Enrofloxacin (ENR, purity ≥ 98%) was procured from Aladdin Biochemical Technology Co., Ltd. (Shanghai, China), while Enrofloxacin Standard (purity ≥ 98%) was obtained from Solarbio Science & Technology Co., Ltd. (Beijing, China). ENR commercially available tablets were obtained from Hebei Kexing Pharmaceutical Co., Ltd. (Shijiazhuang, China). mPEG-PLLA was sourced from Daigang Biomaterial Co., Ltd. (Jinan, China). Acetone came from Kemiou Chemical Reagent Co., Ltd. (Tianjin, China), and high - performance liquid chromatography (HPLC) - grade acetonitrile was purchased from Thermo Fisher Scientific (Waltham, USA). All other chemicals and solvents of analytical grade were commercially acquired.

Escherichia coli (E. coli) and Salmonella typhi (S. typhi) strain were sourced from China General Microbiological Culture Collection Center (Beijing, China). Healthy male beagles (weighing 11–16 kg) were obtained from Marshall Biotechnology Co., Ltd. (Beijing, China).

ENR-m were prepared using the solvent evaporation method, as described previously (21), with appropriate modifications. In this method, mPEG-PLLA (3 mg) and ENR (6 mg) were dissolved in 3 mL of acetone (oil phase) using an ultrasonicator (SB5200DT, Ningbo Scientz Biotechnology Co., Ltd., China). The mixture was injected into water (aqueous phase) at a rate of 0.3 mL/min using a syringe pump (LSP01-1A, Baoding Ditron electronic technology Co., Ltd., China), followed by solvent evaporation with a rotary evaporator (RE-52AA, Shanghai Yarong Biochemical Instrument Co., Ltd., China). After complete evaporation, a dried film was obtained. The film was then hydrated with 15 mL of ultrapure water. Finally, the mixture was filtered to remove excessive unencapsulated drug and stored at 4°C for subsequent analysis.

The DL and EE of the ENR-m formulations were measured using the centrifugation technique described in another report (23). In brief, 1 mL of the ENR-m was centrifuged at 12000 rpm for 10 min to obtain the supernatant. The supernatant was diluted with the mobile phase, and it was demulsified by ultrasound for 10 min. The amount of ENR in the micelles was determined by HPLC at 273 nm. The DL (%) and EE (%) values were determined using the equations described in reference (24), and each sample was assayed in triplicate. The DL (%) and EE (%) of the ENR-m were calculated using the following Equations 1, 2.

A three-factor, three-level BBD was employed to optimize the ENR-m. As shown in Supplementary Table S1, three main components were included as independent variables, along with their low, medium, and high levels. DL (%) and EE (%) were used as dependent responses to determine the optimal formulation conditions. Subsequently, Design-Expert® (version 13.0.1.0, State Ease Inc., Minneapolis, MN, USA) was used to generate a 17-run BBD. The optimized blank and ENR micelles were lyophilized at −80°C for pending further analysis.

To assess storage stability, accelerated stability environment testing of the ENR-m was performed. Briefly, ENR-m powders were stored at 40°C. A relative humidity (RH) of 75% was achieved within the stability chamber. An appropriate amount of the ENR-m was taken at 0, 15, 30, and 60 days, and the ENR content was determined by HPLC. DL (%) and EE (%) were used as key indicators.

The in vitro drug release behavior of ENR and the ENR-m was studied using the dialysis method (25) as follows: the ENR-m (5 mL) and ENR (4.5 mL, equivalent to 5 mL of the ENR-m) were placed in a dialysis bag with a molecular weight cutoff of 12,000 Daltons. The dialysis bag was incubated in 300 mL of phosphate buffer (37°C, pH 6.8) with shaking at approximately 100 rpm. At the predetermined time intervals, 1 mL of the solution was withdrawn from the release medium, and an equal volume of fresh medium was added to maintain the sink condition. The ENR content was quantitatively determined using UV–Vis spectrophotometry (UV-2450, Shimadzu, Japan) at 273 nm. The cumulative release rate (Q%) of ENR and the ENR-m was calculated using the following equation:

In the above equation, V₀ is the total volume of the release medium (mL); C_n is the concentration of enrofloxacin (mg/mL) measured during n time sampling; V_i is the volume of each sampling (mL); C_i is the concentration of enrofloxacin at i time point (mg/mL); W is the total content of enrofloxacin (mg).

