Biochemical Journal

Research article

Myosin regulatory light chains are required to maintain the stability of myosin II and cellular integrity

Inju Park , Cecil Han , Sora Jin , Boyeon Lee , Heejin Choi , Jun Tae Kwon , Dongwook Kim , Jihye Kim , Ekaterina Lifirsu , Woo Jin Park , Zee Yong Park , Do Han Kim , Chunghee Cho


Myosin II is an actin-binding protein composed of MHC (myosin heavy chain) IIs, RLCs (regulatory light chains) and ELCs (essential light chains). Myosin II expressed in non-muscle tissues plays a central role in cell adhesion, migration and division. The regulation of myosin II activity is known to involve the phosphorylation of RLCs, which increases the Mg2+-ATPase activity of MHC IIs. However, less is known about the details of RLC–MHC II interaction or the loss-of-function phenotypes of non-muscle RLCs in mammalian cells. In the present paper, we investigate three highly conserved non-muscle RLCs of the mouse: MYL (myosin light chain) 12A (referred to as MYL12A), MYL12B and MYL9 (MYL12A/12B/9). Proteomic analysis showed that all three are associated with the MHCs MYH9 (NMHC IIA) and MYH10 (NMHC IIB), as well as the ELC MYL6, in NIH 3T3 fibroblasts. We found that knockdown of MYL12A/12B in NIH 3T3 cells results in striking changes in cell morphology and dynamics. Remarkably, the levels of MYH9, MYH10 and MYL6 were reduced significantly in knockdown fibroblasts. Comprehensive interaction analysis disclosed that MYL12A, MYL12B and MYL9 can all interact with a variety of MHC IIs in diverse cell and tissue types, but do so optimally with non-muscle types of MHC II. Taken together, our study provides direct evidence that normal levels of non-muscle RLCs are essential for maintaining the integrity of myosin II, and indicates that the RLCs are critical for cell structure and dynamics.

  • myosin essential light chain
  • myosin heavy chain
  • myosin regulatory light chain
  • non-muscle myosin
  • short interfering RNA (siRNA)-mediated knockdown


Myosins are a large family of contractile proteins. Myosin II, the conventional myosin originally identified in muscle, is also found in non-muscle cells. Myosin II is composed of two MHC (myosin heavy chain) IIs, two RLCs (regulatory light chains), and two ELCs (essential light chains). Regulation of cardiac and skeletal myosin IIs differs from that of NM II (non-muscle myosin II). In non-muscle cells, the action of myosin II is controlled by RLCs. In particular, RLCs undergo phosphorylation at the Ser19 and Thr18 sites, which increases the Mg2+-ATPase activity of MHC in the presence of actin. A number of kinases, including MLCK (myosin light chain kinase) and ROCK (Rho-associated kinase), are known to phosphorylate RLCs [1]. Active myosin II proteins form actomyosin complexes, which play roles in cell structure, contractility, adhesion and migration [1,2]. Genes encoding RLCs in cardiac and skeletal muscles have been knocked out in mice and the resulting mutant animals displayed severe abnormalities, with lethality evident at the embryonic or neonatal stage [3,4]. No functional study of non-muscle RLCs using gene knockout or knockdown has been reported yet in mice.

Previously, we identified a novel gene from the mouse cardiac muscle UniGene library ( [5]. During the course of our earlier study, the gene was named myosin light chain regulatory B-like. Our in silico analysis revealed that the gene was an orthologue of human myosin light chain 12A (MYL12A) [6]. Thus we denote the mouse gene as Myl12a in the present study. By further in silico analysis using NCBI (National Center for Biotechnology Information) databases, we found two other genes highly homologous with Myl12a. These are termed myosin light chain 12B regulatory (Myl12b) and myosin light polypeptide 9 regulatory (Myl9) [7]. The mammalian orthologues of Myl12a and Myl12b have been considered to be non-muscle RLC genes [6].

Although information on the expression and interaction properties of the mammalian orthologues of the RLC genes has been reported [611], experimental data on the expression patterns and loss-of-function phenotypes at the cellular level in mice are lacking. In the present paper, we report the first investigation of RLC functions in mice, providing detailed and significant information on transcription, RLC-interacting proteins in different tissues and the cellular functions of RLCs. In particular, we present direct evidence that, at the cellular level, RLCs encoded by Myl12a and Myl12b are essential for the stability of MHC IIs and ELCs, including complexes involving MYH9 (NMHC IIA), MYH10 (NMHC IIB) and MYL6, in fibroblasts studied using siRNA (short interfering RNA)-mediated knockdown.


Sequence analysis of mouse RLCs

All RLC sequences were obtained from the GenBank® database dbEST ( and the UCSC Genome Browser ( The GenBank® accession codes of Myl12A, Myl12B and Myl9 are NM_026064, NM_023402 and NM_172118 respectively. All three genes have highly conserved coding sequences (>90%), but unique UTRs (untranslated regions). Multiple sequence alignment of RLC proteins was performed using ClustalX [11b].

