MEK/ERK pathway mediates cell-shape-dependent plasminogen activator inhibitor type 1 gene expression upon drug-induced disruption of the microfilament and microtubule networks.

Abstract

Changes in cellular morphology induced as a consequence of direct perturbation of cytoskeletal structure with network-specific targeting agents (i.e. microfilament- or microtubule-disrupting drugs) results in the stimulated expression of a specific subset of genes. Transcription of c-fos, collagenase, transforming growth factor-beta, actin, urokinase plasminogen activator and its type-1 inhibitor (PAI-1) appears to be particularly responsive to shape-activated signaling pathways. Cytochalasin D (CD) or colchicine treatment of contact-inhibited and serum-deprived vascular smooth muscle (R22) cells was used, therefore, as a model system to evaluate morphology-associated controls on PAI-1 gene regulation in the absence of added growth factors. PAI-1 transcript levels in quiescent R22 cells increased rapidly and in a CD-concentration-dependent fashion, with kinetics of expression paralleling the morphological changes. Colchicine concentrations that effectively disrupted microtubule structure and reduced the cellular 'footprint' area (to approximately that of CD treatment) also stimulated PAI-1 synthesis. Shape-related increases in PAI-1 mRNA synthesis were ablated by prior exposure to actinomycin D. Unlike the mechanism of induction in growth-factor-stimulated cells, CD- and colchicine-induced PAI-1 expression required on-going protein synthesis (i.e. it was a secondary response). Although PAI-1 is a TGF-beta-regulated gene and TGF-beta expression is also shape dependent, an autocrine TGF-beta loop was not a factor in CD-initiated PAI-1 transcription. Since CD exposure resulted in actin microfilament disruption and subsequent morphological changes, with uncertain effects on interactions between signaling intermediates or 'scaffold' structures, a pharmacological approach was selected to probe the pathways involved. Signaling events leading to PAI-1 induction were compared with colchicine-treated cells. CD- as well as colchicine-stimulated PAI-1 expression was effectively and dose dependently attenuated by the MEK inhibitor PD98059 (in the 10 to 25 microM concentration range), consistent with the known MAP kinase dependency of PAI-1 synthesis in growth-factor-stimulated cells. Reduced PAI-1 mRNA levels upon exposure to genistein prior to CD addition correlated with inhibition of ERK1/2 activity, implicating a tyrosine kinase in shape-dependent MEK activation. Src-family kinases, moreover, appeared to be specific upstream elements in the CD- and colchicine-dependent pathways of PAI-1 transcription since both agents effectively activated pp60(c-src) kinase activity in quiescent R22 cells. The restrictive (src-family) kinase inhibitor PP1 completely inhibited induced, as well as basal, ERK activity in a coupled immunoprecipitation myelin-basic-protein-phosphorylation assay and ablated shape-initiated PAI-1 mRNA expression. These data suggest that PP1-sensitive tyrosine kinases are upstream intermediates in cell-shape-associated signaling pathways resulting in ERK1/2 activation and subsequent PAI-1 transcription. In contrast to the rapid and transient kinetics of ERK activity typical of serum-stimulated cells, the ERK1/2 response to CD and colchicine is both delayed and relatively sustained. Collectively, these data support a model in which MEK is a focal point for the convergence of shape-initiated signaling events leading to induced PAI-1 transcription.