1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References
  8. Supporting Information

Malaria's cycle of infection requires parasite transmission between a mosquito vector and a mammalian host. We here demonstrate that the Plasmodium yoelii Pumilio-FBF family member Puf2 allows the sporozoite to remain infectious in the mosquito salivary glands while awaiting transmission. Puf2 mediates this solely through its RNA-binding domain (RBD) likely by stabilizing or hastening the degradation of specific mRNAs. Puf2 traffics to sporozoite cytosolic granules, which are negative for several markers of stress granules and P-bodies, and disappear rapidly after infection of hepatocytes. In contrast to previously described Plasmodium bergheipbpuf2 parasites, pypuf2 sporozoites have no apparent defect in host infection when tested early in salivary gland residence, but become progressively non-infectious and prematurely transform into EEFs during prolonged salivary gland residence. The premature overexpression of Puf2 in oocysts causes striking deregulation of sporozoite maturation and infectivity while extension of Puf2 expression in liverstages causes no defect, suggesting that the presence of Puf2 alone is not sufficient for its functions. Finally, by conducting the first comparative RNA-seq analysis of Plasmodium sporozoites, we find that Puf2 may play a role in directly or indirectly maintaining the homeostasis of specific transcripts. These findings uncover requirements for maintaining a window of opportunity for the malaria parasite to accommodate the unpredictable moment of transmission from mosquito to mammalian host.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References
  8. Supporting Information

Malarial infections in humans are initiated by the transmission of five species of Plasmodium parasites through the bite of infected Anopheles mosquitoes. The resulting disease affects over 200 million people annually, and causes over half a million deaths per year (WHO, 2011). Understanding the requirements for efficient parasite transmission is important in order to identify new potential targets for intervention (reviewed in Kappe et al., 2010).

To this end, work has been undertaken to elucidate how Plasmodium parasites develop within, and are transmitted to and from the female Anopheles mosquito (reviewed in Aly et al., 2009). Briefly, the mosquito takes up the sexual stages of the parasite from an infected host's bloodstream, and in the mosquito's midgut the male and female gametes fuse to form a zygote. The zygote then becomes a motile ookinete form, and invades the midgut epithelial wall to form a sessile oocyst adjacent to the basal lamina. Within each oocyst, the parasite undergoes sporogony to produce thousands of sporozoites. These sporozoites then emerge into the haemocoel of the mosquito and upon encountering the salivary glands, actively colonize them and become highly infectious. The actual event of transmission from the mosquito vector into a new mammalian host occurs when the mosquito probes for another blood meal, and concomitantly deposits its anti-coagulant saliva (along with sporozoites) in the skin of the host. This transmission event, along with the resulting migration through the skin, the invasion of the vasculature, and ultimately the infection of the liver has recently been reviewed (Lindner et al., 2012).

The parasite's transition from vector to mammalian host challenges the parasite with an entirely new environment, and it appears that the parasite prepares for this while still in the mosquito salivary glands. Early characterizations of mRNAs that are differentially expressed in oocyst sporozoites versus highly infectious salivary gland sporozoites identified 124 significantly upregulated transcripts in salivary gland sporozoites, which have been termed the upregulated in infectious sporozoites (UIS) transcripts (Matuschewski et al., 2002; Mikolajczak et al., 2008a). Deletion of several UIS genes (e.g. UIS3, UIS4, P52) showed that UIS play a critical role in establishing the parasites intracellular niche in hepatocyte infection (Mueller et al., 2005a,b; Labaied et al., 2007; VanBuskirk et al., 2009). Deletion of UIS genes severely attenuates the parasite, rendering it unable to develop as liver stages (also called exoerythrocytic forms, EEFs). Some UIS deletion strains have been successfully used as live-attenuated sporozoites for vaccination and indeed confer complete sterile protection against infectious sporozoite challenge in mice (reviewed in Vaughan et al., 2010). Interestingly, recent proteomics efforts have determined that many UIS transcripts appear not to be translated in sporozoites and therefore might be translationally repressed until the productive infection of hepatocytes (Lasonder et al., 2008; Lindner and Swearingen et al., in press). Further clues that translational repression and protection of transcripts are important to sporozoite transmission came from studies with pysap1 parasites, which showed that several UIS transcripts necessary for early liver stage infection (e.g. UIS2, UIS3, UIS4) were specifically degraded in the absence of SAP1 (Aly et al., 2011), rendering the parasite completely non-infectious to mice. Additional evidence for the role of translational repression in sporozoites is also apparent from previous transcriptomic and proteomic analyses (Le Roch et al., 2004; Hall et al., 2005) (Lindner and Swearingen et al., in press). Consistent with this, many groups have implicated translational repression as an important facet of the host/vector transmission event, as translation of specific mRNAs present in unactivated gametocytes only occurs after activation (cdc20; Guttery et al., 2012) or upon zygote fertilization (p25, p28; Paton et al., 1993). Additionally, studies of the Ddx6 RNA helicase DOZI and the CITH protein have elegantly demonstrated their role in repressing the translation of p25 and p28, as well as other transcripts present in gametocytes (Mair et al., 2006; 2010).

While Plasmodium has evolved stage-restricted specific transcription factors (e.g. ApiAP2 proteins) to drive its growth and development, its use of translational repression to overcome the challenges presented by its transmission strategy is similarly adaptive, as it allows differences in external stimuli to immediately trigger the start of the parasite's stage transitions by the translation of repressed transcripts (Painter et al., 2011). Many of the mechanisms of translational repression have recently been elucidated in model organisms (e.g. Saccharomyces cerevisiae, Drosophila melanogaster, Caenorhabditis elegans), and have been reviewed in detail (Balagopal and Parker, 2009). Most of the key proteins of these processes, including members of the Puf family of proteins, traffic to cytosolic granules [e.g. stress granules and/or Processing bodies (P-bodies)] in order to silence and protect targeted mRNAs, or alternatively, to actively hasten mRNA degradation. Recent work has demonstrated that these mechanisms are likely conserved in Plasmodium species as well. For instance, as mentioned above, UIS transcripts are specifically protected from degradation in the sporozoite by SAP1 (Aly et al., 2011). SAP1 was also found in cytosolic granules in the sporozoite, which taken together with its function is consistent with a putative role of inhibiting the degradation machinery typically found within P-bodies. Furthermore, studies by Gomes-Santos et al. and Muller et al. have shown a role for Puf2 in the control of parasite dedifferentiation from a sporozoite to a EEF (Gomes-Santos et al., 2011; Muller et al., 2011). The Puf family of proteins is commonly implicated in the control of metazoan development through the sequence specific binding of target RNAs near or in their 3′ untranslated region (3′ UTR), leading to silencing/protecting transcripts and/or targeting transcripts for degradation (reviewed in Wickens et al., 2002; Quenault et al., 2011). Puf proteins all contain a canonical RNA-binding domain (RBD) comprised of eight RNA recognition helices structurally arranged in a concave arc, inside of which sits the bound RNA molecule (Wang et al., 2002). Each recognition helix typically uses three amino acids to interact with a single ribonucleotide specifically by base stacking and hydrogen bonding mechanisms (Qiu et al., 2012). While members of the Puf family all have similar modular RBDs, their binding sequence preferences, and thus their mRNA targets, differ greatly between family members but are evolutionarily conserved among orthologues. Indeed, recombinant Plasmodium falciparum Puf2 RBD was able to specifically bind the Nanos Response Element (NRE) that D. melanogaster PUM binds (Fan et al., 2004). Despite this, the in silico identification of transcripts that contain the highly A-T rich NRE consensus binding sequence (UGUAAA/C/UAU) that might be bound by Puf2 in vivo has proven difficult due to the high A–T content of Plasmodium genomes (Gomes-Santos et al., 2011). Many transcripts have been shown by RT-PCR and microarray analyses to be affected by the genetic disruption of Plasmodium berghei Puf2, but it remains unclear which (if any) of these are direct effects of the lack of Puf2, or are indirect effects due to the proclivity of these parasites to prematurely transform into EEFs (Gomes-Santos et al., 2011).

In order to determine the role of Puf2 in regulating successful transmission, we sought to uncover the attributes of Puf2 that are necessary to produce and maintain the infectivity of sporozoites. We used reverse genetics of the rodent-infective Plasmodium yoelii parasite to delete, or modify the Puf2 ORF or perturb its endogenous expression. We found that in contrast to the knockout of P. berghei Puf2 (pbpuf2), PyPuf2 is dispensable for the infectivity of sporozoites that have colonized the salivary glands at early time points of mosquito infection. However, with prolonged salivary gland residence, sporozoite infectivity dropped precipitously and ceased within eight days, which occurred prior to the extensive premature transformation into EEFs. Immunofluorescence microscopy indicated that Puf2 traffics to cytosolic granules that do not colocalize with canonical P-body (XRN1) or stress granule (eIF2α) markers in sporozoites. After productive infection of a hepatocyte, Puf2 storage granules rapidly dissolve during dedifferentiation of the parasite from an invasive form into a trophic form in wild-type (WT) parasites. Moreover, expression of Puf2's RBD alone is necessary and sufficient for all essential Puf2 functions, and parasites that express only the RBD retain normal infectivity, traffic to storage granules, and exhibit no phenotypic defects. In addition, by replacing the endogenous Puf2 promoter with that of circumsporozoite protein (CSP) or UIS4, we have found that the timing of Puf2 expression is critical: as premature expression is detrimental but extended expression has no significant effect upon the parasite. Finally, by conducting the first RNA-seq analyses of Plasmodium sporozoites, we have compared WT and pypuf2 sporozoites. Significant differences in the abundances of transcripts specific to a subset of UIS genes and to RNA metabolic processes might be directly or indirectly caused by the lack of Puf2. Taken together, we present a model of Puf2's role in maintaining RNA homeostasis in sporozoites to preserve and extend infectivity.

