Adaptive Transcriptional Responses by CRTC Coactivators in Cancer

Abstract

Adaptive stress signaling networks directly influence tumor development and progression. These pathways mediate responses that allow cancer cells to cope with both tumor cell-intrinsic and cell-extrinsic insults and develop acquired resistance to therapeutic interventions. This is mediated in part by constant oncogenic rewiring at the transcriptional level by integration of extracellular cues that promote cell survival and malignant transformation. The cAMP-regulated transcriptional coactivators (CRTCs) are a newly discovered family of intracellular signaling integrators that serve as the conduit to the basic transcriptional machinery to regulate a host of adaptive response genes. Thus, somatic alterations that lead to CRTC activation are emerging as key driver events in the development and progression of many tumor subtypes.

CRTC Coactivators Direct Adaptive Cell Signaling Responses Involved in Oncogenesis

Cells experience a constant barrage of signals that must be interpreted and integrated. Multicellular organisms have further evolved an elaborate network of intracellular signal transduction cascades to process and transduce these inputs into diverse outputs affecting gene expression, cell proliferation, differentiation, and apoptosis [1]. To a large extent, these adaptive response-signaling networks are sequentially activated to maintain homeostasis and prepare cells to respond to subsequent signaling and/or new stimuli [2]. These adaptations are crucial for cells and tissues to dynamically endure unexpected and changing environmental pressures. However, this biological phenomenon creates an unfortunate liability since oncogenic mutations that arise in cancer cells evolve to co-opt these adaptive regulatory networks [3]. This hypothesis was recently confirmed by high throughput clustered regularly interspaced short palindromic repeat (CRISPR) screens where validated cancer gene pathways were specifically enriched in the subset of context-specific essential genes and not within core essential genes [4]. Furthermore, while >1000 genes are required for cell division under stable conditions [5], inspection of recurrent somatic mutational events in human tumorigenesis reveals a surprising paucity of mammalian cell cycle targets focused on a narrow CDK4/6/cyclin D/retinoblastoma pathway that integrates growth factor signaling [6,7]. Thus, malignancy appears to arise when cells exploit or aberrantly deregulate normal adaptive stress signaling mechanisms to sustain proliferation, evade growth suppression and apoptosis, avoid immune destruction, and promote inflammation, angiogenesis, invasion, and metastasis [8].

Malignant transformation imposes extreme pressures on cellular processes that under normal homeostatic conditions might push cells towards terminal differentiation, growth arrest, quiescence, or death. Normal human cells are highly protected against neoplastic transformation and clonal expansion, and require the sequential disruption of at least five parallel and interconnecting pathways that include adaptive stress responses that reset energy sensing and growth factor requirements [9]. This oncogenic rewiring reflects a dynamic interplay between the cancer cell and adaptive stress responses that have been usurped to establish a new homeostatic state that promotes cell survival, metastatic expansion, and acquired resistance to therapeutic interventions [10,11]. Thus, the field of cancer research has shifted from focusing solely on tumor cell dynamics, to rather a more integrated approach that includes contextual cues as a critical parameter influencing each step of the tumorigenic process. In this aspect, our understanding of the role that the host microenvironment, external cell signaling, and immune response play in cancers is being revised [11,12].

A central theme underlying adaptive stress signaling, conserved from single cell to complex multicellular organisms, is the ability to store and retrieve energy sources to ultimately allow DNA replication under unpredictable external conditions. For example, cell surface receptors are key transducers of extracellular cues emanating from within the tumor microenvironment. The β-adrenergic receptors are a class of G-protein-coupled receptors (GPCRs) present in human tissues that respond to stress hormones and regulate numerous downstream pathways involved in health and disease. While normally mediating physiological stress responses [13], activation of adrenergic receptors results in profound changes to energy utilization [14,15] and many tumors evolve to exploit these adaptive pathways [13,16]. The cAMP-regulated transcriptional coactivators (CRTCs) are a newly discovered family of intracellular signaling integrators that regulate anabolic metabolism and form part of the β-adrenergic axis. While CRTC1 was initially identified as a fusion oncogene in mucoepidermoid carcinoma [17], the CRTC1–3 family was independently discovered in unbiased functional screens as the most potent coactivators of the cAMP-response element-binding protein (CREB) transcription factor [18,19]. Over the past decade, the scope of functions for the CRTC family members has expanded dramatically to include roles in lifespan prolongation, biology of learning and memory, regulation of fatty acid and glucose metabolism, as well as roles beyond transcriptional coactivation [20]. In this review, we discuss the physiological roles and functional consequence of the CRTC family in CREB target gene transcription, define the mechanisms governing CRTC deregulation in cancer, and summarize key findings on tumor cell-extrinsic roles and context-specific interactions of CRTCs. A comprehensive understanding of the mechanisms used by CRTCs at various stages of tumor development and progression may lead to novel cancer therapies that also aid in restoring systemic homeostasis.

