Chemical characterization of Central American pitaya (Hylocereussp.) seeds and seed oil

Abstract

Central American red pitaya (Hylocereussp.) seeds were studied for their chemical and nutritional traits with particular reference to the fatty acid (FA) profile of the seed oil. Proximate seed composition averaged 352 g/kg, 302 g/kg, 296 g/kg, 206 g/kg, 126 g/kg, and 21 g/kg for total carbohydrates, dietary fiber, fat, protein, moisture, and ash, respectively. FA composition of pitaya seed oil determined by GC–MS and quantified by GC-FID was dominated by unsaturated FAs (753 g/kg). Interestingly, linoleic acid was found to be the only polyunsaturated and most abundant FA (466 g/kg), whereas palmitoleic (3 g/kg), oleic (239 g/kg), and cis-11-vaccenic acid (45 g/kg) were the major monounsaturated FAs. Palmitic (182 g/kg), stearic (49 g/kg), and arachidic acid (18g/kg) represented the saturated FA fraction. Iodine absorption value, saponification number and free fatty acid value amounted to 105.6 g I₂/100 g, 235.7 mg KOH/g, and 1.9 mg KOH/g, respectively.

Semillas del fruto de pitaya roja (Hylocereussp.) cultivada en Centroamérica se analizaron químicamente, con énfasis en el perfil de ácidos grasos (AG) del aceite extraído de éstas. Del análisis proximal de las semillas se obtuvo 352 g/kg, 302 g/kg, 296 g/kg, 206 g/kg, 126 g/kg y 21 g/kg de carbohidratos, fibra dietética, grasa, proteína, humedad y ceniza, respectivamente. El perfil y la cuantificación de AG se realizaron mediante GC-MS y GC-FID. Se encontró un alto porcentaje de AG insaturados (753 g/kg), dominando el ácido linoleico con un 4.66 g/kg y siendo éste el único AG polinsaturado. Los ácidos palmitoleico (3 g/kg), oleico (239g/kg) y cis-11-vaccénico (45 g/kg) fueron los principales AG monoinsaturados encontrados. En cuanto a AG saturados, se encontraron palmítico (182 g/kg), esteárico (49 g/kg) y ácido araquídico (18 g/kg). El índice de absorción de yodo, el de saponificación y el valor de AG libres ascendieron a 105,6 g I₂/100 g, 235,7 mg KOH/g y 1,9 mg KOH/g, respectivamente.

Keywords:

• Hylocereussp.
• pitaya seeds
• vegetable oil
• fatty acid composition

Palabras clave:

• Hylocereussp.
• semillas de pitaya
• aceite vegetal
• composición de los ácidos grasos

Introduction

Vine cacti, native to tropical regions of North, Central, and South America, are known in Latin America as pitahaya or pitaya. The epiphytic plant has elongated three-ribbed stems that climb to trees and rocks (Nerd & Mizrahi, 1997; Nerd, Tel-Zur, & Mizrahi, 2002). Commercial pitaya cultivation in tropical regions, like Central America and the Far East, aim at the production of fruits for fresh consumption differing in size, shape, taste and color. The most widely cultivated vine cactus is the red pitaya, Hylocereus undatus[(Haworth) Britton & Rose] (red skin, white flesh). Although red pitaya is native to Central America, fruits for the Asian and European markets are produced in Vietnam and Israel (Nerd et al. 2002). Further commercial species include H. polyrhizus[(F.A.C. Weber) Britton & Rose] and H. costaricensis[(F.A.C. Weber) Britton & Rose], both exhibiting red skin and purple flesh, which have been suggested as a viable betalain source (Mizrahi, Nerd, & Nobel, 1997; Nerd et al., 2002; Stintzing & Carle, 2006). In Central America, which the fruits are native to, particular names have been assigned to Hylocereusgenotypes depending on their fruit morphology, scale number and shape. “Orejona”, “Rosa”, and “San Ignacio” are common genotypes in Nicaragua, where pitaya production is high. However, only scattered information on these genotypes is available and the assignment of these genotypes to definite species remains unclear (Esquivel, Stintzing, & Carle, 2007; Le Bellec, Vaillant, & Imbert, 2006).

