If I had to pick a characteristic of the natural world that distinguishes it from the man-made one, it would be: the closer you look, the more you see. That obtains for some very good underlying reasons. The first is that the natural world can often operate at a finer level of detail than we humans can manage by design, at least outside of an advanced chip fabrication facility. The second is that with living systems, we are looking at the accumulation of a couple of billion years of evolutionary adaptation, which has had plenty of time to encrust everything with interlocking, overlapping rococo epicyclic curclicues, but in a complete nonhuman manner that can be a source of constant bafflement and surprise.
And that leads to a third reason behind the complexity of the natural world: we don’t know a priori what it contains, how many features it has. A human-designed system has some annotations and assumptions built along with it. You can be pretty sure that Shakespeare’s sonnets will not turn out to be imbedded letter by letter in the source code of the latest version of Windows and that it does not contain sections that will for some reason crank out endless variations on muffin recipes or recapitulate the residential telephone directory of Kankakee, Illinois for 1957. But evolution has provided examples of things that seem nearly as detailed, surprising, and pointless (at first glance).
Of course, non-evolved systems can have a similar bottomlessness to them as well: consider the analysis of (say) mineral specimens. You start out with things like color and density and hardness. Then you realize that they have different crystal forms, and you develop enough mathematics to know exactly how many of those there are – how many there can be, actually. And you develop enough knowledge of chemistry to understand the complex molecular formulae that give rise to all these neat crystals, and then realize that their colors are often due to tiny amounts of impurities in what would otherwise be a clear sample. And that the ratios of those subtle traces can tell you that a particular ancient ruby came from this deposit over in this part of the world as opposed to that one a thousand miles away. And that the isotopic distribution of the heavy atoms point to that element having been synthesized in a particular rare type of supernova explosion somewhere else in the galaxy, whose debris helped make up our star systems’ protosolar nebula. And on, and on.
Well, I enjoyed writing all that widescreen intro, but you probably could have gotten it more succinctly in Robert Graves’ “Warning to Children”. Anyway, it’s all a leadup to talking again about post-translational modification of proteins, a topic that has come up here many a time, most recently with this hitherto unknown example. Here’s a pretty good list of PTMs, and I thought I’d heard of most of them.
But there’s a new paper that sheds some light on one of these that was honestly new to me: the modification of Cys side chains to persulfides (R-S-S-H). It’s not on that list I just linked to, for starters. It’s part of a whole set of oxidative modifications to that thiol group (and to the S-methyl in methionine), many of which I think we can safely call “poorly understood”. But that is not to be confused with “unimportant”, because Cys groups really hit above their weight in the protein world, being crucial ingredients in active sites, regulatory domains, and tertiary structures. The fact that a whole list of important proteins regularly goes through something as funny-looking as persulfidation suggests that we are missing something of interest.
The paper linked to identifies the first known enzyme that accomplishes this process, 3-mercaptopyruvate sulfur transferase (MPST), which was already known to be involved in generating hydrogen sulfide as an intracellular signaling molecule (one of several such “gasotransmitters”, and not the only one that is also a deadly poison under other conditions). But MPST can also form these persulfides (the paper demonstrates this for , and it’s a safe bet that it’s not going to be the only enzyme in that class, either. This result does explain why MPST knockout organisms were found to be more sensitive to oxidative stress. There are other “sulfur transferases” out there, which might also be doing this protein regulatory work on the side, and it’s also possible that some of these modifications are nonezymatic, and perhaps part of a general emergency stress response to keep Cys groups from getting torn up across the proteome. There’s a complicated sulfur storage and release system inside cells that we don’t fully understand, and while some of it is certainly involved in on-the-spot generation of hydrogen sulfide, some of it is clearly regulating protein function by SSH groups, too.
And probably other things we’ve never heard of yet! Reading about this, I found myself wondering about amino acids like serine and threonine and whether they have post-translational modifications to peroxide groups. Protein peroxidation is certainly real, but I’m not aware yet of a “controlled” regulatory pathway through these intermediates. But I wouldn’t have pegged hydrogen sulfide (or carbon monoxide) as intracellular signaling candidates either, would I? The closer you look, the more you see.
