Oral Ingestion of Bacterially Expressed dsRNA Can Silence Genes and Cause Mortality in a Highly Invasive, Tree-Killing Pest, the Emerald Ash Borer

RNA interference (RNAi) is a naturally occurring process inhibiting gene expression, and recent advances in our understanding of the mechanism have allowed its development as a tool against insect pests. A major challenge for deployment in the field is the development of convenient and efficient methods for production of double stranded RNA (dsRNA). We assessed the potential for deploying bacterially produced dsRNA as a bio-pesticide against an invasive forest pest, the emerald ash borer (EAB). EAB feeds on the cambial tissue of ash trees (Fraxinus spp.), causing rapid death. EAB has killed millions of trees in North America since its discovery in 2002, prompting the need for innovative management strategies. In our study, bacterial expression and synthesis of dsRNA were performed with E. coli strain HT115 using the L4440 expression vector. EAB-specific dsRNAs (shi and hsp) over-expressed in E. coli were toxic to neonate EAB after oral administration, successfully triggering gene silencing and subsequent mortality; however, a non-specific dsRNA control was not included. Our results suggest that ingestion of transformed E. coli expressing dsRNAs can induce an RNAi response in EAB. To our knowledge, this is the first example of an effective RNAi response induced by feeding dsRNA-expressing bacteria in a forest pest.


Introduction
RNA interference (RNAi) regulates gene expression at the post-transcriptional level by degrading specific messenger RNAs (mRNA), thus blocking translational efficiency [1]. RNAi using exogenous dsRNA is emerging as a novel means of pest suppression [2]. After introduction into cells, dsRNA is recognized by the RNase III enzyme dicer and processed into small interfering RNAs (siRNAs). These siRNAs then bind to the Argonaute protein and form an RNA-induced silencing complex (RISC), and the RISC complex binds to the complementary mRNA molecule, thus blocking gene expression [3].
Coleopteran insects are known to exhibit robust RNAi responses [2,4,5]. RNAi efficiency varies between insect species, insect life stages, target genes, and modes of dsRNA delivery [6]; dsRNA can be delivered in several ways, including by injection, orally, and through absorption [7]. While RNAi is emerging as an attractive option for insect pest control, convenient and efficient methods to produce and deliver dsRNA to target insects is challenging.
The emerald ash borer (EAB), Agrilus planipennis Fairmaire, is an exotic beetle that was accidentally introduced from China into North America in the mid-to late 1990s [8]. Adult beetles feed on ash, Fraxinus spp., and foliage and cause little damage, but larvae feed on cambial tissue beneath the bark,

Target Gene Selection, Total RNA Extraction, cDNA Synthesis, PCR Amplification, and Construction of Recombinant L4440 Vector
To assess the insecticidal activity of bacterially-expressed dsRNA, candidate genes shibire (shi) and heat shock protein-70kDA (hsp) were chosen due to their effectiveness in RNAi-induced EAB mortality by in vitro produced dsRNA [15]. Total RNA was extracted from EAB larvae using Trizol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions; RNA concentration and purity were evaluated using a Nano Drop 1000 (Thermo Fisher, USA). cDNA was synthesized from 1 µg of total RNA using a M-MLV reverse transcriptase kit (Thermo Fisher, USA). Target sequences of shi and hsp were amplified using gene-specific primers with restriction enzymes (Table 1). PCR conditions were as follows: 94 • C for 5 min, followed by 35 cycles of initial denaturation at 94 • C for 30 s, annealing at 60 • C for 30 s and extension at 67 • C for 1 min, and finishing with extension at 67 • C for 8 min. The L4440 plasmid (Addgene plasmid 1654) comprising two T7 promoters in an inverted position flanking multiple cloning sites was used to clone target genes. Restriction (Xba I and XmaI) digested Insects 2020, 11, 440 3 of 10 amplicons were ligated into the Xba I and XmaI digested L4440 vector, respectively. The recombinant vectors were validated by colony PCR [16] and restriction digestion (Xba I and XmaI).

