Introduction
The Hedgehog (Hh) signaling pathway plays critical roles for embryonic development and postnatal tissue homeostasis in organisms ranging from insects to mammals [
1]. The activation Hh pathway is initiated by binding of the secreted Hh proteins, including Sonic, Indian and Desert Hedgehog (SHh, IHh and DHh, respectively) to the 12-transmembrane receptor Patched (Ptch), thereby liberating the Ptch-mediated inhibitory effect on Smoothened (Smo), a 7-pass transmembrane protein and a key component in the Hh signaling pathway. This subsequently causes the accumulation of Smo in the primary cilium and a series of consequent intracellular events, finally resulting in the activation of the canonical transcriptional factor Gli which consists of Gli1, Gli2 and Gli3 [
2]. Analogous to other pathways active during embryonic development, inappropriate Hh pathway activity has been demonstrated to be critical for the initiation and progression of various kinds of tumors. Aberrant Hh pathway activity for tumors may occur either by mutations in key components of Hh pathway or by the production of Hh ligands in tumor cells in an autonomous and non-autonomous manner [
3].
Given that the addiction of many types of tumors to aberrant Hh pathway activity, a variety of antagonists targeting Hh pathway have been developed for the treatment of cancers. Among them, the majority function as inhibitors of Hh pathway by targeting Smo, a critical component for canonical Hh pathway [
4]. In this regard, dissecting the characteristics of signal transduction elicited from Smo is crucial and an area of intense investigation, as it may help us with the development of antagonists targeting Smo and its downstream effectors.
Smo, which possesses a structural similarity with classic G-protein coupled receptors (GPCR), has long been suggested to couple with heterotrimeric G proteins [
5‐
7]. Indeed, it has been shown that Smo may interact with Gαi and subsequently acitivate the transcriptional activity of Gli (Ci in drosophila cells) in Drosophila Cl8 cells, Sf9 cells, and NIH3T3 cells, indicating the requirement of Gαi for the activation of Gli mediated by Smo [
8,
9]. However, this argument is challenged by the observations that
pertussis toxin (PTX), which may ADP-ribosylate and consequently uncouple Gi from GPCR, fails to impact the Gli-dependent biological events such as chick neural tube patterning and some patterning events in zebrafish embryonic development [
5,
10]. On the other hand, after coupled to Gαi, Smo may exert a variety of biological activities independently of Gli, such as migration of murine embryonic fibroblasts, tubulogenesis of endothelial cells, and calcium spike activity of embryonic spinal cells [
11‐
13]. Moreover, recent study indicate that Smo may contribute to the survival of diffuse large B-cell lymphoma cells by coupling to Gαi and Gα12 and subsequently activate NF-κB independently of Gli [
14]. These studies suggest that the association between heterotrimeric Gαi proteins and Smo remains far from being fully understood, especially in the context of cancer biology. Meanwhile, in the case of canonical signal transduction of GPCR, ligand binding causes conformational changes in the structure of GPCRs, endowing them with abilities to function as a guanine nucleotide exchange factor (GEF). The exchange of GDP for GTP at the Gα subunit induces its dissociation from Gβγ dimmer [
15]. To our knowledge, whether and how Gβγ, after dissociated from Gα subunit, may impact the Smo dependent Gli activity remains as well unclear.
Although great achievements have been made for the molecular-targeted anti-cancer drugs, traditional chemotherapy is still one of the most efficient approaches for treatments of cancers. Many studies have shown that Hh signaling pathway activity plays critical roles in maintaining the chemoresistant phenotype of acquired chemoresistant cancer cells [
16‐
23]. In this study, utilizing well established acquired chemoresistant cancer cell lines and their respective parental ones, we provide a series of complementary evidences to show that Smo may promote acquired chemoresistance by activating Gli through Gαi and Gβγ-JNK signaling axis, therefore revealing that GPCR-like signaling elicited from Smo is involved in the canonical Hedgehog-Gli signaling pathway activation and the acquired chemoresistance.
