Shigeto Yoshida, Division of Medical Zoology, Department of Infection and Immunity, Jichi Medical University, 3311-1 Yakushiji, Shimotsuke, Tochigi 329-0498, Japan. Tel.: +81 285 58 7339; fax: +81 285 44 6489; e-mail: firstname.lastname@example.org
‘Flying vaccinator’ is the concept of using genetically engineered hematophagous insects to deliver vaccines. Here we show the generation of a transgenic anopheline mosquito that expresses the Leishmania vaccine candidate, SP15, fused to monomeric red fluorescent protein (mDsRed) in its salivary glands. Importantly, mice bitten repeatedly by the transgenic mosquitoes raised anti-SP15 antibodies, indicating delivery of SP15 via blood feeding with its immunogenicity intact. Thus, this technology makes possible the generation of transgenic mosquitoes that match the original concept of a ‘flying vaccinator’. However, medical safety issues and concerns about informed consent mitigate the use of the ‘flying vaccinator’ as a method to deliver vaccines. We propose that this expression system could be applied to elucidate saliva–malaria sporozoite interactions.
Stable germline transformation systems have been achieved in a number of insect species within the last decade (O'Brochta & Handler, 2008). These systems provide the basis for the development of powerful tools in hematophagous insect vectors for the control of parasite transmission and management of insect populations in the field. Mosquitoes are vectors of pathogens such as malaria, dengue and yellow fever, and recent advances in the genetic engineering of mosquitoes have raised expectations for their use in new strategies for infectious disease control (Chen et al., 2008). The generation of anopheline mosquitoes refractory to malaria by transgenesis has been intensively studied (Marshall & Taylor, 2009), and some mosquito lines expressing malaria-refractory molecules to rodent and avian malaria parasites have been established (Ito et al., 2002; Moreira et al., 2002; Jasinskiene et al., 2007). Recently, we reported that a transgenic mosquito expressing the haemolytic C-type lectin CEL-III from sea cucumber impairs the Plasmodium transmission dynamics of human malaria (Yoshida et al., 2007). These studies support the use of such transgenic mosquitoes in population replacement strategies (Terenius et al., 2008). Alternately, ‘release of insects carrying a dominant lethal (RIDL)’ has been proposed as a new sterile insect technique (SIT) for the control of insect populations using genetically engineered insects (Heinrich & Scott, 2000; Thomas et al., 2000; Fu et al., 2007; Alphey et al., 2008).
An intriguing use of transgenesis technology was proposed as a concept, termed the ‘flying vaccinator’. This idea shifts a hematophagous insect from a pest to a beneficial insect (Crampton et al., 1999). The ‘flying vaccinator’ strain was not developed previously in mosquitoes because until recently, a salivary gland-specific expression system was not available. We identified and characterized an abundant protein in the salivary glands of the malaria vector mosquito, Anopheles stephensi. As the saliva protein binds directly to collagen and inhibits collagen-induced platelet aggregation (Yoshida et al., 2008), it was named anopheline anti-platelet protein (AAPP). The upstream region of the aapp gene open reading frame has the capacity to direct the strong female salivary gland-specific and blood meal-inducible expression of a reporter gene. A transgenic A. stephensi mosquito harbouring a gene encoding tetrameric red fluorescent protein (tDsRed) under the control of the aapp promoter specifically expressed tDsRed protein as a cytoplasmic form in the secretory cells of the distal–lateral lobes of the female salivary glands (Yoshida & Watanabe, 2006). These transgenic mosquitoes facilitated investigations of the mechanisms of sporozoite invasion into the salivary glands using in vivo imaging. Thus, we have established a robust salivary gland-specific expression system in an anopheline mosquito.
