SHH signaling transfers signals from the extracellular environment and activates the expression of genes involved in cell survival and proliferation (28). A schematic presentation of the pathway is shown in Fig. 1A and B, and the core elements of the pathway are briefly presented in Table I.

SHH signaling is triggered by the cell membrane binding of the functional SHH glycoprotein. It acts as a classic morphogen during embryonic development, where it is involved in the crucial phases, such as patterning of the ventral neural tube, the anterior-posterior limb axis and ventral somites (20). Germinal mutations of the SHH gene, located at 7q36.3, lead to congenital defects, such as holoprosencephaly (31-33). Recent research on genomic DNA of patients affected by holoprosencephaly has revealed that eight synonymous single-nucleotide variants in the SHH gene are associated with a reduced level of SHH protein (34). A recent in vivo study on Cre-modified mice demonstrated that SHH expression was also crucial for proper fetal development of the tongue and mandible (35). SHH is the most well-known among other hedgehog family proteins, comprising Desert Hedgehog (DHH) and Indian hedgehog (IHH) molecules (36). Although all hedgehog family members can bind to the PTCH1 receptor, their tissue distribution and roles are different (20,37). It has been proven that SHH protein plays a significant role in central nervous system development (38). The activity of IHH in skeletal tissue formation has been reported, whereas DHH is present only in granulosa cells of ovaries and Sertoli cells of the testis (20,32). Post-translational modifications of all three hedgehog protein family members are required for their attachment to the PTCH1 receptor (15). During this molecular process, the full-length SHH protein (~45 kDa) undergoes autoproteolysis and cleavage into the C- (C-SHH; ~25 kDa) and N- (N-SHH; ~19 kDa) terminal domains (39,40). C-SHH is an auto-processing molecule that participates in the attachment of cholesterol to the C-terminal end of N-SHH. Furthermore, the N-terminal end of N-SHH binds to palmitic acid moiety through the reaction induced by hedgehog acyltransferase (HHAT), which is necessary for its full biological activity (41,42). The activity of HHAT may be blocked by the use of RU-SKI inhibitors (RU-SKI 41, 43, 101 and 201; not shown in the figures) (43); however, overall cytotoxicity was observed for RU-SKI 41, 43 and 101 in an in vitro study (44). Currently, there are no data available regarding the use of RU-SKI inhibitors in clinical studies, at least to the best of our knowledge. Finally, through its interaction with dispatched resistance-nodulation-division (RND) transporter family member 1, modified N-SHH is secreted to the extracellular matrix (ECM) and may act as a biologically active upstream regulator of the SHH pathway (Fig. 1B) (15,45,46). Therefore, the binding of N-SHH may present another target in cancer drug studies. For that reason, 5E1 antibody against N-SHH (Fig. 1B) was analyzed in a mouse model of pancreatic cancer, and was found to have a promising effect in the reduction of tumor size and angiogenesis (47). The final interaction between N-SHH ligand and PTCH1 may occur in either an auto- or paracrine way (40), which will be discussed below in the present review.

The PTCH1 receptor, encoded by the PTCH1 gene at 9q22.32, is composed of 1,447 amino acids, arranged in 12 transmembrane helices, two extracellular domains (1 and 2) that can attach extracellular ligand N-SHH and one cytoplasmic carboxyl-terminal domain (48). Mutations in the PTCH1 gene lead to an autosomal dominant, multisystem disorder known as Nevoid basal cell carcinoma syndrome, also known as the Gorlin-Goltz syndrome (49). The patched protein family also includes PTCH2 receptor (50). Although both PTCH1 and PTCH2 can bind to Hedgehog ligands with the same affinity, PTCH2 appears to have a lower ability to inhibit the SMO protein (20,45,51). PTCH1 acts as a negative regulator of the SHH pathway by inhibiting SMO protein from translocating to the plasma membrane (Fig. 1A). The mechanism of this inhibition is not yet fully understood; however, recent studies have suggested the involvement of cholesterol or another sterol lipid in this regulation (52,53). Following SHH pathway activation, the blockade of PTCH1 is abolished and the receptor undergoes internalization, while SMO protein is exposed on the cell surface in the primary cilium (54). PTCH1 subsequently undergoes endocytosis, followed by ubiquitination and lysosomal degradation through E3 ubiquitin ligase SMAD specific E3 ubiquitin protein ligase 1/2 (Smurf1-2; Fig. 1B).

