Brg1 Determines Urothelial Cell Fate during Ureter Development

Abstract

Developing and adult ureters express the epigenetic regulator Brg1, but the role of Brg1 in ureter development is not well understood. We conditionally ablated Brg1 in the developing ureter using Hoxb7-Cre and found that Brg1 expression is upstream of p63, Pparγ, and sonic hedgehog (Shh) expression in the ureteral epithelium. In addition, epithelial stratification in the basal cells required Brg1-dependent p63 expression, whereas terminal differentiation of the umbrella cells required Brg1-dependent Pparγ expression. Furthermore, the loss of ureteric Brg1 resulted in failure of Shh expression, which correlated with reduced smooth muscle cell development and hydroureter. Taken together, we conclude that Brg1 expression unifies three aspects of ureter development: maintenance of the basal cell population, guidance for terminal differentiation of urothelial cells, and proper investment of ureteral smooth muscle cells.

Development of the metanephric kidney begins with the initiation of the ureteric bud (UB), an outgrowth from the caudal Wolffian duct. The proximal UB branches into the future collecting duct system of the kidney, whereas the distal UB forms the ureter.¹ The early ureter is an epithelial monolayer that differentiates into a stratified urothelium consisting of basal cells sitting on the urothelial basement membrane and umbrella cells at the luminal interface with intermediate cells lying in between. Ureter development can therefore be separated into three major events: proliferation and maintenance of a urothelial stem or progenitor cell population, differentiation of daughter cells to form the stratified urothelium, and investment of ureteral smooth muscle cells surrounding the urothelial basement membrane. In many stratified epithelia, including the bladder, basal cells can self-renew and thus serve as progenitors for the more apical cell layers.^2,3

Expression of sonic hedgehog (Shh) by the basal cells has been shown to activate Shh receptor patched 1 (Ptch1) on the adjacent stromal cells and induce smooth muscle differentiation and investment of the ureter.^4,5 In contrast, the exact identity of the urothelial stem cell and factors that control urothelial stratification and differentiation are less well understood. In vitro studies suggest that the nuclear receptor Pparγ can induce uroplakin expression and thus promote urothelial differentiation.^6,7 Even less is known about the ureteral urothelial stem cell. Shin et al. demonstrated that bladder urothelial basal cells, which are derived from endoderm, are the stem cells for the bladder epithelium.³ Even though ureteral urothelial basal cells are derived from the intermediate mesoderm, their functional and structural similarity to bladder basal cells makes it possible that they serve a similar function as stem cells for the ureteric epithelium.

Mounting evidence suggests that the Switch/Sucrose nonfermentable (Swi/Snf) complex is essential for maintenance of stem cell pluripotency and self-renewal in many organs during development.^8–13 The Swi/Snf complex is composed of multiple proteins that serve to regulate protein-nucleosome contacts and thus RNA polymerase II DNA binding and gene transcription. This multimeric complex includes tissue-specific factors that provide gene specificity along with a central Swi/Snf family protein that provides ATPase and DNA helicase activities. Brg1 is one of the prototypic members of the Swi/Snf family, and global loss of Brg1 results in early embryonic lethality due to failure of embryo implantation.¹⁴

Using conditional ureteral ablation of floxed Brg1 via the Hoxb7-Cre, we show that loss of Brg1 leads to diminished basal cell numbers and failure of normal stratification. We provide evidence that p63, a functionally distinct homolog of p53, is highly expressed by urothelial basal cells in a Brg1-dependent manner, and is required for maintenance of the basal cell compartment. Pparγ expression is also reduced in Brg1 null ureters with accompanying loss of umbrella cell differentiation. Finally, urothelial Shh expression is decreased, which is accompanied by reduced smooth muscle investment of the ureter. These changes lead to massive hydroureter in the newborn mice and demonstrate that the Brg1 containing Swi/Snf complex plays indispensible roles in normal ureter development by orchestrating basal cell maintenance, terminal differentiation, and smooth muscle development.

