
==== Front
J Ind Microbiol Biotechnol
J Ind Microbiol Biotechnol
jimb
Journal of Industrial Microbiology & Biotechnology
1367-5435
1476-5535
Oxford University Press

39152090
10.1093/jimb/kuae028
kuae028
Original Paper
Fermentation, Cell Culture and Bioengineering
Jimb/5
AcademicSubjects/SCI01150
AcademicSubjects/SCI00540
Optimizing dsRNA engineering strategies and production in E. coli HT115 (DE3)
https://orcid.org/0000-0002-7895-0300
da Rosa Juliana Department of General Biology, Londrina State University, Celso Garcia Cid Road, PR 445, km 380, University Campus, 86057-970 Londrina, PR, Brazil
Embrapa Soja, Carlos João Strass Highway, Acess Orlando Amaral, District of Warta, 86085-981 Londrina, PR, Brazil

Viana Américo José Carvalho Embrapa Soja, Carlos João Strass Highway, Acess Orlando Amaral, District of Warta, 86085-981 Londrina, PR, Brazil
Arthur Bernardes Foundation, Headquarters Building, no number - University Campus, 36570-900 Viçosa, MG, Brazil

Ferreira Fernando Rafael Alves Embrapa Soja, Carlos João Strass Highway, Acess Orlando Amaral, District of Warta, 86085-981 Londrina, PR, Brazil
Arthur Bernardes Foundation, Headquarters Building, no number - University Campus, 36570-900 Viçosa, MG, Brazil

Koltun Alessandra Embrapa Soja, Carlos João Strass Highway, Acess Orlando Amaral, District of Warta, 86085-981 Londrina, PR, Brazil

Mertz-Henning Liliane Marcia Embrapa Soja, Carlos João Strass Highway, Acess Orlando Amaral, District of Warta, 86085-981 Londrina, PR, Brazil

Marin Silvana Regina Rockenbach Embrapa Soja, Carlos João Strass Highway, Acess Orlando Amaral, District of Warta, 86085-981 Londrina, PR, Brazil

Rech Elibio Leopoldo Embrapa Genetic Resources and Biotechnology, National Institute of Science and Technology in Synthetic Biology, 70770-917 Brasilia, DF, Brazil

Nepomuceno Alexandre Lima Embrapa Soja, Carlos João Strass Highway, Acess Orlando Amaral, District of Warta, 86085-981 Londrina, PR, Brazil

Correspondence should be addressed to: Juliana da Rosa at ju_lianarosa@hotmail.com
2024
16 8 2024
16 8 2024
51 kuae02827 5 2024
14 8 2024
05 9 2024
© The Author(s) 2024. Published by Oxford University Press on behalf of Society of Industrial Microbiology and Biotechnology.
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs licence (https://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any medium, provided the original work is not altered or transformed in any way, and that the work is properly cited. For commercial re-use, please contact journals.permissions@oup.com

Abstract

Producing double-stranded RNA (dsRNA) represents a bottleneck for the adoption of RNA interference technology in agriculture, and the main hurdles are related to increases in dsRNA yield, production efficiency, and purity. Therefore, this study aimed to optimize dsRNA production in E. coli HT115 (DE3) using an in vivo system. To this end, we designed a new vector, pCloneVR_2, which resulted in the efficient production of dsRNA in E. coli HT115 (DE3). We performed optimizations in the culture medium and expression inducer in the fermentation of E. coli HT115 (DE3) for the production of dsRNA. Notably, the variable that had the greatest effect on dsRNA yield was cultivation in TB medium, which resulted in a 118% increase in yield. Furthermore, lactose induction (6 g/L) yielded 10 times more than IPTG. Additionally, our optimized up-scaled protocol of the TRIzol™ extraction method was efficient for obtaining high-quality and pure dsRNA. Finally, our optimized protocol achieved an average yield of 53.3 µg/mL after the production and purification of different dsRNAs, reducing production costs by 72%.

Improved production of double-stranded RNA (dsRNA) in E. coli HT115 (DE3) bacteria resulted in significant yield increases and cost reductions through optimized methods.

Graphical Abstract

Graphical Abstract Improved production of double-stranded RNA (dsRNA) in E. coli HT115 (DE3) bacteria resulted in significant yield increases and cost reductions through optimized methods.

Bacterial dsRNA production
Double-stranded RNA
Lactose induction
RNAi
Conselho Nacional de Desenvolvimento Científico e Tecnológico 10.13039/501100003593 465603/2014-9 Conselho Nacional de Desenvolvimento Científico e Tecnológico 10.13039/501100003593
==== Body
pmcIntroduction

RNA interference (RNAi) is a defense system present in eukaryotic cells. This mechanism is triggered by the entry of double-stranded RNA (dsRNA) molecules into the cell cytoplasm (Timmons and Fire, 1998). Once in the cytoplasm, dsRNA is cleaved by ribonuclease III (Dicer) into small interfering RNAs (siRNAs) of 21–24 bp (Ding, 2010). These siRNAs are incorporated into an enzymatic complex, the RNA-induced silencing complex, which uses one of the siRNA strands as a guide to cleave complementary mRNAs, degrading them (Baulcombe, 2004; Dalakouras et al., 2019). In agriculture, dsRNA can be used to control insects, fungi, nematodes, viruses, and weeds (Hendrix et al., 2021; Mitter et al., 2017; Zotti et al., 2018). However, one of the bottlenecks in developing RNAi-based pesticides is the scalable and high-quality production of these molecules.

Double-stranded RNA can be produced in vitro or in vivo. Both methods use T7 RNA polymerase to transcribe sequences intended to produce dsRNA (Voloudakis et al., 2015). In vitro production is generally performed using commercial kits. However, such systems are expensive and have low yields (Das and Sherif, 2020). In vitro production is recommended for proof of concept or small tests at the initial research stage. For larger test scales or manufacturing of RNAi-based products, other dsRNA production methods better serve these goals, such as in vivo production via liquid fermentation. Several species can be applied for the in vivo production of dsRNA, such as Saccharomyces cerevisiae, Corynebacterium glutamicum, and Escherichia coli (Ahn et al., 2019; Guan et al., 2021; Hashiro et al., 2021).

