Preclinical to clinical translation of cenerimod, a novel S1P1 receptor modulator, in systemic lupus erythematosus.

INTRODUCTION

Both B and T lymphocytes have been implicated in playing a major role in the pathogenesis of systemic lupus erythematosus (SLE), a complex and heterogeneous autoimmune disease of unknown aetiology, characterised by the production of high titres of pathogenic autoantibodies, infiltrating lymphocytes and tissue damage across multiple organ systems.[1]

In SLE, T lymphocytes orchestrate pathogenesis by directly driving organ damage and additionally by supporting B lymphocyte activation and proliferation, thus enabling the maturation of autoantibody-producing B lymphocytes, so-called antibody-secreting cells (ASC). ASC represent proliferative plasmablasts and plasma cells which are the predominant effector B lymphocytes in pathogenic conditions of autoimmunity.[2 3] Increased numbers of ASC correlate with various aspects of disease activity in patients with active SLE.[4-6]

Concomitant to the changes in immune lymphocyte populations observed in patients with SLE, the inflammatory environment in SLE is reminiscent of a chronic interferon (IFN)-driven immune response associated with viral infections. These include sustained expression of type I IFNs, often referred to as the ‘IFN signature’,[7-9] and an increase in IFN-responsive chemokines such as CXCL9 and CXCL10. Additionally, increases in modulators associated with SLE pathology such as interleukin 6 (IL-6) and tumor necrosis factor (TNF) exacerbate the inflammatory environment in SLE.[10]

One key aspect of the inextricable network between immune lymphocytes and the inflammatory environment is the interaction between lymphocytes and chemokine gradients, underlying the tightly controlled process of immune lymphocyte migration, which is fundamental in driving an adaptive immune response. The sphingosine-1-phosphate (S1P) chemotactic gradient is the predominant factor for lymphocyte egress out of peripheral lymphoid organs into the lymphatic and vascular circulation via S1P/S1P1 receptor engagement on B and T lymphocytes.[11] S1P1 receptor modulators are approved to treat multiple sclerosis and have shown clinical benefit in other diseases, providing evidence that this mode of action is of importance in autoimmune conditions.[12-14]

Therefore, cenerimod, a novel potent, selective and orally active S1P1 receptor modulator[15] was evaluated for efficacy in the MRL/lpr mouse model of SLE and clinically in patients with SLE.

MATERIALS AND METHODS

Details of materials and methods are available in online supplemental section.

RESULTS

Cenerimod reduces circulating and organ-infiltrating lymphocytes, decreases proteinuria and increases survival in the MRL/lpr mouse model of SLE

The MRL/lpr lupus mouse model reflects autoimmunity and organ pathology associated with SLE[16] and was therefore used to evaluate the preclinical efficacy of cenerimod. Seven-week-old MRL/lpr mice, an age at which B cell abnormalities are already detectable,[17] were treated with vehicle or cenerimod food admix (figure 1A). At treatment week 11, 6/20 mice in the vehicle group had died, achieving the predefined endpoint of at least 20% morbidity/mortality in one group. At this time point, all cenerimod-treated mice were still alive (figure 1B). Analysis of cenerimod-treated animals revealed a significant decrease of peripheral CD19+ B lymphocytes (−78.9%), and CD4+ (−98.9%) and CD8+ (−90.4%) T lymphocytes when compared with vehicle control (figure 1C).

Figure 1

Cenerimod reduces circulating and organ-infiltrating lymphocytes, decreases proteinuria and increases survival in the MRL/lpr mouse model of SLE.

Brain and kidney pathology are commonly associated with severe complications in patients with SLE,[10] aspects of which can be addressed in the MRL/lpr model. In the brain, cenerimod treatment led to a significant reduction of T cell infiltrates (figure 1D) and total IgG (figure 1E). Incidence of brain pathology was significantly decreased in cenerimod-treated (20%) compared with vehicle-treated mice (71%) (figure 1F), displaying reduced lymphocyte infiltrate in the choroid plexus and tela choroidea (figure 1G).

In the kidney, a significant reduction of T cell infiltrates (figure 1H) and a 25% decrease in kidney weight (online supplemental figure 1, p<0.0001) was observed in cenerimod-treated compared with vehicle-treated mice. Importantly, the decrease in urine albumin/creatinine ratio (figure 1I) was associated with reduced kidney pathology. The severity of histological damage (membranoproliferative glomerulonephritis associated with chronic tubulo-interstitial nephritis and pyelitis) was significantly lower in the cenerimod-treated group (median severity score of 1.0; IQR: 1.0–2.0) when compared with vehicle (median severity score of 2.5; IQR: 1.0–3.25) (figure 1J,K).

