Differentiation of grapevine (Vitis vinifera L.) conculta members based on molecular tools

Twenty-seven grapevine (Vitis vinifera L.) varieties within 12 putative berry colour variation groups (conculta) were genotyped with 14 highly polymorphic microsatellite (simple sequence repeats (SSR)) markers. Three additional oligonucleotide primers were applied for the detection of the Gret1 retroelement insertion in the promoter region of VvMybA1 transcription factor gene regulating the UFGT (UDP-glucose: flavonoid 3-O-glucosyltransferase) activity. UFGT is the key enzyme of the anthocyanin biosynthetic pathway. SSR results proved that the analysed cultivars can be grouped only into nine concultas, the other three putative berry colour variant groups consist of homonyms as a consequence of misnaming. In the case of Sárfehér-Sárpiros, Delaware red-Delaware white and Járdovány fekete-Járdovány fehér, it was attested that they are not bud sports, but homonyms. Some conculta members could be differentiated according to the presence or the absence of the Gret1 retroelement (Chasselas, Furmint and Lisztes), while others, Bajor, Bakator, Gohér and Traminer conculta members, remained indistinguishable either by the microsatellites or the Gret1-based method.


Introduction
Mutations affected important traits of horticultural plants to a great extent during domestication of these species including fruit colour variations. [1] Zhukovsky [2] introduced the term of conculta for the colour mutant varieties. The Hungarian ampelographer M arton N emeth [3] also adopted and extended this over-cultivar taxonomic category for grapevine. According to his theory, the grapevine conculta members originate from blue-berried ancestors as a consequence of bud mutation. The difference between the members can be recognized only by the colour of the berry, the autumn leaf colouration and the prostrate hairs of the shoot tips. Based on morphological traits N emeth [4] differentiated 26 concultas, grown in Hungary.
The identification of the cultivars is very important at all phenological stages. [5] Besides ampelographic descriptors DNA fingerprints are significant tools in varietal characterization in grapevine. Among these molecular methods microsatellite (SSR ¼ simple sequence repeats) analyses became widely used due to their reliability and high reproducibility. [6][7][8] Although the SSR markers made it possible to detect even clonal variations, [9][10][11] they are not applicable in all cases of somaclonal variations like bud mutations. [12] Bowers et al. [6] and Regner et al. [12] concluded that the SSR markers are not suitable for the discrimination of the members of Pinot group (Pinot blanc, Pinot gris, Pinot noir). Other molecular methods like Restriction Fragment Length Polymorphism (RFLP) and Random Amlified Polymorphic DNA (RAPD) also proved to be ineffective for differentiating berry colour variants. [13][14][15][16] Hal asz et al. [14] analysing autochthonous varieties in the Carpathian Basin with six polymorphic SSR markers could not discriminate the members of the Bajor, Bakator, Goh er and Lisztes concultas. Additionally they concluded that because of the high allelic differences the cultivar Bakator k ek cannot be the member of the Bakator group. The name Bakator is a homonym confirming that the similarity in names does not certainly mean a bud mutation event. Existence of homonyms and synonyms is very frequent in the nomenclature. [8,17] Slinkard and Singleton [18] suggested that the white berried cultivars originate from the coloured ones by lossof-function mutations. The berry colour is determined by anthocyanin accumulation in the skin, which varies greatly in concentration and composition depending on the grape cultivar. [19] The key enzyme of anthocyanin biosynthesis is the UDP-glucose-flavonoid 3-O-glucosyltransferase (UFGT). This enzyme does not express in the white berried cultivars, in spite of the fact that there are no differences in both VvUFGT promoter and coding region between white and coloured cultivars. [19,20] The anthocyanin biosynthesis is controlled by a transcription complex including the Myb genes, which activates the UFGT gene. [21] The ancient wild grape had coloured berries and the nowadays-cultivated varieties derived from the ancient form. The white cultivars arose mostly from red-berried parents by different mutations in two adjacent Myb genes, VvMybA1 and VvMybA2. [22,23] Among these mutations, insertion of a retroelement, the Gret1 retrotransposon into the promoter region of the VvMybA1 gene leading to transcriptional inactivation of VvMybA1 was first identified in cultivars Italia and Muscat of Alexandria. [22] This mutant allele was named VvMybA1a, while the functional allele of the coloured cultivars is VvMybA1c. White berried cultivars are homozygous for Gret1 insertion, whereas the colour-skinned varieties contain at least one functional allele. In several white cultivars deletion of Gret1 from promoter region was observed resulting in a functional allele, VvMybA1b, containing only a short part, the 3 0 -LTR region of the retrotransposon, thus these types of red cultivars derived from their white-skinned progenitor. [22,23] Single nucleotide polymorphism in VvMybA2 coding region also could result in white berries. [11] Based on the results of Mitani et al. [24] wild Vitis species do not carry the VvMybA1 locus even if they are white berried.
Yakushiji et al. [25] showed that the deletional mutation of functional VvMybA1c from Pinot noir resulted in Pinot blanc. At the same time the other members of the Pinot conculta, Pinot gris and Pinot noir are undistinguishable with the retroelement based method. [26] Kobayashi et al. [20] and Giannetto et al. [26] concluded that the colour mutations can be bidirectional: black-to-white and white-to-red/pink between the conculta members. These facts confirmed the result of Walker et al. [23] identifying two pale coloured mutations of the Cabernet sauvignon (Malian and Shalistin) as a deletion consequence in two regulatory genes of the berry colour locus.
In this paper we combine the knowledge of the cultivar characterization with SSR markers and the application of the Gret1 retroelement for discriminating 16 local Hungarian and 11 international putative or already proven bud sports.

