Regrowth-delay body as a bacterial subcellular structure marking multidrug-tolerant persisters

Abstract

Bacteria have long been recognized to be capable of entering a phenotypically non-growing persister state, in which the cells exhibit an extended regrowth lag and a multidrug tolerance, thus posing a great challenge in treating infectious diseases. Owing to their non-inheritability, low abundance of existence, lack of metabolic activities, and high heterogeneity, properties of persisters remain poorly understood. Here, we report our accidental discovery of a subcellular structure that we term the regrowth-delay body, which is formed only in non-growing bacterial cells and sequesters multiple key proteins. This structure, that dissolves when the cell resumes growth, is able to be viewed as a marker of persisters. Our studies also indicate that persisters exhibit different depth of persistence, as determined by the status of their regrowth-delay bodies. Our findings imply that suppressing the formation and/or promoting the dissolution of regrowth-delay bodies could be viable strategies for eradicating persisters.

Fig. 1. The cell division protein FtsZ in the late stationary-phase E. coli cells exists in cell-pole granule likely as a folded form.

a Immunoblotting results for detecting endogenous FtsZ, EF-Tu, or GroEL in the total cell lysate (total), supernatant (sup.) and pellet (pel.) of the log-phase or late stationary-phase wild-type E. coli cells, probed with the indicated antibodies. b Fluorescence and bright field microscopic images of the log-phase (top) and late stationary-phase (bottom) E. coli cells in which FtsZ-mNeonGreen was heterologously expressed. Scale bars, 1 μm. c Fluorescence microscopic images of the log-phase and late stationary-phase ftsZ-mNeonGreen-dnaK-mCherry or ftsZ-mNeonGreen-clpB-mCherry cells. Scale bars, 1 μm. d Fluorescence microscopic images of the late stationary-phase ftsZ-mNeonGreen cells in which the FtsZ inhibitor protein CbtA was expressed (left panel) Scale bars, 1 μm; the corresponding immunoblotting results for detecting FtsZ in the indicated cell lysate fractions, probed with anti-FtsZ antibodies (right panel)

Fig. 2. When the non-growing cells exit their regrowth lag, the cell-pole granules dissolve to release the FtsZ for re-functioning, but maintain unaltered otherwise.

a Fluorescence microscopic images of the re-cultured late stationary-phase ftsZ-mNeonGreen cells present in fresh LB medium lacking rhamnose, as obtained at the indicated time points. Note: one of the examined cells divided into two daughter cells at 120 min (circled by pink dashed lines). Scale bars, 1 μm. b Fluorescence microscopic images of the late stationary-phase ftsZ-mNeonGreen cells re-cultured to the log phase (OD₆₀₀ ~0.5) in liquid LB medium lacking rhamnose. Scale bar, 1 μm. c Fluorescence microscopic images of the late stationary-phase ftsZ-mNeonGreen cells re-cultured to the indicated time points in fresh LB medium that lacked rhamnose and contained the antibiotic chloramphenicol. Scale bars, 1 μm

Fig. 3. The regrowth-delay bodies are formed heterogeneously in the cell population and the degree of their formation correlates with the duration of regrowth lag and the level of multidrug tolerance.

a Fluorescence microscopic images of ftsZ-mNeonGreen cells cultured to the indicated time points in LB medium containing 0.02% rhamnose (to induce the production of FtsZ-mNeonGreen). Scale bars, 1 μm. b Percentage of cells containing regrowth-delay bodies that were cultured to the indicated time points. c Percentages of cells retaining their regrowth-delay bodies when the particular stationary-phase cell samples were re-cultured for 30 min in fresh medium containing chloramphenicol (red columns). Percentages of cells containing regrowth-delay bodies for more than 3 h (pink columns) were calculated from Fig. 3b. d Re-division T_id (the average initial doubling time) values of wild-type cells that were pre-cultured to the indicated time points. The T_id values were calculated based on the increase in cell numbers within the first 30 min of re-culturing (after diluting 40-fold) in fresh medium at 37 °C (for details, see Methods). e Survival rates of the indicated re-cultured stationary-phase wild-type cells that were treated with ofloxacin (5 μg/ml) or ampicillin (200 μg/ml) for 2 h (in fresh LB medium at 37 °C). The survival rates were calculated according to the equation: [colony-forming units (CFU) of the antibiotic-treated cells]/[colony-forming units of the untreated cells] ×100. f Live-cell fluorescence microscopic images merged with bright field of the re-cultured late stationary-phase ftsZ-mNeonGreen cells in the fresh ampicillin-containing LB medium (at 37 °C), as obtained at the indicated time points. Representative cells that exited (eventually became lysed) or maintained (unaltered) the regrowth lag is indicated by the white or red arrows, respectively. Scale bars, 3 μm. The symbol * in d and e denotes a significant difference between the compared pair of samples (P-value < 0.05, t-test). At least three biological replicates were analyzed in obtaining each value

