CASZ1 Regulates Late Epidermal Differentiation

The epidermis is a constantly renewing stratified epithelium that provides crucial barrier functions against insults and dehydration. During epidermal development and tissue homeostasis, a monolayer of keratinocyte progenitors located at the base of the epidermis differentiate and give rise to a stratified epidermis. Disrupted skin barrier due to impaired epidermal differentiation characterizes common skin diseases, such as psoriasis, atopic dermatitis, and squamous cell carcinoma (Lopez-Pajares et al., 2013Lopez-Pajares V. Yan K. Zarnegar B.J. Jameson K.L. Khavari P.A. Genetic pathways in disorders of epidermal differentiation.Trends Genet. 2013; 29: 31-40Google Scholar). Thus, it is essential to uncover the regulatory network controlling epidermal differentiation. While work from labs in the field have identified IRF6, GRHL3, NOTCH1, MAF/MAFB, PRDM1, OVOL1, ZNF750, KLF3 and KLF4, as essential transcription factors (TFs) involved in this process (Jones et al., 2020Jones J. Chen Y. Tiwari M. Li J. Ling J. Sen G.L. KLF3 Mediates Epidermal Differentiation through the Epigenomic Writer CBP.iScience. 2020; 23101320Google Scholar, Lin et al., 2020Lin Z. Jin S. Chen J. Li Z. Lin Z. Tang L. et al.Murine interfollicular epidermal differentiation is gradualistic with GRHL3 controlling progression from stem to transition cell states.Nat Commun. 2020; 11: 5434Google Scholar, Lopez-Pajares et al., 2013Lopez-Pajares V. Yan K. Zarnegar B.J. Jameson K.L. Khavari P.A. Genetic pathways in disorders of epidermal differentiation.Trends Genet. 2013; 29: 31-40Google Scholar, Magnusdottir et al., 2007Magnusdottir E. Kalachikov S. Mizukoshi K. Savitsky D. Ishida-Yamamoto A. Panteleyev A.A. et al.Epidermal terminal differentiation depends on B lymphocyte-induced maturation protein-1.Proc Natl Acad Sci U S A. 2007; 104: 14988-14993Google Scholar, Schwartz et al., 2024Schwartz B. Levi H. Menon G. Maria R. Upcher A. Kotlovski Y. et al.ZNF750 Regulates Skin Barrier Function by Driving Cornified Envelope and Lipid Processing Pathways.J Invest Dermatol. 2024; 144: 296-306 e3Google Scholar), the full transcriptional regulatory network governing epidermal differentiation is yet to be uncovered. In search for regulators of epidermal differentiation, we analyzed published RNA sequencing (RNA-seq) data comparing undifferentiated primary human keratinocytes (KCs) with differentiated KCs (Kretz et al., 2013Kretz M. Siprashvili Z. Chu C. Webster D.E. Zehnder A. Qu K. et al.Control of somatic tissue differentiation by the long non-coding RNA TINCR.Nature. 2013; 493: 231-235Google Scholar). Interestingly, the expression of Castor zinc finger 1 (CASZ1) was highly induced during KC differentiation, alongside other TFs known to regulate KC differentiation (Figure 1a). In order to validate this observation, we placed primary normal human epidermal KCs in differentiation conditions for six days and analyzed CASZ1 expression using reverse transcription followed by quantitative PCR (RT-qPCR), which showed a significant increased expression as KC differentiation progressed (Figure 1b). In line, immunofluorescence staining of normal human skin further demonstrated predominant CASZ1 expression in the upper suprabasal layers of the epidermis (Figure 1c). CASZ1, also known as ZNF693, is a C2H2-type zinc finger TF known to be involved in vast developmental processes, including neural and cardiovascular tissue development and disease (Liu et al., 2023Liu T. Li T. Ke S. Role of the CASZ1 transcription factor in tissue development and disease.Eur J Med Res. 2023; 28: 562Google Scholar). Interestingly, mutation frequencies in the CASZ1 gene were recently reported to differ significantly between subgroups of psoriasis patients with different medical history (Xing et al., 2020Xing J. Zhao X. Li X. Wang Y. Li J. Hou R. et al.Variation at ACOT12 and CT62 locus represents