Introduction

Spiroplasma citri the first spiroplasma to have been cultured since 1970 is the causal agent of citrus stubborn disease [34]. S. citri is able to infect many plant species other than citrus including horse radish and Madagascar periwinkle plants [10], sesame [37], carrots [26], and safflower [23] in which it induces brittle root, wilting, stunting, and necrotic yellows. It is a wall-less phloem-inhabiting bacterium belonging to the class Mollicutes. The bacterium is naturally transmitted by several species of leafhoppers such as Circulifer tenellus, in California and Circulifer haematoceps, in the Mediterranean area in a circulative–propagative manner [10, 27].

The genome of S. citri is known as one of the largest among Mollicutes, the wall-less bacteria with low G+C content. In addition to its 1,820 Kbp circular chromosome [11] the reference strain GII-3 of S. citri also harbors 7 plasmids, pSci A and pSci 1 to pSci 6 [36]. Genetic diversity in Spiroplasma variants may involve acquisition and loss of DNA elements, replication and repair in DNA, homologous recombination and transposition with no relation between genetic diversity of the strains and isolation time or geographical regions [29]. Indeed, in addition to being vertically inheritable, pSci 1–6 plasmids can transfer between cells by conjugation [8]. S. citri has also acquired in its chromosome a great number of viral sequences through evolution [11]. S. citri strains may be infected by filamentous, DNA phages related to genus Plectrovirus [33]. Southern blot hybridizations carried out with ribosomal and spiroplasma viral probes, PCR-Restriction Fragment Length Polymorphism (RFLP) [23] electrophoretic mobility of spiralin gene [17] and Random Amplification of Polymorphic DNA (RAPD) by PCR [29] are used as possible tools for efficiently differentiating S. citri strains.

Spiralin lipoprotein is one of the most thoroughly characterized S. citri membrane proteins [13–17, 42]. This amphiphilic lipoprotein is specific to the genus Spiroplasma. Spiralin of S. citri GII-3 has 83.5, 85.1, and 88.9 % identity with spiralins of S. kunkelii, S. phoeniceum, and Spiroplasma melliferum, respectively. The protein possesses a typical signal peptide of eubacterial lipoproteins [19] upstream of a cysteine residue and a highly conserved central region surrounded by two short repeated sequences [18]. A model in which spiralin could form a protein carpet covering the entire spiroplasma surface was proposed [12] and its requirement for efficient transmission of S. citri by its leafhopper vector was evidenced [16]. Spiralin could play a key role in the transmission of S. citri by mediating spiroplasma adherence to epithelial cells of the gut or the salivary glands of the insect vector. Recently, the GII-3 spiralin was shown to act in vitro as a lectin binding to two C. haematoceps glycoproteins of 50 and 60 kDa and therefore it might function as a ligand able to interact with uncharacterized insect surface protein receptors [24]. We report in this paper the results concerning the spiralin diversity in different S. citri isolates in the Iranian Fars province.

Materials and Methods

Source of Strains, Culture, and ELISA Detection of S. citri

Suspected S. citri infected citrus and non-citrus plants were collected from several regions in the Fars province. Spiroplasmas from citrus trees in different localities: leaves (midribs) and fruits (columella and aborted seeds) were isolated and cultivated in LD10 medium [25] (Table 1). S. citri cultures from fruit columella were preferentially used for further molecular analysis. Leafhoppers (C. haematoceps) collected on different plants were caged individually on young Madagascar periwinkle seedlings (Catharanthus roseus) for 2 weeks and midribs from yellow plants were used as the source of S. citri for culture assays (Table 1). Midribs of sesame and safflower plants with yellowing symptoms were also used for culture of S. citri (Table 1). For all different samples, upon a change in the color, LD10 medium was checked by electron microscopy for the presence of spiral particles. Spiroplasmas in culture were triply cloned and stored at −20 °C. S. citri was detected in plants and leafhoppers by ELISA according to the method used for S. citri [35]. S. citri antiserum was from a rabbit that had been immunized with S. citri isolate Fasa III.

Table 1 Origin of the Iranian S. citri cultures used in this study
Full size table

DNA Extraction

DNA was extracted from spiroplasmal cells cultured in LD10 medium. 10 mL of the culture medium were centrifuged at 13,400g and DNA was extracted from the pellet using CTAB procedure [14]. For each sample, concentration of DNA was adjusted to 60 ng/μL. DNA samples were stored at −20 °C until use. The isolate Shiraz I hardly grew during the first passage in LD10 medium and attempts for sub-culturing the organism with the same medium were unsuccessful. Therefore, for this isolate, the bacterial cells following the first passage in LD10 were centrifuged. The harvested pellet was re-suspended in sterile water and incubated at 80 ºC for 15 min before being used as a source of DNA for PCR experiments.

