Teliospores of Thecaphora frezii were collected in 2019 from hypertrophic peanut pods naturally infected in a single commercial field located at Criadero El Carmen, General Cabrera, Córdoba province, Argentina (32°49'44.2"S, 63°52'31.8"W). This region has a long history of peanut smut occurrence and fungicide use. According to local records, fungicides belonging to the triazole, SDHI, and strobilurin classes have been applied in this area since at least 2016 (Paredes, 2017). The collected material consisted exclusively of teliospores, confirmed by microscopic observation.

Teliospores from 20 infected pods were pooled, and total RNA was extracted using TRIzol reagent (Invitrogen) following the manufacturer’s instructions. RNA pellets were resuspended in 20 µL of RNase-free water and incubated at 55–60°C for 10 min. RNA integrity and purity were verified using agarose gel electrophoresis, a NanoPhotometer spectrophotometer (Implen), and quantification with the Qubit RNA assay kit and Qubit 2.0 Fluorometer (Life Technologies) (Díaz et al., 2024).

Complementary DNA (cDNA) libraries were prepared using the NEBNext Ultra RNA Library Prep Kit for Illumina (NEB) at INDEAR (Agrobiotechnology Institute of Rosario, Argentina). mRNA was purified using poly-T oligo-attached magnetic beads, fragmented using divalent cations, and reverse-transcribed into first-strand cDNA using random hexamer primers. Second-strand synthesis was performed with DNA polymerase I and RNase H. The resulting cDNA was size-selected, adaptor-ligated, and enriched. Index-coded samples were clustered using the cBot Cluster Generation System with the TruSeq PE Cluster Kit v3-cBot-HS (Illumina). Sequencing was performed on an Illumina HiSeq 1500 platform, generating 2 × 150 bp paired-end reads.

Raw reads were quality-checked using FastQC and assembled de novo using standard transcriptome pipelines. Read counts were normalized as counts per million (CPM), and gene expression levels were estimated. Differential expression analysis was conducted using a General Linear Model implemented in EdgeR (Robinson et al., 2010) within the R environment (R Core Team, 2018).

Genes associated with fungicide resistance in plant-pathogenic fungi were selected based on their well-established roles:

Protein sequences of Ustilago maydis, Pseudozyma flocculosa, Kalmanozyma brasiliensis, Sporisorium reilianum, and Moesziomyces antarcticus were retrieved from the NCBI protein database and used as queries in BLASTp searches to identify orthologous genes. Genomic data of Thecaphora thlaspeos (GenBank accession: GCA_900291925.1) (Courville et al., 2019) and T. frezii (ASM2628400v1) (Arias et al., 2023) were retrieved from NCBI. It is worth noting that T. thlaspeos is a urocystidial fungus within the Urocystidales, is phylogenetically distinct from T. frezii. Gene predictions were performed using AUGUSTUS, a web-based gene-prediction platform. Protein-coding genes analyzed in this study and their corresponding GenBank accession numbers are listed in Table 1.

Predicted protein sequences from T. frezii were aligned with orthologs from closely related fungi within the Ustilaginales and with sequences from Cercospora beticola and Botrytis cinerea, which have well-documented fungicide-resistance mutations (Secor et al., 2010; Fernández-Ortuño et al., 2012). Alignments were generated using Clustal Omega. Proteins were considered homologous if they shared ≥25% identity over >80 amino acids (Sander & Schneider, 1991). Amino-acid substitutions were evaluated based on conservation and physicochemical properties.

The structural and functional impact of identified variants was assessed using the HOPE web platform (Venselaar et al., 2010), which predicts changes in protein stability and function resulting from amino-acid substitutions.

Three conserved variants (R164K, K256R, I339L) were identified in the CYP51 gene (Figure 1). These substitutions involve residues with similar physicochemical properties and are thus unlikely to alter enzyme activity or fungicide binding significantly. A non-conserved substitution (R336L) was also detected exclusively in the transcriptome of T. frezii. This variant replaces a positively charged arginine with a hydrophobic leucine in the ligand-binding region.

