Chemical characterization of odour-active volatile compounds during lucuma (Pouteria lucuma) fruit ripening

ABSTRACT Lucuma (Pouteria lucuma) is an important exportation fruit in Peru. In this work, the odour-active volatile changes during fruit ripening were studied by using a sensomics approach (GC-O in combination with GC-MS analyses). The volatile compound extracts from unripe and ripe fruits were obtained by solvent-assisted flavour extraction (SAFE) with dichloromethane. By application of the aroma extract dilution analysis (AEDA) on the ripe fruit extract, 16 odour-active compounds were detected as responsible of sweet, green, and rancid odour-notes, characteristics of this fruit. Based on odour activity value (OAV) calculation, 2,3-butanedione, methional, (Z)-3-hexenal, (E)-2-hexenal, (Z)-β-ocimene, and 3-methyl butanoic acid were identified as key-aroma compounds in this fruit. These compounds were also quantitated in the unripe fruit SAFE extract finding that their amount increased with during the ripening, such contributing to the development of characteristic odour notes.


Introduction
Lucuma (Pouteria lucuma) is a tropical fruit belonging to Sapotaceae family ( Figure 1) and is widespread in the western South America. Peru is one of the main producers and exporters worldwide (SIICEX, 2019), because its different microclimates allow to have fruit throughout all year. This fruit is usually cultivated from 1500 to 3000 m above sea level, with a temperature range of 8-27°C and a relative humidity value of 80-90%. It was used by Inca culture, as one of the main ingredients of their diet, since much evidence was found in images moulded in "huacos" (ceramic pieces) of the Mochica culture (Ministerio de Agricultura del Perú, 2019a). The production of lucuma has remained almost constant during the last 3 years, reaching 14,040 tons in 2017 (Ministerio de Agricultura del Perú, 2019b).
The fruit pulp exhibits an intense yellow colour and characteristic aroma that allow us to use it for the preparation of different foods. Among them, ice creams and bakery products are the most consumed. Pouteria lucuma fruit has a low-moisture content (56.0-72.3% during consumption maturity stage) compared to other fruits; in contrast, protein content is high, varying from 1.5 to 3.3% (Erazo, Escobar, Olaeta, & Undurraga, 1999). This fruit is also rich in carotenoids; minerals, such as, calcium, phosphorus, and iron (Ministerio de Salud del Perú, 2009); as well as vitamins such as, thiamine, riboflavin, niacin and large amount of vitamin C (0.14-1.07 mg/100 g dry matter) (Fuentealba et al., 2016;Yahia & Gutierrez-Orozco, 2011). Lucuma pulp is the main presentation to export this fruit, which stimulates the development of processing technologies to preserve its properties.
Triterpenes, phenolic compounds and carotenoids are the main secondary metabolites in fruit extracts of Pouteria species, a tropical and subtropical genus (da Silva, Gordon, Jungfer, Marx, & Maia, 2012). Besides, in vitro antioxidant and antihyperglycemic (specifically α-amylase and α-glucosidase inhibitory activities) properties have been studied and were found dependent on ripening stage (da Silva Pinto et al., 2009;Fuentealba et al., 2016). Despite attractive sensorial characteristics of P. lucuma fruit, there is no any study in literature related to the chemical composition of aromaactive volatile compounds. Only the VOC profile of lucuma fruits exported to Italy was recently analysed by using PTR-MS-TOF measurements, tentatively identified more than 50 compounds (Taiti, Colzi, Azzarello, & Mancuso, 2017).
Usually, this fruit is harvested in unripe stage and after ca. 2 weeks the ripe stage is reached, which is characterized by a strong and pleasant aroma. Thus, the aim of this work was to identify the odour-active volatile compounds in this fruit, evaluate their contribution to the whole aroma and their change from unripe to ripe stages. The so-called molecular sensory approach (Schieberle & Hofmann, 2011) was used, what combines instrumental (GC-MS) and sensory (GC-O) analyses to focus the identification efforts only in those compounds that are aroma-active.

Fruits
Pouteria lucuma (biotype Dos Marron) fruits were harvested from a farm located in Fundo Huayquina, Chincha, Peru, July 2017. Fruits were randomized collected from 10 trees (yellow peel colour under the calix, unripe stage), and stored until they reached fully ripe stage (soft texture and brown-greenish peel colour). Physicochemical parameters, such as pH, total soluble solid content (°Brix), and acidity, were measured in both stages. For this purpose, each sample lot consisting of 4 units (ca. 1 Kg) of fruits was homogenized using a commercial stainless-steel blender. The pH of the fruits was determined by using a C6820 pH-meter (Schott Gerate, Berlin, Germany). The total soluble solids content (SS) was measured with an Atago PAL-alfa refractometer at 20°C. Titratable acidity (TA) was determined from 10 g of puree diluted with 50 mL of water, titrated with 0.1 N NaOH and calculated as percent of citric acid (AOAC, 2006). All measurements were done in triplicate.