Bacterial Transformation and Expression of dsRNA
The RNase III-deficient E. coli strain HT115 (DE3), obtained through the CGC at the University of Minnesota, was grown in Luria broth (LB) medium with ampicillin (100 µg/mL) and tetracycline (10 µg/mL). The recombinant L4440 vector was transformed into E. coli HT115 (DE3) competent cells. Single bacterial colonies were cultured in LB broth maintained on a shaker incubator at 37 • C (225 rpm) overnight. Cultured broth was then transferred to 50 mL fresh broth medium containing 100 µg/mL ampicillin and cultured at 37 • C until colony growth reached the late exponential phase, with OD 600 = 0.4-0.6. Expression of T7 RNA polymerase was induced by adding a final concentration of 0.5 mM of isopropyl-β-D-1-thiogalactopyranoside (IPTG). Bacteria with dsSHI and dsHSP were then incubated at 37 • C, 30 • C, and 25 • C for up to 4 h to evaluate dsRNA expression. Based on this optimization, further experiments were conducted at 37 • C for dsSHI and 30 • C for dsHSP. Total bacterial RNA was extracted with Trizol reagent and the presence of synthesized dsRNA was confirmed by electrophoresis using a 1% agarose gel.

Biological Activity of Recombinant Bacteria Expressing dsRNA
Laboratory-reared EAB eggs were placed in petri dishes with moistened filter paper and maintained at 24 ± 1 • C in an incubator ( Figure 1). Newly hatched neonates were used in all bioassays. To determine the biological activity of the recombinant bacteria expressing dsRNA, neonate EAB larvae were fed using a modified droplet feeding bioassay [5], where 1 ml of bacterial culture was centrifuged at 3000 rpm for 15 min and the pellet was dissolved in 100 µL of 1% sucrose solution with green tracking dye (Kroger, Co., USA). For each assay, 3µL of bacterial suspension were fed to individual neonate larvae using the droplet assay. Cellular density of the bacterial culture was determined by considering that an optical density of 1 at 600 nm corresponds to 10 8 bacterial cells/mL [17]. As the control, HT115 (DE3) bacteria were used as a treatment [18]. Neonate EAB larvae were fed dsRNA-expressing bacteria for five consecutive days. On day 6, larvae were fed with 1% sucrose lacking dsRNAs for two days. Assays were maintained in an incubator at 26 ± 1 • C, under a 14:10 (L:D) photoperiod. Mortality was measured on day 5 (the last day of dsRNA feeding) and on day 7 (the final day of bioassays). Each treatment was replicated three times, and for each replication, 10-15 larvae were used. Mortality (%) was calculated and the mean values of the experimental replicates were analyzed using a one-way ANOVA, with Tukey's post-hoc t-test to evaluate differences.

Bacterial Transformation and Expression of dsRNA
The RNase III-deficient E. coli strain HT115 (DE3), obtained through the CGC at the University of Minnesota, was grown in Luria broth (LB) medium with ampicillin (100 µ g/mL) and tetracycline (10 µ g/mL). The recombinant L4440 vector was transformed into E. coli HT115 (DE3) competent cells. Single bacterial colonies were cultured in LB broth maintained on a shaker incubator at 37 °C (225 rpm) overnight. Cultured broth was then transferred to 50 mL fresh broth medium containing 100 µg/mL ampicillin and cultured at 37 °C until colony growth reached the late exponential phase, with OD600 = 0.4-0.6. Expression of T7 RNA polymerase was induced by adding a final concentration of 0.5 mM of isopropyl-β-D-1-thiogalactopyranoside (IPTG). Bacteria with dsSHI and dsHSP were then incubated at 37 °C, 30 °C , and 25 °C for up to 4 h to evaluate dsRNA expression. Based on this optimization, further experiments were conducted at 37 °C for dsSHI and 30 °C for dsHSP. Total bacterial RNA was extracted with Trizol reagent and the presence of synthesized dsRNA was confirmed by electrophoresis using a 1% agarose gel.

