Recent developments in production and purification of malaria antigens: Evidence for environmental modulation of gametocytogenesis in Plasmodium falciparum in continuous culture*

Buleujn of the WorldHealth Organization. 57 (Suppl. 1). 37-52 (1979) Evidence for environmental modulation of gametocytogenesis in Plasmodium falciparum in continuous culture* RICHARD CARTER' & Louis H. MILLER2 With the introduction ofcontinuous culture ofPlasmodium falciparum it has become possible to study thefactors involved in gamnetocyte production in vitro and thus eliminate the uncontrollable in vivo variables of the host. The authors have developed a methodfor measuring quantitatively the rate ofproduction ofgametocytes at any time in such cultures. The method is based on an estimation of the percentage of ringforms that develop into stage II gametocytes. Using this approach, it wasfound Ihat dilution ofcultures withfresh red blood cells so as to lower the parasitaemia led to rapidfall in the rate of conversion to gametocytes. The conversion ratessubsequently rose again to levels typically in the order of10% afterseveral days ofgrowth in the new culture. In the parental culturesfrom vhich the dilutions were made, conversion rates remained high at all times. Thispattern was consistently observed in three different isolates of P. falciparum from Africa and the results indicate that the reduction ofparasitaemia by addition offresh cells was responsiblefor reducingproduction ofgametocytes and that conditions associated with a period of growth in culture induced renewedgametocytogenesis. The authorsconclude, therefore, thatenvironmentalconditions directly modulate the rate ofgametocyteproduction by P. falciparum in culture. After 1 ½1 years in culture, parasites have retained theirability toproduce gametocytes and the gametocytes to undergo exflagellat:on. Asexual multiplication of malaria parasites takes place within the host red blood cells (RBCs) by the process of schizogony and results in the production of merozoites which break out of infected cells and rapidly invade other RBCs. During the course of the continuing cycles of asexual reproduction some parasites convert into sexual forms, the male and female gametocytes. The factors controlling the formation of gametocytes (gametocytogenesis) are unknown. For instance, is the course of gametocytogenesis in malaria parasites predetermined as it appears to be in the related coc- cidian parasites? Do environmental conditions within the host modulate gametocyte production? Can gen- etic differences among parasites of the same species influence the parasites' capacitv for gametocyto- genesis? In the present report we describe an approach to the study of these questions as they relate to Pkastmodiuin Fromn the Malana Section, Laboratory of Parasitic Disea.ves, National Institute of Allergy and Infectious Diseascs, National Insiu- tutes of Health, Bethesda, MID 20205, USA. Visiting Scientist SecEion Head falciparu,n in culture and we provide evidence that environmental conditions do modulate the rate of gametocyte production in this species of malaria parasite. MATERIALS AND METHODS Isolates of P. falciparum Three isolates of P.falciparum (designated Z, GI, and G2) were used in this study. The derivation of the Z isolate has already been described (6). Originating from an infectcion acquired by a nonimmune North American travelling in Zaire in December 1976, lines of this parasite have been maintained in culture since January 1977. The GI and G2 isolates were derived from the peripheral blood of two Gambian children wvith parasitaemias of the order of 1% .a Both GI and G2 were stored in glycerol (9) under liquid nitrogen immediately after drawing blood. Stabilates of these I The blood wvas collected in December 1977 at the Br lish Medical Research Council Laboralories, Fajara, The Gambia. The authors are indebrea to Dr R. S Bra:, Dr M. E Smalley, and Dr G. Pasvol ior their kinC cooperation. 3867 - 37- 38 R. CARTER & LE H. MILLER isolates were introduced into culture on 18 April and 5 May 1978, respectively, and have been maintained in continuous culture since that time. Continuous culture of P. falciparum The method for culturing P.falciparum was de- iived from the methods described by Trager & Jensen (20) and Haynes et al. (13). Medium RPMI 1640 with added HMPES buffer was made up as described by Trager & Jensen. Immediately prior to use 4.2 ml of 5/o NaHCO3 was added to each 100 ml of RPMI 1640 medium with HEPES. Sucb a medium will be referred to as incomplete medium. Complete medium was prepared by supplementing the incomplete medium with 100 ml of heat inactivated human serum (from donors of blood typeA Rh+, orAB Rh±) per litre. Both complete and incomplete media were fil- tered through 0.45-jim Nalgen fllters immediately prior to use. Cultures were established from isolates of infected human blood stored under liquid nitrogen, as follows. Stabilates were thawed at room temperature and deglycerolized according to the method of Diggs et al. (9), using incomplete medium as a washing solu- tion. The RBCs were then made up to a 2% haemato- crit in complete medium with type AB serum. As a source of fresh uninfected RBCs, blood drawn from type A Caucasian male donors and stored in CPD (citrate-phosphate-dextrose) at 4°C for up to 8 weeks was washed three times in incomplete medium and then made up to a 201o haematocrit in complete medium wiLh type AB serum. The two suspensions of RBCs so prepared (one containing the cells infected with P.falciparum, the other containing the uniSl- fected cells) were mixed together in a ratio of 1: 10 by volume. 10-ml portions of this final mixture were dispensed into 30-ml tissue culture flasksb and gassed for 10 seconds with a mixture of 3% CO2: 7% 02: 90% N2 flowing at a rate of 3 litres/min. The flasks were tightly sealed and placed on their broad sides in an incubator at 37°C. Medium was replaced daily with 10 ml of fresh complete medium and the flasks gassed as before. Sub-lines from established cultures were made as follows. After routine replacement of culture medium, as described above, the cells in culture were evenly resuspended; a small portion of the cell sus- pension (generally 0.5-2 ml) was transferred into a new culture flask containing 8-10 ml of a 2%o suspen- sion of fresh RBCs in complete medium (prepared as described above). The proportion of cells added from a donor culture was usually chosen so as to acbieve a starting parasitaemia in the daughter culture of about b Corning, 25 cm2, polystyrene. 