Next Article in Journal
Engineering of g-C3N4 for Photocatalytic Hydrogen Production: A Review
Previous Article in Journal
Preclinical Evaluation of Soluble Epoxide Hydrolase Inhibitor AMHDU against Neuropathic Pain
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
IJMS
Volume 25
Issue 16
10.3390/ijms25168840
ijms-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

High-Density Genetic Map Construction and Quantitative Trait Locus Analysis of Fruit- and Oil-Related Traits in Camellia oleifera Based on Double Digest Restriction Site-Associated DNA Sequencing

by
Ping Lin
1,2,
Jingyu Chai
1,2,
Anni Wang
1,2,
Huiqi Zhong
1,2 and
Kailiang Wang
1,2,*
1
State Key Laboratory of Tree Genetics and Breeding, Research Institute of Subtropical Forestry, Chinese Academy of Forestry, Hangzhou 311400, China
2
Zhejiang Key Laboratory of Forest Genetics and Breeding, Research Institute of Subtropical Forestry, Chinese Academy of Forestry, Hangzhou 311400, China
*
Author to whom correspondence should be addressed.
Int. J. Mol. Sci. 2024, 25(16), 8840; https://doi.org/10.3390/ijms25168840
Submission received: 2 July 2024 / Revised: 9 August 2024 / Accepted: 12 August 2024 / Published: 14 August 2024
(This article belongs to the Section Molecular Plant Sciences)
Browse Figures
Versions Notes

Abstract

:
Camellia oleifera, an important tree species and source of edible oil in China, has received significant attention owing to the oil’s high unsaturated fatty acid content, which has benefits for human health. However, the mechanisms underlying C. oleifera yield and oil quality are largely unknown. In this study, 180 F1 progenies were obtained from two parents with obvious differences in fruit- and oil-related traits. We constructed a high-density genetic map using a double digest restriction site-associated DNA sequencing (ddRAD-Seq) strategy in C. oleifera. This map spanned 3327 cM and anchored 2780 markers in 15 linkage groups (LGs), with an average marker interval of 1.20 cM. A total of 221 quantitative trait loci (QTLs) associated with fruit- and oil-related traits were identified across three years’ worth of phenotypic data. Nine QTLs were detected simultaneously in at least two different years, located on LG02, LG04, LG05, LG06, and LG11, and explained 8.5–16.6% of the phenotypic variation in the corresponding traits, respectively. Seventeen major QTLs were obtained that explained 13.0–16.6% of the phenotypic variance. Eleven and five flanking SNPs of major QTLs for fruit- and oil-related traits were detected which could be used for marker-assisted selection in C. oleifera breeding programs. Furthermore, 202 potential candidate genes in QTL regions were identified based on the collinearity of the genetic map and the C. oleifera “CON” genome. A potential regulatory network controlling fruit development and oil biosynthesis was constructed to dissect the complex mechanism of oil accumulation. The dissection of these QTLs will facilitate the gene cloning underlying lipid synthesis and increase our understanding in order to enhance C. oleifera oil yield and quality.

1. Introduction

Camellia oleifera, a species belonging to the genus Camellia (Theaceae), originates from China and has a cultivation history of over 2000 years in south China [1]. It is widely recognized as a significant source of edible oil in China, with a cultivated area exceeding 4 million hectares [2,3]. It is also widely used in cosmetics, ink, lubricants, etc. [4,5]. The oil extracted from the seeds of C. oleifera, known as Camellia oil, exhibits a distinctive chemical composition and possesses notable medicinal and healthcare properties. It is rich in unsaturated fatty acids (comprising over 90% of the total oil), of which the monounsaturated fatty acid, oleic acid, accounts for about 80% of the total oil content [6]. Camellia oil also contains squalene, tea polyphenol, and other bioactive substances [7] which are beneficial for reducing the level of blood cholesterol and treating cardiovascular and cerebrovascular diseases [7,8]. For these reasons, C. oleifera has received a significant amount of attention, and the yield and quality of the seed oil have been continuously selected as the primary targets in C. oleifera breeding [1].
Due to the long breeding cycle, C. oleifera poses unique challenges for genetic improvement, and there is an urgent need to improve upon traditional breeding methods [9]. High-resolution genetic maps are indispensable to many genetic studies and molecular breeding programs, such as molecular-marker-assisted selection, fine-scale quantitative trait locus (QTL) mapping, genome scaffolding and assembly, and comparative genomic analysis. Advances in the genetic map construction and QTL analysis of Brassica napus, soybean [10,11,12], oil palm [13], and other plants have revealed novel findings in fruit development and the oil biosynthesis pathway and provided valuable genomic resources for genetic improvement in oil crops [14,15,16,17]. The application of molecular-marker-assisted breeding is anticipated to expedite C. oleifera breeding cycles. However, the complexity of C. oleifera’s genetic background, a scarcity of molecular markers, and the challenges in hybrid population creation have hindered the progress of genetic map construction and molecular-marker-assisted selection studies. C. oleifera is a self-incompatible tree with a highly heterozygous genome (heterozygosity 2.52%), and most cultivars are hexaploid (2n = 6x = 90) [18]. Therefore, the construction of high-density genetic maps for C. oleifera has always been an enormous challenge. Recently, chromosome-level diploid oil-Camellia genomes (C. oleifera var. ‘Nanyongensis’ (CON), C. lanceoleosa and C. chekiangoleosa, 2n = 2x = 30) have also been reported, and genome sizes of 2.95 Gb, 3.00 Gbs and 2.73 Gb were obtained, respectively [18,19,20]. The diploid “CON” is considered to be the ancestor species of the hexaploid C. oleifera [18], and the “CON” genome sequence is advantageous to hexaploid C. oleifera breeding. To date, only one genetic map of C. oleifera has been constructed, which has poor marker density, with only 300 SNP markers [18]. To date, no research on QTLs associated with C. oleifera traits has been published. C. oleifera often has larger fruit, a thinner pericarp, and higher oil content in the kernel, allowing for a boost in seed oil yield [18]. However, an understanding of the molecular basis of the regulation of fruit and seed development and oil and fatty acid biosynthesis in C. oleifera is limited [21,22,23,24]. Marker development and key gene mining based on a high-density genetic map and fine-scale QTL mapping are necessary for boosting the genetic improvement of oil traits in C. oleifera.
For the construction of the high-density genetic maps, single nucleotide polymorphism (SNP) markers representing the most abundant sources of variation in the genome are being increasingly utilized [25]. In terms of genotyping methods, double digest restriction site-associated DNA sequencing (ddRAD-Seq) is a way for de novo SNP discovery and genotyping in non-model species, even for polyploid plants with high heterozygosity and without reference genomes, that is efficient and cost-saving [26]. In addition, due to this complexity, researchers often use a double pseudo-testcross mapping strategy to construct genetic maps. ddRAD-Seq has already been applied in high-density genetic linkage map construction and QTL mapping in many oil plants, such as peanut [27,28], Brassica napus [29], oil palm [30], and sesame [31]. Based on this, nine QTLs for leafspot resistance in peanut were identified and the phenotypic variance explained (PVE) ranged from 11% to 24%; moreover, 30 QTLs for yield-related traits were identified in which three major QTLs showed a PVE of over 10% in sesame. Miao et al. examined the overlapping QTLs of the seed lignin content, cellulose content, hemicellulose content, and oil content in Brassica napus, which were found to have a PVE ranging from 9.87 to 48.50% [14]. These results provided enriched QTLs and potential candidates for oil crops yield and quality, as well as new information for understanding the complex genetic mechanism underlying yield formation and oil accumulation.
In this study, we constructed a high-density genetic linkage map using an F1 population consisting of 180 progenies and provided high-density molecular markers for genetic breeding. The variances of eleven fruit- and oil-related traits were analyzed based on three years’ worth of data of this F1 population. We also performed QTL mapping of these eleven traits. A subset of major QTLs involved in fruit development and oil biosynthesis were obtained. The potential candidate genes in QTL regions were identified based on the collinearity of the linkage map and the C. oleifera “CON” genome [18]. This work reveals the pivotal genetic basis underlying fruit development and seed oil biosynthesis and provides insights into genetic improvements in tree breeding.

2. Results

2.1. Variation and Correlation Analysis of Fruit- and Oil-Related Traits

Eleven fruit- and oil-related traits of the mapping population were measured for three consecutive years. As expected, most traits had abundance variance and followed an approximately normal distribution (Figure 1 and Figure S1). These traits were suitable for QTL mapping. For example, the fruit yield per plant (FY) ranged from 0 to 11.34 kg with a mean of 2.62 kg in 2017 and the ratio of dry seed to fresh fruit (RSF) was from 15.43% to 36.36% with a mean value of 26.04%. The oil content of dry kernel (OC) ranged from 16.1% to 52.17% (mean 45.04%). The palmitic acid (C16:0) and oleic acid (C18:1) content in oil, which accounted for over 90% of all fatty acids in Camellia oil, ranged from 6.91% to 9.84% (mean 8.39%) and from 75.20% to 84.12% (mean 80.70%), respectively (Table S1). The variances of eleven traits in 2015 and 2016 resembled the scenario in 2017 (Table S1, Figures S1 and S2).
The analysis of variance (ANOVA) showed that the eleven traits differed significantly among the years (Table 1 and Figure 2). C. oleifera fruit yield has the major or minor year phenomenon and it is a major year in 2017. The average FY in 2017 was 2.62 kg, significantly higher than that of the other years. The ratios of dry kernel to dry seed (RKS) and C16:0 and linoleic acid content (C18:2) were significantly higher, and the C18:1 was lower in 2017 than in 2015 and 2016, respectively. A Pearson’s correlation analysis of the eleven traits was performed using the mean values of three years. The fruit- and oil-related traits showed significant correlation in the mapping population (Figure 3). RKS1 and RKS2 were significantly positively correlated (r = 0.965, p < 0.01), and both were positively correlated with OC with r = 0.406 and 0.497 (p < 0.01), respectively. In addition, RKS1 and RKS2 were both positively correlated with C16:0 (p < 0.01). The RSF was significantly negatively correlated with C18:0 (p < 0.01). The six kinds of fatty acid content detected in this study showed significant correlations with each other. According to the correlation, the six fatty acid content traits could be divided into two groups. One group included C16:0, C18:2, and linolenic acid (C18:3), and the other group consisted of stearic acid (C18:0), C18:1, and cis-11-eicosenoic acid (C20:1). Significant positive intra-group pairwise correlations and negative inter-group pairwise correlations were observed (Figure 3). Similar to the previous study results in the association population [6], C18:1 and C18:2 showed significant negative correlations (r = 0.903), while C16:0 and C18:2 were positively correlated (r = 0.567) in this linkage population (Figure 3).

