Wine grapes are considered the most important fruit crop in the world in terms of production and economic importance (Alston and Sambucci, 2019). It has been reported that there are nearly 8 million hectares of vineyards worldwide and the global annual production have reached approximately 90 million tons (http://faostat.fao.org). In the United States alone, which ranks fourth in the volume of wine production behind Italy, France, and Spain (Stevenson, 2005), wine, grapes, and grape products contribute $276 billion to the economy in 2022 (https://wineamerica.org/economic-impact-study/2022-american-wine-industry-methodology/). The importance of grape cultivation for wine production, however, goes beyond its bare contribution to the economy. Wine consumption has moved from a source of nutrition to a cultural phenomenon with a large tourist industry associated with it. For this reason, the wine industry has helped fix local populations in rural areas by diversifying the job markets in such regions (https://wineamerica.org/economic-impact-study/2022-american-wine-industry-methodology/). The majority of cultivated grapes belong to Vitis vinifera subsp. vinifera; but the cultivation of other Vitis sensu stricto species, including hybrids, and the related subgenus Muscadinia are also common in regions where the climate and/or disease pressure are not suitable for V. vinifera (Hickey et al., 2019), despite climate change. Early breeding efforts done in the Southern United States using American grape species, led to the production of many modern cultivars such as the muscadine and bunch grapes, that can still be found today (Stafne et al., 2015), and more recently to the modern rootstocks being used worldwide (Ollat et al., 2016; Santos et al., 2020). However, many bunch grape hybrids derived from V. vinifera, and other non-resistant species are susceptible to Pierce’s disease (PD). Thus, the breeding focus has mainly been on producing resistant cultivars, such as those released by the University of California-Davis (https://ucdavis.app.box.com/s/dte06een1orc7uzfucccwdaqwepqp90y).

Climate change is expected to severely affect the major viticultural regions of the world by reducing the areas where most grapevine cultivars can be cultivated economically, due to an increase in abiotic stress pressure (Diffenbaugh et al., 2011), and in the incidence of pests and diseases (Gullino et al., 2018). The long domestication and breeding history of V. vinifera in particular, for wine and fresh and dried fruit consumption has led to desirable traits such as berry color, sugar content, and berry size (Aradhya et al., 2003; Myles et al., 2011; Zhou et al., 2017). Traditional breeding attempts to cross wild non-vinifera species with V. vinifera has generated some resistant/tolerant genotypes (Morales-Cruz et al., 2021; Ruiz-García et al., 2021), although these genotypes can be compromised by negatively perceived flavors, prominently in wine production (Liu et al., 2015; Yang et al., 2016). While this is only true in F1 crosses between vinifera and wild species, due to the protracted nature of traditional grapevine breeding programs, the future success of the wine industry will require the utilization of new breeding technologies for the development of novel varieties better suited to the climatic conditions predicted under the scenario of climate change (Töpfer and Trapp, 2022).

Plants have acquired many adaptation strategies, activated and controlled by changes in gene expression and nuclear organization (Budak et al., 2015) to cope with ever-changing environmental conditions. Progress in plant molecular biology has enable the identification of major stress response pathways, leading into a deeper understanding of the plant responses that constitute such strategies (Hirayama and Shinozaki, 2010). The availability of the complete grapevine genome sequence has allowed the identification and characterization of various stress-inducible genes, cis-regulatory elements and transcription factors (Jaillon et al., 2007). More recent studies have shown that epigenetic mechanisms, some with the potential to be inherited, play an important role in plant response to environmental stress (Miryeganeh, 2021). Although the current knowledge on the role of epigenetic regulation in response to the environment in the grapevine is still limited, the demonstration of the involvement of epigenetic mechanisms in model plants has led to an increased interest in their role in crop resilience to environmental stress (Varotto et al., 2020).

Here we summarize the current knowledge on, environmental factors that affect grape and wine qualities, transcriptomic approaches that have been utilized to study the effect of environmental factors on grapevine, and finally recent studies focusing on epigenetic mechanisms, particularly those involved in plant response to environmental changes, which have led to proposing epigenetic breeding as a new tool for the generation of climate resilient grapevines.

