Identical Distribution of Fluorescently Labeled Muscle Actins in Living Cardiac Fibroblasts and Brain and Myocytes

We have investigated whether living muscle and nonmuscle cells can discriminate between microinjected muscle and nonmuscle actins. Muscle actin purified from rabbit back and leg muscles and labeled with fluorescein isothiocyanate, and nonmuscle actin purified from lamb brain and labeled with lissamine rhodamine B sulfonyl chloride, were co-injected into chick embryonic cardiac myocytes and fibroblasts. When fluorescence images of the two actins were compared using filter sets selective for either fluorescein isothiocyanate or lissamine rhodamine B sulfonyl chloride, essentially identical patterns of distribution were detected in both muscle and nonmuscle cells. In particular, we found no structure that, at this level of resolution, shows preferential binding of muscle or nonmuscle actin. In fibroblasts, both actins are associated primarily with stress fibers and ruffles. In myocytes, both actins are localized in sarcomeres. In addition, the distribution of structures containing microinjected actins is similar to that of structure containing endogenous F-actin, as revealed by staining with fluorescent phalloidin or phallacidin. Our results suggest that, at least under these experimental conditions, actin-binding sites in muscle and nonmuscle cells do not discriminate among different forms of actins. The actin family consists of multiple polypeptides that vary slightly in amino acid sequences (1). In muscle cells, the predominant isoform is a-actin; in nonmuscle cells, ¢~ and actins predominate (2, 3). These isoforms are similar biochemically (4). However, even across phylogenetically disparate species, the nonmuscle actins are more similar to each other than they are to a-actin (5, 6). It is, therefore, important to determine whether the different forms of actin have different functions in cellular processes. One of the most direct ways to examine possible differences in function is to microinject fluorescently labeled muscle and nonmuscle actins into living cells (7). Muscle actin, which was the first cytoskeletal protein to be fluorescently labeled and microinjected (8), can be incorporated into normal actincontaining structures in many cell types, including gizzard cells (9), fibroblasts (10, l l), ameba (7), and macrophages (12). Participation of injected actin in the assembly of de novo structures, and in the reorganization of existing structures had also been documented (13). Thus, labeled muscle actin is considered to be an accurate tracer of endogenous pools of actin (8-13). However, it is still not clear whether muscle actin actually behaves in a way identical to nonmuscle actin in nonmuscle cells. Nor is it known whether nonmuscle actins can be utilized by muscle cells. The experiments described in this paper ask whether embryonic chick cardiac fibroblasts and myocytes can distinguish between co-injected muscle and nonmuscle actins. Since the muscle and nonmuscle actins are co-injected into the same cell, comparison of the fluorescent images should reveal the presence of structures that preferentially bind one or the other. In addition, we examine whether injected muscle and nonmuscle actins participate in all F actin-containing structures or whether either actin is excluded from some specific structures. MATERIALS AND METHODS Cell Culture, Microinjection, and Microscopy: Monolayer cultures of cardiac fibroblasts and myocytes were obtained by trypsinizing hearts of 7-d chick embryos (14). Cells were injected 24 h to 20 d after plating as described by Wang (13). Many myocytes continued to beat during and after injection~ Approximately 5-20% of the cell volume were injected. Thus, the concentration o f injected actin per cell was between 0,15 and 1.0 mg/ml. No variations in the distribution ofac t in were detected within this range. Injection volumes much smaller than 5% resulted in very faint structures that were difficult to resolve and photograph, and large volumes frequently caused cell damage. Phallotoxin staining was performed as described by Amato et al. (12). Fluorescent images of injected cells and phalloidinor phallacidin-stained cells were observed using epi-illumination and a 63 × oil immersion objective NA 1.25 or a 100 x oil immersion objective NA 1.30. Filters with narrow bandwidths were used so that no crossover of fluorescence was observable in THE JOURNAL OF CELL BIOLOGY • VOLUME 100 JANUARY 1985 292 296 292 © The Rockefeller University Press . 0021-9525/85/01/0292/05 $1.00 on Jauary 0, 2018 jcb.rress.org D ow nladed fom experiments involving comparisons of different fluorophores. Images were detected by a Venus DV-2 image intensifier, and photographed off a video monitor. Preparation of Fluorescent Analogues: Muscle actin was purified from rabbit back and leg muscles according to Spudich and Watt (15). Nonmuscle actin was purified from lamb brain according to Ruscha and Himes (16). Lissamine rhodamine B sulfonyl chloride (LRBJ; Molecular Probes, Junction City, OR) or fluorescein isothiocyanate (FITC; Research Organics, Cleveland, OH) was dissolved in 100 mM borate, 100 mM KCI, 0.4 mM MgCI2, pH 8.8 (LRB), or pH 9.0 (FITC). An equal volume of F actin in 2 mM Tris, 100 mM KCI, 2 mM MgCI2, 0.2 mM CaCI2, 0.2 mM ATP, 0.05 mM dithiothreitol, pH 8.0 was added to obtain a final dye to protein mass ratio of 0.09 (LRB) or 0.9 (FITC). The LRB-actin solution was stirred at room temperature for 15 min; the FITC-actin solution was incubated 2 h at room temperature. Subsequent procedures were essentially as described previously (8). The final actin conjugate had a dye to protein molar ratio of 1.0 to 2.5 (LRB) or 0.7 (FITC), estimated using a molar extinction coefficient of 13,000 at 550 nm for LRB, and a molar extinction coefficient of 64,000 at 492 nm for FITC. The range of the final ratios did not affect our results. Before injection, the actin conjugates at a concentration of 3-5 mg/ml were dialyzed against 0.5 mM PIPES, 0.05 mM MgCI2, 0.2 mM ATP, 0.l mM dithiothreitol, pH 6.95, then clarified. Protein purity and absence of unbound dye were determined by gel electrophoresis. LRB-ovalbumin (Sigma Chemical Co., St. Louis, MO) was prepared as described by Amain et al (12).