Anti‐CD146 Autoantibodies: The First Biologic Markers Associated With Occupational Exposure in Systemic Sclerosis

Abstract

Objective

Systemic sclerosis (SSc) is a severe autoimmune disease, with occupational exposure being a significant risk factor. Because CD146 was recently identified as a driver of fibrosis in SSc through regulation of the Wnt/reactive oxygen species interplay, we hypothesized that it is a major autoimmune target in this disease.

Methods

We developed an in‐house ELISA test to detect anti‐CD146 autoantibodies (AACD146), which was confirmed by immunoprecipitation and Western blotting. AACD146 positivity was assessed in the sera of patients with SSc compared with healthy controls. A validation cohort of workers exposed to asbestos or silica was evaluated and compared to patients with pulmonary cancer and healthy controls without any occupational exposure.

Results

Detection of AACD146 was assessed by ELISA and confirmed with Western blot and an absorption test. In the first cohort, the prevalence of positive AACD146 was significantly higher in patients with SSc (n = 14 of 93; 15%) than in controls (n = 2 of 40; 5%). Interestingly, among patients with SSc, positive AACD146 were associated with male sex (P = 0.04) and occupational exposure to silica (P = 0.009), with a sensitivity of 57% and specificity of 88% for occupational exposure. Results were confirmed in a validation cohort, in which positive AACD146 were found in 57% (n = 13 of 23) of patients with professional exposure. The frequency of AACD146 was significantly higher compared to controls (P = 0.03) and to patients with a history of cancer (P = 0.02).

Conclusion

We demonstrated that AACD146 are detectable in patients with SSc and are linked to male workers with occupational dust exposure. AACD146 are the first biomarkers associated with occupational exposure in SSc, with potential implications for preventive medicine.

INTRODUCTION

Systemic sclerosis (SSc) is a severe autoimmune disease characterized by vascular damage, autoimmunity, and fibrosis of the skin and internal organs, causing severe morbidity and premature mortality. SSc autoantibodies may be linked to specific clinical phenotypes: anticentromere with limited cutaneous SSc, anti‐topoisomerase I with diffuse cutaneous disease and interstitial lung disease (ILD), and anti‐RNA polymerase III with renal crisis and cancers. ¹ SSc pathogenesis involves genetic and environmental factors, including occupational exposure to solvents or silica, ² which are recognized as occupational hazards in some health systems. ³ Silica exposure induces oxidative stress, generating reactive oxygen species (ROS) that modify autoantigens and trigger autoimmunity. ⁴ , ⁵ We recently identified CD146 as a novel player in SSc fibrosis, regulating both ROS and the Wnt pathway. ⁶ CD146, an endothelial adhesion molecule, exists in membrane‐bound and soluble forms (sCD146), generated by membrane cleavage or splicing. ⁷ Physiologically, CD146 and sCD146 regulate angiogenesis and inflammation. ⁷ In a bleomycin‐induced SSc model, CD146 expression status modified skin fibrosis via regulation of the noncanonical Wnt pathway and ROS production, suggesting that oxidative stress modifies CD146 expression, making it a potential neoantigen. ⁸ , ⁹ In patients with SSc, lower sCD146 levels are associated with disease progression. ⁸ CD146 is absent in healthy dermal fibroblasts but is expressed on SSc dermal fibroblasts when stimulated by ROS. ⁸ These findings led us to hypothesize that CD146 plays a role in SSc and the development of a ROS‐mediated autoimmune response against CD146 in patients with SSc.

SIGNIFICANCE & INNOVATIONS.

• Anti‐CD146 autoantibodies (AACD146) are detectable in systemic sclerosis (SSc) and are associated with occupational exposure.

• Workers exposed to silica or asbestos show detectable AACD146 antibody levels.

• AACD146 is a new antibody for SSc diagnosis and may be evaluated in preventive medicine.

To this end, we assessed anti‐CD146 autoantibodies (AACD146) in patients with SSc sera compared to those of healthy controls. For further validation, we compared a second cohort of occupationally exposed workers with patients with pulmonary cancer and nonexposed controls.

