Introduction

Sugarcane (Saccharum spp) is a major crop widely grown for its high-sucrose culms used to produce sugar, ethanol and other by-products. Brazil is the largest sugarcane producer in the world with an annual production of around 620 million tons (CONAB, 2020). Owing to the tropical nature of the crop, sugarcane is vulnerable to several insect pests, with Sphenophorus levis (Coleoptera: Curculionidae), particularly in Brazil affecting productivity up to 30 tons or more/ha. The insect larvae feed on the sugarcane clumps, boring irregular galleries causing plant death (Fonseca et al. 2012; Santos et al. 2015). Unfortunately, current control measures are largely inefficient necessitating novel control strategies to combat this insect pest.

Transgenic technology has been exploited widely to develop insect-resistant crops. Among the transgenic approaches, Bt-technology has garnered worldwide success, in which genes encoding for the crystal toxin from Bacillus thuringiensis are genetically engineered into plants, particularly for resistance against lepidopteran and coleopteran pests. However, cases of Bt-resistance have been reported in the target insect pests necessitating search of advanced strategies to control insect pests (Tabashnik et al. 2013). Apart from Bt, transgenic crops expressing cystatins have been reported to enhance insect resistance resulting in reduced crop losses (Chen et al. 2014; Nogueira et al. 2012; Ribeiro et al. 2006). Earlier, we overexpressed a sugarcane cystatin gene (CaneCPI-1) in sugarcane and the resulting transgenic lines showed reduced growth and survival of S. levis larvae when fed on them. The transgenic lines also exhibited reduced stem damage caused by larval attack (Schneider et al. 2017).

RNA interference (RNAi), or gene silencing, performs two major functions in plant, defense against viruses, transposons and aberrant genes (endogenous or introduced transgenes) and regulation of endogenous genes involved in plant growth, development and stress responses (Pattanayak et al. 2013). RNAi-mediated approach has been widely utilized as an efficient tool in deciphering gene functions and crop improvement (Saurabh et al. 2014). Gene function studies in insects have led to the discovery of essential genes that take part in pivotal roles in growth and development (Tomoyasu and Denell 2004). RNAi-mediated silencing of such key genes led to lethal phenotypes in insects indicating that these genes could be targeted for insects–pests control (Asokan et al. 2013; Wang et al. 2011).

Based on transcriptomic analysis of S. levis larvae (unpublished data), we identified V-ATPase E-subunit gene as one of the key genes that are essential for the insect’s growth and survival. V-ATPase is a highly conserved enzyme that acts by pumping protons that lead to acidification of organelles such as lysosomes and endosomes in all eukaryotic cells. V-ATPases are also found in specialized animal tissues, and in insects they play an important role in maintaining the pH gradient in the intestinal lumen and also regulate the secretion of fluids in Malpighi's tubules (Wieczorek et al. 2009). Despite being a gene having a highly conserved sequence, the severity of the phenotype observed for gene silencing in Diabrotica virgifera will serve as a positive control for the tests in S. levis.

We chose the subunit E of the enzyme V-ATPase for cloning and silencing tests, since it has less homology with other insects. In this study, an attempt was made to test the dsRNAs of V-ATPase E on S. levis larvae to evaluate the efficacy of silencing and its impact on the insect survival. Based on these results, we developed RNAi sugarcane lines expressing V-ATPase E dsRNA and demonstrated that the resulting RNAi lines were resistant to S. levis larvae showing reduced stem damage, reduction in larval weight and increased mortality.

Materials and methods

Insect culture and diet

Larvae of S. levis were cultured on artificial diet described by Degaspari et al. (1987). The cultures were maintained in a versatile environmental test chamber (Model MLR352, Panasonic, Japan) with 70% relative humidity and 12-h photoperiod at 26 °C.

Cloning and in vitro transcription of dsRNAs

For cloning, the S. levis V-ATPase E gene fragment was amplified from cDNAs, using the specific primers containing restriction sites for the enzymes BamHI and PstI in the sense and antisense orientation, respectively. As a control for the experiments, dsRNA synthesized from the gene encoding the neomycin phosphotransferase nptII (Kanamycin resistance) was used with primers sense and antisense containing restriction sites for XbaI and XhoI, respectively and the primers used for cloning are given in Table 1. The amplicons were individually cloned in pLITMUS 28i (New England Biolabs, Beverly, MA, USA) and confirmed by Sanger sequencing. The sequence was submitted in NCBI with Ac. No MT941867.

