Monitoring the viable grapevine microbiome to enhance the quality of wild wines

Keywords: fermentation, metagenomics, micro-biodiversity, microbiome, microbiota, wild, wine.

The process of producing wine from grape juice was discovered to be the result of microbial organisms in the nineteenth century by Louis Pasteur. With advancements in biochemistry and microbiology, the understanding behind this complex fermentation process has been investigated extensively. As such, winemaking is one of the oldest human utilisations of microorganisms through the fine-tuned control of these complex fermentation reactions. The fermentation process is mostly completed during ‘primary fermentation’, typically by inoculated yeasts, wherein most of the alcohol content is produced by the alcoholic fermentation of glucose and fructose into ethanol. This is followed by ‘secondary fermentation’, which leads to microbial stability and advance the wine’s sensory profile (its flavour and aroma) through processes such as malolactic fermentation.¹ Two categories of microorganisms, fungi (predominantly yeasts) and bacteria, are recognised as the driving force of the primary and secondary fermentation processes, as well as the more recently recognised ‘spontaneous fermentation’ process performed by wild yeasts and bacteria, which originate from the plant’s microbiome.^2–6

The grapevine microbiome, similar to other, complex microbiomes such as the human gut microbiome, is separated into distinct compartments with each hosting its own diversity of microbes. The main subdivisions of the grapevine microbiome are the rhizosphere microbiome (the area surrounding the roots), the endosphere microbiome (the area within the plant’s tissues) and the phyllosphere microbiome (the surface of the aerial portion of the plant). Of these, the rhizosphere has been the main focus since its coinage⁷; however, current literature is beginning to recognise the importance of the phyllosphere and endosphere microbiomes because of their role in vine health as well as wine fermentation.⁸ Both phyllospheric and endospheric microorganisms inhabit the grape berry tissues, collectively known as the carposphere microbiome, and as such act as diverse pools of fungal and bacterial taxa that can be utilised as wild inoculations to be utilised during spontaneous fermentation^9,10 (Table 1).

Table 1.  Top 10 most abundant bacterial and fungal genera present in the grape carposphere microbiome identified across multiple regions globally.¹⁰

Currently, within the Australian wine industry, a movement is underway with some wineries adopting traditional practises for winemaking through the production of ‘wild wines’. Wild wines are fermented exclusively through spontaneous fermentation, using wild microorganisms derived from the grapevine microbiome to ferment wines to completion as opposed to exogenously inoculated Saccharomyces cerevisiae cultures (Fig. 1). By utilising the grapevine microbiome for wine fermentation, winemakers are capable of producing complex wines that have unique and diverse sensory traits.^6,11 Furthermore, many methods of maintaining optimal micro-biodiversity in vineyards, such as bio-dynamic and low-input approaches, align with the Australian wine industry’s promotion of ‘organic’ and pesticide-free products for health-conscious target consumers and more sustainable viticultural practises.¹²

Fig. 1.  Inoculated wines v. wild wines.

Although wild wines are becoming favourable, there is a multitude of challenges that must be overcome to successfully complete fermentation and produce wines with palatable characteristics. Two methods of producing wild wines currently exist. The first utilises vine-borne microbial isolates that can be inoculated into wine ferments to drive and support the complex fermentation reactions.^13–15 Although this method does not use specialised microorganisms such as S. cerevisiae, it also does not completely depend on the endogenous grapevine microbiota, and thus does not require the involved viticultural processes, such as monitoring, which come along with wild ferments. The second, and more-traditional method, utilises the vineyard’s own micro-biodiversity, which is cultivated on the grapevine that makes its way into the wine ferment during the crushing process. This can produce complex sensory traits that vary between seasons and regions depending on a variety of different factors (natural, human-derived and environmental).

As with most organisms, a healthy grapevine microbiome is one that holds a richness and diversity specific to its host organism,¹⁶ with the opposite typically being an indicator of a disease state within the host vine (e.g. one pathogenic taxon dominating the niche).^16–18 As the health of a plant can be dictated by that state of its microbiome, as well as the outcomes of wild fermentation, suitable micro-biodiversities must be maintained within vineyards for effective wine production. Although the occurrence of each taxon is largely host-dependent,^10,18 these microbial communities are also affected by various stressors from the environment that shape the structure of the grapevine microbiome. The frequency of each taxon is also dependent on season, geographic location, water availability, UV exposure and human intervention.¹⁶ Controlled cultivation of these specific communities is crucial to the host plant’s development; however, for wild winemaking, effective monitoring of the micro-biodiversity is also of great importance in order to encourage the development of beneficial species and hinder the growth of unwanted microorganisms such as those related to wine spoilage.

