A calpain unique to alveolates is essential in Plasmodium falciparum and its knockdown reveals an involvement in pre-S-phase development

A calpain unique to alveolates is essential in Plasmodium falciparum and its knockdown reveals an involvement in pre-S-phase development Ilaria Russo, Anna Oksman, Barbara Vaupel, and Daniel E. Goldberg1 Departments of Medicine and Molecular Microbiology, Howard Hughes Medical Institute, and Washington University School of Medicine, 660 South Euclid Avenue, St. Louis, MO 63110 Edited by Anthony Cerami, Warren Pharmaceuticals, Ossining, NY, and approved November 21, 2008 (received for review July 18, 2008)

Plasmodium falciparum encodes a single calpain that has a distinct domain composition restricted to alveolates. To evaluate the potential of this protein as a drug target, we assessed its essentiality. Both gene disruption by double cross-over and gene truncation by single cross-over recombination failed. We were also unable to achieve allelic replacement by using a missense mutation at the catalytic cysteine codon, although we could obtain synonymous allelic replacement parasites. These results suggested that the calpain gene and its proteolytic activity are important for optimal parasite growth. To gain further insight into its biological role, we used the FKBP degradation domain system to generate a fusion protein whose stability in transfected parasites could be modulated by a small FKBP ligand, Shield1 (Shld1). We made a calpain-GFP-FKBP fusion through single cross-over integration at the endogenous calpain locus. Calpain levels were knocked down and parasite growth was greatly impaired in the absence of Shld1. Parasites were delayed in their ability to transition out of the ring stage and in their ability to progress to the S phase. Calpain is required for cell cycle progression in Plasmodium parasites and appears to be an attractive drug target. We have shown that regulated knockdowns are possible in P. falciparum and can be useful for evaluating essentiality and function. cell cycle 兩 cysteine protease 兩 essentiality 兩 inducible 兩 malaria

P

lasmodium falciparum, the causative agent of severe malaria, has a complex life cycle involving numerous biological events, such as migration, host cell invasion, metabolism, and cell cycle progression and egress. Cysteine proteases have been implicated in most of these processes (1), but the function of many of the estimated 35 members of this protease class encoded in the genome has not been defined. Among them a single calpain-like protease is recognized (2). Calpains belong to the C2 family of cysteine proteases. Calpains are found in organisms from bacteria to mammals and exhibit great divergence of sequence and domain structure but have homologous catalytic domains. Calpains have been implicated in diverse processes (3–6), such as muscle function, cell signaling, migration and attachment, death, transformation, cell-cycle regulation, differentiation and development, and fungal alkali tolerance, although the precise physiological role of many of these calpains is still poorly understood (3, 7). Calpains can be divided into typical (resembling calpain 1) and atypical (lacking domain IV) (4, 5). In typical calpains, domain IV mediates Ca2⫹ binding and consequently activity regulation and dimerization (8). In P. falciparum, only 1 calpain gene (Pcalp) is apparent in the genome (2), and it is atypical. It has an unusually long coding sequence, and outside of the catalytic domain, has only distant homology with characterized calpains. We found that Pcalp contains several subdomains that are highly conserved in related species and confer uniqueness to the protein. One of them determines nucleolar localization, and another regulates calpain movement in and out of the nucleus through reversible palmitoylation (I.R., unpublished data). Surprised by the location of Pcalp, we 1554 –1559 兩 PNAS 兩 February 3, 2009 兩 vol. 106 兩 no. 5

undertook a genetic characterization to assess its essentiality and gain insight into its function in the cell. We report that Pcalp is a distinct type of calpain restricted to alveolates. Using multiple approaches, we show that the gene is essential to intraerythrocytic parasites. By developing a conditional knockdown system, we have been able to examine gene function. Pcalp plays a critical role in cell cycle progression during trophozoite development. Results Pcalp Is a Distinct Type of Calpain Restricted to Alveolates. We

identified a single calpain gene (Pcalp) in P. falciparum, MAL13P1.310 locus. The predicted coding region is 6,147 bp, encodes a 242-kDa protein, and was confirmed by extensive sequence analysis of the genomic DNA and cDNA amplified from mature mRNA (SI Materials and Methods). Orthologs to Pcalp were found in all Plasmodium species for which sequence data are available (Table S1). By phylogenetic analysis we confirmed that Pcalp has greatest homology to Aspergillus PalB (2), limited to the catalytic domain. However, the low reliability of its node derivation indicates substantial divergence (Fig. 1). In fact, aside from a central calpain catalytic domain (IIa-b) and a C-terminal domain homologous to calpain domain III, Pcalp possesses a long N terminus that has no significant homology to any known protein. Pcalp is a distinct type of calpain for its domain composition, not common to any other identified calpain class (Fig. 1). All sequenced apicomplexan parasite genomes encode a single calpain with N-terminal homology, although for some of them the gene predictions result in a shorter encoded N terminus (Fig. 1). Two ciliate calpains, from Tetrahymena and Paramecium, also have N termini with some homology to Pcalp. Bootstrap analysis and domain composition indicate that Pcalp belongs to a distinct clade of calpains that is restricted to Apicomplexa and other alveolates (Fig. 1). The diatom Thalassiosira pseudonana, which belongs to the chromalveolate kingdom, has an encoded calpain with distant N-terminal homology to the alveolate sequences (not shown). Pcalp Is Expressed in Early Intraerythrocytic Stages. Because pub-

lished Pcalp mRNA assessments differ from each other (9, 10), we analyzed cDNA by semiquantitative RT-PCR through asexual development. Amplifying both the N-terminal and C-terminal Author contributions: I.R. and D.E.G. designed research; I.R., A.O., and B.V. performed research; I.R. and D.E.G. analyzed data; and I.R. and D.E.G. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession nos. EF432831, EF432832, and EF432833). 1To

whom correspondence should be addressed. E-mail: [email protected].

