Extraction Methods, Chemical Characterization, and In Vitro Biological Activities of Plinia cauliflora (Mart.) Kausel Peels

Plinia cauliflora (Mart.) Kausel, popularly known as jabuticaba, possesses bioactive compounds such as flavonoids, tannins, and phenolic acids, known for their antioxidant, antibacterial, wound healing, and cardioprotective effects. Therefore, this study aimed to standardize the P. cauliflora fruit peel extraction method, maximize phenolic constituents, and evaluate their antioxidative and antimicrobial effects. Various extraction methods, including vortex extraction with and without precipitation at 25, 40, and 80 °C, and infusion extraction with and without precipitation, were performed using a completely randomized design. Extraction without precipitation (E − P) showed the highest yield (57.9%). However, the precipitated extraction (E + P) method displayed a yield of 45.9%, higher levels of phenolic derivatives, and enhanced antioxidant capacity. Major compounds, such as D-psicose, D-glucose, and citric acid, were identified through gas chromatography–mass spectrometry (GC-MS) analysis. Ultra-high-performance liquid chromatography–tandem mass spectrometry (UHPLC-MS/MS) analysis identified citric acid, hexose, flavonoids, tannins, and quercetin as the major compounds in the extracts. Furthermore, the extracts exhibited inhibitory effects against Bacillus subtilis, Staphylococcus aureus, Pseudomonas aeruginosa, and Escherichia coli bacteria. In conclusion, the E + P method efficiently obtained extracts with high content of bioactive compounds showing antioxidant and antimicrobial capacities with potential application as a dietary supplement.


Introduction
Plinia cauliflora (Mart.) Kausel, popularly called jabuticaba, has globose berry fruits up to 3 cm in diameter, with white mucilaginous and bittersweet pulp mainly composed of sugars. These fruits contain one to four seeds and have a dark reddish peel, with promising effects on human and animal health [1,2]. P. cauliflora peel extracts have been shown antioxidant [3], antimicrobial [4], and wound-healing properties [5]. Moreover, cardioprotective [6] and hepatoprotective effects [7] were described. Phytochemical analyses predominantly show phenolic compounds, including tannins and organic acids [8].
According to Palozi et al. [3], hydroethanolic extracts of P. cauliflora peel contain ellagic acid, gallic acid, O-desoxy-hexosyl quercetin, and O-hexosyl cyanidin anthocyanins. In According to Palozi et al. [3], hydroethanolic extracts of P. cauliflora peel contain ellagic acid, gallic acid, O-desoxy-hexosyl quercetin, and O-hexosyl cyanidin anthocyanins. In the same study, preclinical tests confirmed the safety of the pharmacological use of these extracts. Owing to their beneficial health effects and lack of toxicity, P. cauliflora peel extracts can be used for daily oral dietary supplementation [8].
The effects of bioactive compounds on health depend on the qualitative and quantitative characteristics and the synergism between these substances, which vary depending on the extraction method used [9]. Therefore, an efficient extraction process can maximize the extraction of bioactive compounds, prevent their degradation, use environmentally friendly technologies, and produce low-cost raw materials [10].
The proper choice of solvent is crucial in producing extracts on an industrial scale, especially when seeking more remarkable purification, requiring inserting a precipitating agent for polysaccharides. In this regard, water is an effective solvent for extraction, allowing easy binding of polysaccharides to water molecules through hydrogen bonds, ensuring their solubility and successful extraction. On the other hand, ethanol is commonly used as a precipitating agent because it reduces the polarity of water, rendering polysaccharides insoluble [11]. Additionally, because of its renewable origin (from sugarcane) and its classification as generally recognized as safe (GRAS), ethanol is suitable for green chemical extraction [12].
Thus, this study aimed to determine the effects of different methodologies on the extraction performance of phenolic derivatives of P. cauliflora fruit peel extracts and identify the phytochemical components and their antioxidant and antimicrobial properties. Overall, this study may pave the way for bioprospecting formulations enriched with P. cauliflora for dietary supplementation.

