Human Molecular Genetics, 2015, Vol. 24, No. 2 doi:10.1093/hmg/ddu459 Advance Access published on September 10, 2014

436–449

Fetal inhibition of inflammation improves disease phenotypes in harlequin ichthyosis Denny L. Cottle1,{, Gloria M. A. Ursino1,{, Sally Chi Ieng Ip1, Lynelle K. Jones1, Tia Ditommaso1, Douglas F. Hacking4,9, Niamh E. Mangan5, Natalie A. Mellett7, Katya J. Henley3, Dmitri Sviridov7, Claudia A. Nold-Petry6, Marcel F. Nold6, Peter J. Meikle7, Benjamin T. Kile3,8 and Ian M. Smyth1,2,∗ 1

Received April 15, 2014; Revised July 14, 2014; Accepted September 2, 2014

Harlequin ichthyosis (HI) is a severe skin disease which leads to neonatal death in ∼50% of cases. It is the result of mutations in ABCA12, a protein that transports lipids required to establish the protective skin barrier needed after birth. To better understand the life-threatening newborn HI phenotype, we analysed the developing epidermis for consequences of lipid dysregulation in mouse models. We observed a pro-inflammatory signature which was characterized by chemokine upregulation in embryonic skin which is distinct from that seen in other types of ichthyosis. Inflammation also persisted in grafted HI skin. To examine the contribution of inflammation to disease development, we overexpressed interleukin-37b to globally suppress fetal inflammation, observing considerable improvements in keratinocyte differentiation. These studies highlight inflammation as an unexpected contributor to HI disease development in utero, and suggest that inhibiting inflammation may reduce disease severity.

INTRODUCTION The ichthyoses are a family of .20 congenital diseases characterized by the development of a thick hyperkeratotic epidermis (1). The most severe form of this disease spectrum is Harlequin ichthyosis (HI) (OMIM #242500) which is caused by mutations in ABCA12 (2,3), a putative lipid transport protein of the ATPbinding cassette (ABC) family. Gene targeting and ENU mutagenesis approaches have generated three different mouse models of the disease. These have helped to establish that loss of ABCA12 causes defective extracellular lipid trafficking and profound intracellular lipid accumulation within keratinocytes (4 – 6). In particular lamellar bodies, the small specialized intracellular organelles responsible for lipid transport in keratinocytes, lack lipid cargos in HI epidermis. Thus, the characteristic extracellular lipid lamellae which surround the cornified cells

upon terminal differentiation are variably disrupted or absent, leading to a profound defect in the barrier function of the skin. Loss of cutaneous barrier function presents particular challenges to the neonate, particularly through dehydration and infection. As a result, the survival of newborn HI individuals is only 50%, although their disease upon delivery is already extreme (7). For those patients who do survive beyond birth, a modest improvement in disease phenotypes is observed, although a lifetime regime of frequent bathing, removal of scales and frequent application of emollient oils is required to manage the disorder (8). In mouse models of HI neonatal mortality is fully penetrant, but grafted fetal skin exhibits an analogous self-improvement which has been attributed to better keratinocyte differentiation (9). Retinoid therapy is the main treatment for HI neonates (7), as they typically promote keratinocyte differentiation and shedding. However, there are a number of undesirable side-effects that limit

∗

To whom correspondence should be addressed. Tel: +61 399029119; Fax: +61 399052500; Email: [email protected] The authors wish it to be known that, in their opinion, these first authors should be regarded as joint First Authors.

†

# The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Department of Biochemistry and Molecular Biology, 2Department of Anatomy and Developmental Biology, Monash University, Wellington Road, Clayton, VIC 3800, Australia, 3Walter and Eliza Hall Institute, 1G Royal Parade, Parkville, VIC 3052, Australia, 4Department of Anaesthetics, Saint Vincent’s Hospital Melbourne, 41 Victoria Parade, Fitzroy, VIC 3065, Australia, 5Centre for Innate Immunity and Infectious Diseases, 6The Ritchie Centre, MIMR-PHI Institute of Medical Research, 27-31 Wright Street, Clayton, VIC 3168, Australia, 7Baker IDI Heart and Diabetes Institute, 75 Commercial Road, Melbourne, VIC 3004, Australia, 8Department of Medical Biology, University of Melbourne, Parkville, VIC 3010, Australia and 9Department of Paediatric Intensive Care, The Royal Children’s Hospital, Melbourne, VIC, Australia

Human Molecular Genetics, 2015, Vol. 24, No. 2

RESULTS Microarray analysis of Abca12lx12/lx12 embryonic epidermis We first profiled the temporal development of keratinocyte differentiation defects in the embryonic skin of the Abca12lx12/lx12 knockout mouse and found epidermal thickening and hyperkeratosis associated with the induction of wounding and hyperproliferative keratins such as Keratin 6 (K6) from E17.5 (Fig. 1A and B). To profile the global gene expression changes associated with these late embryonic phenotypes, microarray analysis was conducted on E17.5 embryonic epidermis (but not dermis) collected from the dorsal back skin of Abca12lx12/lx12 and littermate controls. Although previous studies have examined gene expression in HI keratinocytes (9), these were only performed on cultured cells. In Abca12lx12/lx12 embryos, we found 216 differentially regulated genes (from n ¼ 178 upregulated and n ¼ 70 downregulated probes; changed by .1.6-fold of P , 0.05 statistical significance). One class of upregulated genes were

those typically associated with hyperproliferation and wounding (Fig. 1C). These included the epithelial mitogen Epigen (Epgn), an EGF family protein, which promotes proliferation in keratinocytes (14), and the wound healing and hyperproliferative Keratin 6 and 216 (15). Consistent with the latter genes, we observed increased expression of Repetin, a filaggrin-like intermediate filament-associated protein which is believed to function in the aggregation of K16 filaments (16). We also observed upregulation of several small proline rich proteins which act as precursor proteins in the formation of the cornified envelope (17) and as reactive oxygen species scavengers in the skin (18) (Fig. 1C). These altered gene profiles highlight hyperkeratotic phenotypes but we also observed a second cohort of gene expression changes which suggest that the program of keratinocyte differentiation is impaired, particularly with respect to the process of terminal differentiation. These changes include downregulation of members of the late cornified envelope (Lce) gene family (19), extracellular matrix 1 (Ecm1) (which controls terminal cell differentiation) (20) and alcohol dehydrogenase 1 (Adh1) (linked to retinoic acid metabolism which strongly promotes keratinocyte differentiation) (21,22) (Fig. 1D). Taken together, these observations support the hypothesis that terminal differentiation of keratinocytes in the developing HI epidermis is defective. We subjected our list of differentially expressed transcripts to gene ontology (GO) analysis using GOStat (23) (Supplementary Material, Fig. S1A). As expected, genes encompassed by the terms GO:0031424 (keratinization) and GO:0001533 (cornified envelope) were enriched (P , 0.009 and P , 0.003, respectively); however, the most significantly altered genes were those related to immune/inflammatory changes. While a basic skin barrier initially forms in utero; to permit rapid fetal growth the fully cornified epidermal barrier to which Abca12 contributes is not completed until just before birth (24). Thus, these in utero inflammatory changes were particularly unexpected as they were triggered before complete barrier maturation or the need for barrier competency. Enriched GO terms included GO:0002376 [immune system processes (P , 1.37e207)], GO:0006955 [immune responses (P , 1.22e205)], GO:0009605 [response to external stimulus (P , 1.28e205)], GO:0008009 [chemokine activity (P , 3e205)], GO:0042379 [chemokine receptor binding (P , 3e205)], GO:0006954 (inflammatory response (P , 0.0008)), GO:0006935/GO:0042330 [chemotaxis/taxis (P , 0.0008)] and GO:0009611 [response to wounding (P , 0.002)]. A sample of genes associated with this innate immune response, inflammation, chemokine and cytokine signature are shown in (Fig. 1E and F). These include Lysozymes (Lyz and Lyz2) which are anti-bacterial glycoside hydrolase enzymes integral to the innate immune system of the skin (25), genes regulating inflammatory IL-1 receptor and NFkB signalling (Il-1rn, Nfkbia and Nfkbiz), many targets suggestive of a type 1 interferon response and a large number of chemokines of the Ccl and Cxcl gene families and the chemokine response associated genes S100a8 and S100a9 (Fig. 1E and F). Chemokine up-regulation and immune recruitment IL-1a and IL-1 receptor signalling have been identified as key players in other related inflammatory skin pathologies (10,26), so we investigated their activity in HI. However, we did not

