Decomposition of Glucose-Sensitive Layer-by-Layer Films Using Hemin, DNA, and Glucose Oxidase

Glucose-sensitive films were prepared through the layer-by-layer (LbL) deposition of hemin-modified poly(ethyleneimine) (H-PEI) solution and DNA solution (containing glucose oxidase (GOx)). H-PEI/DNA + GOx multilayer films were constructed using electrostatic interactions. The (H-PEI/DNA + GOx)5 film was then partially decomposed by hydrogen peroxide (H2O2). The mechanism for the decomposition of the LbL film was considered to involve more reactive oxygen species (ROS) that were formed by the reaction of hemin and H2O2, which then caused nonspecific DNA cleavage. In addition, GOx present in the LbL films reacts with glucose to generate hydrogen peroxide. Therefore, decomposition of the (H-PEI/DNA + GOx)5 film was observed when the thin film was immersed in a glucose solution. (H-PEI/DNA + GOx)5 films exposed to a glucose solution for periods of 24, 48 72, and 96 h indicated that the decomposition of the film increased with the time to 9.97%, 16.3%, 23.1%, and 30.5%, respectively. The rate of LbL film decomposition increased with the glucose concentration. At pH and ionic strengths close to physiological conditions, it was possible to slowly decompose the LbL film at low glucose concentrations of 1–10 mM.

We have recently reported that hydrogen peroxide (H2O2) induced the decomposition of LbL films composed of hemin-modified poly(ethyleneimine) (H-PEI) and DNA [25]. Hemin is an iron porphyrin molecule and the iron porphyrin produces more reactive oxygen species (ROS), such as hydroxy radicals (OH), by reaction with H2O2 [26,27]. The ROS cause non-specific DNA cleavage [28][29][30], and H-PEI/DNA LbL film was consequently decomposed by the addition of H2O2. H2O2 is generated by the reaction of substrates and oxidases. LbL films composed of oxidases have been applied in biosensors and stimuli-responsive devices [31][32][33]. In the present work, we report the preparation of thin LbL films consisting of H-PEI and DNA with glucose oxidase (GOx), and the glucose-induced decomposition of these LbL film ( Figure 1). Glucose-sensitive materials respond to the blood glucose level of a diabetic patient; therefore, LbL films composed of GOx can be applied to an insulin DDS [34,35]. The decomposition of LbL films reported in this study is slow due to the stepwise reactions (enzymatic reaction, generation of ROS by hemin, and nonspecific cleavage of DNA). If decomposition of the membrane is slow, then the drug encapsulated in a capsule membrane can be released gradually. The glucose-induced decomposition of the LbL film in this work was achieved by the addition of glucose and the decomposition's dependence on the concentration of glucose was investigated.

Materials
Hemin and poly(ethyleneimine) (PEI, molecular weight; 60,000-80,000) were obtained from Tokyo Chemical Industry Co. (Tokyo, Japan). PEI has a random branched structure with the ratio of primary, secondary, and tertiary amino groups being nominally ca. 1:2:1. DNA (calf thymus) was purchased from Nacalai Tesque Inc. (Tokyo, Japan). GOx was obtained from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). All other reagents used were of the highest grade and used without further purification.

Apparatus
A quartz crystal microbalance (QCM; eQCM 10M Garmry, Warminster, UK) was employed for the gravimetric analysis of LbL films consisting of H-PEI and DNA. An 8 MHz AT-cut quartz resonator coated with a gold (Au) layer (surface area 0.2 cm 2 ) was used as a probe. The Au surface layer of the quartz resonator was cleaned using piranha solution (a mixture of H2O2 and H2SO4, 1:3 v/v) and thoroughly rinsed in pure water before use (CAUTION: piranha solution should be handled with extreme care). All QCM operations used flow QCM cells (cell volume ca. 120 µL; EQCM flow cell kit, BAS, Tokyo, Japan). Atomic force microscopy (AFM; AFM5200S, Hitachi High-Technologies Co., Tokyo, Japan) images were acquired in contact mode at room temperature in air. UV-vis spectroscopy measurements were conducted using a V-560 (Jasco, Tokyo, Japan) spectrometer.

Preparation of LbL Films
H-PEI/DNA films were prepared on the cleaned quartz resonator for QCM analysis. The quartz resonator was immersed in 0.1 mg/mL H-PEI solution in 10 mM HEPES buffer containing 150 mM NaCl (pH 7.4) for 15 min to deposit the first H-PEI layer by the hydrophobic force of attraction. After being rinsed in buffer for 5 min to remove any weakly adsorbed H-PEI, the quartz resonator was immersed in 0.1 mg/mL DNA solution for 15 min to deposit DNA by electrostatic interaction. At that time, 0.1 mg/mL GOx was mixed in the DNA solution. A second H-PEI layer was deposited similarly on the surface of the quartz resonator. The deposition steps were repeated to build up LbL films. Circular glass slides (18 mm diameter) and quartz slides (50 × 9 × 1 mm) with LbL films were prepared in the same manner. UV-vis absorption spectra of the LbL films in the working buffer were recorded on a UV-vis spectrometer. For dry AFM observations, the circular glass slides used to prepare each of the (H-PEI/DNA + GOx)5 films were rinsed with milli-Q water and dried for 24 h in a desiccator. AFM images were acquired in AC mode using an Arrow-NCR probe (Toyo Corporation, Tokyo, Japan) at room temperature in air.