CRTCs: Form and Function of a Newly Discovered Family of Signal Integrators

CRTCs are a newly discovered, conserved family of cell signaling integrators [18,21]. There are three CRTCs (CRTC1–3) in humans sharing common domains [22], while Caenorhabditis elegans and Drosophila melanogaster have only one functional CRTC homolog each, termed CRTC-1 and crtc respectively [22,23]. Structurally, CRTCs contain a conserved N-terminal coiled coil that forms the CREB-binding domain (CBD), a central regulatory domain (RD), a serine/glutamine-rich splicing domain (SD), and a C-terminal transactivation domain (TAD) [22,24] (Figure 1, left). An LKB1/AMPK signaling axis regulates RD phosphorylation, which mediates CRTC subcellular localization and activation state in response to converging signaling cascades [25,26]. For example, canonical RD phosphoregulation at CRTC2 S70, S171, and S274 by members of the AMPK family of Ser/Thr kinases, or these conserved residues in CRTC1 and CRTC3, are inactivated by promoting cytoplasmic sequestration through interactions with 14-3-3 proteins in most cell types investigated to date [27]. Activation of cAMP and/or Ca²⁺ signaling activates PKA and calcineurin, respectively, leading to CRTC dephosphorylation, release from 14-3-3 proteins, and nuclear localization where CRTCs interact with NONO/p54^nrb and bind to CBP/p-CREB complexes to mediate transcription-dependent and transcription-independent processes [20,22,23,28]. However, these phosphoregulatory modifications can occur at alternative Ser/Thr residues present within specific CRTC family members, while yet other regulatory post-translational modifications (PTMs) such as acetylation, methylation, O-glycosylation, and ubiquitination may also occur at additional residues (Figure 1, right) [26,27,29–38]. Moreover, these PTM codes may occur in a context-dependent manner such that specific PTMs and cytosolic–nuclear patterns may be modified or even inversed in different tissues. For example, CRTC2 is dephosphorylated in the basal, fed state of neuronal hypothalamic nuclei (arcuate, ventromedial, and paraventricular) and hormonal cues released during fasting induce phosphorylation and cytoplasmic translocation [39]. This subcellular relocalization pattern is opposite to that seen in other cell types including hippocampal neurons [40–42], hepatocytes [25,43,44], pancreatic β islets [45], and skeletal muscle where CRTCs shuttle from the cytoplasm to the nucleus when stimulated [46]. Finally, in contrast to AMPK/SIK/MARK-mediated phosphorylation and inactivation, noncanonical MEKK1-mediated phosphorylation of the RD has been reported to activate CRTC1 by induction of nuclear translocation [47]. GSK-3 Ser/Thr kinase functions have been shown to promote CRTC2 interactions with CREB that may also occur via noncanonical RD phosphorylation, although the biochemical mechanism for this induced association and effect on CRTC subcellular localization remains elusive [48]. Thus, the regulatory networks governing CRTC localization and activity can vary, such that similar inputs induce different responses depending on tissue and cell-type context.

Figure 1. Post-translational Modifications Regulate CRTC Function.

Schematic of the CRTC1–3 family protein structures with known functional domains denoted with the locations of key regulatory PTMs (left). Summary of the types of PTMs, amino acid residue position, enzyme responsible for the PTM, and the effect of each PTM on CRTC13 function (right). Abbreviations: A, acetylation (grey), CBD, CREB binding domain; CRTC, cAMP-regulated transcriptional coactivator; M, methylation (red); P, phosphorylation (blue); PTM, post-translational modification; RD, regulatory domain; SD, serine/glutamine-rich splicing domain; TAD, transactivation domain; U, ubiquitination (green).

CRTCs in Cell Biology and Physiology: From Modulators of CREB Transcription to Multimodal Signaling Regulators

CRTCs were initially named transducers of regulated CREB activity (TORCs) [18,21]. This term reflects the initial conceptualized spectrum of CRTC activities in normal hepatocyte and pancreatic β-islet cell biology and physiology, which is mediated by CREB and CREB-regulated target gene transcription. However, nearly a decade and a half following their discovery, our understanding of these CRTC/CREB-affiliated networks has increased and the scope of their physiological role and interactions has expanded beyond these tissues and CREB, respectively (Table 1). A growing list of CREB-dependent functions of CRTCs has revealed that CRTC/CREB signaling is part of a broader network that regulates global energy homeostasis [25,30,36,37,43–45,49–53]. In hepatocytes, several studies have reported that CRTC2 is a downstream target of the insulin and glucagon pathways where fasting-induced glucagon release activates CRTC2 and promotes gluconeogenesis and insulin resistance in diabetes, while insulin inactivates CRTC2 by promoting phosphorylation and ubiquitin-mediated degradation [30,36,37,52]. CRTCs role in maintaining glucose homeostasis, however, is context-dependent as CRTC2 in pancreatic β-cells promotes insulin secretion [45], and CRTC2 and CRTC3 activity in macrophages mediates PGE2/cAMP-induced M2 differentiation, which is associated with increased insulin sensitivity [51], while CRTC2 activation in adipocytes mediates insulin resistance [54]. Systemic glucose homeostasis is also indirectly maintained by CRTC-mediated regulation of skeletal muscle functions. Specifically, CRTC2 and CRTC3 direct anabolic adaptations to high intensity exercise-induced release of catecholamines that activate adrenergic signaling and promote increased glucose mobilization into glycogen, in addition to triglyceride storage [46]. Activation of CRTC1–3 upregulates PGC1α expression in muscle cells to increase mitochondrial oxidative capacity and consequently CRTC2 inactivation is associated with PGC1α-dependent muscle atrophy [55,56]. Neural CRTC2 has also been reported to mediate mitochondrial fragmentation and suppress the positive lifespan-related effects downstream of AMPK signaling in a distal manner on muscle cells in C. elegans [57]. CRTCs also appear to play different roles in regulating adipogenesis and lipid homeostasis. Both CRTC2 and CRTC3 promote adipogenesis in an overlapping manner [58]. In brown adipose tissue, thermogenesis is facilitated by LKB1/AMPK-mediated CRTC3 inactivation [59], while CRTC2 activity is associated in hepatocytes with lipid droplet accumulation [52], cholesterol synthesis [60], and lipophagy under nutrient-deprived states [61]. Besides mammals, the fly crtc homolog (D. melanogaster) and its upstream modulator SIK regulate lipid and glucose homeostasis in neural cells [53]. Lastly, CRTC2 and CRTC3 induce SOCS3 expression, which disrupts JAK-mediated STAT3 phosphorylation in bone marrow (BM) hematopoietic cells and subsequently decreases G-CSF expression [62]. Thus, CRTCs regulate BM homeostasis through modulating cytokine expression in BM stroma under the PGE2-R/PKA pathway [62].

