Plasmodium berghei in vivo


Description of the characteristics of infections in a number of laboratory rodents after mosquito infection or are found in references #1-6. (Small) differences in the course of infections in laboratory animals can result from a) genetic differences between isolates or laboratory lines of P. berghei, b) genetic differences between different strains and species of rodents and c) environmental differences (see for more information the previous chapter) . Here we describe a number of characteristics of parasites of the ANKA strain of P. berghei in laboratory mice and rats and in the mosquito Anopheles stephensi, with the emphasis on infection features that may be important for experimental procedures in the study of the developmental biology of P. berghei. Many details on the life cycle of P. berghei are also described in 'The life cycle of P. berghei'.

Infection of rodent hosts

Laboratory rodents, such as mice, rat and hamsters are sensitive to infection with P. berghei, both through the bites of infected mosquitoes and through artificial injection of infected blood (mechanical transmission).

Infection by bites of infected mosquitoes
Infection characteristics are dependent on both the mosquito species and the rodent species. Here we describe characteristics of P. berghei in Anopheles stephensi and infection of laboratory mice. Anopheles stephensi is infectious for mice from day 14 after a blood meal on an infected animal and remain infectious for several weeks. When fed on infected mice with a parasitemia between 1% and 5% (on day 4-5 of infection, see below), oocyst numbers between 50-300 per mosquito can easily be obtained, resulting in heavily infected salivary glands that can contain thousands of sporozoites per salivary gland (10.000 - 100.000 salivary gland sporozoites have been recovered per A. stephensi mosquitoes). Per bite probably only 20-50 sporozoites are delivered to the host. Between different strains of (laboratory) rodents, large differences exist in susceptibility to mosquito infection. The following infective sporozoite inocula have been reported: >104, 10 and 4 for BALB/c mice, C57Bl6 mice and thicket rats Thamnomys, respectively. To infect animals, mosquitoes are usually allowed to feed on day 18-25 after their infectious blood meal. The number of mosquitoes that have to feed (inject sporozoites) to establish an infection is dependent on the susceptibility of the mouse strain.
Sporozoites can be observed inside hepatocytes from a few minutes to several hours after inoculation. Within the hepatocyte the sporozoite develops within 47-52 hour via the trophozoite stage into the mature schizont that can contain 1500-8000 merozoites. The total number of merozoites per mature schizont can vary in different hosts. The blood stage infection thus starts around 50h after the infectious bites of mosquitoes. The time point that parasites can be detected in the blood after mosquito infection (prepatent period) by the method of thin blood film examination is dependent on a number of factors. For example, the number of sporozoites injected (dependent on the number of sporozoites per mosquito and the number of mosquitoes that were allowed to feed); the sensitivity of the rodent host to sporozoite infection (see above); the number of merozoites per liver schizont and the asexual multiplication rate in the blood. This latter feature is relatively stable and varies around 10 times per 24 hour during the first phase of blood stage infection (see below) in laboratory rodents such as mice and rats. If one want to compare sporozoite infectivity (viability) based on observations on the course on blood stage infections it is important to choose rodent strains that are highly susceptible to sporozoite infection of the liver and to use artificial intra-venous injection of (low numbers of) isolated salivary gland sporozoites.

Infection by injection of infected blood
Laboratory animals can be infected by intravenous (i.v.) or intraperitoneal (i.p) injection of infected erythrocytes containing asexual blood stages, such as ring forms, trophozoites and schizonts. A routine procedure used by many laboratories is to collect tail blood from an infected mouse with a parasitemia between 1 and 10% and inject 105 to 108 i.p into a naive animal. In our experience about 10% of i.p. injected parasites will survive and enter the blood stream. Since P. berghei (ANKA strain) in laboratory mice/rats has a multiplication rate of about 10x per 24 hour during the first phase of blood stage infection (see below) one can calculate the course of parasitemia after infecting with different doses of infected erythrocytes.
To clone parasites we inject i.v. a single infected erythrocyte in mice (obtained by the method of limiting dilution). In these animals the parasitemia reaches levels of 0.5-1% on day 8 after infection, indicating an average of a 10x multiplication rate per 24h. This multiplication rate is a very stable feature of the parasites in the cloning procedure.

