Virulence-inhibitory activity of the degradation product 3-hydroxybutyrate explains the protective effect of poly-β-hydroxybutyrate against the major aquaculture pathogen Vibrio campbellii

The bacterial storage compound poly-β-hydroxybutyrate, a polymer of the short-chain fatty acid 3-hydroxybutyrate, has been reported to protect various aquatic animals from bacterial disease. In order to obtain a better mechanistic insight, we aimed to (1) investigate whether 3-hydroxybutyrate is released from poly-β-hydroxybutyrate within sterile brine shrimp larvae, (2) determine the impact of 3-hydroxybutyrate on the virulence of Vibrio campbellii to brine shrimp larvae and on its cell density in the shrimp, and (3) determine the impact of this compound on virulence factor production in the pathogen. We detected 3-hydroxybutyrate in poly-β-hydroxybutyrate-fed brine shrimp, resulting in 24 mM 3-hydroxybutyrate in the intestinal tract of shrimp reared in the presence of 1000 mg l−1 poly-β-hydroxybutyrate. We further demonstrate that this concentration of 3-hydroxybutyrate does not affect the growth of V. campbellii, whereas it decreases the production of different virulence factors, including hemolysin, phospholipase and protease activities, and swimming motility. We hypothesize that by affecting all these virulence factors at once, 3-hydroxybutyrate (and thus also poly-β-hydroxybutyrate) can exert a significant impact on the virulence of V. campbellii. This hypothesis was confirmed in a challenge test showing that 3-hydroxybutyrate protected gnotobiotic brine shrimp from pathogenic V. campbellii, without affecting the number of host-associated vibrios.


Determination of 3-hydroxybutyrate levels in poly-β-hydroxybutyrate-fed brine shrimp larvae.
We previously hypothesised that the beneficial activity of poly-β-hydroxybutyrate is mediated by the release of water-soluble 3-hydroxybutyrate monomers 12 . In order to investigate this, we set-up sterile brine shrimp cultures with different concentrations of poly-β-hydroxybutyrate particles added to the rearing water, and then determined the concentration of 3-hydroxybutyrate in brine shrimp larvae harvested after 24 h. The concentration of 3-hydroxybutyrate in the brine shrimp samples increased with increasing doses of poly-β-hydroxybutyrate in the rearing water ( Table 1). The 3-hydroxybutyrate concentrations in samples from the 100 and 1000 mg l −1 poly-β-hydroxybutyrate treatments were significantly different from those in samples taken from brine shrimp cultures that had not received poly-β-hydroxybutyrate. Assuming that all 3-hydroxybutyrate originated from the intestinal tracts of the sampled larvae, we further calculated the concentration of 3-hydroxybutyrate in the brine shrimp gut, and found that 3-hydroxybutyrate concentrations increased to 24 mM in the presence of 1000 mg l −1 poly-β-hydroxybutyrate ( Table 1).

Impact of 3-hydroxybutyrate on the growth of Vibrio campbellii.
We investigated the impact of 3-hydroxybutyrate on the growth of V. campbellii in both nutrient-rich and nutrient-poor conditions, and at pH 6 or 7. At pH 6, the growth of V. campbellii was inhibited at the highest concentration of 3-hydroxybutyrate (125 mM), with a significantly longer lag phase in nutrient-rich conditions and a complete inhibition of growth in nutrient-poor conditions ( Fig. 1A and C). At pH7, on the other hand, only a relatively small effect was observed at 125 mM 3-hydroxybutyrate, both in nutrient-rich and nutrient-poor conditions ( Fig. 1B and D).
