Molecular mechanism of activation-triggered subunit exchange in Ca(2+)/calmodulin-dependent protein kinase II

Activation triggers the exchange of subunits in Ca²⁺/calmodulin-dependent protein kinase II (CaMKII), an oligomeric enzyme that is critical for learning, memory, and cardiac function. The mechanism by which subunit exchange occurs remains elusive. We show that the human CaMKII holoenzyme exists in dodecameric and tetradecameric forms, and that the calmodulin (CaM)-binding element of CaMKII can bind to the hub of the holoenzyme and destabilize it to release dimers. The structures of CaMKII from two distantly diverged organisms suggest that the CaM-binding element of activated CaMKII acts as a wedge by docking at intersubunit interfaces in the hub. This converts the hub into a spiral form that can release or gain CaMKII dimers. Our data reveal a three-way competition for the CaM-binding element, whereby phosphorylation biases it towards the hub interface, away from the kinase domain and calmodulin, thus unlocking the ability of activated CaMKII holoenzymes to exchange dimers with unactivated ones.

DOI:http://dx.doi.org/10.7554/eLife.13405.001

How does memory outlast the lifetime of the molecules that encode it? One enzyme that is found in neurons and has been suggested to help long-term memories to form is called CaMKII. Each CaMKII assembly is typically composed of 12 to 14 protein subunits associated in a ring and can exist in either an “unactivated” or “activated” state. In 2014, researchers showed that CaMKII assemblies can exchange subunits with each other. Importantly, an active CaMKII can mix with an unactivated CaMKII and share its activation state. CaMKII may use this mechanism to spread information to the next generation of proteins – thereby allowing activation to outlast the lifespan of the initially activated proteins. However the molecular mechanism that underlies this process was not clear.

Now, Bhattacharyya et al. – including some of the researchers involved in the 2014 work – address two questions about this mechanism. How do subunits exchange between CaMKII assemblies? And how does the activation of CaMKII initiate subunit exchange?

A closed-ring hub ties the subunits of CaMKII together, similar to the organization of the segments in an orange. To undergo subunit exchange, the hub must open up to release and accept subunits. Bhattacharyya et al. have now uncovered an intrinsic flexibility in the hub that is triggered by a short peptide segment in CaMKII. This segment, which is exposed in activated CaMKII but not in the unactivated form, can crack open the hub ring by binding between the hub subunits, like a finger separating the segments of an orange. This allows the hub to flex and expand, and once open, the hub’s flexibility allows room for subunits to be released or accepted.

Although this subunit exchange mechanism could be a powerful means for spreading the activated state throughout signaling pathways, the biological relevance of this phenomenon has not been clarified. However, the mechanistic framework provided by Bhattacharyya et al. may allow new experiments to be performed that test the consequences of subunit exchange in live cells and organisms. It could also enable investigations into the importance of subunit exchange in long-term memory.

DOI:http://dx.doi.org/10.7554/eLife.13405.002

Abstract POSTER-BIOL-1350: Dissecting cellular and molecular consequences of disrupting cyclooxygenase-1 activity in ovarian cancer

Introduction: Cyclooxygenase-1 (COX-1) and COX-2 catalyze the rate-limiting step in the biosynthesis of prostaglandins and thromboxanes, which contribute to tumorigenesis by promoting tumor proliferation, escape from apoptosis, and angiogenesis. COX-1 is overexpressed in multiple human and mouse models of ovarian cancer and that COX-1 inhibition reduces ovarian cancer cell viability in preclinical models; in contrast COX-2 is the principle COX isoform expressed in most solid tumors. Our group has an ongoing interest in developing COX-1-targeted molecules for molecular imaging and therapeutic use.

Hypothesis: We hypothesize that COX-1 is a viable molecular target for the treatment of ovarian cancer. To test this hypothesis, we (i) compared expression of COX-1 and COX-2 in The Cancer Genome Atlas (TCGA) ovarian tumors and in an ovarian cancer tissue microarray (TMA) generated in our laboratory, and (ii) determined cellular and molecular effects of disrupting COX-1 activity in ovarian cancer cells.

Methods: RNA-seq mRNA data for COX-1 and COX-2 in the PANCAN 12 tumor panel of TCGA was accessed through the Firehose repository (Broad Institute, Cambridge, MA). COX-1 and COX-2 protein expression in our TMA was determined by immunohistochemistry, and percentage of COX-1 and COX-2-positive tumor cells quantified. An ovarian cancer cell line model of disrupted COX-1 activity was generated by the stable integration of shRNA targeting COX-1 in OVCAR-3 cells (ShCOX1). Cell growth in ShCOX1 cells compared to control cells expressing scrambled shRNA (ShSCR) was determined by sulphorhodamine B assay. Differential gene expression between ShSCR and ShCOX1 cells was determined by RNA-seq.

