Cross-protective effect of acid-adapted Salmonella enterica on resistance to lethal acid and cold stress conditions

Authors

  • H. Xu,

    1.  Division of Biomaterials Engineering, Kangwon National University, Chuncheon, Gangwon, Republic of Korea
    Search for more papers by this author
  • H. Y. Lee,

    1.  Division of Biomaterials Engineering, Kangwon National University, Chuncheon, Gangwon, Republic of Korea
    2.  Institute of Bioscience and Biotechnology, Kangwon National University, Chuncheon, Gangwon, Republic of Korea
    Search for more papers by this author
  • J. Ahn

    1.  Division of Biomaterials Engineering, Kangwon National University, Chuncheon, Gangwon, Republic of Korea
    2.  Institute of Bioscience and Biotechnology, Kangwon National University, Chuncheon, Gangwon, Republic of Korea
    Search for more papers by this author

Juhee Ahn, Institute of Bioscience and Biotechnology, Kangwon National University, Chuncheon, Gangwon 200-701, Republic of Korea. E-mail: juheeahn@kangwon.ac.kr

Abstract

Aims:  To evaluate the cross-protected Salmonella enterica cells under acid and cold stress conditions.

Methods and Results:  The acid-adapted S. enterica cells were exposed to pH 4·0 at 4 and 20°C. Recovery of sublethally injured cells was estimated by the difference between the counts obtained on trypticase soy agar (TSA) and xylose lysine desoxycholate (XLD) agar. The survival curves of nonadapted and acid-adapted S. enterica cells at pH 4·0 were fitted with Weibull distribution model. The recovery behaviour of injured S. enterica cells was estimated by the modified Gompertz parameters. Acid-adapted S. enterica were more resistant to subsequent acid shock than the nonadapted cells. The numbers of nonadapted S. enterica cells were decreased by 4·57 and 7·55 log CFU ml−1 at 4 and 20°C after 12-day acid challenge, respectively. The acid adaptation induced cross-protection and viable nonculturable (VBNC) state against low acid and cold stresses. The 7-h adaptation showed the least recovery of injured cells.

Conclusion:  The results suggest that acid-adapted S. enterica cells induced acid tolerance response and VBNC state.

Significance and Impact of the Study:  These results provide useful information for understanding the induction of cross-protected and VBNC pathogens under various stresses, which might be needed in designing new food preservation strategies.

Introduction

As the control of microbial contamination has been a great concern to the food industry, various techniques have been employed for improving food quality and safety, including heat, pressure, radiation, osmotic shock, acids, salts, antimicrobials and biocontrol (Jay et al. 2005; Rodriguez-Romo and Yousef 2005). However, foodborne pathogens can be adapted under these preservative and processing methods, leading to enhanced resistance and potentially serious clinical diseases. Adaptation of foodborne pathogens to food processing-related stresses may result in a significant increase in the food safety risk (Koutsoumanis et al. 2003). Therefore, food processing-related stresses on adaptive tolerance of the foodborne pathogens have raised public awareness of food safety.

Acids and low temperature are among the most commonly used as food preservatives encountered by foodborne pathogens (Foster and Spector 1995; Greenacre et al. 2003). For instance, acetic acid is widely employed as a food acidulant for inhibiting microbial growth and extend the shelf life of foods. Numerous gram-negative and -positive bacteria including Escherichia coli O157:H7 (Cheng et al. 2003; Greenacre et al. 2003; Bjornsdottir et al. 2006), Listeria monocytogenes (Phan-Thanh and Montagne 1998; Greenacre et al. 2003; Koutsoumanis et al. 2003), Salmonella spp. (Leyer and Johnson 1992; Greenacre et al. 2003; Liao and Fett 2005; Tosun and Gonul 2008), Campylobacter jejuni (Murphy et al. 2003) and Vibrio vulnificus (Bang and Drake 2005) have been reported to have the ability to adapt to acid stress conditions. Micro-organisms adapted to mild acid stress may also survive against the same type of stress as a severe acid and to different types of lethal stresses, known as a specific adaptive response and a multiple adaptive response (also termed cross-protection), respectively (Foster and Hall 1990; Rodriguez-Romo and Yousef 2005).

