Lymphocyte Activation Gene 3 Regulation of Profibrotic Cytokines and Type I Collagen Production in Patients With Systemic Sclerosis

Abstract

Objective

Diffuse cutaneous systemic sclerosis (dcSSc) is characterized by skin and internal organ fibrosis. The regulatory aspects of the immune system involved in maintaining tolerance are still poorly understood in dcSSc. The late checkpoint inhibitory lymphocyte activation gene 3 (LAG‐3) is implicated in T cell inhibition by interacting with its ligand, major histocompatibility complex class II. Its role in dcSSc has yet to be established.

Methods

Plasma soluble LAG‐3 (sLAG‐3) levels of 35 patients with dcSSc were compared with the levels of 20 healthy controls (HCs). Peripheral blood mononuclear cell (PBMC) samples from patients with dcSSc (n = 8) and HCs (n = 8) were analyzed by flow cytometry for the surface expression of LAG‐3 on CD4+ and CD8+ T cells after CD3/CD28 stimulation. The dcSSc PBMCs were stimulated (anti‐CD3/CD28), kept as monocultures, or cocultured with autologous dermal fibroblasts and treated with a LAG‐3 agonistic antibody or isotype control. The supernatants were analyzed for proinflammatory cytokines, extracellular matrix proteins, and bioactive type I interferon (IFN).

Results

Patients with dcSSc had a two‐fold decreased sLAG‐3 level compared with HCs (P < 0.0001). Increased surface expression of LAG‐3 was observed in activated CD4+ T cells and CD8+ T cells in patients with dcSSc compared to HC PBMCs. The LAG‐3 agonistic antibody decreased IFNγ, interleukin‐4, and tumor necrosis factor α levels in supernatants from PBMCs and dermal fibroblast cocultures. Furthermore, we observed decreased pro collagen, fibronectin, and IFNα/β bioactivity in the cocultures.

Conclusion

LAG‐3 is immunoregulatory in dcSSc, and a LAG‐3 agonistic antibody is a potential treatment modality.

INTRODUCTION

Lymphocyte activation gene 3 (LAG‐3) is a coinhibitory molecule expressed on activated immune cells. LAG‐3 plays a critical role in regulating the signaling pathways of T lymphocytes and antigen‐presenting cells (APC), particularly on antigen‐activated T cells and Treg cells. It has structural similarities to CD4 and binds to stable peptide–major histocompatibility complex (MHC) class II with a higher affinity than CD4. ¹ When bound to MHC class II, LAG‐3 leads to a bidirectional inhibition of APC and T cell function. ² In addition to CD4 T cells, LAG‐3 is expressed by CD8 T cells, Treg cells, B cells, plasmacytoid dendritic cells, and natural killer cells. ³ LAG‐3 can be cleaved from the cell surface, and the soluble LAG‐3 (sLAG‐3) is considered a surrogate marker for the cell‐expressed LAG‐3. High levels of sLAG‐3 are associated with poor disease outcomes in cancer. ⁴ sLAG‐3 has been proven biologically active, as adding Fc‐bound LAG‐3 has an anti‐apoptotic effect in vitro. ⁵ Recent evidence suggests a potential role for LAG‐3 in autoimmune rheumatic diseases such as rheumatoid arthritis (RA) and juvenile idiopathic arthritis (JIA). Here, the frequency of LAG‐3–expressing T cells is high within the inflamed joints, and plasma sLAG‐3 is increased. ⁶ In both RA and JIA synovial cell cultures, adding an agonistic LAG‐3 antibody decreased cytokine production. ⁶ , ⁷ Diffuse cutaneous systemic sclerosis (dcSSc) is associated with immune activation and increased proinflammatory cytokines, type I interferons (IFNs), and collagen synthesis. ⁸ T cell subsets in dcSSc resemble “functionally adapted” senescent T cells responding to chronic stimulation, leading to checkpoint receptor exhaustion and loss of tolerance. ⁹ Reports suggest increased LAG‐3 expression in immune cells in dcSSc, ¹⁰ but its potential as a therapeutic target for regulating the immune response has yet to be explored. This study aimed to investigate the role of LAG‐3 in dcSSc and understand its influence on the complex microenvironment during immune‐stromal crosstalk.

