Allele-specific Effects of Thoracic Aortic Aneurysm and Dissection α-Smooth Muscle Actin Mutations on Actin Function*

Twenty-two missense mutations in ACTA2, which encodes α-smooth muscle actin, have been identified to cause thoracic aortic aneurysm and dissection. Limited access to diseased tissue, the presence of multiple unresolvable actin isoforms in the cell, and lack of an animal model have prevented analysis of the biochemical mechanisms underlying this pathology. We have utilized actin from the yeast Saccharomyces cerevisiae, 86% identical to human α-smooth muscle actin, as a model. Two of the known human mutations, N115T and R116Q, were engineered into yeast actin, and their effect on actin function in vivo and in vitro was investigated. Both mutants exhibited reduced ability to grow under a variety of stress conditions, which hampered N115T cells more than R116Q cells. Both strains exhibited abnormal mitochondrial morphology indicative of a faulty actin cytoskeleton. In vitro, the mutant actins exhibited altered thermostability and nucleotide exchange rates, indicating effects of the mutations on monomer conformation, with R116Q the most severely affected. N115T demonstrated a biphasic elongation phase during polymerization, whereas R116Q demonstrated a markedly extended nucleation phase. Allele-specific effects were also seen on critical concentration, rate of depolymerization, and filament treadmilling. R116Q filaments were hypersensitive to severing by the actin-binding protein cofilin. In contrast, N115T filaments were hyposensitive to cofilin despite nearly normal binding affinities of actin for cofilin. The mutant-specific effects on actin behavior suggest that individual mechanisms may contribute to thoracic aortic aneurysm and dissection.

in the absence of systemic findings of a connective tissue disorder, complicating diagnosis (3). Twenty percent of affected patients inherit the disorder, and the vast majority have an autosomal dominant pattern of inheritance. However, variable penetrance further hobbles patient identification in patients with familial thoracic aortic aneurysm and dissection (TAAD) 2 (4).
Multiple genes and loci have been associated with familial TAAD (5,6). Mutations in ACTA2, which encodes ␣-smooth muscle actin (␣-SM actin), are the most common genetic cause of familial TAAD, responsible for 15% of cases (7). Interestingly, mutations in ACTA2 are associated with an array of cardiovascular diseases from congenital to premature acquired heart disease. Clinical problems include patent ductus arteriosus, occlusive strokes, and coronary artery disease in addition to thoracic aortic aneurysms (8). ␣-SM actin is one of six highly homologous isoactins expressed in mammals. Mutations in the actin family have recently been associated with disease, the vast majority of which behave in a dominant fashion. Over the past decade, mutations in ␣-cardiac (ACTC), ␣-skeletal (ACTA1), and ␥-cytoplasmic actin (ACTG1) have been associated with cardiomyopathies, skeletal myopathies, and deafness, respectively. As such, the role for the ␣-SM actin in human disease is not surprising.
In vascular smooth muscle cells, ␣-SM actin has many roles, including maintenance of cell wall integrity, force transduction, and regulation of vascular smooth muscle cell proliferation (9). ␣-SM actin is the most abundant protein in vascular smooth muscle cells, making up 40% of total cellular protein and over 70% of the total actin (10). The other two actin isoforms in the vascular smooth muscle cells are ␤and ␥-cytoplasmic actin. Histopathology from individuals with ␣-SM actin mutations noted loss or disorganization of the vascular smooth muscle cells in the tunica media of the aorta (8). Analysis of aortic vascular smooth muscle cells showed diminished ␣-SM actin, with substantially fewer polymerized actin fibers that did not extend completely across the cell body. ␣-SM actin staining did not co-localize with polymerized actin fibers and was found clumped along the cell wall or nuclear wall.
To gain insight into how ACTA2 mutations cause disease, the functional changes in the context of the structure of the actin molecule must be established. The actin monomer can be divided into two large domains. An adenine nucleotide binds between the two halves, bridging the cleft to impart stability to the protein, and plays a major role in controlling protein dynamics. Each of the domains can each be further divided into two subdomains, as shown in Fig. 1. Actin polymerizes to form a polar filament with what is termed a barbed and a pointed end. The barbed end is the preferred site for monomer addition during filament elongation and, as such, is where much of actin assembly regulation occurs. The asymmetry can be imparted to the actin monomer as well. Subdomains 1 and 3 constitute the monomer barbed end, whereas subdomains 2 and 4 make up the pointed end.
At least 22 autosomal dominant missense mutations in ␣-SM actin have now been identified in familial TAAD, and they are distributed across all four subdomains (7,(11)(12)(13). The focus of this paper is on two of these mutations, N117T and R118Q, which are adjacent to one another in a secondary structural element, near the barbed end of the protein. The location is predicted to be involved in intermonomer interactions, as shown in Fig. 1. Of interest, the two mutations have different clinical phenotypes. The N117T mutation has been found in both ␣-skeletal muscle and ␣-SM actins and is associated primarily with aneurysms of the thoracic aorta (7,8,14). R118Q is the second most common mutation identified in ␣-SM actin, and remarkably, 75% of patients with this mutation have premature coronary artery disease in addition to aortic aneurysms. The varied clinical phenotypes suggest that although all 22 of these mutations cause TAAD, the molecular mechanisms by which different mutations lead to aortic disease are allele-specific.
Understanding of the effects of ␣-SM actin mutations at the molecular level in the context of the aortic wall is hampered by the difficulty in obtaining sufficient samples from patients for biochemical studies. Even if one were to introduce the mutations into model animals to establish cultured vascular smooth muscle cell preparations, the amount of material available for biochemical analysis would be limited. Additionally, the presence of three essentially unresolvable isoforms of actin in the vascular smooth muscle cells would hinder direct assessment of the effects of the mutations on ␣-SM actin function. An attractive model system for studying the effects of actin mutations is the budding yeast, Saccharomyces cerevisiae. Yeast actin is 86% identical and 94% similar to ␣-SM actin and is encoded by a single essential gene, ACT1 (15). More importantly, the two residues at which the TAAD mutations occur in ␣-SM actin are identical in yeast actin (N117T and R118Q in ␣-SM actin correspond to N115T and R116Q in yeast actin, and for the remainder of this work, the yeast numbering system will be used). In addition, actin regulation is well conserved between species. Many of the actin-binding proteins in mammalian cells are also found in yeast (16,17), and yeast actin will interact with many of the mammalian isoforms of actin-binding proteins due to the high degree of homology between actins (18,19). Mutations can be readily introduced into the actin gene by site-directed mutagenesis to express mutant actin as the sole actin isoform. The impact of mutations on actin function in the cell can be assessed cytologically, and the actin can be purified for studies of the biochemical effects of the mutations on the molecule. The in vivo and in vitro data can be analyzed for alterations in molecular behavior to provide insight into the mechanisms underlying aortic disease. We have previously applied  (47), with the N115T and R116Q mutations color-highlighted and labeled as described above. The trimer model highlights the location of the mutations along the intermonomer interface. The numbers denote the actin subdomains of the individual monomers comprising the trimer. The symbols indicate the individual monomers within the filament; *, cyan monomer; no symbol, red monomer; and #, gray monomer. Subdomains 1 and 3 constitute the barbed end of the actin molecule, whereas subdomains 2 and 4 constitute the pointed end; this polarity is also consistent within the filament. this approach to the analysis of mutations in ␥-nonmuscle actin that cause autosomal dominant early onset deafness (20 -22).
In this study, either N115T or R116Q mutant actin was expressed as the sole actin in the cell. The effects of the TAAD mutations on cell behavior and actin-related cellular functions were characterized. Actin was purified, and the impact of the mutations on actin monomer and polymer function in vitro was assessed.

