Systematic HPLC/DAD/MS n study on the extraction efficiency of polyphenols from black goji: Citric and ascorbic acid as alternative acid components in the extraction mixture

Abstract

BACKGROUND:

The beneficial effects of the fruit of Lycium rhutenicum (black goji) have been linked to their polyphenolic profile.

OBJECTIVE:

Systematic examination of the extraction efficiency of polyphenols from cultivated black goji from Macedonia was carried out using 25 different solvent mixtures containing methanol, acetone or water combined with different acids (hydrochloric, acetic, citric and ascorbic acid).

METHODS:

An HPLC/DAD/MS ⁿ method was used for identification and quantification of phenolic acids, flavonoids, anthocyanins and also spermines and spermidines.

RESULTS:

The extraction solvent composition was found to have a significant effect on the yield of total as well as specific polyphenols. Pure methanol was found to be more efficient solvent for extraction of total phenolic compounds than pure water or acetone. Ascorbic acid in methanol (2%, m/v) was found to be the most efficient extraction solvent for total phenolic compounds. Aqueous solutions of citric and ascorbic acid gave the highest yield of phenolic acids, spermidines and flavonoids. The anthocyanin content in these extracts was somewhat lower in comparison with the one obtained with methanol/water/ascorbic acid (70 : 28 : 2). The qualitative analysis of the fruits cultivated in Macedonia showed similar polyphenolic pattern and anthocyanin content to the native plant growing in China.

CONCLUSIONS:

Citric and ascorbic acid can be used as alternative acid components in the extraction mixture.

Keywords

black goji polyphenols extraction phenolic acids spermine spermidine anthocyanins citric acid ascorbic acid

1 Introduction

Goji berries have been used in China as food and for medicinal purposes for millennia [1]. Due to their high morphological similarity, fruits of two distinct species belonging to the family Solanaceae, Lycium barbarum and Lycium chinense (Chinese boxthorn), are usually marketed together as goji berries [2]. A third closely related species, a wild Lycium ruthenicum (black wolfberry) morphologically distinct from goji berries, is also popular in the traditional folk Chinese medicine [2].

Goji berries in traditional Chinese medicine have been used for more than 2000 years, but rapid increase in consumption and cultivation in Europe has been noticed at the beginning of the twenty-first century. There are several reports in the literature about the cultivation of goji berries in Switzerland [3], Romania [4], Italy [5], Bulgaria [6] and North Macedonia [7]. In most of the cases they were claimed to be L. barbarum species.

The fruit of L. ruthenicum (black goji) is morphologically distinct from L. barbarum (red goji or only goji), but also the differences in flavonoid pattern and content were also found [8 –10]. The ripe L. ruthenicum fruit has deep purple color due to a high content of anthocyanins, whereas the pigmentation of ripe L. barbarum fruit is reddish due to the high accumulation of carotenoids [8].

The fruits of L. ruthenicum are edible and have been used as a remedy for the treatment of hypertension, ureteral stones, tinea and furuncle, and gingival bleeding [11]. They contain a variety of bioactive ingredients, in particular, polyphenols such as anthocyanins, phenolic acids, flavonols and also spermine and spermidine derivatives [12]. Studies have shown the presence of more than 30 anthocyanins, of which petunidin glycosides being the most abundant [11]. More than 20 spermidine alkaloids were also reported, most of them containing caffeoyl or dihydrocaffeoyl moieties and some being glycosylated [13].

Furthermore, as compared to the red goji berry, detailed studies of the polyphenolic profiles for specific phenolic acids and flavonoids in black goji berry are limited, and there are only few papers reporting the presence of caffeic, ferulic, sinapic, chlorogenic acid and quercetin-3-O-rutinoside [12, 13]. A literature search reveals that research in the past was mainly focused on the extraction methods and measurement of the total anthocyanins [14, 15] or total phenolic content [12]. It is noteworthy that no comprehensive study (including analysis of pattern and content) has been conducted to explore the efficiency of extraction of specific polyphenols from L. ruthenicum in different solvent mixtures.

Many factors influence to the final results of polyphenolic profile as: the chemical nature of the phenolic compounds, the extraction mixture used, the extraction technique employed, the sample particle size, storage time and conditions, as well as the presence of interfering substances affect the efficiency of the extraction [16].

It is well known that anthocyanins are soluble in many polar solvents and commonly extracted by aqueous mixtures with organic solvents, such as methanol, ethanol, or acetone. Anthocyanins have good stability in solutions with low pH (below pH 2 there are in flavylium form), so the addition of small amounts of hydrochloric acid or formic acid is helpful to prevent the degradation of the nonacylated compounds [17]. With the development of technology, traditional extraction methods have been enhanced by using ultrasound or microwaves [18].

Several different procedures for extraction of polyphenolic compounds from black goji fruits are reported in the literature. It is important to stress that there is no single extraction method that may be suggested as the most efficient since there are no results from a systematic study of extraction efficiencies exploiting various extraction solvents. Yang et al. [13] and Jin et al. [19], used 70% ethanol adjusted to pH 2.5 with hydrochloric acid (HCl), Wang et al. [20] used methanol and formic acid (98:2, v/v), Tang & Giusti [21] used acetone, water, HCl (70:30, 0.01, v/v), Islam et al. [9] used mixture of acetone, water, acetic acid (70:29.5:0.5, v/v) and Tian et al. [22] used mixture of water and formic acid (98:2, v/v).

The main objectives of this study were: (1) to systematically examine the efficiency of extraction of polyphenols from L. ruthenicum in different solvent mixtures. For that purpose, solvent mixtures containing methanol, acetone or water combined with different acids (hydrochloric, acetic, citric and ascorbic acid) were used to evaluate the effect of the extraction solvent composition on the yield of different polyphenols and to find out the most efficient one; (2) to determine the polyphenolic profile of cultivated black goji from North Macedonia because this is a first report for analysis of the polyphenolic pattern of L. ruthenicum cultivated in Europe, and (3) to compare the qualitative and quantitative polyphenolic composition of these cultivated samples with the ones wild growing in China.

2 Materials and methods

2.1 Reagents and standards

Acetic acid, hydrochloric acid, citric acid, ascorbic acid, methanol, sodium carbonate and water were purchased from Merck (Darmstadt, Germany). Standard of cyanidin-3-glucoside chloride was purchased from Phytolab (Vestenbergsgreuth, Germany); quercetin 3-O-rutinoside was from Sigma (Darmstadt, Germany); caffeic and chlorogenic acid were from Genay (Lyon, France). Sodium fluoride was purchased from Alkaloid (Skopje, Macedonia).

2.2 Plant material and sample preparation

Berries of Lycium ruthenicum were collected during August 2020 from organic plantations located near the city of Skopje, North Macedonia (coordinates: 42° 01’ 33.8’ N 21°22’11.9’E). Goji cultivation started in this area in 2015 from plant cultivars of Lycium rhutenicum selected by Zoran Velkovski, Superfoods NO 1 DOO Skopje, BGB HEALTH. Amounts of 1-2 kg fruits from 7–10 plants were sampled in the same day from the top, central and bottom part of the plant. Fresh fruits were transported to the laboratory in polypropylene bags and stored at –80°C until analysis.

Around 100 g of frozen material was homogenized and precisely 1 g (to 0.1 mg) was weighed for each extraction performed in triplicate. Immediately, 10 mg of solid NaF (corresponding to 2 mM) was added to each extract to inactivate polyphenol oxidases and prevent phenolic degradation, and then the samples were extracted with 10 mL of the extraction solvent mixture (30 min, at 25°C). Twenty-five different solvent mixtures containing methanol, acetone or water with hydrochloric, acetic, citric and ascorbic acid were tested (Table 1). All extracts were sonicated for 30 minutes, centrifuged for 15 minutes at 3000 rpm at room temperature and the supernatant collected. The residue was then re-extracted with 10 mL of the same extraction solvent mixture, and the procedure was repeated. The supernatants were combined and filled to a final volume of extracts of 25 mL. The pH values of the combined extracts were measured (Table 1). Finally, the extract was filtered through a 0.45μm pore nylon membrane filter before analysis. Extractions and analyses were made in triplicate. All extracts were analyzed utilizing HPLC/DAD/MS ⁿ .

