International Journal of Food Microbiology

International Journal of Food Microbiology 161 (2013) 164–171 Contents lists available at SciVerse ScienceDirect International Journal of Food Micro...
Author: Gary Johns
0 downloads 1 Views 785KB Size
International Journal of Food Microbiology 161 (2013) 164–171

Contents lists available at SciVerse ScienceDirect

International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro

Weak-acid preservatives: pH and proton movements in the yeast Saccharomyces cerevisiae Malcolm Stratford a, Gerhard Nebe-von-Caron b, Hazel Steels b, Michaela Novodvorska a, Joerg Ueckert c, David B. Archer a,⁎ a b c

School of Biology, University of Nottingham, University Park, Nottingham NG7 2RD, United Kingdom Mologic Ltd., Bedford Technology Park, Thurleigh, Bedford MK44 2YP, United Kingdom Unilever Research and Development Vlaardingen, Biosciences, The Netherlands

a r t i c l e

i n f o

Article history: Received 28 August 2012 Received in revised form 19 November 2012 Accepted 6 December 2012 Available online 28 December 2012 Keywords: Spoilage Sorbic acid Acetic acid H+-ATPase PMA1 Proton efflux

a b s t r a c t Weak-acid preservatives commonly used to prevent fungal spoilage of low pH foods include sorbic and acetic acids. The “classical weak-acid theory” proposes that weak acids inhibit spoilage organisms by diffusion of undissociated acids through the membrane, dissociation within the cell to protons and anions, and consequent acidification of the cytoplasm. Results from 25 strains of Saccharomyces cerevisiae confirmed inhibition by acetic acid at a molar concentration 42 times higher than sorbic acid, in contradiction of the weak-acid theory where all acids of equal pKa should inhibit at equimolar concentrations. Flow cytometry showed that the intracellular pH fell to pH 4.7 at the growth-inhibitory concentration of acetic acid, whereas at the inhibitory concentration of sorbic acid, the pH only fell to pH 6.3. The plasma membrane H+-ATPase proton pump (Pma1p) was strongly inhibited by sorbic acid at the growth-inhibitory concentration, but was stimulated by acetic acid. The H+-ATPase was also inhibited by lower sorbic acid concentrations, but later showed recovery and elevated activity if the sorbic acid was removed. Levels of PMA1 transcripts increased briefly following sorbic acid addition, but soon returned to normal levels. It was concluded that acetic acid inhibition of S. cerevisiae was due to intracellular acidification, in accord with the “classical weak-acid theory”. Sorbic acid, however, appeared to be a membrane-active antimicrobial compound, with the plasma membrane H+-ATPase proton pump being a primary target of inhibition. Understanding the mechanism of action of sorbic acid will hopefully lead to improved methods of food preservation. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Many yeast species grow in media of high osmotic strength and low pH, enabling growth and spoilage on a variety of food products. Such foods include confectionary, preserved/dried fruits, pickles, butter and cheese, salted, dried and smoked meats, and stored fruit and vegetables. Yeast growth on foods is regarded as a cause of spoilage rather than a safety issue. The symptoms of yeast spoilage are formation of clouds, particulates, or white colonies, alteration of flavour and odour (ethanolic taint being common), and causing “blown” containers due to excess gas formation (Stratford, 2006). The vast majority of yeast species rarely cause food spoilage. It has been estimated that only 10–12 yeast species are responsible for the great majority of spoilage of foods that have been processed and packaged according to good manufacturing practice (Pitt and Hocking, 1997). Several

⁎ Corresponding author. Tel.: +44 115 951 3313; fax: +44 115 951 3251. E-mail addresses: [email protected] (M. Stratford), [email protected] (G. Nebe-von-Caron), [email protected] (H. Steels), [email protected] (M. Novodvorska), [email protected] (J. Ueckert), [email protected] (D.B. Archer). 0168-1605/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijfoodmicro.2012.12.013

weak acids have been approved for use in foods within the EC, some of which are legally designated as preservatives (Anon, 1995). These include sorbic acid (2,4-hexadienoic acid), benzoic acid, propionic acid, and sulphites. These, together with acetic acid used as an acidulant in pickles, dressings and mayonnaise, are commonly termed the weak-acid preservatives. Research into preservative resistance has largely been carried out on the yeast Saccharomyces cerevisiae, recently reviewed by Piper (2011). The “classical weak-acid theory” of inhibition of microbes by preservatives is defined largely by the parameters of physical chemistry. Weak acids in aqueous solution partially dissociate leading to a dynamic equilibrium between molecular acids and charged anions/ protons. At low pH, the equilibrium increasingly favours molecular acids while at neutral pH, charged anions predominate. Molecular weak-acids of preservatives are lipid-soluble, unlike the charged anions, and at low pH, are able to penetrate cells by simple diffusion through the lipids of the plasma membrane into the cytoplasm. Diffusion is rapid, being fully complete within 1–3 min (acetic acid, Conway and Downey, 1950; benzoic acid, Macris, 1975; sulphite, Stratford and Rose, 1986). The apparent cessation of transport after 3 min is really a highly dynamic exchange of weak acids, effluxing