Materials
DNase I (grade D) was purchased from Worthington. DE52 DEAE-cellulose was obtained from Whatman. Micro Bio-Spin P-30 Tris columns and Affi-Gel 10-activated resin were purchased from Bio-Rad. ATP was purchased from Sigma. ⑀-ATP, rhodamine-phalloidin, FM4-64, and 4Ј,6-diamidino-2-phenylindole (DAPI) were purchased from Molecular Probes. The QuikChange site-directed mutagenesis kit was from Stratagene, and oligodeoxynucleotides were purchased from Integrated DNA Technologies. Yeast cakes for wild type actin preparations were purchased from a local bakery. All other chemicals were of reagent grade quality.

Construction of Mutant Yeast Strains
Mutations were introduced into the centromeric plasmid pRS314 (21) containing the yeast actin coding sequence driven by the ACT1 promoter using the QuikChange mutagenesis kit according to the manufacturer's instructions. Plasmids containing the desired mutations were introduced into a recipient yeast strain containing a deleted chromosomal ACT1 gene and a plasmid expressing wild type actin (pCENWT) as described previously (23). Plasmid shuffling yielded viable haploid strains for each of the mutations. The plasmids containing the mutant constructs were reisolated from these strains and sequenced to confirm the presence of the desired mutation.

Growth Behavior in Liquid Culture
Cells were grown in complete liquid YPD medium (1% yeast extract, 2% peptone, and 2% dextrose) at 30°C on a shaking platform. Growth was determined by diluting an overnight culture of each strain into fresh medium at an initial A 600 of 0.1 and following growth at 30°C with agitation. The absorbances of the cultures were recorded as a function of time. The absorbances were back-calculated following the appropriate dilutions to lower the cell density to the linear range of the spectrometer.

Growth under Stress Conditions
Temperature sensitivity of mutant actin was examined by plating four serial 10-fold dilutions of the cultures on YPD plates followed by incubation at 24, 30, or 37°C. Colony size was assessed as a function of time. To assess mitochondrial function, cells were grown on media containing glycerol as the sole carbon source. Cultures were plated on YPG medium (YPD medium with the dextrose replaced with 2% glycerol) and incubated at 30°C. To test for hyperosmolar sensitivity, cells were plated on YPD plus 0.9 M NaCl agar plates and incubated at 30°C.

Cytology
Cell structures were imaged with an Olympus IX81 microscope and a Hamamatsu (model C10600-10B-H) camera. Camera control and image enhancement were performed using MetaMorph version 4.5 software (Universal Image Corp., Downingtown, PA). Presentation of cell images was done using CorelDRAW 11. All cellular statistical analysis was based on cell counts of Ͼ100 for each sample. To measure cell size, mounted samples were visualized by differential interference contrast microscopy. The long axis of the cell was measured using ImageJ (Wayne Rasband, National Institutes of Health, Bethesda, MD).
Mitochondria were visualized in living cells by expressing a fusion protein of green fluorescent protein (GFP) conjugated to the mitochondrial signal sequence of citrate synthase kindly provided by Dr. Liza A. Pon (24). Cells expressing the plasmid were grown to an A 600 of 0.3-0.5 in Ura Ϫ synthetic medium to force retention of the URA3-marked plasmid in the otherwise Ura3 Ϫ cells. An aliquot of cells was resuspended in VECTASHIELD mounting medium (Vector Laboratories, Burlingame, CA), and the cells were then observed by fluorescence microscopy as described above. Z-sections through the cell were obtained at 0.15-m intervals. Out-of-focus light was removed by deconvolution using MetaMorph software, and each series of deconvolved images was further rendered with ImageJ.
The actin cytoskeleton was visualized by fluorescence microscopy after staining fixed cells with rhodamine-phalloidin as described previously (25). Cytoskeletal analysis focused on budding cells when the daughter cell is less than one-half the size of the mother cell. Vacuoles were imaged following exposure of the cells to the dye FM4-64 as described previously (26). Nuclear and mitochondrial DNA was stained with DAPI as described previously (25).

