Mixtures of phenolic extracts of Plinia jaboticaba, Euterpe edulis and Vaccinium myrtillus fruits and identification of the anthocyanins by HPLC-DAD-ESI/MSⁿ

Abstract

A blend of phenolic extracts from jabuticaba (Plinia jaboticaba), jussara (Euterpe edulis), and blueberry (Vaccinium myrtillus) has been proposed as a viable approach for incorporating bioactive compounds into food products. In this study, concentrated phenolic extracts from these three fruits were combined to formulate a mixture with a high content of phenolic compounds, total anthocyanins, and antioxidant capacity. The anthocyanins profile of the fruits was analyzed by HPLC/MS. The phenolic extracts were blended based on a mixture experimental design (M1−M10). Multiple regression models were applied to evaluate the relationships between the mixture components and the response variables, which included the content of anthocyanins and phenolic compounds, antioxidant capacity, and color parameters. The blueberry extract exhibited the greatest diversity of anthocyanins, with cyanidin-3-glycoside was the majority. The M9 mixture, containing 0.2167:0.5667:0.2167 (v/v/v) of jabuticaba: jussara: blueberry extracts, presented the highest total anthocyanins, total phenolics and antioxidant capacity. The resulting mixtures were predominantly purple, largely influenced by the jussara extract. The M2 blend, with the highest proportion of the jussara extract, exhibited color values of a* 0.43 and h* -56.51. The combination of jussara, jabuticaba, and blueberry extracts led to an increased in the concentration and diversity of anthocyanins and other phenolic compounds.

Highlights

• Fourteen anthocyanins were identified in blueberry fruit (Vaccinium myrtillus).

• Malvidin-3-O-galactoside was the principal anthocyanin in blueberry.

• The mix of the fruit extracts increased anthocyanin levels.

Introduction

Brazil hosts a significant diversity of native fruits that remais commercially underexploited despite possessing pleasant flavors and aromas. Among these are jabuticaba (Plinia jaboticaba (Vell.) Kausel) and jussara (Euterpe edulis Mart.), which are native to Brazil and commonly found in the Atlantic Forest. In contrast, blueberry (Vaccinium myrtillus L.), native to the northern hemisphere, has been successfully cultivated in Brazil, benefiting from the tropical climate, which allows for year-round production. These three fruits are characterized by their high concentrations of phenolic compounds, particulary anthocyanins, and their notable antioxidant capacity (Inada et al. 2020; Pires et al. 2020; Vannuchi et al. 2021).

The jabuticaba fruit is similar in shape to a grape and exhibits a deep purple to black color. Its skin serves as a potential source of natural antioxidants. Jabuticaba fruit matures rapidly, requiring immediate consumption after harvest. Although jabuticaba is popular thoughout Brazil, its commercial value remains low due to its high perishablility, typically deteriorating within tree days post-harvest. This degradation is attributed to water loss, microbial spoilage, and pulp fermentation. The anthocyanins, particularly cyanidin-3-glucoside, delphinidin-3-glucoside and peonidin-3-glucoside, are the primary constituents of jabuticaba, abd are predominanthy concentrated in the skin (Inada et al. 2020; Neves et al. 2018a).

The jussara fruit is round, with a purple pulp enveloping a single seed. Despite its abundance, jussara is infrequently consumed. The primary anthocyanins present in jussara fruits are cyanidin-3-glycoside and cyanidin-3-rutinoside, with smaller quantities of cyanidin-3-sambubioside, pelargonidin-3-glycoside, cyanidin-3-rhamnoside and pelargonidin-3-rutinoside are found. Most of the anthocyanins of jussara are similar to those found in the Amazonian açaí (Carvalho et al. 2022; Neves et al. 2022).

Blueberry is regarded as one of the five healthiest foods for humans due its high concentration of phenolic compounds, particularly anthocyanins. The primary anthocyanins identified in blueberry fruits are delphinidin, malvidin and petunidin glycosides, which account for over 90% of the total anthocyanin content (Liu et al. 2021; Yang et al. 2022).

Anthocyanins, water-soluble pigments classified as phenolic compounds, are naturally found in various plant organs, including flowers, fruits, stems, leaves, and roots. These pigments are responsible for the red, purple, and blue colors observed in fruits and flowers. With their potent coloring properties and health benefits—stemming from their high antioxidant capacity—there is a pressing need to incorporate them into food products. This need is driven by the increasing consumer demand for healthier foods and the replacement of synthetic additives with natural alternatives. As a result, numerous studies have focused on incorporating anthocyanins into products such as beverages, confectionery, dairy, and bakery items (Granados-Balbuena et al. 2024; Lin et al. 2023).

