Introduction

The polycomb group (PcG) genes play a centrally important role in mammalian development. Bmi-1 (B-cell-specific Moloney murine leukemia virus integration site 1) is the mammalian form of a Drosophila PcG protein that regulates stem cell maintenance and renewal (Lessard et al., 1999; Ohta et al., 2002; Park et al., 2003). Bmi-1 is an epigenetic regulator that is thought to act, in some systems, by suppressing ink4a/ARF locus gene expression. This locus encodes the cell cycle progression suppressor proteins p16^ink4a and p19^arf (p14^arf in humans). Bmi-1-dependent suppression of p16^ink4a and 19^arf level, and the level of other cell cycle suppressor proteins, leads to enhanced stem cell survival and proliferation (Iwama et al., 2004). Disruption of Bmi-1 action has profound effects on development (Iwama et al., 2004). For example, Bmi-1^-/- mice are unable to efficiently renew hematopoietic stem cells, a response that has been attributed to the absence of normal, tonic Bmi-1-dependent suppression of p16^Ink4a and p19^Arf levels (Jacobs et al., 1999a; Molofsky et al., 2003; Park et al., 2003). Reduced stem cell renewal is also observed for leukemic and neuronal stem cells in Bmi-1^-/- mice (Lessard and Sauvageau, 2003; Molofsky et al., 2003). In addition to these effects on stem cells, it has been suggested that Bmi-1 may enhance tumor cell proliferation/survival by repressing transcription of p16^ink4 and p19^arf (Kim et al., 2004) and via Bmi-1-dependent increased expression and activity of telomerase reverse transcriptase (Dimri et al., 2002). Telomerase reverse transcriptase is the catalytic protein subunit of telomerase, which is responsible for maintenance of telomere length, and telomerase reverse transcriptase activation is associated with cell immortalization (Dimri et al., 2002). Taken together, these findings suggest an important role for Bmi-1 in regulating normal cell physiology and as a pro-survival/proliferation factor in hyperproliferative diseases and epithelial cancer (Vonlanthen et al., 2001; Dimri et al., 2002; Kim et al., 2004).

The epidermis is a stratifying multilayered epithelium in which proliferating cells, located in the basal layer, undergo regulated cell division (Fuchs and Byrne, 1994). This cell division results in the production of daughter cells that move into the suprabasal epidermal layers and undergo differentiation (Eckert et al., 1997). These differentiating cells lose the ability to proliferate, but they remain viable until the final stage of the differentiation process. Understanding the process whereby these cells cease proliferation and are maintained during the differentiation program is a major goal. In this study, we explore the role of Bmi-1 as a regulator of keratinocyte function. We show that Bmi-1 is expressed in cultured epidermal keratinocytes and in human epidermis, and that overexpression of Bmi-1 in cultured normal human keratinocytes promotes increased cell numbers. In addition, Bmi-1 protects keratinocytes from challenge with stress agents (UVB, okadaic acid (OA), etc.), and Bmi-1 levels are elevated in psoriasis and in squamous cell cancer, suggesting a role in promoting tumor cell survival. These effects are associated with changes in cell cycle regulatory protein expression and suppression of keratinocyte apoptosis. Based on these findings, we propose that Bmi-1 plays an important role in maintaining proliferation of basal keratinocytes and in viability of differentiating suprabasal keratinocytes. These findings suggest that Bmi-1, which was originally identified as a stem cell maintenance factor, may have a non-stem cell role in epidermis in maintaining the survival of differentiating suprabasal keratinocytes.

Results

Bmi-1 is expressed in keratinocytes

A major goal of this study was to assess whether Bmi-1 has a role in regulating normal human epidermal keratinocyte survival. Bmi-1 has been reported to be expressed in epidermal keratinocytes maintained on 3T3 feeder cells and also in cultured hair follicle-derived stem cells (Claudinot et al., 2005; Maurelli et al., 2006). We first confirmed that Bmi-1 is expressed in cultured epidermal keratinocytes when maintained in keratinocyte serum-free medium and cultured on plastic. As shown in Figure 1a, Bmi-1-encoding mRNA, as assessed by reverse transcription-PCR (RT-PCR) analysis, is expressed. Moreover, as shown in Figure 1b, in which total keratinocyte extracts were incubated in the presence (+) or absence (-) of anti-Bmi-1, Bmi-1 protein is also present. As PcG gene products, including Bmi-1, act in the nucleus (Cohen et al., 1996), we monitored the subcellular localization of Bmi-1. Keratinocytes, grown on glass coverslips, were immunostained with anti-Bmi-1. As shown in Figure 1c, endogenous hBmi-1 is detected in the nucleus.

Figure 1.

Bmi-1 is expressed in the nucleus of cultured normal human keratinocytes. (a) Total RNA (500 ng), prepared from near-confluent cultures of normal human epidermal keratinocytes, was amplified by RT-PCR in the presence or absence of 400 nM of the appropriate primers, and the resulting cDNA product was detected by ethidium bromide staining (arrow). M indicates the migration of DNA size markers (nucleotides). -Actin RNA was amplified to a similar level in each sample (not shown). (b) Total cell extract, prepared from near-confluent cultures of normal human keratinocytes, was electrophoresed on an 8% polyacrylamide gel for immunoblot. Parallel lanes were incubated in the presence (+) or absence (-) of mouse monoclonal anti-Bmi-1 (diluted 1:500) followed by addition of the secondary antibody and visualization using enhanced chemiluminescent reagent. -Actin levels were measured to normalize protein loading. (c) Keratinocytes, growing on glass coverslips, were fixed and incubated in the presence or absence of mouse monoclonal anti-Bmi-1 (diluted 1:25) followed by incubation with Cy3-conjugated sheep anti-mouse IgG (1:1,000). The arrows indicate nuclear Bmi-1. Similar results were observed in each of the three separate experiments.

