The Effects of Glucose Concentration on the Reciprocal Regulation of rRNA Promoters in Plasmodium falciparum*

The developmental progression of Plasmodium falciparum is remarkably sensitive to glucose concentration. We have investigated the effects of glucose concentration on the parasite development cycle as reflected by changes of ribosomal RNA (rRNA) transcription. We showed that glucose starvation differentially affects transcriptional control of the rRNA genes by sharply repressing transcription from those loci involved with asexual development of the parasite while up-regulating transcription at those loci involved with sexual development of the parasite. Temperature change also effects regulation of transcription. We found that the effects of temperature and glucose were synergistic. We identified and compared the upstream region of the transcription start sites of each gene. These putative promoter structures are considerably different from one another and contain structures remarkably similar to rRNA control elements in other organisms.

The developmental progression of Plasmodium falciparum is remarkably sensitive to glucose concentration. We have investigated the effects of glucose concentration on the parasite development cycle as reflected by changes of ribosomal RNA (rRNA) transcription. We showed that glucose starvation differentially affects transcriptional control of the rRNA genes by sharply repressing transcription from those loci involved with asexual development of the parasite while up-regulating transcription at those loci involved with sexual development of the parasite. Temperature change also effects regulation of transcription. We found that the effects of temperature and glucose were synergistic. We identified and compared the upstream region of the transcription start sites of each gene. These putative promoter structures are considerably different from one another and contain structures remarkably similar to rRNA control elements in other organisms.
The number of ribosomes present in a cell is directly related to its protein-synthesizing activity and size of the cell. The ribosomes may make up 45% of the dry weight of an actively growing cell; thus, control of the synthesis of the molecules that make up the ribosome (e.g. rRNA) is essential to balance the needs of the cell with its external environment. Control of growth and function of a cell by regulation of ribosome production is somewhat analogous to controlling a vehicle with accelerator and brakes. Speed has to be monitored and regulated in conjunction with driving conditions for the sake of both fuel efficiency and safety. Likewise, the organism must have the means to speed up, slow down, or remain constant developmentally in response to its changing environment. Any one of a variety of metabolic rates may have an adaptive advantage depending on the situation.
Although Plasmodium is a eukaryote, its rRNA gene is unlike any other eukaryote. In most eukaryotes, the rRNA gene units are identical in sequence and tandem arrayed in large numbers. Plasmodium species are encoded by a small number of rRNA gene units that are physically separated in the genome and variable in sequence (1,2). Transcription studies indicate that eventually all of the genes are expressed over the course of the developmental cycle (3)(4)(5)(6). Distinct types of mature 18 S rRNA have been shown to be associated with asexual parasites in the blood (A-type rRNA) (4,5), gametocytes (this study), and developing sporozoites (S-type rRNA) (3,7). Although different ribosomes perform the same basic function-translation of mRNA-functional variations exist between the types (8). Temperature is known to be one factor involved with transcriptional control in the developing sporozoite-stage rRNA (9), but regulation of the other rRNA copies is still elusive.
Glucose is related to rRNA transcription control in a number of other microbes (10,11) and has been implicated in developmental transitions in Plasmodium (12). There are dramatic drops in glucose concentration during transmission of the parasite from vertebrate to insect (13) as well as significant changes in glucose concentration in an infected human, especially within the microcirculation (14). One can anticipate that some of the changes in transcriptional patterns will parallel changes in glucose concentrations, even in culture. Some changes will be effected as a direct result of glucose starvation. Other changes involved in developmental transitions may well involve glucose but require additional factors including temperature change (9). Glucose concentration, the most important factor associated with the parasite growth, is addressed in this study as related to transcriptional control of rRNA.

EXPERIMENTAL PROCEDURES
Parasite Culture and Separation Stages-Plasmodium falciparum 3D7 was cultured according to standard methods (15). The parasites were synchronized twice with 5% sorbitol, and different asexual stages of parasites were collected. Gametocytes were induced and harvested as described (16). Oocysts and sporozoites of the parasite were obtained by feeding laboratory-reared Anopheles gambia mosquitoes with P. falciparum-infected red blood cells (17). The parasite used for glucose starvation is the unsynchronized mixed stages of parasite.