This article contains supporting information online at www.pnas.org/cgi/content/full/ 0806926106/DCSupplemental. © 2009 by The National Academy of Sciences of the USA

www.pnas.org兾cgi兾doi兾10.1073兾pnas.0806926106

A

0.05

68 77

62

63 63

58 95

72 50

40

100

37

80 52 98 48

66 100

62 26 64 58 56 91 18

80 62

99

Domains-> N-terminal Catalytic IIa IIb III Drosoph_CalpC Schisto_Calp Human_Calp9 Human_Calp3 Murine_Calp8 Human_Calp2 Human_Calp1 Human_Calp11 Drosoph_CalpA Drosoph_CalpB C.elegans_clp-7 Human_Calp12 Human_Calp13 C.elegans_Tra3 Human_Calp5 Human_Calp6 Human_Calp10 ZnF_RBZ SOLH Drosoph_CalpD Human_SOLH TMs Arabid._Dek1 LyM Entamoeba_Calp Trypanosome_Calp MIT PBH Human_Calp7 C.elegans_clp-2 PCalp Asperg_PalB C.neo_PalB Candidal_PalB S.cerevisiae_Rim13 HemolysinCabind Gloeob._Glr4140 P.ging_Tpr C.elegans_F44F1.3

B IV 0.05

Domains-> N-terminal Catalytic IIa IIb III P. falciparum

100 100

P. reichenowi P. gallinaceum

99 55

100

P. chabaudii P. yoelii

100

P. berghei P. vivax

100 100

C2

P. knowlesi N. caninum

100 100

T. gondii E. tenella

98 70

B.bigemina

97 100

T. annulata T. parva

C.parvum 100

P. tetraurelia T. therm.

?

Fig. 1. Phylogenetic analysis reveals Pcalp as a distinct type and clade of calpains. (A) Phylogenetic analysis of Pcalp (catalytic domain) aligned with representatives of each documented type of calpain is shown (left side) together with the domain composition of each protein (right side). Bootstrap values are shown at the tree joints. Domains are color coded and from top to bottom named at their first appearance (SMART code). The Pcalp N terminus has subdomains conserved among Plasmodium species (green) (I.R., A.O., and D.E.G., unpublished data). Scale is fractional change per unit distance. (B) Phylogenetic analysis and domain composition of alveolate calpains. Crossed boxes denote incomplete sequence information. Sequence data are listed in Tables S1 and S2 and alignments in Figs. S1 and S2.

coding regions and normalizing against actin, we found a peak of induction (2⫻) in the late ring/early trophozoite stages [18–24 hours postinvasion (hpi), Fig. 2A], consistent with the data of Bozdech et al. (9). Despite very low expression levels, we were able to detect Pcalp by Western blotting after immunoprecipitation from a large number of parasites (5 ⫻ 108). We used a parasite line (C3) in which the endogenous calpain ORF is tagged with GFP (see Materials and Methods), so that the anti-GFP antibody could be used for visualization. The asexual cycle was monitored for morphology and DNA content (Fig. 2B). Pcalp was detected from rings through late trophozoites, showing a peak in the amount per cell in midtrophozoites and in concentration (normalized to BiP) in rings (Fig. 2 C and D). In these experiments, the majority of the cells entered the

S phase around 30 hpi (Fig. 2B), corresponding to the peak amount of Pcalp. The Pcalp Gene and Its Activity Are Essential for Optimal Growth. We

assessed the essentiality of Pcalp for optimal P. falciparum asexual growth by using multiple approaches (Fig. 3A). The first one is a classical gene knockout approach using double cross-over with positive/negative selection (11). We used a double positive selection by fusing HcRed to hDHFR and obtained red fluorescent parasites that stably maintained plasmid under drug pressure (Fig. S4). However, we were unable to recover integrants after negative selection despite multiple attempts, assayed by Southern blotting (Fig. 3B) and nested PCR (data not shown). For gene disruption by 5⬘ single cross-over, we transfected 3D7 parasites with constructs containing 1.3 or 1.6 kb of the 5⬘ end of

