never say never

If I had to use one word to describe all of Nature, I'd be torn between "diverse" and "unpredictable". The first adjective is a fairly obvious descriptor: after all, there's a lot of stuff out there! Antlers, peacocks, narwhals, methanotrophic bacteria . . . However, the second adjective is a more loaded term. It is obviously a reflection of the observer, the one trying to predict.

We humans like to generalize. I don't know if we even like it so much as we can't help it. It might be hard-wired into our brains. It's an extremely useful ability for coping with our surroundings. You want to be able to look at a dog, any dog, and know that it is a dog, rather than seeing every dog as a totally unrelated individual. Or seeing the same dog on different days, or from different angles, and not knowing it is the same animal. I seem to recall there's a cognitive disorder that breaks the sufferer's ability to generalize in this way. (Anyone know what it's called?)

So we look at Nature, and greedily gather her into all our little compartments, niches, species, and populations. But Nature loves to fool us, to lull us into thinking she really does come in these convenient categories, before coyly presenting us with a complete categorical conundrum.

Here's a classic example. It's useful to casually define mammals as animals that give live birth. Unfortunately, this generalization breaks in both directions: not all mammals give live birth, and not all animals that give live birth are mammals. While the egg-laying monotremes can be labelled "primitive mammals" and neatly sectioned away from marsupials and placentals, who play by the rules, it is somewhat difficult to classify viviparous sharks as anything other than fish who simply happen to give live birth. No one has been quite foolish enough to try classifying them as mammals. (And sharks are by no means the only non-mammals to figure out vivipary. My favorite viviparous animals are the lovely onychophorans, a.k.a. velvet worms.)

Another classic example is endothermy, long thought to be the sole domain of mammals and birds, until tuna were discovered to warm their blood as well.

So, what once were sensible, educated statements to make, such as "All mammals give live birth," or "Only birds and mammals are warm-blooded," must now be--not thrown out, for although the absoluteness is wrong, the guidelines are interesting and useful--but presented as tendencies. "Most mammals are viviparous," etc.

Okay, sorry to slog through that, since it's all pretty familiar these days--at least to folks who like animals. My point, however, in going over the obvious stuff, is this: we still haven't learned our lesson! Rather, I think we are actually incapable of learning this particular lesson, or at least incapable of modifying our behaviors accordingly. Two human traits are at work here:

1. We like making rules.
2. We like breaking rules.

We the members of the human race seem practically compelled to be constantly making and breaking rules. This is very Popperian! (Karl Popper was one of the most influential philosophers of science in the last century. He suggested that the best way to move science forward was to make bold conjectures (the bolder the better) then try with all might and main to disprove them.)

So, I want to discuss some more current (and more obscure yet very fun) examples of rule-making and breaking. I'll start with cellulase. Cellulose is made by plants and it is really tough to digest--just imagine chewing on wood. You need a special enzyme to break it down, called cellulase*. You don't have the enzyme, trust me. Who does? Well, who eats wood? Wood rots, and rotting is just another word for "being broken into very small pieces by bacteria," so bacteria clearly produce cellulase. Who else? Termites come obviously to mind, but in fact, termites have symbiotic bacteria living in their guts to produce cellulase. So the conventional wisdom about termites is that they can only digest wood because of these prokaryotic companions. In fact, this wisdom is so conventional that I recently heard an eminent professor comment that only bacteria make cellulase. This elicited some interesting discussion in the class he was teaching (and I was attending), so I decided to go to Google Scholar and verify it.

Surprise! It turns out that quite a few eukaryotes actually do make their own cellulase, but this information has not yet percolated from obscure primary literature to textbook knowledge. Eukaryotic cellulase was actually first discovered in (get this) the defaunated guts of termites**. After people realized that termites could make their own cellulase, in addition to borrowing that of their symbionts, they started looking elsewhere. Cellulase enzyme activity, as well as cellulase genes, have since been found in parasitic nematodes, beetles, cockroaches, crayfish, and mussels. (Mussels?? Don't ask me!)

As long as we're talking about endosymbionts (I was, really) let's turn to corals. Most folk have heard of coral bleaching. That's what it's called when corals lose their endosymbionts--tiny photosynthesizing algal cells that fix carbon into sugar and share it with their coral hosts. But most of the time, happy, healthy corals can get up to 90% of their nutritional needs met by these algae. In a later session of the same class that hosted the discussion of cellulase, we started wondering how every new generation of corals gets their algae. Transmission of endosymbionts can be either vertical (passed directly from parent to child) or horizontal (every new generation has to acquire them anew from their environment). Another casual blanket statement was made at this point (admittedly, not by a coral biologist), to wit: "Corals all have horizontal transmission of symbionts."