Table 1 Primer sequences used for PCR screening, cloning and real-time PCR; the restriction sites are underlined
Full size table

The dsRNAs were prepared using the in vitro transcription system MEGAscript® T7 High Yield Transcription Kit (Ambion, Grand Island, NY, USA) according to the manufacturer’s instructions. T7-RNA polymerase was used in transcription reactions, with target template linearized with appropriate restriction enzymes to generate single-stranded RNA (ssRNA). Each ssRNA was precipitated by adding equal amounts of lithium chloride and nuclease-free water and re-suspended in 50 μl 20 mM Tris, pH 7.0. Aliquots of ssRNA were analyzed by agarose gel electrophoresis to check the integrity of samples. Finally, equal amounts of ssRNA were added together and annealed by heating the reaction to 80 °C and allowing it to cool to room temperature overnight.

The resulting dsRNAs were quantified using Nanodrop ND-1000 (Uniscience, USA) and precipitated, as described earlier, for concentration of samples. The dsRNA was re-suspended in 20 mM Tris, pH 7.0.

Bioassay with dsRNA injection

The dsRNA injections in S. levis were performed by injecting 10-d old larvae under an MZ6 stereomicroscope (Leica, Germany) using the FemtoJet injector (Eppendorf, Germany). The capillaries were connected to the injector bayonet, which was manually manipulated, without fixing it on a support. To facilitate the visualization of the injections, 49 μl of dsRNA was mixed with 1 μl of food color (MC Cornich, USA). Injection volumes ranged from 0.25 to 0.5 μl, and approximately 0.5–1 μg of dsRNA was injected in the dorsolateral region of the posterior half of S. levis larvae. After the injections, the larvae were kept in the same flasks, containing artificial diet, at room temperature (24–28 ºC), under natural photoperiod.

Quantitative real-time PCR

The qRT-PCR was performed to determine expression levels of V-ATPase E during S. levis development, in different tissues of 30-day-old larvae, transgenic sugarcane leaves, larvae in bioassay injections and inoculated in the transgenic plants.

For developmental stages (eggs, 10-, 20- and 30-day-old larvae, pre-pupa, pupa and adults) and insect parts (integument, head, midgut, hindgut, fat body, hemolymph) V-ATPase assessment, material from three individuals was pooled for RNA extraction. To assess V-ATPase suppression in larvae injected with dsRNA, midgut from three larvae was dissected and pooled for RNA extraction. For gene expression of transgenic sugarcane, approximately 600 ng of fully developed leaves were used. To evaluate the efficacy of the RNAi suppression in transgenic sugarcane lines, the larvae inoculated in the transgenic and control plants were collected for qPCR evaluation. Each larva was individually dissected and the gut was isolated manually and washed in phosphate buffer thrice. Total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA), following the manufacturer's instructions. The extracted RNAs were quantified on a Nanodrop ND-1000 (Uniscience, USA) and their integrity was verified on a 1.5% agarose gel ran using 0.5 X TBE buffer. Each RNA sample was individually treated with RNAse-free DNAse I (Invitrogen, Carlsbad, CA, USA) and cDNA synthesis was performed using the Improm II Reverse Transcription System kit (Promega, Madison, WI, USA), 600 ng of RNA per reaction and 0.5 μg of Oligo dT (× 15) in a final volume of 20 μl. The primers used for qRT-PCR were designed using the Primer 3 v4.0 program (www.frodo.wi.mit.edu/primer3) with specific parameters for qRT-PCR (Table 1). RT-PCR reactions were performed using 5 μL Platinum SYBR Green qPCR UDG Supermix (Invitrogen, Carlsbad, CA, USA) mixture, 1 μL of cDNA diluted ten times and 0.4 μM of forward and reverse primers in 10 μL of final volume. The reactions were carried out in the Pixo Real-Time PCR System (Helixis Pixo, California, USA) and the reaction cycle used: 50 °C for 2 min, 95 °C for 2 min, followed by 40 cycles of 95 °C for 30 s, 54 °C for 30 s and 72 °C for 40 s. For the melting curve, the samples were cycled at 95 °C, 55 °C and 95 °C for 15 s at each temperature.

β-actin and GAPDH genes were used as calibrators to confirm silencing in S. levis larvae injected with dsRNA for V-ATPase E. For sugarcane, GAPDH and ubiquitin 1 (UBQ1) genes were used as references (Andrade et al. 2017). Quantitative RT-PCR analyses to assess differential expression were performed using the 2−ΔΔCT method (Livak and Schmittgen 2001). The measurements were taken in triplicates (technical replicates) and standard deviations are indicated with error bars in the figures. For all reactions, no template and RNA treated with DNAse negative controls were included.

