Basal cell carcinomas in mice arise from hair follicle stem cells and multiple epithelial progenitor populations

Abstract

Uncontrolled Hedgehog (Hh) signaling leads to the development of basal cell carcinoma (BCC), the most common human cancer, but the cell of origin for BCC is unclear. While Hh pathway dysregulation is common to essentially all BCCs, there exist multiple histological subtypes, including superficial and nodular variants, raising the possibility that morphologically distinct BCCs may arise from different cellular compartments in skin. Here we have shown that induction of a major mediator of Hh signaling, GLI2 activator (GLI2ΔN), selectively in stem cells of resting hair follicles in mice, induced nodular BCC development from a small subset of cells in the lower bulge and secondary hair germ compartments. Tumorigenesis was markedly accelerated when GLI2ΔN was induced in growing hair follicles. In contrast, induction of GLI2ΔN in epidermis led to the formation of superficial BCCs. Expression of GLI2ΔN at reduced levels in mice yielded lesions resembling basaloid follicular hamartomas, which have previously been linked to low-level Hh signaling in both mice and humans. Our data show that the cell of origin, tissue context (quiescent versus growing hair follicles), and level of oncogenic signaling can determine the phenotype of Hh/Gli-driven skin tumors, with high-level signaling required for development of superficial BCC-like tumors from interfollicular epidermis and nodular BCC-like tumors from hair follicle stem cells.

Figure 1. BCC-like skin tumors arise from hair follicle stem cells.

(A) General scheme showing triple-transgenic model combining Cre-lox and tet/doxycycline-regulated technologies to achieve tight spatial and temporal control of transgene expression. The 3 components include mouse strains carrying: (i) a tissue-specific promoter (TSP) driving expression of a hormone-inducible Cre allele fused to a steroid binding domain (CreSBD); (ii) a Cre-inducible “tet on” rtTA, under the control of the ubiquitously expressed ROSA26 promoter (R26-LSL-rtTA) (37); and (iii) an oncogene downstream of multiple tetO sequences and minimal CMV promoter that is induced by rtTA when doxycycline (doxy) is present. (B) Cellular compartments in a quiescent (telogen) hair follicle. Most hair follicle stem cells are localized to the bulge and secondary hair germ compartments. (C) Compartments in which K15-CrePR1 mice can drive recombination in K15-CrePR1;R26-LSL-rtTA;tetO-GLI2ΔN (iK15;rtTA;GLI2ΔN) mice to activate rtTA expression. (D) Synchronized hair growth cycle in postnatal mouse skin, with approximate mouse ages and timing of doxycycline treatment/GLI2ΔN transgene induction indicated. (E) Spontaneous development of microscopic skin tumors from telogen hair follicle stem cells in dorsal skin of iK15;rtTA;GLI2ΔN mice, after 3 weeks of doxycycline treatment. Note tumor development from lowermost follicle in a region corresponding to the secondary hair germ. Original magnification, ×200 (left panel); ×400 (right panels). (F) Spontaneous tumor development from K15⁺ follicle stem cells at other body sites 3 weeks after GLI2ΔN induction. Original magnification, ×100 (tail, dorsal paw, and snout); ×200 (ear).

Figure 2. Nodular skin tumors arising from K15⁺ stem cells express BCC markers.

(A) Immunostaining showing BCC-like keratin profile (K5 and K17 expressed, K1 and K6 not expressed). Tumor cells also express the BCC marker Sox9 and are highly proliferative, based on Ki67 immunostaining. BCC-like tumors also express MYC-tagged GLI2ΔN transgene product. Original magnification, ×200. (B) In situ hybridization showing transcripts for Hh target genes Gli1 and Ptch1 in early-stage microscopic tumors, along with GLI2ΔN, detected using transgene-specific riboprobe against the SV40 poly(A) tail. Original magnification, ×200.

Figure 3. Robust BCC-like skin tumor development from the Lgr5⁺ subset of follicle stem cells.

(A) Lgr5-CreER–driven recombination is limited to the secondary hair germ and lowermost bulge region. (B) Hair growth cycle and timing of doxycycline treatment/GLI2ΔN expression. TAM, tamoxifen. (C) Spontaneous development of microscopic skin tumors from telogen hair follicle stem cells in dorsal skin of iLgr5;rtTA;GLI2ΔN mice after 3 weeks of doxycycline treatment. Note early tumor development from the secondary hair germ. Original magnification, ×200 (left panels); ×400 (right panels). (D) Spontaneous nodular BCC-like tumor development from Lgr5⁺ follicle stem cells at other body sites. Original magnification, ×200.

Figure 4. Impaired responsiveness of follicle bulge stem cells to acute induction of GLI2ΔN.

