Saxitoxin: Arun P. Thottumkara, William H. Parsons, and J. Du Bois

DOI: 10.1002/anie.201308235

I have to admit, I really wanted to read the minireview instead on microfluidic chemistry... Okay, maybe I will read it anyways since it's ten pages, and I'm still alive after getting through that Saxitoxin review. Which is more than can be said for anyone ingesting 1 mg of the stuff.

Admittedly, the review starts off kind of easy with a good historical background on this toxin and some color pictures of poisonous living things and structures of some other related toxins. But, since I read the abstract and I'm somewhat familiar with what the group from which this review came from does (and it's not my field, once again), I was ready for slogging through total synthesis schemes which, while I might see how you got from point A to point B with this reagent, I have no idea of the mechanism really, since it's been a while since my Advanced Orgo grad course and the hardest things I had to make didn't have as many nitrogens in it as saxitoxin. Which reminds me, that thing has a lot of nitrogens. Together with the four oxygens, it has 11 heteroatoms, more than the 9 carbons. Still, it'll kill you dead with only 1 mg, and that is because of the two guanidinium residues, that makes it a dication actually. That quality makes it really good for blocking the ion transport channels, and mainly the sodium transport channels in mammalian, and especially primate, cells. That means the review will have lots of other things such as talking about the structure of the Na transport proteins and how exactly its function is blocked, plus the synthesis of saxitoxin analogues that will attack a specific Na transport protein, something that will only block the ones responsible for pain but not the ones that keep your heart muscles moving for example. Humans have a lot of these different type of membrane proteins and some of these analogues can be pretty selective for just one of them (a several hundred fold difference in activity, which ideally means 1 mg could stop pain, but you would need 200 mg to end it all).

Still, to get to all of that, you do have to get through the total synthesis part, which if you're not a total synthesis person, or a pharma researcher working on saxitoxin analogues, will be probably painful. Still, this blog is part of my self-challenge to make it through one review a week, reading the whole thing faithfully (up to a point), so, here's hoping there will only a bit of total synthesis since I actually enjoy reading through one or two of them once in a while.

First comes the biosynthesis, but beyond looking at one change and another, and seeing arginine turn into ornithine, I didn't really get much from it. The authors weren't probably too inspired by that 1 page section either. Maybe drawing out the structure of ornithine would have been better. I'm not going to look it up when I have a ton of total synthesis to still go through. Speaking of that total synthesis, I have to say that I did enjoy reading through the first two. I was wondering how easy it was to separate all the intermediate products on the column, since they are highly polar and have a ton of nitrogens probably. The first one by Kishi in 1977 answered that question by introducing the guanidium units as late as possible. I guess after you have preparative HPLC, it gets a bit easier, but back in 1977 there was a shortage of that instrument in academic labs. The second synthesis by Jacobi in 1984 has a very inspired transformation which you can see in Figure 10, where a ring expansion occurs through an NN cleavage of a hydrozide. Technically the word 'inspired' came from the review authors, but I thought that was a very neat transformation as well. Plus this transformation of compound 12 to 11 occurs with Na in NH3... always a reaction I wanted to try someday.

After this, the total synthesis got more modern and more involved, with all them fancy reagents. I couldn't really follow it so well beyond where the authors guided me through some mechanism that resulted in particular distereoselectivity. Especially the DuBois total synthesis started cheating on me, by talking about analogues, and I was getting stuck for so long on each scheme, my eyes started to glaze over at some point and I just power-read through it. Which means I probably didn't retain much.

Anyways, no reason to make my review of a review longer than necessary. On page 13 we once again come to the biochemistry part. Saxitoxin and friends are apparently used to figure out binding sites to these proteins. Since these are membrane proteins, they are very difficult to crystallize and there is not structure of them as of yet. Probably the structure highly depends on the membrane environment anyways. Everyone is really certain that these toxins target the channel through which the cations pass, and not some other part of the protein (that is responsible for sensing Na+ levels or receiving some other substrate that opens the channel), but they are probably right since I don't know anything. Plus the guanidiniums are cations like Na+ and people made models of them sticking in the channels (where they presume that that is the structure of the channel), so let's just go with it. Maybe the positive charge of saxitoxin sticking in there repels Na+ electrostatically, since from the picture it doesn't seem that they are blocking the entire thing so well.

What's really interesting, is that just one amino acid mutation, but often a couple, can have a dramatic effect on mortality, and it explains that while all living things and especially multicellular organisms (nerve communication) need to transport sodium in order to live, the from and the mollusk seem to be fine while humans die from a very small amount. It effects them too of course, as at a big enough concentrations even other forms of the Na transport proteins get blocked up, but they can deal up to a point.

The review then talks about computational attempts to guess the structure of the sodium channel part of the protein by sticking residues that they think make up that channel around the toxin structure that binds the best, but who knows if that is going to be even close to the real thing, once a crystal structure is available. While I was reading this of course, I thought of using saxitoxin for probing how many channels a cell has by attaching some sort of fluorescent probe. Well, the people working in the field thought of that long before I did at the end of reading the review. All these total synthesis introduced a route where you can modify a part right at the end and attach a fluorescent probe, then send it to the cell and gunk up all its channels and watch them move for a bit until the cell dies I guess. This is how they found that these channels move around a bit and are swallowed up by endocytosis once in a while. Still, maybe they were swallowed up because a cell decided it was gummed up by a toxin and needed replacing...

Interesting. I have to admit that I actually enjoyed reading through the last part, even if I don't understand much about the particulars of biochemistry and don't really know my amino acids by heart.

Okay, that's long enough. If you read this far, you should probably read the review of Saxitoxin by the DuBois group. Maybe I will read that microfluidics micro-review.