Rho-associated kinases 1 and 2 (ROCK1/2) are Rho-GTPase effectors that control key aspects of the actin cytoskeleton, but their role in proliferation and cancer initiation or progression is not known. Here, we provide evidence that ROCK1 and ROCK2 act redundantly to maintain actomyosin contractility and cell proliferation and that their loss leads to cell-cycle arrest and cellular senescence. This phenotype arises from down-regulation of the essential cell-cycle proteins CyclinA, CKS1 and CDK1. Accordingly, while the loss of either Rock1 or Rock2 had no negative impact on tumorigenesis in mouse models of non-small cell lung cancer and melanoma, loss of both blocked tumor formation, as no tumors arise in which both Rock1 and Rock2 have been genetically deleted. Our results reveal an indispensable role for ROCK, yet redundant role for isoforms 1 and 2, in cell cycle progression and tumorigenesis, possibly through the maintenance of cellular contractility.
Escherichia coli is the most extensively used host for the production of recombinant proteins. However, most of the eukaryotic proteins are typically obtained as insoluble, misfolded inclusion bodies that need solubilization and refolding. To achieve high-level expression of soluble recombinant human interferon alpha (rhIFNalpha) in E. coli, we have first constructed a recombinant expression plasmid (pGEX-hIFNalpha2b), in which we merged the hIFNalpha2b cDNA with the glutathione S-transferase (GST) coding sequence downstream of the tac-inducible promoter. Using this plasmid, we have achieved 70% expression of soluble rhIFNalpha2b as a GST fusion protein using E. coli BL21 strain, under optimized environmental factors such as culture growth temperature and inducer (IPTG) concentration. However, release of the IFN moiety from the fusion protein by thrombin digestion was not optimal. Therefore, we have engineered the expression cassette to optimize the amino acid sequence at the GST-IFN junction and to introduce E. coli preferred codon within the thrombin cleavage site. We have used the engineered plasmid (pGEX-Delta-hIFNalpha2b) and the modified E. coli trxB(-)/gor(-) (Origami) strain to overcome the problem of removing the GST moiety while expressing soluble rhIFNalpha2b. Our results show the production of soluble and functional rhIFNalpha2b at a yield of 100 mg/l, without optimization of any step of the process. The specific biological activity of the purified soluble rhIFNalpha2b was equal to 2.0 x 10(8) IU/mg when compared with the WHO IFNalpha standard. Our data are the first to show that high yield production of soluble and functional rhIFNalpha2b tagged with GST can be achieved in E. coli.
Cellular behavior is strongly influenced by the architecture and pattern of its interfacing extracellular matrix (ECM). For an artificial culture system which could eventually benefit the translation of scientific findings into therapeutic development, the system should capture the key characteristics of a physiological microenvironment. At the same time, it should also enable standardized, high throughput data acquisition. Since an ECM is composed of different fibrous proteins, studying cellular interaction with individual fibrils will be of physiological relevance. In this study, we employ near-field electrospinning to create ordered patterns of collagenous fibrils of gelatin, based on an acetic acid and ethyl acetate aqueous co-solvent system. Tunable conformations of micro-fibrils were directly deposited onto soft polymeric substrates in a single step. We observe that global topographical features of straight lines, beads-on-strings, and curls are dictated by solution conductivity; whereas the finer details such as the fiber cross-sectional profile are tuned by solution viscosity. Using these fibril constructs as cellular assays, we study EA.hy926 endothelial cells' response to ROCK inhibition, because of ROCK's key role in the regulation of cell shape. The fibril array was shown to modulate the cellular morphology towards a pre-capillary cord-like phenotype, which was otherwise not observed on a flat 2-D substrate. Further facilitated by quantitative analysis of morphological parameters, the fibril platform also provides better dissection in the cells' response to a H1152 ROCK inhibitor. In conclusion, the near-field electrospun fibril constructs provide a more physiologically-relevant platform compared to a featureless 2-D surface, and simultaneously permit statistical single-cell image cytometry using conventional microscopy systems. The patterning approach described here is also expected to form the basics for depositing other protein fibrils, seen among potential applications as culture platforms for drug screening.
