Characterization and Optimization of the Ultrasound-Assisted Extraction Process of an Unexplored Amazonian Drupe (Chondrodendron tomentosum): A Novel Source of Anthocyanins and Phenolic Compounds

Abstract

This study presents the first comprehensive physicochemical and bioactive characterization of the fruit of Chondrodendron tomentosum Ruiz & Pav. (Menispermaceae). Biometric and physicochemical parameters were characterized across three fruit ripening stages (green, turning, ripe). Additionally, proximate composition was determined in ripe fruits, and methanol concentration (25–75%), ultrasonic amplitude (30–70%), and time (1–15 min) were optimized using response surface methodology with a Box–Behnken design. During ripening, weight increased by +47.7% (3.89 to 5.74 g; p < 0.0001), TSS by +26.1% (7.00 to 8.83 °Brix), pH decreased by 32.0% (6.28 to 4.27), and acidity increased by 276% (0.25 to 0.94%). The quadratic models demonstrated high predictive accuracy (R² > 96.5%; p < 0.004). Optimal conditions (57% methanol, 70% amplitude, and 15 min) maximized total anthocyanin content (120.71 ± 1.89 mg cyanidin-3-glucoside/L), total phenols (672.46 ± 5.84 mg GAE/100 g), and DPPH radical scavenging capacity (5857.55 ± 60.20 µmol Trolox/100 g) in ripe fruits. Unripe fruits do not contain anthocyanins, reaching 46.01 mg C3G/L in turning fruits and 120.71 mg/L in ripe fruits (162% higher than turning fruits). Principal component analysis (90.6% variance) revealed synchronized co-accumulation of anthocyanins and phenols, enhanced by vacuolar acidification. These results suggest ripe C. tomentosum fruits as a potential source for natural colorants, nutraceuticals, and functional foods, pending prior development of green, human-safe extraction processes.

1. Introduction

Chondrodendron tomentosum Ruiz & Pav. (Menispermaceae), locally known in northeastern Perú as “uva de montaña” (mountain grape), is an Amazonian woody liana that produces dark purple fleshy drupes arranged in infructescences of 3–6 units [1,2]. According to localized ethnobotanical information collected in Andean–Amazonian communities (Supplementary Material S4), the ripe fruits are occasionally consumed in a restricted local context, where they are described as having a sweet-tart flavor and ripening between July and September in montane forests at 1500–2000 m above sea level (m a.s.l.) [3]. These observations do not imply general edibility or toxicological safety. Despite this localized knowledge, the physicochemical and bioactive characterization of C. tomentosum fruits remains largely unexplored in the scientific literature, classifying this species as understudied from a food-processing and bioactive-compound perspective.

Historically, the bark, stems, and roots of C. tomentosum have been used by indigenous American peoples as arrow poison due to their capacity to induce rapid immobilization through neuromuscular blockade [4,5], associated with the content of d-tubocurarine, a monoquaternary alkaloid [5,6].

Wild Amazonian and Andean fruits constitute rich sources of phenolic compounds [7,8], including anthocyanins (in pigmented species) [9,10], organic acids, and vitamins [8,10], with recognized antioxidant [7,8], anti-inflammatory [7], and anticancer properties [11]. These secondary metabolites confer nutraceutical value and position wild species as candidates for functional food development [12]. Additionally, their resistance to adverse climatic conditions and natural immunity to pathogens represent significant agronomic advantages for sustainable production systems [13].

C. tomentosum remains absent from indexed scientific literature: as of November 2025, no studies on fruit composition, bioactive compounds, or extraction methodologies have been reported in databases such as Scopus or Web of Science. Ethnobotanical and taxonomic records exist in grey literature; however, these do not report physicochemical or bioactive characterizations of the fruit, which constitute the focus of the present study.

Ultrasound-assisted extraction (UAE) is gaining in importance as a green technology for extracting bioactive compounds from plant matrices by enhancing extraction time, yield, and selectivity over traditional techniques [14,15]. The acoustic cavitation of the ultrasonic probe system, breaking down cell walls, increases mass transfer and reduces thermal degradation of labile metabolites [16,17]. Various berries have shown promising efficiency in collecting anthocyanins and total phenols: Prunus spinose [18], Aronia melanocarpa [19], Myrciaria cauliflora [20], Myrtus communis [21], and Morus nigra [22].

Various solvents have been employed in the UAE of fruits, with methanol standing out for its high efficiency in extracting anthocyanins and total phenolic compounds. Although it is a class 2 solvent due to its toxicity [23], it is recommended for analytical purposes due to its favorable physicochemical properties [24]. Therefore, the results should be considered as a reference study to understand the process and guide future industrial proposals.

The goal of this research was to analyze the biometric and physicochemical properties of C. tomentosum at three maturity stages (green, turning, and ripe), as well as to assess them along with the proximate characterization of ripe fruits. Ultrasound-assisted extraction (UAE) of phenolic compounds, anthocyanins, and antioxidant capacity from ripe fruits was optimized through a Box–Behnken design, evaluating the effect of methanol concentration (25–75%), ultrasonic amplitude (30–70%), and extraction time (1–15 min). Once optimized, the optimal UAE conditions were validated and applied on green and turning stages to evaluate the variation of bioactive compounds during ripening. This study constitutes the first detailed scientific report addressing the biometric, physicochemical, proximate, and phytochemical characterization of this species, together with the optimization of green extraction technologies for bioactive compound recovery.

