Turns out we can express most of proteins, some of the time, and then isolate them. This includes enzymes, when isolated these can do things like they naturally do but now in flask, but also they do things that aren't remotely natural but are useful for us. These things are pretty fragile usually so then some of these can be modified so that they are resistant to higher temperatures, detergents etc. This is not only the nerdy shit like advanced chemical synthesis - lots of dishwasher tablets and and washing powders contain enzymes that cut proteins into pieces (like subtilisin), so in some cosmic sense dishwasher digests your leftover food off plates

Enzymes are still proteins, and have all problems of proteins. Turns out, you can just take the most important part out of enzyme, make it, or something functionally similar out of completely synthetic parts, and it still works. Sure, it's not as active or selective, most of the time, but it's resistant to things that would absolutely shred proteins. This is called organocatalysis and it was subject of 2021 Nobel Prize

Sometimes you want to take an enzyme and make it not work. We also have a tool for that: first you have to get structure of that enzyme, or some receptor protein, and by looking how a small set of random molecules lodges in it you can make a very selective, very potent ligand, sculpting it atom by atom with no knowledge other than protein structure. If you have time and resources, this can be made to work for almost any protein (that can be crystallised)