ABSTRACT
After transmission by Anopheles mosquitoes, Plasmodium sporozoites travel to the liver, infect hepatocytes, and rapidly develop as intrahepatocytic liver stages (LS). Rodent models of malaria exhibit large differences in the magnitude of liver infection, both between parasite species and between strains of mice. This has been mainly attributed to differences in innate immune responses and parasite infectivity. Here, we report that BALB/cByJ mice are more susceptible to Plasmodium yoelii preerythrocytic infection than BALB/cJ mice. This difference occurs at the level of early hepatocyte infection, but expression levels of reported host factors that are involved in infection do not correlate with susceptibility. Interestingly, BALB/cByJ hepatocytes are more frequently polyploid; thus, their susceptibility converges on the previously observed preference of sporozoites to infect polyploid hepatocytes. Gene expression analysis demonstrates hepatocyte-specific differences in mRNA abundance for numerous genes between BALB/cByJ and BALB/cJ mice, some of which encode hepatocyte surface molecules. These data suggest that a yet-unknown receptor for sporozoite infection, present at elevated levels on BALB/cByJ hepatocytes and also polyploid hepatocytes, might facilitate Plasmodium liver infection.
INTRODUCTION
Malaria is caused by eukaryotic parasites of the genus Plasmodium. They have coevolved with their mammalian hosts, significantly shaping both parasite and host genomes. It is the most deadly parasitic infection in the world, with approximately 300 million clinical episodes annually, mainly in the developing world. Initiation of preerythrocytic infection commences after parasite transmission by the bite of an infected mosquito, when the sporozoite travels to the liver and invades hepatocytes. During invasion, the parasite surrounds itself with a parasitophorous vacuole membrane (PVM) that protects the intracellular niche and permits liver stage (LS) development. Parasites leave the liver as first-generation red blood cell infectious merozoites and initiate the symptomatic erythrocytic stages of infection by cyclic intraerythrocytic replication and concomitant destruction of erythrocytes (1).
Host factors that mediate malaria parasite erythrocytic infection (2–4) have been uncovered in part by examining genetics of susceptible and nonsusceptible populations (5). For example, African populations are largely protected against Plasmodium vivax blood stage infection and also lack Duffy blood group determinants. This has led to the finding that Duffy blood group determinants are critical to allow P. vivax merozoite invasion of erythrocytes (6). However, human populations resistant to preerythrocytic Plasmodium infection have not been identified. Therefore, mouse strains differing in susceptibility to preerythrocytic infection might provide a useful model for the identification of host factors that are important for hepatocyte infection.
Differences in susceptibility to rodent malaria preerythrocytic infection have been documented for mice. Plasmodium berghei sporozoites are much more infectious to C57BL/6 mice than they are to BALB/cJ mice (7). Yet BALB/c mice are 2,000 times more susceptible to Plasmodium yoelii sporozoite infection than P. berghei infection (8). While it is difficult to ascertain whether Plasmodium species which infect humans exhibit distinct infectivity profiles in vivo, hepatocytes from different human donors differ in susceptibility to preerythrocytic stage infection in vitro (9). Unfortunately, mechanistic insights into the causes for these differences are lacking.
Candidate-based approaches have revealed several host cell receptors that facilitate sporozoite infection of hepatocytes, such as the tetraspanin CD81 (10), scavenger receptor B1 (SR-BI) (11, 12), and heparin sulfate proteoglycans (HSPGs) (13, 14). However, none of these receptors are entirely essential for infection in vivo, arguing that an unbiased investigation aimed at identifying novel hepatocyte receptors required for sporozoite infection is needed. To begin this process, we took advantage of genetically similar mice with pronounced differences in susceptibility to preerythrocytic P. yoelii infection.
BALB/c mice are highly susceptible to P. yoelii preerythrocytic infection (8), but in some cases, hundreds of generations of mice have been bred separately, leading to founder effects within substrains. These breeding arrangements have led to the establishment of BALB/cJ and BALB/cByJ substrains (15). Here we show that these two substrains exhibit differential susceptibility to preerythrocytic P. yoelii infection and that this is determined by differences in early hepatocyte infection. None of the described receptors for sporozoite infection are, however, differentially expressed in BALB/cJ and BALB/cByJ hepatocytes. Our data suggest that additional hepatocyte receptors are required for the effective sporozoite infection of hepatocytes.
