Abstract
Wild tomato species have been widely used for improvement of tomato disease resistance but have not been extensively explored for health-related traits. In this work, three interspecific populations derived from backcrosses between cultivated tomato and Solanum pimpinellifolium (LA1589), S. habrochaites (LA1223), and S. peruvianum (LA2172) were analyzed for water-soluble antioxidant activity, phenolic content, vitamin C content, and basic agronomic traits including fruit weight, shape, and color. The wild species accessions significantly exceeded S. lycopersicum for all three antioxidant traits with only one exception: vitamin C content in S. habrochaites LA1223. Several populations and traits showed transgressive segregation indicating that the backcross populations contained individuals with allele combinations that allowed antioxidant activity/content to exceed that of both parents. The S. habrochaites LA1223 population provided the best starting material for improvement of water-soluble antioxidant activity and phenolics content with 20% and 15% of the population, respectively, significantly exceeding the parental values for these traits. Moreover, the S. habrochaites population contained individuals that had nearly 2-fold more water-soluble antioxidant activity and phenolic content than cultivated tomato. The S. peruvianum LA2172 population was best for improvement of vitamin C content with 3-fold variation for the trait and individuals, which had twice as much vitamin C as cultivated tomato.
Tomato, Solanum lycopersicum (syn. Lycopersicon esculentum), is a member of the Solanaceae, a plant family of more than 3000 species including the economically important crops potato, eggplant, pepper, and tobacco. Tomato ranks second after potato as the most consumed vegetable of this family. Currently, tomato is grown in almost all countries worldwide. However, the origins of tomato are believed to be in western South America: Chile, Bolivia, Ecuador, and the coastal region of Peru (Razdan and Mattoo, 2007) where wild tomato species, including S. pimpinellifolium, S. pennellii, S. habrochaites, S. peruvianum, S. chmielewskii, and S. chilense, are still found (Peralta et al., 2006).
Wild tomato species are sources of both desired and undesired traits. S. peruvianum (syn. L. peruvianum) is genetically and morphologically the most diverse of the wild tomatoes (Peralta et al., 2005). It is so polymorphic that S. peruvianum was divided into varieties including dentatum (Dunal) and humifusum (Müller). S. peruvianum has green fruits, is typically self-incompatible, and allogamous. This species is also known for its virus, bacteria, fungi, aphid, and nematode resistances and has been widely used to introduce these traits into cultivated tomato (Robertson and Labate, 2007). S. habrochaites (syn. L. hirsutum) is another wild species used in the improvement and breeding of tomatoes. It has green fruits like S. peruvianum and is self-incompatible. S. habrochaites has a wide range of biotic and abiotic stress tolerances such as bacterial (Francis et al., 2001; Kabelka et al., 2002), fungal (Zhang et al., 2003), viral (Momotaz et al., 2005), insect (Musetti and Neal, 1997; Snyder et al., 1993), and cold (Veneema et al., 2000) resistances. Many of these resistances have been incorporated into cultivated tomato by backcross breeding (Hajjar and Hodgkin, 2007; Kalloo, 1991). S. pimpinellifolium (syn. L. pimpinellifolium) is a self-compatible tomato with red fruit. It is autogamous and facultatively allogamous. It is closely related to S. lycopersicum (Peralta and Spooner, 2007) and has been used in breeding to improve fruit color and quality (Causse et al., 2007). To date, most of the work using wild species for improvement of tomato has focused on biotic stress resistance. In comparison, only superficial and partial assessment of genetic variation for abiotic stress and nutritional factors such as antioxidant compound content has been made in wild tomato species (Causse et al., 2007).
Tomato is rich in various vitamins, antioxidants, and secondary metabolites (Adalid et al., 2004). The most abundant of these molecules are vitamin C, lycopene, β-carotene, and phenolics. These compounds are all known antioxidants, which are reported to be important for human health as well as a plant defense to abiotic and biotic stresses (Khachik et al., 2002). However, during domestication and later breeding, farmers and plant breeders focused on agronomically important traits of tomato such as fruit weight, color, shape, and disease resistance. Now that we know the genetic mechanisms of some of these phenotypic traits, it is possible to breed agricultural plants with more precision for what is needed. Often the best source for improvement of a crop plant is its wild species (Tanksley and McCouch, 1997). Wild species retain allelic (trait) diversity, which has been lost during domestication and breeding.