Full text Figures and data Side by side Abstract eLife digest Introduction Results Discussion Materials and methods Statistical analysis Data availability References Decision letter Author response Article and author information Metrics Abstract Rho-associated kinases 1 and 2 (ROCK1/2) are Rho-GTPase effectors that control key aspects of the actin cytoskeleton, but their role in proliferation and cancer initiation or progression is not known. Here, we provide evidence that ROCK1 and ROCK2 act redundantly to maintain actomyosin contractility and cell proliferation and that their loss leads to cell-cycle arrest and cellular senescence. This phenotype arises from down-regulation of the essential cell-cycle proteins CyclinA, CKS1 and CDK1. Accordingly, while the loss of either Rock1 or Rock2 had no negative impact on tumorigenesis in mouse models of non-small cell lung cancer and melanoma, loss of both blocked tumor formation, as no tumors arise in which both Rock1 and Rock2 have been genetically deleted. Our results reveal an indispensable role for ROCK, yet redundant role for isoforms 1 and 2, in cell cycle progression and tumorigenesis, possibly through the maintenance of cellular contractility. https://doi.org/10.7554/eLife.12203.001 eLife digest Animal cells contain a structure called the cytoskeleton, which helps give the cells their shape. This structure can rapidly disassemble and reassemble, which enables cells to change their shape, move and divide into two. Many proteins are involved in controlling these processes. In particular, two proteins called ROCK1 and ROCK2 are known to be important for helping cancer cells move. However, investigations into the exact roles of these proteins have so far produced contradictory results. Kümper et al. have now developed a more refined method of studying what ROCK1 and ROCK2 do in cells. This involves genetically engineering mice in a way that makes it possible to control whether ROCK1 and ROCK2 are produced in specific cell types and tissues. Studying cells that had been taken from these mice revealed that cells that lacked both proteins could not contract. Moreover, these cells became bigger and flattened out. This change in appearance went hand in hand with the cells becoming unable to divide and form new cells. However, cells that lacked just one type of ROCK protein were still able to divide and proliferate. As tumors form as a result of cells dividing and proliferating uncontrollably, Kümper et al. then studied how the ROCK proteins affect tumor development in mice that are susceptible to lung or skin cancer. Although cancerous cells were found that contained just one type of ROCK protein, no tumor cells were found that lacked both ROCK1 and ROCK2, further confirming that having one ROCK protein is essential for tumor formation. Overall, it appears that within the systems studied ROCK1 and ROCK2 perform the same roles, and that ROCK proteins are indispensable for cell proliferation and hence tumor development. The next challenge will be to identify the tumor types that are highly dependent on processes driven by the ROCK proteins. Further work could then investigate whether drugs that inhibit the activity of the ROCK proteins block the growth and spread of these tumors. https://doi.org/10.7554/eLife.12203.002 Introduction The Rho-associated coiled-coil-containing protein serine/threonine kinases ROCK1 and ROCK2 are downstream effectors of the Rho subfamily of small GTPases. They are activated by interaction with Rho GTPases and act through a number of pathways to regulate the actin cytoskeleton and thus cell migration, cell-cell adhesion and cancer cell invasion (Itoh et al., 1999; Olson and Sahai, 2009; Thumkeo et al., 2013). Substrates for ROCKs include the myosin-binding