Vascular tumors are a highly diverse group of aberrant growths which include various benign hemangiomas, borderline malignant hemangioendotheliomas, and malignant hemangiopericytomas and angiosarcomas. Benign vascular tumors display a range of characteristics, from well-defined, non-invasive small vessels to less defined, locally invasive large vessels [
]. These tumors are relatively abundant in the human population, with infantile hemangiomas being the most common tumor in children and cavernous hemangiomas affecting approximately one in every one hundred people. Treatment is not necessary for most benign vascular tumors unless they threaten bodily functions; however radiotherapy and/or embolization have been used with limited success for very large hemangiomas, and beta blockers, which target catecholamine-stimulated beta adrenergic receptor signaling, are considered a highly effective treatment option for pediatric patients with life threatening infantile hemangiomas [
]. In contrast, their malignant vascular tumor counterparts such as angiosarcomas can be highly lethal tumors, and are composed primarily of aberrant lymphatic or vascular endothelial cells [
]. Treatment of angiosarcomas involves radiation, surgery, and neoadjuvant and/or adjuvant chemotherapy with doxorubicin or taxanes, yet the five year survival rate for these patients is abysmally low [
]. Despite their vascular origin, even the addition of novel anti-angiogenic drugs has shown a minimal to absent response in angiosarcoma patients [
], though similar to infantile hemangiomas, beta blockade has recently emerged as a potential therapy against angiosarcomas [
]. Effective treatments are desperately needed to increase the progression free survival or overall patient survival in individuals suffering from this highly aggressive sarcoma.
The Rho associated protein kinases (ROCK) 1 and 2 are serine/threonine kinase protein paralogs identified in the 1990’s as direct downstream effectors of Rho-GTPase signaling and are responsible for regulation of the actin cytoskeleton through phosphorylating numerous downstream targets including LIM kinase, myosin regulatory light chain, and the myosin binding subunit of myosin light chain phosphatase [
]. Since that time, the ROCK paralogs have been shown to be involved in a variety of cellular processes far beyond regulation of cytoskeletal dynamics, including cell proliferation, apoptosis, and cell differentiation [
]. The role of ROCK proteins in cancer development, progression, and metastasis has been well established in the literature. Regulation of ROCK’s kinase activity is altered in many cancers through modulation of these proteins’ activation processes, altered subcellular localization, and disrupted interactions with regulatory molecules [
]. Elevated protein expression of the ROCK paralogs has been reported across several cancers including hepatocellular carcinoma, osteosarcoma, and breast, colon, and bladder cancers, and the expression of ROCK1 has been shown to have strong prognostic value in colorectal, breast, and bladder cancer [
]. Mutations in both ROCK genes have been identified in multiple cancer genomes and some of these mutations result in enhanced kinase activity of the proteins [
]. Given their central roles in regulating major oncogenic processes, inhibition of ROCK activity has shown efficacy against tumors in a large number of pre-clinical studies [
]. The success of these pre-clinical studies has the potential to translate clinically given that small molecule inhibitors targeting the kinase activity of these proteins are currently in the clinical pipeline against solid tumors, including AT13148 from Astex Pharmaceuticals (currently in Phase I clinical trials). In addition to performing central roles in tumorigenesis, ROCK proteins and their associated signaling pathways have been heavily implicated in regulating angiogenesis, including pathological angiogenesis in a variety of tumors [
]. This suggests that not only does inhibition of ROCK activity directly target tumor cell function, but it also limits the blood supply to tumors through disrupting aberrant tumor angiogenesis.
ROCK1 and 2 share a high degree of homology and modulate the activity of many common substrates, however a number of studies have revealed that ROCK1 and 2 additionally play unique and non-overlapping roles in processes such as stress fiber and focal adhesion formation, phagocytosis, apoptosis, inflammation, and multiple aspects of organ and tissue development [
]. Our lab has previously used a combination of silencing RNA (shRNA)-mediated gene expression knockdown and a haplo-insufficient animal model to demonstrate that ROCK1 and 2 play unique and overlapping roles in regulating multiple aspects of endothelial function and angiogenesis, with ROCK2 acting as the dominant paralog in normal endothelial cells [
]. More investigations on the individual functions of the ROCK paralogs are needed to elucidate their underlying mechanisms and to determine the predominant paralog in normal and diseased tissues. In the current study, we examined the protein expression patterns of ROCK1 and 2 in a panel of diverse vascular tumors and subsequently employed a shRNA driven approach to elucidate the role of ROCK1 and 2 in a vascular tumor xenograft model.
