Nutrient enrichment increases virulence in an opportunistic environmental pathogen, with greater effect at low bacterial doses

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Nutrient enrichment increases virulence in an opportunistic environmental pathogen, with greater effect at low bacterial doses © 2024 the Authors Published version Pulkkinen, Katja; Taskinen, Jouni Pulkkinen, K., & Taskinen, J. (2024). Nutrient enrichment increases virulence in an opportunistic environmental pathogen, with greater effect at low bacterial doses. FEMS Microbiology Ecology, 100(4), Article fiae013. https://doi.org/10.1093/femsec/fiae013 2024 FEMS Microbiology Ecology , 2024, 100 , fiae013 DOI: 10.1093/femsec/fiae013 Ad v ance access publication date: 1 February 2024 Research Article Nutrient enrichment increases virulence in an opportunistic environmental pathogen, with greater effect at low bacterial doses Katja Pulkkinen * and Jouni Taskinen Department of Biological and Environmental Science, P.O. Box 35 (Survontie 9), University of Jyväsk ylä, Jyväsk ylä, Finland ∗Corresponding author. Department of Biological and Environmental Science, P.O. Box 35 (Survontie 9), University of Jyväsk ylä, Jyväsk ylä, Finland. Tel: + 358-40-8469359; E-mail: [email protected] Editor: [Ville-Petri Friman] Abstract Eutrophication of aquatic ecosystems is associated with an increased risk of pathogen infection via increased pathogen growth and host exposure via increased pathogen doses. Here , w e studied the effect of nutrients on the virulence of an opportunistic bacterial patho gen of fish, Flav obacterium columnar e , in c hallenge experiments with r ainbow trout fingerlings. We hypothesized that removing all nutrients by washing the bacteria would reduce virulence as compared to unwashed bacteria, but adding nutrients to the tank w ater would incr ease the virulence of the bacterium. Nutrient addition and increase in bacterial dose increased virulence for both unwashed and washed bacteria. For unwashed bacteria, the addition of nutrients reduced the survival probability of fish challenged with low bacterial doses more than for fish challenged with higher bacterial doses, suggesting acti v ation of bacterial virulence factors. Washing and centrifugation reduced viable bacterial counts, and the addition of washed bacteria alone did not lead to fish mortality. However, a small addition of nutrient medium, 0.05% of the total water volume, added separately to the fish container, restored the virulence of the washed bacteria. Our results show that human-induced eutrophication could trigger epidemics of aquatic pathogens at the limits of their survi v al and affect their ecology and e volution b y altering the dynamics between strains that differ in their growth c har acteristics. Ke yw ords: aquatic environment; bacterium; eutrophication; fish disease; Flavobacterium columnare ; nutrient Introduction Eutrophication of aquatic ecosystems is one of the most serious envir onmental pr oblems, leading to man y negativ e effects suc h as algal blooms, oxygen depletion, fish kills, and loss of biodiversity (Smith and Schindler 2009 ). One of the main concerns associated with anthropogenic nutrient enrichment is the increased risk of infection (McKenzie and Townsend 2007 , Johnson et al. 2010 , Aalto et al. 2015 ). Nutrient enrichment has been associated with increased infections in various aquatic organisms, such as fungal infections in corals (Voss and Richardson 2006 ), helminth infections in amphibians (Johnson et al. 2007 ), and myxozoan infections in fish (Bailey et al. 2018 , Lauringson et al. 2021 ). Many pathogens that cause disease epidemics in aquatic organisms can survive and replicate in the aquatic environment without the host (Arias et al. 2012 , Sundberg et al. 2014 , Ben Hamed et al. 2018 , Zhang et al. 2020 ). In natural waters without nutrient or organic pollution, the nutrient le v els ar e usuall y low (Wetzel 2001 ), and numbers of microbial pathogens are so low that successful host invasions are unlikely. Many aquatic microbes are adapted to persist in low-nutrient water while retaining the ability to r epr oduce under nutrient r eplete conditions (Elser et al. 1995 , Cotner et al. 2010 ). The eutrophication of aquatic ecosystems or the accumulation of organic nutrients in the water can create conditions that favor disease outbreaks in aquatic organisms (Smith and Schindler 2009 , Kinnula et al. 2017 ). Increased nutrients can promote pathogen growth and consequently increase the risk of infection by increasing the infectious pathogen dose encountered by the host (Smith and Schindler 2009 , Wedekind et al. 2010 , Kinnula et al. 2017 ). Incr eased or ganic and inor ganic nutrient enrichment of the environment outside the host may affect the gr owth and c har acteristics of envir onmentall y gr owing opportunistic pathogens, including their virulence (Brown et al. 2012 ). Furthermor e, if the r esources av ailable in the outside-host envir onment ar e nutritionall y close to the host tissues, the pathogen could become pr eada pted for host exploitation (Brown et al. 2012 , K etola et al. 2016 ), possibl y via activ ation of virulence factors (Penttinen et al. 2016 , Kinnula et al. 2017 ). Understanding the effects of nutrient enrichment on the dynamics and evolution of aquatic pathogens, both in natural waters and in aquaculture, is important