Experimental and discussion

Peanut or groundnut (Arachis hypogaea L.) is an allotetraploid species (2n = 4x = 40) and one of the five most important oilseed crops in the world with a production of about 32 million metric tons of seeds (http://www.usda.gov/nass/pubs/agstas.htm). Peanut seeds contain 44–56% oil and 22–30% protein (Pancholy and Despande, Reference Pancholy, Despande and Krall1978), and are therefore mainly used for edible oil production and as a high-protein food for human consumption, especially in developing countries where people have limited access to protein sources. Peanut seeds also contain useful phytochemicals, such as trans-resveratrol, which can contribute to human health (Jang et al., Reference Jang, Cai, Udeani, Slowing, Thomas, Beecher, Fong, Farnsworth, Kinghorn, Metha, Moon and Pezzuto1997). trans-Resveratrol (3,4′,5-trihydroxystilbene) is a polyphenolic compound initially classified as a phytoalexin (Ingham, Reference Ingham1976) due to its antifungal activity. trans-Resveratrol has recently drawn a great amount of attention from nutraceutical and pharmaceutical companies due to its antioxidant, anti-inflammatory and anticancer activities, as well as chemopreventive, cardioprotective and estrogenic effects (Baur and Sinclair, Reference Baur and Sinclair2006). Although resveratrol has been identified in peanut seeds, variation in resveratrol content of peanut germplasm accessions was unknown. The objectives of this study were to (1) determine the concentration and variability of resveratrol in peanut accessions of the USDA germplasm collection by high-performance liquid chromatography (HPLC) and (2) detect whether there is a correlation between resveratrol content and seed weight in peanut.

Twenty accessions were selected from the USDA peanut germplasm collection (http://www.ars-grin.gov/npgs/) and their botanical variety type and collection site are listed in Table 1. Seeds from each accession were planted in a 10-foot-long plot at Byron Georgia in 2007. There were no major diseases observed in the field. Matured pods were harvested, air-dried, shelled by hand and then the seeds were stored at 4°C until analysis. Dried seeds were counted, weighed and recorded in two replicates. A standard curve was generated with a trans-resveratrol reference purchased from Sigma (St Louis, MO, USA). Under bright light, trans-resveratrol could easily convert to cis-resveratrol. trans-Resveratrol was extracted by following the procedures described by Sanders et al. (Reference Sanders, McMichael and Hendrix2000), which were performed under yellow light. Approximately, 8 g of air-dried seeds were ground to a fine powder in a coffee blender. Ground seed tissues (3 g) were transferred into 50 ml Falcon tubes and homogenized with 9 ml of 80% ethanol using a Power Gen 125 homogenizer (Fisher Scientific, Loughborough, UK). The homogenized samples were centrifuged (Eppendorf, 5415D, Hamburg, Germany) at 12,000 g for 3 min. Two millilitres of the supernatant were taken and cleaned by solid-phase extraction using Poly-Prep chromatography column (0.8 cm × 4 cm, Bio-Rad) packed with ~1 ml mixture (1:1 w/w) of Al₂O₃ (EM Industries Inc., Gibbstown, NJ, USA) and silica gel 60 RP-18 (EMD Chemicals Inc., Hawthorne, NY, USA). The packed column was conditioned with 80% ethanol. The supernatant was applied to the equilibrated column and the effluent was collected into a 4 ml vial. The column was washed with an additional 2 ml of 80% ethanol and the effluent was collected into the same vial. The collected solvent was evaporated at 50°C to dryness with a nitrogen gas stream. The extracted compounds were dissolved in 1 ml of 20% acetonitrile and filtered (at 0.45 μm filter) prior to injection for HPLC analysis. Separation of metabolites was performed on RP-HPLC system (Agilent 1100 series) using a C₁₈ column (4.6 mm × 150 mm, 5 μm; Agilent Technologies, Santa Clara, CA, USA) at 40°C with a binary pump and autosampler. The mobile phase consisted of (A) filtered sterile water containing 0.1% formic acid at pH 2.5 and (B) HPLC-grade acetonitrile. The flow rate was 1.5 ml/min with the following gradient: 10% B for 2 min, 10–30% B for 8 min, 30% B for 1 min, followed by column wash at 95% B for 6 min and 10% B for 9 min before next injection. The volume for sample injection was 30 μl and the analytes were monitored with a diode array detector at 310 nm absorbance. trans-Resveratrol in the extract of each accession was quantified at 310 nm by reference to the peak area of an external authentic standard of resveratrol. Two replicates were conducted for determination of trans-resveratrol concentration. For each replicate, two sample extractions were conducted after grinding for data collection. The average from two extractions per sample was used for data analysis. One-way analysis of variance was conducted using statistical analysis system (SAS OnlineDoc^® 9.1.3, 2004) to analyse the data, and Fisher's protected least significant difference (LSD) test was used to separate means. Pearson's correlation coefficient analysis was performed to determine the interrelationship between resveratrol concentration and seed weight.

