Psoriasis is a chronic auto-immune skin disease that affects 2–3% of the European population. Disease characteristics include epidermal hyperplasia, infiltration of leukocytes in the dermis and erythema (Dainichi et al., 2018). Long-term activation of the immune system in the skin can also lead to tissue damage in other organs, and psoriasis is generally regarded as a systemic disease. Consequently, numerous comorbidities, including psoriatic arthritis, inflammatory bowel disease (IBD) and cardiovascular disease, are associated with psoriasis (Greb et al., 2016).

Genome-wide association studies have revealed over 60 loci that contribute to the development of psoriasis. These include candidate genes involved in skin barrier function and the immune response (Liang et al., 2017a). It is thought that many low risk genetic variants together with environmental factors increase the predisposition of individuals to psoriasis development. Psoriasis vulgaris (PV), the most common form, is characterised by stable erythematous scaly plaques, predominantly located on knees and elbows, in which the adaptive immune system predominates. In contrast, generalised pustular psoriasis (GPP), a very rare form of the disease that is characterised by widespread diffuse skin inflammation and sub-corneal pustules (Munro’s microabscesses and spongiform pustules of Kogoj), involves massive infiltration of neutrophils in the skin. GPP patients also experience acute exacerbations (flare ups) involving development of new skin lesions which can cover >70% of the body surface, fever, cheilitis (scaling and superficial ulceration of the lips), diarrhoea and dehydration (Ly et al., 2019). GPP exacerbations, which often meet the criteria for systemic inflammatory response syndrome, involve neutrophil infiltration in multiple organs, are difficult to treat and associated with significant morbidity and mortality (3–7%) (Gooderham et al., 2019).

Gain-of-function, highly penetrant (>90%) and dominantly acting mutations within the CARD14 gene can trigger the development of PV or GPP (Jordan et al., 2012b). CARD14 (CARMA2) is a member of the CARMA family of scaffolding proteins that includes CARD11 (CARMA1) and CARD10 (CARMA3) (Lu et al., 2019). Each of these proteins has a similar domain structure, comprising an N-terminal CARD domain, followed by a coiled-coil (CC) domain, and a C-terminal MAGUK domain (PDZ-SH3-GUK). CARD11 and CARD10 play critical roles in the activation of NF-κB transcription factors following ligation of antigen receptors and G-protein-coupled receptors, respectively (Lu et al., 2019). NF-κB, composed of dimers of Rel polypeptides, regulates gene expression by binding to κB elements in the promoters and enhancers of multiple target genes that control immune and inflammatory responses (Zhang et al., 2017).

The structural similarity to CARD11 and CARD10 suggests a role for CARD14 in NF-κB activation. Consistent with this, the highly penetrant CARD14^E138A psoriasis-associated mutation, first identified as a sporadic mutation in a child suffering from GPP (Jordan et al., 2012b), induces CARD14 interaction with BCL10 (B cell lymphoma protein 10) and the paracaspase MALT1 (mucosa-associated lymphoid tissue lymphoma translocation protein 1) (Howes et al., 2016). This complex triggers activation of the IκB kinase (IKK) complex, the central activator of NF-κB transcription factors, leading to constitutive activation of NF-κB and expression of pro-inflammatory cytokine/chemokine genes in keratinocytes in the absence of exogenous stimulation.

Recent studies have investigated how psoriasis-associated CARD14 mutations induce skin inflammation by generating knock-in mice expressing the mouse equivalent CARD14 variants (Mellett et al., 2018; Sundberg et al., 2019; Wang et al., 2018). These mice develop psoriasiform skin inflammation that is partially dependent on the cytokines IL-17A and IL-23, which play important roles in human psoriasis (Greb et al., 2016). Although these studies have confirmed the importance of CARD14 mutations in inducing skin inflammation, the constitutive nature of the knock-in mutations generated have precluded detailed study of disease pathogenesis. Furthermore, the constitutive CARD14^E138A mutation is associated with embryonic or peri-natal lethality, preventing analysis of disease in adult skin. To circumvent these limitations, we generated a new knock-in mouse allele Card14^LSL-E138A that allows conditional expression of the human CARD14^E138A mutation in the homologous mouse locus following tamoxifen injection. The inducible nature of this mutation allowed us to study the direct effects of CARD14^E138A signalling, rather than potentially indirect secondary or tertiary consequences of constitutively expressed CARD14^E138A. Tamoxifen-induced expression of Card14^E138A in adult mice also ruled out any developmental effects of CARD14^E138A signalling on skin inflammation.

We present evidence that conditional heterozygous expression of CARD14^E138A in adult mice rapidly promoted psoriasiform skin inflammation, which was independent of adaptive immune cells and due to signalling in keratinocytes. However, homozygous expression of CARD14^E138A induced rapid weight loss, temperature reduction and extensive neutrophil infiltration in multiple organs. This severe systemic phenotype resulted from CARD14^E138A signalling in keratinocytes and showed striking similarities to an acute flare of GPP in human patients. Heterozygous and homozygous phenotypes were both ameliorated by antibody neutralisation of tumour necrosis factor (TNF). Our findings suggest that specific blockade of the CARD14 signalling pathway in keratinocytes would inhibit the systemic inflammatory syndrome that can develop in GPP patients with activating CARD14 mutations.

Human and mouse CARD14 proteins are highly homologous, with 77% amino acid identity overall and 80% identity in the coiled-coil region in which the CARD14^E138A mutation is located (Figure 1—figure supplement 1A). To study how CARD14^E138A mutation induces skin inflammation, we developed a conditional knock-in mouse strain, Card14^{LSL-E138A/LSL-E138A}, in which a floxed minigene (exons 4–22) encoding wild type (WT) C-terminus of CARD14 linked to a 3xFLAG tag was inserted between exons 3 and 4 of the Card14 locus and an E138A point mutation was introduced into endogenous Card14 exon 5 (Figure 1—figure supplement 1B). In the absence of Cre-mediated recombination, Card14 was expressed from exon 3 and the inserted minigene to produce WT CARD14-3xFLAG. After Cre-mediated recombination, the minigene was excised, allowing transcription of Card14 from the endogenous exons and expression of CARD14^E138A.

CARD14 is expressed at high levels in differentiated keratinocytes of the skin epidermis

In order to understand the effects of Card14^E138A mutation, it was first important to know where CARD14 protein is normally expressed. High levels of CARD14-3XFLAG were detected in the back skin and ears, as expected, and also in the colon, small intestine and caecum (Figure 1A). Low levels of CARD14-3XFLAG were additionally found in the liver and kidney.

Figure 1 with 3 supplements see all

Download asset Open asset

RNAscope analysis revealed that Card14 mRNA expression in healthy mouse skin was restricted to the upper (most differentiated) keratinocytes, the outer root sheath of the hair and sebocytes in the skin (Figure 1B). These data contrasted with earlier immunohistochemistry results suggesting CARD14 expression in the basal layer of human skin (Fuchs-Telem et al., 2012; Jordan et al., 2012b). However, the RNAscope analyses were in line with published RNA sequencing data from different layers of the mouse epidermis (Figure 1—figure supplement 1C) and single cell sequencing of murine epidermis (http://kasperlab.org/mouseskin) which also indicated Card14 mRNA expression in differentiated keratinocytes (Asare et al., 2017). Consistent with this, CARD14 protein expression was induced in mouse keratinocytes following in vitro differentiation in culture medium containing CaCl₂ (Figure 1—figure supplement 1D; Bikle et al., 2012). CARD14 mRNA and CARD14 protein expression was similarly induced in cultured human keratinocytes concurrent with the keratinocyte differentiation marker involucrin (Figure 1—figure supplement 1E and F). The CARD14 mRNA expression data agree with published transcriptomic datasets of differentiating human keratinocytes (Bin et al., 2016) and organotypic epidermis (Lopez-Pajares et al., 2015; Figure 1—figure supplement 1G).

CARD14^E138A signalling in keratinocytes induces skin inflammation independently of the adaptive immune system

CARD14 is expressed at high levels in skin keratinocytes but is also expressed at lower levels in intestinal epithelial cells and other tissues (Figure 1A). To test whether the skin inflammation in the Card14^LSL-E138A/+Rosa26^CreERT2/+ mice was caused by keratinocyte-intrinsic CARD14^E138A signalling, Card14^{LSL-E138A/LSL-E138A} mice were crossed with Krt14^CreERT2 mice. Tamoxifen injection of the resulting Card14^E138A/+ Krt14^CreERT2/+ mice induced expression of CARD14^E138A in keratinocytes (Figure 2—figure supplement 1A). This induced acanthosis, hyperkeratosis, infiltration of S100A9⁺ myeloid cells and increased expression of chemokines and proinflammatory cytokines in the ears and the back skin at 5d after tamoxifen injection (Figure 2A,B,C,D and Figure 2—figure supplement 1B,C,D and E). Some Card14^E138A/+ Krt14^CreERT2/+ mice presented a visible skin phenotype without tamoxifen injection, which could be due to leaky expression of Cre allowing basal expression of CARD14^E138A, these were excluded from experiments. Skin expression of psoriasis-associated genes at 5d was increased in Card14^LSL-E138A/+Krt14^CreERT2/+ mice compared to controls, similar to Card14^LSL-E138A/+Rosa26^CreERT2/+ mice (Figure 2—figure supplement 2A). However, the fold changes for the majority of these transcripts were higher on the Rosa26^CreERT2 background. Nevertheless, these experiments showed that CARD14^E138A signalling in keratinocytes alone was sufficient to induce skin inflammation.

