Dual responsive block copolymer coated hollow mesoporous silica nanoparticles for glucose-mediated transcutaneous drug delivery

Yaping Wang ,Songyue Cheng ,Wendi Fan ,Yikun Jiang ,Jie Yang ,Zaizai Tong,*,Guohua Jiang

1 College of Materials Science and Engineering and Institute of Smart Biomedical Materials,Zhejiang Sci-Tech University,Hangzhou 310018,China

2 College of Science,Institute of Materials Physics and Chemistry,Nanjing Forestry University,Nanjing 210037,China

Keywords:Diabetes Transcutaneous microneedles Stimuli-responsive drug release Hollow mesoporous silica nanoparticles Block copolymer

ABSTRACT A self-regulated anti-diabetic drug release device mimicking pancreatic cells is highly desirable for the therapy of diabetes.Herein,a glucose-mediated dual-responsive drug delivery system,which combines pH-and H2O2-responsive block copolymer grafted hollow mesoporous silica nanoparticles (HMSNs)with microneedle (MN) array patch,has been developed to achieve self-regulated administration.The poly[4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl acrylate]-b-poly[2-(dimethylamino)ethyl methacrylate] (PPBEM-b-PDM) polymer serves as gate keeper to prevent drug release from the cavity of HMSNs at normoglycemic level.In contrast,the drug release rate is significantly enhanced upon H2O2 and pH stimuli due to the chemical change of H2O2 sensitive PPBEM block and acid responsive PDM block.Therefore,incorporation of anti-diabetic drug and glucose oxidase (GOx,which can oxidize glucose to gluconic acid and in-situ produce H2O2) into stimulus polymer coated HMSNs results in a glucose-mediated MN device after depositing the drug-loaded nanoparticles into MN array patch.Both in vitro and in vivo results show this MN device presents a glucose mediated self-regulated drug release characteristic,which possesses a rapid drug release at hyperglycemic level but retarded drug release at normoglycemic level.The result indicates that the fabricated smart drug delivery system is a good candidate for the therapy of diabetes.

1.Introduction

Diabetes mellitus is a metabolic disease characterized by chronic hyperglycemia with multiple etiology [1,2].According to the etiology classification system,there are four types of diabetes including type 1 and 2 diabetes as well as gestational diabetes and other types of diabetes [3-6].Currently,diabetes treatment still has been received multiple subcutaneous insulin injections to control blood glucose levels,but this method is painful and requires strict adherence by the patients [7,8].As a new option,microneedles (MNs) have been widely used as a promising noninvasive tool to enhance percutaneous administration [9-13].Compared with traditional subcutaneous injection,the microchannels produced by MNs can be used for simple and painless treatment.To maximize its therapeutic ability,various formulations in terms of MNs have been developed [14-17].Among these,soluble MNs have been extensively studied due to the ability of rapid and controlled drug release to achieve enhanced drug delivery[9,17].In addition,open-loop hypodermic insulin administration often fails to adequately regulate blood sugar levels,leading to very serious disorders such as hypoglycemia,coma and even death[18].Therefore,many researches have been focused on designing glucose-sensitive drug delivery systems that can monitor glucose levels to achieve the goal of closed-loop drug delivery [19-26].

Recently,hollow mesoporous silica nanoparticles(HMSNs)have been received extensive attention in nano-scale drug delivery systems because of their adjustable porous structure,stable skeleton structure,large specific surface area and surface functional availability [27-30].With their unique superior performance,surface functionalized HMSNs become an excellent drug delivery vehicle with intellectual sensitivity [27].Compared with organic compounds [31,32] or inorganic nanoparticles [33-35] modified HMSNs,the responsive polymers have a special application prospect to functionalize HMSNs due to their diversity,versatility and excellent drug control ability [36,37].Therefore,polymergrafted HMSNs have been promising candidates as drug storage for stimulus-responsive drug release.

