Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Nov 5.
Published before final editing as: J Invest Dermatol. 2026 Oct 9:S0022-202X(26)03771-1. doi: 10.1016/j.jid.2026.09.856

Functional Roles of Wnt/β-catenin Pathway Components in Pre-Basal Cell Carcinoma-like Cell Proliferation and Survival

Xuming Zhu 1,*, Andrzej A Dlugosz 2,3, John T Seykora 4, Sarah E Millar 1,5,6,*
PMCID: PMC13734293  NIHMSID: NIHMS2216697  PMID: 42854951

Abstract

Wnt/β-catenin signaling supports basal cell carcinoma (BCC)-like tumor initiation in mice. However, nuclear β-catenin is frequently detected in well-developed BCCs and correlates with increased tumor cell proliferation, suggesting that Wnt/β-catenin signaling may also promote BCC development. Our analysis of published single-cell transcriptomic datasets from human BCCs revealed that multiple Wnt pathway components, including Wntless (WLS) and the Wnt receptor Frizzled2 (FZD2), are enriched in specific BCC cell populations with active Hedgehog (Hh) signaling. Using an inducible mouse model with forced epithelial expression of the Hh activator GLI2ΔN, we found that Wnt-responsive Axin2+ epidermal basal cells contribute to BCC-like tumor formation. Deletion of epithelial β-catenin or WLS several days after GLI2ΔN expression significantly reduced tumor cell proliferation without affecting BCC marker expression, Hh pathway activity, or apoptosis. These findings indicate that Wnt/β-catenin signaling acts downstream of Hh signaling to promote pre-BCC-like cell proliferation. Furthermore, Wnt receptor Fzd2 depletion reduced tumor cell proliferation and induced apoptosis, suggesting that FZD2 promotes proliferation and survival through β-catenin-dependent and -independent mechanisms. These findings identify Wnt pathway components as key regulators of pre-BCC-like cell proliferation and survival.

Keywords: Basal cell carcinoma, Hedgehog, Wnt/β-catenin, Wntless, Frizzled2

Introduction

Basal cell carcinoma (BCC) is the most common form of skin cancer, characterized by aberrant activation of the Hedgehog (Hh) signaling pathway (Bakshi et al., 2017, Roky et al., 2025). Under physiological conditions, Hh signaling is initiated by the binding of Hh ligands to their receptor Patched1 (PTCH1), which relieves inhibition of the signal transducer Smoothened (SMO) and activates GLI transcription factors to regulate the expression of downstream target genes (Zhang and Beachy, 2023). In contrast, oncogenic Hh signaling is ligand-independent and is often driven by PTCH1 loss-of-function or SMO gain-of-function mutations (Jing et al., 2023). SMO inhibitors such as Vismodegib have been developed and approved for the treatment of diseases driven by oncogenic Hh activity, including advanced or metastatic BCC (Axelson et al., 2013, Nguyen and Cho, 2022). Although the majority of BCC patients benefit from such drugs, a significant subset exhibits only partial responses (Metcalfe and de Sauvage, 2011). These are marked by initial tumor shrinkage followed by regrowth upon drug withdrawal (Dong et al., 2018), suggesting that additional factors or signaling pathways may contribute to BCC development and persistence.

The Wnt/β-catenin pathway plays essential roles in embryonic development, postnatal tissue homeostasis, and disease (Liu et al., 2022). This pathway is activated by the binding of Wnt ligands to Frizzled (FZD) receptors, leading to stabilization of the signaling mediator β-catenin in the cytoplasm. Stabilized β-catenin translocates into the nucleus to regulate the expression of target genes such as LEF1, AXIN2, and LGR5 (Carmon et al., 2012, Jho et al., 2002, Xue et al., 2025). Involvement of Wnt/β-catenin signaling in BCC is suggested by the presence of nuclear β-catenin, which correlates with increased tumor cell proliferation (El-Bahrawy et al., 2003, Saldanha et al., 2004). Additionally, LEF1, which is minimally expressed in normal human epidermis, is highly upregulated at the invasive front of human BCC lesions (Kriegl et al., 2009). These findings imply that the Wnt/β-catenin pathway may promote tumor growth, progression, and metastasis in established BCCs.

Supporting this notion, a slow-cycling population of tumor cells expressing LGR5 and exhibiting active Wnt signaling has been identified within BCCs. Ablation of LGR5+ cells or inhibition of Wnt signaling in combination with Vismodegib treatment leads to complete tumor eradication (Sánchez-Danés et al., 2018), suggesting that Wnt signaling is essential for maintaining these quiescent Vismodegib-resistant BCC cells. However, whether Wnt signaling also directly regulates the proliferation and survival of BCC cells remains unclear.

Previous mouse genetic studies using ectopic overexpression of an activated SMO mutant (M2SMO) have shown that epidermal inhibition of the Wnt/β-catenin pathway, either by β-catenin deletion or DKK1 overexpression, completely prevents BCC-like tumor initiation (Yang et al., 2008, Youssef et al., 2012). At the cellular level, Wnt/β-catenin signaling reprograms adult epidermal stem cells into embryonic hair follicle progenitors (Youssef et al., 2012), forming hair germ-like structures that resemble the budding morphology of human superficial BCCs (Yang et al., 2008). However, because these studies involved simultaneous Hh activation and Wnt inhibition, they do not address the role of Wnt/β-catenin signaling after Hh activation.

In this study, we first examined the expression of Wnt/β-catenin pathway components in the various cell populations within human BCC and mouse BCC-like tumors by analyzing published single-cell RNA sequencing (scRNA-seq) datasets. The results revealed that multiple Wnt/β-catenin pathway components are upregulated in cells with active Hh signaling, a finding further validated by RNAscope assays. Using lineage tracing of Wnt-responsive, Axin2-expressing epidermal basal cells, we demonstrated that these cells contribute to BCC-like tumor formation following forced epithelial expression of constitutively active GLI2ΔN in the hairless plantar epidermis. Disruption of the Wnt/β-catenin pathway by depleting β-catenin or Wntless (WLS) several days after GLI2ΔN induction markedly impaired the proliferation of GLI2ΔN-expressing cells without affecting Hh pathway activity or inducing apoptosis. Furthermore, we identified FZD2 as a functional Wnt receptor that mediates the pro-proliferative effects of Wnt/β-catenin signaling and also permits cell survival during pre-BCC-like tumor development.