MATERIALS AND METHODS

Patient material

Plasma samples from patients with dcSSc (n = 35) were obtained from patients attending the Royal Free Centre for Rheumatology. All included patients were diagnosed with dcSSc without overlapping connective tissue diseases. The demographic and clinical characteristics are detailed in Supplementary Table 1. Plasma from age‐ and gender‐matched healthy controls (HCs; n = 20) was retrieved for comparison. Peripheral blood mononuclear cells (PBMCs) from patients with dcSSc (n = 8) and HCs (n = 8) were isolated using Ficoll–Paque density gradient centrifugation. Primary dermal fibroblast cultures were established from skin explants of patients with dcSSc (n = 8) according to a previously described method. ¹¹ Dermal fibroblasts at passages three to seven were used in this study. Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing GlutaMAX and 25 mM glucose (Gibco, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (FBS), 1 mM sodium pyruvate, 50 units/mL penicillin, and 50 μg/mL streptomycin (culture medium). The demographic and clinical characteristics of the patients whose cells were isolated are presented in Supplementary Table 1.

Table 1.

Clinical characteristics of included patients with dcSSc (total), examined for sLAG‐3 (ELISA) and surface expression of LAG‐3 (FLOW)^*

	Total	ELISA cohort	FLOW cohort
	(N = 43)	(n = 35)	(n = 8)
Age, mean (SD), y	52.20 (13.1)	51.5 (13.7)	52.9 (12.5)
Disease duration, mean (SD), mo	32.1 (37.5)	22.2 (11.6)	76.5 (70.8)
MRSS, mean (SD)	19.6 (10.6)	22.4 (9.0)	9.4 (10.4)
FVC in % of predicted, mean (SD)	94.1 (21.0)	95.7 (20.9)	88 (21.8)
CRP in mg/L, mean (SD)	7.1 (9.4)	6.3 (8.1)	10.6 (13.7)
ESR in mm/h, mean (SD)	15.1 (15.8)	12.8 (11.6)	28.2 (29.0)
Pulmonary fibrosis, n (%)	18 (42)	11 (31)	7 (88)
Antibodies, n (%)
Anti‐RNA polymerase	41 (51)	18 (51)	0 (0)
Anti‐topoisomerase I	21 (49)	13 (37)	8 (100)
Anti‐U3RNP	1 (2)	1 (3)	0 (0)
Treatment, n (%)
Cyclophosphamide	4 (9)	2 (6)	2 (25)
Mycophenolate mofetil	19 (44)	16 (46)	3 (38)
Methotrexate	7 (16)	5 (14)	2 (25)
Prednisolone	10 (23)	9 (26)	1 (13)
Abatacept	2 (5)	0 (0)	2 (25)

^*Values are expressed as mean (SD) for continuous variables or numbers (%) for categorical variables. CRP, C‐reactive protein; dcSSc, diffuse cutaneous systemic sclerosis; ELISA, enzyme‐linked immunosorbent assay; ESR, erythrocyte sedimentation rate; FVC, forced vital capacity; LAG‐3, lymphocyte activation gene 3; MRSS, modified Rodnan skin thickness score; sLAG‐3, soluble LAG‐3.

Protein and cytokine measurements

Quantification of sLAG‐3 (LAG‐3 Human Enzyme‐Linked Immunosorbent Assay [ELISA] kit # BMS2211; Invitrogen), type 1 procollagenα (Human Pro‐Collagen I alpha 1 DuoSet ELISA Cat.DY6220‐05; R&D Systems), and fibronectin (Human Fibronectin DuoSet ELISA Cat. DY1918‐05; R&D Systems) in the plasma and supernatants was performed according to the manufacturer's protocol. Samples were diluted 1:20 and supplemented with immunoglobulin to block heterophilic antibodies. Samples were analyzed in duplicate using the average optical density values.