MATERIALS AND METHODS
Mosquito rearing and sporozoite production.For P. yoelii sporozoite production, female 6- to 8-week-old Swiss Webster (SW) mice (Harlan, Indianapolis, IN) were injected with blood-stage P. yoelii (17XNL) parasites to begin the growth cycle. Animal handling was conducted according to Institutional Animal Care and Use Committee-approved protocols. Infected mice were used to feed female Anopheles stephensi mosquitoes after gametocyte exflagellation was observed. Salivary gland sporozoites were isolated according to standard procedures at day 14 or 15 post-blood meal. For each experiment, salivary glands were isolated in parallel in order to ensure that sporozoites were extracted under identical conditions.
Quantification of liver burden by real-time RT-PCR.Total RNA was extracted using TRIzol reagent (Invitrogen). cDNA synthesis was performed using the Super Script III Platinum two-step quantitative reverse transcription-PCR (qRT-PCR) kit according to the manufacturer's instructions (Invitrogen). All PCR amplification cycles were performed at 95°C for 30 s for DNA denaturation and 60°C for 4 min for primer annealing and DNA strand extension. Parasite 18S was amplified using primers with sequences 5′-GGGGATTGGTTTTGACGTTTTTGCG-3′ and 5′-AAGCATTAAATAAAGCGAATACATCCTTAT-3′. Mouse glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was amplified using sequences 5′-CCTCAACTACATGGTTTACAT-3′ and 5′-GCTCCTGGAAGATGGTGATG-3′. For quantitative PCR (qPCR), a standard curve was generated using 1:4 dilutions of a reference cDNA sample for PCR amplification of all target PCR products. Experimental samples were compared to this standard curve to give a relative abundance of transcript.
Real-time in vivo imaging of LS development in whole bodies of live mice.BALB/cJ or BALB/cByJ mice were infected with Py-GFP-luc salivary gland sporozoites by intravenous (i.v.) tail vein injection. For i.v. injections, salivary gland sporozoites were enumerated and suspended in RPMI medium prior to injection of 50,000 sporozoites per mouse. Mice were injected with 100 μl of RediJect d-luciferin (PerkinElmer) intraperitoneally prior to being anesthetized using the isoflurane anesthesia system (XGI-8; Caliper Life Sciences, USA). Luciferase activity in animals was visualized through imaging of whole bodies using the IVIS Lumina II animal imager (Caliper Life Sciences). Animals were kept anesthetized during the measurements, which were performed within 5 to 10 min after the injection of d-luciferin. Bioluminescence imaging was acquired with a 10-cm field-of-view (FOV), medium binning factor, and an exposure time of 1 to 5 min. Quantitative analysis of bioluminescence was performed by measuring the luminescence signal intensity using the region-of-interest (ROI) settings of the Living Image 3.0 software. ROIs were placed around the abdominal area at the location of the liver. ROI measurements are expressed as total flux (photons/s).
Hepatocyte isolation.Hepatocytes were isolated from mice using a two-step perfusion in situ. Portal veins of anesthetized mice were cannulated with warmed perfusion buffer (1× Hanks balanced salt solution [HBSS] without calcium, magnesium supplemented with 8 mM HEPES buffer and 0.5 mM EDTA) at low flow rate. Once cannulated, the inferior vena cava was severed to blanch the liver. The buffer flow rate was increased to 5 ml/min and allowed to perfuse for 5 min, with occasional clamping and releasing of the vena cava to inflate the liver. The liver was then perfused with 2% collagenase II (Worthington) in collagenase buffer (HBSS supplemented with 8 mM HEPES and 0.5 mM CaCl2) for 5 min at 5 ml/min. The liver was then removed from the abdominal cavity and the gallbladder excised. The liver was placed in a dish of Dulbecco modified Eagle medium (DMEM) and gently pushed through a 100-μm cell strainer to dissociate the hepatocytes, which were then collected with a wide-bore syringe.
