Introduction

Parasitic arachnids such as ticks and mites are well known to transmit a variety of bacteria, many of them pathogenic, to and between hosts (Houck et al. 1991; Jaenike et al. 2007). The genus and species of mites present in Drosophila culture vary, and differ in destructiveness from those that eat the culture food to those that prey upon the flies themselves (Ashburner and Thompson 1978). Good Drosophila husbandry requires that fly stocks be maintained free of mites (Ashburner and Thompson 1978), yet this situation prohibits detection of rare horizontal transmission events that might be vectored by mites. Further, maintaining Drosophila as monocultures does not accurately mimic conditions in nature. There is experimental evidence that mites can serve as vectors of both mobile elements and bacteria. Mites are thought to be the vector that introduced P-elements into Drosophila (Houck et al. 1991). Mites have also been shown to be able to transfer Spiroplasma endobacteria from one species of Drosophila to another (Jaenike et al. 2007); Macrocheles subbadius mites acquired Spiroplasma infection after feeding on the hemolymph of infected D. nebulosa and were then able to transmit the infection to D. willistoni (Jaenike et al. 2007). This provides evidence that horizontal transmission of endobacteria via ectoparasitic mites is possible in Drosophila.

The endobacterial species Wolbachia pipientis is an obligate intracellular α-proteobacterium, a relative of the ancient α-proteobacterium that gave rise to mitochondria. Wolbachia mainly infect arthropod species and are thought to infect more than 20 % of all insect species (McGraw and O’Neill 2004; Tram et al. 2003; Iturbe-Ormaetxe and O’Neill 2007; Hilgenboecker et al. 2008). As obligate intracellular bacterial species, Wolbachia are passed vertically to the offspring through the egg cytoplasm (Serbus and Sullivan 2007). As a result, Wolbachia have evolved numerous means to manipulate host reproduction to enhance their vertical transmission, so to ensure propagation to the next generation. While the effect of Wolbachia infection ranges from parasitic to symbiotic, depending on the Wolbachia strain, titer, and the host organism (Serbus and Sullivan 2007), these alterations of host biology can cause deleterious effects. These effects include altering the female-male sex ratios by inducing parthenogenesis, feminization of genetic males or male killing, or by impacting offspring production and survival through cytoplasmic incompatibility, the most widely studied reproductive consequence of Wolbachia infection (Harris and Braig 2003; McGraw and O’Neill 2004; Iturbe-Ormaetxe and O’Neill 2007). These reproductive effects are caused by modifications to sperm and oocyte chromatin, which implies that Wolbachia can have widespread epigenetic effects on host gene expression (Clark et al. 2005; McGraw and O’Neill 2004).

Wolbachia are not usually considered to be susceptible to horizontal transmission—direct transfer between host individuals of the same generation. However, horizontal transfer of Wolbachia can be achieved experimentally through microinjection of infected egg cytoplasm between strains (Boyle et al. 1993; Schilthuizen and Stouthamer 1997; Van Meer and Stouthamer 1999; Dobson et al. 2002; Merçot and Charlat 2004; Xi and Dobson 2005). The presence of similar Wolbachia genomes in distantly related arthropods suggests horizontal transfer can also occur in nature (O’Neill et al. 1992; Zhou et al. 1998; Vavre et al. 1999; Schulenburg et al. 2000; Casiraghi et al. 2001; Tram et al. 2003). As Wolbachia typically induce reproductive isolation in their hosts, it is reasonable to postulate that these instances of horizontal transmission involve a vector (Vavre et al. 1999).

Parasites have been shown to be able to acquire Wolbachia from their insect hosts; Trichogramma and Leptopilina parasitic wasps can become infected with Wolbachia when parasitizing Wolbachia-infected hosts (Heath et al. 1999; Huigens et al. 2004) and in woodlice, Wolbachia has been transferred by contact with an infected host (Rigaud and Juchault 1995). Additionally, a monophyletic relative of Wolbachia, Rickettsia, also obligate endobacterium, can be horizontally transmitted to mammals through arachnid vectors such as ticks (Anderson and Karr 2001). Thus, it is possible that parasites such as mites might not only acquire Wolbachia from Wolbachia-infected hosts but also transmit it to other host individuals, allowing occasional horizontal transfer of Wolbachia.