Figure 2 with 2 supplements see all

Download asset Open asset

To investigate the role of the adaptive immune system in the development of CARD14^E138A-induced skin pathology, Card14^LSL-E138A/+Rosa26^CreERT2/+Rag1^-/- mice were generated that lacked T and B cells. Tamoxifen induced a skin phenotype very similar to Card14^LSL-E138A/+Rosa26^CreERT2/+ Rag1^+/+ controls, with acanthosis, hyperkeratosis, S100A9⁺ myeloid cell infiltration and expression of chemokines and proinflammatory cytokines at 5d following tamoxifen injection (Figure 2E,F,G,H and Figure 2—figure supplement 1F,G,H and I). Consequently, the adaptive immune system was not required for acute skin inflammation induced by ubiquitous CARD14^E138A expression.

Neutralisation of TNF efficiently blocks skin inflammation induced by CARD14^E138A

Two of the most common and efficient treatments for moderate to severe psoriatic patients are antibody blocking of TNF and IL-17A (Greb et al., 2016). We next assessed whether the inflammatory skin phenotype induced by CARD14^E138A could be ameliorated by neutralising either of these cytokines following the protocol shown in Figure 3A.

Figure 3 with 3 supplements see all

Download asset Open asset

Anti-TNF substantially reduced acanthosis in ears and back skin of tamoxifen-induced Card14^LSL-E138A/+ Rosa26^CreERT2/+ mice such that epidermal thickness was similar to that detected in Card14^+/+ Rosa26^CreERT2/+ control mice after tamoxifen injection (Figure 3B and Figure 3—figure supplement 1A). Skin expression of mRNAs encoding chemokines and proinflammatory cytokines was also significantly reduced by anti-TNF in the ears and the back skin (Figure 3C and Figure 3—figure supplement 1B). Flow cytometric analyses demonstrated that anti-TNF reduced infiltration of neutrophils, DCs and eosinophils into the ears (Figure 3D).

Expression of Card14 mRNA in the skin of Card14^LSL-E138A/+Rosa26^CreERT2/+ mice 5d after tamoxifen injection was preferentially localised to the suprabasal layers of the epidermis (Figure 3—figure supplement 2A). RNAscope analysis showed that Tnfa mRNA had a staining pattern resembling that of Card14 mRNA (Figure 3—figure supplement 2B). Co-staining with specific antibodies revealed that Tnfa mRNA was largely produced by keratinocytes, with some contribution by neutrophils (Figure 3—figure supplement 3A). Tnfa mRNA was not detected in T cells in the skin. These results suggest that keratinocytes are the main source of TNF driving skin inflammation.

Blocking of IL-17A also resulted in a significant reduction in epidermal thickness of the back skin, but did not significantly reduce epidermal thickness in the ears (Figure 3—figure supplement 1A and Figure 3B). The expression of pro-inflammatory cytokines and chemokines in the ears was not altered by anti-IL-17A, although S100a9 mRNA was reduced (Figure 3C). Anti-IL-17A reduced Cxcl3 and S100a9 mRNAs in the back skin of Card14^LSL-E138A/+ Rosa26^CreERT2/+ mice, but did not significantly alter the mRNA expression of the other cytokines and chemokines assayed (Figure 3—figure supplement 1B).

Together these results indicated that the acute inflammatory skin phenotype induced by tamoxifen injection of Card14^LSL-E138A/+ Rosa26^CreERT2/+ mice was predominantly driven by TNF produced by keratinocytes, while IL-17A appeared to play a less important role. These findings were consistent with the lack of requirement for adaptive immune cells for Card14^E138A to induce acanthosis and expression of pro-inflammatory genes in the skin, since αβ and γδ T cells are a major source of IL-17A in psoriasis (Prinz et al., 2020). Consistent with this, Il17a mRNA levels were not increased in ears of Card14^LSL-E138A/+Rosa26^CreERT2/+Rag1^-/- mice following tamoxifen induction (Figure 2H).

Comparison of the inflammatory effects of acute versus chronic CARD14^E138A expression

Histological and flow cytometric analyses of ears revealed clear evidence of acanthosis and skin inflammation, respectively, 5d following tamoxifen injection of Card14^LSL-E138A/+ Rosa26^CreERT2/+ mice (Figure 1), but externally the ears appeared normal (Figure 4A). However, one month (1 m) after CARD14^E138A induction, large scales with sharply demarcated edges were clearly evident on the ears of the mutant mice (Figure 4A). No obvious external back skin phenotype was visible at 5d or 1 m. Histologic examination of ears revealed increased acanthosis at 1 m compared with 5d (Figure 4B). In contrast, epidermal thickening of the back skin, although still detected, was significantly reduced at 1 m compared to 5d (Figure 4B).

Figure 4 with 1 supplement see all

Download asset Open asset

Flow cytometry was used to compare immune cell infiltration in the ears and back skin at 5d and 1 m post-tamoxifen injection (Figure 4C). At 5d, the immune cell infiltrates in the ears and skin was largely composed of innate cells, with substantially increased numbers of neutrophils, macrophages, dendritic cells and NK cells. The number of αβ T cells was only modestly increased (approximately two-fold) at 5d in ears, but not altered in back skin. However, the numbers of αβ T cells, including Tregs, were substantially increased in both ears and back skin at 1 m. CARD14^E138A expression did not increase numbers of γδ T cells in ears or back skin at either time point.

Analysis of serum levels of TNF, IL-6 and IL-17A also revealed clear differences between 5d and 1 m post tamoxifen injection of Card14^LSL-E138A/+ Rosa26^CreERT2/+ mice. Each of these cytokines was elevated at 5d, but reduced to WT levels at 1 m (Figure 4D). Paralleling the early systemic overproduction of pro-inflammatory cytokines, Card14^LSL-E138A/+ Rosa26^CreERT2/+ mice suffered a decrease in body weight between 6-7d post-tamoxifen, returning to control levels by 10d (Figure 4E).

To investigate the role of the adaptive immune system in skin pathology induced by prolonged CARD14^E138A signalling, Card14^LSL-E138A/+Rosa26^CreERT2/+Rag1^-/- mice were analysed 1 m after tamoxifen injection. The absence of an adaptive immune system did not decrease ear and skin acanthosis, the abundance of S100A9⁺ myeloid cells or the levels of proinflammatory mRNA transcripts induced by CARD14^E138A (Figure 4—figure supplement 1A–H). Indeed, despite the absence of T and B cells, Card14^LSL-E138A/+Rosa26^CreERT2/+Rag1^-/- mice compared to Card14^LSL-E138A/+Rosa26^CreERT2/+Rag1^+/+ controls had significantly higher numbers of infiltrating NK, eosinophils and monocytes, while numbers of neutrophils and macrophages were similar (Figure 4—figure supplement 1I). Thus, the adaptive immune system was not required for skin inflammation induced by chronic CARD14^E138A signalling.

Taken together, these results suggest that acute CARD14^E138A signalling in keratinocytes induced the transient systemic release of pro-inflammatory mediators that produced body weight loss. This early inflammatory burst was accompanied by a skin infiltrate predominantly composed of innate immune cells. After this acute phase, the skin lesion evolved and external ear skin morphology became similar to plaque psoriasis in humans. Chronic CARD14^E138A signalling involved a prominent influx of T cells, although this was not required for induction of skin inflammation (from analysis of Card14^LSL-E138A/+Rosa26^CreERT2/+Rag1^-/- mice), and reduced levels of pro-inflammatory cytokines systemically. The progression of the pathology in the Card14^LSL-E138A/+ Rosa26^CreERT2/+ mice resembled the flares suffered by psoriatic patients, which eventually undergo spontaneous resolution (Greb et al., 2016).

CARD14^E138A signalling induces dynamic changes in the skin transcriptome

To gain more insight into the mechanisms by which CARD14^E138A induced skin inflammation, RNA was extracted from the ears of Card14^LSL-E138A/+Rosa26^CreERT2/+ and Card14^+/+Rosa26^CreERT2/+ mice 5d and 1 m after tamoxifen injection and subjected to RNA sequencing analysis. A principal component analysis indicated that the main difference between sample groups was the genotype (PCA1); CARD14^E138A expression induced a genetic response clearly differentiated from the control groups (Figure 5A). The gene responses in the mutant mice were clearly distinct between 5d and 1 m, resulting in signalling duration being identified as the second cause of variation (PCA2) between sample groups (Figure 5A).

Figure 5 with 3 supplements see all

Download asset Open asset

458 genes were found to be upregulated and 88 genes downregulated more than 2-fold in the Card14^LSL-E138A/+Rosa26^CreERT2/+ compared to the controls at 5d. Interestingly, the pattern of gene expression changed between 5d and 1 m, as shown by the different colour intensities in the heat map, indicating a change in transcriptional pattern with prolonged CARD14^E138A signalling (Figure 5B). Pathway analysis revealed that TNF signalling via NF-κB was one of the pathways most strongly upregulated by CARD14^E138A at both 5d and 1 m (Figure 5C). Strikingly, pathways related to extracellular matrix formation were also largely reduced in the Card14^LSL-E138A/+Rosa26^CreERT2/+ mice at 5d compared to the other groups. Expression of keratinisation genes was enriched in the Card14^LSL-E138A/+Rosa26^CreERT2/+ mice relative to controls, especially at 1 m.

Among the genes more strongly upregulated by CARD14^E138A at 5d compared to 1 m were Il6, Il1a, Osm, Mmp8, Il22, Ccl20, Cxcl5, Ccl17 and Il17c (Figure 5D). Several collagen genes (Col11a1, Col6a6, Col13a1, Col14a1 and Col16a1) were downregulated by CARD14^E138A at 5d and returned to WT levels by 1 m. These changes may be due to the effects of IL-17A (and possibly IL-17C) on the extracellular matrix (Nakashima et al., 2012). Cornification genes (Klk8, Casp14, Flg and Krt6a) were upregulated at 1 m compared to 5d, which was likely a secondary effect of chronic CARD14^E138A signalling. Krt6a is implicated in psoriasis pathogenesis and is associated with the wound healing/regenerative phenotype seen in psoriasis (Lessard et al., 2013). Genes related to immune cells changed in line with the types of immune cell infiltrates seen at different time points (Figure 4C). Ly6g (a marker for neutrophils) was markedly upregulated at 5d, while markers for T cells (Cd4, Cd8a and Foxp3) were only upregulated at 1 m.

Together the results in this section indicated that conditional expression of CARD14^E138A in adult mice allowed the evolution of CARD14^E138A-induced skin inflammation to be monitored. This involved downregulation of some key proinflammatory genes at 1 m, which may reflect a homeostatic response to chronic activation of CARD14^E138A signalling pathways. Importantly, several genes relevant to the pathology of psoriasis (S100a9, Il1f9, Cxcl2, Il17a and Tnfa) remained significantly elevated in Card14^LSL-E138A/+Rosa26^CreERT2/+ mice at 1 m and may be necessary in maintaining the inflammatory skin phenotype promoted by chronic CARD14^E138A signalling.

Homozygous expression of Card14^E138A induces a severe systemic illness

The majority of psoriasis-associated CARD14 mutations identified are heterozygous suggesting that homozygous expression of such CARD14 mutations might exceed a threshold of NF-κB activation that is compatible with life. To investigate whether the inflammatory effects of CARD14^E138A were dose dependent, the responses of Card14^{LSL-E138A/LSL-E138A}Rosa26^CreERT2/+ (homozygous) mice to tamoxifen injection were compared with Card14^LSL-E138A/+Rosa26^CreERT2/+ (heterozygous) mice.

Tamoxifen injection of Card14^LSL-E138A/+Rosa26^CreERT2/+ mice did not result in any significant weight changes up to 4d (Figure 6A). In contrast, Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice started losing weight around 3d after tamoxifen induction. By 5d, homozygous mice had lost near to 20% of their initial body weight (Figure 6A), reaching humane end-points and requiring sacrifice. Upon general inspection, homozygous mice at d5 appeared scaly, emaciated, inactive and hunched, while heterozygous mice remained in good health.

Figure 6 with 2 supplements see all

Download asset Open asset

The fraction of body surface covered by visible scales appeared to be greater in the homozygous mice than their heterozygous littermates. The middle and lower back skin of the heterozygous mice was largely clear of visible scaling, although areas of skin that were easily accessible to the mice (upper back and beneath the chin) often presented fine scales. In contrast, large areas of back skin in the homozygous mice, including sections less easy to scratch, were covered in fine scales (Figure 6—figure supplement 1A). The skin of some homozygotes also appeared erythrodermic. The ears of the homozygous mice were visibly scalier than those of the heterozygotes, although the foot pads of both genotypes were similarly affected (Figure 6—figure supplement 1B and C). Of particular note, the lips of the d5 homozygous mice were consistently chapped (Figure 6—figure supplement 1D), while the lips of the heterozygous mice appeared normal.

On d5, homozygous mice were cold to the touch and measurement of rectal temperature confirmed hypothermia (Figure 6B). In contrast, rectal temperatures of heterozygous mice were similar to WT controls. Frequently, homozygous mice developed diarrhoea and dissection revealed that small and large intestines and caecum were filled with runny yellow fluid and occasional pockets of gas (Figure 6—figure supplement 2A). Despite this apparent gut phenotype, histological analysis of small intestine (not shown) and colon (Figure 6C) did not reveal any gross tissue damage. Although homozygous mice appeared emaciated and dehydrated, the stomachs were full of food and much enlarged. The livers of the homozygous mice often appeared unusually pale. Histological examination of the ear and lower back skin revealed a similar extent of acanthosis present in heterozygous and homozygous mice (Figure 6D). Hence, while body surface involvement, as measured by visible external scaling, was greater in the homozygous mice, epidermal thickening was comparable between genotypes.

Flow cytometry showed increased neutrophil infiltration into the ears of the homozygous mice as compared to the heterozygotes (Figure 6E). Moreover, qRT-PCR analysis of d5 ear mRNA revealed a gene dosage effect of Card14^E138A for the majority of genes analysed and the expression of key psoriasis associated genes was higher in homozygous mice than in heterozygous mice (Figure 6F). Immunohistochemical analyses revealed significantly higher infiltration of S100A9⁺ myeloid cells in kidney, liver, colon, and lung in homozygous mice compared to heterozygotes (Figure 6C, lower panels and 6G). The mRNA expression of several pro-inflammatory genes, including Ccl20, Cxcl2, S100A9 and Il17c, was also significantly increased in the lungs of homozygous mice, but not heterozygous mice, compared to WT controls (Figure 6F). Analyses of sera demonstrated that circulating levels of TNF and IL-1α were increased in homozygous mice compared to heterozygotes and there was a trend toward increased IL-6 levels (Figure 6H). Similarly, serum ALT (alanine aminotransferase), GLDH (glutamate dehydrogenase) and inorganic phosphate were significantly elevated in the homozygous, but not heterozygous mice (Figure 6I). Increased circulating levels of ALT and GLDH are indicative of liver damage, while increased inorganic phosphate is indicative of defective kidney function (Vervloet et al., 2017). These results suggested multi-organ dysfunction in d5 homozygous mice.

In conclusion, the results in this section indicated that homozygous expression of Card14^E138A in adult mice caused a severe systemic illness, in addition to widespread skin inflammation. This phenotype had striking similarities to the systemic phenotype displayed in exacerbations of GPP (Bachelez, 2020).

Keratinocyte-intrinsic CARD14^E138A signalling induces systemic inflammation independently of adaptive immune cells

Immunoblotting analyses demonstrated that the skin and gastrointestinal tract were the main sites of WT CARD14 expression (Figure 1A). RNAscope analysis of the colon revealed Card14 expression to be largely restricted to the single layer of epithelial cells lining the gut lumen (Figure 6—figure supplement 2B). Furthermore, Card14 mRNA expression was found to increase from proximal to distal ends of the colon (Figure 6—figure supplement 2B–D).

The colons of Card14^LSL-E138A/+ Rosa26^CreERT2/+ mice exhibited infiltration of S100A9⁺ myeloid cells at 5d after tamoxifen and increased expression of mRNAs encoding proinflammatory cytokines (Figure 6—figure supplement 2E–F). Given the unusual gut phenotype observed in homozygous Card14^{LSL-E138A/LSL-E138A}Rosa26^CreERT2/+ mice following tamoxifen injection, it was possible that gut-specific expression of CARD14^E138A was driving the severe systemic phenotype.

To investigate this, Card14^{LSL-E138A/LSL-E138A} Villin^CreERT2 were generated to restrict CARD14^E138A expression to the epithelial cells of the gut. Tamoxifen injection of the resulting Card14^{LSL-E138A/LSL-E138A} Villin^CreERT2 mice, induced expression of CARD14^E138A only in intestinal epithelial cells (Figure 6—figure supplement 2G). Card14^{LSL-E138A/LSL-E138A} Villin^CreERT2 mice exhibited no gross phenotype at d5 after tamoxifen and were healthy (Figure 7A), with no signs of skin disease or organ dysfunction, and presented a histologically normal colon even 1 m after induction (Figure 6—figure supplement 2H). Tamoxifen injection induced expression of Ccl20, Tnfa and Nos2 mRNAs in the gut of Card14^{LSL-E138A/LSL-E138A} Villin^CreERT2 mice resulting from CARD14^E138A signalling in gut epithelial cells (Figure 6—figure supplement 2I). These results suggested that CARD14^E138A signalling in the gut was not sufficient to promote skin inflammation or the systemic inflammatory phenotype.