subunit of myosin phosphatase (MYPT1) and the myosin regulatory light chain (MLC), which regulate actomyosin contractile forces (Kawano et al., 1999; Kureishi et al., 1997). Actomyosin contractility is generated when myosin regulatory light chain (MLC) is phosphorylated, allowing myosin II interaction with actin filaments to generate mechanical force (Aksoy et al., 1983; Kureishi et al., 1997). ROCKs may either directly phosphorylate MLC (Amano et al., 1996) or indirectly regulate MLC phosphorylation by phosphorylating the myosin-binding subunit of myosin phosphatase (MYPT1) and inactivating it (Kimura et al., 1996). Myosin phosphatase inactivation results from phosphorylation of two inhibitory sites, Thr696 and Thr850, on MYPT1 (Feng et al., 1999; Kawano et al., 1999; Muranyi et al., 2005). Furthermore, ROCKs phosphorylate and activate LIM kinases 1 and 2 (LIMK1, 2), which inhibit the actin-depolymerizing protein Cofilin (Yang et al., 1998). ROCK1 and 2 exhibit 65% overall identity and 87% within the kinase domain, and some studies suggest differential roles for the isoforms. Using RNA interference, ROCK1 was reported to be important for stress fiber formation in fibroblasts, whereas ROCK2 controls cortical contractility and phagocytosis (Yoneda et al., 2005; 2007). Distinct roles for ROCK1 and 2 have also been described in the regulation of keratinocyte differentiation (Lock and Hotchin, 2009) and cell detachment (Shi et al., 2013). Mice in which Rock1 has been genetically deleted are born, but have defects in eyelid as well as ventral body wall closure (Shimizu et al., 2005). Ninety percent of Rock2 null mice die in utero due to defects in the placental labyrinth layer. This indicates that ROCK1 cannot compensate for a loss of ROCK2. However, the few Rock2 null mice, that are born, display defects similar to those described in Rock1 null mice (Thumkeo et al., 2003). This indicates some level of functional redundancy (Thumkeo et al., 2005). In addition to their role in cell migration, ROCKs have been reported to modulate apoptosis (Coleman et al., 2001; Sebbagh et al., 2005) and cell proliferation (Croft and Olson, 2006; Samuel et al., 2011; Zhang et al., 2009). The precise role of ROCKs in cell proliferation is not clear: some reports suggest ROCK function is required for G1/S progression (Croft and Olson, 2006; Zhang et al., 2009), but others suggest ROCK is only required for anchorage-independent growth of transformed cells (Sahai et al., 1999; Vigil et al., 2012). One in vivo study reported that over-activation of ROCK, by expressing the kinase domain of ROCK2 in mouse skin, led to hyperproliferation and epidermal thickening (Samuel et al., 2011). In order to investigate the roles of ROCK1 and 2 in tumorigenesis, we have generated conditional Rock1 and 2 knockout mice and studied these in vivo, using genetically engineered mouse models of non-small cell lung cancer (NSCLC) and BrafV600E mutant models of melanoma, as well as in vitro, using cells isolated from these mice. Results Depletion of both ROCK1 and 2 leads to multi-stage cell cycle arrest through defects in actomyosin contractility It is well understood that ROCKs control key aspects of the actin cytoskeleton such as actomyosin contractility, but their role in cell proliferation is not known. Because Rock null mice die early due to developmental defects, we generated Rock1 and 2 conditional alleles (Rock1f and Rock2f) by inserting LoxP sites flanking exon 6 in the Rock1 locus and exons 5 and 6 in the Rock2 locus (Figure 1—figure supplement 1A). These exons are located within the kinase domain and their deletion results, through frameshifts, in the absence of ROCK proteins. We first generated cells lacking ROCK1 (Rock1∆/∆), ROCK2 (Rock2∆/∆) or ROCK1 and 2 (Rock1∆/∆;Rock2∆/∆) by isolating mouse embryo fibroblasts (MEFs) from embryos with the following genotypes: Rock1f/f;Rock2wt/wt, Rock1wt/wt;Rock2f/f or Rock1f/f;Rock2f/f and infecting them with Adenovirus-expressing Cre recombinase (Ad-Cre) or GFP (Ad-GFP). Depletion of ROCK1 and ROCK2 (Rock1∆/∆;Rock2∆/∆) resulted in a strong impairment of cell proliferation in vitro, which was not observed in cells lacking either ROCK1 (Rock1∆/∆) or ROCK2 (Rock2∆/∆) (Figure 1A), demonstrating that