Article
Reference
Transfected muscle and non-muscle actins are differentially sorted by cultured smooth muscle and non-muscle cells
MOUNIER, N., et al.
Abstract
We have analyzed by immunolabeling the fate of exogenous epitope-tagged actin isoforms introduced into cultured smooth muscle and non-muscle (i.e. endothelial and epithelial) cells by transfecting the corresponding cDNAs in transient expression assays. Exogenous muscle actins did not produce obvious shape changes in transfected cells. In smooth muscle cells, transfected striated and smooth muscle actins were preferentially recruited into stress fibers.
In non-muscle cells, exogenous striated muscle actins were rarely incorporated into stress fibers but remained scattered within the cytoplasm and frequently appeared organized in long crystal-like inclusions. Transfected smooth muscle actins were incorporated into stress fibers of epithelial cells but not of endothelial cells. Exogenous non-muscle actins induced alterations of cell architecture and shape. All cell types transfected by non-muscle actin cDNAs showed an irregular shape and a poorly developed network of stress fibers. beta- and gamma-cytoplasmic actins transfected into muscle and non-muscle cells were dispersed throughout the cytoplasm, often accumulated at the [...]
MOUNIER, N., et al . Transfected muscle and non-muscle actins are differentially sorted by cultured smooth muscle and non-muscle cells. Journal of Cell Science , 1997, vol. 110, no. 7, p. 839-846
PMID : 9133671
Available at:
http://archive-ouverte.unige.ch/unige:11048
Disclaimer: layout of this document may differ from the published version.
1 / 1
INTRODUCTION
Actin is one of the most abundant and conserved structural proteins of eukaryotic cells. It participates in a myriad of functions, including contractility, cell movement, intracellular transport, maintenance of cell shape, and signal transduction.
Actin is encoded by a multigene family which produces in mammals six closely related isoforms. They have been classi- fied according to their pattern of expression: four muscle actins, two in striated muscle (α-skeletal and α-cardiac), two in smooth muscle (SM) (α-vascular and γ-enteric, also called α- and γ-SM), and two non-muscle or cytoplasmic actins present in every cell, the β- and γ-isoforms. The isoactins exhibit a very high degree of protein sequence identity (99.5 to 93.5%) and muscle actins resemble each other more than do non-muscle actins (Vandekerckhove and Weber, 1978; Buck- ingham et al., 1984). This distinction between muscle and non- muscle actins appears to be a general feature in higher animals since it is observed in vertebrates and insects as well (Mounier et al., 1992).
In spite of their strong similarity, isoactins do not appear to be interchangeable in vivo but rather to reflect functional diversity. Isoforms are differentially expressed in different cell types, and within a single cell several isoforms may co-exist and segregate to distinct regions (Rubenstein, 1990; Herman,
1993; North et al., 1994). An unsolved question is how isoactins are sorted within the cell. Mechanisms controlling the localization and sorting of mRNA and protein are probably involved. Indeed the mRNAs of α-cardiac, β- and γ-cytoplas- mic actin genes have been shown to be differently localized within cells (Hoock et al., 1991; Hill and Gunning, 1993;
Kislauskis et al., 1993). The localization of the β-actin mRNA and its translation is induced by serum (Latham et al., 1994;
Hill et al., 1994). Actin sorting and assembly may involve the subtle conformational differences existing between isoforms, resulting from the few amino acid differences in the protein sequence, which would favor isoform-specific and actin binding protein-specific interactions (Mounier and Sparrow, 1993).
In this paper we have addressed the question of the utiliza- tion of a particular isoactin introduced by transfection of an expression plasmid containing the corresponding cDNA into different cultured cells in transient expression assays. Each cDNA contained at the end of the protein coding sequence a sequence coding for an epitope tag, which allowed us to dis- criminate exogenous from endogenous actins (von Arx et al., 1995). As muscular cell target, we used smooth muscle cells (SMCs) from rat aorta, characterized in vivo by the strong expression of specific genes such as α-SM actin gene and a remarkable plasticity. In response to physiological or patho-
Printed in Great Britain © The Company of Biologists Limited 1997 JCS9513
We have analyzed by immunolabeling the fate of exogenous epitope-tagged actin isoforms introduced into cultured smooth muscle and non-muscle (i.e. endothelial and epi- thelial) cells by transfecting the corresponding cDNAs in transient expression assays. Exogenous muscle actins did not produce obvious shape changes in transfected cells. In smooth muscle cells, transfected striated and smooth muscle actins were preferentially recruited into stress fibers. In non-muscle cells, exogenous striated muscle actins were rarely incorporated into stress fibers but remained scattered within the cytoplasm and frequently appeared organized in long crystal-like inclusions. Trans- fected smooth muscle actins were incorporated into stress fibers of epithelial cells but not of endothelial cells.
Exogenous non-muscle actins induced alterations of cell
architecture and shape. All cell types transfected by non- muscle actin cDNAs showed an irregular shape and a poorly developed network of stress fibers. ββ- and γγ-cyto- plasmic actins transfected into muscle and non-muscle cells were dispersed throughout the cytoplasm, often accumu- lated at the cell periphery and rarely incorporated into stress fibers. These results show that isoactins are differ- ently sorted: not only muscle and non-muscle actins are dif- ferentially distributed within the cell but also, according to the cell type, striated and smooth muscle actins can be dis- criminated for. Our observations support the assumption of isoactin functional diversity.