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

their long-term use (10). Their effectiveness as an HI treatment is also under review amid debate about whether the perceived improvement in disease is a function of retinoid activity or the product of better disease management (8). In skin diseases such dermatitis and psoriasis, where compromised barrier function plays a role in the development of hyperkeratotic phenotypes, inflammatory immune responses are thought to be important for the progression of disease (11). Keratinocytes themselves express a range of cytokines which act to recruit components of the innate immune system (12). In the context of the ichthyoses, a limited number of studies have suggested a role for inflammatory dysregulation in disease development. This is particularly the case in lamellar ichthyosis (LI; caused by mutations in trans-glutaminase 3; TGM3) wherein interleukin 1 alpha (IL-1a) receptor antagonists have been shown to improve epidermal differentiation in in vitro models (10). Similarly, transgenic overexpression of the proinflammatory transcription factor NRF2 was found to phenocopy LI through a mechanism involving the response to the presence of reactive oxygen species in the epidermis (13). The availability of HI mouse models provides an opportunity to characterize and delineate the molecular pathways impacted by Abca12 mutation. In particular, they facilitate the investigation of the processes underlying the development of in vivo and in utero phenotypes which contribute to the high incidence of neonatal death in HI patients. Using these models, we have profiled profound defects in keratinocyte differentiation and an underlying and unexpected upregulation of a number of pro-inflammatory chemokines and cytokines. Transgenic expression of the broad spectrum inflammatory suppressor IL-37b largely ameliorated inflammation in utero, effecting a considerable improvement in the defects ordinarily apparent in keratinocyte differentiation in HI. Our studies indicate that the low rates of survival in HI patients at birth result in part from an inflammatory phenotype in the skin which contributes to aberrant keratinocyte differentiation; combined with lipid defects in barrier function. We show that ameliorating these inflammatory processes results in a considerable improvement in keratinocyte differentiation and thus lowers disease severity in models of this disease.

437

438

Human Molecular Genetics, 2015, Vol. 24, No. 2

observe elevation of Il-1a mRNA in Abca12lx12/lx12 samples (at E17.5), despite the sample integrity being validated by downregulation of Abca12 mRNA (Supplementary Material, Fig. S1B and C). Instead, we did confirm the upregulation of Il-1rn mRNA [encoding the endogenous IL-1 receptor antagonist (IL-1RA)] and Nfkbia (encoding the NFkB inhibitor, IkBa), consistent with our array results (Supplementary Material, Fig. S1D and E). These changes are qualitatively different from the activation of the classic IL-1 inflammatory pathway identified in in vitro rat models of LI (10) and suggest that the inflammatory circuits activated as a consequence of ABCA12 dysfunction are distinct from LI (and hence are unlikely to be a simple product of defective barrier function). In assessing other potential pro-inflammatory mechanisms, we were struck

by the upregulation of a number of chemokines, in particular Ccl7, Ccl9, Ccl11, Cxcl1, Cxcl12 and Cxcl16. These changes reflect the results of the GO analysis [chemokine activity (P , 3e205), chemokine receptor binding (P , 3e205)]. We confirmed upregulation of Ccl9, Ccl11, Cxcl1 and Cxcl16 by qRT-PCR (Fig. 2A– D) and/or by western blot analysis (Fig. 2E – H) in Abca12lx12/lx12 skin relative to wild-type mice and examined the protein expression of CCL9, CCL11 and CXCL16 by immunofluorescence. CCL9 was predominantly detected in K14-expressing cells with a perinuclear to nuclear distribution and increased numbers of K14/CCL9 expressing cells were noted in Abca12lx12/lx12 embryonic skin (Fig. 2I). The expression of CCL11 was expanded in granular cell layers of mutant epidermis compared with wild-type controls

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Figure 1. Microarray analysis of Abca12 knockout embryonic epidermis. (A) Haematoxylin and Eosin (H&E) stained E18.5 embryonic skin sections of wild-type and lx12/lx12 HI mice. Dashed line marks boundary of epidermis and dermis. (B) Keratin 6 (K6- red) and Keratin 14 (K14-green) immunostainings of E18.5 embryonic skin sections of wild-type and lx12/lx12 HI mice. Nuclear stain 4′ ,6-diamidino-2-phenylindole (DAPI) is shown in blue. Scalebars are 100 mm. (C) Selection of keratinocyte differentiation-related genes significantly upregulated at least 1.6-fold (P , 0.05) in microarray mRNA analysis of E17.5 epidermis, in 3 lx12/lx12 HI mice relative to three wild-type mice. (D) Selection of keratinocyte differentiation-related genes significantly downregulated at least 1.6-fold (P , 0.05) in microarray mRNA analysis. (E and F) Selection of immune and inflammatory-related genes significantly upregulated (P , 0.05) in microarray mRNA analysis.

Human Molecular Genetics, 2015, Vol. 24, No. 2

439

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Figure 2. Confirmation of elevated chemokine expression in HI embryos. (A–D) qRT-PCR analysis of Ccl9, Ccl11, Cxcl1 and Cxcl16 mRNA expression in E18.5 lx12/lx12 HI whole skin relative to wild-type mice (defined as 1) (n ¼ 6– 8). (E–H) Western blot analysis of CCL9, CCL11, CXCL1 and CXCL16 protein expression and loading control b-tubulin, in lx12/lx12 HI and wild-type E18.5 embryos (n ¼ 3). (I–K) Immunostaining of CCL9, CCL11 and CXCL16 protein expression in E18.5 embryonic skin, counterstained with DAPI and K14 (n ¼ 4). (L–O) FACS analysis of immune cell recruitment in the dermis of E18.5 lx12/lx12 HI skin (n ¼ 3 –8) relative to wild-type and lx12 heterozygous mice (n ¼ 8, defined as 1). Errorbars represent SEM. Scalebars are 50 mm.

(Fig. 2J). CXCL16 expression was increased throughout the epidermis but especially in the more differentiated epidermal layers of Abca12lx12/lx12 mutants (excluding the cornified envelope)

(Fig. 2K). We were unable to detect CXCL1 expression by immunofluorescence using available antibodies. Given that CCL9 recruits monocytes (27), CCL11 recruits all granulocytes

440

Human Molecular Genetics, 2015, Vol. 24, No. 2

including eosinophils, neutrophils and basophils (28) and both CXCL1 and CXCL16 are chemo-attractants for neutrophils (29,30) we examined immune cell recruitment to either the dermis or epidermis of fetal skin. No significant changes were observed in the epidermal fraction; however, the percentage of granulocytes (CD11bHiGR1+) and inflammatory monocytes (CD11bHiLy6C+) present in the dermis were significantly elevated (Fig. 2L – O), indicating increased chemokine activity. Grafted HI model

We next sought to examine the impact of these changes on the development of disease phenotypes that may contribute to neonatal lethality. Because the inflammatory response we observed was both fetal and pleiotropic (making drug selection and delivery difficult) we chose to modulate inflammation via a genetic approach. This was achieved by crossing the Abca12lx12 loss-of-function allele with mice constitutively overexpressing IL-37b. This enigmatic cytokine has been shown to possess a remarkable capacity to suppress inflammatory phenotypes (32). The transgenic mouse strain we employed uses a CMV promoter to drive low levels of ectopic IL-37b expression which has been shown to lower plasma expression of pro-inflammatory cytokines (IL-6 and IL-1b) and chemokines (including CCL2, CCL5, CCL11, CXCL1, CXCL2, CXCL9 and CXCL10) and which protects against LPS-induced endotoxic shock and colitis (31,32). The full mechanism of action for IL-37b is currently unknown; however, it is thought to act in part through Smad3 to limit signalling downstream of a number of activated receptors that respond to a variety of pro-inflammatory stimuli (32). To confirm the efficacy of IL-37b expression in repressing pro-inflammatory chemokines in our model, we analysed the protein levels of CCL9, CCL11, CXCL1 and CXCL16 in samples derived from E18.5 embryos. By western blot, mice carrying the IL-37b transgene exhibited reduced levels of CCL9, CXCL1 and CXCL16 (but not CCL11) in whole skin (Fig. 4A– D). Downregulation of CCL9 and CXCL16 was also confirmed by immunohistochemistry (Fig. 4E – G). These results demonstrate that IL-37b expression is able to repress at least a subset of the pro-inflammatory chemokines upregulated in HI.

