Materials and Methods

The experimental site was at Jiangpu experimental farm located in Nanjing Agricultural University, Jiangsu Province of China (lat. 32°3′4.96′′N, long. 118°36′38.78′′W). The study was carried out during the growing seasons of Autumn and Spring 2013 and 2014. Low and high temperatures in the spring growing season were March (3 to 15 °C), April (9 to 21 °C), May (15 to 26 °C), and June (19 to 30 °C). The autumn season was August (25 to 33 °C), September (18 to 29 °C), October (12 to 23 °C), and November (5 to 17 °C). The soil at the experimental site is a clay loam.

The experimental design was a completely randomized block design where mature seeds of 46 cucumber cultivars and one wild species of cucumber were selected from the laboratory germplasm bank. The seeds were first germinated in small tubes and then transplanted into the experimental greenhouses with drip irrigation and polythene mulch. Five replicates of each cultivar were transplanted. The cultivars were conventionally grown and a blended fertilizer with micronutrients (12N–5.2P–10K) was applied at 30 g for each plant every 3 weeks. The plants were supported with strings and pruned to a single stem. Fertilization of the flowers was done manually for three fruit on each plant and any other fruit growing was removed to ensure optimal seed filling of the fertilized fruit.

Seed harvesting was done when the cucumber fruit was mature and had turned yellow and soft, at 30 and 45 d after fertilization in spring and autumn, respectively. Thus, harvesting in spring was at the end of June while autumn fruits were harvested in mid-November. Consequently, the cucumber fruit from fertilization to maturity in spring had approximate average temperatures of 19 to 30 °C and in autumn 12 to 23 °C. Seeds were then extracted from the fruit, washed, and sun dried. The seeds from individual fruit were counted, recorded, and kept in separate bags. The seeds were then selected from each bag and bulked in a separate bag for each cultivar for seed oil and fatty acid extraction and analysis.

Seed sample preparations for each cultivar were obtained from the bulked seeds and dried in an oven at 80 °C for 16 h to constant moisture. The Ultra-Turrax machine (Ika-Werke, Staufen, Germany) was used to grind seed samples in triplicates for seed oil extraction as well as fatty acid methylation. The grinder was washed with ethanol after each sample preparation to avoid contamination.

Seed oil extraction was determined according to the Association of Official Agricultural Chemists (AOAC) method (1990). The extraction was performed with diethyl ether solvent. An automated Soxhlet extractor (Universal Extraction System B-811/B-811 LSV; Büchi Labortechnik, Flawil, Switzerland) was used to extract cucumber seed oil. To achieve this, 10 g of ground seed powder was put in a folded filter paper and placed inside a 250-mL extractor device. The sample was extracted for 8 h at the solvent’s boiling point temperature. The proportion of diethyl ether was 150 mL in a 250-mL Soxhlet flask at 60 °C. The sample was put in an oven for 3 h at 105 °C and then weighed again to get the oil content:

Ground powder of 1000 mg of each cultivar from the bulked seeds was each immersed in petroleum ether: absolute ether in the ratio of 3 mL : 3 mL in a 10-mL centrifuge tube with a stopper which was vortexed and left for 16 h at room temperature. The crude oil supernatant was then extracted and put in a 10-mL centrifuge tube where 2 mL of KOH in methanol (0.5 mol Na per liter of methanol) was added for methyl esterification. Samples were then vigorously shaken for 1 min at room temperature, left for 2 h when 10 mL distilled water was added and left for another 1 h (AOAC, 1990). Finally, the supernatant was transferred into the auto sampler vials for gas chromatography-mass spectrometry (GC-MS) analysis in triplicates for each cultivar.

The GC-MS analysis of the methyl esters of the fatty acids (2 μL) was performed in a 7890 Series gas chromatograph (Agilent Technologies, Little Falls, DE) equipped with a flame-ionizing detector and a fused silica capillary column (HPINNOWAX, 19091N-133, 30.0 m × 250 μm × 0.25 μm; Agilent). This process was conducted at an oven temperature of 180 °C, which was then raised to 210 °C at a rate of 5.0 °C·min⁻¹ and then kept at 205 °C for 1 min. The oven temperature was raised to 230 °C at a rate of 3.0 °C·min⁻¹ and then kept for 4 min. The injector and detector temperatures were 250 and 300 °C, respectively. The carrier gas was nitrogen (99.999%) at a flow rate of 29.2 mL·min⁻¹ and Constant column was 0.8 mL·min⁻¹.

Peak identification was performed by comparing the relative retention times with those of a commercial standard mixture of FAME (Sigma-Aldrich, St. Louis, MO). The fatty acid quantity of linoleic, palmitic, oleic, stearic, linolenic, behenic, arachidic, lignoceric, eicosenoic, palmitoleic, myristic, and other fatty acids were determined using a computing integrator and recorded as a percentage of total fatty acids. Fatty acid percentages were then converted to milligrams per 100 g to get the amount of each individual fatty acid.

