The Human Malaria Parasite Plasmodium falciparum Is Not Dependent on Host Coenzyme A Biosynthesis*

Pantothenate, a precursor of the fundamental enzyme cofactor coenzyme A (CoA), is essential for growth of the intraerythrocytic stage of human and avian malaria parasites. Avian malaria parasites have been reported to be incapable of de novo CoA synthesis and instead salvage CoA from the host erythrocyte; hence, pantothenate is required for CoA biosynthesis within the host cell and not the parasite itself. Whether the same is true of the intraerythrocytic stage of the human malaria parasite, Plasmodium falciparum, remained to be established. In this study we investigated the metabolic fate of [14C]pantothenate within uninfected and P. falciparum-infected human erythrocytes. We provide evidence consistent with normal human erythrocytes, unlike rat erythrocytes (which have been reported to possess an incomplete CoA biosynthesis pathway), being capable of CoA biosynthesis from pantothenate. We also show that CoA biosynthesis is substantially higher in P. falciparum-infected erythrocytes and that P. falciparum, unlike its avian counterpart, generates most of the CoA synthesized in the infected erythrocyte, presumably necessitated by insufficient CoA biosynthesis in the host erythrocyte. Our data raise the possibility that malaria parasites rationalize their biosynthetic activity depending on the capacity of their host cell to synthesize the metabolites they require.

of the intraerythrocytic stage of P. lophurae, and proposed that avian malaria parasites are incapable of metabolizing pantothenate to CoA, and instead rely on CoA synthesized by the host erythrocyte. In support of this proposal, CoA biosynthesis enzymes are readily detectable in duck erythrocytes, but appear to be absent from P. lophurae parasites isolated from their host erythrocyte (7,8). Pantothenate is therefore required by the P. lophurae-infected duck erythrocyte for CoA biosynthesis within the host cell and not the parasite itself.
By contrast with nucleated avian erythrocytes, mammalian erythrocytes are thought to be incapable of CoA biosynthesis. In the only study on the subject, Annous and Song (9) reported that although pantothenate is phosphorylated within rat erythrocytes (the first step in CoA biosynthesis), there is no evidence for the subsequent steps of the CoA biosynthesis pathway. Saliba et al. (10) confirmed that human erythrocytes similarly phosphorylate pantothenate, but did not investigate whether CoA synthesis also occurs in the cells. A lack of CoA biosynthesis in mammalian erythrocytes would seemingly place the burden of CoA synthesis squarely on malaria parasites that infect mammals (such as P. falciparum), contrary to the situation in birds. Although Saliba et al. (10) have shown that P. falciparum is capable of performing the first step in CoA biosynthesis, it remains to be established whether the parasite can metabolize the 4Ј-phosphopantothenate generated from pantothenate to CoA or, like P. lophurae, relies on CoA synthesized in the host erythrocyte for its normal growth and replication.
In this study we followed the metabolism of pantothenate within uninfected human erythrocytes, P. falciparum-infected human erythrocytes, and isolated P. falciparum parasites. We provide evidence that both uninfected erythrocytes (which we show do take up pantothenate, albeit very slowly) and P. falciparum-infected erythrocytes synthesize CoA from pantothenate. CoA biosynthesis is, however, dramatically higher in the P. falciparum-infected cell. Furthermore, we show that P. falciparum parasites synthesize CoA in the absence of the host erythrocyte, and hence, by contrast with avian malaria parasites, the human malaria parasite does not rely on the host erythrocyte for CoA.