Smoothened protein belongs to the F-class of the G-protein-coupled receptor superfamily and is a key intracellular positive SHH pathway regulator (55,56). Research on SHH signaling in D. melanogaster has indicated that biochemical processes, such as phosphorylation or sumoylation, are required to obtain full SMO activity (57). In mammalian cells, ciliary localization of this molecule appears to be crucial for SMO activation, as well as post-translational SMO modifications, which are analogous to those in Drosophila cells (58). The ciliary translocation of SMO occurs following phosphorylation by a β-adrenergic-receptor kinase (G protein-coupled receptor kinase 2) and is followed by an interaction between cytosolic β-arrestin (ARBB) and clathrins (59). Following ciliary translocation, SMO is further phosphorylated by casein kinase 1α, and then SMO-β-arrestin complex recruits motor protein kinesin family member 3A (Kif3A), which consequently interacts with the kinesin family member 7 motor protein (KIF7)-SUFU-GLI2/3 ciliary located complex (Fig. 1B) (60). The SMO protein is the main anticancer treatment target among all SHH pathway proteins (61). Several SMO inhibitors are in clinical trials, and three of them (vismodegib, sonidegib and, in November 2018, glasdegib) have been approved for selected cancer treatment by the US Food and Drug Administration (FDA) (61,62). However, this molecular treatment has certain limitations, including the development of drug resistance due to frequent SMO mutations, as well as the presence of alternative SMO-independent mechanisms of GLI transcription factor activation (63,64), which are discussed in the following paragraphs.

Cubitus interruptus has been identified as a transcription factor of the Hedgehog pathway in D. melanogaster (41,65). In mammals, this protein has three analogs (the GLI1, GLI2 and GLI3 molecules), which belong to the Kruppel zinc-finger transcription factor family (66). The lack of a repressor domain in GLI1 protein structure suggests that this molecule may act only as a transcription activator, while GLI2 and GLI3 possess both repressive (GLI2/3^R; Fig. 1A) and activating (GLI2/3^A; Fig. 1B) properties (20). Furthermore, several isoforms of GLI1 and GLI2 (called GLIΔN or tGLI), which are the products of alternative splicing, have been identified in human tissues (65). tGLI1 has been detected only in cancer samples and been associated with aggressive behavior of the disease (67-69). In the absence of SHH, GLI2/3 are attached to the SUFU molecule in the ciliary location by KIF7 (Fig. 1A). GLI2/3 underwent phosphorylation by glycogen synthase kinase (GSK)-3β, protein kinase A (PKA) and CK1, which is triggered by cyclic AMP produced by G protein-coupled receptor 161 (Gpr161). Such action causes proteolytic cleavage of GLIs' C terminus by cullin 1 (CUL1) and β transducin repeat-containing protein (β-TrCP), leading to the removal of their transcriptional activation domain. Cleaved GLI2/3, in the form of GLI2^R and GLI3^R, translocates to the nucleus and act as inhibitors/repressors after binding to regulatory regions of SHH target genes (21,70).

Following the activation of the canonical SHH pathway, the SMO-β-arrestin complex inhibits Gpr161 and cyclic adenosine monophosphate-dependent PKA (Fig. 1B) (71), which blocks the phosphorylation and proteolytic cleavage of GLI2/3. Subsequently, the GLI2/3 KIF7/SUFU/GLI2/3 complex dissociates, and full-length GLIs undergo several posttranslational modifications, including phosphorylation, ubiquitination and sumoylation (70), and may simultaneously undergo proteolysis mediated by CUL3 and speckle-type POZ protein (72). The activity of GLI2^A and GLI3^A transcription factors may then be upregulated by various cytoplasmic factors on their way to the nucleus. There are several protein kinases [casein kinase II (CK2), protein kinase B (AKT), extracellular signal-regulated kinase 1/2 (ERK1/2), ribosomal protein S6 kinase 1 (S6K1), dual specificity tyrosine phosphorylation regulated kinase 1B (DYRK1B) or unc-51 like kinase 3 (ULK3)] (73), which phosphorylate GLI2/3^A (Fig. 1B), thus promoting GLI translocation into the nucleus. Acetylation/deacetylation of GLIs is another important factor that regulates their transcriptional activity (70). Acetylation of GLI1/2 by p300/CBP complex prevents GLIs from attaching to DNA and provides nuclear export through exportin 1 and LAP2 proteins (70,74). On the contrary, GLI deacetylation by histone deacetylase HDAC1 enables them to interact with genomic DNA (75). Of note, HDAC1 is upregulated by GLIs; therefore, the HDAC1-GLIs interaction forms a positive feedback loop with the SHH pathway (70,75). A significant role of primary cilium in the functioning of the GLI proteins has been recently reported. In the absence of the SHH ligand, GLI2/3-SUFU complexes are transported to the tip of the cilium by kinesin through the microtubule cytoskeleton, while GLI2 translocation to the cell nucleus following ligand stimulation occurs through dynein-2 (76). The in vivo studies by Wong et al (77) and Han et al (78) revealed that the removal of the Kif3a allele, which is essential for cilia formation, leads to the inhibition of both BCC and medulloblastoma, respectively. However, this effect was observed only in lesions overexpressing the SMO gene, but not the constitutively active GLI2 gene. Therefore, the primary cilium components could be a molecular target for SMO-dependent neoplasms (77,78).

