Erb-041, an Estrogen Receptor-β Agonist, Inhibits Skin Photocarcinogenesis in SKH-1 Hairless Mice by Downregulating the WNT Signaling Pathway

See related commentary by Yao et al., p. 182

Nonmelanoma skin cancers (NMSC), which include basal cell carcinomas and squamous cell carcinoma (SCC), are the most commonly diagnosed cancers in the United States. Their incidence exceeds the combined incidence of cancers of the breast, prostate, lung, and colon (1). UVB radiation (280–320 nm) from the sun and tanning beds are the main etiologic cause of skin cancer (2). UVB induces DNA damage, inflammatory response, and alters multiple cell signaling events, which altogether lead to initiation, promotion, and progression of epidermal neoplasm (3). During the past decade, a number of attempts have been made to understand the pathogenesis of these cancers and to identify novel molecular targets to intervene the disease progression. In this regard, we and others have demonstrated the involvement of p53, ornithine decarboxylase, cyclooxygenases, retinoid receptor signaling, oxidative stress, etc., besides many others in the molecular pathogenesis of these cancers (3–8). Strategies have also been developed to modify these targets to prevent NMSCs both in humans and in experimental animals (5, 9, 10). However, these approaches have been only partially successful.

The modulation of estrogen receptor (ER) activity has proved therapeutically valuable for the treatment of various epithelial cancers in experimental models (11–15). The ERs exist in two distinct forms ER-α and ER-β. Their splice variants, which are also biologically active, have been identified (13). Estrogens exert their tissue-specific responses via ER-α or ER-β or their splice variants by activating diverse signaling pathways that mediate both genomic and nongenomic events (11). It is interesting that despite remarkable similarities in the two receptors, ER-α and ER-β are often antagonistic in nature. Altered ratio of ER-α/ER-β in a cell is the major determinant of responses of the cell to estrogen. ER-α/ER-β–mediated activation or deactivation is dependent on the effects of coactivator and corepressor proteins on estrogen-responsive element (14, 15).

ER-β is a member of the nuclear receptor superfamily (13) and is produced from eight exons. Upon ligand activation, it regulates gene expression by modulating transcription factors, such as NF-κB, activating protein-1 (AP-1), and stimulating protein-1 (SP-1) through transcription factor cross talk (16, 17). The nongenomic effects of ER-β are regulated by the activation of protein kinases (A and C) and MAPK (mitogen-activated protein kinase) signaling pathways (18). The expression of ER-β is considered an important determinant of tumor phenotype and has also been suggested as a useful biomarker in the rheumatoid disease progression (19). ER-β–selective agonists have been shown to possess anticarcinogenetic and anti-inflammatory properties in experimental model systems (20, 21). Loss of ER-β expression has been reported in various cancers, such as prostate, colorectal, thyroid carcinoma, etc. (22–24). Methylation of CpG islands in the promoter of ER-β is considered as one of the putative mechanisms involved in the loss of its expression (25). Erb-041, a selective ER-β agonist, has been reported to possess strong anti-inflammatory activity and is under clinical trial for its potential use in rheumatoid arthritis (20, 26, 27).

In this study, we investigated the cancer chemopreventive effects of Erb-041 on the UVB-induced skin photocarcinogenesis using SKH-1 hairless mice. We observed a potent cancer chemopreventive activity of Erb-041 in this experimental animal model. Erb-041 affects the growth of UVB-induced murine SCCs. We show that the mechanism by which this ER-β agonist manifests cancer chemopreventive effects involves inhibition of the WNT/β-catenin–dependent signaling pathway.

Erb-041 (C₁₅H₁₀FNO₃) was procured from IRIX Pharmaceuticals Inc. and was supplied by the NCI chemoprevention branch. Details of antibodies used in this study are provided as Supplementary Table S1.

Fresh skin tumor samples were collected according to our approved Institutional Review Board (IRB) protocol (N081204004) for undesignated samples. Human samples were carefully handled according to IRB guidelines.