Fruit and wine quality are determined by the interaction between the cultivar(s) (including the interaction between rootstock and scion), the local environmental conditions (climate, topography, soil, etc.), and the viticultural and enological practices implemented to grow the grapes and produce the wine (Van Leeuwen et al., 2004). Such interaction has been traditionally termed terroir (Seguin, 1986) (Figure 1).

Stress can be classified into biotic and abiotic. Biotic stresses are caused by biological agents such as fungi, bacteria, viruses, and insects, whereas abiotic stresses are caused by physical environmental factors. Common abiotic factors unfavorable for plant growth and crop yield include drought, saline soils and irrigation, heat, and cold. Worldwide, extensive agricultural losses result from heat stress, often in combination with drought (Vogel et al., 2019). It is expected that the effects of combined drought and heat stress will become more severe as the climate continues to warm (Zhao et al., 2017; Raza et al., 2019), as it is predicted that an increase in global temperature of 1.5°C will cause more extremely hot days on land, and an increase in the intensity and frequency of drought and precipitation deficits (IPCC, 2018).

Agriculture is highly dependent on climatic conditions. Climate determines the ability to successfully grow a particular variety and can greatly affect the value of the fruit produced (Jones and Davis, 2000; Jones, 2006; Bai et al., 2022). Grape production is vulnerable to environmental stress as the environmental conditions occurring during one growing season contribute to the quality and yield of the next vintage (Mullins et al., 1992; Edwards and Clingeleffer, 2013; Martínez-Lüscher and Kurtural, 2021). Viticulture is commonly practiced in regions with a Mediterranean climate (Cs climate according to the updated Koppen-Geiger climate classification (Peel et al., 2007)), where the growing season is characterized by low rainfall, the majority occurring in winter, and by high air temperature and evaporative demand (Fraga et al., 2012). In addition to the coastal regions of the Mediterranean Sea, this includes, the West coast of the Iberia Peninsula, the Pacific coast of Chile and the United States, Cape Town region in South Africa, and portions of the West and South Coast of Australia (Peel et al., 2007). Recent studies have shown that temperature rise is highly correlated with an earlier onset of many growth stages in the grapevine (Alikadic et al., 2019). It has been proposed that an increase in ambient temperatures will constitute the primary cause of water shortages for viticulture due to increased evaporative demand (Schultz, 2010), and may eliminate production in many areas (White et al., 2006; Diffenbaugh et al., 2011). On the other hand, temperature rise is not without benefits, Bunting et al. (2021) showed that changes in climate in Michigan (MI), United States has helped the state to overcome the challenge for grape cultivation due to low growing season temperature, short growing seasons, and excessive precipitation. Similarly, Cabré et al. (2016) suggested that Argentina has a great potential for expansion into new suitable vineyards due to climate change. Climate change is also expected to affect plant-pathogen interactions causing severe damage to grapevine and leading to extensive yield and quality losses (Yu et al., 2012; Gullino et al., 2018). The maintenance of stable and high-quality supplies of grapes and derived products will demand the implementation of measures such as relocation of vineyards to northern zones or higher altitude areas with lower average temperature (White et al., 2006) or the development of novel and faster breeding programs.

