Use of flow cytometry and total viable count to determine the effects of orange juice composition on the physiology of Escherichia coli

Abstract Orange juice (OJ) contains numerous compounds some of which are known to play key roles in growth and survival of bacteria. This study aimed to investigate the effects of natural or processing‐induced variations in OJ composition on the physiology of Escherichia coli. OJ and model OJ (MOJ) samples containing various sugars, organic acids, amino acids, or ascorbic acid were inoculated with E. coli K‐12 MG1655 in different growth phases. The culturability, viability, and physiology of the cells were investigated during storage using plate counting and flow cytometry. Generally, stationary‐phase cells displayed the greatest survival in both MOJ and OJ. Increase in incubation temperature from 4 to 22.5ºC caused a significant decrease in both healthy and culturable cell populations. Supplementation of MOJ with ascorbic acid and amino acids increased both the viability and culturability of the cells. Similar trends were observed in amino acid‐supplemented OJ, albeit at a slower rate. In contrast, variations in sugar or organic acid composition had negligible effects on the physiological status of the cells. In summary, natural variation in ascorbic acid or amino acid concentrations could potentially have an adverse effect on the microbiological safety of orange juice.

With regard to amino acids present in OJ, arginine and glutamate are the major components of the acid resistance mechanisms in E. coli (Richard & Foster, 2004). Moreover, it has been suggested that proline, the most abundant amino acid in OJ, could increase the survival of E. coli in an acidic model apple juice (Reinders, Biesterveld, & Bijker, 2001). In addition, ascorbic acid has been shown to act not only as an antioxidant but also an antimicrobial against E. coli particularly in combination with organic acids (Padayatty et al., 2003;Tajkarimi & Ibrahim, 2011).
Flow cytometry (FCM) is a rapid, single-cell analysis technique that allows analysis of bacterial viability and physiology without the requirement for growth on agar plates (Comas-Riu & Rius, 2009). Samples of bacteria suspended in a liquid (e.g., a liquid foodstuff, a fermentation broth or water) are passed in a stream in front of a Laser beam, where particles suspended in the liquid are illuminated one at a time, up to several thousand per second. Scattered light is detected using sensors in line with and perpendicular to the Laser beam, indicating particle size and "granularity" (a measure of particle optical complexity), respectively. In addition, particle fluorescence is measured; staining samples with fluorescent dyes allows determination of aspects of bacterial physiology and viability (Bridier, Hammes, Canette, Bouchez, & Briandet, 2015). The major advantages of FCM are speed of analysis (minutes); detection of subpopulations present in a sample (e.g., numbers of live, dead, and injured bacteria); and the lack of reliance on growth on agar plates, thus allowing analysis of not only culturable but also nonculturable, and thereby also VBNC (viable but nonculturable), bacteria (Nebe-Von-Caron, Stephens, Hewitt, Powell, & Badley, 2000).
The objective of this study was to test whether changes in concentrations of components of OJ brought about by natural variations (seasonal changes in composition or changes in cultivar) or processing-induced changes in composition could influence the physiology of E. coli. A model orange juice (MOJ) solution (Table 1) was used in order to investigate the role of each component independently of the inherent variability in the composition of real fruit juice (Reinders et al., 2001) and to test known sugar or acid compositions mimicking seasonal and/or cultivar variability of OJ (Kelebek & Selli, 2011;Villamiel & Martõ, 1998). Furthermore, using a model system eliminated the need for chemical analysis of OJ composition. We tested the physiological response of E. coli to this simple MOJ, to OJ as a comparison, and to MOJ with altered composition with regard to organic acids, sugars, and amino acids to model variations in OJ composition. We utilized flow cytometry as a rapid technique for assessing bacterial physiology and viability in a nongrowth-dependent manner (Anvarian, Smith, & Overton, 2016 Niu et al. (2008), Kelebek, Selli, Canbas, and Cabaroglu (2009) Kelebek and Selli (2011) and Villamiel and Martõ (1998). c From Robards and Antolovich (1995). d From Capilla, Navarro, Sendra, and Izquierdo (1988), Niu et al. (2008), Kelebek et al. (2009) and Kelebek and Selli (2011). e Measured in this study for freshly squeezed OJ, filtered through 1.2μm filter, mean ± standard deviation of 3 samples.
A single colony of E. coli was taken from a nutrient agar (Oxoid) plate inoculated into 20 ml of 2 × LB (Luria-Bertani broth; 20 g/L tryptone, 10 g/L yeast extract, 10 g/L NaCl) in a 250-mL conical flask and was grown for 18 hr at 37°C and 150 rpm shaking. The overnight culture was subsequently diluted (1:1,000) into 50 ml of fresh 2 × LB medium in a 250-ml conical flask and grown under the same conditions until the desired OD 650 was reached. Cultures at an OD 650 of 0.5 were used for experiments with supplementation of MOJ and OJ with organic acids, amino acids, and ascorbic acid. Cells were harvested by centrifugation (10 min at 3,256 g), washed, and dispersed in PBS (phosphate-buffered saline; 8.0 g/L NaCl, 0.2 g/L KCl, 1.15 g/L Na 2 HPO 4 , and 0.2 g/L KH 2 PO 4 , pH 7.3). Then, 50 μl of cell suspension containing 3 × 10 9 E. coli cells was added to 15 ml of either MOJ or OJ. Samples were taken for analysis immediately prior to addition to OJ or MOJ, immediately after addition, and after specified time intervals.

