Ribosome Reconstruction during Recovery from High-Hydrostatic-Pressure-Induced Injury in Bacillus subtilis

HHP treatment is used as a nonthermal processing technology in the food industry to inactivate bacteria while retaining high quality of foods under suppressed chemical reactions. However, some populations of bacterial cells may survive the inactivation. Although the survivors are in a transient nongrowing state due to HHP-induced injury, they can recover from the injury and then start growing, depending on the postprocessing conditions. The recovery process in terms of cellular components after the injury remains unclear. Transcriptome analysis using vegetative cells of Bacillus subtilis revealed that the translational machinery can preferentially be reconstructed after HHP treatment. We found that both Mn²⁺ and Zn²⁺ prolonged the growth-arrested stage of HHP-injured cells by delaying ribosome reconstruction. It is likely that ribosome reconstruction is crucial for the recovery of growth ability in HHP-injured cells. This study provides further understanding of the recovery process in HHP-injured B. subtilis cells.

ABSTRACT

Vegetative cells of Bacillus subtilis can recover from injury after high-hydrostatic-pressure (HHP) treatment at 250 MPa. DNA microarray analysis revealed that substantial numbers of ribosomal genes and translation-related genes (e.g., translation initiation factors) were upregulated during the growth arrest phase after HHP treatment. The transcript levels of cold shock-responsive genes, whose products play key roles in efficient translation, and heat shock-responsive genes, whose products mediate correct protein folding or degrade misfolded proteins, were also upregulated. In contrast, the transcript level of hpf, whose product (Hpf) is involved in ribosome inactivation through the dimerization of 70S ribosomes, was downregulated during the growth arrest phase. Sucrose density gradient sedimentation analysis revealed that ribosomes were dissociated in a pressure-dependent manner and then reconstructed. We also found that cell growth after HHP-induced injury was apparently inhibited by the addition of Mn²⁺ or Zn²⁺ to the recovery medium. Ribosome reconstruction in the HHP-injured cells was also significantly delayed in the presence of Mn²⁺ or Zn²⁺. Moreover, Zn²⁺, but not Mn²⁺, promoted dimer formation of 70S ribosomes in the HHP-injured cells. Disruption of the hpf gene suppressed the Zn²⁺-dependent accumulation of ribosome dimers, partially relieving the inhibitory effect of Zn²⁺ on the growth recovery of HHP-treated cells. In contrast, it was likely that Mn²⁺ prevented ribosome reconstruction without stimulating ribosome dimerization. Our results suggested that both Mn²⁺ and Zn²⁺ can prevent ribosome reconstruction, thereby delaying the growth recovery of HHP-injured B. subtilis cells.

IMPORTANCE HHP treatment is used as a nonthermal processing technology in the food industry to inactivate bacteria while retaining high quality of foods under suppressed chemical reactions. However, some populations of bacterial cells may survive the inactivation. Although the survivors are in a transient nongrowing state due to HHP-induced injury, they can recover from the injury and then start growing, depending on the postprocessing conditions. The recovery process in terms of cellular components after the injury remains unclear. Transcriptome analysis using vegetative cells of Bacillus subtilis revealed that the translational machinery can preferentially be reconstructed after HHP treatment. We found that both Mn²⁺ and Zn²⁺ prolonged the growth-arrested stage of HHP-injured cells by delaying ribosome reconstruction. It is likely that ribosome reconstruction is crucial for the recovery of growth ability in HHP-injured cells. This study provides further understanding of the recovery process in HHP-injured B. subtilis cells.

INTRODUCTION

High-hydrostatic-pressure (HHP) treatment is a nonthermal food processing technology that can extend the shelf life of food via microbial inactivation while maintaining nutrients, flavors, and colors, which are often deteriorated by heat treatment (1). On the other hand, some populations of HHP-treated cells can survive the treatment, which is typically carried out at 400 to 600 MPa. Although these survivors are in a transient nongrowing state due to HHP-induced injuries, they can recover from the injury during storage or incubation (2–4).

HHP treatment inhibits various cellular processes, including motility, cell division, replication, transcription, and translation, thereby causing pleiotropic effects on cells (5–8). Previous studies have revealed that ribosomes could be a primary target for HHP-induced cell death in bacteria (9, 10). Proteomic and transcriptomic studies of Escherichia coli showed that heat shock proteins (HSPs), such as DnaK and GroESL, and cold shock proteins (CSPs), such as CspA and CspB, were induced in the cells under sublethal HHP at around 50 MPa (11, 12). In addition, sublethal heat shock was shown to induce pressure resistance in E. coli (13), while other experiments demonstrated that cold shock affected pressure resistance in Staphylococcus aureus and Listeria monocytogenes (14, 15). It is likely that HSPs and CSPs help protect cells against HHP-induced stresses. As previously demonstrated by VanBogelen and Neidhardt (16), ribosome-targeting antibiotics can induce HSPs or CSPs, leading these authors to suggest that ribosomes can act as cellular thermosensors. Based on these previous studies, ribosome damage associated with HHP treatment may trigger the inductions of HSPs and CSPs.

It has also been demonstrated that HHP treatment can induce endogenous oxidative stress in E. coli (17). Since increased levels of intracellular reactive oxygen species (ROS) can cause damage to nucleic acids, proteins, and membrane lipids, ROS scavengers, such as catalase and superoxide dismutase, are considered to play important roles in HHP resistance in E. coli (17). Thus, any comprehensive analysis of HHP-treated E. coli cells should include the effects of both HHP and oxidative stresses.

In a Gram-positive model bacterium, Bacillus subtilis, much effort has been devoted to optimizing the HHP conditions necessary for spore inactivation or germination (18, 19). In contrast, relatively few studies have examined the effect of HHP on B. subtilis vegetative cells. We have previously reported the effect of HHP on vegetative cells of a sporulation-deficient B. subtilis strain (2). The characteristics of HHP-injured B. subtilis cells are different from those of E. coli cells. Oxidative stress contributes to HHP-induced cell death in E. coli (17), while ROS scavengers had little or no effect on the viability of HHP-treated B. subtilis cells (2). Furthermore, HHP-injured B. subtilis cells are highly sensitive to salts, with cell lysis occurring in the presence of NaCl or KCl at concentrations of 100 mM or higher (2). These results suggest that HHP-induced damage and subsequent recovery may differ between E. coli and B. subtilis. Our research goal is to understand how bacteria can recover from HHP-induced injury. In this study, DNA microarray analysis was performed to investigate gene expression patterns and the ribosome reconstruction process was analyzed in HHP-injured vegetative cells of B. subtilis.

