Nε-lysine acetylation of the histone-like protein HBsu influences antibiotic survival and persistence in Bacillus subtilis

Abstract

N^ε-lysine acetylation is recognized as a prevalent post-translational modification (PTM) that regulates proteins across all three domains of life. In Bacillus subtilis, the histone-like protein HBsu is acetylated at seven sites, which regulates DNA compaction and the process of sporulation. In Mycobacteria, DNA compaction is a survival strategy in response antibiotic exposure. Acetylation of the HBsu ortholog HupB decondenses the chromosome to escape this drug-induced, non-growing state, and in addition, regulates the formation of drug-tolerant subpopulations by altering gene expression. We hypothesized that the acetylation of HBsu plays similar regulatory roles. First, we measured nucleoid area by fluorescence microscopy and in agreement, we found that wild-type cells compacted their nucleoids upon kanamycin exposure, but not exposure to tetracycline. We analyzed a collection of HBsu mutants that contain lysine substitutions that mimic the acetylated (glutamine) or unacetylated (arginine) forms of the protein. Our findings indicate that some level of acetylation is required at K3 for a proper response and K75 must be deacetylated. Next, we performed time-kill assays of wild-type and mutant strains in the presence of different antibiotics and found that interfering with HBsu acetylation led to faster killing rates. Finally, we examined the persistent subpopulation and found that altering the acetylation status of HBsu led to an increase in persister cell formation. In addition, we found that most of the deacetylation-mimic mutants, which have compacted nucleoids, were delayed in resuming growth following removal of the antibiotic, suggesting that acetylation is required to escape the persistent state. Together, this data adds an additional regulatory role for HBsu acetylation and further supports the existence of a histone-like code in bacteria.

Keywords: acetyl; acetylation; antibiotic resistance; bacteria; nucleoid-associated protein; persisters; tolerance.

Figure 1

Wild-type cells compact their nucleoids in response to kanamycin challenge. (A) Wild-type cells (BD630) were pre-grown in LB media for 2 h, then incubated with or without 5 μg/mL kanamycin for 20 min, and nucleoids stained with DAPI. Representative microscopy images are displayed. DIC, Differential interference contrast. (B) Cumulative distribution plots are displayed, where the 50th percentile represents the median of the population distribution. Nucleoid areas of at least 800 cells were analyzed. The distributions with and without kanamycin were significantly different (p value = 2.54e⁻⁷), as determined by the Kolmogorov–Smirnov test.

Figure 2

The impact of HBsu acetylation on nucleoid compaction following kanamycin exposure. hbsK3Q (BD8577, A), hbsK18Q (BD8219, B), hbsK37Q (BD8119, C), hbsK41Q (BD8147, D), hbsK75Q (BD8398, E), hbsK80Q (BD8576, F), and hbsK86Q (BD7493, G) cells were pre-grown in LB media for 2 h, then incubated with or without 5 μg/mL kanamycin for 20 min, and nucleoids stained with DAPI. Cumulative distribution plots are displayed, where the 50th percentile represents the median of the population distribution. Nucleoid areas of at least 800 cells were analyzed. The population distributions of nucleoid area in all strains, except for hbsK3Q (p value = 0.0682), were significantly different ± kanamycin, with the hbsK75Q p value = 0.005 and for all others, p values <2.2e⁻¹⁶.

Figure 3

The impact of HBsu deacetylation on nucleoid compaction following kanamycin exposure. hbsK3R (BD8387, A), hbsK18R (BD8190, B), hbsK37R (BD8120, C), hbsK41R (BD8148, D), hbsK75R (BD8333, E), hbsK80R (BD7484, F), and hbsK86R (BD7506, G) cells were pre-grown in LB media for 2 h, then incubated with or without 5 μg/mL kanamycin for 20 min, and nucleoids stained with DAPI. Cumulative distribution plots are displayed, where the 50th percentile represents the median of the population distribution. The population distributions of nucleoid area in all strains were significantly different ± kanamycin: hbsK3R (p value = 0.005), hbsK18R (p value = 3.16e⁻¹²), hbsK37R (p value < 2.2e⁻¹⁶), hbsK41R (p value < 1.1e⁻¹⁵), hbsK75R (p value < 2.2e⁻¹⁶), hbsK80R (p value < 2.2e⁻¹⁶), and hbsK86R (p value < 2.2e⁻¹⁶).