We observed that when initially Wolbachia-free stocks were placed into our non mite-free room, they could become Wolbachia-infected with distressing frequency, often within a generation or two. Because there is abundant speculation (Jaenike et al. 2007) and empirical evidence (Rozhok et al. 2011) in the literature that mites can transmit Wolbachia, we tested the Tyrophagus putrescentiae (Astigmata: Acaridae) mites from these Wolbachia-infected and uninfected stocks to determine whether these mites could transmit Wolbachia infection between fly stocks. We found that the Drosophila-specific wMel strain of Wolbachia could be isolated from the mites infesting Wolbachia-infected Drosophila stocks. We then tested these mites for their ability to invade non-mite infested and non Wolbachia-infected Drosophila stocks and transmit Wolbachia to these flies. We found that mites were able to invade the mite-free stocks, and this was coupled with novel Wolbachia infection in the previously Wolbachia-free stocks; this infection occurred within 5 fly generations. This provides evidence that Wolbachia transfer between Drosophila laboratory populations can occur via T. putrescentiae. Mite-vectored Wolbachia transmission may also occur in natural settings and as Wolbachia has manifold effects on host reproduction and gene expression, mite-vectored bacterial transfer could be more biologically important that generally appreciated.

Methods

Drosophila stock maintenance

All D. melanogaster stocks were maintained at 25 ± 5 °C in 55 ml glass vials stoppered with foam bungs, all autoclaved prior to fly introduction to destroy any contaminating mites. The bungs are porous enough to allow passage of mites. The stocks were cultured on a standard cornmeal-agar-yeast-molasses medium containing methylbenzoate (0.125 %) to inhibit mold growth. All fly stocks were routinely stored on paper towel soaked with 0.1 % benzyl benzoate to inhibit mite transfer between vials. This step was omitted for the infection transfer experiment.

Mite identification and collection

Mite species was determined morphologically by H. Proctor, University of Alberta. Multiple attempts at molecular species identification by CO1 barcoding using “universal” invertebrate primers (Folmer et al. 1994) served to only detect Drosophila CO1 sequence, presumably from the mite guts. Live T. putrescentiae, were located and scraped from the food and vial surfaces of Wolbachia-infected and uninfected Drosophila stock vials using fine forceps, a fine paintbrush or a spatula. Mites were transferred to 1.5 ml microfuge tubes for DNA extraction or to a fly-containing vial for testing of infection transfer. Mites were not sorted for sex or developmental stage.

Infection transfer experiment

Wolbachia infection transfer by mites between Drosophila strains was tested using uninfected In(1)w m4 (FBst0000807) and infected O.R strains (FBst0002376). Infection status was confirmed by PCR as described below. Mites were removed from fly stocks by rapid serial transfer to fresh vials as recommended by Ashburner and Thompson (1978). The stocks were then visually checked for mite contamination. To begin the experiment, approximately 20 mites from a mite-infested Drosophila stock were transferred to the O.R mite-free stock; no mites were added to subsequent fly generations but mites were passively transferred to new vials as the flies were transferred into new food vials each generation (Fig. 1). The O.R vial with mites and the w m4 vial (initially) without mites were held together with an elastic band and placed in a plastic basket in a mite-free room at 25 °C. The outsides of both vials, the basket, and the elastic were wiped with ethanol to remove any invading mites. The flies were allowed to proceed to further generations in a mite-free incubator without any other fly stocks. Each generation of flies was visually inspected to ensure that no flies had transferred between stocks, and to check if mites were present.

Fig. 1
figure 1

Experimental design. Two mite-free fly stocks, Wolbachia-infected O.R (red eyes) and Wolbachia-uninfected w m4 (white eyes), were kept together in a mite-free location. Mites were introduced to the O.R stock and over 5 generations, the w m4 stock was examined for mite and Wolbachia contamination. This test of transmission was performed in triplicate, with all triplicates performed simultaneously, but physically isolated in different areas of a mite-free room. (Color figure online)

Full size image

DNA extraction

DNA was extracted using the AquaGenomic (MultiTarget Pharmaceuticals) Drosophila protocol adapted from the manufacturer’s tissue protocol with the following amendments; 30 µl of 1 mM Tris–HCl (pH 7.6) was added after air drying the pellet, DNA was re-suspended at 60 °C for 1 h. Extracted DNA was stored at −20 °C until further use. Five whole female flies and approximately 20 whole mites were used for DNA extraction. Any contaminating mites were removed from the flies before each DNA extraction.