25100. 0.1 'fo. After two such subcultures had been made, the parasites were routinely maintained using medium prepared with serum from type A Rh+ donors. Thin blood films were made daily from each cul- ture, fixed with methanol and stained with Giemsa stain. Description and clsification ofsexual and asexual parasites in culture We have classified the parasites under eight dif- ferent stage headings: ring forms, trophozoites, schizonts, and gametocytes divided into stages I-V as first proposed by Hawking et al. (12). The parasites were observed under bright-field illumination; the distribution and shape of pigment granules, an im- portant character in stage identification, was also ob- served using polarized light. Representative parasites of each of these stages are illustrated in Fig. 1. We have dermed the stages as seen in continuous culture, with emphasis on points of distinction between similar stages, as follows: (a) Ringforms (Fig. 1, 1-3). The ring form persists from the time of merozoite invasion until it becomes possible to distinguish between asexual trophozoites and stage 1 gametocytes. Ring forms are characterized by a clearly recognizable central "vacuole"C sur- rounded by blue cytoplasm with a single or double dot of chromatin. At this stage the parasites vary in shape from forms represented by a fine, almost hair-like circle of cytoplasm, to parasites with thicker, more extensive, often irregular, and almost amoeboid cyto- plasm. Ring forms are without pigment, with the ex- ception of the late forms which may contain a small granule. These parasites reach diameters slightly ex- ceeding half that of a normal RBC. (b) Trophozoites (Fig. 1, 4-6). The trophozoite is the earliest identifiable stage in asexual development. The cytoplasm of a trophozoite stains a strong blue and may be amoeboid or slightly vacuolated but no longer has a clearly identifiable central vacuole. A single nuclear region is present, staining a deep crim- son. The pigment is characteristic; it appears in the form of a single gold-brown nugget that increases in size as the parasite grows. (c) Schizonts (not illustrated). These stages extend from the first nuclear division in the trophozoite up to the mature segmenter separated into individual mero- zoites. A detailed description of the schizonts is not relevant here. The single nugget-like collection of pigment granules remains a distinctive feature of the asexual parasite through to the completion of schizogony. c lt has beca demonstrated by transmisston electron microscopy that this is not a true vacuole (1). Nevertheless, because of its appear- ance on blood filrms the term has been retained. GAMETOCT0GENESIS OF P. FALCIPARL'M IN CULTURE (d) Stage I ganmetocytes (Fig. 1, 7-9). This is the earliest stage at which a gametocyte may be dis- tinguished from an asexual trophozoite. The stage I gametocyte is a small rounded parasite i-+ of the diameter of an RBC. The parasite has a lightly stain- ing or almost clear cytoplasm and is never vacuolated or amoeboid. The nuclear area is not so distinct as in the trophozoite and stains more faintly. The distri- bution of pigment granules in stage I gametocytes most clearly distinguishes them from trophozoiEes. Rather than the single nugget seen in trophozoites, the pigment in stage I gametocytes is distributed in the form of small needle-like granules which increase in number as the parasite grows. The granules extend over a large proportion of the parasite and tend to be parallel to one another. (e) Stage II gametocytes (Fig. 1, 10-12). The gametocyte reaches this stage when one side becomes extended giving the parasite an overall teardrop shape (Fig. 1, 10). Elongation along one side continues as the parasite progresses to a form that has the appear- ance of a half moon; one side is smoothly rounded, the other straight, and the two ends are sharply pointed (Fig. 1, 11 and 12). At this stage the parasite occupies up to half of the area of the RBC which is by now partially depleted of haemoglobin, but is not dis- torted. This represents the upper limit of development of the stage 11 gametocytes. The cytoplasm remains pale, as does the nucleus which is somewhat spread out. Pigment formation is almost completed, the granules tending to spread out along and parallel to the axis of elongation of the parasite. (f) Stage III gametocytes (Fig. 1, 13-15). This stage begins when the pointed ends of the stage II gametocyte begin to become bluntly rounded. If still visible, the RBC becomes markedly distorted along the axis of the steadily lengthening parasite; the host cell may, however, be no longer visible on stained smears, being by now almost fully dehaemoglobi- nized. As stage Ill development continues the discrep- ancy in the curvature of the two sides is lost, both adopting a moderate convex or almost straight aspect. The parasite now has the appearance of a lozenge. No further pigment formation will take place; the gran- ules remain spread out along the length of the para- site. Other features of cytoplasm and nucleus are similar to those seen in the stage II gametocyEe. (g) Stage IV gametocytes (Fig. 1, 16 and 17). The parasite continues to grow along the direction of its axis and the two ends may become once again rather pointed (as the two sides acquire a simnilar gentle con- vex curve). Pigment granules are still widely spread over the length of the parasite, which remains un- curved or bent. The cytoplasmic and nuclear areas may begin to stain rather more intensely than at the earlier stages. Distinction between the sexes remains difficult. In stage IV, the gametocytes achieve the full size of the mature parasites. (h) Stage V gametocytes (Fig. 1, 18 and 19). This stage represents the point at which the gametocytes reach full morphological maturity and the male and female gametocytes can be readily distinguished. In both sexes maturation is accompanied by rounding of the ends of the gametocyte, and the development of a more oi less marked axial curve to the body of the gametocyte (accounting for the common practice of referring to the mature gametocytes of P.falciparum as cresccnts). Centralization