Regulation of Plasmodium sporozoite motility by formulation components

Mosquitoes were infected by feeding on infected mice (female OF1 mice, 6–7 weeks old; Charles River, Leiden, The Netherlands) as described previously [30]. The animal experiments of this study were performed in accordance with Dutch welfare regulations and were approved by the institutional Animal Ethics Committee of the Leiden University Medical Center (DEC PE.18.005.001). Plasmodium berghei spz (PbANKA-mCherry_hsp70 + Luc_eef1α; line RMgm-1320, http://​www.​pberghei.​eu) which contain the fusion gene mcherry under control of the strong hsp70 promoter integrated into the neutral 230p gene locus (PBANKA_0306000) were obtained by microsurgical dissection of the salivary glands of infected female Anopheles stephensi mosquitoes. To allow the spz to mature dissections were performed 21–24 days following infection of the mosquitoes. During this period mosquitoes were kept at a temperature of 21 °C and 80% humidity. The salivary glands were collected and crushed in the different formulations mentioned in the next section. The free spz were counted in a Bürker counting chamber using phase-contrast microscopy. On average, dissection yielded 55 k spz/mosquito. The mosquitoes and the spz were kept on ice (0–4 °C) for a maximum of 3 h until use.

To investigate which components are needed to induce motility of spz, the mosquitoes were dissected and the obtained salivary glands were crushed in eleven different formulations: (1) Phosphate buffered saline which contains potassium, sodium, chloride and phosphate (PBS; Life Technologies Inc.), (2) Hanks’ balanced salt solution which contains besides the salts present in PBS also glucose, calcium, magnesium, sulphate and bicarbonate (HBSS; Life Technologies Inc.), (3) Roswell Park Memorial Institute medium (RPMI; Life Technologies Inc.) which contains besides the components of HBSS also amino acids and vitamins, (4) PBS enriched with 3.5 mg/ml bovine serum albumin (BSA; Sigma-Aldrich), (5) HBSS + 3.5 mg/ml BSA, (6) RPMI + 3.5 mg/ml BSA, (7) PBS enriched with 10% fetal bovine serum (FBS; Life Technologies Inc.), (8) HBSS + 10% FBS, (9) RPMI + 10% FBS, (10) RPMI without amino acids (Caisson Labs) + 3.5 mg/ml BSA and (11) RPMI without glucose (Life Technologies Inc.) + 3 mg/ml BSA. All eleven formulations had a physiological pH of 7.0–7.5 and a viscosity of < 1.1 cP. For imaging of the spz, 10 µl of the spz solution was pipetted on the cover slip of a confocal dish without any precoating (ø14 mm; MatTek Corporation), covered with another cover slip (ø12 mm; VWR Avantor) and imaged within 45 min. The set-up is schematically depicted in Additional file 1: Fig. S1.

Images of the spz were taken on a Leica TCS (true confocal scanning) SP5 or SP8X WLL (white light laser) microscope (Leica Microsystems, Wetzlar). The spz expressed mCherry which was excited at 587 nm and the emission was collected between 600 and 650 nm. The movies were recorded with a frame rate of 35 frames per minute, 400 frames per movie, three movies per condition. The resolution of the images was set at 1024 × 1024 pixels. For imaging of the spz a 40× objective (Leica HCX PL APO CS 40×/1.25–0.75 na OIL) was used and the experiments were performed at room temperature, unless otherwise stated. The movies were recorded using the Leica software (LAS X version 1.1.0.12420; Leica Microsystems, Wetzlar).

Maximum projections of the recorded microscopy movies were generated using Fiji software [31]. The movies were further processed using SMOOT_{In vitro}, an in-house developed graphical user interface (GUI), written in the MATLAB programming environment (version r2017b, The MathWorks Inc.). Via SMOOT_{In vitro} analysis the spz could be segmented per movie frame, based on their fluorescence signal intensity, size and crescent shape by applying a binary threshold, a median filter, skeleton adjustment and spot removal. Colliding spz and spz circling on the edge of the field of view were excluded, since the locations of these spz on the consecutive frames cannot be stitched together in a reliable way. The value of these segmentation parameters can be set in the GUI. The direct visual feedback provided by the GUI during the setting of the parameters was used to find the optimal segmentation settings. Over time the median pixel locations of segmented spz visible on different frames could be connected into full spz tracks, which were subsequently separated in segments based on changes in movement pattern. The movement pattern of the spz was classified per segment as floating, stationary or circling. The velocity and turn angle at frame level, the turn direction (clockwise (CW) or counter-CW (CCW)) at segment level and the length of the travelled path (expressed as number of frames and displacement per frame) at track level of the circling spz were determined. On average ~ 100 spz were analysed per condition.

SMOOT_{In vitro} was able to discern three main spz movement patterns in the microscopic spz movies: floating (red), stationary (green) and circling (blue; Fig. 1). The floating spz could not attach to the surface and as such displayed random movement based on fluidics (Fig. 1a). Overall, 3% of the total number of spz imaged were considered to be stationary, either by full length or partial surface attachment. Uniquely the latter results in a “waving” phenomenon (47% of stationary spz, Fig. 1b). Furthermore, motile spz were observed which attached to the surface and turned in circles (Fig. 1c). The majority of circling spz moved in a CCW direction, only 2% of them turned CW. Whilst the CCW movement was often sustained during the whole movie, CW-turning spz only turned on average 1.5 circles (range 1–7 circles) before detaching from the surface. The different formulations did not influence the ratio between CCW and CW movement (Additional file 2: Table S1).

