Darwin was probably wrong, the flagellum tale

A little while back I was between scientific projects looking for inspiration. I tend to find ideas by looking at the opinions of people deemed kooks and “deniers”. Their arguments are less restricted and more interesting than people working on some minutiae of some sub theory of a sub theory which is completely wrong anyway. (These minutiae are about 90% of biology research nowadays).

And I stumble upon Michael Behe. He testified in a US Federal Court case called “Dover” which was notorious in the science world. This case was about whether intelligent design should be taught in schools in conjunction with Darwinian evolution. He testified on behalf of intelligent design.

Michael Behe is a professor of biochemistry at Lehigh University, and the most eminent advocate of intelligent design. He coined the term “Irreducible Complexity” which is the first serious challenge to evolution I’ve heard and one which I haven’t been able to let go.

I work in a field called directed evolution. The basic idea is you crank up the rate of mutation of some gene in the context of a system which requires that gene to become more functional. Imagine you have 1,000 cats and the only food available is grass. Whichever cat can eat grass will survive, the rest…. Now the survivor cat will make more cats, and some of them will be even better at eating grass. So on, so forth. Eventually you’ll get an herbivore cat. This is Darwinian evolution.

When you really get down to the nitty gritty on a gene level you find some interesting problems. Say we look at the gene which allows breakdown of grass. The normal way of thinking about Darwinian evolution is that these genes are being randomly mutated and sometimes that makes the gene more efficient at that breakdown. Most of the time mutations make genes worse at their function, but sometimes they make them better.

Now, say that those herbivore cats are taken to a place full of mice. If the grass digesting gene isn’t being used and gets mutated randomly a few times it will break. Once it’s broken it’s unlikely it will ever come back. There’s some examples of this in humans. We used to be able to make vitamin C from other nutrients, but once we started eating food with vitamin C those genes broke down and now we have to eat oranges sometimes or we’ll die.

There’s a concept called “sequence space” which describes the combinatorial possibilities of a protein. If you’ve ever read the short story “Library of Babel” by Borges, it’s a good proxy. In that story there’s a near infinite library which contains 300 page books with random letters. Somewhere in that library is every book by Shakespeare, Steinbeck, Tolstoy, etc. Somewhere in there is those same books in a perfect translation in Spanish, French, Polish. Somewhere in there is those translations but written with every word backwards, and with the main character identical in every way other than being a 3-month, 4-month, 6-month old kitten. But because the letters are random, odds are each book will read like this: “Thanirging enrfinn eaiaena aigntkoarng anerianegamr antia antianramsdm gitwntk aing…” and so on.

Proteins are kind of the same. Composed of 20 different letters, the number of possible proteins in a 100 amino acid protein (pretty small) is 10^130. There are 10^80 atoms in the universe. Only a miniscule fraction of those possibilities are functional, and if you change just a few letters they’ll be completely broken.

In comes Professor Behe.

His challenge to Darwininan evolution centers on the implications of the numerical issue here. The example of this that really fucked me up is the flagellum.

The flagellum is a bacterial tail. It allows bacteria to swim away from predators and towards food. It’s an engineering marvel. It’s composed of 40 different proteins, which come together to make a structure which:

·        Piece by piece makes a stiff rod with a curve at the top allowing it to generate force

·        Shuttles that rod through a little hole in the outside of the bacteria

·        Grabs on to it with motor proteins which spin at 15,000 RPM (your car is revving at 2-3,000 RPM most of the time)

·        Form a clutch which allow the motor to detach from the rod when the bacteria doesn’t need to swim

·        Can also reverse the direction of the spin

·        Integrate with some sensing mechanism telling the bacteria how to move

Behe uses it as an example of something that’s impossible to explain with Darwinian evolution, and he’s right. The problem is all of those 40 proteins have to be there for the flagellum to be useful. Imagine the bacteria can make a rod but it has to swim all the time, dead. Swims randomly, undirected, dead. Makes a rod but it doesn’t have a curve, dead.

So in those books that are all “nagrkr aierngamAAS antamat anweaw” you have open 40 books at the same time and find a set which is all in the same language and has a sequential plot with minimal typos. It’s numerically impossible. The contention is that evolution works step by step, but in the case of multiple things evolving at exactly the same moment it doesn’t work.  

Instead of the cat needing to learn to eat grass, it has to learn to grow, cut, cook, and season a tasty salad. Sorry to say that cat is never going to make it.

The evolution of the flagellum is a mystery. Scientists around for the “Dover” case often get pretty upset if you bring it up. The “consensus” at the time was that it evolved from the Type III secretory system, but that has since been largely disproven.

Even today, new grants are being given for people to just try to come up with ideas on how it happened. Here’s a quote from a 2021 grant press release:

“For evolutionary biologists, the flagellum is an enduring mystery. This ‘molecular machine’ consists of many different parts that are all essential for the flagellum to work. How can gradual, Darwinian evolution build such a structure if not by inventing all necessary parts at once?”

The flagellum is not the only example of this phenoma. There are hundreds of examples of systems which require many connected parts. The clotting system in humans, most of our immune system, the eye, among others. The scientific community is wrong to dismiss the idea without serious consideration. Because science is a consensus machine, its ego is very fragile.

Behe says that there’s some kind of god-engineer which directs evolution, I don’t really agree. I emailed him to express admiration for the challenge and it seems like he was trying to win me over into religion and the conversation fizzled out after that.

Regardless, this is a real challenge to Darwin, and a question I’ve been obsessed with since I heard it over a year ago. I’ve emailed a lot of professors to get their takes and I haven’t found anyone with a satisfying idea. I’ve subjected most of my friends, people I meet at bars, grocery store clerks, my poor girlfriend to this rant. It genuinely made me consider we’re in a simulation, which is funny because that’s basically god for tech bros.

I have some ideas about the degeneracy of biology, and the connections in sequence space which I think will help explain it. They will be controversial and ruffle some feathers, which is fun. They still won’t come close to explaining flagella, but they’ll help a bit, if I’m right. I’m planning on getting a tattoo of the flagellum when I publish my first related paper. That’s what I’m working on now.