M. Udvardi1, T. Altmann1, B. Essigmann1, G. Colebatch1, S. Kloska1, P. Smith2, B. Trevaskis1
'Max Planck Institute of Molecular Plant Physiology, Am Mithlenberg 1, 14476 Golm,
2Dept Botany, Univ. of West. Aust., Crawley, WA 6009, Australia 1. Introduction
Functional genomics brings together high-throughput genetics with multi-parallel analyses of gene transcripts, proteins, and metabolites to answer the ultimate question posed by all genome-sequencing projects: What is the biological function of each and every gene? Functional genomics is driving a shift in the research paradigm away from vertical analysis of single genes, proteins, or metabolites, towards horizontal analysis of full suites of genes, proteins, and metabolites. By identifying and measuring many of the molecular players that participate in a given biological process, functional genomics offers the prospect of obtaining a truly holistic picture of life.
Technologies for high-throughput analysis of transcripts, proteins, or metabolites have developed rapidly over the last decade. Several approaches are now available for quantifying transcript levels for thousands of genes simultaneously, including DNA microarray analysis (Gress et al. 1992), and sequencing-based methods such as serial analysis of gene expression (SAGE, Velculescu et al. 1995) and massively parallel signature sequencing (MPSS; Brenner et al. 2000). Approaches that utilize mass spectrometry enable hundreds of proteins or metabolites to be identified and quantified in a single experiment (Fiehn et al. 2000; Washburn et al. 2001). However, there may be hundreds of thousands of different proteins, and tens of thousands of different metabolites in a given organism. Thus, current methods can provide information on only a small fraction of an organism's entire proteome or metabolome. This is not the case for transcriptome analysis, where DNA microarrays representing entire genomes are already being used to provide a comprehensive profile of gene activity in some bacterial species as well as the yeast, Saccharomyces cerevisiae. It is expected that comprehensive analysis of the Arabidopsis thaliana transcriptome will become possible in the near future as a result of the completion of the Arabidopsis genome sequence late in 2000. However, useful transcriptome analysis of other plant species need not wait until their genomes are completely sequenced. In fact, libraries of expressed sequence tags (ESTs) derived from cDNA clones are already being put to good use in this regard, including those of several legume species.
Lotus japonieus is a legume with a number of attributes that make it useful for molecular genetics and functional genomics (Handberg, Stougaard 1992). It is a self-fertile, diploid species with a relatively small genome (approximately 450 Mbp), is easily transformed using Agrobacterium tumefaciens, and produces large amounts of seed in a relatively short time. Lotus japonieus has become a model species for studies on symbiotic nitrogen fixation. We are using Lotus to identify genes and proteins that play important roles in nodule primary metabolism and nutrient exchange between the plant and nitrogen-fixing bacteroids. To this end, we are producing an EST database from Lotus japonieus nodule cDNA clones, and are using it not only for gene discovery, but also for the production of DNA microarrays for transcriptome analysis. Here, we describe the status of our Lotus EST project and present the first results of transcriptome analysis in this species.
2.1. Production of cDNA arrays. Two Lotus japonicus nodule cDNA libraries were used to generate clones for array analysis. The first library was provided by Dr. Jens Stougaard (Aarhus, Denmark), and constructed in the XZAP-XR vector (Stratagene). The second library was constructed in the pZLl vector using the Superscript™ Lambda system for cDNA synthesis and X cloning (Invitrogen™ Life Technologies). cDNA from 2307 bacterial plasmid clones was PCR amplified in 30 |il reactions in 384-well plates. For each clone, three independent reactions were performed and then pooled to reduce variation in PCR efficiency and to increase the concentration of product. PCR products were spotted in two positions onto Nytran® SuPerCharge nylon transfer membranes (Schleicher and Schuell, Germany) using the BioGrid robotic system (Bz'oRobotics Ltd, England). DNA on membranes was denatured with 0.5 M NaOH and baked at 80°C for two hours.
2.2. Reference hybridization. To estimate the amount of DNA spotted for each clone, filters were hybridized to a reference oligonucleotide that was complementary to vector sequence. For the XZAP-XR vector the oligonucleotide sequence was GCTGCAGGAATTC, and for the pZLl vector the oligonucleotide sequence was ACGCGTGGGTCGA. Oligonucleotides were end labeled with y33P-ATP using T4 polynucleotide kinase (NEB). Filters were pre-incubated in SSARC (24% Sarcosyl NL30, 4xSSC, 4 mM EDTA) for 2 hours at 5°C before addition of the probe and hybridization overnight at 5°C. Filters were washed for 30 min at 5°C and exposed to a phosphor screen (Fujifilm) for 4-6 hours, after which the imaging plate was scanned by the BAS-1800 II phosphor imager (Fujifilm) at a resolution of 50^m per pixel. Reference oligonucleotide probes were removed by stripping the filters at least twice in diluted SSARC (1/10 in 1 mM EDTA) for 30 min at 80°C, leaving the filters ready for use in complex hybridizations.
