The resistance of Bacillus atrophaeus spores to the bactericidal activity of peracetic acid is influenced by both the nature of the solid substrates and the mode of contamination


Murielle Naïtali, AgroParisTech, INRA, UMR1319 Micalis, Equipe Bioadhésion, Biofilm et Hygiène des Matériaux, 25 avenue de la République, 91300 Massy, France. E-mail:


Aims:  To evaluate the impact of the mode of contamination in relation with the nature of solid substrates on the resistance of spores of Bacillus atrophaeus -selected as surrogates of Bacillus anthracis- to a disinfectant, peracetic acid.

Methods and Results:  Six materials confronted in urban and military environments were selected for their different structural and physicochemical properties. In parallel, two modes of contamination were examined, i.e. deposition and immersion. Deposition was used to simulate contamination by an aerosol and immersion by an extended contact with liquids. A pronounced difference in the biocontamination levels and spatial organization of spores was observed depending on the mode of contamination and the nature of the solid substrate considered, with consequences on decontamination. Contamination by immersion led to lower efficiency of peracetic acid decontamination than contamination by deposition. Infiltration of spores into porous materials after immersion is one reason. In contrast, the deposition mode aggregates cells at the surface of materials, explaining the similar disinfecting behaviour of porous and nonporous substrates when considering this inoculation route.

Conclusions:  The inoculation route was shown to be as influential a parameter as material characteristics (porosity and wettability) for decontamination efficacy.

Significance and Impact of the Study:  These results provide comparative information for the decontamination of B. atrophaeus spores in function of the mode of contamination and the nature of solid substrates.


A variety of microbiological agents may be used in a biological warfare or bioterrorism attack such as Brucella sp., Burkholderia malei, Francisella tularensis and Yersinia pestis. Special attention should be given to spore-forming pathogens because of the high resistance of spores to environmental stress and decontamination treatments. Hazard of Bacillus anthracis spores has been shown by the events of 2001 in the United States. Anthrax spores were intentionally released using the postal system. Spores were found widely dispersed into areas surrounding one delivery bar code sorter machine that had sorted the contaminated envelopes, and four inhalational anthrax cases occurred in the mail distribution centre involved (Sanderson et al. 2004). This event prompted extensive clean-up efforts and increased public awareness as well as a growing interest in B. anthracis detection methods, sampling and inactivation within the United States and all over the world (Canter 2005).

In order to react rapidly and efficiently against a bioterrorism attack and ensure the protection of soldiers and civilians, increased knowledge in decontamination is needed. Nowadays, the decontamination efficiency of a disinfectant solution is evaluated on planktonic cells or on cells that are deposited on glass or steel as the carrier materials according to standard test methods (Rastogi et al. 2009). In these standards, biocontamination of materials is carried out by deposition of calibrated bacterial suspensions (ASTM 2005; CEN 2005; AFNOR 1988a.). This route of contamination can simulate an aerosol of large liquid particles. However, contamination of materials can also be achieved by an extended contact with bacterial suspensions, which can be simulated by contamination by immersion. Furthermore, in urban and military domains, along with glass and steel, other materials such as plastics, wood, tissue and fibres can be met. They differ notably according to their porous/nonporous characters and their hydrophilic/hydrophobic properties. Some studies have been conducted taking into account the nature of the solid substrates (Rogers et al. 2007; DeQueiroz and Day 2008; Rastogi et al. 2009) while considering only one mode of contamination. Recently, a study has shown that the latter parameter has an influence on the repartition of spores (Edmonds et al. 2009), with potential consequences on decontamination efficiency.

In this context, our study aimed at determining the efficacy of a liquid disinfectant (peracetic acid) according to the nature of the solid substrate as well as the mode of contamination. The influence of the mode of contamination on the biocontamination capacity of spores and their spatial organization on different surfaces was first considered. Then, the impact of the nature of the solid substrates on disinfection efficiency was evaluated following conditions similar to those defined in standard tests such as EN 14561, i.e. a contamination by deposition at a level between 107 and 108 spores. Next, the impact of the mode of contamination on disinfection efficiency was studied for various solid substrates using similar levels of biocontamination, as this parameter can be influential (Johnston et al. 2000; Virto et al. 2005; Kamgang-Youbi et al. 2008; Rastogi et al. 2009). All experiments were conducted with spores of Bacillus atrophaeus (previously named Bacillus globigii), a classical nonpathogenic surrogate of B. anthracis warfare, which is employed in testing of liquid disinfecting agents (AFNOR 1988b.). Its spores have shown similar sensitivity to chemical decontamination as those of virulent B. anthracis (Rogers et al. 2005; Sagripanti et al. 2007).

