More than 95% of individuals in populations of Bacillus subtilis spores killed approximately 95% by hydrogen peroxide or hypochlorite germinated with a nutrient, although the germination of the treated spores was slower than that of untreated spores. The slow germination of individual oxidizing agent-treated spores was due to: (i) 3- to 5-fold longer lag times (Tlag) between germinant addition and initiation of fast release of spores' large dipicolinic acid (DPA) depot (ii) 2- to 10-fold longer times (ΔTrelease) for rapid DPA release, once this process had been initiated; and (iii) 3- to 7-fold longer times needed for lysis of spores' peptidoglycan cortex. These results indicate that effects of oxidizing agent treatment on subsequent spore germination are on: (i) nutrient germinant receptors in spores' inner membrane (ii) components of the DPA release process, possibly SpoVA proteins also in spores' inner membrane, or the cortex-lytic enzyme CwlJ; and (iii) the cortex-lytic enzyme SleB, also largely in spores' inner membrane. This study further indicates that rapid assays of spore viability based on measurement of DPA release in spore germination can give false-positive readings.
Significance and Impact of the Study
This work shows that with Bacillus subtilis spore populations in which approximately 95% of individual spores were killed by several oxidizing agents, >95% of the spores in these populations germinated with nutrients, albeit slowly. This is important, as assay of an early germination event, release of dipicolinic acid, has been suggested as a rapid assay for spore viability and would give false-positive readings for the level of the killing of oxidizing agent-treated spore populations. Analysis of the germination kinetics of multiple individual untreated or oxidizing agent-treated spores also provides new information on proteins damaged by oxidizing agent treatment, and at least some of which are in spores' inner membrane.
Spores of some Bacillus species can cause food spoilage or foodborne disease, and Bacillus anthracis spores are a potent biological weapon (Setlow and Johnson 2012). As a consequence, there is much interest in developing methods to kill such spores and to understand the mechanism(s) of spore killing as well as potential limitations in various spore killing regimens. Spores can be killed by high temperatures and UV or γ-radiation, but some regimens for spore inactivation or decontamination use toxic chemicals. Oxidizing agents in particular are used for spore killing, in particular in applied settings, where agents such as hydrogen peroxide (H2O2) and sodium hypochlorite are used (Al-Adham et al. 2013). The mechanism whereby these latter two agents kill spores has been reasonably well studied, and despite the potential genotoxicity of these agents, they do not kill spores by DNA damage (Young and Setlow 2003; Cortezzo et al. 2004; Setlow 2006; Setlow and Johnson 2012). Interestingly, spores treated with either of these agents often germinate, especially if the treatment is not too harsh (Cortezzo et al. 2004). However, in most cases, the germinated-treated spores do not outgrow and exhibit minimal metabolism, suggesting that one or more key spore proteins have been severely damaged by these oxidizing agents, as has also been suggested for wet-heat-treated spores of Bacillus species (Coleman et al. 2007, 2010; Coleman and Setlow 2009). Unfortunately, the fact that these oxidizing agent-treated spores do germinate is a potential problem, even if the germinated spores are actually dead, since some rapid assays of spore viability monitor changes that take place in spore germination, in particular, the release of the spore core's large depot of pyridine-2,6-dicarboxylic acid (dipicolinic acid (DPA)) (Yung et al. 2007; Yang and Ponce 2011). Thus, there is the potential for false-positive readings of spore viability with such assays in spore populations.
As noted above, the germination of populations of spores treated with oxidizing agents has been reasonably well studied. However, analysis only of the germination of spore populations can obscure important information, as germination of individuals in spore populations is extremely heterogeneous (Setlow 2003; Setlow et al. 2012). In recent years a number of techniques have been developed for the analysis of the germination of large numbers of individual spores in populations (Zhang et al. 2010a,b; Kong et al. 2011). Such techniques have also been applied to the analysis of the germination of Bacillus subtilis spores killed by wet heat, and this analysis has revealed steps in the germination pathway that are especially sensitive to wet heat (Wang et al. 2011a,b, 2012). In the current work, we have extended this work to compare the kinetic parameters of the germination of B. subtilis spore populations that are either untreated, or treated with H2O2 or hypochlorite to kill approximately 95% of the population, with spore killing measured by the ability to form colonies on nutrient plates. The results of this comparison have given new insight into the effects of oxidizing agents on spores.
