But how a paramecium does it turns out to be a long and tangled tale. It was finally untangled last month in the cover story of the 22 May Nature journal.
Back before the Beatles, when I was taking high school biology, scientists already knew that paramecia, and other ciliate protozoans (they didn't call them protists yet) lead lively sex lives. There are no identifiable males or females, but every paramecium belongs to one of two mating types, arbitrarily designated even and odd, or E and O. Every now and then, one E and one O will lock membranes and randomly swap one half of their genetic material; when the swap is complete, they break apart, and each promptly divides twice to produce four daughter cells. All the E's daughters are type E; all the O's daughters are type O.
(I might note that there's a heckuva sci-fi story here, about a society in which any sexual act results in each partner walking away with one half of their own identity and one half of their mate's. The narrative implications would be staggering. For one thing, how will they ever be able to divvy up their CD collections now that they have half of one another's tastes?)
Ah, but wait a minute! If the genetic swap was really half and half, and really random, any gene for E-ness or O-ness, you'd think, could have wound up with either parent. So the long standing puzzle was: how does each parent hang on to his, her, or its sexual identity?
It all hangs on some surprisingly elaborate machinery. And it affords wonderful fodder for believers in intelligent design - so long as they're allowed to postulate that the designer in question was Rube Goldberg. It also spells out the very particular conditions under which a paramecium will ever mate with a paramecium of the same gender.
The sordid details are beyond the orange slipper animalcule.
The first thing to understand is that every ciliate is blessed with not one nucleus, but two nuclei. The micronucleus is the repository of the genome. The macronucleus - which at least in principle contains the same DNA, but exploded into a plethora of short chromosomes, which makes it many times the size of the other - is the somatic nucleus. Only it will interact with the cell's messenger RNA to produce proteins.
When two paramecia conjugate, only the micronucleus is involved in the swap. And it's carried out in such a way that when the swap is over, the two individuals have exactly the same genetic contents in their micronuclei. They are identical twins. (See footnote 1 if you're curious how that gets arranged.)
So, you might figure, what remembers the parent's gender must be the macronucleus. Unfortunately for that theory, as soon as the parents separate, the macronucleus disintegrates. Then a new one is built by copying the DNA from the brand new micronucleus. (Okay, that's a very slight oversimplification. First the little nucleus splits three times - by mitosis. Now there are eight of them. Four of them become macronuclei, undergoing a lot of hanky panky that splits chromosomes apart, edits out some DNA and scrambles the rest. Then in the process of division one little nucleus and one big one are doled out to each daughter cell.)
Now we come to the new information. There is a single protein, dubbed mtA, which the paramecium can display on the surface of its membrane, making it into mating type E. The O gender doesn't display it. Both genders, however, have the gene for it. That gives us a clue: we should look to the cell's gene-silencing personnel, the short RNAs in the cytoplasm.
Here is what happens:
- As soon as the new micronucleus is established, it throws off tens of thousands of short RNAs - a pool representing every bit of its genetic material.
- The parent match: Before the parent's original macronucleus is destroyed, those are compared to the parent's DNA, and whatever short RNAs match are pulled out of the pool.
- The non-parent deletion: As the new macronucleus is being built, whatever properly matches remaining short RNAs is edited out. (Here the article was not quite explicit enough for me to follow. It seems that the editing machinery mostly ignores matching genes, deleting pieces that look like viral transposons, and can affect gene expression.) So the new macronucleus will in many ways mimic the parent.
- The gene for mtA (that type E protein) requires an attached stretch of DNA that acts as a promoter, permitting the expression of the gene. Both genders have that promoter in their micronucleus, but it's missing in the macronucleus of the O type.
- If the parent was O, the short RNA that can destroy the promoter was left intact during the parent match. If the parent was E, it was culled. So the promoter is destroyed in the type O progeny, and retained for type E.
- Type E expresses the mtA gene on its cell surface, and can now mate with Os. Type O does not, and can now mate with Es.
In sum: the entire (big) nucleus gets matched, massively modifies its RNA world, it is destroyed and a new nucleus built from scratch, in a process that requires another massive match. All this, just to ensure that one short DNA stretch designating the parent's gender gets preserved for the next generation. The complexity is dazzling. Intelligent Design enthusiasts might just think of it as "irreducibly complex". But the other term that springs to mind is "pointlessly complex"...
All this got me to thinking about the possibilities for same-sex unions among paramecia. Apparently no bug with mtA on its surface ever mates with another of the same type, and no two lacking the protein ever mate. However! From time to time, the gene for mtA will mutate in such a fashion that it can no longer present the proper protein. Then you'll have a cross-dresser - an individual who is actually type E, but presents as type O, and will mate with other Es.
In this case, though, so wildly unlike our own, there is a "cure" for homosexuality - and that cure is same-gender sex. Because half the time that our type E mates with a "straight" type E, it will receive a working copy of the mtA gene. Since it has always had the intact promoter in its macronucleus, it will proceed to lead a straight life-style thereafter - but it will have passed its mutated mtA gene, and therewith its cross-dressing proclivities, on to its partner.
And so ends my little excursion into the wild and wacky world of one-celled epigenetics. I hope you found it as clumsy, elegant, and intriguing as I did.
[Footnote 1: Why the two paramecia are genetically identical twins after they have sex (but obviously not epigenetically identical, or this post would never have been born.) When conjugation begins, the micronucleus undergoes two divisions, into four haploid gametes, the standard meiotic sequence. Three of the four disintegrate. The survivor undergoes mitosis, so now there are two identical copies of each chromosome. One identical copy goes to the partner, and one of the partner's identical copies is received - so the happy couple part on extremely compatible terms.]