Immunohistochemical (IHC) studies were performed on 5 μm thick, formalin fixed, paraffin-embedded sections. These sections were taken from the scrambled control, ROCK1 shRNA, or ROCK2 shRNA xenograft tumors or from a commercially obtained tumor tissue array (US Biomax, Inc.; #SO8010) consisting of 6 cases of angiosarcoma, 2 malignant hemangiopericytomas, 6 borderline malignant hemangioendotheliomas, 6 capillary hemangiomas, 3 granulomatous hemangiomas, 46 cavernous hemangiomas, and 10 normal (aortic or carotid artery) blood vessel tissues. The pathological features of each tumor were confirmed independently by a University Medical Center Pathologist. Sections were deparaffinized, rehydrated, and treated for antigen retrieval using Trilogy (Cell Marque). Nonspecific binding was blocked with background block solution (Cell Marque). Antigens were detected with antibodies purchased from Abcam as follows: ROCK1 (#ab45171), ROCK2 (#ab71598), and Ki67 (#ab15580). Sections were then incubated with the CytoScan Alkaline Phos Detection System (Cell Marque) and detected using the DAB substrate kit (Cell Marque). All slides were counterstained with Hematoxylin. Immunopositivity was quantified blindly using two metrics: the percentage of tissue with positive staining (<25%, 25–50%, 50–75%, or >75%) and the staining intensity (0 = no staining, + = weak staining, ++ = moderate staining, +++ = high staining). IHC scores were determined by multiplying the staining intensity (0 = 0, + = 1, ++ = 2, +++ = 3) by the percent of tissue stained (<25% = 1, 25–50% = 2, 50–75% = 3, >75% = 4) based on previously described methods [
]. For statistical analysis, the Mann-Whitney rank sum test was used. Statistical significance was determined if the two-sided
value of the test was <0.05.
Cell culture and treatment
SVR cells (ATCC; #CRL-2280) were maintained in Dulbecco’s modified Eagle’s media (DMEM) supplemented with 10% fetal bovine serum (FBS), 80 U/ml penicillin, and 50 μg/ml streptomycin C. SVR cells have been used extensively as a model for angiosarcoma given that no reliable human angiosarcoma cell lines are currently capable of forming tumors that recapitulate the human disease [
]. shRNA vectors (SABiosciences) were transfected using Lipofectamine 2000 and cell pools were stably selected using puromycin. The sequences and efficacy of each shRNA and scrambled control vector used in this study have been previously validated by our lab and published [
] (scrambled control: GGAATCTCTCATTCGATGCATAC; ROCK1 shRNA: GCGCAATTGGTAGAAGAATGT; ROCK2 shRNA: AACCAACTGTGAGGCATGTAT). Y-27632 (trans-4-[(1R)-1-aminoethyl]-N-4-pyridinyl-cyclohexanecarboxamide; Santa Cruz Biotechnology) was utilized at 10 μM.
For qPCR, total RNA was purified using the Purelink RNA mini kit (Ambion) and converted to cDNA using the Verso cDNA synthesis kit (Thermo-Scientific). qPCR was performed in triplicate using SYBR Green-based probes against ROCK1 (SABiosciences; #PPM04660B), ROCK2 (SABiosciences; #PPM36940C), or GAPDH (SABiosciences; #PPM02946E). Assays were run on an ABI7900HT real time PCR system (Applied Biosystems).
Angiosarcoma xenograft model
All xenograft experiments were approved by and performed in accordance to Texas Tech University Health Sciences Center Institutional Animal Care and Use Committee (IACUC) regulations for the care and use of animals in experimental procedures (IACUC protocol # 11035), and all efforts were made to minimize suffering. Animals were housed 4 per cage in a temperature-controlled animal facility on a 12 h–12 h light-dark cycle. Animals had free access to chow and water. Xenograft angiosarcoma tumors were generated by subcutaneous injection of 1 × 105 SVR cells (scrambled control, ROCK1 shRNA, or ROCK2 shRNA) into the dorsolateral flanks of 4 week old female mice as previously described [
]. Body weight and tumor volume of the animals were measured once a week to ensure health of the animals. The mice were observed daily for ulceration, abdominal swelling, emaciation and/or other signs of distress, and tumor burden did not interfere with the ability of the mice to move freely. When the scrambled control tumors reached approximately 1 cm in diameter, the mice were sacrificed by CO2 asphyxiation followed by cervical dislocation, and the tumors from all treatment groups were collected and weighed. Statistical significance in tumor weight was determined using an unpaired two-tailed t-test with Graphpad Prism version 6.05.
Hybridization and analysis of the high throughput antibody arrays were performed on tumor lysates using the Phospho-Explorer Antibody Array contract service offered by Full Moon Biosystems (Sunnyvale, CA).
H NMR analysis was performed on tumor lysates using the contract service offered by Chenomx Inc. (Edmonton, Canada). Normalized heatmap data was generated in Cluster 3.0 software (
) using unsupervised hierarchical clustering analysis with an uncentered correlation similarity metric and centroid linkage. Heatmaps were visualized using Java Treeview software (
). Physical and functional associations of the omics data were performed using Metacore Pathway Analysis Software (Thompson Reuters, New York City, NY). For both the proteomics and metabolomics analysis, independent biological samples were tested in triplicate.