Figure 7

Download asset Open asset

To determine whether CARD14^E138A signalling in keratinocytes was driving the severe systemic phenotype, homozygous Card14^{LSL-E138A/LSL-E138A}Krt14^CreERT2 mice were generated. Tamoxifen injection of these mice rapidly induced a pronounced loss of weight, comparable to that in Card14^{LSL-E138A/LSL-E138A}Rosa26^CreERT2/+ mice (Figure 7A). Card14^{LSL-E138A/LSL-E138A}Krt14^CreERT2 mice also became hypothermic following tamoxifen administration and displayed significant infiltration of S100A9⁺ myeloid cells in the lungs (Figure 7B and C). Upon dissection, these mice also showed signs of organ dysfunction, most notably swollen stomach and intestines (data not shown), akin to those seen in the Card14^{LSL-E138A/LSL-E138A}Rosa26^CreERT2/+ mice (Figure 6—figure supplement 2A). These results indicated that keratinocyte-intrinsic CARD14^E138A signalling was sufficient to induce severe systemic inflammation.

To determine whether the adaptive immune system contributed to the development of systemic disease induced by homozygous CARD14^E138A expression, Card14^{LSL-E138A/LSL-E138A}Rosa26^CreERT2/+ Rag1^-/- mice were generated. The absence of T and B cells in these mice did not ameliorate the systemic phenotype induced by homozygous Card14^LSL-E138A, as judged by weight loss, hypothermia and extensive infiltration of S100A9⁺ myeloid cells into the lungs (Figure 7A–C). Hence, the rapid development of systemic disease in homozygous Card14^{LSL-E138A/LSL-E138A}Rosa26^CreERT2/+ mice following tamoxifen injection was not dependent on the adaptive immune system.

Anti-TNF ameliorates severe disease in homozygous Card14^{LSL-E138A/LSL-E138A}Rosa26^CreERT2/+ mice

Blocking TNF significantly reduced skin inflammation in heterozygous Card14^LSL-E138A/+Rosa26^CreERT2 mice following tamoxifen injection (Figure 3). Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice were treated with anti-TNF blocking antibody prior to and post tamoxifen induction to determine whether the severe systemic disease was also mediated by TNF. Weight loss was significantly reduced by anti-TNF treatment compared to isotype control, but not completely blocked (Figure 8A). However, anti-TNF prevented the hypothermic response of Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice to tamoxifen-induction (Figure 8B), which correlated with the mice appearing more active than isotype controls. Recruitment of S100A9⁺ myeloid cells into the kidneys, liver and lung at d5 post-tamoxifen was also significantly reduced by anti-TNF compared to isotype control (Figure 8C).

Figure 8

Download asset Open asset

Anti-TNF treatment significantly decreased ear mRNA expression of Ccl20, Cxcl2 and Cxcl3 (Figure 8D), the products of which are chemoattractants for neutrophils (Chiang et al., 2019). The reduction in expression of these genes by anti-TNF was even more pronounced in the lungs. mRNA expression of S100a9, a marker of myeloid cell infiltration, was also significantly reduced (Figure 8D).

Reactive oxygen species (ROS) generation, degranulation and formation of NETs (neutrophil extracellular traps) by neutrophils in the skin contribute to the inflammatory pathogenesis of psoriasis (Chiang et al., 2019). Anti-TNF reduction in S100A9⁺ myeloid cell infiltration into the liver, kidneys and lungs of Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice suggested that reduction in myeloid cell-mediated inflammation might explain in part the protective effect of anti-TNF treatment. To investigate this, Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice were administered anti-GR1 antibody, which depleted neutrophils and monocytes. Treatment with anti-GR1 before and during CARD14^E138A induction significantly reduced weight loss in the Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice, although not to the same extent as anti-TNF treatment (Figure 8A). However, at d5, anti-GR1 treated mice were completely protected from hypothermia (Figure 8B) and had significantly decreased S100A9⁺ myeloid cell infiltration into kidney, liver and lung (Figure 8C).

Biochemical analyses revealed that treating homozygous mice with either anti-TNF or anti-GR1, normalised ALT, GLDH and inorganic phosphate levels in the serum (Figure 8E). This suggested that these blocking antibody treatments protected homozygous mice from organ dysfunction.

Together these results indicated that in Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice, CARD14^E138A signalling in keratinocytes induced systemic release of TNF, which promoted recruitment of myeloid cells into vital organs via increased expression of chemokines. Myeloid cells in combination with upregulated expression of pro-inflammatory genes led to organ dysfunction and severe disease.

Gain-of-function CARD14 mutations are linked to the development of plaque psoriasis and generalised pustular psoriasis (GPP). CARD14 variants are also associated with pityriasis rubra pilaris, a rare papulosquamous disorder phenotypically related to psoriasis (Israel and Mellett, 2018). In the present study, a conditional knock-in Card14^E138A mouse strain was used to determine how the highly penetrant psoriasis-associated CARD14^E138A mutation induces inflammation in adult mice. The effects of CARD14^E138A signalling were found to be dose-dependent. Tamoxifen induction of CARD14^E138A in heterozygous Card14^LSL-E138A/+Rosa26^CreERT2 mice induced skin inflammation phenotypically similar to psoriasis, while homozygous Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice developed skin inflammation and a severe systemic disease. Both phenotypes were driven by CARD14^E138A signalling in keratinocytes and dependent on TNF.

Earlier immunohistochemical studies showed that CARD14 protein was expressed in the basal layer in human skin (Jordan et al., 2012b). In contrast, our analyses of mouse skin found that CARD14 was expressed at low levels in undifferentiated keratinocytes, and its expression increased with differentiation both at the RNA and protein level. These results suggest that the localisation of CARD14 expression differs between mouse and human skin. However, the present study also found that CARD14 protein expression increased in human keratinocytes differentiated in vitro. Furthermore, a similar trend was found in publicly available RNA sequencing datasets in which human keratinocytes were differentiated in vitro (GSE73305) and in organotypic epidermal tissue (GSE52954). It will be important to determine whether CARD14 protein levels similarly increase as keratinocytes differentiate in human skin. Nevertheless, regardless of apparent differences in CARD14 distribution between mouse and human, the predominant localisation of CARD14 in the skin is consistent with a function in sensing harmful stimuli, such as pathogens, at the interface between the organism and the environment.

Psoriasis-associated CARD14 mutations are not linked with inflammatory gut phenotypes, although CARD14 is also expressed at high levels in the gastro-intestinal tract in humans (Fuchs-Telem et al., 2012). RNAscope analyses showed that CARD14 is expressed in intestinal epithelial cells (IECs). Consistent with this, tamoxifen induction of CARD14^E138A in Card14^LSL-E138A/+Rosa26^CreERT2 mice promotes the expression of pro-inflammatory genes and recruitment of S100A9⁺ myeloid cells in the colon. However, CARD14^E138A expression in the colon was not sufficient to cause any gut tissue damage. These results are similar to the effects of expressing a constitutively active IKK2 mutant in IECs in IKK2[EE]^IEC mice, which also induces expression of pro-inflammatory cytokines in the gut and inflammatory cell infiltration without causing overt tissue damage (Guma et al., 2011). IKK2[EE]^IEC mice, however, develop destructive acute inflammation, with intestinal barrier disruption and bacterial translocation following dextran sulfate sodium (DSS) or lipopolysaccharide (LPS) administration (Guma et al., 2011; Vlantis et al., 2011). Thus, additional insults might be required for CARD14^E138A signalling to induce a gut pathology.

Psoriasis can present as a mild stable disease with sharply demarcated psoriatic plaques in the skin at the prototypical body sites (e.g. elbows, knees, scalp) that predominantly involves the adaptive immune system. Alternatively, psoriasis can be an active severe disease with extensive body surface coverage, involving elevated levels of cytokines and inflammatory markers in the blood and systemic inflammation (Christophers and van de Kerkhof, 2019; Schön, 2019). This is mainly driven by cells of the innate immune system (Liang et al., 2017a). Acute induction of CARD14^E138A in Card14^LSL-E138A/+Rosa26^CreERT2/+ mice induced widespread skin lesions in back skin and ears that lacked defined borders. This primarily involved infiltration of innate immune cells into the skin and did not require adaptive immune cells. Prolonged CARD14^E138A signalling in Card14^LSL-E138A/+Rosa26^CreERT2/+ mice promoted the development of clearly delineated plaques on the ears with reduced acanthosis in back skin, suggesting less severe skin disease overall. These changes correlated with decreased skin expression of pro-inflammatory genes, decreased serum levels of key pro-inflammatory cytokines and more extensive infiltration of T cells into the skin. Acute expression of CARD14^E138A in Card14^LSL-E138A/+Rosa26^CreERT2 and Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice, therefore, appeared to model aspects of active psoriasis. In contrast, chronic CARD14^E138A signalling in Card14^LSL-E138A/+Rosa26^CreERT2 mice resulted in a phenotype more similar to mild, stable disease in psoriatic patients. Further research using this model will help to elucidate what factors mediate the transition from acute to chronic lesion.

