Arabidopsis thaliana and Nicotiana benthamiana plants were grown in pots with autoclaved vermiculite and watered with Hoagland solution. The growth condition is at 24°C under a 16 h light/8 h dark cycle. A. thaliana plants for HR and disease resistance assays were grown under 8 h light/16 h dark cycle condition. All the lines of A. thaliana are in Columbia-0 background. The pldδ mutant 12B (SALK_023247C) were obtained from ABRC. The mutant line rpm1-3 and the transgenic line pRPM1::RPM1-Myc/rpm1-3 (AT5) were gifts from Dr. Jeff. Dangl (University of North Carolina at Chapel Hill, Chapel Hill, NC, United States).

The gateway system was used to construct vectors. For transient expression, The CDSs of AtPLD genes were cloned into the expression vector pEarleyGate 101 containing the constitutive high-expression CaMV 35S promoter, and YFP-HA tag (Karimi et al., 2007). HA tag was used for protein detection. RPM1-Myc was cloned into the pGWB2 vector to obtain the 35S::RPM1-Myc expression construct (Nakagawa et al., 2007). RPM1(D505V)-Myc was cloned into the pMDC7 vector under the control of the estradiol-inducible promoter to obtain the Est:: RPM1(D505V)-Myc expression construct (Karimi et al., 2007). For the bimolecular fluorescence complementation (BiFC) assay, we modified the pEarleyGate 101 vector into vectors that could express the two complementary parts of YFP, the N terminus (nYFP) and the C terminus (cYFP) (Walter et al., 2004; Citovsky et al., 2006). The expression constructs of 35S:: RPM1 -nYFP-HA and 35S:: PLDδ-cYFP-HA were made respectively.

The expression constructs were electro-transformed into Agrobacterium tumefaciens GV3101. Agrobacteria were cultured overnight at 28°C with suitable antibiotics. Overnight-grown Agrobacteria were centrifuged and re-suspended in the induction buffer (10 mM MES pH 5.6, 10 mM MgCl₂, 150 μM acetosyringone). Samples were infiltrated into the leaves of 5-week-old N. benthamiana with a 1-mL needless syringe at desired OD_{600 nm} values. Agrobacteria containing the P19 plasmid was always co-infiltrated with the samples at an OD_{600 nm} of 0.2. The transiently expressed proteins were extracted and analyzed at 48 h after infiltration (Pitino et al., 2016).

For protein expression, three leaf disks (7 mm diameter) were ground with 110 μL of the protein extraction buffer (20 mM Tris-HCl PH 8.0, 5 mM EDTA, 10 mM DTT, 1% SDS). The supernatants were collected with centrifugation at 8,000 × g for 3 min and mixed with fourfold protein loading buffer. Actin was used as an internal protein loading control. For the fractioning experiments, leaf tissues were ground in a mortar with the extraction buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl, 20% glycerol, 1 mM DTT, 10 mM EDTA, 1 mM PMSF, protease inhibitor cocktail). Samples were centrifuged at 6,000 × g for 10 min at 4°C to remove the debris. The supernatant containing total proteins (T) was separated into the cytosolic fraction (S) and the microsomal fraction (M) with centrifugation at 20,000 × g for 60 min at 4°C. The plasma-membrane localized H⁺-ATPase was used as the microsomal marker. The big subunit of Rubisco was stained with Ponceau S as the cytosolic marker.

For Co-IP experiment, because PLDδ and RPM1 are membrane associated proteins, the microsomal fractions were used for Co-IP. The protein content of RPM1-Myc in the microsomal fractions were determined with Western blots, and the microsomal fractions were re-suspended with the suspension buffer (40 mM Tris-HCl PH 7.5, 150 mM NaCl, 1 mM DTT, 5 mM EDTA, protease inhibitor cocktail, 1% Triton X-100) to equalize RPM1-Myc protein levels between the control and test samples. The samples were rotated at 4°C for 1 h and centrifuged at 20,000 × g for 1 h to remove the insoluble proteins. 30 μL of the anti-GFP conjugated agarose beads (MBL, #D153-8) were added into 1 mL of the supernatants and incubated at 4°C for 3 h. Beads were washed three times with the suspension buffer. Immunoprecipitates were eluted with 50 μL of the elution buffer (2% SDS) at 37°C for 5 min.