Hepatocyte processing.Hepatocytes processed for flow cytometry were spun by benchtop centrifuge at 50 × g for 3 min to pellet hepatocytes. The supernatant, containing nonparenchymal cells and lymphocytes, was aspirated, and the remaining hepatocytes were washed in DMEM. The spins and washes were repeated until the supernatant was clear, usually 3 washes. Cells were then fixed and stained as described below. Hepatocytes processed for plating or RNA extraction were spun and washed once as described above and then resuspended in DMEM at a volume of 10.4 ml. To this was added 10 ml of 90% Percoll (GE Health Sciences) in phosphate-buffered saline (PBS), followed by spinning at 50 × g for 10 min with no brake. Pelleted cells were given a final wash with DMEM, then resuspended in warm InvitroGro HI hepatocyte medium (Bioreclamation IVT), and checked for density and viability using trypan blue staining.
Immunofluorescence microscopy.A total of 1.5 × 105 primary hepatocytes were seeded as described above in each well of a type I collagen-coated glass slide (BioCoat; BD Biosciences). Cells were infected with 5 × 104 P. yoelii sporozoites. Slides were centrifuged for 3 min at 515 × g in a hanging-bucket centrifuge to aid in sporozoite invasion. After 90 min, medium containing extracellular sporozoites was aspirated, and fresh medium was added. LSs developed for 48 h, at which time cells were fixed with 4% paraformaldehyde, blocked, and permeabilized for 1 h in PBS with the addition of 0.1% Triton X-100 and 2% bovine serum albumin (BSA). Staining steps were performed in PBS supplemented with 2% BSA. Cells were stained using antisera to Plasmodium heat shock protein 70 (HSP70), upregulated in infectious sporozoites 4 protein (UIS4) 4°C overnight, and then visualized with the use of Alexa Fluor-488 goat anti-mouse and Alexa Fluor-594 goat anti-rabbit secondary antibodies (Invitrogen). Cells were stained with 4′,6-diamidino-2-phenylindole (DAPI) to visualize both hepatocyte and parasite nuclei. Sporozoites that had not invaded and/or developed in hepatoma cells were distinguished by UIS4 noncircumferential staining and morphology.
Quantification of LSs by flow cytometry.Cells were cultured as described above. A total of 1.5 × 105 cells were plated in each well of a 24-well plate and infected with 105 P. yoelii sporozoites per well. At the desired time point, cells were trypsinized and fixed with Perm/Fix buffer (BD Biosciences). Cells were blocked in Perm/Wash buffer (BD Biosciences) supplemented with 2% BSA. Additional staining steps were also performed in Perm/Wash buffer supplemented with 2% BSA. Cells were stained with monoclonal antibody to circumsporozoite protein (CSP) conjugated to Alexa Fluor-488 or Alexa Fluor-647 at 25°C. Parasitized hepatocytes were identified by flow cytometry performed on a BD LSRII. Flow cytometric analysis was performed using FlowJo software (TreeStar). All experimental conditions were tested in biological triplicate.
Analysis of DNA content.Mouse hepatocytes were isolated and processed for flow cytometry as described above. After blocking, cells were resuspended in a mixture containing PBS, 5 mM EDTA, 0.1 mg/ml of RNase A, and SYTOX blue DNA dye (Invitrogen). DNA content was measured by flow cytometry on an LSRII and analyzed by FlowJo (Tree Star).
In vivo and in vitro analysis of CD74 expression.For analysis of CD74 expression by ploidy, mouse hepatocytes were isolated and processed for flow cytometry as described above. Cells were fixed in 1% paraformaldehyde and then washed and blocked in PBS plus 2% BSA. Cells were stained with anti-mouse CD74 conjugated to fluorescein isothiocyanate (FITC; BD Pharmingen) at a concentration of 5 μg/ml overnight at 4°C and then stained for DNA analysis as described above. Cells were gated on ploidy state by DNA dye fluorescence, and then the median fluorescence of FITC was measured for each ploidy state. To analyze CD74 expression in infection, Hepa1-6 cells were infected as described above for quantification of LSs by flow cytometry. After blocking, cells were stained with anti-P. yoelii CSP conjugated to Alexa Fluor-647 as well as anti-mouse CD74 conjugated to FITC. Flow cytometric analysis was performed using FlowJo software (TreeStar). All experimental conditions were tested in biological triplicate.