2. Materials and Methods

2.1. Plant Material

Wild fruits of C. tomentosum Ruiz & Pav. were collected at green (0), turning (1), and ripe (2) maturity stages (Figure 1) from a montane forest located in the village of Villa Rica, San José de Lourdes district, San Ignacio province, Cajamarca, Peru (UTM 733700.38 m E, 9,440,384.16 m N; Zone 17S; 1626 m a.s.l.) (Figure 2). An initial collection was conducted in August 2024, during which taxonomic identification was performed, and a reference specimen was deposited at the Isidoro Sánchez Vega Herbarium (ISV), Universidad Nacional de Jaén, Peru (Supplementary Material S1), resulting in the issuance of the Botanical Determination Certificate No. 05-2024 (Supplementary Material S2). Samples were transported under refrigeration (4 ± 1 °C) to the Nutrition and Toxicology Laboratory of the Instituto de Investigación Tecnológico Agroindustrial (IITA), Universidad Nacional del Santa, Chimbote, Perú, for immediate conditioning and processing. Subsequently, in December 2025, a second collection was carried out at the same georeferenced locations to enable taxonomic re-identification and to confirm the consistency of species determination across both sampling periods (Supplementary Material S3). Only healthy fruits free from mechanical damage, bruising, or visible signs of insect or pathogen attack were selected.

Additionally, ethnobotanical interviews were conducted at the same collection site with local inhabitants of the village of Villa Rica to verify key traditional knowledge related to the fruits. These interviews aimed to document the customary forms of fruit consumption, the length of time the fruits have been consumed by the local population, and whether any cases of intoxication or consumption-related illnesses have been reported. The information obtained provided contextual evidence supporting local use patterns and perceptions of safety and was analyzed qualitatively. A detailed description of the interview protocol, informed consent procedure, and anonymized testimonies is provided in Supplementary Material S4.

2.2. Chemical Reagents

Folin-Ciocalteu reagent, 2,2-diphenyl-1-picrylhydrazyl (DPPH), (±)-6-hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid (Trolox), HPLC-grade methanol, potassium chloride, sodium acetate trihydrate, hydrochloric acid (37%), and glacial acetic acid were supplied by Sigma-Aldrich (St. Louis, MO, USA). Anhydrous sodium carbonate was purchased from J.T. Baker (Phillipsburg, NJ, USA), and gallic acid from Merck KGaA (Darmstadt, Germany). Sodium hydroxide and phenolphthalein indicators were of analytical grade.

2.3. Classification of Maturity Stages and Sample Preparation

Fruits were washed with abundant running water, disinfected with sodium hypochlorite solution (80 mg/L) for 2 min, and rinsed with distilled water. Classification by maturity stage was established through a visual colorimetric scale adapted from the maturity index (MI) of olives Olea europaea L. [25], defining three categories according to epicarp pigmentation: green (index 0)—uniform green epidermis without pigmentation; turning (index 1)—epidermis with 25–75% purple-reddish pigmentation; ripe (index 2)—uniform dark purple epidermis with >95% pigmented surface (Figure 1).

2.4. Morphological Characterization

Visual observations were performed to describe the C. tomentosum plant, as well as the shape, color, and texture of the fruit and seed at three ripening stages [26,27]

2.5. Biometric Characterization

Fifty fruits per maturity stage were evaluated (n = 50). The recorded variables were fresh weight (g), polar diameter (cm), and equatorial diameter (cm). Fresh weight was determined on a precision balance (LX-4200C, Precisa Gravimetrics AG, Dietikon, Switzerland). Diameters were measured with a stainless steel vernier caliper (Pretul, Truper S.A. de C.V., Mexico City, Mexico) [28]. Polar diameter corresponded to the longitudinal peduncle-style axis; equatorial diameter was measured perpendicularly at the widest zone of the fruit.

2.6. Physicochemical Characteristics

Fruits from each maturity stage were homogenized in a mortar to obtain a purée. Vitamin C content was determined by titration using the AOAC 967.21 (2019) [29] method. pH (AOAC 981.12), total soluble solids (AOAC 932.12), and titratable acidity, expressed as citric acid equivalents (AOAC 942.15), were developed following standard AOAC methods (2023) [30]. pH was measured with a benchtop pH meter (Hanna, HI2211, Romania) previously calibrated with pH 4.0 and 7.0 buffer solutions. A manual refractometer (Wmeters, REF108, China) was used to estimate total soluble solids (TSS), and values were expressed in °Brix at 20 °C. Titratable acidity (TA) was determined by titrating the sample with an acid-base solution with 0.1 N NaOH to pH 8.2, and results were expressed as a percentage of citric acid. The maturity index (MI) was calculated as the TSS/TA ratio [31].

2.7. Proximate Composition

Proximate characterization was performed only on ripe fruits (n = 3). Moisture was quantified by gravimetry after drying samples in an oven (POL-EKO/SLW 240, Poland) following the AOAC 945.15 (2023) [30] method. Ash was determined by incineration in a muffle furnace (Protherm Furnaces, PLF 110/6, France) at 550 °C using AOAC 223.03 (2023) [30]. Crude protein was estimated using a Kjeldahl apparatus (Labconco, 6030001, USA) following AOAC 954.01 (2019) [32] with 6.25 as the nitrogen conversion factor. Crude fiber was determined by acid–base hydrolysis following NTP 205.003 (2011) [33], and crude fat by Soxhlet extraction (Marconi, MA-491, Brazil) with petroleum ether as solvent using AOAC 930.09 (2019) [34]. Carbohydrate content was estimated by difference according to the AOAC 986.25 (2023) [30] method using Equation (1).

$$Carbohydrates (\%)=100-(moisture+protein+fat+fiber+ash)$$

(1)

2.8. Ultrasound-Assisted Extraction (UAE)

The UAE protocol was adapted from Christou et al. [35] and Razola et al. [36]. Seeds were manually removed, and the pericarp (pulp and peel) was homogenized in a mortar. A total of 1.00 g of sample was weighed into a 50 mL beaker, adding the methanolic solution at a 1:30 (w/v) ratio. Extraction was performed using a USCG 1800 ultrasonic homogenizer (Infitek, China) equipped with a titanium sonotrode (Ø 18 mm, frequency 24 kHz, nominal power 2500 W). The probe was immersed to approximately 50% of the volume without contact with the vessel walls, operating in continuous mode. Methanol (%, v/v), amplitude (%), and time (min) conditions varied according to the Box–Behnken design. Subsequently, extracts were centrifuged (Sigma 4-16KS, Osterode, Germany) at 4200 rpm, 4 °C for 20 min. Supernatants were filtered (Whatman No. 42 paper), protected from light in aluminum foil-wrapped Falcon tubes, and stored at −20 °C until respective analyses.

