Indole-3-acetic acid (IAA) biosynthesis in the smut fungus Ustilago maydis and its relevance for increased IAA levels in infected tissue and host tumour formation

Abstract

Infection of maize (Zea mays) plants with the smut fungus Ustilago maydis is characterized by excessive host tumour formation. U. maydis is able to produce indole-3-acetic acid (IAA) efficiently from tryptophan. To assess a possible connection to the induction of host tumours, we investigated the pathways leading to fungal IAA biosynthesis. Besides the previously identified iad1 gene, we identified a second indole-3-acetaldehyde dehydrogenase gene, iad2. Deltaiad1Deltaiad2 mutants were blocked in the conversion of both indole-3-acetaldehyde and tryptamine to IAA, although the reduction in IAA formation from tryptophan was not significantly different from Deltaiad1 mutants. To assess an influence of indole-3-pyruvic acid on IAA formation, we deleted the aromatic amino acid aminotransferase genes tam1 and tam2 in Deltaiad1Deltaiad2 mutants. This revealed a further reduction in IAA levels by five- and tenfold in mutant strains harbouring theDeltatam1 andDeltatam1Deltatam2 deletions, respectively. This illustrates that indole-3-pyruvic acid serves as an efficient precursor for IAA formation in U. maydis. Interestingly, the rise in host IAA levels upon U. maydis infection was significantly reduced in tissue infected with Deltaiad1Deltaiad2Deltatam1 orDeltaiad1Deltaiad2Deltatam1Deltatam2 mutants, whereas induction of tumours was not compromised. Together, these results indicate that fungal IAA production critically contributes to IAA levels in infected tissue, but this is apparently not important for triggering host tumour formation.

Figure 1

Trp‐dependent IAA biosynthesis. The scheme shows IAA pathways and key intermediates proposed for plants and microorganisms. Trp, tryptophan; IPA, indole‐3‐pyruvic acid; IAAld, indole‐3‐acetaldehyde; IAA, indole‐3‐acetic acid; Tol, indole‐3‐ethanol; TAM, tryptamine; N‐TAM, N‐hydroxyl tryptamine; IAOx, indole‐3‐acetaldoxime; IAN, indole‐3‐acetonitrile; IAM, indole‐3‐acetamide. Enzymes involved in these pathways are: Trp aminotransferase (1), IPA decarboxylase (2), IAAld dehydrogenase (3), IAAld reductase (4), Trp decarboxylase (5), flavin monooxygenase‐like enzymes (6), cytochrome P450 enzymes (7); nitrilase (8); Trp monooxygenase (9); IAM hydrolase (10). Possible intermediates or by‐products of IAA formation thus far reported for fungi are shaded grey (see Discussion). Enzymes for the conversion of intermediates connected by dashed arrows are elusive. U. maydis genes involved in IAA formation are indicated. Tol emerges as a by‐product from IAAld in the absence of IAAld dehydrogenase activity.

Figure 2

Carbon source‐dependent expression of iad2, tam1, tam2 and pad16. U. maydis strains FB1 (lanes 1, 3, 5) and FB2 (lanes 2, 4, 6) were cultivated in either CM without an additional carbon source (CM), CM/Glc or CM/Ara. For RNA‐blot analysis, total RNA (about 10 µg per lane for iad2, tam1 and tam2, and about 40 µg per lane for pad16) was loaded, and hybridized with the ³²P‐labelled probes as indicated. Radioactive signals were quantified. Filters were additionally hybridized with the constitutively expressed ppi gene (shown for pad16) to calculate the ratios of gene‐specific to ppi signals.

Figure 3

Sequence alignment of Iad2 with Iad1 and Pad2. Identical amino acids are boxed and shaded. Gaps have been inserted to increase the number of identities. Asterisks mark highly conserved amino acid residues described to be implicated in catalytic activity (E₂₅₃), substrate binding (C₂₈₈) and NAD binding (K₁₈₀, G₂₃₁, G₂₃₆; numbers refer to Iad2). Dots mark amino acids shared between Iad1 and Iad2 and absent from all remaining predicted aldehyde dehydrogenase sequences of U. maydis (see Discussion).

