ABSTRACT
Melanins are structurally complex pigments produced by organisms in all domains of life from bacteria to animals, including humans. In insects, melanins are essential for survival and have key roles in cuticle sclerotization, wound healing and innate immunity. In this study, we used a diverse set of molecular, biochemical, and imaging approaches to characterize mosquito melanin involved in innate immune defense (melanotic capsules). We observed that melanotic capsules enclosing Plasmodium berghei ookinetes were composed of an acid-resistant and highly hydrophobic material with granular appearance, which are characteristic properties of melanins. Spectroscopical analyses reveal chemical signatures of eumelanins and pheomelanin. Furthermore, using an enrichment approach for the proteomic analysis, we identified a set of 14 of acid-resistant mosquito proteins embedded within the melanin matrix possibly related to an anti-Plasmodium response. Among these, AgMesh, a highly conserved protein among insect species that contains domains suggesting a role in immune recognition and function. We targeted AgMesh for further study using a RNAi-based gene silencing approach in mosquitos and challenged them with two Plasmodium spp. Surprisingly, AgMesh gene silencing in mosquitos was associated with reduced parasite infection, implying an important role in facilitating vector infection by Plasmodium spp. Our results provide a new approach to study aspects of insect melanogenesis that revealed proteins associated with melanotic capsule, one of which was strongly implicated in the pathogenesis of Plasmodium spp. mosquito infection. Given the conservation of AgMesh among disease-transmitting insect vector species, future analysis of this protein could provide fertile ground for the identification of strategies that block transmission of disease to humans.
AUTHOR SUMMARY Vector-borne diseases account for 17% of the global burden of all infection diseases and only malaria kills nearly 500,000 individuals per year, most of them children. Climate change is impacting the geographical distribution and incidence of these diseases, therefore novel ecological-friendly strategies for vector control are needed. Here, we adapted methodologies to study fungal melanogenesis and used them to explore the melanin-based immune response of Anopheles gambiae against malaria parasites. We revealed that pheomelanin together with eumelanin is incorporated in the melanotic capsules against Plasmodium. We also found that melanin-encapsulated Plasmodium is associated to acid-resistant mosquito gut proteins and identify several putative factors of the melanin-mediated immunity. Disruption of a surface-associated protein, found highly conserved among other mosquito vectors, demonstrates its ability to impaired parasite development within the mosquito gut. Our study provides a new approach to investigate the melanin-based defense mechanism in insects, which identified a new host molecule as a potential avenue for developing novel universal pest management schemes.
INTRODUCTION
Malaria is a devastating parasitic disease that has afflicted humans since ancient times. Globally, it is responsible for approximately half-million deaths per year, 90% of them occurring in Sub-Saharan Africa, and mostly in children less than 5 years of age [1]. The disease is caused by Plasmodium parasites, which must undergo a complex developmental cycle inside a mammalian host and an obligatory insect vector, the anopheline mosquitoes [2]. Mosquito midgut infection by Plasmodium motile ookinetes and their subsequent differentiation into oocysts (encysted form) is a crucial stage for successful parasite development and malaria transmission [3]. Several reports have demonstrated that ookinetes establish multiple protein-protein and protein-glycan interactions with surface molecules of the midgut epithelial cells to mediate their adhesion, attachment, and invasion [4–8]. Conversely, during the parasite’s development in the mosquito gut, it is targeted by innate immune defenses that rely on systemic and cellular effector mechanisms mediating parasite killing; the midgut is therefore regarded as a bottleneck in the malaria transmission cycle [9, 10]. A major defense system against invading pathogens is the melanin-based immune response which also can target Plasmodium during the ookinete traversal of the midgut epithelium, albeit this defense is rarely observed in natural infections with Plasmodium [11] because of parasite immune-evasive mechanisms [12–14].
Melanins are high molecular weight, insoluble, and acid-resistant biopolymers that exhibit a stable free radical signature [15]. These pigments are found across all kingdoms of life and play key roles in diverse physiological processes of insects, such as cuticle hardening, wound healing, eggshell resistance to desiccation, and immune defenses (reviewed in [16]). As a defense mechanism, melanization is the result of an enzymatic reaction that oxidizes tyrosine into melanin precursors followed by additional oxidation and crosslinking to proteins. This results in the formation of melanotic capsules, melanin-layered structures that surround and sequester invading pathogens [17]. The melanin biosynthesis pathway involves a multistep proteolytic cascade that turns pro-phenoloxidases (PPOs) zymogens into active phenoloxidases (POs), which generate free radicals and toxic quinone intermediate radicals as secondary metabolites. PPO enzymes are stored in both circulating and sessile hemocytes (insect immune cells), which play a central role in melanogenesis (reviewed in [18]). Unlike mammals and fungi, where the melanization reaction is compartmentalized within a vesicle called melanosome [19–21], defense-mediated melanization in insects occurs in an infected body cavity, and consequently requires tight regulation to avoid the deleterious effects of toxic by-products. To restrict damage, several invertebrates have developed an alternative strategy. POs can form activation complexes associated with a functional amyloid-scaffold, which provides a site-specific to sequester and accelerate melanin precursor polymerization, anchor melanin, and promote hemocytes adhesion during the pathogen encapsulation [22–24]. In malaria-transmitting mosquitoes the complex and tightly regulated process of pathogen recognition by the complement-like protein TEP1, responsible for the elimination of Plasmodium ookinetes through lysis or melanization, has been extensively documented [25–27]; however, the factors and mechanism(s) mediating melanin accumulation and hemocyte recruitment to successfully target the pathogen surfaces and minimize damage to the host, are poorly understood. Current knowledge about melanin-related defenses in insects suggest that formation of a functional amyloid-scaffold is required for enhanced melanin accumulation and anchorage upon hemocyte-targeted intruders [22, 28].
