The Kunitz-like modulatory protein haemangin is vital for hard tick blood-feeding success

Abstract

Ticks are serious haematophagus arthropod pests and are only second to mosquitoes as vectors of diseases of humans and animals. The salivary glands of the slower feeding hard ticks such as Haemaphysalis longicornis are a rich source of bioactive molecules and are critical to their biologic success, yet distinct molecules that help prolong parasitism on robust mammalian hosts and achieve blood-meals remain unidentified. Here, we report on the molecular and biochemical features and precise functions of a novel Kunitz inhibitor from H. longicornis salivary glands, termed Haemangin, in the modulation of angiogenesis and in persistent blood-feeding. Haemangin was shown to disrupt angiogenesis and wound healing via inhibition of vascular endothelial cell proliferation and induction of apoptosis. Further, this compound potently inactivated trypsin, chymotrypsin, and plasmin, indicating its antiproteolytic potential on angiogenic cascades. Analysis of Haemangin-specific gene expression kinetics at different blood-feeding stages of adult ticks revealed a dramatic up-regulation prior to complete feeding, which appears to be functionally linked to the acquisition of blood-meals. Notably, disruption of Haemangin-specific mRNA by a reverse genetic tool significantly diminished engorgement of adult H. longicornis, while the knock-down ticks failed to impair angiogenesis in vivo. To our knowledge, we have provided the first insights into transcriptional responses of human microvascular endothelial cells to Haemangin. DNA microarray data revealed that Haemangin altered the expression of 3,267 genes, including those of angiogenic significance, further substantiating the antiangiogenic function of Haemangin. We establish the vital roles of Haemangin in the hard tick blood-feeding process. Moreover, our results provide novel insights into the blood-feeding strategies that enable hard ticks to persistently feed and ensure full blood-meals through the modulation of angiogenesis and wound healing processes.

Figure 1. Primary structure and expression of Haemangin.

(A) Sequence alignment of Haemangin with other venom basic proteinase inhibitors belonging to the BPTI/Kunitz family of SPIs. CLUSTALW (1.83) alignment of BPTI/Kunitz-type SPIs: Nn, N. naja (P20229); Pa, Pseudodechis australis (AAT45402); Va, V. ammodytes ammodytes (P00992); Em, Eristocophis macmahonii (P24541); Hl, H. longicornis (AB434485). The signal peptides are underlined. Identical amino acid residues are marked with asterisks. The six conserved cysteine residues are in bold with numbers. Dashes indicate gaps inserted to optimize the alignment. (B) Endogenous expression of Haemangin in adult H. longicornis SGs by immunofluorescence. Freshly obtained SGs of a partially fed adult tick were washed in 0.1% PBS-T, fixed in acetone, and incubated with primary anti-Haemangin antibody. FITC-labeled goat anti-mouse IgG was used as a secondary antibody (left panel). H. longicornis SGs treated with mouse pre-immune sera are regarded as control (right panel).

Figure 2. Inhibition of vascular sprout formation in chick aorta in vitro.

(A) Chick aortic rings (∼1.0 mm) obtained from 12-day-old embryos were placed in a 96-well tissue culture plate coated with 20 µl Matrigel. Then, 100 µl of EBM-2 media supplemented with penicillin and streptomycin was added, followed by the addition of an increasing concentration of Haemangin or HlSGE/OmSGE. The plate was incubated at 37°C and at 5% CO₂. Vessel sprout formation was observed for 3 days and images were taken (magnification×100). (B) Dose-dependent response curves. Data on vessel sprout formation in the presence of an increasing concentration of Haemangin or HlSGE compared to media alone (control, 100%) are expressed as percentages. Error bars indicate the standard error of the mean (n = 5).

Figure 3. Inhibition of capillary tube formation in vitro.

(A) HUVECs were seeded (7.5×10³ cells/well) into a 96-well tissue culture plate coated with 50 µl Matrigel. Then, an increasing concentration of Haemangin or HlSGE/OmSGE was added. Cells were incubated in HUVEC growth medium at 37°C, 5% CO₂. Tube formation was observed for 6 h and images were taken (magnification×100). (B) Dose-dependent response curves. Data on tube formation in the presence of an increasing concentration of Haemangin or HlSGE compared to media alone (control, 100%) are expressed as percentages. Error bars indicate the standard error of the mean (n = 3).

Figure 4. Inhibition of angiogenesis in CAM in vivo.

(A) Thermanox discs, 13 mm in diameter, loaded either with PBS only or test compounds (Haemangin 500 ng/embryo and HlSGE 2.5 µg/embryo) were applied to CAMs of 8-day-old chick embryos. After 48 h of incubation at 37°C, negative or positive responses were assessed. Images were taken using a digital camera (Canon). (B) Quantification of inhibition of angiogenesis by Haemangin and HlSGE. Numbers of vessels were from 7–12 CAMs in each group. **p<0.001 and *p<0.01 versus control. Error bars indicate the standard error of the mean.

Figure 5. Inhibition of cell proliferation and wound healing.