Due to frequent mutations in the SMO receptor, which lead to cancer resistance to previously mentioned SMO-targeted drugs (64), blocking cytoplasmic/nuclear GLI-activator proteins is one of the recently identified targets (30,79). It has been found that CK2, DYRK1B and S6K1 protein kinases, as well as HDAC1, do not require SMO-dependent activation of the SHH pathway to activate GLIs (64). This observation provides reasoning for examining the activity of several potential drugs, including CIGB-300 and CX-4945 targeting CK2 (80), CCI-779 and RAD001 targeting S6K1 (81), BVD-523 targeting ERK1/2 (82), MK2206 targeting AKT (83), AZ191 inhibiting DYRK1B (84) and SU6668 targeting ULK3 (85). It has also been reported that HDAC1 deacetylation activity is successfully blocked by 4SC-202 (17), with GLI-DNA-interaction inhibitors (glabrescione B and GANT61) as well as GLI2 destabilizers (arsenic trioxide and pirfenidone) (Fig. 1B) (17). It is worth noting that the GLI proteins are also involved in other cancer-related pathways, which are discussed below in the present review.

Kidney cancers, otherwise known as renal cell cancers (RCCs), are a group of histologically different tumors (132), which rank 14th in incidence among other neoplasms worldwide (8). Clear cell RCC (ccRCC) is the most common subtype (6) and is associated with unfavorable outcomes (133). ccRCC development is strongly associated with the inactivation of the von Hippel-Lindau tumor suppressor (VHL) gene, which can be hereditary (VHL syndrome) or occurs spontaneously during life (10,134-136). Other alterations in genes such as PBRM1 or mTOR have been identified; however, no specific prognostic or predictive molecular markers of RCC can be recommended for clinical use (6,10). RCC therapy includes surgical and pharmacological treatment in the advanced stages of the disease, including tyrosine kinase inhibitors (TKIs) with sunitinib, the first such drugs to be introduced, and mTOR kinase inhibitors (everolimus) and several others, introduced into clinical treatment over the past decade (6,137).

The first report regarding the expression of SHH pathway components in ccRCC was published in 2009 (138). Dormoy et al (138) found that the SHH signaling genes were expressed at the mRNA level in various RCC cell lines, independently of VHL gene status. In that study the overexpression of SMO and GLI1 mRNAs was also revealed by RT-qPCR in RCC tumor tissues, compared with corresponding normal kidney samples in the group of 8 patients. Furthermore, incubation with cyclopamine (SMO inhibitor) decreased ccRCC cell proliferation and increased apoptosis, as well as induced the regression of ccRCC tumors in nude mice (138).

Further studies conducted on human RCC samples have demonstrated the association between SHH signaling and cancer progression. A study on 140 ccRCC specimens derived from patients with non-metastatic disease revealed a significantly elevated DHH, SHH, PTCH1 and GLI3 protein immunoreactivity in samples assessed as G3 or G4 in Fuhrman's grading system [grades 3 and 4 in International Society of Urologic Pathologists (ISUP) grading (139)] than in those with grade G1 or G2 (ISUP grades 1 and 2) (140). An elevated immunoreactivity of the GLI2 transcription factor was found to be associated with a poor prognosis in a group of 39 patients with metastatic ccRCC treated with sunitinib (141). In addition, in vitro experiments revealed a decrease in the GLI2 protein level by western blot analysis in ACHN cells treated with sunitinib, but not in sunitinib-resistant ACHN cells. Therefore, these results suggested that GLI2 protein may be involved in the mechanism of drug resistance associated with TKI inhibitors in RCC (141). Behnsawy et al (142) demonstrated an association between the activity of SHH signaling and EMT, an important step of cancer progression, in RCC cell lines. The recombinant SHH ligand (r-SHH) not only significantly increased proliferation in RenCa and ACHN cells, but also reduced the mRNA level of E-cadherin, the epithelial marker of EMT, suggesting a stimulating role of the SHH pathway in EMT (142).

