Species-specific regulation and switching of transcription between stage-specific ribosomal RNA genes in Plasmodium berghei.

Malaria parasites (Plasmodium spp.) differentially express structurally distinct sets of rRNA genes in a stage-specific manner. The four rRNA genes of the rodent malaria parasite, P. berghei, form two classes of 2 units that are genetically unlinked and termed A-type and S-type. Through Northern analysis and in situ hybridization, expression of the units was demonstrated in synchronized parasite preparations covering the developmental pathway from the initiation of the blood-stage asexual cycle to the production of mature ookinetes. A-type units were transcribed in direct response to cell growth in bloodstage asexual parasites yet were differentially regulated during male (inactive) and female (active) gametocytogenesis. S-type expression was not confined solely to the mosquito stages and exhibited a finite period of expression in a subset of bloodstage trophozoites that was significantly elevated in gametocyte-producing parasites. Unlike in the human parasite, P. falciparum, there was no evidence for accumulation of precursor forms of the S-type transcripts in gametocytes. No significant rRNA transcription was observed in cultured, fertilized ookinetes until ∼20 h of development when S-type transcription was initiated. The results further demonstrate that in Plasmodium the expression of the different rRNA units is linked to developmental progression but in a species-specific manner.

One of the remarkable features of Plasmodium parasites is the developmentally regulated ribosome characterized by transcription in a stage-specific manner of discrete repertoires of unlinked nuclear rRNA genes, now termed A-type (from Asexual) and S-type (from Sporozoite) (1). The exceptional pattern of transcription of these genes was initially discovered in the rodent parasite, Plasmodium berghei, where it was broadly demonstrated that the asexual blood stages transcribe the Atype units, whereas the S-type units are transcribed in the mosquito stages of the parasite, including the sporozoite (2). A similar pattern of rRNA unit transcription has been confirmed for every species of Plasmodium investigated (3)(4)(5). Subse-quent work in different parasite species demonstrated that the switches in rRNA unit transcription were linked to parasite development associated with transitions between the vector and host. Thus, it was shown in the human parasite, Plasmodium falciparum, that the switch from A-to S-type transcription occurred during gametocytogenesis and zygote formation and involved accumulation of S-type rRNA precursor forms (6). In P. berghei, the switch back to A-type transcription occurred early in the liver stages following invasion of cultured hepatocytes by sporozoites (7). Recent work in Plasmodium vivax has shown that the transition to S-type rRNA occurred much later in the period of development within the mosquito, at the point in oocyst development when sporozoites were being formed and contrasting with the observations in P. falciparum (5).
Lately, a third type of developmentally regulated nuclear rRNA gene has been demonstrated in P. vivax, a finding that may prompt a re-evaluation of all previous data. Termed Otype, their transcription was first detectable in ookinetes, and their period of maximum transcription was in oocysts separating those of the A-and S-types. O-type genes are structurally significantly different from both the A-and S-type. Transcripts of the O-type appeared to be confined to the oocyst structure and were not detected in salivary gland sporozoites. 1 These observations and the fact that the O-type has not yet been demonstrated in P. berghei indicated that there may be species-specific features of the switch and the types of genes involved. The four rRNA units in the nuclear genome of P. berghei are currently divided into two pairs representing the A-type (A and B units) and S-type (C and D units) (8,9), which somewhat simplifies analysis. We present here a description of the steady-state transcriptional activity of these stage-specific units through the use of probes directed against the relatively rapidly evolving A and C unit external transcribed spacer (ETS) 2 regions. We found that transcription of the P. berghei nuclear rRNA genes was also developmentally regulated. Specific and unexpected periods of rDNA unit transcription were observed, demonstrating a complex pattern of transcriptional regulation that could be linked to the various differentiation pathways of the parasite.

MATERIALS AND METHODS
Parasites-The following cloned lines of the ANKA strain of P. berghei were used in this study: 8417 (gametocyte producer); clone 2.34 (producer); 8458 (low producer); and clone 2.33 (nonproducer). In addition, the nonproducer clone 1 of the K173 strain was used (10).
