The western corn rootworm (WCR), Diabrotica virgifera virgifera LeConte, is an annual pest of maize in the United States Corn Belt. Larval feeding on the root system can promote significant yield loss through reduced water and nutrient uptake and decreased plant stability. Various management tactics, including crop rotation, insecticides, and transgenic crops expressing Bacillus thuringiensis Berliner proteins, have been used to manage WCR densities. However, resistance has evolved to each of these tactics in local areas, highlighting the need for new management strategies. The use of RNA interference (RNAi) technology for WCR management represents the next phase of species-specific pest management.
1. Introduction
The western corn rootworm (WCR),
Diabrotica virgifera virgifera LeConte (Coleoptera: Chrysomelidae), is one of the most destructive insect pests of maize (
Zea mays L.) in the United States (U.S.) [
1,
2]. Native to Central America and first identified as a pest of cultivated maize in Colorado in 1909 [
3,
4], populations are now found throughout the midwestern U.S. and Europe [
5,
6]. Once achieving pest status, various control tactics have been used to reduce damage caused by WCR larval feeding including crop rotation, soil-applied insecticides, and maize hybrids expressing insecticidal proteins from the soil bacterium
Bacillus thuringiensis Berliner (
Bt) [
2,
7,
8]. However, managing WCR has been historically challenging due to its remarkable ability to evolve resistance to all available management tactics throughout various local areas in the U.S. Corn Belt [
2,
7,
9,
10,
11,
12,
13,
14]. Economic analyses estimate costs associated with control strategies and yield loss exceed USD $2 billion annually [
1,
15]. Four insecticidal
Bt proteins are currently available for WCR management: Cry3Bb1, mCry3A, eCry3.1Ab, and Cry34Ab1/Cry35Ab1 (now classified as Gpp34Ab1/Tpp35Ab1 [
16]) [
17,
18,
19,
20]. Field-evolved resistance to Cry3Bb1 and mCry3A was first reported in 2011 [
21] and has subsequently been confirmed in various areas of the U.S. Corn Belt [
22,
23,
24,
25,
26,
27]. Cross-resistance between Cry3Bb1, mCry3A, and eCry3.1Ab has been widely demonstrated [
21,
23,
25,
26,
28]. More recently, resistance to Cry34Ab1/Cry35Ab1 was identified [
27,
29,
30], highlighting the urgency for alternative approaches for WCR management. Recent review articles highlight the history, use of, and evolution of resistance to synthetic insecticides [
7] and
Bt traits [
8].
In planta expression of double-stranded RNA (dsRNA) in the shape of a hairpin RNA triggers an RNA interference (RNAi) response within the insect, representing a new mode of action for WCR control [
31,
32]. In June 2017, the U.S. Environmental Protection Agency (EPA) registered the first transgenic maize product with an RNAi-based plant-incorporated protectant (PIP) for WCR management [
33]. This product expresses three
Bt proteins (Cry3Bb1 and Cry34Ab1/Cry35Ab1) and a dsRNA [
32].
