This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 138/3/1734    most recent pp.105.062851v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (13) Citing Articles via Google Scholar Google Scholar Articles by Salmi, M. L. Articles by Roux, S. J. Search for Related Content PubMed PubMed Citation Articles by Salmi, M. L. Articles by Roux, S. J. Agricola Articles by Salmi, M. L. Articles by Roux, S. J.

GENETICS, GENOMICS, AND MOLECULAR EVOLUTION

Profile and Analysis of Gene Expression Changes during Early Development in Germinating Spores of Ceratopteris richardii^1,[w]

Molecular Cell and Developmental Biology, University of Texas, Austin, Texas 78751

	ABSTRACT

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 
Analysis of an expressed sequence tag library with more than 5,000 sequences from spores of the fern Ceratopteris richardii reveals that more than 3,900 of them represent distinct genes, and almost 70% of these have significant similarity to Arabidopsis (Arabidopsis thaliana) genes. Eight genes are common between three very different dormant plant systems, Ceratopteris spores, Arabidopsis seeds, and Arabidopsis pollen. We evaluated the pattern of mRNA abundance over the first 48 h of spore development using a microarray of cDNAs representing 3,207 distinct genes of C. richardii and determined the relative levels of RNA abundance for 3,143 of these genes using a Bayesian method of statistical analysis. More than 900 of them (29%) show a significant change between any of the five time points analyzed, and these have been annotated based on their sequence similarity with the Arabidopsis proteome. Novel data arising from these analyses identify genes likely to be critical for the germination and subsequent early development of diverse cells and tissues emerging from dormancy.

Ceratopteris has been used to study such diverse processes as sex determination and differentiation (Wen et al., 1999), hormone responses (Hou et al., 2004), photomorphogenesis (Kamachi et al., 2004), and gravity-directed polar development (Edwards and Roux, 1998; Chatterjee et al., 2000). Imbibed spores remain dormant until exposed to light, at which time they begin a documented series of developmental steps, including the production of a detectable polar calcium current that peaks 6 h after light exposure (Chatterjee et al., 2000), migration of the nucleus 24 h after light exposure, a polar cell division 48 h after light exposure, and subsequent primary rhizoid emergence in the direction determined by the nuclear migration 72 h after light exposure. As demonstrated by Edwards and Roux (1998), the spores fix the direction of their nuclear migration and primary rhizoid emergence as a response to gravity in the 6 to 18 h after light exposure. Here, we provide an initial analysis of the mRNA expression that accompanies these physiological changes.

Partial sequencing of cDNA clones as expressed sequence tags (ESTs) is an alternative to more extensive genome sequencing efforts. Large-scale EST sequencing projects (>40,000 ESTs) have been described in a number of plant systems, including maize (Zea mays), tomato (Lycopersicon esculentum), moss (Physcomitrella patens), soybean (Glycine max), and clover (Trifolium repens; Fernandes et al., 2002; Rensing et al., 2002; Shoemaker et al., 2002; Van der Hoeven et al., 2002; Sawbridge et al., 2003).When cDNA clones are randomly selected for sequencing, the abundance of ESTs for the same gene is related to the expression level of that gene in the cDNA library. This relationship enables expression analysis of genes within a sample or comparisons between libraries constructed from biologically distinct samples using only the EST abundance data generated from sequencing and analysis (Fernandes et al., 2002). Additionally, ESTs are useful for other expression analysis techniques, including cDNA microarrays, which rely upon known sequence information.

DNA microarray technology has provided a method of monitoring the expression profiles of almost any biological system during developmental changes (Hennig et al., 2004), under conditions of stress response (Schenk et al., 2000; Marin et al., 2003), and as a means of mutant analysis (Mandaokar et al., 2003). Until now, researchers have used microarrays to determine the mRNA complement of Arabidopsis seed (Girke et al., 2000) and Arabidopsis pollen (Honys and Twell, 2003), but there has been no analysis of the expression changes occurring during development of these systems. This study evaluates and documents the gene expression changes that occur during the emergence from dormancy and early development of a fern spore, using thoroughly replicated microarray data analyzed by a Bayesian model for determining statistically significant changes. We present data from 34 array hybridizations that include at least eight replications of four different developmental time point comparisons. Analysis of these data reveals more than 900 genes with relative changes in transcript abundance over the first 48 h of spore development. Several changes in expression observed in microarray analysis have been independently verified by real-time quantitative reverse transcription (RT)-PCR.

	RESULTS

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

Functional and Compartmental Categorization of the C. richardii Tentative Unique Genes

Clustering and assembly of the ESTs yielded 3,930 tentative unique genes (TUGs), composed of 513 contigs and 3,417 singletons. Contigs are consensus sequences generated from two or more ESTs that are determined to represent the same expressed gene, while singletons are ESTs with no strict similarity to other ESTs in the collection.

The estimate of the total number of genes being expressed in spores 20 h after light initiation varies slightly depending on the sample size used in the calculation. Over the range of sample sizes we used (a single 384-well sequencing plate, half a plate [192 wells], or two plates [768 wells]), we estimate there to be between 14,317 and 15,297 unique genes expressed 20 h after light initiation of spore germination. This estimate indicates that the current 3,930 TUGs represent approximately 25.7% to 27.6% of those sequences.

The TUGs were identified by BLAST analysis against the Arabidopsis proteome, yielding 2,710 TUGs with significant similarity (E value 1.0 x 10^–10) to Arabidopsis proteins. Using the Gene Ontology terms of the matching Arabidopsis loci (Berardini et al., 2004), functional and localization assignments were made for each of these C. richardii TUGs, and only genes with an assignable function or subcellular localization are presented. In order to determine what may represent typical functional and compartmental distributions, as well as provide a basis for comparison, genetic loci expressed in Arabidopsis seed, pollen, and leaf tissue were similarly analyzed.

The functional expression patterns seen in C. richardii spores were generally similar to those found in various Arabidopsis tissues previously sampled (Fig. 1). In each of the sets of loci, the broad categories of metabolism and protein metabolism were the most abundant, accounting for more than 45% of the genes with assignable functions (Fig. 1). The compartmental distribution of genes was more variable between the Ceratopteris EST collection and the Arabidopsis tissues (Fig. 2). The collection of genes with their localization classified as other membranes shows the largest difference: C. richardii spores had a proportion 5% to 10% smaller than that typically seen in Arabidopsis. The other membranes compartmental category includes membrane proteins, excluding those that localize to the plasma membrane. The only other difference between the C. richardii spores and the three Arabidopsis tissues occurs in genes associated with the ribosome, which occurred 1.5- to 4-fold more frequently in the spores than in the Arabidopsis tissues (Fig. 2).

