All the chemicals and solvents used in the experiments were of HPLC and analytical grades.

The fruits of Salvadora persica were obtained from experimental salt farm area of CSIR-CSMCRI, Bhavnagar, Gujarat, India (latitude 21° 47.3060 N and longitude 72° 7.4170 E). The fruits were collected from at least five different randomly selected plants of S. persica. Fruits that were mature, firm and free from any damage were sorted and stored at –80°C until extraction and estimation of various parameters.

Standard metabolite data reporting protocol was used for the extraction of the samples for the metabolic profiling (Fernie et al., 2011). The samples were extracted following the method described by Lisec et al. (2006). Hundred micrgram of deseeded fruit pulp was weighed and homogenized into fine powder using a chilled mortar and pestle using liquid nitrogen. The metabolites extraction was carried out by adding 1.4 mL of pre-chilled methanol and with subsequent addition of internal standard (100 μL of ribitol, 1 mg/mL). The mixture was incubated at 70°C for 10 min. After incubation the mixture was cooled and centrifuged at room temperature (11000 × g) for 10 min. Then HPLC grade chloroform (750 μL) and water (1.4 mL) was added to the supernatant. The resulting solution was mixed properly and centrifuged for 15 minutes at 22,000 × g at room temperature. Hundred and fifty microliter of aqueous layer was taken and vacuum dried.

For derivatization, the residues collected after vacuum drying, were re-dissolved in 40 μL of methoxylamine hydrochloride in pyridine (20 mg/mL) and derivatized at 37°C for 2 h. Then it was mixed with 70 μL of O-Bis(trimethylsilyl)trifluoroacetamide(BSTFA) and incubated at 37°C for 30 min. The GC column was then injected with 2 μL of sample. The GC/MS system (GCMS-QP2010, Shimadzu, Kyoto, Japan) was equipped with an auto-sampler. The ion source was tuned to 250°C and the transfer line was set as 300°C at a rate of 14.5°C s^-1. The scanning range for the mass spectra was 70–700 mass-to-charge ratios and it was recorded at a rate of eight scans per second. The compounds were identified based on mass fragmentation spectra and relative retention time of those standards and NIST 2014 and WILLEY 2014 libraries. The quantification of the metabolites was carried out using ribitol as an internal standard.

The amino acid profiling of S. perica fruit was carried out according to the method described by Kwanyuen and Burton (2010). Fruit pulp (10 mg) dried and powdered was hydrolyzed in a hot-air oven with 500 μl HCl (6 N) for 24 h in a glass vessel at 110°C and the hydrolyzed sample was dried in a vacuum descicator. After hydrolysis, the samples were neutralized by adding 500 μl of a reaction mixture containing ethanol-water–TEA (v/v, 2:2:1) along with an amino acid standard (10 μl, AAS18, Sigma, USA) to the vessels containing hydrolyzed samples. The mixtures were vortexed and vacuum-dried. For derivatization, added 500 μl of derivatization mixture containing TEA–water–PITC-ethanol in the ratio1:1:1:7(v/v) and vortexed thoroughly. After an incubation of 20 min at room temperature, the reaction mixture was vacuum-dried and re-dissolved in 400 μl of 5 mM Na₂HPO₄ buffer (pH 7.4) containing 5% (v/v) acetonitrile. HPLC system (Waters Alliance model, 2695-seperation module, 2996-photodiode array detector with an auto-sampler, USA) was used for the analysis of the amino acid composition after filtering the sample with 0.22 μm syringe filter. The separation of the amino acid was carried out by a gradient resulted from mixing of eluent A and eluent B with a flow rate of 1 ml min^-1 throughout. The eluent A contained 0.05% TEA, 150 mM CH₃COONa.3H₂O and acetonitrile (6%) having pH 6.4, however, eluent B has a composition of water and acetonitrile in the ratio 4:6 (v/v). The amino acids eluted were recorded at 254 nm. The amino acid composition was calculated from the relative proportion of the peak area and expressed on dry weight basis.

The results of all the experiments were stated as mean ± standard deviation (S.D., n = 5) and 5 different plants were selected for experimental sampling.