E. coli HT115 (DE3) is the strain most widely used to produce dsRNA. It was originally developed for gene silencing work in Caenorhabditis elegans (Timmons et al., 2001). E. coli HT115 (DE3) has a suppressed ribonuclease III, allowing dsRNA accumulation in the cell cytoplasm, as well as the T7 polymerase-encoding gene inducible by isopropylthiogalactoside (IPTG) (Timmons et al., 2001). The pL4440 plasmid contains two T7 promoters in an inverted orientation flanking a multiple cloning site (MCS). E. coli HT115 (DE3) associated with the expression vector pL4440, has become one of the most commonly used expression systems to produce dsRNA. Therefore, efficient microorganism/vector systems to produce dsRNA can contribute to production on an industrial scale.

In the in vivo dsRNA production process, the induction of transcription through the addition of an inducer and an optimized culture medium is indispensable. In systems using T7 RNA polymerase, synthetic allolactose IPTG is a widely applied inducer in the small-scale production of dsRNA (Guan et al., 2021; Mo et al., 2022). IPTG can lead to metabolic stress and is both toxic and costly (Khani and Bagheri, 2020; Neubauer et al., 1992). Therefore, exploring other inducers that are more efficient, less expensive, and non-toxic becomes imperative for improving this process (Delgado-Martín and Velasco, 2021). Additionally, the choice of culture medium to be used is crucial for cell growth and metabolism and directly impacts dsRNA production (Papić et al., 2018; Thammasorn et al., 2015).

Regarding the steps involved in dsRNA production in bacteria, dsRNA purification requires optimization at both bench and industrial scales. The bench-scale extraction of dsRNA after in vivo cultivation includes approaches commonly used to extract other nucleic acids, such as phenol-chloroform extraction (Chen et al., 2019), silica column-based methods (Nwokeoji et al., 2016), TRIzol™ reagent protocols (Ongvarrasopone et al., 2007), alcohol extraction (Posiri et al., 2013), and phenol/chloroform/isoamyl alcohol extraction, followed by DNase and RNase digestions (Ahn et al., 2019). For dsRNA applications, integrity during RNA extraction and purity are crucial (Nwokeoji et al., 2017). In this sense, dsRNA extraction can impact the future development of RNAi-based products for use in agriculture.

To improve the dsRNA production methodology in E. coli HT115 (DE3) bacteria and the dsRNA purification process, we optimized several parameters. First, we designed a new pCLone_VR_2 expression vector. We carried out assays with different types of culture media and concentrations of the lactose inducer. Additionally, we compared two extraction protocols commonly used in dsRNA extraction and scaled the TRIzol extraction method. Furthermore, we analyzed the quality and quantity of the dsRNAs produced and analyzed the yield and cost-benefit. Finally, after the optimized protocol, we tested the production and extraction of different dsRNA molecules.

Materials and Methods

Selection of the Target Genes

dsRNA sequences from the genes Phytoene Desaturase (PDS), 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS), and Magnesium Chelatase, subunit D (MgD) of Digitaria insularis, sourgrass, were selected to be cloned into the vectors. Primers were designed with the addition of BglII and XhoI restriction sites for cloning the PDS sequence into L4440, while for cloning into pClone_VR_2, AgeI (Anza™) and SalI (Anza™) were used. Additionally, to clone the PDS1, PDS2, PDS3, MgD1, MgD2, MgD3, and EPSPS sequences into pClone_VR_2, Esp3I (Anza™) was used. Three guanines (GGG) were also added to provide binding space for the restriction enzymes. The sequences of the primers used can be found in Supplementary Table S1. The sequences were amplified using Platinum™ Taq DNA Polymerase (Thermo Fisher Scientific) following the manufacturer's instructions.

Polymerase chain reaction (PCR) amplification was performed in a final volume of 50 µL containing 1.5 mM MgSO4, 400 µM each dNTP, 0.2 µM each primer, and 1.5 U of Taq DNA polymerase in 60 mM Tris-SO4 buffer (pH 8.9) with 180 mM (NH4)2SO4 and sourgrass cDNA as the template. The reaction proceeded with the following program: 5 min at 95°C; 35 cycles of 30 s at 94°C, 30 s at 58°C, and 30 s at 68°C; and a final cycle of 7 min at 68°C. Aliquots of the reactions were run on a 1% agarose gel. Since they had single amplification bands of the expected size, the amplification products were purified from the PCR using the Wizzard® SV Gel and PCR Clean-Up System Kit (Promega) according to the manufacturer's instructions.

Bacterial Strain and Plasmid Construction

dsRNA synthesis was carried out in E. coli HT115 (DE3) transformed with two expression vectors: pL4440 and pCloneVR_2. pL4440 (plasmid #1654, Addgene) contains two T7 promoters in inverted orientations flanking a MCS (Fig. 1a). We designed a new vector, pCloneVR_2, with two T7 promoters and terminators flanking the MCS (Fig. 1b). The sequence of pCloneVR_2 is presented in the supplementary data (Supplementary Table S2).

Fig. 1. Expression vectors and dsRNA production in E. coli HT115 (DE3). (a) Scheme of the L4440 vector showing the multiple cloning site and bidirectional T7 promoters. (b) Scheme of pCloneVR_2 showing the multiple cloning site flanked by bidirectional T7 promoters and terminators. (c) Schematic diagram of dsRNA production in the E. coli HT115 (DE3) system. The target gene fragment is inserted into the multiple cloning site of L4440 or pClone_VR_2. Next, the vector is transformed into E. coli HT115 (DE3), an RNase III-deficient strain. Bacteria grow in a liquid medium, and with the addition of the inducer, the expression of T7 RNA polymerase occurs. Next, T7 polymerase transcribes the dsRNA strands. Subsequently, the produced dsRNA can be extracted and purified. *MCS = multiple cloning site.