We report here the development of ‘flying vaccinator’ strains of transgenic A. stephensi that secrete a transgene product as a component of saliva. We chose the gene encoding the SP15 antigen, which is a 15 kDa saliva protein of the sand fly Phlebotomus papatasi, as the expressed vaccine candidate molecule. P. papatasi saliva enhances Leishmania infection when the parasites are transmitted to their vertebrate hosts by sand fly feeding (Kamhawi et al., 2000). Valenzuela and colleagues showed in a murine model that vaccination with SP15 conferred protection to a subsequent Leishmania infection (Valenzuela et al., 2001) and are developing a DNA vaccine containing the sp15 cDNA. We generated transgenic mosquito lines expressing SP15 fused to a secretory form of monomeric DsRed (mDsRed) in the salivary glands. Mice bitten repeatedly by the transgenic mosquitoes raised anti-SP15 antibodies, indicating the achievement of the original concept of the ‘flying vaccinator’. Although it is unlikely that this concept will be developed as a public health measure for delivering vaccines, the salivary gland-specific expression of relevant proteins is expected to be a powerful tool for the elucidation of saliva–malaria sporozoite interaction.
Results and discussion
Generation of transgenic lines
A Minos-based transfer vector comprising a gene cassette consisting of the sp15 gene fused to the mDsRed gene (mDsRed-sp15) under the control of the aapp promoter (Fig. 1A) was injected together with a Minos helper plasmid into A. stephensi embryos. One-hundred and thirty-five embryos of the 505 injected (27%) hatched normally G0 generation and 54 adults were eclosed. Three independent transgenic lines were established. A single integration event was confirmed by Southern blot analysis (Fig. 1B). The three transgenic lines have been stably maintained by blood meals on mice for over 10 generations.
Expression of mDsRed-SP15 in the salivary glands
Immunoblot analysis of all three transgenic lines with anti-DsRed and anti-SP15 antibodies verified the presence of expressed mDsRed-SP15 protein [molecular mass (Mr) = 44 kDa] in the salivary glands (Fig. 2). Anti-DsRed antibody recognized a protein (Mr = 28 kDa) in the salivary glands of the tDsRed transgenic mosquitoes, which express the tDsRed gene without the sp15 gene, and no signal was detected with anti-SP15 antibody. The expression levels of mDsRed-SP15 protein in the salivary glands were indistinguishable among the three lines following calibration with anti-AAPP antibody-recognition of native AAPP (Mr = 37 kDa). This is consistent with the interpretation that there are no gross position effects based on this analysis. The amount of mDsRed-SP15 protein expressed is estimated at ∼20 ng per pair of salivary glands, as quantified by comparison with commercially available tDsRed protein. This expression level was equivalent to as much as 4% of the total salivary gland protein, which was similar to that of tDsRed protein in the corresponding transgenic mosquitoes (25 ng per pair of salivary glands; Yoshida & Watanabe, 2006).
Location of mDsRed-SP15 protein in the salivary glands
Strong red fluorescence was observed in the distal–lateral lobes of dissected female salivary glands of both the tDsRed and mDsRed-SP15 transgenic mosquitoes. This is consistent with the endogenous expression profile mediated by the aapp promoter (Fig. 3A–D). A detailed analysis using 3D confocal microscopy revealed distinct differences in the distribution of DsRed proteins between the two transgene constructs. mDsRed-SP15 protein was observed in the cytoplasm and secretory acini of the cells in the lobes (Fig. 3E), whereas tDsRed protein was detected only within the secretory cells (Fig. 3F). Furthermore, incubation of the distal–lateral lobes with SYTO-13 green-fluorescent dye, which stains DNA and RNA, facilitated visualization of the shape of the secretory cells (Figs 3G–J, S1). Overlaying the red and green fluorescence patterns showed that tDsRed protein was located in the secretory cells (yellow), indicating a cytoplasmic form. In contrast, mDsRed-SP15 protein was located in the secretory cavities and ducts of the salivary glands, consistent with secretion into the acini.
To obtain supportive data on the localization of the expressed mDsRed proteins, the dissected salivary glands of the transgenic mosquitoes were physically ruptured under fluorescent microscopy. Although a trace of tDsRed protein spilt out from the sites of rupture, the intensity of red fluorescence remained strong (Fig. 4, Videos S1 and S2). In contrast, a large amount of mDsRed-SP15 protein was clearly released into the dissecting medium following damage. These results are consistent with the interpretation that mDsRed-SP15 protein is secreted and accumulates in the secretory acini and duct as a component of saliva. The release of saliva with red fluorescence during salivation from the proboscis supports this conclusion (Fig. 5A and Video S3). In contrast to mDsRed transgenic mosquitoes, saliva with red fluorescence was not released from proboscis of tDsRed transgenic mosquitoes. As a result, mice bitten repeatedly by tDsRed transgenic mosquitoes induced no antibody response to DsRed (data not shown). The difference in the localization between the two transgene products is likely to be because of protein structure, including a signal leader peptide, and conformation. tDsRed protein is a tetramer with no signal leader sequence, resulting in a cytoplasmic form with low solubility. mDsRed-SP15 protein is a monomer with a signal leader peptide, resulting in a secreted form with high solubility. Transgenic mosquitoes harbouring the gene encoding nonfusion SP15 with the same signal peptide under the control of the aapp promoter expressed undetectable levels of SP15, although the SP15 transcripts were detected in the salivary glands by real time-PCR (Fig. S2). So far, we have failed to express non-mDsRed fusion constructs at levels equivalent to the mDsRed-SP15 protein. These results support the conclusion that the mDsRed-fusion expression system facilitates not only the monitoring of secretion of transgene products but also stabilizes and/or chaperones mDsRed-fusion proteins prior to their secretion.
Antibody responses to mDsRed-SP15 in mice exposed to transgenic mosquitoes
Immunoblot analyses showed that sera from four individual mice bitten repeatedly by ∼1500 mDsRed-SP15 mosquitoes recognized Glutathione S-transferase (GST)-fusion mDsRed-SP15 protein (Fig. 5B). Anti-SP15 antibody titres were determined by enzyme-linked immunosorbent assay (ELISA) using recombinant SP15 protein (rSP15) and SP15 synthetic peptides (sySP15) as antigens. Levels of anti-SP15 antibodies induced in all of the four sera were not high, but statistically different from sera obtained from mice bitten by wild-type (WT) mosquitoes, which had undetectable levels of anti-SP15 antibodies (Fig. 5C). One of the four experimental mice (mouse no. 4) had sera that reacted with the sySP15 (Fig. 5D). These combined results support the conclusions that immunogenically active mDsRed-SP15 was injected as a component of saliva via blood feeding and induced SP15-specific antibodies.
A critical question remains to be answered for the present study concerning the flying vaccinator. It is whether mice exposed to mDsRed-SP15 mosquitoes are protected when challenged by Leishmania infection. Challenge studies using Leishmania-infected P. papatasi are required to answer this question. The mice immunized with SP15 conferred protection to Leishmania major infection, and SP15 is recognized by people naturally exposed to sand flies in endemic areas of Le. major (Valenzuela et al., 2001). Additionally, following exposure to another sand fly, Lutozomia longipalpis, lymphocytes reduced the infection rate of Leishmania in human macrophages in vitro (Vinhas et al., 2007). Therefore, exposure to mDsRed-SP15 mosquitoes may confer resistance in humans to Leishmania infection. We are now proceeding to evaluate vaccine efficacy induced by the mosquito bites against challenge with Leishmania-infected P. papatasi in a murine model.
In summary, we have demonstrated that transgenic mosquitoes secreting a foreign protein as a component of saliva can elicit immune responses, and therefore demonstrate proof-of-principle for the concept of the ‘flying vaccinator’. Although the model we developed targets a pathogen, Leishmania, for which a candidate vaccine exists, mosquitoes secreting antigens to combat malaria would be preferred. However, the concept of a ‘flying vaccinator’ transgenic mosquito is not likely to be a practicable method of disease control, because a ‘flying vaccinator’ is an unacceptable way to deliver vaccine without issues of dosage and informed consent against current vaccine programmes. These difficulties are further complicated by the issues of public acceptance to release of transgenic mosquitoes. Therefore, we intend only that the present study makes available a model system using a salivary gland-specific promoter as a potential tool to elucidate the saliva–malaria sporozoite interactions.
The A. stephensi mosquito strain, SDA500, was maintained at Jichi Medical University. Female C57/BL6 mice were obtained from Japan SLC (Hamamatsu, Shizuoka, Japan) and used at 7–8 weeks of age.
