An ex vivo system for investigation of Plasmodium berghei invasion of the salivary gland of Anopheles stephensi mosquitoes

ABSTRACT

Plasmodium sporozoites associated with the midgut and in the hemolymph of mosquitoes differ from sporozoites in the secretory cavities and ducts of the insects’ salivary glands in their transcriptome, proteome, motility, and infectivity. Using an ex vivo Anopheles stephensi salivary gland culture system incorporating simple microfluidics and transgenic Plasmodium berghei with the fluorescent protein gene mCherry under the transcriptional control of the Pbuis4 promoter whose expression served as a proxy for parasite maturation, we observed rapid parasite maturation in the absence of salivary gland invasion. While in vivo Pbuis4::mCherry expression was only detectable in sporozoites within the salivary glands (mature parasites) as expected, the simple exposure of P. berghei sporozoites to dissected salivary glands led to rapid parasite maturation as indicated by mCherry expression. These results suggest that previous efforts to develop ex vivo and in vitro systems for investigating sporozoite interactions with mosquito salivary glands have likely been unsuccessful in part because the maturation of sporozoites leads to a loss in the ability to invade salivary glands.

Introduction

The sporozoites of malaria-causing Plasmodium parasites are asexual forms of the parasite that begin their development in the hemocoel of a mosquito vector and end it with their invasion and infection of hepatocytes of a vertebrate host [1,2]. While the morphology of the parasites remains relatively constant during this time, the quantity and specific patterns of gene and protein expression vary markedly along with the metabolic activity and motility of the parasites and their potential to interact with, invade, traverse and exit from multiple cell types [1,3–6]. Sporozoite-based approaches in the development of malaria interventions span the vaccine, drug development, and vector domains [1,7–9].

Current understanding of sporozoite biology in the mosquito while extensive in terms of descriptions of patterns of gene and protein expression is somewhat limited in its mechanistic understanding of the interactions between sporozoites and salivary glands. Functional genomics studies have enabled some parasite genes essential for salivary gland invasion to be identified and a few mosquito proteins have been implicated in parasite interactions [1,2,10,11]. However, the interactions between sporozoites and salivary glands represent a substantial knowledge gap in our understanding of the biology of vertebrate malarias. In vitro systems for studying the vertebrate liver and erythrocytic phases of Plasmodium development have been significant enabling technologies which, unfortunately, do not have analogs for studying the mosquito phases of Plasmodium development [12,13]. There is no ex vivo or in vitro systems for investigating ookinete/midgut or sporozoite/salivary gland interactions and these are well-recognized critical limitations [1,10,11].

As one of first attempts toward investigating the sporozoites invasion of An. stephensi salivary glands and to visualize the molecular developmental differences between sporozoites obtained from mature oocysts and sporozoites obtained from salivary glands, an ex- salivary gland vivo system is used which could render this complex biological event amenable to experimental analysis. Here, while attempting to establish a workable ex vivo culture system to study sporozoite/salivary gland interactions we discovered that coculturing of salivary glands of Anopheles stephensi and Plasmodium berghei sporozoites stimulated the rapid precocious maturation of the sporozoites in the absence of salivary gland invasion into parasites resembling those that would be found within salivary glands. Salivary gland invasion by sporozoites results in the parasites gaining their full ability to infect a vertebrate host while losing the ability to invade mosquito salivary glands[14]. Preventing this precocious maturation will be important in the successful development of an ex vivo model system for investigating the mosquito phases of Plasmodium development.

Material and methods

Mosquitoes

Anopheles stephensi SDA 500 were used throughout and is a strain which was selected to be highly susceptible to Plasmodium falciparum [15,16]. All life stages were reared and infected at the Institute for Bioscience and Biotechnology researches (IBBR) in a Conviron environmental chamber at 28°C ± 2, with 80% relative humidity (RH) under a 12-hour light/dark cycle. Larvae were fed pulverized fish food (TetraMin® Tropical Flakes) ad libitum while adults were fed 10% sucrose solution continuously. Mated adult females were blood-fed on 4–6 week old BALB/c females Mus musculus.

