Relative expression of genes related with cold tolerance in temperate rice at the seedling stage

Low temperature is one of the main abiotic stresses affecting rice yield in Chile. Alterations in phenology and physiology of the crop are observed after a cold event. The objective of this work was to study the relative expression of genes related with cold stress in Chilean cultivars of rice. For this, we analyzed the expression of candidate genes using real-time polymerase chain reaction (rtPCR), after exposure to cold of the rice cultivars Diamante-INIA and Zafiro-INIA and one experimental line from INIA ́s Rice Breeding Program in Chile, Quila 241701, with known high cold tolerance. For analysis, the Spanish cultivar, Susan, was used as check tolerance and Oryzica 1 as susceptible check. Oxidative stress was measured through lipid peroxidation. To find mechanisms of cold tolerance in Chilean cultivars, we determined the relative expression of genes related with oxidative stress, such as superoxide dismutase (SOD), glutathione reductase (GR), and catalase (CAT). Lipid peroxidation allowed the measurement of the physiological stress level of the genotypes under study. The results indicate that lipid peroxidation in Oryzica 1 was higher than in the other genotypes. No differences were observed in expression levels of gene-encoding SOD and GR between genotypes. Contrary to expected results, high level of the gene-encoding CAT enzyme in Oryzica 1 after cold stress was observed. Future experiment, related with enzymatic activity and non-enzymatic antioxidant mechanism, are necessary to elucidate the relationship between cold stress and expression levels of gene-encoding antioxidant enzyme in Chilean rice germplasm.


