Fingerprinting analysis of Oenococcus oeni strains under stress conditions

Authors


  • Editor: Linda Bisson

Correspondence: Maria Anna Sico, Dipartimento di Biologia, Difesa e Biotecnologie Agro-Forestali, Università degli Studi della Basilicata, Viale dell'Ateneo Lucano, 10, 85100 Potenza, Italy. Tel.: +39 971 205 572; fax: +39 971 205 503; e-mail: mariaanna.sico@unibas.it

Abstract

Oenococcus oeni strains from traditional Italian red wines of the Basilicata region were investigated on the basis of their physiological and molecular response to different temperatures and ethanol concentrations. All strains were highly resistant to different ethanol concentrations and it has been observed that 7% ethanol was able to stimulate the growth of strains in wine, and 12–13% of ethanol allowed their proliferation. Moreover, strain tolerances to 18 and 42 °C were observed. Fingerprinting analysis with fluorescent differential display-PCR and investigation of changes in gene expression during the tolerance process were carried out. The expression gene pattern reflects mechanisms involved in tolerance to environmental conditions. This study establishes and validates a method that enables, with a high reproducibility, different gene expression identification under stress conditions in lactic acid bacteria.

Introduction

In winemaking, Oenococcus oeni is the most common species of lactic acid bacteria (LAB) associated with malolactic fermentation (MLF), a process basically consisting of malic acid decarboxylation in wine after completion of the alcoholic fermentation performed by yeasts. The hostile environment of wine can lead to a delay in starting MLF; different physicochemical factors such as ethanol, temperature and pH are known to affect the growth of LAB in wine and this may lead to a number of processing problems, time consumption and alteration of wine. Oenococcus oeni is recognized as the best adapted bacterial species to grow in wine and it is frequently associated with spontaneous MLF. However, inoculation of O. oeni starter cultures directly into wine leads to significant cell mortality and, consequently, failure of MLF. More information about the mechanisms involved in the adaptation of O. oeni to stress conditions is required, particularly under winemaking conditions. A better knowledge of stress physiology may be useful to optimize the survival of starter cultures of O. oeni.

Several mechanisms enable O. oeni to withstand stress conditions: the generation of a proton motrice force (Salema et al., 1996), activation of membrane-bound H+-ATPase (Carretèet al., 2002), modification of membrane fluidity (Tourdot-Maréchal et al., 2000, Silveira et al., 2003) and stress protein synthesis (Guzzo et al., 2000). Many of these proteins function as molecular chaperons or protease that could participate in the refolding or degradation processes of denatured proteins in the cell (Craig et al., 1993).

The adaptation processes have been shown to enhance the survival of O. oeni cells under stress conditions in wine: heat shock at 42 °C (Guzzo et al., 1994) or growth in the presence of ethanol (8% v/v) (Silveira et al., 2003). The increase in cell survival is linked to stress response mechanisms: heat treatment has been shown to increase the synthesis of heat shock proteins (HSPs), notably a small HSP named Lo18 (Guzzo et al., 1997), and growth of O. oeni in the presence of ethanol leads to a modification in membrane composition (Silveira et al., 2003). Stress-induced proteins are molecular markers for the fitness of starter cultures and could be used as positive indicators for a culture that is fully adapted to resisting an upcoming stress condition (Sanders et al., 1999).

Previous studies about O. oeni stress response and MLF have used physiological and biochemical techniques (Guzzo et al., 2000; Tourdot-Maréchal et al., 2000; Silveira et al., 2003). This study presents a molecular approach to characterize O. oeni gene expression under several adverse conditions – fingerprinting analysis. Random amplification of polymorphic DNA (RAPD)-PCR was carried out on cDNA to analyse the gene expression pattern of O. oeni during the stress response to obtain a greater understanding of the behaviour of bacteria and to characterize the wild O. oeni strains for use as a starter culture during the winemaking process to improve the efficacy of the MLF.

