Báo cáo khoa học: hhLIM is a novel F-actin binding protein involved in actin cytoskeleton remodeling ppt

In vitro analyses showed that green fluorescent protein GFP-tagged hhLIM pro-tein accumulated in the cytoplasm of C2C12 cells and colocalized with F-actin, indicating that hhLIM is an act

cytoskeleton remodeling

Bin Zheng, Jin-kun Wen and Mei Han

Department of Biochemistry and Molecular Biology, Hebei Medical University, Shijiazhuang, China

The actin cytoskeleton is a highly organized and

dynamic structure present in all eukaryotic cells, where

it plays a central role in many processes including

intracellular transport and cell growth, signaling, and

division Many of the actin-binding proteins affect the

cytoskeletal structure and architecture by mediating

the association of actin filaments into cables and

bun-dles and cross-linking these structures into complex

networks [1] The data presented here demonstrate that

human heart LIM protein (hhLIM) is an actin-binding

protein that participates in remodeling of the actin

cytoskeleton, possibly by promoting actin bundling

The LIM domain

[CX2CX16–23HX2CX2CX2-CX16–21CX2(C⁄ H ⁄ D), where X denotes any amino

acid] is a cysteine-rich zinc-finger motif found in a

large family of proteins and now recognized as a key component of the regulatory machinery of the cell [2–4] Recent studies have indicated that proteins containing LIM domains have diverse cellular roles as regulators of gene expression, cytoarchitecture, cell adhesion, cell motility and signal transduction [3,5] hhLIM, also named hLIM3 (GenBank AF121260), was cloned by three-element PCR-select cDNA subtrac-tion from the embryo heart cDNA library [6] Using insulin-like growth factor-1 and endothelin-1 as con-trols, our previous studies have shown that: (a) expres-sion of the hhLIM gene is tightly linked to cardiac and skeletal specification, (b) hhLIM plays an impor-tant role in cardiac hypertrophy, (c) hhLIM can shuttle between the nucleus and the cytoplasm and initiate

Keywords

actin-binding protein; cytoskeleton; F-actin;

hhLIM; LIM domain

Correspondence

M Han, Department of Biochemistry and

Molecular Biology, Hebei Medical

University, No 361, Zhongshan East Road,

Shijiazhuang 050017, China

Fax: +86 311 8669 6826

Tel: +86 311 8626 5563

E-mail: hanmei@hebmu.edu.cn

(Received 21 November 2007, revised 15

January 2008, accepted 30 January 2008)

doi:10.1111/j.1742-4658.2008.06315.x

Human heart LIM protein (hhLIM) is a newly cloned protein In vitro analyses showed that green fluorescent protein (GFP)-tagged hhLIM pro-tein accumulated in the cytoplasm of C2C12 cells and colocalized with F-actin, indicating that hhLIM is an actin-binding protein in C2C12 cells Overexpression of hhLIM–GFP in C2C12 cells significantly stabilized actin filaments and delayed depolymerization of the actin cytoskeleton induced

by cytochalasin B treatment Expression of hhLIM–GFP in C2C12 cells also induced significant changes in the organization of the actin cytoskele-ton, specifically, fewer and thicker actin bundles than in control cells, sug-gesting that hhLIM functions as an actin-bundling protein This hypothesis was confirmed using low-speed co-sedimentation assays and direct observa-tion of F-actin bundles that formed in vitro in the presence of hhLIM hhLIM has two LIM domains To identify the essential regions and sites for association, a series of truncated mutants was constructed which showed that LIM domain 2 has the same activity as full-length hhLIM To further characterize the binding sites, the LIM domain was functionally destructed by replacing cysteine with serine in domain 2, and results showed that the second LIM domain plays a central role in bundling of F-actin Taken together, these data identify hhLIM as an actin-binding protein that increases actin cytoskeleton stability by promoting bundling of actin filaments

Abbreviations

CRP, cysteine-rich protein; GFP, green fluorescent protein; GST, glutathione S-transferase; hhLIM, human heart LIM protein; MLP, muscle LIM protein.

cardiac hypertrophy, and (d) hhLIM is a member of

the group of cytosolic LIM proteins and interacts with

skeletal a-actin in the cytoplasm However, little is

known about the mechanism whereby hhLIM interacts

with skeletal a-actin and regulates the organization

and rearrangement of the actin cytoskeleton [7]

