The structure of nonvertebrate actin: implications for the ATP hydrolytic mechanism

The structures of Saccharomyces cerevisiae, Dictyostelium, and Caenorhabditis elegans actin bound to gelsolin segment-1 have been solved and refined at resolutions between 1.9 and 1.75 A. These structures reveal several features relevant to the ATP hydrolytic mechanism, including identification of the nucleophilic water and the roles of Gln-137 and His-161 in positioning and activating the catalytic water, respectively. The involvement of these residues in the catalytic mechanism is consistent with yeast genetics studies. This work highlights both structural and mechanistic similarities with the small and trimeric G proteins and restricts the types of mechanisms responsible for the considerable enhancement of ATP hydrolysis associated with actin polymerization. The conservation of functionalities involved in nucleotide binding and catalysis also provide insights into the mechanistic features of members of the family of actin-related proteins.

F-actin-like ATPase activity in a polymerization-defective mutant yeast actin (V266G/L267G)

Polymerization increases a low level G-actin ATPase activity yielding ADP-P(i) F-actin and then ADP F-actin following release of P(i). By monitoring P(i) release, we explored the relationship between the ATPase activity and polymerization characteristics of a mutant yeast actin, GG. In this mutant, two hydrophobic residues at the tip of a proposed hydrophobic plug between actin subdomains 3 and 4, Val(266) and Leu(267), were mutated to Gly. Although GG-actin does not polymerize by itself in vitro, GG cells are viable. We show that GG-actin ATPase activity increases under normal polymerization conditions, although stable filaments do not form. A plot of P(i) release rate versus actin concentration yields an apparent critical concentration, like that seen for actin polymerization, of approximately 8 microm for Mg(2+) GG-actin and 11 microm for Ca(2+) GG-actin. In contrast to WT-actin, P(i) release from GG-actin is cold-sensitive, reflecting the temperature sensitivity associated with mutations that decrease hydrophobicity in this region. Thus, under polymerization conditions, GG-actin exhibits a continuous F-actin-like ATPase activity resulting from the temperature-sensitive formation of unstable cycling F-actin oligomers. Tropomyosin limits the extent and rate of this activity and restores polymerization by capturing and stabilizing these oligomers rather than enhancing filament nucleation.

Tropomyosin-dependent filament formation by a polymerization-defective mutant yeast actin (V266G,L267G)

A major function of tropomyosin (TPM) in nonmuscle cells may be stabilization of F-actin by binding longitudinally along the actin filament axis. However, no clear evidence exists in vitro that TPM can significantly affect the critical concentration of actin. We previously made a polymerization-defective mutant actin, GG (V266G, L267G). This actin will not polymerize alone at 25 degrees C but will in the presence of phalloidin or beryllium fluoride. With beryllium fluoride, but not phalloidin, this polymerization rescue is cold-sensitive. We show here that GG-actin polymerizability was restored by cardiac tropomyosin and yeast TPM1 and TPM2 at 25 degrees C with rescue efficiency inversely proportional to TPM length (TPM2 > TPM1 > cardiac tropomyosin), indicating the importance of the ends in polymerization rescue. In the presence of TPM, the apparent critical concentration of actin is 5.5 microm, 10-15-fold higher than that of wild type actin but well below that of the GG-actin alone (>20 microm). Non N-acetylated TPMs did not rescue GG-actin polymerization. The TPMs did not prevent cold-induced depolymerization of GG F-actin. TPM-dependent GG-actin polymerization did not occur at temperatures below 20 degrees C. Polymerization rescue may depend initially on the capture of unstable GG-F-actin oligomers by the TPM, resulting in the strengthening of actin monomer-monomer contacts along the filament axis.

Interaction of the gonococcal porin P.IB with G- and F-actin

The invasion of epithelial cells by N. gonorrheae is accompanied by formation of a halo of actin filaments around the enveloped bacterium. The transfer of the bacterial major outer membrane protein, porin, to the host cell membrane during invasion makes it a candidate for a facilitator for the formation of this halo. Western analysis shows here that gonococcal porin P.IB associates with the actin cytoskeleton in infected cells. Using the pyrene-labeled Mg forms of yeast and muscle actins, we demonstrate that under low ionic strength conditions, P.IB causes formation of filamentous actin assemblies, although they, unlike F-actin, cannot be internally cross-linked with N,N'-4-phenylenedimaleimide (PDM). In F-buffer, low porin concentrations appear to accelerate actin polymerization. Higher P.IB concentrations lead to the formation of highly decorated fragmented F-actin-like filaments in which the actin can be cross-linked by PDM. Co-assembly of P.IB with a pyrene-labeled mutant actin, S(265)C, prevents formation of a pyrene excimer present with labeled S(265)C F-actin alone. Addition of low concentrations of porin to preformed F-actin results in sparsely decorated F-actin. Higher P.IB concentrations extensively decorate the filaments, thereby altering their morphology to a state like that observed when the components are copolymerized. With preformed labeled S(265)C F-actin, P.IB quenches the pyrene excimer. This decrease is prevented by the F-actin stabilizers phalloidin and to a lesser extent beryllium fluoride. P.IB's association with the actin cytoskeleton and its ability to interact with and remodel actin filaments support a direct role for porin in altering the host cell cytoskeleton during invasion.

Cross-linking constraints on F-actin structure

The DNase I binding loop (residues 38-52), the hydrophobic plug (residues 262-274), and the C terminus region are among the structural elements of monomeric (G-) actin proposed to form the intermonomer interface in F-actin. To test the proximity and interactions of these elements and to provide constraints on models of F-actin structure, cysteine residues were introduced into yeast actin either at residue 41 or 265. These mutations allowed for specific cross-linking of F-actin between C41 and C265, C265 and C374, and C41 and C265 using dibromobimane and disulfide bond formation. The cross-linked products were visualized on SDS-PAGE and by electron microscopy. Model calculations carried out for the cross-linked F-actins revealed that considerable flexibility or displacement of actin residues is required in the disulfide cross-linked segments to fit these filaments into model F-actin structures. The calculated, cross-linked structures showed a better fit to the Holmes rather than the refined Lorenz model of F-actin. It is predicted on the basis of such calculations that image reconstruction of electron micrographs of disulfide cross-linked C41-C374 F-actin should provide a conclusive test of these two similar models of F-actin structure.

Structural transition at actin's N-terminus in the actomyosin cross-bridge cycle

Force and motion generation by actomyosin involves the cyclic formation and transition between weakly and strongly bound complexes of these proteins. Actin's N-terminus is believed to play a greater role in the formation of the weakly bound actomyosin states than in the formation of the strongly bound actomyosin states. It has been the goal of this project to determine whether the interaction of actin's N-terminus with myosin changes upon transition between these two states. To this end, a yeast actin mutant, Cys-1, was constructed by the insertion of a cysteine residue at actin's N-terminus and replacement of the C-terminal cysteine with alanine. The N-terminal cysteine was labeled stoichiometrically with pyrene maleimide, and the properties of the modified mutant actin were examined prior to spectroscopic measurements. Among these properties, actin polymerization, strong S1 binding, and the activation of S1 ATPase by pyrenyl-Cys-1 actin were not significantly different from those of wild-type yeast actin, while small changes were observed in the weak S1 binding and the in vitro motility of actin filaments. Fluorescence changes upon binding of S1 to pyrenyl-Cys-1 actin were measured for the strongly (with or without ADP) and weakly (with ATP and ATPgammaS) bound acto-S1 states. The fluorescence increased in each case, but the increase was greater (by about 75%) in the presence of MgATP and MgATPgammaS than in the rigor state. This demonstrates a transition at the S1 contact with actin's N-terminus between the weakly and strongly bound states, and implies either a closer proximity of the pyrene probe on Cys-1 to structural elements on S1 (most likely the loop of residues 626-647) or greater S1-induced changes at the N-terminus of actin in the weakly bound acto-S1 states.

