Isolation and Characterization of Poecistasin, an Anti-Thrombotic Antistasin-Type Serine Protease Inhibitor from Leech Poecilobdella manillensis

Antistasin, first identified as a potent inhibitor of the blood coagulation factor Xa, is a novel family of serine protease inhibitors. In this study, we purified a novel antistasin-type inhibitor from leech Poecilobdella manillensis called poecistasin. Amino acid sequencing of this 48-amino-acid protein revealed that poecistasin was an antistasin-type inhibitor known to consist of only one domain. Poecistasin inhibited factor XIIa, kallikrein, trypsin, and elastase, but had no inhibitory effect on factor Xa and thrombin. Poecistasin showed anticoagulant activities. It prolonged the activated partial thromboplastin time and inhibited FeCl3-induced carotid artery thrombus formation, implying its potent function in helping Poecilobdella manillensis to take a blood meal from the host by inhibiting coagulation. Poecistasin also suppressed ischemic stroke symptoms in transient middle cerebral artery occlusion mice model. Our results suggest that poecistasin from the leech Poecilobdella manillensis plays a crucial role in blood-sucking and may be an excellent candidate for the development of clinical anti-thrombosis and anti-ischemic stroke medicines.


Introduction
Protease inhibitors play important roles in the biological purposes of venomous animals, for example, predation and defense [1,2]. There are nine classes of proteases already reported in previous studies, including named molecules (Aspartic, Cysteine, Glutamic, Metallo, Asparagine, Serine and Threonine), unknown molecules (Unknown) and molecules of mixed catalytic mechanisms (Mixed) [3,4]. For each protease class, inhibitors are already described [3,4]. Serine protease inhibitors represent a diverse class of proteins that have been subdivided into many distinct families [5], such as The fraction of previous step exerts FXIIa inhibitory activity was further purified by C8 RP-HPLC by monitoring at 280 nm. The protein peak exerts FXIIa inhibitory activity is indicated by an arrow. The dashed line represents a line gradient of acetonitrile from 30 to 60% over 35 min. (D) The peaks of "C" was used to test FXIIa inhibitory activity. (E) The peak of previous step exerts FXIIa inhibitory activity were further purified by a Mono S TM 5/50 GL column connected to AKTA FPLC system by monitoring at both 215 and 280 nm. The blue and green line represents the conductivity and NaCl concentration, respectively. The protein peak that was used for liquid chromatography coupled with tandem mass spectrometry (LC−MS/MS) is indicated by a red arrow. (F) The peaks of "E" were used to test FXIIa inhibitory activity. Control means a lack of addition of any peak. (B,D,F) are representative of at least five independent experiments.

Primary Structure of Poecistasin
The eluted peak 3 ( Figure 1E) of FPLC containing FXIIa inhibitory activity was collected and lyophilized. Peptide sequence was determined by LC−MS/MS. The sequence of mature peptide of poecistasin (48 amino acids) is "ADCGGKTCSGGQVCSDGVCVCTKLRCRLLCRNGFLKDENGCEY PCTCA" (Figure 2A). Sequence alignment showed it was similar to hirustasin (Hirudo medicinalis; identity: 63%), Bdellastasin (Hirudo medicinalis; identity: 57%), piguamerin (Hirudo nipponia; identity: 69%) and guamerin (Hirudo nipponia; identity: 65%) which are antistasin-type serine protease inhibitors containing only one domain ( Figure 2B). Matrix-assisted laser desorption ionization The secretions of P. manillensis were separated by Sephadex G-50 column by monitoring at both 215 and 280 nm. The fraction exerts FXIIa inhibitory activity is indicated by an arrow. (B) The fractions of "A" were used to test FXIIa inhibitory activity. (C) The fraction of previous step exerts FXIIa inhibitory activity was further purified by C 8 RP-HPLC by monitoring at 280 nm. The protein peak exerts FXIIa inhibitory activity is indicated by an arrow. The dashed line represents a line gradient of acetonitrile from 30 to 60% over 35 min. (D) The peaks of "C" was used to test FXIIa inhibitory activity. (E) The peak of previous step exerts FXIIa inhibitory activity were further purified by a Mono S TM 5/50 GL column connected to AKTA FPLC system by monitoring at both 215 and 280 nm. The blue and green line represents the conductivity and NaCl concentration, respectively. The protein peak that was used for liquid chromatography coupled with tandem mass spectrometry (LC−MS/MS) is indicated by a red arrow. (F) The peaks of "E" were used to test FXIIa inhibitory activity. Control means a lack of addition of any peak. (B,D,F) are representative of at least five independent experiments.

