Molecular mechanism underlying biofloc formation of Acidovorax soli Q11 and its application potential Authors: Yaqi Liu, Na Gao, Jingcheng Dai, Shuangyuan Liu, Ansi Jiang, Yan Huang, Chang Wu, Dongru Qiu Published: 2026 ABSTRACT AbstractBiofloc technology (BFT) is an emerging microbial-based technology in aquaculture that enhances system productivity by assimilating excessive nitrogenous pollutants, such as ammonia nitrogen and nitrite, into biofloc biomass via the activity of biofloc-forming microbes, rather than relying on conventional nitrification-denitrification or anammox processes. Traditionally, BFT has been implemented by enhancing the carbon/nitrogen ratio (C/N) to stimulate the uptake of ammonia and the biosynthesis of amino acids and proteins of biofloc-forming microbes, whose biomass could be filter-fed and digested by filter-feeding aquatic animals for nutrients. In this study, we isolated a series of biofloc-forming bacteria from multiple aquaculture systems and systematically investigated the molecular mechanisms governing biofloc formation in Acidovorax soli Q11, isolated from a freshwater sturgeon aquaculture pond. Molecular genetic analyses conducted on strain Q11, combined with whole genome sequencing, revealed that biofloc formation is likely regulated by cellulose biosynthesis (bcs) genes. Further characterization of extracellular polymeric substances (EPS) via Fourier transform infrared spectroscopy (FTIR) and scanning electron microscopy (SEM) demonstrated that the secreted EPS exhibited high physicochemical similarity to bacterial cellulose. However, the bioflocs displayed pronounced resistance to cellulase degradation. Notably, the genome of strain Q11 contains two distinct bcs gene clusters, corresponding to Type Ic (harboring bcsD) and Type IIc (harboring bcsE and bcsG), suggesting that this dual bcs operon system may contribute to the enhanced structural stability of the cellulose-like matrix. Furthermore, strain Q11 demonstrated efficient removal of both ammonium (NH4+) and nitrite (NO₂⁻), achieving complete removal of initial NH4+ (39 mg/L) within 36 h and NO₂⁻ (65 mg/L) within 77 h, without detectable accumulation of toxic nitrogenous intermediates. This strain also exhibited versatile carbon source utilization, including effective growth on substrates commonly applied in aquaculture systems, such as glucose, brown sugar, fermentation-derived carbon sources, and rice bran. Collectively, these findings provide new insights into the mechanism of biofloc formation and highlight the potential of Acidovorax soli Q11 as a promising candidate for bioaugmentation in BFT systems, with potential applications in sustainable aquaculture and the treatment of nitrogen-rich wastewater.Key points• Biofloc formation mechanism in Acidovorax soli Q11 is mediated by a dual bcs operon system.• EPS secreted by strain Q11 exhibits high structural stability and cellulase resistance.• Strain Q11 enables efficient ammonium and nitrite removal, supporting wastewater bioaugmentation.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00253-026-13873-3. Introduction The escalating global population has driven a substantial increase in demand for aquatic products, a critical protein source (FAO 2022). To meet this need, intensive farming systems have been developed in the aquaculture industry, yet this shift has resulted in the significant precipitation and accumulation of undesirable metabolic byproducts and undigested protein-rich feed residues (Gutierrez-Wing and Malone 2006; Choudhury et al. 2022). Consequently, elevated ammonia nitrogen and nitrite concentrations in aquaculture water have deteriorated water quality and subsequently caused disease outbreaks and increased mortality rates among cultured species (Choudhury et al. 2022). Conventional aquaculture practices relied on extensive water exchange and purification to mitigate these issues, but this method incurs significant operational costs (Gutierrez-Wing and Malone 2006). By contrast, biofloc technology (BFT), a self-purification technology requiring little or no water exchange, has emerged as a novel, sustainable, and environmentally-friendly application in the aquaculture industry (Schryver et al. 2008; Ahmad et al. 2017). However, prolonged operation of low- or zero-water-exchange BFT systems may result in the accumulation of refractory dissolved organic matter (DOM) and conservative inorganic salts, which may compromise long-term chemical stability. BFT, first proposed by Professor Yoram Avnimelech, is primarily based on manipulating the C/N ratio in aquaculture water to promote the growth of heterotrophic microorganisms (Avnimelech 1999). This microbial-driven process facilitates the conversion of nitrogenous toxic waste into microbial biomass, thereby simultaneously achieving water purification and nutrient recycling promotion (Azim and Little 2008). Notably, the formation of stable biological flocs is central to the effectiveness of BFT, wherein biofloc-forming bacteria provide habitats and attachment sites for protozoa and other particulate matter through extracellular polymeric substances (EPS) production (Fang and Liu 1988; Qiu et al. 2019). At present, the molecular mechanisms underlying biofloc formation remain poorly understood, particularly in aquaculture-derived systems. Biofloc technology shares conceptual similarities with the activated sludge method, in which floc-forming bacteria constitute the structural backbone of microbial aggregates. In activated sludge systems, floc formation in most dominant floc-forming genera, such as Zoogloea and Thauera, is mainly mediated through coordinated mechanisms involving extracellular polysaccharides, PrsR-PrsK two-component systems, and PEP-CTERM domain-containing proteins, whereas in Shinella, floc formation has additionally been reported to be mediated through extracellular cellulose polymers (Gao et al. 2018, 2022). Despite increasing studies on microbial community composition in BFT systems, mechanistic understanding at the molecular level remains scarce. Currently, BFT faces several challenges, including complex microbial composition, difficulty in controlling bacterial community structure, and potential pollution from nitrate