Introduction

Oilseed rape (B. napus L.) is an economically important crop that provides more than 13% of the world’s vegetable oil production [1]. B. napus is also a potential resource for the sustainable production of renewable bioenergy. However, to solve the challenges faced in the genetic improvement of seed fatty acid components, it is essential to fully understand oil biosynthesis [2, 3]. It is known that the GPATs as an initial and rate-limiting enzymes in the acylation reactions of de novo triacylglycerol (TAG) biosynthesis [2, 4, 5]. In previous study, the AtGPAT9 was identified to encode an endoplasmic reticulum sn-1 GPAT in TAG synthesis of seed [6, 7]. In addition, the decrease of oil content and increase of total polyunsaturated fatty acids in the seeds of the AtGPAT9 knockout mutant independent lines [8]. To date, only a few membrane-bound GPATs in B. napus (BnaGPATs) have been identified because of inconvenient purification and detection methods [9, 10].

The allotetraploid B. napus originated from the interspecific hybridization of B. oleracea and B. rapa [11, 12]. During evolution, early allopolyploidy and chromosome duplication resulted in genes with multiple homologues, which increases the difficulty of studying homologous gene functions. Moreover, functional divergence occurred in both paralogous and homoeologous genes during long-term domestication [9]. In addition, the redundancy of homologous genes affects transcription and functional divergence in genetic engineering with polyploid backgrounds. Fortunately, B. napus genomes and pan-genome have been assembled as whole-genome sequencing research has increased, and provide large amounts of data that can be used to mine for genes with vital quality and agronomic traits [13,14,15,16]. However, it should be noted that caution in applying the sequence data from progenitor genomes rearrangements (B. rapa and B. oleracea), even mostly functionally conserved in homologous gene [17]. Fatty acid elongase (FAE1 gene-encoded) catalyses two successive condensation reactions using oleoyl-CoA as the substrate in erucic acid biosynthesis [18]. Loss of function in FAE genes in domesticated B. napus results in rapeseed varieties with low erucic acid levels [19], whereas the contributions of GPATs to the de novo biosynthesis of fatty acids are unknown. Furthermore, it also unknown if there exist GPATs lacking the conserved acyltransferase domains in plants. Therefore, we have constructed a yeast genetic complementation system for the screening of eukaryotic GPATs [20]. By constructing the exogenous gene into the yeast expression vector yADH1-pYES2-KanV2 and transforming it into the yeast strain ZAFU1, we can quickly identify whether the exogenous gene has acyltransferase activity based on whether it can restore the growth of yeast on a medium supplemented with glucose. In this study, to examine the homologous functional divergence in glycerolipid biosynthesis, four homologous BnaGPAT9-encoded genes were characterized by yeast genetic complementation and transformation of seed-specific expression.

Results

Characterization of homologous GPAT9 genes in B. Napus

Four homologous GPAT9 genes (i.e., BnaGPAT9-A1, C1, A10, C9) was amplified from B. napus (advanced line 21L10). All of these sequencing data have been uploaded into the NCBI repository, and the GenBank accession numbers of BnaGPAT9-A1 (OR521147), C1 (OR521149), A10 (OR521148), C9 (OR536417) are listed in Supplementary Table S1. Table 1 shows the CDS and protein sequence identity of the four BnaGPAT9 genes, and 97.2% similarity was found between the orthologous gene pairs BnaGPAT9-A1 and BnaGPAT9-C1 and BnaGPAT9-A10 and BnaGPAT9-C9. Moreover, protein sequence similarities of 100% and 98.4% were found between A1 and C1 and between A10 and C9, respectively. The sequence differences in CDS and proteins among the four homologues are illustrated in Supplementary Figure S1 and S2, respectively.

