Genetic transformation of sweet oranges with the D 4 E 1 gene driven by the AtPP 2 promoter

The objective of this work was to produce transgenic 'Pêra' and 'Valência' sweet orange plants using the D4E1 gene driven by the Arabidopsis thaliana phloem protein (AtPP2) promoter and to quantify transgene expression in different transformation events. Genetic transformation experiments were carried out with epicotyl segments co‐cultivated with Agrobacterium tumefaciens. Six plants from 'Pêra' sweet orange and seven plants from 'Valência' sweet orange were confirmed as different transgenic events by means of the polymerase chain reaction (PCR) and the Southern blot techniques. Transgene expression was quantified using real‐time quantitative PCR. D4E1 gene expression levels vary from 5 up to 50 times among different transformation events.


Introduction
Brazil is the world's largest producer and exporter of frozen concentrated orange juice.It exports 98% of the juice produced in the country, which represents 85% of the juice consumed worldwide.The Brazilian citrus industry is directly and indirectly responsible for more than 230,000 jobs in the country (Neves et al., 2010).The primary scion cultivars used in Brazilian orchards are 'Hamlin','Pêra','Valência',and 'Natal';therefore,approximately 90% of the orchards comprise these cultivars, which implies low genetic variability and a contribution to increased vulnerability, especially to diseases (Machado et al., 2010).
Huanglongbing (HLB) is a disease associated with phloem-restricted, Gram-negative bacteria.To date, three species of HLB-associated bacteria were identified: "Candidatus Liberibacter africanus" (CLaf), "Ca.L. asiaticus" (CLas), and "Ca.L. americanus" (CLam).These bacteria were classified according to the continent where they were first identified.The CLam and CLas species are found in Brazil, and the latter is the most frequently observed in the country.The bacteria are transmitted through propagation by infected budwood or by the Diaphorina citri psyllid, which is the disease vector in Brazil (Bové, 2006).
Gene insertion through genetic transformation is an alternative for controlling this disease given the lack of resistance to HLB in the citrus genus, low effective control measures, and rapid disease dissemination (Belasque Júnior et al., 2010).Antimicrobial peptides have been identified in several organisms, from bacteria to humans, and they are recognized as key components of the defense system.D4E1 is a synthetic peptide with 17 amino acids and a beta sheet conformation in solution that does not change when interacting with cell membranes (De Lucca et al., 1998); it originates from and operates similarly to cecropins by forming pores in the cell membrane of Gram-positive and Gram-negative bacteria.Tests to evaluate hemolytic activity for D4E1 in mammalian cells showed that this peptide has low toxicity at the concentrations used for phytopathogen elimination (Stover et al., 2013).
The expression of synthetic peptides in transgenic plants offers advantages over natural analogs, including greater specificity, efficiency at low concentrations and a smaller size compared with natural peptides, without contamination or toxicity for the transgenic plant tissue; furthermore, it better resists degradation compared with natural cecropin (Cary et al., 2000).Transgenic plants from a Populus tremula L. x P. alba L. hybrid that expressed the D4E1 peptide showed a decrease in symptom severity for diseases caused by Agrobacterium tumefaciens and Xanthomonas populi pv.populi (Mentag et al., 2003).The genetic transformation of tobacco with a gene encoding the synthetic antimicrobial peptide D4E1 controlled the growth of Aspergillus flavus and Verticillium dahliae fungi (Cary et al., 2000); whereas the D4E1 gene expression in cotton plants reduced disease symptoms caused by Thielaviopsis basicola (Rajasekaran et al., 2005).In a recent work, Stover et al. (2013) reported excellent results regarding in vitro control by the D4E1 peptide of Proteobacteria, which belong to the same family of HLB-associated bacteria (Ca.Liberibacter spp.).Presumably, the expression of the D4E1 peptide in phloem might be effective for controlling HLB-associated bacteria in citrus.
The expression of an antimicrobial peptide in the citrus plant phloem, particularly in young leaves, can aid in controlling HLB because this disease is associated with bacteria restricted to phloem cells, which are primarily spread by psyllids while feeding on young leaf phloem (Dutt et al., 2012).Another advantage of preferential expression in phloem is that it decreases the expression of the transgene in other tissues (Zhao et al., 2004).
This work is pioneering in Brazil in relation to quantification by quantitative PCR (qPCR) of gene expression driven by phloem specific promoter in citrus, especially in sweet orange.Although these techniques have already been applied in other studies, the genetic transformation of citrus is a complex issue, since this woody species shows different responses depending on the cultivars and gene constructs used.
The objective of this work was to produce transgenic 'Pêra' and 'Valência' sweet orange plants using the D4E1 gene driven by the AtPP2 promoter and to quantify transgene expression in different transformation events.