Production of the Various Stages of the Parasite-Synchronous asexual blood stages were obtained from infections in rats that were syn-* This work was financially supported in part by Grants TS3*-CT93-0116 and ERB3514PL950457 from the STD3 program of the European Union. 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.
§ To whom correspondence should be addressed. chronized as has been described (11)(12)(13). Parasites synchronized in this way typically have a window of development no greater than 3 h. Blood containing ring forms was collected from the rats, and leukocytes were removed as described and incubated under standard in vitro culture conditions (11) for 21-24 h to allow the development of the rings into mature schizonts. From these cultures, samples were taken at the appropriate time points for RNA extraction (see below). Immature gametocytes were purified from blood collected from rats with synchronous infections as described (14). These immature gametocytes were incubated for a period of 6 h under standard culture conditions (11) for complete maturation. The purified gametocyte preparation consisted of 95% male and female gametocytes in a 1:1 ratio mixed with 5% mature schizonts and ring forms.
Zygotes and ookinetes were obtained from in vitro cultures as described (15). In this study, we used two different sources of gametocytes to start the cultures: 1) purified gametocytes, which were collected as described above (in these cultures, normally 50 -70% of the females were fertilized, and there was a 5% schizont contamination (10)); and 2) infected blood obtained from mice with a high proportion of gametocytes (16). From this material, ookinete smears were prepared for in situ hybridization or were purified from unfertilized gametes, young gametocytes, and trophozoites by density gradient centrifugation, resulting in ookinete preparations consisting of mature ookinetes, uninfected cells, and ring forms.
Nucleic Acid Techniques: RNA Extraction-RNA was prepared from purified parasites as reported previously (10). RNA was fractionated in formaldehyde containing agarose gels and blotted to nylon membrane (Hybond Nϩ, Amersham Corp.) according to standard protocols (17).
Definition of the rRNA Units and Probes Used in This Study-The four rRNA gene units of P. berghei were originally defined as A-D according to their size upon digestion of genomic DNA with EcoRI (8). This definition has survived the discovery of the specific developmental transcription of the different units, which are now defined as A (Asexual vertebrate host), S (Sporozoite), and more recently, O (Oocyst). This work confirms what was evident from the early restriction mapping that the four units form two groups of two units, currently termed A and S, and the individual units will continue to be referred to as A, B, C, and D. As discussed, it is not yet possible to distinguish between the C and D units, although restriction analysis (8,18) and preliminary sequence data 3 indicate that this should be possible. The S units may yet have to be further redefined into an O and S type. New probes generated during the course of this study are referred by numerical designation according to use in the laboratory with lettering A or S according to the type of unit they were derived from and detect.
Probes-The probes were derived by subcloning or PCR amplification from the following two genomic DNA clones that have been described previously (8,19). A-unit probes were derived from clone pPbS5.6 (8), a 5.6-kilobase pair EcoRI/HindIII fragment of P. berghei genomic DNA, which contains the A-unit SSU gene and 3.0-kb of upstream sequence. S-unit probes were derived from clone pPbSL8.8 (19), an 8.8-kb EcoRI/ HindIII fragment of the C P. berghei genomic DNA that contains part of the C-unit LSU gene, the 5.8S gene, the SSU, and ϳ1 kb of sequence upstream of the SSU. These clones and the regions used in this study as probes are illustrated in Fig. 1.
Oligonucleotides used in the PCR amplification reactions were: 365R, 5Ј-TAA TAC GAC TCA CTA TAG GGA GAG ATC AAC CAG GTT-3Ј, which corresponds in an antisense fashion to the 5Ј end of the mature 18 S unit of the A and C rDNA units but also includes the T7 RNA promoter for the optional generation of antisense RNA probes (this was used for the amplification of ETS probes from both the A and C units); 38A, 5Ј-TAT GCA TCA TAA TAA AGA-3Ј, which corresponds in a sense fashion to position Ϫ395 with respect to the 18 S gene of the A unit 4 and when used with 365R, generated the 101A probe of 438 nt; and PUC R reverse sequencing primer, 5Ј-AGC GGA TAA CAA TTT CAC ACA GGA-3Ј, was used with 365R to amplify 99S, an ϳ1-kb probe containing sequences homologous to the C unit ETS region. PCR conditions were 4 cycles of 94°C for 30 s, 40°C for 30 s, and 72°C for 1 min, followed by 25 similar cycles but with an elevated annealing temperature of 54°C. The Taq polymerase used in these studies was purchased from Life Technologies, Inc. The antisense oligonucleotides TM4 and TM3 are specific for the A-and S-type SSU rRNA, genes, respectively and have been described before (2). Final wash conditions for these probes was at T m (melting temperature) Ϫ5°C in 3 ϫ SSC, 0.5% SDS (1 ϫ SSC is 0.3 M NaCl, 0.03 M sodium citrate).
