Author for correspondence: Kirankumar S. Mysore Tel: +1 580 224 6740 Email: email@example.com
•Successful genetic transformation of plants by Agrobacterium tumefaciens requires the import of bacterial T-DNA and virulence proteins into the plant cell that eventually form a complex (T-complex). The essential components of the T-complex include the single stranded T-DNA, bacterial virulence proteins (VirD2, VirE2, VirE3 and VirF) and associated host proteins that facilitate the transfer and integration of T-DNA. The removal of the proteins from the T-complex is likely achieved by targeted proteolysis mediated by VirF and the plant ubiquitin proteasome complex.
•We evaluated the involvement of the host SKP1/culin/F-box (SCF)-E3 ligase complex and its role in plant transformation. Gene silencing, mutant screening and gene expression studies suggested that the Arabidopsis homologs of yeast SKP1 (suppressor of kinetochore protein 1) protein, ASK1 and ASK2, are required for Agrobacterium-mediated plant transformation.
•We identified the role for SGT1b (suppressor of the G2 allele of SKP1), an accessory protein that associates with SCF-complex, in plant transformation. We also report the differential expression of many genes that encode F-box motif containing SKP1-interacting proteins (SKIP) upon Agrobacterium infection.
•We speculate that these SKIP genes could encode the plant specific F-box proteins that target the T-complex associated proteins for polyubiquitination and subsequent degradation by the 26S proteasome.
Agrobacterium tumefaciens is a soil-borne phytopathogenic bacterium that elicits the crown gall disease in many plant species. Unlike other plant pathogenic bacteria that commonly affect the host defenses and physiology by secreting effectors, toxins and growth regulators, A. tumefaciens in addition to altering host defenses, also directly modifies the genetic material of its hosts. This genetic modification is achieved through the transfer and integration of a specific subgenomic component, often referred to as the transferred DNA (T-DNA), from the bacterial tumor-inducing (Ti) plasmid into the plant genome. Plant genetic transformation by A. tumefaciens requires the interplay of two essential components located on the Ti plasmid: the activation of the virulence (vir) genes and the translocation of the T-strand and a number of Vir proteins (VirD2, VirD5, VirE2, VirE3 and VirF) into the plant cell. Additionally, a set of A. tumefaciens chromosomal virulence (chv) genes participates in the early stages of bacterial chemotaxis and attachment to plant cells (Gelvin, 2000, 2003; Zhu et al., 2000; Tzfira & Citovsky, 2002). This unique ability of Agrobacterium to genetically transform plants has made it the workhorse of modern plant molecular biology, allowing us to generate transgenic plants.
The Ubiquitin–proteasome system is a key proteolysis pathway and the primary mechanism in eukaryotic cells for degrading unwanted and misfolded proteins (Ciechanover et al., 2000). Through the cascade of E1 ubiquitin activating, E2 ubiquitin conjugating and E3 ubiquitin ligase enzymes, ubiquitin monomers are attached sequentially to the target proteins. The polyubiquitinated proteins are then recognized by the 26S proteasome that removes the ubiquitin chain and degrades the proteins into short peptides. The SCF class of ubiquitin ligase is one of the four different E3 ligases (Vierstra, 2003). The SCF-E3 ligase is typically composed of four different polypeptides: SKP1 that serves as an adaptor protein; Cullin (or CDC53); RING-finger protein RBX1 (or ROC1 and HRT1) that interacts with the Cullin and E2 conjugating enzymes; and an F-box protein that is responsible for recruiting substrates (Mazzucotelli et al., 2006). The Arabidopsis genome contains at least 11 Cullins (five functional genes; Risseeuw et al., 2003), two RBX1 genes, 21 Skp1-like (ASK) genes (Farrás et al., 2001; Marrocco et al., 2003), plus > 700 genes that encode F-box proteins (Gagne et al., 2002; Risseeuw et al., 2003). F-box proteins, named for the conserved 40-amino acid motif (F-box motif) are responsible for binding to ASK/SKP1 (Gagne et al., 2002). Ubiquitination controls several critical cellular processes in plants and as well as responses involving plant–microbe interaction and plant defense signaling (reviewed in Sullivan et al., 2003; Moon et al., 2004; Angot et al., 2007; Kurepa & Smalle, 2008).
Previous studies have demonstrated that VirF from Agrobacterium interacts with Arabidopsis ASK protein, a component of the SCFVIRF complex (Schrammeijer et al., 2001) and plays a role in targeting the cognate proteins on the T-complex to the host proteolysis machinery for degradation (Tzfira et al., 2004). Agrobacterium was shown to induce the expression of a plant-specific VIP1-interacting F-box protein (VBF) that functionally replaces the function of VirF, while regulating the levels of VirE2 and VIP1 via a VBF-containing SCF complex (Zaltsman et al., 2010a,b). These studies prompted us to investigate the role of other components of the SCF complex and its accessory proteins in Agrobacterium-mediated plant transformation. In this study, we investigated the differential expression of genes that encode components of SCF complex in Nicotiana benthamiana and Arabidopsis following infection with transfer competent Agrobacterium strain. We showed that few genes encoding components of SCF (ASK1, ASK2) and its accessory proteins, including SGT1a and SGT1b, are specifically induced following Agrobacterium infection. Using virus-induced gene silencing (VIGS) and Arabidopsis mutants, we demonstrate that SGT1 and SKP1/ASK are required for stable transformation in both N. benthamiana and Arabidopsis. We also observed the differential expression of a set of genes encoding SKP1-interacting proteins (SIPK; Risseeuw et al., 2003) following Agrobacterium infection and we speculate that these SKIP genes encode additional F-box proteins involved in the targeted proteolysis of the T-complex.