2.2. Analysis of ddRAD Data and SNP Markers

In total, 182 ddRAD libraries (including 180 F1 individuals and two parents) were sequenced, which generated approximately 657.84 Gb data [18] with a Q30 of over 94.56%. 4.41 Gb and 5.42 Gb of data were generated for the female and male parents, respectively, with an average 3.60 Gb for every F1 individual. The GC contents were ranged from 38.31% to 47.96%, with a mean value of 42.40%. The clean reads were aligned with the reference genome, and over 90% reads could be mapped to the reference genome in 172 samples. The average mapping rate of the reads reached 96.10% for all samples, and the average sequencing depth of the mapped reads reached 10.39X (Table S2).
Next, 1,371,271 and 2,313,900 SNPs were identified in the female and male parents with a heterozygosis rate of 43.20% and 48.35%, respectively, using the C. oleifera “CON” genome as a reference [18]. The numbers of SNPs in F1 individuals ranged from 240,081 to 2,583,535 with an average of 1,272,218 (Table S3). In all SNPs, the nucleotide transition was significantly more than transversion. The rates of nucleotide transition to transversion (Ti/Tv) were 3.43 and 3.33 in two parents and from 2.87 to 3.75 in F1 individuals, respectively (Table S3). It was found that the base substitution types of C→T and G→A were the most common types, accounting for 22.23% and 22.22% of all SNPs, respectively (Figure 4A). A→G and T→C were the second common types, accounting for 14.68% of all SNPs, respectively. The remaining eight substitution types were relatively few, only accounting for 2.46–4.33% of all SNPs (Figure 4A). The SNP data were further filtered, and a total of 253,700 SNPs were identified with eight kinds of segregation patterns (Figure 4B). In addition, 250,715 SNP markers with the segregation patterns of lm × ll, nn × np, hk × hk, and ef × eg were used for the genetic map construction based on the double pseudo-testcross strategy. The details of all the SNPs used for genetic map construction can be found in S1 Data by Lin et al. [18].

2.3. Genetic Map Construction and Its Basic Characteristics

The SNPs were further filtered for <70% integrity using a Chi-squared test with a threshold of p < 0.01. As a result, a total of 5772 high-quality markers, including 2871 lm × ll, 2371 nn × np, 527 hk × hk, and three ef × eg, were ultimately obtained for map construction (genotype data are listed in Table S4). These SNPs were distributed into 15 chromosomes according to their physical locations on the C. oleifera “CON” genome. Then, the marker order and genetic distances of the SNP markers were evaluated. The final genetic map consisted of 15 LGs with 2780 markers and covered a total of 3327.021 cM with an average interlocus distance of 1.20 cM (Figure 5). Information of all the markers on the map can be found in Table S5, including marker IDs, genetic position (cM), chromosomes, and physical location (bp) in the “CON” genome. In all the markers on the map, there were significant differences (p < 0.01) among the SNP numbers of different segregation patterns. The lm × ll and nn × np markers were the most common and accounted for 47.45% and 41.40% of all markers, respectively. There were only two ef × eg markers (Figure 6A). The ratio of SNP numbers with different segregation patterns in each LG was analyzed and found to be similar to that for the whole genetic map (Figure 6B).
On average, each LG contained 185 markers that spanned an average length of 221.80 cM (Table 2). The genetic length of 15 LGs ranged from 196.04 cM (LG13) to 319.60 cM (LG07). The LGs were numbered according to the chromosome numbers of the “CON” reference genome. Among the 15 LGs, LG11 was the most saturated, containing 262 markers with an average marker density of 0.86 cM. LG04 contained the largest intervals of 40.67 cM between adjacent markers. The longest linkage group, LG07, harbored 258 markers, covering a length of 319.60 cM with an 1.24 cM average interlocus distance. The shortest linkage group, LG13, contained only 171 markers, spanning a length of 196.04 cM, with an average interlocus distance of 1.15 cM (Table 2).

2.4. QTL Analysis of Fruit- and Oil-Related Traits Based on the High-Density Genetic Map

Based on the high-density genetic map and phenotypic data of three consecutive years, the QTLs for eleven fruit- and oil-related traits were identified, and the proportions of the QTLs explaining the phenotypic variance were calculated. In this study, overlapped QTLs or adjacent QTLs with a distance of less than 5 cM were classified into the same loci. Based on this rule, a total of 221 QTLs were identified for the eleven traits of C. oleifera, which were distributed among 15 linkage groups, with 10–32 for each phenotypic trait (Table 3 and Table S6). In addition, 670 markers were associated with all the QTLs (Table S6), and 17 major QTLs were found for the eleven traits with a PVE ≥13.0% and LOD ≥4.00.

2.4.1. Fruit-Related Traits

The QTLs for four fruit-related traits, FY, RSF, RKS1, and RKS2, were analyzed. In total, 23 QTLs were identified for FY in all LGs except for LG08, 12, and 14, with an LOD from 3.00 to 4.07 and a PVE from 7.4% to 9.9%, respectively (Table S6). There were 69 markers that were closely linked to these QTLs.
For the RSF, we found 32 QTLs in LG02 (23.95–24.46 cM, 112.21–113.10 cM), LG03 (131.77–132.34 cM, 184.77–195.40 cM), LG04 (31.57–75.40 cM), LG06 (68.91–70.79 cM, 184.01–186.42 cM), LG07 (119.02–163.38 cM), LG08 (97.76–167.49 cM), LG09 (75.76–89.30 cM, 191.98–211.33 cM), LG10 (63.51–64.59 cM, 138.77–143.93 cM), LG11 (104.21–210.15 cM), LG12 (80.44–81.16 cM), LG13 (54.74–149.15 cM), and LG14 (66.61–119.86 cM), which explained 8.20~14.00% of the observed RSF variation. There were 84 markers closely linked to these QTLs (Table S6). Two major QTLs for RSF were detected in LG03 (184.77–195.40 cM) and LG11 (150.45–151.33 cM), which explained 13.90% and 14.00% of the observed RSF phenotypic variation, and their LODs were 4.07 and 4.1, respectively (Figure 7A,B). Their flanking molecular markers were Chr03_112839138 and Chr11_81012082.
For RKS, a total of 46 QTLs were detected: 24 QTLs for RKS1 and 22 for RKS2. Additionally, 13 QTLs overlapped because of the high correlation of the RKS1 and RKS2 phenotypic data. These QTLs were located in LG01 (78.14–79.49 cM, 141.77–143.23 cM), LG02 (73.23–73.96 cM, 129.13–156.20 cM), LG03 (71.01–72.66 cM), LG04 (74.28–75.40 cM, 135.72–138.50 cM), LG05 (54.24–68.66 cM), LG06 (50.72–83.73 cM, 122.18–122.80 cM), LG07 (66.08–68.19 cM, 149.59–150.46 cM), LG08 (4.88–21.04 cM, 62.70–70.61 cM, 131.37–132.79 cM), LG09 (138.33–140.44 cM), LG10 (5.18–25.18 cM, 143.93–156.49 cM), LG11 (113.78–116.10 cM, 160.84–167.34 cM), LG13 (29.64–32.13 cM, 94.16–107.04 cM, 166.78–168.64 cM), LG14 (75.34–81.06 cM, 137.72–139.27 cM), and LG15 (38.19–42.83 cM, 188.96–206.47 cM) with LODs from 3.00 to 5.57 and the PVEs of 7.50~16.60% (Figure 8 and Table S6). There were 95 markers that were closely linked to these QTLs.
Further, we found five major QTLs for RKS in LG01 (78.14–79.49 cM), LG03 (71.01–72.66 cM), LG04 (135.72–138.50 cM), LG06 (50.72–64.53 cM), and LG07 (149.59–150.46 cM). These QTLs exhibited LODs from 4.13 to 5.57 and explained 13.80–16.60% of the observed RKS phenotypic variation, respectively (Figure 7C–G). Nine flanking SNPs were identified, including Chr01_39071315, Chr03_108338909, Chr04_77743434, Chr06_142599472 (Chr06_112292496, Chr06_88890750, Chr06_25329856 and Chr06_111554923), and Chr07_133837987.
In conclusion, a total of 101 QTLs controlling fruit-related traits and seven major QTLs were detected (Table S6).

2.4.2. Oil-Related Traits

The QTLs for OC and six kinds of fatty acid contents were mined. Thirteen QTLs for OC were detected, of which three were located in LG01 and LG13, with two genetic intervals of 12.65 cM and 63.74 cM in LG01 and 20.46 cM and 28.63 cM in LG13; two in LG11, with a genetic interval of 41.63 cM; and one each in LG02, LG03, LG10, LG14, and LG15 (Figure 8). The LODs of these QTLs were from 3.03 to 4.37, and the PVEs were 7.50–13.30% (Table S6). There were 43 SNP markers closely linked to these QTLs. Two major QTLs for OC located in LG01 (78.141–79.493 cM) and LG14 (103.179–109.183 cM) were found with LODs of 4.28 and 4.37 (Figure 7H,I). These QTLs explained 14.5% and 13.3% of the observed OC variation, respectively. The flanking SNP markers of these two QTLs were Chr01_39071315 and Chr14_62200937 (Chr14_49268251).
Twenty-six QTLs for C16:0 in LG01 (78.035–78.844 cM, 91.466–92.726 cM), LG02 (58.745–60.291 cM, 162.07–163.893 cM), LG03 (34.404–36.251 cM, 65.24–66.66 cM), LG04 (77.03–78.179 cM, 107.093–108.543 cM, 128.952–131.849 cM), LG06 (20.548–28.486 cM), LG07 (92.854–94.457 cM, 130.315–132.196 cM, 141.16–142.925 cM, 173.807–176.519 cM), LG08 (69.597–82.121 cM), LG10 (110.824–111.642 cM, 165.323–167.372 cM), LG11 (74.883–76.89 cM, 95.209–96.752 cM, 164.001–169.537 cM, 189.399–191.663 cM), LG12 (189.484–199.058 cM), LG13 (103.325–104.692 cM), LG14 (96.397–98.324 cM), and LG15 (107.254–111.151 cM, 166.456–171.564 cM) were identified with a PVE of C16:0 from 7.90% to 14.60% (Figure 8 and Table S6). There were 87 markers closely linked to these QTLs. Two major QTLs located in LG07 and LG11 were identified with LODs of 4.31 and 4.58 and PVEs of 14.6% and 13.9%, respectively (Figure 7J,K). Chr07_96522586 and Chr11_64280487 were the flanking SNPs of these two QTLs.
Ten QTLs for C18:1 were identified in LG05 (104.487–104.896 cM), LG07 (130.315–132.196 cM), LG08 (69.597–72.195 cM), LG10 (94.045–94.404 cM, 110.824–112.279 cM), LG11 (192.03–193.251 cM), LG12 (189.484–199.058 cM), and LG15 (38.187–39.621 cM, 67.133–69.77 cM, 107.254–111.151 cM) (Figure 8). The LODs ranged from 3.00 to 4.30, and these QTLs explained 10.40–14.50% of the C18:1 content variation (Table S6). There were 32 markers that were closely linked to these QTLs. Only one major QTL located in LG12 was identified with an LOD of 4.30, and it explained 14.50% of the observed C18:1 variation (Figure 7L). The flanking SNP of this QTL was Chr12_74070604.
Moreover, 13, 14, 28, and 16 QTLs for C18:0, C18:2, C18:3, and C20:1 were also detected in the whole genetic map, respectively (Table S6). The markers linked to these QTLs ranged from 41 to 94 for these four oil-related traits. The LODs were from 3.01 to 4.57, and these QTLs could explain 7.70–15.4% of the corresponding phenotypic variations. A total of five major QTLs had LOD values over 4.0 and PVE ≥ 13.0%, which included two for C18:3 and three for C20:1, respectively (Figure 7M–Q). The major QTL for C20:1 located in LG15 had an LOD of 4.57 and explained 15.4% of the observed C20:1 variation, which was the QTL with the highest LOD in this group of QTLs.
In conclusion, a total of 120 QTLs controlling oil-related traits and 10 major QTLs were detected. Of all the QTLs, nine could be detected simultaneously using at least two years’ worth of trait data, and 48 markers were associated with these OTLs (Table 4 and Table S6). The nine QTLs were composed of two for RKS2, RKS1, and C18:2, and one for C18:3, C16:0, and C18:0, and with a 7.5–16.6% PVE for the corresponding traits. The PVEs showed significant differences among the three years, and the PVEs detected based on 2017 phenotypic data were significantly lower than those based on the 2015 and 2016 data. The lengths of these QTLs ranged from 1.159 cM (C18:2-q5-1) to 13.81 cM (RKS2-q6-1), and 25 SNPs were located in these QTL regions. Of these 25 markers, four loci had three genotypes in the linkage population and the parents were both heterozygotes. The other loci had two genotypes in the linkage population; one of parents was a heterozygote and the other was a homozygote. The mean values of the phenotype differed somewhat between the different genotype groups (Table 4). Furthermore, the dominant and additive effects of the four SNPs, with three genotypes in the linkage population, were calculated (Table 5). It was shown that the modes of gene action were different based on the data from different years. The marker–trait pair Chr06_25329856-RKS2 acted with the mode of under- or overdominance (|d/a| > 1.25) in 2015 and 2016 and with the mode of dominance (0.50 < |d/a| < 1.25) in 2017, which resembled the scenario for the Chr05_93231328-C18:2 pair. Chr06_25329856-RKS1 acted with the mode of dominance in 2015, under- or overdominance in 2016, and codominance (additive) gene action (|d/a| < 0.50) in 2017, respectively. The Chr02_98558620-C18:0 pair acted with the mode of under- or overdominance (|d/a| = 2.33) in 2015 and dominance in 2016 and 2017 (|d/a| = 0.86, 0.50), respectively (Table 5).

