Plasmodium berghei EXP-1 interacts with host Apolipoprotein H during Plasmodium liver-stage development

Significance

The clinically silent intracellular development of Plasmodium parasites in the host liver is a prerequisite for the onset of malaria pathology. Liver stages can be completely eliminated by sterilizing immune responses and are promising targets for urgently needed antimalarial drugs and/or vaccines. The parasite is separated from the host cell cytoplasm by a parasitophorous vacuole (PV). We show that the PV membrane protein exported protein 1 interacts specifically with host Apolipoprotein H. The characterization of this protein–protein interaction revealed an essential role for both molecular partners during intrahepatic parasite growth. Our results improve our understanding of cell-biological aspects of host–pathogen interactions and may also help to develop new strategies to control Plasmodium infections.

Abstract

The first, obligatory replication phase of malaria parasite infections is characterized by rapid expansion and differentiation of single parasites in liver cells, resulting in the formation and release of thousands of invasive merozoites into the bloodstream. Hepatic Plasmodium development occurs inside a specialized membranous compartment termed the parasitophorous vacuole (PV). Here, we show that, during the parasite’s hepatic replication, the C-terminal region of the parasitic PV membrane protein exported protein 1 (EXP-1) binds to host Apolipoprotein H (ApoH) and that this molecular interaction plays a pivotal role for successful Plasmodium liver-stage development. Expression of a truncated EXP-1 protein, missing the specific ApoH interaction site, or down-regulation of ApoH expression in either hepatic cells or mouse livers by RNA interference resulted in impaired intrahepatic development. Furthermore, infection of mice with sporozoites expressing a truncated version of EXP-1 resulted in both a significant reduction of liver burden and delayed blood-stage patency, leading to a disease outcome different from that generally induced by infection with wild-type parasites. This study identifies a host–parasite protein interaction during the hepatic stage of infection by Plasmodium parasites. The identification of such vital interactions may hold potential toward the development of novel malaria prevention strategies.

Malaria remains the most important vector-borne disease worldwide, leading to particular devastation in sub-Saharan Africa. Malaria pathology is caused by the blood stages of single-celled parasites of the genus Plasmodium. However, before the symptomatic infection of red blood cells, Plasmodium parasites undergo an obligatory and clinically silent developmental phase in the liver, which constitutes an ideal target for disease prevention (1, 2). The liver stage of Plasmodium infection occurs after sporozoites are injected into the skin of the mammalian host upon a blood meal of an infected female Anopheles mosquito (3). Injected sporozoites eventually reach the liver, where they undergo a dramatic transition to form invasive first-generation merozoites that are released into the bloodstream. Hepatic Plasmodium infection comprises distinct developmental phases. After successful penetration of the endothelial barrier in the liver sinusoid (4) and traversal of several liver cells (5), the infectious sporozoite eventually invades a hepatocyte with the formation of a membranous replication-competent niche, the parasitophorous vacuole (PV) (6). The intracellular parasite then transforms into round exoerythrocytic forms (EEFs), which undergo repeated closed mitosis, ultimately leading to the formation of several thousand progenies. This development is exceptional for an obligate eukaryotic intracellular pathogen and likely depends on the extensive acquisition of lipids and nutrients from its host cell, while also relying on the parasite‘s own metabolism to ensure its survival and replication within host cells (7, 8).

Despite being metabolically active itself, the parasite has been shown to scavenge a plethora of host-cell molecules, such as glucose, cholesterol, fatty acids, phosphatidylcholine, or lipoic acids (8–12). Because Plasmodium parasites do not reside freely in the host cell cytoplasm or in endocytic compartments, but, rather, inside a vacuole formed de novo during the active invasion process, required nutrients have to cross the parasite plasma membrane as well as the PV membrane (PVM).

It is generally suggested that the PVM is central to nutrient acquisition, host-cell remodeling, waste disposal, environmental sensing, and protection of the intracellular pathogen from innate immune defenses (13). However, little is known about intrahepatic Plasmodium stages with regard to interactions between the parasite and the host hepatocyte and their potential for nutrient uptake and/or exchange. Small molecules (up to 800 Da) can cross the PVM freely via specialized transport channels (14), whereas larger molecules might reach the parasite via association and possibly fusion of late endosomes, lysosomes, or amphisomes with the PVM (15–19).

Several PVM-resident proteins have been identified, the largest family being the early transcribed membrane proteins (ETRAMPs), of which seven are present in the rodent malaria parasite P. berghei (Pb) (20). The most prominent members of this family, named up-regulated in infective sporozoites (UIS) 3 and 4, are expressed exclusively in sporozoites and liver stages (21, 22). Depletion of these proteins leads to developmental liver-stage arrest, and UIS3 was shown to interact with the host hepatocyte protein liver fatty acid binding protein 1 (LFABP1), suggesting a role in fatty acid scavenging during Plasmodium hepatic infection (23–25). Although another two ETRAMPs of the human malaria parasite, P. falciparum (Pf), are known to bind host-cell proteins, the function of these protein–protein interactions and that of most PVM-resident proteins remains largely elusive (26, 27).

Exported protein 1 (EXP-1) was the first protein described to associate with the PVM (28). It harbors a classical N-terminal signal peptide, and, upon trafficking via the endoplasmic reticulum–Golgi transport route, is inserted with its transmembrane domain into the PVM of both blood and liver parasite stages (29–33). It was shown later that the protein’s N-terminal region faces the PV lumen, whereas its C terminus (CT) extends into the host-cell cytosol (34, 35). The same membrane topology was reported for other small PVM-resident proteins of the ETRAMP family, which are, together with EXP-1, organized in nonoverlapping oligomeric arrays within the PVM (36, 37). Because EXP-1 was found to be continuously trafficked to the PVM throughout at least the first half of liver-stage development, there was speculation as to whether the protein might also be trafficked back from the PVM to the PV or even the parasite cytosol (38). PfEXP-1 is one of the most abundantly transcribed loci during ring and early trophozoite-stage development; it is therefore not surprising that it seems to exert vital functions during blood-stage development and was shown to be refractory to gene deletion (39–41). Recently, EXP-1 was predicted to possess GST activity, and it was demonstrated to conjugate glutathione onto hematin in vitro (42). Whether EXP-1 also exerts GST activity during Plasmodium intrahepatic development remains to be investigated.

Because recruitment of host-cell proteins to the parasite–host interface during liver-stage development could be a possible function for a PVM-resident protein such as EXP-1, we aimed at identifying potential host-cell interaction partners of this protein. We found that the C-terminal portion of PbEXP-1 specifically interacts with host Apolipoprotein H (ApoH) and that this interaction is pivotal for successful liver-stage development of the parasite, both in vitro and in vivo. Our data suggest that Plasmodium liver stages use EXP-1 to specifically recruit host–hepatocyte ApoH to the parasite–host interface and to potentially mediate uptake of ApoH and/or ApoH-associated proteins or lipids.

Results

PbEXP-1 Interacts with Rn ApoH.

To address the functionality of PbEXP-1 (PBANKA_0926700) during Plasmodium intrahepatic development, we carried out a yeast two-hybrid (Y2H) screen to identify novel host-cell molecular partners of this parasite protein. Plasmodium EXP-1 is a small, single-pass transmembrane protein with a classical signal peptide at its N-terminal portion (Fig. 1A), which localizes to the PVM (34, 37). Because EXP-1 CT is known to be exposed to the host cell cytoplasm (Fig. S1 A–C) (34, 35), this portion of the protein was used as bait in a Y2H screen against a Rattus norvegicus (Rn) hepatocyte cDNA library. The screen identified several putative interacting partners of EXP-1, with rat ApoH (RnApoH; NP_001009626.1), also known as beta-2-glycoprotein I (Fig. 1B), emerging as the most frequent hit (Table S1).

Fig. 1.

EXP-1 CT interacts with host cell ApoH. (A) Primary structures of PbEXP-1 and PfEXP-1 with signal peptide (light gray box) and transmembrane domain (black box). C-terminal fragments CT, C1, and C2 used as baits in the Y2H screening and cotransformation experiments are shown below. (B) Primary structure of RnApoH with signal peptide (light gray box), CCP (gray), and Sushi-2 (dark gray) domains. Two prey cDNA sequences (Y2H prey 1 and 2) of rat ApoH were identified in the Y2H screening, and the corresponding protein fragments are shown. (C) PbEXP-1 CT and C2 fragments interact with ApoH of rat, human, or mouse origin, whereas the C-terminal domain of P. falciparum EXP-1 did not interact with rat, human, or mouse ApoH, as demonstrated by yeast cotransformation assays. (D) HA-tagged PbEXP-1 CT (HA-PbEXP-1 CT) was immunoprecipitated by using an anti-HA antibody (Sigma, H9658), followed by immunoblotting with an anti-ApoH antibody (Abcam, ab108348). HEK293T cells were transfected with plasmids expressing HA-PbEXP-1 CT, HA-tagged PbUIS4 (HA-PbUIS4), and Myc-RnApoH in various combinations. The ApoH–EXP-1 complex was pulled down in cells cotransfected with plasmids expressing both PbEXP-1 CT and RnApoH (lane 6). No interaction was observed between PbUIS4 and RnApoH (lane 5). (E) Far Western blot analysis and antibody inhibition assay of rat ApoH binding to EXP-1. Protein extracts of liver stages derived from Pb-infected Huh7 cells (lane 1) and of blood stages from Pb-infected mouse erythrocytes (lane 2) were separated by SDS/PAGE and blotted onto a nitrocellulose membrane, followed by incubation with ApoH-containing rat serum and ApoH detection with an anti-ApoH antibody. A band corresponding to ApoH was detected at the expected location of EXP-1 (∼20 kDa), as confirmed by the similarly treated extracts of Pf-infected human erythrocytes (lane 3) and detection with an anti–PfEXP-1 antibody. Incubation of the blots as described above with an anti–EXP-1 antibody before incubation with ApoH-containing rat serum rendered EXP-1 inaccessible to ApoH binding (lanes 4–6). (F) Immunofluorescence microscopy analysis of ApoH-stained (green) and EXP-1–stained (red) stained WT Pb parasites 48 hpi of Huh7 cells shows partial EXP-1 colocalization with ApoH and uptake of host ApoH into the parasite. (Scale bar: 10 µm.) (G) Quantification of EXP-1 colocalization with ApoH measured at various time points of liver-stage development using an anti-ApoH and an anti–EXP-1 antibody raised against full-length PbEXP-1 (anti–PbEXP-1^FL). Each symbol represents one parasite (n = 20, except for 15 h, where n = 15). *P < 0.05; **P < 0.01; ***P < 0.001 (unpaired t test). (H) Quantification of EXP-1 colocalization with ApoH at 24 and 48 hpi using an anti-ApoH antibody and either the anti–PbEXP-1^FL or the anti–PbEXP-1^c antibody, which exclusively detects the C-terminal region of EXP-1. Each symbol represents one parasite (n = 20). For G and H, data are shown as mean (indicated by red line) ± SD.

