RACK1 Prevents the Premature Differentiation of Epidermal Progenitor Cells by Inhibiting IRF6 Expression

TO THE EDITOR

To properly maintain barrier function, epidermal stem and progenitor cells, which reside in the basal layer of the skin, continually proliferate and differentiate to contribute to the outermost layers. Perturbations in the balance between stem cell self-renewal and differentiation leads to a variety of skin disorders(Blanpain and Fuchs, 2009). Thus, it is critical to understand the factors that govern basal layer cell fate decisions. We and others have identified IRF6, ZNF750, KLF4, and GRHL3 as transcription factors necessary to transition basal layer cells to a differentiated one(Oberbeck et al., 2019, Segre et al., 1999, Sen et al., 2012, Ting et al., 2005). However, less is known about how the levels of these transcription factors are regulated to prevent premature differentiation of stem and progenitor cells. To identify putative factors enriched in epidermal progenitor cells, we mined our previous gene expression profiling data comparing undifferentiated and differentiated primary human keratinocytes(Sen et al., 2010). Interestingly, the receptor for activated C kinase 1 also known as RACK1 (GNB2L1) was highly downregulated upon epidermal differentiation.

RACK1 is a WD40 containing scaffolding protein that was originally identified as an anchoring protein for activated protein kinase C(Ron et al., 1994). As a scaffolding protein, it has been found to be associated with a large variety of proteins to affect a wide range of cellular processes (Li and Xie, 2015). RACK1 knockout mice are embryonic lethal at the gastrulation stage, which has precluded further examination of RACK1 in normal tissue homeostasis(Volta et al., 2013). Currently, it is unclear whether RACK1 has any role in stem cell fate decisions.

To answer this question, we use human skin as a model system. Staining of adult human skin showed that RACK1 is expressed primarily in the basal layer (Figure 1a). Notably, the vast majority of RACK1 did not co-localize with keratin 1 (K1) which is expressed in the differentiated layers of the skin (Figure 1a). In primary human keratinocytes, RACK1 mRNA and protein levels are downregulated upon induction of differentiation with calcium (Figure 1b). To test the function of RACK1, we knocked down RACK1 using two distinct shRNAs (RACK1i-A and RACK1i-B). Significant depletion of RACK1 as compared to scrambled shRNA controls (CTRi) was achieved on both the mRNA and protein levels (Figure 1c). Loss of RACK1 expression resulted in the spontaneous upregulation of epidermal differentiation genes and inhibited cell proliferation without impacting apoptosis (Figure 1d-f). Together, these data suggest that the loss of RACK1 led the cells to prematurely differentiate and exit the cell cycle.

Figure 1. RACK1 is required for maintaining epidermal progenitor cell function.

(a) Immunofluorescent staining of human skin with antibodies against RACK1 (red) and Keratin 1 (K1: green), a marker of the suprabasal layer. Merged imaged includes Hoechst 33342 staining of nuclei. n=3. Scale bar=10um. (b) RACK1 mRNA and protein expression in proliferating (−Ca⁺²) and differentiated (+Ca⁺² day 3) primary human keratinocytes maintained in 2D culture. For RT-qPCR analysis of RACK1 mRNA levels, expression was normalized to L32. n=3. Data represented as mean ± SD. ** p<0.01 vs proliferating (−Ca⁺²) cells (unpaired t-test). For Western blot of RACK1 protein levels, β-actin is shown as loading control. Representative image is shown and n=3. (c) Primary human keratinocytes were knocked down with control (CTRi) or RACK1 (RACK1i) shRNAs and the remaining RACK1 mRNA and protein levels were measured by RT-qPCR and Western blotting respectively. Two separate shRNAs (RACKi-A, RACK1i-B) targeting different regions of RACK1 were used. RACK1 mRNA expression was normalized to L32 levels. n=3. Data represented as mean ± SD. *** p<0.001 vs CTRi (unpaired t-test). For Western blot of RACK1 protein, β-actin is shown as loading control. Representative image is shown and n=3. (d) RT-qPCR analysis of the relative mRNA expression of epidermal differentiation genes in CTRi and RACK1i cells. Cells were cultured in proliferation conditions. Expression of differentiation genes were normalized to L32 levels. n=3. Data represented as mean ± SD. * p<0.05 vs CTRi (unpaired t-test). (e) Proliferation assay of CTRi vs RACK1i cells. Plotted values represent the relative increase in fluorescence at each timepoint relative to starting point (day 0). n=3. Data represented as mean± SD. ** p<0.01 vs CTRi (1-way ANOVA). (f) Percent of apoptotic cells, comparing CTRi and RACKi cells, were analyzed by flow cytometry. Only cells positive for Annexin-V and negative for Propidium Iodide was considered as apoptotic. 6000 cells per measurement. n=3. Data represented as mean ± SD. ns= not significant.

Next, we analyzed the gene expression profile of RACKi in comparison to CTRi keratinocytes grown in proliferation conditions. The loss of RACK1 resulted in the differential expression of 1,221 genes (573 upregulated and 648 downregulated) (Supplementary Figure S1a and Supplementary Table S1). Genes with decreased expression were enriched in gene ontology (GO) terms such as DNA replication and cellular macromolecule biosynthesis (Supplementary Figure S1b). Upregulated genes were enriched in GO terms such as epidermis development and establishment of skin barrier (Supplementary Figure S1b). Premature expression of these differentiation genes may be due to the increased expression of transcription factors such as IRF6, KLF4, GRHL3, and ZNF750 (Supplementary Figure S1c, Supplementary Table S1). ARCHS4 analysis of the 573 genes with increased expression upon RACK1 depletion identified IRF6, GRHL3 and ZNF750 as the top 3 transcription factors to be co-expressed with this expression profile (Supplementary Figure S1d)(Lachmann et al., 2018). To confirm that RACK1 is regulating the transcriptional levels of these proteins (as opposed to post-transcriptional regulation), we quantified the formation of newly synthesized IRF6, GRHL3, KLF4, and ZNF750 RNA in CTRi and RACKi cells by 5-ethynyluridine (EU) labeling. The loss of RACK1 resulted in a significant increase in nascent IRF6, GRHL3, and ZNF750 mRNA (Supplementary Figure S1e). Thus, RACK1 acts as a transcriptional repressor of transcription factors required for epithelial cell differentiation.