His(73), often methylated, is an important structural determinant for actin. A mutagenic analysis of HIS(73) of yeast actin

His(73), has been proposed to regulate the release of P(i) from the interior of actin following polymerization-dependent hydrolysis of bound ATP. Although it is a 3-methylhistidine in the vast majority of actins, His(73) is unmethylated in S. cerevisiae actin. We mutated His(73) in yeast actin to Arg, Lys, Ala, Gln, and Glu and detected no altered phenotypes associated with the mutations in vivo. However, they significantly affect actin function in vitro. Substitution of the more basic residues resulted in enhanced thermal stability, decreased rate of nucleotide exchange, and decreased susceptibility to controlled proteolysis relative to wild-type actin. The opposite effects are observed with the neutral and anionic substitutions. All mutations reduced the rate of polymerization. Molecular dynamics simulations predict a new conformation for the His(73) imidazole in the absence of a methyl group. It also predicts that Arg(73) tightens and stabilizes the actin and that Glu(73) causes a rearrangement of the bottom of actin's interdomain cleft leading possibly to our observed destabilization of actin. Considering the exterior location of His(73), this work indicates a surprisingly important role for the residue as a major structural determinant of actin and provides a clue to the impact caused by methylation of His(73).

Interaction in vivo and in vitro between the yeast fimbrin, SAC6P, and a polymerization-defective yeast actin (V266G and L267G)

A mutant yeast actin (GG) has decreased hydrophobicity in a subdomain 3/4 hydrophobic plug believed to be involved in a hydrophobic cross-strand "plug-pocket" interaction necessary for actin filament stability. This actin will not polymerize in vitro but is compatible with cell viability. We have assessed the ability of Sac6p, the yeast homologue of the actin filament stabilizing and bundling protein fimbrin, to restore polymerization in vitro and to facilitate GG-actin function in vivo. Sac6p rescues GG-actin polymerization at 25 degrees C but not at 4 degrees C. The actin polymerizes into bundles at room temperature with a fimbrin:actin molar ratio of 1:4. At this ratio, every actin monomer contacts a Sac6p actin binding domain. Following cold-induced depolymerization, actin/Sac6p mixtures repolymerize beginning at 15 degrees C instead of the 25 degrees C required for de novo assembly, because of the presence of residual actin-Sac6p nuclei. Generation of haploid Deltasac6/GG-actin cells from either diploid or haploid cells was unsuccessful. The facile isolation of cells with either mutation alone indicates a synthetic lethal relationship between this actin allele and the SAC6 gene. Sac6p may allow GG-actin function in vivo by stabilizing the actin in bundles thereby helping maintain sufficient levels of an otherwise destabilized actin monomer within the cell.

A highly conserved 3-methylhistidine modification is absent in yeast actin

To identify a protein histidine methyltransferase from Saccharomyces cerevisiae, we examined purified actin for the presence of the highly conserved 3-methylhistidine residue at position 73 by amino acid analysis of the whole protein and by amino acid analysis and mass spectrometry of the corresponding tryptic fragment. Surprisingly, we found that His-73 is not modified. A similar lack of modification was also found in actin from the yeast Candida albicans, while rabbit muscle actin revealed the expected 3-methylhistidine residue. Phylogenetic analysis of actin sequences suggests that this modification was introduced in evolution after the divergence of yeast from higher eukaryotic organisms, including unicellular eukaryotes such as Acanthamoeba, Dictyostelium, and Physarum, whose actins contain 3-methylhistidine. Our methodology for the analytical determination of 3-methylhistidine in actin offers an improved approach for investigating histidine methylation in proteins.

Structure determination and characterization of Saccharomyces cerevisiae profilin

The structure of profilin from the budding yeast Saccharomyces cerevisiae has been determined by X-ray crystallography at 2.3 A resolution. The overall fold of yeast profilin is similar to the fold observed for other profilin structures. The interactions of yeast and human platelet profilins with rabbit skeletal muscle actin were characterized by titration microcalorimetry, fluorescence titrations, and nucleotide exchange kinetics. The affinity of yeast profilin for rabbit actin (2.9 microM) is approximately 30-fold weaker than the affinity of human platelet profilin for rabbit actin (0.1 microM), and the relative contributions of entropic and enthalpic terms to the overall free energy of binding are different for the two profilins. The titration of pyrene-labeled rabbit skeletal actin with human profilin yielded a Kd of 2.8 microM, similar to the Kd of 2.0 microM for the interaction between yeast profilin and pyrene-labeled yeast actin. The binding data are discussed in the context of the known crystal structures of profilin and actin, and the residues present at the actin-profilin interface. The affinity of yeast profilin for poly-L-proline was determined from fluorescence measurements and is similar to the reported affinity of Acanthamoeba profilin for poly-L-proline. Yeast profilin was shown to catalyze adenine nucleotide exchange from yeast actin almost 2 orders of magnitude less efficiently than human profilin and rabbit skeletal muscle actin. The in vivo and in vitro properties of yeast profilin mutants with altered poly-L-proline and actin binding sites are discussed in the context of the crystal structure.

Modulation of yeast F-actin structure by a mutation in the nucleotide-binding cleft

Although the actin sequence is very highly conserved across evolution, tissue-specific expression of different isoforms in high eukaryotes suggests that different isoforms carry out different functions. However, little information exists about either the differences in filaments made from different actins or the effects on filament structure caused by the various mutations in actin that have been introduced to gain insight into actin function. Using electron microscopy and three-dimensional reconstruction, we have studied the differences in the filaments made by yeast and rabbit skeletal muscle actin, two proteins with 88% homologous sequences, and we have assessed the changes in filament structure caused by the introduction of the S14A mutation into yeast actin. Elimination of the S14 hydroxyl group, assumed to bind to the gamma-phosphate of actin-bound ATP, results in a 40 to 60-fold decrease in actin's affinity for ATP. We show that yeast actin displays less extensive contacts between the two long-pitch helical strands than does muscle actin, and displays the large cooperativity within filaments previously observed for muscle actin. Finally, we demonstrate that the S14A mutation narrows the cleft between the two lobes of the actin subunit and strengthens the inter-strand connections in F-actin.