Key words: Actin isoform, Stress fiber, Microfilament, Cytoskeleton SUMMARY
Transfected muscle and non-muscle actins are differentially sorted by cultured smooth muscle and non-muscle cells
Nicole Mounier1,3, Jean-Claude Perriard2, Giulio Gabbiani3 and Christine Chaponnier3
1Centre de Génétique Moléculaire et Cellulaire, Université Lyon 1, 69622 Villeurbanne Cedex, France
2Institute for Cell Biology, Swiss Federal Institute of Technology, 8093 Zürich, Switzerland
3Department of Pathology, University of Geneva-CMU, 1211 Geneva 4, Switzerland
*Author for correspondence (e-mail: giulio.gabbiani@medecine.unige.ch)
840
logical signals, they modulate their phenotype by down regu- lating a set of contractile protein genes and thus exhibit a poorly differentiated phenotype, this de-differentiation being reversible. Cultured SMCs resemble the de-differentiated phenotype observed in vivo (for review see Desmoulière and Gabbiani, 1995; Owens, 1995). The non-muscle cells used in these experiments were bovine aorta endothelial cells and the epithelial cell line PtK2which express exclusively cytoplasmic actins (Kocher and Madri, 1989; Vandekerckhove and Weber, 1981). We show here that transfected muscle and non-muscle actins are distinctly distributed within the cells and that some of the tested cell types discriminate striated from SM actins.
Moreover, transfected cytoplasmic actins produce shape changes in all cell types.
MATERIALS AND METHODS Actin constructs
The six actin constructs used in our experiments were described by von Arx et al. (1995). They consist of full-length cDNA clones encoding the chicken α-skeletal and α-cardiac muscle actins, rat α- and γ-SM actins, human β- and rat γ-cytoplasmic actins subcloned in an eucaryotic expression vector composed of the cytomegalovirus (CMV) and T7 RNA polymerase promotors, and a rabbit β-globin genomic sequence containing an intron, splice sites and a poly(A) addition signal. Since the protein sequence of each isoactin is strictly conserved in humans, rodents and chicken (for review see Sheterline and Sparrow, 1994), we considered that the animal species from which the actin cDNAs and cells were isolated, was not relevant for our experiments.
To discriminate the transfected actin from endogenous actins, each cDNA was modified by inserting a tag just before the stop codon (Soldati and Perriard, 1991). This tag is a 36 nucleotide sequence encoding a foreign epitope allowing a specific immunolabeling. The epitope is composed of 11 amino acids of the C terminus of the vesicular stomatitis virus (VSV) G-protein against which polyclonal and monoclonal antibodies have been raised (Kreis, 1986; von Arx et al., 1995).
Cell culture
SMCs were isolated from thoracic aortic media of 6-week-old Wistar rats by enzymatic digestion (Bochaton-Piallat et al., 1992). They were plated on culture dishes in Dulbecco’s modified Eagle’s medium (DMEM, Gibco, Basel, Switzerland) supplemented with 10% fetal calf serum (FCS, Biological Industries, Bet-Haemek, Israel). SMCs were used for transfection experiments at passage 4-7.
Endothelial cells were isolated from bovine aorta and cultured as previously described (Gabbiani et al., 1984). The epithelial cell line PtK2 was kindly provided by Dr Pascale Cossart, Institut Pasteur, Paris, France, and cultured in modified Eagle’s medium (MEM) sup- plemented with 1% non essential amino acids, 1 mM sodium pyruvate (all from Gibco), and 10% FCS.
DNA transfection
Transfection was carried out by cationic liposome-mediated methods using Lipofectamine or LipofectACE reagents and lipopolyamine- mediated methods using di-octadecylamido-glycyl-spermine (DOGS) as described by the manufacturer (Gibco) and Barthel et al. (1992).
Cells were seeded into 35 mm culture dishes and grown to mid-con- fluence. For each transfection, 2 µg of DNA purified using the Qiagen plasmid kit (Qiagen AG, Basel, Switzerland) and 12 µl of Lipofect- amine or LipofectACE reagents, or 4 µl of a 2 mM DOGS solution in ethanol (kindly provided by Dr Jean-Philippe Loeffler, Strasbourg, France), were used. Cells were incubated with the complexes for 5-6
hours at 37°C and the transgene activity was analyzed after 48 hours of transient expression.
Antibodies and immunofluorescence staining
Double immunofluorescence staining was performed directly in the culture dish. After transfection, cells were either fixed at room tem- perature in 3% paraformaldehyde for 10 minutes and permeabilized with 0.1% Triton X-100 for 2 minutes or fixed in cold methanol for 5 minutes.
The tagged actin was detected using either a rabbit polyclonal antibody (von Arx et al., 1995) or the mouse monoclonal antibody P5D4 (a gift from Dr Thomas Kreis, Geneva, Switzerland) directed against the VSV G-protein. F-actin was revealed by rhodamine- or fluorescein isothiocyanate (FITC)-phalloidin. We also used a mouse monoclonal antibody specific for the N terminus of α-SM actin (Skalli et al., 1986b) and a rabbit polyclonal antibody specific for β-cyto- plasmic actin. This latter was generated by immunizing rabbits with the corresponding hemocyanin-coupled NH2 terminal nonapeptide and purified by affinity chromatography (Yao et al., 1995), by using the nonapeptide covalently linked to Mini-Leak beads (Kem-En-Tec A/S, Copenhagen, Denmark) according to the manufacturer’s descrip- tion. The secondary antibodies were FITC-conjugated goat anti-rabbit IgG or anti-mouse total Ig to visualize tagged actins and tetramethyl rhodamine isothiocyanate (TRITC)-conjugated goat anti-mouse total Ig for α-SM actin (Jackson, Immuno Research Laboratories, West Grove, PA, USA).
Photographs were taken with a Zeiss Axiophot photomicroscope (Carl Zeiss, Oberkochen, Germany) equipped for epi-illumination and specific filters for fluorescein and rhodamine, by using an oil immersion plan neofluar objective (×40/1.3) on Kodak T-MAX 400 film (Eastman Kodak, Rochester, NY, USA).
