Expression of stress-related genes in zebrawood (Astronium fraxinifolium, Anacardiaceae) seedlings following germination in microgravity

Seeds of a tropical tree species from Brazil, Astronium fraxinifolium, or zebrawood, were germinated, for the first time in microgravity, aboard the International Space Station for nine days. Following three days of subsequent growth under normal terrestrial gravitational conditions, greater root length and numbers of secondary roots was observed in the microgravity-treated seedlings compared to terrestrially germinated controls. Suppression subtractive hybridization of cDNA and EST analysis were used to detect differential gene expression in the microgravity-treated seedlings in comparison to those initially grown in normal gravity (forward subtraction). Despite their return to, and growth in normal gravity, the subtracted library derived from microgravity-treated seedlings was enriched in known microgravity stress-related ESTs, corresponding to large and small heat shock proteins, 14-3-3-like protein, polyubiquitin, and proteins involved in glutathione metabolism. In contrast, the reverse-subtracted library contained a comparatively greater variety of general metabolism-related ESTs, but was also enriched for peroxidase, possibly indicating the suppression of this protein in the microgravity-treated seedlings. Following continued growth for 30 days, higher concentrations of total chlorophyll were detected in the microgravity-exposed seedlings.


Introduction
Plants have evolved under constant gravitational conditions and even transient exposure to microgravity is unnatural. Gravisensing is one of the most important factors in the regulation of plant growth and development, where plant shoots grow upward (negative gravitropism) and roots grow downward (positive gravitropism) (Morita and Tasaka, 2004). However, seed-to-seed growth experiments performed with plants such as Brassica rapa, Arabidopsis thaliana and peas have shown that microgravity may not be an impediment to development and completion of the life cycle (Musgrave et al. 2000;Laurinavicius et al., 2001;Sychev et al., 2007). Microgravity, experienced by plants during spaceflight, as distinct from simulated microgravity in clinostat or random positioning machine experiments, may cause physiological and ultrastructural changes that can provoke acceleration of growth and differentiation of cells and their aging as a result (Kordyum, 1994). Recent evidence from transcriptome profiling of seedlings and cultured cells confirms the fundamental hypothesis that survival in the spaceflight environment requires adaptive changes that are both governed and displayed by alterations in gene expression, primarily of heat shock-related and stress-related genes . cal traits analyzed appeared to be highly variable (Aguiar et al., 2004). The adaptability of the species has attracted interest for its use in the reforestation of degraded soils in Brazil (Miranda et al., 2011). The seeds are tolerant of osmotic (-10 Bars), and anoxic (1% O 2 ) stress, with germination that is both rapid and homogenous (A.N. Salomão, unpublished data). These characteristics made the germination of zebrawood seeds an attractive choice for one of eight experiments to be carried out by the first Brazilian astronaut, Lt. Col. Marcos Cesar Pontes, on the International Space Station (ISS), where the timeframe available to obtain seed germination results was strictly limited by the proposed six-day orbital flight.
On their return to Earth, the zebrawood seedlings were grown on for a further six days, along with parallel control batches of seedlings grown solely under terrestrial conditions. The germination process was both faster and more homogeneous in microgravity conditions, both for seeds maintained in the presence of light, and for those kept in the dark. A differential gene expression analysis was then conducted after the total 12-day growth period. To our knowledge, these are the first experiments of their kind to be carried out on a tropical tree species.

Materials and Methods
Germination Experiment GSM formed part of the scientific research program performed by the Brazilian Astronaut, Marcos C. Pontes on the Russian Segment of the ISS in the framework of the CENTENARIO Brazilian Soyuz Mission Project. The zebrawood germination experiment was designed by researchers at Embrapa Genetic Resources and Biotechnology, who germinated identical seeds on the ground using the same materials as those used aboard the ISS. Four replicates of 10 A. fraxinifolium seeds were fixed to a blotting paper substrate by Kapton tape in sealed plastic bags. Two plastic bags were exposed to light and the other two were put into an aluminium foil coated bag. Activation of the germination process was performed by seed wetting using a water-filled syringe, incorporated in the germination kit, which was injected by the astronaut when on the ISS. To monitor the germination rate and phenotype, photography by an onboard digital camera was used, where the germination kits kept in the dark were periodically removed from their foil-coated bags to facilitate this. Following the nine day spaceflight, the germinated seeds were returned to earth. On arrival in the laboratory three days later, development was again monitored (12 days postwetting) and samples of the microgravity-treated seedlings and terrestrial controls reserved, where roots and any emergent shoot tissue were dissected and frozen in liquid nitrogen for later RNA extraction.
