Targeting of ROCK2 Inhibits Angiosarcoma Tumor Growth

By Clarissa Amaya, BS and Brad A. Bryan, PhD

Introduction

Angiosarcomas are highly lethal tumors composed primarily of aberrant lymphatic or vascular endothelial cells and account for over ½ of all radiation-induced sarcomas of the breast.¹ In addition to breast tumors, angiosarcomas manifest in several locations throughout the body with a predominance of cases occurring in the head and neck region. The majority of angiosarcomas rapidly metastasize due to their vascular cell composition, and unlike many epithelioid sarcomas, angiosarcomas generally metastasize directly to the lungs via the vascular system.^2,3 Treatment of angiosarcoma involves radiation, surgery, and neoadjuvant and/or adjuvant chemotherapy with doxorubicin or taxanes, yet with this treatment regimen the five year survival rates for these patients is still less than 30%.⁴ Despite their vascular origin, even the addition of novel anti-angiogenic drugs has shown a minimal to absent response in angiosarcoma patients.⁵ Thus very little therapeutic progress has been made over the past several decades to increase the progression free survival or overall patient survival of individuals suffering from this sarcoma, and effective therapeutics against this disease are desperately needed.

ROCK2 is a serine/threonine kinase that serves as a key regulator of the actin cytoskeleton and cell motility.⁶ Previously published data from our lab has shown that pharmacological targeting ROCK2 activity inhibits the growth of several tumor types.^7,8,9 Moreover, we have previously shown that pharmacological targeting of small hairpin RNA (shRNA)-mediated gene expression knockdown of ROCK2 inhibits several properties of endothelial cells including capillary network formation.^9,10,11 Based on our lab’s cumulative findings, we hypothesized that targeting of ROCK2 in angiosarcomas (which express tumor and vascular characteristics) could lead to inhibition of tumor growth and aggression.

Results

We proposed two aims in our funded grant submission to the Liddy Shriver Sarcoma Initiative titled “ROCK inhibition as an anti-angiosarcoma therapeutic” The first aim was to test the effect of ROCK inhibition alone or in combination with chemotherapy on angiosarcoma tumor growth. The second aim was to examine the effects of ROCK inhibition on angiosarcoma vascular development.

shRNA knockdown of ROCK2 inhibits angiosarcoma tumor growth.

We knocked down the expression of ROCK2 in SVR mouse angiosarcoma cells using stable transfection of a panel of shRNA vectors. As a control we utilized a stably transfected scrambled shRNA vector. RNA was collected from the engineered cells and the ROCK2 level in each of the cell lines was detected using quantitative PCR, revealing between 20-50% knockdown depending on the shRNA construct used.

The scrambled control and ROCK2 knockdown angiosarcoma cells were injected into the flanks of nude mice (200,000 cells/subcutaneous injection; N=30 mice per condition) and allowed to grow until the control primary tumors reached approximately 1 cm³. As can be seen in Figure 1, shRNA knockdown of ROCK2 led to an over 80% reduction in tumor volume compared to the scrambled control tumors, suggesting that ROCK2 is essential for angiosarcoma tumor growth.

To understand the mechanism by which ROCK2 inhibits angiosarcoma tumor growth, we performed antibody array analysis comparing the proteomic signatures of scrambled control versus ROCK2 shRNA angiosarcoma tumors for over 1300 proteins and post-translational modifications. Our analysis revealed significant alterations in protein levels and phosphorylation status in a large number of proteins between the two conditions. Most notable were changes in key apoptotic and cell cycle regulators including p53, Bcl-Xl, IRS-1, 4E-BP, and CDC25.

Combination treatment of ROCK2 shRNA and doxorubicin in angiosarcoma tumors.

The current standard of therapy for angiosarcomas involves chemotherapeutic treatment with the anthracyclin antibiotic doxorubicin which intercalates DNA and inhibits the proliferation of tumor cells.¹² The most dangerous side effect of doxorubicin is cardiomyopathy leading to congestive heart failure.¹³ Unfortunately, the incidence of cardiomyopathy is dependent on the cumulative dose of this drug, thus there is a lifetime dose limit of doxorubicin that a patient may receive which greatly limits its efficacy against resistant angiosarcomas.

