A macromolecular approach to eradicate multidrug resistant bacterial infections while mitigating drug resistance onset

Abstract

Polymyxins remain the last line treatment for multidrug-resistant (MDR) infections. As polymyxins resistance emerges, there is an urgent need to develop effective antimicrobial agents capable of mitigating MDR. Here, we report biodegradable guanidinium-functionalized polycarbonates with a distinctive mechanism that does not induce drug resistance. Unlike conventional antibiotics, repeated use of the polymers does not lead to drug resistance. Transcriptomic analysis of bacteria further supports development of resistance to antibiotics but not to the macromolecules after 30 treatments. Importantly, high in vivo treatment efficacy of the macromolecules is achieved in MDR A. baumannii-, E. coli-, K. pneumoniae-, methicillin-resistant S. aureus-, cecal ligation and puncture-induced polymicrobial peritonitis, and P. aeruginosa lung infection mouse models while remaining non-toxic (e.g., therapeutic index-ED₅₀/LD₅₀: 1473 for A. baumannii infection). These biodegradable synthetic macromolecules have been demonstrated to have broad spectrum in vivo antimicrobial activity, and have excellent potential as systemic antimicrobials against MDR infections.

Fig. 1

Synthesis of monomers and polymers. a General scheme for the modular synthesis of Boc-protected guanidine-functionalized cyclic carbonate monomers (MTC-OX-BocGua) bearing various hydrophobic spacer groups; b general scheme for synthesis of guanidinium-functionalized polycarbonates (P(MTC-OX-Gua)) by metal-free ROP followed by acid-mediated removal of Boc groups; c GPC chromatograms of representative polycarbonates (prior to deprotection)

Fig. 2

Antimicrobial (MIC) and hemolytic (HC₅₀) activities. R₁₀ and R₂₀ represent arginine peptide with 10 and 20 amino acids, respectively. ^aAt the highest concentration (8000 µg/mL) tested, extent of hemolysis: <5%. The extent of hemolysis for pPh_20 at the threshold limit of 8000 µg/mL: ∼40%

Fig. 3

Antibacterial activity of guanidinium-functionalized polycarbonates. a MIC and selectivity (HC₅₀/MIC) of polymers against clinically isolated multidrug-resistant bacteria (A. B.: A. baumannii; K. P.: Klebsiella pneumoniae; MIC of R₁₀ and R₂₀: >512 µg/mL against all the strains of bacteria tested); b killing kinetics of A. baumannii 10073 at different concentrations as specified; c killing kinetics of E. coli 56809 at different concentrations as specified. The antibiotic imipenem did not exert significant bactericidal activity against A. baumannii and E. coli after 1 and 3 h treatment, respectively, while the polymers eradicated the bacteria rapidly. An increased polymer concentration led to faster killing efficiency. Error bars represent s.d. for n = 3

Fig. 4

Analyses of mechanisms of the antimicrobial polymers. Scanning electron microscopic (FE-SEM) (a, b) and transmission electron microscopic (TEM) (c, d) images of A. baumannii 10073 with and without polymer treatment. Treatment conditions: 4 × MIC, 2 h. Scale bar: SEM—5 µm in a and 1 µm in b; TEM—1 µm in c and 0.5 µm in d. e Ratio of absorbance at 260 nm of supernatant in A. baumannii 10073 suspension treated with pEt_20 or polymyxin B for 2 h at different concentrations as specified (bacterial density: 2.08 × 10⁹ CFU/mL). Ratios above 1.0 reflect release of cytoplasmic materials of bacteria. MBC for pEt_20 and polymyxin B: 32 and 256 µg/mL, respectively. f Drug resistance development profiles of A. baumannii 10073 after exposed to the polymers and the clinically used antibiotic imipenem at sub-MIC concentrations. At the effective dose (MBC), polymer treatment did not cause significant release of cytoplasmic materials while polymyxin B disrupted cell membrane, leading to significant leakage. This is consistent with SEM observation, where the majority of the cells did not see membrane lysis although a few cells had distorted membrane. Precipitation of cytoplasmic materials was seen in most cells with an intact membrane under TEM. Taken together, these observations suggested that antimicrobial mechanism is mainly based on membrane translocation followed by precipitation of cytoplasmic materials through interactions between the polymer and proteins/genes in the cytoplasm. This unique antibacterial mechanism prevented drug resistance development. e Error bars represent s.d. for n = 3. f The data are representative of three replicates

