A ligation-independent cloning tobacco rattle virus vector for high-throughput virus-induced gene silencing identifies roles for NbMADS4-1 and -2 in floral development

Abstract

Virus-induced gene silencing (VIGS) is a widely used, powerful technique for reverse genetics. VIGS vectors derived from the Tobacco rattle virus (TRV) are among the most popular for VIGS. We have developed a TRV RNA2 vector that allows the insertion of gene silencing fragments by ligation-independent cloning (LIC). This new vector has several advantages over previous vectors, particularly for applications involving the analysis of large numbers of sequences, since TRV-LIC vectors containing the desired insert are obtained with 100% efficiency. Importantly, this vector allows the high-throughput cloning of silencing fragments without the use of costly enzymes required for recombination, as is the case with GATEWAY-based vectors. We generated a collection of silencing vectors based on 400 tomato (Solanum lycopersicum) expressed sequence tags in this TRV-LIC background. We have used this vector to identify roles for SlMADS1 and its Nicotiana benthamiana homologs, NbMADS4-1 and -2 in flowering. We find that NbMADS4-1 and -2 act nonredundantly in floral development and silencing of either gene results in loss of organ identity. This TRV-LIC vector should be a valuable resource to the plant community.

Figure 1.

TRV2-LIC vector. A, Construction of the TRV2-LIC vector. The TRV2 vector, pYL170, was used to generate the new TRV2-LIC vector by inserting a cassette containing adapters and two PstI sites in two digestion and ligation reactions. B, LIC cloning of inserts into TRV2-LIC. Briefly, the TRV2-LIC vector is digested with PstI and treated with T4 DNA polymerase. The EST carrying the relevant adapter sequences is generated by PCR and also treated with T4 DNA polymerase. The vector and insert are then mixed and used to transform Escherichia coli cells. C, TRV-LIC does not interfere with viral replication and spread as shown by the presence of TRV RNA1 (lane 1) and modified RNA2 (lane 2) in upper leaves of N. benthamiana plant. Lane M contains DNA size marker. D, TRV2-LIC allows efficient silencing of NbPDS as shown by the photobleaching of silenced plant (right). Control plant infiltrated with empty TRV-LIC vector shown for comparison (left). E, EST cloning efficiency into TRV-LIC is 100% as shown by PCR on colonies obtained from transformation with the vector and PCR product mixture. All 48 of the colonies tested here contain insert. Lane M contains DNA size marker.

Figure 2.

Selected phenotypes observed with silencing of tomato ESTs. A, Silencing of SlRibosomal protein L11 like, Slα-tubulin, SlHSP70, and SlPolyubiquitin causes death. B, Silencing SlMAP3K epsilon disrupts leaf development resulting in severe crinkling. C, Reduction in SlGeranylgeranyl reductase levels causes yellowing of leaves. D, Reduced SlPlastidic aldolase causes photobleaching and a variegated appearance. E, Decreased SlCyclin-dependent protein kinase, p34cdc2, levels cause severe stunting. F, Homeotic transformations of floral tissue in SlTAG1-silenced plants. G, mRNA levels in control (left) versus silenced (right) tissue indicate effective silencing of genes indicated. Number of PCR cycles is indicated above lanes. Lane M is the DNA size marker. C is no RT control. Actin is a loading control.

Figure 3.

Silencing SlMADS1. A, SlMADS1-silenced plants (right) show loss of apical dominance compared to control empty vector-treated plants (left). B, Flowers on control empty vector-treated plants show distinct white petals surrounded by sepals. C, Flowers on SlMADS1-silenced plants have enlarged sepals and white petals are largely absent. D, Sepals separated from inner whorls from control (left) and SlMADS1-silenced (right) flowers. E, Magnified view of inner whorls of flower dissected in D. Note reduced white petals and the presence of flower-like organs. F, Indeterminate flowers of SlMADS1-silenced plants. G, Magnified view of one of the flowers in F.

Figure 4.

NbMADS4-silenced petals compared to controls. A, Abaxial surface of control N. benthamiana petal. Note absence of trichomes. B, Abaxial surface of SlMADS1-silenced petal contains both trichomes and stomata (arrows). C, Abaxial surface of control N. benthamiana leaf contains many stomata. Scale bar represents 50 μm.

Figure 5.

Silencing NbMADS4-1 and NbMADS4-2. A, Control N. benthamiana flower. B and C, Silencing NbMADS4-1 using fragment 1 (B) or fragment 2 (C) causes enlargement of sepals and the loss of large white sepals. D and E, Silencing NbMADS4-2 causes identical phenotypes to silencing with either NbMADS4-1 fragment. F, The reduction of NbMADS4-1 transcript levels was confirmed by semiquantitative RT-PCR in tissue that had been silenced with fragment 1 (row 2) or fragment 2 (row 3). The levels of NbMADS4-2 mRNA were not altered in NbMADS4-1 fragment 1 silenced plants (row 4) and NbMADS4-1 fragment 2 silenced plants (row 5). The transcript levels of the related genes NbMADS5 and NbMADS11 were also examined to confirm specificity of silencing (rows 6 and 7, respectively). The phenotypes shown are not due to the silencing of another MADS-box transcription factor NbAG1 (row 8). Lanes to the left of the marker (M) contain control tissue and those on the right contain tissue from silenced tissue. C is no RT control. Numbers at the top of each lane indicate number of PCR cycles.

Figure 6.

Analysis of NbMADS4 tissue-specific expression. NbMADS4-1 and NbMADS4-2 transcripts were detected by semiquantitative RT-PCR. Lane 1 is tissue from sepals, 2 is petals, 3 is petioles, 4 is carpels, 5 is roots, 6 is stems, 7 is leaves, and 8 is seedlings. Lane M is the size marker. EF1α was used as a loading control.