Mechanisms of killing of Bacillus subtilis spores by Decon and Oxone^TM, two general decontaminants for biological agents

Abstract

Aims:  To determine the mechanisms of Bacillus subtilis spore killing by and resistance to the general biological decontamination agents, Decon and Oxone^TM.

Methods and Results:  Spores of B. subtilis treated with Decon or Oxone^TM did not accumulate DNA damage and were not mutagenized. Spore killing by these agents was increased if spores were decoated. Spores prepared at higher temperatures were more resistant to these agents, consistent with a major role for spore coats in this resistance. Neither Decon nor Oxone^TM released the spore core's depot of dipicolinic acid (DPA), but Decon- and Oxone^TM-treated spores more readily released DPA upon a subsequent normally sublethal heat treatment. Decon- and Oxone^TM-killed spores initiated germination with dodecylamine more rapidly than untreated spores, but could not complete germination triggered by nutrients or Ca²⁺-DPA and did not degrade their peptidoglycan cortex. However, lysozyme treatment did not recover these spores.

Conclusions:  Decon and Oxone^TM do not kill B. subtilis spores by DNA damage, and a major factor in spore resistance to these agents is the spore coat. Spore killing by both agents renders spores defective in germination, possibly because of damage to the inner membrane of spore.

Significance and impact of study:  These results provide information on the mechanisms of the killing of bacterial spores by Decon and Oxone^TM.

Introduction

A variety of biological agents may be present in a biological spill or used in a biological warfare or bioterrorism attack. Consequently, there is a need for easy to use treatments that can destroy many different biological agents. Spores of Bacillus anthracis are one potential biowarfare or bioterrorism agent, as has been shown all to well by recent events in the United States. These spores as well as those of other Bacillus spp. are metabolically dormant and very resistant to many treatments that kill growing cells, including harsh chemicals, UV radiation and wet and dry heat (Russell 1990; Bloomfield and Arthur 1992; McDonnell and Russell 1999; Nicholson et al. 2000; Setlow et al. 2000). Given this high level of spore resistance, usually higher than that of other biowarfare agents, a major test of any general biological decontamination agent is the ability of the agent to kill spores.

What makes spores so resistant to the various treatments noted above? The first line of defence of the spore against toxic chemicals is the thick proteinaceous coat (Russell 1990; Bloomfield 1999; McDonnell and Russell 1999; Nicholson et al. 2000; Setlow et al. 2000). A second protective barrier is the inner membrane that has very low permeability to small hydrophilic molecules, and thus provides further protection against toxic chemicals (Setlow 2000). The spore core has an extremely low water content and this is of major importance in spore resistance to wet heat (Gerhardt and Marquis 1989). In addition, in the core, the α/β-type small, acid-soluble spore proteins (SASP) saturate and protect the spore DNA from damage caused by heat, UV radiation and some genotoxic chemicals such as nitrous acid (Setlow 2000; Tennen et al. 2000). Damage to dormant spore DNA can also be repaired by a number of mechanisms that become operative early in spore germination and outgrowth (Setlow and Setlow 1996; Nicholson et al. 2000; Setlow 2000).

Chlorine dioxide and hypochlorite are two agents that are effective in killing spores of Bacillus spp., in particular spores of B. subtilis, a species whose spores are often used as a non-pathogenic surrogate for B. anthracis spores (Young and Setlow 2003). Indeed, both chlorine dioxide and hypochlorite were used in the clean up after B. anthracis spores contaminated the Hart Senate Office Building in Washington, DC, USA. However, there are significant concerns over the widespread use of these two agents, as hypochlorite solutions are corrosive and chlorine dioxide is a hazardous chemical (http://www.inchem.org/documents). An ideal decontaminant should kill its biological target rapidly, have no or minimal deleterious effects on the environment or associated materials, be uncomplicated in its application and be easily removed after use. Two general decontamination agents that have recently been introduced are Decon and Oxone^TM (DuPont Chemicals, Wilmington, DE, USA). Decon (MDF-200) is a three-part formulation consisting of surfactants, fatty alcohols, hydrogen peroxide and a propellant. This agent was developed by Sandia National Laboratories (http://www.sandia.gov) and can be deployed as a foam, liquid or fog. Oxone^TM, a mild commercial oxidizer, has potassium peroxymonosulphate as the active ingredient. This agent can be used in combination with a silica gelling agent to decontaminate biological spills (Rabner and McGuire 2002). Both Decon and Oxone^TM are effective not only against biological agents but also against a variety of chemical warfare agents (http://www.sandia.gov) (Rabner and McGuire 2002). While these two general decontaminants show great promise, little is known about the mechanism whereby these agents kill spores. For example, do these agents truly kill spores or are the spores simply rendered superdormant and perhaps remain capable of being revived at some later date. Consequently, in this work we report studies on the mechanism of killing of spores of B. subtilis by Decon and Oxone^TM.

Materials and methods

Strains and spore preparation

The B. subtilis strains used in this work are isogenic and are derived from strain 168. The wild-type strain is PS533, containing a kanamycin resistance marker on plasmid pUB110. Strain PS578 (termed α⁻β⁻) differs from PS533 in the absence of the genes coding for the two major α/β-type SASP of B. subtilis, sspA and sspB (Setlow and Setlow 1996). Strains PS2318 (recA) and PS2319 (α⁻β⁻recA) are derivatives of PS533 and PS578, respectively, and also carry both chloramphenicol and erythromycin resistance markers, one of which inactivates the recA gene responsible for much DNA repair in B. subtilis (Yasbin et al. 1993; Setlow and Setlow 1996). Strains PS3394 (cotE) and PS3395 (α⁻β⁻cotE) are derived from strains PS3328 and PS3329, respectively (Paidhungat et al. 2001), and contain plasmid pUB110. Strain PS3379 is isogenic with PS533 but lacks pUB110 and carries the luxAB genes from Vibrio harveyi under the control of the forespore-specific sspB promoter (Loshon et al. 2001), and strain PS2307, a derivative of strain PS832, has a mutation in the cwlD gene and forms an altered spore cortex that is not degraded during spore germination (Popham et al. 1996).