accumulation. In particular, bioflocs often exhibit poor stability, reflected in compromised structural integrity, erratic sedimentation behavior, and reduced resistance to hydrodynamic shear stress under dynamic aquaculture conditions. Such instability may be further exacerbated by microbial succession and variations in EPS composition (Cardona et al. 2016). These challenges highlight the need to identify biofloc-forming bacteria with more stable floc-forming performance under aquaculture conditions. Our colleagues Liu et al. (2019) previously attempted to enhance biofloc formation in aquaculture by adding the dominant floc-forming bacterium Zoogloea resiniphila MMB strain that we isolated from activated sludge (AS). However, the inability of this strain to metabolize glucose, a critical carbon source used for sustaining microbial consortia in their BFT systems, may be the main reason for the failure of bioaugmentation (An et al. 2016; Liu et al. 2019). Compared to activated sludge-derived bacteria, the indigenous biofloc-forming bacteria in aquaculture ponds may exhibit greater ecological fitness for biofloc formation due to their prior adaptation to aquaculture-specific environmental conditions, including local nutrient composition, carbon source availability, and microbial community interactions, while better preserving native microbial community stability. Based on this hypothesis, indigenous biofloc-forming bacteria were isolated from a freshwater aquaculture system, and a strain belonging to the genus Acidovorax capable of forming stable bioflocs and utilizing glucose was isolated from the sturgeon farming pond. Notably, the genus Acidovorax has been reported not only as one of the 28 core bacterial taxa strongly linked to activated sludge performance, but also as an abundant genus in BFT systems with tapioca starch as a carbon source, supporting its ecological relevance and potential competitive advantage in aquaculture environments (Deng et al. 2018; Wu et al. 2019). This study aimed to elucidate the molecular mechanisms underlying biofloc formation in the isolated strain and to evaluate its potential for ammonia and nitrite removal, thereby providing mechanistic insights for developing targeted strategies to improve floc stability, enhance settling performance, and optimize microbial community control in BFT systems. Materials and methods Isolation and identification of bacteria and genome annotation Freshwater aquaculture samples were collected from the sturgeon aquaculture ponds in Yichang City, Hubei Province, PR China (N 29°56′, E 110°15′). Biofloc-forming bacteria used in the study were isolated from freshwater samples using the conventional dilution spread plate method, followed by screening based on their biofloc formation phenotypes in liquid culture. The isolated strains were stored in our lab at −80 °C with 40% (v/v) glycerol and taxonomically identified via 16S rRNA gene amplification using universal primers 27 F and 1492R (Liu et al. 2022). During this process, a strain with stable floc formation was selected and designated Q11. The genomic DNA of the isolated strains was extracted using a commercial Bacterial DNA Extraction kit (Omega, USA) according to the manufacturer’s instructions and subsequently sequenced using the Illumina HiSeq2000 platform. The draft genomes were de novo assembled using SPAdes (version 3.13.0) and functionally annotated using the RAST (Rapid Annotation using Subsystems Technology) server. To improve annotation reliability, the predicted protein-coding genes were further verified by BLASTp searches against the NCBI non-redundant (NR) database with an e-value cutoff of 1e − 5. The identification of bacterial cellulose biosynthesis (bcs) gene clusters was based on sequence homology and functional annotation, and their classification was determined according to gene composition and organization, following the criteria described by Romling and Galperin (2015). Culture conditions The bacterial strains and plasmids used in this study are listed in Table S1 of the Supplementary Materials. Escherichia coli WM3064 was cultured at 37 °C in Luria-Bertani (LB) medium supplemented with diaminopimelic acid (DAP) at a final concentration of 50 µg/mL. All other bacterial strains used were grown at 28 °C in R2A medium. Antibiotic selection was performed with 15 µg/mL of gentamicin (Gm) or 50 µg/mL of kanamycin (Km) as required. Construction of transposon insertional mutant library and genetic complementation To investigate the molecular mechanisms underlying biofloc formation, a transposon insertional mutant library comprising biofloc formation-deficient mutants was constructed, followed by genetic complementation of the disrupted genes. In this study, the plasmids used for transposon mutagenesis and genetic complementation were pMiniHimar RB1 (Km resistance) and pBBR1-MCS5 (Gm resistance), respectively (Kovach et al. 1995; Bouhenni et al. 2005). To generate the transposon insertional mutant library, Escherichia coli WM3064 carrying the pMiniHimar RB1 transposon delivery suicide plasmid was cultured in LB-DAP-Km medium for 36 h and used as the donor strain (An et al. 2016). The biofloc-forming bacterium Acidovorax soli Q11, isolated in this study, was used as the recipient for biparental conjugation. Biofloc formation-deficient mutants were screened by observing the biofloc-related phenotypic characteristics of transconjugants in R2A-Km liquid medium, and the identified mutants were further characterized by high-efficiency thermal asymmetric interlaced PCR (TAIL-PCR) to determine the transposon insertion sites as previously described (Wang et al. 2019; Liu and Chen 2007). For genetic complementation analyses, the target genes were PCR-amplified using wild-type genomic DNA from strain Q11 as the template, ligated into the pBBR1-MCS5 vector, and subsequently transferred into the corresponding mutant strains to assess functional restoration (Gao et al. 2018). In parallel, the pBBR1-MCS5 empty vector was introduced into the Q11 wild-type and mutant strains as controls, and phenotypic changes were monitored through liquid cultivation and microscopic observation. The primers used for PCR amplification are listed in Table S2. Extracellular