Results

Wnt/β-catenin pathway components are overexpressed in tumor cells with active Hh signaling in human BCC

To investigate the expression of Wnt/β-catenin pathway components in BCC tissues at a single-cell level, we reanalyzed the published scRNA-seq dataset GSE141526, which profiled mixed BCC subtypes and peri-BCC non-cancerous skin samples (Guerrero-Juarez et al., 2022). UMAP analysis identified 10 cell clusters, including four epidermal cell populations (KRT5/KRT14+), one hair follicle cell population (KRT23/KRT7+), two dermal cell populations (TWIST2/COL1A1+), one melanocyte population (TYRP1/DCT+), and two immune cell populations (CXCR4/CD3D+ and CD74/CD93+) (Supplementary Figure S1a, b). All cell types were present in both BCC and peri-BCC skin samples, although epidermal and dermal cell proportions differed between conditions (Supplementary Figure S1c).

We then subclustered the epidermal cell populations and assessed the expression of known BCC marker genes (GLI1, PTCH1, HHIP, GLI2, SOX9, KRT17) and several representative Wnt pathway components (WLS, LEF1, AXIN2, FZD1, FZD2, WNT7B, WNT10A), which were defined as BCC_score and WNT_score, respectively. Notably, epidermal cells with high WNT_scores were rare but consistently exhibited high BCC_scores (Figure 1a). Among these, cluster 5 was enriched for both BCC marker genes and Wnt pathway components (Figure 1b). In BCC samples, these cells expressed higher levels of BCC markers and Wnt components compared to peri-BCC unremarkable skin (Figure 1c), indicating activation of the Wnt pathway in epidermal cells with robust Hh activity.

Figure 1. scRNA-seq analysis reveals enrichment of Wnt/β-catenin pathway components in human BCC cells with active Hedgehog signaling.

Figure 1.

Open in a new tab

(a) Feature plots of epidermal cells from GSE141526 showing a subset of cells (yellow dots) with high WNT_scores (red) and high BCC_scores (green). (b) Dot plot depicting the expression of representative BCC marker genes and Wnt pathway components in epidermal cell populations within the GSE141526 dataset. (c) Violin plots illustrating the expression of the indicated BCC markers and Wnt pathway genes.

To validate these findings, we analyzed the published scRNA-seq dataset GSE181907 that profiled human infiltrative BCC samples (Yerly et al., 2022). UMAP analysis identified 15 cell populations, including four tumor cell clusters characterized by SOX9 and K17 expression (Supplementary Figure S2a, b). Consistent with the analyses described above, tumor cells with high WNT_scores were limited in number but showed high BCC_scores (Supplementary Figure S2c). Cluster 0 tumor cells were particularly enriched for BCC markers and Wnt components, including FZD2 and WLS (Supplementary Figure S2d).

To determine whether similar expression patterns also characterize mouse BCC-like tumors, we examined the published scRNA-seq dataset GSE186184 from K14-CreER Ptch1+/− p53fl/fl mice (Haensel et al., 2022). UMAP analysis identified five tumor clusters expressing BCC markers Ptch1, K17, and Sox9 (Supplementary Figure S3a, b). As in human BCC, cells with high WNT_scores were rare but strongly overlapped with high BCC_scores (Supplementary Figure S3c), and cluster 2 was enriched for both BCC markers and Wnt components (Supplementary Figure S3d).

To verify these bioinformatic findings, we performed RNAscope assays on three human nodular and two superficial BCC samples containing both tumor and peri-tumoral epidermis. WLS and FZD2 expression were upregulated in BCC tumors relative to non-cancerous epidermis, coinciding with elevated expression of the Hh target GLI1 (Figure 2a–y; Figure Supplementary S4a–r). Collectively, these results indicate that Wnt components are overexpressed in BCC cells with strong Hh activity and suggest potential roles for WLS and FZD2 in BCC development.

Figure 2. WLS and FZD2 are upregulated in human BCC tissues.

Figure 2.

Open in a new tab

(a–c) RNAscope analysis of BCC#1 samples showing WLS and GLI1 expression in tumor and peri-tumoral epidermis. Both genes are overexpressed in BCC lesions compared with adjacent unremarkable epidermis. (d–f) Enlarged views of the yellow boxed regions in (a–c). (g–i) RNAscope data showing increased FZD2 and GLI1 expression in BCC#1 samples compared with adjacent epidermis. (j–l) Enlarged views of the yellow boxed regions in (g–i). (m–o) RNAscope images demonstrating elevated WLS and GLI1 levels in BCC#2 samples compared with the surrounding epidermis. (p–r) Enlarged views of the yellow boxed regions in (m–o). (s–u) RNAscope images illustrating elevated FZD2 and GLI1 expression in BCC#2 samples compared with adjacent unremarkable epidermis. (v–x) Enlarged views of the yellow boxed regions in (s–u). White and orange dashed lines outline the epidermis–dermis boundary and BCC, respectively. Scale bar: 100 μm. (y) Quantification of RNAscope signals in the BCC and adjacent epidermis samples. N=5; bilateral paired Student’s t-tests were applied for P-value calculation.

Forced expression of GLI2ΔN in non-hairy plantar epidermis induces BCC-like tumors

To investigate the role of Wnt/β-catenin signaling in BCC after tumor initiation, we established an inducible adult mouse model that mimics superficial BCC by overexpressing activated human GLI2 (GLI2ΔN) under the control of the epithelial-specific promoter K5 driving the reverse tet transactivator rtTA. Because high GLI2ΔN levels caused severe, lethal phenotypes due to widespread epithelial tumors, we optimized a low-dose doxycycline regimen that induced BCC-like tumors without early lethality. After three weeks of low-dose doxycycline induction (Supplementary Figure S5a), we observed pre-BCC-like lesions in non-hairy plantar epidermis (Supplementary Figure S5b, c). IF and RNAscope confirmed GLI2ΔN expression (Supplementary Figure S5d, e) and increased Gli1 levels (Supplementary Figure S5f, g). These lesions were highly proliferative (Ki-67+) and expressed BCC markers SOX9 and K17 (Supplementary Figure S5j–m). Longer induction times produced well-formed BCC-like tumors in plantar skin (Supplementary Figure S5n–p). Sustained GLI2ΔN expression maintained elevated Gli1 levels and proliferation, with tumors continuing to express SOX9 and K17 (Supplementary Figure S5q–z). These results confirm that controlled GLI2ΔN expression recapitulates superficial BCC-like tumor development in adult hairless plantar epidermis.