Latrunculin A Sensitivity
Sterile filtered discs (0.5 cm in diameter) were presoaked in 2 l of DMSO (control) plus 0.1, 0.5, or 1 mM of latrunculin A (9). Soaked discs were placed on YPD plates containing 100 l of evenly spread wild type or mutant cells (A 600 ϭ 0.1). The plates were incubated at 30°C for 48 h, and the zone of growth inhibition around the drug eluting latrunculin A disc was assessed.

⌬Aip1p Deletion Strain Studies
Using an ⌬aip1:pCENWT strain as the host (26), pRS314 plasmids containing the promoter region, the TRP1 gene, and the coding sequence for wild type or mutant yeast actin were transformed into this haploid yeast strain (27). Transformants were selected on tryptophan-deficient medium and then subjected to plasmid shuffling to eliminate the wild type actin gene. Growth curves, growth characteristics, and cytology were determined as described above for wild type and each mutant in the ⌬Aip1p deletion strain.

Protein Purification
Wild type and mutant yeast actins were purified from lysates of frozen yeast cells using a combination of DNase I-agarose

Effects of ␣-Smooth Muscle Actin Mutations on Actin Function
affinity chromatography, DEAE-cellulose chromatography, and polymerization/depolymerization cycling as described previously (27). The quality of actin preparations was assessed using SDS-PAGE and Coomassie Blue staining of the gels. The concentration of G-actin was determined from the absorbance at 290 nm using an extinction coefficient of 0.63 M Ϫ1 cm Ϫ1 . Actin was stored as G-actin in G buffer (5 mM Tris-HCl, pH 7.5, 0.1 mM ATP, pH 7.5, 0.2 mM CaCl 2 , and 0.2 mM dithiothreitol). All actins were stored at 4°C and used within 4 days of purification. Yeast cofilin was purified from Escherichia coli carrying a recombinant construct for the protein according to Lappalainen et al. (20,28), and the concentration of the purified cofilin was determined by absorption at 280 nm with an extinction coefficient of 14,650 M Ϫ1 cm Ϫ1 .

Actin Polymerization
Polymerization of 4.8 M G-actin in a total volume of 120 l was induced by the addition of MgCl 2 and KCl to final concentrations of 2 and 50 mM, respectively (F-salts). Polymerization at 25°C was monitored by following the increase in light scattering of the sample in a FluoroMax-3 fluorescence spectrometer outfitted with a computer-controlled thermostatted fourposition multisample exchanger (HORIBA Jobin Yvon Inc.). The excitation and emission wavelengths were both set to 360 nm with the slit widths set at 1 nm. To determine the effects of cofilin on preformed actin filaments, the desired amount of cofilin was added to the polymerized actin sample, and the resulting change in light scattering was monitored. For experiments examining the effects of different molar fractions of mutant actin on overall actin behavior, wild type and mutant actin were combined with a final total actin concentration of 4.8 M before induction of polymerization. All polymerization experiments were performed at least three times with at least three different actin preparations.

Actin Depolymerization Rates
Actin was polymerized to steady-state levels followed by the addition of DNase I in a 1:1 actin/DNase molar ratio. Depolymerization was monitored as a decrease in light scattering over time. Depolymerization rates were determined by fitting the data to a single exponential expression using BioKine version 3.1.

Circular Dichroism Measurements
The apparent melting temperatures of wild type and mutant actins were determined using circular dichroism by following the change in ellipticity of the G-actin sample at 222 nm as a function of temperature between 25 and 90°C as described previously (21). Measurements were made on an Aviv 62 DS spectropolarimeter. Data were fit to a two-state model, and the apparent T m value was determined by fitting the data to the Gibbs-Helmholtz equation to approximate the temperature at which 50% of the actin was denatured.

⑀-ATP Exchange
The ability of G-actin to exchange bound nucleotide was assessed by first loading the actin with ⑀-ATP and quantifying the rate of displacement from actin in the presence of a large excess of ATP as described previously (21). Exchange rates were determined by fitting the data to a single exponential expression using BioKine version 3.1.

Cofilin Binding Assay
G-actin-Pyrene-labeled G-actin was made according to Feng et al. (29). Increasing amounts of cofilin were added to a 1.5-ml sample containing 1 M 100% pyrene-labeled G-actin, and the cofilin-dependent increase in pyrene fluorescence was recorded on a FluoroLog3 fluorescence spectrometer outfitted with a computer-controlled thermostatted sample exchanger with continuous sample mixing (HORIBA Jobin Yvon Inc.). All experiments were performed in G-buffer containing 50 mM KCl. Note that this concentration of KCl will not induce polymerization of yeast actin because polymerization of yeast actin requires Mg 2ϩ , unlike muscle actin (30). The excitation and emission wavelengths were 344 and 386 nm, respectively, with the corresponding slit widths of 1 and 2 nm. Using Microsoft Excel, experimental data were fit to the quadratic binding isotherm, where ⌬F represents the observed fluorescence change of the actin-cofilin complex after the fluorescence of the G-actin alone has been subtracted; F max is the maximum fluorescence change at complete saturation of actin with cofilin; [A] and [C] are the concentrations of G-actin and cofilin, respectively; and K d is the observed dissociation constant. The solver function was used to minimize the difference between the experimental data and the best fit to produce the K d .
F-actin-Pyrene-labeled G-actin was made according to Feng et al. (29). Polymerization of 4.8 M G-actin (10% pyrenelabeled and 90% unlabeled) in a total volume of 120 l was induced by the addition of F-salts. The change in actin-pyrene fluorescence due to actin polymerization was recorded with a FluoroLog3 fluorescence spectrometer (Jobin Yvon-Spex). The excitation wavelength was 365 nm. The change in fluorescence intensity at emission wavelength 386 nm was recorded over time for kinetics analyses. After steady state was reached, an increasing amount of cofilin was added to the sample, and the cofilin-dependent quenching of the pyrene fluorescence was recorded. The excitation and emission wavelengths were 344 and 386 nm, respectively, with the corresponding slit widths of 1 and 2 nm. Using Microsoft Excel, experimental data were fit to the quadratic binding isotherm (as above).