M. Stratford et al. / International Journal of Food Microbiology 161 (2013) 164–171

and influxing through the membrane (Stratford and Rose, 1986). Low concentrations of preservatives are concentrated many-fold within the cytoplasm due to ionization, leading to early claims that weak-acid permeation was active transport. In reality, weak-acid molecules entering the neutral pH of the cytoplasm, dissociate into charged anions, which are not able to diffuse out of the cell. The neutral cytoplasm acts, in effect, as a pH-mediated sink for preservatives. Dissociation of weak acids also releases protons in equimolar proportions, and high concentrations of preservatives release sufficient protons to cause substantial cytoplasmic acidification. Cellular metabolism is then inhibited through acidification (Krebs et al., 1983; Pearce et al., 2001). Removal of cells from preservatives causes an immediate diffusive efflux of weak acids, and consequent rise in cytoplasmic pH. The classic weak-acid theory was independently proposed, and cytoplasmic acidification was verified for acetic acid (Neal et al., 1965), benzoic acid (Krebs et al., 1983) and sulphite (Pilkington and Rose, 1988; Stratford, 1983). Resistance to weak-acid preservatives in yeast has been reported to involve ejection of protons via the plasma membrane H +-ATPase proton pump, encoded by PMA1 (Carmelo et al., 1997; Cole and Keenan, 1987; Holyoak et al., 1996). This removes protons from the cytoplasm with a normal stoichiometry of 1 proton ejected/ATP (Cid et al., 1987) although this may decline to 0.1 proton/ATP in starved cells (Venema and Palmgren, 1995). In addition, it has been demonstrated in S. cerevisiae that Pdr12p has a major effect on weak-acid resistance (Hatzixanthis et al., 2003; Piper et al., 1998). It has been proposed that this plasma-membrane pleiotropic drug resistance pump causes ejection of preservative anions from the cytoplasm into the external media. The assumption that all weak-acid preservatives act identically in causing inhibition has recently been questioned from several perspectives. Examination of genetic and transcriptional responses to a variety of weak-acid preservatives showed extensive specific responses (Abbott et al., 2007) but very limited consistent up-regulation and the authors concluded that weak acids did not have a common generic response in S. cerevisiae. Theoretical calculations of the proton release from S. cerevisiae by weak-acid preservatives showed that while acetic acid and sulphite released high cytoplasmic concentrations of protons; sorbic acid did not (Stratford and Anslow, 1997). These calculations were confirmed by direct measurement of intracellular pH drop in germinating spores of the mould Aspergillus niger (Stratford et al., 2009). It was concluded that in mould spores, acetic acid inhibition was due to intracellular acidification, but that inhibition by sorbic acid was not. In the present paper, these studies are extended to S. cerevisiae to determine the mechanism of sorbic acid inhibition of this spoilage yeast. 2. Materials and methods 2.1. Yeast strains The yeast strains used in this study are listed in Table 1 together with their source of isolation. The identity of all strains was confirmed by sequencing of the D1/D2 region of the 26S rDNA using the method described by Kurtzman (2003). The plasma-membrane H+-ATPase proton pump encoded by PMA1 is essential in S. cerevisiae and PMA1 deletion is lethal. However, the DAmP-PMA1 strain has reduced PMA1p activity caused by diminished mRNA abundance due to insertion of a kanamycin-resistance cassette into the 3′ untranslated region (UTR) immediately following the PMA1 ORF (Breslow et al., 2008). DAmP-PMA1 strain YSC5094-99851795, clone ID YGL008C (Decreased Abundance by mRNA Perturbation), was obtained from Open Biosystems (https:// www.openbiosystems.com/Query/?i=0&q=YGL008C). The presence of the disruption cassette was verified by PCR using specific primers (Breslow et al., 2008). Deletion of subunits of the vacuolar H+-ATPase proton pump causes misdirection of Pma1p away from the plasma membrane and results in reduced activity of Pma1p (Kane, 2006;

165

Table 1 Strains of Saccharomyces cerevisiae used in this study and their origins. NCYC strains are available from the National Collection of Yeast Cultures, Norwich, UK. Others were from a collection (Mologic strain number) assimilated over several years from the food industry. All strains were confirmed in identity by D1/D2 rDNA sequencing. Weak-acid preservative resistance, sorbic acid and acetic acid, was measured in YEPD pH 4.0 at 103 cells/ml and incubated at 25 °C for 2 weeks at pH 4.0. Data are the lowest concentration of weak acids (mM) to completely inhibit growth, and the concentration ratio (acetic:sorbic). Strain

Origin

Acetic acid

Sorbic acid

Ratio

22 47 48 49 51 56 61 62 63 65 125 174 176 244 253 282 291 292 308 359 632 633 634 635 636

Spoilage, carbonated soft drink, UK Spoilage, apple concentrate, UK Spoilage, fruit drink, UK NCYC 366, ale strain NCYC X2180-1B, lab strain Spoilage, mayonnaise, UK NCYC 87, distillery strain Wine strain Wine strain Baker's yeast Spoilage, soft drink, France Raspberry juice Euroscarf BY4741 Soy sauce, Netherlands Soy sauce, Netherlands Spoilage, soft drink, Netherlands NCYC 3253, spoiled soft drink, UK Spoilage, sugar syrup, UK Spoilage, soft drink, Belgium Factory isolate, Turkey Spoilage, natural yoghurt, UK Spoilage, Greek yoghurt, UK Spoilage, Greek yoghurt, UK Spoilage, fruit yoghurt, UK Spoilage, fruit yoghurt, UK

145 140 140 165 140 110 115 155 110 140 180 145 120 125 170 165 145 155 170 150 115 110 115 110 115

4 3.9 3.7 3.6 3.5 3.7 3.1 4.3 2.3 3.9 3.9 3.6 3 2.8 2.6 2.8 3.5 3.4 3.3 3.3 3 3.1 3 3 2.9

36.25 35.90 37.84 45.83 40.00 29.73 37.10 36.05 47.83 35.90 46.15 40.28 40.00 44.64 65.38 58.93 41.43 45.59 51.52 45.45 38.33 35.48 38.33 36.67 39.66

Tarsio et al., 2011). Δvma2 strains lack V-ATPase activity and Pma1p activity is 65–75% lower (Martinez-Munoz and Kane, 2006). Mutant Δvma2 strain Y03266 (Euroscarf gene reference YBR127c) was obtained from Euroscarf (http://web.uni-frankfurt.de/fb15/mikro/euroscarf/). Yeast strains were stored in glycerol on ceramic beads at −80 °C (Microbank™), and maintained short term on MEA (malt extract agar, Oxoid) slopes at 4 °C.

2.2. Growth medium and conditions The growth medium used in this study was YEPD, containing glucose 20 g/l, bacteriological peptone (Oxoid) 20 g/l, and yeast extract (Oxoid) 10 g/l, adjusted to pH 4.0 with 10 M HCl prior to heat sterilization. Starter cultures comprised 10 ml YEPD pH 4.0 in 28 ml McCartney bottles, inoculated with yeast from MEA slopes, and incubated for 48 h at 25 °C. Experimental cultures comprised either 10 ml YEPD pH 4.0 in 28 ml McCartney bottles, or 40 ml YEPD pH 4.0 in 100 ml conical flasks shaken at 130 rpm at 25 °C.

2.3. Measurement of weak-acid resistance (MIC tests) Resistance of yeast strains to weak-acid preservatives was investigated by determination of the minimum inhibitory concentration (MIC) of each acid to completely prevent growth. Series of McCartney bottles were prepared with 10 ml aliquots of YEPD, each containing a progressively higher concentration of preservative. The pH of all media was back-titrated to pH 4.0 following acid addition. Bottles were inoculated with yeast at a final concentration of 10 3 cells/ml and incubated for 14 days at 25 °C. The MIC was the lowest concentration of preservative at pH 4.0, at which no growth was detectable at 14 days.

