Food Bioprocess Technol (2014) 7:2267–2280 DOI 10.1007/s11947-013-1197-2
ORIGINAL PAPER
Effect of Hyperbaric Pressure Treatment on the Growth and Physiology of Bacteria that Cause Decay in Fruit and Vegetables Pansa Liplap & Vicky Toussaint & Peter Toivonen & Clément Vigneault & Jérôme Boutin & G. S. Vijaya Raghavan
Received: 11 July 2013 / Accepted: 22 September 2013 / Published online: 3 October 2013 # Springer Science+Business Media New York 2013
Abstract The response of bacteria to hyperbaric pressure treatment was investigated. Three selected bacteria which cause fruit and vegetable decay (i.e., Pseudomonas cichorii, Pectobacterium carotovorum, and Pseudomonas marginalis) were inoculated onto BIOLOG microplates and subjected to different pressure and temperature conditions including 100, 200, 400, 625, and 850 kPa at 20 °C and 100 kPa at 4 °C. Changes in microplate color, which corresponds to carbon source utilization of bacteria or their growth, were monitored every 24 h for 7 days. Results showed that the bacterial growth was affected by both hyperbaric pressure and temperature. As hyperbaric pressure increased, the bacterial growth significantly decreased and the extent was dependent on bacterial species. The 850-kPa pressure treatment reduced maximum growth by 71, 56, and 43 % for Pseudomonas cichorii, Pectobacterium carotovorum, and Pseudomonas marginalis, respectively. Among these bacteria, Pseudomonas cichorii was the most pressure-sensitive, while the most temperaturesensitive was Pectobacterium carotovorum . In general, an increase in hyperbaric pressure caused bacteria to utilize carbon sources similar to those when they were exposed to low temperature. Overall, hyperbaric treatment has the potential to P. Liplap (*) : C. Vigneault : G. S. V. Raghavan Department of Bioresource Engineering, Macdonald Campus, McGill University, Sainte-Anne-de-Bellevue, Montreal, QC, Canada H9X 3V9 e-mail:
[email protected] P. Liplap : V. Toussaint : C. Vigneault : J. Boutin Horticulture Research and Development Centre, Agriculture and Agri-Food Canada, Saint-Jean-sur-Richelieu, QC, Canada P. Toivonen Pacific Agri-Food Research Centre, Agriculture and Agri-Food Canada, Summerland, BC, Canada
directly reduce bacterial growth in fruit and vegetables after harvest. Keywords Low temperature . BIOLOG . Carbon source utilization . Bacterial growth
Introduction Several studies have shown the close relationship between a high consumption of fruit and vegetables and lower incidence of some chronic diseases (Lir et al. 2000; Van't Veer et al. 2000). Thus, demand for fresh fruit and vegetables has increased over recent decades. In contrast, their availability in the form of processed fruit and vegetable products is falling behind the gross consumption (Barth et al. 2009). However, increased fruit and vegetable production is not the solution to this problem of availability because approximately 15–40 % of the production volume is lost between harvest and consumption (James and James 2010) with spoilage being one of the major causes for these losses (Kantor et al. 1997). Since fruit and vegetable tissues provide environments conductive for survival and growth of microorganisms, they will inevitably succumb to microbiological attack and spoilage. Although decay-causing microorganisms may be present on fruit and vegetables at harvest, rots generally develop during transport and storage (Barth et al. 2009). Loss due to decay by rotcausing pathogens during postharvest handling of fruit and vegetables is estimated to 20–25 % in developed countries, and the situation is much worse for developing countries (EI-Ghaouth et al. 2004; Korsten 2006; Sharma et al. 2009). Given this situation, improving management of freshly harvested fruit and vegetables to control rots is crucial to provide commodities with high quality standards.
