Both authors contributed equally to this study.
Evidence for folate-salvage reactions in plants
Article first published online: 30 MAR 2006
The Plant Journal
Volume 46, Issue 3, pages 426–435, May 2006
How to Cite
Orsomando, G., Bozzo, G. G., de la Garza, R. D., Basset, G. J., Quinlivan, E. P., Naponelli, V., Rébeillé, F., Ravanel, S., Gregory, J. F. and Hanson, A. D. (2006), Evidence for folate-salvage reactions in plants. The Plant Journal, 46: 426–435. doi: 10.1111/j.1365-313X.2006.02685.x
- Issue published online: 30 MAR 2006
- Article first published online: 30 MAR 2006
- Received 29 October 2005; revised 12 December 2005; accepted 20 December 2005.
- folate salvage;
Folates in vivo undergo oxidative cleavage, giving pterin and p-aminobenzoylglutamate (pABAGlu) moieties. These breakdown products are excreted in animals, but their fate is unclear in microorganisms and unknown in plants. As indirect evidence from this and previous studies strongly suggests that plants can have high folate-breakdown rates (approximately 10% per day), salvage of the cleavage products seems likely. Four sets of observations support this possibility. First, cleavage products do not normally accumulate: pools of pABAGlu (including its polyglutamyl forms) are equivalent to, at most, 4–14% of typical total folate pools in Arabidopsis thaliana, Lycopersicon esculentum and Pisum sativum tissues. Pools of the pterin oxidation end-product pterin-6-carboxylate are, likewise, fairly small (3–37%) relative to total folate pools. Second, little pABAGlu built up in A. thaliana plantlets when net folate breakdown was induced by blocking folate synthesis with sulfanilamide. Third, A. thaliana and L. esculentum tissues readily converted supplied breakdown products to folate synthesis precursors: pABAGlu was hydrolysed to p-aminobenzoate and glutamate, and dihydropterin-6-aldehyde was reduced to 6-hydroxymethyldihydropterin. Fourth, both these reactions were detected in vitro; the reduction used NADPH as cofactor. An alternative salvage route for pABAGlu, direct reincorporation into dihydrofolate via the action of dihydropteroate synthase, appears implausible from the properties of this enzyme. We conclude that plants are excellent organisms in which to explore the biochemistry of folate salvage.
Tetrahydrofolate (THF) and its one-carbon (C1) substituted derivatives – collectively termed folates – are vital cofactors for various C1 transfer reactions. They are tripartite molecules consisting of pterin, p-aminobenzoate (pABA), and glutamate moieties, and typically have a short, γ-linked polyglutamyl tail attached to the first glutamate (Figure 1a). Plants, bacteria and fungi synthesize folates de novo whereas higher animals need a dietary supply (Cossins, 2000; Green et al., 1996).
Folates undergo spontaneous oxidative or photooxidative scission of the C9–N10 bond at physiological pH to yield a pterin and p-aminobenzoylglutamate (pABAGlu) or its polyglutamyl forms (pABAGlun) (Figure 1a) (Gregory, 1989). Such non-enzymatic cleavage is thought to be the main way folates break down in all organisms, although proteins may sometimes facilitate the reaction (Scott, 1984; Suh et al., 2001). Folates vary in susceptibility to cleavage, THF and its oxidation product dihydrofolate (DHF) being among the most vulnerable (Gregory, 1989; Hillcoat et al., 1967; Reed and Archer, 1980). For THF and DHF, the first pterins formed in the reaction are, respectively, tetrahydro- and dihydropterin-6-aldehyde; further oxidation can convert the tetrahydro to the dihydro form, and both to the fully oxidized, aromatic form (Hillcoat et al., 1967; Reed and Archer, 1980; Whiteley et al., 1968) (Figure 1a). Still further oxidation converts pterin-6-aldehyde to pterin-6-carboxylate and perhaps other end products (Lowry et al., 1949).
The pABAGlu and pterin moieties from folate breakdown are excreted in the urine in mammals (Scott, 1984). Nothing is known about the fate of these products in plants, but fragmentary data from microorganisms suggest their re-use in folate synthesis (Figure 1b). Thus (i) enzymes that release pABA from pABAGlu are known from bacteria (Hussein et al., 1998; Sherwood et al., 1985); (ii) the protozoan parasite Leishmania and bacterial cell extracts use pterin-6-aldehyde for folate synthesis, suggesting the capacity to reduce both the pterin ring and the aldehyde side chain to give the folate synthesis intermediate 6-hydroxymethyldihydropterin (Figure 1a) (Brown et al., 1961; Nare et al., 1997; Shiota, 1959); and (iii) enzyme activities mediating these reductions are known in Leishmania and Escherichia coli (Bello et al., 1994; Mitsuda and Suzuki, 1971). For pABAGlu, another potential salvage route is via the folate-synthesis enzyme dihydropteroate synthase (DHPS), which can use pABAGlu in place of pABA, giving DHF. The DHPS enzymes of E. coli and Plasmodium have Vmax/Km values for pABAGlu that are approximately 100-fold lower than for pABA, suggesting that the pABAGlu reaction is insignificant in vivo (Ferone, 1973; Swedberg et al., 1979). However, a Pneumococcus DHPS shows far less preference for pABA over pABAGlu (Ortiz, 1970), and the Lactobacillus enzyme shows none (Shiota et al., 1969), so in these cases salvage via DHPS may be important. However pABAGlu is recycled, its polyglutamyl tail may first need to be removed by γ-glutamyl hydrolase (Figure 1b) (Orsomando et al., 2005).