Data were analyzed using GenStat statistical package (VSN International, Hemel Hempstead, UK) using a two-way combined analysis of variance (ANOVA) for two seasons (spring and autumn) both in 2013 and 2014. All the effects were considered random for seed number, oil content, fatty acids, and the interaction of cultivars and seasons. Effects were considered statistically significant at P ≤ 0.01. Correlation was done using Spearman’s correlation as provided by GenStat. Season means were further compared pairwise for significance using the Bonferroni test.

Results

Results of the combined ANOVA indicated significant variations for the four seasons (Spring and Autumn 2013 and 2014) in seed numbers, oil content as well as all the fatty acids. Results were also highly significant for the 46 cultivars and the wild species in all the assessed parameters. Significant effects were also observed in the season and cultivar interaction for all parameters except for myristic acid (C14:0), cultivar, and season interaction at P ≤ 0.001 (Table 1).

Table 1.Combined analysis of variance mean squares for seed number, oil content, and individual fatty acids in cucumber cultivars grown in Spring and Autumn 2013 and 2014.

View Fullscreen

The fatty acids found in cucumber were linoleic, palmitic, oleic, stearic, linolenic, behenic, arachidic, lignoceric, eicosenoic, palmitoleic, myristic, and other fatty acids. Seed oil was significantly different in the study cultivars as shown in Table 1. ‘Hazerd’ had the highest seed oil (41.07%) while the wild species (C. anguria) had the lowest at 23.3%. The lowest of the cucumber cultivars was Lubao at 29.24%. ‘National Pickling’ cucumber had the highest seed number with 379.8 seeds and the lowest was ‘Chaoyou NO.3’ with 85.2 seeds. Myristic had significant variation in the study cultivars with the highest being in Marketmore 2000 (4.89 mg/100 g) and many cultivars did not have detectable amounts. The highest quantity of palmitic was in ‘627G’ (526 mg/100 g) and lowest in ‘Xintaimici’ (382 mg/100 g), although C. anguria had a lower measure at 238 mg/100 g. Palmitoleic was 6.28 mg/100 g in ‘CC3a’ and in ‘Pingwang’ was 2.02 mg/100 g, while C. anguria had 1.63 mg/100 g of the fatty acid. ‘EC5 Green’ with 292.62 mg/100 g had the highest stearic while ‘Xishuangbanna’ and C. anguria had 145.92 and 113.78 mg/100 g of the fatty acid, respectively (Table 2).

Table 2.Cucumber cultivar means for seed number, oil content, and individual fatty acids for four seasons from combined analysis of variance.

View Fullscreen

In Table 2 oleic acid ranged from 557.21 mg/100 g in ‘JC3’ to 249.61 mg/100 g in ‘Xishuangbanna’, and C. anguria had 142.99 mg/100 g. Highest amount of linoleic was 2460.55 mg/100 g in ‘Erzaozi’, while the lowest was in ‘Lubao’ at 1664.42 mg/100 g. C. anguria had 1634.63 mg/100 g. Linolenic acid ranged from 37.06 mg/100 g in ‘Marketmore 2000’ to 12.12 mg/100 g in ‘Xishuangbanna1’. Arachidic ranged from 24.14 mg/100 g in ‘Amir’ to 6.68 mg/100 g in ‘Xishuangbanna 2’. C. anguria had 5.36 mg/100 g. Highest amount of eicosenoic was 31.6 mg/100 g in ‘Pingwang 2’ and least was 5.07 mg/100 g in ‘PI 426169’. Behenic acid ranged from ‘Xintaimici’ (36.16 mg/100 g) to ‘Marketmore 2000’ (7.03 mg/100 g), while lignoceric varied from 21.31 mg/100 g in ‘Xishuangbanna1’ to 5.4 mg/100 g in ‘Lubao’. Other fatty acids were highest in ‘Amir’ (159.73 mg/100 g) and lowest in ‘Wynia1’ at 12.62 mg/100 g.

Correlations were tested for the different parameters as shown in Table 3. Palmitic acid had a positive relation with seed oil and fatty acids, including palmitoleic, stearic, oleic, linolenic, arachidic, and eicosenoic. Palmitoleic had also a positive relation with seed oil and fatty acids, including stearic, oleic, arachidic, and lignoceric. Stearic had a positive relation with seed oil and oleic acid.

Table 3.Spearman’s rank correlation (r) of seed number, oil content, and fatty acids in cucumber cultivars.

View Fullscreen

Oleic had a strong negative relation with linoleic and a positive correlation with seed oil, arachidic, eicosenoic, and lignoceric. Linoleic had a strong positive relationship with seed oil. Linolenic had a positive relation with seed oil as well as a negative relation with arachidic. Arachidic was positively related to eicosenoic while eicosenoic had a positive correlation with lignoceric. Lignoceric had a weak positive relation with seed oil.