Immunostained transfected cells were scanned with a high sensi- tivity Photonic Science Coolview colour camera (Carl Zeiss) through the photomicroscope using the oil immersion objective ×40/1.3. The camera was connected to a 486DX2/66 Intel PC and the images were captured using the software package Image Access 2.04 (Imagic, Zürich, Switzerland). Image processing was performed with Adobe Photoshop 3.0 (Adobe Systems Inc., Mountain View, USA) and the resulting images were photographed on a Kodak Ektachrome E100SW slide film with a freeze-frame digital camera (Focus Graphics, Geneva, Switzerland).
Evaluation of the staining distribution
An evaluation of the staining distribution was performed on cells transfected by non-muscle actin cDNAs since the obtained pheno- types were variable. The shape of 700 cells transfected by β- and γ- cytoplasmic actin cDNAs in, respectively, 4 and 3 independent exper- iments was examined and classified as: (1) large and flat; (2) small and condensed; and (3) unmodified, in comparison to the general aspect of non-transfected cells. The network of stress fibers was examined in 210 non-overlapping transfected cells and defined as normal and poorly developed when more and less than 10 stress fibers were present per cell, respectively.
In vitro polymerization of tagged actin and immunoblot analysis
Rabbit skeletal muscle actin was purified from acetone powder by the method of Spudich and Watt (1971) in buffer A (0.5 mM ATP, 0.5 mM DTT, 0.2 mM CaCl2, 1 mM NaN3, 2 mM Tris-HCl, pH 7.8) at 4°C. Two cycles of polymerization were performed prior to gel fil- tration on Sephacryl S-300 (Pharmacia, Dübendorf, Switzerland).
Control and transfected SMCs were trypsinized, washed three times in DMEM without FCS, then 100,000 cells were resuspended in 150 µl buffer A containing protease inhibitors (1 mM PMSF, 0.2 mg/ml aprotinin) and extracted with several passages through a 25G5/8 gauge needle. Cell extracts were centrifuged for one hour in a Beckman Airfuge (Beckman Instruments, Morges, Switzerland) for N. Mounier and others
30 minutes at 30 psi using an A-100 fixed angle rotor. Actins from cell supernatants were polymerized with 30 µl of 22 µM α-skeletal actin in the presence of 100 mM KCl and 2 mM MgCl2(final con- centration) for 30 minutes at room temperature. After a second airfuge centrifugation step in the same conditions as above, supernatants con- taining unpolymerized actin, and pellets containing polymerized actin were run on a 10% SDS-PAGE gel and transferred onto a nitrocellu- lose membrane. The blot was incubated with the monoclonal antibody specific for the tag epitope and with goat anti-mouse IgG conjugated to horseradish peroxidase (Jackson), and developed with the ECL western blotting system (Amersham, Buckinghamshire, UK).
RESULTS
Cultured cells and transfection
Adult rat arterial SMCs express muscle-specific genes such as α-SM actin, SM myosin, and desmin. When they are placed in culture, there is a de-differentiation process with a decrease in muscle-specific gene expression and an increase in cytoplas- mic actin gene expression. At the 5th passage, 70% of SMC still contain α-SM actin, while only 3% contain SM myosin and desmin is not expressed. The average actin isoform content per total actin, as seen on two-dimensional gels, is 6% α-SM actin, 64% β-cytoplasmic actin and 30% γ-actins, no distinc- tion being possible by this technique between γ-cytoplasmic and γ-SM actins (Skalli et al., 1986a; Bochaton-Piallat et al., 1992). These cells actively replicate and present a well organized and abundant network of stress fibers stained by phalloidin and total actin antibody, and in most cells by α-SM actin specific antibody (Skalli et al., 1986b).
Endothelial cells derived from bovine aorta and the epithe- lial cell line PtK2express exclusively the β- and γ-cytoplasmic actins which are present at a ratio of approximately 70:30%
and 85:15%, respectively (Gabbiani et al., 1984; C. Chapon- nier, unpublished observations). Endothelial cells show an important peripheral ring of microfilaments and a few randomly distributed actin cables (Gabbiani et al., 1984) while PtK2cells, in addition to a well developed peripheral ring of microfilaments, exhibit numerous stress fibers (Huckriede et al., 1990).
All the actin cDNAs transfected in our experiments were driven by the same promotor (CMV and T7 polymerase promotors) and utilized the same signals from the rabbit β- globin gene for intron splice sites and poly(A) addition. We tested on SMCs another heterologous promotor, intron and poly(A) signals derived from SV40, and obtained identical results.
The different actin cDNAs were transfected into cultured cells with cationic liposomes. Comparable results were obtained when using Lipofectamine, LipofectACE reagents or DOGS. All methods produced 1-5% of transfected cells.
Expression of an additional actin gene controlled by an het- erologous promotor and 3′untranslated region was not lethal for the cells tested, since transfected cells were observed in all cell types and with each of the six actin constructs.
Expression of muscle actins
The expression of additional muscle actins in SMCs did not influence their cytoarchitecture and results obtained with trans- fected α-skeletal, α-cardiac, α- and γ-SM actin cDNAs were identical. The shape and size of transfected cells were similar
to those of non-transfected cells and the stress fiber network was as well organized as in control cells (Fig. 1B,D,F). Trans- fected muscle actins were preferentially recruited into stress fibers and the tagged actin containing stress fibers were also stained by phalloidin (Fig. 1A-B, C-D, E-F). When a cell presented only a few stress fibers incorporating tagged actin, the F-actin network revealed by phalloidin was equally poorly developed. Among cells transfected by the α-SM actin cDNA, about 50% showed tagged actin efficiently integrated into the stress fibers and about 50% presented a weak incorporation of this actin into the microfilamentous apparatus (compare, in Fig.
1, tagged cells of C with those of A and E). Such a difference in tagged actin utilization was not observed in cells transfected by the α-cardiac and γ-SM actin cDNAs.
