Of Plasmodium Berghei* R.E

The development and routine application of high-density exoerythrocytic-stage cultures of Plasmodium berghei* R.E. Sinden, A. Suhrbier, C.S. Davies, S.L. Fleck, K. Hodivala, & J.C. Nicholas

Methods are reviewed for the culture of the exoerythrocytic stages of Plasmodium berghei wherein development reproducibly reflects growth observed in vivo in laboratory rodents. The combination of these methods with the culture of both asexual and sexual blood stages has allowed the completion of the entire vertebrate phase of malaria development in vitro. The development of new methods for high-density exoerythrocytic-stage culture combined with robust statistical analysis ofparasite growth by morphological (light microscopy), or DNA probe methods now allows the critical and precise evaluation of chemothera- peutic or immunological treatments. These methods are illustrated by data obtained on pyrimethamine, primaquine and a hydroxynaphthoquinone. Some of the new avenues ofresearch made feasible by the high- density cultures, e.g., direct immunization to produce monoclonal antibodies and biochemical studies are discussed.

Introduction mature schizonts and merozoites infective to mice. Despite the very significant advances made in the last Yet further advances were made by the introduction decade, the exoerythrocytic (EE) stages ofmammalian of the hepatoma cell line HepG2-A16 (7). Some labor- malarial parasites remain the least understood of all atories have pursued the use of primary hepatocytes stages of the parasite's life-cycle. The major limita- from susceptible natural (8) and laboratory hosts tions in their analysis have been (1) the lack of para- (9, 10). Primary hepatocytes have the advantage that sites in quantities to support direct immunological or they may be obtained from animals that are histocom- biochemical studies, and (2) the absence of an ad- patible to experimental immunized hosts. In addition equately characterized culture system that is repro- they may allow the testing of drugs requiring faithful ducible enough to allow biologically and statistically metabolic activation, although qualitative and quant- robust analysis of parasite growth under a variety of itative inter-species differences commonly exist in the chemotherapeutic and immunological treatments. pathways of hepatic drug metabolism. However, The earliest and most prolific cultures of EE primary hepatocytes also suffer from a number of stages have been achieved using the avian parasites major disadvantages: namely; (1) the metabolic profile Plasmodiumfallax, P. gallinaceum and P. cathemerium of the hepatocyte will vary during the period of cultured'in chicken embryos or in vitro (1-3). Regret- culture; (2) the density of parasites obtained using tably the biological disparity between the EE cycle in these cells is relatively low; (3) for technical reasons avian and mammalian malarial parasites, together preparations will vary from experiment to experiment, with the differences in host immune systems, and the and between laboratories; and (4) preparations 'will unique role of the hepatocyte in detoxification all vary depending upon the nutritional status and other combine to devalue significantly the logical applica- characteristics of the host. Thus, recognizing that all tion of such a robust culture system as a model for the existing culture methods have individual inadequacies major malarial parasites of man, even on very basic we concluded a useful contribution could be made by areas of parasite biology. the development and characterization of a repro- Exoerythrocytic schizonts of mammalian mal- ducible and defined culture system for the EE stages aria parasites were first cultured successfully in em- of P. berghei to act as a robust model for future bryonic rat fibroblasts (4,5). Use of the cell line WI 38 studies. significantly improved the results (6), producing The objectives ofthe work described in this paper were therefore: (1) To develop a reproducible system in which a * From the Molecular and Cellular Parasitology Group, Depart- cloned parasite grows to maturity in a defined host ment of Pure and Applied Biology, Imperial College, London SW7 cell at the same rate and with the same morphology as 2BB, England. Correspondence to Dr R.E. Sinden. that observed in a convenient laboratory host.

Bulletin of the World Health Organization. 66 (Suppl.): 115-125 (1990) 115 R.E. Sinden et al.

