Inactivation of spores and vegetative cells of Bacillus subtilis and Geobacillus stearothermophilus by pulsed light

Innovative Food Science and Emerging Technologies 28 (2015) 52–58

Contents lists available at ScienceDirect

Innovative Food Science and Emerging Technologies journal homepage: www.elsevier.com/locate/ifset

Inactivation of spores and vegetative cells of Bacillus subtilis and Geobacillus stearothermophilus by pulsed light Mari Luz Artíguez ⁎, Iñigo Martínez de Marañón AZTI, Food Research, Parque Tecnológico de Bizkaia, Astondo Bidea, Edificio 609, 48160 Derio, Bizkaia, Spain

a r t i c l e

i n f o

Article history: Received 14 August 2014 Received in revised form 24 November 2014 Accepted 7 January 2015 Available online 2 February 2015 Keywords: Decontamination Total fluence Bacterial spores Population cell density Light exposure Transmittance Physiological state

a b s t r a c t The effect of pulsed light (PL) on the inactivation of vegetative cells and spores of Bacillus subtilis and Geobacillus stearothermophilus at different cell densities was evaluated. The antimicrobial effect of PL decreased when population density increased, both for vegetative cells and spores of B. subtilis and G. stearothermophilus. For low cell densities, vegetative cells were more sensitive to PL than spores. However, lower reductions in vegetative cell counts were shown for higher cell densities, which could be attributed to the fact that vegetative cell suspensions transmitted less amount of light than spores. Concerning the resistance of both microorganisms, lower reduction in G. stearothermophilus than B. subtilis counts were found for the same cell density. When cell suspensions with similar light transmittance were compared, vegetative cells of B. subtilis were found to be more sensitive than the ones of G. stearothermophilus, while the spores of G. stearothermophilus were less resistant to PL. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction Pulsed light (PL) consists of a successive repetition of short duration and high power flashes of broadband emission light (190–1000 nm) with approximately 40% of the emitted light corresponding to the ultraviolet (UV) region (Wekhof, 2000). This novel technology appears as a promising alternative to conventional heat preservation processes to ensure the microbial quality and safety of food products. It has been shown to be effective in inactivating a wide range of microorganisms involved in food spoilage and food borne pathogens including bacteria, fungi, viruses and protozoa (Elmnasser, Federighi, Bakhrouf, & Orange, 2010; Gómez-López, Devlieghere, Bonduelle, & Debevere, 2005; Huffman, Slifko, Salisbury, & Rose, 2000). However, one of the main challenges for the application of non-thermal technologies in the food industry is the inactivation of bacterial spores. These microbial forms are extremely resistant to many stresses, including toxic chemicals and biocidal agents, desiccation, high pressure, heat treatment, ionising radiation or ultraviolet (UV) processing (Nicholson, Munakata, Horneck, Melosh, & Setlow, 2000; Setlow, 2006). Their high resistance to UV radiation has been mainly attributed to altered DNA photochemistry caused by the binding of small acid-soluble proteins (SASP), an efficient repair pathway specific for their photoproducts, ⁎ Corresponding author. E-mail addresses: [email protected] (M.L. Artíguez), [email protected] (I. Martínez de Marañón).

http://dx.doi.org/10.1016/j.ifset.2015.01.001 1466-8564/© 2015 Elsevier Ltd. All rights reserved.

the accumulation of dipicolinic acid in the dormant spore core or a low core water content and the presence of a thick spore protein coating (Nicholson et al., 2000; Setlow, 2006). Regarding PL, several previous works have reported its use in bacterial spore inactivation (Buschnell, Cooper, Dunn, Leo, & May, 1998; Chaine, Levy, Lacour, Riedel, & Carlin, 2012; Rice & Ewell, 2001). However, only a few studies have pointed out the higher sensitivity to PL of vegetative cells compared to spores (Dunn et al., 1989; Levy, Aubert, Lacour, & Carlin, 2012), so that the impact of the physiological state of the microorganisms on PL effectiveness is not fully elucidated. PL may have potentiality to decontaminate liquid food products (water, milk, juice …), the surface of solid food products (eggs, meat and fish products …), packaging and processing equipment. The application of PL for liquid food treatment is limited by the penetration capacity of the incident light, which determines the exposure of the microbial cells to the effective wavelengths and therefore, its decontamination effectiveness (Artíguez, Arboleya, & Martínez de Marañón, 2012). Since the microbial population density could determine the degree of light penetration through the liquid, one of the aims of this study was to evaluate its influence on the antimicrobial effectiveness of PL. In addition, the resistance of vegetative cells and bacterial spores was compared to assess the impact of the bacterial physiological state on the effectiveness of PL for microbial inactivation. For these purposes, the antimicrobial effect of PL was evaluated on Bacillus subtilis and Geobacillus stearothermophilus, bacteria of great concern to the food processing industry due to their ability to form highly resistant spores