hhLIM contains two LIM domains and is most

homologus to the cysteine-rich protein (CRP) family,

which comprises three members (CRP1, CRP2 and

CRP3) hhLIM displayed nuclear, actin-associated,

and nuclear plus actin-associated distributions similar

to those of CRPs But the one-LIM motif Drosophila

protein (DMLP1) displayed a diffuse cytosolic pattern

in subset of cells [8] The LIM homeo-domain protein

Apterous and Isl-1 almost exclusively accumulated in

the nucleus [9] The nuclear functions of CRPs have

been studied over the past two decades and it is now

well established that this subset of LIM proteins are

important regulators of cell differentiation and

tran-scription By contrast, their actin cytoskeleton-related

roles have remained obscure CRPs were first believed

to interact with actin filaments in an indirect manner

through the intermediation of actin-binding protein

partners such as a-actinin or zyxin [10] However, in

agreement with our data on hhLIM, it has been

dem-onstrated that CRP1 and CRP2 have the ability to

interact with actin filaments in a direct manner

Impor-tantly, CRP1 has been shown to induce actin filament

bundling in vitro, as well as in transformed rat

embry-onic fibroblasts [11,12] Taken together, these data

strongly suggest that CRPs and CRP-related LIM

pro-teins participate in regulation of the actin cytoskeleton

architecture [13] Understanding the mechanism of

actin filament stabilization and bundling triggered by

hhLIM and CRPs requires, in the first instance,

identi-fication of their actin-binding domains To date, none

of the actin-binding domain sequences registered in

databases is present in hhLIM or CRPs The goals of

this study were to define the actin-binding properties

of hhLIM and determine the precise actin-binding sites

of hhLIM Our results show that hhLIM binds to

fila-mentous (F) actin and the second LIM domain of

hhLIM plays a central role in this interaction

Results

hhLIM interacts and colocalizes with F-actin

in the cytoplasm of C2C12 cells

Using confocal microscopy we have identified that

hhLIM is colocalized with actin filaments [7] To

fur-ther confirm this interaction, coimmunoprecipitation

and a pull-down assay were performed C2C12 cells

transfected with Myc-tagged hhLIM and GFP-tagged actin were incubated in 2% horse serum to induce dif-ferentiation Extracts were incubated with anti-Myc or anti-GFP Sepharose, and interacting proteins were analyzed by western blotting with antibody specific to actin or GFP antibody Figure 1A shows that actin was specifically immunoprecipitated together with hhLIM To demonstrate that endogenous hhLIM and actin can form a complex in vivo, actin was precipitated from C2C12 cell lysates and the immuno-precipitates were analyzed by western blot using anti-hhLIM Ig The data showed that actin was specifi-cally immunoprecipitated together with endogenous hhLIM, whereas protein A–agarose did not precipitate hhLIM Lysates were immunoprecipitated with anti-hhLIM Ig and detected by anti-actin Ig, and results showed the same specific interaction between endogenous hhLIM and actin, which indicated that the interaction of these two proteins is not an artifact of hhLIM overexpression (Fig 1B) The glutathione S-transferase (GST) pull-down experiment also demon-strated a direct interaction between GST–hhLIM and actin GST or GST–hhLIM fusion proteins were bound

to glutathione–Sepharose and incubated with purified rabbit skeletal muscle actin or lysates from hhLIM-expressing cells After extensive washing, Sepharose pellets were immunoblotted with anti-actin Ig to detect actin in fusion protein or the pellets with anti-GST Ig

to demonstrate equal loading of fusion protein As shown in Fig 1C, both purified actin and endogenous actin bound to GST–hhLIM but not GST

hhLIM bundles F-actin directly

In order to identify whether hhLIM and actin interact directly, we investigated the activities of hhLIM bind-ing to actin usbind-ing a co-sedimentation assay Purified F-actin was incubated with recombinant hhLIM pro-tein, and pelleted by centrifugation at 10 000 g, which allows pelleting of heavy, cross-linked F-actin only Controls for this series of experiments included SM22a, a known actin cross-linking protein, and BSA, which does not interact with or cross-link actin In the absence of hhLIM, the majority of actin remained in the supernatant (S) and only a small amount was detected in the pellets (P) The addition of hhLIM sig-nificantly enhanced the amount of actin present in the pellets (P) compared with samples with actin alone or with the BSA control (Fig 2A) These data indicated that hhLIM binds to and has a bundling effect on actin Figure 2B shows that, in the absence of hhLIM, 20% of the total actin was detected in the pellet By contrast, in the presence of hhLIM, the amount of

actin in the pellet increased along with the increment

of the hhLIM, indicating that hhLIM induces F-actin

bundling Maximum actin bundling occurred when

molar ratios of hhLIM (2 lm) to actin (8 lm) were

> 1 : 4 Indeed, when the concentration of hhLIM exceeded 4 lm, 60% of total actin was detected in the pellet (Fig 2B) Cumulative data from several indepen-dent experiments demonstrated that the co-sediments

of hhLIM and F-actin was greater than that of actin alone (Fig 2C) In order to directly analyze the effect

of GST–hhLIM on actin filament bundling, we per-formed electron microscopy on negatively stained actin filaments As shown in Fig 2D, in the absence of hhLIM, actin filaments formed a uniform meshwork

of fine filaments The inclusion of BSA had no effect

on the ability to bundle actin, however, when actin was polymerized in the presence of hhLIM, higher order structures were observed Although single actin filaments were still present, most of the actin filaments were recruited into thick and long actin bundles, confirming the cross-linking activity of hhLIM To determine whether hhLIM also binds to monomeric (G)-actin, GST pull-down assays were performed with GST–hhLIM versus GST alone Although actin was pulled down with GST–hhLIM, there was no signifi-cant difference between samples containing GST– hhLIM and GST alone (Fig 2E) Thus, this approach suggests that hhLIM does not bind to monomeric actin

hhLIM stabilizes F-actin in C2C12 cells

To further determine whether hhLIM modulates the actin cytoskeleton in C2C12 cells, we studied the effects of hhLIM overexpression on the actin stress fibers Overexpression of hhLIM induced actin poly-merization (data not shown) We have established that overexpression of hhLIM may increase the expression