Six- to 8-week-old SKH-1 hairless female mice were used for this study. Animals were housed in groups of five in each cage under conditions of constant temperature of 24°C ± 2°C and relative humidity of 50% ± 10%, and were maintained on a 12 hours light/12 hours dark cycle with food and drinking water ad libitum. The animal studies described here were approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Alabama at Birmingham (Birmingham, AL).

Human immortalized keratinocyte (HaCaT) and human epidermoid carcinoma (A431) cells were purchased from the American Type Culture Corporation and SCC13 cells were gifted by Dr. S. K. Katiyar (University of Alabama at Birmingham, Birmingham, AL). These cells were routinely cultured in the recommended growth medium containing 10% FBS, 100 U/mL of penicillin, and 100 μg/mL of streptomycin in humidified incubators at 37°C under 5% CO₂. Cells (60%–70% confluent) were treated with Erb-041 or WNT signaling inhibitor or vehicle (dimethyl sulfoxide, DMSO) in complete culture medium. After 24 hours of treatment, medium was removed and the cells were washed and harvested to prepare cell lysates.

The UVB light source was a UVA/UVB Research Irradiation Unit (Daavlin Co.), which is fitted with an electronic controller to regulate dose of irradiated UVB. The UVB irradiation procedure was identical to that described earlier (7).

Animals were randomly divided into three groups of 20 mice each. Group I animals received topical treatment with ethanol and served as age-matched vehicle control (negative control). Group II and III animals were irradiated with UVB (180 mJ/cm²; twice/week) for 30 weeks. In addition, group II received vehicle and group III animals received topical treatments with Erb-041 (2 mg/mouse in 200 μL ethanol), 30 minutes before UVB irradiation. The tumor number and size were recorded weekly using electronic Vernier Caliper as described earlier (7). Data were presented as mean ± SE and plotted as a function of weeks on test. After 30 weeks, the experiment was terminated and all mice were euthanized as per the IACUC recommendations. Skin and tumor tissues were harvested and processed for histologic and biochemical analysis as described in the following sections.

Of note, 10% neutral-buffered formalin-fixed tissues were embedded, and cut in the serial sections of 5 μm. For histologic evaluation, tissues were stained with hematoxylin and eosin (H&E). Immunohistochemical and immunofluorescence staining were performed as described earlier (7). The Vector Red Alkaline Phosphatase Substrate Kit (Cat no. SK5100) was also used according to the manufacturer's guidelines for immunohistochemistry (IHC). TUNEL (terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling) assay was done using an in situ cell death detection, fluorescein kit from Roche Applied Science (Cat. no.1684795) following the manufacturer's guidelines.

Myeloperoxidase activity in the skin samples was determined as described earlier (28). The change in absorbance was recorded at 460 nm using a PerkinElmer 1420 Multilabel Counter Victor 3. The data are expressed as mean myeloperoxidase U/mg protein per minute.

Tissues were lysed in ice-cold lysis buffer containing 50 mmol/L Tris pH, 1% Triton X 100, 0.25% NaF, 10 mmol/L β-glycerophosphate, 1 mmol/L EDTA, 5 mmol/L sodium pyrophosphate, 0.5 mmol/L Na₃VO₄, 10 mmol/L dithiothreitol, 1% phenylmethylsulfonylfluoride, and protease inhibitors cocktail. For Western blot analysis, proteins (60–80 μg) were resolved on 10% to 15% SDS–PAGE and transferred onto a nitrocellulose membrane (Bio-Rad) as described previously (7). Membrane was stripped and reprobed with anti-β-actin antibody to confirm equal protein loading. In instances, in which a blot was stripped multiple times and probed with different antibodies but the data are presented as a part of more than one figure, the same β-actin image was placed multiple times to represent loading controls in the figures.

Extraction of total RNA, cDNA preparation and real-time PCR (RT-PCR) were performed as described previously (29). Relative quantification of the steady-state target mRNA levels was calculated after normalization of total amount of cDNA to GAPDH (glyceraldehyde-3-phosphate dehydrogenase) endogenous reference. The list of primers used in this study is described in Supplementary Table S2.