The fold changes of the majority of proinflammatory genes analyzed at d5 in Card14^LSL-E138A/+Krt14^CreERT2/+ with respect to controls were lower than the fold changes of the Card14^LSL-E138A/+Rosa26^CreERT2/+ mice. The Krt14^CreERT2 driver can induce ‘leaky’ expression of Cre (Vasioukhin et al., 1999) consistent with some Card14^E138A/+ Krt14^CreERT2/+ mice presenting with an inflammatory skin phenotype without tamoxifen injection. Although such mice were excluded from our analyses, it is possible that low concentrations of active Cre in uninduced mice resulted in the expression of limiting amounts of CARD14^E138A which chronically activated NF-κB without promoting overt skin inflammation. This may have reduced the fold changes after tamoxifen injection due to negative feedback regulation as seen at 1 m post-tamoxifen injection in Card14^LSL-E138A/+ Rosa26^CreERT2/+ mice. It is also possible that reduced tamoxifen-induction of skin gene expression in Card14^E138A/+ Krt14^CreERT2/+ mice was due to differences to different expression levels of Cre protein between the two Cre-drivers, which use distinct promoters. Regardless, the induced genes were the same in both Cre drivers and the phenotype of both het and hom mice in the Krt14^CreERT2/+ background was similar to the Rosa26^CreERT2/+ histologically and systemically demonstrating that CARD14^E138A signalling in keratinocytes alone is sufficient to induce skin and systemic inflammation.

Earlier studies have described the phenotypes of knock-in mice constitutively expressing activated CARD14 variants (Mellett et al., 2018; Sundberg et al., 2019; Wang et al., 2018). Heterozygous Card14^ΔE138/+, Card14^E138A/+ and Card14^ΔQ136/+ mice develop skin inflammation similar to heterozygous Card14^LSL-E138A/+Rosa26^CreERT2/+ mice following tamoxifen induction. In both Card14^ΔE138/+ and Card14^ΔQ136/+mice, blocking the IL-17A/IL-23 axis ameliorates skin inflammation. Anti-IL-17A also reduced the acanthosis that develops in the back skin of Card14^LSL-E138A/+Rosa26^CreERT2/+ mice 5d after tamoxifen induction, but did not reduce acanthosis in ears or profoundly alter expression of pro-inflammatory genes. Acute induction of skin inflammation by CARD14^E138A, therefore, is relatively independent of IL-17A and rather driven by overproduction of TNF by keratinocytes. Furthermore, inducible expression of CARD14^E138A promoted the acute and chronic development of skin inflammation independently of adaptive immune cells (and T cell production of IL-17A). It will be important to test whether anti-TNF treatment could prevent chronic skin inflammation induced by CARD14^E138A.

The absence of a requirement for adaptive immune cells for the development of skin inflammation in Card14^LSL-E138A/+Rosa26^CreERT2/+ mice contrasts with a published study by Wang et al. where the inflammatory skin phenotype in mice constitutively expressing CARD14^ΔQ136 was reduced by RAG1 deficiency. However, our data resemble other murine models of skin inflammation originating from constitutive signalling in keratinocytes where the absence of adaptive immune cells does not ameliorate skin disease and can even exacerbate it (Leite Dantas et al., 2016; Uluçkan et al., 2019). Accordingly, in the absence of Tregs, innate immune cells play a more significant role in the imiquimod-induced psoriasis-like skin pathology (Hartwig et al., 2018; Stockenhuber et al., 2018). Card14^ΔQ136/+ mice only displayed a histopathologic phenotype in tail skin, but not ear skin. It is possible that the adaptive immune system plays different roles in these different anatomical regions. Alternatively, the inflammatory skin phenotype that develops in Card14^ΔQ136/+ mice may involve developmental alterations that occur in utero that require adaptive immune cells.

Systemic disease was not described in mice constitutively expressing the CARD14^E138A variants (Mellett et al., 2018; Sundberg et al., 2019; Wang et al., 2018), with all analyses focusing only on inflammatory skin phenotypes . However, heterozygous Card14^E138A/+ mice have reduced survival after birth, while Card14^{ΔE138A/ΔE138A} mice die in utero or shortly following birth. It is possible that variant CARD14 signalling during early life induces a severe, systemic disease in these strains that decreases survival. In addition, in utero expression of variant CARD14 could mean that, by the time of analysis of surviving adult mice, the disease state may have already developed from severe/active to mild/stable. In line with this, the inflammatory back skin phenotype, which coincided with severe disease in Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice, was not evident in constitutive Card14^ΔQ136/+ mice, and only briefly detected (between 5 and 7 days of age) in constitutive Card14^ΔE138A/+ mice (Wang et al., 2018).

Our analyses of Card14^LSL-E138A/+Rosa26^CreERT2 mice and Card14 ^{LSL-E138A/ LSL-E138A} Rosa26^CreERT2/+ mice revealed a Card14^E138A gene dosage effect, with increased CARD14^E138A expression inducing a more severe acute phenotype. This presumably explains why CARD14 psoriasis-associated mutations are generally heterozygous. Weaker NF-κB activating variants, such as CARD14^G117S, have been found in their homozygous forms in children with severe psoriasis suggesting that the overall level of NF-κB activation is important (Craiglow et al., 2018; Israel and Mellett, 2018; Sugiura et al., 2015). Genetic background also contributes to the penetrance of gain-of-function CARD14 mutations in promoting psoriasis (Bhalerao and Bowcock, 1998), as recently confirmed in mouse studies in which the survival of mice with a constitutive Card14^E138A mutation is increased by crossing C57BL/6 mice onto a 129S4 background (Sundberg et al., 2019).

The severe, systemic disease that developed in homozygous Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice following tamoxifen induction had similarities with exacerbations in GPP patients. These included skin disease that covered the majority of the body surface and hypothermia, the mouse equivalent of human fever (Gordon, 2005). The extensive scaling of mutant mouse lips also appeared to resemble cheilitis, a common manifestation of GPP (Ly et al., 2019). Although pustules were not macroscopically visible (perhaps reflecting differences between human and mouse skin), histological analysis revealed extensive pustule formation within the epidermis and stratum corneum following CARD14^E138A expression. Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice also presented with myeloid cell infiltration in multiple vital organs and evidence of multi-organ dysfunction, similar to patients with GPP. The skin transcriptome of Card14^LSL-E138A/+Rosa26^CreERT2 mice also shared similarities with the skin transcriptome of the GPP patient expressing the CARD14^E138A variant. The increased expression of Steap1 and Steap4 mRNAs in the skin of Card14^LSL-E138A/+Rosa26^CreERT2 mice after tamoxifen injection is particularly interesting since these genes have been found to be upregulated in several types of pustular psoriasis and are known to be important in neutrophil chemotaxis (Liang et al., 2017b). To our knowledge, Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice are the first to mimic the phenotypes of GPP and might be useful in future studies to understand more about the etiopathogenesis of this severe form of psoriasis.

Both GR1 antibody (to deplete neutrophils and monocytes) and TNF antibody significantly reduced disease severity in Card14^{LSL-E138A/LSL-E138A} Rosa26^CreERT2/+ mice following tamoxifen administration. Current treatments for GPP patients include granulocyte/monocyte apheresis and anti-TNF (Fujisawa et al., 2015; Koike et al., 2017) suggesting that Card14^{LSL-E138A/LSL-E138A}Rosa26^CreERT2/+ mice could be used for pre-clinical testing of new medicines for GPP treatment. The present study indicates that constitutive CARD14^E138A activation of NF-κB in keratinocytes is sufficient to promote skin inflammation and systemic inflammatory disease. This raises the possibility that it may be feasible to develop topical medicines that block CARD14 signalling in the skin to treat CARD14-induced psoriasis and systemic inflammation.

The RNA-Seq data generated in this article was deposited in the GEO repository (GSE149880).

1. 1. Manils J
   2. Webb L
   3. Boeing S

   (2020) NCBI Gene Expression Omnibus

   ID GSE149880. Expression analysis of WT or Card14E138A ears 5 days and 1 month after injecton of tamoxifen.

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE149880

1. 1. Asare A
   2. Levorse J
   3. Fuchs E

   (2017) NCBI Gene Expression Omnibus

   ID GSE75931. A spatio-temporal characterization of the transcriptional landscape of epidermal development.

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE75931
2. 1. Bin L
   2. Deng L
   3. Yang H
   4. Zhu L
   5. Wang X
   6. Edwards MG
   7. Richers B
   8. Leung DYM

   (2016) NCBI Gene Expression Omnibus

   ID GSE73305. RNA-sequencing transcriptome profiling of normal human keratinocytes differentiation.

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE73305
3. 1. Vanessa L-P
   2. Kun Q
   3. Jiajing Z
   4. Dan EW
   5. Brook CB
   6. Zurab S
   7. Brian JZ
   8. Lisa DB
   9. Rios EJ
   10. Shiying T
   11. Markus K
   12. Paul AK

   (2014) NCBI Gene Expression Omnibus

   ID GSE52954. A LncRNA-MAF/MAFB transcription factor network regulates epidermal differentiation.

