Adaptive responses to purine starvation in Leishmania donovani

Summary

Starvation of Leishmania donovani parasites for purines leads to a rapid amplification in purine nucleobase and nucleoside transport. Studies with nucleoside transport-deficient L. donovani indicate that this phenomenon is mediated by the nucleoside transporters LdNT1 and LdNT2, as well as by the purine nucleobase transporter LdNT3. The escalation in nucleoside transport cannot be ascribed to an increase in either LdNT1 or LdNT2 mRNA. However, Western analyses on parasites expressing epitope-tagged LdNT2 revealed a marked upregulation in transporter protein at the cell surface. Kinetic investigations of LdNT1 and LdNT2 activities from purine-replete and purine-starved cells indicated that both transporters exhibited significant increases in V_max for their ligands under conditions of purine-depletion, although neither transporter displayed an altered affinity for its respective ligands. Concomitant with the increase in purine nucleoside and nucleobase transport, the purine salvage enzymes HGPRT, XPRT and APRT were also upregulated, suggesting that under conditions where purines are limiting, Leishmania parasites remodel their purine metabolic pathway to maximize salvage. Moreover, qRT-PCR analyses coupled with cycloheximide inhibition studies suggest that the underlying molecular mechanism for this augmentation in purine salvage occurs post-transcriptionally and is reliant on de novo protein synthesis.

Introduction

In parasitic protozoa, as well as in higher eukaryotes, equilibrative nucleoside transporters (ENTs) have emerged as the predominant route for nucleoside translocation across cell membranes (Baldwin et al., 2004, Carter et al., 2001). As well as being key metabolic substrates, nucleosides are also important regulators of many physiological processes (Baldwin et al., 2004). While the functional characterization of nucleoside transport has been extensively undertaken both at the molecular and cellular level, the regulatory components governing nucleoside transport activity and expression remain to be uncovered.

The physiological regulation of nucleoside transport may be particularly important in parasitic protozoa. Parasitic protozoa are unable to synthesize the purine ring de novo and consequently, are obligatory scavengers of purines from their host (Carter, 2003, Carter et al., 2008). Due to their intricate and varied milieu parasitic protozoa likely encounter and must surmount considerable fluctuations in extracellular purines, necessitating adaptations in purine metabolism and transport for maintenance of their intracellular purine homeostasis. Indeed, depletion of extracellular purines triggers a substantial upregulation in nucleoside uptake in the parasitic protozoa, Crithidia spp. and Trypanosoma brucei (de Koning et al., 2000, Alleman & Gottlieb, 1996, Hall et al., 1996, Liu et al., 2005).

While these previous studies suggest that nucleoside transport is intimately linked to the extracellular purine milieu, little is known about the molecular mechanisms governing these changes. In Crithidia spp., enhanced nucleoside transport appears to be regulated at a post-transcriptional level (Hall et al., 1996, Liu et al., 2005), but whether these changes are mediated by an upregulation of nucleoside transporter protein at the cell surface, by a modification of its activity, or by some other as yet undisclosed mechanism, is not known. In mammalian cells, human equilibrative nucleoside transporter 1 (hENT1) activity can be modulated by protein kinase C (δ and ε subtypes) (PKCδ/ε). Pharmacological stimulation of PKCδ/ε leads to an increase in hENT1 activity, while inhibition of PKCδ/ε activity leads to diminished hENT1 transport (Coe et al., 2002). Likewise inhibition of PKCε by either hypoxia or by phorbol esters in the murine cardiomyocyte cell line, HL-1, diminishes mENT1 activity, the murine counterpart to hENT1, and this is accompanied by a decrease in mENT1 RNA levels (Chaudary et al., 2004).

Nucleoside transport in Leishmania donovani, which is mediated by two high-affinity, discriminating ENTs, Leishmania donovani nucleoside transporters 1 and 2 (LdNT1 and LdNT2) (Carter et al., 2000, Vasudevan et al., 1998), is also influenced by the extracellular purine concentration (Seyfang & Landfear, 1999). LdNT1 activity, which mediates the uptake of adenosine and pyrimidine nucleosides, is repressed when increasing exogenous concentrations of adenosine are added to the culture medium, although LdNT2 activity, which mediates inosine and guanosine transport, appears unchanged (Seyfang & Landfear, 1999). The lack of complexity in L. donovani for nucleoside transport (Carter et al., 2000, Liu et al., 2005) (other trypanosomatids possess a multiplicity of nucleoside transporters with overlapping substrate-specificities) (Maser et al., 1999, Sanchez et al., 2002, Sanchez et al., 1999, Vasudevan et al., 1998), coupled with the availability of Δldnt1 and Δldnt2 nucleoside transport-deficient mutants (Liu et al., 2006) provides an excellent background in which to dissect the regulation of nucleoside transport within an intact biological system.

As a first step towards this goal, we have comprehensively characterized the effects of purine withdrawal on L. donovani parasites. Here we report that there is a dramatic and rapid escalation in purine transport in purine-starved L. donovani resulting from an upregulation of the nucleoside transporters, LdNT1 and LdNT2, as well as from the purine nucleobase transporter, Leishmania donovani nucleobase transporter 3 (LdNT3), which is orthologous to LmaNT3 from L. major (Sanchez et al., 2004). Studies with episomally expressed epitope-tagged LdNT2 suggest that this escalated activity arises from a rapid increase in transporter protein at the cell surface of the parasite. These changes are also accompanied by corresponding amplifications in the activity and protein levels of the three major purine salvage enzymes, hypoxanthine-guanine phosphoribosyltransferase (HGPRT), adenine phosphoribosyltransferase (APRT), and xanthine phosphoribosyltransferase (XPRT). Preliminary analyses to uncover the molecular mechanism governing these changes suggests that this upregulation occurs at a post-transcriptional level and likely involves the enhanced translation of the nascent polypeptides encoding for various transport and metabolic components of the purine salvage pathway.

Results

Purine starvation significantly enhances purine nucleobase and nucleoside transport in L. donovani wild type cells

Initially the effects of purine withdrawal on purine transport in wild type L. donovani were monitored over a 72 h period. The uptake of 10 μM [³H]hypoxanthine, [³H]inosine, and [³H]adenosine were all significantly enhanced in purine-starved cells and the degree of magnification correlated with the length of time that purines had been withdrawn from the media (Fig. 1A - C). After 72 h of purine starvation [³H]hypoxanthine, [³H]adenosine and [³H]inosine transport was amplified by 82, 45, and 46-fold, respectively (Fig. 1A - C, closed bars). By contrast, no significant changes in nucleoside transport were evident in cells that were incubated in purine-replete media for the entire 72 h time course (Fig. 1A - C, open bars). The adaptive response in purine transport upon purine withdrawal was rapid and evident by the first time point of 6 h, although this increase was modest and not of the magnitude observed after prolonged purine removal (Fig. 1A – C, closed bars). Note that in all subsequent experiments cells were purine-starved for 72 h, unless otherwise specified.