Glucose oxidase(GOx)is a widely used enzyme that can convert the glucose to gluconic acid and futher generate H2O2in situ,resulting in a low pH and high concentration of H2O2at local environment[38-41].Thus,pH-and H2O2-responsive drug release systems were normally constructed by incorporation of GOx with stimuli-responsive nanoparticles.pH-responsive polymers such as poly(acrylic acid) and poly(2-(dimethylamino)-ethyl methacrylate) (PDM) can be anchored on HMSNs to achieve acid triggered drug release [42-46].On the other hand,phenylboronic ester(PBE)is quite sensitive to H2O2,which is widely used in the design of H2O2responsive drug release systems [47-49].On the basis of above factor,herein a PPBEM-b-PDM diblock copolymer with H2O2and pH dual responsiveness was grafted onto the surface of HMSNs as a drug delivery vehicle.After loading GOx and antidiabetic drug metformin (Met),the drug-loaded nanoparticles((Met+GOx)@HMSNs@(PPBEM-b-PDM)) were deposited into the tips of microneedles made from soluble polyvinylpyrrolidone(PVP) matrix for transdermal drug delivery (Fig.1).Bothin vitroandin vivoresults showed this MN device presented a glucose mediated self-regulated drug release characteristic,which possessed a rapid drug release at hyperglycemic level but retarded drug release at normoglycemic level.

2.Experimental

2.1.Preparation of (PPBEM-b-PDM)-coated HMSNs (HMSNs@(PPBEM-b-PDM))

HMSNs-NH2was modified by PPBEM-b-PDM via an amidation reaction in the presence of EDC and NHS.The carboxyl group on PPBEM-b-PDM (1059.0 mg,0.20 mmol) was activated by incubation in a solution of EDC (92.0 mg,0.48 mmol) and NHS (55.2 mg,0.48 mmol)in anhydrous DMF(15 ml)for 30 min at room temperature.After the reaction between the amino group and activated carboxyl group for 7 days at room temperature,the PPBEM-b-PDM polymer was immobilized onto the surface of HMSNs-NH2(0.30 g).The crude product (HMSNs@(PPBEM-b-PDM)) was ultracentrifuged (10,000 rpm,5 min) for three times with DMF to remove the ungrafted polymer,and another several times with ethanol and finally dried under vacuum to yield the polymer coated HMSNs,HMSNs@(PPBEM-b-PDM).

2.2.Preparation of drug-loaded nanoparticles

The prepared HMSNs@(PPBEM-b-PDM) (5.0 mg) was dispersed in HCl solution (0.01 mol·L-1,pH 2.0,2 ml) followed by the addition of metformin (4.0 mg) and the glucose oxidase (GOx,0.8 mg).The solution was moderately stirred at room temperature in the dark for 48 h.The resultant drug-loaded nanoparticles were then separated from the solution by centrifugation(10000 r·min-1,6 min)and washed repeatedly with deionized water to remove the adsorbed drug.The drug-loaded nanoparticles were named as(Met+GOx)@HMSNs@(PPBEM-b-PDM).

Rhodamine 6G was used as a fluorescent model drug instead of metformin to observe the diffusion behavior of drug in skin tissues.Similar preparation process was performed to prepare R6G-loaded HMSNs ((R6G+GOx)@HMSNs@(PPBEM-b-PDM)).Metformin loaded HMSNs (Met@HMSNs) and metformin loaded polymercoated HMSNs ((Met@HMSNs@(PPBEM-b-PDM)) were also prepared using the same method for a better comparison.

The amount of encapsulated metformin was measured through the UV-vis absorption at 233 nm by subtracting the amount of unloaded metformin in the collected supernatant from the amount of original feeding metformin.The drug loading content (DLC,%(-mass)) and drug loading efficiency (DLE,%(mass)) were calculated according to the following formula:

Fig.1.Schematic illustration of preparation of dual responsive (Met+GOx)@HMSNs@(PPBEM-b-PDM) MNs for transdermal drug delivery.

2.3.In vitro drug release

Thein vitrorelease of metformin from the nanoparticles (Met+GOx)@HMSNs@(PPBEM-b-PDM) triggered by biological stimuli was studied at 37°C in PBS solution (8 mg·ml-1) with different pH values (pH 2.0,5.0,and 7.4),different concentrations of H2O2(0 mmol·L-1,1 mmol·L-1,and 5 mmol·L-1) and different concentrations of glucose.(Met+GOx)@HMSNs@(PPBEM-b-PDM) suspension was first placed into a dialysis bag and dialyzed against corresponding buffer medium at 37°C in the dark.At predetermined time intervals,4.0 ml of the release medium was taken out and determined by absorbance measurement and using the standard curve of metformin,while the same amount of new buffer solution was added back to the original release medium.The experiments were carried out in triplicate and the mean value was used as the final result.