Genomic DNA and RNA Isolation-Genomic DNA from cultured lines of P. falciparum was isolated as described previously (18). Total RNA from the various parasite stages was isolated using Trizol (Invitrogen).
Inverse PCR-Inverse PCR was performed as described (19). The restriction enzymes HindIII and AflIII were used to clone fragments spanning the 18 S rRNA and promoters. The amplified PCR fragments were cloned into vector pCRII-TOPO (Invitrogen) and sequenced.
Primer Extension and 5Ј-RACE 1 -Primer extension was used to map the transcriptional start sites of four rRNA genes as described (20). The RNA for A1 and A2 primer extension is from blood-stage parasites cultured at 37°C. The RNA for S1 primer extension is from gametocytes. The RNA for S2 primer extension is from blood-stage parasites cultured at 26°C for 3 h. Gametocytes were chosen for S1 initiation site determination due to the abundance of S1 type rRNA in gametocyte stage (see "Results"). The blood-stage parasites cultured at 26°C were chosen for S2 initiation site determination due to high transcription level of S2 at this temperature (9). 5Ј-RACE (Invitrogen) was used as an alternative method to determine the transcriptional start site.
Real-time PCR-One microgram of DNase-treated RNA was converted to cDNA using a SuperScript first-strand synthesis system (Invitrogen). 5% of the cDNA was used for real-time PCR using a Light-* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF503871, AF503870, AF503869, and AF503868.
Underlined letters indicate conserved regions and similar GC-rich regions between A1 and A2 rRNA genes. C, comparison of promoter structures of A-type rRNA genes in P. falciparum with promoter structure of rRNA gene in E. coli. Pink letters indicate conserved sequences between A-type rDNA promoter and E. Coli rDNA promoter.
Real-time PCR conditions for each gene was according to the Roche Applied Science protocol, except that the extension temperature was 60°C (21). Specific cDNAs were quantified using a standard curve based on the known concentration of the plasmids that contained the same DNA fragment. The products were cloned into pCRII vector and sequenced. The sequences were as predicted.

RESULTS
Cloning rRNA Gene Precursor and Promoter Regions-It has been shown that both A-type rRNA and S-type rRNA genes in P. falciparum have multiple copies and different loci properties (18). The sequences preceding two mature A-type rRNA genes (A1 and A2, GenBank TM accession numbers AF503871 and AF503870, respectively), and two mature S-type rRNA genes (S1 and S2, GenBank TM accession numbers, AF503869 and AF503868, respectively) were identified by inverse PCR. A schematic of each precursor is shown in Fig. 1; transcriptional patterns of each rRNA gene (rDNA) are described in ''Characterization of Precursors and Promoters of rRNA Genes.'' The sequence data obtained here are the same as those in the recently released Plasmodium data base (PlasmoDB www.plasmodb.org).
Characterization of Precursors and Promoters of rRNA Genes-RNA from different blood-stage or low temperaturetreated parasites was used as a template for mapping the initiation sites of different rRNA genes. The transcriptional start point for each of the four transcripts ( Fig. 2A) was revealed by primer extension experiments and confirmed by 5Ј-RACE technique. Distances from the transcription starting points of each rRNA gene to the 5Ј-end of mature 18 S rRNA are: A1, 1145 bp; A2, 1042 bp; S1, 1184 bp; and S2, 1318 bp (Fig. 1). The full-length precursors of A1 and A2 rRNA contain a similar 530-nucleotide region preceding the mature 18 S rRNA, although the sequences upstream to this region are diverse. In contrast, the only homologous sequence relating S1 and S2 in the precursor is a 20-bp region immediately 5Ј to the mature 18 S RNA sequence that is similar to the corresponding sequence in the A genes.