The remaining seedlings germinated in both environments were also removed from their packs and transferred to a germinator, maintained at a temperature of 25°C, with a photoperiod of 74.98 mMm -2 .s -1 /12 h for about 30 days, for the development of leaflets. Chlorophylls and carotenoids were then extracted, separated, identified and quantified by scanning spectrophotometer at 180-600 nm, adapting methods of Gomes et al. (2003) and Rodriguez-Amaya (1999).

RNA methods and construction of subtractive cDNA libraries
Because of the extremely limited amounts of biological material available from the spaceflight samples, we opted to produce and sequence two cDNA libraries of expressed sequence tags (ESTs) from a subtractive hybridization experiment, to detect differential gene expression under microgravity and normal gravity, mixing light-and dark-grown seedlings together.
The roots and shoots from five dark-germinated, and five light-germinated seedlings from either microgravity or terrestrial samples were ground to a fine powder in liquid nitrogen with a mortar and pestle. Total RNA was then extracted using a mini-column purification kit (Invisorb spin plant RNA mini kit; STRATEC Molecular GmbH, Berlin, Germany). Messenger RNAs were purified using Oligo (dT) 25 magnetic beads (Dynabeads; Dynal -Life Technologies Corp., Carlsbad, USA) and quantified spectrophotometrically using a Nanodrop instrument (Thermo Scientific, Wilmington, USA).
Messenger RNA transcripts enriched either by microgravity or by terrestrial conditions were detected by PCR suppression subtractive hybridization Gurskaya et al., 1996), using the PCR-Select cDNA Subtraction Kit (Clontech Laboratories Inc., Mountain View, USA), following the manufacturer's protocols. A forward subtracted experiment was set up using cDNA from the seeds germinated in space as the tester and cDNA from the terrestrially germinated samples as driver, utilizing a 10-fold excess of driver cDNA over tester cDNA. A reverse subtracted experiment was also set up using the converse configuration of tester and driver cDNAs. For the primary subtraction the cDNAs were denatured at 98°C for 1.5 min, and hybridized at 68°C for eight hours. Samples were then secondarily subtracted for a further eight hours using an additional 10-fold excess of denatured driver cDNA. PCR amplified subtracted cDNAs were cloned in the pGEM T-easy vector (Promega, Madison, USA) and used to transform E. coli DH5a competent cells. Positive clones were unidirectionally sequenced using the Big Dye 3.1 kit and 3700 sequencer (Applied Biosystems -Life Technologies Corp., Carlsbad, USA). Sequences from the subtracted clones were stripped of vector and adapter sequence and of low-quality regions and assembled into contigs using the high sensitivity/medium setting of Geneious (v.5.4.3., Biomatters Ltd. Auckland, New Zealand). 82 Inglis et al.

EST analysis
EST analysis was carried out using the Blast2GO gene annotation and ontology assignment pipeline (Conesa and Götz, 2008). BLAST searches of the NCBI nucleotide database (nr) used the BlastX algorithm, with an ExpectValue of 1.0E-3 and HSP length cutoff of 33. The 20 most significant BlastX hits per sequence were saved, and the top hits then annotated with their gene ontology (GO)-terms, using the default annotation configuration and evidence code weights. Annotation augmentation (ANNEX) was applied, and the results of an InterProScan and GO-EnzymeCode mapping steps added. The gene ontologies were also simplified using the GO-Slim Plant ontology for comparative purposes. Overlap in GO-terms between the libraries was calculated and visualised using the BioVenn web application (Hulsen et al., 2008).