166

M. Stratford et al. / International Journal of Food Microbiology 161 (2013) 164–171

2.4. Measurement of weak-acid resistance (population diversity) Resistance of individual cells in populations of yeast cells was determined by colony growth in liquid culture in 96 well microtitre plates (Steels et al., 2000). 20 ml aliquots of YEPD containing progressively higher concentration of preservatives were inoculated at 15–30 cells/ ml and dispensed into microtitre plates at 200 μl/well. Plates were sealed, lidded, double-bagged to prevent evaporation, and incubated at 25 °C for 14 days. Yeast cells fall to the base of wells and surviving cells grow into visible colonies. Yeast colonies were counted every 2 days, as preservatives slow yeast growth. 2.5. Measurement of cellular internal pH by flow cytometry The method used for determination of cellular internal pH by flow cytometry was fully described in Stratford et al. (2009). Briefly, exponentially-growing yeast cells were obtained from 16 h shake flasks (OD 1.0). Cells were harvested by centrifugation, washed and resuspended three times in buffer (KH2PO4 50 mM, succinate 50 mM, pH 6.0). CFDASE (carboxyfluoresceindiacetate succimidyl ester) was added at 10 μg/ml final concentration and cells were incubated at 38 °C for 12 min for stain uptake. Uncharged CFDASE, colourless and non-fluorescent, passes into the cell where it is cleaved intracellularly by esterases. The fluorescent succimidyl ester binds to proteins, ensuring retention of the dye within the cell. The internal pH of populations of individual fluorescent cells was determined from the linear ratio of the 575 nm (largely pH-independent) and 525 nm (pH-dependent) emission signals. Calibration was carried out using cells of defined pH, permeated using 2 mM 2,4-dinitrophenol in 500 mM acetate, and 100 μM nigericin in 500 mM acetate.

strain, YSC5094-99851795, clone ID YGL008C. Duplicate exponentiallygrowing yeast cultures (300 ml YEPD in 1 l conical flasks, OD 2.2) were treated with 1 mM sorbic acid and 40 ml samples removed at selected time points. Each sample was analysed in duplicate. Cells were rapidly harvested by filtration (3 μm, 50 mm filters), washed 2 times with water, and snap-frozen in liquid nitrogen. Frozen cells were mixed with glass beads and broken in a Sartorius dismembranator (2 min, 2000 rpm). Total RNA was then purified using the Plant/Fungi total RNA Purification Kit (Norgen Biotek, Canada) including the on-column DNAase treatment step. The concentration and quality of RNA for each sample was determined by UV spectrometry (Nanodrop ND-1000 spectrophotometer) and integrity checked using 1.25% agarose gels. qRT-PCR amplifications were carried out using the Applied Biosystems 7500 Fast Real-Time PCR system. Total RNAs SuperScript™ III reverse transcriptase (Invitrogen) was used to prepare cDNA. The PCR reaction mixture (10 μl) contained 100 ng cDNA, specific primer sets (200 nM final concentration), and FAST SYBR-Green Master Mix (Applied Biosystems). PCRs were carried out for 40 cycles; denaturation at 95 °C for 15 s, annealing at 60 °C for 30 s, and extension at 60 °C for 60 s using primers for the PMA1 gene: PMA1-qRT UP 5′-ATTCCAT CCATTTGACCCTG-3′ and PMA1-qRT DOWN 5′-CATAACACCCAAGATTT CCCA-3′. Experiments were independently conducted in duplicate for each RNA sample. The specificity of primer sets used for qRT-PCR amplification was evaluated by melting curve analysis. The Standard Curve Method was used for data evaluation (Liu et al., 2009).

3. Results 3.1. Strain variation in S. cerevisiae

2.6. Measurement of proton efflux Exponentially-growing yeast cells were obtained from 40 ml shaken cultures at OD 2.2 (measured OD 0.2 following an 11-fold dilution in water). Cells were rapidly harvested by filtration (3 μm, 50 mm filters), washed 4 times with water, twice with 100 mM glucose, resuspended in 10 ml 100 mM glucose, and equilibrated with rapid stirring within a water jacket at 25 °C, for 3 min. The time taken from shaken flask to experimental initiation was b 5 min. The resuspended yeast concentration was determined by optical density (OD 6.6–7.2). Proton efflux was measured using a pH probe attached to an Electrolab FerMac 360 fermenter system set to record every 6 s, over an experimental run of 20 min. At the end of each run, exact proton efflux was determined by rapid back-titration with 10 mM NaOH to the original pH value. Where indicated, sorbic acid was added from a 500 mM stock solution in methanol (controls showed no effect by methanol), or from a 500 mM potassium sorbate solution in water, or from a pH 4.0 10 mM sorbic acid solution (adjusted with NaOH) in water. Acetic acid was added from a pH 4.0 100 mM stock in water (adjusted with NaOH). Where proton efflux was measured after prolonged exposure to sorbic acid, sorbic acid was added to exponentially growing cultures in YEPD pH 4.0. To prevent sorbic acid efflux, after 3 h treatment, cells were washed 6 times in 100 mM glucose pH 4.0 containing sorbic acid at the same concentration applied to the flask, and proton efflux was measured in the presence of the same sorbic acid concentration. In other experiments, activation of proton efflux was determined by addition of sorbic acid to growth flasks, followed by 6 washes with 100 mM glucose without sorbic acid, and proton efflux measured in the same medium. 2.7. Measurement of PMA1 transcript levels by qRT-PCR Measurement of PMA1 transcript levels in cells treated with sorbic acid was carried out in wild-type strain BY4741, and the DAmP-PMA1

It is a frequent comment that studies using laboratory strains bear little relationship to industrial strains and conditions. In this paper, strains from diverse origins were chosen to ensure that results are industrially relevant, rather than applicable only to potentially atypical laboratory strains. The identity of all strains was confirmed as S. cerevisiae by D1/D2 rDNA sequencing. Tests were carried out on the sorbic and acetic acid resistance of all strains in YEPD at pH 4.0 (Table 1). Results showed considerable variation in resistance to sorbic acid, from 2.3 mM to 4.3 mM, and in resistance to acetic acid, from 110 mM to 180 mM. In all strains examined, sorbic acid inhibited growth at a much lower concentration than acetic acid. The mean sorbic acid resistance was 3.33 mM at pH 4.0 and mean acetic acid resistance was 138 mM. The mean MIC molar ratio between these two weak-acid preservatives was 42.01 ± 7.98. Since sorbic acid and acetic acid have near-identical pKa values, 4.76, they should affect cytoplasmic acidification equally, and inhibit growth at equimolar concentration. Results clearly confirm earlier tests (Stratford and Anslow, 1997) that this is incorrect. The origin of yeast strains appeared unrelated to robustness, with wine strains amongst the least- and most-resistant strains. The average of spoilage strain characteristics, with regard to weak acids, was very similar to that of all other strains of S. cerevisiae. Laboratory strain BY4741 was therefore chosen for further research into pH effects by sorbic acid and acetic acid.