Results

Ablation of Brg1 Leads to Ureter Malformation

Brg1 is highly expressed in the epithelial cell layer of the developing ureter (Figure 1A). To study its role in ureter development, Brg1^fl/fl mice were bred with Hoxb7-Cre mice to generate Hoxb7-Cre⁺;Brg1^fl/fl mice with specific loss of Brg1 in UB derived structures (Figure 1B).^14,15 Whereas no macroscopic ureteral abnormalities were identified until E15.5, approximately 50% of Hoxb7-Cre⁺;Brg1^fl/fl mutants develop hydroureter at E16.5, with 100% penetrance within 2–3 weeks after birth (Figure 1C and Supplemental Table 1). The null mice progressively lose renal parenchyma, develop renal failure as judged by elevated blood urea nitrogen (BUN) levels and die before weaning (data not shown). Blue dye injected into the renal pelvis of P1 Hoxb7-Cre⁺;Brg1^fl/fl pups moves antegrade into the bladder, while dye injected into the bladder does not reflux into the ureter or kidney at 150 cm hydrostatic pressure (Supplemental Figure 1). Histologic examination of the Brg1 null ureter reveals failure of urothelial stratification and diminished smooth muscle investment detectable at E16.5 and persisting until birth (Figure 1D). These data suggest that loss of urothelial Brg1 leads to hydroureter due to failure of normal ureteral development rather than complete obstruction or vesicoureteral reflux.

Figure 1.

Conditional ablation of Brg1 leads to hydroureter. (A) Fluorescent immunostaining analyses of Brg1 and p63 on cross-sections of wild-type ureters from five different ages are shown. All pictures are taken under the same exposure scheme. (B) Double immunostaining for Brg1 and GFP (expression from the Hoxb7-Cre-IRES-GFP) in ureteral cross-sections of E16.5 embryos. Arrow in h points to a cell that is GFP⁻/Brg1⁺. (C) Photograph demonstrating the normal (a and c) and hydroureter (b and d) in E16.5 embryos (a and b) and P1 pups (c and d). Arrow in b points to the initial site of ureteral dilation. Bladders in c and d are filled with urine before tissue dissection. (D) Histology of proximal ureters of E16.5 and P1 controls and mutants. Whereas control urothelium has multiple cell layers, the mutant urothelium remains as a monolayer (b, e, and f); d and f are close-up view of boxed areas in c and e, respectively. con, control; Brg1 cKO, conditional knockout of Brg1 using the Hoxb7-Cre-IRES-GFP mouse; UE, ureteral epithelial (urothelium); SMC, smooth muscle cell; GFP, green fluorescent protein. Original magnification, ×400 in A and B.

Brg1 Is Required for p63 Expression and Subsequent Stratification of Ureteral Basal Cells

p63, a homolog of p53, has been found to be important for regenerative potential in stratified epithelium, including the urothelium.^16–21 p63 has two major isoforms, TAp63 and ΔNp63.²² The predominant isoform of p63, ΔNp63, is detectable in the normal urothelium on E13.5 just before initial urothelial stratification, and increases through E16.5 with highest expression maintained in the basal cell layer (Figure 1A). However, in ureters from Hoxb7-Cre⁺;Brg1^fl/fl mice, ΔNp63 mRNA expression in the ureter is significantly reduced at E13.5 when no obvious ureteral phenotype can be detected either macroscopically or histologically (Figure 2A and data not shown). Immunofluorescence reveals a marked reduction in ΔNp63 protein at E14.5 (Figure 2B) and E16.5 (Figure 2C, quantified in Figure 2D) in Hoxb7-Cre⁺;Brg1^fl/fl mice that persists throughout life (data not shown). In contrast, mRNA expression of the TAp63 isoform of p63 is extremely low in both normal and Brg1 null ureters (Supplemental Figure 2), with protein expression being undetectable (data not shown).

Figure 2.

p63 is downstream of Brg1 and plays indispensable roles in ureter development. (A) Quantification of p63 expression in E13.5 ureters. (B) p63 protein is almost not detectable in Brg1 null E14.5 ureters. (C) Analysis of the mutant phenotype by fluorescent immunostaining on E16.5 proximal ureters using anti-E-cadherin and anti-p63 antibodies. All anti-p63 immunostaining images presented in this article are acquired using the antibody that detects the ∆Np63 isoform. (D) Quantification of p63⁺ basal cells per ureter cross-section (n=7 and n=9 for con and Brg1 cKO, respectively). (E) Histologic comparison of wild-type, heterozygote, and homozygous GFP knock-in (p63 KO) of E16.5 ∆Np63 GFP knock-in embryonic ureters. Whereas the controls (wt and het) have formed multiple urothelial cell layers, the KO contains a mostly monolayer urothelium (arrow) with extensive folding. The green asterisk points to the smooth muscle layer. (F) Expression of Brg1 is not affected by loss of ∆Np63. The GFP signal comes from the knock-in construct, which labels the urothelial cells affected by the knock-in construct. *** P<0.001. (G) Identification of proliferative cells in urothelium by Ki-67. White dashed line on the Ki-67 images demarcates the boundary of urothelium with its outside mesenchymal compartment. GFP signal comes from the knock-in construct. *P<0.05; ^#P<0.001. con, control; Brg1 cKO, conditional knockout of Brg1 using the Hoxb7-Cre-IRES-GFP mouse; GFP, green fluorescent protein; wt, wild-type; het, heterozygote. Original magnification, ×400.