Cenerimod suppresses the systemic and local inflammatory environment in the MRL/lpr mouse model of SLE

Cenerimod treatment significantly reduced plasma anti-dsDNA and immunoglobulin (Ig) levels (figure 2A), and plasma IFN-associated, proinflammatory-associated and B cell-associated biomarkers (figure 2B). Importantly, the systemic effect of cenerimod was reflected in a statistically significant reduction of several of these disease relevant biomarkers in the brain (figure 2C).

Figure 2

Cenerimod suppresses the systemic and local inflammatory environment in the MRL/lpr mouse model of SLE.

In summary, cenerimod showed a consistent and pronounced suppression of the inflammatory environment in the MRL/lpr mouse model both systemically (figure 2D) and locally in brain tissue (figure 2E).

B and T lymphocytes in both healthy subjects and patients with SLE are responsive to cenerimod in vitro

In a non-interventional exploratory study of 10 healthy subjects and 10 patients with SLE (figure 3A), leucocyte populations were quantified (online supplemental figure 2A). Patients with SLE had decreased T lymphocytes, specifically central memory CD4+ and CD8+ T cells (online supplemental figure 2B), and plasmacytoid dendritic cells compared with healthy subjects while other populations including B lymphocytes were unchanged (online supplemental figure 2A).

Figure 3

Lymphocytes in both healthy subjects and patients with SLE are responsive to cenerimod in vitro, and patients with SLE present with a heterogeneous inflammatory phenotype.

Blood B (CD19+) and T lymphocytes (CD4+ and CD8+) were isolated and analysed for surface expression of S1P1 receptor using flow cytometry (online supplemental figure 2C). Mean surface expression of the S1P1 receptor was similar in healthy subjects and patients, and expression was threefold higher in B compared with T lymphocytes in all individuals (figure 3B). In vitro, cenerimod induced a consistent internalisation of the S1P1 receptor on B and T lymphocytes (figure 3C). Importantly, cenerimod had a similar potency (EC50, half-maximal effective concentration) on B (10.8 nM vs 15.3 nM), CD4+ T (16.5 nM vs 14 nM), and CD8+ T lymphocytes (13.2 nM vs 19.6 nM) from healthy subjects and patients (figure 3D). In summary, peripheral B and T lymphocytes in both healthy subjects and patients with SLE treated with standard of care were responsive to cenerimod-induced S1P1 receptor modulation.

Patients with SLE present with a heterogeneous inflammatory phenotype

Currently, no clinically approved biomarker exists to monitor SLE disease activity (monitoring), to predict disease progression (prognostic), or to capture disease heterogeneity. Therefore, a panel of known exploratory biomarkers were used to characterise patients with SLE. In the non-interventional exploratory study, these SLE-associated biomarkers were all increased in patients compared with healthy subjects (figure 3E). Specifically, type-1 IFN signature molecules such as pan-IFN-α,[18] and Galectin-9[19] as well as the IFN-responsive chemokines CXCL9[20] and CXCL10,[21] and the inflammation biomarkers TNF-α[22] and IL-6,[23 24] were statistically significantly elevated. In addition, B cell-activating factor (BAFF), a critical survival factor for mature B-lymphocytes and promoter of B-lymphocyte proliferation and differentiation, IgG class switching[25] and anti-dsDNA production[26] was increased. IL-10, a cytokine associated with B-lymphocyte hyperactivity and autoantibody production,[27 28] was also statistically significantly increased (figure 3E). A heatmap of these plasma biomarkers demonstrated a heterogeneity in the inflammatory severity between patients. Within individual patients, the overall biomarker phenotype was consistent, ranging from broadly elevated to levels observed in healthy subjects (figure 3F). The correlation between the individual biomarkers was high in patients with SLE (figure 3G). The heterogeneity between patients was confirmed in a larger patient cohort (online supplemental figure 2D). Additionally, there was a positive correlation between the majority of individual biomarkers across the cohort (online supplemental figure 2E).