Materials and Methods
Plant material and bacterial culture preparation
Seeds of Nicotiana benthamiana Domin were germinated and maintained under glasshouse conditions as described (Anand et al., 2007b). One- to three- week old plants were used for silencing experiments. The Arabidopsis thaliana (L.) Heynh. mutants and overexpressors described in the Results section were kindly provided by Profs Jane Parker (Max-Planck-Institute, Germany) and Hong Ma (Pennsylvania State University). The culture media, propagation and induction of all the Agrobacterium strains before infection assays were performed as described previously (Anand et al., 2007a,b). A bacterial concentration of 1 × 109 CFU ml−1 was used for both tumorigenesis and transient transformation assays in N. benthamiana, while 1 × 107 or 1 × 109 CFU ml−1 was used in Arabidopsis for root tumorigenesis and transient transformation assays.
Plasmid construction and VIGS
pTRV2 VIGS vectors containing NbSGT1, NbSKP1 and NbRAR1 (Liu et al., 2002b) were obtained from Dr Dinesh-Kumar, Yale University. The ORFs corresponding to NbCUL1c and NbRBX1 were amplified by RT-PCR (Supporting Information Table S4) from N. benthamiana cDNA with an adapter (attB1 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCT-3′ and attB2 5′-GGGGACCACTTTGTACAAGAAAGCTGGGT-3′). The PCR products were cloned into the VIGS vector pTRV2 (Liu et al., 2002a,b) by GATEWAY cloning according to the manufacturer’s instructions (Invitrogen Life Technologies, Carsland, CA, USA). The constructs were confirmed by DNA sequencing and introduced into A. tumefaciens GV2260 by electroporation. Agroinoculations for VIGS were performed as described (Anand et al., 2007b) with minor modifications. Two to three plants were inoculated with each clone during initial screening, and in the subsequent screenings, 12 separate plants were inoculated with each clone. We detected the transcript of the virus and the recombinant genes by reverse transcriptase PCR (RT-PCR) using the coat-protein specific primers 5′-CTGGGTTACTAGCGGCACTGAATA-3′ (forward primer) and 5′-TCCACCAAACTTAATCCCGAATAC-3′ (reverse primer) and the GATEWAY adaptor primers attB1 and attB2 2–3 wk postinoculation.
In planta tumor and leaf disk transformation assays
The protocols for in planta tumor and leaf disks transformation assay have been described in detail (Anand et al., 2007a,b). Briefly, for in planta tumor assays, we inoculated, 3 wk postsilencing, the shoots of the gene silenced and the control (TRV::GFP ) plants by puncturing the stems using a needle with a suspension culture of an oncogenic strain Agrobacterium tumefaciens A348. Tumors incited on the infected stems were observed 4 wk after infection. In the leaf disk transformation assays, axenic leaf disks (15–20 per treatment) were incubated with different Agrobacterium strains (detailed in the Results section) for 15 min, blotted on sterile filter paper and co-cultivated with the bacteria at 25°C for 2 d in the dark; transferred onto either Murashige-Skoog (MS; Gibco-BRL, Gaithersberg, MD, USA) medium for tumorigenesis assay or to callus-inducing medium (CIM; 4.32 g l−1 MS minimal salts, 1 ml l−1 vitamin stock, 100 mg l−1 myo-inositol, 20 g l−1 glucose, 0.5 mg l−1 2,4-dichlorophenoxyacetic acid, 0.3 mg ml−1 kinetin, 5 mg l−1 indole acetic acid, and 0.8% phytagar with antibiotics) containing cefotaxamine (200 μg ml−1) and tricarcillin (100 μg ml−1) for stable and transient transformation assays. Stable transformation assay was achieved by including 5 μg ml−1 glufosinate ammonia (GF) in CIM. The cultures were incubated at 25.0 ± 2.0°C; 16 h photoperiod; 70% humidity at 150 μmol s−1 m−2 light intensity. Transient transformation assays (histochemical GUS staining and quantification of GUS activity) were performed as described (Anand et al., 2007b). Data from leaf disk transformation experiments represent the mean of three experiments with a minimum of 150 leaf disks each per treatment with their standard error values shown as error bars.
Agrobacterium-independent transformation assays
The efficacy of Agrobacterium-independent transformation methods in the NbSGT1 and NbSKP1 silenced leaves was tested by particle bombardment with either a 35S::GFP (pCAMBIA) or 35S::uidA (pAHC20) as described (Anand et al., 2007a). The leaves were collected 48–72 h postbombardment, stained with X-gluc staining solution (50 mM NaH2PO4, 10 mM Na2.EDTA, 300 mM mannitol, and 2 mM X-Gluc, pH 7.0) or directly seen under BioRad confocal microscope, for visualizing the GUS and GFP expressing spots. Furthermore, the leaves of the Arabidopsis mutants were also biolistically transformed with 35S::GFP plasmid and visualized for the expression of the GFP-protein.
Characterization of SCF mutants
Seeds of Arabidopsis wild-type Col-0, WS, Ler, and the mutant sgt1a (WS-sgt1a-1; Austin et al., 2002), sgt1b (Col0-edm1, a deletion mutant, Tor et al., 2002; Col0-eta3, EMS mutant, Gray et al., 2003), ask1 (heterozygous Ler-ask1-1/ask1-1; Liu et al., 2004a), and ask2 (WS-ask2-1; Liu et al., 2004a) were germinated and the roots subjected to transient and stable Agrobacterium-mediated transformation assays as described (Nam et al., 1999; Zhu et al., 2003a; Anand et al., 2007a). Tumorigenesis assays were performed on the axenic root segments by infecting with oncogenic strain A. tumefaciens A208 containing a nopaline type Ti plasmid, co-cultivated for 48 h in the dark at room temperature, transferred to a hormone-free MS media supplemented with cefotaxamine and tricarcillin, and the tumor numbers and phenotypes were recorded 4–5 wk postinfection. Transient and stable GUS expression assays and the GF-resistant calli assay were performed as detailed earlier (Anand et al., 2007a) using the disarmed strain A. tumefaciens GV3101 containing either pBISN1 or pCAS1. For all the above-mentioned assays, roots from several (c. 25) identical plants were combined in each experiment and the experiment was repeated at least twice.
RNA extraction, differential gene expression analyses and qRT-PCR
RNA extraction, first strand cDNA synthesis, semi-quantitative RT-PCR and qRT-PCR were performed using standard protocols as described (Nam et al., 1999; Zhu et al., 2003a; Anand et al., 2007a,b). For differential gene expression analyses of the genes encoding proteins associated with the SCF complex, N. benthamiana plants were Agro-infiltrated with an avirulent Agrobacterium strain (A136; non-oncogenic, lacking a Ti plasmid), a T-DNA transfer competent Agrobacterium strain (At804; non-oncogenic, containing pBISN1 with a Ti plasmid facilitating the transfer of T-DNA and Vir proteins) at a concentration of 1 × 107 CFU ml−1 or the infiltration buffer. Samples were collected at different time points as detailed in the results section and qRT-PCRs were performed with gene specific primers and elongation factor as loading control (see Table S4 for primer details). The experiments were repeated two times.
The Affymetrix microarrays (Arabidopsis ATH1 genome array) were used for expression profiling studies in Col-0 ecotype following infection with disarmed A. tumefaciens GV3101 (O.D-0.2), and samples were collected at 0, 48 and 72 h postinfection (hpi), as described (Anand et al., 2007a). A few differentially expressed genes that encode proteins associated with SCF complex identified from the transcriptome analyses were selected for validation by qRT-PCR (see Table S4 for primer details). RNA collected from three independent biological replicates treated with either the disarmed strain of Agrobacterium or mock-infected (buffer), along with three technical replicates for each time point (0, 48 and 72 hpi) was used for quantification.
Agrobacterium attachment and germline transformation assays
Arabidopsis seeds were germinated aseptically and seedlings were grown on solidified MS medium. Root segments were cut from plants c. 10–12 d after germination and washed briefly with distilled water. About 15 ± 25 segments were then suspended in 20 μl of A. tumefaciens GV2260 or chvB− attachment deficient strain (Douglas et al. 1985) harboring the binary plasmid pDSK-GFPuv (Wang et al., 2007) and transferred to solidified MS basal medium. After 24 h, the root segments were picked, washed gently 3–4 times with phosphate-saline buffer and briefly vortexed at very low speed (20–30 rpm) for 30 min. Fluorescent bacteria attached to the cut surfaces and inside the root tissues were visualized using a Leica TCS SP2 AOBS confocal system (Leica Microsystems, Wetzlar, Germany) as described in Anand et al. (2007b). Image acquisitions were carried out using excitation at 488 nm and collecting emitted light from 500 to 600 nm with a 63× water immersion objective (NA 1.2). The experiments were repeated three times.
For the germ line transformation, bulked (25–30 plants per pot) Arabidopsis seeds were germinated and maintained in pots for 4–6-weeks until bolting. The primary and secondary inflorescences were then dip inoculated as described in Clough & Bent (1998) with a disarmed strain A. tumefaciens GV3101 carrying the binary plasmid pBISN1. The inoculated plants were drained of excess bacteria, placed in tray and covered with plastic hoods to maintain high humidity and kept in dark for 2 d. Two days later the pots were uncovered and moved into growth chambers for seed production. Seeds were collected after the seed pods had dried. Seeds were germinated on MS medium containing 100 μg ml−1 timentin, 250 μg ml−1 cefotaxamine and 50 μg ml−1 kanamycin. Transgenic plants were scored after 1 wk for both antibiotic resistance and GUS gene expression.
Data were subjected to analysis of variance using JMP software v.9 (SAS Institute Inc., Cary, NC, USA). When a significant F-test was obtained at P =0.05, separation of treatment means was accomplished by Fisher’s protected least significant difference (LSD).