2268
Several different strategies have been used to control decay pathogens in fresh fruit and vegetables. Among them, biological control of postharvest disease using microbial antagonists like yeasts and bacteria has been reported to inhibit decay of fruit effectively (Sharma et al. 2009; Korsten 2006; Droby et al. 2003). However, the application of antagonistic microorganisms is limited due to difficulty in management of the storage environment and other postharvest treatments required to achieve sufficient protection from the biocontrol agent (Janisiewicz and Korsten 2002). Synthetic chemicals, including pesticides and preservatives, have been widely applied to fresh produce. Nevertheless, use of chemicals is not the most desirable means of disease control because they are generally expensive, can cause environmental pollution, and may induce pathogen resistance. But most importantly, the presence of chemical residues leads to harmful effects on the human health by interfering with the reproductive systems and fetal development as well as in their capacity to cause cancer and asthma (Gilden et al. 2010). Therefore, use of chemical treatments as postharvest treatments has been so far restricted. Recently, physical treatments, either alone or in combination, such as heat (such as water and vapor), irradiation, UV-B/C, modified/controlled atmosphere, pressure, etc., have drawn great attention in the postharvest field because they not only control insect and storage diseases but can also maintain or even improve eating qualities of stored fruit and vegetables (Charles and Arul 2007; Vigneault et al. 2012) Of the physical means, hyperbaric treatment has shown to improve quality preservation on some horticultural produce, i.e., tomato (Goyette et al. 2012; Liplap et al. 2013b, c), lettuce (Liplap et al. 2013a), mume fruit (Baba et al. 1999), cherry, and grape (Romanazzi et al. 2008). The basic operation consists of subjecting produce to a pressurized air environment ranging from 0.1 to 1.0 MPa (1–10 atm), in which the proportion of air composition is maintained (Goyette et al. 2007). It is different from conventional high pressure treatment, where the pressure is set between 400 and 1,200 MPa (Ahmed and Ramaswamy 2006). An application of such high pressures is generally not suitable for fruit and vegetables because they can cause irreversible damage to cell structure (Ahmed and Ramaswamy 2006; Baba and Ikeda 2003). The main advantage of hyperbaric pressure treatment over the other physical treatments is the homogeneity of response, as it acts instantaneously and uniformly throughout a mass of product independently of the size and the shape (Goyette et al. 2007). The inhibitory effects of high pressure on the growth of several microorganisms have been widely reported (San Martín et al. 2002). In general, very high pressure is required to kill or inactivate the growth of microorganisms. Nevertheless, recent studies have shown that microorganisms were affected by hyperbaric treatment as well. Romanazzi et al. (2008) showed that hyperbaric treatment (150 kPa) applied for 4 h significantly reduced the incidence of brown rot, gray and blue molds, and
Food Bioprocess Technol (2014) 7:2267–2280
total rots (including Alternaria sp., Rhizopus sp., and Cladosporium sp. rots) of sweet cherries. They also found that wounded table grapes, subjected to 150 kPa pressure for 24 h and then inoculated with Botrytis cinerea, had significantly fewer berries infected and lesions were less severe. Similar results were observed from our previous study (Liplap et al. 2013a), where the lettuce was treated with varying pressures ranging from 100 to 850 kPa at 20 °C for 5 days. The development of decay was dramatically delayed when lettuce was stored under hyperbaric pressure, especially at 850 kPa, in comparison with ambient pressure. Nevertheless, the mode of action of hyperbaric treatment on the growth of microorganisms has not been clearly elucidated. It is probably due to the direct impact of elevated pressure itself on the microorganisms (Segovia-Bravo et al. 2012), or the elevated O2, causing O2 toxicity to microorganisms, bacteria, yeasts, and molds (Kader and Ben-Yehoshua 2000; Bean 1945), or the enhancement of the host pathogen defense compound synthesis, induced by mild stress as demonstrated for UV-C and heat treatments (Charles and Arul 2007; Lu et al. 2010). To date, very limited information is available concerning hyperbaric pressure effects on decay-causing microorganisms of fruit and vegetables. Studies are required to understand the mechanisms of hyperbaric treatment on microorganism inactivation. In this study, the physiology of three selected bacteria causing fruit and vegetable decay (i.e., Pseudomonas cichorii, Pectobacterium carotovorum, and Pseudomonas marginalis) under different hyperbaric pressures (from 100 to 850 kPa) was investigated. BIOLOG microplates (GEN III) containing 71 carbon sources were used to follow bacteria development under the different hyperbaric pressure treatments, thus eliminating the potential interference of a plant self-defense mechanism. The hyperbaric treatments were applied at 20 °C and their carbon source utilization behaviors were compared to that of storage at 4 °C at ambient atmospheric pressure.
