The proteomic content of Varroa destructor gut varies according to the developmental stage of its host
The proteomic content of Varroa destructor gut varies according to the developmental stage of its host
Abstract
The nutritional physiology of parasites is often overlooked although it is at the basis of host-parasite interactions. In the case of Varroa destructor, one of the major pests of the Western honey bee Apis mellifera, the nature of molecules and tissues ingested by the parasite is still not completely understood. Here, the V. destructor feeding biology was explored through artificial feeding, dissection of the mite’s gut and proteomic analyses. More specifically, the proteome of guts extracted from starved mites and honey bee-fed mites was compared to highlight both the parasite proteins likely involved in food processing and the honey bee proteins actually ingested by the mite. We could identify 25 V. destructor candidate proteins likely involved in the parasite digestion. As the host developmental stages infested by the mite are diverse, we also focused on the identity and on the origin of honey bee proteins ingested by the mite when it feeds on larvae, pupae or adults. We highlighted profiles of consumed honey bee proteins and their variations throughout the V. destructor life cycle. These variations matched the ones observed in the honey bee hemolymph, showing that this tissue is an important part of the mite’s diet. Based on the variations of abundance of the most consumed honey bee proteins and on their functions, the potential implication of these key candidate nutrients in V. destructor reproduction is also discussed.
Author summary
Varroa destructor is one of the major parasitic pests in modern beekeeping worldwide. Since it shifted host from Eastern to Western honey bees, it was shown to weaken colonies by feeding on both immature and adult stages while transmitting several deadly viruses in the process. Nutrition, an overlooked aspect of parasite biology, is thus a key to comprehend the V. destructor life cycle and its impacts on its honey bee host. We explored the feeding physiology of this ectoparasite by analysing the protein content of its isolated gut to compare it with the protein composition of the honey bee tissues ingested. We highlighted several mite proteins probably involved in digestion and many honey bee derived proteins acquired during the feeding. Honey bee hemolymph is an important part of the mite’s diet although the diet could vary throughout the cycle, especially when mites feed on adult bees. The abundance of the most regularly ingested honey bee proteins such as Vitellogenin or Hexamerin varies throughout the bee development and could directly impact the parasite physiology. The analysis and identification of key proteins required for the mite’s survival and reproduction will pave the way for the development of more specific control strategies.
Introduction
The study of parasite interactions with their animal host often focuses on the impact of host nutrition on parasites’ survival or behaviour [1]. The nutrition of the parasite, on the other hand, tends to be overlooked or limited to the damage it causes to the host. As in any other species, the diet and feeding biology of a parasite is crucial to obtain the necessary nutrients for reproduction and survival [2–4]. Identifying the host tissues consumed and the molecules ingested when the parasite feeds is thus a first step towards a better understanding of its physiological needs and ecology [5]. This is especially important when the host species has an essential environmental or agroecological role.
With hundreds of crop plants and agricultural products depending on honey bee pollination, the Western honey bee Apis mellifera is a perfect example of such essential insect species [6,7]. As other eusocial insects, honey bees suffer from many parasites and pathogens that benefit from their social organisation to spread within colonies [8]. In the past few decades, Varroa destructor became the major ectoparasite of Western honey bees [9,10]. Since this native pest of the Eastern honey bee (A. cerana) shifted host, it has caused damage to A. mellifera colonies in the Northern hemisphere [11]. The impact of the parasite is both direct through the parasitization of honey bees, and indirect through the transmission of deadly viruses [12–14]. Varroa destructor females infest and feed on adult bees during the dispersal phase and on juvenile stages, namely larvae and pupae, during the reproductive phase. Regardless of the host developmental stage being parasitized, it was thought that the mite fed solely on honey bee hemolymph until a study showed that the adult fat body was also part of the parasite diet during the dispersal phase [15]. More recently, the work of Han and colleagues focused on the honey bee hemolymph proteins found inside the whole body of V. destructor. The diet of the mite would in fact change from fat body to hemolymph between the dispersal and the reproductive phase [16], as suggested by the survival of mites fed only on honey bee larva hemolymph [17]. The change in diet composition throughout the mite life cycle is of prime interest as it could in fact play a critical role in the mite’s physiology. The protein intake from the diet is indeed considered essential in the reproduction of arthropods [18]. In the case of V. destructor, in addition to the classic protein intake, females seem able to directly use their undigested host proteins to improve the energetic stocks needed for the reproduction [19–21]. Several specific honey bee proteins obtained through feeding on its pupal or adult host are thus crucial in the parasite metabolism [19,21]. The exploration of the honey bee proteins ingested throughout the parasitic life cycle could shed light both on key nutrients required by this major ectoparasite, on their origin in the host body and on their function.
