Let’s talk about why multicellular creatures aren’t puddles of goo.
From each one’s initial single-celled state, plain old cell division by itself doesn’t make a creature, merely more cells. You get a creature via development, the differences in what these cells – all with the same genome – are doing depending on here-or-there. Something must be happening here: how the systems of the body and specific modifications or activities of their cells are related to their location, conditions, and time at any given stage of development.
The primary responses include (from roughly most drastic to least) death, rate of division, migratio, and adhesion, all typically carried out by a bunch of cells at a particular spot and time. As the bodily mass becomes bigger, you can observe an axis of some kind, various edges rolling into tubes, empty spaces, and distinctive layers and clumps of cells. Here’s the biggest picture – for any given piece, and also roughly overall, two stages:
- Morphogenesis: the movement, growth, and clumping of stem cells and barely-altered stem cells into the shape and divisions of the eventual body – you can think its composition as living play-dough getting into position
- Differentiation: the local changes in cell shape and activity that occur once morphogenesis has shaped a given organ or sheet of the body, establishing and beginning the operations we call its function
How did such a process evolve? And how do its properties, as such, impinge upon processes like selection? Thinking about it has been a very hard road in biology. As always, much abides in the dialogues of the 19th century. One productive angle on embryology (as it was called) was comparative, yielding Ernst Haeckel’s seminal work on similarities among vertebrate embryos and much else, quickly integrated with Darwin’s ideas, along with Alfred Russell Wallace’s work in ecology and biogeography. Another, slightly older angle led by the estimable Karl Ernst von Baer was procedural, delving into the moments of change in a single individual’s personal experience of becoming a working organism, which did not become so integrated into the evolutionary theory of the time.
Von Baer is one of the towering presences of early modern biology, including his discovery of the ovum (without which, no cell theory), his observations of the germ layers (those “-derm” words you had to memorize) and their fates in animal organ systems, and also stating principles from those observations which – as I see it – demanded mechanics-based investigation. Yet this work didn’t blossom like ecology and genetics did, not for a long time, to the extent of neglect. Even Haeckel’s ideas, what he called the biogenetic law, underwent significant revision at this point, and although I agree with the intellectual details of doing so, it also served to mark development “down” into the role of genetic handmaiden. Colleagues reading this will see that I could go right into how otherwise it couldn’t be easily fit into the phenotypic application of the Hardy-Weinberg equilibrium model and the New Synthesis. (But you can bet a New Synthesis post is coming some day.)
Research on development during most of the twentieth century reflected its isolated status, being concerned medical-physiological topics, none of which is trivial but which also remained, for a long while, restricted to those applications. Only a few significant details of natural history appeared, including metamorphosis, with the nifty insight that most of the larval body dies and the new body is mostly constructed de novo from a bank of stem cells; and observations concerning heterochrony, especially neoteny, the retention of a infant appearance in adults in some species. As late as the 1980s, the basic bio texts presented little more than a quick paraphrase of von Baer’s work, over a century and a half old, tucked away in its own little chapter as a set of arbitrary terms of memorize.
I was fortunate to encounter Aron Moscona as one of my early university mentors, especially when, bored and frustrated by molecular labwork, I was seriously reconsidering my goals in college. His response to that, which is to say, to me as an undergrad struggling in his graduate course, has never left my thoughts. His course broke open for me the whole “handmaiden of genetics” role for development, when I learned that not only did cells do things to one another, or that they creepy-crawled all over the embryo into various places to become specific things, but that cell membranes were carrying out operative, significant management roles over whole sectors of developmental outcomes. Not genes. This pointed to the notion that location and condition, all around the developing embryo, signaled the genes in the cells there what to do – an outside-to-in causal process which had definitely not been included in my basic training. Wow, I said, this is huge. How does it relate to genes-and-traits outside of medical anomalies?