Fig. S1.

Membrane topology of EXP-1 in infected erythrocytes and alignment of Pb and PfEXP-1 protein sequences. (A) Incubation of infected erythrocytes (red) with Streptolysin O (SLO) selectively permeabilizes the erythrocyte plasma membrane (EPM). Human HSP70 (HsHSP70) can thus be found in the supernatant (S) fraction, whereas the pellet (P) fraction contains parasite proteins SERP and Aldolase, which are retained in the PV (yellow) or parasite cytoplasm (blue), respectively. The two potential orientations of EXP-1 within the PVM are depicted on the left side (CT toward the erythrocyte cytoplasm) and right side (CT toward the PV lumen). After SLO treatment, cells were incubated with Proteinase K (+PK) or without Proteinase K (−PK). (B) After SLO treatment of infected erythrocytes, only HsHSP70 can be detected in the S fraction. The PV protein SERP and the parasite’s cytoplasmic protein Aldolase are detected in the P fraction, indicating intact PVM and parasite plasma membrane (PPM), respectively. (C) An antibody raised against the EXP-1 CT detects the protein in the parasite PVM after SLO treatment of infected erythrocytes in the absence of Proteinase K (−PK). Upon Proteinase K treatment (+PK), the EXP-1 CT is digested and is no longer detected, demonstrating that it is facing the erythrocyte cytoplasm and is accessible to PK digestion. (D) Alignment of Pb and Pf EXP-1 amino acid sequences. Identical amino acids between the two proteins are highlighted with black boxes. The locations of the signal peptide and transmembrane domain are indicated by red and blue bars, respectively. The overall amino acid identity between Pb and PfEXP-1 is 38.8%. The C2 region (green bar) was the least similar, with only 16.1% identity with the other sequence. Sequences were aligned by using ClustalW, and protein identity was calculated with Biology Workbench (Version 3.2).

Table S1.

List of prey genes identified by Y2H screening with PbEXP-1 CT as bait protein against rat liver cDNA library

Gene name	Gene ID (NCBI)	No. of times observed	Prey start–prey end, amino acids
AFG3-like AAA ATPase 2	307,350	1	647–788
Alpha-2HS-glycoprotein	25,373	1	*
ApoH (beta-2-glycoprotein I)	287,774	2	88–345/87–345
Argininosuccinate synthetase 1	25,698	1	364–412
GST mu 2	24,424	1	*
Hemopexin	58,917	1	329–460
NSFL1 (p97) cofactor (p47)	83,809	1	178–370
Serpin peptidase inhibitor, clade A, member 6	299,270	1	1–278
Signal sequence receptor, beta	295,235	1	*
Zinc finger, DHHC-type containing 6	361,771	1	293–414

NCBI, National Center for Biotechnology Information.

^*Prey sequence in UTR.

Next, we performed direct interaction assays to confirm the interaction with rat ApoH and to identify the portion of the EXP-1 CT responsible for the interaction. To this end, cotransformation experiments were conducted (Fig. 1C), by using the complete CT of PbEXP-1 (amino acids 103–166), as well as fragments corresponding to the PbEXP-1 amino acids 103–136 (C1 region) and 137–166 (C2 region), as bait proteins (Fig. 1A). Rn, Mus musculus (Mm; NP_038503.4), and Homo sapiens (Hs; NP_000033.2) ApoH were used as preys in these Y2H assays (Fig. 1C). Our results not only confirmed the interaction between PbEXP-1 and ApoH of rat, mouse, and human origin, but further identified the C2 region of PbEXP-1 as responsible for this interaction (Fig. 1C).

Similar to the studies with PbEXP-1, we used the complete CT of PfEXP-1 (amino acids 107–162) of the strains 3D7 and GC-06, which represent the two different naturally occurring variants of the protein that differ only at position 136 (31), as bait proteins. Alternatively, we used fragments of PfEXP-1 3D7 corresponding to amino acids 107–134 and 135–162 (C1 and C2 region, respectively) (Fig. 1A). In contrast, PfEXP-1 (PF3D7_1121600) CT differs substantially from its PbEXP-1 ortholog (Fig. S1D), and indeed, when performing direct interaction assays using different PfEXP-1 CT variants (Fig. 1A), no interaction with ApoH of rat, mouse, or human origin could be detected in three independent experiments (Fig. 1C and Table S2).

Table S2.

The C-terminal portion of PfEXP-1 does not interact with human, rat, or mouse ApoH

Bait	Empty prey plasmid control, mean (±SD) cfu/µg DNA on		RnApoH, mean (±SD) cfu/µg DNA on		HsApoH, mean (±SD) cfu/µg DNA on		MmApoH, mean (±SD) cfu/µg DNA on
	-Trp-Leu control plates	-Trp-Leu-Ade-His interaction plates	-Trp-Leu control plates	-Trp-Leu-Ade-His interaction plates	-Trp-Leu control plates	-Trp-Leu-Ade-His interaction plates	-Trp-Leu control plates	-Trp-Leu-Ade-His interaction plates
PbEXP-1 C2	4.34 × 104 (±4.3 × 103)*	0*	2.83 × 104 (±2.3 × 103)	1.53 × 104 (±4.1 × 103)	1.89 × 104 (±9.7 × 103)	3.48 × 104 (±3.3 × 104)	1.95 × 104 (±4.9 × 103)	9.8 × 103 (±3.7 × 103)
PfEXP-1 CT 3D7	1.89 × 104 (±7.9 × 103)*	0*	2.08 × 104 (±1.8 × 104)	0	2.06 × 104 (±2.0 × 104)	0	1.07 × 104 (±5.4 × 103)	0
PfEXP-1 CT GC-06	2.32 × 104 (±4.6 × 103)*	0*	1.46 × 104 (±5.9 × 103)	0	2.54 × 104 (±2.6 × 104)	0	1.19 × 104 (±2.5 × 103)	0
PfEXP-1 C1	3.34 × 104 (±3.4 × 104)*	0*	2.56 × 104 (±3.0 × 104)	0	2.72 × 104 (±3.1 × 104)	0	7.94 × 104 (±1.5 × 103)	0
PfEXP-1 C2	3.64 × 104 (±2.1 × 104)*	0*	3.84 × 104 (±3.4 × 104)	0	3.3 × 104 (±3.5 × 104)	0	1.45 × 104 (±3.5 × 103)	0

^*Mean (±SD) of four independent experiments; all other values show mean of three independent experiments.

The PbEXP-1 and RnApoH protein–protein interaction was further confirmed by coimmunoprecipitation studies using an HA-tagged version of the C-terminal region of PbEXP-1 (HA-PbEXP-1 CT) and Myc-tagged RnApoH (Myc-RnApoH). In these experiments, the complex was successfully pulled down with two independent α-HA antibodies and detected by immunostaining with either α-Myc (Fig. S2A) or α-ApoH antibodies (Fig. 1D), respectively. No interaction with RnApoH was observed with HA-tagged PbUIS4 (HA-PbUIS4), used as a control in these experiments (Fig. 1D). The interaction between PbEXP-1 and RnApoH was further validated by far Western blot and antibody inhibition assays. In these experiments, protein extracts of Pb-infected Huh7 cells or Pb-infected C57BL/6 mouse blood were separated by SDS/PAGE and blotted onto a nitrocellulose membrane that was subsequently incubated with ApoH-containing rat serum and washed to remove any unspecific bound proteins. Binding of ApoH to EXP-1 was demonstrated by staining with an α-ApoH antibody, which revealed a band of ∼20 kDa, corresponding to the size of the PbEXP-1 protein (Fig. 1E). Detection of PfEXP-1 in Pf blood-stage cultures with α-PfEXP-1 antibody, which we have shown to be fully reactive against the PbEXP-1 protein (Fig. S2B), was used as control (Fig. 1E). Incubation of the membrane with α-PfEXP-1 antibody, which was raised against the protein’s CT, before incubation with rat serum completely blocked binding of ApoH, again confirming the specificity of the interaction (Fig. 1E). Moreover, we carried out ELISAs where wells coated with recombinant RnApoH were incubated or not with lysates of Pb-infected or noninfected Huh7 cells, followed by washing and incubation with an α-PbEXP-1 antibody. Our results showed a clear increase in the signal detected only in samples incubated with lysates of Pb-infected cells, solidifying the interaction between PbEXP-1 and RnApoH (Fig. S2C). Immunofluorescence microscopy colocalization studies were conducted from 10 to 48 h postinfection (hpi) by using a commercial α-ApoH antibody and an α-PbEXP-1 antibody (Fig. 1F). Our results showed that the two proteins colocalize in the PVM region of the parasite. Interestingly, colocalization of ApoH and EXP-1 increased during EEF development, with, on average, 40% of the total EXP-1 signal in EEFs analyzed at 48 hpi colocalizing with ApoH (Fig. 1G). Finally, we performed colocalization studies 24 and 48 hpi using antibodies specifically raised against the full-length PbEXP-1 (PbEXP-1^FL) or against its C-terminal regions (PbEXP-1^c, raised against amino acids 103–166; see also Fig. 1A) of PbEXP-1 (Fig. 1H). These results clearly confirm the colocalization of ApoH with the C-terminal domain of EXP-1 facing the hepatocyte cytoplasm. Overall, our data demonstrated, by a variety of independent methods, a parasite–host molecular interaction between the PbEXP-1 and RnApoH proteins, prompting further investigation of the physiological role of this interaction during hepatic infection by Plasmodium parasites.