Modulating inflammation alters lipid profile but fails to rescue barrier function defects Having shown we could dampen the inflammatory response, we first examined whether this translated into an improvement in skin barrier function. We assessed this using an ‘outside-in’ toluidine blue dye assay and found that expression of IL-37b on the background of ABCA12 loss of function was unable to prevent overt defects in the skin’s capacity to exclude dye (Fig. 5A). To examine whether a more subtle rescue was effected, we also undertook trans-epidermal water loss (TEWL) assays to assess ‘inside-out’ loss of barrier function, using samples of skin taken from embryos just prior to birth. We confirmed the persistent loss of barrier function and showed no improvements (Fig. 5B). These results were not surprising, given that the central lipid dysregulation defect caused by loss of ABCA12 is unlikely to be corrected simply by altering inflammation. However, we did wish to determine whether the inflammatory response in HI embryos could feedback and impact lipid profiles in any way. To this end, we finely analysed the profile of .400 lipid species in E18.5 epidermis using mass spectrometry. We saw dramatic changes in several lipid classes between control and HI embryos, but notably observed subtle changes in the lipid profile between Abca12lx12/lx12 and Abca12lx12/lx12;IL-37b+ embryos, with increases in dihydroceramides, ceramides, acylceramides and sphingomyelin (Fig. 5C;

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Mouse models of HI are uniformly characterized by perinatal lethality (4 –6), making it difficult to assess whether the inflammatory changes that we observed in the fetus are maintained or modulated after birth. We therefore attempted to graft E17.5 skin from Abca12lx12/lx12 embryos onto Nude mice (Supplementary Material, Fig. S2A and B) but in our hands, these grafts were not viable (Supplementary Material, Fig. S2C). As an alternative, we employed the ENU mutagenesis derived strain Abca12el12/el12, which carries a missense mutation in one of the ABCA12’s transmembrane domains. In contrast to Abca12lx12/lx12 grafts, those from Abca12el12/el12 embryos were 100% viable after 14 days (Fig. 3A and B; Supplementary Material, Fig. S2C), suggesting that this might be a hypomorphic allele. These grafts demonstrated all the typical properties of HI previously observed, including loss of lipid in the stratum corneum (stained via Nile Red, Fig. 3C), loss of lipid in lamellar bodies (Fig. 3D) and premature expression of differentiation markers alongside K14 (Fig. 3E; Supplementary Material, Fig. S2F – H). Close comparison of Abca12el12/el12 embryonic skin at E17.5 and after 14 days after grafting, highlighted some important changes in Abca12el12/el12 grafted skin. While embryonic el12 skin has initial subtle differences to lx12 [lack of K6, normal proliferation (4)], these are lost upon grafting (Supplementary Material, Fig. S2D and E) suggesting that el12 represents a hypomorphic allele (albeit a very severe one) that requires slightly more time to manifest its full HI phenotype. Premature expression of filaggrin (FIL), involucrin and loricrin (LOR) also increased in grafts (Supplementary Material, Fig. S2F – H, see arrowheads). Importantly, we noted improvements in grafted skin in that basal K10 was not detected (Fig. 3E, see arrowheads) and FIL expression was detected in the cornified envelope (Supplementary Material, Fig. S2F). These changes are consistent with the self-improvement of HI keratinocyte differentiation previously reported (9) and with the moderate reduction in disease severity in HI patients in the weeks and months following birth. We next examined if the chemokine signature identified in embryonic Abca12lx12/lx12 skin was also apparent in embryonic and grafted Abca12el12/el12 skin. CCL9, CCL11 and CXCL16 were indeed detected with increased expression in embryonic Abca12el12/el12 skin and CCL9 and CXCL16 remained elevated in HI grafts (Fig. 3G – I, compare brackets). However, we observed a reduction in granular layer CCL11 expression (range and intensity) when comparing Abca12el12/el12 14dpg grafts with fetal skin (Fig. 3H, compare brackets). These observations indicate that this embryonic inflammatory signature is a feature of different HI mouse models and that at least some elements of the pro-inflammatory response precipitated by loss of Abca12 persist in adult HI.

Suppression of inflammatory cytokines via cross with IL-37b transgenic mice

Human Molecular Genetics, 2015, Vol. 24, No. 2

441

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Figure 3. Grafted HI model. (A) E17.5 embryonic skin from el12/el12 HI mice and control siblings were grafted onto the backs of Nude mice and analysed 14 days post graft (14 dpg) (n ¼ 5 –9). (B) H&E stained section of 14 dpg skins. (C) Nile red lipid staining of the 14 dpg epidermis with stratum corneum (SC) indicated by bracket. (D) Transmission electron micrograph (TEM) of 14 dpg stratum corneum with lamellar bodies indicated by arrowheads. (E– I) E17.5 and 14 dpg skin sections, immunostained as indicated (n ¼ 3). Arrowheads indicate basal keratinocytes expressing K10. Brackets show expression/distribution of indicated chemokine for comparison. Scalebars are 100 mm (B and C) and 50 mm (E –I), unless otherwise indicated (as in TEM).

Supplementary Material, Fig. S3A). While these lipids did not normalize to WT levels (nor did we expect them to, given the overarching primary lipid transport defect), this general increase in the ceramide production pathway is consistent with the elevated ceramide levels observed upon self-improvement in grafting models (9). Although these changes are subtle, they do indicate that the inflammatory response initiated in keratinocytes lacking ABCA12, influences pathways which regulate lipid homeostasis in the organ.

In utero suppression of inflammatory chemokines promotes improvement of keratinocyte differentiation Although barrier function defects remain in IL-37b rescue animals, we wished to determine whether expression of IL-37b was able to correct any other HI skin features we had characterized prior to birth. Double transgenic offspring still demonstrated the epidermal thickening evident in HI (Fig. 6A and B); however, the distribution of cells within the various

442

Human Molecular Genetics, 2015, Vol. 24, No. 2

layers of the epidermis was considerably improved. In double transgenic embryos, the expansion of the K14+ layer was reduced relative to Abca12lx12/lx12 embryos, as detected and quantified by IHC (Fig. 6C; Supplementary Material, Fig. S3B), western blot and densitometry (Fig. 6D). Importantly, the proliferation of keratinocytes did not change appreciably as a result of IL-37b expression on an Abca12lx12/lx12 background (Fig. 6E and F) nor were changes in apoptosis a significant feature (Supplementary Material, Fig. S3E), indicating that the reduction in K14+ keratinocytes is most likely a function of improved differentiation. To investigate this possibility further, we measured expression of other markers of epidermal differentiation and wounding. The expression of the wounding keratin, K6, was considerably reduced by expression of IL-37b (Fig. 6G and H). While K10 protein expression levels were unchanged (Supplementary Material, Fig. S3C), there was also

improvement by way of a reduction in aberrant co-expression of basal (K14) and spinous (K10) layer markers (Fig. 7A and B). The distribution of FLG also shifted from the granular layer towards the cornified envelope as a consequence of suppressed inflammation (Fig. 7C); however, the expression and processing of the protein remained unchanged (Fig. 7D). Similarly, LOR expression in the granular layer was more compact (Fig. 7E and G), despite no change in total levels of protein expression (Fig. 7F). Finally, we measured the thickness of the cornified envelope and observed an increase in thickness in Abca12lx12/lx12; IL-37b+ embryos relative to Abca12lx12/lx12 skin (Fig. 7H). These observations clearly demonstrate that suppressing inflammatory responses in an HI model is able to considerably correct the defects caused by loss of ABCA12 function through a significant normalization of the keratinocyte differentiation program (Fig. 7I).