A pharmacokinetic study was conducted based on previous research from our laboratory (26). A total of nine healthy male beagles were housed in an environment-controlled room with access to fresh food and water, and they were then randomly divided into three groups (n = 3). The powder samples were suspended in 0.5% CMC-Na, and then the uniform drug suspension was orally administrated to the beagles as a single dose of 10 mg/kg pure ENR, 11.11 mg/kg ENR-m, and 10.42 mg/kg commercial ENR tablets. At 0.083, 0.167, 0.25, 0.583, 0.75, 1, 1.5, 2, 4, 6, 8, 2, and 24 h post-administration, 1 mL of blood was drawn from the forelimb vein. All blood samples were transferred to heparin sodium-anticoagulated tubes and centrifuged at 6000 rpm for 10 min. The resulting supernatant was then collected and stored at −20°C for subsequent analysis. The animal experimental protocol and procedures used in this study were approved by the Animal Care and Use Committee of Hebei Agricultural University (Baoding, China; protocol number 2021058; approval date 28 February 2021) and were conducted in accordance with relevant guidelines and regulations.

Furthermore, 100 μL of plasma was combined with 300 μL of methanol. The mixture was vortexed for 3 min and centrifuged at 12000 rpm for 10 min. Subsequently, the supernatant was filtered and analyzed using HPLC. DAS 2.0 software (Mathematical Pharmacology Professional Committee of China, Shanghai, China) was used to calculate key pharmacokinetic parameters using the non-compartment method. These parameters included the maximum ENR plasma concentration (C_max), time to reach C_max (T_max), half-life (t_1/2), and the area under the ENR plasma concentration-time curve (AUC). One-way analysis of variance was performed using SPSS 19.0 to evaluate the significance of differences between the groups.

The ENR concentration was determined using an HPLC system (Model 1,525, Waters Corporation, MA, U. S. A). The setup included a Waters 2,998 PDA detector and a Waters C18 column (5 μm, 4.6 mm × 250 mm). The column was kept at 37°C. The mobile phase, a mixture of 0.025 M aqueous phosphoric acid (pH 2.5 ± 0.1 with triethylamine) and acetonitrile (82:18, v/v), was pumped at a flow rate of 1 mL/min. A total of 20 μL of the sample was injected for analysis.

The agar diffusion method was employed to observe the antibacterial activity of ENR and the ENR-m against E. coli and S. typhi (27). All bacterial strains were cultivated in nutrient broth (Hopebiol Ltd., Shandong, China) and incubated at 37°C for 24 h. A bacterial suspension of each strain (100 μL) was inoculated onto nutrient agar. On each inoculated plate, four stainless steel cylinders (F6621, Beyotime Institute of Biotechnology, Shanghai, China) of the same size (8 × 6 × 10 mm) were placed. ENR and the ENR-m were, respectively, dissolved in NaOH (0.1 mol/L) and sterile water, and the cylinders were filled with the sample solutions (2.5, 5, and 10 μg/mL). NaOH (0.1 mol/L) and sterile water were used as control groups, respectively. The agar plates were incubated at 37°C for 24 h, and a digital caliper was used to measure the diameters of the growth inhibition zones.

According to the single-factor experiments, EE (%) and DL (%) showed significant variations with changes in the concentrations of mPEG-PLLA, the water-to-oil ratio, and the feed ratio. In the BBD experiment, a total of 17 runs were conducted to optimize the formulation. The concentration of mPEG-PLLA (A) ranged from 0.1 to 3 mg/mL, the water-to-oil ratio (B) ranged from 10:2 to 10:4, and the feed ratio (mPEG-PLLA: ENR; C) ranged from 1:1 to 1:3. DL (Y1) and EE (Y2) were used as dependent variables (responses). Supplementary Table S2 displays the results of the BBD experiments conducted to evaluate the three independent variables.

The results in Supplementary Tables S3, S4 present the analysis of variance for the two responses. The lack-of-fit test for DL (%) resulted in p = 0.3321 and R² = 0.9971, while for EE (%), the values were p = 0.4207 (p > 0.05) and R² = 0.9950. This indicated an excellent goodness of fit at p < 0.0001. The coefficients of variation for DL (%) and EE (%) were 0.87 and 2.48, respectively.

To visualize the effects of the three factors on the drug loading of the ENR-m, response surface analysis was conducted to describe the regression equations that explain the relationship between these factors and the response. Both 2D contour plots and 3D response surface graphs were generated to assess the impact of single and multiple factors on the response (28). As depicted in Figures 1A–C, 2A–C, the 3D response surface graphs show the significant effects of the polymer concentration (A), water-to-oil ratio (B), and feed ratio (C) on the response of DL (%) and EE (%). This is also observed in the corresponding 2D contour plots (Figures 1D–F, 2D–F), which illustrate how the three factors worked together and interacted with each other to influence changes in DL (%) and EE (%).

By analyzing the 2D contour plots and 3D response surface graphs, the optimal formulation conditions were obtained as follows: mPEG-PLLA concentration of 0.17 mg/mL, water-to-oil ratio of 15:3.4 (w/w), and feed ratio of 1:2.4 (w/w). The optimized formulation was prepared. Then, the experimental results were compared with the predicted ones to validate the optimization process. As a result, the predicted values of DL (%) and EE (%) were 68.38 ± 0.22% and 88.40 ± 0.91%, respectively, in the calculated model, which showed no significant difference from the experimental values (p > 0.05). Thus, the BBD optimization of the ENR-m was adequately validated.