In this work, antioxidant and agronomic traits of three wild tomato species accessions, S. peruvianum LA2172, S. habrochaites LA1223, and S. pimpinellifolium LA1589, were compared with cultivated tomato. Antioxidant traits are of interest to consumers and plant breeders for their health-related contributions as well as their importance in plant resistance to biotic and abiotic stresses. The main aim of this work was to determine the potential of these wild species accessions for improvement of different traits. These results can be used in designing a breeding program to improve elite tomato lines for antioxidant characteristics.
Materials and Methods
The S. peruvianum advanced backcross BC2F2 population was obtained by crossing S. lycopersicum (TA496) as the cultivated recurrent parent with S. peruvianum (LA2172) as the wild donor parent. TA496 is the processing tomato line E6203 with the addition of tomato mosaic virus resistance. Because crossability barriers between S. peruvianum LA2172 and S. lycopersicum prevent normal development of F1 seeds, ovule culture was used to rescue the hybrids as described in Doganlar et al. (1997). F1 hybrids were then backcrossed to TA496 for two generations with no further need for ovule culture. Self-pollination of the BC2F1 families resulted in the BC2F2 families that were used in this work. Overall, the Peruvianum population contained 88 lines.
For the Habrochaites studies, a BC2F2 population was developed by crossing S. lycopersicum (TA1166) as the recurrent parent and S. habrochaites (LA1223) as the donor parent. TA1166 is a fresh-market inbred line (5TSS802 from Harris Moran). Population development was the same as for the Peruvianum population except that ovule culture was not necessary. The Habrochaites population contained 151 lines.
The Pimpinellifolium studies were performed with a BC2F9 population that was developed by crossing recurrent parent S. lycopersicum (TA209) and S. pimpinellifolium (LA1589), the wild donor parent. TA209 is processing line E6203 without virus resistance. One F1 hybrid was selected from the interspecific cross and backcrossed to TA209. A total of 27 BC1 plants were then selected based on presence of the sp locus on chromosome 6 so that all progeny plants had determinate growth. These selected individuals were backcrossed to TA209 to produce the BC2 population. BC2F1 individuals were then selfed for eight generations to produce the BC2F9 population. The Pimpinellifolium population contained 109 lines.
The Peruvianum population and parental lines were grown in the field at MultiTarim Seed Co., Antalya, Turkey, from late March to June 2006. The Habrochaites population and parents were also field-grown in Antalya from early April to July 2007. During the growing season, mean minimum temperature ranged from 11 to 22 °C, whereas mean maximum temperature ranged from 21 to 34 °C. Ten plants of each population and parental line were grown and were transplanted in double rows with 140-cm wide rows and 50-cm narrow row spacing. Double-row spacing is common in coastal regions and provides shade for fruit. Plants were spaced at 40 cm within rows. Soil was loamy and well-drained and fertilized with 15:15:15 fertilizer and 50 t composted manure per hectare before transplanting. Plants were drip-irrigated with fertigation (1.4 dS·m−1 electrical conductivity value) during the growing season (1:2:1 until first fruit set, 2:1:1 until first ripe fruit, 1:1:2 after first ripe fruit).
The Pimpinellifolium population was grown in the field at the Aegean Agricultural Research Institute, Izmir, Turkey, from late April to early Aug. 2006. During the growing season, mean minimum temperature ranged from 11 to 22 °C, whereas mean maximum temperature ranged from 21 to 33 °C. Ten plants of each population and parental line were grown in single rows with 140 cm between rows and 40 cm between plants within rows. The soil at the Izmir site was sandy loam and was fertilized with 30 kg·d−1 potassium sulfate, 25 kg·d–1 di-ammonium phosphate, and 10 kg·d−1 ammonium sulfate before transplanting. Plants were drip-irrigated with 30 kg·d−1 ammonium nitrate and 10 kg·d−1 potassium nitrate fertilizer. Plants were treated three times with 2% calcium nitrate to prevent blossom-end rot during fruit development. At both locations, fruit were harvested at the normal ripe market stage with samples bulked from the different plants for each line.