Materials and Methods
The study was conducted at the Laboratório de Biotecnologia de Plantas Hortícolas, Escola Superior de Agricultura Luiz de Queiroz, Universidade de São Paulo, SP, Brazil.For the genetic transformation experiments, the strain of A. tumefaciens EHA 105 was used, which expressed the binary vector pCAMBIA 2201/D4E1 driven by the AtPP2 promoter and the nopaline synthase (NOS) terminator, as well as the nptII gene for kanamycin-resistance selection (Figure 1).'Pêra' and 'Valência' sweet orange cultivars were selected as the genetic material for the transformation experiments.Seed extraction, explant preparation, transformation, and plant acclimation procedures were carried out according to Miyata et al. (2012).
The samples were amplified in a thermocycler programmed for the following conditions: 94°C for 2 min; 40 cycles at 94ºC for 30 s, 58°C for 30 s, and extension at 72°C for 30 s; and the reaction was terminated by maintaining the temperature at 72°C for 4 min.The amplification product size of 353 bp was assessed through electrophoresis on a 1.5% agarose gel stained with ethidium bromide (0.03 µg mL -1 ) and photographed under UV light.
Southern blot analysis was performed as described by Southern (2006) to assess whether T-DNA was integrated into the plant genome, using a 275 bp probe with the D4E1 gene and NOS-T terminator (D4E1/ NOS) (Table 1).
RNA extraction from leaves was performed using TRIzol (Life Technologies, Carlsbad, CA, USA) and the extraction protocol recommended by the manufacturer for performing expression analysis using real-time qPCR.Subsequently, the RNA was purified using the RNeasy Plant Mini Kit (Qiagen, Hilden, Germany) followed by a DNase treatment using the RNase-Free DNase Set (Qiagen, Hilden, Germany) according to the manufacturer's instructions.The NanoDrop 2000c spectrophotometer (Thermo Scientific, Wilmington, DE, USA) was used to quantify and assess the RNA.RNA integrity was assessed using 1.5% agarose gel at 60 V for 40 min.Reverse transcription was performed in a total volume of 20 μL, with total RNA (1 μg), oligo (dT) primer (10 µmol L -1 ), and Moloney murine leukemia virus (M-MLV) enzyme (1 µL).The reaction was carried out according to the manufacturer's instructions (Life Technologies, Carlsbad, CA, USA).The cDNA samples were standardized to 6.6 ng µL -1 upon reaction completion and stored at -20ºC.The relative expression was quantified through real-time qPCR analysis upon cDNA synthesis completion.
Two reference genes, cyclophilin (cyc) and ubiquitin (ubq), were used to analyze gene expression through real-time quantitative PCR and to normalize the D4E1 gene expression levels via specific primers (Table 1).Different concentrations of primers and samples were used to optimize the reaction efficiency.Real-time qPCR analyses were performed in 15 µL solution with 5 µL of cDNA (30 ng of cDNA), 0.3 µL forward and reverse primers, 7.5 µL Fast Sybr Green Master Mix (Life Technologies, Carlsbad, CA, USA), and 1.9 µL nuclease-free water.Samples were amplified using three replicates per sample, including samples from non-transgenic plants and a control sample without DNA.The qPCR analyses were performed in a 7500 Fast System (Life Technologies, Carlsbad, CA, USA) using the comparative method and the fast mode in the device's default program.The amplicons were assessed at the end of the reaction through melting analysis, which identifies the average melting point of the amplicons formed during the reaction.The detection of a single dissociation peak for the PCR product strands was regarded as evidence for reaction specificity.AGAGTATGCAGAGGAATGG/GTCCTTAACAGAAGTCCGT 107 ubq (1)  TTCGTCAGTTGACTAATCCT/GTTGCTGTGTTGACTGTG 95 D4E1 CCATGGGATTTAAGTTGAGAG/GGTCACCTTACAACTTAATCTTAGC 69 (1) Boava et al. (2011).
The quantification cycle (Cq) value was detected for the target gene and for the two reference genes in each transformation event.The data generated during the qPCR analyses were evaluated to assess efficiency using the LinRegPCR software (Ramakers et al., 2003).This software performs a linear regression for each amplification curve and generates an efficiency value for each reaction, as well as an average efficiency for each gene.After this calculation, the Cq data produced through qPCR analysis were entered into the GenEx software, version 5.3.6.170(Kubista et al., 2006), and the average efficiency calculated from LinRegPCR was used to correct the Cq value for each gene.After correcting the efficiency, the software compared the data using the following relative quantification formula: RQ = 2 -∆∆Cq (Schmittgen & Livak, 2008).The calibrator was the lowest D4E1-expressing transformant.

