Further improvements of the P. falciparum humanized mouse model.
ABSTRACT It has been shown previously that it is possible to obtain growth of Plasmodium falciparum in human erythrocytes grafted in mice lacking adaptive immune responses by controlling, to a certain extent, innate defences with liposomes containing clodronate (clo-lip). However, the reproducibility of those models is limited, with only a proportion of animals supporting longstanding parasitemia, due to strong inflammation induced by P. falciparum. Optimisation of the model is much needed for the study of new anti-malarial drugs, drug combinations, and candidate vaccines.
We investigated the possibility of improving previous models by employing the intravenous route (IV) for delivery of both human erythrocytes (huRBC) and P. falciparum, instead of the intraperitoneal route (IP), by testing various immunosuppressive drugs that might help to control innate mouse defences, and by exploring the potential benefits of using immunodeficient mice with additional genetic defects, such as those with IL-2Rγ deficiency (NSG mice).
We demonstrate here the role of aging, of inosine and of the IL-2 receptor γ mutation in controlling P. falciparum induced inflammation. IV delivery of huRBC and P. falciparum in clo-lip treated NSG mice led to successful infection in 100% of inoculated mice, rapid rise of parasitemia to high levels (up to 40%), long-lasting parasitemia, and consistent results from mouse-to-mouse. Characteristics were closer to human infection than in previous models, with evidence of synchronisation, partial sequestration, and receptivity to various P. falciparum strains without preliminary adaptation. However, results show that a major IL-12p70 inflammatory response remains prevalent.
The combination of the NSG mouse, clodronate loaded liposomes, and IV delivery of huRBC has produced a reliable and more relevant model that better meets the needs of Malaria research.
- [show abstract] [hide abstract]
ABSTRACT: Models currently occupy the crucial first step in the research flow for the development of new drugs and vaccines. Some animal models are better at reflecting the host-pathogen interaction in humans than others; this depends on the pathogen and its host specificity. Data gathered from what are often poorly adapted models provide a mosaic of sometimes contradictory information, yet there is little incentive to better delineate the relevance of models or to exploit recent advances to develop improved ones. This review reports on three particularly intractable human pathogens - Mycobacterium, Plasmodium and Schistosoma - and reflects that the extent to which these model systems mimic infection and protection processes in humans might not be sufficiently well defined.Trends in Microbiology 02/2002; 10(10 Suppl):S38-46. · 8.43 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Vaccines against the pre-erythrocytic stages of malaria hold the greatest promise as an effective intervention tool against malaria, as shown by immunization with radiation-attenuated sporozoites over four decades ago. To date, however, the development of subunit vaccines, while generating high expectations and investment, has not lived up at all to the promise. This path has been characterized by insufficient research into both identification of key defense mechanisms in humans and the discovery of better antigens, focusing rather on a technological race of how to present mainly a single antigen. The lack of success has also led, perhaps from desperation, to a revival of the live attenuated sporozoite approach, handicapped, however, by major bottlenecks in production, safety, and regulatory issues. It should now be clear that the field can no longer continue to succeed in mice and fail in the clinic. We advocate here in favor of a third option, relying on an understanding of the basis of attenuated sporozoite immunity in humans, to provide leads to the discovery of critical immunogens and the use of models with validated relevance to the human situation in order to rationalize and renew the promise of pre-erythrocytic subunit vaccines.Current Opinion in Microbiology 09/2007; 10(4):371-8. · 8.23 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: P.berghei ANKA infection in CBA or CB57BL/6 mice is used widely as a murine 'model' of human cerebral malaria (HCM), despite markedly different histopathological features. The pathology of the murine model is characterised by marked inflammation with little or no intracerebral sequestration of parasitised erythrocytes, whereas HCM is associated with intense intracerebral sequestration, often with little inflammatory response. There are now more than ten times as many studies each year of the murine model than on HCM. Of 48 adjunctive interventions evaluated in the murine model, 44 (92%) were successful, compared with only 1 (6%) of 17 evaluated in HCM during the same period. The value of the mouse model in identifying pathological processes or therapeutic interventions in human cerebral malaria is questionable.Trends in Parasitology 11/2009; 26(1):11-5. · 5.51 Impact Factor
Further Improvements of the P. falciparum Humanized
Ludovic Arnold.¤a, Rajeev Kumar Tyagi., Pedro Meija¤b, Claire Swetman, James Gleeson, Jean-Louis
Pe ´rignon, Pierre Druilhe*
Malaria Vaccine Development Laboratory, Institut Pasteur, Paris, France
Background: It has been shown previously that it is possible to obtain growth of Plasmodium falciparum in human
erythrocytes grafted in mice lacking adaptive immune responses by controlling, to a certain extent, innate defences with
liposomes containing clodronate (clo-lip). However, the reproducibility of those models is limited, with only a proportion of
animals supporting longstanding parasitemia, due to strong inflammation induced by P. falciparum. Optimisation of the
model is much needed for the study of new anti-malarial drugs, drug combinations, and candidate vaccines.
