Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Sep 29.
Published in final edited form as: J Invest Dermatol. 2025 Apr 30;145(12):2998–3010.e3. doi: 10.1016/j.jid.2025.03.043

Unrestrained MyD88 signaling in A20-deficient keratinocytes triggers T cell dependent Psoriatic Arthritis like disease

Ryan Tobias 1, Sugandh Kumar 1, Jared Liu 1, Nika Lenci 2, Ameneh Gharabi 1, Dorothea Stibor 2, Rommel Advincula 2, Philip Achacoso 2, Zhi-Ming Huang 1, Christopher Bowman 3, Roberto Ricardo-Gonzalez 1, Mary C Nakamura 2,4, Wilson Liao 1, Barbara A Malynn 2, Averil Ma 2,*, Bahram Razani 1,4,*
PMCID: PMC13618700  NIHMSID: NIHMS2205635  PMID: 40316204

Abstract

Polymorphisms in A20 (TNFAIP3), a negative regulator of ubiquitin-mediated immune signaling, are strongly associated with Psoriasis and PsA. The tissue-specific roles of A20 in preventing these diseases are poorly understood. As cutaneous psoriasis typically precedes PsA by several years, skin inflammation may represent a key driver of joint disease. We now find that keratinocyte-specific deletion of A20 in normally developed adult mice spontaneously triggers both psoriasiform skin and joint disease, demonstrating a crucial role for epidermal A20 in restricting PsA-like pathology. Mice with A20-deficient keratinocytes that lack T cells were protected from PsA-like disease, showing a key role for epidermally-triggered lymphocytes in driving joint inflammation. Early gene expression analysis following keratinocyte A20 deletion identified activation of MyD88 and antiviral signaling, reflecting spatial transcriptomic changes of human psoriatic epidermis. Keratinocyte-specific loss of A20 together with MyD88, but not germline disruption of interferon receptors, in vivo protected mice from skin and joint pathology. A20-deficient primary keratinocytes from both mice and Crispr-edited human cells spontaneously produced inflammatory cytokines and chemokines in vitro in a MyD88-dependent manner. A20-deficient murine keratinocytes also directly triggered IL17A-secretion from wildtype T cells. Together, our data demonstrate that keratinocyte A20 is critical for preventing T cell dependent PsA-like disease.

Keywords: Psoriasis, Psoriatic Arthritis, Innate Immunology, Signaling, NF-kB, Inflammation

INTRODUCTION

Nearly a third of patients with Psoriasis develop Psoriatic Arthritis (PsA) (Ritchlin et al. 2017). Psoriatic skin disease typically precedes arthritis by 5–7 years leading to the hypothesis that cutaneous inflammatory dysregulation can trigger or promote joint inflammation (Scher et al. 2019; Schett et al. 2022). However, the immunological mechanisms connecting skin and joint disease in psoriasis are not well understood. GWAS studies have identified polymorphisms in the TNFAIP3 locus, encoding A20 protein, as strongly conferring susceptibility to both cutaneous psoriasis and PsA (Liu et al. 2008; Nair et al. 2009; Stuart et al. 2015; Tsoi et al. 2012). Polymorphisms of A20 are associated with lower expression or function in human cells (Adrianto et al. 2011; Wang et al. 2013). Furthermore, independent of germline polymorphisms, levels of A20 transcription are epigenetically reduced in the skin of patients with psoriasis compared to healthy individuals (Aki et al. 2017; Devos et al. 2019; Sahlol et al. 2019). Thus, A20 is a key factor maintaining immune homeostasis and preventing psoriatic pathology in humans.

Broadly expressed in immune and stromal cell types, A20 possesses multiple domains that act on ubiquitin post-translational modifications (Razani et al. 2020a). By doing so, A20 is able to restrict signaling downstream of multiple innate immune signaling pathways (Razani et al. 2020a). Using germline knock-in mice, we recently showed that mice bearing homozygous mutation of A20’s zinc finger 7 motifs results in PsA-like pathology (Razani et al. 2020b). However, the cell types by which A20 prevents psoriasiform skin and joint disease is unclear. Emerging evidence has begun to highlight the role of keratinocytes in orchestrating inflammation (Simmons and Gallo 2024). We hypothesized that A20 may play a key role in maintaining immune homeostasis of keratinocytes and preventing psoriatic pathology.

To approach this question, we used an inducible model of keratinocyte A20 deletion in adulthood, thus avoiding ectodermal defects associated with constitutive keratinocyte A20 deletion (Lippens et al. 2011). We discover that deletion of A20 in adult keratinocytes results in spontaneous development of psoriasiform skin inflammation and distal digit PsA-like disease. Remarkably, although triggered by keratinocyte immune dysregulation, psoriatic skin and joint disease in these mice required the presence of T cells. We identified early activation of MyD88-dependent and antiviral signaling within the epidermis following keratinocyte A20 loss, finding that unrestrained MyD88 signaling in keratinocytes is critical for driving psoriasiform skin and joint inflammation following loss of A20.

RESULTS

Keratinocyte-specific deletion of A20 in adult mice leads to psoriasiform skin and PsA-like disease

To investigate the role of keratinocytes in the pathogenesis of Psoriasis and PsA, we genetically crossed A20F/F mice with those carrying the Keratin 5CreERT2 transgene (A20F/F Keratin 5CreERT2+). These mice allow for temporally-inducible and keratinocyte-specific deletion of A20 in adulthood with tamoxifen injection, avoiding ectodermal defects associated with constitutive deletion of A20 in keratinocytes(Lippens et al. 2011). We henceforth refer to these mice as A20KIKO mice (Keratinocyte Inducible Knockout) and tamoxifen-injected littermate control A20F/F Keratin 5CreERT2− mice as A20WT. Immunoblot for A20 in keratinocytes from epidermal sheets of A20KIKO mice one week following tamoxifen injection showed successful deletion of A20 protein in 8 week old adult mice (Fig 1a). These mice may bear particular clinical relevance to adult psoriasis patients that display epigenetically reduced cutaneous A20 expression(Aki et al. 2017; Devos et al. 2019; Sahlol et al. 2019).

FIGURE 1: A20 in keratinocytes restricts spontaneous psoriasiform skin and PsA-like disease.

FIGURE 1:

Open in a new tab

A) Keratinocytes from epidermal sheets of A20WT and A20KIKO mice 1 week after tamoxifen injection were immunoblotted for indicated proteins.

B) Representative gross photos of ears (top) and tail (bottom) of A20WT and A20KIKO mice 6 weeks following tamoxifen injection showing visible scaling of skin in A20KIKO mice.

C) Representative (H&E) images of ears (top) and tail (bottom) from A20WT and A20KIKO mice 6 weeks after tamoxifen injection. A20KIKO mice show acanthosis of epidermis (>), neutrophilic micro-abscesses (▼), and lymphohistiocytic dermal infiltrate (*). Scale bar = 200μm (top), = 100μm (bottom).

D) Representative gross photos of paws from A20WT and A20KIKO mice 6 weeks following tamoxifen injection showing clinical dactylitis and nail loss in A20KIKO mice.

E) Longitudinal quantification of digits displaying dactylitis and nail loss in weeks following tamoxifen injection in A20WT and A20KIKO mice. Week 0 indicates date of tamoxifen injection. Statistical significance was assessed by two-way ANOVA with Šídák's multiple comparisons test.

F) Representative histological (H&E) from A20WT mice at 6 weeks following tamoxifen injection presented for purposes of comparison to G. Scale bar = 200μm.

G) Representative histological (H&E) images of digits from A20WT and A20KIKO digits 6 weeks following tamoxifen injection. A20KIKO digits show enthesitis (>), dactilyitis (*), acanthosis and neutrophilic microabscesses (▼),distal osteolysis (□), and arthritis (△). The A20KIKO digits show loss of the nail with significant inflammation at the site of the former nail bed with marked bone erosion of the distal digit. Scale bar = 200μm.

In all assays, statistical significance is indicated as follows: *P < 0.05; **P < 0.01; ***P < 0.005; ****P < 0.0001.