Fig. 2. Expression profiles of Pcalp along the asexual ER LR ET MT MT/LT Sc 12 cycle. (A) Semiquantitative RT-PCR analysis. Expression IP: R- -PCalp 10 WB: M- -GFP of Pcalp mRNA in asexual stages was measured by 8 kDa quantifying amplified regions encoding the N (light 6 250 gray) and C (dark gray) termini (map of region in Fig. 4 75 S3). The signals were normalized to actin. ER and LR, 2 IP inputs early and late rings; ET and LT, early and late trophoWB: R - -BiP 0 hpi 25 29 33 43 9 19 LT/Sc Sc ER LR ET LT zoites; Sc, schizonts. (B) Time course of Pcalp-GFP (clone C3) cycle. Parasites were sampled every 30 min Sc ER LR ET MT MT/LT 100 and analyzed by flow cytometry for DNA content. 8 90 7 Populations selected for the analysis are shown in Fig. 80 6 70 S3. The percentage of total parasites in pre-S phase, S 5 60 phase, and mature schizogony are represented by 50 4 black, red, and green lines, respectively. Zero time 40 3 30 corresponds to the first peak of mature schizonts. Dots 2 20 indicate harvest times for protein analysis (MT, midtro1 10 0 0 phozoites) and the arrow marks the approximate start 8 12 16 20 24 28 32 36 40 44 10 15 20 25 30 35 40 45 50 55 0 -5 5 time (hpi) time (hpi) of S phase. (C) C3 parasites were harvested at the indicated time points, Pcalp was immunoprecipitated with rabbit anti-Pcalp antiserum and then protein was blotted with mouse anti-GFP antibody. BiP content was also measured. Each lane corresponds to equivalent cell numbers at different stages. (D) Pcalp levels were quantified from blots and plotted with time as absolute values (red) and ratio of calpain to BiP (black). The first point of the ratio is omitted because of minimal BiP expression in early rings.

C

calpain/actin rel intensity

A

Russo et al.

PNAS 兩 February 3, 2009 兩 vol. 106 兩 no. 5 兩 1555

MICROBIOLOGY

D

arbitrary units

% of total parasites

B

S

nat 5’ N-term

1

GFP Ahsp86-3’ N

,hDHFR

Ci

Pcalp 3'

NS(I)

NS(II)

S

nat 5’

D a.

2

1

*

S

4

4

3D7

Q G F V/V A D C/C S F L S S L T CAAGGTTTTGTTGCGGATTGTTCTTTTCTGTCTTCATTAAC C C S

-

b.

+

-

+

-

+

-

Q G F V A D C S F L S S L T CAAGGTTTTGTTGCGGATTGTTCTTTTCTGTCTTCATTAAC

+

kb 1.6 1.0 0.8 0.5 0.3 0.1 BstUI

ii

v

7

iv

3D

5

kb 8 7 6 5

P

Pcalp

iii

7

kb 8 7 6

3D

m

ns f.

as

pl

S

ii

2

N P

iv

*

BstUI

iii

kb 11 9 8 7 6 5

X

TK

ii

Bi

3’ region

7

HcRed-hDHFR

tr a

5’region

3D

A i

kb 8 7 6 5 4

NS

Plasmids 3D7

S

1

-

iii

+

-

+

-

+

-

kb 1.6 1.0 0.8 0.5 0.3 + BstUI

clone 1 clone 2 clone 3 clone 4

2

NS

-

+

-

+

-

+

-

kb 1.6

1.0 0.8 0.65 0.5 + BstUI

Fig. 3. Attempted genetic strategies to disrupt calpain functionality. (A) Diagrams of the vectors. (i) For double cross-over gene knockout, a positive selection cassette containing HcRed-hDHFR fusion and flanked by Pcalp ORF 5⬘ and 3⬘ ends was inserted into a thymidine kinase (TK) negative selection vector. (ii) For single cross-over gene disruption, a sequence homologous to Pcalp ORF 5⬘ end was cloned upstream to hDHFR cassette. Two different lengths (1.3 and 1.6 kb) were used. Control constructs are identical but contain 733 bp of upstream sequence (nat 5⬘, brackets). (iii) For allelic replacement, a sequence homologous to 3.8 kb of the Pcalp ORF 3⬘ end was used for homologous recombination. Two plasmids were created. Both have a mutation creating a new restriction site (BstUI). Two codons downstream in the active site cysteine codon (*), one has a synonymous change, whereas the other has a nonsynonymous change. Predicted products of integration of ii and iii are shown in Fig. S4. (iv) For reference the Pcalp gene is shown. Primers used for PCR/restriction in the allelic replacement strategy are indicated (semiarrows). Restriction sites: X, XmaI; N, NsiI; S, SphI; P, PacI. (B) Southern blot analysis. (i) Double cross-over. Transfected parasite DNA (X/P digested) shows the uninterrupted endogenous gene (open arrow) and episomal plasmid (gray arrow), but no evidence of any integration event (filled arrows). Each panel is a different exposure time. (ii–v) Single cross-over. DNA was restricted with N/S. All 4 transfections with the control vector show the predicted integrations (ii and iii; panels are from 2 blots). None of the 4 transfections with the vectors for gene disruption (iv and v) have evidence of integration events (panels are from the same blot). Arrows as in i. 1.6-kb 5⬘ ORF in ii and iv; 1.3-kb 5⬘ ORF in iii and v; nat 5⬘ in ii and iii. 3D7, genomic DNA from the parental isolate. (C) PCR and restriction screening. DNA was prepared from parasite culture transfected with the synonymous (S) or nonsynonymous (NS) vector and selected for integrants. (i) PCR was performed using primer 2 to amplify the transcribed locus but not plasmids. Product was incubated with (⫹) or without (⫺) BstUI (below) and arrows indicate the expected fragments. The S vector transfectants showed evidence of upstream cross-over (introduction of the restriction site), but neither of the NS vector transfectant pools did. (ii) PCR was performed using primer 1 to amplify all copies (including plasmid) from transfectants as well as isolated plasmids (1:1 mix of S and NS). Both isolated and transfected episomal plasmids are restrictable. 3D7, as in B. (iii) Examples of isolated S clones. PCR was performed with primer 2; 1 of the 4 clones shown has crossed-over upstream. (D) Sequencing of PCR products amplified from S (i) and NS (ii) integrant pools. Results of control sequence analysis of plasmids and wild-type genomic amplicons are in Fig. S4.