of both chromatin and pigment granules occurs in both sexes but to different degrees. The cytoplasm acquires a more deeply stain- ing quality than at any earlier stage. The RBC envel- ope, virtually invisible on stained blood smears from the late stage III gamerocyLe onwards. in fact persists around the mature gametocyte and may be recognized by phase contrast or interference microscopy as a "bib" extending across the concave side of the para- site. Features distinguishing the sexes are as follows. The mature macrogametocyte (female) is characterized by a blue-staining cytoplastn and is more elongated than the microgametocyte (male). The chromatin and pig- ment granules of the female are closely aggregated at the centre of the parasite. The pigment frequently appears so densely clustered that it almost obliterates the chromatin from view. Other areas of the parasite's cytoplasm are virtually featureless. Because of the pink-staining quality of their cytoplasm, the distinc- tion between the cytoplasm and the nucleus is often vague in the microgametocytes. Both chromatin and pigment granules, while tending to centralize, are nevertheless more spread out than in the female. While the elongated macrogametocyte frequently ap- pears to be folded upon itself as though cut out of paper (not illustrated here), the microgametocyte has a more stumpy appearance, the axial curve is less marked, and the parasite is never folded. Macro- gametocytes usually predominate over microgameto- cytes in our cultures in a ratio of at least 4 to I but sometmes by a much greater margin. For comparison with the culcure forms, Fig. I also shows mature gametocytes ofPRfakciparum as seen in the peripheral blood of a human subject (painted by the same artist and reproduced from the Primate malarias (8) by kind permission of M. Warren). Degenerate parasites in culture Observations on degenerate asexual parasites formed the basis of our assessment of the state of health of cultures. The descriptions and illustrations presented above apply to typical healthy specimens of parasites in culture. Many variations of each stage 39 R. CARTER & L. H. NMlLLER may be encountered, most of which can be recognized without difficulty to conform to one or other of the stages categorized. Nevertheless the forms that we recognize as unhealthy or degenerate parasites war- rant special mention. Such parasites tend to be mis- shapen, often condensed, internally disorganized, and to have pyknotic nuclei, either excessively dark staining or faded cytoplasm, and, in gametocytes, clumped pigment. A more detailed account of the de- generate types of the various stages is as follows. Schizonts provide the best indication of the state of health of a culture. The typical healthy schizont is a rounded object of about the size ofa normalRBC and packed with clearly stained merozoites, 12-20 being typical numbers. In less healthy cultures the schizonts appear first as misshapen, elongated, or bean-shaped objects with fewer merozoites. As the state of health of the culture declines further the schizonts appear as smaller bodies with fewer and often smaller mero- zoites with faint cytoplasm and pyknotic nuclei. In their least healthy but still recognizable state the schizonts are typically represented by 2 or 3 sick- looking merozoites. The trophozoites undergo equally marked degener- ative changes in progressively less healthy cultures, becoming darker and smaller, and lose their clear structure. The ring stages are generally the least affected by the health of a culture except under the worst conditions. They appear fainter and less amoeboid, and the proportion of them decreases relative to the trophozoites and schizonts. As gametocytes mature in culture the percentage of degenerate forms increases. Stage V gametocytes are usually degenerate and appear as dark, lumpy, con- stricted, or broken parasites with pigment often dis- persed in irregular clumps. Stage IV degenerate forms are dark-staining and excessively narrow and pointed, spindle-like objects. Degenerate forms of stages II and III also tcuid to be darker, narrower, and more pointed forms. Degenerate stage I gametocytes pre- sumably occur; we are unable to recognize them, however, and assume that they resemble degenerate trophozoites. We have also noted the occurence of Garnham bodies (11, 19) in association with stage It and III gametocytes. These are darkly staining, rather thick, thread-like bodies extending in curved or folded lines into the cytoplasm of the host cell. They usually orig- inate from one of the ends of the gametocyte and often extend along the straight side. Occasionally, they appear to be free in the cytoplasm of the host cell. Jensen (14) reported failure to observe these bodies in gametocytes ofP.fakciparum in culture. It is interesting to note that virtually all of the degenerate and atypical forms of P.fakciparum gametocytes that we have observed in culture have been described among the gametocytes seen in humans infected with this parasite (11, 19, 10). Estimation of the rate of conversion ofasexual parasites to gametocytes A certain percentage of the ring forms that appear in culture develop into gametocytes; this percentage represents the rate of conversion of asexual parasites to garnetocytes. We have estimated the conversion rate from the number of stage II gametocytes that appear in culture 2 days after the appearance of the ring forms from which they developed. The conver- sion rate can thus be expressed by the following formula: rate of conversion of asexual parasites to gametocytes no. of stage II gametocytes - per 100 RBCs counted 48 h after nng forms x 100 no. of ring forms per 100 RBCs The conversion rate refers to the entire 24-h period im- mediately preceding the taking of the blood smear for the ring-form count. Our calculations depend upon the assumption that we can determine from a single blood film the number of rings and stage II gametocytes appearing in culture during the previous 24 h. When, as we have assumed for each of these stages, the residence time (seeFig. 