Mice

All mice (Mus musculus) used in this study were 4–6 weeks old BALB/c females were obtained from Taconic Biosciences (Rensselaer, New York). The use of mice and all procedures involving mice were done under the authority and with the approval of the University of Maryland, College Park’s Institutional Animal Care and Use Committee operating in compliance with the guidelines of the National Institutes of Health’s Office of Laboratory Animal Welfare (PROJECT REFERENCE #: R-15-78).

Plasmodium berghei

The transgenic Plasmodium berghei line PbANKA-Cherry-2204cl was obtained from the Leiden Malaria Research Group (Leiden University Medical Center, Leiden, Netherlands). PbANKA-Cherry 2204cl contains an mCherry expression cassette flanked by the 5’ and 3’ UTRs of P. berghei UIS4 integrated into the 230p locus of P. berghei GIMO-PbANKA. mCherry expression is only detectable in sporozoites after they have entered the salivary glands of the mosquito host. A description of PbANKA-Cherry-2204c1 can be found in the Rodent Malaria genetically modified Parasites Data Base (RMgm-1339) [17,18].

Plasmodium berghei infection of An. stephensi

Three to four days old female mosquitoes were allowed to feed on anesthetized BALB/c mice infected with PbANKA-Cherry-2204cl with a parasitemia of 25–30%. Fed mosquitoes were maintained on a solution of water containing 8% w/v dextrose in 0.05% w/v para-amino benzoic acid at 24° C ± 2 and 70% RH in 12:12 light/dark cycles.

Midguts and salivary glands Isolation

The infected mosquitoes (10–14 days post infection and 15–22 days post infection) were anesthetized by chilling at 4°C ± 2, until immobilized, quickly surface sterilized with 70% ethanol and washed in 1X phosphate buffered saline and kept chilled on ice prior to dissection. Mosquitoes were dissected in cold RPMI 1640 (Corning Life Sciences, Tewksbury, Massachusetts). Midguts and salivary glands were dissected as described [19] and placed in 50 μl of cold RPMI 1640. For microscopic observations infected midguts and salivary glands were placed on a glass microscope slide in a small volume of VECTASHIELD® Mounting Medium (Vector Laboratories, Burlingame, California) and covered with a glass coverslip. All microscopic observations were made using a Zeiss Axio Imager.A1 using phase contrast or fluorescence illumination to visualize mCherry protein.

Ex vivo culture system

Freshly dissected salivary glands were cultured in RPMI media on a channel slide with a height of 0.4 mm width of 5 mm and length of 50 mm that enabled continuous flow of culture media (ibidi µ-Slide I ^0.4 Luer; 80,176, ibidi GmbH, Gräfelfing, Germany). Midgut sporozoites were isolated from oocysts in infected mosquitoes at 10–14 days post infection by dissecting midguts of infected mosquitoes (30–50) and pooling them in 50ul of cold RPMI. Pooled midguts were gently crushed in an additional 100 ul of RPMI 1640 with a plastic pestle to release sporozoites. Sporozoites (15–25 × 10³) were added to a culture slide containing salivary glands (approximately 10) and incubated at 24 ± 2°C either under static culture conditions or under perfusion conditions (using a syringe pump) whereby the culture media with sporozoites flowed across the salivary gland at approximately 16ul/sec. The flow was reversed every 5–10 seconds to create an oscillatory laminar flow. Salivary glands and sporozoites were co-cultured for 4hrs after which time the glands and sporozoites were examined microscopically.

Sporozoites counting

In vitro treated sporozoites’ samples were examined using Zeiss Axio Imager.A1 under phase contrast and fluorescence illumination to visualize mCherry. The total number of sporozoites was counted for all samples, as well as mCherry-expressing sporozoites (at least 5-fields for each slide, 3-slides from each tube and the whole experiment was repeated at least 3 times).

