Mechanics and Tissue Remodeling Integrating Computational and Experimental Systems (MATRICES)

National Institute Of Biomedical Imaging And Bioengineering, Bethesda

Investigators

Abstract

In year three of the MATRICES lab, now that we have been up and running with equipment, focus shifted to hiring and developing our scientific workforce to conduct our studies on sickle cell disease, proteolytic networks and tissue remodeling, and cancer metastasis. Project 1: Sickle Cell Disease Mediated Arteriopathy We have been testing the central hypothesis that inflammation and disturbed flow caused by sickle cell disease (SCD) mediates damage to the large arteries and promotes cell aggregation and adhesion that can further disturb flow, leading to the development of aneurysms, stenoses, and weakened artery walls that progressively elevate stroke risks. By measuring expansive artery remodeling, we found thinning of the medial layer of carotid arteries and imaged this longitudinally with magnetic resonance angiography (MRA) in a mouse-specific manner to reduce subject to subject variability. Further, we found that early bone marrow transplants (BMT) in mice at 2 months of age were protective and prevented this expansive remodeling, but if waiting until 4 months for BMT, the arterial damage had accumulated and was not resolved, even when mice were no longer generating sickling RBCs. This work was published in Science Translational Medicine this year with the title “Bone Marrow transplant protects mice from sickle-cell mediated large artery remodeling.” From the magnetic resonance imaging collected, we also began a collaboration to apply radiomics analysis to carotid arteries. We applied radiomics analysis of the MR data to extract quantitative imaging features and test if radiomics could differentiate between carotid arteries from SCD mice (SS) and heterozygous littermate controls (AS). 112 radiomic features extracted from segmented carotid artery images using PyRadiomics software were used to develop models predictive of SCD genotypes. We found that at one-month of age, four radiomics features yielded accuracy of 74%. At three-months, a single feature achieved 76% accuracy. Histological review confirmed that incorrectly identified mice carotid arteries resembled the incorrectly predicted genotype: larger luminal areas for AS and smaller luminal areas for SS, reflecting the biological variability impacted radiomic feature predictions. This study demonstrates the feasibility of radiomics in discriminating arterial features between SCD and control mice and may offer non-invasive and translational approach to assess arterial changes in SCD and motivates validating these findings clinically for prognostic utility in SCD management in humans. This paper is currently under review for publication. We are also investigating mechanisms of cathepsin K mediated arteriopathy in sickle cell disease with a double transgenic mouse we generated that has sickle cell disease but is null for cathepsin K. The working hypothesis is that sickle cell disease-mediated inflammation upregulates cathepsin K, which drives accelerated elastin and collagen degradation in the artery walls, but that a mouse null for cathepsin K but still with sickle cell disease would be protected from this accelerated remodeling and damage. From our results that we are compiling and preparing for publication, absence of cathepsin K is more protective in male mice than female mice although the absence of cathepsin K preserves collagen, reduces elastin breaks, and preserves the perimeter of the artery closer to that of wildtype arteries without sickle cell disease. The protective effects in the absence of cathepsin K were also reflected in biomechanical testing of the arteries indicating cathepsin K was mediating elastin and collagen damage that increased compliance pathologically, risking hemorrhage or rupture, but in its absence, arteries retained mechanical properties closer to that of AA or AS arteries. These data indicate a role for targeting cathepsin K pharmacologically or with the advent of gene therapy for sickle cell disease, might suggest additional modifications to be made to the graft prior to transplant. Project 2: Computational Fluid Dynamics Reveals Flow Disturbances in Carotid Arteries of Mice with Sickle Cell Disease Disturbed blood flow is a biomechanic influence that, independent of sickle cell disease, causes arterial remodeling. Hemoglobin polymerization inside of red blood cells due to the sickle cell mutation causes changes to density of red blood cells and can alter hemodynamics. Additionally, the adhesiveness of the cells can also stick to endothelium transiently that can locally disturb flow and initiate remodeling and proteolytic programs in the artery wall. To investigate how changes in the geometries of arteries due to sickle cell disease causes regions of disturbed flow, we used magnetic resonance angiography (MRA) to longitudinally image the common carotid arteries of Townes sickle cell transgenic mouse models of wild-type (AA), sickle-trait (AS), and sickle cell (SS) genotypes as they age. We then used RadiAnt DICOM Viewer, Materialise Mimics, and Materialise 3-Matic to reconstruct the MRA data into 3-dimensional finite element mesh models of the common carotid artery. These models were then simulated with SimVascular using computational fluid dynamics (CFD) under steady-state, no-slip boundary conditions. Using ParaView, blood flow and wall shear stress (WSS) profiles were visualized to show regions of flow recirculation over time. Arterial geometries were analyzed using centerlines to determine how changes in cross-sectional area due to stenoses or curvature affected hemodynamics. Computational fluid dynamics (CFD) and hemodynamic analysis was completed on SS and AS mice at 4, 12, and 24 weeks of age and our key findings are that SS mice have increased surface areas of low wall shear stress even at the straight parts of the internal carotid arteries that are deemed protected from disturbed flow and subsequent remodeling programs in arteries without sickle cell disease or atherosclerosis. Further, although the mean velocities were higher in the internal carotid artery (ICA), the highest maximum velocities were found in the anterior (ACA) and middle cerebral arteries (MCA). were lower time average mean maximum velocities (TAMMV) in the SS cerebral arteries, although there were regions of the highest velocities identified in the ACA or MCA of SS mice. This work is being submitted for publication. We are currently completing paperwork and approvals to obtain human MRA scans to apply CFD analysis to validate these results in human patients. Project 3: Mechanisms underlying sickle cell bone disease and sex-specific differences in severity A collection of bone pathologies, including osteonecrosis, osteoporosis, osteopenia and vaso-occlusive bone pain, known as sickle bone disease (SBD) are among the most common complications of SCD, which progresses from adolescence and occurs in 50% of individuals by age 35. Nearly 41% of adults with SCD have bone loss due to osteonecrosis, particularly in the femoral or humeral head. We recently established the Townes’ sickle mouse as a model system for the study of SBD pathologies and identified several trabecular bone features that distinguish the male and female SBD phenotype. We have been working to determine the role of cathepsin K on sex dependent femoral head damage and trabecular bone loss in the femurs using our double transgenic mouse with sickle cell disease but null for cathepsin K. This work involves micro-CT analysis of femurs, and study of the isolated mouse femurs for osteoclast presence, activity, bone quality, and mechanical testing. Project 4: Proteolytic networks of cathepsin cannibalism interactions Our project continues investigating the system of cysteine cathepsins working with and against each other in proteolytic networks to better enable the deployment of small molecule inhibitors that have been failing human clinical trials, and can be helpful to inhibit cathepsins in diseases such as breast cancer and metastatic bone cancer where overexpression of cathepsins is a target. This builds on previous work from our group that used a bioinformatic approach to identify multiple, putative cathepsin cleavable sites on spike protein that was validated with molecular biology studies. With assistance of NCI’s Protein Expression Laboratory, we have generated several cathepsin mutants with site-specific mutagenesis and purified them for our enzymatic and biochemistry studies in the lab. Particularly, we have been testing mutants of cathepsin L that are resistant to cathepsin K cleavage, and that active stie dead cathepsins K and S alter the cleavage of other proteases in the network interactions. Mass spectrometry is also being used to sequence the protein fragments generated by cathepsin-on-cathepsin cleavage to confirm that cleavage sites predicted by our algorithm are being validated biochemically.