susceptibility to psoriasis in Han population.Mol Genet Genomic Med. 2020; 8e1098Google Scholar), which together with the existing gap of knowledge further prompted us to uncover whether CASZ1 plays a role in epidermal differentiation. To address this question, we used short hairpin RNA (shRNA) sequences to silence the expression of CASZ1 (CASZ1i) in primary normal human epidermal KCs, and determined the effect of CASZ1 silencing on epidermal differentiation gene expression. KCs were differentiated for six days and efficient silencing of CASZ1 (CASZ1i), as compared with that of control shRNA scrambled sequence (CTLi), was confirmed using RT-qPCR (Supplementary Figure S1a). Next, samples were subjected to RNA-seq and differential gene expression analysis. We observed a total of 1,715 differentially expressed genes (DEGs) in CASZ1i versus CTLi differentiated KCs, from which 997 genes showed decreased expression and 718 genes showed increased expression (Figure 1d and Supplementary Table S1). Of the 997 DEGs that demonstrated decreased expression in CASZ1i, 72% overlapped with DEGs upregulated during KC differentiation (Figure 1e). In line, Gene Ontology (GO) analysis of the 997 genes downregulated due to CASZ1 silencing demonstrated significant enrichment in epidermal differentiation-related GO terms, such as KC differentiation, keratinization and programmed cell death (Figure 1f). Together, these data suggest that CASZ1 is necessary to induce the expression of late epidermal differentiation genes. To further understand the regulatory role played by CASZ1 in late epidermal differentiation, we mapped the genome-wide binding of CASZ1 in differentiated KCs on day 6 of differentiation using cleavage under targets & release using nuclease (CUT&RUN) analysis followed by high-throughput sequencing (Supplementary Table S2). CASZ1 binding was mostly enriched at gene bodies and promoter regions of genes that were co-marked by the active histone marks H3K4me4 and H3K27ac (Figure 2a, Supplementary Figure S2a-c, and Supplementary Table S3), suggesting that CASZ1 regulates the expression of active epidermal genes. Moreover, testing for overlap between RNA-seq and CUT&RUN data sets further demonstrated a significant enrichment between CASZ1 binding and DEGs downregulated in CASZ1i, with genomic binding of CASZ1 at about 60% of the 997 DEGs downregulated in CASZ1i versus CTLi in differentiated KCs (Figure 2b). CASZ1-dependent genes included epidermal differentiation TFs such as KLF3 and PRDM1 (Jones et al., 2020Jones J. Chen Y. Tiwari M. Li J. Ling J. Sen G.L. KLF3 Mediates Epidermal Differentiation through the Epigenomic Writer CBP.iScience. 2020; 23101320Google Scholar, Magnusdottir et al., 2007Magnusdottir E. Kalachikov S. Mizukoshi K. Savitsky D. Ishida-Yamamoto A. Panteleyev A.A. et al.Epidermal terminal differentiation depends on B lymphocyte-induced maturation protein-1.Proc Natl Acad Sci U S A. 2007; 104: 14988-14993Google Scholar), as well as epidermal keratinization genes such as late cornified envelope (LCE) family members LCE3D and LCE3E (Jonca and Simon, 2023Jonca N. Simon M. The Cornified Envelope: A Versatile Contributor to the Epidermal Barrier.J Invest Dermatol. 2023; 143: 1335-1337Google Scholar), all of which are known to play key roles in the process of epidermal differentiation and the establishment of barrier functions (Figure 2c and Supplementary Table S2). We validated our RNA-seq data for several epidermal differentiation TFs and keratinization genes using RT-qPCR, and confirmed their significant reduction in CASZ1i versus CTLi in differentiated KCs (Figure 2d). Interestingly. among CASZ1-dependent targets we identified several well-known differentiation TFs including KLF3, KLF4, OVOL1, and PRDM1 (Jones et al., 2020Jones J. Chen Y. Tiwari M. Li J. Ling J. Sen G.L. KLF3 Mediates Epidermal Differentiation through the Epigenomic Writer CBP.iScience. 