Journal Logo

Letters to the Editor

Reevaluation of GLI1 Expression in Skin Tumors

Kervarrec, Thibault MD, PhD*,†,‡; Berthon, Patricia PhD; Thanguturi, Soumanth*; Guyétant, Serge MD, PhD*,†; Macagno, Nicolas MD, PhD‡,§; Jullie, Marie-Laure MD, PhD‡,¶

Author Information
The American Journal of Dermatopathology: October 2021 - Volume 43 - Issue 10 - p 759-761
doi: 10.1097/DAD.0000000000001917
  • Free

To the Editor:

During mammals embryonic life, the Sonic Hedgehog (SHH) pathway is involved in skin morphogenesis and notably contributes to the development of hair follicles.1 In 1996, identification of recurrent germline mutations of the PATCH (PTCH) gene in patients with nevoid basal cell carcinoma, so-called Gorlin syndrome (OMIM # 109400) provided the first indication of a potential SHH pathway contribution to basal cell carcinoma (BCC) oncogenesis.2–4 Indeed bi-allelic inactivation of PTCH in such patients leads to the activation of the SHH pathway, and expression of its downstream targets notably “GLI family zinc finger 1” GLI1,5,6 finally resulting in cancer susceptibility and multiple BCC formation.2,3

Accordingly, recurrent mutations in PTCH or in other members of the SHH pathway (namely “smoothened, frizzled class receptor » SMO and “SUFU negative regulator of hedgehog signaling » SUFU) were further observed in up to 85% of sporadic BCC cases together with inactivating mutations of TP53.7

Thus, because ectopic overexpression of GLI1 in the epidermis of transgenic mice additionally leads to BCC and follicular tumors formations,8,9 SHH activation is expected to be a main contributor to BCC development in human. These data also suggested that the constitutive activation of SHH pathway and GLI1 overexpression represented a potential diagnostic marker of BCC, thus requiring further validation of its diagnostic performance in daily practice.10

In this context, we wanted to congratulate Dr Gradecki et al11 who addressed this issue in their recent paper published in the American Journal of Dermatopathology.

Indeed, their evaluation of the cytoplasmic expression of GLI1 by immunohistochemistry in a cohort of 48 BCC in comparison to 88 other skin epithelial tumors (all potential BCC mimickers) provided important data concerning its diagnostic performance. Indeed, it revealed that the cytoplasmic GLI1 expression carried poor specificity to distinguish BCC from other entities. Albeit a diffuse expression of GLI1 was present in all BCC cases, most of the mimickers investigated also displayed GLI1 cytoplasmic expression except Merkel cell carcinoma.

Importantly, their work focused on GLI1 expression within the cytoplasm. Because GLI1 acts as a transcription factor, the active form of GLI1 protein is expected to be located in the nucleus.12,13 Indeed, it has been previously reported that inactivating mutations of PTCH or SUFU genes and activating mutations of SMO in BCC resulted in GLI1 nuclear relocation and overexpression,14 thus suggesting to assess the nuclear expression of GLI1. To further confirm these findings, we first ectopically expressed the GLI1 protein in normal human epidermal keratinocytes using a lentiviral transduction as previously described.15 Anti-GLI1 immunohistochemistry (clone C68H3, Ozyme) performed on the GLI1-transducted keratinocytes confirmed its diffuse expression within the cytoplasm, whereas its expression in the nucleus was more heterogeneous (Fig. 1). Nuclear GLI1 expression was subsequently confirmed in vivo on normal human skin FFPE samples using a previously described protocol15: most of the epithelium was negative or showed only a weak and random expression (score 1) in the epidermis, whereas the germinofollicular portion of the hair follicles displayed increased and diffuse nuclear expression of GLI1 (Fig. 1).

Expression of GLI1 in keratinocytes and normal skin by immunohistochemistry. A, Transduction of GLI1 in human keratinocytes induced both nuclear and cytoplasmic expression of the protein. B, Expression of GLI1 in normal skin: although lack of/weak staining was observed in the nuclei of a subset of keratinocytes in the interfollicular epidermis, nuclear expression of GLI1 was increased in the germinofollicular portion of the hair apparatus (*).

