Most of the data highlights in particular the crosstalk between ERα and lncRNAs in BC. In BC, ERα can be activated by estrogen in luminal subtypes, but its aberrant activity in a hormone-independent manner has also been described in basal-type subtypes (29). Numerous estrogen-regulated lncRNAs have been identified in BC. HOX antisense intergenic RNA (HOTAIR) is transcriptionally induced by E2 and it is involved in the modulation of chromatin-modifying enzymes in addition to the regulation of gene silencing (30). ERs along with a series of coregulators (histone methylases MLL1, MLL3 and CREB-binding protein/p300) bind to the promoter of HOTAIR, containing multiple functional EREs, in an E2-dependent manner (30). This interaction potentially contributes to BC evolution and progression (31,32). Gupta et al (33) described that aberrant expression of HOTAIR can affect cell motility and matrix invasion in BC cells, and its silencing in doxorubicin-resistant cells decreases cell proliferation and increases apoptosis (Fig. 1) (33). However, estrogen promoted HOTAIR through its membrane receptor GPR30 by the suppression of miR-148a (34). This highlights that the mutual modulation mechanism between ERα and HOTAIR is very complex and implies the simultaneous activation of other molecular networks.

Similar to HOTAIR, metastasis-associated lung adenocarcinoma transcript-1 (MALAT1) is regulated by estrogen and it associates with EREs along with the polycomb repressive complex 1 (PRC1) (35). MALAT1 binding to chromatin and other components can be reduced in presence of estrogen, indicating it has an inhibitory role in regulating estrogen target genes. Zhao et al (36) demonstrated that high-concentration E2 treatment largely decreased MALAT-1 RNA level. Silencing of MALAT1 showed similar effects on the proliferation, migration and invasion of BC cells. These data highlighted that E2 treatment affects BC cells proliferation, migration and invasion in an ERα-independent, via a dose-dependent way by decreasing the MALAT1 RNA level (Fig. 1) (36). Previously, a meta-analysis study revealed that high MALAT1 expression is associated with poor relapse-free survival (37) and its aberrant expression can be related to Oncotype DX 21-gene test, to predict the recurrence score, in patients with ER⁺ tumors and early BC (38). These data strongly indicated the role of MALAT-1 as an important prognostic marker in patients with BC.

Myocardial infarction-associated transcript (MIAT) is overexpressed in ER⁺ BC tissues especially in aggressive ductal BC (39). In these cells, MIAT expression is induced by E2 through ERα activation. Silencing MIAT inhibited cell cycle progression and induced apoptosis, whereas its upregulation increased cell proliferation in MCF-7 cells (39). Moreover, Luan et al (40) underlined that MIAT promotes BC progression and functions as competing endogenous RNA to regulate dual specificity phosphatase 7 (DUSP7) gene expression by sponging miR155-5p in BC. Diethylstilbestrol led to a dose- and time-dependent upregulation of MIAT in MCF-7 cells that were dependent on ERα. MIAT silencing and pharmacological inhibition decreased cell proliferation by perturbing G1 to S phase cell cycle transition (41). The functional studies could suggest the targeting of MIAT for the optimization of therapeutic strategies in Erα⁺ patients with BC.

The lncRNA LINC00472 promoter has an ERα binding site and can be upregulated by ERα. When overexpressed, it can interact with the NF-kB transcription factor modulating the downstream pathway (42). LINC00472 overexpression can perturb tumor growth and reverse the aggressive phenotype of breast tumor cells in vitro. A meta-analysis, performed on clinical studies, described overexpression of LINC00472 in ER⁺ tumors, associated with improved outcomes of patients, and its low/absent expression in ER⁻ tumors (42).

Even LINC01016 can be to be directly regulated by ERα, having an ERα binding site on its promoter, and thereby its expression is rapidly induced by estrogen treatment. Silencing of LINC00160 results in reduced proliferation (43). ER⁺ breast tumors exhibit high expression levels of LINC01016 and patients with higher LINC01016 expression are related to poor overall survival (43). Understanding the molecular mechanisms and molecular pathways involved in the crosstalk between these lncRNAs and Erα could therefore suggest new potential prognostic-predictive markers in patients characterized by ERα⁺. However, all these data should be confirmed on a large series of patients.

