The ATP-dependent ClpQY system is a prokaryotic proteasome-like multi-subunit machinery localized in the mitochondrion of malaria parasite. The ClpQY machinery consists of ClpQ threonine protease and ClpY ATPase. In the present study, we have assessed cellular effects of transient interference of PfClpQ protease activity in Plasmodium falciparum using a trans-dominant negative approach combined with FKBP degradation domain system. A proteolytically inactive mutant PfClpQ protein [PfClpQ(mut)] fused with FKBP degradation domain was expressed in parasites, which gets stabilized by Shield1 drug treatment. We show that the inactive PfClpQ(mut) interacts with wild-type PfClpQ and associates within multi-subunit complex in the parasite. Stabilization of the PfClpQ(mut) and its association in the protease machinery caused dominant negative effect in the transgenic parasites, which disrupted the growth cycle of asexual blood stage parasites. The mitochondria in these parasites showed abnormal morphology, these mitochondria were not able to grow and divide in the parasite. We further show that the dominant negative effect of PfClpQ(mut) disrupted transcription of mitochondrial genome encoded genes, which in turn blocked normal development and functioning of the mitochondria.
Malaria remains a major parasitic disease in the tropical and sub-tropical countries causing 1–2 million deaths globally every year Snow et al., 2005; Hay et al., 2010). There is an urgent need to define new drug targets and develop new pharmacophores against the disease. Therefore, one of the aims of the post-genomic era is to identify potential drug/vaccine targets from the Plasmodium falciparum genome/proteome database and characterize those targets using different reverse genetic tools. The metabolic pathways in two of the parasite organelles, the mitochondrion and the apicoplast, are considered as important drug target due to their possible prokaryotic origin. Indeed, antibiotics such as doxycycline and clindamycin, which target some of the prokaryotic metabolic pathways, have been shown to possess anti-parasitic efficacies (Waller and McFadden, 2005; Goodman et al., 2007; Schlitzer, 2007; Dahl and Rosenthal, 2008). Characterization of novel metabolic pathways in these organelles and understanding their functional role in the parasite may help us to identify novel drug targets.
ATP-dependent protease machineries are large multi-subunit protein complexes that include eukaryotic 26S proteasome system (De Mot et al., 1999; Ciechanover, 2005). The 26S proteasome machinery plays essential roles in controlling the levels of key regulatory proteins and in the elimination of abnormal polypeptides in eukaryotic cells. These tasks are carried out by the ClpAP, ClpXP and ClpQY (HslVU) proteases in archea and eubacteria (De Mot et al., 1999). The ClpQY machinery is a multimeric ATP-dependent protease system in the prokaryotes that resembles the eukaryotic 26S proteasome; it consists of two stacked hexameric rings of ClpQ protease which are capped on one or both sides with hexameric ring of AAA type ATPase, ClpY. The ATPases act as chaperons to unfold the substrate proteins which subsequently get degraded by the protease component (Bochtler et al., 2000). However, the ClpQY machinery is functionally distinct machinery as compared with 26S proteasome. The substrate protein of 26S proteasome are tagged by a chain of Ubiquitin, the poly-ubiqutinated proteins are recognized by 26S proteasome for degradation (Ciechanover, 2005); whereas there is no clear tag known for substrate recognition by ClpQY protease.
A number of proteases in the malaria parasite have been considered as potential drug targets. Detailed functional characterization of some of these proteases showed that these proteins play crucial roles in different metabolic pathways in the parasite (Blackman, 2000; Shenai et al., 2000; Rosenthal et al., 2002; Dasaradhi et al., 2005; Koussis et al., 2009; Moura et al., 2009; Russo et al., 2009). We have earlier characterized ClpQ protease and ClpP protease systems in the parasite (Ramasamy et al., 2007; Rathore et al., 2010). The ClpQ protease is localized in the parasite mitochondria whereas the ClpP is localized in the apicoplast, relict plastid in the parasite (Rathore et al., 2010; Tschan et al., 2010). We earlier showed that disruption of interaction of ClpQ with its ATPase partner ClpY in the parasite leads to onset of parasite cell death (Rathore et al., 2011); suggesting that this machinery is important for parasite survival. In the present study, we have studied the cellular effects of interference of the ClpQ protease activity P. falciparum by using trans-dominant negative approach. We show that transient interference in the PfClpQ activity causes hindrance in development of functional mitochondria in parasite and disrupts the asexual blood stage parasite cycle; these results ascertain the functional importance of prokaryotic ClpQ protease system in P. falciparum for survival of the parasite.
Active site mutant PfClpQ is proteolytically inactive
To gain insights into cellular role of PfClpQ we studied loss-of-function phenotypes using inactive version of PfClpQ. For this we first ascertained that mutations in the potential active site residues [Thr (38) and Ser (160); Fig. 1A] would inactivate PfClpQ protease. The mature protease region of PfClpQ and its mutated version (PfClpQ-mut) were cloned and corresponding recombinant proteins were expressed. The recombinant PfClpQ and PfClpQ(mut) proteins were expressed as soluble proteins in cytosol of the E. coli BL21(DE3) cells and were purified by affinity chromatography. Both the proteins separated as ∼ 23 kDa protein on the SDS-PAGE (Fig. 1B). The purified recombinant PfClpQ and PfClpQ(mut) proteins were assessed for the protease activity using an in vitro protease assay. As expected the PfClpQ protease showed threonine protease-like activity using the synthetic peptide substrate Suc-GGL-AMC in these assays as evident by continuous release of cleaved AMC (Fig. 1C). However, the PfClpQ-Mut protein showed no protease activity in these assays, suggesting that mutation of the potential active site residues lead to loss of protease activity (Fig. 1C).
To ascertain that the mutant protein is able to bind with the substrate, we assessed the activity of wild-type PfClpQ protein in presence of PfClpQ(mut) protein using peptide substrate. As shown in Fig. 1D, presence of PfClpQ(mut) protein in the assay effectively reduced activity of PfClpQ. However, when substrate concentration was doubled in the assay, PfClpQ was able to overcome this competition and PfClpQ activity level was restored; similarly, when PfClpQ concentration was doubled in the assay, PfClpQ activity level was also restored. Addition of a non-specific protein PfAARP in the assay showed no effect on the PfClpQ activity. These data show that there is a direct competition between PfClpQ and PfClpQ(mut) for binding with the substrate. Overall, these results suggest that the PfClpQ(mut) is able to bind with the substrate but it is not able to cleave it.
Expression of destabilizing-domain tagged inactive PfClpQ in transgenic parasites
An inactive version of PfClpQ was expressed as a transgene in the parasites to generate a dominant-negative phenotype and understand the functional significance of PfClpQ. The full pfclpQ gene with mutations in the active site residues was cloned in pHADD vector in frame with 3× HA-tag. Since overexpression of the inactive protein could be deleterious for the parasite, the protein levels in the parasite was regulated by recently developed destabilizing domain (DD) system (Armstrong and Goldberg, 2007; Dvorin et al., 2010); a ddFKBP gene was fused at the C-terminus of the transgene (Fig. 2A). The DD fusion proteins are unstable and are expected to be rapidly degraded in the cell; whereas in presence of the ligand Shield-1 (Shld1) the degradation is mitigated and the fusion protein gets stabilized. Transgenic parasite line was generated using plasmid construct pHADD-ClpQ(mut) to express inactive PfClpQ protease (PfClpQ-mut) and labelled as ClpQ(mut)–DD parasite line. The transfected parasites were selected over positive selection in absence of Shld1 drug. In addition, two other transgenic parasite lines generated to use as control, one using pHADD-ClpQ to express wild-type PfClpQ protease and labelled as ClpQ–DD parasite line and the second using pHADD vector and labelled as Vector–DD.
To identify the appropriate Shld1 drug concentration to be used, we first assessed toxicity of Shld1 drug on wild-type 3D7 parasites. The Shld1 drug showed no significant parasite growth inhibition till 1.0 μM concentration (see supplementary data, Fig. S2). Therefore, for all the assays Shld1 was used at a concentration of 1.0 μM. We first ascertained expression of PfClpQ(mut)–HA in the transgenic parasites after Shld1 treatment by Western blot analysis and immunofluorescence assay. The Shld1 treated parasites showed expression of PfClpQ(mut)–HA fusion protein of expected size (∼ 35 kDa) in the transgenic parasites by Western blot analysis (Fig. 2B). Further, the immunofluorescence assay using anti-HA antibodies also detected the fusion protein in the treated parasite (Fig. 2C); the immunofluorescence staining overlapped with the MitoTracker staining, suggesting that the fusion protein is localized in the mitochondria which is the site of action of native protein.
Inactive PfClpQ interacts with wild-type PfClpQ and associates within multi-subunit ClpQ protease complex in the parasite
Since ClpQ subunits interact with each other to form a large multi-subunit complex, we studied the interaction of inactive mutant-PfClpQ with wild-type PfClpQ. For in vitro interaction studies using recombinant proteins, 6× histidine tagged wild-type PfClpQ (PfClpQ-His) and S-tagged mutant-PfClpQ [PfClpQ(mut)-S] proteins were coexpressed in E. coli (Fig. 3A). Affinity chromatography was carried out using Ni-NTA matrix to pull out the PfClpQ-His protein from cell lysate; Western blot analysis showed that PfClpQ(mut)-S gets co-eluted with PfClpQ-His in these experiments (Fig. 3B). The lysate from E. coli expressing PfClpQ(mut)-S alone was used as control; in this experiment, eluates obtained in similar manner form Ni-NTA matrix did not show presence of PfClpQ(mut)-S (Fig. 3B).