Critical Concentration Determination
To measure the critical concentration (C c ) of each actin, the net change in light scattering of an actin polymerization reaction was measured as a function of increasing actin concentration. Polymerization of G-actin, at concentrations between 4.8 and 1 M, was induced by the addition of F-salts and monitored by light scattering. The final increase in light scattering for each actin concentration was recorded and plotted versus actin concentration. The critical concentration of actin was obtained by fitting the points to a linear regression trend line and determining its intersection on the x axis.

Electron Microscopy
To visualize actin filament morphology, samples of 4.8 M F-actin were deposited onto carbon-coated Formvar grids, negatively stained with 1% uranyl acetate, and imaged with a JOEL 1230 transmission electron microscope (University of Iowa Central Electron Microscopy Facility) equipped with a Gatan UltraScan 1000 CCD 2000 ϫ 2000-pixel camera. Accelerating voltage of the transmission electron microscope was 100 kV. Image J was used to process the images and measure the individual filament lengths for N115T, R116Q, and wild type actins (20).

Statistical Analysis
All data are presented as means Ϯ S.D. Results for wild type and mutant actins were compared by using a paired t test with a p value of Ͻ0.05 considered significant.

Effect of Mutations on Yeast Cell
Growth-To assess the effects of missense mutations on actin function in vivo, the basic biological functions dependent on actin in yeast were examined, including cell growth, cell division, and response to temperature or osmotic stress (31)(32)(33). Yeast cells expressing N115T or R116Q actin as the sole actin in the cell were compared with wild type. There were no differences in doubling time (ϳ2 h) or extent of growth with the mutant isoforms under normal growth conditions (data not shown). Cells expressing mutant actin were 25% larger than wild type cells, however. The diameter of wild type cells was 7.8 Ϯ 0.98 m, whereas N115T and R116Q cells measured 9.88 Ϯ 1.4 and 9.87 Ϯ 1.18 m, respectively (p Ͻ 0.001). In addition, the N115T strain exhibited temperature-sensitive growth at 37°C on YPD medium, and both N115T and R116Q strains had defective growth in hyperosmolar medium, indicative of an impaired actin cytoskeleton ( Fig. 2A).
A functional cytoskeleton is required for mitochondrial function and inheritance (24,34,35). To test for defects in mitochondrial integrity, the ability of mutant cells to grow on glycerol as a sole carbon source was assessed. Yeast requires mitochondrial glycerol-3-phosphate dehydrogenase to utilize glycerol for glycolysis (36,37). As such, actin-dependent deficiencies in mitochondrial function can lead to impaired growth FIGURE 2. Effect of TAAD mutations on cell growth and actin-dependent cellular structures. A, comparison of wild type or mutant yeast strain growth on complete solid medium with either 2% dextrose (YPD at 30 and 37°C), in the presence of 0.9 M NaCl, or 2% glycerol as the sole carbon source. Cell cultures were diluted and spotted onto YPD-or glycerol-containing agar plates and grown at 30°C unless otherwise specified, as described under "Experimental Procedures." Pictures were taken 48 h subsequent to plating the cells. B, fluorescence microscopy of cells expressing wild type, N115T, and R116Q yeast actin. Cells assessed were those in which the bud was between one-half and one-third the size of the mother cell. Quantified results are based on assessment of Ͼ100 cells for each sample. The cytoskeleton was visualized after staining fixed cells with rhodamine-phalloidin. Vacuoles were observed following exposure of the cells to the dye FM4-64. Mitochondria were visualized with GFP as described under "Experimental Procedures." Scale bar, 10 m. Larger fields of cells for each assay can be found in supplemental Fig. 1. C, the bar height indicates the percentage of the cell population that exhibited normal structures. Error bars, S.D. Differences were statistically significant as compared with wild type actin (p Ͻ 0.05). on glycerol. Both the N115T and R116Q strains exhibited mildly diminished growth on glycerol medium ( Fig. 2A), indicating relatively preserved mitochondrial function.

