Genome-scale identification of MLO domain-containing genes in soybean ( Glycine max L . Merr . )

In plants, powdery-mildew-resistance locus o (Mlo) genes encode proteins that are calmodulin-binding proteins involved in a variety of cellular processes. However, systematic characterization of this gene family in soybean (Glycine max L. Merr.) has not been yet reported. In this study, we identified MLO domaincontained members in soybean and examined their expression under phytohormone treatment and abiotic stress conditions. A total of 20 soybean Mlo genes were identified (GmMlo1-20), which are distributed on 13 chromosomes, and display diverse exon-intron structures. Phylogenetic analysis indicated that the Mlo family can be classified into four subfamilies. Sequence comparison was used to reveal the conserved calmodulin-binding domain (CaMBD) in GmMLO proteins. The expression of GmMlo genes was influenced by various phytohormone treatments and abiotic stresses, suggesting that these Mlo genes have various roles in the response of soybean to environmental stimuli. Promoter sequence analysis revealed an overabundance of stress and/or phytohormone-related cis-elements in GmMlo genes. These data provide important clues for elucidating the functions of genes of the Mlo gene family.

Recently, Mlos have attracted increasing attention because of their association with the stress responses in plants.In barley (Hordeum vulgare), homozygous recessive mutant alleles of the Mlo gene can confer durable and broad-spectrum pathogen resistance to the biotrophic powdery mildew fungus (Jørgensen, 1992;Piffanelli et al., 2002Piffanelli et al., , 2004)).Loss of Arabidopsis MLO2 plant produced resistance against multiple powdery mildew species (Consonni et al., 2006(Consonni et al., , 2010)).The same defense mechanism was found in tomato MLO1 (Bai et al., 2008).Another main role of Mlo in plants is in the regulation of cell death; for example, mutant mlo plants exhibit spontaneous mesophyll cell death associated with partly accelerated leaf senescence (Wolter et al., 1993;Piffanelli et al., 2002;Stein and Somerville, 2002;Consonni et al., 2010).An increasing number reports suggest that Mlos also have additional functions in response to abiotic stimuli.For example, there are some evidences to suggest that Mlo transcript abundance increased in response to wounding, paraquat treatment, a wheat powdery mildew-derived carbohydrate elicitor, salt stress and mannitol treatment (Piffanelli et al., 2002;Feechan et al., 2008;Konishi et al., 2010).However, despite the apparent association of MLOs with plant abiotic stresses, their roles in plants remain elusive.
Soybean (Glycine max L. Merr.) is one of the most economically oil crops worldwide.Based on a comparison with MLOs in other plant species, those in soybean should also be encoded by the same multi-gene family.However, Mlo genes have not yet been identified in soybean.Sequencing of the soybean genome provides an opportunity to identify previously uncharacterized genes.In this study, a bioinformatics approach was used to identify all members of the Mlo gene family in soybean.Their genomic organization, Ca 2+ -binding motif location, phylogenesis, cis-elements in promoter regions and the expression profiles under various abiotic and hormone stresses were analyzed in detail.The results presented here provide important information for future function studies of Mlo genes.

Identification of Mlo members in soybean
The HMMER (version 2.3.2) program was used to search all MLO domain-containing protein genes in the soybean genome databases (http://www.phytozome.net/soybean.php,phytozome, Release v4.0) with the HMM (Hidden Markov Model) profile of the MLO domain (PF03094) (http:// pfam.sanger.ac.uk/Software/Pfam;Finn et al., 2006) and with an E-vlaue < 1.0e-200 and more than 450 amino acid residues as the cutoff.The predicted MLO sequences were then compared against the conserved MLO domain in the Pfam and SMART (http://smart.embl-heidelberg.de; Letunic et al., 2004) databases.
Database search and sequence analysis Expressed sequence tags (ESTs) were subjected to a BLAST search against the identified Mlo sequences to determine genuine EST hits.The exons-introns structures in the differ-ent genes were determined by comparing of the cDNAs with their corresponding genomic DNA sequences.Open reading frame (ORF) analyses were performed with the ORF Finder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html).The exons and introns were determined by using the Gene Structure Display Server (GSDS) tool (http:// gsds.cbi.pku.edu.cn/).TM segments were predicted by using the TMHMM2 tool (http://www.cbs.dtu.dk/services/TMHMM/; Krogh et al., 2001).Motifs were identified statistically using online MEME (http://meme.sdsc.edu).CaM-binding motif analysis was identified according to Kim et al. (2002b).The chromosomal positions of GmMlo genes were also retrieved from phytozome soybean genome databases (http://www.phytozome.net/soybean.php,phytozome).Ten proteins-encoding genes from regions flanking of each GmMlo gene were retrieved from phytozome soybean genome databases to identify segmental replication events of GmMlos.
Protein sequence alignments and phylogenetic analysis Multiple alignments of protein sequences were performed using Clustal W (Accelrys Inc., USA).A phylogenetic tree was constructed using neighbour-joining (NJ) algorithms.Bootstrapping was performed 1000 times to obtain support values for each branch.

cis-elements identification
To analyze putative cisacting regulatory DNA elements (i.e.cis-elements) in the promoters of GmMlo genes, online PLACE 26.0 (http:// www.dna.affrc.go.jp/PLACE/index.html;Lescot et al., 2002) was used to analyze a 2,000-bp sequence upstream of the full-length cDNAs or predicted CDS extracted from the genomic sequences.

