Letter | Published:

A cell identity switch allows residual BCC to survive Hedgehog pathway inhibition

Naturevolume 562pages429433 (2018) | Download Citation

Abstract

Despite the efficacy of Hedgehog pathway inhibitors in the treatment of basal cell carcinoma (BCC)1, residual disease persists in some patients and may contribute to relapse when treatment is discontinued2. Here, to study the effect of the Smoothened inhibitor vismodegib on tumour clearance, we have used a Ptch1Trp53 mouse model of BCC3 and found that mice treated with vismodegib harbour quiescent residual tumours that regrow upon cessation of treatment. Profiling experiments revealed that residual BCCs initiate a transcriptional program that closely resembles that of stem cells of the interfollicular epidermis and isthmus, whereas untreated BCCs are more similar to the hair follicle bulge. This cell identity switch was enabled by a mostly permissive chromatin state accompanied by rapid Wnt pathway activation and reprogramming of super enhancers to drive activation of key transcription factors involved in cellular identity. Accordingly, treatment of BCC with both vismodegib and a Wnt pathway inhibitor reduced the residual tumour burden and enhanced differentiation. Our study identifies a resistance mechanism in which tumour cells evade treatment by adopting an alternative identity that does not rely on the original oncogenic driver for survival.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Data availability

RNA-seq and ATAC–seq data supporting the findings of this study have been deposited in the Gene Expression Omnibus (GSE116966). All other data are available from the corresponding author upon reasonable request.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. 1.

    Sekulic, A. et al. Efficacy and safety of vismodegib in advanced basal-cell carcinoma. N. Engl. J. Med. 366, 2171–2179 (2012).

  2. 2.

    Sofen, H. et al. A phase II, multicenter, open-label, 3-cohort trial evaluating the efficacy and safety of vismodegib in operable basal cell carcinoma. J. Am. Acad. Dermatol. 73, 99–105.e1 (2015).

  3. 3.

    Wang, G. Y., Wang, J., Mancianti, M. L. & Epstein, E. H. Jr. Basal cell carcinomas arise from hair follicle stem cells in Ptch1+/– mice. Cancer Cell 19, 114–124 (2011).

  4. 4.

    Rubin, A. I., Chen, E. H. & Ratner, D. Basal-cell carcinoma. N. Engl. J. Med. 353, 2262–2269 (2005).

  5. 5.

    Robarge, K. D. et al. GDC-0449-a potent inhibitor of the hedgehog pathway. Bioorg. Med. Chem. Lett. 19, 5576–5581 (2009).

  6. 6.

    Von Hoff, D. D. et al. Inhibition of the hedgehog pathway in advanced basal-cell carcinoma. N. Engl. J. Med. 361, 1164–1172 (2009).

  7. 7.

    Basset-Seguin, N. et al. Vismodegib in patients with advanced basal cell carcinoma (STEVIE): a pre-planned interim analysis of an international, open-label trial. Lancet Oncol. 16, 729–736 (2015).

  8. 8.

    Vidal, V. P., Ortonne, N. & Schedl, A. SOX9 expression is a general marker of basal cell carcinoma and adnexal-related neoplasms. J. Cutan. Pathol. 35, 373–379 (2008).

  9. 9.

    Feil, R., Wagner, J., Metzger, D. & Chambon, P. Regulation of Cre recombinase activity by mutated estrogen receptor ligand-binding domains. Biochem. Biophys. Res. Commun. 237, 752–757 (1997).

  10. 10.

    Folgueras, A. R. et al. Architectural niche organization by LHX2 is linked to hair follicle stem cell function. Cell Stem Cell 13, 314–327 (2013).

  11. 11.

    Nijhof, J. G. et al. The cell-surface marker MTS24 identifies a novel population of follicular keratinocytes with characteristics of progenitor cells. Development 133, 3027–3037 (2006).

  12. 12.

    Pichery, M. et al. Endogenous IL-33 is highly expressed in mouse epithelial barrier tissues, lymphoid organs, brain, embryos, and inflamed tissues: in situ analysis using a novel Il-33-LacZ gene trap reporter strain. J. Immunol. 188, 3488–3495 (2012).

