**1. Introduction**

One of the main characteristics of yeas<sup>t</sup> affecting their oenological use is their capacity to ferment sugars. Non-*Saccharomyces* yeasts have traditionally been considered bad fermenters. For that reason, selected strains of *S. cerevisiae* have been used in the oenological industry to ensure that complete fermentation occurs [1,2]. Recently, however, wineries have been encouraged to apply new, non*-Saccharomyces* species in winemaking processes to provide distinguishable flavours within wines [3–6]. Non-*Saccharomyces* species have been used to produce different aromas and flavours, compared with *Saccharomyces* strains [7,8]. Therefore, many efforts have been made to identify non-conventional yeas<sup>t</sup> strains for oenological purposes [4,8,9].

The selection of oenological yeasts is commonly accomplished by identifying species from raw material. Spontaneously fermented grape musts are the niche that is most commonly used to identify novel strains that are both capable of fermenting sugars and confer desirable flavours to wines [10,11]. *Hanseniaspora* is the most abundant genus on grapes and grape juices. Studies have shown that up to 75% of the yeas<sup>t</sup> population during the early stages of fermentation is made up of *Hanseniaspora* species [12,13]. After the first 48–72 h of spontaneous fermentation, the percentage of

> 19

*Hanseniaspora* species present decreases and the *S. cerevisiae* strains correspondingly increase. However, some *Hanseniaspora* species have been detected throughout the fermentative process [14].

Researchers have maintained that observed changes in yeas<sup>t</sup> populations during fermentation occur, at least in part, because *Hanseniaspora* species are sensitive to ethanol [15]. *S. cerevisiae* is able to produce high quantities of ethanol rapidly. Therefore, this species dominates the fermentation process until sugar is completely depleted. Recent studies have shown the e ffects of antimicrobial peptides secreted by *S. cerevisiae* throughout the fermentation [16,17] inhibiting the growth of non-*Saccharomyces* yeasts. Therefore, the reputation of *Hanseniaspora* species as poor fermenters may be due to the presence of other inhibitors and not directly related to their reduced capacity to ferment sugars.

Not all the *Hanseniaspora* strains have the same properties. Species of the genus produce di fferent secondary metabolites and exhibit di fferent fermentative behaviours. In fact, di fferences could even be detected between strains belonging to the same species [18,19]. *H. vineae* is an epiphyte yeas<sup>t</sup> that is not easily isolated from fruit, a feature it shares with all the *S. cerevisiae* strains [20]. *H. vineae* can be isolated from samples after one or two days of spontaneous fermentation in wines and other fruit beverages such as cider [19]. This highlights the distinct behaviour of the species, compared with the majority of *Hanseniaspora* apiculate yeasts, which are commonly isolated from the skin of grapes or grapevine soil. The ability of strains identified as *H. vineae* to complete grape juice fermentation has been demonstrated via single inoculation [7]. Moreover, selected strains of *H. vineae* contribute positively to wine aromas by providing floral and honey notes, even when sequential inoculation with *S. cerevisiae* strains was performed [21]. Our assessment of *H. vineae* showed that levels of phenylpropanoid flavour compounds synthetized from grape must were elevated compared with other yeasts. In *H. vineae*, the presence of metabolic pathways that actively transform aromatic amino acids explains the elevated phenyl acetate ester and benzenoid derived compounds synthesis compared to other yeasts and these flavour compounds provide fruity and flowery aromas [21–23].