Polymerase chain reaction for Wolbachia

The Wolbachia infection status of the starting (G0) and subsequent generations (G1, G2, G3, G4, G5) of the w m4 and O.R stock was determined by PCR. Wolbachia was detected by 2 rounds of amplification of 2 sections of the Wolbachia surface protein gene (wsp) using forward primer 5′-GACCCAGCAAATACTATTGCAGACA-3′, and reverse primer 5′-AGCGGGTTCCAAAGGAGTGC-3′, with an annealing temperature of 53 °C, producing a 150 bp amplicon (Jaenike et al. 2010), and the forward primer 5′-TGGTCCAATAAGTGATGAAGAAACTAGCTA-3′ with the reverse primer 5′-AAAATTAAAGTACTCCAGCTTCTGCAC-3′ with an annealing temperature of 61.6 °C amplifies a 548 bp fragment, which with the flanking primers, generates a 605 bp amplicon (Braig et al. 1998). A portion of the Drosophila garnet gene was amplified as a positive control for DNA integrity using the forward primer 5′-CTCTTTGAGTTTGGGAAATGC-3′ and the reverse primer 5′-TACAAATGCTGGGCTACGAC-3′ with an annealing temperature of 58 °C generating a 501 bp amplicon.

All amplifications used GoTaq Green Master Mix (Promega), and used 95 °C for 5 min for denaturation, 35 cycles at 30 s at 95 °C, 30 s at 57 °C along a gradient of ±5 °C to satisfy the annealing temperatures of the primers, and 72 °C for 40 s, followed by a final elongation (72 °C for 5 min). The second round of amplification of Wolbachia DNA from mites used 5 μl of the first round PCR product for each sample, with the same primer sets as the first round; the negative control used water in place of DNA. The PCR products were electrophoresed on 1.2 % agarose gel with 1× sodium borate buffer stained with SYBR Safe DNA gel stain (Cedarlane). Sequencing was performed by Genome Québec at McGill University. Sequence alignments and analyses were performed using SeaView, CodonCode Aligner, and MEGA.

Drosophila/mite video preparation

Day old Drosophila and mite carcasses, killed by freezing overnight, live mites and live Drosophila larva were placed, in various combinations, in the well of a 9.5 cm2 culture plate with 2 ml 2 % agar on the bottom. Sequential still images were obtained using a EZ4D dissecting scope (Leica) and integrated camera programmed with LAS EZ through Macro Express to take photographs every 30 s, illuminated by an L2 fiber-optic illuminator (Leica). For photographs taken under red light, the illuminator lenses were masked with 3 M lithography tape and otherwise performed identically. The images were rendered and edited using Photoshop CS4 (Windows) at 10 frames/s, therefore every second of video represents 300 s of real time.

Results and discussion

The goal of this study was to determine whether T. putrescentiae mites could transmit Wolbachia between D. melanogaster individuals.

Wolbachia DNA is present in mites

If mites can transmit Wolbachia, one would expect mites present in a Wolbachia-infected Drosophila stock to, themselves, contain Wolbachia. Mites were collected from a Wolbachia-infected Drosophila stock, their DNA extracted and Wolbachia DNA amplified. Two Wolbachia surface protein (wsp) fragments of approximately 550 and 150 bp were amplified from a sample of 15–20 whole mites, indicating the presence of Wolbachia DNA within the mites (Fig. 2). Sequence analysis of the 150 bp product showed that it aligned to bases 223–369 of the wsp gene obtained from Wolbachia strain typically found in Drosophila (Saridaki et al. 2011) (Fig. 3). Evidence of a Drosophila-specific strain of Wolbachia in mites is completely consistent with the results of Rozhok et al. (2011) who also found that 635 bp of the wMel wsp gene could be isolated from mites living with Wolbachia-infected Drosophila. Mites have been implicated in other horizontal transfer events, the most well known being the presumed transfer of P-elements from D. willistoni to D. melanogaster (Houck et al. 1991) and the evidence for their role in this transfer was, similarly, the presence of Drosophila DNA in mites. However, to extend this analysis, we initiated a more rigorous test of the ability of mites to transfer Wolbachia between Drosophila strains.

Fig. 2
figure 2

Wolbachia DNA isolated from mites. Wolbachia DNA was amplified from mites by two sets of primers specific for Wolbachia surface protein (to produce the 550 bp wsp amplicon fragment a and 150 bp wsp amplicon fragment b). Negative controls for each of the two amplifications are shown to the right (negative control fragment a and fragment b)

Full size image
Fig. 3
figure 3

The consensus sequence of Wolbachia DNA amplified from mites. The consensus sequence of Wolbachia DNA amplified from mites, aligned from bases 223 to 369 of the GenBank sequence for wMel Wolbachia surface protein (wsp)

Full size image

Evidence of Wolbachia transfer

Amplification of the wMel strain of Wolbachia from mites could represent genuine infection of mites with this strain of Wolbachia, or alternatively, simply Wolbachia-infected Drosophila tissue associated with the mite bodies, either externally or in their gut. To determine if the mite-associated Wolbachia were viable and capable of being transmitted, mite infested and Wolbachia-infected Drosophila stock vials were placed in close proximity to uninfected Drosophila stock vials and the previously uninfected stocks monitored for new Wolbachia infection occurring concomitantly with mite invasion.