Table 1
Composition of extraction mixtures in % (v/v for hydrochloric and acetic acid, m/v for citric and ascorbic acid)

Extract Methanol (M) Acetone (A) Water (W) Hydrochloric acid (H) Acetic acid (Ac) Citric acid (C) Ascorbic acid (Asc) pH of extract

M 100 6.31

98M2H 98 2 0.72

98M2Ac 98 2 3.53

98M2C 98 2 2.61

98M2Asc 98 2 3.72

70M 70 30 5.76

70M2H 70 28 2 0.41

70M2Ac 70 28 2 2.17

70M2C 70 28 2 2.49

70M2Asc 70 28 2 3.25

A 100 5.48

98A2H 98 2 0.43

98A2Ac 98 2 4.65

98A2C 98 2 2.58

98A2Asc 98 2 3.63

70A 70 30 5.01

70A2H 70 28 2 0.68

70A2Ac 70 28 2 3.32

70A2C 70 28 2 2.31

70A2Asc 70 28 2 3.25

W 100 4.61

98W2H 98 0.70

98W2Ac 98 2 2.53

98W 2C 98 2 2.12

98W2Asc 98 2 2.61

2.3 LC/DAD/ESI-MSⁿ analysis

Chromatographic separations were carried out on a 250 mm×4.6 mm, 5μm, Supelco C18 Discovery column (Sigma-Aldrich, Germany). The mobile phase consisted of water-formic acid (2%, v/v) (A) and acetonitrile (B). Gradient elution method was used (0–5 minutes, 5% B; 5–45 minutes, 5% –35% B; 45–60 minutes, 35% –100% B, 60–65 minutes 100% –5% B, 65–75 minutes 5% B). The flow rate was 0.4 mL/min and the injection volume was 20μL. The HPLC system was equipped with an Agilent 1100 series diode array detector and ion trap mass detector in series (Agilent Technologies, Waldbronn, Germany). It consisted of G1312A binary pump, G1329A autosampler, G1379B degasser, and G1315D photodiode array detector, controlled by a ChemStation software (Agilent, v.08.03). Spectral data from all peaks were accumulated in the 190 to 600 nm range and chromatograms were recorded at 320 nm for hydroxycinnamic acids, spermines and spermidines, at 350 nm for flavonoids, and at 520 nm for anthocyanins. The mass detector was a G2449A ion-trap mass spectrometer equipped with an electrospray ionization (ESI) system and controlled by LCMSD software (Agilent, v.6.1.). Nitrogen was used as the nebulizing gas at a pressure of 50 psi and the flow was adjusted to 12 L/min. The heated capillary and the voltage were maintained at 325°C and 4 kV, respectively. The parameters for the capillary exit offset, skimmer 1, and skimmer 2 were 100, 40, and 6 V, respectively, and compound stability was 100%. Mass spectrometric data were acquired in the positive and negative ionization modes. The full scan covered the mass range at m/z 50 to 1200.

Peak assignment of various classes of polyphenols in the chromatograms was based on the comparison of their retention behavior and UV-Vis spectra to those of the authentic compounds and literature data [13 , 23–25]. The conjugated forms of polyphenolic compounds were further characterized by ESI-MS detection in the positive ionization mode for spermines, spermidines and anthocyanins and in the negative ionization mode for the phenolic acids, flavonoids and also for spermines and spermidines.

Quantification was performed by HPLC/DAD using 5-point regression curves (R²≥0.999) of authentic standards. Anthocyanins were quantified with cyanidin 3-O-glucoside chloride at 520 nm, flavonoids at 350 nm as quercetin 3-O-rutinoside equivalent, whereas hydroxycinnamic acids, spermine, and spermidine derivatives were determined at 320 nm using caffeic acid and chlorogenic acid as external standards.

2.4 Statistical analysis

Statistical treatment of the data was performed using Excel 2010 for calculations of calibration curves, mean and standard deviation. Samples were analyzed in triplicate, and one-way analysis of variance (ANOVA) was performed using STATISTICA, version 7. The Newman-Keuls post hoc test (at p < 0.01) was used to determine the significant differences between the results obtained for the individual compounds with different extraction solvents. Principal component analysis was performed using the software TANAGRA 1.4.28 (Lyon, France).

3 Results and discussion

In order to study the influence of organic solvent on extraction efficiency on one side, and the effect of using different type of acid on the other side, twenty-five different solvent mixtures containing pure methanol, acetone and water or mixtures of methanol, acetone and water with hydrochloric, acetic, citric and ascorbic acid were tested (Table 1).

According to literature data, solvents diluted with water have been more effective in the extraction of red goji berry phenolic compounds, than the single-solvent systems, since glycosylated phenolic compounds are more soluble in water [26]. For this reason, in our experiment, we used extraction mixtures with 70% aqueous methanol and 70% aqueous acetone.

The results of our previous study on the effect of acids in solvent mixtures for extraction of polyphenolic compounds from Aronia melanocarpa [27] clearly demonstrated that 2% or 5% acid in the extraction mixture significantly improves its efficiency for extraction of total phenolics and anthocyanins in comparison with lower content of acid. Consequently, in this study, extraction mixtures with 2% acid were used to test their efficiency for the extraction of individual polyphenolic compounds in L. ruthenicum. The effect of utilizing different acids such as hydrochloric, formic and acetic acid in solvent mixtures for extraction of phenolic compounds has been studied [13 , 22], whereas to the best of our knowledge, studies on the influence of citric and ascorbic acid have not been reported. Moreover, citric and ascorbic acid are natural antioxidants and non-toxic substances that are often added in preparation of different types of goji products and in general, plant derived food products.

The polyphenolic compounds in all the prepared L. ruthenicum extracts have been identified by their UV spectra, their deprotonated or protonated molecular ions and the corresponding ion fragments using LC/DAD/ESI-MS ⁿ . The results from the qualitative analysis are shown in Table 2. Typical chromatograms obtained with the extraction mixture 70M2H (methanol/water/HCl, 70:28:2, (v/v/v) and with the extraction mixture 70M2Asc (methanol/water/ascorbic acid, 70:28:2, v/v/v) are presented in Fig. 1.

Table 2
Retention times, UV/Vis maxima and MS data for the polyphenolic compounds in the analyzed extracts

Compound t_R/min UV max/nm MW [M–H]^– → MS² [M + H]⁺ → MS²

Phenolic acids

PA1 quinic acid 6.7 280 192 191 → 178, 193

173, 148, 111

PA2 syringic acid 11.7 274 198 197 → 182, 167 –

PA3 3-O-caffeoylquinic acid 12.9 242, 296, 326 354 353 → 191, 179, 135 –

PA4 1,5-di-O-caffeoylquinic acid 13.4 240, 294, 324 516 515 → 353 –

PA5 3,4-di-O-caffeoylquinic acid 14.3 242, 296, 324 516 515 → 353, 335 –

PA6 3,5-di-O-caffeoylquinic acid 16.2 242, 298, 324 516 515 → 353 –

PA7 5-O-caffeoylquinic acid 17.9 242, 298, 326 354 353 → 172, 191 –

PA8 4-O-caffeoylquinic acid 18.3 240, 298, 326 354 353 → 191, 179, 135 –

PA9 feruloylquinic acid 24.7 242, 290, 322 368 367 → 191,179, –

161,136

Spermine and spermidine

S1 N¹,N¹⁴-bis(dihydrocaffeoyl) 17.3 280, 316 530 529 → 407, 531 → 513, 293

spermine (kukoamine A) 365, 243, 121

S2 N¹,N¹⁰-dihydrocaffeoyl 19.6 286, 312 635 634 → 617, 636 → 619, 475, 385

spermidine hexoside 473, 382, 310

S3 N¹- dihydrocaffeoyl, N¹⁰- 20.4 286, 316 633 632 → 469, 334, 291 634 → 617, 472,

caffeoyl spermidine hexoside 293, 222

S4 N¹,N¹⁰-bis(dihydrocaffeoyl) 20.7 284, 318 473 472 → 350,308 474 → 457, 222

spermidine

S5 N¹,N¹⁰-bis(caffeoyl) 21.7 286, 320 633 632 → 469, 334 634 → 617, 472

spermidine hexoside

S6 N¹-dihydrocaffeoyl, N¹⁰- 22.6 284, 316 471 470 → 334, 291 472 → 455, 310, 220

caffeoyl spermidine

S7 N¹-caffeoyl, N¹⁰- 23.2 286, 318 471 470 → 334, 291

dihydrocaffeoyl spermidine

Flavonoids

F1 taxifolin-glucosyl- 12.5 260, 350 630 629 → 611, 513, –

dihydrocaffeate 465, 393, 303

F2 quercetin-3-O- 28.4 268, 358 610 609 → 301 611 → 303

rutinoside (rutin)

Anthocyanins

[M +]

