|Home | About | Journals | Submit | Contact Us | Français|
Non-muscle myosin II (NMII) motor proteins are responsible for generating contractile forces inside eukaryotic cells. There is also a growing interest in the capacity for these motor proteins to influence cell signaling through scaffolding, especially in the context of RhoA GTPase signaling. We previously showed that NMIIA accumulation and stability within specific regions of the cell cortex, such as the zonula adherens (ZA), allows the formation of a stable RhoA signaling zone. Now we demonstrate a key role for Coronin 1B in maintaining this junctional pool of NMIIA, as depletion of Coronin 1B significantly compromised myosin accumulation and stability at junctions. The loss of junctional NMIIA, upon Coronin 1B knockdown, perturbed RhoA signaling due to enhanced junctional recruitment of the RhoA antagonist, p190B Rho GAP. This effect was blocked by the expression of phosphomimetic MRLC-DD, thus reinforcing the central role of NMII in regulating RhoA signaling.
Non-muscle Myosin II (NMII) is a fundamental determinant of cell and tissue organization in vertebrates and invertebrates.1,2 As the dominant force generator in non-muscle cells, NMII binds to F-actin and elicits contractility through filament sliding, which in turn influences diverse processes ranging from cell migration and adhesion to cytokinesis and cell proliferation.1,2 In epithelia, prominent contractile networks are found in the apical poles of embryonic epithelial cells (medioapical networks3) and at the cortices of cell-cell adherens junctions (AJ).4-6 At the AJs, actomyosin networks interact with E-cadherin-based adhesions that couple adjacent cells together to form cohesive monolayers. These structures are especially apparent in polarized epithelial cells, where prominent actomyosin bundles lie adjacent to the E-cadherin rings found at the apical region of cell-cell junctions, also known as the zonula adherens (ZA).4-6 The physical coupling of actomyosin to E-cadherin adhesions results in contractile tension at junctions, which supports tissue cohesion, epithelial integrity and morphogenesis.7-13
The assembly and activity of junctional actomyosin is subject to regulation by numerous cell signaling pathways. Of these, the best understood is the canonical RhoA-ROCK pathway, where ROCK mediates signaling by active, GTP-loaded RhoA, to promote NMII activity through phosphorylation of its regulatory light chain.14,15 However, it has become increasingly apparent that, apart from being activated by upstream cues, NMII can also exert influence upon these signaling pathways.1,7,16-18 For example, NMII can associate with RhoA regulators, and, indeed, inactivating NMII with blebbistatin was reported to activate GEFs that promoted Rac signaling.18 Recently, we identified a novel positive feedback loop for NMII-dependent RhoA signaling, where junctional NMIIA supports GTP-RhoA at the ZA.7 This occurs through anchorage of ROCK1 to the ZA by NMIIA, which then prevents the RhoA antagonist, p190B Rho GAP, from associating with the junctional cortex and down-regulating RhoA signaling. This enables junctional NMIIA to effectively maintain RhoA signaling at the ZA following activation by exchange factors, such as Ect2.7,9 An important implication of this positive feedback network lies in the prediction that the cortical stability of NMII spatially defines where a stable RhoA zone may be established. Indeed, computational modeling suggested that NMIIA stability, relative to that of cortical RhoA itself, was an important determinant of the RhoA signaling zone. Since NMIIA was more stable at the ZA than in other regions of cell-cell junctions, this defined where the junctional RhoA zone was established.7
In addition to activation by RhoA-ROCK signaling2,19, the junctional localization of NMIIA can also be influenced by the F-actin networks with which it associates.20,21 Actin regulators comprise actin polymerization factors as well as regulators of F-actin architecture. Both these groups of proteins can modulate NMII localization and cortical stability by either altering cellular F-actin content or through reorganization of filament networks to facilitate the incorporation of NMII. We recently identified an important role for the actin regulator, Coronin 1B, in the assembly of the actomyosin apparatus at the ZA.20 We found Coronin 1B was necessary for the architecture of non-aligned F-actin filaments generated by Arp2/3 to be reorganized into aligned bundles, and thereby promote the effective incorporation of NMII to generate contractility and maintain junctional tension.20 Thus, depletion of Coronin 1B, while dramatically altering F-actin architecture, also reduced junctional NMII localization at the ZA. Given the capacity for NMII to feedback to regulate RhoA,20 we then wondered if this capacity for Coronin 1B to influence NMII recruitment at the junction, might translate to affect RhoA signaling itself.