The particle sizes of blank micelles and ENR-m were determined using DLS. The results showed (Supplementary Table S5 and Figures 3A,B) that the average particle sizes of blank micelles and ENR-m were 109.03 ± 4.29 nm and 133.67 ± 3.10 nm, respectively, with an excellent PDI of 0.11 ± 0.07 and 0.13 ± 0.03. The results indicated that ENR-m had a homogeneous micelle system because of the low PDI value (<0.3) (29). The average size of micelles was slightly greater after ENR loading (from 109.03 nm to 133.67 nm).

The blank micelle and ENR-m were found to possess a spherical appearance with smooth surfaces, as shown by the TEM images (Figures 3C,D), which correlated well with the narrow particle size distribution. Spherical particles were found with no obvious aggregation. The particle sizes of the blank micelle and ENR-m were approximately 100 nm and 120 nm, respectively.

The FT-IR spectra of ENR, blank micelle, and ENR-m are provided in Figure 3E. The characteristic peaks of ENR appeared at 1736 cm⁻¹ (C=O stretching), 1,623 cm⁻¹ (C=O stretching of the pyridine group), 1,507 cm⁻¹ (benzene group), and 1,252 cm⁻¹ (C-F stretching of the benzene group) (30). The absorption peak from 3,000 to 2,700 cm⁻¹ was assigned to the stretching of methyl and methylene groups (31). In the blank micelles, strong characteristic peaks at 1760⁻¹ and 1,100 cm⁻¹ corresponded to the C=O stretching of the carboxyl group (a characteristic peak of the PLLA polymer segment) and the C–O stretching of the ester linkage in the polymer, respectively (32, 33). The broad absorption band observed between 4,000 and 3,500 cm⁻¹ in ENR and the blank micelles was assigned to the stretching of the O-H group. After forming ENR-m, strong characteristic peaks were observed at 1760 cm⁻¹ (C=O stretching from mPEG-PLLA), 1,623 cm⁻¹ (C=O stretching of pyridine rings from ENR), 1,507 cm⁻¹ (C-F stretching of the benzene ring from ENR), 1,252 cm⁻¹ (C-F stretching of the benzene ring from ENR), and 1,097 cm⁻¹ (C-O stretching of the ester linkage from mPEG-PLLA).

The accelerated stability of ENR micelles was evaluated at 40°C and 75% RH, and the major evaluation indicators were DL (%) and EE (%). As shown in Table 1, the DL (%) and EE (%) of ENR-m showed no significant changes over 0–60 days (p > 0.05). It revealed that ENR-m exhibited excellent chemical stability under environmental conditions of high temperature and humidity and could be stored for at least 60 days.

The in vitro release profile of ENR and ENR-m in the PBS solution (pH 6.8) is shown in Figure 3F. The cumulative release of ENR was approximately 57% at 6 h and 69% at 24 h. Pure ENR exhibited extremely low cumulative release and released only 69% over 24 h of the study, while the ENR-m released 100% within 6 h in pH 6.8 PBS. The results indicate that the solubility and in vitro release performance of ENR were improved by the ENR-m.

The mean plasma concentration-time curves of pure ENR, ENR-m, and commercial ENR tablets are shown in Figure 3G, and the pharmacokinetic parameters calculated using the non-compartment model are shown in Table 2. The commercial ENR tablet and pure ENR showed similar peak plasma concentrations and bioavailability. The C_max of ENR in the ENR-m was approximately 165% higher than that of pure ENR. Based on the values of AUC_0–24, the bioavailability of ENR from the ENR-m was approximately 160% greater than that of pure ENR.

The antibacterial activities of ENR and ENR-m against E. coli and S. typhi were determined using the agar diffusion method. The diameter of the inhibition zone was used to evaluate the strength of the antibacterial activity. The results are shown in Figures 4A–D, and the values of the inhibition zone diameters are presented in Supplementary Table S6. Around the agar wells inoculated with E. coli and S. typhi, the clear inhibition zones produced by the ENR-m were larger than those of the pure ENR and the control groups at the same concentration. As shown in Figure 4E, compared to ENR, the in vitro antibacterial activity of ENR-m against E. coli and S. typhi was significantly enhanced (p < 0.01). For E. coli, the inhibition zones of ENR-m increased by 16.43, 10.13, and 7.79% at concentrations of 2.5, 5, and 10 μg/mL, respectively, compared to those of pure ENR. Similarly, for S. typhi, the inhibition zones of ENR-m increased by 18.35, 13.23, and 14.49% at concentrations of 2.5, 5, and 10 μg/mL, respectively, compared to those of pure ENR.