2.4.In vitro cell cytotoxicity

Thein vitrocytotoxicity of HMSNs@(PPBEM-b-PDM),Met@HMSNs@(PPBEM-b-PDM)and (Met+GOx)@HMSNs@(PPBEM-b-PDM) nanoparticles towards mouse embryonic fibroblast (3T3-L1) cells was evaluated by Themethyl-thiazolyldiphe nyl-tetrazolium (MTT) assays.3T3-L1 cells were seeded in 96-well plates with a density at 5×103cells per well and cultured for 24 h.Then,the cells were incubated with HMSNs@(PPBEM-b-PDM),Met@HMSNs@(PPBEM-b-PDM) or (Met+GOx)@HMSNs@(PPBEM-b-PDM) nanoparticles (50,100,150,200,250 μg·ml-1)for 24 h.Finally,optical intensities of the solutions at 490 nm were measured by using a PowerWave XS/XS2 microplate spectrophotometer.Each group was repeated 3 times.

2.5.Hypoglycemic effect in vivo

Thein vivoefficacy of MNs with (Met+GOx)@HMSNs@(PPBEMb-PDM) as fillers for diabetes treatment was evaluated in STZ (40 mg·kg-1)-inducedadult diabetic rats (male Sprague Dawley rats weighing (200±20) g) [S4].All animal procedures are adopted and guided by Animal Ethics Committee of Zhejiang Sci-Tech University and Experimental Animal Center of Zhejiang Academy of Medical Sciences (China).The back hair of rats was removed using an electric shaver for MNs group experiments.The following groups of diabetic rats(three rats for each group)were studied:(1)Control group,without other treatment on diabetic rats;(2) Injection group,with a single dose of metformin (1.0 mg) injected into the abdominal skin of diabetic rats using a hypodermic needle;(3)Met@HMSNs MNs group,with Met@HMSNs MNs(loaded with 1.0 mg of metformin in Met@HMSNs without PPBEM-b-PDM coating)applied to the back skin of diabetic rats;(4) (Met+GOx)@HMSNs@(PPBEM-b-PDM)MNs group,with(Met+GOx)@HMSNs@(PPBEM-b-PDM) MNs (loaded with 1.0 mg of metformin in (Met+GOx)@HMSNs@(PPBEM-b-PDM)) applied to the back skin of diabetic rats.The blood glucose levels(BGLs)were measured from tail vein blood samples of rats using a blood glucose meter (Sinocare Inc.,Changsha,China) until there was a return to stable hyperglycaemia.

2.6.Acute toxicity evaluation

Representative tissues (heart,liver spleen,lung,kidney) of experimental rats in group (V) were taken out and made into H&E-stained paraffin-embedded sections to evaluate the acute toxicity of the systemin vivo.

Fig.2.Morphological characterizations of HMSNs@(PPBEM-b-PDM).TEM micrograph of(A)HMSNs,(B)HMSNs-NH2 and(C)HMSNs@(PPBEM-b-PDM).(D)Zeta potential of HMSNs,HMSNs-NH2 and HMSNs@(PPBEM-b-PDM).(E)FT-IR spectra of HMSNs,HMSNs-NH2 and HMSNs@(PPBEM-b-PDM).(F)TGA curves of PPBEM-b-PDM polymer(mass loss: 82.6%),HMSNs (mass loss: 11.3%) and HMSNs@(PPBEM-b-PDM) (mass loss: 38.5%).