Total water-soluble antioxidant (WAOX) activity, total vitamin C (VITC) content, and total phenolic (PHE) content were determined for each line in each population. Ten ripe tomato fruits were selected from the harvested fruit. Care was taken to ensure that the selected tomatoes were of average size, color, and ripeness. The tomatoes were diced, mixed, and stored at –20 °C. All analyses were done within 6 months. For analysis, 100 g ripe tomato fruits were homogenized in 200 mL ice-cold distilled water for 2 min. For each 10 mL of mixture, 15 mL of ice-cold distilled water was added to the homogenate. The mixture was filtered using nylon cloth and filtrate was centrifuged at 3000 g at 4 °C for 20 min. The supernatant was prepared for antioxidant analyses by filtration through four layers of nylon cloth. WAOX measurements were performed as described by Re et al. (1999) using the ABTS cation decolorization assay and Trolox as a standard. WAOX activity was expressed as μmol Trolox/kg fresh weight. VITC was measured by the AOAC 967.21 titrimetric method as described by Augustin (1994) using commercial L-ascorbic acid as control and expressed as mg·kg−1 fresh weight. PHE was measured with the method developed by Singleton and Rossi (1965) and expressed as gallic acid equivalents (mg·kg−1 fresh weight). Each measurement was performed with three replicates. Averages of the replicates were taken and used for statistical analysis.
Fruit weight (FW) was determined by weighing at least 10 fruits per line. Fruit shape (FS), firmness (FIRM), external (EC) and internal (IC) color were scored by visual inspection of at least 10 ripe fruits per line. A scale of 1 to 5 was used for scoring with a different meaning for each trait (FS: 1 = round, 5 = elongated; EC: 1 = yellow or orange, 5 = dark red; FIRM was determined by manual squeezing of fruit (1 = very soft, 5 = very firm). IC was determined by examination of at least 10 transversely cut fruits. IC was scored with the same scale as EC.
Summary statistics of the data for each species were calculated. Means for quantitative data were compared with Student’s t test with a P value threshold of 0.05. All population values were then divided by the S. lycopersicum parental values to obtain normalized contents. Such normalization is a way to minimize variability in metabolite content resulting from environmental conditions (Perez-Fons et al., 2014). Histograms for antioxidant and agronomic traits measurements were drawn using the normalized data. Significant differences among bins in the histograms were determined by using the raw data for each bin and Dunnett’s method to compare each bin to the S. lycopersicum bin (P ≤ 0.05). Correlation tables were constructed using the raw data obtained from the measurements of fruits of each Solanum species. All statistical analysis was performed with the JMP 7.0 statistical program (SAS Institute, Cary, NC).
Results
Total water-soluble antioxidant activity.
WAOX activity of the S. peruvianum accession (8764 ± 110 μmol Trolox/kg) was highest followed by the S. pimpinellifolium (6989 ± 249 μmol Trolox/kg) and S. habrochaites (3925 ± 11 μmol Trolox/kg) accessions (Fig. 1). For all three species, WAOX activity of the interspecific populations showed continuous variation as expected for a quantitative trait. WAOX activity of S. peruvianum fruit was significantly higher than S. lycopersicum (closed arrow in histogram) with a 1.8-fold difference in values. The population mean was only slightly higher than the value for S. lycopersicum. The WAOX activity of the Peruvianum population showed a 1.8-fold range. In addition, 32% of the population had significantly higher WAOX activity than S. lycopersicum. WAOX activity of fruit from the S. habrochaites accession was significantly higher (1.5-fold) than the WAOX activity of S. lycopersicum. The WAOX activity of the Habrochaites population showed a 3.2-fold range. The population mean was closer to the wild parental mean than the mean of S. lycopersicum. In addition, 91% of the population had significantly higher WAOX activity than S. lycopersicum. Moreover, 20% of the population had significantly higher activity than both parents indicating substantial transgressive segregation for this trait. WAOX activity of fruit from the S. pimpinellifolium accession was 1.2-fold higher than for S. lycopersicum, a significant difference. The WAOX activity of the Pimpinellifolium population showed a 2-fold range. The mean of the population was closer to S. lycopersicum. Only 5% of the population had significantly higher activity than S. lycopersicum.
Water-soluble antioxidant activity for (A) S. peruvianum LA2172, (B) S. habrochaites LA1223, and (C) S. pimpinellifolium LA1589-derived populations relative to S. lycopersicum. Cultivated parent, wild-type parent, and mean of the population are shown with open, closed, and spherical arrows, respectively. Asterisks indicate bins with real values significantly different from S. lycopersicum as determined with Dunnett’s method for means comparison (P ≤ 0.05).
Citation: HortScience horts 49, 8; 10.21273/HORTSCI.49.8.1003
Phenolic content.