RT-PCR (reverse transcription PCR) and Northern blot analysis

Total RNA was isolated from various tissues of adult ICR male mice. After immediate freezing in liquid nitrogen, the tissues were homogenized and resuspended in the TRI Reagent (Molecular Research Center). For RT-PCR, total RNA was reverse transcribed into cDNAs using Omniscript Reverse Transcriptase (Qiagen), and then normalized with respect to the level of Gpd1 (glycerol-3-phosphate dehydrogenase) transcription. To perform the Northern blot analysis, 10 μg of total RNA was incubated at 65 °C for 5 min and then separated on 1.2% (w/v) agarose gels containing 1.8% (v/v) formaldehyde. Separated RNA was transferred on to Hybond-XL membranes (Amersham) by a capillary method. The blots were hybridized with 2.5 ng/ml probe solutions derived from gene-specific PCR products of the 3′-UTR at 68 °C for 2 h. Probe DNAs were labelled with [32P]dCTP (PerkinElmer) using a Prime-It random priming kit (Stratagene). The blots were rinsed twice with 2× SSC/0.05% SDS at RT (room temperature) for 10 min and twice with 0.1× SSC/0.1% SDS at 68 °C for 10 min. The blots were developed with Hyperfilm (Amersham), using intensifying screens, at −80 °C.

Our work conformed to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health.

Calcium overlay assay

GST (glutathione transferase)-tagged recombinant full-length Myl12A protein was transferred on to nitrocellulose membranes. After washing three times in buffer (60 mM KCl, 5 mM MgCl2, 10 mM imidazole/HCl, pH 6.8) for 30 min, membranes were probed with 1 μM [45Ca]Cl2 for up to 1 h. Next the membranes were washed in 67% (v/v) ethanol, dried and subjected to autoradiography for 3 days at −80 °C. The calcium overlay technique has been described previously [12].

Cell culture and gene silencing using RNA interference

Cells of the mouse fibroblast cell line, NIH 3T3 (American Type Culture Collection), were cultured in DMEM (Dulbecco's modified Eagle's medium; Invitrogen) supplemented with 10% (v/v) fetal bovine serum (GIBCO). The growth medium for the mouse cardiac muscle cell line HL-1 was the Claycomb medium (Sigma) with 10% (v/v) fetal bovine serum. For immunocytochemistry, cells on coverslips were fixed for 10 min in 4% (v/v) paraformaldehyde in PBS at RT. Fixed cells were permeabilized by incubation with 0.1% Triton X-100 for 10 min. After blocking with 3% (w/v) BSA for 1 h, primary antibody (2 μg/ml) was applied directly to a coverslip and incubated for 1 h at RT. Rhodamine-conjugated secondary antibody was added for detection of primary antibody and then incubation continued for 40 min. For F-actin staining, Alexa Fluor® 488 phalloidin (Invitrogen) was employed at 1:400 dilution. Imaging was performed using a Leica DMLB microscope (Bensheim) and images were captured by a digital CCD (charge-coupled device) camera (Photometrics CoolSNAP). Image data were collected using QED software (QED Imaging). siRNA-mediated knockdown of non-muscle RLC genes was achieved using a mixture of siRNA duplexes (Supplementary Table S1 at, delivered using the DharmaFECT3 transfection reagent (Dharmacon), according to the manufacturer's protocol. The siRNA concentration used was 50 nM. The effects of knockdown were analysed in cells grown in DMEM for 12–96 h. After transfection, the growth medium was changed every 24 h. Reductions in mRNA levels were confirmed by RT-PCR using NIH 3T3 cDNAs prepared after 45 h of knockdown. Morphological changes in cells subjected to knockdown were measured using MetaMorph software (Universal Imaging Corp.). For the contraction assay, a fibrin gel was formed by reaction of 4 μg/μl fibrinogen in 0.9% saline with 20 units of thrombin for 30 min at 37 °C. After 70 h of knockdown, the gel was detached and stained using the Ponceau S reagent. Cell migration velocity was measured from time-lapse imaging data with photographs taken at 10 min intervals over 12 h by an Olympus IX81-ZDS camera.

Antibodies and reagents

To generate an antibodies against MYL12A, MYL12B and MYL9, a synthetic peptide (the C-terminus of MYL12A, FTRILKHGAKDKDD) was used to immunize rabbits. After three immunizations, polyclonal antibodies were purified from the serum using GST–MYL12A (full-length) protein immobilized on a column using an AminoLink Immobilization kit (Pierce). This antigenic peptide is common to all non-muscle RLCs, including Myl12A, Myl12B and Myl9 of the mouse.

Thus the antibody raised against it specifically detects all non-muscle RLCs, but not any other RLCs or ELCs. The following pAbs (polyclonal antibodies) and mAbs (monoclonals antibodies) were used to detect proteins, anti-MYH9 and anti-MYH10 pAbs (Cell Signaling Technology); anti-α-tubulin and anti-β-actin mAbs (Sigma–Aldrich); anti-MYL6 and antiMYH14 (NMHC IIC) pAbs, and anti-MYH6/7 and anti-MYH1/2 mAbs (Abcam); and anti-GAPDH (glyceraldehyde-3-phosphate dehydrogenase) pAb (Lab Frontier). Horseradish peroxidase-conjugated secondary antibodies were obtained from Jackson ImmunoResearch Laboratories. S(−)-blebbistatin and Y-27632 were purchased from Calbiochem. Each drug was added to complete medium at several concentrations ranging from 0 to approx. 50 μM prior to cell growth for 4 to approx. 20 h, followed by urea/glycerol PAGE of cell lysates according to a previous procedure [13,14].