Studying the regulation of gene expression can provide a deeper understanding of the molecular regulation of the physiological and metabolic mechanisms used by grapevine to respond to various stresses such as elevated temperatures (heat) or drought. Earlier efforts included the use of Expressed Sequence Tags (ESTs), which resulted in the development of a microarray containing a set of 3,200 Unigenes from V. vinifera to study grape development (Terrier et al., 2001; Terrier et al., 2005). The number of unigenes present on the microarray rapidly increased with newer technologies such as the Operon (Camps et al., 2010) or Affymetrix (Deluc et al., 2009) grape arrays. The complete sequence of the grapevine genome became available after the sequencing and assembly of the PN40024 line (Jaillon et al., 2007). With that being available, NimbleGen microarrays were utilized to study grape transcriptome (Pastore et al., 2017). With the advance of technology, full coverage of the grapevine transcriptome was made possible by next-generation sequencing, named RNA-sequencing (Zenoni et al., 2010). Since then, both genome wide-microarrays and RNA-sequencing have been widely used to characterize the response of grapevine to various stress. Some examples include heat (i.e., Rienth et al., 2016), drought (i.e., Berdeja et al., 2015), and UV-B stress (Du Plessis et al., 2017). The high-throughput sequencing technology has been proven useful in revealing potential key stress response genes, which could be highly beneficial for breeding new grape cultivars that can better adapt to the changing environment. Examples of the key genes that have been characterized as playing a role in grapevine stress response, include leafy cotyledon1-like (LEC1) and somatic embryogenesis receptor kinase (SERK) (VvL1L and VvSERK, respectively in grapevine), which are key regulators of grapevine development and stress response (Maillot et al., 2009). Abscisic acid-insensitive 3 (ABI3), a gene that is involved in abscisic acid (ABA) signaling and drought response (Mittal et al., 2014; Rattanakon et al., 2016). Various calcium-dependent protein kinases (CDPKs), such as VaCPK20 and VaCPK29 identified from V. amurensis have been shown to involved in drought and cold tolerance, and to heat and osmotic stresses respectively, when being overexpressed in transgenic grape cell cultures and in Arabidopsis thaliana (Dubrovina et al., 2015; Dubrovina et al., 2017). Several dehydration responsive protein associated genes and transcription factors regulated by ABA, including dehydration responsive element-binding protein1a (DREB1A), have been identified as regulators of stress-responsive genes against drought tolerance (Cardone et al., 2019), while apoptosis related-proteins genes were shown to be involved in the regulation of programmed cell death and defense against biotic stress (Repka, 2006). The exact role and mechanism of action of these genes can vary depending on the type of stress and the grapevine genotypes being studied and that they are often a part of a much more complex stress signaling pathways. Additionally, Zha et al. (2020) used transcriptomic analysis to study grapevine response to heat stress and identified two important genes central to grapevine’s response to heat stress, heat shock factor a2 and a7 (VvHSFA2 and VvHSFA7, respectively). Cochetel et al. (2020) showed that more drought tolerant wild genotypes are more responsive transcriptionally in terms of ABA signaling and biosynthesis than less drought tolerant ones. The authors also identified core genes to drought stress as well as gene clusters and sub-networks that are associated with drought tolerance in grapevine.

The experimental designs for transcriptomic analyses are not without limitations. Rienth et al. (2014) showed that the transcriptome of grapevine plants under heat stress can vary drastically depending on the time of the day the stress is being applied. The results from this study suggested that future grapevine transcriptomic analyses should rely standardized experimental designs. Additionally, the quantitation of the applied stress factor and the physiological impact on the plant should be measured carefully (Berdeja et al., 2015). Moreover, a large body of research has suggested the need to go beyond classical differentially expressed gene (DEG) analysis, more detailed tools and analyses such as weighted gene co-expression network (WGCNA) and cluster analysis. Those will provide more in-depth knowledge on stress response by revealing co-regulated gene modules and potential master switch/hub genes that might be key for abiotic stress responses in plants (Palumbo et al., 2014; Hopper et al., 2016; Cochetel et al., 2017). Moreover, although stress conditions in the natural environment often occur in combination (e.g., heat and drought stress tend to occur simultaneously in grapevine cultivating regions), a majority of grapevine transcriptomic studies deal with only one abiotic stress factor, where such a factor is often applied in controlled or semi-controlled conditions. Therefore, it has been suggested that transcriptomic studies should integrate stress combinations in their experimental design (Gomès et al., 2021). We integrated these recommendations in our most recent global transcriptomic and gene co-expression network analysis to reveal core genes central to grapevine response to combined heat and drought stress (Tan et al., 2023). Interestingly, this work also identified that epigenetic chromatin modifications may play an important role in grapevine responses to combined drought and heat stress through the establishment of an epigenetic memory of stress.