Multiplexed protein expression in cell supernatants was measured using the Meso Scale Discovery (V‐PLEX Proinflammatory Panel 1 Human Kit), according to the manufacturer's guidelines (Meso Scale Diagnostics). The analysis of interleukin‐8 (IL‐8) levels was excluded as all the values were above the detection range. Values below or above the detection limit were assigned to the lowest or highest value of the detection range.

LAG‐3 agonist antibody and isotype control

Dr Frederic Triebel (Immutep) generously donated the agonist LAG‐3 antibody IMP761 and the corresponding human IgG4 antibody control. ¹²

Flow cytometry

PBMCs were stained for flow cytometry using the following antibodies: Brilliant Violet 605 anti‐human CD4 (clone OKT4, Cat. 317437), PE/Cyanine7 anti‐human CD3 (clone UCHT1, Cat. 300420), Brilliant Violet 785 anti‐human CD8 (clone SK1, Cat. 344740), Brilliant Violet 711 anti‐human CD223 (LAG‐3) (clone 11C3C65, Cat. 369319), APC anti‐human CD19 (clone 4G7, Cat. 392504), all from Biolegend. Dead cells were visualized using LIVE/DEAD fixable near‐IR (Molecular probe, Cat.L10119; Invitrogen). All samples were fixed and analyzed within 24 hours using a NovocyteQuanteon Flow Cytometer (Beckman Coulter), and data were processed using FlowJo version 10.9.0 software (Tree Star Inc). For the analysis of LAG‐3 expression on activated T cells, 0.5 million PBMCs were activated with Dynabeads human T activator CD3/CD28 antibodies (Gibco, Cat. 11131D; Thermo Fisher Scientific) at a bead‐to‐cell ratio of 1:2 for 24 hours at 37°C and 5% CO₂. Positive and negative cells were gated using fluorescence minus one control and blanks (Supplementary Figure 1).

PBMC monocultures and cocultures with dermal fibroblasts

To establish cocultures, dermal fibroblasts were seeded into 48‐well plates (20,000 cells/well). When the fibroblasts were attached and reached 70% confluency, anti‐CD3/CD28–activated autologous dcSSc PBMCs (100,000 cells/well) were added. Fibroblast + PBMCs cocultures were incubated with 0.5 μg/mL agonist LAG‐3 antibody (IMP761) or its isotype control for 48 hours (n = 8). Fibroblasts and PBMCs originated from the same patient. For monocultures, activated PBMCs (n = 8) were incubated for 48 hours with agonistic anti–LAG‐3 antibody or isotype control. All supernatants were harvested and stored at −80°C until further analysis. MTT assays were performed in parallel to ensure cell viability.

Cell‐based reporter assay

To measure bioactive type I IFN levels, HEK‐Blue IFNα/β cells (Cat. hub‐ifnabv2; Invitrogen) were cultured in 20% FBS containing DMEM. The selective antibiotics blasticidin (15 μg/mL) and zeocin (10 μg/mL) were added to maintain the selection of the cells. Supernatants from PBMCs and fibroblast cocultures were added to the HEK‐Blue cells and incubated for 24 hours. Secreted embryonic alkaline phosphatase in the supernatant was detected using QUANTI‐Blue (InvivoGen) and quantified by measuring optical density at 620 nm.

Statistical analysis

Statistical analyses and graphs were generated using GraphPad Prism version 7.0 for Mac (GraphPad Software). When data were normally distributed, samples were compared using t‐tests and represented as means with SDs. If not fitted to a normal distribution, data were analyzed using the Mann–Whitney test and expressed as median with interquartile range. The correlations were tested using simple linear regression. The level of statistical significance was a two‐sided P value < 0.05.