Table 4 shows the season × cultivar interaction which is significant at P ≤ 0.001 for all measured variables except myristic acid as shown in Table 1. Table 4 illustrates interaction means for 2013 and 2014 during spring and autumn as derived from the two-way ANOVA. In Table 4 we observe that cultivars with consistently high seed numbers throughout the four seasons were National Pickling, Xishuangbanna 2, Beijingxiaoci, and Xintaimici. The cultivars with the highest seed oil quantities in all the four seasons were Hazerd, Amir, Beijingxiaoci, CC3b, and Marketmore 2000. Consistently high linoleic was in ‘Erzaozi’, ‘Hazerd’, ‘Pingwang 2’, and ‘PO2’ throughout the four seasons. In all the seasons oleic was highest in ‘JC3’, ‘Beit Alpha’, ‘P01’, ‘CC3b’, and ‘Beijingxiaoci’. Cultivars with consistently high palmitic were 627G, Poinsett 97, CC3a, PO2, and Lubao. ‘EC5 Green’, ‘Shuchunmeiya’, ‘PO3’, and ‘CC3b’ had consistently high stearic in all the four seasons.

Table 4.Comparison of Spring and Autumn 2013 and 2014 interaction means for season and cultivar in seed number, oil content, and individual fatty acids in cucumber cultivars.

View Fullscreen

Cultivar and season interaction means are shown in Table 5, with arrows showing the highest means in each variable. To further separate and get the most significant differences, the season means were compared pairwise using the bonferroni test as shown in Table 6. The pairwise comparison showed that in Spring and Autumn 2013 and 2014, the means were significantly different with exception of myristic, arachidic and behenic. In spring, cucumbers had higher oil content in addition to high palmitic, palmitoleic, stearic, oleic, eicosenoic, and lignoceric fatty acids while in autumn there were more seeds, higher linoleic, linolenic, and other fatty acids.

Table 5.Seasonal mean variation for seed number, oil content, and fatty acids in cucumber cultivars in Spring and Autumn 2013 and 2014.

View Fullscreen

Table 6.All pairwise mean verses mean comparisons for Spring and Autumn 2013 and 2014 using Bonferroni test for cucumber cultivars.

View Fullscreen

Discussion

In previous studies the seed oil contents in cucumber cultivars showed wide ranges. For instance; 33.2% (Matthaus et al., 2003), 41.7% (Lal et al., 1983), and 25.85% (Mariod et al., 2009). This implies that the seed oil content is influenced by the type of cultivar. This is consistent with our results which showed significant seed oil variation within cultivars. The detected fatty acids were linoleic, palmitic, oleic, stearic, linolenic, behenic, arachidic, lignoceric, eicosenoic, palmitoleic, myristic, and other fatty acids. Like in the seed oil, the fatty acid content also ranged widely within the cultivars. A previous study by Matthaus et al. (2003) detected similar fatty acids as well as vaccenic acid (C18:1n-7). Mariod et al. (2009) detected arachidonic which we did not detect. Another study detected similar fatty acids in one cucumber cultivar but did not detect myristic acid (Achu et al., 2006). Our results showed varying fatty acid profiles in different cultivars. Similar studies on other crops have shown that fatty acid profiles are not wholly similar in all cultivars of the same crop. For example, in seeds of bitter gourd [Momordica charantia (Ali et al., 2008)], pumpkin [Cucurbita pepo (Younis et al., 2000)], jute [Corchorus olitorius (Siriamornpun et al., 2006)], and rape [Brassica napus (Novickienė et al., 2010)] the fatty acid profiles were different. This can possibly explain why other studies that used different cultivars got varying fatty acid profiles as in our case where myristic and lignoceric were not detected in all cultivars and in some seasons since the contents were too low to be detected.

This study detected lignoceric acid that has not been reported before in cucumber. In an Uruguayan cucumber cultivar, fatty acids such as myristoleic (C14:1), caprylic (C8:0), capric (C10:0), and lauric (C12:0) as well as other unknown fatty acids were reported by Grompone (1988). This later report could explain why some cultivars had high amounts of other fatty acids, and thus supporting the findings that genotypic peculiarities partly govern fatty acid composition (Novickienė et al., 2010). This is further supported by the consistently low seed oil and fatty acids in C. anguria throughout all the seasons in our study.

The present results indicate that seed oil content and composition are primarily determined by the cultivar genotype due to similarities in related cultivars. Seed numbers were also cultivar dependent and did not affect seed oil content or composition. However, a cultivar with a high seed number coupled with high oil content will give higher oil yields. According to the data in this study many cultivars had sufficient oil content as well as seed numbers per fruit. Oil content and fatty acid profiles in cucumber are similar to those in other edible oils such as H. annuus oil which has a mean of about 37% (Pérez-Vich et al., 1998).

