A Medicago sativa haem oxygenase gene is preferentially expressed ...

Journal of Experimental Botany, Vol. 55, No. 394, pp. 43±47, January 2004 DOI: 10.1093/jxb/erh020 Advanced Access publication 17 November, 2003

FOCUS PAPER

A Medicago sativa haem oxygenase gene is preferentially expressed in root nodules Emmanuel Baudouin1, Pierre Frendo1, Marie Le Gleuher1 and Alain Puppo1,* 1

Laboratoire de Biologie VeÂgeÂtale et Microbiologie, CNRS FRE 2294, Universite de Nice-Sophia Antipolis, Parc Valrose, F-06108 Nice ceÂdex 02, France Received 27 May 2003; Accepted 11 August 2003

Abstract Haem oxygenases (HO) are ubiquitous enzymes catalysing the oxidative degradation of haem into biliverdin, iron and carbon monoxide. Whereas animal HOs participate in multiple cellular functions including haemoglobin catabolism, antioxidant defence and iron homeostasis, to date, plant HOs have so far only been involved in phytochrome metabolism. The expression of the HO1 gene was studied in Medicago sativa, especially during the interaction with its symbiotic partner, Sinorhizobium meliloti. Transcript accumulation was higher in mature root nodules than in roots and leaves and was correlated to HO1 protein immunodetection. The analysis of HO1 expression following alfalfa root inoculation with S. meliloti indicates that transcripts do not accumulate during the early steps of symbiosis, but rather in the mature nodules. These results correlate with the expression of the leghaemoglobin gene, which encodes the major haem-containing protein present in the nodule. Contrary to its animal counterpart, alfalfa HO1 was not induced by prooxidant compounds including H2O2, paraquat and sodium nitroprusside, suggesting that it is not involved in the antioxidant defence. The results suggest that HO1 could play a role in the alfalfa mature nodule and its involvement in leghaemoglobin metabolism is hypothesized. Key words: Haem, leghaemoglobin, Medicago oxygenase, reactive oxygen species, symbiosis.

sativa,

Introduction Haem oxygenases (HO: EC 1.14.99.3) catalyse the oxidative conversion of haem to biliverdin IXa (BV),

with the concomitant release of carbon monoxide and free iron. HO activity has been detected in different organisms including bacteria, animals and plants (Terry et al., 2002). A major role for HO in animals is associated with haemoglobin degradation and haem recycling in senescent red blood cells. HO is also an important component of the antioxidant cell machinery through the generation of BV, which is rapidly reduced to the potent antioxidant bilirubin (for review, Ryter and Tyrrell, 2000). Indeed, HO1de®cient mouse cells present a hypersensitivity to oxidative stress (Poss and Tonegawa, 1997). Moreover, HO1 expression is induced by an array of pro-oxidant compounds (H2O2, cadmium, paraquat). In addition to these inducers, HO1 expression is activated by haemin, the major product of haemoglobin catabolism (for review, Otterbein and Choi, 2000). Whereas animal HOs act in a catabolic pathway, plant HOs have so far been associated with the biosynthetic pathway leading to phytochrome chromophore formation. This function has been revealed by the analysis of mutants affected in light response in different plant species including Pisum sativum, Lycopersicon esculentum and Arabidopsis thaliana (Chory et al., 1989; Terry and Kendrick, 1996; Weller et al., 1996). In this last species, the hy1 mutant presents long hypocotyls and a chlorotic phenotype when grown in light. HY1 protein is homologous with animal HOs and is able to catalyse haem oxygenation (Muramoto et al., 1999). Moreover, HY1 presents a functional plastid targeting sequence and was localized in chloroplasts. The complete sequence of the Arabidopsis genome revealed the existence of a small HO gene family with three additional members designated HO2, HO3 and HO4. So far, although tissue and/or developmental stage speci®cities have been suggested, the speci®c functions of these different HOs remain unclear. Although haem level is naturally low in plant organs, it may be increased a few hundred fold in leguminous root

* To whom correspondence should be addressed. Fax: +33 4 92 07 68 38. E-mail: [email protected] Journal of Experimental Botany, Vol. 55, No. 394, ã Society for Experimental Biology 2004; all rights reserved