Materials/Methods: We investigated the possibility of improving previous models by employing the intravenous route (IV)
for delivery of both human erythrocytes (huRBC) and P. falciparum, instead of the intraperitoneal route (IP), by testing
various immunosuppressive drugs that might help to control innate mouse defences, and by exploring the potential
benefits of using immunodeficient mice with additional genetic defects, such as those with IL-2Rc deficiency (NSG mice).
Results: We demonstrate here the role of aging, of inosine and of the IL-2 receptor c mutation in controlling P. falciparum
induced inflammation. IV delivery of huRBC and P. falciparum in clo-lip treated NSG mice led to successful infection in 100%
of inoculated mice, rapid rise of parasitemia to high levels (up to 40%), long-lasting parasitemia, and consistent results from
mouse-to-mouse. Characteristics were closer to human infection than in previous models, with evidence of synchronisation,
partial sequestration, and receptivity to various P. falciparum strains without preliminary adaptation. However, results show
that a major IL-12p70 inflammatory response remains prevalent.
Conclusion: The combination of the NSG mouse, clodronate loaded liposomes, and IV delivery of huRBC has produced a
reliable and more relevant model that better meets the needs of Malaria research.
Citation: Arnold L, Tyagi RK, Meija P, Swetman C, Gleeson J, et al. (2011) Further Improvements of the P. falciparum Humanized Mouse Model. PLoS ONE 6(3):
Editor: Lisa Ng Fong Poh, Agency for Science, Technology and Research - Singapore Immunology Network, Singapore
Received December 20, 2010; Accepted February 18, 2011; Published March 31, 2011
Copyright: ? 2011 Arnold et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was funded, in part, by a grant from the French Ministry of Research. The funders had no role in study design, data collection and analysis,
decision to publish, or preparation of the manuscript. No additional external funding was received for this study.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
. These authors contributed equally to this work.
¤a Current address: Immunite ´ et Infection, UMR-S 945 INSERM Universite ´ Pierre et Marie Curie, Faculte ´ de Me ´decine, Paris, France
¤b Current address: Department of Genetics and Complex Diseases, Harvard School of Public Health, Boston, Massachusetts, United States of America
The development of a small laboratory model capable of
tolerating and sustaining human malaria infection has almost
unlimited applications in areas such as parasite biology, novel drug
development and vaccine discovery.
Currently, the majority of in vivo investigations into malaria
biology are performed using rodent malaria species such as P.
berghei, and P. yoelii that are much easier to handle however, their
relevance to human malaria has been questioned [1,2,3]. A
convenient model capable of sustaining P. falciparum would
undoubtedly be beneficial. For example, such a model could
serve to harmonise in vitro studies that use P. falciparum in culture,
with the in vivo models that presently use rodent species. While
these approaches are complementary, the current mismatch of
Plasmodium species is a very serious limitation. With the
development of resistance against all existing anti-malarial drugs,
the need for better tools to discover and develop novel classes and
combinations of anti-malarials is imperative [4,5].
