Experiments were performed using approved protocols in accordance with the UK. Animals (Scientific Procedures) Act 1986. Male CD mice were treated with phenylhydrazine two days prior to infection with either P. berghei ANKA (clone 2.34; wild-type), P. berghei CTRP-KO 14, or P. berghei SOAP-KO 12 by the inoculation of parasites obtained from a donor mouse between the 2^nd and 6^th passage from cryopreserved stock.

Blood from infected mice was collected by cardiac puncture into a heparinized syringe two days post-infection, after the confirmation of an infection by Giemsa staining of thin blood smears. Blood was diluted 1:10 into RPMI-1640 medium containing 10% foetal calf serum (FCS), 14.7 mM sodium bicarbonate, 0.367 mM hypoxanthine, 1000 U/ml penicillin, and 1 mg/ml streptomycin (pH 8.4) prior to incubation for 18 h at 19°C as detailed in 3334.

Ookinetes were harvested using a MidiMacs LS magnetic column as described previously 3235. Briefly, following centrifugation at 648 g for 5 min all but 20% of the medium was removed and the blood mixture was passed through a magnetic column. Following several column washes, ookinetes were eluted with 3 ml of oocyst medium (supplemented Schneider's insect cell culture medium).

Plasmodium berghei wild-type (WT), PbCTRP-KO or PbSOAP-KO ookinetes were immediately seeded into 8-well LabTek™ chamber slides at a density of approximately 1 × 10⁴ in 400 μl of oocyst medium (see 32 for medium details). For each experiment, ookinetes were seeded in chamber slides either previously coated with 80 μl Matrigel™ or with no coating. In order to minimize parasite loss from the wells, medium was not removed during culture. However, to counter the effect of evaporation, 100 μl of fresh medium was added to each well every 48 or 72 h post-culture. Triplicate wells were assessed at each time point and the experiment was repeated three times.

The presence of a transitional stage between ookinete and oocyst in P. berghei ANKA WT, called the took, was recently reported 32. To demonstrate clearly the presence of a took stage during the development of oocysts of the null mutants, cultures were collected 21 h post-culture in oocyst medium. Parasites were fixed in 1:1 methanol: acetone, washed and labelled with an anti-P28 monoclonal antibody derived from a hybridoma cell line maintained in the laboratory. Following a 30 min incubation the slides were washed and incubated for 30 min with a goat anti-mouse fluorescein isothiocyanate-conjugated secondary antibody 32. The slides were washed and mounted in VectaShield (Vector Laboratories) containing 4', 6-diamidino-2-phenyl indole (DAPI) before viewing samples with a Leica DM IRB Fluorescence Microscope (DAPI excitation/emission: 340–380 nm/425 nm; FITC: 450–490 nm/535 nm). Appropriate controls were included for each experiment.

Ookinetes have a propensity to clump together in culture, so well contents were mixed vigorously by pipetting to ensure even distribution within the well prior to counting ookinetes/oocysts. Measurements and counts were conducted on days 1, 3 and 5 post-culture for wells that were not coated and day 1 for wells coated in Matrigel. Parasites were counted in alternate fields of view along a cross representing two diameters of the well, at a magnification of 630×, viewed under phase contrast. In all, 30 fields of view were examined per well (3.58% of the whole well). In order to assess transformation of ookinetes to oocysts, the proportion of P. berghei WT, PbCTRP-KO or PbSOAP-KO ookinetes, tooks and oocysts was calculated for each culture condition at each time point. Only parasites that were fully rounded, translucent, with pigment visible and a diameter greater than 5 micrometers were regarded as having transformed into oocysts 32.

Parasites were collected by centrifugation for 1.5 min at 2,000 g. The pellet was resuspended in 5 μl of PBS to which 5 μl of 0.1% Erythrosin B was added to assess parasite viability. This was immediately spotted onto a microscope slide, a coverslip applied and a minimum of 100 oocysts and any remaining ookinetes and tooks were examined at magnification of 400× and scored as viable or having a compromised membrane. Three repeat experiments were performed in which triplicate wells were assessed on each of the days chosen for observation.

