A crucial role for the C‐terminal domain of exported protein 1 during the mosquito and hepatic stages of the Plasmodium berghei life cycle

Abstract Intracellular Plasmodium parasites develop inside a parasitophorous vacuole (PV), a specialised compartment enclosed by a membrane (PVM) that contains proteins of both host and parasite origin. Although exported protein 1 (EXP1) is one of the earliest described parasitic PVM proteins, its function throughout the Plasmodium life cycle remains insufficiently understood. Here, we show that whereas the N‐terminus of Plasmodium berghei EXP1 (PbEXP1) is essential for parasite survival in the blood, parasites lacking PbEXP1's entire C‐terminal (CT) domain replicate normally in the blood but cause less severe pathology than their wild‐type counterparts. Moreover, truncation of PbEXP1's CT domain not only impairs parasite development in the mosquito but also abrogates PbEXP1 localization to the PVM of intrahepatic parasites, severely limiting their replication and preventing their egress into the blood. Our findings highlight the importance of EXP1 during the Plasmodium life cycle and identify this protein as a promising target for antiplasmodial intervention.

when female Anopheles mosquitoes deposit sporozoites (spz) in their skin during a blood meal. Spz then travel in the bloodstream and eventually arrest in the liver sinusoids. After traversing several cells, spz productively infect a hepatocyte, initiating the asymptomatic but obligatory liver stage of infection. Intrahepatic parasites are enclosed within a parasitophorous vacuole (PV) membrane (PVM), inside which they differentiate into round-shaped exoerythrocytic forms. A process of massive replication ensues, culminating in the formation of thousands of blood-infective merozoites, which are eventually released into the blood stream, initiating the blood stage of infection (Prudencio, Rodriguez, & Mota, 2006). During this phase, merozoites cyclically invade, replicate inside, and burst red blood cells, leading to pathology.
Concomitantly, a few blood stage parasites differentiate into male and female gametocytes, the parasite's sexual forms that can be ingested by an Anopheles mosquito during a subsequent blood meal. The sexual stage of the Plasmodium life cycle occurs inside the mosquito and is initiated by the differentiation of gametocytes into gametes, which fuse to generate zygotes. These forms then transform into ookinetes, which penetrate the mosquito's midgut (MG) wall and develop into oocysts.
The latter undergo a process of maturation, during which spz are formed, eventually leading to oocyst rupture and sporozoite release.
Free spz eventually invade the mosquito's salivary glands (SGs), where they remain ready to initiate a new mammalian infection (Meibalan & Marti, 2017).
Plasmodium falciparum (Pf) exported protein 1 (EXP1), previously known as antigen 5.1 (Hope, Hall, Simmons, Hyde, & Scaife, 1984), circumsporozoite-related antigen (Coppel et al., 1985), or antigen QF116 (Kara et al., 1988), is a 17.1-kDa PVM protein with a 162 amino acid (aa) long sequence, which is highly conserved among the Plasmodium species (Doolan et al., 1996). EXP1 is expressed by the parasite's blood stages (Hope et al., 1984;Simmons, Woollett, Bergin-Cartwright, Kay, & Scaife, 1987) and is essential for parasite proliferation during this phase of infection (Maier et al., 2008). EXP1 has recently been shown to localize to dense granules in merozoites and to translocate to the PVM after erythrocyte invasion (Iriko et al., 2018), where it has been proposed to function as a glutathione S-transferase (GST; Lisewski et al., 2014). EXP1 is also expressed by liver stage parasites (Sanchez, Rogers, Mellouk, & Hoffman, 1994) and localizes to the PVM, to which it is continuously transported throughout at least the first 30 hr of intrahepatic development (Hanson et al., 2013). We have recently shown that Plasmodium berghei (Pb) EXP1 binds to host Apolipoprotein H (ApoH) during the hepatic phase of infection and that this interaction plays a pivotal role throughout parasite development inside liver cells (Sa et al., 2017).
It has recently been reported that PfEXP1 trafficking to the PVM happens independently of the putative Plasmodium translocon of exported proteins (PTEX; de de Koning-Ward et al., 2009) and does not require unfolding of the protein (Tribensky et al., 2017).
We have previously shown that the C2 region of PbEXP1's CT domain mediates the interaction with, and internalisation of, host ApoH during the hepatic stage of infection (Sa et al., 2017) Figure S1a Figure S2c). Collectively, our results confirm the non-essentiality of PbEXP1's CT domain for parasite growth in the blood, while suggesting a possible role for PbEXP1 on malaria pathology and disease severity.

