Regulation of Limbal Keratinocyte Proliferation and Differentiation by TAp63 and ΔNp63 Transcription Factors

Abstract

purpose. To examine the effects of TAp63 and ΔNp63 on the proliferation and differentiation of rabbit limbal keratinocytes cultured on human amniotic membrane.

method. Real-time Q-RT-PCR was used to quantify the relative abundance of TAp63 and ΔNp63 transcripts in limbal, peripheral corneal, and central corneal epithelia. Antisense oligonucleotides were designed specifically to suppress the expression of TAp63 or ΔNp63 in limbal keratinocytes, and their effects on cell proliferation and differentiation were examined. Immunofluorescence was used to examine the expressions of p63 and keratin-3 and -14.

results. The expressions of TAp63 and ΔNp63 transcripts appeared to be site specific. TAp63 was expressed at the highest level in limbus, decreased by approximately 10-fold in peripheral cornea and was undetectable in the central cornea. ΔNp63 was also expressed at the highest level in limbus, decreased by approximately 35% in peripheral cornea, and was undetectable in the central cornea. Suppression of TAp63 expression inhibited limbal keratinocyte proliferation but promoted differentiation. Suppression of ΔNp63 expression also inhibited cell proliferation but had no obvious effect on cell differentiation.

conclusions. TAp63 and ΔNp63 affect the proliferation of limbal keratinocytes by a different mechanism. The inhibition by TAp63 antisense oligos appeared to be secondary to the promotion of cell differentiation. In contrast, the inhibition by ΔNp63 antisense oligos appeared to be independent of cell differentiation.

The p63 transcription factor exhibits high levels of sequence and structure homologies with p53, it consists of a transactivation domain in the amino terminus, a DNA-binding domain in the central region, and an oligomerization domain in the carboxyl terminus. ¹ In contrast to p53, the p63 gene contains two transcriptional start sites, and alternative usage of the promoters leads to the production of two major isoforms: one with and one without the amino terminal transactivation domain. ² The p63 proteins containing the amino terminal transactivation domain are named TAp63 and those without the transactivation domain are named ΔNp63. Transient transfection studies have shown that p63 can either activate or suppress transcriptions of the reporter genes with the p53 DNA-binding site. ¹ Based on the structure similarities between p53 and p63, it has been suggested that TAp63 can transactivate the p53 target genes and that ΔNp63 probably acts to suppress the transcription of the p53 target genes in a dominant-negative manner. ¹  

In contrast to the tumor suppressor role of p53, ΔNp63 has been found to be overexpressed in many human squamous cell carcinomas, ^{3 4 5} suggesting a role in oncogenesis. Recent studies also suggested that p63 functions primarily in ectodermal differentiation during development and in the maintenance of stratified epithelial progenitor cells. Mills et al. ⁶ showed that p63^−/− mice exhibit striking developmental deficiencies in limbs, hair follicles, teeth, and mammary glands during various stages of their embryonic development. Moreover, Yang et al. ⁷ showed that mice without p63 display severe defects in limb, craniofacial, and epidermal development, suggesting that p63 plays a critical role in the morphogenesis and differentiation of these stratified epithelial structures. In a recent report of their study, Koster et al. ⁸ suggested that TAp63 is necessary for the initiation of the epithelial stratification program and the ΔNp63 is necessary for the maintenance of epidermal differentiation. Thus, p63 may be essential for the homeostasis of the epithelial tissue. 

Corneal epithelial cells are derived from the limbus, which is located in the transition zone between the cornea and the conjunctiva. In contrast to the cornea, the basal cells of the limbal epithelium are devoid of the keratin pair k3 and k12, indicative of the presence of more primitive cells. ⁹ In the past decade, several protein molecules, including p63, have been suggested to be limbal stem cell markers; however, self-renewing and slow-cycling are still the two most accepted characteristics of the keratinocyte stem cells. ¹⁰ In this regard, the bromodeoxyuridine (BrdU) label-retaining cells (LRCs), the slow-cycling cells, have been found to reside in the basal layer of the limbal epithelium. ¹¹ Recently, Pellegrini et al. ¹² showed that the p63 transcription factor is selectively expressed in the basal cells of the human limbus, but not in the corneal epithelium, and they suggested p63 to be a specific nuclear marker for the limbal stem cell. Concordant with their report, we also showed that a human amniotic membrane (AM)-based culture system can preserve and propagate p63-positive epithelial cells of the rabbit limbus. ¹³ However, in a more recent report, we showed that the expression of p63 in rat limbal–corneal epithelium changes with the animal’s postnatal age, and that many p63-positive cells are not slow cycling. In rat cornea, p63 is expressed in stem cells and young transient amplifying cells (TACs), with higher levels of expression in the latter. ¹⁴ This suggests that p63 could not be a specific marker for limbal epithelial stem cells. To date, the roles of TAp63 and/or ΔNp63 in the regulation of limbal epithelial stem cell proliferation and differentiation have not been fully elucidated. 