Table 1.

CRTC Family Members in Normal Physiology

Isoform	Tissue/cell type	Process	Refs
CRTC1	Salivary	Regulates embryonic salivary gland morphogenesis	[133]
	Muscle	Induces PGC1α transcription to promote mitochondrial biogenesis	[56]
	Brain	Mitochondrial metabolism and longevity	[57,122]
		Senses synaptic activity to mediate neuromodulation	[40]
		Regulates formation of contextual long-term memory	[41]
		Strengthens new and reactivated memories	[42]
		Facilitates long-term memory formation during fasting	[68]
		Induces per1 to regulate the suprachiasmatic circadian clock	[69]
	Epithelial/HeLa, RKE3	Associates with AP-1 to promote proliferation	[63]
CRTC2	Lymphoid, hematopoietic/T366A, PBL, LCL, RAMOS, NALM6	Induces differentiation and decreases proliferation	[51,67,134,135]
		Controls bone marrow homeostasis	[62]
		Regulates MMR genes to preserve genomic integrity	[64]
	Pancreas/MIN6, HIT, INS1	Regulates islet b-cell proliferation and function	[27,45]
		Induces Bcl2 to inhibit islet b-cell apoptosis	[66]
	Liver	Mediates glucose homeostasis by regulating gluconeogenesis	[25,29,30,43,49,54]
		Associates with ATF6 to regulate ER stress and gluconeogenesis	[43]
		Represses gluconeogenesis by enhancing insulin signaling	[63]
		Induces autophagy gene targets to regulate lipophagy	[61]
		Facilitates cholesterol biosynthesis by increasing SREBP2	[60]
		Regulates lipogenesis and insulin sensibility by disrupting SREBP1	[36]
		Mediates insulin resistance and lipid homeostasis	[44,52,54]
		Mediates infection-induced hypoglycemia	[50]
	Adipose, preadipocytes/3T3-L1	Regulates lipid homeostasis	[54]
		Interacts with STATs to activate Cebpd and promote adipogenesis	[58]
	Muscle	Induces PGC1α transcription to promote mitochondrial biogenesis	[46,56]
		Mediates myofiber hypertrophy and anabolism	[46]
	Melanocytes	Induces MITF to regulate UV-dependent pigmentation	[72]
CRTC3	Hematopoietic/RAW264.7, THP-1	Controls bone marrow homeostasis	[62]
		Induces KLF4 to mediate M2 macrophage polarization	[51]
		Induces differentiation and decreases proliferation	[38]
	Adipose, preadipocytes/3T3-L1	Interacts with STATs to activate Cebpd and promote adipogenesis	[58]
		Regulates thermogenesis	[49,59]
	Skin	Induces MITF in melanocytes to regulate UV-dependent pigmentation	[72]
	Intestine	Promote ISC proliferation	[65]
Crtca	Brain	Regulates neuronal energy homeostasis	[53]
		Facilitate long-term memory formation during fasting	[68]
		Regulates per and tim expression to regulate circadian rhythm	[70]

^aThis section refers to work performed with Drosophila and/or C. elegans model systems.