The course of blood stage infections: asexual multiplication

The course of infection is usually monitored by examination of Giemsa stained thin blood smears of infected blood. The development of blood stages of P. berghei in laboratory rodents is asynchronous. However, in our experience the development of parasites of the ANKA strain is partly synchronous. If animals are kept at a normal day-night light regime, most parasites (> 80%) undergo schizogony between 2:00 and 6:00 a.m. in the first phase of the infection (see below), resulting in the presence of mainly ring forms/young trophozoites during the day (9.00-14.00 hour). During schizogony parasites disappear from the peripheral circulation and sequester in the capillaries of the inner organs, such as lungs and spleen. P. berghei has a predilection for invading reticulocytes but it can also invade mature erythrocytes. Between strains and lines of P. berghei small differences may exist in the strength of the reticulocyte preference. The preference for invading reticulocytes can also differ in different laboratory rodents. For example invasion of merozoites of parasites of the ANKA strain is more restricted to reticulocytes in rats then in mice. The reticulocyte preference influences the course of blood infections (see below)

Between experiments, or in different mice, parasites of the same ANKA clone can either switch to normocyte invasion or remain restricted to invasion of reticulocytes. Infections in mice then have two typical courses of parasitemia. Initially all infections have a typical reticulocyte restricted course of infection until parasitemias reach 0.5-2%; mice infected with parasite that make the switch to invading normocytes than rapidly  increase in parasitemia from 0.5-2% to 15-25% within 2 days (at which stage most mice will succumb to experimental cerebral malaria (ECM) in ECM-sensitive mice). In other infections/mice the parasites remain reticulocyte restricted and in these infections there is an actual small drop in parasitemia at around 3-5% parasitemia as consequence of a shortage of reticulocytes in circulation. In this phase the percentage of multiple infected red blood cells often increases. A multiple invaded erythrocyte containing more than two parasites will usually not allow normal development of the parasites into mature schizonts, thereby decreasing the overall multiplication rate. After this short second phase of a reduced multiplication the parasitemia then again rapidly increase, resulting from a 'strong wave’ of reticulocyte production. In these infections mice usually do not die of experimental cerebral malaria (ECM) but eventually die from a fulminating parasitemia/anemia. In this phase the course of infection is much less predictable as a result of different factors, such as the percentage of reticulocytes, the presence of high numbers of multiple invaded erythrocytes, anemia and other pathological complications (and immunological reactions). The relative contribution of these factors on the course of infection is also dependent on the strain of P. berghei, the host species/strain and the age of the host. Outbred Swiss mice that do not develop ECM usually follow a ‘reticulocyte-restricted’ course of parasitemia, whereas Swiss mice that do die from ECM make the switch to invading normocytes. 

Genotypic/phenotypic changes during asexual multiplication
As a result of mutations and other (large scale) DNA rearrangements during mitotic multiplication of the blood stages, parasites arise with altered genotypes/phenotypes resulting of mixed populations of parasites. Examples of such parasites are parasites that lost gametocyte production (see below), parasites with altered karyotypes (see 'The genome of P. berghei'), parasites with altered susceptibilities to drugs. In our laboratory we therefore culture parasite lines for a period of maximal 5-10 weeks. After this period we start infections from 'fresh parasites' obtained from the original parasite clones that are stored in liquid nitrogen.

Gametocyte production and infectivity during blood stage infections

In each blood stage cycle 5-25% of the blood stage parasites of cloned lines of the ANKA strain stop with asexual multiplication and switch to (sexual) differentiation into male and female gametocytes. Since the development into mature gametocytes is only slightly longer (26-30 hour) then the asexual cycle (22-24 hour), gametocytes are continuously present during the first phase (see above) of infection. In the second and third phase of infection the production of gametocytes is less predictable as a result of multiple factors. Multiple invaded erythrocytes with more then two parasites will not allow for maturation of gametocytes; in mature erythrocytes fewer parasites develop into gametocytes then in reticulocytes (whether this is due to a lower commitment to sexual differentiation or a less efficient maturation of gametocytes in mature erythrocytes is not clear). Gametocytes have a survival time of 24-30 hour after which they degenerate and are removed from the circulation. Gametocytes are most infective to mosquitoes in the first phase and the beginning of the second phase of the blood stage infection (see also below). The lower infectivity in the later phases of infection is not due to the loss of 'innate' infectivity of the gametocytes but is the result of inhibitory environmental factors in the blood.