We injected 8–10 week old adult A20KIKO and control A20WT mice with tamoxifen and longitudinally followed these mice for 6 weeks. Ears and tails of mice became grossly scaly with histology showing thickening of the epidermal layer (acanthosis), hyperkeratosis, and neutrophilic microabscesses, characteristic of Psoriasis (Fig 1b, c). Remarkably, in addition to skin changes, digits of A20KIKO mice showed clinical dactylitis along with nail dystrophy and nail loss, resembling key features of human PsA (Fig 1d) (Taylor et al. 2006; Wright and Moll 1971). These changes progressively affected all digits over the course of 6 weeks after tamoxifen injection (Fig 1e). Histological analysis of paws at 6 weeks following tamoxifen injection showed key features of PsA, including enthesitis, dactylitis, distal osteolysis, and arthritis (Fig 1f, g).

A20 possesses multiple domains which display distinct enzymatic and non-enzymatic functions on ubiquitin chains (Razani et al. 2020a). Our and others’ prior work using germline knock-in mice implicated the seventh zinc finger of A20, which non-enzymatically binds to linear ubiquitin chains, as the key domain which restricts spontaneous inflammatory pathology by preventing aberrantly extended innate immune signaling (Martens et al. 2020; Razani et al. 2020b). To determine if A20’s zinc finger 7 domain acts within keratinocytes to restrict psoriatic skin and joint pathology, we generated mice with A20ZF7 and A20F alleles (A20ZF7/F) that carry the Keratin 5CreERT2 transgene. Because wildtype A20 protein can compensate for zinc finger 7-mutated A20, A20ZF7/F mice do not develop PsA-like disease spontaneously (Supplementary Fig S1c). However, the presence of the Keratin 5CreERT2 transgene, together with tamoxifen injection, results in deletion of the A20F allele solely within keratinocytes, leaving only the A20ZF7 allele specifically expressed in these cells. Indeed, A20ZF7/F, Keratin 5CreERT2+ mice injected with tamoxifen developed PsA-like disease over 6 weeks, similar to A20KIKO mice, demonstrating that non-catalytic linear ubiquitin binding by A20’s zinc finger 7 domain within keratinocytes is critical for restricting PsA-like pathology (Supplementary Fig S1d). Importantly, neither A20ZF7/F, Keratin 5CreERT2− nor A20F/+, Keratin 5CreERT2+ mice develop similar pathology (Supplementary Fig S1b, c).

Together, these studies demonstrate that A20’s presence and ubiquitin-binding function within keratinocytes not only prevents psoriasiform dermatitis but also pathological inflammation of tendons and joints that resembles PsA.

Psoriatic molecular pathways drive inflammation in A20KIKO mice

To determine the immune pathways that trigger skin and joint pathology in A20KIKO mice, we investigated the roles of IL23, IL17A, and TNF, cytokines implicated in human psoriatic disease (Ritchlin et al. 2017). Whole ears were harvested 3 weeks following tamoxifen injection in A20WT and A20KIKO mice and evaluated for expression of Il23a, Il17a, and Tnf. Skin from A20KIKO mice showed elevation of Il23a, Il17a, and Tnf transcripts (Fig 2a). We evaluated numbers of IL17A-secreting lymphocytes in the skin of A20KIKO mice by flow cytometry three weeks following tamoxifen injection, finding an expansion of IL17A-expressing cells among both TCRβ+ αβ-T cells and TCRγ/δ+ γδ-T cell populations in A20KIKO mice (Fig 2b). In addition, we also identified an increased number of IL17A-expressing cells among lymphocytes (CD45+, Thy1.2+) that were negative for NK1.1, CD19, TCRβ, or TCRγ/δ markers, an immunophenotype consistent with that of IL17A-expressing ILC3s present in mouse and human skin (Fig 2b) (Kobayashi et al. 2020). IL17A and TNF promote psoriatic inflammation by inducing myeloid inflammation. We thus performed flow cytometric analysis of A20WT and A20KIKO ear skin 3 weeks following epidermal A20 deletion and found a profound infiltration of neutrophils, similar to human psoriatic skin (Fig 2c). Additionally, chemokines involved in recruitment of neutrophils (CXCL1) and IL17A-secreting T cells (CCL20) were elevated in the ear skin of A20KIKO mice 3 weeks following tamoxifen injection (Fig 2d). Reflecting signs of systemic inflammation associated with psoriatic disease, which include elevated circulating neutrophils and IL17A levels, A20KIKO mice displayed expansion of splenic neutrophils and macrophages as well as elevated serum IL17A (Fig 2e and f) (Liu et al. 2023; Rodriguez-Rosales et al. 2021). These results suggested that keratinocyte A20 restricts spontaneous activation of the IL23-IL17A-TNF cytokine axis.

FIGURE 2: Inflammation in A20KIKO mice molecularly resembles that seen in psoriatic disease.

FIGURE 2:

Open in a new tab

A) Quantitative PCR for indicated transcripts in whole ears at 3 weeks following tamoxifen injection in A20WT and A20KIKO mice. Statistical significance by Unpaired T-test.

B) Flow cytometry of whole ears at 3 weeks following tamoxifen injection in A20WT and A20KIKO mice showing total cells per ear of IL17A-expressing αβ-T cells (CD45+, Thy1.2+, TCRβ+), γδ-T cells (CD45+, Thy1.2+, TCRγ/δ+), as well as other IL17A-expressing cells with ILC-like phenotype (CD45+, Thy1.2+, CD19−, NK1.1−, TCRβ−, TCRγ/δ−). Statistical significance by Unpaired T-test.

C) Flow cytometry of ears 3 weeks after tamoxifen injection in A20WT and A20KIKO mice showing neutrophils (CD45+, Thy1.2−, CD11c−, CD11b+, Ly6G+) per ear. Statistical significance by Unpaired T-test.

D) Protein levels of indicated chemokines by ELISA within whole tail skin of A20WT and A20KIKO mice 3 weeks after tamoxifen injection. Total chemokine (pg) as ratio to total protein (μg) extracted. Statistical significance by Unpaired T-test.

E) Flow cytometry of spleens 3 weeks following tamoxifen injection in A20WT and A20KIKO mice showing neutrophils (CD11b+, Ly6G+, CD11c−) and macrophages (CD11b+, Ly6C+, Ly6G− ,CD11c−). Statistical significance by Unpaired T-test.

F) Serum IL17A by ELISA in A20WT and A20KIKO mice 3 weeks following tamoxifen injection. Statistical significance by Unpaired T-test.

G) Tamoxifen-injected A20KIKO mice were treated with either isotype control or anti-IL23 (p19) antibody weekly. Clinical dactylitis and nail loss were quantified longitudinally for 6 weeks after tamoxifen injection (indicated at Week 0). Statistical significance by two-way ANOVA with Šídák's multiple comparisons test.

H) Representative clinical (top) and histological (H&E) images (bottom) of paws six weeks following tamoxifen injection of A20KIKO mice treated with either isotype control or anti-IL23 (p19) antibody weekly. Scale bar = 200μm.

I) Representative histological (H&E) images of ear skin as in Panel H. Scale bar = 200μm.

J) A20KIKO, A20KIKO IL17A−/−, and A20KIKO TNF−/− mice were injected with tamoxifen. Clinical dactylitis and nail loss were quantified longitudinally for 6 weeks after tamoxifen injection (indicated at Week 0). Statistical significance by two-way ANOVA with Dunnett's multiple comparisons test comparing each compound mutant with A20KIKO mice.

K) Representative clinical (top) and histological (H&E) images (bottom) of paws 6 weeks following tamoxifen injection of A20KIKO with compound alleles of IL17A−/− (left) or TNF−/− (right). Scale bar = 200μm.

L) Representative histological (H&E) images of ear skin as in Panel K. Scale bar = 200μm.

In all assays, statistical significance is indicated as follows: *P < 0.05; **P < 0.01; ***P < 0.005; ****P < 0.0001.

To determine the pathogenic requirement for IL23, we pre-treated A20KIKO mice with anti-IL23 (p19) antibody or isotype control, followed two days later by tamoxifen injection to induce keratinocyte A20 deletion. We longitudinally followed these mice for 6 weeks while treating weekly with the above antibodies. A20KIKO mice treated with anti-IL23 were strongly protected from clinical signs of dactylitis and nail loss (Fig 2g). Histological examination of digits and skin after 6 weeks showed strong protection from both PsA-like inflammation in digits and psoriasiform inflammation in ear skin of anti-IL23-treated mice (Fig 2h, i).