Because ΔNp63 expression was detected on E13.5 before stratification on E14.5 in normal ureters, and the loss of Brg1 led to a significant reduction of both ΔNp63 expression and urothelial cell stratification, ΔNp63^gfp/gfp mice that selectively lack the ΔNp63 isoform in all cells were analyzed for ureteral stratification defects. These mice are known to exhibit severe epithelial abnormalities in other organs including a hypoplastic and disorganized epidermis.²¹ Histologic analysis at E16.5 reveals that the ureters from ΔNp63^gfp/gfp mice fail to undergo normal epithelial stratification compared with ureters from wild-type and heterozygous littermates, with many areas of the urothelium remaining as a monolayer (Figure 2E). This failure of stratification occurred even though Brg1 was normally expressed in the ΔNp63^gfp/gfp ureters, placing ΔNp63 expression downstream of Brg1 in the developing urothelium (Figure 2F). Consistent with a role for p63 in the proliferative self-renewal of epithelial progenitors,²⁰ the urothelium of ΔNp63^gfp/gfp mice exhibits less proliferation than that seen in wild-type controls (Figure 2G).

Brg1 Null Ureters Have Reduced Shh Expression and Diminished Smooth Muscle Development

Although the defect in urothelial stratification in ΔNp63^gfp/gfp mice was similar to that seen in Hoxb7-Cre⁺;Brg1^fl/fl mice, the smooth muscle layer is relatively intact in ΔNp63^gfp/gfp mice (Figure 2E) and these mice do not exhibit detectable hydroureter up until the time of death (P1). Yu et al. reported that conditional knockout of urothelial Shh using Hoxb7-Cre impairs the development of ureteral smooth muscle cells and leads to hydroureter.²³ Consistent with this, the expression of Shh is markedly reduced in the urothelium of Brg1 null ureters at E16.5 (Figure 3A), correlating with the near absence of the α-smooth muscle actin (α-Sma)–positive smooth muscle layer in these ureters by both immunofluorescence (Figure 3B) and quantitative PCR (Figure 3C). Quantitative PCR analysis of other genes involved in the Shh signaling pathway revealed no change in expression of the transcription factor Tshz3 in Brg1 null ureters, whereas myocardin, the transcriptional coactivator upstream of Sma expression, was downregulated in the absence of Brg1 (Figure 3C). In contrast, loss of ΔNp63 expression does not detectably alter Shh expression (Figure 3D) or development of the α-Sma–positive layer of smooth muscle cells in the ureter (Figure 3E).

Figure 3.

Shh is downstream of Brg1 and not affected by loss of p63 in developing ureter. (A) Analysis of Shh expression in E16.5 control and Brg1 null ureters. (B) Detection of SMA expression in E16.5 control and Brg1 null ureters. (C) Quantitative analysis of Shh, Tshz3, MyoCD, and αSma expressions in E16.5 ureters. n=6 each. (D) Shh expression in E16.5 control and p63 KO ureters. (E) SMA expression in E16.5 control and p63 KO ureters. *P<0.05. MyoCD, myocardin; NS, not significant. Original magnification, ×400.

Brg1 Is Required for Urothelial Pparγ Expression and Terminal Differentiation

In addition to failure of normal stratification, Brg1 null ureters also demonstrate a marked reduction in expression of uroplakins and cytokeratin 20 (Krt20), markers of the differentiated umbrella cell layer, whereas levels of E-cadherin remain unchanged (Figure 4A; E16.5). Previous in vitro studies have shown that uroplakin expression depends on nuclear activation of Pparγ, and that Pparγ is a direct upstream regulator of uroplakins.^6,7,24 Pparγ can be first detected in the urothelial monolayer at E13.5 by immunostaining and is upregulated in the more apical cells during stratification between E14.5 and E16.5 (Supplemental Figure 3). Analysis of Pparγ expression in Brg1 null ureters revealed that Pparg mRNA levels are reduced as early as E13.5 (Figure 4B), and that the Pparγ protein was markedly reduced in the urothelium at E14.5 (Figure 4C). The mRNA and protein levels of Pparγ remain low at E16.5 (Figure 4, D and E) even though E-cadherin expression is unchanged (Figure 4, A and C).

