Xieru Bao ,Jie Wu,c,d,** ,Guanghui Ma,c,d,*

a State Key Laboratory of Biochemical Engineering,Institute of Process Engineering,Chinese Academy of Sciences,Beijing,China

b School of Chemical Engineering,University of Chinese Academy of Sciences,Beijing,China

c PLA Key Laboratory of Biopharmaceutical Production and Formulation Engineering,Institute of Process Engineering,Chinese Academy of Sciences,Beijing,China

d Jiangsu National Synergetic Innovation Center for Advanced Materials,Nanjing,China

Keywords:Pickering emulsion Wound healing Chitosan nanoparticles High viscosity Antibacterial activity

ABSTRACT Effective treatment of wound injuries can prevent infections and promote healing.In this work,a sprayed Pickering emulsion composed of chitosan nanoparticles,tea tree oil and curcumin was formulated.The shear thinning behaviour allowed this Pickering emulsion to spray even with a high viscosity provided by the aqueous phase containing chitosan nanoparticles.Moreover,the Pickering emulsion could recover to the high viscosity after spraying onto the wounds,leading to the long residence time and intimate contact.In addition,the Pickering emulsion provided a carrier to encapsulate the tea tree oil,which caused sustained-release effects and further strengthened tea tree oil's antibacterial activity towards the Gram-negative Escherichia coli (E.coli),Gram-positive Staphylococcus aureus (S.aureus) and Gram-positive methicillin-resistant Staphylococcus aureus(MRSA).Besides,this multi-phase system offered a platform to load with hydrophobic curcumin,leading to better synergistic healing effects.Meanwhile,the Pickering emulsion avoided the irritation and toxicity related to surfactants,which was far safer than the classical emulsion especially when used in the defective skin.All these excellent properties endowed this sprayed Pickering emulsion with promising prospects for wound healing.

1.Introduction

In recent years,sprays have been used as new dosage forms for wound healing[1,2].Compared to other formulations,sprays are more portable and easier to use[3].Additionally,sprays can be easily wiped from wound tissues when changing dressings.All these properties make sprays the superior choice for timely treatment of injuries [3–5].At present,commercial sprays are usually aqueous formulation with poor adhesion and have relatively limited ingredients.Hence,it is desirable to explore a valid method to improve sprays’ adhesion strength,increase their antibacterial efficacy and preferably endow them with more beneficial effects on wound healing.

Pickering emulsions are those stabilized by solid particles[6].They have been reported to have shear thinning behaviour,which refers to the tendency of fluid adhesion to decrease with increasing shear rate[7,8].This suggests that Pickering emulsions can provide a novel route to formulate wound sprays.And Pickering emulsions are able to spray and recover to high viscosity after spraying onto the wounds.Importantly,high viscosity can provide a long residence time and intimate contact with the wounds.In addition,Pickering emulsions stabilized by solid particles possess higher biosafety and biocompatibility than classical emulsions stabilized by conventional surfactants [9].These properties facilitate the Pickering emulsions’ application in biomedical material field [10,11].Especially when used as wound dressings contacted directly with the defective skin,Pickering emulsions avoid the hemolytic,irritating reaction and toxicity accumulation related to surfactants [11].Meanwhile,the special structure with a solid particle shell may endow Pickering emulsions with sustained-release effects.Moreover,the multicomponent system composed of particles,water phase and oil phase enables Pickering emulsions to be loaded with both hydrophilic and hydrophobic drugs to improve their effects for healing[12].Based on these properties,Pickering emulsions are promising to be one of the ideal choices as sprayed wound healing formulations.

In this study,chitosan was selected to prepare nanoparticles for stabilizing Pickering emulsions.Because it possesses excellent adhesion property and many other positive effects for wound healing,such as stimulating hemostasis,accelerating tissue regeneration and promoting fibroblast proliferation [13–15].Furthermore,chitosan nanoparticles also have been reported to have good biocompatibility,bioadhesivity and biodegradability [16].Meanwhile,we explored several natural essential oils as oil phase,due to their high antibacterial efficacy,analgesic effect and scar suppression property [17,18].Moreover,when used as antibacterial agents,natural essential oils won't result in any drug resistance,emerging a more promising candidate than conventional antibiotics [19].