MATERIALS AND METHODS

Patient selection

We conducted a retrospective monocentric study of patients with SSc from the Internal Medicine Department of Assistance Publique – Hôpitaux de Marseille, diagnosed according to the 2013 EULAR/American College of Rheumatology criteria ¹⁰ with local ethics approval and written consent. Control sera were collected from healthy donors via the French Blood Establishment (biobank DC 2012 − 1704; approval 2011‐A00095‐35). Clinical, radiologic, and biologic data were extracted from medical records. For validation, a second cohort included asbestos‐ and silica‐exposed workers, patients with lung cancer, and unexposed controls recruited from the Rennes University Hospital (ethics approval ID‐RCB 2020‐A01990‐39).

Anti‐CD146 enzyme‐linked immunosorbent assay autoantibody detection

Microtiter plates were coated with recombinant human sCD146 (10 μg/mL, Biocytex), blocked with 10% phosphate‐buffered saline and fetal bovine serum for two hours at 37°C, and incubated with sera (1:100) or the s‐endo1 positive control (1:5,000), followed by a horseradish peroxidase (HRP)–conjugated anti‐human Fc antibody. After the addition of the 3,3′,5,5′‐tetramethylbenzidine substrate, absorbance was measured at 450 nm. AACD146 positivity was defined as a sample or control absorbance ratio >1.

Absorption experiments of anti‐CD146 antibodies with the enzyme‐linked immunosorbent assay plate

As previously described, ¹¹ positive AACD146 sera were incubated on CD146‐coated plates for two hours at room temperature, which was repeated up to six times. Absorption effectiveness was confirmed by anti‐CD146 enzyme‐linked immunosorbent assay (ELISA) performed on the absorbed sera.

Western blotting detection of anti‐CD146 antibodies

Recombinant sCD146 (1 μg) was run on sodium dodecyl‐sulfate polyacrylamide gel electrophoresis, transferred to a nitrocellulose membrane, and probed with healthy or AACD146‐positive sera (1:50). After incubation with an HRP‐conjugated secondary antibody, bands were visualized using enhanced chemiluminescence and analyzed with ImageJ.

Statistical analysis

Data are expressed as mean ± SEM. Statistical analyses were performed using Prism (GraphPad Software, Inc.): Mann‐Whitney U test for two‐group comparisons and one‐way analysis of variance with Dunn's test for multiple groups. P < 0.05 was considered significant.

Statement of ethics

Written informed consent was obtained in accordance with the Declaration of Helsinki, and the study was approved by the local ethics committee review board in Marseille. Control samples were selected from healthy blood donors at the French Blood Establishment. Serum remnants were used from a declared biobank (DC 2012‐1704). For the validation cohort, a written informed consent was obtained, and the study was approved by the Institutional Review Board (ID‐RCB 2020‐A01990‐39).

Data availability statement

The data and/or related materials that support the findings of this study are not publicly available due to French law restrictions because they contain information that could compromise the privacy of research participants. However, they are available from the corresponding author (JB) upon reasonable request.

RESULTS

Identification of AACD146 in patients with SSc

We initially investigated the presence of AACD146 in patients with SSc (n = 93) compared with controls (n = 40 healthy blood donors) by assessing their potential binding using an in‐house ELISA coated with recombinant sCD146. The prevalence of AACD146 was significantly higher in patients with SSc (P < 0.0001), with 15% testing positive (n = 14 of 93) compared with 5% of healthy controls (n = 2 of 40) (Figure 1A).

Figure 1.

