Plant materials

Three G. barbadense cultivars, 3-79, Pima 90-53 and Hai 7124, were kindly supplied by Professor Luo Xiaoli from the Institute of Cotton Research, Shanxi Agricultural Academy of Science. Seeds were germinated and grown in a greenhouse. Seedlings from cotyledon expansion to 4^th true leaf emergence were used for VIGS assays. Infiltrated plants were grown in a growth chamber at 25°C with a 16/8 h light/dark photoperiod.

Gene cloning and generation of recombinant VIGS vectors

Cotton gene sequences were obtained from the EST database of GenBank. Candidate gene fragments were cloned by PCR and inserted into the pYL156 vector [9] to generate pYL156 derivatives. The gene fragment sequences, fusing gene fragment sequences, and primers for cloning are listed in File S1 and Table S1, respectively.

Plasmids containing the binary TRV vectors, pTRV-RNA1, pYL156 and the pYL156 derivatives, were transformed into Agrobacterium tumefaciens strains GV3101, LBA4404 and EHA105 by electroporation.

Agrobacterium infiltration

Agrobacterium cultures were grown overnight at 28°C in LB medium containing antibiotics (50 µg/ml kanamycin, 25 µg/ml gentamicin), 10 mM MES and 20 µM acetosyringone. The cells were pelleted by centrifugation at 1200 × g at room temperature for 8 min, and resuspended in MMA (10 mM MES, 10 mM MgCl₂, 200 µM acetosyringone) solution to a final OD₆₀₀ of 1.5. Cell suspensions were incubated at room temperature for at least 3 h without shaking. Agrobacterium cultures containing pTRV-RNA1 and pYL156 or its derivatives were mixed at a 1:1 ratio and infiltrated into two fully expanded cotyledons or true leaves using a needleless syringe. To facilitate the infiltration, small holes were punched with a needle on the underside of the cotyledons or leaves. We used separate pots for different TRV treatments to ensure that treated plants were not contaminated by a different TRV vector. After completion of the experiment, the plants and soil were autoclaved before disposal to prevent release of the virus into the environment. VIGS experiments were repeated three times, with eight plants for each construct per repeat.

Optimization of conditions for TRV VIGS in G. barbadense

The agroinfiltrated plants were divided into three groups and grown under different light intensities, 100 µmol m^-2 s^-1, 300 µmol m^-2 s^-1 and 500 µmol m^-2 s^-1, to evaluate the effect of light intensity on photobleaching due to PDS silencing. Three different photoperiods, 16/8, 12/12 and 8/16 h (day/night), were evaluated using the same experimental procedure as above. The leaves of plants at different ages (cotyledon, 1^st leaf, 2^nd leaf, 3^rd leaf, 4^th leaf, and 5^th leaf stages) were agroinfiltrated to test the effect of PDS silencing. To evaluate the effect of Agrobacterium strain and culture concentration on the photobleaching of leaves by PDS silencing, three strains of Agrobacterium, GV3101, LBA4404 and EHA105, were infiltrated into plants at four concentrations (OD₆₀₀ values 0.5, 1.0, 1.5 and 2.0). Sixteen plants formed a group in these experiments, which were repeated three times. Data were analyzed by one-way ANOVA in SAS [21]. Means were compared using Fisher’s least significant difference (LSD) test [21].

RNA extraction and quantitative real-time PCR (qPCR) analysis

Total RNA was extracted from 100 mg fresh leaves using Trizol reagent according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA, USA) and treated extensively with RNase-free DNase I (Promega, Madison, WI, USA). First-strand cDNA was synthesized from 2 µg total RNA with the SuperScript^TM First-Strand Synthesis system (Invitrogen). To further quantify target gene expression in the transgenic plants, qPCR was employed. Real-time PCR assays were performed using the SYBR Green Real-Time PCR Master Mix (Toyobo, Osaka, Japan) and the DNA Engine Opticon 2 Real-Time PCR Detection System (MJ Research, Hercules, CA, USA). The cotton GhUBI gene was used as an internal control to normalize expression. For each construct, three plants with obvious phenotype were sampled. All reactions were performed in triplicate; negative controls were run without reverse transcriptase. Primer sequences are listed in Table S1. Data were processed to determine relative transcript abundances using the Opticon Monitor software (Bio-Rad, Hercules, CA, USA).

