HDAC6 Regulates the MRTF-A/SRF Axis and Vascular Smooth Muscle Cell Plasticity

Visual Abstract

I t is well documented that even highly differentiated mature vascular smooth muscle cells HDACs were not investigated in vascular cells until relatively recently (2,(8)(9)(10). Most of the studies are pharmacology-based using pan inhibitors that do not distinguish individual HDACs. Likely for this reason, reports are often contradictory, even using the same HDAC inhibitor in the same animal or cellular model. For example, trichostatin A (TSA), a pan inhibitor of class I and class II HDACs, was reported to abate neointima in the balloon injury model of rat carotid artery (11). However, in another report using the same model, TSA potentiated neointima formation (12).
Moreover, whereas earlier reports indicated that HDACs 2, 4, and 5 repress SMC contractile gene expression (15), a more recent study showed that TSA treatment diminished SMC contractile proteins in human primary cells (16).
These disparities inspired us to explore differential functions of individual HDACs. We chose the unique HDAC6 from class IIb and HDAC3 from class I for contrast. Whereas HDAC3 is typically chromatinassociated, HDAC6 is essentially a cytosolic protein although its nuclear presence has also been reported (7). Both are implicated in vascular diseases; however, the knowledge of their regulations of SMC marker (contractile) proteins is conspicuously lacking. Our data show that HDAC6 inhibition elevates, whereas HDAC3 inhibition reduces, SMC marker protein levels. With an initial objective to compare these  To determine cell migration, scratch (wound healing) assay was performed as described in our previous report (17). Briefly, SMCs were cultured to a 90% confluency in 6-well plates and then starved over-    Figure 1A, PDGF-BB treatment substantially down-regulated both a-SMA and SMHC in SMCs, as generally observed in the literature (15).
Interestingly, pre-treatment with tubastatin A (5 m) averted PDGF-induced down-regulation of a-SMA, maintaining it at the basal level (no PDGF-BB).
Tubastatin A also increased SMHC, albeit not to a  (17). We thus next determined how HDAC6 and HDAC3 influence these pathogenic phenotypes. Our data show that both tubastatin A and RGFP966 inhibited PDGF-stimulated SMC proliferation and migration ( Figures 1B and 1C). Although the inhibition of SMC proliferation by tubastatin A was slightly more effective than that by RGFP966, PDGFstimulated SMC migration was better attenuated by RGFP966 than by tubastatin A of the same concentration (5 mmol/l).
We also measured the expression of proinflammatory chemokines and cytokines as well as matrix metalloproteinases (MMPs) because their expression levels often change concomitantly with inflammation ( Figure 1D). All these markers constitute      Figures 1 and 2.
Zhang et al. We found that 2 weeks after balloon angioplasty, treatment with tubastatin A decreased IH (measured as the intima/media area ratio) by 40% and increased lumen size by up to 60% compared with vehicle control (Figures 4A and 4B). There was no difference in the overall vessel size measured as EEL length.
These data indicate that perivascular application of tubastatin A effectively mitigated restenosis without causing constrictive remodeling (reduced EEL length). In contrast, RGFP966 slightly increased IH and decreased lumen size, although the changes were not statistically significant ( Figures 4C and 4D). We also tried a different hydrogel (Pluronic gel) to deliver RGFP966, and Apicindin, another commonly used selective HDAC3 inhibitor, but neither experiment resulted in neointimal inhibition (data not shown). Therefore, the results indicate that HDAC6 inhibitor, but not HDAC3 inhibitor, effectively mitigates IH.
Finally, the effect of HDAC inhibition on a-SMA in vivo was determined by immunostaining artery cross sections. Consistent with the in vitro result ( Figure 1A), treatment with tubastatin A increased a-SMA in the arterial wall (including media and neointima) by w30% relative to vehicle control, albeit without statistical significance ( Figures 4E and 4F). On the contrary, treatment with RGFP966 significantly reduced a-SMA by w50%.

DISCUSSION
In the present study, we made an unanticipated finding that inhibition of HDAC6, which is best known as cytosolic localized, enhances SRF nuclear activity and preserves SMC contractile protein levels. Mechanistically, another novel finding was that HDAC6 inhibition increases acetylation and total protein levels of MRTF-A, a powerful co-factor that resides in HDACs. In these studies, SMC proliferation and migration were determined, but SMC inflammation and dedifferentiation were generally left unexplored.
Given our finding that inhibiting HDAC6 (class IIb) up-regulates SMC markers, it is seemingly contradictory that TSA reduced these proteins in a previous study (16). However, our data also showed that inhibiting HDAC3, a class I member, reduced a-SMA, consistent with the previous report. As the outcome of using a pan-HDAC inhibitor represents the sum of the effects mediated by all the targeted HDACs, the gross effect of a pan inhibitor may vary depending on multiple factors (e.g., concentration or treatment duration of the inhibitor). We hence speculate that in this previous study using TSA (16) However, treatment with tubastatin A only reinstated a-SMA that was diminished by PDGF-BB in cultured cells ( Figure 1A) and only slightly increased a-SMA either ex vivo in artery explants (Figure 3) or in vivo in the balloon-injured arterial wall ( Figure 4).