(2) To extend current methods for EE culture to Assays for growth of EE stages achieve routine invasion of erythrocytes in vitro fol- (1) Light microscopic analysis of Giemsa-stained lowed by full erythrocytic schizogony and sexual cultures. Equal numbers of sporozoites were added to development, thus reproducing the entire vertebrate host cell cultures and 48 hours later the cultures were phase of the malarial life-cycles in culture in continuo. fixed in Bouin's fluid and stained in Giemsa. All (3) To improve existing EE culture methods to identifiable parasites, irrespective of morphology, produce quantities of parasite that will support direct were counted under the light microscope and the biochemical and immunological studies. mean number of parasites in 4 replicates was deter- (4) To define a routine and reproducible assay mined. The average diameter (mean of longest and for the growth of EE stages which is both objective shortest axes) of 100 schizonts was also measured, and and statistically robust. their nuclear number estimated. Materials and methods (2) Measurement of DNA content. The quantity of parasite DNA in EE cultures was measured using a Parasite radioactively-labelled specific DNA probe (P. berghei sub-telomeric DNA clone H27 (15). DNA was pre- P. berghei ANKA was maintained as an isolate, and as pared from EE cultures using a simplified, rapid a cloned line (2.34L) by cyclic passage in Theiler's procedure (16). Samples were applied to a nylon Original (T.O.) mice and Anopheles stephensi. Salivary membrane (Gene Screen Plus®, NEN) using a dot- gland sporozoites were dissected aseptically from blot manifold (Millipore). mosquitos which had been immobilized by cooling on DNA was hybridized with the P. berghei clone ice or which had been stunned mechanically. They H27 labelled with 32P-dATP by nick-translation (17), were then placed in complete tissue culture medium or by an adaptation of the random oligonucleotide (described below) and were kept on ice during lengthy primer technique (18). Following standard hybridiza- dissections. tion and washing procedures the nylon membranes were air-dried and prepared for autoradiography Host cells using Fuji RX-Xray film and Cronex lightning-plus® Primary hepatocytes were prepared from T.O. mice intensifying screens at -80°C for various periods. by the method of Klaunig et al. (11) and from Wistar Following autoradiography, the membranes were or Brown Norway rats by the method of Seglen et al. cut up and the bound radioactivity associated with (12). Isolated hepatocytes were cultured in William's each sample was determined by liquid scintillation medium E supplemented with 10% fetal calf serum counting. penicillin (F.C.S.), (50 ug/ml), streptomycin (3) Immunofluorescent staining. Parasites were (100 pg/ml) neomycin (50,ug/ml) and media supple- stained at different times with a variety ofmonoclonal ments (L-glutamine, 1 mmol/l; non-essential amino antibodies including anti-PMMSA (19). acids (1 x Flow Mixture); dexamethazone, 1 Mmol/l) (13). All supplements were added fresh to the stock (4) Viability of mature parasites. Mature parasites medium immediately prior to use. were injected intraperitoneally into recipient mice and WI.38 cells were cultured in BME medium; and the mice were subsequently monitored for the appear- HepG2, HuH7 (14) and FHL human hepatoma cell ance of a patent parasitaemia. Alternatively, the lines were cultured in MEM. Both MEM and BME viability of in vitro cultured EE merozoites may be were supplemented as above except that the dexa- ascertained by the in vitro invasion of reticulocytes methasone was omitted and additional glucose was (13). added to the MEM (3 mg/ml). Pre-confluent HepG2 cell monolayers were irradiated with 3000 rad from a gamma (Cobalt 60) source. The medium was then Results replenished and the cells were left for 2-24 hours before infection with sporozoites. Sporozoites were Development and charactrization of EE inoculated in a small volume onto cultures and fresh development medium was added 3 hours later. Thereafter the We have examined a variety of potential host-cell medium was changed twice daily. lines and the results are given here briefly. Of those Cells were cultured in a wide variety of vessels cell lines that supported growth, those with an un- including microtitration plates, glass and thermanox differentiated phenotype like FHL were poor host coverslips, Titre-tek multiwell slides, Costar 24-well cells giving only very few, small parasites which did culture plates and 25 cm2 Falcon flasks, with equal not mature fully. Using parasite density, ultimate size, success. speed of maturation, and the viability of the mature

lie WHO Bulletin OMS: Supplement Vol. 68 1990 High-density liver-stage culture of P. berghel