Fig. S2.

Coimmunoprecipitation of PbEXP-1 CT and rat ApoH, Western blot analysis of PfEXP-1–antibody binding to Pb EXP-1, and ELISA analysis of PbEXP-1 binding to recombinant ApoH. (A) Immunoprecipitation of HA-PbEXP-1 CT using an anti-HA antibody (Abcam, ab9110) followed by immunoblotting with an anti-cMyc antibody (Clontech, 631206). HEK293T cells were transfected with plasmids expressing HA-PbEXP-1 CT, HA-tagged PbUIS4 (HA-PbUIS4), and Myc-RnApoH in various combinations. The ApoH–EXP-1 complex was pulled down only in cells cotransfected with plasmids expressing both PbEXP-1 CT and RnApoH (lane 6). No interaction was observed between PbUIS4 and RnApoH (lane 5). Expression of cMyc-tagged rat ApoH was confirmed by anti-cMyc staining of total cell lysates. (B) Rabbit antiserum raised against Pf EXP-1 (provided by Klaus Lingelbach) also detects Pb EXP-1 in sporozoite (spz), liver-stage (LS), and blood-stage (BS) parasite lysates. Pf blood-stage lysates were used as positive control. (C) Pb-infected Huh7 cell lysates were added to 20 µg/mL purified ApoH-coated ELISA plates. Relative binding of PbEXP-1 was visualized by using rabbit PbEXP-1 antiserum and HRP-conjugated secondary antibody. Noninfected Huh7 cell lysates were used as negative control. Binding of EXP-1 to ApoH was only observed in ApoH-coated wells incubated with infected cell lysates. Data are shown as mean ± SD.

ApoH Plays an Important Role in Hepatic Infection by Plasmodium Parasites.

We next used RNA interference (RNAi) to specifically silence the expression of the gene encoding endogenous ApoH to examine both the biological and physiological relevance of this particular parasite–host interaction for intrahepatic development of the parasite. To this end, the expression of ApoH in Huh7 cells was down-modulated by specific siRNA oligonucleotides (siRNAs ApoH_1–3; Fig. S3A and Table S3), and the resulting effect on infection by Pb sporozoites was analyzed comprehensively (43). First, RNAi-mediated knockdown of ApoH expression in hepatocytes was confirmed by quantitative real-time PCR (qRT-PCR) and Western blotting (Fig. S3 A and B, respectively) and shown to be stable over the 48-h period of infection (Fig. S3C). Because siRNA ApoH_1 showed the strongest phenotype, it was used in all experiments where a single siRNA was used. Pb infection of Huh7 cells after down-modulation of ApoH expression was then compared with that of control cells, transfected with scrambled siRNA (Neg siRNA). Initially, cells were infected with either luciferase-expressing or wild-type (WT) Pb sporozoites, and infection was measured by bioluminescence (44) or qRT-PCR, respectively (Fig. 2 A and B). Our results showed that down-modulation of ApoH expression leads to a significant decrease in overall infection (Fig. 2 A and B), which was not due to impaired invasion of sporozoites (Fig. S3D). We then used immunofluorescence microscopy to assess the effect of ApoH down-modulation on parasite numbers (Fig. 2C) and sizes (Fig. 2 D and E). A comparison of the resulting infection rates clearly showed that ApoH down-modulation results in a significant decrease in the development of Plasmodium liver stages (Fig. 2D). This finding suggests a specific requirement of host-cell ApoH for successful intrahepatic Plasmodium development.

Fig. S3.

Stability of ApoH gene silencing by siRNAs. (A) Effect of siRNA transfection of Huh7 cells on ApoH mRNA levels, measured by qRT-PCR. Cells transfected with a scrambled siRNA (Neg) were used as negative control. Expression of ApoH target gene was normalized to the HPRT housekeeping gene. Results were normalized to negative control (100%). Three different siRNAs (ApoH_1–3) were used (Table S3). (B) Efficiency of ApoH knockdown at the protein level assessed by Western blot analysis of cell lysates extracted 36 h after transfection. Both a lysate of nontransfected cells and a sample of purified ApoH protein were used as controls. β-Actin was used as loading control. (C) ApoH knockdown stability measured by qRT-PCR starting 24 h after transfection. Knockdown stability was measured over the following 48 h. Cells transfected with a scrambled siRNA (Neg) were used as negative control. Expression of ApoH target gene was normalized to HPRT housekeeping gene. Results were normalized to negative control (100%). (D) Knockdown of ApoH had no influence on Pb WT sporozoite invasion rates 2 hpi. Knockdown of SRB1 was used as positive control. For A, C, and D: Data are shown as mean ± SD.

Table S3.

List of primer, shRNA sequences, and siRNA IDs used in this study

Experiment	Gene (gene no./ID)	Primer name	Sequence
pexp-1∆C2/ctrl plasmid	PbEXP-1 (PBANKA_092670)	3′UTR PbEXP-1 for	5′-CCATCGATGTATCATAAAAAGTTTCGACTC-3′
	PbEXP-1 (PBANKA_092670)	3′UTR PbEXP-1 rev	5′-GGGGTACCGAGTAAATACCCCTACATATG-3′
	PbEXP-1 (PBANKA_092670)	5′UTR PbEXP-1 ORFdC2 for	5′-ATAAGAATGCGGCCGCGCCGTTTAACTCTTAATTTAC-3′
	PbEXP-1 (PBANKA_092670)	5′UTR PbEXP-1 ORFdC2 rev	5′-CGGGATCCTTAGGATGAAAAACCCAAGGATGATTC-3′
	PbEXP-1 (PBANKA_092670)	5′UTR PbEXP-1ORFctrl rev	5′-CGGGATCCTCATTGTTGAAGATTTGGCATGTTAAG-3′
Genotyping ∆C2/ctrl parasite line	5′ integration	PbEXP-1 int test for	5′-GCCACTCTTCTGTAATACCTATGC-3′
	5′ integration	b3D+ rev	5′-CCTTGCTCATTTACCTGCTAATACGATTGC-3
	3′ integration	Tg for	5′-CGCATTATATGAGTTCATTTTACACAATCC-3′
	3′ integration	PbEXP-1 int test rev	5′-GCTTATATCGATTTGTGCTAACTGG-3′
	WT	PbEXP-1 int test for	5′-GCCACTCTTCTGTAATACCTATGC-3′
	WT	PbEXP-1 int test rev	5′-GCTTATATCGATTTGTGCTAACTGG-3′
Immunoprecipitation	RnApoH (287774)	RnApoH FWD EcoRI pCMV-myc	5′-CCCGAATTCCAATGATTTCTCCGGCGC-3′
	RnApoH (287774)	RnApoH REV XhoI pCMV-myc	5′-AAACTCGAGTTATTCTGCTATTTCAGTTGA-3′
qRT-PCR	18S rRNA	PbA18S for	5′-AAGCATTAAATAAAGCGAATACATCCTTA-3′
	18S rRNA	PbA18S rev	5′-GGAGATTGGTTTTGACGTTTATGTG-3′
	MmGAPDH (14433)	MmGAPDH for	5′-CGTCCCGTAGACAAAATGGT-3′
	MmGAPDH (14433)	MmGAPDH rev	5′-TTGATGGCAACAATCTCCAC-3′
	MmHRPT (15452)	MmHRPT for	5′-CATTATGCCGAGGATTTGGA-3′
	MmHRPT (15452)	MmHRPT rev	5′-AATCCAGCAGGTCAGCAAAG-3′
	MmApoH (11818)	MmGAPDH for	5′-GCCACACTTTGCCATGATCG-3′
	MmApoH (11818)	MmGAPDH rev	5′-AGGACCAAGTTCCCGTCTTG+-3′
	HsApoH (350)	MmHRPT for	5′-TTGCAGGACGGACCTGTCCCAAG-3′
	HsApoH (350)	MmHRPT rev	5′-ACACATAGCCCGGCTTGCAGGA-3′
Y2H plasmids	PbEXP-1 (PBANKA_092670)	PbEXP-1 CT for	5′-AAAGAATTCAGACATCCATTCCAAATTGGC-3′
	PbEXP-1 (PBANKA_092670)	PbEXP-1 CT rev	5′-AAAGTCGACCATTGTTGAAGATTTGGCATG-3′
	PbEXP-1 (PBANKA_092670)	PbEXP-1 C1 for	5′-GGAATTCCATATGAGACATCCATTCCAAATTGGC-3′
	PbEXP-1 (PBANKA_092670)	PbEXP-1 C1 rev	5′-CGGGATCCGGATGAAAAACCCAAGGATGATTC-3′
	PbEXP-1 (PBANKA_092670)	PbEXP-1 C2 for	5′-GGAATTCCATATGGAAGGAATACAAAATTTTGCTTC-3′
	PbEXP-1 (PBANKA_092670)	PbEXP-1 C2 rev	5′-CGGGATCCCATTGTTGAAGATTTGGCATGTTAAG-3′
	PfEXP-1 (PF3D7_1121600)	PfEXP-1 CT for	5′-CGGAATTCAGACACCCATTCAAAATAGG-3′
	PfEXP-1 (PF3D7_1121600)	PfEXP-1 CT rev	5′-CGGGATCCGTGTTCAGTGCCACTTACGG-3′
	PfEXP-1 (PF3D7_1121600)	PfEXP-1 C1 for	5′-CGGAATTCAGACACCCATTCAAAATAGG-3′
	PfEXP-1 (PF3D7_1121600)	PfEXP-1 C1 rev	5′-CGGGATCCATTTGGTTCTCCATTGGATTC-3′
	PfEXP-1 (PF3D7_1121600)	PfEXP-1 C2 for	5′-CGGAATTCGCAGGCCCACAAGTTACAGC-3′
	PfEXP-1 (PF3D7_1121600)	PfEXP-1 C2 rev	5′-CGGGATCCGTCTTCAGTGCCACTTACG-3′
	RnApoH (287774)	RnApoH for	5′-CGGAATTCGGAATCTTAGAAAATGGAGTTG-3′
	RnApoH (287774)	RnApoH rev	5′-CGGGATCCAATGAGACATTCAATTTTGATTCTG-3′
	HsApoH (350)	RnApoH for	5′-CGGAATTCGGAATCTTAGAAAATGGAGTTG-3′
	HsApoH (350)	RnApoH rev	5′-CGGGATCCAATGAGACATTCAATTTTGATTCTG-3′
	MmApoH (11818)	RnApoH for	5′-CGGAATTCGGAATCTTAGAAAATGGAGTTG-3′
	MmApoH (11818)	RnApoH rev	5′-CGGGATCCAATGAGACATTCAATTTTGATTCTG-3′
shRNA target sequences	MmApoH (11818)	shRNA 176	5′-GAGGGATGAGACGGTTTACCT-3′
	MmApoH (11818)	shRNA 725	5′-GCCCAGAAGAAGCGGAATGTA-3′
		Non silencing 1	5′-GTAACGACGCGACGACGTAA-3′
Recombinant protein for antibody production	PbEXP-1 (PBANKA_092670)	PbEXP-1 CT for	5′-CGGGATCCAGACATCCATTCCAAATTGGC-3′
	PbEXP-1 (PBANKA_092670)	PbEXP-1 CT rev	5′-CCGCTCGAGTTATTGTTGAAGATTTGGGCATGTTAAG-3′
siRNA IDs	HsApoH (350)	ApoH_1	6678 (Ambion)
	HsApoH (350)	ApoH_2	146892 (Ambion)
	HsApoH (350)	ApoH_3	146894 (Ambion)
	HsSRB1 (949)	SRB1	s2649 (Ambion)

