Index References

Potato and its wild relatives (tuber-bearing Solanum species of subsection Potatoe G. Don) were classified into seven cultivated species and over 200 wild species, which were further grouped into 19 series based on morphology, geographical distribution, cross-compatibility and ploidy (Hawkes, 1990). The variability in potato chloroplast DNA (ctDNA) has been investigated at various levels for disclosing inter-species and –series relationships (Hosaka et al., 1984; Spooner et al., 1991; Spooner and Sytsma, 1992; Spooner and Castillo, 1997) and the maternal ancestry of cultivated potatoes (Hosaka, 1995; Provan et al., 1999; Ames and Spooner, 2008). The common potato, a tetraploid crop grown worldwide (S. tuberosum L. ssp. tuberosum, 2n = 4x = 48), has a unique ctDNA, named T-type ctDNA (Hosaka, 1986), characterized by a 241 bp deletion (Kawagoe and Kikuta, 1991). The T-type ctDNA is predominant in worldwide potato varieties (Hosaka and Hanneman, 1988; Waugh et al., 1990; Powell et al., 1993; Bryan et al., 1999; Provan et al., 1999; Lössl et al., 2000). The maternal ancestor conferring T-type ctDNA was identified as some populations of a diploid wild species S. tarijense Hawkes (Hosaka, 2003). Besides T-type ctDNA, W-, C-, S- and A-type ctDNAs have been identified by a restriction fragment length polymorphism (RFLP) analysis among cultivated potatoes (Hosaka, 1986).

The RFLP analysis of mitochondrial DNA (mtDNA) revealed several mtDNA types among S. tuberosum and diverse wild species (Lössl et al., 1999). The mtDNA type ‘β’ was associated with T-type ctDNA, while ‘α’-type mtDNA was characteristic to S. stoloniferum Schlechtd. et Bché. and S. demissum Lindl., these two types obviously differing from those of Andean cultivated potato species (Lössl et al., 1999). Their mtDNA typing, however, did not likely reflect the inter-series-relationships unlike the ctDNA typing, although at least in a survey of potato varieties, some correlation between mitochondrial and chloroplast genomes was found, suggesting coevolution of both organelles (Lössl et al., 1999). However, little is known about the coevolutionary relation for a little wider species group including cultivated potato species, the putative ancestral wild species and their closely related wild species. In this study, the same set of accessions of cultivated and wild species as those previously investigated by nuclear DNA (nDNA) and ctDNA analyses were examined by the mtDNA analysis. Specific objectives are to disclose 1) the level of variability of newly developed mtDNA markers designed for simple sequence repeats (SSRs) with mononucleotide in tobacco mtDNA, 2) the maternal parent of the common potato suggested by the mtDNA analysis, and 3) to what extent the differentiation of mtDNA is correlated with that of ctDNA. The species relationships based on the mtDNA similarity were not particularly discussed here because they were supplementary to, and basically coincided with our earlier findings (Sukhotu and Hosaka, 2006).

