The CchGLP cDNA sequence contains a 24 bp leader section, poly A tail, and an ORF encoding a predicted protein of 203 amino acids. Figure 1 shows a phylogenetic tree of amino acid sequences from several solanaceae and Arabidopsis thaliana plant proteins, which according to BLAST analysis were the most related to CchGLP. The deduced protein of CchGLP was more related to a GLP of Nicotiana tabacum (93% identity, accession number: BAH15357).

The CchGLP protein sequence was submitted to bioinformatic analysis in order to characterize it in more detail. A highly conserved cupin 2 domain (not shown) and structure that contains three boxes representing the germin domains according to Lu et al. [14] are present in CchGLP (Figure 2). In addition, CchGLP was predicted to contain a N-terminal signal peptide of 18 amino acids, and one potential N-glycosylation site from amino acids 44–47 (NVTV) (Figure 2). Moreover, a putatively important sequence KGD (or sometimes KGE) that has been detected in over 50% of GLPs, but not in germins was detected in CchGLP (Figure 2).

The number of amino acids in the deduced protein as well as sequence features were typical of those reported in plant GLPs elsewhere [2,4,6]. Glycosilation in germins appears to be essential in protein-protein interactions but not for enzyme activity as demonstrated by Pan et al. [15].

To carry out a more detailed analysis of CchGLP, the recombinant protein was produced in the E. coli system. A protein with the expected molecular weight for CchGLP protein without the signal peptide (20.5 kDa) was mainly detected in the soluble fraction (over 20% of total proteins). The recombinant protein was observed 4 h after IPTG induction, but was not observed in the control test without induction (Figure 3A). Recombinant CchGLP protein was further verified with Western blot analysis using specific antibodies against the His-tail tag added to the protein (Figure 3B). After purification of the recombinant protein through Ni-columns, the elution fractions showed a band of 20.5 kDa with a purity of above 95% (Figure 3C). This purified protein was used for enzymatic activity determinations. It is worth mentioning that in contrast to reported problems of expression of germin proteins in E. coli due to protein aggregation [16], in this work we did not have these problems that appeared to reduce expression in this system.

Proteins belonging to the group of “germins” show an OXO activity mainly related to their role in defense [2]. On the other hand, “germin-like” proteins can also accumulate during pathogen attack although they are totally devoid of OXO but display SOD activity [2]. Moreover, several germin proteins expressed in Pichia pastoris might form multimers which are required for activity [15] and homohexameric barley germin, for which the 3D structure has been solved, displays both OXO and SOD activities [17,18]. Thus, in order to determine whether CchGLP displayed either OXO or SOD activity (or both activities), these assays were carried out. OXO activity was not detected for CchGLP (data not shown), instead the CchGLP purified protein displayed SOD activity (Figure 4A). Inhibition assays of SOD activity using either H₂O₂ (inhibits Fe-SOD activity) or KCN (inhibits Cu/Zn-SOD activity) are used to identify the type of SOD activity [19]. On the other hand, Mn-SOD activities are resistant to both compounds [19]. Thus, inhibition assays of SOD activity revealed that CchGLP was not inhibited by either compound, indicating Mn-SOD activity for this protein (Figure 4B,C).

Plant Mn-SODs have been reported to occur extracellularly as well as in mitochondria and peroxisomes [5,18]. Moreover, Mn-SOD activities have previously been associated with defense against biotic stress in plants [13,18].

The CchGLP cDNA sequence was compared with a genomic sequence obtained using PCR with specific oligonucleotides. No differences were detected between the sequences, indicating no introns are present in the CchGLP genomic sequence. Hybridization assays using total genomic DNA from C. chinense BG-3821 (geminivirus-resistant) and UX-SMH-55 (geminivirus-susceptible, data not shown), showed either a single or two bands when genomic DNA was digested with restriction enzymes cutting outside or once inside the CchGLP gene, respectively (Figure 5A). CchGLP was not detected in hybridization assays with Capsicum annuum cv. “Sonora Anaheim” genomic DNA (data not shown). These results indicated a single copy of the CchGLP gene in both C. chinense accessions, and the absence of the CchGLP gene in C. annuum. On the other hand, CchGLP expression was studied in C. chinense BG-3821 during compatible and incompatible interaction with Pepper golden mosaic geminivirus (PepGMV) as well as two plant-pathogen signal molecules, namely ethylene and jasmonic acid. It was observed that CchGLP was induced at least from 12 h to 48 h in inoculated leaves post-inoculation or application, either by PepGMV, PHYVV, salicylic acid or ethylene (Figure 5B,C). CchGLP induction was also observed in newly emerged systemically geminivirus-infected leaves 15 days post-inoculation (Figure 5B). On the other hand, neither P. capsici infection nor jasmonic acid application induced CchGLP expression in C. chinense BG-3821 (Figure 5B,C). Compatible interaction in a geminivirus-susceptible C. chinense accession (UX-SMH-55) did not induce CchGLP when inoculated neither with PepGMV, PHYVV, P. capsici nor Et, JA or SA applications (not shown).