Table 1 CDS and protein sequence identity of the four homologous BnaGPAT9 genes
Full size table

Identification of BnaGPAT9 activity

The full-length CDS of BnaGPAT9-A1/C1/A10/C9 were cloned into the vector yADH1-pYES2-Kan V2 (Supplementary Fig. S3), and were transformed into the yeast mutant strain (ZAFU1) to identify acyltransferase activity. The exogenous gene (AtGPAT1) in episomal vector was driven by galactose induced promoter, that can support the growth of the ZAFU1 on SC-Ura-His-Leu + galactose medium. On the other hand, the candidate GPAT in another episomal vector was driven by glucose induced promoter, that can rescue the growth of the ZAFU1 on SC-Ura-His-Leu + glucose medium. According to yeast genetic complementation, BnaGPAT9-A1 and BnaGPAT9-C1 rescued the growth of ZAFU1 on SC-Ura-His-Leu + glucose solid medium, but BnaGPAT9-A10 and BnaGPAT9-C9 did not rescue growth (Fig. 1). Thus, BnaGPAT9-A1/C1 have acyltransferase activity.

Fig. 1
figure 1

Only one pair of homeologous BnaGPAT9 genes rescues yeast growth under glucose. Yeast genetic complementation with heterologous expression of BnaGPAT9-A1/C1 could rescue the mutant strain ZAFU1 growth on SC-Ura-His-Leu + glucose solid medium, but BnaGPAT9-A10/C9 did not rescue growth. P: positive control (glycerol-3-phosphate acyltransferases 1 gene of Saccharomyces cerevisiae, ScGAT1); N: negative control (empty vector). The independent colonies were serially diluted (1:5) to an initial optical density (OD600) of 1.0

Full size image

Tissue-specific expression of BnaGPAT9-A1/C1

The expression patterns of BnaGPAT9-A1/C1 were analysed by qRT‒PCR in roots, stems, leaves, and developing seeds (Supplementary Fig. S4). The relative expression of BnaGPAT9-A1 in seeds at early stages of development (15 and 20 days after flowering) was higher than that of BnaGPAT9-C1, whereas at later stages of development (25, 30, 35, and 40 days after flowering), the relative expression of BnaGPAT9-C1 was higher than that of BnaGPAT9-A1 (Fig. 2). Analyses of the expression patterns shown that these two genes transcript was associated with seed development.

Fig. 2
figure 2

Relative expression patterns of BnaGPAT9-A1/C1 in different tissues of ‘Zhongshuang 11’. The relative expression level of BnaGPAT9-A1/C1 genes were higher in development seeds than that of in roots, stems, and leaves. At early stages of development seeds, the BnaGPAT9-A1 relative expression level was higher than that of BnaGPAT9-C1, whereas on the contrary at later stages of development seeds. The values are the means ± SDs, n = 3

Full size image

Effect of single nucleotide polymorphisms on BnaGPAT9-C1 enzyme activity

A single nucleotide polymorphism (SNP) occurred in BnaGPAT9-C1 (GA1124C or GT1124C) that induced a change from aspartic acid (D) into valine (V) at the 375 aa site. The sequencing data of BnaGPAT9-C1 from commercial varieties Qinyou2 (OR536420), Zheyou50 (OR536421), Zhongshuang11 (OR536422) and advanced lines 20CP75 (OR536418), 634 (OR536419) has been uploaded into the NCBI repository (Supplementary Table S1). Notably, the amino acid sequence with 375 V exhibited no GPAT enzyme activity to rescue the growth of ZAFU1 on SC-Ura-His-Leu + glucose solid medium (Fig. 3). Furthermore, based on the SNP, 44 materials (14 commercial varieties and 30 advanced lines) were classified into three haplotypes by the PARMS genotyping platform (Fig. 4). The results shown that 18.2% and 65.9% of the varieties/lines belonged to the H2 (A/T heterozygous) and H3 (T/T homozygous) haplotypes, respectively (Table 2).