Figure 4.

Pparγ is downstream of Brg1 and required for the expression of uroplakins. (A) Expression levels of uroplakins and cytokeratin 20 (Krt20) are significantly reduced in E16.5 Brg1 null ureters, whereas changes in E-cadherin levels were not significant. n=5 and n=3 for con and Brg1 cKO, respectively. (B) Quantification of Pparg expression in E13.5 ureters. n=8 each. (C) Analysis of Pparγ protein expression in E14.5 control and Brg1 null ureters. (D) Quantification of Pparg mRNA levels in E16.5 ureters. n=9 and n=7 for con and Brg1 cKO, respectively. (E) Analysis of Pparg expression in E16.5 ureters. *P<0.05; ***P<0.001; ^#P<0.005. con, control; Brg1 cKO, Brg1 conditional knockout. Original magnification, ×400.

To define the in vivo relationship between Pparγ and uroplakin expression, we performed conditional Pparg knockout by crossing the Hoxb7-Cre mouse with the Pparg floxed mouse.^15,25 The vast majority of the Hoxb7-Cre⁺;Pparg^fl/fl mice were fertile and healthy up to 26 weeks. Of a total of 65 Hoxb7-Cre;Pparg^fl/fl adult mice, 5 (7.7%) developed hydroureter, whereas the remainder exhibited abnormal infoldings of the basal cell layer creating a starfish pattern of the urothelium. No abnormalities were found in 22 littermate control mice (Figure 5A and Supplemental Table 2).

Figure 5.

PPARγ is essential for terminal differentiation of urothelial cells in vivo. (A) Histology of P1 (a and b) and adult ureters (c–e). Green asterisks in controls (a and c) point to the pink staining of uroplakins plaques, whereas the arrow in d points to the absence of plaque staining in the knockout ureters (Pparg cKO). (B) Fluorescent immunostaining of Hoxb7-Cre⁺;Pparγ^fl/fl P1 ureters: Pparg and p63 (a–h), Shh and Upk2 (i–p), and αSma and p63 (q–x). Dashed white lines in i and red lines in j demarcate the umbrella (apical) cells with lowest Shh expression in i and strongest uroplakin 2 in j, in sharp contrast to that in m and n. (C) Counting of p63+ basal cells per P1 ureter cross-sections. n=10 and n=17 for con and Pparg cKO, respectively. (D) BrdU labeling of P1 ureter. White arrows in e–h point to one example of the p63⁺/BrdU⁺ basal cells. (E) Manual counting of proliferative basal cells. n=13 and n=14 for con and Pparg cKO, respectively. (F) Identification of Brg1 and ∆Np63 expression in control (con) and Pparg null ureters (Pparg cKO) in P1 pups. *P<0.05. Original magnification, ×400 in Aa–Ad, Ae inset, and F; 100× in Ae.

The loss of Pparγ in the urothelium did not abolish the expression of Shh, or investment of Sma⁺ cells around the ureter, suggesting that Pparγ is not important for smooth muscle differentiation (Figure 5, A and B). In fact, the Pparg null urothelium showed a significantly increased number of p63⁺ basal cells at P1 (Figure 5B, quantified in Figure 5C) with a selective increase in basal cell proliferation as reflected by bromodeoxyuridine (BrdU) uptake (Figure 5D, quantified in Figure 5E) and confirmed by Ki-67 staining (data not shown). No change in apoptotic rate was observed (data not shown). Moreover, conditional knockout of Pparg did not prevent expression of Brg1 or ΔNp63 (Figure 5F).

Basal cell markers including cytokeratin 5 (Krt5) and Shh were detected even in the most apical cell layers of the Pparg null urothelium (Supplemental Figures 4 and 5), arguing that Brg1-induced Pparγ expression is required for basal cell differentiation. Consistent with this model, there was significantly reduced apical cell expression of all uroplakins as well as cytokeratin 20 (Krt20) in P1 pups and adults, although E-cadherin expression was maintained (Figure 5, Bn, and Supplemental Figures 6 and 7). These observations are in agreement with reports indicating that Pparγ activation can induce stem cell differentiation and inhibit proliferation and self-renewal.^26,27 In support of Pparγ and ΔNp63 serving distinct roles downstream of Brg1, analysis of ureters from ΔNp63^gfp/gfp mice revealed normal expression of Pparγ and the uroplakins (Supplemental Figure 8).

