(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



Apoptosis and tissue thinning contribute to symmetric cell division in the developing mouse epidermis in a nonautonomous way [1]

['Arad Soffer', 'Department Of Cell', 'Developmental Biology', 'Sackler Faculty Of Medicine', 'Tel Aviv University', 'Tel Aviv', 'Adnan Mahly', 'Krishnanand Padmanabhan', 'Jonathan Cohen', 'Orit Adir']

Date: 2022-08 

Mitotic spindle orientation (SO) is a conserved mechanism that governs cell fate and tissue morphogenesis. In the developing epidermis, a balance between self-renewing symmetric divisions and differentiative asymmetric divisions is necessary for normal development. While the cellular machinery that executes SO is well characterized, the extrinsic cues that guide it are poorly understood. Here, we identified the basal cell adhesion molecule (BCAM), a β1 integrin coreceptor, as a novel regulator of epidermal morphogenesis. In utero RNAi-mediated depletion of Bcam in the mouse embryo did not hinder β1 integrin distribution or cell adhesion and polarity. However, Bcam depletion promoted apoptosis, thinning of the epidermis, and symmetric cell division, and the defects were reversed by concomitant overexpression of the apoptosis inhibitor Xiap. Moreover, in mosaic epidermis, depletion of Bcam or Xiap induced symmetric divisions in neighboring wild-type cells. These results identify apoptosis and epidermal architecture as extrinsic cues that guide SO in the developing epidermis.

Funding: This work was supported by the U.S-Israel Binational Science Foundation (BSF grant number 2019230 to C.L. and S.E.W.), the Israel Science Foundation (ISF grant number 1174/20 to C.L.), and the Richard Eimert Research Fund on Solid Tumors of the Faculty of Medicine, Tel Aviv University To (C.L.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Although β1 integrin is a well-studied major regulator of epidermal biology [ 27 – 32 ], the role of BCAM in epidermal development is unknown. Therefore, in the present study, we investigated the roles of BCAM in epidermal development in the mouse. Unexpectedly, we found that BCAM activity is dispensable for β1 integrin distribution, cell adhesion, and BM assembly; however, BCAM was shown to play an essential role in cell survival, tissue architecture, and balanced SO in the epidermis.

Basal cell adhesion molecule (BCAM, also known as Lu/Lutheran blood group) is a member of the immunoglobulin superfamily and functions as a β1 integrin coreceptor by modulating its interaction with the BM protein laminin [ 22 – 24 ]. BCAM is a transmembrane glycoprotein that was first identified in red blood cells; however, it is highly expressed in epithelial cells of most organs [ 22 , 23 , 25 ]. Deletion of Bcam does not hinder mouse viability or fertility; however, in the kidney, the BM was abnormally thick and capillary number was decreased, while in the intestine, BM appeared normal but the smooth muscle coat was abnormally thick and disorganized [ 26 ].

The molecular machinery that executes SO is highly conserved [ 17 , 18 ] and has been extensively characterized in several model systems including the mammalian epidermis. Several proteins have been shown to play crucial roles in SO, including the apical polarity complex Par3–Par6–aPKC, which interacts with LGN–Gα(i)–NuMA–dynein–dynactin, a complex that directs positioning of the spindle [ 10 , 11 , 12 , 16 , 19 , 20 ]. Adhesion proteins also play an important role in SO. The adherens junction proteins vinculin, α-catenin, and afadin are essential for telophase reorientation and oriented cell division fidelity [ 21 ]. Similarly, β1 integrin is also essential for SO, presumably through its function in the maintenance of cell–extracellular matrix adhesion [ 19 ].

Studies in the developing epidermis showed that SO is essential for survival and barrier formation [ 10 – 12 ] and additionally plays a role in cell competition, a selection process that optimizes epidermal development [ 13 ], and the specification of hair follicle stem cells [ 14 ]. In the adult epidermis, SO protects the tissue against oncogene-induced hyperproliferation [ 15 ], regulates hair follicle morphogenesis [ 10 , 14 ], and controls the fate of hair follicle stem cells [ 16 ] and matrix cells [ 10 ].

