL-40 level is also related to vascular defects in psoriasis patients [72]. Two of them investigated YKL-40's legitimacy in evaluating treatment response, but their results were inconsistent. There were two studies showing no correlation of sYKL-40 level with presence of disease and psoriasis severity [74]. However, in the psoriatic arthritis group, YKL-40 level had an association with disease severity and treatment response [74].
Two studies which compared YKL-40 levels between SLE patients and healthy controls showed concordant findings [36,79]. They revealed that SLE patients had higher YKL-40 levels than healthy controls [79], but YKL-40 level had no correlation with disease activity or the severity of joint involvement [79]. There was an in vitro study that showed reduced reactivity of T cells from SLE patients in response to YKL-40 [80].
Two studies describing the association between BD and YKL-40 were identified [84,85]. Two studies revealed that YKL-40 was significantly higher in BD patients [82]. But it was controversial whether YKL-40 level is related to BD activity [84,85].

Conclusion
In conclusion, this systemic review indicates that YKL-40 is a very promising molecule as a biomarker of RA both in diagnostic and therapeutic aspects. Further studies on YKL-40 for autoimmune diseases as Psoriasis, SLE, BD, IBD and other diseases with immune mechanism are essential to fully elucidate its clinical significance and utility.

Consent for publication
All the authors checked and gave their approval of this version to be published.

Funding
No financial support was provided for research conduct and/or preparation of the article.

Author contributions
All authors made substantial contributions to all of the following; (1) conception and design of the study, data acquisition, or analysis and interpretation of data; (2) drafting or critical revision of the article for intellectual content; and (3) final approval of version to be submitted.