Salmonella is one of the most significant pathogens causing systemic diseases such as gastroenteritis, septicemia, osteomyelitis, pneumonia, meningitis and arthritis (Trevejo et al. 2003; Ngwai et al. 2007). The genus Salmonella includes over 2400 serovars based on their somatic (O), flagellar (H) and capsular (Vi) antigens, of which Salmonella Typhimurium and Salmonella Enteritidis are the most common foodborne pathogens (Adams and Moss 2000). However, presently little is known about the protective effect of acetic acid adaptation on resistance of Salmonella enterica in lethal acid condition combined with low temperature. Most of the available information with regard to the stress adaptation includes a specific adaptive response. However, from a realistic viewpoint of food processing, foodborne pathogens are likely to be exposed to multiple stresses concurrently or sequentially. Therefore, understanding the behaviour of survival, adaptation and resurrection of foodborne pathogens against various stress conditions is essential for developing strategies for controlling cross-protection in foods.

The objectives of this study were to evaluate the cross-protection of acid-adapted S. enterica cells against severe acid and cold stress conditions and to investigate the recovery of acid-adapted and injured S. enterica cells in enriched broth.

Materials and methods

Bacterial strains and culture conditions

Strain of S. enterica subsp. enterica serovar Enteritidis (KACC 10763) was provided by the Korean Agricultural Culture Collection (KACC; Suwon, Korea). The strain was cultivated aerobically in trypticase soy broth supplemented with 0·1% yeast extract (TSBYE; BD, Becton, Dickinson and Co., Sparks, MD, USA) at 37°C for 20 h. After cultivation, the cultures were harvested at 3000 g for 20 min at 4°C. The pellets were resuspended to the original volume with 0·1% sterile buffered peptone water (BPW) for acid adaptation.

Acid adaptation

Nonadapted S. enterica cells were grown in TSBYE (pH 7·3) for 18 h at 37°C. To prepare acid-adapted cells, c. 3 × 107 CFU ml−1 of S. enterica grown in TSBYE (pH 7·3) were suspended for 2 and 7 h in TSB adjusted to pH 5·0 by the addition of acetic acid at 37°C. No significant changes in cell populations were observed during acid adaptation at pH 5·0.

Lethal acid challenge

After the acid adaptation, the non-adapted and acid-adapted S. enterica cells were collected at 3000 g for 20 min at 20°C and exposed to TSBYE (pH 4·0) adjusted with 1 mol l−1 hydrochloric acid (HCl) for 0, 1, 2, 3, 4, 6, 8, 10 and 12 d at 4°C and 20°C.

Viable counts

The viable counts were determined by the pour plating method. Each cultured sample was serially (1 : 10) diluted with 0·1% BPW and pour-plated on nonselective trypticase soy agar (TSA). The plates were incubated at 37°C for 48 h. Recovery of sublethally injured S. enterica cells was estimated by plating on selective xylose lysine desoxycholate (XLD) agar.

Microbial heterogeneity

The population heterogeneity of the injured S. enterica cells was determined by using ThermoMax Kinetic Microplate Reader (Molecular Devices, Menlo Park, CA, USA). Salmonella enterica cells, cultured in pH 4·0 at 4 or 20°C following unadaptation at pH 7·3 or acid adaptation at pH 5·0, were diluted to the desired inoculum (10 CFU ml−1) in double-strength TSBY. A 200-μl aliquot of the diluted cultures was transferred to microtitre plates and incubated at 37°C. The growth patterns of the injured S. enterica cells were observed at 630 nm at every 6-h interval up to 60 h. Maximum growth rates and lag times were estimated by the modified Gompertz model.

Inactivation and growth kinetics

The Weibull model (Chen 2007) was used to describe acid resistance of S. enterica cells at pH 4·0.

image(1)

where N0 and N represent the count of initial inoculum and the count of survivors after being exposed to pH 4·0 for a specific time (t). The b and n values are the scale and shape factors, indicating a measure of resistance and a degree of curvilinear, respectively.

The kinetic parameters for the bacterial growth were analysed using the modified Gompertz model (Linton et al. 1995):

image(2)

where G, A, B, C and M are, respectively the viable counts (optical density unit) at time t, inoculum level (unit), relative growth rate at the inflection point (h−1), number of log cycles of cell growth (unit) and the time required to reach the maximum growth rate (h). The lag phase duration (LPD = MB−1), maximum growth rate (GR = B·Ce−1) and maximum population density (MPD = A+C) were calculated.

Statistical analysis

The microbial inactivation and growth curves were analysed using Nonlinear Curve Fitting Function of Microcal Origin® 7·5 (Microcal Software Inc., Northampton, MA, USA). Data were analysed by the generalized linear model (GLM) procedure of the Statistical Analysis System (SAS). Least significant difference (LSD) procedure was used to determine significant differences among acid adaptation treatments at the 5% significance level.