ROCK function is required for cell proliferation and that ROCK1 and ROCK2 act redundantly. Infection of cells with adenovirus is not 100% efficient, and when the growth of Rock1∆/∆;Rock2∆/∆ cells was monitored over a longer period of time, these cells eventually recovered their ability to proliferate (Figure 1—figure supplement 1B), but western blot analysis revealed that these cells express ROCK1 and 2 in equivalent levels to wild type cells (data not shown) and thus likely originated from uninfected cells. Figure 1 with 1 supplement see all Download asset Open asset Depletion of ROCK1 and 2 leads to defects in cell proliferation in vitro and in vivo. (A) Proliferation curves of MEFs with different genotypes over 6 days. The cells were seeded 3 d after adenovirus infection. Graphs show total number of cells and SD from 5 independent experiments each carried out in triplicates. p-values were calculated using Student’s t-test: ** p<0.005; *** p<0.001. (B) Rock1f/f;Rock2f/f control and Rock1∆/∆;Rock2∆/∆ MEFs were cultured for 3 days and wild-type cells were treated with H1152, inactive blebbistatin (+) or active blebbistatin (+/-) for 48 hr. Cells from all conditions were then subjected to a colony formation assay and grown for a further 7 days. (C–F) Rock1f/f;Rock2f/f MEFs transformed with Trp53 DD and HRas V12 were treated with Ad Cre to generate ∆. Cells were injected subcutaneously into CD1 nude mice and growth analyzed. The graph shows average tumor volume in mm3 and SEM for Rock1∆/∆ and control (C), Rock2∆/∆ and control (D), Rock1∆/∆;Rock2∆/∆ and control (E). p values were calculated by ANOVA and are as indicated. (F) Tumors with stated genotypes were immunoblotted with indicated antibodies. https://doi.org/10.7554/eLife.12203.003 As genetic depletion abrogates ROCK function long term, we investigated whether long-term treatment of cells with ROCK inhibitors caused proliferation defects. Cells treated for 48 hr with the ROCK inhibitor H1152 (Sasaki et al., 2002) had reduced proliferation (Figure 1B). Similar results were observed with other ROCK inhibitors, such as GSK269962A, AT13148, GSK429286A and chroman1 (data not shown). However, the much-used ROCK inhibitor Y-27632 (Narumiya et al., 2000) had a much weaker effect on cell proliferation than H1152 (Figure 1—figure supplement 1C). This is consistent with previous studies, where we have shown that Y-27632 is a less effective ROCK inhibitor than H1152 (Sadok et al., 2015). To determine whether defects in proliferation were due to ROCK-mediated effects on actomyosin contractility, cells were treated with blebbistatin, an inhibitor of myosin II ATPase, for 48 hr. Treated cells displayed a similar proliferation defect to that observed in Rock1∆/∆;Rock2∆/∆ cells (Figure 1B), arguing that the requirement for ROCK in cell proliferation is mediated through its role in maintaining actomyosin contractility. To test whether Rock1∆/∆, Rock2∆/∆ orRock1∆/∆;Rock2∆/∆ MEFs can proliferate in vivo, MEFs isolated from Rock1f/f, Rock2f/f or Rock1f/f;Rock2f/fmice were immortalized with retrovirus expressing a dominant-negative form of Trp53 (Trp53 DD) (Hahn et al., 2002), and subsequently transformed with a retrovirus expressing mutant H-Ras (H-RasV12). The cells were then treated with Ad-Cre or control virus and, 5 days later, injected subcutaneously and tumor growth monitored. No differences in tumor growth were observed when transformed MEFs, derived from Rock1f/for Rock2f/f cells, infected with Ad-Cre were compared to control virus (Figure 1C,D). In contrast, transformed Rock1∆/∆;Rock2∆/∆ MEFs grew significantly slower in vivo compared to controls (Figure 1E). Western Blot analysis revealed the retention of ROCK1 and 2 protein in tumors derived from Rock1∆/∆;Rock2∆/∆ (Trp53 DD; H-RasV12) MEFs but complete deletion of ROCK1 in Rock1∆/∆ and ROCK2 in Rock2∆/∆ tumor samples (Figure 1F, data not shown), in agreement with the observed results on proliferation in vitro. This result shows that injected Rock1∆/∆;Rock2∆/∆ cells are unable to proliferate and form tumors but the un-infected cells can proliferate and form tumors thereby accounting for the delay in tumorigenesis and implying that ROCK function is essential for proliferation and hence tumor initiation