Clone 315A is an EcoRI/NsiI subclone from pPbs5.6 (18) encompassing 2.6 kilobase pairs of the region upstream of the mature SSU A unit (Fig. 1). The clone of the P. falciparum A-type SSU rRNA gene has been described before (3).
Probes 104S and 105A are PCR-generated probes specific for the internal transcribed spacer region 1 of the C and A units, respectively, and are shown in Fig. 1. Final wash conditions for all complex probes was 0.1 ϫ SSC, 0.5% SDS at 65°C. Templates containing a 3Ј T7 RNA polymerase initiation site for the synthesis of antisense RNA probes by in vitro transcription were generated by two rounds of PCR amplification as follows.
A template for the transcription of a P. berghei rRNA-specific probe was generated by amplification of a 190-nucleotide fragment of clone pPbS5.6, containing P. berghei A SSU rRNA, using the 5Ј primer CTTAACCTGCTAATTAGC in conjunction with the 3Ј primer GACT-CACTATAGGGCCACCGTTCCTCTAAG, which includes 15 nucleotides corresponding to the T7 RNA polymerase initiation site. PCR amplification was carried out in the presence of the 3Ј primer alone for 5 cycles at 94°C for 1 min; 36°C for 1 min; and 72°C for 1 min; then, in the presence of both primers, for 30 cycles at an annealing temperature of 48°C. To complete the T7 RNA polymerase initiation site and add 10 bp of protecting sequence, a second round of amplification was carried out under the same conditions, using 0.5 l of the first amplification reaction mix in the presence of the same 5Ј primer and the 3Ј primer ACAAGCTTCTAATACGACTCACTATAGGGC.
A template for the transcription of a P. berghei rRNA S-ETS-specific probe was generated in the same way by amplification of a 246-bp fragment of the C-unit ETS region within clone pPbSL8.8 in the presence of the 5Ј primer ATTCACTTCCGTAATTTG and, during the first round of amplification, the 3Ј primer GACTCACTATAGGGCTCATT-TACCGTGATC. The annealing temperatures during the first round of amplification were 5 cycles at 32°C and 30 cycles at 42°C, and during the second round were 5 cycles at 36°C and 30 cycles at 46°C. A template for the transcription of a P. berghei rRNA A-ETS-specific probe was generated by amplification of a 438-bp fragment of the A-ETS region within clone pPbS5.6 in the presence of primers 365R and 38R as described above.
Chromosome Analysis-Chromosomes of the different strains and clones of P. berghei were fractionated by pulsed field gel electrophoresis, transferred to nylon, and hybridized using conditions described previously (20).
In Situ Hybridization to Malaria Parasites-The methodologies describing the preparation of P. berghei parasites and suitable RNA and DNA probes for in situ RNA hybridization and their application have been described thoroughly elsewhere (21)(22)(23). Probes were labeled either through the direct incorporation of fluorescein-UTP or the incorporation of digoxigenin-UTP/dUTP as appropriate and detection with fluorochrome-linked specific antibodies. All chemicals and reagents for these processes were purchased from Boehringer Mannheim. The method of visualization used to generate the data shown is indicated in the text and figure legends.