of actin [7] The actin fractionation assay showed that the F-actin fraction (csk) was increased compared with the G-actin fraction (sol) in cells overexpressing hhLIM Silencing of hhLIM expression by siRNA had the opposite result (Fig 3A) If the expression of GFP–hhLIM could increase actin filament bundling, then GFP–hhLIM would be expected to redistribute

to the Triton X-100-insoluble cytoskeletal fraction As shown in Fig 3B, the insoluble hhLIM fraction increased with in a dose-related manner So, we pre-dicted that hhLIM might participate in F-actin forma-tion and stabilizaforma-tion of actin filaments In order to test whether hhLIM could affect the stability of the actin cytoskeleton following its ectopic expression in C2C12 cells, actin depolymerization was induced by cytochalasin B in hhLIM–GFP–transfected C2C12 cells The actin cytoskeleton was visualized by TRITC– phalloidin staining before adding cytochalasin B, and

(a)

A

B

C

(b)

IP with actin Ab – +

hhLIM

Actin

hhLIM

IP with hhLIM Ab – +

Actin

hhLIM

Actin

– –

+

IP with Myc Ab

GFP

Myc

GST GST-hhLIM

1 2 3 4

Actin

GST

Fig 1 Actin interacts with hhLIM in C2C12 cells

Coimmunopre-cipitation of GFP-tagged actin with myc-tagged hhLIM Lysates of

C2C12 cells transfected with full-length myc-tagged hhLIM and

GFP-tagged actin was immunoprecipitated (IP) by anti-myc Ig

cou-pled to Sepharose, and interacting proteins were separated by

SDS ⁄ PAGE and blotted with anti-GFP or anti-myc Ig (B) (a) Cell

lysates of C2C12 cells were immunoprecipitated with anti-actin Ig

or protein A–agarose as indicated Immunoprecipitates and total

lysates were analyzed by western blotting using actin and

anti-hhLIM Ig; (b) cell lysates were immunoprecipitated with anti-anti-hhLIM

Ig or protein A–agarose and detected using hhLIM and

anti-actin Ig Whole-cell extracts of each group were harvested as a

control to demonstrate proper expression of each protein These

experiments were repeated three times (C) GST pull-down assay.

Purified recombinant GST (lane 1) or GST–hhLIM fusion protein

(lanes 2–4) coupled to glutathione–Sepharose was incubated with

rabbit muscle actin (lane 2) cell extracts from C2C12 cells

transfect-ed with hhLIM expression plasmids (lanes 1 and 3) or transfecttransfect-ed

with pcDNA plasmid (lane 4) After extensive washing, Sepharose

beads were analyzed by SDS ⁄ PAGE and immunoblotted using

anti-actin (upper) or anti-GST (lower) Ig 1, Extracts from C2C12 cells

transfected with hhLIM expression plasmid; 2, rabbit muscle actin

protein; 3, extracts from C2C12 cells transfected with hhLIM

expression plasmid; 4, extracts from C2C12 cells transfected with

pcDNA plasmid.

10 and 30 min after treatment (Fig 3C) As early as

10 min after cytochalasin B application, partial

depolymerization of the actin cytoskeleton occurred in

nontransfected cells, whereas hhLIM-expressing cells

showed an unaffected actin network (data not

shown) After 30 min of treatment, most of the non-transfected cells showed a fully depolymerized actin cytoskeleton By contrast, the morphology of hhLIM– GFP-expressing cells remained normal, indicating that the cytoskeleton was existent and supported the

S

P

S P S P S P S P hhLIM + + – – – – – – Actin + + + + + + + + SM22α – – – – + + – – BSA – – – – – – + +

Actin (8 µ M ) hhLIM Actin (8 µ M ) hhLIM

0 0.25 0.5 1 2 4 8 16 µ M [hhLIM]

BSA Actin hhLIM SM22α

0

2

Actin hhLIM + actin SM22 + actin BSA + actin

*

*

0 0.4 0.8 1.2 1.6

2

hhLIM (µ M )

Actin – + – +

Actin

GST GST GST-hhLIM

A

B

D

C

E

Fig 2 Functional interaction between hhLIM and F-actin (A) Coomassie Brilliant Blue stained SDS ⁄ PAGE gel showing typical actin co-sedi-mentation assay hhLIM, SM22a or BSA were incubated with actin for 30 min in F-actin buffer containing ATP and Ca2+and then centri-fuged at 10 000 g for 30 min Proteins in the pellets (P) and supernatants (S) were analyzed by SDS ⁄ PAGE Densitometry was performed to determine the actin P ⁄ S ratios of three independent experiments to quantify the effect of hhLIM on actin sedimentation *P < 0.05, com-pared with the control (B) Actin at 8 l M alone or in the presence of different concentrations of hhLIM (0.25–16 l M ) was polymerized and centrifuged Proteins in the pellets (P) and supernatants (S) were analyzed by SDS ⁄ PAGE and stained with Coomassie Brilliant Blue (C) Quantitation analysis for GST–hhLIM association with F-actin at different concentrations of GST–hhLIM The F-actin concentration was 8 l M After SDS ⁄ PAGE and staining, gels were scanned and the amount of protein that was present in the pellet and supernatant was quantified The concentration of actin-bound hhLIM was plotted against the concentration of free hhLIM Values are means ± SEM for three indepen-dent experiments (D) Electron microscopy morphology of the filaments assembled from the GST–hhLIM-actin complex Electron microscopy