  13. 13.

    Joost, S. et al. Single-cell transcriptomics reveals that differentiation and spatial signatures shape epidermal and hair follicle heterogeneity. Cell Syst. 3, 221–237.e9 (2016).

  14. 14.

    Ge, Y. et al. Stem cell lineage infidelity drives wound repair and cancer. Cell 169, 636–650.e14 (2017).

  15. 15.

    Wang, S. et al. Target analysis by integration of transcriptome and ChIP-seq data with BETA. Nat. Protocols 8, 2502–2515 (2013).

  16. 16.

    Kim, T. H. et al. Broadly permissive intestinal chromatin underlies lateral inhibition and cell plasticity. Nature 506, 511–515 (2014).

  17. 17.

    Adam, R. C. et al. Pioneer factors govern super-enhancer dynamics in stem cell plasticity and lineage choice. Nature 521, 366–370 (2015).

  18. 18.

    Yang, S. H. et al. Pathological responses to oncogenic Hedgehog signaling in skin are dependent on canonical Wnt/β3-catenin signaling. Nat. Genet. 40, 1130–1135 (2008).

  19. 19.

    Youssef, K. K. 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. 14, 1282–1294 (2012).

  20. 20.

    Jho, E. H. et al. Wnt/β-catenin/Tcf signaling induces the transcription of Axin2, a negative regulator of the signaling pathway. Mol. Cell. Biol. 22, 1172–1183 (2002).

  21. 21.

    Hsieh, J. C. et al. A new secreted protein that binds to Wnt proteins and inhibits their activities. Nature 398, 431–436 (1999).

  22. 22.

    Lim, X. et al. Interfollicular epidermal stem cells self-renew via autocrine Wnt signaling. Science 342, 1226–1230 (2013).

  23. 23.

    Tian, H. et al. Opposing activities of Notch and Wnt signaling regulate intestinal stem cells and gut homeostasis. Cell Reports 11, 33–42 (2015).

  24. 24.

    Niemann, C. Differentiation of the sebaceous gland. Dermatoendocrinol 1, 64–67 (2009).

  25. 25.

    Yauch, R. L. et al. Smoothened mutation confers resistance to a Hedgehog pathway inhibitor in medulloblastoma. Science 326, 572–574 (2009).

  26. 26.

    Kahn, M. Can we safely target the WNT pathway? Nat. Rev. Drug Discov. 13, 513–532 (2014).

  27. 27.

    Oser, M. G., Niederst, M. J., Sequist, L. V. & Engelman, J. A. Transformation from non-small-cell lung cancer to small-cell lung cancer: molecular drivers and cells of origin. Lancet Oncol. 16, e165–e172 (2015).

  28. 28.

    Vasioukhin, V., Degenstein, L., Wise, B. & Fuchs, E. The magical touch: genome targeting in epidermal stem cells induced by tamoxifen application to mouse skin. Proc. Natl Acad. Sci. USA 96, 8551–8556 (1999).

  29. 29.

    Kasper, M. et al. Wounding enhances epidermal tumorigenesis by recruiting hair follicle keratinocytes. Proc. Natl Acad. Sci. USA 108, 4099–4104 (2011).

  30. 30.

    Marino, S., Vooijs, M., van Der Gulden, H., Jonkers, J. & Berns, A. Induction of medulloblastomas in p53-null mutant mice by somatic inactivation of Rb in the external granular layer cells of the cerebellum. Genes Dev. 14, 994–1004 (2000).

  31. 31.

    Wu, T. D. & Nacu, S. Fast and SNP-tolerant detection of complex variants and splicing in short reads. Bioinformatics 26, 873–881 (2010).

  32. 32.

    Law, C. W., Chen, Y., Shi, W. & Smyth, G. K. voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 15, R29 (2014).

  33. 33.

    Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).

  34. 34.

    Srinivasan, K. et al. Untangling the brain’s neuroinflammatory and neurodegenerative transcriptional responses. Nat. Commun. 7, 11295 (2016).

  35. 35.

    Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 9, R137 (2008).

  36. 36.

    Robinson, M. D. & Oshlack, A. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 11, R25 (2010).

