Wheat non-specific lipid transfer protein genes display a complex pattern of expression in developing seeds

Biochimica et Biophysica Acta 1730 (2005) 114 – 125 http://www.elsevier.com/locate/bba

Wheat non-specific lipid transfer protein genes display a complex pattern of expression in developing seeds Freddy Boutrot, Anne Guirao, Re´mi Alary, Philippe Joudrier, Marie-Franc¸oise Gautier* INRA, UMR Polymorphismes d’Inte´reˆt Agronomique, 2 place Viala, 34060 Montpellier Cedex 01, France Received 4 April 2005; received in revised form 31 May 2005; accepted 23 June 2005 Available online 13 July 2005

Abstract Nine cDNA clones encoding non-specific lipid transfer proteins (nsLTPs) were isolated from Triticum aestivum and Triticum durum cDNA libraries and characterized. One cDNA is predicted to encode a type 2 nsLTP (7 kDa) while others encode type 1 nsLTPs (9 kDa). All encoded proteins contain an N-terminal signal sequence and possess the characteristic features of nsLTPs. The genomic structures of the wheat nsLtp genes show that type 2 TaLtp7.1a, TaLtp7.2a and type 1 TaLtp9.2b genes lack introns while the other type 1 genes consist of one intron. Construction of a phylogenic tree of Poaceae nsLTPs shows that wheat nsLTPs can be divided into eleven distinct groups and are closely related to barley sequences. Using reverse transcriptase-polymerase chain reaction (RT-PCR) analysis, the expression patterns of nine nsLtp genes were studied during wheat seed development and germination. We identified three different profiles of nsLtp gene transcript accumulation. Whereas TdLtp7.1a, TdLtp9.4a and TdLtp9.7a transcripts were detected during all maturation stages, TdLtp7.2a, TdLtp9.2a, TdLtp9.3a, TdLtp9.5a and TdLtp9.6a transcripts were only present in the first and TdLtp9.1a in the last stages of seed development. Moreover, these nine wheat nsLtp genes are not seed-specific and are also expressed in the coleoptile of young seedlings. The present study revealed the complexity of the wheat nsLtp gene family and showed that the expression of nsLtp genes is developmentally regulated in the seeds, suggesting a specific function for each of the corresponding proteins. D 2005 Elsevier B.V. All rights reserved. Keywords: Evolution; Gene expression; nsLtp gene; RT-PCR; Seed development; Wheat

1. Introduction Plant non-specific lipid transfer proteins (nsLTPs) are a structurally related family of proteins that were originally defined by their in vitro ability to transfer lipids between membranes [1]. Eight conserved cysteine residues are involved in the formation of four disulfide bridges. The bonds stabilize four to five alpha-helices that enclose a hydrophobic cavity. NsLTPs have been reported in most plant species; they are 9 or 7 kDa basic proteins that describe type 1 and 2 nsLTPs, respectively [2]. NsLTPs were first suggested to be involved in membrane biogenesis [1]. However, all plant nsLTPs so far identified possess an N-terminal signal peptide that would direct the proteins to the endoplasmic reticulum and lead their entry to the * Corresponding author. Tel.: +33 499612366; fax: +33 499612348. E-mail address: [email protected] (M.-F. Gautier). 0167-4781/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bbaexp.2005.06.010

secretory pathway [3]. As they are extracellulary located, nsLTPs could be involved in cutin synthesis carrying hydrophobic cutin monomers [4]. On the other hand, numerous studies emphasize that nsLTPs protect plants from infections by fungi [5], bacteria [6] or viruses [7]. Since the nsLtp genes are induced in the early stages of plant –pathogen interaction, the related proteins could be involved indirectly in plant defense by creating a mechanical barrier of cutin and involved directly due to their intrinsic antibiotic properties. It has been recently demonstrated that a sunflower LTP is able to induce the permeabilization of fungal spores [8]. NsLTPs may also be involved in cell signaling as shown for a wheat nsLTP1 that interacts with tobacco plasma membrane receptors previously been identified as elicitin receptors [9,10]. An arabidopsis nsLTP could be involved in the transport of a systemic acquired resistance mobile signal [11] whereas a Zinnia nsLTP may act as an inhibitor of proteasome to

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

protect cells from proteases released during the differentiation of tracheary elements [12]. Plant nsLTPs are encoded by multigenic families with differential expression patterns in a variety of tissues, developmental stages and physiological conditions, as shown in arabidopsis [13], cotton [14], Euphorbia lagascae [15], orange [16], pepper [17] or Poaceae as barley [18], rice [19], sorghum [20] and wheat [21,22].The expression of nsLtp genes is also induced in response to abiotic stresses such as drought [23], low temperatures [24 – 26], saline treatments [23], wounding [17,25] and exposure to heavy metals [27]. NsLtp gene diversity and their involvement in many aspects of cell growth and development suggest specific functional roles for each of the nsLTPs. Maldonado et al. [11] highlighted the individual function of an arabidopsis nsLTP because with more than 30 nsLtp genes in its genome, the disruption of only one of them leads to failure of the plant’s systemic acquired resistance. In order to understand the increasing complexity of the proposed functions of nsLTPs, it is necessary to characterize each nsLtp gene expression. The objective of this work was to further the analysis of the organization of the nsLtp gene family and of its expression in wheat seeds. Here, we report that at least nine wheat nsLtp genes are transcribed during seed development and that, depending on the gene, different patterns of transcript accumulation are observed. Both gene structure and phylogenetic data underline the complexity of the nsLtp gene family expressed in wheat seed.

2. Materials and methods 2.1. Plant materials Triticum durum Desf. cv. Ne´odur was grown in the field at the INRA experimental station in Melgueil (France). Ears were tagged at anthesis and subsequently collected at different post-anthesis stages (4 to 48 DPA). They were immediately frozen in liquid nitrogen and stored at 80 -C. Seed germination was carried out as previously described [21], time zero corresponding to seeds after imbibition. Seeds, roots and coleoptiles were removed from 1- and 3day-old seedlings, immediately frozen in liquid nitrogen and stored at 80 -C. Leaves were from Triticum aestivum L. cv. Apache. 2.2. DNA extraction Genomic DNA was extracted from 1 g of T. aestivum leaves using the DNeasy Plant Maxi kit (Qiagen, Valencia, CA, USA). Plasmid DNA was extracted using the QiagenTip 100 kit (Qiagen) under conditions specified by the supplier. DNA was checked for quality by gel electrophoresis and quantified by spectrophotometry (Ultrospec 3000, Amersham Biosciences, Buckinghamshire, UK).