Germination
Radicle protrusion and growth of rootlets appeared both faster and more uniform in seeds sent to the ISS. Seven days after the start of the experiment, 60% of the seeds in the presence of light and 30% of those kept in darkness had germinated on the ISS. For seeds kept in the laboratory, germination values were 5% and 10% in the presence and absence of light, respectively, though these values did not reach statistical significance, due to the large variances obtained ( Figure 1). On the day the experiment returned to Earth (nine days after the start of the experiment), the percentages of seeds germinated in microgravity were 65% (light) and 85% (dark) and the seeds germinated in the laboratory were 80% (light) and 60% (dark). At 12 days after the start of the experiment, the germination of the material from the ISS was 100% in both the presence and absence of light, with rootlets having an average length of 1.5 cm and starting to show positive geotropic curvature. The germination rate of the terrestrial control was also 100% by the 12th day, but the average length of the radicle was 0.5 cm. During later seedling development for 30 days in the germinator, no difference between the growth of the shoots derived from seeds exposed to microgravity and the control seedlings was observed (not shown). However, the microgravity-treated seedlings of seeds had greater root length and number of secondary roots.

Pigments content
The ratio of Chlorophyll a to b was 2.5 in leaflets of seedlings that started germination both in microgravity and in the laboratory ( Figure 2). However, leaflets of plantlets that germinated in microgravity showed a significantly (p < 0.001) higher concentration of total chlorophyll (581.47 mg/100 g fresh weight) than those germinated in the laboratory (381.84 mg/100 g fresh weight), following 30 days of terrestrial growth. The concentration of a-carotene in leaflets of plantlets that initiated germination in microgravity was not significantly different (p > 0.05) from those germinated terrestrially, while the concentrations of both b-carotene (p < 0.05) and xanthophils (p < 0.01) were both significantly higher in the microgravity-treated samples ( Figure 3).

Suppression subtractive hybridization library EST analysis
Of the microgravity SSH ESTs, 704 of 741 high quality sequence reads were assembled into 56 contigs, leaving 36 singletons. Filtering of ribosomal RNAs using Aesculus pavia rRNA 26S, 18S and 5.8S sequences found a massive 405 ESTs assembling to the 26S rRNA gene and a further 17 ESTs assembling to the 18S rRNA gene. Of the terrestrial SSH ESTs, 588 of 681 high quality reads were assem-Seed germination in microgravity 83 Figure 1 -Germination rates of seeds of Astronium fraxinifolium in microgravity (ISS) and terrestrially (Lab). There was no significant difference (p > 0.05) between the percentage germination in microgravity and terrestrial conditions, and in both light and dark, according to ANOVA with Bonferroni test). Error bars = SD.
Figure 2 -Synthesis of chlorophylls in 30-day plantlets derived from seeds germinated in microgravity (ISS) and terrestrially (Laboratory). The chlorophyll content was measured relative to leaflet fresh weight, and differences are statistically significant (p < 0.001) according to ANOVA with Bonferroni test for all three pairs of measurements between the ISS and the terrestrial laboratory. Error bars = SD.
bled into 101 contigs, leaving 93 singletons. Ribosomal RNA filtering found 136 ESTs matching 26S rRNA and 1 EST matching 18S rRNA. Cross-assembly between the two libraries, with the exception of the rRNA hits, found only three mixed-library contigs, demonstrating that the PCR suppression subtractive hybridization protocol had otherwise been highly effective. The species distribution of the top BLAST hits for the rRNA filtered ESTs was similar for both microgravity and terrestrial SSH libraries, and the combined results are shown in Figure 4. The great majority of hits were with genes from other dicotyledonous plants, as might be expected. The top hits for both microgravity and terrestrial library contigs (excluding singletons) are given in Table 1 and Table 2 respectively.
The microgravity SSH library was notably rich in ESTs of stress-related genes, where seven hits with several different classes of heat shock proteins were noted (Table 1), as well as an additional six hits among the singleton reads (not shown). In contrast, there was just one hit for heat shock protein 70 in the terrestrial library (Table 2). Also related to a stress-response, there were three separate hits for glutathione transferase among the contigs in the A. fraxinifolium microgravity SSH library (Table 1) and an additional hit among the singleton reads (not shown), whereas no hit for this protein was found in the terrestrial library. Similar stress-related hits included polyubiquitin and 14-3-3-like protein. Notably enriched in the terrestrial SSH library was peroxidase (Table 2).