Materials and methods

Preparation of spores

Bacillus atrophaeus CIP 7718 was grown on a sporulation medium agar-agar (Difco, Pessac, France) 20 g l−1, nutrient broth (Difco) 25 g l−1, manganese sulfate 50 mg l−1, calcium chlorure 130 mg l−1) at 30°C for at least 7 days. Once spores had been collected by scraping the surface of agar, they were suspended in 1·5 × l0−1 mol l−1 NaCl before being washed four times and stored at 4°C (concentration 1010 cells ml−1). A thermal treatment (30 min, 90°C) was performed on an aliquot of each stored suspension in order to determine the percentage of spores. It was superior to 95%. The stored suspensions were diluted before utilization. Three stocks of independently produced spores were used.

Inert surfaces

Six materials, representing porous or nonporous substrates were used (Table 1). Chips of each material with an apparent surface of 4·5 cm2 were autoclaved (20 min, 120°C), dried and stored in the dark before utilization. Physicochemical characteristics of materials are given in Table 1. The results were the mean of at least nine measurements at different places of the substrate performed on two different substrates. Materials presented different surface roughness (perthometer M2, Mahr, Igny, France). In our experimental conditions of preparation, glass is a hydrophilic solid substrate, whereas polytetrafluoroethylene (PTFE) and painted stainless steel are hydrophobic as determined by water contact angle (contact angle measuring system G10; Krüss, Palaiseau, France).

Table 1.   Materials used in this study
Solid substrateProvenancePorosityWater contact angle θ (°)Average roughness Ra (μm)Maximum roughness Rmax (μm)
  1. ND, not determinable.

  2. *Water droplet penetrated into porous materials.

  3. †Glass fibres were too rough for the perthometer.

Painted stainless steel covered by a polyurethane paintCentre d’Etudes du Bouchet, DGA, FranceNonporous122 ± 32·0 ± 0·214·1 ± 1·7
PolytetrafluoroethyleneGoodfellow, United KingdomNonporous106 ± 10·5 ± 0·16·1 ± 3·1
GlassErie Scientific, USANonporous23 ± 40·03 ± 0·020·3 ± 0·3
Tissue, 100% cottonLa Redoute, FrancePorousND*8·4 ± 1·958·9 ± 13·7
Pinewood, pinhus pinasterLeroy Merlin, FrancePorousND*3·6 ± 1·129·8 ± 10·4
Glass fibresWhatman, EnglandPorousND*ND†ND†

Biocontamination of inert surfaces by deposition or immersion

For deposition, five small droplets of 10 μl of microbial suspensions containing approximately 109 or 107 CFU ml−1 depending on the experiments were deposited on the solid surfaces and then dried for 1h30 at 20°C. For immersion (or static adhesion), the solid substrates were placed in Petri dishes (5·5 cm in diameter) and covered with 10 ml of a spore suspension, the concentration of which depend on the experiment and the solid substrate considered. After 2 h at 20°C, materials were rinsed to remove nonsticking spores (5 × 5 ml of 1·5 × l0−1 mol l−1 NaCl). In some cases, a drying step (1h30 at 20°C) was added.

To determine the level of biocontamination, deposited and adherent spores were evaluated by duplicate plating on Trypticase Soy Agar (TSA; bioMérieux, Marcy l’Etoile, France) after detachment by vortex-mixing (30 s) with glass beads in 10 ml of a recovery solution (30 g l−1 Tween 20 or a quenching solution as detailed in the following paragraphs). Experiments were performed at least three times on different stocks of spores. The contaminated solid substrates were also used immediately for scanning microscopy observations and disinfection assays.

