Author for correspondence: Dieter Ernst Tel: +49 89 31874440 Fax: +49 89 31873383 Email: email@example.com
• Radiocaesium is one of the main anthropogenic sources of internal and external exposure to β- and γ-radiation (e.g. from global fallout of atmospheric atomic bomb testing and from the Chernobyl reactor accident). Here we investigated gene expression by suppression subtractive hybridization (SSH) and reverse transcription–polymerase chain reaction (RT–PCR) in Arabidopsis thaliana, which was induced by the root uptake of 134Cs.
• SSH analysis resulted in the isolation of 46 clones that were differentially expressed at 30 Bq cm−3 134Cs. Most of the expressed sequence tags identified belonged to genes encoding proteins that were involved in cell growth, cell division and the development of plants, and in proteins controlling translation, general metabolism and stress defence, including a DNA excision repair protein.
• The accumulation of caesium in plant material was measured in plants grown for 5 wk on agar contaminated by up to 60 Bq cm−3 134Cs. 134Cs was found to accumulate, in particular, in leaf rosettes and was dependent on the activity concentration in the growth media.
• The data indicate that low-level ionizing radiation influences important cellular responses, resulting in a changed gene-expression profile.
The exposure of plants to environmental stressors, such as ultraviolet (UV) radiation, air pollutants, drought, cold, metal ions or ionizing radiation, results in the induction of plant stress response networks and in modified plant fitness (Sandermann, 2004). Early plant responses towards a stress factor include a changed transcriptional profile, resulting in the synthesis of proteins and metabolites that act to counteract the unfavourable conditions. Plant responses to diverse environmental stress have been studied in detail (Smallwood et al., 1999; Inzé & Van Montagu, 2002; Sandermann, 2004). However, reports on the transcriptional effects of low-level ionizing radiation are rather rare. Ionizing radiation is known to have a number of different effects on plants, ranging from growth stimulation up to severe leaf, trunk, yield, etc., effects, depending on the radiation dose, as well as on the plant species used (IAEA, 1992; UNSCEAR, 1996; Holst & Nagel, 1997; Streffer et al., 2004). In terrestrial plant populations, no adverse effect was observable below chronic dose rates of up to 10 mGy d−1, and the radiation sensitivity of different species was found to be inversely related to their respective chromosome mass (Sparrow et al., 1967). By using large external γ-radiation exposures of up to 100 Gy, transcriptional changes in Arabidopsis were induced (Deveaux et al., 2000; Doucet-Chabeaud et al., 2001; Pierrugues et al., 2001). Even larger γ-ray doses (1–8 kGy) resulted in the appearance of certain transcripts and proteins, and in a reduction of the abundance of others (Ferullo et al., 1994). Low chronic doses of external ionizing radiation resulted in the induction of antioxidative enzyme activities in Stipa capillata that could be increased by additional growth on 134Cs-treated soil (Zaka et al., 2002).
Radicaesium is an important artificial radionuclide that has mainly been introduced into the environment by the atmospheric testing of nuclear weapons, mostly in the period between 1945 and 1963, and by accidental release from nuclear facilities, such as the Chernobyl reactor accident in 1986 (Karaoglou et al., 1996; Yamashita & Shibata, 1997). Molecular aspects of plant adaptation in the Chernobyl zone indicate complex processes involving changed gene expression, epigenetic effects and genome stability (Kovalchuk et al., 2004). Caesium can be easily taken up by plant roots and then transferred to above-ground plant parts. This is a key process for the entry of 137Cs into the food chain (Witherspoon & Brown, 1965; Guivarch et al., 1999) and thus leads to internal exposure of humans in the long term after contamination of soils. Genetic variation in 137Cs uptake is evident from the wide range of transfer factors observed in different plants (Broadley & Willey, 1997; Lasat et al., 1998; Zhu & Smolders, 2000; Payne et al., 2004). Caesium toxicity in Arabidopsis, and a Cs-changed transcriptional profile, were recently described (Hampton et al., 2004; Sahr et al., 2005).