2.5. Identification of Fruit- and Oil-Related Candidate Genes and Potential Regulatory Network

Based on the collinearity relationship between the genetic linkage map and the C. oleifera “CON” reference genome, 89 and 113 potential candidate genes in C. oleifera were identified as being involved in fruit development and oil biosynthesis, respectively (Table S7). Some noteworthy potential fruit-related candidates were identified in the regions of FY-, RSF- and RKS-QTLs. For instance, Chr11_gene_1254.12 and Chr15_gene_52.4 corresponding to CoERF053 and CoERF061, the transcription factors that regulate vegetable oil biosynthesis [32,33,34,35], were located in the confidence interval (CI) region of RKS1-q11-2 and FY-q15-2 QTLs, respectively. Chr05_gene_78.2 corresponding to CoSAUR67, the transcription factor that regulates fruit development and maturation [36], was located in the CI region of FY-q5-1. Chr10_gene_428, which is homologous to PLC2 (phosphoinositide phospholipase C2), an enzyme involved in TAG assembly in the endoplasmic reticulum, was mapped to the region of RKS1-q10-2.
A number of potential oil-related candidates were identified in the regions of OC-, C16:0-, C18:0-, C18:1-, C18:2-, C18:3-, and C20:1-QTLs. Chr15_gene_748.45, which was homologous to WRI1 (ethylene-responsive transcription factor), a transcriptional activator involved in the activation of a subset of sugar-responsive genes and the control of carbon flow from sucrose import to oil accumulation in developing seeds [37], was mapped to the overlapping region of four QTLs: C16:0-q15-2, C18:0-q15, C18:1-q15-3, and C18:3-q15-1. Chr02_gene_151.2, corresponding to CoARF2B, was located in the CI region of C18:0-q2. Chr06_gene_1017.3, corresponding to CoARF23, was located in the overlapping region of C16:0-q6 and C18:2-q6. CoARF2B and CoARF23 were the transcription factors belonging to the ARF family and were integral to the regulation of fruit set [38]. Chr02_gene_806.11, corresponding to CoSAD, a desaturase involved in the de novo biosynthesis process of free fatty acids in plastid, was located in the region of C18:2-q2-2. Chr03_gene_1405.12, corresponding to CoDOF3.6, the transcription factor that may augment the lipid content of plant seeds by upregulating genes associated with the biosynthesis of fatty acids [39], was located in the CI region of C16:0-q3-2.
To further dissect the fruit development and oil accumulation mechanism in C. oleifera, an interaction network of the potential candidate genes in QTL regions was constructed based on Arabidopsis thaliana orthologs (Figure 9). The whole network, including 34 enzymes and 15 transcription factors, was incorporated into 49 nodes and 114 edges. A number of important genes and regulatory factors involved in growth and development; floral development and gametogenesis; and the accumulation of oil, cellulose, hemicellulose, and lignin in seeds were observed, such as WAK2, BT1, BCCP1, CSE, LAC14, and WRI1 (Figure 9). WAK2 might interact with MIK2 or RLK5 to regulate the pollen tube perception of the female signal and floral organ abscission. WAK2 might also interact with CSI1 to function in primary cell wall biosynthesis and cellulose microfibril organization. Four transcription factors, ZAT5, IDD7, ARF23, and CHR28, might associate with BT1 to participate in the regulation of bloom and gametophyte development. WRI1 might interact with BCCP1 to control the carbon flow from sucrose import to oil accumulation in the developing seeds. CSE might interact with LAC14 to function in the biosynthesis and degradation of lignin. CSE might also interact with BCCP1 to regulate oil accumulation.

3. Discussion

3.1. Mapping Population and Phenotypic Variation

The selection of mapping population is important for constructing a high-density map and further QTL analyses. Recombinant inbred line (RIL) or F2 populations are commonly used to construct genetic maps for most crops and vegetables [28,31,40,41,42]. The challenges in hybrid population creation and conservation have hindered the progress of tree genetic population construction. F1 populations are usually used to construct genetic maps for trees that have high heterozygosity, a long generation cycle, and from which it is difficult to obtain pure lines [43,44,45,46,47]. In this study, we used the F1 population, derived from a cross between CL53 and CL81, for constructing a high-density genetic map and subsequent QTL mapping of fruit- and oil-related traits. In previous studies, this population was used to construct the first linkage map of C. oleifera containing 300 SNP markers [18]. Fruit yield, oil content in the kernel, and the fatty acid component were the important determinants of the economic and meal value of C. oleifera. The improvement in oil yield and quality, including fruit yield, the rate of seed to fruit, kernel to seed, oil content in the kernel, and main fatty acid component, can enhance the economic and meal value [6,18,48]. Variances in the yield- and oil-related traits were abundant in the various populations of C. oleifera [6,18,48]. In this F1 population, FY, RSF, RKS, OC, and six kinds of fatty acid contents all showed abundant phenotypic variation among individuals and years (Figure 1 and Figure 2). RKS was positively correlated with OC, C16:0, and C18:2 and negatively correlated with C18:1. These results were identical to previous reports noting that C18:2 had a positive correlation with C16:0 and a negative correlation with C18:1 in the C. oleifera natural population [6,18,49]. It is noteworthy that C16:0 was highly positively correlated with C18:2 (r = 0.567, p < 0.01) (Figure 3).

3.2. Density and Accuracy of the C. oleifera Genetic Map

The construction of genetic linkage maps and the identification of important trait-related QTLs will facilitate future genetic and breeding studies in C. oleifera. The number and distribution of markers in the genome determine the uniformity and coverage of the markers in the genetic map [9,50]. It has been demonstrated that increasing marker density can improve the resolution of genetic maps and QTL in a given mapping population [51,52]. However, the existence of linkage disequilibrium (LD) implies a limitation on the effectiveness of increasing marker density to improve the resolution of genetic maps. A suitable marker density in the genetic map could be theoretically saturated [53]. To date, only one C. oleifera genetic map has been constructed, but the marker density is low (300 markers), with an average interlocus distance of 6.46 cM [18]. No high-density genetic map for any oil-tea Camellia species had been generated, thus preventing the use of QTLs from C. oleifera in breeding programs. ddRAD is a rapid, efficient, and cost-effective strategy for SNP development, genetic linkage map construction, marker-based complex trait selection, and draft genome assembly in many species with or without reference genomes [25,28,46,54,55,56,57,58]. In the present study, we generated 657.84 Gb of raw data using the ddRAD-seq technique, with an average of 3.60 Gb for each F1 individual. Here, 1,371,271 and 2,313,900 SNPs were detected from the two parents. The genetic map was divided into 15 LGs and comprised 2780 SNP markers, which covered 3327.021 cM with an average marker interval of 1.20 cM. This represented the most saturated linkage map of C. oleifera to date. The high-density SNP-based linkage map provides detailed information on genomic structure and is a valuable resource for fine-scale QTL mapping, MAS, and genomic studies, which should contribute to the genetic improvement of C. oleifera.

3.3. QTL Analysis of Economic Traits in C. oleifera

Fruit yield, oil content in the kernel, and fatty acid component are complex quantitative traits controlled by multiple genes and environmental factors. We have some understanding of the effect of individual locus on the economic traits of C. oleifera and their position in the genome [6,18]. However, these loci discovered in C. oleifera natural populations explained only 1.87–17.93% of the variation in the OC and fatty acid component [6,18]. QTL mapping is an efficient tool to screen candidate genes that control quantitative traits. However, few reports have identified the loci that control C. oleifera traits using QTL analysis. In our QTL mapping analysis, we developed 221 QTLs from eleven economic traits, which explained 7.40–16.60% of the phenotypic variation, respectively (Table 3 and Table S6). We also found that 101 genomic loci distributed on all 15 LGs were associated with fruit-related traits and 120 loci were associated with oil-related traits (Table 3 and Table S6). These regions may contain the primary candidate genes that control the variation in fruit- and oil-related traits.
Some chromosome segments may be associated with different traits for different growth periods and environmental factors, and QTLs can be detected throughout various stages of plant development [59,60,61]. Some QTLs are conditional and are only found in specific growth stages or environments [9,62,63,64]. The RKS1-q6-1 located in the interval 54.07–64.53 cM on LG06 appeared to be unconditional and was detected in all three years’ measurements in this study (Table S6). The other eight QTLs were reproduced in two years, including RKS2-q4, RKS2-q6-1, RKS1-q4-2, C18:3-q11-2, C18:2-q5-1, C18:2-q11-1, C18:0-q2, and C16:0-q11-3. The remaining 95.93% of QTLs were only detected in one year (Table S6). In this study, the 26 QTLs associated with C16:0 content were distributed on all LGs except LG05 and LG09 (Figure 8 and Table S6). Ten QTLs for C18:1 content were detected on the seven LGs, including LG05, LG07, LG08, LG10, LG11, LG12, and LG15 (Figure 8 and Table S6). It is worth noting that five out of the ten QTLs affecting C18:1 content, located in the interval 130.32–132.20 cM on LG07, 69.60–72.20 cM on LG08, 110.82–112.28 cM on LG10, 189.48–119.06 cM on LG12, and 107.25–111.15 cM on LG15, were wholly or partially overlapped with the QTLs for C16:0 content (Figure 8). A possible explanation for the many overlapping QTLs for different traits is that these economic traits are regulated by the interactions of many genes that may act in the same metabolic pathway.

