INTRODUCTION
The study of microbial acetic acid tolerance is relevant in different fields of applied microbiology. Acetic acid, like other weak acids, such as sorbic acid and lactic acid, traditionally has been used as a preservative agent in food and beverages, where it prevents microbial spoilage by arresting the growth of yeasts and other fungi (
1). However, certain strains of the species
Zygosaccharomyces bailii and
Saccharomyces cerevisiae still grow in the presence of relatively highly weak acid concentrations (
2,
3), and, therefore, it is crucial to understand the underlying tolerance mechanisms in order to avoid food spoilage more effectively. More recently, understanding acetic acid tolerance of the platform yeast
S. cerevisiae became important in the field of industrial biotechnology once hydrolysates of lignocellulosic biomass were considered renewable feedstock for microbial fermentations (
4). Notably, the acetic acid concentrations in those hydrolysates can reach up to 133 mM (8 g liter
−1) (
5–7), at which the acid becomes a strong inhibitor of microbial growth and fermentation, especially at the low medium pH values typically used in industrial batch fermentations. Therefore, an understanding of the molecular mechanisms underlying
S. cerevisiae tolerance to acetic acid is important for the generation of robust industrial strains that are able to ferment lignocellulosic hydrolysates efficiently.
The inhibitory effect of acetic acid is associated predominantly with its undissociated form, which can diffuse across the plasma membranes of cells mainly by simple diffusion (
8). Once inside the cytoplasm, acetic acid (p
Ka = 4.76) dissociates into a proton and its counterion, resulting in a decrease in intracellular pH and an accumulation of acetate anions. The yeast
S. cerevisiae has developed several mechanisms by which it can counteract the harmful effects that acetic acid exerts on the cells. In general, adaptation to acetic acid has been associated with the abilities to recover intracellular pH (
3,
9–11), to inhibit further uptake of acetic acid (
12), to activate multidrug transporters to pump out acetate anions (
3,
13), and to adjust the membrane lipid profile (
14). Among these mechanisms, recovery of intracellular pH is thought to be of predominant importance in the responses of
S. cerevisiae to acetic acid (
9). In fact, exposure of cells to acetic acid has been shown to increase the activities of plasma membrane and vacuolar H
+-ATPases, which pump protons out of the cytosol (
3,
11,
13,
15). Another indication for the importance of pH homeostasis in weak acid tolerance is given by two studies that investigated interspecies diversity with regard to short-term changes in intracellular pH upon exposure to weak acid. It has been suggested that the higher tolerance of the species
Z. bailii and
Candida krusei compared to that of
S. cerevisiae is a consequence of their ability to preserve physiological pH better after shifting to acid-containing medium (
16,
17).
Although
S. cerevisiae has an innate tolerance to acetic acid, moderate to high concentrations have been shown to affect the cell's physiology negatively (
18,
19). A frequently reported effect is significant prolongation of the latency phase in the presence of inhibitory acetic acid concentrations (
20–23). This effect was demonstrated recently to be attributable to the fact that only a relatively small fraction of cells in the entire population are able to resume proliferation in the presence of acetic acid (
20). The size of this fraction was shown to decrease with increasing acetic acid concentrations, in a strain-dependent manner. The occurrence of such a fraction was observed previously in populations of
S. cerevisiae cells exposed to other weak acids (
24).
The fact that single cells of a genetically uniform population show different levels of stress tolerance is not a novel observation (
25–27). In fact, phenotypic cell-to-cell heterogeneity is assumed to be a strategy for microbial populations to survive unanticipated environmental changes. Although research regarding the molecular basis of cell-to-cell heterogeneity is still in its infancy, a few studies provide the first insights into the possible mechanisms involved (
25,
27–29). With regard to weak acid tolerance, it has been suggested that cell-to-cell heterogeneity could be explained by the variations in cytosolic pH (pH
c) among individual cells at the moment when the cells are exposed to the acid (
24,
30). It is assumed that only cells with pH
c values around neutrality start cell division in the presence of the acid (
24). More recently, a study of
Z. bailii proposed that tolerance to weak acids is due to a subpopulation of cells showing low pH
c values of around 5.5 (
30). As the uptake of weak acids appears to involve mainly a simple mechanism based on diffusion of the protonated form, lower pH
c values would allow the outside/inside equilibrium to be reached at lower intracellular concentrations of acid, thus reducing the accumulation of anions and the amounts of protons that have to be removed in order to restore the pH
c. However, no experimental evidence based on correlating pH
c kinetics with the cell's ability to proliferate in the presence of acid has been provided so far.
In the current study, we recorded pHc changes of individual cells in S. cerevisiae populations during a shift from nonstress to acetic acid stress conditions, and we correlated both the initial pHc and the maximal pHc drop with the cell's ability to proliferate in the presence of the acid. Data are provided for two different S. cerevisiae strains which significantly differed in acetic acid tolerance.
DISCUSSION
In this study, we showed that cell-to-cell heterogeneity in acetic acid tolerance could be at least partly explained by the variations in pHc values of individual cells at the moment immediately before acetic acid exposure. However, the acetic acid tolerance of an S. cerevisiae cell is also affected by genetic factors, as became obvious when S. cerevisiae strains differing in acetic acid tolerance were studied.
The first part of the current work focused on the cell-to-cell heterogeneity in acetic acid tolerance in the well-studied prototrophic laboratory strain CEN.PK113-7D. We clearly demonstrated that the initial pHc of individual S. cerevisiae cells determined the magnitude of the pHc drop after exposure to acetic acid. Indeed, cells with low initial pHc values (ranging from 6.6 to 7.1) experienced less severe drops in pHc than did cells with high initial pHc values (ranging from 7.1 to 7.9).