However, the food industry faces several challenges in incorporating anthocyanins into products. Among the primary challenges is ensuring a reliable and consistent supply of raw materials that provide sufficient pigment, particularly due to seasonal variability. Maintaining pigment stability during food processing, transportation, and storage also poses significant hurdles (Granados-Balbuena et al. 2024). As anthocyanins are highly unstable, their stability is influenced by the degrees of glycosylation and acylation of the molecules (Yan et al. 2023).

One potential solution to these challenges is using phenolic extracts derived from various raw materials. This approach could reduce reliance on a single anthocyanin source, enhance the concentration of anthocyanins in the extract, and improve the diversity of anthocyanin compounds. Such diversification may lead to different compositions and varying degrees of stability.

Thus, the present study blended the phenolic extracts of blueberry, jussara, and jabuticaba skins to obtain a mixture of the extracts with a high content of phenolic compounds, total anthocyanins, and antioxidant capacity. The anthocyanins were quantified and identified by high-performance liquid chromatography/mass spectrometry (HPLC/MS).

Materials and methods

Reagents and materials

The jussara and jabuticaba fruits were collected in the rural area of Viçosa, Minas Gerais, Brazil (latitude 20°45′17.52′′S, longitude 42°52′42.88′′W), while the blueberry fruits were sourced from Barbacena, Minas Gerais, Brazil (latitude 21°13′18.95′′S, longitude 43°46′14.44′′W). Folin-Ciocalteu reagent, 2,2’-azinobis-3-ethyl-benzothiazoline-6-sulfonate (ABTS), sodium carbonate and Trolox were obtained from Sigma-Aldrich (St. Louis, MO, USA). Additionally, analytical-grade reagents were utilized, including formic acid, ethanol and potassium persulfate (Vetec Química Fina, Ltda, Rio de Janeiro, Brazil), gallic acid (Analyticals Carlo Erba, Paris, France), hydrochloric acid (HCl; Synth, São Paulo, Brazil) and acetonitrile (Brazil).

Ultrasound-assisted extraction

The blueberries and the jabuticaba skins were ground in a mixer, while the jussara fruit was immersed in water at 40 ± 5 ºC and allowed to stand for 20 min. Following this, the water was drained, and the fruit was macerated to remove the core. The crushed jabuticaba skins and the macerated jussara fruit were weighed at 1:10 (m/v) ratio, whereas the crushed blueberries were weighed at 1:5 (m/v) ratio. All fruit were then added to 60% (m/v) ethanolic solutions, acidified (pH 2.0 ± 0.1) using HCl, and subjected to ultrasonication (40 kHz) at 40 °C for 60 min using an ultrasonic bath (Ultracleaner 1400 A, Unique, Indaiatuba, Brazil). The resulting extracts were vacuum filtered through Whatman No. 1 filter paper, then concentrated using a rotary evaporator (Ika RV 10, LAbScience, Germany) at 45 °C to eliminate residual ethanol. All extracts were concentrated to 6 ºBrix and stored in amber bottles under refrigeration at 5 ± 1 ºC for subsequent analysis.

Identification and quantification of anthocyanins by HPLC-DAD-ESI/MSⁿ

Concentrated phenolic extracts from jabuticaba peel, blueberry, and jussara fruit were prepared using a 60% (w/v) ethanolic solution at a 1:2 (m/v) ratio, following the extraction procedure outlined in Sect. 2.2. The phenolic extracts were analyzed using an Agilent 1100 Series HPLC system (Agilent, Germany), equipped with a diode array detection system (G1315B), an Agilent Trap VL (G2445C VL) LC/MSD ion trap, and an electrospray ionization multi-stage mass spectrometer (ESI-MSⁿ). The system was coupled to an Agilent ChemStation data processor (version B.01.03), while the MS data were processed by Agilent LC/MS software version 5.3. The analyses procedure followed the methodology described by Rebello et al. (2013).

