Compositional Diversity among Blackcurrant ( Ribes nigrum) Cultivars Originating from European Countries.

Berries representing 21 cultivars of blackcurrant were analyzed using liquid chromatographic, gas chromatographic, and mass spectrometric methods coupled with multivariate models. This study pinpointed compositional variation among cultivars of different origins cultivated in the same location during two seasons. The chemical profiles of blackcurrants varied significantly among cultivars and growing years. The key differences among cultivars of Scottish, Lithuanian, and Finnish origins were in the contents of phenolic acids (23 vs 16 vs 19 mg/100 g on average, respectively), mainly as 5- O-caffeoylquinic acid, 4- O-coumaroylglucose, ( E)-coumaroyloxymethylene-glucopyranosyloxy-( Z)-butenenitrile, and 1- O-feruloylglucose. The Scottish cultivars were grouped on the basis of the 3- O-glycosides of delphinidin and cyanidin, as were the Lithuanian cultivars. Among the Finnish samples, the content of myricetin 3- O-glycosides, 4- O-caffeoylglucose, 1- O-coumaroylglucose, and 4- O-coumaroylglucose were significantly different between the two green-fruited cultivars and the black-fruited cultivars. The samples from the studied years differed in the content of phenolic acid derivatives, quercetin glycosides, monosaccharides, and citric acid.

Introduction

Horticultural plants have been used for food, fiber, biofuel, medicine, and other products to sustain and enhance human life in the recent years.[1−3] As a species of family Grossulariaceae, blackcurrants (Ribes nigrum) are a rich source of bioactive metabolites and flavor compounds, including sugars, acids, and phenolic compounds.[4−7] Some of the compounds have significant physiological effects on maintenance of cardiovascular health, restriction of cancer growth, control of blood glucose levels, and other physiological functions in in vitro models.[8,9] This leads to a commercial exploitation of blackcurrant as food products and nutritional supplements.

The contents and profiles of bioactive metabolites and flavor compounds are not present constantly in R. nigrum berries. A number of previous studies have confirmed that environmental factors affect the chemical composition.[10−12] For example, in the berries of the cultivar ‘Vertti’, the concentration of phenolic compounds, especially the conjugates of hydroxycinnamic acids, was dependent on the latitude of the growing site.[10] Strong correlation with temperature and radiation was found in the content of some phenolic compounds such as delphinidin 3-O-glucoside, delphinidin 3-O-rutinoside, and myricetin-3-O-glucoside in ‘Melalahti’, ‘Mortti’, and ‘Ola’.[11] The genotype is another major factor influencing chemical profile of R. nigrum berries. Vagiri et al. studied blackcurrant berries of Scottish, Swedish, and Russian origins, revealing large variations in polyphenols, ascorbic acid, and soluble sugars among the genotypes.[13] Mikulic-Petkovsek found that the contents of acids, sugars, and main groups of phenolics varied significantly during fruit ripening among blackcurrant cultivars ‘Rosenthal’, ‘Tenah’, and ‘Titania’.[14] Similarly, differences were observed among the cultivars ‘Titania’, ‘Triton’, ‘Tsema’, and ‘Cacanska crna’.[15] The impact of genotype can also be seen in juice processing where the juice produced from a single cultivar maintains its typical sensory characteristics during the process.[6]

Due to extensive industrial demand, new cultivars of blackcurrant are always requested. The main goal of commercial breeding of new cultivars usually focuses on the adaptation of plants to abiotic and biotic environment, as well as their cropping potential.[16,17] For breeders, breeding is a long and exacting work, making use of previous breeding results and even the achievements of the previous breeders’ generations. It is thus of high importance for breeders to have better knowledge about the fruit quality characteristics of cultivars. Since some common ancestors are typically used in cultivar development, it is even possible that there are some limitations related to fruit quality in breeding populations.

Likewise, food industry needs fruits with specific properties to meet the requirement of processing or to reach the target quality of final products. These may not be achieved by using cultivars traditionally grown by its raw material producers. The chemical composition of blackcurrant fruits has been traditionally less emphasized when new cultivars are selected. The previous studies on chemical profiles of blackcurrants have focused on either a limited number of compounds or only a few cultivars. Therefore, it is necessary to investigate systematically the compositional difference among a collection of blackcurrant cultivars that are bred and cultivated in different countries. The results of our study provide new knowledge to help breeders, trade, and food industry to ensure success in providing targeted quality of blackcurrant fruit and fruit products.

In this study, we investigated and compared the composition of 21 cultivars of blackcurrants originating from five different countries. All cultivars were planted in 2009 at the same location and treated with the same cultivating practice. Samples collected during two consecutive years were analyzed in order to get an idea of the possible seasonal variation. The variations in the compositions of various phenolic compounds, simple sugars, and acids were determined using liquid chromatographic (LC), gas chromatographic (GC), and mass spectrometric (MS) methods, followed by comparison of data sets with partial least-squares (PLS) regression models. Our aim was to pinpoint the main groups or individual compounds separating different cultivars and origins. This knowledge will assist plant breeding as well as providing guidance for food industry in selection of raw materials and farmers in selecting cultivars.