Fig. 2.

Down-regulation of host ApoH severely impairs parasite development in vitro and in vivo. (A) Parasite load of Huh7 cells transfected with three anti-ApoH siRNAs (ApoH_1–3) and infected with luciferase-expressing Pb parasites, measured 44 hpi by luminescence. Results were normalized to the infection load of cells transfected with a scrambled siRNA (Neg) used as negative control. (B) Parasite load of Huh7 cells transfected with ApoH_1 siRNA and infected with WT Pb parasites measured 48 hpi by qRT-PCR. Results were normalized to the infection load of cells transfected with Neg siRNA (negative control). Cells transfected with a siRNA targeting the SRB1 gene (72) were used as positive controls. Human HPRT was used as a housekeeping gene. (C) EEF numbers in Pb-infected Huh7 cells previously transfected with ApoH_1 siRNA measured by immunofluorescence microscopy 48 hpi. Results were normalized to the number of EEFs in cells transfected with Neg siRNA (negative control). Three independent experiments were performed, and data from one representative experiment are shown in A and B. Data in C correspond to the pool of two independent experiments. (D) EEF sizes in Pb-infected Huh7 cells previously transfected with ApoH_1 siRNA (n = 173) measured by immunofluorescence microscopy 48 hpi. Pb-infected cells transfected with Neg siRNA (n = 326) were used as negative controls. Three independent experiments were performed, and data from one representative experiment are shown in D, where each symbol represents one parasite. For A–D: *P < 0.05; **P < 0.01 (Mann–Whitney test). ns, not significant. Data are shown as mean (in D indicated by red line) ± SD. (E) Representative images of Pb EEFs in Huh7 cells transfected with Neg siRNA or ApoH_1 siRNA at 48 hpi stained with anti-HSP70 (red) and anti-UIS4 (green) antibodies. Nuclei were stained with DAPI. (Scale bars: 10 µm.) (F) qRT-PCR analysis of ApoH expression in the livers of mice. Five C57BL/6 mice per group were i.v. injected with 5 × 10¹⁰ rAAV particles carrying different shRNA constructs and 14 d later infected with 10,000 Pb WT sporozoites. Forty-two hours after Pb infection, livers were isolated. The no-rAAV group received only the parasite injection. Results were normalized to the no-rAAV group, which was set to 1 (dotted line). Empty vector (Neg) and nonsilencing (Ns1) shRNAs were used as negative control; 176 and 725 are MmApoH-specific shRNAs. Mouse GAPDH was used as a housekeeping gene. (G) Pb liver infection in the same mice as in F was normalized to the no-rAAV group. All mice were age-matched. For F and G: *P < 0.05; ***P < 0.001 (one-way ANOVA, Dunnett‘s multiple comparison test). Data are shown as mean (indicated by red line) ± SD. Each symbol represents one mouse (n = 5).

To further assess the role of ApoH in an in vivo setting, ApoH expression was specifically down-regulated in livers of mice before they were infected with Pb parasites. To this end, we made use of recombinant adeno-associated viral (rAAV) particles of serotype 8, which exhibit a characteristic hepatotropism (45–48), to deliver anti-MmApoH–specific short hairpin RNAs (shRNAs) to the mouse livers. At 14 d after rAAV injection, mice were infected with Pb WT sporozoites. qRT-PCR measurements of the knockdown efficiency of anti-MmApoH shRNAs (176 and 725; Table S3) 42 hpi showed that both shRNAs significantly reduced ApoH expression levels in comparison with the control group that received no rAAV (Fig. 2F). Interestingly, injection of any rAAV elevated parasite liver load compared with the no-rAAV group, which only received the parasite injection (Fig. 2G). This effect seemed to be solely due to the presence of rAAVs, because parasite liver burden was also elevated in the empty vector control group (Neg), which did not encode any shRNA (Fig. 2G). Most importantly, parasite liver burden of mice injected with either anti-ApoH shRNA was significantly reduced compared with the nonsilencing shRNA (Ns1) or empty vector control groups (Fig. 2G). Together, our results show that ApoH plays a crucial role during hepatic infection by Pb parasites, both in vitro and in vivo.

Truncation of the C2 Region of EXP-1 Impairs ApoH Colocalization and Internalization.

Our data so far indicated that the host protein ApoH contributes to Pb development, presumably through its interaction with the C2 region of its interaction partner, EXP-1. Despite several attempts, a full knockout of PbEXP-1 could not be obtained by using either conventional standard Pb transfection or PlasmoGEM (PbGEM_332867) vectors (Fig. S4 A–D) (49). Moreover, a promoter-exchange approach did not result in transgenic parasites, which would have expressed EXP-1 only during blood-stage and late liver-stage development (Fig. S4 E and F). Because our data using the yeast system highlighted the importance of the C2 region of EXP-1 for the interaction with ApoH (Fig. 1C), transgenic Pb parasite lines expressing a C2-truncated version of PbEXP-1, PbEXP-1∆C2, were generated in both the Pb and PbGFP-luc_con parasites (clone 2 or clones 3 and 4, respectively) (Fig. 3A). Interestingly, even though EXP-1 appeared to be refractory to gene deletion because of its essentiality for the intraerythrocytic development, we were able to isolate this EXP-1 truncation mutant after successful transfection and limited dilution. The resulting PbEXP-1ΔC2 parasites completed their life cycle indistinguishably from WT parasites up until and including hepatocyte invasion (Table S4). This result indicates that the carboxyl-terminal C2 part of EXP-1 is not essential for intraerythrocytic parasite development because we were able to generate deletion mutants lacking this particular portion of the PVM-resident protein. The characteristic movement of infectious sporozoites and their invasion capacity (Fig. S5 A and B) were also not impaired in the PbEXP-1ΔC2 parasites. Additionally, immunofluorescence microscopy analysis of Huh7 cells infected with clonal PbEXP-1∆C2 lines and stained with antibodies against the PVM proteins UIS4 and EXP-1 indicated that neither the integrity of the vacuolar membrane nor the localization of EXP-1 were compromised in transgenic parasites (Fig. 3B). However, a significantly decreased colocalization of EXP-1 and ApoH was observed in additional immunofluorescence experiments using PbEXP-1∆C2–infected Huh7 cells, compared with WT Pb-infected cells (Fig. 3 C and D). Furthermore, quantification of accumulated ApoH in liver stages of PbWT and PbEXP-1∆C2 parasites showed a significantly reduced uptake of ApoH in intrahepatic stages of parasites expressing the C-terminally truncated version of EXP-1 (Fig. 3E). This finding clearly suggests that the interaction between the two proteins is severely compromised when the C2 region of EXP-1 is truncated, in agreement with our data shown in Fig. 1. Additionally, it points to a hitherto-undescribed instance of an interaction between a Plasmodium protein at the PVM and a host protein, leading to the accumulation of the latter inside the parasite.

Fig. S4.