Fig. 3
figure 3

The advanced lines 21L10, 20CP75 and 634 with BnaGPAT9-C1 (GA1124C encodes aspartic acid 375D) have acyltransferase activity that can rescues the ZAFU1 growth on the SC-Ura-His-Leu + glucose solid medium, whereas commercial varieties Qinyou2, Zheyou50 and Zhongshuang11 with BnaGPAT9-C1 (GT1124C encodes valine 375 A) did not. ScGAT1: S. cerevisiae glycerol-3-phosphate acyltransferase 1; Empty vector: yADH1-pYES2-Kan V2; 21L10, 634, 20CP75: advanced lines; Zhongshuang 11, Zheyou 50, Qinyou 2: commercial varieties

Full size image
Fig. 4
figure 4

44 genotypes of B. napus were divided into three haplotypes with single nucleotide polymorphism in position of the BnaGPAT9-C11124 by a penta-primer amplification refractory mutation system (PARMS). Blue dots (seven materials) represent homozygous haplotypes (A/A); red dots (eight materials) represent heterozygous haplotypes (A/T); green dots (twenty nine materials) represent homozygous haplotypes (T/T); grey dots represent negative controls. HEX: green fluorescence universal primer; FAM: blue fluorescence universal primer

Full size image
Table 2 44 commercial varieties/advanced lines were divided into three haplotypes by a single nucleotide genotyping at 1124 bp site of the BnaGPAT9-C1 gene in B. napus. Haplotype analysis of BnaGPAT9-C11124 using a penta-primer amplification refractory mutation system (PARMS)
Full size table

Association of haplotype and erucic acid (C22:1) content in seeds

The seed fatty acid compositions were analysed by gas chromatography in 44 genotypes of B. napus (Supplementary Table S2). The analysis showed that the erucic acid content of all H1 haplotypes (A/A) and H2 haplotypes (A/T) was higher than 10%, whereas that of 62.1% (18/29) of the H3 (T/T) haplotypes was less than 10%, especially in all commercial varieties (Fig. 5 and Supplementary Fig. S5). In addition, 11 advanced lines of the H3 haplotype had erucic acid contents higher than 10% (Fig. 5). These results implied that GPAT enzyme activity could increase the erucic acid (C22:1) content of seeds and other regulatory pathways in erucic acid biosynthesis for genetic improvement.

Fig. 5
figure 5

Distribution of three haplotypes with different erucic acid content. In H1 haplotype, the erucic acid content of all advanced lines was more than 10%. In H3 haplotype, the erucic acid content of all commercial varieties was less than 10%. H1: homozygous haplotype (A/A), H2: heterozygous haplotype (A/T), H3: homozygous haplotype (T/T)

Full size image

Seed-specific expression of BnaGPAT9-C1 1124A for regulating the erucic acid content in ‘Zhongshuang 11’ seeds

A 1131-bp target fragment was detected by double enzyme digestion in the recombinant vector pNapin-BnaGPAT9-C11124A-Nos (Fig. 6A) which were used for A. tumefaciens-mediated genetic transformation. The morphology of explants, calli, induced shoots and transgenic plantlets at various stages of regeneration is shown in Fig. 6B and F. In seed-specific expression of BnaGPAT9-C11124A transformations, 13 positive transformants were obtained by specific PCR identification, and a 1027-bp target fragment amplified (Supplementary Fig. S6). Moreover, the expression level of BnaGPAT9-C11124A in the transformants was higher than that in the wild type, as revealed by qRT‒PCR (Fig. 7). The fatty acid content of T1 seeds of four transformants (lines 6, 11, 12, and 13) was analysed, and all of the components are shown in Table 3. The contents of oleic acid (C18:1) in the seeds of these four transgenic lines were significantly lower than that of the wild type, but the opposite results were found for the linolenic acid (C18:3) and erucic acid (C22:1) contents (Table 3).

Fig. 6
figure 6

Construction of pNapin-BnaGPAT9-C11124A-Nos transformants in the ‘Zhongshuang 11’ genetic background. (A) Identification of the recombinant vector. 1: Double enzyme digestion of the recombinant vector, the BnaGPAT9-C1 fragment was 1131 bp; 2: no enzyme digestion of the recombinant vector; M: 1-kb DNA ladder. (B–F) Genetic transformation of hypocotyls. B: hypocotyl; C: callus; D and E: induced shoots; F: transgenic plantlets. Scale bar = 1 cm

Full size image
Fig. 7
figure 7

Relative expression level of pNapin-BnaGPAT9-C11124A-Nos in independent T0 transgenic lines. The values are the means ± SDs, n = 3. WT: untransformed ‘Zhongshuang 11’ with the BnaGPAT9-C1 homozygous haplotype (T/T) at the 1124 bp site. N-1 to N-13: independent pNapin-BnaGPAT9-C11124A-Nos transformant lines in the ‘Zhongshuang 11’ genetic background