The proximate composition of edible part of S. persica fruit are presented in Table 1. All the nutrient and proximate compositions were assessed on dry weight basis. The edible part was 73.83 ± 5.08% of the whole ripen fruit (Figure 1). The ripe fruit was sufficiently dehydrated containing an average moisture content of 70.00 ± 0.97 g 100 g^-1 FW (Table 1). The proximate analysis showed that the fruit contained comparable moisture content to that reported for Salvadora oleoides (Duhan et al., 1992) but slightly higher as compared to that reported for Durio zibethinus (58–69 g 100 g^-1 FW) (Charoenkiatkul et al., 2016). The protein content of the fruit (5.92 ± 0.66 g 100 g^-1 DW) was at par with that reported for Syzygium cumini (Raza et al., 2015) but was found be comparable to that of Durio zibethinus (Charoenkiatkul et al., 2016) and was higher than Cola parchycarpa (Ene-Obong et al., 2016). The carbohydrate content of the S. persica fruit (73.66 ± 2.17 g 100 g^-1 DW) was in comparison to the high carbohydrate containing durian variety Durio zibethinus and was found to higher than that reported for leguminous plants Bracystegia eurycoma and Pipper guineense (Bolanle et al., 2014) and Cola parchycarpa (Ene-Obong et al., 2016). The carbohydrate content was also in comparison with some conventional carbohydrate sources like cereals (72-90 g 100 g^-1 DW) (Adewusi et al., 1995), and thus S. persica fruit can serve as a potential source of carbohydrate. The lipid and dietary fiber content of the fruit was also higher than the Durio zibethinus, Bracystegia eurycoma, Pipper guineense, and Cola parchycarpa signifying that the fruit is rich in essential nutrients and can be utilized for supplementing the human diet for better health. The relatively high energy content of the fruit 141.35 kcal 100 g^-1 DW could fulfill the daily calorie intake of the people. The results presented here demonstrate that the fruit of S. persica could be an important source of essential nutrients and energy that could be utilized for supplementing and enhancing the diet and health of humans.

The ascorbic acid (vitamin C) has the high antioxidant activity and it helps in maintaining the cellular membrane integrity. Besides this, vitamin C has also been reported to prevent formation of cancer causing N-nitroso compounds from nitrites and nitrates present in water and meat (Kaur and Kapoor, 2001). A considerably high amount of ascorbic acid was found in S. persica fruit (67.99 ± 15.86 mg 100 g^-1 DW). The vitamin C content was at par with that reported for Ziziphus mauritiana (76 mg 100 g^-1 DW) and was found higher than some important non-conventional fruits like Feronia limonia, Capparis decidua, and Balanites aegyptiaca (Duhan et al., 1992). As compared to other species of Salvadora like Salvadora oleoides found in the arid and semi-arid region of India, the vitamin C content was much higher indicating that a high amount of vitamin C in this plant food makes it, an important source of vitamin C in the diet of people in arid and semi-arid areas.

Carotenoids are a family of fat soluble plant pigment considered as a good source of antioxidant. Besides its esthetic role of providing color to the fruit they are also considered beneficial in reducing risk of diseases like cancer, eye and heart diseases (Johnson, 2002). The carotenoid content (11.371 ± 2.48 μg 100 g^-1 DW) in fruits of S. persica was at par with the carotenoid content reported for some underutilized fruits of Garcinia prainiana, Durio kutejensis, Baccaurea reticulate, and Baccaurea macrocarpa (Khoo et al., 2008). In numerous studies, however, it was proven that supplementation of carotenoids through fruits and vegetables is more significant in reducing disease risks and bringing about the health benefits of carotenoids (Fraser and Bramley, 2004; Voutilainen et al., 2006). Therefore, S. persica fruit seems a promising choice for increasing the dietary intake of carotenoids.

Plants are the rich source of phenolic substances of different origins and functions. Most of them are biologically active compounds of plant origin having anti-cancerous, antiviral, antibacterial properties (Manach et al., 2004). However, very less reports on the phenolic contents of non-traditional fruit plants and their antioxidant activities are available (Stratil et al., 2006). The total amount of phenolics in the fruit extract was estimated to be 539.61 ± 120.38 mg 100 g^-1 DW (Table 1), which was in comparison with wild fruit of Saba senegalensis, Sclerocarya birrea, and Diospyros mespiliformis but was higher than Diospyros mespiliformis, Ficus sycomorus, Lannea microcarpa, and Parkia biglobosa, however, the value was lower than that reported wild medicinal plants, Ziziphus mauritiana and Tamarindus indica (Lamien-Meda et al., 2008). Similarly, the total phenolic content was at par with that reported for blueberries (670.9 mg 100 g^-1 DW) and sour cherry (429.5 mg 100 g^-1 DW) and was higher than green and red pepper (Marinova et al., 2005).