Results

Effect of mild acid adaptation on resistance to lethal acidic condition

The population of S. enterica remained constant during the 7 h of acid adaptation period (data not shown). The survival of nonadapted and acid-adapted S. enterica after exposure to lethal pH at 4 and 20°C is shown in Fig. 1. In general, the enumerated populations of nonadapted and acid-adapted S. enterica cells decreased as the exposure time at pH 4·0 increased, regardless of adaptation time. Nonadapted S. enterica cells grown in TSB (pH 7·3) lost viability significantly when exposed to pH 4·0 at both 4 and 20°C (Fig. 1). Compared with the nonadapted S. enterica, the acid-adapted cells showed increased tolerance at both 4 and 20°C. The numbers of the nonadapted, 2 h acid-adapted, 7 h acid-adapted S. enterica cells were reduced by 4·42, 1·17 and 2·29 log CFU ml−1, respectively, after 12 days of incubation at 4°C. The nonadapted and acid-adapted S. enterica cells at 20°C of incubation were more sensitive to acid shock (pH 4·0) than those at 4°C of incubation. While nonadapted and acid-adapted cells still survived in significant numbers (>3·5 log CFU ml−1) at 12-day postacid shock at 4°C of incubation, no viable populations of nonadapted, 2 h acid-adapted, 7 h acid-adapted S. enterica cells were detected after 8, 12 and 10 days, respectively. As shown in Fig. 1, the proportion of injured S. enterica cells increased with increasing incubation time at both 4 and 20°C. Although significant differences were observed in counts between nonselective and selective plates, the numbers of acid-resistant S. enterica cells were decreased in the same manner as those of the total viable S. enterica cells. The kinetic parametric values were estimated from the Weibull model (Eqn 1), as shown in Table 1. Regression coefficients (R2) were larger than 0·91 for the survival curves of nonadapted and acid-adapted S. enterica cells. The survival curves for acid-adapted cells showed linearity (= 1) at both 4 and 20°C, whereas those for nonadapted cells exhibited nonlinear pattern, showing a noticeable upward concavity (< 1). The survival curves for the nonadapted cells declined dramatically at 20°C of incubation as compared with 4°C, which is described by a fast decline followed by a slight tail. There were also significant differences in the scale factor (b) between nonadapted and acid-adapted cells. Preadaptation at pH 5·0 resulted in an increase in resistance, showing that the b values were smaller for acid-adapted S. enterica cells.

Figure 1.

 Survival of Salmonella enterica at pH 4·0 after nonadaptation at pH 7·3 (•, ○), prior to acid adaptation at pH 5·0 for 2 h (bsl00001, □) and prior to acid adaptation at pH 5·0 for 7 h (bsl00066, Δ) during incubation at 4ºC (a) and 20ºC (b). Closed and open symbols represent cells recovered on trypticase soy agar and xylose lysine desoxycholate agar, respectively.

Table 1.   Inactivation kinetic parameters obtained from the fitted Weibull model for the survival curves of nonadapted or acid-adapted Salmonella enterica cells after subsequent acid challenge at pH 4·0
TreatmentSubsequent acid challenge at
4°C20°C
bnbn
  1. *Means (±SD) with different subscript letters within a column are significantly different at P < 0·05.

Nonadaptation (pH 7·3)0·75 ± 0·08a*0·74 ± 0·06a2·96 ± 0·09a0·42 ± 0·01b
Preadaptation at pH 5·0 for 2 h0·11 ± 0·06b0·99 ± 0·26a0·93 ± 0·06c0·86 ± 0·02a
Preadaptation at pH 5·0 for 7 h0·18 ± 0·05b1·14 ± 0·14a1·15 ± 0·04b0·82 ± 0·01a