Ethical committee approval

The study received ethical approval from the NHS Research and Ethics Committee (number 6398) in the UK, and informed consent covering participation was obtained from all patients, which was approved by the London‐Hampstead (IRAS 270295) and London‐Fulham Research Ethics Committees (IRAS 279682). In Denmark, the study was approved by the Regional Ethics Committee (reference number 700779) and Regional Data Protection Agency (1‐16‐02‐542‐20). All participants provided informed consent to participate in the study and provided clinical data and samples for research.

RESULTS

sLAG‐3 plasma levels are decreased in dcSSc

Plasma levels of sLAG‐3 were significantly reduced in patients with dcSSc compared to HCs (368.4 vs 740.8 pg/mL; P < 0.0001; Figure 1A). Circulating levels of sLAG‐3 were inversely associated with disease duration (Slope −0.021; P = 0.018) and correlated with modified Rodnan skin thickness score (MRSS; Slope = 0.015; P = 0.0493). No correlation was found between sLAG‐3 levels and forced vital capacity (FVC), C‐reactive protein (CRP), and erythrocyte sedimentation rate (ESR) (data not shown; Figure 1B). Stratifying sLAG‐3 levels into a high and low group based on the median value did not add additional clinical value.

Figure 1.

Altered levels and T cell expression of LAG‐3 in Diffuse cutaneous systemic sclerosis (dcSSc). (A) Soluble plasma levels of LAG‐3 were measured in patients with dcSSc (n = 35) and compared to plasma levels in HCs (n = 20). Patients with dcSSc had significantly lower sLAG‐3 plasma levels compared to HCs. (B) In patients with dcSSc, sLAG‐3 plasma levels were inversely associated with disease duration and showed a positive correlation with mRSS. There was no correlation between soluble plasma levels of LAG‐3 and FVC, as well as CRP. (C) Surface expression of LAG‐3 on CD3+CD4+ T cells and (D) CD3+CD8+ T cells from patients with dcSSc (n = 8) and HCs (n = 8). PBMCs from both groups were stimulated with CD3/CD28 for 24 hours before flow cytometry analysis. Samples of CD3+CD4+ T cells and CD3+CD8+ T cells from patients with dcSSc had a significantly higher percentage of LAG‐3–expressing cells than the samples of HCs. Data are presented as median with IQR (*P < 0.05, **P < 0.01, ****P = 0.0001). (E) Flowcytometric representation of LAG‐3 expression on dcSSc and HC CD4+ T cells and CD8+ T cells. CRP, C‐reactive protein; dcSSc, diffuse cutaneous systemic sclerosis; FVC, forced vital capacity; HC, healthy control; IQR, interquartile range; LAG‐3, lymphocyte activation gene 3; mRSS, modified Rodnan skin thickness score; PBMC, peripheral blood mononuclear cell; sLAG‐3, soluble LAG‐3.

Increased surface expression of LAG‐3 on T helper cells and cytotoxic T cells in dcSSc

We next investigated the expression of LAG‐3 on the surface of unstimulated PBMCs from patients with dcSSc and HCs. However, we were unable to detect LAG‐3 expression by flow cytometry in either of these groups (data not shown).

Subsequently, we activated the PBMCs by anti‐CD3/CD28 and observed that the percentage of LAG‐3–expressing cells was significantly higher in dcSSc PBMCs than in HC PBMCs. This observation was made on CD4+ (25.7% vs 12.4%; P = 0.04) and CD8+ (14.9% vs 6.3%; P = 0.002) T cells (Figure 1C–E). These results reinforce the observation that dcSSc PBMCs have an increased ability to up‐regulate LAG‐3 in response to antigen. However, the increased cellular expression of LAG‐3 on CD4+ and CD8+ T cells did not correlate with clinical disease parameters, including disease duration, MRSS, FVC, ESR, or CRP (data not shown).