The choice of this molecular approach was suggested by the latest employment of this PCR-based method targeting RNA, such as differential display PCR (DD-PCR), which has recently been used to identify differentially expressed genes in eukaryotic organisms (Hakki & Akkaya, 2001). As no poly(A)-dependent locking primer is necessary for DD-PCR, this method may prove useful in prokaryotic experimental systems. In fact, this technique has been successfully used with prokaryotic cells to evaluate differences in gene expression due to environmental changes (Wong & McClelland, 1994).

The aim of this work was to study the effect of ethanol and temperature on microbial growth and also to investigate a molecular method for the analysis of changes in the profiles of O. oeni that are well adapted to stress conditions for the purpose of formulating a starter culture for direct inoculation in wine.

Materials and methods

Bacterial strains and growth conditions

Five O. oeni wild strains (D29, D30, S1, S11 and S12) isolated previously from Aglianico wines produced in the Basilicata (southern Italy) were used in this study (Lechiancole et al., 2006; Sico et al., 2008). All strains were maintained as freeze-dried stocks in reconstituted (11% v/v) skim milk, containing 0.1% (v/v) ascorbic acid (J.T. Backer, Deventer, Holland) and routinely cultivated in Man–Rogosa–Sharpe broth (MRS; Oxoid Ltd, Basingstoke, Hampshire, UK) containing 20% (v/v) tomato juice (MRSb-TJ; Fluka Chemie, Sigma-Aldrich Chemie GmbH) and adjusted to pH 4.8. Samples were incubated at 30 °C under strict anaerobic conditions for 72 h.

Effect of temperature and ethanol on microbial growth

For each strain, a subculture in MRSb-TJ broth was obtained from the active stock culture using 1% (v/v) inoculum and incubation at 30 °C for 72 h.

The subculture was incubated at 18 and 42 °C anaerobically to evaluate the temperature effect and supplemented with ethanol to a final concentration of 7–12–13% (by volume) at 30 °C anaerobically to evaluate the ethanol effect on microbial growth. Control samples were incubated without ethanol at 30 °C. Bacterial growth was followed by OD600 nm up to 7–12 days. Absorbance data were plotted vs. time for each strain and for each tested condition.

RNA extraction and reverse transcription

After stress treatment, total RNA was extracted using the RNA isolation kit supplied by Gentra System Inc. (Minneapolis, MN) from 5 mL of MRSb-TJ culture. The quality of the RNA samples was verified on a 1% (v/v) agarose gel, and the concentration was determined by measuring the A260 nm using the NanoDrop® ND-1000 Spectrophotometer (NanoDrop Technologies Inc.).

To carry out the reverse transcription, 2 μg of total extracted RNA was treated with 2 U of DNAse (Ambion) as described by the manufacturer and then the synthesis of cDNA was carried out using the ProSTARTM First-Strand RT-PCR kit (Stratagene, La Jolla, CA).

Genotyping analysis

A 2-μL aliquot of first-strand cDNA synthesis reaction was used as a template for fluorescent differential display (FDD)-PCR amplification. This was performed using the random primer M13 marked at 5′ with the reactive fluorescent dye 6-carboxylfluorescein (6-FAM) (Metabion). The thermal cycling conditions were designed as follows: initial denaturation at 94 °C for 2 min, followed by 45 cycles at 94 °C for 30 s, at 35 °C for 36 s and at 72 °C for 7 s. The PCR products were separated by electrophoresis on 2% (w/v) agarose gels at 100 V for 4 h. Gels were stained with 0.5 μg mL–1 ethidium bromide (Serva) for 30 min. A 1-kb DNA ladder (EuroClone) was used as molecular weight and normalization gel standard.

In addition, the PCR products were analysed with the capillary sequencer ABI 3730 to obtain the number, size and peak intensity of each set of amplification product bands.

Compare lane analysis

Gel images were captured with GelDoc 2000 Apparatus (Bio-Rad) and digitized in diversity database software (Bio-Rad). Each PCR amplification was carried out in duplicate for each stressed strain and the banding patterns were analysed by lane to lane comparison.

Statistical analysis

Growth experiments were carried out at least in triplicate, and growth values under stress conditions were analysed statistically by the t-test. Means of the values were considered significantly different when P<0.05.