The ETS Probes Distinguish between the A-and S-type rRNA
Classes-The reader is referred to "Materials and Methods" for the definition of the RNA units. The ETS regions of the A and C rRNA units were isolated from the genomic clones described previously (18) by a combination of PCR and subcloning procedures. These and other probes used in this study are illustrated in Fig. 1, which is a schematic representation of the A and C units showing the gene units and the ETS and internal transcribed spacer regions (adapted from Ref. 19). These probes were used to establish the chromosomal location and sequence type of the units and their patterns of transcription in the various stages of the life cycle reported here. A probe for the small subunit rRNA gene hybridized to chromosomes 5, 6, 7, and 12, showing that the four rRNA units are unlinked in the genome. The A-type ETS probes (101A and 315) hybridized only to chromosomes 7 and 12, whereas the S-type ETS probe (99S) hybridized to chromosomes 5 and 6 ( Fig. 2). This demonstrated that probes directed against the ETS regions of the A and C units were type specific, consistent with the observed similarity of respective restriction maps of the four units (8,24). The sequence of these regions will be the subject of another report that identifies and characterizes the transcription start site of the rRNA gene units. 5 The internal transcribed spacer region 1 probes (Fig. 1) have the same A-or S-type specificity (data not shown).
Transcriptional Activity of the rRNA Genes in Asexual Blood Stages and Gametocytes-The typical rDNA unit is transcribed as a single large primary transcript, which is then processed through a series of characteristic steps to generate the three mature rRNA species of SSU rRNA (ϳ2050 nt for Plasmodium), 5.8S (ϳ155 nt), and the (naturally nicked in P. berghei) LSU rRNA (ϳ800 and 3000 nt) (8,18). The ETS probes used in this study will, therefore, detect the primary transcript, and any secondary processed precursor species that contain the ETS but will not hybridize directly to the mature rRNA species. The Northern and in situ hybridization analyses demonstrated that the ETS probes hybridized only to precursor rRNA species and that these were confined to the nucleus (Fig. 3). The processing of rRNA precursor transcripts is a rapid process, and the RNA species have extremely short half-lives. Nuclear run-on analyses using blood-stage parasites confirmed that the A-type units are actively transcribed in these cell types (data not shown). Probes detecting these intermediates could, therefore, be used to measure steady-state rRNA unit transcription. This was performed using synchronous populations of distinct clones of P. berghei, which either could or could not produce gametocytes.
Four time points were investigated representing each of the morphologically distinct phases of the asexual blood-stage cycle (5 hpi, consisting of ring stages/young trophozoites; 16 hpi, old trophozoites; 25 hpi, mature schizonts/young gametocytes; and 28hpi, mature schizonts/mature gametocytes). RNA from equivalent numbers of parasites were loaded in each lane, which was quantitated by hybridization of an oligonucleotide (TM4) to the mature SSU rRNA species. In the gametocyteproducing clone (8417), the A-type units were transcribed throughout the asexual cycle, and the production of two major precursor species of 2700 and 2900 nt was observed hybridizing to probe 101A (Fig. 3A). A faint band of ϳ7.0 kb also hybridized to 101A and is interpreted as the size of the primary rRNA transcript of the complete unit or an early processing intermediate (data not shown). The 2900-nt precursor was only observed in the first 16 h after erythrocyte invasion, and both major precursors were equimolar during this period. However, the 2900-nt precursor was not observed during schizogony or in mature gametocytes. In the gametocyte nonproducing clone, the 2900-nt precursor is in the same parasite forms and to the same extent as in the producer clone (Fig. 3A, compare lanes 1,  2, 6, and 7). However, A-type transcription was barely observable in mature schizonts (Fig. 3A, compare lanes 3, 4, 8, and 9). We conclude that the 2700-nt precursor is produced by gametocytes, and these forms are responsible for the signal observed in the schizont samples taken from the gametocyte-producing parasite clone. Mature schizonts do not, therefore, transcribe rRNA genes. This pattern of transcription was observed using both 101A or TM4 (data not shown). Preliminary data indicated that the observed precursors consisted solely of the mature SSU and ETS regions. 6 These conclusions were confirmed and clarified by in situ hybridization of 101A to blood-stage parasites, which demonstrated that transcripts containing the ETS region were confined to the nucleus, consistent with the exclusive nuclear processing of eukaryotic rRNA primary transcripts (Fig. 3B). A-type transcription was not observed in newly invaded rings (data not shown) but was observed in all asexual mononuclear stages from 5 hpi. Furthermore, A-type transcription decreased with increasing schizont maturity and was barely detectable in fully mature schizonts (16n) and released merozoites (1n). Atype transcription in gametocytes was shown to be sex-specific. Female gametocytes continued to transcribe the A-type units, whereas the males were silent (Fig. 3B).