Unlike PbSOAP-KO, for which it has been reported that between 15–40% of ookinetes transform into oocysts in vivo 12, no PbCTRP-KO oocysts have been observed in vivo 141619. In order to confirm the lack of a role of CTRP in transformation, in vitro ookinete cultures were injected into the haemocoel of 4 day old An. gambiae KIL mosquitoes. The preparation and injection protocols were adapted from Paskewtiz and Shi 36. Briefly, PbCTRP-KO ookinetes were cultured for 18 h at 19°C in ookinete medium. Cultures were centrifuged at 648 g for 5 min at 4°C. The majority of red blood cells were removed by resuspending the culture pellet in 20 volumes of 0.17 M ammonium chloride and incubating for 10 min on ice. All subsequent steps were performed at 4°C. An equal volume of PBS was added to stop red blood cell lysis and cultures were immediately centrifuged. The pellet was washed once in 40 volumes of PBS, resuspended in 1 ml of PBS and counted on a haemocytometer.

Fifty ookinetes in 69 nl of PBS were injected per mosquito with a Nanoject II automatic nanoliter injector (Drummond Scientific Company, USA). A total of 50 mosquitoes were injected and maintained at 20°C for seven days prior to dissection. Mosquitoes were dissected in PBS and the contents of the abdomen were immediately stained in 0.05% mercurochrome. Tissue was viewed on a Leica DM IRB inverted microscope and images were acquired with a Leica DC 300F digital camera (Leica Microsystems, Germany).

All data sets were analysed using the statistical software package Minitab^® Release 14 (Minitab, Inc.). Where necessary, data were normalized with an arcsine transformation prior to analysis with a General Linear Model (GLM).

It has been proposed that SOAP and CTRP may play a role in ookinete-to-oocyst transformation in vivo as the null mutants for either gene produce very few, or no oocysts respectively 121419. Moreover, both of these molecules have been reported to bind to laminin and it has been proposed that interactions between SOAP and/or CTRP with the basal lamina, and laminin in particular, may trigger transformation 1230. However, it was found that when P. berghei CTRP- and SOAP-KO ookinetes were cultured in vitro, plus or minus Matrigel (a source of laminin), the same proportion transformed to oocysts as did the wild-type. Furthermore, transformation proceeded via a took stage (Figure 1) as recently described by Carter et al 32.

The percentage of transformed ookinetes (i.e. oocysts) was calculated on days 1, 3, and 5 post-culture after counting the number of ookinetes, tooks and oocysts per well. Over all conditions, a mean of 93% (SE ± 1.33) of ookinetes had transformed into oocysts by day 1. There was no significant difference in the percentage of ookinete to oocyst transformations between parasites lines (GLM, F_(2,1,2,2) = 1.64, P = 0.213) or the presence or absence of a basal lamina substitute (GLM, F_(2,1,2,2) = 0.60, P = 0.446) (Table 1). As the majority of parasites had transformed by day 1 and no differences were observed between culture conditions, data for day 3 and 5 was only collected from wells that were not coated with Matrigel. For all parasite genotypes, nearly 100% of ookinetes had transformed by day 5 (Table 1).

The large percentage of ookinete-to-oocyst conversion in these experiments was surprising as it has previously been reported that over 30% of ookinetes cultured in vitro are dying by an apoptosis-like mechanism by 24 h post-culture 37. This suggests that ookinetes harvested at 18 h post-culture and immediately placed in oocyst medium, which supports transformation and growth (see 32), may be rescued from triggers that induce cell death. It was previously observed that the proportion of ookinetes displaying markers typical of apoptotic cells was under 10% after 12 h in culture 37. Current findings confirm this observation, but also show that the proportion of dying or dead ookinetes varies considerably between cultures, possibly as a result of factors from the murine host (Hurd, Ali and Arambage, personal observations).

