Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


The ylide has landed

The enzyme co-substrate SAM has long been known to have two chemically distinct roles. A study of the CmoA enzyme suggests that SAM has a third trick up its sleeve — it forms species known as ylides. See Letter p.123

Rarely has nature made such efficient use of a compound as it has of the biomolecule S-adenosylmethionine (SAM). SAM contains a positively charged sulphur atom known as a sulphonium group, which means that this molecule is often used as an electrophile — a polar species that is attracted to electron-rich centres. The compound also initiates a host of non-polar biochemical transformations that are mediated by free radicals1,2. But sulphonium groups have another ability that so far has not been observed in biochemical transformations: they can promote reactions by forming dipolar 'ylide' intermediates, which act as nucleophiles by reacting with electron-poor centres. In this issue, Kim et al.3 (page 123) report strong evidence that an ylide intermediate is formed from SAM in the biosynthesis of a modified nucleotide, 5-oxyacetyl uridineFootnote 1.

Ylides contain two opposing charges on adjacent atoms. In most ylides, a carbon atom containing an unshared pair of electrons is bonded to a positively charged atom, usually nitrogen, phosphorus or sulphur. Sulphonium-containing ylides are routinely used in synthetic organic chemistry, particularly to prepare molecules that contain small rings of atoms. Although they have been proposed as intermediates in several biochemical transformations4,5,6,7, there has been no compelling evidence for this role.

5-Oxyacetyl uridine (cmo5U) is formed by the post-transcriptional modification of uridines (RNA bases) that occupy 'wobble' positions in several bacterial transfer RNAs. This modification allows a single transfer RNA to decode all the possible codons (three-nucleotide sequences) that could encode a particular amino acid. Genes that encode the SAM-dependent enzymes CmoA and CmoB are required for the biosynthesis of cmo5U, and 5-hydroxy uridine (ho5U) is probably an intermediate in the pathway8 (Fig. 1a). Inactivation of the cmoA gene results in the formation of the incompletely modified tRNA bases ho5U and methoxy uridine, whereas only ho5U is formed on inactivation of cmoB. Perplexingly, genetic and biochemical studies9 indicate a role in the generation of cmo5U for a metabolite that originates from the chorismate biosynthetic pathway, which is key to the formation of certain amino acids and for other crucial metabolites.

Figure 1: An ylide biosynthetic intermediate.

a, In the biosynthesis of the post-transcriptional modification 5-oxyacetyl uridine (cmo5U), the enzymes CmoA and CmoB catalyse the addition of an acetate group (blue and red) to the hydroxyl group (OH) of 5-hydroxy uridine (ho5U). C2 of the acetate (blue) comes from the co-substrate S-adenosylmethionine (SAM). Kim et al.3 propose that C1 (red) comes from another co-substrate, prephenate (tRNA, transfer RNA). b, The authors suggest that prephenate loses its carboxylate group (CO2) as a molecule of carbon dioxide in a process that results in the formation of a SAM ylide (a species in which positive and negative charges reside on adjacent atoms). Phenylpyruvate is generated as a side product. The ylide reacts with the CO2 to form carboxy-SAM, in which an acetate is attached to the sulphur atom of SAM. The acetate is then transferred to ho5U in the presence of CmoB to form cmo5U (reaction not shown). Curved arrows indicate electron movement.

Formally, the biosynthesis of cmo5U from ho5U involves the attachment of the second carbon atom (C2, in the methyl group) of an acetate unit to the hydroxyl group (OH) of ho5U, rather than the first carbon atom (C1, in the carboxylate group; Fig. 1a). But these groups are unlikely to react with each other because they both tend to form nucleophilic, negatively charged species that are chemically incompatible. Moreover, isotopic labelling studies9 have shown that the C2 carbon of the acetate moiety in cmo5U derives from the methyl group of SAM, which, in turn, derives from the amino acid methionine. This begs the question: what is the origin of the C1 carbon?

Kim et al. have solved the X-ray crystal structure of CmoA from the bacterium Escherichia coli and found an unexpected treasure buried in the enzyme's active site: a carboxy-SAM molecule, in which a carboxylate group (CO2) is covalently attached to the methyl group of SAM. So how did the carboxylate group become attached?

The sulphur atom of SAM is bonded to carbon atoms in three chemical entities: a methyl group, a 5′-deoxyadenosyl group and a 3-amino-3-carboxypropyl group (Fig. 1b). The positive charge also causes the hydrogens on these adjacent carbon atoms to be weakly acidic — they can be removed as protons (H+ ions) to form ylides. Carboxy-SAM could therefore be generated if the ylide that forms by deprotonation of SAM's methyl group attacks some electrophilic source of a carboxylate group.

