cDNA-AFLP analysis

For each of the 43 primer combinations, 40-100 different transcript derived fragments (TDFs), which ranged from 50 to 800 bp, were visualized as bands (Figure 2).

Gene ontology analysis of Mexican lime tree transcripts modulated by witches broom infection

Each of the 51 sequenced transcript was annotated functionally through careful analysis of the scientific literature and the Gene Ontology Databases. Out of 51 sequenced DE-TDFs, 36 (80%) could be assigned to one of the following functional groups: stress response/defense (10 TDFs), cell Metabolism (4 TDFs), protein synthesis/destination (4 TDFs), signal transduction (3 TDFs), transporter (2 TDFs), transcription (1 TDFs), cytoskeleton (1 TDFs) and unknown function (11 TDFs). The molecular function of each individual protein is given in Table 1. The stress response/defence group contained 27.7% of the DE-TDFs and constituted the largest functional group (Figure 3).

Verification of representative genes by real-time RT-PCR

To verify the expression patterns that were identified in the cDNA-AFLP study, the expression level of four DE-TDFs was analyzed by real-time RT-PCR. These TDFs were chosen to represent the various functional categories identified (Table 1). Two of the selected TDFs (serine/threonine-protein kinase and importin β) were more abundant in infected plants, whereas two TDFs (autophagy protein 5 and RNA polymerase β) showed higher expression in healthy plants. The 18 s RNA gene of Mexican lime tree was used as a reference gene for data normalization, as described previously [12]. Real-time PCR analysis showed that the expression of the selected genes agreed well with the profiles determined by cDNA-AFLP (Figure 4).

Infection with "Ca. Phytoplasma aurantifolia" causes widespread gene repression in Mexican lime trees

Sixty-seven percent of the identified DE-TDFs were down-regulated in response to infection, whereas only 33% were up-regulated in response to infection which could reflect the exploitation of cellular resources and the suppression of defence responses by the phytoplasma [13].

Responses to external stimuli and defence

Several genes that were modulated in Mexican lime trees by infection with "Ca. Phytoplasma aurantifolia" were related to defence, cell walls, and response to stress. The expression of autophagy protein 5 was repressed. Autophagy is a survival mechanism that protects cells against unfavourable environmental conditions, such as microbial pathogen infection, oxidative stress, nutrient starvation, and aggregation of damaged proteins [14]. It has been shown that carbohydrate starvation induces the expression of autophagy genes [15] and stimulates the formation of reactive oxidative species (ROS) in plants [14]. Biochemical analyses on phytoplasma-infected tobacco plants have indicated that soluble carbohydrates and starch were accumulated in source leaves, whereas sink organs showed a marked decrease in sugar levels. Similar changes in carbohydrate metabolism have been described in coconut palms infected with the lethal yellowing phytoplasma [16]. It is likely that the accumulation of carbohydrate reduces the expression of autophagy genes in the host and limits the burst of ROS burst (hypersensitivity reaction). These effects might result in reduced host resistance to phytoplasma and create a suitable conditions for phytoplasma survival in the host.

We also identified a cell wall hydroxyl proline-rich protein (GT222039) that was induced in response to the pathogen. Proline-rich proteins are among the major structural proteins of plant cell walls. Environmental stresses can alter the composition of the plant cell wall markedly [17]. It has been demonstrated that mechanical wounding, infection, or elicitors obtained from microbial cell walls or culture fluids caused accumulation of specific hydroxyl proline-rich glycoproteins and other antimicrobial cell wall proteins [17]. It has been reported that elicitors cause an H₂O₂-mediated oxidative cross-linking of preexisting structural cell wall proteins that precedes the activation of transcription-dependent defences. The induction of the hydroxyl proline-rich protein in the present study might reflect a defence mechanism of Mexican lime tree in response to phytoplasma infection.

Another induced protein (GT222056) contained a lysine domain that is found in several enzymes that are involved in degradation of the bacterial cell wall [18]. The role of this gene in the response of Mexican lime trees to the pathogens remains to be determined.

Two of repressed genes (GT222036 and GT222036) were identified as a modifier of snc1 (MOS1). Plant resistance (R) genes encode immune receptors that recognise pathogens directly or indirectly and activate defence responses [19]. The expression levels of R genes have to be regulated tightly due to costs to the fitness of plants that are associated with maintaining R-protein-mediated resistance. Recently, it has been reported that MOS1 regulates the expression of SNC1 which encodes a TIR-NB-LRR-type of R protein in Arabidopsis.

It has been shown that mos1 mutations reduce the expression of endogenous snc1, which results in the repression of constitutive resistance responses that are mediated by snc1 [20]. It is likely that down-regulation of Mexican lime tree MOS1 in response to the pathogen reflects a reduction in plant resistance responses to phytoplasma infection.

Cell Metabolisms

Lipid-derived molecules act as signals in plantpathogen interactions, and the roles of jasmonic acid and related oxylipins that are produced from membrane-derived fatty acids through beta-oxidation, are particularly important [21]. During infection, low level defence responses can be activated in susceptible plants [22, 23]. Therefore, it is likely that well-established "Ca. Phytoplasma aurantifolia" infections involve the up-regulation of genes that encode components of the lipid metabolism pathway, such as phosphatidyl glycerol specific phospholipase C-like (GT222018). This enzyme regulates the phosphatidylglycerol content via a phospholipase C-type degradation mechanism [24]. Another gene involved in lipid metabolisms, glycerophosphoryl diester phosphodiesterase (GT222042) was repressed during the infection. This enzyme has both phosphoric diester hydrolase and glycerophosphodiester phosphodiesterase activity and is involved in the metabolism of glycerol and lipids [25].

Protein synthesis and destination

We identified several TDFs that were related to protein metabolism in our study. Among these were genes that encoded ribosomal proteins and enzymes involved in degradation. The expression of two ubiquitin-protein ligases (GT222065 and GT222065) and one 50 S ribosomal protein L15 (GT222023) were repressed, whereas another 50 S ribosomal protein L15 (GT222024) was induced. This suggests that the infection results in a general induction of protein turnover, which could reflect an adaptive response in the plants to remove misfolded proteins that have accumulated as a result of stress [23].

Signal transduction

Three of the modulated genes had signal transduction and/or gene regulation functions. They corresponded two transducin family protein (GT222030 and GT222029) that were repressed by infection and a serine/threonine protein kinases (GT222061) that was induced during infection. Serine/threonine protein kinases are a group of enzymes that catalyse the phosphorylation of serine or threonine residues in proteins, with ATP or other nucleotides acting as phosphate donors. The phosphorylation of proteins on serine, threonine, or tyrosine residues is an important biochemical mechanism to regulate the activity of enzymes and is used in many cellular processes [26].

The two down-regulated proteins were identified as members of the transducin family and contained WD40 domain. This domain is found in several eukaryotic proteins that with wide variety of functions, which include adaptor/regulatory modules in signal transduction, together with proteins involved in pre-mRNA processing, and cytoskeleton assembly [27]. It is unclear how these changes contribute to the response of Mexican lime tree to infection.

Plant material and inoculation

Ten healthy 1-year-old Mexican lime trees grown in the greenhouse were used in this experiment. Specimens from Mexican lime trees infected with witches' broom were grafted to healthy trees, and specimens from healthy Mexican lime trees were grafted to other healthy trees. The grafted plants were covered for 1 month with plastic bags to increase humidity and were arranged randomly on the greenhouse bench. They were kept under natural light conditions at a temperature of 25-28°C. The branches infected with witches' broom were sampled 20 weeks after inoculation and used for RNA extraction. As a control, RNA was extracted from non-grafted healthy plant leaves that has been grown under similar conditions.

Detection of Phytoplasma infection by nested PCR

Total DNA was extracted from leaf samples (vascular tissues from leaf veins and petioles) using the method described originally by Daire et al[28] with some modifications [29]. Samples of tissue (1 g) were homogenised at room temperature in 7 ml of cetyl trimethyl ammonium bromide (CTAB) buffer (3% CTAB, 1 M Tris-HCl pH 8.2, mM EDTA, 1.4 M NaCl), with addition of 0.2% 2-mercaptoethanol, in disposable plastic bags using a ball-bearing device. Aliquots of 1 ml of homogenate were transferred to Eppendorf tubes and incubated in a water bath at 65°C for 20 min. After extraction with 1 ml of chloroform, nucleic acids were precipitated from the aqueous phase with an equal volume of isopropanol, collected by centrifugation, washed with 70% ethanol, dried, dissolved in 150 ml of TE buffer (10 mM Tris, 1 mM EDTA, pH 7.6) and stored at -20°C until use. The region of the phytoplasma 16 S rRNA gene was amplified by PCR in a total reaction volume of 25 μl in an Applied Biosystems thermal cycler. The first set of PCR primers was P1 (5'-AAGAGTTTGATCCTGGCTCAGGATT-3') [30] and P7 (5'-CGTCCTTCATCGGCTCTT -3') [31]. The resulting P1-P7 amplicons were then used as template DNA in a nested-PCR amplification with the universal primer pair for phytoplasmas r16r2/r16F2n [32]. The purified PCR products were cloned into the pGEM-T Easy vector (Promega), and sequenced at the fluorescent automated sequencing facility at Fazabiotech (Tehran, Iran). The phytoplasma strains were classified using iPhyClassifier, as described by Zhao et al[33].