A431 and SCC13 cells were treated with and without Erb-041 for 0, 24, 36, and 48 hours. The cells were trypsinized, washed, and fixed with ice-cold 70% ethanol at −20°C overnight. Thereafter, the cells were washed and incubated with 20 mg/mL RNase A and 200 mg/mL propidium iodide in PBS at room temperature for 30 minutes, and subjected to flow cytometry using the BD Accuri C6 or FACSCalibur flow cytometer. Cell-cycle distribution was analyzed and provided as percentage of the G₁, S, and G₂–M phase of cells.

A431 and SCC13 cells (500 cells/well) were seeded into 6-well plates and were allowed to grow overnight. Cells were treated with and without Erb-041 for 24 hours and incubated in humidified chamber at 37°C for additional 10 days. Cell colonies were fixed with 4% paraformaldehyde for 5 minutes and stained with 0.5% crystal violet for 30 seconds, and cell colonies were counted (30).

Briefly, A431 and SCC13 cells were allowed to grow up to 90% to 100% confluence, and a fine scratch was made using a sterile pipette tip. Then, these cells were treated with and without Erb-041 and incubated at 37°C for 24 hours. The cell motility was observed at 12 and 24 hours using an Olympus CK2 microscope with Olympus DP20 digital camera.

HaCaT, A431, and SCC13 cells were grown in 24-well plate on round glass cover slips with or without Erb-041 slides. The cells were fixed with 4% paraformaldehyde for 15 minutes at RT. Cells were permeabilized and blocked with 1% bovine serum albumin, 10% goat serum, 0.3 mol/L glycine, and 0.1% Tween X for 1 hour at RT. Then, cells were incubated with primary antibodies for 2 hours at RT. After washing, the cells were incubated with appropriate Dylight 488 or Alexa Fluor 594 secondary antibodies for 1 hour at RT in humidified chamber, washed, and mounted with 4′,6-diamidino-2-phenylindole, and observed using Olympus BX51TRF microscope with an Olympus DP71 digital camera.

Relative density of Western blot analysis bands was analyzed by using ImageJ software downloaded from http://rsbweb.nih.gov/ij/. All values are expressed as mean ± SE. Statistical analysis was performed using Microsoft Excel software 2007. The significance between two test groups was determined using the Student t test. A P value of <0.05 was considered to be significant.

Topical treatment with Erb-041 substantially diminished UVB-induced skin tumor development in SKH-1 hairless mice as compared with vehicle-treated and UVB (alone)-irradiated mice. At the time of termination of experiment at week 30, the percentage of mice bearing tumors, tumors per mouse and tumor volume per mouse were significantly reduced in Erb-041–treated mice. The tumor incidence was 75% in the Erb-041+UVB group, whereas it was 100% in UVB-irradiated (alone) mice (Fig. 1A). The number of tumors per mouse was reduced to 3.3 ± 0.62 per mouse from 8.95 ± 0.94 per mouse in the UVB (alone) group, which represents >60% inhibition (Fig. 1B). Similarly, a 50% reduction (P < 0.001) in the number of tumors/tumor–bearing mouse was observed (Supplementary Fig. S1A). About 84% reduction in tumor volume (P < 0.005) was noted in the Erb-041–treated group (Fig. 1C). Erb-041 treatment increased latency period of tumor induction from 17 to 21 weeks. Overall, the number of SCCs per mouse was also reduced by 86% (P < 0.001; Supplementary Fig. S1B). To analyze tumor burden in these animals, we divided each group with respect to the number of animals bearing 0 to 5, 6 to 10, 11 to 15, or 16 to 20 tumors per mouse. Of note, 15% of UVB-irradiated mice were bearing 0 to 5 tumors per mouse, 45% 6 to 10 tumors per mouse, 30% 10 to 15 tumors per mouse, and 10% 16 to 20 tumors per mouse. However, in the Erb-041 treatment group, 70% of mice were bearing 0 to 5 tumors per mouse, whereas 30% had 6 to 10 tumors per mouse (Fig. 1D and E). Histologically, SCCs at week 30 were characterized as a mix of poorly-differentiated SCCs (pSCC), moderately-differentiated SCCs (mSCC), and well-differentiated SCCs (wSCC). We also observed a few invasive keratoacanthomas. In the UVB (alone) group, SCC spectrum comprised of mice with 19% pSCC, 17% mSCC, and 14% (wSCC) of the total tumors, whereas in the Erb-041 treatment group, only 1% pSCC, 6% mSCC, and 11% wSCC were observed (Fig. 1F). UVB-irradiated pSCCs were distinguished by the absence of keratin pearls, aggressive spindle cells with hyperchromatic pleomorphic nuclei, and invasion of dermis. However, wSCCs were characterized by the frequent presence of well-defined keratin pearls (Fig. 1G).