The vectors and the amplified DNA fragment were cut with the restriction enzymes. The fragments and vectors were ligated using T4 DNA ligase at 4°C for 16 h. The sizes of the cloned dsRNA fragments are listed in Supplementary Table S1. The ligation reaction was transformed into E. coli DH5α cells via thermal shock. After transformation, the bacterial cells were inoculated on plates with solid Luria Bertani (LB) medium containing ampicillin (100 µg/mL) and incubated at 37°C for 16 h. Recombinant colonies were identified through PCR with the aforementioned specific primers. Subsequently, the selected colonies were inoculated into liquid LB medium containing ampicillin (100 µg/mL) and incubated at 37°C under agitation at 180 rpm for 16 h. Then, a miniprep was performed to obtain plasmid DNA using the QIAprep Spin Miniprep Kit (Qiagen) following the manufacturer's recommendations. The recombinant colonies were stored at −80°C in LB with ampicillin (100 µg/mL) and 25% glycerol.

Finally, the recombinant vectors were transformed into E. coli HT115 (DE3) by thermal shock, and recombinant colonies were selected using solid LB supplemented with the antibiotics tetracycline (12.5 µg/mL) and ampicillin (100 µg/mL).

dsRNA Production

The dsRNA production system in E. coli HT115 (DE3) is shown in Fig. 1c. The bacteria containing the expression vector were cultured in 5 mL of liquid LB medium with tetracycline (12.5 µg/mL) and ampicillin (100 µg/mL) and incubated overnight (16 h) at 37°C under agitation at 180 rpm. Then, an aliquot (1 mL) of each culture grown overnight was added to 50 mL of LB medium with the same antibiotics in 500 mL flasks. The culture was incubated until the mid-exponential phase (OD600 nm of 0.4) was reached. The next step was the induction of T7 promoter release; the inducing agent IPTG (0.4 mM standard) or lactose at concentrations of 3 g/L (8.3 mM), 6 g/L (16.6 mM), 9 g/L (25 mM), 12 g/L (33.3 mM), or 15 g/L (41.6 mM) was added. After the beginning of lactose induction, 10 mL of culture medium, antibiotics, and lactose were added at the same concentrations every 1 h to implement a fed-batch fermentation. The same approach was applied to IPTG induction; IPTG was added at the start of induction, while the complete medium containing antibiotics and IPTG was replenished during feedings to maintain nutrient supply. We also tested initiating induction with lactose at the beginning of fermentation, which was conducted with a total induction time of 5 h. Finally, the bacterial suspension was collected by centrifugation at 4 000×g for 5 min at room temperature, the supernatant was discarded, and the bacterial pellet was kept in a freezer at −20°C until extraction.

To establish the best medium to produce dsRNA, three media were tested: LB medium, containing 10 g/L peptone, 5 g/L yeast extract, and 5 g/L sodium chloride; TB (Terrific Broth) medium, 24 g/L yeast extract, 12 g/L tryptone, 5% glycerol, and potassium phosphate (0.17 M monobasic and 0.72 M dibasic); and M9 medium (15 g/L KH2PO4, 64 g/L Na2HPO4.7H2O, 2.5 g/L NaCl, and 5.0 g/L NH4Cl) supplemented with 5% glycerol.

dsRNA Extraction Using TRIzolTM

In the TRIzolTM extraction protocol (Ongvarrasopone et al., 2007), 40 mL of bacterial suspension was pelleted into 50 mL Falcon tubes. The bacterial pellets were dissolved in 800 µL of 0.1% SDS in 1× PBS buffer, and the tubes were transferred to a water bath at 100°C for 2 min. Then, 800 µL of TRIzolTM (Invitrogen) was added and incubated at room temperature for 5 min. In the next step, 200 µL of chloroform was added and incubated at room temperature for 2 min, followed by centrifugation at 12 000×g for 10 min. The supernatant was removed and transferred to a new tube, which received 0.7 volume of isopropanol, and after incubation at room temperature for 10 min, it was centrifuged at 12 000×g for 10 min. Next, two washes with 75% ethanol were performed. Then, the pellet was completely dried and resuspended in 180 µL of nuclease-free water.

dsRNA Alcoholic Extraction

In the alcoholic extraction method (Posiri et al., 2013), with adaptations by the author, the bacterial pellet was resuspended in a 70% ethanol solution with 1× PBS, incubated at 4°C for 5 min and then centrifuged at 10 000×g for 10 min at 4°C. The new pellet was resuspended in 1 mL of 150 mM NaCl and incubated at 4°C for 1 h, after which the centrifugation was repeated. The supernatant was collected, 1 volume of isopropanol was added, and after incubation at room temperature for 10 min, the mixture was centrifuged at 12 000×g for 10 min. Then, the pellet containing the nucleic acids was washed in 70% ethanol, dried, and resuspended in 180 µL of nuclease-free water.

DNase and RNase Treatment and dsRNA Purification

After extracting the dsRNA, the samples were subjected to treatment with TURBO™ DNase (Ambion) and RNase A (Invitrogen) to remove single-stranded DNA and RNA (Ahn et al., 2019). The dsRNA samples extracted using the TRIzolTM or ethanol methods were diluted in 180 µL of water, treated with 2 µL of turbo DNAse and 20 µL of RNase A, and incubated for 1 h at 37°C.

After treatment with DNase and RNase, the samples were purified (Ahn et al., 2019). In each sample, 200 µL of phenol: chloroform: isoamyl alcohol (25:24:1) was added, and after vigorous stirring, centrifugation was performed at 12 000×g for 15 min. The upper phase was transferred to another tube, 100 µL of NH4OAc (7.5 M) and 100 µL of isopropanol were added, and the samples were centrifuged at 12 000×g for 45 min at 4°C. This step was followed by two washes with 70% ethanol and centrifugation at 12 000×g for 10 min. The pellet was dried and resuspended in nuclease-free water. The samples were quantified in a Nanodrop spectrophotometer at 260 nm, and the quality was verified by electrophoresis in a 1% agarose gel.

Upscaled TRIzolTM Method

For this assay, 40 mL of bacterial suspension was pelleted into 50 mL Falcon tubes. Then, 5 mL of 0.1% SDS in 1× PBS buffer was added, and the mixture was vortexed until complete dissolution of the bacterial pellet. Then, the tubes were placed in a water bath at 100°C for 5 min. Afterward, 4 mL of TRIzolTM was added, the mixture was manually homogenized for 2 min, and the samples were incubated at room temperature for 5 min. Subsequently, 2 mL of chloroform was added, and the mixture was stirred by inversion and incubated at room temperature for 2 min. The tubes were subsequently centrifuged at 12 000×g for 30 min at 4°C. The supernatant was transferred to 15 mL tubes and placed in a shaker. To improve nucleic acid precipitation, the tubes were kept at −20°C for 16 h.

