Is Saglin a mosquito salivary gland receptor for Plasmodium falciparum?

Anopheles gambiae G3 and An. stephensi SDA 500 were used in these studies and were reared as adults at 28 °C, 80% humidity. Larvae were fed ground TetraMin^® Tropical Flakes ad libitum. Adults were fed 10% sucrose solution and adult females fed on mice to support mosquito colony maintenance.

Mosquito transformation was performed in the University of Maryland’s Insect Transformation Facility (http://​ibbr.​umd.​edu/​insect_​transformation) using standard protocols [37, 38]. A piggyBac vector with a 3xP3DsRed marker gene and a transgene consisting of the promoter from the anopheline antiplatelet protein (AAPP) gene from An. stephensi regulating the expression of the Saglin open reading frame and the 3′ untranslated region (3′UTR) of the SV40 virus was constructed and introduced into the genome of An. stephensi [39, 40]. AAPP is expressed specifically in the distal lateral lobes of the salivary glands of female An. stephensi [40]. The region 5′ of the open reading frame of AAPP (nucleotides 4–1687; accession number A212871) which includes the promoter was isolated from An. stephensi using PCR and was used to ensure that Saglin is expressed specifically in the distal lateral lobes of the salivary glands of female mosquitoes. A 1541 bp fragment of DNA was synthesized consisting of 1239 bp (412 aa) of the Saglin open reading frame (nucleotides 1–1239; AY846632) and 302 bp containing the 3′ UTR of SV40 (GenScript Corporation, Piscataway, NJ). The promoter and transcription unit were assembled in pSL1180fa and then transferred as single fragment into the piggyBac 3xP3DsRed vector [41].

The presence and location of integrated gene vectors in transgenic mosquitoes was determined by inverse PCR [42]. Genomic DNA from the equivalent of one mosquito was digested to completion with restriction endonuclease MspI or HaeIII according to the manufacturer’s recommendation (New England Biolabs) in a reaction volume of 50 μL. On completion of the digestion step, the restriction endonuclease was heat inactivated. The digested genomic DNA was then ligated under dilute conditions to bias the formation of monomeric circular products. The final ligation reaction was 300 μL and contained all the digested DNA from the 50 μL restriction digest and 40 units of ligase (New England Biolabs). Ligation was performed overnight at 16 °C. The DNA in the ligation reaction was precipitated, washed and resuspended in 15 μL of sterile distilled water. One half of the DNA was then used as a template in two identical PCR reactions except for the presence of different primers. To amplify the 5′ terminus of the integrated piggyBac element and flanking genomic DNA the primers used were 5′-TCT TGA CCT TGC CAC AGA GG-3′ and 5′-TGA CAC TTA CCG CAT TGA CA-3′. To amplify the 3′ terminus of the integrated piggyBac element and flanking genomic DNA the primers used were 5′-AAA CCT CGA TAT ACA GAC CGA TAA AAC AC-3′ and 5′-CAT TTG CCT TTC GCC TTA TTT TAG A-3′. The PCR cycle conditions were: one cycle at 95 °C for 3 min, followed by 30 cycles of 95 °C for 30 s, 52 °C for 30 s, 72 °C for 2 min and finally one cycle at 72 °C for 5 min. The DNA sequences of the PCR products were determined.

Plasmodium falciparum strain NF54 cultures were initiated from vials of cryopreserved parasites and maintained at 5% haematocrit in type O+ human red blood cells (RBCs) and complete growth medium (RPMI-1640 supplemented with type O+ human serum and hypoxanthine) with passage of cultures every 3–4 days [25, 43]. Gametocytes were induced by maintaining the cultures with daily complete growth medium replacement but without the addition of uninfected erythrocytes for 14–18 days. Cultures were selected for feeding at 14–18 days post gametocyte induction based upon the proportion of stage V gametocytes present in the culture. Stage V gametocytes were combined with uninfected type O+ human erythrocytes and human O+ serum and fed to mosquitoes via an artificial membrane feeder. Mosquitoes were then maintained at 25 ± 2 °C and 70–80% relative humidity on water and solid sugar until dissection to estimate P. falciparum infection rates at either the oocyst (days 7–9 post feeding) or PfSPZ (days 16–18 post feeding) stage.

Transgenic An. stephensi with the AAPP-Saglin transgene (As-AAPPSaglin) were outcrossed to wild type An. stephensi and transgenic heterozygous progeny were selected and separated according to sex. These heterozygotes were again outcrossed to wild type An. stephensi and female heterozygous transgenic and wild type progeny were infected with P. falciparum. Approximately 300 female mosquitoes (~ 150 heterozygous transgenic, ~ 150 non-transgenic) were placed in each of three cages and fed P. falciparum gametocytes as described above. After 7–9 days at 25 ± 2 °C, approximately 150 mosquitoes were removed from each cage, separated into those expressing DsRed (transgenic) and those without DsRed expression (non-transgenic) and then dissected to remove the midguts. The midguts were then examined by microscopy for the presence of oocysts and the numbers recorded. At 16–18 days post-feeding the remaining mosquitoes in the three cages were separated based on their genotypes (transgenic vs non-transgenic), salivary glands were isolated and the numbers of PfSPZ in each were counted using a hemocytometer.

