Abstract
Introduction:
It is commonplace in East Africa for 100% of cassava fields to be infected with Cassava mosaic disease (CMD) and/or Cassava brown streak disease (CBSD), resulting in annual losses of more than US$1.25 billion and reduced food and economic security for farming households. The vector of both diseases is the African cassava species of the whitefly Bemisia tabaci. Since the late 1990s, there has been an unprecedented increase in whitefly populations, to the extent that they are referred to as “super-abundant”. Research efforts since the late 1990s has focused mainly on developing plant resistance to the viral pathogens and paid scant attention to understanding the root causes of disease epidemics or the control of whitefly infestation.
Methods:
Here, we aimed at developing long-term whitefly-control solutions using an in-planta RNA interference (RNAi) approach. First, transcriptome analysis identified candidate genes that play ‘key’ roles in whitefly biology: osmoregulation, sugar metabolism and transport, symbiosis with endosymbiotic bacteria and detoxification of phytotoxins. Then, fifteen RNAi inverted repeat constructs were produced, designed to target the candidate genes and 140 independent transgenic lines were generated in cassava variety NASE 13.
Results:
Whole plant bioassays showed insecticidal activity of transgenic plants, reaching 58% lethality for adults within 7 days and 75-90% lethality of nymphs after 25 days, compared to control plants. Target genes were confirmed to be downregulated by up to 2.5-fold in adult whiteflies and nymphs.
Discussion:
We used population dynamics modeling to predict the potential of the RNAi technology to control whiteflies under field conditions in East Africa. These analyses indicated that the developed technology offers a realistic option for obtaining durable control of cassava whitefly in African cassava fields.
1 Introduction
Cassava is recognized as a strategic food-security crop for smallholder farmers, particularly in sub-Saharan Africa where it supplies over half of caloric intake for about 200 million people and it is cultivated predominantly on farms smaller than one hectare (Obong’o et al., 2024). The importance of cassava for food security stems from its high carbohydrate yield, tolerance to drought and poor soils, and capacity to remain in the ground for extended periods, which together make it an “insurance crop” that buffers households against climatic and market shocks (Adebayo, 2023). The contribution of cassava to small-holder food security is increasingly constrained by viral diseases, vectored by the cassava-whitefly Bemisia tabaci. Since the 1990s, an unprecedented increase in the abundance of the cassava-whitefly, sub-Saharan Africa 1 (SSA1) and SSA2 species (Mugerwa et al., 2021a), has occurred across the cassava growing regions of East and Central Africa (Colvin et al., 2004; Legg et al., 2014; Macfadyen et al., 2018, Macfadyen et al., 2021). The cassava whitefly is responsible for vectoring the causal viruses of Cassava mosaic disease (CMD) and Cassava brown streak disease (CBSD). CBSD and CMD are the most important diseases of cassava in Africa, causing estimated production losses as high as 47%, equivalent to US$1.25 billion annually (Mugerwa et al., 2021b). In addition to vectoring virus diseases, super-abundant whitefly populations can cause up to 40% yield loss through direct feeding, and from reduced photosynthesis caused by black molds (sooty mold) that develop on leaves as a result of whitefly-produced honeydew (Omongo et al., 2012).
To date, efforts to combat the effects of cassava whitefly have focused mostly on developing cassava varieties with robust resistance to the viral pathogens, with significantly less attention given to host plant resistance against the whitefly itself (Parry et al., 2020; Mukiibi et al., 2019). Both must be achieved if the detrimental impact of this pest is to be negated. Breeding programs have had some success in selecting new varieties with elevated tolerance to whiteflies (Atim et al., 2024; Sam et al., 2024). However, these varieties do not combine elevated tolerance to whiteflies with other agronomic traits, such as high storage root yields, good dry matter content and the organoleptic qualities important for widespread adoption by farmers (Manze et al., 2021). It is also known that whitefly populations evolve to overcome tolerance (Tadmor et al., 2022). New and robust sources of resistance to cassava whiteflies are therefore required for breeders to pyramid with disease resistance in farmer-preferred germplasm.
We report here a biotechnological approach utilizing plant-expressed RNA interference (in-planta RNAi) to target genes involved in multiple processes essential to the cassava-whitefly life cycle. Like other hemipteran insects, whiteflies can ingest dsRNAs from phloem tissues, which trigger RNAi responses in the insect’s gut and downregulate the target genes (Rosa et al., 2018; Zhu and Palli, 2020). Three previous studies have demonstrated the potential of using transgenic, plant-mediated RNAi to control whiteflies. Thakur et al. (2014) observed an enhanced level of B. tabaci resistance in transgenic Nicotiana tabacum plants expressing dsRNA targeting the vacuolar‐type H+‐ATPase (v‐ATPase A) gene. Transcripts levels of v‐ATPase A in B. tabaci feeding on plants expressing this dsRNA was reduced by 62%, leading to 84% adult mortality after six days. Eakteiman et al. (2018) targeted the glutathione S-transferase (GST) detoxification gene, BtGSTs5. B. tabaci feeding on Arabidopsis thaliana plants expressing dsRNA derived from BtGSTs5 showed significant down regulation of this target gene in their gut. An associated delay in nymph development by up to 3.5 days was observed, leading to significant reduction in population build-up. In the third case, Xia et al. (2021) targeted a glucoside malonyltransferase gene (BtPMaT1) which enables B. tabaci to neutralize phenolic glucosides. Adults feeding on transgenic tomato plants significantly reduced BtPMaT1 expression and resulted in almost 100% adult mortality.
Previous genomic and transcriptomics analyses by our group and others (Jing et al., 2016; Luan et al., 2015; Luo et al., 2017; Malka et al., 2018, Malka et al., 2021) enabled us to develop four strategies for insecticidal RNAi activity. These target essential processes and pathways in the cassava-whitefly: osmoregulation, carbohydrate homeostasis, symbiosis and detoxification. First, to impair osmoregulation (ability to tolerate high osmotic pressure of ingested phloem-sap), we targeted an aquaporin (AQP) that mediates water cycling across the gut (Shakesby et al., 2009), and three GH13 α-glucosidases (SUCs) that contribute to osmotic balance via sucrose isomerization/oligomerization (Jing et al., 2016; Luo et al., 2017; Mathew et al., 2011; Wintraube et al., 2025). Second, to downregulate carbohydrate homeostasis and thereby reduce available energy and stress tolerance, we silenced genes encoding UTP-glucose-1-phosphate uridylyltransferase (UGP) and trehalose-6-phosphate synthase (TPS), which are central to UDP-glucose, trehalose, and glycogen biosynthesis (Thompson, 2003), as well as a sugar transporter (ST) required for sugar flux across membranes. Third, because B. tabaci relies on the obligate endosymbiont ‘Candidatus Portiera aleyrodidarum’ for the synthesis of essential nutrients, we targeted five horizontally acquired insect genes expressed in the bacteriocytes: arginosuccinate lyase (argH), chorismate mutase (CM), diaminopimelate decarboxylase (LysA), diaminopimelate epimerase (DapF), and biotin synthase (BioB), involved in the synthesis of arginine, phenylalanine, lysine, and biotin (Kaweesi et al., 2024). Finally, to compromise detoxification of plant defense phytotoxins, we silenced three cassava-whitefly genes coding for detoxification proteins: two UDP-glucosyltransferases (UDPGTs) involved in phase II conjugation and a gene coding for an ATP-Binding Cassette (ABC) transporter that exports neutralized compounds (Després et al., 2007).
Inverted repeat expression cassettes were generated from sequences targeting the above genes and their expression was directed to phloem tissues in transgenic cassava plants using a phloem-specific promoter. Whole plant bioassays enabled assessment of insecticidal impact on adults and nymphs. Data generated was incorporated into population dynamics modeling to assess potential of these RNAi strategies to suppress build-up of cassava whitefly populations in the field under different regional cropping mixtures of RNAi-protected and non-protected varieties.
2 Materials and methods
2.1 dsRNA design
dsRNA sequences were designed using the Snapdragon platform (Hu et al., 2021; https://www.flyrnai.org/cgi-bin/RNAi_find_primers.pl). The dsRNA regions were selected using an siRNA-design workflow in which high-quality candidate 21-mers were first predicted (DSIR), mapped back to the target transcript, and then used to place each dsRNA within regions of higher predicted siRNA density, with preference (when feasible) for regions proximal to the start codon. To screen for inter-species off-targets, all possible overlapping 21-nt k-mers were generated, using a 1-nt sliding window (splitter -size 21 -overlap 20), for each designed dsRNA sequence. Non-target organism datasets (RefSeq-based datasets) were searched with USEARCH against both strands, counting only exact matches (0 mismatches; -maxdiffs 0). A panel of 10 reference species was selected to represent key relevant non-target groups in East African cassava agroecosystems, including humans, domestic animals, pollinators, and the host plant cassava. Specifically, the following RefSeq assemblies were used: Homo sapiens (GCF_000001405.40), Bos taurus (GCF_002263795.3), Sus scrofa (GCF_000003025.6), Canis lupus familiaris (GCF_011100685.1), Ovis aries (GCF_016772045.2), Capra hircus (GCF_001704415.2), Gallus gallus (GCF_016699485.2), Manihot esculenta (GCF_001659605.2), Apis mellifera (GCF_003254395.2) and Frieseomelitta varia (GCF_011392965.1). Across all comparisons, no 21-bp perfect matches were detected in any of reference species, indicating the absence of major off-target candidates. All designed dsRNA sequences are available in Supplementary Table 1.
