A Cell’s Eye View of Evolution, Part 3

(The image for this week is an illustration from Waddington’s 1957 paper The Strategy of the Genes, which is often used to explain canalization. It shows a landscape with shallow, forking grooves and a ball rolling down that landscape. Although the ball’s path isn’t fully determined, the existing impressions in the landscape constrain it to one of a few likely paths)

This is part three of a three-part series. You can read it on its own, but to get the whole story, you should start from the beginning.

Darwin explained part of the great mystery of life: how complexity and intelligence can evolve from randomness. When DNA was discovered, this seemed to “seal the deal.” DNA is the molecule that describes an organism’s nature, makes traits heritable, and carries mutations that are fodder for natural selection. That seemed to explain everything, at first, but I would argue that’s just the beginning. The DNA molecule itself lies at the heart of an incredibly complex system of processes that manage its care, use, and replication. These systems are collectively studied as “epigenetics,” and science is just beginning to understand how they work and the impact they have on evolution.

It’s important to remember that DNA is an inert molecule that does nothing by itself. It needs a cell, a sort of organic micro-robot, to interpret that DNA and turn it into form and behavior. So, in a sense, every living thing is made up of two evolved programs: the DNA and the cell. Both are made of physical matter, which are subject to mutations. Both share the same selective pressures and reproductive fate. They evolve together, but they have different purposes. For the most part, the DNA program is what determines the organism’s lifestyle. The cell decides how to read that program, and how to make changes to that program over a lifetime and across generations.

One of the most important ways cells influence their own programming is through mutations. These happen naturally. As molecules bang against each other and get exposed to UV radiation from the sun, they sometimes spontaneously change shape. When those molecules represent critical information for a species’ survival, that could be disastrous. For this reason, life invests a ton of energy into detecting and correcting errors. But this process is never perfect, and it can’t be. If life always copied itself perfectly, there would be no variation for selection to act on, and no evolution. Not only that, getting the error rate much lower than it already is would be prohibitively expensive. So, life strikes a healthy balance, allowing just enough mutation to be useful, but not enough to be dangerous.

Interestingly, that finely-tuned mutation rate is not constant and universal. There are some stretches of DNA that get extra error correction, always triple checked to ensure they stay as stable as possible. On the other hand, some stretches of DNA get actively shuffled, injecting randomness into things like the immune system, creating diversity that makes the population as a whole more robust. Perhaps most remarkable is that when cells get stressed out, they divert energy to other things, and away from error correction. This may just be an accidental side effect of the cell breaking down, but it might also be a survival strategy. To anthropomorphize, perhaps cells get creative when times are tough, trying out crazy ideas in the desperate hope that one might save them.

That said, a single DNA copying error can be devastating, so how does life cope? Remarkably, it can often just work around the problem. Living systems have a lot of redundancy, with many mutually supportive ways of doing basically the same thing. This leads to a phenomenon called “canalization.” The more critical some behavior is to life, and the longer it persists over many generations, the more redundancy builds up around it. This means that single errors may alter the behavior a little, change how it works, or make it less efficient, but probably won’t break it entirely.

When errors are too severe to recover from, an organism might just fail to thrive and die, but sometimes it actually notices the failure and decides to self-terminate. That may seem bizarre, but in multicellular organisms it makes a lot of sense. If the error is in a single cell, then removing that cell lets others take over its job. If the error would prevent the whole embryo from developing into a healthy adult, then it’s better to scrap the work in progress, recycle those materials, and start over from scratch. In other words, life has its own Quality Assurance processes, at multiple levels, which minimize investment into evolutionary dead ends.

Cells can also swap genes with each other, sharing useful recipes and trying them out in new combinations. Sex is one way to do it, aligning and remixing two complete genomes in an incredibly complex way that ensures the resulting DNA program is still valid. Simpler organisms like bacteria don’t do this, but swap genes in a much more free-form process called “horizontal gene transfer.” Basically, cells sometimes leave scraps of DNA lying around, or pick up those scraps and integrate them into their own programming. This can let a new behavior (like resisting some toxin or eating some food) spread very rapidly through a bacterial colony. Either way, randomly adopting genes that have proven successful in another organism is a much safer and more powerful way to create useful diversity than mutation alone.