The lncRNA Down syndrome cell adhesion molecule antisense RNA 1-AS1 (DSCAM-AS1) is an estrogen-responsive lncRNA differentially expressed between benign and malignant BC cells (44) and it can distinguish features between luminal subtypes of BC (45). Its overexpression in patients with BC was associated with poor overall survival (46,47). DSCAM-AS1 can induce G1/S transition of cell cycle and represents a poor prognostic factor in luminal patients with BC treated with hormone therapy (48). In vitro studies highlighted that DSCAM-AS1 interacts with nuclear ribonucleoprotein (hnRNPL) in BC cells, promoting tumor progression and inducing tamoxifen resistance (49). More recently, Yadav et al (50) demonstrated that progesterone also can induce expression of DSCAM-AS1 and its silencing mimics the effect of progesterone in impeding cell migration and invasion in PR⁺ BC cells. Additionally, DSCAM-AS1 sponges miR-130a that regulates the expression of ESR1 by binding to its 3′-untranslated region (UTR), thereby mediating the effect of progesterone in BC cells (50). These data underlined once again that the reciprocal modulation mechanisms are very complex and that they can also involve other mediators such as progesterone.

Numerous other lncRNAs can be deregulated under the effect of estrogen and therefore by the modulation of Erα: i) Eosinophil granule ontogeny transcript (EGOT) dysregulation is associated with tumor size, lymph node metastasis progression and worse survival in patients with BC. Estrogen can reduce the expression of EGOT in a dose-dependent manner, thereby modulating autophagy in BC cells (51). ii) Two lncRNAs, lncRNA152 and lncRNA67, exhibited significant overexpression in ER⁺ breast tumors in comparison to non-cancerous breast tissue; moreover, after estrogen treatment, it can regulate cell proliferation and mitogenic division in ER⁺ cell lines (52). iii) A recent study, analyzing lncRNAs that are either dysregulated in response to estrogen treatment in BC cells, identified ~90 upregulated lncRNAs and 160 downregulated lncRNAs by E2 stimulus. In ER⁺ patients with BC, lncRNA LINC00160 demonstrated a strong correlation with the survival rate and its neighboring E2-responsive coding gene RUNX1 (53).

Numerous other lncRNAs estrogen signaling-dependent are associated with endocrine therapy resistance in BC. BC cells can develop different mechanisms to promote intrinsic and acquired resistance to endocrine therapy; between these, the ligand-independent transactivation of the ERα is one of the main mechanisms responsible for acquired resistance, in which numerous lncRNAs play an essential role. In particular, it is known that aberrant ERα signaling is strongly involved in tamoxifen resistance and several lncRNAs could modulate and promote this process (54). H19 lncRNA level was higher in ER-positive BC tissues than in ER-negative tumors suggesting that it is an estrogen-inducible gene and plays a key role in estrogen-induced cell proliferation (55). Treatment of BC cells with tamoxifen or fulvestrant increases H19 expression while its silencing leads to drug resistance. H19 expression can be modulated by Notch and c-MET inhibition can reverse resistance to tamoxifen and fulvestrant in an H19-dependent manner in these cells (56). Moreover, altered ERα expression can also modify H19 levels thereby modulating the apoptosis response, through pro-apoptotic gene BCL2 interacting killer (BIK), to chemotherapy in BC cells (57).

lncRNA in non-homologous end joining (NHEJ) pathway 1 (LINP1) expression is upregulated in tamoxifen-resistant BC cells and its knockdown resulted in increased tamoxifen-induced apoptosis. LINP1 transcription is suppressed in the presence of estrogen, and tamoxifen treatment restores its expression (56).

BC anti-estrogen resistance 4 (BCAR4) is inversely associated with the development of resistance to anti-estrogens in BC cells and associated with worse survival and also a higher risk of metastasis in other cancers (58). Recently, a polymorphic variant of the BCAR4 gene associated with BC susceptibility has been identified (59).

The lncRNA DLGAP1 antisense RNA 2 (DLGAP1-AS2) can promote ERα signaling, binding to the AFF3 protein and thereby the inhibition of the degradation of it. It is significantly upregulated in BC and tamoxifen can induce DLGAP1-AS2 expression (60). In the same manner, the lncRNA ERLC1 is involved in ERα signaling in BC tissues. Functionally, ERα induces ERLC1 transcription, which in turn stabilizes the ESR1 transcript by interaction with miR-129 and FXR1. Its expression became aberrant in tamoxifen-resistant BC cells, and its silencing restored sensitivity to tamoxifen and increased the efficacy of palbociclib or fulvestrant therapy (61).

The lncRNA estrogen inducible lncRNA (ERINA) is aberrantly expressed especially in ER⁺ BCs where it was inversely associated with survival and sensitivity to CDK inhibitor. Mechanistically, ERINA interacts with the E2F transcription factor 1 (E2F1), which prevents the binding of E2F1 to the tumor suppressor retinoblastoma protein 1 (RB1), thereby promoting cell-cycle progression. ERINA acts as an ER-responsive gene, and an intronic ER-binding site was identified as an enhancer that mediates the transactivation of ERINA (62). With a similar mechanism of action, Yu et al (63) described the lncRNA actin gamma 1 pseudogene 25 (AGPG) as a regulator of E2F1 activity in endocrine resistant BC. AGPG physically interacted with purine rich element binding protein A, thus releasing E2F1 and promoting its signaling activation in ERα⁺ BC cells.

