Last week, I wrote about Takashi Hamaji’s new paper characterizing the mating-type/sex-determining loci in Eudorina and Yamagishiella. That paper showed that the sex-determining region of anisogamous Eudorina is, surprisingly, considerably smaller than the mating-type loci of isogamous Chlamydomonas, Gonium, or Yamagishiella. Because only one gene, MID, is present in the male version of the sex-determining region in Eudorina, Hamaji and colleagues concluded that
…the evolution of males in volvocine algae might have resulted from altered function of the sex-determining protein MID or its target genes.
I commented that
…we’re still left with two (non-mutually exclusive) possibilities: changes to the MID gene itself may have changed which genes it interacts with (or how it interacts), or there may have been changes in the genes whose expression is controlled by MID.
Now Sa Geng and colleagues have provided at least a partial answer. In a new paper in Development, they swapped versions of MID among different volvocine species* (unfortunately, no unpaywalled version of the paper is currently available; I will add a link when I find one). We already knew that MID is necessary and sufficient for male development: genetically male Volvox carteri colonies that have MID expression turned off produce eggs, and genetically female colonies transformed with MID produce sperm packets (“Sex change (in Volvox)”). But that’s MID from the same species. It’s somewhat surprising that a single gene can cause Volvox to switch sexes, but at least Volvox MID evolved side-by-side with the genes whose expression it controls.
What would be really surprising is if MID from other species, species that diverged from the Volvox lineage ~200 million years ago, worked in Volvox. It would be extraordinarily surprising if MID from a species that doesn’t even have males could control their development in Volvox. It won’t work. Waste of time; don’t bother trying.
That is, I think, pretty much what I would have said if Geng and colleagues had told me what they were planning. I might have rolled my eyes. I might have used the word “crazy”. Fortunately, they were bolder than I’d have been, because all these surprising things turned out to be true.
Geng and colleagues took the MID gene from male Pleodorina starrii and put it into genetically female Volvox carteri, which subsequently produced functional sperm packets. The transformed Volvox were actually hermaphrodites, producing sperm and eggs in various ratios. We know the sperm packets are functional, because they fertilized the eggs.
What they’re calling “Eve:: PsMID-BH” (panels D & E) is the female Volvox strain (Eve) transformed with the MID gene from male Pleodorina starrii (PsMID) and some markers (BH; see the paper for technical details).
This is the less surprising result. Pleodorina starrii has males and females (they are anisogamous, in other words) and they diverged from Volvox carteri on the order of 100 million years ago. Nor did the Pleodorina MID do a perfect job of controlling male development:
…PsMID-BH-expressing strains had spermatogenesis-related developmental defects, including sperm cell morphological abnormalities and delayed sexual development. They also showed delayed or incomplete hatching of sperm packets from their vesicles that was more severe than the delayed vesicle hatching in VcMID-BH transformants. However, unlike sperm packets from VcMID- BH transformants that were eventually released from their mother spheroid, most of the PsMID-BH transformant sperm…were unable leave the mother spheroid and remained trapped within a partially dissolved sperm packet vesicle.
VcMID-BH transformants are the same strain of Volvox carteri (Eve) transformed with MID from Volvox carteri males, so essentially what they are saying here is that MID from the same species does a better job of controlling male development than MID from Pleodorina starrii.
This is on the spectrum of what we might expect. It wouldn’t be terribly surprising to learn that most of the changes to MID and the genes whose expression MID controls happened around the time that the Volvox carteri / Pleodorina starrii lineage evolved anisogamy. In that case, Volvox MID and Pleodorina MID might not be all that functionally different, allowing one to substitute for the other. There are bound to be some differences, though, and those would explain the developmental defects observed in the transformed Volvox.
More surprising, to me at least, is that MID from Gonium can also control male development in Volvox. Gonium and Volvox last shared a common ancestor around 200 million years ago, twice as long as did Pleodorina and Volvox. More importantly, though, Gonium is isogamous. There are no male and female Gonium, only morphologically indistinguishable mating types (see “Origins of the sexes: isogamy and anisogamy“). Gonium never produces sperm, and there’s no reason to think that they or their ancestors ever did. Nor is there any reason to think that Gonium would have a genetic pathway for producing sperm.
Nevertheless, when Geng and colleagues transformed the MID gene from Gonium pectorale into genetically female Volvox carteri, the transformed colonies produced sperm packets!
The terminology here is similar to that for Figure 2 above: Eve::GpMID is the female Volvox strain (Eve) transformed with MID from Gonium pectorale (GpMID), and BH denotes added markers. cDNA just means that that strain was transformed with DNA complementary to a mature mRNA, in other words lacking introns (see the paper for more detail on the Methods).
As with the Pleodorina-transformed strains, the Gonium-transformed strains developed as hermaphrodites, that is they produced both sperm packets and eggs. The resulting developmental defects were more severe than in the Pleodorina-transformed strains:
Besides formation of ectopic sperm packets and self-fertility (Figs. S7A-S7C), the additional phenotypes included sperm cell developmental abnormalities and incomplete or delayed hatching from the sperm vesicle that were even more severe than what we observed for PsMID transformants: In GpMID transformants most sperm packets never hatched from their vesicle, but instead dissolved within it; and we never observed sperm or sperm packets released from parental spheroids…
Besides the sperm development and hatching abnormalities observed in GpMID-expressing transformants, all of the different GpMID constructs we tested caused additional sexual-stage and vegetative-stage phenotypes that were never observed in PsMID or VcMID transgenic strains. These included smaller-sized vegetative and sexual spheroids with fewer vegetative gonidia or sexual germ cells than in control strains. We also observed around 30-50% of the vegetative and sexual spheroids that expressed GpMID had unevenly spaced and disorganized somatic cells, and around 30-40% of the vegetative-phase spheroids from the same strains had a misshapen appearance that may have been caused by incomplete or aberrant inversion—the post-cleavage reversal of embryo curvature that turns the spheroid right-side out.
Even so, I found this result astounding. How could a gene from a species that doesn’t even have males cause another species, from which it diverged 200 million years ago, to produce sperm? In retrospect, I probably shouldn’t have been as surprised as I was. In the larger context of evo-devo, stranger things have happened, as Geng and colleagues remind us in the Discussion:
We note that functional divergence and loss of inter-specific compatibility between transcription factor orthologs is not an inevitable outcome over the time scale of 200-300 MY during which volvocine algae diversified. For example, the the well-known metazoan Eyeless/PAX6 transcription factors retained function in eye development between fruit flies and mice over 800 MY despite the evolution of anatomically completely different eye structures between arthropods and vertebrates. Similarly, an ortholog of the circadian transcription factor CONSTANS (CO) from Chlamydomonas was able to complement an Arabidopsis CO mutant, though the two CO orthologs share significantly less sequence similarity (32% identity/38% similarity) than the most diverged Mid orthologs in volvocine algae (41% identity/52% similarity), and exhibit a much deeper phylogenetic divergence of around 1000 MY. [references omitted]
So what’s it all mean? Going back to the question I raised last week, these results considerably narrow the range of answers. Before this study, it seemed entirely possible that the evolution of anisogamy in the volvocine algae was mainly due to changes in the MID gene itself. MID from isogamous Chlamydomonas reinhardtii does not cause female Volvox to produce sperm, which is consistent with this idea. But if that were the case, we would not expect MID from Gonium, which has never produced sperm, to cause sperm production in Volvox:
…the inability of ectopic CrMID or chimeras between CrMID and VcMID to induce spermatogenesis in V. carteri females led to a simple hypothesis that the evolution of sexual dimorphism and oogamy in the volvocine lineage was due to molecular changes in the Mid protein (Geng et al., 2014). The experiments described here rule out the simplest version of this hypothesis because GpMid from the isogamous species G. pectorale could induce spermatogenesis when ectopically expressed in V. carteri females, albeit at reduced efficiency…
Our data on cross species complementation with Mid instead support a phyletic model in which Mid functional divergence was not coupled to the evolution of gamete dimorphism, but occurred in a more gradual manner that was proportional to divergence times between lineages. Under this model the Mid protein from the Chlamydomonas lineage has accumulated too many changes for it to function in the multicellular volvocine taxa (and vice versa), but the core interactions between Mid, its target DNA sequences, and other transcriptional regulators have been sufficiently conserved within the multicellular taxa (Volvox, Pleodorina, Gonium) for it to retain basic function across these genera.
Thus it now seems more likely that the genetic changes underlying the evolution of anisogamy were mostly changes not to MID itself but the the genes whose expression it controls. It will be fascinating to learn what those changes were and how they cause the production of sperm. I would argue that the volvocine algae are already the model system in which we know the most about how the sexes evolved (see here, here, here, here, and here), but we still have a ways to go to a complete understanding.
*EDIT (2019-01-14): The Development paper is freely available from the doi link. I was mistaken about it being paywalled, so I have struck that text.
Geng, S., De Hoff, P. and Umen, J.G. 2014. Evolution of sexes from an ancestral mating-type specification pathway. PLoS Biology, 12, e1001904. doi: 10.1371/journal.pbio.1001904
Geng, S., Miyagi, A. and Umen, J.G. 2018. Evolutionary divergence of the sex-determining gene MID uncoupled from the transition to anisogamy in volvocine algae. Development, dev.162537. doi: 10.1242/dev.162537