polymeric substances (EPS) purification The extracellular polymeric substances (EPS) from Acidovorax soli Q11 were purified using an alkali treatment method adapted from bacterial cellulose purification protocols (Bilgi et al. 2016). For the preparation of the seed inoculum, A. soli Q11 was cultivated in a 250 mL Erlenmeyer flask containing 100 mL of R2A medium at 28 °C with shaking at 180 rpm for 24 h. The resulting inoculum (10%, v/v) was then transferred into fresh R2A medium and incubated statically at 28 °C for 7 days to facilitate EPS formation (Fan et al. 2016). After cultivation, the EPS-rich biomass was harvested and thoroughly rinsed with deionized water to remove residual culture medium. To eliminate bacterial cells and reduce labile non-cellulosic EPS fractions (e.g., proteins and soluble polysaccharides), the biomass was boiled in a 4% (w/v) NaOH solution for 30 min, which may also preferentially enrich cellulose-like fractions and thereby potentially influence subsequent FTIR and SEM characterizations. Subsequently, the sample was sequentially rinsed with deionized water, 0.1% (v/v) acetic acid, and deionized water until neutral pH was achieved, resulting in a semi-transparent EPS preparation (Bilgi et al. 2016). Fourier transform infrared spectroscopy (FTIR) and scanning electron microscopy (SEM) analyses For FTIR analysis, the purified EPS was freeze-dried and analyzed using a Nicolet 5700 FTIR spectrometer to characterize its chemical functional groups and structural features. The scanning conditions were set as follows: a spectral range of 4000–400 cm−1, 24 scans, and a resolution of 4 cm−1. For SEM analysis, the centrifuged bacterial cells or purified EPS were fixed in 2.5% glutaraldehyde fixative (for electron microscopy) for 4 h at 4 °C. The samples were then washed three times with 0.1 M phosphate buffer (PB, pH 7.4) for 15 min each. Subsequently, the samples were subjected to a graded ethanol dehydration series (30%, 50%, 70%, 80%, 90%, 95%, 100%, 15 min at each concentration). After dehydration, the samples were freeze-dried, sputter-coated with gold, and examined using a field-emission scanning electron microscope (Crossbeam 350, Zeiss, Germany). Effects of cellulase on biofloc formation and stability As many microorganisms ubiquitously distributed in natural environments possess cellulase-producing potential, and given that the EPS secreted by strain Q11 is associated with bacterial cellulose synthesis mediated by the bcs gene cluster, evaluating its feasibility as a candidate strain for biofloc technology (BFT) requires assessing the structural stability of the formed bioflocs. To this end, this study conducted cellulase degradation experiments to systematically investigate the stability of EPS-derived bioflocs. The biofloc-forming strain Shinella zoogloeoides ATCC 19623T was selected as a positive control, as it exhibits a bcs-regulated biofloc formation mechanism similar to that of strain Q11 (Gao et al. 2022). The cellulase used in this study was a commercially available preparation (10,000 U/g), which was diluted to a working activity of 1000 U/mL for the experiments. Cellulase treatment was applied both before and after biofloc formation, and the stability of the bioflocs was evaluated by monitoring phenotypic changes in liquid cultures. For the pre-biofloc degradation assay, different volumes of cellulase (0, 5, 50, 100, 200, and 500 µL) were added to 3 mL of R2A medium to assess cellulase tolerance and its effect on biofloc formation. For the post-biofloc degradation assay, different volumes of cellulase (0, 50, and 500 µL) were incubated with preformed bioflocs (after removal of the culture medium) in 50 mM sodium citrate buffer (pH 4.8) at 45 °C to evaluate the enzymatic disruption and stability of mature bioflocs, as described previously (Nwamba et al. 2021; Gao et al. 2022). Evaluation of removal efficiency for high concentrations of ammonium and nitrite In this study, the removal efficiency of strain Q11 toward high concentrations of ammonium (NH4+) and nitrite (NO2−) was evaluated in defined culture medium (C/N ratio of 18–20), which contained NaNO2 (0.24 g/L) or NH4Cl (0.2 g/L) as the sole nitrogen source, sodium succinate (C4H4Na2O46H2O, 5.5 g/L) as the carbon source, K2HPO4 (1.0 g/L), KH2PO4 (1.0 g/L), MgSO47H2O (0.2 g/L), and trace element solution (2 mL/L) at an initial pH of 7.0 ± 0.3. The composition of the trace element solution (1L) was as follows: ZnSO47H2O 0.1 g, MnCl2·4H2O 0.4 g, H3BO3 1.24 g, CoCl2·2H2O 0.5 g, CuCl2·2H2O 0.5 g, NiSO4·6H2O 0.02 g, NaMoO4·2H2O 0.4 g, KI 0.2 g, CaCl2 5.0 g, FeSO47H2O 1.1 g, and EDTA-2Na 10.0 g. The bacterial strain Q11 preserved at −80 °C was streaked onto R2A agar plates for reactivation and purification. A single colony was then inoculated into liquid medium and incubated at 28 °C for 24 h. Cells (1 mL) were harvested by centrifugation (8000 × g, 5 min) and washed twice with sterile deionized water to remove residual medium components. The resulting cell suspension was transferred into fresh defined medium containing different nitrogen sources (160 mL of medium per flask) and incubated at 28 °C with shaking at 180 rpm for 3–5 days. During cultivation, samples were collected at intervals of 6–8 h and filtered through a 0.22 µm membrane. The concentrations of ammonium (NH4+) and nitrite (NO2−) in the supernatant were measured using standard colorimetric methods with a UV–Vis spectrophotometer (UV-1800, MAPADA), in accordance with Chinese National Standards (GB/T 7479–1987 and GB/T 7493–1987), as previously described by Liu et al. (2021). Briefly, NH4+ was quantified based on the formation of a yellow-brown complex with Nessler reagent, while NO2− was determined via diazotization reaction with N-(1-naphthyl) ethylenediamine dihydrochloride (NED). All experiments were performed in triplicate, and data are presented as mean ± standard deviation (SD) of three independent replicates. These values were used to plot degradation curves with error bars. Statistical analysis All statistical analyses were performed using one-way ANOVA followed by Tukey’s post-hoc test using the Statistical Package for Social Sciences (SPSS) (19.0). Data are expressed as the mean ± standard deviation (SD). Differences among groups were considered statistically significant at P < 0.05 and indicated by different lowercase letters. Carbon source utilization analysis Carbon source utilization analysis of Acidovorax soli Q11 was evaluated using a Biolog GEN III Microplate (Biolog, USA). Briefly, colonies grown on R2A agar plate were suspended in Inoculating Fluid A (IF-A), and the cell density was adjusted to approximately 90% transmittance using a turbidimeter, according to the manufacturer’s instructions. The resulting cell suspension was then inoculated into the GEN III Microplate at a volume of 110 µL per well and incubated at 28 °C for up to 7 days to monitor substrate utilization–associated color development. A purple color development was interpreted as a positive (+) reaction, whereas the absence of color change was recorded as a negative (−) reaction. Results Identification and genetic complementation of mutant strains In this study, a biofloc-forming strain, designated Q11, was isolated from a sturgeon aquaculture pond in Yichang, Hubei Province, China, and its microscopic morphology is shown in Figure S1. The 16S rRNA gene sequence of strain Q11 exhibited the highest similarity to Acidovorax soli BL21T (99.37%), and a phylogenetic tree constructed using the neighbor-joining method is presented in Figure S2. Our previous studies have demonstrated that both extracellular polysaccharides biosynthesis and the expression of PEP-CTERM domain-containing proteins are essential for floc formation in Zoogloea resiniphila MMB and other floc-forming bacteria (An et al. 2016; Gao et al. 2018; Wang et al. 2019; Liu et al. 2020). However, whole-genome sequencing analysis revealed that strain Q11 lacks several key genes previously reported to be involved in floc formation, including asnH, asnB, prsT, and epsB2, which have been identified in Z. resiniphila MMB and Pseudoduganella eburnea YN12 (An et al. 2016; Gao et al. 2018; Liu et al. 2020). These observations suggest that strain Q11 may employ a distinct biofloc formation mechanism. Based on this hypothesis, the biofloc-forming mechanism of strain Q11 was investigated by constructing a transposon insertion mutant library. Through phenotypic screening, a total of 35 biofloc formation-deficient mutants were identified. Among these mutants, eight independent transposon insertions were mapped to the same gene encoding cellulose synthase operon protein C (bcsC), three insertions were mapped to the gene encoding the cellulose synthase catalytic subunit A (bcsA), two insertions were mapped to the gene encoding the cellulose biosynthesis protein (bcsQ), and two insertions were mapped to the gene encoding the cellulose synthase regulator protein (bcsB). In addition to these bcs-related genes, several mutants were also mapped to genes involved in polysaccharide precursor synthesis and protein secretion, including UTP-glucose-1-phosphate uridylyl-transferase and components of the Type II secretion system (gspD, gspF, gspI, and gspE). The transposon insertion sites of these mutants are summarized in Table 1. Table 1The list of biofloc formation deficient mutantsMutantsProteinTransposon insertion sitesAx2Cellulose synthase catalytic subunit BcsACGGCTGTGTTACCTCAACGCAx8GCCCGCCCCTACGTGGTGCTAx20CCACTGGCTTACCTGTTTTTTAx3Cellulose synthase operon protein C BcsCGGCGCTGTATGGCATGGCGCAx9GCGGGGCTCTACATCCCCGAAx10CATGGGGGCTACTTCAGCCCAx12CGGCGAGCTATTACCCGGAAAx15GACTATGCCTATGCCGAGTCAx16GGCGCTGTATGGCATGGCGCAx18CGGCGGCATGTACTGGCATGAx22GCGGCCCCCTACTTCCCCAAAx4UTP-glucose-1-phosphate uridylyl-transferaseGCAAGCGCTACGACTGCGGCAAx7TCCGGGAATATCTGAAGGCCCTAx5Cellulose biosynthesis protein BcsQAGGCCCTGGCCTATGACCAGAx14TGGTACATACCTATTGCACAAx6Type II secretion system protein D GspDAGGACACCTACTCGCAGGGCGAx26GCAGGCCTGTACAAGGTGGTGCAx19Type II secretion system inner membrane protein GspFCCGACATCTACTGCGCCGTGATAx21Cellulose synthase regulator protein BcsBGGGCACCTACCTGGGGCTCAAx27GCGCAGTGGTACCTGTCGCGAx23Type II secretion system minor pseudopilin GspIACCCGAGCTTTAGGCGCATCGAAx25Type II secretory pathway ATPase GspECCCGCCGTAGAGGACTTGCTGG Table 1The list of biofloc formation deficient mutantsMutantsProteinTransposon insertion sitesAx2Cellulose synthase catalytic subunit BcsACGGCTGTGTTACCTCAACGCAx8GCCCGCCCCTACGTGGTGCTAx20CCACTGGCTTACCTGTTTTTTAx3Cellulose synthase operon protein C BcsCGGCGCTGTATGGCATGGCGCAx9GCGGGGCTCTACATCCCCGAAx10CATGGGGGCTACTTCAGCCCAx12CGGCGAGCTATTACCCGGAAAx15GACTATGCCTATGCCGAGTCAx16GGCGCTGTATGGCATGGCGCAx18CGGCGGCATGTACTGGCATGAx22GCGGCCCCCTACTTCCCCAAAx4UTP-glucose-1-phosphate uridylyl-transferaseGCAAGCGCTACGACTGCGGCAAx7TCCGGGAATATCTGAAGGCCCTAx5Cellulose biosynthesis protein BcsQAGGCCCTGGCCTATGACCAGAx14TGGTACATACCTATTGCACAAx6Type II secretion system protein D GspDAGGACACCTACTCGCAGGGCGAx26GCAGGCCTGTACAAGGTGGTGCAx19Type II secretion system inner membrane protein GspFCCGACATCTACTGCGCCGTGATAx21Cellulose synthase regulator protein BcsBGGGCACCTACCTGGGGCTCAAx27GCGCAGTGGTACCTGTCGCGAx23Type II secretion system minor pseudopilin GspIACCCGAGCTTTAGGCGCATCGAAx25Type II secretory pathway ATPase GspECCCGCCGTAGAGGACTTGCTGG The list of biofloc formation deficient mutants To further verify the role of these genes in biofloc formation, the bcsC gene was PCR-amplified using gene-specific primers, cloned into the broad-host-range shuttle plasmid pBBR1-MCS5, and introduced into the corresponding mutant strain via conjugation. The resulting complemented strain (pBBR1-MCS5-bcsC) fully restored the biofloc formation phenotype, confirming that bcsC gene is directly involved in biofloc formation in strain Q11 (Fig. 1). In addition, genetic complementation analyses demonstrated that the bcsA, bcsQ, and bcsB genes are also functionally associated with the biofloc formation phenotype of strain Q11.Fig. 1The genetic complementation analysis of the bcsC gene in transposon mutant Ax9. A The biofloc-forming wild-type strain Q11 carrying the empty pBBR1-MCS5 vector under differential interference microscope. B The gene bcsC-deficient mutant Ax9 carrying the empty vector. C The Ax9 mutant carrying the recombinant vector pBBR1-MCS5-bcsC. D Agitated cell culture of A, B, and C (from left to right). E Settled cell culture of A, B, and C (from left to right). Bar: 10 µm The genetic complementation