Figure 1.
SM1 binds to a salivary gland protein.
In all experiments the lectin (WGA) was labeled with FITC (green fluorescence) and binding of the biotinylated SM1 peptide was detected with Texas red-labeled streptavidin (red fluorescence). (A) Schematic diagram of the lifecycle of malaria parasite in the mosquito vector. After initial development in the midgut lumen, motile ookinetes (ook) cross the midgut epithelium and differentiate into sessile oocysts (ooc). Mature oocysts release sporozoites (spo) into the hemocoel and these specifically recognize the salivary glands (circled in the inset) resulting in invasion. (B) SM1 structure. The 12-amino acid SM1 peptide has two cysteine residues at positions 2 and 11 that when connected by disulphide bond form an 8-amino acid loop. (C) Binding of SM1 to salivary glands does not involve surface carbohydrates. Control salivary glands treated with buffer without periodate showed strong lectin binding (a). Panels (b–d) show an experimental gland treated with periodate and then incubated with a mixture of WGA and SM1. WGA bound poorly to this gland [b], indicating efficient removal of carbohydrates, while SM1 bound strongly (c), indicating that carbohydrate removal did not hinder SM1 binding. A differential interference contrast (DIC) image of the same gland is shown in (d). (D) Protease treatment abrogates SM1 binding. Salivary glands were treated with 0 (control; a, b), 25 µg/ml (c, d) or 100 µg/ml (e, f) of trypsin and were incubated with a mixture of SM1 and WGA. Protease treatment interfered with SM1 but not with lectin binding.
Figure 2.
An anti-SM1 antibody recognizes a protein component of midgut sporozoites.
(A) Western blots with the anti-SM1 antibody. The same number of midguts from Plasmodium-infected (carrying mature oocysts) or non-infected mosquitoes (mock) was used for gradient purification of sporozoites. Either 5×105 sporozoites (lanes 1 and 3) or an equivalent amount of material from the mock purification (lanes 2 and 4) was fractionated by SDS-PAGE and blotted. One blot was incubated with anti-SM1 antibody (immune) and the other with pre-immune serum, as indicated. The double arrows point to the bands that are specifically detected by the immune serum. (B) Indirect immunofluorescence staining (IFA). Fixed sporozoites were incubated with anti-SM1 or pre-immune serum followed by incubation with an FITC-labeled (green) secondary antibody. Nuclei were labeled with DAPI and appear blue. (a–c) Methanol-fixed (permeabilized) midgut sporozoites incubated with anti-SM1 antibody. (d–e) Paraformaldehyde-fixed (non-permeabilized) midgut sporozoites incubated with anti-SM1 antibody. (g–i) Methanol-fixed (permeabilized) midgut sporozoites incubated with pre-immune serum. Panel (i) is a DIC image of the same field as (g) and (h). (C) Western blots of 2D gels with the anti-SM1 antibody. Two aliquots of 106 (samples 1 and 2) and one aliquot of 107 (sample 3) sporozoites from the same enriched sporozoite preparation were fractionated in parallel by 2D gel electrophoresis. Gels from samples 1 and 2 were blotted onto a membrane and incubated with an anti-SM1 antibody (panel C) or with pre-immune serum (not shown). Sample 3 was stained with Coomassie Blue (not shown). Three protein spots (A–C) with mobility of ∼90 kDa (c.f. panel B) that reacted with the anti-SM1 antibody were excised from the stained gel for protein sequence analysis. (D) Western blots with anti-TRAP and anti-SM1 antibodies. Equal amounts of recombinant GST (lanes 1 and 3) and GST-tagged recombinant TRAP domain-A (lanes 2 and 4) proteins were fractionated by SDS-PAGE and the proteins blotted. The blot shown in the left panel was incubated with anti-TRAP domain-A antibody (the antibody was made against a GST-tagged recombinant protein) and the blot shown on the right panel was incubated with anti-SM1 antibody. The double arrows point to the TRAP domain-A bands recognized by the both antibodies.
Figure 3.
Pull-down of salivary gland proteins that interact with SM1.