Functional maturity of gametocytes/ookinete conversion rate
In later phases of infection high numbers of gametocytes can be present that are not infectious for mosquitoes but they are fully capable of producing gametes, to fertilise and to form ookinetes under in vitro conditions. We have defined this ability to fertilize in vitro as the 'functional maturity' of gametocytes. Functional maturity is thus not the same as infectivity.

Gametocyte conversion rate
To compare the rate the gametocyte production of different clones/lines of P. berghei we defined the 'gametocyte conversion rate'. The conversion rate is the percentage of blood stage ring forms that develop in synchronized infections under standardized conditions into gametocytes. In the reference clones of the ANKA strain of P. berghei, 15-25% of asexual parasites develop into gametocytes under these standardized conditions. These clones are called 'high gametocyte producers'. We never obtained well-defined clones with an intermediate or low level of gametocyte production. (Uncloned) populations of parasites can show low or intermediate levels of gametocyte production as a result of differing ratio's of non-producer parasites and high gametocyte producer parasites (see below)

Loss of gametocyte production
During asexual multiplication malaria parasites can loose their capacity to produce gametocytes, which is a well-known phenomenon. These non-producer mutant parasites can overgrow the originally gametocyte producer parasites. In our laboratory a number of clones have been obtained from the reference clone 8417HP of the ANKA strain of P. berghei that (irreversibly) have lost the capacity to produce gametocytes ('non-gametocyte producer' clones). In our hands the rate of the loss of gametocyte production is not a stable and predictable feature of clones.

Synchronous (experimentally induced) blood stage infections

Methods are available to establish synchronous development of the P. berghei blood stages during 2 cycles at a parasitemia between 1 and 10%. Infections are started by i.v. injection of purified mature schizonts. After two cycles the synchronicity is lost as a result of aberrant development of parasites in multiple infected erythrocytes. The course of parasite development in synchronous infections is as follows: purified schizonts are i.v. injected at 0 h. Between 0h and 4h all schizonts burst and merozoites invade red blood cells resulting in a 1-3% parasitemia of ring forms. Ring forms develop within 22-24h into mature schizonts, which give rise to the second developmental cycle of the parasites. The old trophozoites/young schizonts disappear from the peripheral circulation at 16-18h and sequester in the capillaries of inner organs. A 'small' percentage of the ring forms do not develop into schizonts but differentiate in 26-30h into mature gametocytes. In our reference 'high producer' clone this percentage is 15-25%. Thus at 24-26h ring forms of the second cycle and (immature) gametocytes of the first cycle are present. The gametocytes of the first cycle survive for a period of 24-26h.

Infection of mosquitoes with P. berghei

The information below is mainly based on experiences with A. stephensi infections and P. berghei infections in mice 

Host feeding of mosquitoes and gametocytes
Mosquitoes can be infected by direct feeding on infected mice in which mature gametocytes are present. In general gametocytes (infected blood) are most infective in first and second phase of a blood stage infection at a parasitemia between 0.5 and 10%. At higher parasitemias infectivity of gametocytes usually decreases. At lower parasitemias the infection rate, as measured by the number of oocysts formed in the mosquito, drops. In our experience there is no clear daily periodicity in gametocyte infectivity. In each asexual blood stage cycle new gametocytes are produced which have a survival time of 24-30 hour. Infection rates of mosquitoes are dependent on the number of gametocytes present in the blood and the infectivity of the gametocytes (see above; in the later stages of an infection infectivity decreases as a result of inhibiting factors present in the blood).

Membrane feeding of mosquitoes with gametocytes/ookinetes
Mosquitoes can be infected 'artificially' by feeding on blood containing gametocytes or mature ookinetes via a membrane. The same membrane-feeding system and feeder types can be used that are developed for membrane feeding of P. falciparum.  Ookinetes for membrane feeding are obtained from in vitro cultures of ookinetes. In general, the infectivity of ookinetes (as measured by counting oocysts) is higher then that of gametocytes when fed via membrane feeding systems.

Mosquitoes and course of infection

A. stephensi is routinely used in many laboratories as a vector for P. berghei. This mosquito species shows 'reasonable' infection rates after feeding on P. berghei infected hosts (see below for more details on infection rates). Breeding of A. stephensi mosquitoes is performed at 26°C and a relative humidity of 70-80% (for details of breeding procedures see ref. # 6). Mosquitoes are usually fed 4 to 7 days post emergence. Host- or membrane-feeding of mosquitoes and maintenance of infected mosquitoes take place at a temperature of 20-21°C.