Animal handling and infection.All animal protocols were approved by the Seattle Biomedical Research Institute IACUC.
Transcriptome sequencing (RNA-Seq) analysis.We used clipped fastq files provided by Expression Analysis. These reads were subjected to quality control (QC) based on the following criteria as provided by Expression Analysis: (i) average Q score across all bases of >25, (ii) removal of any single base with a quality of less than 7 (if enough bases are removed that the sequence is less than 25 bases in length, the read is removed), (iii) removal of homopolymers, (iv) removal of Illumina adapters (if the cleaned sequence is less than 25 bases in total length, the whole read is removed), and (v) the presence of no more than 4 N′s in any given sequence. These reads were then mapped to the reference genome using Tophat2; then differential gene expression was calculated using Cuffdiff2. Genes with a Benjamini-Hochberg corrected P value of ≤ 0.05 and absolute log2 ratio of ≥0.95 were considered significantly differentially expressed between the BALB/cByJ and BALB/cJ mice.
Validation of differentially transcribed genes by qPCR.RNA was extracted from whole BALB/cJ and BALB/cByJ mouse livers using TRIzol reagent (Invitrogen) and a Direct-Zol RNA extraction kit (Zymo Research). cDNA was generated using the QuantiTect reverse transcription kit (Qiagen) and then used as the template for quantitative PCR. Primers for genes tested are listed in Table S1 in the supplemental material. For qPCR, a standard curve was generated using 1:4 dilutions of a reference cDNA sample for PCR amplification of all target PCR products. Experimental samples were compared to this standard curve to give a relative abundance of transcript. Abundances of genes of interest were standardized to the GAPDH abundance from the same sample. The relative measurements from five mice were averaged for each strain, and the ratio of BALB/cJ to BALB/cByJ mice was calculated for each gene tested.
RESULTS
The study of preerythrocytic Plasmodium infection has been facilitated by the use of suitable mouse models, perhaps the most susceptible of which is the BALB/c mouse for the study of P. yoelii (16). We asked if susceptibilities are similar across BALB/c-derived substrains. BALB/cJ or BALB/cByJ mice were infected with 106 P. yoelii sporozoites and sacrificed 44 h after infection. LS burden was assessed by histological analysis. We observed that LS density in infected BALB/cByJ mouse livers was approximately five times higher than in BALB/cJ mouse livers (Fig. 1A and B). To test if frequencies of infection and/or survival or the LS growth rates were different between BALB/cJ and BALB/cByJ mice, we measured the cell size of LS in sections of infected BALB/cJ and BALB/cByJ mice. We did not observe a significant difference in LS size, indicating that it is frequency of liver infection and/or survival of the parasite in the liver that causes the observed differences in LS burden (Fig. 1C).
BALB/c mouse substrains show increased susceptibility to P. yoelii preerythrocytic infection. (A) Five mice were infected with 1 M P. yoelii sporozoites by intravenous injection. Livers were excised at 44 h postinfection. LS parasites were visualized by hematoxylin and eosin (H&E) staining. (B) Density of LSs in tissue quantified by microscopic counting. Data are representative of those from at least three independent experiments. (C) LS size does not vary by strain. Three BALB/cByJ or BALB/cJ mice were infected with 1 M P. yoelii sporozoites and stained with H&E as illustrated in panel A. The size of 10 LS parasites was quantified in each sample. (D) LS burden in mice infected with 100,000 P. yoelii sporozoites. Parasites expressed GFP-luciferase fusion protein, and burden was quantified by light output using an in vivo imagining system (IVIS) 24 h after infection.