View larger version (28K): [in this window] [in a new window]   Figure 1. Functional classification of gene products expressed in C. richardii spores. Ceratopteris TUGs were annotated by BLAST comparison with the Arabidopsis proteome, and the functional classification of each TUG was done according to The Arabidopsis Information Resource (TAIR) Gene Ontology database of the resulting best BLAST match. The functional classification distribution of genes expressed in Arabidopsis seed, pollen, and leaf tissue is also indicated.

View larger version (24K): [in this window] [in a new window]   Figure 2. Localization of gene products expressed in C. richardii spores. Ceratopteris TUGs were annotated by BLAST comparison with the Arabidopsis proteome, and the compartmental classification of each TUG was done according to the TAIR Gene Ontology database of the resulting best BLAST match. The compartmental distribution of genes expressed in Arabidopsis seed, pollen, and leaf tissue is also indicated.

Ceratopteris spores share similar biological and physiological characteristics with Arabidopsis pollen and seeds. Therefore, we examined which genes were expressed in all three developmental stages of these plant model organisms. In order to limit the comparison to genes that show relatively specific patterns of expression rather than genes that are broadly expressed, we first screened each of these gene sets with a set of vegetatively expressed genes derived from analysis of more than 32,000 ESTs from Arabidopsis roots, shoots, and leaves. Of the genes included in the seed, spore, or pollen sets, 50% to 60% are also expressed in vegetative tissues (Fig. 3). The Arabidopsis genes exhibiting pollen- and seed-specific expression were then compared with the genes expressed in spores to determine which genes are shared. Nearly 9% of the genes expressed in seeds or pollen are also expressed in C. richardii spores (Fig. 3; Table I), and eight genes are expressed in all three tissues (Table I). Not surprisingly, the number of times an EST for a particular gene was found in the seed and spore libraries often varied (Table I). For example, in Arabidopsis Genome Initiative (AGI) number AT4G25650, four ESTs were found in Arabidopsis seeds and two ESTs were found in C. richardii spores. Note that pollen expression was based on oligonucleotide arrays; therefore, no EST data are available.

View larger version (34K): [in this window] [in a new window]   Figure 3. Comparison of the C. richardii spore TUGs to genes expressed in Arabidopsis seeds and pollen. A, Identification of tissue-specific gene expression in Ceratopteris spores and Arabidopsis seeds and pollen. The proportion of genes present in the seed, spore, and pollen sets that were also expressed in vegetative tissue is indicated in light gray. The remaining tissue-specific genes, indicated in dark gray, are used for the expression comparisons. Numbers indicate the number of genes included in each proportion. B, Overlapping expression of genes expressed in seeds, spores, and pollen. Numbers in overlapping regions indicate number of genes shared between the respective sets. Several relatively abundant genes present in each overlap are also indicated.

Early Developmental Transcription Profile in C. richardii Spores

To analyze changes in gene expression during early spore development, we did four pair-wise time point comparisons with a minimum of eight replications for each comparison; 0 h versus 24 h, 6 h versus 24 h, 12 h versus 24 h, and 48 h versus 24 h. At least five different total RNA samples from each time point were used to generate probes for these comparisons. The supplemental data provide complete data that adhere to Minimum Information About a Microarray Experiment (MIAME) standards (Brazma et al., 2001) for all arrays included in this study.

Relative abundance of transcripts at 0, 6, 12, 24, and 48 h after light exposures was determined using Bayesian Analysis of Gene Expression Level (BAGEL) software (Townsend and Hartl, 2002). For statistical analysis by BAGEL software, an array spot must meet the quality criteria described in methods for a minimum of three array data sets for each of the four time point comparisons. Seventy-two spots were omitted from analysis because of this type of insufficient replication. Because of the comparative, quantitative nature of microarray analysis, BAGEL software presents the treatment (or developmental time point in this study) with the lowest expression level with the value one and presents all other treatments as fold increases over one. This analysis also provides 95% credible interval range for all features at all treatment conditions (time points) included.

Changes in transcript abundance are defined as nonoverlapping 95% credible interval between any two time points. Of the TUGs analyzed, 70% showed no significant change at any time point over the first 48 h of development. Altogether, 922 TUGs (29%) showed a significant difference between at least two of the developmental time points analyzed (Supplemental Table I). Of TUGs that showed a change in transcript abundance, 138 (15%) were significantly more abundant 48 h after light exposure than upon germination initiation by light or were up-regulated during the first 48 h of development. Included in that list are 35 TUGs that were significantly more abundant 48 h after light exposure than at all other developmental time points analyzed (Table II). Altogether, 203 TUGs (22%) were more abundant at the time of initial light exposure (0 h) than 48 h after light exposure, or they were down-regulated over the first 48 h of development.

View larger version (22K): [in this window] [in a new window]   Figure 4. Specific expression patterns of C. richardii TUGs. Sequences are identified by accession number. The identity of the best BLASTX match to the Ceratopteris sequence is included. Times after initial light exposure tested were 0, 6, 12, 24, and 48 h. Transcript abundance at the time point with the lowest mean expression level is presented as 1, and all other time points are relative to 1. Error bars represent 95% credible interval for relative expression levels. Significant difference in RNA abundance between time points is defined as nonoverlapping 95% credible interval. Ceratopteris accession numbers are as follows: A, Mago nashi-1 (BE641602), Mago nashi-2 (BE642256), and SIN-like family protein (BE643380); B, Ras-related GTP-binding protein (BE642932), NPH3 family protein (BE641485), and catalase 1 (BE642028).

The TUG encoding a protein with high sequence similarity (E value of 1 x 10^–56) to Ras-related GTP-binding protein in Arabidopsis shows steady expression throughout the first 24 h of development, with no significant difference in message abundance between any of these time points, followed by a doubling in abundance 48 h after initial light exposure. Abundance of the mRNA for a protein with high sequence similarity (E value of 3 x 10^–23) to the NPH3 family protein from Arabidopsis is 2-fold higher at the 0 h time point than 48 h after light exposure, with a steady decrease in abundance between the extreme points in development that were analyzed. Lastly, the TUG that encodes a protein with high sequence similarity (E value of 8 x 10^–96) to catalase chain 1 of upland cotton (Gossypium hirsutum), as well as strong similarity to catalase genes from several other plants, is significantly up-regulated between 0 and 48 h after light exposure.

