Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 13 (3), 203-14

Developmental Roles for Srf, Cortical Cytoskeleton and Cell Shape in Epidermal Spindle Orientation

Affiliations

Developmental Roles for Srf, Cortical Cytoskeleton and Cell Shape in Epidermal Spindle Orientation

Chen Luxenburg et al. Nat Cell Biol.

Abstract

During development, a polarized epidermal sheet undergoes stratification and differentiation to produce the skin barrier. Through mechanisms that are poorly understood, the process involves actin dynamics, spindle reorientation and Notch signalling. To elucidate how epidermal embryogenesis is governed, we conditionally targeted serum response factor (Srf), a transcription factor that is essential for epidermal differentiation. Unexpectedly, previously ascribed causative defects are not responsible for profoundly perturbed embryonic epidermis. Seeking the mechanism for this, we identified actins and their regulators that were downregulated after ablation. Without Srf, cells exhibit a diminished cortical network and in mitosis, they fail to round up, features we recapitulate with low-dose actin inhibitors in vivo and shRNA-knockdown in vitro. Altered concomitantly are phosphorylated ERM and cortical myosin-IIA, shown in vitro to establish a rigid cortical actomyosin network and elicit critical shape changes. We provide a link between these features and Srf loss, and we show that the process is physiologically relevant in skin, as reflected by defects in spindle orientation, asymmetric cell divisions, stratification and differentiation.