To measure changes in TRV virus biomass, relative quantification of the TRV RNA2 included in the pYL156 vector was performed. Total RNA of systemic leaves from plants inoculated with pYL156 and its derivatives was isolated, cDNA was synthesized and qPCR was carried out as described above. The primers for TRV RNA2 are listed in Table S1. The cotton GhUBI gene was used as an internal control to normalize expression.

Chlorophyll and histochemical assay

Total leaf chlorophyll content was measured 14 d after plants were inoculated with silencing vectors according to the method described by Tang et al [22].

Proanthocyanidin (PA) staining was carried out using the DMACA (p-dimethylaminocinnamaldehyde)-HCl protocol described by Li et al [23]. Fresh leaves were immediately soaked in ethanol-glacial acetic acid solution (3:1, v/v) for decoloration. The decolorized leaves were stained for about 20 min with 0.3% (w/v) DMACA in a cold mixture of methanol and 6 M HCl (1:1, v/v), and then rinsed with several changes of 70% (v/v) ethanol. PA containing cells were stained blue.

Optimal conditions for TRV VIGS in G. barbadense

In our initial experiment to develop an Agrobacterium-mediated VIGS system in G. barbadense, we investigated the optimal conditions under which TRV-based VIGS in G. barbadense cv 3-79 was effective by attempting to silence the G. barbadense Phytoene desaturase (GaPDS) gene as a marker. PDS has been used as a marker for the effectiveness of VIGS in several instances [8,9,24]. The silencing of PDS typically results in white leaves caused by photobleaching, which occurs in the absence of the gene product. In addition, the VIGS vector tobacco rattle virus (TRV) is able to spread vigorously throughout the entire plant but produces only mild symptoms [4]. G. barbadense gene sequences with homology to Arabidopsis genes encoding PDS were used for functional analysis. Partial fragments of putative cotton PDS were cloned into the pYL156 vector to give pYL156:PDS. Plants infiltrated with the empty pTRV-RNA vector (pTRV-RNA1 + pYL156) were used as controls in this study. All plants infiltrated with empty vector (henceforth referred to as vector control plants) grew normally and showed normal phenotypes (Figure 1A). No phenotypic differences were observed between uninfected plants and vector control plants (Figure 1A); this is one of the important advantages of TRV-mediated gene silencing.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Optimal factors for Agrobacterium-mediated GaPDS VIGS in G. barbadense.

A, The photobleaching phenotype of cotton leaves triggered by GaPDS VIGS. The pYL156 vector was used as a vector control; WT, wild type. B–E, The percentage of plants showing photobleaching was affected by light intensity, photoperiod, seedling age, and OD value of Agrobacterium cultures. Means ± standard deviation labeled with different letters are significantly different at the 0.05 level.

https://doi.org/10.1371/journal.pone.0073211.g001

Next, we began to optimize factors for TRV VIGS in G. barbadense. We compared the number of plants showing the PDS-silenced phenotype after growth under different light intensities and photoperiods. Among the seedlings grown under high light conditions (300 and 500 µmol m^-2 s^-1), 90% and 97.5% of plants, respectively, displayed photobleaching at 2 weeks post-inoculation, compared with only 8% of those grown under low light (100 µmol m^-2 s^-1) (Figure 1B). For different photoperiods, the number of plants showing symptoms of leaf photobleaching under long-day conditions was significantly higher than those grown under short-day conditions. Under a 16/8 h photoperiod, 100% of plants displayed white leaves, compared to 92.5% and 62.5%, respectively, of those grown under 12/12 and 8/16-h photoperiods (Figure 1C).

After testing seedlings of different ages, we found that silencing of PDS was effective in seedlings inoculated from the expanding cotyledon stage to the one- to two-leaf stage, 90–100% of which exhibited photobleaching symptoms. The number of plants displaying the PDS-silenced phenotype decreased by 55% when seedlings at the five-leaf stage were injected (Figure 1D). Therefore, younger plants were better for TRV-based silencing in G. barbadense. In addition, the extent of leaf photobleaching was greater by inoculating cotyledons rather than true leaves.

We also investigated whether the strain and culture concentration of A. tumefaciens used to introduce the VIGS vectors affected the outcome of silencing in G. barbadense. The three strains of A. tumefaciens tested, GV3101, LBA4404 and EHA105, showed no significant difference in the number of the plants with the VIGS phenotype (data not shown). However, the concentration of the cultures used for agroinfiltration had a significant effect on silencing of PDS. Cultures resuspended at OD₆₀₀ = 1.5 or above exhibited the greatest effect on VIGS in G. barbadense (Figure 1E).