In the phase 2 SLE study, cenerimod treatment dose-dependently reduces blood B and T lymphocytes

In the phase 2 SLE study, patients with SLE were treated for 12 weeks with multiple doses of cenerimod (figure 4A).[29] A dose-dependent reduction in blood CD19+ B, CD4+ and CD8+ T lymphocytes was observed following 12 weeks of cenerimod treatment compared with placebo control groups (figure 4B). The median reduction was most pronounced in CD4+ T (−94%) and B lymphocytes (−90%) compared with CD8+ T lymphocytes (−63%). A granular analysis of T lymphocytes revealed a dose-dependent reduction of all investigated subsets with the most pronounced effect on CD4+ T lymphocytes in general and central memory lymphocytes in particular (figure 4C).

Figure 4

In the phase 2 SLE study, cenerimod treatment dose-dependently reduces blood B and T lymphocytes.

In the phase 2 SLE study, cenerimod treatment normalises blood antibody-secreting cell numbers which correlate with IFN-associated biomarkers

Based on the high expression of S1P1 receptor on blood B lymphocytes, the increase in B cell-associated biomarkers in SLE, and a cenerimod-driven blood B lymphocyte reduction, an unbiased analysis of B lymphocyte populations in the phase 2 SLE study part A was performed (figure 5A–C). The self-organising map algorithm used, identified 16 clusters based on surface marker expression. Of the B cell-specific clusters, clusters 9 and 11 were statistically significantly reduced following cenerimod treatment (figure 5A).

Figure 5

In the phase 2 SLE study, cenerimod treatment normalises blood antibody-secreting cell numbers which correlate with IFN-associated biomarkers.

Cells in cluster 9 had higher expression of CD19, CD20, CD27, CD38 and CXCR3 when compared with the whole lymphocyte population; cells in cluster 11 had a similar expression pattern to cluster 9 except for lower expression of CD19 and higher expression of CD138 (figure 5B,C). Clusters 9 and 11 represented a subset of B lymphocytes resembling previously identified ASC which are known to be central to autoimmunity.[5] Enumeration of these cenerimod-responsive cell clusters confirmed a dose-dependent reduction in absolute numbers in the cenerimod-treated groups compared with placebo control (figure 5C).

In the non-interventional exploratory study, patients with SLE showed a significant elevation in ASC numbers compared with healthy subjects (figure 5D). In the phase 2 SLE study, cenerimod treatment dose-dependently normalised ASC numbers to levels observed in healthy subjects (figure 5D and online supplemental figure 3A). The role of ASC in the production of pathogenic antibodies is well described[5] and the increase in ASC was reflected in an increase of anti-dsDNA in patients from the non-interventional exploratory study (online supplemental figure 3B). Treatment with 4 mg cenerimod led to a significant decrease in systemic anti-dsDNA antibodies (online supplemental figure 3B) and total IgG (online supplemental figure 3C).

The relationship between ASC numbers and the inflammatory environment in patients with SLE was investigated. In the non-interventional exploratory study, ASC numbers correlated positively with the IFN-associated biomarkers CXCL9 (r=0.7, p=0.04) and CXCL10 (r=0.7, p=0.03); in the phase 2 SLE study prior to cenerimod treatment, ASC numbers correlated positively with pan-IFN-α (r=0.4, p=0.01), Galectin-9 (r=0.4, p=0.01) and CXCL10 (r=0.3, p=0.04) (figure 5E).

The phenotypically defined ASC from the non-interventional exploratory study were characterised by RNA sequencing analysis. ASC from patients with SLE had 601 differentially expressed genes (DEG) (false discovery rate <0.05 and fold change >1.5) compared with ASC from healthy subjects. Pathway analysis revealed IFN signalling to be the most significantly increased pathway (online supplemental figure 4A). Genes representing the previously reported 60-gene IFN signature identified in whole blood (WB) of patients with SLE[30] were strongly upregulated, confirming the dominance of the IFN pathway in ASC (online supplemental figure 4B).

In summary, cenerimod treatment in patients with SLE normalised ASC numbers which correlated with the IFN-phenotype.

A whole blood gene expression signature of patients with SLE reflects an elevated IFN-associated phenotype

A WB gene expression analysis in patients with SLE versus healthy subjects was performed in the non-interventional exploratory study to investigate its relationship to the elevation of IFN-associated biomarkers and ASC numbers.