3.2. Cell/cell variation and population diversity in S. cerevisiae It has been shown that resistance to sorbic acid in the spoilage yeast Zygosaccharomyces bailii was largely due to small numbers of highly resistant cells within the population (Steels et al., 2000). In S. cerevisiae strain BY4741, all cells were able to grow in sorbic acid over the range 0–2.4 mM with a declining proportion able to grow at concentrations up to 2.8 mM.

M. Stratford et al. / International Journal of Food Microbiology 161 (2013) 164–171

Population resistance by DAmP-PMA1 and Δvma2 strains showed that both strains were sensitive to sorbic acid, the bulk populations being inhibited by just over 1 mM sorbic acid at pH 4.0. 3.3. Intracellular pH effects of weak-acid preservatives Intracellular pH, pHi, measurements were made with CFDASEstained yeast cells, assayed by flow cytometry over a time-course of 30 min. Results showed a mean population pHi close to pH 6.7, with a single peak in frequency distribution of the fluorescence ratio of individual cells in the population, that was stable over 30 min, in cells at pH 4.0 in the presence of glucose. Addition of sorbic acid at 3 mM (MIC level) or 2.5 mM (bulk population inhibition) caused an immediate fall in cytoplasmic pH (Fig. 1a). Within 30 s the pH had fallen to b pH 6.3, then drifted to a slightly lower value over the next 10 min. In contrast, acetic acid (Fig. 1b) at 120 mM (MIC level) also caused a rapid fall in intracellular pH, but to ~ pH 5.0 over 1.5 min. At 2.5 min, the pH had stabilized at close to pH 4.8. It is evident from these results that at the MIC level, acetic acid caused a very substantial fall in intracellular pH while sorbic acid did not. Examination of pHi results at a range of sub-inhibitory weak-acid concentrations confirmed these results. When sorbic acid was applied over the concentration range 0.5–3 mM (Fig. 1a) the intracellular pH fell immediately to ~ pH 6.3–6.4, with relatively small differences discernable between different sorbic acid concentrations. Acetic acid

167

concentrations, 30–150 mM showed a much wider spread of responses, behaving much more akin to a chemical titration of the cytoplasm (Fig. 1b). All showed an initial fall but this ranged at 30 s from ~ pH 6.5 at 30 mM, to ~ pH 5.25 at 150 mM. The initial pH fall was slower than with sorbic acid, but was complete at all concentrations within 3 min. The removal of protons from the cytoplasm was suggested by the pHi recovery seen in 0.5 mM sorbic acid and 30 mM acetic acid. Tests were therefore carried out to determine the extent of H +-ATPase proton pumping in the presence of weak acids. 3.4. Proton efflux controls Activity by the plasma-membrane H+-ATPase proton pump can be measured either by ATPase activity or by proton efflux. Since ATPase activity can be non-stoichiometric with proton efflux (Venema and Palmgren, 1995) H+-ATPase activity was determined directly by proton efflux. In the absence of glucose, no proton efflux was discernable and protons flowed into yeast cells at ~40 nmol/mg dry weight/h (Fig. 2a). In the presence of 10 mM glucose, proton efflux was strongly evident, the rate increasing to ~120 nmol/mg dry weight/h in 100 mM glucose. No further increase in proton efflux was evident at 200 mM glucose, so all further experiments were carried out in 100 mM glucose. Proton efflux by S. cerevisiae cells can have 4 possible causes, CO2 build-up, H+-ATPase activity, K +/H+ exchange and organic acid secretion, but is predominantly due to H+-ATPase activity (Lapathitis and Kotyk, 1998). In the current experiments, rapid stirring eliminated CO2 build-up and absence of potassium in experimental solutions minimised proton efflux by K+/H+ exchange, as demonstrated by lack of efflux in the absence of glucose (Fig. 2a). Acid secretion was measured in control experiments, carried out where yeast suspensions were sampled at the beginning of experiments and after 30 min. Yeast cells were removed and the suspending fluid was back-titrated over the range pH 3.0–pH 7.0. Near-identical titration curves showed no unexpected buffering peaks and eliminated the possibility of significant proton efflux due to CO2 build-up or organic acid secretion over this time scale, and confirmed that proton efflux was very largely due to H+-ATPase proton pump activity. The PMA1 and PMA2 genes encode H+-ATPase proton pumps in S. cerevisiae, but under normal growth conditions, PMA2 expression is negligible (Fernandes and Sa-Correia, 1998). In the current paper, use of the DAmP-PMA1 mutant strain confirmed that a 4-fold reduction in transcript levels caused a similar decrease in proton efflux (see below). 3.5. Sorbic acid effects on proton efflux

Fig. 1. Time course of the mean internal pH of CFDASE-stained S. cerevisiae BY4741 populations determined from the fluorescence ratio at pH 4.0. a. Sorbic acid treatment, open circles — control (no sorbic acid), closed circles — 0.5 mM sorbic acid, open squares — 1 mM, closed squares — 1.5 mM, open triangles 2 mM, closed triangles — 2.5 mM, open diamonds — 3 mM. b. Acetic acid treatment, open circles — control, closed circles — 3 mM, open squares — 30 mM, closed squares — 60 mM, open triangles 90 mM, closed triangles — 120 mM, open diamonds — 150 mM. Prediction of pHi from the CFDASE ratio becomes increasingly less accurate below pH 5.0.

Addition of 1 mM sorbic acid to yeast cells in water caused an initial drop in suspending media pH, pHex, followed by a complete cessation of proton efflux (Fig. 2b). The initial rapid drop in pHex was shown to be an artefact caused by addition of acid, rather than being due to proton efflux from yeast cells. Experiments where 1 mM sorbic acid was added from a stock solution in methanol caused a pH fall but, when 1 mM potassium sorbate was added from a stock solution in water, the pH rapidly increased (Fig. 2c). When a stock solution of sorbic acid/sorbate pre-adjusted to pH 4.0 was used, no initial pH change occurred. Similar pH changes occurred in control experiments, in the absence of yeast cells. Using sorbic acid and the acid/anion pH 4.0 mixture, proton efflux from yeast cells was inhibited, or even went into reverse for about 10 min, before a slight recovery in efflux. When potassium sorbate was used, proton efflux continued at a lower rate (Fig. 2c), perhaps due to K +/H + exchange, or to the raised pH reducing the effectiveness of the weak-acid preservative. The inhibition of H +-ATPase activity proton efflux by sorbic acid was confirmed by a series of experiments in which the concentration of sorbic acid was varied (Fig. 3a). Sorbic acid is a buffer, being a weak acid used within one pH unit of the pKa, and as a buffer, the presence