In the present study we analyzed the uptake of 134Cs and studied the effects of chronic low-level ionizing radiation on A. thaliana. The combination of suppression subtractive hybridization (SSH) and microarray technology resulted in the identification of numerous genes that showed a changed transcriptional profile upon application of 134Cs at an activity of 30 Bq cm−3.
Materials and Methods
Plant growth conditions
A. thaliana ecotype Columbia (Col-0) wild-type seeds were surface sterilized and then transferred into 0.1% (w/v) agar (Sahr et al., 2005). For 134Cs-uptake studies, plants were grown in an agar medium, according to Gibeaut et al. (1997), containing 0.7 mm K+ and the different activities of 134CsCl (Risoe National Laboratory, Roskilde, Denmark). Plants were grown in normal daylight and at room temperature. For molecular biological analysis, 3-d-old seedlings were transferred to Erlenmeyer flasks containing 30 ml of a liquid medium (Gibeaut et al., 1997) without 134Cs (control), or supplemented with 134Cs and cultivated, under the conditions stated earlier in this paragraph, on a gyratory shaker at 60 rpm. Plants were separated into root and leaf tissue and the roots were frozen in liquid nitrogen, homogenized with a mortar and pestle, and stored at −80°C for further analysis, whereas the leaf material was discarded.
134Caesium was determined by direct gamma-spectrometry by using a high-purity germanium detector and a multichannel analyser (Canberra Eurisys, Rüsselsheim, Germany). The activity of 134Cs was corrected for losses by sum-coincidences during counting. Errors in measurements were far below the standard deviation (SD) of biological repetitions and therefore not taken into account.
RNA isolation and cDNA synthesis
Total root RNA was isolated by using TRIzol® reagents (Invitrogen, Karlsruhe, Germany). Poly(A)+ was isolated from total RNA by using the Oligotex™ mRNA kit (Qiagen, Hilden, Germany). The RNA yield and quality was determined by spectral photometry at 260 and 280 nm.
Two micrograms of total RNA was reverse transcribed by using the Superscript™ II reverse transcriptase, according to the manufacturer's instructions (Invitrogen).
Suppression subtractive hybridization
Suppression subtractive hybridization was performed by using the PCR Select cDNA subtraction kit (Clontech, Heidelberg, Germany) with 2 µg of poly(A)+ RNA (Heidenreich et al., 2001; Sahr et al., 2005). Three different subtractive cDNA libraries were constructed. Control (no 134Cs) and treated (30 Bq cm−3 134Cs) plants were used alternately as driver and tester.
Array production, hybridization and data analysis
Isolated cDNA clones were amplified by using flanking vector DNA sequences. Array production, hybridization and data analysis were exactly as recently described (Sahr et al., 2005). Primary data were processed and normalized by using the arrayvision 6.0 software (InterFocus, Mering, Germany). Clones that hybridize in all replicates to the forward-subtracted and unsubtracted tester probes, but not (or at least with ratios of > 2 for induced or < 0.5 for suppressed genes) to the reverse-subtracted or unsubtracted driver probes, were considered as an indication of a significant change of expression. DNA sequencing of positive clones was carried out by Medigenomix (Martinsried, Germany), and for Blast analysis the MIPS A. thaliana database was used (MAtDB; http://mips.gsf.de/proj/thal/db/index.html).
Semiquantitative reverse transcription–polymerase chain reaction (RT–PCR) was carried out in the presence of GeneAmp® RNA pAW109 (Applied Biosystems, Darmstadt, Germany). Amplification of the cDNA was carried out as follows: incubation at 95°C for 5 min; then 33–39 cycles at 95°C for 1 min, 53°C for 30 s and 72°C for 5 min. The artificial standard cDNA was amplified by using the primer pair DM151/DM152. After separation on 2% (w/v) agarose gels, intensities of the corresponding bands were quantified by using the MultiAnalyst software (Biorad, München, Germany). An induction factor was calculated according to the following induction ratio: intensitysample/standard : intensitycontrol/standard (Sahr et al., 2005). Two independent biological repetitions at 60 Bq cm−3 were carried out. For each repetition, two to three cDNA preparations were obtained that were used for RT–PCR. Together, four to six independent values were used to calculate an induction factor that was statistically analyzed by using the t-test (s-plus 6.2 Professional; Insightful Corporation, Seattle, WA, USA).