A1 petunidin-3-O-(glucosyl- 24.7 280, 296, 532 1095 1095 1095 → 933, 479, 317

trans-p- coumaroyl)-

rutinoside-5-O-glucoside

A2 petunidin-3-O-(glucosyl- 25.8 280, 296, 532 1095 1095 1095 → 933, 479, 317

cis-p-coumaroyl)-rutinoside-

5-O-glucoside

A3 petunidin-3-O-(caffeoyl)- 27.9 280, 300, 528 949 949 949 → 787, 641, 479, 317

rutinoside-5-O-glucoside

A4 petunidin-3-O-(trans-p- 29.7 278, 304, 534 933 933 933 → 771, 479, 317

coumaroyl)-rutinoside-5-

O-glucoside

A5 malvidin-3-O-(p-coumaroyl)- 31.4 280, 308, 532 947 947 947 → 785, 639, 493, 331

rutinoside-5-O-glucoside

A6 petunidin-3-O-(p- 33.1 282, 312, 530 771 771 771 → 317

coumaroyl)-rutinoside

	Compound	t_R/min	UV max/nm	MW	[M–H]^– → MS²	[M + H]⁺ → MS²
Phenolic acids
PA1	quinic acid	6.7	280	192	191 → 178,	193
					173, 148, 111
PA2	syringic acid	11.7	274	198	197 → 182, 167	–
PA3	3-O-caffeoylquinic acid	12.9	242, 296, 326	354	353 → 191, 179, 135	–
PA4	1,5-di-O-caffeoylquinic acid	13.4	240, 294, 324	516	515 → 353	–
PA5	3,4-di-O-caffeoylquinic acid	14.3	242, 296, 324	516	515 → 353, 335	–
PA6	3,5-di-O-caffeoylquinic acid	16.2	242, 298, 324	516	515 → 353	–
PA7	5-O-caffeoylquinic acid	17.9	242, 298, 326	354	353 → 172, 191	–
PA8	4-O-caffeoylquinic acid	18.3	240, 298, 326	354	353 → 191, 179, 135	–
PA9	feruloylquinic acid	24.7	242, 290, 322	368	367 → 191,179,	–
					161,136
Spermine and spermidine
S1	N¹,N¹⁴-bis(dihydrocaffeoyl)	17.3	280, 316	530	529 → 407,	531 → 513, 293
	spermine (kukoamine A)				365, 243, 121
S2	N¹,N¹⁰-dihydrocaffeoyl	19.6	286, 312	635	634 → 617,	636 → 619, 475, 385
	spermidine hexoside				473, 382, 310
S3	N¹- dihydrocaffeoyl, N¹⁰-	20.4	286, 316	633	632 → 469, 334, 291	634 → 617, 472,
	caffeoyl spermidine hexoside				293, 222
S4	N¹,N¹⁰-bis(dihydrocaffeoyl)	20.7	284, 318	473	472 → 350,308	474 → 457, 222
	spermidine
S5	N¹,N¹⁰-bis(caffeoyl)	21.7	286, 320	633	632 → 469, 334	634 → 617, 472
	spermidine hexoside
S6	N¹-dihydrocaffeoyl, N¹⁰-	22.6	284, 316	471	470 → 334, 291	472 → 455, 310, 220
	caffeoyl spermidine
S7	N¹-caffeoyl, N¹⁰-	23.2	286, 318	471	470 → 334, 291
	dihydrocaffeoyl spermidine
Flavonoids
F1	taxifolin-glucosyl-	12.5	260, 350	630	629 → 611, 513,	–
	dihydrocaffeate				465, 393, 303
F2	quercetin-3-O-	28.4	268, 358	610	609 → 301	611 → 303
	rutinoside (rutin)
Anthocyanins
					[M +]
A1	petunidin-3-O-(glucosyl-	24.7	280, 296, 532	1095	1095	1095 → 933, 479, 317
	trans-p- coumaroyl)-
	rutinoside-5-O-glucoside
A2	petunidin-3-O-(glucosyl-	25.8	280, 296, 532	1095	1095	1095 → 933, 479, 317
	cis-p-coumaroyl)-rutinoside-
	5-O-glucoside
A3	petunidin-3-O-(caffeoyl)-	27.9	280, 300, 528	949	949	949 → 787, 641, 479, 317
	rutinoside-5-O-glucoside
A4	petunidin-3-O-(trans-p-	29.7	278, 304, 534	933	933	933 → 771, 479, 317
	coumaroyl)-rutinoside-5-
	O-glucoside
A5	malvidin-3-O-(p-coumaroyl)-	31.4	280, 308, 532	947	947	947 → 785, 639, 493, 331
	rutinoside-5-O-glucoside
A6	petunidin-3-O-(p-	33.1	282, 312, 530	771	771	771 → 317
	coumaroyl)-rutinoside

Fig. 1

HPLC/DAD chromatograms corresponding to extracts: 70M2H (methanol/water/HCl, 70 : 28 : 2, (v/v/v)) (in red) and 70M2Asc (methanol/water/ascorbic acid, 70 : 28 : 2, v/v/v) (in green): a. 280 nm, b. 350 nm, c. 520 nm. Peak assignments as in Table 1.

In total, 25 phenolic compounds were identified and classified as follows: nine phenolic acids (quinic acid, PA1; hydroxybenzoic acid, PA2; caffeoylquinic acids, PA3-PA8; feruloylquinic acid, PA9) (Fig. 2a), seven spermine or spermidine derivatives (Fig. 2b), two flavonoids (one flavanonol and one flavonol) (Fig. 2c) and six anthocyanins (Fig. 2d).

Fig. 2

Chemical structures of identified compounds extract from L. ruthenicum.

The identified phenolic compounds were quantified by HPLC/DAD. The quantitative data for each compound and for the total content for each of the above-mentioned four groups obtained with solvent mixtures were calculated and are given in Table 3 (the results are expressed as mg/100 g fruit, fresh weight, FW) and a graphical presentation of the obtained results for the total content of polyphenolic compounds with the contribution of each class in it is given in Fig. 3.

Table 3

Contents (mg/100 g Fresh weight, FW) of phenolic acids, spermine and spermidine, flavonoids and anthocyanins in L. ruthenicum obtained with extraction using 25 solvent mixtures

