Intracellular Folate and PRPP

I mentioned that the intracellular total folate concentration varies inversely with the intracellular PRPP concentration, and the reason for this is unknown. (I will fill in or provide some references for these statements in subsequent postings, but I don't have time to now.) One hypothesis discussed in the literature is that, at least in proliferating cells, in which the activity of the enzymes responsible for de novo purine biosynthesis are high, folate repletion somehow leads to the increased consumption of PRPP by purine salvage enzymes. Researchers have found that folate deficiency does, in fact, impair purine salvage by cells in the liver (by the liver as a whole, when researchers look at the homogenized liver tissues), but would this explain the magnitude of the reduction in PRPP? There are some major flaws with this conventional explanation.

People with folate or B12 deficiencies do sometimes have elevations in urinary AICAR levels, indicating that the de novo formation of IMP from AICAR is, in fact, being suppressed by folate deficiency. And exogenous purines, usually hypoxanthine, have generally been found to rescue, in combination with thymidine, whose formation is also folate-dependent, folate-deficient red blood cell precursors (erythroblasts) or lymphocytes (white blood cells). This would suggest that de novo purine formation is being suppressed, by folate deficiency, at the level of AICAR transformylase, in the proliferating cells. But if folate were to simply increase IMP formation and "fix the purine depletion" by that mechanism, the extra purines would suppress the de novo pathway at the rate-limiting step, by suppressing amidophosphoribosyltransferase (AMPRT) and PRPP synthetase. AMPRT is suppressed by almost all purines, and PRPP synthetase is suppressed by AMP and ADP, if memory serves. This would be expected to increase PRPP levels or perhaps decrease them, to the extent that PRPP synthetase inhibition would decrease PRPP levels. The argument could be made that folate repletion would "first" increase AICAR transformylase activity and thereby replenish the purine pools. The purines would then suppress AMPRT and, by this suppression, increase PRPP availability for the purine salvage enzymes hypoxanthine-guanine phosphoribosyltransferase (HGPRT) and adenine phosphoribosyltransferase (APRT). But this seems unlikely, for various reasons. One reason is that de novo purine formation consumes PRPP. This could be replaced by the oxidative or nonoxidative pentose phosphate cycles, but it's not clear that this could occur efficiently, particularly under apoptotic ("programmed cell-death") conditions that have been induced by folate-deficiency-induced purine nucleotide depletion and DNA damage.

One reason that this concept of de novo purine formation disinhibiting purine salvage is inadequate is that the concentrations of hypoxanthine that are required to rescue cultured neurons or erythroblasts from folate deficiency is very large, in the range of 50 micromolar ("uM"). It seems unlikely that de novo purine biosynthesis per se, a process that is achingly slow and "lousy," to speak technically, could provide all of this hypoxanthine (IMP is inosine 5'-monophosphate, and inosine is hypoxanthine riboside). Even if the hypoxanthine formed de novo were converted into guanosine and adenosine nucleotides that could suppress AMPRT and increase PRPP availability for purine salvage, what about all the other pathways consuming PRPP? Plasma uridine has been shown to be decreased in folate deficiency, and folate-induced thymidine formation will tend to drive the ATP-consuming replication of DNA. Uridine formation requires PRPP, and so, by the indirect consumption at the level of nicotinamide phosphoribosyltransferase (involved in NAD formation), does the increase in the activity of poly(ADP-ribose) polymerase (PARP), an NAD-consuming process, that accompanies DNA replication. One could then argue that DNA synthesis consumes ATP and decreases the available ATP for PRPP formation. This is plausible, but the inverse relationship between PRPP and folate seems to be rather predictable and pronounced. There probably is a mixture of mechanisms, but DNA replication requires that the purine pool first be elevated. This brings us back to the question of why folate repletion would be able to compensate for the large amount of hypoxanthine, in the range of 50 or even 100 uM, that is required to rescue folate-deficient, proliferating cells. Especially in the brain, the activities of enzyme in the de novo pathway are extremely low. In the literature, the assumption seems to be that folate increases the activity of the de novo pathway (implying AMPRT activity increases), but one group of researchers actually found, as in Lesch-Nyhan syndrome, that folate repletion actually decreased AMPRT activity and thereby decreased the overall de novo pathway. This is because purine depletion disinhibits AMPRT and leads to AICAR accumulation. So it depends on what a person means by the "de novo" pathway. There's IMP formation and then there's AMPRT activity, and not-always the twain shall meet or something.

The effects of folate repletion on AICAR levels also seem to be too small and inconsistent to account for the dramatic, inverse relationship between intracellular total folates and PRPP levels. As the concentration of extracellular folate is increased from something like 30 nM to 2,400 nM, as a culture condition, the PRPP level decreases by 10 to 20-fold.

One partial explanation for these inconsistencies, an explanation offered by Rosensweig and his colleagues (http://www.ajcn.org/cgi/content/abstract/28/6/648), could relate to the suggestion, by Rosensweig, that the long-ignored folate-dependent enzyme, glutamate formiminotransferase [part of the glutamate formiminotransferase-cyclodeaminase (GFITCD) enzyme complex, if I remember correctly], may in some way be responsible for the activation of glycolytic enzymes (the activities of fructose bisphosphatase, pyruvate kinase, and phosphofructokinase). GFITCD is very interesting and mysterious and is one step in the breakdown, or rather recycling, of histidine, an amino acid. GFITCD can also become an autoantigen in some liver diseases and was shown to mediate the attachment of glutamate to microtubules (some effect like that). There may be some kind of protein-protein interaction between GFITCD, or the protein may cause post-translational glutamylation of other proteins (as has been shown for brain-specific tubulin--I think this is a representative article showing polyglutamylation of tubulin: http://lib.bioinfo.pl/pmid:9677387). GFITCD is an enzyme of the Golgi body, too, an organelle that is involved in protein trafficking. There's also some research on folate's effects on GLUT4, I think, that may help clarify this. But it's interesting that ribose-1,5-bisphosphate, produced by two sequential reactions of PRPP synthetase, I think, is an activator of glycolysis. If folate increased the production of this somehow, this would be expected to potentially deplete PRPP. Methotrexate, an antifolate, and AICAR, which may increase in folate depletion, also inhibit fructose bisphosphatase (forget the details), effects that are opposite to the effect of folate repletion (as would be expected). The overall concept is partly related to the metabolic effects of ribose, which, depending on the availability of the pool, which may be influenced by folate status, can apparently act as a kind of switch for initiating glycolysis. Folate could somehow drive ribose into the nonoxidative pentose cycle and, in conjunction with the increases in the activities of other glycolytic enzymes by folate repletion, both deplete PRPP and provide more ATP for other processes. This could account for the increase in purine salvage induced by folate repletion, given that, actually, a low PRPP to ATP ratio is necessary for purine salvage to proceed normally (http://www.jbc.org/cgi/content/full/277/12/9865). Ischemia and anoxia tends to elevate PRPP levels, and folate may help drive the PRPP into ribose-1,5-bisphosphate and thereby help activate glycolysis. It should be noted that PRPP levels tend to fluctuate drastically, particularly in proliferating cells, in association with the demands on the pool for DNA replication, etc. It may be that folate has one effect on PRPP levels in postmitotic neurons and another in rapidly proliferating erythroblasts or lymphocytes.