In addition to these CREB-dependent functions, CREB-independent interactions also occur where CRTCs serve as coactivators of other basic-region leucine zipper (bZIP) transcription factors including AP-1 (c-Jun/c-Fos) and activating transcription factor (ATF)-6 [43,63]. Hepatic endoplasmic reticulum (ER) stress promotes CRTC2 dephosphorylation and interactions with proteolytically cleaved, nuclear ATF-6, thereby stimulating ER-mediated cell viability and glucose homeostasis [43]. All three CRTC family members and their homolog in D. melanogaster (crtc) can function to integrate different signaling cues that regulate critical aspects of the cell cycle including DNA damage response [64], proliferation [45,65], and apoptosis [66]. The extent to which CRTCs bind to other bZIP family members relative to CREB is unclear, however. With respect to specific CRTC members, activation and association of CRTC1 with AP-1 promotes cell proliferation by increasing the number of cycling cells [63] while CRTC2 activation within intestinal stem cells, pancreatic β-islet cells, and tonsillar lymphoid tissue promotes proliferation and blocks apoptosis by inducing expression of the antiapoptotic gene BCL2 [45,66,67]. GSK-3α or GSK-3β regulation of MEIS1-dependent CREB/CRTC2 interactions induces HOX-mediated transcription in myeloid progenitors and sustains cell proliferation [48]. CRTC2 also induces expression of several key mismatch repair genes (EXO1, MSH6, PMS1, and POLD2) to preserve DNA integrity in response to single-base DNA mismatches that occur during DNA replication [64]. Notably, these studies were performed in HeLa cells, where LKB1 is mutated, leading to the dephosphorylation and constitutive activation of CRTCs. Future studies should reveal whether the role of CRTC2 in the DNA damage response is conserved in other cell contexts.

Aside from transcriptional regulation of the cell cycle, cell survival, and metabolism genes, CRTCs have also been associated with regulating other cellular processes in neural cells [40–42,57,68–70], melanocytes [71,72], and macrophages [38,51]. More specifically, in neural tissue CREB/CRTC1 complexes have been identified as mediators of neural activity and participate in long-term memory formation processes both in mammals and D. melanogaster [40–42,68]. Also, CREB/CRTC-mediated gene transcription regulates the circadian clock with similar mechanisms in mammals and Drosophila (CREB/CRTC1 targets period1 and CREB/crtc targets per and tim) [69,70]. In melanocytes, CRTC2 and CRTC3 upregulate MITF, which promotes melanin synthesis [71,72]. Adiponectin inhibits CRTC via AMPK activation and mediates depigmentation [72]. In macrophages, they are part of the network that regulates differentiation and phenotype interconversion, which affects insulin sensitivity and inflammation [38,51]. PGE2 signaling promotes regulatory macrophage (M2) polarization through CREB/CRTC2/3 complexes that target KLF4 [51]. SIKs, which are known CRTC inhibitors, have been associated with CRTC3 inactivation in macrophages and attenuate M2 differentiation [38].

Despite these CREB-dependent and CREB-independent transcriptional coactivator roles as robust regulators of adaptive stress-response genes, additional transcriptional coactivator-independent functions have been described for CRTCs, which include directing alternative splicing of selected CREB-target genes [20,36]. Context-specific variables including cell type and TATA versus TATA-less promoter architecture have been identified as determinants of CRTC effects on alternative exon skipping and alternative 3′ splice-site selection, and these mechanisms are disrupted in pathophysiological states. For example, the splicing domain present in full-length endogenous CRTC1 is deleted in the CRTC1/MAML2 fusion oncogene and this chimera consequently displays defects in pre-mRNA processing, despite retaining the ability to bind CREB and robustly activate transcription. Additionally, since CRTCs are devoid of conserved RNA-binding domains common to splicing factors, the CRTC splicing domain may serve as a scaffold for the assembly of factors involved in pre-mRNA factors and the differential recruitment in specific cell contexts may influence growth factor receptor, oncogene, and/or tumor suppressor gene functions [73]. Thus, CRTCs should be considered crucial mediators of several signaling pathways that influence many fundamental biological processes relevant for tumor initiation, development, and progression [48,63].

Oncogenic and Tumor-Suppressor Functions of CRTCs in Human Cancers

A growing body of studies has implicated CRTCs in signaling networks that regulate cell growth and differentiation, proliferation, survival, DNA damage repair, and apoptosis. Therefore, it is no surprise that multiple links between aberrant activation of CRTCs and carcinogenesis have begun to emerge in association with a growing list of different cancers (Table 2). While the mode of CRTC activation may vary, the primary mechanism appears to involve downregulation or inactivation of the upstream negative regulator LKB1 or the direct upregulation of CRTC1–3 mRNAs, although additional somatic alterations are observed within known regulatory residues of the CRTCs at variable frequencies in specific cancers (Figure 2). At present, direct relationships have been identified between activated CRTCs and acquisition of biological capabilities of malignant cells, known as the hallmarks of cancer [8] (Figure 3). These links include: proliferation [63,67,74–84], cell death resistance [66,75,77,78,81,85–87], replicative immortality [63,67,74–84], invasion and metastasis [75,76,79,83,88–91], growth suppression evasion, and genome instability and mutation [64]. Besides the hallmarks of cancer, CRTCs have been also associated with acquisition of chemoresistance [81,92]. Since CRTC nuclear localization and transcriptional coactivation are context dependent in different cell types [45,51,54], more studies are needed to define their role either as oncogenes or as tumor suppressors in different cancers [79,86].

Table 2.