Generation of transgenic sugarcane plants

For raising RNAi transgenic sugarcane, a 334 bp cDNA fragment of V-ATPase subunit E from S. levis was amplified and cloned as hairpin into pANIC8B vector (Mann et al. 2012) by Gateway technology (Invitrogen, Carlsbad, CA, USA). The primers used for the vector construction are listed in Table 1. The recombinant plasmid pANIC8B-vATP was verified by Sanger sequencing and mobilized into Agrobacterium strain EHA105 by freeze thaw method (Sambrook et al. 1989).

Sugarcane cultivar SP80-3280, a Brazilian commercial hybrid, was used for stable plant transformation. Agrobacterium-mediated sugarcane transformation procedure described by Arvinth et al. (2010) was followed and transformed calli were selected on media containing 30 mg L−1 hygromycin. After 60 days of stringent selection, embryogenic calli that survived were transferred to regeneration media (supplementary Table 1) and incubated at 16 h photoperiod in a Phytotron chamber. Fully grown transgenic sugarcane lines were transferred to rooting media (supplementary Table 1). Rooted plants were transferred to pots and maintained in green house until further analysis.

PCR screening of RNAi lines

The presence of V-ATPase RNAi cassette in the presumed transgenic lines was confirmed by PCR. Genomic DNA was extracted from sugarcane leaves using the CTAB method (Murray and Thompson 1980), and 100 ng was used per reaction as a template for amplification. Primers were designed to amplify the V-ATPase gene fragment with an expected amplification product of 334 bp. PCR was performed in the Veriti thermo cycler (Applied Biosystems, CA, USA) under the following conditions: 94 °C for 5 min, 35 cycles at 94 °C for 45 s, 57 °C for 30 s, 72 °C for 45 s and final extension at 72 °C for 7 min.

Insect bioassays

Five transgenic lines with high transcript expression and untransformed control sugarcane plants were vegetatively propagated and maintained in a greenhouse until they reached a height of 2 m. Feeding bioassays were performed as described by Schneider et al. (2017). Five transgenic lines with ten plants per line were used for the in vivo feeding bioassay and to each plant, a S. levis larva was inoculated. After 30 days, the plants were analyzed for the damage caused by larval feeding. The weight of the larvae before inoculation and after 30 days was determined and the stem damage was photographed. The larvae were then collected for qRT-PCR analysis.

Statistical analysis

Statistical analysis and graphs were made with the aid of GraphPad Prism 6.01 for Windows.

Results

V-ATPase E relative in vivo expression

V-ATPase E expression during insect development and in different parts of the larvae was analyzed. V-ATPase E expression during the insect development does not show high levels of variation (Fig. 1a). Regarding the analysis in different parts of the larvae, V-ATPase was expressed in all parts of the insect, presenting high levels of expression in midgut and hindgut (Fig. 1b), and low expression in other parts. The higher levels were used as calibrator (1x).

Fig. 1
figure 1

Relative expression of V-ATPase E in S. levis. a: Relative V-ATPase E expression levels in developmental stages. Highest expression levels are detected for adults and were considered 1x (one fold). b: V-ATPase expression levels in different parts of 30-d-old larvae

Full size image

Bioassay with dsRNA injection

The bioassays of S. levis larvae injected with endogenous V-ATPase E dsRNA showed that four out of six larvae died within three weeks of injection indicating 66.6% mortality. However, the control larvae injected with kanamycin dsRNA developed normally reaching the pupal and adult stages (Fig. 2).

Fig. 2
figure 2

Sphenophorus levis larvae injected with kanamycin dsRNA. The larva (a) and pupa (b) developed normally. Only one insect reached adulthood (c), due to growing conditions. d Larvae of S. levis after 44 days of treatment with V-ATPase E dsRNA. Four of the six larvae injected with V-ATPase dsRNA died by day 20 after the injections (n = 6)

Full size image

Quantitative real-time PCR

The effects of dsRNA injection in S. levis were evaluated by qRT-PCR. The qRT-PCR analysis revealed that S. levis larvae injected with V-ATPase E dsRNA showed 33.75% and 46% reduction in transcription levels using the reference genes GAPDH and β-actin, in relation to larvae injected with kanamycin dsRNA (Fig. 3).

