- Top of page
- Materials and methods
- Conflict of interest
The production of wine involves a variety of microbial transformations comprising a complex succession of various yeast and bacterial species. Oenococcus oeni has long been reported as the main lactic acid bacteria (LAB) species associated with malolactic fermentation (MLF), crucial step for winemaking process that can occur spontaneously. However, this step can start randomly, and any delay could lead to an alteration of wine quality. These delays are due to very harsh environmental conditions in the wine for bacterial survival and growth and various physical/chemical factors, such as ethanol, pH and temperature, are known to affect the growth of the LAB responsible for the MLF of wine. O. oeni is the LAB species most resistant to the presence of ethanol in wine and able to grow in the hostile environment of wine by the activation of several mechanisms (G-Alegrìa et al. 2004).
A better knowledge of stress physiology may be useful to optimize survival of starter cultures of O. oeni that the winery practices recommend to use for direct inoculation into wines to improve the control of MLF (Nielsen et al. 1996). However, induction of MLF by inoculation with malolactic starters is not effective in ‘difficult wines’ (e.g. in wines having a pH below to 3·2) because of significant cell mortality (da Silveira et al. 2002). To overcome this problem, adaptation processes have been shown to enhance the survival of O. oeni cells to stress conditions in wine, and the increase in cell survival is linked to stress response mechanisms (da Silveira et al. 2002).
The complete genome sequence of O. oeni strain PSU-1 was previously studied to advance the study of O. oeni and to provide fundamental information on the genetic endowment of this micro-organism (Zé-Zé et al. 1998, 2000; Mills et al. 2005).
Moreover, the differential expression of genome in O. oeni strains under stress conditions was previously studied by the development, and the optimization of the fluorescent differential display (FDD) technique that allows the identification of gene expression changes, associated with differential microbial behaviour under different stress conditions with a better stress response definition and a better discrimination of starter cultures (Sico et al. 2009).
The genome analysis reveals potential survival strategies, as well as metabolic properties that enable O. oeni to effectively compete in the wine environment. The existence of such unique features can be viewed as evolutionary adaptation to the wine environment (Beltramo et al. 2006). So a combined knowledge of genome features and specific gene expression is required for understanding the adaptive mechanisms of O. oeni to the wine environment (Beltramo et al. 2006). Although several studies (Carreté et al. 2002; Bourdineaud et al. 2003; da Silveira et al. 2003; Grandvalet et al. 2005, 2008; Olguìn et al. 2009) analysed few mechanisms that enable O. oeni to withstand stress conditions, more information about the mechanisms involved in the adaptation of O. oeni to stress conditions is required.
Quantitative real-time PCR (qPCR) has become a routine technique for gene expression analysis and a better understanding of these patterns is expected to provide insights into complex regulatory networks to obtain the identification of genes relevant in adaptation processes (Vandesompele et al. 2002). There is still no consensus about appropriate normalization of qPCR raw data, essential to compensate for experimental error that can be introduced at various stages throughout the procedure. The most commonly used strategy is the use of an internal reference gene, a so-called housekeeping gene. Although several studies have used only one gene for normalization, recent studies compared the use of a single and multiple reference genes for normalization of qPCR expression data finding that the use of a single reference gene was unreliable and revealed a risk of misinterpretation of expression data (Vaudano et al. 2011; Sumby et al. 2012). Therefore, these studies validated the use of multiple internal control genes, but revealed no ideal reference genes because the expression of many genes, which have been used as internal controls in qPCR experiments, is influenced by metabolic conditions, growth phase or experimental conditions (Theis et al. 2007). Therefore, each internal control gene needs to be validated within a given experimental set-up before it can be used for the normalization of qPCR data to avoid the achievement of erroneous results (Theis et al. 2007).
In this study, ten potential internal control genes were tested under different culture and stress conditions. A comparative study of their stability has been performed to select the most adapted internal control for further studies. These internal controls were then used to study the expression level of geranylgeranyl pyrophophate synthase (ggpps) gene in O. oeni under ethanol stress conditions.
