Introduction
It is indeed rare these days that a totally new coenzyme or vitamin conjugate is identified. This seems to be the case, however, with the report by Bettendorff et al., who identified a conjugate (dubbed adenosine thiamine triphosphate, or AThTP) that is formed between the 
-phosphoryl group of ATP and the 
-hydroxyethyl side chain of thiamine (Th) or vitamin B1 (ref. 1 and Fig. 1). In an exciting study, the authors isolated the conjugate and solved the structure, which they confirmed by de novo synthesis. The data suggest that the newly discovered compound is produced in response to carbon starvation. Though conjugates of the vitamins riboflavin, nicotinamide, and panthothenic acid with adenine nucleotides (FAD, NAD+/NADH, and coenzyme A, respectively) have long been discussed in biochemistry courses, the discovery of a nucleotide analog of thiamine indicates that we still have more to learn about vitamin metabolism.
Figure 1: Possible metabolic fates of AThTP.
The newly discovered AThTP may serve as a chemically and enzymatically more stable 'storage' form for ThTP or ThDP.
Full size image (18 KB)The history of the role of thiamine and its phosphates in physiology is at least 80 years old, with the mono- (ThMP), di- (ThDP) and triphosphates (ThTP) of thiamin frequently referenced2. Whereas the diphosphate ThDP has well-known coenzyme function as an electrophilic covalent catalyst both in the decarboxylation of 2-oxo acids and in carboligation and lyase-type reactions (reminiscent of benzoin condensations)3, 4, the function of ThMP and ThTP is less clear. The fact that enzymes exist to interconvert these various forms has been reported for many years, and ThTP has long been suspected of having functions in the nervous system2, 4.
Research on thiamine and its derivatives has progressed along two tracks over the years: (i) the coenzymatic functions of ThDP3, 4, and (ii) its biosynthetic pathways in a variety of cells, including Escherichia coli and yeast. The coenzyme functions have kept many groups busy, who are in no small part fascinated by the varied chemistry afforded by these enzymes, both oxidative and nonoxidative. For example, several of the ThDP enzymes have provided excellent synthetic tools to produce chiral compounds, especially 
-ketols, with high enantiomeric excess4. At the same time, interest in the biosynthesis of ThDP in different cell types has also been rekindled, with notable contributions made by Chatterjee et al.5 and Spenser et al.6.
So where does the newly revealed derivative of thiamine with the 'nucleotide handle' fit in? The case for thiamine derivatives having biological functions beyond the classical role of ThDP as a coenzyme has received considerable support by Bettendorff and collaborators, among others. The compound AThTP seems to be produced in E. coli under stress created by carbon starvation, and once the cell is provided with glucose, ThTP concentrations increase and the concentration of AThTP diminishes. The authors identified AThTP in E. coli, yeast, some plant roots and mammals, but unlike ThTP, the AThTP is more abundant in mammals in several organs, and less so in the brain. Bettendorff and co-workers have long been on the trail of ThTP-metabolizing enzymes, and they have made important contributions on the identification of a ThTP triphosphatase from a variety of mammalian tissues, including rodent brain7. Additionally, Nghiem et al. showed that ThTP can act as phosphoryl donor in certain kinases in Torpedo marmorata8. Those results suggest that ThTP and AThTP may participate in different functions or pathways.
Future research should focus on placing the newly identified AThTP into the proper metabolic context—that is, researchers must determine the biosynthetic route leading to it, its metabolic fate, and, of course, its biological role. Whereas the current study focuses on a 'conjugate' of ATP with the intact vitamin, another paper by Chatterjee et al.5 on the biosynthesis of vitamin B1 reports a related ADP-thiazole conjugate with similar attachment of the diphosphate to the -hydroxyethyl side chain. Chatterjee et al. have suggested that the biosynthesis of this thiazole-ADP conjugate is derived from NAD+ (ref. 5), which raises the interesting possibility that NAD+ may serve as a biosynthetic intermediate, in addition to its coenzyme function. Chemically, it is plausible that AThTP is a precursor to ThTP (as a less reactive storage form of the latter), and is converted to ThTP by a single enzyme—an enzyme that is induced only when there is a carbon source available. Of course, AThTP may also serve a similar 'storage' function for ThDP (Fig. 1). Whatever the ultimate answers to these questions may be, the discovery of a new metabolic intermediate is always exciting and is bound to generate new research in the field.