EXPERIMENTAL PROCEDURES
Cell Culture and Preparation-Erythrocytic stage P. falciparum parasites (strain 3D7) were maintained within human erythrocytes in continuous culture as described elsewhere (11). Immediately prior to experimentation, cultures of predominantly trophozoite stage P. falciparum-infected erythrocytes were concentrated to Ͼ95% parasitemia using magnet-acti-vated cell sorting essentially as described elsewhere (12). Briefly, infected erythrocytes were passed through a CS column (attached to a 21-gauge flow resistor) mounted in a Miltenyi Biotec VarioMACS magnet. The trophozoite-infected erythrocytes retained on the column were washed (to remove uninfected erythrocytes) and subsequently eluted (following removal of the column from the magnetic field) with pantothenate-free medium (RPMI 1640 culture medium prepared without pantothenate and supplemented with 25 mM HEPES, 11 mM glucose, 200 M hypoxanthine, 24 g/ml gentamicin, and 0.6% (w/v) Albumax II), pH 7.4. The cells were washed (four times) and suspended in pantothenate-free medium, pH 7.4. For experiments with parasites isolated from their host erythrocyte, following enrichment to Ͼ95% parasitemia, trophozoite stage P. falciparum-infected erythrocytes in suspension were treated with saponin (0.05% (w/v)) essentially as described previously (10). Following treatment with saponin, which permeabilizes the erythrocyte and parasitophorous vacuole membranes that enclose the parasite, while leaving the parasite plasma membrane intact, the cells were washed (four times) and suspended in pantothenate-free medium, adjusted to pH 7.1 to match the pH in the cytosol of a human erythrocyte infected with a trophozoite stage P. falciparum parasite (13).
In preparation for experiments involving uninfected erythrocytes, the erythrocytes were washed to ensure minimal contamination with leukocytes. Briefly, the erythrocytes were suspended in RPMI 1640 culture medium supplemented with 25 mM HEPES, 11 mM glucose, 200 M hypoxanthine, and 24 g/ml gentamicin, and then centrifuged (1500 ϫ g for 8 min). The supernatant and cells at the interface between the medium and the packed erythrocytes were removed, and the wash was repeated twice more. The washed cells were then incubated in the same manner as the P. falciparum-infected erythrocyte cultures for 12-18 h to ensure that the uninfected and infected erythrocytes were comparable. Immediately prior to experimentation the cells were washed (an additional three times) and suspended in pantothenate-free medium, pH 7.4.
Leukocyte Quantitation-The number of contaminating leukocytes in uninfected erythrocyte preparations was estimated by manual counting using a hemocytometer and Turk reagent. Turk reagent (2% (v/v) acetic acid, 0.01% (w/v) crystal violet) effectively lyses the erythrocytes and stains the leukocytes purple, facilitating their localization on the hemocytometer grid and allowing their enumeration. Erythrocytes, obtained from the blood bank as "leukocyte reduced" packed cells, were washed (as described in the previous paragraph) and suspended in saline (137 mM NaCl, 2.7 mM KCl, 10 mM phosphate buffer, pH 7.4) at a density of 3.7-5.1 ϫ 10 8 cells/ml. An aliquot (1 ml) was centrifuged (16,300 ϫ g for 1 min) and 950 l of the supernatant removed. The cells were then vortexed to resuspend them in the remaining supernatant and 950 l of Turk reagent was added to lyse the erythrocytes and stain the leukocytes. Following a 1-min incubation at room temperature the samples were centrifuged (16,300 ϫ g for 1 min) to concentrate the leukocytes and 950 l of the supernatant was removed. The pelleted leukocytes were then gently resuspended in the remaining supernatant (by pipetting) and used for cell counting. The total number of leukocytes present in 0.9 l (the volume above the entire hemocytometer grid) of the resultant 50 l of concentrated leukocyte preparations were counted. The leukocytes in each batch of washed erythrocytes were counted twice on separate days.
Pantothenate Uptake and Phosphorylation Measurements-Pantothenate uptake and phosphorylation were measured essentially as described by Saliba et al. (10) using [ 14 C]pantothenate (American Radiolabeled Chemicals). [ 14 C]Pantothenate (0.1 Ci/ml; 1.8 M; a concentration within the normal human whole blood total pantothenate concentration range (14)) was added to uninfected and P. falciparum-infected erythrocytes suspended in pantothenate-free medium, pH 7.4, and the reactions incubated at 37°C on a horizontally rotating shaker in an atmosphere of 96% nitrogen, 3% carbon dioxide, and 1% oxygen. Aliquots (200 l) of the suspensions (3.8 -5.1 ϫ 10 8 cells/ml for uninfected erythrocytes and 3.0 -3.7 ϫ 10 7 cells/ml for P. falciparum-infected erythrocytes) were removed in duplicate at appropriate time intervals and centrifuged immediately through 300 l of oil (dibutyl phthalate) layered over 50 l of 15% (w/v) perchloric acid at 15,800 ϫ g for 2 min to terminate the uptake and intracellular metabolism simultaneously. Samples were processed as described by Saliba et al. (10) for quantitation of total [ 14 C]pantothenate uptake and phosphorylated [ 14 C]pantothenate using the Somogyi reagent (15), which precipitates phosphorylated compounds from solution.
