A novel homozygous nonsense mutation in CAST associated with PLACK syndrome

Peeling skin syndrome is a heterogeneous group of rare disorders. Peeling skin, leukonychia, acral punctate keratoses, cheilitis and knuckle pads (PLACK syndrome, OMIM616295) is a newly described form of PSS with an autosomal recessive mode of inheritance. We report a 5.5-year-old boy with features of PLACK syndrome. Additionally, he had mild cerebral atrophy and mild muscle involvements. Whole exome sequencing was performed in genomic DNA of this individual and subsequent analysis revealed a homozygous c.544G > T (p.Glu182*) nonsense mutation in the CAST gene encoding calpastatin. Sanger sequencing confirmed this variant and demonstrated that his affected aunt was also homozygous. Real-time qRT-PCR and immunoblot analysis showed reduced calpastatin expression in skin fibroblasts derived from both affected individuals compared to heterozygous family members. In vitro calpastatin activity assays also showed decreased activity in affected individuals. This study further supports a key role for calpastatin in the tight regulation of proteolytic pathways within the skin.


Introduction
Peeling skin syndrome (PSS; OMIM 270300) is characterized by continuous shedding of the stratum corneum of the epidermis accompanied by diverse clinical findings including erythema, vesicular lesions, or other ectodermal features (Kurban and Azar 1969;Levy and Goldsmith 1982;Hashimoto et al. 2000;Judge et al. 2004). It is classified into two clinical forms; either acral PSS (APSS; OMIM 609796) or generalized PSS (GPSS; OMIM 270300) (Bowden 2011). APSS involves mainly skin fragility of the palmar, plantar and dorsal part of hands and feet. Previously, mutations in genes encoding for transglutaminase (TGM5) and cysteine protease inhibitor cystatin A (CSTA; OMIM 184600), which are known to have Part of this research was presented as a poster presentation at the American Academy of Dermatology Annual Meeting 2018, San Diego and the abstract was published in the Journal of American Academy Dermatology (JAAD) (2018): Volume 79, Issue 3, Supplement 1, Page AB16, DOI: https://doi.org/10.1016/j.jaad.2018.05.108. key roles in cross-linking of cornified cell envelope proteins in epidermis, have been associated with APSS (Cassidy et al. 2005;Kharfi et al. 2009;Krunic et al. 2013). GPSS can present with severe pruritis, atopy, patchy feeling of entire skin, food allergies, repeated episodes of angioedema, urticaria and asthma, which can be further divided into inflammatory or non-inflammatory GPSS. Mutations in corneodesmosin (CDSN; OMIM 602593) and CHST8 (OMIM 610190), encoding N-acetylgalactosamine-4-O-sulfotransferase have been associated with inflammatory GPSS (Oji et al. 2010;Israeli et al. 2011;Cabral et al. 2012).
Recently, a new recessive form of GPSS with peeling skin, leukonychia, acral punctate keratoses, cheilitis and knuckle pads (termed PLACK syndrome) has been described and is associated with the loss of function mutations in the CAST gene (OMIM 114090; Refseq NM_001042440.4) encoding calpastatin Alkhalifah et al. 2017). To date, homozygous mutations of c.607dupA (p.Ile203Asnfs * 8), c.424A>T (p.Lys142 * ) and c.1750delG (p.Val584Trpfs * 37)  and, recently, a homozygous 4-base insertion c.461dupGCAT (p.Ser154Cysfs*6) (Alkhalifah et al. 2017) have been reported in the CAST gene leading to loss of function of calpastatin. The aim of this study is to investigate a family displaying features of PLACK syndrome.