Biological Activity of Recombinant Bacteria Expressing dsRNA
Laboratory-reared EAB eggs were placed in petri dishes with moistened filter paper and maintained at 24 ± 1 °C in an incubator ( Figure 1). Newly hatched neonates were used in all bioassays.
To determine the biological activity of the recombinant bacteria expressing dsRNA, neonate EAB larvae were fed using a modified droplet feeding bioassay [5], where 1 ml of bacterial culture was centrifuged at 3000 rpm for 15 min and the pellet was dissolved in 100 μL of 1% sucrose solution with green tracking dye (Kroger, Co., USA). For each assay, 3μL of bacterial suspension were fed to individual neonate larvae using the droplet assay. Cellular density of the bacterial culture was determined by considering that an optical density of 1 at 600 nm corresponds to 10 8 bacterial cells/mL [17]. As the control, HT115 (DE3) bacteria were used as a treatment [18]. Neonate EAB larvae were fed dsRNA-expressing bacteria for five consecutive days. On day 6, larvae were fed with 1% sucrose lacking dsRNAs for two days. Assays were maintained in an incubator at 26 ± 1 °C, under a 14:10 (L:D) photoperiod. Mortality was measured on day 5 (the last day of dsRNA feeding) and on day 7 (the final day of bioassays). Each treatment was replicated three times, and for each replication, 10-15 larvae were used. Mortality (%) was calculated and the mean values of the experimental replicates were analyzed using a one-way ANOVA, with Tukey's post-hoc t-test to evaluate differences.

Molecular Validation of Gene Silencing
Following ingestion of dsRNA-expressing bacteria, total RNA was isolated from 5-6 EAB larvae at two time intervals (72 h and 120 h) using Trizol reagent. Total RNA was treated with DNase I to degrade any genomic contamination. cDNA was synthesized using a M-MLV Reverse Transcriptase Kit (Thermo Fisher, USA), and was used as a template for gene expression studies. The expression analysis of the target gene was performed using SYBR™ Green Master Mix (Applied Biosystems, Figure 1. Emerald ash borer egg hatch at 24 ± 1 • C and neonate larva.

Molecular Validation of Gene Silencing
Following ingestion of dsRNA-expressing bacteria, total RNA was isolated from 5-6 EAB larvae at two time intervals (72 h and 120 h) using Trizol reagent. Total RNA was treated with DNase I to degrade any genomic contamination. cDNA was synthesized using a M-MLV Reverse Transcriptase Kit (Thermo Fisher, USA), and was used as a template for gene expression studies. The expression analysis of the target gene was performed using SYBR™ Green Master Mix (Applied Biosystems, USA). qPCR reactions were performed using StepOnePlus Real-Time PCR system (Life Technologies, USA). All reactions were carried out in triplicate with a final volume of 10 µL. A melting curve was generated at the end of each reaction to confirm single product (target) amplification. In order to eliminate undesirable amplification from input recombinant plasmids and/or dsRNAs, primers for qPCR were designed to detect target mRNAs by amplifying only sequences that lay outside of the insert interfering sequences. The TEF1α gene (Table 1) was used as the reference gene ( [19]; Supplemental Material), and the 2 −∆∆Ct method [20] was used to calculate expressions of the target gene relative to the control. A two-tailed t-test was used for statistical analysis to compare the means of a single variable.