Although several phenotypic studies have been carried out throughout wine fermentation using non-*Saccharomyces* species, there is a lack of information regarding the genetic basis of observed characteristics in non-*Saccharomyces* strains [8]. Due to the development of next generation sequencing, genomes of *Hanseniaspora* species from wine have been recently sequenced [24–27]. Further work will be needed to determine which genes are responsible for each function. In previous studies, the aromatic profile of *H. vineae* was correlated with genomics and transcriptomics data [22]. However, genes involved in glycolysis and fermentative behaviour in the species remain unknown. In *S. cerevisiae*, genes necessary for fermentation have been reported using mutant analysis. All of these, were grouped in a "fermentome" [28]. The genome of *H. guilliermondii* has been recently analysed [27] and the presence and absence of genes involved in the glycolytic and fermentative pathways compared with *S. cerevisiae* and other *Hanseniaspora* species were reported. Moreover, the *H. uvarum* glycolytic pathway has been assessed in a study that revealed the catalytic potentials of enzymes involved in the route [29]. The authors showed that the main glycolytic enzyme of *H. uvarum*, pyruvate kinase, had a 15-fold lower enzymatic activity than that of the *S. cerevisiae* enzyme.

The aim of this work is to establish the di fferences and similarities between *H. vineae*, other *Hanseniaspora* species, and *S. cerevisiae* regarding glycolytic and fermentative behaviour. In the present study, a comparative analysis of the fermentative capacity of *H. vineae* was performed using genetic and transcriptomic data. Characterization of the glycolytic and fermentative potential of *H. vineae* will enhance our understanding about the mechanisms and the regulation of the fermentative process in a non-*Saccharomyces* yeast. *Hanseniaspora* genus studies might help reveal new signs of *S. cerevisiae* domestication mechanisms for wine production.

#### **2. Materials and Methods**

#### *2.1. Yeast Strains*

Yeast strains used for this study are listed in Table 1.



#### *2.2. Fermentation in Natural Grape Must*

Chardonnay grape must containing 300 mg N/L and 200 g/<sup>L</sup> of sugars at pH 3.5 was treated with 200 mg/<sup>L</sup> dimethyldicarbonate to prevent microorganism growth. Pre-cultures of *H. vineae* T02/19AF, *H. vineae* T02/05AF, *H. uvarum* AWRI1280, *H. osmophila* AWRI3579, and *S. cerevisiae* ALG804 were isolated from the Chardonnay grape must and incubated at 25 ◦C for 12 h in a rotary shaker at 150 rpm. Then, 125-mL Erlenmeyer flasks closed with cotton plugs used to simulate microaerobic conditions were inoculated with 75 mL of must containing 1 × 10<sup>5</sup> cells/mL. Static batch fermentations were conducted at 20 ◦C to simulate winemaking conditions.

#### *2.3. Growth Kinetics in Di*ff*erent Types of Media*

Six types of growth media were prepared using yeas<sup>t</sup> nitrogen base (YNB) (Difco, Detroit, MI, USA) as a sole nitrogen source (6.7 g/L). Media were supplemented with the following carbon sources: Glucose, fructose, sucrose, xylose, glycerol, and maltose (2% w/w). YNB that lacked a carbon source was used as a negative control.

Chardonnay must used in the fermentation analysis was also used to measure the growth kinetics of yeas<sup>t</sup> strains tested. Moreover, synthetic media that mimicked must fermentations at pH adjusted to 3.5 and ethanol concentrations of either 5% or 10% were used (20 g/<sup>L</sup> glucose, 4 g/<sup>L</sup> tartaric acid; 0.134 g/<sup>L</sup> sodium acetate; 5 g/<sup>L</sup> glycerol; and 1.7 g/<sup>L</sup> YNB) (v/v).

Pre-cultures of *H. vineae* T02/19AF, *H. vineae* T02/05AF, *H. uvarum* AWRI1280, *H. osmophila* AWRI3579, and *S. cerevisiae* ALG804 were prepared in yeas<sup>t</sup> extract peptone dextrose (YPD) media (1% yeas<sup>t</sup> extract and 2% peptone, 2% glucose) via incubation for 12 h in a rotary shaker at 150 rpm and 25 ◦C. These pre-cultures were used to inoculate fermentations carried out in microtitler plates at a final volume of 250 μL. Inoculates producing 1 × 10<sup>5</sup> cells/mL in media were used for all strains and treatments. All conditions were tested in triplicate. Absorbance at 620 nm was measured at 30-min intervals for 48 h at 25 ◦C using an automatic plate reader (Tecan, Männedorf, Switzerland) and data were acquired with the Magellan software for further statistical analyses.