We performed double knockdown of RACK1 and IRF6 to determine if premature differentiation due to loss of RACK1 can be rescued. IRF6 was chosen because it is an early epidermal differentiation transcription factor directly downstream of P63(Moretti et al., 2010). Depletion of RACK1 and IRF6 prevented the increase in IRF6 mRNA and protein levels observed with just RACK1 knockdown (Figure 2a-b). Preventing the increase in IRF6 levels in RACK1i keratinocytes also blocked the expression of differentiation genes such as TGM1, DSG1, SPPR1A, LCE3D, FLG, GRHL3, KLF4, and ZNF750 (Figure 2c). To assess the extent of rescue as a result of IRF6 knockdown in RACK1i cells, we assayed 60 additional genes that were differentially expressed upon RACK1 depletion (Supplementary Figure S2 and Supplementary Table S1). Of the 36 genes increased, 92% (33/36) were restored similar to control (CTRi) levels upon RACK1 and IRF6 double knockdown (Supplementary Figure S2a-b). The 3 genes that were still increased in expression included MTRNR2L2, CALB1, and PAK3 (Supplementary Figure S2a-b). Of the 24 genes downregulated upon RACK1 loss, 67% (16/24) were restored to control levels upon double knockdown (Supplementary Figure S2c-d). The genes that were not rescued include MCM7, ASF1B, ETV1, PLPP2, PEG10, PLLP, APLN, and CPVL (Supplementary Figure S2c-d). Overall, 82% (49/60) of the differentially expressed genes were restored similar to control levels upon double IRF6 and RACK1 knockdown. To explore if knockdown of IRF6 and RACK1 will rescue the RACK1i phenotype in a tissue setting, we regenerated human skin for all three groups. We chose an early timepoint (Day 3) of skin regeneration where terminal differentiation proteins filaggrin (FLG) and loricrin (LOR) are not yet expressed in control tissue (Figure 2d). In the absence of RACK1, the skin was prematurely differentiated as FLG and LOR could be readily detected. Depletion of IRF6 and RACK1 simultaneously prevented the expression of the terminal differentiation proteins, suggesting that RACK1 prevents premature differentiation in epidermal progenitor cells by repressing IRF6 expression (Figure 2d).

Figure 2. RACK1 inhibits IRF6-mediated differentiation of epidermal progenitor cells.

(a) RACK1 and IRF6 mRNA expression in primary human keratinocytes transfected with control (CTRi), RACK1 (RACK1i), or RACK1(RACK1i) plus IRF6 (IRF6i) siRNA in 2D culture. Cells were cultured in proliferation conditions. For RT-qPCR analysis, L32 was used for internal normalization. n=3. Data represented as mean ± SD. * p<0.05, ** p<0.01 (unpaired t-test). (b) Western blot of samples described in (a) for RACK1 and IRF6 protein levels. β-actin is shown as loading control. Representative image is shown and n=3. (c) RT-qPCR analysis of the relative mRNA expression of epidermal differentiation genes for samples described in (a). L32 was used as an internal normalization control. n=3. Data represented as mean ± SD. * p<0.05, ** p<0.01, *** p<0.001 vs CTRi (1-way ANOVA). (d) Immunofluorescent staining of late differentiation marker filaggrin (FLG: green) and loricrin (LOR: red) in regenerated human skin (day 3). Nuclei was stained with Hoechst (blue). Scale bar =20um. Quantification of FLG and LOR signal intensity is shown to the right of their respective images. n=3. Data represented as mean ± SD. * p<0.05 vs CTRi (1-way ANOVA).

IRF6 is upregulated during epidermal differentiation by P63 however it is unclear how IRF6 levels are repressed in undifferentiated keratinocytes(Moretti et al., 2010). Here we show that RACK1 suppresses IRF6 expression in epidermal progenitor cells where RACK1 protein levels are the highest. Upon knockdown of RACK1, levels of newly synthesized IRF6 mRNA is significantly increased. Furthermore, RACK1 prevents premature differentiation of epidermal stem and progenitor cells partly by repressing IRF6 expression since loss of IRF6 and RACK1 rescues ~82% of the RACK1i differentially expressed genes that were tested. Since double knockdown of IRF6 and RACK1 doesn’t completely rescue all the genes affected by RACK1 depletion, it is likely that RACK1’s phenotype is due to the premature expression of a combination of transcription factors. These other factors could include GRHL1, OVOL1, KLF4, GRHL3, ZNF750, and DSP which all increase in mRNA expression and are co-expressed with the genes differentially expressed upon RACK1 loss (Supplementary Figure S1c-d, Supplementary Table S1).

In summary, we have found a basal layer enriched protein that is critical for preventing premature differentiation by inhibiting the expression of differentiation promoting transcription factors.

DATA AVAILABILITY

Datasets related to this article can be found at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE181265, hosted at the Gene Expression Omnibus (GEO): GSE181265.