As IL23 signaling is a key upstream activator of IL17A and TNF-mediated pathology, we investigated whether these cytokines play pathogenic roles in A20KIKO mice(Hawkes et al. 2018). We genetically interbred A20KIKO mice with IL17A−/− or TNF−/− mice and longitudinal followed tamoxifen-injected mice for clinical disease. A20KIKO mice with compound mutations in either IL17A or TNF were strongly protected from development of clinical dactylitis and nail loss (Fig 2j). At 6 weeks following tamoxifen injection, we analyzed mice for histological signs of digit and skin inflammation, finding that A20KIKO IL17A−/− and A20KIKO TNF−/− compound mutants were protected from PsA-like inflammation in digits and psoriasiform inflammation in ears (Fig 2k, l).

These data demonstrate that epidermal immune dysregulation in the context of keratinocyte A20 deficiency can trigger psoriatic skin inflammation by activating the IL23-IL17A-TNF axis of cytokines. The requirement for IL23 and IL17A in the development of PsA-like disease in A20KIKO mice implicated a key role for T cells in potentiating pathogenic signals from keratinocytes.

PsA-like disease triggered by keratinocyte A20 loss requires T cells

To determine the pathogenic role of T cells in A20KIKO mice, we interbred these mice with TCRδ−/− or TCRβ−/− mice, which lack αβ- or γδ-T cells, respectively. We longitudinally followed A20KIKO TCRδ−/− and A20KIKO TCRβ−/− mice for 6 weeks following tamoxifen injection. With similar kinetics to A20KIKO mice, A20KIKO TCRδ−/− and A20KIKO TCRβ−/− mice developed clinical dactylitis and nail loss (Fig 3a). Histological analysis of digits from TCRδ−/− and TCRβ−/− compound mutant mice showed PsA-like digit inflammation as well as psoriasiform dermatitis (Fig 3b, c).

FIGURE 3: PsA-like disease in A20KIKO mice requires T cells.

FIGURE 3:

Open in a new tab

A) A20KIKO, A20KIKO TCRδ−/−, and A20KIKO TCRβ−/− mice were injected with tamoxifen and longitudinally followed. Digits showing signs of clinical dactylitis and nail loss were quantified for 6 weeks following tamoxifen injection (indicated at Week 0). Statistical significance was assessed by two-way ANOVA with Dunnett's multiple comparisons test comparing each compound mutant with A20KIKO mice.

B) Representative clinical (top) and histological (H&E) images (bottom) of paws six weeks following tamoxifen injection of A20KIKO with compound alleles of TCRδ−/− (left) or TCRβ−/− (right). Scale bar = 200μm.

C) Representative histological (H&E) images of ear skin six weeks following tamoxifen injection of A20KIKO with compound alleles of TCRδ−/− (left) or TCRβ−/− (right). Scale bar = 200μm.

D) A20KIKO and A20KIKO TCRβ−/− TCRδ−/− mice were injected with tamoxifen and longitudinally followed. Digits showing signs of clinical dactylitis and nail loss were quantified for 6 weeks following tamoxifen injection (indicated at Week 0). Statistical significance was assessed by two-way ANOVA with Šídák's multiple comparisons test.

E) Representative clinical (top) and histological (H&E) images (bottom) of paws six weeks following tamoxifen injection of A20KIKO TCRβ−/− TCRδ−/− mice. Scale bar = 200μm.

F) Representative histological (H&E) images of ear skin six weeks following tamoxifen injection of A20KIKO with combined compound alleles of TCRβ−/− TCRδ−/−. Scale bar = 200μm.

G) Flow cytometry of distal digits from A20WT, A20KIKO, A20KIKO TCRβ−/− TCRδ−/− mice 4 weeks after tamoxifen injection quantifying neutrophils (CD45+, CD11b+, Ly6G+, CD11c−) and macrophages (CD45+, CD11b+, Ly6C+, Ly6G−, CD11c−). Significance assessed by one-way ANOVA with Tukey’s Multiple Comparison’s test.

H) Protein levels of indicated cytokines as assessed by ELISA within digits of A20WT, A20KIKO, A20KIKO TCRβ−/− TCRδ−/− mice 4 weeks following tamoxifen injection. Significance assessed by one-way ANOVA with Tukey’s Multiple Comparison’s test.

In all assays, statistical significance is indicated as follows: *P < 0.05; **P < 0.01; ***P < 0.005; ****P < 0.0001.

Given potential functional redundancy between αβ- or γδ-T cells in the disease processes observed in A20KIKO mice, we generated A20KIKO mice with combined compound deficiency of TCRβ−/− TCRδ−/−, which lack all T cells, but retain ILC3s. Tamoxifen-injected A20KIKO TCRβ−/− TCRδ−/− mice were longitudinally followed for 6 weeks and were strongly protected from development of clinical dactylitis and nail loss (Fig 3d). Histological analysis of digits from A20KIKO TCRβ−/− TCRδ−/− mice at 6 weeks following tamoxifen injection showed protection from PsA-like inflammation (Fig 3e). Additionally, flow cytometry for neutrophils and macrophages in distal digits 4 weeks after tamoxifen injection showed significantly reduced numbers of these inflammatory myeloid cells in digits of A20KIKO TCRβ−/− TCRδ−/− mice compared to A20KIKO mice (Fig 3g). Congruently, levels of inflammatory cytokines IL6 and IL1β were reduced in A20KIKO TCRβ−/− TCRδ−/− mice compared to A20KIKO mice (Fig 3h). These data demonstrate a key requirement for T cells in mediating distal digit PsA-like disease. The preserved digit pathology in subset-specific T cell deficient mice (A20KIKO TCRδ−/− and A20KIKO TCRβ−/−) suggests that either αβ- or γδ-T cells are sufficient for mediating pathogenic epidermal signals to joints. Surprisingly, although protected from PsA-like disease, A20KIKO TCRβ−/− TCRδ−/− mice continued developed psoriasiform skin inflammation similar to A20KIKO mice (Fig 3f).

Loss of keratinocyte A20 triggers early activation of epidermal antiviral and MyD88 pathways, reflecting human psoriatic epidermis

The temporally inducible and keratinocyte-specific origin of pathology in A20KIKO mice allows an opportunity to identify the earliest endogenous immune pathways responsible for triggering psoriatic skin and joint disease. To identify these early pathways, we performed bulk RNASeq on epidermal sheets from A20WT and A20KIKO mice 1 week following tamoxifen injection, a time point at which A20 is undetectable within the epidermis but there is no histologically detectable inflammation within the skin or digits of A20KIKO mice (Fig 1a).

RNASeq showed spontaneous elevation of 456 genes within the epidermis of A20KIKO mice (Fig 4a). Pathways analysis identified Defense Response to Virus and Response to Interleukin-1 as over-represented among these early elevated genes (Figure 4b and highlighted in blue and red respectively in Figure 4a). These results raised the possibility that A20 within keratinocytes restricts epidermal antiviral signaling and signaling from MyD88, the central signaling mediator downstream of IL1-like Receptors and Toll-like Receptors (TLRs).

FIGURE 4: Loss of keratinocyte A20 triggers early activation of epidermal antiviral and MyD88 pathways, reflecting human psoriatic epidermis.

FIGURE 4:

Open in a new tab

A) Volcano plot of significantly differentially expressed genes (log2Fold > 1 and adjusted p value < 0.05) from RNASeq of epidermal sheets from A20KIKO vs A20WT mice 1 week after tamoxifen injection. Antiviral pathway and MyD88 response pathways genes highlighted in blue and red, respectively

B) Pathways Analysis of elevated DEGs in Panel A highlighting top enriched biological pathways including Defense Response to Virus and Response to IL1 pathways.

C) UMAP of spatial transcriptomic data from skin biopsies of healthy volunteers as well as skin biopsies of lesional and non-lesional skin of patients with psoriasis derived from Castillo et al dataset(Castillo et al. 2023). A total of 24 samples were analyzed (‘Healthy = 8’, ‘Lesional = 8’, and ‘Non-Lesional = 8’).