Effect of the Mutations on Actin Cytoskeletal Patterns and Organelle Morphology-Alterations in morphology of the cell and actin cytoskeleton could result from mutations in actin.
Disruption of the actin cytoskeleton may lead to loss of polarization during budding, resulting in aberrant cell size and actin organization (21). Actin filaments in yeast are organized as cables and patches. Cables facilitate movement of organelles to the daughter cell during division. Patches are distributed in a cell cycle-dependent manner and localize at sites of endocytosis (38,39). Cellular phenotypes were determined by analysis of more than 600 cells/actin isoform by two masked reviewers. Both mutations affected cytoskeletal morphology with a high frequency of abnormalities compared with wild type (p Ͻ 0.01) (Fig. 2, B and C). Inappropriately distributed patches or lack of appropriate cables was found with high frequency in budding cells expressing mutant actin, in which the bud was between one-half and one-third the size of the mother cell (63% of N115T cells, 38% of R116Q cells). A subjective phenotypic difference between mutant actin isoforms was appreciated. N115T cells displayed patches in nonpolar distributions in both mother and daughter cells in addition to cables inappropriately aligned within the mother cell. R116Q cells also displayed patches in a nonpolar distribution in mother and daughter cells as well as cables that were less prominent but organized in a polarized fashion. These subjective cytoskeletal differences warrant further quantitative analysis.
Actin regulates organelle movement from the mother cell to the bud along polarized actin cables. Given the abnormalities seen in the actin cytoskeleton, organelle morphology and distribution in the mutant cells were assessed. Vacuole morphology, regulated by actin, was visualized by staining with FM4-64. Both mutations resulted in abnormal vacuole morphology (p Ͻ 0.01 compared with wild type) but in an allele-specific manner (Fig. 2, B and C). Wild type cells most commonly have 1-4 vacuole lobes. Ninety-one percent of wild type cells displayed normal vacuole morphology, whereas only 68% of N115T and 64% of R116Q cells exhibited the normal phenotype. Of interest, the abnormal vacuole morphology found in mutant cells differed between alleles (p Ͻ 0.01). Hypervesiculation with 5-10 small vacuoles was the most common anomalous phenotype, found in 59% of abnormal N115T cells. In contrast, an oversized lobe, filling more than half of the cell was the dominant aberrant morphology seen in 70% of the abnormal R116Q cells (supplemental Fig. 1).
To assess the effect of mutations on DNA distribution, genomic and mitochondrial DNA were imaged by staining cells with the DNA-intercalating dye DAPI. We found no differences in DNA distribution between cells expressing the mutations and wild type (data not shown). The normal findings of a large single spot consistent with nuclear DNA and multiple diffuse, faint extra nuclear spots consistent with mitochondrial DNA had similar prevalence in all three cell types.
Disorganization of the polarized actin cytoskeleton can lead to abnormal fission/fusion events that regulate mitochondrial morphology and function (40). To visualize the mitochondria, cells were transformed with a plasmid containing GFP fused to the mitochondrial targeting sequence of citrate synthase. Normal mitochondrial tubular structures are arranged in an orderly, polar manner running from the mother cell to bud. This normal mitochondrial pattern was present in 94% of wild type cells compared with only 43% of the N115T cells and 37% of the R116Q cells (p Ͻ 0.05) (Fig. 2, B and C). Both the N115T and R116Q cells demonstrated fragmented and aggregated mitochondrial tubules. Despite these morphological effects, mitochondrial function remained relatively normal as assessed by growth behavior as described above.
The cytoskeletal dysfunction in mutant cells prompted evaluation of cytoskeletal stability in vivo. Latrunculin A is a drug that sequesters actin monomers and reversibly promotes rapid depolymerization of actin filaments (9,(41)(42)(43). Greater sensitivity of cells to latrunculin A can indicate an increased monomer off-rate contributing to an unstable actin cytoskeleton. Using a disc diffusion assay, the sensitivity of cells to latrunculin A was assessed. Discs saturated with a range of drug concentrations were placed on cultures for 48 h, and the zone of growth inhibition surrounding the disc was compared (Fig. 3). Both mutants displayed increased growth inhibition compared with wild type cells, with R116Q demonstrating the most sensitivity to latrunculin A.
Effect of the Mutations on in Vivo Phenotype in a ⌬Aip1 Background-The in vivo findings support overall cytoskeletal instability. If such instability exists in the cytoskeleton, decreasing the effects of the machinery in the cell responsible for filament turnover might rescue the phenotype. One such regulator is the F-actin-severing protein, cofilin. Cofilin is an essential protein in yeast; however, its activity can be attenuated in vivo by eliminating Aip1p, an enhancer of cofilin. The N115T and R116Q mutations were expressed in ⌬aip1:pCENWT host cells, and the above cytological assays were repeated (Fig. 4). In the absence of Aip1p, the reduction in cofilin activity rescued the mild growth defects in hyperosmolar media seen with the R116Q mutation (compare with Fig. 2), but the frequency of cytoskeletal abnormalities persisted. In contrast, N115T cells displayed a marked improvement in cytoskeletal morphology (p Ͻ 0.005). The frequency of normal cytoskeletal structure doubled in the background of attenuated cofilin activity. No differences were seen, however, in actin-dependent vacuole morphology or mitochondrial morphology (supplemental Fig. 2).
Effects of the Mutations on Actin Monomer Characteristics in Vitro-To gain insight into the molecular basis for the in vivo phenotypes described above, wild type or mutant yeast actin was purified to assess the biochemical effects of the mutations in G-and F-actin function. No difference in the amount of expressed actin was noted in mutant cells compared with wild type based on protein purification yield. Thermal stability of G-actin was determined by quantifying the unfolding of the monomer as a function of temperature as detected by a change in circular dichroism. Both mutants showed decreased thermal stability. The T m for N115T actin, 55 Ϯ 1°C, and R116Q, 53 Ϯ 2°C, were both significantly lower than that for wild type actin (59 Ϯ 1°C, p Ͻ 0.005 for both) ( Table 1).
Mutation-dependent effects on G-actin dynamics can also alter the rate of exchange of the adenine nucleotide from deep within the protein. The kinetics of fluorescent ⑀-ATP exchange from wild type and mutant protein actin was determined. Both mutants had prolonged t1 ⁄ 2 values as compared with wild type: wild type, 36 Ϯ 3 s; N115T, 43 Ϯ 2 s; R116Q, 54 Ϯ 8 s (p Ͻ 0.0001 for both). The delayed exchange implies that mutant monomers are more rigid and/or assume a more closed conformation around the nucleotide than wild type actin.
Effects of Mutations on Actin Polymerization-A potential cause for the abnormal vascular smooth muscle cell morphology found in patients is a mutation-dependent alteration of actin polymerization, leading to delayed or altered dynamic adaptation of the contractile or cytoskeletal apparatus. Polymerization kinetics are a factor in this dynamic response. The location of the two mutations along the barbed end of the monomer, where interactions facilitate monomer addition to the filament, suggests that mutations might alter polymerization behavior characterized by changes in nucleation or elongation

