Failure to cure ovarian cancer relates to the persistence of dormant, drug-resistant cancer cells following surgery and chemotherapy. “Second look” surgery can detect small, poorly vascularized nodules of persistent ovarian cancer in ~50% of patients, where >80% are undergoing autophagy and express DIRAS3. Autophagy is one mechanism by which dormant cancer cells survive in nutrient poor environments. DIRAS3 is a tumor suppressor gene downregulated in >60% of primary ovarian cancers by genetic, epigenetic, transcriptional and post-transcriptional mechanisms, that upon re-expression can induce autophagy and dormancy in a xenograft model of ovarian cancer. We examined the expression of DIRAS3 and autophagy in ovarian cancer cells following nutrient deprivation and the mechanism by which they are upregulated. We have found that DIRAS3 mediates autophagy induced by amino acid starvation, where nutrient sensing by mTOR plays a central role. Withdrawal of amino acids downregulates mTOR, decreases binding of E2F1/4 to the DIRAS3 promoter, upregulates DIRAS3 and induces autophagy. By contrast, acute amino acid deprivation did not affect epigenetic regulation of DIRAS3 or expression of miRNAs that regulate DIRAS3. Under nutrient poor conditions DIRAS3 can be transcriptionally upregulated, inducing autophagy that could sustain dormant ovarian cancer cells.

Autophagy can protect cancer cells from acute starvation and enhance resistance to chemotherapy. Previously, we reported that autophagy plays a critical role in the survival of dormant, drug resistant ovarian cancer cells using human xenograft models and correlated the up-regulation of autophagy and DIRAS3 expression in clinical samples obtained during “second look” operations. DIRAS3 is an imprinted tumor suppressor gene that encodes a 26 kD GTPase with homology to RAS that inhibits cancer cell proliferation and motility. Re-expression of DIRAS3 in ovarian cancer xenografts also induces dormancy and autophagy. DIRAS3 can bind to Beclin1 forming the Autophagy Initiation Complex that triggers autophagosome formation. Both the N-terminus of DIRAS3 (residues 15–33) and the switch II region of DIRAS3 (residues 93–107) interact directly with BECN1. We have identified an autophagy-inhibiting peptide based on the switch II region of DIRAS3 linked to Tat peptide that is taken up by ovarian cancer cells, binds Beclin1 and inhibits starvation-induced DIRAS3-mediated autophagy.

The persistence of dormant, drug-resistant cancer cells after cytoreductive surgery and combination chemotherapy is a major challenge that contributes to poor outcomes for ovarian cancer patients. Despite normalization of CA125 and negative imaging following primary treatment, “second look” surgery can detect small, quiescent, poorly vascularized nodules of persistent ovarian cancer in ~50% of patients. After positive second look operations, persistent ovarian cancer may take months or years to become clinically evident, consistent with tumor dormancy. Autophagy is one mechanism by which persistent, potentially dormant cancer cells could survive in a nutrient poor environment. Autophagy or “self-eating” is a catabolic process by which organelles and misfolded proteins are degraded and recycled to provide energy that could protect dormant cancer cells from acute starvation in the face of an inadequate blood supply. Autophagy can be induced by multiple mechanisms. Our group has found that, in contrast to primary ovarian cancers, the cancer cells in >80% of positive second look biopsies exhibit autophagy and express DIRAS3 (ARHI). DIRAS3 is an imprinted tumor suppressor gene that is downregulated in >60% of primary ovarian cancers by multiple mechanisms including loss of heterozygosity, CpG promoter methylation, transcriptional repression (E2F1/4) and miRNA regulation (miR-221/222). Re-expression of DIRAS3 prevents ovarian cancer growth, inhibits motility, induces autophagy and establishes tumor dormancy in xenografts. When DIRAS3 is upregulated in human ovarian carcinoma cell lines from a doxycycline-inducible promoter in nu/nu murine xenografts cancer cells remain dormant until DIRAS3 is downregulated permitting cell cycling, vascularization and progressive growth. Autophagy is required for survival of dormant xenografts in that treatment with chloroquine, a functional inhibitor of autophagy, markedly delays outgrowth of ovarian cancer xenografts when DIRAS3 is downregulated. Mechanisms by which DIRAS3 is upregulated and autophagy induced in positive second look specimens remains poorly understood. We have found that amino acid deprivation upregulates DIRAS3 and induces autophagy. Knockdown of DIRAS3 markedly decreases induction of autophagy following nutrient deprivation. We have identified that amino acid deprivation results in decreased E2F1/E2F4 expression and transcriptional upregulation of DIRAS3, but changes in promoter methylation or miRNA regulation are not observed. Knockdown of E2F1 or E2F4, both genetically and pharmacologically increase DIRAS3 expression and autophagy. Taken together these results suggest that nutrient deprivation results in transcriptional upregulation of DIRAS3-mediated autophagy and thus likely presents a mechanism by which DIRAS3 is upregulated in dormant, second look ovarian cancer tumor specimens. Citation Format: Margie Nicole Sutton, Gilbert Y. Huang, Jinhua Zhou, Zhen Lu, Robert C. Bast. DIRAS3 (ARHI) is required for amino acid-mediated autophagy and nutrient deprivation in dormant ovarian cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1325.

