Preparation, characterisation, and controlled release of sex pheromone-loaded MPEG-PCL diblock copolymer micelles for Spodoptera litura (Lepidoptera: Noctuidae)

Sex pheromones are important for agricultural pest control. The main sex pheromone components of Spodoptera litura are (Z,E)-9,11- and (Z,E)-9,12-tetradecadienyl acetate (Z9,E11-14:Ac; Z9,E12-14:Ac). In this study, we investigated the optimal conditions for encapsulation of S. litura sex pheromonesin micelles via the self-assembly method using monomethoxy poly (ethylene glycol)-poly (ε-caprolactone) (MPEG-PCL) as a biodegradable wall-forming material with low toxicity. In the L9(34) orthogonal experiment, 3 amphiphilic block copolymers, with different hydrophilicity to hydrophobicity ratios, were examined. Optimal encapsulation conditions included stirring of MPEG5000-PCL2000 at 1000 rpm at 30°C with 2.5:1 wall-forming: core material mass ratio. S. litura sex pheromone-loaded MPEG5000-PCL2000 micelles presented a homogeneous spherical morphology with apparent core-shell structure. The release kinetics of optimized MPEG5000-PCL2000 micelles was best explained by a first-order model. Encapsulated Z9,E11-14:Ac and Z9,E12-14:Ac were released slowly, not suddenly. Methyl oleate (MO) was used as an agent to control micellar release performance. When MO content equalled block content, micelle half-life could be prolonged, thereby controlling the release speed. Overall, our results showed MPEG-PCL as a promising controlled-release substrate for sex pheromones.


Introduction
Spodoptera litura Fabricius (Lepidoptera: Noctuidae), a type of polyphagous pest with an aggressive eating pattern, has a wide range of hosts, encompassing approximately 200 kinds of plants [1] including those of various vegetables, fruits, baccies, cotton, corn, tea, and other cash crops [2]. However, due to the overuse of chemical agents to prevent and control this a1111111111 a1111111111 a1111111111 a1111111111 a1111111111

Critical micelle concentration
The critical micelle concentration (CMC) of MPEG-PCL was measured by a UV spectrophotometer (TU-1810, Beijing Purkinje General Instrument Co., Ltd.). Three kinds of MPEG-PCL, with different proportions of hydrophobic and hydrophilic components were dissolved in deionized water to obtain a stock solution of concentration 1.000 g/L. The absorption maxima of the different concentrations of MPEG-PCL (0.001, 0.005, 0.008, 0.01, 0.04, 0.05, 0.1, 0.2, 0.35, 0.5, 0.7, and 1 g/L) was recorded and mapped with the values of lgA-lgC (A: absorption, C: concentration); critical micelle concentration of block copolymer corresponded to the concentration at which the first derivative curve reached zero [37].

Orthogonal experimental design
An L 9 (3 4 ) orthogonal table (Table 1) was adopted for this test. The investigated factors included wall-forming materials (W), mass ratio of sex pheromone to wall-forming materials (W/S ratios), reaction temperature (T), and stirring speed (S); encapsulation efficiency of the micelles (EE) was considered as the assessment index. The optimised formulation was prepared in triplicate.