In addition to stress fibers, tagged muscle actins were also distributed throughout the cytoplasm, delimiting clearly the cellular edges, and forming granular accumulations, especially in the perinuclear region where the cell is thicker. These tagged granules had variable sizes and were not or were only poorly stained by phalloidin (compare Fig. 1A-B, C-D, E-F). They were probably rich in polymerized actin and resulted from the transgene expression since no such aggregates were present in non-transfected cells.
Fig. 2 shows the superimposition of the anti-tag and phal- loidin stainings in SMCs transfected by γ-SM actin cDNA.
Tagged and F-actins were colocalized in stress fibers and the granular accumulations of tagged actin particularly abundant in the perinuclear region, were not stained by phalloidin.
Most SMCs had a well developed network of stress fibers containing α-SM actin. In cells transfected by muscle actin cDNAs, the network of tagged stress fibers was also stained by the α-SM actin antibody (Fig. 1G-H). However, 5-10% of cells transfected with striated and γ-SM actin cDNAs presented an abundant network of tagged stress fibers which did not contain α-SM actin (Fig. 1I-J). Possibly, these cells were not express- ing the gene coding for α-SM actin when the transfection procedure occurred, but they were still able to utilize the trans- fected actin in the same way as differentiated SMCs.
We investigated whether the tagged muscle actins can poly- merize in vitro. Actin obtained from the soluble fractions of non-transfected and α-cardiac actin cDNA transfected SMCs were mixed with α-skeletal muscle actin. After polymeriza- tion, more than 80% of the total actins, i.e. endogenous and exogenous actins, were recovered in the pellet fraction from both control and transfected cells (Fig. 3A). After western blotting, anti-tag staining was positive in the fractions from transfected cells with a strong signal in the pellet fraction, indi- cating that the transfected α-cardiac actin copolymerizes in vitro with α-skeletal actin (Fig. 3B).
Interestingly, transfected striated muscle actins were correctly recognized and targeted to SMC stress fibers and the resulting network was as abundant and well organized as in non-transfected cells, although these cells never express genes coding for striated muscle actins during their ontogeny (McHugh et al., 1991).
We investigated whether transfected muscle actins were utilized by cells which never express muscle actin genes. The four muscle actin cDNAs were transfected into bovine aortic endothelial cells and the epithelial cell line PtK2. In all cases transfected cells presented a shape similar to that of non-trans- fected cells. Endothelial cells transfected by striated muscle
842
actin cDNAs presented thin, straight, or curved inclusions which were strongly tag-positive and not or only poorly stained by phalloidin (Fig. 4A-B). These crystal-like inclusions probably contained unpolymerized actin and did not apparently interfere with the cell architecture and actin stress fiber organ- ization. The rest of the tagged actin was scattered within the cytoplasm as small granular accumulations. Similar crystal- like inclusions were observed in PtK2 cells transfected by striated muscle actin cDNAs and also in rare cases (1-2%) in SMCs transfected by α-cardiac actin cDNA, but in this latter case only 1 to 3 short and thin inclusions per cell were present (data not shown).
SM actins transfected into endothelial cells were distributed throughout the cytoplasm as granules and not or exceptionally integrated into actin filaments (Fig. 4C-D). On the contrary in PtK2cells, transfected SM actins were well incorporated into stress fibers and also dispersed in a granular pattern within the cytoplasm (Fig. 4E-F).
Expression of non-muscle actins
Transfection of non-muscle actin cDNAs had a rather severe impact on shape and architecture of SMCs. The pattern of expression did not present any significant difference when the β- or γ-cytoplasmic actin cDNAs were used. Transfected cells presented variable aspects: about 20% were similar to control cells (Fig. 5A), about 20% presented a large and flattened shape (Fig. 5C) and most frequently (about 60%) exhibited a small and condensed cell shape with local and irregular protuber- ances at the cell periphery (Fig. 5E,G).
Tagged cytoplasmic actins were distributed in granular accu- mulations scattered within the cytoplasm and large deposits of tagged actin were often accumulated at the cell periphery. These accumulations of transfected actin were not or were only poorly stained by phalloidin (Fig. 5A-B, C-D, E-F). They were stained by β-cytoplasmic actin antibody in cells transfected by the β-actin plasmid (Fig. 5G-H); these cells were more heavily positive to the β-actin antibody than adjacent non-transfected cells (Fig. 5H).
N. Mounier and others
Fig. 1. Double immunofluorescence staining of SMCs transfected by cDNAs coding for α-cardiac (A-B, G-H, I-J), α-SM (C-D) and γ-SM (E-F) actins. The pattern of expression of transfected α-skeletal muscle actin cDNA was identical to that obtained with the other muscle actin cDNAs (data not shown). (A,C,E,G,I) Cells stained by the antibody specific for the tag epitope. (B,D,F,H,J) The same cells stained by either phalloidin (B,D,F) or α-SM actin antibody (H,J). In each panel, one or several transfected cells are shown close to non-transfected cells. Exogenous tagged muscle actins were incorporated into stress fibers and granular accumulations scattered within the cytoplasm. Tagged stress fibers were stained by phalloidin (B,D,F) or α-SM actin antibody (H). In some cases, however, stress fibers of the transfected cells were devoid of α-SM actin (J).
Tagged granular accumulations were not stained by phalloidin (B,D,F) nor by α-SM actin antibody (H). The arrow in C shows cells transfected by the α-SM actin cDNA with a network of stress fibers presenting a weak anti-tag staining when compared to that of the surrounding transfected cells, although their microfilament network was normally developed as shown by the phalloidin staining (arrow in D). Bar, 10 µm.