Immunoprecipitation and LC (liquid chromatography)–MS/MS (tandem MS)

Either cells or tissues were lysed in a buffer containing the nonionic detergent Nonidet P40 (1%, v/v), 10 mM MgATP and protease inhibitors (Halt™ Protease Inhibitor Cocktail, Thermo) as described previously [15]. Cell extracts (200 to approx. 500 μg) were incubated with 0.5 to approx. 1 μg of anti-MYL12A, anti-MYL12B or anti-MYL9 antibody and Protein A–Sepharose (Amersham Biosciences) for 4 h at 4 °C. Protein A beads were washed in lysis buffer three times at 4 °C. Protein elution from antibody beads was performed using 8 M urea in 100 mM Tris/HCl (pH 8) buffer solution. Eluted proteins were reduced with TCEP [tris-(2-carboxyethyl)phosphine] (final concentration of 5 mM) at RT for 30 min, and then alkylated with iodoacetamide (final concentration of 25 mM) in the dark for 30 min at RT. The sample was diluted with 2 M urea using 100 mM Tris/HCl (pH 8) and buffer solution. CaCl2 was added to a final concentration of 5 mM, and the proteins were digested in digestion buffer [100 mM Tris/HCl (pH 8) and 5 mM CaCl2] using sequencing grade modified trypsin (Promega) for 14 h at 37 °C with an enzyme/substrate ratio of 1:25. Trypsin digestion of eluted proteins was quenched by adding 90% formic acid (final concentration of 5%). Digested proteins were loaded on to fused silica capillary columns (100 μm inner diameter and 360 μm outer diameter) containing 7.5 cm of 5 μm particle Aqua C18 reversed-phase column material (Phenomenex). The column was placed inline with an Agilent HP1100 quaternary LC pump and a splitter system was used to achieve a flow rate of 250 nl/min. Buffer A (5% acetonitrile and 0.1% formic acid) and buffer B (80% acetonitrile and 0.1% formic acid) were used to create a gradient over 80 min. The gradient profile started with 3 min of 100% buffer A, followed by 5 min from 0% to 15% buffer B, 57 min from 15% to 55% buffer B, and 15 min from 55% to 100% buffer B. Eluted peptides were directly electrosprayed into an LTQ Ion Trap mass spectrometer (ThermoFinnigan) by applying 2.3 kV of DC voltage. A data-dependent scan consisting of one full MS (400–2000 m/z) and ten data dependent MS/MS scans was used to generate MS/MS spectra of eluted peptides. A normalized collision energy of 35% was used throughout the data acquisition period. MS/MS spectra were compared with the Mouse IPI protein database (version 3.31 [15b]) using SEQUES (Thermo Scientific) and Linux Cluster Systems (14 nodes). DTASelect [15c] was used to filter the search results, and the following Xcorr and delta Cn values applied to different charge states of peptides with a fully tryptic digested end requirement of 1.8 for singly charged peptides, 2.2 for doubly charged peptides and 3.2 for triply charged peptides, with a delta Cn value of 0.08 for all charge states. Manual assignments of fragment ions in all filtered MS/MS spectra were followed to confirm the protein database search results.


Myl12a, Myl12b and Myl9 are highly homologous RLC genes

We examined the sequences and genomic structures of the murine Myl12a, Myl12b and Myl9 genes using the NCBI databases. Myl12a and Myl12b encode RLC proteins with a 99% amino acid sequence identity. The genes are very similar in exon–intron structure and lie adjacent in the proximal region of mouse chromosome 17, suggesting that the two genes may have arisen by gene duplication. MYL9 has an amino acid sequence very similar to those of MYL12A (96%) and MYL12B (95%), but is encoded on a different chromosome (Figures 1A and 1B). All of these genes have orthologues in a variety of mammalian species [6,7,11]. Comparison of Myl12a, Myl12b and Myl9 with genes encoding other mouse RLCs revealed relatively low levels (43–55%) of amino acid sequence similarity (Figure 1A). In general, RLCs are activated by both Ca2+-activated kinases and Ca2+ binding to EF hand motifs [16,17]. The amino acid sequences of MYL12A, MYL12B and MYL9 contain conserved phosphorylation sites (Ser1, Ser2, Thr9, Thr18 and Ser19) and EF hand motifs (Figure 1B). To test the Ca2+-binding activities of proteins encoded by these genes, we performed a [45Ca2+] overlay assay using the GST–MYL12A protein. As shown in Figure 1(C), the recombinant MYL12A protein was found to bind Ca2+, suggesting that the protein has a Ca2+-dependent function.

