All organisms need folates for one-carbon donations and other vital functions. Plants, most protists, and most bacteria synthesize them de novo. A few bacteria do not make folates and must acquire them from external sources by using energy-requiring transport systems. Bacteria with both synthesizing and transporting capabilities are rare (Hitchins, 1962; Huennekens et al., 1978). In E. coli (Green et al., 1996) the pathway to tetraihydrofolate, the biologically active form, is by way of 6-hydroxymethyl-dihydropterin pyrophosphate (hereafter called dihydropterin) (starting with GTP) and p-aminobenzoate (starting with chorismate). These moieties are then united by dihydropteroate synthetase to form dihydropteroate. This is the step that is inhibited by sulfonamides. Sulfonamides combine with dihydropterin instead of p-aminobenzoate, thus draining off the cell's supply of the folate precursor. Dihydropteroate is then converted to tetrahydrofolate by reduction and addition of a glutamate residue. Tetrahydrofolate may then acquire additional glutamate residues and / or be transformed into potential one - carbon donors such as 5,1 0 - methylenetetrahydrofolate.

Folate metabolism in Chlamydiaceae has been discussed by Moulder (1991) and McClarty (1994; 1999). Although exceptions were noted at the outset, it has been known for a long time that Chlamydia spp. are generally susceptible to growth inhibition by sulfonamides whereas Chlamydophila spp. are not. This difference was first interpreted as meaning that sulfa - sensitive chlamydiae made folates de novo and that resistant chlamydiae got folates from their hosts. However, experiments with chlamydia - infected cell cultures revealed that, of several chlamydiae tested, both sulfa - sensitive and sulfa - resistant strains had strain - specific complements of folates and synthesized folates de novo (Colon and Moulder, 1958; Colon, 1960; Fan et al., 1992). These findings have been corroborated by whole genome sequence determinations. The genomes of C. trachomatis (sensitive), C. muridarum (sensitive)., and C. pneumoniae (resistant) have the same five genes of folate biosynthesis [see: Genome comparisons for references]. The small number is not unusual. Although the pathway to tetrahydrofolate in E. coli involves more than 10 genes (Green et al., 1996; Blattner et al., 1997), the genomes of many other bacteria, including those of other genome - challenged epicellular and intracellular bacteria (See Table 2 for references) have far fewer genes of folate biosynthesis. For example, rickettsiae also have five folate synthesis genes, but not the same set found in chlamydiae. In Chlamydiaceae, genes for biosynthesis of dihydropterin and p - aminobenzoate are lacking, so these essential intermediates must be supplied by host cells. Transport of dihydropterin and p - aminobenzoate into other bacteria has been noted, but the mechanisms are obscure. The sequenced chlamydial genomes all contain the gene for dihydropteroate synthetase, so if the as yet un-sequenced genomes of other Chlamydiaceae spp. also have it, then all Chlamydiaceae are potentially subject to growth inhibition by sulfonamides. Why then are some chlamydiae sulfa - sensitive and other sulfa-resistant? 

The answer to this question may lie in the experiments of Fan et al., (1992). They found that the meningopneumonitis (Francis) strain of C. psittaci grew equally well in Chinese hamster ovary cells maintained under either folate - rich or folate - depleted conditions. However, its growth was inhibited by sulfisoxazole only when the host cells were cultivated in folate - depleted media. They concluded that chlamydial susceptibility to sulfonamides is determined, not by the ability to make folates, but by the ability to obtain active forms of the vitamin from host cells. This conclusion assumes that de novo folate biosynthesis is down - regulated in the presence of preformed folate.

If the results with this one strain of C. psittaci hold for other "sulfa - resistant' strains, a plausible picture of folate metabolism in Chlamydiaceae may be drawn. The familial LCA had already lost the genes for synthesis of p - aminobenzoate and dihydropterin but had retained the genes for making active forms of folate from these precursors. Like most other Gram - negative bacteria, the LCA did not utilize exogenous folates and was therefore sulfonamide - sensitive. Then as Chlamydiaceae radiated into different niches [see: radiation and speciation], some populations acquired the genes for folate transport. They became indifferent to sulfonamides, but only when exogenous folate was available. Because sensitive chlamydiae are preponderantly Chlamydia spp. and resistant ones are preponderantly Chlamydophila spp., it is tempting to conclude that acquisition of the genes of folate transport occurred after divergence of the two genera. However, exceptions such as the well-known susceptibility to sulfonamides of the 6BC strain of C. psittaci, an otherwise typical Chlamydophila (indeed the type species of the genus!) makes me wary of so simple a conclusion.

Although the LCA seems to have lost the genes for p - aminobenzoate and dihydropterin synthesis when it learned to get them from its hosts, its sulfa - resistant descendents did not lose the capacity for folate biosynthesis when they learned to utilize host folates. Perhaps this is a sign that folate transport is a recently acquired attribute. Is there a fitness advantage in retaining both routes to tetrahydrofolate? Will folate biosynthesis genes eventually disappear from the genomes of folate - importing chlamydiae? Which is the most efficient way of obtaining folates; importing p - aminobenzoate and dihydropterin and then making folate, or importing preformed folate in the first place? In view of the complete absence of information about the energetics of the transport mechanims involved, who knows?.

Where did these newly acquired folate transport systems come from? They are either the product of tinkering with existing genetic material or they have come from other organisms. Folate - utilizing bacteria are usually Gram- positives like lactobacilli or streptococci (Huennekens et al., 1978). Horizontal transfer between Gram - positives and chlamydiae seems unlikely, but think of the C. muridarum cytotoxin with homology to the large clostridial toxin (Belland et al., 2001; Genome comparisons). Other possible source are Leishmania and Plasmodium spp. (McClarty, 1999). Unlike most protists, these obligate intracellular parasites have folate transport systems. What is the relation between the folate transport systems in chlamydiae and the protists? Is it common origin or convergent evolution?
[JWM]