Plasmodium vivax Liver Stage Development and Hypnozoite Persistence in Human Liver-Chimeric Mice

Resource

Plasmodium vivax Liver Stage Development and

Graphical Abstract

Highlights

d The FRG KO huHep mouse model supports Plasmodium

vivax liver stage development

d The mouse model supports hypnozoite formation and

reactivation

d P. vivax liver stage to blood stage transition occurs in FRGKO

huHep mice

d Testing of drugs for P. vivax radical cure can be carried out in

the mouse model

Mikolajczak et al., 2015, Cell Host & Microbe 17, 526–535April 8, 2015 ª2015 Elsevier Inc.http://dx.doi.org/10.1016/j.chom.2015.02.011

Authors

Sebastian A. Mikolajczak,

Ashley M. Vaughan, ...,

Jetsumon Sattabongkot,

Stefan H.I. Kappe

[email protected] (S.A.M.),[email protected](S.H.I.K.)

In Brief

Plasmodium vivax malaria has been

difficult to study in the absence of animal

models. Mikolajczak et al. show that the

human liver-chimeric FRG KO huHep

mouse supports complete development

of P. vivax liver stages, including

hypnozoite formation and persistence

in vivo.

Cell Host & Microbe

Resource

Plasmodium vivax Liver StageDevelopment and HypnozoitePersistence in Human Liver-Chimeric MiceSebastian A. Mikolajczak,1,7,* Ashley M. Vaughan,1,7 Niwat Kangwanrangsan,2 Wanlapa Roobsoong,3

Matthew Fishbaugher,1 Narathatai Yimamnuaychok,3 Nastaran Rezakhani,1 Viswanathan Lakshmanan,1 Naresh Singh,4

Alexis Kaushansky,1 Nelly Camargo,1 Michael Baldwin,1 Scott E. Lindner,6 John H. Adams,4 Jetsumon Sattabongkot,3

and Stefan H.I. Kappe1,5,*1Seattle Biomedical Research Institute, 307 Westlake Avenue North, Suite 500, Seattle, WA 98109, USA2Faculty of Science3Mahidol Vivax Research Center, Faculty of Tropical Medicine

Mahidol University, Bangkok 10400, Thailand4Department of Global Health, College of Public Health, University of South Florida, Tampa, FL 33620, USA5Department of Global Health, University of Washington, Seattle, WA 98195, USA6Department of Biochemistry and Molecular Biology, Center for Malaria Research, Pennsylvania State University, University Park,

PA 16802, USA7Co-first author

*Correspondence: [email protected] (S.A.M.), [email protected] (S.H.I.K.)

http://dx.doi.org/10.1016/j.chom.2015.02.011

SUMMARY

Plasmodium vivax malaria is characterized by peri-odic relapses of symptomatic blood stage parasiteinfections likely initiated by activation of dormantliver stage parasites—hypnozoites. The lack of trac-table P. vivax animal models constitutes an obstaclein examining P. vivax liver stage infection anddrug efficacy. To overcome this obstacle, we haveused human liver-chimeric (huHep) FRG KO miceas a model for P. vivax infection. FRG KO huHepmice support P. vivax sporozoite infection, liverstage development, and hypnozoite formation. Weshow complete P. vivax liver stage development,including maturation into infectious exo-erythrocyticmerozoites as well as the formation and persistenceof hypnozoites. Prophylaxis or treatment with theantimalarial primaquine can prevent and eliminateliver stage infection, respectively. Thus, P. vivax-infected FRG KO huHep mice are a model to investi-gate liver stage development and dormancy andmayfacilitate the discovery of drugs targeting relapsingmalaria.

INTRODUCTION

The majority of human malaria is caused by infections with two

Plasmodium parasite species: Plasmodium falciparum and

Plasmodium vivax. Research efforts are predominantly focused

on P. falciparum malaria because of the high mortality the dis-

ease causes in sub-Saharan Africa. However, P. vivax malaria

affects more people in a wider geographical range (95 countries)

and puts 2.85 billion people at risk of disease every year (Guerra

526 Cell Host & Microbe 17, 526–535, April 8, 2015 ª2015 Elsevier In

et al., 2010). Furthermore, recent studies indicate that P. vivax in-

fections are more pathogenic than previously appreciated (Price

et al., 2009). Two major attributes contribute to P. vivax’s unique

epidemiology: first, its ability to develop in mosquitoes at lower

temperatures; and second, the existence of dormant liver stages

termed hypnozoites that can be activated weeks, months, or

even years after the primary mosquito-transmitted infection.

Activated hypnozoites are thought to complete liver stage devel-

opment, leading to a relapse of symptomatic blood stage infec-

tion (White, 2011). Thus, it is of great importance to develop

experimental animal models that allow for the study of the bio-

logical features associated with the unique epidemiology of

this parasite.

Unfortunately, studies of the complex P. vivax liver stage

biology are encumbered by the parasite’s strong preference

for human and nonhuman primate tissue. The initial studies

describing P. vivax liver stages were performed on either human

liver biopsies of a patient undergoing experimental malaria fever

therapy for neurosyphilis (Shortt et al., 1948), or the liver biopsies

of chimpanzees infected by intravenous inoculation of a large

numbers P. vivax sporozoites (Krotoski et al., 1982b; Rodhain,

1956). It was the latter study (Krotoski et al., 1982b) that demon-

strated the existence of small, nonreplicating forms—hypno-

zoites (Markus, 2011)—in the P. vivax-infected liver. Since these

studies were undertaken, P. vivax liver stage research has been

sparse and mostly limited to in vitro studies in primary hepato-

cytes (Mazier et al., 1984) or hepatoma cell lines (Hollingdale

et al., 1985; Sattabongkot et al., 2006). Overall, little additional

knowledge has been gained to date that has yielded a better un-

derstanding of the biology of P. vivax hypnozoites and their role

in malaria relapse. These shortcomings negatively impact the

development of new antimalarial drugs, and as a result prima-

quine, an 8-aminoquinoline, is still the only licensed drug that

eliminates hypnozoites and offers causal prophylaxis and radical

cure treatment for P. vivax infection (Fernando et al., 2011). Addi-

tionally, P. vivax blood stages only replicate in reticulocytes, and

c.

Figure 1. Development of P. vivax Liver Stages in the FRG KO huHep

Mouse

(A) Liver stage parasites in infected liver sections 3, 5, and 7 days after

sporozoite infection were visualizedwith differential interference contrast (DIC)

imaging (left panel), with a mAb specific for P. vivax circumsporozoite protein

(CS; VK247) (central panel), and DAPI for DNA content (right panel). To visu-

alize the DNA content in the small 3-day-old parasite, the area inside the

dotted box is enhanced in the top right corner.

(B) Liver stages at 3, 5, and 7 days after sporozoite infection were analyzed

with monoclonal mouse antibodies to P. vivax Upregulated in Infectious

Sporozoites protein 4 (UIS4, localizes to the PVM), acyl carrier protein (ACP,

localizes the apicoplast), heat shock protein 60 (HSP60, localizes to the

mitochondria), and binding immunoglobulin protein (BiP, localizes to the ER).

Scale bar, 10 mm.

continuous in vitro blood stage culture remains extremely chal-

lenging. This further impedes studies of the parasite life cycle

(Carlton et al., 2011). Consequently, researchers have in the

past turned to the relapsing, nonhuman primate malaria parasite

Plasmodium cynomolgi to model the biology of hypnozoites

(Galinski et al., 2013). P. cynomolgi is genetically closely related

to P. vivax, and research on its liver stages led to the identifica-

tion of hypnozoites (Krotoski et al., 1982c; Shortt and Garnham,

1948). Recently, an improved in vitro culture system for

P. cynomolgi liver stages and hypnozoites was described (Dem-

bele et al., 2014). Further refinements of such systems will

certainly contribute to drive a better understanding of the biology

of hypnozoites.

In search of P. vivax in vivo liver stage models, we took advan-

tage of a mouse that supports engraftment and long-term

survival of human primary hepatocytes (Azuma et al., 2007).

The severely immunocompromised FRG KO mouse (with dele-

tions in fumarylacetoacetate hydrolase [FAH], recombination-

activating gene 2 [Rag2], and interleukin-2 receptor subunit

gamma [Il2rg] gene deletions) can be transplanted with human

hepatocytes (FRG KO huHep). We have recently shown that

this mouse model supports the complete development of

P. falciparum liver stages, culminating in transition to blood stage

infection (Vaughan et al., 2012). Here we use the FRG KO huHep

mouse to show complete P. vivax liver stage development as

well as the formation and persistence of hypnozoites in vivo.