In parallel, during the working process, we screened for potential non-target effects on parasitoids of whiteflies in Africa, once reliable genomic data became available. Surveys conducted in East Africa indicated that parasitism can be substantial and that Encarsia sophia and Eretmocerus spp. are among the most common parasitoids of whiteflies on cassava (Guastella et al., 2015; Otim et al., 2005). The only relevant chromosome-level genome assemblies available to date are that of Eretmocerus hayati (BioProject PRJNA951478; assembly GCA_029851415.1) and E. sophia (BioProject PRJNA1131600; assembly GCA_041053325.1) published in 2023 and 2024, respectively (Zhong et al., 2023; Man et al., 2024). Of all dsRNAs tested under our screening criteria, only one match (21-nt identity) was identified between E. hayati and the GH13 gene targeted by DWF15 and two matches were identified between E. sophia and the TPS gene targeted by DWF 60 (one 23-nt match) and the UDPGT gene targeted by DWF 66 (one 20-nt match). In future work, these limited matches can be readily mitigated through careful dsRNA design and sequence optimization, allowing selection of target regions with minimal similarity.
2.2 Production of RNAi constructs for plant transformation
RNAi gene constructs DWF10–17 and DWF55–68 (Supplementary Table 2) were generated in which the AtSUC2 promoter drives expression of inverted repeats of sense and antisense sequences separated by the pdk intron (Supplementary Figure 1A) (Rosche et al., 1994). The binary vector p8384, a modified derivate of p5000 (Chauhan et al., 2015), was used as the backbone. Both the cassette and vector were digested with SbfI and AscI (NEB), ligated, and used for cloning. T-DNA cassettes, containing an AtSUC2 promoter, inverted repeat and NOS terminator, were synthesized and cloned into the p5000 vector. The gene cassettes were excised with XhoI/SbfI and the AtSUC2 promoter with XhoI/AscI, p5000 linearized with SbfI/AscI and a three-way ligation performed. All gene constructs were electroporated into Agrobacterium tumefaciens strain LBA4404 and used for genetic transformation of cassava variety NASE 13.
2.3 Production of RNAi transgenic NASE 13 cassava plants
Cassava genetic transformation was performed following Segatto et al. (2022). Organized embryogenic structures (OES) were induced from leaf explants of micropropagated cv NASE 13 cultured on Murashige and Skoog basal medium (Murashige and Skoog, 1962) supplemented with 20 g/l sucrose (MS2) and 50 μM picloram. Friable embryogenic callus (FEC) was generated from OES by sequential culture on Gresshoff and Doy medium (Gresshoff and Doy, 1974) containing 20 g/l sucrose and 50 μM picloram (GD2 50P). A. tumefaciens strain LBA4404 carrying the binary vector of interest was grown to OD600 0.6 and used to infect FEC for 45 min in medium supplemented with 200 µM acetosyringone. After four days co-culture, FECs were transferred to resting medium consisting of GD2 50P supplemented with 125 mg/l cefotaxime and cultured for 14 days. Transgenic callus tissues were recovered by subsequent subculture to GD2 50P medium containing 125 mg/l cefotaxime and 25 μM paramomycin. Growing callus colonies were transferred to MS2 medium supplemented with 5 μM NAA to induce development of mature embryos, after which cotyledon-stage embryos were transferred to MS2 containing 2 µM 6-benzylaminopurine (BAP) to induce germination. Regenerated plantlets were micropropagated and maintained on MS2 medium (Segatto et al., 2022).
2.4 Production and visual assessment of plants transgenic for the GUS reporter
The p8670 binary vector was produced by inserting the AtSUC2 promoter and uidA visual reporter gene into the modified pCAMBIA 2300 backbone which carries the nptII selectable marker gene under control of the 35S promoter (Supplementary Figure 1B). This construct was used to transform FEC of cassava cv. 60444 (Segatto et al., 2022). Leaves from regenerated in-vitro plants, plus hand cut sections of stem and storage roots were obtained from 12-week-old greenhouse-grown plants and used for GUS assays following Jefferson et al. (1987), with modifications. Tissues were incubated in X-Gluc solution (1 mM X-Gluc, 100 mM sodium phosphate buffer pH 7.0, 10 mM EDTA, 0.5 mM potassium ferricyanide, 0.5 mM potassium ferrocyanide, and 0.1% Triton X-100) at 37 °C for 12 h, then cleared in 70% ethanol. Sections were mounted on glass slides and imaged using a Nikon SMZ1500 fluorescence microscope and QImaging Retiga 1300 digital camera.
2.5 Plant establishment and growth in the greenhouse
Micropropagated in-vitro plants were transferred to 7.6-cm pots containing potting soil (Conrad Fafard, Agawam, MA, USA) (Segatto et al., 2022). After acclimatization on a mist bench, plants were grown on the open greenhouse bench under conditions of 28 °C/27 °C day/night temperatures and 70–95% relative humidity. Plants were watered with reverse-osmosis water twice daily, and fertilized twice weekly with Jack’s professional fertilizer at 100 µg/g dry weight.
2.6 Determination of the construct-specific RNA expression in RNAi transgenic plants
Transgenic in vitro plantlets were characterized for construct-specific RNA expression by quantitative real-time PCR (qRT-PCR) as described previously (Narayanan et al., 2021), using the primer pairs listed in Supplementary Table 3. Total RNA was isolated using the Fruitmate kit (Fruitmate™, Takara, Shiga, Japan) per manufacturer instructions, followed by on-column DNAse I treatment (Sigma Aldrich, St. Louis, MO, USA). One microgram of total RNA was reverse transcribed using the SuperScript® III-First-Strand Synthesis System (Invitrogen, Waltham, MA, USA). RNA expression in transgenic and non-transgenic control plants was quantified by qRT-PCR using the primer pairs listed in Supplementary Table 3, plus 10 ng reverse transcriptase template and SsoAdvanced™ Universal SYBR® Green Supermix (Bio-Rad laboratories Inc., Hercules, CA). The endogenous cassava gene PP2A was used as an internal control (Moreno et al., 2011). qRT-PCR cycling conditions included an initial denaturation holding stage at 95 °C for 30 s, followed by 39 cycles at 95 °C for 5 s, 61 °C for 30 s, and a melt curve stage from 65 °C to 95 °C with 0.5 °C increments for 5 s, followed by final extension at 95 °C for 5 min. For each sample, reactions were set up in triplicates to ensure reproducibility. Quantification of the relative transcript levels was performed using the comparative CT (threshold cycle) method (Livak and Schmittgen, 2001). As the endogenous cassava gene PP2A always appeared in an earlier cycle in the qRT-PCR than the transformed inverted repeats, lower Delta CT values (CT inverted repeat – CTPP2A) reflected higher expression levels. To determine construct-specific RNA expression in insect-challenged transgenic plants, covered top leaves (protected from insect infestation) from transgenic and control plants subjected to insect challenge for 96 hours were collected. RNA extraction and PCR procedures were carried out using the same methodology described above.
2.7 Detection of transgenically-derived siRNAs by Northern blot analysis
The accumulation of transgenically-derived siRNAs in transgenic plants was assessed using Northern blot analysis. Total RNA was isolated from 100 mg of fresh leaf tissue collected from in-vitro grown plantlets using TRIzol reagent (Ambion, Houston, TX, USA). Northern blotting, hybridization, probe preparation, and siRNA detection were carried out as described by Beyene et al. (2017). Thirty micrograms of total RNA were resolved on 15% Criterion™ TBE-Urea Precast Gels (Bio-Rad, Hercules, CA, USA) and transferred to Hybond™ N+ membrane (GE Healthcare Ltd) using a semi-dry transfer apparatus (Bio-Rad, Hercules, CA, USA). DIG-labeled RNA probes were synthesized using the DIG RNA Labeling Kit and SP6/T7 polymerase, following the manufacturer’s protocol (Roche Applied Science, Indianapolis, IN, USA). Signal intensity was visualized by scanning the membranes and quantified using ImageJ software version 10.2 (Rasband, 1997–2019).