RESULTS AND DISCUSSION

Transient growth arrest of B. subtilis cells treated by HHP.

To investigate the effect of HHP on B. subtilis vegetative cells, the sporulation-deficient strain TI465, which lacks the sporulation-specific σ factor F (2), was used in this study. When stationary-phase cells (6 h of cultivation in NaCl-free L medium [10 g of tryptone and 5 g of yeast extract, both per liter]) were exposed to 250 MPa of HHP (25°C for 10 min) and subsequently diluted into fresh recovery medium (NaCl-free L medium), the cells showed a prolonged growth lag. The average growth delay (the time difference between HHP-treated and untreated culture) at an optical density at 600 nm (OD₆₀₀) of 1.0 was 160 ± 21 min. Under the experimental condition, a 10-fold dilution of untreated control culture (1.0-log reduction in viable cell counts) resulted in a delay of 70 ± 13 min. Accordingly, the apparent reduction in viable cells after HHP treatment therefore can be expressed as (DT_HHP/DT_CONT) log CFU per milliliter, where DT_HHP and DT_CONT indicate the delay times for HHP-treated culture and for 10-fold-diluted control culture, respectively. Based on this calculation, the apparent reduction in viable-cell counts after 250 MPa treatment was 2.3 log CFU/ml. However, the actual reduction in viable-cell counts as determined by colony formation was approximately 0.3 log CFU/ml. The different viability values extracted from growth delay and colony counts indicate that the HHP-treated B. subtilis culture included nongrowing cells as previously demonstrated for E. coli (3, 4). Hence, we concluded that a large population of viable cells after 250 MPa of HHP treatment was temporarily in the nongrowing state (growth arrest phase).

Transcriptome analysis of HHP-injured B. subtilis cells.

To understand the recovery process in HHP-injured B. subtilis cells, gene expression patterns following HHP treatment were investigated using DNA microarray analysis. The cells treated at 250 MPa were transferred into the recovery medium and incubated with vigorous shaking. Total RNAs were extracted from the cells incubated for 0 h (immediately after transferring the HHP-treated cells into the recovery medium), 2 h (growth arrest phase), and 4 h (regrowth phase) and then subjected to DNA microarray analysis. Principal-component analysis (PCA) showed that the gene expression profile at each time point was clearly different from the others (see Fig. S1 in the supplemental material). The microarray data were analyzed using the rank products method (20) to identify differentially expressed genes (DEGs) that were either up- or downregulated after HHP treatment. With a false-discovery rate (FDR) of less than 5%, 77 genes were identified as DEGs (Tables 1 and 2). To confirm the DNA microarray results, reverse transcription-quantitative PCR (RT-qPCR) analysis was carried out, focusing on one or two genes from each transcription unit of all DEGs (Fig. 1 and Fig. S2).

TABLE 1.

Genes induced during the recovery phase of HHP-injured B. subtilis

Gene	Gene product (function)	Expression ratioa
		2 h/0 h	4 h/0 h	4 h/2 h
citB	Aconitase	2.89
infA	Translation initiation factor IF-1	3.46
infC	Translation initiation factor IF-3	2.88
map	Methionine aminopeptidase	2.67
mtlD	Mannitol-1-phosphate 5-dehydrogenase	3.01
pdhA	Pyruvate dehydrogenase (E1 alpha subunit)	2.32
pstA	Phosphate ABC transporter (permease)	4.03
pstS	Phosphate ABC transporter (binding protein)	3.94
rplE	Ribosomal protein L5	3.04
rplJ	Ribosomal protein L10	2.68
rplN	Ribosomal protein L14	3.33
rplP	Ribosomal protein L16	3.47
rplR	Ribosomal protein L18	3.02
rpmC	Ribosomal protein L29	3.48
rpmE	Ribosomal protein L31	2.16
rpsN	Ribosomal protein S14	2.94
rpsQ	Ribosomal protein S17	3.15
tuaG	Biosynthesis of teichuronic acid	2.89
ylbN	Unknown	3.16
yubF	Unknown	2.81
yugI	General stress protein found in the ribosome fraction	3.14
clpE	Clp ATPase	4.98	1.52
cspB	Cold shock protein	2.77	3.52
rplA	Ribosomal protein L1	2.78	2.95
rplS	Ribosomal protein L19	3.63	2.93
rpmJ	Ribosomal protein L36	3.38	2.56
tig	Trigger factor	2.89	3.72
yneF	Unknown	5.68	5.24
yozC	Unknown	2.42	3.06
cspD	Cold shock protein	2.71	4.58	1.87
groES	10-kDa chaperonin	3.74	5.27	1.53
abh	Transcriptional regulator		3.61	2.30
acpP	Acyl carrier protein		4.33	2.61
feuA	Siderophore binding protein		3.59	4.38
ydbN	Unknown		5.40	3.94
yfiY	Siderophore uptake?		4.15	3.67
ykuN	Flavodoxin		3.30	4.07
yolA	Unknown; SP-β protein		4.04	4.47
yqgA	Unknown; cell wall binding protein		4.37	3.22
ysbA	Murein hydrolase regulator LrgA; essential for pyruvate utilization		5.92	6.58
ysbB	Antiholin-like protein LrgB		4.93	5.67
hag	Flagellin protein			3.53
hbs	Nonspecific DNA-binding protein HBsu			2.39
rbsA	Ribose ABC transporter (ATP-binding protein)			3.88
spoVG	Stage V sporulation protein G (spore cortex synthesis)			2.79
ykuP	Flavodoxin			4.18
yolB	Unknown; SP-β protein			3.34
ywsB	Unknown; cell wall binding protein			1.92

^aExpression ratio indicates the average value of the log₂ ratio between two independent microarray results.

TABLE 2.