2020; 23101320Google Scholar, Lopez-Pajares et al., 2013Lopez-Pajares V. Yan K. Zarnegar B.J. Jameson K.L. Khavari P.A. Genetic pathways in disorders of epidermal differentiation.Trends Genet. 2013; 29: 31-40Google Scholar, Magnusdottir et al., 2007Magnusdottir E. Kalachikov S. Mizukoshi K. Savitsky D. Ishida-Yamamoto A. Panteleyev A.A. et al.Epidermal terminal differentiation depends on B lymphocyte-induced maturation protein-1.Proc Natl Acad Sci U S A. 2007; 104: 14988-14993Google Scholar), suggesting for a central role played by CASZ1 in coordinating the epidermal differentiation program. Altogether, here we demonstrated that CASZ1 both binds to and is necessary for the expression of a large subset of epidermal differentiation genes. Notably, a parallel study in differentiating KCs by the group of Xiaomin Bao have identified CASZ1 as a p63/p300-dependent factor that regulates epidermal differentiation genes (Droll et al., 2024Droll S.H. Zhang B.J. Levine M.C. Xue C. Ho P.J. Bao X. CASZ1 is Essential for Skin Epidermal Terminal Differentiation.J Invest Dermatol. 2024; Google Scholar). Together these studies elucidate the role played by CASZ1 in epidermal KCs, demonstrating an essential role for CASZ1 in promoting late epidermal differentiation gene expression programs. Further studies are expected to shed light on the potential roles of CASZ1 in common skin diseases characterized by impaired epidermal differentiation. RNA-seq and CUT&RUN datasets related to this article can be found at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE261573, hosted at the Gene Expression Omnibus (GEO) datasets with accession number GSE261573. Authors state no conflict of interest. For help, advice, and critical suggestions, we thank Ran Taube, Dan Levy, Ramon Birnbaum, Deyou Zheng, and Roi Gazit. LO-R is supported by a returning scientist scholarship from the Israeli Ministry of Aliya and Integration. This work was supported by a grant to IC from the Israel Science Foundation (grant number 1908/21). AUTHOR CONTRIBUTIONS Conceptualization: LO-R, IC; Formal Analysis: LO-R, NTM, HL, LL; Funding Acquisition: IC; Visualization: LO-R, NTM; Writing - Original Draft Preparation: LO-R, IC. Writing - Review and Editing: LO-R, IC. Download .xlsx (1.68 MB) Help with xlsx files Cell cultures Primary normal human epidermal keratinocytes (KCs) were purchased from Lonza (Walkersville, MD), and grown in KC growth medium (KGM-Gold) containing 0.1 mM CaCl2 supplemented with growth factor bullet kit (Lonza). Medium was changed every 2–3 days. For differentiation, KCs were cultured in KGM-Gold growth media supplemented with 1.4 mM CaCl2. HEK293T cells were cultured in DMEM, supplemented with 10% heat-inactivated fetal bovine serum and 1% pen/strep/nystatin solution (Sartorius, Israel). shRNA-mediated gene silencing To knockdown the expression of CASZ1 in KCs, lentiviruses expressing short hairpin RNA against CASZ1 were used. shRNA lentiviral vectors were generated by cloning oligos into the pGreenPuro vector (System Biosciences, CA). The sequence for shRNA targeting CASZ1i-(shRNA #1) and CASZ1i-(shRNA #2) is 5'-GAACGGGAGCACCTACAAGAA-3' and 5'-ACGTGATGACCCACGAGAACT-3', respectively. The scrambled control shRNA sequence is 5'-CAACAAGATGAAGAGCACCAA-3'. The lentiviral constructs were then co-transfected with helper plasmids HIV gag/pol, pRev, pTat and VSV-G, into HEK293T cells by calcium phosphate. Viral supernatant was harvested 48 hours post transfection, filtered through a 0.45μm filter, and used to infect KCs. Cells were incubated in the viral supernatant, supplemented with 8 μg/ml polybrene, and centrifuged at 200 g for 1 hour at 32oC. Cells were then washed twice with Dulbecco's Phosphate Buffered Saline (PBS, Sartorius Israel) and grown in KGM-Gold. After 24 hours the infection process was repeated once more, and after 48 hours selection was performed in the presence of puromycin (2 