| Orange juice and model orange juice
Freshly squeezed OJ was obtained from a local retailer. The total viable count (TVC) of the freshly squeezed OJ was <10 CFU/g. Freshly squeezed OJ was centrifuged at 17,696 g for 40 min to remove pulp.
The supernatant (pulp-free OJ) was filtered through sterile 1.2μm filter paper to permit analysis by flow cytometry. A model orange juice (MOJ) was developed containing the major components of OJ (Table 1). The constituents of MOJ were dissolved in deionized water and filter sterilized using a 0.22μm filter (Millipore).

| Analysis of bacterial viability and physiology
Viability was determined using TVC; bacteria were diluted in maximum recovery diluent (8.5 g/L NaCl, 1 g/L peptone) and plated on nutrient agar medium. Plates were incubated at 37°C for 48 hr before counting. Flow cytometry was used as an alternative technique for measuring bacterial physiology. A BD Accuri C6 (BD, Oxford, UK) flow cytometer was used. Samples were diluted in PBS and stained with propidium iodide (PI; Sigma) and Bis-(1,3dibutylbarbituric acid) trimethine oxonol (DiBAC 4 (3) or BOX; Life Technologies) to determine viability and membrane potential, respectively (Anvarian et al., 2016

| Effect of growth phase and incubation temperature on E. coli physiology in orange juice and model orange juice
The physiology and culturability of E. coli in different growth phases, from early log phase to late stationary phase, was compared over time in orange juice at 4°C (indicative of refrigerated storage) or 22.5°C (to model ambient storage). In order to compare total viable count data with viability and physiology data generated by flow cytometry, orange juice was filtered, as the cloud of OJ interferes with bacterial detection using FCM. The effects of filtering OJ with filters of different pore sizes on E. coli are previously described (Anvarian et al., 2016). In addition, a model orange juice (MOJ) was developed to allow modelling of OJ composition. Previous studies have mainly used MOJ to study browning (e.g., (Shinoda, Komura, Homma, & Murata, 2005)) but not effects on bacterial viability or physiology.
The composition of MOJ mimics the composition of OJ found in the literature ( Table 1). The MOJ used in this study initially consisted of only the main sugars, organic acids, and minerals of OJ (Robards & Antolovich, 1995).
Three general trends were observed in the TVC data ( Figure 1).  (Table 1). This is indicative of other components of OJ having a protective effect on E. coli. Even when passed through a 1.2μm filter, OJ contains many compounds that might have a protective effect on E. coli (Anvarian et al., 2016). This phenomenon has been observed before in other studies comparing foodstuffs and model foodstuffs (Nualkaekul & Charalampopoulos, 2011;Uljas & Ingham, 1998).

First, incubation in
Finally, stationary phase E. coli tended to be more resistant to losses in culturability than exponential phase bacteria. This trend was most marked in OJ samples incubated for 48 hr at 22.5°C ( Figure 1d). This is a widely reported phenomenon (Arnold & Kaspar, 1995;Benjamin & Datta, 1995;Chung, Bang, & Drake, 2006), thought to be due to a combination of lower growth rate and induction of stress responsive and detoxification pathways in stationary phase.
As well as TVC analysis, samples incubated in 1. thereby alive and with a membrane potential; Figure 2), and numbers of healthy cells, permitting comparison with TVC data (Supporting information Figure S1).
Upon addition to MOJ or OJ, the percentage of healthy cells decreased in most samples (0 hr sample); this decrease was larger in MOJ than OJ (Figure 2 panels a &  Comparison of FCM healthy cell numbers (Supporting information Figure S1) with TVC data (Figure 1)

| Effects of altering the sugar and organic acid composition of model orange juice
As the concentrations of sugars present in OJ can fluctuate according to season, variety of orange, and other factors, the effect of changes in MOJ sugar concentrations (Supporting information  Figure S2a).  (Lanford, 1942)) as possible. The concentration of potassium in MOJ solutions (0.97-3.25 g/L) and the molar ratio of citric acid to malic acid (2.63-3.13) were close to the range reported for OJ (Robards & Antolovich, 1995;Saccani et al., 1995). With the exception of the composition of MOJ solutions, the method used for this experiment was identical to that described above for sugar experiments (Supporting information Figure S2b).
While there were no significant differences between viability or culturability of bacteria incubated in medium and low acid MOJ, or medium and high acid MOJ, there was a significant difference in the numbers of healthy bacteria as determined by FCM and culturable bacteria (as determined by TVC) between low and high acid MOJ.
As such, increasing organic acid concentration decreased viability and culturability. The biochemical and physiological reason for this is presently unknown, but as altering organic acid concentrations did not significantly change the pH of MOJ, the organic acid concentrations themselves are probably the cause of loss of viability and culturability.