For each phenolic extract, the samples were diluted with 0.1 N HCl in a 1:5 (v/v) ratio and injected into a Zorbax Eclipse XDB-C18 reverse phase column (2.1 × 150 mm, particulate 3.5 μm, Agilent, Germany), thermostated at 40 °C. The mobile phase flow was 0.19 mL min^− 1, using mobile phase A: acetonitrile/water/formic acid (3.0:88.5:8.5, v/v/v) and mobile phase B: acetonitrile/water/formic acid (50.0:41.5:8.5, v/v/v). The elution gradient was as follows: 0 min (97% A and 3% B), 20 min (72% A and 28% B), 34 min (57% A and 43% B), 36–42 min (100% B) and 45 min (97% A and 3% B). To identify anhocyanins, the ESI/MS-MS detector operated in positive ionization mode, using the following parameters: N₂ gas at 8 L min^− 1 flow and 325 °C, and nebulization with N₂ at 50 psi. Ionization and fragmentation were optimized by the direct insertion of standard solutions of malvidin 3,5-diglycoside in the positive ionization mode. Identification was based on spectroscopic data (UV-Vis and MS/MS) and the detected compounds were quantified by chromatographic data at 520 nm using calibration curves constructed authentic standards.

Experimental design of mixtures

The concentrated phenolic extracts of jabuticaba, jussara and blueberry were combined to optimize the total phenolic compounds and anthocyanins content. A simplex-centroid mixture design was employed, and the blends of the three components, i.e., pseudo-components, were performed. All mixtures contained the three components, namely, the jabuticaba pells extract (X₁), the jussara extract (X₂) and the blueberry extract (X₃). Ten randomized trials were conducted, including three replicates at the central point. The proportions of the components in the mixtures (X) ranged from 0.1 to 0.8, as shown in Table 1.

Table 1.

Randomized mixtures obtained by simplex-centroid design {3,3}

Mixtures	Extract ratios (v/v)
	Jab (X1)	Jus(X2)	Bb(X3)	PS1	PS2	PS3
M8	0.333	0.333	0.334	0.333	0.333	0.333
M7	0.670	0.165	0.165	0.566	0.217	0.217
M2	0	1	0.000	0.100	0.800	0.100
M3	0	0	1.000	0.100	0.100	0.800
M6	0.5	0	0.500	0.450	0.100	0.450
M8	0.333	0.333	0.334	0.333	0.333	0.333
M4	0.500	0.500	0	0.450	0.450	0.100
M1	1.000	0.000	0	0.800	0.100	0.100
M10	0.165	0.165	0.670	0.216	0.217	0.567
M9	0.165	0.670	0.165	0.217	0.567	0.217
M5	0.000	0.500	0.500	0.100	0.450	0.450
M8	0.333	0.333	0.334	0.333	0.333	0.333

Jab: phenolic extract of jabuticaba peels; Jus: phenolic extract of jussara; Bb: phenolic extract of blueberry; PS: pseudocomponents

Multiple regression models, linear [Eq. 1] and quadratic [Eq. 2], were employed to analyze and interpret the data. These models were used to evaluate the relationships between the mixture components and the response variables, such as total phenolic content and anthocyanin concentration. The linear model provided a first-order approximation, while the quadratic model accounted for potential interactions and non-linear effects between the phenolic extracts.

1

2

Where Y represents the response function of the experimental data for total phenolics, total anthocyanins, antioxidant capacity and colorimetry, β are the estimated coefficients, and X is the levels of the independent variables.

The models were evaluated with analysis of variance (ANOVA), and for those that did not present a significant lack of adjustment (P < 0.05), the response surface was constructed. Calculations and graphs were performed using Statistica 7^® software (v.7.0, Statsoft Copyright, Inc.).

Analysis of total anthocyanin

The total anthocyanin content of the extract mixture (mg·mL^− 1) was quantified using the method described by Lees and Francis (1972). An aliquot of the extract was diluted with an ethanol:1.5 mol L^− 1 HCl solution (85:15) (v/v), and the absorbance was measured at 535 nm with a UV/vis 1601 PC spectrophotometer (Shimadzu, Kyoto, Japan). The dilution of the samples was adjusted to yield absorbance values between 0.2 and 0.8, in accordance with the Beer−Lambert law.

Analysis of total phenolic compounds

The total phenolic content was determined using the method described by Singleton and Rossi (1965), with modifications. Briefly, a 0.6-mL aliquot of the extract was mixed with 3.0 mL of the Folin-Ciocalteu reagent pre-diluted with distilled water (1:10) (v/v). The mixture was allowed to react for 3 min in the dark at room temperature (25 ± 1 °C). Subsequently, 2.4 mL of 7.5% saturated Na₂CO₃ solution (m/v) was added, and the reaction was allowed to proceed for one hour at room temperature, protected from light. Absorbance was measured at 760 nm. A standart curve for gallic acid was generated, and the results were expressed as mg of gallic acid equivalents (GAE) per mL of extract (mg GAE·mL^− 1).