Full size image
Table 3 Analysis of the seed fatty acid composition of independent T1 transgenic lines with pNapin-BnaGPAT9-C11124A-Nos
Full size table

Discussion

With the sequencing of genomes that have undergone complex genomic changes such as whole/partial genome duplications, and interspecific hybridization we can study the functional evolution of duplicated genes. As a result, B. napus has become an important model for investigations of the consequences of polyploidy, which have provided intriguing insights into genome restructuring resulting from duplicated selective forces during crop evolution. However, little is known about the functional divergence between paralogous or homologous genes in the evolution of genomic segmental duplications. In this study, four homologous BnaGPAT9 genes were identified. Based on sequence identity, we could able to determine the BnaGPAT9-A1/A10 and BnaGPAT9-C1/C9 origin from B. rapa and B. oleracea, respectively. Nevertheless, BnaGPAT9-A1/C1 exhibited GPAT enzyme activity and thus rescued the growth of the double knockout mutant strain (ZAFU1) in yeast genetic complementation assays, whereas BnaGPAT9-A10/C9 did not rescue ZAFU1 growth (Fig. 1). The results indicated that functional divergence occurred, and through this process, duplicated genes lose a redundant function during crop evolution [9, 21].

In addition, homologous genes have different developmentally regulated mechanisms in polyploidy crops [9, 22]. For example, the BnaGPAT4-A1 and BnaGPAT4-C1 shown significantly opposite expression level in maturing embryo and developing seed coat [9]. In our study, the seed-specific expression patterns indicated that BnaGPAT9-C1 might be important in fatty acid biosynthesis during later stages of seed development (25–40 days after flowering), whereas BnaGPAT9-A1 might be important in early stages (15–20 days after flowering) (Fig. 2). The possible mechanisms are supposed to be important for polyploid crops adaptation during the domestication processes [23].

In this study, an SNP of BnaGPAT9-C11124T resulted in GPAT activity that was too low to rescue ZAFU1 (Fig. 3) and produced a phenotype consisting of a low erucic acid content in seeds, especially in commercial varieties (Fig. 5). It is speculated that the emergence of multiple haplotypes may be due to the improved adaptability of crops during domestication processes [23]. Fatty acid analysis showed that transformants with seed-specific expression of BnaGPAT91124A exhibited an increase in the erucic acid content compared with that of ‘Zhongshuang 11’, a commercial variety with a low erucic acid content. In summary, the results of this study imply that BnaGPAT9-A1/C1 has substrate preferences in triacylglycerol biosynthesis. For further step, we intend to use another line with high erucic acid as the genetic background or in vitro enzyme assays to check the preference of BnaGPAT9.

Conclusions

Functional divergence in BnaGPAT9 genes was identified using the yeast mutant strain ZAFU1 and expression pattern analysis. The BnaGPAT9-A1/C1 homologues but not the BnaGPAT9-A10/C9 homologues encoded functional GPAT enzymes. In addition, an SNP of BnaGPAT9-C1 (A or T at a 1124-bp site) that occurred during evolution and domestication processes was associated with enzyme activity and contributed to the erucic acid content. Moreover, seed-specific expression of BnaGPAT9-C11124A increased the erucic acid content in the seeds of the transformants. Thus, this work will be of great interest to breeders working on the genetic and breeding improvement of oil crops.

Materials and methods

Plant materials

B. napus seeds of 44 genotypes (14 commercial varieties and 30 advanced lines) were used in the experiments. Twenty-five of the advanced lines were kindly provided by Prof. Weijun Zhou (Zhejiang University, Hangzhou, China), whereas four were provided by Dr. Qi Peng (Jiangsu Academy of Agricultural Sciences). Information on these materials is listed in Supplementary Table S3.