Results and discussion

Tolerance to temperature and ethanol

Oenococcus oeni strains from Aglianico wines previously characterized by Lechiancole et al. (2006) and Sico et al. (2008) were analysed. These five strains (D29, D30, S1, S11 and S12) present a very interesting MLF technological performance, and the specificity of indigenous strains of this type of wine could be of interest to local producers as a way to induce MLF with native starter culture (Sico et al., 2008). The first step in the development of the fingerprinting approach for the analysis of change in the profile expression of wild O. oeni strains best adapted to stress conditions was the determination of the tolerance to temperature and ethanol concentration. The growth of the strains was followed under different stress conditions: all strains showed maximal growth rate and maximal population when grown in an ethanol-free medium at 30 °C (Fig. 1). When O. oeni strains were incubated at 18 °C, growth rates decreased 50% and there was a short lag phase of up to 72–96 h. The growth at 42 °C was very reduced, but after 72 h of adaptation the lag phase was nevertheless reached (Fig. 1), suggesting that the synthesis of stress proteins was induced. This mechanism permits the microbial cell to survive and to proliferate. Indeed, it was shown in previous studies that the survival of O. oeni in wine and its ability to induce MLF were improved after direct inoculation with cells pretreated at 42 °C (Lounvaud-Funel, 1996; Coucheney et al., 2005).

Figure 1.

 Effect of ethanol and temperature on growth of Oenococcus oeni strains D29, D30, S1, S11 and S12. Cells were incubated in MRSb-TJ broth under the following conditions: ◆, 30°C without ethanol; ▪, with 7% ethanol; ▴, with 12% ethanol; ▵, with 13% ethanol; □, 42°C without ethanol; •, 18°C without ethanol. The results are expressed in OD600 nm over time (h).

When O. oeni strain populations are grown in a medium to which 7% ethanol has been added, a lag phase of 48–72 h appeared, and an increase in ethanol concentration [12% and 13%, the normal alcohol content of the original red wines from which strains were isolated (Aglianico wines)] produced a decrease in bacterial populations with respect to strains grown at 7% (v/v), but produced no major decrease with respect to strains grown at 18 °C without ethanol. This indicates the high degree of adaptation of our five O. oeni strains (D20, D30, S1, S11 and S12) to growth in red wine. Sampling points were the average of three independent experiments corresponding to the growth of the five strains studied.

The tolerance to ethanol varies from strain to strain and it is generally accepted that all O. oeni strains grow in a medium containing 10% ethanol and that small quantities of ethanol [3–5% (Britz & Tracey, 1990) or 7% (G-Alegìa et al., 2004)] can stimulate their growth.

In addition, the combined effect of the presence of 7–12% and 13% ethanol in the medium and a low incubation temperature of about 10 °C prevented bacterial growth in all strains studied (Fig. 2). Britz & Tracey (1990) and G-Alegìa et al. (2004) show, respectively, that the combination of 15 °C and 13% ethanol and the combination of 10 °C and 12% reduced, and in some cases prevented, growth of O. oeni strains under stress conditions.

Figure 2.

 Effect of ethanol and temperature on growth of Oenococcus oeni strains D29, D30, S1, S11 and S12. Cells were incubated in MRSb-TJ broth under the following conditions: ◆, 30°C without ethanol; ▪, 18°C without ethanol; ▴, 10°C without ethanol; ▵, 10°C with 7% ethanol; □, 10°C with 12% ethanol; •, 10°C with 13% ethanol. Sampling point is the average of five independent experiments corresponding to the growth of the five strains analysed in this work. The results were expressed in OD600 nm over time (h).