Due to their high pigment content, the potential of pitaya as a source of natural colorants has frequently been highlighted, particularly since a recent study related several synthetic azo-dyes with hyperactivity in children (McCann et al., 2007; Stintzing & Carle, 2006; Vaillant, Perez, Davila, Dornier, & Reynes, 2005).

However, the production of natural colorants from pitaya fruit is technologically challenging due to poor process yields caused by the fruit mucilage, which firmly encapsulates the seeds, thus hindering their separation (Stintzing & Carle, 2007). Therefore, an enzyme-assisted liquefaction procedure at low temperature has been recently developed for the extensive degradation of pitaya mucilage. Using this procedure simple seed separation and excellent juice yields were achieved (Schweiggert, Villalobos-Gutierrez, Esquivel, & Carle, 2009), thus enabling easy concentration of betalainic food colorants. Hence, Schweiggert et al. (2009) proposed the recovery of pitaya seed oil as a valuable by-product of fruit processing into coloring preparations, since the seeds come to 0.46 up to 1.47 g/kg of the fresh pulp depending on Hylocereusgenotypes (Esquivel et al., 2007; Lim, Tan, Karim, Ariffin, & Bakar, 2010).

Characterization of seed oil components, e.g. their ratio of unsaturated to saturated fatty acids (FA), is of utmost nutritional and technological interest, since replacement of saturated by mono- or polyunsaturated FAs in human nutrition was shown to lower low-density lipoprotein (LDL) blood levels associated with lower incidence for coronary heart diseases (Mensink & Katan, 1992). Moreover, dietary intake of some FAs like linolenic and linoleic acid is essential for human health, since these FAs are metabolic precursors of eicosanoid hormones like prostaglandins (Whelan, 2008). Previous studies on red and white Malaysian Hylocereusseeds, after thermal degradation of fruit flesh for seed separation, revealed high contents of linoleic acid and nutritionally favorable low ratios of saturated fatty acids (Ariffin et al., 2009; Lim et al., 2010).

Generally, highly unsaturated seed oils such as those from wheat grain, sunflower, or almonds, contain considerable amounts of tocopherols preventing mono- and polyunsaturated FAs from oxidation during storage and food processing (Fisk, White, Carvalho, & Gray 2006; Frankel, 1996; Slover, 1970). As reported by Lim et al. (2010), H.undatusand H. polyrhizusonly contained moderate amounts of vitamin E, mostly  α-tocopherol (240–319 mg/kg seed oil) in their seeds. As expected,  α-tocopherol contents in wheat germ ranged between 1810 and 3518 mg/kg, while values from 523 to 749 mg/kg were reported for sunflower. However, sesame seed oil, which is also rich in linoleic acid, only contained 136 mg/kg (Slover, 1970; Tasan & Demirci, 2005; Wang & Johnson, 2001).

Due to its high content of unsaturated fatty acids, like linoleic acid and the presence of vitamin E, pitaya seed oil may represent an interesting source for the food and cosmetic industries. Applying the novel fruit liquefaction procedure proposed by Schweiggert et al. (2009), this study aimed at the characterization of the seeds of Central American pitaya fruits, since seeds were recovered without thermal or acidic mucilage degradation, this work should provide data on pitaya seed oils obtained by very gentle processing. Additionally, to assess the potential use of expeller cake, after seed oil extraction, the nutritional composition of the seeds, e.g. protein, fat, carbohydrates and fiber content should be investigated. Furthermore, a detailed fatty acid profile and the characteristic acid, iodine and saponification values of the seed oil should be determined.

Materials and methods

Plant material

Pitaya fruits (Hylocereussp. [Weber] Britton & Rose) of the genotype Rosa harvested in 2008 were obtained from APPINIC (La Concepción, Masaya, Nicaragua). Fruits were collected when their peel color started to change from green to red (about 28–32 days after anthesis), fruits were subsequently stored at room temperature and processed after full color development, which was completed after four days.

Solvents and reagents

All reagents and solvents were purchased from Merck (Darmstadt, Germany) and were of analytical or GC grade. Boron trifluoride–methanol complex solution (13–15% BF₃in methanol) was obtained from Riedel-de-Haen (Taufkirchen, Germany). FA and FA methyl esters (FAME) standards were purchased from Sigma-Aldrich (Steinheim, Germany).