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Figure 4. Suppression of inflammatory chemokines via cross with IL-37b transgenic mice. (A– D) Western blot and densitometry analysis of CCL9, CCL11, CXCL1 and CXCL16 protein expression, standardized to loading control b-tubulin, in Wild type, IL-37b+, lx12/lx12 HI and lx12/lx12; IL-37b+ HI whole skin protein extracts (n ¼ 3–4). Embryos were generated from intercross of lx12 and IL-37b transgenic mouse strains. Values are relative to wild type defined as 1. Errorbars represent SEM. (E–G) Immunostaining of CCL9, CCL11 and CXCL16 protein expression in E18.5 embryonic skin, counterstained with DAPI and K14 (n ¼ 3). Scalebars are 50 mm.

Human Molecular Genetics, 2015, Vol. 24, No. 2

443

DISCUSSION HI presents as a severe disease from birth, and neonatal lethality associated with the condition is very high. By using mouse models of this disorder, we have profiled a significant increase in pro-inflammatory signalling within the fetal epidermis. That this occurs in utero means that it does so without the considerable influence of dehydration and infection which characterizes the organ after birth. Our localization studies and the array itself (which was undertaken only on the epidermis) instead suggest that keratinocyte-derived expression of pro-inflammatory chemokines is an intrinsic feature of HI. This raises the question as to what is driving this pre-natal response in the skin? We propose that this is a consequence of the very considerable lipid dysregulation we have characterized downstream of defective ABCA12 function. In particular, recent observations that accumulation of intracellular ceramide directly promotes development of ichthyosis-like phenotypes (33) are particularly prescient. Significantly, many of the microarray hits in our study are in common with that of Amen et al. (33), including Repetin, Cxcl1, Keratin 6, Epigen and several Sprrs. Intracellular ceramides may trigger inflammation and impair keratinocyte differentiation in many ways. Intracellular ceramide-driven ichthyosis is thought to occur through induction of the lipid-regulated transcription factor, PPARb/d (33), as ceramides have been shown to both induce Pparb/d mRNA and to generate endogenous ligands to promote PPARb/d activity (34). Furthermore, PPARb/d is an attractive candidate to link ceramides to skin dysfunction, as it often behaves as a pro-inflammatory retinoic acid signalling transcription factor that regulates granular layer formation (and hence barrier function) and it has been shown to drive psoriasis-like para-keratotic skin differentiation (34– 37). We also observed upregulation of Pparb/d mRNA in our mouse HI model (Supplementary Material, Fig. S3D), coinciding with the considerable increase in the

levels of ceramide species in extracts of keratinocytes. While we did not see Pparb/d mRNA levels reduce upon IL-37b expression, PPARb/d activity can be reduced by inflammatory inhibitors such as aspirin (38) and it will be interesting to examine if IL-37b impacts PPARb/d activity or downstream inflammatory mediators. Similarly, links are emerging between the production of chemokines and cell differentiation. One such example is Caspase-14, whose expression is downregulated in response to LPS-induced inflammatory signals (39) and which encodes a protease critically required for FLG processing during keratinocyte differentiation and barrier formation (40). Although the precise links between inflammation and differentiation require further elucidation, the results of this study support a model in which the accumulation of ceramides, their derivatives and/or other dysregulated lipid species elicit the production of chemokines in keratinocytes prior to birth which then impacts their differentiation. This model therefore splits the HI phenotype into two distinct driver pathways, the barrier dysfunction caused by loss of extracellular lipids in the stratum corneum, and inflammatory responses due to lipid dysregulation (Fig. 7J). The involvement of defective lipid transport in HI compared with other forms of ichthyosis which derive from ‘structural’ epidermal defects (such as TGM3 mutations in LI) might explain the differences we note in the type of inflammatory response elicited (10,26) (Supplementary Material, Fig. S1B – E). Notably, we demonstrate an increase in signalling pathways mediated by members of the CCL/CXCL chemokine families. The specific impact of each of these different proteins on keratinocyte differentiation and disease aetiology remains to be determined; however, several of these proteins have already been linked to inflammatory skin disease. CXCL1 (GROa) is a chemoattractant for neutrophils and is enriched in the upper epidermal layers in the skin diseases verruca vulgaris, psoriasis, keratoacanthoma and squamous cell carcinoma (29). CXCL1 is also upregulated during chronic skin inflammation triggered by epidermal

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Figure 5. Assessment of barrier function. (A) Toluidine Blue dye exclusion analysis was performed on a litter of intact E18.5 embryos. (B) TEWL assays were performed on E18.5 back skin with skins weighed for water loss every half an hour over 5 h (n ¼ 5– 8). (C) Lipid analysis of E18.5 epidermis. Control samples (pool of WT and lx12/+) are defined as 1 and HI samples are shown relative to this baseline. Errorbars represent SEM.

444

Human Molecular Genetics, 2015, Vol. 24, No. 2

deletion of BLIMP1, a transcriptional repressor expressed in cells of the granular layer (41). Similarly, CXCL16 is upregulated in psoriasis (42) and has anti-microbial activity (43). CCL11 expression by differentiating keratinocytes recruits granulocytes including eosinophils, neutrophils and basophils to the site of infection or inflammation (28) and over expression in associated with chronic atopic dermatitis (44). A key question is to what extent anti-inflammatories might present a potential therapeutic for ameliorating disease severity? We have addressed this question using a pleiotropic genetic approach, through the expression of the anti-inflammatory molecule, IL-37b. This led to a significant decrease in the expression of many of the chemokines we had previously shown were upregulated as a consequence of ABCA12 mutation, and to a number of significant improvements in disease phenotype. Principal among these was a correction in keratinocyte differentiation defects; namely a reduction in basal layer thickness, a decrease in the expression of wounding-associated keratins, a correction

in the level of co-expression of keratins that are normally sequentially activated during differentiation and a reduction in abnormal thickening of the granular layer that occurs as a consequence of failed compaction. Taken together, these findings indicate that inflammatory signals which arise from loss of ABCA12 function are central to the defective keratinocyte differentiation observed in HI. Quantitatively dissecting the individual (or combinatorial) effects of these different chemokines will be important in assessing the mechanism(s) underlying the correction of these defects. There has been considerable recent interest in the emerging role that anti-inflammatory cytokines such as IL-10, IL-27, IL-35 and IL-37 play in moderating immune responses. The ability of these proteins to limit inflammatory reactions in a number of disease contexts suggests considerable promise as a new type of therapeutic approach by which to treat a number of different diseases. The work presented here is one of only a handful of studies that have employed IL-37b in vivo to modulate

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Figure 6. In utero suppression of inflammatory cytokines promotes signs of improvement of keratinocyte differentiation. (A) H&E stained section of E18.5 embryos generated from intercross of lx12 and IL-37b transgenic mouse strains. Dashed line marks boundary of epidermis and dermis. (B) The total epidermal thickness was quantified in cross sections of H&E stained skin across the four genotypes indicated (n ¼ 6– 8). (C, E and G) Immunostaining as indicated in E18.5 embryonic skin (n ¼ 3–8). (D) Western blot and densitometry analysis of K14 protein expression (n ¼ 4) standardized to loading control b-tubulin, in Wild type, IL-37b+, lx12/lx12 HI and lx12/lx12; IL-37b+ HI whole skin protein extracts. (F) The number of PCNA+ve keratinocytes was scored per epidermal column (defined as five basal keratinocytes wide) (n ¼ 3– 8). (H) Western blot and densitometry analysis of K6 protein expression (n ¼ 4) standardized to loading control actin, in Wild type, IL-37b+, lx12/lx12 HI and lx12/lx12; IL-37b+ HI whole skin protein extracts. Errorbars represent SEM. Scalebars are 50 mm.