Further, we investigated whether the episomally expressed PfClpQ(mut) protein associates in a complex with the native PfClpQ protein. The anti-HA antibodies were used for immunoprecipitation of PfClpQ(mut) and associated protein from total parasite lysate of ClpQ(mut)–DD parasites treated with Shld1 or solvent alone. Western blot analysis of immunoprecipitated complex showed that the PfClpQ gets co-immunoprecipitated with PfClpQ(mut)–HA fusion protein from parasite lysate (Fig. 3C), suggesting their existence as a protein complex in the parasite. The PfClpQ was not detected in eluates from control set or in reactions where anti-HA antibodies were replaced with non-specific antibodies.
Dominant negative effect of expression of inactive PfClpQ on parasite growth
We next investigated dominant negative effect of expression of inactive PfClpQ on parasite growth and development in vitro. Parasite growth was estimated as development of new ring stage parasite at 48 and 96 h after treatment with the Shld1 drug and compared with the parasite cultures treated with solvent alone. The ClpQ(mut)–DD parasite line in control set showed about 6–7 times increase in percentage new ring stage parasites on day 2 as compared with starting parasitaemia, which further showed 6 times increase on day 4 (Fig. 4A). However, the Shld1 treated ClpQ(mut)–DD parasite line showed no significant increase in new ring stage parasites on day 2 and on day 4; overall the Shld1 treated ClpQ(mut)–DD parasite cultures showed ∼ 97% growth reduction as compared with untreated cultures (Fig. 4A). The control parasite lines, ClpQ–DD and Vector–DD, showed similar increase in percentage of new ring stage parasite for treated and untreated cultures which was comparable to untreated ClpQ(mut)–DD parasite cultures (see supplementary data).
Expression of dominant negative inactive PfClpQ disrupts intra erythrocytic parasite developmental cycle
To study the dominant negative effect of inactive PfClpQ on parasite development and morphology synchronized ring stage parasite cultures of ClpQ(mut)–DD parasite lines were treated with Shld1 or solvent alone and parasite developmental-stage profile as well as total parasitaemia was analysed for next 5 days (∼ 120 h). In ClpQ(mut)–DD control set, the parasites developed from ring to trophozoites (day 1) to mature schizonts and subsequently merozoites released from these schiozont invaded new RBCs and developed into ring stage parasites 48 h (day 2) after treatment, this led to about 7 times increase in the total parasitaemia (Fig. 4B). These parasites further developed into late trophozoites and schizont on day 3, which subsequently formed new rings on day 4, which further increased the parasitaemia (Fig. 4B and C). Almost all of these ring stage parasites converted into trophozoites on day 5 (Fig. 4C). In the Shld1 treated ClpQ(mut)–DD parasite lines [ClpQ(mut)–DD + Shld1], about 80% of the ring stage parasites were able to develop into trophozoite stage parasites on day 1 as in the control sets; however, these trophozoites were not able to develop into mature schizonts (Fig. 4C and D). On day 2 most of the parasites were seen as stressed trophozoites or abnormal unruptured schizonts (Fig. 4D); therefore there was no significant increase in parasite even on 2 days after treatment. A small fraction of these schizonts showed delayed development and formed new ring stage parasites on day 3 (72 h after the treatment) which resulted in increase in parasitaemia (Fig. 4B–D). About 70% of these ring stage parasites were able to develop into trophozoite stage parasites on day 4 (Fig. 4B–D); however, these trophozoites also failed to develop into mature schizonts and ultimately died, so that there was no significant increase in parasitaemia even on day 5 after treatment (Fig. 4B–D). No significant difference was observed in growth of Shld1 treated ClpQ–DD set as compared with control set (Figs S3 and S4).
Dominant negative effect of inactive PfClpQ disrupts mitochondrial development
Given that the ClpQY machinery is functional in the parasite mitochondria, we studied the effect of expression of inactive PfClpQ in the parasite on development of its mitochondria. The ClpQ(mut)–DD parasite cultures at ring stage parasites (8–10 h post invasion) were treated with Shld1 or solvent alone, growth and morphology of the mitochondria was observed after different time points using MitoTracker stain. Based upon mitochondrial morphology and staining pattern five parasite groups were identified, and percentage parasites in each group were calculated (Fig. 5A and B; Fig. S5). The mitochondria in control ClpQ(mut)–DD set showed normal growth and development; in majority of these parasites at trophozoite stage (24 h after treatment) the mitochondria appeared as elongated structures, in late trophozoite/early schizont stage (36 h after treatment) most of these mitochondria showed branched structure (Fig. 5A and B). In ClpQ(mut)–DD + Shld1 set ∼ 25% of parasites showed elongated mitochondria at 24 h after treatment; out of these only a small fraction was of mitochondria were able to develop into branched structure 36 h after treatment. Majority of parasites in ClpQ(mut)–DD + Shld1 set showed no clear mitochondrial development. These parasites showed either diffused staining of MitoTracker dye or punctate staining with no clear organelle structure (Fig. 5A and B) at 24 and 36 h after treatment; these may represent under developed and/or dysfunctional mitochondria. These results suggest that development of functional mitochondria is affected in these parasites with dominant negative phenotype of inactive ClpQ protease. No significant difference was observed in development of mitochondria of Shld1 treated ClpQ–DD set as compared with control set (Figs S5 and S6).