It’s also worth noting that how a cell reads its DNA can change over a lifetime. Cells annotate their program with notes (ie, methylation) that indicate which recipes to avoid or use more of, depending on context. This is how single celled organisms adapt their behavior to a changing environment, and how cells in multicellular organisms differentiate into different kinds of tissues. Importantly, these notes are sometimes passed down across generations. For instance, an organism might survive near starvation by tuning down its metabolism, staying smaller, slower, and using less energy. That change is heritable. The next few generations will also have a slower metabolism, and if that serves them well, it could lead to long-lasting behavioral changes that eventually get encoded into the DNA itself.

So far, I’ve talked about how life modifies itself, but it also modifies the environment. Organisms can build caches, nests, and tools that make life easier, and these get passed on, too, both as hand-me-downs and as lessons. Organisms form an ecosystem, full of mutualistic relationships between species that make life easier. Over geological time, this has transformed our planet from a barren rock to a lush world full of possibility. Life cultivates a supportive environment for future generations, shaping their behavior and evolutionary fitness. Child care might be the most visible example, protecting each new life when it’s most fragile, then sending off the new generation in a good direction informed by the parent’s life experience.

In the basic Darwinian story, evolution is something that happens to organisms. Accidental changes occur randomly, and nature chooses which ones will persist. But as we’ve just seen, life does not leave things up to chance. Randomness plays a key role in biological evolution, but life manages that randomness carefully and uses it selectively. Life also does everything in its power to influence the next generation, in ways that are not random, but “purposeful” in a sense. A cell can’t understand why it does these things, but it does them for a reason: they worked well in the past, got selected for, and ended up in the cell’s programming.

This leads to a powerful realization: when it comes to influencing evolution, cells don’t understand what they’re doing, but many higher organisms do, at least a little. For instance, an animal can apply its full cognitive capacity, mind and all, to choosing a mate and raising its offspring. In this way, the cell has moved from blindly repeating what worked in the past, to making evolutionarily relevant decisions intentionally, with forethought and analysis. A dog may not understand genetics or think about the future of her species, but she certainly has strong opinions about who would make a good mate, when / where / how to raise her puppies, and which pups to give more or less attention to. She uses her senses, her instincts, and her big brain to make big decisions that shape evolution. She may not see the big picture, but she cares and makes informed choices nonetheless.

This helps explain the paradox of how life managed to become so incredibly smart just by randomly banging molecules together for a few billion years. Life may have started off randomly, but it quickly became more directed. Life harnessed Darwinian evolution to build a more powerful evolutionary algorithm, one that’s opinionated and shapes its own search space. At first, these evolution-shaping behaviors were simple and rigid, just tricks repeated by rote because they tended to make the next generation more successful. Then, as life became more intelligent, it started to apply that intelligence to shaping itself, creating a runaway process of recursive self-improvement.

Conclusion

The main takeaway from all this is that Darwin is the beginning of the story of evolution, not the end. Life uses all of its intelligent capacity to influence its own evolution. This led to a virtuous cycle. Increased intelligence gave life greater influence over evolution, which it used to become more intelligent, which gave it greater influence over evolution. For this reason, I think it’s better to say that life designed itself than to say it evolved by chance. The process of “design” here was more stochastic, collective, and unthinking than we normally associate with that word, but in the end, the result is the same.

This story is still uncertain. The science around autonomous robots, intelligent collectives, and epigenetics is relatively new, and changing all the time. Plenty of biologists push back hard against the idea of any sort of agency or direction in evolution, partly because they’ve been fighting against the theory of Intelligent Design for so long. Others believe we’re overdue for a new story about evolution, and are trying to find the right narrative and the evidence to back it up. I hope my research into evolutionary algorithms might be a useful contribution to that effort. If you’d like to dig deeper into this topic, Evolution in Four Dimensions is an excellent overview of the field of epigenetics.

What do you think? Did reading this make you think of life, cells, or evolution any differently? Any new ideas? Does anything I said sound wrong or misleading? Do you have other ways of looking at it? This post is more speculative than usual, and represents some of the ideas I hope to pursue in my PhD research, so I’m very interested in criticism and feedback. If you have any thoughts, please let me know in the comments!