  37. 37.

    Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576–589 (2010).

  38. 38.

    Robinson, J. T. et al. Integrative genomics viewer. Nat. Biotechnol. 29, 24–26 (2011).

  39. 39.

    Ramírez, F., Dündar, F., Diehl, S., Grüning, B. A. & Manke, T. deepTools: a flexible platform for exploring deep-sequencing data. Nucleic Acids Res. 42, W187–W191 (2014).

  40. 40.

    Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102, 15545–15550 (2005).

  41. 41.

    Tian, H. et al. A reserve stem cell population in small intestine renders Lgr5-positive cells dispensable. Nature 478, 255–259 (2011).

  42. 42.

    Jensen, K. B., Driskell, R. R. & Watt, F. M. Assaying proliferation and differentiation capacity of stem cells using disaggregated adult mouse epidermis. Nat. Protocols 5, 898–911 (2010).

  43. 43.

    Greco, V. et al. A two-step mechanism for stem cell activation during hair regeneration. Cell Stem Cell 4, 155–169 (2009).

  44. 44.

    Garza, L. A. et al. Bald scalp in men with androgenetic alopecia retains hair follicle stem cells but lacks CD200-rich and CD34-positive hair follicle progenitor cells. J. Clin. Invest. 121, 613–622 (2011).

  45. 45.

    Blanpain, C., Lowry, W. E., Geoghegan, A., Polak, L. & Fuchs, E. Self-renewal, multipotency, and the existence of two cell populations within an epithelial stem cell niche. Cell 118, 635–648 (2004).

  46. 46.

    Snippert, H. J. et al. Lgr6 marks stem cells in the hair follicle that generate all cell lineages of the skin. Science 327, 1385–1389 (2010).

  47. 47.

    Page, M. E., Lombard, P., Ng, F., Göttgens, B. & Jensen, K. B. The epidermis comprises autonomous compartments maintained by distinct stem cell populations. Cell Stem Cell 13, 471–482 (2013).

  48. 48.

    Yang, H., Adam, R. C., Ge, Y., Hua, Z. L. & Fuchs, E. Epithelial-mesenchymal micro-niches govern stem cell lineage choices. Cell 169, 483–496.e13 (2017).

  49. 49.

    Yaari, G., Bolen, C. R., Thakar, J. & Kleinstein, S. H. Quantitative set analysis for gene expression: a method to quantify gene set differential expression including gene-gene correlations. Nucleic Acids Res. 41, e170 (2013).

  50. 50.

    O’Neill, J. W. & Bier, E. Double-label in situ hybridization using biotin and digoxigenin-tagged RNA probes. Biotechniques 17, 870, 874–875 (1994).

  51. 51.

    Beer, T. W., Shepherd, P. & Theaker, J. M. Ber EP4 and epithelial membrane antigen aid distinction of basal cell, squamous cell and basosquamous carcinomas of the skin. Histopathology 37, 218–223 (2000).

Download references

Acknowledgements

We thank J. Diaz for mouse colony management; the IVP team for animal dosing; S. Flanagan and V. Nunez for necropsy support; the Pathology core for help with histology; S. Biswas for help with LCM; the NGS laboratory for RNA-seq and ATAC–seq; I. Caro for help with human samples; and J. Svärd, R. Toftgard, and S. Teglund for Ptch1fl/fl mice.

Reviewer information

Nature thanks R. Shivdasani and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Author information

Affiliations

  1. Department of Molecular Oncology, Genentech, San Francisco, CA, USA

    • Brian Biehs
    • , Gerrit J. P. Dijkgraaf
    • , Soufiane Boumahdi
    •  & Frederic J. de Sauvage
  2. Department of Bioinformatics and Computational Biology, Genentech, San Francisco, CA, USA

    • Robert Piskol
  3. Department of Translational Oncology, Genentech, San Francisco, CA, USA

    • Bruno Alicke
    •  & Stephen E. Gould
  4. Department of Research Pathology, Genentech, San Francisco, CA, USA