Verification of Microarray Patterns by Quantitative Real-Time RT-PCR

As an independent confirmation of RNA expression patterns, we performed real-time RT-PCR on six genes showing significant expression changes. Message levels for Ceratopteris sequences with the accession numbers BE642028, BE642763, BE642932, BE643392, BQ087159, and BE642674 were compared to those of two control genes, BQ086953 (-tubulin) or BE640734 (adenine phosphoribosyltransferase form 1 [APT1]). RNA was isolated in the same manner as samples used for microarray experiments. Melting curve analysis showed discrete peaks for all samples, indicating amplification of single targets. Lack of detectable signals in mock RT controls demonstrated absence of genomic DNA contamination (data not shown). Expression pattern comparisons between microarray and real-time RT-PCR are presented (Supplemental Fig. 1). General expression trends were corroborated by at least one control gene comparison. Specific fold changes were reasonably close to those estimated by the microarray results considering the longer linear range of sensitivity for real-time RT-PCR.

	DISCUSSION

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

Comparison of Gene Expression in Spores, Seeds, and Pollen

While the number of ESTs analyzed in this project is similar in scale to that of other libraries of specific tissues and developmental stages (Fernandes et al., 2002; Shoemaker et al., 2002; Sawbridge et al., 2003), this library is unique because its biological source is a single cell and represents a plant gametophytic generation. The first cell division in C. richardii typically takes place 48 h after light initiation of germination, approximately 24 h after the gravity-directed downward migration of the spore nucleus. Because the cDNA used for EST sequencing is based on RNA collected 20 h after initiation of germination, this collection of ESTs represents C. richardii gene expression at a single-celled stage of development.

Spores at this stage of development are transitioning from a dormant, desiccated state to a metabolically active one, analogous to the process of seed germination in angiosperms. This similarity may be more than superficial, as both processes appear to involve the relatively specific expression of similar proteins, including those related to desiccation and dormancy, as well as several Asp or Cys proteases (Table I). In developing seeds, Asp and Cys proteases process storage proteins into mature subunits upon their transport to specialized protein storage vacuoles. Proteases are also involved in seed germination to initiate and complete metabolism of the storage proteins (Muntz, 1996; Gruis et al., 2002, 2004). Because transport to the protein storage vacuole is critical, the shared expression of a seed-specific vacuolar processing enzyme may be an indicator that a similar protein storage strategy operates in both seeds and spores. It has been hypothesized that seed plants could have coopted genes used for spore dormancy to develop post-embryonic seed dormancy (Banks, 1999). The common, tissue-specific expression of genes in both Ceratopteris spores and Arabidopsis seeds supports this hypothesis.

Biologically, fern spores are part of the haploid gametophytic generation and are equivalent to angiosperm microspores and megaspores, which divide to produce mature pollen grains and embryo sacs, respectively. Using an oligonucleotide array, Honys and Twell (2003) identified approximately 1,000 genes expressed in Arabidopsis pollen and estimated that the total number of pollen-expressed genes is approximately 3,500. The serial analysis of gene expression approach (Lee and Lee, 2003) predicts a similar number for Arabidopsis pollen. Estimates from our EST analysis place the number of genes expressed in germinating C. richardii spores at more than 14,000. The substantial difference in estimated gene diversity between these two relatively simple germinating systems may reflect the relative physiological complexity of the free-living photosynthetic gametophytes in homosporous ferns as compared to the reduced, nonphotosynthetic gametophytes of angiosperms.

The unique group of genes found in Arabidopsis seeds, Arabidopsis pollen, and Ceratopteris spores, but not included in cDNA libraries from Arabidopsis shoot, root, or leaf (Fig. 1; Table I), can illuminate the physiological processes shared by these three stages; namely, maintenance of and emergence from a desiccated, metabolically dormant state. Due to the specificity of the comparisons carried out, certain expected commonalities may be missing. While several of these genes have been previously implicated in processes related to germination, to our knowledge, the presence and functional roles of these genes in all three of these unique systems have not been commented on before.

The shared expression of a eukaryotic translation initiation factor, locus AT1G54290, in Arabidopsis seed, pollen, and Ceratopteris spores comes as no surprise. It is well documented that inhibition of translation by treatment with the drug cycloheximide inhibits pollen germination (Fernando et al., 2001; Metcalf et al., 2004) as well as seed germination (Rajjou et al., 2004). We have also found that spores treated continuously with cycloheximide do not germinate (S.C. Stout and S.J. Roux, unpublished data).

These three systems are poised for extensive and rapid growth through cell division and/or cell expansion; therefore, the inclusion of a subunit of the mitochondrial NADH-ubiquinone oxidoreductoase complex (Heazlewood et al., 2003), locus AT2G02050, in these three systems is also not surprising. In keeping with the theme of increased cellular metabolism, locus AT2G17370 encodes HMG-CoA reductase 2, another gene found in all three systems. This enzyme catalyzes the synthesis of mevalonate, a precursor of plant isoprenoids, which are a diverse group of compounds including the hormones abscisic acid and gibberellins, sterols, components of the electron transport chain, and some plant defense agents (Enjuto et al., 1994).

Full-length cDNA sequence similarity identifies the Arabidopsis locus AT2G25110 as an MIR domain-containing protein, similar to stromal cell-derived factor 2 precursor of Homo sapiens. The stromal cell-derived factor 2, a secreted protein, acts as a chemoattractant in mammalian immune system cells (Hamada et al., 1996), and this report of its expression in Arabidopsis seed, pollen, and Ceratopteris spore is, to our knowledge, the first indication of a potential role in plants.

SNAP- and SNARE-type targeting protein systems control membrane trafficking in most eukaryotes. These systems have multiple small peptide components that provide specificity, and the Arabidopsis genome contains several of these genes, including 14 synaptobrevin, or VAMP, family proteins (Sanderfoot et al., 2000). Vesicular trafficking and vacuolar sorting are processes important in seed germination and pollen tube elongation, and we have identified one synaptobrevin family gene, locus AT2G32670, that has shared expression in seed and pollen and a likely homolog expressed in spores. This shared expression may help in identifying the specific roles of various SNAP/SNARE targeting components in plant growth and development.

