FormalPara Key Points

MYC deregulation occurs in a large number of tumors across multiple tissue types, making this oncogenic master transcription factor a highly desirable therapeutic target.

Genetic models indicate that MYC inhibition may be well-tolerated and lead to sustainable tumor regression.

Despite lacking targetable structural domains, several novel therapeutic strategies have emerged in an attempt to inhibit MYC activity clinically, including inhibition of transcriptional and post-translational regulatory events.

1 Introduction

In an attempt to move away from toxic and non-specific chemotherapeutic agents, a global effort to develop targeted therapeutic strategies to inhibit oncogenic drivers has dominated the cancer biology field. By interrogating tumor cells at the DNA, RNA, and protein level, we have been able to identify specific cancers or cancer subtypes where a significant percentage of patients express a dominant oncogenic driver. In these cases, researchers have shown that the loss of this dominant driver leads to tumor cell death, and multiple targeted therapeutic agents based on this principle have shown great clinical success. For example, the BCR/ABL1 inhibitor Gleevac® has increased the 8-year survival of patients with chronic-phase chronic myeloid leukemia (CML) from 6 to ~ 90% and represents one of the most successful targeted kinase inhibitors to date [1]. Similarly, HER2 (human epidermal growth factor receptor 2) overexpression or amplification has been shown to occur in ~ 20% of breast cancer patients and anti-HER2 therapies such as trastuzumab and lapatinib have significantly increased patient survival in this subset of patients [2]. These clinical successes have helped fuel translational studies and highlight the potential of utilizing targeted therapies in the clinic. Unfortunately, a large number of tumors are driven by a small number of common oncogenic proteins that lack structural regions amenable to therapeutic inhibition [3, 4]. Prototypic examples of this are KRAS and MYC, where mutational activation and deregulated oncogenic expression are common driver events in cancer progression in many tissues and, therefore, these oncoproteins are considered highly desirable therapeutic targets [3, 5,6,7]. However, despite their significant contribution to disease states, these factors are commonly thought to be ‘undruggable’. The generation of therapeutic compounds that could effectively target these drivers would significantly alter the clinical outcome of an extraordinary number of patients. Here we review the biology of MYC deregulation in cancer that supports innovative strategies for therapeutic targeting and the potential for translating these strategies to the clinic.

1.1 MYC Deregulation in Cancer

The MYC transcription factor family consists of c-, L-, and N-MYC. The aberrant expression or activity of any one of these family members has been shown to contribute to tumor development, although the latter two seem to be restricted to specific tissues, most prominently lung and neural, respectively [8,9,10,11]. MYC family proteins function as potent transcription factors that regulate multiple cellular processes, including proliferation, differentiation, adhesion, and survival [9, 10, 12]. Several studies have demonstrated that MYC functions as a master transcriptional regulator, binding to the majority of regulated genes in the genome [10, 13, 14]. Given the prolific role of MYC in transcriptional regulation, expression of MYC proteins is tightly regulated at the transcriptional, translational, and post-translational levels in normal tissues, with a half-life of ~ 20 min [15, 16]. The major MYC protein domains include an N-terminal transactivation domain (TAD), MYC box domains (MB0-IV), a PEST domain (Proline, glutamic acid [E], Serine and Threonine rich), a nuclear localization sequence (NLS), and the carboxy-terminus basic-helix-loop-helix-leucine zipper (bHLHZ) [17,18,19,20,21]. Each of these domains facilitates interactions between MYC and a diverse set of binding partners in order to regulate MYC function and gene target specificity. The MB0-II domains are essential to MYC protein stability and activity, and facilitate MYC’s association with co-factors, such as PIN1, FBW7, and P-TEFb [19, 22,23,24,25,26]. MBIII and MBIV regulate the apoptotic function of MYC, as well as protein turnover [27,28,29,30]. Finally, the bHLHZ domain facilitates MYC’s interaction with its transcriptional co-factor MYC-associated protein X (MAX), allowing for DNA binding [8, 17, 18, 31]. Although the complex MYC interactome creates unique challenges for the development of MYC-specific inhibitors, each of these functional domains provides potential points of regulation that can be exploited to reduce the oncogenic function of MYC. Since all three MYC family proteins contain homology in these functional domains and their bHLHZ domains, several of the proposed therapeutic agents are likely to function against multiple MYC proteins.