Transcript levels of mitochondrial-encoded genes are reduced in transgenic parasites expressing inactive PfClpQ
We assessed any effect of disruption of the ClpQ machinery on the transcriptional regulation in the mitochondria of the parasites. The ClpQ(mut)–DD parasites at ring stages were treated with Shld1 or solvent alone for 24 h. The total cDNA from trophozoite stage parasite was assessed by quantitative PCR to assess any change in the transcript levels of mitochondrial genome encoded genes, cox 1, cox 3 and cyt b. The ClpQ(mut)–DD transgenic parasites treated with Shld-1 drug showed ∼ 80% reduction in the transcript levels of all three genes transcribed from mitochondrial genome as compared with untreated parasites (Fig. 6). To rule out possibility of global downregulation of transcription machinery, transcript level of nuclear encoded pfclpP gene was also studied. The pfclpP transcript levels shows no change in the ClpQ(mut)–DD Shld-1 treated parasites as compared with the control set, these results show specific downregulation of mitochondrial transcription (Fig. 6). To ascertain that the reduced transcription levels are not due to lack of mitochondrial DNA, we assessed levels of mitochondrial DNA in control and Shld1 treated parasites by quantitative PCR using total DNA samples. The genomic equivalents of the mitochondrial genome encoded genes, cox 1, cox 3 and cyt b showed only marginal variation in the Shld1 treated parasites as compared with control; main genome encoded pfFal-2 gene showed similar results (Fig. S7).
The metabolic pathways in the mitochondrion and the apicoplast, two parasite organelles of prokaryotic origin, are considered as suitable drug targets in the parasite. The P. falciparum ClpQ protease (PfClpQ), homologue of prokaryotic ClpQ, is one such potential drug target. The PfClpQ has a 37-amino-acid-long pro-domain so that after cleavage of the pro-domain, the conserved threonine residue, Thr (38) of active site triad becomes the first residue at the N-terminal of the mature protease. In addition other conserved residue Ser (160) forms the active site with this threonine residue. The ClpQY machinery is a multimeric system consisting of a small barrel of two stacked hexameric rings of ClpQ protease with the active site facing inside the barrel; the protease barrel is capped on one of both sides with a hexameric ring of AAA type ATPase, ClpY (Bochtler et al., 2000). The ATPases act as chaperons to unfold the substrate proteins which subsequently get degraded by the protease component. Only unfolded protein can enter in the protease barrel and comes out as small peptide after cleavage.
Attempts to knock out the ClpQ gene using double cross over homologous recombination were not successful (see supplementary data, Fig. S1A–F). These results may suggest that deletion of the pfclpQ gene may be lethal for the parasite. The pfclpQ gene locus is not recalcitrant for recombination, since the homologous recombination by single cross over has been obtained earlier for tagging of the native gene (Tschan et al., 2010). We earlier also showed that disruption of interaction of ClpQ protease with its ATPase partner ClpY using peptide based inhibitor caused severe affected parasite growth and survival (Rathore et al., 2011). These peptide based inhibitor affects were quite rapid which lead to onset of parasite death within 3–4 h of treatment (Rathore et al., 2011). To understand the cellular reason of this parasite cell death, we have now utilized a new system to transiently hinder the activity of ClpQ protease in the parasite and studied its cellular effects across the complete asexual parasite cycle.
In this study, we generated loss-of-function phenotypes in the parasite using the dominant negative approach. We first showed that mutation of the active site residues results in generation of proteolytically inactive protein, although this protein can efficiently bind with the substrate. We then episomally expressed the mutated pfclpQ gene in the transgenic parasites to produce the dominant negative phenotype. Since constitutive expression of inactive ClpQ could be lethal for the parasite, we used the recently developed DD system using a ddFKBP gene fusion (Armstrong and Goldberg, 2007; Herm-Götz et al., 2007; Dvorin et al., 2010) to regulate the protein levels in the parasite. The DD fusion proteins are unstable and gets rapidly degraded in the cell; whereas in presence of the ligand Shield-1 (Shld1) the degradation is mitigated and the fusion protein gets stabilized. Therefore, the fusion protein levels can be regulated using Shld1 drug. These systems have been utilized in regulating the episomally expressed transgene in the parasites (Armstrong and Goldberg, 2007); similarly trans-dominant negative effect of ddFKBP-tagged mutant Rab11 was studied in Toxoplasma gondii (Agop-Nersesian et al., 2009). Treatment of transgenic parasite lines ClpQ(mut)–DD with Shld1 drug showed stabilization of PfClpQ (mut) protein and confirmed its correct localization in the parasite mitochondria.