Microscopic observations

Material topography, spore morphology and cell organization on surfaces were visualized by Scanning Electron Microscopy (SEM S-4500; Hitachi, Tokyo, Japan) after fixation, dehydration and gold sputtering (JFC-1100E ion sputter; Jeol, Tokyo, Japan), as previously described (Kamgang et al. 2007). Micrographs were performed using the MiMA2 microscopy platform. At least 20 images were made for each substrate at different places and typical images were selected.

Disinfection assays of contaminated surfaces

Peracetic acid (32 % in dilute acid acetic; Sigma, Munich, Germany), an oxidative agent, was used as disinfectant after dilutions in distilled water. Peracetic acid dilutions were prepared just before use. The tolerance of deposited and adherent spores to the bactericidal activity of peracetic acid was evaluated according to adaptations of the European standard EN 14561 (CEN 2005).

Briefly, the contaminated solid substrates were immersed in 10 ml of peracetic acid diluted solution at 20°C for 5 min. The survivors on both the solid substrates or detached in the disinfecting solution were then evaluated. For this evaluation, the action of the biocide was stopped by the simultaneous transfer of the disinfected support to 10 ml of a quenching solution (Naitali et al. 2009) containing glass beads and 1 ml of the disinfectant solution to 9 ml of the quenching solution. The cells were detached from solid substrates by vortex-shaking (30 s).

Survivors were evaluated by the drop counting method (Chen et al. 2003) in which 6 × 10 μl drops of 1/10 serial dilutions in 1·5 × l0−1 mol l−1 NaCl were placed on the surface of agar plates (TSA). A volume of 2 × 1 ml of the two most concentrated suspensions was also directly plated to lower the threshold, making it equal to 50 CFU per solid substrate. Plates were incubated at 30°C for 24 and 48 h. It has previously been shown that the drop-counting method gives similar results to those of the classical plate method (data not shown). The results were the mean of survivors (sum of those from the support and those found in the disinfectant solution) of at least three experiments performed on different stocks of spores. For each disinfection test, the number of spores present before disinfection was evaluated. The absence of toxicity of the quenching solution and its efficacy was verified.

Statistical analysis

Analyses of variance (anova) were performed using Statgraphics software (ManugisticsTM, Rockville, MD, USA). The P-values tested the statistical significance of each of the factors through F-tests. When P-values were lower than 0·05, these factors had a statistically significant effect at the 95 % confidence level.


The nature of the solid substrate and the mode of contamination influence biocontamination

In the standard test EN 14561, a target level of approximately 5 × 107 (7·7 log) spores per material is recommended. When contamination was performed by deposition of 1·3 × 108 spores, between 4·1 × 107 (7·7 log) and 9·8 × 107 (8·0 log) spores were recovered per solid substrate in agreement with the specification of the standard test. No significant difference in the level of contamination between porous and nonporous substrates was observed (P > 0·05). Using SEM makes it possible to study the spatial organization of spores. They are found on the surface of all solid substrates and they highly aggregate on the surface after deposition, as illustrated in Fig. 1 for three materials. A circular delimitation of deposition of spores was observed for nonporous materials, probably because of droplet drying. Within this delimitation, spore repartition is more homogeneous for glass than for PTFE (Fig. 1).

Figure 1.

 Scanning Electron Microscopy images of material contaminated at approximately 5 × 107 CFU per chip. Scale: the white bars correspond to 300 μm for the first line, 30 μm for the second line and 7·5 μm for the last line. ND: not done.

Contamination by immersion was carried out by settling of a suspension with an initial concentration of approximately 108 CFU ml−1 on the solid substrate (Table 2). In such conditions, the level of contamination depends on the substrate considered and ranges from 4·8 to 7·4 log CFU per material. The number of sticking bacteria was found to be higher on porous surfaces (< 0·05) (pinewood, tissue and glass fibres) than on nonporous (glass, painted stainless steel and PTFE). Among nonporous substrates, painted stainless steel presented a significantly higher level of contamination (P < 0·05). Glass fibres presented the highest level of contamination among all the substrates tested. Fewer spores were observed on the surface of glass fibres after immersion than after deposition (Fig. 1), whereas the level of contamination was similar. Spores were observed in depth, trapped between glass fibres after immersion.