In the present study, we present data showing that, in human AM-based rabbit limbal keratinocyte culture, TAp63 may be involved in the maintenance of the limbal epithelial cells in a less differentiated state, whereas ΔNp63 may be involved in cell-cycle control, to sustain the proliferating activity of the limbal keratinocytes. 

Materials and Methods

Dulbecco’s modified Eagle’s medium (DMEM), Ham’s F-12 nutrient, trypsin-EDTA, fetal bovine serum (FBS), dispase II, and extraction reagent (TRIzol) were purchased from Invitrogen/Gibco Corp. (Carlsbad, CA); dimethyl sulfoxide (DMSO) and bovine insulin from Sigma-Aldrich (St. Louis, MO); cholera toxin A subunit (type Inaba 569B, azide free) from Merck Biosciences AG/Calbiochem (Oakland, CA); mouse receptor-grade epidermal growth factor (EGF) from Upstate Biotech Inc. (Waltham, MA); antibodies to p63 nuclear protein (4A4 clone), keratin 3 (AE5 clone), keratin 14 (LL002 clone), and all the fluorescent-dye–conjugated secondary antibodies from Chemicon International, Inc. (Temecula, CA); and plastic cell culture ware from Corning Incorporated Life Sciences Co. (Acton, MA). 

Partial Sequencing of the Rabbit TAp63 and ΔNp63

Total RNA was isolated from rabbit skin epidermis by guanidine isothiocyanate/acid-phenol extraction (TRIzol; Invitrogen/Gibco Corp., Carlsbad, CA). Because rabbit genomic information was not available, we used isoform-specific reverse primers of human p63 transcript ¹⁵ as the gene-specific primer (GSP) oligonucleotides for the cDNA 5′ end rapid amplification reaction (5′-RACE). The 5′-RACE reaction was performed with a commercial 5′-RACE system according to the manufacturer’s protocol (Invitrogen Corp.). In brief, the first-strand cDNA synthesis was primed using human p63 GSP. After cDNA synthesis, the first-strand product was purified from unincorporated dNTPs and GSP, and homopolymeric tails were added to the 3′ ends of the cDNA by termini deoxynucleotidyl transferase (TdT). The tailed cDNA was then amplified by PCR with a mixture of human GSP and a combination of complementary homopolymer-containing anchor primers and corresponding adapter primers that permit amplification from the homopolymeric tails. The products of the 5′-RACE amplification reaction were used to construct rabbit cDNA clones by a TA cloning system (Invitrogen Corp.). Fifteen clones were selected and sequenced, and the sequencing results were compared with the cDNA sequences of the human and mouse p63 isoforms. An N-terminal fragment encompassing 246 and 185 bases, respectively, of the rabbit TAp63 and ΔNp63 cDNA were obtained (Fig. 1) . Based on our sequencing data, the N termini of the rabbit TAp63 and ΔNp63 exhibit a 96% and 90% homology, respectively, to its human counterpart. 