Results and discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References
  8. Supporting Information

Puf2 is essential only for maintaining sporozoite infectivity during protracted salivary gland residence

Translational repression of specific transcripts is an important regulatory process in Plasmodium parasites. For instance, the DOZI and CITH proteins are essential translational repressors during the fusion of male and female gametes to form a zygote, as their deletion disrupts parasite development at this point (Mair et al., 2006; 2010). Thus, the identification of additional translational repressors that are essential to the development and transmission of Plasmodium would be of interest. Puf2, a putative translational repressor based upon its family's characterized functions in other species, was found to be transcriptionally upregulated and present in infectious sporozoite proteomes (Mikolajczak et al., 2008a; Lindner and Swearingen et al., in press). Work on Puf2 in P. falciparum indicated that Puf2 plays a role in gametocytogenesis but its role in sporozoite biology was not analysed (Miao et al., 2010). Studies on Puf2 in P. berghei showed a role in regulating the transformation of salivary gland sporozoites into early EEFs (Gomes-Santos et al., 2011; Muller et al., 2011). We sought to characterize Puf2's role in the vector/host transmission event in the rodent-infective malaria species P. yoelii, because its developmental conditions in the mosquito more closely align with that of P. falciparum, sharing both an optimal temperature (24°C) and developmental duration period in the mosquito midgut (14–16 days).

We found that the currently annotated gene structure of PyPuf2 (ID: PY04369, available at, v9.1; Carlton et al., 2002; Vaughan et al., 2008) differed from that of all other Plasmodium species. Experimental determination of the actual PyPuf2 exon/intron boundaries by RT-PCR and sequencing (data not shown) matched that of its orthologues (Fig. S1A) and with the recently released predicted gene structure in the P. yoelii YM strain (data not shown). Using this re-annotated gene structure, we generated a transgenic knockout parasite (pypuf2) by double-cross-over recombination whereby the entire ORF of PyPuf2 was replaced with a plasmid bearing a human dihydrofolate reductase (HsDHFR)-expression cassette. Successful replacement of this locus in two independent clones was confirmed by genotyping PCR across both recombination regions (Fig. S1B). To assess Puf2's role throughout the parasite life cycle, A. stephensi mosquitoes were infected with WT Py17XNL or pypuf2 P. yoelii parasites. We observed no significant change in the male : female gametocyte ratio present in the infected mouse in contrast to what was previously reported in P. berghei and P. falciparum (data not shown), nor did we observe a difference in the number of pypuf2 oocysts, oocyst sporozoites or the number of salivary gland sporozoites present in the mosquito when compared to WT. We next sought to assess Puf2's role in regulating sporozoite functions after salivary gland infection. To this end, salivary gland sporozoites were isolated from infected mosquitoes early after their arrival in the salivary glands (14 days post infection) or at two-day increments thereafter (16, 18, 20 or 22 days post infection). Sporozoite infectivity was assessed by measuring the time required to the onset of blood stage patency following sporozoite injection into naïve BALB/cJ mice (Fig. 1). WT sporozoites isolated at time points 14 and 22 days completed liver stage of development and caused patent blood stage parasitemia within 3 days. Even WT sporozoites isolated as late as 30 days post infection of the mosquito induced patent parasitemia within 3 days (Table 1). Strikingly, pypuf2 sporozoites progressively lost infectivity after arrival in the salivary glands, becoming non-infectious within the following 8 days as determined by this patency assay [Logrank test, P (one-sided) = 0.001234, χ2 = 17.98, d.f. = 4]. Loss of infectivity was approximately 10-fold per day, as a one-day delay in the time to blood stage patency roughly correlates with a 10-fold reduction in infectivity (Gantt et al., 1998).


Figure 1. pypuf2 sporozoites lose infectivity during prolonged salivary gland residence. Wild-type (WT) and pypuf2 salivary gland sporozoites were isolated from mosquitoes at the indicated times post blood meal, and 10 000 sporozoites were injected intravenously (iv) into four mice per group. The time to blood stage patency (defined as > 1 parasite per 20 000 RBCs) was determined microscopically by daily Giemsa-stained thin blood smears and was plotted as a Kaplan–Meyer survival curve. WT parasites produced blood stage patency 3 days post infection at all time points between t = 14 and 22 days, as well as at 30 days. Logrank test of statistical significance of the loss of infectivity over time: P (one-sided) = 0.001234, χ2 = 17.98, d.f. = 4. Hazard ratios: 14 versus 16 days = 2.0; 14 versus 18 days = 3.1; 14 versus 20 days = 9.2.

Download figure to PowerPoint

Table 1. Measurements of the time to blood stage patency indicate the necessity and sufficiency of Puf2's RNA-binding domain for infectivity
Parasite genotypeNo. of days post blood mealNo. of sporozoites injectedNo. of mice infectedNo. of mice patent (days of patency)
  1. Salivary gland sporozoites were isolated from mosquitoes at 14, 22 and 30 days post blood meal and injected iv into four BALB/cJ mice per group in duplicate. Time to blood stage patency (defined as > 1 parasite per 20 000 RBCs) was determined microscopically by daily Giemsa-stained thin blood smears for up to 14 days post injection. Data for pypuf2 parasites were not determined (N/D) due to the nearly complete premature transformation of sporozoites into early exoerythrocytic forms.

Wild-type1410 00044 (3)
2210 00044 (3)
3010 00044 (3)
pypuf21410 00044 (3)
2210 00040 (N/A)
Puf2-RBDmyc1410 00044 (3)
2210 00044 (3)
3010 00044 (3)
UIS4prom:Puf21410 00044 (3)
2210 00044 (3)

However, in contrast to pbpuf2 sporozoites, which exhibited a defect in infectivity at all time points examined (Muller et al., 2011), pypuf2 sporozoites were as infectious as WT parasites when taken early after salivary gland invasion (day 14), inducing patency within 3 days (Fig. 1). The knockout genotype of the resulting blood stage parasites was confirmed by PCR to ensure trace amounts of WT parasites were not contaminating the population (Fig. S1C). This difference in the infectivity phenotype between the two species is perhaps explained by the difference in sporozoite development and their peak arrival time in the salivary glands, as P. berghei sporozoites arrive later than do P. yoelii sporozoites (18 versus 14 days respectively). It is tempting to speculate that if pbpuf2 sporozoites were isolated at earlier time points, these parasites might also behave similarly to WT parasites. Our finding that Puf2 is dispensable for sporozoite infectivity early after salivary gland invasion provides important clues about its function, as its role appears not to regulate infectivity per se but to maintain infectivity during prolonged salivary gland residence perhaps by modulating the homeostasis of RNAs present in the salivary gland sporozoite (see below).

Loss of pypuf2 sporozoite infectivity precedes premature transformation

Malaria parasites have evolved to adapt to the mosquito vectors' blood feeding behaviour for their transmission between vertebrate hosts, and therefore out of necessity must have made provisions to adapt to the unpredictable timing at which mosquitoes have the opportunity for a blood meal. We and others have demonstrated that sporozoites can remain in the salivary glands for over 2 weeks and maintain their full infectivity (Table 1, Porter et al., 1954; Muller et al., 2011). As pbpuf2 parasites were previously shown to both lose their infectivity over time and prematurely transform into a non-infectious trophic EEFs while inside the salivary glands, we undertook a similar analysis with pypuf2 parasites. While very little premature transformation of P. yoelii WT sporozoites was observed even 30 days post mosquito infection, a significant and progressively increasing fraction of pypuf2 sporozoites prematurely transformed [Wilcoxon rank sum test, WT versus pypuf2, P (one-sided) < 0.025]. Starting 18 days post mosquito infection, premature transformation significantly increased by approximately 20% during each of the subsequent 2-day increments tested [Fig. 2, Jonckheere–Terpstra Test, P (one-sided) = 4.55 × 10−9, W* = 5.747]. This progressively increasing number of prematurely transforming sporozoites was comparable to what was reported with pbpuf2 parasites (Gomes-Santos et al., 2011; Muller et al., 2011). However, if matched with the loss of infectivity data (Fig. 1), we observed that the loss of infectivity in fact preceded the premature transformation of sporozoites and is most clearly demonstrated with 22-day sporozoites, as nearly half of the parasites still retained sporozoite-like morphology but were non-infectious to mice (Figs 1 and 2). This is an important point, as premature transformation of an invasive sporozoite into a trophic EEF could account for the inability of such a parasite to productively infect a new host. However, as the loss of infectivity and occurrence of premature transformation appeared uncoupled, the lack of Puf2 expression in the sporozoite likely leads to the loss or gain of specific mRNA populations that adversely affect infectivity first, with the induction of transformation as a downstream event.


Figure 2. Progressive premature transformation of sporozoites of pypuf2 parasites is completely rescued by Puf2-RBDmyc expression. Salivary gland sporozoites (sg-spz) were isolated from mosquitoes at the indicated times post blood meal, and were assessed by IFA (anti-PyCSP (clone 2F6) and DAPI) for indications of premature transformation into exoerythrocytic forms (EEFs). Transformation was defined as any exhibition of a medial bulge around the nucleus, deflation of sporozoites ends and/or rounding up of the parasite. Representative images of the differently scored forms are provided at the right of the graph. Observations of 500 parasites were made in triplicate. Scale bar is 5 μm. Jonckheere–Terpstra test of statistical significance of progressive premature transformation of pypuf2 parasites: P (one-sided) = 4.55 × 10−9, W* = 5.747. Wilcoxon rank sum test of statistic difference in progressive premature transformation between WT or Puf2-RBDmyc versus pypuf2 parasites at 22 or 30 days: P (one-sided) < 0.025.

Download figure to PowerPoint

Next, we assessed whether pypuf2 sporozoites simply have a higher proclivity to transform into EEF by using either WT or pypuf2 sporozoites that had recently invaded the mosquito salivary glands (14 days) and placing them in axenic culture (Kaiser et al., 2003). We measured no significant difference in the axenic transformation potential between WT and pypuf2 sporozoites (24.4% ± 0.916 and 24.1% ± 0.597 respectively) (Table S1). This finding stands in contrast to pbpuf2 sporozoites (18 days), which were more prone to transform axenically when compared to WT sporozoites (Gomes-Santos et al., 2011).