Materials and Methods

Materials

Blackcurrant cultivars originating from Scotland (9 cultivars), Lithuania (5 cultivars), Latvia (1 cultivar), Poland (1 cultivar), and Finland (5 cultivars) were cultivated in the test site of Natural Resources Institute (Luke) in Piikkiö, Kaarina, southwest Finland (latitude 60°23′ N, longitude 22°33′ E, altitude ca. 5 m). The propagation material of plants was provided by the breeding institute of each cultivar, to guarantee the true-to-type of the cultivars. One-year old transplants were planted in 2009 in three rows, with a distance of 4 m between rows and 1 m between plants within a row. Two plants of each cultivar were planted in a plot randomized in each row. Berries for the analyses were sampled from the total harvest of each plot in August 2014 and 2015, each representing one replicate sample of each cultivar. The soil was silt moraine rich in organic matter. Irrigation via a trickle tape, fertilization, and other cultivation methods were according to the Finnish standard guidelines.[18] The harvesting time of each cultivar was defined by experienced horticulturist, the definition being based on the color, flavor, and structure of optimally ripe berries. The samples collected in year 2014 were first stored in a freezer at −70 °C for 1 year and then transferred to −20 °C together with the samples harvest in 2015 for 5 months. All frozen samples were then delivered from Luke to University of Turku and stored at −20 °C for a maximum period of 15 months until all analyses were completed. The information on the samples is shown in Supplemental Table 1, including cultivar names, origin countries, and harvesting dates.

Weather Conditions

Data on climatic conditions were collected in the meteorological station in the Luke Kaarina test site and provided by the Finnish Meteorological Institute (Helsinki, Finland), to give information on the climatic differences between the fruit ripening periods of the two years. The main climatic factors with the one-month time interval during July 20–August 20 are shown in Supplemental Table 4. The time interval was chosen to cover the harvesting period of all cultivars in 2014 and all but two very late cultivars in 2015, and at least 12 days preceding the earliest harvest date.

Dry Matter Content

Approximately 5 g of currant samples were weighed accurately and cut with a blade in a watch glass. The residue on the blade was rinsed into the watch glass with Milli-Q water. The samples were dried in the oven (Oy Santasalo-Sohlberg Ab, Helsinki, Finland) at 105 °C overnight until their weights reached a constant value.

Phenolic Compounds

Phenolic compounds were identified using a Waters Acquity Ultra performance liquid chromatography (UPLC) system equipped with a 2996 DAD detector, an electrospray ionization interface (ESI), and a Waters Quattro Premier mass spectrometer (Waters Corp., Milford, MA, U.S.A.). All phenolics were characterized by comparing LC retention time and typical mass fragments with reference compounds and literature.[6,19−29] Mass spectrometry was set in both negative- and positive-ion modes, the condition of which was reported in our previous study.[20]

Two methods were applied for analysis based on the types of phenolic compounds. For anthocyanins, 5 g of frozen berries were crushed into slurry and extracted with 15 mL of acidic methanol (MeOH/HCl 99:1), followed by ultrasonication (10 min) and centrifugation (4420g for 10 min). The extraction was carried out three times. The three supernatants were combined, and the total volume was set to 50 mL with acidic methanol. The samples were filtered through a 0.2 μm syringe filter before UPLC-DAD-ESI-MS analysis. The analysis of anthocyanins was conducted according to the method previously reported by Mäkilä and co-workers.[19] The signal of anthocyanins in the LC analyses was monitored at the wavelength of 520 nm.

Other phenolic compounds were extracted from crushed materials (15 g) with 10 mL of ethyl acetate. Ultrasonication (15 min) and centrifugation (4420g for 15 min) were applied in the four-time extraction. The combined supernatant was evaporated at 36 °C; the residue was dissolved with 3 mL of methanol and filtered through a 0.2 μm syringe filter. Liquid chromatographic separation was performed with a Phenomenex Aeris peptide XB-C18 column (150 × 4.60 mm, 3.6 μm, Torrance, CA, U.S.A.) at room temperature. The injection volume was 10 μL and the total flow was kept at 1 mL/min. The mobile phase was a combination of Milli-Q water (A) and acetonitrile (B), both containing 0.1% (v/v) of formic acid. The gradient applied was the following: 0–15 min with 8–10% solvent B, 15–20 min with 10–13% B, 20–25 min with 13–16% B, 25–30 min with 16–18% B, 30–35 min with 18–20% B, 35–40 min with 20–22% B, 40–45 min with 22–25% B, 45–50 min with 25–60% B, 50–55 min with 60–8% B, 55–57 min with 8% B. The chromatograms were recorded at three different wavelengths (360 nm for flavonols, 320 nm for phenolic acids, and 280 nm for flavan-3-ols and other phenolic compounds).

The quantification of the phenolics was performed using a Shimadzu LC-10AT liquid chromatograph system, coupled with a SPD-M20A VP photodiode array (Shimadzu Corp., Kyoto, Japan). The chromatographic conditions were the same as in the corresponding qualitative analyses. The concentration of the compounds identified was determined using an external standard method as described previously.[20] The compounds lacking corresponding reference standards were quantified by the calibration curves of compounds with closest structures. For instance, cyanidin 3-O-(6″-coumaroyl)-glucoside was quantified by the calibration curve of cyanidin 3-O-glucoside (y = 3 × 10–8x + 0.0026, R2 = 0.9990). The detailed information on external standards is given in Supplemental Table 6.