Fluorescence probing of yeast actin subdomain 3/4 hydrophobic loop 262-274. Actin-actin and actin-myosin interactions in actin filaments

Residues 262-274 form a loop between subdomains 3 and 4 of actin. This loop may play an important role in actin filament formation and stabilization. To assess directly the behavior of this loop, we mutated Ser265 of yeast actin to cysteine (S265C) and created another mutant (S265C/C374A) by changing Cys374 of S265C actin to alanine. These changes allowed us to attach a pyrene maleimide stoichiometrically to either Cys374 or Cys265. These mutations had no detectable effects on the protease susceptibility, intrinsic ATPase activity, and thermal stability of labeled or unlabeled G-actin. The presence of the loop cysteine, either labeled or unlabeled, did not affect the actin-activated S1 ATPase activity or the in vitro motility of the actin. Both mutant actins, either labeled or unlabeled, nucleated filament formation considerably faster than wild-type (WT) actin, although the critical concentration was not affected. Whereas the fluorescence of the C-terminal (WT) probe increased during polymerization, that of the loop (S265C/C374A) probe decreased, and the fluorescence of the doubly labeled actin (S265C) was approximately 50% less than the sum of the fluorescence of the individual fluorophores. Quenching was also observed in copolymers of labeled WT and S265C/C374A actins. An excimer peak was present in the emission spectrum of labeled S265C F-actin and in the labeled S265C/C374A-WT actin copolymers. These results show that in the filaments, the C-terminal pyrene of a substantial fraction of monomers directly interacts with the loop pyrene of neighboring monomers, bringing the two cysteine sulfurs to within 18 A of one another. Finally, when bound to labeled S265C/C374A F-actin, myosin S1, but not tropomyosin, caused an increase in fluorescence of the loop probe. Both proteins had no effect on excimer fluorescence. These results help establish the orientation of monomers in F-actin and show that the binding of S1 to actin subdomains 1 and 2 affects the environment of the loop between subdomains 3 and 4.

The effects of severely decreased hydrophobicity in a subdomain 3/4 loop on the dynamics and stability of yeast G-actin

The hydrophobicity of the subdomain 3/4 hydrophobic loop (262-274) has been implicated to be essential for actin's function. We previously showed (Kuang, B., and Rubenstein, P. A. (1997) J. Biol. Chem. 272, 1237-1247) that a mutant yeast actin (V266G/L267G) with markedly decreased hydrophobicity in this loop conferred severe cold sensitivity to its polymerization. Here we further tested the mutational effect on the conformation and function of G-actin. This GG mutation caused no significant changes in overall secondary structure or in the microenvironment around actin's tryptophan residues, nor did it alter the dissociation constant of G-actin for ATP. However, it lowers the intrinsic ATPase activity and the melting temperature for Mg-GG actin from 51 to 33 degrees C and transforms the conformation of subdomain 2 and the central cleft of G-actin into an F-monomer-like structure. The results suggest that the hydrophobic plug may not only play a role in actin filament stabilization but also may be important for controlling the stability of G-actin and for promoting the conformational change of the monomer needed for addition to a growing actin filament.

Beryllium fluoride and phalloidin restore polymerizability of a mutant yeast actin (V266G,L267G) with severely decreased hydrophobicity in a subdomain 3/4 loop

Holmes proposed that in F-actin, hydrophobic residues in a subdomain 3/4 loop interact with a hydrophobic pocket on the opposing strand resulting in helix stabilization. We have determined how a decreased hydrophobicity of this plug affects yeast actin function. Cells harboring only the V266G, V266D, V266F, L267G, L269D, or L269K actins appear normal, although V266G cells display an altered budding pattern. However, V266G,L267G (GG) double mutant cells are cold-sensitive with randomly oriented thick actin assemblies seen in rhodamine phalloidin-stained GG cells. V266D actin polymerizes slower than wild-type actin at room temperature. At 4 degrees C, not only is polymerization slowed, but there is also an effect on critical concentration. However, the polymerization defects are milder than those associated with substitution of Asp for the neighboring Leu267. Purified GG-actin does not polymerize in vitro alone or in the presence of wild-type F-actin seeds. GG-actin polymerization can be restored by larger amounts of wild-type actin, beryllium fluoride, or phalloidin at room temperature, although at 4 degrees C only phalloidin is effective. These results suggest that the diminished hydrophobicity of the plug in GG-actin leads to filament destabilization. However, the V266D actin results require a modification of the original Holmes filament model.

Mutational analysis of the role of the N terminus of actin in actomyosin interactions. Comparison with other mutant actins and implications for the cross-bridge cycle

Yeast actin mutants with acidic residues at the N terminus either neutralized (DNEQ) or deleted (delta-DSE) were used to assess the role of N-terminal acidic residues in the interactions of actin with myosin in the contractile cycle. Cosedimentation experiments revealed an approximately 3-fold decrease in the binding constant for DNEQ and delta-DSE actins to myosin subfragment-1 (S1) relative to that of wild type actin both in the presence of MgATP and in the absence of nucleotides (strong binding). DNEQ and delta-DSE actins protected S1 from tryptic digestion as well as the wild type and rabbit actins. The activation of S1 ATPase by DNEQ and delta-DSE actins (up to 50 microM) was very low but increased greatly after cross-linking these mutant actins to S1 by dimethyl suberimidate. Thus, the increased dissociation of mutant actins from S1 in the presence of ATP is the main cause for the low acto-S1 ATPase activities. At low-ionic strength conditions and in the presence of methylcellulose, the DNEQ and delta-DSE actins moved in the in vitro motility assays at a mean velocity similar to that of wild type actin (3.0 microns/s). Yet, the sliding velocity of the N-terminal and D24A/D25A and E99A/E100A mutant actins decreased relative to that of the wild type at all levels of external load introduced into the assay and at low densities of heavy meromyosin (HMM) on the cover slip. This indicates a lower relative force generation with the mutant actins. In contrast, the force generated under the same conditions with the 4Ac mutant actin (with four acidic charges at the N terminus) was higher than with wild type actin. At higher-ionic strength conditions (I = 150 mM), the sliding of the DNEQ and delta-DSE as well as that of the D24A/D25A and E99A/E100A actins ceased even in the presence of methylcellulose, while I341A actin (deficient in strong binding to myosin) still moved. These results indicate the importance of electrostatic actomyosin interactions under physiological salt conditions and show functionally distinct roles for the different myosin binding sites on actin.

The effect of the S14A mutation on the conformation and thermostability of Saccharomyces cerevisiae G-actin and its interaction with adenine nucleotides

The actin Ser14 hydroxyl is one of a number of ligands that binds to the gamma-phosphate of ATP thereby stabilizing the actin.ATP complex. In yeast actin, conversion of Ser14 to Ala (S14A), causes a temperature-sensitive phenotype in vivo and temperature-sensitive polymerization defects in vitro (Chen, X., and Rubenstein, P. A. (1995) J. Biol. Chem. 270, 11406-11414). Here, using a new luciferase-based procedure, we show that the mutation results in a 40-60-fold decrease in actin's affinity for ATP. The mutation causes a decrease in the intrinsic ATPase activity of both Ca- and Mg-G-actin at 30 degrees C and alters the protease susceptibility of sites on subdomain 2. Ca-S14A-actin but not Mg-S14A-actin binds etheno-ATP at 37 degrees C. Intrinsic tryptophan fluorescence measurements show that at 37 degrees C, Mg-S14A-actin but not the calcium form unfolds. CD measurements show the mutation causes a decrease in the apparent denaturation temperature for Ca-actin from 57 to 45 degrees C and for the magnesium form a decrease from 52 to 40 degrees C. Based on a re-examination of actin's crystal structure coordinates, we propose that the Ser14 hydroxyl forms a polar bridge between the ATP gamma-phosphate and the amide nitrogen of Gly74, thus conferring additional stability on the actin small domain.