Cases and clinical data
The proband is a 5.5-year old boy who initially presented with skin fragility to the Department of Dermatology Uludag University, Faculty of Medicine, Bursa, Turkey. Clinical assessment identified fragile skin, wooly hair, sparse eyelashes and brows, palmoplantar punctate keratoderma, follicular hyperkeratosis, knuckle pads and cheilitis. Cerebral atrophy and mild muscle involvement were observed on magnetic resonance imaging (MRI) and nerve conduction velocity (NCV)/electromyography (EMG), respectively. Full neurologic assessment by a pediatric neurologist revealed just a slight activation of the deep tendon reflexes. Family members were recruited to the study and their personal medical history was reviewed and physically examined.
This revealed that the proband's aunt had similar clinical cutaneous features but in milder form with in addition nail dystrophy. Though not evaluated by MRI and NCV/EMG due to pregnancy, his aunt's neurological assessment was normal. Pedigree analysis revealed that the condition was segregating as an autosomal recessive trait. Written informed consent for the study was obtained from all participants and blood was collected from the two affected and five unaffected family members.
The study and data collection were in accordance with the principles of the Declaration of Helsinki and the study had the approval of the local ethical committee.

Whole exome sequencing and sanger sequencing
Genomic DNA was isolated from blood samples using Wizard® Genomic DNA Purification Kit, Promega. Whole exome sequencing (WES) and analysis were performed on 3 samples (proband, father and mother) at the core facility of Advanced Genomics and Bioinformatics Research Center (IGBAM, TUBITAK, Gebze, Turkey) using Illumina technology.
DNA samples were prepared for massively parallel sequencing by using an Illumina TruSeq Sample Preparation kit. Exonic regions were captured by a NimbleGen SeqCap EZ Human Exome Library v3.0 Kit. Illumina TruSeq PE Cluster Kit v3-cBot-HS was used for pairedend cluster generation and a TruSeq SBS Kit v3-HS reagent kit was used for sequencing the post-capture libraries. Initial clustering was performed on an Illumina cBot machine. Paired-end sequencing was done on an Illumina HiSeq 2500 system with a read length of PE 2 × 104. All procedures were carried out according to the manufacturer's instructions. Base calling and image analysis were done using Illumina's Real Time Analysis software version 1.13 with default parameters.
Raw sequencing data were aligned to the hg19 reference human genome using BWA (Li and Durbin 2009) with standard parameters in paired-end (PE) mode. SAMtools (Li et al. 2009) was then used to remove PCR duplicates. To calculate the coverage of targeted exome regions, BEDtools (Quinlan and Hall 2010) was used. To perform local realignment around indels, Genome Analysis Toolkit v1.6 (GATK) (DePristo et al. 2011) IndelRealigner was used. Then, SNPs and small indels were called by using GATK UnifiedGenotyper. SnpEff (Cingolani et al. 2012) was used for functional annotation of variants such as gene/exonic region, minor allele frequency, segmental duplications and effect of variants. HomSI (Görmez et al. 2014) was used for homozygosity mapping analysis. Homozygous candidate variants were sifted using FMFilter (Akgün et al. 2016).
The homozygous rare/novel sequence variants identified in CAST and VPS13C were verified by Sanger sequencing in family members using the GenomeLab™ GeXP Genetic Analysis System (Beckman Coulter, Germany). DNASTAR sequence analysis software and sub-components were used for sequence assembly and analysis of Sanger data (DNASTAR Inc., USA). CAST sequencing primers, forward 5′ TTTGGTAGCAAGGG AATTGG and reverse 5′ AGATCAGGCTCTGG AAAGCA; and VPS13C sequencing primers, forward 5′ TTAGGGAACAGCAGAAACTCA and reverse 5′ AACATCCCACTTGATTACGC oligonucleotide primers, were used for PCR amplification following DNA sequencing.

Histogical and immunohistochemical analysis
Hematoxyline-eosine staining was applied to the paraffin-embedded skin tissue sections of the proband (III:7). Immunohistochemical staining was performed with a calpastatin antibody (1:100, sc-376547, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) in the paraffin-embedded skin tissue sections of the proband (III:7) and the control of a skin age and sex matched tissue, available in the Pathology Department of Near East University.

