Control of Parasitophorous Vacuole Expansion by LYST/Beige Restricts the Intracellular Growth of Leishmania amazonensis

The intracellular protozoan Leishmania replicates in parasitophorous vacuoles (PV) that share many features with late endosomes/lysosomes. L. amazonensis PVs expand markedly during infections, but the impact of PV size on parasite intracellular survival is still unknown. Here we show that host cells infected with L. amazonensis upregulate transcription of LYST/Beige, which was previously shown to regulate lysosome size. Mutations in LYST/Beige caused further PV expansion and enhanced L. amazonensis replication. In contrast, LYST/Beige overexpression led to small PVs that did not sustain parasite growth. Treatment of LYST/Beige over-expressing cells with vacuolin-1 reversed this phenotype, expanding PVs and promoting parasite growth. The opposite was seen with E-64d, which reduced PV size in LYST-Beige mutant cells and inhibited L. amazonensis replication. Enlarged PVs appear to protect parasites from oxidative damage, since inhibition of nitric oxide synthase had no effect on L. amazonensis viability within large PVs, but enhanced their growth within LYST/Beige-induced small PVs. Thus, the upregulation of LYST/Beige in infected cells functions as a host innate response to limit parasite growth, by reducing PV volume and inhibiting intracellular survival.


Introduction
Infections with the trypanosomatid protozoan Leishmania cause a broad spectrum of human diseases throughout the world. Depending on the parasite species, and on the genetic and immunological composition of the host, the clinical form can range from self-healing cutaneous lesions to severe visceralizing disease. The parasites enter mammalian hosts through the bite of sandflies, and replicate intracellularly as amastigotes. Although macrophages are considered the major host cell type for Leishmania, fibroblasts also harbor parasites and are thought to play an important role during latent infections [1]). In both macrophages and fibroblasts, intracellular amastigotes replicate within parasitophorous vacuoles (PV) that share several properties with late endosomes/lysosomes, including low luminal pH and the presence of lysosome-specific membrane proteins and acidic hydrolases [2,3].
An important question in the study of Leishmania pathogenesis is how parasites persist indefinitely in the host, even after the development of immunity to reinfection [4]. Their exclusively intracellular life style suggests that amastigotes possess mechanisms to avoid killing by the abundant microbicidal products produced by activated host cells. The mechanisms of in vivo persistence are of particular interest in relation to L. amazonensis, a species that causes cutaneous leishmaniasis in the new world. Several lines of evidence indicate that L. amazonensis is particularly adept at surviving intracellular killing mechanisms, when compared to other Leishmania species [5][6][7]. Interestingly, the morphology of the PVs harboring L. amazonensis amastigotes (and other species from the L. mexicana complex) also differs dramatically from PVs containing other Leishmania species, such as L. major and L. donovani. Amastigotes of L. mexicana and L. amazonensis replicate within very large, communal PVs that continuously undergo fusion with lysosomes and phagolysosomes. In contrast, PVs containing L. major and L. donovani amastigotes partition as the parasites replicate, resulting in small compartments containing only one parasite per vacuole [2,8]. It was suggested that PV expansion might protect L. amazonensis from host killing mechanisms, by diluting microbicidal molecules [9]. Here we directly investigated this hypothesis, by examining the expression pattern and role in infection of LYST/Beige, a gene known to regulate lysosome size in mammalian cells. Our results show that L. amazonensis infections upregulate LYST/Beige transcription, resulting in the control of PV expansion and inhibition of intracellular growth.

Results/Discussion
LYST/Beige mRNA transcription is upregulated in macrophages infected with Leishmania Human mutations in LYST (also known as lysosomal trafficking regulator) are responsible for the Chediak-Higashi syndrome (CHS), an autosomal recessive disease characterized by severe immune deficiency, partial albinism and recurrent bacterial infections. Cells from CHS patients and their mouse counterparts, beige, have abnormally enlarged lysosomes and lysosome-related organelles [10][11][12]. LYST/Beige overexpression reduces the size of lysosomes, suggesting that this large cytosolic protein (,430 kDa) is involved in regulating the size of these organelles [13]. Although the LYST/Beige mechanism of action is still unknown, yeast two-hybrid screens [14], expression of truncated constructs [15], and functional analysis of homologs containing similar BEACH domains [16,17] suggest that it may provide an anchoring scaffold for kinases and other molecules controlling membrane fusion/fission reactions. To investigate a possible role of LYST/Beige in the regulation of L. amazonensis PVs, we initially focused our investigations on the expression levels of this gene in infected macrophages.