A reverse genetics approach based on VIGS identified genes that encode proteins of SCF complex to be involved in Agrobacterium-mediated plant transformation
We previously showed VIGS to be a versatile reverse genetics tool for characterizing plant genes involved or predicted to be involved in Agrobacterium-mediated plant transformation (Anand et al., 2007a,b; Anand et al., 2008). Using this approach we silenced a set of genes which encode proteins that associate or interact with the SCF-ubiquitin complex. NbSGT1, NbSKP1, NbRAR1 (Liu et al., 2002a), NbCUL1c (homolog of Cullin 1c; EH371388), and NbRBX1 (putative ring box1; CK284376) were silenced in N. benthamiana using the TRV-VIGS vector (Liu et al., 2002a). As a control, plants were infected with TRV::GFP (has no effect on Agrobacterium infection).
In order to investigate the role of SCF-E3 ligase complex during Agrobacterium infection, we performed in planta tumor assays individually on the stems of SGT1, SKP1, RAR1, CUL1c, RBX1 silenced and control (TRV::GFP) plants of N. benthamiana as described (Anand et al., 2007b) using oncogenic strain A. tumefaciens A348 (pCC113; pTiA6NC, Garfinkel et al., 1981). We observed smaller tumors on the shoots of SGT1, SKP1 and RAR1 gene silenced plants when compared to the tumors on control plants (Fig. 1a). RBX1 and CUL1c silenced plants produced approximately the same sized tumors as the control plants. The down regulation of all the genes used for gene silencing studies in N. benthamiana was confirmed either by real-time quantitative RT-PCR (qRT-PCR) or by semiquantitative RT-PCR (Fig. S1).
Leaf disk transformation assays suggest that SGT1 and SKP1 are involved in early steps of Agrobacterium-mediated plant transformation
The decreased susceptibility of SGT1, SKP1 and RAR1 gene silenced plants to Agrobacterium infection was further confirmed by performing quantitative leaf disk and transient transformation assays as described (Anand et al., 2007a,b). Briefly, in the leaf disk transformation assays, we tested the ability of the gene silenced plants to develop tumors on leaf disks following inoculation with oncogenic strain A. tumefaciens A348 (bacterial dilution 108 CFU ml−1). We previously showed that this particular bacterial strain and dilution produced large green tumors on the leaf disks of N. benthamiana plants within 4 wk of infection (Anand et al., 2007a,b). The tumor phenotypes were scored both visually and by measuring the biomass of leaf disks, 4 wk postinfection (Fig. 1b–d). The tumor-inducing capability was severely attenuated in SGT1 and SKP1 silenced plants and infected leaf disks were approximately four- to six-fold lower in biomass when compared to control (TRV::GFP) and wild-type plants. The RAR1 gene silenced leaf disks also showed reduced tumor formation when compared to the control plants. However, the differences in the tumor phenotypes between the control and RAR1 gene silenced plants were relatively minor and therefore we concluded that RAR1 might not have a major role in Agrobacterium-mediated plant transformation. Because gene silencing of RBX1 and CUL1c did not significantly affect tumor formation (Fig. 1c,d) and we failed to achieve higher silencing efficiency for these genes (Fig. S1), we excluded these genes from future analyses. On the basis of the above observation we concluded that SGT1, and SKP1 are involved in Agrobacterium-mediated plant transformation in N. benthamiana and further characterized them.
In order to identify the step at which SGT1 and SKP1 are involved in Agrobacterium-mediated transformation, we inoculated leaf disks derived from SGT1 and SKP1 silenced and control plants with a disarmed strain A. tumefaciens GV2260 containing the binary vector pBISN1 (carries on its T-DNA a uidA-intron gene encoding β-glucoronidase (GUS; Nam et al., 1999). The infected leaf disks were collected at 2, 4, 7 and 10 d postinfection (dpi), stained with 5-bromo-4-chloro-3-indolyl β-D-glucuronide (X-Gluc) and GUS activity was measured as described (Anand et al., 2007b). Leaf disks derived from SGT1 and SKP1 silenced plants showed significantly less X-Gluc staining and reduced GUS activity when compared to leaf disks derived from control inoculated plants at 2 dpi (Fig. 2a,b). The detection of GUS expression at early time points most likely represents transient expression of the uidA gene, and our data suggest that the SGT1 and SKP1 silenced plants are significantly blocked at the early stages of transformation. Furthermore, the X-Gluc staining and GUS activity continued to be significantly lower in the SGT1 and SKP1 silenced plants at later time points after infection (4, 7 and 10 dpi, representing a combination of transient and stable transformation; Fig. 2a,b). Thus, we concluded that silencing of SGT1 and SKP1 in N. benthamiana probably blocked the early stages (T-DNA transfer and import) of Agrobacterium-mediated plant transformation.
To rule out the possibility that the reduction in the number of tumors produced in SGT1 and SKP1 silenced plants could have resulted from the down-regulation of gene(s) involved in phytohormone responses, we inoculated leaf disks from the gene silenced plants and control plants with a disarmed strain A. tumefaciens GV2260 containing the binary vector pCAS1 (Nam et al., 1999) that contains a bar gene as a selectable marker. Fig. 2(c,d) shows the results of the stable transformation assays on representative leaf disks. Silencing of SGT1 and SKP1 produced lower percentage of disks with glufosinate ammonium (GF)- resistant calli (8.9% and 15.2%, respectively), when compared to the wild-type or control (85.6% and 91.1%, respectively; Fig. 2c,d).