Full figure and legend (69K)

To confirm Bmi-1 expression in human epidermis, keratinocyte RNA was isolated and Bmi-1 mRNA was detected by RT-PCR. As shown in Figure 2a, Bmi-1-encoding mRNA is detected in foreskin epidermis. Moreover, incubation of foreskin epidermal extracts with anti-Bmi-1 reveals the presence of Bmi-1 protein (Figure 2b). We then examined foreskin sections using immunohistological analysis to monitor Bmi-1 tissue distribution. As shown in Figure 2c, although Bmi-1 is detected in some of the basal layer proliferative cells, immunoreactivity is also detected, in a dispersed distribution, in the suprabasal epidermal layers. Little or no Bmi-1 expression is detected in the dermis. The presence of suprabasal Bmi-1 is consistent with a recent report (Reinisch et al., 2007). The fact that Bmi-1 is expressed at high levels throughout the suprabasal layers is interesting as it indicates that Bmi-1 may have a non-stem cell role in this tissue.

Figure 2.

Bmi-1 is expressed in human epidermis. (a) Total RNA (500 ng), prepared from dispase-separated human foreskin epidermis, was amplified by RT-PCR in the presence or absence of 400 nM of each primer, and the resulting DNA product was detected by ethidium bromide staining (in vivo) (arrow). The sample indicated "K" is the RT-PCR product detected using RNA isolated from tAd5-hBmi-1-infected cultured human keratinocytes. M indicates the migration of DNA size markers (nucleotides). (b) Total cell extract was prepared from dispase-separated human foreskin epidermis (in vivo) or from cultured normal human keratinocytes (K) and electrophoresed on an 8% polyacrylamide before immunoblot using anti-Bmi-1 (1:500) and HRP-conjugated sheep anti-mouse IgG secondary antibody (1:5,000). Secondary antibody binding was visualized using enhanced chemiluminescent reagent. -actin levels were measured to normalize protein loading. (c) Human epidermal tissue sections were paraffin embedded, deparaffinized, and hydrated through xylene and a graded alcohol series. Following incubation in 0.3% H₂O₂ for 30 minutes, the tissue sections were washed with PBS for 5 minutes and blocked for 1 hour in PBS containing 1% normal goat serum and 1 mg/ml bovine serum albumin. The primary antibody (mouse monoclonal anti-Bmi-1, diluted 1:50) was added for 2 hours at room temperature followed by incubation for 30 minutes with the secondary antibody (biotinylated horse anti-mouse, 1:1,000) and the signal was visualized using the Vectastain kit (Vector Laboratories, Burlingame, CA). The sections were washed and then counterstained with hematoxylin (Fisher, Pittsburgh, PA) and sealed under coverslips. The top panel (control) was incubated with secondary antibody and the bottom panel (anti-Bmi-1) was incubated with both primary and secondary antibodies. The arrows indicate the epidermal basal layer. Bar=50 m. This experiment was repeated three times with similar results in each experiment.

Full figure and legend (101K)

Bmi-1 expression promotes keratinocyte proliferation

The above studies demonstrate that Bmi-1 is present in the nucleus of cultured keratinocytes and is also expressed in epidermis. To assess its functional role, Bmi-1 was overexpressed using an adenoviral vector. As shown in Figure 3a, infection of primary keratinocytes with tAd5-hBmi-1 results in a fivefold increase in intracellular Bmi-1 level. To assess the impact on cell number, keratinocytes were infected with tAd5-EV or tAd5-hBmi-1 and cell number was monitored after 72 hours. In tAd5-EV-infected cells, keratinocyte cell number increased threefold at 72 hours post infection owing to normal cell growth over the 3-day period. In contrast, cell number increased sevenfold in Bmi-1-expressing cultures (Figure 3b). To assess whether this increase in cell number is reflected at the level of DNA synthesis, we monitored [³H]thymidine incorporation into the DNA of cells treated as above. As shown in Figure 3c, the increase in cell number is associated with a parallel increase in DNA synthesis, indicating that more cells are progressing through S phase in the Bmi-1-expressing cells.

Figure 3.

Bmi-1 expression increases keratinocyte cell number. (a) Thirty percent confluent 50 cm² dishes of normal human keratinocytes were infected with 5 MOI of tAd5-hBmi-1 or tAd5-EV in the presence of 5 MOI Ad5-TA virus. After 24 hours, fresh virus-free medium was added for an additional 24 hours. At 48 hours post infection, the cells were harvested and extracts were prepared for electrophoresis and immunoblot with anti-Bmi-1. -actin was detected in parallel as a protein normalization control. (b) Keratinocytes were treated with virus as indicated above, and at 48 hours post infection they were harvested and counted. The open bar (control) is the cell number on day zero at the time of initiation of virus treatment and the shaded bars are the cell counts at 72 hours post infection. Note that the cells grow at a normal rate in the presence of tAd5-EV, such that cell number has increased several fold by 72 hours. As assessed by the Student's t-text, the difference between the means of the tAd5-EV and tAd5-hBmi-1 groups is significant at the 95% confidence interval. (c) Keratinocytes were treated with virus as above, and during the final 2 hours of the 48 hours growth period they were pulsed with 1 Ci/ml [³H]thymidine. The cells were then washed twice with ice-cold PBS, twice with ice-cold 5% (w/v) trichloroacetic acid, and then dissolved in 0.2 M NaOH. Incorporated [³H]thymidine was measured by liquid scintillation counting. Identical results were observed in each of the three experiments.

Full figure and legend (63K)

Bmi-1 protects keratinocytes from challenge with stress agents

We next monitored the impact of Bmi-1 on keratinocyte survival following stress agent challenge. Keratinocytes were infected with tAd5-EV or tAd5-hBmi-1, and, after 24 hours, treated with 12-O-tetradecanoylphorbol-13-acetate (TPA) or OA. At 48 hours post infection, the cells were examined. Our previous studies show that treatment with TPA or OA reduces keratinocyte survival (Welter et al., 1995; Efimova et al., 2003). As shown in Figure 4a, expression of Bmi-1 produces a marked increase in the survival of TPA-treated keratinocytes. Figure 4b shows the impact of Bmi-1 on the response to treatment with OA. Figure 4c provides a quantitative assessment of the change in cell number for the treatments shown in Figure 4a and b. This experiment shows that the reduction in cell survival observed in cells treated with 2 or 10 nM can be reversed by Bmi-1; however, the Bmi-1-dependent protection is less when the cells are challenged with the higher (10 nM) OA concentration. This suggests, as could reasonably be expected, that the ability of Bmi-1 to protect cells is limited and that high concentrations of stress agent may overwhelm the protective effects of Bmi-1.

Figure 4.