The RNA-binding domain of Puf2 is necessary and sufficient for its functions

Having observed strong phenotypes associated with prolonged salivary gland residence of pypuf2 sporozoites, we sought to use this phenotypic profile to genetically dissect the functionally important domains of Puf2. The predominant feature of Puf2 is its RBD, which is comprised of eight canonical RNA recognition helices and spans all four exons (AA132-478, Fig. S1). The N-terminal portion of Puf2 (AA1-131) has no predicted PFAM domains, nor any significant similarity/identity to non-orthologous proteins. To test what functions, if any, the RBD alone provided, we produced a transgenic parasite that lacked the N-terminal domain by fusing the Puf2 5′ UTR (including its start codon) to the predicted RBD (Fig. S2A). As a full-length variant of Puf2 could accommodate a C-terminal epitope tag without affecting its localization or function (data not shown), we also appended a C-terminal 4× Myc tag to the Puf2 RBD. Generation of transgenic parasites (Puf2-RBDmyc) was confirmed by genotyping PCR (Fig. S2B), and Puf2-RBDmyc was used to infect mosquitoes. As with the pypuf2 parasites, no decreases in the number of oocysts, oocyst sporozoites or salivary gland sporozoites were observed with Puf2-RBDmyc (data not shown). Salivary gland sporozoites were isolated from mosquitoes either shortly after salivary gland invasion (14 days), or 8 (22 days) or 16 days later (30 days) and their infectivity and proclivity to transform prematurely were assessed as before. Interestingly, we found that the RBD alone was sufficient to completely maintain infectivity and the ability of the parasite to maintain its sporozoite morphology, phenocopying WT parasites at all time points tested [Fig. 2, Wilcoxon rank sum test, Puf2-RBD versus pypuf2 P (one-sided) < 0.025]. Moreover, sporozoites expressing only Puf2-RBDmyc were able to preserve their infectivity even 30 days post mosquito infection (Table 1). The genetic identity of the blood stage infections induced by 14-day sporozoites was confirmed by genotyping PCR, indicating that the phenotypic rescues were in fact attributable to expression of the RBD alone (Fig. S2C). Taken together, these data indicate that the RBD of Puf2 is necessary and sufficient to fulfil all of the functions required to maintain sporozoite infectivity and its cellular organization. Furthermore, if any essential interactions occur between Puf2 and other unknown proteins, they necessarily must occur via its RBD. This assertion is consistent with observations on Puf family orthologues from S. cerevisiae, C. elegans and H. sapiens, showing that the RBD alone is sufficient for all repressive activities (Goldstrohm et al., 2006). Further evidence has shown that the RBDs of these proteins are sufficient to make protein-protein interactions with other regulatory proteins (such as CPEB family members and Pop2p) via their RBDs using extended loops and the opposing face of this domain (Goldstrohm et al., 2006; Campbell et al., 2012). This might hold true for P. yoelii Puf2, and it will be important to determine what proteins interact and partner with Puf2 either by binding its external convex face, by binding the Puf2/RNA interface (thus providing for a stepwise assembly program), or indirectly through other factors.

Puf2 forms discrete cytosolic storage granules in salivary gland sporozoites that lack markers of stress granules and P-bodies

Previous work indicated that PbPuf2 is expressed in salivary gland sporozoites as part of granules that are perinuclear but also distributed throughout the cytoplasm of the sporozoite (Gomes-Santos et al., 2011). Using anti-PbPuf2 antisera to stain WT P. yoelii sporozoites, we were able to detect weak Puf2 expression in oocysts and oocyst sporozoites, which increased in salivary gland sporozoites and exhibited similar granular patterns as described for P. berghei (Gomes-Santos et al., 2011; Fig. S3A, Fig. 3). We next used a C-terminally appended 4× Myc epitope tag to localize Puf2-RBDmyc within the salivary gland sporozoite and observed similar granules (Fig. 3). Due to the higher specificity of monoclonal antibodies to c-myc, the resolution of Puf2 structures was greatly improved, and comparison of its localization with other proteins was achieved. We have previously shown that SAP1 is also contained within cytoplasmic granular structures in the salivary gland sporozoite, and plays a critical role in inhibiting the degradation of UIS transcripts (e.g. UIS2) that are important for early liver stage development (Aly et al., 2011). SAP1 localization and its role in either the direct or indirect inhibition of transcript degradation, most likely places it in P-bodies, which are granules that contain the canonical mRNA degradation machinery. By co-staining Puf2-RBDmyc salivary gland sporozoites with PySAP1-specific antisera and anti-c-myc antibodies, we observed that these two types of granules constitute distinct populations in sporozoites (Fig. 3A). In agreement with this, co-staining of Puf2-RBDmyc salivary gland sporozoites with anti-c-myc and antibodies raised against Toxoplasma gondii XRN1 (a key exonuclease in mRNA turnover found in P-bodies) also showed mutually exclusive staining patterns (Fig. 3B) (Jones et al., 2012). Another cytoplasmic granule commonly found in eukaryotes is the stress granule, which translationally represses mRNAs in response to various stress stimuli (Wek et al., 2006). The eukaryotic translation initiation factor eIF2α is commonly found in stress granules, and thus can serve as a marker. Using antisera raised against either total eIF2α from T. gondii, or against a peptide from the same protein that contains the phosphoserine regulatory mark (Ser 51), we also observed that the 4× Myc-tagged Puf2-RBD did not colocalize with eIF2α (Fig. 3C and D). Taken together, Puf2 localized to distinct cytoplasmic granules that lack these common P-body or stress granule markers. It remains possible, however, that Puf2 might traffic to a subset of P-bodies and/or stress granules that do not contain these components. Until evidence arises to place Puf2 in an already characterized granule type, we propose calling these Puf2 containing structures ‘storage granules’ due to the role of Puf proteins in other species to bind, store and protect specific mRNAs (Wickens et al., 2002; Anderson and Kedersha, 2008).


Figure 3. Puf2 traffics to cytosolic storage granules that are distinct from stress granules and P-bodies. Representative IFA images are shown of a Puf2-RBDmyc salivary gland sporozoite stained with anti-c-myc (clone 9E10), DAPI and either anti-PySAP1 (A), anti-TgXRN1 (B), anti-Tg eIF2α (C) or anti-Tg eIF2α-P (D). Differential interference contrast (DIC) images demonstrate overall sporozoite morphology. A magnification of the merged images is provided to demonstrate the two distinct non-overlapping granule types in each case. Scale bar is 5 μm.

Download figure to PowerPoint

Puf2 storage granules dissolve immediately after hepatocyte invasion

Puf2 plays an important role in maintaining the infectivity of salivary gland sporozoites, and if deleted, sporozoites suffer loss of infectivity followed by premature transformation into EEFs. We predicted that after hepatocyte infection, endogenous Puf2 must release relevant transcripts from translational repression and/or abandon any role in hastening RNA degradation to allow transformation to trophic EEFs. We therefore analysed the expression and localization of Puf2 in intrahepatocytic, dedifferentiating EEFs. Mice were infected by iv injection with 3 × 106 WT or transgenic sporozoites that express Puf2-RBDmyc and were sacrificed 1.5 h later. The rapid dedifferentiation of early EEFs was observed by IFA using antibodies directed against PyCSP. Most parasites exhibit medial bulges and deflated/retracted ends, while some had completed the dedifferentiation process and had assumed the rounded trophozoite shape (Fig. 4A). Co-staining with anti-c-myc revealed extremely weak Puf2 expression in dedifferentiating parasites, and a lack of expression in rounded liver stage trophozoites, indicating that Puf2 is rapidly degraded after hepatocyte infection. In order to observe the progressive events of the dissolution process, we utilized an in vitro infection of HepG2 + CD81 cells. As with the in vivo experiment, WT or transgenic sporozoites that express Puf2-RBDmyc were allowed to infect HepG2 + CD81 cells and were analysed after 1.5, 3 or 6 h. Similar to what we observed in vivo, sporozoites that exhibited no or little dedifferentiation upon infecting hepatocytes retained Puf2 expression, whereas progressive dedifferentiation correlated with reduced levels of Puf2 until no Puf2 was detectable in liver stage trophozoites (Fig. 4B, left-to-right). The observations of this dedifferentiation process in P. yoelii show that it occurs with kinetics similar to what was described for P. berghei (Jayabalasingham et al., 2010). These data indicate that Puf2 is rapidly degraded in early EEFs, presumably allowing translation of the silenced transcripts or effects upon RNA homeostasis that are needed for early EEF development.


Figure 4. Puf2 storage granules dissolve during dedifferentiation immediately after hepatocyte invasion. Representative IFA images are shown of a Puf2-RBDmyc salivary gland sporozoite/EEF 1.5 h after in vivo infection of BALB/cJ mice by tail vein injection (A) or at 1.5, 3 or 6 h after in vitro infection of HepG2 + CD81 cells (B). Parasites were stained with anti-c-myc (clone 9E10), anti-PyCSP (clone 2F6) and DAPI. Specific Puf2 staining is rapidly lost during parasite dedifferentiation in both assays. Scale bar is 5 μm.

Download figure to PowerPoint

The temporal expression dynamics of Puf2 are critical to proper sporozoite development and salivary gland invasion, but not to liver stage development

Having delineated the temporal dynamics of Puf2 expression in WT salivary gland sporozoites and early EEFs, we next assessed if perturbation of endogenous Puf2 expression would positively or adversely affect the parasite at any point in its life cycle. To this end, we first introduced the 5′ UTR of the circumsporozoite protein (CSP) upstream of the Puf2 ORF into the endogenous locus (‘CSPprom:Puf2’, Fig. S4A) in order to increase Puf2 expression in oocysts and oocyst sporozoites. Clonal transgenic parasites were isolated by limiting dilution and the identity of the modified locus was confirmed by genotyping PCR (Fig. S4B).

To assess if the CSPprom:Puf2 parasites did in fact express more Puf2 transcript, we employed qRT-PCR with primers specific to Puf2, the fused CSPprom:Puf2 transcript, or 18S to normalize across samples. Specific amplification of the fused CSPprom:Puf2 transcript only occurred with cDNA generated from the transgenic oocysts (Fig. S4C). As expected, the transcript abundance of Puf2 in CSPprom:Puf2 transgenic oocysts increased significantly compared to WT parasites (3.1 ± 1.65-fold increase).