D) Spatial spots assigned to epidermal keratinocytes (highlighted in C) were evaluated for antiviral and MyD88-pathway genes. Dot plot shows elevation of antiviral and MyD88-pathway gene expression in psoriatic lesional skin compared to non-lesional skin from the same patient or skin from healthy controls. Gene expression values were scaled to a range from −1.0 to 1.0, with a value of 0.0 representing mean expression across all samples.

To determine if similar immune pathways are elevated in human psoriatic epidermis, we analyzed a recent 10X Visium spatial transcriptomics atlas of skin biopsies from both lesional and non-lesional skin of psoriatic patients as well as healthy controls (Castillo et al. 2023). We specifically focused our analysis to UMAP clusters of 50 μM resolution spots with marker genes corresponding to epidermal keratinocytes and other cell types (Fig 4c). We then evaluated for the expression of antiviral and MyD88 pathway genes, as identified in recently published transcriptomic data sets, within spots assigned to human epidermal regions (Swindell et al. 2018; Tsoi et al. 2019). We focused our evaluation on human homologs of genes that were differentially expressed in A20KIKO mouse epidermis which included the antiviral genes Ifit1, Ifit2, Ifit3, Isg15, Mx1, Oasl, Rsad2, Usp18 and the MyD88-pathway genes Ccl2, Ccl5, Ccl19, Ccl20, Cxcl8, Cxcl10, Tnfaip3.

Compared to skin from healthy individuals or non-lesional skin from psoriatic patients, psoriatic lesional skin displayed greater numbers of epidermal spots expressing transcripts from both antiviral and MyD88 pathway genes (Figure 4d). Furthermore, within epidermal spots that displayed detectable transcript, the average expression of both antiviral and MyD88 pathway genes were increased in psoriasis lesional skin compared to non-lesional skin or healthy control skin (Figure 4d). Interestingly, TNFAIP3 (A20), which is an NF-κB response gene, proved to be an exception to the pattern of elevated expression of MyD88 pathway genes in psoriatic epidermis. Epidermal regions of skin from individuals with psoriasis, whether in lesional or non-lesional areas, displayed lower expression of TNFAIP3 (A20) compared to epidermal regions from healthy individuals, analogous to recent reports using laser capture microdissection of human epidermis (Devos et al. 2019). Taken together, our findings show that loss of A20 within keratinocytes leads to early and spontaneous induction of antiviral and MyD88 pathway genes, reflecting transcriptomic changes observed in human psoriatic epidermis.

Antiviral signaling is not required for skin and joint disease following keratinocyte A20 loss

We next dissected the contributions of antiviral and MyD88-dependent immune pathways to the pathogenesis of psoriatic disease in A20KIKO mice. Antiviral gene expression is induced by interferon receptors which include the broadly-expressed Type I Interferon receptor, IFNAR, as well as the epithelial-restricted Type III interferon receptor, IFNLR(Lazear et al. 2019). We thus generated A20KIKO compound mutant mice with those carrying IFNAR−/− and IFNLR−/− alleles.

To determine if spontaneous elevation of antiviral signature transcripts seen in A20KIKO epidermis was lost with compound deletion of IFNAR or IFNLR, we injected A20KIKO IFNAR−/− and A20KIKO IFNLR−/− mice with tamoxifen and harvested epidermal sheets one week later for quantitative PCR of antiviral genes Ifit1, Oasl2, and Rsad2. A20KIKO compound mutants with IFNAR−/− or IFNLR−/− were significantly protected from spontaneous elevation of these antiviral transcripts, demonstrating that epidermal antiviral gene expression triggered by loss of keratinocyte A20 is non-redundantly mediated by Type I and Type III Interferons (Fig 5a). We then longitudinally followed A20KIKO IFNAR−/−, A20KIKO IFNLR−/−, and control mice for 6 weeks following tamoxifen-induced keratinocyte A20 deletion. Similar to A20KIKO mice A20KIKO IFNAR−/− and A20KIKO IFNLR−/− develop progressive dactylitis and nail loss (Fig 5b, c).

FIGURE 5: Antiviral interferon receptor signaling does not contribute to psoriatic skin and PsA-like disease in A20KIKO mice.

FIGURE 5:

Open in a new tab

A) Quantitative PCR of antiviral pathway genes (Ifit1, Oasl2, Rsad2) in epidermal sheets one week following tamoxifen injection in indicated genotypes of mice. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple comparison’s test.

B) A20KIKO, A20KIKO IFNAR−/−, and A20KIKO IFNLR−/− mice were injected with tamoxifen and longitudinally followed. Digits showing signs of clinical dactylitis and nail loss were quantified for 6 weeks following tamoxifen injection (indicated at Week 0). Statistical significance was assessed by two-way ANOVA with Dunnett's multiple comparisons test comparing each compound mutant with A20KIKO mice.

C) Representative clinical images of paws six weeks following tamoxifen injection of A20KIKO IFNAR−/− and A20KIKO IFNLR−/− mice.

D) Quantitative PCR of antiviral pathway transcripts (Ifit1, Oasl2, Rsad2) in epidermal sheets one week following tamoxifen injection in indicated genotypes of mice. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple comparison’s test.

E) Quantitative PCR of MyD88 pathway transcripts (Cxcl1, Ccl20, Tnf) in epidermal sheets one week following tamoxifen injection in indicated genotypes of mice. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple comparison’s test.

F) A20KIKO and A20KIKO IFNAR−/− IFNLR−/− mice were injected with tamoxifen and longitudinally followed. Digits showing signs of clinical dactylitis and nail loss were quantified for 6 weeks following tamoxifen injection (indicated at Week 0). Statistical significance was assessed by two-way ANOVA with Šídák's multiple comparisons test.

G) Representative clinical images of paws six weeks following tamoxifen injection of A20KIKO IFNAR−/− IFNLR−/− mice.

H) Representative histological (H&E) image of paws six weeks following tamoxifen injection of A20KIKO IFNAR−/− IFNLR−/− mice. Scale bar = 200μm.

I) Representative histological (H&E) images of ear skin six weeks following tamoxifen injection of A20KIKO IFNAR−/− IFNLR−/− mice. Scale bar = 200μm.

In all assays, statistical significance is indicated as follows: *P < 0.05; **P < 0.01; ***P < 0.005; ****P < 0.0001.

Given the non-redundant role of IFNAR and IFNLR in mediating antiviral gene expression in A20KIKO epidermis, we generated A20KIKO compound mutants with a combination of IFNAR−/− IFNLR−/− alleles, eliminating antiviral signaling from both Type I and III interferon receptors. Quantitative PCR on epidermal sheets of A20KIKO IFNAR−/− IFNLR−/− mice one week following tamoxifen injection showed potent protection from spontaneous elevation of antiviral genes Ifit1, Oasl2, and Rsad2 (Fig 5d). We also evaluated several select MyD88-pathways genes (Ccl20, Cxcl1, and Tnf) and found spontaneous elevations of these genes similar to that in A20KIKO mice (Fig 5e). Thus, MyD88 pathway activation does not appear to be dependent on antiviral response pathways.

Similar to A20KIKO mice, A20KIKO IFNAR−/− IFNLR−/− mice developed clinical dactylitis and nail loss affecting all digits by 6 weeks following tamoxifen injection (Fig 5f, g). Histological analysis at 6 weeks showed PsA-like inflammation in digits and psoriasiform inflammation in ear skin (Fig 5h, i). These data show that keratinocyte A20 restricts epidermal antiviral gene activation via Type I and Type III interferons but that eliminating antiviral signaling does not protect A20KIKO mice from psoriatic skin or joint pathology in this keratinocyte-driven model of disease.