After precipitation, the tubes were centrifuged at 12 000×g at 4°C, and the supernatant was discarded. The pellets were washed with 5 mL of 95% ethanol and centrifuged at 12 000×g for 5 min at 4°C. The pellets were dried in a vacuum concentrator, resuspended in 900 µL of RNase-free water, and incubated for 25 min at 37°C in an oven for complete dissolution of the pellet. Then, 100 µL of 10× Turbo DNase buffer was added and incubated at room temperature for 30 min, and 4 µL of Turbo DNase and 10 µL of RNase A were added and incubated for 30 min. Then, 1 mL of phenol/chloroform/isoamyl alcohol (25:24:1) was added, and the mixture was vigorously stirred for 20 s. The tubes were centrifuged for 15 min at 12 000×g at room temperature. The upper phase was transferred to 2 mL tubes, and 500 µL of 7.5 M ammonium acetate and 500 µL of isopropanol were added, followed by vigorous shaking and incubation at −20°C for 16 h.

Finally, the tubes were centrifuged for 60 min at 12 000×g at 4°C, and the supernatant was discarded. The pellet was washed twice with 500 µL of 70% ethanol and centrifuged at 12 000×g for 10 min at 4°C. The pellet was dried and dissolved in 1 mL of RNase-free water. The samples were quantified in a spectrophotometer, and a quality check was performed in a 1% agarose gel.

Experimental Design and Statistical Analysis

Statistical analysis and graphing of the data were performed using Prism Software. The data obtained from the comparison between extraction protocols were analyzed by Student's t-test with p < .05. Data from the inducer, culture media, and different dsRNA experiments were evaluated using analysis of variance, followed by Tukey's test with p < .05.

Results

Expression Vectors

First, we compared the L4440 vector, commonly used to produce dsRNA in the laboratory (Ahn et al., 2019; Papić et al., 2018; Timmons et al., 2001), with a new vector designed by our group, pCloneVR_2, in which two T7 terminators were inserted upstream of the T7 promoters (Fig. 1a). The two vectors received the insertion of the sequence of the PDS gene from sourgrass. The dsRNAs produced using the pCloneVR_2 and L4440 vectors are shown in the agarose gel in Fig. 2. Following extraction using the TRIzolTM method, treatment with DNase and RNase, and purification of the dsRNA produced with the HT115/L4440 system, bands larger than the expected size were observed (Fig. 2a). However, this phenomenon did not occur when utilizing the HT115/pCloneVR_2 system (Fig. 2b).

Fig. 2. Production of dsRNA from the PDS gene using the vectors (a) L4440 and (b) pCloneVR_2 produced with E. coli HT115 (DE3) system. M: 1 kb plus DNA ladder.

The pCloneVR_2 vector was designed by incorporating two terminators downstream of the T7 promoter. This strategy was also used by other authors to design vectors for dsRNA production (Chen et al., 2019; Nwokeoji et al., 2017). The purpose behind adding two T7 terminators upstream of the T7 promoter was to induce transcription termination by T7 RNA polymerase, which does not occur in the L4440 vector. Thus, when the T7 polymerase passes through the cloned fragment, transcription proceeds unhindered, resulting in the production of long dsRNA fragments derived from the vector backbone (Ma et al., 2020). The production of dsRNA containing these regions is undesirable due to their potential to induce off-target effects. These unintended effects occur when siRNAs possess sequences similar to those of non-target genes (Chen et al., 2021). One of the biggest challenges in designing an RNAi-based pesticide strategy involves preventing the generation of off-target sequences (Naito et al., 2005). Therefore, the newly tested vector pCloneVR_2 demonstrates efficiency in dsRNA production, addressing this concern.

Optimization of the Extraction Method

In our work, we tested two different extractions based on TRIzolTM with adaptations (Ongvarrasopone et al., 2007) and ethanol (García et al., 2015; Posiri et al., 2013). This step was followed by treatment with turbo DNase and RNase A and purification with phenol: chloroform: isoamyl (25:24:1) (Ahn et al., 2019).

Both methods efficiently extracted nucleic acids, as shown in Fig. 3a. The use of crude extract is not indicated since other nucleic acids, such as genomic DNA and bacterial RNA, are present. In the ethanol extraction method, the single-stranded RNA (ssRNA) band was less intense than that in the TRIzolTM extraction (Fig. 3a), requiring a smaller amount of RNase A for the treatment, thus reducing costs. Furthermore, the TRIzolTM extraction method resulted in a higher yield (3.9 µg/mL) than did the ethanol extraction method (2.5 µg/mL) (Fig. 3b).

Fig. 3. dsRNA extraction using TRIzolTM and ethanol extraction methods. (a) Agarose gel after dsRNA extraction from liquid fermentation in the HT115-L4440 system and lactose induction. M: 1 kb plus ladder; Lanes 1, 2, and 3 are the triplicates of ethanol extraction; Lanes 4, 5, and 6 are the triplicates of TRIzolTM extraction; Lanes 7, 8, and 9 are the triplicates after treatment with DNase and RNase A; Lanes 10, 11, and 12 are the triplicates after DNase and RNase A treatment. The arrow indicates a dsRNA size of approximately 341 bp. (b) Yield of dsRNA-PDS after extraction with TRIzolTM and ethanol extraction methods. * indicates a significant difference according to Student's t-test, p < .05.

Therefore, in this study, we optimized the standard TRIzolTM extraction protocol to improve the efficiency of dsRNA extraction and purification. The Upscale TRIzolTM protocol includes increasing the amount of TRIzolTM and other reagents to enhance extraction capacity. Additionally, it includes improved steps to minimize dsRNA loss, such as optimizing centrifugation and precipitation times to ensure maximum dsRNA recovery. Using the standard protocol found in the literature (Posiri et al., 2013), we obtained a yield of 4 µg/mL dsRNA (Fig. 3); from the upscale TRIzolTM protocol, we reached 14.2 µg/mL from the production in LB medium (Fig. 5a). These results highlight the significant influence of the extraction and purification method on the final yield of dsRNA molecules from bacterial culture.