We further analyzed at what point during preerythrocytic infection the difference in susceptibilities arose. We infected BALB/cJ or BALB/cByJ mice with 5 × 104 P. yoelii parasites that express a green fluorescent protein (GFP)-luciferase fusion protein under the control of the EF1α promoter (17) and monitored LS burden over the course of liver infection by bioluminescent imaging (Fig. 1D). The difference in LS burden remained constant between BALB/cByJ and BALB/cJ mice at 16, 24, and 43 h postinfection (Fig. 2A). These data suggest that the increased susceptibility of BALB/cByJ mice is established at a time point between infection and 16 h after infection. Furthermore, the consistent difference in LS burden throughout schizogony (approximately 16 h after infection through the completion of LS development) suggests that once infection is established, mice do not differ in their capacity to support further LS development.
Differences in susceptibilities of BALB/c mice to sporozoite infection are largely explained by differences in hepatocyte infection. (A) BALB/cJ or BALB/cByJ mice (n = 10 for each substrain) were injected with 100,000 P. yoelii GFP-Luc sporozoites. Total LS burden was quantified by IVIS 16, 24, or 43 h after infection. (B) Primary hepatocytes were isolated from BALB/cByJ or BALB/cJ mice. Plated hepatocytes from three separate animals were infected with 50,000 P. yoelii sporozoites in culture. Ninety minutes postinfection, hepatocytes already displayed distinct differences in susceptibility to infection, which was maintained throughout in vitro LS development. (C and D) LS parasites 48 h after infection do not appear morphologically different in their in vitro development between infected BALB/cJ (C) and BALB/cByJ (D) mice. All data are representative of those from three independent experiments.
Genetic variation of the host could lead to a variant phenotype in multiple cell types, which might impact preerythrocytic infection. To evaluate if host hepatocytes are responsible for the difference in susceptibility to infection in vivo, we isolated primary hepatocytes from BALB/cByJ and BALB/cJ mice via collagenase-mediated perfusion and infected them in vitro with P. yoelii sporozoites. We then assessed infection levels using flow cytometry (18). Since hepatocytes are isolated from nonparenchymal cells by Percoll gradient, this experiment strictly assessed hepatocyte susceptibility to infection without the potential impact of other cell types. We found that BALB/cByJ mouse hepatocytes were dramatically more susceptible to P. yoelii preerythrocytic infection as early as 90 min after sporozoite infection (Fig. 2B). This suggests that differences in hepatocytes are the primary cause of the differential susceptibility to preerythrocytic infection observed in vivo. Neither rates of LS development nor their size varied between BALB/cJ and BALB/cByJ mouse primary hepatocyte cultures (Fig. 2C and D). This supports the notion that the difference in in vivo infection arises from differential susceptibility of hepatocytes to initial sporozoite infection.
Both CD81 (10) and scavenger receptor B1 (SR-BI) (11, 12) have been implicated in facilitating sporozoite infection of hepatocytes. Furthermore, p53 levels are suppressed in infected hepatocytes, and boosting p53 levels reduced preerythrocytic infection (19). We asked if differences in BALB/cByJ and BALB/cJ mouse hepatocyte susceptibility to infection would be explained by differences in the basal levels of one of these factors. Expression of CD81 on hepatocytes derived from BALB/cJ or BALB/cByJ mice showed no significant difference (Fig. 3A). Furthermore, no significant differences in SR-BI and p53 levels of mRNA (Fig. 3B and C) or protein (Fig. 3D and E) was observed in BALB/cJ and BALB/cByJ mouse hepatocytes. Therefore, differential expression of previously identified host hepatocyte factors did not account for the distinct BALB/cJ and BALB/cByJ mouse susceptibilities to preerythrocytic infection.
Differences in hepatocyte susceptibility cannot be explained by known infection-related host factors. (A) CD81 surface expression levels are similar in BALB/cJ and BALB/cByJ mouse hepatocytes, as assessed by surface staining of CD81 followed by flow cytometry. (B and C) Transcript levels of scavenger receptor B1 (B) and p53 (C) do not vary between hepatocytes of BALB/cJ and BALB/cByJ mice. (D and E) Protein levels of SR-BI and p53 do not vary in hepatocytes between BALB/cJ and BALB/cByJ mice. Transcript levels were determined by qPCR and compared to the transcript of the given gene and normalized to mouse GAPDH. Protein levels were assessed by immunoblotting, and quantities were normalized to mouse β-actin. Data are representative of those from three independent experiments.