Several lines of evidence shows that GLPs are important in plant protection against pathogens and pests, or induced with applications of chemicals such as salicylic acid, ethylene or hydrogen peroxide [20–22]. A previous report showed that CchGLP was induced in incompatible interactions between C. chinense BG-3821 and PHYVV, Xanthomonas campestris pv. vesicatoria and not in a compatible interaction with oomycete Phytophthora capsici [11]. In this study, it was shown that CchGLP was also induced in incompatible, but not in compatible, interactions with another geminivirus affecting pepper in Mexico (PepGMV) in C. chinense accessions. The phenotype of susceptibility to geminiviruses in C. chinense UX-SMH-55 evaluated, might, at least partially, be explained by the absence of CchGLP induction in all infections and plant signal molecules applications evaluated. Defense response genes are also uninduced in other susceptible plants as showed elsewhere [2,11,12]. The fact that the CchGLP gene was not detected in the genome of the geminivirus-susceptible C. annuum cv. “S. Anaheim”, suggest either a low homology with this gene or absence in the genome of this pepper species, as occurred with other genes previously identified in C. chinense BG-3821 [12]. Moreover, CchGLP was also demonstrated to be induced by SA and Et applications in the absence of pathogens [11]. In this study, it was demonstrated that Et but not JA induced local and systemic CchGLP gene expression in the absence of pathogens. The SA and JA signaling pathways are mutually antagonistic, whereas SA and Et pathways appear to interact [23,24]. Further studies are necessary to determine the role of this signal routes in this plant-microbe interaction.

Based upon the CchGLP sequence (accession number DQ677335), two oligonucleotides to amplify the CchGLP open reading frame without signal peptide were designed. Oligonucleotide sequences included restriction sites for Msc I and Xho I enzymes in the forward and reverse primers, respectively. The sequence of these primers (restriction sites are written in lower case) was as follows: (Forward) 5′- tggccaGCTGTTCAAGATTTCTGCGTCGC-3′, (Reverse) 5′-gagctcCTTAATCGTAGCTTCATCAAGG-3′. PCR reaction mix contained the following components: 0.75 μL each dNTPs (2.5 mM), 2 μL of oligonucleotides (50 ng/mL), 0.5 μL of Taq DNA polymerase (6 U/μL), 1 μL of DNA from CchGLP cloned in PCR^TM 4-TOPO^R (Invitrogen) in a reaction volume of 50 μL. PCR conditions to amplify the CchGLP gene were: 94 °C, 1 min; 51 °C, 1 min and 72 °C, 2 min for 35 cycles. PCR products were visualized on 1.5% agarose gels using a digital image system (1D Image Analysis Software, version 3.02; Kodak Digital System, Rochester, NY, USA). This amplified product (containing Msc I and Xho I restriction sites) was purified from agarose gel using a gene clean commercial kit (QIAGEN, Mexico, DF, Mexico) and cloned into PCR^TM 4-TOPO^R. This construct was digested with the aforementioned enzymes and the CchGLP fragment was purified again from agarose gel. The pET22b (+) protein expression vector (Novagen), was digested with Msc I and Xho I and then ligated with the CchGLP fragment using T4 ligase (Promega) for 3 h at 16 °C. The resulting construct was named as pET22b(+)-CchGLP, and further evaluated using protein expression and enzymatic activity assays.

