The multivesicular body-localized GTPase ARFA1b/1c is important for callose deposition and ROR2 syntaxin-dependent preinvasive basal defense in barley

Abstract

Host cell vesicle traffic is essential for the interplay between plants and microbes. ADP-ribosylation factor (ARF) GTPases are required for vesicle budding, and we studied the role of these enzymes to identify important vesicle transport pathways in the plant-powdery mildew interaction. A combination of transient-induced gene silencing and transient expression of inactive forms of ARF GTPases provided evidence that barley (Hordeum vulgare) ARFA1b/1c function is important for preinvasive penetration resistance against powdery mildew, manifested by formation of a cell wall apposition, named a papilla. Mutant studies indicated that the plasma membrane-localized REQUIRED FOR MLO-SPECIFIED RESISTANCE2 (ROR2) syntaxin, also important for penetration resistance, and ARFA1b/1c function in the same vesicle transport pathway. This was substantiated by a requirement of ARFA1b/1c for ROR2 accumulation in the papilla. ARFA1b/1c is localized to multivesicular bodies, providing a functional link between ROR2 and these organelles in penetration resistance. During Blumeria graminis f sp hordei penetration attempts, ARFA1b/1c-positive multivesicular bodies assemble near the penetration site hours prior to the earliest detection of callose in papillae. Moreover, we showed that ARFA1b/1c is required for callose deposition in papillae and that the papilla structure is established independently of ARFA1b/1c. This raises the possibility that callose is loaded into papillae via multivesicular bodies, rather than being synthesized directly into this cell wall apposition.

Figure 1.

Identification of Barley ARF-GTPases. Phylogenetic analysis of barley and Arabidopsis members of the ARF-GTPase protein family. Rooted neighbor-joining tree. Bootstrap values are shown above each branch, and the tree was collapsed when the value was below 50. The Arabidopsis (At) proteins are named as by Vernoud et al. (2003), and the barley (Hv) proteins are named accordingly. The ScYpt51p RAB5 GTPase from yeast is used as an outgroup. The alignment used to generate this tree is available as Supplemental Data Set 1 online.

Figure 2.

Identification and Expression Analysis of a Barley ARF GTPase Involved in Penetration Resistance against Bgh. (A) Relative haustorial index after transformation with the indicated 35S-driven ARF RNAi constructs. (B) Relative haustorial index after transformation with constructs for 35S-driven expression of wild-type or mutant versions of ARFA1b/1c. (A) and (B) The constructs were cotransformed with a GUS reporter construct into barley epidermal tissue, followed by inoculation with Bgh 2 d after bombardment. Haustorial indices (number of cells with haustoria divided by total number of cells) of GUS-stained, transformed epidermal cells was determined 48 HAI by light microscopy. The indices are given relative to empty vector controls. (C) Transcript levels of ARF genes in whole leaves during Bgh infection, detected by quantitative PCR. ARF transcripts were normalized to GAPDH. Time point 0 was set to 1. Data shown are mean values; n = 6 (A) and 3 ([B] and [C]). Error bars are ±sd.

Figure 3.

ARFA1b/1c-GTPase and ROR2 Syntaxin Act on the Same Pathway Mediating Penetration Resistance. Impact on relative haustorial index (see legend for Figure 2) of transformation of barley cultivar Ingrid wild type (wt) and ror2 with ARFA1b-T31N. The construct was cotransformed with a GUS reporter construct into barley epidermal cells, followed by inoculation with Bgh 2 d later. Haustoria frequency of GUS-stained, transformed epidermal cells was determined 48 HAI by light microscopy. The empty vector was used as reference. A polyubiquitin RNAi construct was used as positive control. Data shown are mean values; n = 3. Error bars are ±sd.

Figure 4.