RESULTS

Infection of FRG KO huHep Mice with Plasmodium vivax

SporozoitesFRG KO huHep mice were injected intravenously with 3.5–5 3

105 P. vivax sporozoites derived from mosquitoes that had

been infected with parasite isolates from Thailand. Mice were

sacrificed 3, 5, and 7 days postinfection. Infected liver tissue

was collected for histological evaluation and immunofluores-

cence assays (IFAs). The liver stages were initially localized

with a mouse monoclonal antibody (mAb) to the P. vivax circum-

sporozoite protein (CS, genotype VK247) (Rongnoparut et al.,

1995). Infections appeared robust, as indicated by the presence

of numerous liver stage parasites in individual liver sections (see

Figure S1A available online). CSwas expressed on the liver stage

parasite plasma membrane (PPM) and strongly confined to

the parasite periphery in a circumferential pattern at day 3 post-

infection (Figure 1A). It showed a more complex distribution at

days 5 and 7 postinfection (Figures 1A and S1B), presumably

due to the invagination of the PPM (Uni et al., 1985), which

precedes the formation of exo-erythrocytic merozoites. In

contrast, CS expression was not reported at late time points of

P. falciparum liver stage development in FRG KO huHep mice

(Vaughan et al., 2012), but it has previously been observed for

P. vivax liver stage development in vitro (Hollingdale et al.,

1985). To establish that P. vivax sporozoites infected and devel-

oped only in human hepatocytes within the mouse liver, we cos-

tained the infected liver sections with a FAH-specific polyclonal

antibody, which only identifies human hepatocytes (FRG KO

mouse hepatocytes lack FAH). Indeed, liver stages were exclu-

sively detected in human hepatocytes at all time points of devel-

opment (Figure S1C).

Cell

The Complex Cellular Organelle and MembraneDevelopment in P. vivax Liver Stages Is Revealedwith AntibodiesTo reveal the cellular features of P. vivax liver stages, we devel-

oped polyclonal rabbit and monoclonal mouse antibodies spe-

cific to P. vivax proteins or conserved Plasmodium proteins

that have a high amino acid sequence identity among orthologs

(Figures 1B and S2A). One prominent feature ofPlasmodium liver

stages is the establishment of the parasitophorous vacuole

membrane (PVM), which separates the parasite from the cyto-

plasm of the host hepatocyte. To visualize the P. vivax PVM,

Host & Microbe 17, 526–535, April 8, 2015 ª2015 Elsevier Inc. 527

Figure 2. Complete Maturation of P. vivax Liver Stages and Blood

Stage Transition in the FRG KO huHep Mouse

(A) Indirect immunofluorescence assay (IFA) of P. vivax liver stages at 10 days

after sporozoite infection using Pv merozoite surface protein 1 (MSP1), poly-

clonal rabbit antibody (green), PvUIS4 mouse monoclonal antibody (red), and

DAPI (blue, DNA).MSP1 expression reveals the presence of differentiated exo-

erythrocytic merozoites within the liver stage schizont. The magnification of

the area in the yellow box shows a MSP1-positive exo-erythrocytic merozoite

(arrow) outside the confines of the mature liver schizont. Scale bar, 10 mm.

(B) IFA of a 9-day-old P. vivax liver stage stained with PvMSP1 polyclonal

rabbit antibody (green) to visualize the merozoite surface, ACP monoclonal

mouse antibody (red) to visualize the apicoplast, and DAPI (DNA stain, blue).

The mature liver stage is in the process of releasing exo-erythrocytic mero-

zoites into the surrounding liver tissue. Scale bar, 20 mm.

(C) Human red blood cells enriched for reticulocyteswere injected into FRGKO

huHep mice 9 and 10 days post-sporozoite infection. Four hours after the

second injection, the blood was removed and microscopically analyzed by

Giemsa-stained thin blood smear. Black arrows point to P. vivax ring stage

parasites within reticulocytes. Scale bar, 10 mm.

we produced a mouse mAb against the P. vivax ortholog

(PvUIS4) of P. yoelii Upregulated in Infectious Sporozoite 4

(UIS4), which was previously shown to localize to the PVM of

rodent malaria parasite liver stages (Mueller et al., 2005). The

PvUIS4 mAb revealed a circumferential staining pattern that

ensconced the liver stages (Figure 1B) and was distinct from

CS (a PPMmarker) (Figure S2B), indicating that PvUIS4 localizes

to the PVM. Furthermore, previously unstudied organelles of

P. vivax liver stages were detected using mouse mAbs to (1)

the relict plastid (apicoplast)-targeted acyl carrier protein

(ACP); (2) the mitochondrial heat shock protein, HSP60; and (3)

the endoplasmic reticulum (ER)-targeted protein binding immu-

noglobulin protein (BiP) (Figure 1B). As liver stage schizogony

progressed, the organelle structures became highly complex,

consistent with their extensive replication. The PV lumen was

localized with a rabbit polyclonal antibody to falstatin (P. vivax

ortholog of P. yoelii falstatin; Pei et al., 2013), the parasite cyto-

plasm was visualized with rabbit polyclonal antibody to macro-

phage inhibitory factor (MIF) (P. vivax ortholog of P. yoelii MIF;

Miller et al., 2012), and the Golgi apparatus localized with anti-

body the ER lumen protein retaining receptor (ERD2) (Elmendorf

and Haldar, 1993) (Figure S2A).

Complete Maturation of P. vivax Liver Stages, InfectiousExo-erythrocytic Merozoite Release, and ReticulocyteInvasionThe duration of P. vivax liver stage development culminating in

the formation and release of infectious exo-erythrocytic merozo-

ites remains imprecisely defined. Analysis of P. vivax liver

stage schizogony in the FRG KO huHep mice at day 7 and day

8 post-sporozoite infection showed that liver stage maturation

was not yet complete. This contrasts with observations of

P. falciparum liver stage maturation in this mouse model where

release of exo-erythrocytic merozoites occurred between 6

and 7 days post-sporozoite infection (Vaughan et al., 2012). In

P. falciparum, the expression of MSP1 precedes exo-erythro-

cytic merozoite formation and clearly identifies the surface of in-

dividual exo-erythrocytic merozoites (Vaughan et al., 2012). We

thus analyzed P. vivax MSP1 expression in liver stage schizonts

and observed expression as early as day 5 post-sporozoite

inoculation, with increased expression at day 8 (Figure S2C).

However, individual merozoites were not observed. Continued

analysis of liver stage schizogony showed numerousmature liver

stage schizonts at day 9 that contained individual exo-erythro-

cytic merozoites (Figure 2A). Some mature schizonts had

released merozoites in tightly packed masses that were

observed in the surrounding tissue, beyond the confines of

what appeared to be a ruptured PVM (Figure S3A). The majority

of mature liver stage schizonts lost their characteristic spherical

structure and released a substantial fraction of their exo-erythro-

cytic merozoites into the surrounding tissue, including the liver

sinusoids (Figure 2B). Mature liver stage schizonts were also de-

tected at day 10 post-sporozoite infection, but they were fewer

when compared to day 9 (data not shown). The remaining liver

stages had clearly differentiated apicoplasts, suggesting that

they were close to maturation (Figure S3B). The decline in liver

stage schizont numbers at day 10 compared to day 9 was likely

the result of mature parasite egress from hepatocytes, leaving

fewer schizonts in the tissue. Therefore, we conclude that the

528 Cell Host & Microbe 17, 526–535, April 8, 2015 ª2015 Elsevier In

complete maturation of P. vivax liver stages and exo-erythrocytic

merozoite release occurs between day 9 and day 10 post-sporo-

zoite infection in the FRG KO huHep mouse model. To test

whether exo-erythrocytic merozoites released from the liver

can invade red blood cells to continue the parasite life cycle,

mice were intravenously injected with reticulocytes at days 9

and 10 after sporozoite inoculation (this experiment was per-

formed twice with two mice per group). Four hours after the sec-

ond reticulocyte injection, blood was removed from each mouse

by cardiac puncture and analyzed for parasites by Giemsa-

stained thin blood smears. Ring stage parasite-infected reticulo-

cytes were readily observed for each mouse tested (at least 20

separate thin blood smears were made from each mouse and

parasite-infected reticulocytes were seen in all smears; Fig-

ure 2C). Thus, P. vivax liver stagematuration and release of retic-

ulocyte-infectious exo-erythrocytic merozoites can be modeled

using FRG KO huHep mice.

Formation of P. vivax Hypnozoites in the FRG KO huHepMouse LiverDuring the analysis of liver stage schizont growth at days 5–8

post-sporozoite infection, we also detected a subset of liver

c.

Figure 3. Characterization of P. vivax Hyp-

nozoites in the FRG KO huHep Mouse and

Comparison to Sporozoites

(A) IFA of P. vivax sporozoites stained with mouse

monoclonal antibody to PvCS (green) and rabbit

polyclonal antibodies to PvUIS4, BiP, PvHSP60,

ACP, and Pv MIF (red).

(B) Hypnozoites at day 7 post-sporozoite infection

were localized with antibodies to PvUIS4 (rabbit

polyclonal and mouse monoclonal) (green),

monoclonal mouse antibody to PvCS (red), and

rabbit polyclonal antibodies to BiP, PvHSP60,

ACP, and PvMIF (red). The yellow arrow points to a

PvUIS4-postive PVMprominence that is notable in

the PVM of all hypnozoites. Scale bar, 10 mm.

stage parasites that were small at all time points and appeared

not to increase significantly in size (Figure 3). We concluded

that these were small, nonreplicating hypnozoite forms and for

the ease of terminology will refer to these small forms as hypno-

zoites. At the time of exo-erythrocytic merozoite release at

days 9 and 10 postinfection, hypnozoites continued to persist

(Figure S3A). To confirm that hypnozoites resided within human

hepatocytes and not to remaining mouse hepatocytes, which

could retard liver stage development, we used an antibody to

detect human FAH in the infected liver sections. Indeed, all

observed hypnozoites were located within human hepatocytes

(Figure S1C).