(A) Histology of epithelial compartments in tail skin of control and K14-rtTA;tetO-GLI2ΔN (K14;GLI2ΔN) bitransgenic mice reveals basaloid cell hyperplasia in epidermis, sebaceous gland, and secondary hair germ (yellow asterisks in right panels), but not in the bulge or central isthmus (marked with black and dotted yellow bars, respectively). Original magnification, ×200 (upper panels); ×600 (lower panels). (B) Immunostaining for MYC reveals GLI2ΔN transgene expression in basal epithelial compartments, with the exception of cells in the central isthmus (white brackets). Coimmunostaining for GLI2ΔN (MYC) and PCNA reveals increased proliferation in all compartments expressing GLI2ΔN, although the fraction of PCNA⁺ cells is lower in the bulge than other compartments (D). Original magnification, ×200 (upper panels); ×600 (lower panels). (C) Coimmunostaining for the bulge marker K15 and PCNA confirms proliferation in bulge stem cells in tail as well as dorsal skin. Original magnification, ×400. (D) Proliferative response to GLI2ΔN is approximately 50% lower in bulge cells than in epidermal or sebaceous gland cells. (E) Progressive increase in number of Ki67⁺ normally quiescent bulge cells 2 and 5 days after activation of GLI2ΔN transgene expression by using doxycycline. Data in D and E are presented as mean ± SEM; *P < 0.05, **P < 0.005. (F) Increased apoptosis in bulge compartment of GLI2ΔN-expressing mice compared with controls, based on immunostaining for activated caspase-3 (arrows). Original magnification, ×600.

Figure 5. Induction of GLI2ΔN expression leads to BCC-like tumors derived from all 3 epithelial skin lineages.

(A) Widespread potential recombination pattern for K5-CreER transgenic driver used to generate iK5;rtTA;GLI2ΔN mice. (B) Treatment with low-dose tamoxifen to drive recombination in a limited number of cells yielded BCC-like tumors in tail arising from epidermis (1), sebaceous gland (2), and lower follicle (3), with the bulge compartment (asterisk) unaffected. Superficial BCC-like tumor development in hairless skin confirms that these tumors can arise from interfollicular epidermis. Original magnification, ×200 (tail and ear); ×100 (dorsal paw and snout); ×400 (volar).

Figure 6. Anagen accelerates GLI2ΔN-driven tumorigenesis.

(A) Hair growth cycle in dorsal skin showing timing of depilation (to induce anagen) and doxycycline treatment (to activate GLI2ΔN expression in either early or mid-anagen). (B) Depilation of control mice activates hair follicle growth, with anagen follicles extending into the subcutaneous adipose layer. Tangential section (right panel) shows outer root sheath compartment of the anagen follicle (arrowheads) and pigmented hair shafts (asterisks). Original magnification, ×100 (left and middle panels); ×400 (right panel). (C) Induction of GLI2ΔN in iK15;rtTA;GLI2ΔN mice during early anagen leads to widespread tumor development from growing hair follicles in 2 weeks, at a time when spontaneous tumors in adjacent skin are rare (left panel). In some areas, tumors are contiguous with and appear to replace the outer root sheath compartment of the anagen follicle (arrowheads in right panel). Original magnification, ×40 (left panel); ×100 (middle panel); ×400 (right panel). (D) GLI2ΔN induction for 1 week in mature hair follicles (mid-anagen) shows tumor derivation directly from the outer root sheath. Immunostaining for transgene (right panel) reveals GLI2ΔN expression in all regions exhibiting BCC-like changes. Dashed lines delineate hair follicles. Typical BCC-like tumors develop after an additional week of transgene expression (lower panels). Original magnification, ×100 (upper-left panel); ×400 (upper-middle and -right panels); ×40 (lower-left panel); ×100 (lower-middle panel); ×400 (lower-right panel).

Figure 7. Oncogenic Smo drives outer root sheath hyperplasia but not nodular BCC-like tumor development.

(A) Schematic depicting timing of depilation and transgene activation in iK15;rtTA;SmoA1 mice. (B) In contrast to analogous GLI2ΔN-expressing mice, which produce nodular BCC-like tumors when the same experimental approach is used (Figure 6C), activation of SmoA1 in early anagen leads to hyperplasia of the outer root sheath (compare distance between arrowheads in upper- and lower-right panels) but no evidence of BCC-like tumors. Original magnification, ×200.

Figure 8. Low-level expression of GLI2ΔN gives rise to basaloid hamartomas instead of nodular BCC-like tumors.

(A) Histology showing nodular BCC-like tumors in iK5;rtTA;GLI2ΔN mouse treated with 1 gm/kg doxycycline in chow and 200 μg/ml doxycycline in drinking water (GLI2ΔN-high), compared with basaloid hamartomas in K5;rtTA;GLI2ΔN mice treated with 2 μg/ml doxycycline (GLI2ΔN-low) in drinking water, which resemble lesions that arise in mice harboring a Cre-inducible M2SMO allele (SMO). (B) Ki67 reveals limited proliferation at the periphery of basaloid hamartomas arising in GLI2ΔN-low and SMO mice, compared with diffuse proliferation in nodular BCCs in GLI2ΔN-high mice. Dashed lines in right panels outline the extent of epithelial cells comprising hamartomas. HF indicates an anagen hair follicle matrix with robust Ki67 immunostaining. (C) MYC immunostaining to detect GLI2ΔN confirms low-level expression in GLI2ΔN-low mice. Original magnification, ×200 (A–C).