EE merozoites as criteria for selection, the following in vivo/in vitro comparison in a laboratory model susceptible cells are listed in increasing order of ac- system. The recorded rate of parasite growth is illus- ceptability for the production of EE stages in quant- trated (Fig. 1). The growth curve for P. berghei in vivo ity: primary mouse hepatocyte, primary rat hepato- (20) is also illustrated. The average number of mero- cyte, WI 38 human embryonic lung cell line, and zoites in mature schizonts cultured in HepG2 cells was HepG2 together with HuH7 human hepatoma cell estimated by light microscopy to be in the region of lines. 1000. This value is equivalent to the number observed Having confirmed early in this study the suit- in the laboratory mouse, but falls below the number in ability ofHepG2 (parent line from which the reported the rat (4-8000) (21, 22) and well below figures HepG2-A16 clone was derived (7)), we have attempted quoted for P. berghei in the natural host Thamnomys to achieve culture conditions wherein growth rates (10-18000) (22). and parasite size were the same as that observed in vivo in laboratory rodents. Of the many media sup- plements tested, the following have been found to be beneficial: additional L-glutamine, non-essential High-density cultures amino acids and glucose. Not only does the quantity High numbers of evenly distributed P. berghei EE and composition of the medium modify parasite forms with good growth characteristics can be ob- growth, but also the density ofhost cells and parasites. tained reproducibly in our HepG2 cell culture system At high densities (e.g., when host cells are inadequate- (parasite infection rates of up to 10-20% can be ly irradiated) metabolically active host cells will rap- achieved). In such cultures, however, heterogeneity in idly deplete the medium resulting in parasites growing both the size and the maturity of EE schizonts was at variable rates in the same preparation. observed (Fig. 2) (23). Although this in part reflects Under the favourable conditions described, both a natural heterogeneity and problems of cell growth rate (parasites maturing at 48-60 hours) and density (see above), we believe that artifacts in vitro, size ofthe parasites (up to 50 pm in diameter) paralled e.g., multiple infections of cells, may also be a contrib- those recorded in vivo (20), thus offering a useful utory factor. Similar heterogeneity in oocyst size has

Fig. 1. The Increase In mean diameter of P. berghel EE stages observed In Glemsa-stalned preparations of HepG2 cells. 0 ---- 0 (16) and in vivo 0 --- * (20) (error bars are standard errors of mean).

40 -

30 -

E

a) E 20 - cc (aa)

10 -

I 0 w 9 w I I 1 0 20 30 40 50 60 70 Time after sporozoite inoculation / hours

WHO Bulletin OMS: Supplement Vol. 68 1990 117 R.E. Sinden et al.

Fig. 2. Indirect fluorescent antibody staining of a high- sporozoites onto cultures. Salivary gland homo- density 50-hour EE culture of P. berghel. (Bar is 30 pm) genates yield replicate cultures with very consistent schizont densities which are then readily subjected to drug and immunological assays at practical and statistically viable levels (see below). Large-scale high-density cultures (Fig. 3) are readily handled by a small team (1-3 workers), and produce enough protein for immunization and bio- chemical studies (19, 29). Completion of mammalian phase of malarial life-cycle by continuous in vitro culture Maturation of the EE schizont of P. berghei in vitro differs from that recorded by micro-cinephotography ifi the avian malarial parasites (30), in that merozoites are released in clumps and not as a continuous stream of individual merozoites (13) (Fig. 4). In vivo it has been shown that merozoites are first released by the rupturing parasitophorous vacuole membrane, into the cytoplasm of the degenerating hepatocyte (31). Only later is the hepatocyte plasmalemma broken, and the hepatocyte cytoplasm, nucleus, merozoites and parasite remnants released into the sinusoid- where they will be a target for both specific and non- specific host defence mechanisms. In vivo the clumped

Fig. 3. Glemsa-stained preparation of a high-density 50- hour EE culture of P. berghel. (Bar is 10 pm) been recorded in high-density infections in mosquitos (24, 25). We have examined a wide range of sporozoite bulk preparation techniques from homogenized mosquitos including gradient centrifugation (26), ion-exchange chromatography (27), and differential centrifugation (28), and consider that all techniques e suffer from two serious limitations: firstly the time N taken for preparation, and secondly the co-isolation ; of contaminants. Yeast, fungal and bacterial contam- ination can be a practical problem and varies signific- antly depending on the source of the mosquito. The following anti-fungal and anti-bacterial agents do not inhibit P. berghei EE development: 5'-fluorocytosine (5 pg/ml) and fungizone (amphotericin), streptomycin, penicillin, neomycin, gentamycin. Nystatin was found to be toxic. Salivary glands are dissected under sterile condi- tions and are quickly washed in fresh medium. The glands may be teased to release sporozoites or, for quantitatve studies, they may be carefully liberated using a small (1 ml) glass/teflon homogenizer. The preparation time is kept as short as possible to max- imize sporozoite viability. Invasion in vitro reaches a ___ 3 plateau approximately 1 hour after inculation of _ 1 W :

Ila WHO Bulletin OMS: Supplement Vol. 68 1990 High-onsity liver-stage culture of P. berghol