(A) Comparison of proportion for positive AACD146 between patients with SSc and healthy blood donors (controls). Gray dots represent patients with occupational exposure. Data are expressed as optical density ratios, with positivity defined by a ratio >1. ***P < 0.001. (B) Absorption of AACD146 on wells coated with soluble CD146 for patients with SSc (n = 3) compared with healthy control. Absorption was assessed up to six times. Data are provided in optical density. *P = 0.03. (C) Western blot analysis on human sera (dilution 1:50) with and without AACD146 detecting sCD146. All blots are made from the same gel, but cropping was necessary to test the different conditions, as explained in the Materials and Methods section. The gel was blocked in tris‐buffered saline with tween 20‐5% bovine serum albumin for 1 hour at room temperature and then probed separately with human sera from two healthy blood donors (controls) and one patient with positive AACD146. Data are reported as a percentage of the mean of controls. (D) Comparison of positive AACD146 between exposed workers, patients with pulmonary cancer, and healthy controls in the validation cohort. Data are expressed as optical density ratios, with positivity defined by a ratio >one. *P < 0.05. AACD146, anti‐CD146 autoantibodies; OD, optical density; SSc, systemic sclerosis.

To confirm ELISA specificity, detection was performed following AACD146 absorption experiments on sCD146 (Figure 1B). A significant decrease in signal was observed in three positive samples (P = 0.03) with no change in the negative sample used as a control. Western blot analysis on sera from healthy donors (n = 2) and AACD146‐positive patients with SSc (n = 1) revealed recognition of recombinant sCD146 at the expected molecular weight of 100 kDa (Figure 1C).

Characteristics of patients with SSc depending on AACD146

To assess the clinical relevance of AACD146 in SSc, we evaluated their association with clinical and biologic characteristics in patients with SSc. In a first cohort, 93 patients were included, predominantly women (89%); the mean age at SSc diagnosis was 50 years (±14), with samples collected at a mean age of 59 years (±14). Most patients had limited cutaneous SSc (64%), with clinical complications including digital ulcers (37%), arthritis (42%), and ILD (33%). A history of cancer was noted in 10 patients (11%) including breast (n = 3), cutaneous (n = 3), uterus and ovarian, lung, and colon cancers. Occupational exposure was identified in 8% of patients (n = 7), primarily to crystalline silica (n = 6; 85%). AACD146 levels were not correlated with disease severity, as assessed by the Rodnan skin score (Pearson correlation, r = 0.287, 95% confidence interval −0.3512 to 0.7351, P = 0.38). Positive AACD146 results were observed in 14 patients with SSc (15%) and were associated with male sex (P = 0.04). AACD146 positivity also showed a significant association with occupational exposure (P = 0.009), with a sensitivity of 57% and specificity of 88% for exposure among patients with SSc. Among AACD146‐positive patients, four had a history of silica exposure (29%; 4 of 14), whereas in AACD146‐negative patients, only three (4%; 3 of 79) had occupational exposure histories, including silica (n = 2) and aluminum (n = 1). No association was found between AACD146 positivity and specific organ involvement such as pulmonary hypertension or cancer history. All data are available in Table 1. Regarding treatment in the SSc cohort, treatment data were extracted from medical records. In the AACD146− group, 14 patients (18%) received steroid therapy compared with 3 patients (21%) in the AACD146+ group, with no significant difference between groups (P = 0.7). With respect to immunosuppressive agents, 13 patients (16%) in the AACD146− group were treated, including with mycophenolate mofetil (n = 8), methotrexate (n = 2), rituximab (n = 1), azathioprine (n = 1), and an unspecified immunosuppressive regimen (n = 1). In the AACD146+ group, three patients (21%) received immunosuppressive therapy, all with methotrexate (P = 0.7).

Table 1.