In this study,a sprayed Pickering emulsion stabilized by chitosan nanoparticles was developed and the natural essential oils with high antibacterial efficacy were used as its oil phase.To achieve the better wound healing effects,curcumin,a hydrophobic drug,was added into the oil phase to enhance the antioxidant and anti-inflammatory effects of this system[20,21].After the Pickering emulsion was formulated and characterized,its antibacterial activity was explored in vitro and in vivo,and the wound healing test was also performed.All these results were expected to demonstrate the promising potential of Pickering emulsions for being used as sprays for wound healing.

2.Materials and methods

2.1.Chemicals

Chitosan powder was purchased from Haizhiyuan Co.,China.Tea tree oil,olive oil,lavender oil,corn oil,clove oil and curcumin were procured from Aladdin Co.,USA.Kangfuxinye,a common commercial spray with aqueous formulation,was purchased from Sichuan Haoyisheng Panxi Pharmaceutical Co.,China.Sodium chloride,tryptone,yeast extract and soya peptone were obtained from Shenggong Co.,China.Sodium hydroxide and absolute ethanol were purchased from Sigma-Aldrich Co.,USA.The bacteria were obtained from Beina Bio Co.,China,including the Gram-negative Escherichia coli (E.coli),Gram-positive Staphylococcus aureus (S.aureus) and Gram-positive methicillin-resistant Staphylococcus aureus (MRSA).All other chemical reagents used in this study were of analytical grade.

2.2.Animals

Balb/c mice,6–8 weeks of age,were obtained from Vital River laboratories (Beijing,China).This study was performed in strict accordance with the Regulations for the Care and Use of Laboratory Animals and Guideline for Ethical Review of Animal (China,GB/T 35892-2018).All animal experiments were reviewed and approved by the Animal Ethics Committee of the Institute of Process Engineering(approval ID:IPEAECA2018168).

2.3.Preparation and characterization of chitosan nanoparticles (CS NPs)

CS NPs were prepared by precipitation.Briefly,4 mol/L NaOH solution was slowly added tochitosan solution until the pH value was greater than 6.5.Because when the pH was higher than the pKaof chitosan (6.5),the solved chitosan would precipitate to form the CS NPs.The concentrations of chitosansolutions were 2,3,4 and 5 mg/mL,respectively.The particle size and zetapotential of CS NPs were measured by dynamic light scattering (DLS) using a Nano Zeta Sizer(Malvern).Size distribution of CS NPs was represented by the polydispersity index (PDI).The smaller the PDI value was,the more uniform-sized the particles were.All measurements were repeated in triplicate,from three independent samples,at 25 °C.

2.4.Preparation of Pickering emulsions

Five kinds of natural essential oils (tea tree oil,olive oil,lavender oil,corn oil and clove oil) were selected as oil phase because of their excellent properties for wound healing,including antibacterial activity,anti-inflammatory property and scar prevention property [22–24].Then the curcumin was accurately weighed and dissolved into these oils.The final content of curcumin in the Pickering emulsion system was set at 10 μg/mL [25].After that,these oils containing curcumin were mixed with CS NPs suspensions to prepare Pickering emulsions with different oil-water ratio.Pickering emulsions were prepared by ultrasonication and the pH of them was adjusted to 7.4 to suit for the wounded skin by adding 4 mol/L NaOH solution slowly.A classical emulsion composed of chitosan solution and the same oil phase stabilized by tween 80 and span 85 was also prepared as comparison.

2.5.Characterization of Pickering emulsions

Stability of freshly prepared Pickering emulsions was checked by centrifugation.Briefly,Pickering emulsions were centrifuged to investigate the volume of the precipitated oil or water.The larger the volume was,the more unstable the Pickering emulsions were.Then the most stable one was stored at room temperature for further investigations.

The droplet size and zeta potential of Pickering emulsions were measured by dynamic light scattering (DLS) using a Nano Zeta Sizer(Malvern).All measurements were repeated three times.

Therheological property of Pickering emulsions was measured by using the parallel-plate rheometer (Malvern) at 25 °C.The shear rate was gradually increased from 0.01 to 100 s-1.In this test,Kangfuxinye and the final selected oil were also measured as comparison.

Storage stability of Pickering emulsions was evaluated by measuring their droplet size and zeta potential every week.