PbEXP-1 knockout and promoter exchange strategies. (A) Replacement strategy for generation of a PbEXP-1 knockout parasite line with b3D+ vector. (B) Genotyping of parental transfected parasite lines (P1–P2) shows no integration of plasmid. (C) Replacement strategy for generation of a PbEXP-1 knockout parasite line with PlasmoGEM vector 332867. (D) Genotyping of parental (P1–P4) and transfer (T1–T4) transfected parasite lines shows no integration of plasmid. (E) Integration strategy for generation of PbEXP-1 promoter exchange, with EXP-1 expression under the control of the PbAMA1 promoter. (F) Genotyping of parental (P1–P5) transfected parasite lines shows no integration of plasmid. Binding sites for primers used for genotyping are indicated by arrows. V, vector; WT, WT gDNA.

Fig. 3.

Truncation of EXP-1 CT leads to both reduced ApoH colocalization and internalization. (A, Upper) Replacement strategy for generation of the PbEXP-1 C2 deletion mutant (∆C2). (A, Lower) Successful integration is shown for three clonal transgenic parasite lines (cl2 in Pb, cl3 and cl4 in PbGFP-luc_con background) generated in two independent transfections. Binding sites for primers used for genotyping are indicated by arrows. V, vector; WT, WT gDNA. (B) Immunofluorescence microscopy analysis of WT Pb and PbEXP-1∆C2 clone 2 liver stages 48 hpi. Parasite cytoplasm is stained with anti-HSP70 antibody (green), the PVM with anti-UIS4 or anti–EXP-1 antibodies (red), and nuclei with DAPI (blue). (Scale bars: 10 µm.) (C) Colocalization of parasite EXP-1 (red) and host ApoH (green) in WT and PbEXP-1∆C2 parasite lines. (C, Left) representative immunofluorescence microscopy images. (Scale bars: 10 µm.) (C, Right) Histograms showing the intensity of the red and green signals across the section indicated in the images in C, Left. (D) Quantification of EXP-1–ApoH colocalization. Colocalization was measured as percent of total EXP-1 that colocalizes with ApoH. (E) Accumulation of ApoH was measured for liver stages of WT and PbEXP-1∆C2 parasites. For D and E: **P < 0.01; ***P < 0.001 (Mann–Whitney test). Two independent experiments were performed, and data from one experiment are shown as mean ± SD.

Table S4.

Analysis of mosquito stages of WT Pb and PbGFP-luc_con and PbEXP-1 ∆C2 and PbGFP-luc_conEXP-1 ∆C2 transgenic clonal parasite lines

Parasite population	Exflagellation observed at day	Oocyst prevalence, % (day 12–13)	Mean no. of mg spz (day 13–16)	Mean no. of sg spz (day 20–22)
PbWT	5	187.5 (±17.7)*	2,800*	19,070 (±6,970)*
PbEXP-1∆C2 cl2	5	81.3 (±26.5)*	104,780*	19,140 (±16,770)*
PbGFP-lucconWT	5.5	55.8 (±29.2)†	28,410 (±11,290)*	8,840 (±7,720)†
PbGFP-lucconEXP-1∆C2 cl3	5.5	69.5 (±9.3)†	23,890 (±15,460)*	7,870 (±6,310)†
PbGFP-lucconEXP-1∆C2 cl4	5.5	81.3 (±26.5)*	59,860 (±3,360)*	12,080 (±500)*

^*Mean of two experiments (±SD).

^†Mean of three experiments (±SD).

Fig. S5.

Gliding motility and invasion assays of salivary gland sporozoites of PbWT and EXP-1∆C2 clonal parasite lines. (A) Sporozoites were allowed to glide on BSA-coated glass slides for 30 min. CSP trails were stained, and percentages of gliding (i.e., more than three circles) vs. nongliding parasites were calculated. Deletion mutants showed no significant difference compared with respective WT strains (two-way ANOVA). (B) Sporozoites were allowed to invade Huh7 cells for 90 min. Percentages of sporozoites outside and inside of Huh7 cells were calculated. Deletion mutants showed no significant difference compared with respective WT strains (two-way ANOVA). n, total number of sporozoites counted. Data are shown as mean ± SD.

Truncation of the C2 Region of EXP-1 Impairs Intrahepatic Plasmodium Development.

We next assessed whether deletion of the utmost CT of PbEXP-1 (PbEXP-1∆C2) would equally result in a reduced intrahepatic development when the parasite binding partner is no longer able to interact with the host-cell factor ApoH at the parasite/host interface, hence presumably severing the liver-stage parasite from its exogenous nutrient supply. A detailed analysis of Huh7 cell infection by PbEXP-1∆C2 mutant lines showed that truncation of the EXP-1 C2 region leads to a phenotype similar to that observed when host ApoH expression was down-modulated by RNAi (Fig. 2 C–E). In fact, our analyses of infected Huh7 cells 48 h after sporozoite addition showed that infection with any of the three PbEXP-1∆C2 clones led to lower EEF numbers (Fig. 4A) and sizes (Fig. 4B) than those yielded by their counterparts containing full-length EXP-1. In addition, we generated a complementation parasite line (PbEXP-1compl) by applying the same genetic strategy as for the PbEXP-1∆C2 transgenic line (Fig. S6A). The PbEXP-1compl parasite displayed an integral vacuolar membrane with similar localizations for UIS4 and EXP-1 (Fig. S6B). Furthermore, its intrahepatic growth efficiency (Fig. S6 C and D) was comparable with that of Pb WT parasites, despite a currently unexplained small, yet significant, increase in EEF numbers of the PbEXP-1compl parasite relative to the WT counterpart (Fig. S6C). The growth defect of PbEXP-1∆C2 parasites observed during intrahepatic development was further demonstrated by quantifying membrane-enclosed first-generation merozoites (merosome formation) in Huh7 cells. Our results indicate that merosome formation is clearly delayed in transgenic PbEXP-1∆C2 parasites compared with their WT parasite controls (Fig. S7A). This finding was shown by the significant difference in number of merosomes detected in the former compared with the latter at 65 hpi, which was no longer observed at the later, 72-hpi time point (Fig. S7A). Collectively, these results provide further support to the notion that the interaction between the C2 region of the parasite’s EXP-1 protein and the host’s ApoH molecule plays an important role during the hepatic stage of Plasmodium infection.

Fig. 4.

Truncation of EXP-1 CT leads to impaired intrahepatic parasite development in vitro and decreased liver infection in vivo. (A) EEF numbers in Huh7 cells infected with three PbEXP-1∆C2 clonal transgenic parasite lines compared with their respective WT controls measured 48 hpi by immunofluorescence microscopy. Results were normalized to WT controls. Data from two independent experiments were pooled and are shown as mean ± SD. (B) EEF sizes in Huh7 cells infected with three PbEXP-1∆C2 clonal transgenic parasite lines compared with their respective WT controls measured 48 hpi by immunofluorescence microscopy (mean indicated by red line). Data from two independent experiments were pooled, and each symbol represents one parasite (PbWT, n = 350; cl 2, n = 362; PbWT (GFP-Luc_con), n = 309; cl 3, n = 244; cl 4, n = 235). (C) Parasite load in the livers of C57BL/6 mice infected with Pb WT or PbEXP-1∆C2 clonal lines measured by qRT-PCR 42 hpi after i.v. injection of 10,000 sporozoites. Parasite 18S rRNA expression was normalized to respective WT controls, which were set to 100% (dotted line). Mouse HPRT was used as the housekeeping gene. Data were pooled from five and two independent experiments for clone 2 and clones 3/4, respectively. Each symbol represents one mouse [PbWT/cl 2, n = 22; PbWT (GFP-Luc_con) clo 3/cl 4, n = 8]. (D) Mean parasitemia (±SEM) of female C57BL/6 mice after i.v. infection with 10,000 sporozoites from either WT or PbEXP-1∆C2 (n = 8). (E) Multiplication rates of WT and PbEXP-1∆C2 intraerythrocytic stages after inoculation of 10⁶ infected red blood cells of each parasite line (n = 3). Growth rates of the two parasite lines were calculated daily and are shown as mean. No significant (ns) difference between WT and PbEXP-1∆C2 was detected. For A–C and E: Data are shown as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001 (Mann–Whitney test).

Fig. S6.

Generation and functional characterization of PbEXP-1compl parasites. (A) Replacement strategy for generation of the PbEXP-1compl parasite line. Successful integration is shown for one clonal transgenic parasite line (cl3.2). Binding sites for primers used for genotyping are indicated by arrows. WT, WT gDNA. (B) Immunofluorescence microscopy analysis of WT Pb and PbEXP-1compl liver stages 48 hpi. Parasite PVM integrity was confirmed by using anti-UIS4 or anti–PbEXP-1^c antibodies (red) and nuclei with DAPI (blue). (Scale bars: 10 µm.) (C) EEF numbers in Huh7 cells infected with PbEXP-1compl and WT. Data are represented as normalized to WT and are shown as mean ± SD. (D) Developmental progression of intrahepatic PbEXP-1compl and WT parasites. EEF sizes were normalized to Pb WT (PbEXP-1compl, n = 234; respective PbWT, n = 216). Mean is indicated by red line. For C and D: ***P < 0.001 (Mann–Whitney test). ns, not significant.

Fig. S7.

Merosome formation and liver burden at late time points after Plasmodium sporozoite infection. (A) Quantification of merosome formation. At indicated time points after invasion, merosome formation was quantified in culture supernatants of Pb WT and PbEXP-1∆C2–infected Huh7 cells per well. Data are shown as mean ± SD. *P < 0.05 (unpaired t test). (B) Parasite load in the livers of C57BL/6 mice infected with Pb WT or PbEXP-1∆C2 clonal line measured by qRT-PCR 42, 72, and 96 hpi after i.v. injection of 10,000 sporozoites. Results were normalized to the liver load of mice 42 h after Pb WT infection, which was set to 100% (dotted line). Mouse GAPDH was used as the housekeeping gene. *P < 0.05 (Mann–Whitney test). (B, Lower) Table shows blood-stage patency of Pb WT and PbEXP-1∆C2 parasites after sporozoite administration.

