Project Summary
Pancreatic ductal adenocarcinoma (PDAC) is characterized by extensive desmoplasia, resulting in occupancy
of most of the tumor mass by a dense stroma consisting of fibroblasts, macrophages, and extracellular matrix
(ECM). Fibrillar collagens, such as type I collagen (Col I), are major ECM components and were reported to
exert poorly understood stimulatory and inhibitory effects on PDAC growth and malignancy. It was also observed
that patients whose stroma is fibrolytic, with a high content of cleaved collagen fragments generated by matrix
metalloproteases (MMP), show poor overall survival after resection, whereas patients with an inert stroma, rich
in intact collagen fibers, have considerably improved survival. The mechanisms underlying these clinical
observations and the tumor stimulatory or inhibitory effects of Col I were heretofore unknown. Using a unique
culture system, preclinical models, and clinical specimens, we discovered that intact (i) Col I fibers and MMP-
generated cleaved (c) Col I fragments have diametrically opposed effects on PDAC metabolism and growth due
to their differential interactions with the collagen receptor DDR1. Whereas cCol I (and other MMP-cleaved
collagens) activates DDR1 and its downstream effectors, which include transcription factors NF-kB and NRF2,
to stimulate PDAC metabolism, mitochondrial biogenesis, and respiration, iCol I does not activate DDR1 and
instead induces its ubiquitin-dependent proteasomal degradation, thereby inhibiting NRF2 activation and PDAC
metabolism. Our goal is to understand how MMP cleavage enables the binding of cCol I to DDR1 and determine
if binding to DDR1 is required for induction of its degradation by iCol I. To fully understand how cCol I and DDR1
signaling control PDAC metabolism and growth, and why Col I ablation is not as inhibitory as prevention of Col
I cleavage, we will test whether the differential effects of cCol I and iCol I on DDR1 expression and activation
apply to other fibrillary collagens (Col III) and define the underlying biochemical mechanisms that enable DDR1
binding. Next, we will investigate how iCol I triggers DDR1 ubiquitination and proteasomal degradation, and the
role of the newly uncovered DDR1 E3 ligase FBXW2 in this process. These studies will include identification of
the mechanism by which iCol I stimulates DDR1 ubiquitination by FBXW2 and development of ligase-recruiting
tool compounds that induce DDR1 degradation, whose biological effects will be evaluated in cultured cells and
mouse models. This will be followed by studying the role of individual MMP isozymes, expressed in PDAC cancer
cells, cancer-associated fibroblasts, and macrophages, in collagen remodeling and PDAC metabolism and
growth. We will determine the suitability of individual collagen cleaving MMPs as drug targets for the treatment
of cCol I enriched PDAC. We will also determine the mechanisms responsible for MMP induction in the PDAC
microenvironment, as they may provide additional therapeutic targets. The proposed studies will employ cutting-
edge technologies, innovative ECM based PDAC cultures, new genetically modified mouse strains that allow the
evaluation of both primary and liver metastatic PDAC and clinical specimens.