A mutation in an ATP-binding loop of Saccharomyces cerevisiae actin (S14A) causes a temperature-sensitive phenotype in vivo and in vitro

The Ser14 hydroxyl group of actin is one of six groups that potentially form hydrogen bonds with the gamma-phosphate of the ATP bound in the cleft separating the two domains of the protein. To understand the importance of this group in actin function, we mutated Ser14 of Saccharomyces cerevisiae actin and studied the effects of these mutations in vivo and in vitro. Substitution of Cys of Gly resulted in cell death. Substitution of Thr for Ser resulted in an actin with wild-type properties in vivo and in vitro. Cells carrying the Ser14-->Ala (S14A) mutation were viable but displayed a temperature sensitive lethality at 37 degrees C preceded by delocalization of actin patches, the appearance of bar-like structures, and finally the disappearance of identifiable actin structures. The mutation caused no effect on the critical concentration of polymerization but resulted in an actin with an increased rate of polymerization, an altered protease susceptibility, and a decreased filament ATPase activity. At 37 degrees C, Mg-, but not Ca-S14A-actin irreversibly lost the ability to polymerize. These results demonstrate the importance of the ATP-Ser14 hydroxyl hydrogen bond in regulating actin function in vivo and in vitro and the magnification of the effects of the mutation when Mg2+ is substituted for Ca2+ in the protein.

Caldesmon, N-terminal yeast actin mutants, and the regulation of actomyosin interactions

N-Terminal yeast actin mutants were used to assess the role of N-terminal acidic residues in the interactions of caldesmon with actin. The yeast actins differed only in their N-terminal charge: wild type, two negative charges; 4Ac, four negative charges; DNEQ, neutral charge; delta DSE, one positive charge. Caldesmon inhibition of actomyosin subfragment 1 ATPase was affected by alterations in the N-terminus of actin. This inhibition was similar for skeletal muscle alpha-actin and the yeast 4Ac and wild-type actins (80%), but much smaller for the neutral and deletion mutants (15%). However, cosedimentation experiments revealed similar binding of caldesmon to polymerized rabbit skeletal muscle alpha-actin and each yeast actin. This result shows that the N-terminal acidic residues of actin are not required for the binding of caldesmon to F-actin. Caldesmon-actin interactions were also examined by monitoring the polymerization of G-actin induced by caldesmon. Although the final extent of polymerization was similar for all actins tested, the rates of polymerization differed. Skeletal muscle and 4Ac actins had similar rates of polymerization, and the wild-type actin polymerized at a slower rate. The neutral and deletion mutants had even slower rates of polymerization by caldesmon. The slow polymerization of DNEQ G-actin was traced to a greatly reduced binding of caldesmon to this mutant G-actin when compared to wild-type and alpha-actin. MgCl2-induced actin polymerization proceeded at identical rates for all actins.(ABSTRACT TRUNCATED AT 250 WORDS)

Yeast actin with a mutation in the "hydrophobic plug" between subdomains 3 and 4 (L266D) displays a cold-sensitive polymerization defect

Holmes et al. (Holmes, K. C., D. Popp, W. Gebhard, and W. Kabsch. 1990. Nature [Lond.] 347: 44-49) hypothesized that between subdomains 3 and 4 of actin is a loop of 10 amino acids including a four residue hydrophobic plug that inserts into a hydrophobic pocket formed by two adjacent monomers on the opposing strand thereby stabilizing the F-actin helix. To test this hypothesis we created a mutant yeast actin (L266D) by substituting Asp for Leu266 in the plug to disrupt this postulated hydrophobic interaction. Haploid cells expressing only this mutant actin were viable with no obvious altered phenotype at temperatures above 20 degrees C but were moderately cold-sensitive for growth compared with wild-type cells. The critical concentration for polymerization increased 10-fold at 4 degrees C compared with wild-type actin. The length of the nucleation phase of polymerization increased as the temperature decreased. At 4 degrees C nucleation was barely detectable. Addition of phalloidin-stabilized F-actin nuclei and phalloidin restored L266D actin's ability to polymerize at 4 degrees C. This mutation also affects the overall rate of elongation during polymerization. Small effects of the mutation were observed on the exchange rate of ATP from G-actin, the G-actin intrinsic ATPase activity, and the activation of myosin S1 ATPase activity. Circular dichroism measurements showed a 15 degrees C decrease in melting temperature for the mutant actin from 57 degrees C to 42 degrees C. Our results are consistent with the model of Holmes et al. (Holmes, K. C., D. Popp, W. Gebhard, and W. Kabsch. 1990. Nature [Lond.]. 347:44-49) involving the role of the hydrophobic plug in actin filament stabilization.

Enhanced stimulation of myosin subfragment 1 ATPase activity by addition of negatively charged residues to the yeast actin NH2 terminus

We examined the effects of yeast actin NH2-terminal mutations on actomyosin interactions and the function of actin in vivo through measurements of actin-activated ATPase activity, cosedimentation with rabbit muscle myosin subfragment 1 (S-1), in vitro motility, and invertase secretion assays. As reported earlier (Cook, R. K., Blake, W., and Rubenstein, P. A. (1992) J. Biol. Chem. 267, 9430-9436), elimination of NH2-terminal acidic residues from yeast actin results in an increased actin bundling, decreased actin-activated S-1 ATPase, and complete inhibition of actin filament sliding over myosin. Here we show that the addition of 2 new acidic residues to the NH2 terminus of yeast actin increased the Vmax value and the catalytic efficiency of the actin-activated ATPase activity of S-1. However, the binding of actin to S-1 in the presence of ATP and the velocities of actin sliding over myosin in the in vitro motility assays were not affected by this mutation. Thus, the number of actin NH2-terminal negative charges is important for actin activation of myosin S-1 ATPase activity, while only a minimum number of acidic residues is required for actin sliding over myosin in vitro. The number of actin NH2-terminal negative charges therefore appears to determine the efficiency with which the energy from ATP hydrolysis is converted to filament sliding.

Control of profilin and actin expression in muscle and nonmuscle cells

Profilin is a small G-actin binding protein implicated in sequestering actin monomers in vivo. We have quantitated profilin and actin expression in human hepatoma HepG-2 cells and in two mouse myogenic cell lines, BC3H1 and C2C12, to determine whether the expression of profilin and the expression of nonmuscle isoactin or total actin are co-regulated. During differentiation of both muscle cell types, profilin and nonmuscle actin expression decrease in a coordinate manner as shown by measurements of steady state mRNA and newly synthesized protein. In human hepatoma HepG-2 cells, the twofold increase in actin synthesis observed after 24 hours of exposure to cytochalasin D did not result in an increase in profilin synthesis. Thus, profilin and actin expression are not co-regulated in all cells. To determine if there is sufficient profilin to sequester a large portion of cellular G-actin, we measured total profilin and G-actin levels in the three cell types. In each case, profilin accounted for less than 10% of the total G-actin on a molar basis. Thus, profilin is not responsible for total G-actin sequestration in these cells. Finally, using poly-L-proline affinity chromatography, we showed that, in the cell types tested, less than 20% of the poly-L-proline purified profilin existed as a complex with G-actin. The profilin in these cells may be interacting with cellular components other than actin.