Uninfected leaf disks of SGT1 and SKP1 silenced plants were able to form calli, at an equal efficiency as that of control plants, on nonselective callus-inducing media (CIM; Fig. S2). These results suggest that silencing of SGT1 and SKP1 does not significantly interfere with essential plant cellular functions pertaining to cell division when cultured on a media supplemented with phytohormones. Taken together these observations suggest that silencing of SGT1 and SKP1 in N. benthamiana attenuates Agrobacterium-mediated transient transformation resulting in deficiency in stable transformation.
Many genes that encode proteins associated with SCF complex showed differential expression after Agrobacterium infection in N. benthamiana and Arabidopsis
The observation that silencing of some genes in the SCF-complex resulted in altered crown gall phenotype lead us to investigate further the differential expression of the genes that encode proteins associated with SCF-complex following Agrobacterium infection. N. benthamiana leaves were separately infected with two different Agrobacterium strains, transfer competent strain At804 (GV2260 plus pBISN1; can transfer both T-DNA and Vir proteins to plant cell), transfer deficient strain A136 (non-oncogenic strain lacking a Ti plasmid; this strain cannot transfer T-DNA or Vir proteins to plants), and the buffer control as described (Anand et al., 2007b). The presence of the binary vector pBISN1 (Narasimhulu et al., 1996) allows monitoring the transformation efficiency by examining GUS activity in the infected N. benthamiana leaves. We sampled uninfected (buffer control) and infected leaf samples with different Agrobacterium strains mentioned above at 12, 24, 36 and 48 hpi and determined the expression, by qRT-PCR, of NbSGT1 and NbSKP1. When N. benthamiana plants were infected with At804 (transfer competent strain), NbSGT1 and NbSKP1 genes were induced approximately three- to four-fold within 36 hpi, and 5- to 7.5-fold higher after 48 hpi, respectively (Fig. 3). The expression of NbSGT1 and NbSKP1 was not significantly altered following infection with the strain A136 or by mock infection with the buffer. Our results suggest that the increased expression of NbSGT1 and NbSKP1 genes is in response to transfer of T-strand and/or the Vir proteins, and not merely due to contact of plant cells with the Agrobacterium cells.
We further compared the above data with the microarray data generated from uninfected and infected Arabidopsis (Col-0) leaves with Agrobacterium strain At804 (Anand et al., 2007a) to characterize the differential expression of all the genes that encode proteins associated with the SCF complex. The microarray experiments suggested that only ASK1, ASK2 and ASK20 were specifically induced (> 1.5-fold), while gene expression of most of the other ASK genes were unaffected (Fig. 4a). Because F-box is a large gene family represented by over 700 genes; we did not scrutinize the expression of all the genes encoding F-box proteins. However, few differentially expressed SKIP (Risseeuw et al., 2003) genes identified from a microarray experiment were further confirmed by qRT-PCR (detailed results are described in the next section).
Among the Arabidopsis genes that encode accessory proteins that interact with the SCF-complex, both Arabidopsis SGT1 paralogs were induced at 48 hpi, with SGT1a showing higher induction (2.4- to 2.7-fold), when compared to SGT1b (1.2- to 1.9- fold) at two different time points following Agrobacterium infection (Fig. 4b). qRT-PCR analyses were performed to revalidate the differential expression of SGT1b and ASK (ASK1, ASK2 and ASK20) genes (Table S1).
In an effort to identify F-box proteins that are likely to play an important role in Agrobacterium-mediated plant transformation, we mined the microarray data (Anand et al., 2007a) for Arabidopsis genes that encode SKP1/ASK-interacting proteins (SKIP; Risseeuw et al., 2003) and differentially expressed after Agrobacterium infection. Many of the differentially expressed genes that encode SKIP proteins (24 out of 35 genes) identified carry a putative F-box motif. Out of all the SKIP genes that were differentially expressed after T-DNA transfer competent Agrobacterium infection, all 24 of them were induced (> 1.5-fold; Fig. 4c). Four SKIP genes – namely SKIP10, SKIP14, SKIP25 and SKIP31 – showed almost four- to seven-fold higher induction at 48 hpi (Fig. 4c). In addition, qRT-PCR analyses suggested that SKIP15, SKIP16, SKIP19 and SKIP32 are induced almost by 5- to 34-fold higher at early time points (48 hpi) following Agrobacterium infection (Table S1). The distinctive induction of the SKIP-genes in the Agrobacterium-infected leaf samples further suggests a putative role of these genes in Agrobacterium-mediated plant transformation.