The TUG found in Ceratopteris spores that is similar to the Arabidopsis gene AT4G25650 found in seed and pollen is described as a Rieske (2Fe-2S) domain-containing protein, similar to cell death suppressor protein lethal leaf spot 1 (Lls1) from maize. The gene, Lls1 from maize, was originally described as encoding a novel protein highly conserved in plants that functions as a cell death suppressor (Gray et al., 1997). Recently, it was determined that the Lls1 and its known ortholog in Arabidopsis, accelerated cell death 1 (Yang et al., 2004), are both genes that encode pheophorbide a oxygenase. Pheophorbide a oxygenase is an enzyme necessary for the catabolism of chlorophyll b and removal of a phototoxic intermediate in the pathway of chlorophyll degradation (Pruzinska et al., 2003). This gene is highly conserved among land plants and its discovery in several cyanobacteria (Gray et al., 2004) predicts the presence of a similar gene in Ceratopteris. What may seem unlikely is the expression of a gene involved in chlorophyll catabolism in the nonphotosynthetic spore, as well as pollen and seed of Arabidopsis. This is parallel to the finding of low level expression of Lls1 in nonphotosynthetic maize embryos, endosperm, and roots and has been explained by the postulate that all plant cells have an ability to degrade chlorophyll, regardless of the presence of chlorophyll in the cell (Yang et al., 2004).

Research has shown the importance of the enzymes located in the peroxisome for lipid metabolism and reactive oxygen species scavenging (for review, see Palma et al., 2002; Titorenko and Rachubinski, 2004). Recently, the works of Murgia and Hu have implicated peroxisome-mediated signaling in the nitric oxide pathway of stress responses (Murgia et al., 2004) as well as in the photomorphogenesis and development of seedlings (Hu et al., 2002). The Arabidopsis gene COMATOSE is a homolog of the human adrenoleukodystrophy protein that is involved in transport of very long chain fatty acids into peroxisomes, and this gene has been implicated specifically in seed germination (Footitt et al., 2002). The presence of peroxisomal targeting signal type 1 (PTS1) receptor gene (PEX5) in seed, pollen, and Ceratopteris spores provides evidence of the importance of peroxisomal activity in the process of emerging from a desiccated dormant state that is common to all three (Table I). The Arabidopsis proteome contains almost 200 genes with PTS1 targeting signals for the peroxisome (Kamada et al., 2003). In tobacco (Nicotiana tabacum), yeast (Saccharomyces cerevisiae), and humans, PEX5 serves as part of the receptor of proteins bound for the peroxisome that contain the PTS1 motif (Kragler et al., 1998). Catalase is an enzyme found in the peroxisome that breaks down hydrogen peroxide very efficiently. A PTS1 signal may target the enzyme to the peroxisome (Kamigaki et al., 2003). Further evidence for the importance of peroxisome activity in germinating systems is the significant up-regulation of a Ceratopteris catalase TUG (Fig. 4B) in the early development of germinating spores.

Analysis of Early Development Expression Profiles

The majority of TUGs analyzed, around 70%, show no credible change between any two time points of development analyzed. The remaining 922 TUGs that show a significant change between at least two time points (Supplemental Table I) were analyzed for trends in expression pattern relevant to the physiological processes of developing spores.

Around 48 h after light exposure, the single cell of the spore undergoes its first division. The 138 TUGs that are up-regulated between initial light exposure and 48 h later may be involved in the spore emerging from its dormant state or preparing for and undergoing this cell division. There is a subset of this group of TUGs that shows a unique expression pattern: 35 TUGs (Table II) show no change in transcript abundance over the first 24 h of development, but their expression level 48 h after light exposure is significantly higher than at all other time points, suggesting that they may be especially important for spore cell division.

One TUG that shows this pattern of change in abundance has strong similarity (E value of 1 x 10^–56) to Arabidopsis Ras-related GTP-binding protein (Fig. 4B). Arabidopsis G-subunit and G protein coupled receptor play a role in the cell proliferation that occurs during seed germination (Jones and Assmann, 2004). The role of G protein signaling in plant cell division and proliferation may not be restricted to angiosperms but may be a fundamental signaling pathway common to all plants as suggested by this expression pattern.

Another interesting pattern of expression includes genes that are possibly involved in maintaining and breaking the dormancy of the spore; 203 TUGs that are significantly more abundant when the spores are first exposed to light (0 h) than 48 h later. This is almost twice the number of TUGs that are significantly up-regulated over the first 48 h of development. The emergence from dormancy of fern spores is a light-activated, phytochrome-mediated response (Cooke et al., 1987). One TUG with this expression pattern has strong sequence similarity (E value of 3 x 10^–23) to the NPH3 family protein from Arabidopsis that is critical for phototropism (Fig. 4B). Several other TUGs with this expression pattern show strong similarity to signal transduction pathway elements and should be studied for their roles in the process of breaking dormancy and germination.

RNA Localization in Spore Polarity Development

C. richardii spores determine the polarity of their subsequent development as a response to the vector of gravity some time between 6 and 18 h after their initial exposure to light. For this reason, genes that undergo changes in transcript abundance during this period of development are of particular interest. The mechanisms by which plant cells determine developmental polarity and cell fate are only beginning to be unraveled (Vroemen et al., 1999; Cove, 2000; Grebe et al., 2001).

The Drosophila melanogaster oocyte is a model system for studying the establishment of cell polarity with well-characterized molecular components. Among these molecular components is a protein called Mago nashi, which appears to be highly conserved across kingdoms (Swidzinski et al., 2001). In Drosophila and Caenorhabditis elegans, Mago nashi family proteins function as structural components of the spliceosome that are important in oogenesis. In Drosophila, Mago nashi protein is necessary for proper localization of at least one other mRNA, Oskar, and the proper localization of this mRNA during polar development is required for normal oocyte development (Micklem et al., 1997; Hachet and Ephrussi, 2004). An Arabidopsis analog of Mago nashi is found in both seed and pollen (Table I), and two distinct Ceratopteris TUGs with significant (1 x 10^–77) similarity to a previously described Mago nashi protein in the fern M. vestita were included in this EST library. One of the Mago nashi TUGs in developing spores (Fig. 4A) is up-regulated early in development, with its peak transcript abundance 12 h after light exposure, followed by steadily decreasing transcript levels at 24 and 48 h. The other Ceratopteris Mago nashi TUG shows a steady increase in transcript abundance over the first 48 h of development, supporting our distinction of this as a separate Ceratopteris gene and indicating that these two structurally similar proteins may be involved in different developmental processes (Fig. 4A). Imaging the subcellular localization of Mago nashi mRNA could provide additional evidence for its role in spore polar development, but available techniques for in situ localization are particularly problematic in Ceratopteris spores because of their high oil content. The role of RNA localization during plant embryogenesis was recently reviewed by Okita and Choi (2002).