Plant material
The cultivars investigated in this study are listed in Table 2. Young leaves of the grapevine varieties were collected from grapevine collections maintained at K aroly R obert College in Eger, Szo lo skert Ltd. in Nagyr ede, Helv ecia and University of P ecs, Institute of Viticulture and Enology.

Short characterization of the putative concultas
Bajor is probably a Hungarian variety. It used to be cultivated in most wine regions of Hungary; presently it has no importance in the Hungarian viticulture -it can be found only in old plantations of quality wine regions (Tokaj Hegyalja, Mecsek). [4,27] Bakator is a Carpathian Basin variety. K ek bakator, Piros, T€ udo sz ınu and Feh er Bakator carry the Bakator name; however, it was shown by SSR analysis that K ek bakator is a different cultivar and not a berry colour variant of Bakators. [14] Nowadays only Bakator piros is cultivated in Hungary on a few hectares. [27] Chasselas: in spite of its French name it derives from Asia, its way to Europe is unknown. The most widely cultivated table grape in Hungary, Chasselas rouge, rose and b1ancs are the members of the conculta. [4,27] Delaware red is of North-American origin, assumably a natural hybrid of Vitis vinifera, Vitis labrusca and Vitis aestivalis. In Hungary it can be found only in old vine yards, however earlier it was a popular cultivar in the Trans-Danubian part of Hungary. Delaware white was supposed to be the progeny of Delaware red as a result of open pollination. [27] Furmint: old Hungarian wine grape with three variants: piros, feh er, v altoz o. Feh er Furmint is the third most widespread cultivar in Hungary, particularly in the Tokaj Hegyalja region, one of the components of the world famous Tokay aszu. Piros and v altoz o Furmints are maintained in gene banks. [27,28] Goh er: old Hungarian variety, used as table and wine grape in the past, it has three variants (feh er, piros, v altoz o). Recently Goh er feh er has been planted at the famous Tokaj wine region. Goh er piros and v altoz o are conserved in gene banks. [27,28] J ardov any: its origin is unknown. It was widespread in the past, but nowadays it can be found mainly in old plantations. J ardov any feh er wine is not a characteristic one, but J ardov any fekete has better quality, therefore it has got permission to be planted as wine grape in some regions of Hungary. [27] Lisztes: old autochthonous variety of the Carpathian Basin, high yield and low wine quality are characteristic of Lisztes; it has no role in the present Hungarian viticulture. [4] Merlot: known since the eighteenth century, derives from France (Bordeaux). It is cultivated all over the world; it was registered in Hungary in 1973. [27,28] Pinot derives from France, where it has been growing for centuries. It is a worldwide cultivated variety with the following berry colour variants: Pinot gris, noir, blanc, rose and violet. [27] S arfeh er is an old white berried Hungarian cultivar. Because of the first syllable of its name (S ar ¼ mud) it can be assumed that the red berried S arpiros is also a berry colour variant. Before the Phylloxera epidemic it was the characteristic variety of Soml o and Neszm ely in Hungary. [27,28] Traminer: its origin is uncertain. Generally it was thought to derive from Tramin (village in South Tirol). Red, blue and white berry colour variants of Traminer are  known, but only the red grape is cultivated in Hungary, the other two exist only in collections. [27] DNA extraction and Polymerase Chain Reaction (PCR) analysis DNA was extracted with Qiagen DNeasy Plant Mini Kit according to the manufacturer's protocol (Qiagen-Biomarker Ltd., G€ od€ ollo , Hungary). The quality and quantity of the isolated DNA was checked on 1.5% agarose gel with electrophoresis and by NanoDrop spectrophotometer (BioScience Ltd. Budapest).
The standardization of the allele sizes was made by using French and Hungarian reference varieties such as Pinot noir, Chardonnay and Irsai Oliv er, Csaba gy€ ongye, and Pozsonyi, respectively. Among these cultivars some parent-progeny relationships exist confirmed by SSR results. [31,35] Results and discussion SSR analysis The SSR profiles of the characterized 27 samples are presented in Table 1. The SSR analysis resulted in 15 different allele profiles including nine concultas. Individual varietal differences could be observed in the case of the following cultivars: Delaware white-Delaware red, S arfeh er-S arpiros and J ardov any feh er-J ardov any fekete meaning that they do not constitute concultas, it can be supposed that these cultivars have only homonym names and they are not the results of bud mutations. Homonymy and synonymy is a common phenomenon in grapevine nomenclature, which can be clarified by microsatellite markers. [36][37][38] At the same time based on 14 microsatellite loci there were no differences in allele sizes within the following concultas: Bajor, Bakator, Chasselas, Furmint, Goh er, Lisztes, Merlot, Pinot and Traminer (Table 1). Our results supported the earlier conclusions of Regner et al. [12] that there are no SSR allele differences between Pinot samples.
Two members of the Chasselas, Merlot and Traminer group were analysed in this work. Based on the SSR profiles, none of the putative conculta members differed from each other at the investigated loci.
According to Galet [39] and Csepregi and Zilai [28] Delaware white is the seedling of Delaware red (V. labrusca L. Â V. aestivalis Michx.) but their SSR profile excludes the possibility of parent-progeny relationship. These cultivars are homonyms and not the results of either bud mutation or paternity.
In the case of the cultivars S arfeh er-S arpiros and J ardov any fekete-J ardov any feh er the name identity can be explained also with homonymy since the genetic profile disclaims the possibility of colour mutation.

Molecular analyses of VvMybA1 locus
Since conculta members were indistinguishable by SSR markers our second aim was not only to differentiate them but also find the reason for the berry colour variation. The method that we applied is based on detecting the presence or absence of the Gret1 retrotransposon in the promoter region of the VvMybA1 transcription factor gene. In majority of the white berried cultivars transcription of the UFGT gene (coding the key enzyme in the anthocyanin biosynthetic pathway) is blocked, because of the insertion of Gret1 retrotransposon into VvMybA1 promoter. [25,34] This et al. [34] reported three white berried cultivars (Avgoulato, Gamay Castille mutation blanche and Sultanina-Gora Chirine) which did not contain the Gret1 in the VvMybA1 promoter.
Kobayashi et al. [22] described three alleles at VvMybA1 locus based on the presence (VvMybA1a) or absence of Gret1 (VvMybA1c) element in the promoter of the gene. The third one VvMybA1b contains a 3 0 -LTR sequence remaining after Gret1 deletion.
The results of PCR analysis of the promoter region of the VvMybA1 transcription factor gene in five proven Hungarian concultas (Bajor, Bakator, Furmint, Goh er and Lisztes) are shown in Figure 1. Amplification of an $1500 bp DNA fragment indicates the presence of VvMybA1a allele in each variety (Figure 1), independently of the actual berry colour except Bajor k ek which does not contain the Gret1 insertion at all (it is homozygous for the VvMybA1c allele). Analysis of the functional VvMybA1 allele resulted in PCR products only in the coloured cultivars due to VvMybA1b and VvMybA1c alleles ( Figure 2). The coloured members of these five Hungarian concultas, Bajor sz€ urke, Bajor feketef aj u, Bakator piros Bakator t€ udo sz ınu , Furmint piros and Lisztes piros are heterozygous for the Gret1 retroelement, therefore the anthocyanin biosynthesis is undisturbed (Figure 2). Interestingly the red berried Goh er piros does not contain any functional VvMybA1 (b, c) alleles. Two members of Bajor conculta (Bajor sz€ urke and feketef aj u) gave identical DNA pattern at the VvMybA1 locus, showing that they contain both the functional VvMybA1c and the non-functional VvMybA1a alleles.
Our results are in accordance with the earlier published works on the correlation between the presence of Gret1 and the loss of the berry colour. [26,34] Although S arfeh er, J ardov any fekete and Delaware red have not proven to be concultas, we analysed their VvMybA1 locus. Red berried S arpiros and J ardov any fekete are heterozygous for the Gret1 insertion.
Surprisingly, in the case of the characterization of the promoter region of the VvMybA1 transcription factor gene, the primers amplified a larger fragment in the same region in Delaware red ( Figure 2).
In Traminer red -similar to Goh er piros -no functional VvMybA1b or c alleles could be detected. The members of the Bajor, Bakator and Goh er concultas were indistinguishable based on the Gret1 retroelement. Thus these data call for a further study in order to find an unambiguous method for distinguishing varieties in the concultas. Our results also revealed that Chasselas rouge -likewise Lisztes piros and Furmint piros -contains a VvMybA1b allele. The possibility of discrimination and identification of the investigated cultivars are listed in Table 2.

Conclusions
The results of the present study demonstrate that the applied 14 SSR markers were appropriate to prove which cultivars constitute concultas. Nine of the 12 putative cultivar groups can be considered as real concultas. Members of these concultas were indistinguishable with the SSR primer set used in this study. Earlier results proved the loss  of berry colouration to be the consequence of Gret1 insertion in the VvMybA1 promoter region. Therefore we analysed the VvMybA1 locus coding a key transcriptional factor of berry anthocyanin biosynthesis. Testing three alleles of VvMybA1 (a, b, c) among the five Hungarian concultas only Furmint piros and Lisztes piros could be differentiated from the white berried variants. Bajor (except Bajor k ek), Bakator and Goh er concultas require further investigations to clarify the genetic background of their berry colour.