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE52954

1. 1. Asare A
   2. Levorse J
   3. Fuchs E

   (2017) Coupling organelle inheritance with mitosis to balance growth and differentiation

   Science 355:eaah4701.

   https://doi.org/10.1126/science.aah4701

   • PubMed
   • Google Scholar
2. 1. Bachelez H

   (2020) Pustular psoriasis: the dawn of a new era

   Acta Dermato Venereologica 100:87–93.

   https://doi.org/10.2340/00015555-3388

   • Google Scholar
3. 1. Bhalerao J
   2. Bowcock AM

   (1998) The genetics of psoriasis: a complex disorder of the skin and immune system

   Human Molecular Genetics 7:1537–1545.

   https://doi.org/10.1093/hmg/7.10.1537

   • PubMed
   • Google Scholar
4. 1. Bikle DD
   2. Xie Z
   3. Tu CL

   (2012) Calcium regulation of keratinocyte differentiation

   Expert Review of Endocrinology & Metabolism 7:461–472.

   https://doi.org/10.1586/eem.12.34

   • PubMed
   • Google Scholar
5. 1. Bin L
   2. Deng L
   3. Yang H
   4. Zhu L
   5. Wang X
   6. Edwards MG
   7. Richers B
   8. Leung DY

   (2016) Forkhead box C1 regulates human primary keratinocyte terminal differentiation

   PLOS ONE 11:e0167392.

   https://doi.org/10.1371/journal.pone.0167392

   • PubMed
   • Google Scholar
6. 1. Breschi A
   2. Gingeras TR
   3. Guigó R

   (2017) Comparative transcriptomics in human and mouse

   Nature Reviews Genetics 18:425–440.

   https://doi.org/10.1038/nrg.2017.19

   • PubMed
   • Google Scholar
7. 1. Chiang CC
   2. Cheng WJ
   3. Korinek M
   4. Lin CY
   5. Hwang TL

   (2019) Neutrophils in psoriasis

   Frontiers in Immunology 10:2376.

   https://doi.org/10.3389/fimmu.2019.02376

   • PubMed
   • Google Scholar
8. 1. Christophers E
   2. van de Kerkhof PCM

   (2019) Severity, heterogeneity and systemic inflammation in psoriasis

   Journal of the European Academy of Dermatology and Venereology 33:643–647.

   https://doi.org/10.1111/jdv.15339

   • PubMed
   • Google Scholar
9. 1. Craiglow BG
   2. Boyden LM
   3. Hu R
   4. Virtanen M
   5. Su J
   6. Rodriguez G
   7. McCarthy C
   8. Luna P
   9. Larralde M
   10. Humphrey S
   11. Holland KE
   12. Hogeling M
   13. Hidalgo-Matlock B
   14. Ferrari B
   15. Fernandez-Faith E
   16. Drolet B
   17. Cordoro KM
   18. Bowcock AM
   19. Antaya RJ
   20. Ashack K
   21. Ashack RJ
   22. Lifton RP
   23. Milstone LM
   24. Paller AS
   25. Choate KA

   (2018) CARD14-associated papulosquamous eruption: a spectrum including features of psoriasis and Pityriasis rubra pilaris

   Journal of the American Academy of Dermatology 79:487–494.

   https://doi.org/10.1016/j.jaad.2018.02.034

   • PubMed
   • Google Scholar
10. 1. D'Erme AM
    2. Wilsmann-Theis D
    3. Wagenpfeil J
    4. Hölzel M
    5. Ferring-Schmitt S
    6. Sternberg S
    7. Wittmann M
    8. Peters B
    9. Bosio A
    10. Bieber T
    11. Wenzel J

    (2015) IL-36γ (IL-1F9) is a biomarker for psoriasis skin lesions

    Journal of Investigative Dermatology 135:1025–1032.

    https://doi.org/10.1038/jid.2014.532

    • PubMed
    • Google Scholar
11. 1. Dainichi T
    2. Kitoh A
    3. Otsuka A
    4. Nakajima S
    5. Nomura T
    6. Kaplan DH
    7. Kabashima K

    (2018) The epithelial immune microenvironment (EIME) in atopic dermatitis and psoriasis

    Nature Immunology 19:1286–1298.

    https://doi.org/10.1038/s41590-018-0256-2

    • PubMed
    • Google Scholar
12. 1. DeLuca DS
    2. Levin JZ
    3. Sivachenko A
    4. Fennell T
    5. Nazaire M-D
    6. Williams C
    7. Reich M
    8. Winckler W
    9. Getz G

    (2012) RNA-SeQC: RNA-seq metrics for quality control and process optimization

    Bioinformatics 28:1530–1532.

    https://doi.org/10.1093/bioinformatics/bts196

    • Google Scholar
13. 1. Dobin A
    2. Davis CA
    3. Schlesinger F
    4. Drenkow J
    5. Zaleski C
    6. Jha S
    7. Batut P
    8. Chaisson M
    9. Gingeras TR

    (2013) STAR: ultrafast universal RNA-seq aligner

    Bioinformatics 29:15–21.

    https://doi.org/10.1093/bioinformatics/bts635

    • Google Scholar
14. 1. Eckert RL
    2. Broome AM
    3. Ruse M
    4. Robinson N
    5. Ryan D
    6. Lee K

    (2004) S100 proteins in the epidermis

    Journal of Investigative Dermatology 123:23–33.

    https://doi.org/10.1111/j.0022-202X.2004.22719.x

    • PubMed
    • Google Scholar
15. 1. el Marjou F
    2. Janssen KP
    3. Chang BH
    4. Li M
    5. Hindie V
    6. Chan L
    7. Louvard D
    8. Chambon P
    9. Metzger D
    10. Robine S

    (2004) Tissue-specific and inducible Cre-mediated recombination in the gut epithelium

    Genesis 39:186–193.

    https://doi.org/10.1002/gene.20042

    • PubMed
    • Google Scholar
16. 1. Fuchs-Telem D
    2. Sarig O
    3. van Steensel MA
    4. Isakov O
    5. Israeli S
    6. Nousbeck J
    7. Richard K
    8. Winnepenninckx V
    9. Vernooij M
    10. Shomron N
    11. Uitto J
    12. Fleckman P
    13. Richard G
    14. Sprecher E

    (2012) Familial Pityriasis rubra pilaris is caused by mutations in CARD14

    The American Journal of Human Genetics 91:163–170.

    https://doi.org/10.1016/j.ajhg.2012.05.010

    • PubMed
    • Google Scholar
17. 1. Fujisawa T
    2. Suzuki S
    3. Mizutani Y
    4. Doi T
    5. Yoshida S
    6. Ogura S
    7. Seishima M

    (2015) Granulocyte and monocyte adsorption apheresis for generalized pustular psoriasis: therapeutic outcomes in three refractory patients

    Therapeutic Apheresis and Dialysis 19:336–341.

    https://doi.org/10.1111/1744-9987.12342

    • PubMed
    • Google Scholar
18. 1. Gauld SB
    2. Gauvin D
    3. Olson L
    4. Leys L
    5. Paulsboe S
    6. Liu Z
    7. Edelmayer RM
    8. Wetter J
    9. Salte K
    10. Wang Y
    11. Huang S
    12. Honore P
    13. McGaraughty S

    (2018) Mechanistic and pharmacological assessment of murine IL-23 mediated psoriasiform dermatitis; implications for drug discovery

    Journal of Dermatological Science 92:45–53.

    https://doi.org/10.1016/j.jdermsci.2018.08.001

    • PubMed
    • Google Scholar
19. 1. Gooderham MJ
    2. Van Voorhees AS
    3. Lebwohl MG

    (2019) An update on generalized pustular psoriasis

    Expert Review of Clinical Immunology 15:907–919.

    https://doi.org/10.1080/1744666X.2019.1648209

    • PubMed
    • Google Scholar
20. Book

    1. Gordon CJ

    (2005)

    Temperature and Toxicology: An Integrative, Comparative, and Environmental Approach

    Taylor & Francis, Boca Raton, FL. xvii.

    • Google Scholar
21. 1. Greb JE
    2. Goldminz AM
    3. Elder JT
    4. Lebwohl MG
    5. Gladman DD
    6. Wu JJ
    7. Mehta NN
    8. Finlay AY
    9. Gottlieb AB

    (2016) Psoriasis

    Nature Reviews Disease Primers 2:16082.

    https://doi.org/10.1038/nrdp.2016.82

    • Google Scholar
22. 1. Guma M
    2. Stepniak D
    3. Shaked H
    4. Spehlmann ME
    5. Shenouda S
    6. Cheroutre H
    7. Vicente-Suarez I
    8. Eckmann L
    9. Kagnoff MF
    10. Karin M

    (2011) Constitutive intestinal NF-κB does not trigger destructive inflammation unless accompanied by MAPK activation

    The Journal of Experimental Medicine 208:1889–1900.

    https://doi.org/10.1084/jem.20110242

    • PubMed
    • Google Scholar
23. 1. Guttman-Yassky E
    2. Krueger JG

    (2017) Atopic dermatitis and psoriasis: two different immune diseases or one spectrum?