Bmi-1-expressing keratinocytes are resistant to stress-induced cell death. (a) Twenty percent confluent cultures of normal human keratinocytes were infected with 5 MOI of tAd5-EV (EV) or tAd5-hBmi-1 (Bmi-1) in the presence of 5 MOI of Ad5-TA. At 24 hours post infection, fresh virus-free medium, containing the indicated concentration of TPA, was added and incubation was continued for an additional 24 hours before the cells were photographed. (b) Twenty percent confluent keratinocyte cultures were treated for 24 hours with virus as outlined above followed by an additional 24 hours in virus-free medium containing 2 or 10 nM OA. The cells were then photographed. (c) Graphical analysis of the change in cell number in the absence or presence of Bmi-1 and challenge with TPA or OA. The experiment shown in panels a and b was repeated three times and the data were plotted. As assessed by the Student's t-text, the difference between the means of the tAd5-EV and tAd5-hBmi-1 groups is significant for each treatment (50 ng/ml TPA, 2 nM OA, and 10 nM OA) at the 95% confidence interval.

Full figure and legend (153K)

Bmi-1-dependent alteration in cell cycle regulatory protein expression

Previous studies in other systems suggest that Bmi-1 may enhance cell survival by regulating the level of cell cycle regulatory proteins (Jacobs et al., 1999a; Molofsky et al., 2003; Park et al., 2003; Kim et al., 2004). We therefore determined whether Bmi-1 produces this response. As shown in Figure 5a, hBmi-1 expression does not alter the expression of the cyclin-dependent kinase inhibitor, p21. In addition, p16^Ink4a level, which is known to be reduced by Bmi-1 in other systems (Jacobs et al., 1999a; Molofsky et al., 2003; Park et al., 2003), is not regulated by Bmi-1 in keratinocytes grown under the present conditions (Figure 5a). The absence of modulation of Bmi-1 level is consistent with a recent report in SCC-4 oral cancer cells, where cell growth was slowed by siRNA knockdown of Bmi-1 expression, but this was not associated with reduced p16^Ink4a level (Kang et al., 2007). In contrast, Bmi-1 expression does increase expression of the cyclin-dependent kinases cdk2 and cdk4, and the level of cyclin D1. These findings suggest that Bmi-1 can influence cell cycle progression via regulation of a host of cell cycle regulators.

Figure 5.

Bmi-1 enhances keratinocyte cell survival. (a) Keratinocytes were infected with 5 MOI of tAd5-EV or tAd5-hBmi-1 in the presence of 5 MOI of Ad5-TA. After 24 hours, fresh virus-free medium was added and the cells were incubated for an additional 24 hours. Total extract was then prepared and electrophoresed for immunoblot to detect the presence of the indicated proteins. The -actin immunoblot was included to assure equivalent protein loading. (b) Keratinocytes were infected with 5 MOI of tAd5-EV or tAd5-hBmi-1 for 24 hours. Fresh virus-free medium was then added and the cells were incubated with 2 nM OA for an additional 24 or 48 hours. Total extract was then prepared and electrophoresed for immunoblot with anti-PARP or anti-pro-caspase 3. A -actin immunoblot was included to assure equivalent protein loading. (c) Keratinocytes were infected with adenovirus as outlined above, and after 24 hours they were treated with 50 mJ/cm² UVB. After an additional 48 hours, the cells were harvested and assayed for PARP and pro-caspase 3 cleavage products by immunoblot. The arrows indicate the intact forms of PARP and pro-caspase 3. The asterisks indicate migration of cleaved forms. This experiment was repeated four times with similar results in each repeat.

Full figure and legend (63K)

A second mechanism whereby Bmi-1 may enhance cell survival is by inhibition of apoptosis. In keratinocytes, apoptosis involves the loss of mitochondrial membrane potential with accompanying release of apoptogenic factors, including cytochrome c, and activation of cysteine-dependent aspartate-directed proteases known as caspases (Denning et al., 2002; Waterhouse et al., 2002; Scorrano and Korsmeyer, 2003). Poly(ADP-ribose) polymerase (PARP), a DNA nick-sensing protein that is involved in DNA repair and is a downstream target of caspase-3, is also cleaved (Scorrano and Korsmeyer, 2003). To assess whether Bmi-1 influences apoptosis, we monitored the extent of pro-caspase and PARP cleavage in cells expressing Bmi-1. Keratinocytes were treated with tAd5-EV or tAd5-hBmi-1 for 24 hours, followed by treatment with OA in virus-free medium for an additional 24 or 48 hours. At 24 and 48 hours post infection, the cells were harvested and assayed for pro-caspase 3 and PARP cleavage (Figure 5b). Cleavage of these proteins is associated with their activation (Adams and Cory, 2002; Shi, 2002). OA treatment of EV-infected cells produces a marked increase in pro-caspase 3 and PARP cleavage products at 24 and 48 hours. In contrast, the presence of Bmi-1 inhibits production of these cleavage products. Ultraviolet light exposure is a particularly important apoptotic stimulus in epidermis that results in the production of apoptotic "sunburned" cells (Bayerl et al., 1995). To assess the impact of Bmi-1 expression on the cellular response to this physiologic apoptotic stimulus, cells were infected with tAd-EV or tAd-hBmi-1 for 24 hours, exposed to 50 mJ/cm² UVB, and then harvested at 48 hours post infection for preparation of cell extracts. As shown in Figure 5b, Bmi-1 expression results in a substantial reduction in UVB-dependent pro-caspase 3 and PARP cleavage. These findings suggest that Bmi-1 inhibits apoptosis.

Bmi-1 expression in skin tumors and in involved psoriatic tissue

As Bmi-1 expression has been reported to immortalize cells (Dimri et al., 2002), and Bmi-1 overexpression has been reported in tumors (Vonlanthen et al., 2001; Kim et al., 2004), we hypothesized that increased Bmi-1 expression may be associated with hyperproliferative epidermal diseases, including psoriasis and cancer. We first examined its expression in transformed keratinocyte-derived cell lines. As shown in Figure 6a, Bmi-1 mRNA is readily detected in A431, HaCaT, and SCC-13 cells. Moreover, the level of expression is markedly increased compared to that observed in normal foreskin keratinocytes (KER_N). To assess the expression pattern in squamous cell carcinoma, we stained tumor sections with anti-Bmi-1. As shown in Figure 6b, Bmi-1 is present in invasive squamous cell carcinoma. Staining is evident throughout the tumor, with less staining apparent at the tumor margins (+anti-Bmi-1). No staining is observed when the primary antibody is omitted (-anti-Bmi-1).