    • Franklin Peale

Authors

  1. Search for Brian Biehs in:

  2. Search for Gerrit J. P. Dijkgraaf in:

  3. Search for Robert Piskol in:

  4. Search for Bruno Alicke in:

  5. Search for Soufiane Boumahdi in:

  6. Search for Franklin Peale in:

  7. Search for Stephen E. Gould in:

  8. Search for Frederic J. de Sauvage in:

Contributions

B.B., G.J.P.D., S.E.G. and F.J.d.S. conceptualized the project, designed experiments and analysed data. B.B. performed IHC, immunofluorescence and ISH on skin sections, isolated tumour nests by LCM and tumour cells by FACS, and performed qRT–PCR. G.J.P.D. performed initial characterization of residual disease. R.P. performed bioinformatics analyses. B.A. monitored mice for tumour burden and coordinated drug dosing. S.B. analysed ATAC–seq data. F.P. performed histopathological evaluation of skin sections. B.B., G.J.P.D. and F.J.d.S. wrote the manuscript.

Competing interests

All authors are employees of Genentech and own shares of Roche.

Corresponding author

Correspondence to Frederic J. de Sauvage.

Extended data figures and tables

  1. Extended Data Fig. 1 Limited apoptosis in tumour nests early during vismodegib treatment and recovery of residual BCCs.

    Related to Figs. 1a–d, 2e–h. ad, Time-course analysis of cell death in our K14CreER;Ptch1fl/fl;Trp53fl/fl mouse model of BCC during vismodegib treatment. Representative images of skin sections collected after indicated length of vismodegib treatment were stained for cleaved caspase 3 (CC3, red) and nuclei (DAPI, blue) (n = 3 per time point). The apoptosis marker CC3 is lacking in untreated BCC (a), but could be detected in rare tumour cells after just 1 day of treatment (b). CC3+ tumour cells were most abundant after 3 days of treatment (c) but were no longer detectable after 7 days of treatment (d), indicating that additional mechanism(s) contribute to the tumour de-bulking that occurs at later time points. e, f, Additional images of Gli1 ISH on skin sections from vismodegib-treated BCC mice shown in Fig. 2a. Note that the robust Gli1 expression endures when animals remain off drug for 6 (e) or 12 days (f). g, h, Representative images of skin sections stained for BrdU (red) and nuclei (DAPI, blue) from untreated BCC mice and mice treated with vismodegib for 7 days (n = 3 per group). g, BrdU dosing before vismodegib treatment labels proliferating cells in BCCs and in the epidermis (above dotted line). h, Label-retaining BrdU+ cells can be found only in residual tumour nests (below dotted line) after 7 days of vismodegib treatment, as epidermal cells continue to proliferate in the presence of drug and dilute out the label. All experiments were performed at least twice. Scale bars, 100 μm; n represents the number of mice.

  2. Extended Data Fig. 2 Vismodegib treatment does not cause BCCs to differentiate into normal epidermis.

    Related to Fig. 1. Detailed time-course analysis of epidermal differentiation marker expression in BCC. Skin samples were collected from BCC mice after indicated length of vismodegib treatment. Representative images of skin sections are shown for n = 4 per time point. Residual tumour nests could be identified with the BCC marker SOX9 throughout treatment (column 1; red). By contrast, the supra-basal marker KRT1 (column 2; red) and the terminal differentiation marker loricrin (column 4; red) were at no point expressed in either untreated or residual BCCs. The supra-basal marker KRT10 was detected in both untreated and residual BCC (column 3; red). Its expression pattern within tumour nests gradually changed from a proximal/dermal location in untreated BCCs to the centre of residual BCCs, where it no longer overlapped with the basal marker KRT5 (green; outlined in 28 d vismo). Scale bars, 100 μm; n represents the number of mice. DAPI nuclear stain is blue.

  3. Extended Data Fig. 3 Identification of differentially expressed genes in BCC after vismodegib treatment and their normal expression pattern in telogen skin.