Genes repressed during the recovery phase of HHP-injured B. subtilis

Gene	Gene product (function)	Expression ratioa
		2 h/0 h	4 h/0 h	4 h/2 h
ahpC	Alkyl hydroperoxide reductase (small subunit)	−3.94
ahpF	Alkyl hydroperoxide reductase (large subunit)	−3.64
bfmBAA (bkdA1)	Branched-chain alpha-keto acid dehydrogenase E1 subunit	−2.58
csbDb	General stress protein	−3.39
gapB	Glyceraldehyde-3-phosphate dehydrogenase	−2.47
hag	Flagellin protein	−4.46
mntA	Manganese ABC transporter (membrane protein)	−2.89
mntB	Manganese ABC transporter (ATP-binding protein)	−3.33
pckA	Phosphoenolpyruvate carboxykinase	−2.92
rbsA	Ribose ABC transporter (ATP-binding protein)	−3.16
spoVG	Stage V sporulation protein G (spore cortex synthesis)	−3.32
yflTb	General stress protein	−3.16
yqgZ (mgsR)	Spx paralog controls a subregulon within general stress response	−3.71
ytxH	General stress protein	−2.55
hpf (yvyD)	Dimerization of 70S ribosome	−4.45
ywsB	Unknown; putative cell wall binding protein	−3.27
citZ	Citrate synthase II	−2.69	−2.14
gsiB	General stress protein	−2.73	−1.74
katA	Vegetative catalase 1	−4.06	−3.52
mntC	Manganese ABC transporter (membrane protein)	−3.40	−1.94
mrgA	Metalloregulation DNA-binding stress protein; E. coli Dps homolog	−3.34	−3.02
srfAA	Surfactin synthetase	−2.45	−2.60
ybyB	Unknown	−4.72	−3.16
yhzC	General stress protein	−2.86	−2.55
ykzI	General stress protein	−3.94	−3.57
appC	Oligopeptide ABC transporter (permease)	−1.97	−3.91	−1.94
phrA	Phosphatase RapA inhibitor	−1.84	−3.76	−1.92
rapA	Response regulator aspartate phosphatase	−2.08	−3.75	−1.67
ybcO (skfA)	Sporulation killing factor	−2.22	−3.62	−1.40
yydF	Inducer of LiaRS	−2.65	−3.83	−1.19
mntD	Manganese ABC transporter		−2.76
nhaX	Regulator of nhaC		−2.67
srfAB	Surfactin synthetase		−2.59
clpE	Clp ATPase			−3.45
pstA	Phosphate ABC transporter (permease)			−3.09
mtlD	Mannitol-1-phosphate 5-dehydrogenase			−3.40
citB	Aconitase			−3.26
tuaG	Biosynthesis of teichuronic acid			−2.39

^aExpression ratio indicates the average value of the log₂ ratio between two independent microarray results.

^bNot reproducible by RT-qPCR analysis.

FIG 1.

Transcriptional analysis of genes affected by HHP treatment in B. subtilis. B. subtilis strain TI465 was subjected to HHP treatment at 250 MPa at 25°C for 10 min. After HHP treatment, the culture was diluted 5-fold in fresh NaCl-free L medium and incubated at 37°C with vigorous shaking. After incubation for 0 h (immediately after dilution), 2 h, or 4 h, cells were collected by centrifugation. Then total RNAs were extracted and used for RT-qPCR analysis. The relative expression was expressed as the relative amount of each transcript per 1,000 units of 16S rRNA (ΔCt). Genes are classified into four groups according to their functions: translation-related genes (A), stress-responsive genes (B), metabolic genes (C), and metal-regulated genes (D). The average values of at least four independent experiments are shown.

(i) Translation-related genes. DNA microarray analysis showed that 12 ribosomal protein genes (rplA, rplE, rplJ, rplN, rplP, rplR, rplS, rpmC, rpmE, rpmJ, rpsN, and rpsQ), 2 genes for translation initiation factors (infA and infC), and a gene for amino-terminal processing of proteins (map) were upregulated during the growth arrest phase (Table 1). Among them, 10 genes (rplE, rplN, rplP, rplR, rpmC, rpmJ, rpsN, rpsQ, map, and infA) belong to an identical ribosomal operon, S10-spc-α, which includes 25 ribosomal genes (21). The entire S10-spc-α operon is transcribed from the S10 promoter located immediately upstream of rpsJ, which is the first gene of this operon. Although three additional weak promoters are found in the internal region of the S10-spc-α operon (21), the S10 promoter is a sole promoter for seven genes that were listed as upregulated genes (rplE, rplN, rplP, rplR, rpmC, rpsN, and rpsQ). Since no significant difference in the rpsJ transcript was detected by DNA microarray analysis, we further validated the transcript level of rpsJ after HHP treatment by RT-qPCR. The result showed that the rpsJ gene was also upregulated in the growth arrest phase (Fig. 1A). The map, infA, and rpmJ genes can be transcribed from not only the S10 promoter but also additional promoters existing in the internal region of this operon. Hence, the infA gene was also employed for RT-qPCR analysis. In addition, the transcript levels of the remaining five genes (rplA, rplJ, rplS, rpmE, and infC), which are transcribed from a different transcription unit, were further analyzed. RT-qPCR analyses confirmed that the transcript levels of these six genes as well as rpsJ were enhanced within 2 h after HHP treatment (Fig. 1A). The hpf (yvyD) gene, whose gene product is required for ribosome inactivation by dimerizing 70S ribosomes (22), was downregulated in the growth arrest phase.

(ii) Stress-responsive genes. As observed in previous studies on E. coli (11, 12), the transcript levels of the CSP genes (cspB, cspD, and tig) as well as those of the HSP genes (groES and clpE) were increased after HHP treatment, as shown in Table 1 and Fig. 1B. CspB and CspD act as RNA chaperones that are essential for efficient translation at both low and optimal growth temperatures (23). The ribosome-associated trigger factor, which is encoded by tig, is known to initiate folding of newly synthesized proteins (24). The chaperonin GroES also supports translation through mediating correct protein folding under both normal and heat shock conditions (25). A heat-inducible Clp ATPase, ClpE, degrades misfolded proteins and a class III heat shock regulator CtsR, thereby inducing CtsR regulon genes (i.e., clpP, clpC, and clpE) (26, 27). These stress-responsive proteins might contribute to efficient translation (CspB, CspD, Tig, and GroES) and protein quality control (ClpE) in HHP-injured cells.