- Top of page
- Materials and methods
- Conflict of interest
Fluorescent differential display profiles were produced carrying out an assay system based on FDD method, by using a single fluorescently labelled random primer for detecting and isolating differentially expressed fragments under different stress situations. Distinct bands in response to different ethanol concentrations were obtained and only the fragments present in the profiles of stress exposed samples and absent in all control profiles were considered indicating a specific response. The resulting fingerprints showed the presence and the absence of specific products between each stressed sample and the control (data not shown).
To identify genes within the resulting O. oeni genome database that are differentially expressed in response to various changes, differentially amplified FDD products were sequenced and results of the early sequence analysis confirmed the effectiveness of the system at identifying differentially expressed transcripts on FDD fingerprints correlated with microbial tolerance to stress situations and revealed that several of the products can be referred to specific genes already known in literature under stress conditions (Guzzo et al. 2000; Beltramo et al. 2004; Bourdineaud et al. 2004; Desroche et al. 2005). In particular, this study was focused on a transcript of 279 bp identified in the response of O. oeni S12 strain to 12% ethanol stress. The sequence alignment analyses of bands revealed a significant similarity to a gene coding a ggpps (NCBI no. ABJ56986) which was originally found to be present in O. oeni PSU-1 complete genome (NCBI no. CP000411).
For the quantification of ggpps gene expression in the different stress conditions, 10 reference genes commonly used as internal control were employed (rpoA, gyrB, gapA, gmk, recA, rpoB, ftsZ, pta, ldhD, rrs) involved in different metabolic aspects. All ten genes were evaluated for their potential as internal control for O. oeni gene expression experiments. qPCR assays were carried out using each primer pair for the ten selected reference genes on each sample. To identify the most stable genes to use for the normalization of the expression level of the target gene ggpps, geNorm, provided by qbase_plus 2.4 software, was used. It was able to supply a gene-stability measure (M) defined as the average pairwise variation of a particular gene with all other potential reference genes (Vandesompele et al. 2002). The genes with the lowest M value were considered the most stable, and they were pta and rpoB, with a M value of 0·38–0·40 (Fig. 1). Instead, the genes with the highest M values were rpoA and rrs, with a M value of about 1·18 and 1·85, respectively (Fig. 1).
Figure 1. Candidate reference genes classified according to their average expression stability values (M) by geNorm analysis. The highest M values characterize the least stable genes.
Download figure to PowerPoint
To find the optimal number of reference genes to use for an accurate normalization of the expression level of a target gene, the value Vn/n+1 was determined (Vandesompele et al. 2002); the lowest Vn/n+1 represented the optimal number of reference genes to use to obtain the most accurate data normalization that is possible. Based on these criteria, the best genes identified by geNorm, in order of stability, were pta, rpoB, gapA, recA and gyrB (Figs 1 and 2). Using the five most stable reference genes, an accurate and reliable normalization of qPCR data was achieved.
Figure 2. Determination of the optimal number of internal control genes showing pairwise variation values (Vn/n+1) by geNorm analysis. Each bar represents the change in normalization accuracy through a stepwise addition of reference genes according to the classification showed in Fig. 1. The lower value of Vn/n+1 represents the better normalization reachable using this particular set of reference genes.
Download figure to PowerPoint
Once stable internal control genes were identified, the expression level of the target ggpps gene was analysed by using the geometric mean of copy numbers as normalization factor.
Using cells grown in absence of ethanol as calibration (control) condition, the expression gene analysis produced various changes in the transcription level of the gene in response to different ethanol concentrations.
Figure 3 shows results of the relative expression levels of ggpps gene in the presence of different concentrations of ethanol. It can be observed that transcription was activated in all the media after 1 h of incubation, and statistically significant differences (P < 0·05) were found comparing transcription in the control condition (0·40 relative quantity) with that in different concentrations tested. The increase in ethanol concentration in the growth medium resulted in an increase in the ggpps gene expression (Fig. 3).