Sugars and Simple Organic Acids

Fifteen grams of frozen berries was crushed with a T25 digital Ultra-Turrax (IKA Werke GmbH & Co. KG, Staufen im Breisgau, Germany) and extracted with 10 mL of Milli-Q water at room temperature. The extraction was assisted with ultrasonication (15 min) and centrifugation (4420g for 15 min). After the supernatant was collected, the residue was extracted with the same procedure three times. The supernatants from the four times of extraction were combined and diluted with Milli-Q water to a final volume of 50 mL. Sugars and simple organic acids in the samples were analyzed as trimethylsilyl (TMS) derivatives by Shimadzu GC-2010 equipped with a flame ionization detector (FID) (Shimadzu Corp., Kyoto, Japan). The compounds were identified on the basis of the retention time of reference standards. A mixed internal standard, consisting of sorbitol (for sugars) and tartaric acid (for acids), was used for quantification. The methods for preparation of samples and standards, as well as gas chromatographic conditions, were the same as described in the previous research.[12]

Statistical Analyses

The quantitative analyses of chemical compounds were performed in triplicates. The results were calculated on the basis of dry weight (mg/g or 100 g of berries) and expressed as mean ± standard deviation (SD). Partial least-squares (PLS) regression with full cross validation was applied to determine the correlation between chemical profile and cultivar/country of origin/growing year by using Unscrambler 10.4 (Camo Process AS, Oslo, Norway). PLS models were established with the concentrations of compounds as the predictors (X-data) and the cultivars (and other factors listed above) as the responses (Y-data).

Results and Discussion

Altogether, 63 chemical compounds were identified from blackcurrant berries, primarily as anthocyanins (15 compounds), flavonols (19), flavan-3-ols (4), phenolic acid derivatives (14), organic acids (4) and sugars (6). The qualitative results and chromatographs are given in Table and Supplemental Figure 1, respectively. In accordance with previous study,[30] most of phenolic compounds present in blackcurrants were anthocyanins, flavonols, flavan-3-ols, and the derivatives of hydroxycinnamic acids (caffeic acid, coumaric acid, and ferulic acid). In addition to delphinidin and cyanidin derivatives as the dominant anthocyanins in the berries, the glycosides of petunidin (peaks 5 and 6), pelargonidin (peaks 8 and 9), peonidin (peaks 10 and 11), and malvidin (peaks 12 and 13) were detected and confirmed based on the typical MS fragmentations. These minor anthocyanins were not reported in previous studies.[22,23] Presence of anthocyanins was not the only difference between black and green cultivars. Some flavonols present in black cultivars were not found in the two green-fruited cultivars, such as myricetin 3-O-arabinoside (peak 19), quercetin 3-O-galactoside (peak 22), quercetin 3-O-arabinoside (peak 24), isorhamnetin 3-O-(6″-malonyl)-galactoside (peak 31), myricetin-hexoside-deoxyhexoside (peak 32), and myricetin aglycone (peak 29). Organic acids in blackcurrants were characterized as malic acid, citric acid, quinic acid, and ascorbic acid. The main sugars in blackcurrants were fructose, glucose, and sucrose.

Table 1

Identification of Phenolic Compounds, Organic Acids, and Sugars in Blackcurrant (Ribes nigrum) Cultivars