(A) Cells were seeded (3×10³ cells/well) into a 96-well tissue culture plate and were incubated in HUVEC growth medium in the absence or presence of Haemangin or HlSGE at 37°C, 5% CO₂ for 72 h. The numbers of cells in each well were determined after the addition of MTS solution, and inhibition of cell proliferation (% of control) was then estimated. Error bars indicate the standard error of the mean (n = 3). (B) Induction of apoptosis of ECs. HUVECs were seeded (1×10⁵ cells/well) into a 6-well tissue culture plate and were incubated in the absence or presence of Haemangin (250 nM) or HlSGE (25 µg/ml) for 48 h as in (A). Cells were stained with Hoechst 33258 (magnification×200). (C) DNA ladder formation in HUVECs. Cells were seeded (2.5×10⁵ cells/flask) into a T-25 tissue culture flask and were incubated in the absence or presence of Haemangin (250 nM) or HlSGE (25 µg/ml) for 48 h as in (A). Cytoplasmic DNA was isolated and analyzed using 1% agarose gel electrophoresis. (D) Inhibition of wound healing in vitro. HUVECs were seeded (1.25×10⁵ cells /well) into a 6-well tissue culture plate and were grown to confluency in HUVEC growth medium. An increasing concentration of either Haemangin or HlSGE was added at the time of scratching, and the wound healing process was observed for 12 h. Wounds were treated with Haemangin (150 nM) and HlSGE (50 µg/ml), respectively (magnification×50). (E) Quantification of wound reduction by Haemangin and HlSGE. Error bars indicate the standard error of the mean (n = 3).

Figure 6. Inhibition of serine proteinases.

(A) Inhibition of hydrolytic activity of trypsin, chymotrypsin, and elastase on synthetic substrates. Enzyme inhibition assays were performed using fluorogenic substrates by measuring the residual hydrolytic activity after pre-incubation with increasing concentrations of Haemangin or HlSGE. The final concentrations of enzyme/substrate used were as follows: trypsin (5.6 nM)/Substrate (10 µM), chymotrypsin (5 µM)/Substrate (10 µM), and elastase (51 nM)/Substrate (10 µM). Each enzyme was pre-incubated with an increasing concentration of Haemangin at 37°C for 30 min. After pre-incubation of enzyme and inhibitor, appropriate fluorogenic substrate was added. The total volume of reaction mixture was 200 µl in individual wells in 96-well flat bottomed plates. The residual enzyme activity was measured by reading the fluorescence over time at excitation and emission wavelengths of 360 and 460 nm, respectively. PMSF was used as a positive control. Percent inhibition by Haemangin is assessed by the following formula: % inhibition = (1–inhibited rate/uninhibited rate)×100. Error bars indicate the standard error of the mean (n = 3). (B) Inhibition of BSA proteolysis by trypsin and chymotrypsin. Proteolysis inhibition assays were done by pre-incubating the enzymes with increasing concentration of Haemangin or HlSGE in 20 mM Tris-HCl (pH 8.0) in a total reaction mixture of 40 µl for 1 h at 37°C. Then, BSA (500 µg/ml) was added to the inhibited and uninhibited enzymes (250 µg/ml) (control) and was further incubated overnight at 37°C. The following morning, tryptic digests were electrophoresed in 12.5% SDS-PAGE under reducing condition. Molecular weight markers in kDa (lane M) are shown on the left. (C) Effects of Haemangin on plasmin degradation during fibrinolysis. Fibrinogen (8 µM) and thrombin (0.83 NIH unit/ml) were incubated in a buffer containing 50 mM Tris-HCl (pH 7.5), 100 mM NaCl, and 5 mM CaCl₂ at 25°C for 30 min. After the fibrin was completely polymerized, lysis of fibrin polymer was initiated by the addition of 20 µM plasminogen and 15 nM t-PA. Reactions were started in the absence or presence of 200 nM (+) and 400 nM (++) Haemangin, and 150 µg/ml (+) and 300 µg/ml (++) HlSGE separately. The reaction mixtures were incubated overnight at 25°C and were subjected to 8% SDS-PAGE under reducing condition. (D) Inhibition of plasmin-dependent fibrin polymer lysis. Fibrinogen (8 µM) and thrombin (0.83 NIH unit/ml) were incubated in a buffer mentioned in (C). After fibrin polymerization was completed, a 100 µl buffer alone or buffer containing plasmin (1.5 µM) without or with an increasing concentration of Haemangin or HlSGE was added to initiate fibrinolysis. Error bars indicate the standard error of the mean (n = 3).

Figure 7. Haemangin's function in blood-feeding in vivo.

(A) A naïve rabbit's ear as source of blood-meals. (B) Blood-feeding progression and engorgement of adult H. longicornis. Semi-quantitative RT-PCR was performed. Expression of Haemangin-specific mRNA is shown. Arrows indicate areas of blood pool. (C) A rabbit ear showing feeding wounds after the engorged ticks dropped off the ear (arrows; left panel) and repaired wounds on day 10 post-dropping (arrowheads; right panel). (D) Gene silencing of Haemangin by RNA interference. Ticks were injected either with E. coli malE dsRNA (control) or with Haemangin dsRNA. RT-PCR analysis was performed. Expression of Haemangin-specific mRNA is shown. β-actin is shown as an internal control. (E) Effect of targeted gene silencing on blood-feeding by adult ticks. Images were captured at 72 h and 120 h post-feeding. Arrows indicate areas of blood pool. (F) Effect of targeted gene silencing on angiogenesis in vivo. H&E- and silver nitrate–stained sections of a rabbit ear at 120 h tick post-feeding, revealing capillaries and venules (arrows) and area of blood pool (arrowheads) (magnification×100).