Wnt/β-catenin activity increases during BCC-like tumor development

To assess the activity of Wnt/β-catenin signaling in GLI2ΔN-induced tumor tissues, we evaluated the presence of nuclear β-catenin and expression of Wnt target genes Lef1 and Axin2. IF assay detected nuclear β-catenin signals in pre-BCC-like tumors but not in control plantar epidermis following three weeks of doxycycline treatment (Supplementary Figure S6a–c). IF and RNAscope further showed that LEF1 was likewise restricted to pre-BCC-like lesions (Supplementary Figure S6d, e), and Axin2 expression was elevated relative to controls (Supplementary Figure S6f, g). Similar expression patterns of these Wnt pathway components were observed in fully developed tumors after six weeks of doxycycline administration (Supplementary Figure S6h–n). Consistent with findings from analyses of human BCCs, Fzd2 and Wls were upregulated in BCC-like tumors (Supplementary Figure S6o–r). Together, these data indicate persistent Wnt/β-catenin activation during BCC-like tumor development.

Axin2-expressing epidermal basal cells contribute to BCC-like tumor formation

To determine whether Wnt-responsive epidermal cells contribute to BCC development, we crossed Axin2-CreERT2 mice with the Rosa26mTmG reporter and traced Axin2-expressing (Axin2GFP+) epidermal basal cells in GLI2ΔN-positive tissues. Forty-eight hours after tamoxifen injection to induce Cre recombinase activity (Figure 3a), sporadic Axin2-expressing cells were detected, predominantly at the periphery of pre-BCC-like lesions (Figure 3b, c). These cells were also Ki-67+, indicating that they were proliferative and likely fueling rapid tumor growth (Figure 3b, c). Three weeks after tamoxifen administration (Figure 3d), Axin2GFP+ cells had expanded to form distinct patches within well-formed BCC-like tumors under prolonged GLI2ΔN induction (Figure 3e, f).

Figure 3. Axin2-expressing epidermal basal cells contribute to BCC-like tumor formation.

Figure 3.

Open in a new tab

(a, d, g, j) Schematics of experimental design. (b, c) IF images showing sporadic Axin2GFP+ cells (yellow arrowhead) at the periphery of pre-BCC-like lesions 48 h after tamoxifen injection in GLI2ΔN-expressing mice treated with doxycycline for three weeks. The GLI2ΔN-expressing cells are Ki-67+, indicating proliferation. (e, f) Three weeks post-tamoxifen, Axin2GFP+ cells have expanded to form clones within well-developed BCC-like tumors under continued GLI2ΔN induction; a subset are Ki-67+ (yellow arrowhead). (h, i) Following doxycycline withdrawal for three weeks, GLI2ΔN expression diminished, but Axin2GFP+ cells persisted and proliferated (yellow arrowhead) within hyperplastic epidermis derived from regressed lesions. (k, l) Upon reactivation of GLI2ΔN for three weeks, Axin2GFP+ cells proliferated (yellow arrowheads) and contributed to BCC-like tumor growth, demonstrating retained proliferative and tumor-repopulating potential despite prior deactivation of GLI2ΔN expression. Three mice were analyzed per condition. Scale bar: 25 μm.

To examine the behavior of Axin2GFP+ cells after withdrawal of Hh activation, we discontinued doxycycline treatment and continued lineage tracing for an additional three weeks (Figure 3g). IF staining revealed that Axin2GFP+ cells continued to proliferate within the hyperplastic epidermis derived from regressed pre-BCC-like lesions (Figure 3h, i). Upon restoring GLI2ΔN expression for three weeks (Figure 3j), Axin2GFP+ cells again proliferated and contributed to BCC-like tumor development (Figure 3k, i). Collectively, these findings demonstrate that Axin2+ epidermal basal cells can contribute to BCC-like tumor formation and retain their long-term proliferative and tumor-repopulating potential even in the absence of Hh signaling.

β-catenin regulates the proliferation of GLI2ΔN-expressing tumor cells

To investigate the role of the Wnt/β-catenin pathway in epidermal keratinocytes after Hh activation, we introduced a β-catenin floxed allele (Ctnnb1fl) into our inducible GLI2ΔN-expressing mouse model, which also carries a doxycycline-inducible tetO-Cre transgene. Upon doxycycline treatment, Cre recombinase is induced, leading to excision of the floxed allele and subsequent depletion of β-catenin in GLI2ΔN-expressing epidermal cells. Four days after doxycycline administration (Supplementary Figure S7a), the plantar epidermis displayed early signs of invagination, GLI2ΔN-expressing epidermal cells, increased proliferation, and elevated expression of BCC markers SOX9 and K17, as confirmed by H&E and IF analyses (Supplementary Figure S7b–j). At this stage, β-catenin protein levels were comparable between control GLI2ΔN-expressing mice and their Ctnnb1fl/fl counterparts (Supplementary Figure S7k, l), indicating that GLI2ΔN expression and keratinocyte overgrowth occur prior to β-catenin deletion in the plantar epidermis. Therefore, this model can be used to probe the effects of β-catenin deletion after Hh activation.

By ten days after doxycycline treatment (Figure 4a), IF staining confirmed efficient epidermal β-catenin deletion (Figure 4d, e), and invagination of pre-BCC-like tumors was impaired (Figure 4b, c). Reduced Wnt/β-catenin signaling activity was evident from the downregulation of LEF1 and Axin2 (Figure 4f–h). In contrast, Hh pathway activity was unaffected, as Ptch1 expression remained similar between control and β-catenin-deficient samples (Figure 4h). Additionally, IF analyses revealed unchanged SOX9 and K17 expression (Figure 4i–l), indicating that β-catenin deletion did not impair BCC identity.

Figure 4. β-catenin regulates the proliferation of GLI2ΔN-expressing pre-BCC-like tumor cells.

Figure 4.

Open in a new tab

(a) Experimental scheme. (b, c) H&E images showing impaired invagination of pre-BCC-like tumors after β-catenin deletion in GLI2ΔN-expressing epidermis. (d–g) IF analysis confirming efficient β-catenin deletion (d, e) and reduced LEF1 expression (f, g) in β-catenin–depleted lesions. (h) qPCR showing significantly reduced Axin2 but unchanged Ptch1 expression. N=3. (i–l) IF images demonstrating unchanged SOX9 and K17 expression upon β-catenin deletion. (m, n, q) Ki-67 staining indicating reduced proliferation after β-catenin deletion. N=5. (o, p) TUNEL assay showing comparable apoptosis between control and β-catenin-deficient lesions, with no apoptosis detected in basal layers. White dashed lines mark the epidermis–dermis boundary. Scale bar: 50 μm.