2. Field Efficacy of RNAi for Insect Control, Insect Resistance Management, and RNAi Resistance
2.1. Field Efficacy of RNAi Traits
The U.S. EPA approved the first RNAi product for insect control in 2017 [
146], representing the first new MOA for WCR control since the release of the Cry34Ab1/Cry35Ab1 binary protein in 2005 [
147]. As previously indicated, this product (SmartStax
® PRO) expresses three rootworm-active
Bt proteins, Cry3Bb1 and Cry34Ab1/Cry35Ab1, as well as
DvSnf7 dsRNA [
33]. SmartStax
® PRO also contains three Lepidopteran-active
Bt proteins (Cry1A.105/Cry2Ab2 and Cry1F) and genes for glyphosate tolerance [
33]. Given reports of field-evolved resistance to Cry3Bb1 [
21,
22,
23,
26,
30], an efficacy evaluation of this product on Cry3Bb1-resistant insects and field populations was performed [
148]. Additionally,
DvSnf7 dsRNA concentration-response larval bioassays were conducted on artificial diet using field-derived populations, a susceptible laboratory population, and a Cry3Bb1-resistant population. Lastly, greenhouse experiments evaluating beetle emergence from plants expressing each trait individually (i.e., Cry3Bb1, Cry34Ab1/Cry35Ab1, and
DvSnf7) and in combination were conducted. The Cry3Bb1-resistant population exhibited a significant 2.7-fold decrease in susceptibility to
DvSnf7 dsRNA compared to the Cry3Bb1-susceptible population. However, the Cry3Bb1-resistant population lowered susceptibility was similar to other WCR field populations in diet bioassays. Cry3Bb1-resistant and susceptible colonies had similar WCR adult emergence from plants expressing
DvSnf7 and Cry34Ab1/Cry35Ab1, and WCR adult emergence from plants expressing
DvSnf7 and
DvSnf7 + Cry3Bb1 was not significantly different when the Cry3Bb1-resistant colony was evaluated [
148]. Collectively, diet assay and
in planta experiments have demonstrated a lack of cross-resistance between Cry3Bb1, Cry34Ab1/Cry35Ab1, and
DvSnf7 [
148]. Previous research has also documented significant variation in susceptibility of WCR larvae from field populations to
DvSnf7, with an LC
50 ranging from 4.07 to 40.51 ng/cm
2 [
148]. This suggests that some populations might already exhibit higher tolerance to
DvSnf7 dsRNA and resistance monitoring will be essential to track susceptibility changes in field populations to promote trait durability.
Further field studies evaluated the efficacy of SmartStax
® (Cry3Bb1 + Cry34Ab1/35Ab1 pyramid) and SmartStax
® PRO maize (also expresses
DvSnf7 dsRNA) against western and northern corn rootworms [
32]. Field trials conducted between 2013 and 2015 across the U.S. Corn Belt demonstrated that SmartStax
® PRO could significantly reduce root damage ratings under high WCR larval densities, in areas of Cry3Bb1 resistance, and in areas with greater than expected injury to Cry3Bb1 or SmartStax
®. SmartStax
® PRO also significantly reduced root damage ratings relative to SmartStax
® in some field trials, indicating that
DvSnf7 dsRNA can provide additional root protection when coupled with
Bt proteins [
32]. In addition to enhanced root protection, SmartStax
® PRO also reduced adult emergence compared to SmartStax
®, single-event Cry3Bb1 and Cry34Ab1/Cry35Ab1 hybrids, and the non-
Bt control [
32]. Collectively, the reduced root damage ratings and decreased adult emergence associated with SmartStax
® PRO suggest that the addition of
DvSnf7 and other RNAi traits could serve as a valuable tool for insect resistance management (IRM) strategies [
32]. Various resistance modeling scenarios indicated that inclusion of
DvSnf7 dsRNA could promote durability of the expressed
Bt proteins and decrease the rate of resistance evolution relative to SmartStax
®, even in areas with suspected resistance to Cry3Bb1 [
32]. However, no work has been performed to evaluate the role of RNAi traits under scenarios with resistance to both Cry3Bb1 and Cry34Ab1/Cry35Ab1. This product and other products with RNAi traits will likely play an important role in WCR population management, especially in locations with high annual WCR densities and confirmed resistance to Cry3Bb1 or Cry34Ab1/Cry35Ab1 [
32].
2.2. Insect Resistance Management
The U.S. EPA requires that registrants of PIPs, such as
Bt proteins and RNAi traits, complete and submit an IRM plan for the target pest before registration [
149]. IRM is the scientific approach to delay the development of resistance in pest populations. SmartStax
® PRO follows the IRM pyramiding strategy and expresses three different MOAs targeting WCR (
DvSnf7 dsRNA, Cry3Bb1 and Cry34Ab1/Cry35Ab1), and stacks traits against Lepidopteran pests (Cry1F, Cry2Ab2, and Cry1A.105) and weeds (cp4/epsps (glyphosate resistance)). The IRM value of the pyramid is significantly reduced if individual components are deployed simultaneously, field-evolved resistance to one or more components is present, and/or cross-resistance between components is observed [
150]. Confirmed WCR field-evolved resistance to Cry3Bb1 and/or Cry34Ab1/Cry35Ab1 has been reported in some populations [
21,
23,
24,
26,
27,
28,
29,
30], which could potentially impact the IRM value of this new pyramid. After this product is commercially available, it will be essential to monitor changes in susceptibility to the three traits, particularly Cry3Bb1 and Cry34Ab1/Cry35Ab1, to ensure the durability of all traits in SmartStax
® PRO.