The cultivar’s genotype influences oil content and fatty acid composition. This is because a genotype is dependent on genes and enzymes that control synthesis of different fatty acids that in turn decrease or increase the composition of fatty acids. This is well demonstrated in Table 2 where similar cultivars such as Xishuangbanna 1 and Xishuangbanna 2, Xishuangbanna 3 and Xishuangbanna 4, PI 419017 and PI 426169, and Poinsett 76 and Poinsett 97 had very similar oil and fatty acid quantities. However, this was not always the case for all similar cultivars such as CC3a and CC3b; PO1, PO2, and PO3; and JC1 and JC3, which were slightly different possibly due to breeding that affected the genes controlling the seed oil biosynthesis.

Our results indicated that linoleic had a strong negative relation with oleic as well as a weak negative relation with stearic which agrees with those of other crops where increased oleic reduced linoleic and stearic percentages in seed oil crops (Izquierdo et al., 2013) as shown in Table 3.

Seed oil content increased slightly in spring which was also true for Glycine max, where high temperatures had little or no effect on oil content in the seeds (Ojo et al., 2002). However, the composition of the seed oil fatty acids were affected by seasonal changes as illustrated in Table 5. High temperatures have been found to alter seed oil composition by increasing oleic and decreasing linoleic and linolenic concentrations (Bellaloui et al., 2013). High temperatures during seed filling increased palmitic and oleic percentages and reduced stearic and linoleic percentages in H. annuus, suggesting some modification of enzymatic activities (Izquierdo et al., 2013) as a possible explanation for seasonal variation in fatty acid quantities.

The spring season plants encountered high temperatures during the seed filling stage hence the increased palmitic, palmitoleic, stearic, oleic, eicosenoic, and lignoceric fatty acids. High temperatures of 40/33 °C (day/night) resulted in higher palmitic and lower stearic acid (Bellaloui et al., 2013). Palmitic and stearic did not show any correlation with temperatures in our work. When the temperatures in Nanjing were 40/33 °C (day/night) in July, the cucumbers had already been harvested, thus missing the high summer temperatures that could have influenced the fatty acid compositions.

Low temperatures reduced oleic quantity (Pritchard et al., 2000) due to the activity of microsomal oleoyl phosphatidylcholine desaturase being enhanced at low temperatures of between 10 and 20 °C. This later observation was explained as being due to increased enzyme synthesis, increasing linnoleate quantities (Garcés et al., 1992). Autumn cultivars had more seed numbers as well as increased quantities of linoleic, linolenic and other fatty acids, and this could be due to lower temperatures during seed filling that occurred when temperatures were decreasing toward winter. This findings agree with results reported by Martínez-Rivas et al. (2003) where linoleic increased at “25 °C day/15 °C night” temperature due to increased activity of microsomal oleate desaturase (FAD2).

The main environmental factor affecting fatty acid composition is whether temperatures are high or low following the onset of flowering (Baux et al., 2008), hence determining the quantity of fatty acids in the seed. An increase in temperature from 10 to 40 °C resulted in an increased synthesis of oleic and a decreased synthesis of linoleic and linolenic in G. max, B. napus, and H. annuus resulting in more monounsaturated oils and less polyunsaturated oils (Schulte et al., 2013).

The results of our study show that the seed oil content in cucumber is sufficient for commercial extraction. The oil content and fatty acid composition is comparable to that of other commercial oils and can therefore be exploited as edible oil, nutraceutical and cosmetic ingredient in addition to other industrial uses. Cucumber cultivars have a high variation in oil content and fatty acid compositions and are therefore suitable for genetic engineering to produce different novel fatty acids for specialized uses.

Cucumber seed oils have similar characteristics to other commercial seed oils, implying that methods used to increase oil content and quality in other vegetable oils can also be successfully applied in cucumber. Seed oil and fatty acid synthesis is a complex interaction between environmental conditions, seed growth, and oil synthesis dynamics (Echarte et al., 2013), and further research in a climate controlled environment can be done to further validate the results obtained.

We conclude that there is a highly significant variation both in oil content (41.07% to 29.24%) and fatty acid composition in the cucumber cultivars that can be used for improvement. The high oil content coupled with the reasonably high concentration of fatty acids makes the seed oil suitable for further development and utilization as a food supplement, cosmetic, pharmaceutical product, insecticide, and other novel industrial uses. Consequently, the information gotten from this study on seed number, oil content, fatty acid composition, and seasonal variation, which was for instance very significant in oleic where spring had 546.3 mg/100 g and autumn 296.3 mg/100 g, will be useful in future cucumber breeding research and seed oil exploitation for relevant industrial development and economic gains from cucumber-based entrepreneurs.