Arginine regulates the mucoid phenotype of hypervirulent Klebsiella pneumoniae

Abstract

Hypervirulent Klebsiella pneumoniae causes severe community-acquired infections, with its mucoid phenotype resulting from altered capsular polysaccharide chain length. While both environmental and genetic factors influence mucoidy, the cues regulating it remain unclear. Here, we show that casamino acids enhance mucoidy without affecting total capsular polysaccharide levels. We show that arginine is both necessary and sufficient in stimulating mucoid expression, activating the rmpADC promoter and increasing rmpADC transcript levels. The arginine regulator, ArgR, is crucial in this process; deleting argR reduces mucoidy and increases capsule chain length diversity. ArgR directly regulates the rmpADC promoter by binding to the ARG box. Loss of argR in vitro increases macrophage association and reduces competitive fitness in lungs, suggesting that ArgR influences adherence and fitness in the lung. Arginine-dependent regulation of mucoidy is conserved in hypervirulent K. pneumoniae isolates, suggesting that this regulatory mechanism broadly controls bacterial cell surface adaptations.

Fig. 1. Specific nutrients dissociate capsular polysaccharide (CPS) from mucoidy.

Wild-type K. pneumoniae strain KPPR1 was cultured in low-iron M9 minimal medium with 20 mM glucose (Glc), 20 mM glycerol (Glyc), or 20 mM fucose (Fuc), with or without 1% casamino acids (CAA), as nutrient sources. A Mucoidy was determined by quantifying the supernatant OD₆₀₀ after sedimenting 0.5 OD₆₀₀ unit of culture at 1000 x g for 5 min. B Uronic acid content of crude CPS extracts were quantified and normalized to the OD₆₀₀ of the overnight culture. Data presented are the mean, and error bars represent the standard error of the mean. Statistical significance was determined using a one-way ANOVA with a Tukey post-test. p-values are displayed above each comparison, # p < 0.0001. The statistics displayed above the data bars represent values relative to M9 + CAA. Statistics connected by lines identify other direct comparisons. Exact P-values: (A) left to right above data bars: <0.0001, <0.0001, <0.0001, <0.0001, <0.0001; connected by lines: <0.0001, <0.0001, <0.0001; (B) left to right: <0.0001. Experiments were performed 3 independent times, each in biological triplicate (n = 9). Source data are provided as a Source Data file.

Fig. 2. Arginine and phenylalanine are necessary for K. pneumoniae to regulate mucoidy.

Wild type K. pneumoniae strain, KPPR1, was cultured in (A, B) low-iron M9 minimal medium with 20 mM sodium pyruvate and all 18 amino acids (M9 + All), or with an individual amino acid absent. In C, D KPPR1 was cultured in low-iron minimal M9+All or M9 medium without arginine, or phenylalanine, or with both absent. A, C Mucoidy was determined by quantifying the supernatant OD₆₀₀ after sedimenting 0.5 OD₆₀₀ unit of culture at 1000 x g for 5 min. Uronic acid content of crude CPS extracts were quantified and (B) normalized to the OD₆₀₀ of the overnight culture or (D) normalized to 10⁹ CFUs. Data presented are the mean, and error bars represent the standard error of the mean. Statistical significance was determined using a one-way ANOVA and Dunnett post-test compared to M9 + All. p-values are displayed above each comparison; * p < 0.05; ** p < 0.01; *** p < 0.001; # p < 0.0001. Exact P-values: (A) left to right: <0.0001, 0.0350, 0.0002, 0.0131, 0.0008; (B) left to right: <0.0001, <0.0001, 0.0077, <0.0001, 0.0001, <0.0001, <0.0001, 0.0013; (C) left to right: 0.0007, 0.0001, <0.0001; (D) left to right: 0.0470. Experiments were performed ≥ 3 independent times (for graphs A-C, n = 9 and for graph d, n = 9 for M9 + All and M9-Arg conditions, but n = 12 for M9-Phe and M9-Arg-Phe conditions), each in biological triplicate. Source data are provided as a Source Data file.

Fig. 3. Arginine is sufficient to induce mucoidy in K. pneumoniae and requires the ArgR regulator.