Discussion

Here we provide the first in vivo data demonstrating that Brg1 plays an indispensible role in regulating ureter development. After ablation of Brg1 in the developing ureter, ΔNp63, Pparγ, and Shh each exhibit markedly reduced expression, correlating with morphologic abnormalities including absence of ureteral stratification and terminal differentiation, reduction of smooth muscle development and massive hydroureter. Using mice in which either ΔNp63 or Pparγ expression is genetically interrupted, we go on to show that ureteral expression of Brg1 separately regulates three key aspects of ureter development: ΔNp63-dependent maintenance of the basal cells, Pparγ-dependent induction of ureteral differentiation, and Shh-mediated investment of the smooth muscle cell layer.

Knockout of p63 has been shown to suppress the regenerative potential of several different stratified epithelial tissues.^16–19 Consistent with this, loss of ΔNp63 in the urothelium prevents normal stratification without altering Brg1 or Pparγ expression (Figure 2, E and F, and Supplemental Figure 8A). In vitro studies have suggested that Pparγ is required for uroplakin expression,^6,7 and our ureteral knockout of Pparg indeed revealed failure of terminal differentiation of the apical cell layer into uroplakin 2-expressing umbrella cells, even though Brg1 and ΔNp63 expression were maintained (Figure 5, B and F). The fact that urothelial Brg1 expression is preserved upon loss of either ΔNp63 or Pparγ indicates that Brg1 is upstream of both factors. In contrast, the maintenance of ΔNp63 in Pparg-null urothelial cells and Pparγ in ΔNp63-null cells argues that these two pathways are operating in parallel downstream of Brg1 (Figure 6).

Figure 6.

A working model of the Brg1 regulatory network in ureter development. (A) A simplified view of the different cell layers of a relatively developed ureter (E16.5 and later). The type size of each gene roughly correlates to its levels of expression. (B) A working hypothesis suggesting that Brg1 is upstream of p63, Pparg, and Shh, although existing data cannot confirm whether these are the direct targets of Brg1.

To guide the proper investment of ureteral smooth muscle cells, the basal cells secret Shh whereas the stromal cells immediately outside of the basal cell layer express the Shh receptor Ptch1. The Shh-Ptch signaling pathway provides the fundamental instructions for stromal cells to differentiate into smooth muscle cells,^4,5 and loss of Shh leads to failure of smooth muscle development and hydroureter.^1,23 In this study, we found that Shh expression is downstream of Brg1, and that loss of Brg1-Shh correlated with failure of smooth muscle differentiation and/or ureteral investment. However, Shh was found to be normally expressed in ureters lacking ΔNp63. This latter finding argues against the proposed role for p63 as a direct regulator of Shh,²⁸ but does not rule out the possibility that ΔNp63 and Brg1 cooperatively define the region of Shh expression. Interestingly, Shh expression was significantly reduced in the Pparg-null ureter even though the failure of normal differentiation in these ureters led to the abnormal maintenance of some Shh expression even in the most apical urothelial cell layer. These findings suggest that Pparγ and Brg1 may both be required for normal expression of Shh by the basal cells, whereas high levels of Pparγ expression in the intermediate and apical cells induces differentiation and acts to suppress Shh expression.

In the bladder, basal cells are believed to serve as progenitors for the intermediate and umbrella cells.³ Our data support this possibility in the ureter as well, and suggest that ΔNp63 and Pparγ may have opposing effects on the maintenance of the basal cell population. There is a decrease in basal cell proliferation while Pparγ expression and terminal differentiation are maintained when ΔNp63 is absent from the ureter, thus depleting the basal cell population. In contrast, the urothelial basal cells exhibit increased proliferation and greater numbers after loss of Pparγ, whereas development of the more apical cell layers is suppressed. Of note, in this study we defined basal cells as those cells adjacent to the ureteral basement membrane that expressed ΔNp63 and/or Krt5. Thus, we cannot rule out the possibility that Pparγ also normally serves to suppress the expression of these basal cell markers.