Figure 4

Demonstration of IAAld activity in vitro. (A) Immunoblot analysis of His‐tagged proteins expressed in E. coli strains GR1 (Iad1; 53.7 kDa), GR2 (Pad2; 56.2 kDa), GR16 (Pad16; 57.4 kDa) and GR5 (Iad2; 53.0 kDa) using an anti‐His antibody. Cultures were incubated under non‐inducing (–) or inducing conditions (+). Protein amounts corresponding to 40 µL of cell culture were loaded per lane. Size markers are indicated on the right. The right panel shows the Coomassie‐stained gel loaded with the same amounts of the same protein preparations. Arrowheads point to the Iad1, Pad2, Pad16 and Iad2 proteins expressed under inducing conditions. M: molecular mass standards. (B) HPLC analysis of in vitro reaction products. Identical volumes of enzyme extracts (corresponding to 0.7 mL of cell culture each and 100 µg protein for GR1, 200 µg for GR2 and 400 µg for GR5) prepared from E. coli strains GR1, 2 and 5 grown in parallel under inducing conditions (+Ara) were applied for the enzyme assay in the presence of IAAld (0.5 mm). Reaction products were separated by HPLC. The IAAld substrate (elutes 1 min earlier than IAA under the applied conditions) was removed from the samples via ethyl acetate extraction. Absorbance was monitored at 220 nm. Arrows denote peaks eluting like IAA. (C) Salkowski analysis of HPLC fractions. Fractions from HPLC runs of reaction products under non‐inducing conditions (open bars) and inducing conditions (closed bars) were collected for detection of IAA with the Salkowski reagent. Absorbance was measured at 530 nm. (D) Co‐factor and pH requirements of Iad2. The desalted extract from induced E. coli strain GR5 was incubated with IAAld (0.5 mm) in the presence of either NAD (1 mm) (left panel) or NADP (1 mm) (right panel). Reactions were allowed to proceed for 5 min (open bars) or 30 min (closed bars). Products were treated with the Salkowski reagent. Absorbance was measured at 530 nm.

Figure 5

IAA and Tol production in Δ iad1/Δ iad2 mutant strains. (A) Strains FB1, FB2, FB1Δiad1, FB2Δiad1, FB1Δiad2 (GRN1), FB2Δiad2 (GRN3), FB1Δiad1Δiad2 (GRN7) and FB2Δiad1Δiad2 (GRN8) were incubated in the presence of 0.2 mm IAAld, and IAA formation was determined in culture supernatants (150 µL) by the Salkowski reagent. Absorbance was measured at 530 nm. (B) HPLC analysis of extracted culture supernatants from strains FB2, FB2Δiad1, GRN3 and GRN8 cultivated as described in A. Absorbance was monitored at 220 nm. Arrows denote peaks eluting like IAA and Tol, respectively. (C) Quantification of IAA and Tol formation. Extracted culture supernatants from strains FB1, FB2, FB1Δiad1, FB2Δiad1, GRN7 and GRN8 incubated in CM/Ara in the presence of 20 µm IAAld were subjected to HPLC analysis and peak areas were quantified for calculation of IAA and Tol concentrations. The mean (± SD) values were calculated from three replicates of a single strain.

Figure 6

Determination of free IAA levels in maize tissue in response to U. maydis infection and analysis of proliferation of Δiad1Δiad2Δtam1Δtam2 mutants in planta. (A) Plants (6 days old) were inoculated with either combinations of FB1/FB2 strains (wild‐type; •), BH7/BH13 strains (Δiad1Δiad2Δnit1; Δ), or injected with water (mock‐control; ). (B) Plants were inoculated with either combinations of FB1/FB2 strains (wild‐type), FB1Δiad1Δiad2Δtam1/FB2Δiad1Δiad2Δtam1 strains (Δtam1), two independent combinations of FB1Δiad1Δiad2Δtam1Δtam2/FB2Δiad1Δiad2Δtam1Δtam2 strains (Δtam1Δtam2), or injected with water (mock‐control). (A,B) Concentrations of IAA are calculated in pmol/g fresh weight. Average values and standard deviations of three data points are given. Leaf material was collected at the time points indicated. For the 12‐h and 48‐h time points, material was collected 0.5–3 cm above ground and 0.5–3 cm below the injection site, respectively. For the 4‐, 6‐ and 9‐day time points, chlorotic or early leaf tumour (4 days) and leaf tumour (6 and 9 days) tissue was collected between the ligule and > 1 cm below the injection site. All parts were exclusively from the third and fourth leaves. Non‐infected control material was isolated correspondingly. For each time point, ten or more tissue samples were collected. (C) Maize plants were inoculated with either mixtures of FB1/FB2 (wild‐type), FB1Δiad1Δiad2/FB2Δiad1Δiad2 (Δiad1Δiad2) or FB1Δiad1Δiad2Δtam1Δtam2/FB2Δiad1Δiad2Δtam1Δtam2 (Δiad1Δiad2Δtam1Δtam2) strains. Two days after inoculation samples from infected leaf blade tissue were stained with Chlorazol Black E. Note the ramification of hyphae throughout the epidermal layer. The bar (10 µm) refers to all panels. (D) Detection of FB1Δiad1Δiad2/FB2Δiad1Δiad2 (Δiad1Δiad2) and FB1Δiad1Δiad2Δtam1Δtam2/FB2Δiad1Δiad2Δtam1Δtam2 (Δiad1Δiad2Δtam1Δtam2) strain combinations in maize tumours. Chromosomal DNA (100 ng) isolated from each of five individual tumours (Tum1–5) 6 days after inoculation with either of these combinations was used as template for PCR to amplify a fungal‐specific DNA fragment (see Experimental procedures). C1, C2: DNA (100 ng) isolated from the respective strain combinations prior to plant infection was used as template. Twenty‐six and 30 cycles (numbers below the lanes) were performed. The expected size of the amplified fragment is 633 bp. All lanes are from the same gel. Phage lambda (M) DNA (500 ng) digested with PstI was used as size marker.