Horie et al (64) revealed that the BC Natural antisense transcript 1 (BNAT1) can induce tamoxifen-resistant in ER⁺ BC cells and its silencing significantly reduced the in vitro and in vivo growth of cancer cells. This mechanism is mediated by binding of BNAT1 binding with ERα and its knockdown reduced ERα expression and its transactivation.

More recently, Chen et al (65) indicated that LINC02568 regulates estrogen/ERα-induced gene transcriptional activation in trans by stabilizing ERα mRNA through sponging miR-1233-5p in the cytoplasm and by regulating carbonic anhydrase CA12 in cis in the nucleus. This crosstalk modulates BC cell growth and tumorigenesis as well as endocrine therapy drug resistance (65).

In conclusion, the ability of numerous lncRNAs related to Erα activity to modulate resistance to therapies makes them important predictive biomarkers, and the evaluation of their aberrant expression could provide an additional tool for the therapeutic stratification of patients with BC.

Some lncRNAs are involved in estrogen signaling in prostate cancer. In particular, HOTAIR and MALAT1 can interact with both ERα/ERβ at the chromatin level (35). HOTAIR and MALAT1 silencing determine the abrogation of estrogen sensitivity of the estrogen target gene pS2. Nanni et al (66) described that MALAT1 has a role in the transcription repression of hormone receptors target genes such as pS2 and PSA in prostate cancer cells and orthotopic models (35). These observations make HOTAIR and MALAT1 important biomarkers and potential therapeutic targets in prostate cancer.

The nuclear enriched abundant transcript 1 (NEAT1) is a specific transcriptional target of ERα in prostate cancer and it is significantly overexpressed in cancer as compared with benign prostate. In prostate cancer cell lines, ERα overexpression and E2 treatment upregulated NEAT1 transcript levels in a time-dependent manner. NEAT1 acts as a transcriptional activator of different PCa genes and its silencing modified ERα target genes, suggesting that NEAT1 is not only a downstream target but also a mediator of ERα signaling in prostate cancer cells. Moreover, in vitro and in vivo NEAT1 expression is associated with aggressive phenotype and drug resistance (67). Additionally, NEAT1 lncRNA could therefore represent an important prognostic biomarker and a predictive marker of therapeutic response in patients with prostate cancer.

As previously described, lncRNA H19 is more sensitive to estrogens, a stimulus with a pro-tumoral role in prostate cancer. Bacci et al (68) described H19-dependent mechanisms by which estrogen and hypoxia signaling promote the acquisition of an aggressive phenotype in prostate cancer cells. H19 can be independently upregulated by estrogen or hypoxia. Moreover, the same authors revealed that H19/estrogen signaling has a potential impact on the TME by the switch from EMT to a β integrin-mediated invasion (68).

Numerous other lncRNAs are capable of modulating and being modulated by ERα also in other neoplastic pathologies. The lncRNA LINC00312, also known as nasopharyngeal carcinoma-associated gene 7, a novel putative tumor suppressor, is essential for ERα binding and suppression activity in nasopharyngeal carcinoma (NPC). It can promote NPC cell invasion by regulation of ERα and the HRAS/p-c-RAF and JNK2/AP-1/MMP1 signaling pathways (69).

Estrogen signaling has long been implicated in epithelial ovarian cancer progression. The lncRNA ElncRNA1 is transcriptionally regulated by E2 through ERα-EREs and it was described as able to promote epithelial ovarian cancer cell proliferation (70). Moreover, Qiu et al (71) reported that E2 significantly dysregulated 115 lncRNAs in ERα-positive SKOV3 ovarian cells compared with E2-untreated controls. Among these, both TC0101441 and TC0101686 were dysregulated by E2 also in other ERα-positive PEO1 ovarian cells and dysregulated in ovarian cancer patients. In particular, high TC0101441 expression was associated with lymph node metastasis, exhibiting a potential involvement in the metastatic progression of ovarian cancer (71).

The lncRNA DSCAM-AS1 overexpression in endometrial cancer (EC) has a significant association with shorter overall survival of patients with EC. In these cells, it is associated with the endometrial tumor promotor prolactin gene and with the expression of ERα together with its target genes trefoil factor 1 and progesterone receptor (PgR). DSCAM-AS1 silencing in EC cell lines reduced cell growth and proliferation by inhibition of NOTCH1, Protein Tyrosine Kinase 2 as well as early growth response factor 1 (EGR1) expression (45).