The current dogma regards MYC amplification as the primary method by which MYC is deregulated in disease states. However, the post-translational regulation of MYC has emerged as an important mechanism, irrespective of amplification, by which MYC is stabilized and activated [32,33,34]. Research has identified two interdependent phosphorylation sites that are critical for the regulation of MYC stability and function. While these sites are conserved across MYC family members, we focus here on c-MYC (‘MYC’, unless otherwise specified). Downstream of growth-stimulatory signals, activation of the RAS/MEK/ERK cascade or cyclin-dependent kinases (CDKs) leads to the phosphorylation of MYC at Serine 62 (pS62-MYC) [32, 33, 35, 36]. This modification supports isomerization of Proline 63 in MYC from the trans to cis conformation by the phospho-serine/threonine-directed peptidyl-prolyl isomerase, PIN1, and these events increase MYC DNA binding and target gene regulation. Phosphorylation of Serine 62 (S62) also primes MYC for glycogen synthase kinase 3 (GSK3)-mediated phosphorylation at Threonine 58 (pT58-MYC), which initiates MYC turnover. Dual phosphorylated MYC (pS62/pT58-MYC) then undergoes a second isomerization by PIN1, returning Proline 63 MYC to the trans conformation. This second isomerization event results in the association of MYC with the trans-specific phosphatase Protein Phosphatase 2A (PP2A), which dephosphorylates the stabilizing S62 residue and targets MYC for ubiquitin-mediated proteosomal degradation through the E3 ubiquitin ligase SCFFBW7 [33, 37,38,39,40]. Considering that MYC has a very short half-life, the balance of these phosphorylation and isomerization states provides controlled activity and rapid turnover of the MYC protein, allowing an expedited response to cellular signals while preventing the persistent expression of gene targets in normal cells.

It is now well-appreciated that a high percentage of cancers develop mechanisms to increase MYC activity in order to globally increase cell survival, proliferation, and invasiveness [9]. In disease states, studies have shown that aberrant MYC expression results in promoter invasion, with MYC binding to both high- and low-affinity consensus sequences, altering the expression of a large number of target genes [41, 42]. Consistent with these results, amplified or high Myc expression can drive tumorigenesis in multiple mouse models and MYC amplification is observed to various degrees in almost every human cancer type [11, 43]. Although amplification or overexpression of MYC commonly occurs in cancers, this is not the only mechanism by which MYC is deregulated. In fact, the majority of solid tumors do not display significant MYC amplification [44]. We and others find elevated levels of pS62-MYC and lower levels of pT58-MYC, consistent with a more active and stable form of MYC, in a large percentage of tested human tumors [33, 45,46,47,48,49,50]. Moreover, mutation of the Threonine 58 (T58) residue (MycT58A) results in constitutive S62 phosphorylation and increased tumorigenic potential compared to wild-type MYC [48, 51]. These studies suggest that the post-translational regulation of MYC in cancer may be wildly underestimated and play a significant role in tumor phenotypes. Importantly, in mouse models, low-level constitutive expression of Myc alone does not induce transformation, but rather exacerbates tumorigenic phenotypes when combined with oncogenes such as HER2 and mutant KRAS that can enhance S62 phosphorylation [51, 52]. Conversely, the genetic loss of Myc can prolong survival in aggressive KRAS-driven tumors, highlighting the contribution of endogenous MYC activity to oncogenic signaling pathways and supporting the rationale for therapeutic inhibition of MYC in a large number of cancers [52,53,54,55].

Given that MYC has been implicated in global gene regulation, one would predict that MYC suppression would result in large toxicities, with decreased proliferation and survival in normal cells. Surprisingly, the genetic inhibition of MYC in mice, through switchable transgenes or expression of a dominant negative form called OmoMYC, has resulted in dramatic losses of tumor phenotypes in lung adenocarcinomas, glioblastomas, skin papillomatosis, and pancreatic tumors with little to no toxicities [12, 54, 56,57,58,59]. OmoMYC is a mutated bHLHZ dimerization domain that is able to form OmoMYC homodimers that bind to DNA and compete with endogenous MYC:MAX complexes, reducing MYC promoter occupancy and effectively suppressing transcription. A recent study from Jung et al. [59] demonstrated that under physiologic levels of MYC, expression of recombinant OmoMYC protein minimally suppresses MYC at high-affinity binding sites. In contrast, the oncogenic, low-affinity MYC binding sites are acutely responsive to OmoMYC expression [59]. This study suggests that therapeutic targeting of oncogenic-specific MYC functions may be possible and highlight the importance of understanding the contribution of MYC signaling to oncogenic phenotypes.