Transmission electron microscopy analysis
Briefly, skin tissue was fixed in 2.5% glutaraldehyde solution in PBS (phosphate-buffered saline) (1×, pH 7.4), pH 7.4, for 4 h and post-fixed for 1 h in 1% osmium tetroxide in 0.1 M 1X-PBS. After washing in 1× PBS, they were dehydrated through ethanol series, treated with propylene oxide and embedded in Araldite/Epon812 (Cat. no. 13940, EMS, Hatfield, PA, USA). After heat polymerization, sections were cut using a microtome. Semi-thin sections were stained with methylene blue-azure II and examined using a light microscope (Leica Microsystems, Germany) with a DC490 digital camera (Leica Microsystems, Germany). Ultrathin sections (Leica ultracut R, Germany) were double-stained with uranyl acetate and lead citrate (Leica EM AC20, Germany). These sections were examined in a JEOL-JEM 1400 (Jeol USA Inc) electron knuckle pads. c Axial and c' sagittal T2-weighted images revealed mild cerebral atrophy and evidence of peripheral CSF areas secondary to atrophy by MRI (indicated by white arrows). d Showing the paternal aunt's (II-3) nail dystrophy, d' follicular hyperkeratosis d" and eroded bullous lesion residuals d"' in her arm microscope and photographed by CCD camera (Gatan Inc., Pleasanton, CA, USA).

Real-time qPCR
Total RNA was extracted from cultured fibroblasts using RNeasy kit (Qiagen, USA). To quantify mRNA expression of CAST gene, qRT-PCR was carried out using following p r i m e r s ( C A S T C o n t F : c a t g a t t t c t g c t g g t g g a g , CASTContR:ccctatgggtttcgaagagtc primer pair that amplify u p s t r e a m o f t h e v a r i a n t r e g i o n , a n d CASTcE182XF:GGACCAGAAGTTTCAGATCCAA, CASTcE182XR:TCCCTGCTGACTGAGCTTTT primer pair that amplify the region that includes c.544G>T (p.Glu182*)) and 1-step QuantiTect SYBR Green qRT-PCR Kit (Qiagen, USA) according to the manufacturer's standard protocol on Light Cycler 480 instrument (Roche, S w i t z e r l a n d ) . T h e h o u s e k e e p i n g g e n e G A P D H (Hs_GAPDH_1_SG QuantiTect Primer Assay, NM_002046, Qiagen) was used for normalization, and relative gene expression levels were calculated using the 2 −ΔΔCt method. Results are shown as fold expression (mean ± SEM, n = 3; bars, SE; *P < .05, **P < .01). Statistical significance was analyzed by using Student's t-tail test on GraphPad Prism 6.2 software.

Calpastatin activity assay
For the calpastatin activity assay, protein concentrations of the fibroblast-derived cell extracts obtained from patients were adjusted to 3 μg/μL. The extracts were heated to 90°C for 5 min and then centrifuged at 16,000g at 4°C for 20 min.
Reactions contained 5 μL of supernatant, 10 nM human erythrocyte calpain 1 (Millipore, USA), 200 μM Suc-Leu-Leu-Val-Tyr-AMC (Abcam, UK) and 195 μL of the homogenization buffer with 5 mM beta-mercaptoethanol. One sample contained no supernatant and was designated as negative control. The reaction was started by adding CaCl 2 to a final

Clinical findings
We report on a 5.5-year old boy who initially presented with generalized peeling skin who was the first-born of IVF twins from a consanguineous marriage. Clinical findings included fragile skin (Fig. 1b), wooly hair, sparse eyelashes and brows, palmoplantar punctate keratoderma, follicular hyperkeratosis, knuckle pads (Fig. 1b') and cheilitis. Mild cerebral atrophy, secondarily to atrophy asimetrically enlargement of lateral ventricles and mild muscle involvement were observed on his MRI (Fig. 1c, c') and NCV/EMG, respectively. His aunt also had similar clinical features but in milder form with, in addition, nail dystrophy ( Fig. 1d-d"').