Determination of entrapment efficiency
Briefly, 0.5 mL of the sex pheromone-loaded micelle solution was fully mixed with 0.5 mL ultrapure water. The solution was extracted with 1 mL n-hexane and completely disrupted using an ultrasound sonicator (Scientz-IID, Ningbo Scientz Biotechnology Co., China) on ice. The encapsulated sex pheromone was dissolved in the hexane after 30 min. The concentration of sex pheromone was determined by gas chromatography (GC, Agilent 7890B, Agilent Technologies, Santa Clara, CA, USA). For GC, a capillary column (HP-5, 30 m × 0.32 mm × 0.25 μm) with a flame ionisation detector and a splitless injector, with nitrogen as the carrier gas, was used. GC conditions were as follows: the column temperature set at 80˚C (held for 5 min), raised to 210˚C at 10˚C/min, and held at 210˚C for 15 min. A standard curve was generated according to the concentration of sex pheromones and peak area; quantity of each component in the sex pheromone was determined from the standard curve. The standard curve regression equations of Z9,E11-14:Ac and Z9,E12-14:Ac were y = 16806x + 50.5 (R 2 = 0.9997) and y = 18672x − 3.4706 (R 2 = 0.9999). The sex pheromone entrapment efficiency was calculated using Eq 1: Characterisation of micelles Particle morphology. The micelle morphology was observed by transmission electron microscope (TEM, HT 7700, Hitachi, Tokyo, Japan), and the speeding voltage during the test was 80 kV. Samples were prepared by dropping the micelle solution on a carbon-coated copper net, followed by air drying, and dyeing with 0.2 wt% phosphotungstic acid.
Determination of particle size. The particle size and its distribution were analysed using a Malvern nanometre particle size analyser (MNPSA) (Zetasizer Nano S90, Malvern Instruments Ltd., Malvern, UK).
Stability of micelles. Micelles were stored at 2, 4, and 8˚C in the dark. In order to evaluate the physical stability of nanoparticles during this storage period, particle size distribution was monitored at time intervals of 0, 15, and 30 days, using the method described in the section "Determination of particle size".

Release performance
Sex pheromone release. To evaluate sex pheromone release, the micelles were transferred to a centrifuge tube and placed in an artificial climate chamber (MGC-450HP2, Shanghai Yiheng Co., China) with controlled temperature in the range of 35 ± 3˚C, light: dark cycle of 12 h:12 h, and relative humidity of 75 ± 5% for a period of 28 days. The samples were taken out of the artificial climate chamber at regular time intervals for sex pheromone examination by GC. To evaluate the release of sex pheromones from micelles prepared under optimal conditions, the samples were examined every day during the first 14 days, and every 7 days during the subsequent 14 days. To evaluate the release from micelles containing the controlled-release agent, the samples were examined every 3 days over a period of 15 days. Three samples were used in each experiment.
Sex pheromone release was expressed as percentage of accumulated release, since this enabled the evaluation of performance of different micelles. Accumulated release was calculated using Eq 2: where W 0 is the sex pheromone content at the initial time and W t is the sex pheromone content at each recorded time. Sex pheromone release kinetics for optimized micelles. For a better understanding of the efficacy of sex pheromones, their release kinetics were studied. Selection of a suitable kinetic model for fitting the sex pheromone release data helped determine the release characteristics. There are a number of kinetic models that describe the overall release of sex pheromone from the vehicle. The most common mathematical models used are: zero-order model (Eq 3), first-order model (Eq 4), Higuchi model (Eq 5), Korsmeyer-Peppasmodel (Eq 6), and Hixson-Crowell model (Eq 7) [39][40][41][42][43][44][45]: where C t -amount of drug released in time t, C0-the initial amount of drug, K0-zero-order kinetic constant, K1-first-order kinetic constant, K H -Higuchi kinetic constant, K KP -Korsmeyer-Peppas release constant, KHC-Hixson-Crowell release constant, n-diffusional release exponent, t-time.
Half-life calculations. Depletion of pheromone components from the micelle formulations was characterised by the first-order kinetic model: lnC t = lnC 0 +K 1 Át. Half-lives (t 1/2 ) for compounds were determined from the exponential equation, substituting calculated values of C 0 and K 1 , and setting (C t /C 0 ) to 0.5 [46]. Statistical analysis. Statistical analysis was done with SPSS 17.0 software package (Chicago, IL, USA). One-way analysis of variance (ANOVA) for independent samples followed by Duncan's multiple range tests were performed to evaluate the quantitative results. Data were obtained from triplicate samples and, expressed as mean ± standard error (SE); values of P 0.05 and P 0.01 were considered statistically significant and extremely significant, respectively.