Taken together, our data place ΔNp63, Pparγ, and Shh downstream of Brg1 during ureteral development, of which Pparg has been shown to be a direct target of the Brg1-containing Swi/Snf complex in adipocytes.^24,29 Additional work will be required to determine whether the ΔNp63 and Shh loci are also direct Brg1 targets. Although the yeast Swi/Snf complex binds primarily to promoter regions and thus target genes can be detected by ChIP or ChIPseq approaches, mammalian Swi/Snf complexes are predominantly found in intergenic areas and can therefore activate or repress target genes at more remote sites.⁸ In either event, our data demonstrate that Brg1 expression activates a program that promotes basal cell Shh expression that is required for smooth muscle development along with both ΔNp63-dependent maintenance of the basal cell compartment and Pparγ-mediated terminal differentiation of a defined population of those basal cells.

Concise Methods

Animals

Mouse protocols were approved by the Yale University Institutional Animal Care and Use Committee. The Brg1^fl/fl mouse was kindly provided by Dr. Pierre Chambon (INSERM U964, France). The Pparg floxed mouse was obtained from the Jackson Laboratory (Bar Harbor, ME). The Hoxb7-Cre-IRES-GFP mouse was a kind gift of Dr. Carlton Bates (University of Pittsburgh, Pittsburgh, PA). The ΔNp63 knockout mice were provided by the Sinha laboratory (Buffalo, NY).

BrdU Labeling and Quantification of Ureteral Basal Cells

Two hours before being euthanized, P1 mice were given 25 μl BrdU (25 mg/ml) by intraperitoneal injection. To label DNA synthesis events in adult ureter, Pparg null and control mice were given 200 μl BrdU solution by intraperitoneal injection 24 hours before being euthanized. To label proliferation events in utero, pregnant females were given 200 μl BrdU and euthanized 2 hours afterward. Rat anti-BrdU antibody (Novus Biologicals, Littleton, CO) was used to detect cells that duplicated their DNA as reflected by BrdU incorporation during the labeling period. Because the basal cell marker cytokeratin 5 (Krt5) is only detectable after E18.5 (Cathy Mendelsohn, personal communication), basal cells were manually counted on the cross-section of proximal ureters by including only the far outside ring of p63⁺ urothelial cells that presumably contact the basement membrane as judged by morphology. This method of counting may miss some true basal cells, but the counting will not falsely include the intermediate cells. Proliferative basal cells are determined by p63⁺/BrdU⁺ from the far outside ring of urothelium.

Obstruction and Reflux Test by Bromophenol Blue Dye Labeling

P1 animals were anesthetized with ketamine and the dye solution was injected in renal medulla/papilla (for the physical obstruction test) and in the bladder (for the vesico-ureteric reflux test) following a standard protocol as described.³⁰ Some internal organs were removed after the test for the clarity of photographing. By the time of photographing, part of the injected dye leaked out so that the blue dye shown on the picture is not as pronounced as it was immediately after injection.

Histology, Fluorescent Immunostaining, and Imaging

After sacrifice, tissues were fixed in 1× PBS buffered formalin for 4–12 hours, embedded in paraffin, and sectioned at 5 μm. To review histology, slides were stained by hematoxylin and eosin. Immunostaining on paraffin-embedded tissue was performed following standard protocols using heat-mediated antigen retrieval. The following polyclonal antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA): Brg1, Pparγ, Upk2, cytokeratin 20 (Krt20), and p63 (which detects the ∆Np63). Rabbit anti-SMA antibody was purchased from Abcam (Cambridge, MA). Rabbit anti-E-cadherin was purchased from Cell Signaling Technology (Danvers, MA). Rabbit anti-Shh antibody was purchased from Bioss USA (Woburn, MA). Rabbit polyclonal anti-TAp63 and anti-cytokeratin 5 (Krt5) antibodies were gifts from the Sinha laboratory.³¹ Hamster anti-Tshz3 was a kind gift from Dr. Alister Garratt (Max-Delbrück Center for Molecular Medicine, Berlin, Germany). All fluorescent images were taken using a Zeiss 710 Duo microscope (confocal/2 photon).

RNA Extraction and Real-Time RT-PCR Analyses

RNA extraction and RT were performed following standard protocols. Real-time PCR analysis was performed using SYBR Green method on a Bio-Rad CFX96 system. The melting curve was always included as part of the program to ensure the quality of single band amplification. Data were presented in the form of Δ-Ct method. All primers have been tested for efficiency. Sequences of all primers are available upon request.

Statistical Analyses

Data are presented as mean ± SD throughout the text unless otherwise specified. Data are analyzed using the t test, with P<0.05 regarded as statistically significant.

Disclosures

None.