Optimization of Cultivation Conditions and Induction

We compared the induction of dsRNA synthesis by IPTG and five different lactose concentrations in E. coli HT115 (DE3) cells containing pClone_VR_2 (Fig. 4a). Lactose addition commenced after the fermentation reached an OD600 of 0.4 and was repeated every hour during the 5-hour fermentation period. The peak dsRNA production (44.8 µg/mL) after 5 hr of fermentation was observed with the addition of 6 g/L of lactose every hour. When comparing dsRNA production with lactose at a concentration of 6 g/L to IPTG (4.3 µg/mL) added only once to the medium, a 10-fold increase in productivity was observed. There was no significant difference between the concentrations of 6 g/L and 9 g/L (33.6 µg/mL). However, due to the similarity in means, the concentration of 6 g/L is preferable, considering resource economy. Providing lactose from the beginning of the culture and subsequently replenishing it resulted in 18% greater dsRNA production than initiating feeding after reaching an OD600 of 0.4 (Fig. 4b).

Fig. 4. Expression inducers IPTG and lactose for dsRNA production in E. coli HT115 (DE3) using the vector pCloneVR_2. (a) dsRNA yield comparison different lactose concentrations and IPTG as inducers, analyzed using the Tukey test. (b) dsRNA yield comparison between induction at the beginning of fermentation and after reaching an OD600 of 0.4, analyzed using the Student's t-test. Error bars represent the standard deviation of the mean. * indicates a significant difference (p < .05).

In addition to the inducer, the culture medium is crucial for optimizing dsRNA production protocols via bacterial culture. In this study, LB, TB, and M9 (supplemented with 5% glycerol) media were evaluated. Using TB, we reached a yield of 31 µg/mL, which represents an increase of 118% in dsRNA yield compared to that of LB (14.2 µg/mL), under the same expression system and process of optimized extraction and purification (Fig. 5a). On the other hand, M9 (0.72 µg/mL) showed a significantly lower yield than LB and TB (Fig. 5a). Fig. 5b shows the integrity of the bands after dsRNA extraction and purification.

Fig. 5. Comparative analysis of dsRNA yield between LB, TB, and M9 media in the HT115/pCloneVR_2 system using the inductor concentration of 6 g/L and upscale TRIzol TM method. (a) Yield of dsRNA-EPSPS in the three media tested. (b) dsRNA quality after extraction and purification. (c) OD600 values during fermentation. M: 1 kb plus ladder. * indicates a significant difference according to Tukey's test, p < .05.

The dsRNA yield is related to bacterial growth, and more nutritious media can provide more suitable development conditions for bacteria. This correlation can be visualized by the OD600 values of TB 2.82, LB 2.49, and M9 0.77 after 5 h of production, which shows that the bacterial biomass from cultivation in the TB medium is superior to that from the other media.

In our study, dsRNA production using TB medium and the upscaled TRIzolTM protocol reduced production costs by 73%. In terms of costs, extraction using alcohol is cheaper than extraction with TRIzolTM, as shown in Table 1. For laboratories with limited resources, alcohol extraction is a suitable choice. However, to achieve the same total amount of dsRNA, nearly twice the cell culture volume is needed with the alcohol extraction method compared to conventional TRIzolTM extraction, and almost six times the volume compared to the scaled-up TRIzolTM protocol. For studies that require high amounts of dsRNA in short period of time, it may be preferable to use a more efficient method like TRIzolTM, even if it is more expensive. This is because obtaining equivalent quantities with alcohol protocol would necessitate larger volume of culture. Our study also incorporates an upscale TRIzolTM protocol to further enhance dsRNA yield. Therefore, the most appropriate protocol depends on the laboratory conditions and the urgency of producing and testing RNAi strategies.

Table 1. Costs of dsRNA Production and Purification in This Study

Culture medium	Extraction protocol	Yield (µg/mL)	Cost of dsRNA (BRL/mg)	Cost of dsRNA (USD/mg)	
LB	Alcoholic	2.5	55.63	11.17	
LB	TRIzol	3.9	77.23	15.51	
LB	UpscaleTRIzol	14.2	78.23	15.71	
TB	UpscaleTRIzol	53.0	21.38	4.29	
The listed prices of reagents and materials reflect current market rates at the time of writing this article.

Comparison of the Yield and Purity of the Optimized Protocol for Different dsRNAs

Finally, we compared the production of dsRNA from different genes and analyzed the quality and yield of our optimized protocol. To do so, we generated sequences of the Phytoene Desaturase genes (PDS1, PDS2, PDS3), Magnesium Chelatase, subunit D (MgD1, MgD2, and MgD3), 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS), and β-galactosidase (LacZ). Our results showed that the yields of different dsRNA sequences did not differ significantly from one another. The production of dsRNA per mL of culture medium for the different dsRNAs were as follows: PDS1 44.4 µg/mL, PDS2 45.0 µg/mL, PDS3 67.3 µg/mL, MgD1 51.4 µg/mL, MgD2 47.0 µg/mL, MgD3 72.2 µg/mL, EPSPS 55.2 µg/mL, and LacZ 44.4 µg/mL (Fig. 6a). On average, the production of dsRNA, regardless of the sequence, was 53.3 µg/mL.

Fig. 6. Yield and quality of different dsRNAs produced using the HT115 (DE3)/pClone_VR_2 system under optimized cultivation and extraction conditions. The black line represents the arithmetic mean of the yield of all dsRNAs (53.3 µg/mL). (b) The 260/280 and 260/230 ratios measured on a Nanodrop spectrophotometer.

The quality of the dsRNA used can affect the effectiveness of molecules in RNAi-mediated silencing (Ahn et al., 2019). The quality of the produced dsRNAs was analyzed by spectrophotometry (260 nm) using the 260/280 and 260/230 ratios (Fig. 6b). The ratios of the dsRNAs evaluated in this study were approximately 2, suggesting that there were no proteins or salt residues after dsRNA purification (Fig. 6b). These data indicate that the optimized protocol using the pClone_VR_2/HT115 system and other production, extraction, and purification parameters generate dsRNAs with quality and purity.