Flavonoids are class of polyphenolic compounds that constitute the major antioxidants in fruits and possess beneficial effect on human health. Although, flavonoids are generally considered as non-nutrients, they are important component for human diet due to their high antioxidant activities (Calado et al., 2015). Different foods, fruits, beverages, herbal drugs are considered as good sources of flavonoids. Earlier studies have stated antimicrobial, anticancer, anti-inflammatory and anti-allergic activities due to effective scavenging of reactive oxygen species comprising of singlet oxygen and other free radicals (Montoro et al., 2005). The flavonoid content of the S persica fruit was estimated to be 317.48 ± 77.59 mg 100 g^-1 DW (Table 1) which was found to be at par with the flavonoid content of plum (366 mg 100 g^-1 DW) and was higher than that reported for blueberries (190.3 mg 100 g^-1 DW) and other commercial fruits like apple, banana, peach, grapes etc. (García-Alonso et al., 2004; Marinova et al., 2005). A positive correlation of the flavonoid content with the high reducing power and antioxidant capacity of the plant was noted, suggesting the role of flavonoids in inferring these properties. Also, this leads to the consideration that there may be possible influence of anthocyanin pigments in this activity which provides dark red color to the ripe fruit. Thus, this fruit enriched with polyphenols and flavonoids, could be a potential source of antioxidants reducing risk of disease caused especially due to oxidative stress by increasing the overall antioxidant capacity of an organism.

The mineral ion compositions of the S. persica fruit are presented in Table 2. The potassium and calcium content was found high (1362.98 ± 81.50 and 1013.72 ± 48.60 mg 100 g^-1 DW, respectively) which was higher than S. oleoides and Dialium guineense (Duhan et al., 1992; Marinova et al., 2005). Previous studies have suggested dietary intake of potassium has beneficial effect on coronary heart diseases by decreasing blood pressure and maintaining adequate Na⁺ and K⁺ ratio (Weaver, 2013). Calcium is a vital mineral nutrients for human diet and it is involved in many aspects of processes like muscle contraction, cell differentiation, neuronal activities and immune responses leading to program cell death (Pu et al., 2016). The calcium level in S. persica fruit was observed to be higher than in traditional fruits like banana, apple, orange, chiku, papaya, etc. and medicinal plant like Caralluma tuberculata (García-Alonso et al., 2004; Rathore, 2009), suggesting it to be a potential calcium source. The fruit extract was found to have significant Na⁺ (263.17 ± 28.32 mg 100 g^-1 DW) concentration. The Na⁺ content was found to be higher than wild fruit D. guineense, Chrysophyllum albidum, Irvingia gabonensis, and Cola millenii (Olayiwola et al., 2013; Ayessou et al., 2014) revealing that S. persica is a good source of Na⁺ which is an imperative mineral for human health required for cellular homeostasis, fluid balance and important physiological processes (Farquhar et al., 2015). Magnesium is a central mineral for retention of vitamin D and calcium into the bones for maintaining bone structure and helps in prevention of osteoporosis and cardiovascular diseases (Chakraborti et al., 2002). Our result showed significant concentration of Mg²⁺ (143.33–159.36 mg 100 g^-1 DW) much higher than that reported for other non-conventional fruits like S. oleides, Prsopsis cineria, Capparis decidua, Ficus glomeratai, and D. guineense but was comparable to I. gabonensis and less than that reported for Spondias mombin (Duhan et al., 1992; Olayiwola et al., 2013; Ayessou et al., 2014) suggesting the utilization of the fruit extract of S. persica in nutrient supplements as a source of minerals.

The fruit of S. persica contained micronutrients Fe²⁺ (137.90 ± 21.57 μg 100 g^-1 DW), Mn²⁺ (8.77 ± 1.79 μg 100 g^-1 DW) and Zn²⁺ (8.95 ± 1.34 μg 100 g^-1 DW) in considerable amount. Iron is vital for survival of an organism as it participate in various metabolic processes like respiration and DNA synthesis (Lieu et al., 2001). The Fe²⁺ content was found to be significantly comparable to C. albidum, I. gabonensis, C. millenii but was found lower than S. oleides and D. guineense (Duhan et al., 1992; Olayiwola et al., 2013). Manganese is an imperative mineral nutrient for human consumption but excessive intake may lead to neurodegenerative disorders (Avila et al., 2013). In accordance with this, the Mn²⁺content was found less in comparison to S. oleiodes and other wild fruits (Duhan et al., 1992) making it suitable for human intake. The trace element Zn²⁺ is required for optimal function of various physiological and biochemical processes. It also helps in healthy aging and mitotic division and DNA synthesis (Tudor et al., 2005). Zn²⁺ content was found to be comparable to that reported for S. oleoides, Capparis decidua, Ziziphus jujube, I. gabonensis (Duhan et al., 1992; Olayiwola et al., 2013) but was found lower than that reported for D. guineense, C. albidum, and S. mombin (Olayiwola et al., 2013; Ayessou et al., 2014).

