Longistatin in tick saliva blocks advanced glycation end-product receptor activation

Abstract

Ticks are notorious hematophagous ectoparasites and vectors of many deadly pathogens. As an effective vector, ticks must break the strong barrier provided by the skin of their host during feeding, and their saliva contains a complex mixture of bioactive molecules that paralyze host defenses. The receptor for advanced glycation end products (RAGE) mediates immune cell activation at inflammatory sites and is constitutively and highly expressed in skin. Here, we demonstrate that longistatin secreted with saliva of the tick Haemaphysalis longicornis binds RAGE and modulates the host immune response. Similar to other RAGE ligands, longistatin specifically bound the RAGE V domain, and stimulated cultured HUVECs adhered to a longistatin-coated surface; this binding was dramatically inhibited by soluble RAGE or RAGE siRNA. Treatment of HUVECs with longistatin prior to stimulation substantially attenuated cellular oxidative stress and prevented NF-κB translocation, thereby reducing adhesion molecule and cytokine production. Recombinant longistatin inhibited RAGE-mediated migration of mouse peritoneal resident cells (mPRCs) and ameliorated inflammation in mouse footpad edema and pneumonia models. Importantly, tick bite upregulated RAGE ligands in skin, and endogenous longistatin attenuated RAGE-mediated inflammation during tick feeding. Our results suggest that longistatin is a RAGE antagonist that suppresses tick bite-associated inflammation, allowing successful blood-meal acquisition from hosts.

Figure 10. Effect of endogenous longistatin on cytokine production in vivo.

RNAi was performed and RNAi-treated or control ticks fed on mice. (A) Mouse skin from both control and RNAi groups was collected, and total RNA was isolated. mRNA of cytokines was checked by qRT-PCR. (B) Total protein was collected from the mouse tissues and evaluated by WB. Mouse β-actin was used as an internal control. n = 5. *P < 0.05; **P < 0.01.

Figure 9. Endogenous longistatin is essential to modulating inflammation during tick feeding.

(A) Endogenous longistatin suppresses inflammation. We attached RNAi-treated or control ticks to tick-naive WT mice. The mice were sacrificed at day 5 of attachment of the ticks, tissues were collected, and sections were prepared and subjected to H&E or special staining. Then cells were counted. Scale bars: 40 μm. (B) Endogenous longistatin suppresses RAGE-mediated inflammation. RNAi-treated or control ticks were allowed to feed on WT or RAGE^–/– mice, tissues were collected, and histopathological analysis was performed. Scale bars: 40 μm. n = 5. *P < 0.05; **P < 0.01.

Figure 8. Expression of RAGE and RAGE ligands at the biting sites of ticks.

Adult unfed ticks were allowed to feed on the shaven backs of tick-naive WT mice. Mice were euthanized at different time points (0–120 hours), and tissues were collected from the feeding lesions. Total RNA was collected, and RAGE ligands and RAGE were analyzed by qRT-PCR. n = 5. *P < 0.05.

Figure 7. rlongistatin prevents RAGE/ligand-mediated inflammation.

(A) Longistatin does not affect carrageenan-induced inflammation in WT or RAGE^–/– mice. Longistatin (100 μg) or PBS was injected into the hind footpad of each WT or RAGE^–/– mouse. One hour later, 2% carrageenan or saline (100 μl) was injected into the same footpad. After 5 hours, mice were euthanized and footpads were collected and fixed, and histopathological analysis was performed. Scale bar: 40 μm. (B) Longistatin attenuates Gla-BSA–induced mouse footpad inflammation. Longistatin (100 μg) or PBS was injected into the hind footpad of each mouse. After 1 hour, Gla-BSA (100 μg) or PBS was injected into the same footpad. Mice were euthanized at the indicated time point, and footpads were collected and subjected to histopathology. Scale bar: 40 μm. n = 5. (C) Longistatin attenuates Gla-BSA–induced pneumonia. Longistatin (50 μg) or PBS was instilled through each nostril. One hour later, Gla-BSA (50 μg) or PBS was instilled. After 48 hours, the mice were euthanized and their lungs were collected and fixed, sections were prepared and stained with H&E, and then cells were counted. Scale bar: 40 μm. n = 6. *P < 0.05; **P < 0.01.

Figure 6. Effect of rlongistatin on RAGE-mediated cell migration in vitro.

(A) Gla-BSA induced cellular migration via RAGE. mPRCs were collected from WT or RAGE^–/– mice and seeded onto the upper chamber of Chemotaxicell in RPMI-1640 medium. The lower chamber contained RPMI-1640 medium with Gla-BSA (0–100 μg/ml) and was incubated. After gentle washing, membrane was stained with Giemsa stain, and cells were counted. (B) Longistatin prevented RAGE-mediated cellular migration. mPRCs from WT mice were treated with longistatin (0–10 μg/ml) or anti-RAGE (1:100) or left untreated, and chemotaxis toward Gla-BSA was determined. n = 4. *P < 0.05.