Patients with SSc characteristics^*

Characteristics	Overall population (n = 93)	Positive AACD146 (n = 14)	Negative AACD146 (n = 79)	P value
Female sex, mean (SD)	83 (89)	10 (71)	73 (92)	0.04
Age at diagnosis, mean (SD), y	50 (14)	50 (14)	50 (13)	0.94
Age at sample collection, mean (SD), y	59 (14)	60 (17)	59 (20)	0.75
BMI, mean (SD) kg/m2	31 (58)	23 (4)	32 (63)	0.64
Smoking history, mean (SD)	25 (27)	5 (33)	20 (25)	0.51
SSc, mean (SD)
Diffuse	31 (33)	6 (43)	25 (31)	0.54
Limited	59 (63)	7 (50)	52 (66)	0.37
Sine scleroderma	3 (3)	1 (7)	2 (3)	0.39
Antinuclear antibody (≥1/160), mean (SD)	93 (100)	14 (100)	79 (100)	1
Anticentromere antibody	35 (38)	7 (50)	28 (35)	0.37
Anti‐topoisomerase I antibody	52 (56)	7 (50)	45 (57)	0.77
mRSS, mean (SD)	11 (9)	13 (11)	11 (9)	0.52
Digital ulcers or gangrene, mean (SD)	34 (37)	8 (57)	26 (33)	0.13
Interstitial lung disease, mean (SD)	31 (33)	3 (21)	28 (35)	0.37
FVC, n (%)	90 (24)	93 (13)	90 (26)	0.38
DLco, n (%)	52 (20)	57 (21)	51 (19)	0.33
Pulmonary hypertension, mean (SD)	25 (26)	1 (7)	24 (30)	0.1
Left heart disease, mean (SD)	3 (3)	0 (0)	3 (4)	1
GI tract involvement, mean (SD)	22 (24)	4 (29)	18 (23)	0.73
Renal disease, mean (SD)	0	0	0	1
Inflammatory arthritis, mean (SD)	39 (42)	6 (43)	33 (42)	1
Cancer history, mean (SD)	10 (11)	1 (7)	9 (11)	1
Occupational exposure, mean (SD)	7 (8)	4 (29)	3 (4)	0.009
Silica	6 (85)	4 (100)	2 (67)	0.42
Treatment, mean (SD)
Steroids	17 (18)	3 (21)	14 (18)	0.7
Immunosuppressive drugs	16 (17)	3 (21)	13 (16)	0.7

^* P values <0.05 are shown in bold. AACD146, anti‐CD146 autoantibodies; BMI, body mass index; DLCO, diffusing capacity for carbon monoxide; FVC, forced vital capacity; GI, gastrointestinal; mRSS, modified Rodnan skin score; SSc, systemic sclerosis.

External validation cohort confirming the association of AACD146 with occupational exposure

To confirm the association of AACD146 with occupational exposure, we investigated its presence in an external cohort of 23 workers with occupational exposure to mineral dusts. This cohort was exclusively composed of men, with a median age of 55 years at sample collection. None had a history of cancer or autoimmune disease, and only one tested positive for antinuclear antibodies. Four had ILD related to occupational exposure (17%). Occupational exposure included silica in seven individuals (30.5%), asbestos in nine (39%), and both in seven (30.5%) (Table S1). Positive AACD146 results were found in 57% (n = 13 of 23) of these workers. Comparison of patients with positive and negative AACD146 results showed no difference in age, duration, or type of exposure. However, there was a trend toward a higher prevalence of ILD in the AACD146‐positive group compared to the negative group (31% vs 0%; P = 0.1). AACD146 frequency in this population of workers was compared with that of healthy donors (n = 3) and patients without occupational exposure but with a history of pulmonary cancer (n = 7) because cancer can induce autoantibody production. ¹² Positive AACD146 results were significantly more frequent in patients with occupational exposure (n = 13 of 23) compared to those with a history of pulmonary cancer but no occupational exposure (n = 0 of 7; P = 0.02) and healthy donors (n = 0 of 3; P = 0.04) (Figure 1D).

DISCUSSION

We identified AACD146 as the first biologic marker for SSc linked to occupational exposure, marking the first identification of an autoimmune response against CD146. AACD146 are associated with male sex and not with a specific clinical phenotype of SSc but are significantly linked to occupational exposure. This was confirmed in a second cohort of exposed workers, showing significantly higher AACD146 in those exposed to silica or asbestos, with no significant difference between these exposures, in comparison to subgroups of controls without occupational exposure.