The PHE content of the S. peruvianum accession was highest (1619 ± 20 mg·kg−1) followed by the S. pimpinellifolium (844 ± 48 mg·kg−1) and S. habrochaites (303 ± 14 mg·kg−1) accessions (Fig. 2). PHE contents of the three populations were distributed as expected for a complex trait controlled by more than one gene. PHE content of the S. peruvianum accession was 3.2-fold higher than cultivated tomato, a significant difference. There was nearly 2-fold variation in PHE content in the population. Approximately 70% of the population had significantly higher PHE values than S. lycopersicum; however, no individuals exceeded the wild parent value. The S. habrochaites accession had significantly higher (1.5-fold) PHE content than S. lycopersicum (Fig. 2). The PHE content of the Habrochaites population showed a 3.2-fold range. The mean value of the PHE content for the population was intermediate between the parental lines. In addition, 15% of the population had significantly higher PHE content than both parental lines as a result of transgressive segregation. The PHE content of the S. pimpinellifolium accession was 1.6-fold higher than S. lycopersicum, a significant difference. Approximately 48% of the population showed significantly higher values than S. lycopersicum.
Phenolic content for (A) S. peruvianum LA2172, (B) S. habrochaites LA1223, and (C) S. pimpinellifolium LA1589-derived populations relative to S. lycopersicum. Cultivated parent, wild-type parent, and mean of the population are shown with open, closed, and spherical arrows, respectively. Asterisks indicate bins with real values significantly different from S. lycopersicum as determined with Dunnett’s method for means comparison (P ≤ 0.05).
Citation: HortScience horts 49, 8; 10.21273/HORTSCI.49.8.1003
Vitamin C content.
VITC content of the S. pimpinellifolium accession was highest (31.5 ± 8.0 mg·kg−1) followed by the S.peruvianum (27.1 ± 2.0 mg·kg−1) and S. habrochaites (16.0 ± 2.0 mg·kg−1) accessions (Fig. 3). All three populations showed continuous variation as expected. VITC content of the S. peruvianum accession was 1.1-fold higher than S. lycopersicum, a small but significant difference. In the population, VITC content showed more than 3-fold variation for the trait. In the Peruvianum population, 82% of the progeny had significantly higher VITC content than both parents. The VITC content of the S. habrochaites accession was 0.9-fold lower than the VITC content of S. lycopersicum. This was the only antioxidant trait–accession combination for which the wild species had a lower value than the cultivated type. The Habrochaites population showed distinct 4-fold variation. Approximately 80% of the population had significantly higher VITC content than both parents and 70% of the population had higher values than cultivated tomato. VITC content of the S. pimpinellifolium accession was only slightly higher than S. lycopersicum. The VITC content in the Pimpinellifolium population showed a 2.7-fold range in values. Overall, 37% of the population had significantly higher values than S. lycopersicum.
Vitamin C content for (A) S. peruvianum LA2172, (B) S. habrochaites LA1223, and (C) S. pimpinellifolium LA1589-derived populations relative to S. lycopersicum. Cultivated parent, wild-type parent, and mean of the population are shown with open, closed, and spherical arrows, respectively. Asterisks indicate bins with real values significantly different from S. lycopersicum as determined with Dunnett’s method for means comparison (P ≤ 0.05).
Citation: HortScience horts 49, 8; 10.21273/HORTSCI.49.8.1003
Fruit weight.
As expected, cultivated tomato fruits were largest, whereas the three wild species accessions all had significantly smaller fruit (ranging from 1.5 to 6.5 g) than the S. lycopersicum accessions. FW of S. lycopersicum was 21-fold higher than the S. peruvianum accession and there was a 2.5-fold range in values in the population (Fig. 4). There was greater variation for FW in the Habrochaites population. The two parental lines showed extremely different values for FW, a 40-fold difference. There was also great variation in the Pimpinellifolium population, a 4-fold range. Weight for S. lycopersicum was 48-fold higher than the S. pimpinellifolium accession. The Pimpinellifolium population was the only one that had individuals that exceeded S. lycopersicum for FW.
Fruit weights of (A) S. peruvianum LA2172, (B) S. habrochaites LA1223, and (C) S. pimpinellifolium LA1589-derived populations relative to S. lycopersicum. Cultivated parent, wild-type parent, and mean of the population are shown with open, closed, and spherical arrows, respectively. Asterisks indicate bins with real values significantly different from S. lycopersicum as determined with Dunnett’s method for means comparison (P ≤ 0.05).