Figure 3. Relative expression levels of the geranylgeranyl pyrophosphate synthase (ggpps) gene in S12 O. oeni cells treated with different ethanol concentrations: in absence of ethanol (control); ethanol 7% (EtOH 7%); ethanol 12% (EtOH 12%); ethanol 13% (EtOH 13%); ethanol 15% (EtOH 15%).
Download figure to PowerPoint
The transcript level of ggpps increased significantly in the presence of higher ethanol concentrations (12, 13 and 15%), showing relative quantities of 1·1, 1·7 and 2·9, respectively, while ggpps showed no significant expression increase in the presence of 7% ethanol, with a relative amount (0·45) very similar to that in the control (0·40) (Fig. 3). The 12% ethanol concentration caused an increase in transcription level of 2·50 times compared with control, while in the presence of 13% of ethanol, the transcription level was higher of almost 4·25 times than that found in the control condition. Finally, the transcription reached the maximum relative quantity (2·9) observed after the stronger stress, due to 15% ethanol presence, that led to a transcription level increase in 7·25 times compared with the control condition (Fig. 3).
- Top of page
- Materials and methods
- Conflict of interest
The previous study of Sico et al. (2009) provided an innovative FDD method, with a high level of reproducibility and quality for studying and probing the knowledge of the relationship between differential genome expression and different stresses tolerance. It proved O. oeni strains respond to stimuli through the differential expression of transcripts in order to survive and adjust to the stresses. In this study, the identification and the sequence analysis of a transcript, differentially expressed in response to ethanol presence, revealed the ethanol effect on the physicochemical state and biological functions of cells and the activation of different devices of cell response by involving various genes, such as that coding a ggpps (NCBI no. ABJ56986) which was originally identified in O. oeni PSU-1 complete genome (NCBI no. CP000411).
Geranylgeranyl pyrophosphates (GGPPs) is an enzyme that belongs to the family of E-prenyl diphosphate (prenyl PP) synthases and catalyses the condensation of isopentenyl diphosphate (IPP) with its allylic isomer, dimethylallyl diphosphate (DMAPP), to produce GGPP, an essential isoprenoid involved in several biosynthetic pathways such as the biosynthesis of terpenoids, carotenoids and membrane stabilizers such as hopanoids, but also the synthesis of quinones and chlorophylls and the prenylation of proteins (Velayos et al. 2003).
Two distinct pathways occur for the biosynthesis of isoprenoids precursors, IPP and DMAPP. Eukaryotes, except plants, perform the mevalonate (MEV) pathway to convert acetyl coenzyme A (acetyl-CoA) to IPP, that is, transformed subsequently in its isomer DMAPP. In mammals, it has been demonstrated that MEV cascade is involved in many biological phenomena and cellular functions in which mevalonate act as a precursor of cholesterol and also of isoprenoids for farnesyl and geranylgeranyl molecules, which have an important signalling function (Tanaka et al. 2000).
In prokaryotes, with a few exceptions, IPP and DMAPP are produced by a mevalonate-independent pathway, the deoxyxylulose-5-phosphate (DXP) pathway, while plants use both MEV and DXP pathways (Wang and Ohnuma 2000). In plants, isoprenoids are important secondary metabolites, highly specific and synthesized in particular development phases that represent chemical substances for adaptation to stresses and defence products (Wang and Ohnuma 2000).
Synthesis of isoprenoids is intrinsic to all organisms and leads to a vast array of metabolites with diverse functions for cell survival. In humans and other mammals, the products of this pathway include essential molecules such as cholesterol, haeme A, ubiquinone, dolichol and farnesoids. The latter products include farnesyl pyrophosphate (FPP) and GGPP, which are precursors for protein prenylation and might serve as nuclear receptor ligands (Kavanagh et al. 2006). The study of Kavanagh et al. (2006) described the structure of human GGPP, and, by carrying out a sequence alignment with GGPP of other organisms (including a bacterial species), shared sequence identity, the presence of conserved regions so a possible common catalytic mechanism.