Using automated SMOOT_{In vitro} movement pattern classification, a matrix of nine different formulations (f1–9) was studied: PBS, HBSS and RPMI (f1–3), which were enriched with either 3.5 mg/ml purified BSA (f4–6) or 10% FBS which contains amongst other things ~ 3.5 mg/ml BSA (f6–9). Changes in the formulation yielded a direct effect on the spz movement patterns. Typically, in formulations without BSA or FBS (f1–3) and also in the condition PBS + BSA (f4), most spz were floating. An example of such an effect is shown in Fig. 2a. In the formulation HBSS + BSA both floating and circling spz were seen (f5, Fig. 2b), whereas in the other conditions with BSA and FBS the majority of spz were circling (f6–9, Fig. 2c).

Quantitative analysis confirmed that the movement behaviour of spz was significantly different between the conditions with and without BSA and FBS (p = 0.007; One-way ANOVA; Fig. 3). Without FBS or BSA, 92% of the spz were floating, only a 7% minority of the spz could attach to the surface and were stationary (Fig. 3: 1–3). These stationary spz were either fully attached (41%) or were waving (59%). Interestingly, waving spz were observed significantly more in the formulations which did not induce circling (Fig. 3: 1–4) than in the formulations which could (Fig. 3: 5–9; resp. 94% and 6%, p = 0.026; independent sample t = test), suggesting that waving and circling are mutually exclusive. It thus seems that spz which are unable to circle can (partially) adhere, but then cannot progress to the next movement phase and as a consequence start waving.

Conforming previous data, BSA proved to be an essential supplement [24] since HBSS and RPMI enriched with 3.5 mg/ml BSA induced circling (Fig. 3: 5–6). However, BSA was not the only essential supplement since PBS + BSA was not able to induce circling (Fig. 3: 4). The following trend was seen for the three formulations enriched with BSA: PBS, a phosphate-buffered sodium chloride solution, did not induce circling, HBSS which contains additional ions (e.g. calcium, magnesium and bicarbonate) and glucose induced 41% circling spz and RPMI, which contains salts, glucose, amino acids and vitamins increased the percentage of circling spz to 76% (Fig. 3: 4–6). This trend revealed that besides BSA, also salts, glucose and certain amino acids and/or vitamins were essential supplements to induce attachment and circling. The role of glucose and amino acids was dissected in more detail by removing these components from the RPMI + BSA formulation (f6). Both depletion of amino acids and glucose resulted in a significant decrease of spz adherence capacity: 22% of the spz were floating in RPMI + BSA, 55% of the spz were floating in the solution without amino acids (Fig. 4: 1; p = 0.002; independent sample t = test) and 63% of the spz were floating in the solution without glucose (Fig. 4: 2; p = 0.005; independent sample t = test).

FBS contains, in addition to its major component BSA (~ 35 mg/ml), many other components like salts, glucose, vitamins and cholesterol. Enrichment of any of the formulations with FBS induced circling of 60–80% of the spz. Significantly more spz were circling in the presence of FBS compared to BSA when it was added to PBS or HBSS (Fig. 3: 4, 5, 7, 8; p = 0.005; independent sample t-test). In contrast, addition of FBS instead of BSA to RPMI did not further increase the amount of circling spz (Fig. 3: 6, 9; p = 0.816; independent sample t-test). Thus, the supplements present in FBS enhanced the percentage of circling spz, however they did not add value on top of RPMI + BSA: albumin, salts, glucose, amino acids and vitamins.

When studying the velocity of spz in the five different formulations that induced circling (Fig. 3: 5–9), overall, spz displayed circular movement with a median velocity of 2.0 ± 1.0 µm/s. When comparing the different formulations, the median velocity was 1.4 and 1.7 µm/s for spz supplemented with BSA (HBSS and RPMI respectively), whereas spz supplemented with FBS exhibited higher velocities; 2.1, 2.5 and 2.7 µm/s (RPMI, HBSS and PBS respectively). This velocity difference persisted over time (Fig. 5a). Besides the components available in solution, also temperature influenced the velocity of the spz. The velocity of spz in RPMI supplemented with FBS nearly doubled after increasing the temperature from room temperature to 37 °C (Fig. 5b; median velocity: 3.8 µm/s at 37 °C versus 2.1 μm/s at room temperature). The spz did not move at a constant speed, their velocity fluctuated along the track, which was visualized by colour-coding (Fig. 6a, b). Yellow sections, corresponding to a high velocity, alternated with purple sections, corresponding to a lower velocity. Given these velocity fluctuations along the track, the velocity of the spz was analysed per frame instead of analysing average track velocities.

Although the range of velocities was similar for spz in the five different solutions (~ 1% of the spz could reach a velocity of 4.0 µm/s under all conditions), between these formulations striking variations could be observed in the distribution of the velocities measured (Fig. 6c: 5–9). Clearly the velocity of the spz was not normally distributed (p < 0.001; Shapiro–Wilk test). Two different trends were seen. First, the spz in medium enriched with FBS moved significantly faster than the spz enriched with BSA (p = 0.002; Mann–Whitney U test). Second, within the three formulations enriched with FBS, spz in PBS moved faster than spz in RPMI (p = 0.050; Mann–Whitney U test). Although less spz started circling in RPMI without amino acids compared to complete RPMI, for the spz which started circling a shift to higher velocities was observed (Fig. 4: 3). Interestingly, this shift was related to a significant decrease in the diameter of the circles (Additional file 3: Fig. S2; 12.3 µm in complete RPMI, 10.3 µm in RPMI without amino acids; p < 0.001; independent sample t-test). Hence, the availability of specific (macro)molecules in the formulation seemed to determine not only spz adherence and their capacity of forward locomotion but also the velocity at which spz were able to move.