Mosquitoes are usually allowed to take a blood meal in the first and second phase of an infection (see above) at a parasitemia between 1 and 10%. Gametocyte infectivity decreases later in infection resulting in lower rate of fertilization and ookinete development in the midgut of mosquitoes. The drop in infectivity results from inhibiting factors, which are present in the blood during the later stages of infection. In in vitro cultures of ookinetes these inhibiting factors are diluted in such an extent that no negative effect is observed on fertilization rate and ookinete development and thus gametocyte infectivity (ref. #6; unpublished observations).
It has been calculated that about 1 macrogametocyte out of 70 forms an ookinete in A. stephensi, although at low blood stage and gametocyte densities the efficiency of fertilization and ookinete development can be much higher. In culture we obtain fertilization rates ranging between 50-90%. The fertilization efficiency is usually higher in vitro then in vivo. In a susceptible line of Anopheles atroparvus we found that 1 macrogametocyte out of 20 to 1 out of 150 develops into an ookinete. Live ookinetes can easily be recognized and counted in standard cell-counters under the light-microscope (40x magnification) or under a fluorescence microscope after staining with labelled Mab's against the surface proteins P28 (=Pbs21) or P25. Fifteen to eighteen hours after the blood meal the ookinetes have the characteristic banana-shaped appearance. From about eighteen hours onwards ookinetes start to penetrate the midgut epithelium.

The ookinete transforms into an oocysts beneath the basal lamina of the midgut but outside the midgut epithelium basement membrane. Oocysts can readily be recognized by phase or interference contrast microscopy from the 5th day of infection. It has been calculated that about 1 of 3500 macrogametocytes produce an oocyst and an average of 70 oocysts have been recovered per A. stephensi mosquito over a period of twenty years of infecting mosquitoes with P. berghei (ref. #6). Efficiency of oocyst formation is however dependent on gametocyte/bloodstage density in the blood and the efficiency rises with lower densities of parasites in the blood (see above). In A. stephensi more then 1.000 oocysts can be produced per mosquito. However, highly crowded oocysts do not mature at uniform rate and numerous aberrant (dying) oocysts are produced. Counts of mature oocysts are used to determine the infectivity of gametocytes and the susceptibility of the vector. Oocysts infections may be recorded by intensity or prevalence or a combination of both. The intensity of infection is usually recorded as the geometric mean number of oocysts per mosquito. Geometric means are preferable because this accounts for the strong negative binomial distributions of oocysts in a population of mosquitoes (ref. #6).

Oocysts rupture from day 12 onward and sporozoites are released in the heamocoele. Sporozoites accumulate in the salivary glands from day 12-14 onwards. Salivary gland infections in A. stephensi can be very heavy (10.000 - 100.000). However, only a small percentage of sporozoites produced by the oocysts reach the salivary glands. Estimates vary between <1-11%.


  1.  Carter, R. and Diggs, C.L (1977) Plamodia of rodents. In: Parasitic Protozoa, vol. III. Pp 359-465.
  2. Killick-Kendrick, R. (1978) Taxonomy, Zoography and Evolution. In: Rodent Malaria (R. Killick-Kendrick and W. Peters, eds.) pp. 1-52.Academic Press, London.
  3. Landau, I. And Boulard, Y. (1978) Life cycles and Morphology. In: Rodent Malaria (R. Killick-Kendrick and W. Peters, eds.) pp 53-84.Academic Press, London.
  4. Landau, I and Chabaud, A. (1994). Plasmodium species infecting Thamnomys rutilans: a zoological study. Advances in Parasitology 33, 50-90.
  5. Nussenzweig, R.S., Cochrane, A.H. and Lustig, H.J. (1978). Immunological Responses. In: Rodent Malaria (R. Killick-Kendrick and W. Peters, eds.) pp 248-307.Academic Press, London.
  6. Sinden (1997). Infection of mosquitoes with rodent malaria. In: Molecular Biology of Insect Disease Vectors: A methods manual (Crampton J.M., Beard C.B. and Louis C., eds.). Chapman and Hall, pp.67-91.