Comparable substitutions, such as Y136F in Blumeria graminis, have been associated with azole resistance (Wyand & Brown, 2005). Moreover, silencing of the CYP51 gene has been shown to alter fungal sensitivity to triazoles, confirming its central role in fungicide resistance (Koch et al., 2013). Although triazole fungicides like cyproconazole have shown good control of peanut smut (>40% reduction in incidence; Paredes et al., 2021), the R336L mutation could represent an early adaptation event that merits further monitoring.

Because T. frezii populations are exposed to fungicides in the field, such mutations may arise from natural genetic variability rather than direct selection. Further population-level studies will be necessary to determine whether these variants confer any fitness advantage under fungicide pressure.

Two amino-acid substitutions were detected in SDHB: R255A and H257R (Figure 2). The latter aligns with the well-characterized H272R mutation in Botrytis cinerea associated with SDHI resistance (Fernández-Ortuño et al., 2012). The R255A substitution, found near this position, replaces a positively charged residue with a small nonpolar alanine, potentially affecting local folding and hydrogen-bond networks. Similar structural effects have been correlated with partial loss of SDHI sensitivity in other fungi (Sierotzki & Scalliet, 2013).

One variant was identified in SDHC (A93I), substituting a small alanine for a bulkier isoleucine near a residue homologous to R87 in B. cinerea (XJQ58198.1), another site linked to SDHI resistance. No variants were found in SDHD.

The relatively poor performance of SDHI fungicides against peanut smut in Argentina (Paredes et al., 2021) could therefore be linked to these subtle alterations. However, reported mutation rates for SDHI resistance in other smut or related fungi remain low (10⁻⁶–10⁻⁸ per generation) (Ishii & Hollomon, 2015), suggesting that T. frezii resistance, if present, is likely in an early stage of development.

Because our isolates were collected from a single field, these findings cannot be extrapolated to the entire pathogen population. Still, they underline the importance of regional monitoring for emerging resistant alleles.

No amino-acid variants were detected in CYTB or TUBB. The absence of CYTB mutations suggests that T. frezii populations have not yet developed resistance to strobilurins. Indeed, strobilurin-based fungicides achieved the highest control efficacy in field trials (Paredes et al., 2021; Table 2). However, studies in other species, such as Monilinia spp. (Hily et al., 2011) and Colletotrichum truncatum (Rogério et al., 2024), have shown that single-nucleotide mutations in CYTB can confer QoI resistance after repeated exposure. Continuous monitoring of T. frezii populations will be crucial to detect early resistance signals.

No variants were found in TUBB, which encodes β-tubulin, the target of benzimidazoles. Nevertheless, the low efficacy of benzimidazoles reported in vitro and in field trials (Paredes et al., 2021) may not necessarily indicate true resistance. It could instead reflect low pathogen sensitivity or limited fungicide uptake, as observed in Fusarium graminearum (Zhou et al., 2016). Furthermore, given our limited sample size, the absence of resistance mutations should be interpreted cautiously.

Fungicide resistance is often multigenic and may involve the accumulation of multiple mutations affecting different components of the target pathway (Lucas et al., 2015). For example, multiple CYTB and SDHB variants have been reported to jointly confer resistance to QoI fungicides in Puccinia horiana (Matsuzaki et al., 2021). The presence of several amino-acid substitutions in T. frezii may therefore reflect natural diversity that, under sustained selection, could evolve toward higher resistance levels.

Notably, Thecaphora thlaspeos, a related urocystidial fungus, is not known to be exposed to agricultural fungicides. To our knowledge, no published evidence of fungicide selection pressure has been reported for T. thlaspeos, which supports the hypothesis that observed polymorphisms in T. frezii are the result of intrinsic genetic variability rather than selection.

Overall, these findings emphasize the necessity of adopting integrated disease management approaches combining resistant cultivars, crop rotation, and rational fungicide use. The detection of conserved and unique variants in T. frezii provides a starting point for future genomic surveys across different regions to evaluate the distribution and frequency of potential resistance alleles.