Despite the low frequency of common EST reads in the two libraries, there was a large overlap in all GO terms extracted from the microgravity and terrestrial SSH library top Blast hits ( Figure 5). This was emphasised in the analysis of the GOSlims, where the great majority of GOs were common to both libraries. Unique plant GOSlims in the microgravity library were: GO:0009856 (fimbrin-like protein 2-like (Pollination)), from contig 29; GO:0030528 (btb and taz domain protein 4 (transcriptional regulator activity)), singleton; GO:0007049 (cell division cycle protein 48 homolog (cell-division cycle)), singleton; GO:0030234 (cystatin (enzyme modulator)), singleton. Because of the 84 Inglis et al.

Figure 3
-Synthesis of carotenoids in 30-day plantlets derived from seeds germinated in microgravity (ISS) and terrestrially (Laboratory). The carotenoid content was measured relative to leaflet fresh weight, and pairs of measurements between the ISS and the terrestrial laboratory are significant at p < 0.05 for b carotene and p < 0.01 for xanthophylls, but were not significant at p > 0.05 for a carotene according to ANOVA with Bonferroni test. Error bars = SD.    low differential resolution offered by the GOSlim terms, further comparative analysis was conducted on the full set of GO terms. The elevated expression of stress-related genes in the spaceflight exposed seedlings was evident in the analysis of the GO terms for the top hits for all contigs and singleton reads. Figure 6 shows the result of a Fisher's Exact Test comparison (Bluthgen et al., 2005) of the enrichment of all GO terms (full set) in the microgravity and terrestrial libraries, where 33% of the microgravity sequences (contigs and singletons) yielded the GO term "response to stress", compared to 13% of the terrestrial sequences. Similarly, the terms "glutathione metabolic process", "glutathione conjugation reaction" and "glutathione transferase activity" were exclusive to the microgravity library. Similar microgravity-exclusive, stress-related terms were "defence response to bacterium" and "response to fungus". Additionally, the term "response to stimulus" was almost doubled in representation in the microgravity library in comparison with the terrestrial sample, and the term "response to oxidative stress" was increased almost four-fold. Conversely, the terrestrial subtracted library was much richer in generalist growth-related terms such as "biological process", "structural molecule activity", and "cellular component biogenesis", among other structural GO terms, and, as stated earlier, possessed nearly double the number of distinct BLAST hits for contigs and an almost three-fold increase in diversity of singleton read hits.

Discussion
The high germination rate and rapid root growth seen in the A. fraxinifolium seeds exposed to spaceflight has been observed in other species, such as Linum 88 Inglis et al.  usitatissimum (Levine et al., 2003), Glycine max (Levine et al., 2000) and Arabidopsis (Millar et al., 2011). The limited time available on the spaceflight precluded reliable observation of any difference in shoot length in the microgravity and terrestrially germinated seedlings, but early shoot growth stimulation has been observed in other, faster developing plants, such as rice seedlings subjected to simulated microgravity, such as clinorotation (Jagtap et al., 2011), where chlorophyll content was also increased. After prolonged further development under normal gravity, shoot growth appeared not to be affected by microgravity treatment in A. fraxinifolium. However, an interesting persistent effect on chlorophyll content was observed, even after 30 days of subsequent terrestrial growth.
The abundance of reads representing rRNA genes in the cDNA libraries is surprising, since these molecules are thought not to be normally polyadenylated, and the library preparation method included both a mRNA purification step and oligo(dT)-primed first strand cDNA synthesis, to avoid spurious rRNA cloning. Possible explanations for the abundance of rRNA genes in our libraries include annealing of the oligo(dT) first strand cDNA synthesis primer to poly-A tracts in the A. fraxinifolium rRNA or self-priming by hairpin formation (Gonzalez and Sylvester, 1997). Although inspection of the unclipped rRNA sequences from our libraries showed no convincing evidence of polyadenylated tails, polyadenylation of rRNA has been periodically reported in the literature, where in plants, rRNA polyadenylation was first observed in Nicotiana tabacum stressed by exposed to cadmium (Lewandowska et al., 2007).