Pcalp (Fig. 3Aii). As controls we used a second pair of constructs, identical to the previous ones but including an additional 733 bp upstream of the Pcalp ORF. These controls, upon integration, recapitulate a functional 5⬘ UTR ahead of the full Pcalp ORF, whereas the disruption constructs are designed to generate a promoterless ORF in addition to the truncated copy. All transfections were repeated twice and each resulted in transformants that were subjected to selection for integrants. Although the promotercontaining constructs showed clear integration (Fig. 3Bii and iii), no integration was detectable for the promoterless constructs (Fig. 3Biv and v). These results strongly suggest that Pcalp is essential for parasite growth. The data also imply that the 733 bp upstream of the ORF is adequate to drive calpain expression and to maintain viability. To further assess Pcalp essentiality and determine if proteolytic activity of the enzyme is required, we generated 2 almost identical allelic replacement plasmids carrying the 3⬘-most two-thirds of the Pcalp ORF (Fig. 3Aiii). Two mutations were made. The first, common to both plasmids, was a silent mutation that creates a new restriction site. The second mutation, 2 codons downstream at the catalytic cysteine triplet, was either synonymous (S) or nonsynonymous (NS) (Cys⬎Ala). We transfected 3D7 parasites and selected for integrants. Cross-over can occur upstream or downstream of the catalytic cysteine codon. Assuming that cross-over is random over the 3,815 bp of homologous sequence, the probability of integration upstream of the new restriction site is 22%, downstream of the active site codon is 78%, and between them is 0.16%. When we specifically PCR-amplified the transcribed calpain locus, the only BstUI-restrictable product derived from the S mutation integrant pool (Fig. 3Ci), indicating that upstream cross-over occurs only when the active site cysteine is not changed. We also directly sequenced the amplicons and did not detect NS mutation at the 1556 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0806926106

active site, whereas about one-third of the S mutation pool had an alteration at this codon (Fig. 3D). As a control, we amplified and digested the active site region of the parental 3D7 genome, and the 2 plasmids as isolated DNA and as episomes after transfections (Fig. 3Cii). We confirmed the absence of restrictability in the parental genome, the introduction of the BstUI sensitivity due uniquely to the plasmids, and the similarity in DNA content among the selected transfectants. We completed our study by isolating clones from each of the transfections to better quantify the proportion of upstream integrants. We isolated nearly 100 clones and screened for PCR product restrictability, as described above (examples in Fig. 3Ciii). Out of 62 NS mutation plasmid-transfected clones, we detected none with upstream integration, whereas ⬇21% of those isolated from the S mutation transfection carried a newly introduced restriction site and consequently a mutated cysteine codon. These data indicate that the essentiality of Pcalp depends on its proteolytic activity. Results of all attempts are summarized in Table 1. Conditional Knockdown of Pcalp. Since disruption of the calpain

gene was unachievable, to study the function of the protein in vivo we attempted to generate a conditional knockdown, taking advantage of the FKBP destabilization domain system (12). The 10-kDa FKBP domain, when fused to a protein of interest, targets it for degradation. However, in the presence of the small-molecule FKBP ligand, Shld1, degradation is mitigated. This system has been shown to work in Plasmodium for episomally expressed constructs (13). To create a conditional protein knockdown, we tagged the Pcalp C terminus with GFP-FKBP by homologous recombination at the 3⬘ end of the endogenous locus (Fig. 4A). After drug cycling with culture constantly exposed to 0.5 ␮M Shld1, we obtained successful integration (Fig. 4B). Clones were isolated and a representative one, A7, was chosen for further study. Russo et al.

Table 1. Calpain gene disruption attempts

Double cross-over Single cross-over Allelic replacement

Purpose

Outcome

Attempts

Gene knockout Gene recapitulation Gene disruption Synonymous mutation Nonsynonymous mutation

0 integrations 4 integrations 0 integrations 7 mutations/33 clones 0 mutations/62 clones

6 4 4 1 2

We analyzed the amount of Pcalp-GFP-FKBP obtained from synchronized midtrophozoites grown in the presence or absence of Shld1. Despite the very low expression level, we were able to visualize a drop in calpain upon removal of Shld1 (Fig. 4C). The experiment was repeated 3 times and was reproducible. The immunoprecipitation was judged successful in the Shld1-minus lane because the IgG precipitated was comparable in amount to that in the Shld1-plus lane. There were no low molecular weight GFPpositive bands that might have indicated degradation during sample workup (not shown). As a further control we used C3, a Pcalp-GFP integrant that lacks the degradation domain. Calpain content in the control cells was not affected. These results show that this strategy can be used to create conditional knockdowns in P. falciparum. Evidence for an Important Regulatory Function of Pcalp in G1. Growth of asynchronous Pcalp-GFP-FKBP-expressing clones was dependent on Shld1. Whenever we removed Shld1 from the cultures we observed a growth inhibition of 40–60% over 4 days (Fig. 5 A and B). The control culture (Pcalp-GFP, C3) showed only a slight long-term toxicity from the drug. With increasing concentrations of Shld1, growth improved, reaching a maximum at ⬇0.4 ␮M (Fig. 5C), which remained constant up to 1 ␮M (data not shown). The phenotypic analysis was extended to synchronized cultures of A7 and C3 clones. Shld1-maintained early schizonts were split into 2 portions (with or without Shld1) and closely monitored through the cell cycle by flow cytometry (Fig. 5 D and E). No differences in invasion efficiency were observed. However, in the A7 culture without Shld1, a growth defect was evident at the trophozoite stage (⬇20% decrease in parasitemia). The loss of parasitemia was maintained through the rest of the cycle and was detected again in the next cycle. No such defect could be detected in the control culture. To better understand the drop in parasitemia during the ring/ trophozoite transition, we tested the viability of the A7 and C3 cultures in this part of the cycle. Three different measurements were made by using combinations of dyes that stain metabolically active cells (CM green), live and dead cells (Syto-59 and Acridine Orange [AcOr]), or dead cells only (Topro3). The signals of each pair of dyes (Syto⫹CM, Topro⫹CM, Topro⫹AcOr) were detected on FITC and APC channels and compared with the staining obtained