Recently, an association between high hepatocyte ploidy and increased susceptibility to preerythrocytic infection was uncovered (20), yet neither CD81 nor SR-BI was upregulated on polyploid hepatocytes (L. S. Austin, A. Kaushansky, and S. H. I. Kappe, unpublished data). Thus, we asked if there was a difference in the frequency of hepatocyte polyploidy between BALB/cJ and BALB/cByJ mice. We collagenase-perfused livers and then stained hepatocytes with a DNA dye, treated them with RNase, and assessed DNA content by flow cytometry. BALB/cByJ mouse hepatocytes exhibited a greater frequency of the polyploid 8n and 16n states than did BALB/cJ mouse hepatocytes (Fig. 4). This suggests that a hepatocyte factor upregulated on both polyploid hepatocytes and BALB/cByJ mouse hepatocytes might associate with increased preerythrocytic infection.
BALB/cByJ mice have an increased frequency of polyploid hepatocytes. (A) Histograms monitor DNA content in hepatocytes in BALB/cJ and BALB/cByJ mouse hepatocytes. (B) Pie charts depict the distribution of hepatocyte ploidy in BALB/cJ and BALB/cByJ mouse hepatocytes. BALB/cByJ mouse hepatocytes are more enriched for 8n and 16n cells than are BALB/cJ mouse hepatocytes.
To uncover novel host factors that associate with increased susceptibility to preerythrocytic infection, we next conducted an unbiased interrogation of differences in target organ gene expression between BALB/cJ and BALB/cByJ mice. We harvested hepatocytes and total livers from three mice of each mouse substrain, isolated RNA from each sample, pooled RNA for each sample type, and performed RNA-Seq. Sequencing reads were mapped by Tophat2 (21), and differential RNA abundance was determined by Cuffdiff2 (22). We identified 129 genes differentially expressed between BALB/cByJ and BALB/cJ mouse livers and 53 genes differentially expressed in hepatocytes only (corrected P value ≤ 0.05; log2 ratio ≥ 0.95) (Fig. 5A; see also Table S2 in the supplemental material). Eighteen of these transcripts were detected as differentially expressed in the total liver samples and purified hepatocyte samples.
RNA-Seq analysis reveals differential gene expression in livers and hepatocytes of BALB/c substrains. (A) Sequencing reads were mapped by Tophat2 (21), and differential gene regulation was determined by Cuffdiff2 (22). This identified 53 genes differentially expressed between BALB/cByJ and BALB/cJ mouse hepatocytes (corrected P value ≤ 0.05; log2 ratio ≥ 0.95). Genes also differentially regulated in total liver (18) were assessed by qPCR. A comparison of transcript levels by RNA-Seq and qPCR is shown. For the complete differential regulated RNA-Seq data set, see Table S2 in the supplemental material. (B) Hepatocytes from naive BALB/c mice were isolated and stained for DNA content and CD74. Cells with higher DNA content also exhibited higher levels of CD74. (C) A total of 300,000 Hepa 1-6 cells were infected with 100,000 P. yoelii sporozoites. Infection rates (1.5 h after infection) within the highest and lowest 25% of CD74-expressing cells (designated CD74high and CD74low, respectively) are illustrated. Infection rate in the total culture is depicted with a dashed line.