Oligonucleotide DNA microarray analysis demonstrated that LYST/Beige transcription was increased in C57BL/6 mouse bone marrow macrophages (BMM) infected with L. amazonensis axenic amastigotes for 48 h (results not shown). These findings were confirmed using real time PCR (qPCR). Infection of BMM with L. amazonensis amastigotes induced a gradual enhancement in LYST/ Beige messenger RNA transcription, reaching a ,3 fold increase 72 h after infection ( Figure 1A). As previously described for L. amazonensis [2], amastigotes replicating intracellularly during this period were visualized attached to the Lamp1-positive limiting membrane of large communal PVs ( Figure 1B and 1C).
The Beige mutation causes expansion of Leishmaniacontaining parasitophorous vacuoles and enhances parasite growth To directly examine the role of LYST/Beige in controlling L. amazonensis PV expansion, we measured the size of parasitecontaining vacuoles in BMM from wild type or beige mice (bg J / bg J ). In wild type BMM infected with L. amazonensis, PVs expanded rapidly between 1 and 48 h after infection, and more slowly between 48 and 72 h ( Figure 1B, wild type). In contrast, PV expansion in infected bg J /bg J BMM appeared to be enhanced, with several significantly larger PVs observed at all time points ( Figure 1B-1D, bg J /bg J ). The PV size distribution in bg J /bg J BMM, however, was heterogeneous ( Figure 1B), with markedly enlarged PVs observed side-by-side smaller ones (arrows, Figure 1C). This pattern was consistent with the overall morphology of the lysosomal compartment in bg J /bg J BMM, which is also heterogeneous in size (results not shown), probably due to the highly dynamic nature of the tubular lysosomes of macrophages [18]. When we analyzed the impact of LYST/Beige deficiency on L. amazonensis intracellular survival and growth in bg J /bg J BMM, we found a small but statistically significant increase in the number of intracellular parasites at 48 and 72 h after infection ( Figure 1E). Although the size heterogeneity of PVs in BMM did not allow a definitive conclusion, these data suggested that PV expansion might favor L. amazonensis survival and replication within host cells.

LYST/Beige overexpression reduces parasitophorous vacuole size and inhibits parasite growth
Fibroblasts have been implicated as important host cells during latent infections with Leishmania [1]. Consistent with this finding, primary and immortalized fibroblast cell lines are susceptible to L. amazonensis infection [3,19]. Examining in parallel primary BMM and murine embryonic fibroblasts (MEF) infected with L. amazonensis axenic amastigotes, we found that both cell types sustain PV expansion and parasite intracellular proliferation (Figure 2A and 2D). We also observed that in fibroblasts PVs expand faster, and are significantly more homogeneous in size ( Figure 2C and 2D, compare to Figure 2A and 2B). This finding allowed us to examine in detail the impact of LYST/Beige expression on L. amazonensis intracellular growth.
Fibroblasts derived from beige mice complemented (YAC-bg J / bg J ) or not (bg J /bg J ) with a wild copy of LYST/Beige [20] were infected with L. amazonensis axenic amastigotes. The size of L. amazonensis-containing PVs in the YAC-bg J /bg J fibroblasts was markedly reduced, when compared to the PVs found in LYST/ Beige-deficient bg J /bg J cells ( Figure 3A and 3B). Importantly, the overexpression of LYST/Beige, which is known to reduce lysosome size in this cell line [13], caused a reduction in PV size that was significantly below the size normally seen in L. amazonensis-infected wild type fibroblasts (compare Figure 3B with Figure 2D). This reduction in PV size had a strong impact on parasite growth, since no intracellular replication was detected in the YAC-bg J /bg J cells overexpressing LYST/Beige, in several independent experiments ( Figure 3C). A cell line derived in parallel from wild type mice also supported L. amazonensis intracellular replication, but the parasite population grew slower than in the mutant bg J /bg J cells ( Figure 3C). Collectively, these results support a role of LYST/ Beige in restricting the intracellular survival of L. amazonensis.