Minos vector construction and germline transformation
PCR reactions were performed with Pfu DNA polymerase (Stratagene GmbH, Heidelberg, Germany). A 366-bp DNA fragment of the SP-15 gene was amplified from pcDNA-SP15 (kindly provided by J. G. Valenzuela) using primers pSP15-F2 and pSP15-R2. The PCR product was cloned into pENTR/D-TOPO (Invitrogen, Carlsbad, CA, USA) to generate pENTR-SP15. A 675-bp DNA fragment of the mDsRed gene was amplified from pDsRed-monomer-C1 (BD Biosciences Clontech, Mountain View, CA, USA) using primers pmDsRed-F1 and pmDsRed-R1. The PCR product was cloned into pENTR/D-TOPO (Invitrogen) to generate pENTR-mDsRed. The mDsRed gene fragment was excised from pENTR-mDsRed by digestion with EcoRI and MunI, and then cloned into the EcoRI site of pENTR-SP15 to generate pENTR-mDsRed-SP15-XhoI. A 1.1 kb DNA fragment of the mDsRed-sp15 gene was excised from pENTR-mDsRed-SP15-XhoI by digestion with EcoRI and XhoI, then cloned into the EcoRI/XhoI sites of pENTR-aappP-tDsRed-antryp1T (Yoshida & Watanabe, 2006) to generate pENTR-aappP-mDsRed-SP15-antryp1T. The transformation plasmid pMinos-EGFP-aappP-mDsRed-SP15-antryp1T (Fig. 1A) was generated by incubation of pMinos-EGFP-RfA-F (Yoshida & Watanabe, 2006) and pENTR- aappP-mDsRed-SP15-antryp1T in the presence of LR Clonase (Invitrogen) as described previously (Yoshida & Watanabe, 2006). Primer sequence information is available in Table S1. Procedures for microinjection into embryos, screening of EGFP-expressing G1 larvae and generation of homozygous lines have been described previously (Catteruccia et al., 2000).
Southern blot analysis
A total of 10 µg of genomic DNA extracted from female mDsRed-SP15 mosquitoes was digested with EcoRI, and separated on 0.8% agarose gels, and then transferred to a Hybond N+ membrane (GE Healthcare Bioscience, Buchingamshire, England). For hybridization, a 0.9 kb EcoRI-XmaI DNA fragment partially encoding mDsRed-SP15 protein was used as a probe. Probe labelling, hybridization, washing and detection of the signals were performed under high stringency conditions using a DIG-High Prime (Roche, Indianapolis, IN, USA) according to the supplier's protocol.
Anti-SP15 antiserum was obtained from rabbit immunized with a synthetic SP15-peptide (sySP15: AKKVKKQAREDSHWLNC). Anti-AAPP antiserum was obtained from rabbits immunized with recombinant AAPP protein (Yoshida et al., 2008). Rabbit anti-DsRed polyclonal antibody was purchased from Clontech Laboratories, Inc. (Mountain View, CA, USA).
Groups of 20 pairs of female salivary glands were homogenized by a plastic homogenizer with 40 µl Laemmili buffer (Laemmli, 1970) containing 5% 2-mercaptoethanol, and then boiled at 95 °C for 5 min. Ten microlitres of each sample (equivalent to five pairs of salivary glands) was separated on a 10 or 12% sodium dodecyl sulphate polyacrylamide gel electrophoresis gel, and transferred to Immobilon Transfer Membrane (Millipore, Bedford, MA, USA). The membrane was incubated either with anti-sySP15, anti-AAPP antiserum, or anti-DsRed antibody. Polypeptides recognized by the antibodies were detected with biotinylated anti-mouse or anti-rabbit immunoglobulin G (IgG) (Vector Laboratories, Burlingame, CA, USA), followed by colour development with 5-bromo-4-chloro-3-indolylphosphate p-toluidine salt/nitroblue tetrazolium chloride substrate (Roche) as described previously (Yoshida et al., 1999).