and formation. We here show that concomitant depletion of ROCK1 and ROCK2 leads to defects in cell proliferation in vitro as well as in vivo. Depletion of a single ROCK isoform however has no effect, suggesting they can act redundantly. ROCK1 and ROCK2 act redundantly to regulate actomyosin contractility and cell shape In order to identify causes for the defects in cell proliferation seen in cells lacking ROCK1 and ROCK2 (Rock1∆/∆;Rock2∆/∆), we characterized their cell morphology. One of the most prominent responses to activated ROCK is the generation of actin stress fibers, which regulate cell shape (Amano et al., 1997; Ridley, 1999). A previous report using RNA interference suggested that ROCK1, rather than ROCK2, is important for stress fiber formation (Yoneda et al., 2005). After thorough analysis of ROCK depletion upon infection with Ad-Cre or Ad-GFP, a slight decrease in protein levels was observed after three days (Figure 2—figure supplement 1A). After four days a further reduction was seen and after five days both ROCK1 and 2 were depleted (Figure 2B–C). Three days after infection with Ad-Cre, Rock1∆/∆;Rock2∆/∆ cells lacked stress fibers and displayed a ‘tail retraction’ phenotype, a well-known characteristic of ROCK inhibition (Somlyo et al., 2000). Cells had long processes and defects in detaching their tails, in contrast to their controls (Figure 2A and Videos 1,2). A similar phenotype was observed in cells treated with the ROCK inhibitor H1152 or the myosin II inhibitor blebbistatin but not in cells lacking either ROCK1 (Rock1∆/∆) or ROCK2 (Rock2∆/∆) (Figure 2A). After five days, cells that lack both ROCK1 and ROCK2 had few apparent central stress fibers, unlike control cells, and surprisingly adopted a flat morphology (Figure 2A and Videos 1,2). This phenotype seemed to appear gradually after prolonged loss of ROCK which was seen three days after adenovirus infection. Interestingly, when random motility was analyzed between Days 3 and 4 after adenovirus infection, using ImageJ to track cells, migration speed and directionality were significantly increased in cells lacking ROCK1 and 2 (Figure 2—figure supplement 1B). This was previously described by Lomakin et al. (Lomakin et al., 2015). Figure 2 with 1 supplement see all Download asset Open asset Only deletion of both ROCK1 and 2 leads to a loss of central stress fibers and a decrease in actomyosin contractility. (A) Images of wild-type, Rock1∆/∆, Rock2∆/∆or Rock1∆/∆;Rock2∆/∆ MEFs, 3 days after Ad-GFP and Ad-Cre-GFP infection, stained for pMLC and phalloidin. Wild-type and Rock1∆/∆;Rock2∆/∆ cells are also shown 5 days after Ad-GFP and Ad-Cre-GFP infection. Scale bars are 50 µm. (B, C) Western blot analyses of ROCK targets in lysates of MEFs with indicated genotypes. Representative blots are shown, quantification of multiple biological replicates can be found in Figure 2—figure supplement 1C. (D) Images of MEFs with indicated genotypes plated on a thick layer of collagen and stained for phalloidin. Scale bars are 50 µm. (E) Images and quantification of collagen gel contraction assay using MEFs of indicated genotypes. Graph shows average data and SD from 5 independent experiments, each carried out in triplicate. p-values were calculated using Student’s t-test: ***p<0.001. https://doi.org/10.7554/eLife.12203.005 Video 1 Download asset This video cannot be played in place because your browser does support HTML5 video. You may still download the video for offline viewing. Download as MPEG-4 Download as WebM Download as Ogg Rock1f/f;Rock2f/f control, 3 days after infection with Ad-GFP, 10X magnification. https://doi.org/10.7554/eLife.12203.007 Video 2 Download asset This video cannot be played in place because your browser does support HTML5 video. You may still download the video for offline viewing. Download as MPEG-4 Download as WebM Download as Ogg Rock1∆/∆;Rock2∆/∆, 3 days after infection with Ad-Cre, 10X magnification. https://doi.org/10.7554/eLife.12203.008 Consistent with the levels of stress fibers, phosphorylation of MYPT and MLC were only affected in cells where both ROCKs had been removed (Figure 2B–C, quantification in Figure 2—figure supplement 1C), showing that ROCK1 and ROCK2 act redundantly in regulating MYPT and MLC