**1. Introduction**

Malolactic fermentation (MLF), the process of biological de-acidification in winemaking, is based on the l-malic acid decarboxylation to l-lactic acid and CO2. It can occur during or after alcoholic fermentation as a result of the metabolic activity of lactic acid bacteria (LAB), which are present in wine at all stages of winemaking. Four genera were identified as the principal organisms involved in MLF: *Lactobacillus*, *Leuconostoc*, *Oenococcus*, and *Pediococcus* [1]. Wine pH is most selective, and, at a pH below 3.5, generally only strains of *Oenococcus oeni* can survive and express malolactic activity. *O. oeni* is probably the best adapted to overcome the harsh environmental wine conditions and therefore most of the commercial MLF starter cultures consist of strains from this species. Traditionally, when selected wine bacteria are used, inoculation is performed at the completion of alcoholic fermentation (AF). Since 1980, researchers have explored the possibility of inoculating wine LAB into the grape must together with the yeas<sup>t</sup> or shortly after the yeas<sup>t</sup> at the beginning of the alcoholic fermentation. Today, we have identified two di fferent timings throughout the winemaking process for inoculating wine LAB into the wine: co-inoculation with yeas<sup>t</sup> (selected wine bacteria added 24 to 72 h after yeas<sup>t</sup> addition) or sequential inoculation, when selected wine lactic acid bacteria are added at the end of, or just after the completion of, AF.

Wine pH has been increasing gradually for the last several years. Red wines with pHs over 3.5–3.6 are more and more frequent. At these pH levels, we can observe very fast growth of various indigenous microorganisms, some of which are spoilage bacteria that can cause loss of wine quality. Among these species, *Lactobacillus plantarum* strains have shown most interesting results for their capacity to induce MLF under high pH conditions, their facultative hetero-fermentative properties that avoid acetic acid production from hexose sugars and their more complex enzymatic profile and di fferent metabolism compared to *O. oeni*, which could play an important role in the modification of wine aromas.

Besides pH, the ethanol produced by the yeas<sup>t</sup> during alcoholic fermentation is another limiting factor for bacterial growth and survival in wine. Radler [2], Peynaud and Domercq [3], and Henick-Kling [4] reported an increasing inhibition above 5% (*v*/*v*). The degree of ethanol tolerance is, however, strain dependent. Specific details of alcohol sensitivity for the various species of wine LAB are contradicting. Davis et al. [5] reported strains of *Lactobacillus* and *Pediococcus* being in general more tolerant to high ethanol concentrations than *O. oeni*. In contrast to this, Henick-Kling [6] reported *O. oeni* being only partially inhibited by ethanol concentrations above 5% (*v*/*v*) and able to tolerate up to 14% (*v*/*v*) alcohol, while the growth of *L. plantarum* stops at ethanol concentrations of 5–6% (*v*/*v*). The first *L. plantarum* starter culture was introduced in the late 1980s in the United States and later also in Europe. Prahl et al. [7,8] proposed to inoculate the grape juice before alcoholic fermentation using a facultative hetero-fermentative *L. plantarum* starter culture. In EP0398957B1 [7], they disclosed a method of introducing an important freeze-dried biomass of *L. plantarum* directly into must or fruit juice to induce MLF without significant consumption of sugars present in the must or fruit juice and substantially without any production of volatile acidity. This malolactic bacteria strain had little alcohol tolerance and had been unable to survive in the fermented wine. The application of this culture had only been recommended for partial malic acid degradation in low pH white wines.

In 2004, Bou and Krieger [9] filed a patent on "Alcohol-tolerant malolactic strains for the maturation of wines with average or high pH" under the international application number PCT/FR2004/001421. The patent relates to alcohol tolerant LAB strains of the genera *Lactobacillus* and *Pediococcus* capable of initiating and carrying out a complete MLF upon direct inoculation in dried, frozen, or lyophilized state into a wine with an alcohol content of 10% (*v*/*v*) or more and an average high pH level. Dating back to 2005, a new selection of *L. plantarum* at Università Cattolica del Sacro Cuore in Italy resulted in a very e ffective *L. plantarum* culture V22, adapted to high pH wines, and showing good alcohol tolerance [10].

