Tight junction dysregulation: Mechanisms and druggability

 

Tight junction barrier loss is present in many diseases of the intestine and other epithelial organs (Nalle and Turner. 2015. Mucosal Immunol). However, the contributions of barrier loss to disease pathogenesis and potential of barrier-restoring therapies are not well defined (Odenwald and Turner. 2013. Clin Gastroenterol Hepatol). Nevertheless, data from our group and others indicate that the pore and leak pathways are both activated in disease as a result of claudin-2 upregulation and myosin light chain kinase activation, respectively.

 

One particularly exciting discovery is that dephosphorylation of a specific site (S408) within the cytoplasmic tail of occludin triggers assembly of a complex of occludin, ZO-1, and claudin-2 that, ultimately, inhibits claudin-2 function (Raleigh et al. 2011. J Cell Biol). Inhibition of the kinase (CK2) that phosphorylates S408 is able to restore barrier loss resulting from IL-13-dependent claudin-2 upregulation (Raleigh et al. 2011. J Cell Biol). Our preliminary data further indicate that CK2 inhibition and subsequent downregulation of claudin-2 pore function has therapeutic potential in immune-mediated colitis. Ongoing studies are aimed at better defining these regulatory processes. One outcome of such work may be development of small molecules that promote assembly of the occludin:ZO-1:claudin-2 complex.

 

Like claudin-2-deficient mice, mice lacking the epithelial long isoform of myosin light chain kinase are protected from immune-mediated colitis (Su et al. 2013. Gastroenterology).  These mice lacking intestinal epithelial myosin light chain kinase also fail to upregulate claudin-2 during immune-mediated colitis (Su et al. 2013. Gastroenterology). This suggests that there may be more to the interplay between pore and leak pathways (Turner. 2009. Nat Rev Immunol), and this remains an important focus of our ongoing work.

 

While the effects of epithelial myosin light chain kinase inhibition or genetic deletion knockout are impressive, these are unlikely to be effective therapies. This is because the catalytic domain of epithelial myosin light chain kinase is identical to that of smooth muscle myosin light chain kinase. Any inhibitor of kinase activity will therefore also inhibit smooth muscle myosin light chain kinase and lead to significant toxicities, including hypotension and visceral paralysis. It will, therefore, be important to  develop therapies that inhibit myosin light chain kinase-dependent tight junction regulation without impacting smooth muscle myosin light chain kinase or interfering with essential myosin light chain kinase functions within epithelia, e.g. cell migration and wound healing. Ongoing work suggests that this may be possible by interfering with myosin light chain kinase trafficking to the tight junction and perijunctional actomyosin ring. In addition to defining the underlying mechanisms of such trafficking, we hope to develop small molecule drugs that can block such trafficking.

CK2-mediated phosphorylation of occludin residue S408 regulated tight junction structure, dynamic behavior, and barrier function.  (A) CK2-mediated phosphorylation of S408 enhances occludin self-association, increases the occludin mobile fraction, and reduces occludin association with ZO-1, claudin-1, and claudin-2. This promotes flux across claudin-2 pores, thereby increasing paracellular cation flux. (B) When dephosphorylated at S408, occludin is stabilized at the TJ through enhanced association with ZO-1 via the U5-GuK domain. ZO-1 also facilitates indirect interactions between occludin and claudin-2, which associates with ZO-1 via the PDZ1 domain, thereby acting as a scaffold to organize a complex. As a result, dynamic behaviors of occludin, ZO-1, and claudin-2 converge, and function of claudin-2 paracellular pores is reduced. From Raleigh et al. 2011. J Cell Biol.

 

Transgenic, intestinal epithelial-restricted expression of constitutively active-MLCK (CA-MLCK) restores MLC phosphorylation, caludin-2 upregulation and clinical features of CD4+CD45RBhi colitis in long MLCK-/- mice. Immunofluorescence (green) of phosphorylated MLC (pMLC) and claudin-2 (CLDN2) before or after CD4+CD45RBhi adoptive transfer in indicated genotypes. Nuclei are blue. Bar = 10 um. From Su et al. 2013 Gastroenterology.

Jerrold R. Turner, MD, PhD
Professor of Pathology and Medicine

Brigham and
Women's
Hospital

Harvard
Medical
School

NRB 730
77 Avenue Louis Pasteur
Boston, MA 02115