Mass spectrometry-based proteomics has emerged as a powerful tool to study the life cycles of parasites such as ticks [22], nematodes [23], or protozoa [24]. In the case of V. destructor nutrition, the use of proteomic analyses allowed Han and colleagues to compare the honey bee hemolymph proteins that are found in mites sampled during dispersal on adults and reproduction on pupae. They highlighted that there was a clear change of protein profiles between pupa- and adult-fed mites and attributed this change to the type of tissue consumed by the mite, namely hemolymph on honey bee pupae or fat body on adults. However, proteomic data coming from the honey bee fat body are still lacking and could not be included in this first analysis. Furthermore, honey bee larva-fed mites should also be considered as this host developmental stage is crucial for the mite’s reproduction activation [25]. Our aim was thus to gain a better understanding of the ectoparasite nutrition in terms of tissue consumed and diet requirements during its development. Contrary to previous studies, we focused on the proteome from mite’s isolated gut as it is the major organ involved in nutrition and digestion in the parasite. Furthermore, the inclusion of artificially starved mites was used as a baseline both for the presence of bee proteins and of V. destructor proteins inside mites’ guts. The presence of such control conditions is informative and many studies about parasite nutrition use starved or partially fed individuals as negative controls [26–30]. By comparing these baseline levels to proteomes from mites fed on honey bee larvae, pupae or adults, the identification of the most frequently ingested host proteins was possible as they appear in higher quantities compared to starved mites. Similarly, putative digestive V. destructor proteins are mobilized as soon as the parasite ingests honey bee tissues and are expected to be upregulated in honey bee-fed mites compared to starved mites. Through artificial feeding and off-gel bottom up proteomics, we thus zoomed in on the proteomic profiles from the guts of mites fed on A. mellifera specimens at different developmental stages (larvae, pupae, and adults). The proteomic data obtained from V. destructor guts were further compared to results of proteomic analyses directly ran on both honey bee tissues putatively consumed (i.e. hemolymph and fat body). The honey bee proteins found inside V. destructor guts are indeed expected to be related to the protein profile from the consumed honey bee tissues. This could further shed lights on the origin of the proteins actually ingested by the mite for each of the three developmental stages selected and on the nutritional variations throughout the parasite life cycle. The nutrients available are indeed expected to vary between the honey bee developmental stages, which could in the end impact the physiology of the mite during its crucial reproductive phase.
Results
Identification of V. destructor proteins putatively involved in the mites food processing
A total of 1,847 Acari proteins could be identified in the extracted guts of adult females V. destructor in our study (S1 Fig). The measurement of the protein abundances and subsequent calculated ratios across feeding conditions were performed using the label-free quantitative (LFQ) mass spectrometry (MS proteomics strategy). Based on these ratios, we highlighted specific Acari proteins differentially regulated (i.e. superior to 2 or inferior to 0.5) when mites were starved compared to mites that fed on their honey bee larval, pupal or adult hosts (Fig 1). In our experimental conditions, a tendency towards down-regulation was detected as approximately 63.2% (134/212) of the Acari proteins detected were down-regulated in honey bee fed mites when compared to starved mites, while 78 (36.8%) proteins were up-regulated (Fig 1). This tendency changed when only considering Acari proteins significantly different in all three categories (larva, pupa or adult fed mites) compared to starved mites. Indeed, 18 Acari proteins were up-regulated in all three categories and only two were always down-regulated (Fig 1 and S1 Table). In addition, an Endochitinase-like protein (XP_022673141.1) not listed in the 18 proteins from Fig 1 was up-regulated in all categories but the larva-fed vs starved factor was slightly inferior to 2 (S1 Table). Four proteins were also differentially regulated in all three categories but varied in different directions (the Ankyrin repeat domain containing protein 13 A0A132A0G5, the Chorion peroxidase T1KYK0, the Aldo_ket_red domain-containing protein T1JXC1 and the uncharacterized protein XP_022662294.1 in the S1 Table).