To test the effects of ROCK2 inhibition in combination with doxorubicin, the scrambled control and ROCK2 knockdown angiosarcoma cells were injected into the flanks of nude mice (200,000 cells/subcutaneous injection; N=30 mice per condition) and allowed to grow until the control primary tumors reached approximately 1 cm³. The tumors were then treated with sham, a low dose of doxorubicin (3 mg/kg), or a high dose of doxorubicin (6 mg/kg) at biweekly intervals over a three week time course. At the end of the time course tumors were harvested. As expected, doxorubicin reduced tumor volume compared to the sham treatment by over 75% in the 3 mg/kg treatment and by over 90% in the 6 mg/kg treatment. Interestingly, ROCK2 shRNA knockdown alone proved more effective than the low dose doxorubicin treatment alone (Figure 2). No additive effect was seen when ROCK2 was combined with the high dose doxorubicin treatment. Collectively, our data suggest that combinations of ROCK2 inhibition with lower doses of doxorubicin may allow effective treatment of patients with angiosarcomas over a longer time course due to less doxorubicin-induced cumulative cardiomyopathy and congestive heart failure.

ROCK2 shRNA mediated effects on the angiosarcoma tumor vasculature.

Our previous in vitro data has demonstrated that ROCK activity is essential for capillary formation, thus we sought to examine the effects of ROCK2 inhibition on capillary formation in angiosarcoma tumors. Analysis of scrambled control and ROCK2 shRNA tumors indicates that inhibition of ROCK2 in angiosarcoma tumors results in increased protein expression of EGFR, FGFR1, FLT3, VEGFR2, and AKT1. Each of these proteins are well established positive modulators of blood vessel formation,¹⁴ therefore we are intrigued to understand how disrupting ROCK2 expression, which has been reported to function in an anti-angiogenic manner using in vitro experiments, positively modulates pro-angiogenic gene expression.  We strongly suspect that ROCK2 shRNA mediated disruption of the tumor vasculature induces a pro-angiogenic response from the tumor. We are currently evaluating tumor vascular density and hypoxia within the scrambled control and ROCK2 shRNA tumors, and these experiments are expected to be completed by summer of 2014. Quantifying tumor vascular density and hypoxia for each condition will provide a solid understanding of how ROCK2 directly affects the tumor vasculature and how this effect contributes to reduced tumor size in ROCK2 shRNA tumors.

Summary

We set out to test the efficacy of ROCK inhibition against angiosarcomas. Our studies successfully demonstrate that ROCK2 is a prime target that should be further studied as a potential therapy for angiosarcoma treatment. Indeed, several pharmaceutical companies have currently developed small molecular inhibitors that inhibit ROCK2 with high specificity and low toxicity. Future studies should involve academic-industrial partnerships to test the efficacy of these novel ROCK2 inhibitors against angiosarcomas.

V11N3 ESUN Copyright © 2014 Liddy Shriver Sarcoma Initiative.

ROCK Inhibition as an Anti-Angiosarcoma Therapeutic

By Brad A. Bryan, PhD

Background

Angiosarcoma is a poorly understood malignant tumor accounting for 1-2% of soft tissue sarcomas, and is characterized by aggressively proliferating, extensively infiltrating aberrant blood vessel (endothelial) cells that form irregular vessel like channels (Figure 1). The majority of angiosarcomas are cutaneous, originating as singular or multifocal "bruise-like" patches on the scalp, upper forehead, or neck that aggressively grow and spread to surrounding tissues and distant parts of the body.¹ Because of their "bruise-like" appearance, cutaneous angiosarcomas are often misdiagnosed at early stages.² Moreover, these tumors are notoriously difficult to treat, displaying a 5 year disease free survival at 20-30%,³ thus more effective therapeutics for this disease are desperately needed.

