Skip to main content

Chemical News

Phosphorylation Without Pain

Phosphorylation is a prime example of a reaction that’s hugely important in biochemistry that organic synthesis struggles with terribly (as opposed to the efficiency and finesse with which it’s handled in by enzymatic processes). The methods used to attach phosphate esters in the flask are frankly pretty crude (all the way up, or down, to good ol’ phosphorus oxychloride), and generally involve hydrolyzing intermediates to reveal the free phosphate groups at the end. You see high temperatures, acid/base conditions, oxidizing reagents, and less-than-impressive yields – as opposed to the biological systems which are whisking around selectively adding and removing phosphate groups from specific protein protein and carbohydrate residues. As usual, watching the enzymes at work is like watching a ballet being performed in mid-air on trapeze; down here on the ground we can only shake our heads.

This new paper, though, gets us a bit closer. The authors (from the Univ. of Tokyo and the Nara Inst. of Science and Technology) are using a biomimetic strategy with phophoenolpyruvate (PEP) as the phosphate donor. This needs to be activated (which is what the enzyme active sites will do for you), and they’ve found that the combination of DMF as solvent and tetrabutylammonium hydrogen sulfate as acid does the trick quite well. The reaction optimization is pretty interesting – a lot of other acids barely work, or not at all. The usual organic-chemist thinking is that protic acids are all fairly similar once you hit the same pKa (as opposed to Lewis acids, which definitely have widely varying behaviors), but this is a good example of how fine-tuning H-acid conditions really can be crucial.

Detailed mechanistic work shows that the real activated phosphorylating agent is the odd-looking (and previously undescribed) mixed anhydride shown at right. The reagent system phosphorylates primary and secondary alcohols in the presence of a wide variety of other functional groups (aldehydes, carboxylic acids, alkynes, phenols, protecting groups such as Fmoc, Boc, and trityl, and more. That phenolic selectivity means that you can take an oligopeptide and phosphorylate serines and/or threonines in the presence of tyrosine residues, and the authors show several examples of this and of carbohydrate substrates as well, yields 50 to 80%.

This makes a person wonder what other phosphoryl donor species might be out there that we haven’t come across. It’s for sure that the authors weren’t aiming at the POSOP system when they started out – they noticed that acidic ion-exchange resin gave them some product (test alcohol plus PEP), and investigated a whole list before finding that the hydrogen sulfate salts stood out, and so on. But that’s how good reagents get found! And this one looks like it will be widely adopted.

16 comments on “Phosphorylation Without Pain”

  1. Barry says:

    prof. Henry Rapoport was known to say that “ammonium chloride is a fine weak acid to quench reactions, except for the ammonium, and except for the chloride” (neither of which is reliably a bystander). He recommended KHSO4 for that role. But even sulfate isn’t wholly passive in all circumstances if you are very clever about looking very closely.

  2. A Nonny Mouse says:

    Reminds me of chloromethyl chlorosulphate which transfers the chloromethyl group (including on to phosphonyl)

    1. Tourettes of Chemistry says:

      Or the Burgess Reagent.

  3. ex lab rat says:

    Nice approach. I think from a practical standpoint, this looks like a nice way of putting one phosphate on one or more relatively unhindered alcohols at the end of a synthesis. There are a number of drugs that are phosphate prodrugs that this might be a nice way of making in one step. In fact, given the one step approach I could imagine a couple programs I worked on would have benefited by just submitting the prodrug for PK analysis because the unprotected alcohol always had poor PK but the phosphate got on board. However, putting a phosphate group on in the middle of a synthesis and relying on prep RP-HPLC to purify everything thereafter actually is a pain.

  4. Reader says:

    Derek- sorry for hijacking a comment spot, but do you plan to write a post on the new coronavirus?