Previously, we have used the environmental fungus Cryptococcus neoformans as a model to study melanin biology and demonstrated that fungal cell-wall constituents like chitin, chitosan, and lipids serve as scaffolding for melanin accumulation [29–31], while the melanin granules themselves are intimately associated to proteins that may be essential in the synthesis and/or structure of the biopolymer [19]. Here, we adapted protocols developed to study fungal melanization and used them to investigate the biophysical nature of melanotic capsules to explore the hypothesis that these structures will trap key mosquito factors involved in establishing the melanin-based immune response against malaria parasites and/or mediating parasite interaction with the midgut epithelium.
RESULTS
Melanotic capsules of Anopheles mosquito against malaria parasites are made of eumelanin and pheomelanin
The genetically selected A. gambiae L3-5 refractory (R) strain is known to melanize most P. berghei parasites in the midgut epithelium because of a dysregulated redox system [32, 33]. In this study, for the ease of accumulation of melanotic capsules, we have used this model system. To isolate melanized Plasmodium parasites from infected mosquito midguts, we employed a modification of the protocol used to obtain “melanin ghosts” from fungal cells, that relies on the acid-resistant property of melanin [34, 35]. This property was critical for obtaining the melanized sample material that we further analyzed as illustrated in (Figure 1A). As the tissues were digested by acid exposure, examination of remaining material by light microscopy demonstrated the recovery of a dark-colored material that tended to clump together (Figure 1B, top; Figure S1). The acid-resistant aggregates were analyzed by dynamic light scattering (DLS), which revealed particles with a mean diameter of 1283.8 nm and 2015.1 nm, at 30- and 60-min post-treatment, respectively, consistent with the size dimensions observed by microscopy. The formation of aggregates correlated with the tough and highly hydrophobic nature of melanins (Figure 1B, bottom). Further studies by scanning electron microscopy (SEM) showed that the aggregated material corresponded to masses of clumped ookinetes (Figure 1C), while a cross-sectional micrograph visualized by transmission electron microscopy (TEM) revealed a single ookinete coated by an electron-dense material with granular appearance highly suggestive of melanin (Figure 1D).
Insect melanotic capsules are assumed to be composed of melanin but this assumption has not been experimentally validated by direct characterization of these structures. Consequently, we examined the physicochemical nature of the aggregated material by optical absorption and fourier transform infrared spectroscopy (FTIR). A well-dispersed suspension of isolated melanotic capsules exhibited the classical monotonic, featureless, and broad band UV-visible spectrum (Figure 1E), similarly to those reported for C. neoformans and synthetic melanins [36, 37]. The infrared spectrum of the previously isolated material was a characterized by: 1) a broad absorption band at 3200 cm−1, which reflects stretching vibrations of -OH and -NH2 groups [38]; 2) bands centered around 2930 cm−1 corresponding to vibrations of CH2 and CH3 aliphatic groups [39]; 3) strong absorption bands at 1600 cm−1 assigned to the vibrations of aromatic groups (C=O or C=C), which are typically seen for a conjugated quinoid structure and represent a key feature for the identification of melanins; and 4) weak absorption bands between 700-600 cm−1 due to C-S stretching vibration, suggesting of presence of pheomelanin [40, 41] (Figure 1F). Except for the last absorption bands corresponding to sulfur groups, all other bands are like those found in C. neoformans eumelanin (Figure 1F).
Altogether, these analyses demonstrate that the melanin-based immune response of A. gambiae mosquitoes against malaria parasites requires synthesis and assembly of a granular polymer composed of both eumelanin and pheomelanin. To our knowledge, this is the first experimental evidence of pheomelanin incorporation together with eumelanin in the insect immune defense mechanism triggering melanotic encapsulation.
Identification of acid-resistant proteins enriched in Plasmodium-infected midguts from A. gambiae’s melanotic capsules
We further explored whether a proteomic analysis of melanin-encapsulated parasites allowed us to identify mosquito proteins involved in the establishment of the melanin-based immune response and/or mediating parasite interaction with the midgut epithelium to promote its invasion. To distinguish proteins associated to the mosquito melanin-based immune response, we aimed to identify proteins that were found enriched in midguts of mosquitoes fed with a P. berghei-infected bloodmeal in comparison to midguts from naïve blood fed mosquitoes, before and after incubation in acid.