To determine if P. vivax hypnozoites display any unique organ-

ellar features that would differentiate them from sporozoites and

replicating liver stage schizonts, we performed IFAs on sporozo-

ites (Figure 3A) and hypnozoites at days 5, 7, and 8 post-sporo-

zoite infection (Figure 3B). Mouse mAbs and rabbit polyclonal

antibodies to PvUIS4, BiP, PvHSP60, ACP, and PvMIF were

used to characterize protein expression and localization. As ex-

pected, the ER marker BiP and the mitochondrial marker HSP60

were localized to the sporozoite interior (Figure 3A), as has been

described for sporozoites of other Plasmodium species. Inter-

estingly, PvUIS4 was not detected in P. vivax sporozoites, in

contrast to observations made for P. yoelii UIS4 (Kaiser et al.,

2004) and P. falciparum UIS4 (Mackellar et al., 2010) but in

agreement with P. berghei UIS4 (Silvie et al., 2014). Expression

of ACP and PvMIF was also not observed in P. vivax sporozoites

(Figure 3A). As opposed to sporozoites, UIS4 was readily de-

tected in hypnozoites in a circumferential pattern, indicative of

PVM localization (Figure 3B). Intriguingly, PvUIS4 staining always

revealed a polarized densely fluorescent prominence of the PVM

(Figure 3B, yellow arrows), which was never observed on the

PVM of liver stage schizonts at all time points examined (Figures

1B and S2A and data not shown). Moreover, the PVM promi-

nence was also not observed on liver stage trophozoites at

Cell Host & Microbe 17, 526–

day 2 post-sporozoite infection (two

mice infected with two independent

P. vivax isolates) (Figure S3C). Thus, the

UIS4-positive PVM prominence might

be a unique feature of P. vivax hypno-

zoites. The localization patterns for BiP,

HSP60, and ACP suggested that biogen-

esis of the ER and initial replication of mitochondria and apico-

plast had commenced in hypnozoites (Figure 3B). However,

the hypnozoite organelles were more limited in their replication

when compared to early liver stage schizonts at day 3 post-

sporozoite infection (Figure 1B). Interestingly, ACP and PvMIF

were expressed in hypnozoites, but not in sporozoites (Figure 3).

Furthermore, exported protein 1 (EXP-1), another PVM-resident

protein (Doolan et al., 1996; Vaughan et al., 2012), was not ex-

pressed in sporozoites, hypnozoites, or day 3 liver stage schiz-

onts but was clearly detected in liver stage schizonts starting

at day 5 post-sporozoite infection (Figure S4A). This expression

pattern contrasted with the PVM protein, PvUIS4, which was ex-

pressed by all P. vivax liver stages observed, including hypno-

zoites (Figure 3B).

Next, the extent of genome replication was analyzed in hypno-

zoites. We encountered difficulties in visualizing hypnozoite DNA

byDAPI due to the high fluorescent intensity of hepatocyte nuclei

in the infected liver sections and comparatively minute DNA con-

tent of hypnozoites. Therefore, to indirectly analyze the extent of

DNA replication in hypnozoites, a histone acetylation-specific

antibody (recognizing acetylated lysine 9 of histone H3,

H3K9Ac) was used for IFA (Figure 4). A single histone-positive

structure was observed in hypnozoites, suggesting that they

had not undergone significant DNA replication. This was in stark

contrast to replicating liver stage schizonts, which exhibited an

increasing number of histone-positive structures with progres-

sive growth over the course of infection, starting as early as

day 3 after sporozoite infection (Figure 4).

The Persistence of P. vivax Hypnozoites and HypnozoiteActivationTo investigate if hypnozoites persisted in the mouse model, we

analyzed livers at days 14 and 21 post-sporozoite infection. At

day 14, persistent hypnozoites were present (Figures S4B and

5B), and importantly, no liver stage schizonts were observed at

535, April 8, 2015 ª2015 Elsevier Inc. 529

Figure 4. P. vivax Liver Stage Genome Replication

Antibodies to acetylated lysine 9 of histone H3 (H3K9Ac) (green) were used to

analyze genome replication of liver stage parasites in infected FRG KO huHep

mice. Schizonts and hypnozoites were also stained with antibody to CS

(VK247) (red). DNA was visualized with DAPI (blue). A single H3K9Ac-positive

structure was detected within liver stage trophozoites 2 days after sporozoite

infection. Hypnozoites (shown here 5 days after sporozoite infection) also

contained a single H3K9Ac-positive structure. Multiple H3K9Ac-positive

structures were observed for replicating liver stage schizonts at days 3, 5, and

7 after sporozoite infection. Scale bar, 10 mm.

Figure 5. P. vivax Hypnozoite Persistence and Activation in FRG KO

huHep Mice

(A and B) Persistent hypnozoites at day 21 after sporozoite infection. Parasites

were visualized with antibodies to PvUIS4 (green) and antibodies to apico-

plast-localized ACP and acetylated lysine 9 of histone H3 (H3K9Ac) or ACP

(red). Host hepatocyte nuclei were visualized with DAPI. The UIS4-postive

PVM prominence is maintained and notable in the PVM of all persistent hyp-

nozoites. Hypnozoites contained a single H3K9Ac-positive structure.

(C) Size comparison of liver stage trophozoites, liver stage schizonts, and

hypnozoites at different time points of infection. The size (liver stage area at the

greatest circumference of the parasite) was calculated. Measurements were

taken for at least ten liver stages at each time point. The average size ± SEM

is shown on the dot plots. Note that hypnozoites show growth over time

but remain smaller than 3-day-old schizonts. No schizonts were detected

in infections at 14 days after sporozoite infection but were again detected at

day 21.

(D and E) Shown are examples of replicating liver stage schizonts at day 21

post-sporozoite infection, suggesting that they originated from hypnozoites

that activated and entered schizogony. Antibodies used for IFA were mouse

monoclonal antibody to PvUIS4 (green) rabbit polyclonal antibodies to ACP in

the left panel and H3K9Ac (red) in the right panel. DNA was stained with DAPI

(blue). Scale bar, 10 mm.

this time point (Figure 5C). Furthermore, no parasite-infected

human reticulocytes were recovered when the day 14 sporo-

zoite-inoculated mice were transfused with human reticulo-

cyte-enriched blood and analyzed as described above (data

not shown). We also observed multiple hypnozoites at day 21

postinfection (five mice evaluated; >20 hypnozoites observed

in 30 sections from each liver) (Figures 5A–5C). Day 21 hypno-

zoites were somewhat larger when compared to hypnozoites

detected during the first 10 days of infection (Figure 5C) but still

showed no evidence of DNA replication based on detection of a

single H3K9Ac-positive structure in each hypnozoite (Figure 5B).

Interestingly, day 21 hypnozoites exhibited multiple, apparently

individualized apicoplasts (Figure 5A). All persistent hypnozoites

showed the unique UIS4-positive PVM prominence (Figures 5A,

5B, and S4B). Hypnozoites were not distinguishable at day 2

postinfection, when all parasites were at trophozoite stage

but could be clearly distinguished from replicating schizonts at

day 3 postinfection onward (Figure 5C). Hypnozoites from all

analyzed time points of infection remained smaller in size than

liver stage schizonts detected at day 3 postinfection (Figure 5C).

Strikingly, we observed a total of three liver stage schizonts

in multiple liver sections analyzed from day 21 infections

530 Cell Host & Microbe 17, 526–535, April 8, 2015 ª2015 Elsevier In

(>30 sections analyzed in five independently infected mice)

(examples are shown in Figures 5D and 5E). These schizonts

were not yet fully mature and were similar in size to schizonts

observed at days 5–7 post-sporozoite infection (Figures 1B

and 5C).

c.

Figure 6. Thai P. vivax Hypnozoite Frequencies and Outcomes of

Experimental Drug Treatments in FRG KO huHep Mice

(A) A graphical representation of hypnozoite and liver stage schizont abun-

dance in infections. Liver stages are denoted as schizonts or hypnozoites. For

quantitative assessment, liver stages observed on ten nonserial liver sections

were totaled for each infected mouse. Results for two Thai P. vivax CS ge-

notypes are shown CS VK247 genotype (two independent experiments with

different patient isolates) and CS VK210 genotype (four independent experi-

ments with different patient isolates). PF NF54, P. falciparum NF54 liver stage

assessment at day 7 post-sporozoite infection. Error bars, standard deviation.