Spores of all strains were prepared on 2xSG medium agar plates without antibiotics at 24°C, 37°C and 46°C, and were cleaned and stored in water as described (Nicholson and Setlow 1990; Paidhungat et al. 2000). Spores prepared at 37°C were used for all experiments except the one comparing resistance of spores prepared at different temperatures. All spore preparations used in this work were free (>98%) of cell debris, germinated spores and growing cells.

Treatment of spores and growing cells with Decon and Oxone^TM

To prepare Decon (Modec, Inc., Denver, CO, USA), 0·5 ml of Decon  1 (surfactants and fatty alcohols) and 0·5 ml of Decon  2 (hydrogen peroxide) were mixed with 20·4 μl of Decon  3 (a propellant) giving a solution of pH 10. Spores (1 ml) at an optical density at 600 nm (O.D._600 nm) of 1 (ca 1·5·10⁸ spores per millilitre) were treated at room temperature (ca 24°C) with freshly prepared 10–30% Decon (diluted in water). The reaction was terminated with catalase as described (Setlow and Setlow 1993), and 7 ml water was added followed by centrifugation at 4°C for 10 min at 9770 g. The spores were then washed with 10 ml water and recentrifuged. After addition of 1 ml water, the sample was sonicated for 15 s to disperse the spore pellet and the suspension was diluted with water to an O.D._600 nm of 0·1, which was diluted serially in sterile phosphate-buffered saline (PBS) [50 mmol l⁻¹ potassium phosphate (pH 7·4) and 100 mmol l⁻¹ NaCl], plated on Luria–Bertani (LB) medium (Sambrook et al. 1989). Plates containing one appropriate antibiotic, incubated at 30°C for 24–48 h and colonies were counted. To determine spore killing by the individual components of the complete Decon formulation (termed Decon), Decon  1, Decon  2 and Decon  3 were used alone, as well as hydrogen peroxide at the concentration of hydrogen peroxide in Decon  2, following the same protocol.

Oxone^TM (Aldrich, Milwaukee, WI, USA) was prepared as a 320 g l⁻¹ solution in water, filter sterilized and used at 100–300 g l⁻¹. The pH of Oxone^TM solutions was 1·1–1·3, but incubation in HCl at pH 1·1 gave no spore killing over the time periods used for Oxone^TM killing (data not shown). Spores incubated for various time periods with Oxone^TM were diluted 10-fold in 10 g l⁻¹ sodium thiosulphate, incubated for 10 s, diluted further in sterile PBS, plated and plates incubated as described above for the Decon-treated spores. All killing experiments were performed at least twice using at least two different spore preparations with similar results. The variation in the slopes of killing curves was ≤25% for individual spore preparations in replicate experiments. To obtain larger quantities of spores killed with the various agents, spores at an O.D._600 nm of 20 (ca 3·10⁹ spores per millilitre) were treated with Decon and Oxone^TM as described above, but the Oxone^TM-treated spores were neutralized with 100 g l⁻¹ sodium thiosulphate. After samples were taken for determination of spore killing, spores were washed twice with PBS and suspended in water. In one experiment, spores were treated with 2·5 g l⁻¹ sodium hypochlorite (pH 11) with or without 2% Decon  3, and spore killing was measured as described (Young and Setlow 2003).

Log phase cells of strain PS533 were grown at 37°C in 2xYT medium (Paidhungat et al. 2000) without antibiotics and harvested at an O.D._600 nm of 1 (ca 2·10⁸ cells per millilitre). After centrifugation the cells were washed with PBS, suspended in PBS at an O.D._600 nm of 20, treated with either Decon or Oxone^TM and cell killing determined as described above.

Analysis of mutagenesis by Decon and Oxone^TM

Levels of auxotrophic or asporogenous mutations generated in spores were determined by transferring colonies formed by spores that survived Decon or Oxone^TM treatment onto minimal medium or sporulation medium plates, and after incubation at 37°C colonies were examined for mutations as described (Fairhead et al. 1993).

Decoating and recovery of decoated spores

Spores at an O.D._600 nm of 10 (ca 1·5·10⁹ spores per millilitre) were decoated by 30 min incubation at 65°C in 100 mmol l⁻¹ NaOH, 100 mmol l⁻¹ NaCl, 100 mmol l⁻¹ dithiothreitol and 5 g l⁻¹ sodium dodecylsulphate followed by 10 water washes (Bagyan et al. 1998). Decoated spores were treated with 25 mg l⁻¹ lysozyme in a hypertonic medium as described (Popham et al. 1996), and lysis of the spore cortex by lysozyme was monitored by observation in the phase contrast microscope. In one experiment decoated Decon- or Oxone^TM-treated spores were plated on LB plates containing lysozyme at 0·5 mg l⁻¹.

Assays of dipicolinic acid and spore cortex fragments

Pyridine-2,6-dicarboxylic acid (dipicolinic acid, DPA) was extracted from dormant and germinated spores and measured as described (Rotman and Fields 1967; Setlow and Setlow 1993). The DPA released from treated or untreated spores that were suspended in water and heated to 80 or 85°C for 30 min was also determined as described (Genest et al. 2002). The total hexosamine content of untreated and treated dormant and germinated spores, and of the supernatant fluid from germinated spores was determined as described (Popham et al. 1996; Tennen et al. 2000).