Rho-kinase proteins (ROCK1 and ROCK2) are serine/threonine kinases that control cytoskeleton dynamics, cell movement/invasion, proliferation, survival, differentiation, and gene expression and have been implicated in angiogenesis and tumor progression. Tumor derived endothelial cells display a strong ROCK-dependent ability to organize into capillary networks,⁴ and disrupting ROCK signaling effectively blocks capillary angiogenesis.⁵ Additionally, ROCK expression is significantly elevated in a variety of cancers^6-9 and deregulation of ROCK signaling is an important step in tumor progression and metastasis. Pharmacological inhibition of ROCK activity shows strong efficacy in the treatment of in vivo malignancies including prostate, breast, glioma, melanoma, and human papillovirus infected tumors^10-14 with little to no adverse side effects. Because angiosarcomas are solid tumors that originate from vascular endothelial cells, we hypothesize that inhibition of ROCK signaling could be particularly effective in targeting their angiogenic and tumor properties.

Purpose of the Specific Investigation

Preliminary data from our laboratory indicate that disruption of ROCK1 or ROCK2 expression inhibits the early stages of angiosarcoma tumor formation (discussed below). The experiments outlined in this proposal will establish how inhibition of ROCK signaling contributes to the disruption of angiosarcoma tumor formation, metastasis, and vascular formation. Findings elucidated from this study will shed light on the molecular roles of ROCK proteins in angiosarcoma tumor formation and establish a therapeutic foundation for what appears to be an excellent anti-angiosarcoma target.

Preliminary Data

Research in our lab has primarily focused on cytoskeletal control of angiogenesis and tumor formation. We have previously reported that inhibition of ROCK1 or 2 blocks multiple aspects of angiogenesis, resulting in a significant disruption of in vitro capillary network formation, cell polarization, and cell migration.¹⁵ Moreover, inhibition of ROCK signaling led to alterations in cytoskeletal dynamics in part due to ROCK1 & 2-mediated reductions in actin isoform expression, and ROCK2-specific reduction in myosin phosphatase and cofilin phosphorylation¹⁵ (Figure 2). Our lab has utilized microarray analysis on ROCK1 & 2 knockdown endothelial cell lines revealing overlapping and unique control of global transcription by the paralogs, and a reduction in the transcriptional regulation of just under 50% of vascular endothelial growth factor (VEGF) responsive genes, suggesting that ROCK protein modulation of cytoskeletal dynamics is essential for growth factor-mediated gene transcription in endothelial cells.^5,15 Moreover, using in vivo tumor models including angiosarcomas, we have shown that inhibition of ROCK signaling is capable of blocking the early stages of primary tumor formation to a significant degree (Figure 3).^14-16 Our initial angiosarcoma tumor experiments are preliminary in that tumors were only collected at one time point, we did not examine how ROCK inhibition alone or in combination with chemotherapy affects overall survival in tumor-bearing hosts, and we did not examine the overall histology of the tumor.

Research Plan and Experimental Design

Our preliminary data serves as the spark for a more thorough understanding of how manipulation of ROCK signaling can be effectively used as an anti-angiosarcoma target. To determine if ROCK inhibition shows efficacy in angiosarcoma treatment regimes, we propose two experimental aims:

Aim #1: To determine the effect of ROCK inhibition alone or in combination with chemotherapy on primary and metastatic angiosarcoma tumor growth.

Aim #2: To examine the effects of ROCK inhibition on angiosarcoma vascular development. Completion of these proposed aims will lay a solid foundation supporting the use of ROCK inhibitors as anti-angiosarcoma agents.

Aim #1—To determine the effect of ROCK inhibition alone or in combination with chemotherapy on primary and metastatic angiosarcoma tumor growth.

Our preliminary results suggest that targeting ROCK signaling could be a valid anti-cancer therapy, however as mentioned previously, there are still many questions to address. The experiments proposed in this aim will test the efficacy of ROCK inhibition as an effective therapy against angiosarcoma by examining the outcomes of ROCK expression knockdown on angiosarcoma tumor formation, metastasis, and host death using live animal in vivo tumor imaging over the time course of the disease.

