Variability of seed oil content, fatty acid composition, and nervonic acid content in Acer truncatum, native to 14 regions of China

Q. Qiao, W. Xue and Z. Feng^*

¹College of forestry Shandong Agricultural University, Taian 271018, Shandong, China

*Corresponding author: fengzn408@qq.com; 1226694758@qq.com

1. INTRODUCTIONTOP

Plant seeds are an important source of oil, which serves as a staple agricultural food product and an important industrial raw material. Moreover, plant seed oils are an increasingly important renewable and environmentally safe bioenergy resource. However, increasing demand for seed oil has resulted in increasingly prevalent and serious oil shortages. To address this problem, efforts have been focused on the development of woody oil trees, which have recently become a major edible oil source in some western European countries. Woody crop oils with a large share of the edible oil market include coconut, olive, palm, almond, walnut, and tea oils. However, there are many potentially excellent woody edible oils in nature that have not been exploited, including those coming from maple tree species.

Acer truncatum, a member of the Aceraceae family, and is a forest tree species native to northern China, Korea and Japan, but can also be found in Europe and North America (Guo et al., 2014). It is a deciduous tree species native to northern and western China, and it is mainly distributed in the Yellow River basin, the northeast, Inner Mongolia, Jiangsu, Sichuan and other regions. This species of maple has been commonly chosen as a landscape tree for its brilliant autumn leaf color (Zhao et al., 2007; Li et al., 2015). A. truncatum is also an ideal source of timber, protein, medicinal substances, and chemical raw materials due to its sturdy texture and high levels of protein, chlorogenic acid, tannin, and flavonoids (Ma et al., 2005). In addition, its abundant oil seeds have been roasted for many years for use as a food source in northeastern China, but have not yet been used on a large scale. Previous studies (Wang et al., 2006; Hu et al., 2017; Sun et al., 2018) have suggested that maple seed oil would be an excellent edible oil because it contains a large percentage of unsaturated fatty acids (92%) and nervonic acid (6.22%). Notably, nervonic acid, an important fatty acid for brain growth and maintenance, has been shown to prevent neural deficits and shows promise for the treatment of neurological disorders such as schizophrenia and psychosis (Akoh et al., 2001; Barcarolo et al., 2003; Chen et al., 2017). It has a high content of vitamin E (125.23mg/100g) (Wang et al., 2006). Therefore, A. truncatum is regarded as a potential species for use in the development of valuable nutritional and medicinal substances.

A. truncatum is also a renewable biomass energy tree species which produces approximately 30 kg of fruit per tree after 20 years and its seed ratio is up to 66.4% (Wang et al., 2006). Currently, it is receiving much attention from researchers in China due to its great potential value for energy and other numerous applications. Consequently, the seed oil was approved as a new food resource by the Chinese Ministry of Health in 2011. A. truncatum seed oil (ATO) is usually obtained by cold pressing and commercialized as a crude oil in China (Hu et al., 2017). At the present time, the A. truncatum artificial cultivation area encompasses 4 × 10⁴ ha and should continue to expand rapidly due to this tree’s numerous industrial prospects. However, as most previous studies have focused on the edible, medicinal, and health care uses of its seed oil, no studies have investigated variations in oil content or fatty acid composition among natural stands. Moreover, only limited information is available regarding the effects of environmental factors on seed oil yield and quality.

In this study, we selected 138 accessions from 14 A. truncatum tree populations growing across the known natural distribution area of this species in China. The objectives were to analyze seed oil variation and investigate the influence of environmental factors on seed oil content and fatty acid composition. The information obtained from these studies should guide breeding programs, aid in selection of the most adaptable lines for large-scale cultivation, and ultimately stimulate further development of the ATO industry in China.

2. MATERIALS AND METHODSTOP

2.1. Plant materialTOP

In October and November 2016, 138 accessions of A. truncatum that grow naturally in fourteen regions were collected from nine provinces of China. About 10 separate pest-free adult plants (each 20 years of age or older) were selected from each sample collection area. In order to minimize other factors affecting seed development, individual tree species were chosen with spacing to other trees of at least 50 m. For each germplasm, 1-2 kilograms of fully matured samaras were randomly selected from multiple locations on each tree to ensure that the sample represented each whole plant. Samaras were stored at room temperature. After a period of time, the seeds are stripped from the Samaras manually. Using GPS to record latitude, longitude and altitude, the meteorological factors were listed using the data from the local meteorological department (Table 1).