Bacterial Transformation and Expression of dsRNA
Bacteria were prepared with the recombinant vector containing fragments of the shi and hsp genes ( Figure 2a). Using colony PCR, we confirmed that 100% of the recombinant bacterial colonies tested contained the insert: dsSHI (483 bp) and dsHSP (468 bp) (Figure 2b), and the IPTG-induced bacteria expressed dsRNA specific to EAB (shi (483 bp) and hsp (468 bp)). Expression of dsSHI was at 37 • C for 4 h and dsHSP was at 30 • C for 4 h (Figure 2c,d). The two genes were successfully synthesized in the bacteria.
Insects 2020, 11, x 4 of 9 USA). qPCR reactions were performed using StepOnePlus Real-Time PCR system (Life Technologies, USA). All reactions were carried out in triplicate with a final volume of 10 μL. A melting curve was generated at the end of each reaction to confirm single product (target) amplification. In order to eliminate undesirable amplification from input recombinant plasmids and/or dsRNAs, primers for qPCR were designed to detect target mRNAs by amplifying only sequences that lay outside of the insert interfering sequences. The TEF1α gene (Table 1) was used as the reference gene ( [19]; Supplemental Material), and the 2 −ΔΔCt method [20] was used to calculate expressions of the target gene relative to the control. A two-tailed t-test was used for statistical analysis to compare the means of a single variable.

Bacterial Transformation and Expression of dsRNA
Bacteria were prepared with the recombinant vector containing fragments of the shi and hsp genes (Figure 2a). Using colony PCR, we confirmed that 100% of the recombinant bacterial colonies tested contained the insert: dsSHI (483bp) and dsHSP (468bp) (Figure 2b), and the IPTG-induced bacteria expressed dsRNA specific to EAB (shi (483bp) and hsp (468bp)). Expression of dsSHI was at 37 °C for 4 h and dsHSP was at 30 °C for 4 h (Figures 2c,d). The two genes were successfully synthesized in the bacteria.

Biological Activity of dsRNA Expressing E. Coli Against EAB
Ingestion of dsRNA-expressing bacteria targeting shi and hsp caused 69.44% and 46.66% mortality, respectively, of neonate larvae at day 7 (Figures 3 and 4). Larvae ingesting dsSHI and dsHSP experienced greater mortality than control larvae ingesting bacteria that lacked the dsRNA, and dsSHI-ingested larvae appeared to grow more slowly based on larval length ( Figure 5). amplified from two individual bacterial colonies), lane HSP: heat shock protein (hsp) gene amplified from two individual bacterial colonies, (c) biosynthesis of dsRNA corresponding to partial sequence of the shibire (shi) gene in the RNAse III deficient bacterial strain (lane M: 1Kb marker, lanes 1 and 2: uninduced dsSHI, lane 3: dsSHI induced by IPTG), and (d) biosynthesis of dsRNA corresponding to partial sequence of the hsp gene in the RNAse III deficient bacterial strain (lane M: 1Kb marker, lane 1: uninduced dsHSP, lane 2: dsHSP induced by IPTG).

Biological Activity of dsRNA Expressing E. Coli Against EAB
Ingestion of dsRNA-expressing bacteria targeting shi and hsp caused 69.44% and 46.66% mortality, respectively, of neonate larvae at day 7 (Figures 3 and 4). Larvae ingesting dsSHI and dsHSP experienced greater mortality than control larvae ingesting bacteria that lacked the dsRNA, and dsSHI-ingested larvae appeared to grow more slowly based on larval length ( Figure 5).

Figure 3.
Insecticidal activity of bacterial dsRNA specific to shi and hsp in EAB larvae is demonstrated by mortality of neonate EAB larvae (mean + Standard Error) following ingestion of dsRNA-expressing bacteria relative to those ingesting HT115 (control), which contained no dsRNA (N = 3). One-way ANOVA, with a post-hoc t-test (Tukey) was used to evaluate differences at p < 0.05.   . Insecticidal activity of bacterial dsRNA specific to shi and hsp in EAB larvae is demonstrated by mortality of neonate EAB larvae (mean ± Standard Error) following ingestion of dsRNA-expressing bacteria relative to those ingesting HT115 (control), which contained no dsRNA (N = 3). One-way ANOVA, with a post-hoc t-test (Tukey) was used to evaluate differences at p < 0.05. Figure 4. Effect of dsRNAs specific to shi and hsp on EAB neonate larval survival (%) 7 d after feeding on dsRNA-expressing bacteria (N = 3). Observations were taken on day 1, day 5, and day 7. One-way ANOVA, with a post-hoc t-test (Tukey) was used to evaluate differences at p < 0.05. by mortality of neonate EAB larvae (mean + Standard Error) following ingestion of dsRNA-expressing bacteria relative to those ingesting HT115 (control), which contained no dsRNA (N = 3). One-way ANOVA, with a post-hoc t-test (Tukey) was used to evaluate differences at p < 0.05. Figure 4. Effect of dsRNAs specific to shi and hsp on EAB neonate larval survival (%) 7 d after feeding on dsRNA-expressing bacteria (N = 3). Observations were taken on day 1, day 5, and day 7. One-way ANOVA, with a post-hoc t-test (Tukey) was used to evaluate differences at p < 0.05.

Molecular Validation of Gene Silencing
Our qPCR analysis showed that bacteria containing dsSHI resulted in a 24.92% reduction in gene expression at 72 h; expression did not differ from controls. However, at 120 h post-exposure there was a 74.14% reduction in the transcript level, which differed significantly from controls (untransformed bacteria) (Figure 6a). Silencing hsp caused 48.67% and 96.94% reductions in the transcript level relative to controls at 72 h and 120 h, respectively (Figure 6b), following exposure to bacterially-expressed dsRNA.

Molecular Validation of Gene Silencing
Our qPCR analysis showed that bacteria containing dsSHI resulted in a 24.92% reduction in gene expression at 72 h; expression did not differ from controls. However, at 120 h post-exposure there was a 74.14% reduction in the transcript level, which differed significantly from controls (untransformed bacteria) (Figure 6a). Silencing hsp caused 48.67% and 96.94% reductions in the transcript level relative to controls at 72 h and 120 h, respectively (Figure 6b), following exposure to bacterially-expressed dsRNA.

Discussion
This is the first time an effective RNAi response in the tree-killing EAB using oral ingestion of dsRNA-expressing bacteria has been demonstrated. We transformed HT115 E. coli to express dsSHI and dsHSP specific to EAB, which caused gene knockdown and showed biocidal activity that resulted in significant mortality of neonate EAB larvae. Use of bacterially-expressed dsRNA to trigger RNAi was first demonstrated experimentally in Caenorhabditis elegans [21], and bacterially-expressed dsRNA has subsequently been used against numerous insect pests. In the coleopteran Leptinotarsa decemlineata, ingestion of bacterially-expressed dsRNA led to effective suppression of five target genes, causing decreases in body weight and significant mortality of treated beetles [22]. In our work the engineered E. coli strain HT115 (DE3), lacking dsRNA-specific RNase III produced EAB-specific dsRNAs and effectively triggered the RNAi pathway upon ingestion by EAB larvae. These features make HT115 (DE3) a promising strain for preparing dsRNA in vivo, providing a less costly and potentially more efficient alternative to in vitro synthesis of dsRNA. However, bacterial dsRNA production can have limitations; recombinant bacterial production of dsRNA is reportedly less effective in causing mortality in Spodoptera exigua (Order: Lepidoptera) than is in vitro synthesized dsRNA [23], perhaps due to a lower expression of target gene(s) in bacteria. These direct comparisons have yet to be made experimentally in EAB.