Figure 1 Genes encoding myosin RLCs in mice

(A) Mouse RLC genes, genomic expression data and functional information on the genes are shown. Myl12a is a new gene symbol assigned in the present study, on the basis of an orthologous relationship of a mouse gene with human MYL12A. The homology level of each gene reflects the percentage of amino acid sequence similarity relative to Myl12a. Cell and tissue expression patterns are based on data from both the present and previous studies. We also show whether loss-of-function information, obtained by gene knockout or knockdown, is available (+) or not (−). C, cardiac muscle; N, non-muscle; PL, precursor lymphocyte-specific; Sk, skeletal muscle; Sm, smooth muscle. (B) Alignment of amino acid sequences of the proteins encoded by the mouse RLC genes. The conserved regions in all RLC proteins are shaded using black boxes. Grey regions represent sequences conserved only in proteins encoded by Myl12a, Myl12b and Myl9. Stars indicate the two phosphorylation sites, Thr18 and Ser19, conserved in all RLCs. EF hand domains are shown by lines. The double line indicates an antigenic region used for generation of the pAb. (C) Calcium overlay assay. Autoradiography to assess [45Ca]Cl2 binding was performed using a GST-tagged recombinant of full-length MYL12A.

Expression patterns of Myl12a, Myl12b and Myl9

To determine the expression pattern of the Myl12a, Myl12b and Myl9 genes, we performed RT-PCR analysis using specific primers and Northern blot analysis employing probes specific for 3′-UTRs unique to the genes (Supplementary Table S2 at In general, the three genes were widely transcribed in diverse tissues, but their expression patterns differed (Figures 2A and 2B). Myl12a and Myl12b were expressed abundantly in most tissues except the brain (Myl12a) and striated muscles (Myl12b) (Figure 2A and 2B). Notably, the expression of Myl9 was highest in smooth muscle such as that of the bladder (Figures 2A and 2B). Thus our results suggest that the genes are differentially regulated in a transcriptional sense. It should be noted that the genes are considered to encode non-muscle RLCs, based on the expression patterns shown in the present results and previous reports in other mammalian species [6,9].

Figure 2 Expression of MYL12a, MYL12b and MYL9 in different tissues

(A) Tissue distribution was explored by RT-PCR analysis using specific primers. Measurement of Gpd1 transcript levels were used to normalize cDNA templates. Bd, bladder; Br, brain; H, heart; K, kidney; Li, liver; Lu, lung; Sk, skeletal muscle; Sp, spleen; T, testis. (B) Northern blot analysis using specific probes corresponding to regions in the 3′-UTRs. Ribosomal RNAs and β-actin mRNA were used as controls. (C) Immunoblot analysis employing an antibody specific for MYL12A, MYL12B and MYL9. α-tubulin and GPD1 were used as loading controls. (D) Staining of NIH 3T3 fibroblasts for MYL12A, MYL12B and MYL9 (rhodamine) and actin stress fibre. Nuclei were stained with Hoechst 33342 (scale bar, 20 μm). (E) Staining of mouse cardiomyocytes for MYL12A, MYL12B and MYL9 (rhodamine, a and e), actin stress fibre protein (b) and α-actinin (Alexa Fluor® 488, f). Nuclei were stained by Hoechst 33342 (c and g). Merged images are shown in d and h. The inset in (h) is magnified in (i) (Scale bar, 20 μm).

We generated polyclonal antibodies that detected the proteins encoded by all three genes, using a peptide common to the C-terminal regions of the three proteins. We denote the antibody that recognizes the three proteins as ‘anti-MYL12A/12B/9’. To determine the specificity of the antibody, we performed immunoblot analysis using various recombinant RLC proteins. We found an absence of cross-reactivity of the anti-MYL12A/12B/9 antibody to the cardiac and skeletal muscle RLCs, suggesting the antibody specificity (Supplementary Figure S1 at Immunobloting analysis of the different tissues showed that MYL12A, MYL12B and MYL9 were widely expressed, with a molecular mass of 20 kDa, comparable with the pattern shown by transcript analysis (Figure 2C). Staining of NIH 3T3 mouse fibroblasts with the anti-MYL12A/12B/9 antibody and phalloidin revealed co-localization of MYL12A, MYL12B and MYL9 with actin stress fibres (Figure 2D). We also stained mouse cardiomyocytes with the antibody and found that MYL12A, MYL12B and MYL9 are present in the Z-lines (Figure 2E) where non-muscle MHC IIs are known to be localized [18] (Supplementary Figure S2A at Co-staining of cardiomyocytes with the anti-MYL12A/12B/9 antibody and an antibody to major MHC IIs (MYH6 and MYH7) in cardiac muscle cells showed that these two groups of proteins are not co-localized in the muscle cells (Supplementary Figures S2A and S2B).