2.8 Rearing experimental whitefly colonies on eggplants
Seeds of eggplant Solanum melongena var. Black Beauty (Kings Seeds, UK) were sown into a 1:1 mixture of loam-based compost (J. Arthur Bower’s John Innes No 2, UK) and coir/perlite 70/30 mix compost (Jiffy Tref propagation compost, UK) in 5x5x8 cm disposable plastic pots. The plants were enclosed in a Bug Dorm cage that was thrips- and mites-proof and maintained in a whitefly-free room at 28 ± 2 °C, 50–60% relative humidity and a 14:10 light:dark photoperiod. The plants were watered twice a week with a nutrient solution containing 18:18:18 + 3 NPK + magnesium + trace elements (1:1:1 | Products | Solufeed). After four to six weeks, when the plants had at least four true leaves, adult whiteflies were released on the eggplant plantlets. The B. tabaci colonies from sub-Saharan Africa subgroup 1 (SSA1-SG1) used in this study were established from field populations collected from cassava in Uganda and Tanzania in 2013 (Mugerwa et al., 2021c). The purity of B. tabaci colonies was checked by sequencing the partial mtCO1 gene amplified using primer pairs C1-J-2195 and TL2-N-3014 (Dinsdale et al., 2010; Vyskočilová et al., 2018). SSA1-SG1 is a cassava-associated but polyphagous B. tabaci species (Namuddu et al., 2025) and can perform well on both cassava and eggplant plant hosts (Mugerwa et al., 2021a). Moreover, our previous transcriptomic studies (Malka et al., 2018, Malka et al., 2021), indicated that B. tabaci polyphagous species, capable of performing well on multiple hosts, largely show constitutive gene expression with very low level of expression plasticity.
2.9 Growing transgenic cassava plants for whole plant bioassays
Transgenic and non-transgenic plants of NASE 13 were imported from the Donald Danforth Plant Science Center (DDPSC), St Louis, MO as in-vitro plantlets growing in Petri dishes on MS2 medium solidified with gelzan. Plants were placed in a growth chamber maintained at 28 ± 2 °C, 50–60% relative humidity and a 14:10 Light: Dark photoperiod. After 5–7 days, plantlets were removed from the growth medium and the roots washed using lukewarm tap water to remove all the gelling agent. The plantlets were transferred to 1:1 mixture of loam-based compost (J. Arthur Bower’s John Innes No 2, UK) and coir/perlite 70/30 mix compost (Jiffy Tref propagation compost, UK) in 5x5x8 cm disposable plastic pots and placed in incubation trays with transparent lids to minimize water loss. After about eight weeks, plants with more than three or four leaves were transferred and enclosed in Lock & Lock whitefly-proof cages at one plant per each cage (Wang et al., 2011). The Lock & Lock cages had two side openings in the upper container covered by 160-µm nylon mesh. Plants were watered twice a week with a nutrient solution containing 18:18:18 + 3 NPK + magnesium + trace elements (1:1:1 | Products | Solufeed).
2.10 Whole plant adult and nymph insect performance bioassays
The laboratory plant bioassays were performed on cassava plants enclosed in Lock & Lock whitefly-proof cages using a B. tabaci SSA1-SG1 colony reared on eggplants. For adult survival assays, adult whiteflies were released onto cassava plants growing in Lock-Lock cages within the NRI insectary at 28 ± 2 °C; 14:10 h light:dark cycle and 50-60% relative humidity. Six clonal replicate plants per transgenic event were each inoculated with fifteen pairs (male and female) of 1–2-day-old, whitefly adults. Survival was recorded on the 7th day after release of the insects. Adult survival was documented by carefully examining the live adult whiteflies feeding on the leaves of the experimental plants in the lock-lock pots with minimal disturbance.
In the nymph development assays, fifteen pairs (male and female) of 1–2-day-old whitefly adults were released onto six-week-old cassava plants in Lock-Lock cages within the NRI insectary (28 ± 2 °C; 14:10 h light:dark cycle at 50-60% relative humidity) for a 24 h egg laying period. Six plants (replications) per transgenic event/line were tested. After 25 days, the developmental status of the progeny was determined. In the first experiments, we calculated the total number of progeny by counting the number of undeveloped eggs, 2nd and 3rd nymphs, red-eyed 4th nymphs (pupae) and number of individuals that completed development and emerged as adults, leaving the remains of their exoskeleton (exuviae). These analyses indicated that there is no difference between the total progeny laid on the control plants, NASE13 expressing dsRNA against GFP, and the experimental plants expressing dsRNA that target insect genes. Moreover, a highly significant correlation was observed between the total number of laid progeny and the total number of nymph and exuviae stages (r = 0.94, P ≤ 0.0001). Therefore, the complete set of bioassays reported in this study focus only on counting 2nd and 3rd nymphs, red-eyed 4th nymphs (pupae) and emerged adults (exuviae).
2.11 Statistical analysis
The significance of the differences between the proportion of adult survival and the proportion of progeny that reached advanced development stages (the proportion of red-eyed 4th nymphs and exuviae from the total number nymph and exuviae stages) was tested using a one-way ANOVA model followed by pairwise comparisons. Prior to the analysis, the proportional data were arcsin-square root transformed. As multiple tests were conducted, a false discovery rate (FDR) correction was applied. Statistical significance was assumed at P ≤ 0.05. All statistical analyses were conducted using JMP Pro 18.0 (SAS Institute, Cary, NC).
2.12 Analyses of target gene expression in whiteflies
Downregulation of target insect genes was determined from insects after feeding on RNAi transgenic, and non-transgenic control plants. Samples consisting of 50 surviving adults (after 7 days of feeding) or 50 nymphs (after 15 days of development), were collected into 1.5 ml Eppendorf tubes, immediately placed into liquid nitrogen, and stored at -80 °C. Total RNA was extracted using an ISOLATE II RNA mini kit (Meridian Bioscience, Ohio) following the manufacturer’s protocol, with DNA contamination removed using a PerfeCTa DNaseI (Quanta Bioscience, Massachusetts) treatment. Reverse transcriptase reaction was performed using 500 ng RNA from each sample and the Verso cDNA synthesis kit with Oligo-dT primer (Thermo Fisher Scientific, Massachusetts). The expression level of each target gene was examined using qRT-PCR. The reactions were performed using the CFX Connect Real-Time PCR System (BIO-RAD, California). A set of primers (Supplementary Table 3) was designed and calibrated for each target gene based on the Thornton and Basu (2011) protocol using the Primer3 software (Kõressaar et al., 2018; http://primer3.ut.ee/). The B. tabaci ribosomal protein L13a was used as the reference gene. The qRT-PCR master mix contained: 5 µl iTaq Universal SYBR Green Supermix (BIO-RAD, California), 0.5 µl of both forward and reverse primers (2 pmol/µl), 2 µl DDW and 2 µl cDNA template. The expression of target and reference genes was tested in triplicate for each sample to ensure the validity of the results. qRT-PCR thermal conditions consisted of one cycle of 95 °C for 2 min, followed by 40 cycles of 95 °C for 5 s and 60 °C for 30 s, and an ending cycle of 95 °C for 5 s, 65 °C for 5 s, and 95 °C for 30 s. Quantification of the transcript’s expression levels was conducted according to the ΔΔCt method. A one-way ANOVA model was used to determine the significance of the differences between the expression means of the treatment and control samples (P ≤ 0.05). As multiple tests were conducted, a false discovery rate (FDR) correction was applied.
2.13 Population dynamics modeling
2.13.1 Model – single population
The life cycle of B. tabaci SSA1-SG1 was modeled with four stages: eggs, nymphs, pupae and adults. Each day, individuals were considered to either remain in their current stage, progress to the next stage or die. Adults could also lay eggs. Transition probabilities depend on temperature, T. To simulate the dynamics, a matrix population model was used (Bodine et al., 2014). The population state was a 4×1 vector giving the numbers of eggs, nymphs, pupae and adults, and daily transitions, given by a 4×4 Leslie matrix, M(T), whose entries are temperature-dependent transition probabilities. Each day the current population vector was multiplied by the temperature-specific matrix. To avoid unrealistic exponential growth and to reflect a carrying capacity of ~500 adults per leaf, density-dependent adult mortality was applied: with each day, an amount of subtracted from the adult population.
2.13.2 Parameterization – single population
The transition probabilities for the wild-type population were obtained based on laboratory measurements of development, reproduction, and mortality rates (Aregbesola et al., 2020). The laboratory data provided the number of individuals, out of a given starting number, that developed to the next stage, along with their median development time in days, mi, where i is the stage index. For our model, the daily development rates were calculated as . Daily mortality rate was calculated as . It follows that the fraction of population remaining in their present state is given by . Four life stages (eggs, nymphs, pupae and adults) were used, although the laboratory experiments separated the early nymph stages into two (2nd and 3rd nymphs). The nymph stages were combined into one stage by using the following calculation: Tj (j = 1, 2, 3) represents the median development time from the jth to (j + 1)th instar nymph, and T4 representing the median development time required for individuals in the 4th instar nymph stage to develop into pupae. The daily development rate for the entire nymph stage is given by . Accordingly, stage-specific mortality was denoted as Nj, and the daily mortality rate of nymphs was calculated as . The detailed calculation of the Matrices is given in the Supplementary Data File. The daily survival rate for the nymphs was calculated as .
Lab-derived parameters were adjusted to reflect field mortality using parameter q. This adjustment reduced survival rates in the egg, nymph, and pupae stages to match observed field adult emergence rates of approximately 10% (Zidon et al., 2016). The population dynamics under two scenarios was compared: with no intervention (wild-type plants), and with intervention (transgenic plants). The rates of the transgenic population were adjusted based on our empirical findings. Based on the findings that a fraction PL of the nymphs and PA of the adults are eliminated in a given generation. Accordingly, we increased the daily mortality rates such that an additional fraction of of adults and of nymphs die each day in transgenic populations, where g is the number of days in a generation.