Plant materials, abiotic and phyhormone treatments
The plant materials used in this study included the soybean cultivars 'Williams'.Plants were grown in a controlled environment chamber (200 μmol photons m -2 s -1 , 14 h light/10 h dark per day at 22 ± 2°C).
In the abiotic treatment, 14-day-old seedlings were treated at either 40°C or 4°C, or with either 5% polyethylene glycol (PEG) or 200 mM NaCl.In the 40°C treatment, the leaves were freeze-stored 1 h after the heat shock treatment.In the 4°C treatment, the leaves were collected 6 h after treatment.For the PEG and NaCl treatments, the leaves were then frozen in liquid nitrogen after 24 h of treatment.Untreated seedlings were used as controls.

Identification of Mlo genes in soybean
To identify all the putative genes of the Mlo gene family in soybean, BLAST was used to search the soybean genome databases (http://www.phytozome.net/soybean.php,phytozome) using the HMM Profile of the MLO domain as a query.By using this approach, 20 putative soybean Mlo genes were identified.The EST database provided confirmative transcript information for all Mlo genes.These genes were designated sequentially from GmMlo1 to GmMlo20, according to their genome location (Table 1).The information of GmMlos, including accession numbers of the full-length cDNAs and proteins, and their physical locations on the chromosomes are listed in Table 1.Diverse exon-intron structures were identified by comparing the full-length cDNAs or predicted CDS with the genomic sequence of GmMlo genes in the soybean genome databases (Fig. 1).GmMlo exons ranged in number from 13 to 16.There were 11 GmMlo genes containing a 5' or 3' un-translated region (UTR).The 20 GmMlo genes were unevenly located on the 13 chromosomes (Fig. 2) and the number of genes ranged from one to three ranged per chromosomes.There was only one group, GmMlo18/19 that had a clustered distribution on the soybean chromosome, whereas the remainder had a scattered distribution.Meanwhile, nine GmMlo genes including GmMlo1, 11, 17, GmMlo2, 3, 18 and GmMlo12, 13, 14 were identified on segmental duplicated regions.
Sequence analysis of GmMLO proteins GmMLO proteins have a typically MLO domain.The length of the MLO proteins varied greatly, from a minimum of 469 (GmMLO5) to 623 amino acids (GmMLO18) (Table 1).The multiple alignments of the amino acids of GmMLOs showed considerable sequence diversity and the overall pairwise sequence identity was generally <60% (data not shown).However, more identity was found among the MLO domain of the GmMLO proteins (Fig. 3).TM structures were predicted by using TMHMM2 and showed that the GmMLO proteins contained five to eight TM segments (Table 1); these putative TM segments were located on GmMLOs, based on sequence alignment (Fig. 3).An exact match for CaMBD, an important motif for the calcium response, was found on the C-terminal cytoplasmic tail of all the GmMLO proteins, with the exception of GmMLO16.Highly conserved tryptophan (Trp) residues and hydrophobic residues located at positions 1, 8 and 14 were observed in the CaMBD-containing GmMLO proteins (Fig. 3).

Phylogenetic analysis of MLO proteins
To gain insight into the potential function of different GmMLO pro-  teins, a phylogenetic tree was constructed to compare the relationship between MLOs from other plant species (20 in soybean; 15 in Arabidopsis; 9 in maize; 11 in rice (but not OsMLO10); 3 in barley; 6 in wheat; and 3 MLO proteins from other species) (Fig. 4).The results showed that the MLO family comprises four subfamilies and a single divergent lineage (AtMLO3).Subfamilies III and IV showed strong bootstrap support for a sister group relationship, and contained in total, 14 GmMlo genes.Four GmMlo genes (GmMlo1,5,11,17) belonged to subfamily III and ten GmMlo genes (GmMlo2,6,8,10,12,13,14,18,19,20) belonged to subfamily IV.Other GmMlo9 and GmMlo15 belonged to subfamily I and four GmMlo genes (GmMlo3, 4, 7, 16) belonged to subfamily II.

Expression of GmMlos under abiotic stresses and phytohormone applications
To find out whether GmMlos are response to abiotic stresses, the expression of GmMlo genes was examined in plants subjected to abiotic treatments (heat, cold, drought and salinity).One GmMlo gene was selected from each subfamily.The expression of four GmMlo genes (GmMlo4,15,17,19) was analyzed (Fig. 5).After heat treatment, the transcription levels of four genes were downregulated (Fig. 5A).Under cold conditions, the expression of GmMlo15 and GmMlo17 was upregulated, whereas that of GmMlo19 was downregulated; however, this treatment had no effect on GmMlo4 (Fig. 5B).As shown in Fig. 5C, GmMlo4 was upregulated under drought conditions, whereas the other three genes were downregulated (Fig. 5C).GmMlo15 and GmMlo17 were downregulated under the salinity treatment, although, this treatment had no effect on either GmMlo4 or GmMlo19 (Fig. 5D).
Given that the abiotic stress response involves phytohormone signaling pathways, the expression of GmMlo genes was examined in response to exogenous application of ABA, 6-BA, IAA and GA 3 .As shown in Fig. 5 (E and  F), four genes were suppressed in leaves were sprayed with either 10 mM ABA or 1 mM 6-BA.GmMlo4 and GmMlo15, 17, and 19 expressions was also up-or downregulated in leaves sprayed with 1 mM IAA (Fig. 5G).GmMlo15 showed no response in leaves sprayed with 10 mM GA 3 , whereas GmMlo4 and GmMlo17 expression was induced, but that of GmMlo19 was downregulated (Fig. 5H).The strong response of GmMlos following abiotic stress and phytohormone treatments suggests that they have various roles in the responses of environmental stimuli.

cis-response element analysis of the promoters of GmMlo genes
The promoter region of the GmMlo genes (2,0-kb upstream of the transcriptional start site) was analyzed using the Plant CARE database to identify putative stress-responsive cis-elements.Sequence analysis identified a total of nine types of stress-related cisregulatory elements among GmMlo genes (Table 2).The defense-responsive cis-regulatory elements contained a putative TC-rich repeats element.Abiotic responsive cis-elements, such as the heat-stress element (HSE), were identified in 12 and 16 promoters of GmMlo genes.The drought responsiveness element, which contains MYB binding site and a dehydration responsive element (DRE), was identified in 14 promoters of GmMlo genes; The ABA responsive element (ABRE), gibberellin-responsive element (containing a GARE-motif, TATC-box and P-box) and auxin-responsive element (AuxRE) were also identified in 12, 7 and 2 promoters of GmMlo genes, respectively (Table 2).
Major stress-responsive genes contain putative corresponding stress-responsive cis-elements in their promoters.
For example, the GA 3 -responsive element was identified in the promoter region of GmMlo4 and GmMlo17, which were induced by GA3 treatment, GmMlo19 were also induced in response to GA 3 treatments, and the corresponding elements was identified in its promoters.However, the putative cis-elements detected in GmMlo genes were not closely related with the stress responsiveness.GARE was identified in the promoter of GmMlo15, but the gene did not show a change in response to GA3 treatment.ABA-responsive elements were identified in the promoter of GmMlo19, despite the fact that the gene was downregulated after in leaves treated with ABA.

DISCUSSION
Mlo is a multi-membered gene family.So far, 15, 12, 9 and 7 Mlo genes have been identified in Arabidopsis thaliana, rice, maize and wheat, respectively.
Mean-   (Devoto et al., 2003;Miklis et al., 2007;Liu and Zhu, 2008;Konishi et al., 2010).In the current study, 20 MLO domain-containing sequences were identified in soybean, based on database searches and analysis (Table 1).The GmMlos were similar in character to other Mlo gene members in Arabidopsis, barley and wheat; for example, in terms of their TM and CaMBD structures.The seven TM structures are thought to be a key characteristic of the Mlo gene family (Devoto et al., 2003;Elliott et al., 2005;Feechan et al., 2008).Seven TM segments were also located on GmMLO members based on sequence alignment.However, as predicted by TMHMM2, the GmMLO proteins could contain five to eight TM segments (Table 1).The various TM structures might be the result of structural and allosteric complexity that is too difficult to be predicted by computational methods (Devoto et al., 1999;Kenakin, 2009).Therefore, the occurrence of seven TM segments is also thought to be a main characteristic among GmMLO proteins.The CaMBD motif located on the C-terminus in MLO family members has been identified as an important Ca 2+ signal response motif that modulates defence reactions against powdery mildew (Kim et al., 2002a(Kim et al., , 2002b)).Among the 20 predicted GmMlo genes, 19 genes have the conserved CaMBD motif.GmMlo6 did not show any detectable CaMBD motif, but it may be a special Mlo gene in soybean genome, also, it may be a pseudogene.GmMlo genes were distributed on 13 soybean chromosomes, with a mainly scattered distribution expect one group existed a clustered distribution.Gene families could arise through tandem amplification or segmental duplication of chromosomal regions.Tandem amplification results in a clustered distribution, whereas segmental duplication results in a scattered distribution (Schauser et al., 2005).Therefore, it is suggested that segmental duplication is the main amplification method that occurs in GmMlo family members; in the current study, three groups of GmMlo genes were identified from one segmental duplication event.The result is similar to AtMlos and OsMlos family members, which also mainly result from segmental duplication (Devoto et al., 2003;Liu and Zhu, 2008).Meanwhile, the existence of multiple Mlo gene copies in soybean is not surprising, soybean is a paleopolyploid, also, paleopolyploid must have been the factor of gene duplication.Moreover, the relationship of segmental duplication with clustering in the phylogenetic tree was appeared, It seems that members derived from segmental chromosome duplication such as GmMlo1, 11, 17 are clustered together in the same subfamily.
The response of Mlo family members to Ca 2+ signals to modulated plant defence reactions against powdery mil-dew and to regulated cell death has been identified (Kim et al., 2002a(Kim et al., , 2002b;;Bai et al., 2008).Increasing evidences suggests that Mlo genes have additional functions.Their transcription can occur in response to wounding, paraquat, and auxins and phytohormone stress (Piffanelli et al., 2002;Feechan et al., 2008).In the current study, four GmMlos were responsive to various abiotic stresses and phytohormone treatments.Multi-type abiotic-and phytohormone-responsive elements predicted in the promoters of the GmMlo genes were also take as evidence of the various stress-responsive functions of GmMlos.The ABA-, gibberellin-and auxin-responsive elements, in particular, were identified in parts of promoters of GmMlo genes.ABA is an important phytohormone that can convert the initial stress signal, such as drought or high salinity into a cellular response (Fujita et al., 2005;Nakashima et al., 2009).Auxins and gibberellins are also important signal molecules in environmental signal transduction (Sun, 2000;Berleth et al., 2004).In addition, crosstalk between phytohormones and Ca 2+ signals in response to various environmental stresses has been reported ( Van der Meulen et al., 1996;Gao et al., 2002;Yang and Poovaiah, 2003;Navarro-Avino and Bennet, 2005;Qudeimat et al., 2008;Siegel et al., 2009;Kang et al., 2010;Zou et al., 2010).However, this evidence comes mainly from Arabidopsis.To date, many questions remain relating to the mechanisms of MLO proteins.Crosstalk between abiotic stresses and phytohormone signaling pathways increases the complexity of MLO function.Isolation and characterization of Mlo genes on a genome scale could pave the way for functional investigations.The results from the current study provide obvious clues for further elucidating the functions of members of the Mlo genes family in response to environmental stimuli.

Fig. 1 .
Fig. 1.Intron-exon organization of 20 GmMlo genes.Exon(s) are shown in black and spaces between the black boxes correspond to introns.

Fig. 2 .
Fig. 2. Genomic distribution of GmMlo genes on soybean chromosomes.The bars represent the loci with the gene name on the right.A white circle on the chromosome indicates the approximate position of the centromeres.Chromosome numbers are indicated at the bottom of each chromosome.

Fig. 3 .
Fig. 3. Sequence alignments of GmMLO proteins.Dark shading indicates invariant residues.Black or white triangles indicate nonvariable Trp residues and highly conserved hydrophobic amino acids in the CaMDB domain.

Fig. 4 .
Fig. 4. Phylogenetic analysis of MLO proteins from soybean and other plant species.The Arabidopsis CRT1 protein (NP_001031199) was selected as the outgroup.The tree shows the four major phylogenetic subfamilies (numbered I to IV and marked with different alternating background to make subfamily identification easier) with high predictive value.At, Arabidopsis thaliana; Br, Brassica rapa; Ca, Capsicum annuum; Hv, Hordeum vulgare; Le, Lycopersicon esculentum; Os, Oryza sativa; Pp, Physcomitrella patens; Ta, Triticum aestivum; Zm, Zea mays.

Fig. 5 .
Fig. 5. Expression of GmMlo genes in response to abiotic treatments and phytohormone applications.(A) heat treatment, (B) cold treatment, (C) drought treatment, (D) salinity treatment, (E) ABA treatment, (F) 6-BA treatment, (G) IAA treatment and (H) GA3 treatment.'0' represents a simulated treatment.The horizontal axis represents the time (h) after either abiotic or phytohormone treatments.Error bars represent the standard deviation from three replicated experiments.

Table 2 .
Putative regulatory cis-element sequences in the promoter of GmMlo genes Mlo genes have been identified and isolated from barley, tomato Lycopersicon, pepper Capsicum and cabbage Brassica, respectively No.20090097120023) and Science and Technology Foundation of Guizhou province (No. [2010] 2089).