A

B

N

N

S

hDHFR N N C-term

Integration S

Discussion We have characterized a distinct type of calpain that is found only in Apicomplexa and other alveolates. It has a unique N-terminal domain, which constitutes half of the entire protein. The marked divergence from mammalian calpains and its presence in a number of clinically relevant apicomplexans make this enzyme an attractive target to be studied.

3D7 Sel1 Sel2 Sel3

C IP: R-

clones

500bp

kb 6.0 5.0

N

4.0

Pcalp S

in fixed cells. We calculated the ratio of the percent of dead parasites detected without versus with Shld1 (Fig. 5F). For A7, the amount of dead cells in the absence of Shld1 was consistently ⬇1.5 times more than for the same culture maintained in Shld1, whereas for the C3 control culture the ratio was steady at ⬇1.0. This difference was confirmed by each of the 3 techniques and correlates with the decreased parasitemia detected. We then analyzed by microscopy the appearance of the dead parasites. Topro3-positive parasites, detected in all cultures with or without Shld1, were mostly mature schizonts that stochastically did not egress properly (data not shown). Surprisingly, when A7 calpain was destabilized by removal of Shld1, we detected the appearance of another type of dead cell (Fig. 5G). These cells were trophozoites that appeared morphologically normal but either extraerythrocytic or inside RBC ghosts, indicating a failure of homeostasis during development. The surviving parasites in the A7 culture without Shld1 do not look morphologically abnormal but are slightly delayed in the transition from ring to trophozoite (Fig. 5H). Therefore, we did careful monitoring of progression through the asexual cycle. Using an automated collector, we harvested and fixed a synchronized culture of A7 every 30 min along multiple cycles. The samples were analyzed for DNA content (Fig. S5) and counted as a percent of parasites in G1 (1N), in S phase (2–15N), or late schizonts (⬎15N). Fig. 5I shows 2 representative cycles of culture in the presence (Upper) or absence (Lower) of Shld1. Although no substantial difference in cycle length was detected, the start of S phase in the absence of Shld1 was consistently delayed by 3–3.5 h. This correlates with the peak of enzyme expression (Fig. 2). Our data indicate that Pcalp knockdown leads to cell death or a delay in transitioning out of G1 phase in trophozoite development.

2.0 GFP-FKBP 1.0

-

C3 + Shield-1

250 IP inputs:

N

+

kDa

3.0

GFP-FKBP

probe

-

-GFP WB: M- -GFP

A7

75 -BiP -PM2 25

Fig. 4. Knockdown of Pcalp. (A) Creation of a calpain-GFP-FKBP chimera by homologous recombination. The diagram shows the strategy to create C-terminally tagged calpain by integration at the endogenous locus. The plasmid contains sequence from the Pcalp ORF 3⬘ end in frame with GFP-FKBP. Relative positions of NsiI (N) and SphI (S) restriction sites, and the probe are indicated. (B) Southern blot of N/S restricted DNA from the 3 drug cycles (Sel1–3) and 3 clones. Arrows: endogenous gene (black), plasmid (gray), and modified calpain locus (open). (C) Western blot of immunoprecipitated Pcalp-GFP (C3) and Pcalp-GFP-FKBP (A7) parasites grown in the presence or absence of Shld1. Samples were processed as in Fig. 2B. Each lane corresponds to equivalent numbers of mid/late trophozoites.

Russo et al.

PNAS 兩 February 3, 2009 兩 vol. 106 兩 no. 5 兩 1557

MICROBIOLOGY

Method

parasitemia (%)

parasitemia (%)

10.5 9 7.5 6 4.5 3 1.5 0

8 7 6 5 4 3 2 1 0 0.5 1

E 1211 parasitemia (%)

parasitemia (%)

B 109 C3

1.5

2 2.5 3 time (days)

3.5

4

dead par (%) in -/+ Shld1

% of total parasites

Ring

0.1

0.3 0.2 Shld1 ( M)

0.4

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

merges

Topro3 Hoechst

Phase

H

- Shld1

+ Shld1

a.

b. Troph

Troph

C3

10 9 8 7 6 5 Ring 4 3 2 1 E Schiz 0 0 20

G

Ring

Ring

Ring

I100

+Shld1

80 pre S-phase = 25.5 0.5h

60

Troph Troph

40 60 time (hours)

80

100

F1.82

C100 90 80 70 60 50 40 30 20 10 0 0

A7

10 9 8 7 6 5 4 Ring 3 2 1 E Schiz 0

%of total parasites

D 1211

13.5

A12 A7

40

Cycle time = 42.5 0.5h

20 0 100 80

-Shld1

60

pre S-phase = 28.5 1h

40

Cycle time = 42 0.5h

20 0 -15 -10 Syto+CM

Topro+CM

Topro+AcOr

-5

0

5

10 15 20 25 30 35 40 time (hpi)