Of the 50 mosquitoes injected with P. berghei CTRP-KO ookinetes only four survived to day 7 post-injection. P. berghei CTRP-KO oocysts were observed in all four mosquitoes. Thus PbCTRP-KO ookinetes are able to transform into oocysts when injected into the haemocoel of 4-day old female An. gambiae KIL mosquitoes. Qualitative observations were made of the infections as it is difficult to quantify oocysts as they can develop anywhere in the haemocoel 36. Upon dissection the majority of oocysts recovered were observed to be floating in the haemolymph. However, oocysts were also found bound to the midgut (Figure 2) and the Malpighian tubules.

These findings confirmed suggestions that both PbCTRP and PbSOAP null mutants have retained their full potential to develop into oocysts. Both microneme proteins have been implicated in ookinete motility 121416192738 and findings presented here suggest that the barrier to further development in vivo must be caused by the necessity to migrate out of the blood bolus, traverse the peritrophic matrix and invade the midgut epithelium. Dessens et al 12 also observed the transformation of PbSOAP-KO parasites in vitro, but surprisingly, in their system ookinetes invaded and developed inside Aedes aegypti Mos20 cells. The axenic culture system used here also shows that cell invasion is not a prerequisite for transformation. Interestingly, Hirai et al 23 recently reported that disruption of the gene encoding PbGCβ also resulted in normal ookinetes that do not produce oocysts in vivo. Here too these transgenic ookinetes transformed into oocysts in an in vitro system based on Al-Olayan et al 37 and ookinete motility was shown to be defective.

Transformation rates of null mutants and wild-type P. berghei were the same in the presence or absence of Matrigel in this study (Table 1). These results suggest that the binding of either one of these micronemal proteins to basal lamina components is not a requirement for transformation. Indeed, the study also confirms that interactions between basal lamina components and any ookinete molecule are unnecessary to initiate conversion to oocysts 32. It has been suggested that the mosquito midgut basal lamina plays an essential role in the development of Plasmodium oocysts 39. Gene silencing technology resulting in the depletion of mosquito laminin greatly reduced oocyst numbers developing in vivo 39 and interactions between laminin and microneme and/or surface proteins support this suggestion 1229. If this idea is correct then, as it is clearly not involved in triggering transformation or supporting very early growth, it may perform a function as a protective coat, disguising the oocyst from the mosquito defence system 3940 or just serve to bind the ookinete in place between the midgut basolateral plasma membrane and the basal lamina, thus preventing the ookinetes from migrating into the haemocoel, as do most insect parasites.

The identification of the molecular composition of the mosquito midgut is rapidly advancing. The major components of the basal lamina, collagen IV and laminin, have been characterized in An. gambiae 4142, as has beta integrin, a receptor for laminin and collagen IV43. The latter has been implicated in ookinete invasion of the midgut and may have a role in facilitating putative interactions between ookinetes and laminin or vice versa 43. However, it now appears that these molecules are more likely to be involved in later phases of oocyst development such as growth, maturation, and sporozoites formation.

In addition to transforming into oocysts, PbCRTP-KO and PbSOAP-KO parasites were able to grow in culture at a comparable rate to P. berghei WT and, here too, no significant differences were found between the presence and absence of the basal lamina substitute (Figure 3). Over the five day period there were significant increases in parasite sizes. Tukey's pairwise comparison day 1 vs day 3 (T = 4.946, P < 0.001), and day 3 and 5 (T = 3.195, P < 0.01) (Figure 3). However a significant decrease in the number of viable oocysts between each of the time points was observed (GLM: F_(2,1,2,2) = 117.32, P < 0.001), but with no significant difference between parasite genotype (GLM: F_(2,1,2,2) = 0.70, P = 0.497) or the presence or absence of Matrigel (GLM: F_(2,1,2,2) = 1.85, P = 0.175) (Figure 4). Although not statistically significant, fewer parasites were viable when incubated with Matrigel at day 5. This could be caused by additional gentamycin, present in Matrigel as well as in the oocyst culture medium.