Discussion

The production of dsRNA of adequate quality and quantity is one of the first steps toward consolidating RNAi technology for pesticide production (Ahn et al., 2019). It is estimated that in field applications, doses of 2–10 g of dsRNA per hectare will be required depending on the target organism, method of application, protection with nanoparticles, and sensitivity of the organism, among other factors (Das and Sherif, 2020; Zotti et al., 2018). To meet this demand in developing new products, supplying dsRNA in good quantity, quality, and low cost is essential. In vivo synthesis is one of the most promising strategies for reducing the cost of dsRNA production (Nwokeoji et al., 2022). In this study, we selected the best fermentation and extraction conditions for producing dsRNA with high yield and purity on a bench scale using the E. coli HT115 (DE3) strain.

In bacterial dsRNA production, the vector/microorganism interaction is essential. Our results showed that the new vector pClone_VR_2 is efficient for dsRNA production. New expression systems can increase the efficiency of dsRNA production in vivo. An example is the bacterium Corynebacterium glutamicum associated with the pPK4H1 vector, which results in a substantial dsRNA yield of 1 g/L (Hashiro et al., 2021). Notably, the pPK4H1 vector has two convergent T7 promoters and terminators, in addition to allowing a high copy number (800 copies). Another interesting system is pET28-BL21 (DE3), which has been shown to be approximately three times more productive than HT115/L4440 (Ma et al., 2020). Therefore, efficient microorganism/vector systems to produce dsRNA, combined with efficient methods of extraction and purification of these molecules, can contribute to production on an industrial scale.

Cultivation conditions within heterologous expression systems require optimization to increase yield and reduce production costs. The production of dsRNA is intricately linked to bacterial growth and the inducer employed (Papić et al., 2018). Here, we investigated the effects of inducers and culture media on dsRNA production.

The use of lactose induction is an attractive alternative for cultivation in both flasks and bioreactors owing to its low cost, absence of toxicity, and carbon-supplying properties (Neubauer et al., 1992). In this study, we demonstrate that a lactose concentration of 6 g/L provides the highest yield of dsRNA compared to concentrations both above and below this value, as well as IPTG. Lactose, in comparison to IPTG, the gold standard inducer in T7 systems, is cost-effective and non-toxic (Hausjell et al., 2020). However, it requires continuous supplementation as it is rapidly metabolized by bacterial cells. Beyond its role as an inducer, lactose serves as a carbon source, undergoing enzymatic breakdown into glucose and galactose (Kram and Finkel, 2015). Notably, lactose influences not only transcription induction but also cell growth. Compared with IPTG fed-batch systems, lactose can increase bacterial growth by up to 20% (Thammasorn et al., 2015).

Culture media are crucial for bacterial production; the medium should provide the necessary nutrients for microbial growth without affecting the production of the molecule of interest. Among the media tested, TB is the richest nutritionally, while M9 is the poorest, evidencing the correlation between bacterial biomass and yield. In addition, in the TB medium, the presence of glycerol as a carbon source, whose metabolism releases acid into the medium, and the addition of potassium phosphate, which acts as a buffer, contribute to maintaining the pH of the medium and consequently to cell viability (Kram and Finkel, 2015; Thammasorn et al., 2015). In this context, our results demonstrated that TB medium was the most suitable for dsRNA production using the pClone_VR_2/E. coli HT115 system, increasing dsRNA yield by 118%.

Productivity can be improved by using bioreactors for dsRNA production in a controlled and optimized way (Hausjell et al., 2020). Optimizations in the different stages of production lead to a significant increase in dsRNA yield (Papić et al., 2018). For instance, using bench fermenters allows for more rigorous control of aeration, temperature, and pH than production in flasks. On a laboratory scale, a 6.2 µg/mL dsRNA yield was obtained using TB medium (Thammasorn et al., 2015). However, in a bioreactor operating in fed-batch mode, this concentration reached 95 µg/mL. In another study using the HT115/pGEM-GT3 system and TB medium under 2 mM IPTG induction, the yield was 16.9 μg/mL dsRNA using the phenol: chloroform method (Mo et al., 2022). The search for alternative and low-cost culture media, combined with the optimization of fermentation bioprocesses in bioreactors, plays a fundamental role in increasing the productivity of the HT115/pCloneVR_2 production system and other systems, consequently leading to a reduction in production costs.

In addition to the production method, the extraction and purification of dsRNA are crucial for the application of RNAi techniques in target organisms. Although crude bacterial extracts containing dsRNA have been used for gene silencing of viruses and plants (Lau et al., 2015; Tenllado et al., 2003), pure dsRNA molecules are more often used in commercial product formulations and agricultural applications (Nwokeoji et al., 2017).

Obtaining dsRNA samples free of genomic DNA and ssRNA, preserving their integrity, and minimizing losses during the extraction process are challenging because yield measurements can be affected by the inefficiency of the extraction and purification methods used. The low yield of ethanol-extracted dsRNA can be attributed to the non-aggressive nature of the method concerning the cell wall, releasing only part of the dsRNA contained inside bacterial cells. Ethanol interacts with the phospholipid bilayer of cell membranes and the bonds between peptidoglycans, destabilizing and increasing membrane fluidity, which leads to subsequent leakage of dsRNA molecules into the buffer (Posiri et al., 2013).

Methods based on cell lysis by heating followed by phenol addition are more efficient at breaking the bacterial cell wall. This method increases the yield of total RNA extracted from cells (Ahn et al., 2019; Verdonckt and Vanden Broeck, 2022). The yield of dsRNA is influenced by the method of pre-treatment of cells, and sonication was found to be the most efficient, yielding 19.5 µg/mL dsRNA, although sonication was observed to cause dsRNA degradation (Ahn et al., 2019).