The primary antibodies used for Western blots were the anti-Myc (GeneScript, #A00704), the anti-HA (Roche, #11867423001), the anti-H⁺-ATPase (Agrisera, #AS07260-100), the anti-β-Actin (Abbkine, #A01050-2), and the anti-T7 (Novagen, #69522).

The nYFP-tagged proteins and the cYFP-tagged proteins were co-expressed in the leaves of N. benthamiana. The fluorescent images were observed using confocal microscopy according to the BiFC protocol (Walter et al., 2004). YFP was excited at 488 nm and fluorescent emissions were observed at 518–540 nm.

Pseudomonas syringae pv. tomato strain DC3000(avrB) was grown on King’s B medium with antibiotics at 28°C for 36 h. For spray inoculation, leaves from 5-week-old plants were sprayed with Pto DC3000(avrB) in 10 mM MgCl₂ with 0.025% silwet L-77 at an OD₆₀₀ of 0.05. Inoculated plants were covered with a plastic dome for 2 days. All experiments were repeated three times. For ion-leakage assay, five leaf disks (8 mm diameter) of A. thaliana or N. benthamiana were floated in a clean wild-mouth tube with 6 ml of distilled water. The conductivity of the samples was measured at the indicated time points with a conductivity meter (Hubert et al., 2009).

Phospholipase D mediates the signal transduction of RPM1 (Andersson et al., 2006). We cloned and transiently expressed all the members of AtPLD family in the leaves of N. benthamiana to determine their interactions with RPM1. Of the 12 PLD members, PLDδ, PLDα2, PLDβ1, PLDβ2, and PLDγ3 had detectable protein expression, and were used for Co-IP assay. We co-expressed 35S::PLD-YFP-HA and 35S::RPM1-Myc in the leaves of N. benthamiana, and immunoprecipitated PLD-YFP-HA with the anti-GFP antibody to determine the Co-IP of RPM1-Myc. Co-expression of 35S::YFP-HA and 35S::RPM1-Myc was used as the negative control. The Co-IP assay showed that PLDα2 and PLDδ interacted with RPM1, while PLDβ1, PLDβ2, and PLDγ3 did not (Table 1). PLDδ was used for subsequent experiments due to stronger interaction with RPM1 (Figure 1A). We validated the interaction with the BiFC assay. The Yellow fluorescence protein (YFP) was split into two complementary parts, the N terminus (nYFP) and the C terminus (cYFP) (Walter et al., 2004; Citovsky et al., 2006). The expression constructs 35S:: RPM-nYFP-HA and 35S:: PLDδ -cYFP-HA were co-expressed in the leaves of N. benthamiana. The results showed clear yellow fluorescence in the samples co-expressing RPM1-nYFP-HA and PLDδ-cYFP-HA, but no fluorescence was detected in the negative controls (Figure 1B). We transformed 35S::PLDδ-YFP-HA into the pRPM1::RPM1-Myc/rpm1-3 plant (AT5) to determine the interaction of RPM1 and PLDδ in A. thaliana. The Co-IP of RPM1-Myc with PLDδ-YFP-HA in the transgenic plants indicated the interaction of RPM1 and PLDδ in A. thaliana (Figure 1C).

Since PLDδ interacted with RPM1, we tested whether PLDδ affected the function of RPM1. We firstly determined whether PLDδ could affect RPM1-induced HR. RPM1(D505V), which mimics the active state of RPM1, is sufficient to induce the cell death (Gao et al., 2011). The estradiol inducible Est::RPM1(D505V)-Myc construct was co-expressed with 35S::PLDδ-YFP-HA or 35S::YFP-HA in the leaves of N. benthamiana to compare the occurrences of the cell death induced by RPM1(D505V). The cell death could be stained with trypan blue or be quantified with ion-leakage assay. Results showed that PLDδ-YFP-HA could obviously suppress RPM1(D505V) induced cell death (Figures 2A,B). We hypothesized that PLDδ either suppressed the activation of RPM1, or reduced the protein level of RPM1 to affect its function.