no.a	tentative identificationb	abbreviationc	UV λmax (nm)	[M–H]−/[M + H]+ (m/z)	[A–H]−/[A+H]+ and other ions(m/z)	identification by
	Anthocyanins
1	delphinidin 3-O-glucoside	De-Glu	276, 524	463/-	301/-	MS, standard, and literature[6,20−23]
2	delphinidin 3-O-rutinoside	De-Rut	276, 525	609/-	301/-	MS and literature[6,20−23]
3	cyanidin 3-O-glucoside	Cy-Glu	280, 516	447/-	285/-	MS, standard, and literature[6,20−23]
4	cyanidin 3-O-rutinoside	Cy-Rut	280, 517	593/-	285/-	MS, standard, and literature[6,20−23]
5	petunidin 3-O-glucoside	Pt-Glu	276, 527	477/-	315/-	MS and literature[20−23]
6	petunidin 3-O-rutinoside	Pt-Rut	276, 527	623/-	315/-	MS and literature[20−23]
7	cyanidin 3-O-arabinoside	Cy-Ara	280, 516	417/-	285/-	MS and literature[20−23]
8	pelargonidin 3-O-glucoside	Pl-Glu	278, 525	431/-	269/-	MS and literature[21−23]
9	pelargonidin 3-O-rutinoside	Pl-Rut	278, 525	577/-	269/-	MS and literature[21−23]
10	peonidin 3-O-glucoside	Po-Glu	280, 517	461/-	299/-	MS and literature[20−23]
11	peonidin 3-O-rutinoside	Po-Rut	280, 517	607/-	299/-	MS and literature[20−23]
12	malvidin 3-O-glucoside	Ma-Glu	281, 522	491/-	329/-	MS and literature[20−23]
13	malvidin 3-O-rutinoside	Ma-Rut	281, 522	637/-	329/-	MS and literature[20−23]
14	delphinidin 3-O-(6″-coumaroyl)-glucoside	De-coGlu	280, 530	609/-	447, 301/-	MS and literature[20−23]
15	cyanidin 3-O-(6″-coumaroyl)-glucoside	Cy-coGlu	280, 524	593/-	447, 285/-	MS and literature[20−23]
	Flavonols
16	myricetin 3-O-rutinoside	My-Rut	255, 265(sh), 355	625/627	317/481, 319	MS, standard, and literature[19,20,24,25]
17	myricetin 3-O-galactoside	My-Gal	255, 265(sh), 355	479/481	317/319	MS, standard, and literature[19,20,24]
18	myricetin 3-O-glucoside	My-Glu	255, 265(sh), 355	479/481	317/319	MS, standard, and literature[19,20,24]
19	myricetin 3-O-arabinoside	My-Ara	255, 265(sh), 355	449/451	317/319	MS and literature[19,20,24,25]
20	myricetin 3-O-(6″-malonyl)-galactoside	My-maGal	256, 266(sh), 356	565/567	521,317/319	MS and literature[19,20]
21	quercetin 3-O-rutinoside	Qu-Rut	255, 265(sh), 355	609/611	301/465, 303	MS, standard, and literature[19,20,24,25]
22	quercetin 3-O-galactoside	Qu-Gal	255, 265(sh), 355	463/465	301/303	MS, standard, and literature[19,20,24]
23	quercetin 3-O-glucoside	Qu-Glu	255, 265(sh), 355	463/465	301/303	MS, standard, and literature[19,20,24,25]
24	quercetin 3-O-arabinoside	Qu-Ara	255, 266(sh), 355	433/435	301/303	MS and literature[20]
25	quercetin 3-O-(6″-malonyl)-glucoside	Qu-maGlu	256, 266(sh), 356	549/551	505,301/303	MS and literature[19,20,24,25]
26	kaempferol 3-O-rutinoside	Ka-Rut	266, 346	593/595	285/449, 287	MS and literature[19,20,24,25]
27	kaempferol 3-O-galactoside	Ka-Gal	266, 346	447/449	285/287	MS and literature[20,25]
28	isorhamnetin 3-O-glucoside	Is-Glu	256, 265(sh), 354	477/479	315/317	MS, standard, and literature[19,20,24]
29	myricetin aglycone	My agly	255, 266(sh), 370	317/319		MS and literature[19,20,24]
30	kaempferol 3-O-(6″-malonyl)-glucoside	Ka-maGlu	265, 465	533/535	489, 285/287	MS and literature[19,20]
31	isorhamnetin 3-O-(6″-malonyl)-galactoside	Is-maGal	256, 265(sh), 355	563/565	519, 315/317	MS and literature[19,20,24]
32	myricetin-hexoside-deoxyhexoside	My-hex-deox	255, 268(sh), 356	625/627	317/319	MS and literature[20]
33	isorhamnetin 3-O-(6″-malonyl)-glucoside	Is-maGlu	256, 265(sh), 355	563/565	519, 315/317	MS and literature[19,20]
34	quercetin aglycone	Qu agly	274, 368	301/303		MS and literature[19,20,24]
	Phenolic Acid Derivatives
35	5-O-caffeoylquinic acid	5-CaQA	295(sh), 325	353/355	191, 179/377, 163	MS, standard, and literature[19,20,26]
36	4-O-caffeoylglucose	4-Ca-Glu	298(sh), 328	341/343	179, 161/365, 163	MS and literature[19,20,26,27]
37	1-O-caffeoylglucose	1-Ca-Glu	296(sh), 324	341/343	179, 161/365, 163	MS and literature[19,20,26,27]
38	coumaroylquinic acid isomer	CoQA	290(sh), 310	337/339	191, 163/361, 147	MS and literature[19,20,26]
39	3-O-coumaroylquinic acid	3-CoQA	292(sh), 314	337/339	191, 163/361, 147	MS and literature[19,20,26]
40	4-O-coumaroylglucose	4-Co-Glu	298(sh), 314	325/327	163/349, 165	MS and literature[19,20,26,27]
41	1-O-coumaroylglucose	1-Co-Glu	298(sh), 314	325/327	163/349,165	MS and literature[19,20,26,27]
42	3-O-caffeoylquinic acid	3-CaQA	295(sh), 325	353/355	191, 179/377, 163	MS and literature[19,20,26]
43	feruloylglucose isomer	Fe-Glu	298(sh), 318	355/357	193, 175/379, 177	MS and literature[19,20,26,27]
44	1-O-feruloylglucose	1-Fe-Glu	298(sh), 318	355/357	193, 175/379, 177	MS and literature[19,20,26,27]
45	(E)-caffeoyloxymethylene-glucopyranosyloxy-(Z)-butenenitrile	Ca-meGlu-B	296(sh), 329	436/438	179, 135/460, 276	MS and literature[19,28]
46	(E)-coumaroyloxymethylene-glucopyranosyloxy-(Z)-butenenitrile	Co-meGlu-B1	290(sh), 314	420/422	163, 119/444, 260	MS and literature[19,28]
47	(Z)-coumaroyloxymethylene-glucopyranosyloxy-(Z)-butenenitrile	Co-meGlu-B2	290(sh), 314	420/422	163, 119/444, 260	MS and literature[19,28]
48	(E)-feruloyloxymethylene-glucopyranosyloxy-(Z)-butenenitrile	Fe-meGlu-B	290(sh), 328	450/452	193, 134/474, 290	MS and literature[19,27,28]
	Flavan-3-ols
49	gallocatechin	GCat	280	305/307		MS
50	epigallocatechin	EGCat	280	305/307		MS
51	(+)-catechin	Cat	280	289/291		MS, standard, and literature[20,29]
52	(−)-epicatechin	ECat	280	289/291		MS, standard, and literature[20,29]
	Other Phenolics
53	aureusidin glucoside	Au-Glu	280, 325(sh)	447/449	285/287	MS and literature[19]
	Organic Acids
54	malic acid	MaA	-	-	-	standard and literature[12,30]
55	citric acid	CiA	-	-	-	standard and literature[12,30]
56	quinic acid	QuA	-	-	-	standard and literature[12]
57	ascorbic acid	AsA	-	-	-	standard and literature[12,30]
	Sugars
58–60	fructose anomers	Fru	-	-	-	standard and literature[12,30]
61,62	glucose anomers	Glu	-	-	-	standard and literature[12,30]
63	sucrose	Sur	-	-	-	standard and literature[12,30]

The number of compounds is referenced in Supplemental Figure 1.