M.L. Artíguez, I. Martínez de Marañón / Innovative Food Science and Emerging Technologies 28 (2015) 52–58

and their spoilage potential (Burgess, Lindsay, & Flint, 2010; From, Pukall, Schumann, Hormazábal, & Granum, 2005). 2. Material and methods 2.1. Bacterial strains and growth conditions Microbial species, culture media and incubation temperatures used in this study are summarised in Table 1. Bacterial strains were stored at − 80 °C in a 20% (w/w) glycerol solution. Thawed stock cultures (100 μL) were transferred to 10 mL of the appropriate broth and precultured at temperatures indicated in Table 1 for 24 h. Each bacterial strain was then inoculated at 103 CFU/mL and cultured at the corresponding temperature for 24 h until early stationary growth phase. Cells were harvested by centrifugation (5804R centrifuge, Eppendorf AG, Germany) at 10,000 ×g for 15 min at 4 °C, washed twice with Potassium Phosphate Buffered Saline (KPBS; 0.01 M K2HPO4, 0.01 M KH2PO4, 0.15 M NaCl; pH: 6.7) and finally resuspended in this buffer at a cell density of 105–107 CFU/mL for G. stearothermophilus and 107–109 for B. subtilis. 2.2. Preparation of spore suspensions For spore preparation, aliquots of 500 μL of bacterial culture prepared as described above were spread on sporulation agar, and then incubated at the appropriate temperatures (Table 1) until N90% of the cells were sporulated. Sporulation was verified by observing the refractile spores using a phase-contrast microscopy. Spore suspensions were not heated to inactivate vegetative cells because thermal treatments could affect the resistance to posterior PL treatments of B. subtilis spores (Artíguez & Martínez de Marañón, 2015). Spores were harvested from agar plates with distilled sterile water. Collected spores were centrifuged at 4000 ×g for B. subtilis and at 8000 ×g for G. stearothermophilus, for 15 min at 4 °C, washed three times with distilled sterile water and cell density adjusted to 105–108 spores/mL for G. stearothermophilus and 107–109 for B. subtilis.

53

Samples were placed on the centre of the quartz shelf at 6 cm from the upper xenon lamp and treated at 2.2 kV. Fluence per pulse was measured with an energy metre (model QE25-LP-H-MB, Gentec, Canada) used with a Solo2 readout unit and was expressed in joules per square centimetre (Artíguez et al., 2012). Pauses of at least 30 s between measurements were allowed to prevent possible overheating of the detector. All fluence measurements were performed in triplicate. Total fluence or amount of photons striking on the sample per unit area, is the most relevant process factor affecting microbial inactivation by PL (Artíguez & Martínez de Marañón, 2014; Lasagabaster & Martínez de Marañón, 2013). Therefore, total fluence (H) was calculated by multiplying pulse fluence by the number of emitted light pulses (n) (H = H0 × n). Samples were treated at fluence in the range of 0.14–12 J/cm2. Each condition treatment was repeated at least three times. Untreated inoculated samples were used as controls. 2.5. Transmittance spectrum of the bacterial suspensions The spectral transmittance of all samples was determined each 1 nm from 190 to 1000 nm with a Genesys 6 UV/VIS spectrophotometer (Thermo Spectronic, Rochester, USA). All measurements were performed in the same cuvettes used for PL treatments (see above). Distilled water was used as the blank. At least three measurements per sample were performed. 2.6. Microbiological analyses Bacterial counts in inoculated untreated (control) and PL treated samples were determined immediately after each treatment. Liquid samples were serially diluted in 1% buffered peptone water (Pronadisa, Spain) and 0.1 mL of appropriate dilutions were surface plated onto the corresponding agar media (Table 1). After incubating Petri dishes at the appropriate temperature for 72 h, colonies were enumerated and the results expressed as Log CFU/mL, 1 Log being the detection limit of the plate count method. 2.7. Data analyses

2.3. Pulsed light device A desktop SBS-XeMaticA-(L + L) device (SteriBeam Systems GmbH, Germany) was used. The polished stainless steel treatment reactor consists of a vertically mobile quartz shelf located between two xenon lamps (upper and lower), which emit high intensity light pulses of 325 μs duration. The emission spectrum includes wavelengths from 190 to 1000 nm, with about 20% of the emitted light in the UV-C, 8% in the UV-B and 12% in the UV-A region (Wekhof, 2000). A fan cooling system is used to prevent overheating of lamps and samples. 2.4. Pulsed light treatment Aliquots of 300 μL of each resulting bacterial suspension (prepared as described above) were transferred to sterile UV-transparent Suprasil quartz cuvettes (optical length 1 mm, Hellma, Germany) and immediately subjected to PL treatment at room temperature (23–25 °C).

Microbial inactivation was calculated as Log (N0/N), where N0 represents initial cell count (untreated samples) and N post-treatment bacterial count (treated samples). Analysis of variance and Tukey's honestly significant difference test were used to determine significant differences among treatments (p b 0.05) (SPSS Inc., IL, USA). 2.8. Inactivation modelling The Geeraerd and Van Impe inactivation model-fitting tool (GInaFiT), a freeware add-in for Microsoft Excel, was used (Geeraerd, Valdramidis, & Van Impe, 2005). The Weibull and the log-linear with shoulder and/or tail model were fitted to experimental data. Results indicated that the log-linear with shoulder and/or tail model, proposed by Geeraerd, Herremans, and Van Impe (2000) best fitted the experimental data (data not shown) and, therefore, it was employed to describe B. subtilis and G. stearothermophilus inactivation as a function of the total