	M	98M2H	98M2Ac	98M2C	98M2Asc	70M	70M2H	70M2Ac	70M2C	70M2Asc
Phenolic acids
PA1	–	0.660±0.120 ^ab	–	0.501±0.031 ^ab	–	1.28±0.12	0.901±0.100 ^c	–	0.730±0.091 ^b	–
PA2	–	–	–	–	–	0.401±0.011 ^bc	0.162±0.051 ^a	0.301±0.011 ^b	0.311±0.011 ^b	0.429±0.021 ^bc
PA3	–	–	0.351±0.020 ^ab	–	0.271±0.021 ^a	–	–	–	–	–
PA4	–	0.330±0.021 ^a	1.13±0.04 ^d	0.470±0.111 ^ab	–	–	0.993±0.121 ^d	0.921±0.050 ^cd	0.832±0.021 ^c	0.561±0.060 ^b
PA5	–	–	0.701±0.010 ^b	–	–	0.521±0.050 ^a	–	0.542±0.021 ^a	–	–
PA6	2.40±0.21 ^b	4.69±0.85 ^d	4.46±0.31 ^d	4.46±0.43 ^d	4.03±0.32 ^cd	3.67±0.22 ^c	5.77±0.65 ^ef	4.96±0.51 ^de	4.89±0.73 ^d	6.31±0.21 ^f
PA7	7.02±1.10 ^c	0.321±0.060	14.9±0.99 ^f	14.2±1.1 ^f	9.12±1.07 ^de	8.51±0.92 ^cd	1.61±0.21 ^a	15.2±0.9 ^f	15.1±1.0 ^fg	10.8±0.7 ^e
PA8	5.34±0.98 ^a	–	–	–	–	6.51±0.65 ^a	–	–	–	–
PA9	–	13.0±1.4 ^a	–	–	–	–	13.6±1.11 ^a	–	–	–
Total	14.8±0.5	19.0±0.6 ^c	21.5±0.4 ^d	19.6±0.5 ^c	13.42±0.54	20.89±0.37 ^d	23.1±0.4	21.9±0.4 ^d	21.8±0.5 ^d	18.1±0.3
Spermine and
spermidine
S1	0.521±0.092 ^a	1.36±0.32 ^b	–	–	–	0.81±0.13 ^a	2.76±0.51 ^de	–	–	–
S2	6.71±0.89 ^f	5.01±0.87 ^c	3.17±0.12 ^ab	–	2.33±0.98 ^a	4.91±0.95 ^c	5.24±0.98 ^cd	4.45±0.54 ^bc	5.12±0.48 ^c	5.73±0.87 ^d
S3	11.2±1.40 ^d	14.8±1.1 ^e	14.2±1.6 ^e	15.9±0.7 ^f	16.8±1.2 ^h	13.8±1.8 ^e	19.2±1.0	15.2±1.0 ^ef	16.1±0.7 ^fh	17.4±1.0 ^g
S4	1.13±0.15 ^a	3.39±0.45 ^c	2.14±0.20 ^b	2.32±0.10 ^b	–	1.39±0.22 ^a	6.09±0.55	2.18±0.12 ^b	3.18±0.25 ^c	–
S5	–	–	–	–	–	1.21±0.65 ^a	–	–	–	–
S6	6.2±0.98 ^d	3.99±0.35 ^b	5.57±0.89 ^cd	5.60±0.98 ^cd	6.02±0.87 ^d	4.29±0.92 ^b	5.65±0.21 ^cd	3.79±0.25 ^b	5.97±0.82 ^d	10.2±0.9 ^g
S7	7.27±0.45 ^c	11.5±1.2 ^e	9.20±1.07	4.19±0.55 ^a	5.83±0.54 ^b	14.8±1.0 ^f	19.3±1.0 ^h	12.4±0.9 ^e	13.0±1.0 ^e	20.0±1.0 ^h
Total S	33.0±0.5 ^ef	40.1±0.4 ^g	34.3±0.6 ^f	28.0±0.4	31.0±0.3 ^e	41.2±0.6 ^g	58.2±0.3	38.0±0.4	43.4±0.3	53.3±0.1 ^h
Flavonoids
F1	2.82±0.12 ^a	3.18±0.92 ^a	4.10±0.32 ^bcd	4.99±0.31 ^cd	4.73±0.55 ^c	3.81±0.25 ^ab	5.04±0.61 ^cd	5.39±0.22 ^d	5.31±0.33 ^d	4.93±0.25 ^cd
F2	4.91±0.63 ^c	2.69±0.36 ^b	5.96±0.62 ^c	5.06±0.45 ^cd	5.25±0.12 ^cd	9.00±0.92	7.52±0.85	12.9±1.00 ^e	10.9±1.0	13.8±0.9 ^e
Total F	7.73±0.36 ^b	5.87±0.40	10.1±0.2	10.1±0.1 ^c	9.98±0.30 ^c	12.8±0.47	12.56±0.17 ^d	18.3±0.6e	16.2±0.5	18.7±0.5 ^e
Anthocyanins
A1	10.7±1.2 ^c	10.7±1.3 ^c	15.1±1.1 ^e	14.3±1.2 ^d	16.0±0.7 ^e	10.2±1.2 ^c	16.2±1.5 ^e	14.2±1.2 ^d	14.1±1.3 ^d	17.8±1.0 ^f
A2	1.47±0.21 ^b	0.791±0.091 ^a	1.47±0.10 ^b	0.909±0.071 ^a	1.50±0.05 ^b	0.82±0.02 ^a	1.38±0.12 ^b	1.34±0.22 ^b	1.27±0.05 ^b	1.62±0.65 ^b
A3	2.54±0.41 ^b	2.17±0.21 ^b	2.06±0.16 ^b	1.95±0.21 ^ab	2.23±0.12 ^b	1.89±0.11 ^ab	2.50±0.07 ^b	1.80±0.08 ^ab	1.63±0.28 ^a	2.11±0.28 ^b
A4	81.4±2.2 ^f	76.3±2.7 ^e	75.4±1.9 ^e	71.9±2.9 ^d	80.5±2.5 ^f	57.6±2.1 ^a	77.1±2.3 ^e	69.2±1.3 ^d	68.1±2.1 ^d	87.9±3.0
A5	1.44±0.21 ^d	1.25±0.12b ^c	1.22±0.07 ^b	1.24±0.24 ^bc	1.32±0.25 ^c	0.991±0.150 ^b	1.53±0.05 ^d	1.30±0.07 ^c	1.37±0.65 ^d	1.78±0.10 ^d
A6	0.391±0.070 ^bc	0.44±0.03 ^c	0.379±0.011 ^b	0.351±0.109 ^b	0.382±0.072 ^b	0.162±0.020 ^a	0.519±0.010 ^c	0.332±0.021 ^ab	0.349±0.051 ^b	0.384±0.031 ^b
Total A	97.9±0.8 ^f	91.7±1.1 ^e	95.7±0.8 ^f	90.7±1.2 ^e	101.9±1.0 ^g	71.7±0.9 ^c	99.1±1.1 ^g	88.1±0.6 ^de	86.8±0.8 ^d	111.6±1.1
Total phenolics	153.4 ± 0.2 ^d	156.7 ± 0.3 ^d	161.5 ± 0.3	148.3 ± 0.5 ^c	156.3 ± 0.3 ^d	146.6 ± 0.2 ^c	193.0 ± 0.4 ^f	166.3 ± 0.1 ^e	168.1 ± 0.2 ^e	201.7 ± 0.5 ^h
A	98A2H	98A2Ac	98A2C	98A2Asc	70A	70A2H	70A2Ac	70A2C	70A2Asc
Phenolic acids
PA1	–	–	–	–	–	–	–	–	–	0.512±0.131 ^ab
PA2	–	–	–	–	–	–	–	–	–	0.189±0.071 ^a
PA3	–	–	–	–	–	–	–	–	–	0.291±0.053 ^a
PA4	–	–	–	–	–	–	–	–	–	–
PA5	–	0.40±0.02 ^a	0.291±0.031	0.409±0.059 ^a	0.441±0.030 ^a	–	0.961±0.149 ^b	0.511±0.120	0.548±0.031	1.16±0.16 ^b
PA6	2.29±0.12 ^b	–	–	–	–	3.31±0.68 ^c	–	–	–	0.881±0.077 ^a
PA7	2.15±0.14 ^a	4.64±0.55 ^b	4.84±0.25 ^b	3.69±0.35 ^a	4.59±0.54 ^b	2.30±0.15 ^a	4.68±0.65 ^b	4.50±0.35 ^b	5.01±0.65 ^b	5.49±0.65 ^b
PA8	2.50±0.15	–	–	–	–	5.87±0.73 ^a	–	–	–	–
PA9	–	–	–	–	–	–	–	–	–	–
Total	6.94±0.02	5.04±0.37 ^ab	5.13±0.16 ^b	4.10±0.21 ^a	5.03±0.36 ^ab	11.48±0.32	5.64±0.35 ^b	5.01±0.16 ^ab	5.56±0.44 ^b	8.52±0.23 ^b
Spermine and
spermidine
S1	1.69±0.78 ^bc	2.09±0.21 ^cd	–	–	2.07±0.06 ^cd	3.12±0.35 ^e	–	–	–	0.901±0.060 ^a
S2	–	–	–	–	–	–	–	–	–	0.222±0.011
S3	3.76±0.65 ^a	4.40±0.33 ^b	5.48±0.65 ^bc	3.13±0.65 ^a	4.72±0.63 ^b	6.39±0.68 ^c	3.40±0.68 ^a	3.61±0.45 ^a	3.83±0.22 ^ab	5.55±0.25 ^c
S4	–	–	–	–	–	0.991±0.021 ^a	1.53±0.03 ^ab	–	–	–
S5	–	–	–	–	–	–	–	–	–	0.519±0.121
S6	8.11±0.87 ^e	7.72±0.68 ^e	8.78±0.98 ^f	10.86±0.88 ^g	9.88±1.05 ^fg	9.12±0.99 ^f	10.1±0.9 ^fg	9.19±0.99 ^f	7.05±0.97 ^e	10.1±1.2 ^g
S7	7.32±0.65 ^c	6.15±0.65 ^bc	6.23±0.75 ^bc	7.49±0.99 ^cd	5.36±0.87 ^ab	5.22±0.51 ^ab	5.09±0.45 ^a	5.71±0.85 ^b	8.18±1.02 ^d	8.08±1.09 ^d
Total	20.88±0.11 ^b	20.36±0.23 ^b	20.49±0.17 ^b	21.48±0.17 ^bc	22.03±0.43 ^c	24.84±0.36 ^d	20.1±0.4 ^b	18.51±0.28 ^a	19.06±0.45 ^a	25.4±0.6 ^d
Flavonoids
F1	–	–	–	–	–	–	–	–	–	–
F2	–	–	–	–	–	1.82±0.32 ^a	2.21±0.12 ^ab	2.13±0.32 ^ab	1.89±0.22 ^a	1.82±0.16 ^a
Total	–	–	–	–	–	1.82±0.32 ^a	2.21±0.12 ^ab	2.13±0.32 ^ab	1.89±0.22 ^a	1.82±0.16 ^a
Anthocyanins
A1	10.1±1.0 ^c	9.18±0.84 ^b	8.82±0.99 ^b	7.81±0.65 ^a	6.36±0.87 ^a	8.62±0.68 ^b	8.89±0.65 ^b	10.3±1.0 ^c	12.4±1.2	13.6±1.2 ^d
A2	–	–	–	–	–	–	–	–	–	–
A3	–	–	–	–	–	–	–	–	–	–
A4	70.1±2.2 ^d	56.6±1.3 ^a	57.2±2.1 ^b	62.2±2.7 ^b	42.2±1.4	45.6±1.1	62.3±2.2 ^b	54.4±2.2 ^a	37.2±2.0	24.7±1.5
A5	–	–	–	–	–	0.928±0.033	0.170±0.020 ^a	0.301±0.010 ^a	0.291±0.023 ^a	0.928±0.130 ^b
A6	–	–	–	0.571±0.030 ^c	–	–	–	–	–	0.861±0.091
Total	80.3±0.8	65.8±0.3 ^b	66.0±0.8 ^bc	70.5±1.4 ^c	48.6±0.4 ^a	55.2±0.5	71.4±1.1 ^c	65.0±0.8 ^b	49.9±1.0 ^a	40.2±0.7
Total phenolics	108.1 ± 0.4	91.2 ± 0.2 ^b	91.6 ± 0.3 ^b	96.1 ± 0.6	75.6 ± 0.2 ^a	93.3 ± 0.1 ^b	99.3 ± 0.5	90.6 ± 12.5 ^b	76.4 ± 0.3 ^a	75.9 ± 0.3 ^a
	W	98W2H	98W2Ac	98W2C	98W2Asc
Phenolic acids
PA1	0.451±0.02 ^ab	0.908±0.121 ^c	0.503±0.151 ^ab	1.01±0.15 ^c	1.06±0.05 ^c
PA2	0.599±0.099 ^d	0.571±0.030 ^cd	0.601±0.160 ^d	0.668±0.021 ^d	0.631±0.032 ^d
PA3	–	0.732±0.101 ^c	0.472±0.029 ^b	–	0.882±0.063 ^c
PA4	1.72±0.20 ^ef	1.57±0.32 ^e	1.71±0.21 ^ef	1.89±0.32 ^f	2.13±0.12
PA5	2.89±0.50 ^c	3.25±0.55 ^c	2.80±0.51 ^c	3.32±0.65 ^c	3.02±0.25 ^c
PA6	5.93±0.60 ^ef	6.16±0.98 ^f	2.66±0.32 ^b	6.23±0.88 ^f	0.767±0.012 ^a
PA7	5.80±0.30 ^b	9.51±1.02 ^de	15.4±1.01 ^fg	18.8±0.7	16.8±0.6 ^h
PA8	14.0±1.2 ^b	15.8±1.2 ^bcd	14.7±0.7 ^bc	16.6±1.0 ^cd	17.0±1.1 ^d
PA9	–	–	–	–	–
Total	31.4±0.4	38.5±0.5 ^e	38.8±0.3 ^e	48.5±0.4	42.3±0.4
Spermine and
spermidine
S1	1.56±0.03 ^b	3.65±0.65 ^f	2.85±0.12 ^e	2.99±0.12 ^e	3.63±0.22 ^f
S2	5.81±0.60 ^d	7.21±0.98 ^f	7.26±0.98 ^f	7.03±0.55 ^f	8.59±0.98
S3	9.73±0.90 ^d	16.8±1.4 ^gh	13.9±1.0 ^e	16.9±1.1 ^gh	14.6±2.1 ^e
S4	3.58±0.22 ^c	4.95±0.15 ^d	4.55±0.98 ^d	5.14±0.12 ^d	5.14±0.50 ^d
S5	1.70±0.32 ^a	–	–	–	–
S6	2.78±0.12 ^a	4.76±0.65 ^c	9.59±0.65 ^f	2.57±0.51 ^a	5.58±0.65 ^cd
S7	7.25±0.65 ^c	15.0±1.2 ^f	18.0±1.2 ^g	15.8±1.1 ^f	17.5±1.1 ^g
Total	32.41±0.32 ^e	52.5±0.4 ^h	56.1±0.4 ⁱ	50.4±0.4	55.0±0.7 ⁱ
Flavonoids
F1	4.27±0.13 ^bc	4.15±0.52 ^bc	4.94±0.25 ^cd	5.23±0.12 ^d	5.02±0.65 ^cd
F2	5.58±0.23 ^d	4.68±0.58 ^c	4.91±0.35 ^c	5.15±0.32 ^cd	4.99±0.32
Total	9.85±0.07 ^c	8.83±0.04 ^b	9.85±0.07 ^c	10.38±0.14 ^c	10.01±0.23 ^c
Anthocyanins
A1	8.98±0.65 ^b	18.0±1.6 ^f	14.2±1.2 ^d	17.0±1.7 ^f	15.0±1.0 ^e
A2	0.671±0.058 ^a	1.57±0.15 ^b	1.17±0.65 ^ab	1.48±0.22 ^b	1.24±0.01 ^b
A3	1.20±0.30 ^a	2.34±0.14 ^b	1.30±0.32 ^ab	2.17±0.21 ^b	1.89±0.20 ^a
A4	55.0±1.9 ^a	76.4±2.1 ^e	53.3±2.3 ^a	65.9±2.2 ^c	65.8±2.7 ^c
A5	0.400±0.031 ^g	1.34±0.54 ^c	0.961±0.012 ^b	1.20±0.15 ^b	1.10±0.05 ^b
A6	0.331±0.061 ^ab	0.477±0.071 ^c	0.391±0.030 ^bc	0.288±0.022 ^a	0.151±0.012 ^a
Total	66.5±0.7 ^b	100.2±0.9 ^g	71.3±0.9 ^c	88.0±0.9d ^e	85.2±1.1 ^d
Total phenolics	140.2 ± 0.3	200.0 ± 0.3 ^gh	176.1 ± 0.3	197.2 ± 0.3 ^g	192.4 ± 0.4 ^f