    Related to Fig. 3. a, b, Volcano plots depicting the number of differentially expressed genes that are either significantly upregulated (green dots) or downregulated (blue dots) in residual disease after 28 days of vismodegib treatment relative to either Ki67+ (a) or Ki67 (b) untreated BCC. Significance was tested with a moderated t-statistic (two-sided) and P values were adjusted for multiple testing with the Benjamini–Hochberg procedure (log2FC ≥ 1, adjusted P ≤ 0.05; n = 5 per group). c, Robust correlation between the two differential gene expression estimates from the comparisons made in a and b. d, QuSAGE of indicated signatures in vismodegib-treated versus Ki67 untreated BCC samples (n = 5 per group). Coloured violins depict the differential expression of the entire gene set quantified by a probability density function calculated using QuSAGE. Each grey dot represents the log2[fold change] in expression of a particular gene in the signature. e, f, GSEA ranking of indicated skin compartment-specific gene signatures using the differential gene expression estimates from the comparisons made in a, b. Changes in expression were determined for n = 5 per group. gi, Expression pattern of select skin epithelial markers in wild-type telogen skin. Representative images are shown for n = 4 per condition. g, Lhx2 mRNA is particularly enriched in the bulge of hair follicles. h, Antibody stain revealing nuclear localization of IL-33 in the isthmus and interfollicular epidermis. i, DefB6 mRNA is particularly enriched in the isthmus region of hair follicles. j, k, Representative images of Defb6 ISH on skin from untreated BCC mice (j) and animals treated with vismodegib for 28 days (k; n = 4 per condition). Arrowheads indicate DefB6+ residual BCCs. l, Relative expression of Lhx2 (hair follicle bulge) and MTS24 (ISTH) in BCCs treated with vismodegib for the indicated length of time. Data are plotted as mean ± s.e.m.; n = 3 per condition. Scale bars, 100 μm; n represents the number of mice. Source data

  4. Extended Data Fig. 4 Residual BCCs reinitiate the hair follicle bulge program when vismodegib treatment is discontinued.

    Related to Fig. 3. ad, Lineage tracing in BCC mice with a Cre-inducible TdTomato reporter allele after 28 days of vismodegib treatment (n = 3 per group). Untreated TdTomato+ BCCs display robust LHX2 antibody staining (a), but completely lack IL-33 antibody staining (c). TdTomato+ residual BCCs lack LHX2 antibody staining (b), but stain positive for IL-33 protein (d). eh, Return of LHX2 expression in residual BCCs after vismodegib treatment. Representative images of skin sections stained for BrdU (red), nuclei (DAPI, blue) and LHX2 (green) from BCC mice treated with vismodegib for 28 days that were allowed to recover for the indicated times (n = 3 per group). e, BrdU and LHX2 co-staining is virtually absent in residual BCC (outlined) immediately after treatment. f, LHX2 staining is weak in BrdU-labelled tumour nests after 3 days of recovery. g, LHX2 staining is more prominent in residual tumours after 6 days of recovery. h, LHX2 staining is the most robust and shows maximum overlap with the BrdU label after 12 days of recovery. i, Proportion of BrdU–LHX2 double-positive tumour cells in skin sections from eh. Graph shows mean ± s.d. Each data point represents the proportion of BrdU-labelled cells that were LHX2-positive within a single tumour nest (n ≥ 12 tumours from 3 mice per time-point). Scale bars, 100 μm; n represents the number of mice except for i, where it represents the number of tumours. Source data

  5. Extended Data Fig. 5 Il-33 expression in BCCs during vismodegib treatment.

    Related to Fig. 3. a, Expression of Il33 in LCM tumours from untreated BCC mice and mice treated with vismodegib for 28 days (n = 5 per group). RNA-seq expression data were quantile normalized and represented as nRPKM. Data plotted are mean ± s.e.m.; data points (dots, squares) indicate Il33 expression in tumours from individual mice. bf, Detailed time-course analysis of IL-33 expression in BCCs during vismodegib treatment. Representative images of skin samples collected from both untreated and treated BCC mice are shown (n = 3 per treatment and time point). IL-33 gradually appears in residual BCCs over the course of treatment. This nuclear factor can readily be detected in the basal layer of the epidermis, but is lacking in untreated BCCs (b) and in tumours treated with vismodegib for just 1 (c) or 3 (d) days. Nuclear IL-33 begins to emerge in a fraction of the tumour nests after 7 days of treatment (e), and most residual BCCs express IL-33 by 14 days of treatment (f). Scale bars, 100 μm; n represents the number of mice.