Optimisation of MPEG-PCL micelle formation
The results of the L 9 (3 4 ) orthogonal experiments using MPEG-PCL nanoparticles are shown in Tables 2 and 3 Conversely, it was much easier to draw a more intuitive conclusion from the results by range analysis of the orthogonal experiment. However, the calculation processes were extensive and could not evaluate the errors; thus, it was necessary to carry out variance analysis of the orthogonal experiment results. It can be seen from the variance analysis tables (Tables 4 and 5) that except for the W/S ratios, all the other factors (including W, S, and T) had significant effects on the experimental results. The order of factors affecting the encapsulation efficiency of Z9,E11-14:Ac and Z9,E12-14:Ac, obtained from variance analysis, was the same as that from the range analysis.
Encapsulation efficiency of the micelles was controlled by the length of hydrophobic or hydrophilic chain (wall-forming materials), W/S ratio, T, and S. Based on the two analyses, it was concluded that the order of effect of W and T on the encapsulation efficiency was different. Since the mass ratio of Z9,E11-14:Ac was much larger than that of Z9,E12-14:Ac, factor S was regarded as the most important factor affecting the encapsulation efficiency followed by T, W, and W/S ratios. S likely played an important role in the formation of micelles, since the sex pheromone should be well mixed in the process of micelle formation, and a certain speed would be required when water is added to the solution to conjugate the hydrophilic ends of the amphiphilic block copolymer. The influence of W was determined by the length of hydrophobic and hydrophilic chains, whereas T likely influenced micellar assembly and speed of sex pheromone volatility to lessen the encapsulation efficiency. However, the influence of W/S ratios on encapsulation efficiency was relatively small.
The optimal conditions for S. litura sex pheromone encapsulation with MPEG-PCL, determined from the above results, involved stirring MPEG 5000 -PCL 2000 at a speed of 1000 rpm at 30˚C with a 2.5:1 mass ratio of wall-forming to core materials. Based on these conditions, three parallel experiments with MPEG-PCL micelles were subsequently conducted ( Table 6). The results consistently showed that entrapment efficiency was the highest among the combinations used in the orthogonal experiments, which verified the utility and feasibility of the conditions.

Characterisation of microcapsules
For fresh MPEG 5000 -PCL 2000 nanoparticles, prepared according to the optimised formulation and preparation conditions, the particle size was 374 ± 5.13 nm by MNPSA (Fig 1). The formation of micellar nanostructures was confirmed by TEM. The MPEG 5000 -PCL 2000 nanoparticles showed a homogeneous spherical morphology, with average diameter of 300 nm, presenting an apparent core-shell structure (Fig 2). The size of the MPEG 5000 -PCL 2000 nanoparticles, measured by TEM, was smaller compared to that from MNPSA measurements, since the former was related to the collapsed nanoparticles after water evaporation, whereas the latter represented their hydrodynamic diameter [47].
After preparation, the micelles were dispersed in aqueous medium. Therefore, stability of their sizes was of great importance, both as a measure of particle structure integrity and as an   indicator of the possible inter-particular associations (aggregation). At sub-zero temperatures, the solution solidified and the micellar structure lost its integrity. For this purpose, we chose 2, 4, and 8˚C as the storage temperatures, at which the particle size was monitored in the dark over a period of 30 days. The variation of micellar size as a function of storage time is shown in Table 7. All the micelles increased slightly in size, throughout the measurement period, at different temperatures. This observation could not be an indicator of aggregation, which usually leads to a several-fold increase in size; instead, copolymer swelling and/or hydration may be responsible for this event [48]. Since the variation of micellar size was less when stored at 2˚C, we chose to store the micelles at 2˚C in the dark for the best storage conditions.

