The cardiovascular effects of proteolysis of high molecular weight basic fibroblast growth factor by inflammatory serine proteases

Cardiovascular disease is characterized by structural and functional changes of blood vessel’s wall that lead to reduced blood flow and eventually occlusion. The integrity of the vascular wall is maintained by homeostatic mechanisms controlled by the endothelium. Stress caused by age, oxidants, mechanical injury, and inflammation can result in endothelial dysfunction leading to remodeling of the vessel wall. We found that thrombin, a key protease of the coagulation cascade and inflammatory response, cleaves the high molecular weight (HMW) forms of basic fibroblast growth factor (FGF-2), a ubiquitous protein with trophic effects on vascular cells. The C-terminal fragment of FGF-2 generated by thrombin is similar to low molecular weight (LMW; 18 kDa) FGF2, and induces vascular cell activation of the mitogen-activated protein kinases ERK-1/2, migration and proliferation. The N-terminal fragment generated by thrombin cleavage of HMW FGF-2 contains a sequence rich in asymmetric-dimethyl-arginine (ADMA) residues. In free form ADMA inhibits key reactions for blood vessel homeostasis such as nitric oxide synthesis, and its serum levels, elevated in diabetes, renal failure, hypertension, and hypercholesterolemia, correlate with a poor prognosis in cardiovascular patients. We found that thrombin cleavage of HMW FGF-2 dramatically upregulates intracellular ADMA levels in cultured cells. Thus, the C-terminal cleavage product of HMW FGF-2 can activate intracellular signaling and control vascular cell functions, while the N-terminal fragment of FGF-2 generates ADMA, a powerful inhibitor of nitric oxide synthesis. We hypothesize that upon vascular injury HMW FGF-2 is processed by inflammatory serine proteases such as thrombin generating molecules that accelerate the development of intimal hyperplasia. This novel mechanism implicates human HMW FGF-2 in the pathogenetic mechanisms of vascular injury occurring in hypertension, diabetes, and dyslipidemia, conditions that are all characterized by elevated serum levels of ADMA. The elucidation of these mechanisms will foster the development of new pharmacological tools for the treatment of the cardiovascular disorders associated with these diseases. Fgf-2 gene expression Basic fibroblast growth factor (FGF-2) is the prototypic member of a large family of proteins with pleiotropic effects initially identified as an 18 kDa protein [1]. Analysis of the human cDNA sequence upstream of the 5’ AUG revealed the existence of at least three (two in rodents) additional CUG initiation sites located on the same mRNA. These alternative initiation codons originate 22, 22.5, and 24 kDa (20.5 and 21 kDa in rodents) forms of FGF-2 referred to as high molecular weight (HMW) FGF-2. An additional 34 kDa form, generally poorly translated, is synthesized under cellular stress conditions [2; 3; 4]. HMW FGF-2 forms are, therefore, colinear extensions of 18 kDa, or low molecular weight (LMW) FGF-2. The various FGF-2 forms are differentially distributed in the cell, which might explain in part their differential biological activity [5]. LMW FGF-2 is mostly cytosolic whereas HMW FGF-2 forms predominately localize to the nucleus; independent of their intracellular localization, all FGF-2 forms are released and associate with the extracellular matrix despite the lack of a classical signal peptide that directs secretion through the ERGolgi pathway [6; 7; 8; 9]. LMW FGF-2 promotes Giuseppe Pintucci The cardiovascular effects of proteolysis 22 cell proliferation and migration by interacting with its cell membrane receptor(s), whereas HMW FGF-2 can induce cell transformation via a receptor-independent mechanism [3; 10]. Furthermore, the effects of HMW FGF-2 appear to depend on their endogenous levels of expression and are cell type-specific; in fact, HMW FGF-2 has also been shown to signal growth arrest [11; 12]. In addition, unlike the LMW form, HMW FGF-2 inhibits cell migration [7; 13; 14]. Several reports have described the presence of HMW FGF-2 in the extracellular matrix; although alternative mechanisms of FGF-2 release from intact cells have been proposed [9; 15; 16; 17] cell damage such as in tissue injury or disruption remains the major mechanism of release for all FGF-2 forms (reviewed in [5]). HMW and LMW FGF-2 may therefore serve different functions by controlling each other’s biological activity depending on their relative concentration and/or localization [12; 18]. The observation that a number of growth factors can be targeted to the nucleus with or without their receptors has generated the hypothesis that FGF-2 may act through alternative mechanisms of action in addition to its interaction with cell membrane receptors [19; 20; 21]. Fig. 1. The N-terminal sequence of HMW FGF-2. The initiating amino acids for each FGF-2 form (L for HMW and M for LMW FGF-2) are underlined (apparent MW are represented below each line). Boldfaced R’s represent arginine residues positively identified as targets of methylation. Thrombin cleavage sites are identified with a red arrow. FGF-2 is implicated in a variety of physiological and pathological conditions including angiogenesis, tumor growth, and vascular stenosis secondary to injury [30]. During injuryinduced vascular remodeling, such as after percutaneous transluminal coronary angioplasty or coronary artery by-pass grafting, FGF-2 appears to promote intimal hyperplasia, the major cause of long-term failure following vascular interventions [31; 32]. The relative concentrations of LMW and HMW FGF-2 are mostly controlled at the translational level [22; 23]. Conditions like stress, exposure to catecholamines or estrogens alter the balance of FGF2 expression in favor of the HMW forms [23; 24; 25]. We have shown that PDGF-BB induces expression and nuclear accumulation of HMW FGF-2 in vascular smooth muscle cells (VSMCs), an effect controlled by the mitogen-activated protein kinases (MAPKs) ERK-1/2 [18]. Interestingly, hyperglycemia has also been shown to increase HMW FGF-2 expression [26]. The existence of these mechanisms suggests that HMW FGF-2 may have an as yet unidentified role in vivo. The Nterminal extension of HMW FGF-2 is uniquely endowed with characteristic arginine-glycine (RG) or arginine-glycine-glycine (RGG) motifs that are methylated by specific protein arginine methyltransferases (PRMTs); such post-translational modification controls their nuclear localization [27; 28; 29]. 24 kDa FGF-2 contains eight arginine methylation sites whereas 22 kDa FGF-2 contains five (Fig. 1) [29]. Thrombin cleaves the HMW forms of