An important consideration for IRM with RNA-based traits is that, in contrast to
Bt proteins, larvicidal dsRNAs can cause mortality in adult WCR [
75,
126,
151,
152]. The adult WCR RNAi response is rapid and can persist throughout most of the life stage. For example, knockdown of
Lac2 generated 76% knockdown 10 hours after ingestion and 86% knockdown 20 days after ingestion [
153]. Expression of
DvSnf7 dsRNA occurs throughout the plant (event MON87411), including two tissues commonly consumed by adults in the field (e.g., pollen, leaves). However, the concentrations expressed
in planta are not sufficient to generate mortality in adults and will provide sublethal exposure (mean of 0.103 ng/g and 33.8 ng/g in fresh weight pollen and leaf tissue, respectively [
154]; the LC
50 of
DvSnf7 previously observed in WCR adults was 60.2 ng/cm
2 [
151]). Therefore, adult sublethal exposure to
DvSnf7 may have implications for resistance management by adding selection pressure benefiting resistant individuals or individuals with resistance alleles [
155,
156]. Movement of adult WCR is common as maize phenology changes, with beetles searching for underdeveloped silks and pollinating maize as a primary food source [
148]. Intrafield adult WCR movement could increase the likelihood of many individuals emerging from a single commercial maize field feeding on SmartStax
® PRO tissue at some point during the life cycle, while interfield movement of adult WCR and subsequent feeding on these plants would increase the risk of exposure to sublethal concentrations of dsRNA [
157]. No fitness costs were observed in WCR adults exposed to
DvSec23 dsRNA LC
25, suggesting that exposure to a sublethal concentration may not affect the fitness of exposed adults and their offspring [
126]. However, due to the unique physiological effects of each dsRNA trait, future studies are important to determine the role sublethal exposure to
DvSnf7 field-relevant concentrations will have on adult WCR.
Reports of
Bt trait field-evolved resistance suggested that the high dose refuge (HDR) approach as a standalone IRM strategy was ineffective at delaying
Bt resistance in WCR. The most recent draft of the U.S. EPA’s “Framework to Delay Corn Rootworm Resistance” requires incorporating integrated pest management (IPM) strategies into IRM plans to mitigate the spread of field-evolved resistance and minimize the risk of resistance evolution [
158]. Suggested IPM strategies include crop rotation, rotation of PIP MOAs, adulticide application, and area-wide management [
158]. Deploying this new dsRNA product within an IPM framework is necessary to increase trait durability and decrease the rate of resistance evolution [
159].
2.3. Rootworm Resistance to RNAi
As in previous
Bt PIPs, selection pressure from continuous exposure to dsRNA will eventually promote resistance evolution. Mechanisms of resistance to dsRNA were initially postulated to include degradation of dsRNA in the gut, reduced dsRNA uptake, alteration in proteins involved in dsRNA transport or formation of the RISC complex, loss of siRNA recognition by the RISC complex, mutation of the target gene, or systemic spread failure [
76,
134,
160]. Identifying the resistance mechanism to
DvSnf7 dsRNA in WCR will inform effective IRM strategies to extend product durability if resistance is related to dsRNA processing, which may confer cross-resistance to other RNAi traits.
Efforts to evaluate the multigenerational effect of
DvSnf7 in field-collected insects resulted in the development of a resistant colony. Khajuria et al. [
75] collected WCR adults emerging from areas planted with transgenic maize expressing
DvSnf7 dsRNA. Field-collected beetles were crossed with a non-diapausing WCR colony and exposed to
DvSnf7 dsRNA for eight generations, creating a population with ≥130-fold resistance to
DvSnf7 dsRNA [
75]. Reciprocal crosses determined that
DvSnf7 dsRNA resistance was recessive, monogenic, and autosomal [
75]. Resistance resulted from reduced uptake of dsRNA in gut cells and, interestingly, was found to be non-sequence specific. Cross-resistance to dsRNAs targeting
v-ATPase A,
COPI B (Coatomer Subunit beta), and
mov34 (26s proteasome) was identified, indicating adaptation occurred within the RNAi mechanism itself. Due to the development of cross-resistance to a variety of dsRNA targets in this study, it is possible dsRNA represents a single MOA in WCR [
75]. However, it is yet unknown how insects will respond and adapt to dsRNA under field conditions. Further studies with other RNAi traits will provide a better understanding of potential resistance mechanisms that might exist in the field.