Isolation and characterization of the rat liver actin N-acetylaminopeptidase

Actins from most eukaryotes undergo a unique post-translational modification of the amino terminus called "processing." Processing consists of the removal of an amino-terminal Ac-Met or Ac-Cys to leave an acidic amino-terminal residue. We have previously demonstrated that this reaction is not catalyzed by the ribosomally associated methionine aminopeptidase or by other previously described acetylaminopeptidases. Here we present the isolation and characterization of the actin N-acetylaminopeptidase (ANAP) from rat liver. A five-step purification protocol achieves a 4100-fold purification of the enzyme with an overall 8% recovery of activity. ANAP is a 77-kDa monomer with a pI of 4.6. Using unprocessed yeast actin as a substrate, the Km of ANAP is 3.5 microM. Purified ANAP was used to generate a polyclonal antibody. The antibody has been used along with activity assays to demonstrate the presence of ANAP in a variety of rat tissues. Finally, evidence is presented that in mammals, ANAP may function with a second, as yet unpurified, component to process actin amino termini.

Splicing of two alternative exon pairs in beta-tropomyosin pre-mRNA is independently controlled during myogenesis

Two known tissue-specific tropomyosin (TM) isoforms are produced from the rodent beta-TM gene. Skeletal muscle beta-TM uses the alternative exons 6b and 9a and the exon 9a-associated poly(A) site. Fibroblast and smooth muscle TM-1 use exons 6a and 9b and the exon-9b associated poly(A) site. We have identified a new skeletal muscle beta-TM isoform, beta-TM2. beta-TM2 contains exon 6b (muscle) and exon 9b (nonmuscle). Full-length beta-TM2 cDNA clones were isolated from a cDNA library of mouse muscle BC3H1 cells. Its mRNA was also found in mouse skeletal muscle tissue but not in other tissues. beta-TM2 mRNA level and protein synthesis are differentiation-dependent, with a transient high level in the early stages of myogenesis both in BC3H1 cells and in mouse embryo limbs. Trace amounts of beta-TM3 mRNA, the other hybrid form (exons 6a + 9a), were found in less differentiated BC3H1 cells, mouse uterus, heart, and 3T3 fibroblasts but not skeletal muscle tissue. Thus, the selection of the two alternative exons appears to be controlled independently. Furthermore, during myogenesis, there is a sequential switch in the internal alternative exon, the terminal exon, and the poly(A) site from the nonmuscle to the muscle type.

Effects of profilin and profilactin on actin structure and function in living cells

Previous studies have yielded conflicting results concerning the physiological role of profilin, a 12-15-kD actin- and phosphoinositide-binding protein, as a regulator of actin polymerization. We have addressed this question by directly microinjecting mammalian profilins, prepared either from an E. coli expression system or from bovine brain, into living normal rat kidney (NRK) cells. The microinjection causes a dose-dependent decrease in F-actin content, as indicated by staining with fluorescent phalloidin, and a dramatic reduction of actin and alpha-actinin along stress fibers. In addition, it has a strong inhibitory effect toward the extension of lamellipodia. However, the injection of profilin causes no detectable perturbation to the cell-substrate focal contact and no apparent depolymerization of filaments in either the nonlamellipodial circumferential band or the contractile ring of dividing cells. Furthermore, cytokinesis of injected cells occurs normally as in control cells. In contrast to pure profilin, high-affinity profilin-actin complexes from brain induce an increase in total cellular F-actin content and an enhanced ruffling activity, suggesting that the complex may dissociate readily in the cell and that there may be multiple states of profilin that differ in their ability to bind or release actin molecules. Our results indicate that profilin and profilactin can function as effective regulators for at least a subset of actin filaments in living cells.

Removal of the amino-terminal acidic residues of yeast actin. Studies in vitro and in vivo

We have examined the role of the acidic residues Asp2 and Glu4 at the NH2 terminus of Saccharomyces cerevisiae actin through site-directed mutagenesis. In DNEQ actin, these residues have been changed to Asn2 and Gln4, whereas in delta DSE actin, the Asp2-Ser-Glu tripeptide has been deleted. Both mutant actins can replace wild type yeast actin. Peptide mapping studies reveal that DNEQ, like wild type actin, retains the initiator Met and is NH2 terminally acetylated, whereas delta DSE has a free NH2 terminus and has lost the initiator Met. Interestingly, microscopic examination of filaments of these two actins reveal the appearance of bundled filaments. The DNEQ bundles are smaller and more ordered, whereas the delta DSE bundles are larger and more loosely organized. Additionally, both mutant actins activate the ATPase activity of rabbit muscle myosin S1 fragment to a lesser extent than wild type. We have also developed a sensitive assay for actin function in vivo that enabled us to detect a slight defect in the ability of these mutant actins to support secretion, an important function in yeast. Thus, although the mutant actins resulted in no gross phenotypic changes, we were able to detect a defect in actin function through this assay. From these studies we can conclude that 1) although NH2-terminal negative charges are not essential to yeast life, the loss of such charges does result in a slight defect in the actins' ability to support secretion, 2) removal of the NH2-terminal negative charges promotes the bundling of actin filaments, and 3) actins lacking NH2-terminal negative charges are unable to activate the myosin S1 ATPase activity as well as wild type actin.

Amino-terminal processing of actins mutagenized at the Cys-1 residue

Most actins examined to date undergo a unique posttranslational modification termed processing, catalyzed by the actin N-acetylaminopeptidase. Processing is the removal of acetylmethionine from the amino terminus in class I actins with Met-Asp(Glu) amino termini. For class II actins with Met-X-Asp(Glu) amino termini, processing is the removal of the second residue as an N-acetylamino acid. Other cytosolic proteins with these amino termini are not processed suggesting that the reaction may be specific for actins. In actin, X is usually cysteine. However, there are some class II actins in which this residue is other than cysteine, suggesting a broader substrate specificity for actin N-acetylaminopeptidase than acetylmethionine or acetylcysteine. We constructed mutant actins in which this cysteine was replaced with serine, asparagine, glycine, aspartic acid, histidine, phenylalanine, and tyrosine and used these to determine the substrate specificity of rat liver actin N-acetylaminopeptidase in vitro. Amino-terminal acetylmethinonine was cleaved from adjacent aspartic acid, asparagine, or histidine, but not serine, glycine, phenylalanine, or tyrosine. Of the acetylated actin amino termini tested, only acetylmethionine and acetylcysteine were cleaved. Histidine was never N-acetylated and was not cleaved. When phenylalanine and tyrosine were adjacent to the initiator methionine, no initiator methionine was cleaved even though it was acetylated. These results suggest a narrow substrate specificity for the rat liver actin N-acetylaminopeptidase. They also demonstrate that the adjacent residue can effect actin N-acetylaminopeptidase specificity.