A. thaliana Col-0 were grown for 5 wk on a medium containing 0.7 mm K+ and very low concentrations of 134Cs (1.5–9.4 × 10−12m). Including the inactive Cs, the total Cs concentration was 1.3–7.9 × 10−10m. These concentrations were far below the known phytotoxic concentrations of 133Cs, which range from 5 × 10−4m to 7 × 10−4m (Sahr et al., 2005). Therefore, all effects observed upon 134Cs application were caused by low chronic doses of ionizing radiation and not by the known effects of Cs ions (Hampton et al., 2004; Sahr et al., 2005). Increasing the external activity of 134Cs resulted in an increased accumulation of 134Cs in leaves, as measured by an increased activity of up to 1 Bq mg−1 fresh weight (f. wt) in the leaves (Fig. 1). The absorbed dose rate of an Arabidopsis plant for external γ-radiation and for internal β- and γ-radiation exposure, as a result of 134Cs, was estimated by using the dose rate conversion coefficients, calculated for reference terrestrial biota, of Taranenko et al. (2004) and Gómez-Ros et al. (2004). The dose rate of the external exposure was only approx. 0.1 µGy h−1 and, thus, negligibly small as compared to the dose rate of the internal exposure amounting to 50–100 µGy h−1. The internal and external exposure caused by natural radionuclides, mainly 40K, was < 1 µGy h−1. Thus, the absorbed dose rate of an Arabidopsis plant was dominated by the internal exposure to 134Cs and higher than the natural background by a factor of approx. 100, but smaller, by several orders of magnitude, than the dose rate caused by γ-irradiation in many other studies. Real et al. (2004), reviewing all data available on the effects induced by ionizing radiation in various wildlife groups, conclude that the threshold for statistically significant effects is a dose rate of approx. 100 µGy h−1.
The internal 134Cs contamination in A. thaliana was much higher than in plants grown on different soil types that were spiked with 134CsCl (Zaka et al., 2002; Tang & Willey, 2003). Although Cs accumulation varies in different plant species (Broadley & Willey, 1997; Tang & Willey, 2003; White et al., 2003), the high 134Cs accumulation in this study was mainly dependent on an increased bioavailability using agar medium. Application of 100 µm 133Cs resulted in a further slight uptake of 134Cs (Fig. 1). This is in accordance with an increased uptake of Cs caused by increasing the external 133/137Cs concentration (Fuhrmann et al., 2002; White et al., 2003; Hampton et al., 2004; Sahr et al., 2005). Up to an activity of 30 Bq cm−3, no influence on morphology and development of plants was found. External activities of 60 Bq cm−3 resulted in a growth reduction of leaves, stem and roots. However, flowering and seed production was not affected at all.
Isolation of genes up- and down-regulated by 134Cs
Abiotic and biotic stress responses, detected by an increased or a decreased level of transcripts, are often not specific, as identical genes might be affected also by other stress factors (Heidenreich et al., 1999; Reymond et al., 2000; Broschéet al., 2002; Inzé & Van Montagu, 2002; and articles therein; Mahalingam et al., 2003). In addition, the dose, as well as the duration of stress given, might affect the stress specificity. Focussing on ionizing radiation, gene-expression profiling of human carcinoma cells revealed overlapping and distinct classes of genes, both of which are induced by internal β- and by external γ-radiation (Marko et al., 2003). Recently, we described the isolation of 133Cs-affected genes in roots of Arabidopsis by SSH (Sahr et al., 2005). To compare these 133Cs effects with the low level of ionizing radiation of 134Cs at the transcriptional level, Arabidopsis plants were grown for 5 wk in a hydroponic medium containing 134CsCl at an activity of 30 Bq cm−3. This value is comparable with realistic outdoor activities in the Chernobyl zone (White & Broadley, 2000). Although 134Cs was incorporated into the plants we cannot distinguish between external and internal radiation effects caused by the high β-energy of 2.059 MeV and the additional γ-radiation, although the dose rate of the internal exposure was much higher than the external exposure.