2.9. Quantification of Total Phenols

Total phenols were determined by the Folin–Ciocalteu spectrophotometric method [37] with modifications. A calibration curve was prepared with gallic acid (0–300 μg/mL in 50% methanol). In amber microtubes, 80 µL of extract was mixed with 160 µL of Folin-Ciocalteu reagent (10% v/v), incubating for 5 min in darkness. A total of 640 µL of Na₂CO₃ (7.5% w/v) was added, and the mixture was incubated for 60 min in darkness at room temperature. Absorbance was measured at 765 nm in a microplate reader (Synergy™ H1, BioTek Instruments, Inc., Winooski, VT, USA). Results were expressed as mg gallic acid equivalents (GAE)/100 g fresh weight basis.

2.10. Quantification of Monomeric Anthocyanins

Total anthocyanins were quantified by the pH differential method [38]. Buffer solutions were prepared: potassium chloride (pH 1.0, 0.025 M) adjusted with concentrated HCl and sodium acetate (pH 4.5, 0.4 M) adjusted with glacial acetic acid. 500 µL of extract was mixed with 500 µL of each buffer independently, homogenizing in a vortex mixer (MS 3 digital, IKA^® Werke GmbH & Co. KG, Staufen, Germany). Absorbances were measured at 510 nm and 700 nm in the microplate reader. Monomeric anthocyanin concentration was calculated using Equations (2) and (3):

$$\mathrm{A} = {(A510- A700\mathrm{nm})}_{\mathrm{pH} 1.0} -{(A510 - A700\mathrm{nm})}_{\mathrm{pH} 4.5}$$

(2)

$$\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{A}\mathrm{n}\mathrm{t}\mathrm{h}\mathrm{o}\mathrm{c}\mathrm{y}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{s}\left({\displaystyle \frac{\mathrm{mg}}{\mathrm{L}}}\right)={\displaystyle \frac{\mathrm{A}\times \mathrm{PM}\times \mathrm{FD}\times 1000}{(Ԑ\times 1)}}$$

(3)

where A is the sample absorbance, MW = 449.2 g/mol (cyanidin-3-glucoside), DF = dilution factor, ε = 26,900 L·mol⁻¹·cm⁻¹, and l = path length in cm. Results were expressed as mg cyanidin-3-glucoside equivalents (C3G)/L.

2.11. Antioxidant Capacity by DPPH

Antioxidant activity was evaluated by the DPPH method [39]. A Trolox stock solution (1 mM) was prepared by dissolving 0.0125 g in 50 mL of methanol. A calibration curve was constructed with Trolox (5–500 µM). A methanolic DPPH solution (1 mM) was prepared, adjusting absorbance to 0.800 ± 0.020 at 515 nm with methanol. In a 96-well microplate, 10 µL of diluted extract (1:5) was mixed with 190 µL of DPPH solution (final volume 200 µL). The mixture was incubated for 30 min in darkness at room temperature, and absorbance was measured at 515 nm in a microplate reader (Synergy™ H1, BioTek Instruments, Inc., Winooski, VT, USA). Results were expressed as µmol Trolox equivalents (TE)/100 g fresh weight basis.

2.12. Experimental Design and Statistical Analysis

RSM using a three-factor Box–Behnken design (15 experimental runs) was applied to optimize UAE. Independent variables were methanol concentration (A: 25, 50, 75%, v/v), ultrasonic amplitude (B: 30, 50, 70%), and extraction time (C: 1, 8, 15 min). Response variables were anthocyanins (mg C3G/L), total phenols (mg GAE/100 g), and DPPH radical scavenging capacity (µmol TE/100 g). All experimental runs were carried out independently in triplicate (n = 3). For RSM modeling purposes, the mean value of each run was used to fit the quadratic polynomial model in coded variables (Equation (4)):

$$y={\beta }_0+{\beta }_{A}A+{\beta }_{B}B+{\beta }_{C}C+{\beta }_{AA}{A}^2+{\beta }_{BB}{B}^2+{\beta }_{CC}{C}^2+{\beta }_{AB}AB+{\beta }_{AC}AC+{\beta }_{BC}BC$$

(4)

where y represents the predicted value of each response variable (anthocyanins, total phenols, or antioxidant capacity), analyzed independently, A, B, and C are coded independent variables, β₀ is the intercept, β_A, β_B, and β_C are linear coefficients, β_AA, β_BB, and β_CC are quadratic coefficients, and β_AB, β_AC, and β_BC are interaction coefficients. Optimal conditions were determined using Derringer–Suich’s desirability function, simultaneously maximizing the three responses. The pure experimental error was estimated from the variability among the replicated center point runs, as is standard practice in Box–Behnken designs. The adequacy of the fitted model was assessed using the coefficient of determination (R²) and the adjusted coefficient of determination (adjusted R²). The optimal conditions predicted by the model were validated through an independent experimental assay performed in triplicate (n = 3). Experimental responses obtained under the optimal conditions were compared with the model-predicted values. Analyses were performed using Statgraphics Centurion software (version XVI.I; Statgraphics Technologies, Inc., The Plains, VA, USA).

Biometric and physicochemical data were analyzed by one-way ANOVA and Tukey’s multiple comparison test (α = 0.05) using Minitab software (version 14; Minitab LLC, State College, PA, USA).