2A for definition) is 24 h, the number of parasites of either stage counted on a blood smear is equal to the number produced during the 24 h preceding the time at which the slide was taken (Fig. 2B). If the residence time of either parasite stage were shorter or longer than 24 h then the number of parasites counted would be proportionately smaller or larger than the number actually produced during the previous 24 h (Fig. 2C and 2D). These principles apply directly to asynchronous cultures regardless of the time at which a blood film is taken. In synchronous cultures, on the other hand, the number of parasites on a blood film will vary de- pending on the relationship between the time the slide was taken and the time the parasites of the relevant stage appeared in the culture (Fig. 2C and 2D). This qualification does not apply, however, when the stage residence time is equal to 24 h; in such a case the number of parasites of that stage counted on a slide is always equal to the number produced in the preceding 24 h whether the culture is synchronous or asyn- chronous (Fig. 2B). We think that the residence time for ring forms (as we defixne them) is close to 24 h; the residence time for stage II gametocytes, however, is probably greater than 24 h but less than 48 h. Our data and those of others (18, 12, 14) are consistent with an interval of about 2 days between the formation of ring forms and the appearance of stage II gametocytes. 40 RI.TROPHOZOITES .': !'. . = ; . S '.t .,' .: ,' ^;, ,,:, '. C ........... ', , --; x; :-. :.:t. t.- :,; ,:.: ::.. .:-::; ..-S sj v*:*;s !;C^' C ' .s,', .+.lr .- ;s S. S . - ......... ox.\e ..''. 4. 5: 6 GAMETOCYTES .STAGE 11 7 8 *9 t 11 12 STAGE III. 13 .14 15 STAGE V a., 18 *STAGE IV 17 MATURE.-GAMETOCYTES. IN PERIPHERALSBLOOD oI .19 .20 21 4 22 *23 Fig. 1. Stages of Pfasmodium faiciparum seen in continuous culture (1 to 19). Maturegametocytes of P. fa/ciparum asseen in the peripheral blood of infected humans (20 to 23) are shown for comparison with the culture forms. itt.. 4 21 3 .STAGE I *RIN.GS. 16 .-.. GAMETOCYTOGENESIS OF P. FALCIPARUM 1N CULTURE B RESIDENCE TIME 24 HOURS 12 STAGES PER 24 HOURS IASYWCrEOUS 2 SYNCHRONO a12 0 1224 DRESIDENCE TIME 36 HOURS 12 STAGES OER 24 HOURS I ASYNOIlDNOUIS 2 SYNCHRONCLIS ==1 I . r-. I 12,2 HOR Fig. 2. The effect of stage residence time on the estimation, from a single blood smear, of the number of parasites of that stage formed over the preceding 24 h period. Initiation ofgametogenesis (exflagellation) from cultured gamelocytes Exflagellation of mature gametocytes from our cul- tures was initiated using the method described by Carter & Beach (6). Cells were resuspended in 10 volumes of fetal bovine serum (FBS) adjusted Lo pH 8.0 with 1.5% NaHCO3. The cells were centrifuged for 30 seconds at 1000 g and resuspended in a mini- mum volume of fresh FBS, pH 8.0, with NaHCO3. A drop of this preparation was placed on a slide and examined under a microscope at 400 x magnification with phase-contrast illumination. RESULTS Growth ofparasites in culture We followed the course of the parasitaemia in six parallel cultures of P.falciparum isolate G2 and for these cultures calculated the daily rate of conversion to gametocytes (Fig. 3). Parasitaemia rose rapidly on the first day from 0.8% to about 2.5%7o reaching a peak of about 6% on day 4. Up to the third day the parasites appeared healthy; however, from day 4 on- wards the proportion of degenerate forms in culture increased steadily. (See "Materials and Methods" for a description of these forms.) The overall para- sitaemia fell gradually to a mlnimum of about4% on day 7, 1 day after the asexual parasites appeared least healthy. Thereafter the health of the cultures im- proved but never recovered to the level seen in the first few days. Despite the presence of degenerate forms in culture, the total parasitaemia increased steadily, reaching a peak, with a mean of about 16% on days 12 and 13. Thereafter, the numbers declined slowly to less than 1007o by day 18. By that time the haema- tocrits in all cultures had fallen considerably below their initial values of 2%o, being then of the order of one-tenth of that value. The concentration of ring forms followed the general pattern of the total parasitaemia but varied in certain characteristic ways. The proportion of ring forms present was highest when the cultures were healthiest and lowest when they were least healthy, this pattern being reflected also in the changing growth rates of the cultures. The pattern of appearance and disappearance of the three groups of gametocytes-stage I, stage II, and combined stages III and IV-rnay be followed in Fig. 3A. Values for stage V gametocytes are not in- cluded; these wvere never present in large numbers during this study. The following features of these curves are worth emphasizing: (a) The moderate levels of stage I gametocytes present at the initiation of the cultures began to fall immediately towards zero during the flrst few days in culture, and rose again only after day 6. (b) In contrast, stage 11 gametocytes experienced a brief rise to a peak on day 3 of culture before they also fell towards zero on day 7. The resurgence of stage II gametocytes occurred 1 day after that of the stage I gametocytes. (c) Stage III and IV gametocytes rose from their initial near zero level 1 or 2 days after the early rise in stage II gametocytes. These stages did not experience the fall seen in the earlier stages between day 3 and day 7. This can be attributed to the much longer com- bined residence time for gametocytes of stages III and IV of the order of4-5 days. The numbers of stages II1 and [V were thus steady between day 6 and day 12, followed by a second rise on day 13 in response to the second rise of stage II gametocytes 5 days previously. Rate ofcon version to ganmetocytes Using the approach described in "Materials and Methods", we calculated the rate of conversion to gametocytes from data for ring forms and stage II A DEFINITION OF RESIDENCE TIME - Ra NCs. Te'ro Timl l...lnoat T ol FomuRm. OiDelobpnre. of Step *rio Nero Slag. C RESIDENCE TIME. 12 HOURS 12 STAGES PER 24 HOUR}S 1 ASYNCINOLUS _4 a 12 24 HOURS 2 SYNCHRONTOUS 1-- I * 12 24 HOURS 43 0 12 24 HCURS R CARTER & L. H. MILLER U' 1011 /- ~050 - u.S 0200.- cc Gamelmacym S53tuwIoo u 010 _ 05 z aOtc !G 5002- 0001 00000.0020 t- ) so.o0 B. late of Coresion to Gametocytes 0 0 e. 10.0 U) 18 20 On. 10 U. 05 m 03 02. 