In search of the primordial actin filament

One paradox of evolution is the actin filament, which is an obligate right-handed, double-stranded helical filament in eukaryotes, yet forms diverse architectures in prokaryotes (1) (Fig. 1). Uncovering the origin of this asymmetrical distribution in filament morphologies is fundamental to understanding the emergence of the domains of cellular life. The variation in actin amino acid sequences magnifies the diversity in filament structures (Fig. 1). Eukaryotic actins are far more highly related than their parent genomes, whereas prokaryotic actins, particularly plasmid-based actins, are uncommonly diverged and are often difficult to identify from sequence-based homology searches (Fig. 1). Despite this variance, two features are preserved between prokaryotic and eukaryotic actins. The first common feature is the ATP-binding site, which operates as an ATP-hydrolysis and phosphate-release controlled conformational switch that is activated by polymerization. The ATP switch acts as a timing mechanism to coordinate depolymerization, conferring on actins the ability to polymerize dynamically and, subsequently, to depolymerize. The second feature is the conservation of in-strand protomer contacts, particularly the association between subdomains 3 and 4, which have been observed in all filament structures determined to date. In contrast, association between strands involves different surfaces of the protofilaments to generate the diversity in multistrand filament architectures. This maintenance of contacts within a strand suggests that the primordial actin filament was single-stranded. In PNAS, Braun et al. (2) determine the structure of the crenactin filament from the Archaea Pyrobaculum calidifontis. Crenactin forms a single-stranded filament, and thus is a candidate present-day record of the lost ancestor of eukaryotic actin.

Fig. 1.

Relatedness of actins. The structures of actin protofilaments (2, 9–15) are shown below a maximum-likelihood phylogenetic tree of the actin protein sequences. The structures are aligned via the central protomer, which is shown as a ribbon (slate blue) with the nucleotide in red. The subdomains are numbered for actin. Subsequent protomers are shown as surfaces (alternating between fawn and slate blue). For dynactin, actin-related protein 1 (ARP1) is in slate blue or fawn, actin is in black, and ARP11 is in red. The number of protofilaments that comprise each filament is indicated in blue. BLAST (16) protein sequence similarities to human skeletal actin are given in red. BLAST (16) searches with the human skeletal actin sequence align less than half of each prokaryotic actin sequence.

Actin sequences that have recently been identified in Archaea are more closely related to eukaryotic actin than to prokaryotic actins (3, 4). Indeed, the closest homologs are found in Lokiarchaeota, which also possess homologs of the actin-severing protein gelsolin, suggesting a potential regulated actin filament system (5). Crenactin (crenarchaeal actin) from P. calidifontis is more distant than the Lokiarchaeota sequences (Fig. 1) and forms a cell shape-determining cytoskeleton. Previous structural studies had revealed that crenactin forms a protomer structure that is highly similar to actin, and a single-stranded “protofilament” was observed in the crystal packing (6, 7). Low-resolution electron micrographs (EMs) of crenactin filaments also indicated a structure that is consistent with either single- or double-stranded filaments (6). Now, Braun et al. (2) definitively show in an 18-Å cryo-EM reconstruction that crenactin can form a single-stranded filament in vitro.

In determining whether crenactin represents a record of the ancestral eukaryotic actin filament, its binding partners and functions must be considered. Crenactin appears to interact with a unique set of proteins, the arcadins (3, 4). Crenactin’s role as a cell shape-determining cytoskeletal element has some parallels with the bacterial actin homologs MreB and FtsA, and the tubulin homolog FtsZ. Thus, crenactin may be a dedicated cell shape-determining actin. Evidence from bacteria suggests that when actin filaments have a single vital function, their sequences and filament arrangements have evolved to become optimized for that function (1). Consequently, one possible reason why crenactin forms a single-stranded filament may be that this conformation is optimal for its role in cell shape determination, and that it represents a highly evolved dedicated structure in the same way as the bacterial actins.

Eukaryotic actin has followed a different path, relative to prokaryotic actins, during evolution, in that one filament system provides force and form to many cellular processes (1). In this system, a common pool of actin is tapped by each cellular process through dedicated nucleation machineries. Thus, the actin pool is maintained at a level above that needed for any individual process. Many of the regulating proteins, including those regulating proteins that maintain the pool (profilin, cofilin, twinfilin, and cyclase-associated protein) (8) and access the pool (formins and ARP2/3) are conserved between humans and yeast (1). Hence, eukaryotic actin is a highly connected hub, which, at least in part, explains its conservation. P. calidifontis does not contain homologs of the eukaryotic actin-regulating proteins, such as profilins or cofilins. Thus, at this stage, it is not possible to determine whether the crenactin structure is a record of the single-stranded ancestor of eukaryotic actin or rather a highly evolved dedicated structure. Many more archaeal genomes will be needed to make that assessment. Nevertheless, the fascinating discovery that crenactin forms a single protofilament (2) demonstrates that the eukaryotic actin-like fold is capable of supporting stable single-stranded filaments.