Multivariate exploration was performed through principal component analysis (PCA) on a standardized data matrix (mean-centered, scaled to unit variance) comprising vitamin C, TSS, pH, TA, MI, anthocyanins, total phenols, and antioxidant capacity. A z-score heatmap was generated with annotation by maturity stage. Analyses were executed in R software (version 4.5.1; R Foundation for Statistical Computing, Vienna, Austria) with factoextra and ggplot2 packages.

3. Results and Discussion

3.1. Morphological, Biometric, and Physicochemical Characterization During Ripening

3.1.1. Plant and Fruit Morphology

C. tomentosum is a climbing woody liana (Figure 3A) characterized by large cordate leaves (Figure 3B) and small whitish-green flowers arranged in axillary inflorescences (Figure 3C) [1,2].

As previously shown in Figure 3, C. tomentosum is a climbing woody liana characterized by large cordate leaves and small whitish-green flowers arranged in axillary inflorescences [1,2]. Figure 4 complements these observations by illustrating the reproductive structures of the species. In the reproductive stage, C. tomentosum produces clusters of green immature fruits directly attached to the liana (Figure 4A), which develop into dark purple fleshy ovoid fruits grouped in clusters of 3–6 units (Figure 4B) [3]. Each fruit contains a single lunate seed with hard testa, characteristic of Menispermaceae (Figure 4C).

The ontogenic transition from green to ripe involved pronounced changes in epicarp pigmentation and mesocarp firmness (

Table 1. Morphological characteristics of mountain grape fruits at three maturity stages.

Characteristic	Green (Index 0)	Turning (Index 1)	Ripe (Index 2)
Images
Shape	Ovoid	Ovoid	Ovoid
Epicarp	Thin, intense green, rough	Thin, yellowish-red, smooth	Thin, blackish-violet, smooth, glossy
Mesocarp	Pale green, firm, darkens upon cutting	Red, semi-fleshy, slightly soft	Garnet, scarce, soft
Endocarp	Lunate seed, whitish-yellowish testa, hard	Lunate seed, light yellow testa, hard	Lunate seed covered with pulp, light yellow testa, hard

). The epicarp evolved from intense green and rough (green) to smooth yellowish-red (turning) and glossy blackish-violet (ripe). This may be due to chlorophyll degradation and anthocyanin synthesis via phenylpropanoid [40,41]. The mesocarp transitioned from pale green and firm (green) to semi-soft red (turning) and garnet (ripe), indicating pectin hydrolysis and accumulation of soluble sugars and organic acids [11,42]. The seed maintained constant lunate morphology and hardness, with testa evolving from whitish yellowish to light yellow.

Visual classification by colorimetric index (green = 0, turning = 1, ripe = 2) is used in drupes such as Olea europaea and Prunus spinosa, where epicarp pigmentation correlates with anthocyanin content, firmness, and physicochemical parameters [18,42]. In C. tomentosum, the chromatic transition green → yellowish-red → blackish-violet represents a phenotypic marker of physiological ripening (

Table 2. Biometric and physicochemical parameters of mountain grape fruits at different maturity stages.

Parameter	Green	Turning	Ripe	p-Value
Biometric characteristics
Polar diameter (cm)	2.34 ± 0.15 b	2.45 ± 0.22 a	2.49 ± 0.17 a	<0.0001
Equatorial diameter (cm)	1.71 ± 0.15 b	1.90 ± 0.18 a	1.97 ± 0.14 a	<0.0001
Weight (g)	3.89 ± 0.86 b	5.24 ± 1.31 a	5.74 ± 1.07 a	<0.0001
Physicochemical characteristics
Vitamin C (mg/100 g fresh sample)	5.12 ± 0.01 b	4.78 ± 0.01 c	10.22 ± 0.01 a	<0.0001
TSS (°Brix)	7.00 ± 0.00 c	7.50 ± 0.00 b	8.83 ± 0.29 a	<0.0001
pH	6.28 ± 0.04 a	4.60 ± 0.02 b	4.27 ± 0.02 c	<0.0001
Titratable acidity (%)	0.25 ± 0.01 c	0.62 ± 0.02 b	0.94 ± 0.02 a	<0.0001
Maturity index	27.82 ± 0.83 c	12.13 ± 0.38 b	9.42 ± 0.49 a	<0.0001

Note. Values expressed as mean ± SD (n = 50). Different superscripts in the same row indicate significant differences (Tukey, α = 0.05). TSS: total soluble solids.

).

3.1.2. Biometric Parameters

Fruit dimensions increased progressively during ripening (Table 2). Polar diameter increased from 2.34 ± 0.15 cm (green) to 2.49 ± 0.17 cm (ripe) (+6.4%), while equatorial diameter increased from 1.71 ± 0.15 cm to 1.97 ± 0.14 cm (+15.2%), accompanied by fresh weight gain from 3.89 ± 0.86 g to 5.74 ± 1.07 g (+47.7%; p < 0.0001). This growth pattern may be due to cell expansion and accumulation of soluble solids characteristic of climacteric ripening in drupes [11,42].