The S-type units were transcribed in a phased manner in blood-stage parasites (Fig. 3C). S-type precursors of an estimated size of 3000 and 2650 nt, respectively were observed with a peak in production at 16 hpi, when all cells are maturing trophozoites. The same pattern of S-type precursors was observed in both gametocyte producer and nonproducer clones, although the level of transcription in the former was significantly higher (Fig. 3C). The level of S-type transcription appeared to vary between clones between 2-and 10-fold, as estimated by densitometry (Fig. 3C) and appeared to be consistently greater in producer clones. The larger precursor was more abundant than the smaller (ratio, 5:1). A third band of Ͼ10 kb was also visible, which is interpreted as being the full-length primary rRNA transcript of the complete unit comparable to the smaller Ͼ7.0kb transcript of the A-type unit. The transitory appearance of the S-type precursors indicated that they did not accumulate as steady-state precursor forms that had been observed in gametocytes of P. falciparum (4), nor was accumulation of the mature transcripts observed through hybridization to the S-type-specific oligonucleotide, TM3 (data not shown). Neither transcription of the S-type units nor accumulation of S-type transcripts in mature gametocytes was evident (Fig. 3C) in clone 8417 illustrated nor in rings, mature trophozoites, nor schizonts from independent nonproducer parasite clones, LK173 and 233 (see "Materials and Methods"; data not shown).
In Situ Hybridization of the S 1 -ETS probe to synchronous blood-stage parasites confirmed that the peak of S-type transcription occurred at 16 hpi during trophozoite development. The peak appeared to be restricted to a subset of mononuclear cells (Ͻ1%). As in Northern analysis, neither S-type transcription nor accumulation was observed in mature gametocytes (Fig. 3D). It is not yet possible to distinguish between 16-hpi trophozoites, which will form schizonts, and those committed to sexual differentiation.
Transcription of the Nuclear rRNA Gene Units during Sexual The blot was stripped between each hybridization. Schematics of the parasite forms are shown above the appropriate lane of the blot, and the age of the ookinete forms is indicated. Sz, schizont; G, gametocyte; NP, gametocyte nonproducer; P, gametocyte producer. B, in situ hybridization of ETS probes to developing ookinetes from asynchronous cultures of mixed parasite forms. A, top row, detection of A-type transcription by dual hybridization of developing zygotes (arrow) and ookinetes to a P. berghei SSU probe (directly labeled with fluorescein-conjugated UTP) and to an A-unit ETS antisense RNA probe (labeled with digoxigenin-UTP and detected with a rhodamine-conjugated anti-digoxigenin antibody; see "Materials and Methods"). The multinuclear hybridizing forms are multiply-invaded cells that occur frequently in this type of ookinete culture and give rise to the A-type transcription signal seen in the Northern analysis of 22-h ookinete RNA. They also provide a positive control for the hybridization. S, middle row, onset of S-type transcription. Dual hybridization of developing zygotes (arrow) and ookinetes to a P. berghei SSU probe (fluorescein-labeled as above) and to a combination of C-unit ETS and internal transcribed spacer 1 antisense RNA probes labeled with digoxigenin-UTP and detected with rhodamine-conjugated anti-digoxigenin antibody. Uninuclear (1n) cells also hybridize to the RNA probes and are rings/trophozoites. Bottom row, the schematics represent (from the left) a zygote (4 h), the retort form (8 h), and the mature ookinete (20 -24 h) forms.