Metagenomic monitoring methods, including many of the PCR-based quantifications and DNA sequencing technologies that are typically applied to microbiome research, must be utilised to properly ensure vineyards are maintaining a healthy micro-biodiversity. The introduction of these methods would allow for incredibly efficient characterisation of the microbial richness and diversity present on a vineyard’s grapevines. Understanding which microorganisms are present in the grapevine microbiomes of a vineyard will allow grape growers to understand what involvement is necessary to cultivate or preserve a healthy microbiota. However, one of the most significant downsides to current microbiome analyses is the inability to discern which organisms were ‘alive’, at the time of sampling.^19,20 Relic DNA makes up all DNA from non-living sources, such as dead cells, compromised cells or environmental DNA (eDNA),²¹ and can be just as easily amplified and sequenced as DNA extracted from viable cells. This indiscriminate nature of PCR can provide false positives to vineyard biodiversity monitoring, which, with the wrong intervention, could lead to undesired impacts on the wild microorganisms and downstream outcomes.

Discrimination between live and dead cells is an important milestone that must be overcome to allow grape growers the ability to determine which taxa are present and associated with their host vines. To combat this, we have utilised the viability selection dye, propidium monoazide (PMA),^20–23 which is capable of covalently binding to DNA when exposed to light. The binding of the PMA dye to DNA prevents it from being amplified by PCR, and thus, cannot be observed. However, the dye cannot cross the intact membrane of live cells, allowing it to only bind to relic DNA from compromised cells or environmental sources. We tested this dye using leaf phyllosphere microbiome samples collected from Vitis vinifera ‘Syrah’ (Shiraz) grapevines and were able to gain sequencing data representative of the living phyllosphere microbiome.

Our PMA-Seq results utilised both 16S and ITS rRNA gene amplicon sequencing to observe the bacterial and fungal communities respectively. Sequencing was conducted on two sets of grape leaf swab samples; both sets were prepared identically with one being treated with PMA (representative of the viable communities) and the other left untreated (representative of the total DNA within the sample) prior to DNA extraction and amplification (B. Welsh, unpubl. data). DNA sequencing of these samples demonstrated the PMA dye’s ability to effectively remove relic DNA from grapevine microbiome samples without sacrificing the quality of the data (Fig. 2). For example, from the untreated sequencing data, we observed a high relative abundance of the fungal family Erysiphaceae, a taxon responsible for powdery mildew (a common grapevine infection); however, this abundance was lower in the samples treated with PMA. This result suggests that the observed Erysiphaceae taxa were, in fact, false positives, and the DNA responsible for those sequences likely came from non-viable sources. In a vineyard setting, if these metagenomic monitoring practices were taking place without the use of PMA viability selection, these results would have suggested a disease instance within the vines, resulting in fungicidal sprays that could affect the established structure of the grapevine microbiome. Responses to false positives could be catastrophic to the micro-biodiversity of vineyards resulting in unfavourable sensory characteristics in wild wines.

Fig. 2.  Stacked bar charts showing the top 15 families of fungi and bacteria collected from leaf phyllosphere samples. Samples were either treated with PMA dye to remove relic DNA or left untreated (representative of the total DNA).

With the growing movement towards wild wines within the wine industry, new viticultural and oenological practices will be required. To produce grapevines with sufficient micro-biodiversity for the fermentation of wine, metagenomic analytical techniques must be adopted for sufficient monitoring of the grapevine microbiome. With the inclusion of metagenomic monitoring in vineyards, considerations must be taken to ensure these microbial communities are observed with extreme scrutiny to account for potential false positives such as those demonstrated in our results. By specifically monitoring the living microbiome, the wine industry can begin to discover and adopt practices for cultivating microbial communities which produce healthier grapevines for more sustainable practices, as well as higher quality wild wines for the current growing market.

The data that support this study will be shared upon reasonable request to the corresponding author.

The authors declare that they have no conflicts of interest.

This work was supported by Wine Australia, with levies from Australia’s grape growers and winemakers and matching funds from the Australian Government.