Vitamin C content doubled from green to ripe phase (10.22 ± 0.01 vs. 5.12 ± 0.01 mg/100 g FW; +99.6%), This could be related to the activation of ascorbate biosynthesis during late ripening. The same pattern has been observed in strawberries at different stages of ripeness. The authors Fecka et al. (2021) [43] report that the highest ascorbic acid content was found in Selvik red fruits (192.3 mg/100 g), while the lowest was found in green Pandora MAR berries (38.8 mg/100 g). The same was found in fruits such as Vaccinium spp. and other Andean berries [8,11,44]. TSS increased from 7.00 to 8.83 °Brix (+26.1%), which can be associated with the conversion of starch to sugars and the accumulation of soluble carbohydrates [45]. At the same time, pH decreased from 6.28 to 4.27 (−32.0%) and titratable acidity increased from 0.25 to 0.94% (+276%), reflecting organic acid accumulation induced by berry ripening [42,46]. The maturity index (TSS/acidity) decreased from 27.82 to 9.42 because the increase in acidity exceeded sugar synthesis, reflecting a species-specific ripening pattern, opposite to that reported for many temperate climate fruit species, but consistent with tropical species such as Rubus and Vaccinium [11,47]. This behavior is characterized by concurrent accumulation of sugars and organic acids, generating an intense sweet-sour sensory profile; therefore, a low MI at harvest indicates optimal quality and not immaturity. This stage coincides with maximum pigmentation and increased respiratory activity, which requires immediate implementation of cold chain to preserve the chemical and sensory balance achieved [11]. Additionally, the acidic pH (<5.0) and elevated TSS in the ripe stage favor stability of the flavylium cation, maximizing the stability, extractability, and chromatic intensity of anthocyanins [42,46,48], positioning ripe fruits as the stage of greatest technological suitability for extracting bioactive compounds.

3.2. Proximate Composition of Ripe Fruits

Ripe fruits presented moisture of 90.45 ± 0.47%, ash of 2.04 ± 0.14%, crude protein of 1.70 ± 0.02 g/100 g, crude fiber of 2.49 ± 0.01 g/100 g, crude fat of 0.05 ± 0.01 g/100 g, and carbohydrates by difference of 3.01 ± 0.49 g/100 g (dry basis). The high water content classifies C. tomentosum as an aqueous matrix comparable to grumixama (Eugenia brasiliensis, 86.2–89.9%) [49] and superior to cauchao (Amomyrtus luma, 75.3%) [8], favoring hydroalcoholic extraction of water-soluble metabolites [7].

The ash fraction exceeded 3.6–6.0 times the values of cauchao (0.57%) [8] and grumixama (0.34–0.49%) [49], suggesting high mineral content associated with montane volcanic soil [50]. Protein content was comparable to cauchao (1.73 g/100 g) [8] but higher than grumixama (0.65–0.83 g/100 g) [49]. The extremely low lipid content (2.3% of cauchao’s value) confirms the characteristic profile of pigmented drupes where phenolic metabolites predominate over structural lipids [7,49]. This high-moisture, low-lipid matrix minimizes apolar interferences and maximizes solvent diffusion toward epidermal vacuoles rich in anthocyanins [7,8].

3.3. Ultrasound-Assisted Extraction Optimization

3.3.1. Box–Behnken Experimental Design

The Box–Behnken design evaluated 15 combinations of methanol (25–75%, v/v), ultrasonic amplitude (30–70%), and time (1–15 min) on anthocyanins, total phenols, and antioxidant capacity (

Table 3. Box–Behnken design and experimental responses for UAE of mountain grape in ripe stage.

Methanol (%)	Amplitude (%)	Time (min)	Response Variables
			Anthocyanins a	Total Phenols b	DPPH Radical Scavenging Capacity c
25 (−1)	30 (−1)	8 (0)	50.50 ± 1.28	431.45 ± 3.74	3106.72 ± 91.39
25 (−1)	50 (0)	1 (−1)	42.64 ± 0.64	372.05 ± 0.79	1869.09 ± 111.22
25 (−1)	50 (0)	15 (+1)	64.87 ± 2.61	608.97 ± 3.70	3784.06 ± 118.17
25 (−1)	70 (+1)	8 (0)	54.02 ± 1.58	525.86 ± 3.94	3491.38 ± 104.96
50 (0)	30 (−1)	1 (−1)	54.45 ± 0.79	386.91 ± 3.34	2688.60 ± 115.43
50 (0)	30 (−1)	15 (+1)	85.34 ± 2.64	576.94 ± 3.80	3758.98 ± 137.04
50 (0)	50 (0)	8 (0)	77.38 ± 0.72	572.04 ± 3.51	3859.33 ± 148.29
50 (0)	50 (0)	8 (0)	71.93 ± 1.68	516.29 ± 2.40	3725.53 ± 142.78
50 (0)	50 (0)	8 (0)	77.96 ± 1.58	535.42 ± 3.97	3625.18 ± 74.98
50 (0)	70 (+1)	1 (−1)	69.63 ± 1.45	441.83 ± 1.00	2872.57 ± 127.04
50 (0)	70 (+1)	15 (+1)	118.58 ± 2.57	677.67 ± 4.12	5799.38 ± 155.35
75 (+1)	30 (−1)	8 (0)	93.41 ± 2.15	538.73 ± 3.24	3449.57 ± 138.29
75 (+1)	50 (0)	1 (−1)	66.53 ± 1.21	448.27 ± 1.46	3098.35 ± 132.38
75 (+1)	50 (0)	15 (+1)	120.82 ± 1.79	667.21 ± 3.13	5581.96 ± 131.48
75 (+1)	70 (+1)	8 (0)	104.36 ± 2.81	547.09 ± 3.32	4469.78 ± 156.46

Note. Response variables expressed as ^a mg cyanidin-3-glucoside/L; ^b mg GAE/100 g; ^c µmol Trolox/100 g. Values corresponded to mean ± SD (n = 3).

). Response ranges were 42.64–120.82 mg C3G/L, 372.05–677.67 mg GAE/100 g, and 1869.09–5799.38 µmol TE/100 g, respectively, demonstrating variability of 2.8, 1.8, and 3.1-fold between extreme conditions. This second-order factorial design allows modeling linear, quadratic, and interaction effects with a reduced number of experiments compared to full factorial designs [19,20,51].

