Arabidopsis thaliana PAD4 encodes a lipase-like gene that is important for salicylic acid signaling

Abstract

The Arabidopsis PAD4 gene previously was found to be required for expression of multiple defense responses including camalexin synthesis and PR-1 gene expression in response to infection by the bacterial pathogen Pseudomonas syringae pv. maculicola. This report describes the isolation of PAD4. The predicted PAD4 protein sequence displays similarity to triacyl glycerol lipases and other esterases. The PAD4 transcript was found to accumulate after P. syringae infection or treatment with salicylic acid (SA). PAD4 transcript levels were very low in infected pad4 mutants. Treatment with SA induced expression of PAD4 mRNA in pad4-1, pad4-3, and pad4-4 plants but not in pad4-2 plants. Induction of PAD4 expression by P. syringae was independent of the regulatory factor NPR1 but induction by SA was NPR1-dependent. Taken together with the previous observation that pad4 mutants have a defect in accumulation of SA upon pathogen infection, these results suggest that PAD4 participates in a positive regulatory loop that increases SA levels, thereby activating SA-dependent defense responses.

Figure 1

Positional cloning and structure of the PAD4 gene. A 5-cM region between CAPS markers m457 and AFC1 was partially spanned with YAC, BAC, and cosmid clones. The number of recombination events between PAD4 and a particular marker among 620 chromosomes tested is shown below the marker. ●, Right ends, and ■, left ends of YAC and BAC clones. Cosmids 7, 8, 21, 23, 24, 30, and 35 and the indicated 5.6-kb BamHI fragment from cosmid 8 complemented the pad4–1 mutation. Shading indicates the region common to all these cosmids. Only 2 of the 13 noncomplementing cosmids are shown here.

Figure 2

Complementation of the camalexin-deficient phenotype (A), enhanced bacterial growth phenotype (B), and the PR-1 transcript accumulation phenotype (C) of pad4–1 by cosmids 8 and 21. Wild-type (Col), pad4–1, and transgenic pad4–1 containing cosmid 8 or cosmid 21 were infected with Psm ES4326. Camalexin levels in infected leaves were determined 48 hr after infection. Bacterial titer was determined 3 days after infection, and PR-1 mRNA levels were determined 36 hr after infection. For A and B, each bar represents the mean and SD of six replicate samples. In C, the 18S rRNA probe was used to evaluate uniform loading. Similar results were obtained in another independent experiment.

Figure 3

Structure of the PAD4 gene showing the position of the intron and all four mutations in the coding sequence and the 3′ untranslated region. Insertion of an extra T at nucleotide position 430 occurs in pad4–2, codon TGG at position 359 is changed to TAG in pad4–1, codon CAA at position 386 is changed to TAA in pad4–3, and a G is missing from codon 513 in pad4–4. The underlined region displays sequence similarity to triacylglycerol lipases and esterases as shown in Fig. 4. cDNA 1 starts at nucleotide 46. PAD4 is located on the sequenced BAC clone F2206 (GenBank accession no. AL050300.1).

Figure 4

Amino acid sequence comparison of the predicted PAD4 protein with other lipase and lipase-like genes. The putative lipase catalytic triad consisting of a serine, histidine, and aspartate is indicated by arrows. RhizoTGL, triacylglycerol lipase precursor 1 from Rhizomucor miehei; FusaTGL, triacylglycerol lipase from Fusarium heterosporum; Rhizolip, triacylglycerol lipase precursor 1 from Rhizomucor niveus; Thermolip, lipase from Thermomyces lanuginosus; AspFAE, ferulic acid esterase A from Aspergillus niger; AtEDS1, A. thaliana EDS1; AtPAD4, A. thaliana PAD4. Invariant residues are indicated in bold letters, and conserved amino acids are underlined.

Figure 5

After infection by Psm ES4326, PAD4 transcript levels are very high in wild-type plants and greatly reduced all pad4 mutant alleles. Leaves from wild-type (Col) and all four pad4 mutants were excised 0, 6, 12, 24, 36, or 48 hr after infection. Mg indicates leaves mock-inoculated with 10 mM MgSO₄ and harvested after 36 hr. Similar results were obtained in another independent experiment.

Figure 6

PAD4 mRNA is induced by SA in wild-type, pad4–1, pad4–3, and pad4–4 but not in pad4–2. Wild-type (Col and Ler) and pad4 plants were treated with 5 mM SA in 0.02% Silwet L-77 (vol/vol) until uniformly wet. Control samples were treated with 0.02% Silwet L-77 (H₂O). (A) Wild-type (Col) plants were sprayed with 5 mM SA, and PAD4 mRNA levels were determined 0, 6, 12, 24, and 36 hr after treatment. (B) Wild-type (Col and Ler) and pad4 plants were treated with 5 mM SA, and PAD4 mRNA levels were determined 0 and 6 hr after treatment. Similar results were obtained in another independent experiment.

Figure 7

PAD4 mRNA induction by Psm ES4326 is SA-dependent but NPR1-independent whereas induction by SA is NPR1-dependent. (A) Wild-type (Col and Ler), nahG, and npr1–1 plants were infected with Psm ES4326. Samples were analyzed for PAD4 mRNA 36 hr after infection. Mg indicates leaves mock-inoculated with 10 mM MgSO₄ and harvested at 36 hr. (B) Wild-type (Col) and npr1–1 plants were treated with 5 mM SA, and PAD4 mRNA levels were determined at 0, 6, and 12 hr. Control samples were treated with 0.02% Silwet L-77 (H₂O). Similar results were obtained in another independent experiment.

Figure 8

Proposed models for the roles of PAD4, SA, and NPR1 in defense gene expression. (Model 1) SA is necessary but not sufficient for activation of expression of defense genes including PAD4. Another component is required—either NPR1 or some unknown factor X from the pathogen. NPR1 also inhibits SA accumulation. (Model 2) Different SA levels modulate PAD4 activity differently. Low SA levels activate and very high SA levels inactivate PAD4. Activated PAD4, in turn, stimulates expression of defense genes and inhibits the repressing activity of NPR1 on the SA amplification loop. Very high SA levels turn PAD4 off. In this situation, NPR1 activity is required for defense gene expression.