Yield of the Extracts
The method with the highest yield ( Figure 1) was extraction without precipitation (E − P) (57.9%), followed by vortex extraction without precipitation at 40 °C (T − P40) (52.8%) and extraction with precipitation (E + P) (45.9%). The extraction method with the lowest yield was vortex extraction with precipitation at 80 °C (T + P80) (28.9%).

Characterization by GC-MS and UHPLC-MS/MS
Gas chromatography coupled with mass spectrometry (GC-MS) identified 83 compounds in the P. cauliflora extracts (Table 1) and classified them into the following categories: phenols and derivatives; sugars and derivatives; carboxylic acids and derivatives; alkaloids, fatty acids, glycerolipids, hydroxy acids, and derivatives; keto acids and derivatives; organooxygen compounds; and quinones. The major compounds identified were D-psicose 5TMS, D-glucose 5TMS, citric acid 3TMS, and glycerol 3TMS. GC-MS analysis revealed a predominance (%) of sugars and derivatives, with D-psicose 5TMS as the main constituent in T + P25, T + P40, and E + P, D-(−)-fructofuranose; pentakis (trimethylsilyl) ether (isomer 1) as the main constituent in T-P80; and D-glucose 5 TMS as the main compound in T + P25, T + P40, and E − P. The compounds were identified by searching the library database of Spectra NIST Mass Spectral Library (version 2014), and the main fragments of major compounds were compared with the literature, as shown in Table A1.  A total of 27 compounds were identified using ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) in the positive and negative modes (Tables 2 and 3). These compounds are classified as flavonoids, phenolic acids, tannins, and their derivatives; sugars and their derivatives; carboxylic acids and their derivatives; and alkaloids. UHPLC-MS/MS analysis in negative mode revealed the predominance of carboxylic acids and their derivatives, with citric acid being the major compound. In contrast, the positive mode showed a predominance of flavonoids, with quercetin being the major compound, followed by sugars and derivatives, with hexose and di-hexoside being the most abundant. The ion chromatogram and MS and MS/MS spectra were visualized using Data Analysis 4.3 software and the compounds were identified based on literature data, according to Romão et al. [6]. Table 2. Identification of the constituents of Plinia cauliflora extracts using ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) in positive mode.

Evaluation of Antimicrobial Activity
All extracts showed the ability to inhibit both Gram-positive and Gram-negative bacteria, but did not exhibit antifungal activity against Candida albicans at the evaluated concentrations (Table 6). Overall, E + P exhibited the lowest minimum inhibitory concentration (MIC), followed by T + P25.  Hierarchical clustering was employed to illustrate the variability in the quantification of phenolic compounds and flavonoids and the antioxidant capacity of the methodologies used for the production of P. cauliflora extracts (Figure 2). The samples were grouped into four main clusters and displayed in a dendrogram obtained using the Unweighted Pair Group Method with Arithmetic Mean (UPGMA) using unweighted arithmetic averages.
The first cluster included the E + P method, indicating that the results obtained were more significant in terms of the quantity of phenolic compounds and flavonoids, as well as the antioxidant capacity of this sample, compared with the other methods.
Principal component analysis (PCA) allowed for the joint evaluation of all variables. Multivariate analysis was applied to assess the antioxidant capacity of the extracts and quantify TPC and TFC, considering all preparation methods. PCA showed a total variance of principal components of 90.52%, of which 65.46% was explained by PC1 and 25.06% by PC2 (Figure 3). Principal component analysis (PCA) allowed for the joint evaluation of all variables. Multivariate analysis was applied to assess the antioxidant capacity of the extracts and quantify TPC and TFC, considering all preparation methods. PCA showed a total variance of principal components of 90.52%, of which 65.46% was explained by PC1 and 25.06% by PC2 (Figure 3).  The E + P and T − P40 methods were relatively displaced compared with the other methods, indicating more and less significant responses, respectively.

Discussion
Fruits are the main dietary source of polyphenols, which, owing to intrinsic and ex- The E + P and T − P40 methods were relatively displaced compared with the other methods, indicating more and less significant responses, respectively.

