Alternative mRNA splicing events and regulators in epidermal differentiation

Abstract

Alternative splicing (AS) of messenger RNAs occurs in ∼95% of multi-exon human genes and generates diverse RNA and protein isoforms. We investigated AS events associated with human epidermal differentiation, a process crucial for skin function. We identified 6,413 AS events, primarily involving cassette exons. We also predicted 34 RNA-binding proteins (RBPs) regulating epidermal AS, including 19 previously undescribed candidate regulators. From these results, we identified FUS as an RBP that regulates the balance between keratinocyte proliferation and differentiation. Additionally, we characterized the function of a cassette exon AS event in MAP3K7, which encodes a kinase involved in cell signaling. We found that a switch from the short to long isoform of MAP3K7, triggered during differentiation, enforces the demarcation between proliferating basal progenitors and overlying differentiated strata. Our findings indicate that AS occurs extensively in the human epidermis and has critical roles in skin homeostasis.

Keywords: CP: Molecular biology; CP: Stem cell research; MAP3K7; alternative splicing; epidermal differentiation; skin; skin homeostasis.

Figure 1.. Alternative mRNA splicing events in human epidermal differentiation

(A) Experimental design to identify mRNA splicing isoforms in progenitor and differentiated epidermal keratinocytes. (B) Number and type of alternative mRNA splicing events in progenitor and differentiated human epidermal keratinocytes. The full list of splicing events is listed in Table S1. (C) RT-PCR of alternatively spliced skipped exon (SE) events identified by rMATS. PSI, percent spliced in, represented by percentage of exon-included isoform. In vitro columns show primary keratinocytes differentiated in cell culture by confluent growth conditions with supplemented calcium (Ca²⁺). 1 and 2 represent biological replicates. In vivo columns show epidermal keratinocytes flow sorted from intact truncal skin of an adult male into beta 4 integrin (ITGB4) high and low populations. KRT1 (keratin 1) is a differentiation-specific transcript control; RPL32 is a ribosomal protein transcript invariant control. (D) Scatterplot showing magnitude of alternative splicing (AS) change (ΔPSI) vs. expression level for epidermal gene transcripts when comparing progenitor vs. differentiated keratinocytes. Blue dots represent transcripts displaying greater than 10% splicing isoform differences (|ΔPSI|>0.1) and <2-fold total gene expression change. (E) Proportion of distinct AS events in epidermal genes.

Figure 2.. Characteristics of cassette exon AS events in epidermal keratinocytes

(A) Heatmap of cassette exon AS events. Identities of representative key epidermal regulator genes are noted. (B) Ontology analysis of genes alternatively spliced between progenitor and differentiated keratinocytes. (C) Proportion of spliced exon events predicted to cause in-frame insertion or frameshift. (D) Overlap of alternatively spliced genes in mouse and human epidermal keratinocytes. The most highly enriched Gene Ontology (GO) term associated with shared vs. organism-specific mouse/human SE genes is shown. (E) Impact of keratinocyte SE events on protein domains. Protein domains affected by 5 or more distinct AS events are shown.

Figure 3.. RNA-binding proteins (RBPs) participating in epidermal AS cassette exon events

(A) Scheme of sequence motif analysis of AS exons in epidermal differentiation. (B) Representative RBPs and their binding motifs enriched at each cassette exon splicing junction. A–D lettering corresponds to the exon-intron junctions shown in (A). The complete list of predicted RBPs and their binding motifs are in Table S2. (C) RNA immunoprecipitation of candidate RBPs and their predicted RNA targets. SRSF1 is a positive control. (D) Quantitative RT-PCR of RNA immunoprecipitation. HPRT is a negative/specificity control RNA. Data are means ± SEM (n = 3). (E) Quantitative RT-PCR and immunoblot of FUS after treatment of keratinocytes with control (siCTRL) or FUS-targeting siRNAs (siFUS). (F) RT-PCR of alternatively spliced SE events associated with FUS depletion. (G) Number and type of alternative mRNA splicing events associated with FUS depletion. (H) Immunofluorescence of control and FUS-depleted epidermal organotypic tissues. Ki-67 marks proliferating keratinocytes (pink, arrowheads); KRT10 is a differentiation-associated protein (orange/red). White asterisks indicate non-specific antibody staining of cornified layer. Scale bars (gray): 50 μm. (I) Quantitative RT-PCR of progenitor-associated (CCNB1, BNC1, CCND1) and differentiation-associated (KRT1, KRT10, FLG) transcripts in epidermal organotypic tissue. Data are means ± SEM (n = 2).