Tagged cytoplasmic actins were not or only poorly incorpo- rated into the stress fibers of transfected cells (Fig. 5A,C and E). The network of F-actin identified by phalloidin in SMCs transfected by non-muscle actin cDNAs was poorly developed (i.e. less than 10 stress fibers per cell) in all transfected cells presenting a condensed shape (Fig. 5F) and in about one third of transfected cells with the unmodified morphology (Fig. 5B).
This stress fiber network was normal (i.e. more than 10 stress fibers per cell) in about two thirds of the transfected cells pre- senting a large and flat shape (Fig. 5D).
Cultured SMCs exhibit a network of stress fibers stained by the antibody specific for β-cytoplasmic actin, however, tagged β-actin was poorly incorporated into stress fibers of transfected cells (Fig. 3G-H). This poor incorporation probably did not result from the impossibility for this isoactin to be integrated into stress fibers but rather indicated a disturbance of the cellular machinery utilized for the incorporation of transfected β-actin into stress fibers.
The pattern of expression of transfected non-muscle actins into endothelial and epithelial cells was similar to that observed in SMCs. Comparable results were obtained in these two cell types when using the β- or γ-cytoplasmic actin cDNAs. The majority of transfected cells also exhibited an irregular shape and tagged actin was scattered within the cytoplasm as granular
accumulations and frequently accumulated as large deposits at the cell periphery (data not shown).
The presence of the tag is probably not responsible for the altered shape of cells transfected by cytoplasmic actins.
Although the actin C terminus in which the tag was inserted is a binding site for several actin binding proteins such as profilin Fig. 2. Double fluorescence staining of SMCs transfected by
γ-SM actin cDNA. Transfected cells were stained with the antibody specific for the tag epitope detected with FITC- conjugated anti-rabbit IgG antibody (A) and with rhodamine- conjugated phalloidin (B). The 2 fluorescent channels of the same image were superimposed (C). Double labeled structures appear in yellow. Note the colocalization of the tagged and F-actins in the stress fibers whereas the tagged actin containing granules, particularly abundant in the perinuclear region, are not stained by phalloidin. Bar, 2 µm.
A B C
Fig. 3. In vitro polymerization of tagged actin. Total proteins were extracted from control (C) and α-cardiac actin cDNA transfected (T) SMCs. Actin obtained from the soluble fraction was mixed with α- skeletal actin for copolymerization. Supernatants (S) and pellets (P) obtained after ultracentrifugation containing, respectively,
unpolymerized and polymerized actin, were run on a 10% SDS- PAGE gel and stained with Coomassie blue (A). About 20 and 80%
of total actins were recovered, respectively, in the supernatant and pellet fractions in both control and transfected cells. Proteins from an identical gel were transferred onto nitrocellulose for anti-tag immunoblotting (B). The anti-tag staining was positive in the S and P fractions from transfected cells, with a strong signal in the P fraction indicating the capability of the tagged actin to polymerize in vitro.
Fig. 4. Double immunofluorescence staining of endothelial cells transfected by cDNAs coding for α-cardiac (A-B) and γ-SM (C-D) actins, and of the epithelial cell line PtK2transfected by the γ-SM actin cDNA (E-F). (A,C,E) Cells stained by the antibody specific for the tag epitope. (B,D,F) The same cells stained by phalloidin. Note in A the strongly tag-positive, long crystal-like inclusions (arrowheads) in endothelial cells transfected by striated α-cardiac muscle actin cDNA and the absence of tagged actin in actin filaments stained by phalloidin in B. Transfected γ-SM actin accumulated into punctiform structures in endothelial and epithelial cells and was incorporated into stress fibers only in PtK2cells (E-F) and not in endothelial cells (C-D). Identical results were obtained with α-SM actin transfected into both cell types (data not shown). Bar, 10 µm.
844
(for review see Sheterline and Sparrow, 1994) and any confor- mational change in this region may be relevant for interactions with ligands, the presence or absence of the tag has been shown
to have no effect on the distribution of transfected γ-cytoplas- mic actins in cardiomyocytes (von Arx et al., 1995). As we used the same constructs, we assume that the observed shape changes did not result from the presence of the tag in trans- fected cytoplasmic actins.
DISCUSSION
Our results indicate that transfected muscle and non-muscle actins are differentially discriminated and utilized by cultured SM and non-muscle cells and have different effects on cell shape as summarized in Table 1. In transient expression assays, the presence of exogenous muscle actins did not generate sig- nificant changes in the shape and morphology of the cell types tested. SMCs efficiently incorporated transfected striated and SM actins into stress fibers and the remainder of the exogenous actin was scattered within the cytoplasm in a granular pattern.
SMCs, which never express striated muscle actin genes, were capable of efficiently integrating striated muscle actins into stress fibers, suggesting that mechanisms utilized for SM actin incorporation were sufficient to sort striated actins. The epi- thelial and endothelial cells studied here, which never express any muscle actin genes, did not efficiently integrate the trans- fected striated muscle actins into their microfilamentous apparatus. Possibly, these ‘foreign’ actins do not find in the cell favorable conditions to polymerize adequately and so accu- mulate as crystal-like inclusions. However, transfected SM actins are incorporated into stress fibers of the epithelial cell line PtK2but not endothelial cells.