WES and sanger sequencing
To search for the putative causative mutation or mutations related with proband's phenotype and presumed autosomal recessive inheritance, we evaluated novel and rare homozygous variants in the exome file and identified eight candidate variants ( Table 1). The most plausible variant associated with the cutaneous phenotype segregating in the family was the c.544G>T nonsense variant in CAST. However, the pathogenic class of all eight variants was evaluated according to ACMG (American College of Medical Genetics) criteria in company with standards and guidelines for the interpretation of sequence variants (Richards et al. 2015). KHDRBS1 and TMCC1 variants were silent substitutions that do not change the encoded amino acid and these genes have not been previously related to any OMIM phenotypes. All of the other five missense variants were predicted as not damaging to protein by at least one of three computational algorithms used. In addition, four of these missense genes, except for VPS13C, have not been previously associated with any OMIM phenotypes. The VPS13C variant was predicted as not damaging to protein by SIFT and Provean predictions.

Transmission electron microscopy findings
Semi-thin sections of proband's lesioned skin showed elongated epidermal rete pegs (Fig. 3d-white asterisk) and dermal papilla (Fig. 3d-black asterisks), thick stratum corneum (Fig.  3d-white asterisk), extensive karyolysis even in stratum basale and loss of chromatine pattern in all epithelial cells (Fig. 3f-red  arrows). Electron micrographs showed widened intercellular Fig. 3 Histological, immunohistochemical and transmission electron microscopy results. a Proband's lesional skin biopsy, photomicrographs showing hyperkeratosis (white asterisk), mild spongiotic changes, H&E. b, c Immunohistochemistry shows absent calpastatin staining in the proband (b) compared to normal staining throughout the epidermis in the control (c). d-f Photomicrographs showing elongated epidermal rete pegs (d-white asterisk) and dermal papilla (d-black asterisk), thick stratum corneum (e-black asterisk), extensive karyolysis even in stratum basale and loss of chromatin pattern in all epithelial cells (f-red arrows) and semi-thin sections T&B. g-j Electron micrographs showing widened intercellular space, heterochromatin in the bleb (g-red asterisk, red arrow, respectively), chromatin condensation and marginization (h-white arrows), nuclear chromatin condensation and fragmentation (i-white asterisk) and the loss of the integrity of the nuclear membrane (j-white arrow) space, heterochromatin in the bleb (Fig. 3g-red asterisk, red arrow, respectively), chromatin condensation and marginization (Fig. 3h-white arrows), nuclear chromatin condensation and fragmentation (Fig. 3i-white asterisk) and the loss of the integrity of the nuclear membrane ( Fig. 3j-

Real-time qPCR analysis
In contrast to the carriers, we could detect only trace expression of CAST mRNA in the affected cases, probably due to mechanisms of nonsense mediated mRNA decay (Fig. 4a, a'). Furthermore, we demonstrated a significant difference between the expression of mRNA upstream of the variant region and the variant region in affected cases (both Normal Skin (NS) and Lesioned Skin (LS)) and Mt/N III-2. Similarly, we found that expression of CAST mRNA was significantly different in affected cases compared to Mt/N II-4 (*P < .05) and Mt/N II-5 (**P < .01).

Protein quantification and immunofluorescence
To compare the abundance of calpastatin in fibroblasts of individuals with homozygous and heterozygous genotypes, total protein samples were resolved on SDS-PAGE and immunoblotted using calpastatin and GAPDH-specific antibodies. A semi-quantitative protein expression assay revealed a drastically low amount of calpastatin in patients' normal and lesioned skin cells as compared to cells from the heterozygous unaffected cell line (Fig. 4b). Hence, we detected more than a two-fold decrease in calpastatin expression levels in the patients homozygous for c.544G>T (Mt/Mt III:1 and Mt/Mt II:3) compared with heterozygous carriers (Mt/N II:4 and Mt/N III:2). Confocal microscopy results also confirmed reduced expression pattern of calpastatin in fibroblasts from affected patients compared to carriers (Fig. 4c-c", d-d", e-e").