Following dissection in phosphate-buffered saline (PBS), the female salivary glands were immediately stained with 0.2 mM SYTO-13 (Invitrogen) in PBS for 5 min, and then washed twice with PBS. Stained samples were mounted on a glass slide. DsRed and SYTO-13 fluorescence in the salivary glands were observed by confocal microscopy [FV1000 (Olympus, Tokyo, Japan) and TCS SP5 (Leica, Wetzlar, Germany)].
For the observation of physical rupture of the salivary gland in two transgenic lines, the female salivary glands were dissected in PBS and put between a slide and cover slip, and then pressure was applied using tweezers under a fluorescent microscope AX80R (Olympus). Real-time imaging was monitored and recorded by DP controller software (Olympus).
To monitor the release of saliva with DsRed fluorescence from the proboscis, mosquitoes were kept on ice, their legs and maxillary palps were removed, and their abdomens were immobilized on a glass slide with adhesion bond. Salivation was observed under a Biorevo fluorescent microscope (Keyence, Osako, Japan) immediately after inserting the proboscis into 70% glycerol.
Immunization of mice with transgenic mosquito bites
mDsRed-SP15 transgenic and WT mosquitoes were allowed to repeatedly feed on individual mice 42 times at 4-day intervals with a total of approximately 1500 bites.
The sp15 gene was excised from pENTR-SP15 by digestion with NcoI and XhoI, and cloned into the NcoI/XhoI sites of pET32-b(+) (Novagen, Madison, WI, USA) to generate pET32-SP15. A recombinant SP15 protein (rSP15), created as a fusion protein with thioredoxin, was expressed in Escherichia coli. A 684-bp DNA fragment of the glutathione s-transferase (gst) gene was amplified from pGEX-6P-1 (GE Healthcare Biosciences) by PCR using primers pGEX6P-F2 and pGEX6P-R2. Primer sequence information is available in Table S1. The PCR product was cloned into pENTR/D-TOPO (Invitrogen) to generate pENTR-GEX6P2. The gst gene fragment was excised from pENTR-GEX6P2 by digestion with NdeI and NcoI, and then cloned into the NdeI/NcoI sites of pET22-b (+) (Novagen) to generate pET22-GSX6P2. A 1.1 kb DNA fragment of the mDsRed-SP15 fusion gene was excised from pENTR-mDsRed-SP15-XhoI by digestion with EcoRI and XhoI, and then cloned into the EcoRI/XhoI sites of pET22-GSX6P2 to generate pET22-GEX6P2-mDsRed-SP15. A recombinant mDsRed-SP15 protein, created as a fusion protein with GST, was expressed in E. coli. Both proteins were purified using a Ni-NTA affinity column (Qiagen, Valencia, CA, USA) as described previously (Daly & Long, 1993), and used as antigens for immunoblotting and ELISA.
Sera from individual mice bitten repeatedly either by mDsRed-SP15 or WT mosquitoes were collected from the tail vein. rSP15- and sySP15-specific antibody titres were quantified by ELISA. Pre-coated 96-well plates with either rSP15 or sySP15 at 100–300 ng/well in carbonate buffer (15 mM Na2CO3, 35 mM NaHCO3, pH 9.6) were incubated with serially diluted sera. Specific IgGs were detected using horseradish peroxidase-conjugated goat anti-mouse IgG (H+L) (Bio-Rad, Hercules, CA, USA). The plates were developed with a peroxidase substrate solution [H2O2 and 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulphonate)]. The optical density at 414 nm of each well was determined using a Multiskan BICHROMATIC microplate reader (Thermo Labsystems, Vantaa, Finland). End-point titres were expressed as the reciprocal of the highest sample dilution for which the optical density was equal to or greater than the mean optical density of naive mouse sera.
We thank J. Sato, K. Watano, H. Araki and T. Yokomine for handling of mosquitoes and mice; S. Mizogami (KEYENCE) and H. Kato (Leica microsystems K.K.) for fluorescent microscopy technical assistance; H. Matsuoka, M. Hirai and G.I. Sano for their hospitality to D. S. Y. We also thank A. A. James for critical review of the manuscript. We are indebted to J. G. Valenzuela providing us with plasmids containing the P. papatasi sp15 gene. This work was supported by grants from the Ministry of Education, Culture, Sports and Science of Japan (21390126).