Real-Time Quantitative RT-PCR

The relative abundance of the TAp63 and ΔNp63 transcripts in the epithelia of the rabbit limbus, peripheral cornea, and central cornea was analyzed by real-time quantitative RT-PCR (real-time Q-RT-PCR). New Zealand White rabbit was anesthetized, and the ocular surface was excised aseptically and placed on a 60-mm culture dish containing phosphate-buffered saline (PBS). The superfluous connective tissue was removed under a dissecting microscope. The limbal zone was defined as the ring with one-mm width from both sides of the cornea-conjunctival junction. The central corneal zone was defined as the circular area with a 2-mm radius from the center of the cornea. The peripheral corneal zone was defined as the ring between the central cornea and limbal zone. ^{13 14} This visual definition of the limbal and peripheral zones was determined by the histologic differences in the underlying stroma matrix. Total RNA was extracted (TRIzol; Invitrogen-Gibco) reagent according to the manufacturer’s specifications. RNA sample was dissolved in RNase-free water and the A260/A280 ratio was measured. One microgram of the RNA was used for real-time Q-RT-PCR. The rabbit p63 isoform-specific primers used were ΔNp63 forward, 5′-AAC AGC ATG GAC CAG CAG ATC-3′; ΔNp63 reverse, 5′-TCT GTG CGT GGT CTG TGT TGT-3′; TAp63 forward, 5′-GAG TCC TGC ATG CGG ATA CA-3′; and TAp63 reverse, 5′-AAA ATG GCG CAA CAA ACA AGA-3′. The thermal cycle parameters of real-time Q-RT-PCR were as follow: 2 minutes at 50°C, 10 minutes at 95°C, 40 PCR cycles of 15 seconds at 95°C, and 1 minute at 60°C. The real-time Q-RT-PCR reaction mixture was prepared in a 50 μL solution, containing 2× master mix (SYBR Green PCR Master Mix) and the desired primers (0.5 nM). The PCR-amplified product was quantitated by a sequence detector (Prism 7000; Applied Biosystems, Inc. [ABI], Foster City, CA). The intensity of the green fluorescent dye emissions increased in direct proportion to the increase of the PCR amplified product. The amplification plot was examined early in the reaction, at a point that represents the logarithmic phase of the product accumulation. The point representing the detection threshold of the increase of the fluorescent signal associated with the exponential growth of the PCR product for the detector is defined as the cycle threshold (C _T). C _T is predictive of the quantity of input target—that is, when the conditions of the PCR are the same, the larger the starting concentration of a transcript, the lower the C _T. A normalization experiment was performed simultaneously for each sample (Eukaryotic 18S rRNA Endogenous Control Kit; cat. no. 4333760T; ABI). The data obtained from real-time Q-RT-PCR were calibrated with those of 18S rRNA in the respective tissue and recalibrated with the data obtained from central cornea to obtain the relative expression levels. To validate the specificity of the Q-RT-PCR reaction, the heat-dissociation kinetics for TAp63 and ΔNp63 were performed. Briefly, after Q-RT-PCR, samples and the respective nontemplate control (NTC) were subjected to a heat-dissociation protocol between 60°C and 95°C. The TAp63 and ΔNp63 Q-RT-PCR samples dissociated, respectively, at 76.8°C and 82.4°C. In contrast, the NTC samples showed no dissociation (Fig. 2) . The single dissociation peak exhibited by the TAp63 and ΔNp63 Q-RT-PCR samples strongly suggested that the amplification is template specific. Polyacrylamide gel (20%) electrophoretic analysis of the reaction mixture also showed the presence of a single product with the expected size range (58 bp for TAp63, and 74 bp for ΔNp63, data not shown). 

Blockade of the p63 Expression by Antisense Oligonucleotides

Antisense oligonucleotides were designed from partial gene sequences presented in Figure 1 . The sequence of anti-TAp63 oligonucleotide was 5′-TAT GCT GGA AAA CCT CAG GAC TGA GGA ATT-3′, and the anti-ΔNp63 oligonucleotide was 5′-TCC ATG CTG TTA CAG GAG CAC CCA GGT TCG-3′. The effect of TAp63 and ΔNp63 on the growth of limbal epithelial cells was examined by adding antisense oligos, either alone or together to the culture medium. The antisense oligos were added to a final concentration of 1 μΜ in limbal stem cell complete medium (LSCCM; DMEM/Ham’s F12 at 1:1 ratio with 0.5% DMSO, 2 ng/mL mEGF, 1 μg/mL bovine insulin, and 0.1 μg/mL cholera toxin) 3 days after tissue implantation. The culture medium was replaced with the same medium containing the desired antisense oligos every day. The areas of the epithelial sheets grown from limbal explants were measured on days 7 and 14 after implantation. Data were calculated from the images on computer (Image Pro-Plus, ver. 4.5; Media Cybernetics, Silver Spring, MD). 

Preparation of Human AM and AM-Based Limbal Keratinocyte Cell Culture

Human AM was obtained from the Department of Obstetrics and Gynecology Chang Gung Memorial Hospital (Linko, Taiwan) with proper informed consent. AM was processed as previously described ¹³ and stored at −80°C in DMEM containing 50% glycerol. AM was denuded by scrapping and placed in a culture dish with the basement-membrane side up at room temperature overnight before use. 

