Divergent roles in Arabidopsis thaliana development and defense of two homologous genes, aberrant growth and death2 and AGD2-LIKE DEFENSE RESPONSE PROTEIN1, encoding novel aminotransferases

Abstract

The disease-resistant Arabidopsis thaliana aberrant growth and death2 (agd2-1) mutant has elevated levels of the defense signal salicylic acid (SA), altered leaf morphology, and mild dwarfism. AGD2 and its close homolog ALD1 (for AGD2-LIKE DEFENSE RESPONSE PROTEIN1) encode aminotransferases that act on an overlapping set of amino acids in vitro. However, kinetic parameters indicate that AGD2 and ALD1 may drive the aminotransferase reaction in opposite directions. ALD1-deficient mutants have the opposite phenotypes from agd2-1, showing reduced SA production and increased disease susceptibility. Furthermore, ALD1 transcript levels are elevated in agd2-1 and are induced in the wild type by bacterial pathogen infection. ALD1 is responsible for some of the elevated SA content and a majority of the disease resistance and dwarfism of agd2-1. A complete knockout of AGD2 renders embryos inviable. We suggest that AGD2 synthesizes an important amino acid-derived molecule that promotes development and suppresses defenses, whereas ALD1 generates a related amino acid-derived molecule important for activating defense signaling.

Figure 1.

Features of AGD2, ALD1, and agd2-T. (A) Complementation of agd2-1. agd2-1 plants were transformed with the genomic clone of AGD2 (AGD2G, at right). The plants shown at the left and in the middle are untransformed wild-type and agd2-1 control plants. agd2-1 plants transformed with the genes adjacent to AGD2 were not complemented (data not shown). (B) Molecular structures of AGD2, ALD1, OsAGD2, and OsALD1. T-DNA insertion sites are indicated by triangles. The asterisk indicates the Pro-to-Ser change caused by the agd2-1 allele. Closed and open boxes represent exons, and closed boxes represent possible chloroplast transit peptides. The closed circles indicate the pyridoxal-5′-phosphate attachment sites. (C) Immature seeds inside a heterozygous agd2-T/AGD2 silique. Aborted dead seeds are indicated by red arrows. (D) Microscopy analysis of aborted and normal seeds in agd2-T/AGD2 heterozygotes. Embryos are indicated by red arrows. Top panel (I to III), dying and dead seeds; bottom panel (IV to VI), normal seeds. IV, Globular stage; V, triangular stage; VI, heart stage. Bars = 80 μm.

Figure 2.

Subcellular Localization of AGD2. The pGFP and pAGD2-GFP under control of the CaMV 35S promoter were transformed into A. thaliana. GFP fluorescence of the protoplasts was observed using a fluorescence microscope. (A) and (B) pAGD2-GFP. (C) and (D) pGFP.

Figure 3.

RNA Expression of AGD2 and ALD1. Tissue from the wild type and/or aldT-1 was used for RNA extraction and RNA gel blot analysis. (A) ALD1 and PR1 expression in the fourth and fifth leaves of 20-d-old wild-type (Col) and agd2-1 plants. (B) ALD1 and PR1 expression in wild-type (Ws) and ald1-T1 plants. Leaves (fourth and fifth) from 18-d-old plants were infected with P. s. maculicola DG3 (OD₆₀₀ = 0.01). (C) Expression of AGD2 and ALD1 during development of the fourth leaf. Leaf length and age of the plant when RNA was extracted are indicated. (D) Expression of AGD2 and ALD1 in the indicated tissues. (E) A time course of steady state AGD2 mRNA accumulation. Twenty-day-old plants were kept in a 16-h-light/8-h-dark (16L/8D) cycle or were shifted to continuous dark (D) for the indicated time after 4 h in the light of the normal light cycle (16-h-light/8-h-dark) and then switched back to light (L) for 1 or 4 h. Although the signal was low for the ALD1 transcript, the probe was verified to be high specific activity. It gave a strong signal on the blots probed in parallel (see [A] and [B]).

Figure 4.