Effects of Poecistasin on FeCl3-Induced Carotid Artery Injury Model
Recombinant poecistasin, which showed similar FXIIa inhibitory activity with native poecistasin, was expressed in E. coli and purified ( Figure S2). As illustrated in Figure S2A, recombinant poecistasin was expressed induced by 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) for 6 h and the fusion poecistasin eluted from Ni 2+ affinity chromatography column was cut by rTEV protease ( Figure S2B). The fraction treated with rTEV protease was loaded onto a RP-HPLC C8 column to purify recombinant poecistasin ( Figure S2C). As illustrated in Figure S2D, the four peaks from previous step were applied to test FXIIa enzymatic inhibitory activity and the "peak 4" with strong FXIIa enzymatic inhibitory activity was analyzed by MALDI-TOF-MS. MALDI-TOF-MS showed the MW of recombinant poecistasin is 5085.7 Da ( Figure S2E). The effect of poecistasin on thrombosis in vivo was evaluated in carotid artery thrombus model induced by FeCl3 in C57BL/6J mice. Different concentrations of recombinant poecistasin (0.2, 5 and 10 mg/kg) and heparin sodium (20 mg/kg) were injected by tail vein 10 min before surgery. After the treatment with 10% FeCl3, blood flow was monitored for 0, 5, 10, 15, 20, 25 and 30 min, respectively. As shown in Figure 4A, B, poecistasin inhibited thrombus formation in a dose-dependent manner with a similar tendency to heparin sodium.

Effects of Poecistasin on FeCl 3 -Induced Carotid Artery Injury Model
Recombinant poecistasin, which showed similar FXIIa inhibitory activity with native poecistasin, was expressed in E. coli and purified ( Figure S2). As illustrated in Figure S2A, recombinant poecistasin was expressed induced by 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) for 6 h and the fusion poecistasin eluted from Ni 2+ affinity chromatography column was cut by rTEV protease ( Figure S2B). The fraction treated with rTEV protease was loaded onto a RP-HPLC C 8 column to purify recombinant poecistasin ( Figure S2C). As illustrated in Figure S2D, the four peaks from previous step were applied to test FXIIa enzymatic inhibitory activity and the "peak 4" with strong FXIIa enzymatic inhibitory activity was analyzed by MALDI-TOF-MS. MALDI-TOF-MS showed the MW of recombinant poecistasin is 5085.7 Da ( Figure S2E). The effect of poecistasin on thrombosis in vivo was evaluated in carotid artery thrombus model induced by FeCl 3 in C57BL/6J mice. Different concentrations of recombinant poecistasin (0.2, 5 and 10 mg/kg) and heparin sodium (20 mg/kg) were injected by tail vein 10 min before surgery. After the treatment with 10% FeCl 3 , blood flow was monitored for 0, 5, 10, 15, 20, 25 and 30 min, respectively. As shown in Figure 4A, B, poecistasin inhibited thrombus formation in a dose-dependent manner with a similar tendency to heparin sodium.

Effects of Poecistasin on Stroke Model
The effect of poecistasin on thrombosis was further investigated by using a transient middle cerebral artery occlusion (tMCAO) mouse model. As illustrated in Figure 5A,B, the cerebral infarct volume was ~32% in the control mice, while different concentrations of recombinant poecistasin (0.2, 1 and 5 mg/kg) decreased infarct volume to ~25, 10 and 2.5%. Edaravone decreased infarct volume

Effects of Poecistasin on Stroke Model
The effect of poecistasin on thrombosis was further investigated by using a transient middle cerebral artery occlusion (tMCAO) mouse model. As illustrated in Figure 5A,B, the cerebral infarct volume was~32% in the control mice, while different concentrations of recombinant poecistasin (0.2, 1 and 5 mg/kg) decreased infarct volume to~25, 10 and 2.5%. Edaravone decreased infarct volume to~2.7%. The functional outcomes reflected by Bederson score and Zea longa score further indicated that poecistasin can alleviate ischemic stroke (IS) (Figure 5C,D). to ~2.7%. The functional outcomes reflected by Bederson score and Zea longa score further indicated that poecistasin can alleviate ischemic stroke (IS) ( Figure 5C,D).