44 Baudouin et al.

nodules. These unique specialized organs result from the symbiotic association between plants belonging to Leguminosae and Rhizobiaceae soil bacteria (for review, Schultze and Kondorosi, 1998). Following recognition and infection processes, bacteria are released into plant cells and subsequently differentiate into bacteroids. Bacteroids have the ability to convert atmospheric dinitrogen into ammonium and, therefore, provide plant cells with a source of reduced nitrogen. This conversion is catalysed by the bacterial enzyme, nitrogenase. To support this process, the actively respiring bacteroids are supplied with oxygen by leghaemoglobin. This haem-containing protein is the most abundant protein found in nodule tissues. The accumulation of leghaemoglobin requires a high level of available haem, and leghaemoglobin synthesis has been correlated with an increased activity of several enzymes involved in the haem metabolic pathway in pea and soybean root nodules (Santana et al., 1998). In this study, the expression of the HO1 gene in Medicago sativa was investigated upon its interaction with Sinorhizobium meliloti, and its possible relation with leghaemoglobin metabolism. Material and methods Plant culture and inoculation Experiments were carried out on Medicago sativa L. var. Europe (alfalfa). For testing the organ speci®city of HO1 gene expression, 1-week-old seedlings grown on vermiculite in controlled greenhouse conditions were inoculated with a Sinorhizobium meliloti strain 1021 culture at 0.1 OD and, subsequently, provided with nitrogen-free nutritive medium. Chemical assays for HO1 induction were carried out on non-inoculated plants provided with KNO3 as a nitrogen source. Short-term infection kinetics were carried out on seedlings grown and inoculated in sterile conditions as previously described (HeÂrouart et al., 1996). Plant organs were harvested at different times, deep-frozen in liquid nitrogen and conserved at ±80 °C. Chemical treatments One-week-old seedlings were harvested from the vermiculite and transferred to Petri dishes containing 20 ml of Fahraeus medium (Fahraeus, 1957) plus the different chemicals. All the chemicals were purchased from Sigma and prepared as stock solutions in water except for haemin, which was prepared in DMSO. Following treatment, roots were excised, deep-frozen in liquid nitrogen and conserved at ±80 °C. Cloning and analysis of HO1 and HO2 cDNA probes M. truncatula expressed sequence tag (EST) collections were screened with the Arabidopsis thaliana HO1 and HO2 protein sequences using the tBLASTN program (Altschul et al., 1997). The sequences obtained were aligned and analysed using the UWGCG software (Devereux et al., 1984). The AW981017 and AL381336 sequences were assembled to construct the HO1 cluster sequence. The AW696919 and BE205128 sequences were assembled to obtain the HO2 cluster sequence. The ampli®cation of the HO1 and HO2 cDNA probes was achieved by reverse transcriptase-polymerase chain reaction (RTPCR) ampli®cation. Total RNA was reverse transcribed using the Omniscript RT kit (Qiagen). PCR ampli®cation was achieved using

HO1 speci®c primers (5¢-GCAATAGGCAATGGCTGCTTC-3¢ and 5¢-GCTGGTCCCTTATCATGATAG-3¢) or HO2 speci®c primers (5¢-ATGTTGTTAACAGCGAAACCCACTCAGC-3¢ and 5¢-CAGAGAGCCCACCAGTCTC-3¢) for 35 cycles of sequential incubations at 95 °C for 0.5 min, 60 °C for 1 min and 72 °C for 1 min in a 50 ml reaction mixture containing 10 pmol of each primer and 5 units of Taq DNA polymerase (Qbiogen). PCR products were inserted in the pGEM-T vector (Promega), sequenced and found to be identical to the corresponding EST clusters. Northern blot analysis RNA was extracted from plant tissues using Trizol (Trizol reagent, Gibco BRL Life Technologies), fractionated on 1% formaldehydeagarose gels, transferred onto Hybond N membranes (Amersham) and hybridized according to standard protocols (Sambrook et al., 1989). Ethidium bromide added to the RNA samples before denaturation was used to assess even loading and transfer of the RNA. Immunodetection of alfalfa haem oxygenase Plant tissues (0.5 g) were ground in an extraction buffer composed of 0.1 M TRIS-HCl, pH 7.5, 1 mM EDTA, 5 mM DTT, 1 mM PMSF, 4 mM leupeptin, and 10% (w/v) glycerol. After centrifugation (1 h, 25 000 g, 4 °C), the soluble protein extracts were used for immunodetection. Proteins were separated by SDS-PAGE (Laemmli, 1970) and transferred to PVDF membranes (Pall). PVDF sheets were blocked in TRIS-NaCl buffer (50 mM TRISHCl, pH 7.5, 150 mM NaCl) containing 5% non-fat milk, and incubated overnight in primary antibodies raised against Arabidopsis thaliana HY-1 (Muramoto et al., 1999), diluted 1:2000 in TRISNaCl buffer plus 1% non-fat milk. The blots were washed and incubated with alkaline phosphatase-conjugated goat anti-rabbit immunoglobulin G. The phosphatase-labelled antigens were visualized with the colorigenic substrate 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium.

Results Medicago sativa HO1 homologue is preferentially expressed in root nodules

The expression of the Medicago sativa HO1 homologue was studied in roots, leaves and nodules (Fig. 1A). Similar levels of transcripts were detected in leaves and roots. Nevertheless, the level of transcripts was signi®cantly higher in 5-week-old nodules, suggesting a preferential expression of HO1 gene in this organ. In the same conditions, the transcripts for Medicago sativa HO2 homologue, another member of the haem oxygenase family, were detected faintly and equally in the three organs tested (data not shown). In order to analyse whether the HO1 gene expression was correlated with the corresponding protein detection, the level of HO protein was studied by western blot hybridization using antibodies raised against Arabidopsis thaliana HO1 protein. As shown in Fig. 1B, a polypeptide with an apparent molecular mass of 30 kDa was detected in the nodule protein extracts. This was similar to the molecular mass of the translation product deduced from the HO1 sequence (28.9 kDa), as well as to the molecular mass of Arabidopsis thaliana HO1 (32.6 kDa). A fainter signal was detected on

Haem oxygenase expression in alfalfa

Fig. 1. Analysis of HO1 expression in M. sativa. (A) Northern blot analysis of HO1 expression. Samples of total RNA (10 mg) were isolated from roots (R), leaves (L) and nodules (N) of 6-week-old M. sativa plants. Visualization of ribosomal RNA (rRNA) was used as the control for RNA loading. (B) Western blot analysis of HO1. Samples of proteins (50 mg) were isolated from the roots (R), leaves (L) and nodules (N) of 6-week-old M. sativa plants.

root protein extracts, which correlated with the results obtained on gene expression. Nevertheless, the signal corresponding to HO had an equivalent intensity in leaf protein extracts, although a signi®cantly lower HO1 transcript level was detected in this organ compared with that in nodules. HO1 expression is not modi®ed during early stages of nodule development

The preferential accumulation of HO1 transcripts in nodules led to the analysis of its expression pattern during the establishment and functioning of M. sativa/S. meliloti symbiosis. As shown in Fig. 2, similar transcript levels were detected in uninoculated roots (0) and in roots harvested 1±14 d after inoculation with Sinorhizobium meliloti. These time points correspond with the stages of symbiosis establishment and nitrogen-®xation initiation. By contrast, a signi®cantly higher level of transcripts is observed in functional, mature nodules harvested 5 weeks post-inoculation. These data were compared with the kinetic of leghaemoglobin Lb1 transcript accumulation. As shown on Fig. 2, Lb1 transcripts are detected slightly after 7 d of inoculation and accumulate strongly in 5-week-old mature nodules. HO1 is not induced by leghaemoglobin degradation products and pro-oxidants

In order to get hints of HO1 regulation, gene expression was analysed following exogenous treatments of alfafa roots with compounds reported to activate HO genes in animal cells. With respect to the role of HO1 in haemoglobin turnover, several reports indicate that haem itself, and the iron released from haem, are potent inducers of HO1 expression. With this knowledge, the expression of HO1 in

45

Fig. 2. Analysis of HO1 expression during the nodule formation process. Northern blot analysis was carried out using samples of total RNA (10 mg) isolated from roots 1, 2, 4, 7, 14, and 35 d postinoculation with S. meliloti. Blots were hybridized with both HO1 (HO1) and leghaemoglobin (Lb1) probes. Visualization of ribosomal RNA (rRNA) was used as control for RNA loading.

Fig. 3. Analysis of HO1expression in response to haemin and iron treatments. Northern blot analysis was carried out using samples of total RNA (10 mg) isolated from roots 3 h and 6 h after treatment with haemin (50 mM) and iron citrate (500 mM). Visualization of ribosomal RNA (rRNA) was used as the control for RNA loading.

alfalfa roots exogenously treated with haemin and iron citrate was analysed. As shown in Fig. 3, a slight increase of transcript level was detected in the control plants. Nevertheless, no signi®cant difference in HO1 transcript accumulation was observed between control and haeminor iron-treated roots 3 h or 6 h after treatment. These data suggest that leghaemoglobin degradation products are not potent inducers of HO1 in alfalfa. HO1 has been widely reported as playing a major part in the antioxidant cell machinery of animal cells, and gene activation was observed upon treatment with a large array of pro-oxidant molecules. To test whether a similar regulation also occurs for plant HO1, the effect of H2O2 and paraquat, an O2- -generating compound, on HO1 expression in alfalfa treated roots was analysed. As shown in Fig. 4, under these conditions a slight accumulation of HO1 transcripts was observed in the control roots, possibly due to the stress generated by the transfer of plantlets to liquid culture medium. Nevertheless, this expression was not signi®cantly modi®ed when either H2O2 or paraquat were added to the incubation medium. Similarly, no signi®cant difference was observed when alfalfa roots were treated with the nitric oxide donor, sodium nitroprusside, a potent activator of animal HO1 (Fig. 4). Taken together, these results indicate that HO1 expression is not induced by reactive oxygen or nitrogen species in alfalfa.

46 Baudouin et al.

Fig. 4. Analysis of HO1 expression in response to different oxidative stresses. For northern blot analysis samples of total RNA (10 mg) isolated from roots 3 and 6 h after treatment with hydrogen peroxide (5 mM H2O2), paraquat (100 mM) and sodium nitroprusside (1 mM SNP) were used. Visualization of ribosomal RNA (rRNA) was used as the control for RNA loading.

Discussion In the present study, the expression of the HO1 gene in the legume plant, Medicago sativa, was analysed. In a ®rst approach, the HO1 expression in different organs was compared. Similar levels of HO1 transcripts were observed in the leaves and roots. Nevertheless, the highest expression was found in root nodules, the highly specialized organs resulting from the symbiotic interaction between legumes and rhizobial bacteria. Interestingly, it was observed that HO2, another member of the haem oxygenase gene family, was expressed at similar levels in leaves, roots and nodules (data not shown). Moreover, the information provided by in silico analysis of Medicago truncatula ESTs (Journet et al., 2002) and available at the M. truncatula website (http://medicago.toulouse.inra.fr/ Mt/EST/DOC/MtB.html) indicates that the Medicago truncatula HO1 homologue is preferentially expressed in mature nodules, and, to a lower extent, in mycorrhizae and elicitor-treated cells. By contrast, HO2 expression is detected in the stems, insect-challenged leaves and nitrogen-starved roots. These data strengthen the hypothesis that the preferential expression of HO1 in nodules is not a general feature of HO genes, but rather a speci®c characteristic of this gene. The immunodetection of the HO protein using antibodies raised against Arabidopsis thaliana HO1 showed that the HO1 protein synthesis is concomitant with the HO1 expression in nodules. However, the experiments also indicated that similar levels of HO protein were detected in the leaves and nodules, although a signi®cant difference in gene expression was observed between these organs. Although the existence of speci®c regulatory mechanisms can not be ruled out, this discrepancy could result from the crossreactivity of HO1 antibodies with different members of the HO family. Indeed, the information deduced from

Arabidopsis thaliana HO sequences indicates that they share a high degree of homology at the protein level. Moreover, they presented similar molecular masses that would not permit their separation by SDS-PAGE. Further analysis would be required to identify fully the different HO proteins detected. In mammals, HO1 has emerged as a major component of cellular stress defence. In particular, a dual antioxidant role has been proposed for this enzyme (Ryter and Tyrrell, 2000). On the one hand, it would limit the pro-oxidant reactions associated with haem accumulation. On the other hand, it indirectly generates bilirubin, which possesses in vitro anti-oxidant properties. Concomitant with this role, many oxidant molecules are potent inducers of HO1 expression. As far as is known, the hypothesis that HOs play a role in antioxidant defence in plants has not been explored. This study indicates that different pro-oxidant molecules were not able to induce HO1 expression in alfalfa roots under the present experimental conditions. These results suggest that HO1 is not regulated by oxidative stress compared with its animal counterpart. Moreover, the production of bilirubin downstream of HO1 activity has not been shown. Although a biliverdin reductase has recently been described in plants, it catalyses the formation of the plant-speci®c bilin 3Z-phytochromobilin, which has not been reported as a potent antioxidant (Frankenberg et al., 2001). Taken together, these data indicate that HO1 is not directly involved in antioxidant defence. They also suggest that, although the production of different reactive oxygen species occurs during Medicago sativa/Sinorhizobium meliloti symbiosis (Santos et al., 2000, 2001), they are not major signals inducing HO1 expression. The role of HOs in plants has only been investigated so far in the context of phytochrome metabolism and light response. Although no information on phytochrome metabolism in nodules has been reported, no ESTs corresponding to PHY genes have been isolated from nodules, suggesting that phytochrome metabolism may be low in these organs. On the other hand, haem metabolism is particularly important for proper nodule functioning, in relation to the synthesis of leghaemoglobin and other haem-containing proteins (O'Brian, 1996). Indeed, the enzymes from the haem biosynthetic pathway are strongly stimulated in mature pea and soybean nodules, leading to increased haem content (Santana et al., 1998). In this context, HO1, which is the ®rst enzyme of haem catabolism in animals, could act in a similar catabolic pathway in plants. The expression data provided are consistent with this hypothesis. Indeed, the HO1 transcript level is steady during the early steps of symbiosis establishment and nodule development, but it increases in 5-week-old nodules. This expression pattern is in accordance with the data reported by Santana et al. (1998), indicating that haem accumulates strongly 4 weeks

Haem oxygenase expression in alfalfa

after infection in soybean. Moreover, a similar pattern of accumulation was observed for leghaemoglobin transcripts, which is only slightly expressed after 2 weeks of inoculation and accumulates strongly thereafter. Leghaemoglobin can represent up to 30% of nodule proteins and a turnover rate of 2 d has been measured in pea nodules (Bisseling et al., 1980). Therefore, HO1 could be involved in leghaemoglobin turnover. Indeed, free haem which is released during leghaemoglobin degradation is toxic, and HO1 could participate in minimizing its deleterious effects. In animals, the ef®ciency of haem degradation is enhanced by the ability of haem itself to activate HO1 transcription (Poss and Tonegawa, 1997). In the present study's experiments, haemin was not able to activate HO1 expression in alfalfa roots. So far, the ability of haemin to induce plant genes has not been demonstrated. Moreover, its ability to penetrate root tissues is unknown. Nevertheless, as preliminary experiments, these data suggest that haemin is not involved in HO1 regulation. The identi®cation of non-symbiotic haemoglobins in a large array of plant species has raised new questions about the role of these proteins in plants (Dordas et al., 2003). Although their function is still unclear, some of them could participate in plant adaptation to stress conditions (Trevaskis et al., 1997). Thus, it would be of interest to investigate further the potential role of HO1 in plant haemoglobin turnover. Acknowledgement We thank Dr T Kohchi for kindly providing the Arabidopsis HO1 antibodies. References Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research 25, 3389±3402. Bisseling T, van Straten J, Houwaard F. 1980. Turnover of nitrogenase and leghemoglobin in root nodules of Pisum sativum. Biochimica et Biophysica Acta 610, 360±370. Chory J, Peto C, Feinbaum R, Pratt L, Ausubel F. 1989. Arabidopsis thaliana mutant that develops as a light-grown plant in the absence of light. Cell 58, 991±999. Devereux J, Haeberli P, Smithies O. 1984. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Research 12, 387±395. Dordas C, Rivoal J, Hill RD. 2003. Plant haemoglobins, nitric oxide and hypoxic stress. Annals of Botany 91, 173±178. Fahraeus G. 1957. The infection of white clover root hairs by nodule bacteria studied by a simple slide technique. Journal of General Microbiology 16, 374±381. Frankenberg N, Mukougawa K, Kohchi T, Lagarias JC. 2001.

47

Functional genomic analysis of the HY2 family of ferredoxindependent bilin reductases from oxygenic photosynthetic organisms. The Plant Cell 13, 965±978. HeÂrouart D, Sigaud S, Moreau S, Frendo P, Touati D, Puppo A. 1996. Cloning and characterization of the katA gene of Rhizobium meliloti encoding a hydrogen peroxide-inducible catalase. Journal of Bacteriology 178, 6802±6809. Journet EP, van Tuinen D, Gouzy J, et al. 2002. Exploring root symbiotic programs in the model legume Medicago truncatula using EST analysis. Nucleic Acids Research 30, 5579±5592. Laemmli UK. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680±685. Muramoto T, Kohchi T, Yokota A, Hwang I, Goodman HM. 1999. The Arabidopsis photomorphogenic mutant hy1 is de®cient in phytochrome chromophore biosynthesis as a result of a mutation in a plastid heme oxygenase. The Plant Cell 11, 335±348. O'Brian MR. 1996. Heme synthesis in the rhizobium±legume symbiosis: a palette for bacterial and eukaryotic pigments. Journal of Bacteriology 178, 2471±2478. Otterbein LE, Choi AM. 2000. Heme oxygenase: colors of defense against cellular stress. American Journal of Physiology Lung Cellular and Molecular Physiology 279, 1029±1037. Poss KD, Tonegawa S. 1997. Reduced stress defense in heme oxygenase 1-de®cient cells. Proceedings of the National Academy of Sciences, USA 94, 10925±10930. Ryter SW, Tyrrell RM. 2000. The heme synthesis and degradation pathways: role in oxidant sensitivity. Heme oxygenase has both pro- and antioxidant properties. Free Radical Biology and Medicine 28, 289±309. Sambrook J, Fritsch EF, Maniatis TA. 1989. Molecular cloning: a laboratory manual, Vols 1, 2, 3. New York: Cold Spring Harbor Laboratory Press. Santana MA, Pihakaski-Maunsbach K, Sandal N, Marcker KA, Smith AG. 1998. Evidence that the plant host synthesizes the heme moiety of leghemoglobin in root nodules. Plant Physiology 116, 1259±1269. Santos R, HeÂrouart D, Puppo A, Touati D. 2000. Critical protective role of bacterial superoxide dismutase in Rhizobium± legume symbiosis. Molecular Microbiology 38, 750±759. Santos R, HeÂrouart D, Sigaud S, Touati D, Puppo A. 2001. Oxidative burst in alfalfa±Sinorhizobium meliloti symbiotic interaction. Molecular Plant±Microbe Interactions 14, 86±89. Schultze M, Kondorosi A. 1998. Regulation of the symbiotic root nodule development. Annual Review of Genetics 32, 33±57. Terry MJ, Kendrick RE. 1996. The aurea and yellow-green-2 mutants of tomato are de®cient in phytochrome chromophore synthesis. Journal of Biological Chemistry 271, 21681±21686. Terry MJ, Linley PJ, Kohchi T. 2002. Making light of it: the role of plant haem oxygenases in phytochrome chromophore synthesis. Biochemical Society Transactions 30, 604±609. Trevaskis B, Watts RA, Andersson CR, Llewellyn DJ, Hargrove MS, Olson JS, Dennis ES, Peacock WJ. 1997. Two haemoglobin genes in Arabidopsis thaliana: the evolutionary origins of leghemoglobins. Proceedings of the National Academy of Sciences, USA 94, 12230±12234. Weller JL, Terry MJ, Rameau C, Reid JB, Kendrick RE. 1996. The phytochrome-de®cient pcd1 mutant of pea is unable to convert heme to biliverdin IXa. The Plant Cell 8, 55±67.