Table 2.   Biocontamination level of Bacillus atrophaeus spores on porous and nonporous surfaces after immersion in 10 ml of a suspension containing 108 spores ml−1
SupportSpores detected (log CFU per material)
Painted stainless steel6·0 ± 0·4
Polytetrafluoroethylene4·8 ± 0·4
Glass4·8 ± 0·6
Tissue7·1 ± 0·1
Pinewood6·7 ± 0·2
Glass fibres7·4 ± 0·4

Solid substrates influence the disinfection efficiency in applying standard mode and level of contamination

The inactivation of deposited spores (approximately 5 × 107 spores per solid substrate for all materials as referred to above) was performed by exposure to 0·2 % peracetic acid for 5 min. The results obtained (Table 3) show a significant impact of the solid substrate on disinfection efficiency (P < 0·05). Three homogeneous groups were determined using anova. Tissue was the most easily decontaminated material; the second group was constituted by glass fibres, painted stainless steel and PTFE; finally pinewood and glass were the substrates that were the most difficult to decontaminate. There was, furthermore, a significant difference between hydrophilic (glass) and hydrophobic (PTFE and painted stainless steel) nonporous substrates (P < 0·05).

Table 3.   Activity of 0·2 % peracetic acid on deposited spores according to the contaminated substrate
MaterialSpores recovered* (log CFU per material)Recovery* (%)Logarithmic reduction after disinfection
  1. *Before disinfection, for approximately 5 × 107 spores deposited.

Painted stainless steel7·7 ± 0·175 ± 264·8 ± 0·5
Polytetrafluoroethylene7·4 ± 0·164 ± 414·3 ± 0·9
Glass7·7 ± 0·369 ± 252·5 ± 0·2
Tissue7·6 ± 0·166 ± 276·4 ± 0·4
Pinewood7·7 ± 0·268 ± 232·9 ± 0·5
Glass fibres7·4 ± 0·142 ± 215·0 ± 0·4

The mode of contamination also influences the disinfection efficiency in relation to the nature of solid substrates

The impact of the mode of contamination on disinfection efficiency can be studied only for identical levels of contamination. Glass fibres were the only material for which experimental conditions were defined that made it possible to obtain the standard level of contamination for the two modes of contamination studied, as previously presented. They were therefore the only material used for studying the impact of the mode of contamination on disinfection efficiency following the recommended level of the standard test. The results reported in Table 4 show that there was a significant influence of the mode of contamination on disinfection efficacy (P < 0·05) for this solid substrate. Peracetic acid was found to be more efficient on deposited spores than on spores obtained after immersion. Furthermore, a drying step after contamination by immersion significantly decreased the efficacy of the disinfectant (< 0·05).

Table 4.   Activity of 0·2 % peracetic acid on spores varying with the mode of contamination of glass fibres
Mode of contaminationSpores recovered (log CFU per material)Logarithmic reduction after disinfection
  1. *Before disinfection, for approximately 5 × 107 spores deposited.

  2. †Before disinfection, for glass fibres immersed in 10 ml of suspension containing approximately 108 spores ml−1.

Deposition7·4 ± 0·1*5·0 ± 0·4
Immersion7·2 ± 0·3†3·7 ± 0·2
Immersion and drying7·1 ± 0·1†3·0 ± 0·3

The disinfection was also tested in the case of a level of contamination which was reached for all solid substrates (i.e., 105 CFU per solid substrate, the lowest level during immersion as referred to above). This level was obtained by adjusting the concentration of the suspensions used to contaminate the substrate. For the deposition method, the utilization of a suspension containing 107 CFU ml−1 proved satisfactory. For immersion, the suspensions were adjusted from 5 × 107 to 108 CFU ml−1 for glass and PTFE, approximately 107 UFC ml−1 for glass fibres and tissue and from 5 × 106 to 107 CFU ml−1 for pinewood and painted stainless steel. The number of cells recovered before disinfection is given in Fig. 2.

Figure 2.