Based on our data, it is tempting to assume that the correlation between the initial pH
c and the magnitude of the pH
c drop can be explained simply by the fact that the equilibrium between undissociated acetic acid and internal acetate is dependent on the pH
c. In fact, a lower initial pH
c results in a smaller proportion of acetic acid that dissociates inside the cell, thereby avoiding excessive acid accumulation, as proposed previously by Stratford et al. (
30). Those authors proposed a correlation between the pH
c and the intracellular accumulation of acetic acid at the average population level for
Z. bailii. However, the possibility that the difference in the pH
c drops observed for cells with low versus high initial pH
c values could also be attributed to other intrinsic physiological factors cannot be excluded, since it was shown previously that pH
c acts as an intracellular signal in yeast (
38). For instance, H
+-ATPase pumps might be more active in cells with lower initial pH values, thus contributing to fast removal of newly produced cytosolic protons. In fact, it has been demonstrated that Pma1, the major yeast H
+-ATPase, is activated when intracellular pH decreases (
39). Further studies are necessary to evaluate these hypotheses.
Valli et al. (
40) demonstrated that
S. cerevisiae mutants with high intracellular pH are those that accumulate the most lactic acid (when expressing heterologous lactate dehydrogenase). Obviously, the high pH was used as an indicator for improved lactic acid tolerance. This result is not in contradiction to our findings. The context of the study by Valli et al. (
40) is very different from that of the current one. In particular, it has to be considered that the cells were continuously producing lactic acid intracellularly and that mutations that led to improved tolerance were caused by an increased ability to maintain physiological pH even though lactic acid accumulated. In the current study, nonstressed cells were challenged by sudden exposure to extracellular acetic acid.
As only a fraction of CEN.PK113-7D cells (approximately 60%) were shown to contribute to growth of the culture in the presence of 96 mM acetic acid (pH 4.5), we also studied whether the initial pH
c determined the cell's ability to resume proliferation. Indeed, only cells with low initial pH
c values were able to recover pH
c to neutral values and resume proliferation in the presence of the acid. Moreover, the time periods required for individual CEN.PK113-7D cells to resume proliferation (lag phases) varied among the cells and were correlated with both their initial pH
c values and their budding status before the addition of acetic acid. The latest observation is congruent with the fact that nonbudding cells are considered more stress resistant (
41).
In contrast to the proliferating CEN.PK113-7D cells, cells with high initial pH
c values experienced severe drops in pH
c and were unable to resume proliferation. Our previous study showed that these nonproliferating cells do not die immediately upon exposure to acetic acid and that they stay viable for a relatively long period of time (
20). A possible explanation for the fact that cells with large drops in pH
c cannot resume proliferation is that these cells are unable to restore their pH
c to neutral values, which might be caused by a deficiency in ATP. In fact, ATP drives the activity of plasma membrane and vacuolar H
+-ATPases, which pump protons out of the cytosol in order to restore the pH
c after weak acid stress (
24). However, a recent study based on measuring average intracellular ATP levels in
S. cerevisiae populations suggests that ATP depletion alone is not the only cause of growth inhibition (determined by optical density measurements) upon acetic acid stress (
42).
As our data obtained from strain CEN.PK113-7D showed that only cells with relatively low initial pHc values were able to resume proliferation, the question arose as to whether a population of cells from a strain with higher acetic acid tolerance would contain a larger fraction of cells with relatively low initial pHc values. The analysis of the highly tolerant strain MUCL 11987-9 showed that the distribution of initial pHc values was indeed broader than for CEN.PK113-7D. Most importantly, a fraction of cells exhibited lower initial pHc values than did the reference strain. These data might explain why a larger fraction of cells in the MUCL 11987-9 population experienced less dramatic pHc drops, which might facilitate their proliferation in the presence of acetic acid. All cells from MUCL 11987-9 were able to proliferate in the presence of 96 mM acetic acid (pH 4.5), and this was even independent of the initial pHc of the cells. Therefore, the larger fraction of cells with lower initial pHc values in strain MUCL 11987-9 cannot be the only parameter that explains the difference in the acetic acid tolerances of the two strains. Obviously, the cells of strain MUCL 11987-9 have a genetically determined greater capability to recover from severe pHc drops. Our results emphasize the relevance of studying weak acid tolerance at the single-cell level, as well as the population level, and can serve as a starting point for developing industrially useful strains that are tolerant to weak acids.
ACKNOWLEDGMENTS
This work was supported by grant PIM2010EEI-00610 (ERA-NET IB, MINECO, Spain), an Ajut 2014SGR-4 award, and an Institució Catalana de Recerca i Estudis Avançats Academia 2009 award to J.A. and the ERA-NET Scheme of the 6th EU Framework Program (INTACT; German Federal Ministry of Education and Research project 0315933 [to E.N.]). M.F.-N. received a personal stipend from Colciencias (Colombia).
We thank Johan Thevelein (KU Leuven, Belgium) and Jack Pronk (TU Delft, The Netherlands) for kindly providing us with the strains MUCL 11987 and CEN.PK113-7D, respectively. We are also grateful to Gertien J. Smits (University of Amsterdam, The Netherlands) for providing us with the plasmid pYES-PACT1-pHluorin (URA3). We thank Nuria Barba, Inka Schönebeck, Martina Carrillo, Solvejg Sevecke, and Meneka Ruvi Rupasinghe for technical support and Mathias Klein for fruitful discussions.