Mechanisms of CRTC Deregulation in Cancer

Isoform	Cancer type	Mechanism of regulation	Refs
CRTC1	Mucoepidermoid carcinoma	t(11;19) translocation generates a CRTC1/MAML2 fusion protein	[93]
	Lung cancer	Loss of LKB1 results in SIK inactivation and hypophosphorylation of CRTC1	[77,79,88]
	Lung cancer	Upregulated LINC00473 interacts with NONO and enhances CRTC1/CREB transcriptional activity	[77]
	Esophageal squamous cell carcinoma	Loss of LKB1 results in SIK inactivation and hypophosphorylation of CRTC1	[90]
	Esophageal adenocarcinoma	SNP	[116,117]
	Acute myeloid leukemia	Loss of miR-22 leads to de-repression of CRTC1 expression	[118]
	Colorectal cancer	Increased protein stability	[81]
	Colorectal cancer	PGE-2 secretion results in Calcineurin and PKA activation and dephosphorylation of CRTC1	[81]
	Melanoma	UV irradiation causes a decrease in SIK2 expression and hypophosphorylation of CRTC1	[136]
CRTC2	Lymphoma	Decreased expression due to epigenetic silencing	[64]
	Multiple cancer types	Gene mutation	[64]
	Breast cancer	Leptin and PGE-2 secretion lead to AMPK-dependent hypophosphorylation of CRTC2	[84]
	Lung cancer	Upregulated MetaLnc9 interacts with NONO and enhances CRTC2/CREB transcriptional activity	[91]
	T cell lymphoma	HTLV-1 protein Tax interacts with CRTC2/CREB	[82]
	B cell lymphoma	Failed CRTC2 inactivation caused by defective DNA damage response	[67]
	Pancreatic cancer	Activating mutations in GNAS cause SIK inactivation and dephosphorylation of CRTC2	[131]
CRTC3	Mucoepidermoid carcinoma	t(11;15) translocation generates a CRTC3/MAML2 fusion protein	[96]

Figure 2. Catalog of CRTC Aberrations in Cancer.

Overview of The Cancer Genome Atlas (TCGA) PanCan data summarizing the frequencies and distribution of somatic mutations, gene fusions, amplifications, deletions, and mRNA expression in CRTC1–3 across a subset of cancers. Mutations in amino acid residues that have known regulatory roles are depicted with large blue circles in the lollipop plots in contrast to mutations with unknown roles that are depicted with small red circles. Abbreviations: CBD, CREB binding domain; CRTC, cAMP-regulated transcriptional coactivator; RD, regulatory domain; SD, serine/glutamine-rich splicing domain; TAD, transactivation domain.

Figure 3. Hallmarks of the CRTC Adaptive Stress Regulators and Cancer.

The impact of CRTCs, on the next generation of hallmarks of cancer, are annotated to highlight specific examples of directly regulated target genes that are involved in tumorigenesis and progression. Each concentric ring moving out from the CRTC core provides information regarding the functional role as either an oncogene or tumor suppressor, the specific CRTC family member involved, and the type of cancer reported. The outer ring represents the cancer hallmarks specifically controlled by the CRTCs. Abbreviations: Ca, carcinoma; CRTC, cAMP-regulated transcriptional coactivator; ESCC, esophageal squamous cell carcinoma; MEC, mucoepidermoid carcinoma; NR, not reported; NSCLC, non-small cell lung cancer; OG, oncogene; TS, tumor suppressor.

Mucoepidermoid Carcinoma

Mucoepidermoid carcinoma (MEC) is the most common salivary gland malignancy. A recurrent t(11;19) translocation creates a CRTC1/MAML2 chimeric protein (C1/M2; mucoepidermoid carcinoma translocated 1, MECT1/MAML2) and is the major driver genetic alteration in over 50% of all MECs [93]. This translocation is considered pathognomonic of MEC but has also been identified in mucoepidermoid-like tumors arising as primary tumors from glandular epithelium scattered throughout the human body [94,95]. A similar CRTC3/MAML2 (C3/M2) fusion, resulting from a t(11;15) translocation, has also been detected in a mutually exclusive pattern from C1/M2, confirming the role of activated CRTC biology in tumorigenesis [96].