(B) Four mice were treated with primaquine (30 mg/kg) at days 1–3 post-

sporozoite infection (PQ prophylaxis, P) or days 3–7 post-sporozoite infection

(PQ treatment, T). Parasites of the CS VK210 genotype were used for the

experiments. Four untreatedmice served as controls (Control). Eight days after

sporozoite infection, all mice were sacrificed, and cDNA was produced from

three separate liver tissue samples for each mouse. Liver stage burden using

qRT-PCR andwas used to normalizeP. vivax 18S rRNA transcription to human

ApoAI transcription. The graph insert shows microscopic quantitation of liver

stage schizonts and hypnozoites observed by IFA. Area of tissue refers to

cumulative area of tissue analyzed for each treatment where similar number of

sections from each technical replicate was processed.

The Frequency of Hypnozoite Formationin Thai P. vivax IsolatesHypnozoites remain nonreplicating and much smaller than repli-

cating liver stage schizonts, and persist after commencement of

the primary liver stage infection (Figure 5C). Hypnozoites also

possess unique cellular characteristics that can be readily visu-

alized by antibody staining (Figure 4). To determine hypnozoite

frequencies for the Thai P. vivax isolates used for infections,

we calculated their fraction in the total number of liver stage

parasites (schizonts + hypnozoites) at days 5–7 of infection.

We observed that infections performed with P. vivax isolates

of the CS VK247 genotype (experiment performed twice with

two mice per independent experiment and individual VK247 iso-

lates) yielded a hypnozoite frequency of approximately 40%

(Figure 6A), closely matching the hypnozoite frequency of 50%

Cell

predicted (but never formally shown) for tropical strains of

P. vivax (White, 2011). This hypnozoite frequency did not

change significantly between 5 and 7 day infections or mice

that were repopulated using different hepatocyte donors, indi-

cating that hypnozoite frequencies are maintained with progres-

sion of infection in the FRG KO huHep model (Figure S5A).

Strikingly, however, in another set of infections with Thai

P. vivax isolates of the CS VK210 genotype, the hypnozoite fre-

quency was drastically different, averaging only 8% (four inde-

pendent infections with at least two mice per infection; Figures

6A and S5A).

Testing P. vivax Liver Stage Drugs in the FRG KO huHepModelWe next explored the FRG KO huHep model as a platform for

testing drugs against P. vivax liver stages. In addition to evalu-

ating liver stage burden by histological examination, we also

developed a quantitative real-time PCR (qPCR) assay based

on the detection of P. vivax 18S ribosomal RNA (normalized

against human apolipoprotein AI to account for hepatocyte

humanization of the liver). Primaquine is currently the only

licensed drug that has demonstrated efficacy to prevent re-

lapses and affords causal prophylaxis as well as treatment for

P. vivax infections of the liver. We treated a group of four mice

with primaquine (30 mg/kg; intravenously) for 5 days, starting

1 day before P. vivax sporozoite infection (CS VK210 genotype)

as a prophylaxis regimen. A second group of mice was treated

with an equivalent primaquine regimen starting on day 3 post-

sporozoite infection, to assess elimination of established hypno-

zoites. The mice were sacrificed at day 8, and the efficacy of

the drug to eliminate P. vivax liver stage infection relative to

control, four untreated P. vivax-infected mice, was evaluated

by qPCR. In untreated control mice, P. vivax liver stage burden

was robustly detected by qPCR (Figure 6B). In seven out of

eight primaquine-treated, P. vivax-infected mice, liver stage

parasites were undetectable by qPCR (Figure 6B). One mouse

in the prophylaxis group showed a very low signal for parasite

18S rRNA. Histological analysis of multiple liver tissue sections

from all mice readily detected P. vivax liver stage schizonts and

hypnozoites in untreated control livers. No schizonts or hypno-

zoites were detected in the livers of primaquine treated mice,

including the liver that showed residual parasite 18S rRNA signal

by qPCR analysis (Figure 6B). We also investigated the effect of

atovaquone, a drug that eliminates replicating liver stages but

does not prevent relapses, on P. vivax liver stages in the FRG

KO huHep mouse model. We treated a group of three mice

with atovaquone (10 mg/kg; orally) for 3 days, starting 1 day

before P. vivax sporozoite infection (Figure S5B). The mice

were sacrificed at day 8, and the effect of the drug on P. vivax

liver stage infection relative to control—two untreated P. vivax-

infected mice—and the primaquine prophylaxis regimen (three

mice in a prophylaxis regimen described above) was analyzed.

Histological analysis of multiple liver tissue sections from mice

treated with atovaquone showed that P. vivax liver stage schiz-

onts were almost completely absent (one schizont detected) in

the mice. However, establishment of hypnozoites was not

affected by atovaquone (Figure S5B). In contrast, primaquine

prevented establishment of hypnozoites and replicating liver

stages (Figure S5B).

Host & Microbe 17, 526–535, April 8, 2015 ª2015 Elsevier Inc. 531

DISCUSSION

Activation of dormant hypnozoites, subsequent liver stage schi-

zogony, exo-erythrocytic merozoite release, and concomitant

blood stage infection (relapse) constitutes a major determinant

of the unique P. vivax epidemiology (White and Imwong, 2012).

Yet, P. vivax hypnozoites and their occurrence, frequency,

dormancy, and activation remain mostly unexplored. Since pio-

neering studies described P. vivax liver stages in humans and

Aotusmonkeys (Collins et al., 1973; Shortt et al., 1948) and hyp-

nozoites in infected chimpanzees (Krotoski et al., 1982b), little

progress has been made in this area of research, despite the

global importance of vivax malaria.

Here, we have shown that the human liver-chimeric FRG KO

huHep mouse model provides a unique opportunity to advance

our understanding of P. vivax liver stage biology. This model sup-

ports robust infection with P. vivax sporozoites and the develop-

ment of liver stages, both nonreplicating hypnozoites that persist

and replicating schizonts that develop to completion and release

infectious exo-erythrocytic merozoites. One of the advantages

of the model is its small size. In consequence, with only a few

hundred thousand P. vivax sporozoites, a robust liver stage

infection can be achieved, and infection is amenable to quantita-

tive analysis by microscopy and PCR. Thus, the model also

enables testing of drugs that prevent the establishment of hyp-

nozoites or that eliminate established hypnozoites. Although

in vitro models for P. vivax liver stage research have improved

with the advent of microscale primary hepatocytes cultures

(March et al., 2013), the FRG KO huHep mouse model will be

extremely valuable to study P. vivax liver stage biology and test

drugs in vivo.

We have presented numerous lines of evidence that P. vivax

infections in the mouse model result in the formation and persis-

tence of hypnozoites. Intriguingly, our data indicate that inde-

pendent P. vivax isolates from the same geographical area in

Thailand differ in their hypnozoite frequencies, and this appears

to be associated with their genotype. It is known that P. vivax iso-

lates from temperate regions relapse with longer time intervals to

first event and also show longer periodicity when compared to

tropical strains, which show short relapse frequencies (reviewed

in White, 2011). It is possible that longer relapse periodicity is a

parasite adaptation to the seasonal availability of vectors for

transmission. The significance of the different hypnozoite fre-

quencies we have seen for Thai P. vivax isolates is currently

not understood, and it will be of interest to further explore the

question of why parasites in the same geographic area would

show heterogeneity in this important biological aspect of liver

infection. The robust P. vivax liver stage infections we have

observed will allow future quantitative investigation of hypno-

zoite frequencies from P. vivax strains of different geographical

origin. This analysis should contribute to unraveling the relation-

ship between clinically observed relapse patterns and hypno-

zoite frequencies in different parts of the world.

Antibodies to a diversity of parasite proteins allowed a detailed

cellular characterization of P. vivax liver stages. Previous studies

relied on immune sera from patients living in P. vivax-endemic

areas (Krotoski et al., 1982b) and more recently CS specific

antibodies (Rosenberg et al., 1989). The antibodies revealed

the PVM, apicoplast, mitochondria, ER, and Golgi throughout

532 Cell Host & Microbe 17, 526–535, April 8, 2015 ª2015 Elsevier In

liver stage development and also allowed for an unprecedented

characterization of hypnozoites. Hypnozoites exhibited a PVM

with a unique UIS4-positive prominence, a feature not observed

in developing liver stage schizonts or early trophozoite stages. It

is tempting to speculate that this structure constitutes a unique

point of interaction between the hypnozoite and its host hepato-

cyte, but this remains to be investigated. Of equal interest,

hypnozoites showed evidence for active cellular processes,

including development of the ER and initial replication of mito-

chondria and the apicoplast. Hypnozoites also increased, albeit

modestly, in size during the time course of infection (3–21 days

after sporozoite infection). Together, these findings challenge

the prevailing notion that hypnozoites are truly dormant. Never-

theless, hypnozoites did not show evidence of DNA replication

as determined by microscopic analysis of their histone content.

Critically, hypnozoites persisted for at least 21 days post-sporo-

zoite infection, and future experiments will determine if even

longer persistence can be observed in this model. Our initial

observation also suggests that hypnozoite activation occurs,

as we detected a new generation of replicating schizonts at

day 21 post-sporozoite infection, but this will require more

detailed study. It might be feasible to combine the persistence

of hypnozoites in the model with the transitioning of exo-erythro-

cytic merozoites to blood stage infection, which would truly

model relapsing infection. Furthermore, it should be feasible

to test in vivo stimuli that can trigger hypnozoite activation.