The amount of [ 14 C]pantothenate trapped in the extracellular space between uninfected erythrocytes centrifuged into the acid layer was estimated simply by mixing [ 14 C]pantothenate (0.1 Ci/ml), which is taken up only very slowly by uninfected erythrocytes ( Fig. 2A), with uninfected erythrocytes (at the same cell density as in the uptake experiment) immediately before centrifuging aliquots through dibutyl phthalate into perchloric acid. The amount of [ 14 C]pantothenate trapped in the extracellular space between the P. falciparum-infected erythrocytes was estimated by mixing [ 14 C]pantothenate and furosemide (100 M), an effective inhibitor of pantothenate uptake by P. falciparum-infected erythrocytes (10), with infected erythrocytes immediately before centrifuging aliquots through dibutyl phthalate into perchloric acid. [ 14 C]Pantothenate distribution ratios (i.e. the concentration of [ 14 C]pantothenate and [ 14 C]pantothenate-derived metabolites inside the cell relative to the concentration of [ 14 C]pantothenate in the extracellular solution) were calculated using an intracellular water volume of 75 fl for both uninfected and P. falciparum-infected erythrocytes (10). The data obtained with P. falciparum-infected erythrocytes (Ͼ95% parasitemia) have been corrected to 100% parasitemia.
Metabolic Radiolabeling and Extraction of Radiolabeled Metabolites-[ 14 C]Pantothenate (0.1 Ci/ml; 1.8 M) was added to uninfected and P. falciparum-infected erythrocytes suspended in pantothenate-free medium, pH 7.4, and to isolated P. falciparum parasites suspended in pantothenate-free medium, pH 7.1. The reactions were incubated as described for the uptake assay. The final concentrations of cells in uninfected erythrocyte suspensions was 4.4 -5.1 ϫ 10 8 cells/ml, and in the P. falciparum-infected erythrocyte and isolated parasite suspensions were 2.7-3.3 ϫ 10 7 cells/ml. Following 3, 72, or 96 h of labeling, aliquots of the cell suspensions were removed (typically 0.4 ml from the uninfected erythrocyte reactions and 9.5 ml from the P. falciparum-infected erythrocyte and isolated parasite reactions) and the cells were harvested by centrifugation at 2,000 ϫ g for 8 min. The cell pellets were washed three times with 1 ml of ice-cold pantothenate-free medium (pH 7.1 or 7.4, as appropriate), to remove extracellular [ 14 C]pantothenate, then resuspended in 200 l of 50 mM Tris-HCl, pH 7.4 (4°C), and triturated (10 times) through a 25-gauge needle to lyse the cells. The radiolabeled lysates were heated at 95°C for 10 min, and the precipitated protein and cellular debris removed by centrifugation (two times at 15,800 ϫ g for 10 min each, at 4°C). To simplify analysis, dithiothreitol (10 mM) was added to all extracts to prevent formation of dimeric species and to convert disulfides and thioesters of CoA to free CoA (16,17). The extracts were stored at Ϫ80°C until high performance liquid chromatography (HPLC) analysis.
The radiolabeled metabolites detected in extracts prepared as described above were compared with those detected in extracts prepared essentially according to a published methanol/chloroform/water extraction procedure (18). Following extraction, the upper methanol/water phase was collected and dried, then re-dissolved in 200 l of 50 mM Tris-HCl, pH 7.4, at 4°C, and 10 mM dithiothreitol was added. All radiolabeled metabolites detected were present in the extracts prepared by both extraction procedures.