Exposure to silica is a significant risk factor for developing SSc, with a 28‐fold increase risk compared to general population. ¹³ Erasmus syndrome, characterized by the onset of SSc following silica exposure with or without silicosis, ¹⁴ has a prevalence from 1% to 15% among patients with SSc, depending on the study and definition criteria. ¹³ , ¹⁵ This syndrome is characterized by a higher prevalence in men, younger age at diagnosis, and increased mortality rate. ¹⁶ , ¹⁷ , ¹⁸ Although SSc is recognized as an occupational disease, no biologic marker has yet been identified. Therefore, AACD146 may represent the first marker leading to the identification of a subtype of SSc linked with occupational exposure. The pathogenesis of SSc may involve environmental triggers causing endothelial injury, leading to microvascular dysfunction, ¹ fibroblast activation, extracellular matrix overproduction, and ultimately fibrosis. ¹ Silica exposure is known to activate lung macrophages, releasing inflammatory mediators and promoting dysregulated autoreactive B cells, which produce autoantibodies. ¹⁹ Given that CD146 is expressed in endothelial cells, a hallmark of the initiation and progression of pulmonary fibrosis, ²⁰ we hypothesize that silica inhalation leads to endothelial activation, causing overexpression of CD146 and sCD146, thereby fostering AACD146 production and sustaining the inflammatory process promoting fibrogenesis. The development of an autoimmune response targeting CD146 may be a mechanism involved in SSc, in which AACD146 production reduce the regulatory role of CD146 in the Wnt and ROS pathways, leading to inflammation and fibrosis. However, further studies are required to elucidate their functional role and determine whether they play a causal or consequential role in the fibrotic process.

AACD146 enables the identification of a subgroup of patients with SSc who have occupational exposure to silica or asbestos, a correlation not previously observed with other autoantibodies. Currently, no biologic marker exists to monitor such exposure, which can lead to late diagnosis at advanced stages of the disease. Early identification of occupational exposure is crucial for timely diagnosis and for preventing further exposure once diagnosed. AACD146 are also found in workers with occupational exposure and could be valuable in occupational prevention strategies. Workers testing positive for these antibodies might benefit from enhanced monitoring to detect potential occupational diseases early or to promptly cease exposure. Our findings suggest that patients with positive AACD146 and occupational exposure history tend to have a higher prevalence of ILD. However, these results were not statistically significant, possibly due to the small sample size. These observations need further attention and evaluation in a larger cohort.

Our study has several strengths: (1) It is the first, to our knowledge, to identify a biomarker of occupational exposure in SSc. (2) Data were validated in a second cohort, highlighting the role of AACD146 in crystalline silica exposure. (3) Comparison between exposed patients and those with lung cancer (none of whom tested positive) reinforces AACD146 specificity in occupational diseases. (4) AACD146 could refine the diagnosis and phenotyping of patients with SSc and help prevent occupational disease in healthy workers. One major limitation of our study is its retrospective design, which inherently restricts our ability to establish temporal relationships between silica exposure and the emergence of autoimmunity. Although all included patients had documented long‐term silica exposure, the precise timing of autoantibody development in relation to exposure remains to be elucidated. Furthermore, our study included a limited number of control patients. To robustly confirm the observed relationships, future investigations with a larger control cohort will be necessary. Nonetheless, our findings provide a proof of concept for the involvement of AACD146 in silica‐associated SSc. To validate and extend these results, we have secured funding for a prospective study investigating the presence and dynamics of AACD146 in construction workers with and without silica exposure.

In conclusion, this study highlights AACD146 as a promising tool for the diagnosis and management of SSc in the context of occupational disease, offering new insight into its pathogenesis. This autoantibody may also be useful in monitoring workers exposed to silica or asbestos. Further investigations in larger populations and longitudinal follow‐up are essential to better characterize AACD146‐positive patients and workers.

AUTHOR CONTRIBUTIONS

All authors contributed to at least one of the following manuscript preparation roles: conceptualization AND/OR methodology, software, investigation, formal analysis, data curation, visualization, and validation AND drafting or reviewing/editing the final draft. As corresponding author, Dr Bermudez confirms that all authors have provided the final approval of the version to be published and takes responsibility for the affirmations regarding article submission (eg, not under consideration by another journal), the integrity of the data presented, and the statements regarding compliance with institutional review board/Declaration of Helsinki requirements.

Supporting information

Disclosure form.

Table S1: External validation cohort characteristics.