2.6.Antibacterial activity test in vitro

The Gram-negative Escherichia coli (E.coli),Gram-positive Staphylococcus aureus (S.aureus) and Gram-positive methicillin-resistant Staphylococcus aureus (MRSA) were chosen for this antibacterial activity test.For the quantitative analysis,the bacterial suspensions previously cultured and diluted were divided into five groups and then were separately treated with (1) Pickering emulsion,(2) Classical emulsion,(3) CS NPs,(4) Tea tree oil and (5) Blank,as shown in Table 1.Then the five groups were incubated at 37 °C under 180 rpm.Finally,the biomass analysis was taken by measuring the optical density at 600 nm (OD600) after 5,12,24,36 and 48 h of co-cultured[26,27].

After the quantitative analysis,the morphological transformation of the bacteria was studied.Bacteria of the Pickering emulsion group and the Blank group were collected by centrifugation and fixed with 4%formaldehyde.After that,the bacteria were further dried with a gradient of ethanol and subsequently observed by scanning electronic microscopy (SEM,JEOL) after sputter-coating with gold.

2.7.Antibacterial activity test in vivo

To assess the antibacterial activity of the Pickering emulsion in vivo,we first created full-thickness skin wounds with diameter of 5 mm onthe back of mice with a sterile corneal drill.Then edges of the wounds were injected with the mixed bacterial suspension of E.coli,S.aureus and MRSA (3 ×107CFU,50 μL) [28].After that,we fabricated sprays of (1) Pickering emulsion,(2) Classical emulsion,(3) CS NPs and (4)Tea tree oil and (5) Blank respectively,also according to Table 1.Finally,the mice were divided into five groups with three mice in each group and were separately treated with these sprays (50 μL each time,twice daily).The wounds were photographed at different intervals by the smart phone camera and were excised on the fifth day.Then the bacterial samples obtained from the wounds were incubated at 37 °C under 180 rpm,and the OD600of them was recorded to evaluate the number of bacteria after 12 h.

Table 1 The experimental groups.

2.8.Wound healing test

Before the wound healing test,five kinds of sprays were fabricated previously according to Table 1.Then the full-thickness skin wounds with diameter of 5 mm were created on the back of mice with a sterile corneal drill.After that,the mice were divided into five groups with three mice in each group and were treated with the five kinds of sprays respectively(50 μL each time,twice daily).Photographs of the wounds were taken at different intervals by the smart phone camera.Then the wound area was measured by Image J software (http://rsbweb.nih.gov/ij/download.html) [25,29].

The relative reduction in wound area was defined as the wound healing rate,calculated as:

Wound healing rate%=[(S0-St)/S0] × 100

Where,S0and Streferred to the wound area at the initial time and at the different time intervals (t),respectively.

Besides,the wound tissues of mice were excised on the tenth day and fixed in 4% paraformaldehyde for further investigation.

2.9.Histology analysis

The wound tissues fixed in 4% paraformaldehyde previously were cut into sections and stained with hematoxylin and eosin (H&E).After that,these slices were observed using an UltraviewVox (PerkinElmer).

2.10.Statistical analysis

GraphPad Prism 6 was used for the statistical analysis.The results were expressed as mean ± s.e.m.Statistical data were analyzed by one-way or two-way analysis of variance (ANOVA).Difference about P ＜0.05 was considered to be significant.

3.Results and discussion

3.1.Preparation and characterization of CS NPs

CS NPs were prepared successfully by adjusting the pH of chitosan solutions with different concentrations of 2,3,4 and 5 mg/mL.It was depicted that the zeta potential of them was approximately +16 mV(Fig.1a).The positive charge on the surface was owing to the presence of protonated amino groups (NH3+).The particle size of CS NPs was similarly around 300 nm–600 nm,without a significant linear correlation(Fig.1b).The PDI value was small when the concentration was 2,3 and 4 mg/mL,indicating the good dispersion and uniform particle size (Fig.1c).While when the concentration was 5 mg/mL,the PDI became broader.This might be due to the high viscosity of chitosan,which made it difficult to be dispersed.Therefore,the CS NPs prepared by chitosan solutions with concentrations of 2,3,and 4 mg/mL were chosen for the next step.