Materials and Methods Species, Bacterial Inoculum, and Microplates Pseudomonas marginalis , Pseudomonas cichorii , and Pectobacterium carotovorum used in this study were isolated from rotting lettuce tissues (Liplap et al. 2013a). For isolation of bacteria, pieces of lettuce from different samples with rotting symptoms were sequentially washed under running distilled water, cut into small pieces, and macerated in 10 mM of phosphate buffer (pH 7.0) for 30 min. Macerate was then diluted 1:10 and streaked onto King's B agar medium (KB) (King et al. 1954) for Pseudomonas spp. recovery and Nutrient Agar (NA) (Becton Dickinson and Company, Sparks, MD, USA) for other species, both media were supplemented with 50 mg L−1 of cycloheximide (Sigma Aldrich
Food Bioprocess Technol (2014) 7:2267–2280
Co., St. Louis, MO, USA) to avoid fungal growth. Bacteria were grown on media at 28 °C for 48 h. After the incubation period, predominant colonies on plates were purified by streaking on KB (for fluorescent colonies) or NA. Fluorescent Pseudomonas isolates were identified based on LOPAT profile and identification confirmed using the GEN III Microbial Identification System (BIOLOG, Hayward, CA, USA). Pectobacterium was identified using GEN III and the pectolytic activity was verified using surface sterilized potato slices (De Boer and Kelman 2001). The bacterial inocula were prepared for each species. A bacterial suspension was adjusted to A 600 =0.09 in 10 mM phosphate buffer pH 7.0 which corresponds to about 108 CFU mL-1. The final concentrations of these inocula were verified by dilution plating onto KB medium. These bacterial suspensions were used immediately to inoculate BIOLOG GEN III microplates. For microplate preparation, the inoculating protocol of the manufacturer for the GEN III was followed as described in the user guide, with the exception that the inoculation fluid was inoculated using 10 μL of the bacterial suspension prepared as described above. The inoculated microplates were then transferred into each hyperbaric vessel with three plates per species and three species per treatment. The experiment was conducted twice. Hyperbaric Treatment System Hyperbaric pressure system used during the tests is illustrated in Fig. 1. The system is similar to the one previously used for horticultural produce storage, where pressure, air flow rate, and relative humidity are controlled (Liplap et al. 2013a, b). Briefly, it consists of a compressed air tank, three low pressure vessels, three high pressure vessels, and an infrared gas analyzer. The low pressure vessels are made from painted steel apparatus (PRO-TEK, Mirabel, Quebec, Canada), 220 mm in height and 265 mm inside diameter, whereas the high pressure vessels are made from stainless steel (GracoTM, Minneapolis, MN, USA), 225 mm high and 235 mm inside diameter. The net volume of each vessel is 10.75 L. A 12.7-mm flat rubber ring is used to ensure air tightness between the cover and the chamber. Each vessel is equipped with a pressure regulator and a flow control needle valve to individually regulate the pressure and air flow rate, respectively. A safety relief valve is used to prevent pressure overload. Two compression fittings are fastened to the vessel to quick connect the airflow inlet and outlet using plastic tubes of 3.2 mm inner diameter. The air inlet of the vessel is connected to a compressed air tank equipped with a regulating manometer. A channel selector (a manifold equipped with controlled valves) is installed to connect the air outlet of the selected vessel to a CO2 infrared gas analyzer (Guardian® Plus, Kirkton Capus, Livingston, UK) and the electronic air flow meter (Bronkhorst TM , Ruurlo, Netherlands). The manifold equipped with controlled valves,
2269
air flow meter, and CO2 gas analyzer are all connected to a data acquisition and control card (Personnel DAQ 3000, Cleveland, Ohio, USA) and a personal computer. Experimental Procedure Prior to the experiment, the hyperbaric system was sanitized to avoid contamination of other microorganisms from the air. The pressurized air was filtered through a series of filters: a 20-μm, a 0.03-μm, and a 0.01-μm filter. The air supply line was disinfected with a 10 % v/v sodium hypochlorite solution. The storage vessels were cleaned with 70 % v/v ethanol solution prior to use. Each of the inoculated BIOLOG GEN III microplates were submitted to one of the following pressure and temperature conditions: 100, 200, 400, 625, or 850 kPa at 20 °C or 100 kPa at 4 °C. The 100 kPa at 20 °C treatment was the control and 100 kPa at 4 °C was the cold treatment. The vessels were then sealed and pressurized for 7 days. The microplates were taken out and read every 24 h up to 168 h (day 7) using a microplate reader (Model ELx808BLG, BioTek instruments, Winooski, VT). Absorbance was read at 590 nm since respiration of the carbon source in each well induces the tetrazolium dye color development (purple) with a peak of absorbance at this wavelength. After the 7th day, the plates were kept at 100 kPa, 20 °C (ambient conditions) and the changes in color were monitored until 240 h (day 10). BIOLOG Data Treatment BIOLOG is a 96-well microplate technique originally developed for a rapid identification of bacteria (Garland and Mills 1991). In addition to that role, it is found useful in investigating the response of bacteria to different nutrients and environmental conditions. The BIOLOG microplate GEN III model analyzes microorganisms in 94 phenotypic tests: 71 carbon source utilization assays and 23 chemical sensitivity assays, with one negative control (no carbon source) and one positive control used as a