Discussion
Several antistasin-type serine protease inhibitors containing two domains have been found from leeches, including antistasin from Haementeria officinalis and ghilanten from Haementeria ghilianii [22,28]. Hirustasin and Bdellastasin containing only one domain were identified from Hirudo medicinalis [29,30]. Piguamerin and guamerin containing one domain was also identified from Hirudo nipponia [31,32]. In this new report, an antistasin-type serine protease inhibitor exerting anti-coagulatory effect named as poecistasin was firstly purified from leech P. manillensis. Poecistasin contains 48 amino acids and MALDI-TOF-MS analysis showed that the MW of native poecistasin was 5027.5 Da. Sequence of native poecistasin shows high similarity to antistasin-type serine protease inhibitors which contain five pairs of disulfide bonds forming with 1-3, 2-4, 5-8, 6-9 and 7-10 patterns [30] by ten cysteine residues. A guamerin-like homolog from P. manillensis [GenBank: ATE50007] that shares high sequence similarities to poecistasin (92% identical) has been reported. However, no functions and biological significances were studied. Enzyme activity test showed that poecistasin inhibited FXIIa, kallikrein, trypsin and elastase but it had no inhibitory effects on FXa and thrombin. The reactive P1 site was reported in previous works with leech antistasin-type inhibitors [30][31][32]. Antistasin, ghilanten, hirustasin, bdellastasin and piguamerin, which inhibit trypsin-like proteases, were reported to contain arginine or lysine at its predicted P1 site [30,31]. Guamerin, which inhibits chymotrypsin-like and elastase-like but not trypsin-like proteases, contains a methionine residue at its predicted P1 site [32]. Based on this works, we infer that putative poecistasin reactive P1 site locates at arginine 27. In accordance with an arginine being in P1 site, poecistasin is an inhibitor of trypsin. Different from other reported antistasin-type serine protease inhibitors from leeches, poecistasin inhibited both kallikrein and FXIIa of the intrinsic coagulation pathway, implying that poecistasin may influence intrinsic coagulation pathway. As

Discussion
Several antistasin-type serine protease inhibitors containing two domains have been found from leeches, including antistasin from Haementeria officinalis and ghilanten from Haementeria ghilianii [22,28]. Hirustasin and Bdellastasin containing only one domain were identified from Hirudo medicinalis [29,30]. Piguamerin and guamerin containing one domain was also identified from Hirudo nipponia [31,32]. In this new report, an antistasin-type serine protease inhibitor exerting anti-coagulatory effect named as poecistasin was firstly purified from leech P. manillensis. Poecistasin contains 48 amino acids and MALDI-TOF-MS analysis showed that the MW of native poecistasin was 5027.5 Da. Sequence of native poecistasin shows high similarity to antistasin-type serine protease inhibitors which contain five pairs of disulfide bonds forming with 1-3, 2-4, 5-8, 6-9 and 7-10 patterns [30] by ten cysteine residues. A guamerin-like homolog from P. manillensis [GenBank: ATE50007] that shares high sequence similarities to poecistasin (92% identical) has been reported. However, no functions and biological significances were studied. Enzyme activity test showed that poecistasin inhibited FXIIa, kallikrein, trypsin and elastase but it had no inhibitory effects on FXa and thrombin. The reactive P1 site was reported in previous works with leech antistasin-type inhibitors [30][31][32]. Antistasin, ghilanten, hirustasin, bdellastasin and piguamerin, which inhibit trypsin-like proteases, were reported to contain arginine or lysine at its predicted P1 site [30,31]. Guamerin, which inhibits chymotrypsin-like and elastase-like but not trypsin-like proteases, contains a methionine residue at its predicted P1 site [32]. Based on this works, we infer that putative poecistasin reactive P1 site locates at arginine 27. In accordance with an arginine being in P1 site, poecistasin is an inhibitor of trypsin. Different from other reported antistasin-type serine protease inhibitors from leeches, poecistasin inhibited both kallikrein and FXIIa of the intrinsic coagulation pathway, implying that poecistasin may influence intrinsic coagulation pathway. As expected, further research showed that poecistasin prolonged APTT in a dose-dependent manner but had no effect on PT. Blood-sucking animals get blood meal by overcoming host's blood coagulation [34][35][36][37]. Actually, coagulation inhibitors have been found from P. manillensis, such as hirudin-like peptide homologs, hirudin is a kind of serine protease inhibitor exerts thrombin inhibitory activity [38]. Poecistasin may have potent function to help P. manillensis to take a blood meal by inhibiting coagulation at the bitten-site.
Mice carotid artery thrombosis induced by FeCl 3 was used to evaluate the effect of poecistasin on thrombosis in vivo. Our results showed that poecistasin strongly inhibited blood flow and thrombus formation in a dose-dependent manner with a similar tendency to heparin sodium. The effect of poecistasin on thrombosis was further investigated by using a tMCAO mouse model and poecistasin alleviated infarct volume in dose-dependent manner and poecistasin showed stronger anti-IS activity than positive drug edaravone. These all imply poecistasin may be an excellent candidate for the development of clinical anti-thrombosis and anti-IS drugs.