3.2.Preparation and characterization of Pickering emulsions

The characteristics of Pickering emulsions were mainly affected by the types of oils,oil-water ratios and particle concentrations [30].Therefore,we first prepared Pickering emulsions by emulsifying the mixture of curcumin dissolved in the selected oils(tea tree oil,olive oil,lavender oil,corn oil and clove oil) into the CS NPs suspension,and adjusted the oil-water ratios.It was observed that there was almost no delamination after centrifugation when tea tree oil containing curcumin was used as the oil phase and the oil-water ratio was 1/9,indicating the stable Pickering emulsion was obtained.However,Pickering emulsions produced by the other oils(olive oil,lavender oil,corn oil and clove oil)were unstable,regardless of the oil-water ratio.Thus the tea tree oil containing curcumin was chosen as oil phase for all further studies.

We found thatthe droplet size of the Pickering emulsion stabilized by CS NPs with different concentrations was 2.2 ± 0.6,1.9 ± 0.5 and 1.7 ± 0.6 μm respectively,presenting a decreasing trend (Fig.2a).This was because that within a certain range,the increasing particle concentration would afford a higher droplet surface coverage and lead to the formation of a network structure around the droplets,which further avoided droplet coalescence and improved the stability of Pickering emulsion [11,31].

The zeta potential of the Pickering emulsion was approximately+24 mV (Fig.2b).It was reported that positive charge might enhance the residence on tissues [10].And we found that the zeta potential of the Pickering emulsion was greater than that of CS NPs,owing to a positive charged shell of particles around the droplet [32].

It was represented that viscosities of Pickering emulsions prepared by CS NPs with concentrations of 2,3,and 4 mg/mL were 1.18,6.36 and 26.46 Pa·s,respectively.The results indicated that the viscosity of the Pickering emulsion increased with the increasing chitosan particle concentration (Fig.2d).This endowed Pickering emulsion excellent adhesion property.The viscosities of tea tree oil and Kangfuxinye were 4.06 Pa·s and 1.02 Pa·s respectively,which were much lower than the Pickering emulsion.Because the higher viscosity could provide a longer residence time on the wounds and lead to more intimate contact with the wound tissues,CS NPs suspension of 4 mg/mL concentration was finally chosen as the water phase.

Meantime,it was observed that the viscosity of Pickering emulsion was decreased when the shear rate gradually increased from 0.01 to 100 s-1(Fig.2d).This shear-thinning phenomenon of Pickering emulsions was associated with the formation of a weak droplet network[8].Besides,fluids with shear-thinning phenomenon were generally considered sprayable,and it was subsequently demonstrated that the Pickering emulsion was sprayable as expected (Fig.2c).

In addition,we used centrifuge tubes to collect the Pickering emulsion after spraying and evaluated its rheological property again.The results were consistent with that before spraying.This indicated that the rheological property of the Pickering emulsion was not affected by the process of spraying.

After two months’ storage,there were little changes in the droplet size and zeta potential of the Pickering emulsion(Fig.2e).This revealed that the Pickering emulsion displayed a good long-term stability,which was attributed to the strong stabilization of CS NPs.

3.3.The antibacterial activity test in vitro

Fig.1.(a) Zetapotential of CS NPs with different concentrations.(b) Particle size of CS NPs with different concentrations.(c) PDI of CS NPs with different concentrations.Data are presented as mean ± standard error of the mean (s.e.m.) over three independent experiments.

Fig.2.(a)Droplet size of Pickering emulsions.(b)Zetapotential of Pickering emulsions.(c)Schematic representation of the spraying process of Pickering emulsions.(d)Shear rate dependence of viscosity for Pickering emulsions.(e)Droplet size and zetapotential of Pickering emulsions after storage for 0 day,1 week,2 weeks and 2 months,respectively.Data are presented as mean ± standard error of the mean (s.e.m.) over three independent experiments.

Fig.3.Optical density measurements of(a)E.coli,(b)S.aureus and(c)MRSA.The samples of 1–5 were bacteria suspensions treated with(1)Pickering emulsion,(2)Classical emulsion,(3) CS NPs,(4) Tea tree oil,(5) Blank,respectively.Data are expressed as mean ± s.e.m.(n=3).＊p ＜0.05,＊＊p ＜0.01,＊＊＊p ＜0.005,＊＊＊＊p ＜0.001.