analysis of the bcsC gene in transposon mutant Ax9. A The biofloc-forming wild-type strain Q11 carrying the empty pBBR1-MCS5 vector under differential interference microscope. B The gene bcsC-deficient mutant Ax9 carrying the empty vector. C The Ax9 mutant carrying the recombinant vector pBBR1-MCS5-bcsC. D Agitated cell culture of A, B, and C (from left to right). E Settled cell culture of A, B, and C (from left to right). Bar: 10 µm Identification of the gene cluster required for biofloc formation Integrated genomic analysis and mutant mapping revealed that the four genes bcsQ, bcsB, bcsA, and bcsC, disrupted by 15 independent biofloc-deficient mutants in strain Q11, are co-localized within a single gene cluster of approximately 17 kb in size (Fig. 2). This cluster also includes the gene encoding cellulose synthase subunit D (bcsD).Fig. 2The bacterial cellulose synthase (bcs) gene cluster in Acidovorax soli Q11 and insertion sites of biofloc formation-deficient mutants The bacterial cellulose synthase (bcs) gene cluster in Acidovorax soli Q11 and insertion sites of biofloc formation-deficient mutants In contrast to the floc-formation mechanism in Z. resiniphila MMB, which involves the exopolysaccharide biosynthesis, PrsK-PrsR two-component system and PEP-CTERM domain-containing protein family (An et al. 2016; Gao et al. 2018), biofloc formation in strain Q11 appears to be regulated predominantly by a series of bacterial cellulose synthase (bcs) genes. This mechanism closely resembles that reported for Shinella zoogloeoides ATCC 19623T, in which the bcsA and bcsB subunits catalyze cellulose polymerization and translocation, while bcsC is predicted to form a β-barrel in the outer membrane (Omadjela et al. 2013; Gao et al. 2022). Taken together, these results strongly indicated that the EPS secreted by strain Q11 is predominantly cellulose in nature. Characterization of extracellular polymeric substances (EPS) To further characterize the EPS secreted by strain Q11, EPS was extracted and subsequently analyzed using Fourier transform infrared spectroscopy (FTIR) and scanning electron microscopy (SEM). The FTIR spectra (Fig. 3) of the EPS secreted by strain Q11 exhibited characteristic absorption bands at 3410 cm−1 (O–H stretching vibration), 2930 cm−1 (C–H stretching vibration), 1640 cm−1 (O–H bending vibration), 1380 cm−1 (C–H deformation vibration), and 1060 cm−1 (C–O stretching vibration) (Oh et al. 2005). The absorption peak at 1640 cm−1 is attributed to the O–H bending vibration of absorbed water, whereas the band at 1060 cm−1 is associated with C–OH and C–O–C stretching vibrations of the pyranose ring, which are characteristic of cellulose (Quintana-Quirino et al. 2019; Ye et al. 2019). Notably, these spectral features are highly similar to those of commercial cellulose and consistent with the characteristic peaks reported for natural cellulose (Antunes et al. 2024), suggesting that the EPS produced by strain Q11 is predominantly cellulose-based. However, given the intrinsic limitations of FTIR as a qualitative analytical technique, minor non-cellulosic components, if present at low abundance, cannot be completely excluded.Fig. 3Fourier transform infrared spectroscopy (FTIR) spectra of extracellular polymeric substances (EPS) secreted by Acidovorax soli Q11 and commercial cellulose Fourier transform infrared spectroscopy (FTIR) spectra of extracellular polymeric substances (EPS) secreted by Acidovorax soli Q11 and commercial cellulose SEM analysis of strain Q11 revealed an aggregated cellular morphology, in which numerous viable cells were coated and interconnected by EPS (Fig. 4A). Similar filamentous or EPS-mediated intercellular connections have been reported in Pseudomonas aeruginosa, Staphylococcus aureus, and Methylobacterium species (Zhang et al. 2020; Park et al. 2020). Notably, genomic analyses have identified at least two bacterial cellulose synthase operons in Methylobacterium species that are implicated in cell aggregation and biofilm formation (Gao et al. 2024). Furthermore, SEM observation of the purified EPS revealed a reticular and fibrous microstructure (Fig. 4B), collectively supporting the conclusion that the EPS secreted by strain Q11 is composed predominantly of bacterial cellulose (Li et al. 2021).Fig. 4Scanning electron microscopy (SEM) analysis of strain Q11 and its EPS. A SEM micrograph showing the aggregated cellular morphology of strain Q11, bar: 500 nm. B SEM micrograph of the cellulose-based EPS secreted by strain Q11, exhibiting a reticular microstructure Scanning electron microscopy (SEM) analysis of strain Q11 and its EPS. A SEM micrograph showing the aggregated cellular morphology of strain Q11, bar: 500 nm. B SEM micrograph of the cellulose-based EPS secreted by strain Q11, exhibiting a reticular microstructure Effects of cellulase on biofloc formation and stability Given that many microorganisms in natural environments possess cellulase-producing potential, the stability of the EPS produced by strain Q11 was evaluated by cellulase degradation assays conducted both before and after biofloc formation. The growth and aggregation behaviors of the tested strains in cellulase-supplemented medium, in which cellulase was added prior to biofloc formation, are shown in Fig. 5A. The addition of 5 µL cellulase completely inhibited biofloc formation in S. zoogloeoides ATCC 19623T, whereas strain Q11 retained the ability to form visible aggregates even in the presence of up to 500 µL cellulase. The corresponding phenotypic changes during cultivation are shown in Figure S3a.Fig. 5Effects of cellulase supplementation on biofloc formation and stability in Acidovorax soli Q11. The cellulase activity was 1000 U/mL. A Phenotypic appearance of cultures supplemented with different volumes of cellulase prior to biofloc formation. B The OD600 values of culture supernatants after 48 h of incubation under different cellulase concentrations. Data are presented as mean ± SD (n = 3). Different letters within the same strain indicate significant differences among treatments based on one-way ANOVA followed by Tukey’s test (P < 0.05). C Microscopic observations of cultures under cellulase-supplemented conditions after incubation for 48 h, bar: 10 