Fig. 1.

Timecourse of the effects of purine withdrawal on purine nucleoside and nucleobase transport in wild type L. donovani.

L. donovani DI700 promastigotes were cultivated in RPMI-L with 100 μM xanthine or without purine for 0 - 72 h. Rates of uptake for 10 μM [³H]hypoxanthine, [³H]inosine, and [³H]adenosine were measured over 60 s as described in Experimental procedures at 0, 6, 12, 24, 48, and 72 h for both purine-replete (open bars) and purine-starved cells (closed bars).

Effect of purine withdrawal on cell morphology and growth of L. donovani promastigotes

In addition to the escalations in purine transport observed, removal of purines from the culture medium also led to changes in the morphology of purine-starved L. donovani promastigotes, producing a significant elongation of the cell body (Fig. 2A). The average length of the cell body in wild type cells (from anterior-most tip to posterior end) was 7.16 ± 0.25 μm (n=63) under purine-replete conditions versus 11.48 ± 0.51 μm (n=41) for cells cultured for 48 h under purine-depleted conditions) (Fig. 2A). These effects were evident as early as 24 h post purine withdrawal. Moreover, L. donovani promastigotes cultured under purine-depleted conditions became growth arrested after approximately 24 h (Fig. 2B), whereas L. donovani promastigotes cultured under exactly the same conditions but with the addition of 50 μM adenine and 50 μM xanthine reached ∼15 times the cell number and had entered stationary phase after ∼4 days (Fig. 2B). In addition to the effects on cell morphology and growth, the effect of purine starvation on cell cycle progression was also analyzed over a 96 h time-period by FACS (Fig. 2C and Fig. S1). Whereas ∼66-70 % of the cells analyzed at various time-points from the purine-replete cultures were in G₁ phase, under purine-depleted conditions the number of cells in G₁ phase had accumulated to ∼86 % by 72 h. The number of cells in S phase and G2/M were also commensurately low in the purine-starved culture, suggesting that the majority of cells were in a quiescent, non-dividing state.

Fig. 2.

Cell morphology and growth of purine-replete and purine-depleted L. donovani promastigotes.

(A) LdBob promastigotes were grown in the presence or absence of purine for 48 h. Cell morphology was visualized by differential interference microscopy (DIC) and cell body length evaluated according to the Experimental procedures.

(B) The growth of LdBob promastigotes in DME-L plus or minus purine was evaluated as detailed in Experimental procedures. Cell numbers for purine-replete (open circles) and purine-starved (closed circles) parasites are shown over the 144 h time-period.

(C) FACS analyses on purine-replete (open bars) and purine-starved (closed bars) LdBob promastigote cultures were performed as described in Experimental procedures. Results are shown as percentage of cells in G1 phase divided by the combined percentages of cells in S and G2 phase (G1/S + G2) versus time (h).

Effect of purine removal on other transport processes

To ensure that the rise in purine influx that occurred upon purine removal was not symptomatic of nutritionally stressed cells with compromised membrane integrity, the transport of two other radiolabeled physiological nutrients, [³H]glucose and [³H]arginine, was measured under both purine-replete and purine-depleted conditions (Fig. 3A). In comparison to [³H]inosine transport, [³H]glucose and [³H]arginine transport were not increased in purine-starved cells, and conversely were either somewhat decreased, as was observed for [³H]glucose transport (1.5 fold reduced), or substantially reduced as for [³H]arginine transport (5.8 fold reduced). Moreover, the upregulation of purine transport did not appear to be a general adaptation to metabolic stress since glucose depletion from the culture medium for 12 h, conditions which were previously determined to not affect cell viability (Rodriguez-Contreras & Landfear, 2006), while provoking an upregulation in [³H]glucose transport (∼2.3-fold increased) led to diminished [³H]adenosine transport (∼4.5-fold decreased) (Fig. 3B inset).

Fig. 3.

Effect of purine and glucose withdrawal on inosine, arginine, glucose and adenosine transport in wild type L. donovani.

(A) L. donovani DI700 promastigotes were cultivated in RPMI-L with 100 μM xanthine or without purine 72 h. Rates of uptake for 10 μM [³H]inosine, [³H]glucose, and [¹⁴C]L-arginine were measured over 60 s as described in Experimental procedures for both purine-replete (open bars) and purine-starved cells (closed bars).

(B) L. donovani DI700 promastigotes were cultivated in RPMI-L replete with 100 μM xanthine in the presence or absence of glucose for 12 h. Rates of uptake for 10 μM [³H]adenosine (see inset) and [³H]glucose were measured over 60 s as described in Experimental procedures for cells cultured under both glucose-replete (open bars) and glucose-depleted conditions (closed bars).

Amplification of nucleoside transport in purine-starved cells is dependent upon LdNT1 and LdNT2

Using the previously characterized nucleoside transport-deficient null mutants Δldnt1 and Δldnt2 (Liu et al., 2006), the contributions of the nucleoside transporters LdNT1 and LdNT2 towards the upregulation of purine nucleoside acquisition upon purine removal were gauged. Under normal cultivation conditions in the presence of 100 μM purine, 1 μM [³H]adenosine and 1 μM [³H]inosine transport were dramatically reduced in Δldnt1 and Δldnt2 cells, respectively, in comparison to wild type cells under purine-replete conditions (Fig. 4A and B, open bars and (Liu et al., 2006)). Although the withdrawal of purines led to dramatic increases in adenosine and inosine transport in wild type cells, no significant upregulation of 1 μM [³H]adenosine and [³H]inosine transport was observed in Δldnt1 or Δldnt2 cells, respectively (Fig. 4A and B, closed bars), suggesting that the alterations observed in purine nucleoside transport accompanying purine withdrawal are largely contingent on the expression of LdNT1 and LdNT2 protein. This result, however, could not be reproduced for the Δldnt2 cell line at 10 μM [³H]inosine, which showed a small but persistent increase in transport under purine-starved conditions (Fig. 5, closed bars), suggesting that under these conditions a portion of inosine uptake occurs through an alternative route.

Fig. 4.

The upregulation of adenosine and inosine transport in purine-starved L. donovani is mediated by the nucleoside transporters LdNT1 and LdNT2.

L. donovani DI700 (wild type), DI700Δldnt1, and DI700Δldnt2 promastigotes were cultivated as described in Experimental procedures in RPMI-L with 100 μM xanthine or without purine for 72 h. Rates of uptake for 1 μM (A) [³H]adenosine and (B) [³H]inosine were measured over 60 s as described in Experimental procedures for both purine-replete (open bars) and purine-starved cells (closed bars).

Fig. 5.

Residual inosine transport in purine-starved L. donovani Δldnt2 cells.

Δldnt2 promastigotes harboring the pXG-GFP2+′ episome were cultivated as described in Experimental procedures in RPMI-L plus or minus 100 μM xanthine for 72 h. Rates of uptake for 10 μM [³H]inosine were measured over 60 s as described in Experimental procedures for both purine-replete (open bars) and purine-starved cells (closed bars).