Arabidopsis sgt1 and ask mutants are impaired in Agrobacterium-mediated plant transformation
In order to further confirm whether SGT1 and SKP1/ASK proteins play an important role in Agrobacterium-mediated plant transformation in another plant species, Arabidopsis, we acquired mutants for SGT1a (WS-sgt1a-1; Austin et al., 2002), SGT1b (Col0-edm1, a deletion mutant; Tor et al., 2002; Col0-eta3, EMS mutant, Gray et al., 2003), Arabidopsis homologs of yeast SKP1 gene, ASK1 (heterozygous Ler-ask1-1/ask1-1; Liu et al., 2004a) and ASK2 (WS-ask2-1; Liu et al., 2004a) in various backgrounds as indicated. In Arabidopsis, there are two SGT1 paralogs (SGT1a and SGT1b; Austin et al., 2002) while the ASK is a large gene family consisting of at least 21 genes (Farrás et al., 2001; Marrocco et al., 2003). Additionally, Arabidopsis ASK1 and ASK2 share a significant sequence homology, functional redundancy and were shown to interact with VirF (Schrammeijer et al., 2001; Liu et al., 2004a). The above mutants were tested for their susceptibility to Agrobacterium-mediated plant transformation. Upon infection of root segments with the oncogenic strain A. tumefaciens A208, sgt1b mutants (eta3 and edm1) and ask mutants (ask1 and ask2) formed tumors at three- to six-fold lower efficiency than that of wild-type plants (Fig. 5a, Table 1). Interestingly, the root segments of sgt1a mutant infected with A208 formed tumors comparable to wild-type with both high (1 × 109 CFU ml−1) and low (1 × 107 CFU ml−1) bacterial concentrations (Fig. 5b, Table S2). We performed an additional stable transformation assay to test the formation of glufosinate ammonium (GF)-resistant calli by the root segments of sgt1a, sgt1b, ask1 and ask2 mutants after infection with disarmed strain A. tumefaciens At872 that contains a bar gene on the T-DNA. Mutants sgt1b (eta3 and edm1) ask 1, and ask2 produced GF-resistant calli at 4- to 12-fold lower efficiency compared to wild-type plants (Fig. 5c, Table 2). However, the sgt1a mutant was equally efficient in producing stable GF-resistant calli when compared to wild-type (data not shown).
Table 1. Root tumorigenesis assay in Arabidopsis sgt1b (eta3 and edm1) and ask (ask1 and ask2) mutants and their respective wild-types (shown in parenthesis)
% Tumor (total no. of roots)
Letters indicate significant differences between genotypes by ANOVA at P = 0.05.
76.8 ± 9.0 (585)a
79.3 ± 6.2 (670)a
82.3 ± 4.2 (453)a
22.3 ± 8.2 (453)c
36.8 ± 9.3 (453)b
14.2 ± 5.0 (646)d
17.6 ± 4.8 (595)d
Table 2. Stable root transformation assay in Arabidopsis sgt1b (eta3 and edm1) and ask (ask1 and ask2) mutants and their respective wild-types (shown in parenthesis)
% GF-resistant calli (total no. of roots)
Letters indicate significant differences between genotypes by ANOVA at P = 0.05.
81.7 ± 15.6 (471)a
83.1 ± 6.7 (409)a
21.1 ± 6.7 (409)c
32.1 ± 8.7 (409)b
6.8 ± 2.0 (423)d
8.7 ± 2.7 (405)d
We further tested transient transformation efficiency of sgt1a, sgt1b (eta3 and edm1), ask1 and ask2 mutants. Upon infection by strain A. tumefaciens At804 (GV3101 containing the binary vector pBISN1; Narasimhulu et al., 1996), sgt1b, ask1 and ask2 mutants showed a two- to eight-fold reduction in transient transformation frequency compared with their corresponding wild-type plants, as indicated by X-Gluc staining (Fig. 6a) and GUS activity (Fig. 6b,c). The transformed root segments of sgt1a showed no deficiency in the transient and stable transformation frequency as determined by their GUS activity at 2 and 7 dpi (Fig. 6b). These results further suggest that SGT1b but not SGT1a is involved in Agrobacterium-mediated plant transformation in Arabidopsis. In summary, transient and stable root transformation assays showed that sgt1b, ask1 and ask2 mutants produced a strong rat (resistant to Agrobacterium transformation; Nam et al., 1999) phenotype.
Overexpression of SGT1b, SGT1a in its wild-type background does not increase Agrobacterium-mediated plant transformation
Based on the strong rat phenotype observed in sgt1b mutants, one can hypothesize that overexpression of SGT1b in a wild-type background might enhance Agrobacterium-mediated plant transformation. This hypothesis was tested by performing root tumorigenesis and transformation assays, as described earlier, using Arabidopsis plants overexpressing SGT1b (2F10 and 2F12; Noel et al., 2007). However, the SGT1b overexpressors produced tumors and transgenic calli at the same frequency as that of wild-type plants (data not shown).
SGT1 and SKP1/ASK are not required for Agrobacterium-independent transformation methods
The leaf disks derived from the NbSGT1, NbSKP1 silenced and empty vector (TRV::00) inoculated plants were bombarded with 35S::GFP construct (pCAMBIA-derivative plasmid) using a biolistic method as described (Anand et al., 2007a). No significant differences were detected in the number of cells transiently expressing GFP in the NbSGT1 and NbSKP1 silenced plants when compared to TRV::00 inoculated plants (Fig. S3). We performed similar experiments using a 35S::uidA (pHAC20) construct and found no significant differences in the transient expression of the GUS in the gene silenced and TRV::00 inoculated plants (data not shown). Furthermore, the ability to transiently transform Arabidopsis sgt1b, ask1 and ask2 mutants by Agrobacterium-independent transformation techniques was also tested through biolistic transformation with a 35S::GFP construct. No noticeable differences in the transient expression of GFP were detected in mutants when compared to their wild-type plants (data not shown).