| Truncation of the CT domain of PbEXP1 impairs parasite sporogony and intrahepatic development in vitro and in vivo
Having shown that Pb parasites expressing a CT-truncated version of PbEXP1 retain their ability to replicate in the blood, we proceeded to examine the effect of the CT deletion on the remaining stages of the parasite's life cycle. To this end, mosquitoes were allowed to feed on the blood of PbEXP1ΔCT-or Pb GFP EXP1ΔCT-infected mice, as well as on mice infected with each of these parasites' parental lines, employed as controls. The numbers of oocysts developing in the mosquitoes' MGs, and of spz present in the mosquitoes' SGs, were then

| Truncation of the CT domain of PbEXP1 compromises the protein's localization and function during the liver stage of infection
We have previously shown that the localization of PbEXP1 on the PVM inside hepatic cells is not compromised in PbEXP1ΔC2 parasites, which lack the C2 region of the protein's CT (Sa et al., 2017). To ascertain the impact of the truncation of the entire CT domain on PbEXP1's ability to localize to the PVM, we employed immunofluorescence microscopy to analyse protein localization in Huh7 cells infected with PbWT, Not surprisingly, we also found that PbEXP1ΔCT parasites fail to internalise ApoH, similarly to what was previously observed for the  Despite being one of the earliest described parasite-derived PVM proteins (Coppel et al., 1985;Kara et al., 1988;Simmons et al., 1987), much remains to be understood about EXP1. In the present study, we sought to dissect the role of the protein's NT and CT domains during the life cycle of the malaria parasite. It has previously been shown that EXP1 is refractory to deletion from both the Pf (Maier et al., 2008) and Pb (Sa et al., 2017)   Impact of CT truncation on the PVM localization of PbEXP1. Colocalization of PbEXP1 (cyan) and UIS4 (red) proteins was assessed by immunofluorescence microscopy at 30 hpi, in Huh7 cells infected with 2 × 10 4 sporozoites of (a) PbWT, (b) PbEXP1ΔC2, and (c) PbEXP1ΔCT parasite lines. (a-c) Upper panels: Representative immunofluorescence microscopy images (scale bar = 10 μm); lower panels: histograms showing the relative intensity of the cyan and red signals across the section indicated in the representative images by an arrow. (d) Quantification of EXP1/ UIS4 colocalization, measured as the percentage of total EXP1 that colocalizes with UIS4. Results shown are from two coverslips infected in one experiment. Each symbol represents one parasite (n = 25). Data were analysed using two-tailed unpaired t test. n.s., non-significant. **** P < .0001. Data are shown as mean ± SEM.
PbEXP1ΔCT parasites are unable to internalise ApoH and display an impairment in hepatic development. These phenotypes are similar to those observed for the PbEXP1ΔC2 parasite and can therefore be ascribed to the C2 region of the EXP1 protein (Sa et al., 2017).
Although the liver stages of Pb parasites lacking the C2 region of EXP1 are no longer able to uptake host ApoH, PbEXP1 retains its ability to localize to the PVM (Sa et al., 2017). We now show that PVM targeting of EXP1 during Pb hepatic development is completely abolished in the absence of the protein's entire CT domain, indicating that EXP1 localization is dictated by the C1 region of the protein.
Importantly, these results are validated by the demonstration that a complementation parasite line (PbEXP-1compl) displays hepatic infection and EXP1 localization/function similar to those of its wild-type counterpart (Sa et al., 2017). Thus, although a topology where EXP1's NT domain is exposed to the host cell cytosol and the CT domain faces the parasite PV lumen has been proposed (Lisewski et al., 2014), our results favour the widely preferred view that the CT protrudes into the host cell cytosol, whereas the NT faces the inside of the PV (Gunther et al., 1991;Tribensky et al., 2017). These results further suggest that the C1 and C2 regions of EXP1 play complementary roles in the protein's function, the former being responsible for protein targeting to the PVM, and the latter being involved in the uptake of host apolipoproteins (Figure 4).

| Cell culture
The Huh7 human hepatoma cell line was cultured in DMEM supplemented with 10% (v/v) FCS and 1% (v/v) antibiotic-antimycotic FIGURE 4 Schematic representation of EXP1's localization and function during the liver stage of infection by Plasmodium berghei. The C2 region of EXP1 interacts with host Apoliprotein H (ApoH), enabling its internalisation and allowing for successful parasite replication that culminates on egress from the liver and subsequent infection of red blood cells. In the absence of the C2 terminal region of EXP1 (PbEXP1ΔC2 parasites; Sa et al., 2017), EXP1 still localizes to the PVM but no longer interacts with and internalises ApoH, resulting in impaired parasite replication but not preventing the appearance of the blood stage of infection. In the absence of the entire CT domain (PbEXP1ΔCT parasites), EXP1 no longer localizes to the PVM, resulting in the developmental arrest of the parasite and its inability to proceed to the blood stage of infection. CT, Cterminus, including the C1 and C2 regions; HPM, hepatocyte plasma membrane; NT, N-terminus; PVM, parasitophorous vacuole membrane; TMD, transmembrane domain.