Since several studies have reported a SHH gene upregulation in RCC (140,142,143), the present review focused on the occurrence of the upstream protein and miRNA regulators of SHH expression in RCC. Shang et al (144) analyzed the mRNA expression rates of the ZIC2 gene in 533 ccRCC and 72 normal kidney samples (TCGA database), and found that the overexpression of ZIC2 mRNA was associated with age, TNM, histological grade and a shorter overall survival; thus, this gene can therefore be used as an independent prognostic factor in ccRCC. Jia et al (145) also analyzed TCGA data from 525 patients with ccRCC focusing on SHH-associated FOX family genes, and it was found that FOXA2 mRNA overexpression was associated with poor outcomes.

As previously mentioned, RCC initiation is strongly associated with the VHL gene status, which is inactivated in a broad range of ccRCC cases. The gene encoding VHL protein, which acts as an E3 ubiquitin ligase, is the enzyme responsible for hypoxia-inducible factor (HIF1α and HIF2α) degradation under normoxic conditions (136). Therefore, Zhou et al (146) investigated the expression of SHH pathway genes in normoxia and hypoxia, as well as the association between the SHH signaling components and HIF2α. The mRNA expression of all SHH signaling genes was significantly elevated in RCC cell lines that were cultured under hypoxia, compared with normoxic control RCC cells. Of note, the re-activation of the SHH pathway under hypoxic conditions was independent of VHL expression, with the dual inhibition of HIF2α and GLI1 activity. Furthermore, the treatment with sh-HIF2a and GLI1 inhibitor GANT61 significantly sensitized RCC cells to ionic radiation. These results demonstrated that the SHH pathway together with HIF2α protein may be involved in the molecular mechanisms of RCC radioresistance. In addition, the SMO inhibitor, cyclopamine, was not found to reduce the observed overexpression of GLI1 under hypoxic conditions, which suggested that GLI1 expression in RCC cells does not depend on upstream SHH signaling components, but could be induced by different molecular signaling (non-canonical activation) (146). Further evidence provided by Zhou et al (147) confirmed this conclusion and demonstrated an involvement of the PI3K/AKT cascade on the main effectors of the SHH pathway in RCC cells. PI3K/AKT signaling stimulation or inhibition induced or decreased the expression of GLI1 and GLI2, respectively. It was also demonstrated in vitro and in vivo that the combination of GANT61 with the AKT specific inhibitor, perifosine, was associated with a significantly enhanced therapeutic potential, compared with that of the use of each substance alone (147).

The efficacy of several other SHH inhibitors on kidney cancer treatment has been under investigation over the past few years. Erismodegib, a SMO antagonist, was previously shown to inhibit the survival of the human 786-O RCC line, either alone or, more effectively, in combination with sunitinib and everolimus (148). This antitumor effect was also observed in sunitinib-resistant RCC cells (786-O SuR cells), revealing a novel research direction for RCC therapy. It was also observed that erismodegib combined with sunitinib or everolimus decreased the tumor volume and increased the survival of nude mice with 786-O SuR cell-derived tumor xenografts, confirming previously described results. However, unlike erismodegib, GANT61 had no inhibitory effect on RCC cells (148), indicating that SMO is a more promising selective RCC therapy target than GLI transcription factors.

In a previous study by the authors (143), the expression of SHH pathway components in 37 ccRCC tissue samples, a significant correlation was identified among the expression of almost all SHH signaling genes at the mRNA level. Although the mRNA level of SHH, SMO and GLI1 was increased in ccRCC samples, compared to the morphologically unaltered kidney tissues, no association was observed between the expression rates of genes and the pathological features of patients. However, at the protein level, western blot analysis of SHH revealed a significant increase of full-length SHH and a decrease of the C-SHH domain in ccRCC tissues (143). This novel observation may suggest an involvement of the SHH ligand in ccRCC development, and indicate changes in the post-translational modification of this protein during tumor progression.