Limbal tissue was procured from healthy eyes of the New Zealand White rabbit, housed, and treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Visual Research and according to an experimental procedure approved by the Committee for Animal Research at Chang Gung University. The limbal biopsy specimen was treated with 2.4 U/mL of dispase II at 37°C for 20 minutes to loosen the epithelial layer from the underlying stroma tissue. Dispase-II treatment improved the migration of keratinocytes from the explant tissue onto the AM surface, therefore, increased the uniformity of the initial epithelial outgrowth. After enzyme digestion, the limbal biopsy was washed with PBS for 5 minutes. The dispase-treated limbal biopsy specimen was implanted on the AM and cultured in LSCCM. ¹³ Culture dishes were incubated in a humidified incubator at 37°C for 7 to 14 days under 95% air-5% CO₂. 

Immunofluorescent Staining and Microscopic Examination

The immunofluorescent staining was performed as previously described. ¹³ Briefly, rabbit limbal tissues and AM-based ex vivo epithelial cultures were fixed with 4% paraformaldehyde and paraffin embedded. The 3-μm sections were incubated with normal serum and then with the primary antibodies: anti-p63, anti-keratin-3 and -14. For dual-color immunostaining, samples were incubated with either cyanine 3 (Cy3)- or FITC-conjugated antibodies. Sections were mounted with anti-fade mounting medium (Gel Mount; Biomeda Co., Foster City, CA), and examined under a fluorescence microscope (Carl Zeiss Meditec, Oberkochen, Germany). All images were acquired from digital photograph system and processed (PhotoShop 7.0; Adobe Systems, Mountain View, CA). 

In Vitro Cell Migration Assay

Cell migration was measured by the razor wound method. ¹⁶ Briefly, limbal keratinocytes were expanded by coculture with mitomycin C-pretreated Swiss 3T3 fibroblasts. The cells were than harvested, plated, and grown in 35-mm dishes until confluent. The confluent cultures were treated with the desired antisense oligos at 1 μM for 24 hours. Cultures were then wounded by gently pressing a sterilized razor through the epithelial sheet into the plastic wall to mark the origin and drawing the razor on one side through the monolayer to remove the cells. After the cells were wounded, the medium was replaced with the same medium containing respective antisense oligonucleotides at the same concentration as indicated. Keratinocyte migration was permitted for up to 6 hours and stopped by fixation with 4% paraformaldehyde. The number of cells across the origin line into the wounded area was counted by image-analysis software. 

Statistical Analysis

All the data were analyzed on computer (SPSS) by one-way analysis of variance (ANOVA) and the Fisher multiple-comparison method. 

Results

Relative Abundance of TAp63 and ΔNp63 mRNA in the Epithelia of Limbus and Cornea

In a prior report, we showed that p63 signal is confined to the basal and suprabasal epithelial cells of rabbit limbus, limbal explant, and limbal epithelial outgrowth on AM. ¹³ Because the antibodies used in our study did not differentiate TAp63 from ΔNp63, the relative abundance of TAp63 and ΔNp63 isoforms in the limbal–corneal epithelium was not compared. To determine whether there is site-specific expression of TAp63 and/or ΔNp63 mRNA transcripts in rabbit limbal–corneal tissues in vivo, we used real-time Q-RT-PCR. The primer pairs were designed specifically to detect the nucleotide sequences between exons 2′ and 4, and exon 3 and exon 4 for TAp63 and ΔNp63, respectively. As stated earlier, the single heat-dissociation peak in the TAp63 and ΔNp63 real-time Q-RT-PCR reaction strongly indicated the presence of a single product and that the amplification is template specific (Fig. 2) . As shown in Figure 3 , TAp63 transcript was expressed at the highest level in the limbus, was decreased by 10-fold in the peripheral cornea, and was undetectable in the central cornea. The ΔNp63 transcript was also expressed at the highest level in the limbus, was decreased by approximately 35% in the peripheral cornea, and was undetectable in the central cornea. In the limbus, TAp63 transcript was relatively more abundant than the ΔNp63 transcript. In contrast, in the peripheral cornea, the TAp63 transcript was five times less abundant than the ΔNp63 transcript. 