The presence of multiple heat-shock protein reads in the microgravity-exposed A. fraxinifolium seedlings agrees with previous experimental findings in plants exposed to gravity perturbations. In this context, Kozeko and Kordyum (2006) showed that HSP70 and HSP90 levels were both significantly increased in pea seedlings grown under horizontal or vertical clinorotation, simulating microgravity. Similarly, levels of HSPs have been found to increase under hypergravity, where the HSPs could be involved in protein stabilization and quality control as well as signal transduction pathways under altered gravity (Kozeko and Kordyum, 2009). Using an agravitropic mutant of Arabidopsis thaliana grown under clinorotation or increased gravity (7 g), it was shown that HSP70 and glutathione s-transferase 6, among other proteins, are part of a generalized stress response to gravitational changes (Tan et al., 2011). The glutathione antioxidant pathway has also been shown to be induced in Xenopus laevis embryos during simulated microgravity, and was suggested to play a protective role (Rizzo et al., 2009). A further type of stress-related gene present among the microgravity ETSs was polyubiquitin (Table 1, 6 ESTs), where in budding yeast, the stress-inducible polyubiquitin gene, UBI4, has been shown to be upregulated in response to oxidative stress, and has been suggested to play an important role in increasing cellular ubiquitin levels to allow cells to survive under toxic stress conditions (Cheng et al., 1994). Furthermore, polyubiquitin has been shown to be upregulated in rat muscle during spaceflight, and was associated with the stimulation of expression of ubiquitin-proteasome pathway genes and resultant muscle atrophy (Ikemoto et al., 2001). Recent evidence suggests that protein ubiquitination also plays a critical role in regulating responses to abiotic stresses in plants (Lyzenga and Stone, 2012).
It has been shown that peroxidase activity can be depressed under weightlessness in pine seedlings (Cowles et al., 1984), in Brassica napus protoplasts (Skagen and Iversen, 2000), and in germinating spores of the aquatic fern, Ceratopteris richardii, where three different genes likely to encode peroxidases were down regulated 1.5 to 2.5 fold during spaceflight (Salmi and Roux, 2008). Peroxidase repression in A. fraxinifolium during spaceflight may, therefore, have been indicated by its corresponding overrepresentation in the terrestrial SSH library, where peroxidase was the most abundant non-rRNA EST (Table 2; Contigs 02 and 08). In Arabidopsis, however, peroxidase expression was shown to be insensitive to six minutes exposure to microgravity on a sounding rocket (Martzivanou et al., 2006). 14-3-3-like protein was also among the more abundant ESTs in the microgravity library (Table 1, contig 9,8 ESTs), and has been associated with the defence response to abiotic and biotic stress. This protein has been shown to interact with ascorbate peroxidase and may play a regulatory role in the stress response on multiple levels. Important mechanisms of regulation by 14-3-3 include shuttling proteins between different cellular locations and acting as scaffolds for the assembly of larger signalling complexes (Roberts et al., 2002). In Drosophila cells, expression of 14-3-3 protein has been shown to be heat shock-related and, in cooperation with Hsp70/Hsp40, was demonstrated to mediate the resolubilization and reactivation of heataggregated citrate synthase, where 14-3-3 protein or Hsp70/Hsp40 alone, lacked the activity (Yano et al., 2006). In the context of an induction of a generalized stress response in the microgravity-treated seedlings, it would not be surprising, therefore, to detect the enrichment of 14-3-3-like protein transcripts along with several other stressrelated ETSs. The enrichment of 14-3-3-like protein ESTs may also be correlated with the longer root length observed in the microgravity-treated seedlings, where these proteins have recently been demonstrated to play an important role in both root and chloroplast development in Arabidopsis.
Here, 14-3-3 defective mutants have shorter roots than wild-type, but show increased root greening (Mayfield et al., 2012). In the context of the observed higher chlorophyll content of the microgravity-treated seedlings, 14-3-3 proteins, together with the Hsp70 molecular chaperone, are thought to play a role in chloroplast development, guiding Seed germination in microgravity 89 phosphorylated chloroplast precursors towards their destinations (May and Soll, 2000). It would appear that a major consequence for germination during spaceflight for the A. fraxinifolium seeds was an at least temporary switch to a "stress" mode of growth and a quantitative reduction in general metabolism, when compared to terrestrially germinated seeds. This was despite the phenotypic effects observed, where the microgravity sample germinated more rapidly and more homogenously. The latter observation may, in part, explain the increased variety of ESTs in the terrestrial sample, which, being less synchronized than the spaceflightsample, may have possessed a greater variety of temporally-expressed mRNAs, quantitatively affecting the subtractive hybridization results.
