Klebsiella pneumoniae peptide hijacks a Streptococcus pneumoniae permease to subvert pneumococcal growth and colonization

Abstract

Treatment of pneumococcal infections is limited by antibiotic resistance and exacerbation of disease by bacterial lysis releasing pneumolysin toxin and other inflammatory factors. We identified a previously uncharacterized peptide in the Klebsiella pneumoniae secretome, which enters Streptococcus pneumoniae via its AmiA-AliA/AliB permease. Subsequent downregulation of genes for amino acid biosynthesis and peptide uptake was associated with reduction of pneumococcal growth in defined medium and human cerebrospinal fluid, irregular cell shape, decreased chain length and decreased genetic transformation. The bacteriostatic effect was specific to S. pneumoniae and Streptococcus pseudopneumoniae with no effect on Streptococcus mitis, Haemophilus influenzae, Staphylococcus aureus or K. pneumoniae. Peptide sequence and length were crucial to growth suppression. The peptide reduced pneumococcal adherence to primary human airway epithelial cell cultures and colonization of rat nasopharynx, without toxicity. We identified a peptide with potential as a therapeutic for pneumococcal diseases suppressing growth of multiple clinical isolates, including antibiotic resistant strains, while avoiding bacterial lysis and dysbiosis.

Fig. 1. Pneumococcal growth in the presence and absence of peptides.

a Growth curves of pneumococcal strain D39 in the presence and absence of 11 amino acid peptide V11A (VNATDEDRWNA) and its longer versions (13- and 15-amino acid length) in chemically defined medium (CDM). b Growth curves of pneumococcal strain D39 in the presence and absence of peptide V11A and peptides with naturally occurring amino acid differences compared to V11A in CDM. Difference at position 3 occurs in strains of K. pneumoniae and Salmonella enterica, at position 4 in strains of E. coli, at positions 3 and 4 in strains of Arsenophonus, E. coli, Moritella, Pasteurellales, and Psychromonas, at position 7 in strains of K. pneumoniae and S. enterica, at position 10 in some strains of Cedecea, Psychromonas and S. enterica. Amino acid differences from V11A are underlined. All peptides at a concentration of 0.5 mg/ml. c Growth curves of trimethoprim-sulfamethoxazole resistant clinical pneumococal isolate 1154.75 (serotype 23F) in the presence and absence of peptide V11A and K. pneumoniae ribosomal peptides AmiA (AKTIKITQTR) or AliA (FNEMQPIVDRQ) ligand in CDM. d Growth curves of pneumococcal strain 1154.75 in the presence and absence of peptide V11A in human cerebrospinal fluid (hCSF) of donor 1 (see Supplementary Fig. 2 for equivalent results in hCSF from four other donors). Bacterial inoculum was twice that used in CDM due to anticipated slower growth in hCSF. e Time-kill assay of S. pneumoniae strain D39. Bacteria were untreated (negative control) or exposed to 0.5 mg/ml peptide V11A or 0.5 mg/ml penicillin G (positive control for bacteriolytic effect). Increasing cfu/ml represents bacterial growth over time, steady cfu/ml represents bacteriostatic effect, and decreasing cfu/ml represents bacteriolytic effect. Growth was measured by optical density OD (450 nm) measurement in peptide-free CDM for (a–c), undiluted human cerebrospinal fluid for (d) or by plating dilutions of bacterial sample onto agar plates at the timepoints indicated and counting colony forming units (cfu) for (e). Results represent three independent experiments, error bars indicate SEM. Peptide concentration = 0.5 mg/ml unless otherwise indicated.

Fig. 2. Peptide V11A uptake via Ami permease.

a Growth curve of S. pneumoniae D39 parent strain and ∆amiC mutant. Growth was measured in peptide-free chemically defined medium (CDM) in the presence and absence of 0.5 mg/ml peptide V11A by measuring optical density OD (450 nm) over time. Results represent three independent experiments, error bars indicate SEM. b, c Microscopy images of S. pneumoniae after incubation with FITC-labelled peptide V11A for D39 parent strain (b) and ∆amiC mutant (c). Representative images were taken at mid-bacterium localization in the z position showing FITC and Brightfield (BF) channels. Scale bar indicates 2 µm for all pictures in (b) and (c).

Fig. 3. Effects of V11A on pneumococcal morphology, chain length and transformation.

a Microscopy images of pneumococci to assess pneumococcal cell morphology. Representative images of S. pneumoniae strains D39, 106.66, 208.41 after incubation with or without 0.5 mg/ml peptide V11A for 2.5 h taken at 100x magnification. Scale bar indicates 2 µm for all images. b Chain length of S. pneumoniae strains D39, 106.66, 208.41. Results represent chain length measurements of three independent experiments after incubation with or without 0.5 mg/ml peptide V11A for 2.5 h, each box represents length measurements of >1200 bacterial chains, *** indicates p value < 0.001 by two-sided Wilcoxon rank sum test for all comparisons. c Transformation rate of S. pneumoniae strain D39 without or with 0.5 mg/ml peptide V11A or control peptide (V11A with amino acid difference at position 4). Results represent at least five independent experiments for each condition represented by grey dots, **** indicates p value  ≤ 0.0001 by t-test.

Fig. 4. Effect of peptide V11A on pneumococcal transcriptome.

Genetic interaction network of significantly downregulated (log2FC ≤ −1.25) and upregulated (log2FC ≥ 1.25) genes in S. pneumoniae strain D39 after 15 min of incubation with 0.5 mg/ml peptide V11A. Interactions between 2 genes (nodes) are depicted as a line (edge) and were imported from the STRING database. Nodes are colour-coded according to the log2FC, with blue representing negative log2FC (downregulated) and red positive log2FC (upregulated). We highlighted Gene Ontologies or keywords of groups of genes in the background.

Fig. 5. Effect of peptide V11A on adherence of S. pneumoniae in vitro and colonization in vivo.

a Pneumococcal adherence to primary human airway epithelial cell (hAEC) cultures. S. pneumoniae strain D39 was added to the apical side of differentiated hAEC cultures without or with 0.5 mg/ml peptide V11A or control peptide (V11A with an amino acid difference at position 10). Results represent 5 biological replicates using hAEC cultures of 3 donors on one and 2 donors on another day for each condition, respectively. *indicates p value = 0.011 by pairwise one-sided Wilcoxon rank sum test. b Pneumococcal adherence to rat nasopharynx. Trimethoprim-sulfamethoxazole-resistant clinical pneumococcal isolate 1154.75 (serotype 23F) was administered intranasally to rat pups without or with 0.5 mg/ml peptide V11A. Results represent a total of 11 individual pups per condition (from 2 different days with 6 and 5 pups, respectively). * indicates p value = 0.01367 by pairwise one-sided Wilcoxon rank sum test. Adhered bacteria to hAEC cultures or rat nasopharynx were quantified 24 h post-inoculation by dilution plating and counting colony-forming units (cfu). Data points are represented by grey dots, error bars indicate SEM.