A20 restricts pathogenic and cell autonomous MyD88 signaling from keratinocytes to prevent psoriatic skin and joint disease

To determine the role of keratinocyte MyD88 signaling in A20KIKO pathology, we interbred A20KIKO mice to MyD88F/F mice. These mice (hence termed A20KIKO MyD88KIKO) allow simultaneous deletion of A20 and MyD88 within keratinocytes following tamoxifen injection. Quantitative PCR of epidermal sheets one week following tamoxifen injection showed that A20KIKO MyD88KIKO mice were protected from spontaneous elevation of chemokines and cytokines such as Ccl20, Cxcl1, and Tnf (Fig 6a). Intriguingly, transactivation of antiviral genes such as Ifit1, Oasl2, and Rsad2 was also prevented, suggesting a role for upstream MyD88 in activating antiviral genes (Fig 6b). Remarkably, tamoxifen-injected A20KIKO MyD88KIKO mice were strongly protected from development of clinical dactylitis and nail loss over 6 weeks (Fig 6c, d). Histological examination showed strong protection from PsA-like pathology within digits as well as psoriasiform inflammation within ears (Fig 6e, f). Thus, A20 prevents psoriatic pathology, both within skin and joints, by restricting pathogenic MyD88 signaling within keratinocytes.

FIGURE 6: MyD88 within keratinocytes is required for skin and PsA-like disease in A20KIKO mice.

FIGURE 6:

Open in a new tab

A) Quantitative PCR of MyD88 pathway transcripts (Cxcl1, Ccl20, Tnf) in epidermal sheets one week following tamoxifen injection in indicated genotypes of mice. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple comparison’s test.

B) Quantitative PCR of antiviral pathway transcripts (Ifit1, Oasl2, Rsad2) in epidermal sheets one week following tamoxifen injection in indicated genotypes of mice. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple comparison’s test.

C) A20KIKO mice and A20KIKO MyD88KIKO were injected with tamoxifen and longitudinally followed. Digits showing signs of clinical dactylitis and nail loss were quantified for 6 weeks following tamoxifen injection (indicated at Week 0). Statistical significance was assessed by two-way ANOVA with Šídák's multiple comparisons test.

D) Representative clinical images of paws six weeks following tamoxifen injection of A20KIKO MyD88KIKO mice. Representative histological (H&E) images of ear skin six weeks following tamoxifen injection of A20KIKO MyD88KIKO mice (right).

E) Representative histological (H&E) image of digit six weeks following tamoxifen injection of A20KIKO MyD88KIKO mice. Scale bar = 200μm.

F) Representative histological (H&E) image of ear skin six weeks following tamoxifen injection of A20KIKO MyD88KIKO mice. Scale bar = 200μm.

G) ELISA for indicated proteins within the supernatants of in vitro cultured keratinocytes derived from A20WT, A20KIKO, and A20KIKO MyD88KIKO mice. Primary keratinocytes were treated with 4-hydroxytamoxifen following isolation. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple comparison’s test.

H) ELISA for IL17A in supernatants of in vitro cultured wildtype splenic T cells incubated for 72hrs with conditioned media from in vitro cultures either containing no keratinocytes (−) or cultures of primary keratinocytes derived from A20WT, A20KIKO, and A20KIKO MyD88KIKO treated with 4-hydroxytamoxifen following isolation. Statistical significance by one-way ANOVA with Tukey’s multiple comparison’s test.

I) Immunoblot for indicated proteins derived from primary human keratinocytes following treatment with CRISPR-Cas9 guides targeting scramble sequence (Ctrl), A20, or A20 in combination with MyD88.

J) Quantitative PCR for indicated transcripts from primary human keratinocytes treated with CRISPR guides targeting scramble sequence (Ctrl), A20 alone, or A20 in combination with MyD88. Concomitant western blot for A20 and MyD88 is seen in Panel G. Statistical significance by one-way ANOVA with Tukey’s multiple comparison’s test.

In all assays, statistical significance is indicated as follows: *P < 0.05; **P < 0.01; ***P < 0.005; ****P < 0.0001.

To determine a cell autonomous role for A20 in restricting pathogenic keratinocyte MyD88 signaling, we cultured keratinocytes from A20WT, A20KIKO, and A20KIKO MyD88KIKO mice. We acutely deleted A20 alone, or in combination with MyD88, in vitro by treating keratinocytes with 4-hydroxytamoxifen. Multiplex ELISA on culture supernatants 48hrs later showed that A20-deficient keratinocytes spontaneously produced key psoriatic chemokines and cytokines, including CCL2, CCL20, CXCL1, CXCL2, GM-CSF, and TNF (Figure 6g). Secretion of these factors by A20-deficient keratinocytes was ablated in the absence of MyD88 (Figure 6g).

We asked if culture supernatants from A20-deficient keratinocytes could directly impact IL17A secretion by T cells. Culture supernatants from the keratinocytes above were incubated with total splenic T cells from wildtype mice for 72hrs. Intriguingly, wildtype splenic T cells treated with conditioned media from A20-deficient keratinocytes spontaneously secreted IL17A whereas those treated with supernatants from wildtype keratinocytes or keratinocytes with combined deletion of A20 and MyD88 did not secrete IL17A (Figure 6h).

We optimized efficient CRISPR-mediated gene deletion of A20 alone or in combination with MyD88 in primary human keratinocytes in vitro (Figure 6i). Remarkably, CCL2, CCL20, CXCL8, CXCL2, CSF2, and TNF transcripts were all significantly elevated in A20-deleted primary human keratinocytes, but not in those with combined CRISPR-mediated deletion of A20 and MyD88 (Figure 6j). Thus, A20 cell autonomously restricts spontaneous MyD88-dependent production of multiple key inflammatory chemokines and cytokines in both mouse and human keratinocytes.

DISCUSSION

A20KIKO mice reveal that A20 within keratinocytes is critical for preventing not only spontaneous psoriasiform skin inflammation but also fully penetrant PsA-like joint disease. We find that A20 deficient keratinocytes require T cells to trigger psoriatic joint pathology and we identify MyD88 signaling within keratinocytes as the key pathway that A20 restricts to prevent psoriatic disease.

The immunological connection between psoriatic skin and joint disease remains poorly understood in humans. Psoriatic joint disease typically follows the presentation of cutaneous psoriasis by several years (Scher et al. 2019). A key question that emerges from these clinical observations is whether psoriatic cutaneous inflammation triggers or potentiates arthritis. The role of skin in driving arthritis has been explored by several mouse models that involve keratinocyte-specific transgenic overexpression of cytokines (IL17A, IL23) (Croxford et al. 2014; Chen et al. 2020), growth factors (Areg) (Cook et al. 2004), secreted enzymes (Klk6) (Billi et al. 2020), or constitutively activated signaling mediators (Stat3C:F759, Rac1V12) (Winge et al. 2016; Yamamoto et al. 2015). These studies elegantly demonstrate that aberrant keratinocyte-specific production or activity of positive regulators of human psoriasis can drive not only cutaneous inflammation but also disrupt immune homeostasis within joints. Few mouse models of psoriatic joint disease have been described that arise from disrupting negative regulators of inflammation in keratinocytes. Such models illustrate that, in the absence of such negative regulation, endogenous processes within keratinocytes can drive both cutaneous inflammation and arthritis. In addition to the A20KIKO model described here, the spontaneous arthritogenic potential of keratinocytes was recently exemplified by mice with keratinocyte-specific defects in Zfp36, an RNA-binding protein that destabilizes endogenously transcribed inflammatory gene transcripts(Andrianne et al. 2017). Thus, multiple layers of mechanistically distinct negative regulation may be required to restrain endogenous signaling within keratinocytes from spontaneously triggering PsA-like disease. Genetic or epigenetic defects in these layered mechanisms, such as those linking A20 polymorphisms with both psoriasis and psoriatic arthritis, may additively lower the threshold for skin-triggered arthritogenic inflammation (Aki et al. 2017; Devos et al. 2019; Liu et al. 2008; Nair et al. 2009; Sahlol et al. 2019; Stuart et al. 2015; Tsoi et al. 2012).

It is noteworthy that many keratinocyte-specific models of psoriasis do not develop arthritis (As reviewed by (Gangwar et al. 2022)), suggesting that dysregulation of specific pathways within keratinocytes may be crucial for connecting epithelial immune dysregulation to arthritis(Gangwar et al. 2022). Our findings demonstrate that unrestrained MyD88 signaling may be one such pathway. MyD88 signaling activates both AP1 and NF-κB transcription factors(Kawai and Akira 2007). Interestingly, prior studies have shown keratinocyte-specific deletion of AP1 transcription factors JunB and JunC (JunBΔep* c-JunΔep* mice) yields spontaneous psoriasiform dermatitis as well as a lymphocyte-dependent arthritis (Zenz et al. 2005). A20 is a key negative regulator of NF-kB signaling by virtue of its ability to non-catalytically bind linear ubiquitin post-translational modifications (Razani et al. 2020b). Thus, loss of homeostasis within keratinocyte AP1 and NF-kB signaling may be a key molecular locus by which skin promotes arthritogenic lymphocytes.