Isolation of volatile extract by SAFE distillation
Ripe and unripe lucuma fruit pulps (300 g) were separately homogenized with anhydrous sodium sulphate (Na 2 SO 4 ) and these mixtures were packed in a glass column, to elute volatile compounds with 300 mL of dichloromethane in each case. The organic phases so obtained, containing the aroma compounds, were subjected to SAFE (Solvent-Assisted Flavour Extraction) (Engel, Bahr, & Schieberle, 1999) and a colourless extract without non-volatile compounds exhibiting the characteristic fruit aroma was obtained. The organic extract was concentrated at 45°C on a Vigreux column (50 cm × 1 cm) to a 0.2 mL volume to be analysed by GC-O and GC-MS.

GC-O and GC-MS analyses
The GC-O analyses were performed in a Hewlett Packard 5890 series II gas chromatograph (Hewlett-Packard, Wilmington, DE, USA), equipped with a TR-FFAP column (Thermo, 30 m x 0.32 mm i.d., 0.25 μm film thickness). The injection (1 μL) was done in split mode (1:10). Helium was used as a carrier gas at flow rate of 1.2 mL/min. The oven temperature programme started at 40°C for 1 min, then the temperature was increased at 6°C/min until 180°C, then at 12°C/min until 230°C and finally held for 10 min. The final end of the column was connected to a deactivated Y-shaped glass splitter (Chromatographie Handel Mueller, Fridolfing, Germany), to divide the effluent into two equal parts (1:1 ratio), one for to the conventional FID detector (230°C) and the other for the sniffing port (230°C) by using deactivated fused silica capillaries of the same length (50 cm × 0.32 mm). Sniffing port consisted of an elbow-shaped aluminium tube (80 mm × 5 mm i.d) self-made. Three trained panellists detected (every 20 min) the olfactory-active regions and described the odour notes perceived from each odour-active compound. The panellists were trained in the recognition of main descriptors of this fruit (green, sweet, and rancid-fermented), by the orthonasally evaluation of the corresponding reference odorants ((Z)-3-hexenal (green), butanoic acid (rancid), and 4-hydroxy-2,5-dimethyl-3(2H)-furanone (sweet), among others).
GC-MS analyses were done in a 5977A mass selective detector (Agilent Technologies Inc., Wilmington, DE, USA) coupled to the Agilent 7890B gas chromatograph, in electron impact ionization mode at 70 eV. Mass spectra were recorded in the range of 30 to 350 u. The SAFE extracts were analysed in a HP-FFAP (Agilent, 50 m x 0.25 mm i.d., 0.32 μm film thickness) column, using the same temperature program described for GC-O analyses. The volatile compound extracts were also analysed on a HP-5MS (Agilent, 30 m x 0.25 mm i.d., 0.25 μm film thickness) column under the following conditions: oven temperature was increased in three steps, first rate at 4°C/min from 50°C until 160°C, second rate of 2.5°C/min until 220°C, and third at 8°C/min until 300°C, where finally was held for 4 min; injector temperature was 250°C; and Helium was used as carrier gas at 1 mL/min.

Aroma extract dilution analysis (AEDA)
The AEDA method was applied to the SAFE extract of ripe lucuma fruit pulp, in order to establish the contribution of  each odour-active compound to the whole P. lucuma fruit aroma. Thus, this SAFE extract was diluted step by step to obtain 2 n dilutions (Schieberle, 1995), and each dilution was analysed by GC-O as it was above-mentioned. The odour activity of each compound was expressed as flavour dilution factor (FD), which was determined as the greatest dilution at which that compound was still detected by comparing all of the runs (Grosch, 1994).

Identification of odour-active volatiles
Odour-active volatile compounds were identified by comparing their mass spectra, retention indexes (RI) calculated by using a C 8 -C 25 alkane-mix (in both columns), and odour notes with those exhibited by the corresponding standards. The correlation of GC-O and GC-MS analyses was done by comparison of RI obtained in each run. The database NIST/ EPA/NIH Mass Spectral Library 2014 (2.2) was used for the cases that the standards were not available.