RNA isolation and cDNA synthesis

Total RNA was extracted from 5 phytoplasma-infected Mexican lime trees and 5 healthy trees using Trizol reagent (Invitrogen, Life Technologies) in accordance with the manufacturer's instructions. cDNA synthesis and cDNA-AFLP analysis were performed for the 10 replicates. First-strand cDNA was synthesised from 2 μg of total RNA using a SuperScript III First Strand Synthesis System (Invitrogen, USA) in accordance with the manufacturer's instructions. Second-strand cDNA was sythesised by adding the first-strand cDNA reaction to a reaction mix that contained 15 μl of 10 × cDNAII buffer, 35 U DNA of Polymerase I (Invitrogen), 3 U of RNase H (Invitrogen), and 1 μl dNTPs (25 mM) in a final volume of 150 μl, and incubating for 2 h at 16°C (). The resulting double-stranded cDNA was purified in accordance with the method of Powell and Gannon [34]. The concentration of the cDNAs was determined using spectrophotometer (Bio-Rad) and their quality was determined by electrophoresis on a 1.2% agarose gel.

cDNA- AFLP

A 500-ng aliquot of double-stranded cDNA was used for AFLP analysis as described by Bachem et al. [35] with the following modifications. The template for cDNA-AFLP was digested with the restriction enzymes, EcoR I/Mse I and Psu I/Mse I (Invitrogen). The Sequence of the primers and adapters used for the AFLP reactions are given in Additional File 2. AFLP reactions were performed in accordance with Bachem et al. [36]. Selective amplification products were separated on a 10% polyacrylamide gel and stained with silver nitrate [37]. The gels were dried onto 3 MM Whatman paper.

Cloning, sequencing and bioinformatic characterisation

To select DE-TDFs, the profiles of infected and non-infected samples were compared between replicates. TDFs that differed in abundance between the two types of sample, namely infected and non-infected plants, were selected only when the same pattern was observed in all replicates.

The cloning of bands of interest was performed as previously described[38]. Briefly, the bands were excised from the gels using a razor blase. Each gel slice was incubated in 10 μl of distilled water for 10 min at 96°C. Aliquots of the eluent were subjected to PCR using the same conditions as for the selective PCR described before. PCR products were separated on 10% polyacrylamide gel to confirm that the correct polymorphic fragments had been selected [39].

After verification, the recovered products re-amplified using primer pair E-0/M-0 and P-0/M-0 to provide sufficient DNA for cloning. The purified PCR products were cloned into the pGEM-T Easy vector (Promega) and then sequenced.

The sequences were compared with those in the non-redundant databases of the National Center for Biotechnology Information (NCBI; http://www.ncbi.nlm.nih.gov/BLAST/) and The Arabidopsis Information Resource (TAIR; http://www.arabidopsis.org/Blast/) using the Blastn, Blastx, and tBlastx sequence alignment algorithms [40]

Real-time RT-PCR analysis

Real-time RT-PCR was carried out using RNA derived from infected and non-infected Mexican lime trees as described previously by Martini et al. [41]. Aliquots of each RNA sample (2 μl) were used to synthesise cDNA using a SuperScript III First Strand Synthesis System (Invitrogen). Primer pairs were designed using OligoTech Software (Additional File 3). Gene expression was assayed using the iCycler iQ, Multicolor Real-Time PCR Detection System (Bio-Rad) and iQ SYBR Green Supermix kit (Bio-Rad). Reaction conditions (20 μl volumes) were optimized by changing the primer concentration and annealing temperature to minimise primer-dimer formation and increase PCR efficiency. The following PCR profile was used: 2 min at 95°C, (95°C for 20 s, 60-63°C for 20 s, 72°C for 20 s) × 45, and 1 min at 72°C followed by recording of a melting curve. The absence of primer-dimers or accumulation of nonspecific products was checked by melting-curve analysis. Each run included standard dilutions and negative reaction controls. The expression levels of each gene of interest and of the 18 S rRNA, which was chosen as a housekeeping gene, were determined in parallel for each sample. Results were expressed as the ratio of the mRNA level of for each gene of interest normalised over housekeeping gene using the difference between threshold cycle values or ΔΔCt method [42, 43]. Ct values for individual target genes were calculated and the ΔCt average for the housekeeping gene was treated as an arbitrary constant and used to calculate ΔΔCt values for all samples. The mean fold induction for the three independent pools for each target gene was determined and the standard error of the mean was calculated. The list of primers used for the real-time PCR analysis is presented in Additional File 3.