Citation: HortScience horts 49, 8; 10.21273/HORTSCI.49.8.1003
Fruit shape.
FS of the S. lycopersicum and S. peruvianum accessions were different from each other. The S. lycopersicum accession had somewhat elongated fruit and the S. peruvianum accession had round fruit. There was good variation ranging between round and elongated types in the Peruvianum population (Supplemental Fig. 1A). Progeny were mostly intermediate between the parents and 21% of the population had rounder fruit than the average. In both S. habrochaites and S. lycopersicum parental lines, FS was spherical (Supplemental Fig. 1B). However, there was also variation ranging between round and elongated types in the Habrochaites population. Approximately 39% of the population had elongated fruit. For the Pimpinellifolium population, the S. lycopersicum accession fruits were somewhat elongated and S. pimpinellifolium fruits were round (Supplemental Fig. 1C). For the population, the range was between round and elongated. Approximately 11% of the population showed more elongated fruits than S. lycopersicum as a result of transgressive segregation.
Firmness.
The S. peruvianum accession had firmer fruit than S. lycopersicum (Supplemental Fig. 2A). In the Peruvianum population, 30% of the population had firmer fruit than S. lycopersicum. For the Habrochaites population, the S. habrochaites accession had similar firmness as cultivated tomato (Supplemental Fig. 2B), whereas a total of 36% of the population exceeded the firmness of the S. lycopersicum fruit. S. lycopersicum fruit were also firmer than S. pimpinellifolium fruit (Supplemental Fig. 2C). In addition, 65% of the population had better firmness than the S. lycopersicum accession.
External and internal color.
EC and IC were determined as a moderate red color for the S. lycopersicum accessions, whereas the S. peruvianum accession had green fruits. The mean EC and IC for the Peruvianum population were medium red (Supplemental Fig. 3A–B). There was a wide range of variation for both characters ranging from orange to dark red. A total of 57% of the population had darker red EC than cultivated tomato and 36% of the population exhibited higher IC than S. lycopersicum. The S. habrochaites accession had green IC and EC; however, none of its progeny had green fruit. Means for both IC and EC for this population were light red (Supplemental Fig. 3C–D). Although 17% of the Habrochaites population had better EC than S. lycopersicum, 13% showed better IC. The S. pimpinellifolium accession had darker red IC and EC than S. lycopersicum (Supplemental Fig. 3E–F). For IC, 6% of the population and for EC 29% of the population had higher values than S. lycopersicum.
Correlation of traits.
The correlation matrices in Table 1 show that each species exhibited a positive correlation between antioxidant traits. The highest correlation for Peruvianum population antioxidant traits was between WAOX activity and VITC content. However, this association was only moderate (r = 0.52). The Habrochaites population showed moderate correlations between WAOX activity and PHE content and between WAOX activity and VITC content. Among the three species, the Pimpinellifolium population had the highest correlations for antioxidant traits. The highest value (r = 0.60) was for the association between WAOX activity and PHE content with slightly lower values for the correlations between total antioxidant activity and VITC content and between PHE and VITC contents.
Antioxidant trait correlation matrices for Peruvianum, Habrochaites, and Pimpinellifolium populations.z
There were very few strong correlations between phenotypic traits (Table 2). The highest correlation in all species was found to be between IC and EC. Each species showed strong IC and EC correlations with the Habrochaites population having the highest value (r = 0.88) followed by the Peruvianum (r = 0.75) and Pimpinellifolium (r = 0.68) populations. Moderate correlations were observed between FW and FS in the Peruvianum and Pimpinellifolium populations. Interestingly, these two traits had a negative correlation in the Habrochaites population (r = –0.31). Antioxidant traits and FW tended to be negatively correlated.