Association of MYH9 (NMHC IIA) and MYH10 (NMHC IIB) with MYL12A/12B/9 in NIH 3T3 fibroblasts

We identified proteins interacting with MYL12A, MYL12B and MYL9 by proteomic analysis. Immunoprecipitation of protein lysates from NIH 3T3 cells was performed using the anti-MYL12A/12B/9 antibody. Precipitates were subjected to tryptic digestion followed by MS analysis, which revealed that MYH9 and MYH10 are major proteins associated with MYL12A, MYL12B and MYL9 (Supplementary Table S3 at In addition MYL6, one of the ELCs, was found complete in these myosin complexes. MYH9 (NMHC IIA) and MYH10 (NMHC IIB), along with MYH14 (NMHC IIC), are known to be non-muscle MHC IIs [19]. We found that MYH9 and MYH10, but not MYH14, were abundant in NIH 3T3 cells. MYH14 was found to be expressed predominantly in the lung (Supplementary Figure S3A at and to be associated with MYL12A, MYL12B and MYL9 in this tissue (Supplementary Figure S3B). When immunoblotted with anti-MYH9, anti-MYH10 and anti-MYL6 antibodies, proteins immunoprecipitated by the anti-MYL12A/12B/9 antibody were found to contain MYH9, MYH10 and MYL6 in NIH 3T3 cells (Figure 3A), thus verifying the proteomic data. Immunostaining of NIH 3T3 cells using anti-MYH9 and anti-MYH10 antibodies showed a pattern highly similar to that obtained when the anti-MYL12A/12B/9 antibody was employed (Figure 3B). It should be noted that co-staining of the cells using the anti-MYL12A/12B/9 antibody with anti-MYH9 or anti-MYH10 antibody was not possible because these antibodies were raised from the same species.

Figure 3 Association of MYL12A, MYL12B and MYL9 with MHC IIs

(A) IP (immunoprecipitation) of MYL12A, MYL12B and MYL9 was performed on lysates of NIH 3T3 cells. An irrelevant IgG and beads were used as negative controls. Immune complexes were analysed by IB (immunoblotting) with anti-MYH9, anti-MYH10 and anti-MYL6 antibodies respectively. (B) Immunostaining for MYH9 and MYH10 (red) in NIH 3T3 fibroblasts. Actin stress fibres (green) and nuclei (blue) are also shown (scale bar, 20 μm).

Depletion of MYL12A and MYL12B results in dramatic changes in cell structure and function and decreases MYH9 and MYH10 levels

To determine the role of MYL12A, MYL12B and MYL9 in cellular functions, we depleted NIH 3T3 fibroblasts of these RLCs using siRNA oligonucleotides. A set of siRNAs drastically down-regulated the expression of two of the three genes, Myl12a and Myl12b (Figure 4A), most probably because of the sequence homology between the genes (Supplementary Table S1). Myl9, which is less conserved in the regions targeted by the siRNAs, was not changed significantly in transcript level. Immunoblotting analysis using the anti-MYL12A/12B/9 antibody showed a dramatic reduction in protein levels after knockdown (Figure 4B), suggesting that MYL12A and MYL12B are major RLCs in NIH 3T3 fibroblasts. siRNA-mediated knockdown of MYL12A and MYL12B resulted in striking changes in cell structure and morphology. We observed that actin stress fibre formation was disrupted in knockdown cells (Figure 4C). Furthermore after 36–60 h of knockdown, several filopodia now protruded from the cell body and the cytoplasm extended along the filopodia (Figure 4C) resulting in increases in cell size and protrusion length (Figure 4D and 4E). During a later period (72–84 h) of knockdown, cells developed additional protrusions from the extended filopodia and the cytoplasmic cell body became shrunken (Figures 4C, 4D and 4F). Despite these major alterations in cell structure, the viability of the knockdown cells did not differ from that of control cells (Figure 4G).

Figure 4 Knockdown of MYL12A/12B expression in NIH 3T3 fibroblasts and the resulting phenotype

(A) Transcript levels of the Myl12a, Myl12b and Myl9 genes in control and KD (knockdown) cells. Real-time PCR was performed on cDNAs obtained from cells after 45 h of knockdown using specific primers for Myl12a, Myl12b and Myl9 (*P<0.001, n=8). Non-targeting siRNA was used as a negative control. (B) Immunoblotting of proteins from control and knockdown fibroblasts 65 h after transfection, using the anti-MYL12A/12B/9 antibody. GPD1 was used as a control. (C) Morphology of control and KD cells (scale bar, 20 μm). (D) Cell size. Cell area measured and analysed using MetaMorph software. More than 30 individual cells were randomly selected for calculation. (E) Length of cellular protrusions. The results are the sums of the lengths of both the filopodia (extending from the cell body) and filopodial branches per cell. (F) The extent of cellular branching. Branching is defined as the number of protrusions divided by the number of filopodia (*P<0.01) (DF). (G) Cell viability measured by a dye (Trypan Blue) exclusion assay. No significant difference in viability was observed between control and KD cells (P>0.1, n=8).

To investigate isoform specificity between MYL12A and MYL12B in the knockdown phenotype, we performed individual knockdown analysis using siRNAs unique to each isoform in NIH 3T3 fibroblasts (Supplementary Table S4 at The Myl12a and Myl12b transcript levels were reduced significantly in Myl12a- and Myl12b-specific knockdown cells respectively (Supplementary Figures S4A and S4B at However, the protein levels of MYL12A, MYL12B and MYL9 were unchanged in isoform-specific knockdown fibroblasts, and the cells did not exhibit the altered phenotype observed in MYL12A/12B knockdown cells (Supplementary Figures S4C and S4D). This suggests the redundancy of these RLCs at the levels of expression and function.