  5. Jb says:

    I once had to make phosphorylated non-natural carbohydrates as a grad student in order to be able to test them in some enzymatic assays, because of course you couldn’t buy them. Holy Toledo, what a nightmare because phosphorylation chemistry is so messy, and it’s almost impossible to purify a phosphorylated carbohydrate out from the harsh mess you have to use to phosphorylate no matter what the literature describes.

    It took me 2 years to crack the synthesis (we weren’t an organic lab, and I did cellular work too, the synthesis was a side project). However, I stumbled upon phosphoramidite methods to phosphorylate things. It is mild and very clean. I was then able to use some hydrogen to clip off the protecting groups leaving behind a very nice phosphorylated sugar that could be used from flask into the enzymatic assays because it came out so clean with no purifications needed other than a quick filter from what I recall. I believe they use amidite reagents to phosphorylate dna/rna molecules specifically because it is quite mild and efficient.

    1. oligo says:

      Hasn’t that been automated to make oligos?

  6. Barry says:

    It has long been known that all Phase Transfer Catalysts are not alike; closer scrutiny may show that they’re sometimes doing quite a bit more than escorting ions

  7. E. D. Thorsett says:

    Many many years ago I used an efficient phosphorylating agent based on 2-phenyl-1,2-dibromoethylphosphonic acid. Treatment with base generated bromide, beta-bromostyrene, and the phosphorylating species- PO3(-2). Phosphorylated alcohols were easily separated from the orgainic by-product in high yield. I came across this reagent in a JCS article dealing with sugar phosphoylation but do not have the exact reference. As I recall, the reagent was commercially available at one time.

  8. FOEFOEENO says:

    I know it’s a (minor) typo, but I giggled at the way “phophoenol..” (missing an s!) sounded in my mind when reading it.

  9. Anonymous says:

    This may be hard to draw in ASCII (“ASCII me a question, I’ll give you an ANSI.”), but they show that the P-OS bond in P-O-S or P-O-S-O-P is more labile than the S-OP bond, hence phosphorylation, not sulfation. Prep of pyrophosphates from ROH usually involves prep of mixtures of RO(P), RO(PP), RO(PPP) that have to be separated by chromatography (normal phase, reverse phase, ion exchange, depending on R). If you could prepare a P(2)-O-P(1)-O-S reagent (or the symmetrical POPOSOPOP) would you able to diphosphorylate by cleavage of the more hindered P(1)-OS preferentially over the less hindered P(2)-OP? Delta-steric hindrance versus delta-bond strength.

    If so, the effort put into making the P-O-P-O-S reagent could be worth it if you are not wasting precious ROH in a lower yield, hard to purify step. Which POPOS to make and how to make it? Use CF3SO2O for the terminal S, P-O-P-OSO2CF3? Potassium triflate + 2 PEP? Methyl sulfate + ammonium phosphate to make P-O-P-O-SO2-OMe? I think you could brute force synthesize several POPOS reagents and see if they P-ylate or (PP)-ylate. If you file a patent, be sure to cite In The Pipeline.

  10. milkshake says:

    It is not completely unprecedented: cyclic triphosphate (trisodium salt) will phosphorylate starch hydroxyls even in water if you add CaCl2 as activator. Some of the polysacharide hydroxyls end up crosslinked by phosphodiester bond.

    For the peptide phosphorylation work they showcase, I would still prefer the system developed by nucleotide chemists: (tBuO)2PNEt2 with unsusbtituted tetrazole in DCM, quenched with tBuOOH, followed by TFA. It is very mild. If the substrate does not tolerate TFA there is also benzyl protected version available

  11. The Grim Reaper says:

    Frank Westheimer wrote a short note in Science in 1987 entitled “Why Nature Chose Phosphate”. It is worth rereading. What a smart man.

    1. Tourettes of Chemistry says:

      And a bit more recently:


  12. R Jason Herr says:

    Hey, thanks for that! I’m struggling with this right now, and hadn’t seen this pre-print. Cool!

  13. zzt says:

    Hopefully we will get a review of antivirals for coronaviruses on this site soon ?

Comments are closed.