Conclusions
In conclusion, poecistasin was purified from P. manillensis secretions after three chromatographic steps-gel-filtration, RP-HPLC C8 and cation exchange, and its primary sequence identified by LC-MS/MS and MALDI-TOF analysis. Poecistasin can inhibit kallikrein and FXIIa of intrinsic coagulation pathway and proteases trypsin and elastase but has no inhibitory activity on FXa and thrombin under the same assay conditions. Interestingly, poecistasin was found to prolong APTT and had no influence on PT. Poecistasin showed anti-thrombosis activity with the similar effect of heparin sodium in FeCl 3 -induced carotid artery thrombosis model. Poecistasin also showed anti-IS effect with stronger activity than edaravone. Overall, poecistasin was the first antistasin-type serine protease inhibitor exerting anti-coagulatory effect from leech P. manillensis and might be an excellent candidate for the development of clinical anti-thrombosis and anti-IS drugs.

Collection of Crude Secretions
P. manillensis leeches were purchased from Guangxi Province of China. The leeches were still alive when they were transported to the laboratory. Leeches were stimulated by an electrical stimulator and the crude secretions were collected. More specifically, leeches were stimulated by an electrical stimulator (6 volt, 2.5~345 Hz) around the mouth for 5~10 s, the secretions were washed for several times with stimulation buffer (150 mM NaCl, 1 mM L-Arginine). The secretions were centrifuged at 12,000× g for 1 h at 4 • C, and the supernatant was collected and stored at −80 • C.

Purification of Poecistasin
The supernatant of leech secretions was lyophilized, dissolved with 0.1 M PB buffer (Na 2 HPO 4 -NaH 2 PO 4 , pH 6.0) and separated by a Sephadex G-50 column (100 × 2.6 cm, GE Health, Chicago, IL, USA) that was previously equilibrated with the same buffer subsequently. By eluting the column with the same buffer, we got the sample fractions. The flow rate of eluting was 0.3 mL/min at 4 • C and fractions were collected once every 10 min. The absorbance of the elution fractions was monitored at both 215 and 280 nm. Fractions that inhibited FXIIa were pooled and lyophilized prior to further purification. The powder from the previous step was dissolved and loaded to RP-HPLC on a C 8 column (30 × 0.46 cm, Hypersil BDS, Bellefonte, PA USA). Elution was carried out with a linear gradient of solution B (Acetonitrile, 0.1% TFA) at a flow rate of 0.7 mL/min and the absorbance of the elution fractions was monitored at 280 nm. The eluted fraction containing FXIIa inhibitory activity was collected. The eluted fraction of C 8 RP-HPLC was lyophilized, resuspended with solvent A (20 mM MES, pH 6.0) and applied to a Mono S TM 5/50 GL column (GE, Chicago, IL, USA) connected to AKTA explorer 10S FPLC system (GE, Chicago, IL, USA). The column was equilibrated with solvent A and the elution was performed with a linear gradient of 0-45% solvent B (20 mM MES, 1 M NaCl, pH 6.0) over 45 min at a flow rate of 1 mL/min and the absorbance of the elution fractions was monitored at both 215 and 280 nm. The fractions of previous step were also used to test FXIIa inhibitory activity. A lack of addition of any peak was used as negative control.