The cell extracts were mixed with unlabeled internal standards (2.5 mM pantothenate, 2.5 mM 4Ј-phosphopantothenoylcysteine, 2.5 mM 4Ј-phosphopantetheine, 0.25 mM dephospho-CoA, and 0.25 mM CoA) prior to 25 l being injected into the column. Extracts were eluted at a flow rate of 1 ml/min. The eluate was monitored at 215 nm and using the on-line radiation analyzer. Subsequent to the initial run, cell extracts were also mixed with 4Ј-[ 14 C]phosphopantothenate (0.05-0.09 Ci/ml) and again, 25 l was injected onto the column. Radiolabeled metabolites present in the extracts were identified on the basis of co-elution with the unlabeled and radiolabeled standards. By spiking the extracts with standard compounds, identification was not compromised by the slight variations in retention times between runs. 4Ј-Phosphopantothenate was synthesized using purified recombinant Escherichia coli pantothenate kinase (PanK). 4Ј-Phosphopantothenoylcysteine and 4Јphosphopantetheine were a gift from Erick Strauss (Stellenbosch University, South Africa), and were synthesized as described previously (21,22). Uninfected erythrocyte extracts were also collected in 1-ml fractions for more sensitive counting in a Beckman LS 6500 Multipurpose Scintillation Counter.
The detection of peaks in the traces recorded by the online radiation analyzer was based on a pre-determined minimum peak height, width, and area. The 1-ml fractions with significant radioactivity were those in which the measured radioactivity was greater than 3 S.D. above the mean background radioactivity. Following subtraction of an average background value, the amount of each metabolite detected in the extracts (in micromoles) was estimated from the integration of the relevant radioactive peaks relative to the integration of radioactive peaks generated following injection of known amounts of [ 14 C]pantothenate or 4Ј-[ 14 C]phosphopantothenate. The data obtained with P. falciparum-infected erythrocytes (Ͼ95% parasitemia) have been corrected to 100% parasitemia. Adjustments have been made to all traces shown to correct for the delay between the UV detec-tor and the radiation detector, and the delay between the UV detector and arrival of the sample at the fraction collector.
Synthesis of 4Ј-Phosphopantothenate-4Ј-Phosphopantothenate was synthesized in both radiolabeled and unlabeled forms. [ 14 C]Pantothenate (1 Ci/ml; 18.2 M) or unlabeled pantothenate (5 mM) was incubated with E. coli PanK (140 g/ml), ATP (5.5 mM), and MgCl 2 (5.5 mM) in 50 mM potassium phosphate buffer, pH 8.0, for up to 3 h at 37°C. The reaction was terminated by incubation at 95°C for 10 min, and the precipitated protein was pelleted by centrifugation at 15,800 ϫ g for 10 min. The yield of the reactions was determined, using Somogyi reagent (15), to be Ͼ91%. HPLC analysis of the unlabeled 4Ј-phosphopantothenate reaction revealed, aside from peaks corresponding to ATP and ADP, a single peak. The product of the unlabeled reaction was confirmed as 4Ј-phosphopantothenate by 1 1H). A single radioactive peak that co-eluted with the unlabeled 4Ј-phosphopantothenate was detected in the radiolabeled reaction.
Statistical Analysis-To test for statistical significance, twotailed Student's t tests were performed (paired or unpaired as appropriate).

Uptake of Pantothenate by Uninfected and P. falciparum-infected
Human Erythrocytes-Previously it was observed that uptake of [ 14 C]pantothenate by uninfected human erythrocytes was negligible over a period of 20 min (10). To determine whether uninfected human erythrocytes are impermeable to pantothenate or instead take up the vitamin very slowly, [ 14 C]pantothenate uptake was measured over a 78-h time course. [ 14 C]Pantothenate was taken up by uninfected human erythrocytes ( Fig. 2A, black circles). The uptake was, however, very slow; on average, 44 h was required for the radiolabel to equilibrate between the intracellular and extracellular environment (i.e. reach a distribution ratio (CPM in / CPM out ) of 1), and after 78 h, pantothenate had only accumulated within the erythrocytes to a distribution ratio of 1.2 Ϯ 0.2 (mean Ϯ S.E.; n ϭ 4). Under similar conditions the uptake of [ 14 C]pantothenate by human erythrocytes infected with mature (trophozoite stage) P. falciparum parasites was rapid (Fig. 2B, black circles). At the earliest time point (30 min) the distribution ratio was already above 1, and after 4.  following a 78-h incubation (0.17 Ϯ 0.02 mol/10 12 cells; mean Ϯ S.E.; n ϭ 4).