The initial model included five temperatures: 16 °C, 20 °C, 24 °C, 28 °C, and 32 °C. Field measurements of temperatures in Kampala, Uganda were used, recorded at 15-minute intervals (Macdonald et al., 2018), and averaged to daily values. When the daily average temperature T fell between two of these values, interpolation using a fractional distance between lower and upper bounds was applied to calculate the matrix for any given temperature when running the simulation with actual temperatures.
2.13.3 Model – two populations
Simulated dynamics was developed for a heterogeneous field containing plots of wild-type and transgenic plants. Migration rates represent the proportion of wild-type and transgenic plants surrounding a focal transgenic plant. The two-population model uses the same temperature-dependent transition structure but extends the 4×4 matrices to 8×8, representing the insects’ dynamics on both plant types. The upper-left 4×4 block describes stages on non-transgenic plants, covering egg, nymphal, pupal and adult stages. The lower-right 4×4 block describes the corresponding stages on transgenic plants. Migration is modelled as adult movement from non-transgenic to transgenic plants. The migration rate appears in the element coupling these two stages in the extended matrix. For non-transgenic fields, migration from transgenic sites was assumed negligible, given the large extent of non-transgenic areas relative to typical dispersal distances of B. tabaci, and was set to zero. Temperature-specific matrices were selected and interpolated as in the single-population model, using the same temperature data.
2.13.4 Parameterization – two populations
Two migration scenarios were modelled based on field-realistic dispersal distances: (1) where maximum migration rate was 1% per day (Byrne, 1999), and (2) where maximum migration rate was 10% per day. Migration rates were scaled to represent realistic field compositions, although the true migration rates are largely unknown.
3 Results
3.1 Transgenic GUS expression under control of the SUC promoter
Transgenic plants of cassava variety 60444 were produced expressing the GUS visual marker gene under control of the SUC2 promoter. Histochemical staining of transgenic plants (Figure 1A) revealed tissue-specific GUS expression localized to the vascular tissues in leaves (Figure 1B) and in transverse sections of stem and storage roots, particularly associated with the phloem (Supplementary Figure 2).
Figure 1
3.2 Production and analysis of insecticidal RNAi transgenic plants in cultivar NASE 13
Fifteen RNAi gene constructs targeting the four functional strategies: osmoregulation, carbohydrate homeostasis, symbiosis and detoxification under control of the AtSUC2 promoter, plus an RNAi GFP control construct, were produced and transformed into cassava variety NASE 13. A total of 140 independent RNAi transgenic events were recovered, with the minimum target of five independent events achieved for all constructs except DWF 64 (DapF) (Supplementary Table 2). Transgenic plant lines were evaluated for their ability to express RNA from the inverted repeat sequence using qRT-PCR (Figure 1C). To show that template was being processed into siRNAs, selected plants were also analyzed for accumulation of small interfering RNA (siRNA) species specific to the inverted repeat sequence (Figure 1C). Higher dsRNA expression levels (lower Ct values – see section 2.6 for clarification) were correlated with higher accumulation of siRNA (Figure 1D). The transgenic plant lines were screened under greenhouse conditions, where no adverse phenotypic effects were observed associated with expression of the inverted repeat sequences (Figure 1E). Stability of transgenic dsRNA expression was also assessed after a two-year period (Figure 1F). These longitudinal analyses demonstrated sustained expression of the dsRNA constructs and a consistent correlation of expression levels across the two-year timeframe (Figure 1G).
3.3 Insecticidal effects of the four strategies on whitefly performance in whole plant bioassays and their correlation with gene silencing
Transgenic NASE 13 cassava plants expressing the 15 different RNAi constructs targeting genes within the four strategies underwent whole plant bioassays to assess their insecticidal effect on adult and nymph whiteflies. The most insecticidal transgenic lines in each strategy caused up to 35-58% decrease in adult survival after seven days feeding when compared to the non-transgenic wildtype (WT) control line NASE 13. Excluding the two constructs DWF59 and DWF60, at least two transgenic events, showed significant (P ≤ 0.05) negative effect (reduced adult survival) when compared to an event expressing dsRNA against the GFP gene (Figure 2A). Of special interest were constructs DWF55, expressing dsRNA targeting an ST gene; DWF56, expressing dsRNA targeting an ABC transporter gene; DWF61, expressing dsRNA targeting the ArgH gene; DWF63, expressing dsRNA targeting the LysA gene and DWF66, expressing dsRNA targeting the UDPGT gene. When using these constructs, 83-100% of the transformed plant lines were capable of causing a significant reduction (P ≤ 0.05) in adult survival, indicating a strong negative effect of the specific targeting RNAi. In contrast, only two transgenic events, DWF66–006 and DWF63-001, out of a total of 104, presented high efficacy capable of reducing the adult whitefly population by more than 50%.
Figure 2
Overall, expression of the 15 different RNAi constructs did not affect the total number of nymphs developing on the experimental plants. When independent events were compared to a control event expressing dsRNA against the GFP gene, only 14 were found to display a significant (P ≤ 0.05) lower total number of developing nymphs. Excluding one construct, DWF62, the effect was considered to be sporadic and event specific as it appeared only in a minority of events per construct. For example, only one from seven events in DWF55 and two from 12 events in DWF66 (Supplementary Table 4).
On the other hand, in the nymph development assays, the best transgenic events in each strategy caused 75-90% delay in development when the normalized proportion of progeny in advanced stages (the proportion of red-eyed 4th nymphs and nymphs that completed development) was compared to the NASE 13 WT control (Figure 2B). Eight of the 15 constructs showed significant (P ≤ 0.05) effect (reduced nymph development), in at least two independent events, when compared to the control expressing RNAi against GFP (Figure 2B). Of special interest were constructs DWF11, expressing dsRNA targeting an AQP gene; DWF55, expressing dsRNA targeting an ST gene; DWF62, expressing dsRNA targeting the CM gene; and DWF65, expressing dsRNA targeting the BioB gene. When using these constructs for transformation, 68-100% of the transgenic events were capable of causing a significant (P ≤ 0.05) delay in nymph development, indicating a strong negative effect of the specific RNAi on immature whitefly stages. Transformations with all four constructs produced two or more independent events that presented high insecticidal efficacy capable of reducing the development of immature stages by more than 50% (Figure 2B).
We employed qRT-PCR to establish a direct association between downregulation of target gene expression and reduced performance of whitefly nymphs and adults. RNA was extracted from nymphs and adults feeding on transgenic cassava plant lines expressing RNAi inverted repeat constructs DWF55, DWF56, DWF57, DWF61, DWF65, DWF66, and DWF68, and from control plants expressing dsGFP. In adult whiteflies, target gene expression was reduced by up to 45% relative to control adults, while in nymphs, downregulation reached 57% compared to controls. A significant positive correlation (P = 0.0117) was observed between the level of target gene silencing and the insecticidal effect on nymph development and adult survival (Figure 3), providing proof-of-principle that suppression of target gene expression is causally linked to reduced performance in both immature and adult stages of cassava-whitefly.
Figure 3
3.4 Assessing the influence of RNAi transgenic cassava on whitefly population dynamics using a temperature-driven model
Using a daily, temperature-dependent 4×4 Leslie matrix that captures transitions among egg, nymph, pupa, and adult - parameterized from laboratory measurements across 16–32 °C and driven by field temperature time series via interpolation, we simulated stage-specific B. tabaci SSA1 whitefly mortality of adults, nymphs, or both. Figure 4 illustrates modelled whitefly populations over the course of one year, excluding immigration and emigration processes. Simulation results indicated that population growth begins approximately 40 days after initialization, corresponding to two generations, in non-transgenic cassava fields with an exponential increase in population size occurring over the subsequent ~40 days (two additional generations), followed by stabilization at a steady-state level after four generations. In transgenic scenarios, RNAi cassava plants were simulated to reduce both adult survival and nymph development (Figure 4A), only nymph development (Figure 4B) or only adult survival (Figure 4C) by 30–80%. These interventions altered population dynamics in two principal ways. First, the initial population build-up was delayed by one to five generations when the combined negative effect on both nymph and adult stages reached 60%, 70%, or 80%, respectively. Second, the steady-state population size was reduced by approximately 33%, 44%, and 68% under the same respective treatment intensities. Two additional insights can be drawn from the simulations presented in Figure 4. First, RNAi technologies targeting immature nymphal stages are predicted to be more effective under field conditions than those primarily affecting adult stages. Second, interventions with a combined negative effect below 50–60% are projected to exert only minor influence on the overall population dynamics of whiteflies in the field.