The advent of several new mouse strains with genetic immune
deficiencies has greatly benefited the development of a small
laboratory malaria model, and results have shown that such a
model is indeed both achievable and useful [6,7,8,9,10].
Experiments performed so far have used mouse strains such as
SCID, NIH III (Beige Xid Nude), and NOD/SCID, together with
pharmacological agents to control their remaining innate defences,
or mouse adapted parasites. It was shown that by grafting them
with either human erythrocytes [6,11] or human hepatocytes 
these animals can support, respectively, the asexual blood cycle, or
the intra-hepatic cycle of the human parasite P. falciparum.
However, on a day-to-day basis they are quite cumbersome to
manage [6,11], and this onus, combined with their currently poor
reproducibility of successful infection, limits their usefulness, and is
likely to have contributed to the persistent use of rodent Plasmodium
PLoS ONE | www.plosone.org1March 2011 | Volume 6 | Issue 3 | e18045
species in the majority of experimental malaria studies - for
reasons of convenience rather than scientific merit.
A main barrier to achieving an improved, workable P. falciparum
mouse model is the strong pro-inflammatory effect of the parasite
itself. In humans the asexual erythrocytic stages of P. falciparum are
known to result in a systemic inflammatory process that is
responsible for many of the symptoms of the disease .
Although the two situations differ in several respects, a similar
inflammatory response was observed in immunodeficient mice,
lacking T and B cells. In this case, as recently described , P.
falciparum triggers strong pro-inflammatory reactions in monocytes
and macrophages, contributing to partial control of the parasite
and the human erythrocyte (huRBC) graft.
A second significant practical problem with all existing models
developed to date is that huRBC are injected by the intra-peritoneal
(IP) route, which relies on the successful migration of huRBC into
the blood stream across the peritoneum. This is a process that is not
properlyunderstood,and, inturn,this preventsanyrational analysis
and improvement of the model. Existing models have been useful in
proving the concept that P.falciparum can survive in a small rodent
model, however they have also shown that unless inordinate
volumes of blood are injected IP daily [11,14], trans-peritoneal
passage to the peripheral blood is uneven, particularly in the long
term. This leads to important variations from one animal to the
other, most likely related to the inflammatory reaction triggered by
high parasite loads, as shown previously .
With these limitations of current models in mind, we thought it
was essential to attempt to improve the P. falciparum humanised
mouse model, particularly in terms of control of inflammatory
reactions, and reproducibility of parasitemia. We decided to
address these issues by using the IV route for huRBC and parasite
administration, and by investigating other means to increase
control over the mouse innate immune response. The use of this
IV model led us to identify, among several factors investigated, the
effect of aging and that of inosine as significant in reducing
inflammatory reactions, and therefore improving P. falciparum
growth. Moreover, after using various strains of immunodeficient
mice, we investigated, as others , the value of the NOD/
SCID/IL-2Rc-null mouse (NSG mouse) which, due to the knock
out of the c-chain of the IL-2 receptor, has been shown to better
tolerate a variety of transplanted human cells [15,16,17,18].
The resulting new IV model using NSG mice presents several
significant advantages over previously available models. It offers
greater reproducibility, with 100% of mice successfully grafted
without the need for mouse-adapted parasites, consistent curves of
parasitemia, and high levels of infection with up to 40–50% of
total erythrocytes infected.
Materials and Methods
All procedures were carried out in line with the European
Community Council Directive, 24thNovember 1986 (86/609/
EEC), and the European Union guidelines. All procedures were
reviewed and approved by the Pasteur Institut Animal ethical
committee (approval number A 75 15–27). Every effort were made
to minimize suffering.
BALB/c, NOD/SCID and NSG mice were purchased from
Charles River. Immunodeficient mice were kept in sterile isolators.
They were housed in sterilized cages equipped with filter tops
during the experimentation, and they were provided with
autoclaved tap water and a c-irradiated pelleted diet ad libitum.
