Mutant vascular actin is a TAAD misbehaving

Actin plays a number of important roles in mammalian cells. Mammals express six different isoforms of actin in a tissue- and temporal-specific manner (1). Mutations in each isoform have been linked to specific diseases in humans, and in most instances a mutation in one of the two copies of the gene is sufficient to cause the disease (2). Mutations in α-smooth muscle isoactin (αSMA) cause vascular defects leading primarily to thoracic aneurysm and dissection (TAAD). In PNAS, Lu et al. (3) addresses two of these mutations, primarily R258C and secondarily R258H, to try to gain insight into how changes induced by these mutations lead to TAAD by affecting actin function. Vascular material from an affected individual carrying the R258C mutation shows that the muscle layer of the vessel wall is abnormal, with disorganized structure and a dearth of contractile filaments compared with normal smooth muscle, suggesting abnormal filament assembly, protein instability, or failure of the assembled cytoskeleton to withstand the forces imposed on it by myosin during contraction of the muscle (4, 5). However, the advanced disease stage at the time of tissue resection prevents early stages of the disease from being addressed. This same site has been mutated in other actin isoforms, in each case leading to a disease. An R256 mutation to either H or L in α-skeletal muscle isoactin causes nemaline myopathy and an R256W mutation in β-nonmuscle isoactin causes Baraitser–Winter Syndrome (2).

Normal blood vessel wall function requires normal kinetics of actin polymerization. The filaments must have normal enough intermonomer contacts to allow formation of structures with normal dynamics. The structure of the filament must allow proper control by the numerous regulatory proteins with which it interacts. Finally, the interaction of smooth muscle myosin with the actin filament must produce enough contractile force for normal regulation of the circulation. An actin mutation can potentially affect any of these parameters. Gaining insight into the effects of these mutations at a biochemical level requires sufficient pure mutant actin isoform for biochemical experimentation. The usual approach incorporates a model expression system in which one can express a mutated protein and isolate it in a chemically and isoform pure state. For studying vascular smooth muscle function, it would be best to use smooth muscle proteins, including αSMA. These proteins have evolved together either within a specific cell type or within an organism to maximize functional efficiency of that particular system. Disease could result from quantitative differences in the output of these systems, so using model hybrid systems based on protein homology could produce misleading data in terms of gaining insight into a disease process.

Because of technological problems in obtaining mutated vascular smooth muscle actin, initial studies into the effect of TAAD-causing actin mutations first used the budding yeast Saccharomyces cerevisiae. Actin from budding yeast is 86% identical to αSMA in terms of amino acid structure (6), and all of the sites at which disease-causing mutations occur in the smooth muscle actin are conserved in the yeast protein. One of the mutations studied with yeast was the R258H mutation (R256H in the yeast nomenclature). Malloy et al. (7) demonstrated that this mutation produced adverse cytoskeletal behavior in vivo and that the mutated actin formed less stable filaments and was subject to abnormal regulation by yeast formin Bni1p. Further observation suggested that R258 was part of an interstrand ionic system within the filament interior, also containing K115 and E197, which could be allosterically regulated by the binding of proteins, such as Arp2/3 and formin, on the filament surface (8).

Lu et al. (3) make a significant achievement in establishing an in vitro system for studying the effects of TAAD-causing mutations in αSMA per se (Fig. 1). Previous attempts to express this isoactin in baculovirus-infected cells had been unsuccessful. This was particularly surprising because α-cardiac, α-skeletal, and β- and γ-nonmuscle isoactins had been expressed using this system. However, by using a system for expressing toxic actins in Dictyostelium (9), Lu et al. produced enough WT and mutant αSMA to carry out their experiments. This achievement not only allowed the investigators to establish the effects of the mutation on the actin actually present in the diseased cell, but it also provided an opportunity to compare the results obtained with yeast vs. smooth muscle actin in cases where similar experiments were performed.

Fig. 1.

Location of the TAAD mutations at R258 in actin. In A and B, the site of the mutation is shown in red. (A) Monomeric or G-actin. The four subdomains of the protein are denoted. The area of the monomer to which profilin binds is denoted. (B) A trimer of actin monomers in the context of an actin filament. Each monomer is denoted by an m number. The binding site for myosin bridges the gap between two longitudinally apposed monomers.