Antioxidant capacity

The antioxidant activity of the extract mixture (µM·Trolox g^− 1) was determined using the Trolox equivalent antioxidant capacity assay (TEAC), employing the ABTS radical, as described by Re et al. (1999).

Colorimetric coordinates

The color of the extract mixture was determined using a ColorQuest XE colorimeter (HunterLab, Reston, VA, USA), with direct measurement of the L* (luminosity), a* (positive = red, negative = green) and b* (positive = yellow, negative = blue) cordinates. The hue (h*) and saturation (C*) parameters were calculated from the a* and b* values, using Eqs. 3 and 4, respectively.

3

4

Results and discussion

Identification of anthocyanins by HPLC-DAD-ESI/MSⁿ

The anthocyanin, flavonoid and phenolic acid contents of jabuticaba, jussara and blueberry fruits were analysed. Chromatograms of the concentrated phenolic extracts of the jabuticaba (A), jussara (B) and blueberry (C) fruits are shown in Fig. 1. The anthocyanins from each extract were individually identified to assess the diversity of anthocyanins present in the mixture of extracts. Anthocyanins were the major phenolic compounds in jabuticaba, jussara and blueberry fruits. In jabuticaba, the highest concentration of anthocyanins was located in the peel, whereas, for the jussara and blueberry fruit, the highest concentrations were found in the whole fruit, with seeds removed (Inada et al. 2015). Each anthocyanin was identified by comparing spectroscopic data, primarily MS/MS data, with spectra obtained from authentic standards.

Fig. 1.

Chromatograms of the anthocyanins of the fruits of jabuticaba (A), jussara (B) and blueberry (C)

Peaks 1 and 2 in Fig. 1A correspond to the main anthocyanins in jabuticaba fruit. Peak one was identified as delphinidin-3-O-glycoside, comprising 13.4% of the total anthocyanins, while peak two was identified as cyanidin-3-O-glycoside (Table 2), the predominant anthocyanin in jabuticaba fruit, accounting for 86.6% of the total. These findings were consistent with those reported by Inada et al. (2015), where cyanidin-3-O-glycoside and delphinidin-3-O-glycoside were found to represent 82.4% and 17.6%, respectively, of the anthocyanins in jabuticaba peel extracts. Additionally, the same study indicated that cyanidin-3-O-glycoside constituted 34% of the total phenolic content in jabuticaba peels. Other studies also reported cyanidin as the most abundant anthocyanin in jabuticaba fruit (Inada et al. 2020, 2020; Neves et al. 2021; Quatrin et al. 2019).

Table 2.

Identification and quantification of anthocyanins of Jabuticaba. Jussara and blueberry fruits

Sample	Peak	Attribution	Rt(min)	Ms/Ms	Concentration (mg/L)
Jabuticaba	1	dp-3-O-glc4	9.6	465/303	347.850
	2	cy-3-O-glc	12.4	449/287	2250.823
Jussara	3	cy-3-O-glc	12.8	449/287	1286.120
	4	cy-3-O-rut	14.6	595/287	2201.448
Blueberry	5	dp-3-О-gal	7.9	465/303	346.808
	6	dp-3-O-glc	9.8	465/303	10.100
	7	cy-3-O-gal	11.1	449/287	284.119
	8	dp-3-O-ara	11.4	435/303	214.568
	9	cy-3-O-glc	12.9	449/287	32.923
	10	pt-3-O-gal	13.7	479/317	489.282
	11	pt-3-O-glc	15.3	479/317	23.558
	12	pn-3-O-gal	16.3	463/301	169.347
	13	pt-3-O-ara	16.6	449/317	157.608
	14	pn-3-O-glc	17.7	463/301	61.794
	15	mv-3-O-gal	18.2	493/331	986.251
	16	pn-3-O-ara	19.1	433/301	77.047
	17	mv-3-O-glc	19.7	493/331	89.823
	18	mv-3-O-ara	20.9	463/331	31.986

Rt: Retention time in minutes; Fragmented ions (m/z values) in Ms/Ms experiments obtained from the precursor ions generated in Ms experiments; Concentration as equivalent to cyanidin-3-glycoside; dp. Delphinidin; Cy. Cyanidin; Pt. Petunidine. Pn. Peonidin. Mv. Malvidin; Glc. Glycoside; Gal. Galactoside; Ara. Arabinoside; Rut: rutinosid

Peaks 3 and 4 in Fig. 1B represent the anthocyanins identified in jussara fruits. Peak three was identified as cyanidin-3-O-glycoside, and peak four as cyanidin-3-O-rutinoside (Table 2). Both anthocyanins are the same as those found in açaí in the Brazilian Amazon region. Cyanidin-3-O-rutinoside was found to be the major anthocyanin in jussara, accounting for 63.12%, while cyanidin-3-O-glycoside represented 36.87% of the total anthocyanins identified. These results agreed with previous studies on the jussara fruit (Silva et al. 2014; Vieira et al. 2017).

