Cysteine-rich protein 1 (CRP1) regulates actin filament bundling
© Tran et al; licensee BioMed Central Ltd. 2005
Received: 02 September 2005
Accepted: 08 December 2005
Published: 08 December 2005
Skip to main content
© Tran et al; licensee BioMed Central Ltd. 2005
Received: 02 September 2005
Accepted: 08 December 2005
Published: 08 December 2005
Cysteine-rich protein 1 (CRP1) is a LIM domain containing protein localized to the nucleus and the actin cytoskeleton. CRP1 has been demonstrated to bind the actin-bundling protein α-actinin and proposed to modulate the actin cytoskeleton; however, specific regulatory mechanisms have not been identified.
CRP1 expression increased actin bundling in rat embryonic fibroblasts. Although CRP1 did not affect the bundling activity of α-actinin, CRP1 was found to stabilize the interaction of α-actinin with actin bundles and to directly bundle actin microfilaments. Using confocal and photobleaching fluorescence resonance energy transfer (FRET) microscopy, we demonstrate that there are two populations of CRP1 localized along actin stress fibers, one associated through interaction with α-actinin and one that appears to bind the actin filaments directly. Consistent with a role in regulating actin filament cross-linking, CRP1 also localized to the membrane ruffles of spreading and PDGF treated fibroblasts.
CRP1 regulates actin filament bundling by directly cross-linking actin filaments and stabilizing the interaction of α-actinin with actin filament bundles.
Stress fibers are bundles of actin microfilaments formed in cells following integrin-mediated attachment and spreading . Regulation of these contractile fibers is critical for cell adhesion and motility. The microfilaments within stress fibers are held together by specialized bundling proteins, such as myosin and α-actinin, which can interact simultaneously with two actin filaments. These classical actin-bundling proteins have been studied extensively leading to a basic understanding of their interaction with actin filaments and regulation of stress fibers. One important discovery was the periodic and alternating association of myosin and α-actinin which is clearly visualized as a beaded pattern along stress fibers in cells stained for immunofluoresence microscopy . Although it is not understood how this alternating association of myosin and α-actinin with the microfilaments is regulated, it is critical for the contractility of the stress fiber. In addition to myosin and α-actinin, actin stress fibers are decorated with numerous other proteins, associated either directly or through interaction with other actin-binding proteins. Determining the function of these ancillary proteins is important for understanding the regulation of stress fibers.
Modulation of the actin cytoskeleton by LIM domain proteins is an active and growing field of research . LIM domains are cysteine-rich sequences of 50–60 amino acid residues that contain two tandem zinc fingers . Several LIM proteins have been demonstrated to interact with and/or regulate α-actinin. Vallenius et al.  have reported that reversion-induced LIM (RIL) protein associates with α-actinin increasing its binding to actin filaments in vitro and may alter actin stress fibers in various cell types. Four and a half LIM domain protein 3 (FHL3) has been demonstrated to disrupt actin stress fibers in C2C12 myoblasts presumably by binding to actin filaments and inhibiting α-actinin bundling . In addition, ENH , ALP , Cypher , CLIM1 , CLP-36 [11, 12], zyxin , and the cysteine-rich protein (CRP) family [14, 15] have also been demonstrated to interact with α-actinin, although it is not clear if these proteins influence α-actinin function.
The CRP family, which includes CRP1, CRP2, CRP3, and the thymus LIM protein (TLP), is a subgroup of LIM domain proteins containing two LIM domains linked to short glycine-rich repeats . CRPs are highly conserved between species with CRP1 from human, chicken, and quail having an amino acid sequence identity greater than 90% . Although the CRPs have different patterns of expression, evidence suggests that they are functionally similar [14, 16]. CRPs are important for cell differentiation presumably by modulating protein-protein interactions involved in transcriptional regulation [16, 18]. Outside of the nucleus, CRPs clearly localize to focal adhesions and the actin cytoskeleton, and have been postulated to play a role in controlling these structures .
CRP1, CRP2, and CRP3, have all been shown to bind α-actinin [14, 15]. Interaction between the two proteins was first reported for CRP1 using affinity chromatography, solution and solid-phase binding assays, and co-immunoprecipitation from cell lysates . A follow up study demonstrated that α-actinin could interact equally with CRP1, CRP2, and CRP3 . In addition, it was determined that the CRPs were interacting with the actin-binding domain of α-actinin  and the α-actinin-binding site was mapped to amino acid residues 62–79 of human CRP1 . Based on the binding assays and co-localization within the cell, it was proposed that the interaction of CRP with α-actinin was responsible, in part, for its localization to the actin cytoskeleton [14, 15]. Recently, Grubinger and Gimona  demonstrated that CRP2 binds directly to actin filaments and suggested that CRP2 does not interact with α-actinin. We have found that there are two populations of CRP1 associated with actin stress fibers, one interacting with α-actinin and one that appears to interact directly with the actin filaments. Although, CRPs have been shown to bind to α-actinin and actin filaments, it is not clear how these proteins regulate the actin cytoskeleton. In this study, we show that CRP1 regulates actin filament bundling by directly cross-linking actin filaments and stabilizing the interaction of α-actinin with actin filament bundles.