Decomposition of LbL Films
The H2O2-induced decomposition of the (H-PEI/DNA + GOx)5 films was studied using UV-vis absorption spectroscopy. The LbL films prepared on quartz slides were exposed to 1, 10, and 100 mM H2O2 solution (pH 7.4) for 30, 60, 90, 120, 180, and 240 min, and then rinsed with the working buffer Hemin H-PEI PEI for 5 min. The films were measured for absorption at a particular time and then subsequently immersed in the H2O2 solution for the next exposure step.
The glucose-induced decomposition of the (H-PEI/DNA + GOx)5 films was monitored in the same manner. The LbL films prepared on quartz slides were exposed to 1, 10, and 100 mM glucose solution (pH 7.4) for 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8, 10, 24, 30, 35, and 48 h. The glucose-induced decomposition of the LbL films was studied by monitoring the resonance frequency change (ΔF) of the 5-bilayer film-coated quartz resonator in the flow-through cell of the QCM. After rinsing the filmdeposited probe with buffer for 5 min, the steady-state frequency was recorded. The film-deposited probe was measured for frequency at a particular time and then subsequently immersed in the glucose solution for the next exposure step. There have been other reports on the preparation of LbL films by electrostatic interaction using QCM [36]. LbL films using DNA or GOx driven by electrostatic interactions have been reported by Zhang et al. [37,38] and Kakade et al. It is considered that DNA and GOx are both present in the negatively charged layers because DNA and GOx adsorb to positively charged polymers.    Figure S1). H-PEI is adsorbed on the surface of the DNA layers, and the LbL film has an absorption maximum derived from hemin. The disappearance of the absorption at 390 nm was confirmed after the (H-PEI/DNA + GOx)5 film was exposed to 100 mM H2O2 solution for 1 h. The iron porphyrin is known to be oxidized by H2O2, according to Equation (1) [39,40].

Results and Discussion
Porphyrin-Fe(III) + H2O2 → Porphyrin-Fe(IV)=O + OH + H + The oxidation of iron porphyrin significantly reduces the absorption at 390 nm ( Figure S1). The iron porphyrin contained in hemin produces ROS. The H2O2 causes the hemin to be degraded, so that the absorption maximum derived from hemin in the film was significantly reduced. Therefore, the absorption at 390 nm is not suitable for the evaluation of LbL film decomposition. The absorbance of the (H-PEI/DNA + GOx)5 film at 260 nm significantly decreased when immersed in 100 mM H2O2 solution. The DNA solution has an absorption at 260 nm, and there is no significant change in absorption at 260 nm after hydrogen peroxide treatment ( Figure S2). The absorbance coefficient of GOx is small compared to DNA and hemin ( Figure S3). Therefore, GOx has a negligible effect on the absorbance, unless the concentration is high. The oxidation of the iron porphyrin slightly reduces the absorbance contribution at 260 nm, whereas the contribution of the (H-PEI/DNA + GOx)5 film to the absorbance is more significant (Table S1). If ROS are generated near the LbL films, then nonspecific cleavage of DNA is triggered. Yoshida et al. reported the decomposition of H-PEI/DNA nanofilm by H2O2 [25]. Similarly, the decomposition of the (H-PEI/DNA + GOx)5 films by H2O2 resulted in a decrease in the absorption maximum derived from DNA. Therefore, changes in the absorption at 260 nm can be an indicator of the degradability of the (H-PEI/DNA + GOx)5 film. The H2O2-induced decomposition of (H-PEI/DNA + GOx)5 films was investigated using UV-vis absorption spectroscopy ( Figure 6). The absorbance at 260 nm is attributed to hemin, DNA, and GOx. Therefore, if the film is decomposed, the constituent components of the film on the quartz slides are lost, and the absorbance at 260 nm decreases. Therefore, the decomposition of the LbL film was evaluated using an absorbance ratio. The extent of the H2O2-induced decomposition of the LbL films was determined using Equation (2). (2) When the (H-PEI/DNA + GOx)5 film was exposed to the buffer for a long time, the absorbance did not change. On the other hand, the (H-PEI/DNA + GOx)5 film immersed in H2O2 solution showed a decrease in the absorbance ratio, and the decrease was larger as the H2O2 concentration was higher. Significant decrease in the absorbance ratio can thus indicate the decomposition of the LbL films by H2O2. In addition, the decomposition of the membrane is affected by the H2O2 concentration. If the same method is used, then the glucose-induced decomposition of (H-PEI/DNA + GOx)5 film can be evaluated. The glucose-induced decomposition of (H-PEI/DNA + GOx)5 films was investigated using UVvis spectroscopy (Figure 7). The glucose-induced LbL films were determined using Equation (3).
Ab /Ab 0 × 100 = absorbance at 260 nm of (H − PEI/DNA + GOx) 5 film when immersed in glucose solution for t min absorbance at 260 nm of (H − PEI/DNA + GOx) 5 film × 100 When (H-PEI/DNA + GOx)5 films were exposed to 1, 10, and 100 mM glucose solutions for 48 h, the absorbance ratio was 76.1%, 70.2%, and 64.0%, respectively. Hemin is an iron porphyrin molecule and is an active cofactor for various enzymes, such as catalase and peroxidase [28]. GOx present in the LbL films reacts with glucose to generate gluconic acid and H2O2. H2O2 present in the LbL films reacts with hemin to generate ROS. The nonspecific cleavage of DNA by ROS promotes the partial degradation of the LbL films. In addition, the rate of LbL film decomposition increased with the glucose concentration. The decomposition rate of the film immersed in the glucose solution was slower than that immersed in the H2O2 solution. However, the influence of the glucose concentration on the decomposition of the (H-PEI/DNA + GOx)5 film was less than that from the H2O2 concentration. GOx and glucose are enzymatic reactions; therefore, H2O2 is generated slowly and the degradation of the LbL film is delayed. Therefore, at pH and ionic strengths close to physiological conditions, it is possible to slowly decompose the LbL films at a low glucose concentration (1-10 mM). The glucose-induced decomposition of (H-PEI/DNA + GOx)5 films was investigated using the QCM (Figure 8). The extent of the decomposition was calculated from the change in the resonance frequency (Equation (4)).
∆F ∶ frequency change of the (H − PEI/DNA) 5 films ∆F ∶ frequency change difference between blank and (H − PEI/DNA) 5 films immersed in glucose solution fot time When the (H-PEI/DNA + GOx)5 film was exposed to 100 mM glucose solution for 24, 48, 72, and 96 h, the extent of decomposition was 9.97%, 16.3%, 23.1%, and 30.5%, respectively. However, the results of Figure 8 show less decomposition of LbL films than expected compared to Figure 7. Compared to QCM operation, the amount of decomposition of LbL films immersed in glucose solution during UV operation is greater. The absorbance ratio is influenced by hemin oxidation; therefore, there was a difference in the amount of membrane degradation. The glucose-induced decomposition of LbL films reported by Sato et al. occurred for 1 h when glucose was added [41]. On the other hand, from the QCM results, a longer immersion time in 100 mM glucose solution resulted in greater decomposition of the LbL films. Depending on the glucose concentration, it is possible to slowly decompose the (H-PEI/DNA + GOx)5 film.