Figure 6.

Bmi-1 is expressed in squamous cell tumors and expression is increased in cancer cell lines. (a) Bmi-1 is overexpressed in tumor cell lines. Total protein extracts were prepared from near-confluent cultures of A431, HaCaT, SCC13 cells, and normal human keratinocytes, and electrophoresed. The resulting blots were incubated with anti-Bmi-1 or anti--actin. (b) Bmi-1 is present in tumor tissue. Tissue sections, derived from a human squamous cell carcinoma, were incubated with secondary antibody alone (-anti-Bmi-1) or in the presence of both primary and secondary antibody (+anti-Bmi-1). Bar=50 m. Bmi-1 is detected in all areas, but not along the margins where the tumor meets the connective tissue.

Full figure and legend (136K)

To assess whether Bmi-1 is expressed in other epidermal hyperproliferative diseases, we measured Bmi-1 expression in psoriasis. As shown in Figure 7a, incubation with anti-Bmi-1 reveals substantial staining in psoriatic tissue (+anti-Bmi-1). This staining is not observed in sections incubated in the absence of primary antibody (- anti-Bmi-1). To study the association between increased Bmi-1 level and increased proliferation, we transplanted human involved psoriatic tissue onto severe combined immunodeficiency (SCID) mice and monitored the level and distribution of Bmi-1 before and after treatment with the immunosuppressive agent, cyclosporine (Nickoloff, 1999, 2000). Cyclosporine treatment normalizes involved psoriatic tissue implanted onto SCID mice (Dam et al., 1999). As shown in Figure 7b, the involved psoriatic tissue displays the characteristic epithelial expansion into the underlying dermis (left panel). Moreover, Bmi-1 staining reveals the presence of focal Bmi-1 expression (arrows). In addition, treatment with cyclosporine normalizes the tissue morphology and reduces Bmi-1 expression. Although Bmi-1 is present at scattered locations in the normalized tissue, the levels are reduced compared to that observed in involved psoriatic tissue. An interesting feature is that the highest level of staining is observed in the suprabasal layers in the tissue, a pattern that is similar to that observed in normal epidermis (see Figure 2c).

Figure 7.

Bmi-1 is expressed in human psoriatic tissue and expression is reduced following tissue normalization. (a) Bmi-1 expression is elevated in psoriasis. Punch biopsies were derived from involved psoriatic tissue and incubated with secondary antibody alone (-anti-Bmi-1) or in the presence of both primary and secondary antibody (+anti-Bmi-1). Bars=50 m. The arrows indicate Bmi-1 staining. (b) Normalization of psoriatic phenotype is associated with reduced Bmi-1 expression. Human involved psoriatic tissue was transplanted onto SCID mice. Parallel grafts were treated with or without cyclosporine A for 3 weeks. The tissue was then harvested, fixed, and stained with anti-Bmi-1. Bars=50m. The arrows indicate Bmi-1 staining. The extent of the epidermis is indicated in parentheses (Epi).

Full figure and legend (127K)

Discussion

The PcG genes encode a family of evolutionarily conserved regulators that were discovered in Drosophila as repressors of Homoeotic genes, which are involved in establishing body segmentation patterns during development. In mammalian systems, PcG proteins regulate genes involved in development and differentiation via epigenetic mechanisms (e.g., chromatin modification). PcG genes are important regulators of organogenesis and development, as the reprogramming of stem cells requires epigenetic alterations (e.g., methylation) to specify and maintain long-term changes in gene expression. Proteins encoded by PcG genes comprise two distinct protein complexes that act coordinately to regulate gene expression – the "Bmi-1 complex" and the "Eed complex". The Eed complex includes Eed, EzH1, and EzH2. The Bmi-1 complex includes Bmi-1, Mel-18, Mph1/Rae28, M33, Scmh1, and Ring1A/B. The Eed-containing complex controls gene repression through recruitment of histone deacetylase. This recruitment leads to local chromatin deacetylation and subsequent methylation of Lys₂₇ of histone H3. The Bmi-1 complex subsequently binds to methylated Lys₂₇ of histone H3, created by Eed complex action, to suppress gene expression and contribute to the maintenance of epigenetic memory (Fischle et al., 2003). Thus, these complexes act coordinately to maintain changes in gene expression during development (Jacobs and van, 2002; Orlando, 2003).

Bmi-1 is expressed in keratinocytes

Because the epidermis is a renewing tissue and the survival of differentiating keratinocytes must be controlled to produce a normal epidermal barrier, we assessed whether Bmi-1 is present and if it has a regulatory role in keratinocytes. Bmi-1 is localized in the nucleus of cultured normal human epidermal keratinocytes, consistent with a previous report indicating a nuclear localization in fibroblasts (Cohen et al., 1996). This localization appears to be necessary for function, as Bmi-1 mutants that cannot localize in the nucleus are inactive (Cohen et al., 1996).

Several studies suggest that the pro-survival, pro-proliferation action of Bmi-1 may be due to its ability to suppress expression of proteins that regulate cell cycle progression. For example, in some cell types, in the absence of Bmi-1, the levels of p16^ink4a and p19^arf increase (Jacobs et al., 1999a, 1999b). p16^ink4a and p19^arf are products of the ink4a-arf locus that function to suppress cell cycle progression. p16^ink4a inhibits cell cycle progression by inhibiting cyclin D-dependent kinases, and thereby prevents phosphorylation of Rb (Serrano et al., 1993). p19^arf (p14^arf in humans) prevents the degradation and inactivation of the p53 tumor suppressor protein by binding to MDM2 (Serrano et al., 1993; Weber et al., 1999). It is interesting that p16^Ink4a is expressed at low levels in keratinocytes cultured under the present conditions and that no change in the level of p16^INK4a is observed in response to Bmi-1 overexpression. This finding is consistent with recent studies showing that Bmi-1 expression does not alter p16^Ink4a expression in oral epithelial cells (Kang et al., 2007). Nor did we observe a change in p21 level. However, changes were observed with other cell cycle regulatory proteins including increased cyclin D1, and cdk2 and cdk4 levels. This is consistent with a recent report suggesting that many Bmi-1 targets may exist (Bracken et al., 2006). Thus, our study suggests that Bmi-1 regulates several molecular targets that control progression through the G1 phase of the cell cycle.