of negatively stained actin filaments was performed with the following combinations of purified proteins: (a) 8 l M actin and 2 l M GST– hhLIM; (b) 8 l M actin; (c) 8 l M actin and 2 l M SM22a; (d) 8 l M actin and 2 l M BSA Bar = 70 nm (E) In vitro binding analysis using nono-meric (G) actin and GST or GST–hhLIM bound to glutathione agarose beads Western blot of GST pull-down assay fractions using an actin antibody showing similar amounts of actin in samples with GST Sepharose versus GST-tagged hhLIM As expected, no signal was detected

in the absence of G-actin Similar results were obtained in three independent experiments.

appearance of the cell (Fig 3C) Finally, 120 min after

cytochalasin B application, almost all the

hhLIM-expressing cells presented a fully disrupted actin

cyto-skeleton (data not shown) In order to test this further,

C2C12 cells were treated with cytochalasin B and phal-loidin for 30 min, and the distribution of hhLIM and actin in the soluble (sol) and cytoskeleton (csk) frac-tions was determined by western blotting As shown in

Con hhLIM hhLIM (–)

Actin (sol)

Actin (csk)

hhLIM (sol)

GAPDH

(csk)

hhLIM

Cytochalasin B

Pholloidin

Con

pcDNA-hhLIM

Sol csk sol csk sol csk sol csk 0

2 4 6 8 10

pcDNA-hhLIM

hhLIM

(sol)

hhLIM

(csk)

Actin (sol)

Actin (csk)

Con CB Phalloidin

A

B

(h) (g)

(f) (e)

D

Fig 3 hhLIM stabilizes F-actin in C2C12 cells Extracts from C2C12 cells transfected with pcDNA, pcDNA3–hhLIM or hhLIM siRNA expression plasmids were separated into cytosolic soluble (sol) and cytoskeleton-associated proteins (csk) Equal amounts were separated by SDS ⁄ PAGE and proteins

in each fraction were detected by immuno-blotting by using anti-actin or anti-hhLIM Ig (B) C2C12 cells transfected with 0.5, 1, or 1.5 lg of hhLIM expression plasmid were lysed, and cytosolic soluble (sol) and cyto-skeleton-associated proteins (csk) were sep-arated for analysis Left, a representative result from three independent experiments

is shown Right, the density of specific band

of csk ⁄ sol was scanned and quantified (C) hhLIM delayed the effect of cytochala-sin B on C2C12 cells (a–c) C2C12 cells transfected with pEGFP–hhLIM were trea-ted with cytochalasin B for 30 min; (d) C2C12 cells were treated with cytochala-sin B for 30 min; (e–g) C2C12 cells trans-fected with pEGFP–hhLIM were treated with phalloidin for 30 min; (h) C2C12 cells were treated with phalloidin for 30 min; (i–k) C2C12 cells transfected with pEGFP– hhLIM; (l) C2C12 cells transfected with pEGFP (D) C2C12 cells were treated with cytochalasin B or phalloidin for 30 min and lysed by lysis buffer and separated into cytosolic soluble (sol) and cytoskeleton-associated proteins (csk) Equal amount were separated by SDS ⁄ PAGE and proteins

in each fraction were detected by immuno-blotting by using anti-actin or anti-hhLIM Ig.

Fig 3D, cytochalasin B led to the release of hhLIM

from the insoluble fractions This was consistent with

the result of immunofluorescence analysis, indicating

that hhLIM participates in actin polymerization

(Fig 3C) However, pelletable hhLIM from

phalloidin-treated C2C12 cells was increased by 100%

Together, these results indicate that modulation of

the actin cytoskeleton induces changes in hhLIM

localization

LIM domain 2 of hhLIM mediates the interaction

between hhLIM and actin

hhLIM has two LIM domains To identify which

domains or sites of hhLIM interact with actin, a series

of truncated mutants was constructed and a GST

pull-down assay was used This showed that the F4 region

(amino acids 41–194), which contains the LIM

domain 2, has almost the same activity as full-length

hhLIM Interestingly, although the F5 region (amino

acids 41–154), with the C-terminus of the F4 region

deleted, is still able to interact with actin, binding

activity is decreased compared with the F4 region By

contrast, the F3 fragment (amino acids 1–120), with

the C-terminus of hhLIM deleted, is not able to

inter-act with inter-actin The data suggest that hhLIM binding

to actin requires a functional LIM domain 2 (Fig 4A)

