Mining the bitter melon (momordica charantia l.) seed transcriptome by 454 analysis of non-normalized and normalized cDNA populations for conjugated fatty acid metabolism-related genes

Abstract

Background: Seeds of Momordica charantia (bitter melon) produce high levels of eleostearic acid, an unusual conjugated fatty acid with industrial value. Deep sequencing of non-normalized and normalized cDNAs from developing bitter melon seeds was conducted to uncover key genes required for biotechnological transfer of conjugated fatty acid production to existing oilseed crops. It is expected that these studies will also provide basic information regarding the metabolism of other high-value novel fatty acids.

Results: Deep sequencing using 454 technology with non-normalized and normalized cDNA libraries prepared from bitter melon seeds at 18 DAP resulted in the identification of transcripts for the vast majority of known genes involved in fatty acid and triacylglycerol biosynthesis. The non-normalized library provided a transcriptome profile of the early stage in seed development that highlighted the abundance of transcripts for genes encoding seed storage proteins as well as for a number of genes for lipid metabolism-associated polypeptides, including Δ12 oleic acid desaturases and fatty acid conjugases, class 3 lipases, acyl-carrier protein, and acyl-CoA binding protein. Normalization of cDNA by use of a duplex-specific nuclease method not only increased the overall discovery of genes from developing bitter melon seeds, but also resulted in the identification of 345 contigs with homology to 189 known lipid genes in Arabidopsis. These included candidate genes for eleostearic acid metabolism such as diacylglycerol acyltransferase 1 and 2, and a phospholipid:diacylglycerol acyltransferase 1-related enzyme. Transcripts were also identified for a novel FAD2 gene encoding a functional Δ12 oleic acid desaturase with potential implications for eleostearic acid biosynthesis.

Conclusions: 454 deep sequencing, particularly with normalized cDNA populations, was an effective method for mining of genes associated with eleostearic acid metabolism in developing bitter melon seeds. The transcriptomic data presented provide a resource for the study of novel fatty acid metabolism and for the biotechnological production of conjugated fatty acids and possibly other novel fatty acids in established oilseed crops.

Figure 1

Fatty acid compositions of lipids and detection of selected transcripts at different stages of bitter melon seed development. (A). Distribution of fatty acids in phospholipids (PL, open circle) and the triacylglycerol (TAG)-enriched neutral lipids (NL, black square) during the development of bitter melon. The fatty acids shown are palmitic acid, 16:0; stearic acid, 18:0; oleic acid, 18:1; linoleic acid, 18:2; linolenic acid, 18:3; and eleostearic acid, ESA. (B). RT-PCR detection of transcripts for the bitter melon conjugase (McConj) and oleosin (McOleo) during bitter melon seed development. β-Tubulin (Mcβ-Tub2)-specific primers were used as a control to assess the quality of first-strand cDNA. The seeds were collected at the designated days after pollination (DAP).

Figure 2

Preparation of normalized cDNA library from total RNA. Abbreviations: ss cDNA, single-stranded cDNA; ds cDNA, double-stranded cDNA.

Figure 3

Illustration of normalized versus non-normalized library on agarose gel electrophoresis. (A). Analysis of equal loadings of non-normalized and normalized cDNA populations on a 1% agarose gel. (B). DNA insert fragments were amplified by PCR from random individual colonies in the normalized cDNA library, and analyzed on a 1% agarose gel. The 1 kb DNA ladder was loaded as control.

Figure 4

Distribution of sequence length (A) and number of reads (> = 2) (B) of contigs in both non-normalized and normalized cDNA populations.

Figure 5

Go function analysis [55]of Arabidopsis gene homologous from both non-normalized and normalized cDNA libraries. Distribution of genes associated with different cellular components, molecular functions and biological processes.

Figure 6

Numbers of reads in the non-normalized and normalized 454 sequence analysis for genes encoding enzymes in the fatty acid and TAG biosynthetic pathways in bitter melon seeds. Numbers after the gene (a/b) are the numbers of reads for the corresponding genes in the 454 analysis of cDNA libraries: a, non-normalized; b, normalized. The dashed arrows in the plastid fatty acid biosynthetic pathway indicate one or more cycles of acyl-chain elongation that is initiated by 3-ketoacyl-ACP synthase (KAS) I or II. Abbreviations: BCCP, biotin carboxyl carrier protein subunit of acetyl-CoA carboxylase; BC, biotin carboxylase subunit of acetyl-CoA carboxylase; α-CT, α-carboxyltransferase subunit of acetyl-CoA carboxylase; ACP, acyl carrier protein; FAB2/SAD, stearoyl-ACP desaturase; FatA, acyl-ACP thioesterase A; FatB, acyl-ACP thioesterase B; GPAT, glycerol 3-phosphate acyltransferase; LPAT, lysophosphatidic acid acyltransferase; PA Pase: phosphatidic acid phosphatase; LPCAT, lysophosphatidylcholine acyltransferase; AAPT, CDP-choline:diacyglyglycerol cholinephosphotransferase; PLC, phospholipase C-type enzymes; DGAT, diacylglycerol acyltransferase; PDAT, phospholipid:diacylglycerol acyltransferase; FAD2, Δ12 oleic acid desaturase; FAD3, Δ15 (ω-3) linoleic acid desaturase; G3P, glycerol-3-phosphate; LPA, lysophosphatidic acid; PA, phosphatidic acid; LPC, lysophosphatidylcholine; PC, phosphatidylcholine; DAG, diacylglycerol; TAG, triacylglycerol.

Figure 7

Amino acid sequence alignments of McDGAT1 and DGAT1 polypeptides from various species. (A) Alignment of various DGAT1 s. Sequences were aligned by Clustal × and displayed using GeneDoc Software. (B) Phylogenetic tree of DGAT1 polypeptides. Sequences of DGAT1 are from the following species: Vitis vinifera (Vv) (gi 225444869), Euonymus alatus (Ea) (gi 54145459), Arabidopsis thaliana (At) (gi15224779), Vernicia fordii (Vf) (gi 86279632), Ricinus communis (Rc) (gi 255546145). Phylogenetic tree was constructed using MEGA 4.0.1 software [52].

Figure 8

Characterization of FAD2 polypeptides and genes in bitter melon. (A) Alignment of amino sequences for divergent FAD2 s (McFAD2, McFAD2v) and the FAD2-related conjugase (McConj) from bitter melon cDNA library. Sequences were aligned by Clustal X and displayed using GeneDoc Software. (B). Phylogenetic tree of FAD2 s and conjugase. Sequences used to construct the tree include Arabidopsis thaliana At FAD2 (gi 21536781), Trichosanthes kirilowii Tk FAD2 (gi 28371823), Tk Conj (gi 28371821), and Sorghum bicolor Sb DES2 (gi 242062720). (C). Production of 16:2 and 18:2 in Saccharomyces cerevisiae by expression of McFAD2 and McFAD2v. As indicated, no 16:2 or 18:2 was detected in yeast cells containing only the vector. (D) Expression of McFAD2, McFAD2v, and conjugase (McConj) during bitter melon seed development as determined by RT-PCR. β-Tubulin (β-Tub) -specific primers were used as a control to assess the quality of first-strand cDNA.