Effect of acid tolerance on resuscitation and re-growth

The recovery characteristics of survivors remaining after the acid shock (pH 4·0) at both 4 and 20°C were observed as shown in Fig. 2. Neither acid-adapted nor subsequent acid challenge was used as the control. The growths of the control increased exponentially up to the OD at 630 nm of 1·0 after 36 h of incubation at 37°C. However, nonadapted and acid-adapted S. enterica cells were recovered slowly throughout the incubation period, except for 2-h acid adaptation, as compared with the control. Poor recovery was observed at the 7-h acid-adapted S. enterica cells exposed to pH 4·0 at both 4 and 20°C (OD630 < 0·2). Unlike the survival curves shown in Fig. 1, the remaining survivors after the subsequent acid challenge showed the highest recovery for 2-h adaptation, followed by nonadaptation and 7-h adaptation. Growth parameters of the control, nonadapted and acid-adapted S. enterica cells were estimated by the modified Gompertz model (Eqn 2) as shown in Table 2. The modified Gompertz equation fitted well with the experimental growth data. Regression coefficients (R2) were larger than 0·96 for the repair curves of acid-injured S. enterica cells. According to the Gompertz parameters, the 2-h adaptation had the largest B, largest C and smallest M values, followed by nonadaptation and 7-h adaptation, indicating that the remaining survivors after subsequent challenges at 4°C were well recovered in TSBYE at 37°C. However, no significant differences in B and M values were observed for the remaining survivors after subsequent challenges at 20°C, as shown in Table 2. Maximum GR, LPD and MPD were estimated from the Gompertz parameters, as shown in Table 3. When compared with the control, nonadaptation and acid adaptation had a lower GR, ranging between 0·004 and 0·036 unit h−1.

Figure 2.

 Heterogeneity in distribution of the control (•; nonadaptation at pH 5·0 and no acid shock at pH 4·0), nonadapted Salmonella enterica (bsl00066), preadapted S. enterica at pH 5·0 for 2 h (bsl00001), and preadapted S. enterica at pH 5·0 for 7 h (◆) during the post-acid shock (pH 4·0) at 4ºC (a) and 20ºC (b). Symbols and solid lines denote the observed and predicted growths of S. enterica, respectively.

Table 2.   Gompertz parametric values for the recovery curves of survivors remaining after subsequent acid challenge at pH 4·0 (= 22)
TreatmentABCMR2
  1. *Control is neither acid adaptation at pH 5·0 nor subsequent acid challenge at pH 4·0.

 Subsequent acid challenge at 4°C
Control*0·0260·1540·94919·5820·998
Nonadaptation (pH 7·3)0·0260·0830·64426·4430·994
Preadaptation at pH 5·0 for 2 h0·0270·0941·02925·6340·993
Preadaptation at pH 5·0 for 7 h0·0270·0500·23534·1560·958
 Subsequent acid challenge at 20°C
Control*0·0270·1690·99918·2220·998
Nonadaptation (pH 7·3)0·0280·1090·33822·9760·989
Preadaptation at pH 5·0 for 2 h0·0240·1300·44220·5270·995
Preadaptation at pH 5·0 for 7 h0·0290·1450·14419·9700·980
Table 3.   Derived maximum growth rate (GR), lag phase duration (LPD) and maximum population density (MPD) of survivors remaining after subsequent acid challenge at pH 4·0
TreatmentSubsequent acid challenge at
4°C20°C
GRLPDMPDGRLPDMPD
  1. *Control is neither acid adaptation at pH 5·0 nor subsequent acid challenge at pH 4·0.

Control*0·05413·0760·9750·06212·2891·026
Nonadaptation (pH 7·3)0·02014·3460·6700·01413·7780·366
Preadaptation at pH 5·0 for 2 h0·03615·0441·0560·02112·8350·466
Preadaptation at pH 5·0 for 7 h0·00414·1540·2620·00713·0730·173

Nonadaptation and acid adaptation showed a longer LPD than the control. Unlike the control, nonadaptation and acid adaptation caused the growth retardation of S. enterica cells during the subsequent acid shock at both 4 and 20°C. Other than the control, the 2-h adapted cells showed the best recovery, which is supported by maximum GR and MPD values.

Discussion

Acid tolerance of nonadapted and acid-adapted cells was investigated in subsequent acid challenge at 4 and 20°C. The experimental conditions represent realistic scenarios that could occur through food processing, including pH gradients (pH 5·0 to pH 4·0), refrigerated storage (4°C) and temperature abuse (20°C).

No significant injured S. enterica cells were observed during the adaptation at pH 5·0 (data not shown), while the numbers of sublethally injured cells were increased rapidly during the subsequent challenge at pH 4·0, indicating that S. enterica cells were more sensitive to the lethal acid. This result was confirmed by comparison with the counts between nonselective TSA and selective XLD agar plates. The acid-adapted cells maintained a constant internal pH, resulting in enhanced resistance to lethal acidic conditions (Leyer and Johnson 1992; Greenacre et al. 2003). The injured cells were dominantly present at the 12-day postacid shock at pH 4·0. Bacterial pathogens can be injured under the food processing-related stresses, and sublethally injured cells may repair sublethal damage and regain viability and pathogenicity under favourable environmental conditions (Wong et al. 1998; Liao and Fett 2005). Therefore, the repair and recovery of injured cells in food systems have become an increasing concern on food safety. The study regarding the fate of injured cells would help not only understand the behaviour of injured cells during food processing but also design preservation processes.