2 Therapeutic Strategies to Target MYC

Studies have shown that transcription factors contain intrinsically disordered regions, which allow for the association with high- and low-affinity DNA binding sites and a diversity of co-factors [60, 61]. Additionally, these disordered regions are common sites for post-translational modifications, underscoring the importance of these mechanisms in regulating transcription factor function and stability [62]. Unfortunately, the inherent flexibility of transcription factors makes the direct therapeutic targeting of these proteins difficult, with most strategies relying on disrupting expression, protein–protein interactions, or DNA binding. Here we discuss the innovative strategies (Fig. 1) and therapeutic compounds (Tables 1 and 2) that have been proposed for the therapeutic targeting of MYC, including the inhibition of MYC transcription, partner protein dimerization, activating post-translational modifications, and turnover.

Fig. 1
figure 1

MYC regulatory pathways and therapeutic points of intervention. Transcriptional (top) and post-translational (bottom) mechanisms that regulate MYC function. Gray boxes indicate therapeutic categories and representative compounds that are being explored to negatively impact MYC activity. The pink box indicates the EBOX sequence. Ac acetylation, ATRA all-trans retinoic acid, BET bromodomain and extra-terminal motif, DUB deubiquinating enzyme, MAX MYC-associated protein X, p phosphorylation, PP2A Protein Phosphatase 2A, SMAPs small-molecule activators of Protein Phosphatase 2A, Ub ubiquitination

Table 1 Targeting MYC transcriptional regulation
Table 2 Targeting MYC post-translational regulation

2.1 Inhibition of Transcription

Since MYC lacks a defined targetable structure, the epigenetic silencing of the MYC gene provides an interesting strategy to reduce MYC expression and activity. The challenge with this strategy is identification of compounds that preferentially target the MYC gene. Inhibitors of histone deacetylases, histone methyltransferases, histone demethylases, DNA methyltransferases, and bromodomain and extra-terminal motif (BET) bromodomains have all shown some efficacy against MYC, with BET inhibitors being the most well-studied [63]. The BET family member BRD4 recruits positive transcription elongation factor b (P-TEFb) to promoters and enhancers, releasing RNA polymerase II and initiating transcriptional elongation [64]. JQ1, a BET inhibitor, has been shown to inhibit BRD4 binding at acetylated histones within the MYC promoter and enhancers, decreasing expression of c-, L-, and N-MYC [65,66,67]. JQ1 treatment reduces tumor cell survival and has anti-tumor effects in vitro and in vivo in multiple models [68,69,70,71,72,73]. There are 15 different BET inhibitors being assessed in the clinic; however, clinical responses have been limited, often result in relapse, and are inconsistent with their effects on MYC expression [74, 75]. These results suggest that as a single agent, BET inhibition may not result in durable responses. In support of this, the majority of OTX015/MK-8628 phase I and II clinical trials resulted in disease progression and termination of the trial. Kurimchak et al. [70] have demonstrated that treatment with JQ1 can induce large-scale reprogramming of signaling pathways leading to resistance. However, these resistant cells were highly sensitive to kinase inhibitors, suggesting efficacy in drug combination [70]. Similarly, a new BET inhibitor now in clinical trials, BI 894999, reduces tumor growth in vivo and synergistically induces cell death when combined with CDK9 inhibitors, supporting the use of these epigenetic inhibitors in combination strategies [76]. Indeed, new dual-function compounds are being explored. The dual-activity phosphoinositide 3-kinase (PI3 K)–BRD4 inhibitor SF2523 reduced the in vivo growth of MYCN-amplified neuroblastoma and pancreatic xenografts and decreased distant metastasis [77]. Further, SF2523 shows reduced toxicities compared to the individual combination of the PI3 K inhibitor BKM120 and the BRD4 inhibitor JQ1 [77]. Similarly, the dual PI3 K–histone deacetylase (HDAC) inhibitor CUDC-907 reduces MYC gene transcription and MYC protein stability, as well as in vivo tumor growth of lymphoma xenografts [78]. While these inhibitors are not necessarily specific to MYC, the potent effects of dual function compounds on MYC expression and phenotypes supports further exploration of this therapeutic strategy for MYC-dependent tumors.