We investigated the effects of Erb-041 treatment on the expression of proliferative biomarkers such as proliferating cell nuclear antigen (PCNA), cyclin D1, and Ki67 in UVB-induced skin tumors. As assessed by IHC as well as Western blot analysis, Erb-041 treatment significantly (P < 0.05) reduced the expression of these proteins (Fig. 2A and Supplementary Fig. S1C). Angiogenesis biomarkers such as CD31/VEGF were assessed in UVB (alone)-irradiated and UVB+Erb-041–treated tumors. As shown in Fig. 2B, the immunostaining for CD31/VEGF was considerably reduced by Erb-041 treatment. The apoptosis in cutaneous tumor tissues was assessed by the presence of TUNEL-positive cells. The number of TUNEL-positive cells was highly increased in the Erb-041 treatment group as compared with the UVB (alone) group (Fig. 2C). Because induction of apoptosis is often correlated with the increased expression of proapoptotic Bax and decreased expression of antiapoptotic Bcl-2, or an increased Bax/Bcl-2 ratio (31), we also assessed these parameters in this study. Erb-041 treatment altered the expression of Bax and Bcl-2 in these tumor lesions (Supplementary Fig. S1D) in such a way that Bax/Bcl-2 ratio was significantly (P < 0.005) increased in tumors (Fig. 2C).

Earlier studies suggested that ER-β is a potent tumor suppressor and plays a crucial role in various cancers (22, 32, 33). Its expression is lost during the pathogenesis of various epithelial neoplasms (33). We, therefore, first assessed its expression in human cutaneous SCCs and tumor cells derived from SCCs. As shown in Fig. 3A, the expression of ER-β in histologically normal human skin was confined to the basal layer of the epidermis. Loss of expression in ER-β was noted in murine SCCs. Interestingly, Erb-041 treatment restored or even enhanced the expression of ER-β not only at protein level but also at transcriptional level in UVB-induced murine SCCs and human SCC cells in culture (Fig. 3B and C). Moreover, its expression was also apparent in the hyperplastic skin adjacent to papilloma and/or SCCs. However, a significant loss of its expression can be seen in human SCCs as well as SCCs-derived A431 and SCC13 cells as compared with immortalized HaCaT keratinocytes (Fig. 3D). Consistent with our in vivo results, Erb-041 treatment induced expression of ER-β in these human cells (Fig. 3E), which was confirmed with immunoblot analysis. Reduced expression of p-c-Jun and SP-1 was also associated with increase in ER-β expression (Fig. 3E).