Salivary glands were dissected in phosphate buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 2 mM KH₂PO₄, pH 7.4), and fixed in 4% paraformaldehyde in PBS for 15 min at ambient temperature. Fixative was removed and the tissues washed for 3 × 5 min in methanol. Methanol was removed and replaced with 200 μL of a 1:10 solution of H₂O₂ and methanol for 15 min at room temperature, then the tissues were washed 3 × 15 min in a 1:1 mixture of methanol and PBS with 0.1% TritonX100 (PBST). The tissues were washed in PBST with 1% bovine serum albumin (PBSBT) 5 × 5 min and then blocked in PBSBT for 1 h at ambient temperature. The tissues were again washed in PBST 3 × 5 min at which point the An. gambiae Saglin-specific mAb2A3, diluted 1:200 in PBST, was added to the tissue and incubated overnight at 4 °C [30]. The mAb2A3-containing solution was removed and the tissues were washed with PBSBT 3 × 5 min followed by 3 × 15 min washes. Alexa Fluor^® 555 goat anti-mouse antibody (A31622, Invitrogen, Eugene, OR) was diluted 1:200 in PBST, added to the tissues and incubated for 2 h at ambient temperature in the dark. The Alexa Fluor^® 555 Goat Anti-Mouse antibody-containing solution was removed and the tissues were washed PBSBT for 3 × 5 min followed by 3 × 15 min washes. When required, the first 15-min wash with PBSBT also contained 300 nM DAPI (4′, 6-diamidino-2-phenylindole, 46190, Pierce, Rockford, IL). The processed salivary glands were mounted on glass slides under coverslips in VECTASHIELD^® Mounting Medium (Vector Laboratories, Burlingame, CA) and observed with either a Zeiss Axio Imager.A1 under phase contrast or fluorescence illumination using the Rhodamin-shift-free filter set (EX BP 546/12, BS FT 560, EM BP 575–640) or with a Zeiss 710 laser scanning confocal microscope in which the 405 nm and 561 nm lasers were used for fluorochrome excitation. Indirect immunofluorescence experiments was repeated at least four times and each experiment always involved dissecting approximately 10–15 salivary glands from females of non-transgenic An. stephensi, transgenic An. stephensi and non-transgenic An. gambiae and all glands were processed in parallel.

Salivary glands from unfed adult female An. gambiae at 4 days post eclosion were dissected in cold PBS. The medial glands were separated from the lateral glands (proximal + distal regions) and placed immediately in TRIzol^® Reagent (Invitrogen). Three pools of glands (each with approximately 10–15 medial or lateral lobes) were collected (biological replicates) and RNA was isolated according to the manufacture’s recommendations (Invitrogen). Complementary DNA (cDNA) was synthesized using the SuperScript^® III First-Strand Synthesis System for RT-PCR according to the manufacturer’s recommendations (Invitrogen). qRT-PCR was performed using an Applied Biosystems 7300 Real-Time PCR System using Saglin primers SG746F: 5′-CGA CCT TGT CCG GCA GTC CG-3′ and SG1089R: 5′-CTG CCG TGC CGC CTT TAC CA-3′, and ribosomal protein, S7 primers (normalization control); AsS7f: 5′-TGC GGC TTC AGA TCC GAG TTC-3′ and AsS7r 5′-TTC GTT GTG AAC CCA AAT AAA AAT C-3′. All cDNA samples were analysed in triplicate (technical replicates) under the following cycle conditions in the presence of SYBR Green: 95 °C for 1 min followed by 39 cycles consisting of 95 °C for 45 s, 55 °C for 45 s, 72 °C for 45 s. Following the PCR reaction, the products were subjected to a dissociation analysis to detect the presence of non-specific primer-dimers and a sample of each reaction was analysed by agarose gel electrophoresis in the presence of ethidium bromide staining to verify visually that products of only the expected sizes were present and to confirm the absence of primer-dimers. The resulting real time PCR data were analysed using the comparative C_T method [44] and statistical analysis was performed using an analysis of variance (ANOVA).