Spindle orientation (SO) is a highly conserved and tightly regulated process that plays a key role in cell fate determination, tissue morphogenesis, and homeostasis (reviewed in [ 1 – 4 ]). In the developing epidermis, basal layer stem cells/progenitors orient their spindles either parallel or perpendicular to the basement membrane (BM). After parallel (symmetric) division, the 2 daughter cells remain in the basal layer and may proliferate; in contrast, following perpendicular (asymmetric) division, 1 daughter cell remains in the basal layer while the other daughter cell becomes suprabasal and begins to differentiate (reviewed in [ 5 – 9 ]).

Results

BCAM depletion does not alter β1 integrin distribution or activity in the developing mouse epidermis We began our investigation of the potential role of BCAM in β1 integrin-dependent and/or β1 integrin-independent functions in epidermal development by examining BCAM localization in mouse embryos. Immunofluorescent staining of dorsal skin from day 16.5 embryos (E16.5) of wild-type CD1 mice revealed high levels of BCAM throughout the cortex of basal layer cells, similar to what has been reported for β1 integrin expression (Figs 1A and S1). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. Bcam depletion induces apoptosis in the developing epidermis. (A) Sagittal views of 10-μm sections of dorsal skin from E16.5 CD1 mouse embryos immunostained for BCAM (red). (B) RT-PCR analysis of Bcam mRNA in primary mouse keratinocytes transduced with scrambled shRNA (Ctrl) or the 2 Bcam-specific shRNAs 1554 and 1690. Data are the mean ± SD of n = 6 experiments per condition from 3 independent experiments. ***P = 3 × 10−4 for control vs. Bcam-1554 and ***P = 3.2 × 10−4 for control vs. Bcam-1690 by unpaired t test. (C) Western blot analysis of BCAM protein expression in primary mouse keratinocytes transduced with shScr (Ctrl), Bcam-1554, or Bcam-1690 shRNAs. GAPDH was probed as a loading control. (D) Quantification western blot analysis shown in (C). Data are the mean ± SD of n = 3 blots **P = 2 × 10−3 for control vs. Bcam-1554; ***P = 8 × 10−4 for control vs. Bcam-1690 by unpaired t test. (E) Sagittal views of 10-μm sections of dorsal skin from Bcam-1554 KD E16.5 mosaic tissue immunostained for BCAM (red). (F) Sagittal views of 10-μm sections of dorsal skin from control, Bcam-1554 and Xiap-2297 KD E16.5 embryos immunostained for the cell proliferation marker Ki67 (red, upper panel) or pulsed for 2 h with EdU (red, lower panel). (G) Quantification of EdU+ cells from the images shown in (F). Data are the mean ± SD of n = 3 embryos per condition. E15.5, not significant for control vs. Bcam-1554 (P = 0.2754); E16.5, ***P = 8 × 10−4 for control vs. Xiap-2297; not significant for control vs. Bcam-1554 (P = 0.223) by unpaired t test. (H) Sagittal views of 10-μm sections of dorsal skin from control, Bcam-1554, Bcam-1690, and Xiap-2297 KD E16.5 embryos immunostained for active caspase 3 (red). Arrows indicate active caspase 3+ cells. Large boxes/ovals show magnification of fragmented nuclei in active caspase 3+ cells. (I) Quantification of active caspase 3+ cells from the data shown in (H and S3 and S4 Figs). Data are the mean ± SD of n = 4 embryos per condition. E14.5, not significant for control vs. Bcam-1554 (P = 0.0981); E15.5, *P = 1.6 × 10−2 for control vs. Bcam-1554; E16.5, ***P = 2 × 10−4 for control vs. Bcam-1554, ***P = 7 × 10−4 for