Among the fruits studied, blueberry contained the highest number of identified anthocyanins. Fourteen characteristic peaks were designated as follows: 5, delphinidin-3-О-galactoside; 6, delphinidin-3-О-glycoside; 7, cyanidin-3-O-galactoside; 8, delphinidin-3-О-arabinoside; 9, cyanidin-3-O-glycoside; 10, petunidin-3-O-galactoside; 11, petunidin-3-O-glycoside; 12, peonidin-3-O-galactoside; 13, petunidin-3-O-arabinoside; 14, 3-O-glycoside; 15, malvidin-3-O-galactoside; 16, peonidin-3-O-arabinoside; 17, malvidin-3-O-glycoside and 18, malvidin-3-O-arabinoside. The major anthocyanin in the blueberry fruit was identified as malvidin-3-O-galactoside (33.14%), while delphinidin-3-O-glucoside was detected at the lowest concentration (0.33%). These results are consistent with previous studies by Bunea et al. (2013), Sun et al. (2012), and Wang et al. (2016), who identified 14 individual blueberry anthocyanins with similar compositions. Bunea et al. (2013) reported 12 anthocyanins in seven blueberry cultivars of Vaccinium corymbosum L., with malvidin-3-O-galactoside being the most abundant in two of the seven evaluated varieties, in agreement with the present study. However, in other varieties, delphinidin-3-O-galactoside, petunidin-3-O-galactoside and malvidin-3-O-glucoside were found to be the predominant anthocyanins.

Variation in anthocyanin composition across different fruits is expected. Genetic factors of the species and cultivar primarily determine which anthocyanins are present in the fruit, while environmental factors influence gene expression, resulting in differences in anthocyanin concentration (Jaakola 2013). The greater complexity of anthocyanin composition in blueberry has previously been confirmed, with levels higher than those found in fruits such as pitanga (Eugenia uniflora), araçá (Psidium cattleianum), blackberry (Rubus sp.), and strawberry (Fragaria × ananassa) (Chaves et al. 2018).

Anthocyanins are well-established phenolic compounds with specific functions, including a high antioxidant capacity that is associated with the preventing of various diseases. When blended, the extracts contain a high concentration and a wider diversity of anthocyanins. For instance, cyanidin-3-O-glycoside, which is common to all three fruits, has its concentration enhanced when the extract mixtures are combined compared to the individual extracts. Additionally, cyanidin-3-O-rutinoside, which is present exclusively in jussara fruit, may benefit from this approach by preparing a blend enriched with the jussara extract.

In addition to anthocyanins, a wide variety of phenolic compounds is present in these three fruits. Jussara contains phenolic acids, including hydroxybenzoic, ferulic, p-coumaric, protocatechuic, and vanillic acids, as well as derivatives of gallic and ellagic acids. Futhermore, flavonoids such as catechin, epicatechin and quercetin are also present (Silva et al. 2017; Madalão et al. 2021; Schulz et al. 2021; Vieira et al. 2017). Notably, jussara exhibits a higher concentration of phenolic compounds than E. precatoria and E. oleracea, which are the açaí species from the Amazon region (Borges et al. 2011). Similarly, jabuticaba is rich in non-anthocyanins phenolic compounds, such as flavonoids, including quercetin, rutin, quercitrin, isoquercitrin, and quercimeritrin, in addition to phenolic, gallic, and ellagic acids, and depsides. These latter compounds, composed of two or more aromatic units linked by an ester bond, exhibit various bioactivities, including antibiotic, anti-human immunodeficiency virus, antiproliferative, and anti-inflammatory properties (Inada et al. 2015).

The wide variety of anthocyanins and other phenolic compounds in blueberry fruit account for its high bioactive quality. Vrhovsek et al. (2012) evaluated nine different blueberry cultivars over three growing seasons and identified 23 flavonol glycosides. Among them, six different aglycons (myricetin, quercetin, isorhamnetin, kaempferol, laricitrin, and syringetin) were identified, along with their conjugation with five different sugars (galactoside, glucoside, rhamnoside, pentoside and rutinoside), in addition to the acetylated glucose moieties. The biosynthesis of these flavonol glycosides was found to be controlled by the genotype. Quercetin was determined to be the most abundant flavonol across all varieties, while the concentration of the other flavonols was variety-dependent. The diversity and complexity of flavonols in blueberries were attributed to at least two factors: the variety of aglycones and the diversity of glycosides.