The transgenic parasites ClpQ(mut)–DD express both the nuclear encoded wild-type PfClpQ as well as episomally encoded inactive PfClpQ protease (PfClpQ-mut). The PfClpQ forms a multi-subunit complex in the parasite (Mordmüller et al., 2006), formation of this multi-subunit complex is essential to have a functional protease machinery in the cell; it is expected that the inactive PfClpQ subunits may also become part of the multi-subunit ClpQ protease barrel in the complete protease machinery. Our results show that indeed PfClpQ(mut) forms a complex with the native PfClpQ in the parasite. Further, the active site mutant protein is not able to cleave the substrate but is still able to bind with the substrates efficiently and compete with the wild-type protease for the substrate. These results suggest that the inactive PfClpQ(mut) protease subunit in the multiunit complex bind with the unfolded substrate as the normal protease subunit in the complex, but it is not be able to cleave these substrate proteins; thus the unfolded substrate would get stuck in the protease barrel and disrupt the functioning of complete machinery. Therefore, expression of the inactive protease in the parasite causes dominant negative effect. Indeed, for a related proteases machinery in E. coli, the ClpXP system, trans-expression of inactive ClpP protease in the cell is shown to trap unfolded protein substrate in the protease barrel (Flynn et al., 2003).
The Shld1 treatment showed significant growth inhibition of the transgenic parasite lines expressing mutant protein, ClpQ(mut)–DD, within first cell cycle of the parasite as compared with the control transgenic parasites lines ClpQ–DD and Vector–DD. The parasite developmental-stage profile showed that the development of trophozoites into schizonts was severely disrupted in the Shld1 treated parasites. Overall results of trans-dominant negative effect of proteolytically inactive PfClpQ showed that the PfClpQ protease plays functionally important role in parasite growth and development.
Parasite mitochondria play important role in critical biosynthetic processes including iron–sulfur cluster biosynthesis, haem biosynthesis and ubiquinone biosynthesis (Mather et al., 2007). The ClpQ protease is localized in the matrix of the mitochondria in P. falciparum (Tschan et al., 2010); similarly, ClpQ orthologue was also localized in the mitochondria in Trypanosoma brucei (Li et al., 2008). Since ClpQY is the only protease machinery localized in the mitochondria, it may be playing important role in the mitochondrial functioning of the apicomplexan parasites. Our results show that trans-dominant negative effect of PfClpQ caused disruption of mitochondrial development in the parasite, further it lead to significant loss of mitochondrial membrane potential (see supplementary data, Fig. S8). These results correlated with the parasite developmental studies; the parasite development was effected after the trophozoite/early schizont stages after disruption of PfClpQ function in these parasites. Loss of mitochondrial membrane potential is usually lethal for the eukaryotic cells. Inhibition of the ATP-dependent Lon protease in mitochondria of eukaryotic cells caused accumulation of aggregated proteins which subsequently lead to cell-death (Bota et al., 2005). Our study also shows that disruption of ClpQ, a potentially regulatory protease in the parasite, deregulates mitochondrial function which leads to death of the parasite.