Expression of exogenous muscle actins has been investi- gated in other models. Microinjected α-skeletal muscle actin was incorporated into sarcomeres of cardiac myocytes and stress fibers of fibroblasts (McKenna et al., 1985; LoRusso et al., 1992). Transfected muscle actins were integrated into stress fibers of fibroblasts (Gunning et al., 1984; von Arx et al., 1995) and when the corresponding actin cDNAs were microinjected into cardiomyocytes, muscle actins and particularly α-cardiac muscle actin were preferentially recruited into myofibrils (von Arx et al., 1995). From these results and our work it appears that muscle actins artificially introduced into muscle and non- muscle cells do not disturb the cell significantly even if they N. Mounier and others
Table 1. Effects of transfected actin isoforms on cell shape and incorporation of tagged actins into stress fibers in
smooth muscle and non-muscle cells
Changes in Incorporation
cell shape into stress fibers Smooth muscle cells
Striated muscle actins No Yes
Smooth muscle actins No Yes
Non-muscle actins Yes No
Endothelial cells
Striated muscle actins No No
Smooth muscle actins No No
Non-muscle actins Yes No
Epithelial PtK2cells
Striated muscle actins No No
Smooth muscle actins No Yes
Non-muscle actins Yes No
Fig. 5. Double immunofluorescence staining of SMCs transfected by cDNAs coding for non-muscle β-cytoplasmic actin (C,E,G) and γ- cytoplasmic actin (A). (A,C,E,G) Cells stained by the antibody specific for the tag epitope. (B,D,F,H) The same cells stained by either phalloidin (B,D,F) or the antibody specific for β-cytoplasmic actin (H).
Each panel presents one or several transfected cells close to non- transfected cells. Transfected cells presented different shapes: normal (A), large and flat (C) and small and condensed (E,G). Transfected non- muscle actins accumulated in a granular pattern within the cytoplasm and as large deposits at the cell periphery. The tagged granules were not or were only poorly stained by phalloidin (B,D,E) and the large aggregates were either not stained by phalloidin (arrow in A-B) or were phalloidin positive (arrowheads in E-F). Tagged actin was weakly incorporated into stress fibers and the microfilament network was poorly developed. Arrowheads in G-H show the corresponding staining of peripheral aggregates by specific tag and β-cytoplasmic actin antibodies in cells transfected by the β-actin cDNA. Bar, 10 µm.
are not utilized and accumulate as crystal-like inclusions.
Muscle actins are not only efficiently integrated into microfila- ments of striated muscle cells (von Arx et al., 1995) and SMCs (this work) as expected, but can also be incorporated into stress fibers of some non-muscle cells such as fibroblasts (von Arx et al., 1995) and the epithelial cell line PtK2, albeit these latter cells incorporate only SM actins (this work). Since fibroblasts in culture develop SMC features similar to those seen in gran- ulation tissue myofibroblasts, in particular the expression of α- SM actin (for review see Desmoulière and Gabbiani, 1995), they may have available the machinery to incorporate SM actins into their stress fibers. The capability of PtK2cells to integrate SM actins into stress fibers may be related to the process of epithelial-mesenchymal transformation through which, in some instances, epithelial cells may modulate their phenotype to become mesenchymal cells (for review see Hay, 1995). Thus, cultured mouse mammary epithelial cells treated with TGFβ modulate from an epithelial to a fibroblastic phenotype (Miettinen et al., 1994). Moreover, a subset of epithelial cells, i.e. myoepithelial cells, is characterized by the expression of α-SM actin and SM myosin heavy chains (Benzonana et al., 1988; Lazard et al., 1993). In contrast, bovine aortic endothelial cells in culture do not express SM markers even after treatment with TGFβcontrary to microvas- cular endothelial cells (Kocher and Madri, 1989). Further work using other sources of epithelial and endothelial cells is needed in order to investigate the capacity of these cell types to dis- criminate muscle actin isoforms.
Our results indicate that, according to the cell type, striated and SM actins can be discriminated by the cellular machinery.
This suggests the presence, or absence, of yet unknown actin binding proteins capable of recognizing a particular actin isoform and to allow its participation in stress fiber formation.
We have recently shown that the N-terminal sequence AcEEED of α-SM actin controls α-SM actin polymerization in vitro and in vivo, via an interaction with a not yet identified partner (Chaponnier et al., 1995). It is conceivable that a mechanism involving the interaction between a specific sequence in each actin isoform and a partner factor promotes or controls the isoform incorporation into stress fibers in the different cell types.
Transfected cytoplasmic actins changed the morphology and organization of SMCs as well as non-muscle cells. Transfected cells exhibited irregular shape and size and tended to have a poorly developed network of stress fibers. Exogenous non- muscle actins were rarely or not integrated into stress fibers, but dispersed within the cytoplasm in a granular pattern and often accumulated at the cell periphery as large aggregates.
These results are in agreement with those reported for trans- fected cytoplasmic actins in other cell types. Transfection of non-muscle actin genes into myoblasts produced changes at the cell surface and in stress fiber organization, with slightly different patterns according to the transfected genes (Schevzov et al., 1992). Microinjection of tagged β- or γ-cytoplasmic actin cDNAs into cardiomyocytes induced severe changes in both morphology and function: the thin filaments of sarcomeres were removed, the exogenous as well as endogenous actins accumulated at the cell periphery and filopodial extensions were observed (von Arx et al., 1995). Similar results were pre- viously obtained when the γ-cytoplasmic actin cDNA construct not containing the tag epitope was microinjected in these cells,
ruling out the possibility of an effect of the tag on the resulting altered cell morphology (von Arx et al., 1995). Thus, cyto- plasmic actin genes or cDNAs transfected into striated muscle, SM and non-muscle cells modify the cell shape and microfil- ament network organization. However, we did not observe a differential effect of the transfected β- and γ-cytoplasmic actin cDNAs.
The final distribution of isoactins within the cell is the result of many processes from gene activity to protein assembly. One of the regulatory levels is in the intracellular localization of the mRNA. Sequences responsible for the β-actin mRNA local- ization in fibroblasts were mapped within the 3′untranslated region of the message and shown to be important for the further organization of actin microfilaments (Kislauskis et al., 1994).
From the results presented here, the 3′untranslated region of muscle actin mRNAs does not appear to be essential for the targeting of the corresponding protein since there was no apparent change in the cytoarchitecture of cells transfected by muscle actin genes. In contrast we cannot exclude the possi- bility that the 3′ untranslated regions of cytoplasmic actin mRNAs might have an effect on the utilization of the corre- sponding proteins.
