Development of advanced bioinformatic profiles to improve the detection and functional understanding of fungal acid phosphatases

ABSTRACT We have retrieved approximately 9,000 protein sequences annotated as fungal acid phosphatase or phytase from the UniProtKB database. Following stringent quality filtering, a curated dataset comprising 3,058 high-confidence sequences was assembled. Phylogenetic analysis resolved these enzymes into eight distinct clades, representing distinct groups of fungal acid phosphatases: purple acid phosphatases, phytases, and groups containing both phytases and acid phosphatases annotations. Based on this classification, we have developed three representative protein profiles referred to as Prf-A-Fungal_phos, Prf-B-Fungal_phos, and Prf-C-Fungal_phos, each designed to capture the phylogenetic and functional diversity of these enzyme families. Heat-map analyses confirmed the breadth and high specificity of these profiles. Application of these profiles to public protein and metagenomic databases enabled the identification of hundreds of previously uncharacterized fungal proteins, with a broad taxonomic distribution and notable prevalence in the Ascomycota and Basidiomycota phyla. Functional validation through heterologous expression of selected candidates in Saccharomyces cerevisiae confirmed their phosphatase activity, supporting the accuracy of the in silico predictions. By integrating large-scale bioinformatics with experimental validation, this study provides robust tools for the discovery of novel fungal phosphatases and for investigation of their ecological roles in nutrient-limited environments. IMPORTANCE Fungal acid phosphatases are critical enzymes in global phosphorus cycling, yet no dedicated bioinformatic tools exist to comprehensively identify and classify them across fungal diversity. Here, we present the first PROSITE generalized profiles specific to fungal acid phosphatases, derived from a curated data set of over 3,000 high-confidence sequences spanning eight phylogenetic groups. These profiles exhibit high specificity and sensitivity, enabling the detection of hundreds of previously uncharacterized proteins from public protein databases. Experimental expression of representative candidates in Saccharomyces cerevisiae confirmed their phosphatase activity, validating our in silico predictions. By bridging large-scale bioinformatics with functional validation, this study delivers robust resources to uncover novel fungal phosphatases and to explore their ecological roles in nutrient-limited environments. The developed profiles will advance metagenomic annotation, support soil and environmental microbiology research, and foster biotechnological innovation in sustainable phosphorus management.