Materials and methods
Drugs
Doxorubicin (Dox), Vincristine (VCR), Etoposide (VP16), Imatinib were purchased from Sigma-Aldrich (St. Louis, MO). The Hedgehog pathway antagonists cyclopamine (cyc), Robotnikinin (Robo) and GANT58 were obtained from Biovision (Milpitas, CA). The Gi antagonist Pertussis Toxin (PTX) was obtained from Invitrogen (Grand Island, NY). The JNK pathway antagonist TAT-TI-JIP was obtained from Calbiochem (Darmstadt, Germany). The agonist of Hh pathway SAG was obtained from Selleck Chemicals (Houston, TX).
Cell culture
The K562 human chronic myelogenous leukemia cell line, KB human epidermoid carcinoma cell line, NIH-3 T3 mouse embryo fibroblast cells, and HEK293T human epithelial kidney cells were purchased from the American Type Culture Collection and cultured according to the manufacturer’s instructions. The Dox selected multidrug tolerant K562/A02 subline was obtained from the Institute of Hematology, Chinese Academy of Medical Sciences (Tianjin, China), which was routinely maintained in medium containing 200 ng/ml of Dox [
24]. The VCR selected multidrug tolerant KB/VCR subline was obtained from Zhongshan University of Medical Sciences (Guangzhou, China) and was routinely maintained in medium containing 200 ng/ml of VCR [
25]. Both resistant cells were authenticated by comparing their fold resistance with that of the parental cells and examining the expression levels of ABC transporters. All experiments using K562/A02, and KB/VCR cells were performed with cells growing in the absence of Dox or VCR for at least 5–7 days to avoid drug associated secondary effects.
Plasmid constructions and lentivirus
8 × Gli-binding site luciferase reporter (8 × GBS-luciferase) and 8 × mutant Gli-binding site luciferase reporter (8 × GBS-luciferase mutant) plasmids were kindly provided by Dr. Hiroshi Sasaki. pRL-Renilla luciferase plasmid was purchased from Promega (Madison, WI). The mutant mouse plasmids SmoA1 (W539L) was generated from pEGFP-mSmo (a kind gift from Dr. Philip Beachy) using QuikChange Site-Directed Mutagenesis kit from Agilent (Santa Clara, CA) and confirmed by sequencing. The SmoA1 was next engineered into pLVX-EGFP-3FLAG-Puro lentivector. The pCDNA3 Flag MKK7-JNK1, pCDNA3 Flag JNK1(APF) and pCDNA3 Flag MKK7-JNK1(APF) plasmids were purchased from Addgene (Cambridge, MA) and confirmed by sequencing. These three plasmids were engineered into pLVX-EGFP-3FLAG-Puro lentivector. The PIRES2-ZsGreen1-Gβ1, PIRES2-ZsGreen1-Gγ2, and PIRES2-ZsGreen1-Gα transducin plasmids were purchased from Yrbio (Changsha, China) and were confirmed by sequencing. The Gα transducin (Gαt) plasmid was engineered into pLVX-EGFP-3FLAG-Puro lentivector.
Transient transfections were performed using Lipofectamine 2000 reagent (Invitrogen, Grand Island, NY) according to the manufacturer’s instructions. The viral stocks were prepared and infections were performed according to previously reported [
26].
Cell viability assay
The MTT assay was conducted as previously described to determine the sensitivity of cells to chemotherapeutic drugs [
27].
Reverse transcription PCR (RT-PCR) and quantitative RT-PCR (QT-PCR)
Total RNA was extracted from cells by RNAiso Plus Kit from TaKaRa (Dalian, China) as the instructions provided by the manufacturer and processed directly to cDNA by reverse transcription using SuperScript III kit (TaKaRa). Semi-quantitative PCR amplification was carried out using Stratagene mx3005p (Agilent Technologies). The quantitative PCR reactions were performed in triplicate with the SYBR-Green kit (TaKaRa) in iCycler iQ system (Bio-Rad; Hercules, CA). After reaction, the PCR products were subjected to electrophoresis to ensure the amplification from mRNA but not contaminated genomic DNA. The mRNA levels of interested genes were normalized to that of TATA. Primers for the genes tested were obtained from Invitrogen (Shanghai, China): SHh: 5′-CAAGCAGTTTATCCCCAATGTG-3′, 5′-TCACCCGCAGTTTCACTC-3′; IHh: 5′-TCAGCGATGTGCTCATTTTC-3′; 5′-AGCCGTAAAGAGCAGGTGAG-3′; DHh: 5′-TGCCGCTACTCTACAAGCAA-3′, 5′-GTTGTAGTTGGGCACGAGGT-3′; Gli1: 5′-GTGGGAAAGGTCTGGGATGT-3′, 5′-TGCGCCTGTCTCAGAGTAAAA-3′; TATA: 5′-ACCCTTCACCAATGACTCCTATG-3′, 5′-TGACTGCAGCAAATCGCTTGG-3′.