Sample preparation

Three batches of about 7 kg of fresh fruits were processed in order to obtain 300 g of seeds, 100 g of the separated seeds were used for the proximal analyses, while the rest was used for oil extraction, with an oil extraction yield of about 2.5 g/kg for each seed batch. Fruits were cut in quarts and the peels were separated manually. Fruit quarters were strained through a 0.125 in. sieve in a finisher (Sterling, Power Systems, Hamilton, Canada). The obtained pulp was stored at −26°C until enzymatic liquefaction. Pitaya seeds were separated through filtration after optimized enzymatic liquefaction of the pulp as described by Schweiggert et al. (2009). The seeds were washed with water, air-dried and kept at room temperature until further processing.

Oil extraction

Dried seeds were ground in a laboratory mill (Grindomix GM200, Retsch, Düsseldorf, Germany) at 8500 rpm for 20 s. The resulting seed powder (5 g) was filled into a cellulose extraction thimble and oil extraction was performed with hexane in a Soxhlet extractor for 2 h. The solvent was subsequently evaporated in vacuousing a vacuum pump (Model V-700, Büchi, Flawil, Switzerland) attached to a rotary evaporator (Model B-480, Büchi, Flawil, Switerland) at 60°C and the residual solvent was removed by heating in an oven (model B5050 Heraeus, Hanau, Germany) at 100°C. The resulting oil was transferred into amber glass vials, flushed with nitrogen and sealed prior to storage at −25°C until analytical investigation.

Seed composition

Moisture (Method 925.09), fat (Method 920.85), protein (Method 920.152), ash (Method 940.26), dietary fiber (Method 985.29) were determined by triplicate using the Association of Official Analytical Chemists (AOAC, 2005) methods for three different seed batches. Carbohydrate contents were calculated as the arithmetically difference between 100% and the sum of the percentage of the analyzed components.

Fatty acid analysis

FAs were converted into their methyl esters by derivatization of 10–12 mg of oil samples according to the official standard procedure as described by Thurnhofer and Vetter (2005). The FA profile was analyzed by GC-MS using an Agilent gas chromatograph 6890 N connected with an Agilent 5975 mass selective detector. FA contents were determined using a Chrompack CP 9001 gas chromatograph (Chrompack, Middleburg, NL) equipped with an CP 9010 auto sampler and a FID detector. Separation was performed using a HP-5 column from Agilent (Waldbronn, Germany) (30 m ×0.25 mm i.d., particle size 0.25 μm) using helium (purity 5.0) as carrier gas at a constant flow of 1.2 mL/min. Oven temperature was set at 170°C, raised to 220°C at a rate of 1.5°C/min and subsequently ramped from 220 to 320°C at a rate of 50°C/min and held for 5 min. Injector and detector temperatures were set at 250°C and 290°C, respectively. Sample size was 1 μL and a split ratio of 20:1 was used. Each sample was injected twice. Heptadecanoic acid was added as an internal standard for samples and calibration curves. Individual FAs were identified and quantified by relating their retention times and peak areas of the samples to external fatty acid methyl ester standards. Data analysis was carried out using Maestro II 2.4 version software.

Seed oil quality parameters

Standard Association of Official Analytical Chemists (AOAC, 2005) methods were used to determine the acid (Method Nr. 940.28), iodine (Method Nr. 920.158) and saponification (Method Nr. 920.160) values of the extracted oil. For seed oil quality parameters, triplicate analyses of the extracted oil from several seed batches were conducted. In order to better describe the extracted oil, the maximal absorption wavelength was determined through a UV-Vis-scan (between 200 and 780 nm) using a photometer (Jasco, Esaton, USA).

Statistical analysis

All experiments were performed in triplicate for each one of the three batches analyzed, with exception of the fatty acid analysis (2.5), where samples were injected only twice for each batch. The results are expressed as the arithmetic average ± standard deviation for each parameter.