The Plasmodium mitochondrion has a linear ∼ 6 kb long genome which encodes three proteins, Cox1, CytB and Cox3, as well as a number of rRNA subunits (Suplick et al., 1990; Feagin, 1992). However, there is no direct evidence of protein synthesis in the Plasmodium mitochondria; it has been suggested that the three proteins encoded by mitochondrial genome are synthesized by using unusual translation machinery or by sharing the apicoplast translation machinery (Vaidya and Mather, 2009). Silencing of ClpQ in T. brucei showed loss of control of replication of minicircle DNA in the mitochondria resulted in reduced parasite growth (Li et al., 2008). Plasmodium mitochondrion does not possess minicircle DNA and the ClpQY machinery may be involved in regulation of proteins involved in other metabolic pathways. The ClpQY machinery in E. coli is shown to control levels of transcription regulator σ32, which directs RNA polymerase to transcribe the heat shock proteins (Kanemori et al., 1997). The ClpQY machinery in E. coli is also shown to regulate a cell division inhibitor, SulA, and thus it is involved in regulation of cell cycle especially during stress (Seong et al., 2000). Similarly, other prokaryotic Clp proteases are also shown to regulate transcription factors in Bacillus subtilis, S. aureus and Mycobacterium tuberculosis (Turgay et al., 2001; Frees et al., 2003; Barik et al., 2010). Overall, the prokaryotic Clp proteases are known to play essential regulatory role during transcription in a precise manner rather than carrying out non-specific protein degradation. We assessed any possible role of PfClpQ protease in regulation of transcription of mitochondrial genome encoded proteins. Our results of quantitative estimation of transcripts showed that disruption of ClpQ protease system in the parasite due to dominant negative effect suppressed transcription of mitochondrial encoded ORFs. These results suggest possible role of ClpQY system in regulating levels of transcription factors in the mitochondria of the parasite. Indeed, ATP-dependent Lon protease in human mitochondria are shown to regulate the levels of mitochondrial transcription factor A (TFAM) and thus plays important role in transcription regulation (Lu et al., 2013). Further, we found that the mitochondrial transcription machinery was disrupted ∼ 24 h after the Shld1 drug treatment, whereas no effect was seen the mitochondrial membrane potential at that time point. The loss in mitochondrial membrane potential was observed at ∼ 36 h after Shld1 treatment. These studies indicate that disruption of ClpQ protease results in transcription suppression leading to deregulation of mitochondrial functioning. Loss of functional mitochondria hindered further development of the trophozoite stage parasites.
Overall, our results showed that the PfClpQ protease plays important role in development of functional mitochondria. We show that disruption of PfClpQ protease function results in deregulation of the transcription machinery in mitochondria which in turn hinders growth and development of parasite mitochondria; subsequently the mitochondria looses the membrane potential leading to death of the parasite. Our study establishes PfClpQ as a potential drug target in the parasite. Regulation of transcription and translation in the parasite mitochondrion is not fully understood, further studies to understand these pathways may lead to development of novel antimalarial strategies.
Parasite culture, plasmid construct for gene knock-out and parasite transfection
Plasmodium falciparum strain 3D7 was cultured with human erythrocytes (4% haematocrit) in RPMI media (Invitrogen) supplemented with 10% O+ human serum using a protocol described previously (Trager and Jensen, 1976). Parasite cultures were synchronized by repeated sorbitol treatment following Lambros and Vanderberg (1979). To generate pCC1 transfection vector construct, a 5′ fragment of pfclpQ gene loci (Gene ID: PF3D7_1230400) was PCR amplified using primer set 779A (forward) and 780A (reverse) (Table S1) with P. falciparum 3D7 genomic DNA as template and cloned upstream of the hDHFR cassette in pCC1 vector (Maier et al., 2006) using restriction sites SacII and SpeI to give the construct pCC1-ClpQ-5′. The 3′ fragment of pfclpQ gene loci was PCR amplified using primer set 781A (forward) and 782A (reverse) and cloned downstream of the hDHFR cassette in pCC1-ClpQ-5′construct using restriction sites EcoRI and NcoI to give the construct pCC1-ClpQ (Fig. S1).
Synchronized P. falciparum 3D7 ring stage parasites were transfected with 100 μg of purified plasmid DNA (Plasmid Maxi Kit, Qiagen, Valencia, CA) by electroporation (310 V, 950 μF) (Crabb et al., 2004) and the transfected parasites were selected over 2.5 nM of WR99210 drug. The P. falciparum 3D7-pCC1-ClpQ parasite line was then subjected to repeated WR99210 drug cycling. Each drug cycle was for 42 days including 21 days in the absence of the drug followed by 21 days under drug pressure. The genome of parasites after every drug cycle (2–5 drug cycles) were analysed by PCR using primer combinations: 838A-839A; 842A-841A; 842A-354A, annealing location of each of the primer in the pfclpQ gene locus or transfection plasmid construct is indicated in Fig. S1A and sequence of primers is given in Table S1.
Transgene expression of inactive ClpQ and P. falciparum growth analysis
To generate pHADD transfection constructs the pfclpQ gene (–15 to 621 bp) with mutations in the codons of the active site residues Thr (38) (A112G) and Ser (160) (T477G) to code for Alanine was cloned. This gene fragment was cloned into pHADD vector between XhoI and KpnI restriction sites to generate the plasmid construct pHADD-ClpQ (mut) (Fig. 2A). In another construct wild-type pfclpQ gene was cloned in similar manner to generate the plasmid construct pHADD-ClpQ, which was used as control. Synchronized P. falciparum 3D7 ring stage parasites were transfected with pHADD-ClpQ(mut), pHADD-ClpQ or pHADD vector as described above and the transfected parasites were selected over 2.5 nM of WR99210 drug; the transgenic parasites obtained are labelled as ClpQ(mut)–DD, ClpQ–DD and Vector–DD respectively. For growth study assay the cultures were synchronized by two sorbitol treatments at 4 h interval. Parasite growth inhibition assays were carried out in 24-well plates using synchronized parasite cultures at ring stage. Each assay was performed in triplicate. Each well was containing 4% haematocrit 0.5 ml of complete media supplemented with 1 μM Shld1 and the parasitaemia was adjusted to ≤ 0.5%. A parallel set was maintained for each parasite line without Shld1. Smears were made from each well at different time points (0, 24, 48, 72, 96 and 120 h after treatment), stained with Giemsa, and the numbers of ring stage parasites per 5000 RBCs were determined and percentage ring stage parasitaemia was calculated to assess the parasite growth for all the sets, ClpQ(mut)–DD + Shld1, ClpQ(mut)–DD solvent control, ClpQ–DD + Shld1, ClpQ–DD treated with solvent alone, Vector–DD + Shld1 and Vector–DD treated with solvent alone.