To determine whether the reduced invagination resulted from altered cell proliferation or survival, we quantified the proliferation of GLI2ΔN-expressing basal cells. The percentage of Ki-67+ cells was significantly reduced in β-catenin-deficient tissues (Figure 4m, n, q). Conversely, TUNEL assays did not show enhanced apoptosis in β-catenin-deficient GLI2ΔN-expressing epidermis compared with controls (Figure 4o, p). Collectively, these findings demonstrate that β-catenin specifically regulates the proliferation, but not the survival, of GLI2ΔN-expressing basal cells in the pre-BCC-like lesions.

Autocrine Wnt ligands support the proliferation of GLI2ΔN-expressing cells

To identify the source of Wnt ligands driving the aberrant growth of GLI2ΔN-expressing cells that form pre-BCC-like lesions, we deleted Wls, which encodes a protein essential for Wnt ligand secretion (Wolf and Boutros, 2023), in epithelial cells. Four days after doxycycline administration (Supplementary Figure S8a), RNAscope assay revealed that Wls mRNA expression was comparable between control GLI2ΔN-expressing mice and their Wlsfl/fl counterparts (Supplementary Figure S8b, c), indicating that GLI2ΔN expression and keratinocyte overgrowth also occur before Wls deletion in the plantar epidermis. Similar to β-catenin deletion, loss of Wls impaired the invagination of GLI2ΔN-expressing epidermal tissue after ten days of doxycycline treatment (Figure 5a–c, h). IF and RNAscope analyses revealed diminished nuclear β-catenin, reduced LEF1 and Axin2 levels in GLI2ΔN-expressing cells (Figure 5d–h), confirming disrupted Wnt/β-catenin signaling in Wls-deficient tissues. In contrast, Ptch1, SOX9, and K17 expression remained unchanged following Wls deletion (Figure 5h–l), suggesting that epithelial Wls and Wnt secretion are dispensable for maintenance of Hh signaling and BCC identity. Wls deletion markedly reduced the proliferation of GLI2ΔN-expressing cells but had minimal impact on apoptosis (Figure 5m–q). These results, mirroring the effects of β-catenin deletion, indicate that GLI2ΔN-expressing cells produce Wnt ligands in an autocrine manner to sustain their own proliferation during pre-BCC-like tumor development.

Figure 5. Epithelial Wnt ligands support the proliferation of GLI2ΔN-expressing cells.

Figure 5.

Open in a new tab

(a) Experimental scheme. (b, c) H&E images showing that epithelial Wls deletion disrupts invagination of pre-BCC-like lesions. (d–g) IF analysis revealing loss of nuclear β-catenin (d, e) and reduced LEF1 (f, g) expression in Wls-deficient tissues. (h) qPCR indicating significantly decreased Wls and Axin2 mRNA, but unchanged Ptch1 expression. N=3. (i–l) IF images showing sustained SOX9 (i, j) and K17 (k, l) expression in Wls-deficient tumors. (m, n, q) Ki-67 staining demonstrating reduced proliferation following Wls deletion. N=5. (o, p) TUNEL assays indicating similar apoptosis levels in control and Wls-deficient lesions, with no basal layer apoptosis in either group. White dashed lines mark the epidermis–dermis boundary. Scale bar: 50 μm.

Wnt receptor FZD2 is required for the proliferation and survival of GLI2ΔN-expressing cells

To determine which Wnt receptor mediates the Wnt/β-catenin signaling during BCC development, we focused on FZD2, whose expression is elevated in both human BCCs and mouse BCC-like tumors. We utilized mice carrying an unconventional floxed allele for Fzd2 (Fzd2fl) that permits Cre-mediated knockdown of Fzd2 mRNA (Zhu et al., 2023). Four days of doxycycline administration had minimal effects on Fzd2 mRNA expression in GLI2ΔN-expressing plantar epidermis (Supplementary Figure S8d, e). After ten days of doxycycline treatment of GLI2ΔN-expressing mice homozygous for Fzd2fl (Figure 6a), H&E staining showed reduced invagination of pre-BCC-like lesions in doxycycline-treated plantar epidermis overexpressing GLI2ΔN (Figure 6b, c). RNAscope and qPCR analyses confirmed efficient Fzd2 depletion (Figure 6d, e, j). Loss of Fzd2 led to loss of nuclear β-catenin and reduced LEF1 and Axin2 expression, whereas Ptch1, K17, and SOX9 expression remained unchanged (Figure 6f–n). IF analyses revealed a significant reduction in the proliferation of GLI2ΔN-expressing cells following Fzd2 depletion (Figure 6o, p, s). Unexpectedly, unlike β-catenin or Wls depletion, Fzd2 knockdown also induced ecto RNAscope and qPCR analyses confirmed pic basal apoptosis in the plantar epidermis (Figure 6q, r, t), suggesting that FZD2 may support the survival of GLI2ΔN-expressing cells through a β-catenin-independent mechanism. Collectively, these findings identify FZD2 as a critical Wnt receptor that may regulate both the proliferation and survival of tumor cells during pre-BCC-like lesion development.

Figure 6. FZD2 is required for the proliferation and survival of GLI2ΔN-expressing cells.

Figure 6.

Open in a new tab

(a) Experimental scheme. (b, c) H&E images showing reduced invagination of pre-BCC-like lesions following Fzd2 deletion. (d, e) RNAscope confirming efficient Fzd2 depletion in the epidermis. (f–i) IF analysis showing loss of nuclear β-catenin (f, g) and reduced LEF1 expression (h, i) in Fzd2-deficient tissues. (j) qPCR indicating significantly decreased Fzd2 and Axin2 mRNA but unaffected Ptch1 expression. N=3. (k–n) IF images showing maintained SOX9 (k, l) and K17 (m, n) expression after Fzd2 depletion. (o, p, s) Ki-67 staining demonstrating markedly reduced proliferation upon Fzd2 depletion. N=5. (q, r, t) TUNEL assay revealing ectopic apoptosis (yellow arrowheads) in the basal layer of Fzd2-deficient lesions, whereas no apoptotic cells were detected in control basal cells. White dashed lines mark the epidermis–dermis boundary. N=5. Scale bar: 50 μm.

Discussion

In this study, we show that Wnt/β-catenin signaling is activated in BCC cells with robust Hh pathway activity and that Wnt-responsive Axin2-expressing epidermal basal cells contribute to BCC-like tumor formation. Using inducible GLI2ΔN-expressing mouse models, we demonstrate that blocking Wnt signaling, by deleting either β-catenin or Wls, significantly reduces the proliferation of tumor cells without affecting Hh pathway activity or BCC identity. Furthermore, we identify FZD2 as a key Wnt receptor required not only for proliferation but also for the survival of GLI2ΔN-expressing cells (Figure 7).