The proposed Arabidopsis Dicer homolog, SIN1/SUS1/CAF, which is essential for embryogenesis (Golden et al., 2002), demonstrates another important role for posttranscriptional modifications in the early development of plants. A Ceratopteris TUG that is significantly up-regulated during the period of polarity determination in spores shows significant sequence similarity (E value of 7 x 10^–10) to an Arabidopsis SIN-like family protein (Fig. 4A). The involvement of SIN-like and Mago nashi-like genes in animal oocyte polar development, the specific expression of Mago nashi RNA in Arabidopsis seed and pollen, and the pattern of Mago nashi- and SIN-like transcript changes in spore germination suggests a role for RNA splicing and localization in polar development of cells in plants as well as animals.

	MATERIALS AND METHODS

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

Plant Material and Experimental Conditions

Spores of the fern Ceratopteris richardii of an inbred diploid strain designated Hn-n were surface sterilized as described in Edwards and Roux (1998). Spores were allowed to soak in sterile water in complete darkness at 29°C for 4 to 7 d to enhance synchronization of germination. After soaking, the water was removed, and spores were sown in a solution of half-strength Murashige and Skoog basal medium (Sigma-Aldrich, St. Louis), pH 6.3, and 0.5% agarose cooled to a temperature below 55°C. A concentration of 1 g of spores in 20 mL of media was used in this study, and aliquots of 5 mL were placed into 100- x 15-mm petri plates, allowed to solidify, and maintained in a fixed orientation. Spores were exposed to continuous white light and kept at 29°C for varying times as specified. Plates were then sealed and stored at –80°C for RNA isolation.

Total RNA Isolation

Frozen plates of spores in agar were removed and quickly ground to homogeneity with a mortar and pestle. An equal volume (5 mL) of buffer (1 M Tris, pH 7.3, 5 mM EDTA, pH 8.0, and 1% SDS) was added with two volumes (10 mL) of acidic phenol:chloroform:IAA (Fisher Scientific, Hampton, NH). The mixture was vortexed and distributed into 12 to 18 1.5-mL epi-tubes. One 3.2-mm stainless steel or tungsten bead was added to each tube, and a Mixer Mill 300 (Retsch, Haan, Germany) was used at top speed (30 oscillations/s) for 5-min intervals two times to disrupt spores. Homogenate was examined under a light microscope to verify that spores were broken open. This mixture was immediately centrifuged at 15000 rpm for 15 min at 15°C. The aqueous layer was removed and extracted with chloroform. The aqueous layer was again removed and one-tenth volume of 3 M sodium acetate and 2.5 volumes of 95% ethanol were added, and the solution was allowed to precipitate overnight at –80°C. The samples were then centrifuged at 1500 rpm for 15 min, and all pellets from one plate of spores were resuspended and combined in diethyl pyrocarbonate-treated water. Resuspended nucleic acid pellets were then treated with Amplification Grade Deoxyribonuclease I (Invitrogen, Carlsbad, CA) following the manufacturer's protocol and ethanol precipitated as described above. The final RNA pellets were resuspended in 50 µL of diethyl pyrocarbonate-treated water, and the RNA concentration was determined spectrophotometrically. The integrity of the RNA samples was verified by electrophoresis on 1.2% denaturing gels (northernMax-Gly denaturing gel; Ambion, Austin, TX).

cDNA Library Construction and EST Sequencing

RNA isolated from spores 20 h after light initiation of germination (approximately 24 h before the first cell division) was used for a commercially prepared cDNA library (Life Technologies, Rockville, MD). Randomly chosen clones were sequenced at the Purdue Agricultural Genomics Facility (Purdue University, West Lafayette, IN), and 5,085 of the resulting singe-pass sequences were used for further analysis. ESTs have been submitted to the NCBI EST database (GenBank accession nos. BE640669–BE643506, BQ086920–BQ087668, and CV734654–CV736151).

EST Assembly

ESTs were filtered for short entries or low complexity sequences using SeqClean (http://www.tigr.org/tdb/tgi/software). The resulting sequences were assembled with The Institute for Genomic Research Gene Indices clustering tools (TGICL; Pertea et al., 2003). Briefly, TGICL clusters all of the ESTs using minimum overlap length (40 bp) and percentage identity (95%) criteria. The initial clusters were then sent to an assembly program (CAP3; Huang and Madan, 1999) that attempts to create one or more contigs from each cluster.

The resulting sequences in the data set are termed TUGs and consist of two types of sequences: contigs and singletons. Contigs are two or more ESTs that are presumed to represent the same transcript, and singletons are ESTs without significant similarity to any other ESTs. The set of TUGs, composed of the sets of contigs and singletons, represents the unique genes found in the EST collection.

TUG Identification and Functional Analysis

The identities of the TUGs were determined using BLASTX (Altschul et al., 1997) against the Arabidopsis (Arabidopsis thaliana) proteome (ATH1_pep_cm_20040228; http://www.arabidopsis.org). Functional and localization categories of the C. richardii TUGs were assigned using the TAIR Gene Ontology terms associated with the locus of the best Arabidopsis BLAST match (http://www.arabidopsis.org).

An estimate of the total number of unique genes expressed in C. richardii spores was made using a nonparametric estimator typically used for the estimation of population size or species richness in ecological studies (Burnham and Overton, 1979; Brose et al., 2003). This calculation makes use of the number of random samples in which each species appears. In our studies, we varied the sample size as a single 384-well sequencing plate, half a plate (192 wells), and two plates (768 wells). For each TUG, the number of samples in which it occurred was tabulated from the number of plates containing its constituent ESTs. Given an expected percent coverage of approximately 25% to 30%, the fourth-order jackknife estimator was used (Brose et al., 2003), and the final percentage of coverage of expressed genes was determined by dividing the number of TUGs by the estimated number of unique genes.

Arabidopsis Tissue-Specific Genes

EST collections from Arabidopsis seed, leaf, root, and shoot cDNA libraries were downloaded from The Institute for Genomic Research (www.tigr.org). The ESTs were pooled together and analyzed as above for production of clusters and singletons. The resulting TUGs were identified by BLAST analysis against the Arabidopsis transcriptome (ATH1_cDNA_cm_20040228; http://www.arabidopsis.org). Only TUGs with transcript matches longer than 100 bp and greater than 97% identity were retained. A set of Arabidopsis genes expressed in pollen was also obtained from Honys and Twell (2003).