We examined the effects of Erb-041 on UVB-induced inflammation and inflammation-regulating MAPK signaling pathways. UVB-induced inflammatory responses in murine skin are characterized by the development of edema and hyperplasia, enhanced leukocyte infiltration in the dermis, leukocytes-secreted inflammatory cytokines, and increased level of COX-2 and prostaglandins (3, 34). Consistently, as shown in Fig. 4A, the chronic exposure of murine skin to UVB-induced epidermal hyperplasia and dermal leukocytes infiltration, which was significantly reduced by Erb-041 treatment. Myeloperoxidase activity, a marker of neutrophil infiltration, was also decreased significantly (P < 0.05; Fig. 4B). Tumor microenvironment–associated inflammatory responses, which are known to accelerate tumorigenesis (35, 36), were found to be attenuated by Erb-041. Thus, a decrease in proinflammatory cytokines [interleukin (IL)-1β, IL-6, and IL-10] in tumor-associated skin was noted in Erb-041–treated mice (Fig. 4C). CD11b⁺/GR1⁺ myeloid cell population and macrophages in the dermis of UVB-irradiated skin as well as in tumor–stroma contribute to proinflammatory skin tumor progression (36, 37). As shown in Fig. 4D, the numbers of CD11b⁺/GR1⁺ myeloid cells and F4/80⁺ macrophages were significantly decreased by Erb-041 treatment. This was also accompanied by a reduction in the phosphorylation-dependent activation of ERK1/2 and p38 MAPKs (Fig. 4E and Supplementary Fig. S2A). Earlier Kim and colleagues reported that chronic UVB irradiation of the skin induces cytokine production, and activates the MAPK signaling pathway (35), which was confirmed this study. UVB-induced inflammation is also known to be associated with NF-κB activation (38, 39). NF-κB exists as a heterotrimeric complex in cytoplasm, which consists of p65, p52/p50 and inhibitory kappa B (IκB) proteins. Phosphorylation of IκB via inhibitor of I-κB kinases (IKKs) leads to release of transcriptionally active p65-p52/p65-p50 complexes and enable them to translocate to the nucleus (38, 39). Transcription activation of NF-κB is also evident by the enhanced expression of its target genes including proinflammatory COX-2 and iNOS (Fig. 4E and F). The Erb-041 treatment suppressed phosphorylation of IκBα resulting in the accumulation of IκBα as a heterotrimeric complex in the cytoplasm. Concomitantly, by inhibiting the activation of NF-κB, Erb-041 also reduced the expression of UVB-induced iNOS and COX-2 in these neoplastic lesions (Fig. 4E and F and Supplementary Fig. S2A). Similarly, nuclear NF-κBp65 and phosphorylated NF-κBp65 were reduced drastically in Erb-041–treated tumors as compared with the UVB (alone) tumors (Fig. 4E and F). These data provide a basis for the anti-inflammatory action of Erb-041 in the skin.

Epithelial–mesenchymal transition (EMT) is a process by which polarized epithelial cells transform to a mesenchymal fibroblast-like cell phenotype via multiple molecular cascades that result into apoptosis-resistance, enhanced migration, and invasiveness. EMT also increases components of extracellular matrix (40, 41). In malignant neoplasm, repression of E-cadherin by transcription factors such as Snail and Twist, ultimately leads to upregulation of mesenchymal marker proteins such as vimentin, fibronectin, and N-cadherin (41). EMT is known to be regulated by multiple mechanisms, including those dependent on phosphoinositide 3-kinase (PI3k)/Akt signaling pathways (7, 41). Therefore, we investigated whether Erb-041 interferes with the EMT progression in UVB-induced tumors. Immunoblot analysis and immunofluorescence analysis confirmed that Erb-041 increased the expression of epithelial biomarker E-cadherin and reduced the expression of mesenchymal markers N-cadherin, Snail, Slug, and Twist (Fig. 5A–C and Supplementary Fig. S2B). This is consistent with the observations that Erb-041 reduces the incidence of poorly-differentiated invasive SCCs in this study. In addition, in a wound-healing in vitro assay, we also found that Erb-041 treatment reduced migration potential of SCC cells (Supplementary Fig. S2C). Erb-041 inhibited about 55% and 71% cell migration when assessed for A431 and SCC13 cells, respectively (Fig. 5D). We also determined the effects of Erb-041 on the phosphorylation-dependent activation of PI3k and AKT in UVB-induced tumor (Fig. 5E and Supplementary Fig. S3A). These proteins are associated with the cell survival signaling pathway (41). UVB-induced pathogenesis of cutaneous neoplasm is known to be associated with the activation of this pathway (7, 41). Interestingly, Erb-041 treatment reduced phosoho-PI3k–AKT axis in UVB-induced tumor tissues.