Fig. 4. Glemsa-stained preparation of an EE schizont Fig. 5. Glemre-stained preparation of erythrocyte sus- released from the cultured HepG2 cell. Note inclusion of pension Incubated with EE merozoltes relased by the host-cell nucleus. (Bar is 10 pm) stirring from mature EE schizonts cultured In vitro. (Bar is 10pm)

material will be quickly massaged and broken down in the circulation releasing the individual merozoites. In vitro this 'massaging' of the parasites is reproduced by vigorous stirring of the cultures-similarly result- Fig. 6. Asexual and sexual blood-stage P. bergh.l para- ing in the release of merozoites (Fig. 5) (13). Thus to sites cultured In erythrocytes Infected In vitro from cul- achieve routine infection of erythrocytes from EE tured EE schizonts. (Bar is 10pum) cultures, the latter were grown in the apparatus of Mons et al. (32) under 5% CO2 and the medium and gas mixture was then changed to those required for blood culture, namely RPMI 1640 with 20% FCS, 0.2% sodium bicarbonate pH 7.2, and a gas mixture of 4% CO2, 4% 02, and 92% N2. The culture was then stirred vigorously (400 rpm) from 48 to 60 hours after sporozoite infection. Reticulocytes (blood from a rat with phenylhydrazine-induced reticulocytosis) were added at 48, 52 and 56 hours and were then transferred to a new vessel stirred at 50 rpm. Medium was replaced every 12 hours and subsequent rounds of blood parasitaemia were initiated at 24-hour inter- vals. The initial blood infection was 1-2%, of which 20% of the parasites matured to gametocytes (Fig. 6). Owing to the asynchronous nature of the release of EE merozoites, the subsequent rounds of asexual schizogony did not significantly increase the para- sitaemia and there was a noticeable decline in the sexual parasites. These experiments clearly support observations made in vivo that gametocytes arise directly from the EE merozoites (33). While we have

WHO Bulletin OMS: Supplement Vol. 68 1990 119 R.E. Sinden et al. not specifically increased the volume of the erythro- pective of size or morphology, as a parameter in the cytic cultures to volumes adequate for membrane evaluation of causal prophylactic drugs. Fig. 7 and feeding of mosquitos, we have no reason to suspect 8 and Table 1 illustrate the effects of pyrimethamine that the gametocytes produced so early in the blood and a hydroxynaphthoquinone (HNQ) on the EE phase would not be infective. The methodology de- development of P. berghei in vitro (16). The numbers veloped is now being applied to both P. vivax and of parasites in 4 replicate cultures were counted and P.falciparum; however, in the former species erythro- good standardization was achieved. Typical sigmoid cytic phase culture is not yet satisfactory (34), and dose-response curves were obtained from which ED50 in the latter species the EE stage cultures are not values of 19.5 and 1.85 nmol were determined for reaching maturity in adequate number. pyrimethamine and the HNQ, respectively. The above method has the disadvantage that parasite counts give no indication of the size or Comparison of assays for EE growth viability of the parasites. Additional information Assays published to date on growth of EE stages concerning the effects of an agent on EE growth may in vivo or in vitro are based either on microscopic therefore be obtained by measuring the sizes of near- counting ofparasites (often very few in number), or on mature (48 hour) parasites (mean of the longest and the use of a quantitative DNA probe (35-38). The shortest axes) under the light microscope. Fig. 9 latter assay has a unique advantage in that it is not illustrates the effects ofpyrimethamine on mean para- dependent upon parasite density to attain a statist- site diameter. These data combined with counts of the ically significant sample of parasites. numbers of parasites in cultures give a more accurate indication of the effects of an agent on parasite devel- (a) Microscopic analysis. Those EE forms which ap- opment but are tedious and time-consuming to pear damaged or 'non-viable' are excluded by some authors from counts, and this introduces a con- collect. siderable subjective element into the assay. We do not (b) DNA probe assays. A large increase in the nuclear consider it possible or practical to distinguish between number of EE parasites occurs during development, 'viable' and 'non-viable' parasites by cytological and measurement of parasite DNA in EE stage cul- means. We have therefore used parasite number, irres- tures therefore provides a sensitive index of both

Fig. 7. The effed of a hydroxynaphthoqulnone on the number of the liver dsag observed in Glemsa-stalned preparadons of P. berghelcultured In 4epG2 cells, expressed as % Inhibition compared to control preparations (16). (Error bars are standard errors)