Scientific REPORTS | (2018) 8:7245 | DOI: 10.1038/s41598-018-25385-w Impact of 3-hydroxybutyrate on the survival of brine shrimp larvae challenged with Vibrio campbellii and on the cell density of shrimp-associated vibrios. In a further experiment, we determined the impact of 25 or 125 mM 3-hydroxybutyrate on the virulence of V. campbellii in a standardised model system with gnotobiotic brine shrimp larvae. We have previously demonstrated that the addition of 1000 mg l −1 poly-β-hydroxybutyrate to the rearing water offers protection against V. campbellii in this model system 13 . Brine shrimp larvae are particle filter feeders that are unable to accumulate dissolved compounds 27 and therefore, the concentration of 3-hydroxybutyrate within the brine shrimp larvae is expected to be the same as the concentration in the rearing water. Larvae challenged with V. campbellii showed a significantly higher survival in the presence of both concentrations of 3-hydroxybutyrate, and the survival of challenged larvae in the presence of 3-hydroxybutyrate was not significantly different from that of unchallenged larvae ( Fig. 2A). We also observed that the survival of non-challenged larvae was higher in the presence of 3-hydroxybutyrate, suggesting a direct beneficial effect on the brine shrimp. Finally, we determined the numbers of V. campbellii that were associated with the brine shrimp larvae and found that there were no significant differences between treatments (Fig. 2B).
Impact of 3-hydroxybutyrate on virulence factor production by Vibrio campbellii. We determined the impact of non-growth inhibitory concentrations of 3-hydroxybutyrate on the production of hemolytic, lipase, phospholipase and protease activities in V. campbellii. Hemolytic activity decreased at 1 mM of 3-hydroxybutyrate or more, with a maximal decrease of 28 and 18% at pH 6 and 7, respectively (Fig. 3A). Lipase activity was not affected by 3-hydroxybutyrate (Fig. 3B). Phospholipase activity decreased in the presence of 25 mM 3-hydroxybutyrate at pH6 and in the presence of 125 mM at pH 7 (Fig. 3C). Protease activity was decreased only at pH 7, with a 26% decrease in the presence of 25 mM 3-hydroxybutyrate, and no clearing zone extending beyond the colony diameter in the presence of 125 mM 3-hydroxybutyrate (Fig. 3D). Swimming motility decreased by 16% in the presence of 25 mM 3-hydroxybutyrate at pH 6, and by 41% in the presence of 125 mM  3-hydroxybutyrate at pH 7 (Fig. 4A). Finally, there was no impact of 3-hydroxybutyrate on biofilm formation and exopolysaccharide production at any of the concentrations tested ( Fig. 4B and C).

Discussion
During recent years, research has revealed that the bacterial storage compound poly-β-hydroxybutyrate, a polymer of the short-chain fatty acid 3-hydroxybutyrate, is able to improve the survival of various aquaculture species 13,[17][18][19][20][21][22] . Because poly-β-hydroxybutyrate is water-insoluble, we have hypothesized that in order to exert a  Asterisks indicate significant differences with the treatment without 3-hydroxybuturate (P < 0.05). Treatment with 125 mM 3-hydroxybutyrate was not tested at pH 6.
Scientific REPORTS | (2018) 8:7245 | DOI:10.1038/s41598-018-25385-w beneficial effect, the polymer must be degraded into water-soluble products (e.g. 3-hydroxybutyrate monomers) in the intestinal tract of aquatic animals, and that these water-soluble products are responsible for the beneficial effect 12 . This hypothesis was based on indirect observations including (1) increased survival of sterile starved brine shrimp in the presence of poly-β-hydroxybutyrate, (2) an improved effect of poly-β-hydroxybutyrate in the presence of poly-β-hydroxybutyrate-degrading bacteria, and (3) a decreased intestinal pH in sterile sea bass larvae receiving poly-β-hydroxybutyrate in their diet 13,16,28 . In this study, we demonstrate for the first time that 3-hydroxybutyrate is indeed released in poly-β-hydroxybutyrate-fed animals. Importantly, we used sterile brine shrimp for this experiment, excluding the possibility that the release of 3-hydroxybutyrate is a result of bacterial poly-β-hydroxybutyrate depolymerase activity. Therefore, the release of 3-hydroxybutyrate must be a result of abiotic decomposition and/or enzyme activity in the larvae. Abiotic decomposition in the rearing water is unlikely to explain the (relatively high) levels of 3-hydroxybutyrate we measured in the shrimp samples since the abiotic release of 3-hydroxybutyrate from poly-β-hydroxybutyrate in aqueous solutions is very limited. For instance, Freier et al. reported that less than 5% of PHB decomposed abiotically in phosphate buffer at 37 °C after 4 weeks 14 .