Agronomic trait correlation matrices for Peruvianum, Habrochaites, and Pimpinellifolium populations.z
Discussion
Tomato is an economically and nutritionally important crop and is produced and consumed in large quantities around the world. The main goal of this project was to explore accessions of three wild Solanum species for improvement of antioxidant and agronomic traits in S. lycopersicum. For this purpose, three different interspecific populations of S. peruvianum, LA2172, S. habrochaites LA1223, and S. pimpinellifolium LA1589, were analyzed for antioxidant and agronomic traits. Antioxidants are substances that are capable of delaying or inhibiting oxidation processes caused by free radicals. Thus, antioxidant compounds effectively help protect organisms from oxidative stress-mediated damage (Percival, 1998). In humans and other animals, antioxidants are reported to decrease the risk of diseases that are associated with oxidative stress such as cardiovascular diseases, cancers, and neurological disorders (Podsedek, 2007; Yao et al., 2004). Moreover, antioxidants affect plant health. Studies showed that there is a correlation between tomato antioxidants and stress tolerance, especially salinity tolerance (Frary et al., 2010; Mittova et al., 2002, 2004). Therefore, development of tomato cultivars with high WAOX can affect both human and plant health.
The S. peruvianum, S. habrochaites, and S. pimpinellifolium accessions all had significantly higher antioxidant compound activities/contents than the S. lycopersicum lines with only one exception: VITC content in S. habrochaites LA1223. This difference between the cultivated and wild species accessions may be explained in two ways. First, domestication and breeding of S. lycopersicum involved selection for agronomic traits and may have resulted in the concomitant loss of favorable antioxidant alleles in cultivated tomato. Alternatively, because antioxidants can provide abiotic and biotic stress tolerance, alleles for higher levels of antioxidant compounds may have accumulated in wild species through natural selection and adaptation to the environment. When the wild species were compared among each other, S. peruvianum LA2172 had the highest WAOX activity and PHE contents. Despite this, the mean and range of WAOX activities seen in the S. peruvianum population were not better than the other species’ populations. In fact, the S. habrochaites population showed the most range in values and higher proportions of positive transgressive segregation for all three antioxidant traits. These factors are important because they are indicators of the usefulness of each species for cultivated tomato improvement.
The results suggest that among the species accessions tested, S. habrochaites LA1223 should be selected as a donor parent for the development of elite lines with high WAOX activity because introduction of alleles from this species gave the best mean improvement in population WAOX activity and the greatest indication that new allele combinations could provide substantial transgressive segregation. Thus, the use of this accession would allow greater incremental improvement of the trait in cultivated tomato. Similarly, S. habrochaites LA1223 should be selected for improvement of PHE content in elite lines. Although introduction of alleles from S. habrochaites LA1223 did not give a higher mean PHE content than the other wild species, the magnitude of potential improvement as indicated by transgressive segregation was higher. Moreover, the mean VITC content in the Habrochaites population was higher than both parents, which could be explained by heterosis. S. peruvianum LA2172 alleles could be used to improve VITC content of cultivated tomato because the Peruvianum population had the most transgressive segregation and individuals with twice as much VITC as the cultivated tomato line.
There were moderate positive correlations between WAOX activity and PHE content, PHE content and VITC content, and WAOX activity and VITC content. VITC and PHE are water-soluble antioxidants; therefore, genes responsible for these traits would also be expected to increase WAOX activity. Previous studies also indicated similar positive correlations between PHE content and WAOX activity in tomato and pepper (Frary et al., 2008, 2010; Hanson et al., 2004; Rousseaux et al., 2005; Toor et al., 2006). A positive correlation between VITC content and PHE content was also described in pepper by Frary et al. (2008). Negative correlations were seen between antioxidant traits and FW as observed by Hanson et al. (2004). In general, previous work reported stronger correlations than the ones observed here. These weaker correlations may be the result of greater experimental error or more heterozygosity in the tested populations. This may also explain why the more homozygous population, Pimpinellifolium, had higher antioxidant trait correlations than the other populations. Interestingly, mapping of antioxidant-related quantitative trait loci indicated that waox6.1, vitc6.1, and phe6.1 were located on the same part of chromosome 6, whereas waox12.1, vitc12.1, and phe12.1 were colocalized on chromosome 12 (Okmen et al., 2011). Colocalization of the genes controlling these traits may also explain why these antioxidant characters are positively correlated.
Improvement of tomato cultivars for antioxidant characteristics cannot be at the expense of yield and quality traits. For this reason, we also examined these traits in the wild species populations. As expected, the FW of the population progeny was negatively affected by introduction of wild species alleles. Interestingly, the Pimpinellifolium population is the only one that showed transgressive segregation with some individuals having larger fruit than the S. lycopersicum parent. IC and EC were most negatively affected in the S. habrochaites population.