Clones obtained by the SSH were further analysed by DNA-microarray analysis and only clones that showed differences in the signal intensities (subtracted vs nonsubtracted) with a factor of > 2 or < 0.5 were taken as positive. After sequencing, redundant clones were eliminated that finally resulted in 46 annotated genes (Tables 1 and 2). These genes were grouped into nine classes according to their function (Table 1). The slight discrepancy between the numbers of genes in Table 1 (50) vs the 46 134Cs-affected genes was a result of double designation, caused by an unclear assignment. Out of the 46 influenced genes, 41 were up-regulated and five were down-regulated (Table 2). As a control for the validity of the microarray data, RT–PCR amplification was carried out for selected genes (Table 3). Arabidopsis plants were grown in the presence of 134Cs (10, 20 and 60 Bq cm−3) and roots were harvested after 5 wk. Duplicate biological samples and up to six independent RT–PCR amplifications for each sample were carried out in order to verify even very weakly induced or repressed genes, respectively (Table 3).
Table 1. Composition of clusters of the isolated 134Cs-affected expressed sequence tags (ESTs) in roots of Arabidopsis thaliana
Major functional categories
Cell growth, division and development
Transcription and translation
Protein synthesis, folding and modification
Transport and homeostasis
Cellular communication and signalling
Defence, stress response and detoxification
Table 2. Up- and down-regulated (marked in italics) transcripts in roots of Arabidopsis thaliana, isolated by using the suppression subtractive hybridization (SSH) method
A. thaliana ecotype Columbia (Col-0) wild-type plants were grown in liquid medium in the absence or presence of 134CsCl (60 Bq cm−3) in normal daylight and at room temperature. Transcripts that were verified by semiquantitative and quantitative real-time reverse transcription–polymerase chain reaction (RT–PCR) are marked in bold. Accession numbers and annotations are according to the MAtDB (http://mips.gsf.de/proj/thal/db/index.html).
Table 3. Up-regulated expressed sequence tags (ESTs) in roots of Arabidopsis thaliana, isolated by the suppression subtractive hybridization (SSH) method and verified by semiquantitative reverse transcription–polymerase chain reaction (RT–PCR) and real-time RT–PCR
Quantitative real-time RT–PCR
A. thaliana ecotype Columbia (Col-0) wild-type plants were grown in liquid medium in the absence or presence of 134CsCl (60 Bq cm−3) in normal daylight and at room temperature. Accession numbers and annotations are according to the MIPS A. thaliana database (MAtDB; http://mips.gsf.de/proj/thal/db/index.html).
Induction factors are according to the t-test for semiquantitative RT–PCR, and for quantitative real-time PCR the calculation was carried out as described by Pfaffl (2001) (P-values are indicated in parenthesis).
Cell cycle control crn protein-like
Mitosis-specific cyclin 1b
Tubulin α-5 chain
Putative auxin-regulated protein
DNA excision repair protein
Cold-regulated protein (kin2/kin1)
Functional classification of the isolated genes
One group of induced transcripts are involved in general metabolism (Tables 1 and 2), including 3-hydroxy-3-methylglutaryl-CoA-reductase (HMGR), O-succinylhomoserine sulfhydrylase and α-ketoacid dehydrogenase. 3-Hydroxy-3-methylglutaryl-CoA-reductase is a key enzyme of isoprenoid biosynthesis that is up-regulated by abiotic and biotic stress factors (Yang et al., 1991; McGarvey & Croteau, 1995). Furthermore, it has been shown that 3-hydroxy-3-methylglutaryl-CoA-synthase, which precedes HMGR, is induced by ozone (Wegener et al., 1997a). As ozone treatment, as well as ionizing radiation, results in the production of reactive oxygen species (ROS) (Riley, 1994; Langebartels et al., 2002; Rugo & Schiestl, 2004; Streffer et al., 2004), this might be a common link of these stress factors. Two other transcripts of this group, a glutamate dehydrogenase and xylose isomerase, important for nitrogen and carbohydrate metabolism, were down-regulated with 134Cs stress (Table 2).
