A rapid virus-induced gene silencing (VIGS) method for assessing resistance and susceptibility to cassava mosaic disease

VIGS clones were generated from the virulent infectious clone EACMV-K201 described previously by Patil and Fauquet [20]. EACMV-K201 was produced from East African cassava mosaic virus (EACMV-KE[KE:Msa:K201:02]), DNA-A GenBank: AJ717541; and DNA-B GenBank: AJ704953 [21]. The EACMV-K201 DNA-A infectious clone was digested with Hin dIII/Eco RI (2082 bp) and Eco RI/Bam HI (1157 bp), and cloned into pBlueScript vector (Stratagene) as Hin dIII/Bam HI in a three-way ligation. The resulting construct was named p8200. p8200 was modified to introduce restriction sites Nhe I and Avr II near the 5′-region, and Sbf I near the 3′-region within the coding sequence of the coat protein (CP) gene using QuickChange Multi Site-Directed Mutagenesis Kit (Agilent Technologies, Inc.) with the primers listed in Table 2. The mutagenized infectious clone was confirmed by sequencing and named p8202.

The Manihot esculenta homolog of A. thaliana SPINDLY (SPY) gene (MeSPY1 accession number Manes.09G052300.1) was cloned into p8202 at the CP site. A 452 bp fragment of MeSPY1 (406–857 counted from start codon ATG) was amplified from plasmid p8103 that harbors the 2781 bp coding sequence of MeSPY1 by introducing restriction sites Nhe I and Sbf I on the forward and reverse primer pairs, respectively (Table 2). The PCR product was cloned into Zero Blunt Topo (Invitrogen) first and positive clones digested and cloned into the Nhe I/Sbf I site of p8202 to generate the p8250 (hereafter named MeSPY1-VIGS). A non-target VIGS control was produced by amplification of 453 bp (235–687 bp counted from start codon ATG) from the erGFP [22] sequence of a binary vector erGFP-pCAMBIA2300 by introducing Nhe I and Sbf I sites on forward and reverse primer pairs, respectively, and cloned into p8202 to generate p8223 (GFP-VIGS). In order to generate a VIGS vector targeting cassava phytoene synthase (MePSY2 gene, Manes.01G124200.1), a 452 bp fragment (350–801 bp counted from start codon ATG) was amplified introducing a Sbf I and Nhe I restriction site and cloned into the Sbf I and Nhe I site of 8202 to produce p8375 (MePSY2-VIGS).

Cassava cultivars used in this study are shown in Table 1 and included the CMD-susceptible cultivars 60444, Ebwanateraka and TME 7S; and CMD-resistant cultivars TME 3, TME 7 (Oko-iyawo), TME 204 and TMS 98/0505. Additional lines included CMD-susceptible plants [15] regenerated from friable embryogenic callus (FEC) of TME 204, TME 3 and TME 7 and CMD-resistant plants of cultivar TMS 98/0505 recovered from FEC following the method described by Chauhan, et al. [23]. All plants were micropropagated and established in the greenhouse as described previously [15, 24].

Four- to 6-week-old greenhouse-grown plants were inoculated with plasmid DNA of MeSPY1-VIGS, GFP-VIGS or MePSY2-VIGS vectors plus the DNA-B component of EACMV-K201 using a Helios® Gene Gun (BioRad, Hercules, California), following Beyene, et al. [15]. Approximately 75 ng each of the VIGS (DNA-A and DNA-B) components were used to inoculate each plant. Symptom and phenotype scoring after inoculation was performed in three manners. Plants challenged with GFP-VIGS were scored for development of CMD symptoms using an established visual scoring system with a scale of 0–5 [25]. Plants were scored starting 7–10 days post inoculation (DPI) and every 3–7 days thereafter for a total of 12 weeks [15]. For plants challenged with MePSY2-VIGS, visual assessment was made for presence or absence of chlorosis/bleaching. In the case of plants challenged with MeSPY1-VIGS, a new scoring system was employed by which plants were assessed for death of the shoot apical meristem (whole plant) starting 7–10 DPI and every 2–3 days thereafter for a maximum of 4 weeks. Data was expressed as incidence of plants showing shoot-tip necrosis/dead plants presented as the percent of the total number of plants inoculated at the end of the 4-week observation period.