CRP1 is a LIM domain containing protein which has been demonstrated to localize in the nucleus, the cytoplasm, along actin filaments, and in focal adhesions. Recent studies have shown that the nuclear population of CRPs mediates protein-protein interactions regulating transcription and differentiation. Although CRPs have been found to bind to α-actinin, zyxin, and actin filaments, little is known about how CRPs regulate the actin cytoskeleton. In this study, we examined the regulation of actin filaments by CRP1 in vitro and in cultured cells.
Expression of CRP1 in fibroblasts resulted in increased actin filament bundling (Fig. 1). In addition, we found that CRP1 directly binds to and bundles actin filaments (Fig. 2 and 4). CRP2 has recently been reported to bind actin filaments ; however, this is the first study to show that any CRP can bundle actin filaments.
After observing an increase in actin filament bundling in the cells expressing CFP-CRP1, we expected to find that CRP1 was enhancing the bundling activity of α-actinin. However, in vitro bundling studies clearly demonstrated that CRP1 had no influence on the ability of α-actinin to bundle actin filaments (Fig. 3). Furthermore, these results showed that CRP1 and α-actinin do not compete for binding to actin filaments and therefore must bind at different sites. Interestingly, approximately twice as much CRP1 protein pellets with the actin filament bundles compared to α-actinin. The significance of these findings is not clear, but may reflect differences in the mechanisms by which CRP1 and α-actinin bind and bundle actin filaments. Further studies are necessary to determine how CRP1 is bundling actin filaments.
Although α-actinin had previously been demonstrated to bind and presumably localize CRP1 to actin filaments [14, 15], the new evidence that CRP1 could directly bind and bundle actin filaments prompted further investigation. Detergent extraction of cells has long been used to improve examination of the cytoskeleton. Previously, we demonstrated that Triton X-100 extraction removes soluble cytoplasmic proteins leaving behind an intact cytoskeleton with associated adhesion and matrix proteins. Extraction with Triton X-100 has also been used to separate and define adhesion and cytoskeletal proteins based on stable association with the actin cytoskeleton . Triton X-100 extraction of fibroblasts co-expressing CFP-CRP1 and YFP-α-actinin allowed us to differentiate two populations of CRP1 along actin stress fibers (Fig. 5). The Triton X-100 resistant population of CRP1 co-localized with α-actinin along stress fibers, whereas the less stable Triton X-100 susceptible population appeared to represent CRP1 which was directly associated with actin filaments. Furthermore, confocal microscopy demonstrated that CFP-CRP1 and YFP-α-actinin were in close enough proximity for FRET between the CFP and YFP (Fig. 6). Since the tags need to be within ~50Å for efficient FRET to occur , the results indicate that CRP1 and α-actinin are bound to each other along actin stress fibers.
Evidence from this and other studies suggests that the different populations of CRP1 mediate protein-protein interactions unique to each cellular compartment. Nuclear CRP1 regulates interactions between transcription factors. Cytoplasmic CRP1 modulates the actin cytoskeleton by two mechanisms: 1) stabilizing α-actinin interaction with actin bundles and 2) cross-linking actin filaments. Further studies are needed to determine how the different populations of CRP1 coordinate to modulate the function of the cell.
α-Actinin was purified from chicken gizzard as previously described . Non-muscle actin (99% pure; 80% β-actin, 20% γ-actin) was polymerized following the manufacturer's protocol (Cytoskeleton, Inc., Denver, CO). The plasmid containing chicken CRP1 cloned into the EcoRI site of pBlueScript II KS (pBSIIKS, Stratagene) was generously provided by Mary C. Beckerle (Univ. of Utah) . The EcoRI insert of CRP1-pBSIIKS was sub-cloned into the enhanced cyan fluorescent fusion protein vector pECFP-C1 (BD Biosciences) and the BamHI-HindIII insert sub-cloned into pProEx HTb (Invitrogen). Nucleotide sequences were confirmed by sequence analysis. His-tagged CRP1 protein was expressed in BL21 bacteria and purified using Ni-NTA resin (Qiagen) following procedures described by the manufacturer. The his tag was cleaved from CRP1 while still bound to the resin using recombinant TEV protease (Invitrogen) following the manufacturer's protocol. The untagged CRP1 protein was concentrated and buffer-exchanged (10 mM HEPES, pH 7.0, 50 mM NaCl, 1 mM EDTA) in a centrifugal filter device (Amicon). YFP-α-actinin was constructed by subcloning the α-actinin gene  into the HindIII restriction site of the enhanced yellow fluorescent fusion protein vector pEYFP-N1 (BD Biosciences).