Impairment of the EXP-1/ApoH Interaction Leads to a Decrease in Liver Parasite Burden and in the Severity of Pathology.

Having established that the truncation of the C2 region of EXP-1 leads to a decreased interaction with ApoH in vitro, we next investigated the in vivo infection of C57BL/6 mice by PbEXP-1∆C2 parasites. Again, similarly to what we observed when ApoH expression was down-modulated in vivo (Fig. 2G), our results showed a decrease in the liver parasite load of mice infected with either of the PbEXP-1∆C2 clones at 42 hpi, compared with their respective controls (Fig. 4C). Moreover, late intrahepatic development (72–96 hpi) was additionally significantly reduced in PbEXP-1∆C2 compared with WT infections (Fig. S7B). We then asked whether an impairment of the EXP-1/ApoH interaction would have any consequences in terms of disease severity. To evaluate this possibility, C57BL/6 mice were infected i.v. with 10,000 WT Pb or PbEXP-1∆C2 sporozoites, and infection was allowed to proceed to the blood stage. Daily monitoring of blood parasitemia (Fig. 4D), disease symptoms, and mouse survival showed a 1-d delay in the patency of blood-stage infection and a significantly increased survival of the mice infected with the transgenic parasites, compared with their WT counterparts (Table 1). Finally, we infected mice with 10⁶ erythrocytic stages of both WT and PbEXP-1∆C2 parasites and followed the course of parasitemia. Our results showed that, when the intrahepatic phase of infection is bypassed, no significant differences are observed in the overall replication rates of the parasites in the blood (Fig. 4E).

Table 1.

Prepatent period of C57BL/6 mice infected with 10,000 PbEXP-1∆C2 sporozoites is prolonged and results in increased survival rates compared with WT controls

Parasite strain	Prepatent period, *	P value (clone vs. WT control)†	No. of surviving/no. of infected animals	P value (clone vs. WT control)†
PbWT	3.7	—	3/8	—
PbEXP-1 ∆C2 cl2	4.7	0.0027	7/8	0.0359
PbGFP-luccon WT	3.3	—	3/8	—
PbGFP-lucconEXP-1∆C2 cl3	4.9	0.0002	8/8	0.0359
PbGFP-lucconEXP-1∆C2 cl4	4.7	0.0018	7/7	0.4385

^*Number of days after sporozoite inoculation until infected erythrocytes were detected in blood smear examination.

^†Statistical analysis was performed by using the log-rank (Mantel–Cox) test with GraphPad Prism.

These observations clearly indicate that the truncation of the region of the EXP-1 protein responsible for the interaction with ApoH has a significant impact solely during the intrahepatic stage of parasite development.

Discussion

In this study, we identified a host–Plasmodium interaction between the parasitic PVM-resident protein PbEXP-1 and host-cell ApoH and demonstrated its importance in facilitating the parasite’s preerythrocytic development. EXP-1 belongs to a short list of parasite-derived proteins that have been shown to localize at the intrahepatic host–parasite interface, which includes LISP1/2, IBIS1, UIS3/4, and components of the PTEX translocon (23, 24, 33, 50–53). Although it was identified >30 y ago and is abundantly expressed during blood- and liver-stage parasite development (28, 32), the function of Plasmodium EXP-1 largely remained elusive. Recently, a study by Lisewski et al. suggested that in Pf erythrocytic stages, EXP-1 functions as a membrane GST, efficiently degrading cytotoxic hematin (42). Whether EXP-1 also acts as GST during intrahepatic development is not known. However, it is tempting to speculate that the GST function may be more important during the parasite’s blood-stage development, when large amounts of toxic hematin are generated, whereas recruitment of host-cell nutrients might be more important during the extensive liver-stage replication phase. In fact, our study strongly suggests that EXP-1 plays distinct roles during the liver and blood stages of infection. As shown by our results with the PbEXP-1∆C2 parasite, the interaction of the EXP-1 CT with ApoH during the hepatic stage fulfills a function that is not essential for blood-stage parasite survival. So far, such a specific host–parasite interaction during intrahepatic development was only described for the UIS3 protein, which binds host LFABP (25). Interestingly, down-regulation of LFABP1 expression impairs liver-stage growth (25), similarly to what we observe when ApoH expression is silenced. However, unlike UIS3, EXP-1 is a cross-stage antigen and might also function as a binding partner for host proteins during erythrocytic development.

Our analysis of the subcellular localization of EXP-1 and ApoH showed not only that the two proteins colocalize in the PVM region, but also that ApoH appears to additionally accumulate inside the parasite. However, the mechanism by which ApoH enters the parasite and crosses both the PVM and the parasite’s plasma membrane remains unknown. Interestingly, Hanson et al. recently reported that both known PVM-resident proteins, EXP-1 and UIS4, must be continuously exported to the PVM, at least during the first 30 h of intrahepatic development. This finding raises the possibility that these proteins may be trafficked back into the PV or even into the parasite’s cytoplasm (38). If EXP-1 is indeed subjected to retrograde trafficking, it might serve as a shuttle protein for ApoH. An alternative entry route would be the PVM PTEX translocon, which is also expressed during liver-stage development (53). EXP-2 is hypothesized to form the membrane-spanning pore of the PTEX translocon and may have several functions apart from protein export, such as nutrient acquisition (53). It could thus present a potential import route for ApoH, with EXP-1 functioning as a bridging or recruiting molecule. Our results show that transgenic liver-stage parasites expressing a truncated EXP-1 protein exhibited decreased colocalization with ApoH at the PVM and, more importantly, decreased uptake of ApoH into the parasite’s cytoplasm.

Our findings revealed an important role for ApoH during liver-stage, but not intraerythrocytic, Plasmodium development, because either reduced ApoH expression or expression of a truncated EXP-1 protein resulted in a significant decrease in parasite hepatic burden, both in vitro and in vivo. ApoH is an abundant plasma protein, which is mainly synthesized in the liver, but also in intestinal tissues and placenta cells (54–56). The protein comprises five repeats known as complement control protein (CCP) repeats, functioning as protein–protein interaction modules (57, 58). Although the first four repeats are structurally related, the fifth domain is more variable and contains, for example, a highly positively charged region and three, instead of two, disulfide bridges (56, 58). Several different functions have been assigned to human ApoH, such as clearing apoptotic bodies and microparticles from the circulation, interacting with oxidized LDL, or scavenging toxic compounds as well as cellular debris (59, 60). Because ApoH is responsible for the clearance of plasma liposomes, which also contain, for example, cholesterols (61), it is tempting to speculate that ApoH might transport lipids and/or cholesterols to the intracellular parasite via its direct interaction with Plasmodium EXP-1, allowing for successful intrahepatic development of the parasite. This speculation raises the additional possibility that the interaction of EXP-1 with ApoH might also serve to internalize the latter and/or provide the parasite with ApoH molecular partners.

In conclusion, our data highlight the importance of the ApoH/EXP-1 interaction for successful liver-stage Plasmodium development and contribute to our understanding of how the PVM serves as a platform for host–parasite interactions. Based on our findings, we propose that other PVM family members may also engage in specific protein interactions with host-cell proteins and thereby serve as docking proteins for important host-cell factors. Interference with such crucial parasite–host interactions may strengthen the search for innovative intervention strategies against malaria.

Materials and Methods

See additional materials and methods in SI Materials and Methods.

Ethics Statement.

All animal experiments were performed according to European regulations concerning Federation for Laboratory Animal Science Associations category B and Society of Laboratory Animal Science standard guidelines. Animal experiments were approved by German authorities (Regierungspräsidium Karlsruhe), § 8 Abs. 1 Tierschutzgesetz (TierSchG) under the license G-260/12 and by the internal animal care committee of the Instituto de Medicina Molecular and were performed according to national and European regulations. For all experiments, female C57BL/6 and outbred NMRI mice were purchased from Janvier or Charles River Breeding Laboratories, and male inbred C57BL/6J mice were purchased from Charles River Breeding Laboratories. All mice were kept under specific pathogen-free conditions within the animal facility at Heidelberg University [Interfakultäre Biomedizinische Forschungseinrichtung (IBF)] and the facilities of the Instituto de Medicina Molecular, Lisbon, respectively.

Cell Culture.

Huh7 cells, a human hepatoma cell line (a gift from Ralf Bartenschlager, Centre for Infectious Diseases, Molecular Virology, Heidelberg), were cultured in DMEM supplemented with 10% (vol/vol) FCS and 1% Antibiotic-Antimycotic (Life Technologies) or RPMI 1640 medium supplemented with 10% (vol/vol) FBS, 1% penicillin/streptomycin, 2 mM glutamine, 1 mM Hepes, and 0.1 mM nonessential amino acids. Human embryonic kidney 293 T cells (HEK293T; ATCC CRL-1573 from LGC Standards (ATCC) were cultured in DMEM supplemented with 10% (vol/vol) FCS, 1% penicillin/streptomycin, and 2 mM glutamine. All cell lines were maintained at 37 °C with 5% CO₂ and were passaged by trypsinization at ∼80% confluence.

Mosquito Rearing and Parasite Production.

For routine passage of blood-stage Pb parasites and for mosquito feedings, mice (4–6 wk; Janvier or Charles River Breeding Laboratories) were infected by i.p. injections. Anopheles stephensi mosquitoes were reared at 28 °C and 80% humidity under a 14-h/10-h light/dark cycle and fed on 10% (wt/vol) sucrose/p-aminobenzoic acid solution. Adult mosquitoes were fed on infected mice with Pb WT (62, 63) or mutant parasites and maintained at 21 °C and 80% humidity.