Oxidative stress has been reported to contribute to HHP-induced cell death in E. coli (17). In B. subtilis, however, disruption of the genes for vegetative catalase (katA and katE) had no significant effect on the viability of HHP-treated cells (2). In agreement with this previous observation, several PerR regulon genes (ahpC, katA, and mrgA), which play important roles in cellular protection against oxidative stress (28–30), were downregulated after HHP treatment (Table 2 and Fig. 1B).

The general stress regulon, which is responsible for the development of nonspecific stress resistance, is induced by nutrient limitation (31). The transcript levels of general stress genes (gsiB, yhzC, ykzI, and ytxH) were reduced in the growth arrest phase (Fig. S2). Although the DNA microarray analysis data showed that two general stress genes, csbD and yflT, were reduced after HHP treatment, this was not reproducible by RT-qPCR analysis (Fig. S2).

(iii) Metabolic genes. Several metabolic genes (pdhA, citB, tuaG, and pstA) were upregulated in the growth arrest phase (Fig. 1C). The pdhA gene is the first gene of the pdhABCD operon, which encodes the pyruvate dehydrogenase multienzyme complex. This enzyme complex catalyzes the irreversible oxidative decarboxylation of pyruvate to acetyl coenzyme A (acetyl-CoA), linking glycolysis with the tricarboxylic acid (TCA) cycle (32). Although the transcript level of the citB gene (which encodes aconitase, the enzyme which catalyzes the interconversion of citrate and isocitrate in the second step of the TCA cycle) was increased during the growth arrest phase, the transcript level of the citZ gene (which encodes the major citrate synthase in the first step of the TCA cycle) was decreased (Fig. 1C). These results suggest that HHP treatment may result in the reduction in the intracellular level of TCA cycle metabolites and the accumulation of acetyl-CoA. Our preliminary metabolome analysis using 250-MPa-treated B. subtilis cells also supported this prediction (unpublished results). Recently, it has been reported that compromising TCA cycle enzymes can affect HHP resistance in E. coli (33). Changes in the intracellular level of TCA cycle metabolites might affect the recovery of HHP-injured B. subtilis cells. Membrane damage is thought to be one of the major causes of HHP-induced death (34). The accumulation of acetyl-CoA, which is important for fatty acid biosynthesis (35), may be required for membrane repair. The transcript level of acpP, encoding an acyl carrier protein which plays an essential role in fatty acid biosynthesis (36), was increased in the regrowth phase (Fig. 1C). It seems that membrane repair would begin after ribosome reconstruction. As previously demonstrated, HHP-treated E. coli cells can be stained with propidium iodide (PI), which enters cells with a compromised membrane (3, 34). In B. subtilis, no obvious incorporation of PI into 250-MPa-treated cells was observed (data not shown). HHP treatment at 250 MPa might not cause severe damage on membrane. The tuaG gene belongs to the tuaABCDEFGH operon, which encodes enzymes required for the polymerization of teichuronic acid, and the pstA gene belongs to the pst operon, which encodes a phosphate transporter. Both operons are induced under phosphate-limited conditions (37, 38). The transcript levels of the tuaG and pstA genes were both increased in the growth arrest phase and then decreased in the regrowth phase (Fig. 1C). HHP treatment may cause phosphate limitation in B. subtilis vegetative cells. Transcription of gluconeogenic genes (e.g., gapB and pckA) is known to be repressed under glycolytic conditions and derepressed under gluconeogenic conditions (39). In HHP-treated cells, gapB and pckA genes were downregulated in the growth arrest phase and upregulated in the regrowth phase (Table 2 and Fig. 1C).

(iv) Metal-regulated genes. Because metal ions, such as iron and manganese, are required as cofactors for a large number of enzymes, these metal ions are also important for cell growth. DNA microarray and RT-qPCR analyses showed that five genes (feuA, ydbN, yfiY, ykuN, and ykuP) belonging to the Fur regulon were upregulated in the regrowth phase (Table 1 and Fig. 1D). Although DNA microarray analysis showed that the transcript level of the mntABCD operon, belonging to the MntR regulon, was reduced throughout the recovery period, RT-qPCR analysis indicated that the transcript level of this operon was also upregulated in the regrowth phase (Table 2 and Fig. 1D). The Fur regulon is induced under iron-limited conditions, while the MntR regulon is induced under manganese limitation (40, 41). Metal ions may be required for the activation of newly synthesized metalloproteins during the regrowth phase of HHP-treated cells.

Ribosomes are reconstructed during the growth arrest phase.

DNA microarray analysis indicated that many translation-related factors were induced during the growth arrest phase. To examine the effect of HHP on ribosomes, we analyzed ribosome profiles using sucrose density gradient sedimentation (Fig. 2A). In the stationary-phase untreated cells grown under atmospheric pressure (0.1 MPa), ribosomes largely formed multimers (dimers or polysomes), as described previously (22). Although 100-MPa-treated cells grew normally (data not shown), accumulations of 30S and 50S ribosomal subunits were observed, indicating that ribosomes partly dissociated (Fig. 2A). In the cells treated with 250 MPa, the peaks corresponding to intact ribosomes (70S, dimers, and polysomes) almost completely disappeared, with a concomitant increase in the 50S peak. In the 400-MPa-treated cells, however, even the amount of 50S decreased compared with that of the 250-MPa-treated cells, suggesting further dissociation or degradation of the ribosomes. Thus, ribosomes were dissociated or degraded after HHP treatment in a pressure-dependent manner.

FIG 2.

Effect of HHP on ribosomes in B. subtilis cells. (A) Pressure-dependent dissociation of ribosomes. B. subtilis strain TI465 was exposed to HHP at 100, 250, or 400 MPa at 25°C for 10 min. After HHP treatment, cells were collected by centrifugation. Then each cell extract was prepared and analyzed using 10% to 40% sucrose gradient sedimentation. Cell extract from an untreated culture (0.1 MPa) was also examined as a control. (B) Change in the ribosome sedimentation profile during recovery after HHP treatment. B. subtilis strain TI465 was subjected to 250 MPa at 25°C for 10 min. After the HHP treatment, the culture was diluted 5-fold in fresh NaCl-free L medium and incubated at 37°C with vigorous shaking. To compare with the ribosome profile in untreated cells, the untreated culture was also prepared by the same method except for HHP treatment. After incubation for the indicated times, cells were collected by centrifugation. Cell extracts were then analyzed using 10% to 40% sucrose gradient sedimentation.