Human Molecular Genetics, 2015, Vol. 24, No. 2

445

an inflammatory response or to study a disease process. In particular, it suggests that IL-37b might be employed effectively as a novel anti-inflammatory therapeutic for the treatment of several skin diseases. We particularly note that IL37b is considerably upregulated in dermatitis, in which the epidermal skin barrier is compromised (45) as well as in other conditions such as colitis where barrier function in the colonic epithelium is disrupted (31). IL-37b has also been shown to ameliorate inflammation in psoriasis (46). While suppressing the inflammatory response in HI does not improve the intrinsic defects in barrier function of the skin nor prevent lipid accumulation (as we

expected, given that the central failure in lipid transport remains), we did observe changes in lipid levels which were mediated by expression of IL-37b. These changes suggest communication between inflammation and lipid homeostasis. In the absence of a therapy to directly correct ABCA12 mutation, our findings indicate that a beneficial strategy to increase HI neonate survival might therefore involve a two-pronged approach (Fig. 7J). First, a lipid emollient would be required to create an artificial external skin barrier to address the loss of water-proofing extracellular lipids in the stratum corneum. Secondly, as we demonstrate in this report, anti-inflammatory

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Figure 7. In utero suppression of inflammatory cytokines promotes signs of keratinocyte self-improvement of differentiation continued. (A) K14/K10 immunostaining in E18.5 embryonic skin (n ¼ 3 –8). Note: the aberrant K14/K10 co-expressing cells in yellow, as quantified in (B) (n ¼ 3). (C and E) Immunostaining as indicated in E18.5 embryonic skin (n ¼ 3). Arrowheads show aberrant K14/FLG co-expressing keratinocytes.(D) Western blot analysis of Filaggrin (FLG) protein expression, and (F) LOR protein expression standardized to loading control actin, in Wild type, IL-37b+, lx12/lx12 HI and lx12/lx12; IL-37b+ HI whole skin protein extracts (n ¼ 4). (G) The thickness of the granular epidermal layer was quantified in cross sections of LOR stained skin across the four genotypes indicated (n ¼ 3). (H) The thickness of the cornified layer was quantified in cross sections of H&E stained skin across the four genotypes indicated (n ¼ 4). (I) Summary of epidermal layer redistribution observed. (J) Summary model showing outcome of in utero suppression of inflammation. Errorbars represent SEM. Scalebars are 50 mm.

446

Human Molecular Genetics, 2015, Vol. 24, No. 2

therapies may then serve to correct defects in keratinocyte differentiation that occur as a consequence toxic lipid dysregulation.

MATERIALS AND METHODS Mouse strains

Histological analysis and microscopy Embryonic skins were excised from E17.5 to E18.5 embryos and fixed for 2 h in 4% paraformaldehyde at room temperature, while grafts were fixed overnight at 48C. Tissues were paraffin imbedded and sectioned at 4 – 7 mm. Antigen retrieval was performed in Citrate buffer pH 6 using rice/pressure cookers, or using an automated (Dako) PT link, with EnVISION FLEX buffer (Dako, Glostrup, Denmark). Antibody staining was performed as described elsewhere (47), and cover slips mounted using MolWiol (Calbiochem, Merck Millipore, Billerica, MA, USA). Nile Red staining was performed as described previously (48). Imaging was using an Olympus Fluoview 500, Leica SP5 confocal laser scanning microscope or Olympus AX 70 Provis research microscope (Monash University, Clayton, Victoria, Australia). Transmission electron microscopy was performed as described previously (4). Antibodies and stains Anti-actin (A2066) 1 : 3000 (WB) Sigma-Aldrich, St Louis, MO, USA, anti-b-tubulin (RB-9249-PO) 1:2000 (WB) Thermo Scientific, Tewksbury, MA, USA. Anti-mouse CCL9/CCL10/ MIP-1 gamma (bs2477R) 1 : 100 (IHC) Bioss Inc, Woburn, MA, USA. Anti-mouse CCL9/MIP-1 gamma (ab9913) 1:2000 (WB) Abcam, Cambridge, UK. Anti-CCL11/Eotaxin 1 (bs1601R) 1 : 100 (IHC) 1 : 500 (WB) Bioss Inc, Woburn, MA, USA. Anti-cleaved Caspase 3 (#9664) 1 : 100 (IHC) Cell Signaling Technology, Beverly, MA, USA. Anti-GRO-Alpha/ CXCL1 (bs0863R) 1 : 500 (WB) Bioss Inc, Woburn, MA, USA. Anti-CXCL16 (bs1441R) 1 : 100 (IHC) 1 : 500 (WB) Bioss Inc, Woburn, MA, USA. Anti-filaggrin (PRB-417P) 1 : 1000 (IHC) 1 : 1000 (WB) Covance, Princeton, NJ, USA. Antiinvolucrin (PRB-140C) 1 : 1000 (IHC) 1 : 1000 (WB) Covance, Princeton, NJ, USA. Anti-keratin 6 (PRB-169P) 1 : 500 (IHC) 1 : 1000 (WB) Covance, Princeton, NJ, USA. Anti-keratin 10 (PRB-159P) 1 : 500 (IHC) 1 : 1000 (WB) Covance, Princeton, NJ, USA. Anti-keratin 14 (LL002) (ab7800) 1 : 250 –1000 (IHC) Abcam, Cambridge, UK. Anti-keratin 14 (PRB-155P)

RNA and microarray analysis Epidermal peels from E17.5 lx12 and wild-type embryos were prepared by dermal separation using 5 mM EDTA for 1 h at 378C. RNA extraction was performed with RNAqueous, RNA purification kit (Ambion, Life Technologies, Carlsbad, CA, USA). RNA integrity was determined using agarose gel electrophoresis and by the Agilent Bioanalyser 2100 NanoChip protocol. A total of 500 ng was labelled using the Illumina Total Prep RNA amplification kit (IL1791, Ambion, Life Technologies, Carlsbad, CA, USA). A total of 1.5 mg of labelled cRNA was prepared for hybridization to the Sentrix MouseWG6v2 Beadchip, processed and analysed according to Illumina protocols. Expression values were analysed using GeneSpring GX 12.5– 6 software. Genes were normalized to the all sample average. Gene changes between Abca12 lx12/lx12 and control littermates were filtered for absent flags, limited to fold changes of at least 1.6-fold or greater, and the resulting gene list further screened for statistical significance using the unpaired t-test with P , 0.05 cut off. Data are available at NCBI’s GEO archive under GSE56125. For qRT-PCR, RNA was isolated from whole skin using Trizol Reagent (Invitrogen, Life Technologies, Carlsbad, CA, USA) as described previously (47) except DNAse treatment was performed using Ambion TURBO DNA free DNase (Invitrogen, Life Technologies, Carlsbad, CA, USA). cDNA synthesis was performed using Supercript Vilo (Invitrogen, Life Technologies, Carlsbad, CA, USA). qRT-PCR was performed using SYBRw Green PCR Master Mix (Applied Biosystems, Life Technologies, Carlsbad, CA, USA) as described previously (49). Primers were as follows: Gene

5′ -Forward-3′

5′ -Reverse-3′

Abca12

ACTGTGTCATTTCCC TCCAAC TCATACTGCCCTCTC CTTCC CAGATGCACCCTGAAAG CCATA AAAAGGTGTCCCCAAGTA

TCTTCATTCCTCCC GACAAC GGACAGGCAGCAA TCTGAAG TGCTTTGTGGCAT CCTGGAC AAGCAGAACTGAA CTACCATCG GACCAGTTCCACAC TCTTTGC GCATGGACTGTGGT CATGAG GTTTGATATTTGGTC CTTGTAAG TTGCAGGTCATTTAA CCAAGTG AAGTGGAGTGGAGT CTGCTG

Ccl9 Ccl11 Cxcl1 Cxcl16 Gapdh icIL-1Ra1 IL-1alpha Nfkbia

CAGGTTCCAGTTGC AGTCC GCCAAAAGGGTCATCA TCTC AGACACTGCCTGGGT GCTCCT GATCTGGAAGAGACC ATCCAAC GACTCGTTCCTGCA CTTGG

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Abca12el12/el12 mice have been previously described (4). We obtained mouse strain ‘Abca12tm1Lex’ NIH-0129 by Lexicon Pharmaceuticals, The Woodlands, TX, USA genetics and refer to them as Abca12lx12/lx12 mice in this study. These mice have a puromycin selection cassette-mediated exon 8 disruption and recapitulate the features of HI akin to other Abca12 mutant strains (4 – 6). Transgenic IL-37b mice have been described previously (32) and were crossed with Abca12lx12 mice. All animal procedures complied with standards set under Australian guidelines for animal welfare and experiments were subject to Monash University animal welfare ethics review panels. At least three mice have been analysed for all conditions unless otherwise stated.