INTRODUCTION
Chile is one of the coldest regions where rice is cultivated in the Southern hemisphere and therefore, low temperature is the most important abiotic stress affecting rice production.Low temperatures are responsible for up to 40% yield loss in rice production in Chile (Alvarado and Hernaiz, 2007).Similar results are observed in temperate areas such as Australia and Japan (Jacobs and Pearson, 1994;McDonald, 1994).The rice area in Chile has a Mediterranean climate and it is located between the Maule (35° S lat) to Biobío region (36º S lat).Rice cultivation occurs from October to March mainly in a flooding regime by using pre-germinated seeds (Alvarado and Hernaiz, 2007).
Lately, the direct sowing started to be used by the farmers.Rice plants are frequently exposed to cold temperatures at germination, seedling and the reproductive stage.Commonly, the minimum temperature is below 12°C, below the optimum (20-30°C) required for rice cultivation (Yoshida et al., 1996).In all cases, the level of damage associated to cold stress will be dependent on the development stage and the intensity of the cold event (Jacobs and Pearson, 1994).For example, at the germination stage, chilling can cause slow growth and severely reduce seedling vigour (Ali et al., 2006); at the vegetative stage may cause, lower tillering (Shimono et al., 2002), high mortality (Mackill and Lei, 1997;Andaya and Mackill, 2003;Baruah et al., 2009), nonuniform maturity of the crop (Shimono et al., 2004), and yellowing of the plants (Shimono et al., 2002;Ali et al., 2006;Baruah et al., 2009).Night temperatures at germination and seedling stages are below 10°C (Alvarado and Hernaiz, 2007).In this condition, Chilean rice cultivars such as Zafiro-INIA and Diamante-INIA have shown high cold tolerance, especially at the seedling stage (Pulver, 2002).
At molecular levels, cold stress can cause an increase of reactive oxygen species (ROS) causing oxidative stress (O'Kane et al., 1996;Mittler, 2002;Apel and Hirt, 2004).The oxidative stress occurs due to an imbalance between oxidative species formation and antioxidant defences of the cell.In abiotic stress the principal causes of oxidative imbalance are the absorption of more photons than can be used by the photosynthetic apparatus (chloroplast) and the over-reduction of electron transport chain (mitochondria) (Møller et al., 2007).Accordingly, levels of ROS in the cell increase rapidly (Tsugane et al., 1999).ROS is produced in the electron transport chains of chloroplasts and mitochondria, some peroxidases and oxidases and excited chlorophyll molecules (Cheeseman, 2007).In the mitochondrial electron transport chain, NAD dehydrogenase and cytochrome reductase generate the superoxide anion, which is reduced to H 2 O 2 by dismutation.H 2 O 2 can react with reduced Fe 2+ and Cu + to produce hydroxyl radicals (HO • ) (Sweetlove and Foyer, 2004).In chloroplasts, ROS is generated by univalent oxygen reduction in the Mehler reaction in the electron transport chain (Asada, 1996;Edreva, 2005).Generated O 2 •-is protonated to HO 2 • and these molecules can initiate lipid peroxidation.On the external stromal membrane surface O 2 •may be dismutated to H 2 O 2 and O 2 (spontaneously or by Superoxide dismutase) (Grant and Loake, 2000).Through the Fenton reaction, H 2 O 2 is transformed to HO • , the most reactive species.To avoid production of HO • , H 2 O 2 is scavenged by the antioxidant machinery (Noctor and Foyer, 1998;Dat et al., 2000).
Plants have developed numerous mechanisms to reduce the effects of the oxidative damage.Some of them are associated with enzymatic ROS scavenger systems, for example, superoxide dismutase (SOD), which catalyzes the conversion of superoxide to H 2 O 2 and O 2 , catalase (CAT), which convert H 2 O 2 to water, glutathione reductase (GR) and the ascorbate peroxidase (APX), which eliminates the ascorbate H 2 O 2 GSH cycle (Saruyama and Tanida, 1995;Hammond-Kosack and Jones, 1996;Huang and Guo, 2005;Ahmad et al, 2008).
Real time PCR is a powerful technique for evaluation gene expression based on the detection and quantification of the fluorescence emitted by PCR products accumulated through the amplification process (Higuchi et al., 1993).This technique can be used to quantify gene expression using two strategies: relative and absolute quantification.In relative quantification, gene expression of target gene is compared with one or more reference genes (Pfaffl, 2004).In the case of rice, the expression of the 18S rRNA gene is widely used as a reference gene (Kim et al., 2003;Welsch et al., 2008).
Tolerance to cold is a characteristic dependent on the genotype, which is controlled by more than one gene (Baruah et al., 2009).It has been demonstrated that cold tolerance is associated with capacity to remove or detoxify ROS (Malan et al., 1990).Therefore, the hypothesis of this research was that expression of genes associated with proteins that detoxify ROS (SOD, GR and CAT) is related to cold tolerant genotypes.The overall objective was to analyze gene transcription that encodes antioxidant enzymes in 3 cultivars and one cold tolerant experimental line, at the seedling stage.

Plant and growth conditions
Seeds of rice genotypes Quila 241701 (Experimental line), Diamante-INIA, Zafiro-INIA, Susan (tolerant control) and Oryzica1 (susceptible control) were germinated on absorbent paper with a solution of the fungicide Benlate (2 ppm) at 28° C in the dark for 3 days.After, germination seedlings were transplanted into plastic pots of 500 mL with clay soil (Vertisol), fertilized with 100mg of Urea, 55 mg of potassium chloride and 45 mg of triple superphosphate (45% of P 2 O 5 ).
Plants were grown in a greenhouse with a photoperiod of 14 h of light and 10 h of dark, at 28°C.The plants were illuminated with artificial light using metal halide lamps with a radiation of 300 µmol photons m -2 s -1 .Seedlings with 3 to 4 expanded leaves were treated at 5°C in a cold chamber in darkness (Baruah et al., 2009) for 72 h.Finally, three plants per pot were sampled and frozen in liquid nitrogen and stored at 80°C until analysis.