| Effects of altering the ascorbic acid content of model orange juice
Ascorbic acid (vitamin C) is a natural component of OJ and is desired for its health benefits, including its antioxidant properties. In order to study the effects of ascorbic acid on the viability and culturability of exponential-phase E. coli, MOJ was supplemented with different concentrations of L-ascorbic acid: 0.5 g/L, chosen to mimic the typical concentration in OJ; and 1, 5, or 10 g/L, to model commercial OJ brands supplemented with high concentrations of ascorbic acid (>7 g/L). The pH and osmolality of MOJ with different concentrations of ascorbic acid are shown in Supporting information Table S2.
Analysis using FCM revealed that increasing the ascorbic acid concentration decreased the percentage of dead (PI + ) bacteria in MOJ at all timepoints (Figure 3a). Compared to un-supplemented MOJ, addition of 0.5 g/L ascorbic acid caused a significant reduction in the dead population. Numbers of dead bacteria were not significantly different in MOJ with 0, 0.5% or 1% ascorbic acid, but were significantly different in MOJ with 5% or 10% ascorbic acid. The decrease in the proportion of dead bacteria observed at increased concentrations of ascorbic acid did not translate to an increase in healthy (PI − BOX − ) bacteria; instead, the percentage of injured bacteria without a membrane potential (PI − BOX + ) increased with increasing ascorbic acid concentration (Figure 3b).
Culturability of bacteria as measured by TVC was significantly higher in MOJ supplemented with 5 or 10 g/L ascorbic acid ( Figure 3c). Culturability was not significantly different in MOJ with zero, 0.5 g/L or 1 g/L ascorbic acid, indicating that protective effects are only visible at ascorbic acid concentrations significantly greater than those naturally found in OJ under these conditions.
Ascorbic acid has been shown to have both positive and negative effects on microbial viability; effects are dependent upon the physicochemical properties of the food matrix, in particular the presence of organic acids, and the bacterial species. Ascorbic acid is known to increase survival of probiotic bacteria (Shah, Ding, Fallourd, & Leyer, 2010) and has been shown to induce catalase expression in E. coli, potentially increasing resistance to oxidative stress (Richter & Loewen, 1981). Indeed, it is known that supplementation of agar plates with catalase can restore culturability of VBNC E. coli (Mizunoe, Wai, Takade, & Yoshida, 1999).
The protective effect of ascorbic acid observed here may reflect induction of catalase and thus an increased resistance to oxidative stress; this should be taken into account when devising microbiological safety strategies for ascorbic acid supplemented OJ products.

| Effect of altering the amino acid concentration of MOJ
The effect of addition of amino acids on the culturability and physiology of exponential-phase E. coli in MOJ at 4°C was determined ( Figure 4). MOJ was supplemented with the eight most abundant amino acids found in OJ at concentrations similar to those found in OJ (Supporting information Table S3).

| Effect of amino acids supplementation of OJ
Considering the apparent beneficial effects of amino acids on the viability and culturability of E. coli in MOJ, the same mixture of amino acids was added to OJ and E. coli physiology was determined ( Figure 4c). The rate of decrease in the size of the healthy population in OJ as determined by FCM was significantly lower compared to MOJ samples. At all timepoints, the mean percentage of healthy cells was greater in amino acid-supplemented samples. Based on these results, it could be suggested that amino acids in OJ exert a beneficial effect on the viability of E. coli in OJ. This was also observed when comparing the TVC of E. coli incubated in OJ with and without amino acid supplementation ( Figure 4d).
Proline, the major amino acid present in orange juice, is a major osmoprotectant in E. coli. Previous studies have shown that proline is protective for E. coli O157:H7 in model apple juice (Reinders et al., 2001). In addition, both glutamate and aspartate have been shown to increase acid tolerance in E. coli (Rowbury, Humphrey, & Goodson, 1999). It is probable that a combination of these two effects is responsible for the protective effect of amino acid supplementation observed here.
These data are potentially important for the microbiological safety of food products containing amino acid-supplemented fruit juices.

This work was supported by a UK Biotechnology and Biological
Sciences Research Council (BBSRC) PhD studentship to AA. The BD Accuri C6 flow cytometer was awarded to TWO by the BD Accuri Creativity Award.

CO N FLI C T O F I NTE R E S T
TWO and AA were paid speakers expenses by BD for speaking at BD Accuri users' events.

E TH I C A L S TATEM ENTS
This study does not involve any human or animal testing.
Informed consent: Not applicable.