Choice of 3' cleavage/polyadenylation site in beta-tropomyosin RNA processing is differentiation-dependent in mouse BC3H1 muscle cells

The rodent beta-tropomyosin (TM) gene produces either a 1.2-kilobase (kb) skeletal muscle beta-TM mRNA or a 1.1-kb fibroblast/smooth muscle TM-1 mRNA through tissue-specific alternative exon splicing and 3' cleavage/polyadenylation at two alternative poly(A) sites. beta-TM mRNA contains exon 6b, 9a, and the poly(A) site immediately following exon 9a, whereas TM-1 mRNA contains exon 6a, 9b, and the poly(A) site following exon 9b. We isolated a novel 2.1-kb beta-TM cDNA clone, pUTM, from a cDNA library of 2-day differentiated mouse BC3H1 muscle-like cells. This cDNA contains the entire sequence of mature beta-TM mRNA with a normal but unused poly(A) site associated with exon 9a. Instead, 3' cleavage/polyadenylation of this cDNA occurred at the exon 9b-associated distal poly(A) site, resulting in the retention of a 1-kb intron and the TM-1 exon 9b. We identified a 2.3-kb functional mRNA, UTM RNA, corresponding to pUTM. UTM RNA appeared early during BC3H1 cell differentiation and gradually decreased as the beta-TM mRNA increased. UTM RNA was also detected in mouse C2C12 muscle cells and in skeletal muscle tissue isolated from mouse leg. Thus, in the processing of beta-TM gene transcripts, selection of alternative terminal exons and alternative poly(A) sites are not necessarily linked as they appear to be in other gene systems.

Unusual metabolism of the yeast actin amino terminus

In this paper we have examined the post-translational modifications of the NH2 terminus of actin from the yeast Saccharomyces cerevisiae. Like actins examined previously, this actin contains an acetylated NH2 terminus. Actins in other organisms undergo a unique post-translational processing event in which the initial amino acid(s) are removed by an actin-specific processing enzyme in an acetylation-dependent reaction. This is defined as actin processing. In yeast, actin retains its initiator Met in vivo and is thus not processed even though a rat liver actin processing enzyme can process yeast actin in vitro. This lack of actin processing appears to be a general property of fungi, as the actin from three other species, Aspergillus nidulans, Schizosaccharomyces pombe, and Candida albicans are not NH2 terminally processed either. Yeast actin is a class I actin; its initiator Met directly precedes an acidic residue. We converted yeast actin to a class II species by inserting a Cys codon between the Met-1 and Asp-2 codons. In normal class II actins the Cys residue is removed as acetyl-Cys during processing. Neither the mutant actin nor chick beta-actin (a class I actin) are processed when expressed in yeast. S. cerevisiae thus appears to be also incapable of processing exogenous actins. Further study of the mutant actin containing a Cys at position 2 shows that 30-40% of this actin is stably unacetylated. This unacetylated actin does not have a shorter half-life than the acetylated form. From these studies we conclude that 1) NH2-terminal actin-specific processing is not required for actin function in yeast and three other fungi, 2) yeast are apparently incapable of processing any type of actin precursor, and 3) the stability of a yeast pseudo-class II actin is not affected by the acetylation state of the NH2 terminus.

The functional importance of multiple actin isoforms

Actin is a protein that plays an important role in cell structure, cell motility, and the generation of contractile force in both muscle and nonmuscle cells. In many organisms, multiple forms of actin, or isoactins, are found. These are products of different genes and have different, although very similar, amino acid sequences. Furthermore, these isoactins are expressed in a tissue specific fashion that is conserved across species, suggesting that their presence is functionally important and their behavior can be distinguished quantitatively from one another in vitro. In muscle cells, they are differentially distributed within the cell and some are specifically associated with structures such as costameres, mitochondria, and neuromuscular junctions. There is also good evidence for specific isoactin function in microvascular pericytes and in the intestinal brush border. However, the necessity of specific isoactins for various functions has not yet been conclusively demonstrated.

Identification of N-acetylmethionine as the product released during the NH2-terminal processing of a pseudo-class I actin

Genes for the various isoactins define two classes of actin. Class I actin genes code for Met-Asp(Glu)-actin, and class II actin genes code for Met-X-Asp(Glu)-actin where X is usually cysteine. Amino termini of both are removed in an acetylation-dependent processing reaction yielding acetyl-Asp(Glu)-actin. Both classes are processed at approximately equal rate (t1/2 = 15 min) in vivo. In vitro, class II actins are 90% processed by endogenous enzymes after 60 min in a rabbit reticulocyte lysate system, whereas class I actins are only minimally processed during this period. Using site-directed mutagenesis of a human skeletal muscle isoactin coupled with in vitro transcription and translation methods, we have synthesized a pseudo-class I actin in which the penultimate cysteine has been changed to an aspartic acid, thus placing a class I amino terminus on an otherwise class II actin molecule. The pseudo-class I actin was less than 20% processed during the translation period as determined by peptide mapping. It was further processed by exogenous processing enzyme at a rate compatible with a class I actin. These results indicate that the major actin determinant controlling differential actin-processing rates is the amino-terminal residue being cleaved, not the remaining structure of the actin molecule. We have also demonstrated for the first time that N-acetylmethionine is the immediately released product from the amino terminus of a pseudo-class I actin during processing.

Synthesis of mammalian profilin in Escherichia coli and its characterization

Profilin is a G-actin binding protein that may have a role in controlling the ratio of G/F actin within the cell. To devise a way for obtaining large amounts of mammalian profilin in an active state, we transfected Escherichia coli with a plasmid containing a full-length rat spleen profilin cDNA adjacent to a promoter inducible by isopropyl thiogalactoside (IPTG). Upon induction, they synthesized a new protein of 15,000 MW constituting approximately 5% of the total cell protein. This protein bound to poly-L-proline Sepharose and could be eluted with 7 M urea, behavior similar to that exhibited by authentic profilin. The protein could be released from the bacteria in soluble form following sonication, and the profilin could then be purified to homogeneity following chromatography on Sephadex G-75 and DEAE A-50 Sephadex. The protein began with an unblocked Ala, indicating that the initiating formyl and methionine residues had been removed. The dissociation of the recombinant profilin from chicken skeletal muscle actin was characterized by a Kd of approximately 2 microM based on gel filtration analysis and actin polymerization assays. These results show that purified active mammalian profilin can be made conveniently in large quantities. This study also demonstrates the feasibility of using bacterially synthesized profilin in structure-function studies involving mutant profilins altered by site-directed mutagenesis.

Nonuniform behavior of multiple isoactins in the same cell is a cell-dependent phenomenon

The functional significance of multiple isoactins in the same cell is still not understood. To address this question, we examined the response of smooth muscle and cardiac muscle alpha-isoactins to a serial extraction procedure applied to both muscle and nonmuscle cell types. We compared these extraction results with results obtained with the beta- and gamma-nonmuscle actin isoforms from the same cells. In differentiated BC3H1 nonfusing muscle cells (smooth muscle alpha-isoactin), in human rhabdomyosarcoma cells (cardiac alpha-isoactin), and in chick skeletal muscle cells (cardiac alpha-isoactin), different fractions were found selectively enriched in either the nonmuscle or the muscle-specific actin isoforms compared with their relative abundance in whole cell extracts. Conversely, when these same isoactins were examined either in undifferentiated BC3H1 cells or in mouse nonmuscle cells stably transfected with a cardiac alpha-isoactin gene, no enrichment of these isoforms above their relative abundance in whole cell extracts was observed. These results indicate that within the muscle or muscle-like cells examined, the different actin isoforms were either selectively utilized or localized. These results further show that isoactin-specific responses observed were apparently related to the cell type in which they were found and not to differences in inherent physical properties such as solubility of the different isoactins examined.

