Structure and Function of ROCK Kinases
ROCK in Development
Genetic deletion of either ROCK1 (Shimizu et al. 2005) or ROCK2 (Thumkeo et al. 2003) in mice resulted in similar phenotypes, which helped reveal that a major function of the ROCK kinases in vivo is the regulation of epithelial cell motility. The homozygous deletion of ROCK1 still allowed for the birth of mice at the expected Mendelian ratios, indicating that there were no major problems during growth and development in utero, but newborns had defects in eyelid and ventral body wall closure that gave rise to eyes-open at birth (EOB) and omphalocele (organs such as the liver and gut not being contained with the abdomen) phenotypes, respectively. EOB and omphalocele were also observed in homozygous ROCK2 knockout mice, but in this case a sub-Mendelian incidence of ROCK2-/- mice resulted from defects in the placental labyrinth layer, causing decreased blood flow to developing ROCK2-/- embryos (Thumkeo et al. 2003, 2005). Given these results, it seems consistent that ROCK1+/-; ROCK2+/- double heterozygous mice also exhibited EOB and omphalocele, indicating that both kinases contribute to the same actin-driven movement and reorganization of epithelial sheets for eyelid and ventral body wall closure (Thumkeo et al. 2005). That being said, homozygous ROCK1 knockout mice were also independently generated, showing no obvious phenotypic differences from wild-type littermates, which suggests that strain background differences may contribute to the penetrance of the ROCK1 deficient phenotype. (Zhang et al. 2006).
ROCK in Disease
The ready availability of potent and selective ROCK inhibitors has made it possible to examine whether ROCK kinases are involved in a wide variety of pathological conditions including cancer, hypertension and cardiovascular disease, neuronal degeneration, kidney failure, clotting diseases, asthma, glaucoma, osteoporosis, erectile dysfunction, and insulin resistance. However, the areas that have attracted the most research effort are cancer, hypertension–cardiovascular disease, and glaucoma.
Interest in ROCK as a cancer target stems from its wide range of activities that contribute to the growth and progression of tumors including proliferation, survival, and metastasis (reviewed in Wickman et al. (2010)). Early studies revealed that Rho GTPs are overexpressed in a variety of tumors, consistent with increased signaling through the ROCK pathway being a contributory factor. In addition, elevated ROCK expression has been reported in bladder (Kamai et al. 2003) and testicular cancers (Kamai et al. 2002) and correlates with poor survival (Kamai et al. 2003). Conditional deletion of both ROCK1 and ROCK2 in mouse models of lung cancer and melanoma revealed that loss of both kinases blocked tumor formation, indicating that they have essential and redundant roles in tumorigenesis (Kümper et al. 2016). Large-scale sequencing efforts directed at the identification of genetic alterations in human cancers revealed a number of activating somatic ROCK1 mutations in human tumors and tumor cell lines (Lochhead et al. 2010). Accordingly, conditional activation of a ROCK transgene within the epidermis strongly promoted tumor progression in a model of cutaneous squamous cell carcinoma, while pharmacological inhibition of ROCK activity suppressed it (Samuel et al. 2011). A current concept is that it would be advantageous to develop ROCK2 selective inhibitors for the treatment of cancer to avoid the pronounced hypotension that is a side effect of ROCK1 inhibition.
The connection between ROCK and hypertension was revealed with the development of ROCK selective inhibitors. Inhibition of ROCK with Y27632 and related compounds was shown to relieve hypertension in rats by inhibiting the calcium sensitization of smooth muscle contraction. Since that time, numerous studies have built on these observations to show that ROCK activity mediates increased smooth muscle contraction principally via modulation of MLC phosphorylation. In particular, ROCK appears to contribute to aberrant vascular contraction, for example, during coronary vasospasm, cerebral vasospasm following subarachnoid hemorrhage, pulmonary hypertension, and Raynaud’s phenomenon, a condition in which the blood supply to distal extremities such as fingers and toes is decreased to the point of numbness or pain and which is the result of vasospasms. Consistent with a role for ROCK in the regulation of vascular contraction and blood pressure, ubiquitous expression of a conditionally activated ROCK2 transgene resulted in cerebral hemorrhagic lesions (Samuel et al. 2016).
Glaucoma is a disease in which damage to the optic nerve progressively leads to impaired vision and possibly blindness. One way that optic nerve damage may occur is through elevated intraocular pressure. ROCK inhibition helps to relieve this pressure by increasing aqueous outflow by reducing MLC phosphorylation in cells lining the trabecular meshwork. Significant research by pharmaceutical companies in this area has resulted in several promising clinical trials.
Within the nervous system, ROCK has been shown to be an important trigger of neuronal growth cone collapse and neurite withdrawal. As a result, there has been considerable interest in the possibility that ROCK inhibition would actually promote neurite outgrowth and dendrite formation. Possible applications for this include recovery from spinal cord injury by assisting the reestablishment of neural connections across the lesion and Alzheimer’s disease treatment through the decreased production of amyloid precursor protein.