From the complete sequence of tobacco mtDNA (Sugiyama et al., 2005), mononucleotides with over 11 copies were found only for adenine and thymine in 25 regions, among which three (T)₁₃ occurred in the same repeat unit scattering three times in the mtDNA (‘repeat 2’). Thus, 23 primer sets were designed by a web-based program Primer3 (Rozen and Skaletsky, 2000), and tested using S. tuberosum ssp. tuberosum cv. Konafubuki as template DNA. Polymerase chain reaction (PCR) was performed in a volume of 10 μl consisting of 10 ng template DNA, 0.3 μM each of primers, 5 μl of Ampdirect^® Plus (Shimadzu Co., Japan) and 0.25 units Taq DNA polymerase (Nova Taq™ Hot Start DNA polymerase, Novagen^®, USA) in a thermal cycler (Gene Amp^® PCR System 9700, Applied Biosystems) with thermal cycling conditions of one cycle of 10 min at 95°C, followed by 40 cycles of 1 min at 94°C, 1 min at 55°C, and 1 min at 72°C, and then, terminated with one cycle of 5 min at 72°C. Sixteen primer sets, including the one for ‘repeat 2’, generated clear single-banded patterns on a 3.0% agarose gel. The sequences amplified using these primer sets were determined, all of which showed more than 80% similarity with those of tobacco. Most of the mononucleotide repeats in tobacco mtDNA were also mononucleotide repeats (9 loci) or those with one base insertion/substitution (5 loci) in potato mtDNA, except for A17-1 and T16-1, which did not retain mononucleotide repeats anymore (Table 1). Thus, these 14 regions were regarded as the potato mtDNA SSR loci. All 16 loci were further evaluated using 30 accessions including those of all cultivated species and their possible ancestral wild species (Table 2). Amplification products were mixed with 10 μl of loading dye (95% formamide, 0.25% Bromophenol blue and 0.25% Xylene cyanol). After heat-denaturation, 5 μl of the sample were separated by electrophoresis in 4.0% denaturing polyacrylamide gels (Sequi-Gen^® GT Nucleic Acid Electrophoresis Cell, Bio-Rad) at 45 W constant power for 2 hr and visualized by silver staining (Bassam et al., 1991). For evaluation of polymorphism, Diversity index (h) (Nei, 1987) was calculated as h = 1 – Σp_i², where p_i is the frequency of the ith marker phenotype. Ten of 16 loci did not exhibit any polymorphism among these samples, thus showing h = 0 in Table 1. Polymorphic loci except A14-1 showed very low degree of polymorphism with h ranging from 0.064 to 0.180. In contrast, A14-1 displayed four different banding patterns in these limited samples and showed h = 0.620. For the further study, three SSR loci with the highest Diversity indices (A14-1, T11-2 and T12-3) were used.

View Details

Table 1 mtDNA SSR markers developed, and their variability among 30 samples as estimated by Diversity index (h)

A large sample set of 302 accessions of seven cultivated species and 174 accessions of 32 wild species closely related to cultivated ones except S. pinnatisectum Dun. and S. stoloniferum were evaluated (Table 2). Their banding patterns of ctDNA using RFLP analysis (Hosaka, 1986), seven SSR and one cleaved amplified polymorphic sequence (CAPS) markers (Sukhotu et al., 2004) were all obtained previously in Sukhotu and Hosaka (2006). To disclose mtDNA polymorphisms, various types of DNA markers including three SSR markers described above were used. Primer sets of ALM_4 (5’-AATAATCTTCCAAGCGGAGAG-3’) and ALM_5 (5’-AAGACTCGTGATTCAGGCAAT-3’) (Lössl et al., 2000), and nad1B (5’-GCATTACGATCTGCAGCTCA-3’) and nad1C (5’-GGAGCTCGATTAGTTTCTGC-3’) (Demesure et al., 1995) generated sequence-characterized amplified region (SCAR) markers. PCR was carried out in a volume of 10 μl consisting of 10 ng genomic DNA, 0.3 μM each of primers, 1 × PCR buffer attached to the enzyme, 200 μM each of dATP, dCTP, dGTP and dTTP, and 0.25 units of Taq DNA polymerase (AmpliTaq^®, Applied Biosystems), using the following parameters: (1) initial denaturation at 94°C for 3 min; (2) 40 cycles of denaturation at 94°C for 30 s, annealing for 30 s at 57 and 55°C for primer sets of ALM_4/ALM_5 and nad1B/nad1C, respectively, and extension at 72°C for 1.5 min; (3) final extension at 72°C for 5 min. Amplification products were separated on a 1.4% agarose gel. A primer set of rpS14 (5’-CACGGGTCGCCCTCGTTCCG-3’) and cob (5’-GTGTGGAGGATATAGGTTGT-3’) (Demesure et al., 1995) was a CAPS marker: after PCR using this primer set with an annealing temperature of 57.5°C, amplification products were ethanol-precipitated, digested with a restriction endonuclease MseI (New England BioLabs Inc.), and separated on a 1.4% agarose gel.