Plants were sampled for nucleic acid extraction by collecting leaves and shoot material 9 DPI with VIGS constructs. Tissues from three plants were pooled in a 50 mL Falcon tube and immediately flash frozen in liquid nitrogen. A total of four pooled samples collected from 12 plants were analyzed per treatment. Frozen samples were ground in a mortar and pestle to fine powder and freeze-dried overnight. Approximately 20–30 mg of freeze-dried tissue were used for nucleic acid extraction. Genomic DNA was extracted using DNeasy Plant Mini Kit (Qiagen, Hilden, Germany) and the resulting genomic DNA quantified and used for virus titer determination by qPCR and Southern blot analysis [15]. Total RNA was isolated using the Spectrum™ Plant Total RNA Kit, followed by on-column DNAse I treatment per manufacturer instructions (Sigma-Aldrich, St. Louis, MO). One microgram of total RNA was reverse transcribed using the PrimeScript™ RT reagent kit with gDNA Eraser (Takara Bio Inc., Japan). Expression levels of MeSPY in MeSPY1-VIGS-inoculated plants and control GFP-VIGS plants were quantified by RT-qPCR [26] using the primer pairs listed in Table 2, with approximately 10 ng reverse transcriptase template and SsoAdvanced™ Universal SYBER® Green Supermix (Bio-Rad Laboratories Inc.). The cassava PP2A gene was employed as an internal control for RT-qPCR [27].

A VIGS system was developed by modifying the CP nucleotide sequence of the DNA-A component of the virulent infectious clone EACMV-K201 [20] to generate the VIGS construct p8202. Mutations were designed to introduce restriction sites Avr II, Nhe I and Sbf I (Fig. 1a) in order to allow introduction of target sequences. The introduced mutations were confirmed by sequencing both strands of the modified CP sequences as shown (Fig. 1a) and further confirmed by restriction analysis (data not shown). To check infectivity of the p8202 VIGS vector, a 452 bp fragment from the coding sequence of the cassava carotenoid biosynthetic gene phytoene synthase (MePSY2) was cloned at the Nhe I/Sbf I site of p8202 to generate MePSY2-VIGS. In addition, a GFP-VIGS DNA-A component was generated by cloning a 453 bp sequence from erGFP [22] into the restriction sites generated within the CP sequence.

Five-week-old plants of the CMD-susceptible cassava line TME 7S were inoculated by Helios® Gene Gun microparticle bombardment of the modified DNA-A component plus the infectious clone of DNA-B. TME 7S is a CMD-susceptible version of TME 7 previously described by Kuria, et al. [28]. Plants inoculated with MePSY2-VIGS developed visible chlorosis/bleaching on the challenged leaves and then subsequently on systemic leaves within 10–15 days after bombardment (Fig. 1b). Bleaching of leaves persisted throughout the experimental period of 12 weeks (Fig. 1b). Plants challenged with GFP-VIGS (which does not have a target gene sequence within the cassava genome) showed typical but mild CMD symptoms on systemic leaves that persisted throughout the study period (Fig. 1b). Response of inoculated plants to MePSY2-VIGS and GFP-VIGS confirms that the VIGS constructs made by modifying EACMV-K201 are both infectious and efficacious in silencing gene expression in cassava.

Using the Arabidopsis SPY gene (AT3G11540.1) [29, 30] as a bait, two sequences, Manes.09G052300.1 (named MeSPY1) and Manes.08G028400.1 (named MeSPY2), were identified from the cassava v6.1 genome sequence [31]. The MeSPY1 and MeSPY2 sequences are 81% identical to Arabidopsis SPY at the amino acid level. Both MeSPY1 and MeSPY2 carry the conserved N-terminal tetratricopeptide repeat (TPR) domain, plus the novel serine and threonine O-linked N-acetylglucosamine transferase (OGT) [30] at the C-terminus of the protein (data not shown). The coding region of MeSPY1 and MeSPY2 genes are 91.80 and 90.78% identical to each other at the nucleotide and amino acid levels, respectively. The target sequence (452 bp) cloned into the VIGS vector from MeSPY1 was obtained closer to the 5′-end, within the conserved TPR region. As MeSPY1 and MeSPY2 share high sequence identity (94.69%) at this selected region, both can be targeted for simultaneous silencing by MeSPY1-VIGS.