The five strains show the capability to survive adverse conditions. They survive and proliferate at 42 °C and also at 18 °C. Moreover, they grow in the presence of 12–13% ethanol, and 7% ethanol resulted in a larger bacterial population than seen in the absence of ethanol in the medium at 18 °C. Thus, adaptation to cold temperature can render LAB cryotolerant (Wouters et al., 2000), and if the moderate temperature of 18 °C is maintained, spontaneous MLF can be managed successfully (Henick-Kling, 1995). The ability to grow at 42 °C can be explained by the expression of the Hsp18 gene and protects the cell membrane under shock conditions (Coucheney et al., 2005). The protective effect of growth in the presence of ethanol is based on modification of the physiochemical state of the membrane. Ethanol-adapted cells displayed increased metabolic capacity, and also it is well established that the survival of microorganisms in a variety of potentially lethal conditions can be improved by pre-exposure to sublethal stress conditions of the same kind (Silveira et al., 2004).

Fingerprinting analysis

The tolerance process was also analysed using the FDD on cDNA to evaluate the different gene expression in O. oeni strains after the exposure to adverse conditions such as the growth at 18–42 °C and 7–12–13% (by volume) of ethanol in the medium. The pattern obtained showed a specific set of bands; some bands present in the stressed samples were absent in the control sample. The FDD-PCR products were analysed using a genotyping approach. The analysis at the sequencer revealed that after the ethanol and temperature treatments, gene expression changed in a particular manner, so that it can be considered a marker for the detection of LAB under stress. Table 1 shows the band sizes of the amplification products. The fingerprinting pattern obtained for each strain is reported. Some bands are present only in the stressed samples and not in the control sample. Other bands are present in the control and in the stressed samples. Furthermore, there are bands that are present only in some of the stressed samples. The selection and the analysis of the bands of major interest were carried out by comparing each lane of the stressed sample with the control and with other lanes. Bands with molecular weights that can be referred to genes previously identified in other studies were not considered (Craig et al., 1993; Guzzo et al., 2000; Tourdot-Maréchal et al., 2000; Silveira et al., 2003).

Table 1.   Genotyping analysis of the five Oenococcus oeni strains
Size*D29D30S1
123456123456123456
157005620 5600 5510  2320    2350    
2  5200520053004990            
34920  4770 4700  2170 21702170 19001900   
4452045704530460046004600 2140         1740
54080411042104130416041102110   2100211016401640    
6  39903990 39301980   19201920      
7387038703830387038703870    18201820    1590 
837003700 3700370037001780   17101780  1560 1530 
93680 3680   1610      1430 143014301430
103530  3550  1600   1600 137013701370137013701320
11348034803480348034803480  15701570 158012801280128012801280 
12328032803300333033303300152015201500   1200     
133160  3160316031501440 14401440 1400   1190 1190
1425002530 25702540 13501330133013301330 116011601160 1160 
15  2440  2440  12401240124012401100 1100 11001100
16   1800  1220122012201220    10401000 1000
171430  14001400 110011001100110011001100980980980 990 
18  1330  1330104010401050105010501000  930   
19     965            
20   694  970970970990990970 900900 900 
21600600                
22420420400  550 920920940940  850850   
SizeS11S12
123456123456
  • Band size of FDD-PCR amplification with 6FAM-M13 random primer.

  • *

    The size is expressed in base pairs, the marker used is the 1-kb ladder.

  • 1, strain growth at 30°C without ethanol; 2, strain stressed for 1 h with 7% ethanol; 3, strain stressed for 1 h with 12% ethanol; 4, strain stressed for 1 h with 13% ethanol; 5, strain stressed for 1 h at 42°C; 6, strain stressed for 1 h at 18°C.

1        5330   
2         5180  
3      52705020   5070
4  4800    4860 48004800 
5      4770 4640  4660
6  470047504700 44704410 430043004440
7  4600 46004390370038003730373037503800
8  430043304300420035003600 360035503550
9 38003990 39903990325032503290330034003330
10  3460 350034503100  315031503180
11  3020   299029902950299029902990
12 2990  299029902880288028802800 2850
13 272027002700270027002770277027602770  
142590251025002500 2500   27002700 
1524402370 230023002300     2660
16     2100  25802600  
17   20402040       
18200020002000    2557    
19  1860 1830 243024302400244024402500
201550155015101510  235023202300225023002350
21  1420 145014002120  21802100 
221300130013001300130013002020204020102020  
23    1200   1950195019901990
24 113011601180    1840 18101830
25        175017001750175
26103010101020 1020 1590 155155015701570
27   940940940   136014601460
28        1180  1180
29        279   

We obtained different bands in response to the various stresses analysed – high ethanol concentration, low and elevated temperatures. Identification of their sequences is currently underway.