View Details

Table 2 The number of Solanum species accessions used in this study, and their characterization using ctDNA and mtDNA markers

The Diversity index for each ctDNA marker varied from 0.359 to 0.763 when all samples were analyzed (Table 3). All mtDNA markers were successfully amplified by PCR from all samples and the variability was shown by h = 0.066 to 0.650 (Table 3). Apparently, two mtDNA SSR loci T11-2 and T12-3 showed extremely low variability with h = 0.085 and 0.066, respectively. Thus, all developed mtDNA SSR loci except A14-1 were none or less polymorphic compared with ctDNA SSR loci and even with the other mtDNA markers. A very low variability in mtDNA SSR regions than in ctDNA SSR regions has also been reported in common wheat and its ancestral species (Ishii et al., 2006). In many plant species, however, a higher diversity in mtDNA than in ctDNA or nDNA (McClean and Hanson, 1986; Palmer, 1992; Ullrich et al., 1997) and in SSR regions than in other DNA sequences (Powell et al., 1995; Provan et al., 2001) has been known. This discrepancy can be explained partly by evolutionary drive forced on mtDNA. Plant mtDNAs rearrange very rapidly, but change very slowly in nucleotide sequence (Palmer and Herbon, 1987; Wolfe et al., 1987; Palmer, 1992). Consequently, the identification of mtDNA types, or the measurement of diversity, in plants has been mostly achieved by means of RFLPs (Neale and Sederoff, 1989; Dong and Wagner, 1993; Strauss et al., 1993; Lössl et al., 1999). However, since the expansion and/or contraction of the repeat region in mtDNA or ctDNA occur by replication slippage mutations (Richards and Sutherland, 1994) and the mutation rate in mtDNA is much slower than in ctDNA or nDNA (Wolfe et al., 1987), the mtDNA SSR regions per se might evolve slowly.

View Details

Table 3 Variability of ctDNA and mtDNA markers among all samples as estimated by Diversity index (h)

Six ctDNA RFLP types (W, W2, C, S, A and T) were found among the samples (Table 2). Combining the other ctDNA markers, a total of 72 different ctDNA banding patterns were identified (Table 3), the combination of which distinguished 129 different ctDNAs. Using six mtDNA markers including three SSR markers, the same sample set generated 40 banding patterns (Table 3), the combination of which distinguished 63 different mtDNAs. Combining ctDNA and mtDNA markers, 164 types (hereinafter, referred to as haplotypes) were distinguishable among the samples. The samples having T-type ctDNA (S. tuberosum and S. tarijense) showed a characteristic mtDNA by the locus T11-2 (Fig. 1), which was shared among all T-type ctDNA holders, two of nine accessions of S. vernei Bitt. et Wittm. (PI 458373 and PI 473306) and one of two accessions of S. × sucrense Hawkes (PI 473506). These S. vernei accessions differed from T-type ctDNA holders by the other mtDNA markers, whereas the S. × sucrense accession was indistinguishable by any mtDNA markers but distinguishable by ctDNA markers. Thus, the T-type ctDNA holders were identified as a unique haplotype, designated with a haplotype identity number of 164 in Table 2 and Fig. 2. Our finding of the characteristic mtDNA shared among all T-type ctDNA holders lends a strong support to S. tarijense having acted as a maternal ancestor of potato (Hosaka, 2003).