3.Results and Discussion

The synthetic route and1H NMR spectra of PBEM monomer,PPBEM macro agent and block copolymer PPBEM-b-PDM were shown in Fig.S1-S3 (Supplementary Material),respectively.The result of1H NMR spectra showed the successful preparation of diblock copolymer,PPBEM7-b-PDM25(Fig.1).On the other hand,HMSNs as drug-loaded storage were prepared according to our previous work [37].The morphology of HMSNs was characterized by transmission election microscopy (TEM).As shown in Fig.2A,the obtained HMSNs had a well-defined hollow structure with shell thickness of about 40 nm and uniform size of 156 nm.After 3-aminopropyltriethoxysilane (-NH2) was bonded to the surface of HMSN,the structural morphology of the HMSNs-NH2(Fig.2B)had no obvious change compared with neat HMSNs.However,after functionalization by PPBEM-b-PDM polymer,it was interesting to note that the channels of HMSNs were difficult to distinguish,which was possibly due to the coating of PPBEM-b-PDM polymer onto the surface of HMSNs (Fig.2C).Moreover,the size distribution of HMSNs after step functionalization were further characterized by dynamic light scattering (DLS) shown as Fig.S4.It showed the size of HMSNs was about 142 nm,while after modification of -NH2and further incorporation of PBEM-b-PDM onto the HMSNs,the size increased to 144 nm and 155 nm,respectively.

Zeta potentials and the FT-IR spectra were further performed to prove the successful polymer modification of HMSNs.As shown in Fig.2D,Zeta potential of naked HMSNs was around-28.2 mV,while it increased to 21.3 mV after amination for HMSNs-NH2because of the protonation of amino groups on the surface of HMSNs.Moreover,the Zeta potential was again reduced to-0.341 mV after coating with PPBEM-b-PDM diblock copolymer,indicating that the block copolymer were successfully anchored on the surface of HMSNs.FTIR measurement further demonstrated the change of chemical components after step functionalization.As shown in Fig.2E.The newly emerging peak in the HMSNs-NH2spectrum at 1553 cm-1compared to that of HMSNs corresponded to the stretching vibrations of N-H flexural vibration of forming primary amines[37].However,after coating with PPBEM-b-PDM polymer,the intensity of peak at 1553 cm-1in the HMSNs@(PPBEM-b-PDM) spectrum significantly decreased compared to that of HMSNs-NH2,indicating that -NH2group almost reacted with carboxyl terminated PPBEM-b-PDM polymer.On the other hand,the newly existing peak at 1800 cm-1corresponding to the stretching vibration of the carbonyl group in the chemical structure of PPBEM-b-PDM was observed,demonstrating the successful modification of polymer to the surface of HMSNs.The X-ray photoelectron spectroscopy (XPS) result also demonstrated that N element increased after PPBEM-b-PDM was grafted onto HMSNs due to the N element containing PDM block.Finally,to determine the mass percentage of PPBEM-b-PDM in the HMSNs@(PPBEM-b-PDM),thermogravimetric analysis (TGA) was carried out.Fig.2F showed that the mass loss of PPBEM-b-PDM,HMSNs and HMSNs@(PPBEM-b-PDM) after heating to 700 °C in nitrogen were 82.6%,11.3% and 38.5%,respectively.According to mass loss of above three samples,the mass content of the PPBEMb-PDM polymer grafted on HMSNs-NH2was estimated to be approximately 38.1%.

In present drug release system,the HMSNs serves as the drug storage while the grafting polymer severs as the gate keeper to prevent drug leakage from hollow cavity of HMSN under physiological condition.However,the grafting polymer,PPBEM-b-PDM,presents dual responsiveness of hydrogen peroxide (H2O2) and pH,in which the PPBEM block is sensitive to H2O2while the PDM block is easily protonated at low pHs.Upon H2O2and pH stimuli,the covered polymer chains will change from hydrophobicity to hydrophilicity,thus the channels of HMSNs are unfolded to release the encapsulated drug.The conformation change of PPBEM-b-PDM from hydrophobicity to hydrophilicity could be reflected to some extent from the size change of HMSNs@(PPBEM-b-PDM) at different conditions.The size change of HMSNs@(PPBEM-b-PDM) in PBS at pH and H2O2stimuli were tested by DLS.As shown in Fig.S6,the size of the HMSNs@(PPBEM-b-PDM) increased as the pH value decreased and the H2O2concentration increased,implying that the chain conformation of PPBEM-b-PDM changed from collapsed form to extended form after pH and H2O2stimuli.