Results and discussion

Pitaya seed composition

A large amount of seeds imbibed in the fruit pulp of pitaya were observed (Figure 1a and 1b). These seeds were small, soft and black colored. Esquivel et al. (2007) reported seed contents among different genotypes of pitaya fruits ranging from 27 to 46 g seed for kg fruit. In this study, a seed yield of 39 g/kg was achieved. The proximate composition analysis of red pitaya seeds (Hylocereussp.) shown in Table 1was determined for three different seed batches. After storage of air-dried pitaya seeds, moisture contents of pitaya seeds were 126 ± 6 g/kg, close to the values (around 100 g/kg) reported for dried date (Phoenix dactylifera L.), passion fruit (Passiflorasp.), and apple (Malus domesticasp.) seeds (Besbes, Blecker, Deroanne, Drira, & Attia, 2004; Liu, Yang, Li, Zhang, Ji, & Hong, 2008; Yu, van de Voort, Li, & Yue, 2007).

Figure 1. (a) Red pitaya fruit, (b) seeds and (c) seed extracted oil.

Figura 1. (a) Fruto de la pitaya roja, (b) semillas y (c) aceite extraído de las semillas.

Table 1. Proximate composition of red pitaya seeds.¹
Tabla 1. Composición proximal de las semillas de pitaya roja.²

Component	g/kg seed (fresh weight base)
Moisture	126 ± 6
Total carbohydrates	352 ± 15
Dietary fiber	302 ± 19
Fat	296 ± 6
Protein	206 ± 6
Ash	21 ± 1
Note: ^1Average value of triplicate analyses (±standard deviation).
Nota: ^2Valor promedio de tres determinaciones (±desviación estándar).

Dry matter of pitaya seeds was mainly composed of 352 g/kg carbohydrates, mostly consisting of dietary fiber (302 g/kg), 206 g/kg protein, and 21 g/kg ash as the minor component. Protein contents of red pitaya seeds exceeded by far the values reported for Opuntiasp. seeds (50 g/kg) and date seeds, ranging from 52 to 54 g/kg (Besbes et al., 2004; Ennouri, Evelyne, Laurence, & Hamadi, 2005). Melon seeds from Citrullus colocynthisL. were reported to have similar protein contents of about 218 g/kg (Milovanović & Pićurić-Jovanović, 2005). Since current trends in the food oil industry aim at sustainable processing by valorization of by-products (Ramachandran, Singh, Larroche, Soccol, & Pandey, 2007; Weisz, Schneider, Schweiggert, Kammerer, & Carle, 2010), the comparably high protein content in pitaya seeds might allow a further economic exploitation by protein recovery from the oil cake. By analogy, by-products from passion fruit juice processing, particularly the seeds, are considered a promising feed stuff due to their high fiber and carbohydrates contents (Liu et al., 2008). Similarly, grape and melon seed oil cakes have been suggested to be used to improve animal feed or as fertilizers (Kamel, Dawson, & Kakuda, 1985; Milovanović & Pićurić-Jovanović, 2005). Consequently, further studies are required to evaluate the complete exploitability of press cake resulting from pitaya seed oil extraction.

The seed oil content (expressed as fat content, Table 1) observed for Central American red pitaya seeds in the present work (296 ± 6 g/kg), is in good agreement with those reported by Ariffin et al. (2009) for H. polyrhizus(295 g/kg) and H. undatus(320 g/kg) for Malaysian fruits. In contrast, Lim et al. (2010) found remarkably lower oil contents in H. polyrhizusseeds (183 g/kg) compared to H. undatusseeds (284 g/kg), also cultivated in Malaysia. Seed oil contents of other cactus species like Opuntia ficus indica(109 g/kg) and Opuntia stricta, (111 g/kg) were significantly lower than those of Hylocereussp. (Ennouri et al., 2005). Compared to other industrial sources of edible oil, fat contents in pitaya seeds even surpass those reported for soybean (164–250 g/kg). Moreover, they are close to the yields observed for sunflower seeds (250–400 g/kg) or other fruit kernels, e.g. passion fruit (Passiflorasp.), papaya (Carica papayaL.), and apple (Malus domesticasp.), amounting to 234, 280, and 277 g/kg, respectively (Jasso de Rodriguez, Phillips, Rodriguez-García, & Angulo-Sánchez, 2002; Liu et al., 2008; Nguyen & Tarandjiiska, 1995; Nodar, Gómez, & Martínez de la Ossa, 2002; Shankar, Agrawal, Sarkar, & Singh, 1997; Yu et al., 2007).