Approximately 15–25 live mites from mite-infested Wolbachia-infected fly stocks were collected and transferred to three vials of mite-free but Wolbachia-infected O. R stocks (Fig. 1). Each of these vials were kept adjacent to a culture vial containing a mite-free and Wolbachia-uninfected In(1)w m4. This experimental design, rather than the transfer of mites directly to a vial of uninfested Wolbachia-free flies, is somewhat analogous to conditions in nature where mites would have to travel along the substrate to encounter flies. These experimental conditions also mimic the conditions that might be occurring in a mite-infested laboratory. Unlike many acarid mites, T. putrescentiae does not have a phoretic nymphal stage (Colloff 2009) and previous observations in our lab had indicated that mites had no difficulty walking along surfaces and invading stoppered vials, and could do so in a matter of hours to days. The flies from each previously mite-free and Wolbachia-uninfected w m4 stock were monitored visually for mite invasion and for Wolbachia using PCR for wsp, from the original to the fifth generation, to determine if the mites were able to transfer Wolbachia.

Visual inspection revealed that mites invaded the previously mite-free w m4 vials by day 7, successfully established themselves in the new vials and established breeding populations that persisted through generations 1–5. PCR of 5 whole flies per vial from the w m4 stock vials showed no evidence of Wolbachia infection in the first, G0, generation (Fig. 4). However, in replicate A, evidence of Wolbachia infection was first seen in the next generation, G1. Evidence of infection was evident in all subsequent generations with the exception of generation 3 (Fig. 4). In replicate B, evidence of Wolbachia infection was present within generations 2 and 3 (Fig. 4). In replicate C, Wolbachia DNA was stably detectable from generations 2 to 5 (Fig. 4). In no case was there evidence of contamination of the w m4 vials by the O.R flies, which would have been evident by finding red-eyed flies in the w m4 vials. Thus, although initially being uninfected with Wolbachia, as determined by amplification of a portion of wsp, all three In(1)w m4 stocks appeared infected within 1–2 generations. These results are consistent with a study by Kozeretska and colleagues who in a preliminary report (Bilousov et al. 2011) and subsequent more detailed report (Rozhok et al. 2011) documented transfer of Wolbachia infection by Tyrophagus noxius mites. In a similar experimental design, but one in which mites from Wolbachia-infected stocks were transferred directly into vials with uninfected flies, they observed stable transfer of Wolbachia by generation 9, but not by generation 2. Despite more direct access to the flies in this experimental design, it appears to have taken longer for Wolbachia to be transferred. This may reflect the number of Wolbachia required to produce an amplicon, as only single round PCR amplification was used in these experiments so that low titer infections in earlier generations might have been missed. Thus, the results presented here and those of Bilousov et al. (2011) and Rozhok et al. (2011) demonstrate that Wolbachia can be transmitted to new hosts by mites.

Fig. 4
figure 4

Changes in Wolbachia infection status of previously uninfected and mite-free w m4 stocks. The initial uninfected status is indicated by no amplification of Wolbachia surface protein in the G0 flies. In subsequent generations, in replicates A, B, and C, Wolbachia DNA is amplified as indicated by a wsp amplicon. In all cases, confirmation of DNA integrity was confirmed by amplification of the same sample with fly-specific primers (“fly”)

Full size image

In at least one of the three replicates (C), the infection appeared to be stable, however, in replicates A and B, Wolbachia infection appeared rapidly but did not persist throughout the experiment. Similar sporadic maintenance of transferred Wolbachia was also found in horizontal transmission in parasitic wasps (Heath et al. 1999; Rigaud et al. 2001; Huigens et al. 2004), presumably due to the failure to efficiently colonize the germ line. The apparent transient infection could represent an infection that failed to establish in the fly germ line, a complex process involving interactions between bacterial and host proteins (Serbus and Sullivan 2007). Alternatively, the number of Wolbachia cells may have been too low for successful amplification in the intervening generations. The intermittent and low abundance of Wolbachia DNA suggests that the number of cells transferred is low and that they may not necessarily establish immediately in the new host. Additional mite colonization events and/or infection from potentially Wolbachia-infected progeny of the originally invading mites are also a possible source of infection of the Drosophila stocks. Similarly, it is not clear whether Wolbachia establish in the germ line of the mites and are transmitted transovarially or whether continued exposure to Wolbachia-infected Drosophila is necessary for Wolbachia to be present in mites. Regardless of the ease of establishing stable vertically-transmitted infections, in all replicates, mites appeared able to transfer Wolbachia to previously uninfected Drosophila hosts. Further, this transfer occurred within a few generations and may have generated Wolbachia infections that were stable over generations.