Figure 4.. AS of MAP3K7 exon 12 alters NF-κB activity in keratinocytes

(A) Sashimi plot showing enriched relative expression of exon-12-containing isoforms of MAP3K7 upon epidermal differentiation. (B) Scheme of RT-PCR primers (red triangles) to discern MAP3K7 isoforms containing (MAP3K7-long) or excluding exon 12 (MAP3K7-short). Purple bars represent isoform-targeting siRNAs for short and long MAP3K7 transcripts. Bottom, RT-PCR of in vitro progenitor (P) and differentiated (D) keratinocytes and in vivo laser capture microdissected skin tissue of basal (Bs) progenitor and suprabasal (Sb) differentiated layers. CCND1 and LOR are control transcripts enriched in progenitor and differentiated states, respectively. (C) RT-PCR evaluating MAP3K7 isoform expression after transfection of isoform-targeting siRNAs against short and long isoforms. RPL32 is an invariant expression control. (D) Immunoblot for proteins in the NF-κB signaling pathway in progenitor keratinocytes after treatment with control or MAP3K7-isoform-targeted siRNAs, in the presence or absence of tumor necrosis factor α (TNF-α; 10 ng/mL), to stimulate NF-κB signaling. TNF-α was applied for 30 min, and protein lysates were harvested 48 h later. (E) Intensity quantitation of phospho-p65/RelA from immunoblot experiments. Phosphorylated p65 was normalized internally to total p65, and the unstimulated siCTRL signal was set to 1 for each biological replicate. Data are means ± SEM (n = 3) (one-way ANOVA with a Tukey’s honestly significant difference [HSD] post hoc test), *p < 0.05. (F) RT-PCR and immunoblot for TAK1, the protein product of MAP3K7, after overexpression (OE) of each isoform. (G) ELISA for phosphorylated p65 in keratinocytes after OE of empty vector or MAP3K7 short/long isoforms. Data are means ± SEM (n = 3) (one-way ANOVA with a Tukey’s HSD post hoc test), *p < 0.05.

Figure 5.. AS of MAP3K7 exon 12 regulates epidermal differentiation

(A) Immunoblot for proteins in the NF-κB signaling pathway after treatment with control or MAP3K7-isoform-targeted siRNAs, in the presence of absence of protein kinase C (PKC), to stimulate NF-κB signaling. (B) Immunofluorescence of Ki-67 and KRT10 in epidermal organotypic tissues generated with CTRL, MAP3K7-long, or MAP3K7-short siRNA-treated primary epidermal keratinocytes. Dotted white lines denote the basement membrane. Arrowheads highlight examples of proliferating keratinocytes detected in suprabasal layers. PKC treatment was applied for 2 h on the day prior to endpoint to evaluate the effect of NF-κB activation on the tissue phenotype. Scale bars (gray): 50 μm. (C and D) Quantitation of (C) Ki-67+ cells and (D) KRT10 signal in epidermal organotypic tissues generated with CTRL, MAP3K7-long, or MAP3K7-short siRNAs. Data are means ± SEM (one-way ANOVA with a Tukey’s HSD post hoc test) *p < 0.05. (E) Quantitative RT-PCR of progenitor-associated genes in epidermal organotypic tissues. Relative expression of each gene is shown for each isoform-specific knockdown relative to its expression in control (gray bars, normalized to 1.0). Data are means ± SEM (n = 3). (F) Quantitative RT-PCR of differentiation-associated genes in epidermal organotypic tissues. Data are means ± SEM (n = 3). (G) Working model of MAP3K7 isoform spatial expression within the epidermis and proposed effects of each isoform on NF-κB signaling and epidermal differentiation.

Figure 6.. Highlights of alternatively spliced genes and processes involved in epidermal homeostasis

Schematic of major biological processes and representative alternatively spliced genes in human epidermis.