The lncRNA LINC00899 overexpression is able to inhibit the viability, proliferation, migration and invasion of cervical cancer cells. miR-944 could competitively bind with LINC00899 and its upregulation can inhibit ER expression, thereby promoting the migration and invasion of cervical cells. These data highlight the role of LINC00899/miR-944/ESR1 axis in the regulation of cervical cancer progression (72).

More recently, Cheng et al (73) revealed that lncRNA maternally-expressed 3 (MEG3) promoter binds to ERα in liver cancer cells. ERα upregulation repressed the proliferation, migration as well as invasion and promoted apoptosis of HepG2 cells under high glucose conditions. ERα silencing decreased MEG3 expression and in turn enhanced proliferation, migration and invasion. These data highlighted that ERα has a protective role in hepatocellular carcinoma (HCC) being able to inhibit cell progression through positive regulation of lncRNA MEG3 (73).

Androgen as well as ARs play a crucial role in the initiation and progression of prostate cancer (88). Numerous lncRNAs associated with prostate cancer play oncogenic roles by modulating AR-mediated activity through different mechanisms (89).

Two lncRNAs, prostate-specific transcript 1 (PCGEM1), and prostate cancer associated ncRNA 1 have been described to directly interact with AR under ligand-stimulated conditions, but the data regarding the two lncRNAs are contradictory (90,91), questioning their real involvement in prostate cancer progression and their ability to interact and modulate the AR signaling pathway.

HOTAIR is involved in cancer progression especially in androgen-sensitive to castration-resistant prostate cancer (CRPC) cells (92). HOTAIR can modulate androgen-independent AR activation and AR-mediated gene transcription. It interacts with AR thereby inhibiting its ubiquitination and degradation (92). Moreover, HOTAIR forms a repressive complex with polycomb proteins and decreases AR transcription which results in increased cell invasion and MMP9 expression (93). These data evidence HOTAIR as a novel target gene playing a critical role in invasiveness of prostate cancer cells.

Ozgur et al (94) revealed that cellular H19 expression strongly decreased after DHT stimulation in hormone-sensitive prostate cancer cells, and AR inhibition by enzalutamide restored the DHT effect as well as increased H19 expression. The present study provided evidence of the involvement of H19 lncRNA in androgen/receptor pathway-dependent prostate cancer development (94).

Prostate cancer gene 3 (PCA3) lncRNA is overexpressed in prostate cancer tissues compared with adjacent non-neoplastic tissues and it can be involved in AR signaling, since its silencing promotes cell growth and viability (95,96). Moreover, PCA3 knockdown modified the expression of PSA and PCGEM1, AR-inducible genes, as well as sensitized PCa cells to enzalutamide (97). PCA3 silencing also highlights its role in EMT, leading to the overactivation of epithelial markers E-cadherin, claudin-3 and cytokeratin-18 and the dysregulation of vimentin (96). Due to its aberrant oncogenic activity, PCA3 could represent a potential therapeutic target in patients with prostate cancer.

Growth arrest-specific 5 (GAS5) lncRNA plays a crucial role in prostate cancer carcinogenesis and its knockdown in prostate cancer cells leads to the inhibition of tumor progression (98). The interaction between GAS5 and AR could perturb cell proliferation of CRPC. GAS5 silencing in CRPC would upregulate the AR axis, while the high activation of AR could further inhibit the transcription of GAS5 (98). GAS5 can interact with both GR and AR regulating GR/AR-mediated transcriptional activity (99).

The lncRNA BCAR4 is overexpressed in prostate cancer tissues and strongly associated with metastatic progression as well as poor prognosis (100). In AR⁺ PCa cells, upregulation of BCAR4 increases cell growth as well as migration and induce resistance to androgen (101). On the contrary, in AR⁻ prostate cancer cells BCAR4 silencing decreased cell growth (101). Therefore, as well as in BC, BCAR4 represents a potential prognostic and predictive biomarker in patients with prostate cancer.

Instead, some lncRNAs behave like tumor suppressor genes. For example, downregulated RNA in androgen independent cells (DRAIC) lncRNA can inhibit the migration of prostate cancer cells. AR activity induction leads to DRAIC transcription reduction and promotes prostate cancer progression (102). Moreover, DRAIC downregulation increased cell invasion and soft agar colony formation by NF-κB activation (103).

Another tumor suppressor lncRNA, prostate cancer associated transcript 29 (PCAT29) suppresses androgen signaling in prostate cancer. PCAT29 silencing promoted prostate cancer growth and migration while its overexpression reduced these processes. PCAT29 downregulation results were associated with recurrence in patients with prostate cancer (104). Several other lncRNAs are involved in reciprocal modulation with AR and in the progression of prostate cancer.

