The principal finding of this study is the demonstration in vitro
of a mechanistic link between exposure of astrocytes to albumin and the increase in expression of MLCK. This increase in MLCK is mediated partially by the TGFβ receptor, and by p38 MAPK, but not by the smad or rho kinase signaling pathways. The demonstration in vivo
of a parallel time course of albumin extravasation with increased MLCK expression following TBI, the prevention of albumin extravasation by inhibition of MLCK, and the expression of MLCK in astrocytes, provides further evidence for the role of MLCK in the mechanisms leading to BBB compromise following TBI. Further, the identification of a role for the TGFβ receptor and for p38 MAPK in the signaling mechanisms which link albumin to MLCK in astrocytes is also consistent with data which implicate albumin in the mechanisms of epileptogenesis (Cacheaux et al., 2009
; Ivens et al., 2007
) and neuronal injury caused by activated glia (Hooper et al., 2009).
A number of lines of evidence implicate MLCK as a pivotal regulator of cytoskeletal rearrangement regulating endothelial barrier integrity. Studies in multiple organs including lung (Rossi et al., 2007
; Wainwright et al., 2003
; Mirzapoiazova et al., 2010
) intestine, skin (Reynoso et al., 2007
) and brain (Luh et al., 2010
), suggest that phosphorylation of MLC by MLCK is a key step in disruption of the endothelial barrier leading to increased vascular permeability. Previous in vitro
studies in a microvascular endothelial cell line (Afonso et al., 2007
) and an co-culture BBB model (Kuhlmann et al., 2009
) have shown that MLCK activation is sufficient to disrupt endothelial structural integrity, leading to compromise of the BBB produced by either human T cell leukemia-infected lymphocytes (Afonso et al. 2007
) or C-reactive protein (Kuhlmann et al., 2009
). Our finding, that inhibition of MLCK reduces the extravasation of albumin following TBI, is consistent with these studies and the findings in a controlled cortical impact TBI model that such inhibition reduces cerebral edema (Luh et al., 2010
Previous studies of cerebral injury have shown activation of MLCK by oxidative stress due to alcohol (Haorah et al., 2005
), hypoxia (Kuhlmann et al., 2007
) and controlled cortical impact (Luh et al., 2010
). Here, we extend these findings by identifying a signaling mechanism by which this increase in activity may be produced, although we examined changes in expression, not activity of the enzyme. Our data indicate that the effect of albumin on MLCK expression in astrocytes involves the TGFβ receptor, but not the TGFβ smad3 signaling pathway. In contrast to our findings, in a squamous cell carcinoma cell line, TGF-β increases MLC phosphorylation through the canonical smad2/3 signaling pathway (Sinpitaksakul et al., 2008
). In a brain slice preparation, albumin uptake into astrocytes is mediated by the TGFβ receptor (Ivens et al., 2007
). TGFβ receptor II has been shown in conjunction with the smad pathway to activate the downstream TGFβ pathway leading to transcriptional changes resulting in epileptiform discharges. (Cacheaux et al., 2009
; Ivens et al., 2007
). In contrast, our data suggest the smad pathway is not required for the increase in MCLK in astrocytes as the effect of the smad3 inhibitor on MLCK expression was only detected for MCLK210 and only at the highest dose used. It is likely therefore that the signaling mechanisms, which regulate MCLK, may vary according to the organ involved, and the nature of the inciting stimulus.
MLCK can also be activated through the rho kinase pathway (Amano et al., 1996
; Totsukawa et al., 2000
; Park et al., 2010
). MLC has been shown to be involved in hypoxia-induced conformational changes in the lung (Eiznhamer et al., 2004
) as well as TGF-β-mediated changes through Rho kinase (Clements et al., 2005
) leading to breakdown of the alveolar barrier in the lung. Our in vitro
data, which show no effect of inhibition of rho kinase on MLCK expression and partial inhibition of MLCK210 induction only at the highest dose of inhibitor, suggest that this pathway does not play a significant role in the albumin-induced increase in MLCK in astrocytes.