Results and discussion
Germination of untreated and oxidizing agent-treated spore populations
Previous work has shown that most individual spores in populations killed approximately 90% with H2O2 or hypochlorite germinated to a significant extent, when germination was assessed by release of DPA (Cortezzo et al. 2004). However, the germinated-treated spores did not exhibit significant metabolism or outgrowth. Questions that were not previously addressed, however, include: (i) do the oxidizing agent-treated spores that did not germinate in the observation periods used previously eventually germinate (ii) if these treated spores do germinate slowly, what step or steps in the germination pathway are slowed by oxidizing agent treatment; and (iii) can the answers to these questions give information on effects of H2O2 and hypochlorite on spores.
To initiate attempts to answer the questions posed above, we treated B. subtilis spores with either H2O2 or hypochlorite to get approximately 95% killing (93–97%, with killing defined as the inability of a spore to form a colony on a nutrient plate), inactivated the oxidizing agent, and compared the nutrient germination of individual untreated spores with that of the treated spores, all as described in Methods. As measured by monitoring DPA release, approximately 95% of spores in untreated populations germinated with the nutrient germinant L-valine in approximately 40 min, and the germination behaviour of the untreated spores was unaffected by the procedures used to inactivate either H2O2 or hypochlorite prior to subsequent analysis (Fig. 1a,b). Notably, a large percentage of spores in populations that had been killed approximately 95% by H2O2 or hypochlorite also germinated, but took much longer to reach this level of DPA release for the treated spores, up to 300 min for the hypochlorite-treated spores and approximately 200 min for the H2O2-treated spores (Fig. 1a,b, and data not shown, and see below). The germination of spores in populations killed >99·9% by hypochlorite or H2O2 was even slower, as only 20–30% of the treated spores germinated in 300 min (data not shown). Together these findings indicate that treatments with H2O2 or hypochlorite killing the great majority of spores in populations does not completely inactivate any key spore germination component, but only slows the action of one or more of these components. However, if the treatment with oxidizing agents is harsh enough, then spore germination, at least on a reasonable time scale, is abolished, and this is also the case for spores treated with wet heat (Coleman et al. 2007, 2010).
Analysis of the germination of multiple individual untreated or oxidizing agent-treated spores
While the results with spore populations were useful, previous work has shown that the heterogeneity in spore germination can obscure many of the important kinetic parameters of spore germination (Zhang et al. 2010a,b; Kong et al. 2011; Setlow et al. 2012). Consequently, we monitored the germination of large numbers of individual untreated and H2O2 and hypochlorite-treated spores in spore populations that had been killed 93 and 97%, respectively (Fig. 2). The results of this experiment were relatively similar to those obtained following the germination of untreated and H2O2 and hypochlorite-treated spore populations by monitoring DPA release, as the oxidizing agent-treated spores germinated slower even though ≥ 95% of the treated spores did eventually germinate.
Analysis of the germination curves of multiple individual spores by DIC microscopy (Fig. 3a–c) showed that as expected, the curves for the germination of both the untreated and oxidizing agent-treated individual spores were quite heterogeneous. Previous work has identified a number of key time points in the germination of individual spores of Bacillus species (Zhang et al. 2010a,b; Kong et al. 2011), as shown for one spore each in Fig. 3a–c. Upon addition of nutrient germinants to an individual spore, there is a variable lag period with no obvious change in spores, as spore DPA content remains relatively constant although can decrease somewhat very slowly. However, at some point, rapid DPA release begins, with the time of initiation of this event termed Tlag. The point when this rapid DPA release is completed is termed Trelease, and almost all DPA is released in ΔTrelease that is equal to Trelease–Tlag. The release of spores' DPA and its replacement in the spore core by water is accompanied by a decrease in a spore's DIC image intensity of approximately 75%. Following Trelease, there is a further decrease in spores' DIC image intensity due to the hydrolysis of the spores' peptidoglycan (PG) cortex by cortex-lytic enzymes (CLEs), and associated core swelling and water uptake. This process ends at Tlys, and ΔTlys, the time for spore cortex hydrolysis equal to Tlys–Trelease.
It was clear even from a cursory examination of the data for multiple individual untreated and oxidizing agent-treated spores (Fig. 3a–c) that the oxidizing agent-treated spores, in particular, those treated with hypochlorite, had much longer Tlag values than did untreated spores. Compilation of the kinetic parameters for the germination of 163–300 individual untreated or oxidizing agent-treated spores showed that as expected, Tlag values increased 3- to 5-fold with agent-treated spores compared to untreated spores, values for ΔTrelease were 2- to 10-fold higher in the treated spores, and ΔTlys values were 3- to 7-fold higher in treated spores (Table 1).
Table 1. Kinetic parameters of germination of untreated and oxidizing agent-treated B. subtilis sporesa
Untreated or oxidizing agent-treated B. subtilis PS533 spore populations were heat shocked and germinated, and kinetic parameters for the germination of multiple individual spores were extracted from the data as shown in Fig. 3a–c. Spore populations treated with H2O2 or hypochlorite were killed 93 and 97%, respectively. Values for the kinetic germination parameters shown are the average values from the individual spores analysed (all of which germinated) ± the standard deviations.