Because
DvSnf7 will not be released as a single trait product, it was important to identify the chromosomal location of resistance alleles and compare with current knowledge on the location of resistance alleles to the
Bt traits in SmartStax
® PRO. The Cry3Bb1 resistance gene(s) is located on linkage group 8 (LG8) [
161] and the resistance locus for
DvSnf7 dsRNA is located on linkage group 4 (LG4) [
75]. The fact that resistance causative elements are located on different chromosomes supports experimental evidence showing a lack of cross-resistance between
DvSnf7 dsRNA and Cry3Bb1 [
75,
148]. Therefore, resistance to both traits would have to develop independently and would have a lower probability of occurring than if a single trait or MOA were used.
Similar work in
L. decemlineata furthers our understanding of how dsRNA resistance could develop in WCR. A laboratory-derived dsRNA-resistant
L. decemlineata cell line and colony were recently established [
105,
162]. In both cases, cross-resistance to multiple dsRNAs was found, supporting results observed in WCR [
75]. In dsRNA-resistant WCR, non-specific dsRNA uptake was disrupted, conferring resistance to all sequences tested. This fits with the transcriptional analysis of
L. decemlineata cells, wherein genes related to uptake,
clathrin light chain,
vha55,
silA, and the novel dsRNA binding protein
staufenC, were downregulated in resistant cells [
105]. However, while resistance in the WCR colony was narrowed to one locus in the genome, the
L. decemlineata dsRNA-resistant colony displayed polygenic inheritance. This indicates that various mechanisms relating to dsRNA uptake could adapt to intense selection pressure. Given that resistance alleles have been detected in natural WCR field populations and those detected to date may confer cross-resistance to other dsRNA sequences, monitoring changes of susceptibility of field populations will be necessary to preserve RNAi technology. The deployment of
DvSnf7 with two additional functional MOAs will also assist in delaying RNAi trait resistance in populations that remain susceptible to Cry3Bb1 and Cry34Ab1/Cry35Ab1.
3. RNAi Mammalian Safety
When evaluating human safety considerations for ingested RNA molecules associated with the use of RNA-mediated transgenic crops, one must consider the following: (1) the history of safe consumption of RNA; (2) biological barriers that limit internal exposure to exogenous RNAs; and (3) the weight of scientific evidence from published safety studies with RNA molecules. Collectively, these considerations enable proper hazard identification and, taken together with exposure assessment, can be leveraged to make an informed risk assessment decision on the use of transgenic maize expressing RNAi traits as a control tactic for corn rootworms.
3.1. RNA and Its History of Safe Consumption
RNAi is a ubiquitous process for gene expression modulation in all eukaryotic organisms. Therefore, longer dsRNA precursors and the small RNA molecules that this process leverages to regulate gene expression are also ubiquitous in widely consumed foods derived from plants and animals. Owing to the ubiquity of these RNA molecules, small RNAs with perfect sequence complementarity to transcripts of human and/or animal genes are evident in soybean, rice, maize, and fruits and vegetables [
195,
196,
197]. These sequences include those important for key biological processes in mammals [
196]. The presence of such sequences in staple food and feed crops demonstrates the safe consumption of small RNAs in the diet and supports the safety of RNAi for uses in transgenic crops, including those intended for control of agricultural pests such as corn rootworms.