We prioritized quantitative RT-PCR (qRT-PCR) validation of genes that showed differences in mRNA abundance in both total liver and isolated hepatocytes of substrains. Minor differences in transcript abundance between BALB/cJ and BALB/cByJ hepatocytes might be masked by nonparenchymal cell transcripts, but the most substantial changes are likely to be detected in both hepatocyte and total liver samples. A majority of transcripts showed differences in mRNA abundance between BALB/cJ and BALB/cByJ livers by qPCR that were consistent with the changes observed via RNA-Seq (Fig. 5A). Taken together, these data provide a compelling set of gene products which warrant examination, as host factors that might be associated with susceptibility to preerythrocytic infection. Interestingly, transcript expression for five cell surface receptors showed differences between BALB/cJ and BALB/cByJ hepatocytes. Levels of epidermal growth factor receptor (EGFR) transcript were elevated in BALB/cByJ hepatocytes and total liver. In addition, neurotrophic tyrosine kinase receptor 2 (NTRK2) transcripts were increased in BALB/cJ mouse hepatocytes, and CD74 transcripts were increased in BALB/cByJ mouse hepatocytes. Because it was exclusively upregulated in susceptible BALB/cByJ mouse hepatocytes, we chose to perform subsequent studies on CD74. We first assessed the levels of CD74 protein in hepatocytes, which varied in DNA content. We found that polyploid hepatocytes had substantially higher levels of CD74 than those with lower DNA content (Fig. 5B). We reasoned that this might make CD74 a prime candidate to facilitate increased infection in both BALB/cByJ hepatocytes and hepatocytes with high ploidy. To validate this finding, we infected Hepa1-6 cells with P. yoelii sporozoites and assessed the infection rate by flow cytometry. We found that the frequency of sporozoite infection in the top 25% of CD74-expressing cells (CD74high) was significantly higher than the infection frequency in cells with the lowest 25% of CD74 levels (CD74low) (Fig. 5C). These data suggest that sporozoites preferentially infect hepatocytes with increased levels of CD74. Thus, CD74 is an intriguing hepatocyte receptor worthy of additional investigation for its potential role in Plasmodium sporozoite infection.
DISCUSSION
The bite of an infected Anopheles mosquito transmits tens to hundreds of Plasmodium sporozoites (16), infecting the liver at extraordinarily low multiplicities of infection. Unlike the related apicomplexan Toxoplasma, the sporozoite of Plasmodium species that infect mammals specifically targets hepatocytes as a host cell. The motile sporozoite has the capacity to traverse multiple cell types (23, 24), including hepatocytes (24), before committing to intracellular residence. This allows the parasite to select its optimal host cell at the point of invasion. From the time of sporozoite transmission to the point of parasite egress from hepatocytes into the bloodstream, a diversity of host defenses can curtail preerythrocytic infection. These factors include but are not limited to (i) inhibition of sporozoite motility in the skin by immune responses (25, 26), (ii) intracellular trapping of sporozoites during traversal of the sinusoidal endothelium (23) or Kupffer cells (27, 28), and (iii) innate immune responses which curtail infection during LS development (29, 30). While mouse models that differ in their innate and adaptive immune responses to Plasmodium are plentiful, we report the first example of genetically similar hepatocytes derived from two substrains of BALB/c mice with dramatically different susceptibilities to initial sporozoite infection.
The hepatocyte-intrinsic differential susceptibility to sporozoite infection between BALB/cJ and BALB/cByJ mice is unexpected since differences in susceptibility between these BALB/c substrains to diseases such as hyperthyroidism and encephalomyelitis are mediated by the immune system (31, 32). Furthermore, differences in susceptibility to preerythrocytic infection between BALB/c and C57BL/6 mice have been mapped to the expression of TREM2 in macrophages (33, 34). These data indicate that genetic differences that give rise to molecular changes in immune components can alter susceptibility to preerythrocytic infection. However, these studies do not implicate specific hepatocyte factors. Interestingly, it has been shown that P. berghei infection is curtailed more potently by innate immune responses during liver stage development than P. yoelii infection. This suggests that when comparing liver stage development between different rodent malarias and different mouse strains, one must take into account not only the initial susceptibility of the host hepatocyte to sporozoite infection but also the innate immune response that is engendered by the infection (35).
Candidate-based approaches have provided insights into a subset of hepatocyte receptors which are important for sporozoite infection. These investigations demonstrate partial overlap with host entry requirements of hepatitis C virus (HCV). Specifically, Plasmodium species and HCV both utilize CD81 (10, 36) and SR-BI (11, 37). Furthermore, heparan sulfate proteoglycan (HSPGs) (14) play a role in switching between the migratory state and the invasive state of sporozoites (13). Interestingly, HSPGs also play a role in HCV infection (36, 38). Yet the dramatic difference in susceptibility to sporozoite infection between BALB/cJ and BALB/cByJ hepatocytes cannot be explained by either CD81 or SR-BI expression, suggesting that other hepatocyte factors involved in sporozoite infection remain to be discovered.