Recently, a methodology for longer-term in vitro cultivation of

P. cynomolgi liver stages inMacaca fascicularis primary hepato-

cytes has been described (Dembele et al., 2014). This enabled

the observation that P. cynomolgi hypnozoites could persist,

activate, and enter liver stage schizogony. Additional pharmaco-

logical manipulation of the system with inhibitors of histone

modification enzymes accelerated the rate of hypnozoite activa-

tion. Thus, it will be of interest to determine if manipulation of

histone modifications leads to activation of P. vivax hypnozoites

in the FRG KO huHep mouse model.

While the duration of complete liver stage development for

other Plasmodium species has been experimentally determined

(for example, 6–7 days for P. falciparum [Vaughan et al., 2012]

and 7–11 days forP. cynomolgi [Krotoski et al., 1982a; Voorberg-

van der Wel et al., 2013]), our work directly determined the time

to complete P. vivax liver stage maturation and release of exo-

erythrocytic merozoites using the FRG KO huHep mouse model.

Previously, this had been extrapolated from the appearance of

blood stage parasites after experimental P. vivax sporozoite

infection of humans and nonhuman primates (Anstey et al.,

2012). In the latter, peripheral blood stage parasitemia was de-

tected as early as day 9 (for the Chesson strain) postinoculation

with very large numbers of sporozoites, suggesting that the

release of exo-erythrocytic merozoites occurred at this time

(Krotoski et al., 1982b). In agreement with this, we observed

mature liver stage schizonts containing exo-erythrocytic mero-

zoites as well as the release of merozoites at day 9 post-sporo-

zoite infection. We also observed mature schizonts on day 10

post-sporozoite infection, but their numbers had declined

when compared to day 9 infections, indicating that the peak of

liver stage schizont maturation occurs around day 9. To add

credence to our observations in the liver, we showed that

P. vivax exo-erythrocytic merozoites released from the liver on

c.

days 9 and 10 after sporozoite inoculation infected reticulocytes.

Thus, the model can be used for studying the liver stage-to-

blood stage transition of P. vivax.

Successful prevention (causal prophylaxis) and treatment

(radical cure) of P. vivax hypnozoite infection require two compo-

nents: (1) prevention/elimination of the blood stage infection that

causes pathology and (2) prevention of future occurrences of

relapses by prevention/elimination of hypnozoites in the liver.

Currently, the only approved drug for causal prophylaxis

and radical cure of relapsing P. vivax infections is primaquine.

Primaquine, however, has a short half-life, requires a 2 week

dosage regimen, and has incompatibility with glucose-6-phos-

phate-dehydrogenase deficiency, which requires prescreening

of drug recipients (Kevin Baird, 2013). Therefore, there is an

ongoing search for drugs that could replace primaquine. We

showed that primaquine causal prophylaxis and treatment of

established hypnozoites prevents/eliminates P. vivax liver stage

infection in the FRG KO huHep model. We also showed that

atovaquone has little effect on hypnozoites. Thus, this mouse

model will help in accelerating the discovery of next-generation

drugs that target hypnozoites.

Taken together, our data demonstrate that the FRG KO huHep

mouse model constitutes an unprecedented, robust small ani-

mal model to support P. vivax sporozoite infection, liver stage

development, complete maturation of schizonts, and infectious

exo-erythrocytic merozoite release, as well as hypnozoite forma-

tion and persistence. The advance is of critical importance,

because questions regarding the biological basis of relapses

and their relationship to hypnozoite frequencies and activation

can now easily be addressed in vivo. This will reveal insights

into the unique biology of P. vivax and might accelerate the

development of interventions for causal prophylaxis and radical

cure of P. vivax infection.

EXPERIMENTAL PROCEDURES

Plasmodium vivax Sporozoite Production

Anopheles dirusmosquitoes (from theMahidol University colonymaintained at

the Faculty of Tropical Medicine laboratories) were infected with blood

collected from patients who were confirmed positive for only P. vivax malaria

via microscopy at local health centers in close proximity to the Kanchanaburi

Campus, Mahidol University. In brief, 150 ml of red blood cell pellet from blood

samples was suspended in pooled normal AB serum to a packed cell volume

of 50%. Then, the suspension was fed for 30 min to 100 female mosquitoes

(5–7 days old) via an artificial membrane attached to a water-jacketed glass

feeder maintained at 37�C. Next, unfed mosquitoes were removed, and fed

mosquitoes were maintained on a 10% sucrose solution and incubated at

26�C and 80% humidity for at least 14 days. Salivary gland dissections were

performed at days 14–19. Production of P. falciparum sporozoites and the

infection of FRG KO huHep mice with P. falciparum were performed as

previously described (Vaughan et al., 2012).

FRG KO huHep Mice

Female FRG KO mice engrafted with human hepatocytes (FRG KO huHep)

were purchased from Yecuris Corporation (Oregon, USA). The FRG KOmouse

is a triple gene knockout (Azuma et al., 2007; Bissig et al., 2007). R stands for

recombination-activating gene 2 (Rag2), and G stands for interleukin-2 recep-

tor subunit gamma (Il2rg). The Rag2�/� and Il2rg�/� phenotype is a severely

immunocompromised mouse lacking B, T, and NK cells that does not reject

xenotransplanted huHep. The F stands for fumarylacetoacetate hydrolase

(FAH). Due to the lack of FAH, the hepatocytes of FAH�/� mice suffer buildup

of intracellular fumarylacetoacetate, resulting in their death. The phenotype is

ablated by the addition of 2-(2-nitro-4-trifluoromethylbenzoyl)-1, 3-cyclohexa-

Cell

nedione (NTBC) to the mouse diet, via the drinking water or food (Grompe

et al., 1995). FRG KO huHep mice are cycled with NTBC, which allows for

the repopulation of the mouse liver with huHeps. Repopulation levels can

reach in excess of 90%. Mice were maintained periodically on NTBC

throughout the experimental period according to the supplier’s methodology.

Enriched Human Reticulocyte Preparations

Adult human whole blood was obtained from the Thai Red Cross and depleted

from white blood cells by passage through Pall RN1 filters. The remaining re-

ticulocytes and red blood cells were washedwith RPMI1640 and concentrated

by centrifugation (1,000 g for 10 min). The blood was then overlaid on a Nycon-

denz preparation (19%) in KCl buffer and centrifuged (3,000 g for 30 min) to

separate reticulocytes from the mature red blood cells. The enriched reticulo-

cyte preparation was collected from the Nycondenz interface, washed, and

concentrated again in RPMI1640. The enrichment of reticulocytes was esti-

mated by methylene blue staining. For intravenous (IV) mouse injections,

300–400 ml of a 20%–40% enriched reticulocyte preparation was used.

In Vivo Sporozoite Infection and Liver Isolation

Mice were injected IV into the tail with 3.53 105 to 13 106 sporozoites isolated

from the salivary glands of infected mosquitoes in 100 ml of RPMI media. Mice

were euthanized at 2 (four mice), 3 (two mice), 5 (two mice), 7 (three mice), 8

(>10 mice), 9 (three mice), 10 (four mice), 14 (two mice), and 21 (three mice)

days after sporozoite infection. Livers were perfused with PBS through the

hepatic portal vein, removed, and separated into lobes. The lobes were fixed

in 4% electron microscopy grade formaldehyde in PBS, which was replaced

after 24 hr with TBS + 0.05% sodium azide. The liver lobes were then stored

at 4�C in TBS containing 0.05% sodium azide. The fixed lobes were subse-

quently sliced into 25–50 mm sections for IFA. The sporozoites used for IFA

were fixed in 4% formaldehyde in PBS for 30 min. After processing for IFA,

the sporozoites were immobilized on glass slides for microscopy.

Immunofluorescence Assay

IFAs were carried out as previously described (Vaughan et al., 2009). Human

hepatocytes were detected with a rabbit anti-FAH antibody (a kind gift from

Yecuris Corporation). P. vivax liver stages were detected with a CS protein

mouse mAb (obtained from MR4, ATCC, VA), as well as rabbit polyclonal

antibodies to BiP (Noe et al., 2000), ERD2 (MR4, ATCC, VA), and MSP-1.

The polyclonal and monoclonal antibodies created for this study were devel-

oped by GenScript. In brief, codon-optimized plasmids encoding the whole or

a fragment of the protein of interest (if not full length, the amino acids used are

denoted with superscript numbers next to the name of the protein) were pro-

duced and used for immunization of animals according to the company pro-

tocols. Using this technology, antibodies to UIS479-166 (PVX_001715), HSP60

(PVX_095000), falstatin (PVX_099035), MIF (PVX_124095), AMA-144-487

(PVX_092275), and EXP-1100-148 (PVX_091700) were created. ProMab devel-

oped the mouse monoclonal antibodies to BiP (Roobsoong et al., 2014) and

ACP (PY04779, whole protein) according to the company’s protocols. For

P. vivax histone detection, anti-acetylated histone H3 rabbit antibody (Milli-

pore) was used. Fluorescence and differential interference contrast (DIC)

images were acquired using an Olympus 1 3 70 Delta Vision microscope

equipped with deconvolution software. All of the plasmids used to produce

the proteins for immunizations as well as all the antibodies described above

will be deposited in MR4 as a resource for the malaria research community.