gametocyte and a trophozoite (T). C, Northern analysis of S-type transcription. The same blot used for the previous hybridizations was sequentially hybridized to the S-type ETS probe, 99S, and to the S-type specific oligonucleotide TM3 (data not shown). Development and Zygote Development-Using the ETS-specific probes, A-type and S-type transcription was examined in mature gametocytes, zygotes, and developing ookinetes (Fig. 4). A-type precursors were still abundant in mature female gametocytes (Fig. 4A, right panel). After fertilization, it appeared that transcription of the A-type units rapidly decreased. In the ookinete preparations, which were derived from purified gametocytes, A-type precursors were hardly visible at 2 and 14 h after fertilization. The strong hybridization at 22 h is due to the presence in this particular preparation of ookinetes of large numbers of asexual ring forms that produce significant amounts of A-type precursors (Fig. 2B). In situ hybridization analysis confirmed that A-type transcription was down-regulated in zygotes and ookinetes during this period and that the nuclear A-type signal resulted solely from asexual parasites present in the culture (Fig. 4B).
In contrast to the rapid decrease in A-type transcription, S-type transcription was up-regulated during ookinete development (Fig. 4A, left and center panels). In mature gametocytes, little if any S-type transcription was seen. S-type transcription could be detected 2 h after fertilization and dramatically increased between 14 and 24 h in the maturing ookinete. Four S-type rRNA precursor bands were observed in ookinetes, the two observed in asexual parasites (3000 and 2650 nt) and two larger forms (3900 and 3400 nt). The 2650-nt precursor was more abundant than the 3000-nt transcript in contrast to the transcription of the S-type units in the bloodstages, where the relative abundance of these two precursors was reversed (Fig. 2C). In situ hybridization analysis confirmed that the S-type transcription could be detected in the nucleus of about 10% of 20-h ookinetes and in Ͼ70% of mature (Ͼ24-h) ookinetes. S-type transcription was not detected with this method in 2-h zygotes or in 4-, 6-, 12-, and 16-h ookinetes ( Fig. 4B and data not shown). DISCUSSION The pattern of rRNA gene transcription in Plasmodium parasites is complex. This is due not only to the presence of different types of rRNA gene units (A-and S-type) that are transcribed in a stage-specific manner (2-5) but also the presence of multiple and unlinked copies of each type. Each individual unit may be differentially transcribed during the pro-gression of the life cycle, complicating the analysis. The pattern of transcription of the rRNA genes can be established relatively easily in the rodent model malaria parasite, P. berghei, because there are only two copies each of the A-type (the A and B units) and the S-type (the C and D units) (8), and the types can be defined by the ETS probes described here. These probes also further define the extent of pairwise homogeneity between the units. The mechanism that underlies the maintenance of this pairwise homogeneity of units on separate chromosomes involves some level of gene conversion (25), yet it must also discriminate between and maintain the different types.
In Plasmodium, rapid growth periods result in the formation of relatively stable parasite forms (mature schizonts/merozoites, gametocytes, and sporozoites) that only undergo further differentiation after a specific transition, e.g. erythrocyte invasion or passage between the host and vector. The work presented here demonstrates that rRNA transcription in two of these stable forms, schizonts and male gametocytes, is downregulated and that only the female gametocyte continued to transcribe the rRNA genes. The female gametocyte can accumulate maternal mRNA that is not translated until after fertilization (e.g. Pbs21 ookinete antigen gene (10,21,22)) and may, therefore, require protracted ribosome biogenesis during its development and may also accumulate ribosomes for use in the zygote, given the absence of appreciable rRNA transcription in P. berghei at that stage. Electron microscopic examination of mature male gametocytes suggested that ribosomes were undergoing visible degradation, which would be consistent with their cessation of rRNA gene transcription (26,27). The steady state pattern of A-type rRNA precursors was complex and probably cell type-specific, implying that the female gametocyte uses a different rRNA processing pathway compared to asexual blood-stage forms, even if the difference only reflects the kinetic stability of the intermediates. Alternatively, the range of rRNA precursors may reflect two specific forms of A-type transcript, one each produced by the A and B units. The currently available probes cannot discriminate between the units, which appear to be highly similar (8).