Engineering Angiosarcoma Cells

For all proposed experiments, we will use SVR mouse angiosarcoma cells, which are SV40 transformed mouse pancreatic endothelial cells isolated from C57BL/6 mice which harbor a constitutively active H-Ras mutation. Subcutaneous injection of SVR cells into mice results in tumors reaching 1 cm in diameter after approximately 10 days, leading to aggressive metastasis and host death due to hemorrhage and anemia with thrombocytopenia . We have previously generated SVR cells stably expressing shRNA plasmids specific for ROCK1 or 2 expression knockdown or for both ROCK1 and ROCK2 knockdown (double knockdown). Additionally, as a control we have already generated SVR cells stably overexpressing a non-targeting shRNA vector. To aid in in vivo tumor imaging, these cell lines will be further transfected with the pGL4.13[luc2/SV40] constitutively expressing luciferase reporter plasmid (containing a selectable geneticin resistance gene), and subsequent stable pools of engineered cells will be selected with geneticin. Cell lines will be designated: SVR-C-Luc (control), SVR-R1-Luc (ROCK1 knockdown), SVR-R2-Luc (ROCK2 knockdown), or SVR-R1/2-Luc (double ROCK knockdown). For all experiments, SVR-R1-Luc, SVR-R2-Luc, and SVR-R1/2-Luc derived tumors will be compared to SVR-C-Luc derived tumors.

Mouse Angiosarcoma Model and Tumor Visualization

We will utilize in vivo chemiluminescent whole animal imaging to compare the ability of SVR-C-Luc, SVR-R1-Luc, SVR-R2-Luc, or SVR-R1/2-Luc derived angiosaroma tumors to grow and metastasize to distal organs. 1x10⁶ engineered SVR cells (per mouse) from each of the engineered angiosarcoma lines will be injected subcutaneously into the dorsolateral flank of 4 week old syngeneic C57BL/6 mice. Each group of mice harboring the engineered cell lines will be subjected to the following conditions: sham saline (weekly intraperitoneal [IP] sham injections), doxorubicin (weekly IP injections at 2 mg/kg body weight), or paclitaxel (weekly IP injections at 5 mg/kg body weight). Doxorubicin and paclitaxel are commonly used anti-angiosarcoma chemotherapies. To observe a statistical effect (p<0.05) with a power of 0.80, statistical power analysis indicates that we will be using a sample size of 18 mice per condition (total of 4 cell lines x 3 treatments x 18 mice per condition=216 mice). At weekly intervals over the lifetime of the host, mice will be injected via an intraperitoneal route with a Luciferin solution (30 mg/kg in buffered saline) allowing distribution throughout their bodies for 10 minutes to activate the recombinant luciferase protein stably expressed off the pGL4.13[luc2/SV40] plasmid in the injected tumor cells and their progeny, by not in any other cell of the mouse. The mice will then be temporarily anesthetized with 3% isofluorane, and chemiluminescent imaging of the engineered tumors will occur for 2.5 minutes per side (dorsal/ventral) using a live animal chemiluminescent imaging system. The size of the primary and metastatic tumors, the number of arising metastatic tumors, and the total tumor volume of both primary and metastatic tumors will be quantified from the captured images and recorded, as well as statistically analyzed using Study Director software. In addition to calculating total angiosarcoma tumor volume over the life of the host, we will collect mortality data and, using a mortality plot generated by Study Director software, statistically compare overall survival in mice harboring the panel of engineered angiosaroma tumors.

We have previously shown that knockdown of ROCK gene expression in angiosarcoma cells inhibits angiosarcoma tumor formation.¹⁵ The data obtained from the experiments proposed in Aim #1 will provide a solid understanding of how manipulation of ROCK expression affects angiosarcoma tumor progression from initial formation to metastasis. Moreover, these data will demonstrate whether ROCK inhibition is capable of altering overall tumor-bearing animal survival. The methodology is very straight forward and we foresee no technical difficulties arising from this study. The SVR cells utilized in this study were originally isolated from C57BL/6 mice, therefore immune rejection of the tumor cells by syngeneic host C57BL/6 mice will not be an issue. Our proposed experiments using live animal tumor imaging offer the significant advantage that we can study the progression of the disease over time on a per animal basis without the need to sacrifice mice at multiple time points.