Isolation of total DNA, RNA, cDNA synthesis and quantitative real-time PCR
The genomic DNA was isolated from in vitro culture of P. falciparum following standard protocol. To assess any change in transcript levels of different parasite genes, P. falciparum ClpQ(mut)–DD parasite line parasites treated with Shld-1 (1 μM) or solvent alone for 24 h. Total RNAs were isolated from the parasite cultures as well as from synchronized 3D7 parasites culture using mini RNA isolation kit (Qiagen). Each sample was treated with DNase (Invitrogen) prior to reverse transcription in order to negate the possibility of amplifying genomic DNA. An aliquot of 50 ng of total RNA was used to synthesize cDNA using cDNA synthesis kit (Bio-Rad) following manufacturer's recommendations. Gene specific primers were designed using Beacon Designer4.0 software, for the genes pfclpP (Gene ID: PF3D7_0307400; primers 589A and 590A), pfcox 1 (Gene ID: mal_mito_2; primers 903A and 904A) and pfcox 3 (Gene ID: mal_mito_1; primers 901A and 902A), pfcyt b (Gene ID: mal_mito_3; primers 905A and 906A); 18S rRNA control primers (18SF and 18SR) were used following Blair et al. (2002).
Change in expression level for genes, pfcox 1, pfcox 3 and pfcyt b, transcribed from mitochondrial genome as well as the nuclear pfclpP gene were assessed by quantitative real time PCR. Each PCR was carried out in triplicate using the iCycler version 3.0 (Bio-Rad); each reaction was containing equal amount of cDNA, 100 ng of both the gene specific primers and 1× SYBR Green PCR mix (Bio-Rad). Threshold cycle (Ct) values were calculated by using iCycler software. Standard curves for each gene were obtained by using different dilutions of wild-type gDNA (100 to 1 ng) as template, and these standard curves were used to determine genome equivalents of Ct values for respective gene and 18S rRNA in each RNA sample (Blair et al., 2002). Genome equivalents of each gene were normalized using that of 18S rRNA for all the RNA samples.
Expression plasmid constructs, expression and purification of recombinant protein and protease activity assays
The recombinant PfClpQ protein corresponding to mature region of PfClpQ protein (112–621 bp; 38–207 aa) was made as previously described (Ramasamy et al., 2007). For active site mutant protein, similar PfClpQ gene fragment (112–621 bp; 38–207 aa) corresponding to the mature part of the protease with mutated active site residues, Thr (38) to Ala and Ser (160) to Ala, was cloned in pET29a vector between NdeI and HindIII sites to give the construct pET 29a-PfClpQ-M. The recombinant plasmids were transformed into Escherichia coli expression cells BL21 (DE3) for expression of recombinant proteins with C-terminal histidine tag. These cells were grown in Luria broth containing kanamycin (25 μg ml−1) and expression of the recombinant protein was induced with isopropyl-β-thiogalactopyranoside (IPTG) at a final concentration of 1 mM. The recombinant proteins were purified from E. coli cell lyste was purified by affinity chromatography as described earlier (Ramasamy et al., 2007). The eluates were analysed on SDS-PAGE and the fractions containing the recombinant protein with a clear single band were pooled and the protein concentration was determined using the Pierce BCA (bicinchoninic acid) protein assay system and a standard curve of bovine serum albumin.
Fluorometric assays for the protease activities were carried out in 200 μl reaction volume containing 0.8 μM of either of the recombinant protein in assay buffer (20 mM Hepes pH 8.0, 10 mM MgCl2, 2 mM CaCl2). The fluorogenic peptide substrate Suc-GGL-AMC was added at 80 μM final concentration and the release of AMC was continuously monitored as the increase of fluorescence (excitation 355 nm; emission 460 nm) for 1–6 h at room temperature using a Victor-3 Fluorometer (Perkin-Elmer).