Kinetic analysis of purine transport in purine-starved cells

Since enhanced purine nucleoside transport in purine-starved cells appeared to be mediated through LdNT1 and LdNT2, the kinetic parameters for LdNT1 and LdNT2-mediated nucleoside transport were measured in cells deprived of purines for 72 h. Large changes were observed in the V_max of transport for all of the purine nucleosides from a 135 and 289-fold change for inosine and adenosine, respectively, to over a 2000-fold change for guanosine (Table 1). Curiously the V_max for another LdNT1 ligand, the pyrimidine nucleoside uridine, was only modestly enhanced (7-fold change) (Table 1). Although the reason for these differences with LdNT1 ligands is unclear, it likely reflects differences in the downstream cellular metabolism of these ligands, which may be escalated in the case of adenosine metabolism (a purine) and unaltered with respect to uridine metabolism (a pyrimidine) under conditions of purine starvation. Accompanying these changes in V_max were remarkable alterations in the apparent K_m for the LdNT2 ligands, inosine and guanosine (Table 1). The apparent K_m for inosine was reduced from 1.4 μM to 250 μM (a 179-fold change), while the apparent K_m for guanosine was so reduced (0.6 versus 800 μM) that transport was not saturated at up to 1 mM of ligand. Conversely, the apparent K_m of LdNT1 for its ligands was unaltered under purine-starved conditions (Table 1). This modification in LdNT2 activity was also accompanied by apparent changes in ligand specificity. Under purine-replete conditions LdNT2 is a highly discriminating transporter, exhibiting a narrow ligand repertoire for inosine and guanosine (Fig. 6B, open bars and (Carter et al., 2000)). However, when cells were cultivated in the absence of purines for 72 h, the ligand specificity of LdNT2 was apparently modified to accommodate the purine nucleobases, hypoxanthine, guanine and adenine (Fig. 6B, closed bars). In contrast, LdNT1 ligand specificity was the same under both purine-replete and purine-starved conditions (Fig. 6A, open and closed bars). Notably neither transporter gained the capacity to accommodate other purine nucleosides.

Table 1.

Kinetic parameters for purine-replete and purine-starved wild type L. donovani.

Ligand	Km (μM)		Vmax (pmol/108 cells/s)
	Replete	Starved	Replete	Starved
Adenosine	1.2	0.8	0.3	86.8
Uridine	6.3	8.1	2.8	18.7
Inosine	1.4	250	1.3	175
Guanosine	0.6	800	0.3	711

Kinetic parameters were calculated as detailed in Experimental procedures.

Fig. 6.

Ligand specificities of LdNT1 and LdNT2 in purine-replete and purine-starved wild type L. donovani.

Wild type DI700 promastigotes were cultivated in RPMI-L plus or minus 100 μM xanthine for 72 h. The ligand specificity of (A) LdNT1 or (B) LdNT2 under purine-replete (open bars) or purine-depleted (closed bars) conditions was determined by measuring the uptake of 1 μM [³H]adenosine (A) or 1 μM [³H]inosine (B) over 60 s in the presence or absence of 100 μM of each competitor. Results are plotted as percent uptake of either the 1 μM [³H]adenosine (A) or 1 μM [³H]inosine (B) competed rate compared to the non-competed rate.

Although these alterations in LdNT2 activity might indicate that the transporter is modified under purine-starved conditions, the significant inosine uptake observed at 10 μM [³H]inosine in the Δldnt2 cell line (Fig. 5) suggests that at least one other route of uptake for [³H]inosine and/or its downstream metabolite exists under conditions of purine withdrawal. In addition to the nucleoside transporters LdNT1 and LdNT2, L. donovani, as well as other Leishmania species, harbor 2 nucleobase transporters, LdNT3, a homolog of LmaNT3 in Leishmania major, which transports all purine nucleobases (Sanchez et al., 2004), and LdNT4, a homolog of LmaNT4, which has a specificity for adenine at pH 7.4 (Ortiz et al., 2009). Given that the amplified inosine activity in wild type cells is inhibited by the LdNT3 ligands, hypoxanthine, guanine and adenine (Fig. 6B, closed bars), we hypothesized that the altered inosine and guanosine transport kinetics observed under purine-starved conditions (Table 1) may be due in part to uptake by LdNT3. Thus, transport kinetics were measured in a Δldnt3 cell line (Liu et al. unpublished) to negate any contribution of LdNT3 to the escalation of transport observed under purine depleted conditions. In general, [³H]adenosine, [³H]inosine, and [³H]guanosine transport was more robust in Δldnt3 cells under purine-replete conditions in comparison to wild type cells (compare replete V_max values in Tables 1 and 2). Moreover, as observed for wild type parasites, the V_max for [³H]adenosine, [³H]inosine, and [³H]guanosine transport in Δldnt3 cells was significantly increased after 72 h of purine depletion (Table 2). However, unlike that observed for wild type cells, the increase in [³H]inosine, and [³H]guanosine transport in the Δldnt3 cell line was not accompanied by changes in the apparent K_m (Table 2). Moreover, no significant inhibition of [³H]inosine uptake was observed by the LdNT3 ligands hypoxanthine, guanine or adenine (Fig. 7). Since the ligand specificity for LdNT3 is restricted to purine nucleobases under purine-replete conditions, these data suggest that under conditions of purine starvation a substantial portion of exogenous inosine and guanosine are hydrolyzed to their respective nucleobases, hypoxanthine and guanine, prior to being transported by LdNT3.

Table 2.

Kinetic parameters for purine-replete and purine-starved Δldnt3 L. donovani.

Ligand	Km (μM)		Vmax (pmol/108 cells/s)
	Replete	Starved	Replete	Starved
Adenosine	0.5	0.7	32	939
Uridine	17	12	111	880
Inosine	0.4	1	11	105
Guanosine	1.4	1.3	70	560

Kinetic parameters were calculated as detailed in Experimental procedures.

Fig. 7.

Ligand specificity of LdNT2 in purine-replete and purine-starved LdBobΔldnt3 cells.

LdBobΔldnt3 promastigotes were cultivated in DME-L plus or minus purine for 72 h as detailed in Experimental procedures. The ligand specificity of LdNT2 under purine-replete (open bars) or purine-depleted (closed bars) conditions was determined by measuring the uptake of 1 μM [³H]inosine over 60 s in the presence or absence of each competitor (100 μM). Results are plotted as percent uptake of the 1 μM [³H]inosine competed rate compared to the non-competed rate.