. Effect of TAAD mutations on cell growth and actin-dependent cellular structures in ⌬Aip1 strain.
A, comparison of cell growth with wild type or mutant actin expressed in ⌬aip1:pCENWT host strain on complete solid medium with either 2% dextrose (YPD at 30 and 37°C), in the presence of 0.9 M NaCl, or 2% glycerol as the sole carbon source. Cell cultures were diluted and spotted onto YPD-or glycerol-containing agar plates and grown at 30°C unless otherwise specified as described under "Experimental Procedures." Pictures were taken 48 h subsequent to plating the cells. B, fluorescence microscopy of cells expressing wild type, N115T, and R116Q yeast actin. Cells assessed were those in which the bud was less than one-third the size of the mother cell. Results are based on assessment of Ͼ100 cells for each sample. The cytoskeleton was visualized after staining fixed cells with rhodamine-phalloidin. Vacuoles were observed following exposure the cells to dye FM4-64. Mitochondria were visualized with GFP as described under "Experimental Procedures." Scale bar, 10 m. Larger fields of cells for each assay can be found in supplemental Fig. 2. C, the bar height indicates the percentage of the cell population that exhibited normal structures. Error bars, S.D. Differences were statistically significant as compared with wild type actin (p Ͻ 0.05).
rates or critical concentration. The effects of the mutations on actin polymerization kinetics were assayed by quantifying the change in light scattering as the actin polymerized (Fig. 5A). Light scattering assays were used rather than pyrene fluorescence because of the inability to label the R116Q mutant actin preventing isoactin comparisons (see below). Both mutants displayed polymerization defects but with different phenotypes. N115T repeatedly demonstrated a biphasic elongation phase. R116Q had a prolonged nucleation phase and a more rapid elongation phase compared with both wild type and N115T actin. R116Q also displayed a lower final extent of polymerization. Electron microscopic examination confirmed that both mutant actins produced a relatively normal appearing filament, although the average filament length of N115T actin was 30% smaller than wild type actin (p Ͻ 0.0001) (supplemental Fig.  3A). Thus, N115T and R116Q polymerization kinetics indicate that mutations destabilize monomer-monomer contact without significantly disrupting filament structure.
TAAD actin mutations in patients are autosomal dominant with expression of both a wild type and mutant allele. To determine the effect of wild type actin on the mutant phenotype in vitro, mutant actin was co-polymerized with increasing amounts of wild type actin, keeping the total actin concentration constant (Fig. 5, B and C). Increases in the proportion of wild type actin caused a commensurate normalization of the polymerization profile, suggesting a proportional rather than synergistic effect of the mutant actin on the wild type actin behavior.
Effects of the Mutations on Filament Stability-To further quantify filament instability that might be associated with these mutations, the minimum concentration of actin monomers required for polymerization, the C c was determined for N115T and R116Q actin ( Table 1). The critical concentration for both mutants was higher than for wild type actin; wild type was 0.47 M Ϯ 0.19 compared with 0.86 M Ϯ 0.37 for N115T and 1.03 M Ϯ 0.36 for R116Q (p Ͻ 0.05 for both). The increase in critical concentration indicates that both mutations cause filament instability.
Effects of Mutations on Actin Depolymerization-One determinant of actin filament stability is the ratio of the on-rate of actin monomers to the filament to the off-rate of monomers leaving the filament. An unstable actin filament would demonstrate either decreased on-rate or an increased off-rate. To assess if a faster off-rate contributes to the filament instability

TABLE 1 Effect of TAAD mutations on actin monomer and filament characteristics
Values are followed by the number of experiments performed, indicated in parentheses. Differences were statistically significant as compared with wild type actin, p Ͻ 0.05, except for binding of cofilin to N115T F-actin. NA, results not available due to denaturation of the R116Q mutant in the non-reducing environment required for the experiment.

Effects of ␣-Smooth Muscle Actin Mutations on Actin Function
APRIL 1, 2011 • VOLUME 286 • NUMBER 13 described above, the depolymerization rate was quantified. DNase I was added to polymerized actin to bind the G-actin that detaches from the filament ends. N115T actin showed an increased off-rate, with a t1 ⁄ 2 3.9 times faster than wild type, whereas R116Q is 7.8 times faster than wild type (p Ͻ 0.001 for both) (Fig. 3). These findings again indicate that both mutations result in filament instability. Alterations in P i Release-The alterations in mutant filament stability and dynamics can manifest by effects on ATP hydrolysis and P i release during polymerization. Yeast actin releases inorganic phosphate immediately following ATP hydrolysis. P i release for the N115T mutation initially followed polymerization, similar to wild type actin, but during the treadmilling phase, when monomer attachment at the barbed end is matched by monomer release at the pointed end, P i release continued more rapidly than wild type actin in a linear manner (Fig. 6). The phosphate release rate of R116Q is slower than that of wild type during elongation, corresponding to its nucleation lag phase, but then dramatically increases during the treadmilling phase. Increased phosphate release during treadmilling implies increased turnover of actin subunits during treadmilling or, alternatively, increased fragment formation, providing more substrate for the continuous addition of monomers onto the newly formed fragments. Both explanations are consistent with less stable mutant actin filaments.