On the other hand,glucose oxidase(GOx)can oxidize glucose to gluconic acid thus decreasing the local pHs and generate H2O2.For this reason,the anti-diabetic drug metformin (Met) along with GOx were encapsulated into the hollow cavity of HMSNs to construct H2O2and pH dual responsive drug release system.After encapsulation,the average drug loading efficiency (DLE) and drug loading content(DLC)of metformin were 65.8%and 31.2%,respectively.And the average DLE and DLC of GOx were 68.1%and 14.2%,respectively,indicating that metformin and GOx are both encapsulated in HMSNs@(PPBEM-b-PDM).The release of Met from (Met +GOx)@HMSNs@(PPBEM-b-PDM)in vitroin response to H2O2,pH and glucose were studied by incubating the drug-loaded nanoparticles in PBS solutions with different stimuli.First,the Met release behaviors were explored at different H2O2concentrations.As shown in Fig.3A,only about 21.6%of Met was released in PBS solutions without stimulus after incubating for 12 h,indicating the drug could diffuse from the channel of HMSNs with a very slow rate.In contrast,about 35.9% of drug was released at 1 mmol·L-1H2O2and it could increase to 80.3%upon 5 mmol·L-1H2O2,indicating the drug release behavior possessed a H2O2concentrationdependent performance.To study the effect of pH on Met release behavior,the (Met+GOx)@HMSNs@(PPBEM-b-PDM) nanoparticles were incubated in PBS solutions with different pHs.As shown in Fig.3B,a burst drug release was detected upon pH of 2.0,while the drug release rate was reduced at a pH of 5.0.And the total drug release levels reached to 39.3%and 80.4%at pH 5.0 and 2.0,respectively.This drug release behavior indicated a pH-dependent performance,which was due to the different protonation degrees of PDM block at low pHs.To evaluate the glucose responsiveness,the drug loaded nanoparticles were incubated at 37°C in solutions with different glucose concentrations,including 400 (a typical hyperglycemic level),200 (a medial glucose level),100 (a normal glucose level)mg·dl-1,respectively.Fig.3C showed the Met release profiles from (Met+GOx)@HMSNs@(PPBEM-b-PDM) at different glucose concentrations.At a normal glucose level of 100 mg·dl-1,about 24.0% of Met was released,indicating a low drug release amount at a normal blood glucose concentration.However,increasing the glucose concentration to 200 or 400 mg·dl-1,a fast drug release was observed and the release amount of Met could reach to around 40.2% and 70.6%,respectively.This glucoserelated drug release behaviors showed the present drug-loaded nanoparticles were promising to apply in treatment of diabetes.On the other hand,by increasing the concentration of glucose from 100 mg·dl-1to 200 mg·dl-1and finally to 400 mg·dl-1,a cumulative kinetic curve of drug release was detect and the drug release profile showed a step-wise behavior.As shown in Fig.3D,the(Met+GOx)@HMSNs@(PPBEM-b-PDM) was first exposed to 100 mg·dl-1glucose concentration,where only approximately 22.7%of Met was released within 3.0 h.When the glucose concentration increased to 200 mg·dl-1,the released Met amount can be improved to 43.6%.Further increasing the glucose concentration to 400 mg·dl-1resulted in a much higher amount of released drug,implying that a fast drug release at a typical hyperglycemic level while a retarded drug release at a normal glucose concentration.The results are well consistent with glucose-dependent release behaviors as monitored in Fig.3C.

Before application in bio-related field,it is necessary to study the biological safety of the as-prepared drug-loaded nanoparticles.Herein,mouse embryonic fibroblast (3T3-L1) cells were used to evaluate cytotoxicity of HMSNs@(PPBEM-b-PDM),Met@HMSNs@(PPBEM-b-PDM) or (Met+GOx)@HMSNs@(PPBEM-b-PDM) by using MTT assays.As shown in Fig.3E,the survival rate of 3T3-L1 cells was above 91%after 24 h of incubation with all nanoparticles at different concentrations ranging from 50-250 μg·ml-1.The result indicated that all the drug-loaded nanoparticles possessed excellent cytocompatibility and were favorable for nanomedicine application.

Fig.3.Met release profiles from (Met+GOx)@HMSNs@(PPBEM-b-PDM) with different stimuli of (A) H2O2,(B) pH and (C) glucose.(D) The Met release profiles from(Met+GOx)@HMSNs@(PPBEM-b-PDM)upon successive stimulus with different glucose concentrations.(E)The cell viability of 3T3-L1 after treatment with HMSNs@(PPBEMb-PDM),Met@HMSNs@(PPBEM-b-PDM) or (Met+GOx)@HMSNs@(PPBEM-b-PDM).Error bars represent standard deviations.