Western blotting

The total protein was extracted from in vitro treated sporozoites, sporozoites isolated from midgut oocysts (negative control) and sporozoites isolated from infected salivary glands (positive control) by homogenizing the same number of sporozoites (≈15–25 × 10³) with 0.1 µl of 1X protease inhibitor cocktail (SC248229A; Thermo Fisher Scientific, Waltham, Massachusetts). Proteins were separated by SDS-PAGE (4–15% gradient) in running buffer (25 mM Tris base, 192 mM glycine, 1% SDS, pH 8.3). Proteins were transferred to pre-activated PVDF membranes using an iBlot 2 gel transfer device (Thermo Fisher Scientific, Waltham, Massachusetts) at 20 V, 0.9A for 6 minutes. A rabbit polyclonal anti-mCherry antibody (ab167453; Abcam, Cambridge, UK) was the primary antibody (1:1000 in 3% milk – TBS) and an anti-rabbit IgG antibody-alkaline phosphatase conjugate (S373B; Promega, Madison, Wisconsin) was the secondary antibody (1:10,000 in 3% milk – TBS T). Blots were imaged with a stabilized substrate for alkaline phosphatase for ~10–20 mins (S3841; Promega, Madison, Wisconsin) and visualized using the Java image processing program ImageJ 1.48 [20].

Statistics

Data were checked for normality by using Shapiro–Wilk normality test at a significance level p < 0.05. The data did not deviate significantly from normality and parametric statistics were used. Comparisons were performed with an unpaired two-tailed t-test. Statistical analysis was performed using GraphPad Prism 7.04 (GraphPad Software, San Diego, California).

Results

Phenotype of PbANKA-Cherry-2204cl in Anopheles stephensi

Oocysts of PbANKA-Cherry-2204cl-infected midguts 10–14 days post infection had no visible evidence of mCherry expression (Figure 1). Likewise, sporozoites within these oocysts also had no visible evidence of mCherry expression (Figure 1). However, salivary glands of PbANKA-Cherry-2204cl-infected mosquitoes 15–22 days post infections had visible evidence of mCherry expression (Figure 2). These results are expected [29] based on the known expression patterns of UIS4 and the reported phenotype of PbANKA-Cherry-2204cl in the Rodent Malaria Genetically Modified Parasites Database (RMgm 1339) [17,18].

Figure 1.

PbANKA-Cherry-2204cl transgenic sporozoites (SPZ) (mCherry reporter gene under the control of Pbuis4 promoter and 3′ UTR) within oocysts (Oo) in vivo. (a) An. stephensi infected midguts on day 12 post-infection visualized using phase contrast optics and fluorescence illumination (b). After crushing midguts and oocysts to release sporozoites the sporozoites were visualized using phase-contrast optics (c) and fluorescence illumination (d). No expression of the Pbuis4: mCherry reporter gene in midgut sporozoites was detectable. Higher magnification of the marked areas in a & b is shown in the insets; (scale bar = 12.5 µm) scale bar = 25 µm).

Figure 2.

PbANKA-Cherry-2204cl transgenic sporozoites (SPZ) in salivary glands (SG) in vivo. (a) An. stephensi infected salivary glands on day 18 post-infection visualized using phase-contrast optics, low magnification (scale bar = 12.5 µm). (b) Higher magnification (scale bar = 25 µm) view of distal lateral lobe of salivary gland shown in (a) under phase contrast optics and under fluorescence illumination to visualize mCherry (c). (d) Sporozoites released from infected salivary glands by gentle crushing under phase contrast optics and fluorescence, showing mCherry expression (e) (scale bar = 25 µm) .