The region upstream to the initiation site is the putative rRNA gene promoter. No conserved sequences were found upstream to the S1 and S2 initiation sites. Surprisingly, the general structure of these A1 and A2 promoters is very similar to the rRNA gene promoter structure in Escherichia coli (22,23). Conserved sequences encoding the putative promoter are seen in A1 and A2. Two conserved boxes similar to the Ϫ10 and Ϫ35 regions of E. coli rRNA gene promoter were identified upstream to the A1 and A2 transcription start sites. The distance between these two boxes is 61 bp in A1 and 64 bp in A2. Two GC-rich regions of 20 bp each, similar to that in the E. coli rRNA gene promoter, were found just preceding the A1 and A2 transcription start site. The distances between the GC-rich regions and the second conserved box are 90 bp and 92 bp, respectively (Fig. 2, B and C). The fact that the P. falciparum genome is more than 80% AT makes these GC-rich boxes notable.
Transcriptional Pattern of Different rRNA Genes-Sequence analysis of the rRNA precursors revealed polymorphisms 5Ј to the mature 18 S rRNA gene in all copies in ''Characterization of Precursors and Promoters of rRNA Genes.'' Their existence allowed us to design experiments to specifically detect transcripts from each of the four genes. To establish a baseline, RNA from various stages of parasite development was purified, and different copies of A-and S-type rRNA were quantified by detecting the transient precursor region by real-time PCR (Fig.  3). The results showed, as expected, that genes A1 and A2 were the predominant forms in asexual parasites. Full-length transcripts encoded by S1 are dominant in gametocytes but decrease in relative amounts as the parasite develops through the mosquito stages. S2 transcripts, first seen in overwhelming amounts in oocysts, are the dominant gene product in sporozoites.
Glucose Concentration and rRNA Gene Transcription-Transcriptional change as a result of glucose variation was measured at 37°C. Aliquoted parasites were cultured at five glucose concentrations (2, 1, 0.5, 0.25, and 0.11 mg/ml) for 3 h. The normal glucose concentration in human sera ranges from 0.7 mg/ml to 1.3 mg/ml. 1 mg/ml represents the normal glucose concentration in human sera. Glucose concentration in severe malaria with hypoglycemia can be as low as 0.12 mg/ml (24) and even lower in a microcirculation system occluded by sequestrated parasites. We show that culture at the lowest glucose concentration (0.11 mg/ml) for 3 h does not affect parasite viability. Relative amounts of transcription from A1, A2, S1, and S2 genes were determined by real-time PCR (Fig. 4). Each experiment was performed on three independent occasions with different sample sources yielding consistent results. Transcription of A1 at 0.11 mg/ml glucose was decreased relative to the control (1 mg/ml) sample by more than 85% (Fig. 4A); transcription of A2 was decreased by 80% (Fig. 4B). No variation in transcriptional levels was observed for S1 (Fig. 4C). Transcription of S2 was found to increase as glucose level decreased (Fig. 4D).
The effect of low glucose on rRNA transcription was observed over a time course. Aliquoted parasites were separately cultured in control media containing 1 and 0.11 mg/ml glucose for 1, 2, and 3 h. Transcription of A1 and A2 decreased continually over time, whereas no changes were observed in S1. S2 transcription continued to increase as the starvation time increased (Fig. 5).
The viability of parasites under our starvation conditions was determined by recovery of A-type rRNA transcription. Blood-stage parasites were starved at 0.11 mg/ml for 3 h and then returned to 1 mg/ml glucose for another 3 h. Fig. 6 shows that the A gene transcription level fully recovered.
A Synergistic Effect of Glucose and Temperature-During transmission of the parasite from the vertebrate host to the insect host, temperature change parallels glucose concentration change. We analyzed the influence of glucose concentrations and temperature change alone and in combinations. The results are shown in Fig. 7. Cultured parasites at 26°C for 3 h down-regulate the transcription level of A-type rRNAs (A1 and A2) close to 50%. Glucose starvation can down-regulate A1 and A2 transcription levels by 85 and 80%, respectively. The glucose starvation in combination with low temperature can further decrease the transcription of A1 to 90% and A2 to 84% (Fig. 7, A and B). The effect of the combination is too small to say anything about a cooperative effect. S1 is regulated differently than the other genes (Fig. 7C). Neither temperature nor glucose starvation have a significant effect the level of transcription.
