The Arabidopsis NHL3 gene encodes a plasma membrane protein and its overexpression correlates with increased resistance to Pseudomonas syringae pv. tomato DC3000

Abstract

The Arabidopsis genome contains a family of NDR1/HIN1-like (NHL) genes that show homology to the nonrace-specific disease resistance (NDR1) and the tobacco (Nicotiana tabacum) harpin-induced (HIN1) genes. NHL3 is a pathogen-responsive member of this NHL gene family that is potentially involved in defense. In independent transgenic NHL3-overexpressing plant lines, a clear correlation between increased resistance to virulent Pseudomonas syringae pv. tomato DC3000 and enhanced NHL3 transcript levels was seen. These transgenic plants did not show enhanced pathogenesis-related gene expression or reactive oxygen species accumulation. Biochemical and localization experiments were performed to assist elucidation of how NHL3 may confer enhanced disease resistance. Gene constructs expressing amino-terminal c-myc-tagged or carboxyl-terminal hemagglutinin epitope (HA)-tagged NHL3 demonstrated membrane localization in transiently transformed tobacco leaves. Stable Arabidopsis transformants containing the NHL3-HA construct corroborated the findings observed in tobacco. The detected immunoreactive proteins were 10 kD larger than the calculated size and could be partially accounted for by the glycosylation state. However, the expected size was not attained with deglycosylation, suggesting possibly additional posttranslational modification. Detergent treatment, but not chemicals used to strip membrane-associated proteins, could displace the immunoreactive signal from microsomal fractions, showing that NHL3 is tightly membrane associated. Furthermore, immunofluorescence and immunogold labeling, coupled with two-phase partitioning techniques, revealed plasma membrane localization of NHL3-HA. This subcellular localization of NHL3 positions it at an initial contact site to pathogens and may be important in facilitating interception of pathogen-derived signals.

Figure 1.

NHL3 overexpression levels correlate with resistance to virulent bacteria. A, Two micrograms of total RNA prepared from wild-type (WT) or four independent transgenic Arabidopsis lines with NHL3 under the control of the constitutive 35S promoter (S6, S7, S5, and S13) was used for northern blots and was hybridized with the indicated radioactively labeled cDNA probes (rRNA, 26S rDNA). B, Four-week-old plants were inoculated with P. syringae pv. tomato DC3000 at a concentration of 1 × 10³ cfu mL^–1. Bacterial titers in leaves at the indicated time points were determined. The figure shows the means and ses of a representative experiment. Line S13 consistently showed less bacterial growth in at least four other experiments. Bacterial growth in lines S7 and S5 are intermediate between S13 and the nonexpressing line S6 or wild-type plants. The four transgenic lines show differential expression of the transgene, and the relative mRNA level of NHL3 in the transgenic lines is indicated at the bottom of the figure. The level in the untransformed wild-type (Columbia [Col-0]) plants is set at a reference value of one. An inverse correlation between transgene expression and bacterial growth can be observed.

Figure 2.

Predicted structural features of the NHL3 protein. TM1 is the most likely transmembrane domain, whereas TM2 is a second putative transmembrane domain but predicted with a lower likelihood. Regions flanking transmembrane domain containing positively charged amino acids that favor cytoplasmic exposure are indicated by plus symbols. The predicted N-glycosylation sites are marked with asterisks. The amplified boxes show the amino acid sequences of the indicated domains.

Figure 3.

Epitope-tagged NHL3 is localized to the membrane fractions of transiently transformed tobacco leaves (A and B) or stable transformants of Arabidopsis (C). A and B, Tobacco leaves were inoculated with A. tumefaciens strain GV3101 carrying the respective cloning plasmids (empty vector), the NHL3-HA construct (in A), or the c-myc-NHL3 construct (in B). Total (T) protein extracts were separated into soluble (S) and 100,000g microsomal pellet (M) fractions, and were subjected to western-blot analysis with the appropriate antibody. A, Leaves were sprayed with DEX 24 h postinfiltration and were harvested 24 h later. B, Leaf tissues were harvested 48 h postinfiltration. C, Leaves of transgenic Arabidopsis (T₃ generation) harboring the control cloning plasmid (empty vector) or the NHL3-HA construct were harvested 24 h post-DEX treatment and were analyzed as described above for A and B. Three independent transgenic lines of each construct showed identical results and only one representative experiment is shown. Numbers on the left indicate positions of protein size markers in kilodaltons.

Figure 4.

NHL3-HA in transgenic Arabidopsis is a glycosylated and membrane-associated protein complex. A, Microsomal fractions of tobacco leaves transiently expressing NHL3-HA were incubated for 30 min with buffer (M-) or with PNGase F (M+) and were subjected to western-blot analysis with the anti-HA antibody. B, Microsomal fractions isolated from leaves of DEX-treated transgenic NHL3-HA Arabidopsis plants were treated and analyzed as described in A. C, Higher mobility complexes (arrows) were seen when the protein extracts were treated with the indicated amount of chemical crosslinker, 3,3′-dithiobis(sulfosuccinimidylpropionate) (DTSSP) before western blotting. D, Microsomal fractions isolated from leaves of DEX-treated transgenic NHL3-HA Arabidopsis plants were treated with the indicated substances for 1 h at 4°C. Membranes were then pelleted by ultracentrifugation and the resulting soluble (S) and membrane proteins (M) were analyzed by western blotting.

Figure 5.

Immunolabeling localizes NHL3-HA to the periphery of the cell. Cross-sections of leaves from DEX-treated transgenic plants harboring the empty pTA7002 vector control (A) or NHL3-HA construct (B–E) were probed with an anti-HA antibody followed by an Alexa488-coupled (Molecular Probes, Eugene, OR) secondary antibody giving green signals (A–D) or by a secondary antibody conjugated with colloidal gold (E). In leaf sections of plants transformed with the empty vector control, only brownish autofluorescence is visible (A), whereas mesophyll (B) and epidermal cells (C) of plants expressing NHL3-HA show label. In both cases, cell walls do not exhibit label (arrow heads in B and C). Moreover, the label is restricted to the peripheral side of the cytoplasm as visible near the nucleus (arrows in C and D), which is visualized by the concomitant 4,6-diamidino-2-phenylindole (DAPI) staining (D). Immunogold labeling of ultrathin sections exhibits the majority of label at the periphery of the cytoplasm adjacent to the cell wall (arrows in E). cw, cell wall; v, vacuole; m, mitochondrion. Bars = 10 μm in A through D; bar = 0.5 μm in E.

Figure 6.

NHL3-HA is highly enriched in the upper phase of two-phase partitioning fraction. Two-phase partitioning was used to separate plasma membrane and internal membranes. Five micrograms of proteins from microsomal (M), upper phase (U), and lower phase (L) preparations was loaded with 2 μL of total (T) extracts (approximately 30 mg of ground tissue boiled in 200 μL of SDS buffer) and was subjected to western blotting. Antibodies specific for amino acids 6 through 51 of AHA2, a plasma membrane-localized proton ATPase, and against the plant-specific endoplasmic reticulum lumenal binding protein BiP were used to show the efficiency of the fractionation process. Identical results were obtained in two different transgenic lines, and only a representative experiment is shown.