Drug Treatment Studies

For primaquine treatment, 30 mg/kg of primaquine phosphate (Sigma) in PBS

was injected intraperitoneally (i.p.) into mice. For primaquine treatments, mice

were injected i.p. with primaquine at days �1 through 3 post-sporozoite inoc-

ulation or at days 3–7 post-sporozoite injection. For Atovaquone (Sigma)

treatment, 10 mg/kg of drug was administered orally (in 100 ml of PEG450 as

a carrier) at days�1 through 1 post-sporozoite inoculation. For all drug exper-

iments, the P. vivax VK210 genotype was used. All mice were sacrificed at day

8 for analysis. NTBC was not used.

qRT-PCR

Total RNA from liver lobe samples was extracted using Trizol (Invitrogen)

and DNase treated using Turbo-DNase (Ambion). First-strand cDNA was

Host & Microbe 17, 526–535, April 8, 2015 ª2015 Elsevier Inc. 533

synthesized from RNA using the Superscript III Platinum RT Kit (Invitrogen).

The resulting cDNA was used for the amplification of human ApoAI and

P. vivax 18S rRNA cDNA (human ApoAI primers (50-AGCGTGACCTCCACC

TTCAG-30 and 50-CCTTCACCTCCTCCAGATCCTT-30; Pv18S primers, 50-GA

AGAAAATATTGGGATACGTAACAG-30 and 50-ATCGGTAGGAGCGACGGG

CG-30). The qPCR was carried out with SYBR green (Invitrogen) using the

Applied Biosystems 7300 Real-Time PCR System and associated software.

Relative copy numbers for the transcripts under study were calculated using

the DCt method. Each qPCR used identical quantities of first-strand cDNA

generated from three small sections of liver. Mice used in the experiments

were littermates, received the same donor hepatocytes, and were injected

with P. vivax sporozoites (0.8 3 106) on the same date.

Study Approval

The human blood collection protocol was approved by the Ethical Committee

of the Faculty of Tropical Medicine, Mahidol University. The study was per-

formed in strict accordance with the recommendations in the Guide for the

Care and Use of Laboratory Animals of the National Institutes of Health,

USA. To this end, the Seattle Biomedical Research Institute has an Assurance

from the Public Health Service (PHS) through the Office of Laboratory Animal

Welfare (OLAW) for work approved by its Institutional Animal Care and Use

Committee (IACUC). The PHS Assurance number is A3640-01. All of the

work carried out in this study was specifically reviewed and approved by the

Seattle Biomedical Research Institute IACUC.

SUPPLEMENTAL INFORMATION

Supplemental Information includes five figures and can be found with this

article at http://dx.doi.org/10.1016/j.chom.2015.02.011.

AUTHOR CONTRIBUTIONS

S.A.M. and A.M.V. designed and performed experiments, analyzed data, and

wrote the manuscript. N.K., W.R., M.F., N.Y., N.R., V.L., N.S., A.K., N.C., M.B.,

and S.E.L. performed experiments and/or were critical to the execution of

experiments. J.H.A. contributed reagents. J.S. and S.H.I.K. designed experi-

ments, analyzed data, and wrote the manuscript.

ACKNOWLEDGMENTS

Wewould like to thank John Bial and ElizabethWilson (Yecuris Corporation) for

assistance with the humanizedmousemodel, andmembers of the Prachumsri

laboratory with help in mosquito rearing and sporozoite isolation. We would

also like to thank Omar Vandal, Richard Elliot, and Brice Campo for helpful dis-

cussions concerning research presented in this manuscript. The research pre-

sented here was funded from Seattle Biomedical Research Institute internal

financial support to S.H.I.K. in addition to a Global Health Grant from the

Bill and Melinda Gates Foundation to S.H.I.K. (#OPP10215171), S.A.M.

(#OPP1041422), J.H.A., and J.P. (#OPP1023643) and the Medicines for

Malaria Venture (MMV).

Received: February 10, 2014

Revised: July 10, 2014

Accepted: February 16, 2015

Published: March 19, 2015

REFERENCES

Anstey, N.M., Douglas, N.M., Poespoprodjo, J.R., and Price, R.N. (2012).

Plasmodium vivax: clinical spectrum, risk factors and pathogenesis. Adv.

Parasitol. 80, 151–201.

Azuma, H., Paulk, N., Ranade, A., Dorrell, C., Al-Dhalimy, M., Ellis, E., Strom,

S., Kay, M.A., Finegold, M., and Grompe, M. (2007). Robust expansion of

human hepatocytes in Fah�/�/Rag2�/�/Il2rg�/� mice. Nat. Biotechnol.

25, 903–910.

Bissig, K.D., Le, T.T., Woods, N.B., and Verma, I.M. (2007). Repopulation of

adult and neonatal mice with human hepatocytes: a chimeric animal model.

Proc. Natl. Acad. Sci. USA 104, 20507–20511.

534 Cell Host & Microbe 17, 526–535, April 8, 2015 ª2015 Elsevier In

Carlton, J.M., Sina, B.J., and Adams, J.H. (2011). Why is Plasmodium vivax a

neglected tropical disease? PLoS Negl. Trop. Dis. 5, e1160.

Collins, W.E., Contacos, P.G., Jumper, J.R., Smith, C.S., and Skinner, J.C.

(1973). Studies on human malaria in aotus monkeys. 3. Exoerythrocytic stages

of the Salvador II strain of Plasmodium vivax. J. Parasitol. 59, 859–866.

Dembele, L., Franetich, J.F., Lorthiois, A., Gego, A., Zeeman, A.M., Kocken,

C.H., Le Grand, R., Dereuddre-Bosquet, N., van Gemert, G.J., Sauerwein,

R., et al. (2014). Persistence and activation of malaria hypnozoites in long-

term primary hepatocyte cultures. Nat. Med. 20, 307–312.

Doolan, D.L., Hedstrom, R.C., Rogers, W.O., Charoenvit, Y., Rogers, M., de la

Vega, P., and Hoffman, S.L. (1996). Identification and characterization of the

protective hepatocyte erythrocyte protein 17 kDa gene of Plasmodium yoelii,

homolog of Plasmodium falciparum exported protein 1. J. Biol. Chem. 271,

17861–17868.

Elmendorf, H.G., and Haldar, K. (1993). Identification and localization of ERD2

in the malaria parasite Plasmodium falciparum: separation from sites of sphin-

gomyelin synthesis and implications for organization of the Golgi. EMBO J. 12,

4763–4773.

Fernando, D., Rodrigo, C., and Rajapakse, S. (2011). Primaquine in vivax ma-

laria: an update and review on management issues. Malar. J. 10, 351.

Galinski, M.R., Meyer, E.V., and Barnwell, J.W. (2013). Plasmodium vivax:

modern strategies to study a persistent parasite’s life cycle. Adv. Parasitol.

81, 1–26.

Grompe, M., Lindstedt, S., al-Dhalimy, M., Kennaway, N.G.,

Papaconstantinou, J., Torres-Ramos, C.A., Ou, C.N., and Finegold, M.

(1995). Pharmacological correction of neonatal lethal hepatic dysfunction in

a murine model of hereditary tyrosinaemia type I. Nat. Genet. 10, 453–460.

Guerra, C.A., Howes, R.E., Patil, A.P., Gething, P.W., Van Boeckel, T.P.,

Temperley, W.H., Kabaria, C.W., Tatem, A.J., Manh, B.H., Elyazar, I.R., et al.

(2010). The international limits and population at risk of Plasmodium vivax

transmission in 2009. PLoS Negl. Trop. Dis. 4, e774.

Hollingdale, M.R., Collins, W.E., Campbell, C.C., and Schwartz, A.L. (1985).

In vitro culture of two populations (dividing and nondividing) of exoerythrocytic

parasites of Plasmodium vivax. Am. J. Trop. Med. Hyg. 34, 216–222.

Kaiser, K., Matuschewski, K., Camargo, N., Ross, J., and Kappe, S.H. (2004).

Differential transcriptome profiling identifies Plasmodium genes encoding pre-

erythrocytic stage-specific proteins. Mol. Microbiol. 51, 1221–1232.

Kevin Baird, J. (2013). Malaria caused by Plasmodium vivax: recurrent, difficult

to treat, disabling, and threatening to life—averting the infectious bite

preempts these hazards. Pathog. Glob. Health 107, 475–479.

Krotoski, W.A., Bray, R.S., Garnham, P.C., Gwadz, R.W., Killick-Kendrick, R.,

Draper, C.C., Targett, G.A., Krotoski, D.M., Guy, M.W., Koontz, L.C., and

Cogswell, F.B. (1982a). Observations on early and late post-sporozoite tissue

stages in primate malaria. II. The hypnozoite of Plasmodium cynomolgi bastia-

nellii from 3 to 105 days after infection, and detection of 36- to 40-hour pre-

erythrocytic forms. Am. J. Trop. Med. Hyg. 31, 211–225.