Parasite Strains.

Pb ANKA (62) and PbGFP-luc_con (63) WT sporozoites, as well as transgenic PbEXP-1compl and PbEXP-1∆C2 sporozoites, were obtained by dissection of salivary glands of infected A. stephensi mosquitoes bred at the Instituto de Medicina Molecular or at the Centre for Infectious Diseases in-house insectaries. After grinding, the suspension was filtered through a 70-μm cell strainer (Falcon) or centrifuged to remove mosquito debris. The Pf 3D7 strain was grown in 5% hematocrit at 37 °C, 5% CO₂, 5% O₂, and 90% N₂.

Generation of Transgenic Parasite Lines.

To generate the PbEXP-1∆C2 parasite line, a 3′ UTR fragment was amplified by using 3′ PbEXP-1for and 3′ PbEXP-1rev primers (Table S3), and a 5′ fragment including the ORF without the last 93 bp of PbEXP-1 was amplified with 5′ PbEXP-1ORFdC2for and 5′ PbEXP-1ORFdC2rev primers (Table S3) from Pb WT genomic DNA (gDNA) and cloned into the b3D⁺ vector (64). A similar strategy was used to generate the PbEXP-1compl parasite line. Briefly, a 5′ fragment comprising the complete PbEXP-1 ORF was amplified by using the primers 5′ PbEXP-1ORFdC2for and 5′ PbEXP-1ORFcompl_rev (Table S3) from Pb WT gDNA and cloned into the b3D⁺ vector (64) containing the 3′ UTR fragment as described above. Before transfection, the targeting vectors were linearized by using restriction enzymes KpnI and NotI. Transgenic Pb (62) and PbGFP-Luc_con (63) parasite lines were generated as described (65). To obtain clonal parasite populations, limited serial parasite dilutions were performed, and one parasite was administered by i.v. injection to each of 10 recipient naive NMRI mice (66). After gDNA extraction, Pb WT, PbEXP-1∆C2, and PbEXP-1compl parasites were genotyped by using specific primers (Table S3). The deletion of the last 93 bp of the PbEXP-1 ORF and the presence of the entire EXP-1 ORF in the PbEXP-1compl parasite were additionally confirmed by sequencing.

Y2H Analysis.

The Y2H technique was carried out according to the yeast protocol handbook and “Matchmaker Library Construction and Screening Kit” manual (Clontech) by using the yeast reporter strain AH109. PbEXP-1 CT fragments were cloned into the pGBKT7 plasmid. PbEXP-1 CT was used as bait to screen a rat hepatocyte cDNA library constructed with the “Make Your Own Mate & Plate Library” system (Clontech). Fourteen colonies were recovered from quadruple dropout (QDO; -Trp-Leu-Ade-His) plates. Direct interaction was confirmed by cotransformation of pGBKT7 and pGADT7-Rec plasmids containing the respective bait and prey sequences, followed by selection on QDO plates. For the generation of PbEXP-1 and RnApoH binding and activation domain fusions, the respective coding regions were amplified by PCR using primers listed in Table S3, inserted into pGBKT7 and pGADT7-Rec plasmids, and sequenced.

Far Western Blot and Antibody Inhibition Assays.

Huh7 cells (50,000) were infected with 30,000 Pb sporozoites and lysed 48 h later at 4 °C in 10 mM Tris⋅HCl (pH 7.4), 100 mM NaCl, 1% Triton X-100, 1 mM NaVO₄, 1 mM NaF, and complete EDTA-free protease inhibitors for 1 h with gentle rotation. After lysis, samples were centrifuged to remove insoluble cellular material. Sporozoite lysates were obtained from 400,000 Pb sporozoites treated as above. Pb- and Pf-infected red blood cell samples (3–5% parasitemia) were lysed in 0.1% saponin solution for 10 min followed by several washes in ice-cold PBS.

For far Western blotting experiments, protein lysates (30–40 μL) were separated by SDS/PAGE and blotted onto nitrocellulose membranes. The membranes were blocked in 5% (wt/vol) dried powdered milk in PBS containing 0.05% Tween 20 for 1 h or overnight at 4 °C. Blots were subsequently overlaid with or without rat serum with gentle mixing. After 4 h, the binding of ApoH was detected by using rabbit α-ApoH antibody (1:12,000; Abcam, ab108348), followed by a horseradish peroxidase (HRP)-conjugated rabbit secondary antibody (1:5,000; Jackson ImmunoResearch). A control lane on the blot was overlaid with α-PfEXP-1 antibody (1:700) provided by K. Lingelbach, University of Marburg, Marburg, Germany.

Antibody inhibition assays were performed as described before, except that the membranes were overlaid for 8 h with α-PfEXP-1 antibody before incubation with rat serum. ApoH binding to EXP-1 was detected by using α-ApoH antibody (1:12,000; Abcam, ab108348), followed by a HRP-conjugated rabbit secondary antibody (1:5,000; Jackson ImmunoResearch).

All antibodies were diluted in 5% (wt/vol) dried powdered milk in PBS containing 0.05% Tween 20, and nitrocellulose blots were overlaid with antibodies for 1 h at room temperature. The blots were developed by the addition of ECL substrate (Thermo Scientific Pierce), a chemiluminescent substrate to detect peroxidase activity from HRP-conjugated antibodies.

Immunoprecipitation.

The pCMV-HA plasmid encoding the whole PbEXP-1 CT (human codon-optimized) was obtained from Clontech. RnApoH was amplified by PCR from a rat hepatocyte cDNA library (Table S3), and the PCR product was cloned into pCMV-myc (Clontech). All constructs were analyzed by DNA sequencing and enzyme digestion. Protein expression was confirmed by Western blotting.

HEK293T cells (3 × 10⁶ cells per plate) were transiently transfected by using FuGENE 6 (Roche Molecular Biochemicals) with expression plasmids carrying PbEXP-1 (5 μg) and RnApoH (5 μg) and cultured for 48 h. They were subsequently washed with ice-cold PBS and lysed in ice-cold lysis buffer [10 mM Tris⋅HCl, pH 7.4, 100 mM NaCl, 1% Triton X-100, 1 mM NaVO₄, 1 mM NaF, and complete EDTA-free protease inhibitors]. Lysates were cleared by centrifugation, and the supernatants were incubated with α-HA antibody (Abcam, ab9110; or Sigma, H9658) at 4 °C for 2 h. Immune complexes were recovered by incubation with protein G-Sepharose beads (Amersham Biosciences) for 1 h at 4 °C. After three washes with ice-cold lysis buffer, protein complexes were dissociated from the protein G by addition of Laemmli's sample buffer and boiled for 10 min at 100 °C. Samples were then centrifuged, and the supernatant was loaded onto polyacrylamide gels. Separated proteins were transferred to nitrocellulose membranes, and immunoprecipitates were then analyzed by immunoblotting using appropriate primary antibodies (1:1,000, Clontech, α-cmyc 631206; or 1:12,000, Abcam, α-ApoH Ab 108348), followed by HRP-conjugated secondary antibodies (1:5,000; Jackson ImmunoResearch) and developed as described above.

Assessment of in Vivo Plasmodium Infection.

For quantification of liver infection, C57BL/6 mice (6–8 wk) were injected i.v. with 10,000 Pb WT or PbEXP-1∆C2 sporozoites. Liver infection load was quantified after 42–46, 72, and 96 hpi by qRT-PCR analysis. To this end, whole livers were homogenized in 3 mL of denaturing solution (4 M guanidine thiocyanate, 25 mM sodium citrate, pH 7, 0.5% N-lauroylsarcosine, and 0.7% (vol/vol) β-mercaptoethanol in diethyl pyrocarbonate-treated water) or 4 mL of RLT buffer (Qiagen) containing 1% β-mercaptoethanol, followed by RNA extraction and qRT-PCR analysis as described.

For assessment of blood-stage infection, C57BL/6 mice (6–8 wk) were injected i.v. with 10,000 Pb WT or PbEXP-1∆C2 sporozoites or with 1 × 10⁶ infected erythrocytes from these parasite lines. Parasitemia was monitored daily by examination of Giemsa-stained blood smears and was used to calculate the number of asexual blood stages in mice for each day after infection. The growth or multiplication rate was determined for each day according to the following formula: (no. of total parasites/no. of parasites injected)^1/d after infection (67, 68).

Mice were observed daily for signs of experimental cerebral malaria (ECM), including the following neurological symptoms: hemiplegia and paraplegia, ataxia, convulsion, and/or coma. Animals exhibiting signs of severe disease and mice that did not develop ECM but developed hyperparasitemia and anemia were euthanized.

siRNA-Mediated in Vitro Gene Expression Silencing.

Huh7 cells (40,000) were seeded and transfected with either of three predesigned ApoH-silencer double-stranded RNAs (Table S3), by using Lipofectamine RNAimax (Life Technologies) according to the manufacturer's instructions. Thirty-six hours later, cells were infected with 10,000 luciferase-expressing Pb parasites. A scrambled siRNA and a siRNA targeting the SRBI (Table S3) gene were used as negative and positive controls, respectively. Down-regulation of ApoH was confirmed by either qRT-PCR or Western blot analysis using α-ApoH antibody (1:12,000; Abcam, ab108348) and HRP-coupled anti-rabbit IgG secondary antibody (1:5,000; Jackson ImmunoResearch).

Infection was assessed by immunofluorescence microscopy, qRT-PCR, or bioluminescence (Biotium). In the latter case, cell viability was assessed 44 hpi by the CellTiter-Blue assay (Promega) according to the manufacturer’s protocol, and parasite load was measured 48 hpi by using a multiplate reader Infinite M200 (Tecan).

rAAV-Mediated in Vivo Gene Expression Silencing.