Phenotypic assessment of the CSPprom:Puf2 parasites compared to WT parasites revealed several stark differences that all indicated the importance of the proper timing and abundance of Puf2 expression (Table 2). First, while mosquitoes were infected with blood stage parasite populations with similar numbers of exflagellating microgametes (data not shown), there were slightly fewer CSPprom:Puf2 oocysts that developed in the mosquito midgut as compared to WT (Table 2). While there was no significant difference in the average maximum diameter of the oocysts, the time required to achieve maximum size was different (WT: 10 days, CSPprom:Puf2: 14 days). Moreover, there were far fewer CSPprom:Puf2 oocysts that exhibited evidence of sporozoite development [WT: 49.4% ± 1.78 versus CSPprom:Puf2: (clone 1) 10.7% ± 8.40, P (two-sided) = 0.027; (clone 2) 8.04% ± 4.71, P (two-sided) = 0.001], which was delayed as well. As might be expected, this translated into a marked reduction in the average peak number of oocyst sporozoites per mosquito [WT: 93 833 ± 14 180 versus CSPprom + Puf2: (clone 1) 22 250 ± 14 496, (clone 2) 15 500 ± 7778]. Most strikingly, no CSPprom:Puf2 sporozoites were observed to have invaded the salivary glands from 14 through 22 days post mosquito infection. Thus, introduction of the CSP promoter into the Puf2 locus demonstrates that the premature overexpression of Puf2 leads to severe deregulation of sporozoite development and phenotypic maturation. This is not attributable to the trans-use of the CSP 5′ UTR in addition to endogenous CSP, as transgenic parasites that use the CSP promoter to drive expression of fluorescent proteins have no observable effects on oocysts, oocyst sporozoites, or salivary gland sporozoites (Natarajan et al., 2001, S. Mikolajczak, unpubl. results).

Table 2. Overexpression of Puf2 causes severe deregulation of sporozoite development in the mosquito
Parasite genotypeNo. of days post blood mealNo. of oocysts/mosquito% Oocysts with evidence of sporozoitesOocyst diameter (μm)No. of oo-spz/mosquitoNo. of sg-spz/mosquito
  1. Midguts of infected mosquitoes (with a prevalence of infection between 60–100%) were isolated at 5, 10 and 14 days post blood meal and observed by light microscopy at 40× magnification to count total numbers of oocysts and to determine the percentage of oocysts with evidence of sporozoite development. The average diameter of oocysts (from at least 20 representative oocysts per condition) was measured by using the softWoRx software package with captured DIC microscopic images. Numbers of oocyst sporozoites (oo-spz) and salivary gland sporozoites (sg-spz) were counted on a haemocytometer. Calculations of statistical differences in the percent of oocysts with evidence of sporozoite development were determined by the student's t-test.

Wild-type580.6 ± 30.60 ± 030.1 ± 3.9N/DN/D
1096.3 ± 1.0249.4 ± 1.7859.9 ± 1.7093 933 ± 14 180N/D
1450.6 ± 27.31.84 ± 0.73834.6 ± 15.3N/D14 463 ± 6177
CSPprom:Puf2 Clone 1545.0 ± 27.50 ± 031.2 ± 4.23N/DN/D
1049.8 ± 38.514.3 ± 8.1541.2 ± 16.434 500 ± 8260N/D
1484.7 ± 44.110.8 ± 8.1552.3 ± 11.722 250 ± 14 4960 ± 0
CSPprom:Puf2 Clone 2552.0 ± 40.40 ± 033.6 ± 5.41N/DN/D
1054.2 ± 52.113.7 ± 0.04944.3 ± 8.2220 400 ± 6800N/D
1467.9 ± 47.87.82 ± 2.7455.7 ± 3.615 500 ± 77780 ± 0

Although WT oocyst sporozoites are approximately 1000- to 10 000-fold less infective than salivary gland sporozoites, we sought to determine if the CSPprom:Puf2 oocyst sporozoites retained any infectivity despite their inability to reach the salivary glands (Vanderberg, 1975). To this end, 105 WT or CSPprom:Puf2 oocyst sporozoites were iv injected into BALB/cJ mice, and the time to blood stage patency measured as before (Table 3). WT oocyst sporozoites were able to consistently produce a patent blood stage infection four days post injection, whereas CSPprom:Puf2 sporozoite injection did not lead to patent parasitemia when isolated after 10 or 14 days of development within the mosquito. In order to determine if these transgenic oocyst sporozoites were non-infectious or alternatively if they arrested in hepatocytes due to overexpression of Puf2, BALB/cJ mice were iv injected with 106 WT or CSPprom:Puf2 oocyst sporozoites and their livers were harvested at six or 18 h later. While WT parasites were readily observed with anti-PyCSP antibodies by IFA, no CSPprom:Puf2 parasites were detected (data not shown). This indicates that increased expression of Puf2 during oocyst development in CSPprom:Puf2 parasites leads to the complete loss of sporozoite infectivity, possibly due to an imbalance of the ratios of Puf2 and mRNAs it might bind or other regulatory effector proteins that might act upon it.

Table 3. Overexpression of Puf2 in oocysts results in non-infectious oocyst sporozoites
Parasite genotypeReplicate No.No. of sporozoites injectedNo. of mice infectedNo. of mice patent (days of patency)
  1. Oocyst sporozoites were isolated from mosquitoes at 10 days (wild-type, CSPprom:Puf2) or 14 days (CSPprom:Puf2) post blood meal and injected iv into four BALB/cJ mice per group in duplicate. Time to blood stage patency (defined as > 1 parasite per 20 000 RBCs) was determined microscopically by daily Giemsa-stained thin blood smears for up to 14 days post injection.

Wild-type (t = 10 days)1100 00044 (4)
2100 00044 (4)
CSPprom:Puf2 clone 1 (t = 10 days)1100 00040 (N/A)
2100 00040 (N/A)
CSPprom:Puf2 clone 1 (t = 14 days)1100 00040 (N/A)
2100 00040 (N/A)
CSPprom:Puf2 clone 2 (t = 10 days)1100 00040 (N/A)
2100 00040 (N/A)
CSPprom:Puf2 clone 2 (t = 14 days)1100 00040 (N/A)
2100 00040 (N/A)

We next sought to determine if the extension of Puf2 expression into the liver stage of development would be permissive or detrimental to the parasite. We tested this by similarly replacing Puf2's promoter with that of UIS4, which is expressed in both salivary gland sporozoites and throughout liver stage development (Mueller et al., 2005a). As before, clonal transgenic parasites (‘UIS4prom:Puf2’) were isolated by limiting dilution and their identity was confirmed by genotyping PCR (Fig. S5B). In contrast to the developmental defects observed with CSPprom:Puf2 parasites, no defect in salivary gland sporozoite numbers was observed with UIS4prom:Puf2 parasites. Surprisingly, upon infecting mice with these sporozoites, no delay in blood stage patency was detected, thus indicating that no defect in sporozoite infectivity or liver stage development was induced (Table 1). Examination of Puf2 protein expression by IFA revealed a punctate staining pattern in 14- and 22-day salivary gland sporozoites that matched the pattern observed with WT parasites (Fig. S5C). As intended, Puf2 expression persisted throughout liver stage development (Fig. S5C and D). These data indicate that rapid degradation of Puf2 in WT early liver stages is not necessary to ensure normal life cycle progression and that the presence or absence of Puf2 does not solely regulate its target transcripts. It is possible that critical additional factors, which themselves could be temporally regulated, might influence Puf2's ability to bind and/or regulate its mRNA targets, with Puf2 simply providing the selectivity for specific transcripts. Alternatively, if Puf2's target transcripts are translated and then degraded shortly after the initiation of the liver stage infection (and/or a similar fate occurs to additional regulatory effector proteins), the absence of these factors would also prevent Puf2 from acting upon them despite Puf2's extended expression.

Comparative RNA-seq reveals a role of Puf2 in maintaining homeostasis of specific sporozoite mRNAs

Puf family proteins function in several distinct capacities in other organisms, including translational repression of transcripts by inhibiting mRNA circularization and/or by targeting transcripts for deadenylation and degradation (reviewed in Miller and Olivas, 2011). As no canonical 4E-binding motifs are present in Puf2 that could facilitate direct binding with eIF4E to prevent mRNA circularization (data not shown), we sought to experimentally determine if Puf2 instead acts by stabilizing or increasing the degradation of specific transcripts in sporozoites. To accomplish this, we conducted the first comparative RNA-seq experiment with Plasmodium sporozoites. We chose to analyse RNA that was extracted from WT and pypuf2 sporozoites 14 days post mosquito blood meal, as both parasites are equally infectious to mice at this time (Table 1). Moreover, examination of this early time point should reduce indirect effects of transcript abundances that could arise due to the premature transformation of the sporozoite at later time points, such as those used for previous microarray experiments with pbpuf2 parasites (Gomes-Santos et al., 2011).

Transcripts were subjected to standard Illumina library creation and sequencing strategies, and high quality reads were mapped onto the P. yoelii YM strain reference sequence due to its superior quality but high similarity to that of the P. yoelii 17XNL strain. Paired end reads assigned to all annotated genes were normalized to account for differences in gene length, and then compared across WT and pypuf2 samples to detect transcripts with significantly affected abundances due to the absence of Puf2. Using stringent cut-off values (P < 0.01), we identified 513 transcripts that were upregulated and 129 transcripts that were downregulated in pypuf2 parasites versus WT parasites (Fig. 5A). A comparative bioinformatics search for the presence of Nanos Response Elements [‘NREs’, which can be bound by Puf2 in vitro (Fan et al., 2004)] between these affected transcripts and all globally annotated ORFs (with boundaries defined as 1.5 kb upstream of the start codon through 1.0 kb downstream of the stop codon) revealed of an over-representation of NREs in the affected transcripts (NREs are present in 31.4% of upregulated transcripts and 37.2% of downregulated transcripts versus 20.9% for global transcripts). Interestingly, we also detected a NRE in Puf2's own transcript, suggesting that Puf2 might autoregulate its own transcript levels.


Figure 5. Comparative RNA-seq implicates Puf2 in affecting RNA homeostasis. RNA from wild-type and pypuf2 salivary gland sporozoites (14 days) was subjected to a comparative RNA-seq experiment to identify significant differences in transcript abundances (P < 0.01).

A. Upregulated (513) and downregulated (129) transcripts are located at positions significantly away from the diagonal line on a scatter plot.