In 2012, Soerensen et al. [11] filed a patent application on "*Lactobacillus plantarum* cells with improved resistance to high concentrations of ethanol" (WO 2012/17200). The invention relates to cycloserine resistant mutants of lactic acid bacteria having improved resistance towards ethanol. The cycloserine resistant mutants of lactic acid bacteria had been proposed for use to induce malolactic fermentation in wine having high alcohol levels, but, as outlined above, alcohol tolerant *L. plantarum* strains can also be isolated from nature.

In 2016, a new highly concentrated *Lactobacillus plantarum* starter culture was introduced to the markets [12]. The new starter culture, called ML Prime ™, is issued from an optimized process that promotes very high malolactic activity as soon as it is added to must. Despite the good alcohol tolerance of this pure *Lactobacillus plantarum* culture with the homo-fermentative metabolism of hexose sugars, its most interesting application is in co-inoculation (inoculation 24 h after the wine yeast) without any risk of volatile acidity production during MLF even under high pH conditions. Due to the very high malolactic enzymatic activity and early inoculation shortly after the selected wine yeas<sup>t</sup> into the fermenting must, MLF is therefore completed in record time (3–7 days) during alcoholic fermentation. This way, wines can be stabilized early and protected from further contamination and thus retain their sensory integrity. More recently, this *L. plantarum* starter culture ML Prime ™ had been also proposed for a specific application in white wine to achieve a partial malolactic fermentation under lower pH conditions. In the white wine application, co- and sequential inoculation can be applied.

The majority of the selected wine lactic acid bacteria starter cultures are the pure single strain cultures, but, in 2008, the Institute for Wine Biotechnology at Stellenbosch University [13] launched a project to study on the possible application of mixed MLF starter cultures consisting of one selected *L. plantarum* and one selected *O. oeni* strain deriving from the Stellenbosch strain collection. In 2010, the first mixed LAB species culture was as Co-Inoculant NT202 and proposed for simultaneous inoculation together with a specific yeas<sup>t</sup> strain NT202. The authors claim the importance of the *L. plantarum* strain in the mix for its sensory contribution and of the *O. oeni* strain for its malolactic enzyme activity. Certain strains within the *L. plantarum* species have been found to possess even more diverse enzymatic activities, which could contribute to the wine aroma profile than *O. oeni*.

#### **2. Biodiversity of Lactic Acid Bacteria in Wine**

Winemaking is a microbiological process involving a very complex system and it involves numerous microbial transformations comprising a complex succession of various yeas<sup>t</sup> and bacterial species. Malolactic fermentation (MLF) can occur during or after alcoholic fermentation and is carried out by one or more species of lactic acid bacteria (Figure 1).

**Figure 1.** Population kinetics of wine lactic acid bacteria from the vineyard to the wine (Modified from Patrick Lucas, 2016, International ML School Lallemand Toulouse).

Di fferent LAB enter into grape juice/wine from the surface of grape berries, stems, leaves, and soil and winery equipment. In the vineyard, LAB species diversity associated with grape surfaces is rather limited, mainly due to their nutritional requirements [14,15]. The population density of LAB is very limited, especially in comparison to the indigenous yeas<sup>t</sup> population found on grapes [16]. *Pediococcus*, *Leuconostoc*, and *Lactobacillus* species occur on grapes more frequently than *O. oeni* [17]. In addition to grape surfaces, bacterial strains can also be isolated from the cellar environment, such as fermentation tanks and barrels and poorly sanitized winery equipment, such as pipes and valves [17,18]. Shortly after crushing and the start of AF, the LAB population in the grape must generally range from 10<sup>3</sup> to 10<sup>4</sup> cfu/mL (colony forming units per milliliter), and the LAB species largely belonging to the species of *Lactobacillus* and *Pediococcus* disappear progressively during the AF [19]. The decrease could be attributed to increased ethanol concentrations, high SO2 concentrations, initial low pH, low temperatures, the nutrient depletion, and/or competitive interactions with the yeas<sup>t</sup> culture [16,20]. During spontaneous MLF, *O. oeni* is the major bacterial species found, however, several species can be occasionally detected, mainly *Lactobacillus*, *Pediococcus,* and *Leuconostoc* [1,19]. In some of the warmer wine growing regions, *L. plantarum* is more frequently isolated from spontaneous malolactic fermentations [13,21–23]. Lerm et al. reported three *O. oeni* and three *L. plantarum strains* from South Africa wine isolates for use as MLF starter cultures. Bergeral et al. [21] had studied the properties of *Lactobacillus plantarum* strains isolated from grape must fermentation Apulian wines in order to select suitable starter for MLF, and Valdés La Hens et al. [22] reported the Prevalence of *L. plantarum* and *O. oeni* during spontaneous fermentation in Patagonian red wines. More recently, <sup>L</sup>ópes-Seijas et al. [23] evaluated malolactic bacteria associated with wine from the Albariño variety. Di fferent to what has been described from other wine growing regions, the predominant species in the region of Val do Salnés in Spain were *L. hilgardii*, *L.* paracasei, and *L.* plantarum. Nevertheless *O. oeni* is most frequently the predominant species at the later stages of vinification (Figure 1), since it is best adapted to the limiting conditions encountered in wine. Over centuries of selective pressure, *O. oeni* has acquired and perfected various adaptive strategies that enable it to outcompete with other wine lactic acid bacteria during the later stages of vinification and thus to dominate in wine [15]. It proliferates in wine and cider during or after the yeast-driven alcoholic fermentation and reaches population levels above 10<sup>6</sup> cells/mL, thus becoming sometimes the only detectable bacterial species [24,25].