An alternative approach to inhibit the transcription of MYC takes advantage of complex DNA structures called G-quadruplexes. These secondary structures occur when hydrogen bonding connects a run of four guanines in a planar quartet. The assembly of two or more of these quartets makes up a G-quadruplex structure, which generally resides upstream of the transcriptional start site and silences gene expression. In contrast to BRD4 inhibitors, which indirectly inhibit MYC transcription, small molecules designed to bind and stabilize the G-quadruplexes associated with individual genes provides a unique way to target potentially undruggable oncogenes [79,80,81]. Studies have shown that small molecules, such as GQC-05, Cz1, IZCZ-3, and DC-34, are capable of binding and stabilizing G-quadruplexes within the nuclease hypersensitive element (NHE) III region of the MYC promoter, resulting in the suppression of MYC messenger RNA (mRNA) and protein and an induction of cytotoxicity [82,83,84,85,86,87]. Similarly, a study by Bouvard et al. [88] demonstrates that the small molecule stauprimide inhibits the transcription factor NME2 from being recruited to the NHE III region, stabilizing the MYC G-quadruplex, and selectively reducing MYC transcriptional gene programs. Despite having clear MYC-dependent phenotypes, the off-target effects of these various compounds are still being interrogated. In addition to these therapeutic strategies, advances have been made for the in vivo identification and tracking of DNA structures using small-molecule fluorescent probes [89]. This technique would allow for the screening of drugs that affect G-quadruplex structures in cancer and help to identify compounds that lead to the stabilization of these structures. While targeting of G-quadruplexes has emerged as a promising MYC therapeutic strategy, i-motifs, which form on the opposite strand of G-quadruplexes, are mutually exclusive with G-quadruplexes and can promote MYC transcription [90]. The stabilization of i-motifs may drive an acute increase in MYC expression leading to apoptosis; however, we need a more indepth knowledge of the dynamic relationship between these two DNA structures in order to effectively target them in cancer cells [79, 91].

2.2 Dimer Disruptors

There are an extensive number of pathways that upregulate MYC expression, increasing the probability that cancer cells will be able to circumvent therapeutics targeting upstream regulation of MYC [9]. Alternatively, compounds that directly bind and inactivate MYC’s downstream function may have better efficacy and reduced acquired resistance. While each MYC domain significantly contributes to MYC function and stability, the bHLHZ domain represents a logical therapeutic target, as it is required for dimerization of MYC to its binding partner MAX and subsequent DNA binding at E-box sequences. Some of the first MYC/MAX dimer disruptors, including 10058-F4 and 10074-G5, were characterized from chemical library screens using systems such as yeast two-hybrid or fluorescence resonance energy transfer (FRET) [92,93,94]. However, many of these compounds display low potency, with half-maximal inhibitory concentration (IC50) values ranging from 20 to 40 μM and potentially off-target effects [95]. Research efforts have focused on chemically improving these base compounds as well as identifying new compounds. Recently, a chemical screen using bimolecular fluorescence complementation (BiFC) was performed and the MYC:MAX Inhibitor MYCMI-6 was shown to bind directly within the bHLHZ domain and disrupt MYC/MAX dimerization at a low micromolar range (Dissociation constant [KD] ~ 1.5–2 μM) [96]. In a panel of 60 cancer cell lines, almost ~ 75% of lines expressing ‘high’ levels of MYC mRNA and/or protein showed sensitivity to MYCMI-6. Importantly, in vivo treatment of the MYCN-amplified neuroblastoma cell line SK-N-DZ with MYCMI-6 significantly induced cell death and reduced proliferation. However, MYCMI-6 does not lead to MYC protein degradation, and, therefore, the MAX-independent functions of MYC will need to be well-understood in this therapeutic setting [97]. Alternatively, Struntz et al. [98] demonstrate that stabilization of MAX:MAX homodimers, using the small molecule KI-MS2-008, leads to MYC degradation and attenuation of MYC transcriptional gene programs both in vitro and in vivo. KI-MS2-008 not only reduces MYC expression and function, but also takes advantage of transcriptionally inert MAX:MAX DNA binding. However, this strategy would impact the binding of other MYC network proteins to E-box sites, and, therefore, further investigation is needed [8]. Together, these results show great promise for compounds that disrupt the transcriptional activity of MYC and indicate that high MYC levels may represent a biomarker for clinical response to MYC/MAX dimer disruptors.