RT-PCR

Primers were generated based on sequence data available at NCBI for the following ESTs: BE641602, BE642746, BQ087334, BE641661, BE642715, BE642350, CV734685, and BE642120 (supplemental data). One microgram of total RNA from dry spores, 20-h spores, 14-d-old gametophytes, and mature sporophytes was used for reverse transcription with oligo(dT)₂₂ primer and SuperScript II reverse transcriptase (Invitrogen) following the manufacturer's protocol. RNA was treated with Amplification Grade Deoxyribonuclease I (Invitrogen) following the manufacturer's protocol just prior to reverse transcription reaction. Twenty-microliter RT reactions were diluted 1:4 with nuclease-free water, and 10 µL of dilute RT was used as template in 25-µL PCR reactions with Taq PCR Master Mix (Qiagen, Valencia, CA).

Microarray Construction

Spotted cDNA microarrays were printed following protocols by Childs et al. (2003). The 3,840 cDNA clones corresponding to the EST library described by Stout et al. (2003) and deposited in the EST database (GenBank accession nos. BE640669–BE643506 and BQ086920–BQ087668) were amplified by PCR with SP6 and T7 vector-specific primers in 96-well plate format. Amplified cDNA from eight yeast clones with no significant sequence similarity to known C. richardii genes were included in arrays to serve as nonspecific binding controls. Sheared and unsheared yeast genomic DNA and C. richardii genomic DNA were also included as nonspecific binding spots.

Poly-L-lysine-coated slides were produced following the protocol of Childs et al. (2003), available at www.microarrays.org. Thirty-two tips in 4 x 8 configuration were used with Array Maker 2.4 software (http://derisilab.ucsf.edu/arraymaker.shtml) for array printing. Arrays were rehydrated, blocked, and post-processed following the protocol of Childs et al. (2003) and used within 3 weeks of post-processing.

Fluorescent Probe Synthesis and Data Acquisition

Amino-allyl dUTP Cy3- and Cy5-labeled microarray probes were synthesized following protocols adapted from DeRisi (2003). Equal quantities of total RNA (15 to 30 µg) from two time points were used for each array probe. Reverse transcription with oligo(dT)₂₂ primer and SuperScript II reverse transcriptase (Invitrogen) with amino-allyl dUTP (Ambion) were used to generate first strand cDNA. The cDNA sample from each time point was conjugated to one of Cy3 or Cy5 dyes (Amersham Biosciences, Buckinghamshire, UK), and dye-swapped probes were included in this study. Two microliters of 10 mg/mL sheared herring sperm DNA was included in each 50-µL probe for nonspecific binding. The probe was applied to an array with 22 x 401 LifterSlip cover slides (Erie Scientific, Portsmouth, NH) and allowed to hybridize for approximately 6 h at 65°C in humid array hybridization chambers (Corning, Corning, NY). Detailed descriptions of each array hybridization included in this study can be found in the supplemental data.

Immediately following hybridization, the chambers were disassembled and arrays washed according to the method at http://chipmunk.icmb.utexas.edu/ilcrc/protocols/index.shtml. In some cases, Dye Saver 1 (Genisphere, Hatfield, PA) was used prior to scanning to prevent unequal degradation of the two dyes. Arrays were scanned following the protocol at http://chipmunk.icmb.utexas.edu/ilcrc/protocols/Scanning.pdf using an Axon 4000 scanner (Molecular Devices, Union City, CA) and Axon GenePix Pro 4.1 or 5.1 software. The automatic flagging feature to identify spots not found was used. After manually grinding arrays to correct for spot identification errors in the automatic spot location feature of genepix, array images, settings, and results files were uploaded into the Longhorn Array Database (Killion et al., 2003) and are available in the supplemental data. In order to compare separate array hybridizations, normalization of arrays was calculated based on median log ratio equal to zero. The only R/G normalized mean values of spots that were retrieved and compiled for analysis using BAGEL software (Townsend and Hartl, 2002) were those that met the following criteria: spots must have been unflagged by genepix, must have a minimum of 50 spot pixels, and must have at least 50% of green- or red-spot pixels greater then background intensity plus two SDs. Any spots not meeting these criteria were omitted from further analysis.

Quantitative Real-Time RT-PCR

Total RNA was isolated and handled as above, with the exception that DNase treatment of the RNA was carried out just prior to the reverse transcription step as opposed to treatment of the entire RNA sample directly after it was isolated. One microgram of RNA from 0, 6, 12, and 48 h time points was reverse transcribed according to the manufacturer's instructions with oligo(dT)₂₂ primer and SuperScript II reverse transcriptase (Invitrogen) to generate first strand cDNA.

Quantitative real-time RT-PCR was performed on six TUGs, and expression changes were compared to microarray analysis. LUX fluorescent primers were designed using Invitrogen's Web-based LUX Designer software (http://www.invitrogen.com/content.cfm?pageid=3978#PrimerDesign) based on the EST sequences of BE640734 (APT1), BQ086953 (-tubulin), BE643392, BE642763, BE642932, BE642028, BQ087159, and BE642674. APT1 and -tubulin were chosen as control genes because of their general use as such in other systems, their lack of significant expression changes in microarray analysis, and their minimal sequence similarity to other Ceratopteris ESTs. Control gene primers were labeled with JOE, while all experimental genes were labeled with FAM, and primer sequences are available (Supplemental Table II). PCR reactions were performed in 96-well polypropylene microplates using Platinum Quantitative PCR SuperMix-UDG (Invitrogen) according to the Invitrogen cycling programs and protocols at one-half final volumes (25 µL). A final concentration of 50 nM of appropriate gene-specific primers and the equivalent of 100 ng of reverse-transcribed RNA were used per reaction. PCR and fluorescence measurements were carried out with the ABI Prism 7900HT sequence detection system (Applied Biosystems, Foster, CA) as an absolute quantification run, and initial data analysis was done with the manufacturer-supplied SDS 2.2 software. Single target amplification of all samples and absence of genomic DNA in mock RT controls was verified via dissociation curve analysis. Fold expression changes were calculated using the comparative C_T method (User Bulletin 2; ABI Prism 7700 sequence detection system) in Microsoft Excel. For comparison with microarray BAGEL expression patterns, real-time RT-PCR expression change values were normalized to the lowest expression value.

Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes, subject to the requisite permission from any third-party owners of all or parts of the material. Obtaining any permission will be the responsibility of the requestor.

Sequence data from this article have been deposited with the EMBL/GenBank data libraries under accession numbers BE640669 to BE643506, BQ086920 to BQ087668, and CV734654 to CV736151. A tabular summary of significant changes in gene expression based on BAGEL is provided in Supplemental Table I. Additionally, all raw microarray data from the 34 arrays included in this analysis are available upon request from the corresponding author.

