Rethinking the pathogenesis of cutaneous lupus

Abstract

Knowledge of the etiology of cutaneous lupus is rapidly evolving. Dissection of the pathologic events in lesional skin has led to knowledge of important cell populations and transcriptional changes contributing to disease. Recently, the study of non-lesional skin in systemic lupus patients has also identified key abnormalities that likely contribute to a propensity for inflammation. These include an elevated type I interferon (IFN) signature, overproduction of IFNs, and an absence of Langerhans cells. These changes promote aberrant inflammation in response to known triggers of disease, such as ultraviolet light. Further research will undoubtedly accelerate our understanding of this disfiguring disease.

Introduction

Cutaneous lupus erythematosus (CLE) is an inflammatory and potentially devastating disease. It can occur without associated systemic autoimmunity and as a manifestation of disease in up to 70% of patients with systemic lupus erythematosus (SLE) (Mikita et al., 2011). CLE and SLE occur with the same frequency(Jarukitsopa et al., 2015), yet until recently, CLE has received less attention than SLE regarding its etiology and treatment. Some may argue that the mechanisms driving SLE may be similar to CLE and thus CLE can be lumped into SLE treatment algorithms; however, in anecdotal clinical practice, CLE can be refractory to SLE-directed therapies or respond to therapies not typically used for SLE, such as dapsone. Making matters more complicated, CLE has various subtypes, each of which may have a unique pathophysiology. Thus, additional studies of CLE, including understanding subtype differences and the mechanisms driving inflammation are needed and warranted.

Characterization of lesional inflammation

Numerous studies have described the inflammatory infiltrates found in CLE lesions (reviewed in (Achtman and Werth, 2015)). These studies are important to understand which specific inflammatory mediators are contributing to tissue damage. Monocytes, macrophages, CD123⁺ plasmacytoid dendritic cells (pDCs) and T cells are all increased in CLE lesions. B cells may also be important as some CLE patients also respond to B cell depleting therapies (Quelhas da Costa et al., 2018). Some studies have identified specific cellular interactions that contribute to lesions, including interactions between pDCs and natural killer (NK) cells at the dermal-epidermal junction (Salvi et al., 2017). Disruption of pDCs has shown benefit for SLE and CLE in murine models (Sisirak et al., 2014), and recent phase I data support a role for targeting pDCs to treat CLE (Furie et al., 2019).

T cells are also important in CLE. Numerous chemokines that recruit T cells to the tissue are produced in CLE lesions (Wenzel et al., 2005). A role for cytotoxic T cells has been speculated in CLE, especially the more cellular discoid lupus erythematosus (DLE), for many years. Recently, a murine model in which the presence of autoreactive T cells coupled with tissue damage and lack of TLR9 signaling resulting in FasL-driven CLE in mice confirmed this (Mande et al., 2018). The success of ustekinumab, an antibody that blocks IL-12 and IL-23 signaling, in phase II studies for SLE-associated CLE (van Vollenhoven et al., 2019) has raised interest in IL-17 signaling in CLE. Some have identified an increase in IL-17+ CD4+ cells in discoid lesions by IHC, but these lesions also have an increase in total T cell numbers (Mendez-Flores et al., 2019b). How to compare this change with less cellular CLE subtypes is unclear. Other studies have not found an increase in Th17 cells or transcriptional signatures in CLE lesions but instead describe a strong Th1 signature centered around an upregulation of IFNγ, especially for discoid lupus erythematosus (DLE) lesions when compared to subacute CLE (SCLE) (Berthier et al., 2019, Jabbari et al., 2014). More quantitative methods will hopefully provide a more definitive answer for T cell subsets in lesional skin in the near future. This will hopefully also clarify whether the benefits of ustekinumab are secondary to inhibition of Th1 (via IL-12 blockade) vs. Th17 effects.

Another cell of interest in CLE is the neutrophil. Neutrophil extracellular traps (NETs) have been reported as increased in CLE lesions (Villanueva et al., 2011) and the presence of low-density granulocytes correlates with presence of skin lesions (Denny et al., 2010). Intriguingly, a new murine model of cutaneous and systemic lupus in which programmed death-1 homolog (PD-1H) is deleted identifies neutrophils as the initiator cell population in the skin prior to lesion development (Han et al., 2019). The role of neutrophils may vary depending on the subtype of CLE lesion with DLE=acute CLE> SCLE (Safi et al., 2019). As ultraviolet light induces early and profound infiltration of neutrophils in the skin, the role of neutrophils in photosensitive responses is also worthy of investigation.

CLE lesions also demonstrate microRNA changes. Circulating microRNAs have been examined and some potential correlations with disease activity have been noted. In particular, lower circulating levels of miR-150 positively correlate with cutaneous disease activity, suggesting an anti-inflammatory role for miR-150 (Mendez-Flores et al., 2019a). In lesional tissue, miR-31 is increased in DLE>SCLE, and when overexpressed, miR-31 contributes to cell death and production of inflammatory cytokines (Solé et al., 2019). The driving factors behind microRNA changes remain to be determined.

