Mechanistic and structural insights into a new pro-metastatic collagen glucosyltransferase

PROJECT SUMMARY-ABSTRACT. Lung cancer is the primary cause of cancer death in western countries, including the United States. Kentucky leads the nation both in lung cancer incidence and mortality. Lung cancer-related mortality is mainly due to metastasis. During metastasis, malignant tumors accumulate collagen fibrils to drive cancer cell dissemination and impede the influx of anti-tumor immune cells. The long-term goal of our lab is to determine the roles of collagen and collagen-modifying enzymes in metastasis to inform new therapeutic development. Toward that goal, we found that the collagen lysyl hydroxylase 2’s long alternatively spliced isoform b (LH2b), carrying an extra motif encoded by exon 13a, is a previously undetected collagen glucosyltransferase (GGT). GGT plays a major role in LH2’s pro-metastatic activity. High expression levels of LH2b predict a worse prognosis, while the expression levels of LH2’s short alternatively spliced isoform a (LH2a) without exon 13a are not prognostic. These findings provide a new basis for the observation made by many others that high LH2 expression has poor-prognostic value and drives the growth and metastasis of multiple cancer types, including lung cancer, breast cancer, and sarcoma. Thus, LH2 GGT is a new therapeutic target of interest. However, how LH2 GGT drives cancer metastasis is unclear and LH2 GGT inhibitor is not available, deficiencies that can be addressed only by gaining mechanistic insight into the structure and function of LH2 GGT. To elucidate how LH2 GGT enzymatic activity is structurally regulated, we solved 3 collagen GGT structures including LH2 GGT and its mimiviral homolog. Our results suggested that the long alternatively spliced isoform LH2b binds UDP-glucose cooperatively and is a disulfide-linked dimer with a continuous collagen-binding site. Based on these findings, we hypothesized that LH2b disulfide-linked dimer allows it to bind collagen and UDP-glucose, modify tumor stroma and drive metastasis. We will test this hypothesis with 3 aims. In the first aim, we will establish a causal relationship among LH2b exon 13a inclusion, disulfide and dimer formation, and UDP-glucose cooperative binding. We will determine the basis of LH2b’s UDP-glucose binding cooperativity. In the second aim, we will define the mechanistic basis of LH2b’s collagen binding and glucosylation. In the third aim, we will determine the structural features of LH2b that drive tumor metastasis. We will introduce dissect-of-mechanism mutations into LH2b to gain mechanistic insight into how LH2b binds UDP-glucose and collagen, glucosylates collagenous tumor stroma, and drives metastasis. The finding from this work will advance the understanding of how LH2 promotes metastasis and identify the structural features of LH2 GGT that drive cancer progression. Such insights will inform LH2 GGT cancer biology and inhibitor development.