45 50 55 60

Fig. 5. Analysis of Pcalp knockdown phenotypes. (A and B) Asynchronous cultures of representative clone A7 (A) or C3 (B) were grown with (F) or without (E) 0.4 ␮M Shld1 and monitored over time by flow cytometry. X axis in A same as in B. (C) Growth of Pcalp-GFP-FKBP (mean of 3 clones, F) or C3 (䊐) over 3 days in different Shld1 concentrations. (D and E) Fine analysis of the Shld1 growth phenotype in synchronized cultures of A7 (D) and C3 (E). Schizonts from a culture grown in 0.2 ␮M Shld1 were isolated and allowed to reinvade fresh RBCs in the presence (F) or absence (E) of Shld1. The breaks indicate an equal subculture event for each culture. X axis in D same as in E. (F) Viability measurements using different dyes to measure dead parasites in A7 (light) and C3 (dark) cultures. The ratio ⫺/⫹ 0.2 ␮M Shld1 is plotted. Flow profiles are in Fig. S5. (G) Appearance of A7 dead parasites (Topro3-positive) by fluorescence microscopy. In the absence of Shld1 a unique dead species was detected: trophozoites that are either extraerythrocytic (a) or within RBC ghosts (b). (H) Giemsa-stained thin smears showed a delay in the morphological transition from ring to trophozoite in synchronized A7 cultures in the absence of Shld1. (I) Representative cycles of A7 with (Upper) and without (Lower) Shld1. Cycle points were acquired every 30 min for ⬇6 days, samples were fixed, and DNA content was analyzed by flow cytometry (Fig. S3). Green, mature schizonts; red, S phase; black, pre-S phase (G1). The cycle time is shown as distance between the 2 peaks of schizonts; the length of pre-S phase is the time separating the initial slope half maxima of the pre-S-phase and S-phase peaks. Red vertical lines indicate S-phase maxima in ⫹Shld1 for visual comparison to ⫺Shld1 culture. A consistent delay in the start of the S phase was detected when calpain was destabilized.

Multiple efforts to generate a parasite clone lacking calpain expression failed, strongly suggesting that the absence of this gene is incompatible with optimal viability during blood stages. We used 3 different techniques. The double cross-over gene knockout approach failed to generate recombinants. Disruption by single cross-over at the 5⬘ end of the endogenous locus, including a control for homologous recombination, is a new approach. Only when the native promoter was recapitulated did the recombination result in viable parasites. It is always possible that technical problems led to failure to isolate gene-disrupted recombinants for these 2 procedures. Therefore, we designed a third approach, an allelic replacement, to overcome this limitation. We were readily able to isolate parasites with cross-overs upstream of the active cysteine codon, at the predicted frequency, when the vector contained a synonymous active site mutation. In contrast, all recombinants were downstream when the nonsynonymous mutation vector was used. Unlike the first 2 strategies, the allelic replacement method generates recombinants even with the missense mutation vector. It is the site of recombination that informs us of the gene essentiality. In addition, the importance of the active site residue informs us that Pcalp activity is necessary for asexual cycle development. These studies give us insight into essentiality but do not help us understand the role of the enzyme in the cell. To pursue this, we developed a method, based on the FKBP degradation domain (12, 13), to generate a regulated knockdown of Pcalp. By fusing a GFP-FKBP tag at the endogenous locus we generated clonal lines 1558 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0806926106

in which calpain stability is regulated by the small molecule Shld1. Destabilization of Pcalp yielded a growth defect of 40–60% over 2 cycles. We measured a delay of entry into S phase. Morphologically, the transition from ring to trophozoite was also delayed. Concomitantly we detected an increased number of nonviable parasites. Peaks of Pcalp transcription and translation in normal parasites were seen in the pre-S-phase stages (ring to mid trophozoite), after which a sudden drop in levels was detected. This correlates nicely with the observed knockdown phenotype. In the absence of Shld1, the fusion protein is degraded, resulting in a knockdown effect. The remarkably scarce transcript and barely detectable protein levels provided challenges in the detection and study of this calpain. The normal level of calpain is so low that we cannot quantify precisely the extent of protein diminution. However, other FKBP fusions have shown a one order of magnitude range in response to Shld1 (ref. 13 and I.R., unpublished results). Assuming that the small number of calpain molecules has a normal Gaussian distribution in all cells, we attribute the death phenotype to those cells (about 20% of the parasites) where the number of calpain molecules is insufficient to overcome the efficiency of the degradation process. Therefore, these cells die because of a loss of calpain function. The rest of the cells may produce just enough calpain to survive, although they show a defect in pre-S-phase development. The total cell cycle time is normal for surviving parasites, suggesting that they are able to catch up. Perhaps a cycle clock further along is able to proceed at the normal time, whereas Russo et al.

determined according to time and morphological analysis by Giemsa staining. RNA was treated with DNase I (Gibco) and checked by PCR for purity. Pcalp cDNA was amplified by one-step RT-PCR with SuperScript (Invitrogen). The RT-PCR products of at least 2 independent reactions for each time point were analyzed on ethidium bromide/agarose gels. Band intensity was in a linear range. Primers are in Table S3. Southern Blots. For Southern blots, 1 ␮g of DNA was digested and analyzed as previously described (25). The double cross-over integrations were screened by PacI/XmaI digestion and probed using Pcalp ORF 5⬘ end. The single cross-over integrations were screened by NsiI/SphI digestion and probing with Pcalp ORF 5⬘ or 3⬘ end.