Western blot analysis
Protein lysates from the xenograft tumors were subjected to SDS-PAGE and transferred to PVDF membrane. Membranes were blocked using 3% bovine serum albumin and probed with the following antibodies as indicated: p53 (Cell Signaling #2524), Chk1 (Cell Signaling #2360), Fadd (Abcam #ab24533), Nfkb-p105/p50 (pSer337) (ThermoFisher #PA5–37658), Casp6 (Cell Signaling #9762), Nfkb-p65 (pSer536) (Cell Signaling #3033), and actin (Santa Cruz Biotechnology #sc7319). Appropriate secondary antibodies and chemiluminescent detection substrate was used for imaging of the bands.
Previous publications from our lab and others have provided evidence that modulation of cell shape and cytoskeletal dynamics plays a major role in regulating key endothelial processes. For instance, manipulation of endothelial cell shape and actin organization results in gene expression alterations of approximately 8% of the global genome potentially through altering chromosomal boundaries within the nucleus [
]. Specific disruption of the activity of the cell shape regulators RhoA and ROCK in endothelial cells blocks a number of developmental and cellular properties such as angiogenesis, vascular formation during embryogenesis, and lung capillary development [
]. In addition to altering physiological vascular properties, inhibition of ROCK activity leads to anti-angiogenic effects on the capillary networks of gliomas and prostate adenocarcinomas [
], suggesting an effective role for ROCK inhibition as an anti-angiogenic agent against solid tumors. Furthermore, the ROCK proteins play a prominent role in the proliferation, invasion, and metastasis of tumor cells through modulating cytoskeletal dynamics and other cellular processes [
], and a wealth of preclinical studies have demonstrated the efficacy of ROCK inhibition in the treatment of a variety of cancers over the last 1.5 decades [
ROCK1 and ROCK2 are ubiquitously expressed across tissues from early embryonic development to adulthood, though preferential expression of these proteins has been observed in some tissues [
]. Our data revealed that both ROCK1 and 2 protein levels were elevated across vascular tumors relative to normal endothelium. Indeed, ROCK proteins have been found to be aberrantly increased in a variety of more common carcinomas [
]. We suspect that overexpression of ROCK proteins in benign and malignant vascular tumors is a key process whereby these tumors hijack normal cytoskeletal processes to increase invasive and metastatic cell behavior, and therefore may be a selectively preferable therapeutic target whose disruption could prove beneficial to enhancing patient treatment. Thus we hypothesized that targeting ROCK activity may show efficacy against vascular tumors. ROCK1 and 2 shRNA xenograft tumors displayed overlapping and unique roles in both protein expression/modification as well as metabolite concentrations. It has been reported that ROCK proteins display overlapping and unique roles [
], and a handful of reports have implicated ROCK proteins in the regulation of metabolism, particularly regarding insulin resistance and glucose metabolism [
]. Further studies are necessary to identify the unique and overlapping roles of ROCK proteins in these particular metabolic processes, and our data suggests that similar to previously reported paralog-specific transcriptional changes [
] and unique protein expression changes reported in the current study, metabolic targets may be differentially regulated by the ROCK proteins as well. Our animal studies revealed that knockdown of ROCK2, but not ROCK1, greatly reduced xenograft tumor volume in an established xenograft vascular tumor model. We suspect the observed reduction in tumor growth in the ROCK2 knockdown tumors is due, in part, to paralog-specific regulation of cell cycle, survival, and checkpoint modulators that contribute to central processes previously shown to be regulated extensively by the ROCK proteins [
ROCK inhibition has strong promise for effective translation into the clinic for the treatment of angiosarcomas and other solid tumors. Indeed, ROCK inhibitors are currently in use or in clinical trials for a variety of diseases including cerebral vasospasm after subarachnoid hemorrhage, hypertension, atherosclerosis, and aortic stiffness [
]. Though no ROCK inhibitors are currently approved for clinical use in the treatment of cancers, the ROCK inhibitor AT13148 is currently in phase I clinical trials, and several other small molecule inhibitors against ROCK proteins have shown efficacy against carcinomas in preclinical tumor models [
]. While several overlapping roles have been identified for ROCK proteins, interest in selectively targeting each of the ROCK paralogs has recently gained popularity due to the unique roles that are reported in the literature regarding these proteins [
]. Thus, a strategy that utilizes specific targeting of ROCK paralogs in a context dependent manner could lead to efficiency in achieving optimal anti-cancer results in the clinic. Drugs such as the potent selective inhibitor of ROCK2, Slx-2119 [
], may pave the way for future selective inhibition of ROCK-specific paralogs to achieve optimized therapeutic efficacy.
We would like to thank Alireza Torabi, MD (Texas Tech University Health Sciences Center, Department of Pathology) for pathological review of the vascular tumor tissue arrays.
The research was funded by grants to BAB from the National Heart, Lung, and Blood Foundation (R15HL098931), Liddy Shriver Sarcoma Initiative, Sarcoma Foundation of America, and Angiosarcoma Awareness Foundation. The funding bodies had no role in the design of the study, the collection, analysis, and interpretation of the data, or in the writing of the manuscript.
Availability of data and materials
All analyzed data from this study are included in this published article and its Additional files. All data generated during the current study are available from the corresponding author on reasonable request.