Phosphorylation of Pantothenate by Uninfected and P. falciparum-infected Human Erythrocytes-In both uninfected and P. falciparum-infected erythrocytes [ 14 C]pantothenate was accumulated to a distribution ratio greater than 1, consistent with the metabolism of [ 14 C]pantothenate into the phosphorylated derivative(s) that become trapped within the cell. Using Somogyi reagent (15), which precipitates phosphorylated compounds from solution, it was determined that at each time point the majority of the [ 14 C]pantothenate taken up into uninfected human erythrocytes is nonphosphorylated (Fig. 2A). A small proportion of the intracellular radiolabel (up to 20%) was, however, precipitated with Somogyi reagent (15), consistent with some pantothenate being metabolized to a phosphorylated form. By contrast almost all of the [ 14 C]pantothenate accumulated within P. falciparum-infected erythrocytes was precipitable and hence phosphorylated (Fig. 2B).
Metabolism of Pantothenate by Uninfected Human Erythrocytes-In many cells, pantothenate is converted to CoA via the common pathway shown in Fig. 1. CoA and all of the intermediates in the CoA biosynthesis pathway are phosphorylated and hence precipitable with Somogyi reagent (15). To investigate the extent to which uninfected human erythrocytes metabolize pantothenate, uninfected human erythrocytes were incubated with [ 14 C]pantothenate for 72 or 96 h before extracts of the radiolabeled cells were prepared and analyzed for [ 14 C]pantothenate-derived metabolites using HPLC. HPLC was performed using a methanol gradient that separates pantothenate, CoA, and intermediates in the CoA biosynthesis pathway (Fig. 3A), and an on-line continuous flow scintillation analyzer to monitor the elution of radioactive compounds.
The predominant radiolabeled metabolite detected in uninfected human erythrocytes labeled with [ 14 C]pantothenate for 72 or 96 h was identified as unmetabolized [ 14 C]pantothenate, based on co-elution with unlabeled pantothenate (Fig. 3B). In two of four experiments an additional peak, close to the detection limit that co-eluted with unlabeled CoA, was observed (Fig.  3B). This prompted us to increase the sensitivity of the system by measuring radioactivity in 1-ml fractions of the eluate. In all extracts prepared from uninfected erythrocytes labeled for 72 (n ϭ 3) and 96 h (n ϭ 3), radioactivity was detected in the fraction coinciding with the elution of unlabeled pantothenate and in the fraction coinciding with elution of unlabeled CoA (Fig. 3B, inset) Fig. 2B), after an equivalent incubation with [ 14 C]pantothenate. Although the erythrocytes were extensively washed to deplete them of contaminating leukocytes, the possibility exists that the [ 14 C]CoA generated in these experiments was synthesized by leukocytes contaminating the erythrocyte samples. To investigate this possibility, we determined the number of leukocytes that were typically present in these experiments. Manual counting of contaminating leukocytes stained with Turk solution revealed that there were only 19 Ϯ 11 (mean Ϯ S.E.; n ϭ 4; range 2-52) leukocytes present in equivalent samples that gave rise to the HPLC traces shown in Fig. 3B. Assuming an average leukocyte water space of 160 fl (based on an average volume of 210 m 3 for lymphocytes and 630 m 3 for granulocytes (neutrophils, eosinophils, and basophils) and monocytes (23), a cell volume to water space conversion factor of 3.13 (24), and a lymphocyte:granulocyte ϩ monocyte ratio of 3:7 (25)) we estimate that if all of the [ 14 C]CoA generated in the experiments with uninfected erythrocytes was synthesized by contaminating leukocytes, the concentration of [ 14 C]CoA in the leukocytes would be 0.7 Ϯ 0.4 M (mean Ϯ S.E.; n ϭ 4; range 0.07-1.95 M). This concentration, however, is 3 orders of magnitude higher than the concentration of total CoA in liver and heart cells (240 -630 M; calculated as the total tissue CoA content per total intracellular water volume (26 -28)), animal cells with the highest reported total CoA content (28,29). It therefore is unlikely that leukocytes contributed significantly to the [ 14 C]CoA generated in the erythrocyte preparations. Hence, our data are consistent with human erythrocytes having the capacity to metabolize pantothenate to CoA.