Our results suggest that, whatever the cell type, modifica- tions of the ratio between muscle actins or between muscle and non-muscle actins do not disturb the cell, but a change in the β/γ-cytoplasmic actin equilibrium produces a profound effect on cell shape and architecture. The β/γ-cytoplasmic actin ratio varies at the levels of both mRNA and protein according to the tissue involved (Vandekerckhove and Weber, 1981; Otey et al., 1987; Skalli et al., 1987; Erba et al., 1988). Changes in this ratio associated with a disruption and/or reorganization of the microfilament system resulting in shape alteration were reported for several transformed cells (for review see Janmey and Chaponnier, 1995). It remains to be understood why trans- fected non-muscle actins tend to largely modify the cytoarchi- tecture. These severe effects may result either from a direct action of the protein itself or from the imbalance of specific actin binding proteins which may interfere with the continuous process of actin polymerization and depolymerization.
This work was performed in the laboratory of G. Gabbiani and supported by the Swiss National Science Foundation (grant nos 31- 40372.94 and 31-43582.95). We are grateful to Dr Pascale Cossart for the PtK2cells, Dr Thomas Kreis for the monoclonal tag antibody, and Dr Jean-Philippe Loeffler for the DOGS solution. We thank Dr Marie- Luce Bochaton-Piallat for help in color image processing, Anita Hilt- brunner for technical assistance, Jean-Claude Rumbeli and Etienne Denkinger for the photographic work and Myriam Vitali for typing the manuscript.
REFERENCES
Barthel, F., Boutiller, A. L., Giraud, P., Demeneix, B. A., Behr, J. P. and Loeffler, J. P. (1992). Gene regulation analysis by lipopolyamine-mediated DNA transfer in primary neurons. Meth. Neurosci. 9, 291-312.
Benzonana, G., Skalli, O. and Gabbiani, G. (1988). Correlation between the distribution of smooth muscle or non muscle myosins and α-smooth muscle actin in normal and pathological soft tissues. Cell Motil. Cytoskel. 11, 260- 274.
Bochaton-Piallat, M. L., Gabbiani, F., Ropraz, P. and Gabbiani, G. (1992).
Cultured aortic smooth muscle cells from newborn and adult rats show distinct cytoskeletal features. Differentiation 49, 175-185.
Buckingham, M. E., Alonso, S., Barton, P., Bugaisky, G., Cohen, A.,
846
Daubas, P., Garner, I., Minty, A., Robert, B. and Weydert, A. (1984).
Actin and myosin genes, and their expression during myogenesis in the mouse. In Molecular Biology of Development (ed. E. H. Davidson and R. A.
Firtel), pp. 275-292. Alan R. Liss, New York.
Chaponnier, C., Goethals, M., Janmey, P. A., Gabbiani, F., Gabbiani, G.
and Vandekerckhove, J. (1995). The specific NH2-terminal sequence Ac- EEED of α-smooth muscle actin plays a role in polymerization in vitro and in vivo. J. Cell Biol. 130, 887-895.
Desmoulière, A. and Gabbiani, G. (1995). Smooth muscle cell and fibroblast biological and functional features: similarities and differences. In The Vascular Smooth Muscle Cell. Molecular and Biological Responses to the Extracellular Matrix (ed. S. M. Schwartz and R. P. Mecham), pp. 329-359.
Academic Press, San Diego.
Erba, H. P., Eddy, R., Shows, T., Kedes, L. and Gunning, P. (1988).
Structure, chromosome location, and expression of the human γ-actin gene:
differential evolution, location, and expression of the cytoskeletal β- and γ- actin genes. Mol. Cell. Biol. 8, 1775-1789.
Gabbiani, G., Gabbiani, F., Heimark, R. L. and Schwartz, S. M. (1984).
Organization of actin cytoskeleton during early endothelial regeneration in vitro. J. Cell Sci. 66, 39-50.
Gunning, P., Ponte, P., Kedes, L., Hickey, R. J. and Skoultchi, A. I. (1984).
Expression of human cardiac actin in mouse L cells: a sarcomeric actin associates with a non muscle cytoskeleton. Cell 36, 709-715.
Hay, E. D. (1995). An overview of epithelio-mesenchymal transformation.
Acta Anat. 154, 8-20.
Herman, I. M. (1993). Actin isoforms. Curr. Opin. Cell Biol. 5, 48-55.
Hill, M. A. and Gunning, P. (1993). Beta-actin and gamma-actin messenger RNAs are differentially located within myoblasts. J. Cell Biol. 122, 825-832.
Hill, M. A., Schedlich, L. and Gunning, P. (1994). Serum-induced signal transduction determines the peripheral location of beta-actin mRNA within the cell. J. Cell Biol. 126, 1221-1229.
Hoock, T. C., Newcomb, P. M. and Herman, I. M. (1991). Beta actin and its mRNA are localized at the plasma membrane and the regions of moving cytoplasm during the cellular response to injury. J. Cell Biol. 112, 653-664.
Huckriede, A., Füchtbauer, A., Hinssen, H., Chaponnier, C., Weeds, A. and Jockusch, B. M. (1990). Differential effects of gelsolins on tissue culture cells. Cell Motil. Cytoskel. 16, 229-238.
Janmey, P. A. and Chaponnier, C. (1995). Medical aspects of the actin cytoskeleton. Curr. Opin. Cell Biol. 7, 111-117.
Kislauskis, E. H., Li, Z., Singer, R. H. and Taneja, K. L. (1993). Isoform- specific 3′-untranslated sequences sort alpha-cardiac and beta-cytoplasmic actin messenger RNAs to different cytoplasmic compartments. J. Cell Biol.
123, 165-172.