We next monitored ribosome profiles during the recovery of HHP-injured cells (Fig. 2B). In untreated cells, inactive ribosome dimers were rapidly dissociated into active 70S within 30 min after cells were transferred into fresh medium (Fig. 2B, left). In contrast, ribosomes dissociated into 30S and 50S subunits immediately after 250-MPa treatment (0 h) (Fig. 2B, right). In the growth arrest phase (2 h of incubation), the amounts of ribosomal particles 30S and 50S apparently decreased, while the amounts of 70S ribosomes remained low. It is likely that free ribosomal particles (30S and 50S) were degraded. Zundel et al. previously reported that free ribosomal particles (30S and 50S) are more susceptible to degradation by RNase(s) than 70S ribosomes (42). Hence, the formation of free ribosomal particles (30S and 50S) can initiate ribosome breakdown. This is a well-known response to generate nutrients from degraded ribosomes under nutrient-limiting conditions (42–44). In HHP-treated cells, the dissociated ribosomal particles also can be degraded and recycled as a nutrient source. In the regrowth phase (after 4 h of incubation), ribosomes formed multimers (dimers or polysomes) similar to that in stationary-phase cells (Fig. 2A, 0.1 MPa), whereas it formed 70S in growing untreated cells (Fig. 2B, untreated cells, 0.5 and 1 h). Since our preliminary metabolome analysis suggested a remarkable reduction in the intracellular pools of nucleotides and amino acids in 250-MPa-treated B. subtilis cells (unpublished results), the nutrient availability might be restricted due to HHP-induced impairments.

Growth recovery of HHP-injured B. subtilis cells is delayed in the presence of Mn²⁺ or Zn²⁺.

Our transcriptional analyses indicated that both the Fur and MntR regulons were upregulated in the regrowth phase (Fig. 1). Since metal ions are required for a large number of metalloproteins, we hypothesized that HHP-injured cells might require several metal ions for their growth. Accordingly, we examined the effects of several metal ions on the growth recovery of HHP-injured cells. Because ribosomes contain Mg²⁺ and Zn²⁺, we tested not only FeSO₄ and MnCl₂ but also MgSO₄ and ZnSO₄. Contrary to our hypothesis, MnCl₂ and ZnSO₄ had inhibitory effects on the growth recovery of HHP-treated cells (Fig. 3), while FeSO₄ (100 μM) and MgSO₄ (10 mM) had no significant effect (data not shown). The addition of MnCl₂ to the recovery medium prolonged the growth arrest phase (Fig. 3A). The average growth delay in Mn²⁺-supplemented medium compared to that in nonsupplemented medium (160 ± 21 min) was extended to 210 ± 5.8 min at 10 μM and 260 ± 5.8 min at 100 μM, respectively. ZnSO₄ also inhibited growth recovery at a concentration of 100 μM but not at 10 μM (Fig. 3B). The average growth delay in the presence of 100 μM ZnSO₄ was 250 ± 29 min. Since neither MnCl₂ nor ZnSO₄ had any significant effect on the cell viability as determined by colony formation (data not shown), it seems that these metal ions inhibited the recovery process. When E. coli W3110 cells were treated under the same condition (250 MPa at 25°C for 10 min) and diluted in fresh L medium supplemented with MnCl₂ (100 μM) or ZnSO₄ (100 μM), no significant effects of these metal ions on the growth recovery were observed compared to growth in nonsupplemented L medium (data not shown).

FIG 3.

Effects of Mn²⁺ and Zn²⁺ on the growth recovery of HHP-treated B. subtilis cells. B. subtilis strain TI465 was treated with 250 MPa of HHP at 25°C for 10 min and diluted 10-fold in a fresh NaCl-free L medium supplemented with MnCl₂ (A) or ZnSO₄ (B). Cell growth was monitored by measuring the OD₆₀₀.

To examine the effects of MnCl₂ and ZnSO₄ on the transcription of ribosomal genes during the recovery phase, we monitored the transcript level of the rpsJ gene by performing RT-qPCR. The results showed a significant delay in the transcript level of rpsJ when HHP-treated cells were incubated in the recovery medium supplemented with MnCl₂ or ZnSO₄ (Fig. 4). Sucrose density gradient sedimentation analysis also indicated that ribosome reconstruction was delayed under these conditions (Fig. 5). In the presence of MnCl₂ or ZnSO₄, the 50S peak was still high even after incubation for 4 h, suggesting that both Mn²⁺ and Zn²⁺ ions might prevent the degradation of dissociated 50S particle, either directly or indirectly. Interestingly, an unusual accumulation of ribosome multimers (dimers or polysomes) was found after incubation for 2 h in the presence of Zn²⁺, but not with Mn²⁺ (Fig. 5K and F). Western blot analysis using antibodies against ribosomal protein L2 and ribosome dimerization factor Hpf revealed that both were detected in this peak (Fig. 6B and C). No such accumulation of Hpf was observed in cells incubated in the Mn²⁺-supplemented recovery medium (Fig. S3). These results suggest that Zn²⁺ can stimulate dimer formation of ribosomes in HHP-treated cells. In contrast, it seems that Mn²⁺ prevented ribosome reconstruction without stimulating ribosome dimerization (Fig. 5).

FIG 4.

Effects of Mn and Zn on transcript level of rpsJ following HHP treatment. B. subtilis strain TI465 was subjected to HHP treatment at 250 MPa at 25°C for 10 min. After HHP treatment, the culture was diluted 5-fold in fresh NaCl-free L medium (white) or the same medium supplemented with 0.1 mM MnCl₂ (gray) or 0.1 mM ZnSO₄ (black). After the incubation for 0 h (immediately after dilution), 2 h, 4 h, or 6 h, cells were collected by centrifugation. Then total RNAs were extracted and used for RT-qPCR analysis. Total RNA in untreated culture immediately before HHP treatment was also prepared. The transcript levels of rpsJ were normalized by adjusting the amount of 16S rRNA in each RNA sample. The relative expression of rpsJ was calculated by dividing the amount of rpsJ mRNA in HHP-treated cells at each time point by that in cells immediately before HHP treatment (ΔΔCt). The average values of three independent experiments are shown. Statistical analysis was performed by using Dunnett’s test. Asterisks indicate significant differences compared with the nonsupplemented control at each time point (P < 0.05).