168

M. Stratford et al. / International Journal of Food Microbiology 161 (2013) 164–171

Fig. 2. a. Time course of pH change in washed suspensions of S. cerevisiae BY4741 in water containing varying concentrations of glucose, open diamonds — 0 mM, closed squares — 10 mM, closed triangles — 100 mM, closed circles 200 mM. pH was determined every 6 s. b. pH fall in two parallel washed suspensions of S. cerevisiae BY4741 in water containing 100 mM glucose (closed diamonds) arrested by addition of 1 mM sorbic acid (open squares). c. Initial pH jump in suspensions of S. cerevisiae BY4741 caused by addition of 1 mM sorbic acid (open circles), 1 mM potassium sorbate (open squares), or 1 mM sorbic acid/sorbate pre-adjusted to pH 4.0 (open triangles).

of sorbic acid will reduce the size of any pH change. However, back-titration with standard alkali to the initial pH will reveal and equate to the exact number of protons, irrespective of any buffering. Sorbic acid at concentration 0–0.2 mM caused no change to the rate of proton efflux, at ~ 120–130 nmol/mg dry wt/h over a 20 min experimental run, but higher concentrations of sorbic acid progressively lowered this to zero when added at a concentration of 1.3 mM (Fig. 3a). Higher concentrations of sorbic acid resulted in a slight influx of protons.

Fig. 3. a. The effect of sorbic acid concentration on the rate of proton efflux in washed suspensions of S. cerevisiae BY4741 in water containing 100 mM glucose measured by back-titration with NaOH after 20 min. Each experimental cell concentration was determined by OD 600 nm and the results are expressed in nmol protons effluxed/mg dry weight of yeast/h. Results are the means of three independent experiments. b. The effect of acetic acid concentration on the rate of proton efflux in washed suspensions of S. cerevisiae BY4741 in water containing 100 mM glucose measured by back-titration with NaOH after 20 min. Results are the means of three independent experiments. c. The effect of 1.5 mM aliphatic weak acids of increasing chain-length on proton efflux in washed suspensions of S. cerevisiae BY4741 in water containing 100 mM glucose measured by back-titration with NaOH after 20 min. Results are the means of three independent experiments.

3.6. Acetic acid and the effects of other aliphatic acids on proton efflux In contrast to sorbic acid over the concentration range of 0–2 mM, acetic acid caused an immediate rise in the rate of proton efflux. This increased with the concentration of acetic acid (Fig. 3b) and

M. Stratford et al. / International Journal of Food Microbiology 161 (2013) 164–171

169

approximately doubled the rate of proton efflux over the 20 min run. Further experiments were carried out with other aliphatic weak-acids spanning 2-carbon acetic acid and 6-carbon hexanoic acid at 1.5 mM (Fig. 3c). The rise in proton efflux caused by acetic acid was mirrored by propionic acid. Butyric acid caused a slight rise in activity, while valeric and hexanoic acids inhibited proton efflux. The similarity between the saturated, hexanoic acid inhibition of proton efflux, and the effect of unsaturated 2,4-hexadienoic acid (sorbic) demonstrated that inhibition was not dependent on any specific structure in sorbic acid, but was shared by all of the more hydrophobic weak acids (higher partition coefficients). Partition coefficients cLogPoct are acetic acid − 0.194, propionic acid 0.335, butyric acid 0.864, valeric acid 1.393, sorbic acid 1.514 and hexanoic acid 1.922. 3.7. PMA1 mutation effects on proton efflux The rate of proton efflux was examined in the DAmP-PMA1 strain, YSC5094-99851795 clone ID YGL008C reduced PMA1 expression and Δvma2 strain, in the presence and absence of sorbic acid. Proton efflux in the DAmP-PMA1 strain was considerably reduced; ~37 nmol/mg dry wt/h compared with ~120–130 nmol/mg dry wt/h in the WT strain, confirming that the bulk of proton efflux was due to PMA1-encoded H+-ATPase activity. Sorbic acid addition, even at 0.2 mM, inhibited proton efflux. The rate of proton efflux in the Δvma2 strain was similar to the wild-type strain in untreated controls (Fig. 4), but was more sensitive to sorbic acid inhibition. While in the wild-type strain, proton efflux ceased at ~1.3 mM sorbic acid, proton efflux ceased at ~0.7 mM sorbic acid in the Δvma2 strain, and ~0.2 mM in the DAmP-PMA1 strain.

Fig. 5. The long-term effect of sorbic acid on proton efflux in S. cerevisiae BY4741. Sorbic acid was added to the medium of exponentially-growing cells in cultures of yeast in YEPD pH 4.0. After 3 h, cells were washed 6 times and resuspended in identical concentrations of sorbic acid with 100 mM glucose. Proton efflux was measured by backtitration with NaOH after 20 min (light histograms). Each experimental cell concentration was determined by OD 600 nm and the results are expressed in nmol protons effluxed/ mg dry weight of yeast/h. Results are the means of three independent experiments. In parallel experiments, cells were treated with sorbic acid for 3 h, then washed six times and resuspended in 100 mM glucose without sorbic acid (dark histograms).

1.5 mM sorbic acid were washed clean, proton efflux was found to be activated, to a level higher than that of un-treated controls (Fig. 5). 3.9. qRT-PCR measurement of PMA1 transcription in sorbic acid-treated cells

3.8. Long-term proton efflux in sorbic-acid-treated cultures Long-term exposure of yeast to sorbic acid was carried out in YEPD at pH 4.0 using exponentially-growing cells. After 3 h, cells were washed 6 times in 100 mM glucose pH 4.0 containing sorbic acid at the concentration applied to the flask, and proton efflux was measured in the presence of the same sorbic acid concentration. It was found that in cells treated with sub-inhibitory 1.5 mM sorbic acid, proton efflux had largely recovered over 3 h, unlike with 3 mM sorbic acid-treated cells, where proton efflux was still inhibited (Fig. 5). This inhibition was largely due to the remaining presence of sorbic acid, as when the sorbic acid was removed by washing, proton efflux was immediately restored to near control levels. When cells treated with

