The conserved oligomeric Golgi complex is involved in penetration resistance of barley to the barley powdery mildew fungus

Abstract

Membrane trafficking is vital to plant development and adaptation to the environment. It is suggested that post-Golgi vesicles and multivesicular bodies are essential for plant defence against directly penetrating fungal parasites at the cell wall. However, the actual plant proteins involved in membrane transport for defence are largely unidentified. We applied a candidate gene approach and single cell transient-induced gene silencing for the identification of membrane trafficking proteins of barley involved in the response to the fungal pathogen Blumeria graminis f.sp. hordei. This revealed potential components of vesicle tethering complexes [putative exocyst subunit HvEXO70F-like and subunits of the conserved oligomeric Golgi (COG) complex] and Golgi membrane trafficking (COPIγ coatomer and HvYPT1-like RAB GTPase) as essential for resistance to fungal penetration into the host cell.

Figure 1

Transient over‐expression of HvCOG3 supports basal penetration resistance. Barley epidermal cells of detached barley cv. ‘Golden Promise’ leaves were transiently transformed with an HvCOG3 over‐expression construct, together with green fluorescence protein (GFP), by particle bombardment. As control, the empty pGY‐1 vector was used. Inoculation with spores of Blumeria graminis f.sp. hordei took place 2 days after transformation and microscopic analysis followed 2 days after inoculation. The columns represent the mean values of the relative penetration rate. The respective control was set to 100%. The mean values are based on seven independent over‐expression experiments. The bar represents the standard error; the difference of the means is significant at P < 0.05 according to Student's t‐test.

Figure 2

Analysis of green fluorescence protein (GFP) secretion in cells expressing the HvCOG3 RNA interference (RNAi) construct. GFP was N‐terminally fused with a signal peptide. The resulting construct (sGFP in pGY‐1) was transiently co‐transformed, together with either the empty vector pIPKTA30N as a control or the HvCOG3 RNAi construct and the transformation marker mCherry, into epidermal cells of detached barley leaves (cv. ‘Golden Promise’) and analysed by confocal laser scanning microscopy 2 days after transformation. (A) The intensity of sGFP in the cytoplasm of transformed cells was normalized against the fluorescence intensity of mCherry. In each of three independent repetitions, 50 cells were examined. All cells were imaged with the same excitation and detection settings. To warrant detection within the dynamic range of the system, the brightly fluorescing mCherry was recorded at low detector gain, whereas weakly fluorescing sGFP was recorded at higher detector gain. Average GFP fluorescence in HvCOG3 deficient cells is significantly higher than in control cells (Student's t‐test, P < 0.05). (B) Confocal laser scanning micrographs of barley epidermal cells expressing sGFP together with either the empty vector (top) or the HvCOG3 RNAi construct (bottom) 2 days after transformation. Photographs were taken using the same detection settings. Size bars, 20 μm.

Figure 3

Analysis of Golgi body frequency in cells expressing the HvCOG3 RNA interference (RNAi) construct. Epidermal cells of detached barley leaves (cv. ‘Golden Promise’) were simultaneously transformed with a green fluorescing Golgi marker (sGFPHDEL) and the empty vector pIPKTA30N or the HvCOG3 RNAi construct. Transformed cells were analysed 3 days after particle bombardment using confocal laser scanning microscopy, and the number of brightly fluorescing Golgi bodies per cell was counted. (A) The cells were classified into four categories with 0–10 (white), 11–20 (light grey), 21–50 (dark grey) and >50 (black) Golgi bodies per cell. Statistical analysis using a χ ² test showed that the expression of the two constructs led to a significantly different distribution concerning the number of Golgi bodies per cell, with a probability of 99.5%. Seventy cells were examined per variant in three independent repetitions. All cells of the same experiment were imaged with the same excitation and detection settings. (B) Confocal laser scanning whole‐cell projections of barley epidermal cells expressing sGFPHDEL together with either the empty pIPKTA30N vector (top two photographs) or the HvCOG3 RNAi construct (bottom two photographs) 3 days after bombardment. The transmission channels show the cell borders of scanned cells.

Figure 4

Functional analysis of the barley conserved oligomeric Golgi (COG) complex subunits in barley interaction with Blumeria graminis f.sp. hordei. Epidermal cells of detached barley leaves (cv. ‘Golden Promise’) were transiently transformed with the marker gene green fluorescence protein (GFP) and the transient‐induced gene silencing (TIGS) construct of one of the eight HvCOG subunit genes or empty pIPKTA30N vector via particle bombardment. Leaves were inoculated 2 days after transformation and the microscopic evaluation took place 2 days after inoculation. The susceptibility index of the empty vector control was set to 100%, and the TIGS results of HvCOG complex subunits are given as the mean values of at least five independent experiments. The bars represent standard errors. In addition to HvCOG3 (please note data for HvCOG3 are identical to those in Table 1, given here for better comparability of the results), TIGS of HvCOG1 significantly enhanced the susceptibility according to Student's t‐test (P < 0.05).

Figure 5

Functional characterization of potential HvCOG3 interaction partners in transient‐induced gene silencing (TIGS) and over‐expression experiments. Detached leaf segments of the barley cv. ‘Golden Promise’ were transiently transformed via particle bombardment. Potential interaction partners were either knocked down by TIGS (A) or over‐expressed (B), together with the transformation marker green fluorescence protein (GFP). The empty vectors pIPKTA30N (in A) and pGY‐1 (in B) served as controls. The leaf segments were inoculated 2 days after transformation and the microscopic evaluation was conducted 2 days after infection. The susceptibility index (frequency of penetrated cells relative to all cells transformed, A) and penetration rate (frequency of penetrated cells relative to all cells attacked, B) of the empty vector control were set to 100%. Columns for the potential HvCOG3 interaction partners represent mean values relative to the control of five to six independent experiments in (A) and three to six over‐expression experiments in (B). Knockdown of HvYPT1‐like, HvCOPIγ‐like and over‐expression of HvVTI1‐like significantly enhanced the susceptibility at P < 0.05 according to Student's t‐test.

Figure 6

Subcellular localization of GFP‐HvYPT1‐like. GFP‐HvYPT1‐like was transiently expressed in barley epidermal cells (cv. ‘Golden Promise’), together with either the cytoplasmic and nuclear marker protein (mCherry) (A, D) or the Golgi marker protein GmMAN1‐RFP (B, C). (C) Enlargements of the boxed areas in (B) showing the overlay of GFP‐HvYPT1‐like with the Golgi marker (yellow pixel in the merge image). (D) Polarization of the cytoplasm (mCherry) and GFP‐HvYPT1‐like‐labelled Golgi bodies in cytoplasmic strands (arrow heads) to the site of nonsuccessful attack by Blumeria graminis f.sp. hordei (arrows). The nucleus (n in the mCherry channel) is also attracted to the site of interaction by 24 h post‐inoculation. Photographs are whole‐cell projections of single optical sections at 2‐μm increments recorded by confocal laser scanning microscopy. Size bars, 20 μm.