3.4. Identification and Analysis of the Candidate Genes

A number of potential genes involved in oil metabolism in C. oleifera have been reported. Lin et al. [18] identified 21 high-confidence candidate genes, including 14 genes associated with oil accumulation and seven involved in fatty acid content. These candidate genes are located on almost all chromosomes, except Chr01, Chr04, Chr09, and Chr14 in C. oleifera. It is worth noting that these candidate genes include not only genes involved in the lipid metabolism pathway, such as malonyl-CoA:ACP transacylase (FabD), ketoacyl-ACP synthase II and III (KASII and KASIII), stearoyl-ACP desaturase (SAD), fatty acid desaturase 2 (fad2), Oleosin 3, and PLC6, but also phytohormone-related transcription factors, for example, auxin-responsive protein (IAA14 and IAA26), auxin response factor (ARF11), and ethylene response factor (ERF4 and ERF5) [6,18,49,65,66,67,68,69]. In the present study, a total of 202 candidate genes involved in fruit development and lipid metabolism were identified (Table S7). Similarly, some functional genes related to fatty acid biosynthesis and some phytohormone-related transcription factors were detected. For example, the transcription factors CoERF053 and CoERF061, which regulate the biosynthesis of vegetable oil [32,33,34,35], were identified as being associated with the RKS and FY in C. oleifera for the first time. CoSAUR67, which governs fruit development and maturation [36], was identified in the FY QTLs in this study. The interaction network analysis showed that a number of important candidate genes and regulatory factors interacted with each other and might play a potential role in fruit development and lipid biosynthesis in C. oleifera.

4. Materials and Methods

4.1. Plant Materials

C. oleifera clone CL53 is a widely cultivated variety in China, with high fruit yield and moderate OC (mean 36.70%) and C18:1 (mean 79.46%). Clone CL81 normally has a lower fruit yield but higher OC (mean 44.53%) and C18:1 (mean 82.02%). A controlled crossing was performed using “CL81” and “CL53” as male and female parents in the autumn of 2009 at Dongfanghong Forest Farm of Zhejiang Province, Jinhua City, China. At the end of October 2010, 220 seeds were harvested and sown in the middle of March 2011. In 2013, two-year-old seedlings were raised from nursery and planted with a row spacing of 3 m, a plant spacing of 2 m, and a randomized complete block design with three replications. We selected two parents and 180 healthy F1 individuals from this population for genotyping and mapping analyses.

4.2. Phenotypic Data Collection

Oil yield and quality, which depend on fatty acid composition, are the important breeding aims of C. oleifera. In this study, phenotypic data related to oil yield and quality were collected when the seeds of the F1 population and two parents matured in the autumns of 2015, 2016, and 2017, respectively. A total of eleven key traits were measured using the following methods: (1) FY (kg), the weight of all fresh fruits in each individual; (2) RSF (%), the average percent of absolute dry seed weight to fresh fruit weight in 30 fruits for one tree; (3) RKS (%), the average percent of absolute dry kernel weight (RKS1) or air-dried kernel weight (RKS2) to absolute dry seed weight in 30 fruits for one tree; (4) OC (%), the average percent of oil weight to absolute dry kernel weight in 30 fruits for one tree. The Soxtec extraction method was performed to measure the OC as described [19]; (5) C16:0 (%), C18:0 (%), C18:1 (%), C18:2 (%), C18:3 (%), and C20:1 (%) contents of C. oleifera oil, the average weight percent of every kind of fatty acid to all fatty acids in 30 fruits for one tree. The total lipid was extracted from the kernels using petroleum ether, and six kinds of fatty acid components were quantified using gas chromatography according to the previous study [6,70,71].
The normal fitting of the phenotypic data, the ANOVA, and the Pearson’s correlation coefficients (r) for the eleven traits were calculated using the Data Processing System (DPS v14.50) [72] and the pairs function in R (https://www.r-project.org/) (accessed on 1 January 2023).

4.3. DNA Extraction

Tender, healthy leaves were harvested from the 180 F1 individuals and their parents, and DNA was extracted using the TaKaRa MiniBEST Plant Genomic DNA Extraction Kit (TaKaRa, Dalian, China) according to the user manual. The DNA was quantified with an ND-2000 spectrophotometer (NanoDrop, Wilmington, DE, USA) and by electrophoresis in 1.2% agarose gels.

4.4. ddRAD Libraries Sequencing and SNPs Identification

The ddRAD library preparation and sequencing, SNP identification, and genotyping were performed by Lin et al. [18]. After filtering, only SNPs with the segregation patterns of lm × ll, nn × np, hk × hk, and ef × eg were used for the genetic map construction based on the double pseudo-testcross strategy [73]. These four kinds of markers were matched with heterozygous (female) × homozygous/absent (male), homozygous/absent (female) × heterozygous (male), heterozygous (female) × heterozygous (male), and heterozygous/absent (female) × heterozygous/absent (male) models in the double pseudo-testcross strategy, respectively.

4.5. Genetic Map Construction

First, the SNP markers used for genetic map construction were further filtered. The significantly distorted segregation markers with p-values less than 0.01 in the chi-square test and a marker integrity < 30% were excluded. Lin et al. constructed the first genetic map with a poor marker density using the “CON” genome as a reference genome [18]. In this study, the genetic map was constructed by a strategy without a reference genome. The marker order and genetic distances in each LG were calculated using JoinMap4.1 [74]. The regression mapping algorithm and Kosambi’s mapping function were used for map construction with the parameters of recombination frequencies ≤ 0.4, logarithm of odds (LODs) ≥ 3.0 and Jump = 5 [75]. A graphic genetic map was generated using a custom perl script (https://github.com/Niuyongchao/Fish_linkage_map, accessed on 20 December 2022).

4.6. QTL Mapping and Genetic Mode Analysis of SNPs

The inclusive composite interval mapping (ICIM) program of QTL Icimapping v4.2 (https://isbreeding.caas.cn/rj/qtllcmapping/294445.htm, accessed on 4 January 2023) [76] was used to detect the QTLs for oil-related traits using the C. oleifera F1 population. Two steps were included in the ICIM. In the first step, stepwise regression was applied to identify the most significant regression variables. In the second step, a one-dimensional scanning or interval mapping was conducted for detecting additive (and dominance) QTLs and a two-dimensional scanning was conducted for detecting digenic epistasis. A window size of 10 cM and a walk speed of 1 cM were chosen for ICIM analysis. The significance levels of the input and output regression equations (PINs) were 0.002, and the LOD threshold was 3.0. QTL scanning was performed for each trait using three consecutive years of phenotypic data, respectively. For each QTL, the position corresponding to the maximum LOD and the part of the phenotypic variation it explained was estimated. The QTLs with PVE ≥ 13.0% and LOD ≥ 4.00 were considered major QTLs.
The ratios of dominance (d) to additive (a) effect were calculated for the significant markers to quantify the mode of gene action according to previous research [77,78].

4.7. Identification of Fruit- and Oil-Related Candidate Genes and Potential Regulatory Networks

The QTLs for fruit- and oil-related traits obtained in this research were aligned into the physical map of the “CON” reference genome [18]. The closely linked markers of QTLs were used to locate the genomic position. The genes within QTL regions that had SNPs in introns, exons, or within 1 kb up- or downstream were regarded as potential candidate genes. Functional predictions of candidate genes were performed according to the annotation of the “CON” genome. In addition, The interaction networks were analyzed and constructed using String software (http://string-db.org/, accessed on 14 December 2023) and visualized by Cytoscape V3.7.1 [79]. The degree centrality were calculated by a network analyzer and mapped as the node size. All the interactions, clusters, and analyses were based on A. thaliana orthologs.

5. Conclusions

The most saturated high-density SNP genetic map for C. oleifera to date was constructed based on an F1 population with 180 progenies. The polymorphic markers were developed using the method of ddRAD-Seq. A total of 657.84 Gb of raw data were generated, with an average of 3.60 Gb for each F1 individual. The genetic map included 2780 markers on 15 LGs and covered 3327.021 cM with an average marker interval of 1.20 cM. Based on this high-density genetic map and the phenotypic data from three consecutive years, a QTL analysis of eleven fruit- and oil-related traits was performed. A total of 221 QTLs were detected with PVEs from 7.40% to 16.60% in which 17 major QTLs had PVEs > 13% and LODs ≥ 4.0. The flanking SNPs adjacent to the major QTLs were identified and could serve as potential molecular markers for the breeding of C. oleifera. A total of 202 high-confidence candidate genes involved in fruit development and lipid metabolism were detected, including the transcription factors CoSAUR67, CoERF053, CoARF23, CoDOF3.6, and CoWRI1 and the function genes CoPLC2 and CoSAD. These results will not only provide a platform for gene/QTL fine mapping, map-based gene isolation, and molecular breeding in C. oleifera but also provide a reference to help position sequence scaffolds on the physical map and assist in the process of assembling the C. oleifera genome sequence.

Supplementary Materials

The supporting information can be downloaded at https://www.mdpi.com/article/10.3390/ijms25168840/s1.