100 -

80 -

70 - c 0 60 - ,_ 50 - -o o- 40 -

30 - 20 - TI

10 - . n v , 9 I . ...1 . .0a. l 0. .1 1 10 1X0 1 000 Concentration of A566C/ nM

120 WHO Bulletin OMS Supplement Vol. 68 1990 Itgh-density fver-stge culture of P. berghI

Fig. 8. The effect of pyrlmethmilne on th number of the liver stages observed In Gem-stanwd preparatIons of P. b.rgh.I cultured In HepG2 cells, expressed as % Inhibiton compared to control prerations (16). (Error bars are standard errors)

100 - 90- 80 - 70 - C. 60 - .2 I 50 - . . . . . c - 40 9 - 30 I

a 20 - a I 10 -

Ia I 0 Ii . . .. a I . . 10 100 1000 Concentration of pyrimethamine/ nM

Table 1: Effect of hydroxynaphthoqulnone on the development of P. b.rghel EE sages In HepG2 cells

Concentration of hydroxy- Mean diameter naphthoquinone Mean number of EE forms % inhibition of EE forms (pm) (nmol/l) ±S.E. (n=4) of development +S.E.

0 586.25+12.07 - 29.12±0.66 0.27 554.25 + 19.858 5.46 25.65 +0.84c 1.36 435.75 + 12.69b 25.67 25.95 0.77c 2.73 108.75 + 7.19 81.45 15.79+ 0.71 13.63 0 100 27.26 0 100 272.26 0 1t0 - 1% Ethanol 355.25 ± 20.62 39.40 20.67 ± 0.55

Not significantly different from control (0.5 > P > 0.1). b Significantly different from control (P < 0.001). c Significantly different from control (0.01 > P > 0.001). parasite growth and number. An advantage of this used by Ferreira et al. (36, 37); 10-50 pg of parasite technique is that a parasite-specific DNA probe is DNA (or about 350-1500 haploid nuclei) could be able to detect parasite DNA even in the presence of a specifically detected by the DNA probe following vast excess ofcontaminating host-cell DNA. Thus the labelling by the random oligonucleotide primer tech- parasite may be enumerated in vivo where even the nique. This sensitivity lies well within the capacity of injection of large numbers of sporozoites produces both in vitro cultures (see below) and in vivo studies negligible parasite biomass in the liver (36). (36-38). Illustrated here (Fig. 10) is the dose-response The DNA probe used here, (clone H27) (15), has analysis of triplicate cultures treated with primaquine the same sequence as the independently cloned probe (16). These results, and similarly the results of others

WHO Bulletin OMS: Supplement Vol. 68 190 121 R.E. Sinden et al.

Fig. 9. The effect of a hydroxynaphthoquinone on the (36-38), indicate the suitability of this assay method mean diameter of P. berghel EE forms observed in Glem- for the assessment of intervention studies on EE sa-stained preparations of cultured HepG2 cells (16). stages. (Error bars are 95% confidence limits, n = 100 except for pyrimethamine concentrations of 25.13 nmol and 50.26 (c) Immunological assays. Monoclonal antibodies nmol for which n = 50 and 4, respectively) have been described (19, 39, 40) that recognize anti- gens expressed in the EE cycle. Apart from allowing the number and size of parasites to be determined, the patterns seen for some monoclonal antibodies (e.g., anti-PMMSA) also give information about the 35- morphology of the parasites (19, 39). One antibody E 17.6.1 has the characteristic of only staining fully 30 - mature EE merozoites (40) and may therefore be used LL to assess parasite maturation in both IFAT and 0 25 - ELISA techniques; however, mature parasites quickly wLU detach from the substratum and float free in the w 20 medium, making quantitative assays in vitro very difficult. 0 0 15 - E (d) Viability assays. The assays described above do 10 not directly assess the number of viable parasites 0 observed, and whether any treatment is therefore c0. 5- protective. Tests on EE stage viability in vivo are complicated by the 'carry-over' of treatment from the / 1. - - - - liver-stage parasite to the blood stages, and complic- 0 -+j -j . T . ated experimental design is required to overcome 0 12.56 25.13 50.26 100.52 this difficulty, e.g., the 'Hill-test' (41), or assays on the Concentration of pyrimethamine/nM course of asexual parasitaemia (29).