2.2. Oil content and fatty acid composition analysisTOP

A Bruker minispec mq20 pulsed nuclear magnetic resonance instrument (pulsed NMR) was used to estimate the seed oil content of each material. The specific method used here follows the official standard method (ISO 5511:1992, GB/T 15690-1995, AOAC, 2005).

Oil extraction was performed using a Soxhlet apparatus with ~5 g of ground seeds and petroleum ether (60–90 °C) solvent according to published methods (Hu et al., 2017). The pure seed oil was transferred to a vial, flushed with nitrogen, and kept at -20 °C until further analysis. The seed oil was methylated twice, the first step being pre-esterification with H₂SO₄-CH₃OH to reduce the acid value to less than 1 mg KOH/g; the second step was trans-esterification with KOH-CH₃OH, according to the published method (ISO 5509:2000, GB/T 17376-2008). The fatty acid methyl ester (FAMEs) profiles obtained for each accession were determined using an Agilent 7890A (Agilent, Palo Alto, CA, USA) gas chromatograph (GC) equipped with a flame ionization detector (FID) using 17:0 FAME as an internal standard. The DB-23 capillary column (length 30 m, internal diameter 0.32 mm, film thickness 0.25 μm) was used for this test. The injector and detector temperatures were 230 °C and 280 °C, respectively. The oven temperature was maintained at 180 °C for 5 minutes and raised 3 °C·min⁻¹ to 230 °C. The carrier gas (helium) was delivered at a flow rate of 1.0 ml·min⁻¹, and 1 ml of the sample was manually injected in a split injection mode. FAME peaks were identified by comparing the retention times of known standards which were run under the same conditions. Peak integration was performed using instrument software.

2.3. Statistical analysisTOP

R software was used for correlation analysis, hierarchical cluster analysis (HCA), and principal component analysis (PCA); Excel 2010 was used for other calculations. Determinations were run in duplicate and the data were reported as mean values.

3. RESULTSTOP

3.1. Oil content and fatty acid compositionTOP

Oil content is an important indicator to measure industrial production. In this work, the oil content of A. truncatum seeds ranged from 17.81%–36.56%, with a mean value of 28.57% for all 138 accessions studied (Table 2). Among the concessions, we selected three with very high oil content (>35%), including YS-6 (36.56%), ABZ-6 (35.71%), and DQTL-8 (35.44%), which would likely be valuable for the development of new varieties with high oil contents.

With regard to A. truncatum, a total of 14 fatty acid components were detected in this study (Figure 1). The results (Table 2) show that ATO is highly unsaturated (89.60%–92.90%), with an unsaturated to saturated ratio of approximately 10:1. The oil contains mainly oleic acid (C18:1) (20.00%–34.31%), linoleic acid (C18:2) (27.08%–36.71%), cis-11-eicosenoic acid (C20:1) (6.48%–9.15%), erucic acid (C22:1) (13.64%–18.86%), nervonic acid (C24:1) (3.90%–7.85%) and linolenic acid (C18:3) (1.60%–4.35%) which is composed of g-linolenic acid (GLA, 0.20%–1.36%) and α-linolenic acid (ALA, 0.97%–3.51%). The saturated fatty acid (SFA) content was low (7.32%–10.31%), but may stabilize the fat. SFA consisted predominantly of palmitic acid (C16:0) (3.78%-6.12%) and stearic acid (C18:0) (1.31%–3.30%). Palmitoleic acid (C16:1), arachidic acid (C20:0), cis-11,14-eicosadienoic acid (C20:2), behenic acid (C22:0), and tetracosanoic acid (C24:0) were also detected, but were not further characterized due to their small amounts (<1%), and therefore were labeled as other fatty acids.

3.2. Variation of oil content and fatty acid composition among populationsTOP

The oil content varied significantly (p < 0.01) among the 14 analyzed populations (Table 2). The highest seed oil content was exhibited by the DQTL population (32.47%), followed by YS (32.09%), with the lowest content observed in the NJ population (24.06%). Therefore, based on these results, the populations of DQTL and YS were chosen as germplasm with high oil contents for screening purposes.