Discussion
This is the first time an effective RNAi response in the tree-killing EAB using oral ingestion of dsRNA-expressing bacteria has been demonstrated. We transformed HT115 E. coli to express dsSHI and dsHSP specific to EAB, which caused gene knockdown and showed biocidal activity that resulted in significant mortality of neonate EAB larvae. Use of bacterially-expressed dsRNA to trigger RNAi was first demonstrated experimentally in Caenorhabditis elegans [21], and bacterially-expressed dsRNA has subsequently been used against numerous insect pests. In the coleopteran Leptinotarsa decemlineata, ingestion of bacterially-expressed dsRNA led to effective suppression of five target genes, causing decreases in body weight and significant mortality of treated beetles [22]. In our work the engineered E. coli strain HT115 (DE3), lacking dsRNA-specific RNase III produced EAB-specific dsRNAs and effectively triggered the RNAi pathway upon ingestion by EAB larvae. These features make HT115 (DE3) a promising strain for preparing dsRNA in vivo, providing a less costly and potentially more efficient alternative to in vitro synthesis of dsRNA. However, bacterial dsRNA production can have limitations; recombinant bacterial production of dsRNA is reportedly less effective in causing mortality in Spodoptera exigua (Order: Lepidoptera) than is in vitro synthesized dsRNA [23], perhaps due to a lower expression of target gene(s) in bacteria. These direct comparisons have yet to be made experimentally in EAB.
Selection of target gene(s) and target regions within gene(s) is crucial for successful gene silencing. Second-generation sequencing can provide information on target gene selection and screening [4,24], with the goal of selecting genes and/or target regions within genes with increasing RNAi efficiency in the target pest, while having no measurable off-target effects. Here we used two target genes, shibire (shi), and heat shock protein-70kDa protein (hsp), which play essential biological functions and are efficacious in EAB [15], and demonstrate their potential for use in bacterially-expressed RNAi-based EAB management. The heat shock-70kDA protein gene (hsp) functions in protein folding and protects cells from stress [25], while the shibire gene (shi) is involved in production of microtubule bundles, endocytosis and other vesicular trafficking processes [26]. The loss in function of either of these target genes in EAB neonates ingesting transformed bacteria causes significant larval mortality (shi:~69% and hsp:~46%) as well as an apparent suppression of larval growth (shi). Ingestion of bacteria producing dsRNAs, specifically double stranded integrin (dsINT), has also been shown to reduce growth of the lepidoptera, S. exigua [23].
We have demonstrated that, when ingested, bacteria transformed to produce EAB-specific dsRNA can silence target genes and kill neonate EAB, which creates additional potential for its use as a biopesticide. Recombinant bacteria with EAB-specific dsRNA could be sprayed on foliage to be ingested by feeding beetles or on ash stems to be ingested by newly hatched neonates. Naked dsRNAs applied topically or through root drenching can be assimilated into and moved through ash plant tissues [27], suggesting crude extract of bacterially-expressed dsRNA may also be able to be translocated through the plant via soil drench, trunk sprays or injection [28]. Recombinant bacteria producing EAB-specific dsRNAs also make the genetic transformation of ash trees with dsRNA-expressing constructs a more plausible goal [29].

Conclusions
We showed that EAB fed with dsRNA-expressing bacteria results in downregulation of selected genes, demonstrating the potential for application of bacterially-expressed dsRNA for controlling EAB; however, in the current experiment a non-specific dsRNA control was not included. Although optimization of bacterial dsRNA production and expression is needed, our observations suggest that RNA interference mediated by bacterial dsRNA could be a convenient and cost-effective approach for managing this invasive pest.
The specificity of these EAB-specific dsRNAs is an essential step towards moving this approach to the deployment phase; these evaluations are under way. Additionally, management of resistance in the pest population is essential. Western corn root worm, Diabrotica virgifera, has shown field-level resistance to DvSnf7 dsRNA [30], necessitating development and selection of new targets. This process is relatively simple, however, and involves screening for and switching to other appropriate dsRNAs, thereby managing for the potential development of resistance. dsRNAs can be designed for a different region in the same target gene or new genes much more quickly and efficiently than developing a more expensive chemical insecticide [22]. There are clearly knowledge gaps that must be addressed before this technology can be deployed in the EAB-ash system. Potential off-target effects in EAB, mutation of RNAi core machinery genes, mutation of target genes, and enhanced dsRNA degradation, not to mention potential effects on non-target organisms, must be more fully understood before deployment of bacterially-expressed dsRNA in EAB management can become practical.
RNAi is an emerging pest management tool with tremendous potential to protect plants against insect pests. Its application continues to expand into crop and vegetable production [2,6,24,31,32], and horticultural [32,33] and forest systems [5,34,35], and there are numerous native and non-native tree