As expected, homogenates from midgut tissues infected with malaria parasites exhibited a different pattern of total proteins in comparison to the homogenates from naïve blood-fed midguts prior to acid treatment (Figure 2A). In contrast, after acid digestion the total protein content for both samples was significantly reduced but a subtle smear was still noticeable demonstrating that a small number of proteins were resistant to acid degradation (Figure 2A, top). Following solubilization of proteins from midgut tissues and using a filter-aided sample preparation (FASP)-based procedure coupled with mass spectrometric analyses allowed us to identify three sets of proteins enriched in the infected sample vs control: A) Pre-acid treatment; B) Post-acid treatment; and C) Subset commonly found pre- and post-acid treatment (Figure 2A, bottom). Before the acid hydrolysis, proteomic analysis identified 336 proteins enriched in response to Plasmodium infection. According to the top ten gene ontology (GO) terms related to biological process, many of these proteins are involved in cell redox-homeostasis, response to oxidative stress, and aerobic respiration (Table S1), which correlates with the elevated generation of mitochondrial reactive oxygen species (ROS) previously reported for A. gambiae R strain [33, 42]. After the acid treatment, a subset of 14 acid-resistant proteins enriched in infected midguts was identified, suggesting that these were shielded from acid degradation by a close association with melanin (Figure 2B, Table S2). According to their cellular localization, 79% are intracellular proteins (N = 11) likely found in the cytoplasm (N = 2), associated to membranes (N = 4) or to the cytoskeleton (N = 2), while the remaining 21% (N = 3) correspond to extracellular proteins (Figure 2B, unshadowed area). Only protein (AGAP004742) was found enriched in both conditions. Notably, the 3 extracellular proteins embedded within Anopheles melanotic capsules are surface-associated molecules [AGAP000550 (AMOP, complement module), AGAP007745 (Peritrophic matrix-associated protein), AGAP010479 (AgEBP)] previously reported as potential targets of transmission-blocking vaccine (TBV) [8, 43, 44].
AgMesh (AGAP000550) influences malaria parasite development
Multiple molecules on the midgut surface are shown to play critical roles in the process of midgut adhesion and invasion by Plasmodium parasites [4–7, 43]. An in-silico analysis on the above 3 proteins has identified one interesting protein for further investigation. The complete sequence of AGAP000550 reveals a single transcript size of 6.08 kb and a predicted 1,407-amino acid polypeptide of unknown function. It shares high amino acid sequence identity across main anophelines vectors of malaria (> 86%) and other vectors of arboviruses Aedes aegypti and Aedes albopictus (~78%) (Figure 3A). Its predicted protein sequence contains multiple domains that have been previously found in various immune factors, such as an immunoglobulin domain [43], NIDO, AMOP, von Willebrand factor type D (VWFD), and CCP (complement control protein) also known as short complement-like repeat or SUSHI repeats (Figure 3B). These domains are also identical to those identified in Drosophila melanogaster Mesh (CG31004), a transmembrane protein critical for the formation and barrier function of septate junctions -functional counterparts of tight junctions in vertebrates [45]. Thus, we refer to AGAP000550 as A. gambiae Mesh (AgMesh). The high degree of conservation across disease-transmitting insect vector species and its domain composition, suggests the likely recognition and immune functions of AgMesh, which might play a potent role in the establishment of Plasmodium infection in the mosquito midgut.
To investigate whether AgMesh is influencing Plasmodium infection of the mosquito midgut and its association with the melanin-based immune defense, we assessed the capacity of both P. falciparum and P. berghei to establish infection in the midgut of the susceptible A. gambiae (Keele strain) (S) and the refractory A. gambiae (L3-5 strain) (R), respectively, using a RNAi-based gene silencing approach (Figure 4). A double-stranded RNA (dsRNA) corresponding to a 151-bp coding region of AgMesh was injected into the mosquitoes thorax to deplete AgMesh mRNA 4 days prior to infection with a Plasmodium-laced bloodmeal; control mosquitoes were injected with dsGFP, an unrelated gen. Silencing of AgMesh in the S mosquitoes resulted in a decreased (median number of oocysts 7 vs 13, p = 0.0002) P. falciparum infection intensity compared to the dsGFP-injected (Figure 4A, left), whereas the infection prevalence was not affected (90% vs 95%) (Figure 4B, left). Depletion of AgMesh in the R strain resulted in a decreased P. berghei infection intensity and prevalence in comparison to control in comparison to control [median number of total parasites 2 vs 12, p < 0.0001 (Figure 4A, right); prevalence 44% vs 93%, p < 0.0001 (Figure 4B, right)]. Interestingly, AgMesh-silencing did not abolish the melanization phenotype in the R strain [median number of melanized parasites 2 vs 4, p = 0.0007 (Figure 4C, right)], but it significantly reduced parasite development within the midgut of both susceptible and refractory mosquitoes. The weakened melanin-based immune response in the AgMesh-silenced R mosquitoes could be due to a lesser number of ookinetes reaching the basal lamina or to a direct implication of AgMesh in the melanization process. Nevertheless, AgMesh disruption in this strain seemed to promote Plasmodium recognition since no midgut exhibited live parasite only in contrast to 19% in the R control (Figure 4D).
To gain a better understanding on the spatial specificity of AgMesh’s influence on infection we assessed whether its depletion may disrupt P. berghei ookinete development at the luminal compartment of the midgut 26 h post infection (hpi) as well as its ability to invade the midgut tissue. An exploratory analysis of the parasite presence in the lumen was performed by dissecting midguts of AgMesh-silenced mosquitoes at 26 hpi. Visual examination with a light microscope of smears from lumen content stained with giemsa revealed several immature stages of the parasite, along with abundant gut microbiota (Figure S2A, red arrow heads). Immunohistochemical analyses of midgut sheets to localize parasites via detection of the ookinete surface protein P28 [46] using the anti-Pbs28 antibody were performed by confocal microcopy (Figure S2B). There results indicated that P. berghei ookinetes could successfully invade the midgut epithelial cells in AgMesh-silenced R mosquitoes. As a noteworthy observation derived from these trials was an apparent loss of gut elasticity and integrity upon AgMesh depletion, further pointing at a function of AgMesh as a structural component. Difficulties with the maintance and acquisition of the A. gambiae L3-5 refractory strain hindered us to conduct further experiments to elucidate the role of AgMesh in Plasmodium melanization.