Nowadays, antioxidants in food are gaining high importance and are highlighted due to their pertinent role in preserving good health by prevention of diseases through scavenging the free radicals responsible for propagation of many and diseases like AIDS, cancer and neurodegenerative disorders. The current study composes the total antioxidant capacity of S. persica fruit extract presented in Figures 2–4 assessed by ABTS, DPPH, hydrogen peroxide scavenging, superoxide scavenging, reducing power, and phosphomolybdenum complex forming assays. DPPH radical scavenging assay evaluates the ability to quench DPPH free radical (Prior et al., 2005). The IC₅₀ for DPPH of the S. persica fruit was higher than guava, banana, star fruit and water apple (Lim et al., 2007) indicating low antioxidant potential in spite of its relatively high total phenolic content (IC₅₀ = 307.06 ± 49.91 μg) (Figure 2B). The possible reason for that may be, first, it has been reported by Bondet et al., 1997 that low readings for reaction of DPPH for antioxidant activity may be due to reversible reaction of DPPH with phenols like eugenol and its derivatives. Another possible reason could be slow reaction rate between substrate molecules and DPPH (Huang et al., 2005). ABTS radical scavenging assay takes into account the reduction of ABTS ion to neutral colorless form for displaying the antioxidant capacity of the extract. The ABTS radical scavenging was found to be 365 ± 94.0 μg which was higher than that reported for guava (Thaipong et al., 2006) but was lower than chiku, strawberry, plum, star fruit (Leong and Shui, 2002) indicating higher antioxidant capacity than these commercial fruits.

Superoxide radical is regarded as a main source of ROS. It creates oxidative stress in the biological system by generating singlet oxygen and hydroxyl radicals and act as a weak oxidant. It arises due to electrolyte leakage from physiological processes and respiratory chain (Saeed et al., 2012). The hydrogen peroxide (H₂O₂) is a weak oxidizing agent which causes lipid peroxidation and damages DNA by production of hydroxyl radical (Hazra et al., 2008). The IC₅₀ for superoxide and H₂O₂ radical scavenging was found to be 34.1 ± 1.2 μg and 85.70 ± 17.04 μg, respectively (Figures 3A,B), which was slightly higher than Terminalia chebula, Terminalia belerica, Emblica officinalis, and Spondias pinnata (Hazra et al., 2008, 2010) indicating significant free radical efficiency of S. persica fruit. The change in color from yellow to green caused by reduction of Fe³⁺ to Fe²⁺ demonstrates the reducing power of the extract, when measured at 700 nm (Saeed et al., 2012). The higher reducing power of S. persica fruit extract was shown by the increase in reducing power with increasing concentration. The correlation coefficient for this linear relationship was noted to be 0.998 (Figure 4A). The phosphomolymdenum assay measures the reduction of Mo (VI) to Mo (V) spectrophotometrically with absorbance maxima at 820nm for determining the antioxidant capacity. The antioxidant capacity was high as indicated by increase in absorbance with increasing concentration having correlation coefficient (R²) of 0.9932 (Figure 4B). The present study demonstrated that methanolic extract of S. persica fruit displayed the high antioxidant potential measured in terms of phosphomolybdate reduction. It has been reported that that the phosphomolybdate scavenging activity of many medicinal plants may be attributed to the flavonoid and related polyphenols present (Esmaeili et al., 2009). The higher redox potential of S. persica can be attributed to higher ascorbic acid and high phenolic content. Thus, owing to higher redox and antioxidant properties, S. persica fruit may serve as a source of antioxidants in pharmaceutical formulations and food supplements.

The GC-MS analysis revealing the metabolite profile of the fruit extract of S. persica is represented in Figure 5. The peak identification, characteristics and quantification of metabolites are presented in Table 3. The GC-MS analysis of the fruit extract identified 22 metabolites having several pharmacological activities, antioxidant properties and nutraceutical potentials. The metabolites detected were grouped into organic acids, organic base, carboxylic acid, sugar alcohols and sugars. The identified metabolites are quantified using ribitol (peak 8, Figure 5C) as internal standard. The following headings discuss, in brief, the different nutritional and pharmacological functions of the identified metabolites from fruit extract of S. persica.

Amino acids are nutrients that are imperative for enormous biosynthesis processes and are building blocks of protein that make up major content of human body. Apart from these basic functions they are also involved in providing other beneficial effects to human body which includes controlling of blood sugar level and building of muscles (Fürst and Stehle, 2004). Amino acids are majorly used as food additives in food industry and are also preferred as pharmaceutical and cosmetic industry due to their moisturizing effect (Banipal et al., 2016). The present study identified 17 amino acids which are listed in Table 4 and representative chromatogram is shown in Figure 6. The identified amino acids can be calssified as essential, conditionally essential and non-essential amino acids and their role in food, pharmaceutical and cosmetic industry are discussed under these headings.