Effect of purine starvation on the activity and levels of episomally produced LdNT2 protein

Antisera raised against either LdNT1 or LdNT2 epitopes do not detect endogenous LdNT1 or LdNT2 protein under normal conditions, probably due to the low abundance of these transporters (see accompanying manuscript from Ortiz et al.). However, we have previously employed Δldnt2 cells harboring episomes containing the LdNT2 gene lacking its endogenous 5′ and 3′ UTR fused to the 3′ end of the gene for green fluorescent protein (GFP), Δldnt2[GFP-LdNT2], to correlate activity to protein levels using GFP as a reporter (Arastu-Kapur et al., 2003). Thus, the effects of purine removal on Δldnt2[GFP-LdNT2] cells, as well as on Δldnt2 cells containing an episome bearing a 5 kb EcoRV fragment harboring influenza hemagglutinin (HA)-tagged LdNT2 flanked by its native 5′ and 3′ UTR, Δldnt2[HA-LdNT2-RV] (see Supporting information) were evaluated. As expected, even under purine-replete conditions [³H]inosine transport is significantly upregulated in Δldnt2 cells harboring episomes containing the LdNT2 gene (Fig. 8A open bars) in comparison to wild type parasites without episomes (Figs. 3 and 4, open bars), although transport was much more robust in cells harboring the pSNAR-HA-LdNT2-RV episome. However, both Δldnt2[GFP-LdNT2] and Δldnt2[HA-LdNT2-RV] exhibited modest but significant increases in [³H]inosine transport after 72 h of growth in purine-deficient media (2.4-fold for Δldnt2[HA-LdNT2-RV] cells and 4.3-fold Δldnt2[GFP-LdNT2] cells) (Fig. 8A closed bars), compared to growth under purine-replete conditions. Significantly, the upregulation in LdNT2 activity was not contingent upon the presence of its 5′ or 3′ UTR. These results were also corroborated in purine-replete and purine-starved cells by Western analyses on biotinylated cell surface protein fractions (see Experimental procedures), that showed that episomally generated LdNT2 transporter was increased by ∼20-fold (HA-LdNT2) and ∼10-fold (GFP-LdNT2) under purine-starved conditions (Fig. 8B).

Fig. 8.

Effect of purine withdrawal on LdNT2 protein levels in L. donovani.

Δldnt2[HA-LdNT2-RV] and Δldnt2[GFP-LdNT2] were cultivated in RPMI-L plus or minus purine for 72 h as detailed in Experimental procedures.

(A) Rates of uptake for 1 μM [³H]inosine were measured over 60 s as described in Experimental procedures for both purine-replete (open bars) and purine-starved cells (closed bars).

(B) The relative abundance of HA-LdNT2 and GFP-LdNT2 protein at the cell surface was evaluated in both purine-replete (+) and purine-starved (-) Δldnt2[HA-LdNT2-RV] and Δldnt2[GFP-LdNT2] cells, respectively as described in Experimental procedures. The HA-LdNT2 and GFP-LdNT2 proteins were detected with anti-HA and anti-GFP antibodies, and antibodies to GP63 were used to control for loading. The positions of HA-LdNT2, GFP-LdNT2, and GP63 are indicated on the right. The relative intensities of the HA-LdNT2, GFP-LdNT2, and GP63 bands were estimated by densitometry as described in Experimental procedures.

Effect of purine starvation on the activity of purine salvage enzymes

Comparison of the inhibition profiles for inosine transport between wild type and Δldnt3 parasites (Figs. 6 B and 7), as well as the discrepant fold changes in V_max observed for the LdNT1 ligands, adenosine and uridine, under purine-depleted conditions (Tables 1 and 2), suggest that withdrawal of purines from the medium also leads to the amplification of other metabolic components of the purine salvage pathway. Thus, the activities of two of the main purine salvage enzymes in L. donovani, HGPRT and APRT were quantified (Fig. 9A and B). Cells that were purine-starved for 72 h showed enhanced activities for both HGPRT (Fig. 9A), 0.64 pmol IMP/μg protein/min in purine-replete cells (open circles) versus 5.66 pmol IMP/μg protein/min in purine-starved cells (closed circles), and APRT (Fig. 9B), 0.38 pmol AMP/μg protein/min in purine-replete (open circles) versus 9.3 pmol AMP/μg protein/min in purine-starved cells (closed circles). Moreover, these changes were accompanied by significant changes at the protein level for the purine salvage enzymes, HGPRT, XPRT, and APRT, compared to the α̃tubulin loading control (Fig. 9C), and these changes were evident as early as 24 h after purines were withdrawn from the medium (Fig. 9C). The effect of purine starvation upon the relative levels of other metabolic markers was also examined. No changes were observed in the relative levels of glycosomal glyceraldehyde-3-phosphate dehydrogenase (GAPDH), arginase, spermidine synthase, or orotidine-5-phosphate decarboxylase/orotate phosphoribosyltransferase (OMPDC/OPRT), suggesting that withdrawal of purines from the medium does not universally impact cellular metabolism.

Fig. 9.

Purine starvation leads to an upregulation in HGPRT, APRT, and XPRT levels in L. donovani.

Wild type DI700 were cultivated in RPMI-L plus or minus purine for 72 h as detailed in Experimental procedures. HGPRT and APRT activities under purine-replete (open circles) and purine-depleted (closed circles) conditions were evaluated by following the rate of incorporation of [¹⁴C]hypoxanthine (A) and [¹⁴C]adenine (B) into the nucleotide pool as described in Experimental procedures. Results are displayed as pmol of phosphorylated product formed per μg of cell protein extract.

(C) Wild type LdBob were cultivated in DME-L plus or minus purine for 0, 24, 48, and 72 h. The relative level of HGPRT, XPRT, APRT, α-tubulin, arginase, OMPDC/OPRT, GAPDH, and spermidine synthase (SPDSYN) protein was evaluated in both purine-replete (+) and purine-depleted (-) cells by Western blot. HGPRT, XPRT, APRT, arginase, OMPDC/OPRT, GAPDH, or SPDSYN protein was detected with monospecific polyclonal antiserum as described in Experimental procedures, and a monoclonal antibody to α-tubulin was used to control for loading. The positions of all proteins are indicated on the right.

Quantitative reverse transcription PCR analysis of purine-starved cells

To evaluate whether the escalation in purine salvage could be ascribed to transcriptional changes or enhanced mRNA stability, total RNA was isolated from a population of purine-replete and purine-starved cells, cDNA synthesized, and quantitative reverse transcription PCR (qRT-PCR) used to assess abundance of the transcripts described in Tables 3 and S2. Notably, a significant decrease in total RNA yield (∼3-fold reduced) was consistently observed from cells starved of purine for 48 h, 76 ± 9 μg/10⁸ purine-replete cells (n=3) versus 28 ± 4 μg/10⁸ purine-starved cells (n=3). For the qRT-PCR experiments, the relative abundance of transcript within the total RNA pool was compared between purine-starved and purine-replete parasites after 48 h and the changes recorded before and after normalization to the levels of transcript for OMPDC/OPRT (Table 3) or aldehyde reductase (Table S2) under purine-starved and replete conditions. We have previously determined from preliminary proteomic analyses that the protein levels for OMPDC/OPRT and aldehyde reductase remain unchanged after purines are removed from the medium for up to 72 h (data not shown), and this result has also been verified for OMPDC/OPRT by Western analysis (Fig. 9C). Despite a decrease in total RNA yield from purine-starved cells (see above), the global expression of purine salvage and recycling genes remained relatively constant during 48 h of purine depletion. However, under purine starvation conditions small, but significant changes, in the relative abundance of the transcripts for XPRT (∼3-fold increased), LdNT3 (∼6-fold increased), and LdNT1 (∼2-fold increased), were observed. Notably, no substantial change in mRNA abundance for either 3′ nucleotidase/nuclease or acid phosphatase was observed at 48 h. In a previous study, the mRNA for 3′ nucleotidase/nuclease was observed to be upregulated under prolonged conditions of purine starvation (Sopwith et al., 2002). However this was likely due to differences in culture densities between the purine-starved and purine-replete parasites, leading to a downregulation of message in late log/stationary phase parasites grown under purine-replete conditions, rather than an overall transcriptional response to purine starvation. In general, however, these data suggest that the changes exhibited at the protein level are not the sole consequence of increased mRNA levels induced under conditions of purine starvation.