 Activity of 0·065 % peracetic acid on spores varying with the solid substrate (a: nonporous, b: porous) and the mode of contamination. Full and striped bars represent the spores detected before and after the disinfection test, respectively, for contamination performed by deposition (inline image and inline image), by immersion (inline image and inline image) and by immersion followed by a drying step (inline image and inline image). The Y-axis begins at 1·7 log, corresponding to the detection threshold.

No survivors were detected for several solid substrates when 0·2% peracetic acid was applied to this level of contamination for 5 min (data not shown). The disinfection assays were thus performed with a lower concentration of peracetic acid (0·065 %) that was selected as being discriminant. The results (Fig. 2) indicate that the logarithmic reductions obtained for the three nonporous substrates were significantly higher than for the porous surfaces, in the case of contamination by immersion (P < 0·05). Additionally, peracetic acid was found less efficient on substrates contaminated by immersion than by deposition (P < 0·05). The difference was less for nonporous than for porous surfaces. When a drying step was added after immersion, the mode of contamination no longer influenced disinfection efficacy.


Microbial contamination of solid substrates is governed by parameters such as the physico-chemical and rugosity properties of both materials and micro-organisms (Vanloosdrecht et al. 1987; Faille et al. 2002). Hydrophobicity, surface charge and cell density have been shown to influence adhesion of Bacillus spores to materials (Husmark and Ronner 1990; Ronner et al. 1990). The present study shows that biocontamination depends as well on the mode of contamination when immersion and liquid deposit are compared. This effect was also recently observed by Edmonds et al. (2009) in the case of liquid deposit and dry aerosol. In our study, immersion seems to hinder cell aggregation and increase spore penetration into porous materials (Fig. 1). This point probably explains the higher level of contamination on porous substrates observed with this inoculation route, as it protects spores from being detached during rinsing. Among nonporous substrates, painted stainless steel presented the highest contamination following immersion. Seale et al. (2008) observed a similar result with a higher contamination on steel than on glass for hydrophilic Geobacillus spores. They attribute this result to the lower Gibbs energy of this solid substrate but highlight that there is no simple relationship between individual physicochemical interactions and spore adherence. In our case, adherence was probably because of the heterogeneity and the roughness of the painted stainless steel, which presented the highest roughness among nonporous substrates (Table 1). It presented crevices and depressions (data not shown) that can hide spores, similarly to the chemical agent-resistant coating-painted stainless steel illustrated by Edmonds et al. (2009).

This study presents, for the first time to our knowledge, the impact of the route of surface contamination together with the nature of the material on the decontamination efficiency of a disinfectant solution. We have observed that, for all cases studied, contamination by immersion led to a lower efficiency of peracetic acid decontamination than a contamination by deposition. Different reasons could be evoked depending on the nature of the solid substrate and on the concentration of disinfectant utilized. When a low concentration of peracetic acid was utilized, the decrease in efficiency was substantial for porous substrates for which the trapped water may dilute the disinfectant (Fig. 2). Accordingly, the disinfecting efficiency was somewhat restored when a drying step was added before disinfection. When a high concentration of peracetic acid was used, the dilution was not the most important phenomenon: the elimination of the water trapped in the glass fibres (drying step) did not increase the lethal effect of the disinfectant (Table 4). In that case, the main phenomenon contributing to the lower efficiency is probably the higher penetration for immersion than deposition of cells in depth in the porous substrates tested (i.e., glass fibres), which both prevents detachment of cells (consequences of detachment on resistance to disinfection are discussed below) and reduces interaction with the disinfectant. In agreement, drying did not restore the disinfecting efficiency and even decreased it. The lower efficiency of peracetic acid observed for dried glass fibres could result from a weaker diffusion of disinfectant molecules within that dried fibres. This can be compared to the results of Watling et al. (2002), for which the decontaminating effect of hydrogen peroxide was enhanced by condensation. Indeed, as discussed by Rogers et al. (2005), condensation could increase penetration of the disinfectant into porous materials, which could thus reach embedded spores. Further investigation is required on this point.