We started by characterizing the localization of endogenous Coronin 1B in MCF-7 cells, a human mammary epithelial cell line. In confluent monolayers, Coronin 1B concentrated at apical junctions along with F-actin and E-cadherin (Fig. 1A), as we had previously demonstrated in Caco-2 colon epithelial cells.20 RNAi-mediated depletion of Coronin 1B in MCF-7 cells (Fig. 1B) perturbed F-actin organization and resulted in the fragmented distribution of the E-cadherin ring at junctions (Fig. 1A), thus reinforcing its role in maintaining the ZA architecture in epithelial cells.20
We then examined the localization of non-muscle myosin II in MCF-7 cells depleted of Coronin 1B, as E-cadherin and F-actin have been shown to influence the recruitment of non-muscle myosin II to junctions.4,6,19 As previously described,4,6,19 NMIIA and NMIIB concentrated prominently at the zonulae adherente of control MCF-7 monolayers (Fig. 1D, E). However, junctional localization for both isoforms was substantially reduced in Coronin 1B siRNA cells while total proteins levels remain unaltered (Fig. 1D–F). Quantification of junctional fluorescence intensity revealed a ~3-fold reduction of NMIIA and NMIIB in Coronin 1B knockdown (KD) cells (Fig. 1G, H). This indicated that Coronin 1B supports the assembly of an actomyosin cortex in MCF-7 epithelial cells.
To understand if the reduced NMII distribution in Coronin 1B KD cells was a consequence of altered NMII stability, we analyzed the junctional dynamics of NMII by fluorescence recovery after photobleaching (FRAP) in both Caco-2 cells and MCF-7 cells (Fig 2). GFP-NMIIA, the dominant NMII isoform in both MCF7 and Caco-2 cells, was expressed in control and Coronin 1B KD cells and fluorescence recovery was quantified in a defined focal area at the apical ZA.4,22,23 Analysis of the fluorescence recovery of GFP-NMIIA in MCF-7 cells revealed a striking increase in the mobile fraction in Coronin 1B depleted cells, suggesting that NMIIA mobility is less restricted in Coronin 1B KD cells (Fig. 2A). Furthermore, this observation appeared consistent across cell lines as junctional NMIIA exhibited a similar recovery profile in Coronin 1B KD Caco-2 cells (Fig. 2B, Fig. 1C). Taken together, these observations indicate that Coronin 1B, through its role in the organization of F-actin, plays a key function in the regulation of NMIIA stability at the ZA.
We then postulated that the effect of Coronin 1B on NMIIA junctional stability could have consequences for RhoA signaling. To test this hypothesis, we immunolabelled for endogenous RhoA, as we previously found that the junctional localization of RhoA requires it to be active and thus can be used as a proxy for RhoA activity.9 We studied this in confluent MCF-7 and Caco-2 cells, expressing either control or Coronin 1B siRNA (Fig. 3A, B). As described previously, RhoA was concentrated at the apical junctions of both cell lines (Fig. 3A, B).7,24 However, Coronin 1B depletion significantly decreased junctional RhoA, as confirmed by fluorescence intensity analyses for both MCF-7 and Caco-2 cells (Fig. 3A, B). This suggested that Coronin 1B is required for RhoA localization at junctions, thereby potentially regulating its activity and signaling.
In considering candidates that might mediate the effect of Coronin 1B on RhoA activity, we focused on Rac1 which can antagonise RhoA signaling in various contexts.25,26 Further, non-muscle myosin itself can negatively regulate Rac activity by modulating the localization of Rac GEFs.16-18 To pursue this, we examined Rac1 localization in the cells depleted of Coronin 1B (Fig. 4A). Rac1 stained prominently at the junctions of MCF-7 cells and this was not altered in Coronin 1B KD cells, suggesting that Coronin 1B-NMIIA does not support RhoA signaling by inhibiting the Rac1 pathway (Fig. 4A). Alternatively, we recently identified a positive feedback mechanism whereby NMIIA, through the anchorage of ROCK-1, inhibits the junctional recruitment of p190B Rho GAP to support RhoA-GTP signaling at the ZA.7 Since Coronin 1B depletion affected NMIIA stability and reduced RhoA localization to the ZA, we wondered if these effects were mediated via the previously established p190B Rho GAP pathway. Indeed, the weak p190B Rho GAP junctional localization visualized in steady-state MCF-7 monolayers was significantly enhanced in Coronin 1B KD cells and line-scan analysis confirmed this observation (Fig. 4B). Coronin 1B KD also reduced ROCK-1 localization (the upstream inhibitor of p190B Rho GAP) thus further explaining the increase in junctional p190B Rho GAP in Coronin 1B KD cells (Fig. 4C). These observations support the notion that loss of Coronin 1B results in the destabilization of NMIIA at the ZA, which leads to the activation of the p190B Rho GAP pathway, thus inhibiting GTP-RhoA signaling.