Alternatively, studies utilizing peptides against MYC have emerged as novel strategies that disrupt MYC/MAX heterodimers in an effort to reduce MYC DNA-binding potential and transcriptional activation [59, 99]. Although peptides have historically been challenging to administer to patients due to their short half-life and low bioavailability, modifications that address these issues have increased their clinical applicability. For instance, fusion of a MYC H1-derived peptide to an elastin-like polypeptide allowed the peptide to cross the cell membrane in vivo and disrupt MYC/MAX dimers in a glioma model [100]. Similarly, Wang et al. [101] demonstrated that the OmoMYC peptide was unable to penetrate cells; however, the addition of an N-terminal functional penetrating Phylomer peptide allowed OmoMYC to enter cells and reduce tumor growth in vivo. More recently, Beaulieu et al. [102] reported in vivo pre-clinical efficacy using a purified OmoMYC mini-protein, which has intrinsic cell penetrating properties and is capable of disrupting MYC-dependent transcription. Together, these studies represent exciting advances towards the clinical application of MYC-targeted peptides.

2.3 Inhibition of MYC Post-Translational Regulation

Given that MYC expression and activity are dynamically regulated by a variety of protein modifications, therapeutic targeting of these post-translational mechanisms provides an innovative, albeit indirect, way to reduce MYC function in cancer. These mechanisms include, but are not limited to (1) kinases that phosphorylate S62-MYC; (2) phosphatases that dephosphorylate S62-MYC; (3) the PIN1 proline isomerase; and (4) enzymes that affect MYC ubiquitin-dependent proteolysis.

2.3.1 Serine 62 Phosphorylation and Dephosphorylation

Kinase inhibitors that affect the active pS62-MYC state include ERK, CDK2, and CDK9 inhibitors [103,104,105,106,107,108,109]. Unfortunately, cancer cells are quite adept at rewiring signaling pathways in response to targeted therapies in order to keep MYC and other signaling substrates active [110,111,112]. An alternative approach to decrease pS62-MYC is through the activation of PP2A, a serine/threonine phosphatase that targets pS62 [113]. PP2A is a heterotrimeric complex composed of a catalytic subunit (PP2A-C), a structural subunit (PP2A-A), and one of 26 different regulatory B subunits, the latter of which is responsible for fully activating the complex and dictating substrate specificity [114,115,116]. During oncogenesis, cancer cells usually acquire mechanisms to suppress PP2A function [117]. The global suppression of PP2A function has been shown to contribute to cancer cell proliferation, transformation, epithelial-to-mesenchymal transition, and resistance to targeted therapies, placing PP2A as a central regulator of oncogenic signaling [111, 118,119,120]. In a short hairpin RNA (shRNA) knockdown screen, decreased expression of the PP2A-B subunit PPP2R5A (B56α) increased anchorage-independent growth in soft agar, implicating this subunit in the regulation of cellular transformation [121]. We found that B56α is the only B subunit able to directly dephosphorylate pS62 MYC and that the loss of B56α leads to increased MYC expression [40]. The activation of PP2A has, therefore, emerged as an attractive therapeutic strategy to target pS62-MYC to decrease MYC activity and protein stability. Currently, there are several compounds, both indirect and direct, that lead to the activation of PP2A and have tumor-suppressor activities.