We investigated further actomyosin-related cell behaviour, such as contractility and migration, in the MYL12A- and MYL12B- knockdown fibroblasts (Figure 5). The cellular contractility of knockdown cells was significantly reduced compared with controls, as shown by an in vitro fibrin gel contraction assay (Figure 5A). In addition, the migration rate was enhanced significantly in knockdown cells after wounding (Figure 5B). It should be noted that the pattern of the cellular changes seen after the depletion of MYL12A and MYL12B was similar to that induced by the absence of MYH9 [20] and MYH10 [21].

Figure 5 Behaviour of MYL12A- and MYB12B-knockdown NIH 3T3 fibroblasts

(A) Reduced contractility of knockdown cells as analysed by a fibrin gel contraction assay. Knockdown for 3 days suppressed the contraction of fibrin gel, as shown by staining with Ponceau S (*P<0.005, n=4). (B) Time-lapse photography assessing the migration of knockdown cells. Migration velocity doubled in such cells relative to controls (*P<0.001, n=30).

To examine whether the knockdown phenotype was directly related to MHC II expression, proteins from knockdown cells were immunoblotted using antibodies directed against heavy and light chains. Remarkably, we found that the levels of MYH9, MYH10 and MYL6 were dramatically reduced in knockdown cells (Figure 6A). To investigate further the loss of relevant proteins in knockdown cells, protein levels were monitored at different times during knockdown. MYL12A, MYL12B and MYL9 expression decreased 12 h after siRNA transfection. By contrast, the reduction in the levels of MYH9, MYH10 and MYL6 were first observed 24 or 36 h after knockdown commenced, indicating that these proteins were affected at a later stage compared with MYL12A, MYL12B and MYL9 (Figure 6B). Together, the results indicate that MYL12A and MYL12B are required to maintain the stability of MYH9, MYH10 and MYL6, thus suggesting that the changes in cell structure and function seen in knockdown fibroblasts are attributable to decreases in the levels of components of the myosin II complex.

Figure 6 Destabilization of myosin complexes in knockdown NIH 3T3 cells

(A) Immunoblotting of proteins extracted from control and knockdown fibroblasts 65 h after transfection. MHC IIs (MYH9 and MYH10) and an ELC (MYL6) were detected using specific antibodies. β-Actin, α-tubulin and GPD1 served as controls. (B) The expression levels of various myosin forms at different times after knockdown in fibroblasts. Each point represents a relative band intensity on immunoblotting of samples taken every 12 h during knockdown. Error bars represent S.E.M.s.

Characterization of the phenotype of MYL12A- and MYL12B- knockdown cells

To characterize further the phenotype of MYL12A- and MYL12B-knockdown cells, we inhibited either RLC phosphorylation or MHC II activity and compared the resulting cellular changes to those caused by MYL12A and MYL12B knockdown. NIH 3T3 fibroblasts were treated with Y-27632, a synthetic inhibitor of ROCK that is known to activate RLC via phosphorylation [22,23]. Most of the MYL12A, MYL12B and MYL9 proteins were dephosphorylated in the presence of Y-27632 (Figure 7B). We found that Y-27632-treated cells showed a degree of structural impairment similar to that observed during the early period (36–48 h) of MYL12A and MYL12B knockdown (Figure 4C), with disrupted actin stress fibres and extended filopodia (Figure 7A). Next, NIH 3T3 fibroblasts were treated with blebbistatin to inhibit MHC II ATPase activity [24]. These cells also underwent knockdown cell-like changes, such as the development of elongated protrusions and the loss of cytoplasm (Figure 7C). Notably, these characteristics were very similar to those of knockdown cells during the late phase of knockdown (72–84 h) (Figure 4C). Thus the results suggest that the integrity of MYL12A, MYL12B and MYL9, and not just complex phosphorylation, is critical for the full activity of MHC II in actomyosin units.

Figure 7 Phenotypes of cells treated with Y-27632 and blebbistatin

(A) NIH 3T3 cells were treated with the ROCK inhibitor Y-27632 (10 or 50 μM) for 4 h. The drug disrupted almost all the actin stress fibres, showing phenotypes comparable with those seen during the early periods of MYL12A and MYL12B knockdown. (B) Urea/glycerol PAGE. Most MYL12A, MYL12B and MYL9 existed as mono-phosphorylated proteins in the absence of Y-27632. However, upon addition of 50 μM Y-27632, a high proportion of MYL12A, MYL12B and MYL9 proteins became dephosphorylated. Di-P, di-phosphorylated; Mono-P, mono-phosphorylated; Non-p, non-phosphorylated. (C) Treatment of NIH 3T3 cells with blebbistatin (an inhibitor of myosin heavy chain action; 10 or 50 μM) for 4 h resulted in formation of many filopodia of similar morphology to those seen during the late phase after MYL12A and MYL12B knockdown (scale bar, 20 μm).