The cost of producing dsRNA is one of the limiting factors for the use of the RNAi technique in agriculture. The use of the optimized method in this study reduced the production costs of dsRNA by 73%, representing a significant improvement in the process. The production of dsRNA using commercial kits can cost up to US$3 000 per 10 mg using the MEGAscript™ RNAi Kit (Thermo Scientific) (Verdonckt and Vanden Broeck, 2022), which applies only in initial studies. However, several studies have developed strategies to reduce the cost of dsRNA production (Ahn et al., 2019; Das and Sherif, 2020; Guan et al., 2021; Hashiro et al., 2021; Nwokeoji et al., 2022; Papić et al., 2018; Posiri et al., 2013; Voloudakis et al., 2015). Low-cost methods for the production and purification of dsRNA will undoubtedly contribute to industrial and large-scale dsRNA production. Over the years, there has been a noticeable reduction in the costs of dsRNA. In 2008, 1 g of dsRNA cost US$12 000 (Zotti et al., 2018). However, costs have decreased through process optimization and the development of new platforms. For instance, RNAGri, a biotech startup, can produce dsRNA for US$1 per gram using fermentation (Guan et al., 2021). Greenlight Biosciences has successfully lowered the cost to US$0.5 per gram by employing a cell-free method (Maxwell et al., 2018; Taning et al., 2020). Therefore, cost reduction is important both for laboratories that are carrying out research and development of RNAi-based products and for future applications in agriculture.

RNAi-based products for agricultural use could start a transformation in agriculture by offering a highly targeted and sustainable approach to pest, pathogen, and weed control. However, the production of large amounts of dsRNA remains one of the bottlenecks for the development of these products. The in vivo production of dsRNA (bioprocess) and the subsequent extraction and purification steps pose challenges for scaling up dsRNA production. With technological advances driving progress in these areas, the application of RNAi-based products in agriculture holds promise for increasing crop yield and promoting sustainable agricultural practices worldwide.

Supplementary Material

kuae028_Supplemental_File

Funding

This work was supported by Empresa Brasileira de Pesquisa Agropecuária (Embrapa), INCT BioSyn (National Institute of Science and Technology in Synthetic Biology) and CNPq (National Council for Scientific and Technological Development). Juliana da Rosa was supported by the Coordination for the Improvement of Higher Education Personnel (CAPES) through the award of a doctoral scholarship (financing Code 001). Elibio Leopoldo Rech is supported by Embrapa Genetic Resources and Biotechnology/National Institute of Science and Technology in Synthetic Biology, National Council for Scientific and Technological Development (465603/2014-9), Research Support Foundation of the Federal District (0193.001.262/2017), and Coordination for the Improvement of Higher Education Personnel.