Table 1 Origin and culture conditions of the tested microorganisms. Species

Codea

Growth mediab

Sporulation mediab

Culture mediab

Incubation temperature (°C)

B. subtilis G. stearothermophilus

DMS 10 CECT43

TSB TSB

NA (1) NA (2)

TSA NA (3)

30 56

NA (1) Nutrient agar supplemented with 1 mg/L MnSO4 and 0.5 g/L CaCl2 (Prentice, Wolfe, & Clegg, 1972); NA (2) Nutrient agar supplemented with 10 mg/L MnSO4:H2O (Finley & Fields, 1962); TSA, Tryptic Soy Agar; NA (3), Nutrient agar. All media were provided by Pronadisa (Spain). a DSM, Deutsche Sammlung von Mikroorganismen und Zellkulturen, Germany; CECT, Spanish Type Culture Collection. b TSB, Tryptic Soy Broth.

54

M.L. Artíguez, I. Martínez de Marañón / Innovative Food Science and Emerging Technologies 28 (2015) 52–58

0

PL fluence. According to this model, the reduction of microbial population can be described as follows: 1þC þ Nres 1 þ C expð−kmax t Þ

C ¼ expðkmax S1 Þ−1

2 ð1Þ ð2Þ

where N is the microbial population (CFU/mL), N0 is the initial microbial population (CFU/mL), C is the amount of protective components in and/or outside the cell (units/cell) defined by Eq. (2), t is exposure time (s), kmax is maximum inactivation rate (1/s), S1 is the parameter representing shoulder length (s) and Nres is the residual microbial population or the parameter representing the tail. Replacing the exposure time by the total pulsed light fluence, Luksiene, Gudelis, Buchovec, and Raudeliuniene (2007) expressed this model as follows:
0 N ¼ ðN 0 −Nres Þ exp −k max H 
0 0 C ¼ exp k max H 1 −1



0

3 Log (N0/N)

Nðt Þ ¼ ðN 0 −Nres Þ expð−kmax t Þ 

1



where H1 is the parameter representing shoulder length (cm2/J), k′max is the maximum inactivation rate (J/cm2) and C′ is the amount of protective components in and/or outside the cell (units/cell) defined by Eq. (4). The confidence interval at a 95% level was calculated for all kinetic parameters describing bacterial inactivation.

0.3

0.5

0.8 1.0 1.3 1.5 Total fluence (J/cm2)

1.8

2.0

0 1 2 3

4 5 6 7 8 9

0

b)

1

2

3 4 5 Total fluence(J/cm2)

6

0 1 2 3 Log (N0/N)

The effect of PL treatments on the inactivation of vegetative and spore forms of B. subtilis at different cell densities was evaluated. For all population densities, reduction in B. subtilis counts (spores and vegetative cells) increased with total fluences (Fig. 1). Concerning the initial microbial population, PL decontamination effectiveness significantly decreased when population density increased, for both vegetative cells and spores of B. subtilis (Fig. 1). These results confirm previous studies with vegetative cells which have shown that PL inactivation efficiency decreased when increasing the initial microbial population in the range tested in this work (Elmnasser et al., 2010; Farrell, Garvey, Laffey, & Rowan, 2009; Gómez-López et al., 2005). Spores were shown to be more resistant to PL than vegetative cells (Fig. 1). As shown in Fig. 1a, for a cell density of 107 cells/mL, a total fluence of 0.5 J/cm2 induced more than 5 Log reductions of vegetative cells, whereas only 3 Log decrease was observed for spores of B. subtilis. Comparable levels of inactivation have been reported previously for spores and vegetative cells of B. subtilis after similar treatment intensities (Levy et al., 2012). By contrast, no differences in the inactivation of spores and vegetative cells of B. subtilis were found in the present study when higher total fluences (≥ 1.1 J/cm2) were applied, which may be influenced by the proximity to the maximum detectable level of inactivation. For a cell density of 108 cells/mL, higher reductions in counts of vegetative cells than spores of B. subtilis were observed for total fluences lower than 1.6 J/cm2 (Fig. 1 b). However, lower inactivations of B. subtilis spores than vegetative cells were shown when increasing the total fluence from 1.6 up to 6 J/cm2. For a population density of 109 cells/mL, no significant differences in the level of inactivation of both physiological states of B. subtilis were found for low total fluences (1 J/cm2), while higher reductions of spores than vegetative cells were shown for total fluences higher than 1 J/cm2 (Fig. 1c). Although some previous published data showed a higher level of inactivation by PL of vegetative cells compared to spores (Dunn et al., 1989;

0.0

a)

3. Results and discussion 3.1. PL inactivation of vegetative cells and spores of B. subtilis at different cell densities

6

9

ð3Þ ð4Þ

5

8



1þC þ Nres 1 þ C expð−k0 max HÞ

4

7

Log (N0/N)



4 5 6 7 8 9

c)

0

1

2

3

4

5 6 7 8 9 10 11 12 Total fluence(J/cm2)

Fig. 1. Influence of total fluence on the PL inactivation of vegetative cells (empty symbols) and spores (filled symbols) of B. subtilis at different cell densities: (a) 107 cells/mL; (b) 108 cells/mL; (c) 109 cells/mL. Error bars indicate the confidence interval at a 95% level. Continuous line shows the maximum level of inactivation.