Metabolism of Pantothenate by P. falciparum-infected Human Erythrocytes-Pantothenate metabolism by P. falciparum-infected erythrocytes was investigated in the same manner as described for uninfected erythrocytes, except that [ 14 C]pantothenate labeling was performed for just 3 h, which was possible because of the relatively rapid uptake of pantothenate by P. falciparum-infected erythrocytes (Fig. 2). Two distinct [ 14 C]pantothenate-derived metabolites, in addition to some unmetabolized [ 14 C]pantothenate, were detected in P. falciparum-infected erythrocytes (Fig. 3C). The predominant [ 14 C]pantothenate-derived metabolite eluted with 4Ј-phosphopantothenate and 4Ј-phosphopantothenoylcysteine. Once the HPLC conditions were modified to resolve 4Ј-phosphopantothenate and 4Ј-phosphopantothenoylcysteine (Fig. 4A) In three of four experiments, a small peak with a retention time of 9.8 min was observed in HPLC traces of extracts prepared from P. falciparum-infected erythrocytes (Fig. 3C). This peak, which represents Ͻ1.3% of the radioactivity in the P. falciparum-infected erythrocyte extracts, did not co-elute with any of the metabolites shown in Fig. 3A, nor with pantetheine/ pantethine (data not shown), and hence was not identified.

Metabolism of Pantothenate by Isolated P. falciparum
Parasites-Data presented thus far are consistent with CoA biosynthesis occurring within P. falciparum-infected erythrocytes at a rate significantly higher than in uninfected erythrocytes. Whether the site of CoA biosynthesis in the P. falciparum-infected erythrocyte is the host erythrocyte cytosol and/or the parasite itself, however, remained to be determined.  . An average background has been subtracted from all radioactive traces, and identified peaks have been shaded. For improved sensitivity, the uninfected erythrocyte extracts were collected from the HPLC in 1-ml fractions and counted in a scintillation counter. The radioactivity measured in the 1-ml fractions is shown in the inset of B. An average background radioactivity has been subtracted from all fractions. All cell extracts were co-injected with unlabeled pantothenate, 4Ј-phosphopantothenoylcysteine, 4Ј-phosphopantetheine, dephospho-CoA, and CoA, and subsequently also with 4Ј-[ 14 C]phosphopantothenate. Elution of the unlabeled standards was detected simultaneously by absorbance at 215 nm. Peaks were identified on the basis of co-elution with the unlabeled and radiolabeled standards shown in A. By spiking the extracts with standards, identification was not affected by slight variations in retention times between runs. Adjustments have been made to all traces shown to correct for the delay between the UV detector and radiation detector, and the delay between the UV detector and arrival of the sample at the fraction collector. The data shown in B are from a single experiment, representative of those obtained in three independent experiments each on uninfected erythrocytes from different donors. The data shown in C and D are from single experiments, representative of those obtained in four and three independent experiments, respectively.
A small peak with a retention time of 9.8 min was, in two of three experiments, observed in HPLC traces of extracts prepared from isolated parasites, as it was in extracts prepared from infected erythrocytes. This unidentified radiolabeled compound corresponded to Ͻ1.7% of the radioactivity detected in isolated parasites.

DISCUSSION
Uptake of Pantothenate into Uninfected and P. falciparuminfected Erythrocytes-In this study it was shown that pantothenate is permeant to normal human erythrocytes. The uptake, however, is very slow; consistent with published data (10) The slow uptake contrasts with the rapid uptake of pantothenate by P. falciparum-infected erythrocytes, which occurs via the new permeation pathways induced in the membrane of the host erythrocyte by the parasite (30 -34). Contrary to other mammalian cells, which take up pantothenate via a transporter-mediated process (35)(36)(37), rat erythrocytes take up pantothenate by passive diffusion (9). The mechanism by which uninfected human erythrocytes take up pantothenate remains to be established.