Table 3.

Analysis of gene expression in cells starved of purine for 48 h.

Gene Name	ΔCT		ΔΔCT	Fold Change
	Replete	Starved
APRT	-0.17 ± 0.02	-0.58 ± 0.03	-0.41	1.3
HGPRT	-1.27 ± 0.06	-1.98 ± 0.04	-0.71	1.6
XPRT	-0.12 ± 0.06	-1.52 ± 0.03	-1.4	2.6
LdNT1	-4.43 ± 0.03	-5.65 ± 0.04	-1.2	2.3
LdNT2A	-5.53 ± 0.02	-5.49 ± 0.04	0.04	0.97
LdNT2B	-5.63 ± 0.04	-5.73 ± 0.05	-0.1	1.1
LdNT2C	-5.66 ± 0.06	-5.69 ± 0.07	-0.03	1.0
LdNT3	-3.30 ± 0.13	-5.81 ± 0.03	-2.5	5.7
LdNT4	-1.37 ± 0.09	-1.81 ± 0.05	-0.44	1.4
3′-Nucleotidase	-0.16 ± 0.07	-0.61 ± 0.03	-0.45	1.4
Acid Phosphatase	-6.00 ± 0.07	-5.92 ± 0.02	0.08	0.95
MTAP	0.33 ± 0.04	-0.21 ± 0.01	-0.54	1.5
AAH	1.29 ± 0.02	0.76 ± 0.02	-0.53	1.4
GMPR	-0.58 ± 0.04	-0.78 ± 0.02	-0.2	1.1

Relative mRNA abundances for genes involved in purine metabolism were determined via qRT-PCR according to the relative abundance method of (Livak & Schmittgen, 2001). For each gene, the ΔC_T for starved and replete samples represents the mean threshold cycle from triplicate technical replicates normalized to the mean threshold cycle of the endogenous control gene OMPDC-OPRT. Standard deviation was determined according to the method described in Applied Biosystems User Bulletin No. 2 (P/N 4303859). ΔΔC_T was determined by subtracting the mean ΔC_T of the starved samples from the mean ΔC_T of the replete samples and the overall fold change was calculated from the formula 2 ^-ΔΔC_T. Primer sequences and gene accession numbers are listed in Table S1 of Supporting information. Subscripts A, B, and C indicate that the qRT-PCR was performed using primer sets F1/R1, F2/R2, and F3/R3, respectively, as described in Table S1. Abbreviations: MTAP = Methylthioadenosine phosphorylase; AAH = Adenine aminohydrolase; GMPR = GMP reductase.

Effects of cycloheximide on purine transport in purine-replete and purine-starved cells

Since it appeared that the upregulation of various purine salvage pathway components induced by purine removal from the medium occurred at a post-transcriptional level, the effect of cycloheximide, a potent inhibitor of cytosolic protein synthesis in eukaryotes, including L. donovani (Reiner & Kazura, 1982), was tested on Leishmania parasites incubated with or without purine for up to 12 h (Fig 10). Longer incubations with cycloheximide were not undertaken because of the toxicity of the drug towards the cells (∼20 % of the culture died in the presence of 10 μM cycloheximide with incubations of 12 h and beyond). 10 μM [³H]inosine transport was reduced in both purine-replete and purine-starved cells when incubated with 10 μM cycloheximide for either 6 or 12 h, suggesting that the steady-state levels of LdNT2 protein were reduced and that the transporter possesses a half-life less than 6 h. However, this effect was most acute in purine-starved cells after 12 h (17-fold reduced), where cycloheximide completely abolished the effects of purine withdrawal, reducing inosine transport to a level similar to that observed in purine-replete cells incubated in the presence of cycloheximide.

Fig. 10.

The effect of cycloheximide on the response to purine starvation in L. donovani. Wild type DI700 promastigotes were grown in RPMI-L with or without purine for 6 - 12 h in the presence or absence of 10 μM cycloheximide. Rates of uptake for 1 μM [³H]inosine were measured over 60 s as described in Experimental procedures for both purine-replete (open bars) and purine-starved cells (closed bars). Cells incubated in the presence of 10 μM cycloheximide are represented by CHX and the light grey (purine-replete) and dark grey (purine-starved) bars.

Discussion

Parasitic protozoa exhibit complex life cycles and likely encounter considerable fluctuations in their nutritional environment throughout their lifespan. As a first step towards understanding the adaptive responses in L. donovani to nutritional depletion, purines were withdrawn from the medium and the consequent effects on purine uptake examined. The data indicate that L. donovani nucleobase and nucleoside transport are altered in response to a depleted purine environment, which manifest as a rapid upsurge in purine uptake and accumulation. Moreover, the results indicate that this adaptive response is contingent upon the presence of the nucleoside transporter proteins, LdNT1 and LdNT2 (Carter et al., 2000, Vasudevan et al., 1998), as well as on the nucleobase transporter LdNT3 (Sanchez et al., 2004) and not due to either an increase in overall membrane permeability or to other secondary transport processes for purines induced under metabolic stress or purine starvation. These changes in purine uptake were also accompanied by an upregulation in the activities and concomitant changes at the protein level for the purine salvage enzymes, HGPRT, APRT, and XPRT, suggesting that purine metabolism is globally remodeled in response to purine starvation.