Figure 7. Model for Wnt pathway involvement in pre-BCC proliferation and survival.

Figure 7.

Open in a new tab

Hyperactivation of the Hh signaling pathway elevates WLS and FZD2 levels, enhancing Wnt secretion by Axin2+ epidermal basal cells. This amplifies Wnt signaling, promoting proliferation and generating progeny of Axin2+ basal cells that persist long-term and contribute to pre-BCC development. FZD2 mediates β-catenin–dependent Wnt signaling for proliferation but supports tumor cell survival through an unknown mechanism.

Oncogenic Hh signaling is a central driver of BCC, making the Hh inhibitor Vismodegib an effective option for patients for whom surgery is not an option. However, Vismodegib monotherapy rarely eliminates all tumor cells, and residual, Hh-resistant tissue often relapses after treatment discontinuation. Published studies have provided insight into the origin of these persisting cells. One study showed that BCCs are organized into basal and suprabasal compartments and that most residual tumors arise from basal cells with weak Notch activity, enabling survival during Vismodegib therapy and relapse afterward (Eberl et al., 2018). Using mouse models, another group reported that Vismodegib drives BCC regression by altering tumor cell fate and promoting differentiation, but leaves behind slow-cycling, label-retaining cells (LRCs) expressing LGR5. Maintained by active Wnt signaling, these dormant LRCs regenerate tumors once treatment ends (Sánchez-Danés et al., 2018). Together, these findings indicate that additional pathways, particularly Wnt/β-catenin signaling, contribute to BCC persistence and should be considered as targets for combination therapies.

In line with the notion that Wnt signaling maintains LGR5+ LRC cells within BCC, our scRNA-seq analysis showed that only a small fraction of high BCC_score cells exhibit a high WNT_score. Notably, specific tumor cell populations enriched for individual Wnt components and BCC marker genes differed slightly between these datasets. Such discrepancies may stem from the fact that the scRNA-seq datasets analyzed in this study were derived either from mixed human BCC subtypes or exclusively from infiltrative BCCs, whereas the mouse BCC-like tumors originated from hair follicle–derived cells in K14-CreER; Ptch1+/−; p53fl/fl mice (Wang et al., 2011). Additionally, or instead, these differences could reflect inherent species-specific variations in tumor biology between human and mouse BCC models.

Unlike the scRNA-seq results showing a low abundance of BCC cells with high BCC_score and WNT_score, our RNAscope analyses revealed widespread WLS and FZD2 expression in human BCC tissues with abundant GLI1 expression. This discrepancy may reflect the limited sensitivity of scRNA-seq in detecting moderately expressed genes or suboptimal quality of the sequencing data (Chen et al., 2019). The broader expression patterns of WLS and FZD2 observed by RNAscope suggest that Wnt signaling may be more pervasive in BCC tissues than scRNA-seq alone would indicate. Therefore, the cells with high WNT_scores in the scRNA-seq datasets likely represent those with the strongest expression of Wnt pathway components.

Although previous studies using simultaneous Hh activation and Wnt/β-catenin inhibition have established an important role for Wnt/β-catenin signaling in early BCC development, they do not adequately address its role after Hh activation, largely due to the lack of appropriate mouse models. Our genetic mouse model allowed us to disrupt the Wnt pathway several days after Hh activation, circumventing this limitation. Our analyses reveal that blocking Wnt signaling significantly reduces the proliferation of GLI2ΔN-expressing cells. These findings suggest that oncogenic Hh signaling may drive BCC cell proliferation, at least in part, by coordinating Wnt/β-catenin activity—a mechanism analogous to that observed in other malignancies or developmental processes (Alvarez-Medina et al., 2009, Deng et al., 2019). Collectively, our data support a model in which Wnt/β-catenin signaling sustains pre-BCC-like tumor growth by promoting tumor cell proliferation and maintaining the long-term proliferative and tumor-repopulating potential of Axin2-expressing epidermal basal cells.

During embryonic development, epidermal cells produce Wnt ligands to regulate their own proliferation (Fu and Hsu, 2013, Zhu et al., 2014). Consistent with this, we show that GLI2ΔN-expressing cells also secrete Wnt ligands to support their own growth, as deleting epithelial Wls markedly reduced tumor cell proliferation. At the receptor level, our mouse genetic studies identified FZD2 as a critical mediator of Wnt/β-catenin signaling during pre-BCC-like tumor development. Given their elevated Wnt signaling demands, BCC cells may upregulate WLS and FZD2 to enhance Wnt ligand secretion and reception. Indeed, RNAscope analysis revealed significantly higher WLS and FZD2 mRNA expression in BCC tissue compared to adjacent non-cancerous epidermis. Furthermore, Fzd2 deletion unexpectedly induced apoptosis of GLI2ΔN-expressing cells. This finding suggests that FZD2 may mediate both β-catenin–dependent and –independent Wnt pathways (Zhu et al., 2023), or it may participate in additional, non-Wnt-dependent mechanisms (Tao et al., 2016), during pre-BCC-like tumor development.

Some limitations should be considered in interpreting our findings. First, the Fzd2fl used in this study is an unconventional allele that results in knockdown of Fzd2 mRNA rather than gene deletion and may potentially influence the expression of other Frizzled family members (Zhu et al., 2023). Therefore, the specific role of FZD2 in BCC development should be further validated using a conventional floxed allele (Michalski et al., 2026). Second, our study focused on pre-BCC-like tumors resembling superficial BCC. Whether Wnt/β-catenin signaling plays a similar role in other BCC subtypes, particularly those originating from hair follicles, remains to be determined. Future studies addressing these points will help to fully elucidate the contribution of Wnt signaling across diverse BCC contexts. Finally, although our inducible GLI2ΔN model recapitulates many features of human superficial BCC, no current mouse model fully captures the complexity of bona fide human BCC (Wong et al., 2026). Therefore, future studies using patient-derived BCC organoids and additional clinical specimens will be important to validate the role of Wnt/β-catenin signaling and FZD2 in human BCC and further strengthen the translational relevance of our findings.