A, B Wild-type (WT) KPPR1 was cultured overnight in low-iron M9 medium supplemented with 18 amino acids (All), 20 mM glycerol (Glyc), or increasing concentrations of arginine (0.025%–0.2%) and 20 mM Glyc. C, D Additionally, KPPR1, astA::kan, argG::kan, and ∆argR mutants were cultured overnight in low-iron M9 minimal media under the following conditions: with 18 amino acids (All), without arginine (-Arg), with 20 mM glycerol (Glyc), or with 20 mM glycerol and 0.2% arginine (Glyc+Arg). (E, F) The strains KPPR1 and ∆argR were transformed with either the empty vector (EV) or argR vector (pargR) and cultured overnight in low-iron minimal M9+All medium. A−E Mucoidy was determined for each of the conditions by sedimenting 0.5 OD₆₀₀ unit of culture at 1000 x g for 5 min and then measuring the supernatant OD₆₀₀. B−F The total capsule abundance was determined by measuring the uronic acid content of crude CPS extracts and normalized to 10⁹ CFUs. Data presented are the mean, and error bars represent the standard error of the mean. In panels (C, D) the cross (†) indicates no growth. A−F Statistical significance was determined using one-way ANOVA with a Dunnett post-test. C, D Statistical significance was determined using two-way ANOVA with a Tukey post-test. The statistics indicated above the data bars represent values relative to M9 + All (A, B) or WT in the same culture condition (C−F). p-values are displayed above each comparison; * p < 0.05; # p < 0.0001. Exact P-values: (A) left to right: <0.0001, <0.0001, <0.0001, 0.0135; (C) left to right: <0.0001, <0.0001, <0.0001, <0.0001; (E) left to right: <0.0001. Experiments were performed 3 independent times, each in biological triplicate (for graphs A,C-F, n = 9 and for graph B, n = 9 for M9 + All, M9+Glyc, and 0.2% arginine conditions, but n = 12 for 0.025%-0.1% arginine conditions). Source data are provided as a Source Data file.

Fig. 4. Arginine regulates mucoidy by increasing rmpADC transcription in an ArgR-dependent manner.

KPPR1 and the ∆argR mutant were cultured under different amino acid conditions (M9 + All, M9-Arg, M9 + Glyc, M9 + Glyc + Arg). A Rmp promoter (P_rmp) activity was measured by GFP expression via a promoter fusion. RNA was isolated from mid-log cultures of KPPR1 in the four media or from ∆argR and the complementation vectors (KPPR1 + EV, argR + EV, or argR+pargR) in M9 + All. B−E The relative abundance of rmpA, rmpD, and rmpC RNA under the different conditions was determined by qRT-PCR and normalized to gap2 transcript abundance. B, C Relative transcript levels in KPPR1 cultured in M9-Arg or M9+Glyc+Arg were compared to M9 + All or M9 + Glyc, respectively, while (D) the ∆argR mutant (cultured in M9 + All) transcript levels were compared to wildtype, KPPR1. (E) The complementation transcript levels were compared to KPPR1 + EV. Data presented are the mean, and error bars represent the standard error of the mean. Statistical significance was determined in panel (A) with a two-way ANOVA with Tukey’s post-test. The statistics displayed above the data bars represent values relative to M9 + All. Statistics connected by lines identify other direct comparisons. In panels (B–E) significance was determined using a two-tailed unpaired t-test to determine. p-values are displayed above each comparison; *p < 0.05; **p < 0.01; ***p < 0.001; # p < 0.0001. Exact P-values: (A) left to right above data bars: <0.0001, <0.0001; connected by lines: <0.0001, <0.0001, 0.0005, <0.0001; (B) left to right: 0.0002, 0.0002, 0.0199; (C) left to right: 0.0007, 0.0012, 0.0018; (D) left to right: 0.0127, 0.0060, 0.0008; (E) left to right: 0.0107, 0.0398, 0.0009, <0.0001. Experiments were performed ≥ 3 independent times (for graph A, n = 9 with biological replicates and for graph B-E, n = 3 with technical replicates). Source data are provided as a Source Data file.

Fig. 5. Arginine and ArgR regulate mucoidy by increasing rmpD transcription and decreasing capsular polysaccharide chain length diversity.