Author Contributions

Conceptualization, K.W. and P.L.; methodology, P.L.; software, J.C. and H.Z.; validation, A.W.; formal analysis, J.C.; investigation, H.Z.; resources, P.L.; data curation, A.W.; writing—original draft preparation, P.L.; writing—review and editing, K.W.; visualization, H.Z.; supervision, K.W. and P.L.; project administration, P.L.; funding acquisition, P.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Key Research and Development Program of China (2023YFD2200305) and Zhejiang Science and Technology Major Program on Agricultural New Variety Breeding (2021C02070-2).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Data is contained within the article and Supplementary Material.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Zhuang, R. Oil-Tea Camellia in China, 2nd ed.; China Forestry Publishing House: Beijing, China, 2012. [Google Scholar]
  2. Gao, D.F.; Xu, M.; Zhao, P.; Zhang, X.Y.; Wang, Y.F.; Yang, C.R.; Zhang, Y.J. Kaempferol acetylated glycosides from the seed cake of Camellia oleifera. Food Chem. 2011, 124, 432–436. [Google Scholar] [CrossRef]
  3. Gong, W.; Song, Q.; Ji, K.; Gong, S.F.; Wang, L.; Chen, L.; Zhang, J.; Yuan, D. Full-length transcriptome from Camellia oleifera seed provides insight into the transcript variants involved in oil biosynthesis. J. Agric. Food Chem. 2020, 68, 14670–14683. [Google Scholar] [CrossRef]
  4. Yu, Y.S.; Ren, S.X.; Tan, K.Y. Study on climatic regionalization and layer and belt distribution of oil-tea Camellia quality in China. J. Nat. Resour. 1999, 14, 123–127. [Google Scholar]
  5. Ohlrogge, J.B. Design of new plant products: Enginnering of fatty acid metabolism. Plant Physiol. 1994, 104, 821–826. [Google Scholar] [CrossRef]
  6. Lin, P.; Yin, H.; Yan, C.; Yao, X.; Wang, K. Association genetics identifies single nucleotide polymorphisms related to kernel oil content and quality in Camellia oleifera. J. Agric. Food Chem. 2019, 67, 2547–2562. [Google Scholar] [CrossRef]
  7. He, L.; Zhou, G.Y.; Zhang, H.Y.; Liu, J.A. Research progress on the health function of tea oil. J. Med. Plants Res. 2011, 5, 485–489. [Google Scholar]
  8. Wu, X.H.; Huang, Y.F.; Xie, Z.F. Health functions and prospective of Camellia oil. Food Sci. Technol. 2005, 8, 94–96. [Google Scholar]
  9. Zhang, R.; Yang, H.; Zhou, Z.; Shen, B.; Xiao, J.; Wang, B. A high-density genetic map of Schima superba based on its chromosomal characteristics. BMC Plant Biol. 2019, 19, 41. [Google Scholar] [CrossRef] [PubMed]
  10. Wang, L.; Wang, J.; Luo, G.; Yuan, X.; Gong, D.; Hu, L.; Wang, S.; Chen, H.; Chen, X.; Cheng, X. Construction of a high-density adzuki bean genetic map and evaluation of its utility based on a QTL analysis of seed size. J. Integr. Agric. 2021, 7, 1753–1761. [Google Scholar] [CrossRef]
  11. Cai, Z.; Cheng, Y.; Ma, Z.; Liu, X.; Ma, Q.; Xia, Q.; Zhang, G.; Mu, Y.; Nian, H. Fine-mapping of QTLs for individual and total isoflavone content in soybean (Glycine max L.) using a high-density genetic map. Theor. Appl. Genet. 2018, 131, 555–568. [Google Scholar] [CrossRef]
  12. Zhang, W.K.; Wang, Y.J.; Luo, G.Z.; Zhang, J.S.; He, C.Y.; Wu, X.L.; Gai, J.Y.; Chen, S.Y. QTL mapping of ten agronomic traits on the soybean (Glycine max L. Merr.) genetic map and their association with EST markers. Theor. Appl. Genet. 2004, 108, 1131–1139. [Google Scholar] [CrossRef]
  13. Zulkifli, Y.; Kamaruddin, K.; Singh, R.; Mustafa, S.; Marjuni, M.; Ting, N.C.; Amiruddin, M.D.; Leslie, L.E.T.; Cheng-Li, O.L.; Sritharan, K.; et al. An integrated linkage map of interspecific backcross 2 (BC2) populations reveals QTLs associated with fatty acid composition and vegetative parameters influencing compactness in oil palm. BMC Plant Biol. 2020, 20, 356. [Google Scholar]
  14. Miao, L.; Chao, H.; Chen, L.; Wang, H.; Zhao, W.; Li, B.; Zhang, L.; Li, H.; Wang, B.; Li, M. Stable and novel QTL identification and new insights into the genetic networks affecting seed fiber traits in Brassica napus. Theor. Appl. Genet. 2019, 132, 1761–1775. [Google Scholar] [CrossRef]
  15. Stein, A.; Wittkop, B.; Liu, L.; Obermeier, C.; Friedt, W.; Snowdon, R.J. Dissection of a major QTL for seed colour and fibre content in Brassica napus reveals colocalization with candidate genes for phenylpropanoid biosynthesis and flavonoid deposition. Plant Breed. 2013, 132, 382–389. [Google Scholar] [CrossRef]
  16. Wang, X.; Wang, H.; Long, Y.; Li, D.; Yin, Y. Identification of QTLs associated with oil content in a high-oil Brassica napus cultivar and construction of a high-density consensus map for QTLs comparison in B. napus. PLoS ONE 2013, 8, e80569. [Google Scholar] [CrossRef]
  17. Wang, X.; Wang, H.; Long, Y.; Liu, L.; Zhao, Y.; Tian, J.; Zhao, W.; Li, B.; Chen, L.; Chao, H.; et al. Dynamic and comparative QTL analysis for plant height in different developmental stages of Brassica napus L. Theor. Appl. Genet. 2015, 128, 1175–1192. [Google Scholar] [CrossRef] [PubMed]
  18. Lin, P.; Wang, K.; Wang, Y.; Hu, Z.; Yan, C.; Huang, H.; Ma, X.; Cao, Y.; Long, W.; Liu, W.; et al. The genome of oil-Camellia and population genomics analysis provide insights into seed oil domestication. Genome Biol. 2022, 23, 14. [Google Scholar] [CrossRef]
  19. Gong, W.; Xiao, S.; Wang, L.; Liao, Z.; Chang, Y.; Mo, W.; Hu, G.; Li, W.; Zhao, G.; Zhu, H.; et al. Chromosome-level genome of Camellia lanceoleosa provides a valuable resource for understanding genome evolution and self-incompatibility. Plant J. 2022, 110, 881–898. [Google Scholar] [CrossRef]
  20. Shen, T.; Huang, B.; Xu, M.; Zhou, P.; Ni, Z.; Gong, C.; Wen, Q.; Cao, F.; Xu, L. The reference genome of Camellia chekiangoleosa provides insights into Camellia evolution and tea oil biosynthesis. Hortic. Res. 2022, 9, uhab083. [Google Scholar] [CrossRef]
  21. Zhang, M.; Fan, J.; Ohlrogge, T.J.B. DGAT1 and PDAT1 acyltransferases have overlapping functions in Arabidopsis triacylglycerol biosynthesis and are essential for normal pollen and seed development. Plant Cell 2009, 21, 3885–3901. [Google Scholar] [CrossRef]
  22. Zhao, Y.; Cao, P.; Cui, Y.; Liu, D.; Li, J.; Zhao, Y.; Yang, S.; Zhang, B.; Zhou, R.; Sun, M. Enhanced production of seed oil with improved fatty acid composition by overexpressing NAD+-dependent glycerol-3-phosphate dehydrogenase in soybean. J. Integr. Plant Biol. 2021, 63, 18. [Google Scholar] [CrossRef] [PubMed]
  23. Kanai, M.; Mano, S.; Kondo, M.; Hayashi, M.; Nishimura, M. Extension of oil biosynthesis during the mid-phase of seed development enhances oil content in Arabidopsis seeds. Plant Biotechnol. J. 2016, 14, 1241–1250. [Google Scholar] [CrossRef] [PubMed]
  24. O’Neill, C.M.; Morgan, C.; Hattori, C.; Brennan, M.; Rosas, U.; Tschoep, H.; Deng, P.X.; Baker, D.; Wells, R.; Bancroft, I. Towards the genetic architecture of seed lipid biosynthesis and accumulation in Arabidopsis thaliana. Heredity 2012, 108, 115–123. [Google Scholar] [CrossRef] [PubMed]
  25. Kai, W.; Nomura, K.; Fujiwara, A.; Nakamura, Y.; Yasuike, M.; Ojima, N.; Masaoka, T.; Ozaki, A.; Kazeto, Y.; Gen, K. A ddRAD-based genetic map and its integration with the genome assembly of Japanese eel (Anguilla japonica) provides insights into genome evolution after the teleost-specific genome duplication. BMC Genom. 2014, 15, 233. [Google Scholar] [CrossRef] [PubMed]
  26. Orlando, L.; Peterson, B.K.; Weber, J.N.; Kay, E.H.; Fisher, H.S.; Hoekstra, H.E. Double digest RADseq: An inexpensive method for De Novo SNP discovery and genotyping in model and non-model species. PLoS ONE 2012, 7, e37135. [Google Scholar]
  27. Liang, Y.; Baring, M.; Wang, S.; Septiningsih, E.M. Mapping QTLs for leafspot resistance in peanut using snp-based next-generation sequencing markers. Plant Breed. Biotechnol. 2017, 5, 115–122. [Google Scholar] [CrossRef]
  28. Zhou, X.; Xia, Y.; Ren, X.; Chen, Y.; Huang, L.; Huang, S.; Liao, B.; Lei, Y.; Yan, L.; Jiang, H. Construction of a SNP-based genetic linkage map in cultivated peanut based on large scale marker development using next-generation double-digest restriction-site-associated DNA sequencing (ddRADseq). BMC Genom. 2014, 15, 351. [Google Scholar] [CrossRef] [PubMed]
  29. Scheben, A.; Severn-Ellis, A.; Patel, D.; Pradhan, A.; Edwards, D. Linkage mapping and QTL analysis of flowering time using ddRAD sequencing with genotype error correction in Brassica napus. BMC Plant Biol. 2020, 20, 546. [Google Scholar] [CrossRef] [PubMed]
  30. Bai, B.; Wang, L.; Zhang, Y.; Lee, M.; Yue, G. Developing genome-wide SNPs and constructing an ultra high-density linkage map in oil palm. Sci. Rep. 2018, 8, 691. [Google Scholar]
  31. Wu, K.; Liu, H.; Yang, M.; Tao, Y.; Ma, H.; Wu, W.; Zuo, Y.; Zhao, Y. High-density genetic map construction and QTLs analysis of grain yield-related traits in Sesame (Sesamum indicum L.) based on RAD-Seq techonology. BMC Plant Biol. 2014, 14, 274. [Google Scholar] [CrossRef]
  32. Shi, M.; Yu, L.; Shi, J.; Liu, J. A Conserved MYB transcription factor is involved in regulating lipid metabolic pathways for oil biosynthesis in green algae. New Phytol. 2022, 235, 576–594. [Google Scholar] [CrossRef]
  33. Zhang, P.; Xin, Y.; He, Y.; Tang, X.; Shen, C.; Wang, Q.; Lv, N.; Li, Y.; Hu, Q.; Xu, J. Exploring a blue-light-sensing transcription factor to double the peak productivity of oil in Nannochloropsis oceanica. Nat. Commun. 2022, 13, 1664. [Google Scholar] [CrossRef]
  34. Yang, J.; Chen, B.; Manan, S.; Li, P.; Liu, C.; She, G.; Zhao, S.; Zhao, J. Critical metabolic pathways and Sad/Fads, WRI1s, and DGATs cooperate for high-oleic acid oil production in developing oil tea (Camellia oleifera) seeds. Hortic. Res. 2022, 9, uhac087. [Google Scholar] [CrossRef]