Phenolic compounds were identified using UPLC-DAD-ESI-MS with the comparison of reference standards and previous literature. Both organic acids and sugars were identified using GC-FID with internal reference standards.

The abbreviation of each compound is used in PLS regression models.

Quantification of the Compounds

Total content of phenolics ranged from 598 to 2798 mg/100g in black cultivars and from 47 to 104 mg/100 g in green ones (Supplemental Table 2). It has been discussed previously that the absence of anthocyanins resulted in the lowest amount of total phenolics in green cultivars.[31] Among all black cultivars, the total content of anthocyanin was 1501 ± 587 mg/100 g, which was lower than previously detected by Mattila et al. (2057 ± 442 mg/100 g dry weight, DW) in 32 Finnish blackcurrant cultivars in a germplasm collection of mainly traditional cultivars.[32] Nour et al. reported that glycosides of cyanidin and delphinidin (3-O-glucoside and 3-O-rutinoside) accounted for 92–97% of total anthocyanins in blackcurrants.[33] Similar percentages were found in the current study. Anthocyanins formed the dominating groups of the phenolics in black-fruited samples, mainly as glycosylated delphinidin (34–66% of sum content of phenolics) and cyanidin (31–52%). The total content of flavonols was 18–60 mg/100 mg dry weight, accounting for 1–6% of sum content of phenolics in black cultivars and 37–39% in green ones. The difference between black and green cultivars was also shown in the profile of flavonols. In accordance with the results published by Mikkonen et al.,[34] myricetin glycosides was the dominant group of flavonols in the black cultivars studied; however, total content of quercetin glycosides was 6–8 times higher than that of myricetin glycosides in the green cultivars ‘Vilma’ and ‘Venny’.

For phenolic acids, the conjugates of coumaric acids (47–74% of total phenolic acid derivatives) were the major components in most of the cultivars, followed by caffeic acid (17–40%) and ferulic acid (9–20%); however, the cultivars ‘Ben Tron’ and ‘Joniniai’ contained more derivatives of caffeic acids and less of coumaric acids in both years. Moreover, the monomers of flavan-3-ols were found at a total content close to 10–20 mg/100 g.

Although the contents of simple organic acids significantly differed among the cultivars, citric acid accounted for 75–97% of the total content of simple acids (Supplemental Table 3) in accordance with a previous report.[35] It was followed by malic acid representing 3–20% of total simple acids. The highest values of ascorbic acid were found in ‘Tisel’ (2.0–2.5 mg/g), ‘Joniniai’ (2.2–2.5 mg/g), and ‘Ben Tirran’ (1.7–2.3 mg/g); however, in Finnish black cultivars, ascorbic acid was found at considerably low contents ranging from 0.2 to 0.6 mg/g. A small quantity of quinic acid was detected in all the samples. As the dominating sugars in all blackcurrant cultivars studied, fructose and glucose contributed 48–60% and 38–47% of total content of sugars, respectively. The concentration of fructose was higher than that of glucose in all the cultivars. Compared with fructose and glucose, sucrose was present at a lower level in the black-fruited currants as suggested by Woznicki.[36] In this study, ‘Dainiai’ showed significantly higher sucrose content (12 mg/g on average) than other cultivars studied. The contents of simple organic acids and sugars found in the samples in the current study deviated considerably from the levels reported in some blackcurrant cultivars studied in previous research studies.[13,35] This difference was likely due to the different genetic background of the cultivars included in these studies. Also, the growth locations were different in these studies; therefore, the environmental factors may have contributed to the difference observed.

Comparison of Blackcurrant Cultivars Growing in Different Years

A large and significant variation in chemical variables was observed within each cultivar between years 2014 and 2015. For phenolic compounds, two green cultivars presented significantly lower sum content of phenolics than the black cultivars. A newly bred Scottish sample, ‘S 18/2/23’, was also low in phenolics (598–745 mg/100 g of dry berries) in both years. Since annual deviation was seen likely due to the response of plants to the environment, a PLS regression model was used to find the distribution of individual compounds in different years. Regarding phenolic compounds, 78% of the chemical variables explained 89% of the variation among the cultivars in 7 factors in Figure a. Samples from 2015 showed higher total amount of flavan-3-ols, quercetins (primarily as quercetin 3-O-rutinoside), kaempferols (kaempferol 3-O-rutinoside), isorhamnetins, and coumaric acid derivatives than berries of the year 2014. The PLS model did not show clear correlation between years and anthocyanins or sum content of phenolics.

Figure 1

PLS models of comparison of blackcurrant cultivars in two different growing years: (a) phenolic compounds (n = 8662), (b) sugars and simple organic acids (n = 1098). Legend of the scores plots: red open circle for the samples harvested in Year 2015, blue open square means the samples harvested in Year 2014. In the loading plots, the growing year is in red bold italic font and the identified phenolic compounds are in blue font. The full names of these compounds are referenced in Table .