Fig. 4. Regrowth-delay bodies selectively sequester multiple key proteins that are released to re-function when cells exit their regrowth lag and resume growth.

a Fluorescence microscopic images of the log-phase (left) and late stationary-phase (right) E. coli cells in which mNeonGreen-fused FtsA or ZapC (both being identified in the regrowth-delay bodies by mass spectrometry analysis, as shown in Supplementary Fig. S9a) was expressed from a plasmid under the control of a constitutive promoter. Scale bars, 1 μm. b Blotting results to analyze the indicated Avi-tagged proteins in the indicated lysate fractions of the log-phase or late stationary-phase wild-type cells, probed with streptavidin-AP. c Fluorescence microscopic images of the re-cultured late stationary-phase cells in which FtsA-mNeonGreen was expressed from a plasmid under the control of a constitutive promoter in fresh medium containing chloramphenicol, obtained at the indicated time points. Scale bars, 1 μm

Fig. 5. Mutant bacterial cells with a reduced formation of regrowth-delay bodies exhibit a shorter duration of regrowth lag as well as a lower tolerance to antibiotics.

a Fluorescence microscopic images of the log-phase or late stationary-phase ftsZ-mNeonGreen cells having a knockdown of either the nuoA or the sdhC gene. Cells expressing a non-targeting CRISPR RNA were analyzed as the control. Scale bars, 1 μm (left panel). The immunoblotting results for detecting FtsZ in the indicated cell lysate fractions, as probed with anti-FtsZ antibodies (right panel). b Re-division T_id values of early (blue bars; cultured to 12 h) or late (red bars; cultured to 24 h) stationary-phase cells of the indicated gene-knockdown strain. Here wild-type cells in which a non-targeting crRNA was expressed from a plasmid were analyzed as the control. c Survival rates of the late stationary-phase wild-type (control), nuoA-knockdown or sdhC-knockdown cells that were re-cultured in fresh medium after being treated with ofloxacin (5 μg/ml) or ampicillin (200 μg/ml). The survival rates were calculated according to the equation: (CFU of the antibiotic-treated cells)/(CFU of the untreated cells) ×100. The symbol * in b and c denotes a significant difference between the compared pair of samples (P-value < 0.05, t-test). At least three biological replicates were analyzed for obtaining each value

Fig. 6. Regrowth-delay bodies are also formed in the late stationary-phase cells of the pathogenic bacteria Salmonella Typhimurium SL1344 and Shigella flexneri serotype 2a 2457T.

a Immunoblotting results for the detection of FtsZ in the indicated cell lysate fractions of the stationary-phase Salmonella Typhimurium or Shigella flexneri cells taken at the indicated time points, probed with antibodies against the E. coli FtsZ protein. b Re-division T_id values of the Salmonella Typhimurium or Shigella flexneri cells that were pre-cultured to the indicated time points of the stationary-phase before being re-cultured in fresh LB medium. c Survival rates of the indicated re-cultured stationary-phase Salmonella Typhimurium or Shigella flexneri cells that were treated with the indicated antibiotics for 2 h. The symbol * in b and c denotes a significant difference between the compared pair of samples (P-value < 0.05, t-test). At least three biological replicates were analyzed for obtaining each value