Quantitative RT-PCR measurements of PMA1 transcript levels were carried out in wild-type strain BY4741, and DAmP-PMA1 strain, YSC5094-99851795, clone ID YGL008C, over 3 h following 1 mM sorbic acid addition (Fig. 6). In wild-type B4741, sorbic acid caused a short-term increase in transcript level, 10–30 min following sorbic acid addition, possibly as a result of feedback from the sorbic-acidinhibited H +-ATPase. Later, 1–3 h, the transcript level fell to a level slightly lower than before sorbic acid addition. In the DAmP-PMA1 strain, the transcript level of PMA1 was approximately 25% that of the wild-type BY4741. In the DAmP-PMA1 strain, sorbic acid did not increase the transcript level of PMA1 as seen in the wild-type strain, but remained fairly constant over 3 h following sorbic acid addition

Fig. 4. Proton efflux in washed suspensions of S. cerevisiae BY4741 (open circles), DAmP-PMA1 (closed circles) and Δvma2 (closed squares), in water containing 100 mM glucose measured by rapid back-titration with NaOH after 20 min. Each experimental cell concentration was determined by OD 600 nm and the results are expressed in nmol protons effluxed/mg dry weight of yeast/h. Results are the means of three independent experiments.

Fig. 6. Time course of PMA1 transcript levels in wild type S. cerevisiae BY4741 and the reduced PMA1 activity DAmP-PMA1 mutant, over 3 h following addition of 1 mM sorbic acid to 300 ml cultures in YEPD pH 4.0. 40 ml samples were removed in duplicate at each time point, and each analysed in duplicate. Wild type transcripts were significantly higher 10 min after sorbic acid addition (T-test b 0.05 compared with time 0).

170

M. Stratford et al. / International Journal of Food Microbiology 161 (2013) 164–171

(Fig. 6). Lack of stimulation by sorbic acid may indicate that PMA1 transcription in this mutant was already at full capacity. 4. Discussion In a previous publication, it was shown that the action of acetic acid on 5 h-germinated conidia of the spoilage mould A. niger was fully consistent with the classic weak-acid theory (Stratford et al., 2009). Acetic acid at the MIC, 80 mM at pH 4.0, caused the intracellular pH to fall dramatically, reaching ~pHi 4.7. In the current study, acetic acid at the S. cerevisiae MIC, 120 mM at pH 4.0, also caused a fall in pHi to ~4.7and is fully consistent with the original proposal for the action of acetic acid (Neal et al., 1965). Low cytoplasmic pH is known to distort metabolic enzyme structures and inhibit activity, particularly in low-pH sensitive enzymes such as hexokinase and phosphofructokinase (Krebs et al., 1983; Legisa and Golic Grdadolnik, 2002; Pearce et al., 2001). In the current study in S. cerevisiae, acetic acid at 3 mM caused no fall in pH, probably due to enhanced proton removal. Sorbic acid at the MIC 3 mM only caused a slight fall in intracellular pH to ~pH 6.3. Such an intracellular pH is not inhibitory. We can therefore conclude that sorbic acid does not inhibit the yeast S. cerevisiae or the mould A. niger via cytoplasmic acidification as a classic weak-acid preservative. Examination of alternative sites of action by sorbic acid suggests lipids and membranes as the most probable site of action. This is based on the evidence that inhibition is non-specific to the chemical structure of sorbic acid but also to any similar compound with a similar lipophilic partition coefficient (Kabara et al., 1972; Kobayashi et al., 1998; Reinhard and Radler, 1981; Sikkema et al., 1995). Sorbic acid is predicted from the partition coefficient to be concentrated within membranes at 60–100 times greater than the aqueous concentration (Stratford et al., 2009). The classic symptoms of membrane damage caused by an antimicrobial agent range from cytoplasmic leakage to cell lysis (Hammond and Lambert, 1978), the speed at which cytoplasmic components leak from cells being inversely related to their molecular size. The initial leakage of small ions (e.g. K +) is followed by the appearance of 260 nm-absorbing material in the media, including leakage of nucleic acids, proteins and certain amino acids (Hammond and Lambert, 1978). In our studies, sorbic acid at 3 mM caused leakage of potassium ions from yeast cells over 30 min, but no 260 nm leakage was detectable (M. Stratford, unpublished results). This suggests that the direct damage to the membrane structure caused by sorbic acid at the MIC is only minor and that the major effect of sorbic acid on membranes may be to alter the activity of membrane proteins. The lipid environment surrounding membrane proteins has been known for some time to affect protein activity. Transport permeases in yeast were reportedly affected by different membrane lipids (Keenan and Rose, 1979; Prasad and Rose, 1986) and ergosterol content has been shown to affect H+-ATPase activity in yeast (Arami et al., 1997). Previous studies have shown that sorbic acid affects the tryptophan permease in S. cerevisiae (Bauer et al., 2003) and the uracil/uridine permease in A. niger (Melin et al., 2008), neither of which are essential processes for cell viability or vitality. Here we report the inhibition of proton efflux via the plasma membrane H+-ATPase proton pump in S. cerevisiae, encoded by PMA1 (Fernandes and Sa-Correia, 1998) an activity vital for cell survival. This action takes immediate effect, and is initiated at sorbic acid concentrations lower than the growth-inhibitory concentration for the bulk of the cell population. This result is in direct contradiction to the previously reported results, reviewed by Piper (2011), which claim the H+-ATPase to be a resistance mechanism to sorbic acid (Holyoak et al., 1996). Examination of the H +-ATPase proton pump literature reveals several contradictions of inhibition and activation, the major difference being whether inhibitors were added to growth media or added to cells during measurement. Cartwright et al. (1987) found that ethanol