To further characterize that LIM domain 2 is sufficient

to interact with actin, the LIM domain was

function-ally destroyed by replacing cysteine with serine in

either domain 1 (mLIM1) or 2 (mLIM2), and an

in vitro GST pull-down assay was used Figure 4B

shows that F-actin was pulled down by full-length

hhLIM and mutant mLIM1, indicating an interaction,

whereas mLIM2 did not pellet with actin To further

identify the LIM domain that mediates the interaction

of hhLIM with actin, we transfected C2C12 cells with

GFP-tagged full-length hhLIM or GFP-tagged LIM

domain-mutated constructs and detected the

distribu-tion of hhLIM The results revealed that mLIM2 is

mainly diffused and fuzzily distributed (Fig 4C) To

characterize further the interaction between hhLIM

mutants and actin, co-sedimentation assays were

per-formed using purified actin and GST–mLIM1or GST–

mLIM2 protein As shown in Fig 4D, full-length

hhLIM and hhLIM mutants co-sedimented with

F-actin, but the amount of sedimented actin is lower

in the presence of mLIM2 than in the presence of

mLIM1 or full-length hhLIM Importantly, mutation

of LIM domain 2 dramatically affected the contraction

of the C2C12 cells compared with cells expressing

hhLIM, which may underlie the dysfunction (Fig 4E)

Taking these factors together, we determined that

tar-geted disruption of the second LIM domain of hhLIM destroys the interaction between hhLIM and the con-tractive ability of C2C12 cells, indicating the important role that LIM domain 2 plays in controlling assembly and organization of the actin cytoskeleton

Discussion The plasticity of the actin cytoskeleton relies mainly

on the ability of actin filaments to form, branch, bun-dle, and disassemble within short timeframes in response to many signals LIM proteins play a critical role in the organization of the actin cytoskeleton WLIM1 was found both to associate with the actin cytoskeleton in a very dynamic manner and to circu-late rapidly throughout the cytoplasm, making it avail-able wherever and whenever it was needed for new actin bundle formation [1,14] WLIM1 protein con-tains two LIM domains, deletion of one of the domains reduced significantly, but did not entirely abolish, the ability of WLIM1 to bind actin filaments Variants lacking the C-terminal or inter-LIM domain were only weakly affected in their F-actin stabilizing and bundling activities, and trigger the formation of thick cables containing tightly packed actin filaments

as does the native protein By contrast, deletion of one

of the two LIM domains negatively impacted both activities and resulted in the formation of thinner and wavier cables [13] Zyxin-related protein 1, which belongs to a family of LIM-containing proteins that includes zyxin and lipoma-preferred partner, partici-pates in the organization of the actin cytoskeleton [15] FHL2 was observed, along with F-actin, to be involved in the focal adhesion of C2C12 and H9C2 myotubes [16] Overexpression of FHL2 promotes differentiation by binding to b-catenin [17] FHL3 reg-ulates a-actinin-mediated actin bundling as an actin-binding protein [18] CRP3 (also called muscle LIM protein–MLP) plays an important role in myogenesis and in the promotion of myogenic differentiation This function has been related to its myofibrillar location in close vicinity to the Z disk and its interaction with a-actinin MLP is highly expressed during differentia-tion in all types of striated muscle, but its expression

in the adult is restricted to cardiac and slow-twitch fibers of skeletal muscle [8,19] Moreover, it has been reported that targeted deletion of MLP in mice causes marked disruption of the myocardial cytoarchitecture, leading to dilated cardiomyopathy and death resulting from cardiac failure [10,20,21] Despite the dramatic consequences associated with loss of MLP expression, the mechanistic details of CRP function in muscles remain speculative The data presented here identify a

member of the CRP family, hhLIM, as a new

F-actin-binding protein whose targeting of actin filaments

stabilizes the actin cytoskeleton and promotes actin

bundle⁄ cable formation We conclude that hhLIMs are

real F-actin-binding protein on the following

observa-tions: (a) hhLIM colocalized with F-actin, (b) hhLIM

showed F-actin-binding activity, and (c) hhLIM

co-sedimented with F-actin The interaction between

hhLIM and actin filaments was previously believed to

be indirect, requiring intermediary proteins such as

a-actinin or zyxin However, it is clearly established

that hhLIM and other members of the CRP family

are autonomous F-actin-binding proteins Our in vitro

investigations provide, for the first time, strong

aug-ments supporting the idea that the LIM domain

parti-cipates in the F-actin binding and bundling activities

displayed by hhLIM

Confocal analyses showed that hhLIM accumulates

in both the nucleus and the cytoplasm, where it

pre-dominantly associates with the actin cytoskeleton [7]

This dual location is in agreement with that reported

previously for members of the CRP family and other

CRP-related proteins, such as MLP [22] Although

CRPs were first believed to interact indirectly with the

actin cytoskeleton via intermediary proteins, such as

zyxin and a-actinin, recent studies have shown that

CRP1 and CRP2 are autonomous actin-binding

pro-teins [11,23] Our in vitro results extend this property

to the hhLIM protein, suggesting that all CRPs and

CRP-related proteins have the ability to associate with

F-actin Here, we demonstrate the ability of a new LIM

protein to interact with F-actin in a direct manner

Formation of higher order actin structures, such as

bundles and cables, is crucial to stabilize the

organiza-tion of transvacuolar strands and maintain overall

cellular architecture As mentioned above, CRP1 may participate in the formation and⁄ or maintenance of long actin cables [12] Consistent with this hypothesis,

we observed that ectopic expression of hhLIM in C2C12 cells stabilizes actin filaments⁄ bundles against cytochalasin B In addition, overexpression of hhLIM

in C2C12 cells induces an increase in the overall amounts of actin and F-actin This prompted us to investigate whether hhLIM stabilizes and bundles actin filaments directly In vitro cytochalasin B experiments demonstrated that hhLIM stabilizes F-actin by itself