Seed oil analysis

The FA profile of the obtained seed oil was determined by GC–MS and quantification was performed by GC–FID analysis. As summarized in Table 2, predominant saturated FAs of pitaya seed oil were palmitic (182 g/kg), stearic (49 g/kg), and arachidic (18 g/kg) acids, which only represented 249g/kg of total FAs. In contrast to previous findings of Ariffin et al. (2009) and Rui, Zhang, Li, and Pan (2009) who reported minor amounts (2–12 g/kg) of myristic acid, the C14:0 derivative could not be found in the present study. Up to 753 g/kg of the total FAs identified in pitaya seed oil consisted of unsaturated FAs. Oleic (239 g/kg) and its isomeric form cis-11-vaccenic acid (45 g/kg) as well as palmitoleic (3 g/kg) were the monounsaturated FAs found in the seed oils. Similar compositions were previously reported for other Hylocereussp. (Ariffin et al., 2009; Lim et al., 2010; Rui et al., 2009). However, the long-chained mono-unsaturated FA eicosenoic (C20:1) and erucic acid (C22:1), as reported by Rui et al. (2009) and Lim et al. (2010), respectively, were not identified in the present study.

Table 2. Fatty acid profile of the oil extracted from red pitaya seeds (g/kg of oil).¹
Tabla 2. Perfil de ácidos grasos del aceite extraído de las semillas de pitaya roja (g/kg de aceite).²

Fatty acid (FA)	g/kg of oil
Palmitic acid C16:0	182 ± 11
Palmitoleic acid C16:1	3 ± 1
Stearic acid C18:0	49 ± 3
Oleic acid C18:1	239 ± 16
cis-11-vaccenic acid C18:1	45 ± 6
Linoleic acid C18:2	466 ± 42
Arachidic acid C20:0	18 ± 2
Saturated FA	249
Total unsaturated FA	753
Monounsaturated FA	287
Polyunsaturated FA (C18:2)	466
Note: ^1Average value of triplicate analyses (±standard deviation).
Nota: ^2Valor promedio de tres determinaciones (±desviación estándar).

Interestingly, high amounts of linoleic acid (466 g/kg) were found in the oil of tropical origin (Table 2). In agreement with our data, Ariffin et al. (2009) and Lim et al. (2010) also reported predominant amounts of linoleic acid in H. undatusand H. polyrhizusseed oil. High PUFA contents in Hylocereussp. were unexpected, since highly unsaturated FAs generally dominate in vegetable oils from plants adapted to moderate climate, like sunflower, corn, wheat germ, sesame and onion oils (Dubois, Breton, Linder, Fanni, & Parmentier, 2007; Wang & Johnson, 2001; Zheljazkov et al., 2008). The high amount of C18:2 FA makes Hylocereussp. seed oils a valuable source for both nutritional, and pharmaceutical applications. Besides its key role in the biosynthesis of prostaglandins, linoleic acid is controversially being discussed in the treatment of health conditions such as anti-inflammatory disorders and in the prevention of skin diseases (Sanders, 1988; Ziboh, Miller, & Cho, 2000). Further PUFAs, like linolenic acid, could not be found in Central American pitaya seed oil, whereas Ariffin et al. (2009) and Lim et al. (2010) described 9.8–13.0 g/kg linolenic acid (C18:3) in the seed oil of Hylocereusspecies. Nevertheless, pitaya seed oil can be considered a healthy source of human diet due to its high ratio of unsaturated and saturated FAs, which was associated with lower incidence of coronary heart diseases (Mensink & Katan, 1992). Moreover, pitaya seed oils were reported to contain  α-tocopherol (240–320 mg/kg) and  γ-tocopherol (116–127 mg/ kg), presumably protecting the highly unsaturated seed oil from lipid oxidation (Frankel, 1996; Lim et al., 2010; Slover, 1970). Diet rich in vitamin E is associated to several health benefits like, e.g. lower cardiovascular disease risks and protection against age-related diseases like cataract and macular degeneration (Viña, Gomez-Cabrera, & Borras, 2007; Yoshihara, Fujiwara, & Suzuki, 2010). Most health benefits are undisputed for food naturally rich in antioxidants like vitamin E, whereas vitamin E supplements are controversially discussed, since dosage-dependent adverse effects have been reported (Heinonen & Albanes, 1994; Heinonen et al., 1998; Kaliora, Dedoussis, & Schmidt, 2006; Steinmetz & Potter, 2003; Viña et al., 2007; Yoshihara et al., 2010).