Thus, G. barbadense seedlings inoculated at the cotyledon expansion to two leaf stages and grown under a 300 µmol m^-2 s^-1 or greater light intensity with a photoperiod of at least 12-h light displayed the photobleaching phenotype indicative of PDS silencing in almost 100% of cases. We employed three endogenous cotton genes, GaPDS, GaCLA and GaANR to further evaluate the VIGS system in G. barbadense using these optimized conditions for agroinfiltration.

GaPDS and GaCLA1 genes provide visible markers for endogenous gene silencing

VIGS has proven to be a powerful tool for gene function studies and functional genomics in various higher plants including G. hirsutum. Having optimized the conditions for Agrobacterium-mediated VIGS in G. barbadense, we next evaluated the VIGS system for analysis of gene function by employing different endogenous marker genes.

The Cloroplastos alterados 1 gene (CLA1) has been used as a marker gene to investigate the VIGS system in many plants. Estevez et al. [25] reported that Arabidopsis CLA1 encodes 1-deoxyxylulose 5-phosphate synthase, the first enzyme of the 2-C-methyl-d-erythritol-4-phosphate pathway involved in chloroplast development. We isolated a homolog of CLA1 from G. barbadense, named GbCLA1. Initiation of a photobleaching phenotype in plants infiltrated with Agrobacteria expressing a partial sequence of GbCLA1 was observed on newly emerging leaves at approximately 8–10 days post-infiltration. Twenty days later, 100% of these plants showed a photobleaching phenotype uniformly distributed across all true leaves (Figure 2A ii). The stems of these plants turned pink or purple, and the plants became severely stunted by the late growth stage (Figure 2A ii, iii), probably due to the strong inhibition of chlorophyll biosynthesis. The vector control plants did not display a photobleaching phenotype and grew normally (Figure 2A i).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Agrobacterium-mediated TRV VIGS of two marker genes, GaPDS and GaCLA1, in G. barbadense.

A, Phenotypes of plants inoculated with pYL156:CLA1 or pYL156:PDS vectors. The pYL156 vector was used as a vector control. B, Three cotton cultivars exhibited the photobleaching phenotype triggered by GaPDS or GaCLA1 gene silencing to differing extents. C, Relative transcript levels of PDS and CLA1 in systemic leaves of plants infiltrated with pYL156:PDS or pYL156:CLA1. The CK value was set at 100%. D, Total chlorophyll content in photobleached leaves. Error bars represent standard deviations (n = 3 biological replicates) in (C) and (D).

https://doi.org/10.1371/journal.pone.0073211.g002

The PDS gene described above is another commonly used marker gene for VIGS and also causes loss of chlorophyll and carotenoids [26,27]. The general pattern of cells and tissues with GaPDS silencing was similar to that of GaCLA1, but the extent of silencing was even greater. The leaves of GaPDS-silenced plants were all white, whereas pale yellow leaves were observed in GaCLA1-silenced plants (Figure 2A, B). To determine whether fragments from different regions of the gene affected the extent of silencing, two GaPDS inserts were amplified near the 5ʹ- and 3ʹ- termini (File S1). The results showed that the extent of silencing achieved was similar for both fragments (Figure 2A iv, v).

To verify the reduction in GaCLA1 or GaPDS mRNA in silenced leaves, a different region of each target gene to that used in the vector was amplified from silenced tissues by qPCR. We collected leaf samples from three independent cotton plants with prominent phenotypes (based on the visible phenotype) for each gene and further analyzed the transcript levels in systemic leaves. We found that CLA1 and PDS transcript levels were reduced by more than 99% in silenced plants compared with vector control plants (Figure 2C). These results from the pYL156:CLA1 and pYL156:PDS plants confirmed the putative functions of these two cotton genes.