Therefore I was at least a bit prepared for the boom of developmental insights that were soon to reconnect the topic to evolutionary biology with a vengeance. Certain other harbingers had arrived as well, especially Stephen Jay Gould’s Ontogeny and Phylogeny. People started to debate adaptive constraint and developmental constraint relative to selection and environment, and even better, the concept of morphogenetic field – first conceived in 1910! – was recovered for basic or general biology. It’s the very thing I just mentioned, that the three-dimensional matrix of a developing embryo’s body is itself a signaling-and-response environment. As a term it is so mellifluous and sexy that it was scooped up by New Agers almost instantly for all manner of nonsense, so I have to explain a bit to dial it back to reality.
Wow. Development in a given broad type of creature, say, “deuterostome animals,” is a game of Go. Not really that many pieces or ‘rules,’ but enough moves and enough consequences to make any single exchange (“game session,” “species”) recognizably different from another. And when you factor the diversity of physical ecology into that difference – just wow. You and a field mouse and a killer whale: very similar game sessions of Developmental Go. All the classes of Vertebrata: an identifiable cluster of starting parameters, just flexible enough to generate equally-identifiable subsets of early play.
Then came along the discovery of the Hox genes, which offered two surprises: first, their structural correspondence on the chromosome to the segmentation of the body, which to my understanding is the only known connection between physical location of genes and their effects; and also their phyletic conservation, which is to say, not only do all mammals use the same Hox genes to arrange their body linearly, but most of the other animals we’ve examined do too, no matter that the butthole of a mammal is the mouth of an insect and vice versa.
Moving into the 1990s, even such blink-inducing discoveries were just the beginning.
Fine, fine, there are some few genes which are close to 1:1 traits – the ones for which the esulting protein itself is undeniably a hard-working item on its own. The most significant for us are in the MHC complex, regarding the details of our active immune system. But, and to get technical for a second, although the phenotypic applications of the Hardy-Weinberg equations work great for these, when it comes to most genes doing most things in biology, we’re gonna have to turn to chaos theory.
Now for the big take-home question: what, now, is a trait? I can point to one end of an evolutionary process (say, selection) very easily: the teeth are shaped like this, and the critter uses the teeth like this upon a substance or other creature like this. It’s often harder to discern these things than one might think, but the concept itself is easy enough. The physical anatomy, physiology, and behavior interacts with the world around it like so, and we can then look at consequences like reproductive success or energy-use or environmental impact, whatever you’d like. The other end is much more difficult to hold in mind: how those real-life consequences show up as changes in the variation in a later generation, via the medium of population-wide reproduction and genetic recombination. It’s not just about allele frequencies as a function of reproductive success any more – it’s about changes in the developmental processes that generate conditions which alter the outcomes of developmental processes which didn’t change. Dammit, chaos theory, quit doin’ that!
Careful now – spend too much time on that harder end and you might convince yourself that only developmental traits exist at all, as far as evolutionary phenomena are concerned, and that tooth anatomy or anything like it is hardly worth mentioning. But the very temptation to think that way is food for thought. The traits like the teeth which interface with the environment are second-order to the mechanisms of inheritance, now that the existence of development as a suite of traits is undeniably clear. And since developmental traits themselves can’t be traced 1:1 to genes either, at least not in the simple one-gene-with-alleles way we’re mostly taught and mostly teaching, hardly any bit of selection as a process turns out to be as straightforward as, say, sickle-cell or other single-gene, drastic inherited-illness problems.
If I were givin’ out grants, you can bet one topic I’d favor is opportunistic colonialism among protists – i.e., viable one-celled creatures who can make a body for a while and also re-make it some other time but with the various cells taking on different roles from the ones they did last time. Obviously the various developmental mechanisms we use evolved in such creatures, and animals, most fungi, and plants represent one or more limited versions – in which one can only make a body and continue to live in that context. In which our allegedly-totipotent stem cells are a dim and minimal version of that prior much more total totipotency.
Next: Genes aren’t scary