Heat shock protein 90.1 plays a role in Agrobacterium-mediated plant transformation.

Dear Editor, Many bacterial proteins are involved in Agrobacteriummediated plant transformation. By contrast, relatively little is known about plant proteins that play key roles in transformation. Some of these host proteins interact with Virulence effector proteins, including VirE2, that are transferred from Agrobacterium to plants (Gelvin, 2010; Pitzschke and Hirt, 2010). A recent study indicated that the plant protein SUPPRESSOR OF G2 ALLELE OF SKP1 (SGT1), a co-chaperone of heat shock protein 90 (HSP90), is required for Agrobacterium-mediated transformation (Anand et al., 2012). These studies suggested the involvement of HSP90 in Agrobacterium-mediated transformation. We investigated whether HSP90.1, a co-chaperone of SGT1, may also be important for Agrobacterium-mediated transformation. We assayed an Arabidopsis hsp90.1 T-DNA insertion mutant and 35S:Myc-HSP90.1-overexpression plants (Supplemental Figure 1A and 1B) for stable root transformation. Compared to controls, the hsp90.1–2 mutant was 1.7-fold less susceptible, whereas 35S:MycHSP90.1 plants were twice as susceptible to transformation (Figure 1A). Thus, decreased or increased HSP90.1 expression resulted in altered transformation susceptibility. Genomic DNA blots showed that the amount of uidA DNA, a transgene on the T-DNA, integrated into the hsp90.1–2 mutant genome was four-fold less than that of control plants. However, uidA integration into HSP90.1overexpression plants was 10.8-fold greater than that of control plants (Supplemental Figure 1C). The greater increase in T-DNA integration than that of stable root transformation may result from silencing of some integrated genes because of epigenetic effects such as DNA methylation (Park et al., manuscript in preparation). VIP1 (VirE2 interacting protein 1) and VBF (VIP1 F-box binding protein) are host proteins that may be important for T-DNA subcellular trafficking and integration (Tzfira et al., 2001; Djamei et al., 2007; Gelvin, 2010; Zaltsman et al., 2010). We reasoned that VIP1 and/or VBF could interact with HSP90.1. To test this hypothesis, we conducted a Bimolecular Fluorescence Complementation (BiFC) assay in which two A. tumefaciens strains harboring in their T-DNAs genes encoding protein fusions with nVenus or cCFP were co-infiltrated into Nicotiana benthamiana leaves. The results showed an interaction between VIP1 and HSP90.1 (nVenus–VIP1 and cCFP-HSP90.1) as a yellow fluorescence signal in both the cytoplasm and the nucleus (Figure 1B), compared to empty vector combinations (nVenus–VIP1 + cCFP, and nVenus + cCFP-HSP90.1) showing no yellow fluorescence signals (Supplemental Figure 2A). Our BiFC assay did not detect interaction of HSP90.1 with VBF (Supplemental Figure 2A, lower panel), the latter of which recognizes and targets VIP1 and its bound VirE2 for degradation (Zaltsman et al., 2010). These data indicate that HSP90.1 interacts in leaves with VIP1 but not with VBF. We next conducted co-immunoprecipitation (co-IP) studies. Three Agrobacterium strains individually harboring in their T-DNA regions 35S:Myc-HSP90.1, 35S:YFP–VIP1, or 35S:YFP–SGT1b were co-infiltrated into N. benthamiana leaves. Input proteins were observed using anti-GFP and anti-Myc antibodies, whereas interacting proteins were detected by anti-GFP antibodies (Figure 1C, left panel). The in vivo interaction between Myc-HSP90.1 (83 kDa) and YFP–VIP1 (70 kDa) is strong. YFP–VIP1 alone served as a negative control and did not react with anti-Myc conjugated beads. Our co-IP analysis also detected interaction between VIP1 and SGT1b in N. benthamiana leaves infiltrated with two Agrobacterium strains harboring in their T-DNA regions 35S:Myc–VIP1 or 35S-YFP–SGT1b (Figure 1C, right panel). These results suggest that SGT1b may work in concert with HSP90.1 to protect VIP1. Finally, we conducted yeast two-hybrid (Y2H) analyses to substantiate further the interaction between HSP90.1 and VIP1 (Supplemental Figure 2B). We chose VIP1 as the prey in our assays because VIP1 is a transcription factor and, therefore, VIP1 fused to the gal4 DNA binding domain (BD) was able to auto-activate expression of the reporter genes without interacting with a protein containing an acidic activation domain (AD). Yeast strain AH109 transformed with the BD-HSP90.1 and AD–VIP1 plasmids grew on triple dropout plates (SD/-Leu/-Trp/-His), indicating interaction of the two tested proteins (Supplemental Figure 2B, left four columns). We also detected the expression of the MEL1 reporter gene using an X-α-gal (5-bromo-4-chloro-3-indolyl α-D-galactopyranoside) assay, confirming the interaction between HSP90.1 and VIP1 (Supplemental Figure 2B, last column). Negative control strains containing BD-HSP90.1 and AD-empty vector plasmids, or BD-empty and AD–VIP1