Identifying the factors that drive arthritogenic lymphocytes remains a key priority in the study of psoriatic pathogenesis. The genetic association of human psoriatic disease with particular MHCI alleles suggests a role for antigen-specific mechanisms (Scher et al. 2019; Schett et al. 2022) Innate immune mechanisms are implicated by the association of human PsA with biomechanical trauma and the responsiveness of arthritis to TNF inhibitors (Scher et al. 2019; Schett et al. 2022). We find that A20KIKO mice lacking T cells are protected from PsA-like disease but that arthritis occurs when either αβ or γδ T cells are present. The distinct mechanisms for antigen recognition by αβ or γδ T cells suggests these processes are unlikely to be due to loss of tolerance to a single antigen and points to bystander mechanisms connecting A20-deficient keratinocytes to arthritis. One such bystander mechanism is suggested by the capacity of cultured A20-deficient keratinocytes to directly trigger T cell IL17 secretion in vitro. It is noteworthy that while PsA-like disease is prevented in A20KIKO mice lacking T cells, psoriasiform skin disease continues to occur. This pattern was also observed in JunBΔep* c-JunΔep* mice and may reflect the pathogenic effect of skin-resident ILC3s that are unable to migrate to joints (Zenz et al. 2005).

A prior study has shown that constitutive deletion of A20 in keratinocytes using A20F/F Krt14-Cre mice causes ectodermal defects of hair, nails, and sebaceous glands (Lippens et al. 2011). These mice display altered anatomy of their distal digits, disheveled hair, as well as sebaceous hyperplasia; however, they do not develop spontaneous skin or joint inflammation(Lippens et al. 2011). Several possibilities may explain the distinct spontaneous inflammatory phenotype of A20KIKO mice, in which keratinocyte A20 is deleted in adults with normal ectodermal and immune development. Developmental defects in the distal digit of A20F/F Krt14-Cre mice may alter anatomic and biomechanical relationships within the distal digit synovio-entheseal-nail complex, a structure implicated in human PsA pathogenesis and the primary site of PsA-like inflammation in A20KIKO mice (McGonagle 2009). Second, sebaceous gland hyperplasia and hair abnormalities in A20F/F Krt14-Cre mice may alter properties of the skin barrier or composition of the skin microbiome, modifying subsequent tissue immune responses. Animal facility dependent microbiome differences may also play a role. Finally, recent studies have shown that immune stimuli during early life can induce sustained tolerogenic responses whereas exposure to identical stimuli during adulthood can be inflammatory (Scharschmidt et al. 2015). It is possible that deletion of keratinocyte A20 during fetal stages in A20F/F Krt14-Cre causes anatomic, microbiome, or immune changes that alter their propensity for spontaneous psoriatic disease, although they do display enhanced inflammation in response to topical imiquimod (Devos et al. 2019).

Notably, we did not observe ectodermal abnormalities in our previous study of constitutive A20 zinc finger 7 knock-in mice, suggesting compensation by other A20 domains to allow for normal early development (Razani et al. 2020b). Spontaneous PsA-like disease also occurs following keratinocyte-specific and tamoxifen-induced loss of heterozygosity in adult A20F/ZF7 Krt5-CreERT2 mice, demonstrating that noncatalytic ubiquitin binding by A20’s ZF7 within keratinocytes is the key molecular mechanism by which A20 restricts spontaneous PsA-like disease.

The spontaneous MyD88 signaling observed in A20-deficient keratinocytes raises the question of which upstream ligands and receptors initiate this signaling. Intriguingly, one commonality among several MyD88 utilizing IL1-like cytokine receptors and TLRs is their roles in the tissue damage response. IL-1α, IL-36α, and IL-36γ are constitutively expressed in keratinocytes, upregulated in inflammatory contexts, and released upon tissue damage, where they function as Damage-Associated Molecular Patterns (DAMPs) (Bronneberg et al. 2007; Carrier et al. 2011; Macleod et al. 2020). TLRs also detect a variety of tissue-specific DAMPs (Piccinini and Midwood 2010). It is notable that psoriatic inflammation in A20KIKO mice occurs in ears, tails, and paws, representing exposed anatomic sites prone to repeated mechanical stress and trauma, hypothesized to trigger psoriatic inflammation in both skin and joints (McGonagle et al. 2009; Schett et al. 2022; Thorarensen et al. 2017). Reduced keratinocyte A20 expression may lower the threshold for a pathogenic outcome in the context of repeated DAMP-releasing mechanical damage that triggers MyD88 signaling.

Using spatial transcriptomic data we find that A20 (TNFAIP3) levels are reduced in the epidermal regions of individuals with psoriasis while MyD88-pathway genes are elevated within psoriatic epidermis. This divergence between TNFAIP3 and MyD88-pathway genes in psoriatic epidermis is striking given the fact that TNFAIP3 is normally strongly induced by NF-κB signaling downstream of MyD88, boosting its negative regulatory function (Razani et al. 2020b; Swindell et al. 2018). The divergence between expression of TNFAIP3 and other MyD88-pathway genes suggests that psoriatic disease may in part emerge when genetic or epigenetic factors prevent a sufficiently robust epidermal A20 response to innate immune stimuli. Intriguingly, rare gain-of-function germline mutations in MyD88 are associated with early onset arthritis(Sikora et al. 2018).

A major goal in psoriatic management is precision identification of psoriasis patients at risk for PsA and their treatment with molecularly targeted therapeutics. Reduced A20 expression in combination with increased MyD88 pathway transcripts within epidermal keratinocytes may indicate an early immunological process that drives both cutaneous disease and its transition to arthritis. This group of individuals may particularly benefit from emerging therapeutics that potently inhibit MyD88 signaling through degradation of the downstream kinase IRAK4 (Ackerman et al. 2023).

MATERIALS & METHODS

Sex as a biological variable:

Both male and female mice were included in all experiments and analyses. No sex-based differences were observed. Experimental groups were balanced for sex where possible.

Mice:

A20fl/fl mice were previously described(Tavares et al. 2010). IFNLR−/− mice were and mice were a generous gift from Megan Baldridge (Washington University – St. Louis). IL17A−/− mice were a generous gift from Yoichiro Iwakura (University of Tokyo) via Sarah Gaffen (University of Pittsburgh). MyD88fl/fl mice were a generous gift from Anthony L. DeFranco (University of California – San Francisco). The following were obtained from Jackson Labs B6N.129S6(Cg)-Krt5tm1.1(cre/ERT2)Blh/J (Keratin 5CreERT2+), B6(Cg)-Ifnar1tm1.2Ees/J (IFNAR-deficient), B6.129S-Tnftm1Gkl/J (TNF-deficient mice). Collection of animal tissue samples for this study was approved as part of the study protocol. This animal study was approved by University of California - San Francisco IACUC.

In vivo mouse studies:

Longitudinal analyses of clinical arthritis and histological analysis were performed as described previously(Razani et al. 2020b). Briefly, digits with visible erythema, swelling, and loss of the nail plate were counted as positive for disease in a blinded manner. For histological analyses, the paws of euthanized mice were collected in 10% phosphate-buffered formalin, with further processing for hematoxylin and eosin staining performed at the UCSF Mouse Histology Core Facility. Antibody inhibition was done with weekly intraperitoneal injection of 1 mg anti-IL23A (p19) antibody (Bio X Cell; G23–8) or isotype antibody (Bio X Cell; HRPN). All mice were housed and bred in accordance with UCSF’s institutional guidelines under approved protocol.