SGT1 and SKP1 are not required for Agro-infiltration and germline transformation methods of Agrobacterium-mediated plant transformation
Previous reports have suggested that NbSGT1 silenced plants can be transiently transformed including rescuing the suppression phenotype in the silenced plants by an overexpression construct (Liu et al., 2002b) and co-expression of R proteins in NbSGT1 silenced plants to elucidate the role of NbSGT1 gene in multiple resistance pathways in N. benthamiana (Peart et al., 2002). Therefore, the gene silenced plants that showed recalcitrance to stable Agrobacterium-mediated plant transformation were tested for their ability to be transformed transiently by Agro-infiltration. The intact leaves of the NbSGT1, NbSKP1 silenced and control plants were individually infiltrated (using a needle-less syringe) with a disarmed strain A. tumefaciens GV2260 containing the binary vector pBISN1 (Nam et al., 1999) at high concentrations (1 × 108 CFU ml−1) as reported earlier (Liu et al., 2002b; Peart et al., 2002). The infiltrated leaves of the NbSGT1 and NbSKP1 silenced plants showed GUS expression comparable to the GUS expression seen in leaves from control plants (Fig. 7a). Furthermore, we also Agro-infiltrated the Arabidopsis sgt1b (eta3) and ask mutants with strain A. tumefaciens GV3101 containing pBISN1 and showed that the mutants are amenable to transient transformation by Agro-infiltration at high concentration (Fig. 7b). However, when lower Agrobacterium concentration (106 CFU ml−1) was infiltrated, NbSGT1 and NbSKP1 silenced N. benthamiana plants (data not shown) and sgt1b (eta3 and edm1), ask1 and ask2 Arabidopsis mutants showed recalcitrance to transient transformation when compared to their respective wild-type or controls (Fig. 7c).
We also checked the germline transformation efficiencies in Arabidopsis sgt1 and ask mutants using a flower dip method (Clough & Bent, 1998). All the SCF mutants tested showed transformation efficiencies comparable to their respective wild-types (Table S3). The above results suggested that even though the sgt1b, ask1 and ask2 mutants produced strong rat phenotype, they could be efficiently transformed by a flower dip protocol. These results suggest that the role of SCF complex in Agrobacterium-mediated plant transformation depends on the target tissue (Yi et al., 2002) and method of transformation. This is consistent with the earlier observations where several rat mutants were transformable by germline transformation (Mysore et al., 2000; Zhu et al., 2003b).
sgt1 and ask mutants showed no deficiency to bacterial attachment
Because sgt1b (eta3 and edm1), ask1 and ask2 mutants were deficient in transient transformation which could result from the lack of Agrobacterium binding to the plant cells; we investigated the ability of Agrobacterium to attach to the root segments of these mutants. Root segments of mutants and their respective wild-types were infected with GFPuv expressing Agrobacterium strain (pDSK-GFPuv; Wang et al., 2007) as described previously (Anand et al., 2007b). There was no deficiency in Agrobacterium attachment to cut ends or surface of the root epithelial tissues or the root hairs in sgt1b, ask1 and ask2 mutants (Fig. S4). Additionally, the confocal images distinctively showed the polar attachment of the Agrobacterium along the root surfaces (Fig. S4b) and the ability to mobilize and colonize inside the root cells of mutant plants (Fig. S4c). Therefore, the lack of transient transformation in sgt1b, ask1 and ask2 mutants cannot be attributed to deficiency in Agrobacterium attachment.
Agrobacterium relies on its virulence proteins and the host machinery, for transfer and integration of the T-DNA into the host genome (reviewed in Gelvin, 2003; Lacroix et al., 2006; Citovsky et al., 2007). The role of host genes in disassembling the protein complex from the mature T-DNA, before its integration into the host genome is not well characterized. The identification of an F-box domain in VirF and specific interaction of VirF, in yeast two-hybrid assays, with Arabidopsis homologs of yeast SKP1 proteins, ASK1 and ASK2 provided the first line of evidence suggesting that SKP1/Cul1/F-box (SCF)-E3 ubiquitin ligase is likely involved in the targeted proteolysis of the Agrobacterium virulence proteins and associated host proteins (Schrammeijer et al., 2001; Tzfira et al., 2004). The above findings indirectly implicated the role of the proteins associated with SCF complex in Agrobacterium-mediated plant transformation. Recently, a plant-specific F-box protein (FBP) that interacts with VIP1 (VBF) was identified and shown to regulate the levels of VirE2 and VIP1 proteins via a VBF-SCF complex (Zaltsman et al., 2010b). However, the role of other components of the SCF-complex still remains unexplored. In this report we provide evidence to suggest that N. benthamiana and Arabidopsis homologs of yeast SKP1 protein and its accessory protein SGT1, which are integral to the SCF-E3 ligase, also play an important role in the genetic transformation of plant cells. We hypothesize that the ask1, ask2 and sgt1b mutants are recalcitrant to transformation because the VirF protein or its plant functional analog would be unable to interact with the normal complement of SCF-E3 ubiquitin ligase complex in these mutants, which would in turn negate formation of an effective SCF complex required for T-complex degradation.