In addition, co-sedimentation assays and the direct observation of in vitro actin filaments that have been polymerized in the presence of hhLIM demonstrated that hhLIM bundles actin filaments in an autonomous manner

hhLIM consists of two LIM domains Targeted dis-ruption of the second LIM domain of hhLIM abol-ished F-actin-binding activity, indicating the important role that LIM domain 2 plays in the control of assem-bly and organization of the actin cytoskeleton

In conclusion, in vitro results show that hhLIM inter-acts with filamentous actin in a direct manner hhLIM enhances the stability of the actin cytoskeleton and pro-motes actin bundling Although the exact contribution made by hhLIM protein to actin cytoskeleton dynam-ics⁄ remodeling remains to be explored, the data pro-vide strong evidence that hhLIM is an actin cytoskeleton organizer An open question is the signifi-cance of hhLIM in the nucleus Several LIM proteins have been shown to shuttle between the cytoplasm and the nucleus and it has been suggested that they mediate communication between both compartments Similar functions for hhLIM proteins cannot be excluded Con-sistent with a nuclear role for hhLIM, it has been

Fig 4 Relationship between the structure and the activation activity of hhLIM (A) Requirement of the C-terminal half of hhLIM for association activity with actin hhLIM and its various derivatives were constructed into PGEX-3X plasmids GST–hhLIM and its derivative proteins are sche-matically depicted on the left Association activities of hhLIM and its derivatives are represented on the right Extracts from C2C12 cells were precleared with GST–Sepharose beads and then incubated with GST–hhLIM Sepharose beads or its derivative proteins Pellets were washed, and interacting proteins were separated by SDS ⁄ PAGE and identified by western blotting (B) Mutation of LIM domain 2 of hhLIM disrupts the association with actin Extracts from C2C12 cells were precleared with GST–Sepharose beads and then incubated with GST–hhLIM Sepha-rose beads, or LIM domain-mutated (mLIM1, GST-mLIM110Cys fi Ser, 13Cys fi Ser, mLIM2, GST-mLIM2120Cys fi Ser, 123Cys fi Ser) Sepharose beads or GST–Sepharose beads Pellets were washed, and interacting proteins were separated by SDS ⁄ PAGE and identified

by western blotting (C) Fluorescence analysis of hhLIM in the C2C12 cells C2C12 cells were transfected with pEGFP–hhLIM, pEGFP– mLIM1(10Cys fi Ser, 13Cys fi Ser), pEGFP-mLIM2(120Cys fi Ser, 123Cys fi Ser) or pEGFP The cells were fixed and examined with an IX71 fluorescence microscope (Olympus) (D) Actin co-sedimentation assay verified the functional interaction between hhLIM and F-actin Purified F-actin was incubated with GST–hhLIM or LIM domain-mutated hhLIM Cross-linked F-actin was pelleted by centrifugation, separated

by SDS ⁄ PAGE, and stained with Coomassie Brilliant Blue (E) Densitometry micrograph was obtained of the agonist-induced contraction of C2C12 cells C2C12 cells were transfected with pEGFP (control), pEGFP–hhLIM, pEGFP–mLIM1(10Cys fi Ser, 13Cys fi Ser) or pEGFP– mLIM2(120Cys fi Ser, 123Cys fi Ser) and maintained in physiological rodent saline (138 m M NaCl, 2.7 m M KCl, 1.8 m M CaCl 2 , 1.06 m M MgCl2, 12.4 m M HEPES, and 5.6 m M glucose, pH 7.3) in a chamber ( 2 mL) mounted on the stage of an inverted microscope The C2C12 cell length was modified by acetylcholine stimulation (100 l M ) *P < 0.05, compared with C2C12 cells transfected with pcDNA3–hhLIM plasmid.

reported to activate brain natriuretic factor (BNP) and

atrial natriuretic factor (ANF) gene expression [7,24]

Identification of further regulatory mechanisms that

trigger the translocation of hhLIM between the

cyto-plasm and the nucleus is an important goal for the future Perhaps the most fruitful area of future research

in LIM biology will involve dissecting the precise roles

of LIM proteins in both the nuclear and cytoplasmic

Actin GST F1 F2 F3 F4 F5 Neg

F1 F2 F3 F4 F5 LIM zinc-binding domain

Neg mLIM1 LIM mLIM2

Actin

GST

Zinc finger C2H2 type domain LIM zinc-binding domain Protein kinase C phosphorylation site