Studies on the role of actin's aspartic acid 3 and aspartic acid 11 using oligodeoxynucleotide-directed site-specific mutagenesis

One or more of the five acidic amino-terminal residues of skeletal muscle actin have been implicated as being important in a number of actin-related processes. We have constructed a series of actins containing mutations at Asp3 and Asp11 and tested these mutant proteins for their ability to bind to DNase I-agarose, polymerize with rabbit skeletal muscle actin, undergo amino-terminal processing, and bind to the myosin-S1 subfragment. The mutant actins were expressed in vitro using a coupled transcription/translation system which involves the synthesis of mutant RNAs with SP6 RNA polymerase followed by their translation in a rabbit reticulocyte lysate. When Asp3 was changed to Ala, His, or Asn there was no difference in the tested properties as compared to wild type actin. These results suggest that an acidic residue at position 3 is not critical for the actin functions measured. When Asp11 was changed to Glu, Asn, or His or if the conserved Asp-Asn sequence at positions 11 and 12 was reversed, the mutants were able to copolymerize with rabbit skeletal muscle actin and be cross-linked to myosin-S1 to nearly the same extent as wild type actin. However, the amount of in vitro-synthesized actin capable of binding to DNase I-agarose with high affinity or undergoing amino-terminal processing was reduced significantly relative to the wild type actin synthesized in vitro. The Asp11 mutants ran anomalously on native polyacrylamide gels suggestive of a conformational change induced in the actin. Together, these results suggest that Asp11 may be important in proper actin folding and function.

Epidermal growth factor controls smooth muscle alpha-isoactin expression in BC3H1 cells

We have examined the effects of epidermal growth factor (EGF), platelet-derived growth factor, and insulin on the differentiation of a mouse vascular smooth muscle-like cell line, the BC3H1 cells. On the basis of cell morphology and smooth muscle alpha-isoactin synthesis, we demonstrate that EGF at physiological concentrations prevents the differentiation of these cells, whereas platelet-derived growth factor has no apparent effect. The induction of alpha-isoactin synthesis by serum deprivation is inhibited by EGF in a dose-dependent manner with a half-maximal effect at 3-5 ng/ml and a maximal inhibition at approximately 30 ng/ml. Northern analysis also shows that EGF blocks the accumulation of alpha-isoactin mRNA normally observed during cell differentiation. Addition of EGF to differentiated cells results in a repression of alpha-isoactin synthesis, a stimulation of beta- and gamma-isoactin synthesis, and the stabilization of the nonmuscle isoactins. The synthesis of creatine phosphokinase, a muscle-specific noncontractile protein, is also regulated by EGF in a similar fashion. Modulation by EGF of alpha-isoactin expression is not affected by aphidicolin and is therefore independent of its mitogenic effect on these cells. Insulin is not required for observation of the EGF-dependent effects but instead seems to promote differentiation. Our results show that EGF can replace serum in controlling the differentiation of BC3H1 cells.

Studies on the role of actin's N tau-methylhistidine using oligodeoxynucleotide-directed site-specific mutagenesis

The primary structure of all actins except that isolated from Naegleria gruberi contains a unique N tau-methylhistidine (MeHis) at position 73. This modified residue has been implicated as possibly being important for the post-translational processing of actin's amino terminus, the binding of actin to DNase I, and in the polymerization of G-actin. We have investigated the potential role of MeHis in each of these processes by utilizing site-directed mutagenesis to change His-73 of skeletal muscle actin to Arg and Tyr. Wild type and mutant actins were synthesized in vivo, using non-muscle cells transfected with mutant cDNAs, and in vitro by translating mutant RNAs synthesized using SP6 RNA polymerase in a rabbit reticulocyte lysate. We have found that actins containing Arg or Tyr at position 73 undergo amino-terminal processing, bind to DNase I-agarose, and become incorporated into the cytoskeleton of a nonmuscle cell as efficiently as wild type actin. Furthermore, using an in vitro copolymerization assay we have found that although there is no difference between the Arg mutant and the wild type actins, the Tyr mutant has a slightly greater critical concentration for polymerization. These results show that MeHis is not absolutely required for any of these processes.

Alternate pathways for removal of the class II actin initiator methionine

Class II actin genes usually specify a polypeptide with a Met-Cys-Asp NH2 terminus, whereas the actin itself begins with an acetyl (Ac)-Asp(Glu). Previous studies with Drosophila actin showed that the first detectable intermediate is one with an Ac-Cys NH2 terminus which is subsequently cleaved in a novel reaction to expose the Asp. The initiator Met was probably removed early in translation as a free amino acid. To determine whether the class II actin initiating Met could also be removed in an acetylation-dependent manner, we translated Drosophila mRNA in a rabbit reticulocyte lysate in which protein acetylation was inhibited. After 60 min, three actin intermediates were detected, NH2-Met-Cys-Asp-actin, Ac-Met-Cys-Asp-actin, and NH2-Cys-Asp-actin. During processing in the presence of acetyl-CoA, three additional species were observed with NH2-terminal Ac-Cys-Asp, NH2-Asp, and Ac-Asp segments. In a time- and acetyl-CoA-dependent fashion, Met-Cys-Asp-actin was processed to the mature actin, presumably through an Ac-Met-Cys-Asp intermediate. Thus, two different pathways for removal of the initiator Met of class II actins, acetylation-dependent and independent, are possible. Since no class II actin intermediate containing the initiator Met is seen in vivo, although in class I actins this intermediate is observed, the most probable pathway for class II actins in vivo is the cotranslational removal of the initiator Met as a free amino acid.

Characterization of actin mRNA levels during BC3H1 cell differentiation

The expression of a vascular smooth muscle specific alpha-actin isoform can be induced in mouse BC3H1 smooth muscle cells by treating confluent monolayers with serum-free medium (Strauch, A. R., and Rubenstein, P. A. (1984) J. Biol. Chem. 259, 3152-3159; 7224-7229). Using blot hybridization techniques, two size classes of actin RNA were identified in BC3H1 cells with the relative amount of RNA in each size class varying according to the developmental state of the cells; a 2100-nucleotide actin RNA was most abundant in myoblasts, whereas a smaller 1500-nucleotide actin RNA was found predominantly in fully differentiated myocytes. Results of in vitro translation experiments suggested that the 2100-nucleotide actin RNA on blots of myoblast total RNA corresponded to a mixture of similar size transcripts encoding both beta- and gamma-actin, while the 1500-nucleotide actin RNA in myocytes was an alpha-actin mRNA. Cell-cell contact and serum withdrawal initiated a 6-fold increase in the level of alpha-actin mRNA in BC3H1 cells that was followed by a 3-fold decrease in the amount of beta- and gamma-actin mRNA when confluent cells were exposed to serum-free medium for prolonged periods. Vascular smooth muscle alpha-actin was the major alpha-actin isoform synthesized in L-[35S]cysteine-labeled BC3H1 myocytes, indicating that the 1500-nucleotide actin mRNA size class in these cells may be enriched for vascular smooth muscle alpha-actin transcripts.