Leveraging RNAi for insect control in a commercial transgenic crop represents a new application of this mechanism; however, RNAi has served as a natural process underlying plant phenotypes in domesticated crops and as a MOA in commercially approved transgenic traits (reviewed by Petrick et al. [
198] and Sherman et al. [
199]). One prominent example of RNA-mediated traits successfully leveraged in commercial transgenic products is the deployment of resistance to the papaya ringspot virus in papaya [
200], a transgenic trait that played a vital role for Hawaiian growers amidst the devastation of the papaya crop. RNAi has also been leveraged commercially for the production of healthier oils in soybean [
201,
202], for reduced browning in apples [
203], and for the reduction of acrylamide and blackspot bruising in potatoes [
204,
205]. Transgenic crops utilizing RNAi as a MOA have received regulatory approvals in several geographies. These include approvals for use in food; feed; and cultivation in crops such as maize, soybean, squash, potato, tomato, alfalfa, plum, apple, bean, and papaya [
206]. The above information on RNA-safe consumption in the diet and its safe use to date in commercialized transgenic crops with RNAi-based traits should also be applicable to those RNA-based traits intended for control of corn rootworms and other insects [
55,
129,
145].
3.2. Biological Barriers to RNA Absorption
Owing to the ubiquitous presence of the RNAi mechanism in eukaryotic organisms, RNA molecules are ingested by vertebrates through the consumption of plants, animals, and fungi. Such RNA molecules include those with double-stranded regions that could initiate RNAi if they were to be absorbed from the diet and reach cells following consumption by the organism. As a further illustration of this point, exogenous dietary RNAs include those with sequence complementarity to vertebrate genes [
195,
196,
197,
207,
208]. As vertebrates are constantly exposed to such RNA molecules, it is not surprising that there are a series of biological barriers that preclude these dietary components from serving in a regulatory capacity, and instead, they are harnessed for nutritional value. These biological barriers are reviewed in the scientific literature [
198,
199,
209].
Ingested RNAs face an acidic digestive environment in the stomach following their initial exposure to nucleases in the saliva [
210]. The efficacy of the low pH and hydrolysis in the stomach for nucleic acid degradation and the removal of bases from the nucleic acid backbone (e.g., depurination) has been well described and reviewed in the literature [
198,
209,
211]. Further digestion of ingested nucleic acids occurs in the small intestine due to the presence of digestive enzymes and nucleases secreted by the pancreas [
209]. Due to this extensive collection of barriers, nucleic acids from the diet are extensively degraded and do not undergo substantive absorption in an intact form, as demonstrated empirically with miRNAs in rodents [
211,
212], rhesus monkeys [
213], and humans [
214]. Thorough reviews of the safety considerations of plant expressed and externally applied RNA molecules for humans and other vertebrates have been published [
198,
215].
Biological barriers to the absorption and function of ingested nucleic acids expand beyond digestive barriers and include cellular membranes impermeable to RNAs that are both highly polar and large [
198,
216]. Each successive series of membrane barriers must be crossed for a dietary RNA to move from the intestinal tract lumen, across endothelial cells, and into the bloodstream. Once in the blood, RNAs are then subjected to nucleases [
217,
218,
219] and rapid renal elimination [
217,
220]. To reach a putative target tissue, any RNA molecule escaping these nucleases would then have to cross through the endothelium and the epithelium within a target tissue to have the capacity to modulate tissue gene expression. Owing to the impermeability of these charged macromolecules across membrane barriers [
198,
216], any RNA surviving the intestinal tract would be unlikely to reach the bloodstream or target tissue. Furthermore, RNAs reaching the cytoplasm of a cell in the consuming organism would subsequently be subjected to sequestration into endosomes that retain a vast majority of exogenous RNA molecules [
221,
222].
This array of physical, chemical, and biological barriers has made the development of pharmaceutical RNA drugs challenging, necessitating their chemical stabilization to limit degradation and specialized delivery formulations to elude barriers to exogenous RNA molecules [
209,
221,
223,
224]. Without such formulations, injection of these exogenous RNA molecules results in their rapid degradation and elimination [
217,
219,
220,
225]. Oral delivery of therapeutic macromolecules represents a desirable route of administration that is even more elusive to drug developers than systemically or locally administered therapeutics due to the aforementioned biological barriers [
221,
224]. These challenges are evident from literature reviews and studies demonstrating the need for specialized formulations and/or chemical modifications to facilitate limited delivery within proof-of-concept oral delivery evaluations of RNA therapeutics [
209,
223,
226,
227].