AR-regulated lncRNA 1 lncRNA plays a crucial role in AR-dependent prostate cancer development as well as CRPC. It is able to interact with AR mRNA transcript and stabilizes it for protein translation (105).

The androgen-dependent lncRNA suppressor of cytokine signaling 2 antisense (SOCS2-AS1) promotes androgen-dependent prostate cancer growth and migration (87). In CRPC cells increased AR signaling has been strongly associated with SOCS2-AS1 upregulation compared with androgen-dependent LNCaP cells. Moreover, SOCS2-AS1 can modulate the AR target TNF Superfamily Member 10 gene (106).

The prostate cancer upregulated lncRNA 1 is overexpressed in prostate cancer cells and its silencing inhibited AR and NKX3-1 expression, whereas AR silencing decreased PlncRNA-1 expression (107).

TMPO antisense RNA 1 (TMPO-AS1) lncRNA is overexpressed in prostate cancer cell lines and tissues associated with a worse prognosis of patients with prostate cancer. AR can directly regulate the TMPO-AS1 levels in the cytoplasm promoting growth, progression, cell migration and inhibition of apoptosis in prostate cancer cells (108).

The AR-regulated lncRNAs, namely CRPC-lncs, are highly expressed in CRPC tissues. Their silencing reduces the expression of AR and AR variants which as a consequence inhibit CRPC tumor growth (109).

MALAT1 is increased in prostate cancer cells after androgen stimulation, as well as its receptor. MALAT1 knockdown inhibited the DHT administration-induced increase in the cell cycle, induced AR expression in these cells and blocked the tumorigenesis of prostate cancer cells in nude mice. Moreover, MALAT1 can bind miR-320b, negatively regulating its expression, and AR is a target of miR-320b. Consequently, changes induced by MALAT1 silencing can be abolished by miR-320b inhibition or AR overexpression (110).

Recently, a large number of lncRNAs have been associated with AR activity and AR inhibitors sensibility: i) LINC00675, overexpressed in androgen-insensitive prostate cancer cell lines and CRPC, patients can directly modulate AR interaction with mouse double minute-2 (MDM2) and block the ubiquitination of AR by binding to it (111), ii) prostate cancer associated transcript 7 (112) and prostate cancer lncRNA at chromosome 16 (113) lncRNAs are directly related with AR signaling and overexpressed in prostate cancer tissues and AR-dependent cell lines. Their silencing suppressed AR signaling, tumor growth and proliferation. iii) Growth hormone secretagogue receptor (GHSR) opposite strand (GHSROS) lncRNA is overexpressed in prostate cancer cell lines and modulates the expression of protein phosphatase 2 regulatory subunit b gamma gene. The loss of GHSROS may drive AR pathway-independent prostate tumor progression (114). iv) FAM83H-AS1 is upregulated in prostate adenocarcinoma and its silencing in prostate cancer cells modulated cell proliferation, cell cycle and migration. AR signaling is involved in upregulating FAM83H-AS1 expression in prostate cancer cells (115). v) Poly(RC) binding protein 1 antisense RNA 1 is involved in CRPC enzalutamide resistance by enhancing the deubiquitination of AR/AR-V7 (116). vi) Negative expression of AR Regulating LncRNA (NXTAR) bound upstream of the AR promoter inhibiting AR/AR-V7 expression, which in turn upregulates NXTAR levels, and compromising enzalutamide-resistant prostate cancer (117). vii) AC016745.3 overexpression inhibits the proliferation and migration of prostate cancer cells and suppresses the expression of AR target genes by binding non-POU domain containing octamer binding protein (118). viii) RP11-1023L17.1 expression can be directly suppressed by the AR as well as promotes proliferation, migration, cell cycle progression and inhibition of apoptosis. RP11-1023L17.1 can suppress F-Box Protein 32 expression, thereby enhancing c-Myc protein stability (119).

AR is overexpressed in normal breast epithelium as well as in a large proportion of BC tissues and cell models. AR is co-localized with ER together with PgR in epithelial breast cells, indicating a combined role in the modulation of mammary epithelial cell proliferation, and is silent in myoepithelium as well as stromal cells (120,121). However, in a normal breast, androgens are involved in inhibiting breast development, whereas, in BC cell lines, androgens and AR signaling can promote breast cell proliferation and growth (122).

A strong relation between AR signaling and lncRNA has been highlighted in BC in particular in triple-negative BC (TNBC). One of the TNBC subtypes called luminal AR (LAR) was characterized by AR expression (123). A strong association between the lncRNA HOTAIR and AR expression was demonstrated in TNBC patient samples (124). HOTAIR overexpression appeared associated with lymph node metastases and with AR expression indicating a role of HOTAIR in TNBC progression by modulation of the AR signaling pathway (124).