Krotoski, W.A., Collins, W.E., Bray, R.S., Garnham, P.C., Cogswell, F.B.,

Gwadz, R.W., Killick-Kendrick, R., Wolf, R., Sinden, R., Koontz, L.C., and

Stanfill, P.S. (1982b). Demonstration of hypnozoites in sporozoite-transmitted

Plasmodium vivax infection. Am. J. Trop. Med. Hyg. 31, 1291–1293.

Krotoski, W.A., Garnham, P.C., Bray, R.S., Krotoski, D.M., Killick-Kendrick, R.,

Draper, C.C., Targett, G.A., and Guy, M.W. (1982c). Observations on early and

late post-sporozoite tissue stages in primate malaria. I. Discovery of a new

latent form of Plasmodium cynomolgi (the hypnozoite), and failure to detect

hepatic forms within the first 24 hours after infection. Am. J. Trop. Med. Hyg.

31, 24–35.

Mackellar, D.C., O’Neill, M.T., Aly, A.S., Sacci, J.B., Jr., Cowman, A.F., and

Kappe, S.H. (2010). Plasmodium falciparum PF10_0164 (ETRAMP10.3) is an

essential parasitophorous vacuole and exported protein in blood stages.

Eukaryot. Cell 9, 784–794.

March, S., Ng, S., Velmurugan, S., Galstian, A., Shan, J., Logan, D.J.,

Carpenter, A.E., Thomas, D., Sim, B.K., Mota, M.M., et al. (2013). A microscale

human liver platform that supports the hepatic stages of Plasmodium falcipa-

rum and vivax. Cell Host Microbe 14, 104–115.

c.

Markus, M.B. (2011). Malaria: origin of the term ‘‘hypnozoite’’. J. Hist. Biol. 44,

781–786.

Mazier, D., Landau, I., Druilhe, P., Miltgen, F., Guguen-Guillouzo, C., Baccam,

D., Baxter, J., Chigot, J.P., andGentilini, M. (1984). Cultivation of the liver forms

of Plasmodium vivax in human hepatocytes. Nature 307, 367–369.

Miller, J.L., Harupa, A., Kappe, S.H., and Mikolajczak, S.A. (2012).

Plasmodium yoelii macrophagemigration inhibitory factor is necessary for effi-

cient liver-stage development. Infect. Immun. 80, 1399–1407.

Mueller, A.K., Camargo, N., Kaiser, K., Andorfer, C., Frevert, U.,

Matuschewski, K., and Kappe, S.H. (2005). Plasmodium liver stage develop-

mental arrest by depletion of a protein at the parasite-host interface. Proc.

Natl. Acad. Sci. USA 102, 3022–3027.

Noe, A.R., Fishkind, D.J., and Adams, J.H. (2000). Spatial and temporal dy-

namics of the secretory pathway during differentiation of the Plasmodium yoe-

lii schizont. Mol. Biochem. Parasitol. 108, 169–185.

Pei, Y., Miller, J.L., Lindner, S.E., Vaughan, A.M., Torii, M., and Kappe, S.H.

(2013). Plasmodium yoelii inhibitor of cysteine proteases is exported to exo-

membrane structures and interacts with yoelipain-2 during asexual blood-

stage development. Cell. Microbiol. 15, 1508–1526.

Price, R.N., Douglas, N.M., and Anstey, N.M. (2009). New developments in

Plasmodium vivax malaria: severe disease and the rise of chloroquine resis-

tance. Curr. Opin. Infect. Dis. 22, 430–435.

Rodhain, J. (1956). Paradoxical behaviour of Plasmodium vivax in the chim-

panzee. Trans. R. Soc. Trop. Med. Hyg. 50, 287–293.

Rongnoparut, P., Supsamran, N., Sattabongkot, J., Suwanabun, N., and

Rosenberg, R. (1995). Phenotype and genotype diversity in the circumsporo-

zoite proteins of Plasmodium vivax in Thailand. Mol. Biochem. Parasitol. 74,

201–210.

Roobsoong, W., Maher, S.P., Rachaphaew, N., Barnes, S.J., Williamson, K.C.,

Sattabongkot, J., and Adams, J.H. (2014). A rapid sensitive, flow cytometry-

based method for the detection of Plasmodium vivax-infected blood cells.

Malar. J. 13, 55.

Rosenberg, R.,Wirtz, R.A., Lanar, D.E., Sattabongkot, J., Hall, T.,Waters, A.P.,

and Prasittisuk, C. (1989). Circumsporozoite protein heterogeneity in the

human malaria parasite Plasmodium vivax. Science 245, 973–976.

Cell

Sattabongkot, J., Yimamnuaychoke, N., Leelaudomlipi, S., Rasameesoraj, M.,

Jenwithisuk, R., Coleman, R.E., Udomsangpetch, R., Cui, L., and Brewer, T.G.

(2006). Establishment of a human hepatocyte line that supports in vitro devel-

opment of the exo-erythrocytic stages of the malaria parasites Plasmodium

falciparum and P. vivax. Am. J. Trop. Med. Hyg. 74, 708–715.

Shortt, H.E., and Garnham, P.C. (1948). Demonstration of a persisting exo-

erythrocytic cycle in Plasmodium cynomolgi and its bearing on the production

of relapses. BMJ 1, 1225–1228.

Shortt, H.E., Garnham, P.C., et al. (1948). The pre-erythrocytic stage of human

malaria, Plasmodium vivax. BMJ 1, 547.

Silvie, O., Briquet, S., Muller, K., Manzoni, G., and Matuschewski, K. (2014).

Post-transcriptional silencing of UIS4 in Plasmodium berghei sporozoites is

important for host switch. Mol. Microbiol. 91, 1200–1213.

Uni, S., Aikawa, M., Collins, W.E., Campbell, C.C., and Hollingdale, M.R.

(1985). Electron microscopy of Plasmodium vivax exoerythrocytic schizonts

grown in vitro in a hepatoma cell line. Am. J. Trop. Med. Hyg. 34, 1017–1021.

Vaughan, A.M., O’Neill, M.T., Tarun, A.S., Camargo, N., Phuong, T.M., Aly,

A.S., Cowman, A.F., and Kappe, S.H. (2009). Type II fatty acid synthesis is

essential only for malaria parasite late liver stage development. Cell.

Microbiol. 11, 506–520.

Vaughan, A.M., Mikolajczak, S.A., Wilson, E.M., Grompe, M., Kaushansky, A.,

Camargo, N., Bial, J., Ploss, A., and Kappe, S.H. (2012). Complete

Plasmodium falciparum liver-stage development in liver-chimeric mice.

J. Clin. Invest. 122, 3618–3628.

Voorberg-van derWel, A., Zeeman, A.M., van Amsterdam, S.M., van den Berg,

A., Klooster, E.J., Iwanaga, S., Janse, C.J., van Gemert, G.J., Sauerwein, R.,

Beenhakker, N., et al. (2013). Transgenic fluorescent Plasmodium cynomolgi

liver stages enable live imaging and purification of Malaria hypnozoite-forms.

PLoS ONE 8, e54888.

White, N.J. (2011). Determinants of relapse periodicity in Plasmodium vivax

malaria. Malar. J. 10, 297.

White, N.J., and Imwong, M. (2012). Relapse. Adv. Parasitol. 80, 113–150.

Host & Microbe 17, 526–535, April 8, 2015 ª2015 Elsevier Inc. 535

Supplemental  Figures:      Figure  S1,  related  to  Figure  1  

P.  vivax  sporozoites  efficiently  infect  human  hepatocytes  in  FRG  KO  huHep  

mice.  (A)  Low  magnification  image  of  a  liver  section  from  an  FRG  KO  huHep  mouse  

infected  with  0.8  x  106    P.  vivax  sporozoites  (VK247).  P.  vivax  liver  stages  at  day  

seven  post  sporozoite  infection  were  visualized  by  immunofluorescence  assay  (IFA)  

with  an  antibody  specific  for  the  P.  vivax  circumsporozoite  protein  (CS;  VK210).  The  

image  shows  numerous  liver  stages  in  red.  scale  bar  –  100  µm.  Table  on  the  right  

shows  numerical  values  for  two  best  infections  with  P.  vivax    isolates  observed  up  to  

date.  Intermittently  a  lower  density  of  infection  can  also  be  observed.  (B)  IFA  of  a  P.  

vivax  in  vivo  liver  stage  at  day  seven.  Liver  sections  were  visualized  by  differential  

interference  contrast  (DIC)  imaging  (left  panel),  staining  with  an  antibody  (red)  

specific  for  P.  vivax  parasite  plasma  membrane-­‐expressed  circumsporozoite  protein  

(CS;  VK247)  (center  panel),  and  DAPI  (blue)  for  DNA  content  (right  panel).  The  liver  

stage  shows  development  of  extensive  membranous  structures,  some  of  them  

connecting  to  the  parasitophorous  vacuole  (PV)  (yellow  arrow).  (C)  IFA  of  P.  vivax-­‐

infected  FRG  KO  huHep  mouse  liver  sections  at  day  five  after  sporozoite  infection.  