For simple acids and sugars, ‘Ben Tirran’ had the highest content of simple acids (53 mg/g in 2014 and 52 mg/g in 2015) among all the cultivars studied. Sugars were abundant in ‘Tauriai’ but poor in ‘Ben Finlay’. In the plot of Figure b, 64% of the chemical variables of simple acids and sugars explained 65% of the variation among the cultivars in 2 factors. Citric acid, fructose, and glucose correlated strongly with the samples collected in year 2015, which explained the higher content of total simple acids and total sugars, respectively, in this year.

In our previous research, the weather condition in the last months of growth before harvest showed special importance for blackcurrant fruit development,[10,11] since several main primary (sugars) and secondary (anthocyanins) metabolites start accumulating in the last stage of ripening of blackcurrant.[37] In the present study, exceptionally high temperatures including both maximum day time and minimum night time temperatures were observed from mid-July to mid-August of year 2014, which was the last month before harvesting (Supplemental Table 4). Zheng et al. reported that the average temperature of July correlated positively with the content of citric acid, fructose, and glucose in the Finnish cultivars ‘Mortti’ and ‘Ola’, based on analysis of berry samples collected in multiple years.[38] In our study, temperatures were higher than those in the study of Zheng et al.,[38] and our results showed the opposite: higher temperatures were related to the reduction of these sugars and citric acid. The phenomenon is commonly seen in other species too. It was shown, for instance, in strawberry (Fragaria ananassa) fruit that sugar content was negatively correlated to the temperature during fruit development,[39] and high temperatures have been shown to reduce the organic acids in berries of grapevine (Vitis vinifera).[40] Yet, our study was not able to determine that the climatic factors resulted in the yearly deviation of chemical composition of blackcurrant berries because of the data limited to two growing years only.

Comparison of Blackcurrant Cultivars Originating from Different Countries

PLS models were applied to investigate the difference among the samples in order to establish correlation between individual compounds and the cultivars. The PLS plots in Figure a show the interactions between chemical compounds, and notably, all cultivars of blackcurrants as 74% of the chemical variables explained 65% of the variation among the cultivars in 7 factors. Sum of phenolics and total anthocyanins correlated negatively with the green cultivars (‘Venny’ and ‘Vilma’) along the PC1. Along with the expected color-related compounds, myricetins, primarily 3-O-glucoside, 3-O-arabinoside, and the free aglycone of myricetin, also represented a negative correlation with the green cultivars. Since there was only one Latvian and one Polish cultivar, comparison was conducted among black-fruited cultivars of Scottish, Lithuanian, and Finnish origins.

Figure 2

PLS models of comparison of blackcurrant cultivars originating from different countries: based on chemical variables (n = X) (a) all cultivars (n = 9760), (b) the black cultivars (n = 6560) originating from Scotland and Lithuanian, (c) the black cultivars originating (n = 5600) from Scotland and Finland, (d) the black cultivars (n = 3840) originating from Lithuanian and Finland. Legend of the scores plots: blue-filled square for Scottish samples, red-filled circle for Lithuanian samples, green-filled triangle for Latvian samples, purple-filled diamond for Finnish black-fruited samples, brown-filled inverted triangle for Finnish green-fruited samples, and yellow-filled star for Polish samples. In the loading plots, the origin of country is in red bold italic font and the identified phenolic compounds are in blue font. The full names of the compounds are referenced in Table .

The Scottish cultivars generally had higher total content of phenolic acid derivatives than the Lithuanian samples (Figure b; 69% of the chemical variables explained 96% of the variation among the cultivars in 6 factors). Scottish cultivars correlated strongly to the derivatives of both coumaric acid (CoA) and ferulic acid (FeA), primarily as 4-O-coumaroylglucose (4-Co-Glu), (E and Z)-coumaroyloxymethylene-glucopyranosyloxy-(Z)-butenenitrile (Co-meGlu-B1 and 2), and 1-O-feruloylglucose (1-Fe-Glu). Positive correlations of Scottish cultivars were also found with gallocatechin (GCat) and catechin (Cat). The conjugates of both caffeic acid (CaA) and coumaric acid, as well as peonidin glycosides, flavan-3-ols, and ascorbic acid (AsA) were the main variables to separate the Scottish from the Finnish cultivars on the first two PCs in Figure c (72% variation in X-data explained 96% of the variation Y-data with 6 factors). Compared with the Finnish samples, the Lithuanian cultivars were richer in ascorbic acid and caffeic acid derivatives, mainly as 5-O-caffeoylquinic acid (5-CaQA) (Figure d; 66% of variation in X-data explained 97% of variation in Y-data with 5 factors). Also, higher amounts of 3-O-coumaroylquinic acid (3-CoQA) and peonidin 3-O-glucoside (Po-Glu) characterized the Lithuanian cultivars.

Comparison among Cultivars within Scottish Origin

The nine Scottish cultivars were classified into three groups as shown in the scores plot of Figure a based on the variation in the chemical variables (87% of the chemical variables explained 70% of the variation in Y-data with 7 factors). Group A contained cultivars ‘Ben Dorain’, ‘Ben Gairn’, ‘Ben Starav’, and ‘Ben Finlay’. Two newly bred cultivars, ‘S 18/2/23’ and ‘9154–3’, belonged to group B; group C consisted of ‘Ben Hope’, ‘Ben Tirran’, and ‘Ben Tron’. Since a single PLS model was not able to differentiate all Scottish blackcurrants, the comparison was performed by groups.

