Biofunctional lipid nanoparticles for precision treatment and prophylaxis of bacterial infections

Abstract

The lack of bacterial-targeting function in antibiotics and their prophylactic usage have caused overuse of antibiotics, which lead to antibiotic resistance and inevitable long-term toxicity. To overcome these issues, we develop neutrophil-bacterial hybrid cell membrane vesicle (HMV)-coated biofunctional lipid nanoparticles (LNP@HMVs), which are designed to transport antibiotics specifically to bacterial cells at the infection site for the effective treatment and prophylaxis of bacterial infection. The dual targeting ability of HMVs to inflammatory vascular endothelial cells and homologous Gram-negative bacterial cells results in targeted accumulation of LNP@HMVs in the site of infections. LNP@HMVs loaded with the antibiotic norfloxacin not only exhibit enhanced activity against planktonic bacteria and bacterial biofilms in vitro but also achieve potent therapeutic efficacy in treating both systemic infection and lung infection. Furthermore, LNP@HMVs trigger the activation of specific humoral and cellular immunity to prevent bacterial infection. Together, LNP@HMVs provide a promising strategy to effectively treat and prevent bacterial infection.

Fig. 1.. Schematic illustrations of the construction and application of LNP-N@HMVs.

(A) Construction of hybrid cell membrane–coated antibiotic delivery system LNP-N@HMVs. (B) Application of LNP-N@HMVs in dual-targeted treatment and prophylaxis of bacterial infections.

Fig. 2.. Characterization of HMVs and LNP-N@HMVs.

(A) TEM images of E. coli–derived OMVs, HL-60 cell–derived NMVs, and HMVs. Scale bars, 200 nm. (B) Fluorescent spectra of OMVs labeled with FRET dye pair (DiO and DiI) after sonication with unlabeled NMVs at different protein weight ratios. (C) Fluorescence colocalization images of HMVs and a mixture of DiO-labeled OMVs and DiI-labeled NMVs. Scale bars, 5 μm. (D) TEM images of LNP-Ns and LNP-N@HMVs. Scale bars, 200 nm. (E) Hydrodynamic size and zeta potential of LNP-Ns, HMVs, and LNP-N@HMVs. n = 3; **P < 0.01; ns, nonsignificance. (F) SDS-PAGE protein analysis of NMVs, OMVs, and LNP-N@HMVs. (G) Nor release profiles from LNP-Ns and LNP-N@HMVs at pH 7.4 over 36 hours (n = 3).

Fig. 3.. In vitro targeting of LNP@HMVs to inflammatory endothelium cells and homologous bacteria.

(A) Confocal microscopic images of nonactivated or TNF-α–activated HUVECs cultured with LNP-N@HMVs (concentration, 20 μg ml⁻¹ NPs) for 0.5 hour, where rhodamine B (red) and DiO (green) were used to stain LNPs and HMVs, respectively. Scale bars, 20 μm. (B) Confocal microscopic images and (C and D) flow cytometry analysis of TNF-α–activated HUVECs cultured with LNP-Ns, LNP-N@NMVs, LNP-N@OMVs, or LNP-N@HMVs for 1 hour (concentration, 20 μg ml⁻¹ NPs). Scale bars, 20 μm. Confocal microscopic images of (E) E. coli and (F) K. pneumoniae and (G and H) flow cytometry analysis of rhodamine B in these bacteria after incubation with LNPs or LNP@HMVs for 1 hour (concentration, 20 μg ml⁻¹ NPs). Scale bars, 2 μm. n = 3; *P < 0.05, **P < 0.01.

Fig. 4.. In vitro antibacterial and antibiofilm activity of LNP-N@HMVs.