Kislauskis, E. H., Zhu, X. C. and Singer, R. H. (1994). Sequences responsible for intracellular localization of beta-actin messenger RNA also affect cell phenotype. J. Cell Biol. 127, 441-451.
Kocher, O. and Madri, J. A. (1989). Modulation of actin mRNAs in cultured vascular cells by matrix components and TGF-β. In Vitro Cell Dev. Biol. 25, 424-434.
Kreis, T. E. (1986). Microinjected antibodies against the cytoplasmic domain of vesicular stomatitis virus glycoprotein block its transport to the cell surface. EMBO J. 5, 931-941.
Latham, V. M., Kislauskis, E. H., Singer, R. H. and Ross, A. F. (1994). Beta- actin mRNA localization is regulated by signal transduction mechanisms. J.
Cell Biol. 126, 1211-1219.
Lazard, D., Sastre, X., Frid, M. G., Glukhova, M. A., Thiery, J.-P. and Koteliansky, V. E. (1993). Expression of smooth muscle-specific proteins in myoepithelium and stromal myofibroblasts of normal and malignant human breast tissue. Proc. Nat. Acad. Sci. USA 90, 999-1003.
LoRusso, S. M., Imanaka Yoshida, K., Shuman, H., Sanger, J. M. and Sanger, J. W. (1992). Incorporation of fluorescently labeled contractile proteins into freshly isolated living adult cardiac myocytes. Cell. Motil.
Cytoskel. 21, 111-122.
McHugh, K. M., Crawford, K. and Lessard, J. L. (1991). A comprehensive analysis of the developmental and tissue-specific expression of the isoactin multigene family in the rat. Dev. Biol. 148, 442-458.
McKenna, N., Meigs, J. B. and Wang, Y. L. (1985). Identical distribution of fluorescently labeled brain and muscle actins in living cardiac fibroblasts and myocytes. J. Cell Biol. 100, 292-296.
Miettinen, P. J., Ebner, R., Lopez, A. R. and Derynk, R. (1994). TGF-β induced transdifferentiation of mammary epithelial cells to mesenchymal cells: involvement of type I receptors. J. Cell Biol. 127, 2021-2036.
Mounier, N., Gouy, M., Mouchiroud, D. and Prudhomme, J. C. (1992).
Insect muscle actins differ distinctly from invertebrate and vertebrate cytoplasmic actins. J. Mol. Evol. 34, 406-415.
Mounier, N. and Sparrow, J. C. (1993). Muscle actin genes in insects. Comp.
Biochem. Physiol. 105, 231-238.
North, A. J., Gimona, M., Lando, Z. and Small, J. V. (1994). Actin isoform compartments in chicken gizzard smooth muscle cells. J. Cell Sci. 107, 445- 455.
Otey, C. A., Kalnoski, M. H. and Bulinski, J. C. (1987). Identification and quantification of actin isoforms in vertebrate cells and tissues. J. Cell.
Biochem. 34, 113-124.
Owens, G. K. (1995). Regulation of differentiation of vascular smooth muscle cells. Physiol. Rev. 75, 487-517.
Rubenstein, P. A. (1990). The functional importance of multiple actin isoforms. BioEssays 12, 309-315.
Schevzov, G., Lloyd, C. and Gunning, P. (1992). High level expression of transfected beta- and gamma-actin genes differentially impacts on myoblast cytoarchitecture. J. Cell Biol. 117, 775-785.
Sheterline, P. and Sparrow, J. C. (1994). Actin. Protein Profile. 1, 1-121.
Skalli, O., Bloom, W. S., Ropraz, P., Azzarone, B. and Gabbiani, G. (1986a).
Cytoskeletal remodeling of rat aortic smooth muscle cells in vitro:
relationships to culture conditions and analogies to in vivo situations. J.
Submicrosc. Cytol. 18, 481-493.
Skalli, O., Ropraz, P., Trzeciak, A., Benzonana, G., Gillessen, D. and Gabbiani, G. (1986b). A monoclonal antibody against alpha-smooth muscle actin: a new probe for smooth muscle differentiation. J. Cell Biol. 103, 2787- 2796.
Skalli, O., Vandekerckhove, J. and Gabbiani, G. (1987). Actin-isoform pattern as a marker of normal or pathological smooth-muscle and fibroblastic tissues. Differentiation 33, 232-238.
Soldati, T. and Perriard, J. C. (1991). Intracompartmental sorting of essential myosin light chains: molecular dissection and in vivo monitoring by epitope tagging. Cell 66, 277-289.
Spudich, J. A. and Watt, S. (1971). The regulation of rabbit skeletal muscle contraction. I. Biochemical studies of the interaction of the troponin- tropomyosin complex with actin and the proteolytic fragments of myosin. J.
Biol. Chem. 246, 4866-4880.
Vandekerckhove, J. and Weber, K. (1978). Mammalian cytoplasmic actins are the products of at least two genes and differ in primary structure in at least 25 identified positions from skeletal muscle actins. Proc. Nat. Acad. Sci. USA 75, 1106-1110.
Vandekerckhove, J. and Weber, K. (1981). Actin typing on total cellular extracts. A highly sensitive protein-chemical procedure able to distinguish different actins. Eur. J. Biochem. 113, 595-603.
von Arx, P., Bantle, S., Soldati, T. and Perriard, J. C. (1995). Dominant negative effect of cytoplasmic actin isoproteins on cardiomyocyte cytoarchitecture and function. J. Cell Biol. 131, 1759-1773.
Yao, X. B., Chaponnier, C., Gabbiani, G. and Forte, J. G. (1995). Polarized distribution of actin isoforms in gastric parietal cells. Mol. Biol. Cell 6, 541- 557.
(Received 10 June 1996 – Accepted, in revised form, 31 January 1997)
N. Mounier and others