Epithelial cell adhesion complex involves binding of E-cadherin/β-catenin/α-catenin complex to F-actin at transmembrane region, and plays a key role in EMT process during tumorigenesis (41, 42). Numerous studies reported that the release of β-catenin in cytoplasm, and then its migration to the nucleus are associated with loss of E-cadherin (41, 43). The β-catenin–dependent WNT signaling pathway is known to play essential roles in the regulation of cell polarity, proliferation, fate, survival, differentiation, and migration (43). In the presence of WNT ligands, the destruction complex containing proteins adenomatous polyposis coli, glycogen synthase kinase 3β (GSK3β), casein kinase 1 (CK1), β-catenin, and Axin gets dissociated. As a consequence, β-catenin releases, which leads to activation of transcription factors TCF/LEF (T-cell factor/lymphoid enhancer factor) and dependent target genes (43). In this study, we observed that augmented expression of WNT3a, WNT7b, FZD1, and β-catenin in UVB-induced skin tumors were reduced following Erb-041 treatment (Fig. 5F and Supplementary Fig. S3B). In addition, in immunofluorescence staining, we noted nuclear localization of β-catenin in UVB (alone)-induced tumor, whereas it was considerably reduced in Erb-041–treated UVB-induced tumors (Fig. 5G).

In an effort to unravel the underlying mechanism of this ER-β agonist, we treated human epidermal immortalized (HaCaT) and A431 and SCC13 cells with various concentration of Erb-041 in vitro. As shown in Supplementary Fig. S4A and S4B, Erb-041 treatment induced expression of cytokeratin 10, a differentiation marker. We next analyzed its effects on cell-cycle progression in these cells. Erb-041 treatment induced G₁ phase cell-cycle arrest in A431 cells, which was associated with the reduction in the expression of G₁ cyclins (D1, D2, and D3) and CDK4. A slight but insignificant reduction in the expression of cyclin B1/E, CDC-2, and CDK2 was also noted (Fig. 6A and B and Supplementary Fig. S4C). In a colony formation assay, consistent with its effects on cell-cycle progression, Erb-041 dramatically reduced the number and size of A431 and SCC13 colonies (Fig. 6C).

Similar to our observations in murine skin, a marked reduction in the expression of inflammation regulatory proteins such as p-NF-κBp65, iNOS, and COX-2 was observed in A431 cells (Fig. 6D and Supplementary Fig. S4D). Erb-041 treatment diminished phosphorylated-PI3K and -Akt, which was associated with the enhancement in E-cadherin expression and reduction in migration of these cells in an in vitro scratch assay (Fig. 6E).

We also observed that Erb-041 dampened the WNT signaling pathway in the murine skin. The WNT signaling pathway is known to be associated with the pathogenesis of skin cancer (37). It is known to be involved in the development of invasive SCCs by modulating EMT at least partially (24, 43). We, therefore, tested whether Erb-041 manifests similar effects in human carcinoma cells. Erb-041 treatment reduced expression of WNT7b, β-catenin, and p-GSK3β (Fig. 6F and G). These changes were accompanied by the diminished nuclear localization of β-catenin (Fig. 6F). Consistently, we also observed a significant reduction in the expression of its downstream target proteins c-Myc and cyclin D1 (Fig. 6H). The activation of the WNT/β-catenin pathway leads to inhibition of axin-mediated β-catenin phosphorylation, leading to the accumulation of nuclear β-catenin and transcription of the WNT pathway–responsive genes (43). To confirm that the reduction in the WNT signaling pathway in epidermal carcinoma cells may decrease EMT, we used a small molecule pharmacologic inhibitor of the WNT signaling pathway, XAV939. XAV939 stabilizes axin through tankyrase inhibition and modulates Wnt target effectors (44). As shown in Fig. 6H, XAV939 treatment of HaCaT, A431, and SCC13 cells dramatically suppressed the expression of the Wnt signaling pathway proteins, WNT3a, WNT7b, FZD1, β-catenin, and GSK3β along with cyclin D1. Importantly, XAV939 treatment did not induce ER-β expression, although, it reduced the expression of ER-β's cofactors SP-1 and p-c-Jun (Fig. 6H, bottom). Earlier, SP-1 and p-c-Jun were shown to be regulated by the WNT signaling pathway (44). XAV939 treatment also ameliorated the expression of EMT-regulating proteins. The expression of E-cadherin was increased, whereas the expression of N-cadherin, Twist, and Slug was decreased (Fig. 6I, top). Interestingly, the expression of inflammatory signaling molecules p-IκBα, p-NF-κBp65, iNOS, and COX-2 were also reduced in all the cell lines tested in this study (Fig. 6I, lower panel). Many of these effects were similar to those manifested by Erb-041 in these cells (26).