Plasmodium sporozoites that infect mammals (with the exception of P. berghei) are highly specific for their hepatocyte host cell in vivo, yet no single described host factor is completely essential for infection. Small interfering RNA screens have identified a small number of hepatocyte molecules which impact sporozoite invasion and/or development (39) in vitro, but inefficient knockdown in hepatocytes has limited the utility of this approach. Genetically similar mice with variable susceptibilities to preerythrocytic infection might thus provide a useful tool for linking specific genetic changes to differences in host susceptibility. Our data suggest that the same factor(s) might explain the differential susceptibility to sporozoite infection between substrains of BALB/c mice and hepatocytes with low and high ploidy (20).
Several genes that are differentially expressed between BALB/cJ and BALB/cByJ mice might play a role in facilitating or inhibiting sporozoite invasion of hepatocytes. The receptor tyrosine kinase (EGFR) is more highly expressed in BALB/cByJ mouse hepatocytes, making it an enticing candidate receptor for sporozoite entry. EGFR is also involved in HCV infection, where its signaling properties facilitate the clustering of the critical entry molecules CD81 and occludin (40). Since CD81 is also important for P. yoelii and P. falciparum invasion, it is possible that EGFR functions in a similar way for sporozoite infection. Alternatively, a direct interaction between the sporozoite and EGFR might facilitate hepatocyte entry, and the expression of proteins containing the epidermal growth factor (EGF) domain in sporozoites (41) might provide the potential for such an interaction.
The increase of CD74 expression in BALB/cByJ hepatocytes, as well as the preference of P. yoelii sporozoites for CD74high hepatocytes, presents a tantalizing clue that warrants further study. It has also been shown that P. falciparum and P. berghei macrophage inhibitory factor (42) bind host CD74. Additionally, P. yoelii MIF is important for LS development in BALB/c mice (43). Taken together, these data suggest that an interaction between CD74 and Plasmodium MIF might facilitate effective preerythrocytic infection. The relationship between parasite MIF and host CD74 as well as their role in sporozoite invasion of hepatocytes is an important future area of investigation. In contrast to EGFR and CD74, the receptor tyrosine kinase NTRK2 is expressed at higher levels in less susceptible BALB/cJ hepatocytes, suggesting that this receptor might play an inhibitory role in sporozoite invasion. Unlike the well-described EGF-like domains, to our knowledge, no Plasmodium protein is similar to the NTRK2 ligand neurotrophin-3.
The impact of host genetics on hepatocyte susceptibility to infection is not restricted to rodent Plasmodium parasites. Hepatocyte donor origin also impacts P. falciparum preerythrocytic infection in vitro (9). Although host factors which facilitate sporozoite hepatocyte entry for Plasmodium spp. vary (10, 44), factors that mediate differential susceptibility to rodent malaria parasites could still be prime targets to study as factors for susceptibility of humans to malaria parasite infection. The dramatic difference in susceptibility between BALB/cJ and BALB/cByJ hepatocytes to sporozoite infection might assist in identifying novel host receptors, which facilitate the initiation of Plasmodium preerythrocytic infection. This could guide novel approaches to prevent human malaria parasite infection.
ACKNOWLEDGMENTS
We are grateful to William W. Betz and Jen C. C. Hume for mosquito and sporozoite production. We thank Heather S. Kain and Marian Abdullahi for excellent technical assistance. We thank the Seattle Biomedical Research Institute vivarium staff for their work with mice.
A.K. is the recipient of an NRSA Ruth L. Kirschstein National Research Service Award (F32 AI091129). L.S.A. is the recipient of an NDSEG award from the Department of Defense and NIH training grant T32 AI007509-13. This work has also been partially funded by grant 1R01GM101183-01A1 to S.H.I.K.
FOOTNOTES
- Received 19 June 2014.
- Returned for modification 1 August 2014.
- Accepted 3 October 2014.
- Accepted manuscript posted online 13 October 2014.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.02230-14.
- Copyright © 2015, American Society for Microbiology. All Rights Reserved.