0.1 00 0 1 2 3 4 5 67 8 9 10 11 12 1314 15 1617 18 19 20 TIME (Do") Fig. 3. Parasite counts and gametocyte conversion rates in Plasmodium falciparum cultures G2/ 11. In order to determine the rate of conversion to gametocytes in P. falciparum in culture we followed the course of parasit- aemias in six parallel cultures of isolate G2. The parasite counts from the cultures followed over 20 consecutive days are rep- resented in Figure 3A. Values forthe six cultures have been combined for each dayand expressed as ± one standard deviation on either side of the means (in the present instance this corresponds closely to 2 standard errors about the means and thus closely approximates the 959% confidence limits forthe mean values). Total asexual parasite counts represent the sum of the ring, trophozoite and schizont counts. Although a variable proportion of rings are obviously incipient gametocytes this pro- portion is generally small compared to the number of manifest asexual parasites (trophozoites and schizonts). We have ignored the contribution of this component in computing counts for total asexual parasites. Using the approach described in "Materialsand Methods" the ratesof conversionto gametocytes were calculated fromthe data for rings and stage 11 gametocytes. Conversion rates were calculated separately for each of the six cultures. The con- version rates have been combined for the six cultures and expressed as ± one standard deviation on either side of the mean value for the six cultures IFig. 3B). As before, this closely approximates the 95% confidence limits forthe value of the mean. 44 GAMETOCYTOGENESIS OF P. FALCIPARUM IN CULTURE gametocyres (Fig. 3B). The conversion rates that were initially of the order of 5°7 fell towards zero by day 4-5 in culture and then rose rapidly on day 6 and day 7 and remained steady thereafter fluctuating be- tween 5% and 20%. These results confirm impressions gained during previous experience with cultures that the process of dilution with fresh cells at the time of the establish- ment of new cultures, thus lowering the parasitaemia, was responsible for reducing gametocyte production, while culture for about 1 week somehow led to re- newed gametocyte production. This hypothesis was 200. A. Total Asexual Para cn 10 0 U) 5.0 20 0 Parental Cultures t t 1.0 a D 0.5 _ z 0.3 02 0.1 -e 50 0 B. Rate of Conversior U 30.0 3 20.0- 0o Pareo-e C- ture:s rCC 10.0\ L O - DX0E 3.0 / 0. cc LU cn 0.3- D 0.2- more rigorously tested by subculturing shortly after the cultures had returned to maximum conversion rates (Fig. 4). Small samples (1 ml of culture suspen- sion) from four of the six G2/1 I cultures were diluted on day 8 to establish four subcultures with fresh cells and fresh medium in new flasks. As seen in Fig. 4A, the total parasitaemia and concentration of ring forms followed courses similar to those of the parental cultures at the time of their establishment. The pro- gression fromn healthy to unhealthy parasites in culture was also similar and occurred at corresponding times in parental and daughter cultures. Most importantly, 0 1 2 3 4 5 6 7 8 9 10 1 TIME lDays) Fig. 4. Parasite counts and rates of conversion to gametocytes in Plesmcodium falciparum cultures G2/1 1 and in daughter cul- turesdilutedfrom them. As in Fig. 3 al valuesare expressed = onestandard deviation about the meanswhich, in this case of4 daughter cultures, represent the 95% confidence limits for the means. 45 R. CARTER & L. H. MILLER however, the expectation that gametocyte production would be reduced following dilution of the subcul- tures was fulfilled (Fig. 4B). Moreover, gametocyte production was restored about 7 days after the dilu- tion, as had been the case in the parental cultures. Throughout this time conversion rates in the parental cultures remained at a steady level of 5%-20%. Similar results have been obtained with other lines of G2 (Fig. 5), with the Z isolate (Fig. 6), and with lines of GI (results not shown). One possible explanation for the fall and sub- sequent rise in conversion rates following subculture was that the freshly added RBCs required a short time in culture in order to become capable of inducing - A. Asexual Parasite Counts Total Asexual Parasites <t>/8 Rings X Total Diluted Asexual Culture Parasites Rings - -I J B. 1 Rate of Conversion to Gametocytes Parental Culture I L I I ,I*' /.'I . I. >tI\ I Diluted I Cufture . I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 TIME (Days) Fig. 5. Parasite counts and rates of conversion to gametocytes in Plasma dium falciparum culture G2/ 10. Culture G2/10 was treated and studied in a manner exactly analogous to that described for G21 1. The data presented for G2/10, however, are derived from the results from a single parental and a single daughter culture flask. 20.0 10.01 (n w U, < L) < cc O- 80LL8L w X 5.0 3.0 2.0 1.0 0.5 0.3 0.2 0.1 0.0 50.0 a c 30.0 ° 20.0 0 0- 10.0 en X) >_ (D 5.0 o 30o LU, 2.0 < UJ ID 0- 1.0 UL0 lr 0.5 w m 0.3 Z 0.2 0.1 L 0.00 p -I 46 I GAMETOCYTOGENESIS OF P. FALCIPARUM IN CULTURE gametocyte formation. This possibility was tested by pre-incubating batches of uninfected RBCs at 37°C for I week (with regular medium changes and gassing procedures as for normal cultures) with either fresh culture medium or culture medium recovered after 24 h over a growing culture of P.falciparum. At the end of the week 3 pairs of cultures of G2 were estab- lished from the two pre-incubated batches of RBCs and a control batch was kept in citrate-phosphate- cn LU en U)I- < ra- LL80o- a:= wL LU z w 0 0 0 cc . CtL en u LL U F- CD a oz ~-8e u LJ m z dextrose solution (CPD) at 4°C during this period. The results for the rate of conversion to gameto- cytes for the three sets of cultures (Fig. 7: pre-incu- bation in (a) fresh medium; (b) recovered medium; (c) CPD at 4°C) show no significant differences between them. Clearly the aging of RBCs at 37 °C with either fresh or once used medium had no effect on their ability to induce or support gametocyto- genesis. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 TIME (Days) Fig. 6. Parasite counts and rates of conversion to garmetocytes in Plasmodium falciparum culture Z182. Culture Z/82 was treated and studied as described for cultures G2/11 and G2f10in Fig. 4 and 5. As for G2/10the results are derived from a single parental and a single daughter culture flask. For unknown reasonsthe overall growth ofZ/82was very slow during the firstweek in culture (Fig. 6A). Parasite countswerethus too low to measurethe earlyconversion rates (Fig. 6B). Thus, therise in conversion rate is apparent in the parental culture but notthe initial fall. It is of interestto notethat in this slow-growing cul- ture the rise in conversion rate is delayed by several days compared with the time of rise in the fast-growing G2 cultures. Fol- lowing dilution ofthe daughter culture thetypical fall followed by a rise in rate of conversion to gametocytes is clearlyevident (Fig. 68). In this case, moreover, the daughter culture was fast growing and Ihe rise in conversion rate occurred soonerthan it had done in the parental culture. 