| Mosquito rearing and parasite production
Anopheles stephensi mosquitoes were routinely reared at 28°C and 80% humidity and fed on a 10% (w/v) sucrose solution with 0.2 μg/ml of para-aminobenzoic acid. Mice were infected by intraperitoneal injection of PbWT or mutant parasites (cryopreserved iRBCs).
Adult mosquitoes were starved and allowed to feed on infected mice and were subsequently maintained at 21°C and 80% humidity for up to 21 days. PbANKA (Hall et al., 2005)
A similar strategy was used to generate the PbEXP1ΔNT parasite line. Briefly, a 5′ fragment without the last 114 bp coding NT domain of EXP1 protein after signal peptide (SP) was amplified using primers 5′UTR SP PbEXP1for and 5′UTR SP PbEXP1rev (Table S1) together with full transmembrane (TM) and CT domains, using primers TM CT PbEXP1for and TM CT PbEXP1rev (Table S1)

from
PbWT gDNA, and cloned into the b3D + vector containing the 3′ UTR fragment, as described above. Before transfection, the targeting vectors were linearised by using restriction enzymes KpnI and NotI.
To obtain clonal parasite populations, limited serial parasite dilutions were performed, and one parasite was administered by intravenous injection to each of 10 recipient naive NMRI mice (Thathy & Menard, 2002). After gDNA extraction, PbWT, PbEXP1ΔCT, Pb GFP EXP1ΔCT, and PbEXP1ΔNT parasites were genotyped by using specific primers (Table S1). The truncations or mutations of each parasite line were additionally confirmed by sequencing.

| Western blot
For Western blot analyses, blood samples were collected from PbWT-and PbEXP1ΔCT-infected NMRI mice. Collected blood was diluted in PBS and centrifuged at 1,500 rpm for 8 min at room temperature (RT). The supernatant was removed, and the remaining erythrocyte pellet was lysed in 0.2% (v/v) saponin in PBS, followed by centrifugation at 2,800 rpm for 8 min at RT. The supernatant was discarded, and the pelleted parasites were resuspended in PBS and centrifuged at 7,000 rpm for 2 min at RT. Following removal of the supernatant, parasites were resuspended in RIPA buffer and incubated at −20°C overnight. Protein lysates were then mixed with Laemmli sample buffer 1:1, heat denatured at 95°C for 5 min, separated by SDS-PAGE on BisTris 4-12% gradient gels, and transferred to methanol-activated PVDF membranes. Membranes were blocked in 5% (w/v) of powdered milk in PBS containing 0.05% (v/v) Tween 20 (PBST) for 1 h at RT. Chicken anti-PbEXP1 FL (1:700; kindly provided by Prof. Heussler, University of Bern, Switzerland), mouse anti-PbHSP70 (1:200;Tsuji, Mattei, Nussenzweig, Eichinger, & Zavala, 1994), and rat anti-PbEXP1 C-terminus (1:100; kindly pro-

| Assessment of Plasmodium sporogonic stages
For assessment of Plasmodium sporogonic stages, female A. stephensi mosquitoes were infected with PbWT and PbEXP1ΔCT parasites as described above. Mosquito MGs were dissected in PBS 10 to 12 days after mosquito feeding. Next, MGs were incubated with 1% (v/v) of NP40 solution in PBS for 20 min at RT, followed by staining with 1% (v/v) of mercurochrome solution in PBS for 30 min. MGs were then washed and transferred onto glass slides, covered with glass coverslips, and analysed by light microscopy with green filter. PbWT or PbEXP1ΔCT spz or with 1 × 10 3 or 1 × 10 6 iRBCs.

| Assessment of in vivo Plasmodium infection
Parasitaemia was monitored daily by microscopic analysis of Giemsastained blood smears. Mice were also monitored daily for signs of ECM and were euthanised in case of severe disease.

| Assessment of gene expression by qRT-PCR
Collected livers were transferred to RLT buffer supplemented with β-mercaptoethanol and homogenised using TissueRuptor from QIAGEN. RNA was extracted from liver homogenates by using the RNeasy kit from QIAGEN, according to the manufacturer's instructions. cDNA was synthesized by reverse transcription using the firststrand cDNA synthesis kit from Thermo Fisher Scientific and employing the following thermocycling parameters: 25°C for 5 min, 37°C for 60 min, and 70°C for 5 min. Liver P. berghei load was quantified by qRT-PCR, employing primers specific for Pb 18S rRNA (Table   S1). Gene expression levels were normalised to the endogenous mouse housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (Table S1)

| Analysis of UIS4-EXP1 colocalization and ApoH accumulation
Confocal microscopy images were obtained by sequential scanning of each channel in order to eliminate chromophore crosstalk. Prior to analysis, the background was corrected by applying a threshold value to all channels. UIS4-EXP1 colocalizing pixels were identified using the "colocalization highlighter" command in ImageJ software and were represented as a percentage of the total EXP1 signal. ApoH accumulation was quantified using the ImageJ software, as previously described (Sa et al., 2017).

| Quantification of in vitro merosome formation
For the in vitro assessment of merosome formation, Huh7 cells (2.5 × 10 4 ) were seeded in eight-well Lab-Tek chamber slides, infected with 1 × 10 4 spz, and incubated at 37°C with 5% CO 2 . At 72 hpi, supernatants were collected and centrifuged at 1,200 rpm for 5 min at RT.
Merosomes were then carefully resuspended and counted using a hemocytometer.

| Statistical analyses
Statistical significance was evaluated by employing a two-tailed unpaired t test. Values of P < .05 were deemed statistically significant.
All statistical analyses were performed by using GraphPad Prism 7 software.