Mass Spectrometric Analysis and Sequencing of Poecistasin
The eluted "peak 3" of FPLC exerting FXIIa inhibitory activity was collected and lyophilized. The MW of the collected peak was analyzed by MALDI-TOF-MS (AXIMA CFR, Kratos Analytical, Shimadzu Corporation, Kyoto, Japan). Peptide sequence was determined by liquid chromatography coupled with tandem mass spectrometry (LC−MS/MS) at Beijing Biotech-Pack Scientific Co., Ltd. (Beijing, China). Specifically, the purified peptide (peak 3, 100 µg) was transferred into Microcon devices YM-3 (Millipore, Billerica, MA, USA). The device was centrifuged at 12,000× g at 4 • C for 10 min. Subsequently, 200 µL of 50 mM ammonium bicarbonate were added to the concentrate followed by centrifugation and repeat once. After reduced by 10 mM DL-dithiothreitol (DTT) at 56 • C for 1 h and alkylated by 20 mM iodoacetamide (IAA) at room temperature in dark for 1h, the device was centrifuged at 12,000× g at 4 • C for 10 min and wash once with 50 mM ammonium bicarbonate. Added with 100 µL of 50 mM ammonium bicarbonate and free trypsin into the protein solution at a ratio of 1:50 and the solution was incubated at 37 • C overnight. The device was centrifuged at 12,000× g at 4 • C for 10 min. 100 µL of 50 mM ammonium bicarbonate was added into the device and centrifuged and then repeat once. Lyophilize the extracted peptides to near dryness. Resuspend peptides in 50 µL of 0.1% formic acid before LC-MS/MS analysis. Nanocolumn (100 µm × 10 cm) packed with a reversed-phase ReproSil-Pur C18-AQ resin (3 µm, 120 Å, Dr. Maisch GmbH, Ammerbuch, Germany) connected to Ultimate 3000 system (ThermoFisher Scientific, Waltham, MA, USA) combined with Orbitrap Elite™ Hybrid Ion Trap-Orbitrap Mass Spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) was used to analyze the digestion product. The raw MS files were analyzed to deduce the peptide sequence.

Poecistasin Recombinant Expression and Purification
DNA encoding mature poecistasin was synthesized and the prokaryotic expression vector was constructed by inserting DNA sequence encoding mature poecistasin (144 bp) into pET-32a (+) vector (Novagen). Recombinant expression in Escherichia coli BL21 (DE3) was induced by 1 mM IPTG for 6 h in an 80-rpm shaker at 28 • C. Following expression, E. coli cells were collected by centrifuging at 12,000 rpm/min for 10 min at 4 • C and resuspended in binding buffer (20 mM Tris, 100 mM NaCl, pH 8.0) and then homogenized by using Ultrasonic Cell Disruption System (XINYI-IID, XinYi, China). Finally, the supernatant was collected by centrifuging for 1 h at 12,000 rpm/min at 4 • C. Ni 2+ affinity chromatography column was equilibrated in advance with binding buffer. The collected supernatant containing fusion protein was subsequently loaded on a Ni 2+ affinity chromatography column at the flow rate of 1 mL/min. The bound fusion proteins were eluted with 5 column volumes of elution buffer (20 mM Tris-HCl, 100 mM NaCl, 1 M imidazole, pH 8.0). The eluted fraction was resuspended in rTEV protease buffer (50 mM NaH 2 PO 4 , 150 mM NaCl) and the salt was removed by using Ultrafiltration device (Millipore, USA). rTEV protease (5 U/µL) was added into rTEV protease buffer containing poecistasin fusion proteins (1 mg/ml) and reacted for 14 h at 28 • C. The fraction from the previous step was loaded to RP-HPLC C 8 column (30 × 0.46 cm) to purify recombinant poecistasin as the method described above. The purified recombinant poecistasin was applied to test FXIIa inhibitory activity. The MW of recombinant poecistasin was also analyzed by MALDI-TOF-MS.

Effects of Poecistasin on Proteases
Effects of poecistasin on proteases including FXIIa, kallikrein, trypsin, elastase, FXa and thrombin were tested by using corresponding chromogenic substrates. The testing enzyme was incubated with different concentrations (0.5, 1 and 2 µM) of native poecistasin in 60 µL of 50 mM Tris buffer (pH 7.4) for 5 min and then a certain concentration of chromogenic substrate was added. The absorbance at 405 nm was monitored and the kinetic curve was recorded by using an enzyme-labeled instrument (Epoch BioTek, Winooski, VT, USA) for 10 min. Trypsin and elastase were all from Sigma and the enzyme concentrations used were 800 and 400 nM, respectively. The corresponding chromogenic substrates (Sigma, St. louis, MO, USA) were Gly-Arg-p-nitroanilide dihydrochloride for trypsin and N-Methoxysuccinyl-Ala-Ala-Pro-Val-p-nitroanilide for elastase, respectively. The concentration of all the substrates in the reactions was 0. A lack of addition of poecistasin was used as negative control. Dixon plot curve was used to calculate the Kis of poecistasin to inhibit the proteases as described [39].