Bmi-1 in adult tissues and malignancy

Bmi-1 function is required for maintenance of adult stem cells. Aging Bmi-1^-/- mice progressively lose stem cells in the leukemic, neuronal, and cerebellar granule cell lineages (Jacobs et al., 1999a; Lessard and Sauvageau, 2003; Molofsky et al., 2003; Park et al., 2003; Leung et al., 2004), and Bmi-1-deficient fibroblasts express elevated levels of p16^ink4a and p19^arf and proliferate slowly (Jacobs et al., 1999b). In addition, Bmi-1 overexpression enhances cell proliferation (Lessard and Sauvageau, 2003), Bmi-1 can cooperate with myc in lymphomagenesis (Haupt et al., 1991), and Bmi-1 is overexpressed in human colorectal cancer (Kim et al., 2004) and human non-small cell lung cancer (NSCLC) (Vonlanthen et al., 2001). These findings suggest that Bmi-1 has a role in regulating both stem cell and adult somatic cell proliferation and survival. Our findings are consistent with a role for Bmi-1 in enhancing the survival of mature keratinocytes. We also show that Bmi-1 is expressed in squamous cell carcinoma and is present in active psoriatic lesions, and is overexpressed in cultured skin cancer cells. Moreover, in a model in which human psoriatic tissue is engrafted onto the SCID mouse, Bmi-1 expression is observed in the active psoriatic tissue, but is reduced or absent when the tissue is normalized by treatment of the mice with cyclosporine. These findings suggest that Bmi-1 may play a role in the development and maintenance of hyperproliferative skin diseases, and is consistent with the observation that Bmi-1 is overexpressed in other tumor types (Vonlanthen et al., 2001; Dimri et al., 2002; Kim et al., 2004).

Bmi-1 expression in epidermis – a non-stem cell survival role for Bmi-1

Based on its reported role in regulating stem cell survival, we had anticipated that Bmi-1 may be specifically expressed in a limited number of cells in the proliferating compartments of the epidermis – in the epidermal basal layer or in the hair follicle bulge (Alonso and Fuchs, 2003; Lavker et al., 2003; Claudinot et al., 2005; Maurelli et al., 2006). Thus, we were surprised that although Bmi-1 is expressed in basal cells, staining is also observed in a scattered pattern in the suprabasal layers. This finding is consistent with the recent report of Reinisch et al. (2007). This is particularly interesting, as cells of the suprabasal layers (spinous and granular), although metabolically active and viable, have limited proliferative potential. These cells are undergoing differentiation, which is associated with the expression of a host of terminal differentiation products including involucrin, loricrin, filaggrin, etc. (Eckert et al., 1997). Ultimately, these cells survive until they reach the transition zone in the upper epidermis, where they undergo terminal differentiation and cornification (Eckert et al., 1997). This process ultimately leads to the enzymatic destruction of all intracellular structures, the assembly of keratin bundles, and the assembly of the cornified envelope (Green, 1980; Eckert et al., 1997). Based on the pattern of Bmi-1 expression in epidermis, we propose that one role of Bmi-1 is to assist in the maintenance of non-proliferative suprabasal keratinocytes to prevent premature death.

Bmi-1 and keratinocyte apoptosis

Our results also suggest that Bmi-1 may modulate the response of the epidermis to stress. Bmi-1 expression maintains survival of cells following challenge with stress agents, including OA, TPA, and UVB. UVB exposure is known to induce caspase-associated keratinocyte cell death, which ultimately results in the formation of sunburn cells – the products of light-induced apoptosis (Conney et al., 1992; Bayerl et al., 1995). These cells express activated killer and executioner caspases (Shimizu et al., 1999; Lu et al., 2002). Our studies show that Bmi-1 expression can reverse or slow this process. Only one report has addressed a possible role for Bmi-1 in regulating apoptosis. In this study, it was suggested that Bmi-1 acts to inhibit the c-myc-dependent increase in p19^ARF as a mechanism to inhibit myc-associated apoptosis (Jacobs et al., 1999b). Our studies show that expression of Bmi-1 attenuates the UVB-dependent increase in pro-caspase activation and PARP cleavage. Thus, the level of Bmi-1 present in the tissue may influence the tendency of the tissue to undergo apoptosis – that is, high levels of Bmi-1 may make the tissue resistant to light-induced apoptosis.

Thus, our studies show that Bmi-1 is expressed in cells of both the basal and suprabasal epidermal layers, with more intense staining in the suprabasal epidermis. It is particularly interesting that Bmi-1 is expressed in the suprabasal cells, as these cells have lost proliferative potential and are undergoing terminal differentiation. This suggests that Bmi-1 may function in epidermis to maintain survival and prevent premature terminal differentiation. It is also significant that Bmi-1 is expressed in a high percentage of cells within tumors and in involved psoriatic epidermis. In this context, the pro-survival function of Bmi-1 may promote cell survival and proliferation leading to tumor development. Its widespread expression in proliferating and non-proliferating tissues in epidermis suggests that Bmi-1 is not strictly a regulator of stem cell survival in this tissue. It will be important in future studies to address this hypothesis in vivo using animals that express varied levels of Bmi-1 in the epidermis.

Materials and Methods

Reagents and antibodies

OA was purchased from Calbiochem (San Diego, CA) and TPA was purchased from Sigma (Milwaukee, WI). Cyclosporine A was obtained from Ben Venue Laboratories (Bedford, OH). It is supplied USP 250 mg per 5 ml (NDC 55390-122-10). One milliliter contains 50 mg cyclosporine, 650 mg Cremophor, and is 33.2% (v/v) absolute alcohol. Rabbit polyclonal anti-caspase 3 (9665) was obtained from Cell Signaling (Beverly, MA). Mouse monoclonal antibodies, including anti-cyclin D1 (554180) and anti-PARP (no. 556494), were obtained from BD Pharmingen (San Diego, CA). Mouse monoclonal anti-Bmi-1 (ab14389) was purchased from Abcam Inc. (Cambridge, MA) and mouse monoclonal anti--actin was obtained from Sigma (A5441). Rabbit polyclonal antibodies against cdk2 (sc-163), cdk4 (sc-601), cyclin E (sc-481), a mouse monoclonal antibody specific for p21^cip1/waf1 (sc-6246), and a rabbit polyclonal antibody for p16^Ink4a (C-20) (sc-468) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Secondary antibodies, including HRP-conjugated sheep anti-mouse IgG (NA931) and HRP-conjugated donkey anti-rabbit IgG (NA934), were purchased from Amersham Biosciences (Arlington Heights, IL). Methyl-³H-thymidine was obtained from Amersham Biosciences (TRK300, 25 Ci/mmol, 1 mCi/ml).