  35. Huang, R.; Liu, M.; Gong, G.; Wu, P.; Bai, M.; Qin, H.; Wang, G.; Liao, H.; Wang, X.; Li, Y.; et al. BLISTER promotes seed maturation and fatty acid biosynthesis by interacting with WRINKLED1 to regulate chromatin dynamics in Arabidopsis. Plant Cell 2022, 34, 2242–2265. [Google Scholar] [CrossRef]
  36. Fenn, M.A.; Giovannoni, J.J. Phytohormones in fruit development and maturation. Plant J. 2021, 105, 446–458. [Google Scholar] [CrossRef]
  37. Cernac, A.; Andre, C.; Hoffmann-Benning, S.; Benning, C. WRI1 is required for seed germination and seedling establishment. Plant Physiol. 2006, 141, 745–757. [Google Scholar] [CrossRef]
  38. Wang, H.; Jones, B.; Li, Z.; Frasse, P.; Delalande, C.; Regad, F.; Chaabouni, S.; Latché, A.; Pech, J.C.; Bouzayen, M. The tomato Aux/IAA transcription factor IAA9 is involved in fruit development and leaf morphogenesis. Plant Cell 2005, 17, 2676–2692. [Google Scholar] [CrossRef]
  39. Wang, H.; Zhang, B.; Hao, Y.; Huang, J.; Tian, A. The Soybean Dof-type transcription factor genes, GmDof4 and GmDof11, enhance lipid content in the seeds of transgenic Arabidopsis plants. Plant J. 2007, 52, 716–729. [Google Scholar] [CrossRef]
  40. Li, H.; Zhao, T.; Wang, Y.; Yu, D.; Chen, S.; Zhou, R.; Gai, J. Genetic structure composed of additive QTL, epistatic QTL pairs and collective unmapped minor QTL conferring oil content and fatty acid components of soybeans. Euphytica 2011, 182, 117–132. [Google Scholar] [CrossRef]
  41. Branham, S.E.; Levi, A.; Farnham, M.W.; Patrick Wechter, W. A GBS-SNP-based linkage map and quantitative trait loci (QTL) associated with resistance to Fusarium oxysporum f. sp. niveum race 2 identified in Citrullus lanatus var. citroides. Theor. Appl. Genet. 2017, 130, 319–330. [Google Scholar] [CrossRef]
  42. Emebiri, L.C.; Tan, M.K.; El-Bouhssini, M.; Wildman, O.; Ogbonnaya, F.C. QTL mapping identifies a major locus for resistance in wheat to Sunn pest (Eurygaster integriceps) feeding at the vegetative growth stage. Theor. Appl. Genet. 2016, 130, 309–318. [Google Scholar] [CrossRef] [PubMed]
  43. Pek, A.; Ylmaz, K.; Skc, P.; Tangu, N.A.; Gülen, H. SNP discovery by GBS in olive and the construction of a high-density genetic linkage map. Biochem. Genet. 2016, 54, 313–325. [Google Scholar]
  44. Ma, J.; Huang, L.; Ma, C.; Jin, J.; Li, C.; Wang, R.; Zheng, H.; Yao, M.; Chen, L. Large-scale SNP discovery and genotyping for constructing a high-density genetic map of tea plant using specific-locus amplified fragment sequencing (SLAF-seq). PLoS ONE 2015, 10, e0128798. [Google Scholar] [CrossRef]
  45. Guo, Y.; Shi, G.; Liu, Z.; Zhao, Y.; Yang, X.; Zhu, J.; Li, K.; Guo, X. Using specific length amplified fragment sequencing to construct the high-density genetic map for Vitis (Vitis vinifera L. × Vitis amurensis Rupr.). Front. Plant Sci. 2015, 6, 393. [Google Scholar] [CrossRef] [PubMed]
  46. Guo, F.; Yu, H.; Tang, Z.; Jiang, X.; Wang, L.; Wang, X.; Xu, Q.; Deng, X. Construction of a SNP-based high-density genetic map for pummelo using RAD sequencing. Tree Genet. Genomes 2015, 11, 2. [Google Scholar] [CrossRef]
  47. Zhang, Z.; Wei, T.; Zhong, Y.; Li, X.; Huang, J. Construction of a high-density genetic map of Ziziphus jujuba Mill. using genotyping by sequencing technology. Tree Genet. Genomes 2016, 12, 76. [Google Scholar] [CrossRef]
  48. Chai, J.; Wang, K.; Xie, Y.; Wang, A.; Zhong, H.; Yao, X.; Lin, P. Combining ability from complete diallel design of Camellia oleifera: Implications for the utility of GCA and SCA in oil-related traits breeding. Ind. Crops Prod. 2023, 205, 117434. [Google Scholar] [CrossRef]
  49. Ye, Z.; Wu, Y.; Ul Haq Muhammad, Z.; Yan, W.; Yu, J.; Zhang, J.; Yao, G.; Hu, X. Complementary transcriptome and proteome profiling in the mature seeds of Camellia oleifera from Hainan Island. PLoS ONE 2020, 15, e0226888. [Google Scholar] [CrossRef]
  50. Song, X.; Xu, Y.; Gao, K.; Fan, G.; Zhang, F.; Deng, C.; Dai, S.; Huang, H.; Xin, H.; Li, Y. High-density genetic map construction and identification of loci controlling flower-type traits in Chrysanthemum (Chrysanthemum × morifolium Ramat.). Hortic. Res. 2020, 7, 108. [Google Scholar] [CrossRef]
  51. Gutierrez-Gonzalez, J.J.; Vuong, T.D.; Zhong, R.; Yu, O.; Lee, J.D.; Shannon, G.; Ellersieck, M.; Nguyen, H.T.; Sleper, D.A. Major locus and other novel additive and epistatic loci involved in modulation of isoflavone concentration in soybean seeds. Theor. Appl. Genet. 2011, 123, 1375–1385. [Google Scholar] [CrossRef] [PubMed]
  52. Zou, G. Identification of QTLs for eight agronomically important traits using an ultra-high-density map based on SNPs generated from high-throughput sequencing in sorghum under contrasting photoperiods. J. Exp. Bot. 2012, 63, 5451. [Google Scholar] [CrossRef] [PubMed]
  53. Zhang, D.; Li, H.; Wang, J.; Zhang, H.; Hu, Z.; Chu, S.; Lv, H.; Yu, D. High-density genetic mapping identifies new major loci for tolerance to low-phosphorus stress in soybean. Front. Plant Sci. 2016, 7, 372. [Google Scholar] [CrossRef]
  54. Fu, X.; Dou, J.; Mao, J.; Su, H.; Jiao, W.; Zhang, L.; Hu, X.; Huang, X.; Wang, S.; Bao, Z. RAD typing: An integrated package for accurate de novo codominant and dominant RAD genotyping in mapping populations. PLoS ONE 2013, 8, e79960. [Google Scholar] [CrossRef]
  55. Amit, A.D.; Larissa, R.; Andrew, G.S.; Marwan, D.; Anoop, S.; Kirstin, B.; Thomas, D.W.; Bunyamin, T.A. Genome wide SNP identification in chickpea for use in development of a high density genetic map and improvement of chickpea reference genome assembly. BMC Genom. 2014, 15, 708. [Google Scholar]
  56. Yin, T.; Wang, Y.; Ning, Z.; Hu, Y.; Chen, J.; Zhao, R.; Chen, H.; Ai, N.; Guo, W.; Zhang, T. Molecular mapping of restriction-site associated DNA markers in allotetraploid upland cotton. PLoS ONE 2015, 10, e0124781. [Google Scholar]
  57. Liu, Z.; Zhu, H.; Liu, Y.; Kuang, J.; Zhou, K.; Liang, F.; Liu, Z.; Wang, D.; Ke, W. Construction of a high-density, high-quality genetic map of cultivated lotus (Nelumbo nucifera) using next-generation sequencing. BMC Genom. 2016, 17, 466. [Google Scholar] [CrossRef]
  58. Subudhi, P.K.; Fan, W.; Zong, J.; Luo, Z.; Chen, M.; Zhao, X.; Zhang, D.; Qi, Y.; Yuan, Z. Development of a RAD-Seq based DNA polymorphism identification software, agromarker finder, and its application in rice marker-assisted breeding. PLoS ONE 2016, 11, e0147187. [Google Scholar]
  59. Zhang, M.; Lu, N.; Zhu, T.; Yang, G.; Qu, G.; Shi, C.; Fei, Y.; Liu, B.; Ma, W.; Wang, J. A bivariate mapping model identifies major covariation QTLs for biomass allocation between leaf and stem growth of Catalpa bungei. Front. Genet. 2021, 12, 758209. [Google Scholar] [CrossRef]
  60. Zhang, M.; Lu, N.; Jiang, L.; Liu, B.; Fei, Y.; Ma, W.; Shi, C.; Wang, J.; Plomion, C. Multiple dynamic models reveal the genetic architecture for growth in height of Catalpa bungei in the field. Tree Physiol. 2022, 42, 1239–1255. [Google Scholar] [CrossRef] [PubMed]
  61. Zhang, M.; Liu, B.; Fei, Y.; Yang, X.; Zhao, L.; Shi, C.; Zhang, Y.; Lu, N.; Wu, C.; Ma, W.; et al. Genetic architecture of leaf morphology revealed by integrated trait module in Catalpa bungei. Hortic. Res. 2023, 10, uhad032. [Google Scholar] [CrossRef] [PubMed]
  62. Yang, G.; Xing, Y.; Li, S.; Ding, J.; Yue, B.; Deng, K.; Li, Y.; Zhu, Y. Molecular dissection of developmental behavior of tiller number and plant height and their relationship in rice (Oryza sativa L.). Hereditas 2006, 143, 236–245. [Google Scholar] [CrossRef]
  63. Pootakham, W.; Jomchai, N.; Ruang-areerate, P.; Shearman, J.R.; Sonthirod, C.; Sangsrakru, D.; Tragoonrung, S.; Tangphatsornruang, S. Genome-wide SNP discovery and identification of QTL associated with agronomic traits in oil palm using genotyping-by-sequencing (GBS). Genomics 2015, 105, 288–295. [Google Scholar] [CrossRef] [PubMed]
  64. Sun, Y.; Li, X.; Ma, Z.; Chen, S. Quantitative trait locus mapping of fruit aroma compounds in cucumber (Cucumber sativus L.) based on a recombinant inbred line population. Hortic. Res. 2022, 9, uhac151. [Google Scholar] [CrossRef]
  65. Song, Q.; Ji, K.; Mo, W.; Wang, L.; Chen, L.; Gao, L.; Gong, W.; Yuan, D. Dynamics of sugars, endogenous hormones, and oil content during the development of Camellia oleifera fruit. Botany 2021, 99, 515–529. [Google Scholar] [CrossRef]
  66. Zhang, F.; Li, Z.; Zhou, J.; Gu, Y.; Tan, X. Comparative study on fruit development and oil synthesis in two cultivars of Camellia oleifera. BMC Plant Biol. 2021, 21, 348. [Google Scholar] [CrossRef] [PubMed]
  67. Ye, Z.; Yu, J.; Yan, W.; Zhang, J.; Yang, D.; Yao, G.; Liu, Z.; Wu, Y.; Hou, X. Integrative iTRAQ-based proteomic and transcriptomic analysis reveals the accumulation patterns of key metabolites associated with oil quality during seed ripening of Camellia oleifera. Hortic. Res. 2021, 8, 157. [Google Scholar] [CrossRef] [PubMed]
  68. Shi, J.; Dai, X.; Chen, Y.; Chen, J.; Shi, J.; Yin, T. Discovery and experimental analysis of microsatellites in an oil woody plant Camellia chekiangoleosa. Plant Syst. Evol. 2013, 299, 1387–1393. [Google Scholar] [CrossRef]
  69. Brkić, K.; Radulović, M.; Sladonja, B.; Lukić, I.; Šetić, E. Application of Soxtec apparatus for oil content determination in olive fruit. Riv. Ital. Sostanze Grasse 2006, 83, 115–119. [Google Scholar]
  70. Lall, R.K.; Proctor, A.; Jain, V.P. A rapid, micro FAME preparation method for vegetable oil fatty acid analysis by Gas Chromatography. J. Am. Oil Chem. Soc. 2009, 86, 309–314. [Google Scholar] [CrossRef]