DISCUSSION
Melanin synthesis in insects is crucial for their survival (reviewed by [16]). These natural pigments are involved in multiple physiological processes including cuticle hardening, wound healing, and innate immune response. In vector mosquitoes such as the malaria vector A. gambiae, the innate immune system actively restricts and neutralizes foreign organisms including those transmitted to the human host. A better understanding of the chemical nature and molecular scaffolds required to establish a robust melanin-based immune response could inform the development of novel strategies to block transmission of vector-borne human pathogens.
In this study, we explored the A. gambiae melanin-based immune response against malaria parasites by exploiting and adapting a toolbox of multidisciplinary strategies used for the analysis of C. neoformans melanin [19, 35]. We took advantage of melanin’s ability to resist degradation by acids and allowed us to isolate A. gambiae melanotic capsules. We then applied a range of microscopic and spectroscopic analyses to investigate the physicochemical nature of these particles. Lastly, we conducted a proteomic analysis to identify mosquito proteins that remained crosslinked to melanin after the acid treatment. These results demonstrate that the isolated material from mosquito guts in response to Plasmodium infection exhibited consistent properties of melanin: highly hydrophobic, acid-resistant, and having a monotonic broadband absorption curve. The melanin coat surrounding an ookinete revealed a unique electron microscopy appearance of tightly packed nanospheres similarly to those reported for melanized cryptococcal cells after extensive acid hydrolysis [19]. These melanotic capsules incorporated both eumelanins and pheomelanins, therefore providing the first experimental confirmation that pheomelanins play a role in insect immunity.
Early biochemical studies of melanogenesis in other organisms [47], deduced that melanotic capsules were composed of an eumelanin and pheomelanin mixture as formation of thiols conjugates from dopaquinone naturally occur 1,000 times faster that its cyclization into leucodopachrome. Nonetheless, reports of sulfur-containing pheomelanins in insects are restricted to cuticular pigmentation analyses in D. melanogaster [48], the grasshopper Sphingonotus azurescens [49], species of bumblebee from the genus Bombus [50], and male parasitoid wasps from the genus Mesopolobus [51]. The functional significance of pheomelanins in the immune eumelanin surrounding Plasmodium may reflect pro-oxidant activity as well as anti-inflammatory and antioxidant properties [52], respectively, to promote the parasite killing while avoiding host damage. Given that dopamine is the main precursor for both eumelanin and pheomelanin in insects [48], it is possible that melanotic capsules exhibit neuromelanin-like properties [53]. These melanin cages may serve as a trapping system that prevents accumulation of toxic quinones produced by dopamine oxidation [54]. Further chemical and structural investigations of the mosquito melanin-based immune response are encouraged to further elucidate these mechanisms.
Malaria parasite infection has been reported to elicit extensive A. gambiae transcriptome responses that mediate an effective control of Plasmodium [55]. Knowing that melanization of Plasmodium ookinetes occurs as early at 16 h post infection (hpi) [32, 56], we decided to dissect midguts 6 days post infection (dpi) to recover fully melanized parasites for our analyses. The rationale for this choice was the following: a) at 32 hpi A. gambiae R strain kills over 90 % of ookinetes, which persist melanized through the mosquito lifetime [56]; b) parasite melanization occurs in parallel to two peaks of TEP1 expression at 24 hpi and 3-4 dpi, in accordance with the passage of ookinetes through the columnar cells of the midgut epithelium and the establishment of oocysts, respectively [56]; and c) to avoid gross contamination of our samples with human blood and serum proteins. In addition, a previous report demonstrated that mosquitoes fed on uninfected blood exhibit elevated expression of immune genes in response to the expansion of the mosquito’s natural gut microbiota [57].
