<b>Objectives:</b> Fibrillins <i>(FBN1-4)</i> determine the structure and function of the extracellular matrix. In malignancies, they drive the cytokine-loaded microenvironment and neoangiogenesis that promote tumor growth. In high-grade serous ovarian cancer (HGSOC), <i>FBN1</i> ranks among the top five genes with robust prognostic signature for survival [Ann. Oncol 31, 1240-50 (2020)]. Fibrillin-1 displays ≈60 globules aligned as ‘string-of-pearls,' ≈45 being epidermal growth factor domains (EGFDs); ≈40 EGFDs are hydroxylated at Asp/Asn residues by AspH, a 2-oxoacid utilizing dioxygenase (2OUD). AspH forms a high-affinity Ca<sup>2+</sup> binding site in-between adjacent EGFDs, stabilizing the elastic alignment of ‘pearls' within fibrillin. We hypothesized that AspH inhibition causes misfolding of fibrillin due to failed Ca<sup>2+</sup> binding, thus flawed alignment of EGFDs and intracellular retention of this intentionally defective fibrillin, similar to intracellular retention of genetically defective fibrillin in Marfan. We aimed to Identify a pioneer fibrillin-1 inhibitor that disrupts fibrillin synthesis preferentially in cancer cells, blocks any fibrillin contribution to the biology of solid tumors and has potential for clinical testing in ovarian serous and uterine serous cancers (USC). <b>Methods:</b> Crystal structure-guided modeling; cultured cell lines with high genomic fidelity to tissue and validation in drug discovery (MRC5 [normal fibroblasts, NF], KURAMOCHI [HGSOC], ARK1 [USC]); RNA-seq analysis; multi-channel flow cytometry. <b>Results:</b> To identify a medically actionable AspH suppressor for testing our hypothesis, we measured the inhibition-pertinent 3D parameters of the HAG mechanism, known to resolve 2OUD catalysis at the scale of atomic orbitals (https://www.researchgate.net/publication/7621997). We applied this HAG parameter grid to the library of FDA-approved drugs and selected deferiprone (DEF) as a candidate AspH suppressor. Globally, thousands of thalassemic children use DEF for a lifetime to alleviate transfusional iron overload. <i>In silico,</i> we found DEF accesses the catalytic pocket of AspH and blocks the two coordination sites at its iron atom that are essential for utilizing molecular oxygen during EGFD hydroxylations in fibrillins. <i>In vitro,</i> ARK1 transcribes <i>FBN1, FBN2,</i> and <i>FBN3.</i> MRC5, KURAMOCHI, and ARK1 produce fibrillin-1, detected by the FITC-labeled monoclonal 11C1.3, specific for a unique epitope within residues 723-909 of fibrillin-1 (exons 18-20 of <i>FBN1).</i> Employed at clinically relevant concentrations (≤ 150 µM), DEF treatment caused dose-dependent intracellular accumulation of 11C1.3-reactive fibrillin-1 material in all cell lines, exceeding untreated controls by up to five-fold after 48 hrs (p≤0.05). After 96 hours on DEF, such accumulation resolved in non-malignant MRC5 but persisted in malignant KURAMOCHI (p=0.04) and increased even further in malignant ARK1 (p≤0.02). Measured simultaneously, cleaved <i>PARP-1</i>, a signature of apoptotic death, was downregulated in DEF-treated MRC5 yet was elevated up to four-fold over controls in DEF-treated KURAMOCHI (p<0.03) and DEF-treated ARK1 (p<0.006). These increases correlated directly with the accumulation of 11C1.3-reactive material (Pearson's r: 0.97 in Kuramochi, 0.99 in ARK1) <b>Conclusions:</b> DEF causes intracellular accumulation of fibrillin-1, mis- or unfolded by deficient hydroxylation, and triggers apoptotic death in HGSOC and USC, but not in NF cells. Our results encourage a DEF pilot trial in HGSOC/USC patients.
Read full abstract