The LAG‐3 agonist is immunomodulatory in dcSSc. & engaging the LAG3 receptor decreases extracellular matrix protein production and IFN bioactivity

We investigated the impact of the LAG‐3 agonistic antibody (IMP761) on proinflammatory cytokine production in a complex microenvironment analogous to the interactions between immune and stromal cells in dcSSc. In monocultures of dcSSc PBMCs, adding a LAG‐3 agonistic antibody caused a significant decrease in the levels of released proinflammatory cytokines. The secretion of IFNɣ (813.6 pg/mL vs 1241 pg/mL; P = 0.03), IL‐1β (18.69 pg/mL vs 28.38 pg/mL; P = 0.02), IL‐4 (2.61 pg/mL vs 4.09 pg/mL; P = 0.01), and tumor necrosis factor α (TNFα; Q21 400.4 pg/mL vs 526.4 pg/mL; P = 0.02) was all significantly decreased upon treatment with the agonistic antibody versus isotype antibody control, respectively. Additionally, the levels of IL‐10 (3.83 pg/mL vs 2.92 pg/mL; P = 0.01) were increased in these cultures. No significant changes were observed in the secretion of IL‐12p70, IL‐13, IL‐2, or IL‐6 (Figure 2A). F2 We continued conducting autologous coculture experiments using PBMCs and fibroblasts. Comparable to the PBMC cultures, the treatment with LAG‐3 agonistic antibody significantly decreased the production of proinflammatory cytokines, including IFNɣ (2677 pg/mL vs 4104 pg/mL; P = 0.003), IL‐12p70 (10.73 pg/mL vs 12.13 pg/mL; P = 0.04), IL‐13 (386.1 pg/mL vs 448.7 pg/mL; P = 0.02), IL‐2 (457.3 pg/mL vs 555.5 pg/mL; P = 0.03), IL‐4 (5.33 pg/mL vs 6.38 pg/mL; P = 0.001), and TNFα (1425 pg/mL vs 1847 pg/mL; P = 0.001); compared to the control IgG, there was still a notable increase in IL‐10 production (8.33 pg/mL vs 6.10 pg/mL; P = 0.01). No significant changes were observed in the production of IL‐1β or IL‐6 (Figure 2B).

Figure 2.

Impact of agonistic LAG‐3 antibody on proinflammatory cytokine secretion by PBMCs. (A) dcSSc (n = 8) PBMCs were prestimulated with CD3/CD28 for 24 hours in monocultures or (B) cocultured with allogeneic dcSSc fibroblasts. The cultures were treated with either an LAG‐3 Q22 agonistic antibody (LAG‐3 Ag) or an isotype control (LAG‐3 Isotype) for 48 hours. In monocultures, the agonistic LAG‐3 Ag significantly reduced the levels of IFNγ, IL‐1β, IL‐4, and TNFα. In cocultures, IFNγ, IL‐12p70, IL‐13, IL‐2, IL‐4, and TNFα were decreased significantly. In both setups, IL‐10 levels were significantly increased by the addition of agonistic LAG‐3 isotype–treated samples. Data are presented as mean with SD Q23 (*P < 0.05, **P < 0.01). Ag, agonistic antibody; dcSSc, diffuse cutaneous systemic sclerosis; IFNγ, interferon γ; IL, interleukin; LAG‐3, lymphocyte Q24 activation gene 3; ns, not significant; NT, no treatment; PBMC, peripheral blood mononuclear cell; TNFα, tumor necrosis factor α (1425 pg/mL vs 1847 pg/mL; P = 0.001); compared to the control IgG, there was still a notable increase in IL‐10 production (8.33 pg/mL vs 6.10 pg/mL; P = 0.01). No significant changes were observed in the production of IL‐1β or IL‐6 (Figure 2B). Agonistic LAG‐3 antibodies suppress type I IFN production and reduce fibroblast matrix production. PBMCs from patients with dcSSc (n = 8) were stimulated with CD3/CD28 for 24 hours and cultured with or without allogeneic dcSSc fibroblasts for 48 hours. The cultures were then treated with either the LAG‐3 Ag or the isotype control (LAG‐3 Isotype). (C, D) The LAG‐3 Ag significantly reduced the production of bioactive IFNα/β in monocultures and cocultures, as measured in HEK‐Blue cells, compared to the LAG‐3 isotype control. (E) In cocultures, levels of type 1 procollagen and (F) fibronectin were significantly reduced by LAG‐3 Ag. Data are presented as mean with SD (*P < 0.05, **P < 0.01). Ag, agonistic antibody; dcSSc, diffuse cutaneous systemic sclerosis; IFN, interferon; LAG‐3, lymphocyte activation gene 3; ns, not significant; NT, no treatment; PBMC, peripheral blood mononuclear cell.