The extracts’ symbols are in accordance with Table 1 given in the Materials and methods section (M: methanol; A: acetone. W: water; H: hydrochloric acid; Ac: acetic acid; C: citric acid and Asc: ascorbic acid). Means (n = 3) (SD in the same row followed by same letters are not significantly different by Newman-Keuls test (p < 0.01).

3.1 Phenolic acids and derivatives

Nine phenolic acids were detected in all of the analyzed extracts (quinic acid was classified in phenolic acids, in spite of not being one, because it is related and forms conjugates with phenolic acids). Quinic acid (PA1), hydroxybenzoic acid (PA2) and seven hydroxycinnamic acid derivatives (PA3-PA9) (Fig. 2a). Syringic acid (PA2) was previously identified in L. barbarum [28], but not in L. rhutenicum, whereas quinic acid (PA1) was previously identified in L. minutifolium [25] but not in L. barbarum or L. rhutenicum.

In our study, quinic acid (PA1) was extracted in solvent mixtures that contain methanol/water (70M), methanol/HCl (98M2H, 70M2H) or methanol/citric acid (98M2C, 70M2C). The most efficient extraction solvent for quinic acid was 70M (1,28 mg/100 g FW). This compound is well extracted in pure water or in aqueous extraction mixtures containing also some acid. Citric and ascorbic acid in combination with water were also efficient for extraction of quinic acid (1.01 and 1.06 mg/100 g FW, for 98W2C and 98W2Asc, respectively).

Syringic acid (PA2) was found in the extracts obtained with solvents containing 70% methanol with or without acid, or water extracts with or without acid. The content of syringic acid was lower in methanol extracts compared to water extracts which was around 0.61 mg/100 g FW. According to the Newman-Keuls test, there were not significant differences (at p < 0.01) in syringic acid content in the water extracts without or with different acids i.e the presence and/or the nature of the acid did not affect the extraction efficiency.

The LC-DAD-MS(-) profiling also led to the identification of esters of hydroxycinnamic acids and quinic acid as the main polyphenolic components. According to mass spectral data, compounds PA3, PA7 and PA8 were identified as 3-, 5- and 4-O-caffeoylquinic acid, respectively. This fragmentation behavior is considered characteristic for neo-chlorogenic, chlorogenic and crypto-chlorogenic acid, as previously reported by Clifford et al. [23]. Chlorogenic acid (5-caffeoylquinic acid) have been previously reported in fruits of L. barbarum, L. chinense, and L. intricatum as well as in L. rhutenicum, where it was described as being the main acid [12, 29], whereas neo-chlorogenic and cryptochlorogenic acid were previously found in L. minutifolium and cultivated samples of L. barbarum from Romania [4] but not in L. rhutenicum.

In our extracts, 5-cafeoylquinic acid was also found as dominant, followed by 4-caffeoylquinic acid, especially in water extracts. The presence of 3-cafeoylquinic acid was limited, and was found in extracts containing ascorbic acid in combination with methanol, acetone or water.

Compounds PA4, PA5 and PA6 were identified as 1, 5-, 3, 4- and 3, 5-di-O-caffeoylquinic acids, respectively, according to their fragmentation pattern. 1, 5-di-O-caffeoylquinic acid was previously found in L. barbarum cultivars from Poland [30], whereas 3,4- and 3,5-di-O-caffeoylquinic acids were not previously identified in red or black goji. The total content of all dicafeoylquinic acids ranged from 0.291±0.021 mg/100 g FW (98A2Ac) to 11.44±0,99 mg/100 g FW (98W2C). The extracts with pure water or water/HCl mixture showed high content of dicafeoylquinic acids 10.5±0.2 mg/100 g FW (W) and 11.0±0.3 mg/100 g FW (98W2H), which were not significantly different at p < 0.01.