Possible mechanism of Wolbachia transfer

Tyrophagus putrescentiae mites are a common pest of stored products and of Drosophila cultures but are not known to prey upon living flies, so it was not entirely clear how they could obtain Wolbachia from flies and how other flies could obtain it from them. However, these mites are known to feed upon dead and decaying bodies (Braig and Perotti 2009) so one possible mechanism for the Wolbachia transfer is through mutual consumption of corpses, possibly compounded by cannibalism by Drosophila larvae. To see if mites were eating Drosophila corpses, and if Drosophila larvae would eat mite-infested Drosophila corpses, we used time-lapse photography to capture the behavior of mites in the presence of dead flies, and Drosophila larvae in the presence of dead mite-infested adult fly corpses (Fig. 5 and Supplemental Videos S1 and S2). These images show evidence of mites consuming fly tissue, including ovaries that typically contain a high titer of Wolbachia (Dobson et al. 1999). Similarly, imaging of second and third instar larva revealed ready ingestion of dead adult flies. Immature eggs within dead female carcass seemed especially attractive. Additionally, third instar larva were imaged ingesting dead mites.

Fig. 5
figure 5

Still images obtained from time-lapse photography of fly and mite interactions, suggesting a potential mechanism for Wolbachia acquisition and transfer. a Mites are observed ingesting fly carcasses. b Second and third instar Drosophila melanogaster larva were observed ingesting fly carcasses including hemolymph and immature eggs. c Third instar larva (observed under red light) ingesting dead mites. (Color figure online)

Full size image

The ingestion of carcasses or eggs could be a possible source of Wolbachia introduction, complementing the usual vertical transmission of these bacteria by inheritance in the egg cytoplasm. As Wolbachia-infected mites contaminate a fly stock, the fly larvae can eat the mites, either dead or alive as they scavenge the corpses of dead adults. The larvae could then become infected with Wolbachia. New mites may be able to acquire this infection by feeding on the corpses and Wolbachia-rich eggs of the adult flies. This scenario relies on Wolbachia surviving, if not necessarily replicating, in extracellular conditions and being able to invade new cells. Rasgon et al. (2006) have demonstrated that Wolbachia are viable for at least a week outside of cells and that these cells were able to establish subsequent infection. An alternative mode of transmission may be direct transfer or injection of Wolbachia-infected host cells by mites during the feeding process, as was proposed for P-element transfer by Proctolaelaps regalis mites (Houck et al. 1991).

Endobacteria such as Wolbachia can provide invaluable insight into the acquisition and evolution of eukaryotic organelles. Because Wolbachia has, for the most part, not yet evolved a symbiotic relationship with the host, it is profoundly valuable in examining the initial molecular negotiations between an endobacterium and its host eukaryotic cell (Clark et al. 2005; Iturbe-Ormaetxe and O’Neill 2007). These complex networks of interactions between the endobacterium and host cell have optimized vertical transmission through host eggs, however, in this work we show that mites can mediate horizontal transmission of Wolbachia. Traditionally, Drosophila mites have been primarily studied in the context of the threat they poses to delicate fly stocks. However, this work shows that mites may pose a greater threat to Drosophila research than previously appreciated. Horizontal transmission of Wolbachia by mites can be problematic for Drosophila research. Wolbachia is capable of inducing a variety of reproductive effects in the fly and has widespread effects on gene expression, including epigenetic effects (Clark et al. 2006; Xi et al. 2008; Zheng et al. 2011). So both the presence of Wolbachia and its capacity for rare horizontal transmission must be taken into account in laboratory experiments. In nature, rare instances of horizontal transmission vectored by mites could also occur, as mites are present on Drosophila individuals in the wild, and Drosophila and mite population densities can be high at feeding and breeding sites. Horizontal transmission of endobacteria, although probably rare in nature, may be evolutionarily significant and the role of mites in mediating this transfer warrants further investigation.