phosphorylation. While MYPT Thr850 phosphorylation was markedly reduced in Rock1∆/∆;Rock2∆/∆ cells, MYPT Thr696 phosphorylation was not (Figure 2B, quantification in Figure 2—figure supplement 1C), suggesting that, while Thr850 is principally regulated by ROCKs, Thr696 is regulated by other kinases such as MRCK (Wilkinson et al., 2005). ROCK1 levels were slightly but not significantly increased in Rock2∆/∆ cells and vice versa (Figure 2—figure supplement 1C). Rock1∆/∆;Rock2∆/∆ MEFs spread when plated on a thick layer of collagen 5 days after adenovirus infection, unlike Rock1∆/∆ or Rock2∆/∆ cells which had a more spindle-like morphology similar to wild-type controls (Figure 2D). This indicates reduced cellular contractility upon loss of ROCK1 and 2. As a functional measure of actomyosin contractility, we used a gel contraction assay (Hooper et al., 2010), demonstrating that cells containing only ROCK1 or ROCK2 were able to contract a collagen gel, but cells lacking both ROCK1 and ROCK2 were not (Figure 2E), supporting the notion ROCK1 and ROCK2 act redundantly to maintain cellular contractility. To further investigate ROCK signaling, we analyzed the temporal profile of changes in the phospho-proteome upon ROCK inhibition using mass spectrometry. Short-term inhibition of ROCK, using H1152, resulted in a significant decrease in phosphorylation of many phospho-sites, including several known targets of ROCK such as Cofilin-1 (Ser-3), Cofilin-2 (Ser-3), Destrin (Ser-3), and Ser-19/Thr-20 of MLC (Supplementary file 1 and Figure 2—figure supplement 1D). However, the majority of these phosphorylation sites, with the exception of MLC Ser-19/Thr-20, were only transiently affected, as their phosphorylation was fully recovered in long-term ROCK inhibited cells (Supplementary file 1 and Figure 2—figure supplement 1E) or Rock1∆/∆;Rock2∆/∆ cells (Figure 2B–C). These results indicate that alternative phosphorylation mechanisms exist for some ROCK substrates, but that no other kinase previously implicated in MLC phosphorylation can substitute for ROCK1/2 in the generation of actomyosin contractility. In summary, we showed that short-term (three days after Ad-Cre infection) depletion of ROCKs induces a previously described ‘tail retraction’ phenotype but surprisingly long-term (from five days after Ad-Cre infection) depletion induces a further change in morphology resulting in a flattened morphology and decrease in central stress fibers. Loss of ROCK-dependent actomyosin contractility leads to cellular senescence The flattened and enlarged morphology observed upon long-term depletion of ROCK1 and ROCK2 is typical of cellular senescence. Indeed, Rock1∆/∆;Rock2∆/∆ MEFs, or MEFs treated with H1152 or blebbistatin for prolonged periods exhibited a significant increase in staining for senescence-associated β-galactosidase (SA-βGal) (Debacq-Chainiaux et al., 2009) than control cells (Figure 3A). Senescence was also observed after prolonged treatment with other ROCK inhibitors such as GSK269962A, AT13148, GSK429286A and chroman1 (Figure 3—figure supplement 1A). To determine whether abrogation of ROCK function causes senescence in vivo, we used the recently described inhibitor AT13148 that has been shown to potently inhibit ROCK and has been previously used in vivo (Sadok et al., 2015; Yap et al., 2012a). 690cl2 mouse melanoma cells were injected intra-dermally into CD1 athymic mice, and the mice treated with AT13148 at 40 mg/kg. Senescence was detected by SA-βGal staining of tumors. We observed a decrease in tumor growth when mice were treated with AT13148 compared to controls (Figure 3—figure supplement 1B). Additionally, we observed an increase in the number of SA-βGal positive cells within tumors (Figure 3—figure supplement 1C). Figure 3 with 1 supplement see all Download asset Open asset Rock1∆/∆;Rock2∆/∆ cells undergo senescence and show a cell cycle block. (A) Images of MEFs treated with H1152, blebbistatin (+), blebbistatin (+/-) for 5 days and Rock1f/f;Rock2f/f control, Rock1∆/∆;Rock2∆/∆ MEFs, 5 days after Ad-GFP and Ad-Cre-GFP infection and followed by SA-βgal staining. Scale bars are 50 µm. Graph shows number of SA-βgal expressing cells divided by total number of cells and SD of >100 cells from three independent