Polymerase Chain Reaction (PCR) and Cloning

Primer pairs F1/R1 and NesF/NesR designed using the Primer3 software were used for amplification of spiralin gene [22] and part of 16S rDNA gene, respectively (Table 2). PCR was carried out in 25 μL reaction mixtures containing 17.75 μL deionized water, 2 μL (12 ng) DNA template, 0.4 μL of 10 mM dNTPs, 2 μL of 10× PCR reaction buffer, 1 μL of each 10 μM primer, 0.1 μL of 5 U/μL Taq DNA polymerase (Roche), and 0.75 μL of 50 mM MgCl2. PCR amplification was performed as follows: initial denaturing at 94 °C for 4 min, 35 cycles of 94 °C for 30 s, annealing temperature (44 °C for F1/R1 or 54 °C for NesF/NesR) for 40 s and 72 °C for 1 min. Final extension at 72 °C was for 10 min. The PCR products were ligated into pTZ57R/T (Fermentas) and cloned in E. coli DH5α according to the manufacturer’s instruction. Recombinant plasmids were extracted using High Pure Plasmids Extraction Kit (Fermentas). From each sample, two different clones were sequenced.

Table 2 Synthetic oligonucleotides used in this study
Full size table

RFLP Analysis

Spiralin genes from different S. citri isolates were amplified by PCR and the amplified products were submitted to restriction by DraI, RsaI, MboI, according to previous protocol [23].

Sequence Analysis

The nucleotide sequences of PCR products were subjected to BLAST in NCBI. MEGA 5 software [39] was used for multiple sequence alignment and phylogenetic tree construction. The sequences of all isolates were putatively translated in the software DNASTAR Lasergene-Editseq-(version 5.00) and amino acid sequences were aligned by MEGA 5 software [39]. Secondary structure of the protein was predicted using Psipred program [9] and the repeated motifs were identified combining RADAR [20] and MEME tools [5].

Results and Discussion

Detection and Isolation

S. citri was detected in citrus trees, and in non-citrus trees species by ELISA (Table 3). Most citrus samples collected on sweet orange and mandarin trees with stubborn symptoms (67 %) were positive for S. citri. In a high percentage of samples from symptomless sweet lime and sour orange trees (72 %) S. citri was detected by ELISA. Likewise, a high proportion of bindweed plants (69 %) with little leaf, chlorosis and stunting showed positive reaction in ELISA. Previously, Nejat et al. [32] have reported infection of bindweed to S. citri in Iran using polyclonal antiserum against S. citri. However, bindweed samples failed to support growth of S. citri in culture medium despite having high ELISA values. Due to the possible infection of bindweed with stolbur phytoplasma [28] twelve bindweed DNA samples were checked for the presence of phytoplasmas using general primer pair P1/P7 [38]. No phytoplasma DNAs were detected in the tested plants. Safflower plants (Carthamus tinctorius) with yellowing and phloem discoloration and periwinkle plants (C. roseus) with yellowing symptoms were over 85 % positive for S. citri (Table 3). A number of plants (45 %) including wild lettuce (Lactuca virosa), yellow sweet clover (Melilotus officinalis), sweet lime (Citrus limettioides), and Russian thistle (Salsola sp.) were positive with ELISA without showing any specific symptoms (Table 3). Asymptomatic host plants may exist [10] and for brassicaceous species, it was suggested that both wild and cultivated plants appear to be important in the establishment of insect populations (feeding and breeding). The Russian thistle was considered as the preferred host of the beet leafhoppers and might be asymptomatic. Attempts to culture S. citri from it always failed [2]. In citrus free growing areas with a semi arid climate like in the Fars region, S. citri was isolated from leafhopper vectors where their preferred host plant S. kali was commonly found. Leafhoppers collected in Darab and Fasa localities are able to transmit S. citri to periwinkles. Four S. citri isolates were obtained in culture from yellowing periwinkles infected by leafhoppers (Table 1). S. citri was isolated in culture medium from five citrus samples collected from different localities in the Fars province (Table 1). From midribs of sesame and safflower affected by yellowing, 2 isolates named Darab XVIII and Saf I respectively were obtained in LD10 medium (Table 1). Infection of sesame (Sesamum indicum) as a non-citrus host of S. citri was first reported in Turkey where sesame and its associated vector, Circulifer opacipennis, have a key role in the epidemiology of citrus stubborn disease [21].