A similar decomposition rate would have resulted in 3-hydroxybutyrate concentrations in the brine shrimp rearing water below 2, 20 and 200 µg l −1 for the 10, 100 and 1000 mg l −1 poly-β-hydroxybutyrate treatments, respectively, at the time of taking our samples. Because brine shrimp are particle filter feeders that are not able to accumulate dissolved compounds 27 , the concentration of dissolved compounds (such as 3-hydroxybutyrate) in the shrimp body will be the same as in the water. This means that the 3-hydroxybutyrate concentration resulting from abiotic decomposition in the rearing water would have resulted in concentrations in the shrimp that are lower than 2, 20 and 200 µg l −1 . This is more than 175 times lower than the concentrations we measured in our samples. Previous research has also shown that poly-β-hydroxybutyrate decomposes in the digestive tract of animals 15 and that digestive enzymes increase the abiotic decomposition of the compound 14 . We calculated that the 3-hydroxybutyrate concentration in the gut of brine shrimp larvae reared in the presence of 1000 mg l −1 poly-β-hydroxybutyrate, increased to 24 mM. This concentration is within the range of concentrations of short-chain fatty acids (including 3-hydroxybutyrate) that have been shown to increase the survival of brine shrimp larvae challenged with V. campbellii LMG 21363 11,13 , and this was confirmed here for V. campbellii ATCC BAA-1116. The effect we observed here for strain ATCC BAA-1116 (complete protection) was more pronounced than what we found previously for strain LMG 21363 13 , which reflects the fact that strain LMG 21363 is more virulent than strain ATCC BAA-1116 29 .
In order to obtain a further mechanistic insight into the protective effect of the poly-β-hydroxybutyrate degradation product 3-hydroxybutyrate, we subsequently determined the impact of the compound on the growth of V. campbellii, both in nutrient-rich and nutrient-poor conditions, the latter of which might be more relevant to the conditions experienced in the rearing water and in a host. We expected that the impact would be stronger in nutrient-poor conditions because short-chain fatty acids are thought to interact with the energy balance in bacteria 10 , and since the cellular energy pool is expected to be lower in nutrient-poor conditions. We found that 3-hydroxybutyrate inhibits the growth of the vibrios in a pH-dependent manner, showing inhibition at 125 mM at pH 6 (but not at pH 7), and this effect was indeed more pronounced in nutrient-poor conditions. The effect we observed here is similar to what has been reported before for other short-chain fatty acids in vibrios 11,13 and other Gram-negative bacteria, such as Salmonella spp 25 .
We previously reported that short-chain fatty acids (including 3-hydroxybutyrate) increased the survival of brine shrimp larvae challenged with V. campbellii LMG 21363 at concentrations that were below the growth-inhibitory concentration 11,13 , and this was confirmed here for V. campbellii ATCC BAA-1116. Furthermore, we showed here that 3-hydroxybutyrate had no effect on the number of vibrios associated with the brine shrimp larvae, further indicating that the beneficial effect is not caused by inhibition of the growth of the pathogen. Because of this, we hypothesized that the compound might decrease the ability of the vibrios to attack their host, and in order to investigate this, we tested the impact of 3-hydroxybutyrate on a number of virulence-related phenotypes. We found that 3-hydroxybutyrate could indeed decrease hemolytic, phospholipase and protease activities, and swimming motility in V. campbellii. Consistent with this, Kiran et al. recently reported that poly-β-hydroxybutyrate degradation products generated by poly-β-hydroxybutyrate depolymerase activity of Vibrio sp. PUGSK8 inhibited hemolytic activity and motility of the bacterium, without affecting its growth 30 . Although the degradation products were not identified in the latter study, poly-β-hydroxybutyrate depolymerases usually degrade poly-β-hydroxybutyrate into 3-hydroxybutyrate oligomers or monomers 31 . Hence, we hypothesize that the effect observed for poly-β-hydroxybutyrate-treated Vibrio sp. PUGSK8 was also caused by 3-hydroxybutyrate. Short-chain fatty acids are considered to affect the energy balance of bacterial cells (leading to growth inhibition at a certain concentration) 10 , and given the fact that the production (and activity) of virulence factors is metabolically costly 32,33 , we hypothesize that subinhibitory concentrations of 3-hydroxybutyrate decrease the cellular energy status to a level at which virulence, but not growth, is decreased, thus exerting a non-specific antivirulence effect.