An important group of up-regulated mRNAs are involved in cell growth, cell division and development (Tables 1 and 2). Transcripts of a cell cycle crn control protein-like and of a mitosis-specific cyclin B1 were induced by a factor of 1.2–1.4 and of 1.5–2.3, respectively (Table 3). Cell cycle proteins are involved in the regulation of mitosis and RNA processing in eukaryotes (Preker & Keller, 1998). Together with cyclin-dependent protein kinases, cyclins are essential for cell cycle control (Weingartner et al., 2003). It is also known in plants that substances such as aphidicolin, which damage or disturb the DNA replication, induce the gene expression of cyclin B1 and lead to a G2 arrest (Culligan et al., 2004). The transcriptional response of Arabidopsis to genotoxic stress by the application of bleomycin and mitomycin C resulted also in an up-regulation of repair cell cycle genes (e.g. cyclin-dependent kinases and CycB1) (Chen et al., 2003). In addition, crn-like proteins are also induced upon Cs, salt, osmotic and UV stress, and cyclin B1 has been shown to be induced by heat stress and induced programmed cell death (Swidzinski et al., 2002; Craigon et al., 2004; http://affymetrix.arabidopsis.info/narrays/geneswinger.pl).
Transcripts for profilin 2, a putative actin-depolymerizing factor and α-5 tubulin were also induced by 134Cs (Tables 2 and 3), indicating a reorganization of the plant's cytoskeleton upon ionizing radiation. Profilins are localized in the cytoplasm, as well as in the nucleus, and are regulatory proteins of actin modulation (Christensen et al., 1996). Profilins are also induced upon exposure to several other stresses, such as Cs, salt, osmotic stress and UV radiation (Craigon et al., 2004; Hampton et al., 2004; http://affymetrix.arabidopsis.info/narrays/geneswinger.pl). Microtubules, together with tubulin-associated proteins, regulate root growth and are important for stabilization of the intracellular localization of parts of the cytoplasma and for cell division (Jacobs et al., 1988; Bibikova et al., 1999). Cyclin B1 has been shown to be localized to microtubules in human cells (Jackmann et al., 1995). This suggests that an up-regulation of α-5 tubulin and cyclin B1 transcripts upon ionizing radiation contributes to functional microtubules.
Transcript levels of two 60S ribosomal protein transcripts and of a G41-like ribosomal protein were increased upon exposure to 134Cs (Table 2). Ribosomal proteins, important for protein synthesis, are also induced upon genotoxic stress (Chen et al., 2003) or UVB radiation (Ernst et al., 2001; Izaguirre et al., 2003; Casati & Walbot, 2004). Gamma-radiation of human carcinoma cells also resulted in an increased amount of a 40S ribosomal protein and a ribosomal S6 kinase (Marko et al., 2003). This indicates that protein synthesis in plant cells is sensitive to low ionizing radiation and not only to UV radiation.
A putative ubiquitin-conjugating enzyme E2 was up-regulated by 134Cs (Table 2). Similarly, a ubiquitin-conjugating enzyme E2 has also been shown to be induced by genotoxic stress (Chen et al., 2003). Ozone fumigation of Scots pine seedlings was found to result in the accumulation of polyubiquitin mRNA (Wegener et al., 1997b), indicating again the involvement of ROS in signalling, at least partially, the effects of ionizing radiation. In addition to the well-known function of ubiquitins in the degradation of damaged proteins, the ubiquitin system is also of importance in DNA repair, by modification of involved enzymes, thus influencing their activities (Hiller et al., 1996; Ciechanover, 1998; Hellmann & Estelle, 2002; Pickart, 2002).
