Tubulin acetyltransferase discovered: Ciliary role in the ancestral eukaryote expanded to neurons in metazoans

The emergence of α- and β-tubulins from an FtsZ-like prokaryotic homolog represented a defining event in eukaryogenesis. It permitted, together with actin and molecular motors, the establishment of a cytoskeleton able to support dynamic events such as cell division and trafficking of vesicles and proteins. It also served as the foundation for the microtubule-based cilium, an organelle with motility and sensory functions that replaced the less versatile prokaryotic flagellum (1). As it turns out, a mechanism to stabilize the ciliary axoneme, involving acetylation of α-tubulin, evolved in the ancestral eukaryote (2–4). This reversible modification situated inside the microtubule lumen still operates in extant ciliated protists and is exploited by metazoans to support ancillary microtubule functions, particularly in neurons (5). The concurrent discovery of an α-tubulin acetyltransferase by two separate groups, published in PNAS (3) and in Nature (4), now paves the way for a detailed understanding of the cellular and physiological roles conferred by tubulin acetylation.

Discovery

Tubulin acetylation was first detected on the microtubule axonemes of Chlamydomonas motile cilia in 1985 (6). It would take nearly 20 y before the activity responsible for tubulin deacetylation, afforded by histone deacetylase 6 (HDAC6) and SIRT2, would be discovered (2, 5). Surprisingly, the factor(s) responsible for adding an acetyl moiety to the ε-amino group of K40 on α-tubulin have remained enigmatic until now (2, 5). The codiscovery of a bona fide tubulin acetyltransferase followed notably different paths.

Using a biochemical approach, Shida and colleagues (3) from the Nachury and Goodman laboratories copurified an uncharacterized protein (C6orf134) with the BBSome, a ciliary trafficking complex implicated in Bardet–Biedl syndrome. C6orf134 has one telltale feature, namely a cryptic Gcn5-related N-acetyltransferase (GNAT) domain found in histone acetyltransferases (HATs). In addition, the researchers had isolated a BBSome-interacting protein, BBIP10, that was important for tubulin acetylation, but unlikely to be the prime catalyst for this modification (7). Their suspicion that C6orf134 is the sought-after tubulin acetyltransferase is confirmed by demonstrating the ability of the human protein to acetylate α-tubulin in vitro, as well as in vivo, through siRNA knockdown studies. Hence, C6orf134 is assigned the name α-tubulin K40 acetyltransferase (αTAT) (Fig. 1).

Fig. 1.

Acetylation of α-tubulin by the tubulin acetyltransferase αTAT/MEC-17 emerged in the ancestral eukaryote, likely to promote cilium formation and stability. Additional functions evolved in metazoans, including neuronal roles in touch sensation. Tubulin acetylation is countered by the deacetylases HDAC6 and SIRT2.

On the basis of a predicted GNAT domain in the Caenorhabditis elegans MEC-17 protein, previously implicated to function in cells harboring acetylated tubulin (touch receptor neurons or TRNs), Akella and coworkers (4) investigated the potential role of diverse orthologs as tubulin acetyltransferases. In vitro studies with murine MEC-17 corroborated such an activity, and disruption of MEC-17 in Tetrahymena, human cells, and zebrafish revealed their requirement for tubulin acetylation in vivo.

Further substantiation of a conserved role for αTAT/MEC-17 in tubulin acetylation is sought by both groups, using C. elegans as a model system. Each realize that αTAT/MEC-17, encoded by a single-copy gene in most organisms, is joined by a paralog, W06B11.1 (atat-2), in the nematode. Indeed, the mec-17 and atat-2 C. elegans genes are shown to function redundantly in acetylating the sole α-tubulin bearing a K40 residue, MEC-12; whereas individual mec-17 or atat-2 mutants have reduced levels of acetylated tubulin (3), the double mutant resembles mec-12 insofar as having none (3, 4).

Of note, the Elongator complex, which contains a subunit (ELP3) with HAT activity, is purported to function as a major tubulin acetyltransferase (8). However, several studies, including those mentioned above, find evidence to the contrary (3, 4, 9). Its potential role in tubulin homeostasis thus necessitates further investigation.

Acetyltransferase Specific for α-Tubulin?

αTAT/MEC-17 consists mainly of a HAT domain, and the tubulin deacetylase HDAC6 includes histones as potential substrates. To address the substrate specificity of αTAT/MEC-17, Shida et al. (3) query its capacity to act on histones. Acetylation assays using purified components reveal that it effectively acetylates tubulin but not histone H3/H4 or a mixture of core histones; in contrast, a histone acetyltransferase, HAT1, does not modify tubulin. Satisfyingly, αTAT/MEC-17 increases axonemal microtubule stability (4), as expected, likely by preferentially acetylating polymerized α-tubulin over free α/β-tubulin heterodimers (3).