Subsequently,(Met+GOx)@HMSNs@(PPBEM-b-PDM)nanoparticles were deposited in the tips of polyvinylpyrrolidone (PVP)microneedles (MNs) using PDMS microneedle molds for transdermal drug release.Fig.S7 showed the morphologies of the asprepared MNs with pyramidal shape prepared by the microneedle mold,indicating the MNs were successfully replicating from the mold.After preparing the MN array patch,the mechanical strength of MN array patch was test to ensure the needles have a sufficient strength to insert into the skin for further application (Fig.S8A).The mechanical property of the as-prepared MNs was analyzed by the failure compressive strength test.As shown in Fig.S8B,when the displacement reached 0.4 mm,the destructive force of(Met+GOx)@HMSNs@(PPBEM-b-PDM)MNs was 0.179 N per needle compared with 0.098 N per needle of pure PVP MNs.Therefore,(Met+GOx)@HMSNs@(PPBEM-b-PDM)MNs prepared in this work have sufficient mechanical strength to penetrate the skin [50].On the other hand,the recovery state of skin spots showed high performance of skin recovery after MN application (Fig.S8C),which was capable for the frequent administration of transdermal drug delivery in the treatment of diabetes.The dissolving ability of(R6G+GOx)@HMSNs@(PPBEM-b-PDM)MNs in the separated skins treated with different time was confirmed by the confocal laser scanning microscopy (CLSM) using R6G as a fluorescent dyein vitro.As revealed in Fig.S9,the pyramid-shaped morphology of MNs with a height of about 600 μm were gradually disappeared after 5 min (Fig.S9A) and the height of each needle decreased to about 350 μm from the observation and analysis of Fig.S9a1to S9a9.After the MNs penetrating into skin for 30 min,the pyramidal morphology of MNs disappeared completely (Fig.S9B) and the deepest fluorescence of MNs was about 80 μm from the observation and analysis of Fig.S9b1to S9b9,showing PVP matrix was dissolved by tissue fluid.

The capabilities of(R6G+GOx)@HMSNs@(PPBEM-b-PDM)MNs to skin insertion and the transdermal delivery behavior were studied using full-thickness Sprague Dawley (SD) rates skinex-vivo.Histological observations of diabetic and healthy SD rat skin confirmed precise skin penetration and drug release from nanoparticles.As demonstrated in bright-field images of Fig.4A and C,the depth of the microcavity on skin surface(yellow arrows)was about 300 μm,which was similar to the depth of dermis layer.In addition,the red fluorescence signal was hardly detected in deeper tissue after incubation in skin of diabetic rats and healthy SD rats for 5 min.The depth of red florescence was about 300-350 μm,indicating no R6G was diffused into the deeper tissue.In contrast,after application of(R6G+GOx)@HMSNs@(PPBEM-b-PDM)MNs on diabetic rats for 30 min,strong fluorescence(green arrow)was clearly observed in the deeper tissues and the depth of red florescence could reach to far 1000 μm (Fig.4D),but no red fluorescence was observed in deeper tissue on healthy rats (only about 400-450 μm shown in Fig.4B).The results suggested that high blood glucose levels in diabetic rats can trigger the rapid release of drugs from nanoparticles,while rarely drug was released from the nanoparticles at a normal glucose level.The histological observations are consistent with thein vitrodrug release behavior.