RT-PCR analysis

RT-PCR was performed using conventional methods. The primers for Bmi-1 are 5'-GGTCTAGACGGATCCCAAGCAGAAATGCATCG and 5'-GGCTCGAGCATCTAGAAAGCTGTAATGGCATG. The primers for amplification of human -action were purchased from Stratagene, La Jolla, CA.

Bmi-1 adenovirus construction

The cDNA encoding human Bmi-1 (hBmi-1) (Alkema et al., 1993) was amplified by RT-PCR using Xba1 site-encoding primers and total RNA derived from primary cultures of human foreskin keratinocytes. The upstream primer was 5'-GGTCTAGACGGATCCCAAGCAGAAATGCATCG. This primer encodes XbaI and BamHI sites (underlined) and the sequence in bold is identical to hBmi-1 (Alkema et al., 1993). The hBmi-1 protein translation start codon is underlined. The downstream primer, 5'-GGCTCGAGCATCTAGAAAGCTGTAATGGCATG, is homologous to the non-coding sequence located downstream of the Bmi-1 coding sequence. The bold sequence is identical to Bmi-1 sequence and the XbaI and XhoI sites are underlined. The hBmi-1 cDNA product, approximately 1060 bp in length, was cloned into the shuttle plasmid, pE1sp1B, as an Xba1 fragment to create pE1sp1B-hBmi-1. A293 packaging cells were transfected with pE1sp1B-hBmi-1 and the pJM17 adenovirus backbone plasmid to produce the hBmi-1-encoding adenovirus, tAd5-hBmi-1. The empty adenovirus, tAd5-EV, and the Ad5-TA helper virus were constructed as previously described (Dashti et al., 2001).

Cell culture

Normal human foreskin keratinocytes were cultured in keratinocyte serum-free medium as previously described (Dashti et al., 2001). A431, HaCaT, and SCC13 cells were cultured in DMEM:F12 (3:1) containing 2 mM L-glutamine, 100 U/ml penicillin, 100 U/ml streptomycin, 10% fetal calf serum, and non-essential amino acids.

Cell proliferation studies

Normal human keratinocytes were seeded into 9.5 cm² dishes and allowed to attach for 24 hours. The cells were then infected with 5 MOI of tAd5-EV (empty vector) or tAd5-hBmi-1 in the presence of 5 MOI of Ad5-TA and 2.5 g/ml polybrene. Ad5-TA is a helper adenovirus that encodes the tetracycline activator (TA). In the native state, the TA protein activates transcription of the hBmi-1-encoding transcription unit. Addition of tetracycline to the medium inactivates TA and results in reduced levels of hBmi-1. At 24 hours post infection, fresh medium was added, and after an additional 48 hours, the cells were harvested using 0.025% trypsin/1 mM EDTA and counted.

Immunoblot analysis

Subconfluent cultures of cells were infected with adenoviruses as mentioned above for 24 hours. Fresh virus-free medium was added and the cells were treated with 0 or 2 nM OA (24 or 48 hours) or UVB (50 mJ/cm²) (FS20-T12 lamps with Kodacel filer, Ultraviolet Resources International, Brookpark, OH). The cells were washed once in phosphate-buffered saline (PBS), and lysates were prepared in 20 mM Tris-HCl (pH 7.5) containing 150 mM NaCl, 1 mM EGTA, 1 mM EDTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM -glycerophosphate, 1 mM sodium vanadate, 1 g of leupeptin, and 1 mM phenylmethylsulfonyl fluoride. Equal quantities of protein were electrophoresed on 10% denaturing and reducing polyacrylamide gels and transferred to nitrocellulose. The membranes were blocked in 10 mM Tris-HCl (pH 7.2) containing 100 mM NaCl, 0.1% Tween 20, and 5% non-fat dry milk, and incubated with the appropriate primary and secondary antibodies. Secondary antibody binding was visualized using a chemiluminescent detection system.

Immunohistological detection of Bmi-1 in keratinocytes

For detection in cultured cells, human foreskin keratinocytes, growing on coverslips, were fixed with 4% formaldehyde for 30 minutes, washed with PBS, and permeabilized for 30 minutes using PBS containing 0.1% Triton X-100 and 1% BSA. The coverslips were then incubated for 1 hour with mouse anti-Bmi-1 (1:100), washed in PBS containing 0.1% Triton X-100 and 1% BSA, and incubated for 30 minutes with Cy3-conjugated sheep anti-mouse IgG (1:1,000) for 30 minutes. The cells were then washed with PBS containing 0.1% Triton X-100 and 1% BSA and mounted in Mowiol 4-88-containing mounting agent. Bmi-1 distribution was monitored by epifluorescence microscopy. For detection of Bmi-1 in human skin biopsies, the tissue was harvested by punch biopsy, fixed in formalin, and incubated with murine monoclonal anti-Bmi-1 (1:100) followed by washing and incubation with HRP-conjugated goat anti-mouse IgG (1:1,000). Antibody binding was visualized using the Vectastain kit.

The human psoriasis skin-engrafted SCID mouse model

Keratome biopsy samples (0.05 cm thick) were collected from involved psoriatic plaques from patients. For collection, three or four individual graft segments (1.0–1.5 cm²) were divided out from each patient biopsy, and a single small trim piece of each slice was used for initial histological characterization. Within 24 hours after collection, the tissue segments were engrafted onto 8-week-old female CB-17 SCID mice (Taconic). In this method, the mice are anesthetized and a 1.0 cm² segment of skin was removed from the lateral thorax. The 1.0–1.5 cm² human skin graft is then secured at this location using a band-aid strip, after applying sterile Vaseline-impregnated gauze. Three to four mice receive tissue from the same patient, and individual mice are normalized to different experimental treatment groups. Approximately 2 weeks are allowed for recovery before initiation of cyclosporine treatment. The mice are treated twice weekly for 3 weeks with or without intra-graft injection of 0.15 mg of cyclosporine A in a 50 l volume (Dam et al., 1999). At 24 days after initiation of cyclosporine A treatment, the grafts are collected for histological analysis and immunostaining with anti-Bmi-1. These studies were approved by the institutional animal and human use boards and all studies were performed according to the Declaration of Helsinki Principles. The patients' written, informed consent was obtained for collection of all tissue samples.