µm Effects of cellulase supplementation on biofloc formation and stability in Acidovorax soli Q11. The cellulase activity was 1000 U/mL. A Phenotypic appearance of cultures supplemented with different volumes of cellulase prior to biofloc formation. B The OD600 values of culture supernatants after 48 h of incubation under different cellulase concentrations. Data are presented as mean ± SD (n = 3). Different letters within the same strain indicate significant differences among treatments based on one-way ANOVA followed by Tukey’s test (P < 0.05). C Microscopic observations of cultures under cellulase-supplemented conditions after incubation for 48 h, bar: 10 µm To quantitatively assess floc disruption and cell dispersion, the OD600 values of culture supernatants were measured after 48 h of incubation (Fig. 5B). Low OD600 values were observed in cultures without cellulase (0.047 for strain Q11 and 0.074 for S. zoogloeoides ATCC 19623T). Upon cellulase supplementation, the OD600 values of strain Q11 increased only moderately (0.056–0.153), whereas those of S. zoogloeoides ATCC 19623T increased sharply (1.264–1.877), indicating extensive floc disruption and cell dispersion in the latter. Interestingly, the OD600 values at 500 µL cellulase addition in S. zoogloeoides ATCC 19623T are lower than those at 200 µL, which may be attributed to changes in particle aggregation or sedimentation behavior following extensive EPS degradation. Microscopic observations further supported these results (Fig. 5C). Strain Q11 maintained compact cell aggregates even under 500 µL cellulase treatment, while S. zoogloeoides ATCC 19623T exhibited a fully dispersed cellular morphology upon exposure to only 5 µL cellulase. To exclude potential interference from culture medium components on cellulase activity, degradation assays were additionally conducted after biofloc formation using 50 mM sodium citrate buffer (pH 4.8). Consistent results were obtained (Figure S3c): pronounced EPS degradation was observed for S. zoogloeoides ATCC 19623T within 30 min at 45 °C, whereas strain Q11 remained floc structural integrity even after 48 h of incubation with 500 µL cellulase. Collectively, these results demonstrate that cellulase effectively inhibits biofloc formation and stability in S. zoogloeoides ATCC 19623T, but only partially affects strain Q11. Overall, the bioflocs formed by strain Q11 exhibit higher structural stability and greater resistance to cellulase treatment compared with those formed by S. zoogloeoides ATCC 19623T. Evaluation of removal efficiency for high concentrations of ammonium and nitrite The accumulation of ammonium (NH4+) and nitrite (NO₂⁻), originating from the decomposition of residual feed and metabolic by-products, represents a critical constraint on water quality in aquaculture systems (Cunha et al. 2020; Hoang et al. 2022). Biofloc technology (BFT), which relies on microbial metabolic activities, has been demonstrated to effectively mitigate these nitrogenous pollutants (Cunha et al. 2020). Therefore, microbial strains capable of degrading NH4+ or NO₂⁻ are of considerable interest for enhancing BFT performance. In this study, the ammonium (NH4+) and nitrite (NO2−) removal capacities of strain Q11 were evaluated using NH4Cl and NaNO2 as the sole nitrogen sources, respectively. As shown in Fig. 6, when NH4Cl served as the sole nitrogen source, strain Q11 initiated NH4+ reduction after 8 h and achieved complete degradation by approximately 36 h, without detectable accumulation of NO2− throughout the process (Fig. 6a). When NaNO2 was provided as the sole nitrogen source, strain Q11 exhibited significant NO2− degradation after 36 h and completed removal around 71 h, while no accumulation of NH4+ was observed during the entire cultivation period (Fig. 6b).Fig. 6Evaluation of nitrogen removal performance of strain Q11 toward ammonium (NH4+) and nitrite (NO₂⁻). a Temporal variations in NH4+ and NO2− concentrations during cultivation with NH4Cl as the sole nitrogen source. b Temporal variations in NH4+ and NO2− concentrations during cultivation with NaNO2 as the sole nitrogen source. Different letters indicate significant differences among time points for NH₄⁺ in (a) and NO₂⁻ in (b), based on one-way ANOVA followed by Tukey’s multiple comparison test (P < 0.05) Evaluation of nitrogen removal performance of strain Q11 toward ammonium (NH4+) and nitrite (NO₂⁻). a Temporal variations in NH4+ and NO2− concentrations during cultivation with NH4Cl as the sole nitrogen source. b Temporal variations in NH4+ and NO2− concentrations during cultivation with NaNO2 as the sole nitrogen source. Different letters indicate significant differences among time points for NH₄⁺ in (a) and NO₂⁻ in (b), based on one-way ANOVA followed by Tukey’s multiple comparison test (P < 0.05) Collectively, these results indicate that Acidovorax soli Q11 possesses high-efficiency removal capabilities for both NH4+ and NO2− without intermediate metabolite accumulation, confirming its potential as a functional strain for BFT applications in aquaculture systems to mitigate nitrogen pollution and prevent the accumulation of toxic nitrogenous compounds. Carbon source utilization analysis of strain Q11 Given that Z. resiniphila MMB, as reported by Liu et al. (2019), is unable to utilize glucose for growth, the carbon source utilization profile of strain Q11 was investigated using the Biolog GEN III microplate system. The results (Table S3) showed that strain Q11 was capable of utilizing a broad spectrum of substrates as sole carbon and energy sources, including dextrin, glucose, mannose, fructose, sorbitol, mannitol, arabitol, glycerol, glycyl-L-proline, L-alanine, L-aspartic acid, L-glutamic acid, L-histidine, L-serine, D-gluconic acid, D-saccharic acid, L-lactic acid, α-ketoglutaric acid, malic acid, bromo-succinic acid, Tween 40, γ-amino-butyric acid, α-hydroxybutyric acid, c-hydroxybutyric acid, α-keto-butyric acid, acetoacetic acid, propionic acid, acetic acid, D-fructose-6-phosphate, D-galactonic acid lactone, D-glucuronic acid, and methyl pyruvate. Furthermore, the growth performance of strain Q11 on carbon sources commonly applied in aquaculture systems was evaluated under defined conditions (C/N = 18, NH4Cl