Among the 3 GTPases in the DIRAS family, DIRAS3/ARHI is the best characterized. DIRAS3 is an imprinted tumor suppressor gene that encodes a 26-kDa GTPase that shares 60% homology to RAS and RAP. DIRAS3 is downregulated in many tumor types, including ovarian cancer, where re-expression inhibits cancer cell growth, reduces motility, promotes tumor dormancy and induces macroautophagy/autophagy. Previously, we demonstrated that DIRAS3 is required for autophagy in human cells. Diras3 has been lost from the mouse genome during evolutionary re-arrangement, but murine cells can still undergo autophagy. We have tested whether DIRAS1 and DIRAS2, which are homologs found in both human and murine cells, could serve as surrogates to DIRAS3 in the murine genome affecting autophagy and cancer cell growth. Similar to DIRAS3, these 2 GTPases share 40-50% homology to RAS and RAP, but differ from DIRAS3 primarily in the lengths of their N-terminal extensions. We found that DIRAS1 and DIRAS2 are downregulated in ovarian cancer and are associated with decreased disease-free and overall survival. Re-expression of these genes suppressed growth of human and murine ovarian cancer cells by inducing autophagy-mediated cell death. Mechanistically, DIRAS1 and DIRAS2 induce and regulate autophagy by inhibition of the AKT1-MTOR and RAS-MAPK signaling pathways and modulating nuclear localization of the autophagy-related transcription factors FOXO3/FOXO3A and TFEB. Taken together, these data suggest that DIRAS1 and DIRAS2 likely serve as surrogates in the murine genome for DIRAS3, and may function as a backup system to fine-tune autophagy in humans.

Altered energy metabolism is a cancer hallmark as malignant cells tailor their metabolic pathways to meet their energy requirements. Glucose and glutamine are the major nutrients that fuel cellular metabolism, and the pathways utilizing these nutrients are often altered in cancer. Here, we show that the long ncRNA CCAT2, located at the 8q24 amplicon on cancer risk-associated rs6983267 SNP, regulates cancer metabolism in vitro and in vivo in an allele-specific manner by binding the Cleavage Factor I (CFIm) complex with distinct affinities for the two subunits (CFIm25 and CFIm68). The CCAT2 interaction with the CFIm complex fine-tunes the alternative splicing of Glutaminase (GLS) by selecting the poly(A) site in intron 14 of the precursor mRNA. These findings uncover a complex, allele-specific regulatory mechanism of cancer metabolism orchestrated by the two alleles of a long ncRNA.

Abstract DIRAS3 (also known as ARHI; Aplasia Ras Homolog Member I) is a potent tumor suppressor that has been well characterized for its role in ovarian cancer and autophagy. DIRAS3 is a maternally imprinted tumor suppressor gene that encodes a 26kDa GTPase which shares 60% homology to Ras and Rap. DIRAS3 is downregulated in many tumor types including ovarian cancer. DIRAS3 is downregulated by multiple mechanisms including loss of heterozygosity, transcriptional regulation by E2F1 and E2F4, hypermethylation of the second allele, and regulation by miRNAs 221 and 222. Re-expression of DIRAS3 at physiologic levels inhibits cancer cell growth, reduces motility, induces autophagy and promotes tumor dormancy. The mechanisms by which DIRAS3 induce autophagic cell death in vitro and tumor dormancy in vivo have been well characterized by previous work in our laboratory demonstrating that DIRAS3 is required for the induction of autophagy, and that upon inhibition of autophagy you can prevent outgrowth of dormant ovarian cancer cells in vivo. Interestingly, mice do not have DIRAS3 as it was lost during evolutionary rearrangement of murine chromosomes 3 and 6 nearly 60 million years ago, yet murine cells can still undergo autophagy. The DIRAS family members, DIRAS1 and DIRAS2 share 50-60% homology with DIRAS3 and are found in the murine genome. These 22kDa GTPases differ from DIRAS3 by the truncation of their N-terminal extension. Although DIRAS1 and DIRAS2 have not previously been characterized in ovarian cancer, TCGA analysis suggests that higher mRNA expression of these genes results in a survival advantage for patients with high grade serous ovarian cancer. In this study we demonstrate that re-expression of DIRAS1 and DIRAS2 inhibit ovarian cancer cell growth in vitro and induce autophagy in both human and murine cells. DIRAS1 and DIRAS2 are required for murine autophagy induced by rapamycin or amino acid starvation. Overexpression of DIRAS1 and DIRAS2 inhibits cancer cell growth and motility, and results in the inhibition of both the PI3K and Ras/MAPK signaling pathways. Thus DIRAS1 and DIRAS2 provide many of the functions of DIRAS3 in normal and malignant murine cells. Citation Format: Margie N. Sutton, Zhen Lu, Hailing Yang, Gilbert Huang, Yan Wang, Weiqun Mao, Robert C. Bast. DIRAS1 and DIRAS2 are novel ovarian cancer tumor suppressors that regulate cell growth, motility and autophagy. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3659.