Seed oil quality parameters

The oil had a yellowish color hue (Figure 1c) having a maximal absorbance at 453 nm, indicating the presence of yellow colored pigments. Further quality parameters of pitaya seed oil are listed in Table 3. Iodine absorption number of 105.6 g I₂/100 g showed comparable degrees of unsaturation as previously reported for H. undatus(101.2 g I₂/100 g) but higher as reported for H. polyrhizus(88.4 g I₂/100 g), these results are in agreement with the contents of unsaturated FA. According to Rui et al. (2009) the iodine absorption value was significantly affected by the extraction method used for oil recovery. Also seed oils from other cactus species, such as Opuntia ficus-indica(101.5 g I₂/100 g) and Opuntia stricta(91.6 g I₂/100 g) showed iodine absorption value within the same range (Ennouri et al., 2005). The saponification value of red-fleshed pitaya seed oil (235.7 mg KOH/g), indicative of the average molecular weight of the FA, was higher than previously reported for white-fleshed pitaya seed oil (194.4 mg KOH/g ) and other sources, like sunflower (188-194 mg KOH/g), apple (186.5 mg KOH/g), grapes (185–192 mg KOH/g), and raspberry (191 mg KOH/g) seeds (Oomah, Ladet, Godfrey, Liang, & Girard, 2000; Rui et al. 2009; Schieber, Müller, Röhrig, & Carle, 2002; Yu et al., 2007).

Table 3. Quality parameters for the oil extracted from red pitaya seeds.¹
Tabla 3. Parámetros de calidad del aceite extraído de las semillas de pitaya roja.²

Parameter	Values
Iodine absorption number (g I2/100 g)	105.6 ± 0.8
Saponification number (mg KOH/g)	235.7 ± 1.0
Free fatty acids (mg KOH/g)	1.9 ± 0.11
Note: ^1Average value of triplicate analyses (± standard deviation).
Nota: ^2Valor promedio de tres determinaciones (± desviación estándar).

Free fatty acids values for purple pitaya seed oil (1.9 mg KOH/g) were lower than those reported by Rui et al. (2009) for white pitaya seed oil (2.34 mg KOH/1.0 g). Since de-acidification may cause significant losses of triglycerides (Zwijnenberg, Krosse, Peinemann, & Cupersus, 1999), it should be also evaluated, by further stability studies, whether pitaya native oils may be used for consumption without raffination to increase the economic and nutritional value of pitaya seed oil.

Considering, the high contents of unsaturated fatty acids (753 g/kg), in particular oleic (239 g/kg) and linoleic acid (466 g/kg), pitaya seed oil represents a valuable source for culinary, cosmetic, and pharmaceutical applications. Oil quality parameters and characteristics of the fatty acid profile described in this study may be used as a tool for authenticity control of pitaya seed oils.

Conclusions

Red-fleshed pitaya seeds originating from Central American fruits proved interesting oil contents with high amounts of unsaturated FA. Also, the press cake resulting from oil extraction is regarded as a promising material for animal feed due the high protein content of the seeds. The amino acid profile of the press cake residual proteins should be considered in further studies in order to determine the nutritional quality of these proteins. Therefore, pitaya seed oil represents a valuable source for food, cosmetic, or pharmaceutical applications. Furthermore, determination of the profile of tocopherols and phytosterols in the lipid fraction could be of importance for authentication and nutritional evaluation of the oil. Also, studies on sensory properties and on the presence of antinutritive compounds are still missing.

Acknowledgments

The authors wish to thank F. Raoul-Duval and E. Hemme for providing the fruits used in this investigation, which was funded by SCRD (Le Havre, France) and the University of Costa Rica (project VI-735-A8-527). Special thanks go to the Baden-Württemberg Foundation, the University of Costa Rica and the Ministry of Science and Technology of Costa Rica (MICIT) for financing a short-term scholarship of one of the authors (M.G.V.) at Hohenheim University. Critical review of the manuscript by V.M. Jiménez is gratefully acknowledged.