Effects of TAp63 and ΔNp63 Antisense Oligos on Cell Growth and Differentiation

Our studies have shown that the expression of p63 in rabbit limbal explant and its epithelial outgrowth on human AM is confined to the basal and suprabasal layers, a pattern that is similar to that of the freshly prepared rabbit limbus. ¹³ This finding is concordant with the hypothesis that p63 protein plays an important role in the homeostasis of corneal epithelium and may serve as a specific marker for LSCs. However, in rat corneal epithelium, we found that p63 is highly expressed in young TA cells as well, ¹⁴ suggesting that p63 could not be a specific marker for limbal keratinocyte stem cells in rat. To explore the roles of TAp63 and ΔNp63 in cell proliferation and differentiation, we designed the antisense oligonucleotides specifically to block p63 mRNA translation and tested their effect on limbal keratinocyte growth and differentiation. We first examined the effect of antisense oligos on the expression of the relative proteins. The limbal epithelial keratinocytes were isolated from rabbit limbus and cultured on plastic dishes in LSCCM for 24 hours, with or without 1 μM of antisense oligos. The cells were fixed with ice-cold acetic acid and alcohol (1:1), followed by incubation with anti-p63 monoclonal antibody (this antibody reacts with both TAp63 and ΔNp63). The results showed that both anti-TAp63 and anti-ΔNp63 oligos effectively reduced the expression of anti-p63 antibody-reactive proteins (Fig. 4) . The effects of antisense oligos on the epithelial outgrowth on AM-based limbal explant cultures were then examined. Figure 5shows that the epithelial outgrowth from the limbal explant was suppressed significantly by both antisense oligos. Quantitative measurements showed more than 80% and 40% suppression, respectively, on days 7 and 14, compared with that of the untreated control group. Coincubation of the explant with both TAp63 and ΔNp63 antisense oligos showed no additive inhibition (not shown). The epithelial sheets grown from the limbal explants in the presence and absence of the antisense oligos were examined for the expressions of K3 and K14 by immunofluorescent staining. Figure 6shows that K3 and K14 expression in the epithelial sheets formed from the untreated and scrambled oligo-treated explants was confined to the superficial and basal layers, respectively. In the presence of TAp63 antisense oligos, K3 was expressed throughout the entire cell sheet, and the K14 expression was minimal. The expression patterns of K3 and K14 in the epithelial outgrowth formed in the presence of ΔNp63 antisense oligos were not obviously altered compared with that of the non–antisense oligo control in the same culture conditions. 

Effect of Antisense Oligos on Cell Migration

To see whether the suppressive effect of the antisense oligos on epithelial outgrowth was through the suppression of cell migration, we used a limbal epithelial cell monolayer culture and razor wound method to evaluate their effect on cell migration. Figure 7shows that the migratory activity of the limbal keratinocytes was not influenced by the TAp63 or ΔNp63 antisense oligos. The results strongly indicate that the inhibition of epithelial outgrowth by antisense oligos is not due to an inhibition of cell migration. 

Discussion

The p63 transcription factor is expressed in a confined manner, with the highest expression found in the basal cells of various epithelial tissues and with ΔNp63 being the most abundant. ^{1 2} It has been shown that DNA damage upregulates the expression of TAp63 ¹⁷ and downregulates ΔNp63, ¹⁸ suggesting different roles of the TA and ΔN isoforms in cell regulation. Whether TAp63 and ΔNp63 play different regulatory roles in limbal epithelium is an important question in cell biology and ophthalmology and deserves some exploration. 

The first step toward understanding the regulatory functions of TAp63 and ΔNp63 in the development of corneal epithelium is to see whether there are site-specific expressions of these two transcription factors. We used real-time Q-RT-PCR to examine the relative abundance of these two transcripts in limbal, peripheral corneal, and central corneal epithelia. We found that TAp63 transcript is mainly expressed in the limbus, and its expression level in the peripheral cornea was decreased drastically to approximately one tenth that in the limbus (Fig. 3) . A further decrease of its expression was observed toward the cornea, where the TAp63 transcript is undetectable. As the limbus is where the undifferentiated, slow-cycling stem cells are located, and the peripheral cornea is where K3-lineage differentiating and rapid-proliferating TA cells are found, it seem reasonable to postulate that TAp63 may be involved in the maintenance of the limbal keratinocytes in the stem cell compartment or in a less-differentiated state. This postulation seems to be consistent with our result obtained in the antisense oligo experiment. As is shown in Figures 5 and 6 , antisense oligos to TAp63 inhibited the growth of limbal keratinocytes and promoted K3 expression. The result strongly suggests that an absence of TAp63 expression allowed limbal keratinocytes to enter the differentiation path leading to K3 expression and that the inhibition of cell proliferation by TAp63 antisense oligos was probably secondary to its promotion of cell differentiation. 

