A Transformed Bacterium Expressing Double-Stranded RNA Specific to Integrin β1 Enhances Bt Toxin Efficacy against a Polyphagous Insect Pest, Spodoptera exigua

Abstract

Background: Oral toxicity of double-stranded RNA (dsRNA) specific to integrin β1 subunit (SeINT) was known in a polyphagous insect pest, Spodoptera exigua. For an application of the dsRNA to control the insect pest, this study prepared a transformed Escherichia coli expressing dsRNA specific to SeINT.

Principal findings: The dsRNA expression was driven by T7 RNA polymerase overexpressed by an inducer in the transformed E. coli. The produced dsRNA amount was proportional to the number of the cultured bacteria. The transformed bacteria gave a significant oral toxicity to S. exigua larvae with a significant reduction of the SeINT expression. The resulting insect mortality increased with the fed number of the bacteria. Pretreatment with an ultra-sonication to disrupt bacterial cell wall/membrane significantly increased the insecticidal activity of the transformed bacteria. The larvae treated with the transformed bacteria suffered tissue damage in the midgut epithelium, which exhibited a marked loss of cell-cell contacts and underwent a remarkable cell death. Moreover, these treated larvae became significantly susceptible to a Cry toxin derived from Bacillus thuringiensis (Bt).

Conclusions: This study provides a novel and highly efficient application technique to use dsRNA specific to an integrin gene by mixing with a biopesticide, Bt.

Fig 1. Construction of a recombinant E. coli expressing dsRNA (‘dsINT’) specific to an integrin β1 subunit fragment (‘SeINT’).

(A) A diagram showing a fragment (410 bp) used for generating dsINT: signal peptide (‘SP’), extracellular domain (‘ECD’), transmembrane (‘TM’), and intracellular domain (‘ICD’). (B) Cloning SeINT into an expression vector of L4440. Lactose promotor (‘Lac Z’) directs the expression of T7 RNA polymerase (‘T7 RP’). SeINT is transcribed by both directions according to T7 RNA polymerase promoters (‘T7’). (C) Production of dsINT with an inducer, IPTG. Transformed E. coli with L4440-SeINT expressed dsINT at the presence of IPTG. (D) Quantification of dsINT according to the transformed bacterial cell numbers. The extracted dsINT was quantified by band intensity based on a standard curved using the known amounts of purified dsINT prepared by in vitro dsRNA construction method. Each measurement was replicated three times.

Fig 2. Insecticidal activity of a transformed E. coli expressing dsRNA (‘dsINT’) specific to an integrin β subunit fragment (‘SeINT’).

(A) Oral toxicity of the recombinant bacteria (10⁷ cells per larva) to young larvae. Each dose test used 10 larvae with three replications. Mortality was measured after 5 days of treatment (‘DAT’). (B) Suppression of SeINT transcript levels measured by RT-PCR in the bacterial treatment. Expression of a ribosomal gene, RL32, was assessed to confirm the cDNA integrity. (C, D) Effects of the recombinant bacteria on developmental rate and toxicity. The recombinant bacteria (10⁷ cells per larva) were independently fed to each instar (‘L1-L5’) with 10 larvae per replication. Untreated instars were fed with fresh cabbage. Each treatment was replicated three times. The developmental period of each treated instar was measured. Mortality was measured at 5 DAT. Different letters above standard deviation bars indicated significant difference among means at Type Ι error = 0.05 (LSD test). ‘NS’ represents not significant, but ‘*’ and ‘**’ indicate significant differences among means at Type I error = 0.05 and 0.01, respectively.

Fig 3. Comparison of two types of dsRNAs prepared by in vitro transcription or by the recombinant bacteria in RNA interference efficacy.

dsRNA of in vitro transcription was orally fed. In dsRNA produced by the bacteria, the recombinant bacteria were orally applied. The number of the bacteria was determined according to the amount of dsRNA produced in the bacteria based on Fig 1B. (A) Suppression of target gene (‘SeINT’) expression levels. At 5 days after treatment (‘DAT’), total RNA was extracted and assessed in the amount of SeINT transcripts. (B) Toxic effects of in vitro dsRNA and bacterial dsRNA on young larvae of S. exigua. Mortality was measured at 5 DAT. Different letters above standard deviation bars indicated significant difference among means at Type Ι error = 0.05 (LSD test). ‘NS’ represents not significant, but ‘*’ indicates significant differences among means at Type I error = 0.05.

Fig 4. Pretreatment effect of dsINT-expressing E. coli on its insecticidal activity.

(A) Two different pretreatments of recombinant bacteria and their survival: heat-killing (100^°C for 100 min) and sonication. After pretreatment, the bacterial suspension was plated on LB+AMP plate and cultured for 15 h at 37^°C. (B) Suppression of SeINT transcript levels with increase of bacterial doses. The expression was assessed by RT-PCR. Expression of a ribosomal gene, RL32, was assessed to confirm the cDNA integrity. (C) Oral toxicity of the recombinant bacteria against young larvae of S. exigua. Each larva was fed 10⁷ bacteria. Mortality was measured after 5 days of treatment (‘DAT’). Different letters above standard deviation bars indicate significant difference among means at Type Ι error = 0.05 (LSD test).

Fig 5. Effect of dsINT-expressing E. coli on development of S. exigua and on the midgut epithelium.

(A) Effect of the recombinant bacteria on larval development. Control used the bacteria transformed with empty vector. Each larva was fed 10⁷ cells and incubated at 25°C. The pictures were taken at 5 days after treatment. (B) Cytotoxic effect of dsINT expressing bacteria on midgut epithelium. The tissues were stained with 0.1% trypan blue. Stained cells indicate “dead”. (C) Disruption of midgut epithelium in cell-to-cell contacts by oral treatment of dsINT-expressing E. coli.

Fig 6. Enhancement of dsINT-expressing E. coli on Bt Cry toxin toxicity against S. exigua.

(A) Toxicity of Cry1Ca-expressing E. coli against young larvae. Each treatment used 10 larvae with three replications. Mortality was measured at 5 days after treatment (‘DAT’). (B) Different susceptibilities of dsINT bacteria (10⁷ cells per larva)-treated larvae to Cry1Ca-expressing bacteria according to the pretreatment period. Mortality was measured at 7 days after dsINT-bacterial treatment. (C) Effect of dsINT-expressing E. coli on toxicity of a commercial Bt (B. thuringiensis subsp. aizawai, Xentari, 1,000 ppm). ‘Bt’ treatment represents no dsINT treatment. ‘Bt+dsINT’ represents a mixture treatment of Bt treatment at 48 h after dsINT bacterial treatment. Mortality was measured at 7 days after dsINT-bacterial treatment. Each treatment used 10 larvae with three replications. Different letters above standard deviation bars indicate significant difference among means at Type Ι error = 0.05 (LSD test).