P

Actin

*

0

1

2

3

con hhLIM mLIM1 mLIM2

A

B

C

D

E

compartments, and deciphering how the role of a LIM

protein that is associated with actin filaments might be

integrated with nuclear functions and vice versa

Experimental procedures

Cell culture and transfection

The C2C12 mouse myoblast line was maintained with

Dulbecco’s modified Eagle’s medium with 10% fetal bovine

serum Differentiation was induced in C2C12 cells by

replacing medium with Dulbecco’s modified Eagle’s

med-ium containing 2% horse serum hhLIM expression plasmid

was gift from KH Chen (National Institute on Aging,

Balti-more, MD, USA) A hhLIM siRNA-expressing plasmid

was constructed using BLOCK-iT U6 RNAi Vector

by subcloning double-stranded oligonucleotides

comple-mentary (5¢-CACCGCAGTGCCATGGAAGGAGTTTC

CACACGAATGTGGAAACTCCTTCCATGGCACTG-3¢)

according to the manufacture’s protocol (Invitrogen,

Carls-bad, CA, USA) Transfections with various DNA

con-structs were performed with lipofectamine 2000 (Invitrogen)

according to the manufacturer’s instructions

Immunoprecipitation and western blotting

C2C12 cells grown in Dulbecco’s modified Eagle’s medium,

supplemented with 10% fetal bovine serum were transfected

with cDNA constructs using Lipofectamine according to

the manufacturer’s protocol Forty-eight hours later, cells

were lysed in lysis buffer [20 mm Tris, pH 7.5, 150 mm

NaCl, 1 mm EGTA, 1 mm EDTA, 1% Triton X-100,

pro-tease inhibitor mixture (Sigma, St Louis, MO, USA), and

1 lm Na3VO4) Lysates were sonicated on ice, and cell

deb-ris was removed by centrifugation Lysates were precleared

with protein A⁄ G–agarose beads (Santa Cruz

Biotechnolo-gies, Santa Cruz, CA, USA), and the proteins were

immu-noprecipitated with the appropriate antibody overnight at

4C followed by incubation with protein A ⁄ G–agarose for

1 h at 4C Immunoprecipitates were washed three times

with lysis buffer, and proteins were separated on

SDS⁄ PAGE Immunoblotting analysis was performed as

described previously [25–28] Primary antibodies used for

the assays were anti-GST polyclonal Ig (1 : 500; Santa

Cruz), anti-hhLIM polyclonal Ig (gift of KH Cheng,

National Institute on Aging, Baltimore, MD), anti-GFP

polyclonal Ig (1 : 500; Santa Cruz), and anti-(skeletal

a-actin) polyclonal Ig (1 : 500; Santa Cruz)

Site-directed mutagenesis of the LIM domain

of hhLIM

Site-directed mutation of each LIM domain was carried out

by PCR using oligonucleotide primers that coded for the

appropriate point substitutions of amino acids The reac-tions were carried out using a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA) Each mutation was verified by DNA sequence analysis PCR primers used in the site-directed mutagenesis of the LIM domain of hhLIM introduced two point mutations into each LIM domain: LIM1(10Cysfi Ser, 13Cys fi Ser):5¢-GGA GGCGCAAAATCTGGAGCCTCTGAAAAGACCGTCTA C-3¢; LIM2(120Cys fi Ser, 123Cys fi Ser): 5¢-GAGAGTCC GAGAAGTCCCCTCGATCTGGCAAGTCAGTCTATG-3¢

Actin fractionation

Cells were scraped, washed with NaCl⁄ Pi, and lysed in buf-fer A (20 mm Tris⁄ HCl, pH 7.5, 1% Triton X-100, 5 mm EGTA, 1 mm phenylmethylsulfonyl fluoride) on ice for

30 min, and then centrifuged at 12 000 g and 4C for

30 min The supernatants (sol) were harvested The pellets (csk) were lysed in buffer B (10 mm Tris⁄ HCl, pH 7.5,

150 mm NaCl, 1% Triton X-100, 0.1% SDS, 1 mm sodium deoxycholate, 2 mm EGTA, 1 mm phenylmethylsulfonyl fluoride) on ice for 30 min, and then centrifuged at

12 000 g for 30 min The supernatants from the lysed pel-lets (csk) were harvested Protein concentration was deter-mined by a modified Lowry protein assay Equal amounts

of the supernatant (sol) and pellet (csk) were separated by 10% SDS⁄ PAGE and stained with an antibody against hhLIM or actin, with visualization by secondary antibodies and enhanced chemiluminescence [29,30]

Fluorescence staining

Fluorescence staining was performed as described previ-ously [12,31] The cells were stained for 20 min with TRITC⁄ phalloidin (1 lgÆmL)1) in blocking solution (1% BSA and 0.1% Triton X-100 in NaCl⁄ Pi) in the dark at room temperature to localize F-actin

GST pull-down assay

In order to produce GST fusion proteins, full-length and domain-specific regions of hhLIM were generated in a pGEX-3X vector inframe with the N-terminal GST tag All new constructs were confirmed by restriction digestion fol-lowed by sequencing Protein expression was induced by reaction with 0.2 mm isopropyl thio-b-d-galactoside at

30C for 3 h Bacterial lysates were purified over glutathi-one–agarose For the pull-down assay, cell lysate was pre-pared by lysing the C2C12 cells transiently transfected with myc-tagged different mutant or site-directed mutagenesis hhLIM that had been precleared with GST Sepharose beads Assay mixtures were then incubated with GST Sepharose beads or with hhLIM⁄ GST Sepharose beads After centrifugation, the pellets were washed, and the

interacting proteins were separated by SDS⁄ PAGE and

identified by western blot with an anti-actin Ig [32]