control vs. Bcam-1690, ***P = 7 × 10−4 for control vs. Xiap-2297 by unpaired t test. (J) shScr- (Ctrl) and shBcam-1554-transduced primary mouse keratinocytes were cultured in low- (50 μM) or high-calcium (1.5 mM) media and then immunolabeled for active caspase 3. (K) Quantification of active caspase 3+ cells from the data shown in (J). Data are the mean ± SD of n = 3 experiments. Low calcium, not significant (P > 0.999); high calcium, **P = 1 × 10−3 for control vs. Bcam-1554 by unpaired t test. (L) RT-PCR analysis of Xiap mRNA in primary mouse keratinocytes transduced with scrambled shRNA (Ctrl) or Xiap-specific shRNA (2297). Data are the mean ± SD of n = 6 experiments per condition from 3 independent experiments. ***P = 3 × 10−3 for control vs. Xiap-2297 by unpaired t test. The data underlying all the charts in the figure are included in S1 Data. Nuclei were stained with DAPI (blue). Dotted lines indicate the dermal–epidermal border, and upper right insets show the transduced cells (H2B-GFP+). Scale bars = 20 μm. BCAM, basal cell adhesion molecule; RT-PCR, reverse transcription PCR. https://doi.org/10.1371/journal.pbio.3001756.g001 To facilitate the analysis of BCAM function, we screened several Bcam-specific short hairpin RNAs (shRNAs) and identified two, Bcam-1554 and Bcam-1690, which depleted Bcam mRNA levels in primary mouse keratinocytes (1°MKs) by 83 ± 7.7% and 96 ± 2.3%, respectively, compared with control scrambled shRNA (shScr) (Fig 1B). Western blot analysis confirmed the mRNA results and showed that BCAM protein levels were similarly depleted in shBcam-expressing compared with control 1°MKs (Fig 1C and 1D). To deplete Bcam during epidermal development, the amniotic sacs of E9 wild-type mouse embryos were injected in utero with lentiviruses encoding Bcam-1554, Bcam-1690, or shScr together with a GFP-tagged histone 2B reporter (H2B-GFP) to identify successfully transduced cells [33]. Immunostaining with BCAM antibody in mosaic tissue confirmed the depletion of BCAM in H2B-GFP+ cells in the dorsal skin of E16.5 embryos (Fig 1E). We first asked whether BCAM depletion affects the localization of β1 integrin in the dorsal skin of E16.5 embryos, but no differences were detected in β1 integrin localization between control and Bcam knockdown (KD) embryos; β1 integrin was present throughout the cortex of basal layer cells with the highest levels observed in the basal region of the cell juxtaposed with the BM (S1 Fig). Moreover, immunostaining with a 9EG7-specific antibody, which recognizes an epitope unique to active β1 integrin [34], revealed comparable distribution and expression levels (fluorescence intensity) in the control and Bcam KD epidermis (S1 Fig). Because β1 integrin is essential for skin BM assembly [35,36] and BCAM itself directly binds to the BM protein laminin α5 [22–24,37], we next examined BM organization by immunostaining the dorsal skin of control and Bcam KD E16.5 embryos for laminin ɣ1 (represented by laminins 511 and 521 in the skin, which both contain α5), laminin 332, and nidogen, all of which are major components of the skin BM [38,39]. Each protein was detected as a thin line between the epidermis and the dermis in both control and Bcam KD epidermis (S1 Fig). Together, these data indicate that BCAM expression is not necessary for the distribution or activity of β1 integrin, or for cell adhesion or BM organization in the developing epidermis.