Compound PA9 (feruloylquinic acid) was detected only in the extracts prepared in solvents containing methanol and hydrochloric acid (98M2H and 70M2H) and it was observed that, in those extracts, the content of compound PA7 (5-caffeoylquinic acid) was significantly lower compared to the other methanol extracts (without HCl). As found in our previous study [27], it was here confirmed again that solvent mixtures containing methanol and HCl cause O-methylation of caffeoylquinic acids transforming them to feruloylquinic acids during the extraction. The experiments with model compounds have revealed that this O-methylation is specific for dihydroxycinnamic acid derivatives containing neighboring hydroxyl groups. This observation suggests that one should be very careful with the selection of solvents in the studies of various polyphenolic compounds since solvent mixtures containing methanol/HCl, very often used for assays of anthocyanins, can affect other polyphenolic compounds and lead to incorrect results.

The content of total phenolic acids ranged from as low as 4.10±0.21 mg/100 g FW in the extraction mixture acetone/citric acid (98A2C) to the highest content of 48.5±0.4 mg/100 g FW in the extraction mixture water/citric acid (98W2C), whereas the extraction mixtures containing methanol/citric acid (98M2C) and methanol/water/citric acid (70M2C) were in the medium range with 19.6±0.5 mg/100 g FW and 21.8±0.5 mg/100 g FW, respectively.

From the results it is evident that the extracts obtained using pure water of water in combination with some acid produced significantly higher yield of total phenolic acids. The water/citric acid mixture (extraction 98W2C) was found as the most efficient, whereas the extraction mixtures containing methanol were found as less efficient, but they were comparable between themselves. The acetone containing mixtures were found as the least efficient. These results are in accordance with our previous published data [31].

3.2 Spermine and spermidine

Spermine and spermidine are characteristic compounds for black goji that are not present in other berries. In our extracts, 7 amines (Fig. 2b), from which one spermine (S1) and six spermidines (S2-S7), including spermidine with dihydrocaffeoyl/caffeoyl or further hexose conjugation, were detected. However, because of the diversity of these compounds [13 , 32–34], some compounds were identified as isomers with identical molecular weight and analogous MS fragmentation pattern (Table 2). As for anthocyanins with sugar conjugation, the hexose linkage to spermidine derivatives may stabilize the compounds. All found compounds were previously identified in L. rhutenicum [13, 34].

Due to their conjugation with caffeoyl and dihydrocaffeoyl moieties they have an absorption maximum from 312–318 nm allowing their quantification with caffeic acid. The total content of spermine and spermidines ranged from 18.51±0.28 mg/100 g FW (70A2Ac) to 58.2±0.3 mg/100 g FW (70M2H). Extraction mixtures 98W2Asc and 70M2Asc, also were efficient for extraction of amines giving 55.0±0.7 mg/100 g FW and 53.3±0.1 mg/100 g FW, respectively.

The content of N¹, N¹⁴-bis(dihydrocaffeoyl) spermine (kukoamine A) (S1) was generally low, and the most efficient extraction mixture for this compound was 98W2Asc.

Compounds N¹- dihydrocaffeoyl, N¹⁰-caffeoyl spermidine hexoside (S3) and N¹-caffeoyl, N¹⁰-dihydrocaffeoyl spermidine (S7) contribute more than 60% to total spermidine content.

The extraction efficiency when using solvents prepared without acid was lower compared to the one with acidified solvents. Extraction mixtures methanol/water/HCl (70M2H) and methanol/water/ascorbic acid (70M2Asc) were found as most efficient for these compounds with yields of 38.5±0.3 mg/100 g FW and 37.4±0.1 mg/100 g FW, respectively. The extraction mixtures water/acetic acid (98W2Ac) and water/ascorbic acid (98W2Asc) were also very efficient for extraction of these two compounds.

From these results we can conclude that the water containing solvents with acetic or ascorbic acid were most efficient for spermidine extraction as well as extraction mixtures methanol/water/HCl.

3.3 Flavonoids

From the class of flavonoids, two compounds were detected. Taxifolin-glucosyl-dihydrocaffeate (F1) belongs to flavanonols and quercetin 3-O-glucoside (F2) belongs to flavonols (Fig. 2c). Taxifolin-glucosyl-dihydrocaffeate (F1) was not found in acetone containing extraction mixtures, whereas quercetin 3-O-glucoside (F2) was found in the one prepared with 70% acetone, but not in the non-aqueous acetone mixtures. Both compounds were quite well extracted in methanol or water containing solvents, and the most efficient were 70M2Ac (18.3±0.6 mg/100 mL) and 70M2Asc (18.7±0.5 mg/100 mL).

3.4 Anthocyanins

The identification results revealed that the six identified anthocyanins are derivatives of two anthocyanidins: malvidin (m/z 331) and petunidin (m/z 317). Specifically, five of the identified anthocyanins were petunidin glycosides (Fig. 2d). This is in agreement with previous studies showing that petunidin derivatives accounted for 95% of the total anthocyanins in L. rhutenicum fruit [34]. Particularly, the single compound petunidin-3-O-rutinoside (trans-p-coumaroyl)-5-O-glucoside (A4) accounts for more than 80% [36]. The next most abundant anthocyanin after A4 was petunidin-3-O-(glucosyl-trans-p-coumaroyl)-rutinoside-5-O-glucoside (A1), which contributed with 10 to 15% to total anthocyanin content.

Anthocyanins contribute with around 50% to total phenolic content in L. rhutenicum and it is very important to use the solvent mixture that is most efficient for their extraction. It is well established that anthocyanins can be found in different chemical forms depending on the pH of the solution. At pH 1, the flavylium cation is the predominant species and contributes to the purple and red color [37]. For that reason, HCl-containing solvent mixtures are widely used and recommended for their extraction [38].

The total anthocyanin content in the analyzed extracts ranged from 40.2±0.7 mg/100 g FW (70A2Asc) to 111.6±1.1 mg/100 g FW (70M2Asc). High anthocyanin content was also found in the extracts 98W2H (100.2±0.9 mg/100 g FW) and 70M2H (99.1±1.1 mg/100 g FW).

These results are very interesting because they demonstrate higher efficiency of ascorbic acid in comparison with HCl in the solvent mixture that can be important for practical reasons since ascorbic acid as antioxidant can be used in preparation of various goji containing products intended for human consumption.

There are limited published data on the quantification of anthocyanins in L. rhutenicum using HPLC, one paper reports higher results [20], but this difference should be attributed to the expression of the content as DW (dry weight), that is not comparable to FW (fresh weight) in our study.

3.5 Total phenolic content

For a better view of the obtained yield of phenolic compounds with different extraction solvents, the total content of phenolic compounds and the contribution of each group (summed individual contents determined by HPLC/DAD) in each extract are presented in Fig. 3.

Fig. 3

Total content of phenolic compounds (mg/100 g FW) with contributions of each group (expressed as mg/100 g fruit, fresh weight, FW) in the extracts, phenolic acids (PA, green), spermines and spermidines (S, blue), flavonoids (F, yellow), and anthocyanins (A, purple).

The total content of phenolic compounds ranged from 75.6±0.2 mg/100 g FW (98A2Asc) to 201.7±0.4 mg/100 g FW (70M2Asc). The high yield of total phenolic compounds in water/acid extracts (98W2H, 98W2C and 98W2Asc) is also observed, with content of 200.0±0.3 mg/100 g FW, 197.2±0.3 mg/100 g FW and 192.4±0.4 mg/100 g FW, respectively.

It is evident from these results that ascorbic acid in methanol or water gives the highest efficiency for extraction of flavonoids, anthocyanins and spermidines, but it is not as efficient in combination with acetone.

There were no available data in the literature for detail quantitative analysis of phenolic compounds in L. rhutenicum to compare. The results from spectrophotometric measurements of total phenolic compounds indicate the maximum yield of total phenolic content (TPC) of 17.92 mg GAE/g under the best extraction conditions (an extraction temperature of 89.38°C, an ethanol concentration of 70% and an extraction time of 13 min) [39], which is in agreement with our results.