Western blot analysis
Cells were lysed in lysis buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 1% NP-40, 1 mM sodium vanadate, 1 mM PMSF, 1 mM DTT, 10 mg/ml of leupeptin and aprotinin) and subjected to immunoblot analysis. Primary antibodies against JNK, c-Jun, p-JNK, p-c-Jun, HA (Cell Signaling Technology; Beverly, MA), Flag, GAPDH (Santa Cruz Biotechnology; Santa Cruz, CA) were used for immunoblot analysis as standard procedure.
Dual-luciferase reporter assay
Cells were seeded into 48-well plates. Twenty-four hours later, cells were cotransfected with luciferase expression constructs as indicated and Renilla luciferase using Lipofectamine 2000 (Invitrogen). Luciferase activities present in cellular lysates after indicated treatments were measured using a Dual-Luciferase reporter assay system from Promega (Madison, WI) according to the manufacturer’s instructions and a luminometer (Molecular Device; Sunnyvale, CA). The firefly luciferase values were normalized to Renilla values.
Statistical analysis
Statistical differences were analyzed by the two-tailed Student’s t test and P < 0.05 was considered as significant. Asterisks denote statistical significance (*P < 0.05; **P < 0.01; and ***P < 0.001).
Discussion
Hh signaling pathway has been shown to be critical for a variety of physiological and pathological conditions, such as embryonic patterning, maintenance of postnatal tissue homeostasis, as well as initiation and progression of cancers [
41], whereas the molecular mechanisms responsible for its signaling transduction remain to be fully understood. Because of the structural homology with classical GPCRs, Smo has been suspected to have the ability to couple with heterotrimeric G proteins Gαi [
5‐
7]. However, the contribution of Gαi to the Hh signaling transduction is quite controversial and unclear, especially in the cancer biology. Information available so far suggests that it is context-dependent and cell-type dependent for the ability of Smo coupling to Gαi and for the subsequent participation of Gli in the biological significance initiated by the interaction of Smo and Gαi [
42,
43]. In the present study, we provide complementary evidences to show that both Gαi and Gβγ are required for the Hh pathway activity and the subsequent acquired chemoresistance by activating its canonical transcriptional factor Gli, confirming the ability of Smo coupling to Gαi and the requirement of Gαi for the Gli-dependent biological significance in the context of acquired chemoresistance. Moreover, we found that Gβγ, after released from Gαi, are also be involved in the Gli activation and acquired chemoresistance through activating JNK. Indeed, by artificially increasing the Hh pathway activity in chemosensitive cancer cells, we determined that both Gαi and Gβγ-JNk signaling axis are required for the Gli activity and Gli-dependent acquired chemoresistance mediated by SmoA1. Our data that GPCR-like signaling mediated by Smo contributes to the acquired chemoresistance by activating Gli improve our interpretations of the underlying mechanisms for the acquired chemoresistance promoted by Hh pathway and help us with improving the chemotherapeutic efficiency by using Hh inhibitors [
44]. Meanwhile, this study shed light on the understanding the nature of signaling transduction of Smo in cancer biology.