    Current Opinion in Immunology 48:68–73.

    https://doi.org/10.1016/j.coi.2017.08.008

    • PubMed
    • Google Scholar
24. 1. Hartwig T
    2. Zwicky P
    3. Schreiner B
    4. Yawalkar N
    5. Cheng P
    6. Navarini A
    7. Dummer R
    8. Flatz L
    9. Conrad C
    10. Schlapbach C
    11. Becher B

    (2018) Regulatory T cells restrain pathogenic T helper cells during skin inflammation

    Cell Reports 25:3564–3572.

    https://doi.org/10.1016/j.celrep.2018.12.012

    • PubMed
    • Google Scholar
25. 1. Hong KU
    2. Reynolds SD
    3. Watkins S
    4. Fuchs E
    5. Stripp BR

    (2004) Basal cells are a multipotent progenitor capable of renewing the bronchial epithelium

    The American Journal of Pathology 164:577–588.

    https://doi.org/10.1016/S0002-9440(10)63147-1

    • PubMed
    • Google Scholar
26. 1. Howes A
    2. O'Sullivan PA
    3. Breyer F
    4. Ghose A
    5. Cao L
    6. Krappmann D
    7. Bowcock AM
    8. Ley SC

    (2016) Psoriasis mutations disrupt CARD14 autoinhibition promoting BCL10-MALT1-dependent NF-κB activation

    Biochemical Journal 473:1759–1768.

    https://doi.org/10.1042/BCJ20160270

    • PubMed
    • Google Scholar
27. 1. Israel L
    2. Mellett M

    (2018) Clinical and genetic heterogeneity of CARD14 Mutations in Psoriatic Skin Disease

    Frontiers in Immunology 9:2239.

    https://doi.org/10.3389/fimmu.2018.02239

    • PubMed
    • Google Scholar
28. 1. Johnston A
    2. Xing X
    3. Wolterink L
    4. Barnes DH
    5. Yin Z
    6. Reingold L
    7. Kahlenberg JM
    8. Harms PW
    9. Gudjonsson JE

    (2017) IL-1 and IL-36 are dominant cytokines in generalized pustular psoriasis

    Journal of Allergy and Clinical Immunology 140:109–120.

    https://doi.org/10.1016/j.jaci.2016.08.056

    • PubMed
    • Google Scholar
29. 1. Jordan CT
    2. Cao L
    3. Roberson ED
    4. Duan S
    5. Helms CA
    6. Nair RP
    7. Duffin KC
    8. Stuart PE
    9. Goldgar D
    10. Hayashi G
    11. Olfson EH
    12. Feng BJ
    13. Pullinger CR
    14. Kane JP
    15. Wise CA
    16. Goldbach-Mansky R
    17. Lowes MA
    18. Peddle L
    19. Chandran V
    20. Liao W
    21. Rahman P
    22. Krueger GG
    23. Gladman D
    24. Elder JT
    25. Menter A
    26. Bowcock AM

    (2012a) Rare and common variants in CARD14, encoding an epidermal regulator of NF-kappaB, in psoriasis

    The American Journal of Human Genetics 90:796–808.

    https://doi.org/10.1016/j.ajhg.2012.03.013

    • PubMed
    • Google Scholar
30. 1. Jordan CT
    2. Cao L
    3. Roberson ED
    4. Pierson KC
    5. Yang CF
    6. Joyce CE
    7. Ryan C
    8. Duan S
    9. Helms CA
    10. Liu Y
    11. Chen Y
    12. McBride AA
    13. Hwu WL
    14. Wu JY
    15. Chen YT
    16. Menter A
    17. Goldbach-Mansky R
    18. Lowes MA
    19. Bowcock AM

    (2012b) PSORS2 is due to mutations in CARD14

    The American Journal of Human Genetics 90:784–795.

    https://doi.org/10.1016/j.ajhg.2012.03.012

    • PubMed
    • Google Scholar
31. 1. Koike Y
    2. Okubo M
    3. Kiyohara T
    4. Fukuchi R
    5. Sato Y
    6. Kuwatsuka S
    7. Takeichi T
    8. Akiyama M
    9. Sugiura K
    10. Utani A

    (2017) Granulocyte and monocyte apheresis can control juvenile generalized pustular psoriasis with mutation of IL36RN

    British Journal of Dermatology 177:1732–1736.

    https://doi.org/10.1111/bjd.15509

    • PubMed
    • Google Scholar
32. 1. Leite Dantas R
    2. Masemann D
    3. Schied T
    4. Bergmeier V
    5. Vogl T
    6. Loser K
    7. Brachvogel B
    8. Varga G
    9. Ludwig S
    10. Wixler V

    (2016) Macrophage-mediated psoriasis can be suppressed by regulatory T lymphocytes

    The Journal of Pathology 240:366–377.

    https://doi.org/10.1002/path.4786

    • PubMed
    • Google Scholar
33. 1. Lessard JC
    2. Piña-Paz S
    3. Rotty JD
    4. Hickerson RP
    5. Kaspar RL
    6. Balmain A
    7. Coulombe PA

    (2013) Keratin 16 regulates innate immunity in response to epidermal barrier breach

    PNAS 110:19537–19542.

    https://doi.org/10.1073/pnas.1309576110

    • PubMed
    • Google Scholar
34. 1. Li B
    2. Dewey CN

    (2011) RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome

    BMC Bioinformatics 12:323.

    https://doi.org/10.1186/1471-2105-12-323

    • PubMed
    • Google Scholar
35. 1. Liang Y
    2. Sarkar MK
    3. Tsoi LC
    4. Gudjonsson JE

    (2017a) Psoriasis: a mixed autoimmune and autoinflammatory disease

    Current Opinion in Immunology 49:1–8.

    https://doi.org/10.1016/j.coi.2017.07.007

    • PubMed
    • Google Scholar
36. 1. Liang Y
    2. Xing X
    3. Beamer MA
    4. Swindell WR
    5. Sarkar MK
    6. Roberts LW
    7. Voorhees JJ
    8. Kahlenberg JM
    9. Harms PW
    10. Johnston A
    11. Gudjonsson JE

    (2017b) Six-transmembrane epithelial antigens of the prostate comprise a novel inflammatory nexus in patients with pustular skin disorders

    Journal of Allergy and Clinical Immunology 139:1217–1227.

    https://doi.org/10.1016/j.jaci.2016.10.021

    • PubMed
    • Google Scholar
37. 1. Lopez-Pajares V
    2. Qu K
    3. Zhang J
    4. Webster DE
    5. Barajas BC
    6. Siprashvili Z
    7. Zarnegar BJ
    8. Boxer LD
    9. Rios EJ
    10. Tao S
    11. Kretz M
    12. Khavari PA

    (2015) A LncRNA-MAF:mafb transcription factor network regulates epidermal differentiation

    Developmental Cell 32:693–706.

    https://doi.org/10.1016/j.devcel.2015.01.028

    • PubMed
    • Google Scholar
38. 1. Love MI
    2. Huber W
    3. Anders S

    (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2

    Genome Biology 15:550.

    https://doi.org/10.1186/s13059-014-0550-8

    • PubMed
    • Google Scholar
39. 1. Lu HY
    2. Biggs CM
    3. Blanchard-Rohner G
    4. Fung SY
    5. Sharma M
    6. Turvey SE

    (2019) Germline CBM-opathies: from immunodeficiency to atopy

    The Journal of Allergy and Clinical Immunology 143:1661–1673.

    https://doi.org/10.1016/j.jaci.2019.03.009

    • PubMed
    • Google Scholar
40. 1. Ly K
    2. Beck KM
    3. Smith MP
    4. Thibodeaux Q
    5. Bhutani T

    (2019) Diagnosis and screening of patients with generalized pustular psoriasis

    Psoriasis 9:37–42.

    https://doi.org/10.2147/PTT.S181808

    • PubMed
    • Google Scholar
41. 1. Mellett M
    2. Meier B
    3. Mohanan D
    4. Schairer R
    5. Cheng P
    6. Satoh TK
    7. Kiefer B
    8. Ospelt C
    9. Nobbe S
    10. Thome M
    11. Contassot E
    12. French LE

    (2018) CARD14 Gain-of-Function mutation alone is sufficient to drive IL-23/IL-17-Mediated psoriasiform skin inflammation in Vivo

    Journal of Investigative Dermatology 138:2010–2023.

    https://doi.org/10.1016/j.jid.2018.03.1525

    • PubMed
    • Google Scholar
42. 1. Nakashima T
    2. Jinnin M
    3. Yamane K
    4. Honda N
    5. Kajihara I
    6. Makino T
    7. Masuguchi S
    8. Fukushima S
    9. Okamoto Y
    10. Hasegawa M
    11. Fujimoto M
    12. Ihn H

    (2012) Impaired IL-17 signaling pathway contributes to the increased collagen expression in scleroderma fibroblasts

    The Journal of Immunology 188:3573–3583.

    https://doi.org/10.4049/jimmunol.1100591

    • Google Scholar
43. 1. Peled A
    2. Sarig O
    3. Sun G
    4. Samuelov L
    5. Ma CA
    6. Zhang Y
    7. Dimaggio T
    8. Nelson CG
    9. Stone KD
    10. Freeman AF
    11. Malki L
    12. Vidal LS
    13. Chamarthy LM
    14. Briskin V
    15. Mohamad J
    16. Pavlovsky M
    17. Walter JE
    18. Milner JD
    19. Sprecher E

    (2019) Loss-of-function mutations in caspase recruitment domain-containing protein 14 (CARD14) are associated with a severe variant of atopic dermatitis

    Journal of Allergy and Clinical Immunology 143:173–181.

    https://doi.org/10.1016/j.jaci.2018.09.002

    • Google Scholar
44. 1. Prinz I
    2. Sandrock I
    3. Mrowietz U

    (2020) Interleukin-17 cytokines: effectors and targets in psoriasis-A breakthrough in understanding and treatment

    Journal of Experimental Medicine 217:e20191397.

    https://doi.org/10.1084/jem.20191397

    • PubMed
    • Google Scholar
45. 1. Schön MP

    (2019) Adaptive and innate immunity in psoriasis and other inflammatory disorders

    Frontiers in Immunology 10:1764.

