Ancient Aminos

Ludueña’s first step was to theorize what this early tubulin ancestor looked like. In tubulin’s canonical role as structural building block, dimers composed of two isoforms—alpha and beta—stack up to form microtubules. The protein has several other isoforms, though, which serve organizational rather than structural roles. By comparing similarities in their amino acid sequences, Ludueña constructed a tubulin family tree, and hypothesized that alpha and beta, as well as a third isoform, gamma, clustered together while the other isoforms branched off.

What then, was this common ancestor’s function, Ludueña wondered? For a possible answer, he took a clue from FtsZ, which forms rings to allow a membrane to pinch off a bacterial cell. In other words, FtsZ can bend—a quality it just might share with alpha, beta, and gamma tubulin. According to still-unpublished data from McIntosh’s lab, which he presented in a recent talk at San Antonio, microtubules made up of tubulin alpha and beta can curl during mitosis. “They’re not forming rings, exactly, but they are bending,” Ludueña says. Similarly, gamma tubulin forms curved structures—specifically, the circular structure from which microtubules extend, and studies suggest that it too can exist in a curved formation. It’s therefore likely that the common ancestor was a bending protein rather than a microtubule-forming protein.

One of Ludueña’s key areas of study is tubulin’s post-translational modifications. Many of these, such as acetylation, are fairly run of the mill. But two—polyglutamylation and polyglycylation—are rare among other proteins. These modifications are quite important for tubulin, though: they create branched structures on its C-terminal end, without which microtubules don’t form cilia or flagella, likely some of microtubules’ most ancient functions. Both of these modifications involve attaching amino acids to glutamate residues without involving the cell’s protein synthesis machinery, and neither seems to target specific glutamates, he notes. “There’s something very primitive” about the loose way this process is regulated, he says.

“What I’m postulating is actually kind of crazy,” Ludueña says: Suppose that the C-terminus was once a separate protein whose role was to help the ancestral tubulin form cilia and flagella. He hopes to publish more data supporting this idea in the next year.

Zac Cande, who studies the evolution of cell division at the University of California, Berkeley, did a postdoc with McIntosh 35 years ago. “It all seems very interesting,” he says, having reviewed Ludueña’s discussion of the theory laid out in a trio of review articles present in a recent volume of the Cancer Drug Discovery and Development book series. FtsZ’s relationship to tubulin was so surprising because the functional link was elusive, he says. “But he has a way of reconciling that”—specifically, the notion of curved filaments—“and I think that’s a neat thing to do.” Cande cautions, though, that eukaryotes emerged about 2 billion years ago, and determining the qualities of a common ancestor so far in the past is a bit like looking into a crystal ball. “This is all unknowable stuff, in a way, but there’s nothing he says here that’s foolish.”