Analysis of interaction between RLCs and various types of myosin

To understand the interactions of MYL12A, MYL12B and MYL9, we examined comprehensively the binding of RLCs to a number of MHC IIs in NIH 3T3 cells, a mouse cardiomyocyte HL-1 cell line and bladder and skeletal muscle using immunoprecipitation and immunobloting analysis (Figure 8A). The expression patterns of MHC IIs in these cell types and tissues varied. Bladder and skeletal muscle did not express non-muscle MHC IIs. HL-1 cells were found to synthesize a wide range of MHC IIs. Interaction analysis revealed that MYL12A, MYL12B and MYL9 were associated with most MHC IIs expressed in cells and tissues, but to various extents. Interestingly, MYL12A, MYL12B and MYL9 in HL-1 showed a stronger interaction with non-muscle MHC IIs (MYH9, MYH10 and MYH14) than with cardiac MHC IIs (MYH6 and MYH7). Similarly, only a small proportion of MYL12A, MYL12B and MYL9 were associated with MYH1 and MYH2 in skeletal muscle (Figure 8A). Based on Figure 8(A), we prepared a schematic diagram to show the interaction profiles of MYL12A, MYL12B and MYL9 in different cell and tissue types (Figure 8B). Together, our results demonstrate that non-muscle RLCs can form myosin complexes with all types of MHC IIs, but non-muscle MHCs are preferred, despite the greater abundance of muscle MHC IIs.

Figure 8 Analysis of interactions between MYL12A, MYL12B and MYL9, and myosins

(A) Immunoprecipitation of extracts of NIH 3T3 cells, HL-1 cells, mouse bladder muscle and mouse skeletal muscle with an anti-MYL12A/12B/9 antibody. Samples of 10% of total input protein (input, lane 1) and immunoprecipitated protein (lane 4) were loaded. Proteins mock-immunoprecipitated with irrelevant IgG (lane 2), or by beads only (lane 3), served as controls. Co-precipitated proteins were analysed by immunoblotting using antibodies specific for various MHC IIs. Bd, bladder; Sk, skeletal muscle. (B) Interaction profiles of MYL12A, MYL12B and MYL9. Band intensity ratios of immunoprecipitated proteins with respect to levels in input lysates were calculated using ImageJ software ( Each circle indicates a myosin component associated with MYL12A, MYL12B or MYL9. The extent of binding preference is shown by both circle size and line thickness. Non-expressed proteins are shown as grey dots.


The present study provides comprehensive information on three murine RLC genes at the transcriptional, protein expression and functional levels. A number of previous studies have identified RLC genes in other mammalian species [6,7,11] and have investigated the phosphorylation of RLCs [23,25,26], regulation of MHC IIs by RLCs [9,27,28] and RLC-interacting proteins [8,10]. However, direct experimental data on the expression of mouse RLC genes has been lacking, and both the precise identities of MHC IIs interacting with RLCs and the phenotypes of cells lacking RLCs have not been described. The Myl12a, Myl12b and Myl9 genes are highly conserved in both gene structure and amino acid sequence of the resulting proteins. We were unable to identify any other genes highly similar to these RLC genes by genome database searches. The three genes show differential expression patterns. Myl12a is expressed widely, Myl12b is silent in striated muscle, whereas Myl9 is expressed strongly in smooth muscle. Despite this notable tissue preference, the three RLC genes are expressed in all non-muscle tissues (Figures 2A and 2B). This suggests functional redundancy of the RLC genes in these tissues, but confirms the importance of the genes as a group. Indeed, our findings indicate that proteins encoded by these genes are sufficient, when functioning as RLCs, to endow myosin complexes of NIH 3T3 fibroblasts with full activity.

LC-MS/MS and biochemical analyses provided direct evidence that MYL12A, MYL12B and MYL9 are associated with MYH9, MYH10 and MYL6 in NIH 3T3 fibroblasts. Three different genes encode non-muscle MHC IIs: Myh9 (NMHC IIA), Myh10 (NMHC IIB) and Myh14 (NMHC IIC). We found that MYH14 was not present in NIH 3T3 cells.

The functions of several myosin members have been investigated using gene ablation. Deletion of the genes encoding myosin proteins resulted in embryonic lethality or neonatal death [3,4,19,2932], making it difficult to study the cellular functions of the proteins. Therefore, we considered that siRNA-mediated knockdown might be a valuable approach to explore the functions of myosin components at the cellular level. The most remarkable finding of the present study is the loss of MHC IIs (MYH9 and MYH10) and an ELC (MYL6) in NIH 3T3 fibroblasts in which MYL12A and MYL12B were depleted by use of siRNA. The function of RLCs is to regulate MHC II activity and it has been considered that this was achieved principally by phosphorylation of RLCs, which increases the Mg2+-ATPase activity of myosin by regulating the conformation of myosin heads [1,28]. In addition, the phosphorylation of RLCs has been shown to promote the assembly of myosin filaments in vitro [1,33,34]. Although phosphorylation of RLCs is apparently crucial for myosin activity, our findings suggest strongly that cells require normal levels of RLCs to maintain intact MHC IIs. This implies the presence of another level of regulation by MYL12A and MYL12B, in which the phosphorylation status of MYL12A and MYL12B is not critical. Alternatively, it is possible that the absence of phosphorylated RLCs, rather than a lack of RLC proteins, is responsible for instability of MHC IIs. However, at present there is no evidence to indicate that inhibition of RLC phosphorylation causes loss of MHC IIs. In addition, as phosphorylation is reversible [1,35], any link between RLC phosphorylation and MHC II stability is unlikely. RLCs and ELCs bind to the α-helical stretch of a region between the head and the rod domains in a myosin heavy chain [1,28]. It has been suggested previously that RLCs stabilize MHCs [36]. A biochemical study has shown that MHC, in the absence of RLC, abnormally self-aggregates in the chicken skeletal myosin complex [37]. Another study demonstrated that the assembly of myofilaments is impaired in the zebrafish heart with a mutation in the cardiac RLC gene [38]. Finally, it has been reported that loss of non-muscle RLC in Drosophila melanogaster results in the aggregation of MHC but not ELC [39]. Nonetheless, experimental evidence for the loss of MHCs and ELC by RLC ablation or suppression in mammalian cells is lacking. Collectively, our results suggest that binding of RLCs to MHC IIs is essential to maintain the normal active conformation of the MHC IIs, and that stability of ELCs requires the presence of intact MHC IIs.