In mammals, whole-body folate-breakdown rates are normally approximately 0.5% per day (Gregory and Quinlivan, 2002). There are no such estimates for plants, but indirect evidence suggests much higher values. Thus post-harvest studies of leaves and fruits point to net breakdown rates of approximately 10% per day at 20°C (Scott et al., 2000; Strålsjöet al., 2003). Similarly, in Arabidopsis thaliana plantlets given the folate-synthesis inhibitor sulfanilamide (Prabhu et al., 1998), the rate of decline in C1 fluxes suggests a possible folate loss rate of ≥10% per day. If plants have high folate-breakdown rates, it is important to know whether – and how – they salvage the fragments. Accordingly, in this work we first demonstrated that breakdown rates can be high, then began an exploration of salvage reactions. The systems studied were Arabidopsis and pea (Pisum sativum) leaves, in which most biochemical work on plant folates has been done, and fruit of tomato (Lycopersicon esculentum), which is a target for folate metabolic engineering (Díaz de la Garza et al., 2004).
Folate-breakdown rates in Arabidopsis leaves and tomato fruit
For Arabidopsis, the folate-breakdown rate was investigated in liquid-cultured plants by blocking folate synthesis with sulfanilamide and monitoring the decline in leaf folate content (Figure 2a). Relative to the untreated control, total folate fell by 49% after 4 days’ treatment.
For tomato, matched pairs of fruit were selected: one was detached at breaker stage and allowed to ripen, the other left on the plant. All fruit reached the red-ripe stage in 7 days, at which time the detached fruit had, on average, only 38% of the folate content of the attached fruit (Figure 2b). Assuming that folate contents of attached tomatoes change little between breaker and red-ripe stages (Basset et al., 2002), it can be estimated that post-harvest breakdown was 62% in 7 days.
pABAGlu pool sizes
Folate breakdown in vivo will produce pABAGlu and pABAGlun, but efficient salvage is expected to keep their pools small compared with the total folate pool. To test this prediction, total pABAGlu (pABAGlu plus pABAGlun) levels were assayed by a novel method that entails measuring the pABA released by carboxypeptidase G treatment. This enzyme removes polyglutamyl tails via exopeptidase action and then cleaves glutamate from pABAGlu (McCullough et al., 1971). The liberated pABA is partitioned into ethyl acetate and quantified by fluorometric HPLC.
Total pABAGlu pools in Arabidopsis, pea and tomato tissues ranged from 0.06 to 0.6 nmol g−1 FW, which is only 4–14% of typical values for folates in these tissues (Figures 2a and 3a). Moreover, these pABAGlu data overestimate in vivo levels because they are inflated by folate breakdown during the assay. Although assay conditions were designed to minimize oxidative folate cleavage, some is inevitable, and tests with folates added to extracts showed that 5-methyl-, 5,10-methenyl-, and 5-formyl-THF underwent up to 4–6% breakdown, and THF up to 26%. The small pABAGlu pools observed can thus be ascribed partly to folate breakdown.
Further evidence for salvage of pABAGlu comes from the Arabidopsis experiment in which net folate breakdown occurred due to blockade of folate synthesis with sulfanilamide (an inhibitor of DHPS) (Figure 2a). The net folate loss was accompanied by little accumulation of pABAGlu, but by a marked rise in pABA and its metabolite pABA glucose ester, suggesting that pABAGlu moieties released by folate breakdown are efficiently hydrolysed to pABA (Figure 2a). Note that pABAGlu hydrolysis was not the only source of the rise in pABA, as the build-up of free pABA and its glucose ester (2.48 and 3.21 nmol g−1 FW, respectively) far exceeded the net loss of folate (0.41 nmol g−1 FW). The extra pABA presumably came from de novo synthesis, perhaps accelerated by loss of feedback inhibition of pABA production as folate levels fell.
pABAGlu and dihydropteroate synthase
While the above evidence is consistent with salvage of pABAGlu and pABAGlun via conversion to pABA, it does not exclude the possibility of salvage by direct incorporation into DHF via the action of DHPS. To investigate this alternative, we first tested pABAGlu and pABAGlu5 as inhibitors of the well characterized enzyme from pea (Rébeilléet al., 1997). Activity was assayed radiometrically using [14C]pABA as substrate. pABAGlu was a moderate inhibitor of [14C]dihydropteroate synthesis, and pABAGlu5 a weak one (Figure 4). Double-reciprocal plots showed that the inhibition was competitive, and gave Ki values of 30–50 μm for pABAGlu and 400–600 μm for pABAGlu5 (and a Km value for pABA of approximately 0.8 μm, in accordance with published data). The difference between the Ki values and the Km for pABA suggests that pABAGlu is – at best – a poor substrate for DHPS, and that pABAGlu5 is an extremely poor one. Semiquantitative radioassays using [3H]pABAGlu as substrate supported the former inference; formation of [3H]DHF was confirmed by HPLC analysis, and the Km for pABAGlu was estimated as approximately 100 μm (not shown). The large amount of [3H]pABAGlu substrate needed in each assay precluded more detailed kinetic studies.
As physiological context for these kinetic data, note that pea leaves contain ≤0.3 nmol pABAGlu moieties g−1 FW (Figure 3a), and that mitochondrial volume is approximately 5 μl g−1 FW (Winter et al., 1994). Thus, even in the unlikely scenario where all the leaf's pABAGlu moieties (mono- and polyglutamyl) are assumed to be in mitochondria, their concentration in the matrix would be ≤60 μm.
Hydrolysis of pABAGlu in vivo
Given the above indirect evidence for salvage of pABAGlu by hydrolysis, we tested for this capacity directly. pABAGlu labelled with 14C in the pABA moiety was supplied to Arabidopsis leaf or tomato pericarp tissue, and after 6 h the absorbed and non-absorbed 14C was analysed. Both tissues took up 80–90% of the [14C]pABAGlu and hydrolysed significant proportions: 5.4% of the absorbed label in leaf tissue; 11% in pericarp (Figure 5a). Hydrolysis was assessed by summing the 14C present in free pABA, its glucose ester (a major metabolite of free pABA; Quinlivan et al., 2003) and its N-glucoside (a pABA adduct formed during extraction; Quinlivan et al., 2003) (Figure 5b). The hydrolysis probably occurred intracellularly, as the non-absorbed label was almost all in the form of pABAGlu (Figure 5a). The amounts of hydrolysis in leaves and pericarp were equivalent to rates of 1.5 and 2.6 nmol g−1 FW day−1, respectively (assuming rates are constant). These estimated rates are high relative to typical total folate pools (2.2 and 1.3 nmol g−1 FW in leaves and pericarp, respectively; Figure 3).