We next examined whether the bg J /bg J fibroblasts that sustain L. amazonensis growth also responded to infection by upregulating LYST/Beige, as seen in macrophages ( Figure 1A). This determination was possible because the mutation causing LYST/Beige functional inactivation in bg J /bg J fibroblasts would lead to a truncated protein, lacking approximately 1400 amino acids [20]. A single transcript corresponding to the truncated form of LYST/ Beige was detected in bg J /bg J fibroblasts ( Figure 3D, left panel, not infected), and infection with L. amazonensis amastigotes markedly enhanced this signal ( Figure 3D, left panel, infected, and Figure 3D, right panel). Thus, L. amazonensis infection triggers in host cells an innate response that upregulates LYST/Beige expression, a process that blocks parasite replication if the expressed protein is functional.

Author Summary
The protozoan parasite Leishmania causes serious infections in humans throughout the world. After being inoculated into the skin through the bite of infected sandflies, the parasites enter host cells and replicate. The lysosome-like intracellular vacuoles where Leishmania amazonensis replicates expand dramatically as the infection progresses. Here we studied the impact of vacuole expansion on the ability of the parasites to survive and replicate inside host cells. We found that the host cell responds to infection with Leishmania amazonensis by upregulating expression of LYST/Beige, a gene that regulates the size of lysosomes and of parasite-containing vacuoles. The parasites replicated more efficiently in the large vacuoles formed in cells that have a mutation in LYST/Beige, whereas the same cells overexpressing functional LYST/Beige generated small vacuoles that were not able to sustain parasite growth. Drug treatments that reduced or enhanced the size of parasite-containing vacuoles had a corresponding effect on intracellular replication, demonstrating that large vacuoles provide a growth advantage to Leishmania amazonensis. Our results indicate that host cells respond to Leishmania infections by producing a protein capable of reducing vacuole size, as a strategy to inhibit parasite growth.
The intracellular survival of Leishmania is regulated by parasitophorous vacuole size The data discussed above suggested that PV expansion facilitates the intracellular survival and growth of Leishmania. To further investigate this possibility, we examined the effects of reducing PV size in bg J /bg J fibroblasts and, conversely, increasing the size of PVs in YAC-bg J /bg J fibroblasts. Previous studies showed that treatment of bg J /bg J cells with the thiol proteinase inhibitor E-64d reduces the size of the enlarged lysosomes normally observed in these cells [21,22]. A reduction in PV size after pre-treatment with E-64d was evident in bg J /bg J fibroblasts ( Figure 4A and 4B), and this coincided with a reduction of ,3 fold in the number of intracellular amastigotes detected after 24 h, when compared to untreated bg J / bg J cells ( Figure 4C, bg J /bg J ). In contrast, there was no apparent effect of E-64d on the smaller PVs within the YAC-bg J /bg J fibroblasts ( Figure 4B, YAC-bg J /bg J ), and no effect on parasite numbers in these cells ( Figure 4C, YAC-bg J /bg J ).
E-64d was washed out prior to infection, to prevent an effect of this irreversible cysteine protease inhibitor on the parasites. However, we cannot completely rule out the possibility that a small pool of the inhibitor remained inside host cells. For this reason, we examined the effect of a drug that has the opposite effect of E-64d, enlarging lysosomes. YAC-bg J /bg J fibroblasts were pre-treated with vacuolin-1, a small molecule that causes rapid expansion of late endosomes/ lysosomes in mammalian cells [23]. We infected bg J /bg J and YACbg J /bg J fibroblasts pre-treated with vacuolin-1 with L. amazonensis, and determined PV size and the number of intracellular parasites after a 24 h period. Vacuolin-1 pre-treatment of bg J /bg J fibroblasts did not further increase the mean PV size, nor did it cause an enhancement in the number of intracellular amastigotes ( Figure 4D and 4F, bg J /bg J ). In contrast, in the LYST/Beige overexpressing YAC-bg J /bg J fibroblasts, vacuolin-1 caused a significant increase in PV size ( Figure 4D-4E, YAC-bg J /bg J ) and a ,3 fold enhancement in parasite numbers ( Figure 4F, YAC-bg J /bg J ).
In these experiments, fibroblasts were pre-treated with E-64d or vacuolin-1 before infection to avoid any adverse effects on the parasites. For this reason, parasite development was not followed beyond 24 h. Since Leishmania has an intracellular doubling time of approximately 12 h, the strong impact of PV size on the number of parasites detected after only 24 h suggests that the mechanism involves direct modulation of early intracellular survival.