  71. Lin, P.; Wang, K.; Zhou, C.; Xie, Y.; Yao, X.; Yin, H. Seed transcriptomics analysis in Camellia oleifera uncovers genes associated with oil content and fatty acid composition. Int. J. Mol. Sci. 2018, 19, 118. [Google Scholar] [CrossRef]
  72. Tang, Q.; Zhang, C. Data Processing System (DPS) software with experimental design, statistical analysis and data mining developed for use in entomological research. Insect Sci. 2013, 20, 254–260. [Google Scholar] [CrossRef]
  73. Grattapaglia, D.; Sederoff, R. Genetic linkage maps of Eucalyptus grandis and Eucalyptus urophylla using a pseudo-testcross: Mapping strategy and RAPD markers. Genetics 1994, 137, 1121–1137. [Google Scholar] [CrossRef]
  74. Liu, D.; Ma, C.; Hong, W.; Huang, L.; Liu, M.; Liu, H.; Zeng, H.; Deng, D.; Xin, H.; Song, J.; et al. Construction and analysis of high-density linkage map using high-throughput sequencing data. PLoS ONE 2014, 9, e98855. [Google Scholar] [CrossRef]
  75. Kosambi, D.D. The estimation of map distance from recombination values. Ann. Eugen. 1943, 12, 172–175. [Google Scholar] [CrossRef]
  76. Wang, J.K. Inclusive composite interval mapping of quantitative trait genes. Acta Agron. Sin. 2009, 35, 239–245. [Google Scholar] [CrossRef]
  77. Wegrzyn, J.L.; Eckert, A.J.; Choi, M.; Lee, J.M.; Stanton, B.J.; Sykes, R.; Davis, M.F.; Tsai, C.J.; Neale, D.B. Association genetics of traits controlling lignin and cellulose biosynthesis in black cottonwood (Populus trichocarpa, Salicaceae) secondary xylem. New Phytol. 2010, 188, 515–532. [Google Scholar] [CrossRef]
  78. Eckert, A.J.; Bower, A.D.; Wegrzyn, J.L.; Pande, B.; Jermstad, K.D.; Krutovsky, K.V.; St. Clair, J.B.; Neale, D.B. Association genetics of coastal Douglas fir (Pseudotsuga menziesii var. menziesii, Pinaceae). I. Cold-Hardiness related traits. Genetics 2009, 182, 1289–1302. [Google Scholar] [CrossRef]
  79. Cline, M.S.; Smoot, M.; Cerami, E.; Kuchinsky, A.; Bader, G.D.; Workman, C.; Christmas, R.; Avila-Campilo, I.; Creech, M.; Gross, B.; et al. Integration of biological networks and gene expression data using Cytoscape. Nat. Protoc. 2007, 2, 2366–2382. [Google Scholar] [CrossRef] [PubMed]
Ijms 25 08840 g001
Figure 1. Phenotypic evaluation of eleven traits for the linkage population in 2017. The Y-axis is the percent of plant individuals to the linkage population. The X-axis is the continuous (A) fruit yield per plant (FY, kg), (B) the ratio of dry seed to fresh fruit (RSF, %), (C) the ratio of air-dried kernel to absolute dry seed (RKS2, %), (D) the ratio of absolute dry kernel to absolute dry seed (RKS1, %), (E) the oil content of absolute dry kernel (OC, %), (F) palmitic acid (C16:0, %), (G) stearic acid (C18:0, %), (H) oleic acid (C18:1, %), (I) linoleic acid (C18:2, %), (J) linolenic acid (C18:3, %), and (K) cis-11-eicosenoic acid (C20:1, %) contents of Camellia oil, respectively. The green line indicates the distribution of plant individuals and the red dashed line indicates the normal distribution curve.
Figure 1. Phenotypic evaluation of eleven traits for the linkage population in 2017. The Y-axis is the percent of plant individuals to the linkage population. The X-axis is the continuous (A) fruit yield per plant (FY, kg), (B) the ratio of dry seed to fresh fruit (RSF, %), (C) the ratio of air-dried kernel to absolute dry seed (RKS2, %), (D) the ratio of absolute dry kernel to absolute dry seed (RKS1, %), (E) the oil content of absolute dry kernel (OC, %), (F) palmitic acid (C16:0, %), (G) stearic acid (C18:0, %), (H) oleic acid (C18:1, %), (I) linoleic acid (C18:2, %), (J) linolenic acid (C18:3, %), and (K) cis-11-eicosenoic acid (C20:1, %) contents of Camellia oil, respectively. The green line indicates the distribution of plant individuals and the red dashed line indicates the normal distribution curve.
Ijms 25 08840 g001
Ijms 25 08840 g002
Figure 2. Phenotypic variance of eleven traits among different years. The X-axis represents the years. The Y-axis indicates the (A) FY (kg), (B) RSF (%), (C) RKS2 (%), (D) RKS1 (%), (E) OC (%), (F) C16:0 (%), (G) C18:0 (%), (H) C18:1 (%), (I) C18:2 (%), (J) C18:3 (%), and (K) C20:1 (%) content of the Camellia oil, respectively. Different lowercase mean the significant differences at p < 0.05.
Figure 2. Phenotypic variance of eleven traits among different years. The X-axis represents the years. The Y-axis indicates the (A) FY (kg), (B) RSF (%), (C) RKS2 (%), (D) RKS1 (%), (E) OC (%), (F) C16:0 (%), (G) C18:0 (%), (H) C18:1 (%), (I) C18:2 (%), (J) C18:3 (%), and (K) C20:1 (%) content of the Camellia oil, respectively. Different lowercase mean the significant differences at p < 0.05.
Ijms 25 08840 g002
Ijms 25 08840 g003
Figure 3. Pearson correlation matrix for eleven fruit- and oil-related traits of the linkage population based on the means of three years. The distribution of the means of eleven traits in three years are shown on the diagonal. To the bottom left are the bivariate scatter plots with best-fit lines displayed. Correlation coefficients are shown above the diagonal. “***”, “**”, and “*” denote significance with p-values of 0.001, 0.01, and 0.05, respectively. Red and blue denote positive and negative correlations, respectively.
Figure 3. Pearson correlation matrix for eleven fruit- and oil-related traits of the linkage population based on the means of three years. The distribution of the means of eleven traits in three years are shown on the diagonal. To the bottom left are the bivariate scatter plots with best-fit lines displayed. Correlation coefficients are shown above the diagonal. “***”, “**”, and “*” denote significance with p-values of 0.001, 0.01, and 0.05, respectively. Red and blue denote positive and negative correlations, respectively.
Ijms 25 08840 g003
Ijms 25 08840 g004
Figure 4. Statistics for variation types (A) and segregation patterns (B) of SNPs in the linkage population. (A): The X-axis indicates the variation types and the Y-axis indicates the SNP numbers for each variation type. (B): The X-axis represents the predicted genotype segregation patterns which are labeled by “female genotype × male genotype”. The Y-axis represents the SNP numbers of each segregation pattern. Only the SNPs with the segregation patterns lm × ll, nn × np, hk × hk, and ef × eg were subjected to genetic map construction using the double pseudo-testcross strategy. Lm × ll, nn × np, hk × hk, and ef × eg were matched with heterozygous (female) × homozygous/absent (male), homozygous/absent (female) × heterozygous (male), heterozygous (female) × heterozygous (male), and heterozygous/absent (female) × heterozygous/absent (male) models in the double pseudo-test-cross strategy, respectively. Details of the SNP genotypes of individuals are shown in Table S4.
Figure 4. Statistics for variation types (A) and segregation patterns (B) of SNPs in the linkage population. (A): The X-axis indicates the variation types and the Y-axis indicates the SNP numbers for each variation type. (B): The X-axis represents the predicted genotype segregation patterns which are labeled by “female genotype × male genotype”. The Y-axis represents the SNP numbers of each segregation pattern. Only the SNPs with the segregation patterns lm × ll, nn × np, hk × hk, and ef × eg were subjected to genetic map construction using the double pseudo-testcross strategy. Lm × ll, nn × np, hk × hk, and ef × eg were matched with heterozygous (female) × homozygous/absent (male), homozygous/absent (female) × heterozygous (male), heterozygous (female) × heterozygous (male), and heterozygous/absent (female) × heterozygous/absent (male) models in the double pseudo-test-cross strategy, respectively. Details of the SNP genotypes of individuals are shown in Table S4.
Ijms 25 08840 g004
Ijms 25 08840 g005
Figure 5. ddRAD based high-density genetic map of 180 full-sibs. The 3327.021 cM map included 2780 SNPs.
Figure 5. ddRAD based high-density genetic map of 180 full-sibs. The 3327.021 cM map included 2780 SNPs.
Ijms 25 08840 g005
Ijms 25 08840 g006
Figure 6. SNP type statistics for the genetic map. (A) The SNP numbers of four segregation patterns in the whole genetic map. (B) The SNP numbers of four segregation patterns in each linkage group.
Figure 6. SNP type statistics for the genetic map. (A) The SNP numbers of four segregation patterns in the whole genetic map. (B) The SNP numbers of four segregation patterns in each linkage group.
Ijms 25 08840 g006
Ijms 25 08840 g007
Figure 7. The results of the major QTL test for RSF (A,B), RKS (C–G), OC (H,I), C16:0 (J,K), C18:1 (L), C18:3 (M,N), and C20:1 (O–Q).
Figure 7. The results of the major QTL test for RSF (A,B), RKS (C–G), OC (H,I), C16:0 (J,K), C18:1 (L), C18:3 (M,N), and C20:1 (O–Q).
Ijms 25 08840 g007
Ijms 25 08840 g008
Figure 8. The results of the QTL test for the RKS, OC, C16:0, and C18:1 content. The left of each linkage group shows the location of the QTLs in units of cM. The right shows the QTL name. In this study, the QTLs were named as follows: “target traits (capital letter)” + “−” + “q” + “linkage group code” + “QTL number”. The full name of the QTL is presented in italics. The QTLs for RKS, OC, C16:0, and C18:1 are in black, red, blue, and green, respectively. RKS1 and RKS2 are the percent of absolute dry kernel weight and air-dried kernel weight to the absolute dry seed weight, respectively. RKS indicates the overlapping QTLs for RKS1 and RKS2. The locations in pink rectangles indicate the overlapping QTLs of two traits.