Pterin-6-carboxylate pool sizes
As with pABAGlu, folate breakdown will produce pterin products, but if salvage reactions exist they should not accumulate. Pterin-6-carboxylate is expected to be diagnostic in this respect as it is a stable end-product formed by oxidation of the first products of folate cleavage (Figure 1a); it is not salvaged by bacteria (Shiota, 1959) or Leishmania (Nare et al., 1997), and probably not by plants (Stakhov et al., 2002). A lack of pterin-6-carboxylate accumulation would thus suggest pterin salvage activity. Pterin-6-carboxylate levels were relatively low, from 0.06 to 2.8 nmol g−1 FW, and were equivalent to 3–37% of typical folate levels (Figure 3b).
Reduction of dihydropterin-6-aldehyde and pterin-6-aldehyde in vivo
Dihydropterin-6-aldehyde is a primary folate-oxidation product that potentially can be salvaged by reduction to hydroxymethyldihydropterin (Figure 1a). This reaction was investigated in vivo by measuring the accumulation of hydroxymethyldihydropterin in tissues fed with dihydropterin-6-aldehyde for 15 h. Hydroxymethyldihydropterin was estimated by fluorometric HPLC from the difference in hydroxymethylpterin peak size between oxidized and non-oxidized portions of each sample. Arabidopsis leaf and tomato pericarp tissues readily converted dihydropterin-6-aldehyde to hydroxymethyldihydropterin, at rates equivalent to 12 and 0.9 nmol g−1 FW day−1, respectively (Figure 6). These rates are similar to, or above, those for pABAGlu hydrolysis.
We also tested whether the later folate-oxidation product, pterin-6-aldehyde, could be salvaged by reducing the aldehyde side chain and the pterin ring (Figure 6). Side-chain reduction (giving hydroxymethylpterin) was readily detectable in leaf and pericarp tissue, but not side-chain-plus-ring reduction (giving hydroxymethyldihydropterin). It is noteworthy that pericarp was more active than leaf tissue in pterin-6-aldehyde reduction, while the reverse was true of dihydropterin-6-aldehyde reduction: in the simplest case, this suggests that more than one enzyme is involved.
Demonstration of salvage reactions in vitro
To corroborate the in vivo evidence for pABAGlu hydrolysis and dihydropterin-6-aldehyde reduction, the same reactions were sought in vitro using desalted protein extracts of Arabidopsis and pea leaves, and tomato pericarp (Table 1). Both reactions were easily detected, with NADPH as cofactor for the reduction. Their relative activities varied, with pABAGlu hydrolysing activity in the order tomato > Arabidopsis > pea; and dihydropterin-6-aldehyde reduction in the opposite order. Even the lowest of these activities, pABAGlu hydrolysis in pea leaves (0.026 nmol h−1 mg−1 protein), was quite high relative to the folate content of the source tissue (approximately 0.2 nmol mg−1 protein; Chan and Cossins, 2003). As with the in vivo results, the reduction of pterin-6-aldehyde to hydroxymethylpterin was also detected, but not the further reduction to hydroxymethyldihydropterin (Table 1). Also as with in vivo data, tomato pericarp was more active than Arabidopsis with pterin-6-aldehyde as substrate, and less active with dihydropterin-6-aldehyde.
|Tissue||pABAGlu hydrolysis (nmol h−1 mg−1 protein)||NADPH-dependent aldehyde group reduction (nmol h−1 mg−1 protein)|
|Pea leaf||0.026 ± 0.002||21.5 ± 0.5||45.7 ± 0.5|
|Arabidopsis leaf||0.137 ± 0.010||1.79 ± 0.03||4.31 ± 0.06|
|Tomato pericarp||1.86 ± 0.09||0.45 ± 0.05||26.0 ± 0.2|
Strikingly little is known about folate salvage in any organism, given the lability of folates and the plentiful data on salvage of more stable cofactors such as nucleotides, NAD+, and pyridoxal 5′-phosphate (Neidhardt, 1996). In the case of plants, this lack of information is unfortunate as fruits and vegetables – major folate sources for humans – are subject to large post-harvest folate losses, and declining salvage capacity could be among the causes. Moreover, leaves endure oxidative and UVB stresses (Fryer et al., 2002; Gorton and Vogelmann, 1996) that destroy folates, pointing to the existence of robust adaptive mechanisms, of which salvage may well be one.
We first designed a method to measure total pABAGlu, as the standard procedure (McPartlin et al., 1992) does not detect pABAGlun and is set up for urine, which has far fewer potential interferants than plant tissues. Our method – built on the specificity of carboxypeptidase G and on the partitioning properties of pABA – worked well on diverse samples and provided the first estimates of pABAGlu pool sizes in plants. To help optimize this method, and to assay pABAGlu hydrolysis, we also developed an efficient radiochemical synthesis for pABAGlu from [14C]pABA. This procedure was shown also to work well for [3H]pABAGlu, thus providing an economical alternative to the usual protocol based on cleaving [3H]folic acid (McPartlin et al., 1992).
The data we obtained strengthen the case for substantial folate-breakdown rates in plant tissues, and indicate that the breakdown products can be salvaged for re-use. While none of the individual strands of evidence is unequivocal, they are collectively quite compelling, as discussed below.