reference for chemical sensitivity assays. The microbial activity in the cells is determined based on the exchange of electrons generated during respiration, leading subsequently to tetrazolium-based color changes. Therefore, an increase in color development indicates the nutrient utilization, which corresponds to bacterial growth. Each microplate provides a set of 95 data per reading; consequently, to allow easier interpretation, the analysis was carried out through groups of substrates: (1) all wells (n =95), all carbon sources (n =71), sugars (n =26), sugar alcohols (n =5), amino acids (n =11), hexose acids (n =9), and the others (carboxylic acids, esters, and fatty acids; n =18). Overall changes in color development were expressed as average well color development (AWCD) based on the classified substrate groups. AWCD at a given time was derived from the mean difference among absorbance values of the selected carbon sources (C ) and the
2270
Food Bioprocess Technol (2014) 7:2267–2280
Fig. 1 Hyperbaric system for evaluation of the effect of different pressure levels on bacterial growth
absorbance value of the control well (R: no carbon source) as shown in Eq. 1. The negative absorbance values obtained from C-R term were set to zero prior to calculating a mean. X AWCDsub:group ¼
ðC−RÞ n
ð1Þ
where: AWCDsub.group C R N
average well color development (AWCD) of each substrate group absorbance value of each well absorbance value of the control well number of wells in the substrate group
Mean AWCD values were plotted over time for each substrate group. As the area under the curve (AUC) corresponds to the growth of bacteria (asymptote, growth rate, lag time, etc.) and explains combined effect of color development (PrestonMafham et al. 2002), it was used to quantify the effect of different treatment conditions. In this study, AUC was estimated using trapezoid approach (Eq. 2). The AUC method is particularly useful when the data do not fit the logarithmic form. Trapezoidal area ¼
n X
½ðV i þ V i−1 Þ ðt i −t i−1 Þ
i¼1
where: v t n
absorbance value (absorbance unit) incubation time (hour) number of reading
ð2Þ
The predictive modeling was performed to describe the behavior of bacterial growth under different hyperbaric treatment conditions. Also, the model allows prediction of microbial safety or product shelf life (Zwietering et al. 1990; Cayre et al. 2005). In general, bacterial growth shows a phase in which the specific growth rate (μ M) starts a value of zero and then accelerates to a maximum value (A) in a certain period of time, resulting in a lag time (λ) (Fig. 2). Due to the sigmoidal shape of the color development, logistic curve is used as the basis for curve fitting models. A number of microbial growth models are found in literature, such as Richards, Stannard, Schute, logistic model and others (Zwietering et al. 1990). In this study, the AWCD of all carbon sources was fitted to the modified Gompertz equation (Eq. 3), a widely used asymmetric logistic model in which the exponential increase is corrected by an exponential term (Zwietering et al. 1990; Vandepitte et al. 1995). The obtained kinetic model parameters (A , μ M, and λ) are useful in investigating the behavior of bacteria. n hμ e io yðt Þ ¼ Aexp −exp M ðλ−t Þ þ 1 A
ð3Þ
where: y (t) A μM λ e
average well color development of all carbon sources (absorbance value) maximum absorbance value (absorbance unit) specific growth rate (absorbance unit per hour) lag time (hour) = exp (1)
Food Bioprocess Technol (2014) 7:2267–2280
Principal component analysis (PCA) was also performed on AUC data of all substrates to distinguish growth pattern of treatment conditions. PCA reduces a multivariate data set and projects original data onto a small number of principal components (PCs). Each PC extracts a portion of the variance from the original data, with the greatest amount of variance extracted by the first axis. Relationships among treatment conditions were obtained by plotting scores of the first two PCs in two dimensions, allowing comparison on the basis of different patterns of carbon source utilization.
wells) utilization is shown in Fig. 3. It is clearly shown that temperature was the most significant factor affecting the growth of bacteria. The bacteria treated at 4 °C grew slowly throughout the 7 days of treatment and did not fully develop to the sigmoidal curve. This is not surprising because of reduced microbial activity at low temperature. Interestingly, an increase in surrounding pressure at 20 °C clearly showed a positive impact on the growth behavior of the bacteria. The higher the pressure applied, the lower the microbial growth. In general, at 20 °C, the curves followed a sigmoidal curve with different extents of maximum growth and their corresponding times. The exponential growth phase at 20 °C ranged from 24 to 72 h and was also dependent on the types of bacteria.
1.0
Pseudomonas cichorii
0.8 AWCD
Fig. 2 An example of bacterial growth curve expressed by average well color development (AWCD). The fitting Gompertz equation parameters A, μ M, and λ indicate maximum rate color development, specific growth rate, and lag time, respectively
2271
0.6 0.4 0.2 0.0 1.0 0.8
The experiments were conducted in a two-way factorial design (bacterium and treatment) with three replications and conducted twice. An analysis of variance was performed on the results (i.e., AUC and fitting Gompertz equation parameters: A, μ M, λ) using the General Linear Model with SAS software (SAS Institute, Cary, NC, USA), and significant differences between treatment means were determined using Duncan's multiple range test with a 95 % confidence interval (P