for predicting and mitigating their impact on aquatic ecosystems and food production. Flavobacterium columnare is an aquatic pathogen that infects a wide range of freshwater fish, including important aquaculture species such as channel catfish and salmonids (Declercq et al. 2013 ). The pathogen attacks the gills, skin, and fins of fish, causing tissue damage and necrosis, that is ultimately fatal to the host. Infection with highly virulent strains can kill the host before visible signs of disease appear. The pathogen can m ultipl y in liv e and dead host fish, and other organic decaying material in the Recei v ed 25 August 2023; revised 29 J an uar y 2024; accepted 31 J an uar y 2024 ©The Author(s) 2024. Published by Oxford Uni v ersity Pr ess on behalf of FEMS. This is an Open Access article distributed under the terms of the Cr eati v e Commons Attribution License ( https://cr eati v ecommons.org/licenses/by/4.0/ ), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Downloaded from https://academic.oup.com/femsec/article/100/4/fiae013/7596547 by Jyvaskyla University user on 17 April 2024 2 | FEMS Microbiology Ecology , 2024, Vol. 100, No. 4 water (Kunttu et al. 2009 , 2012 ). Flavobacterium columnare can be isolated from lake water (Kunttu et al. 2012 ), and has been shown to survive in lake water and distilled water for long periods in the laboratory (Arias et al. 2012 , Sundberg et al. 2014 ). In the natur al envir onment, the bacterium is likel y to encounter mor e oligotrophic conditions than in the aquaculture facilities, where water can contain high concentrations of nutrients from fish feed and excreta (Naylor et al. 1999 ). Increased nutrient concentrations in the growth medium (Penttinen et al. 2016 ) or in tank water (Kinnula et al. 2017 ) have been shown to increase the virulence of F. columnare in challenge tests with rainbow trout ( Oncorhynchus m ykiss ) and zebr afish ( Danio rerio ). Or ganic nutrients in the water of fish farms could ther efor e pr omote disease epidemics caused by F. columnare on farms. Here, we studied the effect of nutrients in the environment outside the host on the virulence of F. columnare in fish challenge experiments using five different doses of bacteria. We compared the virulence of fresh, unwashed bacteria with bacteria that had been centrifuged, washed, and diluted in distilled water to r emov e growth media. We then examined how the virulence of these bacteria responded to a small addition of growth media (0.05%), ad ded se par atel y to the fish tank. We hypothesized that removing all nutrients by washing would reduce virulence but adding nutrients to the tank w ater w ould increase virulence of the bacterium to rainbow trout fingerlings. We show that organic nutrients in the water increase virulence, and especially so at low bacterial doses, and in washed and possibly damaged bacterial cells. Our results suggest that environmental nutrient enrichment may be more important than previously thought in maintaining and initiating disease epidemics in aquatic organisms. Materials and methods Flavobacterium columnare strain B067 stored frozen in −80 ◦C with 10% fetal calf serum and 10% gl ycer ol was r esuspended in modified Shieh medium (Song et al. 1988 ), further modified by excluding yeast extract and replacing peptone with casamino acids, and cultur ed ov ernight at r oom temper atur e . T he o v ernight cultur e was subcultured for a further 18 h by adding 2 ml of the overnight culture into 20 ml of fresh medium. After combining all subcultur es, bacterial cultur e was divided into fiv e 50 ml tubes. Four of the tubes were centrifuged (3500 g , 4 ◦C, and 12 min), the liquid supernatant was r emov ed, and the pellets wer e diluted bac k to 50 ml of distilled sterile water. This pr ocedur e was repeated twice, after which the pellets from all four tubes were combined and diluted in 21 ml of distilled sterile water. One remaining unwashed 50 ml tube of the bacterial culture was used as such in the fish challenge experiments (see below). Both the washed and unwashed bacterial solutions were diluted at 1:2, 1:5, 1:10, and 1:20 with distilled sterile water. Including the undiluted cultur es, ther e wer e 10 separate bacterial treatments in total to be used in fish challenges. Viable counts of undiluted solution and solutions diluted to 1:2 and 1:20 were then determined from plate counts of serial dilutions (10 0 –10 −7 ) on Shieh agar plates . T he viable counts for the 1:5 and 1:10 dilutions were calculated based on counts from the mor e concentr ated cultur es assuming that the change in bacterial numbers was proportional to the dilution of the liquid. The fish challenge experiments tested the effect of washing the bacteria, bacterial dose, and n utrient ad dition on bacterial virulence. Rainbow tr out ( O. m ykiss ) fingerlings wer e obtained fr om a fish farm using groundwater and were therefore disease-free. The fish were maintained in the laboratory in aerated well water with a 12:12 L:D cycle at 17 ◦C and fed with commercial pellets. Prior to the experiment, the water temper atur e in the fish holding tanks was gr aduall y incr eased fr om 16.4 to 25.5 ◦C ov er 8 days to acclimate the fish to higher temper atur e. Higher temper atur e was used to ensure rapid onset of the disease to meet the humane endpoints of the experiment. Rainbow trout fingerlings (0.8 ±0.3 g) wer e