Gene expression analysis of WB revealed 830 DEG with 642 genes upregulated and 178 genes downregulated in patients with SLE versus healthy subjects. The most prominent canonical pathways perturbed in patients with SLE represented genes associated with IFN signalling, estrogen-mediated S-phase entry, and cyclins and cell cycle regulation/proliferation (figure 6A). Genes from the 60-gene IFN gene signature[30] were strongly upregulated confirming the prominence of the IFN pathway, depicted as a heatmap (figure 6B), and captured in an overall score (figure 6C). In addition, increased plasma IFN-α protein levels correlated with the elevated IFN gene signature score (r=0.8, p<0.0001) (supplemental figure 4C). A correlation analysis demonstrated the WB-SLE gene signature to correlate stronger with the IFN-associated biomarkers than the IFN gene signature (figure 6D). Furthermore, the WB-SLE gene signature correlated with blood ASC numbers (r=0.6, p=0.006, online supplemental figure 4D).

Figure 6

Whole blood gene expression signature of patients with SLE reflects an elevated IFN-associated phenotype which is decreased after cenerimod treatment.

Cenerimod treatment dose-dependently decreases IFN-associated biomarkers

Analysis of the IFN-associated biomarkers in the phase 2 SLE study following 12 weeks of cenerimod treatment demonstrated a numerical, dose-dependent reduction in absolute plasma pan-IFN-α levels (figure 6E) and in per cent to baseline (online supplemental figure 4E). Analysis of the IFN-associated biomarkers in the phase 2 SLE study after 12 weeks of cenerimod treatment also demonstrated a consistent reduction in CXCL9, CXCL10 and Galectin-9 (figure 6F, online supplemental figure 4F).

DISCUSSION

Despite significant improvements in disease understanding and management of patients with SLE, remission and preservation of organ function is still considered challenging due to overreliance on corticosteroids and restricted options of effective drugs.[31 32] In this study, cenerimod, a novel, potent, selective and orally active S1P1 receptor modulator was characterised in the context of SLE.

The MRL/lpr mouse model presents with a systemic autoimmune phenotype, which possesses pathophysiological hallmarks of SLE, and is amenable to study histopathology in target organs typically associated with SLE.[16] Cenerimod treatment decreased blood B and T lymphocyte numbers, consistent with its mechanism of action in sequestering lymphocytes in secondary lymphoid organs and is consistent with previous findings describing S1P receptor modulators in models of SLE.[33-37] Importantly, the reduction of circulating lymphocytes translated to a reduction in organ-infiltrating lymphocytes in the kidney and brain. Consequently, organ-specific pathology and proteinuria were attenuated and culminated in improved survival. Cenerimod treatment also significantly reduced IFN-associated, inflammatory, and B cell-associated biomarkers. In summary, the preclinical data demonstrated that cenerimod treatment, by sequestering B and T lymphocytes, led to an efficient suppression of the inflammatory phenotype associated with SLE. Important for the translatability, the inflammatory phenotype in the blood and target organ of MRL/lpr mice, which could be reduced by cenerimod treatment, was also observed in the blood of patients with SLE.

Cenerimod was further investigated in patients with SLE following the promising results obtained in the MRL/lpr lupus mouse model. Cenerimod internalised the S1P1 receptor with equal potency on B and T lymphocytes isolated from healthy subjects and patients with SLE, thus rendering the lymphocytes non-responsive to S1P gradients. The results demonstrated that despite patients being treated with antimalarials and corticosteroids, cenerimod’s mechanism of action remained functional.

Patients with SLE present with a heterogeneous clinical and molecular phenotype.[38] To capture this heterogeneous molecular phenotype, we characterised biomarkers associated with SLE immunopathogenesis. Patients showed a statistically significant increase of biomarkers related to B lymphocyte activity, the IFN pathway, and inflammation. B lymphocytes are an important source of IFN-α,[39 40] which represents a relevant driver of SLE pathology, and has been employed for patient stratification in clinical studies.[41] Whereas individual patients at time of analysis had no inflammation, others showed pronounced inflammation, reflected in elevation of all the evaluated biomarkers. These data highlight the interconnectivity of these biomarkers and indicate that the IFN pathway is only one component of the general inflammatory phenotype in SLE.