The activation of MAPKs produced by albumin and the link between MAPK activation and MLCK expression in primary culture is consistent both with our previous study of the effects of albumin on glia (Ralay Ranaivo and Wainwright, 2010
). p38 MAPK signaling regulates vascular inflammation and epithelial barrier dysfunction in a radiation induced colitis model (Mihaescu et al., 2010
). ERK2 has been implicated in the pathological and functional deficits following spinal cord injury (Yu et al., 2010
). The downstream pathways which link p38 to MLCK activation in astrocytes are not known, but precedent from an endothelial-astrocyte co-culture system (Kuhlmann et al., 2009
) indicates that a p38-activated increase in reactive oxygen species is linked to increased MLCK activity.
The MLCK inhibitor used in the in vivo
experiments, ML-7, may affect both MLCK isoforms. Both forms of MLCK are regulated by calcium calmodulin and are responsible for phosphorylation of MLC, leading to cell contraction. However, MLCK 210 possesses an amino-terminal extension that displays enhanced interaction with the actin cytoskeleton compared to MLCK 130KDa (Kudryashov et al., 2004
). The endothelial form of MLCK (MLCK210) has been implicated in barrier dysfunction in lung (Rossi et al., 2007
; Mirzapoiazova et al., 2010
; Wainwright et al., 2003
) and microvascular injury (Reynoso et al., 2007
). The functional significance of the increase in both isoforms found in the astrocyte studies is not clear, but precedent from other studies of barrier injury suggest a primary role for MLCK210.
These results suggest that TBI produces a compromise in BBB integrity, allowing extravasation of albumin. Albumin then activates MLCK in astrocytes, leading to further dysfunction of the BBB, consistent with the established role of MLCK in the disruption of vascular barrier integrity in other organs besides the brain. There are a number of limitations with the present study, which will require further investigation. This study does not distinguish between the contribution of stretch-induced tissue injury produced by the primary insult, and the subsequent effects of albumin on MLCK. Second, MLCK is present in multiple cell types including cerebral vasculature endothelial cells, neurons, and astrocytes (Abbott 2002
; Edelman et al., 1992
; Kuhlmann et al., 2007
; Baorto et al., 1992
), and the specific role of different MLCK isoforms in astrocytes forming the BBB leading to compromise of BBB integrity remains to be determined. Last, the study of MLCK in astrocyte primary cultures does not recapitulate the interactions with other cell types possible in co-culture models of the BBB.
Our findings add to other published data, which implicate MLCK in the mechanisms of neurologic injury produced by stroke and TBI. In a controlled cortical impact model, inhibition of MLCK reduced brain edema formation following TBI, although there was no improvement in functional neurologic outcome (Luh et al., 2010
). A reduction in cerebral edema associated with MLCK inhibition has also been reported in a cerebral ischemia model (Kuhlmann et al., 2007
). In cell-based studies of tissue injury, including isolated microvascular endothelial cells (Haorah et al., 2005
), and in an in vitro
BBB model (Kuhlmann et al., 2006
) inhibition of MLCK maintained barrier function.
In summary, our findings are a further step toward identifying specific extra- and intracellular signaling mechanisms by which TBI leads to an increase in MLCK. The identification of the expression of MLCK in astrocytes and the increase in MLCK produced by albumin is consistent with an emerging role for albumin in the mechanisms of neurologic injury after TBI. The demonstration of a specific role for the TGF-β receptor and p38 MAPK signaling in these responses is consistent with the role for these pathways in other mechanisms of neurologic injury or glial activation. Taken together, these data add to evidence for a role for MLCK in the mechanisms of barrier dysfunction in other organs and insults common to critical care medicine (Reynoso et al., 2007
; Rossi et al., 2007
). Understanding mechanisms by which MLCK regulates cytoskeletal integrity and astrocyte function following TBI may advance the development of new therapeutic approaches to the prevention of cerebral edema (Schmidt et al., 2003
), a major determinant of mortality following TBI.