Materials and Methods

Western Blots. Protein analysis to detect Pcalp after immunoprecipitation was conducted as described in SI Materials and Methods.

Reagents. Polyclonal antiserum (no. 19) was raised in rabbits against Pcalp peptide 41–54. We also used rabbit anti-GFP ab6556 (Abcam), mouse anti-GFP JL8 (BD), rabbit anti-BiP (MR4), and rabbit anti-plasmepsin II (18) antibodies. All other reagents were purchased from either Sigma or NEB unless indicated.

Microscopy Techniques. Live parasites were observed in the presence of Hoechst 33342 and Topro3. Images were collected with an Axioskop epifluorescence microscope (Zeiss) as described elsewhere (25).

Cell Cultures and Transfection. P. falciparum (3D7) was cultured as previously described (19). Parasite synchronization was obtained by using 5% D-sorbitol treatment (20), Percoll (21), and magnetic separation (22). For transfections, 160 ␮l of 50% RBCs was transfected by electroporation (23) with ⬇100 ␮g of purified vector DNA and then infected with 3D7 schizonts. After 72–90 h, 10 nM WR99210 was added to the medium. To select for integration, parasites were cycled twice on/off drug (24). Double cross-over recombination was selected as described in ref. 11. Sequence and Phylogenetic Analysis. See SI Materials and Methods. Construction of Vectors. Genomic DNAs were extracted from P. falciparum by using a Blood Mini Kit (Qiagen). Primers and restriction sites are listed in Table S3. All cloning steps were confirmed by sequencing. For the gene knockout strategy, the plasmid pHHTTK (11) was modified by inserting the HcRed coding region upstream to hDHFR. Into this backbone, we cloned the 5⬘ and 3⬘ ends of the Pcalp ORF. Four constructs were made for gene disruption by single cross-over. The calpain N-terminal coding regions (1.3 and 1.6 kb) with or without 733 bp of 5⬘ UTR were PCR amplified from genomic DNA and cloned in pPM2GT (25). The vectors for allelic replacement were generated by cloning 3.8 kb of Pcalp 3⬘ ORF into pPM2GT (25). Mutagenesis of the allelic replacement vectors was done with the QuikChange kit (Stratagene). The vectors for 3⬘ tagging were constructed by cloning 1.1 kb of the 3⬘ Pcalp ORF into pPM2GT (25) to generate Pcalp-GFP or replacing GFP with GFP-FKBP to generate Pcalp-GFP-FKBP. The FKBP domain was amplified from pBMN YFP-L106P (12).

Flow Cytometry. For cell cycle analysis, highly synchronous 2% hematocrit cultures, at 1% parasitemia, were cultivated in gently rocked Roboflasks (Corning) to ensure cell suspension. An automated system, controlled by 3 timers and composed of 3 synchronized peristaltic pumps and a fraction collector, was set up. Each 30 min, ⬇150 ␮l of sample culture was extracted by the system through the flask septum, and each sample was fixed and was collected at 4 °C in a 96-well plate. For subculture, fresh, prewarmed RBCs (2% hematocrit) were added through the septum without removing the flask from the incubator. The fixative solution (1.5 vol per sample) was 3.2% formaldehyde and 0.01% glutaraldehyde in PBS. After 5min of permeabilization with 0.1% Triton X-100 and 15 min of incubation with ⬇75 ␮g/ml RNase A, cell DNA was stained with 0.5–1 ␮M Topro3. The analysis was conducted on a BD FACSCanto flow cytometer, monitoring fluorescence profiles of infected RBC (Fig. S3). For viability analysis, we used the following dyes in pairs: 1.25 ␮g/ml Cell tracker-green (CM-green), 2.5 ␮M Syto59, 0.5 ␮M Topro3, and 0.4 ␮g/ml AcOr (Molecular Probes). Metabolically active cells incorporated CM-green during a 15-min incubation at 37 °C. Then they were analyzed by flow cytometry in the presence of DNA dyes, cell-permeant Syto59, or cell-impermeant Topro3. The third viability measurement was based on the use of 2 DNA dyes at the same time, AcOr and Topro3. Control dead cells were produced by fixing some of the culture. The analysis was conducted recording the green and far-red fluorescence profiles of infected RBCs (Fig. S5).

Semiquantitative RT-PCR. 3D7 RNA was collected from saponin-released parasites by using TRIzol reagent following the manufacturer’s protocol (Invitrogen). RNA was collected at 6 stages of development, ER (6 –10 hpi), LR (12–18 hpi), ET (20 –24 hpi), LT (24 –28 hpi), LT/Sc (28 –32 hpi), and Sc (34 – 40 hpi). Stages were

ACKNOWLEDGMENTS. We thank the Sanger and Broad Institutes for DNA sequences; Jacobus Pharmaceutical for WR99210; T. Wandless (Stanford Univerisity, Palo Alto, CA) for Shld1 and ddFKBP; MR4 and J. H. Adams (University of South Florida, Tampa, FL) for anti-BiP; A. Cowman (Walter and Eliza Hall Institute of Medical Research, Melbourne) for pHHTTK; C. Armstrong (Washington University, St. Louis) for sharing results before publication; and M. Drew for critical reading of the manuscript. This work was supported by National Institutes of Health Grant AI-047798.