Lu et al. (3) show that the R258C mutation in smooth muscle actin resulted in a decreased rate of polymerization compared with that of WT actin, a less stable filament, and increased susceptibility to severing by the filament-severing protein cofilin (10). Earlier yeast work showed that polymerization rates and filament stability were similarly affected by the corresponding R258H mutation. Although in the yeast paper, cofilin-severing was not studied, earlier studies with two other TAAD mutations at positions 115 and 116 in a helix leading to the interface involving R258 produced filament weakening and increased cofilin susceptibility. Lu et al. (3) emphasize that the mutation produces effects far distant from the actual site of the mutation, implicating a high degree of protein allostery. Earlier, Malloy et al. (7) and later Rubenstein and Wen (2) proposed an allosteric model in which a helix containing residues 115 and 116 acted as an allosteric conduit connecting the filament surface with the triad of residues containing R258, E197, and 115 on the other strand, which act as a major determinant of filament stability.

Lu et al. (3) did find a striking difference between the yeast and smooth muscle actin regulation by formin. With yeast mutant but not WT actin, the actin-binding portion of formin acted as a strong filament capper. Lu et al. did not observe such behavior with the R258C or R258H mutation in smooth muscle actin in the presence of the FH1-FH2 fragment from mammalian mDia1 formin. It is not known, however, whether this difference in results reflects inherent differences in the yeast vs. muscle actins or in the yeast vs. muscle formins or both. This result shows that where a disease is the focus of investigation, it is always better to use the actual proteins found in those cells to lessen concerns about potential quantitative differences that may lead to errant conclusions regarding disease mechanism.

Lu et al. (3) obtained a very surprising result in their study of how the R258C mutation affects the actin-profilin interaction. Profilin is a low molecular weight protein that forms a 1:1 complex with monomeric actin, either facilitating its incorporation into a filament or acting as a buffering agent for actin monomers, depending on the actin:profilin ratio (10). The mutant actin bound much more tightly to profilin than did the WT actin, a significant result because the profilin binding site is distant from the site of the mutation on the actin surface. Such allostery is consistent with earlier findings about actin based on results with HD exchange experiments (11). Based on the entirety of this work, one can infer that mutation-induced effects on actin filament assembly and stability could either interfere with the proper construction of the smooth muscle contractile apparatus or the filament’s ability to withstand myosin force exerted during contraction.

Because a mutant actin-dependent defective actomyosin interaction could cause vessel wall weakening, Lu et al. (3) used the smooth muscle myosin heavy myromyosin fragment to address this question. In an in vitro motility assay where filaments are moved by tethered myosin without being restrained by a resisting load, little if any difference was evident between the WT and R258C or R258H actins alone. Addition of smooth muscle tropomyosin to the actin filaments resulted in increased motility of the WT filaments vs. WT filaments alone. Tropomyosin, which binds along the actin filament, slowed the movement of either mutant filament compared with the mutant filament alone. No difference was observed, whether the R258C or H mutant actin was used. In laser-trap experiments, which show how applied resistance affected the velocity by which myosin could move the filaments, there was little if any difference in the force observed with WT or mutant actin alone or WT+tropomyosin. However, velocity was greatly reduced at all applied loads with the tropomyosin-coated mutant actin filaments. These results are highly significant, considering the high abundance of tropomyosin in the vascular smooth muscle cell.

The dominant nature of TAAD mutations means that there is one copy of a WT αSMA gene and one copy of the mutated gene per cell. Therefore, to a first assumption, there is likely to be equal amounts of the mutant and WT smooth muscle actins present. Also likely, based on in vitro studies, is that the actins will copolymerize into the same filament unless prevented from doing so by compartmentalization or by isoform-specific regulatory proteins. It is thus important to determine to what extent, if any, the WT actin can restore the mutant actin to more normal behavior. Lu et al. (3) show that a 50:50 mixture of the two actins seemed to polymerize faster than what would be predicted from the rates of each actin alone. In the total internal reflection fluorescence assay, velocity slowed as the percent mutant actin increased from 0 to 75%. However, it was difficult from the data to ascertain if the rate of movement was linearly dependent on the mutant percentage in the sample. The potential adverse effects of the mutation could be dampened even more by the nonmuscle actins also present in the cell.

The use of pure proteins defines the parameters that must contribute to the disease process in the cell. Ultimately, though, understanding the molecular mechanisms leading to vascular wall malfunction will require being able to follow the development of the disease from its earliest phases. Such experiments will require an animal or cultured vascular smooth muscle cell model. A whole-animal αSMA knockout mouse has been made that lives, reproduces, and displays hemodynamic defects, but aneurysms were not reported (12). A mouse carrying a homozygous or heterozygous TAAD mutant αSMA has not been reported. Nor is there yet a report for any cultured rodent aortic smooth muscle cell system that would allow investigation in the development of the disease at the cellular level. Ultimately, an understanding of the molecular mechanisms underlying actin-dependent TAAD is going to require a combined in vitro approach as represented here by Lu et al. (3) coupled with research at the cellular and animal levels as the appropriate model systems are developed.