Deregulated Hh signaling has been implicated in a wide range of cancers, such as medulloblastoma, basal cell carcinoma, glioblastoma, leukemia, breast cancer, pancreatic cancer, prostate cancer, lung caner, colon cancer, to name a few [
45]. Aberrant Hh pathway activity may result from gain-of-function and loss-of-function mutations in key components in Hh pathway, such as PTCH, Smo, and Sufu [
3]. Hh pathway may as well be activated in tumors by overexpression of Hh ligands functioning in a cell-autonomous or non-cell autonomous manner. However, the activation of Hh pathway in tumor cells via a cell-autonomous manner is challenged by many controversial observations and remains to be fully elucidated [
3], for example, the inability of mutationally activated Smo expressed in pancreatic epithelial to initiate pancreatic cancer [
46]. In this study, we used the well established acquired chemoresistant cancer cell lines as an experimental model system for investigating the contribution of heterotrimeric G proteins and their downstream effectors to Gli activation mediated by Smo. Our data clearly demonstrate that acquired chemoresistant cancer cells harbor aberrant Hh pathway activity in a cell-autonomous manner, thus increasing our knowledge about the mechanisms behind Hh activation in cancers. On the other hand, many studies have shown the loss of Hh pathway activity in cancer cells possessing elevated Hh pathway activity after cultured
in vitro[
47‐
49], arguing against the use of
in vitro cultured cancer cell lines for many kinds of investigations related to Hh pathway in cancer biology, ranging from dissecting molecular mechanisms underlying Hh signaling transduction to preclinical evaluation of Hh inhibitors. In this regard, our data that acquired chemoresistant cancer cells harbor aberrant Hh pathway activity in a cell-autonomous manner identify acquired chemoresistant cancer cell lines as potential and useful
in vitro experimental model systems for investigations related to Hh pathway in cancer biology.
How does the Smo-coupled Gαi signaling link the transcriptional factor Gli in chemoresistant cancer cells? In the case of classical GPCR signaling transduction, the exchange of GDP for GTP at Gαi subunit results in the activation of Gαi, thereby repressing the adenyl cyclase and subsequently decreasing the conversion of ATP to cAMP. Reduced cAMP level implies downregulation of the activity of PKA [
15]. Considering that PKA is the key determinant for proteasome proteolysis of Gli by phophorylating it at multiple sites [
50,
51], we can envision that Gαi after activated by Smo signaling may protect Gli from proteasome degradation by inhibiting the activity of PKA in chemoresistant cancer cells in despite of required further verifications. On the other hand, in the case of classic GPCR signaling transduction, the Gβγ dimmer after releasing from Gαi may stimulate a couple of downstream effectors, such as PKC, PI3K and JNK [
35]. Data from other labs indicate that dissociated Gβγ dimmer initiated by Smo signaling may potentially promote the activation of Gli via PKC and PI3K in chemoresistant cancer cells [
52‐
55]. However, in the present study, we provide complementary evidences showing that Smo may as well promote the activation of Gli via Gβγ-JNK signaling axis. Hence, our data together with that from other labs suggest that Smo utilizes the G protein signaling to its full potential for activating the transcriptional factor Gli.
JNK, a key member of the family of MAPKs, is also called stress activated protein kinase (SAPK) and can be activated by environmental and genototoxic stress and other extracellular stimulus [
56]. JNK activation has also been linked to acquired chemoresistance by promotion of chemoresistance or by reversal of chemoresistance, relying on the duration and strength of the signaling [
56,
57]. Here, we show that JNK may function as a downstream effector of Gβγ for transmitting the signaling from Smo to Gli, thereby promoting the Gli-dependent acquired chemoresitance. Thus, this finding will help us with better understanding the role of JNK in acquired chemoresistance. Similar to ERK1/2, another critical member of MAPKs, JNK signaling is as well deregulated in many types of cancers [
56,
58]. However, the contribution of JNK in cancer development is complex and far from being fully elucidated, in other words, exhibiting context-specific and cell type-specific manner. JNK has been well known to confer the positive impact on proliferation and survival of cancer cells via its target AP1, a transcriptional factor composing Jun and Fos [
59]. Of interest, data in our study imply that Gli represents a putative downstream target of JNK, thus facilitating our better interpretation of the molecular mechanisms responsible for promoting the development of cancers by JNK. Although inhibitors of membrane protein Smo have been approved for treatment of basal cell carcinoma, the early acquired resistance to such inhibitors proposes the need for additional downstream targets [
60]. Hence, our data imply JNK as a new target for the treatment of the tumors with acquired resistance to Smo inhibitors. In this regard, how JNK promotes the activation of Gli is quite interesting, and is currently being investigated in our lab.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
XZ, JW, YL and YP conducted the experiments and were involved in data analysis. XZ helped with drafting the manuscript. WT designed the study, analyzed, and interpreted data, and drafted the manuscript. All authors read and approved the final manuscript.