Ex vivo culture

The culture system used here was designed to facilitate the passive movement of sporozoites and increase contact between parasites and salivary glands by mimicking the flow of hemolymph in the insect. PbANKA-Cherry-2204cl with mCherry under the transcriptional control of the Pbuis4 promoter was used to facilitate the detection of sporozoite invasions since UIS4 expression is restricted to infectious sporozoites in the salivary gland [11,29]. Therefore, this system provided a sensitive way to detect even small numbers of invading sporozoites. While sporozoite invasion of the salivary gland was not observed in this system under any conditions, sporozoites expressing mCherry were abundant in the culture media. Under the conditions of this experiment sporozoites newly isolated from oocysts appear to rapidly mature to exhibit the early expression of Pbuis4::mCherry. The activation of Pbuis4::mCherry expression suggests that the transcriptional profile of infectious sporozoites found within salivary glands has been activated (Figure 3A and 3B).

Figure 3.

Effect of salivary gland (SG) on PbANKA-Cherry-2204cl transgenic sporozoite (SPZs) maturation in vitro. PbANKA-Cherry-2204cl sporozoites were isolated from oocysts on day 10 − 14 post-infection. Sporozoites were incubated with freshly dissected uninfected salivary glands in vitro for 4 hrs., then analyzed by phase and fluorescence microscopy for mCherry (red) expression. Sporozoites in vitro are shown in pairs of images, one taken using phase contrast optics (a, c & e) showing the oocyst sporozoites and under red fluorescence illumination (b, d & f). b & d show expression of the Pbuis4::mCherry reporter gene outside of salivary glands, and F shows oocyst sporozoites with no expression of Pbuis4::mCherry (Magnification ×200). (g) The total sporozoite count and the fluorescence sporozoite count after 4 hrs. incubation in control and treated groups. (h) Western blot analysis of in vivo oocyst and salivary gland sporozoites and in vitro treated sporozoites, showing the expression of the mCherry. ** P < 0.01 using an unpaired two-tailed t-test. (scale bar = 25 µm).

Co-culture of PbANKA-Cherry-2204cl sporozoites and salivary glands in vitro

Oocyst sporozoites isolated from infected An. stephensi 10–14 days post-infection were exposed to intact freshly dissected salivary glands from An. stephensi for 4 hrs after which time the sporozoites were examined microscopically for mCherry expression.

Sixty-seven percent (67%) of PbANKA-Cherry-2204cl sporozoites exposed to salivary glands expressed mCherry although they had not invaded the salivary gland and remained extracellular. In comparison, less than 9% of the PbANKA-Cherry-2204cl sporozoites in the control sample that was not exposed to whole salivary glands but were otherwise treated the same expressed mCherry (Figure 3). The difference in the observed stimulatory effects of exposure to whole salivary glands was statistically significant (p < 0.01) (Figure 3G). These results were corroborated by a Western blot analysis (Figure 3H).

Discussion

Plasmodium sporozoites that emerge from oocysts in the hemocoel of the mosquito while competent to infect the salivary glands of the insect host are not immediately highly infectious to their vertebrate host relative to sporozoites that have successfully invaded the salivary glands and reside in the gland’s secretory cavities and ducts [14]. While multiple studies have shown that salivary gland invasion is not absolutely required for infectivity of the vertebrate host, salivary gland invasion is important [21,22]. For example, while Al-Olayan et al. demonstrated the vertebrate infectivity of P. berghei sporozoites developed in vitro, the degree of infectivity of those sporozoites was significantly lower than the infectivity of sporozoites from the salivary glands of an infected mosquito [22,23]. While sporozoites within the hemocoel of the mosquito have been found to be less infectious relative to sporozoites in the salivary gland their infectiousness increases with time, indicating that there are time-dependent elements to sporozoite development [21]. Nonetheless, there are major differences in the transcriptomic and proteomic profiles of pre-salivary gland invasion sporozoites and post-salivary gland invasion sporozoites [3,5]. Interestingly, once in the salivary glands sporozoite maturation results in changes in the parasite that result in them being incapable of reinvading the salivary gland after their isolation and reinjection into the hemocoel of a mosquito [14].