3.3.2. ANOVA Fit and Regression Coefficients

The quadratic models were significant (p ≤ 0.0037) with R² of 97.62% (anthocyanins), 98.28% (phenols), and 96.57% (antioxidants), without residual autocorrelation (Durbin–Watson: 1.16–2.64; p > 0.05), confirming robust predictive fit comparable to studies in Aronia melanocarpa (R² = 0.99) [19], Myrciaria cauliflora (R² = 0.98) [20], and Prunus domestica (R² = 0.97) [51]. Regression coefficients (see

Table 4. Estimated regression coefficients (β), standard errors (SD), and p-values for quadratic models fitted for bioactive compound extraction and antioxidant capacity.

Term	Anthocyanins (mg Cyanidin-3-Glucoside/L)			Total Phenols (mg GAE/100 g)			Antioxidant Capacity (uMol Trolox/100 g)
	β	SD	p-Value	β	SD	p-Value	β	SD	p-Value
Intercept	75.755	–	–	541.251	–	–	3736.68	(–)	–
A (Methanol)	21.635	−2.179	0.0002	32.871	−7.196	0.006	543.55	−111.294	0.0045
B (Amplitude)	7.862	−2.179	0.0154	32.302	−7.196	0.0065	453.656	−111.294	0.0096
C (Time)	19.544	−2.179	0.0003	110.217	−7.196	<0.0001	1049.47	−111.294	0.0002
A2	−8.466	−6.415	0.2442	−27.183	−21.184	0.2557	−303.831	−327.641	0.3963
B2	8.1	−6.415	0.2624	−33.761	−21.184	0.1719	89.198	−327.641	0.7963
C2	4.387	−6.415	0.5245	−7.067	−21.184	0.7522	−2.787	−327.641	0.9935
AB	1.857	−3.082	0.5731	−21.512	−10.176	0.0882	158.884	−157.394	0.3591
AC	8.016	−3.082	0.0482	−4.495	−10.176	0.6771	142.16	−157.394	0.4078
BC	4.515	−3.082	0.2028	11.454	−10.176	0.3115	464.109	−157.394	0.0319
	R2: 97.623%	Adjusted R2: 93.345%		R2: 98.279%	Adjusted R2: 95.180%		R2: 96.573%	Adjusted R2: 90.404%

Note. A: methanol concentration (% v/v), B: sonication amplitude (%), C: extraction time (min). Bold p-values < 0.05.

) demonstrated the linear effect of methanol (A), amplitude (B), and time (C) on all (p ≤ 0.015) of the three responses, showing a monotonic increase. In summary, the non-significant terms showed no significance in quadratic variables (p ≥ 0.17), indicative of optimal conditions at experimental space limits consistent with Fernández et al. [20] and Vázquez et al. [19]. The A × C interaction enhanced anthocyanins (p = 0.048), and B × C was found to provide the enhanced DPPH radical scavenging capacity (p = 0.032), confirming solvent-time and amplitude-time synergy associated with ultrasonic cavitation mechanisms [51,52]. The predictive equations in coded variables were as follows:

$$Anthocyanins =75.755+21.635A+7.862B+19.544C-8.466A^2+8.1B^2+4.387C^2+1.857AB+8.016AC+4.515BC$$

(5)

$$Total phenols=541.251+32.871A+32.302B+110.217C-27.183A^2-33.302B^2-7.067C^2-21.512AB-4.495AC+11.454BC$$

(6)

$$Antioxidant capacity=3736.68+543.55A+453.656B+1049.47C-303.831A^2+89.198B^2-2.787C^2+158.884AB+142.16AC+464.109BC$$

(7)

Linear effects of methanol (A), amplitude (B), and time (C) were significant for all three responses (p ≤ 0.015), indicating a monotonic increase toward upper levels, like UAE optimizations in berries [20,51,52]. Quadratic terms lacked significance (p ≥ 0.17), suggesting the absence of an interior optimum and confirming that maximum conditions are located at the boundaries of the experimental space [19,53]. Two interactions were relevant: A × C increased anthocyanins (p = 0.048), and B × C increased antioxidant capacity (p = 0.0319), demonstrating synergy between solvent-time and amplitude-time, respectively, consistent with ultrasonic cavitation mechanisms where prolonged times potentiate methanol and mechanical energy effects on cell disruption [20,52].

The absence of statistically significant quadratic terms (p ≥ 0.17), together with the dominance of positive linear effects, indicates the lack of an interior optimum within the studied domain, suggesting that the optimal conditions are located near the boundaries of the experimental space.

3.3.3. Response Surfaces and Mechanistic Interpretation

The 3D response surfaces (Figure 5, Figure 6, Figure 7 and Figure 8) illustrate the combined effects of methanol and amplitude at a fixed time of 15 min on the three responses. The anthocyanin surface (Figure 5) shows progressive increase from 60 mg C3G/L (25% methanol, 30% amplitude) to 151 mg C3G/L (75% methanol, 70% amplitude), with color gradient transitioning from blue (low values) to red (high values). The elongated diagonal contours toward the upper right corner indicate a synergistic effect between solvent concentration and ultrasonic energy. This pattern confirms that elevated methanol concentrations improve solubilization of glycosylated anthocyanins by reducing medium polarity, while higher amplitudes intensify acoustic cavitation that disrupts epidermal vacuoles rich in pigments [20,51].

However, there is a limitation in quantifying anthocyanins only as cyanidin-3-glucoside equivalents using the differential pH method [7,47]. It is recognized that this assumption must be verified in new species to confirm the predominant profile. Therefore, future research should perform specific structural elucidation by HPLC-DAD or LC-MS to accurately characterize the unique pigments of this species.

The high moisture content, characteristic of wild fleshy fruits, does not entail immediate degradation if appropriate postharvest management measures are applied [54]. In this research, fruits were processed immediately after collection, maintained under refrigerated conditions, and protected from light [55,56], procedures that could minimize enzymatic and microbial activity prior to analysis, ensuring stability of the anthocyanins and other phenolic compounds evaluated [7,47].