However, the crystal structure described by Kim and colleagues reveals no group in the carboxy-SAM-binding pocket that could deprotonate the methyl group of SAM. Moreover, when the authors tested common electrophilic sources of carboxylates in the in vitro reaction of CmoA, none was effective. The chorismate biosynthetic pathway is known to be important in the formation of cmo5U; on this basis, the authors tested chorismate (an amino-acid precursor) as a potential carboxylate donor. Sure enough, they observed the slow formation of carboxy-SAM.

The researchers also observed another product, phenylpyruvate. They hypothesized that this product formed in two steps: an uncatalysed rearrangement of chorismate, which forms a compound called prephenate; then a CmoA-catalysed decarboxylation reaction in which prephenate releases a molecule of carbon dioxide (Fig. 1b). When the authors tested prephenate as a carboxylate donor, the CmoA reaction was much faster and higher yielding than it was with chorismate, and proceeded without the lag period observed with chorismate. Moreover, when chorismate labelled with carbon-13 was used in the reaction, the carbon label transferred to the carboxylate of carboxy-SAM.

On the basis of the above observations, Kim et al. propose that prephenate loses CO2 and eliminates a hydroxide ion (OH), which is sufficiently basic to remove a proton from the methyl group of SAM, generating a nucleophilic ylide (Fig. 1b). The ylide then reacts with the liberated CO2 to give carboxy-SAM, in which the C2 carbon of the acetate moiety — the same carbon that was nucleophilic in the ylide — is electrophilic. The authors went on to perform studies with radiolabelled SAM, providing evidence to support a reversible ylide formation that requires hydroxide ions.

The researchers then conducted the CmoA reaction in the presence of CmoB and tRNA purified from a cmoB-deficient strain of E. coli — that is, ho5U-containing RNA — and observed the formation of cmo5U. The generation of this product involves the attack of the nucleophilic hydroxyl group of ho5U on carboxy-SAM. Overall, it seems that nature cleverly uses the sulphonium moiety of SAM to invert the reactivity of the C2 of an acetate unit from nucleophilic to electrophilic, so that a suitable nucleophile can attack it.

The decarboxylation of prephenate to generate both a strong nucleophile (OH) and a strong electrophile (CO2) could cause disastrous side reactions. The way in which the active site carefully orchestrates its catalytic sequence will therefore be of utmost interest to biologists. Kim and co-workers used computational modelling to dock prephenate in the active site of CmoA, and found that the methyl group of SAM is sandwiched between the hydroxyl and carboxylate groups of prephenate. However, the hydroxyl group is poorly positioned for nucleophilic attack on the methyl group, although it is appropriately positioned for proton removal. Further X-ray structures of CmoA in complex with prephenate and/or unreactive SAM analogues might shed light on the details of crucial subtleties in the reaction.

The other interesting aspect of the cmo5U post-translational modification is the mechanism by which a hydroxyl group is attached to uridine to form ho5U, especially in the absence of molecular oxygen. An electrophilic source of the hydroxyl group would be required for this reaction — it would be amazing if this group were supplied by another unexpected metabolite.


  1. 1.

    *This article and the paper under discussion3 were published online on 15 May 2013.


  1. 1

    Challand, M. R., Driesener, R. C. & Roach, P. L. Nat. Prod. Rep. 28, 1696–1721 (2011).

    CAS  Article  Google Scholar 

  2. 2

    Booker, S. J. Curr. Opin. Chem. Biol. 13, 58–73 (2009).

    CAS  Article  Google Scholar 

  3. 3

    Kim, J. et al. Nature 498, 123–126 (2013).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Iwig, D. F., Grippe, A. T., McIntyre, T. A. & Booker, S. J. Biochemistry 43, 13510–13524 (2004).

    CAS  Article  Google Scholar 

  5. 5

    Iwig, D. F. & Booker, S. J. Biochemistry 43, 13496–13509 (2004).

    CAS  Article  Google Scholar 

  6. 6

    Kinzie, S. D., Thern, B. & Iwata-Reuyl, D. Org. Lett. 2, 1307–1310 (2000).

    CAS  Article  Google Scholar 

  7. 7

    Tokiwano, T., Watanabe, H., Seo, T. & Oikawa, H. Chem. Commun. 6016–6018 (2008).

  8. 8

    Nasvall, S. J., Chen, P. & Björk, G. R. RNA 13, 2151–2164 (2004).

    Article  Google Scholar 

  9. 9

    Hagervall, T. G., Jonsson, Y. H., Edmonds, C. G., McCloskey, J. A. & Björk, G. R. J. Bacteriol. 172, 252–259 (1990).

    CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Squire J. Booker.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Landgraf, B., Booker, S. The ylide has landed. Nature 498, 45–47 (2013).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing