South American species Solanum alandiae Card. and S. okadae Hawkes et Hjerting as potential sources of genes for potato late blight resistance

For several decades, wild species of Solanum L. section Petota Dumort. have been involved in potato cultivar breeding for robust resistance to pests and diseases. Potato late blight (LB) is caused by oomycete Phytophthora infestans (Mont.) de Bary, and the genes for race-specific resistance to P. infestans (Rpi genes) have been introgressed into cultivated potatoes by remote crosses and trans- or cisgenesis, first from S. demissum Buk. and, more recently, from other wild species, such as S. bulbocastanum Dun., S. stoloniferum Schlechtd. et Bché, and S. venturii Hawkes et Hjerting (according to the nomenclature by Hawkes, 1990). Most wild species already involved in breeding for LB resistance came from North and Central Americas: series Bulbocastana (Rydb.) Hawkes, Demissa Buk. and Longipedicellata Buk., and some Rpi genes of these species have been already characterized in much detail. Rpi genes of South American species, including the series Tuberosa (Rydb.) Hawkes, have not been sufficiently investigated. Among the latter, this study focuses on the Rpi genes of S. alandiae Card. and S. okadae Hawkes et Hjerting. Four accessions of S. alandiae, one accession of S. okadae and 11 clones of interspecific potato hybrids comprising S. alandiae germplasm from the VIR collection were PCR-screened using specific SCAR (Sequence Characterized Amplified Region) markers for eight Rpi genes. SCAR amplicons of five Rpi genes registered in this study were validated by comparing their sequences with those of prototype genes deposited in the NCBI Genbank. Among the structural homologues of Rpi genes found in S. alandiae and S. okadae, of special interest are homologues of CC-NB-LRR resistance genes with broad specificity towards P. infestans races, in particular R2=Rpi-blb3, R8, R9a, Rpi-vnt1 and Rpi-blb2 (94–99, 94–99, 86–89, 92–98 and 91% identity with the prototype genes, respectively). Our data may help to better understand the process of Rpi gene divergence along with the evolution of tuberbearing Solanum species, particularly in the series Tuberosa.

For several decades, wild species of Solanum L. section Petota Dumort. have been involved in potato cultivar breeding for robust resistance to pests and diseases. Potato late blight (LB) is caused by oomycete Phytophthora infestans (Mont.) de Bary, and the genes for race-specific resistance to P. infestans (Rpi genes) have been introgressed into cultivated potatoes by remote crosses and trans-or cisgenesis, first from S. demissum Buk. and, more recently, from other wild species, such as S. bulbocastanum Dun., S. stoloniferum Schlechtd. et Bché, and S. venturii Hawkes et Hjerting (according to the nomenclature by Hawkes, 1990). Most wild species already involved in breeding for LB resistance came from North and Central Americas: series Bulbocastana (Rydb.) Hawkes, Demissa Buk. and Longipedicellata Buk., and some Rpi genes of these species have been already characterized in much detail. Rpi genes of South American species, including the series Tuberosa (Rydb.) Hawkes, have not been sufficiently investigated. Among the latter, this study focuses on the Rpi genes of S. alandiae Card. and S. okadae Hawkes et Hjerting. Four accessions of S. alandiae, one accession of S. okadae and 11 clones of interspecific potato hybrids comprising S. alandiae germplasm from the VIR collection were PCR-screened using specific SCAR (Sequence Characterized Amplified Region) markers for eight Rpi genes. SCAR amplicons of five Rpi genes registered in this study were validated by comparing their sequences with those of prototype genes deposited in the NCBI Genbank. Among the structural homologues of Rpi genes found in S. alandiae and S. okadae, of special interest are homologues of CC-NB-LRR resistance genes with broad specificity towards P. infestans races, in particular R8,R9a,94 99,86 89,92 98 and 91% identity with the prototype genes, respectively). Our data may help to better understand the process of Rpi gene divergence along with the evolution of tuberbearing Solanum species, particularly in the series Tuberosa.

Introduction
Late blight (LB) caused by oomycete Phytophthora infestans (Mont.) de Bary is among the major obstacles on the road to sustainable potato production. To limit this disease, breeders deploy the technologies of remote crosses and trans-and cisgenesis to transfer the genes of resistance to P. infestans (Rpi genes) from wild Solanum species into susceptible potato cultivars, which are demanded by the market. However, due to rapid genome evolution and migration of new strains, severe outbreaks of the pathogen have been relentlessly overcoming plant defense barriers built up by breeders and negating, literally within a few days, their many years of effort (Cooke et al., 2012;Fry, 2016). By combining (pyramiding) several Rpi genes in one potato plant, breeders provide for robust and durable resistance of new cultivars to a broad range of P. infestans pathotypes. Therefore, it is an urgent task of plant biologists to constantly seek for new sources of resistance, predominantly in wild Solanum genotypes, which have not been as yet introduced into potato breeding, and expand the scope of Rpi genes thoroughly characterized and documented with molecular methods Bethke et al., 2019).
All presently characterized Rpi genes belong to the CC-NB-LRR type: their protein products comprise coiled-coil, nucleotide-binding and leucine-rich region domains (Hein et al., 2009;Jupe et al., 2012;Rodewald, Trognitz, 2013). The mechanisms of immediate molecular interactions between the products of P. infestans avirulence genes and the products of Solanum Rpi genes, that is, effector recognition by receptor kinase, have been sufficiently researched only in few cases, such as Avr3a -R3a interaction. However, such lack of information on the Rpi genes does not hinder the genetic studies aimed at mining Solanum collections for new Rpi genes and introducing these genes into breeding for durable LB resistance. Following the marker-assisted mapping and cloning of Rpi genes, recent years have seen a considerable progress in this direction. New methodologies, such as effectoromics, allele profiling and diagnostic resistance gene enrichment sequencing (dRenSeq), tremendously facilitated the search for new Rpi genes and new alleles of Rpi genes already characterized in other species of Solanum L. within the section Petota Dumort van Weymers et al., 2016;Chen et al., 2018;Jiang et al., 2018;Armstrong et al., 2019). Currently some of these genes have been introgressed into commercial potato cultivars (Haesaert et al., 2015).
The potato genetic collection maintained at the N.I. Vavilov Institute of Plant Genetic Resources (VIR) in St. Petersburg is a major source of prospective Rpi genes. These genes can be found in the collection accessions of wild potatoes (Zoteyeva, 2012) and in the multiparental interspecific hybrids comprising numerous Rpi genes of broad race specificity toward P. infestans. It is most significant that such hybrids could keep Rpi genes already lost from the collections of wild species (Fadina et al., 2017;Rogozina et al., 2018).
Most Rpi genes already characterized in sufficient detail and employed by breeders (Hein et al., 2009;Rodewald, Trognitz, 2013;Vossen et al., 2014;Aguilera-Galvez et al., 2018;Rogozina et al., 2018) come from wild Petota species growing in North and Central America: series Bulbocastana (Rydb.) Hawkes, Demissa Buk. and Longipedicellata Buk. South America hosts three fourth of unique Petota species (Spooner et al., 2014); nevertheless, most of these species are terra incognita as far as Rpi genes are concerned, with the best known exception of the Rpi-vnt1 gene from S. venturii Hawkes et Hjerting. This fully applies to two species from the series Tuberosa, S. alandiae Card. (accepted by Spooner et al., 2014, as S. brevicaule Bitter) and S. okadae Hawkes et Hjerting. We screened these species with specific SCAR (Sequence Characterized Amplified Region) markers of eight Rpi genes and found several structural homologues of Rpi genes with broad specificity toward P. infestans races: R2 / Rpi-blb3, R8,R9,. Some data presented below have been reported at the XVII EuroBlight Workshop (Muratova [Fadina] et al., 2019).

Materials and methods
Plant material from the VIR potato genetic collection included the lines isolated from S. alandiae accessions k-18473, k-20408 and k-21240. The S. okadae line was isolated from the accession k-25397 derived from P. infestans resistant accession CGN 18279, which was kindly provided by Roel Hoekstra (Wageningen University & Research, Wageningen [WUR]). In addition, our study included potato interspecific hybrids containing S. alandiae germplasm and several potato cultivars and hybrids employed as references in LB assessments and as positive and negative controls in the marker analysis. These hybrids and cultivars are maintained in the collection of the All-Russian Research Institute of Phytopathology, Moscow Province, Russia (http://www. vniif.ru).
Resistance of wild species to P. infestans was evaluated in the laboratory tests with detached leaves according to Fadina et al. (2017) using a highly virulent and aggressive isolate of P. infestans 161 (races 1.2.3.4.5.6.7.8.9.10.11, mating type A1) collected in Moscow Province (the collection of the Institute of Phytopathology), and cv. 'Santé' as a reference. The LB resistance of potato hybrids and cultivars was assessed in the field trials at the Institute of Phytopathology and VIR under natural infestation by registering the area under the disease progress curve (AUDPC) against several cultivars used as references. All experimental data for LB resistance were converted to 1-9-point scores.
Plant DNA samples were PCR-screened for eight Rpi genes: R1, R2/Rpi-blb3, R3a, R3b, R8, Rpi-blb1 = Rpi-sto1, Rpi-blb2 and Rpi-vnt1 (Fadina et al., 2017). Specific SCAR markers are based on the sequences of Rpi prototype genes, that is, the genes extensively characterized by their structure and function. The markers were validated against Solanum genotypes comprising the functional alleles of the corresponding genes. Methods for DNA isolation, PCR primers ( Fig. 1; Table 1) and protocols for PCR analysis, cloning and sequencing of DNA fragments as well as bioinformatic procedures were described previously (Fadina et al., 2017; Muratova [Fadina] et al., 2019). SCAR amplicons were verified by comparing their sequences to those of prototype genes deposited in the NCBI Genbank.

Results and discussion
Several accessions of S. alandiae were reported to manifest considerable resistance to P. infestans (Perez et al., 2001;Bhardwaj et al., 2018;Zoteyeva, 2019); however, the Rpi genes in S. alandiae have not been researched extensively. By screening S. alandiae accessions and S. alandiae hybrids with potato cultivars we found the structural homologues of several Rpi genes of broad specificity toward P. in-    Table 3). Comparison of marker sequences to those of the prototype genes showed that the S. alandiae genome comprised the structural homologues of R2/Rpi-blb3, R8,R9a,94 99,86 89,92 98 and 91% identity with the prototype genes, respectively). The marker of the R1 gene, found only in potato cultivars and their hybrids, was apparently derived from that of S. demissum. The markers of the R3b and Rpi-blb1 = Rpi-sto1 genes were not found in the analyzed geno types. This is the first report on the structural homologue of the Rpi-vnt1 gene in S. alandiae. S. okadae amplicons sequenced in this study were 97-98% identical to the Rpi-vnt1.3 homologue from the S. alandiae genome ( Table 3). The Rpi-vnt1.2 allele was reported in the accession CGN 18279 (Pel, 2010), corresponding to S. okadae k-25397; however, Pel and his associates (Pel, 2010) presumed that this accession belonged to the species S. venturii.
The structural homologues of the Rpi-vnt1 gene from S. alandiae and S. okadae differed from the prototype gene   Rpi-vnt1.1 of S. venturii (FJ423044) by several single nucleotide polymorphisms (SNPs) and one three-nucleotide insertion. Rpi-vnt1 clones from S. alandiae (accession k-18473) revealed two diverse variants of sequences with 90% identity: the sequence Rpi-vnt1 S. alandiae 2 comprised a 24-nucleotide deletion, which was also characteristic of the Rpi-vnt1.1 sequences in S. okadae, the prototype gene from S. venturii (FJ423044) and the marker Rpi-vnt1.3-612 (MH297492) cloned from cv. Alouette (Fig. 2). One of the SNPs in Rpi-vnt1 from S. alandiae 1 resulted in a stop codon and, as a consequence, a shortened protein (Fig. 3), which was 83% identical to Rpi-vnt1-like amino acid sequence from S. okadae (ADB85624) and only by 77% to the corresponding sequence from S. venturii (ACJ66596). The amino acid sequences were identical to the corresponding Rpi-vnt1 sequence (ACJ66596) encoded by the functionally active Rpi-vnt1.3 allele of S. venturii (FJ423046). Thus, we can assume that at least some regions of Rpi-vnt1 in S. alandiae and S. okadae are translatable. In addition, S. alandiae comprised sequences resembling the Rpi-vnt-like, which were different from the functional gene Rpi-vnt1. The presence of the Rpi-vnt1.3-612 marker in S. alandiae and potato hybrids was significantly related to elevated LB resistance (the Spearman's correlation coefficient 0.54 at 5% confidence level).

Conclusion
The comparative SCAR marker analysis of S. alandiae and S. okadae accessions in the VIR collection discovered several structural homologues of already known Rpi genes. When analyzing these data, two caveats are appropriate. First, we dealt with rather short DNA fragments, which never covered more than one third of the complete gene sequence. Secondly, even when these fragments were translatable, the proof of functionality of Rpi homologues in S. alandiae and S. okadae must await further studies by independent methods. Our data match the evidence that the Rpi-vnt1 gene is specific to South American species of series Tuberosa (Pel, 2010) and is absent in Mexican S. verrucosum (Chen et al., 2018), despite the fact that the latter species is often grouped with South American Tuberosa germplasm (Hardigan et al., 2015). Other Rpi homologues found in S. alandiae and S. okadae resemble the R2, R8, R9 and Rpi-blb2 genes of Mexican species S. demissum and S. bulbocastanum. An Rpi-mcd1 gene orthologous to R2 but distinct in its chromosomal localization was reported in South American species S. microdontum (Hein et al., 2009;Rodewald, Trognitz, 2013;Vossen et al., 2014;van Weymers et al., 2016;Aguilera-Galvez et al., 2018); however, the sequence of Rpi-mcd1 has not yet been published. South American homologues of Rpi genes reported here for the first time are notably distinct from their Mexican prototypes. Among the Rpi homologues found in S. alandiae and S. okadae, especially promising for potato breeders are those resembling the genes of broad specificity toward P. infestans races, such as R2 / Rpi-blb3, R8, R9, Rpi-vnt1 and Rpi-blb2. In this context, it seems proper to mention our evidence that the presence of an Rpi-vnt1 marker significantly correlated with superior LB resistance.
Our search for new sources of Rpi genes among wild Solanum species takes us to two entwined issues of evolutionary genomics of the tuber-bearing Solanum: genome and species evolution in the section Petota and the origin of Rpi genes associated with the evolution of P. infestans -potato interactions often resulting in LB outbursts. In this context, the genes for defense against pests and diseases in S. alandiae and S. okadae are especially interesting as regards the evolution of the tuberbearing Solanum. J. G. Hawkes (1990) emphasized close relations of these two species with S. microdontum and resemblance of S. okadae to the Argentinian endemic S. venturii. The areas of distribution of the two latter species are located in neighboring highlands of Bolivia and northwestern Argentina. S. alandiae is a Bolivian endemic from the bordering Departments of Cochabamba, Santa Cruz and Chuquisaca, with its plants growing in contrasting climates of cold and hot dry grasslands and warm and wet territories (Fuentes, 2014). The area of S. okadae is disjunct. In Bolivia, it is an endangered species, with few tiny islands in the Departments of Cochabamba and La Paz. Here, its natural habitat is alpine wet meadows and forests, where plants are affected by LB (Coca Morante, Castillo Plata, 2007). The Argentinian part of the S. okadae area of distribution is in the Provinces of Jujuy and Salta. On the basis of molecular evidence, the Argentinian S. okadae accessions have been recently identified as S. venturii (PBI Solanum Project, 2020). These data presume independent evolution of S. alandiae and S. okadae / S. venturii.
All presently characterized Rpi genes belong to the CC-NB-LRR structures, with their evolution widely researched in tuber-bearing Solanum species (Hein et al., 2009;Pel, 2010;Jupe et al., 2012;Rodewald, Trognitz, 2013;Aguilera-Galvez et al., 2018). These species are primarily found in the Mexican and Andean centers of Petota diversity (Spooner et al., 2014;Hardigan et al., 2017), where they successfully cohabit with local P. infestans races (Grünwald, Flier, 2005;Fry, 2016). Many aspects of Petota evolution and Rpi gene geography are hotly debated Spooner et al., 2014;Hardigan et al., 2015;Hardigan et al., 2017), and our data may supplement the research into divergence of Rpi genes as related to the evolution of Solanum species, emergence of tuberbearing forms and their distribution between two Americas. An impressive illustration of the latter issue is the presumably reciprocal segregation of the Rpi-blb1 and Rpi-vnt1 genes between the Mexican and Andean species, respectively.