Despite the few studies involving the mixture of phenolic extracts from different raw materials; it is suggested that increased phenolic complexity may enhance intermolecular pigmentation reactions. These reactions, defined as non-covalent interactions between anthocyanin and non-anthocyanin molecules, result in the formation of pigment-copigment complexes driven by van der Waals interactions. Such interactions have been shown to increase pigment absorption in the visible spectrum (hyperchromic effect) and shift the visible absorption maximum to a longer wavelength (bathochromic effect). This provides a broader range of colors for exploration and serves as a method to improve pigment stability, thereby promoting its application as a food colorant (Zhao et al. 2020).

Effect of extract mixtures on total anthocyanins, total phenolics, antioxidant capacity, and color

The effects of the blended phenolic extracts from jabuticaba, jussara and blueberry on total anthocyanins content, total phenolics content, antioxidant capacity, and colorimetric coordinate values were evaluated (Table 3). The mixtures M9 and M2 exhibited the highest total anthocyanin contents. The M2 mixture, characterized by the highest proportion of jussara phenolic extract, followed by the M9 which contained the highest concentration of anthocyanins, demonstrated this trend. A similar pattern was observed for the total phenolics content, highlighting the significant contribution of jussara extract in enhancing the phenolic compounds and anthocyanins within the mixtures.

Table 3.

Mixtures of the concentrated phenolic extracts obtained through simplex centroid planning {3.3}

Mixtures	Proportions of Extracts (v/v)			Variables Answers
	Jabuticaba (X1)	Jussara (X2)	Blueberry (X3)	Anthocyanins Totals (mg·mL− 1)	Phenolic Totals (mg EAG·mL− 1)	Antioxidant Capacity (µM·g− 1)	L*	a*	b*	C*	h*
M1	0.800	0.100	0.100	0.92	5.29	65.57	27.18	4.52	0.29	4.52	3.67
M2	0.100	0.800	0.100	4.44	8.97	93.88	25.02	0.43	-0.65	0.78	-56.51
M3	0.100	0.100	0.800	0.85	2.27	45.22	26.5	3.1	0.57	3.15	10.42
M4	0.450	0.450	0.100	2.55	6.73	114.19	26.44	0.66	-0.04	0.66	-3.47
M5	0.100	0.450	0.450	2.73	6.05	58.93	26.4	0.66	-0.06	0.66	-5.19
M6	0.450	0.100	0.450	1.02	5.22	130.04	27.17	3.35	0.79	3.44	13.27
M7	0.567	0.217	0.217	1.52	5.18	93.97	26.89	2.00	0.62	2.09	17.22
M8	0.333	0.333	0.333	2.06	6.22	102.00	28.81	1.15	0.14	1.16	6.94
M8	0.333	0.333	0.333	2.45	6.73	114.36	26.75	1.04	0.13	1.05	7.13
M8	0.333	0.333	0.333	2.53	7.19	129.33	26.69	1.09	0.23	1.11	11.90
M9	0.217	0.567	0.217	5.00	13.07	282.47	25.01	0.56	0.08	0.56	8.13
M10	0.217	0.217	0.567	2.30	6.08	48.40	26.15	1.73	0.81	1.91	25.09
Jab*	1.000	0.000	0.000	0.47	5.00	142.33	29.63	4.35	16.47	17.03	14.79
Jus	0.000	1.000	0.000	6.55	10.98	152.04	26.54	-0.85	0.17	0.86	-78.69
Mir	0.000	0.000	1.000	0.44	1.16	12.32	27.23	0.15	3.91	3.92	2.19

* Jab: extract of skins of jabuticaba; Jus: extract of the fruit of jussara; Mir: extract of the blueberry fruit

The highest antioxidant capacity was observed in the M9 moisture, confirming the relationship between phenolic compounds and antioxidant capacity. A positive correlation was noted between the concentration of phenolic compounds and antioxidant capacity, with higher phenolic content corresponding to greater antioxidant capacity.