In summary, our study demonstrates that Wnt/β-catenin signaling is activated in BCC cells with robust Hh pathway activity and is essential for both tumor cell proliferation and the contribution of Wnt-responsive epidermal basal cells to BCC-like tumor development. We further identify FZD2 as a critical Wnt receptor mediating proliferation and survival of GLI2ΔN-expressing cells, highlighting an autocrine Wnt signaling loop that may sustain pre-BCC-like lesion growth. These findings reveal a vital role for the Wnt/β-catenin pathway in pre-BCC-like cell proliferation and survival and suggest the potential clinical relevance of targeting the Wnt pathway in combination with Hh pathway inhibitors for the treatment of BCC.

Materials and Methods

Bioinformatic analysis

Human BCC scRNA-seq datasets, including GSE141526 (encompassing three BCC samples with discernible subtypes, GSM4205686, GSM4205687, GSM4205689, and two peri-tumor skin samples GSM4205690 and GSM4205691) and GSE181907 (encompassing five infiltrative BCC samples), along with the mouse BCC-like tumor scRNA-seq dataset GSE186184 (encompassing untreated and DMSO-treated samples), were assessed through the GEO website (https://www.ncbi.nlm.nih.gov/geo/). Data integration and cell clustering were conducted using Seurat (Stuart et al., 2019) (v5.0.1). The CellCycleScoring and SCTransform functions of Seurat were employed to mitigate the cell cycle heterogeneity effect and integrate the individual samples within the datasets. The resolutions applied in GSE141526, GSE181907, and GSE186184 were 0.2, 0.2, and 0.3, respectively.

Mice

The following mouse lines were used: Krt5-rtTA (#017519, Jackson Laboratories) (Diamond et al., 2000); tetO-Cre (strain #006234, Jackson Laboratories) (Perl et al., 2002); Axin2-CreERT2 (Xu et al., 2017); tetO-Gli2ΔN (Grachtchouk et al., 2011); Rosa26mTmG (strain #007676, Jackson Laboratories); Wlsfl/fl (strain #012888, Jackson Laboratories); β-cateninfl/fl (strain #004152, Jackson Laboratories); and Fzd2fl (Kadzik et al., 2014). Mice were assigned to experimental or control groups based on their genotypes, with control mice included in each experiment. Investigators were aware of the genotype to facilitate allocation and animal handling. Up to five mice were maintained per cage in a specific pathogen-free barrier facility on standard rodent laboratory chow (Purina, 5001). All mice were maintained on a mixed-strain background, and all animal experiments were performed under approved animal protocols according to institutional guidelines established by the Icahn School of Medicine at Mount Sinai IACUC committee. At least three mice of each genotype and experimental condition were analyzed for all the experiments in this study, and male and female mice showed comparable phenotypes.

Tamoxifen induction and doxycycline administration

One dose of 2 mg Tamoxifen, dissolved in corn oil (10 mg/ml), was injected intraperitoneally to perform lineage tracing. To induce DNA recombination or activate transgene expression, mice at postnatal day (P) 49 were fed with high-dose doxycycline (Sigma-Aldrich) water (200 μg/mL in 5% sucrose) or low-dose doxycycline water (20 μg/mL in 5% sucrose) for the indicated time. The doxycycline water was refreshed every 5 days.

Histology, immunofluorescence (IF), and TUNEL assay

Tissue was fixed with 4% paraformaldehyde (PFA)/PBS solution (Santa Cruz Biotechnology) overnight at 4°C, followed by paraffin-embedding and sectioning at 5 μm. The sections were then dewaxed and rehydrated through xylene substitute (Sigma-Aldrich) and graded ethanol solutions. Subsequently, sections were subjected to antigen retrieval using the unmask solution (Vector Labs). For IF staining, sections were blocked with 1% BSA and then incubated with primary antibodies overnight at 4°C, followed by PBS+0.1% Tween 20 (PBST) washing and incubation with fluorophore-conjugated secondary antibodies (Invitrogen and Vector Labs) and mounting medium (Invitrogen). To detect apoptosis, rehydrated sections were incubated with TUNEL labeling mix (Roche) at 37°C for 1 hour, washed with PBST, and mounted with DAPI-containing medium (Invitrogen).

Antibodies

The following primary antibodies were used: rat anti-Ki-67 (eBioscience, 14-5698-82, 1:200); Rabbit anti-GFP (Abcam, ab290, 1:1000); Mouse anti-Myctag (Cell signaling Technology, 2276S, 1:1000); rabbit anti-LEF1 (Cell signaling Technology, 2230s, 1:200); mouse anti-β-catenin (#C7207, Sigma-Aldrich, 1:1000); rabbit anti-KRT17 (Abcam, ab109725, 1:200); rabbit anti-SOX9 (Millipore, AB5535, 1:200).

The following secondary antibodies were used: donkey anti-rabbit IgG, Alexa Fluor 555 (Invitrogen, A-31572, 1:800); donkey anti-mouse IgG, Alexa Fluor 488 (Invitrogen, A-21202, 1:800); donkey anti-mouse IgG, Alexa Fluor 647 (Invitrogen, A-31571, 1:800); Streptavidin Fluorescein (Vector Labs, SA-5001, 1:500); Streptavidin Texas Red (Vector Labs, SA-5006, 1:500).

RNAscope

RNAscope was performed on paraffin sections according to the protocol provided by Advanced Cell Diagnostic (ACD)/Bio-Techne Spatial using RNAscope Multiplex Fluorescent Reagent Kit v2 (ACD #323100) and probes for Mm-Fzd2 (ACD #565781), Mm-Axin2 (ACD #400331), Mm-Wls (ACD #499191), Mm-Gli1 (ACD #311001-C2), Hs-FZD2 (ACD #497071), Hs-WLS (ACD #415851), Hs-GLI1 (ACD #310991-C2). The sections were observed and photographed using a Leica Microsystems DM5500B fluorescent microscope equipped with a digital camera. The mean fluorescence intensity (MFI) of signals in human BCC and adjacent epidermal tissues was quantified using ImageJ (v2.1.0/1.53c).

qPCR

Plantar skin samples were excised using forceps and scissors and incubated in 0.25% trypsin (Gibco) at 37 °C for 30 min. The epidermis was then separated from the dermis with forceps. Total RNA was isolated using TRIzol reagent (Thermo Fisher Scientific), further purified with the RNeasy Kit (Qiagen), and treated with RNase-free DNase (Qiagen) according to the manufacturers’ protocols. cDNA was synthesized from purified RNA using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Quantitative real-time PCR was performed on a StepOnePlus Real-Time PCR System (Applied Biosystems) using SYBR Green. Gapdh served as the internal control, and relative gene expression levels were calculated using the −ΔCT method. The following primers were used: Gapdh-F: 5′-GAGAGGCCCTATCCCAACTC-3′, Gapdh-R: 5′-GTGGGTGCAGCGAACTTTAT-3′; Wls-F: 5’-AGCACC AGGAAAAGAGGGCA-3’, Wls-R: 5’-TGTTTTATGGTGGTTCTTACGGAC-3’; Ptch1-F: 5’-CTGAGCCGCCTCCAAGTGTC-3’, Ptch1-R: 5’-CCCTGAGCGGTCCCTATTG-3’; Axin2-F, 5′-GCTGGTTGTCACCTACTTTTTCTGT-3′, Axin2-R, 5′-GGGGAGCACTGTCTCGTCGTC-3′; Fzd2-F, 5′-CTTCACGGTCACCACCTATTT-3′, Fzd2-R, 5′-AACGAAGCCCGCAATGTA-3.