individuall y tr ansferr ed to 500 ml plastic containers filled with borehole water (25.5 ◦C) aerated immediately before the experiment. Fish wer e c hallenged using the continuous exposur e method (Kinnula et al. 2015 ). Briefly, 500 μl of each bacterial dose (1:0, 1:2, 1:5, 1:10, and 1:20) of either washed or unwashed bacteria was added to the containers, for a total of 10 treatments. A total of 10 additional treatments (each of the five doses for both washed and unwashed bacteria) r eceiv ed an addition of 500 μl of fresh Shieh medium. Specifically, the bacteria were not mixed with the nutrient medium prior to addition to the fish containers, but the medium was added separ atel y after the addition of the bacteria on the opposite side of the container. There was no aeration in the containers, so bacteria and nutrients were mixed to the water only due to fish mo vements . After exposure , the fish containers were placed in random order on the shelves in a temperaturecontr olled r oom at 25.5 ◦C. Eac h bacterial dose x nutrient treatment (without or with nutrient addition) combination was replicated ten times, except for the treatment with unwashed bacteria and nutrient addition with the lo w est dose, which had only nine fish, resulting in 199 challenged fish. In addition, 30 fish received a 500 μl addition of unwashed Shieh medium, and 20 fish r eceiv ed a 500 μl addition of sterile water. A total of 249 fish were used in the experiment. Fish were monitored hourly for signs of disease and those showing signs of disease (moribund, unbalanced swimming, and unr esponsiv e to external stim uli) wer e euthanized with Tricaine mesylate solution (MS-222, Sigma) and killed by decapitation with scissors. Bacterial samples from fish gills were plated on Shieh agar plates containing tobramycin (Decostere et al. 1997 ) to verify infection with F. columnare . Fish w eights w er e r ecorded . The experiment was terminated after 30 h, after which all surviving fish were killed with an overdose of MS-222. The experiment w as conducted accor ding to the Finnish Act on the Use of Animals for Experimental Purposes . T he license was granted by the National Animal Experiment Board at the Regional State Administr ativ e Agency for Southern Finland (permission number ESAVI/10184/04.10.07/2014). Da ta anal yses For fish challenges with the unwashed bacteria, the fish survival was analyzed with Cox survival analysis, with nutrient addition as categorical, and bacterial dose (CFU ml −1 , log-transformed, with mean standardized to zero) and fish weight as continuous variables. Model selection was based on the Akaike information criteria starting from the full model including a three-way interaction and all 2-way interactions. The proportional hazards assumption was c hec ked with Sc hoenfeld r esiduals test ( P > .05 for all tested factors) (In and Lee 2019 ). For fish challenges with the washed bacteria, the lack of mortality in treatment without added nutrients led to problems with model convergence in cox-regression, hence Ka plan–Meier log-r ank test was used for comparing the two nutrient le v els. Statistical anal yses wer e performed using R-studio version 2022.12.0.353 (RStudio Team 2020 ), with pac ka ge “surviv al” (Therneau 2023 ). Due to the statistical limitations in performing survival analyses, mean fish longevity was further analyzed. For each bacterial dose and nutrient treatment, one-tailed one-sample t -tests Downloaded from https://academic.oup.com/femsec/article/100/4/fiae013/7596547 by Jyvaskyla University user on 17 April 2024 Pulkkinen and Taskinen | 3 Figure 1. Ka plan–Meier surviv al pr obabilities for r ainbow tr out ( O. m ykiss ) fingerlings c hallenged with F. columnare . Fish wer e c hallenged with fiv e different doses (CFU ml −1 ) of (A) unwashed bacteria without nutrients, (B) unwashed bacteria with nutrient addition, (C) washed bacteria without nutrients, and (D) washed bacteria with nutrient addition. Note that the doses differ between unwashed and washed bacteria. were used to test whether the mean survival time was less than the maxim um surviv al time in the experiment (30 h). Bonferronicorr ection was a pplied to adjust the probability of increased risk of type I error associated to multiple testing. In addition, the percentage of survived fish was calculated for each bacterial dose and nutrient treatment, but as most of these were either 0% or 100%, they were not statistically analyzed, but are shown for comparison. Results Based on plate counts of unwashed and washed bacteria, washing resulted in ∼1000 times lo w er bacterial doses administered to fish. The doses at dilutions 1:0, 1:2, 1:5, 1:10, and 1:20 for the unw ashed bacteria w ere 3.5 ×10 4 , 1.8 ×10 4 , 7.1 ×10 3 , 3.6 ×10 3 , and 1.8 ×10 3 CFU ml −1 , while those for the washed bacteria were 4.2 ×10 1 , 2.1 ×10 1 , 8 ×10 0 , 4 ×10 0 , and 2 ×10 0 CFU ml −1 . Due to this large difference in CFU ranges for the washed and unwashed bacteria, the data wer e anal yzed separ atel y for fish c hallenged with unwashed and washed bacteria. Of the 20 control fish with the addition of 500 μl of sterile water, one died at time point 19 h, and of the 30 control fish with the addition of 500 μl of fresh Shieh growth medium, one died at time point 27 h. Howe v er, F. columnare was not r ecov er ed fr om the bacterial gill samples of these two fish. The best model explaining fish survival in challenge experiment with unwashed bacteria included nutrient addition, bacterial dose, and their interaction. Fish weight had no effect on surviv al pr obability. Nutrient addition decr eased fish surviv al pr obability (Exp(b):604.2; 95% CI: 78–4657; Wald: 6.146, P < .001). Surviv al pr obability also decr eased with incr ease in bacterial dose (CFU) (Fig. 1A and B; Exp(b):93.4 for log-transformed values; 95% CI: 19–466; Wald: 5.530, P < .001). The interaction between nutrient addition and bacterial dose was statistically significant, indicating that when nutrients were added, bacterial dose had less effect on the surviv al pr obability (Fig. 1A and B; Exp(b):0.06; 95% CI: 0.01– 0.27; Wald: −3.587, P < .001), i.e. also the fish challenged with low bacterial doses died. The addition of nutrients also reduced the survival probability of fish challenged with washed bacteria, (Fig. 1 C and D; log rank χ2 = 30.5, df = 1, P < .001). When nutrients were not added with the washed bacteria, bacterial dose had no effect on fish survival (Fig. 1 C; log rank χ2 = 4, df = 4, P = .4), but when nutrients were ad ded, survi val probability decreased with increasing bacterial dose (Fig. 1 D; log rank χ2 = 65.4, df = 4, P < .001). For fish challenged with unwashed bacteria the mean survival time was not different from 30 h (maximum survival to the end of the experiment) for the three lo w est bacterial doses (Fig. 2 A; onesample t -tests, one-sided P > .05), i.e. all fish survived in these treatments (Fig. 2 C). At the two highest bacterial doses, the mean survival time was statistically significantly less than 30 h (Fig. 2 A; dose 3.5 ×10 4 CFU ml −1 , t = −15.5, df = 9, P < .001 and dose 1.8 ×10 4 CFU ml −1 , t = −4.4, df = 9, P < .01), and fish survival was less than 10% (Fig. 2 C). When nutrients were added with unwashed bacteria, the mean survival time was statistically significantly less than 30 h for all bacterial doses (Fig. 2 A; dose 3.5 ×10 4 CFU ml −1 , t = −23.7, df = 9, P < .001; dose 1.8 ×10 4 CFU ml −1 , t = Downloaded from https://academic.oup.com/femsec/article/100/4/fiae013/7596547 by Jyvaskyla University user on 17 April 2024 4 | FEMS Microbiology Ecology , 2024, Vol. 100, No. 4 Figure 2. Survival times of individual fish and mean survival time ±SE (h) and survival percentage (%) of rainbow trout ( O. mykiss ) fingerlings challenged with F. columnare . Fish were challenged with five different doses (CFU ml −1 ) of (A), (C) unwashed F. columnare and (B), (D) washed F. columnare , both without (open symbols) or with nutrient addition (filled symbols). Note, that the x -axis scales (doses) are different for the unwashed (left panel) and the washed bacteria (right panel). −12.9, df = 9, P < .001; dose 7.1 ×10 3 CFU ml −1 t = −33, df = 9, P < .001; 3.6 ×10 3 CFU ml −1 ; t = −14.6, df = 9, P < .001; 1.8 ×10 3 CFU ml −1 ; t = −5.5, df = 8, P < .01). None of the fish challenged with unwashed bacteria and added nutrients survived (Fig. 2 C). Nutrient addition, thus decreased fish survival relatively more at the three lo w est bacterial doses than at the two highest doses, as at the three lo w est bacterial doses no mortality occurred without ad ded n utrients. The mean survival time of fish challenged with washed bacteria did not differ from 30 h (maximum survival time to the end of the experiment) for any of the five bacterial doses given (Fig 2 B; one-sample t -tests, one-sided P > .05). Mortality occurr ed onl y at the highest dose (Fig. 2 D). For fish challenged with washed bacteria and n utrient ad dition, the mean survival time was not differ ent fr om 30 h for the three lo w est bacterial doses (one-sample t -tests, one-sided P > .05), but was statistically significantly lo w er than 30 h for the two highest doses (dose 4.2 ×10 1 CFU ml −1 , t = −10.2, df = 9, P < .001 and dose 2.1 ×10 1 CFU ml −1 , t = −30, df = 9, P < .001). When nutrients were added to the washed bacteria, mortality occurred at all but the lo w est bacterial dose (Fig. 2 D). Discussion Here, w e sho w ed that a small addition of nutrient medium (0.05% of water volume), added separ atel y to the fish container, r estor ed the virulence to washed F. columnare bacteria, whereas without n utrient ad dition the washed bacteria were unable to infect fish. In addition, when challenged with unwashed bacteria, the addition of nutrients reduced fish survival probability and survival time r elativ el y mor e at the lo w est doses of F. columnare than at the higher bacterial doses. In natural, unpolluted water environments, bacterial abundances are generally low (Wetzel 2001 ) and pathogens r epr esent onl y a fr action of the bacteria. Hence, the doses of microbial pathogens encountered by the hosts in natural w aters are lo w as compared to the pathogen doses r equir ed for confirmed infections under experimental conditions (Kinnula et al. 2015 , Sundberg et al. 2016 ). The curr ent r esults suggest that Downloaded from https://academic.oup.com/femsec/article/100/4/fiae013/7596547 by Jyvaskyla University user on 17 April 2024 Pulkkinen and Taskinen | 5 when pathogenic bacteria are in poor condition or when their numbers are low in natural unpolluted water environment, presence of organic nutrients in water may have a critical importance in restoring the virulence of the pathogenic bacteria. The results support pr e vious findings on the effect of human-induced envir onmental nutrient enric hment in maintaining and triggering disease epidemics in aquatic organisms (Wedekind et al. 2010 , Kinnula et al. 2017 , Pulkkinen et al. 2018 , Zhang et al. 2020 ). Our results from the challenges with the different doses of unwashed bacteria confirmed the findings of pr e vious studies on the dose-dependent infection probability of F. columnare (Kinnula et al. 2015 , 2017 ). Pr e vious studies hav e shown that addition of extra nutrients in the growth medium or aquarium water increases the virulence of aquatic bacterial pathogens (Wedekind et al. 2010 , Penttinen et al. 2016 , Kinnula et al. 2017 ). Increased nutrient availability is likely to increase bacterial replication and hence the pathogen dose contacted by the host, thereby increasing the likelihood of infection. A pr e vious study demonstr ating incr eased virulence associated with n utrient ad dition used 0.64%–1.28% nutrient addition with fresh, unwashed cultures that may also contain nutrients from the original growth medium (Kinnula et al. 2015 ). In our experiment, the nutrient addition (0.05%) was an order of magnitude lo w er, and it w as ad ded se par atel y to the other side of the fish container, so that the added bacteria were not in direct contact with the nutrient broth when dispersing into the water in the container solely due to fish mo vements . T herefore , it is unlikely that the increased virulence due to nutrient addition in our experiment was ac hie v ed solel y via incr eased gr owth of the bacteria. In current experiments, fish survival was relatively lo w er at the low doses than at the high doses of the unwashed bacteria when nutrients were added. The study by Kinnula et al. ( 2017 ) did not find an interaction between bacterial dose and nutrients but did find increased virulence in a low virulent strain in the highest nutrient treatment. Our results support the suggestion that increased virulence in the presence of outside-host nutrients is ac hie v ed in part thr ough activ ation of virulence factors (Kinnula et al. 2015 , Penttinen et al. 2016 ). In F. columnare , increased nutrient le v els hav e been shown to activ ate the putativ e virulence factors c hondr oitinase (cslA) and collagenase (Penttinen et al. 2016 ). Based on plate counts, there was an extensive loss of cell viability after washing, centrifugation, and subsequent dilution in distilled water in our experiment. Washing and subsequent centrifugation are standard laboratory procedures used to remove external contaminants from cells under investigation (Peterson et al. 2012 ). High-speed centrifugation (15 000 ×g ) is known to dama ge Gr am-negativ e bacterial cells due to the shear forces caused by cell collisions (Pembrey et al. 1999 , Peterson et al. 2012 ). On the other hand, the centrifugation speed in our experiment was below the le v el at whic h significant cell dama ge has been found (5000 ×g ) and which is used as a reference method in experiments (Pembrey et al. 1999 ). It is possible that the long rod-like cell morphology (0.3–0.5 μm wide ×3–10 μm long) of F. columnare cells makes them mor e sensitiv e to damage during centrifugation than the coccoid cells or short rods previously tested (Pembrey et al. 1999 , Declercq et al. 2013 ). Dilution of bacterial cells to distilled water alone may not have been detrimental to F. columnare cells, as F. columnare can remain viable in distilled water for months (Arias et al. 2012 , Sundberg et al. 2014 ). Long-term survival in distilled water has been suggested to be based on the recycling of nutrients from dead and decaying cells (Kunttu et al. 2009 ), but cell division of starved cells under the microscope has not been demonstrated (Arias et al. 2012 ). Flavobacterium columnare cells appear to remain culturable e v en under pr olonged starv ation (Arias et al. 2012 ), unlike some other aquatic pathogenic bacteria such as Aeromonas hydrophila (Rahman et al. 2001 ), F. psychrophilum (Madetoja et al. 2003 ), or Francisella noatunensis (Duodu and Colquhoun 2010 ), which enter a nonculturable but viable state. Instead, F. columnare cells undergo a change in cell morphology from the long rod shape to a coiled resting form (Arias et al. 2012 ). Diversification of colony morphology detected during pr olonged starv ation also suggests that cell mor phology under goes c hanges during starv ation (Sundber g et al. 2014 ). The loss of virulence in fish challenge experiments after washing and centrifugation was pr obabl y due to damage to cell surface molecules, whic h ar e important virulence factors for F. columnare (Kunttu et al. 2021 ). The importance of surface molecules for virulence in F. columnare has been demonstrated, for example, in experimental infections of F. columnare cells with pha ges, whic h led to a change in colony morphotype with a concomitant loss of virulence (Laanto et al. 2012 , Kunttu et al. 2021 ). Surface molecules affect bacterial properties such as adhesion, colony morphotype, and biofilm formation, all of which may contribute to fish epithelial colonization and hence bacterial virulence (Decostere et al. 1997 , Kunttu et al. 2011 , Oliv ar es-Fuster et al. 2011 , Lange et al. 2017 ). In addition to surface molecules, bacterial virulence is affected by proteins and polymeric substances (EPS) that bacteria excrete outside the cells (Newton et al. 1997 , Laanto et al. 2014 , Lange et al. 2017 , Zhang et al. 2017 , Mekasha and Linke 2021 ). The EPSs ar e excr eted via specialized secr etion systems, of which T9SS has been associated with se v er al virulence factors in F. columnare (Kunttu et al. 2021 , Mekasha and Linke 2021 ), such as c hondr oitin AC l yase and metallopr oteases that degr ade connectiv e and m uscle tissue (Newton et al. 1997 ). Centrifugation and washing may hav e r emov ed these substances or altered their expression, and hence the virulence properties of the bacterial cells. Inter estingl y, the pr esence of fish and the potential nutrients in fish excreta in the water alone were not sufficient to induce infection when the washed F. columnare cells were added to fish containers without added nutrients. Fish mucus can support the growth of F. columnare (Shoemaker et al. 2018 ) and F. columnare biofilms (Lange et al. 2020 ) and increase the expression of some extracellular pr oteins (Star oscik and Nelson 2008 , Almeida et al. 2019 ), all of which could aid in the infection of fish hosts. In addition, the nutrients in fish excreta could support bacterial growth. According to the plate count, the doses of the unwashed bacteria wer e thr ee orders of magnitude higher than doses of the washed bacteria, but when nutrients were added, the survival times of the fish challenged with the two highest doses (4 ×10 1 and 2 ×10 1 CFU ml −1 ) of the washed bacteria were not m uc h lo w er (20–25 h) than that of the unwashed bacteria (15–25 h), and none of the fish survived. The plate count may have underestimated the abundance of the washed cells because there was a longer interval between preparation of the washed solution and plating than betw een w ashing and challenge of the fish. Another possibility is that washed bacteria formed a ggr egates, r esulting in higher doses added to fish containers than detected by plate counting. Aggregation of F. columnare cells has been detected after Ca 2 Cl 2 addition to TYES medium (Evenhuis et al. 2021 ) or after mucin addition to Shieh medium (Almeida et al. 2019 ), as well as in our experience after dilution of the Shieh nutrient medium. Ho w e v er, the dilution series of washed bacteria with n utrient ad dition also a ppear ed to function in a dose-dependent manner, based on reduction in mean host longevity and proportion of fish surviving with the increase in bacterial dose. Downloaded from https://academic.oup.com/femsec/article/100/4/fiae013/7596547 by Jyvaskyla University user on 17 April 2024 6 | FEMS Microbiology Ecology , 2024, Vol. 100, No. 4 In conclusion, our results suggest that nutrients in the outside host environment of the bacteria may play a crucial role in maintaining and initiating disease epidemics. In F. columnare , there is a positive association with growth and virulence (Pulkkinen et al. 2010 , 2018 , Kinnula et al. 2017 ), as well as a dose-dependent increase in host infection risk (Kinnula et al. 2017 ), suggesting that the virulent strains generally win in the competition for host inv asion. Her e , we ha v e shown that or ganic nutrients can r estor e virulence to bacterial cells that are on the v er ge of losing viability in water, and that nutrients increase virulence r elativ el y mor e at the low bacterial doses than at the high doses. In the natural, less eutrophic waters, some pathogenic bacteria may persist in low numbers at the limits of their viability. Anthropogenic nutrient enrichment of the environment could promote the growth of these slo w-gro wing pathogens to r eac h the doses r equir ed for host inv asion and incr ease their contribution to epidemics, thus changing the dynamics between pathogen strains. Nutrient enrichment may ther efor e hav e a greater impact on the ecology and evolution of pathogens in aquatic organisms than pr e viousl y thought, both in natural waters and in aquaculture. Author contributions Katja Pulkkinen (Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Validation, Visualization, Writing –original draft, Writing –review & editing), and Jouni Taskinen (Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Writing –review & editing). Ac kno wledgments The authors want to thank Lotta-Riina Sundberg for donating the bacterial strain used in the experiment and Marjo Aaltonen, Sari Aaltonen and Nina Honkanen for assistance in the laboratory. Conflict of interest : None declared. Funding This w ork w as supported b y the Academy of Finland (project number 260704 to J.T.). References Aalto SL , Decaestecker E, Pulkkinen K. A three-way perspective of stoic hiometric c hanges on host-par asite inter actions. Trends Parasitol 2015; 31 :333–40. https:// doi.org/ 10.1016/ j.pt.2015.04.005 . Almeida GMF , Laanto E, Ashr afi R et al. Bacteriopha ge adher ence to mucus mediates preventive protection against pathogenic bacteria. mBio 2019; 10 :1–12. https:// doi.org/ 10.1128/ mBio.01984-19 . Arias CR , Lafrentz S, Cai W et al. Ada ptiv e r esponse to starv ation in the fish pathogen Flavobacterium columnare : cell viability and ultr astructur al c hanges. BMC Microbiol 2012; 12 :266. https://doi.org/ 10.1186/1471218012266 . Bailey C , Rubin A, Str eppar av a N et al. Do fish get wasted? Assessing the influence of effluents on parasitic infection of wild fish. PeerJ 2018; 6 :e5956. https:// doi.org/ 10.7717/ peerj.5956 . Ben Hamed S , Tav ar es Ranzani-P aiv a MJ, Tac hibana L et al. Fish pathogen bacteria: adhesion, parameters influencing virulence and interaction with host cells. Fish Shellfish Immunol 2018; 80 :550– 62. https:// doi.org/ 10.1016/ j.fsi.2018.06.053 . Brown SP , Cornforth DM, Mideo N. Evolution of virulence in opportunistic pathogens: generalism, plasticity, and control. Trends Microbiol 2012; 20 :336–42. https:// doi.org/ 10.1016/ j.tim.2012.04.005 . Cotner JB , Hall EK, Scott JT et al. F reshw ater bacteria are stoichiometrically flexible with a nutrient composition similar to seston. Front Microbio 2010; 1 :132. https:// doi.org/ https:// doi.org/ 10.3389/ fmicb .2010.00132 . Declercq AM , Haesebr ouc k F, Van den Br oec k W et al. Columnaris disease in fish: a r e vie w with emphasis on bacterium-host interactions. Vet Res 2013; 44 :27. https:// doi.org/ 10.1186/ 1297971644 -27 . Decoster e A , Haesebr ouc k F, De vriese LA. Shieh medium supplemented with tobramycin for selective isolation of Flavobacterium columnare ( Flexibacter columnaris ) from diseased fish. J Clin Microbiol 1997; 35 :322–4. Duodu S , Colquhoun D. Monitoring the survival of fish-pathogenic F rancisella in w ater microcosms. FEMS Microbiol Ecol 2010; 74 :534– 41. https:// doi.org/ 10.1111/ j.1574-6941.2010.00973.x . Elser JJ , Stabler LB, Hassett RP. Nutrient limitation of bacterial growth and rates of bacterivory in lakes and oceans: a compar ativ e study. Aquat Microb Ecol 1995; 9 :105–10. https:// doi.org/ 10.3354/ ame009 105 . Evenhuis JP , Lipscomb R, Birkett C. Growth of Flavobacterium columnare genomovars in the presence or lack of supplemental cations. Curr Microbiol 2021; 78 :2474–80. https:// doi.org/ 10.1007/ s0 0284021025078 . In J , Lee DK. Survival analysis: part II—applied clinical data analysis. Kor J Anesthesiol 2019; 72 :441–57. https:// doi.org/ 10.4097/ kja.1918 3 . Johnson PT , Chase JM, Dosch KL et al. Aquatic eutrophication promotes pathogenic infection in amphibians. Proc Natl Acad Sci USA 2007; 104 :15781–6. https:// doi.org/ 10.1073/ pnas.0707763104 . Johnson PT , Townsend AR, Cle v eland CC et al. Linking environmental nutrient enrichment and disease emergence in humans and wildlife. Ecol Appl 2010; 20 :16–29. https:// doi.org/ 10.1890/ 08-0633. 1 . K etola T , Mik onr anta L, Laakso J et al. Differ ent food sources elicit fast changes to bacterial virulence. Biol Lett 2016; 12 :20150660. https: // doi.org/ 10.1098/ rsbl.2015.0660 . Kinnula H , Mappes J, Valkonen JK et al. Higher r esource le v el pr omotes virulence in an envir onmentall y tr ansmitted bacterial fish pathogen. Evol Appl 2017; 10 :462–70. https:// doi.org/ 10.1111/ eva. 12466 . Kinnula H , Mappes J, Valkonen JK et al. The influence of infective dose on the virulence of a generalist pathogen in rainbow tr out ( Oncorhync hus m ykiss ) and zebr a fish ( Danio rerio ). PLoS ONE 2015; 10 :e0139378. https:// doi.org/ 10.1371/ journal.pone.0139378 . Kunttu, Jokinen EI , Valtonen ET, Sundberg LR. Virulent and nonvirulent Flavobacterium columnare colony morphologies: characterization of c hondr oitin AC l yase activity and adhesion to pol ystyr ene. J Appl Microbiol 2011; 111 :1319–26. https:// doi.org/ 10.1111/ j.1365-2 672.2011.05149.x . Kunttu, Runtuvuori-Salmela A , Sundell K, Wiklund T et al. Bacteriopha ge r esistance affects Flavobacterium columnare virulence partl y via mutations in genes related to gliding motility and the type IX secr etion system. A ppl Environ Microb 2021; 87 :e0081221. https: // doi.org/ 10.1128/ AEM.00812-21 . Kunttu, Sundberg LR , Pulkkinen K, Valtonen ET. Environment may be the source of Flavobacterium columnare outbreaks at fish farms. Environ Microbiol Rep 2012; 4 :398–402. https:// doi.org/ 10.1111/ j.17 58-2229.2012.00342.x . Kunttu, Valtonen ET , Jokinen EI, Suomalainen LR. Sa pr ophytism of a fish pathogen as a transmission strategy. Epidemics 2009; 1 :96–100. Downloaded from https://academic.oup.com/femsec/article/100/4/fiae013/7596547 by Jyvaskyla University user on 17 April 2024 Pulkkinen and Taskinen | 7 Laanto E , Bamford JK, Laakso J et al. Pha ge-driv en loss of virulence in a fish pathogenic bacterium. PLoS ONE 2012; 7 :e53157. https: // doi.org/ 10.1371/ journal.pone.0053157 . Laanto E , Penttinen RK, Bamford JK et al. Comparing the different morphotypes of a fish pathogen–implications for k e y virulence factors in Flavobacterium columnare . BMC Microbiol 2014; 14 :170. ht tps:// doi.org/ 10.1186/ 1471218014170 . Lange MD , Farmer BD, Abernathy J. Vertebrate mucus stimulates biofilm de v elopment and upr egulates ir on acquisition genes in Flavobacterium columnare . J Fish Dis 2020; 43 :101–10. https://doi.or g/ 10.1111/ jfd.13103 . Lange MD , Farmer BD, Declercq AM et al. Sic keningl y sweet: lrhamnose stimulates Flavobacterium columnare biofilm formation and virulence. J Fish Dis 2017; 40 :1613–24. https:// doi.org/ 10.1111/ jfd.12629 . Lauringson M , Nousiainen I, Kahar S et al. Climate c hange-driv en disease in sympatric hosts: temporal dynamics of parasite burden and pr olifer ativ e kidney disease in wild br own tr out and Atlantic salmon. J Fish Dis 2021; 44 :689–99. https:// doi.org/ 10.1111/ jfd.13 330 . McKenzie VJ , Townsend AR. Parasitic and infectious disease responses to changing global nutrient cycles. EcoHealth 2007; 4 :384– 96. https:// doi.org/ 10.1007/ s1039300701313 . Madetoja J , Nystedt S, Wiklund T. Survival and virulence of Flavobacterium psychrophilum in water microcosms. FEMS Microbiol Ecol 2003; 43 :217–23. https:// doi.org/ 10.1111/ j.1574-6941.2003.tb0 1061.x . Mekasha S , Linke D. Secretion systems in gram-negative bacterial fish pathogens. Front Microbiol 2021; 12 :782673. https:// doi.org/ 10 .3389/fmicb .2021.782673 . Naylor SJ , Moccia RD, Durant GM. The chemical composition of settleable solid fish waste (manure) from commercial rainbow trout farms in Ontario, Canada. North Am J Aquacult 1999; 61 :21– 6. https:// doi.org/ 10.1577/ 1548-8454(1999)061  0021:Tccoss  2.0.C o;2 . Newton JC , Wood TM, Hartley MM. Isolation and partial c har acterization of extracellular proteases produced by isolates of Flavobacterium columnare derived from channel catfish. J Aqua Anim Health 1997; 9 :75–85. https:// doi.org/ 10.1577/ 1548-8667(19 97)009  0075:Iapcoe  2.3.Co;2 . Oliv ar es-Fuster O , Bullard SA, McElwain A et al. Adhesion dynamics of Flavobacterium columnare to channel catfish Ictalurus punctatus and zebrafish Danio rerio after immersion challenge. Dis Aquat Org 2011; 96 :221–7. https:// doi.org/ 10.3354/ dao02371 . Pembrey RS , Marshall KC, Schneider RP. Cell surface analysis techniques: what do cell pr epar ation pr otocols do to Cell surface pr operties?. A ppl Environ Microb 1999; 65 :2877–94. https://doi.org/ 10.1128/aem.65.7.2877-2894.1999 . Penttinen R , Kinnula H, Lipponen A et al. High nutrient concentration can induce virulence factor expression and cause higher virulence in an envir onmentall y tr ansmitted pathogen. Microb Ecol 2016; 72 :955–64. https:// doi.org/ 10.1007/ s0024801607811 . Peterson BW , Sharma PK, van der Mei HC et al. Bacterial cell surface damage due to centrifugal compaction. Appl Environ Microb 2012; 78 :120–5. https:// doi.org/ 10.1128/ AEM.06780-11 . Pulkkinen K , Pekkala N, Ashrafi R et al. Effect of resource availability on evolution of virulence and competition in an environmentally transmitted pathogen. FEMS Microbiol Ecol 2018; 94 :94. https://doi. org/ 10.1093/ femsec/fiy060 . Pulkkinen K , Suomalainen LR, Read AF et al. Intensive fish farming and the evolution of pathogen virulence: the case of columnaris disease in Finland. Proc R Soc B 2010; 277 :593–600. https://doi.org/ 10.1098/rspb.2009.1659 . Rahman MH , Suzuki S, Kawai K. Formation of viable but nonculturable state (VBNC) of Aeromonas hydrophila and its virulence in goldfish, Car assius aur atus . Microbiol Res 2001; 156 :103–6. https: // doi.org/ 10.1078/ 0944501300084 . RStudio Team . RStudio: integrated Development for R. RStudio , Boston: PBC, 2020. https:// doi.org/ http:// www.rstudio.com/ . Shoemaker CA , LaFrentz BR, Peatman E et al. Influence of native catfish mucus on Flavobacterium columnare growth and pr oteol ytic activity. J Fish Dis 2018; 41 :1395–402. https:// doi.org/ 10.1111/ jfd.1283 3 . Smith VH , Schindler DW. Eutrophication science: where do we go fr om her e?. Trends Ecol Evol 2009; 24 :201–7. https:// doi.org/ 10.101 6/j.tree.2008.11.009 . Song YL , F ry er JL, Rohovec JS. Comparison of six media for the cultivation of Flexibacter columnaris . Fish Pathol 1988; 23 :91–4. https: // doi.org/ 10.3147/ jsfp.23.91 . Staroscik AM , Nelson DR. The influence of salmon surface mucus on the growth of Flavobacterium columnare . J Fish Dis 2008; 31 :59–69. https:// doi.org/ 10.1111/ j.1365-2761.2007.00867.x . Sundber g LR , K etola T, Laanto E et al. Intensiv e aquacultur e selects for increased virulence and interference competition in bacteria. Proc R Soc B 2016; 283 :20153069. https:// doi.org/ 10.1098/ rspb.201 5.3069 . Sundberg LR , Kunttu HM, Valtonen ET. Starvation can diversify the population structure and virulence strategies of an environmentall y tr ansmitting fish pathogen. BMC Microbiol 2014; 14 :67. https: // doi.org/ 10.1186/ 147121801467 . Therneau T . A pac ka ge for surviv al anal ysis in R. R pac ka ge v ersion 3.5-7. CRAN, 2023. Voss JD , Richardson LL. Nutrient enrichment enhances black band disease pr ogr ession in cor als. Coral Reefs 2006; 25 :569–76. https: // doi.org/ 10.1007/ s0033800601318 . Wedekind C , Gessner MO, Vazquez F et al. Ele v ated r esource av ailability sufficient to turn opportunistic into virulent fish pathogens. Ecology 2010; 91 :1251–6. https:// doi.org/ 10.1890/ 09-1067.1 . Wetzel RG . Limnology . 3rd edn. Philadelphia: Saunders College Publishing, 2001. Zhang, Xu DH , Shoemaker CA, Beck BH. The severity of motile Aeromonas septicemia caused by virulent Aeromonas hydrophila in channel catfish is influenced by nutrients and microbes in water. Aquaculture 2020; 519 :734898. https:// doi.org/ 10.1016/ j.aq uaculture.2019.734898 . Zhang, Zhao L , Chen W, Huang Y et al. Complete genome sequence analysis of the fish pathogen Flavobacterium columnare provides insights into antibiotic resistance and pathogenicity related genes. Microb Pathog 2017; 111 :203–11. https:// doi.org/ 10.1016/ j.micpath. 2017.08.035 . Recei v ed 25 August 2023; revised 29 J an uar y 2024; accepted 31 J an uar y 2024 ©The Author(s) 2024. Published by Oxford Uni v ersity Pr ess on behalf of FEMS. This is an Open Access article distributed under the terms of the Cr eati v e Commons Attribution License ( https://creativecommons.org/licenses/by/4.0/ ), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is pr operl y cited. Downloaded from https://academic.oup.com/femsec/article/100/4/fiae013/7596547 by Jyvaskyla University user on 17 April 2024