Plants have developed various mechanisms to adapt to daily environmental conditions, and the regulation of gene expression through both transcriptional and post-transcriptional regulation is particularly important to their survival. Among those strategies are a suit of molecular mechanisms studied under the umbrella term of epigenetics. Waddington (1942) first proposed the term epigenetics, referring to the study of the interactions between genes and the environment. The current definition of the term refers to potentially heritable changes in gene function without changes to their underlying DNA sequence (Wu and Morris, 2001) that are usually mediated by three main types of changes: DNA methylation, histone post-translational modifications (PTMs), and the expression of certain small RNAs (sRNAs) (Agarwal et al., 2020).

Fortes and Gallusci (2017) proposed grapevine as a model to study epigenomics in perennial woody plants of agricultural importance due to its characteristics, including a genome and methylome more alike to other crops than those of the most widely used model plant, Arabidopsis thaliana (Lee and Kim, 2014), in addition to a set of important agronomic characteristics which have been previously associated with epigenetic mechanisms: (1) grapevine is considered one of the models for non-climacteric fruit development (Fortes et al., 2015); (2) the usage of grafting and vegetative propagation (Lewsey et al., 2016); (3) vine age and vineyard location (Grigg, 2017; Xie et al., 2017; Grigg et al., 2018) have been traditionally associated with fruit production and quality; and (4) grapevine flower development has been shown to be programmed and affected by the environmental conditions one year in advance (Guilpart et al., 2014).

Although multiple studies have shown that the main driver of DNA methylation variability in grapevine is the genotype (Dal Santo et al., 2018; Varela et al., 2021), recent studies have suggested that the growing environment can have a significant effect on the methylome, and that such environmentally induced epigenetic changes could be the molecular basis of terroir in grapevine. In Xie et al., 2017 showed that the main contributor to differences in DNA between 22 V. vinifera cv. Shiraz vineyards in six sub-regions of South Australia was geographic distance (with 9% of the identified differentially methylated genes being associated with response to environmental stimulus), followed by vineyard management and altitude. A later study comparing DNA methylation patterns in two V. vinifera cultivars (i.e., Merlot and Pinot Noir) grown in contrasting climatic regions showed that a significant amount of DNA methylation variability (roughly 80% and 71% of Merlot and Pinot Noir, respectively) was associated to geographical location (Baránková et al., 2021).

The regulation of the biosynthesis of metabolites and the accumulation of phenolic compounds in grapevine is also found to be associated with epigenetic mechanisms. In V. amurensis cell cultures treated with 5-Azacytidine, a demethylating agent, the methylation level of a stilbene synthase gene was significantly reduced, while the gene expression of the same gene and the synthesis of resveratrol was significantly increased, which led to high level of resveratrol compared to the control cell culture, suggesting that the DNA methylation may be involved in the control of resveratrol biosynthesis during in vitro culture (Kiselev et al., 2013). Although these results might not be directly comparable to what happens in vivo, DNA methylation has also been reported to have a role in the production of anthocyanins, a pathway that competes with the biosynthesis of resveratrol (He et al., 2010), during berry maturation (Jia et al., 2020). In addition, UV-B was associated with flavonol accumulation in V. vinifera cv. Malbec berries and hydroxycinnamic acids in early fruit shoots, and those changes can be DNA methylation-dependent (Marfil et al., 2019). Interestingly, in a study that analyzed ten different grape varieties, a negative correlation between gene body methylation and gene expression variation between grapevine varieties was observed. The authors proposed that a higher number of transposable elements (TEs) within the grapevine genes may be responsible for this negative association between gene body methylation and expression (Magris et al., 2019). Pereira et al. (2022) were able to characterize nine grapevine DNA methyltransferase genes and suggested that changes in grapevine genome methylation are associated with the establishment of compatible and incompatible interactions with Plasmopara viticola. A following study by Azevedo et al. (2022) observed that DNA methylation is affected by P. viticola inoculation and that differences in the DNA methylation levels are related to the different susceptibility to P. viticola. These studies provided useful insights into the role of epigenetic mechanisms in grapevine defense against downy mildew and their potential implications for future breeding programs such as improving tolerance to powdery mildew in grapevine and reducing the massive current and recurring use of chemicals. Additionally, the use of DNA methyltransferases blockers (including but not limited to 5-azacytidine, 5-aza-2’-deoxycytidine, 1-beta-D-arabinofuranosyl-5-azacytosine and dihydro-5-azacytidine) has been proposed as an approach to generate epigenetic variation for crop improvement (Amoah et al., 2012), as recently implemented in the development of drought-tolerant sugarcane epimutants (Koetle et al., 2022).