Color is recognized as a critical factor in food preference, necessitating its thorough analysis. Based on the colorimetric data presented in Table 3, all blends displayed a predominantly purple hue, primarily attributed to jussara extract’s influence. The hue angle (h*) varies between 0 and 360º, where 0º and 360º correspond to pure red, 90º to pure yellow, 180º to pure green, and 270º to pure blue. The M2 mixture, which contained the highest proportion of jussara extract, exhibited an h* -56.51 and a* = 0.43. This corresponds to a hue angle of 303.49º (360 + -56.51), indicating a purple coloration in the M2 powder. No significant variations were observed in the L* values, which ranged from 25.01 to 29.63 across the extracts. All mixtures exhibited low L* values (0 = black, 100 = white), reflecting a darker appearance. The M9 mixture, containing the second-highest concentration of jussara extract, showed the lowest L* and C* values, indicating a darker and less saturated color compared to the other mixtures.

All tested models were found to be significant. However, some showed a lack of significant adjustment (p < 0.05). Consequently, only the models that did not show significant lack of fit (p > 0.05) were considered predictive, and contour plots were generated for these models (Fig. 2).

Fig. 2.

Contour diagram of responses that showed a non-significant lack of fit (p < 0.05). 2a: total anthocyanins; 2b: total phenolics compounds; 2c: b* (yellow vs. green); 2d: C* (saturation); 2e: h* (hue)

Graphical analysis (Fig. 2) revealed that, for both total anthocyanins and total phenolics, the most influential factor was the phenolic extract from jussara, followed by the phenolic extract from jabuticaba skins. As illustrated in Fig. 2a and b, increasing the concentration of the jussara phenolic extract (right inferior vertices) resulted in a tendency for total anthocyanins and total phenolics to reach their maximum values. In contrast, increasing the concentration of blueberry phenolic extract (top apex) reduced these values.

This result does not imply that blueberry is not a valuable source of phenolic compounds. On the contrary, it exhibited the highest diversity of anthocyanins among the fruits analyzed, containing 14 different anthocyanins (Table 2). The observed differences can likely be attributed to variations in the preparation of the extracts. Only the peel was used for jabuticaba, while jussara was analyzed without the core. Conversely, the intere blueberry fruit was utilized in the analysis, which may explain the lower values of anthocyanins and total phenolics values observed in the blueberry extract compared to those from jabuticaba and jussara.

For the colorimetric coordinates, the b* values (positive = yellow, negative = blue) were found to be maximized when the blend contained a higher proportion of blueberry phenolic extract and a lower proportion of jussara phenolic extract (Fig. 2c). For the C* coordinate, the optimal region was identified with approximately 0.7 jussara phenolic extract, 0.05 blueberry phenolic extract and 0.20 of jabuticaba skin phenolic extract. Thus, increasing the content of the jussara phenolic extract resulted in a decrease in vector values within the colorimetric plane.

The pitch angle values further corroborated the C* results. As the concentration of jussara phenolic extract increased, the h* values decreased, producing negative values and positioning the samples in the fourth quadrant (270−360°), confirming the dominant influence of jussara extract on the color of the mixtures.

In Fig. 2e, the optimal region for the h*value can be observed, corresponding to approximately 0.50 blueberry phenolic extract, 0.35 jabuticaba skin phenolic extract, and 0.15 jussara phenolic extract. Lower concentrations of the jussara phenolic extract resulted in higher hue angles, causing the powders to exhibit a redder hue.

Table 4 presents the correlation coefficients between the variables. The results indicated that the jussara phenolic extract was highly correlated with the total anthocyanins (r = 0.95), total phenolics (r = 0.88), and antioxidant capacity (r = 0.78). These findings confirm that the jussara phenolic extracts exerted a more significant influence on the bioactive composition of the blended extracts compared to jabuticaba and blueberry extracts.

Table 4.

Correlation between the dependent and independent variables evaluated during the study

	Jab	Jus	Blu	TA	TP	AC	L*	a*	b*	C*	h*
Jab	-	0.36	0.36	0.52	0.66	0.65	0.77	0.84	0.57	0.83	0.32
Jus	0.36	-	0.36	0.95	0.88	0.78	0.74	0.29	-0.10	0.31	-0.38
Mir	0.36	0.36	-	0.54	0.58	0.54	0.76	0.70	0.71	0.70	0.42
TA	0.52	0.95	0.54	-	0.98	0.91	0.87	0.49	0.19	0.51	-0.10
TP	0.66	0.88	0.58	0.98	-	0.96	0.93	0.62	0.33	0.64	0.04
AC	0.65	0.78	0.54	0.91	0.96	-	0.86	0.59	0.38	0.60	0.16
L*	0.77	0.74	0.76	0.87	0.93	0.86	-	0.81	0.53	0.82	0.18
a*	0.84	0.29	0.70	0.49	0.62	0.59	0.81	-	0.73	1.00	0.35
b*	0.57	-0.10	0.71	0.19	0.33	0.38	0.53	0.73	-	0.72	0.82
C*	0.83	0.31	0.70	0.51	0.64	0.60	0.82	1.00	0.72	-	0.32
h*	0.32	-0.38	0.42	-0.10	0.04	0.16	0.18	0.35	0.82	0.32	-