E.coli,S.aureus,and MRSA were selected for this antibacterial activity test,as they are the most prevalent bacterial species in wound infections and could stand for both the Gram-negative and Gram-positive bacteria[28,33].Quantitative analysis was taken by measuring the OD600of these bacterial suspensions co-cultured with different groups(Fig.3).Compared to the blank group,suspensions of the Pickering emulsion group,the classical emulsion group and the tea tree oil group showed smaller OD600,indicating effective inhibition of bacteria.Whereas the OD600of the CS NPs group was similar with that of the blank group,indicating there was still large amount of bacteria in the co-cultured suspension.This could be explained by the fact that the neutral environment of the suspension would greatly limit the antibacterial activity of CS NPs [34,35].It revealed that the antibacterial activity of this Pickering emulsion system was mainly derived from the tea tree oil.Additionally,we found thatthe OD600of the Pickering emulsion group was lower than that of the tea tree oil group.This indicated that the antibacterial activity of tea tree oil was further enhanced after being encapsulated into the core of the Pickering emulsion.It was mainly because the Pickering emulsion provided a delivery vehicle to improve the aqueous solubility of oil and prevent active ingredients degradation[36,37].Furthermore,the OD600of the Pickering emulsion group was lower than that of the classical emulsion group especially at 48 h for MRSA,presenting a relatively long-lasting antibacterial activity.This was because the solid particle shell endowed the Pickering emulsion with better sustained-release effects [11].

Furthermore,the morphological transformation of the bacteria was explored.It was illustrated that the surface of E.coli was smooth and plump before treatment (Fig.4a).After co-incubation with the Pickering emulsion for 5 h,depressions were presented on the surface(Fig.4b).The morphological transformation of S.aureus(Fig.4c and d)and MRSA(Fig.4e and f)was similar with that of E.coli,indicating that the Pickering emulsion was destructive toE.coli,S.aureus and MRSA.

3.4.The antibacterial activity test in vivo

After injected with the mixed bacteria suspension subcutaneously,the wounds of the CS NPs group and the blank group presented purulence (Fig.5a).Whereas the wounds treated with the Pickering emulsion,the classical emulsion and the tea tree oil did not have significant purulence.Furthermore,it was observed that the wounds treated with the Pickering emulsion could heal normally and had the smallest area on the fifth day.The wound tissues were subsequently excised for the quantitative analysis (Fig.5b).It showed that there was still large amount of bacteria in the wound tissues obtained from the blank group,revealing severe infections.Whereas the sample treated by the Pickering emulsion was with little bacteria,indicating almost complete inhibition of bacteria.These results revealed that the Pickering emulsion displayed excellent killing ability to bacteria in vivo and could effectively avoid wound infections.

3.5.The wound healing test

The images intuitively reflected the wound healing process of the different groups (Fig.6a).It was observed that the wound treated with the Pickering emulsion had the smallest area on the fifth day,and was completely healed on the tenth day.This result was further confirmed by the wound healing rate as shown in Table 2,which of the Pickering emulsion group was 54.92 ± 1.07% on the fifth day and 95.06 ± 1.37% on the tenth day.Meantime,it was revealed that the Pickering emulsion significantly reduced the wound area compared with the blank group (12.13 ± 1.47%,62.41 ± 1.83%).In addtion,the Pickering emulsion displayed better healing effects than the CS NPs(31.97 ± 1.24%,80.28 ± 1.71%) and the tea tree oil(15.71 ± 1.19%,84.31 ± 2.26%).This was mainly attributed to the synergistic effects of tea tree oil,curcumin and CS NPs[17,38,39].The tea tree oil displayed an important role as antimicrobial agents,and was reported with anti-inflammatory property and scar prevention effect[40].The curcumin with great antioxidant and anti-inflammatory effects could effectively reduce the production of reactive oxygen during the inflammatory phase and promote the wound repair [25,41].And the CS NPs had hemostatic effects and could promote fibroblast proliferation [16].Moreover,the Pickering emulsion provided a platform to encapsulate these materials within a self-assembled particle shell that might further enhance these effects [36,37].Furthermore,the wound healing rate of the classical emulsion (24.56 ± 1.47%,82.93 ± 1.93%) composed of the same ingredients was also smaller than that of the Pickering emulsion.This could be attributed to the advantages of the Pickering emulsion over the classical emulsion,including the more intimate contact with wounds and better sustainedrelease effects.Besides,the existence of surfactants in the classical emulsion might induce some negative effects and potential toxicity.

Table 2 The wound healing rate.