Because the use of different antibodies and protocols may result in different immunohistochemical results, we further evaluated the nuclear expression of GLI1 by immunohistochemistry in a large cohort of epithelial cutaneous tumors embedded in tissue microarrays. Only cases with interpretable staining were included in this study (n = 176). Results are available in Table 1/Figure 2. Overall, diffuse, nuclear GLI1 expression (score 2) was observed in all BCC specimens (n = 25/25), in 71% of the trichoblastoma specimens (n = 5/7), and in 11% of the squamous cell carcinoma cases (n = 8/69). Moreover, weak or heterogeneous positivity (score 1) was often present in the desmoplastic trichoepitheliomas (60%, n = 3/5), in a minority of SCCs (11%, n = 8/69), in one case of Paget Disease (20%, n = 1/5), and in one case of adnexal adenocarcinoma not otherwise specified (50%, n = 1/2). All remaining specimens were negative (n = 128).

TABLE 1. - Nuclear Expression of GLI1 in Skin Epithelial Tumors
GLI1 Expression Score Score 0, n (%) Score 1, n (%) Score 2, n (%)
BCC 0 0 25 (100)
 Nodular 0 0 17 (100)
 Superficial 0 0 4 (100)
 Infiltrating 0 0 4 (100)
Squamous cell carcinoma 53 (77) 8 (11) 8 (11)
 In situ SCC/Bowen disease 10 (50) 5 (25) 5 (25)
 Well differentiated 16 (100) 0 0
 Moderately differentiated 11 (92) 0 1 (8)
 Poorly differentiated 16 (76) 3 (14) 2 (10)
Tumors with follicular differentiation 15 (60) 5 (20) 5 (20)
 Trichoblastoma 0 2 (29) 5 (71)
 Desmoplastic trichoepithelioma 2 (40) 3 (60) 0
 Trichilemmoma 4 (100) 0 0
 Pilomatricoma 6 (100) 0 0
 Proliferating trichilemmal tumour 2 (100) 0 0
 Pilomatrical carcinoma 1 (100) 0 0
Tumor with sebaceous differentiation 8 (100) 0 0
 Sebaceous adenoma 6 (100) 0 0
 Sebaceous carcinoma 2 (100) 0 0
Tumors with eccrine/apocrine differentiation 47 (96) 2 (4) 0
 Poroma 7 (100) 0 0
 Spiradenoma 12 (100) 0 0
 Cylindroma 2 (100) 0 0
 Hidradenoma 7 (100) 0 0
 Poroid hidradenoma 3 (100) 0 0
 Paget's disease 4 (80) 1 (20) 0
 Myoepithelioma 3 (100) 0 0
 Chondroid syringoma 2 (100) 0 0
 Porocarcinoma 1 (100) 0 0
 Primary cutaneous mucinous carcinoma 2 (100) 0 0
 Cystic adenoid carcinoma 2 (100) 0 0
 Adnexal adenocarcinoma NOS 1 (50) 1 (50) 0
 Ductal squamoid carcinoma 1 (100) 0 0
Gli1 expression was scored as absent (score 0), weak or heterogeneous (score 1) and diffuse/intense (score 2). Results are expressed in number and percentages.
NOS, Not otherwise specified.

Representative illustration of GLI1 immunohistochemical detection in skin tumors. An invasive BCC, a desmoplastic trichoepithelioma, an invasive cutaneous squamous cell carcinoma, and a cylindroma are depicted. Intense and diffuse nuclear expression of GLI1 was observed in BCC tumor cells, whereas weaker GLI1 expression was detected in desmoplastic trichoepithelioma. GLI1 expression was absent from most of the squamous cell carcinoma specimens and sweat glands tumors.

Therefore, in line with previous investigations of GLI1 transcription levels in skin tumors,10 our results suggest GLI1 immunohistochemical detection may represent a relevant ancillary technique for the diagnosis of BCC in current practice (Sensitivity: 100%; Specificity: 73%). Accordingly, and if further studies confirm our results, negativity of GLI1 could help to rule out BCC diagnosis, whereas GLI1 diffuse nuclear expression would allow the distinction between BCC and most of the adnexal tumors with eccrine/apocrine differentiation. To note, our identification of a nuclear GLI1 positivity in some squamous cell carcinoma specimens may be in accordance with previously reported PTCD mutations in a minority of these tumours.16,17

Finally, and in accordance with the literature,10,15 we have confirmed an activation of the SHH pathway in BCC and in trichoblastoma/trichoepithelioma, although PTCH mutations are rarely detected in the latter.18

Our study harbors some limitations. In particular, the tissue micro-array approach is a source of sampling bias. However, the presence of a clonal alteration that is, a mutation activating the SHH pathway in virtually all of the BCC tumor cells probably limits such bias. This is confirmed by the diffuse nuclear expression of GLI1 in samples observed on whole slides.