2.3.1.1 Indirect Protein Phosphatase 2A (PP2A) Activation

Consistent with the tumor suppressor role of PP2A, the PP2A inhibitors, inhibitor-2 of protein phosphatase-2A (SET) and cancerous inhibitor of PP2A (CIP2A), are overexpressed in a variety of cancers [122,123,124,125,126]. These proteins function to prevent PP2A-B subunits from binding the PP2A A-C core complex, decreasing global PP2A activity and contributing to therapeutic resistance [127, 128]. Interestingly, while CIP2A can broadly inhibit PP2A, it has been shown to preferentially inhibit MYC-associated PP2A in order to increase MYC stability and function [128, 129] Additionally, CIP2A is stabilized when bound to the PP2A B56α subunit, highlighting the importance of this protein to MYC activity [130]. Unfortunately, the therapeutic targeting of CIP2A remains an important and understudied area of research, as there are few therapeutic compounds shown to inhibit CIP2A activity [117]. Currently, bortezomib, a proteasome inhibitor with CIP2A-inhibiting activities, is US Food and Drug Administration (FDA) approved for multiple myeloma and mantle cell lymphoma and is being assessed in other cancers in phase I, II, and III clinical trials. Similar to CIP2A, SET contributes to cancer cell survival and tumor progression. Knockdown of SET reduces MYC phosphorylation and expression levels in breast and pancreatic cancer cells, and leads to decreased cell survival, supporting the use of SET inhibitors as important therapeutic strategy [47, 49]. OP449 is an oligopeptide that binds to SET and sequesters it from the PP2A complex, indirectly activating PP2A [131]. Similar to SET knockdown, OP449 treatment led to decreased pS62-MYC and reduced in vivo tumor growth in pancreatic and breast cancer cells [47, 49]. In CML, elevated levels of ABL lead to increased SET expression and PP2A inhibition [132]. Studies have shown that OP449 induces a cytotoxic response in acute myeloid leukemia (AML) and CML cells, including patients that are resistant to ABL1 kinase inhibitors [133]. FTY720, a sphingosine analog, has also been shown to have PP2A-activating properties [134, 135]. However, FTY720 functions primarily through immunosuppression by internalizing and activating the sphingosine 1 phosphate receptor (S1PR) [136]. FTY720 analogs, such as SH-RF-177, induce cell death in part through PP2A activation without the activation of S1PR, increasing the clinical significance of these compounds [137].

2.3.1.2 Direct PP2A Activation

More recently, small-molecule activators of PP2A (SMAPs) have emerged as novel, first-in-class therapeutic agents that directly activate PP2A. These compounds were generated based on the established functional groups of tricyclic antipsychotics, which have been shown to have PP2A-activating properties at high concentrations [138]. Using binding assays and photoaffinity labeling, Sangodkar et al. [139] demonstrated that SMAPs bind directly to the PP2A-Aα subunit, causing a conformational change that alleviates negative inhibition and leads to PP2A activation. Treatment with SMAPs reduces tumor growth and proliferation in pancreatic, lung, and castration-resistant prostate cancer, in vivo and in vitro [111, 139, 140]. These results are associated with attenuated oncogenic signaling, with significant decreases in active ERK, SRC, CDK, and MYC. Recently, Kauko et al. [111] demonstrated that knockdown of PP2A contributes to kinase inhibitor resistance, in part due to the induction of high MYC levels. Consistent with these studies, we recently demonstrated that select kinase inhibitors can function synergistically with SMAPs in breast and pancreatic cancer cells [140]. Specifically, the combination of mammalian target of rapamycin (mTOR) inhibitors and SMAPs synergistically reduced MYC levels beyond the capabilities of either single agent and potently induced cell death. This combination was also associated with increased suppression of AKT signaling, a common resistance mechanism to mTOR inhibitors. Similar to the inhibition of MYC using OmoMYC, SMAPs show little to no toxicity. There are studies, however, that suggest not all PP2A-B subunits function as tumor suppressors. Specifically, a study by Zhang et al. [141] demonstrates that the PP2A-B55α subunit is able to bind MYC with the help of the transcription factor EYA3 and dephosphorylate pT58, leading to increased stability of MYC. Despite the complex roles of PP2A-B subunits in disease states, the aggregate activation of PP2A by SMAPs appears to be detrimental to cancer cells and indicates a unique susceptibility of cancer cells to PP2A activation.

2.3.2 PIN1 Inhibition

PIN1 is a prolyl isomerase that causes a cistrans or transcis conformational change at proline resides that follow phosphorylated serine/threonine sites (pS/T-P sites) [142]. PIN1 isomerization has significant effects on the localization, stability, and activation of target proteins that regulate a variety of cellular processes including proliferation, survival, and epithelial-to-mesenchymal transition [143]. Similar to MYC, PIN1 is tightly regulated in normal cells, but is aberrantly upregulated in a variety of cancers, including prostate, breast, lung, ovary and cervical tumors, and melanoma [144], and is associated with poor patient outcomes [145, 146]. Further, PIN1 cooperates with aberrant expression of HER2 and RAS to drive tumorigenesis, placing PIN1 as a central mediator of common oncogenic signals and an attractive therapeutic target [147, 148].