Lipid peroxidation
The protocol described by Dionisio-Sese and Tobita (1998), with modifications, was used to determine lipid peroxidation; the main purpose was to measure malondialdehyde (MDA) concentration.
The leaves of the seedlings were ground into a fine powder with liquid nitrogen and weighed.A volume of 500 µL of 50 mM potassium phosphate buffer at pH 7 was added to the sample that was stirred in a Vortex and homogenized with a homogenizer (Heidolph, Diax 900).The extract was centrifuged at 16,000 g for 30 min at 4°C.This process resulted in a 300 µL aliquot of extract that was mixed with 300 µL 0.5% thiobarbituric acid (TBA) in 20% trichloroacetic acid (TCA).Subsequently, the mixture was heated at 95°C for 30 min and reaction was stopped by placing tubes on ice.After, the samples were centrifuged at 10,000 g for 10 min.Absorbance of 532 and 600 nm was determined in the supernatant with spectrophotometer.The MDA concentration was determined by the extinction coefficient, which is equal to 155 mM -1 cm -1 using the standard equation for weight in grams for each of the samples: MDA = [(A532 -A600)/ 155 mM -1 cm -1 ] [nmol g-1 FW] FW = fresh weight tissue (g)

Analysis of genes by real-time PCR
Total RNA was extracted from 0.1 g of ground tissue using RNA-Solv Reagent (Omega Bio-tek).For removal of genomic DNA, DNase I (Applied Biosystems) treatment was applied and cDNA was retro-transcribed from 2 µg of RNA using a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems).Real time PCR was performed with a Stratagene Mx3000p system using GoTaq Green Master Mix (Promega).Expression was calculated using the formula 2 -Ct (Livak and Schmittgen, 2001) with 18S rRNAas -the Housekeeping gene (Table 1).For amplification, GoTaq Mix from Promega was used according to manufacturer's instructions.For this, 3 ng of cDNA was used for amplification of 18S rRNA, 12.5 ng of cDNA for amplification of SOD, and GR, and 25 ng of cDNA for CAT amplification.Furthermore, dissociation curves were performed in each case to determine the absence of primer dimers and/or nonspecific products.All primers are described in Table 1.

Experimental design and statistical analysis
A completely random experimental design was used with three biological replicates.Also, for all Real time PCR analysis, three technical replicates were used.Two way ANOVA (treatment and genotype as factor of variations), was used for the statistical analysis of the MDA concentration.One way ANOVA was used for the analysis of the relative expression of each gene studied.All data were analyzed using Infostat® software (Di Rienzo et al., 2012).