1. Rosenthal PJ (2004) Cysteine proteases of malaria parasites. Int J Parasitol 34:1489 –1499. 2. Wu Y, Wang X, Liu X, Wang Y (2003) Data-mining approaches reveal hidden families of proteases in the genome of malaria parasite. Genome Res 13:601– 616. 3. Croall DE, Ersfeld K (2007) The calpains: Modular designs and functional diversity. Genome Biol 8:218. 4. Goll DE, Thompson VF, Li H, Wei W, Cong J (2003) The calpain system. Physiol Rev 83:731– 801. 5. Sorimachi H (2004) Handbook of Proteolytic Enzymes (Elsevier, London), pp 1206 – 1225. 6. Sorimachi H, Suzuki K (2001) The structure of calpain. J Biochem (Tokyo) 129:653– 664. 7. Perrin BJ, Huttenlocher A (2002) Calpain. Int J Biochem Cell Biol 34:722–725. 8. Blanchard H, et al. (1997) Structure of a calpain Ca2⫹-binding domain reveals a novel EF-hand and Ca2⫹-induced conformational changes. Nat Struct Biol 4:532–538. 9. Bozdech Z, et al. (2003) The transcriptome of the intraerythrocytic developmental cycle of Plasmodium falciparum. PLoS Biol 1:E5. 10. Le Roch KG, et al. (2003) Discovery of gene function by expression profiling of the malaria parasite life cycle. Science 301:1503–1508. 11. Duraisingh MT, Triglia T, Cowman AF (2002) Negative selection of Plasmodium falciparum reveals targeted gene deletion by double cross-over recombination. Int J Parasitol 32:81– 89. 12. Banaszynski LA, Chen LC, Maynard-Smith LA, Ooi AG, Wandless TJ (2006) A rapid, reversible, and tunable method to regulate protein function in living cells using synthetic small molecules. Cell 126:995–1004. 13. Armstrong CM, Goldberg DE (2007) An FKBP destabilization domain modulates protein levels in Plasmodium falciparum. Nat Methods 4:1007–1009. 14. Janossy J, et al. (2004) Calpain as a multi-site regulator of cell cycle. Biochem Pharmacol 67:1513–1521.

15. Zhang W, Lu Q, Xie ZJ, Mellgren RL (1997) Inhibition of the growth of WI-38 fibroblasts by benzyloxycarbonyl-Leu-Leu-Tyr diazomethyl ketone: Evidence that cleavage of p53 by a calpain-like protease is necessary for G1 to S-phase transition. Oncogene 14:255–263. 16. Janse CJ, et al. (2003) Malaria parasites lacking eEF1a have a normal S/M phase yet grow more slowly due to a longer G1 phase. Mol Microbiol 50:1539 –1551. 17. Boisvert FM, van Koningsbruggen S, Navascues J, Lamond AI (2007) The multifunctional nucleolus. Nat Rev Mol Cell Biol 8:574 –585. 18. Francis SE, et al. (1994) Molecular characterization and inhibition of a Plasmodium falciparum aspartic hemoglobinase. EMBO J 13:306 –317. 19. Trager W, Jensen JB (1976) Human malaria parasites in continuous culture. Science 193:673– 675. 20. Lambros C, Vanderberg JP (1979) Synchronization of Plasmodium falciparum erythrocytic stages in culture. J Parasitol 65:418 – 420. 21. Dluzewski AR, Ling IT, Rangachari K, Bates PA, Wilson RJ (1984) A simple method for isolating viable mature parasites of Plasmodium falciparum from cultures. Trans R Soc Trop Med Hyg 78:622– 624. 22. Paul F, Roath S, Melville D, Warhurst DC, Osisanya JO (1981) Separation of malariainfected erythrocytes from whole blood: Use of a selective high-gradient magnetic separation technique. Lancet 2:70 –71. 23. Fidock DA, Wellems TE (1997) Transformation with human dihydrofolate reductase renders malaria parasites insensitive to WR99210 but does not affect the intrinsic activity of proguanil. Proc Natl Acad Sci USA 94:10931–10936. 24. Wu Y, Kirkman LA, Wellems TE (1996) Transformation of Plasmodium falciparum malaria parasites by homologous integration of plasmids that confer resistance to pyrimethamine. Proc Natl Acad Sci USA 93:1130 –1134. 25. Klemba M, Beatty W, Gluzman I, Goldberg DE (2004) Trafficking of plasmepsin II to the food vacuole of the malaria parasite Plasmodium falciparum. J Cell Biol 164:47–56.

Russo et al.

PNAS 兩 February 3, 2009 兩 vol. 106 兩 no. 5 兩 1559

MICROBIOLOGY

Pcalp affects a previous one. Involvement of a calpain-like activity in G1 3 S transitions has been suggested in other systems on the basis of indirect evidence (14, 15). An S-phase delay has also been observed in P. berghei parasites lacking 1 of 2 eEF1a genes, but in that case, the total cell cycle was prolonged (16). Given the role in cell cycle progression, it is interesting that this enzyme is found in the nucleolus (I.R., A.O., and D.E.G., unpublished data), which, in addition to rRNA synthesis and ribosome assembly, has key functions in cell cycle regulation (17). In conclusion, by developing a method for regulated knockdown of Pcalp, we can begin to define its cellular role and further develop this interesting target for antiparasitic chemotherapy.