1 : 500 (IHC) 1 : 1000 (WB) Covance, Princeton, NJ, USA. Anti-loricrin (PRB-145P) 1 : 1000 (IHC) 1 : 1000 (WB) Covance, Princeton, NJ, USA. Anti-PCNA (C-20) (sc-9857) 1 : 50 (IHC) 1 : 250 (WB) Santa Cruz Biotechnology, Dallas, TX, USA. Alexa Fluorw A488/555/568/647 secondary antibodies raised in Donkey, against rat/rabbit/mouse, were used at 1 : 400– 600, from Invitrogen Nuclei stains used were TOPRO3 1 : 600 (Invitrogen, Life Technologies, Carlsbad, CA, USA) and DAPI (Sigma-Aldrich, St Louis, MO, USA) 1 : 1000.

Human Molecular Genetics, 2015, Vol. 24, No. 2

447

Determination of lipid levels by mass spectroscopy

Whole skin pieces were collected to ice-cold RIPA buffer containing protease inhibitors and homogenized. The protein supernatant was collected after centrifugation and stored at 2808C. Twentyfive micrograms of aliquots of protein were prepared in 4× reducing sample buffer (0.25 M Tris–HCl at pH 6.8, 8% SDS, 30% glycerol, 0.02% bromophenol blue with a reducing agent, 10% 2-ME) and boiled for 5 min before loading and performing SDS–PAGE and western blotting. Signals were detected using ECL. Densitometry was performed as described previously (50).

Epidermal fractions from E18.5 lx12 litters were collected and immersed in 0.3 ml of 2.5% dispase/0.125% trypsin in DMEM at 378C for 1.5 h. Following incubation, the epidermis was mechanically peeled from the dermis and homogenizised in 150 – 200 ml PBS and stored at 2808C. Protein concentration was quantified using a BCA protein assay (Pierce Biotechnology, Thermo Fisher Scientific, Rockford, IL, USA) following the manufacturer’s protocol. Lipid analysis was performed as described previously (52), but briefly, involved the analysis of 400 lipid species of the lipid classes and subclasses: dihydroceramide, ceramide, acylceramide, glucosylceramide, acylglucosylceramide, sphingomyelin, diacylglycerol, triacylglycerol, cholesterol, cholesterol esters, dihexosylceramide, trihexosylceramide, GM3 ganglioside, lysophosphatidylinositol, lysophosphatidylethanolamine, phosphatidylglycerol, bismonoacylglycerol phosphate, lysophosphatidylcholine, lysoalkylphosphatidylcholine, phosphatidylcholine, phosphatidylinositol, phosphatidylethanolamine and phosphatidylserine. Data are presented as the sum total of each lipid family and output was normalized to the abundant cellular membrane lipid, phosphatidylcholine.

Cell isolation and flow cytometry E18.5 belly skin was placed in 2.5% dispase + 0.125% trypsin in DMEM for 1 – 1.5 h at 378C. Samples were vortexed then mechanically peeled with forceps. The dermal peel was returned to dispase/trypsin solution and an equal volume of 0.2% Collagenase IV in DMEM added for 1 – 1.5 h at 378C. The epidermal peel was incubated in 0.25% trypsin for 1 – 1.5 h at 378C. 1 ml of Foetal Bovine Serum was added and each sample passed through 70 mm cell strainer, and rinsed with 9 ml DMEM. The cell solution was spun at 1100 rpm (Sorvall Heraeus 75006445 rotor), rinsed in 1 ml PBS and pelleted again before resuspension in 300ml PBS. Isolated cells were fixed by addition of 1 ml pre-warmed 1.5% PFA in PBS for 10 min at room temperature. Fixation was terminated with 1 ml of ice-cold PBS and fixed cells pelleted at 1100 rpm (Sorvall Heraeus 75006445 rotor) for 5 min at 48C, then washed with ice-cold PBS and pelleted again. The final cell pellet was resuspended in 1 ml ice-cold 90% methanol, added drop wise while swirling. Cells were placed on ice for 20 min then stored at 2808C until use. Flow cytometry of cells was performed as described elsewhere (51) but briefly cells were resuspended in staining buffer (2% FCS, 0.5 mM EDTA in PBS) and incubated for 2 h on ice, to rehydrate. Fc receptors were blocked using anti-CD16/32 (eBioscience). Cell populations were first gated for CD45 and then characterized as: CD11bhiGr1+ granulocytes or CD11bhiLy6C+ inflammatory monocytes. Gates were determined using fluorescence-minus-one gating and stained spleen cells for control staining due to low cell numbers in the skin samples. Cells were acquired using an FACS Canto II flow Cytometer (BD) and analysed using Flow Jo software.

Barrier function assays TEWL and toluidine blue dye exclusion assay were performed as described previously (4). Quantification and statistical analyses All quantification of the thickness of various strata of the epidermis was carried out using ImageJ (Java) software. In all cases, the average of three measurements from three separate areas of each sample was taken, giving a total of nine measurements per sample. These were then averaged across at least three independent biological samples per genotype to allow for comparison across the four genotypes used in this study. The data throughout this study were analysed by Student’s t-test.

SUPPLEMENTARY MATERIAL Supplementary Material is available at HMG online.

ACKNOWLEDGEMENTS Skin transplantation Abca12el12/el12 mice and Abca12lx12/lx12 fetal skin transplants were performed as follows: full thickness dorsal embryonic epidermis was removed from E18.5 embryos using sterile scissors, and skin grafts were immersed in a PBS on ice. Recipient BALBc nude mice were anaesthetized by isoflurane and injected subcutaneously with analgesic carprofen. Embryonic skins were tailored with a sterile scalpel to match the area removed from recipient mice and were immediately grafted by placing the embryonic skin onto the recipient area and spread in place. Grafts were secured with latex Band aids and enforced with a layer of elastic adhesive plaster tape. Mice and grafts were observed daily and bandages were removed 7 days after surgery. Recipient mice were euthanized by lethal injection with Nembutal (Sigma-Aldrich, St Louis, MO, USA).

We thank Jerick Guinto from the Australian Genome Research Facility (AGRF) (Parkville, Australia) and Vivian Vasic Monash Health Translation Precinct (MHTP) Medical Genomics Facility (MIMR-PHI, Clayton, Australia) for assistance with microarray analysis. We also thank Douglas J. Hilton and Charles A. Dinarello for providing mice. We also thank the Monash Histology Platform for their assistance in this project. Conflict of Interest statement. None declared.