#Jab: concentrated phenolic extract of jabuticaba skins; Jus: concentrated phenolic extract of jussara; Blu: phenolic extract of blueberry concentrate. TA: Total anthocyanins; TP: Total phenolics; AC: Antioxidant Capacity

While the phenolic extracts of jabuticaba and blueberry were positively correlated with all measured parameters, the correlation values for the blueberry phenolic extract were lower than 0.60 for total phenolics, anthocyanins, and antioxidant capacity. In the case of jabuticaba, non-anthocyanin compounds contributed more significantly to antioxidant capacity than anthocyanins. These findings align with the work of Neves et al. (2018), who also reported a positive correlation between non-anthocyanins phenolic compounds and the antioxidant capacity of jabuticaba.

The phenolic compounds demonstrated a strong correlation with both antioxidant capacity and total anthocyanins, with correlation coefficients of 0.96 and 0.98, respectively. Despite the strong correlation between the phenolic compounds and anthocyanins; these two variables exhibited a non-significant adjustment (p > 0.05), which accounts for the variability observed in the data across the different mixtures.

The phenolic extract of jabuticaba exhibited a strong correlation with the a* parameter, whereas the jussara extract showed a weaker correlation (Table 4). For the b* and h* parameters, the jussara extract displayed negative correlations. Total anthocyanins were poorly correlated with the parameters b* and h*. A high anthocyanin content typically corresponds with a low extract pH (~ 2.0), leaging to an intense red appearance and, consequently, reduced h* and b* values.

All significant factors (p < 0.05) were found to influence the reponse variables of blended extracts, either positively or negatively. The jussara phenolic extract all evaluated responses, positively impacting total anthocyanins, total phenolic contents, and C* values while negatively affecting the b* and h* values.

The variables listed in Table 4 showed a non-significant adjustment (p < 0.05). All models were influenced by the presence of the jussara extract, which contributed a high content of anthocyanins and phenolic compounds.

Based on the data obtained and the enhanced understanding of each raw material’s contributions to the characteristics of the extract mixtures, it is feasible to develop mixtures tailored to meet the specific requirements of different food formulations, such as color, antioxidant capacity, or total phenolic compound concentration. Consequently, it is recommended that future studies include stability tests for each mixture to elucidate better the interactions among the phenolic extracts of these three fruits.

Conclusions

The blending of jussara, blueberry, and jabuticaba phenolic extracts resulted in mixtures with an increased total anthocyanins and phenolic compounds content, along with a greater diversity of anthocyanins non-anthocyanin phenolic compounds. Among the identified anthocyanins, cyanidin-3-glycoside was predominated and common to all three fruits. Thus, utilizing this blend of phenolic extracts as a natural dye presents a viable alternative for the development of healthy-promoting foods and the valorization of jussara, blueberry and jabuticaba skins.

Biological and chemical

ABTS

(2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonate))

HCl

Chloridric Acid

GAE

Gallic Acid Equivalent

TEAC

Trolox Equivalent Antioxidant Capacity

Trolox

6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid

Instrumental techniques

HPLC

High Performance Liquid Chromatography

DAD

Diode Array Detector

ESI

Electrospray Ionization

MS

Mass Spectrometry

LC

Liquid Crhromatografy

UV

Ultraviolet

Vis

Visible

a*

Positive = red, negativa = green byCIELab system

b*

Positive = yellow, negative = blue by CIELac system

C*

Color Saturation

h*

hue

L*

Luminosity

Author contributions

JCGR and NAN carried out the experiments; LLRB and ALAAN analyzed the data and wrote the manuscript; FARB, SGA and PCS idealized the work, provided technical support and supervised the work.

Funding

This work did not have financial support from any funding agency, except for granting a doctoral scholarship to Juliana de Cássia Gomes Rocha, given by the Brazilian agency CAPES “Coordination of Superior Level Staff Improvement.”

Data availability

Not applicable.

Code availability

Not applicable.

Declarations

Ethical approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Conflict of interest

The authors declare no conflict of interest.