FIG 5.

Effects of Mn²⁺ and Zn²⁺ on ribosome reconstruction in HHP-treated B. subtilis cells. B. subtilis strain TI465 was subjected to 250 MPa at 25°C for 10 min. After the HHP treatment, the culture was diluted 5-fold in fresh NaCl-free L medium (C to E) or the same medium supplemented with 0.1 mM MnCl₂ (F to J) or 0.1 mM ZnSO₄ (K to O). After each indicated time, the cells were collected by centrifugation. Cell extracts were then analyzed using 10% to 40% sucrose gradient sedimentation. The cell extract just before the HHP treatment is also shown (A). The cell extract at 0 h (B) was prepared from the HHP-treated culture before dilution with the fresh medium.

FIG 6.

Western blot analysis of a Zn²⁺-dependent unusual peak. A B. subtilis strain TI465 culture was treated with 250 MPa at 25°C for 10 min. After the HHP treatment, the culture was diluted 5-fold in fresh NaCl-free L medium supplemented with 0.1 mM ZnSO₄ and incubated for 2 h at 37°C with vigorous shaking. Cell extract was then fractionated using a sucrose density gradient (A). Each fractionated sample was analyzed by Western blotting using anti-L2 (B) or anti-Hpf (C) rabbit antisera. The arrow indicates the peak that appeared in the presence of ZnSO₄.

Hpf is involved in Zn²⁺-induced dimerization of ribosomes in HHP-treated cells.

We analyzed whether Zn²⁺-induced dimerization of ribosomes can occur even in a Δhpf mutant (Fig. 7). In the hpf⁺ strain, ribosomes formed the dimers within 1 h after the HHP-treated cells were transferred into a Zn²⁺-supplemented medium (Fig. 7B). In contrast, the ribosome dimers were no longer observed in the Δhpf mutant, indicating that Hpf was required for the Zn²⁺-induced dimerization in the HHP-treated cells (Fig. 7F to J). In addition, the accumulation of 70S in the Δhpf mutant was observed after 4 h of incubation, whereas no accumulation was observed with the hpf⁺ strain (Fig. 7E and J). Moreover, the Δhpf mutant showed an approximately 70-min shorter growth delay than that of the hpf⁺ strain (Fig. 8). These results suggest that Zn²⁺ can prevent the growth recovery of HHP-injured cells by stimulating Hpf-dependent dimerization. In this study, we used stationary-phase cells that contain Hpf protein (22). Since the addition of ZnSO₄ to the recovery medium had no effect on the hpf expression (data not shown), preexisting Hpf protein can be involved in Zn²⁺-induced dimer formation. We could not observe the Zn²⁺-induced dimer formation in untreated cells (data not shown). Conformational changes caused by HHP might be required for Zn²⁺-induced dimer formation. B. subtilis has several ribosomal proteins with the Zn-binding motif CXXC (45, 46). These Zn-binding motifs might be involved in the Zn²⁺-induced dimer formation. It should be noted that the inhibitory effect of Zn²⁺ ions on the growth was still observed even in the Δhpf mutant (Fig. 8). This indicates that Zn²⁺ also has a different inhibitory effect on the growth recovery of HHP-injured B. subtilis cells. Zn-sensing transcriptional regulator Zur represses at least three Zn-containing ribosomal protein paralogs (L31 paralog YtiA, L33 paralog RpmGC, and S14 paralog YhzA) under Zn-rich conditions (45–48). Moreover, an elevated level of Mn²⁺ results in repression of not only the MntR regulon but also the Fur and PerR regulons (47, 48). We therefore cannot exclude the possibility that Mn²⁺ and Zn²⁺ ions affect the growth recovery of HHP-injured B. subtilis cells through the function of these metal-sensing regulators.

FIG 7.

Effect of Hpf on Zn²⁺-dependent dimerization of ribosomes in HHP-treated B. subtilis cells. B. subtilis strain TI477 (hpf⁺ [A to E]) and TI478 (Δhpf [F to J]) cultures were treated with 250 MPa at 25°C for 10 min. After the HHP treatment, each culture was diluted 5-fold in fresh NaCl-free L medium supplemented with 0.1 mM ZnSO₄. After incubation for 0 h (immediately after dilution), 1 h, 2 h, 3 h, and 4 h, the cells were collected by centrifugation. Cell extracts were then analyzed using 10% to 40% sucrose gradient sedimentation. The arrows indicate peaks corresponding to ribosome dimers.

FIG 8.

Effects of Hpf and Zn supplementation on the growth of HHP-treated B. subtilis. B. subtilis strain TI477 (wild type [WT]; hpf⁺) and TI478 (Δhpf) cultures were treated with 250 MPa for 10 min at 25°C. After the HHP treatment, the cultures were diluted 10-fold in fresh NaCl-free L medium (open symbols) or the same medium supplemented with 0.1 mM ZnSO₄ (closed symbols). Cell growth was monitored by measuring the OD₆₀₀.

Concluding remarks.

In this study, we investigated the ribosome profile during the recovery of HHP-injured B. subtilis cells. When B. subtilis cells were treated with HHP at 250 MPa or higher, intact ribosomes (70S, dimer, and polysome) were largely dissociated or degraded (Fig. 2A). It is therefore considered that >250-MPa-treated cells cannot grow until active ribosomes are reconstructed.