Only two transcripts for cellular communication and signal transduction were found to be induced by 134Cs: a bax inhibitor-1 like protein and a putative auxin-regulated protein (Tables 2 and 3). The bax inhibitor-1 protein is a mammalian apoptosis suppressor and the auxin-regulated protein might have protein kinase activity (Xu & Reed, 1998; Müllauer et al., 2001). Putative auxin-regulated proteins are also induced upon exposure to several other factors such as Cs, salt, osmotic stress and UV radiation (Craigon et al., 2004; Hampton et al., 2004; http://affymetrix.arabidopsis.info/narrays/geneswinger.pl). Recently, it has been shown that an expressed sequence tag of this putative auxin-regulated protein (At2g33830) was up-regulated in A. thaliana by bacterial colonization (Cartieaux et al., 2003). Auxin, an important plant growth regulator, is associated with cell division, growth, maturation and cell differentiation (Trewavas, 2000). Auxin is known to regulate gene expression via protein degradation, including the ubiquitin system (Dharmasiri & Estelle, 2004; Weijers & Jürgens, 2004), and a putative ubiquitin-conjugating enzyme E2-like protein mRNA (At2g33770) was induced upon 134Cs application (Table 2).
Genes coding for cell defence, stress response and detoxification (Table 1) include a DNA-excision repair protein with an induction factor of up to 1.4 (Table 3). DNA-excision repair proteins are involved in the recognition of DNA single-strand breakages that will be excised and then replaced. In addition, we analyzed the Mre11 protein transcript that is part of the DNA recombination complex, important for the repair of DNA double-strand breakages (Paull & Gellert, 1998). Semiquantitative RT–PCR of Mre11 protein transcripts resulted in a significant induction factor of 1.3 (P = 0.03) with a 134Cs activity of 60 Bq cm−3. The importance of genes involved in DNA double-strand repair has been shown in Arabidopsis mutants that were sensitive to γ-radiation (Hefner et al., 2003). The up-regulation of thioredoxin and catalase mRNAs (Tables 2 and 3) indicates again the involvement of oxidative stress and the production of ROS following treatment with 134Cs. Interestingly, the identical thioredoxin gene was also up-regulated following treatment with bleomycin plus mitomycin C (Chen et al., 2003). The idea of ROS production is further supported by semiquantitative RT–PCR for superoxide dismutase (SOD) transcripts that resulted in the up-regulation of ROS transcripts by a factor of 1.4 (P = 0.02). For SOD and catalase transcripts, an ozone-induced accumulation is well known (Langebartels et al., 2002). Similarly, small heat shock proteins were induced upon oxidative stress and γ-radiation in tomato, as well as by ozone treatment of parsley plants (Eckey-Kaltenbach et al., 1997; Banzet et al., 1998). In addition, the induction of an extensin-like protein transcript (Table 2) was also found upon ozone treatment of Scots pine, Norway spruce and European beech (Schneiderbauer et al., 1995). Finally, genes involved in temperature and salt stress were also induced by 134Cs (Table 2; Gong et al., 2001; Xiong et al., 2002). In a latter group, genes are summarized that cannot exactly be classified.
Our study, on the impact of low levels of ionizing radiation on gene expression in A. thaliana, will not provide a global view on transcriptional changes of potentially 134Cs-responsive genes. However, the low cost of the SSH method, in combination with quantitative real-time RT–PCR, will still contribute to a transcriptional profiling also with regard for organisms lacking complete genomic sequences. Comparing 134Cs- and 133Cs-induced transcripts, only five genes were found to be induced by both stress factors (Fig. 2; Sahr et al., 2005). These genes encode SOD, thioredoxin, phi1-like protein, extensin/extensin-like protein and a putative glycine-rich protein. This clearly indicates the involvement of ROS as a common signal for both Cs isotopes. However, only phi1-like protein had an identical accession number, and an induction of SOD and thioredoxin was shown by RT–PCR. Regarding the noncommon-induced transcripts, 133Cs resulted mainly in an up-regulation of genes involved in transport and homeostasis, in protein synthesis, folding and modification (Sahr et al., 2005). Similarly, transporters and defence genes were also strongly induced in Cs-intoxicated Arabidopsis (Hampton et al., 2004). In contrast, 134Cs showed an up-regulation of genes involved in the DNA excision and repair system, and in homologous recombination events. In addition, genes influencing the cell cycle and the cytoskeleton were only induced with ionizing radiation (Tables 2 and 3; Sahr et al., 2005). These results demonstrate that even the low chronic ionizing radiation of 134Cs influences important cellular responses that differ from transcriptional effects influenced by the incorporated nonradioactive stable isotope, 133Cs.