Spore germination

Spores were germinated at 37°C and an O.D._600 nm of 1 in either 2xYT medium (Paidhungat et al. 2000) with 4 mmol l⁻¹l-alanine or in 50 mmol l⁻¹ Tris–HCl (pH 8·4) with 8 mmol l⁻¹l-alanine. In some experiments, spores were incubated in 60 mmol l⁻¹ Ca²⁺-DPA in 50 mmol l⁻¹ Tris–HCl (pH 8) at room temperature to induce spore germination (Paidhungat et al. 2000). Spores were also germinated at 50°C and an O.D._600 nm of 1 in 1 mmol l⁻¹ dodecylamine (Sigma, St. Louis, MO, USA) and 20 mmol l⁻¹ KPO₄ (pH 7·4), and DPA release was monitored by centrifugation of 1 ml samples of germinating spores and measuring the O.D._270 nm of the supernatant fluid using a Genesys  5 spectrophotometer as described (Setlow et al. 2003).

Measurement of light production from spores carrying the luxAB genes

For analysis of light production during germination of spores carrying the luxAB genes from V. harveyi, spores at an O.D._600 nm of 1 were germinated at 37°C in 2xYT medium plus 4 mmol l⁻¹l-alanine. At various times 500 μl aliquots were mixed with 500 μl aliquots of medium containing 100 mg l⁻¹ dodecanal, and a Turner TD-20/20 luminometer (Turner Designs, Sunnyvale, CA, USA) was used to measure light production as described (Hill et al. 1994; Ciarciaglini et al. 2000; Loshon et al. 2001).

Core wet density of spores

The core wet density of decoated dormant spores or intact germinated spores was determined by equilibrium density centrifugation in Nycodenz gradients (Lindsay et al. 1985; Popham et al. 1996; Melly et al. 2002a).

Staining and microscopic analysis of spores

The BacLight bacterial viability stain (Molecular Probes, Eugene, OR, USA) and 1 μg l⁻¹ 4′,6′-diamino-2-phenylindole (DAPI) were used to stain dormant and germinated spores (Melly et al. 2002a). The percentage of spores that became phase dark during incubation under germination conditions was determined as described (Wyatt and Waites 1975; Tennen et al. 2000).

Results

Spore killing by Decon and Oxone^TM

Spores of B. globigii are killed by Decon and 7 logs of killing in 1 h is claimed in literature from the Sandia National Laboratories (http://www.sandia.gov). We found that incubation of wild-type B. subtilis spores in 30% Decon gave 1 log of killing in 20 min, and spores of the α⁻β⁻ strain were killed at a similar rate (Fig. 1a). Spores that lacked the recA gene whose product is responsible for much DNA repair in B. subtilis (Yasbin et al. 1993), exhibited very similar Decon resistance to that of both wild-type and α⁻β⁻ spores (Fig. 1a). However, spores lacking both the α/β-type SASP and the recA gene were killed more rapidly (Fig. 1a). When used at 100%, Decon is composed of 49% Decon  1, 49% Decon  2 and 2% Decon  3. In contrast to the killing of ca 4 logs of wild-type spores in 50 min by 30% Decon, there was no killing (<10%) of wild-type spores upon incubation for 50 min in either 30% Decon 1, 30% Decon  2 or 2% Decon  3 [data not shown; note that these concentrations of the individual components are twofold (Decon  1 and 2) to threefold (Decon  3) higher than when complete Decon was used at 30%]. Decon  2 used at 50% or a comparable concentration of hydrogen peroxide (4%) also gave no spore killing (data not shown). Surprisingly, a 1:1 mixture of Decon  1 and 2 used at 30% but without Decon  3 did not kill spores effectively, indicating that Decon  3 may potentiate killing by Decon  1 and 2 (Fig. 1b). Subsequent work (see below) suggested that the potentiating effect of Decon  3 on spore killing is due to the effect of this agent on the spore coats.

Figure 1. Resistance of spores of various Bacillus subtilis strains to (a) Decon, (b) indvidual Decon components, or (c) Oxone^TM. (a, c) Spores of strain PS533 (wild-type) (•), PS578 (α⁻β⁻) (), PS2318 (recA) () and PS2319 (α⁻β⁻recA) (▾) were incubated in either (a) 30% Decon or (c) 300 g l⁻¹ Oxone^TM; or (b) spores of strain PS533 (wild-type) were incubated in 30% Decon (•) or 30% Decon without Decon 3 (); and spore survival determined as described in ‘Materials and Methods’

In a fumed silica gel, 0·8 N Oxone^TM was effective in killing B. globigii spores sprayed onto varnished wood, glass and fibre filler (Rabner and McGuire 2002). Oxone^TM also killed B. subtilis spores in water; however, a 1-log kill required >3 h, more than 10-fold longer than was required for similar killing by Decon (Fig. 1c). The α⁻β⁻ spores were slightly more sensitive to Oxone^TM than wild-type spores (Fig. 1c), as has been seen previously with a number of other oxidizing agents (Setlow and Setlow 1993; Loshon et al. 2001; Genest et al. 2002; Young and Setlow 2003). The recAα⁻β⁻ spores were similar to wild-type spores in their sensitivity to Oxone^TM, and the recA spores were slightly more sensitive (Fig. 1c).

If the mechanism of spore killing by Decon and Oxone^TM damage were to spore DNA, greater than or equal to fivefold increased sensitivity to these agents would be expected in spores lacking the α/β-type SASP and/or the recA gene, as has been seen with a number of genotoxic agents (Setlow and Setlow 1996; Setlow 2000; Tennen et al. 2000). Therefore, the results suggest that Decon and Oxone^TM do not kill spores by damaging spore DNA, although spores lacking both α/β-type SASP and the recA gene were killed up to threefold faster by Decon than were wild-type spores. To further examine effects of Decon or Oxone^TM on spore DNA, survivors of wild-type and α⁻β⁻ spores treated with these agents were analysed for auxotrophic and asporogenous mutations. However, treatment with neither agent caused any appreciable increase in mutations in the survivors (Table 1). In contrast to these results, genotoxic agents such as nitrous acid or alkylating agents generate mutations in 10–15% of the survivors of spores treated with these chemicals (Setlow et al. 1998; Tennen et al. 2000).