In summary, our work shows that wild tomato species harbor alleles that could be useful for improvement of antioxidant traits in cultivated tomato. In fact, some of the backcross progeny from the populations used in this study could be used for breeding of these traits. However, as known from more than 50 years of work with tomato wild species, care must be taken that yield and fruit quality traits are not adversely affected by linkage drag. Thus, breeding should be facilitated by use of marker-assisted selection. This is possible because antioxidant traits have been mapped in S. habrochaites (Okmen et al., 2011), S. piminellifolium (Tamer, 2010), and S. pennellii (Frary et al., 2010; Rousseaux et al., 2005) with strong indications that such loci locations are conserved across species (Frary et al., 2010; Okmen et al., 2011).
Literature Cited
Adalid, A., Roselló, S. & Nuez, F. 2004 Breeding tomatoes for their high nutritional value Recent Res. Dev. Plant Sci. 2 33 52
Augustin, J. 1994 Vitamin analysis, p. 249–260. In: Nielsen, S.S. (ed.). Introduction to the chemical analysis of foods. Jones and Bartlett Publishers Intl., Boston, MA
Causse, M., Damidaux, R. & Rousselle, P. 2007 Traditional and enhanced breeding for quality traits in tomato, p. 153–192. In: Razdan, M. and A. Mattoo (eds.). Genetic improvement of Solanaceous crops: Volume 2: Tomato. Science Publishers, Inc., Enfield, UK
Doganlar, S., Frary, A. & Tanksley, S.D. 1997 Production of interspecific F1 hybrids, BC1, BC2 and BC3 populations between Lycopersicon esculentum and two accessions of L. peruvianum carrying new root-knot nematode resistance genes Euphytica 95 203 207
Francis, D.M., Kabelka, E., Bell, J., Franchino, B. & St. Clair, D. 2001 Resistance to bacterial canker in tomato (Lycopersicon hirsutum LA407) and its progeny derived from crosses to L. esculentum Plant Dis. 85 1171 1176
Frary, A., Göl, D., Keleş, D., Ökmen, B., Pınar, H., Şığva, H.Ö., Yemenicioğlu, A. & Doğanlar, S. 2010 Salt tolerance in Solanum pennellii: Antioxidant response and related QTL BMC Plant Biol. 10 58
Frary, A., Keçeli, M.A., Ökmen, B., Şığva, H.Ö., Yemenicioğlu, A. & Doğanlar, S. 2008 Water-soluble antioxidant potential of Turkish pepper cultivars HortScience 43 631 636
Hajjar, R. & Hodgkin, T. 2007 The use of wild relatives in crop improvement: A survey of developments over the last 20 years Euphytica 156 1 13
Hanson, P.M., Yang, R.-Y., Wu, J., Chen, J.-T., Ledesma, D., Tsou, S.C. & Lee, T.-C. 2004 Variation for antioxidant activity and antioxidants in tomato J. Amer. Soc. Hort. Sci. 129 704 711
Kabelka, E., Franchino, B. & Francis, D. 2002 Two loci from Lycopersicon hirsutum LA407 confer resistance to strains of Clavibacter michiganensis subsp. michiganensis Phytopathology 92 504 510
Kalloo, G. 1991 Breeding for environmental resistance in tomato, p. 153–165. In: Kalloo, G. (ed.). Genetic improvement of tomato. Springer-Verlag, Berlin, Heidelberg, Germany
Khachik, F., Carvalho, L., Bernstein, P.S., Muir, G.J., Zhao, D.-Y. & Katz, N.B. 2002 Chemistry, distribution, and metabolism of tomato carotenoids and their impact on human health Exp. Biol. Med. 227 845 851
Mittova, V., Guy, M., Tal, M. & Volokita, M. 2002 Response of the cultivated tomato and its wild salt-tolerant relative Lycopersicon pennellii to salt-dependent oxidative stress: Increased activities of antioxidant enzymes in root plastids Free Radic. Res. 36 195 202
Mittova, V., Guy, M., Tal, M. & Volokita, M. 2004 Salinity upregulates the antioxidative system in root mitochondria and peroxisomes of the wild salt-tolerant tomato species Lycopersicon pennellii J. Expt. Bot. 55 1105 1113