The antioxidant capacity surface (Figure 6) exhibits analogous behaviour, increasing from 3200 µmol TE/100 g (minimum conditions) to 6200 µmol TE/100 g (maximum conditions), representing a 2.0-fold increase. The linear increase reflects that reducing capacity measured by DPPH increases proportionally with the extraction of total phenolic compounds and anthocyanins, main contributors to antioxidant activity [19,52].

Antioxidant capacity was evaluated exclusively by the DPPH assay, which measures electron transfer to synthetic radicals [8,57]. This approach presents limitations, such as restricted specificity that may not capture other essential antioxidant mechanisms, such as metal chelation or deactivation of reactive oxygen species in complex matrices [8,58].

Total phenols (Figure 7) present a surface with an increase from 500 mg GAE/100 g to 670 mg GAE/100 g, showing greater influence of methanol (coefficient β = 32.87; p = 0.006) than amplitude (β = 32.30; p = 0.007). The more widely spaced contours in the direction of the methanol axis evidence its dominant effect on phenolic recovery. The optimal region is located at the upper boundary of the experimental space (≥65% methanol, ≥60% amplitude), confirming the absence of an interior maximum. Mechanistically, methanol facilitates rupture of ester bonds between phenolics and cell wall polysaccharides, while ultrasound accelerates mass transfer by localized microagitation in bubble collapse zones of cavitation [51,53].

3.3.4. Multi-Response Optimization and Validation

The desirability surface (Figure 8) identifies the optimal region at 57% methanol, 70% amplitude, and 15 min (desirability = 0.9996), represented by the red zone in the upper right corner where the three optimal responses converge. Desirability transitions from 0.4 (blue, minimum conditions) to 1.0 (red, optimal conditions), demonstrating simultaneous maximization without trade-offs between objectives. Experimental validation (

Table 5. Experimental validation of the optimal point (57% MeOH, 70% amplitude, 15 min).

Response	Prediction	Experimental	Error (%)	Lower 95.0% Limit	Upper 95.0% Limit
Anthocyanins (mg C3G/L)	122.79	120.71 ± 1.89	1.69	109.045	136.971
Total Phenols (mg GAE/100 g)	675.65	672.46 ± 5.84	0.47	629.499	721.713
Antioxidants (µmol TE/100 g)	5981.00	5857.31 ± 60.20	2.07	5269.8	6696.03

) confirms 120.71 ± 1.89 mg C3G/L, 672.46 ± 5.84 mg GAE/100 g, and 5857.31 ± 60.20 µmol TE/100 g, with deviations of 1.69%, 0.47%, and 2.07% from predicted values, within acceptable limits (< 5%) for RSM models [52].

Studies on berries and fruit by-products have demonstrated that acidified ethanol–water systems combined with UAE, after optimization, achieve anthocyanin and phenol yields comparable to methanol–water. Through RSM and BBD, moderate conditions of EtOH, temperature, and time have been identified with adequate yields for formulation and valorization [59,60].

3.4. Application of Optimal Conditions Across Maturity Stages

Under optimized UAE conditions (57% methanol, 70% amplitude, 15 min), the maturity stage determined significant differences in all three responses (ANOVA, p < 0.0001;

Table 6. Bioactive compounds extracted under optimal UAE conditions (57% MeOH, 70% amplitude, 15 min) at three maturity stages of mountain grape.

Maturity Stage	Anthocyanins (mg C3G/L) *	Total Phenols (mg GAE/100 g) **	Antioxidant Capacity (µmol TE/100 g) ***
Green (0)	0.00 ± 0.00 c	647.83 ± 3.55 b	5492.63 ± 54.30 b
Turning (3)	46.01 ± 0.70 b	458.86 ± 5.66 c	3581.11 ± 39.80 c
Ripe (7)	120.71 ± 1.89 a	672.46 ± 5.84 a	5857.55 ± 60.20 a

Note. Values represent mean ± SD (n = 3). Different letters in the same column indicate significant differences (Tukey, α = 0.05). * Expressed as cyanidin-3-glucoside equivalents; ** expressed as gallic acid equivalents on fresh weight basis; *** expressed as Trolox equivalents on fresh weight basis.

). No anthocyanins were found in unripe fruits, reached 46.01 ± 0.70 mg C3G/L in turning fruits, and 120.71 ± 1.89 mg C3G/L in ripe fruits (+162% vs. turning; p < 0.0001). This progressive increase could be associated with the activation of the phenylpropanoid biosynthetic pathway during ripening, responsible for anthocyanin synthesis through sequential action of chalcone synthase, flavanone 3-hydroxylase, and UDP-glucose:flavonoid 3-O-glucosyltransferase [41,61,62], possibly, the softening of the cell wall mediated by pectinases and cellulases in mature states facilitated mechanical disruption by ultrasonic cavitation, increasing the release of vacuolar anthocyanins [18,20,63].

Total phenols exhibited a nonlinear pattern during ripening. The turning stage registered the minimum value (458.86 ± 5.66 mg GAE/100 g), while ripe reached the maximum (672.46 ± 5.84 mg GAE/100 g; +46.7% vs. turning; p < 0.0001), slightly exceeding green (647.83 ± 3.55 mg GAE/100 g; +3.8% vs. green). The transient depression in turning coincides with the rapid cell expansion phase, where dilution by water accumulation and metabolic reconfiguration toward anthocyanin biosynthesis temporarily reduce the total phenolic pool [11,44,56]. In contrast, the ripe stage combines net accumulation of structural phenols (hydroxycinnamic acids) with newly synthesized anthocyanins, resulting in maximum phenolic content [7,46].