3.6.The histology analysis

The process of wound healing is composed of four overlapping phases,that is,hemostasis,inflammation,proliferative (consisting of granulation,contraction and epithelialization) and the remodeling phases[42–44].The wound tissues excised on the tenth day were dealt with H&E,and then were observed.For comparison,uninjured tissues in the same location of the mice were excised and stained as the negative control group (Fig.7a).It was found that the result of the group treated with the Pickering emulsion was similar with the negative control group,where there was a complete epithelial covering (epidermis) and no granulation tissue (Fig.7b).This result further confirmed that the wounded tissues treated with the Pickering emulsion had been repaired completely on the tenth day.

4.Conclusions

Fig.4.Representative images of(a)the untreated E.coli,(b)the E.coli after 5 h of treatment with the Pickering emulsion,(c)the untreated S.aureus,(d)the S.aureus after 5 h of treatment with the Pickering emulsion,(e) the untreated MRSA and (f) the MRSA after 5 h of treatment with the Pickering emulsion.

Fig.5.(a) Representative images of the wounds.(b)Optical density measurements of the wound tissues after 5 days of treatment.The samples of 1–5 were in accordance with the wounds treated with (1)Pickering emulsion,(2) Classical emulsion,(3) CS NPs,(4)Tea tree oil,(5)Blank,respectively.Data are expressed as mean ± s.e.m.(n=3).＊p ＜0.05,＊＊p ＜0.01,＊＊＊p ＜0.005,＊＊＊＊p ＜0.001.

Fig.6.(a) Representative images of the wounds at different time intervals.(b) Wound healing rate at different time intervals.The samples of 1–5 were in accordance with the wounds treated with (1)Pickering emulsion,(2) Classical emulsion,(3) CS NPs,(4)Tea tree oil,(5)Blank,respectively.Data are expressed as mean ± s.e.m.(n=3).＊p ＜0.05,＊＊p ＜0.01,＊＊＊p ＜0.005,＊＊＊＊p ＜0.001.

In summary,a sprayed Pickering emulsion with high antibacterial efficacy and excellent healing effect was fabricated successfully.This Pickering emulsion was stabilized by chitosan nanoparticles and composed of water and tea tree oil.We took advantage of the presence of oil phase to load the hydrophobic drug curcumin to achieve better healing effects.Meantime,this Pickering emulsion was demonstrated to be sprayable,which made them portable,easy to use.Besides,the results of rheological property test showed that the Pickering emulsion had high viscosities in the absence of high shear rates,which could result in more intimate contact with the wound tissues.In addition,the Pickering emulsion had effective inhibition towards both the Gramnegative bacteria (E.coli) and Gram-positive bacteria (S.aureus and MRSA) in vitro and in vivo,as proved by biomass analysis,bacterial morphology and a wound infection model.Moreover,it could greatly promote the process of wound healing,which was demonstrated by the wound morphology,wound healing rate and histopathological examination of a full-thickness skin defect model.All the results indicated that Pickering emulsion is a promising candidate for sprayed wound dressings.And its unique structure provides it with great potential for improvement,such as being compounded with hydrophilic growth factor to further promote fibroblast proliferation.Notably,further systematic research can be carried out on the other characteristics of Pickering emulsions,such as the particle layer covered on the surface of droplets,the flexible oil core and the pH-responsive ability.These characteristics may also have beneficial effects and potential applications on wound healing.For instance,the flexible oil core of Pickering emulsions endows them with stress deformation,which may increase the contact area between Pickering emulsions and wound tissues.

Fig.7.H&E staining of(a)the uninjured tissue section,(b)the wound section after 10 days of treatment with Pickering emulsion,(c)the wound section after 10 days of treatment with classical emulsion,(d)the wound section after 10 days of treatment with CS NPs,(e)the wound section after 10 days of treatment with tea tree oil and (f) the untreated wound section after 10 days.Sections are with magnification × 20.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by National Key R&D Program of China(No.2018YFC0311303),China,the Key Research Program of the Chinese Academy of Sciences (Grant No.KFZD-SW-218),China,the International Cooperation Project of the Chinese Academy of Sciences(Grant No.122111KYSB20180021),China and National Natural Science Foundation of China (Grant No.21776287).

Appendix A.Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.pnsc.2020.08.001.

Author contributions

The manuscript was written through contributions of all authors.All authors have given approval to the final version of the manuscript.