Figures

Figure 1
Figure 1
Srf is expressed in embryonic basal cells and when absent, morphological abnormalities originate at the basal to suprabasal juncture. (a) Anti-Srf immunohistochemistry of skins of Srf-cKO and wild-type littermate embryos at embryonic ages indicated. Nuclear Srf is lost in some Srf-cKO basal cells by E15.5 (arrows), but in nearly all basal cells by E16.5, when Srf is strongly expressed in the wild-type basal layer. Scale bar,100 μm. (b) Corresponding transmission electron micrographs of ultrathin sections of back-skin epidermis. Arrows in E16.5-cKO skin denote cells with basal morphology that seem to be suprabasal. Arrows in E17.5- and E18.5-cKO skin indicate vacuoles not seen in normal epidermis. Note also that the basal–suprabasal demarcation is completely disrupted by E17.5. Der, dermis; BL, basal layer; Sp, spinous layer; Gr, granular layer; SC, stratum corneum. Dotted lines denote dermal–epidermal boundary. Scale bars,10 μm.
Figure 2
Figure 2
Alterations in biochemical markers of the basal–suprabasal switch are obvious by E17.5 in Srf -null epidermis. Immunofluorescence microscopy and/or immunoblot analyses of Srf-cKO and wild-type skins at ages indicated. Colour-coding is according to secondary antibodies used. (a,b) Basal markers β4-integrin and K14 and spinous layer differentiation marker K10 reveal disruption in the basal–suprabasal switch by E17.5. (c) K6 is known to be induced in spinous cells whenever there is an imbalance in epidermal homeostasis, as revealed here at E17.5 in Srf -null epidermis. (d) Immunoblot of total epidermal lysates from newborn epidermis probed with the same antibodies as in ac and also hypoxanthine phosphoribosyl-transferase (HPRT, control). (e,f) Notch signalling, another sign of the basal–suprabasal switch, is also markedly perturbed in E17.5 Srf-cKO skin, as evidenced by diminished Notch3 and a Notch target Hes1. Dotted lines denote dermal–epidermal boundary. Scale bars,20 μm. Uncropped images of blots are shown in Supplementary Fig. S9.
Figure 3
Figure 3
Cell adhesion remains intact during embryonic development in Srf-cKO skin. (a) Despite alterations in epidermal architecture in the absence of Srf, E-cadherin immunofluorescence was localized at sites of cell–cell contacts throughout development. Scale bar,20 μm. (b) Immunoblot analyses of total epidermal lysates from newborn pups. (c,d) Ultrastructural analyses of newborn epidermis reveals no major differences in cellular junctions between wild-type and Srf-cKO. c shows hemidesmosomes and d shows desmosomes, both readily identified by their electron density and their attachment to keratin filament bundles (arrows). The basement membrane (BM) is the electron dense line directly beneath the epidermis. BL, basal layer. Scale bar,500 nm. (e,f) Immunofluorescence microscopy of wild-type and Srf-cKO skins, labelled for ABP markers. In wild-type basal cells, Par3 is enriched at the apical cortex and the centrosomal marker Pericentrin (PC) is localized apically. As judged by these markers and in agreement with ultrastructure, Srf-cKO-mediated perturbations in ABP were first evident at E17.5, although many cells still showed normal distribution of these markers (arrowheads). Dotted lines denote dermal–epidermal boundary. Scale bar,20 μm. Uncropped images of blots are shown in Supplementary Fig. S9.
Figure 4
Figure 4
Conditional Srf ablation results in an early reduction in cortical actin, which seems to be directly responsible for an associated ability of early mitotic cells to round up during mitosis. (a) Verification of the array results on actin genes from Table 1. qPCR of mRNAs are from FACS-purified E16.5 basal progenitors. n=2. (b,c) Confocal microscopy and fluorescence intensity quantifications of phalloidin-stained E16.5 skin sections; warmer-coloured pixels indicate higher fluorescence intensity. Basal cell cortical F-actin intensity is significantly decreased in Srf-cKO versus wild-type counterparts (* indicates P=0.013,n=3). (d,e) Whole-mount fluorescence microscopy of E16.5 embryos labelled for F-actin and DAPI and imaged in the plane of the basal layer. Note: cKO image was exposed longer than wild type, to visualize cortical actin and cell shape. Quantifications are of the axial ratios of interphase versus early mitotic (prophase to metaphase) cells, which are reduced in wild-type, but not in Srf-cKO, embryos (* indicates P< 0.05,n=3; wild-type interphase versus early mitotic cells, P =0.004; wild-type versus Srf-cKO early mitotic cells, P =0.02). (fj) F-actin inhibitor defects are similar to Srf loss of function. Although exposure of mouse embryos to 2.5 μM latrunculin for 1 h disrupted ABP, exposure to 0.25 μM did not (fh). Under these milder conditions, early mitotic basal cells failed to round up properly, as imaged by whole-mount fluorescence microscopy (i) and as quantified by measuring the axial ratios of interphase versus early mitotic (prophase to metaphase) cells (j) (* indicates P <0.05,n=3; DMSO (control) interphase versus early mitotic cells, P =0.0003; DMSO versus 0.25 μM latrunculin in early mitotic cells, P =0.008). Dotted line denotes dermal–epidermal border. Arrow denotes early mitotic cell. Error bars represent s.d. Scale bar,20 μm.
Figure 5
Figure 5