Table 3 Results of S. citri detection by Enzyme-linked Immunosorbent Assay (ELISA) in citrus and non-citrus plants collected in the Fars province in Iran
Full size table

PCR Amplified Spiralin Genes of 11 S. citri Strains

As expected, primer pair NesF/NesR amplified a 16S rDNA fragment of 1,311 bp in all isolates (Fig. 1a). Spiralin gene amplification with primer pair F1/R1 was performed on 11 DNA extracted from different S. citri strains isolated from different localities in Iran where the stubborn disease occurs (Fig. 1b). In the reference strain GII-3 primer F1 is located about 62 bp upstream of the spiralin gene start codon and primer R1 is located 29 bp downstream of the spiralin stop codon and amplification leads to a product of 817 bp length. As expected, in 10 strains tested the length of the amplified fragment (Fig. 1b tracks 2–11) is almost the same and around 817 bp. An unusually long product of 1,135 bp was amplified with the Shiraz I isolate (Fig. 1b track 12). Among the set of isolates for which the spiralin gene has been sequenced in the present work, only Shiraz I was significantly impaired during growth in axenic medium and could not survive beyond the second passage in LD10 medium. The factors responsible for this growth defect are unknown, as this strain may not have solely undergone a modification in the spiralin gene sequence, but also mutations in genes necessary for a rapid and efficient adaptation to the host-cell free medium. Yet, it should be noted that there does not appear to be any contradiction between the growth defect in LD10 medium and the capacity of this strain to multiply in the infected fruit environment. Indeed, a single PCR product was obtained from the Shiraz I isolate, strongly suggesting that the strain bearing the longest spiralin gene was predominant, if not alone, in the symptomatic fruit, from which it was isolated. The PCR experiments further indicate that the sub-culturing trial did not allow the selection of a subpopulation of S. citri carrying a shorter spiralin gene and best suited for growth in LD10 medium. The possibility that an internal duplication of spiralin gene occurred in LD10 medium can thus reasonably be eliminated. These observations strongly argue in favor of an ability of Shiraz I strain to multiply in the citrus environment, but not in LD10 under our experimental conditions.

Fig. 1
figure 1

Electrophoresis pattern of PCR products with S. citri samples using primer pairs NesF/NesR (a) and F1/R1 (b). lane: 1, negative control lanes 2–12, S. citri isolates: Darab II (lane 2), Darab IV (lane 3), Darab XVIII (lane 4), Fasa I (lane 5), Fasa II (lane 6), Fasa III (lane 7), Fasa VII (lane 8), Kafr II-2 (lane 9), Firouzabad III (lane 10), Saf I (lane 11), and Shiraz I (lane 12). M, GenRuler DNA ladder

Full size image

Sequence and Phylogenetic Analysis of S. citri Isolates

For 4 strains the nucleotide sequences of the amplified fragments (Fig. 1b tracks 6, 8–10) revealed a length of 817 bp while the sequence of PCR products obtained from the other 6 strains (Fig. 1b tracks 2–5, 7, and 11) revealed a longer fragment size of 826 bp. Nucleotide sequence alignments of PCR products (817 bp) from strains Fasa II [23], Fasa VII, Kafr II-2, and Firouzabad III showed that the spiralin gene had 99–100 percent identity with the spiralin gene of the S. citri reference strain GII-3. The spiralin coding sequence contained in the nucleotide sequences of the longer PCR products (826 bp) of strains Darab II, Darab IV, Darab XVIII, and other previously reported isolates Fasa I, Fasa III [24], and Saf I [23] had more than 92.7 % identity to the GII-3 spiralin gene. In Shiraz I isolate (GeneBank Acc. No. JN860712), spiralin gene was 1,041 bp long due to the duplication of nucleotides 153–468. RFLP typing of the spiralin gene, useful in distinguishing S. citri strains [17, 22] was applied to the Shiraz I spiralin gene. The RFLP pattern revealed the presence of three DraI sites at positions 197, 512, and 990, two RsaI sites at positions 273 and 588, and two MboI sites at positions 426 and 741 (data not shown). Comparison between spiralin gene restriction maps of S. citri strains (belonging to the six groups defined previously) and Shiraz I strain confirmed the sequence duplication in the Shiraz I spiralin sequence. In addition to sequence results, RFLP studies made it possible to define a new group (group seven) including at this time only the Shiraz I strain. For further characterization of S. citri isolates, nucleotide sequences of the 16S rDNA and spiralin genes were aligned with those available in the GenBank database. This analysis showed variation in both genes among Iranian isolates at the nucleotide level, with lower sequence diversity in 16S rRNA gene as compared to spiralin gene (data not shown). The bootstrapped phylogenetic tree (Fig. 2) indicated that spiralin genes from Iranian strains were distinct from those of S. melliferum, S. kunkelii, and S. phoeniceum and formed two separate clusters. In cluster one, Khafr II-2, Fasa II, Firouzabad III, and Fasa VII isolates were identical or strikingly similar to S. citri GII-3 and to several strains that were previously described [23] including the Khafr I isolate from sesame (Acc. No. JN974243) and Saf II and Saf III from safflower (Acc. No. JN974240, JN974241, respectively) not shown on the phylogenetic tree. Isolates Firouzabad III and Fasa II (Acc. No. JN974242) from citrus could not be discriminated in the phylogenetic tree and formed together another distinct subgroup in this cluster. Isolate Shiraz I stood alone in a separate branch in this tree. The second cluster included the other isolates (Darab IV, Saf I, Darab II, and Darab XVIII) which were strikingly similar to S. citri Fasa I and Fasa III isolated from C. haematoceps, recently reported to constitute group 6 [22] with the closest relation to Palmyre strain (group 5). The new members in this cluster 2 have been isolated from sesame (Darab XVIII), safflower (Saf I), and C. haematoceps (Darab II and Darab IV) but despite many attempts, we have not been able to isolate them from citrus species.