Several putative virulence factors have been proposed in vibrios belonging to the Harveyi clade. These include flagellar motility, which enables the pathogens to colonize the host, exopolysaccharides that enable pathogens to form biofilms and to attach to host tissues and protect them from the host's defense system, and lytic enzymes that degrade host tissues and enable pathogens to obtain nutrients 34 . Amongst the most well-known lytic enzymes are hemolysins (damaging blood cells), lipases (degrading lipids), phospholipases (degrading phospholipids) and proteases (degrading proteins). We previously reported that there is a positive correlation between hemolysin and protease expression, and virulence towards brine shrimp larvae 35 , and that swimming motility is required for full virulence of V. campbellii 36 . Furthermore, hemolytic and phospholipase activity have also been implicated in virulence of strains belonging to the Harveyi clade 37,38 . Hence, by simultaneously decreasing the production of these virulence factors (even though some of the effects were relatively modest), 3-hydroxybutyrate (and thus also poly-β-hydroxybutyrate) can exert a significant impact on the overall virulence of V. campbellii in the host. Furthermore, Baruah et al. recently reported that poly-β-hydroxybutyrate induces protective innate immune responses in gnotobiotic brine shrimp larvae challenged with V. campbellii, including an increase in phenoloxidase activity, one of the major innate immune responses of crustaceans 39 . It would be highly interesting to unravel which property of poly-β-hydroxybutyrate is responsible for this effect as it might also be mediated by a degradation product of the polymer (such as 3-hydroxybutyrate). Hence, we think that the strong protective effect of poly-β-hydroxybutyrate against bacterial infections can be explained by both a decreased virulence of the pathogens (caused by the degradation product 3-hydroxybutyrate) and an increased activity of the immune system of the host.
In conclusion, our data demonstrated that 3-hydroxybutyrate is released from poly-β-hydroxybutyrate in sterile brine shrimp, and that this compound, at subinhibitory concentrations, inhibits several virulence-related phenotypes in pathogenic V. campbellii. Together with an induction of immune responses of the host, these data explain the protective effect of poly-β-hydroxybutyrate against infections.

Methods
Bacterial growth conditions. V. campbellii ATCC BAA-1116 was routineously grown in Marine Broth (Difco Laboratories, Detroit, USA) at 28 °C with shaking (100 rpm). Cell densities were determined spectrophotometrically at 600 nm. All  Two hundred mg of cysts were hydrated in 18 ml of tap water during 1 h. Sterile cysts and nauplii were obtained via decapsulation, adapted from the protocol described previously 40 . Briefly, 660 µl of NaOH (32%) and 10 ml of NaOCl (50%) were added to the hydrated cyst suspension. The decapsulation was stopped after 2 min by adding 14 ml of Na 2 S 2 O 3 .7H 2 O (10 g l −1 ). During the reaction, 0.22 µm filtered aeration was provided. The decapsulated cysts were washed with autoclaved artificial seawater containing 35 g l −1 of Instant Ocean synthetic sea salt (Aquarium Systems Inc., Sarrebourg, France). The cysts were resuspended in a 50 ml tube containing 30 ml of autoclaved artificial seawater and hatched for 30 h on a rotor (4 min −1 ) at 28 °C with constant illumination (2000 lux).

Determination of 3-hydroxybutyrate levels in poly-β-hydroxybutyrate-fed brine shrimp larvae.