C1/M2 functions have been associated with several aspects of cancer biology both in vitro and in patient-derived xenografts (PDXs) [75,85]. The fusion comprises CRTC1 fused to the NOTCH transcriptional coactivator MAML2, which initially suggested important roles for both NOTCH and CREB pathways in MEC. However, translocation disrupts MAML2 signaling to NOTCH by replacing the NOTCH binding domain with the CREB binding domain of CRTC1. While the role NOTCH plays in MEC is unclear, the same C-terminal exons of MAML2 or MAML3 have been detected in gene fusions of non-MEC tumors [97–100], suggesting that a primary role for these sequences is to provide a transcriptional activation domain and disrupt NOTCH signaling. A recent structure–function study suggests that CRTC1/3–MAML2 allows binding of CBP to CREB independently of phosphorylation in S133 [101]. These results are in sharp contrast with previous studies [102,103] that have shown that phosphorylation of CREB in S133 is necessary for the formation of the CREB/CRTC/CBP complex and CREB-mediated transcription. We and others have shown that C1/M2 functions as a constitutively active CREB coactivator that, unlike forced expression of CRTC1 alone, is sufficient to induce transformation pointing to novel properties, including aberrant regulation of pre-mRNA splicing and deregulation of MYC target genes [78,101]. These oncogenic properties result from absence of the splicing domain in the C1/M2 translocation product and direct C1/M2–MYC interactions that manifest via gain-of-function properties inherent to the C1/M2 fusion protein, respectively. Combined with CREB, the target genes regulated by interactions with MYC are predictive of several malignancies with CREB and MYC involvement [78]. In addition to CREB and MYC, C1/M2 binds to AP-1 to promote malignant transformation and is associated with increased proliferation, invasion, migration, and metastasis [63,75,76,101]. Collectively, AREG, NR4A2/Nurr1, C4.4a, MMP10, NR4A3, and the long non-coding RNA (lncRNA) LINC00473 are downstream targets upregulated by C1/M2 that are necessary for MEC growth and maintenance [75,80,85] (Table 3).

Table 3.

Role of Specific CRTC1-3 Interactions and Transcriptional Targets in Cancer

Isoform	Cancer type	Interaction partner or induced target	Process	Refs
CRTC1	Mucoepidermoid carcinoma	c-Fos (AP-1; c-Jun:c-Fos)	↑Proliferation; malignant transformation	[63]
		LINC00473	↑Survival; ↑proliferation	[77]
		CBP/p300	↑CREB-regulated target transcription	[101]
		PCK1; AREG; C4.4A; MMP10; NR4A3/NOR1*	↑Proliferation; ↑invasion; ↑migration	[76]
		LINC00473; DMBT1; STC1; PDE4B; RUNX3; PTGS1; PDE4D; CA9; ODC1; TGFB2; AREG; TYMS; CDK6	↑Survival; ↑proliferation; ↑invasion; ↑migration	[75]
		MYC and CREB target genes	↑CREB- and MYC-regulated target transcription ↑Survival; malignant transformation; ↑tumor growth	[78]
	Lung cancer	LINC00473	↑Survival	[77]
		PTGS2 (COX2)	↑Tumor growth; ↑migration;	[79]
		NR4A2/Nurr1	↑Tumor growth	[80]
		NEDD9	↑Invasion; ↑metastasis	[88]
	Esophageal squamous cell carcinoma	LYPD3	↑Invasion; ↑migration	[90]
	Mixed lineage leukemia	c-Fos	HOX mediated tumorigenesis	[48]
	Colorectal cancer	NR4A2, COX2, AREG, IL6	↑Viability; ↑proliferation; ↓chemosensitivity	[81]
CRTC2	Lung cancer	c-Jun	EMT (E-cadherin, vimentin); Invasion, migration (MMP-2,-9)	[89]
		MetaLnc9 (LINC00963)	↑Invasion; ↑migration	[91]
	T-cell lymphoma	EXO1, MSH6, PMS1, POLD2	↓Mutation rate; ↓microsatellite instability	[64]
		CCND1 (CyclinD1)	↑Proliferation	[82]
	B-cell lymphoma	TCL1	↑Proliferation	[67]
		TCL1	↓Apoptosis	[87]
		BIM	↓Differentiation; ↓apoptosis	[86]
	Breast cancer	AROM (Aromatase)	↑Proliferation	[84]
	Ovarian, Prostate, Breast cancer	BRCP	↑Chemoresistance	[92]
	Not reported	Bcl-2	↑Proliferation; ↓apoptosis	[66]
CRTC3	Mucoepidermoid carcinoma	CBP/p300	↑CREB-regulated transcription	[38]

In the past, C1/M2-positive MECs have been associated with increased patient survival and improved clinicopathological outcomes [75,78,85,104]. However, more recent studies have failed to confirm these associations [105,106] and it is likely that most C1/M2- or C3/M2-negative MEC cases are misdiagnoses that rather represent high grade adenosquamous cell carcinomas with distinct pathobiology and genetics [104,107]. It is now becoming evident that C1/M2 and C3/M2 status is a reliable diagnostic marker for MEC; however, prognostic data depend on the accumulation of additional somatic mutational events such as loss of CDKN2A/p16 orTP53 expression [108].

Lung Cancer

Lung cancer is one of the most common types of cancer in both men and women. LKB1 (STK11) is an evolutionarily conserved master regulator of AMPK members from yeast to humans [109], and germline transmission of an LKB1 null allele results in familial Peutz–Jeghers syndrome, which is associated with a variety of spontaneous benign and malignant phenotypes including adult onset adenocarcinoma [110–112]. Consequently, LKB1 is also one of the most common targets for somatic mutational inactivation in non-small cell lung carcinoma (NSCLC), and LBK1-null or LBK1-deficient tumor cells exhibit aberrant CRTC1 hypophosphorylation and cotranscriptional activation of CREB targets [80,113,114]. For example, CRTC1 induces expression of downstream target genes such as NR4A2, NEDD9, PTGS2 (COX-2), and LINC00473 in LKB1-null or -deficient NSCLC cells, which contribute to cell growth and survival and tumor development [77,79,80,88] (Table 3). In a positive feedback loop mechanism, LINC00473 promotes CRTC1/CREB/NONO complex assembly and enhances expression of CREB-regulated target genes [77]. Finally, bioinformatics analyses of the LKB1-null mRNA gene signature in 53 annotated lung cancer cell lines were correlated with drug compounds using the Broad Institute Connectivity Map (CMAP) analysis portal. Remarkably, all 17 top scoring compounds such as colforsin (a water-soluble forskolin analog) are known CRTC activators and include six different β-adrenergic analogs and six different PGE analogs [79]. These data further suggest that CRTC activation is a major determinant of the LKB1 null signature in lung cancer.