The transcription level of S2 is affected by a combination of glucose starvation and low temperature in a synergistic manner. The glucose starvation alone can up-regulate S2 transcription by 50% whereas low temperature alone can increase S2 transcription 16-fold. Glucose starvation in combination with low temperature has a synergistic effect increasing S2 transcription by 49-fold (Fig. 7D) in what appears to be a collaborative association. DISCUSSION Developmental transitions in eucaryotic microorganisms relating to glucose starvation have been extensively studied in yeast. In parallel with the situation in P. falciparum, when yeast encounters glucose starvation or nitrogen starvation, sexual development is triggered: the mating ratio of yeast increases, and the cells perform conjugation, meiosis, and sporulation (25). Transmission of Plasmodium from vertebrate host to mosquito likewise involves development through the sexual cycle and meiosis. Changes in rRNA transcription levels are integral to the yeast transformation. We have investigated Plasmodium for similar responses. The transcription pattern of different copies of rRNA indicates that two copies of A-type rRNA have the same type of transcriptional regulation. S-type rRNA, however, is not one type but two with regard to transcriptional control; one is expressed in gameteocytes and the other in sporozoites. Dramatic sequence differences between the two copies of S-type genes occur in the 5.8 S and 28 S regions. The sequences now can be confirmed by the recently released P. falciparum genomic data base. The synergistic effects of glucose starvation and low temperature on down-regulation of A1 and A2 and up-regulation of S2 are consistent with this stage specific transcription characteristic. The gametocyte, like a bridge connecting asexual stage and sexual stage of parasite life, encounters both environments of vertebrate and insect. As a primarily expressed rRNA in gametocyte, S1 should not respond to the change of environmental factors dramatically.
In all organisms, nutrients are primary regulators of signaling pathways that control transcription and translation (26 -28). Carbon and amino acids are basic nutrient sources for cellular organisms. They supply energy for metabolism. Organisms have distinct sensing and signaling pathways to regulate gene expression in response to the type and quantity of their energy sources. The change in transcriptional patterns of rRNA resulting from the low glucose concentrations indicates that the parasite changes its energy source when it transmitted from vertebrate to mosquito host. Actually, it has been proposed that amino acids are the main energy source of Trypanosoma brucei upon transition to its insect vector, whereas glucose is the main energy source of the blood stage parasite (29).
With regard to the molecular mechanism of control over rRNA transcription in many multi-cellular eukaryotes, rRNA transcription is tightly controlled and can be either silenced or opened. The ratio of activating and inactivating rRNA genes is tissue-specific (30). It has been shown that silencing of rRNA genes is often methylation-dependent (31). We tried to determine if the regulation of rRNA genes in P. falciparum is methylation-dependent by using methytransferase inhibitor, 5-azacytidine, which can inhibit the inactivation of rRNA genes in mouse cell. We were unable to detect a relationship between methylation and transcriptional regulation of rRNA gene in the parasite.
There are a number of parallels with the type of control described here for P. falciparum with that initially described for E. coli and is referred to as the "stringent response." Carbon source starvation or amino acid starvation in E. coli can cause the dramatic inhibition of the synthesis of rRNA, tRNA, and mRNA for ribosomal proteins as well as selective pattern changes in other mRNAs (10). The definition of a stringent response has been broadened to include organisms such as yeast (11), which also selectively reduce the transcription of ribosomal RNA and mRNAs of ribosomal protein in response to external signals such as amino acid starvation. Interestingly, it is now known that the stringent response in prokaryotes is dependent on the GC-rich region just preceding the initiation site of transcription (32)(33)(34). The similar GC-rich sequences in the promoters of A-type rRNA genes in P. falciparum suggest that the reciprocal regulation of different types of ribosome RNA genes in P. falciparum under glucose starvation fits into this category. Identifying the specific proteins or enzymes involved would be beneficial to the development of new drug target.