We have demonstrated that PIN1 dynamically regulates MYC activity, with isomerization influencing both MYC activation and degradation. In normal cells, PIN1 helps to balance the activation of MYC with its degradation at select target genes; however, in oncogenic states, where MYC turnover is commonly suppressed through multiple mechanisms and PIN1 expression is high, the predominant effect of PIN1 is to promote MYC activation and regulation of genes involved in tumorigenesis [149]. Consistent with these results, mice with genetic loss of Pin1 (PIN1 knockout [KO] mice) are developmentally normal aside from male sterility, but display increased resistance to tumorigenesis, suggesting that PIN1 predominantly functions as a tumor promoter in this context [150]. Helander et al. [19] demonstrate that PIN1 is capable of binding the MB0 domain of MYC in a potentially priming event to transcriptional activation. Upon S62 MYC phosphorylation, this interaction is stabilized, increasing the association of PIN1 with the MB1 domain where it can promote the isomerization of P63, increasing MYC transcriptional and transforming activity [19]. Interestingly, Su et al. [151] recently demonstrated that serum stimulation localizes pS62-MYC to the nuclear pore basket in a PIN1-dependent manner, where it binds to target genes that regulate proliferation and migration. These studies suggest that the subcellular localization of MYC may be an important regulatory mechanism of MYC transcriptional activity and target gene selection. In addition to direct MYC regulation, PIN1 also influences the activity of proteins that alter MYC’s post-translational modification state, including ERK, CDK, GSK3, and the deSUMOylase SENP1 [152]. Consistent with these results, overexpression of PIN1 increases the transforming potential of MYC, suggesting that PIN1 functions predominantly as a tumor promoter in cancer cells and that therapeutic targeting of PIN1 may be a viable approach to reduce MYC activity [149].

Inhibition of PIN1 prolyl isomerase activity with therapeutic compounds, such as juglone or PiB (diethyl-1,3,6,8-tetrahydro-1,3,6,8-tetraoxobenzo[lmn]3, 8 phenanthroline-2,7-diacetate), have shown efficacy in a variety of tumors [153,154,155,156,157,158,159,160]. However, both of these compounds are able to reduce proliferation in a Pin1 null mouse, indicating that PIN1 is not their only target [158, 161]. Several screens have been performed to try and identify selective inhibitors to PIN1; however, the majority of these studies have led to false positive or non-selective compounds [154]. Wei et al. [162] identified all-trans retinoic acid (ATRA) as a novel PIN1 inhibitor using a fluorescence polarization-based high-throughput screening to identify compounds that bind to the active form of PIN1. Importantly, treatment with ATRA significantly reduces in vivo growth of triple-negative breast cancer xenografts and leads to the degradation of PIN1 protein [162]. Similar results were seen in hepatocellular carcinoma xenografts using a slow-release, poly L-lactic acid microparticle containing ATRA [153]. Recently, a small molecule, KPT-6566, has been shown to both covalently bind PIN1’s catalytic site and target PIN1 for degradation, while simultaneously releasing a quinone-mimicking drug that induces DNA damage and cell death [160]. Treatment with KPT-6566 results in cytotoxicity specifically in a panel of cancer cell lines, as compared to normal cells. These results, together with the PIN1 KO mouse results, suggest that normal cells can tolerate the loss of PIN1 while cancer cells rely on PIN1 for specific oncogenic functions important for their survival, particularly under stressed conditions as occurs in vivo. Importantly, treatment of Pin1 null Mouse Embryonic Fibroblast (MEFs) with KPT-6566 had no effect on cell proliferation, indicating that this compound may have a higher specificity for PIN1.