Estrogen signaling particularly that regulated by ER-β is considered important in the pathogenesis of various cancers. ER-β expression is often lost during the progression of epithelial cancers (22, 23). This signaling is not only mediated through the estrogen response elements but also impacts cellular growth by modulating various transcription factors AP-1, SP-1, NF-κB, etc. (16, 17). Consistently, we also observed a decreased in p-c-Jun and SP-1 by Erb-041 in UVB-induced cutaneous tumors. Although the loss of expression of ER-β receptor may occur through multiple mechanisms, promoter methylation of ER-β is considered as an important downregulator of its expression (25). The importance of upregulation of ER-β was shown by the studies in which valproic acid-mediated demethylation of ER-β, which restored its expression in cancer cells, led to antiproliferative effects (45). Similarly, small molecule antagonists of ER-β, BAG1, and BAG2 resulted in tumor growth arrest and shrinkage (15). However, our results provide additional novel effects of ER-β agonist, Erb-041. Erb-041 not only restored or augmented the expression of ER-β in murine SCCs and in human carcinoma cells but reduced in proliferation and induced differentiation and apoptosis in these models of skin carcinogenesis. Significantly, these effects together led to a profound reduction in the growth of SCCs and the residual SCCs were found to be mostly highly differentiated carcinoma types.

A link between tumor growth and inflammation is now well established (37, 38). Inflammatory immune cells are recruited to cancer sites and lead to development of a conducive neoplastic environment, which is responsible for facilitating tumor progression (37, 39). These inflammatory hematopoietic cells by virtue of their capabilities to supply soluble growth factor, matrix remodeling enzymes, and other bioactive molecules influence cancer cell proliferation, angiogenesis, invasion, and metastasis (36, 37, 39). Interestingly, we found that Erb-041 not only reduced cutaneous hyperplasia but also reduced cytokine production, including those of IL-1β, IL-6, and IL-10. These changes were associated with a significant decrease in the number of GR1/CD11b–positive myeloid cells, F4/80 macrophages, and neutrophils as ascertained by significant decrease in myeloperoxidase activity. Thus, these results provide evidence that Erb-041 acts by modulating proinflammatory tumor microenvironment.