According to the GLOBOCAN database, there were 549,393 new cases of bladder cancer in 2018, which renders this type of cancer as the 11th most common type of cancer worldwide (149). Approximately 90% of bladder carcinomas are derived from the transitional epithelium (1). Several studies have proven the significance of nicotine and industrial gases in the pathogenesis of this type of cancer (150,151). The disease risk assessment is performed using clinical patient examination, medical imaging and microscopic examination of the resected tumor tissues. The bladder cancer guidelines recommend the tumor-node-metastasis (TNM) system as an appropriate classification system for tumor staging. The treatment of bladder cancer includes transurethral resection of the bladder tumor for initial bladder neoplasms; however, for more advanced tumors, radical cystectomy with lymphadenectomy and additional radio- or chemotherapy are required (1). Genetic and epigenetic alterations of bladder cancer cells, which may be useful prognostic factors or targets for personalized therapy, are under investigation. The molecular profile of non-muscle invasive bladder cancer (NMIBC) differs significantly from muscle invasive bladder cancer (MIBC). In addition, genetic alterations characteristic of low-grade NMIBC, such as fibroblast growth factor receptor 3 (FGFR3) or RAS mutations, can be distinguished among NMIBCs. FGFR1, FGFR3, PNEN, CCND1 or MDM2 proto-oncogene genes have been identified as potential therapeutic targets, whereas TSC complex subunit 1 or phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA) mutations may be predictive targets for mTOR or PIK3CA/mTOR inhibitors, respectively (1,152).

In 2012, He et al (153) performed immunohistochemistry (IHC) on 118 human bladder cancer samples. The expression of proteins encoded by the SHH, PTCH1 and GLI1 genes was significantly elevated in tumor tissues, compared to 30 adjacent normal bladder tissues. The increased immunore-activity of SHH pathway proteins was observed in samples derived from patients with a high pathological stage, the presence of venous invasion and lymph node metastasis. Patients with a positive SHH, PTCH1 and GLI1 expression also exhibited poorer disease-free survival rates, according to Kaplan-Meier analysis (153). Further studies suggested the prognostic value of the SHH pathway protein level in bladder cancer. Nedjadi et al (154) revealed that a high SHH protein immunoreactivity in urothelial bladder cancer tissues was associated with the presence of lymph node metastasis; however, no association was identified between SHH expression and other clinicopathological parameters or patient survival. SHH overexpression can be associated with the upregulation of MEIS2 (an upstream SHH gene regulator) in bladder cancer lymph node metastasis, as observed by Xie et al (155) in a clinical study on 104 patients with bladder cancer.

The significant role of SHH pathway proteins was observed in the EMT of bladder cancer cells. An HTB-9 transitional bladder cancer cell line, with acquired mesenchymal features due to TGFβ1 stimulation (T-HTB-9), exhibited an overexpression of the SHH and GLI2 genes at the mRNA and protein levels. Furthermore, following incubation with cyclopamine, and GDC-0449, SMO and GLI1-3 inhibitors, a decrease in the migration, invasion and clonogenicity of T-HTB-9 cells was observed (156). This evidence suggested that inhibitors of the SHH pathway may effectively decrease bladder cancer invasive potential and may thus prove to be useful to bladder cancer treatment. Islam et al (156) examined 22 specimens derived from human bladder cancer. An elevated immunoreactivity of SHH, GLI2, Ki-67 proliferation marker and N-cadherin (mesenchymal cell marker), and a decrease in E-cadherin (epithelial cell marker), were observed in high-grade tumors compared with low-grade tumors, further confirming the participation of SHH signaling proteins in the EMT of human bladder cancer cells (156). Another analysis concerning the association between EMT and the SHH pathway was performed on muscle-invasive T24 and 5637 bladder cancer and non-muscle-invasive KK47 cell lines. The incubation of the cells with recombinant SHH protein decreased the expression of E-cadherin and enhanced that of N-cadherin and vimentin in all three cell lines. Cyclopamine was found to inhibit cell proliferation and invasiveness; however, the effect was more pronounced in T24 and 5637 cell lines. In vivo studies on nude mice with induced bladder cancer revealed a significant inhibition of muscle-invasive-derived tumor development, which indicated the potential benefits of using SHH pathway-targeted therapy in advanced stages of bladder cancers (157). Of note, Kim et al (158) found that the CpG hypermethylation-induced decrease in SHH gene expression in bladder cancer cells led to an increase in tumor invasiveness. The lack of SHH ligand decreased the activity of SHH signaling in stromal cells, inhibiting the expression of bone morphogenetic proteins and ultimately stimulating bladder cancer progression (158). Furthermore, the pharmacological inhibition of DNA methylation inhibited the initiation of invasive urothelial carcinoma at the premalignant stage of progression, through the increase in SHH expression in cancer cells (158). These findings were not consistent with previously presented results; thus, further research on the cell-to-cell interactions between bladder cancer and stromal cells in bladder tumors would improve the understanding of the molecular basis of the role of the SHH pathway in bladder cancer (158).