Melanotic capsules from A. gambiae against the malaria parasite demonstrated an association with 14 proteins, which were found enriched in infected midguts post-acid hydrolysis, suggesting a possible role on the anti-Plasmodium response and close localization to the melanin-coated parasite. A single protein identified as pyruvate carboxylase (AGAP004742) known to be critical for gluconeogenesis and fatty acid synthesis was shared among the two sets. Functional and physiological studies using the melanizing A. gambiae R strain demonstrated that parasite encapsulation is due to increased levels of reactive oxygen species which are further augmented by blood feeding [58, 59]. In agreement with the dysregulation of three key biological processes in this refractory mosquito strain [59], half of the acid-resistant proteins enriched in infected midguts are involved in mitochondrial respiration (AGAP000881, AGAP002884, AGAP012081, AGAP005134), redox-metabolism (AGAP007790, AGAP012407), and lipid metabolism (AGAP004742) [42]. Two additional acid-resistant proteins (AGAP004877, AGAP010147) cytoskeleton-related were also found enriched in infected midguts possibly reflecting major cytoskeletal rearrangements due to blood engorgement and/or ookinete parasite invasion the midgut epithelium [60, 61]. Furthermore, AGAP001151 orthologs in D. melanogaster [62] and Ae. aegypti [63] (96 and 93% identity, respectively) play a critical role in the innate immune response. Multiple studies have shown that Plasmodium melanization occurs upon ookinete contact with the mosquito hemolymph when the parasite reaches the basal side of the epithelial cells [32, 56, 64, 65]; our finding of previously reported extracellular surface-associated molecules (AGAP000550, AGAP007745, AGAP010479) in close association with the melanotic capsules was in agreement with these earlier studies. Nonetheless, a recently published live in vivo imaging of Plasmodium invasion of the mosquito midgut clearly depicted three stages of this process: i) lumen localization; ii) contact with the microvillar surface, and iii) cell invasion [60]. Hence, “parasite coating” with the identified surface components was not unexpected. AGAP010479 (AgEBP) [8] is known to be important ookinete-interacting proteins that mediate midgut invasion through its apical side. AGAP007745 is an abundant peritrophic matrix (PM)-associated protein predicted to be heavily glycosylated and containing a proline-rich domain, which may play a structural role in the PM [44]. AGAP000550 (AgMesh) is a secreted uncharacterized mucin-related protein also found associated with the midgut brush border microvilli (BBMV) proteome of A. gambiae and A. albimanus [43, 66]. Its extracellular domains NIDO and AMOP are found in cell adhesion molecules possibly mediated by the presence of cysteines and tyrosine residues involved in disulfide and dityrosine bonds, respectively. The mucin VWFD is required for blood clotting factor VIII binding and normal multimerization of VWF [67, 68]. Interestingly, phenoloxidases are humoral procoagulants that catalyze clot hardening serving both immune and structural functions [reviewed in [24]]. Furthermore, SUSHI (CCP) is an evolutionary conserved domain essential for complement-mediated immune functions and plays an important role in insect-microbe interactions [69]. We showed that AgMesh silencing in both A. gambiae S and R female mosquitoes significantly reduced their susceptibility to the malaria parasite infection. Particularly in the R strain, Mesh-silenced mosquitoes exhibited a 2-fold decreased of parasite ookinete melanization and severe impairment of parasite development within the midgut epithelium.
Many hematophagous insects including A. gambiae are known to control overproliferation of commensal microbiota via dual oxidase (Duox) activity at midgut epithelium in response to a bloodmeal [70–72]. Duox is a transmembrane protein that contributes to gut homeostasis by secreting ROS [73] and promoting the formation of a dityrosine network (DTN) that stabilizes the peritrophic membrane (PM) [74, 75]. The PM is an acellular semipermeable layer composed of chitin polymers. In A. gambiae, a Duox/Peroxidase (IMPer) system catalyzes formation of the DTN by protein cross-linking in the mucin layer, which is in the ectoperitrophic space between the PM and the midgut epithelium [71, 76]. The DTN forms a permeability barrier that prevents immune responses of mosquito epithelial cells against gut microbiota, thus favoring Plasmodium parasites development in the midgut lumen. Based on the results obtained in this study, we hypothesize that tyrosine residues from extracellular domains NIDO and AMOP of AgMesh serve as substrate to Duox/IMPer to form covalent dityrosine bonds that give rise to the DTN. Reduced parasite infection exhibited in S and R AgMesh-silenced mosquitoes may be due to a sustained induction of microbicidal effector genes as PM integrity has been compromised as it was demonstrated when either Duox or IMPer are silenced [71]. The enhanced levels of ROS and chronic state of oxidative stress known for the R strain [58, 59] could explain the stronger decrease of parasite intensity and no ability of Plasmodium to develop inadvertently, which was evidenced by the absence of midguts only exhibiting live oocysts. The lack of PM integrity could also correlate with the fragility noticed on blood-filled guts of AgMesh-silenced mosquitoes at 26 hpi possible related to a dysregulation of the gut microbiota homeostasis [77, 78]. Furthermore, AgMesh AMOP domain contains 8 invariant cysteine residues predicted to be involved in the formation of intermolecular disulfide bonds suggesting a beta-sheet organization and possible protein multimerization [79, 80]. AgMesh mucoid-coating of those parasite ookinetes that successfully invaded the gut epithelium and reached the basal lamina could serve as a molecular scaffold (amyloid-like) mediating deposition and accumulation of the site-specific melanin-based immune response. This hypothetical model provides a rational explanation for AgMesh association with the mosquitoes melanotic capsules. The absence or reduction of AgMesh avoids formation of the dityrosine-based barrier therefore increasing gut permeability and diffusion of immune elicitors triggering the host immune defenses against Plasmodium parasites. Additional analyses to validate this theory are being performed using the A. gambiae G3, parental strain of L3-5.
Our experimental approach to investigating components associated with insect melanotic capsules was inspired by the success of similar work in melanotic yeasts [19]. The ability of melanin to resist acid digestion allows its recovery in granular form and our results show that it is associated with numerous proteins. We caution that close association does not necessarily imply a role in either melanotic capsule formation or host defense. However, the fact that these proteins resisted acid digestion does imply some form of tight linkage, or protection by the melanin pigment, which in turn hints at a possible involvement in host defense or Plasmodium pathogenesis. By choosing one of these proteins with unknown function and then showing that interference with its expression in mosquitoes reduced Plasmodium spp. infection, our results provide a proof of principle that this approach can lead to proteins involved in process of parasite infection.