BCAM depletion does not hinder cell–extracellular matrix adhesion in cultured keratinocytes Focal adhesions are integrin-based structures that mediate cell adhesion to the extracellular matrix [40]. β1 integrin is a major regulator of focal adhesions in many cell types including cultured 1°MKs [28,35]. To examine whether BCAM activity alters β1 integrin levels, activity, and focal adhesion organization, 1°MKs were transduced with lentiviruses encoding shScr (control) or shBcam-1554 and analyzed by confocal microscopy. Immunostaining analysis of β1 integrin and the 9EG7 epitope levels showed that while overall levels of β1 integrin were comparable in control and Bcam-depleted cells, the levels of the 9EG7 epitope increased by 23% in Bcam-depleted cells (S2 Fig). Immunostaining for the focal adhesion protein paxillin [41,42] detected comparable numbers of focal adhesions; however, average focal adhesion area increased by 8.8% in Bcam-depleted cells (S2 Fig). Together, these data indicate that BCAM loss-of-function does not hinder cell–extracellular matrix adhesion in cultured 1°MKs; instead, it results in a modest increase in β1 integrin activity and focal adhesion area.

BCAM depletion induces apoptosis We next determined whether the coreceptor function of BCAM is necessary for β1 integrin contribution to epidermal growth, a process that involves cell proliferation, differentiation, senescence, and death [43,44]. To this end, the dorsal skin of control and Bcam KD E16.5 embryos was immunostained for the cell proliferation marker Ki67, but no differences in the pattern of Ki67+ cells were observed (Fig 1F). To quantify proliferation, we injected embryos with shScr or Bcam-1554 at E9 and then pulsed the pregnant mice on E15.5 and E16.5 for 2 h with the uridine analog EdU (5-ethynyl-2′-deoxyuridine), which incorporates into S-phase cells. Quantification of EdU+ cells in the embryonic dorsal skin sections confirmed a similar level of proliferation in Bcam-depleted compared with control epidermis (E15.5: 33.4 ± 1.2%, control and 32.4 ± 0.6%, Bcam KD; E16.5: 30.6 ± 0.8%, control and 28.6 ± 1.7%, Bcam KD; Fig 1G). We next examined whether BCAM was required for epidermal differentiation by immunostaining dorsal skin sections of E14.5, E15.5, and E16.5 embryos for the epidermal cell markers keratin 14 (K14, basal layer), K10 (suprabasal layers), and loricrin (granular layer) (S3–S5 Figs). These analyses revealed normal differentiation of Bcam-depleted epidermis. Moreover, induction of K10 was detected when Bcam-depleted 1°MKs were induced to differentiate by increasing calcium levels in vitro (S2 Fig). Together, these results indicate that BCAM is not required for normal epidermal differentiation. Although β1 integrin plays a role in cell senescence pathways [45], we did not detect a difference in expression of senescence-associated β-galactosidase [33,46] between the epidermis of control and Bcam KD E16.5 embryos (S5 Fig), suggesting that BCAM does not contribute to this function. In striking contrast, we identified a key role for Bcam in cell apoptosis. Notably, immunostaining of E14.5, E15.5, and E16.5 dorsal skin sections for the active, cleaved form of the proapoptotic enzyme caspase 3 revealed very few positive cells in control epidermis (<1 positive cell per dorsal skin section). However, a progressive increase in active caspase 3+ cells was observed in Bcam KD epidermis beginning at E15.5, culminating in an approximately 8-fold increase by E16.5 (Figs 1H, 1I, S3, and S4). Since caspase 3 can have nonapoptotic roles in the skin [47], we confirmed that Bcam-depleted, active caspase 3+ cells were apoptotic by their nuclear condensation and fragmentation, a classical hallmark of an apoptotic cell [48] (Figs 1H, S3, and S4). Moreover, apoptosis was detected only in infected (GFP+) basal layer cells, confirming this effect on apoptosis is cell autonomous, and is restricted to the proliferative compartments of the epidermis. To determine whether the defect in apoptosis can be recapitulated in tissue culture, 1°MKs were transduced with lentiviruses encoding shScr (control) or shBcam-1554, cultured in low- or high-calcium media (50 μM and 1.5 mM, respectively), and immunostained for active caspase 3. In low-calcium media in which keratinocytes cannot form cell–cell junctions and cannot differentiate, very few control or Bcam-depleted cells were caspase 3+. However, in high-calcium media, conditions that allow cell–cell adhesion and induce keratinocytes differentiation [49], we detected a 3.5-fold increase in caspase 3+ in Bcam-depleted cells (Fig 1J and 1K). To verify that these findings are indeed indicative of apoptosis in the Bcam-depleted epidermis, we performed the same analyses in embryos depleted of x-linked inhibitor of apoptosis (XIAP), which has been shown to function to suppress apoptosis in hair follicle stem cells [50]. E9 embryos were injected in utero with XIAP-targeted shRNA (shXiap-2297), which we confirmed could effectively deplete Xiap mRNA levels (90 ± 12.7% in 1°MKs; Fig 1L), and dorsal skin sections of E16.5 embryos were immunostained for active caspase 3 (Fig 1H). As expected, Xiap depletion increased the number of active caspase 3+ cells in the epidermis, consistent with an increase in apoptosis; in fact, the number of apoptotic cells in Xiap-depleted epidermis was remarkably similar to that observed in Bcam-depleted epidermis (Fig 1I). However, unlike Bcam-depleted embryos, E16.5 Xiap-depleted epidermis exhibited an approximately 30% increase in EdU incorporation compared with the control epidermis (30.6 ± 0. 8% versus 39.1 ± 1.2%), suggesting that while both BCAM and XIAP have antiapoptotic functions, they have distinct effects on epidermal growth (Fig 1G). Taken together, these data demonstrate that BCAM is not required for cell proliferation, differentiation, or senescence, but unexpectedly, it is required for cell survival and is a potent inhibitor of apoptosis in the developing epidermis and in cultured keratinocytes.