We found that the addition of Mn²⁺ or Zn²⁺ to the recovery medium apparently delayed ribosome reconstruction, preventing the growth recovery of HHP-injured cells (Fig. 3 and 5). In the recovery medium supplemented with Mn²⁺ or Zn²⁺, the dissociated 50S particle was accumulated for a longer time than in nonsupplemented medium (Fig. 5). Moreover, Zn²⁺, but not Mn²⁺, stimulated Hpf-dependent dimer formation of ribosomes (Fig. 6 and 7). Since the dimerized ribosome is considered more resistant to cleavage by ribonucleases than free ribosomal particles 30S and 50S, Zn²⁺ can prevent ribosome breakdown and the subsequent release of nutrients such as nucleotides and amino acids, thereby delaying ribosome reconstruction. In fact, the inhibitory effect of Zn²⁺ on the growth recovery of HHP-treated cells was partially relieved by disruption of the hpf gene (Fig. 8). In contrast, it seems that Mn²⁺ prevented ribosome reconstruction without stimulating ribosome dimerization (Fig. 5). Although the underlying mechanism is still unclear, we propose that the degradation and recycling of dissociated ribosomes are important for ribosome reconstruction in the early stage of the recovery process in HHP-injured B. subtilis cells.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

All B. subtilis strains used in this study were derived from strain 168. Strain TI465 (trpC2 spoIIAC::kan) was used because it lacks the sporulation-specific σ factor F (2). The strain RIK2508 (trpC2 Δhpf) was previously constructed from strain 168 (22). Strains TI478 (trpC2 spoIIAC::kan Δhpf) and TI477 (trpC2 spoIIAC::kan) were obtained by transforming strains RIK2508 and its parent strain with genomic DNA of TI465. To confirm the effect of MnCl₂ or ZnSO₄ in HHP-treated E. coli cells, strain W3110 was purchased from the National Bio-Resource Project (NBRP) of the National Institute of Genetics (strain ID; ME9780) and used.

The growth media used were L medium (10 g of tryptone, 5 g of yeast extract, and 5 g of NaCl, all per liter) and NaCl-free L medium (10 g of tryptone and 5 g of yeast extract, both per liter). Kanamycin (3 μg/ml) was used to select the transformants described above.

HHP treatment.

Cells were grown in NaCl-free L medium for 6 h at 37°C with vigorous shaking. For sucrose density gradient sedimentation analysis, the culture was concentrated by centrifugation followed by removal of approximately 95% of the supernatant. Then the culture was packed into a sterile plastic bag that was heat sealed following manual exclusion of visible air bubbles and then subjected to HHP treatment using an HHP food processor (Dr. CHEF; Kobe Steel, Kobe, Japan) at 25°C for 10 min as previously described (2). The rate of increasing and decreasing pressure was set at 200 MPa/min.

Growth delay assay.

Cells were subjected to HHP treatment at an appropriate pressure as described above. Following HHP treatment, 2 ml of the treated culture was inoculated into 18 ml of the recovery medium (NaCl-free L medium) and incubated at 37°C with vigorous shaking (200 rpm). To investigate the effect of metal ions, the recovery medium was supplemented with MnCl₂ or ZnSO₄ at concentrations of 10 or 100 μM. Cell growth was monitored by measuring OD₆₀₀ using an OD-Monitor A&S (TAITEC, Japan). The measurable range of this system was 0.01 to 2.55. The growth delay time at an OD₆₀₀ of 1.0 was expressed as the time difference between test culture and control culture. Under this experimental condition, a 10-fold dilution of untreated culture resulted in a growth delay of 70 ± 13 min (data not shown). At least three independent experiments were performed to assess the reproducibility.

DNA microarray analysis.

Cells were subjected to HHP treatment at 250 MPa as described above. Following HHP treatment, the culture was diluted 5-fold in fresh recovery medium and incubated at 37°C with vigorous shaking. After incubation for 0 h (immediately after dilution), 2 h, or 4 h, cells were collected by centrifugation (5,000 × g, 5 min, and 25°C). Then cells were treated with killing buffer (20 mM Tris-HCl [pH 7.5], 5 mM MgCl₂, 20 mM NaN₃), collected (again using centrifugation), and stored at −80°C until RNA extraction was performed. Total cellular RNA was obtained using Isogen reagent (Nippon Gene, Japan) as previously described (49). cDNA synthesis, labeling, hybridization to the Affymetrix Bacillus subtilis genome array GeneChip, and scanning were carried out by TaKaRa Bio (Japan). All microarray data were MIAME (Minimum Information About a Microarray Experiment) compliant.

An outline of the DNA microarray data analysis was composed of the three steps as follows: data quantification, overall gene expression profile evaluation, and identification of differentially expressed genes (DEGs). The raw microarray data (CEL files) were quantified and processed using R (https://www.r-project.org/) (50), which is a statistical programming language, with relevant R packages provided by Bioconductor (http://www.bioconductor.org/) (51). Quantification was implemented using the distribution-free weighted method (DFW) (52). Global gene expression profiles of all samples were evaluated by principal-component analysis (PCA) (53) using the “prcomp()” function in R. To identify DEGs, the rank products method (20) was applied to the three pairs of groups (0 versus 2 h, 0 versus 4 h, and 2 versus 4 h) selected from the DFW-quantified data, with the number of permutations set at 1,000. Probe sets with a false-discovery rate (FDR) of <0.05 were regarded as having different transcript levels between the two groups (i.e., they were differently expressed). In order to show the changes in expression values intuitively, the expression ratios of the DEGs between the selected pairs of groups were calculated using conventional Mas5-quantified microarray data. The annotation file for the B. subtilis genome array was downloaded from the NetAffx website (http://www.affymetrix.com/analysis/index.affx; as of 23 October 2015). Microarray analysis was performed two times independently for each sample, and the expression ratio was indicated with average values in the log₂ ratio between duplicated independent experiments.

RT-qPCR.

For reverse transcription-quantitative PCR (RT-qPCR), cells were subjected to HHP treatment at 250 MPa as described above. Following HHP treatment, the culture was diluted 5-fold in fresh recovery medium and incubated at 37°C with vigorous shaking. After incubation for the desired time, cells were collected by centrifugation (5,000 × g, 5 min, and 25°C). Then total cellular RNA was prepared as previously described (49). First-strand cDNA was synthesized using PrimeScript RT reagent kit with genomic DNA (gDNA) eraser (TaKaRa Bio, Japan) with random-hexamer primers. Real-time PCR was performed using Thunder Bird SYBR qPCR mix (Toyobo, Japan) with an ABI 7300 real-time PCR system (Applied Biosystems, USA). The oligonucleotides used for qPCR are listed in Table 3. 16S rRNA was used for normalization of RNA levels.

TABLE 3.

