Generation and analysis of expressed sequence tags from a cDNA library of the fruiting body of Ganoderma lucidum

The fruiting body of G. lucidum was obtained from the co-author Jin Lan, who has long been engaged in Ganoderma research in the Institute of Medicinal Plant Development, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China. She authenticated the G. lucidum using the morphological identification approach and referred to the Fungi Identification Manual [25]. Fifty (50) days after growing on the basswood medium at 25-30°C in a shade shelter, approximately 0.5 g was harvested and frozen in liquid nitrogen immediately. The mRNA of thefruiting body was isolated and purified directly with a Dynabeads (R) mRNA DIRECT™ kit (Invitrogen, USA) according to the manufacturer's recommendations. The cDNA library was constructed from purified mRNA with a Creator™ SMART™ cDNA Library Construction kit (Clontech, USA). The double-stranded cDNA was directionally ligated into the Sfi I restriction site of the pDNR-lib vector (Clontech, USA) and electroporated into a DH5α Escherichia coli strain (TakaRa, Japan).

A total of 1,023 randomly selected clones were cultured in liquid LB medium containing 34 mg/l chloramphenicol and incubated overnight at 220 rpm in rotation and 37°C. Plasmid DNA was prepared with an Axyprep-96 plasmid kit (Axygen, USA). The plasmid DNA was submitted for direct sequencing from the 5' end with an M13 forward primer on an ABI 3730 DNA sequencer using BigDye 3.1 sequencing chemistry (Applied Biosystems, USA).

The ABI-formatted chromatogram sequences were processed automatically with a local EST analysis pipeline. The Phred/Phrap program was applied for trace files conversion and for base calling with quality assessment [26, 27]. The vector and low-quality regions were removed from the sequence with the Cross Match typically included in the Phred/Phrap program. The short sequences (less than 100 bp) and poly A/T tails were filtered from the EST database. The high quality ESTs were assembled into contigs (clusters of assembled ESTs) and singletons (sequences found only once) by Phrap [28].

The unique sequences were searched against public databases including the SwissProt [29], NCBI non-redundant protein (Nr) [30] and non-redundant nucleotide (Nt) [31] databases using BLAST [32] algorithm, with a E-value cut-off at 10^-5. The functional categories of these unique sequences were classified by a broad category, including metabolism, energy production, cell signalling, cell defence and stress response, cell structure and growth, transcription, protein synthesis, protein degradation, transport and secretion as well as unclassified and unknown function.

The detection of simple sequence repeats (SSRs) from the total high-quality ESTs of the fruiting body of G. lucidum was performed with the Simple Sequence Repeat Identification Tool (SSRIT) [33]. The SSRIT accepts FASTA-formatted sequence files and reports the sequence ID, SSR motif, number of repeats (di- and tri-nucleotide repeat units), repeat length and position of the SSR and the total length of the sequence in which the SSRs were found [34]. The search parameters for the maximum motif-length group were set to hexamer and those for the minimum number of repeats were set to five.

A cDNA library was constructed from the fruiting body of G. lucidum for the identification of the transcripts and the expression profiles involved in its cellular development and biosynthesis of secondary metabolites. The cDNA library had a titre of 1.25×10⁶ colony forming units per millilitre (ml). A total of 1,023 cDNA clones were randomly selected from the library for sequencing, yielding 879 (85.9%) high-quality ESTs after vector screening and short sequence (<100 bp) filtering. These ESTs were assembled into 82 contigs and 518 singletons for a total of 600 unique sequences (Table 1). The average sequence length of these unique sequences was 288 bp, ranging from 0.15 kb to 1.5 kb. Over 63.4% contigs had two sequences, followed by 22.0% having three to four sequences and 14.6% having five to 40 sequences. Approximately 31.75% of the ESTs were redundant. A total of 427 unique sequences (71.1%) displaying no similarities to any sequences in the public databases were probably new transcripts. The redundancy (31.75%) of the ESTs suggests considerable potential for new transcripts in continued sequencing of random colonies from this cDNA library. The sequenced transcripts have been deposited in the GenBank database (GO447131-GO448009).

The expressed profile of the unique sequences identified in the G. lucidum fruiting body is shown in Table 2. Among 600 unique sequences, 518 (86.3%) unique sequences were sequenced only once; 72 (12%) unique sequences 2-5 times; five (0.8%) unique sequences 6-10 times and five (0.8%) unique sequences 12 times or more. The most abundantly expressed unique sequences in the G. lucidum fruiting body were coded for the hypothetical proteins (48 ESTs) and the cell wall-associated hydrolase (36 ESTs) (Table 3). Moreover, the unique sequence consisting of five ESTs and with sequence similarity to hydrophobin 2 of Lentinula edodes (Xianggu) was identified for the first time in G. lucidum (Table 3). Moreover, the unique sequences matched the elongation factors and the ribosomal proteins were also expressed at high levels (Table 3).

The list of the annotated ESTs found in the fruiting body of G. lucidum is shown in Additional file 1. Sixty-two (62) ESTs showed sequence similarities to uncharacterized genes encoding hypothetical proteins that were omitted from the list. The unique sequences from this cDNA library were analyzed for similarities by performing BLAST searches against public databases, including SwissProt [29], Nr [30] and Nt [31]. A total of 139 (23.2%) and 67 (11.2%) unique sequences were assigned a putative identity based on significant sequence similarities to at least one sequence in the Nr and Nt databases, respectively. These annotated unique sequences provide an available resource for application and basic microbiology. Furthermore, among the 879 ESTs only three (0.3%) ESTs were identified as homologues of previously reported nucleotides from G. lucidum in the GenBank database, indicating that the vast majority of the ESTs in our dataset were unique and new. The three unique sequences showed similarities to cytochrome c oxidase subunit 2 (GO447869), glyceraldehyde-3-phosphate dehydrogenase (GO447698) and FPS (GO447502) (Additional file 1).