To study competition between wild-type PfClpQ and PfClpQ(mut) for the available substrate, the protease activity assays were carried out using equimolar concentration of each of the recombinant protein. Activity in these reactions was expressed as percentage activity as compared with a parallel assay with PfClpQ alone. In control reactions, PfClpQ(mut) was replaced with non-specific recombinant protein.
To study in vitro interaction of wild-type PfClpQ and PfClpQ(mut), both the proteins were coexpressed in E. coli. The wild-type PfClpQ was cloned in NdeI and HindIII site of pET29a to express with 6× histidine tag (pET-ClpQ construct) and PfClpQ(mut) was cloned in NdeI and XhoI sites in pETDuet-1 to express with S-tag [pETDuet-1-ClpQ(mut) construct]. Both the plasmids were transformed into E. coli expression cells BL21 (DE3) for expression of recombinant proteins as described above.
Co-immunoprecipitation of PfClpQ multi-subunit complex in the parasite lysate and Western immunoblotting
Infected RBCs were collected from ClpQ(mut)–DD parasite culture treated with Shld1 or solvent alone. The parasites were harvested by lysis of the infected RBCs by 0.15% saponin in RPMI for 10 min on ice. Parasite pellet was suspended in TEN buffer (Tris 50 mM pH 7.4, 20 mM EDTA) containing protease inhibitor cocktail (Roche) and then lysed by three freeze-thaw cycles. The lysate was clarified by centrifugation at 12 000 g for 30 min at 4°C and pre-cleared with Protein-A agarose beads. The pre-cleared parasite extract was incubated with Protein-A agarose beads having immobilized anti-HA antibodies (Roche) overnight at 4°C with gentle agitation. After incubation the beads were washed extensively in TEN buffer containing 0.1% Tween-20. The immune complex were dissociated by boiling the beads in SDS-PAGE buffer and centrifuged to separate the beads. The supernatant was separated on SDS-PAGE and analysed by immunoblotting using anti-PfClpQ antibodies.
Western blot analyses were carried out using parasite lysate or purified protein complexes. Parasite pellets were washed with PBS, suspended in Laemmli buffer, boiled, centrifuged. The supernatant obtained from parasite lysate or the purified proteins were separated on 12% SDS-PAGE. The fractionated proteins were transferred from gel onto the nitrocellulose membrane (Amersham) and blocked in blocking buffer (1× PBS, 0.1% Tween-20, 5% milk powder) for 2 h. The blot was washed and incubated for 1 h with primary antibody [rabbit anti-PfClpQ (1:1000)] diluted in dilution buffer (1× PBS, 0.1% Tween-20 and 1% milk powder). Later, the blot was washed and incubated for 1 h with anti-rabbit secondary antibody (1:2000) conjugated to HRP, diluted in dilution buffer. Bands were visualized by using ECL detection kit (Amersham).
Fluorescence microscopy and indirect immunofluorescence assay
Plasmodium falciparum ClpQ(mut)–DD transgenic parasites were synchronized by two consecutive sorbitol treatments 4 h apart. The ring stage parasite cultures were treated with 1 μM of Shld-1 drug or solvent alone and allowed to grow further for 24 or 36 h. Parasites at different developmental stages were collected from the culture for fluorescence microscopy and stained with DAPI at a final concentration of 2 μg ml−1 for 30 min at 37°C prior to imaging. To visualize the mitochondria, the transgenic parasites were stained with MitoTracker Red CMXRos (Invitrogen) at a final concentration of 20 nM in 1× PBS for 15 min at 37°C., the cells were subsequently fixed in 4% paraformaldehyde.
Indirect immunofluorescence assays were performed on P. falciparum ClpQ–DD parasite lines as described earlier (Wickramarachchi et al., 2008). Briefly, the Shld-1 treated parasite samples were stained with MitoTracker and fixed with PFA as described above. Fixed parasites were incubated with rat anti-HA (1:100) antibodies (Roche) and subsequently with FITC-linked goat anti-rat antibodies (1:100, Sigma) as secondary antibody with intermittent washing. The parasite nuclei were stained with DAPI (2 μg ml−1). The parasites stained with MitoTracker or by immunostaining were viewed using a Nikon TE 2000-U fluorescence microscope.
We are grateful to Guy Schiehser and David Jacobus for the drug WR99210; Alan Cowman and Alex Maier for pCC1 and pHADD vectors. We thank Rotary blood bank, New Delhi for providing the human RBCs. S.J. and G.D. are supported by research fellowships from CSIR, Government of India. S.R. and M.A. are supported by research fellowships from ICMR, Government of India. M.E.H. is supported by international research fellowship from ICGEB. The research work in AM's laboratory is supported by Program Support Grant from Department of Biotechnology, Government of India, and Indo-Swiss Joint Research Programme from Department of Science & Technology, Government of India. A.M. is a recipient of National Bioscience Award for Career Development from Department of Biotechnology, Government of India.