115

2.3. Isolation of cDNAs clones encoding wheat nsLTPs We have previously described the construction of cDNA libraries from T. durum [28] and T. aestivum [29] midmaturation seeds. A T. aestivum cDNA library was constructed from 4 DPA seeds in the same conditions as previously described [29]. pTd18 [21], pTd4.90 [30] and pTd6.48 were isolated from the 22 DPA T. durum library. pTd6.48 was isolated by colony hybridization using the maize clone 9C2 [3] as described for pTd4.90 [30]. pTa8.3 was identified from the 23 DPA T. aestivum seeds cDNA library by random sequencing. All the other clones were identified from the 4 DPA T. aestivum seed cDNA library by random sequencing. To obtain non-specific lipid transfer protein cDNA clones spanning the 5V end, pTa4.90, pTa6.48 and pTa232 were isolated by PCR using plasmid DNA prepared from an aliquot of the appropriate cDNA library (corresponding to 3109 cfu) as template. PCRs were done with nsLTP specific antisense-strand primers and the universal T7 promoter sequence (5V-TAATACGACTCACTATAGGG-3V) as sense-strand primer. The antisense-strand primer was 5V-CTACATAGGGATAGAGGCTGCGC-3V for the pTa4.90 sequence, 5V-TTAGAGCTGATCAGCTG-3V for the pTa6.48 sequence and 5V-ACATTACATAGACACGAACGTAGT-3V for the pTa232 sequence. The primers were obtained from MWG Biotech AG (Ebersberg, Germany). Reactions were performed in 100 AL containing 50 ng of library plasmid DNA, 40 pmol of each primer, 250 M of each dNTP, 1X Taq DNA polymerase reaction buffer (Qbiogene, Illkirch, France), 10% (v/v) glycerol and 2 units of Taq DNA polymerase (Qbiogene). In the first step, the samples were denatured at 93 -C for 4 min. This was followed by 36 cycles of 1 min denaturation at 94 -C, 90 s annealing at 60 -C and 2 min elongation at 72 -C. A final cycle with an extension step of 10 min at 72 -C completed the reaction. PCR reactions were analyzed on 1.5% agarose gel and blotted onto Hybond N+ membrane (Amersham Biosciences). PCR fragments that hybridize to the nsLtp probes (pTd4.90, pTd6.48 or pTa232 inserts) were cloned into the pGEM-T vector (Promega, Madison, WI, USA). 2.4. DNA sequencing and protein analysis All cDNA clones were sequenced on both strands using either the Taq Dye Primer Cycle Sequencing kit, the Taq Dye Deoxy Terminator Cycle Sequencing kit or the BigDye Terminator Cycle Sequencing kit (Applied Biosystems, Foster City, CA, USA). Samples were analyzed on an Applied Biosystems 377A DNA Sequencer and sequences were determined using the ABI 373 autoanalysis program. The N-terminal signal sequences of wheat nsLTPs were determined using the SignalP 1.1 neural network using the hidden Markov model [31]. The subcellular location of nsLTPs was predicted with the TargetP 1.0 tool [32]. Sequence identities and similarities were calculated on mature nsLTPs with the program Stretcher [33].

116

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

the pGEM-T Easy Vector (Promega) and sequenced on both strands.

2.5. PCR Coding regions for nsLtp genes were amplified using either plasmid (20 ng) or genomic (100 ng) DNA as template. Reactions were performed in 25 AL containing 200 AM of each dNTP, 0.8 AM of each primer (MWG Biotech AG), 3 mM MgCl2, 1X Taq DNA polymerase reaction buffer (15 mM Tris – HCl pH 8.0, 50 mM KCl), 0.625 unit of Taq DNA polymerase (AmpliTaq Gold, Applied Biosystems), 10% (v/v) glycerol and 0.5% (v/v) Tween 20. Initial template denaturation was at 95 -C for 10 min, followed by 35 (plasmid DNA) or 40 (genomic DNA) cycles of 30 s denaturation at 95 -C, 30 s primer annealing and 1 min extension at 72 -C, ending with 5 min of extension at 72 -C. The primer sequences and annealing temperatures are listed in Table 1. PCR products were fractionated through a 2% agarose gel and visualized by UV fluorescence. The genomic amplicons were purified from agarose (Qiaquick Gel Extraction Kit, Qiagen), cloned into

2.6. RNA extraction and RT-PCR Total RNA was extracted from frozen T. durum seeds at different days post-anthesis (4, 13, 20, 27, 34 and 48 DPA) and post-germination (0, 1 and 3 DPG) as described previously [21]. All total RNA preparations were treated with RNase-free DNase (Promega) in presence of RNasin ribonuclease inhibitor (Promega). RNA quality was checked by gel electrophoresis and then quantified with the RiboGreen RNA quantitation kit (Molecular Probes, Leiden, Netherlands) using a LS50B luminescence spectrophotometer (Applied Biosystems). Total RNA (1 Ag) was reverse transcribed using 20 pmol oligo-(dT)18 and 200 units of M-MLV reverse transcriptase (Advantage RT for PCR kit, Clontech, Palo Alto, CA, USA) under the conditions specified by the supplier. The final

Table 1 Wheat nsLtp-specific primers used for PCR and RT-PCR cDNA origina

Primerb

Nucleotide sequence (5V to 3V)

Detection (gene name)

pTd18

7.1F1 7.1R1 7.1F2 7.2F1 7.2R1 7.2F2 7.2R2 9.1F1 9.1R1 9.1F2 9.1R2 9.2F1 9.2R1 9.2F2 9.2R2 9.3F1 9.3R1 9.3F2 9.4F1 9.4R1 9.4F2 9.4R2 9.5F1 9.5R1 9.6F1 9.6R1 9.6F2 9.6R2 9.7F1 9.7R1 9.7F2 gapdhF1 gapdhR1

CTGCAAGAGCGAGAGCGTGAAC TTCATTCCAGTACCAGCAAAGTC CGTGAGAGTGTCAGACTATAGCT AGGAGTAGACCAAACGTACGTAC GGATCGGACAGAAACATGCGTG GCTGACGGTGTGCATGC GCGGATCCCTAGCAGTGCGGG TACGGTGCACTGTTAGCTACAGACC GATGGATGAGCTGAGCACCAGCTTG TGGGCGACGATCCGTCAAGCTG AAGTCCATAAACACTGGGAGTG GCTCACTACCACTACTATTGCTAGCTTG ATGGCAACAGCGATGGCAG GAGCACTTGCTGCCATCGCTGTT GCTGAGTACACAATCCATATATCAA GCAAGCAAGCCGAAGCACTAG CTGCTCGATCTGTAACAAGGAC AACCTGGCGGGGTCGTTCAAT CGATGGCTCGTCTCAACAGCAAGGC GACATTAAGCTCCACACAAGCGCG CCTGCATGTGTTATTTCAGTAGTGT CTATTCATTAATTGCCTGAACAAG CTCTATCCCCATCAGCACCAAA ATCTGCTCGATCTGTAATAAGGG TAGAGTACAGCACAATGGCGCC TTCACCCATGACTACAGTAGTAGC CGACAAGATACAGTGAGCGACCG CGAAATACAGGTATCGTAGCTACAG CTCCAAGTGCAACGTCG ACATTACATAGACACGAACGTAGT CCGTGTGTCTGTCGACGTACG TAACTGCCTTGCTCCTCTTGCT GTTTCCCTCAGACTCCTCCTT

gDNA (TaLtp7.1a) and cDNA gDNA (TaLtp7.1a), cDNA and RT (TdLtp7.1a) gDNA (TaLtp7.2a) and cDNA gDNA (TaLtp7.2a) and cDNA RT (TdLtp7.2a) RT (TdLtp7.2a) gDNA (TaLtp9.1b) and cDNA gDNA (TaLtp9.1b) and cDNA RT (TdLtp9.1a) RT (TdLtp9.1a) gDNA (TaLtp9.2b) and cDNA gDNA (TaLtp9.2b) and cDNA RT (TdLtp9.2a) RT (TdLtp9.2a) gDNA (TaLtp9.3a) and cDNA gDNA (TaLtp9.3a), cDNA and RT (TdLtp9.3a) gDNA (TaLtp9.4a) and cDNA gDNA (TaLtp9.4a) and cDNA RT (TdLtp9.4a) RT (TdLtp9.4a) gDNA (TaLtp9.5a), cDNA and gDNA (TaLtp9.5a), cDNA and gDNA (TaLtp9.6a) and cDNA gDNA (TaLtp9.6a) and cDNA RT (TdLtp9.6a) RT (TdLtp9.6a) gDNA (TaLtp9.7a) and cDNA gDNA (TaLtp9.7a), cDNA and RT (TdLtp9.7a) RT (TdGapdh) RT (TdGapdh)

pTa176

pTa4.90 and pTd4.90

pTa6.48 and pTd6.48

pTaD2

pTa8.3

pTa260 pTa360

pTa232

Gapdh d a b c d

Ta (-C)c 66 RT (TdLtp7.1a) 64 68 68 60 68 70 66 66 RT (TdLtp9.3a) 68 72 65 RT (TdLtp9.5a) RT (TdLtp9.5a)

66 62 66 65

RT (TdLtp9.7a)

cDNA clone used to design primers. pTa and pTd cDNA clones were obtained from Triticum aestivum and T. durum libraries, respectively. F: Forward primer; R: Reverse primer. Ta: annealing temperature of the PCR reaction. Gapdh: Glyceraldehyde-3-phosphate dehydrogenase (EMBL accession number AF251217).