OC ranks 8th in incidence and mortality among cancers affecting women (18th and 14th in both sexes, respectively) worldwide, with almost 300,000 cases and 185,000 deaths in 2018, according to the GLOBOCAN data (8). Epithelial OC accounts for >90% of all ovarian malignancies and is classified into five histological subtypes: Serous, mucinous, endometrioid, undifferentiated and clear cell subtypes (162), while OC advancement is based on the International Federation of Gynecology and Obstetrics (FIGO) staging. Molecular patterns of SHH upstream regulators in OC have only been analyzed by a few studies.

One of the first studies on SHH pathway components in OC was conducted by Levanat et al in 2004 (163). Although an upregulation of GL1 mRNA expression was not observed in a group of 11 ovarian fibromas and 15 ovarian dermoids, higher mRNA levels of SMO and SHH were observed. A frequent mutation of the PTCH1 gene was also identified in the majority of ovarian fibromas, but it was not found to be associated with the expression level of this gene (163). Marchini et al (164) observed the overexpression of ZIC2 in the malignant form of epithelial OC (n=193), compared to low-malignant potential OCs (n=39). In OC cell lines, ZIC2 overexpression was found to increase the growth rate and foci formation of NIH3T3 cells and stimulate anchorage-independent colony formation (164). The data on FOXA2 expression in OC are inconclusive: Salem et al (165) found that its lower mRNA levels promoted OC tumorigenesis, while Peng et al (166) reported high FOXA2 levels in OCs. Loss of heterozygosity of the PTCH1 gene was a frequent observation in OC (167,168), suggesting that the mechanism of SHH pathway activation in OC is type I. Moreover, the studies regarding somatic mutations in SHH signaling components counted 14% frequency in a MyPathway study (169).

Further studies identified the association between SHH signaling and OC progression; Liao et al (170) observed the overexpression of SHH and patched proteins (assessed by IHC in 80 patients with OC) and GLI1 mRNA (quantified by qPCR in 37 OCs) in tumor specimens, whereas no changes were observed in ovarian tissue. In addition, the observed molecular alterations were associated with the poorer outcome of OC patients. Liao et al (170) also performed a GLI1 ectopic expression experiment on SKOV3 and OVCAR3 OC cell lines and reported the upregulation of tumorigenesis-related genes (i.e., BCL2, VEGF and genes encoding vimentin and E-cadherin). The incubation of SKOV3, OVCAR3 and OVCA433 cells with KAAD-cyclopamine, an inhibitor of SMO protein, suppressed cancer cell viability, induced apoptosis, and decreased the expression of the aforementioned cancer-related genes (170). However, contrasting results were obtained by Yang et al (171), who did not report higher levels of SHH pathway components nor the target genes in 34 OC tumor samples. Based on the SHH, PTCH1, GLI1, HHIP, SMO and SUFU mRNA semi-quantification results (assessed by PCR and qPCR for GLI1), as well as patched 1, GLI1 and HHIP proteins (assessed by IHC), the results of that study suggested infrequent involvement of the SHH pathway in OC development (171). In a study by Schmid et al (172), inconclusive results of the expression of SHH signaling and target genes in OC were obtained. In a group of 16 FIGO stage III serous tumors, various expression levels of SHH genes (GLI1/2, PTCH1, SHH and SMO; assessed by qPCR) were observed, while IHH and PTCH2 genes were upregulated in the majority of cases (172).

More recent data have confirmed, however, the impact of the SHH pathway in the progression of OC; Ozretić et al (97) analyzed SHH pathway genes in 23 OCs, including 16 carcinomas (CA) and 7 atypical proliferative (borderline) tumors. However, higher mRNA levels of GLI1 and SUFU were observed in OCs, and SUFU levels were found to decrease with increasing FIGO stages. Moreover, a strong positive correlation was observed between the SMO and GLI1 mRNA levels. In the primary culture of tumor cells obtained from a high-grade ovarian tumor sample (FIGO IIIC), cyclopamine exerted an inhibitory effect on cell proliferation, but only in the first 24 h, whereas GANT61 decreased the proliferation rates of both primary and SKOV-3 cell lines after 72 h (97). Furthermore, GANT61, unlike cyclopamine, led to the downregulation of GLI2 transcription factor in the cells at the molecular level, rendering it a more effective SHH signaling inhibitor in OC treatment (97).