We then predicted that, if the effect of Coronin 1B on RhoA-signaling is mediated via NMIIA, then restoration of junctional NMIIA should be able to rescue RhoA localization in Coronin 1B KD cells. The phosphomimetic form of the myosin regulatory light chain (MRLC-DD) has been previously shown to restore NMII contractility to cells when its upstream activators are inhibited.27,28 Indeed, expression of MRLC-DD in Coronin 1B KD MCF-7 cells restored junctional RhoA localization. (Fig. 5A and B). We further confirmed this finding by using a location biosensor for GTP-RhoA (GFP-AHPH), which comprises the C-terminus of anillin tagged with GFP.7 The robust junctional localization of GFP-AHPH observed in controls was significantly attenuated in Coronin1B KD cells (Fig. 5C and D), confirming that the observed changes in endogenous RhoA immunostaining paralleled change in the active form of RhoA. However the expression of MRLC-DD in Coronin1B KD cells was able to restore junctional GFP-AHPH to control levels (Fig. 5C and D). Together, these findings indicate that Coronin 1B supports junctional RhoA signaling by maintaining the junctional accumulation of NMII.
Thus, our results identify an essential role for Coronin 1B in sustaining RhoA signaling at epithelial cadherin junctions through stabilization of NMIIA at the ZA (Fig. 6). NMIIA in turn antagonizes the junctional recruitment of p190B Rho GAP, thus suppressing RhoA-inhibitory signals and keeping RhoA GTP-bound. Accordingly, we observed that loss of Coronin 1B led to RhoA inactivation by enhancing the p190B Rho GAP pathway and this could be counteracted by driving the activity of NMII. While the stability of NMII can be regulated by various factors, including its activation via RhoA kinases or the phosphorylation of its heavy-chain,2,29 actin architecture can also play an important role in determining NMII organization and cellular contractility.21,30 In support of this, we identify a necessary role for the actin regulator, Coronin 1B, which strengthens NMIIA localization at ZA through the organization of junctional F-actin networks into bundles. As a result, this apically stabilized NMIIA provides a cortical landmark for a feedback network that confers robustness for RhoA signaling.7
RhoA signaling is a master regulator of actomyosin cytoskeleton and thus has been implicated in a variety of morphogenetic processes.31-36 While the role for active RhoA signaling in facilitating essential cellular processes has been well-established9,37-41 there are circumstances where RhoA signaling must be downregulated, either physiologically or pathologically to elicit a biological response.41,42 Given the importance of the actomyosin cytoskeleton in positively regulating RhoA signaling,7,43-45 targeting actin-regulators such as Coronin 1B might be one of the approaches to achieve a precise spatiotemporal regulation of RhoA. In summary, our analysis contributes to the growing notion that Rho-GTPases and cytoskeleton regulators act in sync to establish epithelial homeostasis and contractility.
MCF-7 cells and Caco-2 cells were procured from ATCC and cultured in DMEM or RPMI respectively; supplemented with 10% foetal bovine serum (FBS), 1% non-essential amino acids, 1% L-glutamine, 100 U/ml penicillin and 100 U/ml streptomycin. For siRNA-mediated knockdown, cells were grown till 50% confluent and then transfected using RNAi max (Invitrogen), according to the manufacturer's instructions. Cells were harvested for western-blotting assay or immunofluorescence assay 48–72 hours post-transfection. For plasmids expression, cells were grown till 80–90% confluent, transfected using Lipofectamine 3000 and harvested 24 hours post-transfection. For imaging live-cells; cells were grown on 29 mm glass-bottom dishes (Shengyou Biotechnology) and imaged in clear Hank's balanced salt solution supplemented with 5% FBS, 10 mM HEPES (pH 7.4) and 5 mM CaCl2.
Coronin 1B was depleted using siRNA designed using Invitrogen Block-iT RNAi designer against the ORF regions. The sequences are as follows:
5′ GGCAGAGCAAAUUCCGGCAUGUGUU 3′
5′ CACCCUGACCUCAUCUACAAUGUCA 3′
Cherry-MRLC-DD was generated using plasmid Plasmid #35682 from Addgene as a template where EGFP was replaced by m-cherry from pmCherry-C1 (Clontech) using AgeI and BsrGI restriction sites. GFP–myosin IIA was procured from Addgene (no. 11347).