experiments and p-values were calculated using Student’s t-test: *** p<0.001. (B) Images of Rock1f/f;Rock2f/f and Rock1∆/∆;Rock2∆/∆ MEFs stained with phalloidin and DAPI. Overlay images shown. Arrows in image indicate binucleate cells. Scale bars are 50 µm. Bar chart shows average data and SD of nuclei in > 150 cells from 5 independent experiments. Values were calculated using Student’s t-test: * p<0.05. (C, D) Analysis of cell division of Rock1f/f;Rock2f/f and Rock1∆/∆;Rock2∆/∆ MEFs in time-lapse movies. (C) Quantification of failed cell divisions resulting in binucleate cells. (D) Quantification of average number of cells dividing. Graphs show average data and SD of > 300 cells from at least five independent experiments. (E) Cell cycle profiles of Rock1f/f;Rock2f/f and Rock1∆/∆;Rock2∆/∆ MEFs. The graph shows the percentage of cells in G2/M (top), S (middle) and G1 (bottom) phase of the cell cycle. Error bars represent SD. Data are from 5 independent experiments and p-values were calculated using Student’s t-test: * p<0.05. https://doi.org/10.7554/eLife.12203.009 Analysis of Rock1∆/∆;Rock2∆/∆ cells by microscopy revealed a significant increase in the number of cells with two nuclei (Figure 3B), indicating a failure of cytokinesis. During cytokinesis an actomyosin-based contractile ring is formed to drive cleavage furrow ingression, which subsequently results in two daughter cells. ROCK and the related Citron kinase have been shown to localize to the cleavage furrow, yet it is not entirely clear which kinase regulates MLC phosphorylation at the furrow to generate actomyosin contractility (Kosako et al., 1999; 2000; Madaule et al., 1998). Analysis of time-lapse movies further confirmed a role for ROCK1 and 2 in cytokinesis as a greater percentage of Rock1∆/∆;Rock2∆/∆ cells failed to divide compare to wild-type cells (Figure 3C and Videos 3,4). In addition to an increased number of binucleate cells, analysis of time-lapse movies of Rock1∆/∆;Rock2∆/∆ cells revealed an overall lower number of dividing cells (Figure 3D and Videos 3,4). These data suggest that cells depleted of ROCK1 and 2 do not initiate division further, thus indicating a cell cycle block. Video 3 Download asset This video cannot be played in place because your browser does support HTML5 video. You may still download the video for offline viewing. Download as MPEG-4 Download as WebM Download as Ogg Rock1f/f;Rock2f/f control, 5 days after infection with Ad-GFP, 20X magnification. https://doi.org/10.7554/eLife.12203.011 Video 4 Download asset This video cannot be played in place because your browser does support HTML5 video. You may still download the video for offline viewing. Download as MPEG-4 Download as WebM Download as Ogg Rock1∆/∆;Rock2∆/∆, 5 days after infection with Ad-Cre, 20X magnification. https://doi.org/10.7554/eLife.12203.012 To analyze the cell cycle profile of Rock1∆/∆;Rock2∆/∆ cells, cells were stained with propidium iodide (PI) along with control cells for analysis by flow cytometry. Fewer double knockout than wild-type MEFs were in G1 or S phase, and a higher percentage of cells in G2/M phase of the cell cycle (Figure 3E). The increased percentage may be due to the fact that binucleate cells in G1 will have the same DNA content as cells in G2/M. To investigate a potential cell cycle block, we treated Rock1∆/∆;Rock2∆/∆ cells and their controls with nocodazole, commonly used to arrest cells with a G2/M phase DNA content (Blajeski et al., 2002). In controls, nocodazole led to an increased proportion of cells with G2/M DNA content as expected (Figure 3—figure supplement 1D). In Rock1∆/∆;Rock2∆/∆ cells, however, distribution of cells in the three cell cycle phases is very similar between control treated and nocodazole treated cells (Figure 3—figure supplement 1D). These data suggest that ROCK depletion causes a cell cycle block prior to G2/M. As there is an overall decrease in S phase cells, the block is likely to be in the G1 phase of the cell cycle. ROCK regulates cell cycle proteins CKS1 and CDK1 We have shown that ROCK depletion or inhibition induces cellular senescence; therefore, we next investigated whether they affected key cell cycle proteins involved in senescence. Two major pathways are activated by senescence