The bundling of F-actin was determined by sedimentation assays as previously described [28, 29]. F-actin (10.4 μM) was incubated with the indicated concentration of CRP1 or α-actinin in bundling buffer (10 mM HEPES, pH 7.0, 50 mM NaCl, 1 mM EDTA) for 30 min at room temperature and centrifuged at 10,000 × g for 30 min. The supernatant and pellet were separated and the proteins analyzed by electrophoresis. Proteins were detected by Gelcode Blue (Pierce) staining and quantified using a KODAK ImageStation 440CF.
F-actin bundles were visualized by fluorescence microscopy following modification of previously described procedures [30, 31]. Briefly, 50 μl of assay solution was incubated on a glass coverslip inside a 12-well tissue culture dish. After 30 min, the proteins were fixed by adding 3% formaldehyde in phosphate buffered saline for an additional 30 min. Coverslips were then stained and processed for microscopy as described previously.
Rat embryonic fibroblasts (REFs) were cultured as described previously . Cells were transfected with pECFP-CRP1 and pEYFP-α-actinin using FuGENE 6 (Roche) following manufacturer's protocols. The expression curve of CFP-CRP1 was carried out by varying the ratio of FuGENE to DNA in a final volume of 100 μL serum-free media. The transfection conditions were: 6 μL FuGENE, no DNA; 3 μL FuGENE, 0.5 μg DNA; 3 μL FuGENE, 1.0 μg DNA; 6 μL FuGENE, 1.0 μg DNA; and 6 μL FuGENE, 2.0 μg DNA. Twenty-four hours after transfection, cells were prepared for fluorescence microscopy or scraped into ice-cold lysis buffer (10 mM Tris, pH 7.4, 150 mM NaCl, 1 mM EGTA, 1 mM EDTA, 2 mM Na3VO4, 1% Triton X-100, 0.5% NP-40, 30 mM sodium pyrophosphate, 50 mM NaF, 1 μg/ml leupeptin, and 1 μg/ml aprotinin) as described previously . The lysates were centrifuged at 10,000 × g for 10 min at 4°C, protein from the supernatant and pellet separated by electrophoresis, and immunoblotted with anti-GFP (Santa Cruz), anti-α-actinin (Chemicon), or anti-actin (Sigma). Proteins were detected by enhanced chemiluminence (Pierce) and quantified using a KODAK ImageStation 440CF. For fluorescence microscopy, cells were fixed for 30 min at room temperature with 3% formaldehyde (Tousimis) in PBS or in Triton X-100 buffer (20 mM Tris, pH 7.4, 50 mM NaCl, 1 mM EGTA, 5 mM EDTA, 100 μM Na3VO4, 50 mM sodium pyrophosphate, 1 μg/mL leupeptin, 1 μg/mL aprotinin, and 0.5 % Triton X-100). Digital images were captured using a Zeiss axiovert 100S microscope equipped with a Photometrics CoolSNAP HQ CCD camera controlled by MetaMorph software. Co-localization studies were carried out using a Zeiss LSM 510 confocal microscope. The scatter plots and correlation coefficients were determined using Zeiss Physiology Software v3.2.
REFs were co-transfected with pECFP-CRP1 and pEYFP-α-actinin, cultured for an additional 24 hrs, fixed using 3% formaldehyde in Triton X-100 buffer, and prepared for confocal microscopy as described above. The FRET assays were carried out following the procedure described by Karpova et al. . Briefly, cells were imaged with a Zeiss LSM 510 confocal microscope operated by Zeiss Physiology Software v3.2 using a 63 × 1.3 NA Zeiss oil immersion lens. The microscopy system was configured in multitracking mode to excite the CFP with a 458 nm and YFP with a 514 nm laser line. A region of interest (ROI) containing actin stress fibers was selected for photobleaching. Using the time series function, 5 images of the cell were collected followed by selective photobleaching of the YFP within the ROI with the 514 nm laser line (typically, 150 iterations at 100% laser power was sufficient), and then the collection of 5 additional images. The FRET efficiency was calculated as a percentage using the following formula E = 100 × (Ipostbleach – Iprebleach)/Iprebleach, where I is the intensity of CFP fluorescence within the ROI. As a control, ROIs were selected from non-bleached regions of the cell.
Four and a half LIM domain protein 3
enigma homologue protein
actinin-associated LIM protein
36kDa carboxyl terminal LIM domain protein
platelet-derived growth factor
cyan fluorescent protein
yellow fluorescent protein
rat embryonic fibroblasts.
The authors thank Dr. Mary Beckerle (Univ. of Utah) for providing CRP1-pBSIIKS. Research was supported by a grant to J.A.G. from the National Institutes of Health (GM 63711). This publication was made possible in part by the Confocal Microscopy (grant number 1S10RR107903-01) and Cell Imaging and Culture Facility of the Environmental Health Sciences Center, Oregon State University, fromgrant number P30 ES00210, National Institute of EnvironmentalHealth Sciences, National Institutes of Health.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.