Recent studies have highlighted the importance of GLI-regulated anti-apoptotic protein survivin (BIRC5) (97,173,174) and matrix metalloproteinase (MMP)-7 (175) as putative markers for OC progression. Zhang et al (175) reported a high immunoreactivity of MMP-7 and GLI2 in tumor tissues from 95 OC patients, and the high expression of MMP-7 protein was found to be associated with poor patient outcomes. The association between the SHH pathway and MMP-7 expression was proven by demonstrating that ectopic stimulation of SHH in an SK-OV-3 OC cell line increased MMP-7 expression (175).

BIRC5 is an anti-apoptotic protein that acts as a negative regulatory protein that prevents apoptotic cell death; the gene is highly expressed during fetal development and in cancer tissues (176). Trnski et al (173) and Vlčková et al (174) analyzed the association between BIRC5 gene activation and the SHH pathway. The first team worked on A549 and the other experimented on SKOV-3 OC cell lines. Based on BIRC5 promoter inactivation by GANT61 rather than cyclopamine, Vlčková et al (174) proved that BIRC5 was regulated by the GLI2 transcription factor. Trnski et al (173) further revealed, by the addition of the GLI1 activator, that GLI3 was not associated with survivin expression.

Recently, the associations between the SHH pathway and CSC have been studied in high-grade serous OC (HGSOC) (177). Sneha et al (177) analyzed the effects of SHH pathway inhibitors on cell viability and spheroid formation through primary cultures of tumor cells from HGSOC and in nine OC cell lines. The treatment of cells with SHH inhibitors reduced the formation of spheroids with the higher efficacy of GANT61, compared with LDE225 (sonidegib) and salinomycin. In a xenograft model, the formation of tumors with an OVCAR3 origin was inhibited by GANT61 treatment. It was also found that the stemness marker, ALDH1A1, was at least partially dependent on the SHH pathway (177). The association between ALDH1A1 and the SHH pathway through the inhibition of GLIs was also observed in bladder (178) and breast (179) cancer. In conclusion, data have demonstrated that the SHH pathway plays an important role in OC development with GLI1/2 downstream effectors as the key points.

The worldwide incidence and mortality numbers of CC in 2018 were approximately 590,000 and 311,000, respectively, with CC ranking fourth in both categories among other malignancies (8). Although it is known that the pathogenesis and progression of CC are associated with human papillomavirus (HPV) infection, the involvement of the SHH pathway has also been described. The study by Rojo-León et al focused on the impact of HPV E6/E7 oncogenes on the SHH pathway in transgenic mice that carry eight GLI1-binding sites bound to the firefly luciferase gene (180). An increased GLI1 expression was observed in the cervix and skin either after exogenous estradiol or E6/E7 oncogene activation (180). Chen et al (181), using a microarray assay, found an increased expression of GLI1, SMO, SHH, PTCH1 and FOXM1 (GLI target gene) in 70 tumor CCs, compared to 10 normal cervical tissues; the expression patterns of those genes were associated with either the clinical or pathological progression of CC.

The majority of studies describing the role of the SHH pathway in CC have been performed using CC cell lines. Vishnoi et al (182) reported a connection between E6/E7 oncoproteins and SHH activation by analyzing HPV-16 positive SiHa CC cells. In SiHa cells, the SHH components, GLI, SMO and PTCH1, were found to be overexpressed, while their reduced expression was observed following either the addition of cyclopamine or siRNA-mediated E6 gene silencing (182). Wang et al (183) demonstrated that, in a hypoxic environment, the GLI1 mRNA level in HeLa cells was increased and was accompanied by an enhanced invasion ability, whereas GLI1 silencing reversed these effects, compromising the invasiveness of HeLa cells. Furthermore, Wang et al (183) observed that the ectopic increase of mir-129-5p resulted in the lower mRNA and protein levels of ZIC2, SHH, GLI1 and GLI2, together with SHH target genes CXCL1, VEGF and ANG2, as well as the inhibition of tumor formation in a mouse xenograft model. These results indicated that mir-129-5p may be a promising target for CC treatment (184). In combination, the available evidence suggested that the SHH pathway is involved in CC progression.