Acid tolerance of S. enterica cells at pH 4·0 increased with increasing pre acid adaptation time, which is in agreement with previous reports in which the adaptation time and pH are the most important factors in inducing acid tolerance response (ATR; Foster and Hall 1991; Leyer and Johnson 1993; Phan-Thanh and Montagne 1998; Greenacre et al. 2003). Acid stress could trigger the acid-adaptive response, which activates the defense mechanism to repair the injured cells. The resistance to acid stress is likely to be increased with increasing acid adaptation time. However, a prolonged exposure to mild acid stress causes loss of homeostasis, resulting in more sensitivity to lethal stress (Greenacre et al. 2003). Acid-adapted S. enterica cells displayed enhanced resistance after exposure to pH 4·0 at 4°C (Fig. 1a). The result suggests that acid adaptation produced cross-protection to cold. Acid-induced ATR resulted in the specific adaptive response and the cross-protection to temperature down-shift (Wong et al. 1998; Rowan 2004; Tosun and Gonul 2008). The cross-protection has important implications for food safety when food preservation methods are developed. During food processing, the exposure to sequential sublethal stresses can induce cross-protective micro-organisms. Therefore, the stress response mechanisms of foodborne pathogens in food systems have recently received much research attention.

Microbial heterogeneity was examined to distinguish the physiological states of cells exposed to pH 4·0 following acid adaptation at pH 5·0. The heterogeneity in distribution of injured and uninjured cells resulted in different recovery behaviour. The status of the cells appears to be divided into two types: one is the re-growth of residual culturable cells consisting of the repair of injured cells and the growth of uninjured cells, and the other is the resuscitation of viable but nonculturable (VBNC) cells. Injured S. enterica cells were able to repair and re-grow in a nonselective medium (Fig. 2). In general, bacterial cells undergo distinct physical and metabolic adaptations in response to unfavourable environmental conditions (Bogosian and Bourneuf 2001; Heim et al. 2002; Nystrom 2003). When in the VBNC state, bacteria cannot grow on conventional agar plate but maintain metabolic activity and pathogenicity (Roszak and Colwell 1987; del Mar Lleo et al. 2000, 2007; Reissbrodt et al. 2002). Factors affecting the VBNC state include temperature, pressure, osmotic pressure, nutrient starvation and pH (Lazaro et al. 1999; Winfield and Groisman 2003). Foodborne pathogens can possibly activate the VBNC state under the food preservatives, which can result in an unpredicted and significant loss of revenues to the food industry and potentially cause serious health problems (Wong and Wang 2004). Therefore, the VBNC in the case of foodborne pathogens may draw significant attention to food safety.

The most significant finding in this study was that the acid-adapted S. enterica cells can exist in the VBNC state. The resuscitation is known as the transition of the VBNC state into the culturable state (Tims and Lim 2003). No significant change in the cell density was observed for the 7-h adaptation throughout the exposure period at pH 4·0 (Fig. 2). The result suggests that the resuscitation may occur during the subsequent acid challenge at both 4 and 20°C. This is in agreement with a previous report that the VBNC does not lead to the growth by increase in cell number (Tims and Lim 2003). The VBNC phenomenon has been reported in numerous pathogens including E. coli O157:H7 (Rigsbee et al. 1997; Kolling and Matthews 2001), L. monocytogenes (Besnard et al. 2002), S. enterica (Gupte et al. 2003), C.. jejuni (Ziprin et al. 2003), V. vulnificus (Linder and Oliver 1989; Brauns et al. 1991) and Aeromonas hydrophila (Maalej et al. 2004). The model-fitting approaches used to evaluate kinetic parameters in this study provide practical information on understanding the induction of cross-protected and VBNC pathogens under food processing-related stresses, which is essential for designing new food preservation methods and reduce their potential health risk. However, the VBNC state in the so-called ‘oxymoron’ or ‘misnomer’ regarding the viability, pathogenicity and identification still remains unclear (Bloomfield et al. 1998; Bogosian and Bourneuf 2001). Therefore, the repair and recovery of cross-protected and VBNC pathogens under different stress conditions are currently under investigation in our laboratory.

Acknowledgements

The authors acknowledge the financial support from the 2007 Research Program for the Development of Forest Science and Technology, Korea Forest Service Research Institute (Daejeon Metropolitan City, Korea) and also thank the Korean Agricultural Culture Collection (KACC) for providing the strain of S. enterica (KACC 10763).

Ancillary