To date, a transformation system for G. barbadense has not been developed, so the ability to silence genes readily in G. barbadense using VIGS as a tool for gene function studies and functional genomics in cotton is desirable. Thus, we further tested the VIGS efficiency on three commercial cultivars that were grown in China, including 3-79, Pima 90-53 and Hai 7124. The photobleaching phenotype was observed in all three cultivars infiltrated with GhPDS-expressing agrobacteria within two weeks after inoculation (Figure 2B). Twenty days later, the photobleaching phenotype had uniformly spread across the entire leaves. The silencing efficiency in all cultivars tested was equally high and consistently reached 100% in multiple independent biological trials. However, for cultivar Hai 7124, the photobleaching phenotype was incomplete, with some green color remaining, although the silencing efficiency was 100% at 20 d after infiltration. Similarly, infiltration of GhCLA1-expressing agrobacteria triggered silencing of GhCLA1 and the extent of photobleaching was greater in 3-79 and Pima 90-53 than in Hai 7124 (Figure 2B). The chlorophyll contents were assayed to determine the extent of photobleaching in leaves. Leaves of cultivar 3-79 were more heavily decolorized than those of the other two cultivars 14 d after inoculation with pYL156:PDS or pYL156:CLA1 (Figure 2D). Taken together, these data suggest that a TRV-based VIGS system could trigger potent silencing of endogenous genes in G. barbadense plants, which are amenable to VIGS experiments for gene loss-of-function studies.

The PA metabolic genes ANS and ANR provide mild marker genes for endogenous gene silencing by VIGS

CLA1 and PDS provided obvious markers for analysis of gene and genome functions by the VIGS system, but plants in which these two genes were silenced died 1–2 months post inoculation because of chloroplast damage and cessation of photosynthesis. Thus, it is necessary to select marker genes that allow plants to grow to maturity for analyzing gene function with the VIGS system, especially for reproduction related genes. PAs are a major class of flavonoids, one of the largest groups of plant secondary metabolites. Most genetic studies on the PA biosynthesis pathway have been performed in Arabidopsis and Medicago truncatula [28,29]. The enzymes ANS and ANR function at branches between anthocyanin and PA biosynthesis. ANS converts the substrate flavan-3,4-diol (leucoanthocyanidin) to anthocyanidin, which serves as a substrate for ANR to produce a major PA, 2,3-cis-flavan-3-ol (epicatechin), in Arabidopsis and Medicago [28,29]. When the expression of these two proteins was repressed in mutants or interruptible plants, plants with lower PA content were able to grow and mature, and could be observed via phenotype and staining with the aromatic aldehyde agent DMACA.

Partial fragments of putative cotton ANS and ANR genes with homology to Arabidopsis genes were cloned into the pYL156 vector to generate pYL156:ANS and pYL156:ANR. No phenotypic differences were observed between uninfected plants and vector control plants. At 8–10 days post inoculation, leaf veins and leaf margins around new systemic leaves began to appear brownish in all of the pYL156:ANR-treated plants. One month later, leaves and leaf petioles (Figure 3C) and stems (Figure 3D) became brownish. This brownish phenotype could be due to anthocyanidin accumulation resulting from the loss of ANR function. In contrast, pYL156:ANS plants showed no visible phenotypic differences to vector control plants (Figure 3A, B, D). However, pYL156:ANS and control plants did show differences after PA staining with DMACA (Figure 3E, F). After staining, the leaves of vector control plants turned blue, indicating a high level of PA accumulation. Leaves from pYL156:ANS plants appeared pale yellow, indicating low anthocyanidin and PA levels, while leaves from pYL156:ANR plants appeared red, indicating anthocyanidin accumulation. Two months later, the brownish phenotype and staining markers could still be detected in organs, indicating that GaANS and GaANR could be used as mild marker genes for gene function analysis by VIGS. In addition, the cotton PAs were mainly insoluble in the cell; soluble PAs comprised about 5% of total PA content in OD₅₅₀ assays. Both soluble and insoluble PAs decreased in systemic leaves of plants agroinfiltrated with pYL156:ANS or pYL156:ANR (Figure S1, File S2). We also analyzed the transcript levels of these genes in systemic leaves and found that ANS and ANR transcripts were reduced by more than 99% in silenced plants compared with vector control plants (Figure 3H).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. TRV-induced silencing of the anthocyanidin and proanthocyanidin biosynthetic genes ANS and ANR.

a–d, Plants infiltrated with the vector control (CK), pYL156:ANS and pYL156:ANR showed different phenotypes in systemic leaves (a–c) and stems (d). e–g, DMACA stained leaves. h, Relative transcript levels of ANS and ANR in systemic leaves of plants infiltrated with pYL156:ANS and pYL156:ANR. The CK value was set at 100%. Error bars represent standard deviations (n = 3 biological replicates). White arrows indicate pink leaf veins (c) and stem (d).