(A) Schematic diagram of the pull-down approach. A double-derivatized SM1 peptide, with a biotin residue (yellow pentagon) at the N-terminus and a UV-crosslinking residue (green star) attached to a phenylalanine residue in the loop (‘F’ in Figure 1B) was incubated in vitro with freshly dissected salivary glands and washed. After shining UV to promote crosslinking of the peptide to the protein to which it was bound, the salivary glands were lysed and the peptide with the crosslinked protein was captured on streptavidin beads. The beads were washed and the retained proteins were fractionated by SDS-PAGE (panel B). (B) Gel electrophoresis of proteins captured on the streptavidin beads. Materials recovered from a pull-down experiment illustrated in panel A were fractionated by SDS-PAGE under reducing conditions and the gel was stained with Coomassie Blue. Lane 1, Materials eluted from streptavidin beads that were not incubated with any proteins. The stained bands are presumed to be bead-derived contaminants. Lane 2, Complete experiment, except that the crosslinking step was omitted. Lane 3, complete experiment, including the crosslinking step (same number of salivary glands as in Lane 2). The four arrows point to protein bands consistently observed only in the complete experiment.
Figure 4.
In vitro and in vivo protein interactions.
(A) SM1 binding to saglin on protein blots. Blots from bacteria carrying a plasmid encoding either histidine-tagged saglin (experimental; panels 1–4) or histidine-tagged ß-galactosidase (control; panels 5–6) were produced from proteins recovered either before (U) or after (I) induction of recombinant protein synthesis. Panel 1: Experimental blot incubated with anti-histidine antibody. Panel 2: Experimental blot incubated with anti-saglin monoclonal antibody. Panel 3: Experimental blot first incubated with the SM1 peptide (10 µg/ml) and then probed with the anti-SM1 antibody followed by incubation with an alkaline phosphatase-tagged secondary antibody. Panel 4: Experimental blot incubated with a linear SM1 peptide and probed with anti-SM1 antibody. In the linear peptide the two cysteines were substituted by alanines. In separate experiments the linear peptide was unable to bind to salivary glands (not shown). Panel 5: Control blot probed with anti-histidine antibody. Panel 6: Control blot first incubated with the SM1 peptide (10 µg/ml) and then probed with the anti-SM1 antibody. The mobility of recombinant saglin is indicated by an arrow in Panel 1. (B) TRAP domain-A binding to saglin on protein blots. Blots from bacteria carrying a plasmid encoding either histidine-tagged saglin (experimental; panels 1–3) or histidine-tagged ß-galactosidase (control; panels 4–5) were produced from proteins recovered either before (U) or after (I) induction of recombinant protein synthesis. Panel 1: Experimental blot incubated with anti-histidine antibody; Panel 2: Experimental blot first incubated with recombinant TRAP domain-A protein (2.5 µg/ml) and then probed with anti-TRAP domain-A antibody, followed by incubation with an alkaline phosphatase-tagged secondary antibody; Panel 3: Same as Panel 2, except that incubation with anti-TRAP domain-A antibody was omitted; Panel 4: Control blot probed with anti-histidine antibody; Panel 5: Control blot subjected to the same treatment as described for Panel 2. The mobility of recombinant saglin is indicated by an arrow in panel 1. (C) In vitro interaction between saglin and TRAP domain-A. Recombinant TRAP domain-A was incubated in wells of an ELISA plate that had been previously coated with recombinant histidine-tagged saglin. The wells were washed and TRAP binding was quantified using a TRAP antibody followed by incubation with an alkaline phosphatase-tagged secondary antibody. Results of the colorimetric alkaline phosphatase reaction are reported. (1) Well not coated with recombinant saglin. (2) Complete protocol. (3) Incubation with TRAP omitted. (4) Incubation with anti-TRAP antibody omitted. (5) Incubation with secondary antibody omitted. (D) Sporozoites bind to saglin in vitro. Purified midgut sporozoites were incubated in wells of an ELISA plate that had been previously coated with recombinant histidine-tagged saglin. The wells were washed and sporozoite binding was quantified by use of an anti-CS antibody followed by incubation with an alkaline phosphatase-tagged secondary antibody. Results of the colorimetric alkaline phosphatase reaction are reported. (1) Well not coated with recombinant saglin. (2) Complete protocol. (3) Incubation with sporozoites omitted. (4) Incubation with anti-CS antibody omitted. (5) Incubation with secondary antibody omitted. (E) SM1 competes with TRAP domain-A for binding to salivary glands. Fixed salivary glands were first incubated with no (a), 0.5 (b), 5 µM (c) or 50 µM (d) of non-biotinylated SM1 peptide followed by incubation with 50 nM of recombinant TRAP domain-A. TRAP domain-A binding was detected by incubation with an anti-TRAP antibody followed by incubation with a FITC-labeled secondary antibody. Note that the SM1 peptide effectively competes with TRAP domain-A for binding to salivary glands. (F) Reverse competition: TRAP domain-A competes with SM1 for binding to salivary glands. Fixed salivary glands were first incubated with no (a), 20 nM (b), 200 nM (c) or 2 µM (d) of recombinant TRAP domain-A protein followed by incubation with biotinylated SM1 peptide (5 µM). Peptide binding was detected by incubation with FITC-labeled streptavidin. Note that the protein effectively competes with the SM1 peptide for binding to salivary glands.