Membrane-bound cGMP-dependent protein kinase (PKG) II is a key regulator of bone growth, renin secretion, and memory formation. Despite its crucial physiological roles, little is known about its cyclic nucleotide selectivity mechanism due to a lack of structural information. Here, we find that the C-terminal cyclic nucleotide binding (CNB-B) domain of PKG II binds cGMP with higher affinity and selectivity when compared with its N-terminal CNB (CNB-A) domain. To understand the structural basis of cGMP selectivity, we solved co-crystal structures of the CNB domains with cyclic nucleotides. Our structures combined with mutagenesis demonstrate that the guanine-specific contacts at Asp-412 and Arg-415 of the C-helix of CNB-B are crucial for cGMP selectivity and activation of PKG II. Structural comparison with the cGMP selective CNB domains of human PKG I and Plasmodium falciparum PKG (PfPKG) shows different contacts with the guanine moiety, revealing a unique cGMP selectivity mechanism for PKG II.

Membrane bound type II cGMP dependent protein kinase (PKG II) is a central mediator of cGMP signaling cascade, which regulates circadian rhythmicity, intestinal water secretion, bone growth and renal functions. PKG II contains an N-terminal regulatory (R)-domain, and a C-terminal catalytic (C)-domain. The R-domain contains tandem cyclic nucleotide binding domains (CNB-A and B) each with different affinities for cGMP, the second messenger that regulates kinase activity of PKG II. While it is known that PKG II needs to be highly selective for cGMP over cAMP to function properly, little is known about its cyclic nucleotide selectivity and the selectivity's role in activation. To understand its cyclic nucleotide selectivity and activation mechanism of PKG II, we first identified CNB-B to be highly selective for cGMP and solved its crystal structure with cGMP. The complex structure revealed that PKG II utilizes an arginine and two aspartate residues on the C-terminal helix to recognize the guanine moiety in cGMP. This is completely different from PKG I, where a conserved arginine from the β barrel of CNB-B specifically binds the guanine moiety of cGMP and imparts cyclic nucleotide selectivity. We are currently testing the roles of PKG II specific interactions in cGMP selectivity and activation of PKG II.

High selectivity of cyclic-nucleotide binding (CNB) domains for cAMP and cGMP are required for segregating signaling pathways; however, the mechanism of selectivity remains unclear. To investigate the mechanism of high selectivity in cGMP-dependent protein kinase (PKG), we determined a room-temperature joint X-ray/neutron (XN) structure of PKG Iβ CNB-B, a domain 200-fold selective for cGMP over cAMP, bound to cGMP (2.2 Å), and a low-temperature X-ray structure of CNB-B with cAMP (1.3 Å). The XN structure directly describes the hydrogen bonding interactions that modulate high selectivity for cGMP, while the structure with cAMP reveals that all these contacts are disrupted, explaining its low affinity for cAMP.

Cyclic guanosine monophosphate (cGMP) and cyclic AMP (cAMP)-dependent protein kinases (PKG and PKA) are closely related homologs, and the cyclic nucleotide specificity of each kinase is crucial for keeping the two signaling pathways segregated, but the molecular mechanism of cyclic nucleotide selectivity is unknown. Here, we report that the PKG Iβ C-terminal cyclic nucleotide binding domain (CNB-B) is highly selective for cGMP binding, and we have solved crystal structures of CNB-B with and without bound cGMP. These structures, combined with a comprehensive mutagenic analysis, allowed us to identify Leu296 and Arg297 as key residues that mediate cGMP selectivity. In addition, by comparing the cGMP bound and unbound structures, we observed large conformational changes in the C-terminal helices in response to cGMP binding, which were stabilized by recruitment of Tyr351 as a "capping residue" for cGMP. The observed rearrangements of the C-terminal helices provide a mechanical insight into release of the catalytic domain and kinase activation.