The fact that most of the [ 14 C]pantothenate taken up by uninfected erythrocytes was not metabolized ( Fig. 2A) is in contrast with the situation observed in isolated P. falciparum parasites (10) and intact P. falciparum-infected erythrocytes (Fig.  2B). It has previously been argued that the parasite quickly phosphorylates pantothenate once it is taken up to "trap" it within its cytosol, making it available for conversion into CoA  (2) is shown in the overlaying traces (broken gray line). An average background radioactivity has been subtracted from all traces recorded by the on-line continuous flow scintillation analyzer. Adjustments have been made to all traces shown to correct for the delay between the UV and radiation detectors. (38). From the data presented in Fig. 2A, it appears that such a strategy is not utilized by uninfected erythrocytes. This may be due to the fact that pantothenate transport across the uninfected erythrocyte membrane is very slow (especially when compared with that across the parasite plasma membrane) and should the extracellular pantothenate concentration become limiting, it would take days for an uninfected erythrocyte to become depleted of pantothenate.
Metabolism of Pantothenate by Normal Human Erythrocytes-Having established that normal human erythrocytes take up pantothenate, albeit very slowly, we went on to provide data consistent with human erythrocytes metabolizing pantothenate to CoA. By contrast, Annous and Song (9)  CoA at all could be detected in rat erythrocytes, only pantothenate, 4Ј-phosphopantothenate, and pantetheine, as determined by quantitation of pantothenate release following treatment of erythrocyte preparations with alkaline phosphatase, pyrophosphatase, and pantetheinase, alone or in combination. They thereby concluded that rat erythrocytes possess only the first of the enzymes required for metabolizing pantothenate to CoA (PanK). The discrepancy between the findings of Annous and Song (9) and the data presented in this study may be due to inherent dissimilarities between rat and human erythrocytes or experimental differences between the two studies. Annous and Song (9) discuss in their article the potential source of a significant pool of pantetheine detected in rat erythrocytes; pantetheine could not be resolved from pantothenate by the method employed for analysis of the radiolabeled metabolites in the [ 14 C]pantothenate-labeled erythrocytes and, as a result, it is unclear whether pantothenate taken up by the erythrocytes is converted to pantetheine (only known to occur indirectly as a result of 4Ј-phosphopantetheine degradation), and/or exogenous pantetheine is taken up. If, however, pantetheine is generated from pantothenate by erythrocytes, it would be indicative of dephospho-CoA and/or CoA (both of which possess the 4Ј-phosphopantetheine moiety) and/or 4Ј-phosphopantetheine itself, and hence possibly additional steps in the CoA biosynthesis pathway being present in rat erythrocytes.
We detected only [ 14 C]pantothenate and [ 14 C]CoA (and none of the intermediates in the CoA biosynthesis pathway) in human erythrocytes. This is consistent with PanK catalyzing the rate-limiting step in CoA biosynthesis in human erythrocytes.
Metabolism of Pantothenate in the P. falciparum-infected Erythrocyte-Herein, we have shown that CoA biosynthesis is significantly higher in erythrocytes infected with P. falciparum, when compared with normal erythrocytes. This follows published work demonstrating the increased permeability of infected erythrocytes to pantothenate, relative to uninfected erythrocytes, and the increased rate at which pantothenate is phosphorylated by lysates prepared from infected erythrocytes relative to uninfected erythrocyte lysates (10).