Srf-cKO early mitotic cells are defective in enriching and activating ERM proteins at the actin cortex and in recruiting myosin-IIA. (a,b) Immunofluorescence microscopy of E16.5 epidermis reveals an enrichment of cortically localized ERM in the basal layer and activation (phosphorylation) of ERM (pERM) in early mitotic cells. Both are perturbed in Srf-cKO skin. Dotted lines denote dermal–epidermal border. Thin line denotes boundary of mitotic cell at right. Scale bars,20 μm. (c) Quantifications reveal a threefold decrease in pERM (* indicates P =0.0007,n=3) at the cortex of early mitotic basal cells. (d) Immunoblot analyses=of total epidermal lysates. HPRT, hypoxanthine phosphoribosyl-transferase loading control. Note that despite reduction in ERM at basal cortex, overall levels of ERM seem to be similar. Note also that ERM activation is clearly reduced following Srf deficiency. (e) Treatment with 0.25 μM latrunculin achieves a similar effect to Srf deficiency with respect to pERM enrichment at the cortex of early mitotic cells (* indicates P =0.002,n=3). (f,g) Planar images are through the basal layer of E16.5 embryos, subjected to whole-mount fluorescence microscopy for F-actin (phalloidin), myosin-IIA and DAPI. Arrows indicate early mitotic cells, enriched in myosin-IIA in wild type, but not Srf-cKO. Insets, reconstructed XZ view of the compressed Z -stack images to show that myosin-IIA is enriched throughout the cortex of wild-type, but not cKO, mitotic cells. Scale bars,10 μm. (g) Quantifications of fluorescence intensity ratios of F-actin and myosin-IIA from mitotic versus non-mitotic cells. * indicates P =0.0066,n=3. (h) Immunoblot analyses of total newborn epidermal lysates. Blots were probed with antibodies against myosin-IIA, pan-myosin heavy chain (MHC), myosin light chain (MLC) and HPRT. (i,j) E14.5 embryos were treated with DMSO or 0.25 μM latrunculin and stained for F-actin, myosin-IIA and DAPI. Shown are planar views of basal layer and quantifications of fluorescence intensity ratios of F-actin and myosin-IIA in mitotic versus neighbouring non-mitotic cells. indicates P =0.008,n=3. Scale bar,10 μm error bars represent s.d. Uncropped images of blots are shown in Supplementary Fig. S9.
Figure 6
Figure 6
Actin and actin regulators downregulated in Srf-deficient basal cells are required for keratinocytes to undergo proper cortical and cell shape changes during mitosis. Epidermal monolayers of cultured wild-type keratinocytes were treated with either DMSO, 0.25 μM latrunculin for 30 min or lentiviruses harbouring Actb, Wdr 1 or scramble shRNAs, as described in the text. Following treatment with calcium to promote adhesion and stratification and nocodazole to arrest cells in mitosis, cultures were then fixed and labelled for F-actin (red), pERM or myosin-IIA (MIIA, green) and DAPI to mark chromatin (blue). (a) DMSO- and latrunculin-treated cultures; note cortical enrichment of pERM and myosin-IIA in DMSO-treated mitotic cells, but the diminished distribution of pERM and myosin-IIA in latrunculin-treated mitotic cells. Arrows denote mitotic cells. (b,c) Morphometric quantifications of mitotic cells revealed normal total area (b), but increased axial ratio (c) in latrunculin-treated cells (* denotes P =0.0005). (d) Real-time PCR to measure mRNA levels of Actb, Actg1 and Wdr1 in the lentiviral-transduced cultures indicated (KD, knockdown; scr, scramble). (e) Immunoblots were probed with antibodies against β-actin, Wdr1 and α-tubulin. (f) Cell morphology, pERM, cortical actin and/or myosin-IIA are perturbed in keratinocytes expressing Actb or Wdr 1 shRNAs, but not scramble control. White lines denote cell borders. (g) Cortical fluorescence intensity quantifications in Actb-knockdown cells reveal significant decreases in cortical actin and myosin-IIA intensities (* denotes P <0.05; F-actin intensity in scramble versus Actb knockdown, P=9.4×10−7; myosin-IIA intensity in scramble versus Actb knockdown, P = 9.03×10−10). (h) Cortical fluorescence intensity quantifications in Wdr 1-knockdown cells reveal significant decreases in pERM intensity (* denotes P = 1.69×10−14). (i) Quantifications of total mitotic cell areas reveal significant increases in both Actb- and Wdr 1-knockdown mitotic cells (* denotes P <0.05; scramble versus actb knockdown, P = 1.07×10−6; scramble versus Wdr 1 knockdown P=6.1×10−7). (i) Quantifications of mitotic cell axial ratios reveal a significant increase on both Actb- and Wdr 1-knockdown (* denotes P < 0.05; scramble versus Actb knockdown, P = 0.0003; scramble versus Wdr 1 knockdown, P =0.024). Scale bars,20 μm for all quantifications n ≥ 50. error bars represent s.d. Uncropped images of blots are shown in Supplementary Fig. S9.
Figure 7
Figure 7
Mitotic abnormalities in spindle orientation and LGN–NuMA localization in E16.5 Srf-cKO basal cells. (a) Representative immunofluorescence micrographs depicting seemingly normal spindles from early mitotic cells of E16.5 embryos. Anti-acetylated tubulin was used to visualize the spindle and DAPI to mark chromatin. Insets show reconstructed XZ view. Scale bar,2 μm. (b) DAPI staining highlights anaphase/telophase chromatin and reveals the angle between daughter nuclei in mitotic basal keratinocytes of E16.5 embryos. To the right of each representative image is a schematic showing the quantifications of spindle plane (n, number of anaphase/telophase cells analysed), measured relative to the underlying basement membrane (dotted line) delineating the epidermal-dermal boundary. Scale bar,20 μm. (c) qPCR analysis for mRNA levels of G-protein signalling modulator 2 (Gpsm2 ; LGN) and Numa1 (NuMA). (d,e) Quantifications of mitotic E16.5 basal cells revealing similar percentages of early mitotic cells (pHH3+) in which LGN can be detected by immunolabelling (d), but perturbations in the normal apical cortical localization of LGN in mitotic cells (e; * indicates P =0 001,n =3). (f) Representative immunofluorescence micrographs of mitotic basal cells labelled for LGN and NuMA. Scale bar,20 μm. (g) LGN mislocalization is concomitant with pERM and myosin-IIA abnormalities, but precedes aberrations in apical Par3. Colour-coding is according to secondary antibodies used. Scale bar,20 μm. (h) A summary of the temporal progression of the Srf-cKO phenotype. (i) Working model consistent with the data presented. BL, basal layer; Sp, spinous layer.

Similar articles

See all similar articles

Cited by 67 articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback