Bactericidal activity of peracetic acid to selected fish pathogens in recirculation aquaculture system water
Peracetic acid (PAA) has been identified as an efficacious alternative to common disinfectants used in aquaculture settings to reduce fish pathogen loads (Farmer et al., 2013; Liu et al., 2017; Pedersen et al., 2009; Sudová et al., 2010). Compared with certain aquaculture disinfectants, such as formalin (Jung et al., 2001), PAA is considered relatively environmentally friendly, given that it degrades to acetic acid, oxygen and water when applied in an aqueous environment, and does not leave toxic residues (Wagner et al., 2002). Additionally, PAA has been shown to be a useful therapeutant in recirculation aquaculture systems (RAS) due to its antimicrobial properties at low doses that, in turn, do not impair biofilter nitrification (Liu et al., 2017; Pedersen et al., 2009). The effectiveness of PAA, however, is impacted to an extent by certain water quality characteristics (Henao et al., 2018; Liu et al., 2014). Given the unique culture environment of RAS, wherein numerous substances such as nitrogen species, waterborne metabolites, nutrients, suspended solids, fine particles, heterotrophic bacterial counts and dissolved metals accumulate over time before reaching a steady state (Davidson et al., 2009; Martins et al., 2009), RAS producers need to consider their specific water chemistry characteristics when developing effective PAA treatment regimens (Pedersen et al., 2013). Previous in vitro studies have demonstrated the toxicity of PAA to a range of fish pathogens (Marchand et al., 2012; Meinelt et al., 2007, 2009, 2015; Muniesa et al., 2019; Straus & Meinelt, 2009); however, little research has been devoted to evaluating the bactericidal effects of PAA on fish pathogens in RAS water. Considering the increasing adoption of RAS technologies in the global aquaculture industry, we sought to characterize in vitro PAA efficacy against three important bacterial fish pathogens in water with representative chemistry profiles of salmonid RAS production. The three selected bacterial pathogens were Yersinia ruckeri, Weissella ceti and Flavobacterium columnare; these pathogens are the causative agents of the diseases enteric redmouth disease, weissellosis and columnaris respectively. Columnaris and enteric redmouth disease have long been identified as major diseases of fish (Davis, 1922; Ross et al., 1966), while weissellosis is currently recognized as an important emerging disease in salmonid aquaculture, with outbreaks reported in China (Liu et al., 2009), Brazil (Figueiredo et al., 2012), United States (Welch & Good, 2013), Mexico (Castrejón-Nájera et al., 2018), Japan (Mitomi et al., 2018), and Peru (Medina et al., 2020). The overall goal of our experiments was to determine the PAA concentration and exposure duration associated with complete pathogen killing, as determined by 6-log colony count reduction, in water from representative salmonid RAS. To carry out PAA experiments within RAS water, initial water samples were collected in sterile 1-litre glass bottles from the culture tank side drain of an on-site experimental-scale freshwater RAS, as described in previously published studies (e.g. Davidson et al., 2009; Good et al., 2009). Briefly, the system consisted of a fluidized sand biofilter, CO2 stripping column, low-head oxygenator, circular dual-drain culture tank, radial flow settler, microscreen drum filter (60 μm), heat exchanger and a 1-HP centrifugal pump. The total RAS water volume was 9.5 m3, and makeup water originated from a freshwater spring source. At the time of sampling, the RAS was operated at low exchange, was stocked with post-smolt Atlantic salmon Salmo salar reared at normal commercial densities and provided with a standard commercial feed at hourly intervals via automated feeders with feeding rates adjusted periodically based on observations related to satiation. Feed loading was approximately 800 L makeup water per kg of feed. RAS water quality characteristics are summarized in Table 1 (Experiment 1). This RAS provided suitable water for successful experiments involving Y. ruckeri and W. ceti; however, this same water hindered the assessment of F. columnare (see details below). Therefore, a second on-site RAS was used as a water source for F. columnare. This second RAS was identical in scale to the initial RAS, but was operated at a higher flushing rate (i.e. relatively more spring makeup water was being added); the RAS water quality profiles are summarized (under Experiment 2) in Table 1. All RAS water samples were kept cold on ice during the laboratory studies and were used within 2 h of collection. The bacterial strains used in these studies were provided by the USDA's National Center for Cool and Cold Water Aquaculture. Y. ruckeri (strain CSF007-82; serotype O1, Biotype 1) was grown at 28°C in Trypticase soy broth (TSB) with shaking at 200 RPM. W. ceti (strain NC36) was grown in DeMan-Rogosa-Sharpe broth (MRS) in static culture with no aeration. For post-disinfection plate counts, both Y. ruckeri and W. ceti were plated onto Tryptic soy agar (TSA) amended with 5% sheep blood (Remel). F. columnare (strain PSFC-81516-1) was grown in tryptone yeast extract salts (TYES) broth supplemented with 1 μg/ml tobramycin, and cultured post-disinfection on TYES agar. Prior to each experiment, each pathogen was grown overnight in their respective culture broths and assessed for optical density (OD) (Koch, 2007) the following day. The PAA used for all experiments was VigorOx® SP-15 (PeroxyChem), a commercial product labelled as containing 15% PAA and 10% H2O2. The product was evaluated upon arrival using a Hach digital titrator (Method LIT2199 – Hach, 2014) and was determined to be 17.67% PAA. A stock solution of 1:100 diluted PAA (1796 mg/L) was then created using sterile distilled water, to serve in all subsequent experiments. For each bacterial pathogen, cells were diluted from the corresponding overnight cultures into RAS water to reach approximately 1.0 × 106 cells per 100 μl (in 100 ml total RAS water) and lightly agitated occasionally to maintain even distribution within solution. Cell numbers were estimated by OD measurements and were confirmed later by plate counting. At this stage of the experiments, it was observed repeatedly that F. columnare cells immediately clumped together (i.e. resisted dispersing into suspension) when added to RAS water; therefore, F. columnare experiments were postponed due to the likelihood that subsequent bacterial colony forming unit (CFU) data would be unreliable. As such, Experiment 1 was carried out for only Y. ruckeri and W. ceti, and Experiment 2 was conducted at the next available opportunity focusing on F. columnare in water from a higher exchange rate RAS (no clumping was observed when F. columnare cells were added to water from this alternate RAS). Based on preliminary range-finding experiments (results not shown) to determine the relative sensitivity of each bacterial pathogen to PAA, the following concentrations and exposure times were selected: 0, 2, 5 or 10 mg/L PAA for 0, 2, 5 or 10 min exposure time for Y. ruckeri and W. ceti, and 0, 1, 2, 3 or 5 mg/L PAA for 0, 2, 5 or 10 min exposure time for F. columnare. Briefly, each selected concentration of PAA was applied to separate flasks of the prepared pathogen-RAS water suspension described above (with sterile phosphate buffered saline administered for the 0 mg/L PAA controls) in triplicate, with each flask gently inverted several times to ensure sufficient mixing. Subsequently, at the specified timepoints, 0.5 ml samples were removed and quickly neutralized by mixing with an equal volume of neutralizing solution (2.5 g sodium thiosulfate in 500 ml of sterile 1x phosphate-buffered saline [PBS]). Preliminary experiments were performed to validate the PAA neutralizing method used and to ensure that it did not affect cell viability. Note that zero timepoints were taken immediately before the addition of PAA and were handled identically. The neutralized samples were then assessed for viable cell numbers by serially diluting each sample in sterile PBS and plating 20 μl aliquots from each dilution on aforementioned nutrient agars. All microbiological work was carried out in a sterile biological safety cabinet. Plates were incubated at 30°C for 24 h, after which colonies were enumerated and CFU/20 μl were calculated and graphed for each PAA concentration and exposure duration, for each pathogen. Complete 6-log reduction to 0 CFU/20 μl was achieved for all three bacterial pathogens examined. In the first experiment using low-exchange RAS water, Y. ruckeri was completely eliminated following 5 min exposure to 5 mg/L PAA, while W. ceti was completely eliminated following 10 min exposure to 10 mg/L PAA (Figure 1). In the second experiment utilizing water from the relatively high-exchange RAS, F. columnare was completely eliminated with 3 mg/L PAA following 5 min exposure time (Figure 2). Our results provide RAS producers with useful information regarding the effectiveness of PAA to reduce populations of specific bacterial pathogens residing in "typical" RAS water and contribute to the growing scientific knowledge-base that will inform regulators when evaluating PAA's potential use in aquaculture settings. Our findings need to be considered in combination with other studies focusing on (i) PAA toxicity to specific aquaculture species (e.g. Straus et al., 2018), (ii) the safety of PAA for use in RAS and its potential effects on biofilter nitrification (e.g. Pedersen et al., 2009), (iii) the impact of differing water quality profiles on PAA efficacy and kinetics (e.g. Henao et al., 2018; Liu et al., 2014) and (iv) the toxicity of PAA to other important fish pathogens (e.g. Farmer et al., 2013; Straus & Meinelt, 2009) in order to develop comprehensive strategies to effectively combat waterborne pathogens in RAS. Similar to hydrogen peroxide (H2O2), which has been shown to be useful and biofilter-safe in commercial RAS (Pedersen & Pedersen, 2012), PAA has strong potential for application as a water disinfectant in operational aquaculture systems (Pedersen et al., 2009); however, PAA is a much more potent disinfectant compared with H2O2 (Block, 2001). Baldry (1983) compared