In summary, our analysis provides a new approach to the study of the A. gambiae melanin-based immune response against malaria parasites by characterizing the melanotic capsule and by identifying several candidate factors of melanin-mediated immunity. The high degree of AgMesh conservation across mosquito vectors of human diseases and its influence on Plasmodium infection suggests it may represent a fertile line of investigation in the search of a universal transmission-blocking strategy, which aims to target the PM for insect pest management. Lastly, the striking resemblances in the molecular scaffolding of polyaromatic polymers (dityrosine and melanin) that are conserved in insects as well as in the fungal kingdom [81], highlight the functional importance for these dark pigments in the tree of life.
MATERIALS AND METHODS
Ethics statement
This project was carried out in accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of National Institutes of Health. The animal protocols and procedures were approved by the Animal Care and Use Committee of the Johns Hopkins University (Protocol MO15H144). Commercial anonymous human blood was used for parasite cultures and mosquito feeding, no informed consent was required.
Biological materials
Plasmodium berghei (ANKA clone 2.34) were propagated in random-bred Swiss Webster female mice. P. falciparum NF54 (Walter Reed National Military Medical Center, Bethesda) infectious gametocyte cultures were provided by the Johns Hopkins Malaria Research Institute Parasite Core Facility and were diluted to 0.03% gametocytemia before feeding to the mosquitoes using an artificial membrane feeder. Anopheles gambiae L3-5 strain and A. gambiae Keele strain mosquitoes were maintained on sugar solution at 27 °C and 70% humidity with a 12-h light to dark cycle according to standard rearing condition.
Isolation of melanotic capsules from mosquito midguts
A. gambiae L3-5 strain female mosquitoes were infected with P. berghei (ANKA clone 2.34) by feeding them on an anesthetized infected mouse. The infectivity of each mouse was established by measuring the parasitemia and visualizing 1-2 exflagellation events/field under the microscope. Blood-fed mosquitoes were kept at 19 °C for 6 d to allow parasite ookinetes invasion of midgut epithelial and development of the mosquito melanin-based anti-Plasmodium immune response. Around 200 midguts were dissected in Dulbecco’s phosphate-buffered saline (DPBS) (Corning), transfer to a 5ml-tube containing 75% ethanol while avoiding contamination from other mosquito body parts (ovaries, thorax, legs, wings, and head), and kept at 4 °C. Dissected guts were washed two times with DPBS and resuspended in distilled water. Midguts were homogenized by sequentially aspirating them through a 26-gauge and 27-gauge needle, 10 times each. Homogenized material (~500 μl) was transferred to a 1.5-ml tube and an equal volume of 12 N HCl was added. Using a Thermomixer C (Eppendorf AG, Germany), homogenate suspended in distilled water was incubated at 80 °C with 800 rpm shaking speed. Aliquots were removed at 5, 10, 20, 30, 40, 50, 60 min, transferred to a clean tube, and centrifuged at 45,000 rpm for 5 min. The pellet from each aliquot, corresponding to melanin aggregates, was resuspended in 50 μl DPBS and examined under a light-contrast microscope.
Dynamic Light Scattering (DLS)
DLS is a non-destructive physical technique that provides information on the size and heterogeneity of a sample by measuring the random fluctuations of scattered light by particles in suspension. Measurement of melanin aggregates particles formed along the melanotic capsules isolation was performed with a 90Plus/BI-MAS Multi Angle particle size analyzer (Brookhaven Instruments). Thus, 10 μl of melanin aggregates suspension was diluted to 100 μl in DPBS. Data are expressed as the average of 10 runs of 1-min data collection each. Particle size distribution by number emphasized the smaller size in the sample, reflecting melanin hydrophobicity and aggregation as exposure time to acid increased.
Electron Microscopy
A 50 μl aliquot of melanin aggregates recovered from A. gambiae infected midguts after 60 min of incubation in 6 M HCl was fixed in 2.5% (v/v) glutaraldehyde, 3mM MgCl2 in 0.1 M sodium cacodylate buffer, pH 7.2 for 1 h at room temperature. For TEM, after buffer rinse, samples were post-fixed in 1% osmium tetroxide, 0.8% potassium ferrocyanide in 0.1 M sodium cacodylate for at least 1 h (no more than two) on ice in the dark. After osmium, samples were rinsed in 100 mM maleate buffer, followed by uranyl acetate (2%) in 100 mM maleate (0.22 μm filtered, 1 h, dark), dehydrated in a graded series of ethanol and embedded in Eponate 12 (Ted Pella) resin. Samples were polymerized at 37 °C for 2-3 days followed by 60 °C overnight. Thin sections, 60 to 90 nm, were cut with a diamond knife on the Reichert-Jung Ultracut E ultramicrotome and picked up with 2×1 mm formvar copper slot grids. Grids were stained with 2% uranyl acetate in 50% methanol followed by lead citrate and observed with a Philips CM120 TEM at 80 kV. Images were captured with an AMT CCD XR80 (8 Megapixel camera - side mount AMT XR80 – high-resolution high-speed camera). For SEM, samples were fixed and rinsed as described above. After buffer rinse, samples were post-fixed in 1% osmium tetroxide in 0.1 M sodium cacodylate buffer (1 h) on ice in the dark. Following a DH2O rinse, samples were dehydrated in a graded series of ethanol and left to dry overnight (in a desiccator) with hexamethyldisilazane (HMDS). Samples were mounted on carbon coated stubs and imaged on the Zeiss Leo FESEM (Field Emission Scanning Electron Microscope) at 1kV.
