An atypical orthologue of 6-pyruvoyltetrahydropterin synthase can provide the missing link in the folate biosynthesis pathway of malaria parasites

Abstract

Folate metabolism in malaria parasites is a long-standing, clinical target for chemotherapy and prophylaxis. However, despite determination of the complete genome sequence of the lethal species Plasmodium falciparum, the pathway of de novo folate biosynthesis remains incomplete, as no candidate gene for dihydroneopterin aldolase (DHNA) could be identified. This enzyme catalyses the third step in the well-characterized pathway of plants, bacteria, and those eukaryotic microorganisms capable of synthesizing their own folate. Utilizing bioinformatics searches based on both primary and higher protein structures, together with biochemical assays, we demonstrate that P. falciparum cell extracts lack detectable DHNA activity, but that the parasite possesses an unusual orthologue of 6-pyruvoyltetrahydropterin synthase (PTPS), which simultaneously gives rise to two products in comparable amounts, the predominant of which is 6-hydroxymethyl-7,8-dihydropterin, the substrate for the fourth step in folate biosynthesis (catalysed by 6-hydroxymethyl-7,8-dihydropterin pyrophosphokinase; PPPK). This can provide a bypass for the missing DHNA activity and thus a means of completing the biosynthetic pathway from GTP to dihydrofolate. Supported by site-directed mutagenesis experiments, we ascribe the novel catalytic activity of the malarial PTPS to a Cys to Glu change at its active site relative to all previously characterized PTPS molecules, including that of the human host.

Fig. 1

The conventional folate (a) and biopterin (b) biosynthetic pathways as found in (a) plants, bacteria and lower eukaryotes that are capable of de novo folate synthesis, and (b) in mammals and other organisms that utilize 5,6,7,8-tetrahydrobiopterin (BH₄) as a cofactor. Certain organisms, such as some fungi, cyanobacteria and pseudomonads, possess both pathways. Pathway (c) involving the P. falciparum PTPS orthologue is demonstrated in this work and (d) shows the substrates (i), (iii) and products (ii), (iv) of conventional DHNA and PTPS enzymes respectively. Underlined product (ii) in pathways (a) and (c) is 6-hydroxymethyl-7,8-dihydropterin (6HMDP), the required substrate for PPPK. Asterisked product (iv) in pathway (c) was identified from its oxidation product (see text). Abbreviations: GTPC, GTP cyclohydrolase I; P, poorly defined phosphatase activity (thought in some systems to first involve loss of pyrophosphate then subsequent removal of the final phosphate); DHNA, dihydroneopterin aldolase; PPPK, hydroxymethyldihydropterin pyrophosphokinase; DHPS, dihydropteroate synthase; DHFS, dihydrofolate synthase; PTPS, pyruvoyltetrahydropterin synthase; SR, sepiapterin reductase.

Fig. 2

Aligned sequences of PTPS orthologues between the conserved His residues that co-ordinate the active site Zn²⁺ ion (arrows). The active site Cys residue conserved in all non-apicomplexan enzymes to date is marked by an asterisk above the sequences; the Glu residue proposed to act as nucleophile instead of Cys in Plasmodium is marked by an asterisk below (see also Fig. S3). Accession numbers for non-apicomplexan enzymes: human, Q03393; rat, P27213; guppy, Q90W95; Drosophila melanogaster, P48611; Caenorhabditis elegans, O02058; Escherichia coli, P65870; Synechocystis, Q55798. Gene loci for malarial sequences (http://www.plasmodb.org): Plasmodium gallinaceum (blast search); Plasmodium berghei, PB000950.03.0; Plasmodium vivax, Pv114505; Plasmodium falciparum, PFF1360w. The Toxoplasma gondii sequence was also determined from blast searching (http://www.toxodb.org).

Fig. 3

HPLC separation of pterin products arising from the reactions of the PTPS orthologues from P. falciparum (Pf₁ and Pf₂; solid line), E. coli (Ec; dotted line) and human (Hs; short-dashed line) using the normal substrate dihydroneopterin triphosphate (DHNTP), followed by oxidation to maximize fluorescence from the pterin ring system (Nichol et al., 1985). This process converts the products of the reaction 6-pyruvoyl-5,6,7,8-tetrahydropterin to pterin, and 6HMDP to 6-hydroxymethylpterin. No substrate (long-dashed line) and no enzyme (dot-dashed line) controls showed no fluorescence in these positions. Vertical arrows indicate the retention times of the pterin and 6-hydroxymethylpterin standards.

Fig. 4

Confirmation of 6HMDP production by the P. falciparum PTPS (Pf) orthologue, but not by those from human (Hs) and E. coli (Ec). Products of the PTPS reaction were coupled to the next two enzymes in the pathway, PPPK and DHPS, in the form of recombinant bifunctional enzyme from T. gondii, and production of ¹⁴C-radiolabelled 7,8-dihydropteroate assayed in triplicate on a Typhoon imager [see Fig. S2, scheme (b), for the reaction set-up and more experimental detail]. Top, Typhoon image; bottom, plot of counts recorded after reaction times up to 45 min and expressed per 1 μl of reaction mix per hour of Typhoon screen exposure (mean ± SD, relative to zero time); (a) PfPTPS with DHNTP substrate, (b) HsPTPS with DHNTP substrate, (c) EcPTPS with DHNTP substrate, (d) 6HMDP, (e) no enzyme, but DHNTP substrate present, (f) PfPTPS, but no substrate present. In the bottom panel, negative controls (e) and (f) are omitted for clarity. In the positive control (d), commercial 6HMDP was provided as substrate for the PPPK step. One hundred counts on the ordinate are equivalent to 0.071 pmol product per μl of reaction mix. Qualitatively similar results were obtained when P. falciparum PPPK-DHPS was used in the coupled reaction instead of T. gondii PPPK-DHPS.

Fig. 5

Formation of ¹⁴C-radiolabelled 7,8-dihydropteroate starting from either DHNTP (black) or DHN (grey) substrate (20 μM, 1 h incubation, 37°C) as recorded on the Typhoon imager (per 1 h screen exposure) from comparable amounts of (a) lysate from untransformed E. coli cells (and thus containing only EcDHNA) and (b) lysate from transformed E. coli cells overexpressing PfPTPS. The same experiment is repeated in (c) using affinity chromatography/ion-exchange-purified PfPTPS obtained from an equal volume of lysate as used in (b). One hundred counts on the ordinate are equivalent to 0.071 pmol product per μl of reaction mix.

Fig. 6

Effect of mutation of the Glu-38 residue in the P. falciparum PTPS orthologue on the reaction products. HPLC separated pterin products arising from the reactions of the wild type (solid line), E38C variant (dashed line), E38Q variant (dotted line) and E38L variant (dot-dashed line) using dihydroneopterin triphosphate (DHNTP) as substrate, followed by processing and calibration as for Fig. 3.