Disease Control and Integrated ManagementFree Access icon

Probing the Application of OmpA-Derived Peptides to Disrupt the Acquisition of ‘Candidatus Liberibacter asiaticus’ by Diaphorina citri

    Affiliations
    Authors and Affiliations
    • Marcus Vinícius Merfa
    • Eduarda Regina Fischer
    • Mariana de Souza e Silva
    • Carolina Sardinha Francisco
    • Helvécio Della Coletta-Filho
    • Alessandra Alves de Souza
    1. Centro de Citricultura Sylvio Moreira, Instituto Agronômico-IAC, Cordeirópolis, SP 13490-970, Brazil

    Abstract

    Huanglongbing (HLB) is currently the most devastating disease of citrus worldwide. Both bacteria ‘Candidatus Liberibacter asiaticus’ (CLas) and ‘Candidatus Liberibacter americanus’ (CLam) are associated with HLB in Brazil but with a strong prevalence of CLas over CLam. Conventionally, HLB management focuses on controlling the insect vector population (Diaphorina citri; also known as Asian citrus psyllid [ACP]) by spraying insecticides, an approach demonstrated to be mostly ineffective. Thus, development of novel, more efficient HLB control strategies is required. The multifunctional bacterial outer membrane protein OmpA is involved in several molecular processes between bacteria and their hosts and has been suggested as a target for bacterial control. Curiously, OmpA is absent in CLam in comparison with CLas, suggesting a possible role in host interaction. Therefore, in the current study, we have treated ACPs with different OmpA-derived peptides, aiming to evaluate acquisition of CLas by the insect vector. Treatment of psyllids with 5 µM of Pep1, Pep3, Pep5, and Pep6 in artificial diet significantly reduced the acquisition of CLas, whereas increasing the concentration of Pep5 and Pep6 to 50 µM abolished this process. In addition, in planta treatment with 50 µM of Pep6 also significantly decreased the acquisition of CLas, and sweet orange plants stably absorbed and maintained this peptide for as long as 3 months post the final application. Together, our results demonstrate the promising use of OmpA-derived peptides as a novel biotechnological tool to control CLas.

    Huanglongbing (HLB) is currently the most devastating disease of citrus worldwide (Gottwald 2010). Affected trees produce low-quality fruits with a reduced yield and have a critically reduced plant lifespan, which together result in significant economic losses to most of the main citrus-producing areas globally (Bassanezi et al. 2020; Bové 2006, 2014; Graham et al. 2020; Zhou 2020). In Brazil, although ‘Candidatus Liberibacter americanus’ (CLam) and ‘Candidatus Liberibacter asiaticus’ (CLas) were initially reported as associated with HLB (Coletta-Filho et al. 2004; Teixeira et al. 2005), CLas has become prevalent over CLam in Brazilian groves over the years and is more efficiently transmitted between citrus plants (Lopes et al. 2009a, b), suggesting higher fitness of CLas. In the United States, Asia, and Central America, only CLas is associated with HLB (Bové 2006, 2014).

    Candidatus Liberibacter spp.’ comprise an important group of unculturable plant pathogenic bacteria and endophytes that are phloem-limited, psyllid-transmitted, fastidious bacteria belonging to the α subdivision of Proteobacteria (Jagoueix et al. 1994; Merfa et al. 2019; Wang and Trivedi 2013). In addition to citrus, they affect an array of other important crops worldwide, including potato, carrot, tomato, and pear (Bové 2014; Nelson et al. 2012; Thompson et al. 2013). HLB-associated bacteria are transmitted from tree to tree by the Asian citrus psyllid (ACP) Diaphorina citri (Bové 2006). In the insect vector, CLas is acquired by uninfected psyllids upon feeding on HLB-symptomatic plants and ingestion of cells. After its acquisition via feeding, CLas colonizes the psyllid host both inter- and intracellularly as a systemic, circulative, and propagative endosymbiont that penetrates several organs and moves within the hemolymph on its way to the salivary gland lumen, from which this bacterium may be secreted into host plants during feeding (Ammar et al. 2011; Ghanim et al. 2017). Moreover, CLas accumulates within gut cells of D. citri in endoplasmic reticulum–associated vacuoles, where it seems to propagate, although putative replication sites have also been observed in several other organs, including the salivary glands (Ammar et al. 2019; Ghanim et al. 2017).

    The ability of CLas to cross multiple membrane barriers within ACPs and cope with different microenvironments suggests that a complex regulatory and mechanistic interplay is required for the successful colonization of the insect vector by the bacterium (Ammar et al. 2011; Ghanim et al. 2017; Merfa et al. 2019). Therefore, disrupting CLas–ACP interactions may be key to improving management of HLB by impeding acquisition and further transmission via the insect vector (Bové 2006). Actually, current HLB management relies mainly on the use of healthy citrus seedlings and vector population control by applying chemical insecticides (Hall et al. 2013). However, the spray approach as a traditionally recommended practice has been mostly ineffective in controlling the introduction and spread of the disease (Bassanezi et al. 2013; Hall et al. 2013; Ichinose et al. 2010) and may lead to development of resistance among ACP populations (Tiwari et al. 2011).

    In recent years, the genomes of CLas and CLam have been sequenced (Duan et al. 2009; Wulff et al. 2014). Curiously, the outer membrane protein OmpA, known to be multifunctional and highly expressed in many enterobacteria (Smith et al. 2007), is absent in CLam (Wulff et al. 2014), suggesting that the absence of this protein in CLam may influence the interaction of this bacterium with its hosts in comparison with CLas. For instance, the infection of ACPs with CLas increases the overall fitness of the insect by increasing traits such as oviposition, egg development, flight, and population growth rate (Galdeano et al. 2020; Martini et al. 2015; Pelz-Stelinski and Killiny 2016; Ren et al. 2016). This suggests that CLas is either a facultative endosymbiont or has a mutualistic association with psyllids. It has been demonstrated that facultative symbiotic bacteria may be harmful or beneficial to an insect vector depending on the sequence of OmpA (Weiss et al. 2008). Moreover, several other functions are attributed to bacterial OmpA, including adhesion and invasion in mammalian cells, recognition by specific host receptors, and biofilm formation (Barrios et al. 2006; Prasadarao 2002; Smith et al. 2007). Nonetheless, OmpA is characterized as an important target in all cases mentioned above for bacterial control (Smith et al. 2007). However, no previous approach has targeted the OmpA of CLas aiming at disrupting its interaction with the insect vector. We have specifically chosen OmpA as target protein in this study because of its presence in CLas, which has competitively excluded the non–OmpA-encoding CLam from Brazilian citrus orchards (Lopes et al. 2009a).

    We generated different peptides from the OmpA sequence of CLas strain Psy62 (Duan et al. 2009) and administered them to ACPs via an artificial diet system and in planta. We hypothesize that these OmpA-derived peptides can interact with specific receptors and/or other structures present in the insect vector and thus prevent their interaction with CLas by blocking recognition/adhesion sites. Treatment with most of the generated peptides significantly reduced or even abrogated the acquisition of the bacterium by psyllids. Our results expand the possibility of identifying new molecules to interfere with the acquisition of CLas by psyllids.

    MATERIALS AND METHODS

    OmpA sequence and sequence and synthesis of peptides.

    The sequence of OmpA from CLas strain Psy62 (locus tag CLIBASIA_04260) was retrieved from NCBI (https://www.ncbi.nlm.nih.gov), and its exclusivity in CLas in comparison with CLam was confirmed using BLAST (Johnson et al. 2008). The alignment of OmpA within Liberibacter species was downloaded from NCBI, processed using T-Coffee Multiple Sequence Alignment Server (Di Tommaso et al. 2011), and finally visualized using the BoxShade webserver. The OmpA sequence was analyzed using the Protean 3D application of the DNASTAR Lasergene software (DNASTAR, Inc., Madison, WI) to identify its hydrophilic and antigenic regions and determine epitopes from which the peptides used in this study were synthesized. After identifying these regions, the peptide sequences of OmpA (named Pep1, Pep2, Pep3, Pep4, Pep5, and Pep6) were sent to the company Aminotech Pesquisa e Desenvolvimento Ltda (Diadema, SP, Brazil) for microscale synthesis.

    D. citri rearing.

    CLas-uninfected ACPs were reared in Murraya paniculata plants, as described elsewhere (de Souza Pacheco et al. 2020). Briefly, psyllids were reared in mesh cages containing M. paniculata at 24 ± 3°C under a 14:10 h (light:dark) photoperiod and 35 to 70% RH. Psyllid populations were maintained at the Centro APTA Citros Sylvio Moreira.

    Delivery of OmpA peptides to ACPs using an artificial diet system.

    The peptides of OmpA were delivered to psyllids using an artificial diet system (Supplementary Fig. S1) similar to that described elsewhere (Galdeano et al. 2017). In summary, 10-day-old adult ACPs were maintained in small Petri dishes (60 × 15 mm; Corning Glass Works, Corning, NY) containing a sachet with 1 ml of artificial diet (30% of sucrose with 0.1% green and 0.4% yellow food dyes, pH 7.4, sterilized by autoclaving) (McCormick & Company, Inc., Hunt Valley, MD) and the correspondent concentration of each peptide (diluted in sterile deionized water). Mock controls without peptide addition were also included. The artificial diet was evenly distributed between two parafilm layers, and the final cage was incubated at 25 ± 2°C under a 14:10 h (light/dark) photoperiod and 35 to 70% RH, with Petri dishes being placed ∼1 m away from the light source. Insects were allowed to feed for 7 days for peptide ingestion. Each cage contained 10 insects per replicate, whereas each experiment consisted of four replicates for peptide ingestion and two replicates for mock control.

    Assessment of ingestion and probing by ACPs and detection of peptides in ACPs and in planta.

    Experiments were conducted similarly as described above for peptide ingestion via the artificial diet system to assess ingestion and probing behaviors of ACPs. However, filter papers were cut to the size of the Petri dishes and placed on the bottom, and 30 adult ACPs were used per cage. After the feeding period of 7 days, the bottom parafilm membrane from each cage was stained with 0.1% safranin for 2 h to observe salivary sheaths from psyllids, which suggests probing. Salivary sheaths were visualized under a light microscope using a 100× objective (Olympus, Tokyo, Japan). In addition, filter papers from each cage were treated with 2% ninhydrin to stain honeydew droplets from ACPs, which suggests ingestion (Ammar et al. 2013). After staining, honeydew droplets are visible to the naked eye.

    Once ingestion of the artificial diet was confirmed, the ingestion of peptides together with the artificial diet by ACPs was assessed through immunoblotting to detect peptides in the honeydew droplets. As proof of concept, experiments were conducted only with the peptide 6 (Pep6; at the final concentration of 50 µM in artificial diet) because it showed a greater ability to impair CLas acquisition by treated insects, as demonstrated below. Thus, experiments were performed as described for ingestion and probing behaviors; however, a Hybond-C nitrocellulose membrane (GE Healthcare, Buckinghamshire, UK) was used instead of filter paper. The antibody used in this assay was obtained by sending the synthesized Pep6 to a third-party company (Célula B, Porto Alegre, RS, Brazil), which generated the polyclonal anti-Pep6 through rabbit immunization. After the feeding period in the artificial diet to ingest the peptide, the nitrocellulose membrane containing the honeydew was recovered and blocked overnight under agitation in tris-buffered saline (TBS) + 0.05% Tween-20 + 5% BSA. Then, the membrane was washed three times for 5 min each with TBS-T (TBS + 0.05% Tween-20 + 0.1% BSA), treated with anti-Pep6 (1:5,000 dilution in TBS-T) for 1 h, washed three times with TBS-T, treated with the goat antirabbit secondary antibody coupled to horseradish peroxidase (Promega Corporation, Madison, WI) at a 1:10,000 dilution in TBS-T for 1 h, and lastly washed again three times with TBS-T (the first wash was for 15 min, with two subsequent 5-min washes). Finally, the membrane was developed using the Amersham ECL Western Blotting Detection Reagent (GE Healthcare, Buckinghamshire, UK) under agitation for 1 min and exposed for 40 min in the ChemiDoc XRS+ System (Bio-Rad Laboratories, Inc., Hercules, CA) for visualization.

    On the other hand, the detection of Pep6 in planta was performed by spraying select branches of sweet orange ‘Pera’ plants with 2 ml of solution containing 50 µM of Pep6. Solutions were composed of either 50 µM of Pep6 in 0.05% Assist adjuvant (mineral oil solution; BASF, Ludwigshafen, Germany) diluted in sterile distilled water or 50 µM of Pep6 in a solution of micronutrients (H3BO3 2.86 g/liter, MnCl2·4H2O 1.81 g/liter, ZnSO4·7H2O 0.21 g/liter, CuSO4·5H2O 0.75 g/liter, Na2MO4·2H2O 0.27 g/liter). Mock controls consisting of 0.05% Assist adjuvant, or the micronutrients solution, both amended with a volume of water corresponding to the volume of Pep6 added to each solution for a final concentration of 50 µM, were also included. Lyophilized Pep6 was always diluted fresh before using. Applications of Pep6 were performed weekly for a total period of 6 months, while leaf samples for Pep6 detection in planta after absorption were taken after 4 and 6 months of the initial application. Additional leaf samples from each treatment were also taken after 3 months of the final Pep6 application to assess whether this peptide was stable in planta for a longer period after the final application. The detection of Pep6 within collected leaf samples was conducted by extracting total proteins and detecting Pep6 through immunoblotting. Thus, for total protein extraction, leaves were washed to remove surface residues and separately ground in liquid nitrogen using sterile mortar and pestle, and 200 mg of each sample was transferred to a 1.5-ml microcentrifuge tube together with 500 µl of extraction buffer (50 mM Tris-HCl pH 6.8, 1% vol/vol 2-mercaptoethanol, 0.2% wt/vol PVP-40, 2 mM PMSF protease inhibitor; filtered in a 0.45-µm syringe filter and stored at –20°C). Then, samples were vortexed at 4°C for 1 h and centrifuged at 10,000 rpm and 4°C for 30 min. After, the supernatant was transferred to a new 1.5-ml microcentrifuge tube and centrifuged once again at the same conditions. Next, the supernatant was transferred to an Ultracel 3K centrifugal filter unit (Merck Millipore Ltd., Carrigtwohill, Ireland) used to concentrate proteins and centrifuged at 14,000 × g and 4°C for 3 h. A 100-µl aliquot of each sample was then transferred to a Hybond-C nitrocellulose membrane (GE Healthcare, Buckinghamshire, UK) using the Minifold I–SRC 96 (Schleicher & Schuell BioScience, Dassel, Germany) dot blot apparatus, and Pep6 was finally detected through immunoblotting as described above for honeydew samples. The concentration of Pep6 in each leaf sample was determined by quantifying the intensity of the signal of each sample (calculated as pixel area) using the Image Lab software (Bio-Rad Laboratories, Inc.) and comparing it to the intensity of the signal of a standard curve containing known concentrations of Pep6 ranging from 0.25 µM to 1 µM (Supplementary Fig. S2). In addition, the stability of Pep6 after solubilization was assessed in silico by calculating its parameters using the Peptide property calculator by Innovagen (Lund, Sweden; https://pepcalc.com).

    A surrogate assay using sterile distilled water was also developed to investigate whether citrus leaves could absorb a solution containing a peptide of OmpA. In this assay, the tip of the petiole of detached young leaves of sweet orange Pera was immersed in 2 ml of sterile distilled water or in sterile distilled water containing 0.1% red food dye (Arcólor, São Paulo, SP, Brazil) and 50 µM of Pep6 for 48 h (period needed for total solution absorption). Leaves were then visualized under a stereomicroscope to observe the red staining of leaf veins, which indicates solution absorption, presumably together with absorption of Pep6. During the incubation period, samples were maintained in 50-ml conical tubes (Kasvi, São José dos Pinhais, PR, Brazil) at 25 ± 1°C under a 16:8 h (light:dark) photoperiod.

    Acquisition of CLas from infected plants.

    To assess the effect that the treatment of ACPs with the OmpA peptides may have on CLas acquisition, adult psyllids treated with peptides using the artificial diet system (described above), as well as mock control insects, were transferred to CLas-infected sweet orange Pera trees (Citrus sinensis (L.) Osbeck) maintained under greenhouse conditions for temperature, humidity, and light and allowed to feed for 10 days. We define acquisition as the ability of CLas-uninfected ACPs to ingest CLas cells and turn into CLas-positive insects (as assessed by quantitative PCR [qPCR]) after feeding on HLB-symptomatic plants. In this assay, ACPs (peptide-treated and mock controls) were recovered from their cages (excluding insects that died during the feeding period) and put in HLB-symptomatic branches of CLas-infected sweet orange plants within mesh-covered cages (Supplementary Fig. S3). Groups of ACPs treated with different peptides as well as mock controls were placed in separate, individual branches, and chosen branches always contained new flushes to allow ACP feeding (Grafton-Cardwell et al. 2013). After the 10-day acquisition period, used branches were entirely cut using a surface-sterilized pruning shear, and the CLas population was quantified within each insect and in a petiole sample (positive control for CLas presence) of each sweet orange branch by qPCR (detailed below). Mock controls were treated following the same protocol for peptide-treated psyllids, and mock control insects were included in every replicate of CLas acquisition assays performed in this study.

    Treatment of ACPs with Pep6 in planta and acquisition of CLas.

    To assess whether adult ACPs could ingest Pep6 while feeding in planta and its effect on CLas acquisition, psyllids were transferred to cages (containing 30 insects each) in healthy sweet orange Pera plants as described above for CLas acquisition and sprayed with 2 ml of solution (per cage) containing 50 µM of Pep6 (as described above for Pep6 detection by absorption in planta). Solutions were composed of either 50 µM of Pep6 in 0.05% Assist adjuvant (BASF) diluted in sterile distilled water or 50 µM of Pep6 in a solution of micronutrients. Mock controls consisting of 0.05% Assist adjuvant, or the micronutrients solution, both amended with a volume of water corresponding to the volume of Pep6 added to each solution for a final concentration of 50 µM, were also included. Lyophilized Pep6 was always diluted fresh before each application. After 24 h of treatment with Pep6 (applied by spraying), ACPs were transferred to branches of HLB-symptomatic sweet orange Pera plants and allowed to feed for another 7 days to acquire CLas as previously described. Then, the CLas population was quantified within each insect by qPCR. Branches of HLB-symptomatic plants used in these assays were previously screened by qPCR to assess their CLas population. This was performed so as to only use branches in CLas acquisition assays that had similar CLas population (1.17 × 105 CLas genome equivalents ml–1 ± 2.77 × 104; standard error).

    DNA extraction and qPCR analyses.

    The DNA of ACPs was extracted from individual insects. In this process, each psyllid was placed separately into a 1.5-ml microcentrifuge tube together with a 5-mm chrome steel bead (QIAGEN, Hombrechtikon, Switzerland) and ground in a beadbeater (TissueLyzer II; QIAGEN, Hilden, Germany) at 15 Hz/s for 30 s. Then, 100 µl of STE solution (10 mM Tris-HCl, 1 mM EDTA pH 8.0, and 25 mM NaCl) was added together with 15 µl of proteinase K 200 µg/ml. Samples were homogenized by vortexing and incubated at 56°C for 30 min. After that step, 150 µl of Nuclei Lysis solution (Promega Corp.) was added and samples were again vortexed and incubated at 80°C for 5 min. Next, samples were cooled to room temperature, and 60 µl of Protein Precipitation solution (Promega Corp.) was added. Samples were vortexed, incubated on ice for 5 min, and centrifuged at 13,000 × g for 5 min. The supernatant (200 µl) was then transferred to a new 1.5-ml microcentrifuge tube together with 1 volume of 100% isopropanol, vortexed, and precipitated at –20°C for at least 2 days. At last, samples were centrifuged at 14,000 × g for 8 min, the supernatant was discarded, and the DNA was suspended in 50 µl of TE + RNase 1/10. On the other hand, the DNA of sweet orange plants was extracted from the petiole of HLB-symptomatic leaves using the cetyltrimethylammonium bromide method (Murray and Thompson 1980).

    Quantification of CLas in each sample was determined as genome equivalents by qPCR. Reactions were carried out using the Las-I-F/Las-I-R/Las-P pair of primers and TaqMan probe set (Lin et al. 2010). Amplifications to detect and quantify CLas were performed in conjunction with primers and probe set targeting the 18S small ribosomal subunit of eukaryotes (Meyer et al. 2007) used as internal control for DNA extraction. DNA amplifications were performed in 13.5-µl reactions containing 1× HOT FIREPol Probe qPCR Mix Plus (ROX) (Solis BioDyne, Tartu, Estonia), 400 nM of each CLas primer, 200 nM of the CLas probe (labeled 5′-6FAM, 3′-BHQ1), 0.2 µl of the 18S primers and probe set, and 3 µl of DNA template and carried out using an ABI Prism 7500 Sequence Detection System (Applied Biosystems, Warrington, UK). Each DNA amplification was performed in duplicate, and positive (DNA from CLas-infected sweet orange plants) and negative (DNA from CLas-uninfected ACPs) controls as well as blank reactions (performed without any DNA template) were included. Standard cycling parameters from the equipment were used. The number of CLas genome equivalents in each sample was calculated by coamplifying the correspondent seven-point standard curve made from 10-fold serial dilutions of the pGEMT-Easy plasmid (Promega Corp.) containing the partial sequence of the target CLas gene amplified by the selected pair of primers and probe set (Lin et al. 2010). Accordingly, we considered CLas-positive insects those with a value of threshold cycle <36, which corresponds to at least six genome equivalents of CLas, as determined by the standard curve (Supplementary Fig. S4).

    Data analysis.

    The acquisition of CLas by adult ACPs (total population of CLas by psyllid) in different peptide treatments in comparison with control treatments was analyzed by the Mann-Whitney U nonparametric test using Jamovi software version 1.6.23 (https://www.jamovi.org). Before performing these analyses, the normality of data was assessed by the Shapiro-Wilk test, also using the Jamovi software. On the other hand, the original data (x) of the percentage of infected insects after treatment with Pep6 and feeding in HLB-symptomatic plants was submitted to a transformation that combines the arcsine and square root functions (y = arcsin(√x/100)). Then, the normality of data was assessed by the Shapiro-Wilk test, and the statistical significance was finally analyzed in comparison with control treatments using the two-tailed Student t test. Significant differences among immunoblotting fluorescence intensity to detect Pep6 as well as the calculated Pep6 concentration within plant samples were determined by one-way analysis of variance followed by Tukey’s honestly significant difference multiple comparisons of means using SigmaPlot software version 11.0 (Systat Software Inc., San Jose, CA). Because of the variable distribution of CLas in planta (Li et al. 2009), which makes it difficult to standardize assays, results from some experiments are shown as representative results from independent experiments. Raw data regarding CLas acquisition by ACPs, following treatment with the different OmpA peptides from each independent replicate, are shown in Supplementary Tables S1 and S2.

    RESULTS

    Analysis of OmpA among Liberibacter spp. and determining the peptide sequences of OmpA.

    To confirm the exclusivity of OmpA in CLas in comparison with CLam, the sequence of this protein from CLas strain Psy62 (CLIBASIA_04260) was compared with other Liberibacter spp. using BLAST, processed through T-Coffee, and then visualized using BoxShade (Fig. 1). It was verified that OmpA is indeed absent in CLam (Fig. 1A), whereas different similarity scores were observed when comparing this protein among Liberibacter spp. (Fig. 1B). As expected, the highest level of similarity was obtained among CLas strains, in which the similarity of OmpA varied from 98.95 to 100%, followed by ‘Ca. Liberibacter africanus’ strain PTSAPSY, which is associated with HLB in Africa (Bové 2006), with a similarity of 61.86%. Next, ‘Ca. Liberibacter solanacearum’ strains that are associated with potato zebra chip (Secor et al. 2009) presented similarities varying from 54.17 to 55.61%, whereas ‘Ca. Liberibacter ctenarytainae’, which is associated with the New Zealand native fuchsia psyllid Ctenarytainia fuchsiae Maskell (Vereijssen et al. 2018), presented a similarity of 42.08%. At last, L. crescens, the only culturable species of the genus Liberibacter (Fagen et al. 2014), showed a similarity of 41.96% (Fig. 1B).

    Fig. 1.

    Fig. 1. Comparison of OmpA sequence within Liberibacter species. A, The alignment of OmpA sequences within Liberibacter species obtained from BLAST in NCBI was processed through T-Coffee and visualized using BoxShade. Different sequences of OmpA are denoted by representative strains within each species. The OmpA sequence from ‘Candidatus Liberibacter asiaticus’ strain Psy62 (locus tag CLIBASIA_04260) was used as query sequence in BLAST. B, Percent identity of OmpA sequences among Liberibacter species in comparison with reference CLas strain Psy62.

    Download as PowerPoint

    After verifying the absence of OmpA in CLam, the sequence of this protein was analyzed using the Protean 3D application of the DNASTAR Lasergene software to identify potential hydrophilic and hydrophobic regions with antigenic properties that could interact with ACPs and presumably block the interaction of these insects with CLas by competition with recognition sites. Two regions comprising the entire antigenic regions of OmpA were predicted within this protein (Fig. 2A). To obtain precise results about which amino acids from the antigenic regions of OmpA could be effectively interacting with cellular receptors in the insect vector, six different small peptides (ranging from 15 to 18 amino acids each) that cover the entire length of these two regions were designed (Fig. 2B; the sequence of each peptide is shown in the figure).

    Fig. 2.

    Fig. 2. Analysis of OmpA sequence from ‘Candidatus Liberibacter asiaticus’ (CLas) using the DNASTAR Lasergene software. A, The OmpA sequence of CLas strain Psy62 was analyzed using the Protean 3D application of the DNASTAR Lasergene software to determine its hydrophilic and antigenic regions. These regions are represented by peaks with positive amplitudes in the figure. B, Six different peptides were designed within the two different antigenic regions found within the analyzed OmpA sequence.

    Download as PowerPoint

    Treatment with peptides of OmpA does not change the ingestion and probing behaviors of ACPs in artificial diet.

    Before analyzing whether the treatment with the designed peptides could disrupt the acquisition of CLas by ACPs, we first evaluated if these peptides could change the ingestion and probing behaviors of psyllids. By using the artificial diet system (Fig. 3A1), the presence of salivary sheaths in the bottom parafilm layer was verified for both nonamended (Fig. 3A2) and Pep5- and Pep6-amended artificial diet (Fig. 3A3 and A4). Accordingly, the honeydew droplets produced by ACPs were collected by a filter paper placed on the bottom of the artificial diet cage (Fig. 3B), with no differences among the psyllids that fed on the artificial diet alone (Fig. 3B3) and those that fed on artificial diet containing 50 µM of either Pep5 or Pep6 (Fig. 3B4 and B5). Pep5 and Pep6 were used as representative peptides for these assays because they presented better results to disrupt acquisition of CLas (shown below). Thus, our results qualitatively demonstrate that ACPs were able to normally probe and ingest artificial diet either by itself or amended with Pep5 or Pep6, demonstrating that peptides of OmpA do not interfere with the production of salivary sheath and probing.

    Fig. 3.

    Fig. 3. Asian citrus psyllids (ACPs) feeding and ingestion of peptides within artificial diet sachets. A, Visualization of the salivary sheaths of ACPs within artificial diet sachets. After a feeding period of 7 days in artificial diet, the salivary sheaths from the stylet of psyllids present in the bottom parafilm layer of the sachets were stained with 0.1% safranin and visualized under a light microscope to assess their probing behavior with peptide treatments. A1, ACPs within an artificial diet sachet. The white arrow is pointing to the double layer of parafilm, which contains the artificial diet between both layers. A2 to 4, Salivary sheaths highlighted by circles of psyllids that fed only on artificial diet, artificial diet + Pep5 (50 µM) and artificial diet + Pep6 (50 µM), respectively. B, Visualization of the honeydew droplets of ACPs within artificial diet sachets. After a feeding period of 7 days in artificial diet, the honeydew droplets deposited in the filter paper placed on the bottom of the sachets were stained with 2% ninhydrin and visualized by the naked eye to assess the ingestion behavior of ACPs with peptide treatments. B1, General setting for the artificial diet experiment with ACPs within an artificial diet sachet (the white arrow is pointing to the filter paper on the bottom of the sachet). B2, Filter paper stained with ninhydrin (negative control). B3 to 5, Ninhydrin-stained filter paper showing the honeydew droplets (arrows) of psyllids that fed only on artificial diet, artificial diet + Pep5 (50 µM), and artificial diet + Pep6 (50 µM), respectively. C, Detection of Pep6 within honeydew droplets of ACPs. After a feeding period of 7 days, the nitrocellulose membrane containing honeydew droplets of psyllids that fed on artificial diet + Pep6 (50 µM) was submitted to immunoblotting with anti-Pep6 to detect this peptide. C1, Nitrocellulose membrane without honeydew droplets (negative control). C2 and 3, Nitrocellulose membrane containing the honeydew droplets of psyllids that fed only on artificial diet and artificial diet + Pep6 (50 µM), respectively. Circles in C highlight areas of the membrane in which Pep6 was detected.

    Download as PowerPoint

    Pep6 is detected within honeydew droplets of ACPs.

    Because psyllids treated with OmpA peptides were able to probe and ingest from the artificial diet, ingestion of the Pep6 peptide within the artificial diet by ACPs was then assessed. In this analysis, the filter paper placed on the bottom of the artificial diet cage to collect honeydew droplets was replaced by a nitrocellulose membrane, which allowed detection of peptides within the honeydew through immunoblotting (Fig. 3C). Presence of peptides in the excrement of ACPs indicates that insects could ingest peptides together with the artificial diet. As proof of concept, only Pep6 was used in this assay. The nitrocellulose membrane containing the honeydew of adult psyllids that ingested the artificial diet with Pep6 (50 µM) presented clear fluorescent signals after developing the immunoblot to detect this peptide (Fig. 3C3). On the other hand, the control membrane (Fig. 3C1) and the membrane collected from adult psyllids that fed only the artificial diet (Fig. 3C2) presented similar basal signal levels of fluorescence, which corresponded to background noise. Therefore, based on the qualitative analysis of membranes, we can infer that ACPs ingested and subsequently expelled the peptide contained in the artificial diet.

    Treatment with OmpA-derived peptides significantly reduces the acquisition of CLas by ACPs.

    We evaluated the ability of the OmpA-derived peptides to disrupt the acquisition of CLas from infected plants by adult ACPs. Adult psyllids feeding on artificial diet containing 5 µM of each tested peptide (Pep1, Pep2, Pep3, Pep4, Pep5, and Pep6, separately) for 7 days were transferred to branches of CLas-infected sweet orange Pera trees for a 10-day CLas acquisition access period. Treatments with Pep1, Pep3, Pep5, and Pep6 significantly reduced the population of CLas acquired by ACPs, as analyzed by qPCR (Fig. 4). Because the treatment of ACPs with Pep5 and Pep6 showed the highest decrease in CLas acquisition (∼20- and 7-fold less CLas acquisition in comparison with the control, respectively), the concentration of these peptides within the artificial diet was increased to 50 µM and the CLas acquisition assay was repeated. The higher concentration of Pep5 and Pep6 led to a total disruption of CLas acquisition by adult psyllids, with no detection of CLas by qPCR in both treatments (Fig. 4D). These results demonstrate that treatment with OmpA-derived peptides can disrupt the acquisition of CLas by ACPs. However, treatment with these peptides did not significantly alter the proportion of CLas-infected insects following acquisition from infected plants nor change the mortality of ACPs throughout the development of assays (Supplementary Tables S3 and S4, respectively).

    Fig. 4.

    Fig. 4. Acquisition of ‘Candidatus Liberibacter asiaticus’ (CLas) by adult Asian citrus psyllids (ACPs) treated with the peptides of OmpA in artificial diet. Adult ACPs were separately treated with the peptides Pep1, Pep2, Pep3, Pep4, Pep5, and Pep6 (5 µM each) and with Pep5 and Pep6 at a higher concentration (50 µM) using the artificial diet system and then allowed to feed for 10 days on huanglongbing (HLB)-symptomatic sweet orange plants to acquire CLas, which was quantified in each insect by quantitative PCR. A, Acquisition of CLas by adult ACPs treated with Pep2 and Pep4. B, Acquisition of CLas by adult ACPs treated with Pep1 and Pep3. C, Acquisition of CLas by adult ACPs treated with Pep5 and Pep6. D, Acquisition of CLas by adult psyllids treated with a higher concentration of Pep5 and Pep6 in artificial diet. Control, psyllids that fed only on artificial diet, without addition of any peptide. Statistical significance was determined using Mann-Whitney U test (*P ≤ 0.05; **P ≤ 0.005 in comparison with the control; n = 5 to 33 technical replicates in each independent replicate). Because experiments were performed independently at different periods using distinct HLB-symptomatic sweet orange plants and their branches, results are presented separately according to the peptides that were analyzed at the same time in each assay.

    Download as PowerPoint

    Sweet orange trees absorb Pep6 applied by spraying.

    Aiming at analyzing whether in planta treatment with OmpA-derived peptides also disrupts the acquisition of CLas by ACPs, the capacity of sweet orange trees to absorb Pep6 when applied by spraying was first evaluated. Only Pep6 was used in this assay because it was able to abolish acquisition of CLas by ACPs when used at a 50-µM concentration in the artificial diet (Fig. 4D) and because it covers almost the same region of OmpA as Pep5 (Fig. 2). In this assay, we sprayed isolated branches of sweet orange Pera plants with 2 ml of solution containing 50 µM of Pep6, which was diluted in a solution of micronutrients or with 0.05% Assist adjuvant. Applications were performed weekly for 6 months. Samples were collected from sprayed branches after 4 and 6 months of the initial application to detect the peptide by immunoblotting, and after 3 months of the final application to evaluate whether Pep6 was stable in planta. Pep6 was detected in all samples following its application in planta (Fig. 5A to C), as observed by quantifying the fluorescence intensity produced by the immunoblot (pixel area) and comparing it to the mock control. Surprisingly, Pep6 was still detected in planta even after 3 months of the final application (Fig. 5C). The concentration in planta ranged from ∼0.15 to 0.34 µM, and there were no significant differences among samples diluted in micronutrients and Assist adjuvant or between the different analyzed time points (Fig. 5D). Besides, the peptide treatment by itself did not disrupt CLas multiplication in planta (Supplementary Table S5). In addition, a surrogate assay performed with detached young sweet orange leaves to evaluate their ability to absorb Pep6 showed absorption of red food dye diluted in water containing 50 µM of Pep6 as fast as 48 h after application (Supplementary Fig. S5). At last, the in silico analysis of the parameters of Pep6 showed that this peptide lacks amino acids that may suffer changes in their side chains, including tryptophan, cysteine, and methionine, and lead to pathways such as hydrolysis, deamidation, oxidation, and diketopiperazine formation. Thus, Pep6 should remain stable in temperatures as high as 30 to 50°C. Moreover, its isoelectric point of 11.39 is far from the pH of the phloem sap of citrus plants, which is usually ∼6.0 (Hijaz and Killiny 2014), and therefore contributes to its solubility in planta. Together, results show that sweet orange plants absorb the peptide that is applied by spraying, which presents some level of stability because it can still be detected for a long period after the final application. The applied peptide can then be ingested by ACPs while feeding on treated plants.

    Fig. 5.

    Fig. 5. Detection of Pep6 within plant samples treated by spraying. Branches of sweet orange ‘Pera’ plants were treated weekly with Pep6 by spraying with 2 ml of solution containing 50 µM of this peptide. Pep6 was either diluted in a solution of micronutrients or in 0.05% Assist adjuvant. Detection and quantification of Pep6 within samples were performed through immunoblotting. A and B, Fluorescence intensity (calculated as pixel area) of samples collected after 4 and 6 months of Pep6 treatment, respectively. C, Fluorescence intensity (calculated as pixel area) of samples collected 3 months after final Pep6 application. D, Concentration of Pep6 (µM) delivered by the micronutrients solution and Assist adjuvant in samples collected after 4 and 6 months of treatment and after 3 months of final application. Data represent means and standard errors. Different letters indicate significant difference among samples as analyzed by one-way analysis of variance followed by Tukey’s honestly significant difference multiple comparisons of means (P ≤ 0.05; n = 2 to 3 independent replicates). A to C, The fluorescence intensity obtained from the positive control Pep6 at 0.25 µM is included as reference, and the background fluorescence intensity from the mock control is considered as the cutoff value for positive samples.

    Download as PowerPoint

    Treatment with Pep6 in planta significantly reduces the acquisition of CLas by ACPs.

    We assessed whether the treatment of ACPs with Pep6 in planta could disrupt the acquisition of CLas via feeding. ACPs were initially treated for 24 h in planta with a 50-µM solution of Pep6 (diluted in micronutrients or Assist adjuvant) by spraying on healthy sweet orange plants and then transferred to HLB-symptomatic plants to feed for another 7 days and acquire CLas. The treatment in planta with Pep6 diluted in micronutrients significantly reduced both the population of CLas within infected ACPs (∼13-fold less in comparison with the control; Fig. 6A) and the proportion of insects that effectively acquired CLas (∼8-fold less in comparison with the control; Fig. 6B). Similar results were also obtained with Pep6 diluted in Assist adjuvant; however, only one independent replicate was performed (Supplementary Fig. S6). Collectively, sweet orange plants were able to absorb Pep6 (Fig. 5), and the treatment in planta disrupted the acquisition of CLas by ACPs (Fig. 6).

    Fig. 6.

    Fig. 6. Acquisition of ‘Candidatus Liberibacter asiaticus’ (CLas) by adult Asian citrus psyllids (ACPs) treated with Pep6 in a solution of micronutrients in planta. Adult ACPs were treated for 24 h with 50 µM of Pep6 (in a solution of micronutrients) in planta by spraying in healthy sweet orange ‘Pera’ plants and then transferred and allowed to feed and acquire CLas from huanglongbing (HLB)-symptomatic sweet orange plants for 7 days. The CLas population within each insect was assessed by quantitative PCR. A, Mean value of CLas population by adult ACP treated with Pep6 in planta, as calculated by genome equivalents. B, Percentage of CLas-infected ACPs after treatment with Pep6 in planta and feeding in HLB-symptomatic plants. Control, psyllids treated only with micronutrients amended with a volume of water corresponding to the used volume of Pep6 but without addition of any peptide. Original data of percentage of CLas-infected insects (treatments; y-axis) was submitted to a transformation that combines the arcsine and square root functions. The Shapiro-Wilk test (W) was then applied to the transformed data, and it supported the assumption of normal distribution (W = 0.87; P value = 0.22). Statistical significance was determined using the A, Mann-Whitney U test and B, Student’s t test. (In both A and B, ** P ≤ 0.001 in comparison with the control; n = 3 independent replicates with an average of 13 insects in each replicate).

    Download as PowerPoint

    DISCUSSION

    Strategies to control vector-borne plant pathogens conventionally target the vector or the pathogen although they seldomly aim at disrupting the molecular interactions among these organisms (Killiny et al. 2012). The development of approaches to disrupt such interactions may, however, reduce disease spread by interfering with key factors required for successful acquisition and then transmission of pathogens. Insect-transmitted plant pathogens may either penetrate and colonize different tissues and organs of their vectors (circulative pathogens) (Ammar et al. 2011, 2019; Nault 1997) or colonize the insect without host internalization by attaching to sections of the foregut (noncirculative pathogens) (Chatterjee et al. 2008; Nault 1997). Nevertheless, molecular interactions among the pathogen and its vector determine the successful colonization and, therefore, pathogen acquisition and transmission in both conditions (Froissart et al. 2002; Hogenhout et al. 2008). Thus, the circulative nature of CLas colonization within ACPs (Ammar et al. 2011, 2019) offers an opportunity to interfere with molecular components of the CLas–psyllid interaction to disrupt the acquisition of this pathogen.

    In bacteria, the outer membrane protein OmpA has been demonstrated to perform many functions and has been proposed as a molecular target for bacterial control (Chen et al. 2021; Smith et al. 2007). Therefore, its presence in the more prevalent species CLas in field conditions suggests that OmpA could be a key target to disrupt its acquisition by ACPs (Lopes et al. 2009a, b; Wulff et al. 2014). Furthermore, recent studies also recently indicated that strategies to neutralize OMP could control HLB (Chen et al. 2021). Therefore, in the present study, we assessed how different peptides generated from predicted antigenic regions of OmpA from CLas strain Psy62 could disrupt the acquisition of this plant pathogen via psyllids from infected sweet orange plants. Treatment of ACPs through feeding using an artificial diet system demonstrated that Pep1, Pep3, Pep5, and Pep6 significantly reduced the acquisition of CLas from HLB-symptomatic trees using as little as 5 µM of each peptide in the artificial diet. On the other hand, increasing the concentration of Pep5 and Pep6 to 50 µM completely abolished the acquisition of CLas. Moreover, in planta treatment with 50 µM of Pep6 significantly decreased the acquisition of CLas by psyllids by as much as 13-fold less, whereas sweet orange plants stably absorbed and maintained this peptide for as long as 3 months after the final application. However, the Pep6 concentration in planta was low, and there were no significant differences among the concentrations of this peptide between the different analyzed time points. This either indicates that Pep6 may get saturated in planta or that the absorption was nonoptimal. In addition, we cannot discard the option that part of the applied peptide can be degraded after application. Furthermore, the ability of the peptide to move systemically within treated plants remains elusive. Nevertheless, the in planta treatment of ACPs with this peptide reduced CLas acquisition. In addition, treatment with Pep5 and Pep6 did not change the ingestion and probing behaviors of psyllids when assessing production of honeydew droplets and presence of salivary sheaths, whereas immunoblotting assays performed with the honeydew showed detection of Pep6, indicating the ingestion of peptides by ACPs during feeding. However, although OmpA-derived peptides showed a great potential to disrupt the acquisition of CLas by psyllids, the molecular basis of the CLas-ACP interaction and processes that are being disturbed by these peptides are still largely unknown.

    One successful example of disrupting the acquisition and transmission of a vector-borne bacterial plant pathogen is using lectins, carbohydrates, antibodies, and peptides to reduce the transmission of Xylella fastidiosa by the leafhopper Graphocephala atropunctata (Killiny et al. 2012; Labroussaa et al. 2016). This bacterium is noncirculative within insect vectors and binds to carbohydrates in the cuticular lining of the foregut of insect hosts through afimbrial adhesins (Killiny and Almeida 2009a, b). Thus, molecules that could compete for binding to the surface of the insect or that could saturate the adhesins of X. fastidiosa successfully decreased acquisition and thus transmission of this pathogen (Killiny et al. 2012; Labroussaa et al. 2016). Similar approaches have been used to disrupt transmission of circulative plant pathogenic viruses. For instance, both the expression of recombinant spike protein from the Rice ragged stunt oryzavirus by transgenic rice lines and feeding of the insect vector Nilaparvata lugens (rice brown planthopper) with this purified protein resulted in disruption of Rice ragged stunt oryzavirus transmission (Chaogang et al. 2003; Guoying et al. 1999). In addition, the use of peptides that bind to the gut cells of insect vectors to interfere with molecular interactions required for the successful establishment and transmission of pathogenic viruses can also be explored. In the Tomato spotted wilt virus, a soluble form of its envelope glycoprotein binds to the thrips vector (Frankliniella occidentalis) midgut and reduces the amount of virus load within insects and transmission to healthy plants (Whitfield et al. 2004, 2008). Expression of this protein by transgenic tomato lines also disrupts Tomato spotted wilt virus transmission (Montero-Astúa et al. 2014). On the other hand, it has been demonstrated that the use of a peptide that binds to the midgut and hindgut of the pea aphid (Acyrthosiphon pisum) reduces the amount of Pea enation mosaic virus in the insect’s hemocoel, which is required for its transmission, and thus likely disrupts this process (Liu et al. 2010).

    Molecules that bind to the gut of insect hosts and disrupt interactions with bacterial colonizers are interesting to our study because (i) CLas has been shown to accumulate, and likely replicate, within endoplasmic reticulum–associated vacuoles in the gut cells of D. citri (Ghanim et al. 2017) and (ii) OmpA has been shown to bind and mediate invasion of different cells of bacterial hosts (Prasadarao 1996, 2002; Smith et al. 2007), including the gut of insect hosts (Hegde et al. 2019; Maltz et al. 2012). Therefore, we hypothesize that OmpA may be one of the factors responsible for mediating the colonization of the gut cells of ACPs by CLas. For instance, OmpA is essential for the survival of the commensal bacterium Sodalis glossinidius within the tsetse fly gut because the absence of biofilm formation in ΔOmpA mutant cells of this bacterium leads to their exposure and elimination by the innate immune system of the insect (Maltz et al. 2012). Similar results have also been observed for the symbiotic bacterium Cedecea neteri colonizing the gut of Aedes aegypti mosquitoes (Hegde et al. 2019). Both studies add to the role of bacterial OmpA within the gut of insect hosts and illustrate how this protein may govern successful colonization. However, the molecular basis of the CLas-ACP interaction mediated by OmpA and whether this protein plays a role in the colonization of the guts of ACPs by CLas are still unknown. Nevertheless, the apparent functional role of the bacterial OmpA for colonization of hosts and its absence in CLam suggest that different receptors may be used by CLas and CLam to colonize the common insect vector D. citri. This opens the possibility of exploring different and new targets within the CLas– and CLam–ACP interactions to disrupt the acquisition of these pathogens by the vector.

    In summary, our results reveal the potential use of OmpA-derived peptides to disrupt the life cycle of CLas, which includes colonization of both ACPs and citrus plants, by blocking the acquisition of this bacterium from infected plants. In addition, in planta treatments suggest a potential use of peptides in field applications to control CLas (at least for Pep6). Moreover, we have investigated disruption of CLas acquisition only in adult ACPs, although nymphs are more efficient in acquiring CLas from infected plants (Ammar et al. 2016; George et al. 2018; Pelz-Stelinski et al. 2010). This was done because adults are the most important life stage of D. citri for the epidemiology process of HLB because they present a higher mobility in the environment, have a longer lifespan than nymphs, and are more efficient in CLas inoculation (Ammar et al. 2020; Hall et al. 2013; Liu and Tsai 2000). In any case, it would be important to assess the ability of the OmpA peptides described here to also disrupt the acquisition of CLas by D. citri nymphs. Nevertheless, the use of small OmpA peptides is a promising novel biotechnological tool to control CLas.

    ACKNOWLEDGMENTS

    We thank Greice Erler from R&D Latin America Insecticides at BASF for all assistance on insect experiments; Franz Josef Braun from Crop Advanced Testing at BASF for direction on the project development; and Gabriel C. Blain from the Ecophysiology and Biophysics Center at the Agronomic Institute (IAC) for statistical analysis assistance. A. A. de Souza and H. Della Coletta-Filho received CNPq fellowships.

    The author(s) declare no conflict of interest.

    LITERATURE CITED

    First and second authors contributed equally to this work.

    Current address of M.V. Merfa: Department of Entomology and Plant Pathology, Auburn University, Auburn, AL 36849, U.S.A.

    Current address of C. S. Francisco: Environmental Genomics Group, Christian-Albrechts University Kiel, Kiel, Germany; and Environmental Genomics Group, Max Planck Institute for Evolutionary Biology, Plön, Germany.

    Funding: This work was supported by research grant from Fundação de Amparo à Pesquisa do Estado de São Paulo (grant 2013/10957-0) and Top Science Program from BASF.

    The author(s) declare no conflict of interest.