μg/ml) for 48 hours. RNA purification, quantitative reverse transcription PCR (RT-qPCR) and RNA-seq library preparation Total RNA was purified using TriRNA Pure Kit (Geneaid Biotech, New Taipei City, Taiwan) according to manufacturer's instructions. Complementary DNA (cDNA) was reverse-transcribed from total RNA using the AzuraQuant cDNA Synthesis Kit (Azura Genetics, Raynham, MA). Samples were analyzed by RT-qPCR using LightCycler 480 SYBR Green I Master Mix (Roche Diagnostics, Basel, Switzerland) on a Lightcycler 480 instrument (Roche Diagnostics). Results were normalized to GAPDH mRNA levels. Primer sequences are available in Supplemental Table S4. Prior to library construction, RNA quality was measured using an Agilent Bioanalyzer, and only samples with RNA integrity numbers greater than eight were used. One hundred nanograms of total RNA per sample was used. Enrichment for poly(A) mRNA was done using the NEBNext Poly(A) mRNA Magnetic Isolation Module (New England Biolabs, Ipswich, MA). The RNA-seq libraries were then constructed following standard protocol for the NEBNext Ultra II RNA Library Prep Kit for Illumina (New England Biolabs). Constructed RNA-seq libraries were sequenced at the Technion Genome Center on an Illumina NextSeq 2000 platform, obtaining 100-nt single-end reads. CUT&RUN Normal human primary epidermal KCs were differentiated for six days, harvested, and resuspended in nuclear extraction (NE) buffer (20mM HEPES pH=7.9, 10mM KCl, 0.1% Triton X-100, 20% Glycerol, 0.5mM Spermidine) supplemented with cOmplete EDTA-free protease inhibitor cocktail (Roche) for 10 minutes on ice. Nuclei were then spun at 600 g for 3 minutes at 4oC, resuspended in fresh NE buffer and subjected to cleavage under targets & release using nuclease (CUT&RUN) analysis using the CUTANA ChIC/CUT&RUN kit (EpiCypher, Durham, NC). Nuclei were attached to concanavalin A beads, permeabilized with digitonin, and aliquots of 5 x 105 nuclei were incubated with anti-CASZ1 (Rockland Immunochemicals, Limerick, PA; #600-401-B62; 1:100), anti-H3K4me3 (Cell Signaling Technology, Boston, MA; #9751S; 1:100), anti-H3K27ac (Cell Signaling Technology; #8173S ; 1:100), and anti-IgG (EpiCypher; 1:100) on nutator overnight at 4oC. The next day, nuclei were washed twice with Digitonin buffer (EpiCypher), incubated with pAG-MNase (EpiCypher) for 10 minutes at room temperature, and washed two more times. Next, MNase was activated with 1mM CaCl2 for 2 hours at 4oC. The reaction was stopped by adding Stop buffer supplemented with Escherichia coli spike-in DNA (EpiCypher). Libraries were constructed following a standard protocol for the CUTANA CUT&RUN Library Prep Kit (EpiCypher). Constructed CUT&RUN libraries were sequenced on an Illumina NovaSeq X platform, obtaining 150-nt paired-end reads. RNA-seq data analysis The analysis of the raw sequence reads was carried out using the NeatSeq-Flow platform (2018). The sequences were quality trimmed and filtered using Trim Galore (v0.4.5) and cutadapt (v1.15). Alignment of the reads to the human genome (GRCh38) was done using STAR (v2.5.2a)(Dobin et al., 2013). The number of reads per gene per sample was counted using RSEM (v1.2.28)(Li and Dewey, 2011). Quality assessment (QA) of the process was carried out using FASTQC (v0.11.8) and MultiQC (v1.0.dev0)(Ewels et al., 2016). Transcripts per million reads mapped (TPM) values were estimated independently and genes with low expression (mean TPM < 1 in both groups under comparison) were excluded from differential expression analysis. We analyzed read counts for differential gene expression using the DESeq2 R package (Love et al., 2014). For CASZ1i versus CTLi, genes with fold change ≥ 1.8 and adjusted p-value < 0.05 were considered as significantly differentially expressed genes (DEGs). To determine DEGs during epidermal differentiation, we analyzed publicly available RNA-seq data with Gene Expression Omnibus