Respecting breakdown, blocking folate synthesis with sulfanilamide and monitoring folate disappearance suggested that the rate in Arabidopsis leaves under such conditions is approximately 10% per day (assuming it to be constant), and that it is normally countered by de novo or salvage synthesis. A similar rate of net breakdown was seen in detached tomato fruit; the true breakdown rate may have been higher as some synthesis could have occurred. Note, however, that sulfanilamide treatment disrupts C1 metabolism (Prabhu et al., 1998), and that this may affect folate-breakdown rates. Likewise, care is needed in extrapolating post-harvest folate-breakdown rates to intact plants as harvest may affect both synthesis and breakdown processes.
The evidence for folate salvage reactions is fourfold. First, pABAGlu moieties did not accumulate greatly in any tissue tested, the levels being certainly no more than 4–14% that of folate, and probably less. Supposing folate-breakdown rates to be approximately 10% per day, such pABAGlu pools would represent, at most, one day's folate breakdown, from which it can be inferred that pABAGlu is salvaged or otherwise metabolized. Similarly, the characteristic pterin end product pterin-6-carboxylate did not accumulate massively, although its levels in three tissues surpassed those of pABAGlu. Although this difference might suggest that pterins are less efficiently salvaged than pABAGlu, it is possible that folate breakdown was not the sole source of pterin-6-carboxylate, as this can, in principle, also come from intermediates of folate synthesis (Rembold, 1985).
A second line of evidence for salvage is that pABAGlu pools in Arabidopsis changed little during sulfanilamide treatment despite a large concurrent folate loss, whereas the free pABA pool expanded. These observations confirm the above inference that pABAGlu is readily metabolized, and are consistent with the view that its main fate is hydrolysis to pABA and Glu.
The last – and most direct – lines of evidence for salvage reactions are that plant tissues readily converted supplied folate breakdown products to folate synthesis precursors in vivo and that protein extracts did so in vitro. pABAGlu was hydrolysed to give p-aminobenzoate, and dihydropterin-6-aldehyde was reduced to hydroxymethyldihydropterin; the oxidized form of the latter, pterin-6-aldehyde, was likewise reduced to hydroxymethylpterin.
Although our observations do not directly prove that folate-breakdown products are salvaged in vivo, it may be strongly inferred that this occurs for pABAGlu at least. Specifically, there is abundant evidence that plants convert free pABA to folates (Imeson et al., 1990; Zheng et al., 1992) and none that they catabolize it. The demonstration that plants release pABA from pABAGlu is thus essentially tantamount to showing that pABAGlu is salvaged.
Neither in vivo nor in vitro tests with pterin-6-aldehyde as substrate demonstrated reduction of the pterin ring to the dihydro form, a salvage reaction well known in Leishmania (Nare et al., 1997). Nor did parallel trials with hydroxymethylpterin show measurable ring reduction (A.D.H. and E.P.Q., unpublished data). It would, however, be premature to conclude that plants lack such a capacity. For example, if the ring-reduction reaction was specific for pterin-6-aldehyde, and much slower than reduction of the aldehyde side chain, our experiments would not have detected it.
Salvage of pABAGlu moieties by direct reincorporation into dihydrofolate via the action of DHPS seems unlikely from the properties of this enzyme. Like its microbial counterparts, plant DHPS can use pABAGlu, but apparently too inefficiently for this reaction to be significant in vivo given the low levels of pABAGlu moieties in plant tissues. Furthermore, in effect DHPS can use only pABAGlu, not its polyglutamylated forms, but this enzyme is located in mitochondria (Rébeilléet al., 1997) whereas the only enzyme known to remove polyglutamyl tails, γ-glutamyl hydrolase, is vacuolar (Orsomando et al., 2005). This arrangement imposes a lengthy trajectory with various transport steps on pABAGlun before it could be recycled by DHPS.
In summary, this work demonstrates that folate breakdown can occur at relatively high rates in plants, and provides unequivocal evidence for reactions that salvage the breakdown products. It further shows that (i) pABAGlu and its polyglutamyl forms are most probably recycled via prior conversion to free pABA, thereby pointing to a central role for pABAGlu hydrolase activity in the salvage process; and (ii) a primary pterin product of folate breakdown, dihydropterin-6-aldehyde, can potentially be recycled by reducing the 6-aldehyde group, thereby implicating a pterin aldehyde reductase. More generally, our data indicate that plants are excellent models to investigate folate salvage, an area of biochemistry that has been almost completely neglected in plants, fungi and bacteria.
Pterins, pABAGlu and pABAGlu5 were from Schircks Laboratories (Jona, Switzerland). 6-Hydroxymethylpterin pyrophosphate was reduced to its dihydro form as detailed (Scrimgeour, 1980). pABA-glucose ester and -N-glucoside were prepared as described (Quinlivan et al., 2003). [Ring-14C]pABA (55 mCi mmol−1) was from Moravek Biochemicals (Brea, CA, USA). Dowex resins were from Bio-Rad (Hercules, CA, USA). TLC plates were from Merck (Darmstadt, Germany).
Radiosynthesis of pABAGlu
Reactions were protected from light. Drying steps were in vacuo without heating. [14C]pABA (10 μCi, 180 nmol) was dried in a 1.5-ml tube and treated for 2 h at 22°C with 50 μl acetic anhydride with occasional mixing. The product was dried, retreated with acetic anhydride, and redried. The N-acetyl-pABA product was then mixed with 8 μl 1 m NaHCO3 (brought to pH 4 with HCl), 8 μl 1 ml-glutamate diethyl ester, and 20 μl 1 mN-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride. After 1 h at 22°C, the reaction mixture was applied to a 2-cm origin on a 10-cm silica gel 60 F254 TLC plate, which was developed with ethyl acetate:methanol:water (77:13:10, v/v/v). The N-acetyl-pABAGlu diethyl ester zone (Rf ≈ 0.6) was located by quenching of the fluorescent indicator or by autoradiography, scraped from the plate, and eluted with 3 × 200 μl 80% methanol. The eluate was dried, dissolved in 60 μl 0.2 m HCl, transferred to a 1-ml glass vial closed with a Teflon-faced septum, and heated for 45 min at 100°C. The reaction products were dried, redissolved in 20 μl water, and applied to a 3-cm origin on a TLC plate as above, which was developed with ethyl acetate:methanol:water (57:23:20, v/v/v). The pABAGlu zone (Rf ≈ 0.4) was located and eluted as above. The eluate was dried and redissolved in 200 μl water. Radiochemical yield was approximately 20% and radiochemical purity was >98%.