Yang et al (125) studied three TNBC cell lines, treated with DHT, to investigate the role of AR-related lncRNAs. They highlighted the downregulation of AR negatively regulated lncRNA (ARNILA), and its role as a negative prognostic marker in patients with TNBC. In vitro and in vivo studies revealed that ARNILA can induce invasion as well as metastatic progression (125). Moreover, it can promote EMT by upregulating SOX4 expression (125).

SRA1-mediated AR signaling pathway regulation may be also involved in BC development, especially in TNBC. Leygue (126) reported that silencing of SRA1 could reduce the invasiveness and motility of TNBC cells (126).

Currently, Huang et al (127) conducted a study to determine the role of the AR signaling pathways (ARSP)-related lncRNAs in the incidence and progression of BC and their relationship with the TME. A total of 340 ARSP-related lncRNAs were detected, of these 203 were upregulated and 137 were downregulated (127). The understanding of the molecular mechanisms beneath the crosstalk between AR and lncRNAs in BC could provide new tools for the prognostic stratification of patients including the development of new therapeutic strategies.

The effects of lncRNAs on AR signaling have also been assessed in other types of cancers.

In bladder cancer, combined overexpression of X inactive specific transcript (XIST) and AR have been associated with the advanced TNM stage. XIST knockdown leads to a decrease in proliferation, invasion and migratory potential through modulation of AR signaling. Moreover, XIST suppresses the expression of miR-124 which target 3'-UTR of AR through direct interaction (128).

Nevertheless, in bladder cancer, another lncRNA, LINC00460, appeared aberrantly overexpressed and its silencing decreased the proliferation of bladder cancer cells through modulation of AR expression. Moreover, LINC00460 overexpression has been associated with poor prognosis for patients with bladder cancer (129).

In esophageal squamous cell carcinoma progression, Wu et al (130) discovered a micro peptide (YY1BM) encoded by Y-linked lncRNA LINC00278, able to block YY1 binding to AR to activate the expression of eEF2K. Recently, He et al (131) showed that AR-mediated transcription activation is responsible for the overexpression of LINC01503, which promoted NPC cell proliferation, migration and invasion in vitro, and facilitated tumor growth as well as metastasis in vivo.

Schmidt et al (132) demonstrated that in melanoma cells lncRNA SRA-like ncRNA (SLNCR) binds to AR and complexes with different transcription factors to mediate invasion or proliferation. More recently, they revealed that AR and EGR1, a transcription factor mainly involved in the processes of tissue injury, immune responses and cancer, bind to the lncRNA SLNCR and increase melanoma proliferation by p21 (133). The same authors later proved that point mutations or oligonucleotides that abrogate AR binding to SLNCR block melanoma invasion, suggesting that targeting lncRNA-protein complexes holds therapeutic promise (134).

Zhai et al (135) reported that suppressing AR in RCC (SARCC) lncRNA establishes a physical interaction with AR acting as a tumor suppressor in RCC cells. SARCC expression appears downregulated in RCC tissues and metastatic samples compared with non-neoplastic adjacent tissues. In addition, its dysregulation is strongly associated with the prognosis of patients with RCC. In vitro, low expression of SARCC associated with cell invasion, migration, proliferation and drug resistance to sunitinib. Mechanistically, SARCC/AR crosstalk inhibited miR-143-3p expression by modulating K-RAS, MMP-13, AKT and P-ERK expression (135). HOTAIR is also capable of interacting with AR in RCC cells, enhancing the expression of VEGFA and platelet derived growth factor receptor alpha (PDGFA) with the promotion of tumor angiogenesis, and inducing tumor stemness phenotype (136).

Recently, You et al (137) showed that AR can increase TWIST1 associated lncRNA regulated via AR expression through binding to the AREs promoting vasculogenesis and metastasis in clear cell RCC via altering TWIST1.

The lncRNA prostate androgen-regulated transcript 1 (PART1) interacts with AR promoting promyelocytic leukemia zinc finger (PLZF) expression, followed by recruitment of EZH2, to mediate epigenetic PDGFB silencing and downstream PI3K/Akt inhibition, and modulating gastric cancer progression (138).

Finally, Qin et al (139) described the upregulation of LINC00667 in HCC tissues and cell lines associated with an unfavourable prognosis of patients with HCC. LINC00667 acts as a molecular sponge in the miR-130s-3p/AR signaling pathway thereby promoting HCC tumor progression (139).