(A) Experimental groups compared in the gut proteomic analysis of Acari proteins. Mites naturally fed on larvae, pupae or adults for 24h were compared to control starved mites that only had ingested PBS. PBS had to be added in the starved condition as V. destructor die from desiccation under 24h if deprived of a water source [44], (B) Venn diagram showing the common number of differentially regulated Acari proteins between larva-fed and starved mites (in blue), between pupa-fed and starved mites (in violet) and between adult-fed and starved mites (in orange). The green and red arrows represent up- and down-regulation, respectively. (C) Amounts of up- and down-regulated proteins in larvae, pupae, adults or in total. For each group, analyses were run on three replicates of 4 mites’ gut pooled together (N = 3 pools of four mite’s gut per condition). Larva (https://doi.org/10.7875/togopic.2022.304) and pupa pictures (https://doi.org/10.7875/togopic.2022.305) were retrieved from the DataBase Center for Life Science and were conceived by Haru Sakai and Hiromasa Ono.
In this reduced pool of mostly up-regulated Acari proteins, the most represented functions were transport or secretion (7/25) with proteins like Vacuolar protein sorting-associated protein 28 homolog, solute carrier family 15 member 1-like protein, Rab GDP dissociation inhibitor, signal recognition particle subunit SRP68-like, FGGY carbohydrate kinase domain-containing protein-like isoform X1, Vesicle-fusing ATPase or Inositol-3-Phosphate synthase [29,31–37]. Immunity or stress response (5/25) were also well represented with Thioredoxin-dependent peroxide reductase-like protein, Solute carrier family 15 member 1-like protein, protein LSM14 homolog A-like, Natterin-4-like and Chorion peroxidase [31,32,38–43] (S1 Table). Furthermore, more than half of the proteins identified (16/25 or 64%) have already been detected in previous studies focusing on the gut or on the feeding status of other ticks, mites or hematophagous insects (S1 Table).
The proteomic exploration of the mite gut content also gave us access to many host-derived proteins ingested by the mite.
Focus on the honey bee proteins ingested by the mite
Starved mites as a baseline to identify ingested host proteins.
After 24h, the starved mites survival reached 77.6% [CI95: 65.8–86.9]. This percentage dropped to 32.8% [CI95: 21.8–45.4] after 48 hours, preventing a longer stay on this artificial medium (S2 Fig). Regarding proteomic analyses, a total of 560 Apis spp. proteins putatively coming from the ingested host tissues could be identified inside the extracted guts of V. destructor females (S1 Fig). Among them, 106 (18.9%) could still be found in the starved parasite gut after 24h without any contact with the host (S1 Data). However, the number of different Apis spp. proteins characterized in the mite’s gut was lower in starved mites when compared to larva-, pupa- or adult fed-mites. The abundance of proteins showed the same trend when compared to the starved control mites and a clear tendency towards over-representation of Apis spp. proteins was observed in honey bee fed groups. More precisely, 90.4%, 86.1% and 92.3% of the differentially expressed honey bee proteins were more abundant in larva-, pupa- and adult fed mites, respectively, than in starved mites (S2 Table).
Richness of honey bee proteins in mites guts and correspondence with the proteomic profile of honey bee hemolymph.
The profiles of Apis spp. proteins found inside V. destructor guts were different when the mite had fed on juvenile stages (spinning larvae or pink-eyed pupae) or on adults (S3 Fig). On the contrary, pupa- and larva-fed mites shared closer Apis spp. protein profiles.
We further compared the Apis spp. proteins found in the V. destructor gut to the proteins found in honey bee hemolymph and fat body. These comparisons were performed for each of the three developmental stages on which the parasite had fed (spinning larvae, light pink-eyed pupae and emerging adults). Naturally, a high number of Apis spp. proteins was found in honey bee tissues, especially in the fat body. Depending on the honey bee developmental stage, 189 (larval stage), 213 (pupal stage) and 298 (adults) of these Apis spp. proteins from honey bee tissues were also found in the mite’s gut (Fig 2).