3.6 Principal component analysis

In order to evaluate the significance of the obtained variables, i.e., nature and content of polyphenolic compounds, for making distinctions between the extraction mixtures studied, principal component analysis (PCA) was applied. The PCA analysis revealed five principal components. The first factor (PC1), which explained 50.49% of the variance was mainly linked to spermidine and anthocyanins, whereas the second principal component, which explained further 24.93% of the total variance was related to hydroxycinnamic acids. The principal component score plot of the variables with PC1 and PC2 based on individual components is presented in Fig. 4. According to PC1 vs. PC2 plot, five groups can be distinguished for all 25 extracts. PC1 clearly separates a group of extracts prepared with pure acetone and acetone/water mixtures with or without acid from the other extraction mixtures (based on methanol or water), due to their lower values of individual and total phenolic compounds. By PC2, the extracts obtained with pure methanol solvent, with acid or without acid were separated from the extracts prepared with water mixtures.

Fig. 4

Principal component analysis score plot of the variables with PC1 and PC2. For sample codes see Table 1.

Among the pure solvents, methanol was the most efficient solvent for extraction of total phenolic compounds, followed by water and acetone. Organic solvent-water mixtures were more efficient in extracting polyphenolic compounds than the respective pure organic solvents, with acid or not, and this result agreed with PCA analysis. According to Cheng et al. [40], the presence of water increases the permeability of cell tissue and thus, enables better mass transfer by molecular diffusion as well as the recovery of water-soluble bioactive compounds.

4 Conclusions

In summary, our study clearly showed that the extraction of phenolic compounds is significantly affected by the choice of the solvent constituents. Methanol was found to be the most efficient solvent for extraction of total phenolic compounds when used as a pure solvent, followed by water and acetone. Addition of 2% (m/v) ascorbic acid in methanol was found to be most favorable with regards to improving the extraction efficiency for total phenolic compounds. Citric and ascorbic acid in combination with water gave the highest yield of phenolic acids, spermidines and flavonoids. The anthocyanin content in these extracts was lower in comparison with the one obtained with the extraction mixture methanol/water/ascorbic acid (70M2Asc), but still satisfactory. In order to obtain fruit extracts to be used as natural antioxidants, the choice of a suitable non-toxic solvent additive must be carefully considered, and from that point, ascorbic and citric acid are a very convenient option.

The qualitative analysis of the fruits of L. rhutenicum cultivated in Europe showed similar polyphenolic pattern to the native ones growing in China. The anthocyanin contents were also comparable, but there are no available quantitative data for comparison of the phenolic acids and spermidines content in L. rhutenicum since this seems to be a first report.

Compared to other fruits and berries growing in Macedonia, cultivated black goji (L. rhutenicum) have shown higher total phenolic content compared to pomegranates (Punica granatum L.) [41], and lower in comparison with bilberries (Vaccinium myrtillus) and bog bilberries (Vaccinium myrtillus) [42], blueberries (Vaccinium corymbosum) [43] and aronia (Aronia melanocarpa) [27]. If we take into account that spermidine derivatives are present only in black goji (L. rhutenicum), which have important biological activities [44], and together with the rich polyphenolic profile, we can conclude that cultivated black goji fruits and the products derived thereof should find appropriate place in the food market. These health beneficial products have potential to be economically viable, but before that, analytical methodology for their quality control should be established focusing on polyphenolic composition and content.

Author contributions: JPS the principal investigator, contributed by designing the experiment, conducted all the experiments, data analyzing, and writing the first draft of the manuscript. MS and JB contributed by designing the experiment, data evaluation and reviewing the manuscript.

Footnotes

Acknowledgment

The authors kindly acknowledge for sample growing and collection provided by Superfoods NO 1 DOO Skopje, BGB HEALTH.

Declaration of competing interest

The authors do not have any conflict of interest to declare.

References

Potterat

. Goji (Lycium barbarum and L.chinense): phytochemistry, pharmacology and safety in the perspective of traditional uses and recent popularity. Planta Med. 2010;76:7–19. https://doi.org/10.1055/s-0029-1186218

Zhao

, Li

, Yin

, An

, Qin

, Wang

, Fan

, Li

, Cao

. Fruit ripening in Lycium barbarum and Lycium ruthenicum is associated with distinct gene expression patterns. FEBS Open Bio. 2020;10:1550–1567. https://doi.org/10.1002/2211-5463.12910

Kosinska-Cagnazzo

, Weber

, Chablais

, Vouillamoz

, Molnar

, Crovadore

, Lefort

, Andlauer

. Bioactive compound profile and antioxidant activity of fruits from six goji cultivars cultivated in Switzerland. J Berry Res. 2017;7:43–59. https://doi.org/10.3233/JBR-160144

Mocan

, Moldovan

, Zengin

, Bender

, Locatelli

, Simirgiotis

, Atalay

, Vodnar

, Rohn

, Crisan

. UHPLC-QTOFMS analysis of bioactive constituents from two Romanian Goji (Lycium barbarum L.) berries cultivars and their antioxidant, enzyme inhibitory, and real-time cytotoxicological evaluation. Food Chem. Toxicol. 2018;115:414–424. https://doi.org/10.1016/j.fct.2018.01.054

Donno

, Beccaro

, Mellano

, Cerutti

, Bounous

. Goji berry fruit (Lycium s): Antioxidant compound fingerprint and bioactivity evaluation. J. Funct. Foods. 2015;18:1070–1085. https://doi.org/10.1016/j.jff.2014.05.020

Dzhugalov

, Lichev

, Yordanov

, Kaymakaov

, Dimitrova

, Kutoranov

. First results of testing goji berry (Lycium barbarum L.) in Plovdiv region, Bulgaria. Sci. Papers Ser. B. Hortic. 2015;LIX:47–50.

Covaci

, Senila

, Leopold

, Olah

, Cobzac

, Ivanova-Petropulos

, Balabanova

, Cadar

, Becze

, Ponta

, Mot

, Frentiu

. Characterization of Lycium barbarum L. berry cultivated in North Macedonia: A chemometric approach. J. Berry Res. 2020;10:223–241. https://doi.org/10.3233/JBR-190450

Liu

, Zeng

, Sun

, Wu

, Hu

, Shen

, Wang

. Comparative analysis of carotenoid accumulation in two goji (Lycium barbarum L. and L. ruthenicum Murr.) fruits. BMC Plant Biol. 2014;14(269):1–14. https://doi.org/10.1186/s12870-014-0269-4

Islam

, Yu

, Badwal

, Xu

. Comparative studies on phenolic profiles, antioxidant capacities and carotenoid contents of red goji berry (Lycium barbarum) and black goji berry (Lycium ruthenicum). Chem Cent J. 2017;11(59):1–8. https://doi.org/10.1186/s13065-017-0287-z

10.

Wang

, Zeng

, Wu

, Lei

, Wang

, Tang

. Interspecies developmental differences in metabonomic phenotypes of Lycium ruthenicum and L.barbarum fruits. J. Proteome Res. 2018;17:3223–36. https://doi.org/10.1021/acs.jproteome.8b00349

11.

Wang

, Li

, Tao

, Zhang

, Gao

, Yong

, Zhao

, Zhang

, Li

, Duan

. Lycium ruthenicum studies: Molecular biology, phytochemistry and pharmacology. Food Chem. 2018;240:759–66. https://doi.org/10.1016/j.foodchem.2017.08.026

12.

Kosar

, Altintas

, Kirimer

, Baser

KHC

. Determination of the free radical scavenging activity of Lycium extracts. Chem Nat Compd. 2003;39:531–5. https://doi.org/10.1023/B:CONC.0000018104.98378.e2

13.

Yang

, Lin

, Jia

, Rehman

, Zeng

, Wang

. Anthocyanin and spermidine derivative hexoses coordinately increase in the ripening fruit of Lycium ruthenicum. Food Chem. 2020;311:1–9. https://doi.org/10.1016/j.foodchem.2019.125874

14.

, Qu

, Lv

, Yuan

. Study on extracting and refining of the pigments from Lycium ruthenicum Murr. Nat Prod Res Develop. 2006;18:650–4.

15.

Luo

, Jin

, Gao

, Chen

. Determination of anthocyanin in Lycium ruthenicum Murr. from different producing areas in Hei River basin by UV. Chin Med J Res Practic. 2015;29:24–7.

16.

Naczk

, Shahidi

. Extraction and analysis of phenolics in food. J Chromatog A. 2004;1054:95–111. https://doi.org/10.1016/j.chroma.2004.08.059

17.

Vankar

, Srivastava

. Evaluation of anthocyanin content in red and blue flowers. Int J Food Eng. 2010;6(4):1–11. https://doi.org/10.2202/1556-3758.1907

18.

Azwanida

. A Review on the extraction methods use in medicinal plants, principle, strength and limitation. Med Aro Plants. 2015;4:1–6. https://doi.org/10.4172/2167-0412.1000196

19.

Jin

, Liu

, Yang

, Wang

, Fu

, Zhang

, Pengc

, Lianga

. Characterization of anthocyanins in wild Lycium ruthenicum Murray by HPLC-DAD/QTOF-MS/MS. Anal Methods. 2015;12:4947–4956. https://doi.org/10.1039/C5AY00612K

20.