All-trans retinoic acid (RA) is the main active metabolite of vitamin A (140). RA signaling is involved in different biological processes, including embryonic development along with organogenesis, cell growth, proliferation, cell differentiation, together with metabolism regulation (141). RA translocates in the nucleus by binding to an intracytoplasmic transporter, cellular RA binding protein type II. Retinoids bind to two different NRs, RARs (RAR-α and RAR-β) and retinoid X receptors (RXRs). RARs function as heterodimers with RXRs. In turn, RXRs can be involved in the formation of heterodimers with other different NRs, such as vitamin D₃, thyroid hormone and PPARs (142).

RAR/RXR and RXR/RXR are localized in the nucleus bound to retinoid hormone response elements (RAREs) in the promoter regions of retinoid-responsive genes thereby inducing the transcription of numerous target genes (143).

There are few examples in the literature regarding crosstalk of RAR signaling and lncRNAs in cancer. In acute promyelocytic leukemia (APL) RA/RARs signaling can induce the expression of the lncRNA HOXA-AS2 antisense, which suppresses the TRAIL pathway and reduces expression of caspases (144). Similarly, RA/RARs signaling induces H19 expression in APL cells and its upregulation perturbs telomerase activity, mediated by hTERC-hTR interaction, during tumor evolution (145). Conversely, the lncRNA RA early transcript 1K pseudogene (RAET1K) can reduce RA/RARs signaling. In lung cancer RAET1K interferes with the protective role of RA, modulating miR-135a-5p, which in turn inhibits Cyclin E1 (CCNE1) expression, a cyclin required for G1-S transition. RA promotes the expression of this miRNA, by blocking lung cancer cells in the G1 phase and thereby tumor cells proliferation and progression. RAET1K silencing can repress CCNE1 expression and inhibit cell cycle progression from the G1 to the S phase (146).

A recent study also highlighted crosstalk between HOTAIR and RA/RARs signaling in acute myeloid leukemia. Different acute myeloid leukemia cell lines (HL-60, NB4, U937 and THP-1) have induced myeloid differentiation by all-trans RA (ATRA). All four cell lines exhibited a higher expression level of HOTAIR after the ATRA treatment suggesting its role in the regulation of RA/RARs pathway and consequently in myeloid cell differentiation (147). The lncRNA RNA HOTAIR myeloid 1 (HOTAIRM1) can mediate ATRA-induced differentiation of ALK cells (148). Upregulation of HOTAIRM1 expression is able to enhance ATRA-induced PML-RARA degradation by affecting autophagic flux and thereby controlling myeloid cell differentiation in APL cells (149). lncRNA NEAT1 is also involved in APL cell differentiation induced by ATRA. C/EBPα binds and transactivates NEAT1 whereas PML/RARα suppresses this process suggesting that PML/RARα could contribute to the pathogenesis of APL by suppressing C/EBPα targets (150).

In glioblastoma multiforme cells ATRA treatment dysregulated different lncRNAs' expression. In particular, ATRA dose-dependently decreased the expression of GAS5 within the microenvironment of the U87MG cell line (151).

RAET1K is necessary for retinoid-induced differentiation in different cells by binding RXRs. RAET1K is upregulated in tumor tissues and is associated with a poor prognosis in patients with lung adenocarcinoma. Aberrant RAET1K expression upregulated cyclin E1 by targeting miR-135, thereby promoting tumor progression (146).

Currently, Wang et al (152) asserted that lncHOXA10 is significantly upregulated in gastric cancer tissues and cell lines, and can promote proliferation, migration as well as invasion of cells. lncHOXA10 can suppress RAR-β expression, involved in the modulation of apoptosis, while ATRA can rescue the expression of RAR-β, inhibiting lncHOXA10 in gastric cancer cells (152). RAR-β is also modulated by lncRNA HAND2-AS1 in human bladder cancer. HAND2-AS1 inhibits cell proliferation and promotes apoptosis in bladder cancer cells by sponging miR-146. The last of these can promote cell proliferation by targeting RAR-β and, in turn, HAND2-AS1 can suppress cell proliferation via releasing RAR-β from miR-146 (153). More recently, Fu et al (154) described that another lncRNA, lymphocytic leukemia 2 (DLEU2), can modulate RAR-β in colorectal cancer cells. DLEU2 induced promoter methylation of RAR-β to downregulate its expression, and its upregulation induced the MAPK signaling pathway promoting colorectal cancer progression (154).