Besides CRTC1, there are data from in vitro and sequencing studies of human samples that indicate the involvement of CRTC2 in lung cancer [89,115], but a complete understanding of its role in lung pathology is missing. Similar to CRTC1, CRTC2 forms complexes with the lncRNA LINC00963 (MetaLnc9) via CREB/NONO interactions, thereby establishing a positive feedback loop to promote migration and invasion [91] (Table 3). CRTC2 also modulates c-Jun activity and induces MMP2, MMP9, E-cadherin, and vimentin expression to promote epithelial–mesenchymal transformation, migration, and invasion [89].

Esophageal Cancer

Esophageal cancers are a common type of aerodigestive cancer that include esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC). Two large genome-wide association studies of SNPs identified CRTC1 as a susceptibility locus for EAC and Barrett’s esophagus [116,117]. In addition, in ESCC, LKB1 loss leads to constitutive CRTC1 activity, induction of LYPD3 gene expression, migration, and invasion [90] (Table 3).

Hematopoietic Malignancies

CRTC proteins can act either as oncogenes or tumor suppressors in different hematopoietic malignancies. In acute leukemia driven by chimeric mixed lineage leukemia (MLL) oncoproteins, GSK-3 promotes association of CREB and its coactivators CBP/p300 and CRTCs with the homeodomain protein MEIS1. These interactions form a transcriptional complex that in turn facilitates HOX-mediated transcription and maintenance of leukemia stem cells [48]. CRTC activity can also be regulated by miRNAs. For example, loss of mir-22 in acute myeloid leukemia (AML) by epigenetic silencing leads to derepression of CRTC1 expression, activation of CREB signaling, and disease progression [118]. In contrast, the role of CRTC proteins in B cell lymphomas is less clear. In B cell lymphomas, double-strand breaks activate LKB1 and the serine/threonine kinase ATM, leading to CRTC2 inactivation and decreased TCL1 expression, which results in proliferation arrest [67] (Table 3). Likewise, TCL1 expression is also down-regulated by CD40 and BCR-mediated inactivation of CRTC2, which results in increased apoptosis. In contrast, 11q23 gain/amplification leads to increased SIK2 expression and activity that inhibits CRTC2 and CREB signaling, decreases BIM expression, and promotes cell survival [86]. These seemingly contradictory findings could be explained under the prism of CRTCs context-dependent interactions. For example, CRTC2 oncogenic functions in germinal center B cell lines versus its tumor-suppressive functions in activated B cell lines suggests that different lymphoid progenitors may be responsible for these varied CRTC roles.

In T cell lymphomas, CRTC2 has been described to synergize with CREB and CBP to promote expression of DNA mismatch repair genes, and thus its inhibition is associated with increased mutation rate and microsatellite instability [64]. However, despite this profound tumor-suppressor role of CRTC2 in T cell lymphomas, the viral oncoprotein Tax interacts with CRTC2/CREB to upregulate CCND1 (cyclin D1) and increase proliferation in HTLV-1 associated T cell lymphomas [82] (Table 3). Previous studies have demonstrated that Tax directly interacts with CRTC1–3 to assemble hybrid host-viral transcriptional complexes that activate HTLV-1 transcription [119,120]. Thus, the role of CRTC2 in different T cell lymphoma subtypes appears to be context dependent but it is tempting to speculate that with HTLV-1 infection, viral Tax may be responsible for directing the switch from CRTC tumor suppressor function to one that is oncogenic by establishing a different intracellular environment conducive to oncogenic CRTC functions.

Other Cancers

The functional roles of CRTC in cancer are constantly expanding. Individual studies report emerging roles of CRTCs in colon [81], breast [84,121], and prostate [83,92] cancer. In colon cancer, COX-2/PGE-2 signaling activates calcineurin and PKA, inducing CRTC1 dephosphorylation, thus inducing AP-1 and CREB interactions that upregulate NR4A2, PTGS2, AREG, and IL-6 gene expression to promote cell viability, proliferation, and chemoresistance [81] (Table 3). In breast cancer, activation of CRTC2 in breast adipose stromal cells provides a link between obesity and breast cancer development. Treatment of adipose stromal cells with leptin leads to AMPK-dependent hypophosphorylation of CRTC2 and increased aromatase expression [84]. Additionally, CRTC2 has also been identified as a target of the RAS/MAPK, PI3IK/AKX, and PKA pathways in breast, ovarian, and prostate cancer. In these malignancies, CRTC2 induces BRCP gene expression and promotes chemoresistance [92]. In prostate cancer, CRTC2 has also been associated with IGF-IR expression and subsequently with proliferation and invasion. Treatment with metformin causes inactivation of CRTC signaling, leading to downregulation of IGF-IR expression and decreased IGF-I sensitivity [83].

