Expanding the cyanuric acid hydrolase protein family to the fungal kingdom

Abstract

The known enzymes that open the s-triazine ring, the cyanuric acid hydrolases, have been confined almost exclusively to the kingdom Bacteria and are all homologous members of the rare cyanuric acid hydrolase/barbiturase protein family. In the present study, a filamentous fungus, Sarocladium sp. strain CA, was isolated from soil by enrichment culturing using cyanuric acid as the sole source of nitrogen. A reverse-genetic approach identified a fungal cyanuric acid hydrolase gene composed of two exons and one intron. The translated spliced sequence was 39 to 53% identical to previously characterized bacterial cyanuric acid hydrolases. The sequence was used to generate a gene optimized for expression in Escherichia coli and encoding an N-terminally histidine-tagged protein. The protein was purified by nickel affinity and anion-exchange chromatography. The purified protein was shown by (13)C nuclear magnetic resonance ((13)C-NMR) to produce carboxybiuret as the product, which spontaneously decarboxylated to yield biuret and carbon dioxide. The protein was very narrow in substrate specificity, showing activity only with cyanuric acid and N-methyl cyanuric acid. Barbituric acid was an inhibitor of enzyme activity. Sequence analysis identified genes with introns in other fungi from the Ascomycota that, if spliced, are predicted to encode proteins with cyanuric acid hydrolase activity. The Ascomycota cyanuric acid hydrolase homologs are most closely related to cyanuric acid hydrolases from Actinobacteria.

Fig 1

Schematic showing the experimental steps from enrichment culturing to gene discovery, protein purification, and protein characterization.

Fig 2

Fractionation of cyanuric acid hydrolase activity from a Sarocladium sp. crude cell extract and MS/MS identification of cyanuric acid hydrolase peptide homologs. (A) SDS-PAGE analysis of the fraction with cyanuric acid hydrolase activity that was eluted from DEAE resin with 0.2 M KCl. MW, molecular weight (in thousands). (B and C) Mass spectra of two cyanuric acid hydrolase peptide homologs identified in the active fraction by database matching (B) or de novo sequencing (C). The y and/or b ions that identified each residue are indicated.

Fig 3

Exon (solid lines) and intron (dashed lines) structures of confirmed and predicted fungal cyanuric acid hydrolase family genes (lengths are proportional). Nucleotide positions of exon boundaries are indicated. (A) The Sarocladium sp. CA authentic splicing pattern, as confirmed by RT-PCR. (B) The G. lozoyensis ATCC 74030 gene was predicted in the unannotated genomic sequence in GenBank (GI:361128271). The Pseudocercospora fijiensis CIRAD86 gene was annotated in GenBank (GI:452984457) as encoding a hypothetical protein in the cyanuric acid hydrolase family. The Mycosphaerella populorum SO2202 gene was predicted from a two-exon gene annotated in GenBank as encoding a hypothetical fusion protein with an N-terminal cyanuric acid hydrolase family domain (GI:453082292). The single exon version shown above uses an in-frame stop codon that was identified with the FGENESH algorithm on the Softberry server.

Fig 4

¹³C-NMR analysis of ¹³C-labeled cyanuric acid hydrolysis by the Sarocladium sp. CA cyanuric acid hydrolase. Shown is the composite spectrum acquired immediately after addition of the enzyme (A) and the composite spectrum acquired the next day (B).

Fig 5

Maximum likelihood phylogenetic tree of cyanuric acid hydrolase/barbiturase family homologs, the consensus of 500 bootstrap replicates, rooted with the Frankia sp. EuI1c sequence. Bootstrap values are indicated at each node. Black triangles represent collapsed branches containing sequences from strains of the same genus or monophyletic homologs (barbiturases). The number of sequences in each set of collapsed branches is identified by an Arabic numeral in parentheses. The two separate subclades of Bradyrhizobium species sequences are identified by bracketed Roman numerals. Phyla of the source strains are indicated by labeled brackets or lines. The GI numbers of sequences are shown in Table S1 in the supplemental material.

Fig 6

Alignment (ClustalW) of the Sarocladium sp. CA cyanuric acid hydrolase (CAH) sequence with AtzD from Pseudomonas sp. ADP showing conservation of the active-site lysine-serine dyads (highlighted) identified in the AtzD crystal structure reported by Peat et al. (42). The black rectangles indicate the highly conserved sequence regions surrounding the catalytic dyad residues.