The main fatty acid composition varied significantly (p < 0.01) among the 14 analyzed populations (Table 2). Unsaturated fatty acids (UFA) are classified into monounsaturated fatty acids (MUFA) and polyunsaturated fatty acids (PUFA) according to the number of fatty acid double bonds. The populations NJ and NX exhibited the highest MUFA levels (> 57.96%), while populations CF and ABZ exhibiting the lowest MUFA levels (< 52.87%). PUFA percentage tended to be inversely correlated with MUFA percentage. Meanwhile, populations DQTL and CF exhibited a high (> 35.08%) content of C18:2 and relatively low content of C18:1, while TA, NJ, LF and NX populations exhibited a high (> 26.32%) C18:1 content and relatively low C18:2 content. Overall, the sum of percentages of C18:1 and C18:2 was consistently 60%, a previously known relationship prevalent in oil crops for which oleic acid is a precursor of linoleic acid (Yu et al., 2017; Darwish, 2014). The highest amount of C18:3 was detected in populations YS (3.12%) and ABZ (3.10%). The content of C20:1 ranged from 7.34% (ABZ) to 8.31% (LF). JY, NX, and PQ samples exhibited the highest amounts of C22:1 (16.72%, 17.50%, and 17.23%, respectively) and C24:1 (6.54%, 6.18%, and 6.29%, respectively). Therefore, JY, NX, and PQ can be regarded as excellent germplasm resources which are especially valuable due to their C24:1 contents. The SFA content levels for populations FZ (9.47%), ABZ (9.30%), and JY (9.05%) were obviously higher than for other populations, as was also observed for C16:0 and C18:0 contents.

In summary, selection of excellent germplasm resources can be used to obtain optimal oil content and composition for various applications. The findings reported here will be beneficial as a starting point for future breeding efforts toward improving the yield and quality of ATO.

3.3. Effects of geography and ecological factors on seed oil content and fatty acid compositionTOP

The correlation analysis is shown in Table 3. Oil content showed significant negative correlations with annual average temperature (p < 0.05, −0.629), annual rainfall (p < 0.01, −0.749), and frost-free season period (p < 0.05, −0.562) and a positive correlation with latitude (0.503) and altitude (0.469), indicating that northern high altitude populations exhibited higher oil contents within the sampling area (such as DQTA and YS). Meanwhile, according to the correlation analysis, SFA and UFA showed a tendency toward longitudinal variation, while C18:2 and C22:1 exhibited latitudinal variation and C16:0 exhibited northeast-to-southwest variation.

Temperature is considered an important factor affecting the composition of fatty acids. It showed a significant negative correlations with PUFA (p < 0.05, −0.584) and C18:2 (p < 0.05, −0.601), although a significant positive correlations with C16:0 (p<0.01, 0.666). Meanwhile, environmental factors other than temperature also play a role in the fatty acid compositions of oils. Precipitation exhibited a significant negative correlation with C18:2 (p < 0.01, −0.790), PUFA (p < 0.01, −0.802) and a significant positive correlation with C24:1 (p < 0.05, 0.592), and MUFA (p < 0.01, 0.741). As a final observation, the frost-free period had the lowest effect on fatty acid composition, only significantly and negatively correlating with C16:0 (p < 0.05, 0.617).

The correlation study among different biochemical and geographical parameters could be used to understand their interrelations for future breeding, and for selecting an ideal location for growing high yield or high quality A. truncatum plants. From this study, we determined that an improvement in fat production and in the quality of ATO can be achieved by planting trees in moderately cold areas, meaning that the northeast and northwest regions of China are the best place to grow.

3.4. Hierarchical cluster analysis and principal component analysisTOP

To evaluate the likely similarities and relationships among different populations, hierarchical cluster analysis (HCA) was performed based on 13 traits selected for this study (Figure 2A) (Szekely and Rizzo, 2005). This analysis provided the basis for understanding variations in fatty acid composition in combination with geographical influences. Based on this analysis, 14 populations of A. truncatum were classified into three groups. The first group consisted of the five populations of TA, LF, NX, JY, and NJ, all rich in MUFA (56.26%–58.79%) and C18:1 (24.93%-27.91%) (Table 2). It should be noted that JY differed from the others in of its exceptionally high level of C16:0 (5.14%). Populations FZ and ABZ, with higher amounts of SFA (> 9.30%) plus a moderate amount of PUFA, were classified in the second group. The third group was further divided into two subgroups. The first subgroup consisted of two populations (DQTL and CF) with high PUFA (> 37.97%) and C18:2 (> 35.08%) compositions, while the CC, PQ, YJ, HY, and YS populations, with higher quantities of UFA (> 91.62%) than other populations, were assigned to the second subgroup.