Discussion
Fruits are the main dietary source of polyphenols, which, owing to intrinsic and extrinsic factors, exhibit varied compositions of these constituents in terms of quantity and quality [13]. High temperatures increase the diffusion rate and solubility of compounds. However, depending on the conditions employed, degradation or partial removal of the active compounds may occur. Additionally, different levels of complexity in the structure of phenolic compounds lead to variations in their sensitivity to extraction conditions. Thus, the yield and composition of the extracts, and consequently their properties, depend on the extraction conditions [14].
The highest yield was obtained using the E − P extraction method. This can be attributed to the non-removal of gums, mucilages, and proteins that occur in other methods in the presence of precipitates through ethanol addition. Consequently, the extraction process was shortened without the need for a second filtration step. However, higher yields in the extraction process do not necessarily indicate higher efficiency [15]. Furthermore, difficulties in handling lyophilized extracts are encountered in these methods because, despite generally having a higher mass without precipitation, the extracts exhibit viscous and adhesive properties. In a previous study [16], the plant-to-solvent ratio was 1:2, whereas, in our study, this ratio was adjusted to 1:10. The obtained results suggest that this modification in the plant-to-solvent ratio also affects the extraction yield, supporting the assertions made by the same authors.
In general, E + P proved to be the most effective compared to others because the extraction temperature employed increased the diffusion rate and enhanced the solubilization of phenolic compounds and flavonoids, concurrently with the precipitation of proteins and polysaccharides by the addition of ethanol. This explains the lower yield of this extraction method compared to that of E − P, although it exhibited superior antioxidant capacity.
The assessment of antioxidant capacity is frequently performed using methods such as FRAP in conjunction with other techniques. Antioxidants derived from natural sources such as fruit peels play a vital role in free radical scavenging and inhibiting iron and copper chain reactions [17]. In this study, the results were directly proportional to the antioxidant capacity exhibited by each extraction process employed to prepare extracts from mature P. cauliflora fruit peels. Furthermore, a remarkable correlation was observed between total phenolic compounds and flavonoids.
Different phenolic compounds scavenge different types of free radicals. Flavonoids, tannins, and condensed tannins contribute to the antioxidant capacity of ABTS, whereas anthocyanins contribute to that of DPPH [18]. Baldin et al. [19] found a lower antioxidant effect in FRAP and DPPH assays for microencapsulated aqueous jabuticaba extracts than in the present study. Different DPPH values occur because of the extraction method, solvents used, and drying process, which can concentrate the components of the extract, and consequently explain the different levels of phenolic compounds observed for each extraction method used in each study [19]. Lenquiste et al. [20] analyzed the aqueous and methanolic extracts of lyophilized jabuticaba peels and found lower antioxidant capacities for ABTS, DPPH, and FRAP than that of E + P.
The use of different chromatographic techniques provides a comprehensive metabolic profile that aids in the identification and quantification of the major chemical markers of plant species. Generally, GC-MS is better for identifying primary metabolites, such as amino acids, fatty acids, carbohydrates, and organic acids [21]. However, UHPLC-MS/MS is more suitable for identifying secondary metabolites, such as alkaloids, saponins, phenolic acids, terpenoids, flavonoids, and glycosides [22].
Through GC-MS and UHPLC-MS/MS analyses, it was possible to obtain a comprehensive view of the chemical composition of the extracts, highlighting the presence of bioactive compounds such as flavonoids, sugars, tannins, and organic acids, which were responsible for the observed antimicrobial and antioxidant activities in this study. However, it is important to acknowledge that these compounds' measures were performed semi-quantitatively, which may have limitations in providing absolute concentrations.