(A) Venn diagram of Apis spp. proteins shared between mites that fed on spinning larvae for 24h and spinning larvae fat body or hemolymph; (B) Venn diagram of Apis spp. proteins shared between mites that fed on pink eyed pupae for 24h and pink eyed pupae fat body or hemolymph.; (C) Venn diagram of Apis spp. proteins shared between mites fed on one day old adult honey bees for 24 hours and one day old adult fat body or hemolymph. (N = 3 pools of four mite’s gut per condition and three pools of three honey bee hemolymph or fat body per condition).
Even though a high proportion of the common proteins were shared between hemolymph and fat body, some were specific to one honey bee tissue. Regardless of the honey bee stage, the number of proteins in common between V. destructor gut and the honey bee hemolymph is higher than the number shared between the mite gut and the honey bee fat body (Fig 2). More precisely, a total of 54, 53 and 69 proteins from the parasite’s gut are specific to the honey bee larval, pupal or adult hemolymph against 15, 15 and 27 proteins for the honey bee fat body, respectively.
To further analyse the origin of these Apis spp. proteins found inside the mite’s gut, we ran generalised linear models (GLMs) to relate the protein counts in honey bee tissues and in mites’ gut. The counts correspond to the frequency of detection of each protein (classified by its accession number) in three biological replicates. Firstly, analyses of correlations show a general significant correlation between the Varroa gut content and the honey bee tissues (GLM Poisson: Hemolymph, Chisq = 1115.98; p<2.2e-16 ***/Fat_body, Chisq = 17.41; p = 3.01e-05 ***). However, regardless of the developmental stage of the honey bee on which mites had fed, the data from V. destructor gut was always more strongly correlated to the honey bee hemolymph than to the honey bee fat body (Tables 1 and 2 and S4 Fig). The relation between the protein content of the mite guts and of the honey bee fat body is even not significant in all cases but one. This observation was valid both when the entire list of Apis spp. proteins was considered to take the absence of proteins into account (Table 1) and when we focused only on the 189, 213 and 298 Apis spp. proteins shared between the mite and the honey bee tissues (Table 2). Multicollinearity risks were checked through the measurement of the variation inflation factor (VIF). Values were always inferior to 1.2 (far below the threshold of 5).
In this first analysis, the entire set of Apis spp. proteins was considered.
In this second analysis, only proteins in common between V. destructor gut and honey bee tissues were taken into account.
Identification and abundance of the most frequently ingested honey bee proteins.
When focusing on the presence or absence of proteins, only 54 to 77 proteins are consistently found in the three biological replicates depending on the developmental stage considered (S2 Table). Several proteins related to energy storage or transport, such as Transferrin, Hexamerin, Vitellogenin, Apolipophorin and Larval-specific very high-density lipoprotein appeared in all our samples from larva, pupa and adult fed mites. In addition, besides enzymes and proteins from the extracellular matrix, Odorant binding proteins (OBP 13–15, 18) also appeared in all samples.
Many of these proteins present in all three biological replicates are also more abundant in bee-fed than in starved parasites, whether they infested larvae, pupae or adults (S2 Table). Transferrin, Vitellogenin, Hexamerin, Larval-specific very high-density lipoprotein, Apolipophorin, and several different OBPs (OBP3, 13–15, 17 and 18) belong to the list of 84 proteins over represented in all guts of mites fed on honey bees when compared to starved mite (Fig 3A and S2 Table). These proteins are thus always ingested in high quantities except for honey bee Vitellogenin in pupa-fed mites (Fig 3B and S2 Table). The peak of expression of Vitellogenin in fact seems to be at the adult stage, which leads to high concentration within the mite gut after feeding. As shown in Fig 3B, the over-represented Apis spp. proteins found inside the mite’s gut do not only vary in comparison to the starved group but also between mites fed on different honey bee stages. The variation of abundance in the mite guts could reflect the same fluctuation in the ingested honey bee tissues.
대화 참여하기