In vitro calpastatin assay findings
In vitro calpastatin activity of fibroblast cell extracts obtained from patients homozygous for c.544G>T (III:1 and II:3) and unaffected heterozygote cases were evaluated indirectly by fluorogenic calpain proteolysis assay using a Suc-Leu-Leu-Val-Tyr-AMC fluorescence probe. Our results demonstrated increased calpain activity and, thus, decreased calpastatin activity in affected individuals compared with control samples (**P < .01) (Fig. 4f).

Discussion
PLACK syndrome has recently been described as a new form of peeling skin syndrome (PSS), characterized by generalized peeling skin, leukonychia, acral punctate keratoses, cheilitis and knuckle pads . To date, four homozygous pathogenic variants (loss of function mutations) in CAST gene (c.424A>T, c.607dup, c.1750delG, c.461dupGCAT) have been reported with PLACK syndrome Alkhalifah et al. 2017). This study presents the fifth family with PLACK syndrome and is associated with a novel homozygous nonsense mutation (c.544G>T p.Glu182*) in the CAST gene. CAST encodes calpastatin, which is an endogenous specific inhibitor of calpain, a calcium-dependent cysteine protease (Goll et al. 2003). Calpastatin can bind to calcium activated calpains (Potz et al. 2016). Calpains are involved in a range of cellular processes, including cell proliferation (Kovács and Su 2014), migration (Franco and Huttenlocher 2005), wound healing (Nassar et al. 2012), apoptosis and survival (Tan et al. 2006). It is also involved in the regulation of epidermal cell differentiation (Carragher and Frame 2004). A proposed mechanism underlying PLACK syndrome is that the absence of a functional calpastatin prevents the inhibition of calpains leading to elevated keratinocyte apoptosis and skin hyperkeratosis Wang et al. 2015).
Modulation of calpastatin expression in mice caused no phenotypic abnormalities under normal conditions indicating that calpastatin may only function as a negative regulator of calpain only under pathological conditions (Takano et al. 2005). In order to determine the functional consequences of CAST mutations in vitro, Lin et al. performed siRNAmediated knockdown of CAST in the immortalized keratinocyte cell line, HaCaT and performed a mechanical induced stress assay. This showed that mechanical stress revealed breakage of the intercellular connections in CAST knock down cell monolayers in contrast to control cell lines . Calpain overactivity is involved in the pathophysiology of a large number of disease processes of the brain, eyes, heart, lungs, pancreas, kidneys, vascular system and skeletal muscle. The situation of stress leads to increased calcium ion release causing over-activation of calpain, which may cause organ dysfunction by promoting cellular apoptosis and degradation of cytoskeletal structure (Potz et al. 2016). Therefore, calpain was suggested as a potential therapeutic target for a wide variety of disease processes. In line with this, inhibition of calpain by over-expression of calpastatin inhibits the abnormal breakdown of cytoskeletal proteins (spectrin, MAP2 and neurofilaments) and ameliorates motor axon loss in amyotrophic lateral sclerosis (ALS) hSOD1G93A transgenic mouse model (Rao et al. 2016). Another study supports the hypothesis that calpastatin may play a role in regulating the initial metastatic dissemination of breast cancer (Storr et al. 2011).
It is noteworthy that our novel mutation and the associated clinical features differ in that leukonychia was absent although it was reported in all of the previous cases. Mild cerebral atrophy and muscle involvement was also present in one of the affected family members, whereas it was not reported in the study by Lin et al. (2015). However, Alkhalifah et al. (2017) did report muscle involvement in their study. The absence of leukonychia and the presence of mild cerebral atrophy in our patient suggest significant phenotypical variability within individuals with CAST mutations.
In this study, we demonstrated that calpastatin expression in fibroblasts from affected family members was reduced at both the mRNA and protein level. In vitro calpastatin activity revealed reduced calpain proteolysis in both affected individuals. The molecular and biochemical studies in conjunction with the described clinical features in this study emphasize the importance of tight regulation of proteolytic pathways by calpastatin in maintaining the structural integrity of the skin.