The random amplification of cDNA when the same strain was stressed or not stressed confirms the activation of different genes identified in previous work (Guzzo et al., 2000); this enabled the bacteria to survive and grow under altered conditions, such as wine during MLF characterized by high alcohol content. The exposure to stress conditions requires some defence mechanism (synthesis of stress proteins) so that bacteria become more tolerant to adverse conditions (G-Alegìa et al., 2004). Further study was focused on the reproducibility of the FDD-PCR approach. Analyses were carried out using the diversity database software, which allows us to compare lane by lane the pattern of genes after stress exposure (Fig. 3). The FDD was applied in duplicate under the same conditions on the cDNA obtained from RNA of the strain grown under stress. Each graphic in Fig. 3 shows the compared lane of the fingerprinting for the strain response to one stress: lane A represents the products of one experiment, and lane B the second experiment carried out under the same conditions. The overlap of each lane displays the high reproducibility of the technique used, although some small differences in the minor bands of the patterns obtained from the same strain could be observed by visual examination. In particular, analysis performed using diversity database software gave 100% similarity for the intensity of peaks, validating the experimental procedure. FDD-PCR was used for the investigation of change in the gene expression during the process of the response to adverse conditions. Thus this study establishes and validates a method that enables, with a high reproducibility, identification of different gene expressions under altered conditions in the LAB O. oeni.

Figure 3.

 Lane comparisons. Compared lanes. Two different FDD-PCR on the cDNA of Oenococcus oeni strain cDNA compared lane by lane using diversity database software (Bio-Rad) to demonstrate the reproducibility of the investigation method.

Results from several studies on genotypic diversity among strains of O. oeni, carried out utilizing different molecular techniques such as DNA fingerprinting, restriction endonuclease analysis–pulsed-field gel electrophoresis, RAPD-PCR and ribotyping suggest that this species is also genomically homogeneous (Viti et al., 1996; Zavaleta et al., 1997; Zapparoli et al., 2000). Arbitrarily primed PCR, or RAPD, is a very useful technique for typing the genomes of bacteria (Welsh & MacLeland, 1990; Williams et al., 1990) and it has been used to characterize microorganisms at the strain level (Martinez-Murcia & Rodrìguez-Valera, 1994) as well as to generate species-specific oligonucleotide probes with known sequences (Martinez-Murcia et al., 1995). As identical RAPD patterns (for a considerable number of randomly chosen primers) are expected only from duplicates of the same strain, the use of this rapid method to evaluate strain authenticity has been suggested (Martinez-Murcia & Rodrìguez-Valera, 1994).

In this present work, the molecular method of arbitrarily primed PCR was used on cDNA to compare the variability of the stress response at molecular level during exposure to environmental stresses; this technique shows a high reliability and reproducibility (Fig. 3). Use of the random primer M13 labelled with fluorescent dye 6-FAM has already been found very suitable; in fact, the amplification products were easily visualized with a capillary sequencer, avoiding the use of radioactivity, the band intensity of PCR products was very high and smaller bands could also be visualized (Ripamonte et al., 2005). Thus, the availability of reliable methods for differentiation of O. oeni strains is indispensable for the study of the population dynamics of strains under different conditions, to monitor the fate of inoculated and autochthonous bacteria and to establish the identity of selected cultures.

The results suggest that the analysis of differential gene expression using the technique of RNA fingerprinting by randomly primed PCR is useful for a preliminary investigation of microbial behaviour involved in food technology. In conclusion, the FDD-PCR appeared to be a rapid and reproducible method to provide more information about the behaviour of malolactic strains under various conditions and could be used for the study of other LAB of industrial interest, such as dairy starters.

Ancillary