PV enlargement may protect Leishmania from nitric oxide (NO)-mediated killing A major mechanism underlying the leishmanicidal activity of murine macrophages is the IFN-gamma and TNF-alpha-mediated activation of nitric oxide synthase (iNOS or NOS2), which generates the potent microbicidal agent nitric oxide (NO) [24,25]. Recent evidence indicates that iNOS is recruited to the membrane of recently formed pathogen-containing phagosomes, in a process thought to be critical for its microbicidal action [26,27]. The strong reactivity of NO implies that it has to be generated in close proximity to targets, in order to be effective. Interestingly, L. amazonensis is significantly less susceptible to IFNgamma-mediated killing, when compared to species that reside in small PVs such as L. major [5,6]. Thus, PV enlargement may have a role in reducing the concentration of NO in direct contact with intracellular amastigotes.
Since L. amazonensis is capable of productively infecting both macrophages and fibroblasts in culture (Figure 2), and PV enlargement enhances the early survival and replication of L. amazonensis in these two cell types (Figures 3 and 4), it is conceivable that PV enlargement protects Leishmania from a microbicidal mechanism common to both macrophages and fibroblasts. Although NO production in response to IFN-gamma and TNF-alpha is much more robust in macrophages, fibroblasts are also capable of mounting this response [1]. To investigate whether NO production plays a role in the enhanced survival of Leishmania in fibroblasts, we quantified intracellular amastigotes in L. amazonensis-infected bg J /bg J and YAC-bg J /bg J fibroblasts pre-treated with the iNOS inhibitors L-NMMA and L-NAME. These inhibitors were previously shown to promote the intracellular growth of Leishmania in macrophages by blocking NO production [25,28]. Quantification of the number of intracellular amastigotes after 24 h showed that these inhibitors did not alter the parasite population size in bgJ/bgJ fibroblasts, which exhibit large PVs ( Figure 5A). However, amastigote survival and replication was significantly enhanced in YAC-bg J /bg J fibroblasts pre-treated with the iNOS inhibitors ( Figure 5B), consistent with the observation of several parasites inside each of the small PVs found in these cells ( Figure 5D). Importantly, the size of the small PVs induced by L. amazonensis in the YAC-bg J /bg J fibroblasts overexpressing LYST/Beige remained unchanged after treatment with the iNOS inhibitors ( Figure 5C), suggesting that inhibition of NO production can promote L. amazonensis survival/growth in the absence of PV expansion. Future studies are required to investigate the possibility that LYST expression might also impact the trafficking of iNOS to L. amazonensis-containing PVs.
This study demonstrates that transcription of LYST/Beige, a gene previously implicated in regulating lysosome size, is upregulated in macrophages and fibroblasts infected by L. amazonensis. Examining the impact of PV expansion on the fate of the parasites, we found that PV size is an important determinant of intracellular survival. Collectively, our findings provide direct experimental evidence for the idea that increased expression of LYST/Beige functions as a host innate response to restrict Leishmania growth, by counteracting PV expansion. Our findings suggest that a search for drugs capable of specifically blocking L. amazonensis PV expansion might lead to novel therapeutic strategies against the human infections caused by this pathogen.