In the second stage of analysis, PCA was performed on all populations and all traits were evaluated in order to explore the interrelationships among and within populations. Most (84.04%) of the variation was explained by the first three principal components (PC1, PC2, and PC3). The result of PCA revealed that the first and second components explained 70.28% of the total variation. The first PC (PC1) accounted for 44.80% of the total variation and had a high positive correlation with MUFA (0.410) and C18:1 (0.362) (Table 4), but a negative correlation with PUFA (-0.394) and C18:2 (-0.386). The PC2 results explained 25.47% of the total variation. A high negative contribution by SFA (-0.516) and C16:0 (-0.478) and positive contribution by UFA (0.509) were also observed.

Further analysis results shown that the four populations of TA, LF, NX and NJ were positively correlated with MUFA, C18:1, C20:1 and C22:1. The population of FZ, ABZ and JY was positively correlated with SFA and C16:0. The remaining 7 populations were positively correlated with oil content, UFA, PUFA and C18:2. Notably, PCA classification confirmed the results of HCA apart from JY (Figure 2B).

4. DISCUSSIONTOP

Yield is the trait most predictive of whether or not an oil is suitable for large-scale production and adaptability to multiple industrial uses. Consequently, crops that contain high oil content at harvest ultimately reduce overall final production costs (Kumar and Sharma, 2011). This mean content (28.57%) value for A. truncatum greatly exceeded values for Glycine max (17%), Olea europaea (20%), and most results for Sapium sebiferum (12–29%) (USDA, 2012; Karmakar et al., 2010), and was comparable to values for Jatropha curcas (20.05%-38.33%) (Kaushik and Bhardwaj, 2013) and wild manihot (17%-31%) (Alvesa et al., 2014). This proves that it is feasible to extract oil from the seed of A. truncatum for industrial purposes.

Seed oil quality and uses are determined mainly by fatty acid composition, with rich variety in fatty acid compositions and content observed among the various species (and even among varieties). Therefore, fatty acid composition and content can be used as fingerprints to identify biological oil sources, as well as to test the authenticity of oils (Li et al., 2011). At the species level, a total of 14 distinct fatty acid components were detected in this study, which is similar to that reported by Hu et al., (2017), and a number similar to that obtained for Xanthoceras sorbifolia by Yu et al., (2017), although their oil content differed from the values reported here. The degree of unsaturation (89.60%–92.90%) of ATO is higher than peanut oil (83.1%), soybean oil (84.4%), olive oil (86.2%), cotton seed oil (74.1%) and other edible oil (USDA, 2012). C18:2 (27.08%–36.71%) is essential for human metabolism due to the lack of enzymes responsible for its biosynthesis (Hanganu et al., 2012), and it contains a special substance named nervonic acid that is an essential fatty acid which helps maintain brain health by its participation in biosynthesis and maintenance of nerve cell myelin (Akoh et al., 2001; Barcarolo et al., 2003). Due to its scarcity, C24:1 draws a high purchase price. Therefore, it is of great significance to find alternative plant sources rich in C24:1. A. truncatum is one of a handful of known species that contains C24:1 in seed oil, including Brassicaceae, Lunaria annua, Borago officinalis, Cannabis sativa, Tropaeolum speciosum, Cardamine graeca, Malania oleifera, and X. sorbifolia (Katavic et al., 2010; Chen et al., 2017; Yu et al., 2017). Therefore, the C24:1 content is an important indicator of ATO value. The content ranged from 3.90% to 7.85%, with a mean value of 5.76% in this study, which is much higher than the content of the current source of medicinal nervonic acid, X. sorbifolia (1.52%–3.04%) (Yu et al., 2017). And the yield per plant is rich (about 30 kg of fruit after 20 years). So, A. truncatum seeds will become a promising raw material to extract nervonic acid.