VDR is a transcription factor that besides interacting with hormonally active vitamin D3, regulates the expression of more than 900 genes involved in numerous different cellular processes. Both vitamin D/VDR signaling and lncRNAs affect numerous genomic and non-genomic processes, the dysregulation of which can be associated with a wide range of diseases including cancer. The principal evidence is associated with skin cancers. Several studies confirmed the protective effect of VDR in cancer initiation and progression (155), especially in keratinocyte carcinoma (156). Jiang and Bikle (157) profiled 90 well-annotated mouse lncRNAs from cultured mouse keratinocytes after deleting VDR. They detected that H19, HOTTIP, mHOTAIR, Malat1, SRA and Nespas were significantly increased, whereas H19 as, Kcnq1ot1 and lincRNA-p21 were decreased (157). Subsequently, other VDR-associated lncRNAs in BC (158) as well as in lung cancer (159) have been identified. VDR, MALAT1 and LINC00511 were significantly upregulated in tumor cells in BC compared with non-cancerous tissues. In addition, VDR and another lncRNA SNHG16 were associated in both tumor and non-tumor breast tissues (158). In lung cancer, the coordinate expression of VDR and the lncRNAs MALAT1, SNHG16, SNHG6, LINC00346 and LINC00511 has also been validated in lung cancer tissues (159).

In oral squamous cell carcinoma (OSCC) Jin et al (160) evaluated crosstalk between vitamin D/VDR signaling and lncRNAs selecting 46 pairs of tumor tissue and adjacent non-tumor tissue as well as two OSCC cell lines. In the latter, they observed 1045 lncRNAs differentially expressed after the 1,25(OH)₂D treatment. In particular lung cancer-associated transcript 1 (LUCAT1) expression was reduced by the 1,25(OH)₂D treatment in vitro and it appeared strongly overexpressed in the OSCC tumor tissues. LUCAT1 silencing in the OSCC cell lines reduced cell growth in association with a reduction in ERK1/2 phosphorylation (160).

The oncogenic lncRNA colon cancer-associated transcript 2 (CCAT2) has been described as overexpressed in ovarian cancer. Wang et al (161) described that 1,25(OH)₂D inhibited CCAT2 expression leading to decreased binding of transcription factor 4 (TCF4) to the MYC promoter in ovarian cancer cell lines. This produced inhibition of proliferation, migration and invasive features of ovarian cells (161). In addition, the lncRNA TOPORS-AS1, overexpressed in ovarian cancer cells and involved in suppression of cell proliferation, migration and invasion, is upregulated by VDR. Crosstalk is essential to interrupt the Wnt/β-catenin signaling thereby defining an improved prognosis in patients with ovarian cancer (162).

Finally, the relation between the lncRNA H19 and VDR has been described in colon cancer. VDR can inhibit H19 expression through the regulation of the c-Myc/Mad-1 network. In turn, H19 upregulation blocked VDR expression by upregulating miRNA 675-5p (163).

Human GR (h-GR) belongs to the nuclear hormone receptor superfamily and is encoded by nuclear receptor subfamily 3C1 gene on chromosome 5q31.3 (164). In the absence of ligands, the GR is sequestered in the cytoplasm by chaperone proteins. Glucocorticoids, such as cortisol, prednisolone and dexamethasone bind to GR leading to both its dimerization and translocation into the nucleus, where it performs its activity as a transcription regulator (164). DNA-bound GR recruits coregulator complexes forming transcription regulatory complexes that can function in both activation and suppression of transcription. Some lncRNAs have been described as regular and can be regulated by glucocorticoid/GR signaling.

GAS5 lncRNA is able to directly interact with the DNA binding domain of GR acting as a molecular decoy by competing with the glucocorticoid-response element for GR binding, thereby inhibiting GR transcription (165). GAS5 is also able to accumulate in cells that have been starved of growth factors by suppressing GR activity, and GAS5 sensitizes cells to apoptosis (165). In the same way, overexpression of GAS5 leads to growth arrest and apoptosis in human breast cell lines (165).

Thyroid hormone action is predominantly mediated by THRs, which are encoded by the THRA and THRB genes (166). THRs act as ligand-activated nuclear transcription factors to regulate different physiologic processes through direct gene regulation and with ncRNAs relation (167,168).

Current evidence indicates a crosstalk between the thyroid hormone/THR pathway and lncRNAs in liver cancer. Brain cytoplasmic RNA 1 (BCYRN1/BC200) is abnormally overexpressed in several tumor types, with a significant overexpression in HCC tissues. Using the Disease-Related Human lncRNA Profiler to identify lncRNAs regulated by thyroid hormone/THR signaling in liver cells, BC200 was identified as a lncRNA downregulated by the thyroid hormone (169).

In HCC cells, the taurine upregulated gene 1 (TUG1) is overexpressed and can induce cell proliferation, invasion, metastatic progression as well as apoptosis by distal-less homeobox 2 (DLX2) activation (170). Alpha-fetoprotein (AFP) is increased in the majority of patients with HCC and TUG1 is involved in THR/AFP signaling. Thyroid hormones suppress TUG1 expression, leading to the downregulation of AFP (171).