65 62

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

solution was brought to 100 AL and 4 AL aliquots from the reverse transcription reaction were used in each PCR reaction using the AdvanTaq Plus PCR kit (Clontech). Sequences of all PCR primers are given in Table 1. Wheat glyceraldehyde-3-phosphate dehydrogenase (Gapdh) gene (EMBL accession number AF251217) was used as internal constitutively expressed control. The amplification was for a total of 30 cycles, which was within the logarithmic range of amplification. Control reactions without reverse transcription were included to ensure that PCR products were not the results of amplification of contaminating DNA. The PCR products were analyzed by running one fifth of each reaction on a 2% agarose gel.

117

using neighbor-joining analysis. Protein distance matrices were computed using Protdist and the Jones – Taylor – Thornton model. The neighbor joining method [36] was implemented using Neighbor. Support for nodes was estimated by the bootstrap procedure using 1000 resamplings of the data. The unrooted phylogenetic trees were graphically displayed using the Treeview 1.6.6 program [37].

3. Results 3.1. Isolation and characterization of cDNAs encoding wheat nsLTPs

2.7. Sequence alignments and phylogenetic analysis A systematic search was performed by examining the EMBL and GenBank databases for Poaceae nsLTPs (http:// www.srs.ebi.ac.uk and http://www.ncbi.nlm.nih.gov, respectively). Including the nine wheat nsLTPs deduced from the cDNA clones isolated in this work, the Poaceae nsLTPs family consists of sixty-eight non-redundant proteins. For the construction of the phylogenic tree, the mature nsLTPs primary sequences were aligned with the ClustalW 1.83 program [34], and the relationship between cereal nsLTPs was then investigated with the Phylip 3.6a3 package [35]

We previously characterized cDNA clones (pTd4.90 and pTd18) encoding wheat nsLTPs isolated from a tetraploid Triticum durum mid-maturation seed (22 DPA) library [21,30]. To explore the diversity of nsLtp genes expressed in wheat seed, we pursued our study on the hexaploid wheat, Triticum aestivum. cDNA clones were isolated from 4 DPA (pTaD2, pTa232, pTa260, pTa360 and pTa176) and 23 DPA (pTa4.90, pTa6.48 and pTa8.3) T. aestivum seed libraries. pTd4.90, pTd6.48 and pTa232 were not full-length cDNAs and contained a 5V truncated open reading frame. In order to obtain full-length cDNA sequences we took

Table 2 Characteristics of nsLTPs isolated from wheat or deduced from cDNA clones NsLTP type

Clone

Nomenclature Gene

2

1

pTd18 / / / pTa176 pTd4.90 LTP1500 / pTa4.90 pTd6.48 pTa6.48 pTaD2 pTa8.3 pTa260 pTa360 pTa232 / / TaLTP1 TaLTP2 TaLTP3

TdLtp7.1a TdLtp7.1b TaLtp7.1a TaLtp7.1c TaLtp7.2a TdLtp9.1a TaLtp9.1a TaLtp9.1a TaLtp9.1b TdLtp9.2a TaLtp9.2b TaLtp9.3a TaLtp9.4a TaLtp9.5a TaLtp9.6a TaLtp9.7a TaWbp1a TaWbp1b TaLtp1 TaLtp2 TaLtp3

Deduced protein Protein

TdLTP7.1a TdLTP7.1b TaLTP2G TaLTP2P TaLTP7.2a TdLTP9.1a TaLTP1500 TaLTP9.1a TaLTP9.1b TdLTP9.2a TaLTP9.2b TaLTP9.3a TaLTP9.4a TaLTP9.5a TaLTP9.6a TaLTP9.7a TaWBP1A TaWBP1B TaLTP1 TaLTP2 TaLTP3

Signal peptide

Mature protein

aa

aa

MM (Da)

pIa

29 / / / 26 23d 26 / 26 13d 25 24 29 24 36 28 / / 25 25 29

67 67 67 67 67 90 90 90 90 90 90 93 93 91 93 92 94 94 90 90 93

6979 7006 6979 7046 6971 9607 9607 9607 9609 8686 8686 9231 9477 8820 9510 9563 9388 9274 8804 8730 9481

9.84 9.84 9.84 9.84 9.95 9.86 9.86 9.86 9.95 11.09 11.09 10.74 10.85 10.76 11.15 10.10 11.00 10.65 10.85 11.09 11.16

n.a. not available. /, no cDNA clone available, sequence comes from sequenced mature protein. a Cysteine residues were not taking into account for the pI determination. b The TdLTP7.1b protein sequence is not indexed in protein databases but only reported in [38]. c EMBL protein accession number. d Truncated signal peptide because of incomplete cDNA. e GenBank protein accession number.

Reference

EMBL primary accession number

[21] [38] [40] [40] this work [30] unpublished [39] this work this work this work this work this work this work this work this work [61] [61] [46] [46] unpublished

AJ297768 n.a.b P82900c P82901c AJ784895 X63669 AF551849 P24296c AJ784902 AJ784903 AJ784901 AJ784899 AJ784900 AJ784896 AJ784897 AJ784898 AAB32995e AAB32996e AY566607 AF334185 AY226580

118

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

advantage of the orientation of the wheat cDNA libraries. PCRs were performed on the T. aestivum cDNA libraries and pTa4.90, pTa6.48 and pTa232 were obtained by PCR using a sense-strand primer that corresponds to the T7 promoter sequence and specific antisense-strand primers deduced from the corresponding partial cDNA clones. The characteristics of the wheat nsLTPs encoded by these cDNA clones are summarized in Table 2. Two clones (pTd18 and pTa176) only code for type 2 nsLTPs, while all the others encode type 1 nsLTPs. The characteristics of nsLTPs isolated from wheat are also described in Table 2. Because the chromosome assignment of wheat nsLtp genes is unknown, we used a gene nomenclature based on the phylogenetic classification of nsLTP sequences with no relationship to wheat genomes or chromosomes. First, Td and Ta allow to identify T. durum and T. aestivum,

respectively. Second, Ltp9 or Ltp7 indicates what type of nsLTP is encoded (9 for type 1 (9 kDa) and 7 for type 2 (7 kDa)). Third, within each type, the Arabic digit specifies a subfamily and fourth the lower-case italic letter indicates sequential series in a subfamily. All wheat nsLTPs are synthesized as preproteins including an N-terminal sequence of 24 to 36 residues that presents the characteristics of a signal peptide (Fig. 1). The N-terminal of TdLTP7.1a, deduced from the pTd18 cDNA clone, has already been validated experimentally on the corresponding purified proteins [38]. The N-terminal of all other deduced nsLTP sequences was identified by homology analysis or using the SignalP program. For the TaLTP9.6a, one potential cleavage site was located at position 30 giving rise to the N-terminal AEAEAAA. However, because the putative cleavage sites

Fig. 1. Multiple alignment of wheat nsLTPs. The amino acid sequences were deduced from plasmid cDNAs (this work) or indexed in the EMBL protein database. Alignment was performed using the multiple sequence alignment program ClustalW [34]. The predicted signal peptides are in italics, the conserved cysteine residues are black boxed, and for each subfamily highly conserved amino acids are grey boxed. Gaps (dashes) are introduced to optimize alignment. Amino acid residues are numbered relative to the N-terminal of mature nsLTPs. The primary accession numbers of the sequences are given in Table 4.

49.5 48.4 74.4 52.7 52.7 51.6 33.3 44.7 53.2 53.2 TaLTP1 91.1 69.9 35.6 37.8 43.6 42.6 54.3 47.9 50.0 55.3 37.9 38.9 95.7 TaWBP1B 67.0 64.9 66.0 30.9 30.9 42.6 41.5 54.3 47.9 48.9 54.3 37.9 41.1 TaWBP1A 96.8 66.0 63.8 64.9 28.7 30.9 41.3 40.2 40.4 41.5 35.1 35.1 33.7 TaLTP9.7a 52.6 52.6 62.8 61.7 60.6 33.0 31.9 28.7 28.7 34.4 35.1 33.0 39.8 TaLTP9.6a 48.4 52.6 52.6 54.8 52.7 51.1 30.1 31.2 43.0 39.8 50.5 79.6 52.7 TaLTP9.5a 55.9 58.5 64.9 64.9 71.4 71.4 72.0 33.7 35.9 43.0 43.0 48.4 55.9 TaLTP9.4a 72.0 51.1 59.6 64.9 66.0 69.9 69.9 97.8 31.2 33.0 37.6 35.5 51.6 TaLTP9.3a 75.3 87.1 52.1 64.9 60.6 61.7 73.1 73.1 75.3 35.1 30.9 47.3 46.2 TaLTP9.2b 75.3 68.8 74.7 52.7 63.8 68.1 69.1 86.7 88.9 68.8 34.8 38.0 95.6 TaLTP9.1b 65.6 60.2 62.4 59.1 43.6 55.4 57.4 57.4 66.7 63.4 62.4 35.6 34.1 TaLTP9.1a 96.7 65.6 62.4 62.4 61.3 45.7 57.6 57.4 57.4 66.7 65.6 62.4 34.4 31.9

A general view of the evolutionary history of the Poaceae nsLTP family was provided by an unrooted phylogenetic tree (Fig. 2) constructed with the neighbor-joining method using the mature amino acid sequences indexed in the EMBL protein database (Table 4). Since the intron is missing, the translation was found to be inaccurate for the Sb18C08.25 protein accession, and so this sequence was curated for construction of the phylogenic tree. Since the nodes are supported by high bootstrap values, the distribution of nsLTPs highlighted the clustering of the 68 proteins in three major clades. Two clades are in agreement with type 1 and type 2 nsLTPs usually reported. The third clade

Table 3 Percentage of sequence identities (upper right triangle) and similarities (lower left triangle) between wheat mature nsLTPs

3.3. Poaceae nsLTP evolution

48.4 45.2 77.8 52.7 49.5 52.7 33.3 42.6 51.1 51.1 84.4 TaLTP2 69.9 34.4 36.7

3.2. Wheat nsLTP homologies

Proteins from all the wheat nsLTP subfamilies are presented. Some proteins from Table 2 have been omitted as they are identical or present high homologies with selected nsLTPs.

41.9 41.9 48.4 58.1 95.7 54.8 30.9 37.2 48.9 50.0 53.8 50.5 TaLTP3 31.2 33.0

20.0 20.0 21.7 23.4 21.5 23.9 21.5 23.4 20.2 21.3 25.6 25.6 21.5 TdLTP7.1a 82.1

are multiple, the N-terminal end of TaLTP9.6a was assigned by alignment with other wheat nsLTPs leading to a 36 residues signal peptide. Mature type 2 nsLTPs contain 67 residues, whereas type 1 nsLTPs contain 90 to 94 residues. Predicted isoelectric points were basic, ranging from 9.84 to 11.16, values that are in good agreement with those of previously described plant nsLTPs.

There is total conservation of the primary structure of the T. durum TdLTP9.1a deduced from clone pTd4.90 [30], the TaLTP1500 sequence translated from T. aestivum genomic DNA and a type 1 nsLTP purified from T. aestivum seeds [39]. Similarly, the TdLTP9.2a and TaLTP9.2b sequences deduced from T. durum and T. aestivum cDNA clones are identical as are those of the TdLTP7.1a deduced from clone pTd18 [21] and TaLTP2G purified from T. aestivum seeds [40]. In contrast, the TdLTP9.1a and TaLTP9.1b differ by 4 residues. Otherwise, wheat type 1 nsLTP primary structure identities ranged from 28.7 to 95.7% and sequence similarities from 43.6 to 97.8% (Table 3). The highest sequence identities were observed between TaLTP9.4a and TaLTP3 (95.7%), TaWBP1A and TaWBP1B (95.7%) and TaLTP9.1a and TaLTP9.1b (95.6%). Type 2 wheat nsLTPs displayed 74.6% sequence identity and 82.1% sequence similarity. Sequence identities between type 1 and 2 wheat nsLTPs were considerably lower and ranged from 18.1 to 26.7%, sequence similarities ranged from 28.7 to 38.0%. As shown in Fig. 1, wheat nsLTPs share conserved regions including a hydrophobic signal peptide, eight cysteine residues in a conserved position and a lack of tryptophan residue. Wheat nsLTPs exhibit a conserved CXC motif, X being a hydrophilic residue either charged (Lys or Arg) or polar uncharged (Thr, Gly or Asn) for type 1 nsLTPs, whereas for type 2 nsLTPs, X is a hydrophobic residue (Phe). This difference is a characteristic of type 2 nsLTPs versus type 1.

119

20.9 22.0 23.9 23.4 21.3 23.9 20.4 22.3 18.1 18.1 26.7 25.6 21.3 74.6 TaLTP7.2a

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

120

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

Fig. 2. Phylogenetic relationships of the Poaceae nsLtp genes based on sequence alignments of the mature encoded proteins. The analysis was based on alignment of sixty-five unique amino acid sequences. The primary accession numbers for the protein sequences used to compile the tree are given in Table 4. The evolutionary tree was constructed by the Neighbor-Joining method and drawn by the TreeView program. The bootstrap values are shown on each branch (% of 1000 resampled data set); only values greater than 50% are shown. The scale bar corresponds to 0.1 substitution per amino acid.

contained 3 proteins related to male-flower-specific proteins [41]. Interestingly, wheat nsLTPs are found almost all over the phylogenetic tree, except in the third clade. They appear in eleven distinct groups and several wheat nsLTPs appear clustered together. With the exception of TaLTP9.6a, TaLTP9.3a, TaLTP9.5a and TaWBP1s, wheat nsLTPs are closely related to barley sequences. Likewise, barley

HvLTP-BR1 and HvLTP3 are not related to the wheat nsLTPs characterized so far. 3.4. Wheat nsLtp gene structure Because introns have been reported in plant type 1 nsLtp genes, primers were designed from cDNAs to amplify all or

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

121

Table 4 Name and primary accession number of the sixty-eight Poaceae nsLTPs indexed in the EMBL database Organism protein

Accession no.

Triticum aestivum TaLTP2G TaLTP2P TaLTP7.2a TaLTP1500 TaLTP9.1b TaLTP9.2b TaLTP9.3a TaLTP9.4a TaLTP9.5a TaLTP9.6a TaLTP9.7a TaWBP1A TaWBP1B TaLTP1 TaLTP2

P82900a P82901a AJ784895 AF551849 AJ784902 AJ784901 AJ784899 AJ784900 AJ784896 AJ784897 AJ784898 AAB32995e AAB32996e AY566607 AF334185

TaLTP3 AY226580 Triticum durum TdLTP7.1a AJ297768 TdLTP7.1b n.a.b TdLTP9.1a X63669 TdLTP9.2a AJ784903 Hordeum vulgare HvLTP2(7)d X69793 HvLTP-NE AF039024 HvLTP-BR1 U880090 HvLTP1 X59253 HvLTP2(9)d X68655 HvLTP3 X68656 HvLTP4 X68654 HvBLT4.2 2115353Ae HvBLT4.6 2115353Be

HvBLT4.9 HvECLTP HvPLTP HvLTP7a2b HvpKG285 Oryza sativa OsLTP-2 OsNsltp2 Osns-LTP OsLTP1 OsLTP2 OsLTP3 OsLTP4 OsLTP5 OsLPT1 OsLPTII

U63993 AF109195 U18127 X96979 Z37115 U16721 P83210a EPRZa U77295 U31766 Z23271 U29176 AF051369 AY327042 AF017359

OsLPTIII OsLPTIV OsLTP6 OsLTP7 OsLTP8 OsLTP9 OsLTP10 OsLTPa15 OsLTPb1 OsLTPb21 OsLT1 OsYY1 Zea mays ZmmLTP2 ZmPLTP ZmLTP

AF017360 AF017361 AF017358 NM_188488c NM_190373c AP004591 AP005828 X83435 X83434 X83433 AY335485 D50575

MZm3-3 AJ224355 ZmLTP3 AJ006702 Sorghum bicolor SbLTP1 X71667 SbLTP1R X71669 SbLTP2 X71668 Sb18C08.25 AF466200 Bromus inermis BiBG14 AY057932 Setaria italica SiLTP AF439446 Eleusine coracana EcLTP P23802a

P83506a J04176 U66105

n.a. not available. a EMBL protein accession number. b The TdLTP7.1b protein sequence is not indexed in protein databases but only reported in [38]. c GenBank nucleotide accession number. d HvLTP2 proteins have been (7)- and (9)-indexed, with respect to their mature molecular mass, because they share the same name but are different proteins. e GenBank protein accession number.

part of the translated sequences (Table 1) and tested for their specificity against all cDNA clones used in this study (data not shown). The comparison of PCR products from genomic and plasmid DNAs showed that all but TaLtp7.1a, TaLtp7.2a and TaLtp9.2b genes do contain an intron and that intron size varies with the gene (Fig. 3A). This was confirmed by sequencing the cloned genomic amplicons whose alignment with cDNA sequences showed that intron size ranges from 89 to 440 bp starting with the typical 5VGU and ending with AG-3V [42] (Fig. 3B). All the wheat

Fig. 3. Structure of wheat nsLtp genes. (A) Analysis on 2% agarose gel of PCR products amplified from cDNAs (c) and T. aestivum genomic DNA (g) as template with primers specified in Table 1. L, DNA ladder. (B) Size of amplicons and sequence of intron – exon boundaries of nsLtp genes. Exon bases are in capitals and intron bases in small letters.

nsLtp gene intron splice phasings are type 2 and the position of this intron 4 to 7 bases upstream of the stop codon is highly conserved. Wheat nsLtp gene exon sequences have a lower AU content (26 to 36%) than intron sequences (51 to 71%) that are 13 to 24% more U-rich than exons. This is in agreement with the report stating that plant introns are AU rich sequences and are about 15% more U-rich than the flanking exons [43]. 3.5. Wheat nsLtp genes are differentially expressed during seed development In a previous study, we reported marked differences for TdLtp9.1a and TdLtp7.1a gene transcript accumulation during seed development [21]. The aim of the present work was to investigate the steady-state level of transcripts of other wheat nsLtp genes during seed maturation and germination. To this end, total RNA was extracted from developing (4 to 48 DPA) and germinating (1 to 3 days postgermination (DPG)) seeds, from coleoptiles and roots of 3 DPG seedlings and then analyzed by RT-PCR. Gene specific primers were designed (Table 1) using the 3V untranslated sequences to maximize their amplification specificity against gene coding region homologies. Except for TaLtp9.3a and TaLtp9.5a genes whose 3V-untranslated sequences are highly conserved (89.6%), the 3V-untranslated nucleotide sequences are only 23.4 –48.1% identical versus 40.2 – 89.9% for the coding sequences. Nevertheless, the specificity of primers against all cDNA clones was successfully tested except 9.1F2/9.1R2 primers that do not enable discrimination between TdLtp9.1a and TaLtp9.1b genes (data not shown). The absence of DNA contamination in the RNA samples was confirmed by PCR analyses in the

122

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

absence of the RT step (data not shown). In order to compare the level of transcripts of the nine wheat nsLtp genes analyzed, the same amount of RT products was used for all PCRs. Amplification products of the expected size were detected as shown in Fig. 4. The steady-state level of TdLtp gene transcripts varies considerably from gene to gene during wheat seed development and these variations may be related to the gene expression pattern as supported by negligible variations in the transcript level of the wheat Gapdh housekeeping gene used as control. All but TdLtp9.1a gene transcripts were detected at 4 DPA and TdLtp9.5a and TdLtp9.3a genes were found not to be transcribed at later stages of seed maturation. While TdLtp7.2a gene transcripts disappear between 13 and 20 DPA, the TdLtp9.1a gene transcription starts during this period and transcripts are found from this stage to dry seed. Between 20 and 27 DPA, gene transcript levels decrease drastically for TdLtp9.2a and TdLtp9.6a genes and only low levels of TdLtp9.2a gene transcripts are detectable at 34 – 48 DPA, when wheat seeds desiccate. The level of TdLtp9.4a and TdLtp9.7a gene transcripts does not undergo major variations from 4 to 48 DPA, and the TdLtp7.1a gene presents a similar pattern. No major variations in nsLtp gene expression patterns were observed using developing seeds of T. aestivum L. cv. Crousty as plant material (data not shown). Steady state transcript levels of TdLtp genes were also compared in germinating seeds (Fig. 4B). Similar to what was observed in dry seeds, neither TdLtp7.2a, TdLtp9.5a

Fig. 4. Steady-state level of nsLtp gene transcripts during wheat seed development. RT-PCRs were performed on total RNA extracted from wheat seeds (T. durum cv. Ne´odur) using primers specified in Table 1. (A) Seeds harvested at different DPA (4 to 48). (B) Seeds from 0- to 3 day-old seedlings. (C) Coleoptiles (c) and roots (r) excised from 3-day-old seedlings. Each lane is load with an equal amount of PCR product obtained from reverse transcription of 8 ng of total RNA. Two percent agarose gel was stained with ethidium bromide. Wheat Gapdh gene was used as a control of the amount and quality of cDNA. Gene names are given on the left. Gapdh, TdLtp7.1a and TdLtp9.1a profiles are reproduced from ref [21] with permission from Elsevier.

nor TdLtp9.6a gene transcripts were observed in germinating seeds indicating that there is no de novo transcription of these three genes early in germination. In contrast, TdLtp9.2a and TdLtp9.3a genes are de novo transcribed at 1 or 3 DPG, respectively. For TdLtp7.1a, TdLtp9.1a, TdLtp9.4a and TdLtp9.7a genes, the level of transcripts varies only slightly compared to their level in dry seeds. Widespread expression of TdLtp genes was observed in coleoptiles of 3-day-old seedlings whereas only TdLtp7.2a and TdLtp9.1a genes transcripts were detected in root seedlings (Fig. 4C).

4. Discussion In this study, we characterized nine cDNA clones from wheat developing seed libraries encoding nsLTPs. Eight of these are type 1 nsLTPs and the other is a type 2 nsLTP. With the already described TdLTP7.1a [21], TdLTP7.1b [38], TaLTP2P and TaLTP2G [40], the TaLTP7.2a constitutes the fifth type 2 nsLTP identified in wheat. Similar to other plant nsLTPs, wheat nsLTPs share the same features with eight cysteine residues, a basic isoelectric point and a signal peptide predicted to target the proteins to the secretory pathway. Our analysis of cDNA and genomic DNA sequences indicates that, excluding the TaLtp9.2b gene, type 1 nsLtp genes are interrupted by a single intron. In contrast, wheat type 2 nsLtp genes appear to have no intron. This observation suggests that the presence of the intron is an ancestral trait contemporary to the split between type 1 and type 2 nsLtp genes and that the loss of this intron is a TaLtp9.2b gene derived character. Regarding the phylogenetic distribution of nsLTPs (Fig. 2), the wheat TaLTP9.2b is closely related to type 1 nsLTPs and supports the hypothesis of intron loss. Interestingly, in the Poaceae nsLTP cladogram, TaLTP9.2b is clustered with type 1 nsLTPs encoded by intronless genes, i.e., the sorghum SbLtp2 gene [20] and the barley HvBlt4.2, HvBlt4.6 and HvBlt4.9 genes [44]. Using PCR specific primers, we verified that wheat TaLTP1 and TaLTP2 genes related to this cluster do not contain an intron (data not shown). Moreover, barley HvPltp, HvLtp4, HvLtp2(9) and HvpKG285 genes that are also related to this cluster may be encoded by intronless genes. Type 1 and type 2 nsLTPs have already been described and our phylogenic representation provides a robust support for this basal split. Nevertheless, our phylogenetic results suggest that three nsLTPs, MZm3-3, OsYY1 and ZmLTP3, are excluded from these groups and could form a third related class of nsLTPs. Beside these three proteins, six more non-Poaceae type 3 nsLTPs belong to this third class (not shown in Fig. 2 that only includes Poaceae nsLTPs) [41]. After the cleavage of their signal peptide, the deduced mature proteins present a molecular mass between 6.4 and 7.3 kDa. All the investigated type 3 nsLTPs are stamenspecific proteins but to date, with no biochemical studies

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

available, it is difficult to link the function of this type 3 nsLTPs with its structure or ligand fixation capacity. Recent divergence between wheat and barley in grass genome evolution [45] suggests the existence of close orthologous genes. This is particularly interesting since in bootstrap analysis, barley nsLTPs failed to cluster with wheat nsLTPs, suggesting that all orthologous wheat nsLTPs have not yet been described. TaLTP2 and TaLTP1 are encoded by distinct genes in the T. aestivum genome [46] and clustered with the barley HvLTP2(9) protein. However, TaLtp1 and TaLtp2 mRNAs have been isolated at the leaf growing stage [46] and HvLtp2(9) mRNA has been characterized from a cDNA library derived from young etiolated leaves [47]. Thus, if the expression of these genes is leaf-specific, we would not expect to identify the orthologs from wheat seed cDNA libraries. Likewise, we did not isolate barley HvLtp-br1 orthologous mRNA whose expression of the related gene has been found inducible in root cells subjected to nutrient deprivation (related note of EMBL accession number HVU88090). The description of the wheat nsLtp genes is not exhaustive since the allohexaploid T. aestivum carries three ancestral diploid genomes, A, B and D. Since these genomes are assumed to have derived from a common ancestral genome [48], each nsLtp gene is supposedly related to two other homoeologous nsLtp genes. In this way, homoeologous genes encoding granule-bound starch synthase [49] and KNOTTED1 homeobox proteins [50] have been described in the wheat hexaploid genome. Consequently, high amino acid homologies between clustered nsLTPs suggest that the corresponding genes share sets of homoeoalleles. Nevertheless, the impact of allopolyploidy on the evolution of the wheat genome results in the elimination of DNA sequences and changes in gene expression (for review, see [51]). As a consequence of local rearrangements, three homoeoalleles for each nsLtp gene would not be expected and because of differential rates of sequence evolution, homoeoalleles could be unclustered in our cladogram. Taken together, these results emphasize the complexity of the nsLTP family in wheat seed that can be linked with the genomic structure of T. aestivum (allohexaploid) and to a lesser extent T. durum (allotetraploid). Both the chromosome assignment and the physical mapping of these nsLtp genes will allow to study the evolution pattern of the different homoeologous nsLtp gene copies in wheat. We are currently investigating this nsLtp gene diversity using a PCR-based strategy. The expression patterns of nine nsLtp genes were analyzed by RT-PCR with RNA template from wheat seeds and seedlings. During seed development, RT-PCR analysis showed marked variations both in the pattern of gene expression and in the transcript levels from gene to gene but not from T. durum to T. aestivum (data not shown). However, it should be kept in mind that although the specificity of primers used for the RT-PCR analysis was verified by crossing PCR on corresponding plasmids, RTPCR products may correspond to transcripts from homoe-

123

ologous genes or to other genes whose corresponding cDNAs have not yet been isolated. Three phases can be roughly distinguished during seed maturation, the cell division phase (anthesis to 13 DPA), the cell expansion phase (13 – 15 to 27 –30 DPA) and the desiccation phase (27 –30 DPA to dry seed). Based on these stages, we identified three different profiles of nsLtp gene transcript accumulation. The first profile corresponds to nsLtp genes whose transcripts are only detected in the first stages of seed development. This profile comprises the closely related TdLtp9.3a and TdLtp9.5a genes and TdLtp7.2a gene whose transcripts are present only during the cell division phase, and TdLtp9.2a and TdLtp9.6a genes whose transcripts are still present at the beginning of the cell expansion phase when storage protein synthesis takes place. The TdLtp9.1a gene whose transcription starts during the cell expansion phase and maintains an abundant level of transcripts until the seed is dry is the only one to display the second profile. The TdLtp7.1a, TdLtp9.4a and TdLtp9.7a genes whose transcripts are present from the cell division phase to the dry seed display the third profile. Complex spatial and developmental expression patterns of nsLtp genes have been described during seed development. Previous works reported that nsLtp mRNA levels decline in nucellar epidermis a few days after pollination [18], whereas transcripts accumulate in the developing epidermal layers of embryos [52] or in the newly formed aleurone cells [53]. Thus, the different gene expression patterns in wheat seeds may be correlated with spatial specificity and correspond to distinct functions for each nsLTP. During the desiccation phase, the most abundant transcripts for type 1 nsLtp genes are those of TdLtp9.1a and TdLtp9.4a genes and those of TdLtp7.1a genes for type 2 nsLtp genes. It is interesting to note that the nsLTPs isolated from wheat flour [39,40] or semolina [21,38] are those encoded by Ltp9.1 and Ltp7.1 genes showing that these proteins are certainly the most abundant in the mature seed. The results of RT-PCR are in a good agreement with the origin of cDNA clones. Indeed, the clones isolated from mid-maturation cDNA libraries correspond to genes that display a high level of transcripts at this developmental stage. Likewise, within the five clones isolated from the 4 DPA cDNA library, four of them (TdLtp7.2a, TdLtp9.3a, TdLtp9.5a and TdLtp9.6a) are specifically or highly expressed early in seed development. Differential expression of nsLtp genes was also highlighted in the 3-day-old seedlings. Early in seed germination, wheat nsLtp genes display different expression patterns. De novo transcription occurs for TdLtp9.2a and TdLtp9.3a genes suggesting that nsLTPs encoded by these two genes may be involved in seed germination. Germination-specific nsLtp genes have been identified in Brassica napus [54] and Euphorbia lagascae [55] seeds indicating that nsLTPs are involved in seed germination. However, their role in the complex process of germination remains to

124

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125

be elucidated. It has been hypothesized than nsLTPs may be involved in programmed cell death in endosperm degradation during germination. NsLTPs may function either as an apoplastic carrier transferring lipids from the dead endosperm cells to growing cotyledon tissues or as a protease inhibitor forming a protective barrier that prevent the cotyledon cells from necrosis [12,15]. In contrast to the results observed in germinated seeds, all the nsLtp gene transcripts studied are found in the developing coleoptile. NsLtp transcripts are usually present in the vegetative aerial organs (for review see [56]), thus the wheat nsLtp transcripts were expected to be found in the young coleoptiles. Finally, only the TdLtp7.2a and TdLtp9.1a gene transcripts were found in the wheat seedling roots. Only a limited number of plant nsLtp genes studied to date exhibit expression in roots. A root-specific nsLtp gene was identified in bean seedling roots [57] whose transcripts accumulate in the cortical tissue [58]. Accumulation of nsLtp transcripts has been evidenced in grapevine roots during nodule development in the rhizobium-nodule symbiosis [59] and in rice roots in response to colonization by a mycorrhizal fungus [60]. Taken together, our RT-PCR results underline the fact that these nine wheat nsLtp genes whose corresponding cDNAs were isolated from seed libraries are not seed-specific but are also expressed in vegetative tissues. Our data support differences in nsLTP primary structure and gene expression patterns raising the question as to why so many nsLtp genes are expressed during seed development and whether the different nsLTPs have distinctive functions depending on the developmental stage or cellular localization. To start to answer these questions, the promoters of seven wheat TaLtp genes have been isolated and their expression pattern is currently being analyzed in transgenic rice using a reporter gene. These data will give us a more precise pattern of expression than the RT-PCR that was performed on whole wheat seed.

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

Acknowledgements FB was the recipient of a fellowship from the French Ministe`re de l’Education Nationale, de l’Enseignement Supe´rieur et de la Recherche.

[17]

[18]

References [1] J.-C. Kader, M. Julienne, C. Vergnolle, Purification and characterization of a spinach-leaf protein capable of transferring phospholipids from liposomes to mitochondria or chloroplasts, Eur. J. Biochem. 139 (1984) 411 – 416. [2] J.-C. Kader, Lipid-transfer proteins: a puzzling family of plant proteins, Trends Plant Sci. 2 (1997) 66 – 70. [3] F. Tchang, P. This, V. Stiefel, V. Arondel, M.-D. Morch, M. Pages, P. Puigdome`nech, F. Grellet, M. Delseny, P. Bouillon, J.-C. Huet, F. Guerbette, F. Beauvais-Cante, H. Duranton, J.-C.

[19]

[20]

[21]

Pernollet, J.-C. Kader, Phospholipid transfer protein: full length cDNA and amino acid sequence in maize, J. Biol. Chem. 263 (1988) 16849 – 16855. P. Sterk, H. Booij, G.A. Schellekens, A. Van Kammen, S.C. De Vries, Cell-specific expression of the carrot EP2 lipid transfer protein gene, Plant Cell 3 (1991) 907 – 921. B.P.A. Cammue, K. Thevissen, M. Hendriks, K. Eggermont, I.J. Goderis, P. Proost, J. Van Damme, R.W. Osborn, F. Guerbette, J.-C. Kader, W.F. Broekaert, A potent antimicrobial protein from onion seeds showing sequence homology to plant lipid transfer proteins, Plant Physiol. 109 (1995) 445 – 455. A. Molina, A. Segura, F. Garcı´a-Olmedo, Lipid transfer proteins (nsLTPs) from barley and maize leaves are potent inhibitors of bacterial and fungal plant pathogens, FEBS Lett. 316 (1993) 119 –122. C.-J. Park, R. Shin, J.M. Park, G.-J. Lee, J.-S. You, K.-H. Paek, Induction of pepper cDNA encoding a lipid transfer protein during the resistance response to tobacco mosaic virus, Plant Mol. Biol. 48 (2002) 243 – 254. M.C. Regente, A.M. Giudici, J. Villalain, L. de la Canal, The cytotoxic properties of a plant lipid transfer protein involve membrane permeabilization of target cells, Lett. Appl. Microbiol. 40 (2005) 183 – 189. N. Buhot, J.-P. Douliez, A. Jacquemard, D. Marion, V. Tran, B. Maume, M.-L. Milat, M. Ponchet, V. Mikes, J.-C. Kader, J.-P. Blein, A lipid transfer protein binds to a receptor involved in the control of plant defence responses, FEBS Lett. 509 (2001) 27 – 30. J.-P. Blein, P. Coutos-The´venot, D. Marion, M. Ponchet, From elicitins to lipid-transfer proteins: a new insight in cell signalling involved in plant defence mechanisms, Trends Plant Sci. 7 (2002) 293 – 296. A.M. Maldonado, P. Doerner, R.A. Dixon, C.J. Lamb, R.K. Cameron, A putative lipid transfer protein involved in systemic resistance signalling in Arabidopsis, Nature 419 (2002) 399 – 403. S. Endo, T. Demura, H. Fukuda, Inhibition of proteasome activity by the TED4 protein in extracellular space: a novel mechanism for protection of living cells from injury caused by dying cells, Plant Cell Physiol. 42 (2001) 9 – 19. V. Arondel, C. Vergnolle, C. Cantrel, J.-C. Kader, Lipid transfer proteins are encoded by a small multigene family in Arabidopsis thaliana, Plant Sci. 157 (2000) 1 – 12. J.-X. Feng, S.-J. Ji, Y.-H. Shi, G. Wei, Y.-X. Zhu, Analysis of five differentially expressed gene families in fast elongating cotton fiber, Acta Biochim. Biophys. Sin. 36 (2004) 51 – 57. D.M. Eklund, J. Edqvist, Localization of nonspecific lipid transfer proteins correlate with programmed cell death responses during endosperm degradation in Euphorbia lagascae seedlings, Plant Physiol. 132 (2003) 1249 – 1259. Z. Wu, J.K. Burns, Isolation and characterization of a cDNA encoding a lipid transfer protein expressed in FValencia_ orange during abscission, J. Exp. Bot. 54 (2003) 1183 – 1191. H.W. Jung, W. Kim, B.K. Hwang, Three pathogen-inducible genes encoding lipid transfer protein from pepper are differentially activated by pathogens, abiotic, and environmental stresses, Plant Cell Environ. 26 (2003) 915 – 928. F. Chen, M.R. Foolad, Nucellar-cell-specific expression of a lipid transfer protein gene in barley (Hordeum vulgare L.), Plant Cell Rep. 18 (1999) 445 – 450. F. Vignols, M. Wigger, J.M. Garcı´a-Garrido, F. Grellet, J.-C. Kader, M. Delseny, Rice lipid transfer protein (LTP) genes belong to a complex multigene family and are differentially regulated, Gene 195 (1997) 177 – 186. F. Pele`se-Siebenbourg, C. Caelles, J.-C. Kader, M. Delseny, P. Puigdome`nech, A pair of genes coding for lipid-transfer proteins in Sorghum vulgare, Gene 148 (1994) 305 – 308. F.-P. Monnet, W. Dieryck, F. Boutrot, P. Joudrier, M.-F. Gautier, Purification, characterisation and cDNA cloning of a type 2 (7 kDa) lipid transfer protein from Triticum durum, Plant Sci. 161 (2001) 747 – 755.

F. Boutrot et al. / Biochimica et Biophysica Acta 1730 (2005) 114 – 125 [22] C.S. Jang, J.Y. Kim, J.W. Haam, M.S. Lee, D.S. Kim, Y.W. Li, Y.W. Seo, Expressed sequence tags from a wheat-rye translocation line (2BS/2RL) infested by larvae of Hessian fly [Mayetiola destructor (Say)], Plant Cell Rep. 22 (2003) 150 – 158. [23] C.S. Jang, H.J. Lee, S.J. Chang, Y.W. Seo, Expression and promoter analysis of the TaLTP1 gene induced by drought and salt stress in wheat (Triticum aestivum L.), Plant Sci. 167 (2004) 995 – 1001. [24] D.A. Gaudet, A. Laroche, M. Frick, R. Huel, B. Puchalski, Cold induced expression of plant defensin and lipid transfer protein transcripts in winter wheat, Physiol. Plant 117 (2003) 195 – 205. [25] E.-M. Yubero-Serrano, E. Moyano, N. Medina-Escobar, J. Mun˜ozBlanco, J.-L. Caballero, Identification of a strawberry gene encoding a non-specific lipid transfer protein that responds to ABA, wounding and cold stress, J. Exp. Bot. 54 (2003) 1865 – 1877. [26] J. Bubier, M. Schlappi, Cold induction of EARLI1, a putative Arabidopsis lipid transfer protein, is light and calcium dependent, Plant Cell Environ. 27 (2004) 929 – 936. [27] B. Hollenbach, L. Schreiber, W. Hartung, K.-J. Dietz, Cadmium leads to stimulated expression of the lipid transfer protein genes in barley: implications for the involvement of lipid transfer proteins in wax assembly, Planta 203 (1997) 9 – 19. [28] M.-F. Gautier, R. Alary, P. Joudrier, Cloning and characterization of a cDNA encoding the wheat (Triticum durum Desf.) CM16 protein, Plant Mol. Biol. 14 (1990) 313 – 322. [29] M.-F. Gautier, M.E. Aleman, A. Guirao, D. Marion, P. Joudrier, Triticum aestivum puroindolines, two basic cystine-rich seed proteins: cDNA sequence analysis and developmental gene expression, Plant Mol. Biol. 25 (1994) 43 – 57. [30] W. Dieryck, M.-F. Gautier, V. Lullien, P. Joudrier, Nucleotide sequence of a cDNA encoding a lipid transfer protein from wheat (Triticum durum Desf.), Plant Mol. Biol. 19 (1992) 707 – 709. [31] H. Nielsen, J. Engelbrecht, S. Brunak, G. von Heijne, Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites, Protein Eng. 10 (1997) 1 – 6. [32] O. Emanuelsson, H. Nielsen, S. Brunak, G. von Heijne, Predicting subcellular localization of proteins based on their N-terminal amino acid sequence, J. Mol. Biol. 300 (2000) 1005 – 1016. [33] E. Myers, W. Miller, Optimal alignments in linear space, Comput. Appl. Biosci. 4 (1988) 11 – 17. [34] J.D. Thompson, D.G. Higgins, T.J. Gibson, CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice, Nucleic Acids Res. 22 (1994) 4673 – 4680. [35] J. Felsenstein, PHYLIP—Phylogeny inference package (version 3.2), Cladistics 5 (1989) 164 – 166. [36] N. Saitou, M. Nei, The neighbor-joining method: a new method for reconstructing phylogenetic trees, Mol. Biol. Evol. 4 (1987) 406 – 425. [37] R.D.M. Page, TREEVIEW: an application to display phylogenetic trees on personal computers, Comput. Appl. Biosci. 12 (1996) 357 – 358. [38] F.-P. Monnet, Caracte´risation d’une prote´ine de fixation de lipides du ble´ dur, purification, se´quenc¸age, ADN comple´mentaire: relations aux prote´ines ve´ge´tales de transfert de lipides et aux inhibiteurs d’amylase/ trypsine des ce´re´ales, PhD thesis, University of Montpellier II, France, 1990. [39] A. De´sormeaux, J.-E. Blochet, M. Pe´zolet, D. Marion, Amino acid sequence of a non-specific wheat phospholipid transfer protein and its conformation as revealed by infrared and Raman spectroscopy. Role of disulfide bridges and phospholipids in the stabilization of the alphahelix structure, Biochim. Biophys. Acta, Prot. Struct. Mol. Enzymol. 1121 (1992) 137 – 152. [40] J.-P. Douliez, C. Pato, H. Rabesona, D. Molle, D. Marion, Disulfide bond assignment, lipid transfer activity and secondary structure of a 7kDa plant lipid transfer protein, LTP2, Eur. J. Biochem. 268 (2001) 1400 – 1403. [41] B. Lauga, L. Charbonnel-Campaa, D. Combes, Characterization of MZm3-3, a Zea mays tapetum-specific transcript, Plant Sci. 157 (2000) 65 – 75.

125

[42] S. Hebsgaard, P. Korning, N. Tolstrup, J. Engelbrecht, P. Rouze, S. Brunak, Splice site prediction in Arabidopsis thaliana pre-mRNA by combining local and global sequence information, Nucleic Acids Res. 24 (1996) 3439 – 3452. [43] G.G. Simpson, W. Filipowicz, Splicing of precursors to mRNA in higher plants: mechanism, regulation and sub-nuclear organisation of the spliceosomal machinery, Plant Mol. Biol. 32 (1996) 1 – 41. [44] A.J. White, M.A. Dunn, K. Brown, M.A. Hughes, Comparative analysis of genomic sequence and expression of a lipid transfer protein gene family in winter barley, J. Exp. Bot. 45 (1994) 1885 – 1892. [45] K.M. Devos, M.D. Gale, Genome relationships: the grass model in current research, Plant Cell 12 (2000) 637 – 646. [46] C.S. Jang, D.S. Kim, S.Y. Bu, J.B. Kim, S.S. Lee, J.Y. Kim, J.W. Johnson, Y.W. Seo, Isolation and characterization of lipid transfer protein (LTP) genes from a wheat-rye translocation line, Plant Cell Rep. 20 (2002) 961 – 966. [47] A. Molina, F. Garcı´a-Olmedo, Developmental and pathogen-induced expression of three barley genes encoding lipid transfer proteins, Plant J. 4 (1993) 983 – 991. [48] E.R. Sears, in: R. Riley, K.R. Lewis (Eds.), Chromosome Manipulation and Plant Genetics, Oliver and Boyd, Edinburgh, 1966, pp. 29 – 45. [49] J. Murai, T. Taira, D. Ohta, Isolation and characterization of the three Waxy genes encoding the granule-bound starch synthase in hexaploid wheat, Gene 234 (1999) 71 – 79. [50] S. Takumi, T. Kosugi, K. Murai, N. Mori, C. Nakamura, Molecular cloning of three homoeologous cDNAs encoding orthologs of the maize KNOTTED1 homeobox protein from young spikes of hexaploid wheat, Gene 249 (2000) 171 – 181. [51] A.A. Levy, M. Feldman, The impact of polyploidy on grass genome evolution, Plant Physiol. 130 (2002) 1587 – 1593. [52] C.W. Vroemen, S. Langeveld, U. Mayer, G. Ripper, G. Jurgens, A. van Kammen, S.C. de Vries, Pattern formation in the Arabidopsis embryo revealed by position-specific lipid transfer protein gene expression, Plant Cell 8 (1996) 783 – 791. [53] K. Skriver, R. Leah, F. Muller-Uri, F.L. Olsen, J. Mundy, Structure and expression of the barley lipid transfer protein gene Ltp1, Plant Mol. Biol. 18 (1992) 585 – 589. [54] I.A. Soufleri, C. Vergnolle, E. Miginiac, J.C. Kader, Germinationspecific lipid transfer protein cDNAs in Brassica napus L., Planta 199 (1996) 229 – 237. [55] J. Edqvist, I. Farbos, Characterization of germination-specific lipid transfer proteins from Euphorbia lagascae, Planta 215 (2002) 41 – 50. [56] J.-C. Kader, Lipid-transfer proteins in plants, Annu. Rev. Plant Physiol. Plant Mol. Biol. 47 (1996) 627 – 654. [57] D.-W. Choi, J.Y. Song, M.-H. Oh, J.S. Lee, J. Moon, S.W. Suh, S.-G. Kim, Isolation of a root-specific cDNA encoding a ns-LTP-like protein from the roots of bean (Phaseolus vulgaris L.) seedlings, Plant Mol. Biol. 30 (1996) 1059 – 1066. [58] J.Y. Song, D.-W. Choi, J.S. Lee, Y.M. Kwon, S.-G. Kim, Cortical tissue-specific accumulation of the root-specific ns-LTP transcripts in the bean (Phaseolus vulgaris) seedlings, Plant Mol. Biol. 38 (1998) 735 – 742. [59] A. Krause, C.J. Sigrist, I. Dehning, H. Sommer, W.J. Broughton, Accumulation of transcripts encoding a lipid transfer-like protein during deformation of nodulation-competent Vigna unguiculata root hairs, Mol. Plant-Microb. Interact. 7 (1994) 411 – 418. [60] I. Blilou, J.A. Ocampo, J.M. Garcı´a-Garrido, Induction of Ltp (lipid transfer protein) and Pal (phenylalanine ammonia-lyase) gene expression in rice roots colonized by the arbuscular mycorrhizal fungus Glomus mosseae, J. Exp. Bot. 51 (2000) 1969 – 1977. [61] G.M. Neumann, R. Condron, I. Thomas, G.M. Polya, Purification and sequencing of a family of wheat lipid transfer protein homologues phosphorylated by plant calcium-dependent protein kinase, Biochim. Biophys. Acta, Prot. Struct. Mol. Enzymol. 1209 (1994) 183 – 190.