The role of proprotein convertases in vascular disease

Abstract

To date, the role of the proprotein convertase subtilisin/kexin (PCSK) family in vascular disease is largely unknown, with the exception of PCSK9 and its key function in lipid metabolism. The PCSK family constitutes of 9 different proteases, of which the first 7 show remarkable structural and functional overlap. While the first 7 members have not been thoroughly investigated in vascular disease, they have been extensively studied in cancer. Primarily, they have been shown to proteolytically cleave various substrates, including cytokines, growth factors, hormones and other proteases, which are all typically involved in vascular disease as well. The only PCSK that has received broad attention in the cardiovascular field before the initiation of this Ph.D. project was PCSK9, due to its involvement in lipid metabolism through regulation of the surface expression of the low-density lipoprotein receptor on hepatocytes. Previously, we demonstrated upregulation of PCSK6 in atherosclerotic plaques, particularly those from symptomatic patients, localization to smooth muscle cells (SMCs) in the fibrous cap and positive correlations with inflammation, extracellular matrix (ECM) remodeling and cytokines. In this Ph.D. thesis we aimed to further expand upon this finding and to investigate the expression and localization of the entire PCSK family in vascular disease, as well as their mechanisms and translational potential.

In Study I we hypothesized that PCSK6 could be involved in flow-mediated vascular remodeling. Using various in vivo and ex vivo models, we evaluated the role of PCSK6 in the physiology of this process. Pcsk6-/- and control mice were subjected to carotid ligation of the left common carotid artery, resulting in increased blood flow in the contralateral right common carotid artery. Using ultrasound and ex vivo wire myography, we detected increased flow in the right common carotid artery, concomitant with a progressive increase in vascular circumference in the knockout mice, but not in control mice. Furthermore, these changes were accompanied by a reduction in active tension in knockout mice, while control vessels exhibited an increase in active tension over time. On a cellular level, these changes could be attributed to decreased expression of the typical contractile SMC markers, SMA, MYH11 and LMOD1.

In Study II we aimed to elucidate the mechanistic role of PCSK6 in SMCs in the context of vascular injury, a process leading to the disruption of ECM and release of the matrix-bound growth factors TGF-b1 and PDGF-BB, also implicated in the early stages of atherosclerosis. Using various human biobanks, we investigated the genetic locus surrounding the PCSK6 gene and identified one SNP in particular, rs1531817 to be associated with cIMT progression, a surrogate marker for atherosclerosis. Next, our previously published results were further extended using microarray analysis of end-stage carotid plaques as well as several publicly available datasets, all showing increased PCSK6 expression in diseased tissues. PCSK6 protein was located particularly to SMCs in the fibrous cap, remnants of the media and mature neovessels. In the rat carotid artery balloon injury model, Pcsk6 expression was low in the initial timepoints, but increasing over-time to reach the peak at day 5 after injury. Based on microarray analysis, we uncovered positive correlations on day 5 after injury between Pcsk6 and matrix metalloproteases Mmp2 and Mmp14. Neointima formation in response to carotid artery ligation was found to be decreased in Pcsk6-/- mice compared to control, concomitant with decreased MMP14 activity. Taken together, our results implicate PCSK6 as a key protease in modulating SMC control in vascular injury. Mechanistically, this likely happens through activation of the PCSK6-MMP14-MMP2 axis, allowing for dedifferentiation of SMCs, degradation of the surrounding ECM and subsequent migration out of the media to contribute to the neointima formation.

Expression and localization of PCSK6 was also associated positively with typical markers of T lymphocytes and macrophages in atherosclerotic plaques. Since PCSK6 is highly expressed in healthy human spleens, this together led to a hypothesis for Study III that PCSK6 may be involved in modulating immune responses, particularly in vascular diseases. At baseline, Pcks6-/- plasma showed enrichment of pro-inflammatory cytokines CCL2, CCL3 and in particular IL-17A and IL-17F. Pcks6-/- spleens had an increased number of germinal centers and contained significantly more CD8+ T cells. Microarray analysis of spleens confirmed that expression of T cell markers Cd4, Cd3e and Cd3g was upregulated in Pcks6-/- mice vs. controls. In vitro, splenocytes isolated from Pcks6-/- mice secreted higher levels of IFN-g, IL-2, IL-17 and IL-10 upon stimulation with a-CD3 and a-CD28 antibodies, and were more proliferative. Moreover, peritoneal macrophages from Pcks6-/- mice secreted more TNF-a, CCL2, IL-6 and IL-10 upon LPS stimulation. Bone marrow-derived macrophages from Pcks6- /- mice were also prone to lipid uptake. Finally, in vivo transplantation of Pcks6-/- bone marrow to Ldlr-/- mice led to increased atherosclerotic plaque burden compared to controls. However, these plaques were more stable, attributed to increased collagen deposition, SMC presence and IL-17 content.

In Study IV we aimed to evaluate the vascular expression and cellular localization of all the members of the PCSK family and assess their therapeutic targeting potential in this context. To first establish a genetic link between the PCSKs and CVD phenotypes, we interrogated publicly available data for any SNPs surrounding the PCSK genetic loci. With the exception of PCSK4, the genetic locus surrounding every PCSK contained SNPs associated with CVD. To further assess the expression and localization of PCSKs within tissue biopsies of pathologies underlying cardiovascular disease, we interrogated three large-scale, independent vascular disease biobanks. We found that PCSK1, PCSK2, PCSK4 and PCSK9 were expressed only in moderate levels in disease biopsies and did not observe differences in expression between healthy and diseased tissues. In carotid plaques vs. control arteries FURIN, PCSK5, MBTPS1 were downregulated, while PCSK6 and 7 were upregulated. In abdominal aneurysms, FURIN, PCSK5, PCSK7, MBTPS1 were downregulated, while PCSK6 was enriched in diseased media. In thoracic aneurysms, only FURIN was significantly upregulated. Cell type correlation analyses and immunohistochemistry showed that PCSK transcripts and protein levels parallel each other, except for PCSK9 where the transcript was not detected, while protein was abundant in vascular biopsies. Integrating our results lead to the development of a novel ‘molecular’ 5D framework for early evaluation of target druggability potential. Our results using this translational pipeline revealed primarily PCSK6, followed by PCSK5, PCSK7 and FURIN, as proprotein convertases with the highest novel therapeutic potential.

Overall, in this thesis we have identified PCSK6 as one of the key players in vascular disease and assessed in detail its function and mechanisms in vascular injury, atherosclerosis and aneurysms. Moreover, we evaluated the expression and localization of the entire PCSK family in various vascular disease pathologies. From a translational perspective, a novel druggability scoring system or ‘molecular’ 5D framework was constructed, aimed at early molecular evaluation of potential novel drug targets in the context of cardiovascular disease.