Figure 3

PLS models of comparison of main groups of Scottish cultivars: (a) all Scottish cultivars (n = 4160), (b) the comparison between groups A and B (n = 2720), (c) the comparison between groups A and C (n = 3200), (d) the comparison between groups B and C (n = 2400). In the loading plots, the names of cultivars and groups are in red bold italic font and the identified phenolic compounds are in blue font. The full names of the compounds are referenced in Table .

In Figure b, 76% of the chemical variables explained 98% of the variation among the cultivars in 5 factors, and the cultivars in group A had higher amounts of sum-content of phenolic and total anthocyanins than group B. Positive correlations of group A were found with cyanidin 3-O-rutinoside (Cy-Rut), petunidin 3-O-glucoside (Pt-Glu), pelargonidin 3-O-glucoside (Pl-Glu), and all glycosides of delphinidin (De) identified. Group B correlated mainly to 4-O-caffeoylglucose (4-Ca-Glu) and 4-Co-Glu. The blackcurrants in group C contained more phenolic acid derivatives (CaA and FeA) and flavan-3-ols than those in Group A (Figure c; 54% of the chemical variables explained 98% of the variation among the cultivars with 3 factors). Many of the minor flavonols, myricetin 3-O-galactoside (My-Gal), quercetin 3-O-galactoside (Qu-Gal), quercetin 3-O-arabinoside (Qu-Ara), and isorhamnetin 3-O-(6″-malonyl)-galactoside (Is-maGal) were not observed in the group C, which also distinguished these cultivars from others (Figure c, Supplemental Table 5). Figure d (84% of the variation in X-data explained 99% of the variation in Y-data with 5 factors) indicated that group B was low in sum content of all studied phenolics compared to group C, which was mostly due to the low content of anthocyanins (including De, Cy, Pt, Pl, and Po compounds) and flavonols (myricetin derivatives).

The variations within the groups A–C of Scottish cultivars observed in Figure were further examined in PLS regression plots in Figure . ‘Ben Dorain’ correlated strongly to citric acid (CiA), fructose (Fru), glucose (Glu), total simple organic acids, and total sugars (Figure a; 91% of the chemical variables explained 99% of the variation among the cultivars in 6 factors). ‘Ben Starav’ correlated positively to both sucrose (Suc) and quinic acid (QuA) in the plot consisting of factor 2 and factor 4 (not present in this paper). Cyanidin 3-O-arabinoside (Cy-Ara) was not found only in ‘Ben Gairn’; however, petunidin 3-O-rutinoside (Pt-Rut), epicatechin (ECat), and 4-Co-Glu were present at higher contents. ‘Ben Finlay’ correlated only to (E)-feruloyloxymethylene-glucopyranosyloxy-(Z)-butenenitrile (Fe-meGlu-B). For minor components, myricetin 3-O-rutinoside (My-Rut), and 3-O-coumaroylquinic acid (3-CoQA) showed negative correlations with ‘Ben Gairn’, but 1-O-coumaroylglucose (1-Co-Glu) correlated positively to ‘Ben Gairn’.

Figure 4

Comparison of Scottish cultivars with PLS regression models based on their chemical composition: (a) the comparison within group A (n = 2400), (b) the comparison within group B (n = 960), (c) the comparison within group C (n = 1140). The groups are based on the model in Figure . The name of cultivars is in red bold italic font and the identified phenolic compounds are in blue font. The full names of the compounds are referenced in Table .

The common difference between two cultivars in group B was that ‘S18/2/23’ was more abundant in citric acid, ascorbic acid, and sucrose, whereas the cultivar of ‘9154-3’ strongly correlated to the total content of flavonols, owing to the high concentration of glycosides of quercetin (Qu), and kaempferol (Ka) (Figure b; 67% of the variation in X-data explained 99% of the variation in Y-data with 2 factors). Phenolic acids in ‘S18/2/23’ were mainly present as the derivatives of caffeic acid, but more ferulic acid conjugates were found in ‘9154–3’.

Figure c showed 96% of the chemical variables explained 100% of the variation among the cultivars in group C with 5 factors. ‘Ben Tirran’ contained the highest amount of citric acid and ascorbic acid. ‘Ben Tron’ exhibited positive correlations with most of the glycosides of anthocyanidins, which explained the highest sum-content of phenolics among the samples in group C. Yet, delphinidin 3-O-(6″-coumaroyl)-glucoside (De-coGlu) and cyanidin 3-O-(6″-coumaroyl)-glucoside (Cy-coGlu) were abundant in ‘Ben Tirran’. ‘Ben Tirran’ was also rich in gallocatechin (GCat), myricetin aglycone (My agly), and ferulic acid derivatives. High concentration of total flavonols and caffeic acid derivatives correlated positively to ‘Ben Tron’, mainly due to the presence of Qu-Glu, 5-CaQA, and Ka-Gal. Moreover, 3-O-coumaroylquinic acids (3-CoQA), quercetin 3-O-(6″-malonyl)-glucoside (Qu-maGlu), and a coumaroylquinic acid isomer (CoQA) were quantified mostly in ‘Ben Hope’.