47 R. CARTER & L. H. MILLER 50.0- Exflagellation of gametocytes from culture coR Exflagellation has been induced only when healthy z * i? ,* stage V gametocytes have been observed in culture. 20.O - In the Z isolate we have observed exflagellation in 8 .1 \ \ four separate cultures over a 1 year period (Fig. 8). _ I</ \ \ Clearly, continuous culture over this period did not fIL 10.0 0/ \\ \ diminish either the ability of the Z isolate to produce D yX \n\].gametocytes or their capacity to reach full maturation o) 5.0 A 1l on t2 (undergo exflagellation).Of. ; v /1/ \\A consistent feature of the production of such 3.0 mature gametocytes has been a fall in haematocrit toXL BX t?l!very low levels (<0.5%) prior to full maturation of 2.0 1 the gametocytes and the maintenance of cultures for ¢ \>S1/1 2-3 weeks without addition of fresh cells. In the light o 1.0 \[' of the present results, this period could be interpreted L 10 tF as the time required for restoration of gametocyte E \YjtR production following establishment of a culture < 0.5 \1 J} (about 1 week) followed by the time required for ga- °L 0.3 L \9\ ,bimetocytes to reach full maturity (about 10 days) (17).0 03 - m 0.2 - Z \ Dtt DISCUSSION Z 0.1 . 1 2 3 4 6 7 8 9 10 11 12 Methodfor estimating the rate of conversion of asexual parasites to gametocytes TIME (Days) Trager & Jensen (19) were the first to show that gametocytogenesis (the formationl of gametocytesFig. 7. Rate of conversion to gametocytes in Plasmodium fromasexualstages) by P.faociparumn couldcontinue falciparum cultured in RBCs incubated at 37 C in used or f fresh culture medium for one week prior to introduction of for long periods in culture, thus enabhng this process parasites (Cultures G2/11). The curves are identified in the to be studied outside the uncontrollable environment text. of the host. In the present studies, we have developed Non-Immune A neaiuTo ns Inimted Wmhixm&n Bf Mosq*iao Me In zo 19411-77 lo7n WW77 SV-78 tzstrw *-_F _w - Ad Pbcad In Culwe At N I N , Fig8.. Isolation and summarized laboratory history of theZ isolate ofPlasmodium falciparum. After isolation from the periph- eral blood of the infected donor, the Z isolate was maintained in culture flasks with passage into new flasks and addition of fresh RBCs as indicated by the arrows. Arrows enclosed in brackets indicate unspecified numbers of successive passages. Arrows not so enclosed represent the establishment of flask cultures on the date indicated. Flasks from which exflageilation was obteined are indicated by the wavy lines, the date(s) on which exflagellation was induced beiing indicated. The last cultureflask indicated in this series, Z/82, isthe culture whose gametocyte production was studied for the present report. 48 GAMETOCYTOGENESIS OF P. FALCIPARUM IN CULTURE a method for estimating the rate of conversion of parasites from asexual reproduction to gametocyte production at any time during the course of a fluc- tuating parasitaemia of P.falciparum in culture. The conversion rates refer to the percentage of ring forms present in culture at a given time that will appear as stage Xl gametocytes 2 days later. We chose the stage 1L gametocyte as the basis for measurement of conver- sion rates for the following reasons. Firstly, it is an early stage in gametocyte development and therefore less subject than later stages to losses due to degener- ation or mortality; secondly, it is easily recognized. Stage I gametocytes were not used for general calcu- lations of conversion rates because of the difficulty, at times, of identifying these parasites. Our method for estimating the rate of conversion to gametocytes is based on certain assumptions: (a) ring forms and stage II gametocytes each persist for about 24 h from the time of their formation until their de- velopment into the next parasite stage; (b) there is an interval of about48 h between the formation of a ring form and its subsequent appearance as a stage II gametocyte; (c) the time intervals assumed in (a) and (b) do not vary; (d) ring forms committed to develop- ment into gametocytes always complete their develop- ment at least as far as stage II. Our first two postulates are assumptions about the specific time intervals in the development of the para- sites. The assumption of 48 h for the interval between ring formation and the appearance of the stage II gametocyte is consistent with our own data and those of others (11, 17). We believe that 24 h is a reasonable estimate for the duration of the ring stage as we define it. The stage II gametocyte, on the other hand, prob- ably persists for a period nearer to 36-48 h. By under- estimating this interval, as we may have done, our conversion rates at 24 h would be overestimaced by a factor of 1.5-2. The times aL which conversion rates appeared to rise and fall would not be affected, however. Our last two postulates are in effect a single assumption that the time intervals in the development of the parasites are constant, i.e., that the stages do not, in significant numbers, grow faster or slower than we have specified, or die under the different con- ditions of culture. This general assumption is certainly more likely to be valid for earlier stages, such as ring forms and stage I and IL gametocytes, than for later stages. Evidence in favour of our assumptions would be obtained if the same values for conversion rates were consistently found using data from two different gametocyte stages. This was found to be true to the extent that the rise and faUl of conversion rates fol- lowed exactly the same timing whether such values were calculated from data for stage I or stage 11 gametocytes (R. Carter & L. H. Miller, unpublished results, 1978). The values from stage I data were. nevertheless, consistently somewhat lower than those obtained from