Table 1.  Levels of mutations in survivors of dormant spores treated with Decon or Oxone^TM*

Strain	Treatment	Killing (%)	Colonies examined (n)	Survivors with mutations (%)
				aux	spo	aux spo	Total
*Spores of various strains were treated, and spore killing and the levels of auxotrophic (aux) and asporogenous (spo) mutants in the survivors were determined as described in ‘Materials and Methods’.
Wild-type (PS533)	None	0	419	0·2	0	0	0·2
α^−β^− (PS578)	None	0	436	0·2	0	0	0·2
Wild-type (PS533)	Decon	99	413	0	0	0	0
α^−β^− (PS578)	Decon	99	430	0	0	0	0
Wild-type (PS533)	Oxone	94	388	0	0	0	0
α^−β^− (PS578)	Oxone	94	543	0	0·4	0	0·4

Exponentially growing B. subtilis cells are killed more than 1000-fold in 15 s by either 30% Decon or 300 g l⁻¹ Oxone^TM (data not shown). As a similar degree of spore killing by these agents requires hours, some features unique to spores must offer protection from Decon and Oxone^TM. One likely protective feature is the thick proteinaceous coat of the spore that is a major factor in spore resistance to many chemicals, including glutaraldehyde, hydrogen peroxide, iodine, ortho-phthalaldehyde, peroxynitrite, the superoxidized water Sterilox^®, hypochlorite and chlorine dioxide (Russell 1990; Bloomfield 1999; Riesenman and Nicholson 2000; Tennen et al. 2000; Loshon et al. 2001; Cabrera-Martinez et al. 2002; Genest et al. 2002; Young and Setlow 2003). The spore coats are also important in spore resistance to Decon and Oxone^TM, as chemically decoated spores exhibited >3 logs of killing in 1 min by either agent (Fig. 2a). Spores carrying a mutation in cotE that codes for a protein essential for spore coat assembly (Driks 1999) were also killed rapidly by either agent, but not as rapidly as chemically decoated spores (Fig. 2b). While chemically decoated spores were equally sensitive to Decon and Oxone^TM, cotE spores were significantly more sensitive to Oxone^TM. The killing curves for chemically decoated α⁻β⁻ spores or of α⁻β⁻cotE spores with either Decon or Oxone^TM were similar to those for the corresponding spores that contained the α/β type SASP (data not shown).

Figure 2. Effects of coat defects on spore resistance. Chemically decoated spores of strain PS533 (wild-type) (a) or spores of strain PS3394 (cotE) (b) were incubated in 10% (chemically decoated spores) or 15% (cotE spores) Decon (•) or 100 g l⁻¹ Oxone^TM (); or (c) intact spores of strain PS533 (wild-type) were incubated with hypochlorite without (•) or with () 2% Decon  3; and spore survival measured as described in ‘Materials and Methods’

The tremendous decrease in Decon resistance of spores with defective coats suggested that perhaps the propellant making up Decon  3 was potentiating spore killing by the ingredients in Decon  1 and 2 by causing or facilitating spore coat removal. This possibility was investigated by examining the ability of Decon  3 alone to potentiate spore killing by hypochlorite, as spore hypochlorite resistance is also extremely dependent on the spore coats (Young and Setlow 2003). Strikingly, spore killing by hypochlorite was greatly stimulated by Decon  3 (Fig. 2c), consistent with Decon  3 decreasing the ability of the spore coat to protect against hypochlorite and perhaps other chemicals.

Spores prepared over a wide temperature range vary in levels of some individual coat proteins, as well as in other characteristics (Melly et al. 2002b). These differences are reflected in the altered sensitivities of spores prepared at higher and lower temperatures to a variety of chemical agents (Melly et al. 2002b; Young and Setlow 2003), and this was also true of Decon and Oxone^TM. Spores prepared at 37 and 46°C were much more resistant to Oxone^TM than spores prepared at 24°C, although the difference in the Decon resistance of spores prepared at 24, 37 or 46°C was much smaller (Fig. 3a,b).

Figure 3. Resistance of Bacillus subtilis spores prepared at different temperatures to (a) Decon or (b) Oxone^TM. Spores of strain PS533 (wild-type) prepared at 24°C (•), 37°C () or 46°C () were incubated with (a) 30% Decon or (b) 300 g l⁻¹ Oxone^TM, and spore survival was measured as described in ‘Materials and Methods’

Properties of Decon and Oxone^TM treated spores

Spores treated with Decon or Oxone^TM were further examined for insights into the mechanisms of spore killing by these agents. Spore preparations killed 85–99% were examined in these experiments in an attempt to examine only initial events in spore killing. As the rates of killing of spores deficient in α/β-type SASP or recA-dependent DNA repair were generally similar to the rates of killing of wild-type spores, only wild-type spores were used in these experiments. Treatment of spores with Decon or Oxone^TM resulted in the loss of the normal brown colour of the spore coat, leaving the spore pellets white (data not shown). This has been noted previously with hypochlorite-treated spores and to a lesser degree with chlorine dioxide-treated spores (Young and Setlow 2003), and probably results from the bleaching of the CotA coat protein that gives purified B. subtilis spores and sporulated colonies their brown colour (Driks 1999). This reactivity of a specific coat protein with Decon and Oxone^TM underlines the importance of the spore coat in resistance to these agents.