DPPH radical scavenging capacity followed a trend analogous to total phenols: turning < green < ripe (3581.11 ± 39.80 < 5492.63 ± 54.30 < 5857.55 ± 60.20 µmol TE/100 g). The +63.6% increase between turning and ripe (p < 0.0001) correlates with synchronized accumulation of anthocyanins (+162%) and total phenols (+46.7%), evidencing additive contribution of both metabolite families to reducing activity measured by DPPH [19,56]. The acidic pH of the ripe stage (4.27 ± 0.02; Table 1) additionally favors stability of the flavylium cation (predominant form of anthocyanins at pH < 5), maximizing its antioxidant activity through electronic resonance of the oxonium ring [42,46].

3.5. Principal Component Analysis

PCA on eight physicochemical and bioactive variables explained 90.6% of total variability (PC1 = 67.4%, PC2 = 23.2%; Figure 9). PC1 ordered samples according to ontogenic gradient: green fruits in negative PC1, ripe in positive PC1, turning in intermediate positions. This separation reflects metabolic transition during ripening [11,44].

TSS, vitamin C, anthocyanins, total phenols, antioxidant capacity, and acidity presented positive loadings on PC1, associated with ripe fruits. pH and maturity index showed negative loadings, associated with green fruits. This configuration agrees with Table 1 data: TSS increased +26.1% (7.00 → 8.83 °Brix), vitamin C +99.6%, acidity +276% (0.25 → 0.94%), while pH decreased 32.0% (6.28 → 4.27) and maturity index −66.1% (27.82 → 9.42). This pattern, where acidity exceeds sugar accumulation, is characteristic of Andean and tropical berries of Vaccinium and Rubus, differing from temperate fruits [7,47,56].

PC2 differentiated the turning stage, which was projected separately from green and ripe. This separation captures the transient metabolic minimum in turning total phenols (458.86 mg GAE/100 g; −29.2% vs. green) and DPPH radical scavenging capacity (3581.11 µmol TE/100 g; −34.8% vs. green; Table 5). This valley coincides with rapid cell expansion, where water dilution and biosynthetic reorientation toward anthocyanins temporarily reduce the extractable phenolic pool [11,44]. Vaccinium corymbosum and Rubus ulmifolius exhibit analogous minima at intermediate stages [46,56].

The standardized z-score heatmap confirmed differentiation by maturity (Figure 10). Ripe fruits presented positive z-scores in anthocyanins, total phenols, antioxidant capacity, vitamin C, TSS, and acidity, and negative z-scores in pH and maturity index. This pattern indicates co-accumulation of primary (sugars, ascorbate) and secondary (phenols, anthocyanins) metabolites, characteristic of pigmented berries where ripening integrates shikimate, phenylpropanoid, and ascorbate-glutathione pathways [41,64,65]. Turning showed minimum z-scores exclusively in phenols and antioxidant capacity, confirming its metabolic transition.

The optimization of bioactive compounds under EAU may have been influenced by vacuolar acidification of ripe fruits (pH 4.27; Table 2). Acidic pH stabilizes the flavylium cation (predominant form at pH < 5), preventing conversion to colourless forms and maximizing chromophoric content and antioxidant activity [42,46]. Furthermore, the acidic environment protonates phenolic hydroxyl groups in hydroalcoholic mixtures, allowing greater solubility in them and in optimizing UAE efficiency [18,20]. This association between ripening physiology and extraction method enables ripe fruit to maximize all three responses at the same time (Table 5): 120.71 mg C3G/L (+162% vs. turning), 672.46 mg GAE/100 g (+46.7% vs. turning), and 5857.55 µmol TE/100 g (+63.6% vs. turning).

The findings demonstrate that C. tomentosum ripening has a progressive bioactive compound accumulation model, probably enhanced by vacuolar acidification, with a transient metabolic trough in the shoot tip. This pattern, along with favourable UAE conditions (57% methanol, 70% amplitude, 15 min), turns ripe fruit into raw materials for natural colorants, nutraceuticals, and functional foods from Amazonian berries.

Industrial implementation of laboratory-optimized UAE conditions requires consideration of scale-up criteria such as reactor type (probe vs. bath or continuous flow systems), power density, and thermal control, given that scaling is not linear and requires adjusting specific power to preserve efficiency and equipment integrity [66,67]. In this context, selection and recovery of solvent (methanol vs. ethanol) acquires relevance for operational, regulatory, and environmental viability of the process [68]. Likewise, although optimal conditions maximize experimental yield, evidence supports the use of slightly milder conditions, which reduce energy consumption and equipment wear with moderate yield losses, favoring optimized or pulsed strategies more efficient and suitable for industrial scaling [14,69,70].

4. Conclusions

Box–Behnken optimization established UAE conditions (57% methanol, 70% amplitude, 15 min) that maximized anthocyanin yield (120.71 mg C3G/L), total phenols (672.46 mg GAE/100 g), and DPPH radical scavenging capacity (5857.55 µmol TE/100 g) from ripe C. tomentosum fruits with high predictive accuracy (R² > 0.95, 2.1% deviation). Ripening from green to ripe stage triggered a 162% increase in anthocyanins and 46.7% increase in total phenols, driven by vacuolar acidification (pH 4.27) that stabilizes anthocyanin structure and enhances UAE extractability. The transitory metabolic valley at the turning stage (−29.2% phenols, −34.8% antioxidant capacity) reflects biosynthetic reorganization during cell expansion, establishing an ontogenic pattern where acidification predominates over sugar accumulation. Proximate composition (90.45% moisture, 2.04% ash, 2.49% fiber, 1.70% protein) positions C. tomentosum as a hydrophilic matrix comparable to Andean berries, favoring hydroalcoholic extraction of water-soluble bioactives. These findings highlight the potential chromatic and bioactive value of ripe C. tomentosum fruits, providing a scientific basis for future studies aimed at evaluating their feasibility as sources of natural pigments or functional ingredients, subject to comprehensive toxicological and regulatory assessment. Additionally, detailed phytochemical profiling at the individual compound level is proposed as a research line for future studies.