#but i know broadly how genes tend to evolve and why and I'm applying that knowledge to this problem

[a-commas-a-pause]

Just had a look and I have a theory, but I'm afraid my theory doesn't really have anything to do with a connection between spots and placentas (though you could draw a connection, if necessary - something about cell-to-cell communication, I suspect). I think it's probably more about the protein in question being a really good switch.

Let me explain what I mean by that. The protein in question is an aminopeptidase, which means that when it's expressed in the cell, what it does is it looks for other target proteins (with a specific structural motif) and then when it finds its targets, it chomps a bit off the end of them. This might result in the degradation of the target, or it might just be required for the target to do something else (like go latch onto DNA and express yet another gene, or like mediating pigment production). The point is, what this particular sort of protein actually does is nothing to do with placentas or spots. What this protein does is that in cells where it's expressed, it goes around turning off (or on) all the other proteins that have a particular motif/flag.

It's kind of like if I had a bunch of unrelated tasks I needed to do, but I gave myself a biscuit for every completed task. The biscuit isn't related to any of the tasks and it also doesn't particularly signify that the tasks have anything to do with each other, but it is still absolutely true that I will not do either task if I do not get my biscuit.

Now, as to why it doesn't affect placental function in cats (or coat colour in humans, I guess)? Well, if I had to guess, I'd say while this gene has been around for a long time, its original job was probably neither of those things. It probably did something else yet again, or several other things.

As long as genes' expression zones aren't crossing over, it's actually fine for them to have multiple roles. If A turns on both B and C, but B is only even expressed in the first place in eyes and C is only expressed in fingers, it's fine that A is crucial to two entirely separate processes, one for making fingers and one for making eyes. You're not going to end up with eyes on your fingers because for that you'd also need B (and X Y and Z probably as well). It's also fine if the zones of expression are only separated temporally - maybe B is expressed early and C late, and A gets turned on twice throughout development, once early to activate B and once later to activate C. And it's turned off in the middle to prevent it from accidentally activating D, which instead needs to be activated by B.

Lots of genes have a huge number of seemingly unrelated roles like this - it's called "moonlighting" and it's turning out to be way more common than we first thought. If you've already got a switch that works, why make another one from first principles? Now this is kind of the equivalent of "well we only have one single bed for two people, but it's fine because one of us works night shifts so we never need it at the same time", that is to say, Not Perfect. But natural selection doesn't optimise for perfect, it optimises for "good enough to not die long enough to successfully pass on your genes" (usually, ish, look this reblog is already really long and we don't need to get into this too right now), and the timeshare solution is plenty good enough, so it's what ends up happening like 90% of the time minimum. It's really really hard to make new things and really really easy to reuse old things, basically, and evolution always takes the easiest option in the same way that streams don't flow uphill.

There are some genes that shouldn't be able to moonlight because they turn on whole SUITES of other genes - genes where if you express them in the wrong place they DO put eyes where your fingers should go, and we call these master regulators. But also, this isn't like. A true dichotomy. Master regulators also do seem to moonlight sometimes. And a gene that's a master regulator of something in one clade might do something completely different in another clade. The GABA genes, which are neurotransmitters in animals, actually have a bunch of other signalling functions in plants, especially to do with establishing cell polarity (how does the cell know what direction to grow in). Turns out GABA's just a pretty handy little signalling molecule that's super versatile for cell-to-cell signalling, and multiple different branches of the eukaryotes have taken advantage of that fact!

was browsing thru papers about animal coloration and learned that a gene that helps make the placenta in humans (the thing that keeps you alive when you're a fetus in the womb) also exists in cheetahs, but in the cheetahs, it just controls the placement of their spots

How does THAT happen?!