SCF genes are induced at early time points after Agrobacterium infection
Gene expression analyses of genes encoding proteins associated with SCF complex in N. benthamiana plants infected with Agrobacterium-transfer competent strain identified SGT1 and SKP1 that are induced upon Agrobacterium infection. A more systematic study relying on the previously published Arabidopsis gene expression data (Anand et al., 2007a) and qRT-PCRs in Arabidopsis leaves infected with a transfer-competent Agrobacterium-strain supported our above observation that specific genes (SGT1, SKP1/ASK and SKIP) encoding proteins associated with SCF-E3 ligase complex are induced upon Agrobacterium infection of plant cells. The above observation suggests that Agrobacterium could modulate the host proteolysis machinery to facilitate the transformation process. There are several studies implicating that the SCF-genes are involved in host defense against pathogen attack and in plant innate immunity (reviewed in Shirsekar et al., 2010). Thus, the SCF-E3 ligase complex presents an attractive component for effector protein-targeting to promote virulence by interfering with host’s proteasomal degradation machinery (Groll et al., 2008; Birch et al., 2009; Spallek et al., 2009; Bos et al., 2010; Price & Abu Kwaik, 2010; Shirsekar et al., 2010). Based on the gene induction during Agrobacterium infection, we speculate that ASK1, ASK2 and ASK20 play a critical role in ubiquitination and proteasomal degradation of proteins important for transformation. In addition, the interaction of ASK1 and ASK2 proteins with VirF (Schrammeijer et al., 2001) strengthens our speculation. ASK20 is a putative new player in the Agrobacterium infection and was not identified earlier as a VirF interacting protein and its role in plant transformation needs further investigation.
Since the discovery of the requirement of VirF for full virulence in certain host plants (Hooykaas et al., 1984; Melchers et al., 1990), the molecular basis for the function of VirF has largely remained elusive. Direct evidence was provided to support the notion that VirF is involved in the intranuclear proteolysis of at least two protein components, VirE2 and VIP1, via the SCFVirF-mediated pathway (Tzfira et al., 2004). F-box proteins are key components of the SCF-type E3 ligase complex, which recruits target proteins for degradation by the 26S proteasome. Genes encoding FBPs (> 700) are one of the largest gene families in the plant genome, and therefore identifying specific FBPs which target the host proteins and bacterial virulence proteins for proteolysis is challenging. A more likely approach towards the identification of host-specific FBPs that play a role in transformation is to characterize genes encoding the F-box domain containing proteins that specifically interact with ASK1 and ASK2 proteins. Accordingly, we observed specific genes encoding SKP1-interacting proteins (SKIP) with F-box motif to be distinctively induced following Agrobacterium infection. These SKIP proteins likely represent the FBPs involved with the targeted proteolysis of the T-complex in plants.
SGT1 and SKP1/ASK are required for Agrobacterium-mediated plant transformation
The involvement of SGT1 and SKP1/ASK in Agrobacterium-mediated plant transformation was shown using a reverse genetics approach involving transient knockdown of the corresponding genes in N. benthamiana through VIGS and through identification of gene knockouts in Arabidopsis. None of the SGT1 and SKP1/ASK gene silenced plants and mutants showed any deficiency to transient transformation either by particle bombardment or by Agro-infiltration (high bacterial concentration) which further suggests that SGT1b, ASK1, and ASK2 are specifically involved during Agrobacterium infection through cut ends. Recent advances in several laboratories have identified SGT1 as a RAR1-interacting protein, and more crucially as a component of R-gene mediated and nonhost resistance (Azevedo et al., 2002; Liu et al., 2002a). Mutant analyses in Arabidopsis and gene silencing experiments in barley and N. benthamiana have revealed that SGT1 is required for responses that are mediated by diverse range of R-gene structural types that confer resistance against a range of pathogens (Austin et al., 2002; Azevedo et al., 2002; Liu et al., 2002b; Tor et al., 2002; Hubert et al., 2003). The conserved function of SGT1 in regulating SCF-activity in plants is supported by complementation of yeast sgt1 mutant by the Arabidopsis SGT1 paralogs (SGT1a and SGT1b), and the observation that SGT1 co-immunoprecipitates with the core SCF subunits in barley and N. benthamiana (Azevedo et al., 2002; Liu et al., 2002b). Consistent with this idea is the finding that the loss of Arabidopsis SGT1b compromised the functions of SCFTIR1 and SCFCOI1 that mediate the ubiquitination-dependent degradation of proteins in response to the phytohormones auxin and jasmonic acid, respectively (Gray et al., 2003; Uppalapati et al., 2011). Earlier, we have shown that Agrobacterium-mediated transformation efficiency increases when plant defense response is compromised (Anand et al., 2008). Based on this, one would expect increased transformation in SGT1 silenced or mutant plants. However, on the contrary, we observed reduced transformation efficiency suggesting a positive role for genes encoding proteins of the SCF complex in Agrobacterium-mediated plant transformation. The association of SGT1b with COP9 signalosomes and chaperones HSP70 and HSP90, which are also implicated in protein degradation by 26S proteasome supports its role as a co-chaperone involved in multiple SCF-regulated pathways (reviewed in Shirasu & Schulze-Lefert, 2003; Schwechheimer & Deng, 2001). Perhaps the co-chaperone activity of SGT1 plays an important role in folding and maturation of proteins associated with the T-DNA, thus targeting them for degradation. Our finding that SGT1 and SKP1/ASK are involved in Agrobacterium-mediated plant transformation adds another dimension to the role of SCF-complex in plant–microbe interactions.
This work was supported by the Samuel Roberts Noble Foundation and grants from the National Science Foundation (grant nos. IOB 0445799 and DBI 0400580).