Turning to the cellular phenotypes seen after MYL12A and MYL12B knockdown, the fibroblasts showed dramatic changes in various cellular characteristics. Knockdown cells were severely altered in cell structure and morphology, showing defective formation of actin fibres and remarkable increases in the number and length of protrusions. In addition, cell dynamics were altered markedly, with a reduction in contractility and enhanced migration rate. Similar changes were also noted in MYH9- (MHC IIA) knockout and knockdown cells [20] and MYH10- (MHC IIB) knockout cells [21].

With regard to the specificity of each RLC, Myl12a and Myl12b, but not Myl9, were decreased in transcript level by siRNAs. This suggests strongly that MYL12A and MYL12B play major roles as non-muscle RLCs in NIH 3T3 fibroblasts and that MYL9 cannot compensate for the functions of MYL12A and MYL12B. In relation to this, it should be noted that MYL9 might exist at a very low level in NIH 3T3 fibroblasts. This is based on our observation that, despite the normal level of the Myl9 transcript, the amount of MYL12A, MYL12B and MYL9 was dramatically reduced in knockdown fibroblasts. On the other hand, the individual knockdown of Myl12a or Myl12b showed that the amount of protein detected by the anti-MYL12A/12B/9 antibody is maintained at a normal level and that the cellular phenotypes are unchanged. Thus this implies the importance of both genes and their functional redundancy.

Finally, it has not been clear whether a particular RLC might bind specifically to a certain MHC II. As MYL12A, MYL12B and MYL9 are expressed in muscle as well as non-muscle cells, we investigated the interaction of MYL12A, MYL12B and MYL9 with various MHC IIs in different cell and tissue types. Our findings indicated that MYL12A, MYL12B and MYL9 are associated with a variety of MHC IIs and that MYL12A, MYL12B and MYL9 bind preferentially to non-muscle MHC IIs. Thus the present study provides detailed information on the interaction of RLCs with MHC IIs.

In summary, we examined comprehensively the non-muscle RLCs of the mouse, providing key information on the expression, interactions and functions of MYL12A, MYL12B and MYL9. Importantly, we provide evidence that MYL12A and MYL12B are crucial for maintenance of the stability of MYH9, MYH10 and MYL6, which leads to normal cell actomyosin function. Thus our work identifies a phosphorylation independent form of myosin II regulation by RLCs in mammalian non-muscle cells.


Inju Park, Cecil Han, Sora Jin, Boyeon Lee, Heejin Choi, Jun Tae Kwon, Dongwook Kim, Jihye Kim and Ekaterina Lifirsu contributed to the esxperimental work. Inju Park, Woo Jin Park, Zee Yong Park, Do Han Kim and Chunghee Cho contributed to the experimental design and data analysis. Inju Park and Chunghee Cho wrote the manuscript. Chunghee Cho directed the project.


This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Ministry of Education, Science and Technology (MEST) [grant number 20100002159] and a GIST Systems Biology Infrastructure Establishment Grant.

Abbreviations: DMEM, Dulbecco's modified Eagle's medium; ELC, essential light chain; Gpd1, glycerol-3-phosphate dehydrogenase; GST, glutathione transferase; LC, liquid chromatography; mAb, monoclonal antibody; MHC II, myosin heavy chain; MYL, myosin light chain, MS/MS, tandem MS; NCBI, National Center for Biotechnology Information; pAb, polyclonal antibody; RT-PCR, reverse transcription PCR; RLC, regulatory light chain; ROCK, Rho-associated kinase; RT, room temperature; siRNA, short interfering RNA; UTR, untranslated region


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 11b.
  13. 12.
  14. 13.
  15. 14.
  16. 15.
  17. 15b.
  18. 15c.
  19. 16.
  20. 17.
  21. 18.
  22. 19.
  23. 20.
  24. 21.
  25. 22.
  26. 23.
  27. 24.
  28. 25.
  29. 26.
  30. 27.
  31. 28.
  32. 29.
  33. 30.
  34. 31.
  35. 32.
  36. 33.
  37. 34.
  38. 35.
  39. 36.
  40. 37.
  41. 38.
  42. 39.
View Abstract