	ACKNOWLEDGMENTS

 
We thank Dr. Jeff Townsend for use of and assistance with BAGEL software. We thank Dr. Alan Lloyd for suggestions on the text of this manuscript and Lane Johnson for proofreading. We appreciate Andy Alverson and all of our lab members for insightful discussion and Dr. Greg Clark for constant encouragement.

Received March 14, 2005; returned for revision April 19, 2005; accepted April 23, 2005.

	FOOTNOTES

 
¹ This work was supported in part by the National Aeronautics and Space Administration (grant nos. NAG2–1586 and NAG10–295 to S.J.R. and training grant nos. NGT5–50371 and NNG04G045H to S.C.S.).

^[w] The online version of this article contains Web-only data.

Article, publication date, and citation information can be found at www.plantphysiol.org/cgi/doi/10.1104/pp.105.062851.

^* Corresponding author; e-mail sroux{at}uts.cc.utexas.edu; fax 512–232–3402.

	LITERATURE CITED

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 
Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389–3402[Abstract/Free Full Text]

Banks JA (1999) Gametophyte development in ferns. Annu Rev Plant Physiol Plant Mol Biol 50: 163–186[CrossRef][ISI]

Berardini TZ, Mundodi S, Reiser L, Huala E, Garcia-Hernandez M, Zhang P, Mueller LA, Yoon J, Doyle A, Lander G, et al (2004) Functional annotation of the Arabidopsis genome using controlled vocabularies. Plant Physiol 135: 745–755[Abstract/Free Full Text]

Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellman P, Stoeckert C, Aach J, Ansorge W, Ball CA, Causton HC, et al (2001) Minimum information about a microarray experiment (MIAME)-toward standards for microarray data. Nat Genet 29: 365–371[CrossRef][ISI][Medline]

Brose U, Martinez ND, Williams RJ (2003) Estimating species richness: sensitivity to sample coverage and insensitivity to spatial patterns. Ecology 84: 2364–2377

Burnham KP, Overton WS (1979) Robust estimation of population size when capture probabilities vary among animals. Ecology 60: 927–936[CrossRef]

Chatterjee A, Porterfield DM, Smith PS, Roux SJ (2000) Gravity-directed calcium current in germinating spores of Ceratopteris richardii. Planta 210: 607–610[CrossRef][ISI][Medline]

Childs G, DeRisi J, Harris T, Holloway A, Hou BH, Massimi A, Murphy M, Sommerville S (2003) Printed spotted glass microarrays. In D Bowtell, J Sambrook, eds, DNA Microarrays: A Molecular Cloning Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp 68–85

Cooke T, Racusen R, Hickok L, Warne T (1987) The photocontrol of spore germination in the fern Ceratopteris richardii. Plant Cell Physiol 28: 753–759[Abstract/Free Full Text]

Cove DJ (2000) The generation and modification of cell polarity. J Exp Bot 51: 831–838[Abstract/Free Full Text]

DeRisi J (2003) Indirect fluorescent labeling of DNA with amino-allyl dyes. In D Bowtell, J Sambrook, eds, DNA Microarrays: A Molecular Cloning Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp 187–193

Edwards ES, Roux SJ (1998) Influence of gravity and light on the developmental polarity of Ceratopteris richardii fern spores. Planta 205: 553–560[CrossRef][ISI][Medline]

Enjuto M, Balcells L, Campos N, Caelles C, Arro M, Boronat A (1994) Arabidopsis thaliana contains two differentially expressed 3-hydroxy-3-methylglutaryl-CoA reductase genes, which encode microsomal forms of the enzyme. Proc Natl Acad Sci USA 91: 927–931[Abstract/Free Full Text]

Fernandes J, Brendel V, Gai X, Shailesh L, Chandler VL, Elumalai RP, Galbraith DW, Pierson EA, Walbot V (2002) Comparison of RNA expression profiles based on maize expressed sequence tag frequency analysis and micro-array hybridization. Plant Physiol 128: 896–910[Abstract/Free Full Text]

Fernando DD, Owens JN, Yu XS, Ekramoddoullah AKM (2001) RNA and protein synthesis during in vitro pollen germination and tube elongation in Pinus monticola and other conifers. Sex Plant Reprod 13: 259–264[CrossRef]

Footitt S, Slocombe SP, Larner V, Kurup S, Wu YS, Larson T, Graham I, Baker A, Holdsworth M (2002) Control of germination and lipid mobilization by COMATOSE, the Arabidopsis homologue of human ALDP. EMBO J 12: 2912–2922[CrossRef]

Girke T, Todd J, Ruuska S, White J, Benning C, Ohlrogge J (2000) Microarray analysis of developing Arabidopsis seeds. Plant Physiol 124: 1570–1581[Abstract/Free Full Text]

Golden TA, Schauer SE, Lang JD, Pien S, Mushegian AR, Grossniklaus U, Meinke DW, Ray A (2002) SHORT INTEGUMENTS1/SUSPENSOR1/CARPEL FACTORY, a Dicer homolog, is a maternal effect gene required for embryo development in Arabidopsis. Plant Physiol 130: 808–822[Abstract/Free Full Text]

Gray J, Close PS, Briggs SP, Johal GS (1997) A novel suppressor of cell death in plants encoded by the Lls1 gene of maize. Cell 89: 25–31[CrossRef][ISI][Medline]

Gray J, Wardzala E, Yang ML, Reinbothe S, Haller S, Pauli F (2004) A small family of LLS1-related non-heme oxygenases in plants with an origin amongst oxygenic photosynthesizers. Plant Mol Biol 54: 39–54[CrossRef][ISI][Medline]

Grebe M, Xu J, Scheres B (2001) Cell axiality and polarity in plants—adding pieces to the puzzle. Curr Opin Plant Biol 4: 520–526[CrossRef][ISI][Medline]

Gruis DF, Schulze J, Jung R (2004) Storage protein accumulation in the absence of vacuolar processing enzyme family of cysteine proteases. Plant Cell 16: 270–290[Abstract/Free Full Text]

Gruis DF, Selinger DA, Curran JM, Jung R (2002) Redundant proteolytic mechanisms process seed storage proteins in the absence of seed-type members of the vacuolar processing enzyme family of cysteine proteases. Plant Cell 14: 2853–2882

Hachet O, Ephrussi A (2004) Splicing of oskar RNA in the nucleus is coupled to its cytoplasmic localization. Nature 428: 959–963[CrossRef][Medline]