Fig. 10. The effect of primaquine on the development of P. berghel In cultures of HepG2 cells as measured by the hybridization of a radiolabelled sub-telomeric P. berghel DNA probe (16). (Error bars are standard errors) 60000 -

50000 -

a) -0 40000 - c E - 0a, 30000 U, 0- 0c 20000 -

10000 -

i O- v ...... '*U- .1 1 0 100 Concentration of Primaquine/LM

122 WHO Bulletin OMS: Supplement Vol. 68 1990 High-density liver-stage culture of P. bergh.l

With the new-found ability to induce blood-stage receptors for cell-mediated immunity studies. In the infections directly from EE cultures in vitro (13), the future, cell lines transfected with appropriate MHC experimental design makes possible the unambiguous genes may become available. At present, primary targeting ofthe treatment to either the EE stage or the hepatocyte cells provide the only workable solution to blood-stage parasites. Thus, a true in vitro assay for this dilemma. causal prophylactic antimalarials is now available. In drug-sensitivity studies hepatoma cell lines We anticipate this assay will be of considerable value may not faithfully reproduce drug metabolism by in the evaluation of re-formulated, or new anti- primary hepatocytes. HepG2 cells can perform cyto- malarial drugs for anti-liver stage activity (16). chrome P-450 dependent mixed function oxidase and conjugation reactions, although differences have been observed in the activities of some of the drug meta- Discussion bolizing enzymes as compared with freshly isolated Since the first partially successful experiments on the primary human hepatocytes (43, 44). Primaquine, growth of mammalian malarial parasite liver stages a which is believed to require activation by the host cell, decade ago (4, 5), the abilities to culture and define has a low activity against P. berghei EE forms cul- parasite growth have increased manyfold. We have tured in HepG2 cells (45) compared to P. yoelii grown developed methods by which parasite growth in vitro in primary Thamnomys hepatocytes (8). This may faithfully reflects that in vivo not only in achieving full therefore reflect the inability of HepG2 to transform maturity of the liver stages, but additionally produ- primaquine to its active state. Inter-species differences cing repeatable infections of erythrocytes in vitro, in drug metabolism also exist and hence primary wherein full asexual and sexual development ensues. rodent hepatocytes may not always be appropriate The high densities obtainable in the system de- for testing potential causal prophylactic drugs for scribed contrasts with the relatively low yield of para- human use. In order to obtain a more complete sites which we, and other workers, obtain in primary picture of the activity of individual drugs several rodent hepatocyte cultures. For example, for P. yoelii different culture systems should ideally be used, the EE stages cultured in primary Thamnomys hepato- identification of alternative human hepatoma host cytes, yields of 0.03-0. 10 parasites per mm2 have been cells (e.g., HuH7) that support full EE development reported (42) compared to 4.5-45 per mm2 in the may be useful in this respect. system described here. With such high density cul- The extension of the methods developed here for tures new areas of research may be pursued; these the culture of asexual and sexual blood-stages directly include the biochemical analysis of liver-stage anti- from the liver-stage parasites in vitro to the human gens and the direct immunization of animals with EE parasites P. vivax and P. falciparum are not without parasites. Furthermore, statistically robust analyses of major technical difficulties (e.g., the culture of the drug or other intervention studies are possible. blood stages of P. vivax (35) or the routine full mat- Light microscopic methods are clearly unsuitable uration of the EE stages of P.falciparum in vitro (46)). for the routine analysis of large-scale intervention However, their successful completion would release studies. Previous studies which assessed the effects of many preliminary experimental studies from their agents on, EE development in low-density cultures possibly wasteful dependence upon the scarce and often failed to record either the absolute number of valuable resource of susceptible primates. parasites observed or the variation between replicates, thus raising considerable doubt about their statistical validity. Even when high parasite densities are achiev- Acknowledgements ed, inaccuracies may still arise if a subjective analysis of parasite morphology is included in the assay. We gratefully acknowledge the financial support of the Medical Research Council of Great Britain, the Science The DNA probe assay is more objective than and Engineering Research Council of Great Britain, the microscopic observation, and now that the procedure UNDP/World Bank/WHO Special Programme for Research for DNA extraction from EE stage cultures has been and Training in Tropical Diseases, and the Leverhulme greatly simplified compared with earlier methods, we Trust. We also thank Dr Makabayashi (Nagoya University, consider it to be suitable for the routine processing Japan) and Dr Monjardino (Department of Medicine, St. of large numbers of samples. DNA methods cannot Mary's Hospital, London) for the cell lines HuH7 and FHL, provide information about morphology or viability, respectively. thus light microscopic and viability studies may also be required to characterize further the effect of any intervention. References One major limitation of using hepatoma cell lines 1. Hawking, F. Tissue culture of malaria parasites (Plas- is that they do not carry appropriately matched MHC modium gallinaceum). Lancet, 246: 693-694 (1944).