Plants and growing conditions
Arabidopsis plants (ecotype Columbia) were grown in potting soil in chambers (8-h days at 18–22°C or 12-h days at 23–28°C, 80–90 μmol photons m−2 sec−1). Plants for in vivo pterin-metabolism experiments were grown in 8-h days for 7–8 weeks; other plants were grown in 12-h days for times given in the text. Arabidopsis plantlets were cultured axenically in 0.5× liquid MS salts containing 10 g l−1 sucrose (Prabhu et al., 1996). Tomato plants (cv. Micro-Tom) were grown as described in a chamber (Díaz de la Garza et al., 2004) or glasshouse (Simkin et al., 2004). Pea plants (cv. Laxton's Progress 9) were grown in a chamber as described (Orsomando et al., 2005).
The procedure entailed: (i) extraction conditions designed to minimize folate breakdown; (ii) an initial step to deplete free pABA by partitioning it at pH ≈ 5.5 into ethyl acetate; (iii) digestion of the pABA-depleted extract at pH ≈ 7.2 by Pseudomonas carboxypeptidase G (Sigma, St. Louis, MO, USA), which removes γ-glutamyl tails from pABAGlun and cleaves pABAGlu to pABA plus Glu (McCullough et al., 1971); (iv) partitioning of the pABA so released into ethyl acetate at pH ≈ 5.5, followed by back-extraction; and (v) quantification of pABA by HPLC. Tissues were pulverized in liquid N2, and 100-mg portions of the frozen powder were extracted for 5 min at 100°C in 0.8 ml N2-sparged 50 mm 2-(N-morpholino)-ethanesulfonic acid (MES)–KOH containing 0.1% (w/v) sodium ascorbate and 10 mmβ-mercaptoethanol, and adjusted to pH 5.5. After cooling and centrifuging, a 0.75-ml aliquot of the extract was readjusted if necessary to pH 5.3–5.8 with 1 m HCl, then shaken for 5 min with 4 ml ethyl acetate that had been pre-equilibrated with 50 mm MES–KOH pH 5.2. The ethyl acetate phase was discarded. The aqueous phase was adjusted to pH 7.0–7.3 with 1 m KOH, mixed with 7.5 μl 10 mm ZnCl2 and 0.1 U carboxypeptidase G (in 10 μl of 50 mm Tris–HCl pH 7.0), and incubated at 37°C for 4–5 h. The pH was then readjusted to 5.3–5.8, followed by partitioning against 4 ml ethyl acetate, as above. The ethyl acetate phase was back-extracted by shaking for 5 min with 0.2 ml 0.3 m KOH. The KOH phase was withdrawn, its volume measured, and 50-μl samples were injected onto a Supelco (St. Louis, MO, USA) Discovery C18 column (4.6 × 250 mm, 5 μm particle size) and quantified by fluorescence (270 nm excitation, 350 nm emission). The column was eluted at 1 ml min−1 with a 6-min linear gradient from 2 to 29% methanol in 50 mm sodium phosphate buffer pH 6.0, containing 8 mm tetrabutylammonium bisulfate. Each batch of assays included a blank without carboxypeptidase, and samples spiked with 50 pmol pABAGlu or pABAGlu5 from which recovery values were calculated and used to correct the data. Recoveries of pABAGlu and pABAGlu5 were typically 40–60%. Representative samples were spiked with 200–250 pmol THF or its 5-methyl, 10-formyl or 5-formyl derivatives to assess the amount of pABAGlu formed by folate breakdown during the assay.
For analysis of endogenous pterin-6-carboxylate, tissue (0.2 g FW) was ground in liquid N2, suspended in 3.6 ml methanol:chloroform:water (12:5:1, v/v/v), heated to 80°C for 5 min, and vortex mixed for 5 min. After clearing by centrifugation, extracts were split into aqueous and organic phases by adding 0.5 ml chloroform and 0.75 ml water, mixed thoroughly, and centrifuged. The aqueous phase was dried in vacuo without heating, taken up in 2 ml water, and fractionated by an ion-exchange procedure based on that of Stea et al. (1980). The 2-ml sample was applied to a 0.6-ml, 7-mm diameter column of AG-50WX8 (H+), and the flow-through and a 15-ml water wash were run straight onto a 0.4-ml, 7-mm diameter column of AG1-X8 (OH−). The AG1-X8 column was then washed with 6 ml water and eluted with 2 ml 1 m HCl containing 14% (v/v) methanol and 10% (v/v) acetonitrile. The eluates were dried as above and redissolved in 0.2 ml water. No reduced pterins remained in plant samples prepared as above, so no oxidation step was needed to convert pterins to their fully oxidized, fluorescent forms before HPLC. Recoveries of pterin-6-carboxylate spikes were 63–86%; these values were used for data correction.