Figure 4

Interfering with HBsu acetylation leads to faster killing in the presence of kanamycin. Cells were pre-grown in LB media, then incubated with or without 5 μg/mL kanamycin for 2 h, with viable counts determined every 30 min. Percent survival at each timepoint was determined as CFUs/mL following kanamycin treatment/initial colony counts (CFUs/mL). (A) Wildtype (BD630), hbsK3Q (BD8577), hbsK18Q (BD8219), hbsK37Q (BD8119), hbsK41Q (BD8147), hbsK75Q (BD8398), hbsK80Q (BD8576), and hbsK86Q (BD7493). (B) Wildtype (BD630), hbsK3R (BD8387), hbsK18R (BD8190), hbsK37R (BD8120), hbsK41R (BD8148), hbsK75R (BD8333), hbsK80R (BD7484), and hbsK86R (BD7506).

Figure 5

Interfering with HBsu acetylation leads to faster killing in the presence of vancomycin. Cells were pre-grown in LB media, then incubated with or without 12.5 μg/mL vancomycin for 2 h, with viable counts determined every 30 min. Percent survival at each timepoint was determined as CFUs/mL following vancomycin treatment/initial colony counts (CFUs/mL). Wildtype (BD630), hbsK3Q (BD8577), hbsK41Q (BD8147), hbsK3R (BD8387), and hbsK41R (BD8148).

Figure 6

Interfering with HBsu acetylation alters survival in the presence of tetracycline. Cells were pre-grown in LB media, then incubated with or without 25 μg/mL tetracycline for 2 h, with viable counts determined every 30 min. Percent survival at each timepoint was determined as CFUs/mL following tetracycline treatment/initial colony counts (CFUs/mL). (A) Wildtype (BD630), hbsK3Q (BD8577), hbsK18Q (BD8219), hbsK37Q (BD8119), hbsK41Q (BD8147), hbsK75Q (BD8398), hbsK80Q (BD8576), and hbsK86Q (BD7493). (B) Wildtype (BD630), hbsK3R (BD8387), hbsK18R (BD8190), hbsK37R (BD8120), hbsK41R (BD8148), hbsK75R (BD8333), hbsK80R (BD7484), and hbsK86R (BD7506).

Figure 7

The influence of HBsu acetylation the formation of persisters. Cells were pre-grown in LB media for 1 h, and then exposed to 5 μg/mL kanamycin for 5 h. Serial dilutions were plated to determine survivors (CFUs/mL), which likely represented persisters. The data shown is an average of at least three independent replicates with standard deviation displayed. (A) Wildtype (BD630), hbsK3Q (BD8577), hbsK18Q (BD8219), hbsK37Q (BD8119), hbsK41Q (BD8147), hbsK75Q (BD8398), hbsK80Q (BD8576), and hbsK86Q (BD7493). (B) Wild-type (BD630), hbsK3R (BD8387), hbsK18R (BD8190), hbsK37R (BD8120), hbsK41R (BD8148), hbsK75R (BD8333), hbsK80R (BD7484), and hbsK86R (BD7506). CFU, Colony forming unit; WT, Wildtype.

Figure 8

HBsu acetylation influences recovery from the persistent state. Persistent cells were prepared in the presence of kanamycin (A,B) and vancomycin (C,D), as described in the Materials and Methods section. Cell pellets were washed three times in LB without antibiotics and added to a 96-well plate. Growth was monitored by OD₆₀₀ determination every 10 min for 8 h in a microplate reader. The curves plotted are the averages of six determinations (three technical replicates for each of two biological replicates). (A,C) Wild-type (BD630), hbsK3Q (BD8577), hbsK18Q (BD8219), hbsK37Q (BD8119), hbsK41Q (BD8147), hbsK75Q (BD8398), hbsK80Q (BD8576), and hbsK86Q (BD7493). (B,D) Wild-type (BD630), hbsK3R (BD8387), hbsK18R (BD8190), hbsK37R (BD8120), hbsK41R (BD8148), hbsK75R (BD8333), hbsK80R (BD7484), and hbsK86R (BD7506).

Figure 9

Model of HBsu response to antibiotic stress. (A) A B. subtilis HBsu structural model was generated in PyMOL using the B. stearothermophilus ortholog as a template (PDB 1HUU). The HBsu homodimer is displayed, with one monomer colored in cyan with the acetylated lysine sites (K3, K18, K37, K41, K75, K80, and K86) labeled in pink, and the other monomer colored purple with the acetylated sites labeled yellow. The left panel is a ribbon diagram, and the right shows a space-filling model. Reproduced and modified with permission from Carabetta et al. (2019). (B) During normal, exponential phase growth, HBsu is acetylated in specific patterns over the chromosome (left). In response to antibiotic stress, possibly only those that induce the formation of ROS, HBsu is deacetylated by an unknown KDAC, especially important at K3 and K75, which leads to a more compacted nucleoid (middle). This compaction may protect against further damage and/or change the transcriptional program to aid in survival. When the stress is removed, an unknown KAT acetylates HBsu and acetylation at K41 is likely an important early event to trigger the reentry into growth (right).