Figure 4
We extended the model to two interacting plots (transgenic and non-transgenic) by doubling the state space (8×8) and allowing daily adult migration between plots, and then varied the fraction of the landscape planted with transgenic cassava while comparing 1% versus 10% daily movement. A constant 60% reduction in both nymphal development and adult survival was assumed for the transgenic plants. Simulation results showed that at a 1% migration rate, population growth dynamics exhibited only minor differences among the four mixed-field scenarios (Figure 5A). However, when the migration rate was increased to 10%, substantial differences emerged in whitefly population dynamics between field mixtures dominated by transgenic plants and those dominated by non-transgenic plants (Figure 5B). Given the absence of reliable estimates for adult SSA1 whitefly migration rates in the current literature, we conservatively limited the migration parameter to a maximum of 10%. Nonetheless, the simulation outcomes suggest that migration rates exceeding 10% could significantly compromise the effectiveness of RNAi transgenic cassava in agricultural landscapes where non-transgenic, whitefly-susceptible varieties remain prevalent.
Figure 5
4 Discussion
We report development of efficacious in planta RNAi technology for controlling a major insect pest of African cassava, the cassava-whitefly B. tabaci. Cassava is a strategic food security crop for smallholder farmers, particularly in sub-Saharan Africa (Obong’o et al., 2024). Integrated management packages of the whitefly vector are currently limited by the relatively low availability of improved disease or insect resistant varieties, and the fact that commercial insecticides and other input-intensive measures are expensive and/or logistically inaccessible for the majority of smallholder farmers (Legg et al., 2022; Bayiyana et al., 2023).
We bring evidence that transgenic cassava plants in the East Africa cultivar NASE 13, expressing RNAi technology that targets essential genes in whitefly biology, has potential to significantly suppress whitefly populations. The most efficacious transgenic events in each of the four reported strategies showed moderate ability (35-58%) to decrease adult survival, but were highly effective (75-90%) in delaying nymph development. In this context, it is important to note that observations during the bioassay indicated that in the majority of cases, developmental delay caused by RNAi prevented successful completion of the insect life cycle, eventually leading to nymph death. The potential of the RNAi technology to achieve significant population suppression of cassava whiteflies was emphasized by two complementing outcomes of the simulation models. First, whitefly control strategies targeting immature nymphal stages were predicted to be more effective under field conditions than those primarily affecting adult stages. Specifically, adult mortality alone would likely not have a marked effect on the population (Figure 4C). Nevertheless, the combination of adult and nymph mortality, as occurs due to transgenic RNAi cassava, may significantly suppress the population (Figure 4A). Second, only control strategies with a proven potential to cause at least 60% reduction in nymph performance should be taken into consideration as possible new strategies for efficient whitefly control. Our simulation models also explored the important effect of adult migration from fields planted with whitefly susceptible, non-RNAi cassava plants, a topic that is neglected in most RNAi studies to date. The simulations indicated that the technology will be most effective if applied at a regional level with high percentage of farmers growing the RNAi varieties.
The differences observed in RNAi efficacy among target genes likely reflect a combination of target gene essentiality, expression level, tissue accessibility after oral delivery, particularly to midgut epithelial cells, and the extent of functional redundancy or physiological compensation. For example, the strong effect of silencing the aquaporin (AQP) gene might be related to the fact that from the eight aquaporins paralogs present in the B. tabaci genome, it is the only one found to code for an effective water-selective channel and to be primarily expressed in midgut and hindgut tissues, likely playing an important role in osmoregulation by shunting excess dietary water from the distal to the proximal end of the gut (Mathew et al., 2011; Van Ekert et al., 2016). Similarly, the relatively strong effects observed for silencing genes coding for a specific sugar transporter (ST) and a specific ABC transporter, although the two genes belong to large host-responsive gene families in B. tabaci (Malka et al., 2021), suggest that the specific target genes play direct important roles in feeding-related physiology, nutrient/osmotic balance, and host-associated stress responses (Malka et al., 2021). ArgH and LysA may likewise represent metabolic functions with more immediate phenotypic consequences under the present assay conditions. These two genes are horizontally transferred genes of bacterial origin, mediate terminal reactions of arginine and lysine biosynthesis respectively, and are highly expressed in target organs where essential amino acid provisioning are regulated (Kaweesi et al., 2024). On the other hand, gene targets that their silencing displayed only little measurable impact on whitefly performance, like the GH13 α-glucosidases (SUCs) targeted by the DWF57 and DWF59 constructs, are likely to be more effectively buffered by the presence in the B. tabaci genome of large number of paralogs (Wintraube et al., 2025) or may require longer time periods to manifest measurable effects.
In smallholder farming systems, such as that of most cassava cultivation in Africa, the main advantage of in planta dsRNA expression is its ability to provide sustainable, seasonal-long, transgenerational protection against whiteflies with no farmer-required inputs. Moreover, once an efficacious RNAi expression cassette is developed it can be introduced into any farmer-preferred variety through direct transformation or conventional backcrossing (Basso et al., 2025). A growing body of studies indicate that RNAi-based insect control is highly sequence-specific and degrades rapidly in soil and water, so that, when target genes are selected using bioinformatic filters to minimize homology with non-targets, the probability of ecologically relevant off-target effects is low to the point of negligible, compared with the negative effects of broad-spectrum chemical insecticides (ChenDe Schutter, 2024). Empirical non-target studies, including recent assessments of dsRNA-expressing cotton on the predatory ladybird Harmonia axyridis (Yao et al., 2025) and dietary exposure assays with multiple beneficial insects, generally report no adverse effects on survival, development or fitness under realistic exposure scenarios (Castellanos et al., 2022; Taning et al., 2021). Finally, in-planta dsRNA expression for pest control is advantageous over repeated dsRNA foliar spraying in marginally profitable and vulnerable crop production systems such as smallholder farms. This is mainly because dsRNA spraying requires access to, and involves recurrent costs of purchasing and applying the formulated dsRNA that are not feasible for small holder farmers (Bramlett et al., 2020; Willow and Veromann, 2022).
There are limitations to the RNAi technology that should be taken into consideration to ensure durable, long-term efficacy. A central concern is the evolution of resistance in target pest populations (Palli, 2023). Two possible scenarios should be considered in this respect. In the first, resistance is specific to a particular RNAi target (e.g. changes in the target gene sequence or its regulation), while the core RNAi machinery remains functional. In this case, resistance management can simply include switching to an alternative sequence region within the same gene or to a different essential gene to restore susceptibility (Parry et al., 2020). In the second scenario, insects evolve resistance that broadly impairs the RNAi pathway itself. For example, increased dsRNA degradation in the insect gut that prevents effective gene silencing (Spit et al., 2017) or reduced dsRNA uptake by the insect, often associated with downregulation or mutations in dsRNA transporters such as Sid-1-like proteins or clathrin-mediated endocytosis components (Duanmu et al., 2025; Khajuria et al., 2018). To overcome the second scenario, resistance management (IRM) frameworks for dsRNA traits recommend pyramiding (stacking) with at least one additional technology that operates through a different insecticidal mode of action (Narva et al., 2025). In this case, only insects that will evolve resistance to both technologies will survive, which is less likely to happen (Ni et al., 2017). If a second technology is not available, the time until resistance to the RNAi technology evolves can be significantly extended by maintaining non-RNAi refuges that produce large numbers of susceptible insects that dilute the resistance alleles (Narva et al., 2025). This approach was successfully applied during the commercialization of the Bacillus thuringiensis (Bt)-derived technology, especially if the evolved resistance is associated with a significant fitness cost (Carrière and Tabashnik, 2023).
There are also few limitations to our study that should be addressed in future research. Because B. tabaci is haplodiploid, any treatment that reduces successful mating or fertilization could indirectly bias progeny sex ratio toward males. In line with this, previous work in B. tabaci showed that young females laid proportionately more female-producing eggs than older females, and that females with access to males throughout their life laid a higher proportion of females than those that mated only once shortly after emergence, consistent with sperm limitation affecting sex ratio (Horowitz and Gerling, 1992). Therefore, as male sterility and mating success were not directly assessed in the present study, we cannot exclude the possibility that RNAi affected offspring sex ratio indirectly through effects on male reproductive performance, including mating success, sperm transfer, or fertility. This possibility should be addressed in future work using dedicated assays of mating success and male fertility.
5 Conclusion
Evidence provided here indicates that in-planta RNAi technology offers a realistic option in developing durable control of cassava whitefly in African cassava fields. Our work and previous studies suggest that targeting genes involved in osmoregulation, such as aquaporin, sugar transporter and α-glucosidase, and/or symbiosis, such as chorismate mutase and biotin synthase, are good targets for markedly reducing the ability of the insect to build large populations in the field (Luo et al., 2017; Wang and Luan, 2023; Wintraube et al., 2025). A logical next step to enhance insecticidal activity is to drive higher and more spatially appropriate dsRNA accumulation by exploiting additional strong phloem-specific promoters such as MeSWEET1-like, and MeSUS1 (Zierer et al., 2022). In parallel, stacking dsRNAs against multiple essential whitefly genes or co-targeting dsRNases and vital genes has been proposed, as a way to increase mortality and slow resistance evolution (Luo et al., 2017; Nitnavare et al., 2021). With these improvements, all achievable in a relatively short period, plant-mediated RNAi will be well positioned to become a core, resistance-management-compatible tool for whitefly control in cassava fields. Still, it is important to note that commercial deployment of such plants will depend on regulatory approval pathways. Under existing biosafety frameworks, transgenic RNAi crops are typically subject to case-by-case assessment for environmental and food/feed safety, including assessment of potential effects on non-target organisms (Akinbo et al., 2021; Barro et al., 2025).