B. Upregulated transcripts that are relevant to RNA metabolism and downregulated UIS transcripts are listed along with their fold change in abundance (pypuf2 versus WT).

Download figure to PowerPoint

In agreement with previously published pbpuf2 microarray experiments, we observed similar trends for a subset of transcripts that were highlighted by those studies (Gomes-Santos et al., 2011), but found these effects were much more pronounced in our dataset (e.g. Exp1: +3.25× versus +19.4×; Exp2: +3.0× versus +29.9×; UIS1: −2.7× versus −7.46×; UIS2: −1.76× versus −5.46×; Puf1: −2.32× versus −8.90×; TLP: −4.77× versus −49.7×). In addition to our observations that match those from P. berghei, our data also reveal a novel role for Plasmodium Puf2 in affecting RNA metabolism, as several transcripts that encode functions involved in these processes (deadenylation, regulation of non-sense transcripts, exonuclease degradation, splicing) were significantly upregulated in the absence of Puf2 (Fig. 5B). Of particular note, members with homology to the CCR4/Not major deadenylase complex were highly upregulated (e.g. CCR4, Not2, Caf16). The differences could indicate that the parasite has either lost an upstream regulator of these transcripts, or alternatively, is attempting to compensate for the deregulation in RNA homeostasis caused by the lack of Puf2 by directly or indirectly increasing the expression of factors required to degrade and regain its preferred mRNA state, or both. These models are further supported by the increased abundance of several RNA-binding proteins, such as Bruno, ALBA3, and especially ALBA4, whose abundance was increased 295-fold in the absence of Puf2. These proteins could function to sequester or affect transcripts that are normally targeted by Puf2, or at least attempt to do so. Finally, we also observed that DOZI, a DDX6 RNA helicase implicated in critical translational repression processes during the host/vector transmission event (Mair et al., 2010), was also significantly upregulated in pypuf2 sporozoites. Genetic disruption of DOZI prevented parasite development beyond the zygote, and thus its potential role in the sporozoite was not previously appreciated. It would be of interest to assess if DOZI plays a similar role during sporozoite vector/host transmission.

Lastly, we also observed that a subset of UIS transcripts (13 of 124) were specifically downregulated in salivary gland sporozoites in the absence of Puf2. Three of these transcripts (UIS1, UIS8 and PY03183) harbour a pair of Nanos Response Elements downstream of the stop codon. Evidence for the translation of these proteins in salivary gland sporozoites was either weak (UIS1) or altogether absent (UIS8, PY03183) in proteomic analyses of this stage of the parasite (Florens et al., 2002; Hall et al., 2005; Lasonder et al., 2008; Lindner and Swearingen et al., in press). While our global comparative RNA-seq experiment did not directly implicate these transcripts as directly bound targets of Puf2, the described observations in combination suggest them as likely candidates that warrant further investigation.

A model of Puf2's role in sporozoite development and vector/host transmission

We propose a reasonable model of Puf2 function in sporozoite development, maintenance of infectivity and in early mammalian infection based upon all available data (Fig. 6). We observed by IFA and qRT-PCR that Puf2 expression is initiated between days 5 and 10 during oocyst development, and that Puf2 abundance continued to increase and peaks in salivary gland sporozoites. Immediately following the productive infection of hepatocytes, the sporozoite dedifferentiates into a liver stage trophozoite and Puf2 is progressively lost in this process. Thus, the precisely timed expression, levels of expression and degradation of Puf2 are well controlled in the WT parasite to help govern the malaria parasite's vector-to-host transition. The formation of Puf2 storage granules presumably stabilizes and/or hastens the degradation of liver infection-relevant mRNAs in salivary gland sporozoites directly or indirectly by affecting proteins that regulate RNA metabolism. Storage granule dissolution could then at least in part contribute to the release of specifically bound mRNAs for selective and immediate translation in early EEFs, although the degradation of Puf2 is not strictly necessary for this process as evidenced by the phenotype of the UIS4prom:Puf2 transgenic parasite. Puf2 expression dynamics appear to precede but then mirror those of many UIS genes, and when coupled with our RNA-seq evidence that the absence of Puf2 leads to the specific downregulation of a subset of UIS transcripts, this suggests that some UIS transcripts might be the direct or early indirect targets of Puf2 stabilization and/or repression. Additional evidence for this comes from the observation that several transcripts from the UIS gene family are highly expressed but are not translated in salivary gland sporozoites (Lindner and Swearingen et al., in press). Although Puf2 is expressed in oocysts and oocyst sporozoites, it is also a UIS gene, which shows much higher expression in salivary gland sporozoites. It is conceivable that Puf2 could regulate the translation of its own transcript as it bears Nanos Response Elements (NREs) in its 5′ and 3′ UTRs, thus establishing a negative feedback loop that fine-tunes its own abundance and providing a mechanism to ensure that targeted mRNAs are sufficiently regulated. A similar negative feedback loop is employed by the DHFR-TS protein, which specifically binds and represses its transcript and leads to the hypersensitivity of Plasmodium to anti-folate drugs (Zhang and Rathod, 2002). This negative feedback would provide an additional layer of regulation upon that offered by specific transcription factors that are upregulated stage-specifically, such as ApiAp2-Sp (PY00247) (Yuda et al., 2009; Lindner et al., 2010). While no direct evidence currently links UIS transcripts specifically to Puf2 storage granules, our composite global analyses provide a reasonable explanation for the silencing and/or protection of a subset of these transcripts.


Figure 6. A model of Puf2 function in sporozoite development and transmission. Parasite development is depicted chronologically from left-to-right with relative abundances of Puf2 protein and UIS gene family transcripts charted below it. The timing of exogenous expression of Puf2 from the CSP promoter (non-permissive) or the UIS4 promoter (permissive) are indicated by red and green bars respectively. Genotypes of parasites characterized in this study are noted, along with the parasite's morphology and infectivity.

Download figure to PowerPoint

We found that Puf2 cytosolic granules do not contain markers of P-bodies (XRN1) nor strictly of stress granules [eIF2α (neither total nor the phosphoserine-containing form)]. Thus, we have chosen to functionally distinguish these as ‘storage granules’ as have been invoked previously in model organisms (Anderson and Kedersha, 2008). It remains possible that these storage granules might be a subpopulation of either P-bodies or stress granules, but such confirmation and reclassification awaits the production and/or availability of additional specific antibodies. It is tempting to speculate that perhaps the canonical eIF2α regulatory events do not control Puf2 storage granules, which contrasts with recent work that has implicated eIF2α and its regulatory kinases in the developmental control of sporozoites and blood stage parasites (Zhang et al., 2010; 2012). It would therefore be of interest to determine experimentally what does regulate these obviously critical granules and their related functions, and how this compares and contrasts to the mechanisms governing SAP1-containing granules, P-bodies and stress granules.

We and others have demonstrated that puf2 sporozoites are unable to maintain their infective state and invasive morphology during prolonged salivary gland residence (Figs 1 and 2, Gomes-Santos et al., 2011; Muller et al., 2011). That Puf2 is employed to allow for sporozoites to remain infectious for longer periods of time might be an adaptation to specifically increase the window of transmission opportunity. We additionally observed that pypuf2 parasites have no detectable defect if isolated and inoculated into mice immediately after salivary gland invasion. To explain this, we propose that a small window of opportunity exists for transmission to occur that does not require the transcript storage, protection, degradation and/or silencing afforded directly or indirectly by Puf2. We hypothesize that the early salivary gland sporozoite might accommodate small perturbations to its RNA homeostasis caused by the absence of Puf2, but as these perturbations increase and accumulate over time (or as they arise due to premature overexpression of Puf2 in the oocyst) they would cause the loss of infectivity and gross deregulation of the parasite and lead to defects such as premature transformation. This can perhaps explain why pbpuf2 sporozoites isolated at 18 days post infection were more prone to transformation in axenic cultures (Gomes-Santos et al., 2011). Thus, it is possible that modulating the abundances of these transcripts would only lead to a phenotypic defect after crossing a critical threshold, explaining why pypuf2 sporozoites taken early after invading the salivary gland have no loss of infectivity but yet exhibit significantly affected transcripts.

To date, no direct evidence implicates Puf2 as a bona fide translational repressor. In fact, our UIS4prom:Puf2 transgenic parasite that overexpresses Puf2 in the sporozoite and liver stages suggests that Puf2 is not sufficient to translationally repress its targets, and might require additional regulatory effector proteins. However, our current data does implicate Puf2 in another activity characteristic of its family: modulation of transcript abundances. We propose that the phenotypes observed in Puf2-deficient-sporozoites results from the absence of Puf2's direct or indirect role in suppressing the abundance of transcripts related to RNA degradation. The absence of Puf2 thus leads to the deregulation of RNA homeostasis due to the increased presence of RNA degradation machinery, which in turn leads to the degradation of transcripts known to be important for infectivity (e.g. UIS transcripts).

Through perturbation of endogenous Puf2 expression via the introduction of the CSP or UIS4 promoters, we have demonstrated that regulated Puf2 expression is critical for proper sporozoite development in the oocyst and infection of the salivary glands, but not in the liver stage. Ectopically expressed Puf2 (in the presence of other potential regulatory effector proteins) aberrantly affects the abundance and/or stability of specific mRNAs prematurely, promiscuously, or both. In contrast, expression of either WT Puf2 or just its RBD from its native promoter produces parasites capable of normal development and infectivity of mice. This suggests that the differences in Puf2 expression timing, abundance, or both can be sufficient to cause severe deregulation of the parasite's life cycle progression, effectively rendering it non-transmissible.

Further determination of the molecular details of Puf2 function, its direct and indirect mRNA targets and protein interactions in storage granules will provide valuable insights into the regulation of parasite transmission, and perhaps provide novel targets to interfere with infection.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References
  8. Supporting Information

Experimental animals and parasite production

Six- to eight-week-old female Swiss Webster (‘SW’) mice from Harlan (Indianapolis, IN) were used for production of transgenic parasites and for parasite life cycle maintenance. Six- to eight-week-old female BALB/cJ mice from the Jackson laboratory (Bar Harbor, ME) were used for assessments of parasite infectivity and indirect immunofluorescence assays (see below). P. yoelii 17 XNL (‘Py17XNL’, a non-lethal strain) WT and transgenic parasites were cycled between SW mice and Anopheles stephensi mosquitoes. Infected mosquitoes were maintained on sugar water at 24°C and 70% humidity. This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Seattle Biomedical Research Institute has an OLAW Animal Welfare Assurance (A3640-01). The protocol was approved by the Seattle Biomedical Research Institute Institutional Animal Care and Use Committee (Protocol #: SK-02).