    https://doi.org/10.3389/fimmu.2019.01764

    • PubMed
    • Google Scholar
46. 1. Seibler J
    2. Zevnik B
    3. Küter-Luks B
    4. Andreas S
    5. Kern H
    6. Hennek T
    7. Rode A
    8. Heimann C
    9. Faust N
    10. Kauselmann G
    11. Schoor M
    12. Jaenisch R
    13. Rajewsky K
    14. Kühn R
    15. Schwenk F

    (2003) Rapid generation of inducible mouse mutants

    Nucleic Acids Research 31:e12.

    https://doi.org/10.1093/nar/gng012

    • PubMed
    • Google Scholar
47. 1. Spanopoulou E
    2. Roman CA
    3. Corcoran LM
    4. Schlissel MS
    5. Silver DP
    6. Nemazee D
    7. Nussenzweig MC
    8. Shinton SA
    9. Hardy RR
    10. Baltimore D

    (1994) Functional immunoglobulin transgenes guide ordered B-cell differentiation in Rag-1-deficient mice

    Genes & Development 8:1030–1042.

    https://doi.org/10.1101/gad.8.9.1030

    • PubMed
    • Google Scholar
48. 1. Stockenhuber K
    2. Hegazy AN
    3. West NR
    4. Ilott NE
    5. Stockenhuber A
    6. Bullers SJ
    7. Thornton EE
    8. Arnold IC
    9. Tucci A
    10. Waldmann H
    11. Ogg GS
    12. Powrie F

    (2018) Foxp3⁺ T reg cells control psoriasiform inflammation by restraining an IFN-I-driven CD8⁺ T cell response

    Journal of Experimental Medicine 215:1987–1998.

    https://doi.org/10.1084/jem.20172094

    • PubMed
    • Google Scholar
49. 1. Suárez-Fariñas M
    2. Arbeit R
    3. Jiang W
    4. Ortenzio FS
    5. Sullivan T
    6. Krueger JG

    (2013) Suppression of molecular inflammatory pathways by Toll-like receptor 7, 8, and 9 antagonists in a model of IL-23-induced skin inflammation

    PLOS ONE 8:e84634.

    https://doi.org/10.1371/journal.pone.0084634

    • PubMed
    • Google Scholar
50. 1. Subramanian A
    2. Tamayo P
    3. Mootha VK
    4. Mukherjee S
    5. Ebert BL
    6. Gillette MA
    7. Paulovich A
    8. Pomeroy SL
    9. Golub TR
    10. Lander ES
    11. Mesirov JP

    (2005) Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles

    PNAS 102:15545–15550.

    https://doi.org/10.1073/pnas.0506580102

    • PubMed
    • Google Scholar
51. 1. Sugiura K
    2. Kitoh T
    3. Watanabe D
    4. Muto M
    5. Akiyama M

    (2015) Childhood-onset PsA in down syndrome with psoriasis susceptibility variant CARD14 rs11652075

    Rheumatology 54:197–199.

    https://doi.org/10.1093/rheumatology/keu419

    • PubMed
    • Google Scholar
52. 1. Sundberg JP
    2. Pratt CH
    3. Silva KA
    4. Kennedy VE
    5. Qin W
    6. Stearns TM
    7. Frost J
    8. Sundberg BA
    9. Bowcock AM

    (2019) Gain of function p.e138a alteration in Card14 leads to psoriasiform skin inflammation and implicates genetic modifiers in disease severity

    Experimental and Molecular Pathology 110:104286.

    https://doi.org/10.1016/j.yexmp.2019.104286

    • PubMed
    • Google Scholar
53. 1. Swindell WR
    2. Sarkar MK
    3. Liang Y
    4. Xing X
    5. Gudjonsson JE

    (2016) Cross-Disease transcriptomics: unique IL-17A signaling in psoriasis lesions and an autoimmune PBMC signature

    Journal of Investigative Dermatology 136:1820–1830.

    https://doi.org/10.1016/j.jid.2016.04.035

    • PubMed
    • Google Scholar
54. 1. Tsoi LC
    2. Rodriguez E
    3. Degenhardt F
    4. Baurecht H
    5. Wehkamp U
    6. Volks N
    7. Szymczak S
    8. Swindell WR
    9. Sarkar MK
    10. Raja K
    11. Shao S
    12. Patrick M
    13. Gao Y
    14. Uppala R
    15. Perez White BE
    16. Getsios S
    17. Harms PW
    18. Maverakis E
    19. Elder JT
    20. Franke A
    21. Gudjonsson JE
    22. Weidinger S

    (2019) Atopic dermatitis is an IL-13-Dominant disease with greater molecular heterogeneity compared to psoriasis

    Journal of Investigative Dermatology 139:1480–1489.

    https://doi.org/10.1016/j.jid.2018.12.018

    • PubMed
    • Google Scholar
55. 1. Uluçkan Özge
    2. Jiménez M
    3. Roediger B
    4. Schnabl J
    5. Díez-Córdova LT
    6. Troulé K
    7. Weninger W
    8. Wagner EF

    (2019) Cutaneous immune Cell-Microbiota interactions are controlled by epidermal JunB/AP-1

    Cell Reports 29:843–859.

    https://doi.org/10.1016/j.celrep.2019.09.042

    • Google Scholar
56. 1. Vasioukhin V
    2. Degenstein L
    3. Wise B
    4. Fuchs E

    (1999) The magical touch: genome targeting in epidermal stem cells induced by tamoxifen application to mouse skin

    PNAS 96:8551–8556.

    https://doi.org/10.1073/pnas.96.15.8551

    • PubMed
    • Google Scholar
57. 1. Vervloet MG
    2. Sezer S
    3. Massy ZA
    4. Johansson L
    5. Cozzolino M
    6. Fouque D
    7. ERA–EDTA Working Group on Chronic Kidney Disease–Mineral and Bone Disorders and the European Renal Nutrition Working Group

    (2017) The role of phosphate in kidney disease

    Nature Reviews Nephrology 13:27–38.

    https://doi.org/10.1038/nrneph.2016.164

    • PubMed
    • Google Scholar
58. 1. Vlantis K
    2. Wullaert A
    3. Sasaki Y
    4. Schmidt-Supprian M
    5. Rajewsky K
    6. Roskams T
    7. Pasparakis M

    (2011) Constitutive IKK2 activation in intestinal epithelial cells induces intestinal tumors in mice

    Journal of Clinical Investigation 121:2781–2793.

    https://doi.org/10.1172/JCI45349

    • PubMed
    • Google Scholar
59. 1. Wang M
    2. Zhang S
    3. Zheng G
    4. Huang J
    5. Songyang Z
    6. Zhao X
    7. Lin X

    (2018) Gain-of-Function mutation of Card14 leads to spontaneous Psoriasis-like skin inflammation through enhanced keratinocyte response to IL-17A

    Immunity 49:66–79.

    https://doi.org/10.1016/j.immuni.2018.05.012

    • PubMed
    • Google Scholar
60. 1. Zhang Q
    2. Lenardo MJ
    3. Baltimore D

    (2017) 30 years of NF-κB: a blossoming of relevance to human pathobiology

    Cell 168:37–57.

    https://doi.org/10.1016/j.cell.2016.12.012

    • PubMed
    • Google Scholar

Author details

1. Joan Manils

   1. The Francis Crick Institute, London, United Kingdom
   2. Department of Immunology & Inflammation, Imperial College London, London, United Kingdom

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Writing - original draft, Writing - review and editing

   Contributed equally with

   Louise V Webb

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-8429-1295

2. Louise V Webb

   The Francis Crick Institute, London, United Kingdom

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Writing - original draft, Writing - review and editing

   Contributed equally with

   Joan Manils

   Competing interests

   No competing interests declared

3. Ashleigh Howes

   National Heart & Lung Institute, Imperial College London, London, United Kingdom

   Contribution

   Conceptualization, Investigation

   Competing interests

   No competing interests declared

4. Julia Janzen

   Department of Immunology & Inflammation, Imperial College London, London, United Kingdom

   Contribution

   Resources, Investigation, Project administration

   Competing interests

   No competing interests declared

5. Stefan Boeing

   1. The Francis Crick Institute, London, United Kingdom
   2. Bioinformatics and Biostatistics, The Francis Crick Institute, London, United Kingdom
   3. Crick Scientific Computing - Digital Development Team, The Francis Crick Institute, London, United Kingdom

   Contribution

   Data curation, Software, Formal analysis

   Competing interests

   No competing interests declared

6. Anne M Bowcock

   1. National Heart & Lung Institute, Imperial College London, London, United Kingdom
   2. Departments of Oncological Science, Dermatology, and Genetics & Genome Sciences, Icahn School of Medicine at Mount Sinai, New York, United States

   Contribution

   Conceptualization, Funding acquisition

   For correspondence

   anne.bowcock@mssm.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-8691-9090

7. Steven C Ley

   Department of Immunology & Inflammation, Imperial College London, London, United Kingdom

   Contribution

   Conceptualization, Formal analysis, Supervision, Funding acquisition, Investigation, Writing - original draft, Writing - review and editing

   For correspondence

   sley@imperial.ac.uk

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-5911-9223

• 2,307

  views

• 311

  downloads

• 23

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.