They were manipulated under pathogen free conditions using
laminar flux hoods.
2. Human Red Blood Cells (huRBC)
Human whole blood was provided by the French blood bank
(Etablissement Franc ¸ais du Sang, Paris, France) and used in
accordance with French legislation. Blood donors had no history
of malaria and all blood-groups were used without observing any
difference on parasite survival. Whole blood was centrifuged thrice
at 9006g, for 5 minutes at room temperature and the buffy coat
was separated in order to eliminate white blood cells and platelets.
Packed RBC were suspended in SAGM (Adenine, Glucose and
mannitol solution) and kept at 4uC for a maximum of 2 weeks.
Before use, huRBC were washed three times with RPMI-1640
medium (Gibco/BRL, Grand Island, N.Y.), supplemented with
1 mg of hypoxanthine per liter (Sigma, St Louis, MO) and
warmed at 37uC.
Blood samples drawn from mice were used to determine the
percentage of huRBC in mouse peripheral blood at regular
intervals by flow cytometry on a FACScalibur (BD biosciences)
using FITC labeled anti-human glycophorin monoclonal antibody
P. falciparum lines 3D7, UPA, and K1 were employed in this
study, along with one clinical isolate taken from a patient at Bichat
Hospital, Paris (used the day after sampling). The Uganda Palo
Alto (UPA) strain employed was the Palo Alto Marburg line, this
was used for all experiments conducted unless otherwise specified.
Parasite cultures were not synchronized and therefore a mix of
various developmental stages was injected to infect mice. Parasites
were maintained under in vitro conditions at 5% hematocrit at
37uC in a candle jar in complete culture medium (RPMI-1640
medium(Gibco/BRL), 35 mM
NaHCO3, 10% albumax (Gibco/BRL) and 1 mg of hypoxanthine
(Sigma) per liter. Parasite samples were cryopreserved using the
glycerol/sorbitol method . The cultures were controlled for
Mycoplasma contamination by using PCR testing.
HEPES(Sigma), 24 mM
4. Mouse infection and immunomodulation protocol
NOD/SCID mice were retro-orbitally injected with 400 ml
huRBC every 3 days to ensure a satisfactory proportion of huRBC
(i.e. chimerism), at the time of infection (<60%). Simultaneously,
0.1 ml of unsized dichloromethylene diphosphonate (Cl2MDP)
encapsulated in liposome (clo-lip) (provided by Nico Van Rooijen)
diluted in 0.4 ml RPMI was intraperitoneally injected. Four
injections at 2–3 day intervals were given before parasite infection.
At the time of the fifth injection, mice were retro-orbitally injected
with 300 ml of a P. falciparum infected huRBC suspension in RPMI
at a parasitemia of 1% (all the developmental forms, i.e. rings,
trophozoites and schizonts were present). After infection huRBC
and clo-lip were supplied every 3 days as described for the pre-
infection step. In some experiments, 250 mg/kg of inosine (Sigma)
was injected intraperitoneally every day as the half-life of inosine is
very short .
In experiments using NSG mice the protocol has been adapted
in order to achieve varying levels of adequate huRBC chimerism
and to avoid overloading the mice. As such, different amounts of
blood were employed, and either 200, 400, 550 or 750 ml huRBC
was injected 3 times per week (i.e. Monday, Wednesday and
Friday) mixed with 250 ml human AB serum, as it has previously
been described that human serum improves huRBC survival in
immunocompromised mice ; 4 injections were done pre-
infection, and clo-lip was injected as described above. Follow-up of
the infection was performed by daily Giemsa stained thin blood
films drawn from the tail vein.