(A) Photographs and quantification of the bacterial colonies after treating E. coli (~10⁵ CFU ml⁻¹) with Nor, LNP-Ns, or LNP-N@HMVs with equivalent Nor concentration at 0.12 μg ml⁻¹ for 8 hours. Bacteria without any treatment were used as control. Limit of detection: 100 CFU ml⁻¹. (B) Confocal microscopic images of E. coli (~10⁸ CFU ml⁻¹) with different treatments at the equivalent Nor concentration of 0.24 μg ml⁻¹ and stained with Live/Dead BacLight bacterial viability kit (green, all the bacteria; red, dead bacteria). Scale bar, 20 μm. (C) Photographs and quantification of the colonies of K. pneumoniae after the same treatments as that of E. coli (A). Limit of detection: 100 CFU ml⁻¹. (D) Confocal microscopic images of E. coli biofilm with different treatments at the equivalent Nor concentration of 0.5 μg ml⁻¹ for 24 hours and stained with Live/Dead BacLight bacterial viability kit. (E) Biomass of E. coli biofilm after treatment with different formulations at various Nor concentrations. n = 3; *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 5.. In vivo targeting and therapeutic effects of LNP-N@HMVs against bacterial infections.

(A) Schematics of the experiments to evaluate targeting function of LNP@HMVs in a mouse muscle infection model. (B) Representative fluorescence images of the infected mice under different treatments at 3, 6, 9, and 24 hours after injection. (C) Mean DiR fluorescence intensity of the infected and the uninfected sites in different treatment groups (n = 3). (D) Schematics of the experiments to evaluate the therapeutic efficacy of LNP-N@HMVs (equivalent Nor dose, 25 μg/20 g) in treating mice with E. coli–induced peritonitis. (E) Quantification of the number of bacteria in peritoneal fluid, spleen, lung, and kidney of the infected mice with different treatments. Infected mice without any treatment were used as control (n = 5). (F) Schematics of the experiments to evaluate targeting and therapeutic efficacy of LNP-N@HMVs (equivalent Nor dose, 1 μg/20 g) in treating mice with K. pneumoniae–induced lung infection. (G) Representative fluorescence images and (H) mean DiR fluorescence intensity of the infected mice under different treatments (n = 3). (I) Quantification of the number of bacteria in the infected lung tissues of mice under different treatments (n = 5). *P < 0.05, **P < 0.01, and ***P < 0.0001.

Fig. 6.. In vitro and in vivo immune responses activated by LNP@HMVs.

(A) Expression levels of CD80⁺ and CD86⁺ in BMDCs (CD11c⁺) and (B) production of IL-6 and TNF-α of the cell supernatant measured by ELISA after different treatments (concentration, 2 μg ml⁻¹ NPs) for 24 hours (n = 3). (C) Schematics of experimental design to evaluate the in vivo short-term immune responses activated by a single dose of various LNPs (20 μg/20 g) via subcutaneous injection. (D) Percentage of CD80⁺ and (E) CD86⁺ cells (gated on CD11c⁺ cells) in the LNs at day 3 after immunization (n = 6). (F) Percentage of CD3⁺ T cells, (G) CD8⁺ T cells (gated on CD3⁺ T cells), and (H) CD4⁺ T cells (gated on CD3⁺ T cells) in the spleen at day 3 after immunization (n = 6). Naive mice without immunization were used as control. *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 7.. In vivo prophylactic effects of LNP@HMVs against E. coli infections.

(A) Schematics of the short-term prophylactic experiment design of a single dose of LNP@OMVs or LNP@HMVs (20 μg/20 g). Quantification of the number of bacteria in (B) peritoneal fluid, (C) spleen, (D) lung, and E) kidney of the infected mice after different treatments. Infected mice without any treatment were used as control (n = 6). (F) Survival curves of the mice intraperitoneally infected with E. coli (1 × 10⁷ CFU/20 g) after different treatments (n = 6). (G) Schematics of the long-term prophylactic experiment design of two doses of LNP@OMVs or LNP@HMVs (20 μg/20 g). (H) Survival curves of the infected mice after different treatments (n = 8). Infected mice without any treatment were used as control. **P < 0.01.