Fourier-transform Infrared Spectra (FTIR)
A dry sample (~ 2 mg) of melanin aggregates recovered from A. gambiae infected midguts after 60 min of incubation in 6 M HCl were IR characterized using a Nicolet Nexus 670 FT-IR spectrometer coupled with Smart Gold Gate KRS-5 accessory (Thermo Fisher Scientific, USA). FTIR spectra were recorded between 4000 and 400 cm−1 in transmittance mode averaging 32 scans. Spectra resolution was 4 cm−1. A standard eumelanin spectrum was also obtained using Cryptococcus neoformans melanin. Data acquisition was done using OMNIC software (Thermo Fisher Scientific).
RNA interference (RNAi) gene-silencing
Primers specific to AgMesh (Table S3) were designed with the T7 polymerase promoter sequence appended to the 5’end of each oligo (5’- TAATACGACTCACTATAGGG-3’). PCR products were purified, and their size verified by agarose gel electrophoresis. Around 1 μg of each PCR product was used as template for dsRNA synthesis using the HiScribe T7 High Yield RNA synthesis kit (New England BioLabs, Cat # E2040S) as per the manufacturer’s protocol. dsRNA samples were adjusted to a final concentration of 3.0 μg/μl in water. A. gambiae female mosquitoes (3-4 d old) were cold-anesthetized and inoculated intrathoracically with 69 nl of dsRNA using a nano-injector (Nanoject, Drummond Scientific, USA). Mosquitoes were maintained at 19 °C for 3 d before infecting with a blood meal. As control, age-matched female mosquitoes were injected with equivalent concentration of GFP (green fluorescent protein gene) dsRNA. Gene silencing was verified 3 to 4 d after dsRNA injection by real-time quantitative RT-PCR, done in triplicate, with the A. gambiae ribosomal S7 gene as the internal control for normalization as previously has been published [57].
Plasmodium infection
A. gambiae Keele strain female mosquitoes were infected artificially by membrane feeding with P. falciparum NF54 gametocyte cultures according to an established protocol [55]. The cultures were washed and diluted to 0.03% gametocytemia in normal 60% human serum plus 40% human RBCs. Infected blood meal was provided to mosquitoes directly from glass, water-jacketed membrane feeders warmed at 37 °C. Mosquitoes were allowed to feed for 30 min, afterward unfed females were removed. Mosquitoes were kept for 8 days at 27 °C to assess parasite infection by oocyst counting.
A. gambiae L3-5 strain female mosquitoes were infected with P. berghei (ANKA clone 2.34) as described previously [55]. Blood-fed mosquitoes were kept at 19 °C for 10 d for oocyst and melanized parasite counting. Midguts were dissected in DPBS (Corning) and stained in 0.2% (w/v) mercurochrome to visualize parasite forms under a light-contrast microscope.
Solubilization of proteins associated with melanotic capsules
For the identification of proteins enriched in Plasmodium-infected midguts and crosslinked with the melanin matrix, sample homogenate was incubated in the acid solution at 80 °C for 30 min as described previously. To neutralize pH, samples were centrifuged at 45,000 rpm for 30 min at 4 °C and resuspended in 2 ml 1M Tris-HCl. Next, sample was washed sequentially with 0.1 M Tris-HCl and PBS, three times per wash with 2 volume of each solution. To solubilize melanin-associated proteins, material was resuspended in 50 μl 2% SDS (end concentration) and incubate in ThermoMixer C (Eppendorf AG, Germany) at 55 °C for 3 h. An aliquot of 5 μl was assess by SDS-PAGE and remaining sample was submitted to the Mass Spectrometry and Proteomic Core Facility (Johns Hopkins Medicine, Baltimore, MD). An enrichment approach to identify proteins expressed in response to the parasite infection was conducted, using as control naïve blood-filled midguts previous and post acid treatment.
LC-MS/MS analysis
Solubilization of proteins associated with the melanin matrix was facilitated by previous treatment with 2% SDS (end concentration). Further studies using a filter-aided sample preparation (FASP)- based procedure was conducted at the Johns Hopkins Mass Spectrometry and Proteomic Facility. To each sample 40 μl of 20 mM ammonium bicarbonate pH 8.0 was added then reduced using 5 μl of 50 mM dithiothreitol for 1 h at 60 °C. Samples were chilled on ice then alkylated using 5 μl of 50 mM iodoacetamide for 15 min in the dark. Samples were then diluted in 400 μl of 9 M sequencing grade urea (Thermo Fisher). For each sample, a 30 kD MWCO spin filter (Amicon) was washed with 300 μl distilled water 3 times using a centrifuge at 14,000 g for 3 min. Each sample was then added to a washed spin filter and spun down at 14,000 g for 5 min. Two successive washes with 400 μl 9 M urea were spun through each filter allowing the SDS in urea to flow through the spin filter. Three additional washes with 400 μl of 20 mM ammonium bicarbonate were used to rinse the urea from the sample prior to trypsinization. To each filter a calculated ratio of 1/50 enzyme to protein in 300 μl of 20 mM ammonium bicarbonate was added and allowed to digest at 37 °C on filter overnight. The next day, samples were each centrifuged at 14,000 g and the flow through of digested peptides was collected. Filters were then rinsed twice with 20 mM ammonium bicarbonate and the flow through added to the digested peptides. The peptides were then acidified and desalted on an Oasis HLB microelution plate (Waters) according to protocol. After FASP digestion, samples were brought up in 2 % acetonitrile, 0.1 % formic acid and separated with an EasyLC into a QE-Plus (Thermo Fisher Scientific) mass spectrometer and eluted over a 90-min gradient from 100% buffer A (2% acetonitrile, 0.1% formic acid) to 100% buffer B (90% acetonitrile, 0.1% formic acid) with an MS resolution of 70,000 and an MS2 resolution of 35,000 running a top 15 DDA method.