Reverse genetics of Plasmodium yoelii 17XNL (Py17XNL) parasites

Gene targeting constructs for transgenic parasite production were designed as previously described with some modifications (Mikolajczak et al., 2008b). Briefly, two regions of the targeted locus were PCR amplified with the Phusion polymerase (NEB) supplemented with 5 mM MgCl2 with specific primers (all oligonucleotides used in this study are listed in Table S2). The two PCR products were gel purified (Gel Extraction Kit, Qiagen), precipitated by ethanol, and fused by Sequence Overlap Extension PCR (‘SOE PCR’) without supplemented MgCl2 by virtue of complementary sequences designed into the primers. This SOE PCR product (‘targeting sequence’) was digested at the 5′ and 3′ ends via exogenous restriction sites designed in the primers, gel purified, precipitated by ethanol, and inserted into a modified pDEF vector designed for genetic disruptions or for gene modifications by locus replacement via double-cross-over recombination. Targeting sequences designed to express only the RBD of Puf2 (‘Puf2-RBDmyc’, AA132-478) or to replace the promoter region of Puf2 with that of Circumsporozoite Protein (‘CSPprom:Puf2’) or UIS4 (‘UIS4prom:Puf2’) were produced by an additional SOE PCR event to seamlessly fuse the promoter with the coding sequence without an intervening restriction enzyme site. Prior to transfection, plasmids were linearized at unique restriction sites designed between the two halves of the targeting sequences, precipitated by ethanol, and dissolved in ddH2O.

Wild-type Py17XNL parasites were genetically modified using standard methods as previously described (Jongco et al., 2006; Lindner et al., 2011). The presence of transgenic parasites was assessed by genotyping PCR. At least two independent clones of each transgenic parasite were isolated by limiting dilution infection of female SW mice.

Assessment of sporozoite infectivity by measuring the time to blood stage patency

Salivary gland sporozoites (WT or transgenic) were isolated by microdissection from mosquito salivary glands at the indicated time points after a single infected-blood meal as previously described (Aly et al., 2011). Ten thousand parasites were injected intravenously (‘iv’) into the tail vein of BALB/cJ mice. Alternatively, 105 or 106 oocyst sporozoites were isolated from mosquito midguts at their peak times (10 days post infection for WT parasites, 14 days post infection for CSPprom:Puf2 parasites), and were similarly iv injected into the tail vein of BALB/cJ mice. The time to blood stage patency (defined as > 1 infected RBCs/10 000 RBCs) was scored microscopically by Giemsa-stained thin blood smears. A one-day delay in the time to blood stage patency correlates with a 90% reduction in sporozoite infectivity (Gantt et al., 1998). Two biological replicates were measured for each condition.

Indirect immunofluorescence assay (IFA)

Salivary gland sporozoites, in vitro infected HepG2 + CD81 cells and in vivo infected hepatocytes were processed for IFA essentially as previously described (Miller et al., 2012). All sample types were stained at room temperature with the following appropriately diluted primary antibodies: mouse anti-c-myc (Clone 9E10), rabbit anti-c-myc (Santa Cruz, SC-789), rabbit anti-SAP1 (Aly et al., 2011), mouse anti-PyCSP (Clone 2F6), mouse anti-PyHep17 (ascites), rabbit anti-Tg eIF2α (prepared as in Sullivan et al., 2004 by Thermo Scientific, Rockford, IL), rabbit anti-Tg eIF2α P-Ser51 (Narasimhan et al., 2008), rabbit anti-TgXRN1 (raised against AA939-1038 with an N-terminal 6xHis tag, ID = TGGT1_041790, Primm Biotech), rabbit anti-PbPuf2 peptide 904 (Gomes-Santos et al., 2011), Alexa Fluor-conjugated secondary antibodies specific to mouse or rabbit IgG (Alexa Fluor 488, 594; Invitrogen), and 4′,6-diamidino-2-phenylindole (DAPI). Samples were covered with VectaShield (Vector Laboratories) and sealed under a coverglass slip. Fluorescent and DIC images were acquired using a DeltaVision Spectris RT microscope (Applied Precision) using a 40× or 100× oil objective and were deconvolved using the softWoRx software package.

Measurement of oocyst and sporozoite development

Mosquitoes infected with WT or transgenic parasites were dissected at t = 5, 10, 14, 16, 18, 20 or 22 days post blood meal to isolate midguts and salivary glands by standard methods. Oocyst burden and evidence of sporozoites development were assessed by light microscopy at 40× magnification. Measurement of oocyst diameters was evaluated by a DeltaVision Spectris RT microscope using the softWoRx software package. Average oocyst and salivary gland sporozoite numbers per mosquito were determined by counting appropriately diluted samples taken from at least 20 mosquitoes on a haemocytometer at 40× magnification.

Measurement of premature transformation

Salivary gland sporozoites were isolated from mosquitoes at different times post blood meal and were assessed by IFA [mouse anti-PyCSP (clone 2F6), Alexa Fluor 488-conjugated anti-mouse IgG and DAPI] for indications of premature transformation. Parasites were scored as having transformed into an early exoerythrocytic form (‘early EEF’) if they exhibited a medial bulge around the nucleus, deflation of its ends and/or rounding up. For each parasite type and time point, three biological replicates of 500 parasites each were scored.

Measurement of axenic transformation potential

Salivary gland sporozoites were isolated by microdissection from mosquito salivary glands 14 days post mosquito blood meal, purified through DEAE-cellulose (Mack et al., 1978), and 5 × 105 sporozoites were incubated in 1× DMEM + gentamicin for 24 h at 37°C. Parasites that had completely transformed to early EEFs (defined as completely rounded cells) were scored by IFA as above for the measurement of premature transformation. For each parasite type, two biological replicates of 500 parasites each were scored.

Quantitative RT-PCR

Oocysts were isolated from mosquitoes infected with WT or CSPprom:Puf2 transgenic parasites 5 days post blood meal. Complementary DNA (cDNA) was produced (Superscript III, Invitrogen) and used as a template for quantitative PCR as per standard methods. Primers specific to 18S RNA, Puf2 or the fused CSPprom:Puf2 transcript were used to compare expression levels in triplicate across samples using an Applied Biosystems 7300 Real Time PCR System with the following cycling parameters: 50°C 2 min, 95°C 10 min, 50× (95°C 15 s, 55°C 15 s, 60°C 1 min). Representative RT-PCRs of the fused CSPprom:Puf2 transcript using cDNA templates or no reverse transcriptase controls were run on 2% agarose gels to demonstrate specificity. Primer sequences are listed in Table S2.

Comparative RNA-seq

Wild-type (Py17XNL) or pypuf2salivary gland sporozoites were isolated as described above, and were purified twice over DEAE cellulose resin to remove contaminating mosquito components. RNA from an equal number of parasites (4 × 106) was prepared using the RNeasy Miniprep Kit with an added DNaseI module (Qiagen). Sample quality and purity were assessed spectrophotometrically and by PCR for genomic DNA contamination respectively. RNA samples were converted into cDNA libraries using Illumina TruSeq RNA sample preparation kit (Illumina #RS-122–2001) by Expression Analysis, Inc (Durham, NC). Briefly, mRNA was purified using poly d(T) magnetic beads, chemically fragmented, and converted into single-stranded cDNA with reverse transcriptase and random hexamer primers. Double-stranded cDNA was prepared using DNA Polymerase I, which was subsequently end repaired to create blunt ends. Addition of a single A nucleotide to the 3′ end facilitated ligation of sequencing adapters (by virtue of a single T base overhang), and PCR amplification of adapter-ligated cDNA was performed to increase the amount of the sequence-ready library. Final cDNA libraries were analysed for size distribution and quality using an Agilent Bioanalyzer (DNA 1000 kit, Agilent # 5067-1504), quantified by qPCR (KAPA Library Quant Kit, KAPA Biosystems # KK4824), and its final concentration was adjusted to 2 nM. Normalized cDNA libraries were used to prepare flow cells using the Illumina TruSeq Paired-End Cluster Kit V3 (Illumina # PE-401–3001). Briefly, libraries were denatured using fresh 0.1 N NaOH, then diluted to 20 pM using chilled hybridization buffer. Libraries were further diluted to 13 pM and aliquots were placed on an Illumina cBot instrument to produce clusters through bridge amplification. Sequencing was conducted on an Illumina HiSeq 2000 using 50 bases paired end, plus a 7 base index cycle. Sequencing data were mapped using Tophat version 1.4.1 (available at, Trapnell et al., 2009) to the P. yoelii YM strain reference sequence (Version 1, available at due to its overall higher sequence quality and better genome assembly than that of the current assembly of the P. yoelii 17XNL strain. Additional quality control was applied to the datasets by clipping or removing poor quality reads and N's by ea-utils toolkit (, with -p 10 and -q 7 settings. The total number of reads retained after clipping was 46 350 420 for P. yoelii 17XNL WT parasites and 63 050 534 for pypuf2 parasites, with 93% and 94% of reads being successfully aligned against the P. yoelii YM reference genome by tophat respectively. Mapped reads were normalized by upper-quartile normalization and the differential transcript abundances between samples were measured in fragments per kilobase of transcript per million mapped fragments (FPKM) using Cuffdiff (Version 1.3.0 available at (Trapnell et al., 2010). RNA-seq differential analysis was visualized by CummeRbund (available at Transcripts with P-values of less than 0.01 and at least fourfold difference in expression (i.e. log fold > 2) were selected to identify the most differentially regulated genes. Coverage plots were generated to examine the RNA distribution of reads across selected genes-of-interest by comparing the number of unique reads present at each nucleotide position. Final RNA-seq datasets for both P. yoelii 17XNL WT and pypuf2 have been deposited in the GEO database under the accession number GSE41873, and are also provided as Table S3. Detection of Nanos Response Elements (NREs, which can be bound by Puf2 in vitro) within these affected transcripts and globally with all annotated ORFs was conducted on the PlasmoDB website (Version 9.1, The Py17XNL genomic DNA reference sequence was searched for the degenerate sequence TGTAHATA (where H = T, C or A) within 1.5 kb upstream of start codons, open reading frames, and within 1.0 kb downstream of stop codons.