1000 mg l −1 stock solutions of poly-β-hydroxybutyrate were prepared by dissolving 500 mg sterile poly-β-hydroxybutyrate particles (98% poly-β-hydroxybutyrate, 2% poly-β-hydroxyvalerate; average diameter 25 µm; Goodfellow, Huntingdon, Ireland) into 500 ml of sterile artifical seawater. The suspensions were sonicated in a sonicater bath for 30 minutes. Axenic brine shrimp larvae were stocked in Schott bottles containing 500 ml sterile artificial sea water and mixed with the poly-β-hydroxybutyrate stock suspension to obtain different concentrations of poly-β-hydroxybutyrate particles. The bottles were provided with sterile (0.22 µm filtered) aeration and incubated for 24 h at 28 °C. Subsequently, the brine shrimp larvae were harvested on a 150 µm sieve and washed with distilled water. The bottom of the sieve was dried with paper cloth to remove excess water, and 350 mg samples of the brine shrimp larvae were transferred into eppendorf tubes, immediately frozen in liquid nitrogen and stored at −80 °C until further use. 3-hydroxybutyrate levels in the samples were determined with the β-Hydroxybutyrate (Ketone Body) Colorimetric Assay Kit (Cayman Chemicals, USA). Briefly, 1 ml of assay buffer (1 metros-HCl, pH 8,5) was added to the brine shrimp samples. The suspensions were mixed for 40 seconds using a vortex mixer (Ultrathurax), and subsequently centrifuged in a table-top centrifuge at 14500 rpm for 5 minutes. The supernatants were filtered over 0.22 µm filters and transferred into new eppendorf tubes. The 3-hydroxybutyrate concentration in the samples were determined as described in the manufacturer's instructions. The 3-hydroxybutyrate concentration in the brine shrimp gut was calculated as follows. Assuming a dry weight of 5 µg per brine shrimp larva 41 , the samples contained approximately 7000 larvae. Further, Gunasekara et al. reported that the gut volume of a brine shrimp larva is approximately 0.002 µl 42 . Therefore, the total gut volume in the samples was 14 µl. The 3-hydroxybutyrate in the brine shrimp gut was then calculated assuming that all 3-hydroxybutyrate that was present in the 1 ml samples originated from the 14 µl total gut volume of the brine shrimp larvae used to prepare the samples.

Determination of the impact of 3-hydroxybutyrate on the growth of Vibrio campbellii. V. campbellii
was grown overnight in Marine Broth on a shaker at 28 °C. Subsequently, the strain was transferred 1:50 (v/v) to either undiluted Marine Broth or 50-fold diluted Marine Broth supplemented with salts to reach the same salt concentration as in undiluted Marine Broth (denoted nutrient-rich and nutrient-poor conditions, respectively). The media were supplemented with β-hydroxybutyrate at different concentrations (0, 1, 5, 25 and 125 mM). The pH of all solutions was adjusted to 6 or 7 by adding HCl. After inoculation, the suspensions were incubated at 28 °C in a static mode. Growth was monitored by measuring the optical density (OD 600 ) during 24 h. Each treatment was performed in triplicate.
Determination of the impact of 3-hydroxybutyrate on hemolytic, lipase, phospholipase and protease activities of Vibrio campbellii. Hemolysin, lipase, phospholipase and protease activities were determined as described previously 43 . Hemolytic activity was determined by supplementing Marine agar containing different concentrations of 3-hydroxybutyrate at pH 6 or 7 with 5% defibrinated sheep blood (Oxoid, Basingstoke, Hampshire, UK). Lipase activity was tested by adding 1% Tween 80 (Sigma-Aldrich) into marine agar containing different concentrations of 3-hydroxybutyrate at pH 6 or 7. Phospholipase activity was assessed on Marine agar supplemented with 1% egg yolk emulsion (Sigma -Aldrich), and different concentrations of 3-hydroxybutyrate at pH 6 or 7. Protease assay plates were prepared by mixing double strength Marine agar containing different concentrations of 3-hydroxybutyrate at pH 6 or 7 with 4% skim milk powder (Oxoid, Basingstoke, Hampshire, UK), autoclaved separately at 121 °C for 5 minutes to prevent denaturation of the protein. Clearing (hemolysin and protease) or opalescent (lipase and phospholipase) zones were measured after 3 days of incubation at 28 °C. The activity zones were divided by the respective colony diameters in order to correct for possible differences in growth. In those cases where there was no activity zone extending beyond the colony diameter, the activity zone was set equal to the colony diameter because in such case it is not possible to accurately determine the activity zone.