The highest expression of ΔNp63 is also found in the limbus; however, its relative abundance was approximately 30% less than that of TAp63. ΔNp63 was expressed in the peripheral cornea at a relatively high level (a mere 35% decrease from that of the limbus) and was also undetectable in the central cornea. The high level of ΔNp63 expression in both the limbus and peripheral cornea implies that it may be essential for cell regulation in these two adjacent tissues. A common characteristic of the keratinocytes in these two regions is cell cycling, being slow in the limbus and rapid in the peripheral cornea. We therefore postulated that ΔNp63 may have a permissive role in the cell cycle and may play little or no role in keratinocyte differentiation. The experimental result with antisense oligos to ΔNp63 appeared to be consistent with our postulation. These oligos inhibited limbal keratinocyte proliferation, but exerted very little or no effect on K3 and K14 expression. Thus, the absence of ΔNp63 expression was inhibitory to limbal keratinocyte proliferation, and such inhibition was independent of cell differentiation. 

Based on the structural similarities, TAp63 has been suggested to have a regulatory role similar to that of p53, whereas ΔNp63 may be a dominant negative inhibitor of p53 and p63 functions. ^{19 20 21} In this study, we clearly show that the absence of TAp63 expression promoted limbal keratinocytes to enter differentiation (as judged from changes in their K3 and K14 expression), whereas the absence of ΔNp63 expression suppressed keratinocyte cell cycling. Because TAp63 and ΔNp63 are both expressed at high levels in the limbal epithelium, it is likely that these two isoforms act in a cooperative fashion to maintain the limbal keratinocyte stem cell compartment. In contrast, the permissive role of ΔNp63 on limbal keratinocyte cell cycling may be exerted independent of TAp63. The molecular mechanisms underlying our observations await further study. 

Our finding is consistent with a report by Koster et al. ⁸ showing that TAp63 and ΔNp63 have different regulatory functions in the development of the stratified epithelium. They suggested that TAp63α, but not ΔNp63α, is necessary for the commitment of single-layered epithelium to initiate a stratification program. De Laurenzi et al. ²² showed that ΔNp63 is downregulated in normal human epidermal keratinocytes during Ca²⁺-induced in vitro terminal differentiation. More recently, it was reported that Ca²⁺-induced differentiation of the primary keratinocyte culture is blocked by TAp63α transfection. ^{8 23} The results indicated that TAp63 plays a role in inhibiting the differentiation of epidermal keratinocytes. ΔNp63α overexpression has been shown to inhibit squamous morphogenesis in keratinocytes and block differentiation- associated growth arrest as well as differentiation-specific protein expression. ²⁴ Considering all evidence, it seems clear that TAp63 and ΔNp63 play different regulatory functions during the terminal differentiation of the epidermal keratinocytes. ΔNp63 has been found to be overexpressed in some lung cancers, squamous cell carcinomas of the head and neck, nasopharyngeal carcinomas, bladder cancers, and oral squamous cell carcinomas (OSCC), ^{5 25 26 27} suggesting a role in carcinogenesis. In a luciferase assay of a full-length p21 promoter construct, Westfall et al. ²⁸ showed that ΔNp63α acts as a transcriptional repressor and regulates the proliferative capacity of the basal keratinocytes, suggesting that ΔNp63 may have a function in maintaining proliferative activity of the epithelial keratinocytes. Thus, the growth regulatory function of ΔNp63 in limbal keratinocytes reported in the current study is in agreement with that which they reported. 

In brief, TAp63 appears to play roles in the initiation of the epithelial stratification program and in the maintenance of the keratinocyte stem cell and/or progenitor cell compartment of the epithelia, whereas ΔNp63 may be necessary for the promotion of the of cell cycle and the proliferative potential of the keratinocytes. However, the precise functions of TAp63 and ΔNp63 in the regulation of the development, differentiation, and maturation of the various stratified epithelia await further exploration. 