Some studies have suggested that D-psicose, a major compound found by GC-MS, can suppress hyperglycemia by exhibiting hypolipidemic and antioxidant activities [23]. Furthermore, it has been indicated as a potential protective agent against type 2 diabetes [24] and its complications, such as cardiovascular diseases and hepatic steatosis, making it an ideal substitute for sucrose [23]. D-psicose is a monosaccharide that can be enzymatically produced from D-glucose via D-fructose catalyzed by D-xylose isomerase and D-tagatose 3-epimerase. However, it is a rare sugar found in nature [25].
Evidenced in large quantities in this study, citric acid plays a vital role in determining the degree of ripeness of the fruit, influencing its flavor and the balance between alkalinity and acidity. However, the activities of these acids as bioactive compounds are variable and not fully understood [26].
According to a study by Gomez-Delgado et al. [27], citric acid has significant benefits in regenerative processes due to its ability to promote the increased release of transforming growth factor beta (TGF-β1). Citric acid is widely used as a flavoring agent and preservative in the food and beverage industry. However, the mechanism of the antimicrobial action of citric acid is not wholly understood [28]. Furthermore, In, Kim, Kim, and Oh [29] have reported that citric acid exhibits weak antibacterial activity against foodborne pathogens.
Quercetin, a flavonoid found in fruits and vegetables, and evidenced as one of the major compounds present in jabuticaba, has various beneficial effects on human health [30]. Scientific studies have investigated the effects of quercetin supplementation on antioxidant and anti-inflammatory activity [31,32]. Furthermore, its potential for cancer prevention [33] and its inhibitory effect against different strains of methicillin-resistant Staphylococcus aureus (MRSA) were also investigated [34].
Regarding antimicrobial activity, the microdilution technique has been used in a limited number of studies with P. cauliflora extracts, as is the case of the present study. Oliveira et al. [14] evaluated jabuticaba peel extracts using four solvents: acetone, water, ethanol, and methanol. None of the tested extracts showed efficacy in inhibiting the growth of the Gram-negative bacteria E. coli and Salmonella choleraesuis. However, the extracts exhibited minimum inhibitory concentrations (MICs) of 250 µg/mL against P. aeruginosa [14].
From another perspective, Fleck et al. [35] demonstrated the antimicrobial activity of the aqueous extract of jabuticaba peel against S. aureus, B. cereus, and E. coli with minimum inhibitory concentrations of 11.22 ± 0, 8.42 ± 2.52, and 2.80 ± 0.11 mg mL −1 , respectively. According to Fleck [34], this activity may be related to the anthocyanins and phenolic acids in fruit peels.
The E + P method exhibited higher antimicrobial activity consistent with the results described in Table 6, although some metabolites were not significantly detected. This observation can be attributed to the precipitation, concentration, and synergism of the bioactive compounds. One notable example is the ability of phenolic compounds to affect the functioning of bacterial cells in various ways. In addition to interfering with enzymatic activity, they can also influence bacterial metabolic processes by forming complexes with metal ions [36].
Anthocyanins are another group of bioactive compounds that stand out for their antimicrobial action mechanisms. Studies conducted by Cisowska and Hendrich [37] demonstrated that the inhibitory effect of anthocyanins present in fruit peels arises from multiple mechanisms and synergies with other compounds, such as weak organic acids, phenolic acids, and their different chemical forms, resulting in various membrane and intercellular interactions that contribute to the antimicrobial action observed in this study.
Therefore, it is important to highlight the use of jabuticaba peel as a source of bioactive compounds that can contribute to the reduction in synthetic chemical use and benefit consumer health. Supplementation with phytochemical complexes based on their proportion and synergism results in more affordable and accessible products for healthy individuals who do not require a medical prescription. Additionally, the use of peel reduces resource wastage and contributes to environmental sustainability by preventing the accumulation of organic residues [14,38,39].