Sex pheromone release kinetics in optimized micelles
The sex pheromone release results of MPEG 5000 -PCL 2000 micelles were used in various mathematical models to evaluate the kinetics and mechanism of release from the micelles. Based on the correlation coefficient (R) value in various models, the one that fit best with the release data was selected; the one with a high 'R' value was considered as the best fit. The release constant was calculated from the slopes of the appropriate models, and the regression coefficient (R 2 ) was determined (Table 8).

Release performance of optimized micelles
The plot of accumulated release from sex pheromone-loaded MPEG 5000 -PCL 2000 micelles indicated that Z9,E11-14:Ac could be released from micelles faster than Z9,E12-14:Ac, in a sustained manner. The two components had a high release rate in the first 3 days, which was attributed to the fact that nanoparticles usually contain sex pheromone not only at the inner core but also on their surface. After this initial loss, sex pheromone release approximated firstorder release rates more closely [49]. Accordingly, following the first burst release period, sex pheromone was released slowly, independent of the initial sex pheromone concentration in the micelles. As shown in Fig 3, from day 4 to 14, the release rate tended to slow down and remained constant. After 14 days, the release rate decreased further and tended to be stable, although the release rate of Z9,E11-14:Ac was less than that of Z9,E12-14:Ac. According to the first-order kinetic model, the half-life of Z9,E11-14:Ac and Z9,E12-14:Ac in the micelle was 5.6 and 7.0 days, respectively. The half-life difference of 1.4 days may have been due to the different proportions of sex pheromone components in the micelle. Based on the results of this study, we found that Z9,E11-14:Ac and Z9,E12-14:Ac were released slowly from MPEG 5000 -PCL 2000 micelles, and that no sudden release occurred throughout the process, thereby indicating that diblock copolymer micelles were suitable for use as a controlled substrate. Our studies indicated that although MPEG-PCL diblock copolymer micelles did not maintain a constant release rate, they met the first-order kinetic model requirements, with adynamic rapid-to-slow release, lasting for almost a month. Other release carriers, such as PVC, have demonstrated equal or better release duration for that pheromone [46]. However, the micelles in this study were in aqueous solution and hence environment friendly; they were physically and chemically stable, non-toxic, and biodegradable. Table 9 shows that while the differences among the tested concentrations of MO were not significant, when the mass of wall-forming materials equalled that of MO (10 mg/mL), the halflife of Z9,E11-14:Ac and Z9,E12-14:Ac in the micelle increased by 3.7 and 4.2 days, respectively, compared to that of the control. With the increased content of controlled-release agent,  the efficiency of controlled release declined, potentially due to the organic liquid which may have affected micelle formation and inhibited the encapsulation efficiency, thereby impacting the release rate. Fig 4 shows that MO, as a controlled-release agent, could retard the overall release rate of micelles, especially in the first 3 days without burst release. Compared to that of the control, release rate of the two components was slower over the first 6 days. From day 7 to 15, the release rates increased relative to that during the first 6 days. Thus, addition of appropriate quantities of MO into the micelle could prolong the half-life and control the release performance.

Conclusions
The optimal preparation conditions of S. litura sex pheromone-amphiphilic block copolymer micelles were shown to involve stirring MPEG 5000 -PCL 2000 at a speed of 1000 rpm at 30˚C with a 2.5:1 mass ratio of wall-forming to core materials. The nanoparticles presented a homogeneous spherical morphology with an apparent core-shell structure, and were free from the inter-micellar adhesion phenomena. The release kinetics of optimized MPEG 5000 -PCL 2000  micelles was best explained by first-order model. Since the release from micelles was slow, without a sudden-release phenomenon, the amphiphilic copolymer was considered suitable for use as a controlled substrate. When the mass of added MO equalled that of wall-forming materials, the half-life could be prolonged, thereby allowing control of the release rate. These results indicated that the diblock copolymer could be a suitable controlled-release substrate, and the micelles could have potential use in the control applications of mass trapping and mating disruption in the field.