In the phase 2 SLE study, blood B and T lymphocytes were dose-dependently reduced, successfully translating cenerimod’s mechanism of action observed preclinically to patients with SLE. Cenerimod was equi-efficacious on CD19+ B and CD4+ T lymphocytes highlighting the essential role of the S1P–S1P1 receptor pathway not only in T cell, but also in B cell trafficking in SLE. The importance of B lymphocytes in patients with SLE is well documented[42] and supported by the approval of an antibody targeting B cell activation.[43] Although profound lymphopenia is considered a risk factor for infection, cenerimod treatment in the 12-week study did not affect rates of infection.[29]

An unbiased multivariate analysis of the B lymphocyte subsets revealed that cenerimod treatment significantly reduced a cell population resembling ASC. This cell population has been shown to be elevated in patients with SLE[5 44 45] and correlates with increased anti-dsDNA levels.[46] A focused phenotypical analysis of ASC confirmed that cenerimod treatment dose-dependently normalised their numbers to a level comparable to healthy subjects and reduced total IgG and anti-dsDNA levels. These data suggest that by normalising ASC numbers, cenerimod treatment may lead to a reduction of pathogenic autoantibodies. The strong effect of cenerimod treatment on reducing blood CD4+ T lymphocytes could provide an additional mechanism to limit B cell help and consequently ASC numbers and anti-dsDNA levels.

The IFN pathway was previously described as an important contributor to the loss of tolerance[47] and in driving the increased frequency of ASC in patients.[48] Furthermore, B lymphocytes were reported to produce significant amounts of IFN-α in patients with SLE,[39 40] indicating that B lymphocytes (including ASC) might contribute to a systemic IFN-α increase. In line with these reported findings, both in the non-interventional exploratory and phase 2 SLE studies, ASC numbers correlated significantly with the IFN-associated biomarkers IFN-α, Galectin-9, CXCL9, and CXCL10. Next-generation sequencing of isolated ASC and WB confirmed a strong expression of IFN-associated genes in patients with SLE, as demonstrated by the striking upregulation of the published IFN gene signature.[30] Elevation of IFN-associated gene expression was reflected in higher plasma IFN-α. The significance of blocking this pathway has been demonstrated in a phase 3 clinical study in patients with 2the non-interventional exploratory study, the WB gene expression signature identified in patients with SLE (WB-SLE) was superior to the IFN gene signature in reflecting the inflammatory phenotype. Besides IFN signalling, the WB-SLE signature also captured additionally activated pathways including estrogen-mediated S-phase entry, which is implicated in immune cell activation and cytokine production,[50] and cyclins and cell cycle regulation which may be why the WB gene signature better reflected the phenotype. These data substantiate the observation that multiple pathways contribute to the overall activation of the adaptive and innate immune system in patients with SLE.[38 51] In the phase 2 SLE study, concomitantly to the reduction in ASC number, cenerimod treatment reduced plasma IFN-α levels and had a small effect on other IFN-associated biomarkers. While cenerimod treatment directly reduced circulating lymphocytes, probably as a consequence it also indirectly dampened IFN-α levels. B cell-associated and inflammatory biomarkers were not impacted in this short 12-week study. It remains to be seen in a long-term clinical study and larger patient cohort if the above-mentioned effects can be confirmed and whether cenerimod treatment can decrease additional inflammatory biomarkers and associated pathologies in patients with SLE.

In summary, cenerimod treatment demonstrated preclinical efficacy in the MRL/lpr mouse model. In patients with SLE, cenerimod treatment reduced blood B and T lymphocytes, normalised ASC numbers, decreased autoantibodies and reduced IFN-associated biomarkers. Cenerimod is being investigated in a multiple-dose clinical study up to 12 months in patients with SLE (CARE: NCT-03742037).

Sphingosine-1-phosphate (S1P)-dependent chemotaxis is a key mechanism of regulating B and T lymphocyte egress from secondary lymphoid organs into the lymphatic and vascular circulation contributing to autoimmune disease pathogenesis.

S1P receptor modulators are approved for the treatment of relapsing-remitting and secondary progressive multiple sclerosis.

Cenerimod, a novel potent and selective S1P1 receptor modulator, showed significant efficacy in the MRL/lpr mouse model of SLE; specifically, by reducing blood B and T lymphocytes, cenerimod treatment reduced proteinuria and tissue pathology, and controlled systemic and local inflammation resulting in increased survival.

In patients with SLE, cenerimod treatment reduced blood B and T lymphocytes, normalised blood antibody-secreting cell numbers, and reduced IFN-associated biomarkers.

This study supports the further development of cenerimod in SLE where this treatment strategy may provide benefit to patients with SLE.