Levy et al., 2012), those authors did not test different treatment intensities and cell densities. The lethal action of PL treatment on vegetative cells has been explained by photothermal and/or photochemical mechanisms (Wang, Macgregor, Anderson, & Woolsey, 2005; Wekhof, 2000). Concerning bacterial spores, inactivation mechanisms responsible for their inactivation have not been elucidated yet. PL treatment was shown to inhibit spore germination (Artíguez & Martínez de Marañón, 2015), a mechanism which could be involved in the inactivation of bacterial spores. On the other hand, as is the case of UV-C treatment, PL could cause the formation of the “spore photoproduct” 5-thyminyl5,6-dihydrothymine and single-strand breaks, double-strand breaks

M.L. Artíguez, I. Martínez de Marañón / Innovative Food Science and Emerging Technologies 28 (2015) 52–58

55

Table 2 Parameters computed after fitting the “Log-linear with shoulder and/or tail” model of Geeraerd et al. (2000) to experimental data of B. subtilis (Fig. 1). Cells/mL

Cell form

N01

H12

k′max3

Nres4

RMSE5

R26

107

Spore Vegetative Spores Vegetative Spores Vegetative

7.2 ± 0.1a 7.1 ± 0.2a 8.3 ± 0.2b 8.3 ± 0.2b 9.3 ± 0.6c 9.2 ± 0.1c

0.13 ± 0.03a – 0.18 ± 0.03b – 1.29 ± 0.8c –

12.24 ± 0.4a 21.64 ± 4.1b 9.21 ± 0.3c 17.05 ± 3.8d 2.07 ± 0.2e 0.73 ± 0.1f

1 ± 0.08a 1.34 ± 0.1b 1.06 ± 0.1a 1.98 ± 0.3cb 1.44 ± 0.3b 2.88 ± 0.4c

0.132 0.383 0.277 0.528 0.335 0.319

0.99 0.93 0.99 0.92 0.97 0.96

108 109

The 95% confidence interval (CI) was calculated for all kinetic parameters. Different letters in the same column indicate values significantly different (p b 0.05). 1 Initial microbial population (Log CFU/mL). 2 Shoulder length (J/cm2). 3 Maximum inactivation rate (cm2/J). 4 Residual microbial population (Log CFU/mL). 5 Root mean squared error. 6 Coefficient of determination.

physiological state of microorganisms but also on other factors such as cell density which could determine microbial exposure to the incident light. 3.2. Effect of transmittance spectrum of the suspensions of B. subtilis cells on the PL inactivation The quantity of light transmitted through a solution, has been previously reported as a relevant factor determining the antimicrobial effectiveness of PL treatments (Artíguez et al., 2012). Therefore, the light transmittance of B. subtilis suspensions cells was measured in the PL emission range (190–1000 nm). As Fig. 2 shows, the light transmittance of B. subtilis cell suspensions (both sporulated and vegetative state) decreased as the cell density increased, which would explain the lower inactivation found when the B. subtilis population was increased. Due to the low light transmittance at high population densities, microorganisms placed in the upper layers (i.e., closer to the lamp) would be exposed to the total incident light but the rest of cells would be shielded. Similar conclusions were reported by Gómez-López, Devlieghere, Bonduelle and Debevere (2005), who found that PL inactivation effectiveness significantly decreased with the increase in Listeria monocytogenes population on the surface of the agar medium. Regarding liquid products, the same shadowing effect was observed when treating high initial populations (Elmnasser et al., 2010). For the same B. subtilis population density, spore suspensions transmitted a higher amount of light in the range from 190 to 1000 nm than vegetative cell ones (Fig. 2). In the case of high cell densities (109 cells/mL), the limited transmittance through vegetative cells suspensions would play a more important role in PL inactivation than the spore resistance, which could explain the lower inactivation