#### *2.4. Fermentation Ability in Di*ff*erent Carbon Sources*

The carbohydrate fermentation capacity was tested using Durham tubes immersed in media to detect gas production. Each type of medium tested was inoculated to produce a final concentration of 1 × 10<sup>6</sup> cells/mL in a final volume of 8 mL performed in triplicate. Results were visually assessed after a 48 and 96 h static incubation period at 28 ◦C.

## *2.5. Genomic Analysis*

Genomic DNA was obtained from *H. vineae* cultures grown in a YPD medium at 30 ◦C using the Wizard Genomic DNA Purification Kit (Promega, NY, USA), according to the manufacturer's instructions.

The Illumina Genome Analyzer Iix platform in paired end mode was used to perform genomic sequencing as described previously [22]. Gene prediction was carried out using Augustus [30] trained with *S. cerevisiae* gene models. Peptide predictions were then annotated using BLASTp (cuto ff for e-value <sup>1</sup>−10) against *S. cerevisiae* proteins, obtained from the *Saccharomyces* Genome Database [31].

A dendrogram was constructed using the sequences of nine genes encoding components of pathways related to glycolysis and fermentation from the *Hanseniaspora* species and *S. cerevisiae*. The genes assessed were *CDC19*, *FBA1*, *PGI1*, *PFK1*, *PFK2*, *HXK2*, *ENO1*, *PGK1,* and *PDC1*. *Schyzosaccharomyces pombe* was used as an external group. Neighbour joining and Kimura 2-parameter methods were carried out using the MEGA version 4 software [32,33].

## *2.6. Transcriptomic Analysis*

Fermentations were performed in triplicate using chemically defined grape (CDG) must with a composition similar to that of natural grape juice, but devoid of grape precursors. Components of CDG must were defined as described in Carrau et al. [34], with modifications. Briefly, glucose and fructose were added in equimolar concentrations until a total sugar concentration of 200 g/<sup>L</sup> was reached. Vitamins and salts were added as previously described [35]. Yeast available nitrogen (YAN) content was adjusted to 100 mg N/L. Of this total, 50 mg N/L corresponded to amino acids and 50 mg N/L corresponded to diammonium phosphate (DAP) supplementation, as described previously [35]. The pH of the media was adjusted to 3.5 using HCl and a final concentration of 10 mg/<sup>L</sup> ergosterol was the only lipid provided.

Pre-cultures of *H. vineae* T02/19AF were prepared in a CDG medium and incubated 12 h in a rotary shaker at 150 rpm and 25 ◦C. The pre-cultures were subsequently used to inoculate fermentation reactions carried out in 250 mL Erlenmeyer flasks that were closed with cotton plugs to simulate microaerobic conditions. For all strains, fermentations were performed using 125 mL CDG and an inoculum to produce 1 × 10<sup>5</sup> cells/mL in the final medium. Static batch fermentations were conducted at 20 ◦C to simulate winemaking conditions.

Wine samples for transcriptomic analyses were taken during the fermentation process at day 1 (exponential growth), day 4 (end of exponential phase), and day 10 (stationary phase of fermentation). For transcriptomic studies, total RNA obtained from *H. vineae* T02/19AF isolated from three replicates sampled from three di fferent fermentation stages (days 1, 4, and 10) were analysed independently. The nine samples were paired-end sequenced using Illumina MySeq. Trinity software was used to assemble raw reads from transcriptomic analyses and further statistical analyses were performed as specified by Giorello et al. [22].

## *2.7. Statistical Analysis*

All the treatments were performed in triplicate and the statistical error was calculated as the standard deviation of all data analysed. To compare growth and fermentation kinetics, variance comparison was performed by the ANOVA test carried out with STATISTICA 7.0 software. Di fferences in the mean absorbance or weight loss were evaluated using the Tukey test.