The germination of individual spores was followed for 60 min (untreated spores), 140 min (H2O2-treated spores) or 300 min (hypochlorite-treated spores).
The results noted above indicate that H2O2 and hypochlorite treatment caused significant damage to one or more spore germination proteins, and this likely indicates some targets for these agents in spores even if these targets are themselves not the lethal targets for these oxidizing agents. One effect of H2O2 and hypochlorite treatment on spore germination kinetics was the 3- to 5-fold increase in Tlag values. Previous work has shown that two major variables determining Tlag values in spore germination are spores' levels of: (i) nutrient germinant receptors (GRs) that recognize and respond to specific nutrient germinants and (ii) the GerD protein essential for proper GR function (Zhang et al. 2010a; Wang et al. 2011b; Ghosh et al. 2012; Ramirez-Peralta et al. 2012). GRs and GerD are located in spores' inner membrane (IM), and previous work indicates that most oxidizing agents kill spores by damaging the IM such that when spores germinate, this membrane ruptures (Setlow 2003; Cortezzo et al. 2004). While the specific spore IM damage caused by oxidizing agent treatment is unknown, it has been suggested to be damage to IM proteins, and perhaps damage to GRs and GerD is a reflection of this.
H2O2 and hypochlorite treatment also resulted in significant increases in times for the rapid DPA release, ΔTrelease, during spore germination. These increases cannot be due to damage to GRs or GerD, as fluctuation in these proteins' levels does not alter ΔTrelease values (Zhang et al. 2010a; Wang et al. 2011b). One group of proteins damage to which might result in increased ΔTrelease values is the SpoVA proteins involved in DPA movement across the spores' IM in germination (Setlow 2003; Wang et al. 2011b; Setlow et al. 2012). Another protein important in determining ΔTrelease values is the CwlJ protein, one of Bacillus spores' two redundant CLEs that degrade spores' PG cortex during germination, as loss of CwlJ results in increases in ΔTrelease values up to 15-fold (Peng et al. 2009; Setlow et al. 2009), and at present, we cannot decide between damage to SpoVA proteins or CwlJ as responsible for the increased ΔTrelease values for oxidizing agent-treated spores, although it appears clear (see below) that at least some CwlJ has been inactivated by oxidizing agent treatment.
Finally, as ΔTlys values also increased markedly in oxidizing agent-treated spores, and either CwlJ or the other redundant CLE, SleB, alone is sufficient for normal values of ΔTlys during spore germination, oxidizing agent treatment must have damaged both CwlJ and SleB significantly. Interestingly, SleB is also largely in spores' IM (Chirakkal et al. 2002), although CwlJ is located at the spore coat-cortex boundary (Setlow 2003) where it will be more accessible to oxidizing agents than IM proteins.
While analysis of spore germination kinetics has given new insight into damage caused by oxidizing agent treatment of spores, it is also important to appreciate that these treated spores can germinate well even if most of these spores are actually dead as measured by the ability of a spore to germinate, outgrow and ultimately form a colony. As a consequence, another major conclusion from the current work is that assays of spore germination measuring DPA release are inappropriate for rapid routine measurement of the viability of oxidizing agent-treated spore populations, as such assays can readily give large numbers of false-positive results, as they also will for spores killed by wet heat which also can germinate, albeit slowly, if the wet-heat treatments of spore populations have not been too harsh (Wang et al. 2011a, 2012).
Materials and methods
The B. subtilis strain used in this work was PS533, a prototrophic 168 derivative that carries plasmid pUB110 encoding kanamycin resistance (10 mg l−1) (Setlow and Setlow 1996). Spores of this strain were prepared on 2x Schaeffer's-glucose plates that were incubated for approximately 3 days at 37°C; following further incubation for 2–3 days at 23°C, the spores were scraped from the plates and purified as described previously (Nicholson and Setlow 1990; Paidhungat et al. 2000), and stored in water at 4°C. No reduction in spore viability was seen in the approximately 3 months during which experiments with these spores were carried out. All spores used in this work were free (>98%) of growing or sporulating cells, germinated spores and cell debris as seen by phase contrast microscopy.