accession number GSE35468 (Kretz et al., 2013) and genes with fold change ≥ 2.0 and adjusted p-value < 0.05 were considered as DEGs. For visualization of RNA signals after z-scoring their variance-stabilized expression values, we used the Heatmapper web tool (https:// www.heatmapper.ca/)(Babicki et al., 2016). Enrichment for gene ontology biological processes was performed using the DAVID tool (Huang da et al., 2009, Sherman et al., 2022). CUT&RUN data analysis The analysis was carried out using the NeatSeq-Flow platform (2018). Raw sequencing data in fastq format were trimmed by Trim Galore v0.4.5 (length= 25, q- 25), reads that were too short were discarded. FastQC v0.12.0 was used for reads quality control. BWA Mapper (version 0.7.12, default parameters t=20, B= 5) (Li and Durbin, 2009) was used to align paired-end reads to reference human genome assembly hg38 and to the spike-in control (Escherichia coli) reference genome assembly. SAMtools (Danecek et al., 2021) was used to remove PCR duplicates and created coordinate-sorted BAM files. Normalization between samples and IgG was performed with a custom script that calculates read scaling factor based on spike-in DNA reads of each sample. The scaling factor was then used in the bamCoverage to normalize each sample according to its spike-in. Bigwig files were created by bamCoverage function from deepTools v1.5.91 (Ramirez et al., 2016) and reads mapped to blacklisted areas designated by ENCODE (Amemiya et al., 2019) were filtered. Reads that refer to off-chromosome locations were removed with BedClip v377 and sorted by Bedtools v2.30.0 (Quinlan and Hall, 2010). Bigwig files were then created with bedGraphToBigWig by deepTools (Ramirez et al., 2016). For each CASZ1 replicate, peaks were called by both MACS (version 3.0.0) (Zhang et al., 2008) ("callpeak" function, options "nomodel", "broadpeak", broad-cutoff = 0.1) and SEACR v1.3 (using the top 1 percentile as the high-confidence peak calibration in "stringent" mode) (Meers et al., 2019). Bedtools (version 2.31.0) (Quinlan and Hall, 2010) was used to intersect the lists of peaks generated by MACS and by SEACR, maintaining only peaks identified by both peak callers. Peaks within 100 bp were then merged and sorted with Bedtools. The union of peaks from CASZ1 replicates were then intersected by Bedtools resulting in a peak list that included only peaks that exist in both replicates. Peaks were viewed in IGV (version 2.16.1) (Robinson et al., 2011) and annotated with Homer v4.11 (Heinz et al., 2010). For each H3K27ac replicate, peaks were called by SEACR v1.3 (using the top 1 percentile as the high-confidence peak calibration in "stringent" mode) (Meers et al., 2019). Peaks within 100 bp were then merged and sorted by Bedtools and unified into a list that included only peaks that exist in both replicates. Peaks were viewed in IGV (version 2.16.1) (Robinson et al., 2011) and annotated with Homer v4.11 (Heinz et al., 2010). Matrices were generated with DeepTools createMatrix and Heatmaps were plotted with deepTools plotHeatmap (Ramirez et al., 2016). Immunofluorescence staining and microscopy Slides containing normal human skin tissue sections were obtained from US Biomax (Tissue Array, Derwood, MD). Slides were deparaffinized in xylene, rehydrated in decreasing concentrations of ethanol and treated with 3% H2O2 in 1xPBS for 10 minutes. Slides were then subjected to antigen retrieval with 10mM citrate buffer pH = 6.0, and then incubated in IF blocking solution (1xPBS supplemented with 1% BSA, 0.01% gelatin, 0.25% normal goat serum, and 0.1% Triton X-100) overnight at 4°C. Primary antibodies against CASZ1 (Abcam, ab244412; 1:250), Keratin 14 (BioLegend, 906001; 1:1000), and E-cadherin (Invitrogen, 13-1900; 1:1000), were diluted in IF blocking solution and incubated for 1 hour at room temperature. The slides were then washed three times with PBST (1x PBS supplemented