View Details

Fig. 1 Banding patterns obtained by the mtDNA SSR marker T11-2. The ctDNA RFLP type for each sample is also shown below the pattern. From left to right; S. stoloniferum (PI 195167), S. chacoense (PI 537025), S. tarijense (PI 473228, PI 473232, PI 473239, PI 473243, PI 498399), S. boliviense (PI 545964, PI 498215), S. megistacrolobum (PI 265874, PI 473361, PI 473356), S. raphanifolium (PI 210048, PI 473371), S. sogarandinum (PI 230510), S. acaule (PI 210030), S. chomatophilum (PI 365327, PI 266387), S. irosinum (PI 568985), S. acroglossum (PI 498204), S. blanco-galdosii (PI 442701), S. acroscopicum (PI 365315, PI 365314), S. ambosinum (PI 365316), and S. brevicaule (PI 498111).

View Details

Fig. 2 Neighbor-joining trees showing relationships among 124 ctDNA types (A), 63 mtDNA types (B), and 164 haplotypes distinguished by combining ctDNA and mtDNA markers (C).

The differences among ctDNA types (129 types), mtDNA types (63 types) and combined types (164 haplotypes) were obtained by counting the number of different banding patterns with a pairwise comparison manner, resulting in three distance matrices. Pearson’s correlation coefficients (r) between the distance matrices of all ctDNA and of all mtDNA markers, and between those of ctDNA SSR and of mtDNA SSR alone, were separately calculated, and the Mantel test (Mantel, 1967) was performed using GenAlEx (Peakall and Smouse, 2001), by which rows and columns in the distance matrix were randomly permutated 1000 times to test whether the original correlation occurred by chance. As the distance matrices generated by all samples were used, correlation was considerably biased because there were some cases in which many samples shared the same haplotype (e.g., haplotype 148 shared in 109 accessions of S. tuberosum ssp. andigena, see Table 2). Thus, correlations were calculated between 164 haplotypes. The correlation between ctDNA and mtDNA differentiation was positive and significant (r = 0.226), but relatively poor when compared with that between ctDNA SSRs and nDNA RFLPs (r = 0.415, Sukhotu et al., 2004). The correlation between ctDNA SSRs and mtDNA SSRs was further lower (r = 0.147), which apparently reduce the applicability of mtDNA markers, particularly mtDNA SSRs to phylogenetic studies at least for a level of cultivated potatoes and their closely related wild species, though not excluding application to such a higher level of species groups as those studied by Scotti et al. (2007).

The distance matrices from ctDNA types, mtDNA types and combined types were separately subjected to the Neighbor-joining method using PAUP (portable version 4.0b10 for Unix) (Swofford, 2003) under the computer operating system Mac OS 10.4.11. The relationships between types were displayed visually as phylograms (Fig. 2). In our previous studies (Sukhotu et al., 2004; Sukhotu and Hosaka, 2006), there was a gap within the diversity range of nDNA among Andean cultivated and closely related wild species, by which various ctDNA types were correspondingly separated into two ctDNA groups; one with ctDNA RFLP types C, S and A, and the other with ctDNA RFLP types W and W2. These two distinct ctDNA groups were again depicted in this study (Fig. 2A). As expected by the relatively poor correlation, the mtDNA phylogram did not exhibit the corresponding two groups (Fig. 2B). Nevertheless, combined information displayed more clearly the two groups (Fig. 2C). Only exceptions were one S. irosinum accession (the haplotype identity number of 1 with W2-type ctDNA), one of two S. acroscopicum accessions (55 with W) and one of two S. multiinterruptum accessions (58 with W), all of which were grouped with those having ctDNA RFLP type C, S or A (Fig. 2C). Thus, it should be more strongly emphasized that there is a clear genetic differentiation between the species with ctDNA RFLP type C, S or A and those with ctDNA RFLP type W or W2, the former being mostly Peruvian species including all Andean cultivated species and their wild progenitors, while the latter being Bolivian or southern Andean species (Sukhotu and Hosaka, 2006).

We thank the US Potato Genebank at Sturgeon Bay, Wis., and the CIP gene bank at Lima, Peru, for providing Solanum materials. We also deeply thank Dr. T. Sukhotu for providing original ctDNA data. This study was partly supported by Calbee Potato Inc.