Self-complementary rAAV vectors expressing shRNAs against murine ApoH under a H1 promoter were generated as described (69). Targeted sense sequences for MmApoH shRNAs 176 and 725 (Table S3) were selected by using siRNA Wizard (InvivoGen), and rAAV vectors expressing nonsilencing (Ns1) shRNAs were used as controls. All rAAV vectors were produced by triple transfection of HEK293T cells (30 plates with 4 × 10⁶ cells per shRNA) with equal amounts (14.6 µg per plate) of AAV helper plasmid encoding AAV8 capsid genes, an adenoviral helper construct and rAAV vector plasmids (70) encoding Ns1 or anti-MmApoH shRNAs. rAAV vector particles were purified by cesium chloride density gradient centrifugation (69).

For in vivo ApoH silencing, 5 × 10¹⁰ rAAV particles were administered to female C57BL/6 mice by i.v. injection. Fourteen days later, mice were infected by i.v. injection of 10,000 Pb WT sporozoites. Liver infection load was quantified 42 hpi by qRT-PCR.

Assessment of Gene Expression by qRT-PCR.

RNA was extracted from hepatic cells or tissues by using the RNeasy Mini kit (Qiagen), and cDNA was obtained by reverse transcription (First-Strand cDNA Synthesis kit; Roche/Thermo Fisher Scientific). For quantification of infection, qRT-PCR was performed by using primers specific for Pb 18S rRNA. For assessment of gene silencing, ApoH-specific primers were used. Gene expression levels were normalized against mouse hypoxanthine guanine phosphoribosyltransferase (HPRT) or GAPDH by using the ΔΔC_T method. Primer sequences are shown in Table S3.

Immunofluorescence Microscopy.

A total of 50,000 and 25,000 Huh7 cells grown on coverslips or eight-well Lab-Tek chamber slides (Nunc), respectively, were seeded 1 d before infection with 10,000 Pb sporozoites and fixed at the indicated time points (10, 15, 18, 21, 24, and 48 hpi) after sporozoite addition. Cells grown on coverslips were fixed with 4% (vol/vol) paraformaldehyde for 15 min, washed three times with PBS, permeabilized in 0.2% (vol/vol) saponin or Triton in PBS for 20 min, blocked for 1 h with 0.1% (vol/vol) Triton in 1% (wt/vol) BSA in PBS, and incubated with the appropriate primary antibodies in blocking buffer for 1 h at room temperature or overnight at 4 °C. Primary antibodies used were mouse monoclonal antibody 2E6 against Pb heat-shock protein 70 (HSP70) (1:100) (71), goat anti-UIS4 antibody (1:1,000), chicken anti–PbEXP-1 antibody (1:700; provided by Volker Heussler, Institute of Cell Biology, University of Bern, Bern, Switzerland), mouse anti–PbEXP-1^c antibody (1:300; raised against the CT part of PbEXP-1, i.e., amino acids 103–166; see also Fig. 1A), and rabbit monoclonal anti-ApoH antibody (1:10; Sigma, HPA003732). After three washes with PBS, cells were incubated for 40 min with secondary antibodies diluted in blocking buffer. Nuclei were stained with DAPI or Hoechst, and coverslips were mounted onto microscope slides with Fluoromount G mounting medium (SouthernBiotech, catalog no. 0100-01).

Cells grown in Lab-Tek chamber slides were fixed with ice-cold methanol for 10 min, washed three times with 1% FCS in PBS, blocked overnight with 10% (vol/vol) FCS in PBS at 4 °C, and incubated with primary antibodies in blocking buffer for 1 h at 37 °C. Primary antibodies used were monoclonal anti-PbHSP70 antibody (1:50), rat anti-PbUIS4 antibody (1:300), or mouse anti–PbEXP-1^c antibody (1:300). After three washes with 1% FCS in PBS, cells were incubated for 1 h at 37 °C with secondary antibodies (1:300) diluted in blocking buffer. Nuclei were stained with DAPI, followed by addition of 50% (vol/vol) glycerol in PBS and sealing with a coverslip.

All images were acquired on Zeiss wide-field Fluorescence Axiovert 200M or on Zeiss LSM 510 META or Zeiss LSM 710 confocal laser point-scanning microscopes. Image analysis was performed by using the LSM image browser and ImageJ.

Analysis of ApoH–EXP-1 Colocalization and ApoH Accumulation.

Confocal microscopy images were obtained by sequential scanning for each channel to eliminate the cross-talk of chromophores. The background was corrected by using the threshold value for all channels. ApoH–EXP-1 colocalization was quantified by using the “co-localization highlighter” command in ImageJ software. The colocalizing pixels identified were represented as a percentage of the total EXP-1 signal. The accumulation of ApoH was assessed by using ImageJ by calculating the difference between the total ApoH signal inside the EEF and the ApoH signal in an area of the same size and shape outside the EEF and normalized to the EEF area.

Quantification of in Vitro Merosome Formation.

Huh7 cells (25,000) were seeded in eight-well Lab-Tek chamber slides (Nunc) and infected 24 h later with 40,000 Pb WT or PbEXP-1∆C2 sporozoites, and merosome formation was quantified at 48, 65, and 72 hpi. At the indicated time points, culture supernatants were collected and centrifuged at 100 × g for 5 min at room temperature. The supernatants were carefully removed, and merosomes in the pellet were counted by using a hemocytometer to estimate the total number of merosomes per well.

Statistical Analyses.

Statistical significance was assessed by using the unpaired t test, two-tailed nonparametric Mann–Whitney test, one-way ANOVA, and Dunnett's multiple comparison test, or log-rank (Mantel–Cox) test. Values of P < 0.05 were deemed statistically significant (in the figures, *P < 0.05; **P < 0.01; and ***P < 0.001). All statistical analyses were performed by using the GraphPad Prism (Version 5.0) or SigmaPlot (Version 13) software.

SI Materials and Methods

ELISA.

ELISA plates were incubated overnight with 20 µg/mL purified recombinant ApoH protein (ProSpec, pro-388) in coating buffer (carbonate buffer, pH 9.6) containing 0.01% N₃Na. Excess unbound purified protein was removed by two washing steps with PBS plus 0.05% Tween 20. Nonspecific binding was blocked with 1% BSA/PBS and 0.05% Tween 20. For direct binding studies, sonicated lysates of uninfected or Pb-infected Huh7 cells in ELISA buffer (1x PBS, 1% wt/vol BSA, and 0.1% Tween-20) were diluted with PBS, added to each well in different dilutions and incubated for 2 h at 37 °C. PbEXP-1 binding to ApoH was detected by addition of rabbit polyclonal antiserum raised against PbEXP-1 (provided by V. Heussler), followed by HRP-conjugated secondary antibody. All antibodies were incubated for 1 h at room temperature in ELISA buffer. Plates were washed, and 50 µL of isotype-specific goat anti-rabbit or -chicken HRP conjugates (1:5,000; Jackson ImmunoResearch) were added. ELISAs were developed with BD optEIA-TMB substrate reagent set (BD Biosciences, catalog no. 555214). OD at 450 nm was determined after 5-min incubation with an automatic ELISA plate reader (Envision Multilabel Reader 2104; Perkin-Elmer). Controls included uncoated wells, and ApoH-coated wells to which no lysates were added and noninfected Huh7 cells as negative control.

Gliding Motility Assay.

The 10,000 salivary gland sporozoites were allowed to glide for 30 min at 37 °C on BSA-coated eight-well glass slides, before they were fixed with 4% (vol/vol) PFA/PBS for 10 min at room temperature. After three washing steps with 1% FCS/PBS and blocking with 10% (vol/vol) FCS/PBS for 15–30 min, circumsporozoite protein (CSP) trails were stained with primary mouse anti-PbCSP antibody (1:300) in blocking buffer for 30 min at 37 °C. After three washing steps, secondary anti-mouse Alexa Fluor 488 (1:300; Life Technologies, A11029) was added for 30 min at 37 °C. Samples were covered with 30% (vol/vol) glycerol/PBS, and coverslips were mounted onto glass slides. Samples were analyzed by fluorescence microscopy, and sporozoites with trails of three or more full circles were regarded as gliding sporozoites.

Invasion Assay.

Huh7 cells (25,000–30,000) were infected with 10,000 Pb WT or PbEXP-1∆C2 sporozoites, and sporozoite invasion was terminated after 90 min by removal of medium and fixation with 4% (vol/vol) PFA/PBS for 10 min at room temperature. After three washing steps with 1% FCS/PBS and blocking with 10% FCS/PBS, extracellular sporozoites were stained with primary mouse anti-PbCSP antibody (1:300). Wells were washed three times, and secondary anti-mouse Alexa Fluor 488 antibody (1:300; Life Technologies, A11029) was added. Samples were washed three times, and cells were permeabilized with ice-cold methanol for 15 min at room temperature. Staining for extracellular as well as intracellular sporozoites was repeated as described above with primary anti-PbCSP antibody (1:300) and secondary anti-mouse Alexa Fluor 546 antibody (1:300; Life Technologies, A11003). During the last antibody-incubation step, nuclei were stained with Hoechst for 3 min. All blocking and antibody incubation steps were performed for 1 h at 37 °C or overnight at 4 °C. Samples were covered with 30% (vol/vol) glycerol/PBS, and coverslips were mounted onto glass slides. Samples were analyzed with Confocal LSM Axiovert 100 M microscope. Sporozoites stained with both Alexa Fluor 488 and 546 antibodies were regarded as extracellular; sporozoites stained only with Alexa Fluor 546 were regarded as intracellular.