Background. Copper sulfate (CuSO4) is a widely used quarantine treatment for saltwater aquarium fishes due to its antimicrobial properties. Treatment consists of gradually dripping a CuSO4 solution into the aquarium until therapeutic copper levels are reached and maintained for 3 weeks. However, exposure to copper may negatively impact microbes essential for nitrification and biofiltration, processes critical to water quality and animal health. Despite its broad application, the susceptibility of these microbial communities to CuSO4, and whether its effects differ between established and naïve aquarium systems, remains unclear.
Methods. Water from two saltwater aquarium systems was sampled longitudinally during and after completion of CuSO4 treatments. One system had never been treated with CuSO4 (naïve), while the other had been running for over a decade with periodic treatments (established). While the naïve system underwent intermittent dosing periods to complete treatment, the established system received continuous dosing. Sampling was carried out over 147-day (n = 97) and 168-day (n = 128) periods, respectively. DNA was isolated using the DNeasy PowerSoil Pro Kit (Qiagen). Shotgun sequencing libraries (n = 225) were prepared using the Illumina DNA Prep Kit and sequenced on an Illumina NovaSeq 6000 platform. Raw sequencing data underwent adapter removal and quality filtering using Trimmomatic. High-quality reads were taxonomically classified with Kraken2 using NCBI’s ‘core_nt’ database. Reads were also mapped to the MEGARes v3.0 database using the AMR++ pipeline. Metagenome-assembled genomes (MAGs) were recovered by assembling contigs with MEGAHIT and binning with METABAT, followed by quality assessment using CheckM. High-quality MAGs (>90% completeness and <5% contamination) were taxonomically classified using GTDB-Tk and annotated using Bakta. To carry out phylogenetic analyses, MAGs classified as nitrifying taxa were used to construct pangenomes using Panaroo. Maximum-likelihood trees were then inferred from core pangenome alignments with IQ-TREE. Finally, to assess functional differences among MAGs, predicted protein sequences from Bakta were annotated using EggNOG-mapper.
Results. Overall microbial community composition differed between system types; however, nitrifying community shifts were less pronounced. Notably, the effects of CuSO4 depended on the timing of exposure to therapeutic copper levels. In the naïve system, the response of nitrifying communities differed between the first-ever and second intermittent exposures, while in the established system, changes in these communities were observed primarily after treatment had concluded. Although the naïve system showed an overall higher prevalence of copper resistance genes, these were not correlated with copper levels in water. Taxa belonging to the Nitrosopumilaceae (AOA) family dominated the ammonia-oxidizing community. Phylogenetic analyses of recovered AOA MAGs revealed that genomes from the first-ever exposure to therapeutic copper levels (early-phase MAGs) in the naïve system clustered separately from those recovered during the second exposure (late-phase MAGs). Furthermore, AOA MAGs recovered post-treatment in the established system clustered closely with the late-phase MAGs from the naïve system. None of these AOA MAGs could be assigned to a known species. Functional analyses further revealed no differences in the presence of copper resistance genes. However, genes encoding multicopper oxidases, potentially involved in ammonia oxidation, were absent in early-phase AOA MAGs but present in all late-phase MAGs.
Significance. By leveraging key data about susceptible and resilient nitrifying taxa in these aquatic ecosystems, proactive approaches could be tailored to sustain nitrification and biofiltration during CuSO4 quarantine treatments.