Cloning and homology analysis of GPAT9 genes from B. Napus

Fourteen-day-old seedlings of the B. napus advanced line ‘21L10’ were used to clone BnaGPAT9-A1/C1/A10/C9 derived from the A and C genomes. The coding sequences of GPAT9s were amplified using the High-Fidelity Enzyme PrimeSTAR Kit (Takara, Japan) with the specific primers listed in Supplementary Table S4. The complementary DNA (cDNA) and protein sequence identity of the GPAT9 genes were analysed using DNAMAN software version 9.0 (Lynnon BioSoft, San Ramon, CA, USA).

Vector construction and yeast genetic complementation

To obtain heterologous expression, BnaGPAT9 genes were cloned and inserted into yADH1-pYES2-Kan V2 vectors with glucose induction using the primers listed in Supplemental Table S4. In the yeast genetic complementation assay, the yeast mutant ZAFU1 (BY4742, gat1Δgat2Δ+[pGAL1::AtGPAT1 LEU2]) was used for screening acyltransferase activity. Subsequently, yeast genetic complementation procedures were conducted according to Lei et al. [20] and Liu et al. [24]. In a serial dilution (1:5) assay, a single clone was inoculated into SC-Ura-His-Leu + galactose or glucose medium with an initial optical density (OD600) of 1.0 and cultured for 5 d at 30 °C.

Reverse-transcription quantitative PCR

Total RNA of ‘Zhongshuang 11’ was extracted using a TransZol Up Plus RNA Kit (Transgene, Shenzhen, China) and detected using a NanoDrop™ One (Thermo Fisher, Waltham, MA, USA). cDNA was synthesized using Hifair® 1st Strand cDNA Synthesis SuperMix for qPCR (Yeasen, Shanghai, China). The expression patterns of BnaGPAT9-A1/C1 in roots, stems, leaves, and developing seeds (15, 20, 25, 30, 35, and 40 days after flowering) were determined by CFX Connect™ Optics Module (Bio-Rad, Hercules, CA, USA) with three biological replications. The 20-µl qRT‒PCR contained 10 µl of 2 x Hieff qPCR SYBR Green Master Mix (Yeasen, Shanghai, China), 0.004 nM forward primer, 0.004 nM reverse primer, and 9.2 µl of cDNA. The reaction program was initiated by predenaturation at 95°C for 5 min, and this step was followed by 40 cycles of denaturation (95°C for 10 s) and annealing (55°C for 30 s). The reference gene was ubiquitin-conjugating enzyme 9 (accession no. XM_013800933) [9], which was used to normalize the expression levels of BnaGPAT9-A1/C1. To distinguish BnGPAT9-A1/C1 by qPCR, primers were designed for the 3’ UTR and 5’ UTR differential regions of the BnaGPAT9-A1 and BnaGPAT9-C1 mRNA sequences, respectively. All primers are listed in Supplementary Table S4.

Haplotype identification

Based on the SNP of BnaGPAT9-A1/C1 in domestication, a penta-primer amplification refractory mutation system (PARMS) was designed to identify the haplotype in the commercial varieties and advanced lines. The standard operation procedure was performed according to Lu et al. [25], and the PARMS primers are listed in Supplementary Table S4.

Seed fatty acid component analysis

To determine seed fatty acid components, 30.0 mg of seed samples were analysed using an Agilent 7890B gas chromatograph (Agilent, Santa Clara, CA, USA). The analysis was conducted according to Ichihara and Fukubayashi [26] with minor modifications. Samples were obtained from three biological duplicates of commercial varieties and advanced lines.

Construction of BnaGPAT9-C1 1124A seed-specific expression transformants in Zhongshuang 11

BnaGPAT9-C11124A was cloned and inserted into a p1300-Napin-Nos vector with SacI and PstI. Genetic transformation with Agrobacterium tumefaciens GV3101 was conducted according to Liu et al. [27] with minor modifications. The T0 generation lines were confirmed by PCR, and T1 seeds from the BnaGPAT9-C11124A high-expression lines were harvested for fatty acid component analysis. All primers are listed in Supplementary Table S4.

Statistical analyses

The seed fatty acid composition was determined on the basis of peak areas. The variance of the mean was analysed by least significant difference tests at the significance level of P ≤ 0.05. The relative expression level was calculated according to the 2−ΔΔCt analysis method [28].