Elucidating pro-metastatic collagen modifying activities of lysyl hydroxylase 2

As they progress, malignant tumors accumulate cross-linked collagen fibrils that enhance matrix stiffness, thereby activating collagen receptors that trigger cancer cell invasion and dissemination. We previously reported that a collagen modifying enzyme called lysyl hydroxylase 2 (LH2) is highly expressed in metastatic

lung cancer cells and promotes metastasis by increasing the amount of a particularly stable type of collagen cross-link called hydroxylysine aldehyde-derived collagen cross-link (HLCC). Although LH family members (LH1-3) have highly conserved LH and glucosylgalactosyltransferase (GGT) domains, LH2 reportedly lacks

GGT activity and is unique among LHs in its ability to hydroxylate lysine (lys) residues on collagen N- and C- termini (“telopeptides”) that are required to generate HLCCs. The structural basis for LH2's unique functional properties remains unclear. The studies proposed herein will address this crucial knowledge gap. On the basis

of collagen LH and GGT domain crystal structures that we recently solved, we show that LH2 has telopeptidyl LH (t-LH) activity owing to a unique basic residue cluster that generates electrostatic interactions with acidic residues on collagen. Furthermore, by using a new collagen GGT activity assay we developed that is more

sensitive than ones reported previously, we show that LH2 has GGT activity owing to an alternatively spliced exon 13a-encoded loop that resides at the entrance of the GGT active site. We show that LH2 isoforms that lack (LH2a) or contain (LH2b) exon 13a are differentially expressed in the TCGA lung cancer cohort, and that

LH2b is the predominant isoform expressed in an orthotopic KMLC model in which LH2 promotes metastasis and causes widespread alterations in intra-tumoral immunity. We developed defined collagen matrices that are deficient or replete in LH2-mediated HLCCs (total or glucosylated) and used these matrices as tools to show

that HLCCs influence lung cancer cell behaviors. On the basis of these preliminary results, we postulate that LH2 drives lung cancer metastasis through dual (LH- and GGT-mediated) collagen modifications and will test this hypothesis by completing 3 specific aims. Aim 1: To demonstrate a causal relationship between LH2's

electrostatic interactions with collagen, HLCC formation, and lung cancer metastasis. Aim 2: To demonstrate a causal relationship between inclusion of LH2's exon 13a, collagen glucosylation, and metastasis. Aim 3: To demonstrate a causal relationship between LH2's dual (t-LH- and GGT-mediated) collagen modifications and

LAIR-1-mediated immunosuppression. In summary, the novelty of our proposal rests in an hypothesis that is based on unique insight into LH2's dual enzymatic activities and the tools we developed to generate that hypothesis (e.g., crystal structures, enzymatic assays, and defined collagen matrices). Findings from these

studies will provide a basis for future testing of selective LH2 antagonists that we have already identified from high-throughput screens and have begun to optimize outside the scope of this application.