Assay for low-speed co-sedimentation of hhLIM

with F-actin

G-Actin (Sigma) was polymerized by incubation at room

temperature for 30 min in a polymerization buffer (20 mm

imidazole⁄ Cl, pH 7.0, 2 mm MgCl2, 1 mm ATP, 0.5 mm

dithiothreitol, 90 mm KCl) The lysates of the

hhLIM-expressing cells were centrifuged at 10 000 g for 30 min,

and the supernatant was used for the assay (F-actin) The

supernatant of the lysates was incubated at room

tempera-ture for 30 min with 0.3 mgÆmL)1F-actin in a solution

con-taining 25 mm imidazole⁄ Cl, pH 7.0, 2 mm MgCl2, 1 mm

ATP, 0.5 mm dithiothreitol, 27 mm KCl and 100 mm NaCl,

and the mixture (50 lL) was placed over a 50 lL cushion

of 30% sucrose in the polymerization buffer After the

sam-ple was centrifuged at 10 000 g for 20 min, the supernatant

and the pellet were subjected to SDS⁄ PAGE, followed by

western blot analysis using the anti-hhLIM and anti-actin

Ig [33–35]

Electron microscopy

Actin (8 lm) was polymerized at room temperature The

actin mixtures were then diluted 1⁄ 8 with Mg-ATP buffer

in the presence of purified GST–hhLIM (2 lm) alone or

with BSA in a final reaction volume of 25 lL These

mix-tures were incubated for 1 h at room temperature The

pro-tein mixtures were adsorbed onto carbon-coated 400-mesh

grids for 1 min Actin filaments were negatively stained

with 2% phosphotungstic acid, pH 7.4, for 15 s Grids were

visualized using transmission electron microscopy (Hitachi

Ltd., Saitama, Japan) at an accelerating voltage of 80 kV

and a nominal magnification of·100 000 [18]

Measurement of contraction

C2C12 cells were transfected with pcDNA3 (control),

pcDNA3–hhLIM, pcDNA3–mLIM1(10Cysfi Ser, 13Cys fi

Ser) or pcDNA3–mLIM2(120Cysfi Ser, 123Cys fi Ser)

and maintained in physiological rodent saline (138 mm

NaCl, 2.7 mm KCl, 1.8 mm CaCl2, 1.06 mm MgCl2,

12.4 mm Hepes, and 5.6 mm glucose, pH 7.3) in a chamber

( 2 mL) mounted on the stage of an inverted microscope

(Olympus, Tokyo, Japan) The C2C12 cell length was

modified by acetylcholine stimulation (100 lm) for 1 min

[25,36,37]

Statistical analysis

To control for day-to-day variations in staining intensity,

untreated cells were always compared with treated cells on

the same microscope slide because cells on the same slide undergo identical culture, fixation, permeabilization, stain-ing and microscopy conditions, allowstain-ing meanstain-ingful com-parisons between samples All data are presented as means ± SE

Acknowledgements

We thank Dr Da-zhi Wang (University of North Caro-lina) for helpful discussions and comments on the manuscript This work was supported by the Program for New Century Excellent Talents in University (No NCET-05-0261), a Key Project of the Chinese Ministry of Education (No.206016), the National Nat-ural Science Foundation of the People’s Republic of China (No.30300132, 30570661) and the Major State Basic Research Development Program of China (No 2005CCA03100)

References

1 Thomas C, Hoffmann C, Dieterle M, Van Troys M, Ampe C & Steinmetz A (2006) Tobacco WLIM1 is a novel F-actin binding protein involved in actin cytoskel-eton remodeling Plant Cell 18, 194–2206

2 Kadrmas JL & Beckerle MC (2004) The LIM domain: from the cytoskeleton to the nucleus Nat Rev Mol Cell Biol 5, 920–931

3 Zhi S, Yao A, Zubair I, Sugishita K, Ritter M, Li F, Hunter JJ, Chien KR & Barry WH (2001) Effects of deletion of muscle LIM protein on myocyte function

Am J Physiol Heart Circ Physiol 280, H2665–H2673

4 Zheng B, Wen JK & Han M (2003) Factors involved in the cardiac hypertrophy Biochemistry Mosc 68, 650–657

5 Bach I (2000) The LIM domain: regulation by associa-tion Mech Dev 91, 5–17

6 Chen H, Zhou Z, Zhang JF & Zhou AR (2000) Screen heart-specific growth genes using three-element PCR-select cDNA subtraction Chin J Biochem Mol Biol 16, 295–300

7 Zheng B, Wen JK, Han M & Zhou AR (2004) hhLIM protein is involved in cardiac hypertrophy Biochim Biophys Acta 1690, 1–10

8 Arber S, Halder G & Caroni P (1994) Muscle LIM pro-tein, a novel essential regulator of myogenesis, promotes myogenic differentiation Cell 79, 221–231

9 Bourgouin C, Lundgren SE & Thomas IB (1992) Apte-rousis a Drosophila LIM domain gene required for the development of a subset of embryonic muscles Neuron

9, 549–561

10 Arber S & Caroni P (1996) Specificity of single LIM motifs in targeting and LIM⁄ LIM interactions in situ Gene Dev 10, 289–300