as the sole nitrogen source), including glucose, sucrose, starch, sodium acetate, sodium succinate, fermentation carbon source, rice bran, and brown sugar. As shown in Fig. 7, growth and biofloc formation of strain Q11 varied among the tested carbon sources. More pronounced growth and biofloc formation were observed in the presence of glucose, sodium succinate, and brown sugar, while relatively moderate growth was observed with fermentation carbon source and rice bran. In contrast, only limited but visible growth was observed when starch and sucrose were supplied, and almost no visible growth occurred with sodium acetate. These results indicate that strain Q11 possesses a broader carbon utilization capacity than Z. resiniphila MMB, supporting its adaptability to diverse and practical carbon sources relevant to aquaculture applications.Fig. 7Growth performance of Acidovorax soli Q11 on different carbon sources commonly used in aquaculture systems Growth performance of Acidovorax soli Q11 on different carbon sources commonly used in aquaculture systems Discussion Analysis of the biofloc formation mechanism in strain Q11 Genomic annotation analysis revealed the presence of two distinct cellulose synthase (bcs) gene clusters in the genome of strain Q11 (Figure S4), each containing the core genes bcsA, bcsB, bcsC, and bcsQ. Bacterial cellulose synthase operons are generally classified into four types based on gene composition and organization (Romling and Galperin 2015). In strain Q11, gene cluster A was classified as a Type Ic bcs operon, which typically contains bcsD, whereas gene cluster B belonged to a Type IIc bcs operon, characterized by the presence of bcsE and bcsG genes (and the absence of bcsD gene). Notably, the bcsZ gene, encoding an endo-β−1,4-glucanase, was absent from gene cluster A in strain Q11. Although the loss of bcsZ has been reported to reduce cellulose production or impair biofilms formation in other bacterial species, its absence did not compromise the biofloc-forming capacity of strain Q11, suggesting the existence of compensatory or alternative regulatory mechanisms. To further investigate the functional relationship between the two bcs operons, genetic complementation experiments were performed. Introduction of the recombinant plasmid pBBR1-MCS5-bcsB2, harboring the bcsB gene from gene cluster B, into a mutant deficient in bcsB gene from gene cluster A failed to restore the biofloc-forming phenotype (Figure S5). Similarly, complementation with bcsC derived from gene cluster B did not rescue biofloc formation in the corresponding bcsC mutant of gene cluster A. These results indicate pronounced functional divergence between the two bcs operons, despite their apparent structural similarity. The coexistence of these two non-redundant bcs gene clusters in strain Q11 therefore enables coordinated regulation of cellulose biosynthesis and secretion, thereby providing enhanced flexibility and robustness in cellulose production and biofilm formation. This functional specialization may also contribute to the observed resistance of EPS secreted by strain Q11 to cellulase-mediated degradation. Comparative genomic analysis of 13 type strains within the genus Acidovorax revealed that only six strains harbored bcs operons, all of which were classified as Type Ic or Type IIc (Fig. 8), indicating that cellulose biosynthesis is not a conserved trait in this genus. This limited distribution suggests that the maintenance of bcs operons, particularly dual-operon systems, may impose a metabolic burden and is therefore selectively retained under specific environmental conditions (Romling and Galperin 2015).Fig. 8Comparative genomic analysis of bacterial cellulose synthase (bcs) gene clusters in Acidovorax type strains Comparative genomic analysis of bacterial cellulose synthase (bcs) gene clusters in Acidovorax type strains In aquaculture ponds, environmental pressures such as intensive aeration, continuous water circulation, complex microbial interactions, and exposure to extracellular enzymes may favor the evolution and maintenance of robust EPS-producing systems (Flemming and Wingender 2010). The dual-operon architecture observed in strain Q11 may therefore represent a potential adaptive strategy, enabling functional partitioning or differential regulation of cellulose biosynthesis pathways. As different bcs operon types have been associated with distinct regulatory and functional roles (Romling and Galperin 2015), it is plausible that one operon supports basal cellulose production, whereas the other is preferentially involved in stress-responsive or structurally specialized cellulose synthesis. Such a division of labor could confer enhanced regulatory flexibility, thereby improving the ability of strain Q11 to maintain biofloc stability under fluctuating environmental conditions, consistent with the established role of EPS in floc formation and structural integrity (Schryver et al. 2008). Notably, the bcs operons identified in strain Q11 exhibited the highest similarity to those in type strain A. kalamii KNDSW-TSA6T, which likewise possesses two bcs operons. However, strain Q11 uniquely encoded the bcsP gene, whose function remains unknown, and neither bcs operon was detected in its closest phylogenetic relative, A. soli BL21T. These differences in genetic architecture suggest that strain Q11, together with A. kalamii KNDSW-TSA6T, may have greater potential for biofloc formation and related bioremediation applications than closely related strains lacking this system. Comparison with previously reported heterotrophic nitrifiers/aerobic denitrifiers (HN-AD) To contextualize the nitrogen removal performance of strain Q11 in aquaculture-relevant systems, a comparison with representative heterotrophic nitrifiers and aerobic denitrifiers reported in BFT-related studies was conducted (Table 2). Direct comparisons should be interpreted with caution due to differences in experimental conditions (e.g., initial nitrogen concentration, carbon source, inoculum size, and cultivation conditions). Overall, strain Q11 appears to exhibit comparable or competitive NH4+ and NO2− removal efficiencies under the tested conditions. Table 2Comparison of NH4+ and NO2− removal