Flow Cytometry:

Flow cytometry was performed on ears cut at their base and digits cut at metacarpophalangeal joint. These tissues were finely minced into ~1mm pieces and digested in a mixture of 0.3mg/mL Liberase TL (Sigma Aldrich), 0.1mg/mL DNAse I (Sigma Aldrich) in DMEM with 5% FCS in an Eppendorf Thermomixer with shaking at 900RPM for 90min at 37C. Ear skin was further processed in Miltenyi C tubes using a Miltenyi GentleMACS Tissue Dissociator (Program C). Antibodies used for cells staining are detailed in Supplementary Text.

T cell Conditioned Media Experiments:

Keratinocyte-conditioned media was collected (See Supplementary Text for Keratinocyte Culture Information), centrifuged, filtered through a 20micron filter, concentrated 3x using Pierce Protein Concentrators (10K MWCO), and subsequently diluted 3 fold in the media of cultured T cells (RPMI 1640, Glutamax, 50μM 2-Mercaptoethanol, 1% Penicillin-Streptomycin; all from Gibco). Total untouched T cells were magnetically sorted (MojoSort™ Mouse CD3 T Cell Isolation Kit) from spleens of adult C57BL/6 mice and cultured for 72 hours in 20ng/mL of recombinant murine IL-7 (Biolegend) with indicated keratinocyte conditioned media.

Protein Analysis:

Immunoblot was performed as previously described with XCell SureLock system and Invitrogen NuPage Bis-Tris pre-cast gels (Razani et al. 2020b). The following antibodies: A20 (Cell Signaling D13H3), MyD88 (Cell Signaling D80F5), GAPDH (Santa Cruz Biotechnology sc-32233) were used for immunoblots. ELISAs were performed according to manufacturer’s instructions CXLC1 (Biolegend 447507), IL17A (Invitrogen 88–7371-22), CCL20 (Invitrogen EMCCL20), IL6 (Biolegend 431304), IL1β Biolegend (432604). Multiplex ELISAs were performed by Eve Technologies.

The following methods are described in Supplementary Text:

Skin transcriptomic analysis, Skin and digit protein preparation, Serum Isolation, Flow Cytometry Antibodies, Sample preparation for RNA-Seq, RNA-seq QC and data processing, Pathway analysis, Spatial transcriptomic analysis, Keratinocyte Culture and Crispr-editing.

Supplementary Material

1

ACKNOWLEDGEMENTS

This work was supported by NIH K08 (K08AR077065), NIH R03 (R03AR082350), Dermatology Foundation Fellowship, Arthritis National Research Foundation Award, American Skin Association Award, UCSF Center for the Rheumatic Diseases, and the Department of Veterans Affairs Health Care System, to B.R. NIH NIAID R01 (R01AI135198) to A.M. NIH/NIAMS P30AR070155, the Russell/Engleman Arthritis Center at UCSF and the Department of Veterans Affairs Health Care System, to M.C.N. NIH UC2 (UC2AR081029), to W.L. B.R. is Guarantor of this work.

Footnotes

CONFLICTS OF INTEREST

The authors have declared that no conflict of interest exists

ETHICS STATEMENT

Deidentified human primary keratinocytes obtained with written, informed consent were purchased from ATCC and PromoCell.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

DATA AVAILABILITY

The data that support the findings of this study are available from the corresponding authors Averil Ma and Bahram Razani (averil.ma@ucsf.edu and bahram.razani@ucsf.edu) upon reasonable request. Datasets related to this article can be found at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE294628, hosted at NCBI GEO (Accession: GSE294628).