2.3.3 Targeting MYC Stability

The primary approaches to alter MYC stability center on increasing MYC ubiquitin-mediated degradation. MYC family proteins are ubiquitinated by a variety of E3 ubiquitin ligases (E3s), most of which stimulate MYC degradation [15, 163,164,165]. Importantly, in cancer, mutations or loss of MYC-directed E3s, such as FBW7, frequently occur, contributing to MYC stability [166]. Conversely, E3 s such as SKP2 and HUWE1 are often overexpressed in cancer and have been shown to positively affect MYC activity, presenting potential targets for indirect MYC inhibition [167,168,169]. For example, Peter et al. [170] demonstrated that inhibition of HUWE1 with either shRNA or small-molecule inhibitors leads to decreased cancer cell viability and suppression of MYC transcriptional activity. However, these findings may be context dependent as other groups have found that HUWE1 has a tumor suppressor function [171, 172]. Therefore, a more indepth understanding of the mechanisms that regulate MYC ubiquitination and turnover are necessary in order to capitalize on therapies that target these factors. An alternative strategy to enhance the activity of E3s that target MYC for degradation is to target MYC deubiquinating enzymes (DUBs). DUBs that deubiquitinate and stabilize MYC family proteins include USP7, USP13, USP22, USP28, USP36, and USP37 [173,174,175,176,177,178,179]. Inhibition of these DUBs has been reported to attenuate MYC-dependent gene transcription and increase MYC turnover, with inhibition of USP36 resulting in a dramatic decrease in c-MYC expression and induction of cytotoxicity [175]. Similarly, USP7 has been shown to bind and stabilize N-MYC and inhibition of this DUB decreases N-MYC driven tumorigenesis in vivo [180]. Another strategy to affect MYC ubiquitination involves modulation of the small ubiquitin-related modifier (SUMO), where we have discovered that inhibition of the deSUMOylation enzyme SENP1, which is overexpressed in human breast cancer cells and increases MYC stability and transactivation activity, stimulated MYC ubiquitination and increased MYC degradation, providing a new strategy for MYC protein degradation [181]. Finally, the Aurora-A kinase inhibitors MLN8054 and MLN8237 have been shown to affect both c-MYC and N-MYC ubiquitin-mediated degradation independent of their kinase activity [182,183,184,185,186,187,188]. Recently, Li et al. [184] demonstrated that elevated MYC expression correlated with MLN8237 response, both in vitro and in vivo, in thyroid cancer. These studies support the role of Aurora inhibitors as potential MYC-destabilizing therapeutics and suggest that MYC expression may be used as a biomarker for patient response. However, despite promising data, clinical trials with MLN8237 have raised concerns about the safety profile of this compound, with a number of trials resulting in significant toxicities and disease progression [189].

3 Perspectives

Biomarkers indicative of the mode of MYC deregulation would be extremely informative in directing strategies to target MYC. For example, for tumors with MYC amplification or high MYC gene transcription, inhibitors of MYC transcription such as BET inhibitors, or G-quadruplex stabilizers may show great promise. Likewise, inhibitors of MYC:MAX DNA binding or dimerization could also be quite efficacious in these settings. For tumors where MYC is post-translationally deregulated, an event that most likely occurs in the majority of human tumors, targeting MYC:MAX DNA binding or dimerization could be effective, but other strategies may also be promising, such as targeting enzymes that control active MYC modifications such as S62-MYC phosphorylation or PIN1-mediated isomerization. Likewise, for tumors where the MYC half-life is extended, determined by discordant MYC protein versus mRNA, inhibition of DUBs or SENP1 may be beneficial. The expression of these MYC-modifying enzymes within tumors could present important biomarkers to direct therapeutic strategies. It is also important to consider that post-translational modifications impart dynamic protein control, and a mechanistic understanding of these dynamics should be considered in targeting strategies. For example, the balance between E3 ubiqutin ligases and deubiquitinating enzymes or PIN1 regulation of MYC, which affects both the temporal and spatial activity of MYC [151]. So far, it appears that cancer cells are particularly vulnerable to deregulation of these precise post-translational control mechanisms, which may favor targeting these modifier enzymes, but also could impact dosing or combination strategies. In addition to determining the mechanism by which MYC is aberrantly activated, careful consideration should be made when selecting compounds that indirectly inhibit MYC activity, as many of these agents target factors other than MYC; although, depending on the outcome and target, these post-translational modifier enzymes often target other oncogenic proteins, potentially increasing their efficacy as targeted anti-cancer agents.

4 Conclusion

Studies over the last several decades indicate that multiple methods of MYC deregulation in cancer exist and not all MYC deregulation is the same, stressing the importance of biomarkers that distinguish these mechanisms, as well as development of diverse therapeutic agents that target unique aspects of MYC oncogenicity.