2.3. Complex hybridization. Total RNA was isolated from nodules of seven-week-old L. japonicus plants which had been inoculated at 7 days with Rhizobium loti, and with total RN55A from roots of seven-week-old uninoculated L. japonicus plants, as described by Jacobsen-Lyon et al. (1995). Five replicate complex hybridizations were performed for each tissue. Complex probes were prepared from 10 |ig of total RNA by reverse transcription using SuperScript II reverse transcriptase (Invitrogen Life Technologies) in the presence of a33P-CTP. Filters were pre-incubated for two hours at 65°C in Church buffer (0.5 M sodium phosphate, pH 7.2, 7% SDS, 1 mM EDTA) plus 100 Hg/ml denatured salmon sperm DNA, after which complex probes were added in fresh Church buffer and hybridized for at least 24 hours at 65°C. Filters were then washed successively for 20 min in lxSSC containing 0.1% SDS and 4 mM sodium phosphate (pH 7.2), 20 min in 0.2xSSC containing 0.1% SDS and 4 mM sodium phosphate (pH 7.2), and 20 min in O.lxSSC containing 0.1% SDS and 4 mM sodium phosphate (pH 7.2), all at 65°C. Air-dried filters were exposed to a phosphor screen overnight, which was then scanned as described above.
2.4. Data analysis. Detection and quantification of the signal intensities were performed using the AIS software package ArrayVision™ (4.0 Rev. 1.7; Imaging Research Inc), resulting in raw data representing the average pixel intensity for each spot. Raw data was normalized in a two-step procedure. Since we observed a linear dependency of the raw signal on the amount of PCR product as well as on the transcript abundance, we included the measurements from both the complex and the reference hybridization into our method. First, we normalized both hybridizations internally to compensate for varying amounts of total probe activity by dividing the background-subtracted signal by the average signal of all spots. In the second step, we divided the internally normalized complex signal by the internally normalized reference signal corresponding to the same spot in order to correct for different amounts of PCR product on the filter. Finally, the average of the two spots representing the same clone was calculated. Raw measurements that were not more than twice as high as empty local background spots were regarded as undetectable and replaced with an estimate of the missing value based on the local detection limit and the average and standard deviation of the replicate measurements of the same clone.
As of June 2001, we had sequenced 6000 Lotus nodule cDNA clones, which represented approximately 3000 different genes. Eighty percent of these encoded polypeptides that were homologous to known proteins. Twelve percent of ESTs appeared to encode enzymes, including an almost complete set of glycolytic enzymes and many enzymes involved in amino acid biosynthesis. Seventeen percent of encoded polypeptides were predicted to have at least one transmembrane domain. Amongst these were homologs to transporters for sugars, nucleotides, amino acids, peptides, and various anions or cations.
The first 2300 EST clones were used to produce DNA microarrays for transcriptome analysis. Microarrays were used to compare the transcriptome of nodules and roots from seven-week-old plants. While little technical and biological variation was apparent when replicate filters were probed with mRNA from the same organ (either root or nodule), there were marked and significant differences in gene expression between nodules and roots (Figure 1).
0.01 0.1 1 10 100 Nodule
Figure 1. Scatter plot of gene activity (relative transcript level) in nodules vs. roots of seven-week-old plants. Panel A shows all genes, panel B shows genes that were expressed at a significantly higher or lower level in nodules than in roots.
Transcripts of 68 genes were significantly up-regulated in nodules compared to roots, and a larger number of genes appeared to be down-regulated during nodule development. Amongst the genes that were up-regulated during nodule development were a number encoding known nodulins, including 6 leghemoglobins, LjNOD21, and homologs of MtN24, pea early nodulin 12, and Enodl6. Several genes involved in C metabolism were up-regulated in nodules, including a sugar transporter (4-fold up) and two different carbonic anhydrases (CA, 3 to 4-fold up). Homologs of CA have been found to be up-regulated in the nodules of other legume species. The transporter may facilitate sugar uptake into nodule cells, which require large amounts of energy and carbon skeletons for nitrogen fixation and assimilation, respectively. The carbonic anhydrases may play a role in CO2 recovery in nodules, a process in which another nodulin, phosphoenolpyruvate carboxylase, has been implicated in the past (Vance et al. 1994). Two genes involved in amino acid biosynthesis, aspartate amino transferase and aspartate kinase, were each up-regulated approximately two-fold in nodules. A number of nodulin genes involved in C and N metabolism in the nodules of other species were not present in the list of up-regulated genes in Lotus nodules. These included phosphoenolpyruvate carboxylase, glutamine synthetase, and glutamate synthase (Vance et al. 1994). Although homo logs of these genes were present on the arrays, it is possible that the particular orthologs of those up-regulated in alfalfa were not represented on our filters. A less likely explanation for their absence is that there are major differences in the way primary metabolism is altered during nodule development in Lotus compared to other species.
Several transporter genes appeared to be up-regulated in Lotus nodules, including three different sulfate transporter genes (between 2- and 20-fold up) and a potassium transporter (2-fold up). It will be interesting to determine whether any of these transporters are involved in nutrient exchange between the plant and bacteroids.
Two genes involved in heme biosynthesis, succinate-CoA ligase and coproporphyrinogen oxidase, were found to be up-regulated 2.5-fold in nodules. The heme moiety is essential for the production of leghemoglobin, the most abundant of nodule proteins. Coproporphyrinogen oxidase was shown to be up-regulated in nodules of soybean and pea in the past (Santana et al. 1998).
Several genes encoding proteins that may have roles in signal transduction were also found to be up-regulated, including a homolog of the transcription factor tga, a MADS-box protein homolog, and a homolog of ER6, an ethylene-inducible serine/threonine receptor kinase. A homolog of the enzyme isoliquiritigenin 2'-0-methyltransferase, which is involved in the biosynthesis of an enhancer of rhizobial nod genes (Maxwell et al. 1993) was up-regulated in Lotus nodules (more than 10-fold). This raises the question of whether this signaling pathway remains active after rhizobia have taken up residence in nodule cells.
Was this article helpful?