WHO Bulletin OMS: Supplement Vol. 68 1990 123 R.E. SInden et al.

2. Dubin, I.N. The cultivation of exoerythrocytic forms of stage of malaria. Am. J. trop. med. hyg., 40: 351-355 Plasmodium gallinaceum in tissue culture. J. infect. (1989). dis., 91: 33-49 (1952). 20. Landau, I. & Boulard Y. Life cycles and morphology. 3. Beaudoln, R.L. & Strome, C.P.A. The feeding process In: Killick-Kendrick, R. & Peters, W., ed. Rodent mal- In the exoerythrocytic stages of Plasmodium lophurae aria. London, Academic Press, 1978, pp. 53-84. based upon observations with the electron micro- 21. Landau, I. & KIllck-Kendrick, R. Rodent Plasmodla of scope. Proc. HelminthoL Soc. Wash., 39: 163-173 the Republique Centrafricaine: the sporogony and (1972) tissue stages of Plasmodium chabaudi and Plasma- 4. Strome, C.P. et al. The cultivation of the exoerythro- dium yoelli. Trans. Roy. Soc. Trop. Med. Hyg., 60: cytic stages of Plasmodium berghei from sporozoites. 633-649 (1966). In Vitro, 15: 531-536 (1979). 22. Yoell, M. et al. Life cycle and patterns of development 5. Slnden, R.E. & Smith, J. Culture of the liver stages of Plasmodium berghei in normal and experimental (exoerythrocytic schizonts) of rodent malaria para- hosts. Milit. med. (suppl.) 131: 900-918 (1966). sites from sporozoites in vitro. Trans. Roy. Soc. Trop. 23. Subhrbler, A. et al. The fate of the circumsporo- Med. Hyg., 74: 134-136 (1980). zoite antigens during the exoerythrocytic stage of 6. Holllngdale, M.R. et al. In vitro cultivation of the Plasmodium berghei. Europ. j. cell biol., 46: 25-30 exoerythrocytic stage of Plasmodium berghei from (1988). sporozoites. Science, 213: 1021-1022 (1981). 24. Batort, J. The biology of rodent malaria with particular 7. Holllngdale, M.R. et al. In vitro cultivation of the reference to Plasmodium vinkei vinkei Rodain 1952. exoerythrocytic stages of Plasmodium berghei in a Ann. Soc. belge Med. trop., 51: 1-204 (1971). hepatoma cell line. Am. J. trop. med. hyg., 32:682-684 25. Slnden, R.E. The cell biology. In: Killick-Kendrick, R. & (1983). Peters, W., ed. Rodent malaria. London, Academic 8. Mazler, D. et al. Inf6station in vitro d'hepatocytes de Press, 1978, pp. 85-168. Thamnomys adulte par des sporozoites de Plasmo- 26. Pachco, N.D. et al. Rapid and large-scale isolation dium yoelii: schizogonie et lib6ration de m6rozoites of Plasmodium berghei from Infected mosquitoes. J. infestants. Comptes rendus Acad. Sci., 294: 963-965 parasitol., 66: 414-417 (1979). (1982). 27. Moser, G. et al. Sporozoites of rodent and simian 9. Lamblotte, M. et al. Developpment de schizontes dans malaria, purified by anion exchangers retain their des h6patocytes de rat adulte en culture apr6s in- immunogenicity and infectivity. J. protozool., 25: festation in vitro par des sporozoites de Plasmodium 119-124 (1978). yoelii. Comptes rendus Acad. Sci., 293: 431-433 28. Turner, D. & Gr.gson, N.A. The cell surface of Plas- (1981). modium gallinaceum sporozoites: microelectrophore- 10. MIllet, P. et al. La culture des schizontes exoerythro- tic and lectin binding properties. Parasitology, 84: cytaires des Plasmodium des rongeurs dans les hep- 227-238 (1982). atocytes: un nouveau modele exp6rimental pour la 29. Suhrbler, A. et al. An antigen specific to the liver stage chimiotherapie du paludisme. Comptes rendus Acad. of rodent malaria recognized by a monoclonal anti- Sci., 301: 403-406 (1985). body. Parasite immunol., 12: 473-481 (1990). 11. Klaunlg, J.E. et al. Mouse liver cell culture, I. Hepatic 30. Huff, C.G. et al. The morphology and behaviour of isolation. In Vitro, 17: 913-925 (1981). living exoerythrocytic stages of Plasmodium gallina- 12. Seglen, P.O. Preparation of isolated rat liver cells. ceum and P. fallax and their host cells. J. blophys. Methods in cell biology, 13: 29-83 (1976). biochem. cytol., 7: 93-102 (1960). 13. Suhrbler, A. et al. The complete development in vitro 31. Moles, J. & Verhave, J.P. Exoerythrocytic development of the vertebrate phase of the mammalian parasite of malarial parasites. Adv. parasit., 27: 1-66 (1988). Plasmodium berghei. Trans. Roy. Soc. Trop. Med. 32. Mons, B. at al. In vitro culture of Plasmodium berghel Hyg., 81: 907-909 (1987). using a new suspension system. Int. j. parasit., 13: 14. Makabmyashl, H. et al. Growth of human hepatoma 213-217 (1983). cell lines with differentiated functions in clinically 33. Kilick-Kendrlck, R. & Warren, M. Primary exoerythro- define medium. Cancer res., 42: 3858-3863 (1982). cytic schizonts of a mammalian Plasmodlum as a 15. Pace, T. et al. Telomeric motifs are present in a highly source of gametocytes. Nature, 220: 191-192 (1968). repetitive element in the Plasmodium berghel gen- 34. Mons, B. at al. Erythrocytic schizogony and invasion ome. Mol. biochem. parasitol., 24: 123-202 (1987). of Plasmodium vivax in vitro. Int. J. parasit., 18: 16. Davles, C.S. In vitro assay of hydroxynaphthoquin- 307-311 (1988). ones against the liver stages of Plasmodlum. Ph.D. 35. Peters, W. & Robinson, B.L. The chemotherapy of thesis, University of London, 1989, pp. 1-177. rodent malaria, XLIII. Indolo (3,2-c) quinoline-N-oxi- 17. Rlgby, P.W.J. et al. Labelling deoxyribonucleic acid to des. Ann. trop. med. parasit., 82: 423-427 (1988). high specific activity with DNA polymerase 1. J. mol. 36. Ferrelra, A. et al. Infectivity of Plasmodium berghei biol., 113: 237-251 (1977). sporozoites measured with a DNA probe. Mol. bio- 18. Felnberg, A.P. & Vogmlsteln, B. A technique for radio- chem. parasitol., 19: 103-109 (1988). labelling DNA restriction endonuclease fragments to 37. Ferrelra, A. at al. Inhibition of development of exo- high specific activity. Anal. biochem., 132:6-13 (1983). erythrocytic forms of malaria parasites by y-inter- 19. Suhrbler, A. et al. Expression of the precursor of the feron. Science, 232: 881-884 (1986). major merozoite surface antigens during the liver 38. Schofield, L. at al. Interferon-y inhibits the intrahep-