data for stage II; this presumably re- flects the constant error involved in underestimating the duration of the stage 1I gametocyte. Evidence for environmentally induced changes in the rate of conversion ofasexual parasites to ganrerocyres in culturre The results prese-nted in Fig. 3 demonstrate that gametocyte production was "turned off" in six paral- lel cultures of Pfalciparurn soon after subculture (addition of fresh RBCs) and "turned on" again after 6 days in culture. The fall in conversion rates early in the cultures and their subsequent rise after a lag period suggests that environmentl influences as- sociated with dilution of parasitized cells during sub- culture are responsible for suppressing conversion to gametocytes. Conversely, prolonged periods in un- diluted cultures appear to stimulate conversion to gametocytes. The direct association of the fall in conversion rate with dilution and subculture was verified by following conversion rates in daughter cultures in parallel with those in the parental cultures from which they were diluted (Fig. 4-6). While conversion rates remained high in parenEal cultures, all daughter cultures experi- enced a rapid drop in conversion rates to near zero followed by a return to parental levels after about l week. These experimenLs demonstrate that the rate of conversion to gametocytes by P.fakciparunm in culture is susceptible to large fluctuations under environ- mental influences. The environmental factors in- volved in mediatLing the changes in conversion rate are not knowvn. Preliminarv experiments suggest.that the effect is not associated witb aging of the RBCs in cul- ture. We believe that changes in the medium associ- ated with a period of growth in culture are probably directly responsible for increasing the conversion rate. Previous studies on variatiorns in gamerocyle production Fluctuations in the rate of production of gameto- cytes from asexual parasites appear to occur in vivo as well as in vitro. The relative importance of environ- ment and innate parasitological factors in modu- lating gametocytogenesis are difficult to separate in in vivo systems. Smalley (17) showed that the proportion of ring forms of P.falciparuwn that developed into gametocytes in naturally infected Gambians varied at different times in an infection. The results of Shute & 49 R. CARTER & L. H. MILLER Maryon (16), who studied drug-controlled infections of P.falciparum in nonimmune subjects, showed that gametocyte production began comparatively early in blood-induced infections, but not until after recur- rence of symptoms in sporozoite-induced infections. Their results suggest that in sporozoite-induced infec- tions the parasites require a period of adjustment before gametocyte production can begin; in blood- induced infections, on the other hand, the parasites may begin gametocyte production almost immedi- ately. Convincing evidence that the immune status or age of the host play a direct role in controlling gameto- cytogenesis is, in our opinion, lacking. For instance, although it is well established that P.falc:parum gametocyte rates in areas of holoendemic malaria are higher in children than in adults (5, 15, 21), this may only reflect the overall higher parasitaemias in children. Indeed, Christophers (7) has shown that there is a direct correlation between gametocytaemna and parasitaemia in human populations infected with P.fakciparum. In addition to the short-term fluctuations in ga- metocyte production seen in the course of individual infections, or in the life of a single culture, malaria parasites may lose their ability to produce gameto- cytes during successive passages in culture (14) or in the blood of laboratory hosts (4, 20, 2). When it occurs such loss appears to be permanent. Reduction in gametocyte production after long periods of culture, or of serial blood passage, is by no means an invariable rule and varies greatly according to the species of parasite host and method of passage. We experienced no reduction in production of ga- metocytes or decline in their capacity to reach maturity (undergo exflagellation) with the Z isolate of Pfakciparum after 1I years in continuous culture. Indeed, one isolate, GI, appears to have increased its level of gametocyte production during 6 months in culture from an almost negligible level to one compar- able to those seen with the Z or02 isolates (R. Carter, unpublished results, 1978). Increased capacity for gametocytogenesis has also been associated with ac- quisition of drug resistance. Thus, Bishop (3) found that lines of P. gallinaceum selected for resistance to sulfadiazine showed marked increases in gametocyte production compared with parasites of the parental line. It would be useful to be able to measure intrinsic (genetic) differences in capacity for gametocyto- genesis between individual isolates or lines of P.fal- ciparum. However, it is clearly not a simple matter to standardize such measurements. We have seen that the laboratory history and environmental conditions in culture can have large effects on the rate of conversion to gametocytes. Hypothetical mechanismsfor gametocytogenesis of P. falciparum We have argued that the rate of gametocytogenesis is not constant during the life of a single culture but is subject to large fluctuations under the influence of environmental conditions. This implies that there is a period in the cycle of development of the asexual parasites when they can be induced by environmental conditions to become commnitted to gametocyto- genesis. In Fig. 9 we propose three models for gameto- cytogenesis based on different times of susceptibility to commitment and different times and patterns of manifestation of that comrmitment. Model I describes a situation in which a merozoite or a ring form is susceptible to influences that deter- mine its development either into an asexual parasite or into a gametocyte. This model assumes that the ring form is uncommitted (neutral) and may develop in either direction. Models II and III describe two distinct types of situ- ation which share the common characteristic that the commitment is made during one asexual generation but is only manifest in the parasites of the next gener- ation. In Model l1 the commitment in one generation detemines whether the progeny will develop entirely into gametocytes or entirely into asexual