Type I interferons (IFNs) are increased in lesional skin and likely contribute to chemokine production and dendritic cell activation (discussed further below). Another area of investigation to consider is the relatively understudied type III interferons, also called IFNλ1–4. IFNλ1 (also known as IL-29) has been identified to be increased in lesional epidermis of CLE biopsies and is also found in the serum of patients with active CLE lesions(Zahn et al., 2011). Type III IFNs are upregulated in keratinocytes quite readily after exposure to endogenous nucleic acids, especially after UVB modification of the nucleotides (Scholtissek et al., 2017). In C57BL/6 mice, IFNλ was identified as highly expressed in the skin after TLR7 agonist stimulation and important for cutaneous inflammation. Intriguingly, absence of IFNλ was protective for not only cutaneous but also renal disease, despite having no impact on autoantibody development (Goel et al., 2020). In contrast, using a similar model in lupus-prone mice lacking the type I IFN receptor, loss of type I IFN signaling was not protective for development of nephritis (Wolf et al., 2018). Type I and type III IFN signaling have synergistic effects for driving skin inflammation (Goel et al., 2020) so the interplay of type I and type III IFNs in cutaneous and systemic lupus should be further investigated.

Examination of mechanisms driving sex bias in CLE has informed disease pathogenesis as well. SLE occurs in a female:male ratio of 9:1 and CLE occurs in a ratio of 4:1(Petersen et al., 2018). Intriguingly, a transcription factor, Vgll3, which is upstream of many genes associated with autoimmune disease, is regulated in a sex biased fashion in normal skin, but is activated in both male and female lesional CLE, suggesting a role in pathology(Liang et al., 2016). Transgenic overexpression of Vgll3 only in the epidermis results in a strong skin phenotype resembling DLE and development of autoantibodies and immune complex deposition in the skin and kidney(Billi et al., 2019). It remains to be seen whether targeting of Vgll3-regualted pathways can alter disease progression.

Predisposition to disease

Autoantibodies are a well-known indicator of CLE risk. Anti-Ro (SSA) positivity is associated with photosensitivity in about 54% of individuals and in these photosensitive patients, 47% have CLE (Popovic et al., 2007). More recently, studies of non-lesional skin in SLE patients have identified a predisposition for inflammation in SLE patient skin itself and have identified potential targets that could be considered for prevention of disease flare. This is important not only for the skin, but also for systemic disease as cutaneous inflammation can drive SLE flares (Clark et al., 2015), especially when the skin is exposed to excessive UVB.

The most consistent and interesting finding in non-lesional skin is an increase in interferon (IFN)-induced genes, the “interferon signature,” from the skin of patients with lupus nephritis (Der et al., 2017, Der et al., 2019) or SLE patients with a history of CLE (Sarkar et al., 2018). In non-lesional SLE keratinocytes, IFN kappa (IFN-κ) is the primary IFN chronically produced, and this leads to a propensity for TLR or UVB-stimulated IL-6 production(Stannard et al., 2017) and cell death after UVB exposure(Sarkar et al., 2018). Similar data from non-lesional skin from CLE-only patients is not yet available. Increased IFNs in SLE skin may also prime for T cell activation following UVB through activation of dendritic cells (Sarkar et al., 2018) and suppression of Treg induction in the draining lymph nodes (Wolf SJ et al., 2019). In addition, IFNs weaken the skin barrier, which contributes to dysbiosis and an increased likelihood of S. aureus colonization(Sirobhushanam et al., 2019). Contributions of the IFN signature from other cell populations in non-lesional skin, such as pDCs, has not yet been explored.

Depletion of Langerhans cells is another interesting feature of non-lesional SLE skin. Diminished Langerhans cell numbers can contribute to T cell activation in autoimmune skin diseases, such as pemphigus (Kitashima et al., 2018). Lack of Langerhans cells also changes the expression profile of keratinocytes, including promoting the upregulation of STAT3 (Su et al., 2020). Intriguingly, Langerhans cells are deficient in SLE non-lesional skin biopsies and this may also contribute to photosensitivity. Langerhans cells promote keratinocyte health through production of EGFR ligands. When this signaling axis is disrupted, the keratinocytes are more prone to die following UV exposure (Shipman et al., 2018). The reason for Langerhans cell depletion in SLE skin remains unknown, but would make a good target for prophylaxis of flares.

Genetic studies have identified polymorphisms in several SLE-related genes, such as TYK2, CTLA4, and IRF5 as conferring higher risk for skin manifestations (Jarvinen et al., 2010). Intriguingly, many of these risk variants for CLE involve the IFN signaling pathways. Further, variants in IFNK, which encodes IFN-κ, may also be associated with CLE development in European and African American patients(Harley et al., 2010). A role in IFN signaling also holds true for the most common polymorphism contributing to CLE risk, found in the ITGAM gene, which encodes for CD11b(Kim-Howard et al., 2010). Risk variants of ITGAM that produce a less functional CD11b molecule associate with IFN scores in SLE patients (Faridi et al., 2017). These data support the need for additional functional genomics analysis in order to understand genetic risk predisposition for skin disease.

Conclusions

The fast-paced research in systemic and cutaneous lupus is unraveling the role of skin changes in nonlesional skin and the specific inflammatory sequella to which these changes lead. It is important to understand the non-lesional and lesional spectrum of disease effect in order to develop better preventative and therapeutic options for our patients.

Table 1:

Comparison of abnormalities between lesional and non-lesional skin in SLE patients

Characteristic	Lesional Skin	Non-lesional Skin
IFN signature	Increased	Increased
IFNs produced	IFNα, IFNβ (low), IFNk, IFNγ (especially in DLE), IFNλ1	IFNk
Cellular Infiltrates	T cells, pDCs, NK cells, neutrophils (early)	Decreased Langerhans cells
microRNAs	Increased mIR-31	Unknown
Cell Death	Mediated by cytotoxic T cells, possibly NK cells	Increased after UVB exposure

Abbreviations used:

CLE

cutaneous lupus erythematosus

DLE

discoid lupus erythematosus

IFN

interferon

pDCs

plasmacytoid dendritic cells

SLE

systemic lupus erythematosus

UV

ultraviolet