Fig. 2
figure 2

Phylogenetic relationship of S. citri isolates based on spiralin gene. The evolutionary history was inferred by using the maximum likelihood method based on the data specific model [31]. The tree with the highest log likelihood (−2249.0779) is shown. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Initial tree(s) for the heuristic search were obtained automatically as follows. When the number of common sites was <100 or less than one-fourth of the total number of sites, the maximum parsimony method was used; otherwise BIONJ method with MCL distance matrix was used. A discrete Gamma distribution was used to model evolutionary rate differences among sites (5 categories (+G, parameter = 0.7137)). The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 18 nucleotide sequences. Codon positions included were 1st + 2nd + 3rd + Noncoding. All positions containing gaps and missing data were eliminated. There were a total of 696 positions in the final dataset. S. melliferum was used as outgroup. Evolutionary analyses were conducted in MEGA5 [39]

Full size image

General Features of Spiralins in Iranian Isolates

For Iranian isolates except for Shiraz I, the deduced amino acid sequences from spiralin gene contained 241 or 244 amino acids. The length of 244 amino acids is shared by all spiralins belonging to group 6 [22] and slightly differs from that of spiralins from strains of groups 1–4 (241 amino acids) and group 5 (242 amino acids) [17]. In isolate Shiraz I, the spiralin deduced sequence is 346 amino acids long (Fig. 3). The putative physiological significance of the additional domain in the spiralin of Shiraz I in the vector transmission of the bacterium is discussed below. All seven spiralin deduced sequences from Iranian isolates were mostly identical for their N-terminal lipoprotein signal sequence to that of S. citri GII-3 (Fig. 3). The cysteine allowing lipid modification was found at position 24. Spiralins from the type strain GII-3, from the 3 Iranian isolates belonging to group 1 (Fasa VII, Fasa II, Firouzabad III) and from Shiraz I isolate share the same signal peptide sequence. Similarly to Palmyre strain (group 5) [17] the isolates Fasa III, Saf I, Darab XVIII from group 6 substituted the aromatic amino acid F at position 11 for L, an aliphatic amino acid of similar size (Fig. 3). In all spiralins the presence of two highly positively charged lysine residues within the first three amino acids (n-region of the signal peptide), and the hydrophobic stretch of 17 amino acid length (h-region) are features recognized in lipid modification of the cysteine at position 24 with a diacyl-glycerol [4, 7]. In addition the distinct sequence (VVAC) at the C-terminal end of the signal peptide corresponding to a lipobox [4] and which constitutes the cleavage site for the lipoprotein-specific SPase could be identified in all spiralins. In addition, before the C-terminal end of spiralin of the 3 spiroplasmas from group 6 and of the strain Palmyre have the same motif (NKKVTP) different from the LAPAN motif present in the other spiralins including the reference strain GII-3.