Statements
Data availability statement
Modelling code and data associated with this paper (Supplementary Data File) was published in Dryad. DOI: https://doi.org/10.5061/dryad.z08kprrv8. Dataset name: “Data and Code for modeling RNAi-technology effect on plant resistance for Cassava Whitefly (Bemisia tabaci)”. Oren Hasson and Adam Lampert.
Ethics statement
The manuscript presents research on animals that do not require ethical approval for their study.
Author contributions
NN: Conceptualization, Data curation, Methodology, Supervision, Writing – original draft, Writing – review & editing, Investigation, Validation. RS: Conceptualization, Data curation, Investigation, Methodology, Validation, Writing – original draft, Writing – review & editing. JG: Investigation, Methodology, Writing – review & editing. TJ: Methodology, Writing – review & editing. SL: Writing – review & editing, Data curation, Investigation. DW: Data curation, Investigation, Writing – review & editing, Methodology. EY: Methodology, Writing – review & editing. OH: Writing – review & editing, Data curation, Investigation. AL: Data curation, Investigation, Writing – review & editing, Conceptualization, Methodology, Supervision, Writing – original draft. JC: Conceptualization, Data curation, Investigation, Methodology, Supervision, Writing – review & editing, Funding acquisition, Project administration. NT: Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing – review & editing, Resources, Writing – original draft. SM: Conceptualization, Data curation, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing, Formal Analysis. OM: Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Project administration, Supervision, Writing – original draft, Writing – review & editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the Gates Foundation, Seattle, WA, through the University of Greenwich (Grant number OPP1058938).
Acknowledgments
We thank Angela E. Douglas for contributions toward the RNAi gene constructs DWF10–15 and her insightful comments during development of the experimental system. The NASE 13 variety utilized here was provided by the National Crops Resources Research Institute, NARO, Uganda. We also thank Ksenia Juravel for her help during Figures’ production.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2026.1822258/full#supplementary-material
Supplementary Figure 1Schematic representation of gene constructs used for the genetic transformation of cassava.
Supplementary Figure 2Transverse sections of stem and storage roots indicating GUS localization, particularly in phloem tissues.
References
1
AdebayoW. G. (2023). Cassava production in Africa: A panel analysis of the drivers and trends. Heliyon9, e19939. doi: 10.2139/ssrn.4257538
2
AkinboO.ObukosiaS.OuedraogoJ.SineboW.SavadogoM.TimpoS.et al. (2021). Commercial release of genetically modified crops in Africa: interface between biosafety regulatory systems and varietal release systems. Front. Plant Sci.12. doi: 10.3389/fpls.2021.605937. PMID:
3
AregbesolaO. Z.LeggJ. P.LundO. S.SigsgaardL.SporlederM.CarhuapomaP.et al. (2020). Life history and temperature-dependence of cassava-colonising populations of Bemisia tabaci. J. Pest Sci.93, 1225–1241. doi: 10.1007/s10340-020-01249-z. PMID:
4
AtimJ.KalyebiA.Bohorquez-ChauxA.Lopez-LavalleL. A. B.OmongoC. A.ColvinJ.et al. (2024). Identifying cassava genotypes resistant to the African cassava whitefly, Bemisia tabaci (Gennadius). Agriculture14, 1016. doi: 10.3390/agriculture14071016. PMID:
5
BarroF.BraeuningA.GoumperisT.LewandowskaA.MoxonS.PapadopoulouN.et al. (2025). Risk assessment considerations for RNAi‐based genetically modified plants. EFSA. J.23, e9321. doi: 10.2903/j.efsa.2025.9321. PMID:
6
BassoM. F.VásquezD. D. N.Campos-PintoE. R.PinheiroD. H.CruzB.MakturaG. C.et al. (2025). Progress and opportunities of in planta and topical RNAi for the biotechnological control of agricultural pests. Agronomy15, 859. doi: 10.3390/agronomy15040859. PMID:
7
BayiyanaI.BuaA.OzimatiA.MugishaJ.ColvinJ.OmongoC. A. (2023). Insecticide use by small-scale Ugandan cassava growers: An economic analysis. Agriculture13, 1043. doi: 10.3390/agriculture13051043. PMID:
8
BeyeneG.ChauhanR. D.IlyasM.WagabaH.FauquetC. M.MianoD.et al. (2017). A virus derived stacked RNAi construct confers robust resistance to cassava brown streak disease. Front. Plant Sci.7. doi: 10.3389/fpls.2016.02052. PMID:
9
BodineE. N.LenhartS.GrossL. J. (2014). Mathematics for the life sciences (Princeton, NJ: Princeton University Press).
10
BramlettM.PlaetinckG.MaienfischP. (2020). RNA-based biocontrols - a new paradigm in crop protection. Engineering6, 522–527. doi: 10.1016/j.eng.2019.09.008. PMID:
11
ByrneD. N. (1999). Migration and dispersal by the sweet potato whitefly, Bemisia tabaci. Agric. For. Meteorol.97, 309–316. doi: 10.1016/s0168-1923(99)00074-x
12
CarrièreY.TabashnikB. E. (2023). Fitness costs and incomplete resistance associated with delayed evolution of practical resistance to Bt crops. Insects14, 214. doi: 10.3390/insects14030214
13
CastellanosN. L.SmaggheG.TaningC. N. T.OliveiraE. E.ChristiaensO. (2022). Risk assessment of RNAi-based pesticides to non-target organisms: Evaluating the effects of sequence similarity in the parasitoid wasp Telenomus podisi. Sci. Tot. Environ.832, 154746. doi: 10.1016/j.scitotenv.2022.154746. PMID:
14
ChauhanR. D.BeyeneG.KalyaevaM.FauquetC. M.TaylorN. (2015). Improvements in Agrobacterium-mediated transformation of cassava (Manihot esculenta Crantz) for large-scale production of transgenic plants. Plant Cell Tissue Organ Cult.121, 591–603. doi: 10.1007/s11240-015-0729-z. PMID:
15
ChenY.De SchutterK. (2024). Biosafety aspects of RNAi-based pests control. Pest Manage. Sci.80, 3697–3706. doi: 10.1002/ps.8098. PMID:
16
ColvinJ.OmongoC. A.MaruthiM. N.Otim–NapeG. W.ThreshJ. M. (2004). Dual begomovirus infections and high Bemisia tabaci populations: Two factors driving the spread of a cassava mosaic disease pandemic. Plant Pathol.53, 577–584. doi: 10.1111/j.0032-0862.2004.01062.x. PMID:
17
DesprésL.DavidJ. P.GalletC. (2007). The evolutionary ecology of insect resistance to plant chemicals. Trends Ecol. Evol.22, 298–307. doi: 10.1016/j.tree.2007.02.010
18
DinsdaleA.CookL.RiginosC.BuckleyY.De BarroP. J. (2010). Refined global analysis of Bemisia tabaci (Gennadius) mitochondrial CO1 to identify species level genetic boundaries. Ann. Entomol. Soc Am.103, 196–208. doi: 10.1603/an09061. PMID:
19
DuanmuH.GuoF.LiuY.HeK. (2025). Developing RNAi pesticides in the post genomic era: a review. Adv. Agrochem.4, 324–335. doi: 10.1016/j.aac.2025.07.002. PMID:
20
EakteimanG.Moses-KochR.MoshitzkyP.Mestre-RinconN.VassãoD. G.LuckK.et al. (2018). Targeting detoxification genes by phloem-mediated RNAi: A new approach for controlling phloem-feeding insect pests. Insect Biochem. Mol. Biol.100, 10–21. doi: 10.1016/j.ibmb.2018.05.008. PMID:
21
GresshoffP. M.DoyC. H. (1974). Derivation of a haploid cell line from Vitis vinifera and the importance of the stage of meiotic development of anthers for haploid culture of this and other genera. Z. Pflanzenphysiol.73, 132–141. doi: 10.1016/s0044-328x(74)80084-x
22
GuastellaD.LulahH.TajebeL. S.CavalieriV.EvansG. A.PedataP. A.et al. (2015). Survey on whiteflies and their parasitoids in cassava mosaic pandemic areas of Tanzania using morphological and molecular techniques. Pest Manage. Sci.71, 383–394. doi: 10.1002/ps.3810. PMID:
23
HorowitzA. R.GerlingD. (1992). Seasonal variation of sex ratio in Bemisia tabaci on cotton in Israel. Environ. Entomol.21, 556–559. doi: 10.1093/ee/21.3.556. PMID:
24
HuY.ComjeanA.RodigerJ.LiuY.GaoY.ChungV.et al. (2021). FlyRNAi.org—the database of the Drosophila RNAi screening center and transgenic RNAi project: 2021 update. Nucleic Acids Res.49, D908–D915. doi: 10.1093/nar/gkaa936
25
JeffersonR. A.KavanaghT. A.BevanM. W. (1987). GUS fusions: β-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J.6, 3901–3907. doi: 10.1002/j.1460-2075.1987.tb02730.x. PMID:
26
JingX.WhiteT. A.LuanJ.JiaoC.FeiZ.DouglasA. E. (2016). Evolutionary conservation of candidate osmoregulation genes in plant phloem sap‐feeding insects. Insect Mol. Biol.25, 251–258. doi: 10.1111/imb.12215. PMID:
27
KaweesiT.ColvinJ.CampbellL.VisendiP.MaslenG.AlicaiT.et al. (2024). In silico prediction of candidate gene targets for the management of African cassava whitefly (Bemisia tabaci, SSA1-SG1), a key vector of viruses causing cassava brown streak disease. PeerJ12, e16949. doi: 10.7717/peerj.16949. PMID:
28
KhajuriaC.IvashutaS.WigginsE.FlagelL.MoarW.PleauM.et al. (2018). Development and characterization of the first dsRNA-resistant insect population from western corn rootworm, Diabrotica virgifera virgifera LeConte. PloS One13, e0197059. doi: 10.1371/journal.pone.0197059. PMID:
29
KõressaarT.LepametsM.KaplinskiL.RaimeK.AndresonR.RemmM. (2018). Primer3_masker: integrating masking of template sequence with primer design software. Bioinformatics34, 1937–1940. doi: 10.1093/bioinformatics/bty036
30
LeggJ. P.Diebiru-OjoE.EagleD.FriedmannM.KanjuE.KapingaR.et al. (2022). “ Commercially sustainable cassava seed systems in Africa,” in Root, tuber and banana food system innovations: value creation for inclusive outcomes. Eds. ThieleG.FriedmannM.CamposH.PolarV.BentleyJ. W. ( Springer International Publishing, Cham), 453–482.