Figure 8. The results of the QTL test for the RKS, OC, C16:0, and C18:1 content. The left of each linkage group shows the location of the QTLs in units of cM. The right shows the QTL name. In this study, the QTLs were named as follows: “target traits (capital letter)” + “−” + “q” + “linkage group code” + “QTL number”. The full name of the QTL is presented in italics. The QTLs for RKS, OC, C16:0, and C18:1 are in black, red, blue, and green, respectively. RKS1 and RKS2 are the percent of absolute dry kernel weight and air-dried kernel weight to the absolute dry seed weight, respectively. RKS indicates the overlapping QTLs for RKS1 and RKS2. The locations in pink rectangles indicate the overlapping QTLs of two traits.
Ijms 25 08840 g008
Ijms 25 08840 g009
Figure 9. Interaction network of potential candidate genes for fruit- and oil-related traits. Network visualization of the interaction of the 49 candidate genes and regulatory factors constructed using String software and observed using Cytoscape_V3.7.1. Genes are presented as nodes, and gene interactions are presented as edges.
Figure 9. Interaction network of potential candidate genes for fruit- and oil-related traits. Network visualization of the interaction of the 49 candidate genes and regulatory factors constructed using String software and observed using Cytoscape_V3.7.1. Genes are presented as nodes, and gene interactions are presented as edges.
Ijms 25 08840 g009
Table 1. Analysis of variance for eleven fruit- and oil-related traits in the linkage population of C. oleifera.
Table 1. Analysis of variance for eleven fruit- and oil-related traits in the linkage population of C. oleifera.
Sources of VariationYearGenotypesError
df2179358
FYMS269.42732.86381.8835
F value143.049 **1.521 **
RSFMS3488.990772.387837.9708
F value91.886 **1.906 **
RKS2MS291.598427.57926.5039
F value44.834 **4.24 **
RKS1MS268.762226.01675.7166
F value47.015 **4.551 **
OCMS75.037527.484615.3854
F value4.877 **1.786 **
C16:0MS102.53230.58380.1932
F value530.75 **3.022 **
C18:0MS4.63550.14030.0551
F value84.199 **2.549 **
C18:1MS657.55754.07071.4652
F value448.777 **2.778 **
C18:2MS245.20112.63650.8135
F value301.403 **3.241 **
C18:3MS0.02640.00180.0008
F value33.577 **2.266 **
C20:1MS0.01430.0040.0015
F value9.765 **2.754 **
MS, Mean squares for phenotypes; **, p < 0.01.
Table 2. Genetic linkage group statistics for the genetic map.
Table 2. Genetic linkage group statistics for the genetic map.
LG IDTotal MarkersTotal Distance (cM)Average Distance (cM)Gaps > 5 cM(%)Max Gap (cM)
LG01231203.790.891.7420.62
LG02187220.891.195.9114.28
LG03197196.621.002.0410.63
LG04138253.021.855.8440.67
LG05192222.041.164.7126.96
LG06187196.191.052.6912.75
LG07258319.601.244.2819.77
LG08164262.681.617.9816.16
LG09118211.331.815.9819.35
LG10186205.401.112.7015.22
LG11262224.300.862.3022.44
LG12181201.071.122.2210.37
LG13171196.041.152.3519.84
LG14145207.601.445.5615.29
LG15163206.471.275.5617.51
Total27803327.021.203.8840.67
Average185415.881.203.88---
Table 3. The statistics of QTLs for fruit- and oil-related traits in C. oleifera.
Table 3. The statistics of QTLs for fruit- and oil-related traits in C. oleifera.
TraitsYearNumber of QTLsLinkage GroupPVE a (%)LOD Value
FY2015101, 2, 5, 6, 10, 137.4–9.43.0–3.87
201653, 5, 7, 97.4–9.03.02–3.7
201783, 4, 7, 11, 13, 157.6–9.93.07–4.07
RSF2015142, 4, 6, 7, 9, 10, 11, 149.3–11.43.0–3.72
201682, 3, 6, 7, 8, 1110.5–14.03.01–4.1
2017103, 9, 10, 12, 13, 148.2–11.23.06–4.25
RKS2201551, 6, 8, 10, 119.8–16.63.17–5.57
2016141, 2, 3, 4, 6, 7, 8, 9, 13, 14, 1510.4–15.13.0–4.48
201754, 5, 6, 10, 137.5–9.03.02–3.66
RKS1201552, 6, 10, 119.6–14.33.09–4.71
2016141, 2, 3, 4, 6, 7, 9, 13, 14, 1510.6–14.53.08–4.27
201784, 5, 6, 7, 8, 10, 147.5–10.03.0–4.08
OC201563, 13, 14, 169.4–13.33.03–4.37
201641, 10, 1111.5–14.53.36–4.28
201731, 2, 117.5–8.33.03–3.35
C16:0201591, 2, 7, 8, 11, 13, 159.5–13.93.07–4.58
2016111, 3, 4, 7, 10, 11, 12, 14, 1510.4–14.63.02–4.31
201773, 6, 7, 10, 117.9–11.93.17–4.89
C18:0201552, 5, 7, 139.7–12.53.13–4.09
201653, 4, 12, 1510.4–12.63.01–3.67
201742, 9, 13, 147.7–8.83.09–3.56
C18:1201545, 7, 8, 1510.1–12.23.26–3.99
2016610, 11, 12, 1510.4–14.53.0–4.3
20170
C18:2201553, 5, 7, 11, 159.6–10.83.07–3.5
201672, 6, 10, 11, 12, 1310.5–13.43.05–3.95
201745, 10, 117.5–9.83.0–3.99
C18:32015141, 3, 4, 8, 9, 10, 11, 14, 159.4–13.03.02–4.26
201693, 4, 5, 7, 11, 12, 1310.5–15.33.05–4.54
201762, 3, 5, 11, 13, 158.0–9.73.2–3.93
C20:1201551, 7, 8, 10, 139.5–10.23.05–3.31
201694, 7, 9, 10, 11, 13, 1510.5–15.43.02–4.57
201721, 117.7–8.13.09–3.27
a Percentage of phenotypic variation explained by each QTL.
Table 4. Summary of QTLs identified in at least two years and the traits in the full-sib population of “ChangLin No. 53” (CL53) × “ChangLin No. 81” (CL81).
Table 4. Summary of QTLs identified in at least two years and the traits in the full-sib population of “ChangLin No. 53” (CL53) × “ChangLin No. 81” (CL81).
TraitQTLStartEndLength (cM)YearLODPVE (%)MarkerCross TypeCL53 (♀)CL81 (♂)GenotypeIndividual NumbersPhenotype Means(%)
201520162017
RKS2RKS2-q4135.721138.5012.7820164.4815.1Chr04_77743434lm × llGTGGGG101 59.7062.05
20173.017.5 GT76 58.2460.33
-3 57.7861.94
RKS2-q6-150.72364.53413.81120155.5716.6Chr06_142599472lm × llGAGGGG9759.25 61.07
20173.669.0 GA8159.12 61.56
-162.77 62.97
Chr06_112292496lm × llGAGGGG9759.31 61.22
GA8059.33 61.58
-354.52 56.69
Chr06_88890750lm × llCTCCCC10059.57 61.35
CT7758.85 61.29
-356.83 60.14
Chr06_25329856hk × hkAGAGAA3557.95 61.19
AG9659.58 61.24
GG3059.01 61.50
-1959.93 61.57
Chr06_111554923lm × llGAGGGG9159.44 61.36
GA8658.87 61.20
-361.85 62.65
RSK1RKS1-q4-2135.721138.5012.7820164.2714.5Chr04_77743434lm × llGTGGGG101 64.7867.21
20173.288.1 GT76 64.6566.97
-3 63.4067.17
RKS1-q6-154.07264.53410.46220154.7114.3Chr06_142599472lm × llGAGGGG9765.6764.48
20163.3711.6 GA8166.7165.26
-169.2768.23
Chr06_112292496lm × llGAGGGG9766.1464.75
GA8066.2364.72
-362.7661.66
Chr06_88890750lm × llCTCCCC10066.3165.23
CT7765.9964.04
-362.9161.79
Chr06_25329856hk × hkAGAGAA3565.2463.62
AG9666.2565.11
GG3066.1964.90
-1966.7764.22
C18:3C18:3-q11-284.5285.6951.17520153.169.8Chr11_27083024nn × npCCCTCC8622.24 24.45
20173.939.7 CT9222.23 24.53
-131.47 33.96
C18:2C18:2-q11-1162.869166.1963.32720153.079.6Chr11_114235680nn × npCCCTCC986.54 7.96
20173.027.5 CT806.90 8.05
-26.23 9.21
Chr11_65073730lm × llAGAAAA786.78 8.08
AG986.59 7.91
-48.16 9.35
C18:2-q5-1104.487105.6461.15920153.510.8Chr05_107820202lm × llAGGGGG956.84 8.22
20173.07.5 AG796.55 7.76
-47.00 7.62
Chr05_93231328hk × hkTCTCTT566.75 7.88
TC676.87 8.13
CC326.37 8.15
-256.58 7.84
C16:0C16:0-q11-3164.001169.5375.53620153.7311.5Chr11_39972487lm × llAGAAAA846.87 8.37
20174.210.3 AG947.02 8.39
-26.68 8.10
Chr11_39972501lm × llAGAAAA836.85 8.35
AG957.03 8.41
-26.68 8.10
Chr11_39972457lm × llCACCCC826.85 8.35
CA967.03 8.40
-26.68 8.10
Chr11_128763300nn × npTTTATT986.88 8.33
TA797.04 8.44
-36.47 8.37
Chr11_10782110nn × npGGGAGG786.95 8.38
GA1006.95 8.38
-26.34 8.36
Chr11_64280835nn × npCCCTCC996.96 8.38
CT786.93 8.37
-36.68 8.46
Chr11_146396217lm × llGAGGGG796.91 8.37
GA1006.97 8.38
-16.84 8.44
C18:0C18:0-q278.59682.1143.51820153.139.7Chr02_116850822lm × llAGGGGG991.85 2.13
20173.448.5 GA761.87 2.17
-51.75 1.94
Chr02_98558620hk × hkAGAGGG411.79 2.11
GA861.89 2.14
AA391.85 2.15
-141.81 2.21
Table 5. List of marker effects for QTLs of traits.
Table 5. List of marker effects for QTLs of traits.
TraitSNP2a 1d 2d/aFrequency 3a4
201520162017201520162017201520162017201520162017
RKS2Chr06_253298561.060.930.311.100.95−0.112.082.05−0.680.48(G)−0.76−0.510.04
RKS1Chr06_253298560.951.280.340.530.850.041.131.330.240.48(G)−0.42−0.46−0.07
C18:2Chr05_932313280.380.160.270.310.370.121.634.620.850.42(C)−0.11−0.17−0.02
C18:0Chr02_985586200.060.140.040.07−0.060.012.33−0.860.500.49(A)−0.030.04−0.01
1 2a, calculated as the difference between the phenotypic means observed within each homozygous class (2a = |GBB − Gbb|, where Gij is the trait mean in the ijth genotypic class). 2 d, calculated as the difference between the phenotypic mean observed within the heterozygous class and the average phenotypic mean across both homozygous classes [d = GBb − 0.5(GBB + Gbb), where Gij is the trait mean in the ijth genotypic class]. 3 Allele frequency of the minor allele. Single nucleotide polymorphism (SNP) alleles corresponding to the frequency listed are given in the parentheses. 4 The additive effect was calculated as a = pB(GBB) + pb (Gbb) − G, where G is the overall trait mean, Gij is the trait mean in the ijth genotypic class, and pi is the frequency of the ith marker allele. These values were always calculated with respect to the minor allele.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Lin, P.; Chai, J.; Wang, A.; Zhong, H.; Wang, K. High-Density Genetic Map Construction and Quantitative Trait Locus Analysis of Fruit- and Oil-Related Traits in Camellia oleifera Based on Double Digest Restriction Site-Associated DNA Sequencing. Int. J. Mol. Sci. 2024, 25, 8840. https://doi.org/10.3390/ijms25168840

AMA Style

Lin P, Chai J, Wang A, Zhong H, Wang K. High-Density Genetic Map Construction and Quantitative Trait Locus Analysis of Fruit- and Oil-Related Traits in Camellia oleifera Based on Double Digest Restriction Site-Associated DNA Sequencing. International Journal of Molecular Sciences. 2024; 25(16):8840. https://doi.org/10.3390/ijms25168840

Chicago/Turabian Style

Lin, Ping, Jingyu Chai, Anni Wang, Huiqi Zhong, and Kailiang Wang. 2024. "High-Density Genetic Map Construction and Quantitative Trait Locus Analysis of Fruit- and Oil-Related Traits in Camellia oleifera Based on Double Digest Restriction Site-Associated DNA Sequencing" International Journal of Molecular Sciences 25, no. 16: 8840. https://doi.org/10.3390/ijms25168840

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop