Gene Expression Analysis Suggests Temporal Differential Response to Aluminum in Coffea arabica Cultivars

Gene Expression Analysis Suggests Temporal Differential Response to Aluminum in Coffea arabica Cultivars Bárbara Regina Bazzo, Ariane de Lima Eiras, Daiane Mariele DeLaat, Walter José Siqueira, Jorge Maurício Costa Mondego, et al. Tropical Plant Biology An International Journal devoted to original research in tropical plants ISSN 1935-9756 Volume 6 Number 4 Tropical Plant Biol. (2013) 6:191-198 DOI 10.1007/s12042-013-9120-6

1 23

Your article is protected by copyright and all rights are held exclusively by Springer Science +Business Media New York. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy Tropical Plant Biol. (2013) 6:191–198 DOI 10.1007/s12042-013-9120-6

Gene Expression Analysis Suggests Temporal Differential Response to Aluminum in Coffea arabica Cultivars Bárbara Regina Bazzo & Ariane de Lima Eiras & Daiane Mariele DeLaat & Walter José Siqueira & Jorge Maurício Costa Mondego & Carlos Augusto Colombo

Received: 14 February 2013 / Accepted: 23 May 2013 / Published online: 11 June 2013 # Springer Science+Business Media New York 2013

Abstract Aluminum (Al) is a limiting factor of crop yields on acidic soils. Ion aluminum (Al3+) acts primarily in plant root system retarding its growth and development, leading to the reduction of lateral roots number, and consequently the decrease of vegetal production. Most of coffee producing areas are located in acidic soils, which have Al3+ contents enough to damage plant development. Despite the advances in the understanding of physiological and genetic mechanisms of Al tolerance/susceptibility, few are known about Al ion action in coffee plants. This report describes the expression analysis of genes related to aluminum stress in germinating seeds of two cultivars of C. arabica (Catuaí Amarelo IAC 62 and Icatu Vermelho IAC 4045) when challenged with Al3+. In silico analyses of Brazilian Coffee Genome Project (BCGP) database were used to select genes previously found to be related with Al-stress. The expression profile of these genes in Catuaí and Icatu was evaluated through Quantitative PCR (qPCR). Based on our data, we suggest that both analyzed cultivars displays mechanisms of resistance or exclusion, which occurs outside the cell excluding Al3+ assimilation, and mechanisms of tolerance that occurs inside the cell after Al3+ absorption. The major difference is the timing of activation of each mechanism. While Catuaí tends to use resistance mechanisms in early stages of stress, Icatu uses tolerance strategies. In late stages, both

Communicated by: Alan Carvalho Andrade B. R. Bazzo : A. d. L. Eiras : D. M. DeLaat : W. J. Siqueira : J. M. C. Mondego (*) : C. A. Colombo (*) Centro de Pesquisa e Desenvolvimento em Recursos Genéticos Vegetais, Instituto Agronômico de Campinas, CEP 13075-630, Campinas, SP, Brazil e-mail: [email protected] e-mail: [email protected] J. M. C. Mondego e-mail: [email protected]

cultivars seem to display tolerance mechanisms, but Icatu also displays Al-exclusion strategy. Keywords Aluminum . Coffee . Gene Expression . Stress . Quantitative PCR Abbreviations APX Ascorbate Peroxidase BCGP Brazilian Coffee Genome Project CAT Catalase CS Citrate Synthase MDH Malate Dehidrogenase OA Organic Acid OAA Oxaloacetate RQ Relative Quantification ROS Reactive Oxygen Species SOD Superoxide Dismutase

Introduction Aluminum (Al) toxicity is one of the major constraints on crop productivity on acidic soils. In soils with pH above 5, Al is precipitated principally in gibsit form (Al(OH)3). When the soil pH drops below 5, Al ion (Al3+) is solubilized into the soil solution becoming phytotoxic. The major symptom of Al3+ toxicity is a rapid inhibition of root growth, which is directly translated into reduced plant vigor and yield (Kochian et al. 2004). It is proposed that plants have developed two mechanisms to overcome Al3+ stress: mechanisms of resistance or exclusion, which occur outside the cell (apoplast) through the avoidance of Al3+ assimilation, and mechanisms of “truetolerance” that occurs inside the cell after Al3+ absorption (Yang et al. 2000; Silva et al. 2001a; Kochian et al. 2004; Maron et al. 2008; Inostroza-Blancheteau et al. 2012). These

Author's personal copy 192

two mechanisms can co-exist and operate in parallel in plants (Ryan et al. 2011). In the case of apoplastic mechanisms, the defense against Al3+ is mediated by the intracellular increasing and consecutive exudation of organic acids (OAs), which are capable of complexing ions Al3+ (Kochian et al. 2005; Delhaize et al. 2012). Curiously, ion magnesium (Mg2+) enhances OA concentration in the root tips and stimulates OA secretion from the roots exposed to toxic Al level (Silva et al. 2001b; Chen et al. 2012). On the other hand, intracellular defense mechanisms against Al3+ are linked to oxidative stress combat, since this ion induces the expression of genes encoding antioxidants enzymes (Boscolo et al. 2003; Vitorello et al. 2005; Panda and Matsumoto 2010). Coffee (Coffea arabica) plants provide one of the most consumed beverages in the world and is ranked as the second world commodity (only supplanted by petroleum oil). Due to its global social-economical value, there is an increasing interest in providing innovative tools for coffee breeding programs, which, at most, are concerned with the augment of coffee production that is essentially affected by biotic and abiotic stress (Waller et al. 2007). Countries that are leaders in coffee production such as Brazil, Vietnam, Indonesia and Colombia are located in tropical areas, which are rich in acidic soils (Kochian et al. 2004). Since coffee plants are considered more productive in soils with pH in the range of 6.0 to 6.5 (Fullin and Dadalto 2001), the application of lime (calcium carbonate) has been used as an efficient way of increasing soil pH in coffee plantations, consequently avoiding damages caused by Al ion. However, this practice is onerous for small farmers, which are responsible for most of world coffee production (Waller et al. 2007). As discussed by Ryan et al. (2011) the combination of lime with the use of Al3+-resistant germplasm would be an interesting management strategy. Some reports described genetic variability of C. arabica plants to Al ion (Braccini et al. 1998; Rodrigues et al. 2006; Mistro et al. 2007). For instance, Rodrigues et al. (2006) showed that upon incubation with Al occurs an increasing of Al3+ concentration in the subsurface of C. arabica cv. Catuaí Amarelo IAC 62, suggesting exclusion mechanisms of resistance. On the other hand, the cv. Icatu Vermelho IAC 4045 maintained the concentration of Al3+ in the roots subsurface, probably by internalizing this ion (Rodrigues et al. 2006). Ramírez-Benitez et al. (2008, 2009, 2011) have been thoroughly studying the metabolic processes that occur in response to aluminum in C. arabica, however using cell suspension lines. In order to further explore the molecular mechanisms of Al3+ tolerance in coffee plants, the major goal of this work was the evaluation of the expression of genes related with Al-stress in two C. arabica contrasting phenotypes in response to Al 3+ stress (Catuaí and Icatu) when germinating seeds were challenged with Al ion.

Tropical Plant Biol. (2013) 6:191–198

Results and Discussion Root Growth Evaluation Based on data of Rodrigues et al. (2006), we decided to check the development of seedlings of C. arabica cv. Catuaí Amarelo IAC 62 and the cv. Icatu Vermelho IAC 4045 under Al-stress. Studies conducted by Eiras (2010) showed that Catuaí and Icatu seedlings had notable phenotypical differences in root growth after eight days of treatment in AlCl3 (370 μM) (data not shown). Such remarkable difference in root growth led us to evaluate the possible early mechanisms responsible for these contrasting phenotypes. For that, we performed two independent time-course experiments at which seedlings of both cultivars, all of them containing 3 cm of root growth, were transferred to nutrient solution containing 370 μM of AlCl3, except control samples at which AlCl3 was not added. After 1, 12 and 48 h of stress, the seedlings were removed from the solution and primary root apexes were collected and frozen in liquid nitrogen. RNA was extracted and cDNA synthesized for qPCR experiments. Data Mining and Primer Design Seven Al-related genes (four encoding antioxidant enzymes, two encoding proteins associated to organic acids synthesis and one encoding a magnesium carrier) were identified and selected for differential expression analysis. Contigs were selected according to similarity to abiotic stress related genes (Table 1). Quantitative PCR (qPCR) Quantitative PCR (qPCR) was used to evaluate gene expression of C. arabica putative Al3+-tolerance related genes in two independent biological replicates. Target genes with RQ (relative quantification; see material and methods) values greater than or equal to 2 were considered significantly up-regulated, because these data demonstrate an increase equal or greater than 100 % in the expression of analyzed genes (Fig. 1). Consequently, only treatments with RQ ≥ 2 were discussed. Replicate experiments concerning Germin-Like protein (GLP, oxalate oxidase) and Magnesium transporter (MGT1) expression indicates very distinct RQ at each time point, leading to high standard deviations. However, the expression profiles of the replicates were strikingly similar. For that reason, we present the expression data of the replicates of each gene separately (Fig. 1). Below, we discuss the hypothetical function of each gene evaluated, according to their functional categories.

Author's personal copy Tropical Plant Biol. (2013) 6:191–198

193

Table 1 Proteins related to Al-stress and orthologues sequences (contigs) from Brazilian Coffee Genome Project (BCGP; http://www.lge.ibi.unicamp.br/ coffea) selected for primer synthesis by PRIMER EXPRESS program (Applied Biosystems) Proteinsa

Contigs of C. arabica databank (BCGP)

ASCORBATE PEROXIDASE (APX) Q42564 (APX3_ ARATH) SUPEROXIDE DISMUTASE (SOD) O81235 (SODM_ARATH) CATALASE (CAT) Q42547 (CATA3_ARATH) CITRATE SYNTHASE (CS) Q9SJH7 (CISY3_ARATH) MALATE DEHYDROGENASE (MDH) Q9ZP06 (MDHM1_ARATH) GERMIN-LIKE PROTEIN (GLP) Q9LEA7 (GL18_ARATH)

1441

1e −121

5351

1e −98

Mg TRANSPORTER (MGT1) Q9S9N4 (MRS21_ARATH) ACTIN (ACT) P93775 (P93775_STRAF) a

E-value tBLASTn BCGP

6750

0.0

17520

0.0

Forward primer sequence

Reverse primer sequence

5′-GAGAAAGTGATGGG CTGCTAAAAC-3′

5′-AAGCTCAACTAAAC GGCGAAATT-3′

5′-TTCGTCTCCTTTTT CCATTTCG-3′

5′-GGGCTTTTCTGGTG ACTAGAGTTC-3′

5′-CGGCCATCGAGTGCTTTC3′ 5′ -TCGTGCAACTATTC CAGATCATG-3′

5′-CAGACGGAGGCAC CAGAATT-3′ 5′ -TGCAAATTGAGTC ATTGGATGAG-3′

5′-ACCAAAATCCCCG AAACACA-3′

5′ -TGGTGAGGATTTA AAAAGGGACTT -3′

3026

1e −158

10522

8e −60

5′- AGCAAATCGACAAG GTTCTAATGTT-3′

5′AAGTGTTGAGCCCA GCTAGATTG-3′

2232

3e −79

5′- CCGGAGGGCCAACCA-3′

5′- CAGCTGCCTTCTTT ATCAATCCA-3′

9902

1e −177

5-′CATGAAGATCCTTA CTGAAAGAGGGT-3′

5′-CGCTCTGCTGAG GTGGTGA-3′

UNIPROT code retrieved from www.uniprot.org

Oxidative Stress Related Genes Contigs 5351, 10522, 6750 and 1441 represent genes potentially related with metabolic pathways of oxidative stress. They are similar to a Superoxide Dismutase (SOD, EC number 1.15.1.1), a Germin-Like protein (GLP, oxalate oxidase. EC number 1.2.3.4), a catalase (CAT, EC number 1.11.1.6) and an Ascorbate Peroxidase (APX, EC number 1.11.1.11), respectively. Al3+ increases the production of reactive oxygen species (ROS), especially superoxide. SOD and GLP are enzymes able to scavenge this superoxide radical, producing toxic H2O2 plus O2. H2O2 is detoxified by its peroxidation performed by CAT and APX enzymes (Apel and Hirt 2004). In our results, SOD expression increases during the first hour of stress, decreases after 12 h in both cultivars and then increases again after 48 h stress only in Icatu (Fig. 2). GLP is only up-regulated in Icatu after 12 h (Fig. 2). CAT expression increases (7-fold) after 12 h in Icatu, then decreases; whereas, CAT expression increases after 48 h (5-fold) in Catuaí (Fig. 2). Finally, APX expression decreases during Al3+ presence in Catuaí, and conversely, increases during treatment in Icatu (Fig. 1). In summary, genes encoding antioxidants enzymes (GLP, CAT, APX) are more expressed in Icatu than in Catuaí during Al3+ stress. It is likely that SOD

alleviates superoxide production during early stress in both plants (Xu et al. 2013). However, its expression decays being replaced by GLP action in Icatu. To deal with the putative intense production of H2O2 triggered by GLP activity in Icatu, CAT expression is enhanced. After CAT decreasing, APX is upregulated for H2O2 detoxification (Fig. 2). Overall, the most pronounced expression of ROS detoxification associated genes in Icatu suggests most intense cellular responses to oxidation in this cultivar in relation to Catuaí. Genes Related with Organic Acids Contig3026 is similar to Medicago sativa NeMDH gene (Nodule Enhanced Malate Dehydrogenase), encoding malate dehydrogenase metabolic enzyme (MDH, EC number 1.1.1.37), involved in malate synthesis. This chemical process involves a phosphoenolpyruvate carboxylation (PEPC) to oxaloacetate with ATP additional formation. Then oxaloacetate is reduced to malate by MDH (Kenyon et al. 1985) (Fig. 2). Transformed alfalfa plants (Medicago sativa) with MDH gene, and submitted to Al3+ stress showed enzyme increasing of activity and concentration (Tesfaye et al. 2001) and also displayed a strong Al3+ tolerance compared with untransformed plants (Wang et al. 2010). This gene has higher expression in Catuaí after 1 h of stress, and then

Author's personal copy 194

Tropical Plant Biol. (2013) 6:191–198

Fig. 1 Expression profiles of genes upon Al ion challenge. Analyses were performed by qPCR and C. arabica actin gene was used as endogenous control to normalize data. Black columns: Catuaí; Gray columns: Icatu. Two replicates are depicted. Dashed lines indicate RQ=2

decreases. On the other hand, after 48 h of stress, MDH expression increases in Icatu, but not in a significantly way taking in account the two replicas (2-fold) in this time point only in Icatu (Fig. 1).

Magnesium-mediated alleviation of aluminum toxicity was observed in many of plant species (Watanabe and Okada 2005; Yang et al. 2007). Mg2+ is an important co-factor of enzymes related to OA synthesis such as malate dehydrogenase and malate synthase (Milne and Cooke 1979; Smith et al. 2003). Recently, Chen et al. (2012) detected that when a putative rice Mg transporter gene (OsMGT1) was knocked out, the tolerance to Al decreased in this plant. The Alinduced inhibition of root elongation in those knockout lines was recovered by addition of Mg. Additionally, the expression of this transporter in roots was rapidly enhanced by Al increasing Mg uptake and Mg concentration in the root cell sap (Chen et al. 2012). Our data show an upregulation of a C. arabica homologue of MGT1 (Contig 9974) after 12 and 48 h specifically in Catuaí plants (Fig. 1). MDH and MGT1 were upregulated in early stages of Alstress (1 h and 12 h, respectively) indicating that the synthesis of malate and the possible increasing of magnesium by

Author's personal copy Tropical Plant Biol. (2013) 6:191–198

195

Fig. 2 Model of response against Al ion in C. arabica cultivars Catuaí and Icatu. Expression of aluminum stress genes at three times (a -1 h; b 12 h; c - 48 h). MDH: malate dehydrogenase; MnSOD: manganese superoxide dismutase; PEP: phosphoenolpyruvate; PEPC:

phosphoenolpyruvate carboxylase; OA: organic acid; MS: malate synthase; Oxo: oxalate oxidase; GLP: germin-like protein; MGT: Magnesium Transporter; CAT: catalase; APX: ascorbate peroxidase; MDHA: monodehydroascorbate; CS: citrate synthase

the uptake of this ion could help alleviate Al effects. On the other hand, CS was upregulated in late responses (48 h) in Icatu, suggesting that citrate exudation and maintenance of Al3+ outside the plant may occur in this cultivar (Fig. 2).

Even though our data corroborate with Rodrigues et al. (2006), which suggest that in Catuaí the mechanisms of compartmentalization and exclusion are acting at the root, while Icatu displays internal tolerance, we speculate that both mechanisms of Al-tolerance are triggered during seed germination of those cultivars, but in different stages. Catuaí displays Al-compartmentalization mechanisms in early stages avoiding Al entrance and its resulting damages to the cell, whereas Icatu tries to deal with Al stress by increasing antioxidative enzymes. In late responses, Catuaí and Icatu express anti-oxidative genes, but Icatu would also produce OAs to block Al access to the cytoplasm. Perhaps, the entrance of Al in the cells of Icatu roots, especially in early germination stages, can retard root growth when we compare it to Catuaí (Eiras 2010). Our data coincide with Ramirez-Benitez et al. (2009) which detected that exists variation of Al-tolerance mechanisms between C. arabica genotypes. Despite the difference between experimental models (C. arabica germination seeds and cell protoplasts) such results support that C. arabica cultivars display differential responses to Al-stress. Our results confirm previous analyses that identified genetic variability among C. arabica cultivars in relation to Al3+ tolerance, but also indicate that such variability can result in differential gene expression. It is known that cv. Icatu Vermelho is result of an introgression of C. canephora in C. arabica cv. Bourbon Vermelho (Fazuoli et al. 2008). Perhaps such difference regarding genetic background influences in the differential Al-stress response of the analyzed cultivars. Further experiments will focus on the physiology of the aluminum tolerance/susceptibility in coffee plants and