Statistical tests

Data were assessed by the Logrank, Jonckheere–Terpstra and Wilcoxon rank sum statistical tests as described (Drinkwater and Denniston, 2011) using Mstat software (v. 5.5.4; available at


  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References
  8. Supporting Information

We would like to thank the members of the Kappe Lab for both technical assistance and critical discussion of this work. We thank S. Khan for providing anti-PbPuf2 antisera and the Wellcome Trust Sanger Institute for provision of the P. yoelii YM strain reference sequence ahead of publication. This investigation was supported by the NIH under the Ruth L. Kirschstein National Research Service Awards (F32GM083438 to S.E.L., F32AI093271 to B.R.J.) and by NIH grants (AI084031 and AI077502 to W.J.S. Jr.). S.E.L. performed research; A.M.V and S.A.M. provided vital experimental support; B.R.J and W.J.S. Jr provided critical reagents and conceptual expertise; S.E.L and S.H.I.K. designed research, analysed and interpreted data and wrote the manuscript.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References
  8. Supporting Information
  • Aly, A.S., Vaughan, A.M., and Kappe, S.H. (2009) Malaria parasite development in the mosquito and infection of the mammalian host. Annu Rev Microbiol 63: 195221.
  • Aly, A.S., Lindner, S.E., MacKellar, D.C., Peng, X., and Kappe, S.H. (2011) SAP1 is a critical post-transcriptional regulator of infectivity in malaria parasite sporozoite stages. Mol Microbiol 79: 929939.
  • Anderson, P., and Kedersha, N. (2008) Stress granules: the Tao of RNA triage. Trends Biochem Sci 33: 141150.
  • Balagopal, V., and Parker, R. (2009) Polysomes, P bodies and stress granules: states and fates of eukaryotic mRNAs. Curr Opin Cell Biol 21: 403408.
  • Campbell, Z.T., Menichelli, E., Friend, K., Wu, J., Kimble, J., Williamson, J.R., and Wickens, M. (2012) A conserved interface between PUF and CPEB proteins. J Biol Chem 287: 1885418862.
  • Carlton, J.M., Angiuoli, S.V., Suh, B.B., Kooij, T.W., Pertea, M., Silva, J.C., et al. (2002) Genome sequence and comparative analysis of the model rodent malaria parasite Plasmodium yoelii yoelii. Nature 419: 512519.
  • Drinkwater and Denniston (2011) Statistical Problems in genetics and molecular biology. CreateSpace 1: 1299.
  • Fan, Q., Li, J., Kariuki, M., and Cui, L. (2004) Characterization of PfPuf2, member of the Puf family RNA-binding proteins from the malaria parasite Plasmodium falciparum. DNA Cell Biol 23: 753760.
  • Florens, L., Washburn, M.P., Raine, J.D., Anthony, R.M., Grainger, M., Haynes, J.D., et al. (2002) A proteomic view of the Plasmodium falciparum life cycle. Nature 419: 520526.
  • Gantt, S.M., Myung, J.M., Briones, M.R., Li, W.D., Corey, E.J., Omura, S., et al. (1998) Proteasome inhibitors block development of Plasmodium spp. Antimicrob Agents Chemother 42: 27312738.
  • Goldstrohm, A.C., Hook, B.A., Seay, D.J., and Wickens, M. (2006) PUF proteins bind Pop2p to regulate messenger RNAs. Nat Struct Mol Biol 13: 533539.
  • Gomes-Santos, C.S., Braks, J., Prudencio, M., Carret, C., Gomes, A.R., Pain, A., et al. (2011) Transition of Plasmodium sporozoites into liver stage-like forms is regulated by the RNA binding protein Pumilio. PLoS Pathog 7: e1002046.
  • Guttery, D.S., Ferguson, D.J., Poulin, B., Xu, Z., Straschil, U., Klop, O., et al. (2012) A putative homologue of CDC20/CDH1 in the malaria parasite is essential for male gamete development. PLoS Pathog 8: e1002554.
  • Hall, N., Karras, M., Raine, J.D., Carlton, J.M., Kooij, T.W., Berriman, M., et al. (2005) A comprehensive survey of the Plasmodium life cycle by genomic, transcriptomic, and proteomic analyses. Science 307: 8286.
  • Jayabalasingham, B., Bano, N., and Coppens, I. (2010) Metamorphosis of the malaria parasite in the liver is associated with organelle clearance. Cell Res 20: 10431059.
  • Jones, C.I., Zabolotskaya, M.V., and Newbury, S.F. (2012) The 5′ --> 3′ exoribonuclease XRN1/Pacman and its functions in cellular processes and development. Wiley Interdiscip Rev RNA 3: 455468.
  • Jongco, A.M., Ting, L.M., Thathy, V., Mota, M.M., and Kim, K. (2006) Improved transfection and new selectable markers for the rodent malaria parasite Plasmodium yoelii. Mol Biochem Parasitol 146: 242250.
  • Kaiser, K., Camargo, N., and Kappe, S.H. (2003) Transformation of sporozoites into early exoerythrocytic malaria parasites does not require host cells. J Exp Med 197: 10451050.
  • Kappe, S.H., Vaughan, A.M., Boddey, J.A., and Cowman, A.F. (2010) That was then but this is now: malaria research in the time of an eradication agenda. Science 328: 862866.
  • Labaied, M., Harupa, A., Dumpit, R.F., Coppens, I., Mikolajczak, S.A., and Kappe, S.H. (2007) Plasmodium yoelii sporozoites with simultaneous deletion of P52 and P36 are completely attenuated and confer sterile immunity against infection. Infect Immun 75: 37583768.
  • Lasonder, E., Janse, C.J., van Gemert, G.J., Mair, G.R., Vermunt, A.M., Douradinha, B.G., et al. (2008) Proteomic profiling of Plasmodium sporozoite maturation identifies new proteins essential for parasite development and infectivity. PLoS Pathog 4: e1000195.
  • Le Roch, K.G., Johnson, J.R., Florens, L., Zhou, Y., Santrosyan, A., Grainger, M., et al. (2004) Global analysis of transcript and protein levels across the Plasmodium falciparum life cycle. Genome Res 14: 23082318.
  • Lindner, S.E., De Silva, E.K., Keck, J.L., and Llinas, M. (2010) Structural determinants of DNA binding by a P. falciparum ApiAP2 transcriptional regulator. J Mol Biol 395: 558567.
  • Lindner, S.E., Llinas, M., Keck, J.L., and Kappe, S.H. (2011) The primase domain of PfPrex is a proteolytically matured, essential enzyme of the apicoplast. Mol Biochem Parasitol 180: 6975.
  • Lindner, S.E., Miller, J.L., and Kappe, S.H. (2012) Malaria parasite pre-erythrocytic infection: preparation meets opportunity. Cell Microbiol 14: 316324.
  • Mack, S.R., Vanderberg, J.P., and Nawrot, R. (1978) Column separation of Plasmodium berghei sporozoites. J Parasitol 64: 166168.
  • Mair, G.R., Braks, J.A., Garver, L.S., Wiegant, J.C., Hall, N., Dirks, R.W., et al. (2006) Regulation of sexual development of Plasmodium by translational repression. Science 313: 667669.
  • Mair, G.R., Lasonder, E., Garver, L.S., Franke-Fayard, B.M., Carret, C.K., Wiegant, J.C., et al. (2010) Universal features of post-transcriptional gene regulation are critical for Plasmodium zygote development. PLoS Pathog 6: e1000767.
  • Matuschewski, K., Ross, J., Brown, S.M., Kaiser, K., Nussenzweig, V., and Kappe, S.H. (2002) Infectivity-associated changes in the transcriptional repertoire of the malaria parasite sporozoite stage. J Biol Chem 277: 4194841953.
  • Miao, J., Li, J., Fan, Q., Li, X., and Cui, L. (2010) The Puf-family RNA-binding protein PfPuf2 regulates sexual development and sex differentiation in the malaria parasite Plasmodium falciparum. J Cell Sci 123: 10391049.
  • Mikolajczak, S.A., Silva-Rivera, H., Peng, X., Tarun, A.S., Camargo, N., Jacobs-Lorena, V., et al. (2008a) Distinct malaria parasite sporozoites reveal transcriptional changes that cause differential tissue infection competence in the mosquito vector and mammalian host. Mol Cell Biol 28: 61966207.
  • Mikolajczak, S.A., Aly, A.S., Dumpit, R.F., Vaughan, A.M., and Kappe, S.H. (2008b) An efficient strategy for gene targeting and phenotypic assessment in the Plasmodium yoelii rodent malaria model. Mol Biochem Parasitol 158: 213216.
  • Miller, J.L., Harupa, A., Kappe, S.H., and Mikolajczak, S.A. (2012) Plasmodium yoelii macrophage migration inhibitory factor is necessary for efficient liver-stage development. Infect Immun 80: 13991407.
  • Miller, M.A., and Olivas, W.M. (2011) Roles of Puf proteins in mRNA degradation and translation. Wiley Interdiscip Rev RNA 2: 471492.
  • Mueller, A.K., Camargo, N., Kaiser, K., Andorfer, C., Frevert, U., Matuschewski, K., and Kappe, S.H. (2005a) Plasmodium liver stage developmental arrest by depletion of a protein at the parasite-host interface. Proc Natl Acad Sci USA 102: 30223027.
  • Mueller, A.K., Labaied, M., Kappe, S.H., and Matuschewski, K. (2005b) Genetically modified Plasmodium parasites as a protective experimental malaria vaccine. Nature 433: 164167.
  • Muller, K., Matuschewski, K., and Silvie, O. (2011) The Puf-family RNA-binding protein Puf2 controls sporozoite conversion to liver stages in the malaria parasite. PloS ONE 6: e19860.
  • Narasimhan, J., Joyce, B.R., Naguleswaran, A., Smith, A.T., Livingston, M.R., Dixon, S.E., et al. (2008) Translation regulation by eukaryotic initiation factor-2 kinases in the development of latent cysts in Toxoplasma gondii. J Biol Chem 283: 1659116601.
  • Natarajan, R., Thathy, V., Mota, M.M., Hafalla, J.C., Menard, R., and Vernick, K.D. (2001) Fluorescent Plasmodium berghei sporozoites and pre-erythrocytic stages: a new tool to study mosquito and mammalian host interactions with malaria parasites. Cell Microbiol 3: 371379.
  • Painter, H.J., Campbell, T.L., and Llinas, M. (2011) The Apicomplexan AP2 family: integral factors regulating Plasmodium development. Mol Biochem Parasitol 176: 17.
  • Paton, M.G., Barker, G.C., Matsuoka, H., Ramesar, J., Janse, C.J., Waters, A.P., and Sinden, R.E. (1993) Structure and expression of a post-transcriptionally regulated malaria gene encoding a surface protein from the sexual stages of Plasmodium berghei. Mol Biochem Parasitol 59: 263275.
  • Porter, R.J., Laird, R.L., and Dusseau, E.M. (1954) Studies on malarial sporozoites. II. Effect of age and dosage of sporozoites on their infectiousness. Exp Parasitol 3: 267274.
  • Qiu, C., Kershner, A., Wang, Y., Holley, C.P., Wilinski, D., Keles, S., et al. (2012) Divergence of Pumilio/fem-3 mRNA binding factor (PUF) protein specificity through variations in an RNA-binding pocket. J Biol Chem 287: 69496957.
  • Quenault, T., Lithgow, T., and Traven, A. (2011) PUF proteins: repression, activation and mRNA localization. Trends Cell Biol 21: 104112.
  • Sullivan, W.J., Jr, Narasimhan, J., Bhatti, M.M., and Wek, R.C. (2004) Parasite-specific eIF2 (eukaryotic initiation factor-2) kinase required for stress-induced translation control. Biochem J 380: 523531.
  • Trapnell, C., Pachter, L., and Salzberg, S.L. (2009) TopHat: discovering splice junctions with RNA-Seq. Bioinformatics 25: 11051111.
  • Trapnell, C., Williams, B.A., Pertea, G., Mortazavi, A., Kwan, G., van Baren, M.J., et al. (2010) Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol 28: 511515.
  • VanBuskirk, K.M., O'Neill, M.T., De La Vega, P., Maier, A.G., Krzych, U., Williams, J., et al. (2009) Preerythrocytic, live-attenuated Plasmodium falciparum vaccine candidates by design. Proc Natl Acad Sci USA 106: 1300413009.
  • Vanderberg, J.P. (1975) Development of infectivity by the Plasmodium berghei sporozoite. J Parasitol 61: 4350.
  • Vaughan, A., Chiu, S.Y., Ramasamy, G., Li, L., Gardner, M.J., Tarun, A.S., et al. (2008) Assessment and improvement of the Plasmodium yoelii yoelii genome annotation through comparative analysis. Bioinformatics 24: i383i389.
  • Vaughan, A.M., Wang, R., and Kappe, S.H. (2010) Genetically engineered, attenuated whole-cell vaccine approaches for malaria. Hum Vaccin 6: 107113.
  • Wang, X., McLachlan, J., Zamore, P.D., and Hall, T.M. (2002) Modular recognition of RNA by a human pumilio-homology domain. Cell 110: 501512.
  • Wek, R.C., Jiang, H.Y., and Anthony, T.G. (2006) Coping with stress: eIF2 kinases and translational control. Biochem Soc Trans 34: 711.
  • WHO (2011) World Malaria Report: 2011. Geneva: World Health Organization.
  • Wickens, M., Bernstein, D.S., Kimble, J., and Parker, R. (2002) A PUF family portrait: 3′UTR regulation as a way of life. Trends Genet 18: 150157.
  • Yuda, M., Iwanaga, S., Shigenobu, S., Mair, G.R., Janse, C.J., Waters, A.P., et al. (2009) Identification of a transcription factor in the mosquito-invasive stage of malaria parasites. Mol Microbiol 71: 14021414.
  • Zhang, K., and Rathod, P.K. (2002) Divergent regulation of dihydrofolate reductase between malaria parasite and human host. Science 296: 545547.
  • Zhang, M., Fennell, C., Ranford-Cartwright, L., Sakthivel, R., Gueirard, P., Meister, S., et al. (2010) The Plasmodium eukaryotic initiation factor-2alpha kinase IK2 controls the latency of sporozoites in the mosquito salivary glands. J Exp Med 207: 14651474.
  • Zhang, M., Mishra, S., Sakthivel, R., Rojas, M., Ranjan, R., Sullivan, W.J., Jr, et al. (2012) PK4, a eukaryotic initiation factor 2alpha(eIF2alpha) kinase, is essential for the development of the erythrocytic cycle of Plasmodium. Proc Natl Acad Sci USA 109: 39563961.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References
  8. Supporting Information