Figure 5. Longistatin reduces adhesion molecules and cytokine production.

(A) Effects on adhesion molecule production. HUVECs were cultured and stimulated, and total RNA was collected. VCAM1, ICAM1, and E-selectin were checked by qRT-PCR using β-actin to normalize the amount of cDNA. (B) Effects on cytokine secretion. HUVECs were cultured and stimulated, and medium was collected. Cytokines were measured by ELISA. n = 3. **P < 0.01; *P < 0.05.

Figure 4. Longistatin prevents nuclear translocation of NF-κB.

(A) In situ detection of NF-κB. HUVECs were grown on chamber slides and stimulated. Cells were treated with anti–NF-κB (1:1000) and a green fluorescent-labeled secondary Ab. Scale bar: 50 μm. (B) Detection of NF-κB in different cellular fractions. Cytoplasmic and nuclear fractions were collected, and NF-κB was determined by ELISA. (C) NF-κB present in the cytosolic or nuclear fraction of different treatment groups was determined by WB. (D) Time-dependent phosphorylation and degradation of IκB-α. Equal amounts (30 μg) of cytosolic protein from different treatment groups were separated by SDS-PAGE, and WB was performed using antibodies against IκB-α, phospho–IκB-α, or β-actin. (E) Inhibition of phosphorylation of IκB-α by longistatin. HUVECs were cultured, treated with longistatin (0–10 μg/ml), and stimulated with Gla-BSA, cytosol was harvested, and phospho–IκB-α was detected by WB. **P < 0.01; *P < 0.05.

Figure 3. Effects of longistatin on cellular ROS.

HUVECs (1 × 10⁵ cells/chamber) were seeded onto chamber slides or 96-well cell-culture plates and allowed to adhere overnight. Cells were treated without or with Gla-BSA (100 μg/ml) in the absence or presence of pretreatment with anti-RAGE (1:100), DPI (20 μM), Mito-TEMPO (10 μM), or longistatin (10 μg/ml) for 2 hours. Cellular ROS were detected with the DCFDA Cellular ROS Detection Kit. Scale bar: 100 μm. **P < 0.01. AFU, arbitrary fluorescent unit.

Figure 2. Longistatin binds with the membrane of HUVECs through RAGE.

(A) Colocalization of longistatin and RAGE. Serum-starved cells were stimulated with TNF-α (1 ng/ml), fixed, and treated with longistatin (5 μg/ml), followed by the treatment with anti-longistatin (1:1000) and anti-RAGE (1:1000). Bound antibodies were probed, and images were merged. Scale bar: 20 μm. (B) Binding of longistatin with RAGE present in HUVEC lysate. Longistatin (5 μg/ml) was coated and reacted with whole-cell lysate. Bound protein was detected with anti-RAGE (1:1600 to 1:100). (C) Pull-down analysis. His-tagged longistatin (10 μg) was trapped with Talon Metal Affinity Resin and incubated with the membrane fraction of HUVECs (100 μg) in buffer A. Proteins were eluted and analyzed by WB using anti-longistatin (1:1000) or anti-RAGE (1:1000). (D) Ex vivo binding of HUVECs with longistatin through RAGE. Ninety-six–well cell-culture plates were coated with BSA, longistatin, or CML-BSA (5 μg/ml) and treated without or with sRAGE (5 μg/ml) for 1 hour at RT. Then stimulated normal or siRNA-treated HUVECs (2 × 10³ cells/well) were seeded and incubated for 2 hours, and adhered cells were counted. Scale bar: 100 μm. n = 4. *P < 0.05; **P < 0.01.

Figure 1. Longistatin binds with RAGE.

(A) Longistatin binds with RAGE in a concentration-dependent manner. RAGE (5 μg/ml) was coated and blocked. Longistatin or BSA (0–384 nM) was added and incubated in buffer A for 1 hour at RT and treated with anti-longistatin (1:1000) or anti-BSA (1:1000) and detected with HRP-conjugated IgG. Amounts of bound longistatin were determined, and K_D value was calculated. (B) RAGE-dependent binding. Longistatin-coated (4 μg/ml) wells were reacted with RAGE/TLR4 (0–5 μg/ml) in 50 μl of buffer A and treated with anti-RAGE or anti-TLR4. (C) Competitive binding of longistatin with the V domain. RAGE (4 μg/ml) was coated, blocked, and treated with longistatin alone or with a mixture of longistatin (1 μg/ml) and each of the other RAGE ligands (5 μg/ml) as indicated, and bound longistatin was detected. (D) Longistatin binds with recombinant V domain. V, C1, or C2 domain (5 μg/ml) was coated and interacted with longistatin, and bound longistatin was detected. (E) Computational docking of longistatin to the V domain of RAGE was performed using ClusPro 2.0. (F) Ca²⁺-dependent RAGE binding. RAGE-coated (5 μg/ml) wells were treated with longistatin or metal-free longistatin (6 μg/ml), and binding was detected. n = 3. *P < 0.05; **P < 0.01.