Lack of NH2-terminal processing of actin from Acanthamoeba castellanii

Acanthamoeba actin is the only actin sequenced to date that has neither an NH2-terminal Ac-Asp nor Ac-Glu residue. The protein begins with an Ac-Gly-Asp and is coded for by a gene that specifies a polypeptide beginning Met-Gly-Asp. Thus, the Acanthamoeba actin gene would appear to specify a class II actin with the usual NH2-terminal Cys replaced with a Gly. Previous studies (Rubenstein, P. A., and Martin, D. J. (1983) J. Biol. Chem. 258, 11354-11360) revealed that for class II actins the Met is probably removed early in translation and the Cys is removed post-translationally as an Ac-Cys residue. Two possibilities might explain why Acanthamoeba actin is not processed in a similar fashion. Either Ac-Gly is not a substrate for the enzyme or the enzyme is absent from the organism. To test these alternatives, Acanthamoeba actin was labeled in vivo with [35S]methionine and incubated with processing enzyme from rat liver, rabbit reticulocytes, and Dictyostelium. In no case did the processing reaction occur, indicating that Ac-Gly is not recognized by the enzyme as a substrate. Furthermore, we could not reproducibly detect the presence of a processing enzyme in Acanthamoeba. We were, however, able to show the presence of such an enzyme in Dictyostelium, the first demonstration of this activity in a lower eukaryote.

Correct NH2-terminal processing of cardiac muscle alpha-isoactin (class II) in a nonmuscle mouse cell

Both mammalian nonmuscle and muscle actins possess an AcAsp(Glu)NH2 terminus. The nonmuscle actin genes code for a polypeptide with a Met-Asp NH2 terminus (class I) whereas the muscle actin genes code for a polypeptide with a Met-Cys-Asp NH2 terminus (class II). Two amino acids must be removed for mature muscle actin synthesis, whereas only the Met must be removed for nonmuscle actin synthesis. We wished to know whether a nonmuscle cell which normally does not synthesize a class I actin can correctly process a muscle actin with its extra NH2-terminal amino acid in vivo. To answer this question we have used L/LK165 cells, a mouse L-cell transfected with a human cardiac muscle actin gene. When these cells were labeled overnight with [35S]Cys, an actin with an NH2-terminal tryptic peptide corresponding to that of mature cardiac muscle actin was detected. When the cells were pulse-labeled for 20 min, a new actin intermediate containing an AcCys-Asp amino terminus was observed which then disappeared with time. Furthermore, the muscle actin was processed as fast if not faster than the nonmuscle actin in these cells. This actin intermediate was also seen in chick myotube cultures. Our results show that the ability to correctly process muscle specific actins is not tissue specific. Furthermore, these results confirm a processing pathway for class II actins proposed by us earlier on the basis of experiments with a cell-free translation system.

A vascular smooth muscle alpha-isoactin biosynthetic intermediate in BC3H1 cells. Identification of acetylcysteine at the NH2 terminus

A fully translated actin biosynthetic intermediate containing N-acetylcysteine at the NH2 terminus has been identified in homogenates of differentiated mouse BC3H1 cerebrovascular smooth muscle cells labeled with L-[35S]cysteine. Thermolysin digestion of the highly acidic NH2-terminal tryptic peptide of this intermediate and electrophoretic analysis of the resulting fragments indicated that the intermediate was a precursor of smooth muscle alpha- isoactin , the major isoactin species in vascular smooth muscle. Carboxypeptidase A digestion of the thermolysin cleavage product corresponding to the first eight amino acid residues of the NH2-terminal tryptic peptide demonstrated an acetylcysteine-glutamate residue at the NH2 terminus. These results imply that the gene for smooth muscle alpha- isoactin , like genes coding for skeletal and cardiac alpha- isoactins , contains a cysteine codon immediately following the initiator methionine codon. Both the methionine and cysteine residues must be removed from the NH2 terminus of the intermediate to yield the mature form of smooth muscle alpha- isoactin . The removal of the cysteine residue in vivo is not direct but apparently involves acetylation of the cysteine and subsequent post-translational cleavage of the resulting acetylcysteine. Such an acetylation-dependent pathway has been demonstrated for removal of cysteine from the NH2 terminus of Drosophila actin synthesized in a cell-free translation system ( Rubenstein , P. A., and Martin, D. J. (1983) J. Biol. Chem. 258, 11354-11360). In vivo pulse-chase experiments indicate that the smooth muscle alpha- isoactin intermediate in BC3H1 cells turns over much more slowly than nonmuscle actin intermediates previously identified in mouse L-cells.

Cerebral microvascular smooth muscle in tissue culture

Cerebral endothelium is being studied rather extensively in tissue culture, but no reports are available describing the tissue culture of cerebral microvascular smooth muscle. The present paper describes for the first time the isolation and culture of non-neoplastic mouse cerebral vascular smooth muscle. Microvessels from a dounce homogenate of mouse brain are plated onto plastic culture dishes in Dulbecco's modified Eagle media plus 20% fetal bovine serum and treated briefly with collagenase. Cells migrate from vessels and proliferate sufficiently to be transferred out of primary culture in 2 to 3 wk. Light microscopy reveals generally broad, polygonal cells that grow collectively in a "hill and valley" pattern. By transmission electron microscopy the cells possess many characteristics of smooth muscle: basal laminas, clusters of pinocytotic vesicles, and bundles of thin filaments. Several ill-defined cell-to-cell junctions are also present. Isoelectric focusing and sodium dodecyl sulfate-electrophoresis of cellular proteins on polyacrylamide gels after pulse labeling cultures with [S-35]methionine demonstrate that these cells actively synthesize a smooth-muscle-specific isoactin, alpha-actin. The identity of alpha-actin is confirmed by analysis of NH2-terminal peptides after actin digestion with trypsin and subsequent peptide cleavage with thermolysin. Both their morphology and active synthesis of alpha-actin strongly suggest that these cells are of smooth-muscle origin. Future studies of their metabolism and interactions with endothelium and astrocytes should provide a better understanding of the cerebral microcirculation.

Induction of vascular smooth muscle alpha-isoactin expression in BC3H1 cells

An isoactin analysis was performed on L-[35S]cysteine labeled BC3H1 cells to determine if these smooth muscle-like cells synthesize vascular smooth muscle actin. Three different NH2-terminal peptides were identified on thin layer electrophoretograms of DNase I-purified and trypsin-digested BC3H1 cell actin. Results obtained from secondary digestion with thermolysin or Staphylococcus aureus V8 protease showed that the most acidic NH2-terminal peptide was derived from vascular smooth muscle alpha-isoactin. Treatment of cell monolayers with serum-free medium caused a 3-fold increase in the level of alpha-isoactin expression and a concomitant decrease in the level of non-muscle beta- and gamma-isoactin. Cell-cell contact was required for induction of alpha-isoactin, and the effects of serum depletion on isoactin expression and cell growth were reversible. The intensity of about 11 out of 500 polypeptide spots on two-dimensional gels of BC3H1 cell polypeptides also was influenced by the culture conditions. The finding that smooth muscle isoactin expression was coupled to cell growth conditions indicate the potential usefulness of BC3H1 cells in studies of isoactin expression and utilization during vascular smooth muscle development.