Conflict of Interest

The authors state no conflict of interest.

References

1. Adams JM, Cory S (2002) Apoptosomes: engines for caspase activation. Curr Opin Cell Biol 14:715–720 | Article | PubMed | ISI | ChemPort |
2. Alkema MJ, Wiegant J, Raap AK, Berns A, van LM (1993) Characterization and chromosomal localization of the human proto-oncogene BMI-1. Hum Mol Genet 2:1597–1603 | Article | PubMed | ISI | ChemPort |
3. Alonso L, Fuchs E (2003) Stem cells of the skin epithelium. Proc Natl Acad Sci USA 100(Suppl 1):11830–11835 | Article | PubMed | ChemPort |
4. Bayerl C, Taake S, Moll I, Jung EG (1995) Characterization of sunburn cells after exposure to ultraviolet light. Photodermatol Photoimmunol Photomed 11:149–154 | PubMed | ISI | ChemPort |
5. Bracken AP, Dietrich N, Pasini D, Hansen KH, Helin K (2006) Genome-wide mapping of Polycomb target genes unravels their roles in cell fate transitions. Genes Dev 20:1123–1136 | Article | PubMed | ISI | ChemPort |
6. Claudinot S, Nicolas M, Oshima H, Rochat A, Barrandon Y (2005) Long-term renewal of hair follicles from clonogenic multipotent stem cells. Proc Natl Acad Sci USA 102:14677–14682 | Article | PubMed | ChemPort |
7. Cohen KJ, Hanna JS, Prescott JE, Dang CV (1996) Transformation by the Bmi-1 oncoprotein correlates with its subnuclear localization but not its transcriptional suppression activity. Mol Cell Biol 16:5527–5535 | PubMed | ISI | ChemPort |
8. Conney AH, Wang ZY, Huang MT, Ho CT, Yang CS (1992) Inhibitory effect of green tea on tumorigenesis by chemicals and ultraviolet light. Prev Med 21:361–369 | Article | PubMed | ISI | ChemPort |
9. Dam TN, Kang S, Nickoloff BJ, Voorhees JJ (1999) 1alpha, 25-dihydroxycholecalciferol and cyclosporine suppress induction and promote resolution of psoriasis in human skin grafts transplanted on to SCID mice. J Invest Dermatol 113:1082–1089 | Article | PubMed | ISI | ChemPort |
10. Dashti SR, Efimova T, Eckert RL (2001) MEK7-dependent activation of p38 MAP kinase in keratinocytes. J Biol Chem 276:8059–8063 | Article | PubMed | ISI | ChemPort |
11. Denning MF, Wang Y, Tibudan S, Alkan S, Nickoloff BJ, Qin JZ (2002) Caspase activation and disruption of mitochondrial membrane potential during UV radiation-induced apoptosis of human keratinocytes requires activation of protein kinase C. Cell Death Differ 9:40–52 | Article | PubMed | ISI | ChemPort |
12. Dimri GP, Martinez JL, Jacobs JJ, Keblusek P, Itahana K, van LM et al. (2002) The Bmi-1 oncogene induces telomerase activity and immortalizes human mammary epithelial cells. Cancer Res 62:4736–4745 | PubMed | ISI | ChemPort |
13. Eckert RL, Crish JF, Robinson NA (1997) The epidermal keratinocyte as a model for the study of gene regulation and cell differentiation. Physiol Rev 77:397–424 | PubMed | ISI | ChemPort |
14. Efimova T, Broome AM, Eckert RL (2003) A regulatory role for p38 delta MAPK in keratinocyte differentiation. Evidence for p38 delta-ERK1/2 complex formation. J Biol Chem 278:34277–34285 | Article | PubMed | ISI | ChemPort |
15. Fischle W, Wang Y, Jacobs SA, Kim Y, Allis CD, Khorasanizadeh S (2003) Molecular basis for the discrimination of repressive methyl-lysine marks in histone H3 by Polycomb and HP1 chromodomains. Genes Dev 17:1870–1881 | Article | PubMed | ISI | ChemPort |
16. Fuchs E, Byrne C (1994) The epidermis: rising to the surface. Curr Opin Genet Dev 4:725–736 | Article | PubMed | ChemPort |
17. Green H (1980) The keratinocyte as differentiated cell type. Harvey Lect 74:101–139 | PubMed | ChemPort |
18. Haupt Y, Alexander WS, Barri G, Klinken SP, Adams JM (1991) Novel zinc finger gene implicated as myc collaborator by retrovirally accelerated lymphomagenesis in E mu-myc transgenic mice. Cell 65:753–763 | Article | PubMed | ISI | ChemPort |
19. Iwama A, Oguro H, Negishi M, Kato Y, Morita Y, Tsukui H et al. (2004) Enhanced self-renewal of hematopoietic stem cells mediated by the polycomb gene product Bmi-1. Immunity 21:843–851 | Article | PubMed | ISI | ChemPort |
20. Jacobs JJ, Kieboom K, Marino S, DePinho RA, van LM (1999a) The oncogene and Polycomb-group gene bmi-1 regulates cell proliferation and senescence through the ink4a locus. Nature 397:164–168 | Article | PubMed | ISI | ChemPort |
21. Jacobs JJ, Scheijen B, Voncken JW, Kieboom K, Berns A, van LM (1999b) Bmi-1 collaborates with c-Myc in tumorigenesis by inhibiting c-Myc-induced apoptosis via INK4a/ARF. Genes Dev 13:2678–2690 | Article | PubMed | ISI | ChemPort |
22. Jacobs JJ, van LM (2002) Polycomb repression: from cellular memory to cellular proliferation and cancer. Biochim Biophys Acta 1602:151–161 | Article | PubMed | ISI | ChemPort |
23. Kang MK, Kim RH, Kim SJ, Yip FK, Shin KH, Dimri GP et al. (2007) Elevated Bmi-1 expression is associated with dysplastic cell transformation during oral carcinogenesis and is required for cancer cell replication and survival. Br J Cancer 96:126–133 | Article | PubMed | ISI | ChemPort |