Treatment of spores with H2O2 and sodium hypochlorite
Spore treatment with H2O2 or sodium hypochlorite at 23°C to determine killing curves was essentially as described previously (Cortezzo et al. 2004), with spores at an optical density at 600 nm (OD600) of 1 in either 50 mmol l−1 KPO4 buffer plus 2·5 mol l−1 H2O2 or in 100 mg l−1 sodium hypochlorite prepared in water (measured pH of 11). At various times, aliquots were diluted 1/100 in 1 ml of either 50 mmol l−1 KPO4 buffer (pH 7·4) with 500 units of beef liver catalase (H2O2) or 10 g l−1 Na2S2O4 (hypochlorite) and incubated at least 15 min at 23°C to inactivate the oxidizing agents as described previously (Cortezzo et al. 2004), and 10 μl aliquots of serial dilutions in sterile water were spotted on LB medium agar plates containing kanamycin (10 mg l−1). The plates were incubated for 24–36 h at 37°C and colonies were counted. Incubation for longer times gave no further increases in colonies. A dead spore was then defined as a spore that does not give rise to a colony in this assay.
To obtain large amounts of spores treated with NaOCl, 20 ml of spores was incubated, and at times determined in preliminary experiments that gave approximately 95% killing of spore populations, aliquots were diluted and spore viability was analysed, all as described above. The remainder of the culture was centrifuged, the pellet suspended in 2 ml of 10 g l−1 Na2S2O4, incubated for approximately 15 min at 23°C, washed 3x with water and stored in water at 4°C. Large amounts of H2O2 treated spores were prepared by incubation of spores at 23°C in 5 ml of 50 mmol l−1 KPO4 (pH 7.4) with spores at an OD600 of 5. At various times, 1-ml aliquots was placed into a capped test tube, 10 μl of beef liver catalase (approximately 500 units) added on the side of the tube, and the tube was capped and mixed. After incubation for approximately 15 min at 23°C, the sample was washed twice by centrifugation, suspended in 1 ml cold sterile water, and spore viability was determined as described above. The various treated spore samples were stored at 4°C.
Monitoring the germination of spore populations
Germination of spore populations was preceded by a heat shock (65°C; 30 min) of spores in water followed by cooling on ice. Germination of heat-shocked spores was at an OD600 of 0.5 in 200 μl of 25 mmol l−1 K-Hepes buffer (pH 7·4)-10 mmol l−1 L-valine-50 μ mol l−1 terbium (Tb+3) chloride, and spore germination was followed in a multiwell fluorescence plate reader, monitoring DPA release by the high fluorescence of Tb-DPA as previously described (Yi and Setlow 2010). The total amount of DPA in spores was determined by its fluorescence with Tb after spores' DPA was released by boiling for 15 min as described previously (Yi and Setlow 2010).
Monitoring the germination of individual spores
Analysis of the germination of multiple individual spores by DIC microscopy was as described previously (Zhang et al. 2010a,b; Kong et al. 2011; Wang et al. 2012). Briefly, heat-shocked spores prepared as described above (1 μl; approximately 108 spores ml−1 in water) were spread on the surface of a microscope coverslip that was then dried in a vacuum desiccator for 5–10 min. The coverslips were then mounted on and sealed to a microscope sample holder kept at 37°C. The DIC microscope was set such that the polarizer and analyser were crossed, and thus, the DIC bias phase was zero. After adding preheated germinant/buffer solution (10 mmol l−1 L-valine-25 mmol l−1 K-Hepes buffer (pH 7·4)) to spores on the coverslips, a digital CCD camera (16 bit, 1,600 by 1,200 pixels) was used to record the DIC images at a rate of 1 frame per 15 s for 60–120 min. These images were analysed with a computation program in Matlab to locate each spore's position and to calculate the averaged pixel intensity of an area of 40 × 40 pixels that covered the DIC image of the whole individual spore. The DIC image intensity of each individual spore was plotted as a function of the incubation time (with a resolution of 15 s); the initial intensity at T0 (the first DIC image recorded after the addition of the germinant) was normalized to 1, and the intensity at the end of measurements was normalized to zero. Invariably, the latter value had been constant for ≥10 min at the end of measurements.
From the time-lapse DIC image intensity, we can determine the time of completion of the rapid fall of approximately 75% in spore DIC image intensity, which is concomitant with the time of completion of spore DPA release (this time is defined as Trelease). DPA release kinetics during germination of individual spores are described by the parameters Tlag and ΔTrelease, where Tlag is the time between the mixing of spores with germinants and the initiation of most DPA release, and ΔTrelease = (Trelease–Tlag). We also defined the additional germination parameters, Tlys and ΔTlys, where Tlys is the time when hydrolysis of the spore's peptidoglycan cortex is completed as determined by the completion of the fall in the spore's DIC image intensity, and ΔTlys = (Tlys–Trelease).
This communication is based on work supported by the U.S. Army Department of Defense Multidisciplinary University Research Initiative through the U.S. Army Research Laboratory and the U.S. Army Research Office under contract number W911F-09-1-0286.