Bone marrow-derived macrophages (BMM) were isolated from the femurs of 8-10 week-old C57BL/6 wild type mice or beige mice (in the C57BL/6 background) and cultured for 7 days in Fibroblasts were infected for 30 min, and PV size was determined microscopically after the indicated time points. The values represent the mean of 50 independent measurements. Asterisks indicate significant differences from the corresponding time points in bg J /bg J fibroblasts: * P,0.05, ** P,0.01 (Student t test). (C) L. amazonensis amastigotes proliferate slowly in wild type and more vigorously in bg J /bg J fibroblasts lacking a functional LYST/Beige protein, but not in YACbg J /bg J fibroblasts that over-express LYST/Beige. Fibroblasts were exposed to L. amazonensis amastigotes for 60 min, washed, and incubated for the indicated periods followed by determination of the number of intracellular parasites. The data corresponds to the mean +/2 SD of triplicates. The results shown are representative of several independent experiments. (D) Upregulation of the mutant LYST/Beige transcripts in bg J /bg J fibroblasts infected with L. amazonensis. The cells were exposed for 30 min to axenic amastigotes, washed, and incubated for the indicated periods, followed by RNA extraction and analysis by qualitative RT-PCR (left) and quantitative real-time qPCR (right). The data on the right panel represents the mean +/2 SD of triplicate determinations. Asterisks indicate significant differences from noninfected samples (NI): * P,0.05 (Student t test). doi:10.1371/journal.ppat.1000179.g003 RPMI 1640 containing 30% L-fibroblast culture supernatant and 20% FBS. Mouse murine embryonic fibroblasts (MEFs) were prepared from day 13.5 mouse embryos [29] and cultured in high glucose DMEM (GIBCO BRL) with 10% FBS, 1% penicillin/ streptomycin, and 2 mM glutamine. Fibroblast lines from wild type C57BL/6J mice and from beige-J mice lacking the Beige gene (bg J /bg J ) (in the C57BL/6J background), and the bg J /bg J line complemented with a yeast artificial chromosome (YAC) carrying the Beige gene (YAC-bg J /bg J ) (NCBI AAL40134 accession number) were generated in J.K.'s laboratory at the University of Utah [20] and maintained in DMEM 10% FBS, 1% penicillin/ streptomycin and 2 mM glutamine (1 mg/ml G418 was kept in the growing medium of the YAC-bg J /bg J line, but removed prior to infection with Leishmania, followed by several washes). (B) Analysis of PV size in infected bg J /bg J or YAC-bg J /bg J fibroblasts pretreated or not with E64d. Fibroblasts were exposed to L. amazonensis amastigotes for 30 min, washed, and further incubated for 24 h followed by microscopic determination of PV size. The values above the symbols represent the mean of 50 independent measurements. Asterisks indicate significant differences from the non-treated controls (NT): ** P,0.01 (Student t test). (C) PV reduction with E64d inhibits the intracellular growth of L. amazonensis in bg J /bg J fibroblasts. After E64d pretreatment the fibroblasts were exposed to amastigotes for 1 h, washed, and further incubated for 24 h, followed by determination of the number of intracellular parasites. The data corresponds to the mean +/2 SD of triplicates. Asterisks indicate significant differences from the non-treated controls (NT): * P,0.05 (Student t test). (D) Analysis of PV size in infected bg J /bg J or YAC-bg J /bg J fibroblasts pretreated or not with 1 mM vacuolin-1. Fibroblasts were exposed to L. amazonensis amastigotes for 30 min, washed, and further incubated for 24 h followed by microscopic determination of PV size. The values above the symbols represent the mean of 50 independent measurements. Asterisks indicate significant differences from non-treated controls (NT): ** P,0.01 (Student t test). (E) Phase contrast images showing that pretreatment with 1 mM vacuolin-1 enlarges the small vacuoles observed in L. amazonensisinfected YAC-bg J /bg J fibroblasts. Arrows point to individual PVs. Bars = 10 mm. (F) After vacuolin-1 pretreatment the fibroblasts were infected for 1 h, washed, and further incubated for 24 h, followed by determination of the number of intracellular parasites. The data corresponds to the mean +/2 SD of triplicates. Asterisks indicate significant differences from the non-treated controls (NT): ** P,0.01 (Student t test). doi:10.1371/journal.ppat.1000179.g004 Affimetrix GeneChip analysis and quantitative real-time PCR C57BL/6 BMM plated at 2.5610 5 in 10 cm dishes were exposed or not to L. amazonensis amastigotes at MOI:10 for 30 min at 34uC. After infection cells were washed 3 times in PBS and further incubated for the appropriate times. Total RNA was isolated using Trizol (Invitrogen). The Affymetrix protocol, used to analyse RNA extracted from BMM infected for 48 h with L. amazonensis amastigotes, was essentially as described (GeneChip Expression Analysis Technical