Effects of the Mutations on the Actin-Cofilin Interaction-The altered monomer-monomer interactions leading to fila-ment instability may also affect the sensitivity of the mutant filaments to the severing activity of the actin-binding protein cofilin. Cofilin is a major regulator of filament dynamics in vivo that binds between two adjacent actin subunits in the filament, promoting disruption of intermonomer bonds. Cofilin also sequesters monomers, preferably when bound to ADP, as is the case when the monomer dissociates from the filament (44 -46). We examined the susceptibility of the wild type and mutant actin to increasing amounts of yeast cofilin (20). Yeast cofilin tends to decorate rather than sever actin filaments at stoichiometric levels, resulting in an increase in light scattering as seen with wild type actin in Fig. 7A. As cofilin was added to wild type actin filaments, light scattering increased at 2:1 and 1:1 actin/ cofilin concentrations, attributable to cofilin binding and decoration of the filament (Fig. 7A). As cofilin concentrations increased beyond 1:1, light scattering decreased, consistent with filament severing and monomer sequestration. At the highest concentrations of cofilin studied, little residual filamentous wild type actin can be identified by electron microscopy (Fig. 7B). The same analyses were done for each mutant. Remarkably, N115T and R116Q showed divergent responses to cofilin (Fig. 7A). R116Q behaved as initially hypothesized and was hypersensitive to cofilin. At a 3:1 actin/cofilin ratio, R116Q showed a 20% drop in light scattering, consistent with filament severing. R116Q light scattering decreased further at increasing concentrations of cofilin, when wild type filaments were abundant, dropping to prepolymerization base-line values at actin/ cofilin ratios greater than 1:1. As an additional control, F-buffer was added instead of cofilin, which verified that changes in light scattering are not due to simple mixing of the samples. In stark contrast, the response of N115T actin to cofilin was nearly opposite that of R116Q. N115T showed an initial decrease in light scattering at lower concentrations of cofilin with a steady drop to 60% of initial light scattering by the benchmark 1:1 concentration (Fig. 7A). However, as the concentration of cofilin was raised, the light scattering increased as well, suggesting cofilin decoration and stabilization of intact filaments. At cofilin concentrations 2.5 times that of actin, well past concentrations inducing total disassembly of both wild type and R116Q filaments, light scattering values were more than 75% of initial light scattering values.
Electron micrographs of wild type, N115T, and R116Q actin filaments confirmed the spectrophotometric data (Fig. 7B). At 1:1 actin/cofilin concentrations, wild type actin filaments were thicker with an irregular surface, indicative of decoration. At higher concentrations of cofilin, only small aggregates were observed. R116Q showed hypersensitivity to cofilin severing, demonstrating no recognizable filaments at cofilin concentrations equivalent to or higher than actin. In contrast, N115T filaments persisted at a 1:1 actin/cofilin ratio and beyond. Filament lengths for each actin were measured from electron micrographs at the varied cofilin concentrations. As expected, the average filament length corresponded to the light scattering values (supplemental Fig. 3).
To ascertain if mutant actin exhibits a dominant effect on the interaction between cofilin and wild type actin, the response of mutant and wild type heteropolymers was assessed, as above. The change in maximum light scattering of polymerized filaments due to the addition of cofilin was quantified for the range of mutant/wild type actin ratios. Similar to the isolated polymerization data, the response of R116Q/wild type actin to cofilin was muted in a dose-related manner by the addition of wild type actin, indicating a proportional effect of the mutant. Likewise, the addition of N115T to wild type actin produced a protective effect proportional to the mutant molar fraction (Fig. 7, C-E). Filament morphology as determined by electron microscopy again supported the spectrophotometric findings (data not shown).
Effects of Mutations on Cofilin Binding-To assess the relative contribution of altered cofilin-dependant severing versus sequestration, the affinities of cofilin for ATP G-actin and F-actin were determined for wild type and N115T mutant actins. The binding of cofilin to R116Q actin could not be analyzed because the elimination of DTT from solution needed for reaction of the actin with pyrene-maleimide resulted in protein denaturation. The K d for wild type G-actin was 0.8 Ϯ 0.4 M, whereas the N115T mutant actin K d was 2.5 times lower at 0.3 Ϯ 0.1 M ( Table 1). The increased affinity did not translate to the cofilin-F-actin interaction because the K d of cofilin to wild type actin filaments, 0.9 Ϯ 0.2 M, was similar to N115T, 0.7 Ϯ 0.2 M. Thus, the persistence of N115T filaments did not result from altered affinity of cofilin for mutant F-actin.