REFERENCES

  1. Ackerman L, Acloque G, Bacchelli S, Schwartz H, Feinstein BJ, La Stella P, et al. IRAK4 degrader in hidradenitis suppurativa and atopic dermatitis: a phase 1 trial. Nat. Med Nature Publishing Group; 2023;29(12):3127–36 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Adrianto I, Wen F, Templeton A, Wiley G, King JB, Lessard CJ, et al. Association of a functional variant downstream of TNFAIP3 with systemic lupus erythematosus. Nat. Genet 2011;43(3):253–8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Aki A, Nagasaki M, Malynn BA, Ma A, Kagari T. Hypomorphic A20 expression confers susceptibility to psoriasis. PLOS ONE. 2017;12(6):e0180481. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Andrianne M, Assabban A, La C, Mogilenko D, Salle DS, Fleury S, et al. Tristetraprolin expression by keratinocytes controls local and systemic inflammation. JCI Insight American Society for Clinical Investigation; 2017;2(11) Available from: https://insight.jci.org/articles/view/92979 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Billi AC, Ludwig JE, Fritz Y, Rozic R, Swindell WR, Tsoi LC, et al. KLK6 expression in skin induces PAR1-mediated psoriasiform dermatitis and inflammatory joint disease. J. Clin. Invest 2020;130(6):3151–7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bronneberg D, Spiekstra SW, Cornelissen LH, Oomens CWJ, Gibbs S, Baaijens FPT, et al. Cytokine and chemokine release upon prolonged mechanical loading of the epidermis. Exp. Dermatol 2007;16(7):567–73 [DOI] [PubMed] [Google Scholar]
  7. Carrier Y, Ma H-L, Ramon HE, Napierata L, Small C, O’Toole M, et al. Inter-Regulation of Th17 Cytokines and the IL-36 Cytokines In Vitro and In Vivo: Implications in Psoriasis Pathogenesis. J. Invest. Dermatol 2011;131(12):2428–37 [DOI] [PubMed] [Google Scholar]
  8. Castillo RL, Sidhu I, Dolgalev I, Chu T, Prystupa A, Subudhi I, et al. Spatial transcriptomics stratifies psoriatic disease severity by emergent cellular ecosystems. Sci. Immunol American Association for the Advancement of Science; 2023;8(84):eabq7991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen L, Deshpande M, Grisotto M, Smaldini P, Garcia R, He Z, et al. Skin expression of IL-23 drives the development of psoriasis and psoriatic arthritis in mice. Sci. Rep Nature Publishing Group; 2020;10(1):8259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cook PW, Brown JR, Cornell KA, Pittelkow MR. Suprabasal expression of human amphiregulin in the epidermis of transgenic mice induces a severe, early-onset, psoriasis-like skin pathology: Expression of amphiregulin in the basal epidermis is also associated with synovitis. Exp. Dermatol 2004;13(6):347–56 [DOI] [PubMed] [Google Scholar]
  11. Croxford AL, Karbach S, Kurschus FC, Wörtge S, Nikolaev A, Yogev N, et al. IL-6 Regulates Neutrophil Microabscess Formation in IL-17A-Driven Psoriasiform Lesions. J. Invest. Dermatol 2014;134(3):728–35 [DOI] [PubMed] [Google Scholar]
  12. Devos M, Mogilenko DA, Fleury S, Gilbert B, Becquart C, Quemener S, et al. Keratinocyte Expression of A20/TNFAIP3 Controls Skin Inflammation Associated with Atopic Dermatitis and Psoriasis. J. Invest. Dermatol 2019;139(1):135–45 [DOI] [PubMed] [Google Scholar]
  13. Gangwar RS, Gudjonsson JE, Ward NL. Mouse Models of Psoriasis – A Comprehensive Review. J. Invest. Dermatol 2022;142(3 Pt B):884–97 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Hawkes JE, Yan BY, Chan TC, Krueger JG. Discovery of the IL-23/IL-17 Signaling Pathway and the Treatment of Psoriasis. J. Immunol 2018;201(6):1605–13 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kawai T, Akira S. Signaling to NF-κB by Toll-like receptors. Trends Mol. Med Elsevier; 2007;13(11):460–9 [DOI] [PubMed] [Google Scholar]
  16. Kobayashi T, Ricardo-Gonzalez RR, Moro K. Skin-Resident Innate Lymphoid Cells – Cutaneous Innate Guardians and Regulators. Trends Immunol. 2020;41(2):100–12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Lazear HM, Schoggins JW, Diamond MS. Shared and Distinct Functions of Type I and Type III Interferons. Immunity. 2019;50(4):907–23 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Lippens S, Lefebvre S, Gilbert B, Sze M, Devos M, Verhelst K, et al. Keratinocyte-specific ablation of the NF- κ B regulatory protein A20 (TNFAIP3) reveals a role in the control of epidermal homeostasis. Cell Death Differ. 2011;18(12):1845–53 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Liu Y, Helms C, Liao W, Zaba LC, Duan S, Gardner J, et al. A genome-wide association study of psoriasis and psoriatic arthritis identifies new disease loci. PLoS Genet. 2008;4(3):e1000041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Liu X-Q, Zhou P-L, Yin X-Y, Wang A-X, Wang D-H, Yang Y, et al. Circulating inflammatory cytokines and psoriasis risk: A systematic review and meta-analysis. PLOS ONE. Public Library of Science; 2023;18(10):e0293327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Macleod T, Ainscough JS, Hesse C, Konzok S, Braun A, Buhl A-L, et al. The Proinflammatory Cytokine IL-36γ Is a Global Discriminator of Harmless Microbes and Invasive Pathogens within Epithelial Tissues. Cell Rep. 2020;33(11):108515. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Martens A, Priem D, Hoste E, Vetters J, Rennen S, Catrysse L, et al. Two distinct ubiquitin-binding motifs in A20 mediate its anti-inflammatory and cell-protective activities. Nat. Immunol 2020;21(4):381–7 [DOI] [PubMed] [Google Scholar]
  23. McGonagle D Enthesitis: an autoinflammatory lesion linking nail and joint involvement in psoriatic disease. J. Eur. Acad. Dermatol. Venereol 2009;23(s1):9–13 [DOI] [PubMed] [Google Scholar]
  24. McGonagle D, Benjamin M, Tan AL. The pathogenesis of psoriatic arthritis and associated nail disease: not autoimmune after all? Curr. Opin. Rheumatol 2009;21(4):340. [DOI] [PubMed] [Google Scholar]
  25. Nair RP, Duffin KC, Helms C, Ding J, Stuart PE, Goldgar D, et al. Genome-wide scan reveals association of psoriasis with IL-23 and NF-kappaB pathways. Nat. Genet 2009;41(2):199–204 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Piccinini AM, Midwood KS. DAMPening Inflammation by Modulating TLR Signalling. Mediators Inflamm. 2010;2010(1):672395. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Razani B, Malynn BA, Ma A. Preserving immune homeostasis with A20. Adv. Immunol 2020a;148:1–48 [DOI] [PubMed] [Google Scholar]
  28. Razani B, Whang MI, Kim FS, Nakamura MC, Sun X, Advincula R, et al. Non-catalytic ubiquitin binding by A20 prevents psoriatic arthritis-like disease and inflammation. Nat. Immunol 2020b;21(4):422–33 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ritchlin CT, Colbert RA, Gladman DD. Psoriatic Arthritis. N. Engl. J. Med. Massachusetts Medical Society; 2017;376(10):957–70 [DOI] [PubMed] [Google Scholar]
  30. Rodriguez-Rosales YA, Langereis JD, Gorris MAJ, van den Reek JMPA, Fasse E, Netea MG, et al. Immunomodulatory aged neutrophils are augmented in blood and skin of psoriasis patients. J. Allergy Clin. Immunol 2021;148(4):1030–40 [DOI] [PubMed] [Google Scholar]
  31. Sahlol NY, Mostafa MS, Madkour LAE-F, Salama DM. Low TNFAIP3 expression in psoriatic skin promotes disease susceptibility and severity. PLOS ONE. 2019;14(5):e0217352. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Scharschmidt TC, Vasquez KS, Truong H-A, Gearty SV, Pauli ML, Nosbaum A, et al. A Wave of Regulatory T Cells into Neonatal Skin Mediates Tolerance to Commensal Microbes. Immunity. Elsevier; 2015;43(5):1011–21 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Scher JU, Ogdie A, Merola JF, Ritchlin C. Preventing psoriatic arthritis: focusing on patients with psoriasis at increased risk of transition. Nat. Rev. Rheumatol Nature Publishing Group; 2019;15(3):153–66 [DOI] [PubMed] [Google Scholar]
  34. Schett G, Rahman P, Ritchlin C, McInnes IB, Elewaut D, Scher JU. Psoriatic arthritis from a mechanistic perspective. Nat. Rev. Rheumatol Nature Publishing Group; 2022;18(6):311–25 [DOI] [PubMed] [Google Scholar]
  35. Sikora KA, Bennett JR, Vyncke L, Deng Z, Tsai WL, Pauwels E, et al. Germline Gain-of-Function MYD88 Mutation in a Child with Severe Arthritis. J. Allergy Clin. Immunol 2018;141(5):1943–1947.e9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Simmons J, Gallo RL. The Central Roles of Keratinocytes in Coordinating Skin Immunity. J. Invest. Dermatol 2024;144(11):2377–98 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Stuart PE, Nair RP, Tsoi LC, Tejasvi T, Das S, Kang HM, et al. Genome-wide Association Analysis of Psoriatic Arthritis and Cutaneous Psoriasis Reveals Differences in Their Genetic Architecture. Am. J. Hum. Genet 2015;97(6):816–36 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Swindell WR, Beamer MA, Sarkar MK, Loftus S, Fullmer J, Xing X, et al. RNA-Seq Analysis of IL-1B and IL-36 Responses in Epidermal Keratinocytes Identifies a Shared MyD88-Dependent Gene Signature. Front. Immunol 2018;9:80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Tavares RM, Turer EE, Liu CL, Advincula R, Scapini P, Rhee L, et al. The ubiquitin modifying enzyme A20 restricts B cell survival and prevents autoimmunity. Immunity. 2010;33(2):181–91 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Taylor W, Gladman D, Helliwell P, Marchesoni A, Mease P, Mielants H. Classification criteria for psoriatic arthritis: Development of new criteria from a large international study. Arthritis Rheum. 2006;54(8):2665–73 [DOI] [PubMed] [Google Scholar]
  41. Thorarensen SM, Lu N, Ogdie A, Gelfand JM, Choi HK, Love TJ. Physical trauma recorded in primary care is associated with the onset of psoriatic arthritis among patients with psoriasis. Ann. Rheum. Dis 2017;76(3):521–5 [DOI] [PubMed] [Google Scholar]
  42. Tsoi LC, Hile GA, Berthier CC, Sarkar MK, Reed TJ, Liu J, et al. Hypersensitive interferon responses in lupus keratinocytes reveal key mechanistic determinants in cutaneous lupus. J. Immunol. Baltim. Md 1950 2019;202(7):2121–30 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Tsoi LC, Spain SL, Knight J, Ellinghaus E, Stuart PE, Capon F, et al. Identification of 15 new psoriasis susceptibility loci highlights the role of innate immunity. Nat. Genet 2012;44(12):1341–8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Wang S, Wen F, Wiley GB, Kinter MT, Gaffney PM. An Enhancer Element Harboring Variants Associated with Systemic Lupus Erythematosus Engages the TNFAIP3 Promoter to Influence A20 Expression. PLOS Genet. 2013;9(9):e1003750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Winge MCG, Ohyama B, Dey CN, Boxer LM, Li W, Ehsani-Chimeh N, et al. RAC1 activation drives pathologic interactions between the epidermis and immune cells. J. Clin. Invest American Society for Clinical Investigation; 2016;126(7):2661–77 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Wright V, Moll JM. Psoriatic arthritis. Bull. Rheum. Dis 1971;21(5):627–32 [PubMed] [Google Scholar]
  47. Yamamoto M, Nakajima K, Takaishi M, Kitaba S, Magata Y, Kataoka S, et al. Psoriatic Inflammation Facilitates the Onset of Arthritis in a Mouse Model. J. Invest. Dermatol 2015;135(2):445–53 [DOI] [PubMed] [Google Scholar]
  48. Zenz R, Eferl R, Kenner L, Florin L, Hummerich L, Mehic D, et al. Psoriasis-like skin disease and arthritis caused by inducible epidermal deletion of Jun proteins. Nature. Nature Publishing Group; 2005;437(7057):369–75 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2205635-supplement-1.pdf (2.9MB, pdf)

Data Availability Statement

The data that support the findings of this study are available from the corresponding authors Averil Ma and Bahram Razani (averil.ma@ucsf.edu and bahram.razani@ucsf.edu) upon reasonable request. Datasets related to this article can be found at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE294628, hosted at NCBI GEO (Accession: GSE294628).

RESOURCES