Plant Material
Ripe fruits of P. cauliflora were collected from a rural property in the city of Esperança Nova (Paraná, Brazil, 23.719864, −53.802104). The specimen was deposited in the herbarium of Universidade Paranaense under number 339 (SisGen number: A672209). After collection, the fruits were washed with running water, and their peels were manually removed and dried by forced air circulation for 5 d at 45 • C. Subsequently, they were pulverized in a knife mill and stored in plastic bags at 2-8 • C until further use.

Extraction Processes
A plant material-to-solvent ratio of 1:10 (w/v) was used to produce the extracts. The effects of temperature, protein precipitation, and carbohydrate precipitation in 95% ethanol and vortex extraction were evaluated, resulting in eight extracts. The study design is illustrated in Figure 4.  Powdered dried peels (25 g) were initially submitted to contact with filtered water (250 mL) at different temperatures (25,40, and 80 °C) for 1.5 h, followed by vortexing for 5 min at a rate of > 2000 rpm according to Islam et al. [16] with modifications. After filtering the solid residues, the first aliquot was frozen and lyophilized. The second aliquot was treated at an extract-to-ethanol ratio of 1:3 (v/v). The ethanol (95% ethyl alcohol) addition reduced the polarity of water, causing the in-solubilization of polysaccharides and proteins. Subsequently, the precipitate was removed by filtration, the solvent was removed using a rotary evaporator (Nova Ética, São Paulo, Brazil), and the resulting fraction was lyophilized (JJ Científica, model LJJ02, São Paulo, Brazil).
The plant material was placed in water at 90 °C for 6 h for preparing infusion. Subsequently, the sample was filtered with or without precipitation, as previously described.
The extraction yield was calculated as the ratio of the extracted mass to the initial mass of the dried peels. Powdered dried peels (25 g) were initially submitted to contact with filtered water (250 mL) at different temperatures (25,40, and 80 • C) for 1.5 h, followed by vortexing for 5 min at a rate of >2000 rpm according to Islam et al. [16] with modifications. After filtering the solid residues, the first aliquot was frozen and lyophilized. The second aliquot was treated at an extract-to-ethanol ratio of 1:3 (v/v). The ethanol (95% ethyl alcohol) addition reduced the polarity of water, causing the in-solubilization of polysaccharides and proteins. Subsequently, the precipitate was removed by filtration, the solvent was removed using a rotary evaporator (Nova Ética, São Paulo, Brazil), and the resulting fraction was lyophilized (JJ Científica, model LJJ02, São Paulo, Brazil).

Phytochemical Evaluation
The plant material was placed in water at 90 • C for 6 h for preparing infusion. Subsequently, the sample was filtered with or without precipitation, as previously described. The extraction yield was calculated as the ratio of the extracted mass to the initial mass of the dried peels.

GC-MS Analysis
For chemical identification by GC-MS before analysis, samples (~40 mg) were derivatized using bis(trimethylsilyl) trifluoroacetamide with trimethylchlorosilane (BSTFA/TMCS) (Sigma-Aldrich, Saint Louis, MO, USA) and pyridine (Anidrol, Diadema, Brazil) (200:200 µL) at 90 • C for 1 h in an oven with air circulation, as indicated by Canini et al. [40], and then diluted in ethyl acetate (Anidrol, Diadema, Brazil) to a final volume of 1 mL. The solution was analyzed by gas chromatography using a mass spectrometer (Shimadzu, GCMS-QP2010 SE, Tokyo, Japan) equipped with an automatic injector (AOC-20i) and capillary column SH-Rtx-5MS (Shimadzu, 30 m × 0.25 mm × 0.25 µm, Tokyo, Japan). Helium (White Martins, purity > 99%) was used as the carrier gas at a flow rate of 1.0 mL min −1 with a split ratio of 1:30, and the injection volume was 3 µL. The column temperature was initially programmed to be 100 • C, heated at 4 • C min −1 to 280 • C, and then heated at 10 • C min −1 to reach a final temperature of 300 • C. The temperature of the injector and MS interface was maintained at 250 • C. Mass spectra were recorded at 70 eV with a mass range of m/z 35-550. The compounds were identified by searching the library database of Spectra NIST Mass Spectral Library (version 2014). The relative abundance of each metabolite was calculated by multiplying the individual area of the compound by 100 and then dividing it by the total area of all identified compounds in the sample.
The UHPLC-MS/MS system was equipped with a UPLC CSH C18 column (Waters, USA, 1.7 µm, 2.1 × 100 mm). The mobile phase was a mixture of solvent A (water with 0.1% formic acid, v/v) and solvent B (acetonitrile with 0.1% formic acid, v/v) at a flow rate of 0.250 mL min −1 . The A:B gradient used was as follows: 3% B from 0 to 1 min, 50% B from 1 to 10 min, 95% B from 10 to 15 min, 95% B from 15 to 19 min, 3% B from 19 to 21 min, and maintained at 3% B from 21 to 25 min at 40 • C, with the last four minutes dedicated to column reconstitution for the subsequent analysis. For negative analysis, solvent mixtures A (water with 0.1% formic acid, v/v) and B (acetonitrile) were used. The gradient mixture for negative analysis was the same as that for positive analysis.
A Q-TOF Impact II mass spectrometer with an electrospray ionization source was used in the auto MS/MS acquisition mode. The acquisition rate was 5 Hz (MS and MS/MS), and the tuning range was m/z 120-1200. Mass spectra were collected using electrospray ionization (ESI) in positive and negative ion modes, with a capillary voltage set at 3.50 kV, source temperature of 200 • C, and a desolvation gas flow rate of 9 L min −1 .
The ion chromatogram and MS and MS/MS spectra were visualized using Data Analysis 4.3 software and compared with the existing literature. This method was developed based on that described by Tolouei et al. [41].