The gene expression responses to microgravity-induced stresses are likely to be variable over time and difficult to standardize between different model-and experimental systems. An example of this was found in rat cells, where the activity of the intracellular antioxidant enzymes, superoxide dismutase, glutathione peroxidase, and catalase, was all significantly increased at 12 h after the microgravity onset, yet decreased at 96 h (Wang et al., 2009). Preparation procedures for spaceflight and the non-ideal environmental conditions on board the ISS subject organisms to additional environmental stresses that demonstrably affect gene expression. The vibrational stresses incurred during space vehicle launch and re-entry and transient hypergravity during acceleration are likely to have a physiological impact on biological systems. The stress gene responses induced by vibration, however, may differ from purely microgravityrelated responses, where, for example, mechanical stresses and vibration did not cause the up-regulation of mRNA for hsp70 and hsp27 in human lymphocytes (Cubano and Lewis, 2001). Importantly, the response of whole plants or seedlings may differ to that of cultured cells. In Arabidopsis thaliana seedlings, up-regulation of TCH (touch) gene expression and an increase in hypocotyl elongation was demonstrated in response to vibration at 50 Hz for 72 h, though the response was weaker than in touch-stimulated plants (Johnson et al., 1998).
Logistical problems are a major complication for gene expression analysis in true spaceflight scenarios, where in the current study, three days of terrestrial development during transit to our laboratory in Brazil was added to the nine day spaceflight time, before samples could be stabilized. Despite this, however, the effects of microgravity on A. fraxinifolium seedlings appear to be persistent, at least for three days following their return to normal gravity, and longer when the increased chlorophyll content in the 30 day plantlets is considered. An example of the long-term effects of stress on plants is priming of defence in response to pathogen attack, which is a strategy employed by stressed plants to enhance resistance against future stress episodes with minimal associated costs on growth. Worrall et al. (2012) showed that tomato seeds treated with the signalling molecules, jasmonic acid or b-aminobutryric acid, displayed increased resistance for up to eight weeks to a range of pests and diseases. A growing body of evidence indicates that stress can induce persistent and substantial changes at the chromatin level in plants, with concomitant changes in gene expression. The long-term duration and heritability of these changes, however, is controversial (Pecinka and Scheid, 2012). Simulated microgravity experiments performed on the ground, under ideal conditions, using a random position machine or clinostat, may show much more subtle effects on gene expression (Herranz et al., 2010).
Although localized hypoxia, caused by the design of the germination kits utilized in our experiment, may be a source of stress for the germinating seedlings, this aspect of the experiment could have been eliminated by the use of identical kits for the terrestrial controls. Differing convection currents and variations in heat and gas exchange during spaceflight and their effects on plants have been investigated (Kitaya et al., 2006), but these effects are, however, difficult to reproduce accurately on the ground in the control kits. Also difficult to control and separate from the direct microgravity-induced effects is the increased exposure of living tissue to cosmic radiation as well as variation in magnetic flux during spaceflight. Despite the fact that reported gene expression effects are frequently similar in real and simulated microgravity experiments, the synergistic effects of the spaceflight environment on microgravity responses cannot be discounted (Beckingham, 2010;Herranz et al., 2010).
Our results largely agree with many other experiments using more widely studied model plants, where a general theme of the induction of multiple stress genes in response to microgravity is usually seen. Since this is the first gene expression analysis of the germination during spaceflight of a tropical tree species, this study represents only the first glimpse of the response of these plants to this environment. Terrestrial clinostat experiments with accompanying EST analyses, involving A. fraxinifolium seedlings, would enable us to partially confirm some of the findings in the present study, where the timing and longevity of the response to gravitational stress would be interesting to investigate in detail, which would be greatly facilitated in purely ground-based experiments. More material reserved for molecular analysis would have allowed us to investigate the microgravity response in greater detail. In particular, it would be interesting to dissect the differential response of shoot and root, in both light and darkness, to microgravity. The logistical problems involved in future true microgravity experiments may be somewhat alleviated by sample fixation by rapid freezing, or more practically using room temperature chemical fixative reagents such as RNAlater (Qiagen), which can be applied by nonspecialists during spaceflight. These measures will make future microgravity experiments much easier to interpret. 90 Inglis et al.