Transcription factor NF-κB is a key regulator of many of inflammatory responses. This transcription factor upregulates the expression of multiple inflammation-linked genes including COX-2, IL-1β, IL-6, p38, iNOS etc. The observations in this study that these proteins are also downregulated by Erb-041 treatment in the skin and in residual tumors provide evidence that Erb-041 may act by modulating the NF-κB–dependent signaling pathway. A significant decrease in the nuclear expression of p65, together with a decrease in its target genes, suggests that ER-β and NF-κB function in coordination to dampen inflammatory signaling and SCC growth in this mouse model. However, it is also known that immunosurveillance is impaired during the progression of tumorigenesis (36, 37) and ER-β has recently been shown to modulate tumor immunosurveillance (19, 20). Therefore, participation of this additional mechanism in the reduction of cutaneous tumorigenesis by Erb-041 cannot be ruled out at this stage. Inflammation is known to augment invasive tumor growth by promoting EMT (46, 47). Earlier, we showed that anti-inflammatory agents not only block UVB-induced inflammation but also reduced EMT progression (7, 41). Parallel to these studies, the observations that Erb-041 treatment reduced inflammation and EMT associated with the enhanced expression of E-cadherin and reduced expression of mesenchymal proteins N-cadherin, Snail, Slug, Twist, and matrix metalloproteinases (MMP) suggest a role of UVB-induced cutaneous inflammation in regulatory EMT in skin SCCs. The reduction in EMT was associated with the diminution of PI3K/AKT signaling provide a molecular basis for the action of Erb-041 for blocking EMT in the malignant cutaneous keratinocytes. Role of PI3K/AKT is already described in EMT (7, 41). Thus, ER-β receptor not only reduced tumorigenesis and inflammation but also diminished progression to an aggressive and invasive tumor phenotype.

The mechanism by which these multitarget effects can occur is not currently well understood. However, recent studies described a crucial role of WNT signaling in connecting inflammatory and tumor promoting responses (47, 48). Autocrine WNT signaling plays a crucial role in the growth and survival of various cancer cells (43, 49). In this study, we found that WNT3a as well as WNT7b are upregulated during the UVB-induced carcinogenesis in experimental animals and in humans. This leads to TCF/LEF–dependent transcriptional activation contributing to the promotion of tumor growth (43). Erb-041 treatment decreased both WNT3a and WNT7b expression in immortalized human skin keratinocytes and SCC cells. This decrease in Wnt ligands was also associated with a decrease in overall nuclear β-catenin and its target genes such as cyclin D1, c-Myc, and SP-1. Earlier, WNT signaling has been shown to regulate the EMT by balancing the expression of E-cadherin and mesenchymal proteins (41, 43). For example, in various epithelial tumors, activation of WNT signaling drives a transcriptional program reminiscent of EMT, which promote cell migration and invasiveness (43). To confirm the role of WNT signaling in regulating ER-β–dependent diminution in EMT and invasive tumor phenotype, we investigated the effects of small molecule XAV939. XAV939 is known to inhibit Wnt signaling (44) and blocks accumulation of β-catenin in colorectal cancer. The mechanism by which this agent acts involves stabilization of axin by inhibiting the poly-ADP-ribosylating enzymes tankyrase 1 and tankyrase 2 (44). In our studies, XAV939 manifested similar results as have been observed by the treatment with Erb-041, suggesting a role of WNT signaling in ER-β–mediated attenuation of EMT in cutaneous SCCs. In summary, our results show that Erb041 is a potent chemopreventive agent that blocks tumorigenesis by inhibiting proliferation and inducing differentiation and apoptosis. The mechanism by which ER-β agonist Erb-041 acts involves diminution of the WNT signaling pathway.

No potential conflicts of interest were disclosed.

Conception and design: S.C. Chaudhary, T. Singh, S.S. Talwelkar, R.K. Srivastava, C.A. Elmets, F. Afaq, L. Kopelovich, M. Athar

Development of methodology: S.C. Chaudhary, T. Singh, S.S. Talwelkar, A. Arumugam, L. Kopelovich, M. Athar

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S.C. Chaudhary, T. Singh, R.K. Srivastava, A. Arumugam, Z. Weng

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S.C. Chaudhary, T. Singh, R.K. Srivastava, A. Arumugam, Z. Weng, F. Afaq, L. Kopelovich, M. Athar

Writing, review, and/or revision of the manuscript: S.C. Chaudhary, T. Singh, R.K. Srivastava, A. Arumugam, Z. Weng, C.A. Elmets, F. Afaq, L. Kopelovich, M. Athar

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S.C. Chaudhary, S.S. Talwelkar, A. Arumugam, Z. Weng

Study supervision: C.A. Elmets, M. Athar

This study was supported by grants NIH/NCI N01-CN-43300 274 and R01 CA138998 from National Cancer Institute (to M. Athar).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.