KPPR1 and the ∆argR mutant were cultured in different amino acid conditions (M9 + All, M9-Arg, M9 + Glyc, M9 + Glyc + Arg) or (M9 + All), respectively. A, C, E Capsular polysaccharides (CPS) were separated on a 4-15% SDS-PAGE gel and stained with Alcian blue and silver stain. Three distinct polysaccharide types emerged: diverse mid- to high-molecular-weight chains (Type A), uniform high-molecular-weight chains (Type B), and diffuse ultra-high-molecular-weight chains (Type C). Data presented are the mean, and error bars represent the standard error of the mean. Representative images of the three independent gels are shown in panels (A, C and E) with the quantification of three independent replicates being shown in panels (B, D, and F) ImageJ was used to quantify each CPS chain type with background subtracted. To minimize gel-to-gel variability, the control and experimental density values were extracted from the same gel and used to calculate fold-change abundance. Changes in CPS type production were determined by calculating the fold-change in experimental density values relative to control density values on three independent gels. Statistical significance was determined using a two-tailed one sample t-test. p-values are displayed above each comparison; * p < 0.05; *** p < 0.001. Exact P-values: (B) left to right: 0.0474; (D) left to right: 0.0010, 0.0289; (F) left to right: 0.0488, 0.0161, 0.0107. Experiments were performed ≥ 3 independent times (n = 3). Source data are provided as a Source Data file.

Fig. 6. ArgR is required for K. pneumoniae to evade immortalized bone marrow-derived macrophages (iBMDMs) association and maintain competitive fitness in the lung.

A Macrophages (BEI NR-9456) were incubated with KPPR1 or ∆argR for 2 hr (MOI = 10). Internalization was determined by a 1 hr gentamicin treatment. Macrophages were washed with PBS, lysed with 0.2% TritonX100, and output CFUs were enumerated and normalized to input. Experiments were performed > 3 independent times, in triplicate (n = 9). Data presented are the mean, and error bars represent the standard error of the mean. B C57BL/6 mice were infected retropharyngeally with 1 × 10⁶ CFUs of a 1:1 mixture of WT + EV and either ∆argR + EV or ∆argR + pargR. The log₁₀ competitive index at 24 h post-infection is shown relative to 0. Each dot represents an individual mouse (open symbols = male; half-filled = female). The horizontal line represents the median; box bounds indicate the minima (25^TH percentile) and maxima (75^TH percentile); and whiskers represent the full range. ∆argR + EV was tested in two independent experiments (n = 8), while ∆argR+pargR was tested once (n = 7). C Arginine concentrations were measured in murine lungs, spleens, livers, and feces. Samples were collected from three mice and experiments were performed 3 independent times, in technical triplicate (n = 3). Data presented are the mean, and error bars represent the standard error of the mean. Statistical significance was determined using an unpaired two-tailed t-test (A), a two-tailed unpaired t-test (B, comparing CI values between ∆argR + EV and ∆argR+pargR), and a two-tailed one-sample t-test (B, comparing CI to a hypothetical value of zero). In (C), a one-way ANOVA with Tukey’s post-test was used. p-values are displayed above each comparison; *p < 0.05; **p < 0.01, ***p < 0.001; #p < 0.0001. Exact P-values: (A) left to right: <0.0001, 0.0287; (B) left to right above data bars: 0.0008, 0.0078, 0.0081; connected by lines: <0.0001; (C) left to right: 0.0033, <0.0001, 0.0327, 0.0003. Source data are provided as a Source Data file.

Fig. 7. A highly conserved ARG box in the rmp promoter is required for ArgR-dependent mucoidy regulation.