Curiosity drove me to the primary literature again, and no one should be surprised to hear that I found exceptions to the rule, including the beautiful Seriotopora.

This discovery led me to consider the validity of yet another (more well-known) rule: Animals don't photosythesize, OR, all photosynthesizers are plants. (According to technical taxonomy, the Plantae are quite a small subset of photosynthesizers, but in the Plant Biology class I'm TAing, the professor covers plants in the broad sense, including algae and phytoplankton. This seems reasonable to me, and for the purposes of my argument, I'll be using this broad sense of plants.)

However, everything you think of as a plant doesn't really photosynthesize all by itself. Instead, plants harbor little machines called chloroplasts inside their cells, and each chloroplast is an entirely self-contained factory for harvesting the sun's energy and using it to build sugars. How self-contained? Well, it depends on which plant you're talking about, but chloroplasts are separated from the rest of the cell by at least two and up to four concentric membranes.

This was one of the clues that led scientists to figure out that chloroplasts were once independent single-celled organisms that were engulfed by larger cells, and, over time, developed into permanent endosymbionts. Their host cells began to take care of everything except photosynthesis for them, so they were slowly pared down to nothing but sugar factories***. (Fun tidbit: some chloroplasts still retain a remnant of their original nucleus, called a nucleomorph.)

The reason that the number of membranes around chloroplasts differ is that (we think) this "engulfing-->endosymbiosis" event happened more than once and in a nested fashion. Let me explain. One cell engulfs a photosynthetic bacterium, and the bacterium becomes a chloroplast. That's primary endosymbiosis and the chloroplast has two membranes. Now another cells comes along and engulfs the first cell (including its chloroplast, of course). The first cell is reduced just as the first bacterium was, until it becomes no more than a chloroplast with four membranes. This is secondary endosymbiosis. The whole thing can happen again in tertiary endosymbiosis and you'd think you'd end up with a six-membrane chloroplast, but that actually doesn't exist. A few membranes seem to have gotten misplaced in the various lineages, so some secondary endosymbionts have only three membranes, and tertiary endosymbionts actually have only four (although it is a different four than the four-membrane secondaries.) Have I thoroughly confused you? Try this; it does a much better job. With pictures!

You'll see on that page that mitochondria (little energy-producing factories in our cells) are thought to be endosymbionts too. That brings me to another fun "never say never" story. I always thought (because I'd been taught) that all eukaryotes, by definition, have mitochondria. Oh! Wrong! It turns out that a couple of our closest companions, the single-celled eukaryotes Giardia and Trichomonas (along with all their cousins), lack mitochondria. They seem to do just fine without them--much to the dismay of people suffering from giardiasis and trichomoniasis.

My apologies, that was a long digression. I was talking about corals and their endosymbionts. Corals are definitely not considered plants. However, they do host photosynthetic plants, and it ought to be interesting to compare these endosymbionts with chloroplasts.

Similarities:
- Chloroplasts and some coral endosymbionts are intracellular (live inside host cells)
- Chloroplasts and some coral endosymbionts are vertically transmitted (direct from parent to offspring)

Differences (as far as I know, but I bet I could be proved wrong, at least in some cases):
- Coral endosymbionts but not chloroplasts can live independently of the coral host
- Chloroplasts but not coral endosymbionts keep some DNA in the host cell's genome
- Coral endosymbionts but not chloroplasts retain all their ordinary cell machinery in addition to photosynthetic machinery (but remember those nucleomorphs!)

So, it seems to me that the intracellular, vertically transmitted coral endosymbionts are well on their way to becoming coral chloroplasts (which, by the way, might be a quaternary endosymbiosis event). At what stage during this process (assuming we're still around) will we decide that we might as well start classifying corals as plants (in the broad sense)?

Or maybe we should just put them on that long, long list of our favorite exceptions to our favorite rules.



* Nerd joke! Parafilm is a fun, sturdy cousin of plastic wrap. It is used in all labs and is nearly as ubiquitous as kimwipes and duct tape. Parafilm comes in a roll that can't be ripped; pieces must be cut off with scissors. However, whenever scissors are placed next to a roll of Parafilm in a prominent location, those scissors disappear. This has led some cunning scientists to rename Parafilm "scissorase".

** Wow. If defaunating a gut isn't an invasive procedure, I don't know what is. Hmmm. That's what we do to ourselves every time we take antibiotics . . .

*** Definitiely the weirdest thought I've had in a while: Some friends of mine work in Silicon Valley for a large company that shall remain nameless. This company is very keen to take care of all their needs: meals, laundry, car repair, etc. Consequently they are spending more and more time at work since they have fewer and fewer reasons to leave. Could this company be slowly turning them into little endosymbiotic organelles--code producing factories living their whole lives inside the company cytoplasm?