signals. Activation of the transcription factor Trp53 induces expression of p21, a cyclin-dependent kinase inhibitor (CDKI) (Brown et al., 1997). The other pathway is induction of the CDKI p16, which prevents CDK4/6-mediated phosphorylation of retinoblastoma (Rb) proteins, thereby blocking E2F-dependent transcription (Brenner et al., 1998; Nevins, 2001). To determine whether blockade of Rb or Trp53 function would overcome senescence resulting from abrogation of ROCK function, we stably infected Rock1f/f;Rock2f/fMEFs with simian virus 40 large T-antigen (SV40 Large T), which is known to inactivate the tumor suppressors Trp53 and Rb (DeCaprio et al., 1988; Kierstead and Tevethia, 1993). The cells were then infected with Ad-Cre, to generate Rock1∆/∆;Rock2∆/∆ cells (Figure 4—figure supplement 1A). Inactivation of Trp53 and Rb did not prevent the senescence phenotype caused by loss of ROCK, and cells displayed a similar growth defect to the one observed in Rock1∆/∆;Rock2∆/∆ MEFs without SV40 Large T (Figure 4—figure supplement 1A). To further confirm that the senescent phenotype did not involve activation of Trp53, MEFs were isolated from Rock1f/f;Rock2f/f;Trp53f/f mice and ROCK and Trp53 depleted in vitro by Ad-Cre infection. Rock1∆/∆;Rock2∆/∆;Trp53∆/∆ MEFs were found to undergo senescence, and were defective in their proliferation (Figure 4—figure supplement 1B). Western blot analysis confirmed depletion of both ROCKs and Trp53 (Figure 4—figure supplement 1B). We then treated Trp53 wild type and null cells with the ROCK inhibitors H1152 and GSK269962A and found an increase in senescence compared to control cells (Figure 4—figure supplement 1C). Therefore, the classical senescence pathways did not seem to be affected in Rock1∆/∆;Rock2∆/∆ cells or cells treated for prolonged periods with H1152. To identify protein changes associated with senescence, we used immunoblotting and quantitative mass spectrometry (Ong et al., 2002). The protein levels of some cell cycle regulators appeared to change upon treatment with Cre recombinase alone, so for proteomics analysis we focused on long-term H1152 treatment. Annotation enrichment analysis (Cox and Mann, 2008) of the proteome-wide changes after long-term ROCK inhibition revealed many cell cycle-related protein categories that were significantly modulated in ROCK-inhibited cells (Supplementary file 2), consistent with a prominent cell-cycle phenotype. However, when we analyzed proteins that are known to control G1 cell-cycle arrest and senescence, we did not detect significant changes in the levels of these proteins. These included known regulators, such as p16, p19, p27, p21, CyclinD1, CyclinE, or phosphorylation of Rb (Supplementary file 2 and Figure 4—figure supplement 2A,B). However, CDK1, CyclinA and CKS1 protein levels were significantly reduced upon loss of and long-term inhibition of ROCK1 and 2 (Supplementary file 2 and Figure 4A–C). Long-term treatment of cells with blebbistatin (blebbistatin +/-), but not the inactive enantiomer (blebbistatin +), also decreased the levels of CDK1, CyclinA and CKS1 (Figure 4B,C), suggesting that it is loss of contractility downstream of ROCK loss that results in down-regulation of these proteins. However, it cannot be ruled out that both ROCK inhibition/depletion as well as myosin II inhibition independently lead to a similar phenotype. Figure 4 with 2 supplements see all Download asset Open asset Downregulation of CKS1, CyclinA and CDK1 on abrogation of ROCK function. (A) MEFs were treated with H1152 or vehicle for 3 days and the proteome analyzed by quantitative mass spectrometry. Graph shows log ratios of identified proteins. Data are from duplicate experiments of reciprocal SILAC labelling. The ‘Significant-B’ outlier test was used to determine significantly regulated peptides or proteins, using a Benjamini-Hochberg FDR rate of 5%. (B, C) MEFs of different genotypes or treated with indicated inhibitors were immunoblotted for pCDK1, total CDK1, Cyclin A, CKS1, ROCK1, ROCK2 and total ERK as loading control (B). (C) Quantification of western blot analyses. Graphs show protein levels of CDK1, CyclinA and CKS1 divided by total ERK protein levels and SEM in in
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