 Supported by Grant NSC-03-2320-B-182-002 from the National Science Council, Taiwan.

 Submitted for publication January 14, 2005; revised March 28, 2005; accepted April 26, 2005.

 Disclosure: D.-Y. Wang, None; C.-C. Cheng, None; M.-H. Kao, None; Y.-J. Hsueh, None; D.H.K. Ma, None; J.-K. Chen, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Alignment and comparison of the human (Hu), mouse (Mu), and rabbit (Rb) p63 N-terminal sequences. Blue: the nucleotide in rabbit that is different from that of human and/or mouse; red: the nucleotide that is found only in rabbit. The underscored sequences in (A) and (B) were used, respectively, for ΔNp63 and TAp63 antisense oligo templates.

Figure 2.

 

The heat-dissociation analysis for TAp63 and ΔNp63 Q-RT-PCR products. After PCR, the samples and their corresponding no-template control (NTC) samples were subjected to a heat-dissociation protocol between 60°C and 95°C. Both TAp63 and ΔNp63 samples exhibited a single dissociation temperature at 76.8°C and 82.4°C, respectively, whereas the NTC sample exhibited no heat dissociation kinetics.

Figure 3.

 

Real-time quantitative RT-PCR analysis of the expressions of TAp63 and ΔNp63 transcripts in rabbit limbal, peripheral corneal, and central corneal epithelia. Total RNA was isolated from the respective epithelium and processed for analysis. Data are the mean ± SD of results in triplicate determinations.

Figure 4.

 

Suppression of TAp63 and ΔNp63 protein expressions by antisense oligos to TAp63 and ΔNp63 mRNA. Limbal keratinocytes were treated with respective antisense oligos or scrambled oligos or were untreated for 24 hours. The cells were then fixed and immunofluorescent staining was performed. Keratinocytes in the untreated (A) and scrambled-oligo–treated (B) cultures showed numerous p63-positive nuclei (pink staining resulted from the simultaneous presence of blue Hoechst staining and red p63 staining), indicating the presence of strong p63-expressing cells. In TAp63 (C) and ΔNp63 (D) antisense oligo-treated cultures, the numerous lighter pinkish blue nuclei represent the cells with a reduced total p63 expression (due to suppression of TAp63 or ΔNp63 expression by antisense oligos, therefore, a lighter red), whereas the nuclei with more intense blue color (Hoechst stain) represent cells with no p63 expression.

Figure 5.

 

Inhibition of epithelial outgrowth from rabbit limbal explants by antisense oligonucleotides to TAp63 and ΔNp63 mRNA. After dispase II treatment, a limbal biopsy specimen (1 × 2 mm) was implanted on human AM in LSCCM. Three days after implantation, explants with a roughly equal amount of epithelial cell outgrowth were chosen for the experiment. The medium was replaced with the same medium containing the desired antisense oligonucleotide at 1 μM, as indicated. Culture medium was changed every 24 hours and contained the same concentrations of the respective antisense oligonucleotides. Cultures were stained with trypan blue, and the amniotic membrane surface that was not covered by keratinocyte sheet was stained blue. Dotted line: boundary of the cell sheet.

Figure 6.

 

Effect of antisense oligonucleotides on limbal keratinocyte differentiation. Keratinocyte sheets grown from limbal explants, with or without the treatment with antisense oligos, were fixed and double stained with K3- and K14-specific antibodies. Keratinocyte sheets formed in the absence of antisense oligos (A) or in the presence of scrambled oligos (B) showed K3-positive cells (red) in the superficial layer and K14-positive cells (green) in the basal layer. The cell sheet that formed in the presence of antisense oligos to TAp63 showed a strong expression of K3 throughout the entire cell sheet, with minimal K14 expression. In the cell sheet formed in the presence of antisense oligos to ΔNp63, the K3 and K14 staining pattern was not obviously altered compared with that in (A) and (B). Yellow: simultaneous presence of K3 and K14; blue: Hoechst stain.

Figure 7.

 

Effect of antisense oligonucleotides on the migratory activity of the limbal keratinocytes cultured on collagen-coated cell culture plasticware. The confluent cells were incubated for 24 hours, with or without antisense oligonucleotides, as indicated. Cultures were then wounded by a sterile razor. Cell migration was allowed to proceed for 6 hours, the cells were fixed, and cell migration across the original razor line was scored. Data are the mean ± SD of triplicate measures.