31
LeggJ. P.SseruwagiP.BonifaceS.Okao-OkujaG.ShirimaR.BigirimanaS.et al. (2014). Spatio-temporal patterns of genetic change amongst populations of cassava Bemisia tabaci whiteflies driving virus pandemics in East and Central Africa. Virus Res.186, 61–75. doi: 10.1016/j.virusres.2013.11.018. PMID:
32
LivakK. J.SchmittgenT. D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔΔCT method. Methods25, 402–408. doi: 10.1006/meth.2001.1262. PMID:
33
LuanJ. B.ChenW.HasegawaD. K.SimmonsA. M.WintermantelW. M.LingK. S.et al. (2015). Metabolic coevolution in the bacterial symbiosis of whiteflies and related plant sap-feeding insects. Genome Biol. Evol.7, 2635–2647. doi: 10.1093/gbe/evv170. PMID:
34
LuoY.ChenQ.LuanJ.ChungS. H.Van EckJ.TurgeonR.et al. (2017). Towards an understanding of the molecular basis of effective RNAi against a global insect pest, the whitefly Bemisia tabaci. Insect Biochem. Mol. Biol.88, 21–29. doi: 10.1016/j.ibmb.2017.07.005. PMID:
35
MacdonaldD. M. J.MillerJ. D.WinterbournJ. B.WarnaarsT. A.MwebembeziL.ArwatP.et al. (2018). “ High resolution meteorological data from a network of five stations in Kampala, Uganda,” in Centre for environmental data analysis. Available online at: https://catalogue.ceda.ac.uk/uuid/a2a8100c46794985b410ea436f24812a/ (Accessed June 30, 2019).
36
MacfadyenS.PaullC.BoykinL. M.De BarroP.MaruthiM. N.OtimM.et al. (2018). Cassava whitefly, Bemisia tabaci (Gennadius) (Hemiptera: Aleyrodidae), in sub-Saharan African farming landscapes: A review of the factors determining abundance. Bull. Entomol. Res.108, 565–582. doi: 10.1371/journal.pone.0204862. PMID:
37
MacfadyenS.TayW. T.HulthenA. D.PaullC.KalyebiA.JacombF.et al. (2021). Landscape factors and how they influence whitefly pests in cassava fields across East Africa. Landscape Ecol.36, 45–67. doi: 10.1007/s10980-020-01099-1. PMID:
38
MalkaO.FeldmesserE.SharonE.Santos-GarciaD.HanW. H.SealS.et al. (2021). The molecular mechanisms that determine different degrees of polyphagy in the Bemisia tabaci species complex. Evol. Appl.14, 807–820. doi: 10.22541/au.159129095.50996321
39
MalkaO.Santos-GarciaD.FeldmesserE.SharonE.Krause-SakateR.DelatteH.et al. (2018). Species-complex diversification and host-plant associations in Bemisia tabaci: A plant-defense, detoxification perspective revealed by RNA-Seq analyses. Mol. Ecol.27, 4241–4256. doi: 10.1111/mec.14865
40
ManX.HuangC.WuS.GuoJ.WanF.FrancisF.et al. (2024). A chromosome-level genome assembly of the heteronomous hyperparasitoid wasp Encarsia sophia. Sci. Data11, 1250. doi: 10.1038/s41597-024-04040-2. PMID:
41
ManzeF.RubaihayoP.OzimatiA.GibsonP.EsumaW.BuaA.et al. (2021). Genetic gains for yield and virus disease resistance of cassava varieties developed over the last eight decades in Uganda. Front. Plant Sci.12. doi: 10.3389/fpls.2021.651992. PMID:
42
MathewL. G.CampbellE. M.YoolA. J.FabrickJ. A. (2011). Identification and characterization of functional aquaporin water channel protein from alimentary tract of whitefly, Bemisia tabaci. Insect Biochem. Mol. Biol.41, 178–190. doi: 10.1016/j.ibmb.2010.12.002. PMID:
43
MorenoI.GruissemW.VanderschurenH. (2011). Reference genes for reliable potyvirus quantitation in cassava and analysis of Cassava brown streak virus load in host varieties. J. Virol. Methods177, 49–60. doi: 10.1016/j.jviromet.2011.06.013. PMID:
44
MugerwaH.ColvinJ.AlicaiT.OmongoA. C.KabaaluP.VisendiP.et al. (2021b). Genetic diversity of whitefly (Bemisia spp.) on crop and uncultivated plants in Uganda: Implications for the control of this devastating pest species complex in Africa. J. Pest Sci.94, 1307–1330. doi: 10.1007/s10340-021-01355-6. PMID:
45
MugerwaH.SseruwagiP.ColvinJ.SealS. (2021a). Is high whitefly abundance on cassava in sub-Saharan Africa driven by biological traits of a specific, cryptic Bemisia tabaci species? Insects12, 260. doi: 10.3390/insects12030260. PMID:
46
MugerwaH.WangH.SseruwagiP.SealS.ColvinJ. (2021c). Whole-genome single nucleotide polymorphism and mating compatibility studies reveal the presence of distinct species in sub-Saharan Africa Bemisia tabaci whiteflies. Insect Sci.28, 1553–1566. doi: 10.1111/1744-7917.12881. PMID:
47
MukiibiD. R.AlicaiT.KawukiR.Okao-OkujaG.TairoF.SseruwagiP.et al. (2019). Resistance of advanced cassava breeding clones to infection by major viruses in Uganda. Crop Prot.115, 104–112. doi: 10.1016/j.cropro.2018.09.015
48
MurashigeT.SkoogF. (1962). A revised medium for rapid growth and bio assays with tobacco tissue cultures. Physiol. Plant15, 473–497. doi: 10.1111/j.1399-3054.1962.tb08052.x. PMID:
49
NamudduA.MalkaO.SealS.van BrunschotS.KabaaluR.OmongoC.et al. (2025). Is polyphagy of a specific cryptic Bemisia tabaci species driving the high whitefly populations on cassava in eastern Africa? J. Pest Sci.98, 853–868. doi: 10.1007/s10340-024-01832-8. PMID:
50
NarayananN.BeyeneG.ChauhanR. D.GrusakM. A.TaylorN. J. (2021). Stacking disease resistance and mineral biofortification in cassava varieties to enhance yields and consumer health. Plant Biotechnol. J.19, 844–854. doi: 10.1111/pbi.13511. PMID:
51
NarvaK.ToprakU.AlyokhinA.GrovesR.Jurat-FuentesJ. L.MoarW.et al. (2025). Insecticide resistance management scenarios differ for RNA-based sprays and traits. Insect Mol. Biol.34, 518–526. doi: 10.1111/imb.12986. PMID:
52
NiM.MaW.WangX.GaoM.DaiY.WeiX.et al. (2017). Next-generation transgenic cotton: pyramiding RNAi and Bt counters insect resistance. Plant Biotechnol. J.15, 1204–1213. doi: 10.1111/pbi.12709. PMID:
53
NitnavareR. B.BhattacharyaJ.SinghS.KourA.HawkesfordM. J.AroraN. (2021). Next generation dsRNA-based insect control: success so far and challenges. Front. Plant Sci.12. doi: 10.3389/fpls.2021.673576. PMID:
54
Obong’oO. E.SainaE.KinyuaM.MainaF. (2024). Social economic factors driving cassava adoption in Teso South, Busia County, Kenya. East. Afr. J. Agric. Bio/Technol.7, 468–483. doi: 10.37284/eajab.7.1
55
OmongoC. A.KawukiR.BellottiA. C.AlicaiT.LeggJ. P.BagumaY.et al. (2012). Resistance to African cassava whitefly, Bemisia tabaci, in African and South American cassava genotypes. J. Integr. Agric.11, 327–336. doi: 10.1016/s2095-3119(12)60017-3. PMID:
56
OtimM.LeggJ.KyamanywaS.PolaszekA.GerlingD. (2005). Occurrence and activity of Bemisia tabaci parasitoids on cassava in different agro-ecologies in Uganda. Biocontrol50, 87–95. doi: 10.1007/s10526-004-0822-4. PMID:
57
PalliS. R. (2023). RNAi turns 25: contributions and challenges in insect science. Front. Insect Sci.3. doi: 10.3389/finsc.2023.1209478. PMID:
58
ParryH.KalyebiA.BianchiF.SseruwagiP.ColvinJ.SchellhornN.et al. (2020). Evaluation of cultural control and resistance‐breeding strategies for suppression of whitefly infestation of cassava at the landscape scale: a simulation modeling approach. Pest Manag. Sci.76, 2699–2710. doi: 10.1002/ps.5816
59
RasbandW. S. (1997–2019). ImageJ (Bethesda, Maryland, USA: U.S. National Institutes of Health). Available online at: https://imagej.nih.gov/ij/.