Human BCC samples

Human BCC tissues containing tumor and peri-tumoral skin were obtained from the Department of Dermatology Skin Biology and Diseases Resource-based Center, University of Pennsylvania, as de-identified specimens discarded from surgical procedures. Samples were paraffin-embedded and sectioned for RNAscope analysis. BCC#1-3 are nodular BCCs, whereas BCC#4-5 are superficial BCCs.

Statistical analysis

Data analysis and visualization were performed using Microsoft Excel 2023 and R (v4.2.1) with ggplot2 (v3.5.2). Data are presented as mean ± SEM. Unpaired bilateral Student's t-tests were used to assess statistical significance between two groups of datasets unless otherwise specified. P<0.05 was considered statistically significant. Each data point in the dot plots in Figures 4, 5, 6, and S7 represents the percentage of Ki-67+ or TUNEL+ basal cells quantified from at least three randomly selected fields and 200 cells per mouse.

Supplementary Material

1

Acknowledgments

We thank Dr. Adam Glick for K5-rtTA mice and Dr. Edward E. Morrisey for Fzd2fl and Axin2-CreERT2 mice. We also thank the Mount Sinai Microscopy and Advanced Imaging CoRE (RRID: SCR_027237) for imaging assistance. This work was supported by NIAMS/NIH grants RO1 AR081322 and R37 AR047709 (SEM); a pilot grant to XZ from the NIAMS/NIH-supported Sinai Skin Biology and Diseases Resource-based Center (SBDRC) P30 AR079200; and the NIAMS/NIH-supported Penn Skin Biology and Diseases Resource-Based Center P30AR069589. XZ is the guarantor for this work.

Footnotes

Declaration of interests statement

The authors state no conflict of interest.

Declaration of Generative Artificial Intelligence (AI) or Large Language Models (LLMs)

No AI or LLM tools were used to prepare this manuscript.

Ethics Statement

De-identified, archived human BCC specimens were used for mRNA in situ hybridization under University of Pennsylvania Institutional Review Board protocol (IRB# 808225), and did not constitute human subjects research.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Data availability statement

The authors declare that the main data supporting the findings of this study are available within the article and its Supplemental Information files. The published scRNA-seq datasets can be accessed through the GEO website under accession numbers GSE141526 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE141526), GSE181907 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE181907), and GSE186184 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE186184). All correspondence and material requests should be addressed to Xuming Zhu or Sarah E. Millar.