Eubacteria, like plants, have the type II synthase: in particular, in O. oeni, this enzyme is implicated in the biosynthesis of secondary metabolites, mainly in terpenoids, steroids and membrane lipids biosynthesis (KEGG PATHWAY Database www.genome.jp/kegg/pathway.html).
Isoprenoids, one of the largest groups of natural compounds, have a variety of roles in respiration, photosynthesis, membrane structure, allelochemical interactions and growth regulation (Sangari et al. 2010). All free-living organisms synthesize isoprenoids from the five carbon precursors IPP and its double-bond isomer dimethylallyl diphosphate (DMAPP). Recently, besides the usual pathways for the isoprenoids precursors biosynthesis, alternative pathways and metabolic intermediates have been proposed to be used for the biosynthesis of isoprenoid precursors in the cyanobacterium Synechocystis PCC 6803 illustrating how limited is still our knowledge of the alternative pathways that can be used by bacteria to synthesize their isoprenoids (Sangari et al. 2010).
Carotenoids are widely produced by plants and micro-organisms in which they accomplish important biological functions (Hagi et al. 2013). Among LAB, some carotenoid-producing enterococci and lactobacilli have been isolated from various origins such as food, plants and human clinical specimens (Garrido-Fernandez et al. 2010; Maraccini et al. 2012; Hagi et al. 2013).
In general, carotenoids contribute to tolerance to stresses, such as oxidative stress, because of their antioxidant ability derived from their conjugated double bonds (Hagi et al. 2013). In Gram-positive bacteria, the effects of carotenoids produced by Staphylococcus aureus on oxidative stress tolerance have been reported (Clauditz et al. 2006), and the heterologous expression of staphylococcal carotenoid biosynthesis genes improves H2O2 stress tolerance in Bacillus subtilis (Yoshida et al. 2009). Carotenoid production in LAB, which is one of the antioxidant mechanisms, is considered to play a role in the elimination of oxygen radicals. Carotenoids are lipophilic agents that are incorporated into the bacterial membrane. Chamberlain et al. (1991) reported that carotenoids could influence cell membrane fluidity and decrease the sensitivity of S. aureus to oleic acid. This finding implies that carotenoid production in LAB also may lead to changes in membrane fluidity. So, it is possible that the change in cell membrane fluidity caused by carotenoids could aid to prevent cell membrane disturbances and cell death induced by ethanol. However, the mechanism of tolerance to ethanol remains unknown.
LAB have been widely used for starters of food fermentation and probiotics; therefore, investigation of the stress tolerance mechanism of LAB is important to optimize its application in food fermentations and probiotics (Hagi et al. 2013). Antioxidant enzymes in LAB, such as glutathione, superoxide dismutase and catalase, have been studied and used to improve oxidative stress tolerance of LAB (Hagi et al. 2013).
A previous work (Garrido-Fernandez et al. 2010) observed that different Lactobacillus plantarum strains, widely used as probiotics in dairy products and dietary supplements, were able to produce high carotenoids amounts in particular environmental conditions. Therefore, carotenoid production should be considered as an important feature for the selection of novel probiotic L. plantarum strains; the use of selected high-carotenoid-producing strains could contribute to increase the total amount of antioxidants supplied in the human and animal diet (Garrido-Fernandez et al. 2010). In all cases, these effects have been related to either the antioxidant properties of carotenoids or their ability to stabilize bacterial cell membranes.
The presence of GGPP in secondary metabolism processes of O. oeni was confirmed by the following investigations, revealing the presence of ggpps gene in S12 O. oeni strain under both stress and optimal conditions. A quantitative analysis was carried out by qPCR technique in order to analyse the expression level of ggpps gene and its changes after exposure to ethanol stresses. The specific gene expression patterns reflect mechanisms involved in adaptation to environmental conditions, and the qPCR method is the best approach to obtain a better understanding of these patterns to provide insights into complex regulatory networks leading to the identification of genes relevant in adaptation processes (Vandesompele et al. 2002).
