Characterization of Andes berry anthocyanins
Figure 1A is the HPLC–PDA chromatogram (518 nm) of Andes berry extract. Although peaks 1 and 2 were unresolved by PDA due to the coelution of anthocyanins, mass spectrometry analysis indicated the presence of 6 anthocyanins with unique elution times, whose extracted ion chromatograms are presented in Figure 2A to 2F. Peaks 1, 5, and 6 were minor peaks representing 5% of the total area while the 3 major peaks (2, 3, and 4) accounted for 40%, 10%, and 45% of the total area, respectively. Percentages were calculated as the individual MS area divided by the total MS area for the 6 anthocyanins. Assuming similar extinction coefficients for the various species, we found that MS predicted the PDA profile in a reliable manner.
Figure 1—. HPLC–PDA chromatogram of Rubus glaucus Benth detected at 518 nm. Peak 1, cyanidin 3-sambubioside; peak 2, cyanidin 3-glucoside; peak 3, cyanidin 3-xylorutinoside; peak 4, cyanidin 3-rutinoside; peak 5, pelargonidin 3-glucoside; peak 6, pelargonidin 3-rutinoside.
Download figure to PowerPoint
Figure 2—. Extracted ion chromatograms of anthocyanins present in Rubus glaucus Benth. (A) cyanidin 3-sambubioside, (B) cyanidin 3-glucoside, (C) cyanidin 3-xylorutinoside, (D) cyanidin 3-rutinoside, (E) pelargonidin 3-glucoside, (F) pelargonidin 3-rutinoside.
Download figure to PowerPoint
Mass spectral data of the anthocyanins in Andes berry are shown in Table 1. This information along with UV-visible spectral characteristics of the compounds was used to confirm their identity. The Abs440/Absλmax ratios for all compounds ranged from 0.31 to 0.49. These values are typical of 3-glycosylated anthocyanidins (Hong and Wrolstad 1990). In addition, only weak UV absorption was detected in the 310 to 320 nm region, which indicates no acylation with hydroxycinnamic acids (Hong and Wrolstad 1990).
Table 1—. Chromatographic and spectroscopic characteristics of anthocyanins detected in Andes berry (Rubus glaucus Benth).
|Peak||tr (min)||Anthocyanin assignment||MS/MS|
|Parent ion M+(m/z)||Fragment ions M+-X (m/z)|
|1||6.58||Cy 3-sambubioside||581.15||287.06 (M+-sam)|
|2||6.60||Cy 3-glucoside||449.11||287.06 (M+-glu)|
|3||6.77||Cy 3-xylorutinoside||727.21||287.06 (M+-xylrut), 581 (M+-rham), 433 (M+-rham-gluc)|
|4||6.94||Cy 3-rutinoside||595.16||287.06 (M+-rut), 449 (M+-rham)|
|5||7.16||Pg 3-glucoside||433.10||271.05 (M+-glu)|
|6||7.59||Pg 3-rutinoside||579.15||271.05 (M+-rut)|
Compound 1 was characterized by MS/MS analysis as cyanidin 3-sambubioside since it exhibited a parent ion of m/z 581 and a daughter ion at m/z 287 corresponding to the aglycon cyanidin. Cyanidin 3-sambubioside has been shown to fragment directly to the aglycone without evidence for intermediate fragments such as cyanidin 3-glucoside (m/z 449). This has been rationalized as due to high relative stability of the xylose–glucose glycosidic bond compared to the glucose to cyanidin bond, which is the preferred fragmentation site (Tian and others 2005b).
Peaks 2, 3, and 4 exhibited a λmax at 516, 521, and 517 nm, respectively, and were characterized as cyanidin derivatives. Peak 2, which was identified as cyanidin 3-glucoside exhibited a molecular cation of m/z 449 and a fragment ion at m/z 287 ([M-glucose]+). Compound 3 was characterized as cyanidin 3-xylorutinoside. It exhibited a parent ion of m/z a 727, a daughter ion of m/z 581 corresponding to the loss of rhamnose ([M- rhamnose]+), an aglycone at m/z 287 (cyanidin), and a fragment at m/z 433 corresponding to the loss of glucose and rhamnose ([M-rhamnose-glucose]+). This last fragment does not match any expected fragments after neutral loss of the constituent sugars as ([M-rhamnose]+) = 581 m/z, ([M-xylose]+) = 595 m/z, ([M-rhamnose-xylose]+) = 449 m/z, and ([M-rhamnose-xylose-glucose]+) = 287 m/z. It has been suggested that 433 m/z may represent cyanidin rhamnoside (Wu and Prior 2005), but considering the structure this fragment should not be possible unless the NMR analysis was in error. Recently, another report has emerged with the structural elucidation of cyanidin xylorutinoside in black raspberry by multidimensional NMR (Tulio and others 2008) and found consistent with previous reports that all 3 sugars are attached to one another with linkage to cyanidin via the glucose moiety. Based on this information, black raspberry can serve as a validated reference standard for cyanidin xylorutinoside and the other major cyanidin glycosides on our system.