In this study we have demonstrated that P. falciparum parasites isolated from their host erythrocyte are not only capable of phosphorylating pantothenate (10), they are fully capable of metabolizing pantothenate to CoA. This capability clearly distinguishes P. falciparum from its avian counterpart, P. lophurae, which must rely on the host erythrocyte for CoA because of its inability to synthesize the cofactor, atleast during the intraerythrocyte stage. We observed that the amounts of 4Ј-[ 14 C]phosphopantothenate and [ 14 C]CoA measured in isolated parasites were significantly lower than the corresponding amounts in P. falciparum-infected erythrocytes. This difference could reflect a significant contribution by the host erythrocyte to CoA biosynthesis in the P. falciparum-infected erythrocyte, however, this seems unlikely because of the similar ratio of 4Ј-[ 14  C]pantothenate in isolated parasites following labeling was therefore calculated using the estimated detection limit, and as such represents only an upper limit. Black asterisks indicate a significant difference (*, p Ͻ 0.05 or **, p Ͻ 0.01) between the amounts of a metabolite detected in uninfected erythrocytes following labeling with [ 14 C]pantothenate for 96 h, P. falciparum-infected erythrocytes following labeling for 3 h, or isolated P. falciparum parasites following labeling for 3 h. A gray asterisk indicates a significant difference (p ϭ 0.03) between the amounts of [ 14 C]CoA detected in uninfected erythrocytes following labeling for 72 or 96 h. lower in P. falciparum-infected erythrocytes when compared with isolated parasites because 4Ј-[ 14 C]phosphopantothenate does not accumulate in uninfected erythrocytes as it does in the parasite (Fig. 5). A more likely explanation is that although isolated parasites have been shown to perform essential biological processes in a manner that is indistinguishable from parasites within intact erythrocytes (10, 11, 38 -40), at least for several minutes following their isolation, they may not be able to do so for the entire 3-h labeling period used in this study. It therefore seems likely that the parasite is the source of most of the CoA synthesized in the P. falciparum-infected erythrocyte, and that P. falciparum, unlike P. lophurae parasites, does not rely on host erythrocyte-generated CoA, perhaps because the limited CoA synthesized by human erythrocytes (Fig. 5) is insufficient to meet the requirements of the parasite. Our data raise the possibility that malaria parasites rationalize their biosynthetic activity on the basis of the capacity of their host cell to synthesize the metabolites they require. Although P. lophurae has been shown to utilize exogenous CoA and therefore must possess a mechanism for CoA uptake, it is currently unknown whether P. falciparum is able to take up CoA or any of the other charged intermediates in the pathway, and hence whether or not any host cell-generated CoA could be utilized by the parasite.
The accumulation of 4Ј-phosphopantothenate in P. falciparum-infected erythrocytes and isolated parasites, but not in uninfected erythrocytes (Fig. 5), suggests that CoA biosynthesis is regulated differently by P. falciparum and erythrocytes, and that the rate of pantothenate phosphorylation does not determine the rate of CoA production by the parasite as it does in other organisms (16,41). PanK activity in P. falciparum lysates is, however, inhibited by CoA (IC 50 ϳ 200 M (42)) and hence is not refractory to feedback inhibition (as shown for Staphyloccoccus aureus PanK (43)). Our data are consistent with the condensation of 4Ј-phosphopantothenate and cysteine to form 4Ј-phosphopantothenoylcysteine, being rate-limiting in P. falciparum CoA production. In isolated perfused rat hearts in which PanK has been stimulated (by omission of exogenous energy substrates such as glucose and palmitate from the perfusate), an insufficient supply of cysteine results in an accumulation of 4Ј-phosphopantothenate, and limits the rate of CoA synthesis (44,45). Whether the accumulation of 4Ј-phosphopantothenate in P. falciparum is a result of regulation of the enzyme catalyzing the condensation of cysteine and 4Ј-phosphopantothenate or limited availability of the relevant substrates (e.g. cysteine), remains to be determined. Irrespective of the mechanism by which high levels of 4Ј-phosphopantothenate are maintained, the physiological importance (if any) of the 4Ј-phosphopantothenate sequestration for erythrocytic stage P. falciparum parasites requires further investigation. One possibility is that the accumulated 4Ј-phosphopantothenate is rationed to daughter merozoites, which, upon infecting erythrocytes and becoming ring stage parasites, would otherwise (until the formation of the parasite-induced new permeation pathways in the erythrocyte membrane, some 15 h postinvasion (46)) be essentially limited to the pantothenate present in the erythrocyte at the time of invasion.
Conclusions-Our data are consistent with pantothenate being taken up and metabolized to CoA by both normal human erythrocytes and human erythrocytes infected with P. falciparum. Both the uptake of pantothenate and its metabolism to CoA are significantly higher in the infected erythrocyte. P. falciparum can synthesize CoA independently of the host erythrocyte, and synthesizes most of the CoA generated in the P. falciparum-infected erythrocyte. This capability distinguishes P. falciparum from the avian malaria parasite P. lophurae.