In conclusion, we are thankful to Dr Gradecki and colleagues for previous exchanges and for providing us the opportunity to compare our results. Although standards of GLI1 immunohistochemical detection remain to be determined, we believe that both of our studies suggest a potential interest of GLI1 detection for the diagnosis of BCC.


1. Oro AE, Higgins K. Hair cycle regulation of Hedgehog signal reception. Dev Biol. 2003;255:238–248.
2. Hahn H, Wicking C, Zaphiropoulous PG, et al. Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome. Cell. 1996;85:841–851.
3. Unden AB, Holmberg E, Lundh-Rozell B, et al. Mutations in the human homologue of Drosophila patched (PTCH) in basal cell carcinomas and the Gorlin syndrome: different in vivo mechanisms of PTCH inactivation. Cancer Res. 1996;56:4562–4565.
4. Levanat S, Gorlin RJ, Fallet S, et al. A two-hit model for developmental defects in Gorlin syndrome. Nat Genet. 1996;12:85–87.
5. Dahmane N, Lee J, Robins P, et al. Activation of the transcription factor Gli1 and the Sonic hedgehog signalling pathway in skin tumours. Nature. 1997;389:876–881.
6. Green J, Leigh IM, Poulsom R, et al. Basal cell carcinoma development is associated with induction of the expression of the transcription factor Gli-1. Br J Dermatol. 1998;139:911–915.
7. Bonilla X, Parmentier L, King B, et al. Genomic analysis identifies new drivers and progression pathways in skin basal cell carcinoma. Nat Genet. 2016;48:398–406.
8. Aszterbaum M, Epstein J, Oro A, et al. Ultraviolet and ionizing radiation enhance the growth of BCCs and trichoblastomas in patched heterozygous knockout mice. Nat Med. 1999;5:1285–1291.
9. Nilsson M, Undèn AB, Krause D, et al. Induction of basal cell carcinomas and trichoepitheliomas in mice overexpressing GLI-1. Proc Natl Acad Sci U S A. 2000;97:3438–3443.
10. Hatta N, Hirano T, Kimura T, et al. Molecular diagnosis of basal cell carcinoma and other basaloid cell neoplasms of the skin by the quantification of Gli1 transcript levels. J Cutan Pathol. 2005;32:131–136.
11. Gradecki SE, Eid MV, Pramoonjago P, et al. Glioma-associated oncogene-1 expression in basal cell carcinoma and its histologic mimics. Am J Dermatopathol. 2020 [epub ahead of print].
12. Carpenter RL, Lo H-W. Hedgehog pathway and GLI1 isoforms in human cancer. Discov Med. 2012;13:105–113.
13. Kogerman P, Grimm T, Kogerman L, et al. Mammalian suppressor-of-fused modulates nuclear-cytoplasmic shuttling of Gli-1. Nat Cel Biol. 1999;1:312–319.
14. Skoda AM, Simovic D, Karin V, et al. The role of the Hedgehog signaling pathway in cancer: a comprehensive review. Bosn J Basic Med Sci. 2018;18:8–20.
15. Kervarrec T, Samimi M, Hesbacher S, et al. Merkel cell polyomavirus T antigens induce Merkel cell-like differentiation in GLI1-expressing epithelial cells. Cancers. 2020;12:1989.
16. Ping XL, Ratner D, Zhang H, et al. PTCH mutations in squamous cell carcinoma of the skin. J Invest Dermatol. 2001;116:614–616.
17. Tarapore E, Atwood SX. Defining the genetics of basosquamous carcinoma. J Invest Dermatol. 2019;139:2258–2260.
18. Hafner C, Houben R, Baeurle A, et al. Activation of the PI3K/AKT pathway in Merkel cell carcinoma. PLoS One. 2012;7:e31255.
Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.