The adaptive response to purine starvation in L. donovani is similar to that described previously in other trypanosomatids (de Koning et al., 2000, Alleman & Gottlieb, 1996, Hall et al., 1996, Liu et al., 2005, Gottlieb, 1985, Sopwith et al., 2002, Yamage et al., 2000, Yamage et al., 2007), where increased nucleoside transport and/or purine salvage enzyme activities were observed under conditions of prolonged purine withdrawal. This intimates that all trypanosomatids possess a common mechanism for adapting to purine starvation and perhaps to any nutritional stress. The precise mechanism governing the adaptation to purine starvation has yet to be delineated in any trypanosomatid. However, the lack of complexity in nucleoside and nucleobase transport in L. donovani (Carter et al., 2008), the availability of purine transporter and purine salvage pathway deletion mutants (Liu et al., 2006, Boitz & Ullman, 2006a, Boitz & Ullman, 2006b), as well as antisera to key purine salvage enzymes (Allen et al., 1989, Jardim et al., 1999, Shih et al., 1998, Hwang et al., 1996, Zarella-Boitz et al., 2004), provides an attractive system in which to begin the dissection of the effects of purine withdrawal at a molecular level. Thus far, the data indicate that these changes in purine metabolism cannot generally be ascribed to the enhanced transcription of purine salvage pathway genes or to an overall increase in the stability of their messages (Tables 3 and S2). Instead, it appears that the upregulation observed in purine transport and metabolism is driven largely by post-transcriptional changes. From the data with the cytosolic protein translation inhibitor cycloheximide (Fig. 10), as well as that in the accompanying manuscript from Ortiz et al., it would appear that this post-transcriptional response is contingent, at least in part, upon the increased synthesis of components of the purine salvage pathway and/or modulators of the translational response.

Similar post-transcriptional responses to other types of nutrient stress have been documented recently in L. donovani and T. brucei, respectively (Darlyuk et al., 2009, Willert & Phillips, 2008). In L. donovani, amino acid starvation led to an increase in L. donovani arginine transporter, LdAAP3, both at the mRNA and protein level (Darlyuk et al., 2009), while perturbations in the polyamine biosynthetic pathway in T. brucei triggered an upregulation of both ornithine decarboxylase and prozyme, the catalytic activator of S-adenosylmethionine decarboxylase, at the protein level (Willert & Phillips, 2008). Since an increase in mRNA abundance could only be detected for one of the two LdAAP3 gene copies, this might imply that signals within the LdAAP3 mRNA itself, and possibly within either the 5′ or 3′ UTR, are important mediators of this effect, possibly by stabilizing LdAAP3 mRNA and/or enhancing the translational response. Whether signals in the 5′ or 3′ UTR of those messages encoding for purine salvage components altered in response to purine stress also facilitate the response to purine starvation remains to be established. Cursory evaluation of the predicted mRNAs for these regulated components does not suggest any obvious commonalities in either sequence or secondary structure. In addition, at least for the episomally expressed GFP-LdNT2 gene in Δldnt2 cells, endogenous LdNT2 5′ and 3′ UTR sequences do not appear necessary to achieve augmented protein levels under conditions of purine starvation (Fig. 8B). It is clear, nevertheless, that the mechanism by which translation of each component is increased under conditions of purine starvation is complex and is likely multifaceted.

Whether other metabolic pathways and cellular networks are also affected by purine withdrawal is unknown and will need to be evaluated by more global approaches beyond the scope of this manuscript. However, from the limited analyses undertaken within this manuscript it would appear that the relative abundances of other metabolic components and markers, e.g. spermidine synthase, OMPDC/OPRT, GAPDH, arginase, and α̃tubulin, are largely unchanged (Fig. 9C), suggesting that purine withdrawal does not universally refashion cellular metabolism in these parasites.

In other eukaryotic cells, such as yeast and mammalian cells, nutrient restriction has been linked to the entrance into a quiescent G₁/G₀ phase within the cell cycle (Smets et al., 2010, Pardee, 1974). The removal of purines from the culture medium also leads to the growth arrest of Leishmania parasites (Fig. 2B) and an increase in the number of cells accumulating in G₁ phase by FACS analysis (Fig. 2C). Whether growth arrest is a trigger for the downstream metabolic changes observed in the salvage pathway is an intriguing possibility. Equally plausible, however, is that growth arrest may simply arise as a consequence of insufficient purine for nucleic acid synthesis and nucleotide metabolism.

Nucleoside transport in mammalian cells is also regulated in response to environmental stressors. Immortalized murine cardiomyocytes under conditions of hypoxic stress exhibit decreased levels of transcript for the mouse equilibrative nucleoside transporter, mENT1, and this correlates with a reduced level of nucleoside transport and a decreased number of binding sites for the mENT1 inhibitor, S-nitrobenzyl-4-thioinosine (Chaudary et al., 2004). A similar result was also observed for hENT1, the human counterpart of mENT1, under conditions of prolonged hypoxia in human umbilical vein endothelia cells (Coe et al., 2002). Although the intracellular signals that transduce this response to hypoxia are not known, the changes in mENT1 and hENT1 activities may be orchestrated in part by protein kinase C (PKC). Whether PKC is involved in modulating the changes in nucleoside transport in purine-starved Leishmania parasites remains to be established. However, the data presented here imply that post-translational modification of either the LdNT1 or LdNT2 protein is unlikely, since the enhanced activity observed for LdNT2, the only transporter evaluated at the protein level in these studies, can solely be explained by its increased synthesis (Fig. 10) and concentration at the cell surface (Fig. 8B) rather than by any modulation of its activity by post-translational modification.

An ability to respond to environmental changes is essential for a successful parasitic lifestyle. Our results, together with those in the accompanying manuscript from Ortiz et al., suggest that Leishmania have evolved a comprehensive mechanism to adapt to purine stress that involves the augmentation of endogenous purine transport proteins as well as purine salvage components. These investigations, together with studies on several other trypanosomatids (Alleman & Gottlieb, 1996, de Koning et al., 2000, Hall et al., 1996, Liu et al., 2005) and a variety of eukaryotic cells (Chaudary et al., 2004, Coe et al., 2002), indicate that the translational control of purine salvage components is modulated by the extracellular milieu and ultimately, may prove pivotal in maintaining intracellular purine homeostasis and the energy balance of the cell under conditions of cellular stress.

Experimental procedures

Chemicals, materials, and reagents

Radiolabeled [2,8-³H]-D-inosine (31.3 Ci mmol^-1), [8-³H]-D-guanosine (7.5 Ci mmol^-1), [2,8-³H]-D-adenosine (36.8 Ci mmol^-1), [5,6-³H]-uridine (37.1 Ci mmol^-1), [8-³H]-hypoxanthine (27 Ci mmol^-1), [2-³H(N)]-D-glucose (20 Ci mmol^-1), and [guanido-¹⁴C]-L-arginine (55 mCi mmol^-1) were purchased from Moravek (Brea, CA). α-[32P]dCTP (3000 Ci mmol^-1) was procured from MP Biomedicals (Irvine, CA). Hygromycin B was purchased from Roche Applied Science (Indianapolis, IN), G418 from BioWhittaker (Walkersville, MD), and phleomycin from Research Products International Corp. (Mt. Prospect, IL). Cycloheximide was purchased from Sigma-Aldrich (St. Louis, MO). Oligonucleotides for the qRT-PCR (detailed in Table S1 of the Supporting information) and for generating the pSNAR-HA-LdNT2-RV episome were purchased from Integrated DNA Technologies (Coralville, IA). The mouse living colors A.v. monoclonal JL-8 GFP antibody was purchased from BD Biosciences (Palo Alto, CA), mouse monoclonal anti-α–tubulin DM1A was purchased from EMD Biosciences (La Jolla, CA), and the mouse monoclonal anti-human influenza hemagglutinin (HA) antibody was obtained from Sigma-Aldrich. Rabbit antisera against the L. donovani HGPRT, APRT, XPRT, and spermidine synthase full-length polypeptides, have been previously described elsewhere (Allen et al., 1989, Jardim et al., 1999, Shih et al., 1998, Hwang et al., 1996, Zarella-Boitz et al., 2004, Roberts et al., 2001). Custom rabbit antisera were generated against the full-length polypeptides of Leishmania mexicana arginase and L. donovani orotidine-5-phosphate decarboxylase/orotate phosphoribosyltransferase/(OMPDC/OPRT) by Sigma-Aldrich. Rabbit antiserum raised to the glycosomal glyceraldehyde-3-phosphate dehydrogenase (GAPDH) protein of L. mexicana (Hannaert et al., 1992) was provided by Dr. P. Michels of de Duve Institute and Laboratory of Biochemistry, Université catholique de Louvain. Sheep antiserum raised to the L. chagasi major surface protease (GP63) protein (Roberts et al., 1995) was obtained from Dr. Mary Wilson of University of Iowa, Iowa City, IA. All other reagents were of the highest grade commercially available.