Fig. 3
figure 3

Partial duplication of spiralin amino acid sequence in S. citri isolate Shiraz I compared to reference strain GII-3, isolates Fasa VII, Fasa II, Firouzabad III (group 1), Palmyre strain (group 5), isolates Fasa III, Saf I, Darab XVIII (group 6). Duplicated sequence in Shiraz I spiralin is shown below alignment. Previously identified short amino acids conserved regions [19] are boxed

Full size image

Repeated Regions in Spiralins

Foissac et al. [12] reported conserved repeated motifs, which are identical in spiralins from different Spiroplasma species. These motifs are located around positions 50 and 160 (Fig. 3) and fit the sequence AnPKQVTnaE (conserved amino acids are in uppercase) in S. citri. Similar repeated elements were identified in all our isolates. These repeated motifs contained substitutions to amino acids sharing similar physico-chemical properties. For instance, isolates of group 6 showed a change of the polar amino acids KQ to the polar pair TN in the second site (Fig. 3). Isolate Darab XVIII showed an additional amino acid change, with L in place of P. The aliphatic and hydrophobic characters of L and P are shared by V found in spiralin at the equivalent position in S. phoeniceum. Taken together these data indicate that these repeated sequences are well conserved among spiralins. While all spiralins in formerly reported isolates of S. citri including wild type strain GII-3 as well as all other novel isolates described in this study contain two of these repeated sequences, the Shiraz I isolate has three conserved repeated regions. In order to determine whether larger less conserved repetitive elements could be evidenced in S. citri spiralins, spiralin repeated motifs for GII-3 and Shiraz I were determined by combining MEME/MAST and RADAR automatic tools. Both methods allowed the identification of at least three degenerate repeated motifs in S. citri spiralins. As shown in Fig. 4a, in Shiraz I the duplication event not only led to the duplication of the short conserved motif previously identified (overlapping motifs 2 and 3 identified using MEME) but also to the duplication of two 29-amino acids long motifs 1 and 2, each being present as two degenerate copies in other spiralins including GII-3 spiralin (Fig. 4b). As a result of the duplication event, repeated motifs 1, 2, and 3 were found to cover a 279 amino acid long sequence in Shiraz I spiralin, as compared to 174 amino acids in GII-3. Using the Predict Protein Meta server [41] the repeated regions in Shiraz I and GII-3 were predicted to be rich in beta strands (data not shown), a feature commonly found in eubacterial lectins [19]. S. citri spiralin is suspected to act as a bacterial sugar-binding adhesin (lectin) [24]. The recognition of eukaryotic cells through repeated regions in lectins and adhesins has been described for several Gram positive bacteria [15, 40]. Repetitive elements present in diverse adhesins of mollicutes have been also shown to allow the bacterial binding to eukaryotic cells [6, 30]. The occurrence of a region rich in repeats and strands in spiralin could support the hypothesis of the involvement of this lipoprotein in the binding to insect vector glycoproteins through repeated elements [24].

Fig. 4
figure 4

Conserved motifs detected in spiralins by MEME/MAST analysis. a Shiraz I; b GII-3

Full size image

In conclusion, Iranian isolates of S. citri showed diversity in both spiralin and part of the 16S rDNA gene, the spiralin gene being less conserved than the 16S rDNA. Our isolates could be assigned to two previously reported RFLP groups 1 and 6. The characterization of the spiralin gene of Shiraz I allowed the establishment of a novel RFLP group (group 7). The phylogenetic tree constructed using the Shiraz I spiralin sequence deleted of the additional duplicated sequence indicated that this truncated sequence clustered with spiralins from strains belonging to cluster one (data not shown), suggesting that Shiraz I isolate might have originated from this group. General common features of spiralins from the different groups included the presence of motifs responsible for N-terminal lipoylation and of internal repeats. The analyses of spiralin sequences for repetitive elements indicated that S. citri spiralins contained 29-amino acids long degenerate repetitive elements. In Shiraz I, internal repetition could afford spiralin enhanced evolutionary prospects due to an increase of its available binding surface area, as proposed for other surface-associated proteins [3]. One might hypothesize that a modification in the length of this putative lectin may be responsible for a change in efficiency of bacterial binding to the vector cells. Enlargement of the protein-exposed surface by internal amplification could also correspond to an alternative to homo-oligomerization [1]. Indeed spiralin has been shown to cover most of the surface of the spiroplasma membrane [12] and is suspected to form oligomers, mainly dimers [43]. Therefore, the putative physiological relevance of Shiraz I spiralin additional domain during transmission by the insect vector will deserve additional research attention in the future. More generally, the present work raises the question of the contribution of partial internal gene duplication in the evolution of spiroplasma genomes and of the possible consequences of such DNA rearrangements in spiroplasmal adaptation to their hosts.