Peak 4 was labeled as cyanidin 3-rutinoside. It exhibited a molecular cation of m/z 595, an aglycon of m/z 287 (cyanidin), and a daughter ion at m/z 449 corresponding to the loss of rhamnose ([M-rhamnose]+).
Peaks 5 and 6 were identified as pelargonidin 3-glucoside and pelargonidin 3-rutinoside. They had an absorbance spectrum with a λmax of 501 and 504 nm, respectively, and a pronounced shoulder in the 400 to 450 nm region typical of pelargonidin derivatives (Hong and Wrolstad 1990). Their mass spectrum contained parent ions of m/z 433 and 579, respectively, and both exhibited a fragment ion at m/z 271 (pelargonidin aglycon).
The major ACNs found for Rubus glaucus Benth (cyanidin 3-glucoside and cyanidin 3-rutinoside) are similar to those present in 18 different blackberry varieties from the United States, Mexico, Chile, France, and Macedonia as reported by Fan-Chiang and Wrolstad (2005). Cyanidin 3-glucoside and cyanidin 3-rutinoside were established as major anthocyanins in those varieties with cyanidin 3-glucoside representing between 43% and 95% of the total peak area and cyanidin 3-rutinoside ranging from trace to 53%. On the other hand, Tian and others (2005a) reported that Rubus occidentalis Jewel (black raspberry) ACNs were predominantly cyanidin rutinoside and cyanidin xylorutinoside while pelargonidin rutinoside was found and has been described as a minor anthocyanin in this commodity (Tian and others 2006).
Total phenolic content
The total phenolic content in the Andes berry extract was 294 ± 37.2 mg GAE/100g FW. This level is lower than that previously reported by Vasco and others (2008) (2167 mg GAE/100 g FW) for Andes berry harvested in the country of Ecuador. However, our results showed that Andes berry from Colombia had a similar level of phenolic content to those of Rubus idaeus (raspberry) and Rubus fruticosus (blackberry) (307 to 320 mg GAE/100 g FW) as reported by Costantino and others (1992), Rotundo and others (1998), and Benvenuti and others (2004). The variance in reported levels for Andes berry could be due to differing methods of extraction. We analyzed the total phenolic content in the extract obtained with pure acetone and acetone: water (70: 30) followed by partition with chloroform while Vasco and others (2008) analyzed the extract obtained first with methanol: water (50: 50, v/v) and then with acetone: water (70: 30, v/v). Our extraction method theoretically determines the maximal amount of ACNs, total phenolics, and antioxidant capacity present in a plant sample (Moyer and others 2002). However, methanol has been reported as better extracting solvent than acetone due to its polarity and good solubility for phenolic components from plant materials (Sobhy and others 2008). Another possible explanation for the difference in values is the variation in phenolic levels according to season and growing location (Connor and others 2005).
The antioxidant activity of the fruit extract determined as mmol TE/100 g FW using the ABTS·+ and FRAP assays was 2.01 ± 0.12 and 4.50 ± 1.22, respectively. Vasco and others (2008) reported values of 5.5 and 6.2 mmol TE/100 g FW for these assays, which is consistent with the higher phenolic content obtained in their study.
Although the ABTS·+ radical scavenging activity of Andes berry in our study is lower than that reported by Vasco and others (2008), it is within the range for other Rubus species (0 to 2.53 mmol TE/100 g FW), which have been recommended for the improvement of nutritional value due to their high antioxidant activities (Deighton and others 2000). According to these researchers, the degree of pigmentation and the phenolic content appear to be important factors in determining the antioxidant capacity of berries. These facts agree with our results that show lower ACN content, phenolic content, and antioxidant activity of Colombian Andes berry as compared to Andes berry from Ecuador. Various factors such as variety, growing condition, maturity, season, geographic origin, fertilizer, soil type, storage conditions, and amount of sunlight received, among others, might be responsible for the observed differences (Al-Farsi and others 2005). The FRAP value expressed as mmoles ferric iron reduced/100 g FW was 8.22 ± 1.50, which is close to the overall mean (7.92 mmoles ferric iron reduced/100 g FW) for 37 Rubus species and cultivars harvested in the state of Oregon, U.S.A. (Moyer and others 2002).
The ABTS method measures the ability of the antioxidant to quench ABTS·+ radicals probably by an electron transfer reaction (Dejian and others 2005) while the FRAP assay measures the potential of an antioxidant to reduce the yellow ferric–TPTZ complex to a blue ferrous–TPTZ complex by electrodonating substances under acidic conditions (Nilsson and others 2005). According to the results, Andes berry is a good electron donor as its extract was able to quench ABTS·+ radicals and reduce the ferric complex to a ferrous complex.