ROCK and Stem Cell Survival
When human embryonic stem cells (hESC) are dissociated or plated at low density, they often undergo apoptotic death. Chemical biology screens to identify agents that would promote survival identified the ROCK selective inhibitor Y27632 as a particularly potent agent. Mechanistic studies revealed that ROCK-mediated actin–myosin contraction makes epiblast-derived hESCs die during low density growth and validate the use of inhibitors of ROCK or actin–myosin contractility for the propagation and genetic manipulation of hESCs for eventual therapeutic use (Samuel and Olson 2010).
HA-1077 (fasudil) and hydroxyfasudil have been in clinical use in Japan for cerebral vasospasm since 1995 (Olson 2008). Since it has been used for such a long period, there is positive post-marketing safety data, which has encouraged trials for a number of indications including angina, acute ischemic stroke, cerebral blood flow, stable angina pectoris, coronary artery spasm, heart failure-associated vascular resistance and constriction, pulmonary arterial hypertension, essential hypertension, atherosclerosis, and aortic stiffness (Olson 2008).
Y-27632 was the first published selective ROCK inhibitor (Uehata et al. 1997), and its ready availability has made it the inhibitor of choice. Although this inhibitor is not strictly specific for ROCK kinases, ROCK1 and ROCK2 siRNA experiments have not revealed significant off-target effects in cells.
H-1152 was developed as an improved version of HA-1077 with greater ROCK selectivity over PKA and PKC. Although also readily available from commercial sources, H-1152 is less often used than HA-1077 to corroborate results from experiments in which Y-27632 was used, despite the improved selectivity.
KD025 (formerly known as SLx-2119) was developed as a potential cancer therapeutic initially but more recently has gone into clinical development for the treatment of diseases including psoriasis vulgaris, chronic graft-versus-host disease, and pulmonary fibrosis (Zanin-Zhorov et al. 2016).
AT13148 was originally developed as an AKT inhibitor but was found to be a broad specificity inhibitor for several related AGC family kinases. Recent research has shown that the high potency of AT13148 for ROCK1 and ROCK2 is chiefly responsible for the antitumor properties of this compound using in vivo models (Yap et al. 2012). AT13148 has entered Phase I clinical trials in patients with advanced solid tumors.
Structurally unrelated inhibitors should produce the same biological effects at concentrations that produce equivalent kinase inhibition. The lowest effective doses should be used to reduce off-target effects.
Where possible additional methods should be used to inhibit ROCK function, such as RNAi-mediated knockdown or CRISPR–Cas9-mediated gene disruption, and the biological effects of inhibitors should be consistent with those observed using these alternative approaches.
Dose–response experiments to establish rank order of potency for a set of inhibitors, i.e., the most potent ROCK inhibitors should be the most effective if a biological response is mediated by ROCK.
Examination of the relationship between ROCK inhibitor dose, substrate phosphorylation, and biological endpoint.
Although there is no doubt that inhibitors are useful and convenient research tools, care should be taken in interpreting the results. The substantial knowledge base of the biological functions of ROCK has been made possible due to the ready availability of such inhibitors. Their greatest utility is actually in excluding a possible involvement of ROCK in specific biological responses when there are adequate positive controls in place.
The most important and central cellular function of ROCK kinases is to regulate cell morphology, largely through actin–myosin contractility. Due to its profound influence on morphology and contractility, ROCK directly influences numerous activities, such as cytokinesis, adhesion, motility, endothelial barrier function, and membrane blebbing. In addition, via direct or indirect pathways, the ROCK kinases also influence biological processes including gene transcription, proliferation, regulation of cell size, and survival. There has been considerable interest in ROCK kinases as potential therapeutic targets for cancer, hypertension–cardiovascular disease and glaucoma, and a number of potent and selective inhibitors have been discovered. In fact, one ROCK inhibitor fasudil has been used clinically in Japan for a number of years for the treatment of cerebral vasospasm. Although there is a substantial literature on ROCK function, largely generated due to the availability of pharmacological inhibitors, greater knowledge at the tissue and organismal levels will result from conditional knockout (Kümper et al. 2016) and conditional-activation mouse models (Samuel et al. 2016, 2009). In vivo experiments using these types of genetically modified models will validate the role of ROCK in various pathological conditions and will highlight additional indications for the use of ROCK inhibitors. Although quite a few ROCK substrates have been identified and well characterized, recent phosphoproteomic studies have identified a large number of previously unknown ROCK substrates. As a result, a significant opportunity awaits to characterize the biological outcomes of ROCK-mediated phosphorylation on these novel substrates and ultimately to determine their possible contributions to human disease.