MS/MS Data analysis
Data were searched against a database compiled from the ReqSeq 2015 database entries for Anopheles gambiae (AgamP4, VectorBase) and Plasmodium berghei (downloaded October 10, 2016) using Proteome Discoverer (version 1.4) and the search engine Mascot (version 2.5.1). The enzyme designation was set to trypsin with one allowed missed cleavage and a mass tolerance of 5 ppm for precursors and 0.02 daltons for fragment ions was used. Variable modifications were allowed for M oxidation and deamidation on N and Q. Carbamidomethyl on C was set to static. The search results were filtered at a 1% FDR using the target/decoy PSM validator for peptides associated with the reported proteins. Scaffold (version 4.10.0, Proteome Software Inc, Portland, OR) was used to validate MS/MS based peptide and protein identifications. Peptide identifications were accepted if they could be established at a greater than 95.0% probability by the Peptide Prophet algorithm [82] and contained at least 2 identified peptides. Normalized total spectral counts from each independent experiment was exported to Microsoft Excel for Windows 365 for analysis. Abundance of expressed proteins identified in control and infected samples was performed based on their consistent detection (at least 2 times across 5 independent experiments) in both experimental conditions (prior and post-acid treatment). This led to the inclusion of 336 and 14 from a total of 1110 proteins for the pre- and post-acid treatment conditions, respectively. Manual annotation of proteins found enriched in infected samples was performed using the conserved domains as predicted by VectorBase (VB-2016-2020), InterPro, Pfam, UniProt, and BLAST. Previous literature regarding genomic analyses of A. gambiae, particularly of the L3-5 strain, were also used as reference [42, 58].
In silico analysis
Bioinformatic analysis to identify predicted domain structures was conducted using PROSITE MyDomains [83]. To estimate amino acid sequences percentage identity analysis was conducted using Geneious Prime Software (version 2020.1.2). Proteomic data were visualized and analyzed using the Scaffold Proteome Software (version 4.10.0).
Statistical analysis
GraphPad Prism 9 (version 9.0.2) software package was used to performed statistical analyses. Test used in a particular experiment is indicated in its respective Figure legend.
Figure S1. Time-course isolation of Anopheles melanotic capsules during acid treatment and heating at 80°C. Light microscopy images at 20X, 40X, and 100X (oil immersion) prior and post 5-, 10-, 20-, 30-, 40-, 50-, and 60-min acid hydrolysis demonstrates acid-resistant and heat-stable aggregated melanin-like material that tends to clump due to its hydrophobic nature. Bars 10 μm
Figure S2. Effect of AgMesh-silencing in the parasite’s development 26 hpi in A. gambiae (L3-5 strain) R mosquitoes. A, Giemsa staining of lumen content of midgut to verifying presence of several Plasmodium parasite immatures stages (red arrow heads) along with gut microbiota. B, Confocal microscopy using anti-Pbs28 against P. berghei to detect parasite (white arrowhead) invasion of the midgut epithelial cells.
Table S1. Subset of proteins from Anopheles gambiae (L3-5 strain) enriched in Plasmodium berghei-infected midguts at 6 dpi (prior to acid treatment).
Table S2. Subset of acid-resistant midgut proteins from Anopheles gambiae (L3-5 strain) identified associated to melanotic capsules against Plasmodium berghei.
Table S3. Primers used in this study.
ACKNOWLEDGMENTS
We would like to thank the Johns Hopkins Malaria Research Institute and Bloomberg Philanthropies. We would like to acknowledge Barbara Smith for her expertise and technical support with transmission EM and SEM at the Microscopy Facility, School of Medicine, Johns Hopkins University. We also gratefully acknowledge Robert N. O’Meally for sample processing and data acquisition related to proteomic analyses (Johns Hopkins Mass Spectrometry and Proteomic Facility, The Johns Hopkins University). We thank Joel Tang for training on FTIR acquisition and data analysis (Johns Hopkins NMR Core Facility, Department of Chemistry). The following reagent was obtained through BEI Resources, NIAID, NIH: Anopheles gambiae, Strain L35, MRA-114, deposited Mark Q. Benedict. EC, DFQS, and AC are funded by the Johns Hopkins Malaria Research (JHMRI) Institute Pilot Grant Casadevall_123; EC, DFQS, and AC are funded by NIAID R01 AI052733; EC was awardee of a Postdoctoral Fellowship from the Johns Hopkins Malaria Research Institute. DFQS has been funded by NIH grants 5T32GM008752-18 and 1T32AI138953-01A1. GD, YD and YA were supported by the National Institutes of Health grants R21AI131574, R01AI101431 and R01AI122743.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.
- 6.
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.
- 21.↵
- 22.↵
- 23.
- 24.↵
- 25.↵
- 26.
- 27.↵
- 28.↵
- 29.↵
- 30.
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