Fig. 3
figure 3

Gene silencing analysis of S. levis V-ATPase using the GAPDH (a) and β-actin (b) genes as reference genes. The reduction in relative transcription levels in relation to insects injected with kanamycin dsRNA was 33.75% and 46%, respectively. The black bars depict V-ATPase and gray bars show kanamycin expression, respectively. The levels of transcription detected for non-injected insects (Day 0), used for normalization, is indicated as 1. Error bars indicate standard deviation. Comparison was calculated using ordinary one-way ANOVA corrected by Dunnet test. *** means P ≤ 0.001; ** means P ≤ 0.01

Full size image

Generation and selection of sugarcane transgenic lines

After selection with hygromycin, 61 plants were regenerated and used for PCR analysis. Of these, 52 were positive for the V-ATPase E gene, which was verified by the presence of an expected amplicon size of 334 bp (Suppl Figure 1). No amplification was observed for the non- transformed control plant.

V-ATPase E expression levels were higher for Lines 7,11,17,19 and 39 (Fig. 4a and b), which were chosen for vegetative propagation. A wild type plant was used for control and showed no expression of V-ATPase E. Line 19 was used for normalization as 1x (one-fold) V-ATPase expression, since it was the line that presented less expression.

Fig. 4
figure 4

Quantitative RT-PCR analysis of transgenic lines. Relative expression. Levels of V-ATPase using (a) GAPDH and (b) UBQ1 as reference genes. As wild type does not express V-ATPase, line 19, which showed the lowest V-ATPase expression levels, was used for normalization as 1x. Standard error bars are indicated in the figure

Full size image

V-ATPase E expression levels were higher for Lines 7,11,17,19 and 39 (Fig. 4a, b), which were chosen for vegetative propagation. A wild type plant was used for control and showed no expression of V-ATPase. Line 19 was used for normalization as 1x (one-fold) V-ATPase expression, since it was the line that presented less expression.

Insect bioassays

In vivo feeding bioassays were conducted to determine the effect of S. levis larvae on transgenic and control lines. The results indicated that the larval mortality was higher in the transgenic lines when compared to control plants. Lines 7 and 39 recorded 70% larval mortality followed by Line 11 (60%). Lines 17 and 19 showed 40% larval mortality (Fig. 5a) However, the larvae fed on control plants recorded only 20% mortality.

Fig. 5
figure 5

Resistance of sugarcane transgenic lines on bioassays with S. levis larvae. a Larvae growth (mg) after 30 days. The dead larvae were represented as 0 growth. b Percent mortality of larvae fed on transgenic pANI8B-V-ATPase lines (n = 10). c Extent of damage recorded in transgenic and control lines after feeding bioassay. The black line indicates 1 cm height

Full size image

Thirty days after feeding bioassay the larval weight was measured which indicated that the larvae fed on Lines 11 and 39 showed reduced weights compared to that of control plants (Fig. 5b). Overall, all the five lines recorded lower larval weights compared to control plants. The larval bioassay indicates that Line 39 showed prominent results in affecting growth and development of S. levis larvae. Figure 4c shows the reduced extent of damage in lines 7 and 39 compared to control.

RT-PCR of larvae fed on RNAi lines

RT-PCR was performed to evaluate the V-ATPase expression in the larvae fed on transgenic and control plants. Our results showed significant suppression of V-ATPase expression in larvae collected from four out of five lines. Larvae from Lines 7, 19 and 39 showed significantly higher suppression compared to control plants. Larva from Line 17 exhibited 0.5-fold suppression compared to control. Larva collected from Line 11 did not show any gene suppression (Fig. 6).

Fig. 6
figure 6

Relative V-ATPase expression in larvae fed on transgenic lines compared to control sugarcane. Error bars indicate standard deviation. Comparison was calculated using ordinary one-way ANOVA corrected by Dunnet test. *** means P ≤ 0.001

Full size image

Discussion

Being a tropical crop, sugarcane is highly susceptible to several insect pests. S. levis or sugarcane weevil, is an important insect pest that hinders sugarcane production in Brazil to a great extent. Sphenophorus levis infestation causes productivity losses of more than 30 tons per hectare. As the larvae thrive inside the sugarcane stem base and the adults live below the soil level, chemical insecticides cannot penetrate efficiently. The existing control methods are largely ineffective for the control of this insect (OECD 2016). Alternative strategies employing advanced tools are necessary to control S. levis. Moreover, the genetic complexity of sugarcane coupled with the non-availability of recognized natural resistance in the germplasm has made conventional breeding for insect resistance difficult.

With the availability of robust tissue culture protocols and genetic transformation procedures for sugarcane, transgenic technology has emerged to be a promising strategy for pest control in this crop (Gao et al. 2016; Wang et al. 2017). Transgenic sugarcane lines resistant to insect pests have been successfully developed using Bt-genes (Arencibia et al. 1997; Arvinth et al. 2010), proteinase inhibitors (Nutt et al. 1999), lectin (Zhangsun et al. 2007) and aprotinin (Christy et al. 2009). Similarly, recent years have demonstrated that RNAi or gene silencing technology has a high potential in generating insect tolerant crops. The key factors that are critical for efficient RNAi are the insect target gene selection and the delivery method (Mamta and Rajam 2017). Genes that have essential functions and are required for various life stages of insects are generally considered ideal for RNAi (Baum et al. 2007; Mao et al. 2007).

Transcriptomic analysis of S. levis performed in our laboratory revealed several key genes that could be possible targets for RNAi (unpublished data). In insects, vacuolar ATPases have been well studied and are known for their role in the processes of osmoregulation, maintenance of the pH gradient and difference in membrane potential as a facilitator of the absorption of solutes and amino acids (Beyenbach and Wieczorek 2006). They are highly conserved proteins among insects. The silencing of a subunit of the V-ATPase should generate the severe and lethal phenotypes already known in insects, as described by Baum and collaborators (2007) and Whyard et al. (2009), serving as a control to evaluate the sensitivity of S. levis to RNAi. In the S. levis cDNA library, only one singlet was identified for each of the A and E subunits of the V-ATPase enzyme, which indicates their lower expression in relation to digestive enzymes. Feeding tests with the Coleoptera Diabrotica virgifera showed similar results for silencing of the enzyme A and E subunits (Baum et al. 2007). In the initial analyses of similarity, the E-subunit of the S. levis V-ATPase showed less identity against the homologous enzyme of T. castaneum (GenBank: XM_965528.1) and was, therefore, chosen for this study aiming at the absence of undesirable effects on other insects.

Bioassays using dsRNA injections of S. levis larvae revealed significant mortality (66%) and the expression analysis revealed reduction in V-ATPase transcripts accumulation indicating that V-ATPase is a potential target for RNAi in S. levis. Rangasamy and Siegfried (2012) reported that D. virgifera when continuously fed on cucurbitacin and V-ATPase dsRNA-treated artificial diet, resulted in more than 95% adult mortality within 2 weeks, whereas mortality in control treatments were below 20%. Similarly, microinjection of V-ATPase dsRNA fragments into pink bollworm larvae led to mortality of 18.9–26.7% (Mohammed 2016). Injection of V-ATPase A and E dsRNAs in bed bugs decreased their survival over 30 days (Basnet and Kamble 2018).

V-ATPase is stably expressed during all stages of insect development. In addition, analysis of the gene expression in several parts of 30-day old larvae reveals it is present in all the insect parts, with different expression levels. Such stability is expected, considering the enzyme’s contitutive role in regulating ion transport in tissues, thus controlling pH. In the S. levis gut, where the protein is highly expressed, H+ active flow through cell membranes is the major mechanism of pH control, besides influencing compounds uptake (Wieczorek et al. 1999).

Based on the dsRNA injection experiment and V-ATPase E expression pattern, we developed RNAi sugarcane lines expressing V-ATPase dsRNA and feeding bioassays performed on five RNAi independent lines showed higher mortality in the transgenic lines compared to control plants. In addition, the weight of larvae was reduced after feeding indicating the suppression of their growth. Moreover, the stem damage was significantly smaller than that of the controls. Our data on feeding bioassays suggest that the transgenic sugarcane lines expressing V-ATPase dsRNA can form sufficient siRNA and could cause mortality of S. levis larvae, thereby providing resistance to this beetle. The qRT-PCR results reiterate the feeding bioassay data which showed knocking down of V-ATPase expression in four of the five lines tested. Line 11 although recorded mortality and reduced larval weight in bioassasays, did not show any suppression in qRT-PCR which may be attributed to the level of down-regulation being too little or transient, to be detected. Transgenic tobacco plants expressing V-ATPase dsRNA delivered sufficient siRNA to whiteflies feeding on them, causing a significant silencing response, leading to their mortality (Thakur et al. 2014). The transcript expression of V-ATPase was reduced in whiteflies feeding on transgenic plants. Liu et al. (2019) reported a similar finding wherein the transgenic maize and soybean lines expressing V-ATPase E dsRNA suppressed the development of A. lucorum upon feeding and could decrease their population.

In conclusion, we identified V-ATPase subunit E as a potential gene target for RNAi-mediated suppression of S. levis larvae. Injections of V-ATPase E dsRNA into S. levis larvae caused significant mortality. Transgenic sugarcane lines expressing V-ATPase dsRNA could cause silencing in the larvae fed on them leading to significant mortality when compared to the control plants. Our data show that plant-mediated RNAi is a promising approach to impart resistance to sugarcane against S. levis. To our knowledge, this is probably the first report of effective plant-mediated RNAi against S. levis.