For pterin metabolism experiments, tissues were ground as above and extracted at 50°C for 5 min in 2 ml methanol:chloroform:water containing 0.1% (w/v) sodium ascorbate and 10 mmβ-mercaptoethanol. After phase-splitting as above, the aqueous phase was concentrated to approximately 0.05–0.1 ml in vacuo without heat, adjusted to 0.2 ml with water, and analysed at once or held under N2 at −80°C. Just before HPLC analysis, 50-μl aliquots were treated for 1 h with 80 μl 1 n HCl containing 1% I2/2% KI (w/v) (which oxidizes reduced pterins) or with 1 n HCl alone (which does not oxidize pterins). All samples then received 5 μl 5% (w/v) sodium ascorbate (to destroy excess I2) and 7 μl 10 n NaOH. Recoveries of 1-nmol spikes of hydroxymethylpterin and its dihydro form were 65 and 31–46%, respectively; these values were used for data correction.
Pterins (40- or 50-μl injections) were separated by HPLC using a 4-μm, 250 × 4.6-mm Synergi Fusion-RP 80 column (Phenomenex, Torrance, CA, USA) eluted isocratically with 10 mm sodium phosphate (pH 6.0) at 1.5 ml min−1. Peaks were detected by fluorescence (350 nm excitation, 450 nm emission) and quantified relative to standards. Hydroxymethyldihydropterin was estimated from the increase in the hydroxymethylpterin peak on oxidation. The identity of the hydroxymethylpterin peak was confirmed by LC–MS from the presence of the molecular ion (M + H+) at m/z 194, and diagnostic fragment ions at m/z 176, 134 and 106. Pterin concentrations in standard solutions were determined using published extinction coefficients (Blakley, 1969; Pfleiderer, 1985), except for dihydropterin-6-aldehyde, for which we estimated a value of 10 400 m−1 cm−1 at 422 nm and pH 7.5. Solutions of dihydropterins contained 10 mmβ-mercaptoethanol and were kept under N2.
Analysis of folates, pABA and pABA glucose ester
Folates were extracted from samples (0.5–1 g FW), deglutamylated, affinity purified, and determined by HPLC with electrochemical detection as described (Díaz de la Garza et al., 2004; Goyer et al., 2005). Free and total pABA were determined as described (Quinlivan et al., 2003), except that the ion-exchange step was omitted. Instead, a final concentration of 50 mm MES–KOH (pH 5.2) was added to each sample (both unhydrolysed and acid-hydrolysed) and the pH was adjusted to 4.6–5.0 with NaOH. The pABA was then extracted into ethyl acetate as before. Using this method, <0.1%pABA glucose ester was converted to pABA in the absence of acid hydrolysis. The difference between free and total pABA values was taken to be pABA glucose ester.
Metabolism of [14C]pABAGlu in vivo
Samples consisted of (i) five Arabidopsis rosette leaves trimmed to 2 cm long and stripped of the mid-vein on the abaxial surface; or (ii) single 1-cm-diameter pericarp disks from mature green tomato fruit, given eight radial cuts on the inner surface. [14C]pABAGlu (92 nCi, 1.7 nmol) in 10 μl water was supplied to the cut surfaces (2 μl per leaf or 10 μl per disk) followed, after uptake, by 2 μl water in the case of leaf sections. Samples were then incubated on moist filter paper in Petri dishes at 26°C for 6 h in fluorescent light (150 μmol photons m−2 sec−1). Non-absorbed 14C was washed out by shaking samples for 15 min in 5 ml 20 μm unlabelled pABAGlu. An aliquot (1.5 ml) of the washing out solution was dried in vacuo for TLC analysis (see below). Tissues were extracted by grinding in 2 ml semi-frozen methanol and warming to 60°C for 5 min, then re-extracted by shaking in 0.5 ml water at 22°C for 5 min. The methanol and water extracts were pooled, and 0.2-ml aliquots were dried in vacuo, redissolved in 5 μl water, mixed with unlabelled standards (pABA, pABA glucose ester, and pABA N-glucoside), and applied to 1-cm origins on TLC plates as above. Plates were developed with ethyl acetate:methanol:water (77:13:10, v/v/v) and autoradiographed to locate labelled zones, which were scraped out for 14C assay by scintillation counting. Unlabelled samples spiked with [14C]pABAGlu were used to correct for slight (approximately 0.5%) chemical decomposition of [14C]pABAGlu during extraction, and for recovery.
Metabolism of pterins in vivo
Tissue samples were as for [14C]pABAGlu experiments. Samples received 20 μl 10 mm potassium phosphate pH 7.5, alone (control) or containing 3.4 nmol dihydropterin-6-aldehyde or 4.9 nmol pterin-6-aldehyde; the dihydropterin solution also contained 10 mmβ-mercaptoethanol. After incubating on moist filter paper in darkness for 14.5 h at 22°C, non-absorbed pterins were washed out by shaking in 5 ml water for 30 min. Extracts were prepared and analysed for hydroxymethyldihydropterin and hydroxymethylpterin by HPLC, as above. Spikes of dihydropterin-6-aldehyde and pterin-6-aldehyde confirmed that unmetabolized aldehydes did not give rise to significant amounts of the corresponding hydroxymethyl compounds during sample work-up.
Tissue extracts and enzyme assays
Tissues were ground in liquid N2. Subsequent operations were at 0–10°C. For pABAGlu hydrolysis assays, the frozen powder was extracted in 3.5 vol 50 mm potassium phosphate pH 8, containing 3% (w/v) polyvinylpolypyrolidone. The brei was centrifuged (10 000 g, 20 min) and the supernatant was brought to 80% saturation with (NH4)2SO4. The precipitated proteins were resuspended in 50 mm potassium phosphate pH 8, desalted on a PD10 column equilibrated with the same buffer plus 10% (v/v) glycerol, flash-frozen, and stored at −80°C. Protein was estimated by dye-binding (Bradford, 1976). pABAGlu hydrolysis was assayed in 10-μl reactions containing 8 μl protein extract (8–100 μg protein), 10 nmol MnCl2, and 46 nCi (0.83 nmol) of [14C]pABAGlu. After incubating for 3 h at 30°C, 0.1 ml 0.1 m sodium citrate buffer pH 5.5 was added and the mixture was shaken for 5 min with 0.6 ml ethyl acetate. The [14C]pABA partitioned into the organic phase was quantified by scintillation counting. Substrate consumption in the assays was <10%.