  6. Extended Data Fig. 6 ATAC–seq of FACS-sorted BCC cells and regulatory potential of accessible chromatin regions.

    Related to Fig. 3. a, b, We crossed our mouse model of BCC to the Lgr5DTRGFP allele41, to enable FACS isolation of both untreated (a) and residual (b) tumour cells. Representative images of skin sections stained for GFP (green) and nuclei (DAPI, blue) are shown for n = 3 per group. c, Gating for low side scatter enriched for epidermal cells after dissociation of back skin samples42. d, Final gating strategy used for the FACS isolation of BCC cells. Lgr5+ hair follicle bulge stem cells were excluded by staining for CD34. e, Heat-map of differentially accessible chromatin regions that either close (top) or open (bottom) in response to indicated length of vismodegib treatment. Data are shown as normalized peak counts per million genomic DNA fragments averaged from n = 3 per condition. f, Matrix of Pearson correlation coefficients showing the overlap in chromatin accessibility between samples based on the 5,000 most variable peak regions. Pearson’s ρ statistic was used to determine each correlation. g, Volcano plots depicting chromatin accessibility in sorted BCC cells after 6 (left) and 14 days (right) of vismodegib treatment relative to untreated (MCT) controls (n = 3 per group). Significance was tested with a moderated t-statistic (two-sided) and P values were adjusted for multiple testing with the Benjamini–Hochberg procedure. Each dot represents a peak (an open chromatin region). Common peaks are grey, and specific peaks are either blue (untreated) or green (6 and 14 days vismodegib). The table summarizes the number of open chromatin regions according to promoters and enhancers. h, Chromatin traces showing averaged ATAC peaks at the Gata6 and Ahr loci; n = 3 per group. Orange rectangles mark TCF binding elements in chromatin that are sensitive to vismodegib. i, Volcano plots depicting the number of differentially expressed genes that are either significantly upregulated (red dots; 150 at 6 d and 794 at 14 d) or downregulated (purple dots; 416 at 6 d and 1,129 at 14 d) in sorted BCC cells after 6 (left) and 14 days (right) of vismodegib treatment relative to untreated controls. Significance was tested with a moderated t-statistic (two-sided) and P values were adjusted for multiple testing with the Benjamini–Hochberg procedure (log2FC ≥ 1, adjusted P ≤ 0.05; n = 3 per group). j, k, BETA analysis graphs depicting the effect of differentially open enhancers (j) and promoter peaks (k) on gene expression in sorted BCC cells after 6 (top) and 14 days (bottom) of vismodegib treatment relative to untreated controls (n = 3 per group). Genes were ranked from high to low according to the regulatory potential of the corresponding chromatin peak. Purple lines represent vismodegib-downregulated genes, while red lines represent vismodegib-upregulated genes. A one-tailed Kolmogorov–Smirnov test40 was used to determine whether the up- and downregulated groups differed significantly (shown as P values in parentheses) from the static group of transcriptionally unchanged genes (dashed lines). Scale bars, 100 μm; n represents the number of mice. Source data

  7. Extended Data Fig. 7 Enrichment of transcription factor binding motifs in differentially regulated enhancers and their regulatory potential.

    Related to Fig. 3. a, b, Enrichment of transcription factor binding sites in non-promoter peaks that are either differentially closing (a) or opening (b) after 6 (top) and 14 days (bottom) of vismodegib treatment relative to untreated BCC cells (n = 3 per group). A hypergeometric test implemented in HOMER was used to identify enriched motifs, which are ranked by P value (top to bottom) over increasingly stringent P value cutoffs for peak calling (left to right). c, d, BETA analysis graphs depicting the effect of differentially open enhancer peaks with indicated transcription factor binding motifs on gene expression in sorted BCC cells after 6 (c) and 14 days (d) of vismodegib treatment relative to untreated controls (n = 3 per group). Genes were ranked from high to low according to the regulatory potential of the corresponding chromatin peak. Purple lines represent vismodegib-downregulated genes, while red lines represent vismodegib-upregulated genes. A one-tailed Kolmogorov–Smirnov test40 was used to determine whether the up- and downregulated groups differed significantly (shown as P values in parentheses) from the static group of transcriptionally unchanged genes (dashed lines). n represents the number of mice. Source data