The susceptible cassava cultivar TME 7S was challenged with MeSPY1-VIGS by Helios® microparticle bombardment. Inoculated plants did not show typical CMD symptoms but instead underwent wilting and withering of one or more leaves just above the inoculation site within 10–15 DPI. This was followed by shoot-tip necrosis within 12–21 DPI. Defoliation then progressed downwards to mature leaves below the inoculation site and along the main stem continuing until most of the plants had died within 2–4 weeks after inoculation with MeSPY1-VIGS (Fig. 2a). TME 7S plants challenged with the control GFP-VIGS showed typical CMD symptom development without shoot tip necrosis (Fig. 2a). In order to determine if this was a cultivar-specific response, the CMD2-type cultivar TME 204 (Table 1) was challenged with MeSPY1-VIGS and GFP-VIGS. Two types of TME 204 plants were used: wild-type TME 204, which shows development of CMD symptoms followed by recovery; and FEC-TME 204 lines. The latter TME 204 plants that were recovered through somatic embryogenesis lost their inherent resistance to CMD and showed no recovery from CMD after challenge with CMGs [15]. After challenge with MeSPY1-VIGS, all plants (10/10) of wild-type TME 204 displayed transient reduction in growth compared to the GFP-VIGS plants and then resumed normal growth without showing typical CMD symptoms. No shoot-tip necrosis was observed. In contrast, all FEC-derived TME 204 plants (8/8) displayed a similar phenotype to that described above for TME 7S, whereby elongation of the shoot was reduced, resulting in compact shoot-tips and wilting and withering of leaves just above the inoculation site. This was followed by death of the shoot-tip and wilting and defoliation of leaves that progressed downwards from the youngest to the oldest leaves. Time-lapse video of the process of shoot-tip necrosis and whole plant death in susceptible FEC-TME 204 and survival of the resistant wild-type TME 204 plant lines is presented (Additional file 1). All plants displayed shoot-tip necrosis with most plants dying within 2–4 weeks after challenge (Fig. 2b). Inoculation with GFP-VIGS resulted in plants of both the susceptible and resistant lines of TME 204 developing CMD symptoms. Wild-type TME 204 plants then underwent recovery from CMD to show no symptoms on newly formed leaves by 16 weeks after challenge. FEC-derived plants remained symptomatic throughout, in a manner consistent with previous observations [15]. This data indicates that silencing of MeSPY is lethal in the two susceptible cassava lines TME 7S and FEC-derived TME 204, but not in the resistant wild-type TME 204.

MeSPY1-VIGS’s potential for use as a screening tool to determine resistance or susceptibility to CMD was investigated further by challenging additional cassava cultivars (Table 1). These included: a) CMD2-type resistant wild-type plants of cultivars TME 7 and TME 3, and CMD-susceptible plant lines derived from FEC of TME 7 and TME 3; b) CMD3-type resistant cultivar TMS 98/0505 (wild-type) and FEC-derived TMS 98/0505; and c) CMD-susceptible cassava cultivars 60444 and Ebwanateraka. Inoculation of these cassava accessions with MeSPY1-VIGS showed that CMD-susceptible germplasm (FEC-TME 3, FEC-TME 7, 60444 and Ebwanateraka) underwent wilting of leaves above the inoculation site and necrosis of shoot-tips within 10–21 DPI. This was followed by defoliation, culminating in death of the majority of plants within 4 weeks after inoculation (Figs. 3 and 4). The CMD-resistant wild-type cultivars TME 3, TME 7 and TMS 98/0505 survived challenge with MeSPY1-VIGS and displayed similar phenotype described above for wild-type TME 204 (Figs. 3 and 4). Plants of FEC-derived TMS 98/0505 behaved in a manner identical to wild-type TMS 98/0505 confirming their known resistance to CMD [15]. Likewise, wild-type and FEC-derived plants of CMD1-type cultivars NASE 3 (TMS 30572) and NASE 14 also survive MeSPY-VIGS (data not shown). All plant lines were also challenged with GFP-VIGS. Data collected for development of and recovery from CMD symptoms correlated with that from plants inoculated with MeSPY1-VIGS such that all plant lines that resisted challenge with MeSPY1-VIGS also showed recovery from CMD symptoms after inoculation with GFP-VIGS (Fig. 5).

Expression of MeSPY after inoculation with MeSPY1-VIGS was quantified by RT-qPCR in the CMD-resistant wild-type cultivar TME 7 and susceptible FEC-TME 7 plants at 9 DPI. At this time, plants challenged with the control GFP-VIGS showed systemic CMD symptoms on young leaves above the inoculation site, while the MeSPY1-VIGS leaves were smaller and shoot-tips were compact relative to the GFP-VIGS. Relative expression of the MeSPY gene was found to be significantly reduced (33%; p < 0.05) in FEC-derived TME 7 plants challenged with MeSPY1-VIGS when compared to TME 7 wild-type plants and to the corresponding GFP-VIGS controls (Fig. 5c). Virus DNA titer was determined by qPCR (Fig. 5b) and Southern blot analysis (Fig. 5a) from the same samples collected at 9 DPI. TME 7 wild-type plants challenged with MeSPY1-VIGS and GFP-VIGS had significantly lower relative virus loads (2.36-fold and 7.23-fold, respectively) compared to the FEC-derived CMD-susceptible TME 7 plants (16.08 and 29.39, respectively). Greater virus DNA was detected in GFP-VIGS inoculated lines than MeSPY-VIGS inoculated lines (Fig. 6a). qPCR virus titer data was seen to be consistent with Southern blot data showing relatively more virus titer in the susceptible FEC-derived TME 7 compared to the wild-type TME 7.