Conclusions
(H-PEI/DNA + GOx)5 LbL films were prepared by alternate immersion of a substrate in H-PEI solution and DNA solution (containing GOx). When the (H-PEI/DNA + GOx)5 films were immersed in a H2O2 solution, partial decomposition of the LbL films was observed. The iron porphyrin in hemin produces more ROS from the reaction with H2O2 [26,27]. The ROS cause non-specific DNA cleavage; therefore, the decomposition of the LbL films composed of DNA was observed. Furthermore, partial degradation was observed when the membrane was immersed in a glucose solution. GOx present in LbL films produced H2O2 from glucose. These LbL films decomposed under physiological conditions with various glucose concentrations, which suggests that a glucose stimuli-responsive nanofilm could be realized. We have developed other glucose-and pH-responsive thin films [42,43]; however, the (H-PEI/DNA + GOx)5 film has a very slow decomposition. If a drug could be encapsulated in this thin film, then there is a possibility that a system capable of drug release over a long time period, depending on the substrate, could be realized.

Supplementary Materials:
The following are available online at www.mdpi.com/xxx/s1, Figure S1: UV-vis absorption spectra for 30 g/mL hemin solution (a) before and (b) after treated with 100 mM H2O2 (pH 7.4) for 1 h. The remaining H2O2 was consumed by dropping 10 L of 0.1 mg/mL catalase into the test solution (2 mL) and letting it stand for 24 h. Figure S2: UV-vis absorption spectra for 13.6 g/mL DNA solution (red) before and (blue) after treated with 100 mM H2O2 (pH 7.4) for 1 h. The remaining H2O2 was consumed by dropping 10 L of 0.1 mg/mL catalase into the test solution (2 mL) and letting it stand for 24 h. Figure S3: UV-vis absorption spectra for 110 g/mL GOx solution (pH 7.4). Table S1: Absorbance difference in hydrogen peroxide treatment.
Author Contributions: K.Y. and K.S. designed the work. K.Y. and Y.K. (Yoshitomo Kashiwagi) collected the materials. K.Y. and Y.K. (Yu Kashimura) conducted experimental works on detailed LbL film data. T.K. and T.S. performed the AFM observations. K.Y., T.D., T.O. and K.S. were involved in providing experimental advice and operations. The manuscript was prepared by K.Y and K.S. All authors have read and agreed to the published version of the manuscript. Decomposed / % Time / h also acknowledged for financial support. The authors thank staff of Tsuruoka College for assisting with the AFM observations.

Conflicts of Interest:
The authors declare no conflicts of interest.