The current study was motivated by an interest in finding ex vivo culture conditions that would enable the direct microscopic observation of sporozoite/salivary gland interactions and parasite invasion – one of the ‘black boxes’ of sporozoite biology as characterized by Frischknecht and Matuschewski [1]. Here, we used a fluorescent protein gene (mCherry) under the transcriptional control of the 5’ regulatory region of Pbuis4 and a simple microfluidics system to simulate the flow of insect blood. UIS4 is essential for vertebrate infections and uis4 expression increases 1500-fold upon entry of the sporozoite into the salivary gland [5]. In PbANKA-Cherry-2204cl mCherry expression was not detectable in oocysts and was only visible in salivary gland sporozoites making mCherry expression a useful proxy for sporozoite maturation. In this study, we did not observe sporozoite invasion of ex vivo cultured salivary glands, but we did observe rapid sporozoite maturation. Following the exposure of sporozoites isolated from oocysts to ex vivo cultured salivary glands, we observed mCherry expression in that absence of gland invasion. These results have a number of implications.

First, there have been no reports of an ex vivo system for culturing mosquito salivary glands that supports sporozoite invasion. This is not due to lack of interest or effort since such a system could provide an experimental platform for investigating this important step in the development and transmission of Plasmodium parasites [1]. Based on the findings reported here, exposure of pre-invasion sporozoites to dissected salivary glands is enough to trigger gene expression that is usually associated with post invasion parasites, suggesting that parasite maturation has been accelerated. It has been reported before that the hemolymph- and oocyst-derived sporozoites attracted to salivary gland homogenate and the attraction was restricted to salivary gland-derived sporozoites [24].

Mature, post-invasion sporozoites are not competent to invade salivary glands and this would account in part for the inability to create an ex vivo culture systems to study salivary gland invasion by these parasites. Salivary gland culture conditions need to be found which, at a minimum, do not stimulate precocious maturation of sporozoites. While it is not known what specifically is triggering the ex vivo maturation of hemocoel sporozoites into salivary gland sporozoites, the observations reported here could serve as the basis for developing assays to identify those factors.

Second, understanding what extra cellular factors are responsible for stimulating the rapid extra-salivary gland maturation of sporozoites could not only help understanding the biology of the parasite but might become a basis for creating malaria transmission incompetent mosquitoes.

Genetic approaches to the control and elimination of vector borne diseases are of increasing interest as technical advances make their development and application increasingly feasible. For example, RNA-guided DNA endonucleases derived from bacterial antiviral systems such as Cas9 are the basis upon which powerful and effective gene drive systems are being constructed and introduced into malaria-transmitting mosquitoes [25]. One proposed application of these technologies is to ‘drive’ genes through standing populations of An. gambiae, for example, that interfere with the development and transmission of Plasmodium falciparum – effectively immunizing the mosquitoes against the parasites [26,27]. Current strategies involving the inclusion of mosquito-expressed single-chain antibodies to Plasmodium surface antigens and/or insect antimicrobial peptides that are directly toxic to the parasite have achieved moderate success [9,28,29].

However, current transgenic mosquitoes are not fully and robustly malaria transmission incompetent. The results of this study suggest an additional strategy for disrupting Plasmodium development in the insect – triggering precocious maturation of sporozoites that are in the hemocoel before they invade the salivary glands by expressing one of more ‘maturation factors in the hemolymph. This might provide an additional strategy for achieving robust neutralization of the mosquito’s ability to transmit malaria parasites that might be propagated within mosquito populations using gene drive technologies.

Funding Statement

This work was supported by the Ministry of Scientific Research, Egypt [Js-3623]; University of Maryland Foundation [805618-1].

Authors’ contributions

Mai I. Hussein performed the whole practical protocols, data analysis and wrote the manuscript; Belal A. Soliman revised the manuscript; Maha K. Tewfick provided direction, contributed in data analysis and reviewed the manuscript; David A. O’Brochta suggested the protocol idea, designed the experiments, supervised the practical work step by step and co-wrote the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).