Manual, Affymetrix). cRNA was hybridized for 16 h to Affymetrix Genechip Mouse Genome 430 2.0 Array (Affymetrix), which contain 45,000 probe sets for the analysis of 39,000 transcripts and variants from over 34,000 mouse genes. Normalization was performed using GeneChip robust multi-array analysis, followed by GC-robust multi-array average (GC-RMA) normalization. Identification of significantly perturbed genes was done using significance analysis of microarrays. The false positive rate was 0.1%. Real-time PCR was performed using a BioRad iQ icycler Detection system (BioRad Laboratories, Ltd) with SYBR green fluorophore (BioRad Laboratories, Ltd) according to the manufacturer's instructions. Oligonucleotide primers 59-AGCAGAAGGTGATAGACCAGAA and 59-CCC- Figure 5. iNOS inhibition rescues Leishmania replication within small parasitophorous vacuoles. (A) Treatment with iNOS inhibitors does not affect L. amazonensis growth in the enlarged PVs of bg J /bg J fibroblasts, but rescues growth in the small PVs of YAC-bg J /bg J fibroblasts. After pretreatment with 50 mM L-NMMA or 50 mM L-NAME, the fibroblasts were infected for 1 h, washed and incubated for 24 h, followed by determination of the number of intracellular parasites. The data corresponds to the mean +/2 SD of triplicates. Asterisks indicate significant differences from non-treated controls (NT): * P,0.05 (Student t test). (B) iNOS inhibitors do not alter PV size in bg J /bg J or YAC-bg J /bg J fibroblasts. Cells were exposed to L. amazonensis amastigotes for 30 min, washed, and further incubated for 24 h followed by microscopic determination of PV size. The values represent the mean of 50 independent measurements. (C) Phase contrast images of bg J /bg J or YAC-bg J /bg J fibroblasts pretreated or not with L-NMMA or L-NAME, exposed to L. amazonensis amastigotes for 1 h and fixed after 24 h. Arrows point to individual PVs. Bars = 5 mm. doi:10.1371/journal.ppat.1000179.g005 ACACTTGGATCATCAATGC were used to amplify a 119 base pair portion of LYST or 59-TCAGTCAACGGGGGACATAAA and 59-GGGGCTGTACTGCTTAACCAG to amplify a 142 base pair of the control cDNA HPRT1. The reaction was incubated for 3 min at 95uC, and then for 45 cycles of 20 s each at 95uC and for 1 min at 55uC. Fluorescence was detected at each annealing step and cycle threshold (C t ) was calculated by determining the point at which the fluorescence exceeded a threshold limit. All reactions were run in triplicate and negative controls (no template cDNA) were included in each experiment. Data were normalized by the level of GAPDH expression in individual samples.

Cell treatments
BMM were plated at 1.25610 5 and fibroblasts at 2.5610 4 on 12-mm diameter glass coverslips in 24-well dishes, 24 h prior to experiments. Vacuolin-1 (compound 5114069, ChemBridge San Diego, CA) (23), was added to fibroblasts at 1.0 mM for 1 h prior to infection. E-64d (Sigma) was added to fibroblasts at 1.0 mM for 1 h prior to infection. Immediately before infection the cells were washed 3 times in PBS. N-methyl-L-arginine (L-NMMA) or Nnitro-L-arginine methyl ester (L-NAME) (Sigma), two nitric oxide synthase inhibitors [26,30], were added to cells at 50 mM for 1 h. Cells were washed in PBS prior to infection.

Leishmania infection assays
For invasion of BMM, purified axenic amastigotes were added at a multiplicity of infection (MOI):1 in RPMI 10% FBS for 1 h at 34uC. For invasion of fibroblasts, amastigotes were added at an MOI = 10 for 1 h in high glucose DMEM supplemented with 10% FBS at 34uC. After invasion, cells were washed 3 times in PBS and incubated for indicated times at the appropriate temperature. Coverslips were then fixed in 2% paraformaldehyde, and host cell and parasite DNA were stained with 10 mg/ml DAPI for 5 min, after permeabilization with 0.05% Triton-X 100 for 10 min. At least 300 host cells, in triplicate, were analyzed for each time point. Images were acquired through a 1006 objective using a Zeiss Axiovert microscope equipped with a Hamamatsu Orca II cooled CCD camera controlled by Metamorph Software (Molecular Devices). For parasite growth curves in fibroblasts the data was expressed as intracellular parasites per microscopic field, to compensate for host cell division [16]. For Lamp-1 immunofluorescence, cells on coverslips were permeabilized with PBS/ 2% BSA/ 0.5% saponin for 20 min, incubated with anti-Lamp1 mAb 1D4B (Developmental Studies Hybridoma Bank) 1:50 in PBS/2% BSA/0.5% saponin for 45 min, followed by Alexa 488goat anti-rat secondary antibodies (Invitrogen). Host and parasite DNA were stained with DAPI. Individual parasitophorous vacuoles (PVs) were measured using the Metamorph draw function. A line was drawn on phase contrast images straight across the length of the PV, and the Metamorph calculate function was used to convert the values to mm. A total of 50 PVs were measured for each sample.