Primers used in this study for RT-qPCR

Forward primer	Sequence (5′→3′)	Reverse primer	Sequence (5′→3′)
abhF	CCGATTGAGTTGAGACGGG	abhR	CCCTCAGGGCTTAGCGTAA
bfmBAAF	TCCGGGTTTGCAAAAGCA	bfmBAAR	TCTCCATACGTCCCGCAA
csbDF	ACAAAATGAAGGGCGGCT	csbDR	TGGATTTCGCCTTTTGCC
groESF	GCCATTAGGTGATCGCGT	groESR	GAACCAGCTGCCACGATT
gsiBF	TGAGCAGAGAAGAAGCAGG	gsiBR	TCGTTGTTGCGGGCGTTT
infCF	TCGGACAAAATGGCGACC	infCR	CAGGCGGTTTTGCATTCG
mrgAF	ATCACAGACGGCGGAAAC	mrgAR	TCCGACAAACAAGTCCGC
mtlDF	GAACCGAATGCACTGGCT	mtlDR	GAATGCCGCACACCTCTT
pdhAF	CCGCTTGCTGTATACGCA	pdhAR	AGGATCGTCACCAGCCAT
pstAF	CTGACAACACCGCGTCTT	pstAR	AGAAATCACCAGCACGGC
tigF	CAAAGGCCTGGGCATTGA	tigR	CAGCTTTTCCGCCTTCGA
tuaGF	ACAGCTTAATGGCGCAGG	tuaGR	TCATCACCGTCAAGCAGC
ydbNF	AAACATGTCAGGAGGGCG	ydbNR	TGCGGTCGATCCTGTCAT
yfiYF	TGCTGAATCTTTGGCCGG	yfiYR	TTGTTTGTCTGGTCGCCC
yflTF	GCAGAAAATGGGTGTTGCG	yflTR	TTGGCTCCGATCGTGTTG
ykuNF	TGTCATGCTGCAAGAGGC	ykuNR	AACCTCTCGCAAACGCTC
ykzIF	TGCTGTCTGGAGACTAGAGG	ykzIR	TCCATTCTTTTCTAAGTCGC
yqgZF	TCCAAGCTGCACGTCATG	yqgZR	AGTCTGGCTTCTTGTCGC
ysbAF	TCGCTTGGTGTGATGCAG	ysbAR	CGCTTCGGTTTTGCGTTT
yugIF	CAAGCTGCGCCTGAGAAA	yugIR	GGTCTTTGCGGTTGGACA
ywsBF	ACTTGCGGCTGGAATGAC	ywsBR	ATGCGGAAGGCTTCGTTC
acpPF	AGATCGCCTTGGCGTTGA	acpPR	TGTAGTTCACAGCGTCGC
citBF	GCGCGAAACTTGTCGGCAA	citBR	TGCAAGCGGCAGTTCAGCA
clpEF	ACGAACTGGCGGCAAAAC	clpER	TGTTCGTCTGCCTGCAGT
cspBF	TCTCTGCTATTCAAGGCG	cspBR	ACGTTAGCAGCTTGTGGTC
cspDF	TGAAGTTGAAGGCGGAGAC	cspDR	AGAAGCTTGAGGTCCACGA
feuAF	TCTCCGGCAAATTCCCTGA	feuAR	GTGCCTGCTGTGCTGATT
hagF	ATGGGTGCTGACGCTCTT	hagR	ACCGCACCAAGCTTAGCA
hbsF	TGCGTGAACGTTCTGCAC	hbsR	TTTTCCGGCAACTGCGTC
infAF	CACTGCCAAACGCGATGT	infAR	ACGTAATCCTGCCACGAGT
rplAF	ACGACGTCTCTGAAGCAGT	rplAR	CGAAAACGAGAACGCGCT
rplJF	AACGCGATCGCATTCAGC	rplJR	AGCCTTCGCGTGATGGAA
rplSF	GTTCCGTCCTGGTGACACT	rplSR	TCGCTGATTCCACCACCA
yqgAF	GGACATTGGCCCGCAATA	yqgAR	ACCTACTGCCCAAGGTGT
ahpCF	ATCGATCCAGACGGCGTT	ahpCR	TCCCATTTAGCCGGGCAA
appCF	TCTGCTGTTTTCGCCCCT	appCR	AAAGAGACACGCGCTCCA
citZF	ACCATGAGCTGAACGCGT	citZR	ACGGCTCAGCGTTTTCCA
gapBF	AAAACAGCTGCCTTGGCG	gapBR	TTCCCGGAGCGGTCAAAA
katAF	CGCGCACGCAACAAAGTA	katAR	TTTTGGACCGCCGAAGCT
mntAF	GCCGGAAAAAGCGCTGAT	mntAR	TGCAACCCGCTGTACAGA
pckAF	AGAAGCGTTTGAGCGGCT	pckAR	CAGCTGCCGCGCAAATAA
rapAF	AGGATTGGAAATCGCCCGT	rapAR	TCATGCGCCAGCTCTTCA
rbsAF	AACGCACGCACGCCATTT	rbsAR	CGTTTGGCGCCAACATCCA
srfAAF	TCATTGGCACAGCAGCGA	srfAAR	TGTCTCGGCGTGCTGTTT
yhzCF	GTCCCGCAACACGCAAAA	yhzCR	TGGCAAGCCGACGGTTAT
ytxHF	TCATCGGTGCCACTACAGC	ytxHR	TTCCTTCGCATCTGCCGT
yvyDF	CGATGACAGATCTGGCGCT	yvyDR	AGCCAAGACCGTTTGCCA
yydFF	CACTAACAATGAGACTGTG	yydFR	ACTTCCAAGAATCCAGCGA
rpsJF	GTATCTGGTCCGATTCCG	rpsJR	GTGGTGTTGGGTTCACAAT

Sucrose density gradient centrifugation.

Cells (>100 AU [milliliter of culture × OD₆₀₀]) were collected by centrifugation (5,000 × g, 25°C, and 10 min) and washed with 20 mM Tris-HCl buffer solution (pH 7.0). Sucrose density gradient centrifugation analysis was performed as previously described (22). At least two independent experiments were performed to assess the reproducibility.

Accession number(s).

All microarray data have been deposited in a MIAME (Minimum Information About a Microarray Experiment)-compliant database, the National Center for Biotechnology Information Gene Expression Omnibus (GEO Series accession number GSE126203), as detailed on the FGED Society website (http://www.fged.org/projects/miame/).