Conflicts of Interest

The authors declare no conflicts of interest.
==== Refs
References

Ahn  S.-J., Donahue  K., Koh  Y., Martin  R. R., Choi  M.-Y. (2019). Microbial-based double-stranded RNA production to develop cost-effective RNA interference application for insect pest management. International Journal of Insect Science, 11 , 1–8. 10.1177/1179543319840323
Baulcombe  D.A. (2004). RNA silencing in plants. Nature, 431 , 356–363. 10.1007/978-3-642-04670-4_15 15372043
Chen  J., Peng  Y., Zhang  H., Wang  K., Zhao  C., Zhu  G., Reddy Palli  S., Han  Z. (2021). Off-target effects of RNAi correlate with the mismatch rate between dsRNA and non-target mRNA. RNA Biology, 18 , 1747–1759. 10.1080/15476286.2020.1868680 33397184
Chen  Z., He  J., Luo  P., Li  X., Gao  Y. (2019). Production of functional double-stranded RNA using a prokaryotic expression system in Escherichia coli. Microbiologyopen, 8 , 1–8. 10.1002/mbo3.787
Dalakouras  A., Wassenegger  M., Dadami  E., Ganopoulos  I., Pappas  M., Papadopoulou  K. K. (2019). Genetically modified organism-free RNA interference: Exogenous application of RNA molecules in plants. Plant Physiology, 182 , 38–50. 10.1104/pp.19.00570 31285292
Das  P. R., Sherif  S. M. (2020). Application of exogenous dsRNAs-induced RNAi in agriculture: Challenges and triumphs. Frontiers in Plant Science, 11 , 1–12. 10.3389/fpls.2020.00946 32117356
Delgado-Martín  J., Velasco  L., (2021). An efficient dsRNA constitutive expression system in Escherichia coli .  Applied Microbiology and Biotechnology, 105 , 6381–6393. 10.1007/s00253-021-11494-6 34415390
Ding  S. W. (2010). RNA-based antiviral immunity. Nature Reviews Immunology, 10 , 632–644. 10.1038/nri2824
García  K., Ramírez-Araya  S., Díaz  Á., Reyes-Cerpa  S., Espejo  R., Higuera  G., Romero  J. (2015). Inactivated E. coli transformed with plasmids that produce dsRNA against infectious salmon anemia virus hemagglutinin show antiviral activity when added to infected ASK cells. Frontiers in Microbiology, 6 , 1–11. 10.3389/fmicb.2015.00300 25653648
Guan  R., Chu  D., Han  X., Miao  X., Li  H. (2021). Advances in the development of microbial double-stranded RNA production systems for application of RNA interference in agricultural pest control. Frontiers in Bioengineering and Biotechnology, 9 , 1–7. 10.3389/fbioe.2021.753790
Hashiro  S., Chikami  Y., Kawaguchi  H., Krylov  A. A., Niimi  T., Yasueda  H. (2021). Efficient production of long double-stranded RNAs applicable to agricultural pest control by Corynebacterium glutamicum equipped with coliphage T7-expression system. Applied Microbiology and Biotechnology, 105 , 4987–5000.34097118
Hausjell  J., Kutscha  R., Gesson  J. D., Reinisch  D., Spadiut  O. (2020). The effects of lactose induction on a plasmid-free E. coli T7 expression system. Bioengineering, 7 , 1–12. 10.3390/bioengineering7010008
Hendrix  B., Hoffer  P., Sanders  R., Schwartz  S., Zheng  W., Eads  B., Taylor  D., Deikman  J. (2021). Systemic gfp silencing is associated with high transgene expression in Nicotiana benthamiana. PLoS One, 16 , 1–15. 10.1371/journal.pone.0245422
Khani  M., Bagheri  M. (2020). Protein expression and purification skimmed milk as an alternative for IPTG in induction of recombinant protein expression. Protein Expression Purification, 170 , 1–4. 10.1016/j.pep.2020.105593
Kram  K. E., Finkel  S. E. (2015). Rich medium composition affects Escherichia coli survival, glycation, and mutation frequency during long-term batch culture. Applied and Environmental Microbiology, 81 , 4442–4450. 10.1128/AEM.00722-15 25911475
Lau  S. E., Schwarzacher  T., Othman  R. Y., Harikrishna  J. A. (2015). dsRNA silencing of an R2R3-MYB transcription factor affects flower cell shape in a Dendrobium hybrid. BMC Plant Biology, 15 , 1–14. 10.1186/s12870-015-0577-3 25592487
Ma  Z. Z., Zhou  H., Wei  Y. L., Yan  S., Shen  J. (2020). A novel plasmid–Escherichia coli system produces large batch dsRNAs for insect gene silencing. Pest Management Science, 76 , 2505–2512. 10.1002/ps.5792 32077251
Maxwell  B., Boyes  D., Tang  J., Rodrigues  T., Desai  S., Cunningham  D., Ramachandriya  K., Abshire  J., Cobb  C. (2018). Enabling the RNA revolution, cell-free dsRNA production and control of Colorado potato beetle. [WWW Document]. URL http://www.globalengage.co.uk/pgc/docs/PosterMaxwell.pdf (accessed 2022, January 3).
Mitter  N., Worrall  E. A., Robinson  K. E., Xu  Z. P., Carroll  B. J. (2017). Induction of virus resistance by exogenous application of double-stranded RNA. Current Opinion in Virology, 26 , 49–55.28778033
Mo  Q., Lv  B., Sun  Y., Wu  X., Song  L., Cai  R., Tang  X. (2022). Screening and production of dsRNA molecules for protecting Cucumis sativus against Cucumber mosaic virus through foliar application. Plant Biotechnology Reports, 16 , 409–418. 10.1007/s11816-022-00750-4
Naito  Y., Yamada  T., Matsumiya  T., Ui-Tei  K., Saigo  K., Morishita  S. (2005). dsCheck: Highly sensitive off-target search software for double-stranded RNA-mediated RNA interference. Nucleic Acids Research, 33 , 589–591. 10.1093/nar/gki419
Neubauer  P., Hoffmann  K, Holst  O, Mattiasson  B., Kruschke  P. (1992). Maximizing the expression of a recombinant gene in Escherichia coli by manipulation of induction time using lactose as inducer. Applied Microbiology and Biotechnology, 36 , 739–744.1369364
Nwokeoji  A. O., Kilby  P. M., Portwood  D. E., Dickman  M. J. (2016). RNASwift: A rapid, versatile RNA extraction method free from phenol and chloroform. Analytical Biochemistry, 512 , 36–46. 10.1016/j.ab.2016.08.001 27495141
Nwokeoji  A. O., Kung  A. W., Kilby  P. M., Portwood  D. E., Dickman  M. J. (2017). Purification and characterisation of dsRNA using ion pair reverse phase chromatography and mass spectrometry. Journal of Chromatography A, 1484 , 14–25. 10.1016/j.chroma.2016.12.062 28088361
Nwokeoji  A. O., Nwokeoji  E. A., Chou  T., Togola  A. (2022). A novel sustainable platform for scaled manufacturing of double-stranded RNA biopesticides. Bioresources and Bioprocessing, 9 , 1–16. 10.1186/s40643-022-00596-2 38647831
Ongvarrasopone  C., Roshorm  Y., Panyim  S. (2007). A simple and cost effective method to generate dsRNA for RNAi studies in invertebrates. Nucleic Acids Research, 33 , 35–39. 10.2306/scienceasia1513-1874.2007.33.035
Papić  L., Rivas  J., Toledo  S., Romero  J. (2018). Double-stranded RNA production and the kinetics of recombinant Escherichia coli HT115 in fed-batch culture. Biotechnology Reports, 20 , 10–13. 10.1016/j.btre.2018.e00292
Posiri  P., Ongvarrasopone  C., Panyim  S. (2013). A simple one-step method for producing dsRNA from E. coli to inhibit shrimp virus replication. Journal of Virological Methods, 188 , 64–69. 10.1016/j.jviromet.2012.11.033 23247053
Taning  C. N. T., Arpaia  S., Christiaens  O., Dietz-Pfeilstetter  A., Jones  H., Mezzetti  B., Sabbadini  S., Sorteberg  H. G., Sweet  J., Ventura  V., Smagghe  G. (2020). RNA-based biocontrol compounds: Current status and perspectives to reach the market. Pest Management Science, 76 , 841–845. 10.1002/ps.5686 31743573
Tenllado  F., Martínez-garcía  B., Vargas  M., Díaz-ruíz  J. R. (2003). Crude extracts of bacterially expressed dsRNA can be used to protect plants against virus infections. BMC Biotechnology [Electronic Resource], 11 , 1–11.
Thammasorn  T., Sangsuriya  P., Meemetta  W., Senapin  S. (2015). Large-scale production and antiviral efficacy of multi-target double-stranded RNA for the prevention of white spot syndrome virus (WSSV) in shrimp. BMC Biotechnology [Electronic Resource], 15 , 5–8. 10.1186/s12896-015-0226-9 25888095
Timmons  L., Court  D. L., Fire  A. (2001). Ingestion of bacterially expressed dsRNAs can produce specific and potent genetic interference in Caenorhabditis elegans. Gene, 263 , 103–112. 10.1016/S0378-1119(00)00579-5 11223248
Timmons  L., Fire  A. (1998). Specific interference by ingested dsRNA. Nature, 395 , 854–854.9804418
Verdonckt  T. W., Vanden Broeck  J. (2022). Methods for the cost-effective production of bacteria-derived double-stranded RNA for in vitro knockdown studies. Frontiers in Physiology, 13 , 1–15. 10.3389/fphys.2022.836106
Voloudakis  A. E., Holeva  M. C., Peter Sarin  L., Bamford  D. H., Vargas  M., Poranen  M. M., Tenllado  F. (2015). Efficient double-stranded rna production methods for utilization in plant virus control. Methods in Molecular Biology, 1236 , 255–274. 10.1007/978-1-4939-1743-3_19 25287509
Zotti  M., dos Santos  E. A., Cagliari  D., Christiaens  O., Taning  C. N. T., Smagghe  G. (2018). RNA interference technology in crop protection against arthropod pests, pathogens and nematodes. Pest Management Science, 74 , 1239–1250. 10.1002/ps.4813 29194942