24. Kim JH, Yoon SY, Kim CN, Joo JH, Moon SK, Choe IS et al. (2004) The Bmi-1 oncoprotein is overexpressed in human colorectal cancer and correlates with the reduced p16INK4a/p14ARF proteins. Cancer Lett 203:217–224 | Article | PubMed | ISI | ChemPort |
25. Lavker RM, Sun TT, Oshima H, Barrandon Y, Akiyama M, Ferraris C et al. (2003) Hair follicle stem cells. J Investig Dermatol Symp Proc 8:28–38 | Article | PubMed | ISI |
26. Lessard J, Sauvageau G (2003) Bmi-1 determines the proliferative capacity of normal and leukaemic stem cells. Nature 423:255–260 | Article | PubMed | ISI | ChemPort |
27. Lessard J, Schumacher A, Thorsteinsdottir U, van LM, Magnuson T, Sauvageau G (1999) Functional antagonism of the Polycomb-Group genes eed and Bmi1 in hemopoietic cell proliferation. Genes Dev 13:2691–2703 | Article | PubMed | ISI | ChemPort |
28. Leung C, Lingbeek M, Shakhova O, Liu J, Tanger E, Saremaslani P et al. (2004) Bmi1 is essential for cerebellar development and is overexpressed in human medulloblastomas. Nature 428:337–341 | Article | PubMed | ISI | ChemPort |
29. Lu YP, Lou YR, Li XH, Xie JG, Lin Y, Shih WJ et al. (2002) Stimulatory effect of topical application of caffeine on UVB-induced apoptosis in mouse skin. Oncol Res 13:61–70 | PubMed | ISI | ChemPort |
30. Maurelli R, Zambruno G, Guerra L, Abbruzzese C, Dimri G, Gellini M et al. (2006) Inactivation of p16INK4a (inhibitor of cyclin-dependent kinase 4A) immortalizes primary human keratinocytes by maintaining cells in the stem cell compartment. FASEB J 20:1516–1518 | Article | PubMed | ISI | ChemPort |
31. Molofsky AV, Pardal R, Iwashita T, Park IK, Clarke MF, Morrison SJ (2003) Bmi-1 dependence distinguishes neural stem cell self-renewal from progenitor proliferation. Nature 425:962–967 | Article | PubMed | ISI | ChemPort |
32. Nickoloff BJ (1999) Animal models of psoriasis. Expert Opin Investig Drugs 8:393–401 | Article | PubMed | ChemPort |
33. Nickoloff BJ (2000) Characterization of lymphocyte-dependent angiogenesis using a SCID mouse: human skin model of psoriasis. J Investig Dermatol Symp Proc 5:67–73 | Article | PubMed | ISI | ChemPort |
34. Ohta H, Sawada A, Kim JY, Tokimasa S, Nishiguchi S, Humphries RK et al. (2002) Polycomb group gene rae28 is required for sustaining activity of hematopoietic stem cells. J Exp Med 195:759–770 | Article | PubMed | ISI | ChemPort |
35. Orlando V (2003) Polycomb, epigenomes, and control of cell identity. Cell 112:599–606 | Article | PubMed | ISI | ChemPort |
36. Park IK, Qian D, Kiel M, Becker MW, Pihalja M, Weissman IL et al. (2003) Bmi-1 is required for maintenance of adult self-renewing haematopoietic stem cells. Nature 423:302–305 | Article | PubMed | ISI | ChemPort |
37. Reinisch CM, Uthman A, Erovic BM, Pammer J (2007) Expression of BMI-1 in normal skin and inflammatory and neoplastic skin lesions. J Cutan Pathol 34:174–180 | Article | PubMed | ISI |
38. Scorrano L, Korsmeyer SJ (2003) Mechanisms of cytochrome c release by proapoptotic BCL-2 family members. Biochem Biophys Res Commun 304:437–444 | Article | PubMed | ISI | ChemPort |
39. Serrano M, Hannon GJ, Beach D (1993) A new regulatory motif in cell-cycle control causing specific inhibition of cyclin D/CDK4. Nature 366:704–707 | Article | PubMed | ISI | ChemPort |
40. Shi Y (2002) Mechanisms of caspase activation and inhibition during apoptosis. Mol Cell 9:459–470 | Article | PubMed | ISI | ChemPort |
41. Shimizu H, Banno Y, Sumi N, Naganawa T, Kitajima Y, Nozawa Y (1999) Activation of p38 mitogen-activated protein kinase and caspases in UVB-induced apoptosis of human keratinocyte HaCaT cells. J Invest Dermatol 112:769–774 | Article | PubMed | ISI | ChemPort |
42. Vonlanthen S, Heighway J, Altermatt HJ, Gugger M, Kappeler A, Borner MM et al. (2001) The bmi-1 oncoprotein is differentially expressed in non-small cell lung cancer and correlates with INK4A-ARF locus expression. Br J Cancer 84:1372–1376 | Article | PubMed | ISI | ChemPort |
43. Waterhouse NJ, Ricci JE, Green DR (2002) And all of a sudden it's over: mitochondrial outer-membrane permeabilization in apoptosis. Biochimie 84:113–121 | Article | PubMed | ISI | ChemPort |
44. Weber JD, Taylor LJ, Roussel MF, Sherr CJ, Bar-Sagi D (1999) Nucleolar Arf sequesters Mdm2 and activates p53. Nat Cell Biol 1:20–26 | Article | PubMed | ISI | ChemPort |
45. Welter JF, Crish JF, Agarwal C, Eckert RL (1995) Fos-related antigen (Fra-1), junB, and junD activate human involucrin promoter transcription by binding to proximal and distal AP1 sites to mediate phorbol ester effects on promoter activity. J Biol Chem 270:12614–12622 | Article | PubMed | ISI | ChemPort |

Acknowledgments

This work utilized the facilities of the Skin Diseases Research Center of Northeast Ohio (NIH, AR39750) and was supported by a grant from the National Institutes of Health (RLE). Dr Adhikary was a trainee on the Visual Sciences Training Grant.