100 90 80

70 60

%T

and cyclobutane pyrimidine dimers (Nicholson et al., 2000; Setlow, 2006). However, PL could also affect other structures of bacterial spores. The shape of the inactivation curves of B. subtilis exposed to PL was dependent on its physiological state. For spores, inactivation curves were typically sigmoidal, showing an initial shoulder followed by an exponential loss of cell culturability and a final tailing tendency. The “log-linear with shoulder and tail” model (Geeraerd et al., 2000), formulated as Eq. (3), was used to describe the inactivation kinetics of B. subtilis spores. Table 2 shows kinetic parameters obtained after fitting this model to experimental data. The shoulder (H1) observed in B. subtilis spore inactivation curves was more pronounced at higher cell densities (mainly at 109 spores/mL), with values ranging between 0.13 and 1.29 J/cm2. The specific rate of inactivation (k′max) decreased from 12.24 to 2.07 cm2/J when the microbial population increased from 107 to 109 spores/mL. The tail appearing after the exponential inactivation phase could be the consequence of the proximity to the maximum detectable level of inactivation as previously pointed out for several bacterial strains submitted to PL treatments (Lasagabaster & Martínez de Marañón, 2012, 2013). In the case of the highest population density (109 cells/mL), however, a deceleration phase in the inactivation curves was shown (Nres = 1.44 CFU/mL) which could indicate the existence of a real tail. This tail could be due to an ineffective exposure of the microbial cells to the incident light caused by the high number of cells in the suspension. In contrast to those found for spores, inactivation curves without shoulder were shown for vegetative cells. The presence/absence of a shoulder in the inactivation curves could be attributed to differences in cell resistance. Whereas more resistant cells would exhibit a shoulder effect in their inactivation kinetic data (spores in this case), this effect would be absent in the inactivation curves of the most sensitive vegetative cells. Accordingly, Farrell et al. (2009) reported differences in the kinetic behaviour of Gram positive and Gram negative bacteria, showing inactivation curves with an initial shoulder for Gram positive (more resistant to PL) and log linear curves for Gram negative (less resistant). After the exponential phase of inactivation appeared a final tail tendency. The “log-linear with tail” model (Geeraerd et al., 2000) was thus fitted to the inactivation data. The specific rate of inactivation (k′max) decreased from 21.64 to 0.73 cm2/J and the residual population (Nres) increased from 1.34 to 2.88 Log CFU/mL when bacterial density increased from 107 to 109 cells/mL (Table 2). Contrary to spores (expect for 109 spores/mL), the deceleration phase at the end of the inactivation curves (Nres = 2.88 CFU/mL) would indicate the existence of a real tail which could be related to a shielding of microbial cells from the incident light, as previously suggested by others (Bialka, Demirci, Walker, & Puri, 2008; Uesugi, Woodling, & Moraru, 2007). Comparing the parameters obtained for both physiological states of B. subtilis, higher k′max values were observed for vegetative cells than for spores at cell density of 107 and 108 cells/mL, whereas the opposite effect was shown at cell density of 109 cells/mL (Table 2). These results suggest that microbial inactivation by PL would not only depend on the

50 40 30 20 10 0

190 290 390 490 590 690 790 890 990 Wavelength (nm) Fig. 2. Transmittance spectrum in the PL emission range (190–1000 nm) for vegetative cells (light lines) and spores (dark lines) of B. subtilis at a density of 107 cells/mL (▬, ), 108 cells/mL (▬ ▬, ) or 109 cells/mL (▪▪, ).

56

M.L. Artíguez, I. Martínez de Marañón / Innovative Food Science and Emerging Technologies 28 (2015) 52–58

0 1

Log (N0/N)

2

3 4 5 6

0.0

0.1

0.3 0.4 Total fluence (J/cm2)

0.5

0.5

0.8 1.0 1.3 1.5 1.8 Total fluence (J/cm2)

2.0

a)

0 1

Log (N0/N)

2 3 4

5 6 7

0.0

0.3

b)

0 1

Log (N0/N)

2 3

4 5 6 7 8

0.0

0.5

1.0

1.5 2.0 2.5 3.0 3.5 Total fluence (J/cm2)

4.0

c)

0 1

Log (N0/N)

2

3 4 5

rate observed for high density vegetative cell suspensions. Due to the poorer light transmission through vegetative cell suspensions compared to spores, a stronger shading effect would occur in vegetative cell suspensions. Minimal variations in the transmittance of 230–290 nm wavelengths, considered the most effective range for microbial inactivation by PL (Artíguez et al., 2012; Wang et al., 2005), would be particularly relevant. Accordingly, higher levels of Listeria innocua inactivation have been previously reported when a slightly higher amount of light was transmitted in this range through different protein solutions (Artíguez et al., 2012). On the other hand, when population density was low (107 cells/mL) and all microbial cells were exposed to the incident light, the lower susceptibility of spores to PL applied would be due to their inherent resistance. For intermediate cell densities (108 cells/mL), results could be explained by a combined effect of factors, the intrinsic spore resistance and the degree of exposure of microbial cells to the incident light. A considerable number of vegetative cells would be exposed to the light and would be more effectively inactivated than spores. However, deceleration in the inactivation rate of vegetative cells, which was pointed out for high total fluences, would probably be due to the lower exposure of microbial cells caused by the insufficient light penetration into vegetative cell suspensions. In cases in which the inactivation effectiveness is limited by the poor penetration capacity of the light, reactors should be modified to ensure a higher exposure of bacteria to the incident light. As previously shown for UV processing (Keyser, Muller, Cilliers, Nel, & Gouws, 2008; Koutchma, Keller, Chirtel, & Parisi, 2004), treatments of a thin layer of liquid or high turbulent flow and mixing conditions could increase the exposure of microbial cells to the effective wavelengths, and therefore, the PL antimicrobial effectiveness. 3.3. PL inactivation of vegetative cells and spores of G. stearothermophilus at different cell densities The effect of the physiological state of bacterial cells and the exposure of microorganisms to the emitted light on the antimicrobial effectiveness of PL was evaluated using a different bacterium, G. stearothermophilus. As mentioned for B. subtilis, the antimicrobial effectiveness of PL technology would depend on the microbial density: the inactivation ratio decreased when the initial population of G. stearothermophilus cells was increased (Fig. 3). The lower inactivation level as the microbial population increased would be due to the lower amount of light transmitted through G. stearothermophilus suspensions (both sporulated and vegetative cell suspensions) when cell density was raised (Fig. 4). Higher reductions of vegetative cells than of spores of G. stearothermophilus were found when suspensions at cell densities of 105 and 106 cells/mL were exposed to total fluence lower than 0.32 J/cm2 and 0.28 J/cm2, respectively (Fig. 3a and b). However, vegetative cells showed lower inactivation levels than spores when treatment intensity was increased until the maximum detectable level of inactivation was reached. Although spores would be more resistant to PL than vegetative cells, due to the lower light transmittance of vegetative cell suspensions compared to spore ones, at equal cell density, vegetative cells would be exposed to less light and would be less effectively inactivated (Fig. 4). For a cell density of 107 cells/mL, no significant differences in the behaviour of vegetative cells and spores were detected (Fig. 3c). In this case, the greater PL resistance of bacterial spores would be offset by the lower transmittance of vegetative cell suspensions (Fig. 4). The highest cell density of G. stearothermophilus tested (108 cells/mL), produced only a slight cell inactivation both in spores and vegetative cells (Fig. 3d). Insufficient transmittance through the microbial suspension,