For pterin-reduction assays, tissues were ground as above and extracted with 2 vol 100 mm potassium phosphate pH 7.5 containing 5 mm dithiothreitol and 3% (w/v) polyvinylpolypyrolidone. After centrifuging, the supernatant was desalted on a PD10 column equilibrated with 50 mm potassium phosphate pH 7.5 plus 5 mm dithiothreitol and 10% glycerol, and frozen as above. Assays (final volume 50 μl, pH 7.5) contained 4–20-μl protein extract (17–29 μg protein), 1.8 μmol potassium phosphate, 0.5 nmol pterin and 50 nmol NADPH; blanks contained no NADPH. After 15–30 min incubation at 30°C, 20-μl samples were treated for 1 h with 10 μl 1 n HCl containing 1% I2/2% KI or with 1 n HCl alone, after which they received 10 μl 5% sodium ascorbate and 60 μl 10 mm sodium phosphate pH 6.0, containing 1% (w/v) Na-ascorbate and 10 mmβ-mercaptoethanol (to stabilize dihydropterins). Hydroxymethylpterin and its dihydro form were assayed by HPLC as above. Pterin utilization in the assays was ≤50% and NADPH consumption was ≤13%.
Dihydropteroate synthase studies
Dihydropteroate synthase activity was partially purified from pea leaf mitochondria as described (Rébeilléet al., 1997). Soluble proteins from purified mitochondria were loaded on a Superdex 200 (Pharmacia, Piscataway, NJ, USA) column (60 × 1 cm) pre-equilibrated with the following buffer: 10 mm potassium phosphate, 10 mm Tris pH 7.2, 1 mm EDTA, 1 mm dithiothreitol, 10 mmβ-mercaptoethanol, 10% (w/v) glycerol. Proteins were eluted with the same buffer (flow rate 0.3 ml min−1). Fractions containing DHPS activity were combined and concentrated to 2 ml by ultrafiltration on a 10-kDa cut-off membrane. DHPS activity was assayed in reaction mixtures (final volume 120 μl) containing 40 mm Tris–HCl (pH 8.0), 20 mmβ-mercaptoethanol, 10 mm MgCl2, 40 μg partially purified protein and various amounts of pABAGlu or pABAGlu5. Two μl 2 mm [carboxyl-14C]pABA (50 mCi mmol−1) were added, then the reaction was started by adding 6-hydroxymethyl-7,8-dihydropterin pyrophosphate (final concentration 100 μm). [14C]7,8-Dihydropteroate formation was estimated by reverse-phase HPLC and scintillation counting as described (Mouillon et al., 2002).
We thank Dr J. Ayling for advice and M. Ziemak for technical help. This work was supported in part by National Science Foundation grant no. MCB-0443709, by a C.V. Griffin, Sr Foundation endowment, and by the Florida Agricultural Experiment Station.
- 2002) Folate synthesis in plants: the first step of the pterin branch is mediated by a unique bimodular GTP cyclohydrolase I. Proc. Natl Acad. Sci. USA, 99, 12489–12494. , , , , , , , and (
- 1994) PTR1: a reductase mediating salvage of oxidized pteridines and methotrexate resistance in the protozoan parasite Leishmania major. Proc. Natl Acad. Sci. USA, 91, 11442–11446. , , , and (
- 1969) Chemical and physical properties of pterins and folate derivatives. In The Biochemistry of Folic Acid and Related Pteridines. New York: Wiley, pp. 58–105. (ed.) (
- 1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. (
- 1961) The biosynthesis of folic acid. I. Substrate and cofactor requirements for enzymatic synthesis by cell-free extracts of Escherichia coli. J. Biol. Chem. 236, 2534–2543. , and (
- 2003) The intracellular distribution of folate derivatives in pea leaves. Pteridines, 14, 67–76. and (
- 2000) The fascinating world of folate and one-carbon metabolism. Can. J. Bot. 78, 691–708. (
- 2004) Folate biofortification in tomatoes by engineering the pteridine branch of folate synthesis. Proc. Natl Acad. Sci. USA, 101, 13720–13725. , , , , and (
- 1973) The enzymic synthesis of dihydropteroate and dihydrofolate by Plasmodium berghei. J. Protozool. 20, 459–464. (
- 2002) Imaging of photo-oxidative stress responses in leaves. J. Exp. Bot. 53, 1249–1254. , , and (
- 1996) Effects of epidermal cell shape and pigmentation on optical properties of antirrhinum petals at visible and ultraviolet wavelengths. Plant Physiol. 112, 879–888. and (
- 2005) 5-Formyltetrahydrofolate is an inhibitory but well tolerated metabolite in Arabidopsis leaves. J. Biol. Chem. 280, 26137–26142. , , , , , , and (
- 1996) Folate biosynthesis, reduction, and polyglutamylation. In Escherichia coli and Salmonella– Cellular and Molecular Biology, Vol. 1, 2nd edn (Neidhardt, F.C., ed.). Washington, D.C., USA: ASM Press, pp. 665–673. , and (
- 1989) Chemical and nutritional aspects of folate research: analytical procedures, methods of folate synthesis, stability, and bioavailability of dietary folates. Adv. Food Nutr. Res. 33, 1–101. (
- 2002) In vivo kinetics of folate metabolism. Annu. Rev. Nutr. 22, 199–220. and (
- 1967) Effect of substrate decomposition on the spectrophotometric assay of dihydrofolate reductase. Anal. Biochem. 21, 178–189. , and (
- 1998) Characterization of mutations that allow p-aminobenzoyl-glutamate utilization by Escherichia coli. J. Bacteriol. 180, 6260–6268. , and (