Wang

, Yan

, Nisar

, Zou

, Yang

, Niu

, Sun

, Guo

. Comparison and multivariate statistical analysis of anthocyanin composition in Lycium ruthenicum Murray from different regions to trace geographical origins: The case of China. Food Chem. 2018;246:233–41. https://doi.org/10.1016/j.foodchem.2017.11.030

21.

Tang

, Giusti

. Black goji as a potential source of natural color in a wide pH range. Food Chem. 2018;269:419–426. https://doi.org/10.1016/j.foodchem.2018.07.034

22.

Tian

, Aierken

, Pang

, Du

, Feng

, Ma

, Gao

, Bai

, Ma

. Constituent analysis and quality control of anthocyanin constituents of dried Lycium ruthenicum Murray fruits by HPLC-MS and HPLC-DAD. J Liq Chromatogr Relat. 2016;39:453–8. https://doi.org/10.1080/10826076.2016.1179201

23.

Clifford

, Knight

, Kuhnert

. Discriminating between the six isomers of dicaffeoylquinic acid by LC-MSⁿ. J. Agric. Food Chem. 2005;53:3821–32. https://doi.org/10.1021/jf050046h

24.

Suh

, Jung

, Lee

. Distinguishing six edible berries based on metabolic pathway and bioactivity correlations by non-targeted metabolite profiling. Front Plant Sci. 2018;9(1462):1–14. https://doi.org/10.3389/fpls.2018.01462

25.

Rodriguez

, Pertino

, Arcos

, Reichert

, Echeverria

, Simirgiotis

, Borquez

, Cornejo

, Areche

, Sepulveda

. Isolation, gastroprotective effects and untargeted metabolomics analysis of Lycium Minutifolium J. Remy (Solanaceae). Foods. 2020;9(565):1–12. https://doi.org/10.3390/foods9050565

26.

Pedro

, Maurer

JBB

, Zawadzki-Baggio

, Ávila

, Maciel

, Haminiuk

CWI

. Bioactive compounds of organic goji berry (Lycium barbarum L.) prevents oxidative deterioration of soybean oil. Ind Crops Prod. 2018;112:90–97. https://doi.org/10.1016/j.indcrop.2017.10.052

27.

Petreska Stanoeva

, Balshikevska

, Stefova

, Tusevski

, Simic

. Comparison of the effect of acids in solvent mixtures for extraction of phenolic compounds from Aronia melanocarpa. Nat Prod Commun. 2020;15:1–10. https://doi.org/10.1177/1934578X20934675

28.

Mocan

, Cairone

, Locatelli

, Cacciagrano

, Carradori

, Vodnar

, Crisan

, Simonetti

, Cesa

. Polyphenols from Lycium barbarum (goji) fruit European cultivars at different maturation steps: extraction, HPLC-DAD analyses, and biological evaluation. Antioxidants. 2019;8(562):1–14. https://doi.org/10.3390/antiox8110562

29.

Yao

, Peng

, Xu

, Li

, Wu

, Xiao

. Phytochemical and biological studies of Lycium medicinal plants. Chem Biodivers. 2011;8:976–1010. https://doi.org/10.1002/cbdv.201000018.

30.

Wojdyło

, Nowicka

, Bąbelewski

. Phenolic and carotenoid profile of new goji cultivars and their antihyperglycemic, anti-aging and antioxidant properties. J Funct Foods. 2018;48:632–42. https://doi.org/10.1016/j.jff.2018.07.061

31.

Kajdžanoska

, Petreska

, Stefova

. Comparison of different extraction solvent mixtures for characterization of phenolic compounds in strawberries. J Agric Food Chem. 2011;59:5272–8. https://doi.org/10.1021/jf2007826

32.

Zhou

, Fan

, He

, Xiao

, Tsoi

, Lan

, Kurihara

, So

, Yao

. , Gao, H. . Lycibarbarspermidines A-O, new dicaffeoylspermidine derivatives from wolfberry, with activities against Alzheimer’s disease and oxidation. J Agric Food Chem. 2016;64:2223–37. https://doi.org/10.1021/acs.jafc.5b05274

33.

, Lv

, Wang

. Characterization of polyphenols from Lycium ruthenicum fruit by UPLC-Q-TOF/MSE and their antioxidant activity in Caco-2 cells. J Agric Food Chem. 2016;64:2280–8. https://doi.org/10.1021/acs.jafc.6b00035

34.

Nzeuwa

IBY

, Xia

, Qiao

, Feng

, Bian

, Liu

, Qu

. Comparison of the origin and phenolic contents of Lycium ruthenicum Murr. by high-performance liquid chromatography fingerprinting combined with quadrupole time-of-flight mass spectrometry and chemometrics. J Sep Sci. 2017;40:1234–1243. https://doi.org/10.1002/jssc.201601147

35.

Zheng

, Ding

, Wang

, Li

, Shi

, Li

, Wang

, Suo

. Anthocyanins composition and antioxidant activity of wild Lycium ruthenicum Murr. from QinghaiTibet Plateau. Food Chem. 2011;126:859–65. https://doi.org/10.1016/j.foodchem.2010.11.052

36.

Chen

, Shao

, Tao

, Mei

, Shu

, Wang

. Main anthocyanins compositions and corresponding H-ORAC assay for wild Lycium ruthenicum Murr. fruits from the Qaidam Basin. J Pharm Technol Drug Res. 2013;2:1–5. https://dx-doi-org.web.bisu.edu.cn/10.7243/2050-120X-2-1

37.

Castañeda-Ovando

, Pacheco-Hernández

MdeL

. , PáezHernández ME, Rodríguez JA, Galán-Vidal CA. Chemical studies of anthocyanins: a review. Food Chem. 2009;113:859–71. https://doi.org/10.1016/j.foodchem.2008.09.001

38.

Silva

, Costa

, Calhau

, Morais

, Pintado

. Anthocyanin extraction from plant tissues: a review. Crit Rev Food Sci Nutr. 2017;57:3072–3083. https://doi.org/10.1080/10408398.2015.1087963

39.

, Du

, Xu

. Extraction optimization of phenolics and antioxidants from black goji berry by accelerated solvent extractor using response surface methodology. Appl Sci. 2018;8:1–13. https://doi.org/10.3390/app8101905

40.

Cheng

, Bekhit

AEA

, McConnell

, Mros

, Zhao

. Effect of extraction solvent, waste fraction and grape variety on the antimicrobial and antioxidant activities of extracts from wine residue from cool climate. Food Chem. 2012;134:474–482. https://doi.org/10.1016/j.foodchem.2012.02.103

41.

Petreska Stanoeva

, Peneva

, Stefova

, Gjamovski

. Identification and quantification of phenolic compounds in pomegranate juices from eight Macedonian cultivars. Maced. J. Chem. Chem. Eng. 2019;38:149–160. https://dx-doi-org.web.bisu.edu.cn/10.20450/mjcce.2019.1775

42.

Petreska Stanoeva

, Stefova

, Baceva Andonovska

, Vankova

, Stafilov

. Phenolics and mineral content in bilberry and bog bilberry from Macedonia. Int J Food Prop. 2017;20:S863–S883. https://doi.org/10.1080/10942912.2017.1315592

43.

Gavrilova

, Kajdžanoska

, Gjamovski

, Stefova

. Separation, characterization and quantification of phenolic compounds in blueberries and red and black currants by HPLC-DAD-ESI-MSⁿ. J Agric Food Chem. 2011;59:4009–4018. https://doi.org/10.1021/jf104565y

44.

Kumar

, Bhatt

, Kumar

. Amides from plants: structures and biological importance, studies in natural products chemistry. Stud Nat Prod Chem. 2018;56:287–333. https://doi.org/10.1016/B978-0-444-64058-1.00009-1

Extract	Methanol (M)	Acetone (A)	Water (W)	Hydrochloric acid (H)	Acetic acid (Ac)	Citric acid (C)	Ascorbic acid (Asc)	pH of extract
M	100							6.31
98M2H	98			2				0.72
98M2Ac	98				2			3.53
98M2C	98					2		2.61
98M2Asc	98						2	3.72
70M	70		30					5.76
70M2H	70		28	2				0.41
70M2Ac	70		28		2			2.17
70M2C	70		28			2		2.49
70M2Asc	70		28				2	3.25
A		100						5.48
98A2H		98		2				0.43
98A2Ac		98			2			4.65
98A2C		98				2		2.58
98A2Asc		98					2	3.63
70A		70	30					5.01
70A2H		70	28	2				0.68
70A2Ac		70	28		2			3.32
70A2C		70	28			2		2.31
70A2Asc		70	28				2	3.25
W			100					4.61
98W2H			98					0.70
98W2Ac			98		2			2.53
98W 2C			98			2		2.12
98W2Asc			98				2	2.61