Eli5: how are anatomical features encoded in DNA?


AFAIK the only output of DNA transcription are proteins but how are things like skeletal structure encoded? I mean things that are not functional specialization like the shape of my nose or the size of my forehead or the length of my legs.

In: 2

Because the DNA contains instructions for what the cells should do *from conception* (from the moment when it’s a single-cell egg with some DNA from the sperm). Instructions for how to go from [a cell](https://biologydictionary.net/wp-content/uploads/2018/03/Embryonic-Stem-Cell.jpg) to the millions of cells in the fully grown YOU (as opposed to a fully grown other animal). Meat is meat, bone is bone, more or less (that’s why we can eat the stuff and USE the digested “components” that are inside the cells, proteins, fats, sugars, etc.), but “form” depends on how the cells divide and which organs and shapes they “grow” into as you develop.

DNA doesn’t encode “how” as much as it encodes “when” cells should begin to specialize or begin to show certain traits. This timing depends both on the number of divisions that occur before a gene is expressed, and on hormonal signals that trigger groups of responsive cells to change expression in subsequent divisions.

They aren’t. There’s nothing in DNA that explicitly encodes “your leg bone is this long.” DNA is a program that needs to be run on a very specific set of hardware (a fertilized egg) in a very specific enviroment in a very specific order. It’s the combination of the cell, the timing, the environment, and the DNA that all comes together to make a human (or a banana or a lizard or…).

If you just dump human DNA into the nucleus of a random cell and grow it in a petri dish you don’t get a human.

But *if* you take a fertilized egg and implant it in a functioning womb (or identical chemical equivalent) and let it grow with the right nutrients a whole series of chemical reactions will happen that will eventually end up with a leg bone of a particular length.

Have you ever played with a cellular automaton like Conway’s Game of Life? Each individual cell works on its own and doesn’t really know where it is. Rather they run on rules like “if you see this chemical and this chemical then you are now a bone cell so make this chemical.” During early fetal development cells figure out where they are and make an approximately human-shaped blob; then they basically grow according to their cell type and the cells directly touching them in the blob.

Getting this right is really hard which is why there are so few basic body plans for animals. They all have four limbs and a head and possibly a tail – because evolution figured that out and stuck to it. All the changes, like tail/no tail, or front legs/arms, are just minor variations on this. Instead of coming up with new signals to make arms instead of front legs, evolution found it easier to just convert the front legs into arms, or make the tail convert into something else or not grow.

Making the legs longer is relatively easy as you just make the chemical that makes legs grow last a little longer so they grow more, but adding another leg is difficult because it fundamentally changes how the legs figure out that they’re legs. It’s probably more likely that humans evolve a tail that works a little bit like a leg, than that they evolve a third actual leg.

Check out the experiments of Michael Levin concerning the modulation of charge and termination of anatomical limbs. Most double stranded DNA is non coding and regulatory. Non coding means not a gene or coding for a protein.

Levin asks the question in his papers and talks how do cells know when to STOP growing? Where is the plan and how might it be shared across great distances between cells? The answer is sometimes via charge or electric potential.

We usually think of nerves using charge in action potentials but he suggests that ion gates are prolific throughout other cell types and might be part of a simple communication network.

He also investigates Basal Cognition.