References

  1. Alvarez-Medina R, Le Dreau G, Ros M, Martí E. Hedgehog activation is required upstream of Wnt signalling to control neural progenitor proliferation. Development 2009;136(19):3301–9. [DOI] [PubMed] [Google Scholar]
  2. Axelson M, Liu K, Jiang X, He K, Wang J, Zhao H, et al. U.S. Food and Drug Administration approval: vismodegib for recurrent, locally advanced, or metastatic basal cell carcinoma. Clin Cancer Res 2013;19(9):2289–93. [DOI] [PubMed] [Google Scholar]
  3. Bakshi A, Chaudhary SC, Rana M, Elmets CA, Athar M. Basal cell carcinoma pathogenesis and therapy involving hedgehog signaling and beyond. Mol Carcinog 2017;56(12):2543–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Carmon KS, Lin Q, Gong X, Thomas A, Liu Q. LGR5 interacts and cointernalizes with Wnt receptors to modulate Wnt/β-catenin signaling. Mol Cell Biol 2012;32(11):2054–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen G, Ning B, Shi T. Single-Cell RNA-Seq Technologies and Related Computational Data Analysis. Frontiers in Genetics 2019;Volume 10 - 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Deng Q, Li P, Che M, Liu J, Biswas S, Ma G, et al. Activation of hedgehog signaling in mesenchymal stem cells induces cartilage and bone tumor formation via Wnt/β-Catenin. eLife 2019;8:e50208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Diamond I, Owolabi T, Marco M, Lam C, Glick A. Conditional gene expression in the epidermis of transgenic mice using the tetracycline-regulated transactivators tTA and rTA linked to the keratin 5 promoter. J Invest Dermatol 2000;115(5):788–94. [DOI] [PubMed] [Google Scholar]
  8. Dong X, Wang C, Chen Z, Zhao W. Overcoming the resistance mechanisms of Smoothened inhibitors. Drug Discovery Today 2018;23(3):704–10. [DOI] [PubMed] [Google Scholar]
  9. Eberl M, Mangelberger D, Swanson JB, Verhaegen ME, Harms PW, Frohm ML, et al. Tumor Architecture and Notch Signaling Modulate Drug Response in Basal Cell Carcinoma. Cancer Cell 2018;33(2):229–43.e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. El-Bahrawy M, El-Masry N, Alison M, Poulsom R, Fallowfield M. Expression of β-catenin in basal cell carcinoma. British Journal of Dermatology 2003;148(5):964–70. [DOI] [PubMed] [Google Scholar]
  11. Fu J, Hsu W. Epidermal Wnt Controls Hair Follicle Induction by Orchestrating Dynamic Signaling Crosstalk between the Epidermis and Dermis. Journal of Investigative Dermatology 2013;133(4):890–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Grachtchouk M, Pero J, Yang SH, Ermilov AN, Michael LE, Wang A, et al. Basal cell carcinomas in mice arise from hair follicle stem cells and multiple epithelial progenitor populations. J Clin Invest 2011;121(5):1768–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Guerrero-Juarez CF, Lee GH, Liu Y, Wang S, Karikomi M, Sha Y, et al. Single-cell analysis of human basal cell carcinoma reveals novel regulators of tumor growth and the tumor microenvironment. Sci Adv 2022;8(23):eabm7981. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Haensel D, Gaddam S, Li NY, Gonzalez F, Patel T, Cloutier JM, et al. LY6D marks pre-existing resistant basosquamous tumor subpopulations. Nat Commun 2022;13(1):7520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Jho EH, Zhang T, Domon C, Joo CK, Freund JN, Costantini F. Wnt/beta-catenin/Tcf signaling induces the transcription of Axin2, a negative regulator of the signaling pathway. Mol Cell Biol 2002;22(4):1172–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Jing J, Wu Z, Wang J, Luo G, Lin H, Fan Y, et al. Hedgehog signaling in tissue homeostasis, cancers and targeted therapies. Signal Transduction and Targeted Therapy 2023;8(1):315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kadzik RS, Cohen ED, Morley MP, Stewart KM, Lu MM, Morrisey EE. Wnt ligand/Frizzled 2 receptor signaling regulates tube shape and branch-point formation in the lung through control of epithelial cell shape. Proc Natl Acad Sci U S A 2014;111(34):12444–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kriegl L, Horst D, Kirchner T, Jung A. LEF-1 expression in basal cell carcinomas. British Journal of Dermatology 2009;160(6):1353–6. [DOI] [PubMed] [Google Scholar]
  19. Liu J, Xiao Q, Xiao J, Niu C, Li Y, Zhang X, et al. Wnt/β-catenin signalling: function, biological mechanisms, and therapeutic opportunities. Signal Transduction and Targeted Therapy 2022;7(1):3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Metcalfe C, de Sauvage FJ. Hedgehog fights back: mechanisms of acquired resistance against Smoothened antagonists. Cancer Res 2011;71(15):5057–61. [DOI] [PubMed] [Google Scholar]
  21. Michalski MN, Diegel CR, Zhong ZA, Marshall ME, Foxa Wiartalla GE, Stevens PD, et al. Clarifying Frizzled 2 function in development through genetically validated mouse models. Disease Models & Mechanisms 2026;19(7):dmm052410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Nguyen NM, Cho J. Hedgehog Pathway Inhibitors as Targeted Cancer Therapy and Strategies to Overcome Drug Resistance. Int J Mol Sci 2022;23(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Perl AK, Wert SE, Nagy A, Lobe CG, Whitsett JA. Early restriction of peripheral and proximal cell lineages during formation of the lung. Proc Natl Acad Sci U S A 2002;99(16):10482–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Roky AH, Islam MM, Ahasan AMF, Mostaq MS, Mahmud MZ, Amin MN, et al. Overview of skin cancer types and prevalence rates across continents. Cancer Pathogenesis and Therapy 2025;3(2):89–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Saldanha G, Ghura V, Potter L, Fletcher A. Nuclear β-catenin in basal cell carcinoma correlates with increased proliferation. British Journal of Dermatology 2004;151(1):157–64. [DOI] [PubMed] [Google Scholar]
  26. Sánchez-Danés A, Larsimont JC, Liagre M, Muñoz-Couselo E, Lapouge G, Brisebarre A, et al. A slow-cycling LGR5 tumour population mediates basal cell carcinoma relapse after therapy. Nature 2018;562(7727):434–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Stuart T, Butler A, Hoffman P, Hafemeister C, Papalexi E, Mauck WM, 3rd, et al. Comprehensive Integration of Single-Cell Data. Cell 2019;177(7):1888–902 e21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Tao L, Zhang J, Meraner P, Tovaglieri A, Wu X, Gerhard R, et al. Frizzled proteins are colonic epithelial receptors for C. difficile toxin B. Nature 2016;538(7625):350–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Wang GY, Wang J, Mancianti ML, Epstein EH, Jr., Basal cell carcinomas arise from hair follicle stem cells in Ptch1(+/−) mice. Cancer Cell 2011;19(1):114–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Wolf L, Boutros M. The role of Evi/Wntless in exporting Wnt proteins. Development 2023;150(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Wong SY, Harms PW, Dlugosz AA. Does Basal Cell Carcinoma Arise from a Precursor Lesion? J Invest Dermatol 2026;146(4):905–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Xu M, Horrell J, Snitow M, Cui J, Gochnauer H, Syrett CM, et al. WNT10A mutation causes ectodermal dysplasia by impairing progenitor cell proliferation and KLF4-mediated differentiation. Nat Commun 2017;8:15397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Xue C, Chu Q, Shi Q, Zeng Y, Lu J, Li L. Wnt signaling pathways in biology and disease: mechanisms and therapeutic advances. Signal Transduction and Targeted Therapy 2025;10(1):106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Yang SH, Andl T, Grachtchouk V, Wang A, Liu J, Syu LJ, et al. Pathological responses to oncogenic Hedgehog signaling in skin are dependent on canonical Wnt/beta3-catenin signaling. Nat Genet 2008;40(9):1130–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Yerly L, Pich-Bavastro C, Di Domizio J, Wyss T, Tissot-Renaud S, Cangkrama M, et al. Integrated multi-omics reveals cellular and molecular interactions governing the invasive niche of basal cell carcinoma. Nat Commun 2022;13(1):4897. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Youssef KK, Lapouge G, Bouvrée K, Rorive S, Brohée S, Appelstein O, et al. Adult interfollicular tumour-initiating cells are reprogrammed into an embryonic hair follicle progenitor-like fate during basal cell carcinoma initiation. Nat Cell Biol 2012;14(12):1282–94. [DOI] [PubMed] [Google Scholar]
  37. Zhang Y, Beachy PA. Cellular and molecular mechanisms of Hedgehog signalling. Nature Reviews Molecular Cell Biology 2023;24(9):668–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Zhu X, Xu M, Leu NA, Morrisey EE, Millar SE. FZD2 regulates limb development by mediating β-catenin-dependent and -independent Wnt signaling pathways. Dis Model Mech 2023;16(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zhu XJ, Liu Y, Dai ZM, Zhang X, Yang X, Li Y, et al. BMP-FGF signaling axis mediates Wnt-induced epidermal stratification in developing mammalian skin. PLoS Genet 2014;10(10):e1004687. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2216697-supplement-1.pdf (18.8MB, pdf)

Data Availability Statement

The authors declare that the main data supporting the findings of this study are available within the article and its Supplemental Information files. The published scRNA-seq datasets can be accessed through the GEO website under accession numbers GSE141526 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE141526), GSE181907 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE181907), and GSE186184 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE186184). All correspondence and material requests should be addressed to Xuming Zhu or Sarah E. Millar.

RESOURCES