A Model for Differential Response to Al3+ in C. arabica Cultivars Based on our data, we propose a hypothetical model that discerns Catuaí and Icatu seedling response to toxic Al. In the first hour, both cultivars increase SOD expression to lead with ROS triggered by Al stress. However, Catuaí increases MDH expression, producing malate that could act as a scavenger of Al outside the cell (Fig. 2a). After 12 h, there is a high increment of GLP in Icatu. It should lead to the increasing of H2O2 in the cytoplasm that has to be detoxified. This fact can explain the high expression of CAT in Icatu after 12 h (Fig. 2b). The augment of anti-oxidative enzymes may cause ROS disturbance in the cell, which leads to cytoplasmatic damages. The increasing of ROS related genes is not observed in Catuaí at this moment, maybe for the action of malate outside the cell (Fig. 2b). Additionally, the toxic effects of Al ions that could enter Catuaí cells can be alleviated by magnesium import by MGT, which is upregulated at 12 h (Fig. 2b). At 48 h after stress, Catuaí continues to upregulate MGT and increases CAT expression for the combat of ROS in this late stage (Fig. 2c). On the other hand, APX is now upregulated in Icatu indicating that H2O2 is still causing damages to this cultivar. Finally, Icatu would activate citrate synthesis to try to complex Al outside the cell (Fig. 2c).

Author's personal copy 196

Tropical Plant Biol. (2013) 6:191–198

in high throughput analyses of transcriptional responses to aluminum in C. arabica.

Methods Plant Material, Growth Conditions and Al Treatment Biological tests were performed with two C. arabica commercial cultivars previously studied concerning Al responses, Catuaí Amarelo IAC 62 and Icatu Vermelho IAC 4045 (Rodrigues et al. 2006). Three thousand and five hundred seeds of each cultivar were used in the experiments (amount of seedlings necessary to select the optimal seedlings, eliminating seedlings nongerminated, contaminated and outliers from our root length pattern - 3 cm). Previously, the seeds were sterilized in 5 % sodium hypochlorite (Meirelles et al. 2007), peeled, treated with 70 % Tyran fungicide, germinated on moistened paper and kept at 26 °C in the dark. After achieving 3 cm long (identical length to all seedlings), 1,200 seedlings were transferred to nutrient solution (Hoagland and Arnon 1950) modified according to Braccini et al. (1998), at 25 °C with a photoperiod of 12 h day/12 h night. After 24 h of acclimatization, AlCl3 (370 μM) (Mistro et al. 2007) was added to solutions (except for negative control). The seedlings were grown until 48 h under the same acclimatization conditions, being the pH daily adjusted to 4.2 with 0.1 mol.L−1 HCl or 0.1 mol.L−1 NaOH. After one hour of treatment, 100 seedlings of each plant material were removed through primary root apexes excision and immediately frozen in liquid nitrogen. This procedure was repeated after 12 h and 48 h, with two replicates per treatment. 100 seedlings of control samples (without AlCl3 - negative control) were collected in the same times described for the two replicates. The experiments were arranged in a completely randomized design. RNA Extraction and cDNA Synthesis Total RNA from the two independent experiments was isolated using the TRIZOL reagent (Invitrogen, Carlsbad, California, USA), according to the manufacturer’s instructions. RNA quality was assessed by electrophoresis on agarose 1 % (w/v) gels stained with GelRed (Biotuim, Inc., Hayward, California, RQ ¼ EðtargetÞ ΔCt targetðCt control–Ct treatmentÞ

USA) and visualized using FLA-3000 (Fluorescent Image Analyzer, Fujifilm Corporation, Tokyo, Japan). RNA quantification was performed by spectrometry at 260–280 nm. To purify of extracted RNA was used RNeasy Mini Kit ® kit (Qiagen, Valencia, California, USA) with specific manufacturer’s protocol to plants. Total RNA (3 μg) was treated with 2 units of DNase I (Invitrogen) and reverse transcribed using the kit Superscript III Reverse Transcriptase (Invitrogen). Data Mining Sequences of proteins associated with plant response to abiotic stress were collected from UNIPROT database (http:// www.uniprot.org) and used as queries in tBLASTN (Altschul et al. 1997) searches against contigs from C. arabica BCGP database (Brazilian Coffee Genome Project; Vieira et al. 2006; Mondego et al. 2011; http://www.lge.ibi.unicamp.br/coffea). The resulting alignments were filtered by a threshold E-value