FUNDING This work was supported by the National Health and Medical Research Council (Australia) by a Project grant (1020281 to I.S., 1012353 and 1043845 to M.F.N. and C.A.N.P.), Program

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

Protein sample, western blot and densitometry analysis

448

Human Molecular Genetics, 2015, Vol. 24, No. 2

Grants (461219, 1016647), an Independent Research Institutes Infrastructure Support Scheme Grant (361646), the Victorian Government’s Operational Infrastructure Support Program to N.E.M., M.F.N. and C.A.N.P. This work was additionally supported by the following fellowships: an Australian Post-graduate Award to SCII, a Monash International Post-graduate Research Scholarship and Monash Graduate Scholarship to T.D.T., an Australian Research Council Australian Research Fellowship to N.E.M., a Fellowship from the Sylvia and Charles Viertel Foundation to B.T.K., Monash University’s Larkins Fellowship to M.F.N., a National Health and Medical Research Council R. Douglas Wright fellowship (384457) an Australian Research Council Future Fellowship (FT100100620) and a Monash University Fellowship to I.S.

REFERENCES

17.

18. 19. 20.

21. 22. 23. 24. 25. 26. 27.

28.

29. 30.

31.

32. 33.

34.

35.

36.

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

1. Akiyama, M. and Shimizu, H. (2008) An update on molecular aspects of the non-syndromic ichthyoses. Exp. Dermatol., 17, 373 –382. 2. Akiyama, M., Sugiyama-Nakagiri, Y., Sakai, K., McMillan, J.R., Goto, M., Arita, K., Tsuji-Abe, Y., Tabata, N., Matsuoka, K., Sasaki, R. et al. (2005) Mutations in lipid transporter ABCA12 in harlequin ichthyosis and functional recovery by corrective gene transfer. J. Clin. Invest., 115, 1777– 1784. 3. Kelsell, D.P., Norgett, E.E., Unsworth, H., Teh, M.T., Cullup, T., Mein, C.A., Dopping-Hepenstal, P.J., Dale, B.A., Tadini, G., Fleckman, P. et al. (2005) Mutations in ABCA12 underlie the severe congenital skin disease harlequin ichthyosis. Am. J. Hum. Genet., 76, 794– 803. 4. Smyth, I., Hacking, D.F., Hilton, A.A., Mukhamedova, N., Meikle, P.J., Ellis, S., Satterley, K., Collinge, J.E., de Graaf, C.A., Bahlo, M. et al. (2008) A mouse model of harlequin ichthyosis delineates a key role for Abca12 in lipid homeostasis. PLoS Genet., 4, e1000192. 5. Yanagi, T., Akiyama, M., Nishihara, H., Sakai, K., Nishie, W., Tanaka, S. and Shimizu, H. (2008) Harlequin ichthyosis model mouse reveals alveolar collapse and severe fetal skin barrier defects. Hum. Mol. Genet., 17, 3075– 3083. 6. Zuo, Y., Zhuang, D.Z., Han, R., Isaac, G., Tobin, J.J., McKee, M., Welti, R., Brissette, J.L., Fitzgerald, M.L. and Freeman, M.W. (2008) ABCA12 maintains the epidermal lipid permeability barrier by facilitating formation of ceramide linoleic esters. J. Biol. Chem., 283, 36624– 36635. 7. Rajpopat, S., Moss, C., Mellerio, J., Vahlquist, A., Ganemo, A., Hellstrom-Pigg, M., Ilchyshyn, A., Burrows, N., Lestringant, G., Taylor, A. et al. (2011) Harlequin ichthyosis: a review of clinical and molecular findings in 45 cases. Arch. Dermatol., 147, 681– 686. 8. Dyer, J.A., Spraker, M. and Williams, M. (2013) Care of the newborn with ichthyosis. Dermatol. Ther., 26, 1 –15. 9. Yanagi, T., Akiyama, M., Nishihara, H., Ishikawa, J., Sakai, K., Miyamura, Y., Naoe, A., Kitahara, T., Tanaka, S. and Shimizu, H. (2010) Self-improvement of keratinocyte differentiation defects during skin maturation in ABCA12-deficient harlequin ichthyosis model mice. Am. J. Clin. Pathol., 177, 106–118. 10. O’Shaughnessy, R.F., Choudhary, I. and Harper, J.I. (2010) Interleukin-1 alpha blockade prevents hyperkeratosis in an in vitro model of lamellar ichthyosis. Hum. Mol. Genet., 19, 2594–2605. 11. Pastore, S., Mascia, F., Gulinelli, S., Forchap, S., Dattilo, C., Adinolfi, E., Girolomoni, G., Di Virgilio, F. and Ferrari, D. (2007) Stimulation of purinergic receptors modulates chemokine expression in human keratinocytes. J. Invest. Dermatol., 127, 660–667. 12. Wang, B., Amerio, P. and Sauder, D.N. (1999) Role of cytokines in epidermal Langerhans cell migration. J. Leukocyte Biol., 66, 33–39. 13. Schafer, M., Farwanah, H., Willrodt, A.H., Huebner, A.J., Sandhoff, K., Roop, D., Hohl, D., Bloch, W. and Werner, S. (2012) Nrf2 links epidermal barrier function with antioxidant defense. EMBO Mol. Med., 4, 364– 379. 14. Strachan, L., Murison, J.G., Prestidge, R.L., Sleeman, M.A., Watson, J.D. and Kumble, K.D. (2001) Cloning and biological activity of epigen, a novel member of the epidermal growth factor superfamily. J. Biol. Chem., 276, 18265–18271. 15. Jiang, C.K., Magnaldo, T., Ohtsuki, M., Freedberg, I.M., Bernerd, F. and Blumenberg, M. (1993) Epidermal growth factor and transforming growth

16.