- 1990) Folylpolyglutamyl derivatives of Pisum sativum L. Determination of polyglutamate chain lengths by high performance liquid chromatography following conversion to p-aminobenzoylpolyglutamates. Plant Cell Physiol. 31, 223–231. , and (
- 1949) Photolytic and enzymatic transformations of pteroylglutamic acid. J. Biol. Chem. 180, 389–398. , and (
- 1971) Purification and properties of carboxypeptidase G1. J. Biol. Chem. 246, 7207–7213. , and (
- 1992) The quantitative analysis of endogenous folate catabolites in human urine. Anal. Biochem. 206, 256–261. , , , and (
- 1971) Enzymatic conversion of 2-amino-4-hydroxy-6-formyl-7,8-dihydropteridine to 2-amino-4-hydroxy-6-hydroxymethyl-7,8-dihydropteridine by cell-free extracts of Escherichia coli B. J. Vitaminol. 17, 5–9. and (
- 2002) Folate synthesis in higher-plant mitochondria: coupling between the dihydropterin pyrophosphokinase and the dihydropteroate synthase activities. Biochem. J. 363, 313–319. , , and (
- 1997) The roles of pteridine reductase 1 and dihydrofolate reductase-thymidylate synthase in pteridine metabolism in the protozoan parasite Leishmania major. J. Biol. Chem. 272, 13883–13891. , and (
- 1996) Escherichia coli and Salmonella– Cellular and Molecular Biology, Vol. 1, 2nd edn. Washington, D.C., USA: ASM Press. (
- 2005) Plant γ-glutamyl hydrolases and folate polyglutamates: characterization, compartmentation, and co-occurrence in vacuoles. J. Biol. Chem. 280, 28877–28884. , , , , , , and (
- 1970) Dihydrofolate and dihydropteroate synthesis by partially purified enzymes from wild-type and sulfonamide-resistant Pneumonococcus. Biochemistry, 9, 355–361. (
- 1985) Chemistry of naturally occurring pteridines. In Folates and Pterins, Vol. 2 (Blakley, R.L. and Benkovic, S., eds). New York: Wiley, pp. 43–114. (
- 1996) 13C Nuclear magnetic resonance detection of interactions of serine hydroxymethyltransferase with C1-tetrahydrofolate synthase and glycine decarboxylase complex activities in Arabidopsis. Plant Physiol. 112, 207–216. , , and (
- 1998) Effects of sulfanilamide and methotrexate on 13C fluxes through the glycine decarboxylase/serine hydroxymethyltransferase enzyme system in Arabidopsis. Plant Physiol. 116, 137–144. , , , and (
- 2003) The folate precursor p-aminobenzoate is reversibly converted to its glucose ester in the plant cytosol. J. Biol. Chem. 278, 20731–20737. , , , , and (
- 1997) Folate biosynthesis in higher plants: purification and molecular cloning of a bifunctional 6-hydroxymethyl-7,8-dihydropterin pyrophosphokinase/7,8-dihydropteroate synthase localized in mitochondria. EMBO J. 16, 947–957. , , , and (
- 1980) Oxidation of tetrahydrofolic acid by air. J. Agric. Food Chem. 28, 801–805. and (
- 1985) Catabolism of pterins. In Folates and Pterins, Vol. 2 (Blakley, R.L. and Benkovic, S.J., eds). New York: Wiley, pp. 155–178. (
- 1984) Catabolism of folates. In Folates and Pterins, Vol. 1 (Blakley, R.L. and Benkovic, S.J., eds). New York: Wiley, pp. 307–327. (
- 2000) Folic acid and folates: the feasibility of nutritional enhancement in plant foods. J. Sci. Food Agric. 80, 795–824. , and (
- 1980) Methods for reduction, stabilization, and analyses of folates. Methods Enzymol. 66, 517–523. (
- 1985) Purification and properties of carboxypeptidase G2 from Pseudomonas sp. strain RS-16. Use of a novel triazine dye affinity method. Eur. J. Biochem. 148, 447–453. , , and (
- 1959) Enzymic synthesis of folic acid-like compounds by cell-free extracts of Lactobacillus arabinosus. Arch. Biochem. Biophys. 80, 155–161. (
- 1969) The enzymatic synthesis of hydroxylmethyldihydropteridine pyrophosphate and dihydrofolate. Biochemistry, 8, 5022–5028. , , and (
- 2004) The tomato carotenoid cleavage dioxygenase 1 genes contribute to the formation of the flavor volatiles beta-ionone, pseudoionone, and geranylacetone. Plant J. 40, 882–892. , , , and (
- 2002) Effect of ultraviolet C irradiation on folate and free amino acid contents in leaves of Pisum sativum. Biofizika, 47, 878–885. , and (
- 1980) Quantitative determination of pterins in biological fluids by high-performance liquid chromatography. J. Chromatogr. 188, 363–375. , , and (
- 2003) Folate content in strawberries (Fragaria × ananassa): effects of cultivar, ripeness, year of harvest, storage, and commercial processing. J. Agric. Food Chem. 51, 128–133. , , and (
- 2001) New perspectives on folate catabolism. Annu. Rev. Nutr. 21, 255–282. , and (
- 1979) Characterization of mutationally altered dihydropteroate synthase and its ability to form a sulfonamide-containing dihydrofolate analog. J. Bacteriol. 137, 129–136. , and (
- 1968) Synthesis of 2-amino-4-hydroxy-6-formyl-7,8-dihydropteridine and its identification as a degradation product of dihydrofolate. Arch. Biochem. Biophys. 126, 955–957. , , and (
- 1994) Subcellular volumes and metabolite concentrations in spinach leaves. Planta, 193, 530–535. , and (
- 1992) The polyglutamate nature of plant folates. Phytochemistry, 31, 2277–2282. , , and (