#### **3. Results and Discussion**

#### *3.1. Fermentative Capacity of H. vineae in Di*ff*erent Media*

*Hanseniaspora* species used a limited number of carbon sources, which may have been related to the reduced competitiveness of the species throughout fermentations [27]. Regarding growth in di fferent carbon sources (Figure 1A), growth of all the *Hanseniaspora* strains tested on both glucose and fructose had kinetics similar to that of *S. cerevisiae* ALG804. The media supplemented with sucrose was fermented by *S. cerevisiae* in a similar manner as that of media containing simple hexose. *H. uvarum* AWRI1280, however, did not grow on media containing sucrose. *H. vineae* T02/05AF, *H. vineae* T02/19AF, and *H. osmophila* AWRI3579 were able to grow on and ferment sucrose to an extent. Invertase gene (*SUC2*) is present in the genome of *H. vineae. SUC2* is highly expressed on day 4 of fermentation reactions, but not day 1 or 10. However, other invertase homologs were not observed in

the genomes of any other *Hanseniaspora* species except *H. osmophila* [25]. Recently, Steenwyk et al. [36] grouped *H. vineae* and *H. osmophila* within the slower-evolving linage of *Hanseniaspora*. In this branch, the *SUC2* gene is present. This is different in species of the fast evolving linage including *H. uvarum*, *H. opuntiae, H. valbyensis,* and *H. guilliermondii,* which might have lost the gene as a result of rapid mutation rates [36]. The same fact was detected with another key gene that show *Saccharomyces* wine yeas<sup>t</sup> adaptations. Increased sulfite tolerance conferred by *SSU1* (Table 2) is present in *H. vineae* and *H. osmophila* and it is absent in the other *Hanseniaspora* species. The presence of *SUC2* and *SSU1* genes are indicators of adaptations to alcoholic fermentation in yeas<sup>t</sup> [37].

Glycerol was not used as a unique carbon source for *H. vineae* in accordance with data reported by Albertin et al. [38]. However, Hv*GUT1* and Hv*GUT2* genes were present in the genomes of both *H. vineae* strains analysed. In addition, xylose was not used by the *H. vineae* strains as expected. A finding that was likely due to the lack of enzymes needed to carry out the xylose conversion. The group of genes were also determined to be absent in *H. guilliermondii*, *H. uvarum,* and *H. opuntiae* [27]. However, *H. vineae* T02/05AF and T02/19AF have the ability to grow weakly when maltose is provided as a sole carbon source, despite the fact that they were not able to ferment the sugar. The same behaviour was also observed for *H. jakobsenii* [36].

**Figure 1.** Capacity of *Hanseniaspora vineae* and *Saccharomyces cerevisiae* to grow and ferment under varied conditions. (**A**) Growth of *H. vineae* and *S. cerevisiae* (G) and the capacity of the species ferment (F) when six different carbon sources (2% w/w) were provided. Yeast nitrogen base (YNB) that lacked a carbohydrate was used as a negative control. Black filled circles indicate that full growth and fermentation were observed, grey circles indicate the moderate capacity of species to grow and ferment, and white circles indicate that the species was not able to grow or ferment. (**B**) Growth kinetics of *Hanseniaspora spp.* and *S. cerevisiae* on the Chardonnay grape juice measure as increased absorbance over a period of 48 h. (**C**) Fermentation kinetics of the three strains in the Chardonnay grape juice after 12 days are shown. Growth and fermentation experiments were performed using independent triplicate samples and error bars express standard deviation.


**Table 2.** Genes involved in sugar transport, glycolysis, and alcoholic fermentation from *S. cerevisiae* and *H. vineae*. Gene copy numbers are detailed in brackets.

As expected, *Saccharomyces* was able to grow and ferment sugars faster than *Hanseniaspora* species and significant differences between the species occurred after 16 h. The growth kinetics of the three *Hanseniaspora* strains tested were similar on grape must (Figure 1B), however fermentation kinetics of *H. vineae* and *H. osmophila* revealed that these species consume sugars significantly faster than *H. uvarum* (Figure 1C).