Hamada T, Tashiro K, Tada H, Inazawa J, Shirozu M, Shibahara K, Nakamura T, Martina N, Nakano T, Honjo T (1996) Isolation and characterization of a novel secretory protein, stromal cell-derived factor-2 (SDF-2) using the signal sequence trap method. Gene 176: 211–214[CrossRef][ISI][Medline]

Heazlewood JL, Howell KA, Millar AH (2003) Mitochondrial complex I from Arabidopsis and rice: orthologs of mammalian and fungal components coupled with plant-specific subunits. Biochim Biophys Acta 1604: 159–169[Medline]

Hennig L, Gruissem W, Grossniklaus U, Köhler C (2004) Transcriptional programs of early reproductive stages in Arabidopsis. Plant Physiol 135: 1765–1775[Abstract/Free Full Text]

Honys D, Twell D (2003) Comparative analysis of the Arabidopsis pollen transcriptome. Plant Physiol 132: 640–652[Abstract/Free Full Text]

Hou G, Hill JP, Blancaflor EB (2004) Developmental anatomy and auxin response of lateral root formation in Ceratopteris richardii. J Exp Bot 55: 685–693[Abstract/Free Full Text]

Hu JP, Aguirre M, Peto C, Alonso J, Ecker J, Chory J (2002) A role for peroxisomes in photomorphogenesis and development of Arabidopsis. Science 297: 405–409[Abstract/Free Full Text]

Huang X, Madan A (1999) CAP3: a DNA sequence assembly program. Genome Res 9: 868–877[Abstract/Free Full Text]

Jones AM, Assmann SM (2004) Plants: the latest model system for G-protein research. EMBO Rep 5: 572–578[CrossRef][ISI][Medline]

Kamachi H, Matsunaga E, Noguchi M, Inoue H (2004) Novel mutant phenotypes of a dark-germinating mutant dkg1 in the fern Ceratopteris richardii. J Plant Res 117: 163–170[CrossRef][ISI][Medline]

Kamada T, Nito K, Hayashi H, Mano S, Hayashi M, Nishimura M (2003) Functional differentiation of peroxisomes revealed by expression profiles of peroxisomal genes in Arabidopsis thaliana. Plant Cell Physiol 44: 1275–1289[Abstract/Free Full Text]

Kamigaki A, Mano S, Terauchi K, Nishi Y, Tachibe-Kinoshita Y, Nito K, Kondo M, Hayashi M, Nishimura M, Esaka M (2003) Identification of peroxisomal targeting signal of pumpkin catalase and the binding analysis with PTS1 receptor. Plant J 33: 161–175[CrossRef][ISI][Medline]

Killion P, Sherlock G, Iyer VR (2003) The Longhorn Array Database (LAD): an open-source, MIAME compliant implementation of the Stanford Microarray Databse (SMD). BMC Bioinformatics 4: 32[CrossRef][Medline]

Kragler F, Lametschwandtner G, Christmann J, Hartig A, Harada JJ (1998) Identification and analysis of the plant peroxisomal targeting signal 1 receptor NtPEX5. Proc Natl Acad Sci USA 95: 13336–13341[Abstract/Free Full Text]

Lee JY, Lee DH (2003) Use of serial analysis of gene expression technology to reveal changes in gene expression in Arabidopsis pollen undergoing cold stress. Plant Physiol 132: 517–529[Abstract/Free Full Text]

Mandaokar A, Kumar VD, Amway M, Browse J (2003) Microarray and differential display identify genes involved in jasmonate-dependent anther development. Plant Mol Biol 52: 775–786[Medline]

Marin K, Suzuki L, Yamaguchi K, Ribbeck K, Yamamoto H, Kanesaki Y, Hagemann M, Murata N (2003) Identification of histidine kinases that act as sensors in the perception of salt stress in Synechocystis sp PCC 6803. Proc Natl Acad Sci USA 100: 9061–9066[Abstract/Free Full Text]

Metcalf JS, Barakate A, Codd GA (2004) Inhibition of plant protein synthesis by the cyanobacterial hepatotoxin, cylindrospermopsin. FEMS Microbiol Rev 235: 125–129[CrossRef]

Micklem DR, Dasgupta R, Elliott H, Gergely F, Davidson C, Brand A, González-Reyes A, St Johnston D (1997) The mago nashi gene is required for the polarisation of the oocyte and the formation of perpendicular axes in Drosophila. Curr Biol 7: 468–478[CrossRef][ISI][Medline]

Muntz K (1996) Proteases and proteolytic cleavage of storage proteins in developing and germinating dicotyledonous seeds. J Exp Bot 47: 605–622

Murgia I, Tarantino D, Vannini C, Bracale M, Carravieri S, Soave C (2004) Arabidopsis thaliana plants overexpressing thylakoidal ascorbate peroxidase show increased resistance to Paraquat-induced photooxidative stress and to nitric oxide-induced cell death. Plant J 38: 940–953[CrossRef][ISI][Medline]

Okita TW, Choi SB (2002) mRNA localization in plants: targeting to the cell's cortical region and beyond. Curr Opin Plant Biol 5: 553–559[CrossRef][ISI][Medline]

Palma JM, Sandalio LM, Corpas FJ, Romero-Puertas MC, McCarthy I, del Rios LA (2002) Plant proteases, protein degradation, and oxidative stress: role of peroxisomes. Plant Physiol Biochem 40: 521–530[CrossRef]

Pertea G, Huang XQ, Liang F, Antonescu V, Sultana R, Karamycheva S, Lee Y, White J, Cheung F, Parvizi B, Tsai J, Quackenbush J (2003) TIGR gene indices clustering tools (TGICL): a software system for fast clustering of large EST datasets. Bioinformatics 19: 651–652[Abstract/Free Full Text]

Pruzinska A, Tanner G, Anders I, Roca M, Hortensteiner S (2003) Chlorophyll breakdown: Pheophorbide a oxygenase is a Rieske-type iron-sulfur protein, encoded by the accelerated cell death 1 gene. Proc Natl Acad Sci USA 100: 15259–15264[Abstract/Free Full Text]

Rajjou L, Gallardo K, Debeaujon I, Vandekerckhove J, Job C, Job D (2004) The effect of alpha-amanitin on the Arabidopsis seed proteome highlights the distinct roles of stored and neosynthesized mRNAs during germination. Plant Physiol 134: 1598–1613[Abstract/Free Full Text]

Rensing SA, Rombauts S, Van de Peer Y, Reski R (2002) Moss transcriptome and beyond. Trends Plant Sci 7: 535–538[CrossRef][ISI]