Results: TCGA RNA-seq data revealed that COX-1 mRNA levels was significantly higher than COX-2 expression in ovarian cancer and higher than 11 other solid tumor types. We confirmed these findings at the protein level in our TMA. Specific COX-1 knockdown in OVCAR-3 cells significantly reduced growth compared to control cells. Pathway analysis from RNA-seq data revealed that COX-1 knockdown significantly downregulates genes involved in cell proliferation, cell migration/invasion, angiogenesis and epithelial-mesenchymal transition.

Conclusions: Our findings confirm that COX-1 is over-expressed in ovarian cancer, and is likely to play a key role in tumor progression through promoting oncogenic tumor growth and progression. Therefore, exploiting high COX-1 expression in ovarian cancer with novel COX-1-targeted molecules has the potential to improve the detection and treatment of this deadly disease.

Citation Format: Andrew J. Wilson, Brian D. Lehmann, Qi Liu, MD Jashim Uddin, Anna Elleman, Brenda Crews, Jeanette Saskowski, Jennifer Pietenpol, Lawrence J. Marnett, Dineo Khabele. Dissecting cellular and molecular consequences of disrupting cyclooxygenase-1 activity in ovarian cancer [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-BIOL-1350.

Abstract 4928: Cyclooxygenase-1 as a target for molecular imaging of high grade serous ovarian cancer

Ovarian cancer is the most lethal gynecologic malignancy. Most women diagnosed with high grade serous ovarian cancer, the most common subtype, present with advanced disease characterized by widespread peritoneal carcinomatosis. Non-invasive imaging plays a critical role in the management of ovarian cancer by confirming sites of disease, informing treatment interventions, and determining the effectiveness of treatment. However, current imaging methods are limited by relatively high false positive and false negative rates. The development of novel molecular imaging agents is urgently required. Emerging evidence from our group and others suggests that cyclooxygenase-1 (COX-1) is an important molecular target in ovarian cancer and distinguishes ovarian cancer from most other solid tumors, where COX-2 is the dominant isoform. COX-1 and COX-2 catalyze the rate-limiting step in the biosynthesis of prostaglandins and thromboxanes, which contribute to tumorigenesis by promoting tumor proliferation, escape from apoptosis, and angiogenesis. In this study, we examined COX-1 expression in high grade serous ovarian tumors using data extracted from The Cancer Genome Atlas (TCGA) and a tissue microarray (TMA) generated in our laboratory from 209 patients. Raw mRNA RNA-seq expression data from TCGA revealed that PTGS1, the gene encoding COX-1, is in the top 10% of genes expressed in ovarian cancer. Furthermore, mRNA expression of the COX-2-encoding gene, PTGS2, in ovarian tumors was significantly lower than COX-1 mRNA levels (p<0.00001, Mann-Whitney test), and COX-1 expression was selectively higher in ovarian tumors than in 11 other solid cancers in the TCGA PANCAN12 panel (p<0.00001 compared to COX-1 levels in ovarian tumors, Mann-Whitney test). We confirmed these findings at the protein level in our TMA. There was high COX-1 expression in a large subset of high grade serous tumors, but markedly lower expression in other epithelial ovarian cancer subtypes (endometrioid, mucinous and clear cell, p<0.0001, Mann-Whitney test). Having confirmed COX-1 as an attractive molecular target in high grade serous ovarian cancer, newly synthesized COX-1-selective compounds were screened in COX-1-expressing OVCAR-3 ovarian cancer cells. Two novel small molecule inhibitors of COX-1 (LM-9954 and LM-9955) inhibited COX-1 in biochemical assays and reduced prostaglandin synthesis in OVCAR-3 cells at nanomolar concentrations. The highly selective expression of COX1 in serous ovarian carcinoma provides a unique molecular screening opportunity and we are currently synthesizing 18F-conjugated forms of these agents to act as radiotracers in positron emission tomography (PET) imaging. Exploiting high COX-1 expression in high grade serous ovarian cancer with novel COX-1-targeted molecules has the potential to improve detection of this deadly disease.

Citation Format: Andrew J. Wilson, Brian D. Lehmann, MD Jashim Ussin, Anna Elleman, Brenda Crews, Jeanette Saskowski, Jennifer Pietenpol, Lawrence J. Marnett, Dineo Khabele. Cyclooxygenase-1 as a target for molecular imaging of high grade serous ovarian cancer. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4928. doi:10.1158/1538-7445.AM2014-4928