Concluding Remarks

CRTCs constitute a new family of signal integrators that normally function to restore homeostasis by responding to metabolic stresses but have also emerged as key targets for aberrant expression or regulation of activity in a growing number of tumor subtypes [122]. Current evidence suggests that CRTC activity is affected in both health [45,51,54] and disease [64,67,82,86,87] by cell type and possibly exogenous viral infections; therefore, the context/tissue-specific roles of CRTCs among cell types and tissues create a challenge that remains unresolved regarding their role in different cancers (see Outstanding Questions). Moreover, the role of CRTCs in various tumor subtypes may be influenced by differential assembly of cotranscriptional complex-binding partners in different cell lineages and by accumulation of distinct somatic mutational landscapes such as LKB1 status. However, it remains unclear whether somatic mutations within CRTC regulatory amino acid residues normally targeted by upstream negative regulators such as AMPK/SIK/MARK also contribute to aberrant CRTC activity in cancer (Figure 2). In addition, although CRTCs can bind and coactivate the transcription factor CREB, which has known roles in melanoma [123], non-small cell lung cancer [124], AML [48,125], and glioblastoma [126], it remains to be determined whether interactions with other transcriptional cofactors and/or related bZIP transcription factors such as the ATF family influence specific functional roles for CRTC as either an oncogene or tumor suppressor. Given the potential impact of assembling context-dependent transcriptional complexes, it will also be important to determine the extent to which viral proteins influence CRTC function, such as HTLV-1 Tax oncoprotein in T cell lymphoma, and if such context-specific Tax–CRTC2 interactions are responsible for the observed switch from a tumor suppressor to an oncogene within infected T cells.

Outstanding Questions.

How do interactions between CRTCs and specific transcription factors and/or cofactors influence transcriptional and/or post-transcriptional activities in different cancers?

What are the context/tissue-specific factors that direct CRTCs functional roles as oncogenes or tumor suppressors?

Do the CRTC1–3 isoforms each have equivalent oncogenic or tumor suppressive potential or are there novel activities and interactions encoded by specific isoforms?

What are the specific driver versus passenger mutations that lead to aberrant CRTC activity in cancers and do differences in mode of activation generate distinct pathobiology?

Can distal rewiring of CRTC activity establish tumor cell-extrinsic activities that support tumor initiation, development, and/or progression?

What is the role of chronic CRTC activation in epidemiology data that link obesity and stress signaling with cancer etiology?

Anabolic metabolism adaptations to the demands of increased proliferative cellular signals are constantly impacted by interactions with both surrounding and distant tissues and cell populations. In this regard, our understanding of the role that tumor microenvironments and distant organ systems play have evolved from that of passive bystanders to active, deregulated niches that crosstalk with cancer cells and modulate their phenotype [127,128]. For example, rewiring and modulation of insulin, glucose, and lipid metabolic pathways, within otherwise normal, nontumorigenic hepatic cells have been detected in patients with advanced lung cancer [129]. These events are proposed to be mediated through the STAT3/Socs3 pathway whose downstream effectors include AMPK, which reprograms energy sensing and stress responses and also serves as the essential upstream negative regulator of CRTC family members. To elucidate the full potential of CRTCs in cancer biology we should explore their potential roles in the distal regulation of metabolism within tissues, organs, and organ systems.

The CRTCs are validated cancer genes since constitutive activation of CRTC1 and CRTC3 by chromosomal translocation underlies the etiology of MEC, and tumor maintenance depends on sustained expression and nuclear localization of the activated CRTC fusion partner [130]. There is now evidence for alternate, non-translocation-based mechanisms of CRTC activation in a growing number of common adult cancers including the important subset of LKB1 null lung cancers. In addition, there are other mutational events that deregulate AMPK/cAMP signaling. For example, mutation of GNAS drives pancreatic tumorigenesis by inducing PKA-mediated AMK/SIK suppression, leading to CRTC and HDAC activation and reprogramming of lipid metabolism in pancreatic tumor cells [131]. The collection of these tumor cell-extrinsic and context-specific roles for CRTCs, whose understanding is limited at present but holds significant translational potential, could offer opportunities for novel clinically actionable approaches [83,132].

Highlights.

CRTCs are key integrators of adaptive stress signaling in health and disease and their spectrum of activities is context dependent.

CRTCs regulate CREB-dependent and CREB-independent transcriptional responses that impact the Hallmarks of Cancer.

Somatic mutations of CRTC1–3 are frequently observed in a diverse panel of neoplasms.

Context-specific interactions determined by the cell of origin and extrinsic signals define the oncogenic versus tumor-suppressive role of CRTCs in different cancers.

The plasticity of CRTC interactions and their importance in cell and cancer biology may offer novel opportunities to exploit clinically.