Tobacco rattle virus vector: A rapid and transient means of silencing manduca sexta genes by plant mediated RNA interference

Abstract

Background: RNAi can be achieved in insect herbivores by feeding them host plants stably transformed to express double stranded RNA (dsRNA) of selected midgut-expressed genes. However, the development of stably transformed plants is a slow and laborious process and here we developed a rapid, reliable and transient method. We used viral vectors to produce dsRNA in the host plant Nicotiana attenuata to transiently silence midgut genes of the plant's lepidopteran specialist herbivore, Manduca sexta. To compare the efficacy of longer, undiced dsRNA for insect gene silencing, we silenced N. attenuata's dicer genes (NaDCL1- 4) in all combinations in a plant stably transformed to express dsRNA targeting an insect gene.

Methodology/principal findings: Stable transgenic N. attenuata plants harboring a 312 bp fragment of MsCYP6B46 in an inverted repeat orientation (ir-CYP6B46) were generated to produce CYP6B46 dsRNA. After consuming these plants, transcripts of CYP6B46 were significantly reduced in M. sexta larval midguts. The same 312 bp cDNA was cloned in an antisense orientation into a TRV vector and Agro-infiltrated into N. attenuata plants. When larvae ingested these plants, similar reductions in CYP6B46 transcripts were observed without reducing transcripts of the most closely related MsCYP6B45. We used this transient method to rapidly silence the expression of two additional midgut-expressed MsCYPs. CYP6B46 transcripts were further reduced in midguts, when the larvae fed on ir-CYP6B46 plants transiently silenced for two combinations of NaDCLs (DCL1/3/4 and DCL2/3/4) and contained higher concentrations of longer, undiced CYP6B46 dsRNA.

Conclusions: Both stable and transient expression of CYP6B46 dsRNA in host plants provides a specific and robust means of silencing this gene in M. sexta larvae, but the transient system is better suited for high throughput analyses. Transiently silencing NaDCLs in ir-CYP6B46 plants increased the silencing of MsCYP6B46, suggested that insect's RNAi machinery is more efficient with longer lengths of ingested dsRNA.

Figure 1. Selection of M. sexta CYPs for plant mediated RNAi (PMRi) and their spatial expression profiles.

(A) Phylogenetic relationship among M. sexta CYPs (complete ORFs) as calculated by the Clustal-W program. CYPs selected for silencing by PMRi are shaded in gray and their most closely related CYPs that were analyzed for off-target co-silencing are in dashed boxes. Thousand bootstrapping trials were conducted (only the bootstrap values >50 displayed). Transcript levels (relative to ubiquitin) of (B) CYP6B46, (C) CYP6B45, (D) CYP4M1, (E) CYP4M3 and (F) CYP4M2 in hemolymph, Malpighian tubules, fat body, foregut, midgut and hindgut of 5^th instar M. sexta larvae feeding on N. attenuata (WT) plants. Bars labeled with different letters indicate the significant differences as determined by one way ANOVAs (p≤0.05).

Figure 2. PMRi of M. sexta CYP6B46 using stably transformed N. attenuata plants.

(A) Northern hybridizations revealed the presence of small RNAs of CYP6B46 in the leaves of two independent lines of stably transformed plants, ir-CYP6B46 (30-2) and ir-CYP6B46 (416-3), and in the midguts of 4^th instar larvae feeding on these plants. RNA samples from leaves of WT and EV (empty vector transformed stable line) plants and from the midguts of larvae feeding on WT and EV leaves were used as negative controls. Similar fluorescence intensity of the ethidium bromide stained 5.8 S rRNA bands reflected the equal loading of LMW RNA. Low molecular weight RNA from leaf or midgut loaded on the gel in each lane was a pool of three biological replicates. smRNA length of 21 b denoted by marker. Transcript abundance (relative to ubiquitin) of: (B) CYP6B46 (target gene) and closely related (C) CYP6B45 (off-target) in the midguts of 4^th instar larvae. (D) Larval mass of 4^th instar larvae feeding on WT, EV, ir-CYP6B46 (30-2) and ir-CYP6B46 (416-3) N. attenuata plants for 14 days. Asterisk indicates the significant differences as determined by one way ANOVAs (p≤0.05).

Figure 3. Efficiency and specificity of CYP6B46, CYP4M1 and CYP4M3 silencing by viral dsRNA-producing system (VDPS).

Northern hybridization showing the smRNAs of (A) CYP6B46 (B) CYP4M1 and (C) CYP4M3 in WT leaves infiltrated with VDPS-EV, -CYP6B46, -CYP4M1 and -CYP4M3 constructs, respectively, as well as in the midguts of 4^th instar larvae feeding on the leaves of plants inoculated with the respective constructs. Lane M shows 21 b oligonucleotide that was used as size marker as well as a positive control for hybridization. Similar fluorescence intensity of the ethidium bromide stained 5.8 S rRNA bands reflected the equal loading of LMW RNA. Transcript abundance (relative to ubiquitin) of the target genes. (D) CYP6B46 in the midguts of 4^th instar larvae feeding on VDPS-EV and -CYP6B46, (F) CYP4M1 in the midguts of 4^th instar larvae feeding on VDPS-EV, -CYP4M1 and -CYP4M3, and (H) CYP4M3 in the midguts of 4^th instar larvae feeding on VDPS-EV, -CYP4M3 and -CYP4M1 plants, respectively. Transcript abundance (relative to ubiquitin) of closely related, off-target genes (E) CYP6B45 in the midguts of 4^th instar larvae feeding on VDPS-EV and -CYP6B46, (G) CYP4M2 in the midguts of 4^th instar larvae feeding on VDPS-EV and -CYP4M1, and (I) CYP4M2 in the midguts of 4^th instar larvae feeding on VDPS-EV and -CYP4M3 plants, respectively. (J) Mass of 4^th instar M. sexta larvae fed for 14 days on VDPS-EV, -CYP6B46, -CYP4M1 and -CYP4M3 plants. Asterisk indicates the significant differences as determined by one way ANOVAs (p≤0.05).

Figure 4. PMRi efficiency is increased after silencing N. attenuata's dicer-like (DCL) genes.

(A) Schematic representation of the silencing of N. attenuata's four DCLs by virus induced gene silencing (VIGS), in ir-CYP6B46 (30-2) stably-transformed plants. ir-CYP (30-2) plants were Agro-infiltrated with pTVDCL harboring cultures (individually, and in all combinations of DCL1, DCL2, DCL3 and DCL4); pTV (EV) was used as a control. (B) Abundance, relative to NaActin, of a 102 bp region of the 5′ end of the 312 bp MsCYP6B46 fragment that the ir-CYP (30-2) N. attenuata plants were harboring in their genome. The plants had been previously Agro-infiltrated with EV or all combinations of vectors designed to silence the expression of the four NaDCLs. (C) Transcript abundance of CYP6B46 (relative to ubiquitin) in the midguts of 4^th instar M. sexta larvae, when fed N. attenuata leaves containing no MsCYP6B46 dsRNA (WT+EV), small (diced) MsCYP6B46 dsRNA (ir-CYP6B46+EV) and on leaves of plants containing higher concentration of longer (102 bp detected by qPCR) MsCYP6B46 dsRNA fragments (ir-CYP6B46+ DCL134 and ir-CYP6B46+ DCL234). See Fig. S4A and Table S1 for the design of the primers used in the transcript quantification. Bars labeled with different letters indicate significant differences as determined by one way ANOVAs (p≤0.05).