Figure 5.
Mutational analysis of TRAP-saglin interactions.
(A) Diagram of the TRAP protein and the Thr126 to Ala mutation in the MIDAS domain. (B) Binding of recombinant wild type or mutant TRAP domain-A to saglin. Saglin was captured on wells of a nickel-coated plate and wild type or mutant recombinant TRAP domain-A were added to the wells, incubated and washed. Bound TRAP domain-A was detected by incubation with anti-TRAP domain-A antibody followed by an alkaline phosphatase-conjugated secondary antibody. (1) Wild type TRAP domain-A; (2) Point mutant TRAP domain-A; (3) No TRAP protein added. (C) Direct visualization of the binding of wild type and mutant TRAP domain-A protein to recombinant saglin. Histidine-tagged saglin was immobilized on Ni-agarose beads, washed and allowed to bind to the purified wild type or mutant TRAP domain-A recombinant protein. Beads were washed and incubated with anti-domain-A antibody. Detection was done after incubation with Alexa Flour-568-conjugated anti-rabbit IgG (red). (a) Wild type TRAP domain-A; (b) Point mutant TRAP domain-A; (c) No TRAP protein added. Panels (d), (e) and (f) show light microscopic images of the fields to their left. (D) Binding of recombinant wild type and mutant domain-A to salivary glands. Fixed salivary glands were incubated with wild type or mutant TRAP domain-A protein, washed and incubated with anti-TRAP domain-A antibody. Detection was with anti-rabbit IgG conjugated to Alexa Flour-488 (green). (a) Wild type TRAP domain-A; (b) Point mutant TRAP domain-A; (c) No TRAP protein added. The lower panels (d, e, f) show light microscopic images of the fields on top (a, b, c, respectively).
Figure 6.
Sporozoites interact with saglin and this interaction is essential for salivary gland invasion in vivo.
(A) Inhibition of P. falciparum sporozoite invasion of An. gambiae salivary glands by anti-saglin antibody. Either anti-saglin monoclonal antibody [1.5∼2 µg, bar 2] or an equivalent amount of an unrelated antibody [control; bar 1] was injected into the haemocoel of 50 P. falciparum-infected An. gambiae female mosquitoes. At 48 h after injection the salivary glands of 22 to 30 surviving mosquitoes were dissected and the number of sporozoites determined. A 95% inhibition was observed in this experiment (p-value, <0.0001). The inhibition in a second experiment (not shown) was 94% (p-value, <0.0001). (B) Down-regulation of saglin mRNA abundance by RNA interference. Either control double-stranded GFP RNA (C) or double-stranded saglin RNA was injected into the hemocoel of An. gambiae female mosquitoes and salivary glands were dissected on days 3 (D3) and 6 (D6) after injection, as indicated. Saglin mRNA abundance was assessed by semi-quantitative PCR using ribosomal protein S7 mRNA primers as controls. (C) Down-regulation of saglin protein abundance by RNA interference. Double-stranded GFP RNA (control; panels a–f) or double-stranded saglin RNA (experimental, panels g–l) was injected into the hemocoel of female An. gambiae mosquitoes and salivary glands were dissected after 1 d (a, b, g, h), 3 d (c, d, i, j) or 6 d (e, f, k, l) for immunofluorescence assays (IFA) with an anti-saglin antibody. The panels pair fluorescent images to the left and DIC images of the same gland to the right. (D) Inhibition of sporozoite salivary gland invasion after down-regulation of saglin expression. Approximately 50 P. falciparum-infected An. gambiae mosquitoes were injected with GFP (1) or saglin (2) dsRNA. Salivary glands were dissected from each of 15∼18 surviving mosquitoes and the number of sporozoites determined. Inhibition was 89% (p-value <0.0008). The inhibition in a second experiment (not shown) was 97% (p-value <0.0001).
Figure 7.
Schematic diagram of the presumed interaction among the various proteins used in this study.
(A) Invasion of salivary glands requires the interaction between the Anopheles salivary gland surface protein saglin with the Plasmodium secreted protein TRAP. (B) The SM1peptide also interacts with saglin because its conformation mimics that of TRAP. In the presence of SM1 all saglin sites are occupied, thus preventing saglin-TRAP interactions and interfering with sporozoite invasion. (C) Because the conformations of SM1 and TRAP are related, an anti-SM1 antibody also recognizes TRAP. This relationship led to the identification of TRAP.