Although these data support the notion that αTAT/MEC-17 is a lysine acetyltransferase for tubulin and not histones, its substrate repertoire might encompass additional proteins. This possibility was suggested by Akella and coworkers (4), albeit with a caveat. Specifically, the group reported that loss of C. elegans αTAT activity in the mec-17;atat-2 double mutant affects TRN mechanosensation to a greater extent than having a nonacetylatable MEC-12(K40R) form of tubulin in the organism—implying additional role(s) for the acetyltransferase independent of tubulin acetylation. However, Shida et al. (3) a priori perform the same experiment but obtain the inverse result. In addition, loss of acetyltransferase activity phenocopied the mec-12 mutant in one study (4) but exhibits a less severe phenotype in the other (3). Further scrutiny of αTAT/MEC-17 in C. elegans and other species should resolve these discrepancies and potentially expose novel substrates.

Phylogenetic Distribution Suggests Ancestral Ciliary Role

The distribution of αTAT/MEC-17 across all eukaryotic clades (3) reveals that it was present in the last eukaryotic common ancestor, which was ciliated (1). Perhaps more revealing, the absence of this protein from organisms that secondarily lost cilia (3) suggests that its primeval and ongoing cellular function is linked to cilium biogenesis, function, and/or maintenance. Indeed, αTAT/MEC-17—but not several other ostensibly key ciliogenic proteins, including intraflagellar transport (IFT) components required for ciliary trafficking—is absolutely conserved in ciliated eukaryotes (3).

Conserved…Yet Nonessential Ciliary Function?

The prevalent population of acetylated microtubules in protists and nonneuronal cells occurs within the ciliary axoneme (2–5). Through guilt by association, a ciliary function for αTAT/MEC-17 is therefore expected, but what is the current evidence for such a role? Based on the available data, an essential ciliogenic role for the acetyltransferase seems unlikely. Although disrupting MEC17 in Tetrahymena abolished tubulin acetylation, no overt defect in cilium formation or motility was noted (4). Similarly, cilia are still present in zebrafish and mammalian cells subjected to morpholino- or siRNA-mediated knockdown of αTAT/MEC-17, respectively (3, 4). Interestingly, however, a timecourse experiment reveals a delay in cilium formation for cultured human cells depleted of αTAT/MEC-17 (3), a phenotype whose cellular or physiological consequence is unclear but deserves further attention.

Intriguingly, most C. elegans-ciliated sensory neurons lack detectable acetylated tubulin. Because MEC-12 is present in TRNs and other (ciliated) neurons (8, 10), this situation could arise because of restricted expression of the acetyltransferases in ciliated neurons. Shida et al. (3) now demonstrate that although mec-17 expression is limited to TRNs; its paralog atat-2 is also expressed in some ciliated neurons (including CEP and OLQ). Indeed, ATAT-2 is needed for acetylating tubulin in dendritic processes and cilia in those neurons (3). Understanding the differential requirement for acetylating distinct axonemal microtubules could prove to be informative.

Neuronal Functions

Adding to our understanding of αTAT/MEC-17 function, Shida et al. (3) and Akella et al. (4) demonstrate distinct neuronal roles for the nematode and zebrafish proteins. First, C. elegans MEC-17 and ATAT-2 are required for body touch sensation, which depends on the nonciliated, acetylated tubulin-containing TRNs (3, 4). ATAT-2 is also needed for nose touch avoidance and wild-type locomotion in the absence of food, behaviors modulated by mechanosensory-ciliated neurons (OLQ, CEP) containing acetylated tubulin that express atat-2 but not mec-17 (3). Second, knockdown of zebrafish MEC-17 causes various developmental phenotypes, some potentially linked to ciliary dysfunction (e.g., curved body shape, hydrocephalus, and small eyes) and another that may not be—a neuromuscular defect that impairs touch sensation (4). Thus, tubulin acetylation supports the function of at least some nonciliated neurons, highlighting an expanded role for αTAT/MEC-17 in metazoans.

Outlook

In the absence of unequivocal ciliary phenotypes following disruption of αTAT/MEC-17 in protists and metazoan cells, the challenge will be to uncover more subtle effects conferred by acetylation of axonemes. Given that the motility of at least one molecular motor (Kinesin-1) depends on tubulin modifications (including acetylation) in neurites (11), the search for a ciliary function could include probing for defects in kinesin-/dynein-driven IFT or, perhaps, transport of components to the base of (growing) cilia; dysfunction of either could potentially explain the slower ciliary growth rate observed upon abrograting αTAT/MEC-17 (3). The functional link between αTAT/MEC-17 and BBSome, and BBIP10, should also prove revealing. Notably, the finding of a ciliary transport-tubulin acetylation connection may parallel the recent discovery of a core IFT protein (DYF-1) associated with polyglutamylation activity—another tubulin modification that confers microtubule stability (12).

The most important discoveries bring with them many more questions than they answer. The unveiling of the tubulin acetyltransferase, coupled with the awareness of two reverse enzymes, HDAC6 and SIRT2, promises to clarify the cellular functions of this ancient modification pathway. For one, tubulin acetylation can now be better understood along with other tubulin modifications, including glutamylation, glycylation, and detyrosination (2, 5). Also, whereas HDAC6 has multiple substrates and binding partners, and acts in a broad range of cellular processes—e.g., cilium disassembly, malignant transformation, and misfolded protein degradation (13)—abrogation of αTAT/MEC-17 may reveal a more limited range of phenotypes uniquely associated with tubulin acetylation. In brief, the discovery of a conserved tubulin acetyltransferase will play an important role in understanding key aspects of the eukaryotic microtubule cytoskeleton.