  8. Extended Data Fig. 8 AXIN2 is also induced in human residual BCC.

    Related to Fig. 4. a, b, Representative images of H&E-stained skin sections from untreated BCC mice (a) and from animals treated with vismodegib for 2 days (b; n = 3 per group). Note that the relatively short treatment has no impact on the overall morphology and size of the tumour nests. ch, Representative images of Ber-EP4 (brown) staining and AXIN2 ISH (red) on skin sections collected from a patient with BCC before and after 12 weeks of vismodegib treatment. The human BCC marker Ber-EP451 was used to identify nodular BCC in a screening biopsy (c) and residual BCC at the tumour site after treatment (f). d, e, g, h, Representative images of AXIN2 ISH on skin sections immediately adjacent to the ones shown in c and f. d, Untreated nodular BCC lacks AXIN2 mRNA. g, Residual BCC contains elevated AXIN2 mRNA levels. e, h, High-magnification views of boxed regions in d and g, respectively. i, Axin2, Wif1 and Lhx2 ISH on skin sections from BCC mice after 28 days of vismodegib treatment. Skin samples were collected 0 (top), 6 (middle) and 12 (bottom) days after the end of treatment, and representative images are shown for n = 3 per group. Note that Axin2 expression rapidly declines during the recovery period, whereas the expression of both Wif1 and Lhx2 progressively increase. All experiments were replicated at least twice. Scale bars: c, f, 90 μm; other panels, 100 μm.

  9. Extended Data Fig. 9 Effects of the vismodegib and anti-LRP6 combination treatment on mouse BCC.

    Related to Fig. 4. ac, Representative images of Axin2 ISH on skin sections from wild-type telogen mice (a) and from BCC mice treated for 28 days with either vismodegib alone (b) or vismodegib and anti-LRP6 (c). Note that co-treatment with anti-LRP6 strongly reduces Axin2 mRNA levels in residual BCCs (n = 3 per group). d, Relative expression of Wnt target genes in the small intestine of BCC mice treated for 28 days with either vismodegib alone (circles) or vismodegib and anti-LRP6 (squares; n = 4 per group). Data are plotted as mean ± s.e.m. Three out of four BCC mice that received vismodegib and anti-LRP6 experienced at least an 80% reduction in expression of all four Wnt target genes that was accompanied by a low residual tumour burden (black squares; 6, 9 and 18 residual tumours per length of skin), while the one animal with poor Wnt pathway modulation had a much higher residual tumour burden (red square; 31 residual tumours per length of skin). eg, Representative images of Oil Red O (orange)-stained skin sections from untreated BCC mice (e) and mice treated for 28 days with either vismodegib alone (f) or vismodegib and anti-LRP6 (g) (nuclear blue counter stain, haematoxylin; n = 4 per treatment). h, Proportion of Oil Red O+ tumours per length of skin in BCC mice from f, g. Average percentages ± s.d. are shown for n = 4 per group. ip, BrdU labelling of residual BCCs and their subsequent growth after cessation of indicated treatment. Representative images of residual BCCs stained for BrdU (red), nuclei (DAPI, blue) and Ki67 (green) are shown (n ≥ 2 per treatment and time point) from BCC mice treated for 28 days with either vismodegib alone (il) or vismodegib and anti-LRP6 (mp). Note that residual BCCs from both treatment groups are fully proliferating by 12 days of recovery and have diluted out the BrdU label. q, Proportion of BrdU+ tumours per length of skin from BCC mice from ip. Average percentages ± s.d. are shown for n ≥ 2 per group. All experiments were replicated at least twice. Scale bars, 100 μm; n represents the number of mice. Source data

  10. Extended Data Table 1 Skin epithelial signatures used in this study

Supplementary information

  1. Supplementary Table

    This file contains Supplementary Table 1

  2. Reporting Summary

Source data

About this article

Publication history

Received

Accepted

Published

Issue Date

DOI

https://doi.org/10.1038/s41586-018-0596-y

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.