Increased Disease Susceptibility in the ald1-T1 Mutant. (A) Disease susceptibility of ald1-T1 plants. Wild-type Ws (open circles) and ald1-T1 (open triangles) plants were infected with P. s. maculicola DG3 (PmaDG3, at left) or with P. s. tomato DC3000 (PtoDC3000, at right) at OD₆₀₀ = 0.0001. Growth of P. syringae in Ws and ald1-T1 was significantly different on days 2 and 3 (P < 0.01, t test, n = 8). Ws and ald1-T1 plants were pretreated with 100 μM BTH or water for 2 d and then subjected to infection with P. s. maculicola DG3. BTH-treated Ws (closed circles) and ald1-T1 (closed triangles) were not significantly different for bacterial growth (P > 0.5, t test, n = 8). cfu, colony-forming units. Bars indicate standard error. In some cases, the symbols obscure the error bars. (B) Complementation of the ald1-T1 mutant phenotype. Ws (open circles), ald1-T1 (open triangles), and ald1-T1 transformed with the ALD1 genomic clone (closed circles) were infected with P. s. maculicola DG3 (PmaDG3, at left) or P. s. tomato DC3000 (PtoDC3000, at right) at OD₆₀₀ = 0.0001. Growth of P. syringae in Ws and ald1-T1 was significantly different on days 2 and 3 (P < 0.01, t test, n = 8), whereas growth in Ws and the transgenic plants was not different (P > 0.7, t test, n = 8). (C) SA levels in Ws and ald1-T1 plants during pathogen infection. Leaves (fourth and fifth) from 18-d-old plants were infected with P. s. maculicola DG3 (OD₆₀₀ = 0.01) and were harvested, extracted, and analyzed by HPLC. The free and total SA values ±sd are averages of three sets of samples. The number of asterisks indicates samples that are significantly different from one another at a given confidence level (one asterisk, P < 0.004; two asterisks, P < 0.0008; three asterisks, P < 0.0002).

Figure 5.

Bacterial Growth in agd2-1 ald1-T2 Plants. Wild-type Col (open circles), ald1-T2 (open triangles), agd2-1 (open squares), and agd2-1 ald1-T2 (closed circles) were infected with P. s. maculicola DG3 at OD₆₀₀ = 0.0001. Growth of P. syringae in Col and agd2-1 ald1-T2 was not significantly different on days 2 and 3 (P > 0.7, t test, n = 8). Bars indicate standard error. In some cases, the symbols obscure the error bars.

Figure 6.

Purification and Relative Activities of AGD2, AGD2^P398S, and ALD1. (A) Electrophoretic patterns of purified recombinant proteins through column elution using Ni²⁺-NTA. (B) Relative aminotransferase activities of AGD2, AGD2^P398S, and ALD1. Substrates were at a 50-mM concentration. Aminotransferase activity was determined by measuring concentration of the reaction product Glu (using 2-oxoglutarate as the cosubstrate) or Ala (using pyruvate as the cosubstrate). The activity of AGD2 with Lys and 2-oxoglutarate was arbitrarily set at 100%. Asterisks indicate no detectable activities. No detectable activity was seen with all of the other standard amino acids tested. Bars indicate standard deviations (n = 3).

Figure 7.

Models for the Enzymatic Functions and Roles of AGD2 and ALD1 in Development and Disease Resistance. (A) Enzymatic reactions that AGD2 and ALD1 may conduct based on the in vitro data. These reactions may be performed in different subcellular compartments. (B) Integrated model for AGD2 and ALD1 functions in vivo. When plants are infected with P. s. maculicola DG3, ALD1 is needed for SA accumulation and activation of SA signaling responses, such as PR1 expression. ALD1 also is probably required for the generation of an SA-independent defense signal. AGD2 is required for normal development. Reduced activity of AGD2 results in developmental cell death and growth phenotypes and SA accumulation possibly because of the generation of an amino acid–derived signal or an amino acid imbalance. NPR1 also is shown as controlling development as an example of a protein that functions in defense and development (Vanacker et al., 2001).