6

0

1

2

3

4 5 6 7 8 9 10 11 12 d) Total fluence (J/cm2)

Fig. 3. Influence of total fluence on the PL inactivation of vegetative cells (empty symbols) and spores (filled symbols) of G. stearothermophillus at different cell densities: (a) 105 cells/mL; (b) 106 cells/mL; (c) 107 cells/mL; (d) 108 cells/mL. Error bars indicate the confidence interval at 95% level. Continuous line shows the maximum level of inactivation.

M.L. Artíguez, I. Martínez de Marañón / Innovative Food Science and Emerging Technologies 28 (2015) 52–58

100

90 80 70 %T

60

50 40 30 20

10 0 190 290 390 490 590 690 790 890 990 Wavelength (nm) Fig. 4. Transmittance spectrum in the PL emission range (190–1000 nm) for vegetative cells (light lines) and spores (dark lines) of B. subtilis at a density of 105 cells/mL (▬, ) 106 cells/mL (▬ ▪, ), 107 cells/mL (▬ ▬, ) or 108 cells/mL (▪▪, ).

in particular from 230 to 290 nm, would explain the limited reduction found at high population densities. Although PL treatment was ineffective for inactivating G. stearothermophilus at high cell density, this technology could be successfully applied to inactivate this microorganism at the cell density found in food products, which would be useful for the food industry due to the extreme heat resistance of G. stearothermophilus spores (Feeherry, Munsey, & Rowley, 1987). As previously shown for B. subtilis cells, the shape of the inactivation curves was dependent on the physiological state of G. stearothermophilus cells. Spores showed typically sigmoidal inactivation curves, with an initial shoulder followed by an exponential loss of cell culturability and a final tailing tendency. Except for the highest cell density tested (108 cells/mL), no initial shoulder was observed in the inactivation curves of vegetative cells due to the fact that the weakest treatment applied caused a large cell inactivation (Fig. 3). Both the “log-linear with shoulder and tail” and the “log-linear with tail” models were then fitted to the experimental data. The kinetic parameters obtained by fitting models are shown in Table 3. Shoulder length (H1) value increased from 0.1 to 0.34 J/cm2 when spore density increased from 105 to 107 spores/mL. The specific rate of inactivation (k′max) decreased from 37.59 to 0.46 cm2/J for spore populations and from 18.94 to 0.32 cm2/J for vegetative cells when bacterial density was increased from 105 to 108 cells/mL, respectively (Table 3). Higher H1 and lower k′max and values were apparent especially for higher cell densities. Except for the highest cell density tested (108 cells/mL), the tailing appearing after the exponential inactivation phase would be due to the proximity to the maximum detectable level of inactivation. In the case of the highest cell density, however, this tailing (Nres value of 8.51 CFU/mL for spores and 7.43 CFU/mL for