3.3. Evaluation of Potential Activity or Adverse Effects of Ingested RNA
The biological barriers described above collectively limit absorption of RNA molecules from the diet. Therefore, it is highly unlikely that ingested RNAs would have the capacity to regulate gene expression or induce adverse effects in a consuming mammalian organism. This concept stems primarily from a lack of significant RNA oral bioavailability [
211,
212], rendering the internal dose, e.g., the number of available copies of a given RNA molecule at the putative site of action, insufficient for regulatory function [
214,
219,
228]. This is made especially challenging as uptake of RNA molecules in mammals is low (e.g., less than one copy per cell), and up to 1000 to 10,000 RNA copies per cell may be required for a functional RNAi response [
229]. Further complicating the possibility of plant RNA activity in the diet is that small RNAs in plants are tightly bound to Argonaute proteins. Bound RNAs are not thought to dissociate from these complexes or exchange into Argonaute proteins in the consuming organism [
228]. These principles collectively provide a solid biological basis for the history of safe dietary RNA consumption described above, even when ingested sequences possess sequence complementarity to transcripts in the consuming organisms.
The weight of evidence supporting the limited potential for functional activity ingested RNAs in mammals [
230,
231,
232,
233] has been called into question by several peer-reviewed publications alleging the opposite. However, a series of reviews and several laboratory studies on absorption and/or activity of ingested RNAs have been published since 2012, and collectively calls into question the potential for significant uptake and bioactivity of ingested RNAs [
197,
211,
212,
213,
214,
215,
228,
234,
235,
236,
237]. For example, work challenging the concept of dietary RNA absorption/activity indicated that very low detection levels could have resulted from laboratory contamination or false-positive PCR results [
213,
238,
239]. Furthermore, it is essential to ensure nutritionally balanced rodent diets when conducting rodent dietary studies. This is evidenced by one of the principal studies asserting uptake and activity of ingested RNAs in mammals [
232]. Measured changes in blood cholesterol concomitant with dietary RNA detection in plasma were ultimately determined to result from dietary imbalances rather than ingested RNA activity [
212]. The most comprehensive analysis of this phenomenon included an assessment of 800 human datasets, the results of which supported the conclusion that detection of small RNAs from exogenous sources in mammalian blood samples likely results from contamination as opposed to dietary uptake [
235].
Toxicological evaluations of orally administered double-stranded RNAs have been conducted in mice to address the potential for oral activity and toxicity of these molecules. A 28-day repeated-dose oral toxicity study was conducted in mice with either a long dsRNA molecule or a pool of four 21 base pair siRNAs targeting the mouse
v-ATPase gene. This proof of concept study for evaluating oral activity/toxicity of RNA was conducted utilizing a known gene target for corn rootworm control when expressed in plants [
55] and to then construct a long dsRNA or a pool of predicted active siRNA sequences with 100% sequence complementarity to the mouse [
237]. These mouse-targeting dsRNA sequences were then administered to mice by oral gavage for 28 consecutive days at doses of ≥48 mg/kg body weight, and traditional toxicology and target gene expression endpoints were evaluated. This study did not identify oral toxicity or suppression of
v-ATPase gene expression in the gastrointestinal tract, kidney, liver, brain, or bone, demonstrating that biological barriers appear to preclude oral activity or toxicity of orally administered dsRNA molecules in mammals.
To further demonstrate RNA oral safety in a product-specific context, the
DvSnf7 dsRNA (240 base pair active dsRNA within a 968 nucleotide RNA sequence) was evaluated in a 28-day repeat-dose oral toxicity study [
197].
DvSnf7 dsRNA at oral doses of up to 100 mg/kg bodyweight for 28 consecutive days did not produce any treatment-related effects on weight, food consumption, clinical observations, clinical chemistry, hematology, gross pathology, or histopathology in mice [
197]. Furthermore, the high dose utilized in this study (No-Observed Adverse-Effect-Level (NOAEL) of 100 mg/kg) was billions of times greater than highly conservative estimates of mean per capita human exposure to
DvSnf7 dsRNA in Europe, the U.S., Mexico, China, Japan, and Korea. This study demonstrates the safety of a dsRNA molecule used to control insects, specifically corn rootworms, illustrating further that transgenic crops using dsRNA for insect control do not pose unique risks to the health of consuming mammals, including humans.
This entry is adapted from the peer-reviewed paper 10.3390/insects13010057