factor alpha specifically induce the activation- and hyperproliferationassociated keratins 6 and 16. Proc. Natl. Acad. Sci. USA, 90, 6786–6790. Presland, R.B., Rothnagel, J.A. and Lawrence, O.T. (2006) Feingold, P. M. Elias and K. M. Feingold (ed), In Skin Barrier. Taylor & Francis, New York, pp. 111 – 140. Hohl, D., de Viragh, P.A., Amiguet-Barras, F., Gibbs, S., Backendorf, C. and Huber, M. (1995) The small proline-rich proteins constitute a multigene family of differentially regulated cornified cell envelope precursor proteins. J. Invest. Dermatol., 104, 902–909. Vermeij, W.P., Alia, A. and Backendorf, C. (2011) ROS quenching potential of the epidermal cornified cell envelope. J. Invest. Dermatol., 131, 1435– 1441. Marshall, D., Hardman, M.J., Nield, K.M. and Byrne, C. (2001) Differentially expressed late constituents of the epidermal cornified envelope. Proc. Natl. Acad. Sci. USA, 98, 13031–13036. Smits, P., Poumay, Y., Karperien, M., Tylzanowski, P., Wauters, J., Huylebroeck, D., Ponec, M. and Merregaert, J. (2000) Differentiationdependent alternative splicing and expression of the extracellular matrix protein 1 gene in human keratinocytes. J. Invest. Dermatol., 114, 718–724. Molotkov, A., Fan, X. and Duester, G. (2002) Excessive vitamin A toxicity in mice genetically deficient in either alcohol dehydrogenase Adh1 or Adh3. Eur. J. Biochem., 269, 2607–2612. Elias, P.M. (1986) Epidermal effects of retinoids: supramolecular observations and clinical implications. J. Am. Acad. Dermatol., 15, 797–809. Beissbarth, T. and Speed, T.P. (2004) GOstat: find statistically overrepresented Gene Ontologies within a group of genes. Bioinformatics, 20, 1464– 1465. Segre, J.A. (2006) Epidermal barrier formation and recovery in skin disorders. J. Clin. Invest., 116, 1150– 1158. Wiesner, J. and Vilcinskas, A. (2010) Antimicrobial peptides: the ancient arm of the human immune system. Virulence, 1, 440–464. Pazyar, N., Feily, A. and Yaghoobi, R. (2012) An overview of interleukin-1 receptor antagonist, anakinra, in the treatment of cutaneous diseases. Curr. Clin. Pharm., 7, 271–275. Mohamadzadeh, M., Poltorak, A.N., Bergstressor, P.R., Beutler, B. and Takashima, A. (1996) Dendritic cells produce macrophage inflammatory protein-1 gamma, a new member of the CC chemokine family. J. Immunol., 156, 3102–3106. Menzies-Gow, A., Ying, S., Sabroe, I., Stubbs, V.L., Soler, D., Williams, T.J. and Kay, A.B. (2002) Eotaxin (CCL11) and eotaxin-2 (CCL24) induce recruitment of eosinophils, basophils, neutrophils, and macrophages as well as features of early- and late-phase allergic reactions following cutaneous injection in human atopic and nonatopic volunteers. J. Immunol., 169, 2712– 2718. Tettelbach, W., Nanney, L., Ellis, D., King, L. and Richmond, A. (1993) Localization of MGSA/GRO protein in cutaneous lesions. J. Cutan. Pathol., 20, 259– 266. Woehrl, B., Klein, M., Rupprecht, T., Schmetzer, H., Angele, B., Hacker, H., Hacker, G., Pfister, H.W. and Koedel, U. (2010) CXCL16 contributes to neutrophil recruitment to cerebrospinal fluid in pneumococcal meningitis. J. Infect. Dis., 202, 1389–1396. McNamee, E.N., Masterson, J.C., Jedlicka, P., McManus, M., Grenz, A., Collins, C.B., Nold, M.F., Nold-Petry, C., Bufler, P., Dinarello, C.A. et al. (2011) Interleukin 37 expression protects mice from colitis. Proc. Natl. Acad. Sci. USA, 108, 16711–16716. Nold, M.F., Nold-Petry, C.A., Zepp, J.A., Palmer, B.E., Bufler, P. and Dinarello, C.A. (2010) IL-37 is a fundamental inhibitor of innate immunity. Nat. Immunol., 11, 1014– 1022. Amen, N., Mathow, D., Rabionet, M., Sandhoff, R., Langbein, L., Gretz, N., Jackel, C., Grone, H.J. and Jennemann, R. (2013) Differentiation of epidermal keratinocytes is dependent on glucosylceramide:ceramide processing. Hum. Mol. Genet., 22, 4164–4179. Tan, N.S., Michalik, L., Noy, N., Yasmin, R., Pacot, C., Heim, M., Fluhmann, B., Desvergne, B. and Wahli, W. (2001) Critical roles of PPAR beta/delta in keratinocyte response to inflammation. Genes Dev., 15, 3263– 3277. Calleja, C., Messaddeq, N., Chapellier, B., Yang, H., Krezel, W., Li, M., Metzger, D., Mascrez, B., Ohta, K., Kagechika, H. et al. (2006) Genetic and pharmacological evidence that a retinoic acid cannot be the RXR-activating ligand in mouse epidermis keratinocytes. Genes Dev., 20, 1525–1538. Hack, K., Reilly, L., Palmer, C., Read, K.D., Norval, S., Kime, R., Booth, K. and Foerster, J. (2012) Skin-targeted inhibition of PPAR beta/delta by

Human Molecular Genetics, 2015, Vol. 24, No. 2

37. 38. 39. 40.

41.

42.

44.

45. Fujita, H., Inoue, Y., Seto, K., Komitsu, N. and Aihara, M. (2013) Interleukin-37 is elevated in subjects with atopic dermatitis. J. Dermatol. Sci., 69, 173– 175. 46. Teng, X., Hu, Z., Wei, X., Wang, Z., Guan, T., Liu, N., Liu, X., Ye, N., Deng, G., Luo, C. et al. (2014) IL-37 ameliorates the inflammatory process in psoriasis by suppressing proinflammatory cytokine production. J. Immunol., 192, 1815– 1823. 47. Cottle, D.L., Kretzschmar, K., Schweiger, P.J., Quist, S.R., Gollnick, H.P., Natsuga, K., Aoyagi, S. and Watt, F.M. (2013) c-MYC-induced sebaceous gland differentiation is controlled by an androgen receptor/p53 axis. Cell Rep., 3, 427–441. 48. Talreja, P., Kleene, N.K., Pickens, W.L., Wang, T.F. and Kasting, G.B. (2001) Visualization of the lipid barrier and measurement of lipid pathlength in human stratum corneum. AAPS PharmSci., 3, E13. 49. Wiradjaja, F., Cottle, D.L., Jones, L. and Smyth, I. (2013) Regulation of PDGFC signalling and extracellular matrix composition by FREM1 in mice. Dis. Model. Mech., 6, 1426–1433. 50. Cottle, D.L., McGrath, M.J., Cowling, B.S., Coghill, I.D., Brown, S. and Mitchell, C.A. (2007) FHL3 binds MyoD and negatively regulates myotube formation. J. Cell Sci., 120, 1423–1435. 51. Nold, M.F., Mangan, N.E., Rudloff, I., Cho, S.X., Shariatian, N., Samarasinghe, T.D., Skuza, E.M., Pedersen, J., Veldman, A., Berger, P.J. et al. (2013) Interleukin-1 receptor antagonist prevents murine bronchopulmonary dysplasia induced by perinatal inflammation and hyperoxia. Proc. Natl. Acad. Sci. USA, 110, 14384–14389. 52. Weir, J.M., Wong, G., Barlow, C.K., Greeve, M.A., Kowalczyk, A., Almasy, L., Comuzzie, A.G., Mahaney, M.C., Jowett, J.B., Shaw, J. et al. (2013) Plasma lipid profiling in a large population-based cohort. J. Lipid Res., 54, 2898–2908.

Downloaded from http://hmg.oxfordjournals.org/ at Seton Hall on March 30, 2015

43.

selective antagonists to treat PPAR beta/delta-mediated psoriasis-like skin disease in vivo. PLoS ONE, 7, e37097. Collins, C.A. and Watt, F.M. (2008) Dynamic regulation of retinoic acid-binding proteins in developing, adult and neoplastic skin reveals roles for beta-catenin and Notch signalling. Dev. Biol., 324, 55– 67. Lopez, J.M., Fernandez, M.A., Pique, M. and Gil, J. (2004) Aspirin-induced apoptosis in jurkat cells is not mediated by peroxisome proliferator-activated receptor delta. Mol. Cell. Biochem., 266, 57– 63. Barber, S.A., Detore, G., McNally, R. and Vogel, S.N. (1996) Stimulation of the ceramide pathway partially mimics lipopolysaccharide-induced responses in murine peritoneal macrophages. Infect. Immun., 64, 3397–3400. Hvid, M., Johansen, C., Deleuran, B., Kemp, K., Deleuran, M. and Vestergaard, C. (2011) Regulation of caspase 14 expression in keratinocytes by inflammatory cytokines--a possible link between reduced skin barrier function and inflammation? Exp. Dermatol., 20, 633– 636. Chiang, M.F., Yang, S.Y., Lin, I.Y., Hong, J.B., Lin, S.J., Ying, H.Y., Chen, C.M., Wu, S.Y., Liu, F.T. and Lin, K.I. (2013) Inducible deletion of the Blimp-1 gene in adult epidermis causes granulocyte-dominated chronic skin inflammation in mice. Proc. Natl. Acad. Sci. USA, 110, 6476–6481. Oh, S.T., Schramme, A., Tilgen, W., Gutwein, P. and Reichrath, J. (2009) Overexpression of CXCL16 in lesional psoriatic skin. Dermatoendocrinology, 1, 114–118. Tohyama, M., Sayama, K., Komatsuzawa, H., Hanakawa, Y., Shirakata, Y., Dai, X., Yang, L., Tokumaru, S., Nagai, H., Hirakawa, S. et al. (2007) CXCL16 is a novel mediator of the innate immunity of epidermal keratinocytes. Int. Immunol., 19, 1095–1102. Taha, R.A., Minshall, E.M., Leung, D.Y., Boguniewicz, M., Luster, A., Muro, S., Toda, M. and Hamid, Q.A. (2000) Evidence for increased expression of eotaxin and monocyte chemotactic protein-4 in atopic dermatitis. J. Allergy Clin. Immun., 105, 1002–1007.

449