57

vegetative cells) could be associated with a shielding of microbial cells from the incident light. In contrast to the B. subtilis populations, higher k′max values were observed for spores than for vegetative cells at cell density of 105 and 106 cells/mL, whereas no differences were found at cell density of 107 or 108 cells/mL (Table 3). These results would suggest that G. stearothermophilus inactivation by PL would be more influenced by the exposure of microbial cells to the light than by the inherent spore resistance. Comparing the sensitivity to PL of both species at the same cell density, populations of G. stearothermophilus seem to be more resistant to PL than B. subtilis cells. One single light pulse of only 0.5 J/cm2 caused a considerable inactivation of spores (N 2 Log) and vegetative cells (N4 Log) of B. subtilis at 108 cells/mL, whereas inactivation was lower than 1 Log in spores and vegetative cells of G. stearothermophilus. Although a higher sensitivity to PL of B. subtilis spores compared to G. stearothermophilus spores at equal cell density has been described previously (Chaine et al., 2012; Levy et al., 2012), inactivation rates in vegetative state were not compared in those previous works. As B. subtilis suspensions (both spores and vegetative cells) transmitted a higher amount of light than G. stearothermophilus suspensions, the higher reduction in cell count observed for B. subtilis could be due to the higher exposure of the microbial cells to the light. When considering vegetative cell suspensions of both microorganisms with similar transmittance in the range from 190 to 1000 nm, and in particular from 230 to 290 nm, but different cell density (B. subtilis at 108 vegetative cells/mL and G. stearothermophilus at 107 vegetative cells/mL for example), a higher specific rate of inactivation (k′max) was found for vegetative cells of B. subtilis than for G. stearothermophilus (Tables 2 and 3). By contrast, when comparing spore population with similar transmittance (B. subtilis at 107 spores/mL and G. stearothermophilus at 106 spore cells/mL for example), B. subtilis spores showed a lower k′max value than G. stearothermophilus spores. Therefore, vegetative cells of B. subtilis were more sensitive to PL than the ones of G. stearothermophilus, considering both cell density and transmittance in the range 230– 290 nm. However, spores of G. stearothermophilus were the less resistant when cell suspensions with similar light transmittance were compared. 4. Conclusions This study indicates that the inactivation effectiveness of PL would not only depend on the physiological state of the cells, but also on their exposure to the incident light, which could be influenced by cell population density among other factors (e.g., presence of particles in the solution). Although the efficiency of PL treatment has been previously demonstrated to inactivate both vegetative and spore forms, this is the first study, to our knowledge, that reveals the importance of both factors, the inherent resistance of microbial cells and the

Table 3 Parameters computed after fitting the “Log-linear with shoulder and/or tail” model of Geeraerd et al. (2000) to experimental data of G. stearothermophilus (Fig. 3). Cells/mL

Cell form

N01

H12

k′max3

Nres4

RMSE5

R26

105

Spore Vegetative Spore Vegetative Spores Vegetative Spores Vegetative

5.3 ± 0.05a 5.1 ± 0.2a 6.1 ± 0.07b 6.2 ± 0.2b 7.3 ± 0.2c 6.88 ± 0.2c 8.3 ± 0.05d 8.3 ± 0.04d

0.12 ± 0.01a – 0.13 ± 0.01a – 0.34 ± 0.1b – – –

37.59 ± 1.3a 18.94 ± 2.63b 28.99 ± 0.8c 12.74 ± 1.4d 4.89 ± 0.9e 5.47 ± 0.6e 0.46 ± 0.4f 0.32 ± 0.06f

1 ± 0.04a – 1.01 ± 0.04a 1.21 ± 0.2a 1.23 ± 0.4a 1.15 ± 0.3a 8.51 ± 0.1b 7.43 ± 0.13c

0.053 0.608 0.074 0.266 0.693 0.456 0.061 0.087

0.99 0.90 0.99 0.98 0.95 0.97 0.97 0.94

106 107 108

The 95% confidence interval (CI) was calculated for all kinetic parameters. Different letters in the same column indicate values significantly different (p b 0.05). 1 Initial microbial population (Log CFU/mL). 2 Shoulder length (J/cm2). 3 Maximum inactivation rate (cm2/J). 4 Residual microbial population (Log CFU/mL). 5 Root mean squared error. 6 Coefficient of determination.

58

M.L. Artíguez, I. Martínez de Marañón / Innovative Food Science and Emerging Technologies 28 (2015) 52–58

transmittance of light across the cell suspension (i.e., the level of cell exposure to light). By taking into account all results, it can be concluded that the feasibility to decontaminate suspensions containing B. subtilis would be easier than G. stearothermophilus ones. Future research should be conducted to investigate the behaviour of bacterial spores in complex food systems, in particular in fluids with limited light transmittance, to determine the effectiveness of PL for the inactivation of these microbial forms at cell densities currently found in liquid food products. Acknowledgments Support for this work was provided by the Department of Environment, Territorial Planning, Agriculture and Fisheries of the Basque Government. M. L. Artíguez was funded by a PhD grant of the Department of Education, Universities and Research of the Basque Government. References Artíguez, M. L., Arboleya, J. -C., & Martínez de Marañón, I. (2012). Influence of β-lactoglobulin and β-casein on Listeria innocua inactivation by pulsed light. International Journal of Food Microbiology, 153(1–2), 224–229. Artíguez, M. L., & Martínez de Marañón, I. (2014). Process parameters affecting Listeria innocua inactivation by pulsed light. Food and Bioprocess Technology, 7, 2759–2765. Artíguez, M. L., & Martínez de Marañón, I. (2015). Effect of pulsed light treatment on the germination of Bacillus subtilis spores. Food and Bioprocess Technology. http://dx.doi. org/10.1007/s11947-014-1433-4. Bialka, K. L., Demirci, A., Walker, P. N., & Puri, V. M. (2008). Pulsed UV-light penetration of characterization and the inactivation of Escherichia coli K12 in solid model systems. Transactions of the Asabe, 51(1), 195–204. Burgess, S. A., Lindsay, D., & Flint, S. H. (2010). Thermophilic bacilli and their importance in dairy processing. International Journal of Food Microbiology, 144(2), 215–225. Buschnell, A., Cooper, J. R., Dunn, J., Leo, F., & May, R. (1998). Pulsed light sterilization tunnels and sterile pass throughs. Pharmaceutical Engineering, 18, 48–58. Chaine, A., Levy, C., Lacour, B., Riedel, C., & Carlin, F. (2012). Decontamination of sugar syrup by pulsed light. Journal of Food Protection, 75(5), 913–917. Dunn, J., Clark, R., Asmus, J. F., Pearlmann, J. S., Boyer, K., Painchaud, F., & Hoffmann, G. A. (1989). Method for preservation of foodstuffs. [07/187281]. Elmnasser, N., Federighi, M., Bakhrouf, A., & Orange, N. (2010). Effectiveness of pulsed ultraviolet light treatment for bacterial inactivation on agar surface and liquid medium. Foodborne Pathogens and Disease, 7(11), 1401–1406. Farrell, H. P., Garvey, M., Laffey, J. G., & Rowan, N. J. (2009). Investigation of critical interrelated factors affecting the efficacy of pulsed light for inactivating clinically relevant bacterial pathogens. Journal of Applied Microbiology, 108, 1494–1508. Feeherry, F., Munsey, D., & Rowley, D. (1987). Thermal inactivation and injury of Bacillus stearothermophilus spores. Applied and Environmental Microbiology, 53, 365–370. Finley, N., & Fields, M. L. (1962). Heat activation and heat-induced dormancy of Bacillus stearothermophilus spores. Applied Microbiology, 10(3), 231–236.