APTT and PT Assays
Plasma from healthy human was collected from Kunming Blood Center. For APTT assay, 50 µL of APTT reagent (F008-1, Nanjing Jiancheng Bioengineering Institute, Nanjing, China) was incubated with 50 µL of plasma added with different concentrations of native poecistasin (0, 80, 120 and 160 nM). After 3 min incubation, 50 µL of CaCl 2 (25 mM) preheated at 37 • C for 5 min was added and the clotting time was monitored at 405 nm using a semi-automatic coagulation analyzer (ThromboScreen 400c, Pacific hemostasis, Middletown, VA, USA). To test PT, 50 µL of plasma added with different concentrations of native poecistasin (0, 80, 120 and 160 nM) was incubated with 100 µL of PT reagent (F007, Nanjing Jiancheng Bioengineering Institute, Nanjing, China), which has been preheated at 37 • C for 15 min. The clotting time was then monitored at 405 nm.

FeCl 3 -Induced Carotid Artery Injury Model
According to the method described [40], the mice (C57BL/6J, female, 23-25 g) were anesthetized and the body temperature was maintained at 37 • C throughout surgery. Different concentrations of recombinant poecistasin (0.2, 5, 10 mg/kg) and heparin sodium (20 mg/kg) were injected by tail vein 10 min before surgery. Saline was used as negative control. One of the carotid arteries was exposed by cervical incision and separated from the adherent tissue and vagus nerve. Thrombosis was induced by applying a piece (2 × 2 mm) of filter paper pre-soaked with 10% (w/v) FeCl 3 solution to the exposed mice carotid artery. The blood flowing of the carotid artery of all groups were measured by the laser speckle perfusion imaging (PeriCam PSI, HR, Stockholm, Sweden) at 0, 5, 10, 15, 20, 25, and 30 min, respectively after FeCl 3 induction. The perfusion unit of the region of interest (ROI) was also recorded.

Stroke Model
The tMCAO model was applied to induce focal cerebral ischemia as the method described [41]. Mice (C57BL/6J, female, 23-25 g) were anesthetized with 2% isoflurane and tied on a heat controlled operating table (Harvard Apparatus, Holliston, MA, USA) to maintain at 37 • C during the whole period of surgery. Following a midline skin incision in the neck, the proximal common carotid artery and the external carotid artery (ECA) were ligated and a standardized silicon rubber-coated nylon monofilament (6023910PK10, Doccol, Sharon, Boston, MA, USA) was inserted and advanced via the right internal carotid artery to occlude the origin of the right middle cerebral artery. One hour later, mice were re-anesthetized, and the occluding filament was removed to allow reperfusion and different concentrations of recombinant poecistasin (0.2, 1 and 5 mg/kg) and edaravone (10 mg/kg) were injected by tail vein 10 min before reperfusion. For determining ischemic brain volume, mice were sacrificed after the induction of tMCAO for 24 h and the brain was quickly removed and cut into 2-mm thick coronal sections using a mouse brain slice matrix (Harvard Apparatus, Holliston, MA, USA). The brain sections were then stained with 2% TTC (T8877-5G, Sigma, St. Louis, MO, USA). The Bederson score and Zea longa score was tested to monitor neurological function.

Animals and Ethics Statement
All animal experiments were approved by the Animal Care and Use Committee at Kunming Institute of Zoology (identification code: SMKX-2016013; date of approval: 15 July 2016). Animal experiments conformed to the U.S. National Institutes of Health Guide for the Care and Use of Laboratory Animals (identification code: No. 85-23; date of approval: 2 January 1996). Mice (C57BL/6J, female, 23-25 g) were purchased from Vitalriver Experiment Animal Company (Beijing, China) and housed in a pathogen-free environment.

Statistical Analysis
For statistical analysis, the data obtained from independent experiments were presented as the mean ± SD. All statistical analyses were two-tailed and with 95% confidence intervals (CI) using GraphPad prism 6. The results were analyzed using an unpaired t-test. Differences were considered significant at p < 0.05.