Cell lines and cultivation

L. donovani strain 1S2D (Dwyer, 1976) were employed in these studies. The studies were undertaken with promastigotes of the clonal derivative DI700 (Iovannisci & Ullman, 1983) and derived mutants, DI700Δldnt1, DI700Δldnt2, DI700Δldnt2[pXG-GFP2+′], DI700Δldnt2[GFP-LdNT2], and DI700Δldnt2[pSNAR-HA-LdNT2-RV], as well as with the visceralizing clonal derivative LdBob (Goyard et al., 2003) and its derived mutant LdBobΔldnt3. Note that the strategy for the construction of the DI700Δldnt1 and DI700Δldnt2 mutants in which both wild type copies of either LdNT1 and LdNT2 have been eliminated by targeted gene replacement and loss-of-heterozygosity has been described elsewhere (Liu et al., 2006), as has the creation of the DI700Δldnt2[pXG-GFP2+′] and DI700Δldnt2[GFP-LdNT2] cell lines (Arastu-Kapur et al., 2003). The construction of the pSNAR-HA-LdNT2-RV episome that contains a 5 kb EcoRV fragment harboring hemagglutinin-tagged LdNT2 flanked by its native 5′ and 3′ UTR is described in the Supporting information. The strategy for the derivation and characterization of the Δldnt3 mutation within LdBob will be described elsewhere. Wild type DI700 parasites were propagated at 26 °C in 5 % CO₂ in RPMI 1640 Leishmania medium (RPMI-L) supplemented with NaHCO₃ (2 g l^-1), D-glucose (2.5 g l^-1), Tween-80 (80 mg l^-1), bovine serum albumin (BSA) (3 g l^-1), penicillin G (0.06 g l^-1), streptomycin sulfate (0.1 g l^-1), hemin (2 ml from a 500× stock containing 2.5 g l^-1) and containing 100 μM xanthine as a purine source. The DI700Δldnt1 and DI700Δldnt2 strains were cultured continuously in RPMI-L supplemented with 50 μg ml^-1 hygromycin to select for the drug resistance marker used in the targeted gene replacement strategies, and the following DI700Δldnt2 derivatives DI700Δldnt2[pXG-GFP2+′], DI700Δldnt2[GFP-LdNT2], and DI700Δldnt2[pSNAR-HA-LdNT2-RV] were maintained with an additional 25 μg ml^-1 G418 to maintain the episome. LdBob promastigotes were cultured at 26 °C in 5 % CO₂ in modified Dulbecco's Modified Eagle-Leishmania (DME-L) medium (Iovannisci & Ullman, 1983) that lacked bovine serum albumin and was supplemented with 5% serum plus (Sigma-Aldrich), 1 mM glutamine, 1× RPMI 1640 vitamin mix, 10 μM folate, hemin (2 ml from a 500× stock containing 2.5 g l^-1), and adenine (50 μM) and xanthine (50 μM) as purine sources. To elicit purine starvation conditions, cells were either grown in RPMI-L (DI700 and derivatives) or DME-L (LdBob and derivatives) lacking purines but with all other media components present.

Purine starvation and transport assays

All cells were grown to mid log phase (∼2 - 5 × 10⁶ cells ml^-1) and washed twice in medium lacking purines. Harvested cells were then resuspended in purine-replete or purine-depleted media (see above) at a density of 1 - 2 × 10⁶ cells ml^-1 and incubated at 26 °C with 5 % CO₂ until ready for analysis. Note that for the qRT-PCR experiments, Western analyses to measure relative protein abundance, and Michaelis-Menten kinetics measurements, the cell densities of the purine-replete cultures were adjusted daily to ensure that all cells were in early log phase. All nucleoside and nucleobase transport measurements were performed as previously described by a modified oil-stop method (Carter et al., 2000, Hasne & Barrett, 2000). Time courses were generated using 10 μM [2,8-³H]-D-adenosine (0.37 Ci mmol^-1), [2,8-³H]-D-inosine (0.31 Ci mmol^-1), 10 μM [8-³H]-hypoxanthine (0.27 Ci mmol^-1). Michaelis-Menten kinetics were measured in purine-replete and purine-starved cells for 0.1-10 μM [2,8-³H]-D-adenosine (0.37 Ci mmol^-1), 2.5-200 μM [5,6-³H]-uridine (0.37 Ci mmol^-1), 0.1-10 μM [2,8-³H]-D-inosine (0.31 Ci mmol^-1) and 0.1-10 μM [8-³H]-D-guanosine (0.08 Ci mmol^-1). Rates of uptake were analyzed and fitted to the Michaelis-Menten equation using Prism v. 4.0 (Graphpad Software, Inc., La Jolla, CA).

Glucose starvation assay

Glucose starvation was induced essentially as described previously (Rodriguez-Contreras & Landfear, 2006). Briefly, mid log phase cells were washed twice in RPMI-L lacking glucose. Harvested cells were then resuspended in glucose-replete (plus 11 mM D-glucose) or glucose-depleted media (minus glucose) at a density of 2 × 10⁶ cells ml^-1 and grown at 26 °C with 5 % CO₂ for 18 h. Transport measurements were performed as described above in the presence of 10 μM [³H]adenosine (0.38 Ci mmol^-1) or 10 μM [³H]glucose (0.20 Ci mmol^-1).