Fig. S1. Genetic disruption of PyPuf2 using a corrected gene structure annotation.

A. The gene structure annotation of PyPuf2 in the current version of PlasmoDB (v. 9.2, available at includes a wrongly called final exon/intron boundary. The corrected gene structure is been confirmed by RT-PCR and sequencing from P. yoelii salivary gland sporozoite cDNA (data not shown, ORF encoded by nucleotides 11 657–11 697 … 11 879–12 024 … 12 126–12 212 … 12 347–13 509, antisense orientation) as well as by the functional expression of C-terminal epitope tags (Figs 3 and 4). A double-cross-over recombination strategy was used to delete the PyPuf2 ORF (pypuf2) using a modified pDEF plasmid and successful production of transgenic parasites was confirmed by genotyping PCR using the indicated primer sets. The RNA-binding domain (RBD) that is comprised of eight RNA recognition helices is also indicated.

B. pypuf2 transgenic populations were subjected to limiting dilution cloning, and clonal parasite populations (as determined by the Poisson distribution) were assessed by genotyping PCR using primers indicated in A. The Molecular Weight (MW) Marker is the 1 kb marker from New England Biolabs. Two independent clones were analysed in parallel with Py17XNL wild-type parasites.

C. Blood stage parasites resulting from infection by wild-type or pypuf2 sporozoites were similarly assessed by genotyping PCR. One clone representative of all samples is shown.

Fig. S2. Deletion of the N-terminal domain of Puf2 by genetic replacement.

A. The correctly annotated wild-type locus was genetically modified by double-cross-over recombination with a plasmid that seamlessly fused Puf2's 5′ UTR (with its start codon) to its RNA-binding domain (RBD) with an appended C-terminal 4× Myc tag (‘Puf2-RBDmyc’).

B. Puf2-RBDmyc transgenic populations were subjected to limiting dilution cloning, and clonal parasite populations (as determined by the Poisson distribution) were assessed by genotyping PCR using primers indicated in (A). The Molecular Weight (MW) Marker is the 1 kb marker from New England Biolabs. Two independent clones were analysed in parallel with Py17XNL wild-type parasites.

C. Blood stage parasites resulting from infection by wild-type or Puf2-RBDmyc sporozoites were similarly assessed by genotyping PCR and compared to the Puf2-RBDmyc blood stage population fed to mosquitoes (Puf2-RBDmyc Clone 1 iRBCs, Input).

Fig. S3. Puf2 is expressed in wild-type oocysts and oocyst sporozoites. Representative IFA images are shown of wild-type oocysts and oocyst sporozoites stained with anti-PyCSP (clone 2F6), anti-PbPuf2 [raised against peptide 904 (Gomes-Santos et al., 2011)] and DAPI. Differential interference contrast (DIC) images demonstrate overall sporozoite morphology. Scale bar is 10 μm for oocyst images and 5 μm for oocyst sporozoite images.

Fig. S4. Introduction of the circumsporozoite protein (CSP) promoter to drive Puf2 expression.

A. The correctly annotated wild-type locus was genetically modified by double-cross-over recombination with a plasmid that seamlessly fused PyCSP's 5′ UTR (with its start codon) to PyPuf2's ORF (‘CSPprom:Puf2’).

B. CSPprom:Puf2 transgenic populations were subjected to limiting dilution cloning, and clonal parasite populations (as determined by the Poisson distribution) were assessed by genotyping PCR using primers indicated in A. The Molecular Weight (MW) Marker is a 1 kb marker (New England Biolabs). Two independent clones were analysed in parallel with Py17XNL wild-type parasites.

C. The fused CSPprom:Puf2 transcript is present in cDNA generated from CSPprom:Puf2 transgenic parasites. RT-PCRs were run using templates generated in the absence or presence of reverse transcriptase and the resulting products were run on a 2% agarose gel. The Molecular Weight (MW) marker is a 100 bp marker (New England Biolabs).

Fig. S5. Replacement of Puf2's endogenous promoter with the UIS4 promoter.

A. The correctly annotated wild-type locus was genetically modified by double-cross-over recombination with a plasmid that seamlessly fused PyUIS4's 5′ UTR (with its start codon) to PyPuf2's ORF (‘UIS4prom:Puf2’).

B. UIS4prom:Puf2 transgenic populations were subjected to limiting dilution cloning, and clonal parasite populations (as determined by the Poisson distribution) were assessed by genotyping PCR using primers indicated in (A). The Molecular Weight (MW) Marker is a 1 kb marker (New England Biolabs). Two independent clones were analysed in parallel with Py17XNL wild-type parasites.

C and D. Representative IFA images are shown of UIS4prom:Puf2 parasites as (C) 14- and 22-day salivary gland sporozoites, a dedifferentiating sporozoite inside a hepatocyte and an early liver stage trophozoite [green: anti-PbPuf2 (904); red: anti-PyCSP (clone 2F6); blue: DAPI; scale bar is 5 μm] or (D) as 24 hr and 40 hr liver stage parasites [green: anti-PbPuf2 (904); red: anti-PyHep17; blue: DAPI; scale bar is 10 μm].

Table S1. No observable difference exists in the axenic transformation potential of pypuf2 sporozoites compared to wild-type.

Table S2. A listing of oligonucleotides used in this study.


Table S3. Comparative RNA-seq data from 14-day P. yoelii 17XNL wild-type and pypuf2 salivary gland sporozoites.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.