ARFA1b/1c Function Is Required for ROR2 Localization to Bgh-Induced Papillae. (A) to (C) Localization analysis of transiently expressed YFP-ROR2 in noninoculated barley epidermal cells 24 h after transformation with 35S:YFP-ROR2 (A), cotransformation with 35S:YFP-ROR2 and 35S:ARFA1b-Q71L (B), or cotransformation with 35S:YFP-ROR2 and 35S:ARFA1b-T31N (C). (D) to (L) Localization analysis of transiently expressed YFP-ROR2 in barley epidermal cells, 12 to 14 HAI with Bgh. Transformation with 35S:YFP-ROR2 ([D] to [F]), cotransformation with 35S:YFP-ROR2 and 35S:ARFA1b-Q71L ([G] to [I]), and cotransformation with 35S:YFP-ROR2 and 35S:ARFA1b-T31N ([J] to [L]). YFP fluorescence images ([D], [G], and [J]) and bright-field images ([E], [H], and [K]) are merged in (F), (I), and (L). YFP-ROR2 localization magnified three times in inserts. Arrows indicate sites of Bgh attack. Bars = 10 μm.

Figure 5.

Callose Timing Is Dependent on ROR2, Whereas Callose Deposition Is Dependent on ARFA1b/1c. (A) Callose deposition in response to Bgh is delayed by mutation in ROR2. Leaves of barley cultivar Ingrid wild type (wt) and ror2 were inoculated with Bgh and at various time points assayed for callose deposition at appressorial attack sites. (B) to (D) Callose depositions in response to Bgh (18 HAI), magnified three times in inserts, after cotransformation of the GUS reporter construct with the empty vector (B) or with 35S:ARFA1b-T31N (C). Quantification of callose deposition in GUS-positive cells after cotransformation with 35S:ARFA1b-T31N or the empty vector (D). Data shown in (A) and (D) are mean values; n = 3. Error bars are ±sd. Bars = 20 μm.

Figure 6.

ARFA1b/1c Organelles Are Recruited to the Penetration Site Prior to Callose Deposition. (A) Localization analysis of transiently expressed ARFA1b-eGFP in noninoculated barley epidermal cells. (B) to (F) Localization of ARFA1b/1c-eGFP–positive organelles at Bgh appressorial attack sites 6 to 7 HAI (asterisks). Barley epidermal cells were inoculated 24 h after bombardment with ARFA1b-eGFP. GFP fluorescence image (B) and bright-field image (C) are merged in (D). GFP fluorescence images ([E] and [F]) were taken 5 and 10 min later than (B), respectively. (G) Quantification of callose deposition at Bgh appressorial attack sites in untransformed cells of the leaves studied in (B) to (F). Arrow indicates the first observation of ARFA1b/1c-eGFP–positive organelle assembly. Data shown are mean values; n = 3. Error bars are ±sd. Bars = 10 μm.

Figure 7.

Localization of Barley ARFA1b/1c in Onion and Barley Epidermal Cells. (A) to (C) Colocalization analysis in onion epidermis transiently expressing barley ARFA1b/1c-eGFP and treated with 5 μM FM4-64 for 45 min. GFP fluorescence image (A) and FM4-64 fluorescence image (B) are merged in (C). Inserts are magnified parts of (A) to (C). (D) to (F) Colocalization analysis in onion epidermis transiently expressing ARFA1b/1c-eGFP and treated with 5 μM FM4-64 for 45 min and 100 μM BFA for 25 min. GFP fluorescence image (D) and FM4-64 fluorescence image (E) are merged in (F). (G) to (I) Colocalization analysis in onion epidermis transiently coexpressing ARFA1b/1c-eGFP and RFP-AtARA7. GFP fluorescence image (G) and RFP fluorescence image (H) are merged in (I). (J) to (L) Colocalization analysis in barley epidermis transiently coexpressing ARFA1b/1c-eGFP and RFP-AtARA7. GFP fluorescence image (J) and RFP fluorescence image (K) are merged in (L). Bars = 10 μM (3 μM in insets).

Figure 8.

ARFA1b-T31N Affects Only Callose Deposition, Not the Papilla Structure. (A) and (B) Barley epidermal cell cotransformed with the empty vector and a GUS reporter construct shown 48 h after Bgh inoculation. (C) and (D) Barley epidermal cell cotransformed with a 35S:ARFA1b-T31N and a GUS reporter construct shown 48 h after Bgh inoculation. Bright-field images of papilla structure (arrows) shown in (A) and (C). Fluorescence images of callose deposition (arrows) shown in (B) and (D). Magnified papillae shown in insets. Bars = 20 μm.