inhibited the H +-ATPase while Monteiro and Sa-Correia (1998) reported its activation by ethanol. Alexandre et al. (1996) described H +-ATPase inhibition and subsequent activation by decanoic acid. Fernandes et al. (1998) reported a similar effect by copper. Inhibition of the H +-ATPase appears to be a common phenomenon by lipophilic antimicrobials, such as 2-methylaminoethyl esters (Lachowicz et al., 1998), nonylphenol (Karley et al., 1997) and β-pinene (Uribe et al., 1985). Examination of all the data suggests that the plasma membrane H +-ATPase is sensitive to inhibition by lipid-soluble compounds but that given time at sub-lethal/sub-growth inhibitory concentrations, the cell responds with a positive feedback to restore activity. It is possible that this effect is mediated via membrane fluidity (either bulk membrane fluidity of in the immediate vicinity of particular membrane proteins) with modification of membrane lipid unsaturation to restore homeostasis. Here we show immediate H +-ATPase activity inhibition by sorbic acid, following a short-term increase in PMA1 gene transcript levels, and much later, an H +-ATPase activity increase at sub-inhibitory concentration if the measurement takes place without sorbic acid. The data on decanoic acid and octanoic inhibition and later activation of H +-ATPase (Alexandre et al., 1996; Viegas et al., 1998) suggests adaptive changes in membrane fluidity via fatty acid and sterol composition, which may restore H +-ATPase activity to normal and increase it upon removal of the inhibitor. This suggestion is supported by the lack of increase in PMA1 transcription shown here at 3 h (Fig. 6), and by a study showing modification of membrane lipids in Penicillium roquefortii grown in the presence of sorbic acid (Sergeeva et al., 2009). Improved efficacy of sorbic acid as a food preservative may then be achieved using other factors affecting membrane composition and fluidity. In conclusion, sorbic acid and acetic acid exert very different modes of inhibition in S. cerevisiae. Acetic acid behaves as a classic weak-acid preservative causing inhibition via intracellular acidification. Inhibition of yeast by sorbic acid is as a membrane-active compound, primarily affecting membrane proteins. A fundamental membrane process, the plasma-membrane H+-ATPase proton pump, has been found to be inhibited by sorbic acid at the growth-inhibitory concentration. Acknowledgements This work was funded by a Defra/BBSRC Link award (FQ128, BB/G016046/1, awarded to D.B.A.) in conjunction with GlaxoSmithKline, DSM Food Specialities and Mologic Ltd. References Abbott, D.A., Knijnenburg, T.A., de Poorter, L.M.I., Reinders, M.J.T., Pronk, J.T., van Maris, A.J.A., 2007. Generic and transcriptional responses to different weak organic acids in anaerobic chemostat cultures of Saccharomyces cerevisiae. FEMS Yeast Research 7, 819–833. Alexandre, H., Mathieu, B., Charpentier, C., 1996. Alteration in membrane fluidity and lipid composition, and modulation of H+-ATPase activity in Saccharomyces cerevisiae caused by decanoic acid. Microbiology 142, 469–475. Anon, 1995. EC Directive 95/2/EC; food additives other than colours and sweeteners. EC Official Journal (L61), 1–40 (March 18th). Arami, S.-I., Hada, M., Tada, M., 1997. Reduction of ATPase activity accompanied with photodecomposition of ergosterol by near-UV irradiation in plasma membranes prepared from Saccharomyces cerevisiae. Microbiology 143, 2465–2471. Bauer, B.E., Rossington, D., Mollapour, M., Mamnun, Y., Kuchler, K., Piper, P.W., 2003. Weak organic acid stress inhibits aromatic amino acid uptake by yeast, causing a strong influence of amino acid auxotrophies on the phenotypes of membrane transporter mutants. European Journal of Biochemistry 270, 3189–3195. Breslow, D.K., Cameron, D.M., Collins, S.R., Schuldiner, M., Stewart-Ornstein, J., Newman, H.W., Braun, S., Madhani, H.D., Krogan, N.J., Weissman, J.S., 2008. A comprehensive strategy enabling high-resolution functional analysis of the yeast genome. Nature Methods 5, 711–718. Carmelo, V., Santos, H., Sa-Correia, I., 1997. Effect of extracellular acidification on the activity of plasma membrane ATPase and on the cytosolic and vacuolar pH of Saccharomyces cerevisiae. Biochimica et Biophysica Acta 1325, 63–70. Cartwright, C.P., Veasey, F.J., Rose, A.H., 1987. Effect of ethanol on activity of the plasma-membrane ATPase in, and accumulation of glycine by, Saccharomyces cerevisiae. Journal of General Microbiology 133, 857–865.

M. Stratford et al. / International Journal of Food Microbiology 161 (2013) 164–171 Cid, A., Perona, R., Serrano, R., 1987. Replacement of the promoter of the yeast plasma membrane ATPase gene by a galactose-dependent promoter and its physiological consequences. Current Genetics 12, 105–110. Cole, M.B., Keenan, M.H.J., 1987. Effects of weak acids and external pH on the intracellular pH of Zygosaccharomyces bailii, and its implications in weak-acid resistance. Yeast 3, 23–32. Conway, E.J., Downey, M., 1950. pH values of the yeast cell. Biochemical Journal 47, 355–360. Fernandes, A.R., Sa-Correia, I., 1998. Transcription pattern of PMA1 and PMA2 genes and activity of plasma membrane H+-ATPase in Saccharomyces cerevisiae during diauxic growth and stationary phase. Yeast 20, 207–219. Fernandes, A.R., Pexoto, F.P., Sa-Correia, I., 1998. Activation of the H+-ATPase in the plasma membrane of cells of Saccharomyces cerevisiae grown under mild copper stress. Archives of Microbiology 171, 6–12. Hammond, S.M., Lambert, P.A., 1978. Antibiotics and Antimicrobial Action. Edward Arnold (Publishers) Limited, London. Hatzixanthis, K., Mollapour, M., Seymour, I., Bauer, B.E., Krafp, G., Schuller, C., Kuchler, K., Piper, P.W., 2003. Moderately lipophilic carboxylate compounds are the selective inducers of the Saccharomyces cerevisiae Pdr12p ATP-binding cassette transporter. Yeast 20, 575–585. Holyoak, C.D., Stratford, M., McMullin, Z., Cole, M.B., Crimins, K., Brown, A.J.P., Coote, P.J., 1996. Activity of the plasma membrane H+-ATPase and optimal glycolytic flux are required for rapid adaptation and growth of Saccharomyces cerevisiae in the presence of the weak-acid preservative sorbic acid. Applied and Environmental Microbiology 62, 3158–3164. Kabara, J.J., Swieczlowski, D.M., Conley, A.J., Truant, J.P., 1972. Fatty acids and derivatives as antimicrobial agents. Antimicrobial Agents and Chemotherapy 2, 23–28. Kane, P.M., 2006. The where, when, and how of organelle acidification by the yeast vacuolar H+-ATPase. Microbiology and Molecular Biology Reviews 70, 177–191. Karley, A.J., Powell, S.I., Davies, J.M., 1997. Effect of nonylphenol on growth of Neurospora crassa and Candida albicans. Applied and Environmental Microbiology 63, 1312–1317. Keenan, M.H.J., Rose, A.H., 1979. Plasma-membrane lipid unsaturation can affect the kinetics of solute accumulation by Saccharomyces cerevisiae. FEMS Microbiology Letters 6, 133–137. Kobayashi, H., Yamamoto, M., Aono, R., 1998. Appearance of a stress-responsive protein, phage-shock protein A, in Escherichia coli exposed to hydrophobic organic solvents. Microbiology 144, 353–359. Krebs, H.A., Wiggins, D., Stubs, M., Sols, A., Bedoya, F., 1983. Studies on the mechanism of the antifungal action of benzoate. Biochemistry Journal 214, 657–663. Kurtzman, C.P., 2003. Phylogenetic circumscription of Saccharomyces, Kluyveromyces and other members of the Saccharomycetaceae, and the proposal of the new genera Lachancea, Nakaseomyces, Naumovia, Vanderwaltozyma and Zygotorulaspora. FEMS Yeast Research 4, 233–245. Lachowicz, T.M., Krasowska, A., Luczynski, J., Witek, S., 1998. Plasma membrane H+ATPase activity in wild type and mutants of Saccharomyces cerevisiae treated with some lysosomatropic drugs. Folia Microbiologica (Praha) 43, 201–203. Lapathitis, G., Kotyk, A., 1998. Different sources of acidity in glucose-elicited extracellular acidification in the yeast Saccharomyces cerevisiae. Biochemistry and Molecular Biology International 46, 973–978. Legisa, M., Golic Grdadolnik, S., 2002. Influence of dissolved oxygen concentration on intracellular pH and consequently on growth rate of Aspergillus niger. Food Technology and Biotechnology 40, 27–32. Liu, Z.L., Palmquist, D.E., Ma, M., Liu, J., Alexander, N.J., 2009. Application of a master equation to quantitative mRNA analysis using qRT-PCR. Journal of Biotechnology 143, 10–16. Macris, B.J., 1975. Mechanism of benzoic acid uptake by Saccharomyces cerevisiae. Applied Microbiology 30, 503–506. Martinez-Munoz, G.A., Kane, P.M., 2006. Vacuolar and plasma membrane proton pumps collaborate to achieve cytosolic pH homeostasis in yeast. Journal of Biological Chemistry 283, 20309–20319.