We continued conducting autologous coculture experiments using PBMCs and fibroblasts. Comparable to the PBMC cultures, the treatment with LAG‐3 agonistic antibody significantly decreased the production of proinflammatory cytokines, including IFNɣ (2677 pg/mL vs 4104 pg/mL; P = 0.003), IL‐12p70 (10.73 pg/mL vs 12.13 pg/mL; P = 0.04), IL‐13 (386.1 pg/mL vs 448.7 pg/mL; P = 0.02), IL‐2 (457.3 pg/mL vs 555.5 pg/mL; P = 0.03), IL‐4 (5.33 pg/mL vs 6.38 pg/mL; P = 0.001), and TNFα (1425 pg/mL vs 1847 pg/mL; P = 0.001); compared to the control IgG, there was still a notable increase in IL‐10 production (8.33 pg/mL vs 6.10 pg/mL; P = 0.01). No significant changes were observed in the production of IL‐1β or IL‐6 (Figure 2B).

Engaging the LAG3 receptor results in a decrease in extracellular matrix protein production and IFN bioactivity

Having observed the LAG‐3 agonistic antibody’s anti‐inflammatory impact on dcSSc PBMCs, we examined its effect on extracellular matrix proteins and type I INF production. We used the HEK‐Blue Q25 reporter cell assay to investigate the production of biologically active type I IFN levels in the supernatants from the activated PBMCs and the PBMC‐fibroblast cocultures. In monocultures of dcSSc vs 0.88 ng/mL; P = 0.01) both compared to the isotype control (Figure E and F). PBMCs, the LAG‐3 agonistic antibody significantly decreased the levels of bioactive IFN/β (75.07 IU vs F3 104.9 IU; P = 0.003) compared to its isotype control (Figure 2C). This effect was also mirrored in PBMCs cocultured with autologous fibroblasts, in which the LAG‐3 agonistic antibody also significantly reduced the levels of bioactive IFN/β (Figure 2D). Upon Q26 the production of extracellular matrix proteins in dcSSc PBMCs and fibroblast cocultures, we observed that the LAG‐3 agonistic antibody caused a significant reduction in the production of pro(aI)collagen (82.01ng/mL vs 94.16 ng/mL; P = 0.001) and fibronectin (0.68 ng/mL. vs 0.88 ng/mL; P = 0.01) both compared to the isotype control (Figure 2E and 2F).

DISCUSSION

Our findings strongly suggest a role for LAG‐3 in the pathogenesis of dcSSc. Previous work indicated that the distribution of inhibitory molecules, including LAG‐3, on T cells may indicate abnormal T cell regulation in autoimmune diseases.