Prostate gland tumors rank 2nd in the worldwide cancer incidence among males (4th among all cancers in both sexes) with almost 1.3 million new cases, and 5th in worldwide cancer mortality in males (8th among all cancers in both sexes) with ~359,000 deaths in 2018 (8). The majority of PCs are associated with defective DNA damage repair molecules, while androgen receptor (AR) signaling also plays an important role in PC pathogenesis, particularly in metastasized cases (187). During fetal life, the AR and SHH pathways play a crucial role in the development of the prostate gland (188,189). Le et al (188) reported that, during prostate development, growth and regeneration, both pathways are indispensable; the AR signaling pathway is superior since, in the murine in vivo model, the expression of AR was essential for urogenital mesenchymal and epithelial cell differentiation, even if the cells overexpressed GLI1.

Yamamichi et al (190) reported that in PC epithelial cells (LNCap) and prostate fibroblast cell lines, normal (NPF) and PC-associated (CPF), dihydrotestosterone (DHT) enhanced cell proliferation in all cell types while the inhibition of SHH signaling by cyclopamine decreased this rate in CPF cells only. The activation of both androgen and SHH signaling enhanced EMT, accelerating PC development, while cyclopamine blocked cancer progression. In addition, DHT (but not SHH) induced the expression of osteonectin, and a high GLI1 expression and stromal osteonectin expression (as found by IHC) in tumor tissues from 25 patients with PC, were associated with PSA recurrence (190).

A recent study by Zhang et al (191) analyzed the AR and SHH pathways in PC clinical cases. In a large group of 443 patients with primary PC and 96 with benign prostatic hyperplasia, the increased immunoreactivity of SHH protein was observed in more aggressive tumors (Gleason score of >7), which was much higher in AR-positive than in AR-negative cancer. Furthermore, SHH was overexpressed in high-grade PC and positively correlated with the expression of both GRP78 (the molecule involved in endoplasmic reticulum stress response) and AR; this suggested that the assessment of SHH protein could be beneficial as a prognostic factor in PC, since SHH overexpression in all patients with PC with AR⁺ tumors was associated with a shorter disease-specific survival (191). Describing the expression pattern of SHH pathway components in PC, Tzelepi et al (192) analyzed SHH, SMO, PTCH, GLI1, VEGF, CD31 and ki67 protein levels using western blot analysis, IHC and tissue microarrays in large groups consisting of 141 hormone-naive primary PC and 53 castrate-resistant bone marrow metastases, compared to 119 prostate non-neoplastic peripheral zone. First, they observed the crosstalk between prostate cells in healthy tissues; SHH and PTCH1 were primarily expressed in epithelial and stromal cells, respectively, while SMO and GLI1 were expressed in both epithelial and stromal cells. This observation suggested paracrine signaling between epithelial (donor) and stromal (acceptor) cells, followed by SHH pathway activation in all cells (192). The expression pattern was continued in primary PCs with higher SHH and SMO protein levels in PC epithelial cells than those in the non-neoplastic peripheral prostate zone. Of note, in PC metastases, a higher PTCH1 expression was observed in epithelial cells compared with that in stromal cells, while the expression of SHH and GLI1 did not differ between the two (192). These results suggested an alteration in the mechanisms of SHH signaling in PC and its metastases, as well as its involvement in PC development.

In combination, the available data demonstrate that the SHH pathway plays an important role in PC development, indicating that SHH pathway-targeting drugs should be introduced into PC treatment. Indeed, two phase I and one phase II clinical trials that used LDE225, vismodegib or itraconazole (SMO inhibitors) have been performed (193-195). Although decreased levels of GLI1 were recorded in tumor tissues from patients treated with vismodegib or LDE255, there was no apparent effect on clinical activity. In addition, vismodegib caused side-effects, such as fatigue or nausea, and LDE255 increased the prostate-specific antigen (PSA) serum level (193,194). Treatment with itraconazole, an FDA-approved antifungal drug, demonstrated that a high dose (600 mg) may be beneficial for progress-free survival. However, such a dose has been found to cause hypokalemia (195,196). In summary, drugs targeting the SHH pathway should be further evaluated as an additional modality of PC treatment, given that more studies associated with the interactions between stromal and PC cells in relation to the AR and SHH signaling pathways are being carried out.