Quantitation of key-aroma compounds
Internal standard (IS) method was used to quantitate the odour-active volatiles in both ripening stages. Ethyl cinnamate (98%, 50 µg, Alfa Aesar, Heysham, UK; 250 μg/kg fruit) was dissolved in dichloromethane (100 mL), and added to the homogenized fruits (300 g), separately for each stage, before extraction by SAFE distillation. To determine the response factor for each volatile compound, calibration lines were constructed using a series of solutions of varying nominal concentrations containing each analyte (IS: analyte, 1:5 until 5:1), where the slope was assumed as the response factor. IS was added to each solution in the same concentration to obtain the corresponding chromatograms (IOFI, 2011). Three replicates were made for each analyte. The comparison of GC-FID signals with those of standards was used to calculate the concentration of each analyte, according to the following equation: where C x is the analyte concentration in μg/kg fruit, A x is the analyte area, A IS is the IS area, and RF is the response factor.

Statistical analyses
Data from the physicochemical characterization and quantitation of odour-active volatiles are reported as the mean ± standard deviation for determinations performed in triplicate. By the Statgraphics Centurion 18 -X64 software (StatPoint Technologies Inc. USA), T-tests for differences of means at probability level p ≤ 0.05 were considered significant.

Results and discussion
Physicochemical characterization of unripe and ripe lucuma fruits is shown in Table 1. As it was expectable, pH value and total soluble solid content increased with ripening, with significant differences among two stages. Pouteria lucuma fruits are climacteric that allows them to mature after harvesting, meaning that they have a living breathing system. One of the major chemical changes that is triggered during the fruit ripening is the flavour development, thus considering the volatile constituents as ripening products (Fischer & Scott, 1997). The trained panellists described the orthonasal aroma of ripe P. lucuma fruit pulp as enriched in green, sweet, and rancid-fermented odour notes. The GC-O analyses of SAFE extract of this fruit showed the presence of 16 odour-active volatile compounds (Table 2) as contributors to the whole aroma. Different kind of chemical compounds were identified: C 6 -compounds were responsible for the green, ketones for the sweet, esters for the fruity, and acids for the rancid odour notes.
The odour-active volatile compounds were detected and quantified in the SAFE extract of unripe P. lucuma fruits (Table 2), finding an increase in the amount of these compounds during ripening, with exception of 4-hydroxy-4methyl-2-pentanone and 3-methyl butanoic acid. Some compounds, such as, methional, ethyl 3-hydroxy-butanoate, and butanoic acid, were only detected in ripe fruits. Figure 3 shows a comparison of odour-active volatile amount between the two ripening stages. The volatile compounds were grouped according to their chemical functions in C 6compounds, ketones, terpenes, sulphur compounds, All data are the mean of three measurements ± standard deviation. Different letters in a file means that there are significant differences based on T-test (p ≤ 0.05). Todos los datos son el promedio de tres medidas ± desviación estándar.
Letras diferentes en una fila significa que existen diferencias significativas basadas en el T-test (p ≤ 0.05).  All data are the mean of three measurements ± standard deviation. In all of the cases, correlation coefficient was higher than 0.975. Different letters (g, h) in a file means that there are significant differences based on Tukey test (p ≤ 0.05).
o Kishimoto, Wanikawa, Kono, and Shibata (2006 aliphatic acids, and esters. With exception of ketones, the other volatile compounds were found in significant major amount in the ripe stage of lucuma fruit, indicating that their biogenesis is activated during this process. Ketones are responsible for sweet, C 6 -compounds for green, and aliphatic acids for rancid characteristic odour notes in this fruit. The unsaturated lineal and branched aliphatic acids are ripening products which are developed through two biogenetic ways: β-oxidation pathway from fatty acids, and amino-acid metabolism, respectively (Fischer & Scott, 1997).

Conclusion
The odour-active volatiles of P. lucuma fruit were identified, suggesting that green, sweet, and rancid odour-notes are predominant in this fruit. The significance of ketones, sulphur, and C 6 -aliphatic compounds on the aroma of lucuma were reported, confirming that methional, (Z)-3-hexenal, (E)-2-hexenal, and (Z)β-ocimene are key odorant compounds in fruits of Sapotaceae family. The ripening process of this fruit triggered the production of volatile compounds, with emphasis on the C 6 -compound and aliphatic acid amount. These results are important to improve the processing of this fruit in order to preserve its authentic aroma.

Disclosure statement
No potential conflict of interest was reported by the authors.