Variability is the result of different environmental selection forces that are the root cause of phenotypical differentiation among populations. In this study, altitude, temperature, and precipitation are the main environmental factors causing variation. The trend is the same in the study of Dewhurst and King (1998), and Darwish (2014), demonstrating higher oil yields in plants collected from relatively higher altitudes. However, the oleic acid content differed from that of previous studies in that it is dependent on mean temperature and can decrease by up to 2% with each degree (ºC) of increase in temperature (Rondanini et al., 2011). However, no significant relationship between mean temperature and oleic acid content in oils was found in this study, but significant negative correlations were observed between annual average temperature and both PUFA (-0.584) and linoleic acid (-0.601). It has been demonstrated that low temperatures increase the PUFA content of plants to maintain the fluidity of biological membranes (Falcone et al., 2004); in a high temperature environment, the increase in SFA could be explained by lipid peroxidation (Arbaoui and Link, 2007). That also explains the observation that palmitic acid exhibits a significantly positive correlation with annual average temperature (0.666).

Laribi et al., (2009) and Rebey et al., (2011) observed water stress effects on lipid metabolism. In our study, precipitation exhibited a significant negative correlation with C18:2 (-0.790) and a significant positive correlation with C24:1 (0.592). In addition, researchers have noted that an increase in altitude has tended to promote greater UFA content (with subsequently lower SFA content), possibly because altitude increases lead to lower temperatures and higher exposure to light (Dewhurst and King, 1998; Lai et al., 2010). However, this relationship is in contrast to the results of this study, which coincide with Vicia sativa by Mao et al., (2012) who found that other environmental factors besides temperature and light intensity, such as soil and air, may also influence the lipid metabolism of plants. This will be the direction of our future work.

In addition to environmental factors, genetics need to be considered. Rahimmalek et al., (2017) suggested that the thymol content in Iranian Ajowan (Trachyspermum ammi) can be affected by both genetic and environmental factors. It is necessary to consider the differences within and among populations when screening promising germplasm resources for oil use, since the biosynthesis of fatty acids can be regulated by both genetic and environmental factors simultaneously. Thus, further research should be conducted to explore the effects of environment and genotype in order to guide in the selection of genotypes which exhibit high oil yield and useful fatty acids.

Results of HCA and PCA showed that within a certain geographic area, populations with similar physiological characteristics can be clustered into a class which is the same as the result of Rahimmalek et al., (2017) on Iranian Ajowan. Within a class, it was observed that the fatty acid composition of ATO assumed variation patterns which are characteristic of geographic isolation. Fatty acid profiles could be roughly correlated with growth in three major geographic areas of China: (I) Eastern central region (TA, LF, NX, JY, and NJ, with high MUFA especially C18:1); (II) Southwest region (FZ and ABZ, with high SFA); (III) Northeast and northwest regions (DQTL, CF, CC, PQ, YJ, HY, and YS, with high UFA, especially PUFA and C18:2). These findings also indicate that it is feasible to purchase seeds in the region which could guide future development of ATO for industrial production.

5. CONCLUSIONSTOP

A. truncatum is a woody oil tree species with high seed oil content (28.57%), high unsaturated fatty acid content (91.47%), and high nervonic acid content (5.76%) ompared to grain and other seed oil sources. Notably, the data presented in this study show patterns of variability in oil content and fatty acid composition which correlate with geographic location: northern high altitude populations exhibited higher oil content; populations in the eastern central region exhibited high levels of MUFA (especially C18:1); populations from the southwest region were rich in SFA; populations from the northeast and northwest regions had high UFA content (especially PUFA, and C18:2).

Seed oil content and fatty acid composition were found to be influenced by ecological factors, especially altitude, temperature, and precipitation. The results suggest that both fat production and quality of ATO may be improved by planting trees in moderately cold areas. Consequently, further studies will incorporate such results to guide in the selection of excellent individual trees which have high specificity (such as high oil, high C24:1) to improve oil yield and quality. The populations of DQTL and YS were chosen as germplasm resources with high oil content, while JY, NX, and PQ were regarded as excellent germplasm resources with high C24:1. This work should lay the groundwork for improving oil yield and quality and should ensure ATO’s future success as an important industrial crop.

ACKNOWLEDGMENTSTOP

This work was supported by Central Fiscal Forestry Science and Technology Extension Demonstration Fund Project (No: 2016LZGC014), a Major Agricultural Project of Shandong Province (No: [2012] 213 and [2014] 96) and Shandong “Double Tops” Program. We are grateful to the Institute of Oil Crops of the Chinese Academy of Agricultural Sciences for providing technical support in the tests.

REFERENCESTOP