Table 1 Selected accessions, seed weight and resveratrol content

* Gram per 100 seeds;

^† a–d: if the letters are the same after values, there is no significant difference.

As an example, compiled chromatograms for detecting the amounts of trans-resveratrol from three peanut varieties using HPLC are shown in Fig. 1. The arrow-marked peak areas correspond to the amounts of trans-resveratrol detected. The accessions of PI 502043, PI 149641 and PI 520600 produced peaks with a small, medium and large area, respectively. In these single extractions, 0.09, 0.70 and 1.46 μg/g of trans-resveratrol, respectively, were detected for these three accessions (Fig. 1). Variation for trans-resveratrol from peanut accessions is statistically significant at P < 0.0001. The variation from replicates was not statistically significant at P < 0.05. Therefore, the variability of trans-resveratrol content detected by HPLC mainly came from differences in the peanut varieties. trans-Resveratrol contents in peanut seeds are summarized in Table 1. Among 20 accessions quantified, the seeds of PI 520600 (TAMRUN 88, a cultivar developed in Texas, USA) contained a significantly higher amount of trans-resveratrol (1.626 μg/g labelled as ‘a’ in Table 1) than the other 19 accessions. The seeds from PI 501297, PI 502043, PI 506419, PI 565448 and PI 629027 contained a significantly lower amount of trans-resveratrol (0.177, 0.125, 0.181, 0.207 and 0.215 μg/g labelled as ‘d’ in Table 1) than the accessions of PI 428261, PI 520600 and PI 632380 (1.088, 1.626 and 1.058 μg/g labelled as ‘ab, a and abc’, respectively, in Table 1). All three accessions (PI 428261, PI 520600 and PI 632380) that contained a high amount of trans-resveratrol in the seeds were from A. hypogaea L. var. hypogaea. The amount of trans-resveratrol in the seeds of Florunner, a released cultivar developed in Florida, USA, had been quantified in this and other studies (Sobolev and Cole, Reference Sobolev and Cole1999). The results from these two separate studies are comparable (0.207 μg/g from this study and 0.273 μg/g averaged from Sobolev and Cole's study).

Fig. 1 Metabolite chromatograms from the peanut extracts analysed by HPLC. Arrows indicate trans-resveratrol peaks in the analysed extracts.

The weights for 100 seeds of the analysed accessions are listed in Table 1. A statistically significant variation in 100 seed weights was observed at P < 0.0001. The average weight for 100 seeds was 52.84 g, ranging from 22.30 to 87.94 g (at least a four times difference). From our study, there was no significant correlation between trans-resveratrol content and seed weight observed. The results from a previous report (Sobolev and Cole, Reference Sobolev and Cole1999) showed decreasing trans-resveratrol with increasing seed size. Seed weight and size are not interchangeable, but a larger seed usually has a higher seed weight. The trend for decreasing trans-resveratrol with increasing seed size was not observed in the correlation between resveratrol content and seed weight. The inconsistent results from these two studies may be explained by the following reason. Seed weights in our study were from different accessions and reflected the difference from genotypes, whereas seed sizes in their study were from one accession (Florunner) and mainly reflected the difference in seed maturity (i.e. mature seeds had a larger size whereas immature seeds had a smaller size).

Significant variability was observed in the amounts of resveratrol among seeds of 20 peanut accessions. There is a potential to increase the trans-resveratrol amount in peanut cultivars. trans-Resveratrol could be added as a selection trait in breeding programmes to develop high resveratrol peanut cultivars. Furthermore, high resveratrol peanuts should be considered for use in peanut product processing. Since consumption of peanuts and/or peanut products containing a high amount of resveratrol can contribute to human health, developing high resveratrol peanut cultivars may eventually benefit peanut farmers and processors as well as consumers.