Primary antibodies used in this study were: rabbit polyclonal antibody (pAb) for non-muscle myosin IIA heavy chain (1:1,000; no. PRB-440P, Covance); rabbit pAb for non-muscle myosin IIB heavy chain (1:1,000; no. PRB-445P, Covance); mouse mAb (1:200; no. A-11120, clone 3E6) against GFP (Molecular Probes/Invitrogen); rabbit pAb (1:300; no. 61-7300, Invitrogen) and mouse mAb against human ZO-1 (1:300, no. 33-9100, clone ZO1-1A12, Invitrogen); mouse mAbs against RhoA (1:100; Santa Cruz Biotechnology, clone 26C4, no. sc418); mouse mAbs against p190B Rho GAP (1:50, no. 611612, clone 54/P190-B) (BD Biosciences); rabbit pAb against ROCK1 (1:300, no. AB134181, Abcam); rat mAb E-cadherin (1:500, no. 13-1900, clone ECCD-2, Invitrogen); mouse mAb Rac1 (1:300, Merck, #05-389); Rabbit antibody against GAPDH (1:4,000; Trevigen; catalog number 2275-PC-100; polyclonal); RFP-ds-Red antibody (#3993-100; 1:200); mouse monoclonal against Rac1 (#05-389, 1:200); mouse monoclonal antibody [3D10] against Coronin 1B (# ab119071).
Secondary antibodies were species-specific antibodies conjugated with AlexaFluor 488, 594 or 647 (Invitrogen, 1:500) for immunofluorescence, or with horseradish peroxidase (Bio-Rad Laboratories, 1:5,000) for immunoblotting.
MCF-7 and Caco-2 cells were fixed with freshly prepared 10% TCA (Sigma # T0699-100 ML) on ice for 15 minutes or with ice-cold methanol for 5 minutes in − 20°C or with 4% PFA in cytoskeleton stabilization buffer on ice for 20 minutes. TCA fixation was used for all immunofluorescence experiments pertaining to RhoA and Rac1. The TCA-fixed cells were subsequently washed thrice with 30mM glycine. The TCA and PFA fixed cells were permeabilized using 0.25% Triton X-100 for 5 minutes at room temperature and washed thrice with PBS. Confocal images were acquired using Zeiss LSM-710 or Zeiss LSM-510 META inverted microscopes with identical laser power, gains and digital offset settings being applied to each experimental group analyzed. For representation purpose, both control and treated images were processed in an identical manner using Fiji by applying a mean filter of 0.5-1-pixel radius or by using rolling-ball background subtraction function of Fiji.
Fluorescence intensity at junctions was calculated using the multi-plot function of ImageJ. Briefly, a line of 10–20 μm in length was drawn orthogonal to the junctions with its center upon junctions. Multi-plot of Fiji was then used to get the numerical values for the fluorescence intensity. Background correction was done by subtracting a constant value from each of the intensity profiles and then fitted to Gaussian function using prism software. Non-linear regression function was used to determine the peak-values. For each experiment, a minimum of 40–50 contacts were analyzed per condition.
For RhoA biosensor experiment, the mean-gray values of GFP-AHPH junctional and cytoplasmic intensity were procured using Fiji and their intensity-ratio was used for statistical analysis. Only junctions between transfected cells were analyzed.
MCF-7 and Caco-2 cells grown on 29 mm glass-bottom dishes (Shengyou Biotechnology) were transfected with Coronin 1B siRNA and 48 hours post-transfection; GFP-NMIIA was transfected and imaged in clear Hank's balanced salt solution. Image acquisition and analysis has been described previously.7,23
No potential conflicts of interest were disclosed.
We thank all our lab colleagues for their support and advice during this work, and M. Naghibosadat for her help with cloning. Confocal and optical microscopy was performed at the ACRF Cancer Biology Imaging Facility, established with the generous support of the Australian Cancer Research Foundation.
This work was supported by grants and fellowships from the National Health and Medical Research Council of Australia (APP1037320, APP1067405, APP1044041), The Australian Research Council (DP150101367) and The Cancer Council of Queensland (APP1086857). GAG was supported by Kids Cancer Project of the Oncology Research Foundation and a University of Queensland Early Career Research Grant. RP was supported by ANZ Trustees PhD Scholarship in Medical Research.