Comparison of Finnish Cultivars

Aside from anthocyanins, the Finnish green cultivars ‘Venny’ and ‘Vilma’ contained high amounts of ascorbic acid, kaempferol glycosides (Ka-Gal and Ka-Rut), and the derivatives of phenolic acids (4-Co-Glu, 1-Co-Glu, Co-meGlu-B2, and 1-Ca-Glu) compared with black ones. Additionally, myricetin was concentrated in black cultivars in forms of both glycosides (3-O-glucoside, 3-O-rutinoside, deoxylhexoside, and 3-O-arabinoside) and aglycone (Figure a). The PLS model in Figure b presents the variation (90% in X-data) among Finnish black cultivars (99% in Y-data with 4 factors). The sum of all phenolic compounds, including the main glycosides of delphinidin and myricetin, were most abundant in the cultivar ‘Marski’. Quercetins correlated strongly with ‘Mikael’ as 3-O-glucoside, 3-O-galactoside, and 3-O-arabinoside. ‘Mortti’ contained the highest levels of sucrose and 3-O-coumaroylquinic acid but the lowest concentrations of cyanidins, peonidins, malvidins, and total flavonols. ‘Venny’ and ‘Vilma’ shared similar compositional characteristics, which was not surprising, both being offsprings of the cultivar ‘Vertti’. ‘Vilma’ highly correlated with the content of sucrose and (E)-feruloyloxymethylene-glucopyranosyloxy-(Z)-butenenitrile (Ca-meGlu-B), whereas ‘Venny’ correlated mainly with malic acid (MaA), ascorbic acid, quinic acid, quercetin 3-O-rutinoside, and gallocatechin (Figure c; 92% of the variation explained 98% of the variation among the two green cultivars with 3 factors).

Figure 5

Comparison of Finnish cultivars with PLS regression models based on their chemical composition: (a) all Finnish cultivars (n = 2400), (b) black cultivars (n = 1440), (c) green cultivars (n = 960). In the loading plots, the name of cultivars is in red bold italic font and the identified phenolic compounds are in blue font. The full names of compounds are referenced in Table .

Comparison of Lithuanian Cultivars

Lithuanian samples were grouped as displayed in Supplemental Figure 2a,b. Group A consisted of ‘Almiai’, ‘Dainiai’, and ‘Gagatai’, presenting higher concentration of anthocyanins (mostly as De, Cy, and Po), myricetin glycosides, and phenolic acids (FeA derivatives) than both ‘Joniniai’ and ‘Tauriai’ in group B. Among the samples in group A, ‘Almiai’ correlated positively to simple organic acids (mainly as CiA); whereas sucrose, malic acid, and quinic acid were abundant in ‘Dainiai’ (Supplemental Figure 2c). The highest level of total anthocyanins was present in ‘Gagatai’, mainly owing to the high content of delphinidin 3-O-rutinoside, delphinidin 3-O-(6″-coumaroyl)-glucoside, and cyanidin 3-O-(6″-coumaroyl)-glucoside. This was in agreement with the results reported by Rubinskiene and co-workers showing higher content of anthocyanins in ‘Gagatai’ than in ‘Joniniai’ and ‘Almiai’.[41] In the present study, ‘Almiai’ correlated negatively to the total content of both cyanidins and myricetins. ‘Dainiai’ contained more (E)-coumaroyloxymethylene-glucopyranosyloxy-(Z)-butenenitrile, myricetin 3-O-galactoside, and myricetin 3-O-(6″-malonyl)-galactoside. Supplemental Figure 2d suggested that ‘Joniniai’ was richer in malic acid, quinic acid, and sucrose than ‘Tauriai’. Positive correlations were found between ‘Joniniai’ and both 3-O-glycoisdes and free aglycones of quercetin and myricetin, as well as some minor phenolics such as epicatechin and 4-O-caffeoylglucose. The total content of coumaric acid derivatives was higher in ‘Tauriai’ because of the presence of two isomers of coumaroyloxymethylene-glucopyranosyloxy-butenenitrile.

To our best knowledge, the present study is the first one revealing systematic information on compositional variation among blackcurrant cultivars originating from different countries. The overall differentiation among cultivars of different origins was highlighted by the concentrations of different phenolic acid derivatives, even after more than a five-year cultivation in the same geographical location with the same climatic condition. The study also found that the contents of organic acids, sugars, and phenolic acid derivatives in blackcurrants correlated strongly with growing year. This may have been caused by different weather conditions during fruit development. The results provide important guidelines for the selection of raw materials in food and beverage processing industry. For example, cultivar ‘Dainiai’ is rich in sucrose, and high levels of ascorbic acid were found in ‘Tisel’, ‘Joniniai’, and ‘Ben Tirran’. ‘S 18/2/23’, and ‘9154-3’ are poor sources of anthocyanins compared with other black-fruited cultivars. The manufacturers can select cultivars accordingly based on the requirements of their products.

In addition, the knowledge of variation in metabolites is essential for breeding new cultivars of blackcurrants. Besides agronomic traits such as yield, fruit size, and environmental resistance, the chemical composition in fruits of new cultivars will be probably more emphasized, when more specific information on human health-related effects of different compounds will be available in the future. Our results suggest that the breeding programs have resulted in variation in chemical quality of currants developed in different countries. The cultivars from the same country may share more similarities than those created in different countries. Therefore, it would be possible for plant breeding to improve fruit quality by introducing new quality characteristics from blackcurrant cultivars originating from different countries.