Differential interference microscopy

5 × 10⁶ purine-starved and purine-replete parasites were affixed to 0.01% poly-L-lysine chamber coverslides for 10 min, washed once in phosphate buffered saline (PBS) to remove unattached cells, and then fixed in 2% paraformaldehyde/formaldehyde for 20 min at room temperature. Cells were visualized on a Zeiss Axiovert 200 M deconvolution microscope using differential interference contrast (DIC). Images of both purine-starved and purine-replete parasites were captured on an AxioCam MRm camera. The “measure tool” in Adobe Photoshop 7.0 was used determine the average cell length of starved and replete parasites. A 10 μm bar placed on each image in Axiovision Release 4.6 was measured for pixel length and the length of each cell in the image was compared to this standard. All purine-starved (n=41) and purine-replete (n=63) cells imaged were counted in this way.

Growth curves and FACS analyses

The ability of L. donovani LdBob promastigotes to grow at 26 °C under purine-depleted conditions was assessed by seeding parasites washed free of purine (see above) at a density of 1 × 10⁶ cells ml^-1 in either purine-replete or purine-depleted DME-L. Parasite numbers were enumerated at 8, 24, 48, 72, and 96 h by hemocytometer. Cells from purine-replete and purine-starved cultures were also prepared for FACS analysis to examine the effects of purine starvation on cell cycle progression. Briefly, 1 × 10⁷ cells were removed from the cultures at 6, 12, 24, 48, and 72 h, washed once in 5 ml of PBS, and resuspended in 1 ml of PBS. Cell suspensions were fixed on ice in 2 ml of ice-cold 70 % ethanol for at least 1 h and the cells pelleted at 1500 × g prior to the addition of 0.5 ml of propidium iodide staining solution (3.8 mM sodium citrate, 50 μg/ml propidium iodide) and 0.5 μg ml^-1 RNase A. Cells were stained overnight at 4°C. Cell cycle analyses were performed by the OHSU Flow Cytometry Shared Resource on a BD FACSCalibur Flow Cytometer.

RNA extraction and qRT-PCR analyses

Total cellular RNA was isolated from 5 × 10⁷ to 1 × 10⁸ L. donovani promastigotes using an RNeasy Mini Kit ((Qiagen Inc., Valencia, CA) following the manufacturer's protocol for on-column DNase I digestion to minimize potential contamination by genomic DNA. First strand cDNA was synthesized with a Superscript III First Strand cDNA Synthesis System for RT-PCR (Life Technologies - Invitrogen, Carlsbad, CA) using 2 μg (experiment 1) or 0.1 μg (experiment 2) of total RNA and random hexamer primers as per the manufacturer's instructions. For qRT-PCR, 5 μl of diluted first strand cDNA reaction corresponding to 2 ng of input RNA was included in a 20 μl reaction with 5 pmol primers (Table S2) and 10 μl 2× SYBR Green PCR Master Mix (Life Technologies - Applied Biosystems) and the reactions were run on an Applied Biosystems 7500 Real-Time PCR System under the following conditions: 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. A final thermal dissociation analysis was performed for each reaction to confirm that the PCR generated a single amplification product. The relative abundance of target amplicons between purine replete and starved samples was determined via the 2^-ΔΔCT method (Livak & Schmittgen, 2001).

Western analyses

Protein samples were fractionated by SDS-polyacrylamide gel electrophoresis on a 10% polyacrylamide gel according to standard conditions (Sambrook, 1989). Immunoblotting was performed as described (Sambrook, 1989). Mouse anti-GFP and anti-HA antibodies were used at a 1:1000 dilution. Rabbit antisera against HGPRT, APRT, XPRT, GAPDH, OMPDC/OPRT, and SPDSYN were used at 1:3000 dilution and antisera against arginase at a dilution of 1:10,000. All antibodies were diluted in PBS containing 0.01% Tween 20 and 3% BSA. The secondary antibody of rabbit anti-sheep IgG conjugated to horseradish peroxidase (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was used at 1:10,000 dilution and prediluted (10 μg/ml) stabilized goat anti-rabbit IgG, and goat anti-mouse IgG horseradish peroxidase conjugates (Thermo Fisher Pierce Protein Research Products, Rockford, IL) were used at a 1:2500 dilution. Signal was detected by the SuperSignal West Dura Extended Duration Substrate (Thermo Fisher Pierce Protein Research Products). The mouse anti-α-tubulin antibody DM1A was diluted to 1:5000 in PBS containing 0.01% Tween 20 and 3% BSA and used as a control to normalize protein loading onto each lane of the gel.

Cell surface labeling

Biotinylation reactions were performed on purine-replete and purine-starved parasites essentially as described in (Arastu-Kapur et al., 2003). Briefly, 1 × 10⁸ cells were washed free of cellular debris and media contaminants with PBS, resuspended in PBS containing 1 mg/ml EZ-Link Sulfo-NHS-LC-Biotin (Thermo Fisher Pierce Protein Research Products), and incubated on ice for 1 h. The biotinylation reaction was quenched by washing the cells three times in 50 mM glycine in PBS. Cell membranes were solubilized on ice in 100 μl of lysis buffer (50 mM Tris-HCl, pH 8.0, 100 mM NaCl, 1% Nonidet P-40, and 10% glycerol) for 1 h. Cell debris was removed by pelleting in a microcentrifuge, and the solubilized membranes removed and incubated for 1 h at room temperature with 20 μl of packed streptavidin-agarose beads (Thermo Fisher Pierce Protein Research Products) prewashed in lysis buffer to isolate the biotinylated membrane fraction. The beads were washed twice with 1 ml of lysis buffer to remove nonspecific contaminants and resuspended in 20 μl of 2× Laemmli Sample buffer (Bio-Rad Laboratories, Hercules, CA). All samples were fractionated by SDS-polyacrylamide gel electrophoresis on a 10% gel and biotinylated protein detected by immunoblotting with anti-GFP, anti-HA, or anti-GP63 (loading control) antibodies as described above. To quantitate the amount of either epitope-tagged LdNT2 protein or GP63 labeled at the cell surface, Western blot films were scanned and densitometry performed on the developed films using AlphaEaseFC software (version 4.0.0, Alpha Innotech Corp., San Leandro, CA). The relative intensities of the epitope-tagged LdNT2 bands on each film were estimated after correcting for loading differences by comparison of the intensities of the GP63 bands.

Phosphoribosyltransferase activity assays

5 × 10⁸ of either purine-replete or purine-starved cells were washed twice in PBS to remove all media contaminants and were resuspended in 0.5 ml of 50 mM Tris.HCl, pH 7.4, 5 mM MgCl₂, 1 mM phosphoribosyl pyrophosphate, and complete EDTA-free protease inhibitor cocktail (Roche Diagnostics Corp., Indianapolis, IN). Pelleted cells were subjected to four rounds of sonication for 15 s and the clarified lysate collected after centrifugation at 16,000 × g at 4 °C. The protein concentration in the lysate was estimated by Bradford assay. Enzyme assays were undertaken with 50 μg of protein within the total cell lysate and the rate of incorporation into the nucleotide and nucleic acid pool of 60 μM of either [8-¹⁴C]hypoxanthine (51 mmol mCi^-1) or [8-¹⁴C]adenine (50 mmol mCi^-1) quantified by binding to DE-81 filter paper as previously described (Iovannisci et al., 1984).