A BL21 E. coli strain was transformed with pET22b (+)-CchGLP. Transformed cells were grown overnight at 32 °C in liquid Basal Salts-Glucose (BSG) medium [25]. Then, a 1 mL aliquot was inoculated into a flask with 100 mL of BSG medium with ampicillin (50 mg/mL) and 1 mM of IPTG, and grown until an OD 0.6 of 32 °C was reached. Aliquots of 5 mL of this medium were evaluated from 3 to 24 h for GLP protein expression. Samples were centrifuged at 10,000 rpm for 10 min and then resuspended in 500 μL saline phosphate buffer (SPB) [26]. Cells were disrupted using a sonicator (Ultrasonic processor) with an amplitude of 30% and 9.9 s pulses during 2 min, then centrifuged at 10,000 rpm for 10 min. Soluble fraction was separated and resuspended in SPB. Protein quantification was carried out using the protein assay kit (Bio-Rad) [27]. Proteins were separated on SDS-PAGE gels (gradient from 4–20%) and stained with Coomasie blue. The molecular size marker used was Benchmark^TM (Invitrogen). CchGLP protein purification under native conditions was carried out using affinity chromatography (ProBond, Invitrogen) following the manufacturer’s recommendations and further evaluated by Western blot analysis using Anti-His (c-term)-AP antibodies (Invitrogen).

Assays for OXO activity were carried out essentially as described in Dumas et al. [28]; the positive control was a commercial oxalate oxidase of barley (SIGMA-Aldrich). SOD activity was determined using the procedure reported by Beauchamp and Fridovich [29]. The gel was incubated in phosphate buffer pH 7.5 containing 2.5 mM of NBT, then the gel was rinsed in distilled water. Afterward, the gel was incubated in phosphate buffer containing 28 μM of Riboflavine and 28 mM TEMED for 20 min, then rinsed with distilled water and placed on a soft light source. Determination of SOD type was carried out including inhibitory concentrations of either H₂O₂ or KCN when evaluating Fe-SOD or Cu-Zn-SOD, respectively [19]. Commercial Fe-SOD from E. coli strain and bovine erythrocyte Cu-Zn-SOD were used as experimental controls (Grupo Nutramex, Mexico, DF, Mexico).

Geminivirus-resistant (C. chinense line BG-3821) and susceptible (C. chinense line UX-SMH-55) plants were inoculated with 2 μg of cloned viral DNA components (cloned in SK+ bluescript plasmid, Stratagene, La Jolla, CA, USA) using a biolistic procedure with a particle delivery system (Model PDS 1000, Dupont, Wilmington, DE, USA) as reported [30]. Both pepper lines were previously identified and characterized in our group based on resistance or susceptibility to geminivirus infections [30]. Before bombardments, PHYVV components A and B, and component B of PepGMV were excised from bluescript plasmid using HindIII. Component A of PepGMV was excised from the plasmid with the enzyme EcoRI in order to increase the percentage of plants infected using the biolistic procedure [11,12,30]. Plants were inoculated at the 4 leaf-stage and maintained in a greenhouse at 26–28 °C as reported [30]. Inoculation of P. capsici was performed according to Barrera-Pacheco et al. [11]. In all cases, newly emerged uninoculated leaves were sampled at 15 days post-inoculation to analyze systemic CchGLP gene induction. As controls, plants were mock-inoculated with the plasmid bluescript or potato dextrose agar blocks. PHYVV and P. capsici inoculations were used as positive or negative controls, respectively, for CchGLP induction in C. chinense BG-3821 plants based on previous reports [11,12,30]. Applications of ethylene, jasmonic acid and salicylic acid were according to Barrera-Pacheco et al. [11]. Mock-inoculated plants for these treatments used water applications only.

The copy number of CchGLP in C. chinense BG-3821 was determined using standard Southern blotting protocols [26]. Gene induction studies of CchGLP in C. chinense BG-3821 and UX-SMH-55 were carried out by Northern blot (slot blot) as described in Barrera-Pacheco et al. [11]. A DNA fragment consisting of the complete ORF of CchGLP was used as a probe in both assays.

Total genomic DNA was extracted following the procedure of Anaya-López et al. [12]. This DNA was used as a template for CchGLP structural region amplification using the specific oligonucleotides aforementioned in Section 3.1. The PCR amplified fragment was directly sequenced and compared to that from CchGLP cDNA, using bioinformatic tools (see below).

Analysis of GLP amino acid sequences was carried out using Lasergene 8.0 software (DNAstar). A phylogenetic tree was generated using the MegAlign tool of Lasergene 8.0. Accession numbers of GLP amino acid sequences evaluated in the tree are: C chinense (ABG76199), A. thaliana (NP_177405), Auxin binding protein from Solanum nigrum (ADW66138), GLP 24 K from Nicotiana tabacum (BAC77634), GLP from Nicotiana tabacum (BAH15357), Auxin binding protein from Solanum nigrum (ADW66150) and auxin binding protein 19a from Prunus persica (Q9ZRA4). Prediction of signal peptide and glycosylation sites was performed using ExPASy proteomics server tools [31].