Acute myeloid leukemia (AML) represents a heterogeneous group of malignancies with huge variability in clinical course and response to therapy. Discovery of new prognostic factors in AML is enhancing our understanding of disease biology and helping to identify new therapeutic targets. Previously, we showed for the first time that Discoidin Domain Receptor 1 (DDR1), a class of collagen-activated receptor tyrosine kinase, was highly expressed on bone marrow derived CD33+ leukemic blasts of AML patients, however, its functional role was not known. Here in, we attempt to understand the role of the bone marrow microenvironment in promoting DDR1 activation and its subsequent functional role in migration of leukemic cells. Initial screening of several leukemic cell lines for DDR1 expression identified K562 as an optimal system for further studies due to high DDR1 expression. In addition to native collagen IV, the current studies also tested for the first time the ability of the cryptic collagen IV in activation of DDR1. Proteolytic remodeling of the collagenous extracellular matrix (ECM) have been shown to result in generation of cryptic collagen epitopes, which under normal physiologic conditions, are masked in its 3-dimensional structure. Previously, the functional cryptic site of collagen IV was shown to be highly expressed within the extracellular matrix (ECM) of malignant tumors and within the sub-endothelial basement membrane of tumor-associated blood vessels. Presently, exposure of K562 cells to denatured collagen IV resulted in statistically significant (p<0.045) migration rates compared to native collagen IV. Further, exposure of denatured collagen IV lead to increased activation of DDR1 in comparison to native collagen IV as assessed via phosphorylation of DDR1 at activation sites of Y792, Y796 and Y797 respectively using phospho-DDR1 sandwich ELISA approach. Importantly, co-incubation with HU-IV 26, a previously characterized monoclonal antibody against cryptic collagen IV epitope significantly (p<0.0003) decreased the phospho-DDR1 levels. Interestingly co-incubation with HU-IV 26 did not significantly alter the migration rates between native and denatured collagen IV. To test the role of matrix metalloproteinases (MMPs) in DDR1 induced migration, MMP2 and MMP9 levels were assayed via zymogram and western blots using conditioned media to assess levels of secreted MMPs. Overall, active and inactive MMP-2 levels were higher when exposed to native or denatured collagen IV compared to serum-free only culture conditions. Currently these outcomes suggest a possible feed forward loop for collagen IV-DDR1-MMP2-cryptic collagen IV in promoting migration of leukemic cells. We next examined the levels of collagen IV and cryptic collagen IV in bone marrow derived leukemic blasts of AML patients (n=24) using HU-IV 26 and a native collagen IV antibody via generating a tissue microarray (TMA) and performing immunohistochemistry (IHC). In comparison to control samples (n=9), native collagen and cryptic collagen IV levels were found in significantly high levels (p<0.019 & p<0.034 respectively). Most importantly, the levels of cryptic collagen IV were significantly higher than native collagen (p<0.05) in AML patients. Significantly, the DDR1 levels were concurrently elevated in the same AML patients with high cryptic collagen IV as examined via TMA and IHC. Taken together, for the first time we demonstrate that cryptic collagen IV is a potent activator of DDR1 and high levels of cryptic collagen IV is expressed in leukemic cells of AML patients. Further understanding of this phenomenon may lead to the development of therapeutic agents that directly modulate the bone marrow microenvironment and attenuate leukemogenesis.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.