parasites. In Model I1I the commitment in one generation is a graded one leading to different degrees of probability that individual merozoites within a schizont will become gametocytes in the next generation. This model differs from Model II in that only a portion of the merozoites of any schizont are likely to become gametocytes. Models II and III share the property that commit- ment could be made, at least in pnnciple, at any time from the earliest stage after invasion until merozoite formation in the generation before the commitment is manifest. In Model I, on the other hand, commitment must be made somewhere in the briefer period that includes merozoite invasion and ring development. Of the three possibilities, we favour Model II pri- marily because of the frequency with which multiple gametocyte infections are observed in culture (14, R. Carter unpublished observations, 1978). This fact supports the model for the following reasons. When a schizont ruptures in stationary culture, surrounding RBCs may become infected by several merozoites from the same schizont. RBCs with multiple ring in- fections are, in fact, commonly encountered. If all merozoites from a schizont were committed to either asexual development or gametocytogenesis, then RBCs containing either multiple asexual parasites or two or more gametocytes should be observed. Indeed, such RBCs, doubly and rarely triply infected with gametocytes, are not uncommonly encountered. so GCAMETOCYTOGENESIS OF P. FALCIPARUM IN CULTURE SUSCEP- ,TBLE IH- IMANIFESTI MODEL .(?1 TROPIOZOITE SCHIZONT RING. (NEUTRAL) u STAGESI STAGE II GAMETOCYTE MANIFEST I ---, SUSCEPTIBLE F________ ~~~~____ _ _ MODEL ASEXUAL RING SCHIZONT 10 10 10 W 1 ASEXUAL TROPHOZOITES SCHIZONTS(3~~~~) ~~RINGS *TROPHOZOITE qS; SCHIZONT10 10. 10SCHIZONT SEXUAL STAGE I STAGE II RINGS GAMETOCYTES SUSCEPTIBLE1- - --- - "--_______ _ _-_1_ MODELD ASEXUAL RING 10-X TROPHOZOITE SCiuZONT Oe X SEXUAL 'RINGS I MANIFEST an Il.-x 0- X T o-X CO-X TROPHOZOtTES SCHIZONTS X X STAGE I STAGE 11 GAMETOCYTES Fig. 9. Models for gametocytogenesis. 51 041 52 R. CARTER & L. H. MILLER ACKNOWLEDGEMENTS We thank Dr David Aling for helping us develop the concepts involved in measuring gametocyte production. We are also greatly indebted toMs Carol Schmidt and Ms Patricia Sigmon for discussion and assistance and to Mrs Gertrude Nicholson who made the illustrations of malarLa parasites. RESUME MISE EN EVIDENCE DES RELATIONS ENTRE LES VARIATIONS DU MILIEU ET LA GAMETOCYTOGENASE DANS UNE CULTURE CONTINUE DE PLASMODIUM FALCiPARUM La mise au point de la culture continue de Plasmodium fakciparum a rendu possible 1'6tude des facteurs n6cessaires A la production de gamEtocytes in vitro et Il'limination des variables incontr6lables qui existent chez Ih6te dans la pro- duction in vivo. Les auteurs ont r6ussi A d6terminer de mani&re precise le taux de production des gam6tocytes A tout moment dans les cultures en applhquant une m6thode fondEe sur l'estimation du pourcentage des tormes annulaires dont le d6veloppement aboutira A des gam6tocytes du stade ll, lequel est atteint dans tous les cas oti ces formes donnent naissance i des gametocytes. Les formes annulaires, comme les gam6tocytes du stade II, sont en effet ais6ment ident'- filables et persistent pendant quelque 24 heures. Une chute rapide du taux de conversion, qui est d'environ 50t* A l'origine, a &t constat&e a la suite de la dilution des cultures par I'adjonction d'trythrocytes frais en vue d'abais- ser la parasit6mie. Dans les nouvelLes cultures, les taux remontaient jusqu'A un niveau de 5 A 200/o apr6s quelques jours. Dans les cultures meres qui avaient 6th utilis6es pour les dilutions, les taux de conversion sont demeur6s eleves A tout moment. Ce tableau a R6 constamment observe pour trois isolats diff6rents de P. fakipurum d'Afrique, et on peut donc conclure que la r&duction de la parasit6mie cons6- cutive A I'adionction de cellules frakches est A l'origine de la r6duction de la production de gamttocytes; il semble d'autre part que les modifications du milieu intervenant dans une pkriode de croissance des cultures soient directement asso- ci6es A une nouvelle hausse du taux de conversion, Les auteurs en deduisent que les conditions de milieu ont une in- fluence directe sur le taux de production des gam6tocytes de P.faciparum cultiv6 in vitro. Apr6s un an et demi de culture, les parasites avaient conserve leur capacit& de produire des gam6tocytes et les gam6tocytes cdlle de liberer des flagella. REFERENCES 1. AIKAWA, M. Experimental parasitology, 30: 284-320 (1971). 2. BAFORT, J. ET AL Nature (London), 208: 1230-1231 (1965). 3. BisHoP, A. Parasitology, 45: 163-185 (1955). 4. BOYD, M. F. American journal of tropical medicine, 25: 293-306 (1945). 5. BRuCE-CiwATrr, L. J. Bulletin of the World Health Organization, 4: 301-327 (1951). 6. CARTER, R. & BEACH, R. F. Nature (London), 270: 240-241 (1977). 7. CHRISTOPHERS, S. R. Indian journal of medical re- search, 12: 273-294 (1924). 8. COATNEY, G. R. ET AL Theprimate malarias, Washing- ton, US Department of Health, Education, and Wel- fare, 1971. 9. DIGGS, C. L. ET AL Amercanjournal oftroprcal medi- cine and hygiene, 24: 760-765 (1975). 10. FIELD, J. W.& SHUTE, P. . The microscopic diagnosis ofhuman malaria. II. A morphologicalstudy of eryth- rocytic parasites. Kuala Lumpur, Government Press, 1956. 11. GARNHAM, P. C. C. Transactions of the Royal Society of Tropical Medicine and Hygiene, 26: 401-403 (1933). 12. HAWKING, F. ET AL Transactions of the Royal Society of Tropical Medicine and Hygiene, 65: 549-559 (1971). 13. HAYNES, J. D. L:T AL. Nature (London), 263: 767-769 (1976). 14. JENSEN, J. B. Journal of protozoology, 26: 129-132 (1979). 15. MUIRHEAD-TiiOMSON, R. C. Transactions of the Royal Society of Tropical Medicine and Hygiene, 48: 208-225 (1954). 16. SHUTE, P. G. & MARYON, M. Transactions of the Royal Society of Tropical Medicine and Hygiene, 44: 421-438 (1951). 17. SMALLEY, M. E. Nature (London), 264: 271-272 (1976). 18. THOMSON, J. G & ROBERTSON, A. Transactions of the Royal Society of Tropical Medicine and Hygiene, 29: 3140 (1935). 19. TRAGER, W. & JENSEN, J. B. Science, 193: 673-675 (1976). 20. VINCKE, I. H. ET AL Annales de la Socierd beige de M&decine tropicale, 33- 269-282 (1953). 21. WiLsoN, D. B. Transactions of the Royal Society of Tropical Medicine and Hygiene, 29: 583-618 (1936).