A K. pneumoniae genomes with intact rmpADC loci (n = 1572) were analyzed for ARG box sequence diversity using Geneious v2023.1.1 and WebLogo v3.7.9. Diversity is shown as percentage and minor allele frequency. Asterisks indicate highly conserved nucleotides. B KPPR1 carrying either a P_rmp-GFP reporter plasmid or a mutated ARG box (P_rmp∆ARG-GFP) was cultured in M9 + All, M9-Arg, M9 + Glyc, or M9 + Glyc + Arg. GFP expression measured promoter activity. C DNA-protein interaction ELISA assessed ArgR binding to the ARG box in P_rmp. Purified His₆-tagged ArgR was incubated in wells coated with DNA oligomers containing either the predicted (P_rmp) or mutated binding site (P_rmp∆ARG). Binding was detected using an HRP-conjugated anti-poly-histidine antibody. OD₄₅₀ values reflect ArgR-DNA binding. D, E KPPR1S or chromosomal ARG box mutant (∆ARGbox) were cultured in the four media conditions. D Mucoidy was assessed by sedimentation assay. E Capsule abundance was measured by uronic acid quantification. F−H KPPR1S and ∆ARGbox mutant were cultured in M9+All. F Transcript levels of rmpA, rmpD, and rmpC were measured by qRT-PCR and normalized to gap2. G Capsular polysaccharides (CPS) were separated by SDS-PAGE and stained with Alcian blue and silver stain; representative image of three independent gels shown. H Quantification was done with ImageJ with background subtraction. Data presented are the mean, and error bars represent SEM. Statistical tests: unpaired two-tailed t-test (B, F); two-tailed Mann-Whitney test (C); two-way ANOVA with Bonferroni’s post-test (D, E); two-tailed one sample t-test (H). p-values: **p < 0.01; #p < 0.0001. Exact P-values left to right: (B): <0.0001, <0.0001, <0.0001, <0.0001; (C): 0.0087, 0.0022, 0.0022, 0.0022, 0.0022, 0.0022, 0.0022; (D): <0.0001, <0.0001; (E): 0.0025, <0.0001, 0.0085, 0.0061; (F): 0.0098, 0.0007, <0.0001; (H): 0.0188. Experiments were performed > 3 independent times (n = 9), in triplicate, except (G) (n = 3) and (C) (duplicates, n = 6). Source data are provided as a Source Data file.

Fig. 8. Multiple hvKp strains regulate mucoidy through the ARG box in response to arginine and host conditions.

A Hypervirulent K. pneumoniae (hvKp) strains NTUH-K2044, Kp4289, Kp4585, and Kp6557 carrying either the P_rmp-GFP reporter plasmid or a mutated ARG box (P_rmp∆ARG-GFP) were cultured in M9+All and M9-Arg. GFP expression measured promoter activity. B NTUH-K2044, Kp4289, Kp4585, and Kp6557 were cultured in M9+All or M9-Arg. Mucoidy was assessed by sedimentation assay. C 20% pooled human heat-inactivated serum was supplemented into 10x M9 salts (M9 + HI Serum). Both M9 + HI Serum and human urine samples were centrifuged to remove debris and proteinase-treated prior to arginine quantification. D KPPR1 carrying either the P_rmp-GFP reporter plasmid or P_rmp∆ARG-GFP was cultured in either M9 + HI Serum or sterile-filtered pooled human urine. Promoter activity was measured by GFP expression. E KPPR1 and ∆argR was cultured in M9 + HI Serum or sterile-filtered pooled human urine and mucoidy was assessed via a sedimentation assay. Data presented are the mean, and error bars represent the standard error of the mean. Statistical significance was determined in panel (A) using a two-way ANOVA with Tukey’s post-test, while in panel (B) a two-tailed t-test was used. In panel (C) a two-tailed Mann-Whitney test was performed and in panels (D and E) a one-way ANOVA with Bonferroni’s post-test was applied. p-values are displayed above each comparison: *p < 0.05; **p < 0.01; ***p < 0.001; #p < 0.0001. Exact P-values: (A), left to right above data bars: <0.0001, <0.0001, 0029; connected by lines: <0.0001, 0.0010, <0.0001, 0.0003, <0.0001, 0.0067, <0.0001, <0.0001; (B), left to right: <0.0001, 0.0019, <0.0001, <0.0001; (C), left to right: <0.0001; (D), left to right; <0.0001, 0.0138; (E), left to right: <0.0001, <0.0001. Experiments were performed > 3 independent times, each in biological triplicate (n = 9). Source data are provided as a Source Data file.

Fig. 9. Model of K. pneumoniae regulation of mucoidy in response to arginine.

When cultured in the presence of arginine, K. pneumoniae increases mucoidy without altering the total capsule abundance. When arginine is brought into the cell, the arginine regulator, ArgR, binds arginine and becomes active. The activated ArgR-arginine complex acts as a transcriptional regulator via binding to ARG boxes. The arginine-ArgR complex binds directly to the P_rmp ARG box and activates the rmpADC promoter, upregulating rmpADC transcription. The RmpD protein then interacts with the Wzc protein, a tyrosine autokinase that regulates high-level capsular polysaccharide (CPS) polymerization. RmpD-Wzc interactions decrease ‘Type A’ polysaccharide chains and increases ‘Type B’, which decreases CPS diversity and presents as the mucoid phenotype resulting in decreased association with macrophages. Created in BioRender. Ryan, B. (2025) https://BioRender.com/r74w727.