RESULTS AND DISCUSSION
Lipid peroxidation showed values between 2.94 and 5.57 nmol g -1 PF in the analyzed genotypes (Figure 1).This result was similar to that found by Cabas (2012), in which lipid peroxidation of different rice genotypes showed variations between 1.39 to 5.56 nmol g -1 .Analysis of variance for MDA concentration showed that seedlings of Oryzica 1 were 1.86 times higher than Susan and 1.37 times higher than Diamante-INIA and 1.71 times higher than the control condition.Also, levels of MDA for Diamante-INIA exposed to cold was 1.28 times greater than the control plants and Quila 241701 exposed to cold.
No differences between Susan and Zafiro-INIA seedlings were found in control conditions and cold.Similar results were also observed by Kim and Tai (2011) and Wang et al. (2013), who found differences in MDA levels between tolerant and susceptible seedlings after cold treatment at 9°C for 14 days and at 5°C for three days, respectively.This suggests that cold treatment was enough to cause oxidative stress in the susceptible check, Oryzica 1 (Yun et al., 2010).
Expression levels of the gene-encoding the SOD enzyme in Susan (cold-tolerant genotype) was 2.7 times higher than Quila 241701, but no differences were observed between other genotypes (Figure 2).The enzymatic activity of SOD showed that cold-tolerant genotypes increase enzyme activity with respect to a susceptible genotype (Wang et al., 2013).SOD is an intracellular enzyme that catalyzes the dismutation of O 2 into H 2 O 2 and O 2 .This enzyme represents the first defense mechanism against ROS in comparison with GR and CAT that act in detoxifying H 2 O 2 .Some studies have revealed that tolerant genotypes express a higher level of antioxidant enzyme activity than sensitive genotypes (Guo et al., 2006).
No differences were observed between genotypes in the analysis of expression levels of the gene-encoding GR.This finding is not in agreement with other results that showed that the activity of the GR increases in the cold tolerant-genotypes (Wang et al., 2013).
The analysis of the level of the expression of the geneencoding the CAT enzyme indicated that Oryzica 1 showed 4 to 10 times greater levels of expression than in the other genotypes (Figure 2).This high expression level of the gene-encoding CAT in Oryzica 1 is in agreement with other studies in rice (Wang et al., 2013), where a significant increase in expression levels of this enzyme was observed in a susceptible genotype.
Several mechanisms can contribute to the reduction or prevention of oxidative stress and its deleterious effects (Bonnecarrère et al., 2011).The first is related to prevention of ROS formation by the dissipation of excess energy via xanthophylls cycle involving synthesis of zeatin (Z) and antheraxanthin (A) from violaxanthin (V) (Demming-Adams and Adams, 1992).
A second mechanism is the detoxification of ROS by carotenoids or by the action of antioxidant enzymes (Saruyama and Tanida, 1995;Huang and Guo, 2005).For example, carotenoids such as lutein and neoxanthin have been associated with the process of elimination of ROS (Dall'Osto et al., 2007).Other studies have compared the contribution of enzymatic and nonenzymatic detoxification of ROS in cold susceptible genotypes (Oryza sativa L. spp.indica) and cold tolerant genotypes (Oryza sativa L. spp.japonica).However, the results have shown that there exist a complexity pattern, where enzymes increase or decrease their activity after cold stress (Bonecarreré et al., 2011).In general, the increase in expression levels of enzymes related with detoxification of ROS, can cause a decrease of lipid peroxidation.For example, the reduction of lipid peroxidation in rice plants after cold acclimation is related with the increase of antioxidant defense system as SOD, GR, and CAT (Kuk et al., 2003;Moradi and Ismail, 2007;Prasad, 1996).However, a high level of CAT transcript did not result in lower levels of lipid peroxidation in Oryzica 1.
The low correlation between mRNA, protein levels and enzyme activity could be explained mainly due to alternative splicing of mRNA precursors and posttranslational modifications of proteins (Gygi et al., 1999;Washburn et al., 2003).The low increase in MDA concentration observed in Susan, Quila 241701 and Zafiro-INIA was not associated with the expression levels of genes-encoding antioxidant enzymes.Therefore, high levels of enzyme activities, non-enzymatic or other mechanisms could be related with cold tolerance of these genotypes.The high level of lipid peroxidation in the susceptible genotype Oryzica 1 could be the cause of the increase in the transcript levels of the gene that encodes the CAT enzyme.However, inefficient protein translation or low activity of CAT enzyme could be also explained by the low level of protection against ROS under cold stress in Oryzica 1.

Conclusions
The high level of lipid peroxidation observed in the genotype Oryzica 1 was associated with the susceptibility to cold stress.On the contrary, the low levels of MDA observed in Susan, Zafiro-INIA and Quila 241701 may indicate that these genotypes have a good mechanism to avoid lipid peroxidation after cold treatment.Unexpectedly, no relationship between lipid peroxidation and SOD and GR enzyme was observed.Also, contrary to expected results, high levels of the gene-encoding CAT enzyme in Oryzica 1 were not related to avoid lipid peroxidation.Future experiment related with enzymatic activity and analyses of non-enzymatic mechanism are necessary to elucidate the mechanism of protection against ROS in Chilean rice germplasm.

Figure 1 .
Figure 1.Lipid peroxidation after cold stress in rice seedlings.Letter and asterisk (*), represent differences between genotypes and treatment, respectively (Fisher LSD test, P≤0.05).Error bars represent the standard error of 3 biological replicates.

Figure 2 .
Figure 2. Relative expression levels of gene-encoding SOD (A), GR (B) y CAT (C).Different letters represent differences in expression levels between genotypes.Error bars represent the standard error of 3 biological replicates.

Table 1 .
Primer sequences used for real-time PCR analysis.