with 0.1% Triton X-100) for 5 minutes each time, and incubated with fluorescent secondary antibodies (Jackson ImmunoResearch, West Grove, PA) diluted 1:1000 in IF blocking solution for 1 hour at room temperature. To visualize nuclei, slides were then washed with PBST and were counterstained with DAPI 1:1000 in 1xPBS for 5 minutes. Slides were imaged using an Olympus BX53 upright microscope (Olympus, Tokyo, Japan) with x20 objective. Statistics To determine the statistical significance between two groups, an unpaired two-tailed t-test was performed. All data in bar graphs are presented as mean ± SD. The number of biological replicates used for comparison is indicated in each figure. Significance levels were defined as P < 0.05 (∗), P < 0.01 (∗∗), P < 0.001 (∗∗∗), or non-significant (n.s). For statistical analyses, GraphPad Prism software was used. Sample size is indicated in figure legends, and statistical methods are indicated in the quantification and statistical analysis paragraph. To calculate the statistical significance of the overlap between two groups of genes, we used nemates.org (http://nemates.org/MA/progs/overlap_stats.html). SUPPLEMENTARY REFERENCES NeatSeq-Flow: A Lightweight High-Throughput Sequencing Workflow Platform for Non-Programmers and Programmers Alike. bioRxiv 2018. Amemiya HM, Kundaje A, Boyle AP. The ENCODE Blacklist: Identification of Problematic Regions of the Genome. Sci Rep 2019;9(1):9354. Babicki S, Arndt D, Marcu A, Liang Y, Grant JR, Maciejewski A, et al. Heatmapper: web-enabled heat mapping for all. Nucleic Acids Res 2016;44(W1):W147-W153. Danecek P, Bonfield JK, Liddle J, Marshall J, Ohan V, Pollard MO, et al. Twelve years of SAMtools and BCFtools. Gigascience 2021;10(2). Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 2013;29(1):15-21. Ewels P, Magnusson M, Lundin S, Kaller M. MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics 2016;32(19):3047-3048. Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol Cell 2010;38(4):576-589. Huang da et al., 2009 Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009;4(1):44-57. Kretz et al., 2013 Kretz M, Siprashvili Z, Chu C, Webster DE, Zehnder A, Qu K, et al. Control of somatic tissue differentiation by the long non-coding RNA TINCR. Nature 2013;493(7431):231-235. Li B, Dewey CN. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics 2011;12:323. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 2009;25(14):1754-1760. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 2014;15(12):550. Meers MP, Tenenbaum D, Henikoff S. Peak calling by Sparse Enrichment Analysis for CUT&RUN chromatin profiling. Epigenetics Chromatin 2019;12(1):42. Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 2010;26(6):841-842. Ramirez F, Ryan DP, Gruning B, Bhardwaj V, Kilpert F, Richter AS, et al. deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res 2016;44(W1):W160-W165. Robinson JT, Thorvaldsdottir H, Winckler W, Guttman M, Lander ES, Getz G, et al. Integrative genomics viewer. Nat Biotechnol 2011;29(1):24-26. Sherman BT, Hao M, Qiu J, Jiao X, Baseler MW, Lane HC, et al. DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update). Nucleic Acids Res 2022;50(W1):W216-W221. Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol 2008;9(9):R137.Supplementary Figure S2. CUT&RUN analysis of CASZ1 genomic binding in differentiated keratinocytes (a) Heatmaps showing CUT&RUN read densities for the two biological replicates of CASZ1, and IgG across CASZ1 genomic peaks in differentiated KCs. The profile of CASZ1 binding is displayed in a -/+ 3kb window from CASZ1 peak center. (b) Graph showing CUT&RUN read density depths for the two biological replicates of CASZ1, and IgG across CASZ1 genomic peaks in differentiated KCs. (c) Pie chart showing the genomic localization of the 22,369 CASZ1-bound peaks identified by CUT&RUN analysis in differentiated KCs. TSS, transcription termination site.