Improved P. falciparum Mouse Model
PLoS ONE | www.plosone.org2 March 2011 | Volume 6 | Issue 3 | e18045
5. Mouse cell isolation
In NOD/SCID mice inflammation was induced by IP injection
of 1 ml 3% thioglycolate (Sigma) diluted in sterile PBS. 4–5 days
after, the peritoneal cavity was washed with HBSS without Ca2+
and Mg2+. The collected cells were washed twice in RPMI
supplemented with L-glutamine, Penicillin (100 U/ml), Streptomy-
cin (100 mg/ml), and 10% Fetal Calf Serum (FCS) and seeded at
36105per well in a 96-well culture plate. Single cell suspensions of
splenocytes were prepared in cold RPMI 1640 medium supple-
mented with 10% FCS and filtered on a 100 mm cell strainer to
remove debris. Erythrocytes were lysed with ACK lysis buffer, and
the splenocytes were washed 2 times with RPMI supplemented with
10% FCS and seeded at 36105per well of a 96 well culture plate.
6. Cytokines/chemokine/chemiluminesence assay
100 ml blood samples were collected from the retro-orbital
plexus with a Pasteur pipette, and sera were stored at 280uC.
Conditioned media obtained after 16 h stimulation of peritoneal
cells and splenocytes with lipopolysaccharide (LPS, 1 mg/ml)
(Sigma) were stored at 280uC. Cytokines and chemokines (IL-6,
MCP-1, IFNc, TNFa, IL-12p70 and IL-10) were quantified using
the BDTMCytometric Bead Array mouse inflammatory kit (BD
biosciences) following the manufacturer’s recommendations on a
FACScalibur (BD biosciences).
Since production of reactive oxygen intermediates (ROI) closely
mirrors the state of activation of macrophages and polymorpho-
nuclear cells, luminol dependent photometric assay was used to
measure ROI. Blood samples were collected from NOD/SCID
mice, washed with HBSS with freshly added Ca2+and Mg2+.
Washed blood was diluted 1/10 in HBSS and 90 ml blood was
added to each well in a 96 well plate (Nunc, Denmark) and
incubated for 30 minutes at 37uC after adding 10 ml PMA (final
concentration 1 mg/ml) to stimulate the cells. 50 ml of luminol
(final concentration 200 mg/ml) solution was added immediately
before measuring emissions.
7. Analysis of deep-seated organs for parasite differential
NSG mice were used for the comparison of parasite differential
counts in the peripheral blood, with that in deep-seated organs.
Four mice were infected with UPA strain, and when a parasitemia
of .10% was reached, a thin smear from peripheral blood was
made before killing the mouse, and harvesting its organs. Kidney,
liver, spleen, lung, and brain were removed from each mouse.
Parasite content was assessed from blots made by repeatedly
spotting sections from each organ. These slides were then stained
with Giemsa. The last blots taken were considered to be the most
representative of the parasite content in the organ’s vascular bed,
and were examined at 1000x magnification to perform differential
counts of each stage (.200 parasites from each organ counted).
Development of an IV model
In a previous study , we have shown that among NOD/
SCID mice inoculated intraperitoneally with huRBC and P.
falciparum (IP-IP model), the majority do not support a stable and
sustained peripheral blood parasitemia. For this reason we
attempted to develop a new model in which both huRBC and
P. falciparum are directly injected into the mouse blood stream (IV-
IV model). With this IV protocol 100% of NOD/SCID mice
proved to be parasitized by day 1 post-inoculation vs only 56%
using the IP protocol (data not shown). Despite this marked
improvement, among the 59 mice treated with the standard
immunomodulatory protocol of innate defences (clo-lip), the
parasitemia duration was short (average 5.5 days) and parasitemia
was stable for $12 days in only 5 mice (Figure 1).
We sought to better understand this large variation in duration
of parasitemia. It was not explained by differences in chimerism.
However, close analysis did show a significant difference in
parasitemia duration between young mice and aged mice.