In this study, we suggest that the regulatory LAG‐3 pathway plays a prominent role in dcSSc. We observed decreased sLAG‐3 levels in patients with dcSSc. This has also been reported in patients with psoriatic arthritis, ¹³ but contrasting with increased sLAG‐3 levels RA and JIA. ⁵ , ⁶ The reduced levels of sLAG‐3 in dcSSc may be attributed to decreased shedding by ADAM10 and ADAM17, whose activity is inhibited by tissue inhibitors of metalloproteinases (TIMPs). Notably, TIMP levels are reportedly elevated in dcSSc, ¹⁴ which may contribute to the observed reduction in plasma sLAG‐3.

Because sLAG‐3 has been linked to disease outcomes in patients with cancer, we explored the association of sLAG‐3 with clinical parameters. We observed an association with disease duration and MRSS in patients with dcSSc. We found no association between sLAG‐3 and CRP or ESR. This could be due to the relatively rapid turnover caused by the solubilization of LAG‐3 and the fact that the clinical progression of dcSSc often occurs over months to years.

Our findings reinforce the importance of LAG‐3 expression on T cells following stimulation. In line with Slevin et al, ¹⁵ who reported elevated IFNγ and IL‐17A production by CD4⁺LAG‐3⁺ T cells in ulcerative colitis, our results further clarify LAG‐3's immunoregulatory role. Activation with an LAG‐3 agonist significantly reduced proinflammatory cytokines (IFNγ, TNFα, IL‐4) linked to fibrosis in dcSSc. In cocultures of activated PBMCs and allogeneic dcSSc fibroblasts, this effect was amplified, with marked reductions in IFNγ, TNFα, IL‐13, IL‐12p70, IL‐2, and IL‐4. These findings suggest that LAG‐3 mediates bidirectional immunomodulation within a complex cellular environment.

Type I IFNs are reported to increase the expression of LAG‐3. Considering that the type I IFN pathway is a primary inducer of fibrosis in dcSSc, ¹⁴ , ¹⁵ it is reasonable to speculate that LAG‐3+ T cells could be essential regulators in the fibrotic process in patients with established dcSSc. Therefore, we continued to investigate the regulation of bioactive type I IFN and extracellular matrix proteins produced in our coculture setups. Here, we demonstrated that LAG‐3 agonism reduced the bioactive type I IFN and the production of extracellular matrix proteins fibronectin and procollagen I. This decreased production of extracellular matrix proteins could either be due to direct inhibition by the LAG‐3 pathway or secondary to a reduced cytokine production. Two studies have demonstrated that the Egr2+ CD4+ LAG3+ cells produce transforming growth factor β 3, causing decreased collagen synthesis and myofibroblast formation in healthy lung fibroblasts. ¹⁶ , ¹⁷ Our data extend these results in a setup mimicking multiple mononucleated cell types in direct contact with the fibrotic tissue in dcSSc, showing that LAG‐3 agonism directly affects the dcSSc immune cell and APC crosstalk, resulting in decreased production of collagen and fibronectin.

Our study has limitations. Based on these data, we cannot conclude that the reduction in type I IFN bioactivity and production of extracellular matrix proteins is solely due to a decrease in proinflammatory cytokines. Our study only focuses on changes influenced by the interaction between the agonistic antibody and the extracellular domain of LAG‐3. It does not explore signaling driven by the intracellular domain of LAG‐3. These aspects will be addressed in future comprehensive analyses.

Overall, this study shows that LAG‐3 significantly regulates proinflammatory cytokines and collagen production in dcSSc. This discovery opens the possibility of investigating LAG‐3 agonistic antibodies as a viable treatment option for dcSSc treatment. By reducing the activity of prominent proinflammatory cytokines and extracellular matrix proteins contributing to fibrosis progression, these antibodies could potentially lead to beneficial outcomes. Further research is required to determine the specific stage and subgroup of the disease that would benefit the most from targeted agonistic LAG‐3 therapeutic intervention.

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 Deleuran 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

Supplementary Figure 1: Gating strategy

Supplementary Table S1: Clinical characteristics of included dcSSc (Total), examined for sLAG‐3 (ELISA) and surface expression of LAG‐3 (FLOW)

Disclosure form.