Aim #2—To examine the effects of ROCK inhibition on angiosarcoma vascular development.

Recruitment of an adequate blood supply is essential for tumor growth and metastasis, and ROCK signaling is required for tumor cell secretion of pro-angiogenic growth factors. To elucidate the ability of ROCK knockdown angiosarcoma tumors to recruit and sustain a functional vascular supply we will utilize live 3D Doppler imaging and immunocytochemistry to quantify vascular density, composition, and blood flow. Using the same mice described in the in vivo angiosarcoma tumor experiments outlined in Aim #1, we will assess blood flow at weekly intervals in the superficial primary tumors using non-invasive contrast enhanced 3D color Doppler sonography, which provides information about the spatial distribution of blood velocities and moving blood volume. Blood circulation will be quantitatively evaluated using two parameters: 1) the vascularization index measuring the number of vessels within the studied tissue volume, and 2) the flow index measuring the intensity of blood flow within the tissue volume.

In addition to endothelial cells, blood vessels are also composed of vascular smooth muscle cells (VSMCs, collectively vascular smooth muscle cells and pericytes), and the physical association (or lack thereof) between endothelial cells and VSMCs regulates vascular development and maturation through the control of endothelial cell motility, proliferation, and differentiation. To assess the role of ROCK signaling in tumor endothelial cell/VSMC interactions in vivo, we will perform immunocytochemical analysis of angiosarcoma tumor vascularization from both primary and metastatic tumors harvested from SVR-C-Luc, SVR-R1-Luc, SVR-R2-Luc, or SVR-R1/2-Luc tumor bearing mice. 1x10⁶ engineered SVR cells from each engineered line will be injected subcutaneously into the dorsolateral flank of 4 week old C57BL/6 mice (20 mice per cell line x 4 cell lines=80 mice total). Each week following initial tumor cell injection for a period of one month, five mice from each condition will be anesthetized with 3% isofluorane and perfused through the aorta with fluorescein (FITC) dextran (2x10⁶ molecular weight) to fluorescently label the functional vasculature. Distribution throughout their bodies will be allowed for 10 minutes, and primary and metastatic tumors will be harvested through dissection and processed for cryosectioning. Cryosections of each tumor will be subjected to immunofluorescent staining and 3-dimensional confocal microscopic imaging to examine endothelial and VSMC density and their cell-to-cell associations. VSMCs will be detected with Alexa Fluor 350-labelled antibodies specific against smooth muscle alpha-actin (a marker for VSMCs). Functional blood vessels will be perfused with FITC-dextran and can be easily detected with fluorescent microscopy. Overlays of blue (VSMCs) and green (perfused blood vessels) fluorescent images will provide a qualitative analysis of tumor vascular composition. Quantitative measurements will include density of functional vasculature (quantifying FITC-dextran perfusion with Image J software), density of VSMCs along the blood vessels, and association (or lack thereof) between VSMCs and functional vasculature, providing an accurate view of the effects of ROCK inhibition on angiosarcoma vascularization.

Impact Statement

Angiosarcomas are particularly aggressive cutaneous soft tissue sarcomas that account for approximately 1-2% of all sarcoma cases, resulting in 0.4 cases per 100,000 individuals. Despite surgical excision, radiation, and chemotherapy, the 5 year survival rate in patients with this cancer is between only 20-30%, indicating the need for better drugs to be developed against this disease. Because angiosarcomas are tumors of vasculature origin, they should be especially susceptible to the anti-angiogenic and anti-tumor properties of ROCK protein inhibition, and the results of this study could be the basis for the next generation of effective anti-angiosarcoma therapies.

V9N4 ESUN Copyright © 2012 Liddy Shriver Sarcoma Initiative.

Grant Funding

The Liddy Shriver Sarcoma Initiative announced the funding of this $45,000 grant in August 2012. The study was made possible, in part, by a generous donation from Laura Somerville and a supporting donation from the Center for Research and Analysis of Vascular Tumors (CRAVAT).