Fig. 5

RNA-seq analysis of bacterial resistance. MDR A. baumannii 10073 was treated with imipenem or pEt_20 for 30 passages at 0.5× MIC that was measured at each passage (treatment duration of each passage: 18 h). a A Venn diagram depicting the overlap of differentially regulated genes either upon pEt_20 or imipenem treatment, relative to untreated controls. The numbers inside the Venn diagram (from left to right) represent the number of differentially regulated genes: found only in pEt_20 treatment (left), common to both groups (venn diagram intersection), only in imipenem treatment (right), respectively. b The three independent heat maps depict genes from the corresponding groups in a (as indicated by the respective black arrows). The heat map on the left shows expression data for genes that are only significant in pEt_20 treatment but were shown in conjunction with the expression data from imipenem treatment (not differentially regulated, i.e., Log₂-fold change in the range between −1 and +1). The heat map in the center depicts significant genes that are shared by both treatment groups. The heat map on the right shows genes that are only significant in imipenem treatment along with Log₂-fold change from pEt_20 treatment (not differentially regulated but shown for comparison). The row color side bars on all three heat maps represent the KEGG categories that the genes represent

Fig. 6

In vivo antibacterial efficacy in various infectious mouse models. a ED₅₀ and ED₁₀₀ of polymers and the imipenem or vancomycin control (a: MDR A. baumannii 10073-caused peritonitis mouse model, intraperitoneal (i.p.) injection at 1 h and 6 h post infection; b: MDR A. baumannii 10073-caused peritonitis mouse model, i.p. injection at 3 h and 8 h post infection; c: MDR E. coli 56809-caused peritonitis mouse model, i.p. injection at 1 h and 6 h post infection; d: MDR K. pneumoniae 8637-caused peritonitis mouse model, i.p. injection at 1 h and 6 h post infection; e: MDR MRSA 25312-induced peritonitis mouse model, i.p. injection at 1 h and 6 h post infection). Colony forming units (CFUs) of A. baumannii 10073 (b) and E. coli 56809 (c) in blood, peritoneal cavity, spleen, liver, and kidney at 24 h post infection. Both polymers were effective against systemic MDR Gram-negative bacterial infections with lower ED₅₀/ED₉₅ values than imipenem (especially A. baumannii 10073- and K. pneumoniae 8637-caused infections), and they also worked against Gram-positive MRSA infection with comparable ED₅₀/ED₉₅ values as compared to vancomycin. The polymers removed bacteria in the blood, peritoneal cavity, and organs more effectively than imipenem especially in the A. baumannii 10073 infection. Means ± s.d., n = 5. One-way ANOVA (Tukey’s post hoc); *p < 0.05; **p < 0.01

Fig. 7

In vivo efficacy in CLP and P. aeruginosa lung infection models. a–c Cecal ligation and puncture (CLP) model; d–f P. aeruginosa (PA14)-caused lung infection model. a Survival of mice in the sham group that went through surgery but without CLP, in the CLP group without treatment, in the gentamicin-treated CLP group (Gen, 10 mg/kg mouse body weight at 1 h and 6 h post infection), and in the pEt_20-treated group (pEt_20, 25 mg/kg mouse body weight at 1 h and 6 h post infection; each dose lower than LD₅ value) (n = 6, p = 0.0112 (<0.05), Log-rank test). b Microbial counts (CFUs) in the blood and peritoneal fluid (c) at 24 h after CLP. WT: control group without surgery or treatment. Means ± s.d., n = 3. d Survival of mice with P. aeruginosa (PA14)-caused lung infection without and with pEt_10 or imipenem treatment (8 mg/kg mouse body weight at 1, 6, and 25 h post infection) (n = 6, p = 0.0214 (<0.05), Log-rank test). e Bacterial counts (CFUs) in the blood and the lung tissues (f). Means ± s.d., n = 3. One-way ANOVA (Tukey’s post hoc); *p < 0.05; **p < 0.01. Control group: without infection or treatment