Spectrophotometric Analysis
The extracts were prepared at a concentration of 1000 µg mL −1 and evaluated in independent triplicates. Total phenolic content (TPC) was quantified using the Folin-Ciocalteu method [42] on a UV/Vis spectrophotometer (Kasuaki Model IL-592) at 765 nm. The results were calculated based on the calibration curve for gallic acid (Sigma-Aldrich) The total flavonoid content (TFC) was determined based on the method described by Woisky and Salatino (1998) [43], and absorbance was measured at 425 nm using a spectrophotometer (Kasuaki Model IL-592). A standard calibration curve of quercetin (y = 81.561x − 126.41; R 2 = 0.9966) was plotted for quantification, and the results were expressed as µg equivalents of quercetin per gram of extract (µg QUE gext −1 ).

Antioxidant Analysis
The extracts were prepared at a concentration of 1000 µg mL −1 and evaluated for their scavenging capacity against DPPH and ABTS radicals and their reducing capacity using the FRAP assay, in independent triplicates.
The FRAP analysis was performed according to the methodology described by Santos et al. [46]. To determine the antioxidant capacity, a calibration curve for ferrous sulfate (Sigma-Aldrich) was plotted (y = 0.6188x − 96.833; R 2 = 0.9926), and the results were presented as µmol of Fe 2+ per gram of extract (µmol Fe 2+ gext −1 ).
The extracts were added to microplates containing Brain Heart Infusion (BHI) broth at concentrations ranging from 125 mg mL −1 to 1.95 mg mL −1 . Microbial suspensions were prepared in sterile water at a McFarland scale of 0.5. The microplates were incubated at 36 • C for 24 h for bacteria and 27 • C for 48 h for yeast. Subsequently, a 2% solution of 2,3,5-triphenyltetrazolium chloride (TTC) was added and incubated for 2 h. A reddish color indicates microbial growth.

Statistical Analysis
The results of different extraction methods were subjected to Levene's test for data homogeneity, followed by an analysis of variance (ANOVA). Means were compared by Duncan's test (p ≤ 0.05) using SPSS software (version 22.0; SPSS Inc., Chicago, IL, USA).
Cluster analysis (CA) and principal component analysis (PCA) were performed to discriminate the composition of the extracts based on the different preparation methodologies, and the variables were analyzed using Statistica v. 13.3.

Conclusions
The compounds identified in extracts from Plinia cauliflora fruit peels, including flavonoids, tannins, and organic acids, reinforce the potential bioactive effects of jabuticaba, including its antioxidant and antimicrobial capacities. Among the extraction methods, extraction with precipitation (E + P) demonstrated higher efficiency when compared to the other extraction methods. It is important to note that this study focused on utilizing phytocomplexes to leverage the synergistic effects among bioactive compounds rather than investigating isolated compounds. This approach is relevant as it may reflect the real-life scenarios of bioactive interactions and potential health benefits when using the whole extract. Therefore, the E + P method emerges as a promising extraction technique for large-scale production in the formulation of food supplements.