Characterized one of the largest collections of grapevine rootstocks (non-vinifera) by molecular markers

Article extracted from the presentation held during Enoforum Web Conference (23-25 February 2021)

The non-vinifera germplasm: the stranger!

Since the first decade of the 19th century, grapevine has been grown on rootstocks, represented by a mix of non-vinifera grapevine species, such as Vitis berlanderi, Vitis rupestris, Vitis riparia and hybrids. Rootstocks contribute to control other soil borne pests (e.g., phylloxera and nematodes) and different environmental conditions (e.g., drought and salinity) (Galet, 1998; Whiting, 2005).

To date, rootstock genetic background is extremely narrow to the extent that less than ten rootstocks are used worldwide for grafting (Keller, 2010). Moreover, these genotypes, released at the end of the 19th century, seem to be inadequate to cope with the emerging challenges in viticulture related to climate change (Huglin and Schneider, 1998).

Since 1985, the DiSAA (Department of Agricultural and Environmental Sciences) research group operating at the University of Milan (UMIL) has been working on the selection of new grapevine rootstocks. M-series (M1, M2, M3 and M4) rootstocks were selected and, recently, registered at the National Register of Vitis varieties. These rootstocks showed tolerance to iron-limited conditions (M1 > M3) (Porro et al., 2013; Vannozzi et al., 2017), moderate resistance to salinity (M2 and M4) (Porro et al., 2013; Meggio et al., 2014), and high tolerance to drought (M4) (Porro et al., 2013; Meggio et al., 2014; Corso et al., 2015; Bianchi et al., 2020b).

Because we are far from solving the climate change challenges, new genotypes dealing with unfavorable environmental conditions should be released in the next future. Our group is keeping working on this in way. In order to select the most promising genotypes for forthcoming rootstock breeding programs, we established one of the largest non-vinifera germplasm collections at UMIL, collecting plant materials from Italy, Spain, and USA, and assessed the genetic variability in the new repository.

The step-by-step methodologies

The first step towards the selection of new pre-breeding materials was to genotype the new non-vinifera collection. In collaboration with the Research Centre for Viticulture and Enology, (Conegliano, Italy) the collection was genotyped using 22 SSR (Simple Sequence Repeat) markers (Migliaro et al., 2019).  These markers are the same universally applied to genotype the vinifera germplasm (Laucou et al., 2011). The SSR profiles were used to identify unique profiles, to investigate the genetic diversity of the collection and to sort out a core collection. Finally, the core collection was genotyped using the new high-throughput genotyping array holding more than 18kK SNP markers (Laucou et al., 2018). The SNP profiles were used to compare the genetic variability between non-vinifera and vinifera compartments (Bianchi et al., 2020a).

The genetic variability of non-vinifera collection at UMIL

The non-vinifera is located in the winegrowing region called Oltrepò Pavese (9°05´E; 44°58´N; 144 m asl) collection and, to date, it accounts for more than 400 accessions. Three hundred and seventy nine individuals were genotyped with 22 SSRs. As expected, approximately 39% of accessions represented redundant germplasm and 232 unique genotyped were detected. Among them, 111 genotypes were identified according to information available in literature and databases cross-checking. More than 50% of genotypes were unknown genotypes. Since the collection hosts genotypes belonging to different species and hybrids, a large number of alleles was detected, as well as a large genetic diversity. The genetic analyses identified three main ancestral groups: i) one cluster included mainly V. rupestris and V. riparia genotypes; ii) one grouped the progenies of the three varieties most used in the past for rootstock breeding programs (e.g. V. berlandieri Rességuiere 2, V. riparia Gloire de Montpellier, V. rupestris du Lot); another one grouped mostly V. labrusca hybrids. A fourth, admixed group accounted for individuals with complex pedigree, where more than two species were involved (Migliaro et al., 2019).

A core collection feasible to handle

Large germplasm collections are difficult to manage for accurate phenotyping. In order to overcome this problem, we extrapolated a core collection capturing the entire genetic diversity in a restricted number of genotypes feasible to handle (Brown, 1995). Our core collection accounted for 70 genotypes. Interesting, only 7% of genotypes were berlandieri × rupestris and berlandieri × riparia genotypes. The majority of core collection individuals were unknown genotypes, with few information in terms of ampelographic and agronomic traits. Moreover, this core collection included the putatively novel pre-breeding materials.

The SNP are a useful tool to study the genetic diversity of non-vinifera germplasm

So far, the Vitis18kSNP genotyping array has been extensively used to investigate the genetic diversity of vinifera germplasm (De Lorenzis et al., 2015, 2019; Mercati et al., 2016; Ruffa et al., 2016; Laucou et al., 2018; Sunseri et al., 2018; D’Onofrio et al., 2021). The array, set up by the GrapeReSeq Consortium, holds around 18k SNP markers. Some of them (around 13.5k) were detected re-sequencing 47 V. vinifera genotypes, while the others (around 4.5k) were detected in the genome of non-vinifera species (Vitis aestivalis, V. berlandieri, Vitis cinerea, Vitis labrusca, Vitis lincecumii, and Muscadinia rotundifolia) (Laucou et al., 2018). Recently, the array has been validated on non-vinifera genotypes as well, using the core collection as training population (Bianchi et al., 2020a). More than 14k loci were successfully amplified, with 61% of markers identified in the V. vinifera genome and around 39% of SNPs identified in the genome of other species, confirming the usefulness of this tool in genotyping non-vinifera germplasm. The uniqueness and preciousness of core collection has been proved by genetic analyses. Four ancestral groups have been identified: i) group 1 clustered only three known genotypes (V. berlandieri × V. riparia, V. vinifera × V. rupestris and V. labrusca); ii) group 2 clustered V. cordifolia, V. labrusca, V. rupestris, and V. vinifera genotypes; iii) group 3 clustered genotypes with V. candicans, V. labrusca, V. longii and V. rupestris in their pedigree; iv) group 4 clustered pure V. riparia genotypes and most of unknown genotypes (Bianchi et al., 2020a). Moreover, when the non-vinifera SNP profiles were compared with the vinifera ones, a clear differentiation among the two compartments has been highlighted (Bianchi et al., 2020a).

The promising novel breeding materials

In order to identify rootstocks breeding materials not commonly used so far, a new non-vinifera collection was established at UMIL. The collection showed high genetic variability, summarized in a core collection encompassing 70 genotypes. Most of this variability has not explored yet and preliminary results highlighted that some genotypes showed promising behavior to cope with drought, with performances sometimes better than the one detected in tolerant reference rootstocks (Bianchi et al., 2018). We are on the good way to get the most promising genotypes for forthcoming rootstock breeding programs.

Acknowledgments

This research was funded by Winegraft srl, SIV (Service for grapevine identification of CREA Viticulture and Enology of Conegliano) and European project titled “FREECLIMB – Fruit Crops Adaptation To Climate Change In The Mediterranean Basin” in the frame of the Programme Partnership For Research And Innovation In The Mediterranean Area (PRIMA; call 2018).