The functions of the proteins that the identified sequences encoded were classified into categories of metabolism, energy production, cell signalling, cell defence and stress response, cell structure and growth, transcription, protein synthesis, protein degradation, transport and secretion as well as unclassified and unknown functions. The functional distribution of identified sequences from the G. lucidum fruiting body cDNA library is shown in Figure 1. The unique sequences associated with metabolism (22%), protein synthesis (22.7%), unclassified and unknown function (12.7%) and energy production (11.3%) were strongly represented, whereas those with cell signalling (2.0%) and cell defence and stress response (3.3%) were not. In summary, a total of 173 unique sequences showed similarities to known genes involved in the biosynthesis of secondary metabolites and developmental regulation. While the EST sequencing scale is limited, it provides some information about the expressed transcript profile of the fruiting body of G. lucidum.

SSRs, also known as microsatellites, are useful genetic markers in molecular biology. A total of 13 SSR motifs were identified from the EST sequences of the fruiting body of G. lucidum (Table 4). The composition of di- and tri-nucleotide SSRs included AC (two), AT (two), CT (one), GA (one), GT (one), TA (one), TC (one), and TG (two). Only two tri-nucleotide repeats with the composition of CGA and GGT were found in the EST dataset (Table 4). Each SSR-containing unique sequence contains only one SSR motif. The lengths of eight repeats were nine bases and the other five repeats were between 11 and 15 bases. In addition, the tetra-, penta- and hexa-nucleotide repeats were not present in this EST dataset. There was a wide variation in the frequency of SSR motifs among species [35].

EST analysis is an important tool to identify secondary metabolite genes in the fruiting body of G. lucidum. Triterpenoids, the major bioactive compounds in G. lucidum, are synthesized from acetyl-CoA in the isoprenoid pathway. While genes involved in the triterpenoid biosynthetic pathway including SQS, GlFPS and Gl-HMGR were cloned from and identified in G. lucidum[9‐11], other genes for the key enzymes in this pathway are to be identified.

According to the studies of the triterpene biosynthesis [10, 36], SE and FPS are rate-limiting enzymes in catalyzing triterpenoid biosynthesis in G. lucidum. The unique sequence (GO447913) with 71% identity (E-value = 1.00^-12) to SE and the unique sequence (GO447502) with 98% identity to FPS (E-value = 4.00^-27) involved in triterpenoid biosynthesis were presented in our EST data. SE acts as an important regulatory enzyme in the triterpenoid biosynthetic pathway. SE (EC 1.14.99.7), a monooxygenase, converts squalene into 2,3-oxidosqualene [36, 37]. The enzyme requires molecular oxygen, flavin adenine dinucleotide (FAD), either NADH or NADPH depending on the organisms [38]. Since the gene encoding SE has not been identified in G. lucidum, the information of the unique sequence (GO447913) will help identify and characterize the SE in G. lucidum. The EST for FPS (GO447502) shows sequence similarity to the GlFPS, suggesting that this EST is the partial sequence of the full-length GlFPS.

The genes encoding key enzymes involved in the triterpenoid biosynthesis, such as SQS, Gl-HMGR and others, are not present in this EST dataset, indicates a low abundance of these genes in the fruiting body of G. lucidum or incomplete sequencing of the library. The absence of ESTs associated with the polysaccharide biosynthesis, which should be abundant in G. lucidum, in this study may be due to the limited sequencing scale.

Several transcripts present in the EST dataset encode the proteins that may be associated with the development processes of the fruiting body of G. lucidum. The unique sequence (GO447641) showed sequence similarity to the inorganic phosphate transporter PHO84 gene which controls the absorption of phosphate nutrition and regulates the development of Saccharomyces cerevisiae[39]. MOB2 is a nonessential yeast gene and plays a role in the maintenance of ploidy [40]. The unique sequences homologous to MOB2 (GO447972) and PHO84 (GO447641) may have the same functions as those in yeast. Hydrophobin is expressed specially in filamentous fungi and is important during the morphogenesis of fungi and the fruiting body development of mushrooms [41]. The unique sequence homologous to hydrophobin 2 of Lentinula edodes in this cDNA library consisted of five ESTs (GO447695, GO447166, GO447364, GO447414, GO447512), suggesting its abundance in the fruiting body of G. lucidum. Suizu et al. (2008) reported that the ESTs for hydrophobins were also most frequently identified in the cDNA library of Lentinula edodes[15]. Profilin is a universal small eukaryotic protein that binds to monomeric actin (G-actin) and is involved in diverse functions such as maintenance of cell structural integrity, cell mobility and growth factor signal transduction [42]. The sequences (GO447955, GO447282) encoding profilin were present in the G. lucidum cDNA library. The important unique sequence encoding an argonaute-like protein (GO447302) may be involved in the RNAi pathway, suggesting a potential gene knock-out by RNA interference in G. lucidum. Cloning and characterization of these candidate genes is under way.

The ESTs sequenced in this study from the fruiting body of G. lucidum were insufficient to cover all functional genes, although this EST dataset showed some characteristics of gene expression in the fruiting body of G. lucidum.