As just discussed for glass fibres, the deposition mode aggregates cells at the surface of materials and prevents infiltration into porous materials that thus have a similar behaviour to nonporous substrates. In agreement, no differences in disinfection results were observed between porous and nonporous substrates in the case of contamination by deposition. On the contrary, using a similar deposition mode of contamination but materials that differ from ours (except wood), Rogers et al. (2005) conclude that the observed logarithmic reductions on porous materials were consistently lower than on nonporous materials. In fact, no general rule concerning the impact of porosity on the efficiency of decontamination can be drawn from the literature as experiments varied in the treatment of several aspects including contamination routes, materials and disinfectants (Sigwarth and Stark 2003; Rogers et al. 2005, 2007; DeQueiroz and Day 2008; Rastogi et al. 2009).

Concerning the impact of the materials, in agreement with Rogers et al. (2005) and Rastogi et al. (2009), we have found that wood appears to be a difficult material to decontaminate. The presence of interfering soils within this complex biological material is probably one of the reasons. More surprisingly, we have also found that glass is the least decontaminated surface when considering standard conditions of contamination (i.e. deposition at 5 107 cells per substrates, Table 2). In contrast to the results of Justi et al. (2001), we have thus found that glass is less easily decontaminated than PTFE. Spatial organization of deposited spores on these two nonporous substrates can be a reason. As Sigwarth and Stark (2003), we have noted that PTFE differs from glass in the spatial repartition of contaminants after drying. This organization clearly depends on material wettability. According to Rauschnabel et al. (2006), lower wettability favours multilayered deposition of micro-organisms and decreases the efficacy of disinfection procedures, a situation not found in this study. In our case, repartition is more homogenous on glass, a nonporous hydrophilic substrate, than on PTFE, a nonporous hydrophobic one. This result can be because of droplets spreading during drying and to the hydrophilic character of the spores used (10% of affinity to hexadecane determined by the MATS method of Bellon-Fontaine et al. (1996)). Cell droplets dried on PTFE come unstuck the solid substrate in contrast to cells dried on glass (Fig. 1). They could also detach and then be treated in their planktonic form and not in an attached state. We did not test the resistance of detached spores, but we found that planktonic (never attached) spores were easier to eradicate than attached ones (a logarithmic reduction greater than six log was observed for disinfection of planktonic spores as compared to the results of Table 3), in line with the recent results of Kreske et al. (2006) on Bacillus spores. This result is certainly because of a larger contact surface between cells and disinfectant in the case of planktonic cells and less probably to physiological modification, as often evoked for vegetative cells (Fux et al. 2004; Kamgang-Youbi et al. 2008). In that case, it is likely that planktonic detached spores present the same resistance as planktonic never-attached spores. We state here that the hydrophobic/hydrophilic character of both substrates and spores have an impact on the adherence behaviour and then on disinfection efficiency. It would then be interesting in further studies to test surrogates whose spores presented surface properties closer to those of B. anthracis than B. atrophaeus. Bacillus cereus strains could be considered as their spores are more hydrophobic and exhibit an exosporium as B. anthracis spores (Zolock et al. 2006).

On the whole, this study highlights different aspects concerning ‘biological agent/material’ interrelations during contamination and decontamination. The inoculation route was shown to be as influential a parameter as material characteristics (porosity and wettability) for decontamination efficacy. The implication of each parameter could vary within materials (Unger et al. 2007). One can also ask if the apparent difference in sensitivity to disinfectants among materials is related to differential recovery (Rogers et al. 2007; Sagripanti et al. 2007). Our recoveries (Table 3) are in line with those previously reported (Rogers et al. 2005, 2007; Edmonds et al. 2009; Rastogi et al. 2009). Even when studies were conducted to optimize recovery rates according to the nature of the material and the mode of contamination (Edmonds et al. 2009; Rastogi et al. 2009), in no case has total recovery been attained. Our further studies will therefore aim at evaluating the disinfection efficiency without recovery phase using in situ observation with confocal laser scanning microscopy and fluorescent probes.


This work was supported by doctoral grants from the French Ministry of Defence (DGA/MRIS) and the National Scientific Research Centre (CNRS). We thank Donald White for revising the English manuscript.