Results and Discussion
The 'Valência' sweet orange explants showed higher bud regeneration and transformation efficiency than those of 'Pêra' sweet orange (Table 2).A similar performance has been observed for 'Pêra' sweet orange in other genetic transformation studies, with lower regeneration and transformation efficiency compared with 'Valência' sweet orange (Boscariol et al., 2003;Fávero et al., 2012).This difference is attributed mainly to the cultivar and to the affinity with A. tumefaciens (Mendes et al., 2002;Almeida et al., 2003;Miyata et al., 2011).Citrus species and cultivars show different responses to genetic transformation.
Therefore, specific protocols must be developed and adapted to each genotype (Donmez et al., 2013).
The AtPP2/D4E1 construct was found in six plants from 'Pêra' sweet orange and in seven plants from 'Valência' sweet orange using PCR analysis (Figure 2).For Southern blot analysis, the transgene copy number in the 'Pêra' sweet orange genome was one for the for the six events, and the integration of two copies in two events and of one copy in the remainder was verified in the 'Valência' sweet orange plants (Figure 3).Noticeably, no event showed gene silencing according to the qPCR analysis results because all transformation events presented gene expression.Gene silencing is typically detected in plants with multiple copies integrated at one or several loci; however, gene silencing was not reported in citrus plants with up to four transgene copies (Dutt et al., 2012;Benyon et al., 2013).
D4E1 gene expression analysis through qPCR showed variable expression, from Log 10 0.76 to Log 10 1.7 , which corresponds to 5.75 (event 7) and 50 (event 6) times compared with the lowest D4E1-expressing transformant (Figure 4).The different expression   levels are related to the number of copies and to the transgene-insertion position in the genome (Elmayan & Vaucheret, 1996).The transgene expression values are similar to those found by Benyon et al. (2013), who observed values from 4.24 to 118.88, corresponding to log 10 0.62 to log 10 2.07 for uidA gene expression driven by the AtSUC2 phloem-specific promoter.Several phloem-specific promoters have been studied for genetic transformation because constitutive transgene expression is unnecessary when localized expression suffices.Miyata et al. (2012) demonstrated specific GUS expression driven by the AtPP2 promoter in phloem tissue of transgenic sweet orange.However, uidA gene expression driven by this type of promoter, quantified by qPCR, is relatively lower when compared to genes driven by constitutive promoters (Dutt et al., 2012;Benyon et al., 2013).

Conclusions
1. 'Pêra' and 'Valência' sweet oranges can be transformed with the D4E1 gene driven by the AtPP2 promoter.
2. D4E1 gene expression levels vary from 5 up to 50 times among different transformation events.

Figure 1 .
Figure 1.Schematic representation of the T-DNA region in the pCAMBIA2201 vector used for genetic transformation.Gene construct with the nptII selection gene controlled by the CaMV35S promoter and the D4E1 gene controlled by the AtPP2 promoter.LB, left border; RB, right border.NcoI BstEII

Table 1 .
Reference and target gene primer sequences and amplicon sizes used in the PCR and qPCR experiments.