performance of strain Q11 with representative heterotrophic nitrifiers/aerobic denitrifiers (HN-AD) reported in aquaculture-related systemsStrainCarbon sourceRemoval of NH4+ and NO2−Key conditionsReferenceAcidovorax soli Q11Sodium succinateNH4+: 98.6% (33 mg/L, 36 h)NO2−: 99.7% (65 mg/L, 77 h)C/N = 18–20, inoculum: 0.63% (v/v)This studyPseudomonas sp. DM02GlucoseNH4+: > 98% (10 mg/L, 12 h)pH 7, C/N = 12, inoculum: 1% (v/v) Deng et al. (2021) Bacillus litoralis N31Sodium succinateNH4+: 86.3% (20 mg/L, 48 h)NO2−:89.3% (20 mg/L, 48 h)pH 7.5, C/N = 10, inoculum: 1% (v/v) Huang et al. (2017) Stutzerimonas frequens TF18Sodium acetateNH4+: 89.21% (20 mg/L, 24 h)NO2−: 99.76% (20 mg/L, 24 h)C/N = 10, inoculum: 1% (v/v) Liu et al. (2023) Acinetobacter sp. LF10Sodium acetateNH4+: 98.02% (51 mg/L, 24 h)NO2−: 86.84% (51 mg/L, 24 h)pH 7.3, C/N = 8–10, inoculum: 1% (v/v) Li et al. (2025) Acinetobacter calcoaceticus TY1Sodium succinateNH4+: 97.47% (106 mg/L, 36 h)NO2−: 97.51% (35 mg/L, 18 h)pH 7.2, inoculum: 7% (v/v), 8 °C Wu et al. (2022) Pseudomonas putida NP5Sodium succinateNH4+: > 99% (100 mg/L, 24 h)NO2−: 99.44% (100 mg/L, 24 h)pH 7.5, C/N = 10, inoculum: 1% (v/v) Yang et al. (2019) Table 2Comparison of NH4+ and NO2− removal performance of strain Q11 with representative heterotrophic nitrifiers/aerobic denitrifiers (HN-AD) reported in aquaculture-related systemsStrainCarbon sourceRemoval of NH4+ and NO2−Key conditionsReferenceAcidovorax soli Q11Sodium succinateNH4+: 98.6% (33 mg/L, 36 h)NO2−: 99.7% (65 mg/L, 77 h)C/N = 18–20, inoculum: 0.63% (v/v)This studyPseudomonas sp. DM02GlucoseNH4+: > 98% (10 mg/L, 12 h)pH 7, C/N = 12, inoculum: 1% (v/v) Deng et al. (2021) Bacillus litoralis N31Sodium succinateNH4+: 86.3% (20 mg/L, 48 h)NO2−:89.3% (20 mg/L, 48 h)pH 7.5, C/N = 10, inoculum: 1% (v/v) Huang et al. (2017) Stutzerimonas frequens TF18Sodium acetateNH4+: 89.21% (20 mg/L, 24 h)NO2−: 99.76% (20 mg/L, 24 h)C/N = 10, inoculum: 1% (v/v) Liu et al. (2023) Acinetobacter sp. LF10Sodium acetateNH4+: 98.02% (51 mg/L, 24 h)NO2−: 86.84% (51 mg/L, 24 h)pH 7.3, C/N = 8–10, inoculum: 1% (v/v) Li et al. (2025) Acinetobacter calcoaceticus TY1Sodium succinateNH4+: 97.47% (106 mg/L, 36 h)NO2−: 97.51% (35 mg/L, 18 h)pH 7.2, inoculum: 7% (v/v), 8 °C Wu et al. (2022) Pseudomonas putida NP5Sodium succinateNH4+: > 99% (100 mg/L, 24 h)NO2−: 99.44% (100 mg/L, 24 h)pH 7.5, C/N = 10, inoculum: 1% (v/v) Yang et al. (2019) Comparison of NH4+ and NO2− removal performance of strain Q11 with representative heterotrophic nitrifiers/aerobic denitrifiers (HN-AD) reported in aquaculture-related systems NH4+: 98.6% (33 mg/L, 36 h) NO2−: 99.7% (65 mg/L, 77 h) NH4+: 86.3% (20 mg/L, 48 h) NO2−:89.3% (20 mg/L, 48 h) NH4+: 89.21% (20 mg/L, 24 h) NO2−: 99.76% (20 mg/L, 24 h) NH4+: 98.02% (51 mg/L, 24 h) NO2−: 86.84% (51 mg/L, 24 h) NH4+: 97.47% (106 mg/L, 36 h) NO2−: 97.51% (35 mg/L, 18 h) NH4+: > 99% (100 mg/L, 24 h) NO2−: 99.44% (100 mg/L, 24 h) Carbon source utilization of strain Q11 and its implications for biofloc performance The versatile carbon metabolism of strain Q11 is essential for its performance in biofloc systems, as the carbon-to-nitrogen (C/N) ratio plays a critical role in regulating microbial growth, extracellular polymeric substance (EPS) production, and nitrogen transformation processes. Notably, cellulose biosynthesis mediated by the dual-bcs operon system may impose an additional carbon demand, potentially limiting the carbon available for microbial growth and the removal of NH4+ and NO2−. However, the broad substrate utilization capacity of strain Q11 may mitigate this trade-off by enabling efficient assimilation of diverse carbon sources under variable conditions, thereby supporting both biofloc structural stability and nitrogen removal performance. Nevertheless, the coupling mechanisms among carbon metabolism, cellulose biosynthesis, and nitrogen transformation in strain Q11 remain unclear. Further investigations are therefore required, and the optimal C/N ratio for simultaneously maximizing biofloc stability and nitrogen removal efficiency warrants systematic evaluation. Conclusion In this study, a biofloc-forming bacterium, Acidovorax soli Q11, was isolated from freshwater aquaculture systems. Integrated whole-genome sequencing, transposon mutagenesis, and genetic complementation analyses demonstrated that biofloc formation in strain Q11 is driven by a cellulose-based extracellular polymeric substances (EPS) synthesis pathway mediated by bacterial cellulose synthase (bcs) genes, which was further supported by FTIR and SEM analyses revealing typical cellulose-like structural characteristics. Notably, the cellulose-based EPS secreted by strain Q11 exhibited pronounced resistance to cellulase degradation, which is likely associated with the presence of two functionally divergent and non-redundant bcs operons in its genome, suggesting a potential mechanistic basis for enhanced EPS structural integrity and biofloc stability compared to previously reported systems. In addition, strain Q11 demonstrated efficient removal of ammonium (NH4+) and nitrite (NO2−) under the tested conditions, without detectable accumulation of intermediate nitrogenous compounds, together with a versatile carbon source utilization capacity, including growth on several substrates commonly applied in aquaculture systems, such as fermentation-derived carbon sources, brown sugar, and rice bran, indicating functional robustness and metabolic adaptability. These synergistic traits suggest that strain Q11 may serve as a potential candidate for bioaugmentation in biofloc-based and wastewater treatment systems, particularly within biofloc technology (BFT)-driven aquaculture applications. However, its role in maintaining long-term chemical stability under prolonged closed-system operation remains to be further verified, and sustained system performance is likely to require integrated management strategies beyond single-strain application. Future studies should therefore focus on evaluating the long-term stability, ecological interactions, and system-level performance of Q11 under realistic operational conditions. Supplementary Information Below is the link to the electronic supplementary material.ESM1(DOCX 3.94 MB) (DOCX 3.94 MB)