DISCUSSION
The data presented begin to elucidate the biochemical basis by which mutations in ␣-SM actin lead to TAAD. This is the first attempt to address the molecular mechanism of this cardiovascular disorder. Yeast actin was used for analyses, given the high homology with ␣-SM actin and opportunity for in vivo/in vitro correlation. Although it would be desirable to study the effect of these mutations in ␣-SM actin per se, we have not yet been able to express this actin isoform in a baculoviral system despite repeated attempts. The two mutations chosen for study, N115T and R116Q, occupy adjacent positions on a secondary structural element in actin thought to be intimately involved in the monomer-monomer interface necessary for actin filament formation (47)(48)(49). Despite their immediate proximity, the mutations have different clinical phenotypes. R116Q is the second most common identified mutation in patients with TAAD and is interestingly associated with premature coronary artery disease. N115T, on the other hand, is a less prevalent mutation and has been associated with stroke but not coronary artery disease (7,8). Thus, our goal was to not only gain insight into positions 115 and 116 in actin behavior but also understand the mechanisms underlying the distinct clinical phenotypes.
In summary, our data indicate allele-specific effects of mutations on actin behavior both in vivo and in vitro. The N115T mutation produces more deleterious effects in yeast cells, but the R116Q mutation leads to more striking effects in vitro. Growth under stress conditions was generally worse with the N115T actin because a greater percentage of cells displayed abnormal cytoskeletal morphology. Conversely, the R116Q mutation results in more dramatic effects on actin monomer behavior and polymerization, especially relative to filament nucleation and the critical concentration needed for polymerization. Although allele-specific differences were observed in vitro, both mutations displayed filament instability based on four criteria. First, both mutant actins exhibited faster disassembly rates in the presence of DNase I. Second, in vivo effects of latrunculin A, a drug that facilitates actin depolymerization, signify decreased cytoskeletal stability in the mutant cells. Third, the rate of treadmilling, based on the rate of P i , was greater for both R116Q and N115T. Finally, both mutant actins have increased critical concentrations.
The differences in clinical phenotypes described above may stem from the role of ␣-SM actin in the vascular wall, to transmit force applied by the myosin head. The severe instability of the R116Q actin monomer-monomer interface, as evidenced by marked alterations in polymerization kinetics, critical concentration, and depolymerization rates, would limit the ability of the actin filament to withstand this force, leading to disruption of the smooth muscle. The decreased stability of R116Q actin filaments in vitro is consistent with the absence of ␣-SM actin-containing filaments and an increased pool of unpolymerized ␣-SM actin in vascular smooth muscle cells from R116Q TAAD patient samples (7). Conversely, the N115T mutation may predominantly affect the interaction of filament regulatory proteins with the actin. A functional cytoskeleton requires rapid, reversible assembly and disassembly of actin filaments in a temporal and spatial pattern. The altered cofilin-N115T actin interactions indicate altered regulation of this dynamic process. Lack of published histopathology involving the N115T mutation makes the correlation between the current model system and the disease unavailable. The differences emphasize that further study of the impact of mutations on actomyosin function is needed.
Vascular smooth muscle cells from TAAD patients should contain a mixture of both wild type and mutant ␣-SM actin due to the autosomal dominant nature of the disease. We assessed the effects of mutant actin on wild type actin behavior to determine if the mutant actin exerted a dominant or proportional influence. Actin behavior was altered at all ratios of mutant and wild type actin in a manner proportional to the molar fraction of mutant actin. At a 1:1 ratio of mutant to wild type actin, the mixture that should be present in TAAD vascular smooth muscle cells, we observed significant effects on filament behavior.
The purification process of the mutant actins revealed that the R116Q mutant actin is more likely to denature than wild type or N115T actins. Further experimentation revealed that the major reason for R116Q destabilization was an increased sensitivity to oxidative damage because omission of the reducing agent DTT from purification steps led to protein denaturation. This observation suggests that a factor contributing to R116Q disease pathology may be increased sensitivity to oxidative stress, especially if inflammation and the attendant production of reactive oxygen species are involved in the dis-ease process. Another well documented case for oxidative damage to actin has been seen in patients with sickle cell disease in which erythrocyte actin undergoes oxidative intramolecular disulfide cross-linking, leading to abnormal cytoskeletal behavior (50).
The alteration of the monomer-monomer interface in mutant F-actin can influence interactions with actin regulatory proteins. Our data indicate that both N115T and R116Q mutations affect actin regulation by the actin-binding protein, cofilin. Cofilin plays a central role in modulating cytoskeletal dynamics in vivo via its ability to bind to and sever F-actin and bind and sequester ADP G-actin. Cofilin has two binding sites on actin, one of which encompasses actin residues 105-132. Furthermore, the 112-125 helix, in which the two mutations are located, interacts strongly with cofilin based on the results of ELISA and fluorescence experiments (51). Unexpectedly, the two mutations had opposite effects on the sensitivity of F-actin to cofilin in vitro. R116Q actin was much more sensitive to yeast cofilin than wild type actin. Conversely, N115T filaments were hyposensitive to severing by cofilin. N115T filaments remained largely intact, even at the highest cofilin concentration tested. Part of the ability of cofilin to sever actin filaments has been attributed to the twist that cofilin binding imparts on the F-actin helix, inducing strain on monomer-monomer interfaces. The cofilin resistance displayed by the N115T mutation may result from alteration in the interface, producing more flexibility and a greater ability to accommodate the additional strain. The persistence of N115T filaments at high concentrations of cofilin cannot be explained by a resistance to cofilin sequestration because the binding affinity of cofilin for the N115T mutant G-actin is tighter than that of wild type. We have observed resistance to cofilin previously in the deafness mutant P332A but to a lesser extent. At an actin/cofilin ratio of 1:2.5, approximately half of the original P332A actin filaments persisted (P332A at 55% versus N115T at 80%) (20).
Examination of the actin filament, derived from the model of the actin trimer of Oda et al. (47,48) based on electron cryomicroscopy, provides clues to the different effects caused by the two mutations. Focusing on N115T effects, Fig. 8 shows that Asn 115 , Lys 113 , and Pro 112 of one monomer (red) form a tightly packed element that interacts with the C-terminal end of helix 191-199 on an adjacent monomer (gray). This prediction is supported by recent HD exchange data from our laboratory, which show decreased amide proton exchange in the 105-132 peptide following actin polymerization. This indicates that when comparing G-actin with F-actin, new contacts are formed between the 105-132 peptide and monomers of the filament such that this peptide is now protected from the solvent (52). Modeling suggests that substitution of threonine for asparagine at residue 115 decreases the packing density in this element (Fig. 8, compare A and B), creating a conformational change that may well alter the interaction with the opposing monomer.
Although adjacent to Asn 115 , the greater extent of filament instability induced by the R116Q mutation suggests that a different mechanism may be involved. The C-terminal peptide of actin, ending with residue Phe 374 , is involved not only in formation of a monomer-monomer interface but also in allosteric regulation of the dynamics of the actin filament (53). The model of Oda et al. (47,48) shows a packed array of residues in subdomain 1 encompassing Glu 107 , Arg 116 , Val 370 , and His 371 (Fig.  9). In this array, Arg 116 appears to participate in an ionic bond with Glu 107 , providing stability to this structural element. An R116Q alteration would not only interrupt this ionic interaction, but the substitution of the smaller glutamine for arginine would also destroy the packing of the element. This combination of effects could exert a substantial effect on actin C-terminal peptide behavior, leading to the altered filament kinetics and stability observed. Our hypothesis is supported by the fact that mutations in Val 370 in this element give rise to two other diseases in humans. In ␣-skeletal muscle isoactin, the V370F mutation causes a severe nemaline myopathy (54), whereas the  The enlargement shown depicts the hypothetical ionic interaction between residues Arg 116 and Glu 107 potentially lost by the R116Q mutation.
V370A mutation in ␥-cytoplasmic isoactin leads to early onset autosomal dominant deafness (55). Our attempts to introduce the V370F mutation into yeast actin resulted in lethality, whereas the V370A mutation resulted in an extremely compromised cell with many of the traits identified in the R116Q cells (21).
Establishing the pathophysiological basis for TAAD caused by mutations in ␣-SM actin is a step toward improving the identification and treatment of patients. Understanding the mechanisms of the disease requires elucidation of the effects of the mutations on the behavior of actin per se, its regulation, and the impact on actin-mediated processes within the cell. Our results demonstrate the utility of using the yeast/yeast actin model system to correlate the effects of the mutation in vitro with the biological changes caused by the mutant actin in vivo as the only actin in the cell. For the two mutants studied, we observed allele-specific effects in actin polymerization and regulation, which may account for the differences in disease phenotypes. Application of this approach to the remaining ␣-SM actin TAAD mutations should greatly improve our knowledge of the molecular effects of these mutations as well as provide insight into the dynamics of wild type actin.