Determination of the impact of 3-hydroxybutyrate on biofilm formation and exopolysaccharide production by Vibrio campbellii. Biofilm formation and exopolysaccharide production were determined as described previously 44 . A grown culture of V. campbellii in Marine broth at an OD 600 of 1 was inoculated 1:10 (v/v) into Marine Broth containing different concentrations of 3-hydroxybutyrate at pH 6 or 7, and 200 µl aliquots of these suspensions were pipetted into the wells of a 96-well plate. Then, the bacteria were allowed to adhere and grow without agitation for 24 h at 28 °C. After that, the cultures were removed and the wells were washed three times with 300 μl sterile physiological saline (9 g l −1 NaCl) to remove all non-adherent bacteria. The remaining attached bacteria were fixed with 150 μl of 99% methanol per well for 20 min, after which the methanol was removed and plates were air-dried. Then, biofilms were stained for 15 min with 150 μl of a 1% crystal violet solution (Pro-lab Diagnostics, Richmond Hill, ON, Canada) per well. Excess stain was rinsed off by placing the plate under running tap water, and washing was continued until the washings were free of the stain. After the plates were air dried, the dye bound to the adherent cells was resolubilized with 150 μl of 95% ethanol per well, and absorbance was measured at 570 nm. Sterile medium served as negative control. For the quantification of exopolysaccharides, Calcofluor white staining (Sigma-Aldrich) was used. In brief, wells were rinsed after 24 h biofilm formation and 100 μl phosphate buffered saline containing 0.5 μl 5 mM Calcofluor white staining dye Scientific REPORTS | (2018) 8:7245 | DOI:10.1038/s41598-018-25385-w was added to the wells. After 60 min, fluorescence (excitation 405 nm and emission 500 nm) was measured with a Tecan Infinite M200 multi-reader.

Determination of the impact of 3-hydroxybutyrate on swimming motility of Vibrio campbellii.
Swimming motility assays were performed on soft agar (Marine Broth containing 0.2% agar) containing different concentrations of 3-hydroxybutyrate (0, 1, 5, 25 and 125 mM) and at pH6 or pH7, as described previously 36 . V. campbellii was grown overnight in Marine Broth, and 5 µl of the culture was spotted in the center of the soft agar plates. Plates were incubated for 24 hours, after which the diameters of the motility halos were measured. All assays were done in triplicate.
Brine shrimp challenge test. Challenge tests were performed as described previously 45 . Briefly, after hatching, groups of 20 nauplii were transferred to new sterile 50 ml tubes that contained 20 ml of autoclaved artificial seawater with different concentrations of 3-hydroxybutyrate (0, 25 and 125 mM, respectively) (Sigma-Aldrich, Bornem, Belgium) and set at pH7. The animals were fed with Aeromonas sp. LVS3 at 10 7 cells ml −1 , and for the challenge treatments, tubes were inoculated with V. campbellii at 10 7 cells ml −1 . Subsequently, the falcon tubes were put back on the rotor and kept at 28 °C, and survival was scored after 2 days. All manipulations were done under a laminar flow hood in order to maintain sterility. Each treatment was done in triplicate.
Quantification of brine shrimp-associated vibrios. Bacterial numbers associated with brine shrimp larvae were determined as described previously 45 . Briefly, rinsed larvae were homogenised with a pestle and sharp sand, followed by 1 min of bead beating. The homogenised samples were serially diluted and spread-plated on Marine agar.

Statistics.
Data analysis was carried out using the SPSS statistical software (version 15). Unless stated otherwise, all other data were compared with independent samples t-tests.