As an universal reference gene of which the expression level remains constant does not exist because the expression of many internal control genes is influenced by metabolic conditions, growth phase or experimental conditions (Dheda et al. 2005; Theis et al. 2007; Bustin et al. 2009; Ritz et al. 2009; Duary et al. 2010; Duquenne et al. 2010; Costantini et al. 2011; Vaudano et al. 2011; Sumby et al. 2012), we used multiple internal control genes applying the statistical algorithm geNorm (Vandesompele et al. 2002).
In this study, the choice of reference genes was based on historical precedents; Desroche et al. (2005) suggested that ldhD (D-lactate dehydrogenase) was the best reference gene among the others that are usually used for qPCR normalization in O. oeni, because of its ability to keep a stable transcriptional levels in different kind of stresses and in different growth stages of the cells, but our study demonstrated an intermediate stability of ldhD, so in this case, it cannot be used for the normalization.
In this work, results showed that, based on Vn/n+1 value obtained from geNorm analysis, the five most stable genes must be used to quantify the expression of ggpps gene, that are pta, rpoB (subunit β of DNA-direct RNA polymerase), gapA, recA (recombinase A) and gyrB in accordance with Sumby et al. (2012) that validated the use of multiple internal control genes to study esterase gene expression in O. oeni, suggesting that the most stable genes were ftsZ (filamentous temperature-sensitive mutant Z), pta (phosphotransacetylase), gapA (D-glyceraldehyde-3-phosphate dehydrogenase) and rpoA (sigma α factor). Moreover, Ritz et al. (2009) proved that rpoA was the most suitable internal control gene to study the stress response in C. jejuni, in contrast with our results that demonstrated rpoA had the lowest stability, and finally, Costantini et al. (2011), established that gyrB (gyrase subunit B) was the most stable gene in absolute to be used as a reference gene to analyse the expression level of mleP gene and two genes involved in the ABC transport system in O. oeni strains. These literature references confirm further that there is not an ideal universal reference gene.
After the characterization of a suitable reference system, the transcription levels of the ggpps gene were analysed and the data obtained suggested that ggpps could play an essential means that O. oeni uses to withstand and adapt to ethanol stresses. The results proved changes for the adaption to a particular stress condition; the ethanol is an agent affecting the physicochemical state and biological functions of cells and leading numerous modifications faced by different devices of cell response. Ethanol acts as a disordering cause of the O. oeni cytoplasmic membrane and the metabolic activities (da Silveira et al. 2003). It interacts with membranes at the lipid–water interface, weakening the hydrophobic barrier to the free exchange of polar molecules, thereby perturbing membrane structure and function (Chu-Ky et al. 2005) with fluidity reduction and changes of cell functionality as membrane rigidity obstructs structural variations indispensable to metabolic activities development (da Silveira et al. 2003).
Our work was effectuated to investigate the effect of ethanol exposure on the expression level of a gene involved in the metabolism of O. oeni to probe the mechanisms of ethanol tolerance correlated with adaptive changes. The analysis of results suggest that O. oeni adjusts the expression of genes to adapt to stress conditions and the high expression level of ggpps would allow a flow of isoprenoid precursors towards the carotenoids and related pathways to stabilize bacterial cell membranes, improving the cell membrane disturbances and preventing cell death induced by ethanol.
In conclusion, the performance of micro-organisms under ethanol stress conditions, such as those prevailing in wine, requires specific cellular features, including modification of metabolic activities to allow survival under such adverse conditions. Improving the knowledge of stress tolerance and adaptation mechanisms of the malolactic bacterium O. oeni is essential to enhance the efficiency of the malolactic starter in wine and to obtain the development of starters able to survive to direct inoculation with a large benefit for wine technology.
The analysis and the investigation of the involvement of ggpps gene in physiological changes of bacterial behaviour confirmed and clarified that the exposure to stress requires the activation of defence mechanism so that bacteria become more tolerant to adverse conditions. qPCR turned out to be an efficient approach to obtain a better understanding of the expression level patterns to provide insights into complex regulatory networks leading to the identification of genes relevant in adaptation processes.