the activity of H2O2 and PAA against bacteria, fungi, and spores and demonstrated similar bacteriostatic activity; however, it was postulated that repeated exposures to H2O2 might lead to resistance of pathogens due to increased catalase activity, whereas pathogen resistance would not occur with PAA as the chemical is not destroyed by catalase. In terms of pathogen killing, Baldry (1983) demonstrated superior bactericidal activity with PAA compared with H2O2. Similarly, Straus et al. (2018) reported PAA to be 100 times more potent as an anti-oomycete compound on channel catfish (Ictalurus punctatus) eggs than H2O2 based on previous research by Mitchell et al. (2009) and Straus et al. (2012). It is unknown what water quality characteristic(s) inhibited F. columnare from fully entering into suspension in the RAS water used during Experiment 1. Previous research has demonstrated a significant water quality effect on the pathogenicity F. columnare, with these authors considering divalent cation concentration (i.e. water hardness) being a major reason for reduction in virulence (Straus et al., 2015). In the present study, hardness likely did not contribute to the observed F. columnare clumping. Although hardness was not measured, alkalinity and hardness tend to correlate and, as such, the hardness in the second trial (in which F. columnare easily entered into suspension) was likely higher than in the first trial, based on alkalinity data and our historical alkalinity/hardness measurements. Among other water quality parameters measured, nitrate-nitrogen, nitrite-nitrogen and total ammonia nitrogen differed significantly between trials; however, little research has been carried out examining the effects of these parameters on F. columnare viability and pathogenicity. Farmer et al. (2011) determined that elevated total ammonia limited F. columnare infection in channel catfish but, as with hardness, total ammonia nitrogen was actually higher in our second, successful experiment. Beyond the water quality profile differences observed between experiments, other speculative chemical and/or biological causes for the apparent inactivation of F. columnare in low-exchange RAS water include the effects of accumulated humic substances (Meinelt et al., 2010) or F. columnare-specific bacteriophages (Laanto et al., 2011). Significant additional research is required to understand the observed effects of low-exchange RAS water on F. columnare viability. The higher resistance of W. ceti to PAA, relative to Y. ruckeri and F. columnare, may be related to W. ceti being a Gram-positive (vs. Gram-negative) bacteria. Gram-positive bacteria tend to have thicker cell walls, ranging from 20–80 nm in thickness (vs. 1.5–10 nm thick in Gram-negative bacteria), and these thick cell walls have been shown to assist Gram-positive bacteria in being resistant to disinfection (Mai-Prochnow et al., 2016), although this is not always the case across different approaches to disinfection (e.g. photoinactivation) (Heyong et al., 2016). The majority of important bacterial fish pathogens are Gram-negative, and as such are likely more susceptible to disinfectants such as PAA; however, further research is required to determine differences in susceptibility between additional important Gram-negative (e.g. Aeromonas salmonicida, Edwardsiella tarda, etc.) and Gram-positive (e.g. Streptococcus iniae, Renibacterium salmoninarum, etc.) pathogens of farmed fish. In conclusion, we provide PAA concentration and exposure duration data for 6-log reduction in three important bacterial fish pathogens while in RAS water. Further research should examine, among other things, additional fish pathogens and/or varying profiles of RAS water quality. Christopher Good: Conceptualization, Investigation, Resources, Data curation, Formal analysis, Writing – original draft, Visualization, Project administration, Funding acquisition. Natalie Redman & Megan Murray: Methodology, Investigation, Resources, Writing – review & editing. David L. Straus: Conceptualization, Writing – review & editing. Timothy J. Welch: Investigation, Methodology, Resources, Data curation, Formal analysis, Writing – review & editing, Visualization. Special thanks are extended to Jeremy Everson, Travis Moreland and Ryan Lipscomb for their assistance. Any use of trade, firm or product names is for descriptive purposes only and does not imply endorsement by The Conservation Fund Freshwater Institute and the U.S. Department of Agriculture. The Conservation Fund and the U.S. Department of Agriculture are equal opportunity employers and providers. This research was carried out by The Conservation Fund Freshwater Institute as an independent third party, with funding support from PeroxyChem LLC (Philadelphia, PA, USA). The authors do not have any conflicts of interest to disclose. As no research animals were directly used in these experiments, The Conservation Fund Freshwater Institute's Institutional Animal Care and Use Committee was not required to review and approve the experimental protocols described in this article. Data are available upon request to the corresponding author.
Read more
3