124 WHO Bulletin OMS: Supplement Vol. 68 1990 High-density liver-stage culture of P. berghel

atocytic development of malaria parasites in vitro. J. Primaquine: pharmokinetics, metabolism, toxicity and immunol., 139: 2020-2025 (1987). activity. Chichester, Wiley, 1987, pp. 139-143. 39. Szarfman, A. et al. Mature liver stages of cloned 43. Dawson, J.R. et al. Induction of drug metabolizing Plasmodium falciparum share epitopes with proteins enzymes in a human liver cell line HepG2. FEBS from sporozoites and asexual blood stages. Parasit. letters, 183: 219-222 (1985). immunol., 10: 339-351 (1988). 44. Grant, M.H. et al. Mixed function oxidase and UDP- 40. Suhrbler, A. et al. Contrasts in antigen expression in glucuronyltransferase activities in the human HepG2 the erythrocytic and exoerythrocytic stages of rodent cell line. Biochem. pharmacol., 37: 411-416 (1988). malaria. Parasitology, 99: 165-170 (1989). 45. Millet, P. et al. In vitro testing of antimalarial exo- 41. HIlil, J. The activity of some antibiotics and long-acting erythrocytic schizontocides in primary cultures of compounds against the tissue stages of Plasmodium hepatocytes. Mem. Inst. Oswaldo Cruz, 81 (suppl. Il): berghei. Ann. trop. med. parasitol., 69: 421-427 135-141 (1986). (1975). 46. Mazher, D. et al. Complete development of hepatic 42. Landau, I. et al. The culture of Plasmodium schizonts stage of Plasmodium falciparum in vitro. Science, 277: in hepatocytes. In: Wernsdorfer, W.H. & Trigg, P.l., ed. 440-442 (1985).

WHO Bulletin OMS: Supplement Vol. 68 1990 125