Figure 1. Effects of aging and inosine on parasitemia in NOD/SCID mice. Parasitemia is shown as a percentage of total erythrocytes found in
mouse peripheral blood measured on giemsa-stained thin smears. Levels are shown for old mice (.15 weeks) receiving inosine (green) or without
inosine (yellow) and young mice (,15 weeks) with inosine (red) or without (blue). P. falciparum 3D7 strain was inoculated on day 0. A 400 ml pellet of
huRBC and clo-lip (100 ml) were administered (IV and IP respectively) 3 times a week to all mice. For the inosine groups, inosine (250 mg/kg) was
administered IP daily. Results represent the mean 6 SEM from 11 different experiments, positive result/success means parasitemia lasting .12 days
a) green curve: n=41, 15 positive, 36,6% success, b) yellow curve; n=30, 5 positive, 16,6% success; c) red curve; n=29, 0 positive, 0% success; d) blue
curve; n=10, 1 positive, 10% success.
Improved P. falciparum Mouse Model
PLoS ONE | www.plosone.org3 March 2011 | Volume 6 | Issue 3 | e18045
Effect of age on inflammation
We define young mice as those less than 15 weeks old and old
mice as greater than 15 weeks old. While 16.6% of the aged mice
were able to support a parasitemia lasting more than 12 days, none
of the young mice supported a parasitemia lasting this long, and
average length of parasitemia for old vs young was 7.1 vs 3.9 days
respectively. The gender of the mice made no difference.
To explain this phenomenon, we analysed the inflammatory
responses of healthy non-infected ‘‘young’’ and ‘‘aged’’ NOD/
SCID mice. The levels of inflammatory mediator production were
investigated in peritoneal cells and splenocytes from these mice in
vitro after stimulation with LPS for 16 hours. The levels of IL-6,
MCP-1, TNFa and IFNc were 1.7, 2.8, 1.8, and 6.4 times lower in
the supernatant of peritoneal cells from aged mice as compared to
that from young mice, respectively. Despite these important
differences, probably due to group size, they were not statistically
significant. The same investigation was performed using spleno-
cytes, and the findings were more modest. The levels of IL-6,
MCP-1, TNFa and IFNc were 1.18, 1.16, 2.14 and 1.54 times
lower in supernatants of splenocytes from aged mice as compared
to young mice, respectively (Figure 2A). Interestingly, cells from
young mice produce very strong inflammatory responses when
incubated in serum from young mice, whereas these responses are
decreased when they are incubated in serum from old mice. This
would imply that these age-related differences are mediated by
Furthermore, a study of O2- free radical emission from PMA-
stimulated peritoneal cells detected by chemiluminescence showed
major differences for young vs aged mice cells, i.e.: 11046354 vs
3086152 CU respectively, and in splenocytes 21006941 vs
13986528 CU respectively. These findings were not statistically
significant either, however (Figure 2B).
The same was addressed in P. falciparum infected NOD/SCID
mice using serum collected 1, 3 and 6 days post-inoculation, either
from young mice with short-lived parasitemia or aged mice with
long-lasting parasitemia. Among the cytokines/chemokines mea-
sured only IFNc differed, particularly on days 1 and 3 (102638 vs
4546168 at day 1, and 2468 vs 3696196 at day 3, in aged vs
young mice, respectively) (Figure 2C).
Effect of immunosupressants on NOD/SCID mice
In order to improve sustainability of parasitemia in the mouse
model further, we investigated immunosuppressive agents that
might help to minimize the innate immune response. Inosine, a
purine nucleoside, proved to be the most effective agent at
improving peripheral blood parasitemia from a group of agents
tested, including TGFb1, dexamethasone, and IL-10.
In vitro experiments found that at a concentration of 1 mM
inosine decreased IL-6, MCP-1, and TNFa secretion from young
NOD/SCID LPS stimulated peritoneal cells by a ratio of 1.72,
3.55, and 2.56 respectively as compared to similar inosine-
untreated cells. This effect was less pronounced on cells from aged
mice, with respective decreases of 1.23, 1.49, and 2.24, although
age related differences were not significant (Figure 3). The other
cytokines/chemokines tested (IL-12p70, IFNc and IL-10) re-
mained unchanged. In the case of NOD/SCID mouse spleno-
cytes, only IFNc release was decreased by treatment with inosine
in vitro. The reduction of IFNc production in young and aged mice
was 2.3 and 1.7 times respectively.