Salivary glands from female An. gambiae, An. stephensi and transgenic An. stephensi containing transgene As-AAPPSaglin were dissected in cold PBS at 7 days post emergence. Approximately 40 medial and lateral (proximal + distal) glands were separated, and placed separately in 50 μL of Lämmli sample buffer (BioRad, Hercules, CA) with 5% β-mercaptoethanol and homogenized. Equal numbers of medial and lateral lobe equivalents (10 or 25) were loaded into wells and fractionated in a 10% Tris HCL polyacrylamide gel (BioRad) using a broad range Prestained Protein Marker (New England BioLabs, Ipswich, MA) as a molecular weight standard and transferred by electro-blotting to a polyvinylidene fluoride membrane (BioRad). Blots were probed simultaneously with mab2A3 (1:500) and an actin antibody that recognizes all actin isoforms (loading control; rabbit pan-actin antibody; 1:1000; 4968; Cell Signaling Technology, Danvers, MA). Blots were subsequently probed with secondary antibodies specific to mouse (Goat αMouse Cy5, PA45009V) or rabbit (Goat αRabbit Cy3, 28901106V) (GE Healthcare Piscataway, NJ) immunoglobulin G, each diluted at 1:2500. Blots were visualized using a Typhoon Trio™ Variable Mode Imager (GE Healthcare).

Saglin in the salivary glands of female An. gambiae displayed consistent temporal and spatial patterns of protein expression. Salivary glands from female An. gambiae 3, 4 and 12 days post emergence showed evidence of abundant Saglin only in the medial and proximal lateral lobes (Fig. 1). Saglin protein was undetectable in the distal lateral lobes of female salivary glands under all conditions tested. The physiological state of females with respect to blood feeding and oogenesis had no detectable effect on the spatial pattern of Saglin expression; expression remained confined to the medial and proximal lateral lobes of the salivary glands (Fig. 1).

A confocal microscopy analysis of Saglin distribution in salivary glands from females revealed a number of distinctive features of this protein. Saglin was clearly localized exclusively to the basal surface of the cells of the medial lobe with no reactivity along the lateral or apical cell surfaces or within the secretory cavity and duct (Fig. 2). The pattern of Saglin distribution was notably different in the cells of the proximal lateral lobes of the salivary gland. The proximal cells of the lateral lobes abundantly expressed Saglin that, based on the relative fluorescence intensity, was comparable to that seen in the medial lobe. Saglin expression terminated abruptly at the transition between proximal cells (no secretory cavities) and distal cells (with secretory cavities) (Fig. 1). Saglin was seen uniformly throughout the proximal cells of the lateral lobes (Fig. 2) and was not confined to the basal cell surface as in the cells of the medial gland.

The number of transcripts in the medial lobe of salivary glands from 5 day old adult female An. gambiae was over 1000-fold higher than of S7 transcripts while in the lateral lobes the number of transcripts of Saglin were only approximately 50-fold higher than those of S7 (F_1,4=189.881, P = 0.00016) (Fig. 3). In these experiments RNA was isolated from the entire lateral lobe of the salivary glands including the morphologically distinct proximal and distal regions. Saglin protein expression, as shown by Western blot, was much higher in the medial lobe of the salivary gland compared to the distal lobes, which was consistent with qRT-PCR and indirect immunofluorescence data.

A transgenic line of An. stephensi was created that contained three inserts of the gene vector containing a transgene consisting of the Saglin open reading frame under the regulatory control of the promoter from AAPP. Inverse PCR was used to isolate and sequence the DNA immediately adjacent to the inverted terminal repeats of the piggyBac vector (Fig. 4). Vector integration occurred through typical piggyBac transposition as indicated by the presence of TTAA direct repeats adjacent to the terminal inverted repeats of the transposable element (Fig. 4). Strong DsRed marker-gene expression was observed in the brain, ventral nerve cord and anal papillae of larvae as is expected based on the known pattern of expression of the 3XP3 promoter in insects. That marker gene expression was observed exclusively in the tissues known to express 3XP3 indicating that this line did not suffer from position effects (Fig. 4). In adult transgenic An. stephensi, Saglin protein was detected only in the cells of the distal region of the lateral lobes (Fig. 4). The fluorescence intensities observed in co-processed salivary glands of transgenic An. stephensi and An. gambiae were visually similar. Saglin was detected exclusively on the basal surfaces of the cells of the distal lateral lobes with no evidence of protein expression along the lateral and apical surfaces or in secretory cavities or ducts (Fig. 5).

In two independent experiments, transgenic and wild type An. stephensi were offered blood meals prepared from the same P. falciparum gametocyte cultures; within each experiment, three pairs of treatments (wild type versus Saglin knock-in An. stephensi) were compared. There were no significant differences in P. falciparum oocyst or PfSPZ infection characteristics of heterozygous transgenic and wild type An. stephensi (Fig. 6). Geometric mean oocyst intensities did not differ significantly between control and experimental groups in any of the feeds (Mann–Whitney U test between control and transgenic mosquitoes for each paired feed). PfSPZ infections of salivary glands were also compared statistically in the same groups of mosquitoes: neither prevalence rates nor infection intensities differed significantly between the wild type and transgenic An. stephensi in any of the three feeds.