From, C., Pukall, R., Schumann, P., Hormazábal, V., & Granum, P. E. (2005). Toxin producing ability among Bacillus spp. outside the Bacillus cereus group. Applied and Environmental Microbiology, 71(3), 1178–1183. Geeraerd, A. H., Herremans, C. H., & Van Impe, J. F. (2000). Structural model requirements to describe microbial inactivation during a mild heat treatment. International Journal of Food Microbiology, 59(3), 185–209. Geeraerd, A. H., Valdramidis, V. P., & Van Impe, J. F. (2005). GInaFiT, a freeware tool to assess non-log-linear microbial survivor curves. International Journal of Food Microbiology, 102, 95–105. Gómez-López, V. M., Devlieghere, F., Bonduelle, V., & Debevere, J. (2005). Factors affecting the inactivation of microorganisms by intense light pulses. Journal of Applied Microbiology, 99(3), 460–470. Huffman, D. E., Slifko, T. R., Salisbury, K., & Rose, J. B. (2000). Inactivation of bacteria, virus and Cryptosporidium by a point-of-use device using pulsed broad white light. Water Research, 34(9), 2491–2498. Keyser, M., Muller, I. A., Cilliers, F. P., Nel, W., & Gouws, P. A. (2008). Ultraviolet radiation as a non-thermal treatment for the inactivation of microorganisms in fruit juice. Innovative Food Science and Emerging Technologies, 9(3), 348–354. Koutchma, T., Keller, S., Chirtel, S., & Parisi, B. (2004). Ultraviolet disinfection of juice products in laminar and turbulent flow reactors. Innovative Food Science and Emerging Technologies, 5(2), 179–189. Lasagabaster, A., & Martínez de Marañón, I. (2012). Sensitivity to pulsed light technology of several spoilage and pathogen bacteria isolated from fish products. Journal of Food Protection, 75(11), 2039–2044. Lasagabaster, A., & Martínez de Marañón, I. (2013). Impact of process parameters on L. innocua inactivation kinetics by pulsed light technology. Food and Bioprocess Technology, 6, 1828–1836. Levy, C., Aubert, X., Lacour, B., & Carlin, F. (2012). Relevant factors affecting microbial surface decontamination by pulsed light. International Journal of Food Microbiology, 152, 168–174. Luksiene, Z., Gudelis, V., Buchovec, I., & Raudeliuniene, J. (2007). Advanced high-power pulsed light device to decontaminate food from pathogens: Effects on Salmonella typhimurium viability in vitro. Journal of Applied Microbiology, 103, 1545–1552. Nicholson, W. L., Munakata, N., Horneck, G., Melosh, H. J., & Setlow, P. (2000). Resistance of Bacillus endospores to extreme terrestrial and extraterrestrial environments. Microbiology and Molecular Biology Reviews, 64(3), 548–572. Prentice, G. A., Wolfe, F. H., & Clegg, L. F. L. (1972). The use of density gradient centrifugation for the separation of germinated from nongerminated spores. Journal of Applied Microbiology, 35(2), 345–349. Rice, J. K., & Ewell, M. (2001). Examination of peak power dependence in the UV inactivation of bacterial spores. Applied and Environmental Microbiology, 67, 5830–5832. Setlow, P. (2006). Spores of Bacillus subtilis: their resistance to and killing by radiation, heat and chemicals. Journal of Applied Microbiology, 101(3), 514–525. Uesugi, A. R., Woodling, S. E., & Moraru, C. I. (2007). Inactivation kinetics and factors of variability in the pulsed light treatment of Listeria innocua cells. Journal of Food Protection, 70, 2518–2525. Wang, T., Macgregor, S., Anderson, J., & Woolsey, G. (2005). Pulsed ultra-violet inactivation spectrum of Escherichia coli. Water Research, 39(13), 2921–2925. Wekhof, A. (2000). Disinfection with flash lamps. PDA Journal of Pharmaceutical Science and Technology, 54(3), 264–276.