To evaluate the efficacy of MN device on regulation of blood glucose concentrationin vivo,the type 2 diabetes induced by streptozotocin were used as animal models (Fig.5A).Briefly,different MNs (Met@HMSNs or (Met+GOx)@HMSNs@(PPBEM-b-PDM)MNs) patches containing metformin (1 mg) were applied to the diabetic rats after the dorsal shaving.And the control group (untreated)and the injection group(single dose of metformin injected into the abdominal skin of diabetic rats using a hypodermic needle) were used for comparison.Fig.5B showed the blood glucose level of SD rats over time after administration with different treatments.For the control group,the blood glucose concentration remained a high level.However,the injection group showed a fast hypoglycemic effect and the blood glucose concentration rapidly decreased to 88 mg·dl-1within 2 h but rapidly increased to the initial level in the next 7 h.On the other hand,the MN patches(Met@HMSNs MNs or (Met+GOx)@HMSNs@(PPBEM-b-PDM)MNs) containing same amount of metformin (1 mg) reached 200 mg·dl-1within 2 h.It should be noted that the blood glucose levels of (Met+GOx)@HMSNs@(PPBEM-b-PDM) MNs group could be maintained in the normoglycemic range for about 4 h,while Met@HMSNs MN group was maintained only for about 0.6 h.More interestingly,the blood glucose concentration of the (Met+GOx)@HMSNs@(PPBEM-b-PDM)MNs group returned to the initial level after 10 h compared with 7 h for subcutaneous injection group and 5 h for Met@HMSNs MN group.The study suggests that MN patches in combination with glucose-mediated dual-responsive HMSNs have potential applications value for the therapy of diabetes.In addition,this dual-responsive HMSNs or integrated MNs can also be used as delivery platforms to provide therapeutic drugs for the treatment of other diseases.The glucose tolerance test of healthy and diabetic rats was conducted.The rats weighting about 200 mg were feed 40 mg dose of aqueous oral dextrose(0.4 g dextrose·ml-1) within 5 min.Blood glucose concentrations were test at 0,30,60 and 120 min.The glucose concentrations reached highest level after 30 min and then decreased for both group.However,the glucose tolerance of diabetic rats was much worse than that of healthy group since they had a larger increase value of glucose concentrations (Fig.S10).

Fig.4.Histological cross-section of the skin penetration site applied with (R6G+GOx)@HMSNs@(PPBEM-b-PDM) MNs for (A,C) 5 min and (B,D) 30 min on healthy and diabetic rates,respectively.

Fig.5. In vivo hyperglycemic effect of (Met+GOx)@HMSNs@(PPBEM-b-PDM) MNs for the therapy of STZ-induced type 2 diabetic rats.(A) Schematic representation of the glucose-mediated (Met+GOx)@HMSNs@(PPBEM-b-PDM) MN-array patch for antidiabetic drugs delivery.(B) Blood glucose changes in diabetic rats treated with an equal amount of metformin (1 mg) by injection;Met@HMSNs MNs;and (Met+GOx)@HMSNs@(PPBEM-b-PDM) MNs over time.Untreated diabetic rats serve as controls.

Fig.6.Microscopic images of H&E-stained tissue sections of rats in (A) control group without drug supply and (B) experimental group MNs loaded with (Met+GOx)@HMSNs@(PPBEM-b-PDM).

Finally,the microscopic images of H&E-stained tissue sections of rats in experimental group MNs loaded with (Met+GOx)@HM SNs@(PPBEM-b-PDM) were shown in Fig.6.Compared with the control group (Fig.6A),no obvious tissue damage and inflammation were observed,which indicated that the drug delivery system had no potential acute toxicity as a release device.Furthermore,we also investigated the degradation behaviors of HMSNs@(PPBEM-b-PDM) in physiological environment,i.e.,in PBS with a normal glucose concentration at 37 °C.The result shown in Fig.S11 demonstrated the HMSNs@(PPBEM-b-PDM) could almost degrade at the 6th day,indicating the as-prepared nanocarriers could be removed from the body.

4.Conclusions

To conclude,a soluble PVP MN array patch integrated with glucose mediated (Met+GOx)@HMSNs@(PPBEM-b-PDM) nanoparticles was designed to achieve painless and controlled transdermal drug delivery.The PPBEM-b-PDM block copolymer on the surface of HMSNs will change its conformation upon pH and H2O2stimulations,which unfolds the channels of HMSNs.Bothin vitroandin vivoresults showed this device presented a glucose mediated self-regulated drug release characteristic of rapid release at high glucose level but retarded release at normoglycemic level.This result indicates that the prepared MN device in this work has potential applications value in the therapy diabetes or other diseaseviatransdermal ingestion.

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 financially supported by the Zhejiang Provincial Natural Science Foundation of China (LY20E030005),Natural Science Foundation of Zhejiang Education Department(Y201942793)and the Opening Project of Jiangxi Province Key Laboratory of Polymer Micro/Nano Manufacturing and Devices(PMND201905).

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2021.07.019.