Wine pH is most selective, and, at a pH below 3.5, generally only strains of *O. oeni* can survive and express malolactic activity, while, in wines with a pH above 3.5, some *Lactobacillus* species have also shown a good ability to conduct MLF. Generally, the most frequent lactobacilli isolated from wine belongs to *Lb. plantarum*, *Lb. brevis*, *Lb. buchneri*, *Lb. hilgardii,* and *Lb. fructivorans*, although their occurrence and that of other species (i.e., *Lb. fermentum*, *Lb. kunkeei*, *Lb. mali*, *Lb. vini*) can be found depending on the grape varieties and typologies of wines [1]. Among them, *Lb. plantarum* is certainly the most important in wine because it is found frequently on grapes and in wine and is often involved in spontaneous MLF under high pH conditions. This versatile bacterium tolerates ethanol up to 14% (*v*/*v*) and can have similar SO2 tolerance like *O. oeni*. Moreover, *Lb. plantarum* has a more diverse array of enzymes and can potentially exert positive e ffects on organoleptic properties of wine [1]. Some selected *L. plantarum* strains have shown interesting results for their capacity to induce MLF under high pH conditions, and, unlike *Oenococcus oeni*, *L. plantarum* has a facultative hetero-fermentative metabolism that prevents acetic acid production from hexose sugars. Due to these characteristics, selected strains of *Lb. plantarum* are currently being commercialized to induce MLF in wine [1,26].

*Pediococcus damnosus* is the other species well represented in the wine environment. It is often found after alcoholic fermentation in wines with rather high pH, along with *Lactobacillus* sp. and *O. oeni*. As it has been identified in most ropy wines, its presence is considered undesirable. In reality, only certain strains of *P. damnosus* are responsible for this spoilage and they are easily identified through polymerase chain reaction (PCR). Little research has been published on the possibility of using these organisms as wine LAB starter cultures. A study in which indigenous strains of *P. damnosus* dominate a starter culture of *O. oeni* and conduct the MLF shows that it is very capable of surviving in wine [27].

#### **3. Selected Wine Lactic Acid Bacteria Starter Cultures and Wine Challenging Factors**

Grape juice and (especially) grape wine contain a challenging matrix, with sugar, ethanol, organic acids, amino acids, fatty acids, other metabolites deriving from the yeas<sup>t</sup> metabolism during alcoholic fermentation, phenol contents, pH, and SO2 determining the growth of wine microorganisms. Various papers reported many factors that influence the occurrence of LAB and MLF in wines. Henick-Kling [28] and Wibowo et al. [29] listed, besides oxygen and CO2, carbohydrates, amino acids, vitamins and minerals, organic acid content, the alcohol level, pH, and SO2 level.. The interrelationships between LAB and wine yeas<sup>t</sup> [30] or other wine microorganisms and the method of vinification have been reported to be the most influential factors to affect LAB growth. The wine pH is one of the most important factors that limits LAB growth and MLF in wine [2,29,31] and determines the type of LAB which will be present. Ideally, for table wines, the pH should be between 3.1 and 3.6 [32], but due to global warming wine pH has increased in recent years in almost all wine regions.

#### *3.1. Well-Known Factors that A*ff*ect Malolactic Fermentation and Bacteria Vitality*

The best understood factors that govern successful MLF are SO2, pH, alcohol, and temperature.