BackgroundcAMP-dependent protein kinase (PKA) and cGMP-dependent protein kinase (PKG) are the main effectors ofdistinct cyclic nucleotide pathways and are preferentiallyactivated by cAMP or cGMP, respectively.We recently characterized the isolated C-terminal cyc-lic nucleotide binding domain (CNB-B) of the humanPKG Ib as highly cGMP-selective (manuscript in pre-paration). In a crystal structure of the CNB-B two novelcGMP-specific interaction sites were identified in addi-tion to the previously described threonine residue (T317)in the phosphate binding cassette [1]. Mutation of eachindividual site resulted in reduced cGMP-selectivity andinterfered with cGMP-dependent activation of PKG Ib.To gain further insight into the molecular basis of cyc-lic nucleotide selectivity, we inserted two cGMP-specificinteraction sites into the CNB-B of human PKA RIa bymutating corresponding residues. We hypothesize thatthis way cGMP-specific interaction contacts can be cre-ated in PKA and thereby modulate cAMP-selectivity[1,2].ResultsWe characterized a deletion construct of the PKA hRIaCNB-B as cAMP-selective using fluorescence polarization(FP) and surface plasmon resonance (SPR).In comparison to the wildtype PKA hRIa CNB-B, singlemutant constructs showed similar affinities for cAMP-and cGMP-analogs, revealing a loss of selectivity. Thecombination of two mutations led to a construct withhigher affinity for cGMP compared to cAMP.Co-crystal structures of this double mutant with cAMPor cGMP, respectively, showed that the cGMP-specificinteraction contacts retained their function in the contextof the PKA hRIa CNB-B.ConclusionThegeneralstructureofcyclicnucleotidebindingdomains is conserved. However, varying amino acids inthe binding pocket enable the distinction between cAMPand cGMP. Here we show that cGMP interaction sitesfound in PKG do restore their specific binding mechan-isms when introduced into PKA.The results underline the relevance of the describednovel binding sites in mediating cGMP-selectivity. Still,other features of CNB domains involved in the specificbinding mechanism as well as the detailed mechanism ofkinase activation need to be investigated.

Background Cyclic guanosine monophosphate (cGMP) is a key secondary messenger that is produced in response to nitric oxide. One of the key mediators of cGMP signaling, cGMPdependent protein kinase (PKG), is activated upon binding to cGMP and phosphorylates downstream substrates in a process required for important physiological processes such as vasodilation, nociception, and memory formation. PKGs are also known to mediate most effects of drugs that increase cellular cGMP levels, including nitric oxidereleasing agents and phosphodiesterase inhibitors, which are used for the treatment of angina pectoris and erectile dysfunction, respectively. It is known that PKG is preferentially activated by cGMP over cAMP roughly 60-100 fold – however, the molecular mechanism by which cGMP is distinguished from a structurally similar messenger, cAMP, is poorly defined. Using competition fluorescence polarization (FP), X-ray crystallography, and in vitro kinase assays, we sought to understand the molecular basis for cGMP selectivity in PKGI.

Background Cyclic GMP-dependent protein kinases (PKGs) are central mediators of the NO-cGMP signaling pathway and phosphorylate downstream substrates that are crucial for regulating smooth muscle tone, platelet activation, nociception and memory formation. As one of the main receptors for cGMP, PKGs mediate most of the effects of cGMP elevating drugs, such as nitric oxide-releasing agents and phosphodiesterase inhibitors which are used for the treatment of angina pectoris and erectile dysfunction, respectively. Methodology/Principal Findings We have investigated the mechanism of cyclic nucleotide binding to PKG by determining crystal structures of the amino-terminal cyclic nucleotide-binding domain (CNBD-A) of human PKG I bound to either cGMP or cAMP. We also determined the structure of CNBD-A in the absence of bound nucleotide. The crystal structures of CNBD-A with bound cAMP or cGMP reveal that cAMP binds in either syn or anti configurations whereas cGMP binds only in a syn configuration, with a conserved threonine residue anchoring both cyclic phosphate and guanine moieties. The structure of CNBD-A in the absence of bound cyclic nucleotide was similar to that of the cyclic nucleotide bound structures. Surprisingly, isothermal titration calorimetry experiments demonstrated that CNBD-A binds both cGMP and cAMP with a relatively high affinity, showing an approximately two-fold preference for cGMP. Conclusions/Significance Our findings suggest that CNBD-A binds cGMP in the syn conformation through its interaction with Thr193 and an unusual cis-peptide forming residues Leu172 and Cys173. Although these studies provide the first structural insights into cyclic nucleotide binding to PKG, our ITC results show only a two-fold preference for cGMP, indicating that other domains are required for the previously reported cyclic nucleotide selectivity.