Momotaz, A., Scott, J.W. & Schuster, D.J. 2005 Searching for silverleaf whitefly and begomovirus resistance genes from Lycopersicon hirsutum accession LA1777 Acta Hort. 695 417 422
Musetti, L. & Neal, J.J. 1997 Resistance to the pink potato aphid, Macrosiphum euphorbiae, in two accessions of Lycopersicon hirsutum f. glabratum Entomol. Exp. Appl. 84 137 146
Okmen, B., Sigva, H.O., Gurbuz, N., Ulger, M., Frary, A. & Doganlar, S. 2011 Quantitative trait loci (QTL) analysis for antioxidant and agronomically important traits in tomato (Lycopersicon esculentum) Turk. J. Agr. For. 35 501 514
Peralta, I.E., Knapp, S. & Spooner, D.M. 2006 Nomenclature for wild and cultivated tomatoes TGC Rpt. 56 6 12
Peralta, I.E., Knapp, S., Spooner, D.M. & Lammers, T.G. 2005 New species of wild tomatoes (Solanum section Lycopersicon: Solanaceae) from Northern Peru Syst. Bot. 30 424 434
Peralta, I.E. & Spooner, D.M. 2007 History, origin and early cultivation of tomato (Solanaceae), p. 1–27. In: Razdan, M. and A. Mattoo (eds.). Genetic improvement of Solanaceous crops: Volume 2: Tomato. Science Publishers, Inc., Enfield, UK
Percival, M. 1998 Antioxidants. Clinical Nutr. Insights 31:1–4
Perez-Fons, L., Wells, T., Corol, D.I., Ward, J.L., Gerrish, C., Beale, M.H., Seymour, G.B., Bramley, P.M. & Fraser, P.D. 2014 A genome-wide metabolomics resource for tomato fruit from S. pennellii Nature Scientific Reports 4 3859
Podsedek, A. 2007 Natural antioxidants and antioxidant capacity of Brassica vegetables: A review LWT-Food Sci. and Technol. 40 1 11
Razdan, M. & Mattoo, A. 2007 Genetic improvement of Solanaceous crops: Volume 2: Tomato. Science Publishers, Inc., Enfield, UK
Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M. & Rice-Evans, C. 1999 Antioxidant activity applying an improved ABTS radical cation decolorization assay Free Radic. Biol. Med. 26 1231 1237
Robertson, L.D. & Labate, J.A. 2007 Genetic resources of tomato (Lycopersicon esculentum Mill.) and wild relatives, p. 25–75. In: Razdan, M. and A. Mattoo (eds.). Genetic improvement of Solanaceous crops: Volume 2: Tomato. Science Publishers, Inc., Enfield, UK
Rousseaux, M.C., Jones, C.M., Adams, D., Chetelat, R., Bennett, A. & Powell, A. 2005 QTL analysis of fruit antioxidants in tomato using Lycopersicon pennellii introgression lines Theor. Appl. Genet. 111 1396 1408
Singleton, V. & Rossi, J.A. 1965 Colorimetry of total phenolics with phosphomolybdic-phosphotungstic acid reagents Amer. J. Enol. Viticult. 16 144 158
Snyder, J.C., Guo, Z., Thacker, R., Goodman, J.P. & Pyrek, J.S. 1993 2, 3-Dihydrofarnesoic acid, a unique terpene from trichomes of Lycopersicon hirsutum, repels spider mites J. Chem. Ecol. 19 2981 2997
Tamer, N. 2010 QTL analysis for fruit traits in Solanum pimpinellifolium inbred backcross lines. MS thesis, Izmir Inst. of Technol., İzmir, Turkey
Tanksley, S.D. & McCouch, S.R. 1997 Seed banks and molecular maps: Unlocking genetic potential from the wild Science 277 1063 1066
Toor, R., Savage, G. & Lister, C. 2006 Seasonal variations in the antioxidant composition of greenhouse grown tomatoes J. Food Compos. Anal. 19 1 10
Veneema, J.H., Eekhof, M. & van Hasselt, P.R. 2000 Analysis of low-temperature tolerance of a tomato (Lycopersicon esculentum) cybrid with chloroplasts from a more chilling-tolerant L. hirsutum accession Ann. Bot. (Lond.) 85 799 807
Yao, L.H., Jiang, Y., Shi, J., Tomas-Barberan, F., Datta, N., Singanusong, R. & Chen, S. 2004 Flavonoids in food and their health benefits Plant Foods Hum. Nutr. 59 113 122
Zhang, L., Lin, G., Nino-Liu, D. & Foolad, M. 2003 Mapping QTLs conferring early blight (Alternaria solani) resistance in a Lycopersicon esculentum × L. hirsutum cross by selective genotyping Mol. Breed. 12 3 19