60
RosaC.KuoY. W.WuriyanghanH.FalkB. W. (2018). RNA interference mechanisms and applications in plant pathology. Annu. Rev. Phytopathol.56, 581–610. doi: 10.1146/annurev-phyto-080417-050044. PMID:
61
RoscheE.WesthoffP.TaylorW. C. (1994). Analysis of the pyruvate orthophosphate dikinase gene in Flaveria species: Evidence for a single gene encoding both C4 and non-C4 isoforms. Plant Mol. Biol.25, 729–738. doi: 10.1104/pp.117.3.821. PMID:
62
SamW.MorrisO. S.TomO.PatrickO.ColvinJ.AbuO. C. (2024). Resistance to cassava whitefly (Bemisia tabaci) among Eastern and Southern African elite cassava genotypes. Insects15, 258. doi: 10.3390/insects15040258. PMID:
63
SegattoR.JonesT.StretchD.AlbinC.ChauhanR. D.TaylorN. J. (2022). Agrobacterium mediated genetic transformation of cassava. Curr. Protoc.2, e620. doi: 10.1002/cpz1.620. PMID:
64
ShakesbyA. J.WallaceI. S.IsaacsH. V.PritchardJ.RobertsD. M.DouglasA. E. (2009). A water-specific aquaporin involved in aphid osmoregulation. Insect Biochem. Mol. Biol.39, 1–10. doi: 10.1016/j.ibmb.2008.08.008. PMID:
65
SpitJ.PhilipsA.WynantN.SantosD.PlaetinckG.Vanden BroeckJ. (2017). Knockdown of nuclease activity in the gut enhances RNAi efficiency in the Colorado potato beetle, Leptinotarsa decemlineata, but not in the desert locust, Schistocerca gregaria. Insect Biochem. Mol. Biol.81, 103–116. doi: 10.1016/j.ibmb.2017.01.004. PMID:
66
TadmorE.JuravelK.MorinS.Santos-GarciaD. (2022). Evolved transcriptional responses and their trade-offs after long-term adaptation of Bemisia tabaci to a marginally suitable host. Genome Biol. Evol.14, evac118. doi: 10.1093/gbe/evac118. PMID:
67
TaningC. N. T.GuiS.De SchutterK.JahaniM.CastellanosN. L.ChristiaensO.et al. (2021). A sequence complementarity-based approach for evaluating off-target transcript knockdown in Bombus terrestris, following ingestion of pest-specific dsRNA. J. Pest Sci.94, 487–503. doi: 10.1007/s10340-020-01273-z. PMID:
68
ThakurN.UpadhyayS. K.VermaP. C.ChandrashekarK.TuliR.SinghP. K. (2014). Enhanced whitefly resistance in transgenic tobacco plants expressing double stranded RNA of v-ATPase A gene. PloS One9, e87235. doi: 10.1371/journal.pone.0087235. PMID:
69
ThompsonS. N. (2003). Trehalose - the insect “blood” sugar’. Adv. Insect Physiol.31, 205–285. doi: 10.1016/S0065-2806(03)31004-5
70
ThorntonB.BasuC. (2011). Real-time PCR (qPCR) primer design using free online software. Biochem. Mol. Biol. Educ.39, 145–154. doi: 10.1002/bmb.20461. PMID:
71
Van EkertE.ChauvignéF.FinnR. N.MathewL. G.HullJ. J.CerdàJ.et al. (2016). Molecular and functional characterization of Bemisia tabaci aquaporins reveals the water channel diversity of hemipteran insects. Insect Biochem. Mol. Biol.77, 39–51. doi: 10.1016/j.ibmb.2016.07.010. PMID:
72
VyskočilováS.TayW. T.van BrunschotS.SealS.ColvinJ. (2018). An integrative approach to discovering cryptic species within the Bemisia tabaci whitefly species complex. Sci. Rep.8, 10886. doi: 10.1038/s41598-018-29305-w
73
WangT. Y.LuanJ. B. (2023). Silencing horizontally transferred genes for the control of the whitefly Bemisia tabaci. J. Pest Sci.96, 195–208. doi: 10.1007/s10340-022-01492-6. PMID:
74
WangP.SunD. B.QiuB. L.LiuS. S. (2011). The presence of six cryptic species of the whitefly Bemisia tabaci complex in China as revealed by crossing experiments. Insect Sci.18, 67–77. doi: 10.1111/j.1744-7917.2010.01381.x. PMID:
75
WillowJ.VeromannE. (2022). Integrating RNAi technology in smallholder farming: accelerating sustainable development goals. Front. Sustain. Food. Syst.6. doi: 10.3389/fsufs.2022.868922
76
WintraubeD.SadehY.Aidlin-HarariO.AmorY.MorinS.MalkaO. (2025). Osmoregulation by sucrose isomerization in the phloem-feeding whitefly Bemisia tabaci involves members of the glycoside hydrolase family 13. Insect Biochem. Mol. Biol.178, 104266. doi: 10.1016/j.ibmb.2025.104266. PMID:
77
XiaJ.GuoZ.YangZ.HanH.WangS.XuH.et al. (2021). Whitefly hijacks a plant detoxification gene that neutralizes plant toxins. Cell.184, 1693–1705. doi: 10.1016/j.cell.2021.02.014. PMID:
78
YaoH.XuH.YangJ.MaW. (2025). Assessment of potential toxic effects of RNAi-based transgenic cotton on the non-target predator Harmonia axyridis. Biology14, 1173. doi: 10.3390/biology14091173. PMID:
79
ZhongY. W.FanY. Y.ZuoZ. Q.ShuR. G.LiuY. Q.LuanJ. B.et al. (2023). A chromosome-level genome assembly of the parasitoid wasp Eretmocerus hayati. Sci. Data10, 585. doi: 10.1038/s41597-023-02450-2. PMID:
80
ZhuK. Y.PalliS. R. (2020). Mechanisms, applications, and challenges of insect RNA interference. Annu. Rev. Entomol.65, 293–311. doi: 10.1146/annurev-ento-011019-025224. PMID:
81
ZidonR.TsuedaH.MorinE.MorinS. (2016). Projecting pest population dynamics under global warming: the combined effect of inter- and intra-annual variations. Ecol. Appl.26, 1198–1210. doi: 10.1890/15-1045.1
82
ZiererW.AnjanappaR. B.LammC. E.ChangS.-H.GruissemW.SonnewaldU. (2022). A promoter toolbox for tissue-specific expression supporting translational research in cassava (Manihot esculenta). Front. Plant Sci.13. doi: 10.3389/fpls.2022.1042379. PMID:
Summary
Keywords
African cassava whitefly, biological processes, cassava, population dynamics modeling, resistance, RNA interference (RNAi)
Citation
Narayanan N, Swamy RAR, Gehan J, Jones T, Lazar S, Wintraube D, Yakir E, Hasson O, Lampert A, Colvin J, Taylor NJ, Morin S and Malka O (2026) Creating resistance to the whitefly Bemisia tabaci in cassava through RNAi-mediated targeting of multiple insect metabolic processes. Front. Plant Sci. 17:1822258. doi: 10.3389/fpls.2026.1822258
Received
03 March 2026
Revised
08 April 2026
Accepted
15 April 2026
Published
22 May 2026
Volume
17 - 2026
Edited by
Alla Yemets, National Academy of Sciences of Ukraine (NAN Ukraine), Ukraine
Reviewed by
Hong Fang, University of Maryland, College Park, United States
Rachid Hanna, University of California, Los Angeles, United States
Updates
Copyright
© 2026 Narayanan, Swamy, Gehan, Jones, Lazar, Wintraube, Yakir, Hasson, Lampert, Colvin, Taylor, Morin and Malka.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Nigel J. Taylor, NTaylor@danforthcenter.org; Shai Morin, shai.morin@mail.huji.ac.il; Osnat Malka, osnat226@gmail.com
†These authors have contributed equally to this work
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.