As damage to the inner membrane permeability barrier of the spore is a likely mechanism of killing by a number of oxidizing agents (Loshon et al. 2001; Genest et al. 2002; Young and Setlow 2003), Decon and Oxone^TM treated spores were tested for damage to their inner membrane. One indication of damage to the inner membrane of the spore is release of DPA, a compound present in high levels in the spore core chelated with divalent cations and whose loss accompanies spore killing by wet heat (Russell 1982; Fairhead et al. 1993). DPA was not released (<10%) from spores killed 85–97% by either Decon or Oxone^TM (data not shown). However, subsequent incubation of treated spores at 80 or 85°C, temperatures that do minimal damage to untreated spores, resulted in the release of virtually all of the DPA from the Oxone^TM-treated spores and the majority of the DPA from the Decon-treated spores (Table 2). In contrast, much less DPA was released from untreated spores incubated at 80 or 85°C. Thus while the initial treatment of spores with Decon or Oxone^TM does not completely disrupt the inner membrane, there must be some damage to this membrane, damage whose effects are magnified by a subsequent heat treatment.

Table 2.  Dipicolinic acid (DPA) release from untreated, Decon- or Oxone^TM-treated dormant spores*

Treatment	Killing (%)	Heat treatment (°C)	DPA release (%)
*Spores of strain PS533 (wild-type) were either not treated or treated with Decon or Oxone^TM. After the level of killing was determined, the spores were washed, incubated in water for 30 min at various temperatures and the DPA released was determined as described in ‘Materials and Methods’. The amount of DPA in untreated spores of strain PS533 was 114 mg g^−1 dry spores, and this value was set as 100%. Incubation of wild-type spores for 30 min at 85°C caused ≤10% spore killing.
None	0	None	<3
Decon	85	None	<3
Oxone^TM	97	None	<3
None	0	80	15
Decon	85	80	58
Oxone^TM	97	80	100
None	0	85	31
Decon	85	85	63
Oxone^TM	97	80	99

Severe damage to the inner membrane permeability barrier might also cause premature hydration of the dormant spore core, which normally has a rather low water content (Gerhardt and Marquis 1989). One method for assessing the level of hydration of the spore core is to measure the core wet density. As expected from the absence of DPA release from spores upon Decon or Oxone^TM treatment, the core wet density of the treated dormant spores was similar to that of untreated spores, indicating that these treatments did not cause spore core hydration (Table 3).

Table 3.  Core wet density of dormant and germinated spores, with or without prior treatment with Decon or Oxone^TM*

Spores analysed	Core wet density (g ml^−1)
*The core wet densities of dormant spores with or without Decon or Oxone^TM treatment were determined after coat removal as described in ‘Materials and Methods’. Spores with or without Decon or Oxone treatment were germinated for 90 (Decon treated) or 120 min at 37°C in 2xYT medium plus 4 mmol l^−1l-alanine (untreated or Oxone^TM treated) and the core wet densities determined without decoating as described in ‘Materials and Methods’. With spores that have not been decoated, the core wet density value obtained is significantly lower than that obtained with decoated spores, because of the incomplete permeation of the density gradient solute into the spore cortex (Lindsay et al. 1985). †This value is almost identical to the value determined for dormant spores that have not been decoated (1·292 g ml^−1) (data not shown).
Dormant decoated spores
Untreated, WT (PS533)	1·332
Untreated, cwlD (PS2307)	1·327
Decon treated (95% killed)	1·346
Oxone^TM treated (98% killed)	1·332
Germinated spores
Untreated, WT (PS533)	1·177
Untreated, cwlD (PS2307)	1·231
Decon treated (94% killed)	1·218
Oxone^TM treated (94% killed)	1·289†

An additional method to assess damage to the inner membrane of the spore is by use of nucleic acid stains. The nucleic acids of the dormant spore, located in the spore core, are not accessible to stains, and DAPI treatment of untreated dormant spores results in only a peripheral staining that gives the spores a doughnut-like appearance (Setlow et al. 2002). As expected, DAPI gave only peripheral staining of Decon- and Oxone^TM-killed spores, although this staining was brighter than in untreated spores (data not shown), perhaps because of some changes in the spore coat or cortex resulting from treatment with these agents. However, the characteristic strong staining of the spore core by DAPI seen with germinated spores or with dormant spores with a disrupted inner membrane (Setlow et al. 2002) was not seen with Decon- or Oxone^TM-treated spores (data not shown).

Germination of Decon- and Oxone^TM-treated spores

While killing of spores by Decon and Oxone^TM is not due to DNA damage or catastrophic rupture of the spore inner membrane, there may be some type of damage to the inner membrane, as noted above. Treatment of spores with a number of chemical agents, including some that appear to cause inner membrane damage, results in the loss of the ability of treated spores to germinate (Williams and Russell 1993; McDonnell and Russell 1999; Tennen et al. 2000; Cabrera-Martinez et al. 2002; Setlow et al. 2002). This also appears to be the case with Decon- and Oxone^TM-treated spores, as upon incubation with nutrients these spores did not turn fully dark, as observed in a phase contrast microscope. However, Decon-treated spores incubated in nutrients became somewhat less bright in the phase contrast microscope and lost almost all of their DPA. In contrast, Oxone^TM-treated spores incubated similarly exhibited no change in appearance in the phase contrast microscope and released very little DPA (Table 4).

Table 4.  Changes during germination of untreated, Decon- or Oxone^TM-treated spores*

Treatment	Killing (%)	Phase dark (%)	Spore components released (%)
			DPA	Hexosamine
DPA, dipicolinic acid; ND, not determined. *Spores of strain PS533 (wild-type) were either not treated or treated with Decon or Oxone^TM, and the level of killing was determined. The spores were incubated for 2 h in either 50 mmol l^−1 Tris–HCl (pH 8·4) with 8 mmol l^−1l-alanine at 37°C or in Ca^2+-DPA for 2 h at room temperature, or for 6 h at 37°C in 2xYT medium plus 4 mmol l^−1l-alanine, and the percentage of spores that became phase dark and the percentages of total spore DPA and hexosamine released were determined as described in ‘Materials and Methods’. †Spores germinated in Tris–alanine. ‡Spores germinated with Ca^2+-DPA. §Spores germinated in 2xYT medium with alanine. ¶Decon-treated spores did not become fully phase dark, but were slightly less bright than dormant spores.
None†	0	97	100	ND
None‡	0	60	ND	ND
None§	0	97	ND	ND
None†	0	97	ND	35
None† (cwlD)	0	0	ND	0
Decon†	89	0¶	87	ND
Decon‡	86	0¶	ND	ND
Decon§	89	6¶	ND	ND
Decon†	92	0¶	ND	2
Oxone^TM†	94	8	ND	ND
Oxone^TM‡	97	2	ND	ND
Oxone^TM§	93	2	23	ND
Oxone^TM†	92	0	ND	4
Oxone^TM†	92	2	19	ND

Spores whose killing is due solely to damage to the nutrient germinant receptors can be recovered by triggering one of several alternative germination pathways that bypass the germinant receptors. One such pathway is triggered by exogenous Ca²⁺-DPA, as germination by this compound does not require the nutrient receptors and requires only one of the two cortex lytic enzymes of the spore (Paidhungat and Setlow 2000; Paidhungat et al. 2001). However, neither Decon- nor Oxone^TM-treated spores became dark in the phase contrast microscope after incubation with Ca²⁺-DPA, nor did they give rise to colonies on plates after this incubation (Table 4; and data not shown). These results, together with the inability of lysozyme treatment to rescue the killed spores (see below), suggest that Decon and Oxone^TM must do damage in addition to any damage to nutrient receptors and cortex lytic enzymes.

An additional non-nutrient germinant for spores of Bacillus spp. is the cationic surfactant dodecylamine (Rode and Foster 1960a,b, 1961). Bacillus subtilis spores incubated with this agent release DPA and initiate spore germination without the need for the germinant receptors or the cortex lytic enzymes, and may involve an effect of this agent on the inner membrane of the spore, possibly the creation or opening of a DPA channel or pore (Setlow et al. 2003). Strikingly, the germination of Decon- and Oxone-treated spores with dodecylamine as measured by release of DPA was significantly faster than that of untreated spores (Fig. 4). However, these treated spores were not revived by germination with dodecylamine (data not shown).

Figure 4. Germination of untreated or Decon- or Oxone^TM-treated Bacillus subtilis spores in dodecylamine. Wild-type spores (PS533) either with or without prior treatment with Decon or Oxone^TM were germinated with dodecylamine and DPA release measured as described in ‘Materials and Methods’. •, untreated spores; , spores killed 92% by Decon; and , spores killed 94% by Oxone^TM

An early event in spore germination is a significant increase in the hydration level of the spore core as the DPA that is released is replaced by water; this is followed by cortex hydrolysis and full core hydration as the core swells, and finally by the initiation of metabolism as seen in the generation of high energy compounds such as ATP and reduced pyridine and flavin nucleotides (Paidhungat and Setlow 2002). Spores that cannot hydrolyze their cortex during germination (i.e. cwlD spores; Popham et al. 1996; Setlow et al. 2001) cannot achieve full core hydration and this is reflected in a higher core wet density of germinated cwlD spores compared with that of germinated wild-type spores (Table 3) (Popham et al. 1996; Setlow et al. 2001). However, note that germinated cwlD spores have a lower core wet density than do dormant spores, reflecting the partial core hydration accompanying the DPA release upon germination of cwlD spores (Table 3). The core wet density of Decon-treated spores did decrease significantly during germination, consistent with the release of DPA by these spores, but the value reached was only slightly lower than that of germinated cwlD spores (Table 3). However, the core wet density of Oxone^TM-treated spores incubated in germination medium remained similar to that of dormant spores (Table 3), undoubtedly reflecting the absence of much DPA release during germination of these spores.

The lack of full core hydration upon incubation in germination medium despite the essentially full DPA release seen with Decon-treated spores could indicate damage by this agent (as well as perhaps by Oxone^TM) to the cortex lytic enzymes needed to degrade the peptidoglycan cortex of the spore and thus complete spore germination (Popham et al. 1996; Paidhungat et al. 2002). Cortex degradation releases hexosamine-containing fragments into the germination medium, but neither Decon- nor Oxone^TM-treated spores were able to hydrolyze their cortex efficiently (Table 4), consistent with the lack of full core hydration seen with these treated spores (Table 3). Bacillus subtilis spores treated with strong alkali, a treatment that inactivates all cortex lytic enzymes and gives spores that appear to be dead, can be revived by chemical decoating followed by incubation in a hypertonic medium with lysozyme, an enzyme that degrades the cortex (Setlow et al. 2002). Bacillus subtilis spores that lack their cortex lytic enzymes because of mutation or have a cortex that cannot be attacked by cortex lytic enzymes (i.e. cwlD spores) can also be revived by this lysozyme treatment (Popham et al. 1996; Setlow et al. 2001). Incubation of decoated Decon- or Oxone^TM-treated spores with lysozyme in a hypertonic medium did result in the hydrolysis of the spore cortex, as the great majority (>85%) of the spores became fully dark in the phase contrast microscope. However, in contrast to spores lacking cortex lytic enzymes that are revived by this treatment (Setlow et al. 2001), Decon- or Oxone^TM-killed spores appeared to lyse shortly after lysozyme treatment and viable colonies were not recovered at a level greater than that expected from the level of killing of the spores prior to decoating and lysozyme treatment (data not shown). Decoated Decon- or Oxone^TM-treated spores were also not revived on rich medium plates containing lysozyme (data not shown). Previous work has shown that spores that appear to be defective in germination after killing by chlorine dioxide, hypochlorite or a number of other chemicals are also not revived by lysozyme treatment in either hypertonic medium or on plates (Wyatt and Waites 1975; Williams and Russell 1993; Bloomfield 1999; Tennen et al. 2000; Cabrera-Martinez et al. 2002; Young and Setlow 2003).

The defect in germination of Decon- or Oxone^TM-killed spores can be further studied using the BacLight viability stain. Dormant spores are stained only peripherally and weakly by the propidium iodide component of this reagent, while live fully germinated spores are stained bright green with the SYTO  9 component and dead fully germinated spores are stained bright red by the propidium iodide (Melly et al. 2002a). The peripheral staining of dormant Decon- and Oxone^TM-treated spores with the BacLight reagent was slightly brighter than that of untreated dormant spores, similar to the results with DAPI noted above (data not shown). However, there was little difference in BacLight staining between the dormant and germinated Decon- or Oxone^TM-killed spores, indicating that neither of components of the BacLight stain penetrated into the germinated spore core (data not shown). This result is consistent with the lack of cortex degradation by Decon- or Oxone^TM-treated spores, as the nucleic acids in the core of spores that have initiated germination and released DPA but have not degraded their cortex are not stained by the dyes in the BacLight viability stain (Setlow et al. 2001).

As a final measure of the germination of Decon- and Oxone^TM-treated spores we assessed the metabolic capability of these spores when they were incubated in germination medium. This work used spores of strain PS3379 that carries the V. harveyi luxAB genes under the control of the forespore-specific promoter of the sspB gene and therefore accumulates luciferase in the dormant spores (Hill et al. 1994). Early in germination with nutrients these spores normally produce light when provided with a long chain aldehyde, indicating the initiation of metabolism and the availability of reduced flavin mononucleotide (Fig. 5) (Hill et al. 1994; Ciarciaglini et al. 2000; Loshon et al. 2001; Genest et al. 2002). However, Decon- and Oxone^TM-treated spores produced little (Decon) or no (Oxone^TM) light under these conditions (Fig. 5). The little or no metabolism in Decon- and Oxone^TM-treated spores upon incubation with nutrients is consistent with the incomplete hydration of the spore core of these treated spores, as full core hydration is required for the onset of metabolism by the germinating spore (Setlow et al. 2001).

Figure 5. Light production during germination of untreated or Decon- or Oxone^TM-treated Bacillus subtilis spores carrying the luxAB genes. Spores of strain PS3379 (luxAB) with or without prior treatment with Decon or Oxone^TM were germinated in 2xYT medium plus 8 mmol l⁻¹l-alanine and light production was measured as described in ‘Materials and Methods’. •, untreated spores; , spores killed 91% by Decon; and , spores killed 99·5% by Oxone^TM

Discussion

It has been discussed that both Decon and Oxone^TM kill spores of Bacillus spp. by causing oxidative damage to the spore DNA. Oxone^TM causes nucleic acid strand breakage (Rabner and McGuire 2002) while Decon causes oxidative attack on the spore DNA after damage to the spore coat (press release, Sandia National Laboratories, 1 March 1999, available at http://www.sandia.gov), with the DNA damage perhaps caused by the hydrogen peroxide component of Decon. However, the concentration of hydrogen peroxide in the undiluted Decon formulation is only 4%, and it has been previously shown that a much higher concentration of hydrogen peroxide is needed to obtain rapid spore killing (Setlow and Setlow 1993). Although damage to spore DNA might be caused by oxidizing agents such as hydrogen peroxide or Oxone^TM, previous work has shown that: (i) wild-type spores are significantly more resistant to hydrogen peroxide than α⁻β⁻ spores, as the α/β-type SASP protect DNA from damage by hydrogen peroxide (Setlow and Setlow 1993); and (ii) killing of wild-type spores by hydrogen peroxide and many other oxidizing agents is not through DNA damage (Setlow and Setlow 1993; Loshon et al. 2001; Genest et al. 2002; Melly et al. 2002a; Young and Setlow 2003). These previous findings as well as the absence of mutagenesis of wild-type or α⁻β⁻ spores by Oxone^TM or Decon and the similar sensitivity of wild-type, recA and α⁻β⁻ spores to these agents all strongly indicate that killing of spores by Decon and Oxone^TM is not through damage to spore DNA.

While Decon and Oxone^TM do not kill spores by damage to the spore DNA, α⁻β⁻ spores were more sensitive than wild-type spores to Oxone^TM and α⁻β⁻recA spores were significantly more sensitive than wild-type spores to Decon. We have no detailed explanation for these differences, but small and reproducible differences in the resistance of wild-type, recA, α⁻β⁻ and α⁻β⁻recA spores to agents that do not kill spores by DNA damage have been seen previously (Setlow et al. 2000; Tennen et al. 2000; Loshon et al. 2001; Genest et al. 2002). It has been found that mutations in these various strains, in particular those resulting in the loss of SASP-α and -β, cause global changes in gene expression during sporulation, and these changes are most notable for coat protein genes (Setlow et al. 2000). Given the crucial role of the coat in spore resistance to Decon and Oxone, it is perhaps not surprising that minor changes in the levels of individual spore coat proteins might result in significant changes in spore resistance to these agents. However, the identities of the specific coat proteins that may be involved in this phenomenon are not clear.

The thick proteinaceous coat of the spore protects the inner layers against many chemicals (Russell 1990; Bloomfield and Arthur 1994; McDonnell and Russell 1999; Riesenman and Nicholson 2000; Tennen et al. 2000; Young and Setlow 2003), although the coat plays no role in spore resistance to other chemicals including acid, alkali and ethanol, and provides only a small component of spore resistance to hydrogen peroxide (Riesenman and Nicholson 2000; Setlow et al. 2002). Treatment with both Decon and Oxone^TM clearly affects the spore coat as both agents bleached this structure, and spore resistance to Decon and Oxone^TM is dependent on the presence of an intact spore coat as shown by the high sensitivity of cotE spores and decoated wild-type spores to these agents. Presumably the chemically decoated spores were more sensitive to these agents than the cotE spores because more of the coat remained in cotE spores than in chemically decoated spores. Also consistent with the coats being a major determinant of spore resistance to Decon and Oxone^TM is that spores prepared at higher temperatures had higher resistance to these agents than spores prepared at lower temperatures. Previous work has shown that spores prepared at different temperatures differ in the levels of a number of coat proteins, and this is consistent with the different levels of spore resistance towards agents against which the coats are a major resistance determinant (Melly et al. 2002b; Young and Setlow 2003), including Decon and Oxone^TM. However, the roles of specific coat proteins in this phenomenon are not known.

Decon is a more complex decontaminant than Oxone^TM, containing cationic surfactants in addition to an oxidizing agent. The action of cationic surfactants, such as those contained in Decon, on bacterial spores was studied more than 40 years ago and reinvestigated recently (Rode and Foster 1960a, 1961; Setlow et al. 2003). Bacillus subtilis spores treated with the cationic surfactant dodecylamine initiate germination events, including release of DPA followed by cortex degradation (Rode and Foster 1961; Setlow et al. 2003). However, these events either do not take place at all with Decon-treated spores (cortex degradation), or take place only after provision of either nutrient germinants or dodecylamine itself (DPA release). Like Decon treatment, dodecylamine treatment also leads to spore killing (Rode and Foster 1960a,b, 1961; Setlow et al. 2003), but this killing takes place only after the spores have germinated as evidenced by DPA release, and Decon-treatment does not cause DPA release. Thus Decon does not appear to induce spore germination and kill the germinated spores.

Damage to the inner permeability membrane of the spore is thought to be the mechanism whereby spores are killed by a number of oxidizing agents, including chlorine dioxide, hypochlorite, Sterilox^® and peroxynitrite (Loshon et al. 2001; Genest et al. 2002; Young and Setlow 2003), and this same type of damage may also be the mechanism of spore killing by Decon and Oxone^TM. Treatment of dormant spores with Decon or Oxone^TM causes neither release of spore DPA, core hydration nor staining of core nucleic acids, indicating that the inner membrane permeability barrier has not been breached in the dormant spore. However, subsequent incubation of these treated spores at 80 or 85°C, normally sublethal temperatures for B. subtilis spores, resulted in release of the majority of spore DPA, indicating that the treated spores have undergone some changes in a permeability barrier such that this barrier is less robust at high temperatures. This latter barrier seems most likely to be the inner membrane of the spore as suggested previously for spore killing caused by other oxidizing agents (Loshon et al. 2001; Genest et al. 2002; Young and Setlow 2003). Also consistent with Decon and Oxone^TM treatment causing some type of damage to the spore inner membrane is the more rapid germination of these treated spores with dodecylamine. Possibly damage of some type to the spore inner membrane makes it easier for dodecylamine to cause DPA passage through this damaged membrane.

In addition to some type of damage to the spore inner membrane, Decon- and Oxone^TM-treated spores have severe germination defects and cannot complete the germination process. When spores were incubated in nutrient media, the Decon-treated spores could initiate the germination process as indicated by normal DPA release. However, the germinated Decon-treated spores reached a core hydration level only slightly lower than that of germinated cwlD spores, and there was little if any cortex degradation by the germinated Decon-treated spores. The germination defect in Oxone^TM-treated spores was more severe, as incubation of these spores in nutrients resulted in little DPA release as well as no cortex degradation.

What mechanisms might cause these germination defects in Decon- and Oxone^TM-treated spores? First, these decontaminants might damage the spore nutrient receptors, but if this was the case these agents must also cause additional damage, because incubation in Ca⁺²-DPA does not rescue the killed spores. Secondly, these agents may damage the cortex lytic enzymes or alter the cortex such that it cannot be degraded by cortex lytic enzymes. Indeed the core wet density of germinated Decon-treated spores is very similar to that of germinated cwlD spores in which the cortex is not degraded although DPA has been released (Popham et al. 1996; Setlow et al. 2001). In terms of their DPA release and lack of cortex degradation in germination media, spores killed by Decon also resemble spores killed by the iodine reagent Betadine (Tennen et al. 2000). However, it is important to note that even if the germination defect in Decon- and Oxone^TM-treated spores is bypassed by appropriate treatment with lysozyme, the treated spores are still not revived. Consequently, the germination defect caused by Decon or Oxone^TM treatment is not the only mechanism contributing to spore killing by these agents. One hypothesis is that agents such as Decon and Oxone^TM, as well as the many other oxidizing agents noted above, significantly damage not only the germination apparatus, including nutrient receptors and cortex lytic enzymes, but also the inner membrane of the spore (Loshon et al. 2001; Genest et al. 2002; Young and Setlow 2003). Perhaps this damage is so severe as to: (i) prevent the action of proteins in this membrane needed for spore germination; and (ii) result in spore lysis if spores are artificially germinated by lysozyme in a hypertonic medium. The inner membrane from spores with and without treatment with the agents used in this communication as well as previously should now be examined in an attempt to correlate specific damage to this membrane with spore killing.

Acknowledgements

This research was supported by a grant from the Army Research Office. We are grateful to Ann Cowan for assistance with the fluorescence microscopy.