BCAM is not involved in cell–cell adhesion or apicobasal polarity in the epidermis Our results thus far identify BCAM’s ability to impact SO. Because cell–cell adhesion and apicobasal polarity have been shown to play important roles in the regulation of SO in the epidermis [16,19,21,53,54], we next investigated whether BCAM expression affects these processes. To this end, we immunostained for the cell adhesion proteins E-cadherin and α-catenin, the apicobasal protein Par3, and the centrosomal protein pericentrin in E16.5 control and Bcam KD embryos. However, the localization and intensity of staining of all 4 proteins were comparable between the embryos (S5 Fig). Moreover, normal cell–cell adhesion was detected in E14.5, and E15.5 Bcam-depleted embryos (S3 and S4 Figs), and in calcium-shifted cultured keratinocytes (S2 Fig), indicating that BCAM expression is not required for normal epidermal cell–cell adhesion and apicobasal polarity. Similarly, upon examination of basal layer cell shape and mitotic rounding, which regulate SO in the epidermis [55–57], we detected no differences in interphase and mitotic basal layer cell shape in Bcam-depleted and control epidermis sections (S7 Fig), indicating that BCAM is unlikely to be involved in these processes.

BCAM involvement reveals a link between apoptosis and spindle orientation/symmetric cell division Having shown that BCAM inhibits apoptosis and is also required for normal SO and balanced symmetric/asymmetric cell division in the developing epidermis, we next asked whether apoptosis and SO might be linked. Given that overexpression of XIAP is known to inhibit apoptosis [58–60], we generated shScr;GFP-Xiap (XIAP overexpression) and shBcam-1554;GFP-Xiap (Bcam KD and XIAP overexpression) viruses and confirmed the ability of the shBcam-1554;GFP-Xiap virus to deplete Bcam with concomitant overexpress of Xiap in 1°MKs (Fig 3A). Next, we transduced E9 embryos with the shScr;GFP-Xiap virus and confirmed that XIAP overexpression does not alter the overall ratio between parallel and perpendicular SO (0°–45° and 45°–90°, respectively) (S6 Fig). In agreement with previous reports showing that Xiap overexpression can suppress apoptosis [58–60], GFP-Xiap overexpression reduced the elevated apoptosis observed in Bcam-depleted embryos to control levels (Fig 3B and 3C. Data from 3C is from Fig 1). Next, we examined SO in the same Bcam KD; Xiap overexpressing embryos by staining dorsal skin sections for survivin. Whereas Bcam depletion alone led to an increase in parallel divisions (Fig 2C), this increase was suppressed by Xiap overexpression (Fig 3D and 3E). Moreover, while the epidermal thickness of Bcam KD; Xiap overexpressing embryos was thinner than control embryos, it increased by approximately 20% compared to shBcam-transduced embryos (Fig 3F). Together, these data suggest that apoptosis and epidermal architecture influence SO in the developing epidermis. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Apoptosis enhances symmetric cell division. (A) RT-PCR analysis of Bcam and Xiap mRNA in primary mouse keratinocytes transduced with shScr; H2B-GFP (Ctrl) or shBcam-1554; GFP-Xiap. Data are the mean ± SD of n = 6 experiments per condition from 3 independent experiments. ****P < 10−4 for Bcam levels in control vs. Bcam-1554 and **P = 7.8 × 10−3 for GFP-Xiap levels in control vs. shBcam-1554;GFP-Xiap by unpaired t test. (B) Sagittal view of a 10-μm section of dorsal skin from a shBcam-1554;GFP-Xiap-transduced E16.5 embryo immunostained for active caspase 3. (C) Quantification of active caspase 3+ cells shown in (B). Quantification of active caspase 3+ cells in Ctrl, shBcam-1554-, shBcam-1554-, shXiap-2279- is shown in Fig 1. Data are the mean ± SD of n = 4 embryos per condition. Not significant (P = 0.7228) for control vs. shBcam-1554;GFP-Xiap by unpaired t test. (D) Sagittal views of 10-μm sections of dorsal skin from shBcam-1554;GFP-Xiap E16.5 embryos immunostained for the cleavage furrow marker survivin (red). White ovals show magnifications of survivin-positive, late-mitotic cells. Quantification of SO is presented to the right of each image. (E) Same data as in (D), plotted as a cumulative frequency distribution. Not significant (P = 0.3364) by Kolmogorov–Smirnov test. (F) Quantification of epidermal thickness from the data shown in (B) and (G). Quantification of epidermal thickness in Ctrl, shBcam-1554-, shBcam-1554, and Xiap-2297 is shown in Fig 2. Data are the mean ± SD of n = 40 microscopic fields from 4 embryos. ****P < 0.0001 for shBcam-1554 vs. shBcam-1554;GFP-Xiap, **P = 1 × 10−3 for control vs. shBcam-1554;GFP-Xiap; not significant (P = 0.5242) for shBcam-1554 mosaic tissue vs. shBcam-1554, ****P < 0.0001 for shBcam-1554 mosaic tissue vs. control (G) Sagittal views of 10-μm sections of dorsal skin from shScr-, shBcam-1554;GFP-Xiap-, shBcam-1554-, shBcam-1690-, and shXiap-2297-transduced E16.5 embryos immunostained for the cleavage furrow marker survivin (red). H2B-GFP+ (green) and H2B-GFP− cells denote infected and uninfected cells, respectively. White circles indicate survivin-positive, late-mitotic, uninfected cells. Quantification of SO is presented to the right of each image. (H) Same data as in (G), plotted as a cumulative frequency distribution. E16.5, *P = 1.66 × 10−2 for control vs. Bcam-1554, *P = 2.7 × 10−2 for control vs. Bcam-1690, *P = 1.91 × 10−2 for control vs. Xiap-2297, and not significant for control vs. shBcam-1554;GFP-Xiap (P = 0.8455) by Kolmogorov–Smirnov test. The data underlying all the charts in the figure are included in S1 Data. Nuclei were stained with DAPI (blue). Dotted lines indicate the dermal–epidermal border. Scale bars = 20 μm. RT-PCR, reverse transcription PCR; SO, spindle orientation. https://doi.org/10.1371/journal.pbio.3001756.g003

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001756

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/