FRNK overexpression limits the depth and frequency of vascular smooth muscle cell invasion in a three-dimensional fibrin matrix

Abstract

Pathological vascular smooth muscle cell (VSMC) behavior after vascular interventions such as angioplasty or bypass is initiated within the 3D environment of the vessel media. Here VSMCs proliferate, invade the surrounding matrix, migrate adluminally, and deposit substantial amounts of matrix, leading to myointimal hyperplasia and decreased blood flow to critical organs and tissue. Since focal adhesion kinase (FAK) mediates many of the VSMC responses to these pathologic events, it provides a reasonable pharmacologic target to limit this invasive VSMC behavior and to better understand the cellular pathophysiology of this disease. Here we quantified the effectiveness of disabling FAK in VSMCs with its dominant-negative inhibitor, FAK-related nonkinase (FRNK), in a clinically relevant 3D assay. We found that FRNK overexpression decreased VSMC invasion (both the length and frequency) in this matrix. These effects were demonstrated in the presence and absence of chemical mitotic inhibition, suggesting that FAK's effect on cellular matrix invasion, migration, and proliferation utilize separate and/or redundant signaling cascades. Mechanistically, FAK inhibition decreased its localization to focal adhesions which led to a significant decrease in FAK autophosphorylation and the phosphorylation of the serine/threonine kinase, AKT. Together these findings suggest that disruption of FAK signaling may provide a pharmaceutical tool that limits pathological VSMC cell behavior.

Figure 1. VSMC invasion of this 3-D fibrin matrix

Panel A depicts cellular invasion of either the ECs (1) or VSMCS (2) that were differentially labeled with PKH-26. There is little EC but robust VSMC invasion of this matrix by PKH-26 labeled cells. Panel B has representative fluorescent and light microscope quantification of GFP and GFP-FRNK infected cellular invasion in co-culture assay (1: LM demonstrating inhibition of the length and frequency of VSMC invasion in the GFP-FRNK infected group in comparison to GFP infected control group (section 3); Fluorescent imaging of these same cells is demonstrated in sections 2 (GFP-FRNK group) and 4 (GFP). The distance scale represents 500 μm, and images were taken at 14 days. Panel C quantifies the effect of GFP-FRNK on VSMC invasion using light microscopy. GFP-FRNK overexpression significantly decreased the length of VSMC invasion compared to GFP-infected and uninfected control groups at all time points out to stagnation of invasion (day 20); n=12; P<.02.

Figure 2. Matrix Invasion into fibrin hydrogels by VSMCs, effect of Mitomycin C on cellular invasion

FRNK overexpression decreased the length (A) and frequency (B) of matrix invasion into fibrin hydrogels over the duration of the experiment compared to GFP control (p < .001). Inhibition of matrix invasion by FRNK persisted in MMC treated VSMCs over the duration of the experiment vs. MMC treated GFP controls (p < .016 and p < .005 for length and frequency of invasion, respectively). Also Figure 2 demonstrates that MMC treatment significantly decreased length and frequency of invasion compared to the cells not treated with MMC.

Figure 3. GFP-FRNK infection increases cellular FRNK content and displaces FAK from focal adhesions in VSMCs

GFP-FRNK infection increases cellular FRNK content 3.5 fold (A) without a discernable difference in total FAK levels in GFP vs. GFP-FRNK treated cells. (B) Immunohistochemical staining of FAK VSMCs using a Rhodamine-tagged secondary antibody that recognizes a FAK epitope not present on FRNK (first column). FAK's normal localization to focal adhesions is demonstrated in the GFP-infected VSMCs (Top row), but it is markedly decreased and less well localized in GFP-FRNK infected VSMCs (Bottom row). In the GFP-channel (middle column), GFP expression is noted diffusely throughout the GFP-infected VSMCs , while GFP-tagged FRNK localizes to focal adhesions (E) in a similar fashion as FAK localized to focal adhesion complexes in the 1^st column. Uninfected cells are demonstrating normal FAK localization. The last column (online supplement in color) demonstrates that co-localization (yellow) of endogenous FAK and GFP-FRNK is a rare event.

Figure 4. GFP-FRNK decreases FAK autophosphorylation and AKT phosphorylation

Representative Western Blot analysis of FAK and AKT levels in GFP and GFP-FRNK infected cells. Bar graph (A) and Western blot (B) illustrating the significant decrease in FAK autophosphorylation in VSMCs after GFP-FRNK overexpression (P<.002) that was not dependent on changes in the level of FAK expression between groups (P>.1). Phosphorylated AKT in FRNK overexpressing cells was decreased to 6.4% of GFP controls when normalized to total AKT levels (p<.001) even though total AKT levels in VSMCs over expressing FRNK were 2.7 fold greater, this was not significant (P=.2). Error bars represent standard deviations from the results of two independent experiments.

Figure 5. α-SMA is not increased in GFP-FRNK infected VSMCs

α-SMA levels were investigated to determine whether GFP-FRNK increased these levels as would be expected if its mechanism was through maintaining or switching VSMCs into the contractile phenotype. The relative α-SMA expression as normalized to percent uninfected VSMCs are shown in part A. Part B shows a representative Western blot comparing uninfected VSMCs with GFP and GFP-FRNK infected VSMCS.