As described in this review, there is growing evidence that epigenetic mechanisms play an important role in increasing crop resilience to stresses and therefore may be an important tool in the development of more resilient grapevine cultivars. However, among the plethora of epigenetic memory of stress and priming studies done in plants, only a small amount of them is perennial-focused – even less on grapevine specifically. Contrary to the limited studies on epigenetic memory of stress and priming, there is no lack of observations of stress priming in grapevine. Some of the more recent studies that observed physiological, transcriptional, and biochemical modifications that are potentially indicative of established stress memory in grapevine include, Babajamali et al. (2022), which showed that drought stress priming improved freezing tolerance in shoot and root tissues of both drought-tolerant and sensitive grapevine cultivars. In addition, a study performed on dry-grown Cabernet Sauvignon suggested the more drought-resilient grapevines with superior vine water status, leaf gas exchange and berry size are likely due to long-term drought stress adaptation via stress priming (Pagay et al., 2022). Spray-induced gene silencing (SIGS) that targets a putative grape glutathione S-transferase (GST) gene (VvGST40) has been shown to prime vines resulting in increased resilience to severe drought (Nerva et al., 2022). In the response to salinity stress, it has been shown that 6-Benzylaminopurine (BAP) primes salt tolerance in V. vinifera, with BAP-primed plants exhibiting higher intrinsic water use efficiency, photosystem-II efficiency, and growth (Montanaro et al., 2022). Moreover, grapevines infected with Grapevine fanleaf virus (GFLV) are more resilient to mild water stress than healthy vines, suggesting that the biotic stress can potentially induce priming in grapevine (Jež-Krebelj et al., 2022). Many more studies including biotic stress priming (e.g., Trouvelot et al., 2008; Verhagen et al., 2010; Perazzolli et al., 2011) and abiotic stress priming (e.g., Tombesi et al., 2018) are good indicators of the familiarity of priming effects in grapevine. Even with the ever-growing research on epigenetic regulations in the grapevine, to date, a limited amount of research is available on how this memory of stress and its underlying epigenetic mechanisms are established, maintained, and even reset. The long lifespan of woody perennials could be used to address some of the prevailing concerns in studies with annual plants, such as whether the period of vegetative growth between the priming treatment and the second stress treatment is long enough to test whether the stress is phenotypically effective and whether the changes in the epigenome are induced by priming treatment (Sani et al., 2013). Therefore, providing valuable insights into how long-term somatic memory is established and if it can be maintained past winter dormancy. Similarly, the connection between vegetative propagation and epigenetic memory of stress establishment and maintenance should also be considered. The use of vegetative propagation (i.e., propagated through cutting or layering) in woody perennials could reveal novel and useful information on how permanent or transient long-term somatic memory is after vegetative propagation (Perrone and Martinelli, 2020). Viticulture could benefit greatly from the understanding of transient or stable modification to the epigenome of stress memory, as it may contribute to developing novel molecular approaches such as targeted, gene-specific modifications for stress adaptation through plant breeding, leading to the production of more resilient grape varieties. It should be considered, when using epigenetic and epigenomics to develop stress resilient crop, that the negative effects of stress memory on breeding in general, as the obtained stress memory could inhibit normal plant growth (Chinnusamy and Zhu, 2009). The prediction and assessment of the impact of stable epigenetic variation on plant phenotype and performance should be explored further, via machine learning and model training as demonstrated in several studies (Colicchio et al., 2015; Hu et al., 2015; N'Diaye et al., 2020).