https://doi.org/10.1371/journal.pone.0073211.g003

GaPDS and GaANR genes were simultaneously silenced in one plant by the VIGS system

In plants, some proteins have the same or similar functions (for example isoenzymes) and some traits, such as resistance to biotic and abiotic stresses, can be determined by several genes. In such cases, the phenotype cannot be uncovered when only a single gene is silenced by VIGS. In addition, large-scale VIGS experiments have been adopted as a fast-forward genetics approach to screen for phenotypes of interest [4,30–32]. While this approach could be used for high-throughput screening of important genes in cotton, it takes considerable space and effort to grow a large number of cotton seedlings for VIGS. Considering these factors, we sought to determine whether one plant could be inoculated with fusing several-gene-silencing vector or mixing multiple gene-silencing vectors. We fused the sequence fragments of GaPDS and GaANR by chemical synthesis and inserted into pYL156 and generated pYL156:PDS-ANR vector. At 10 d after inoculation with pYL156:PDS-ANR vector, the two distinct symptoms, photobleaching and brownish coloration, were both obvious in the one plant. However, when we mixed three different Agrobacterium cultures containing pTRV-RNA1, pYL156:PDS and pYL156:ANR at a 2:1:1 ratio to inject a single plant, the symptoms were not well observed (Figure 4A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Simultaneous silencing of GaPDS and GaANR in a single plant with the VIGS system.

A, The phenotypes of plants inoculated with pYL156 (i), pYL156:PDS + pYL156:ANR (ii), pYL156:PDS-ANR (iii), and pYL156:PDS (iv), and pYL156:ANR (v). B, Relative transcript levels of PDS and ANR in systemic leaves of plants infiltrated with pYL156:PDS, pYL156:ANR, and pYL156:PDS + pYL156:ANR, and pYL156:PDS-ANR. The CK value was set as 100%. C, Relative levels of TRV RNA2 in systemic leaves of plants infiltrated with pYL156:PDS, pYL156:ANR, and pYL156:PDS + pYL156:ANR, and pYL156:PDS-ANR. The CK value at 10 d post-inoculation (dpi) was set at 1. Error bars represent standard deviations (n = 3 biological replicates) in (B) and (C).

https://doi.org/10.1371/journal.pone.0073211.g004

Leaf samples from three independent cotton plants with prominent phenotypes were collected, and the transcript levels of GaPDS and GaANR were analyzed. We found both a 99% reduction in transcript level for these genes in the silenced plants with fusing two-gene-silencing vector compared with vector control plants (Figure 4B). However, the transcript levels of the two genes in one plant simultaneously inoculated with mixed pYL156: PDS and pYL156:ANR were obviously higher than in plants inoculated with either pYL156:PDS or pYL156:ANR. The reductions of transcript level were respectively 72% and 79% for PDS and ANR gene compared to vector control plants (Figure 4B). This result suggested that it is useful and feasible to simultaneously silence two genes in a single plant through inoculation with fusing two-gene-silencing vector, but not with mixing two gene-silencing vectors.

To test whether the differences in GaPDS and GaANR expression among plants agroinfiltrated with fusing two-gene-silencing vector, mixing two gene-silencing vectors, and a single gene-silencing vector were correlated with virus growth in the host tissue, measurements of virus RNA2 in the infected leaves were made. Total RNA was extracted from systemic leaves of plants inoculated with different vectors. Virus RNA2 and the cotton internal control gene GaUBI were then quantified by real-time PCR. At 10 and 20 days after inoculation, the amount of virus RNA2 was similar among all inoculated plants at each time point. However, the levels at 10 d post-inoculation (dpi) were higher than those at 20 dpi. This result indicated that the amount of virus in a cell was stable regardless of whether one or multiple gene silencing vectors were introduced by the TRV-based VIGS system. Taken together, these results indicated that under fusing these genes condition, two or three genes could be simultaneously silenced in one plant for analysis of gene and genome functions by TRV-based VIGS although the total virus amount in a cell was stable. Thus, we suggest that it would be useful and feasible to silence multiple genes in one plant via fusing these genes silencing vector. The genes identified in this VIGS system will provide valuable resources for targeted breeding programs and genetic engineering of cotton cultivars with desirable traits in G. barbadense.