Antibodies  used  recognized  P.  vivax  CS  (in  red)  and  human  fumarylacetoacetate  

hydrolase  (hFAH)  (in  green)  and  DAPI  was  used  for  DNA  detection  (blue).  FAH  is  

only  detected  in  human  hepatocytes  since  the  mouse  FAH  gene  is  deleted  in  the  FRG  

KO  mouse  (Azuma  et  al.,  2007).The  top  panel  shows  a  hypnozoite  (yellow  arrow)  

and  the  bottom  panel  a  liver  stage  schizont.  Both  parasites  are  growing  within  a  

human  hepatocyte.  For  reference,  the  DIC  images  are  shown  to  the  right.    

Figure  S2,  related  to  Figure  1  and  2  

Cellular  features  of  the  P.  vivax  liver  stages  detected  in  FRG  KO  huHep  mouse  

infections.  (A)  IFA  of  P.  vivax  liver  stage  development  at  days  three,  five  and  seven  

in  the  FRG  KO  huHep  mice  was  analyzed  with  rabbit  polyclonal  antibodies  to  P.  vivax  

falstatin  (localizes  to  the  PV),  mouse  monoclonal  antibodies  to  circumsporozoite  

protein  (CS)  (localizes  to  the  parasite  plasma  membrane  (PPM)),  rabbit  polyclonal  

antibody  to  P.  vivax  macrophage  migration  inhibitory  factor  (MIF)  (localizes  to  the  

cytoplasm)  and  endoplasmic  reticulum  lumen  protein  retaining  receptor  (ERD2)  

(localizes  to  the  Golgi  apparatus);  Scale  bar:  10    µm.  (B)  IFA  of  a  seven  day  old  P.  

vivax  liver  stage  using  antibodies  to  CS  (green)  and  UIS4  (red).  The  magnification  of  

the  area  in  the  yellow  box  showed  a  lack  of  co-­‐localization  between  CS  and  UIS4.  The  

arrows  point  to  an  area  where  the  separation  of  the  PVM  and  parasite  plasma  

membrane  is  clear.  (C)  IFA  with  and  antibody  to  MSP-­‐1  (green)  showed  that  MSP-­‐1  

liver  stage  expression  (green)  was  weak  at  day  five  and  strong  at  day  seven.  This  is  

in  agreement  with  other  Plasmodium  liver  stage  parasites,  where  MSP-­‐1  is  

expressed  late  in  schizogony  during  merozoite  formation;  Scale  bar:  10    µm.  

Figure  S3,  related  to  Figure  3  

Detection  of  mature  P.  vivax  liver  stage  schizonts  and  persisting  hypnozoites  

in  FRG  KO  huHep  mice.  (A)  IFA  of  a  P.  vivax  liver  stage  at  day  nine  after  sporozoite  

infection  using  antibody  to  UIS4  (green)  to  visualize  the  PVM,  antibody  to  acyl  

carrier  protein  (ACP)  (red)  to  visualize  the  apicoplast  and  DAPI  to  localize  DNA  

(blue).  The  mature  liver  stage  was  in  the  process  of  releasing  exo-­‐erythrocytic  

merozoites  into  the  surrounding  liver  tissue.  Note  the  gap  in  the  PVM  at  the  point  of  

merozoite  egress  (lack  of  UIS4  signal  in  the  bottom-­‐center  of  the  liver  stage  

parasite).  In  the  right  bottom  corner  of  the  image,  a  non-­‐replicating  hypnozoite  was  

also  present.    Scale  bar  –  10  µm.  (B)  IFA  of  a  P.  vivax  liver  stage  at  day  10  using  

antibodies  to  UIS4  (green)  and  ACP  (red);  DAPI  localized  DNA.  ACP  staining  reveals  

the  completion  of  apicoplast  maturation  since  individual  apicoplasts  that  segregate  

with  each  merozoite  were  present.  Scale  bar  –  10  µm.  (C)  IFA  showing  a  

representative  two  day  P.  vivax  liver  stage  localized  with  antibody  to  USI4  (green),  

ACP    (red).  DNA  was  localized  with  DAPI  (blue);  Scale  bar:  10    µm.  

Figure  S4,  related  to  Figure  3  

P.  vivax  exported  protein  1  (EXP-­‐1)  is  expressed  in  liver  stage  schizonts  at  day  

five  onwards  but  not  in  non-­‐replicating  hypnozoites  and  early  schizonts.  

(A)  IFA  was  performed  on  P.  vivax  liver  stage  schizonts  at  days  three,  five,  seven  

after  sporozoite  infection  and  a  day  seven  hypnozoite  using  antibodies  to  EXP-­‐1  

(red)  and  UIS4  (green).  EXP-­‐1  was  detected  in  day  five  and  seven  schizonts  but  was  

absent  in  day  three  schizonts  and  day  seven  hypnozoites.  Conversely,  UIS4  protein  

was  detected  on  all  liver  stages.  Co-­‐localization  of  EXP-­‐1  and  UIS4  suggests  that  

EXP-­‐1  is  also  expressed  on  the  PVM,  in  agreement  with  other  Plasmodium  species.  

DNA  was  localized  with  DAPI;  Scale  bar  –  10  µm.  (B)  IFA  showing  an  example  of  a  

hypnozoite  detected  at  day  14  post  sporozoite  inoculation.  The  hypnozoite  was  

visualized  using  antibodies  to  UIS4  (green)  and  ACP  (red);  DAPI  localized  DNA  

(blue).  Scale  bar  –  10  µm.  

Figure  S5,  related  to  Figure  5  and  6  

P.  vivax  hypnozoite  frequencies  and  outcomes  of  Atovaquone/Primaquine  drug  

treatment.  (A)  A  tabulated  representation  of  hypnozoite  and  liver  stage  trophozoite  

abundance  in  infections  with  percentage  hypnozoite  to  total  infection  calculated.  Liver  

stages  are  denoted  as  schizonts  or  hypnozoites.  For  quantitative  assessment,  liver  

stages  were  scored  on  a  minimum  of  ten  non-­‐serial  liver  sections  for  each  infected  

mouse.  Results  are    shown  for  two  P.  vivax  Thai  isolates  (CS  VK247  (three  independent  

experiments)  and  CS  VK210  (three  independent  experiments)).  Additionally,  for  two  

infections  with  CS  VK247  isolates,  independent  analysis  of  day  five  and  day  seven  

infections  were  performed,  or  mice  that  were  repopulated  using  different  hepatocyte  

donors  (hep  A  -­‐  HHF16001;  hep  B  -­‐  HHF17006;  primary  hepatocyte  reference  numbers  

provided  by  Yecuris),  are  shown.  (B)  For  the  atovaquone  experiment,  three  mice  were  

treated  with  atovaquone  (10  mg/kg)  at  days  -­‐1  thru  1  post  P.  vivax  sporozoite  infection  

(AQ  (P)),  three  mice  were  treated  with  primaquine  (30  mg/kg)  at  days  -­‐1  thru  3  post  P.  

vivax  sporozoite  infection  (PQ  (P))  (CS  VK210  genotype).  Two  untreated  mice  served  as  

controls  (Control).  Eight  days  after  sporozoite  infection,  all  mice  were  sacrificed  and  

liver  stage  burden  was  analyzed  for  hypnozoite  and  schizont  presence  in  liver  section  by  

IFA.  Area  of  tissue  refers  to  cumulative  area  of  tissue  analyzed  for  each  treatment  where  

similar  number  of  sections  from  each  technical  replicate  was  processed.  

Reference:    Azuma,  H.,  Paulk,  N.,  Ranade,  A.,  Dorrell,  C.,  Al-­‐Dhalimy,  M.,  Ellis,  E.,  Strom,  S.,  Kay,  M.A.,  Finegold,  M.,  and  Grompe,  M.  (2007).  Robust  expansion  of  human  hepatocytes  in  Fah-­‐/-­‐/Rag2-­‐/-­‐/Il2rg-­‐/-­‐  mice.  Nat  Biotechnol  25,  903-­‐910.    

Figure S1!

A)!

B)!

C)!

P.  vivax  isolate  

#  mice  

#  liver  stages  (100  

mm2)  

VK247(1)   2   160  

VK247(2)   3   132  

VK210(1)   2   11  

Figure S2!

A)!

B)!

C)!

Figure S3!

A)!

B)!

C)!

Treatment   Area    (mm2)  

#  Schizonts  

#  Hypnozoites  

Control   1060   49   7  

PQ  (P)   1245   0   0  

AQ  (P)   1149   1   6  

Isolate   #  Mice    

#  Schizonts    

#  Hypnozoites  

%  Hypnozoites  

VK247  2  (day  5)   31   19  

37  2  (day  7)   32   18  

VK247  2  (hep  A)   60   40  

39  2  (hep  B)   62   38  

VK247   2   120   40   25  

VK210   2     92   4   4  

VK210   2     116   16   12  

VK210   3     385   12   3  

A)!

B)!

Figure S5!