We tested whether PLDδ could affect the protein level of RPM1(D505V) in the leaves of N. benthamiana. RPM1(D505V) is unstable due to the cell death induced by the protein, and it is hard to accumulate detectable protein level. Lanthanum chloride (LaCl₃) can inhibit RPM-mediated HR, and prevent the degradation of the activated RPM1 or RPM1(D505V) (Grant et al., 2000; El Kasmi et al., 2017). We co-expressed Est::RPM1(D505V)-Myc and 35S::PLDδ-YFP-HA or 35S::YFP-HA in the leaves of N. benthamiana, and induced the expression of RPM1(D505V) with estradiol at 2 days after infiltration. The leaves were treated with 4 mM of LaCl₃ at the same time of the induction. The protein levels of RPM1(D50V) were detected at 3, 5, and 8 h after induction. Our results showed that PLDδ-YFP-HA obviously reduced the protein level of RPM1(D505V) compared with the negative control YFP-HA (Figure 3A).

RPM1(D505V) is a plasma-membrane associated protein. We collected the samples at 5 h after induction, and separated total proteins into the soluble and the microsomal membrane fractions using ultracentrifugation. Certain amount of RPM1(D505V) participated in the soluble fraction of the sample co-expressed with PLDδ, while less amount of RPM1(D505V) participated in the soluble fraction of the sample co-expressed with the negative control YFP-HA (Figure 3B).

PLDδ is an oleate and phosphatidylinositol 4,5-bisphosphate (PIP2) activated enzyme. PLDδ(R410D) and PLDδ(R622P) are two activity-deficient mutants due to loss oleate binding or PIP2 stimulated activity respectively (Wang and Wang, 2001). We made PLDδ(R410D/R622P) double mutant to determine whether PLD activity was required to affect the function of RPM1(D505V). Contrast to PLDδ, the activity deficient mutant did not affect the autoactivity and the protein level of RPM1(D505V) (Figures 4A,B).

We transformed 35S::PLDδ-YFP-HA into AT5 plant to determine the effects of PLDδ on RPM1 in A. thaliana. Three independent transgenic lines PLD-8, PLD-9, and PLD-21 were used to assay the protein level of RPM1. The results showed that the protein levels of RPM1 were lower in the PLDδ transgenic plants (Figure 5A). According to our assay, PLDγ3 did not interact with RPM1 in N. benthamiana. We transformed 35S::PLDγ3-YFP-HA into AT5 plant to determine the protein levels of RPM1 in three independent PLDγ3-OE plants. The results showed that the protein levels of RPM1 in the transgenic plants were the same as that in AT5 (Figure 5B). The results suggested that the interaction of PLD with RPM1 was necessary to reduce the protein level of RPM1.

Since the protein level of RPM1 reduced in the PLDδ-OE plants, we reasoned that the function of RPM1 should be deficient in the PLDδ-OE plants. We compared the functions of RPM1 in PLD-8 and PLD-9 with those in AT5 plant. The loss-of-function mutant rpm1-3 was used as a negative control, and the pldδ-KO mutant (SALK_023247C) was included for the assays. RPM1 can be activated by its corresponding effector AvrB, and thus restrict the growth of the avirulent pathogen Pto DC3000(avrB). We spray-inoculated the leaves of the plants with Pto DC3000(avrB), and counted bacteria numbers at 0 and 3 days after inoculation. The result indicated that PLD-8 and PLD-9 plants displayed obviously less disease resistance than AT5 (Figures 6A,B). The pldδ-KO mutant displayed the same bacteria growth restriction as AT5, which was consistent with previous report. We quantified the RPM1-mediated HR by measuring the ion-leakage of the plants. The ion-leakage data showed that the PLD-8 and PLD-9 had weaker HR than AT5 (Figure 6C). Based on above results, we concluded that overexpressing of PLDδ reduced the protein level of RPM1, and thus negatively affected the function of RPM1.

Because PLD is enzyme, we want to determine whether PLD activity is required to affect the protein level of RPM1. ABA was able to stimulate the activity of PLD (Jacob et al., 1999), so we treated the leaves of AT5 with or without 100 μM of ABA, and compared the protein levels of RPM1 between the samples. The protein levels of RPM1 assayed at 3 h after treatments, and the results showed that ABA treatment obviously reduced the protein level of RPM1 (Figure 7A). Since PLD activity leads to the synthesis of PA, we treated the plants with n-butanol to block the synthesis of PA and thus the PLD enzymatic function. When the plants were treated with both n-butanol and ABA at the same time, ABA-induced RPM1 reduction was abolished (Figure 7B). The results indicated that PLD activity and its derived PA were required to mediate the ABA-induced RPM1 reduction.