171

Melin, P., Stratford, M., Plumridge, A., Archer, D.B., 2008. Auxotrophy for uridine increases the sensitivity of Aspergillus niger to weak-acid preservatives. Microbiology 154, 1251–1257. Monteiro, G.A., Sa-Correia, I., 1998. In vivo activation of yeast plasma membrane H+ATPase by ethanol: effect on the kinetic parameters and involvement of the carboxyl-terminus regulatory domain. Biochimica et Biophysica Acta 1370, 310–316. Neal, A.L., Weinstock, J.O., Lampen, J.O., 1965. Mechanisms of fatty acid toxicity for yeast. Journal of Bacteriology 90, 126–131. Pearce, A.K., Booth, I.R., Brown, A.J.P., 2001. Genetic manipulation of 6-phosphofructo1-kinase and fructose 2,6-bisphosphate levels affects the extent to which benzoic acid inhibits the growth of Saccharomyces cerevisiae. Microbiology 147, 403–410. Pilkington, B.J., Rose, A.H., 1988. Reactions of Saccharomyces cerevisiae and Zygosaccharomyces bailii to sulphite. Journal of General Microbiology 134, 2823–2830. Piper, P.W., 2011. Resistance of yeasts to weak organic acid preservatives. Advances in Applied Microbiology 77, 97–113. Piper, P.W., Mahe, Y., Thompson, S., Pandjaitan, R., Holyoak, C., Egner, R., Muhlbauer, M., Coote, P., Kuchler, K., 1998. The Pdr12 ABC transporter is required for the development of weak organic acid resistance in yeast. EMBO Journal 17, 4257–4265. Pitt, J.I., Hocking, A.D., 1997. Fungi and Food Spoilage, 2nd edn. Blackie Academic and Professional, London, Weinhein, New York, Tokyo, Melbourne, Madras. Prasad, R., Rose, A.H., 1986. Involvement of lipids in solute transport in yeasts. Yeast 2, 205–220. Reinhard, L., Radler, F., 1981. The action of sorbic acid on Saccharomyces cerevisiae. 1. Effect on growth and on the aerobic and anaerobic metabolism of glucose. Zeitscrift fur Lebensmittel- Untersuchung und -Forschung 172, 278–283. Sergeeva, Ya.E., Galanina, L.A., Kochkina, G.A., Feofilova, E.P., 2009. The effect of the preservative sorbic acid on the lipid composition of the ascomycete fungus Penicillium roquefortii Thom. Microbiology (Russia) 78, 630–635. Sikkema, J., de Bont, J.A.M., Poolman, B., 1995. Mechanisms of membrane toxicity of hydrocarbons. Microbiology Reviews 59, 201–222. Steels, H., James, S.A., Roberts, I.N., Stratford, M., 2000. Sorbic acid resistance: the inoculum effect. Yeast 16, 1173–1183. Stratford, M., 1983. Sulphite metabolism and toxicity on Saccharomyces cerevisiae and Saccharomycodes ludwigii. Ph.D. thesis, University of Bath, UK. Stratford, M., 2006. Food and beverage spoilage yeasts. In: Querol, A., Fleet, G.H. (Eds.), Yeasts in Foods and Beverages. Springer Verlag, Berlin, Heidelberg, pp. 335–379. Stratford, M., Anslow, P.A., 1997. Evidence that sorbic acid does not inhibit spoilage yeasts as a “classic weak-acid preservative”. Letters in Applied Microbiology 27, 203–206. Stratford, M., Rose, A.H., 1986. Transport of sulphur dioxide by Saccharomyces cerevisiae. Journal of General Microbiology 132, 1–6. Stratford, M., Plumridge, A., Nebe-von-Caron, G., Archer, D.B., 2009. Inhibition of spoilage mould conidia by acetic acid and sorbic acid involves different modes of action, requiring modification of the classical weak-acid theory. International Journal of Food Microbiology 136, 37–43. Tarsio, M., Zheng, H., Smardon, A.M., Martinez-Munos, G.A., Kane, P.M., 2011. Consequences of loss of Vph1-containing (vacuolar) V-ATPases for overall pH homeostasis. Journal of Biological Chemistry 286, 28089–28096. Uribe, S., Ramirez, J., Pena, A., 1985. Effects of β-pinene on yeast membrane functions. Journal of Bacteriology 161, 1195–1200. Venema, K., Palmgren, M.G., 1995. Metabolic modulation of transport coupling ratio in yeast plasma membrane H+-ATPase. Journal of Biological Chemistry 270, 19659–19667. Viegas, C.A., Almeida, P.F., Cavaco, M., Sa-Correia, I., 1998. The H+-ATPase in the plasma membrane of Saccharomyces cerevisiae is activated during growth latency in octanoic acid-supplemented medium accompanied by the decrease in intracellular pH and cell viability. Applied and Environmental Microbiology 64, 779–783.