Some challenges to the utilization of epigenomics to design environment resilient grapevine, such as the stability and heritability of the epigenetic variation, are important for the potential transmission to the progeny (Eichten et al., 2014; Iwasaki and Paszkowski, 2014; Vriet et al., 2015). Most of the stress-induced epigenetic modifications return to basal levels when the stress is removed, but some of the modifications can be inherited mitotically and meiotically in plants (Sudan et al., 2018). Such epigenetically mediated stress memory can later lead to longer-term adaptation, an indication of the possibility of using epigenetics as a tool to combat environmental stress. It is important to note, however, further study is needed for understanding various factors that might affect epiallele stability in order to avoid inducing epialleles that might be unstable during the breeding process (Hofmeister et al., 2017). Moreover, epigenetic variation can also be maladaptive and become an epigenetic trap (Consuegra and Rodríguez López, 2016), not only if the changes they induce do not match the environment experienced by the offspring, but also due to the energetic cost associated to the maintenance of the acquired epigenetic state, which could negatively impact the plant growth and development, and ultimately affect crop yield (Chinnusamy and Zhu, 2009).

Another major challenge in creating epigenetic populations in crops is the uncertainty of whether epigenetic changes (i.e., alteration of DNA methylation patterns) induced by approaches developed in model species such as A. thaliana can be transferable to crops, since so few to none viable equivalent mutants have been produced in crop species (Hu et al., 2014; Li et al., 2014; Kawakatsu and Ecker, 2019). An alternative approach such as epimutagenesis and targeted epigenome editing can be utilized, as demonstrated in A. thaliana (Johnson et al., 2014; Springer and Schmitz, 2017). However, it will require advancement and innovation in both technical and biological disciplines to bring out the full potential of epigenomic variants and use them efficiently in the breeding of better stress-adapted crops (Varotto et al., 2020). Similarly, for the integration of epigenetics and epigenomics in crop, or more specifically, grapevine breeding, more knowledge needs to be acquired on stress induced epigenetic memory in perennials. Such acquisition of knowledge should move beyond describing the correlation between epigenetic variation and the desired trait, towards the demonstration of the functional association between acquired epialleles and enhanced tolerance to stress.

Despite the gap in knowledge in stress memory establishment and maintenance, the advancement in technology and the employment of multi-omics approaches have allowed epigenetic breeding (epi-breeding) to be successful in various steps of the process (Rajnović et al., 2020). Such as the generation of mutant lines (e.g., Yang et al., 2015), recurrent epi-selection (e.g., Hauben et al., 2009; Greaves et al., 2014), and epigenome editing (e.g., Park et al., 2016), as well as the usage of priming/stress memory (e.g., Lämke and Bäurle, 2017). One of the successful examples is through suppressing the nuclear-encoded MutS HOMOLOGUE 1 (MSH1). The success of the MSH1 system has been reported in A. thaliana and tomato, where the phenotypic changes that led to improved growth vigor and yield were linked to DNA methylation, as 5-AzaC can repress those improvements, while METHYLTRANSFERASE 1 (MET1) and HISTONE DEACETYLASE 6 (HDA6) played an important role in those phenotypic changes (Yang et al., 2015; Kundariya et al., 2020; Yang et al., 2020). In soybean, epigenetic selection has led to yield improvement for at least three generations (Raju et al., 2018). Moreover, after crossing the msh1 mutant to the wild type, the epi-population created was shown to possess multiple yield-related traits both in the greenhouse and in the field (Raju et al., 2018). Many other examples showing the potential of epi-breeding for plant adaptation to various stress including the usage of eustressors have been reviewed by Kakoulidou et al. (2021), and Villagómez-Aranda et al. (2022). For these successful examples to serve as future grapevine improvement strategies, its inherent characteristics (long-living perennial, highly heterozygous, high inbreeding depression) must be considered.

If the grape industry, and by extension other perennial crop industries, want to benefit from the potential use of epi-breeding approaches to produce climate resilient varieties, future multi-omics studies should be custom designed to unravel how environmentally induced epigenetic mechanisms interact with gene expression to affect the phenotype, and to determine if environmental stress is followed by the establishment and maintenance of a memory of stress in grapevine. Such studies will lay the foundation for the development of comprehensive models integrating plant response to stress, the establishment of transcriptional and epigenetic memory of stress, and their maintenance, over time and during vegetative propagation in perennial plants.

Both authors contributed to manuscript writing and revision, read, and approved the submitted version.