In vivo inoculation of inosine confirmed the above data in both
young and aged mice. In young NOD/SCID mice daily intra-
peritoneal injection of inosine at a dose of 250 mg/kg led to 10% of
inosine-free group. Using the combination of aged mice and inosine
administration resulted in 37% of mice sustaining parasitemia for
.12 days, and an average length of parasitemia of 9.8 days.
Development of an improved IV model using NSG mice
Experiments were performed in NSG mice using various
amounts of huRBC (200, 400, 550 or 750 ml of RBC pellets)
injected IV 3 times a week, and also using the IV route to
introduce the parasite infection. Clo-lip was delivered IP on the
same day as huRBC administration. This is called the NSG-IV
Injecting 750 ml of huRBC resulted in a high proportion of
huRBC in mouse peripheral blood (chimerism), ranging from 83%
to 92% of total erythrocytes (Figure 4A). This high level of
chimerism was stable over a number of months and supported
rapid, optimal growth of P. falciparum (Figure 4B) with abundant
and very healthy parasites (Figure 5). Indeed peripheral blood
parasitemia levels reached 30–35% within 3 weeks (Figure 4B) and
went on to reach a maximum of 56.2% (which corresponds to a
parasitemia of 67% in the huRBC subset of mouse peripheral
blood). The total haematocrit in mouse peripheral blood was on
average 70% using this protocol (data not shown).
Use of a lower dose of huRBC (400 ml) resulted in an initial
chimerism of about 60% (Figure 6A), and parasitemia of up to
7.2% (ca. 12% of huRBC parasitized). Use of this lower huRBC
volume however, was followed by a decrease in chimerism most
likely due to inflammation induced by the parasite, producing a
parallel reduction in parasitemia, which fluctuated at an
intermediate level, i.e. without complete parasite clearance
An analysis of inflammatory markers was performed in mice
from the latter protocol to investigate the observed decrease seen
in chimerism. The results revealed substantial differences from
NOD/SCID mice. The decreased huRBC chimerism and
parasitemia seen in NSG mice correlated with the release of
TNFa and IL-12p70. In contrast IL-6 and IFNc remained almost
unchanged (Figure 6B), whereas in NOD/SCID mice IL-6 and
IFNc levels increased at the time of parasite decrease . Hence
the inflammation produced by NSG mice is reduced in
comparison to that seen in NOD/SCID, and this is likely to be
one of the reasons for improved P. falciparum survival.
We observed from the variety of protocols experimented with,
that provided the chimerism is maintained above 10–20%,
parasitemia can be supported for a number of months (i.e. the
mouse life-span). So, to avoid volume overload in the mice,
where 550 ml of huRBC was injected three times a week (n=47),
witha mean 6 SD chimerism of 74%67% over the first month and
half, decreasing to 56% over the second and third month, and long-
lasting parasitemia reaching up to 20–40% (25.6% to 51.3% of
huRBC parasitized); ii) when 550 mL of huRBC was injected only
twice a week, a chimerism of 40–60% was achieved and this
supported a moderate 5%–12% parasitemia during the first month.
In this manner, adjusting the quantity of huRBC injected leads
to various levels of prolonged, stable parasitemia. The level can be
tailored in this manner to the needs of each experiment.
In this model there was no need for preliminary adaptation of
the parasites, contrary to what has been reported for another
model based upon the NSG mouse . All strains tested were
supported (UPA, 3D7, K1), and parasite growth was readily
apparent from the first days post-inoculation. Naturally, there
were some variations in the maximum parasitemia reached with
each strain, particularly for 3D7, which produced the lowest. Mice
were also infected with a fresh clinical isolate of P. falciparum. With
Improved P. falciparum Mouse Model
PLoS ONE | www.plosone.org4March 2011 | Volume 6 | Issue 3 | e18045
Improved P. falciparum Mouse Model
PLoS ONE | www.plosone.org5March 2011 | Volume 6 | Issue 3 | e18045