In most organisms, transcription of rRNA genes occurs during the growth phase of the cell (28). This work has shown that FIG. 5. Schematic representation of the pattern of transcription of the A-and S-type rRNA units. The transcription profiles of the Aand S-type rRNA units of P. berghei are indicated with reference to the stage of development of the parasite. This progression begins from asexual blood-stage forms through gametocytogenesis, gamete formation, fertilization, and subsequent development to ookinetes, and the illustrated rRNA transcriptional activity results from the work described in this study. The up-regulated transcription of the S-type units in the oocyst and sporozoite mosquito-borne stages is taken from Gunderson et al. (2) and will be described in more detail. 7 The switch from S-to A-type transcription occurs in the hepatocyte in the rodent host and is taken from Zhu et al. (7). The transcription is indicated in a purely qualitative fashion and is not intended to imply relative quantitation. Schematics of the various parasite life forms are given at the appropriate point in the life cycle progression.
transcription of the A-type rRNA genes of P. berghei is maximal in the rapidly growing blood-stage forms and, therefore, under a similar type of control. In addition, we have shown that at least one of the S-type units, normally mainly transcribed during oocyst growth (2), is also transcribed during the bloodstage development in a small subset of P. berghei trophozoites and to a greater extent in parasite clones that produce gametocytes. Bloodstage transcription of the S-type units may indicate a temporary requirement for S-type ribosomes during blood-stage development, implying the existence of a specific mRNA population or cell type. This subset of cells may be committed to sexual development, but in the absence of additional early markers of gametocyte development, this cannot be confirmed. Unlike the observations made in P. falciparum (4), however, the P. berghei blood-stage S-type transcription did not result in the accumulation of precursor forms in gametocytes; in addition, there was no evidence of mature S-type SSU rRNA in any blood-stage form. This species-specific S-type rRNA gene transcription may well reflect the marked difference in the kinetics of development and morphology of P. falciparum gametocytes (26), which take a minimum of 8 days to mature, whereas in other malaria species, gametocytogenesis is of a similar length to asexual blood-stage development. Alternatively, it is possible we observed a general transcriptional event that is so transient that at any time point of analysis, transcription appears to be limited to a subset of the synchronized populations that have a developmental window of about 4 h. The developmental timing of the switch between A-and S-type transcription in the mosquito also differs in P. vivax, where there is a lag phase between days 2 and 6 after infection where neither class of rRNA is detectable (5).
The recent report of a third type of developmentally regulated nuclear rRNA gene transcribed during this lag phase in oocysts of P. vivax 1 raises the question of whether a similar type exists in P. berghei. This could only result from a subdivision of the very similar but distinguishable S-type C and D units because we have demonstrated that the A types are barely transcribed throughout the mosquito stages and are almost identical (2). Characterization of the differences between the C and D units will allow this question to be addressed.
In conclusion, these studies demonstrate that transcription of the stage-specific rRNA genes of P. berghei is highly regulated and involves different periods of activation demonstrably linked, in the case of the A-type, to periods of cell growth. In addition to this general conventional pattern, it was observed that switching of transcriptional activity between the two different rRNA gene types was linked to the various differentiation pathways of the parasites. Superimposed upon this general pattern of transcription were transient bursts of rRNA transcription, which may yet prove to play a role in parasite differentiation (see Fig. 5 for summary of known transcription patterns of rRNA throughout the life cycle of P. berghei). The general pattern is directly analogous to the transcriptional activity of rRNA genes found in the well studied eukaryotic models. Typically, transcriptional regulation or rRNA units in these models results from different types of competitive associations of both transcriptional promoter (e.g. SL1, UBF, TBP, and REB1) and repressor proteins (e.g. Rb) with both enhancer and promoter DNA elements (29,30). These associations can be affected by the phosphorylation status of protein factors (31). Additionally, the status of the chromatin and the phasing of nucleosomes in the region of the rDNA units can influence transcription (32,33). All of these mechanisms may play a role in the regulation of the individual rDNA units of Plasmodium, and their interplay may then control the relative expression of each unit. The work described here implies that the general transcription of the rDNA units of Plasmodium will, however, be regulated in a similar conserved fashion to those of other eukaryotes, and the particular interest will lie in examining the supplementary mechanisms that effect the observed stagespecific control of rDNA transcription.