Isolation and characterization of native probiotics for fish farming

Among animal food sources, fish are considered particularly important, providing high quality protein and essential nutrients such as polyunsaturated fatty acids (PUFA), minerals and vitamins. Undoubtedly, in the context of stagnating fisheries landings combined with increasing per capita consumption of fish worldwide, the demand for fish can only be met by a sustainable development of the aquaculture industry, where resources are by far more efficiently used [1, 2]. Here, limited feedstuff – most importantly fishmeal and fish oil – challenges future expansion. Thus, innovations in fish nutrition may act as drivers for the development of this industrial sector. In this context, functional diets that provide benefits by targeting specific physiological mechanisms to improve the health and immune status or optimize growth and feed conversion have a huge potential. Among dietary supplements, probiotics have been widely assessed in functional diets of terrestrial livestock but have an expandable potential in fish nutrition.

In fish nutrition, the mentioned limitations in fishmeal and the rapid growth of the aquaculture sector led to numerous research efforts in exploring replacements for fishmeal as protein source. Several studies focusing on protein-enriched agricultural by-products such as press cakes from plant oil production or residuals from starch production [3–5]. If plant-based proteins are used, anti-nutrients are a major constraint [6–8] affecting the performance of the farmed species, including reduced growth, feed conversion as well as manifold pathological alterations [3–5]. Often, plant derived ingredients such as soy or other legumes cause intestinal enteritis via detrimental secondary metabolites like saponin [9–12]. For example, most fish species do not tolerate less than 50% fishmeal replacement by soybean protein, commonly assigned to the anti-nutritional effects of saponin [11].

In 1905, Metchnikoff was the first to point out the positive role of bacteria in milk and yoghurt products for human health. Probiotics have therefore been defined as bioactive, living microbial food/feed additives, which have a positive influence on the digestion and, moreover, the microflora of the gastrointestinal tract (GIT) in general [13–15], improving nutrition as well as disease resistance. Among those beneficial effects, probiotics have been reported to improve a) disease resistance & immunity [16–18], b) nutrition & feed utilization [7, 19], c) reproduction & development [20–23], e) gut morphology and functioning [24, 25] as well as f) counteracting spinal curvature [26, 27].

In the past, aquaculture research has focused on well-known probiotic strains derived from terrestrial hosts, ignoring fundamental differences in the physiology of cultured species – mammals or birds versus fish – as well as those differences related to the environment where the respective microbial communities actually evolved - aquatic versus terrestrial environment. Therefore, probiotic bacteria isolated from the respective fish host are expected to perform better in their natural habitat than those derived from terrestrial hosts [28]. Thus, it is not surprising that in most cases, effects observed during the actual application, vanish rather fast if administration is terminated [29, 30]. Only recently, native probiotic candidates have been explored, mostly in classical fish feeding trials focusing on the evaluation of growth and feed conversion as well as immune modulation [31–34]. Such feeding trials are costly, time consuming and only feasible with a few candidates. Thus, fast in vitro screening strategies have to be established, allowing the selection of the most promising isolates. Ideally, in vitro screening would allow the identification of beneficial effects and reduce the risk of negative impacts. Furthermore, such research may provide new insights on the biology and ecology of autochthonous bacteria and improve the knowledge of microbial host interactions [35–37].

Used as prophylaxis or therapeutic treatment, probiotics act via competitive exclusion of the respective pathogen or directly inhibiting its propagation [38, 39]. Therefore, probiotics represent an alternative in disease prophylaxis and treatment, particularly in the context of emerging antibiotic resistance reported from aquatic environments in general [40–43] as well as from aquaculture sites [44, 45].

Nutritional effects reported upon probiotic administration include a better supply of nutrients such as essential fatty acids and their precursors [14]. PUFA namely EPA and DHA provide health benefits for the consumer, including optimal development of the nervous system during early ontogeny, reduced risk of abnormal heart rhythms, heart failure or strokes as well as anti-inflammatory effects. Consequently, probiotics may improve product quality, e.g., enrichment of essential fatty acids.

In this study, we evaluated candidate probiotics native to closely related flatfish species using an in vitro screening approach. Therefore, morphologically different colonies were isolated and characterized based on 16S rRNA sequence information. Consequently, isolates that could be stored as cryo-culture were evaluated using specific in vitro assays on 1) the antagonism towards three relevant fish pathogens Tenacibaculum maritimum, Edwardsiella tarda, Listonella anguillarum, 2) the synthesis of essential fatty acids EPA, DHA and precursors such as octadecatrienoic acid (ALA), octadecadienoic acid (LA), eicosatetraenoic acid (ETA) and 3) the metabolization of saponin as an example of a plant derived anti-nutrient. Finally, a cost-effective method for the supplementation of probiotics to fish feed was evaluated, assessing the viability of the respective bacteria upon top coating during storage.

Probiotics have the potential to improve immune status as well as performance of farmed fish but the use of non-native bacteria derived from endothermic terrestrial species may undermine a successful colonization of the GIT of farmed fish. Thus, isolation of native candidates is promising if in vitro screening for beneficial effects allow the extraction of a feasible number of candidates for in vivo evaluation in classical feeding trial or challenge test. Ideally, probiotic candidates should exhibit high growth rates at the respective rearing temperatures, be easily stored in cryo-cultures, oxygen tolerant and effectively supplemented to the diet assuring sufficient viability. Although several demanding strategies such as micro- and bio-encapsulation of the probiotics have been described [46, 47], we present a simple and cost-effective method for the coating that only affected the integrity of the pellets to an acceptable degree.

In the present study several probiotic candidates were identified addressing antagonism towards a major pathogen of turbot, Tenacibaculum maritimum, synthesis of essential FA and metabolization of plant-specific anti-nutrients, using saponin as a model substance. Among the autochthonous microbial community closely associated with the tissue surface of the fish intestine, a few pathogens were identified here. This confirms that pathogens are part of the teleost microbiom, even in fish that do not exhibit any symptoms of disease [36, 48]. Molecular characterization may allow the exclusion of pathogenic taxa but in several groups 16S rRNA sequencing does often not provide species-specific identification as observed in this study (Table 2). For example, isolates assigned to Acinetobacter and Psychrobacter comprise beneficial candidates as well as recorded pathogens, as presented below.

In this study, bacterial isolates were considered as closely associated with the intestinal wall (autochthonous), since faeces were removed and loosely associated, transient bacteria were washed away before detaching candidate bacteria with a non-ionic surfactant (Triton X 100), which is commonly used for this purpose [49, 50]. The ability to attach to the mucosal surface, pH and the resistance to bile acids are particularly important to allow a successful colonization of the intestine [51, 52]. Although there is evidence for a core gut microbiota in teleost fish, the diversity of identified intestinal bacteria is enormous and depending on various biotic and abiotic factors [48, 53]. According to Xing et al. 2013 [36], the predominant bacterial phylum in turbot is Proteobacteria accounting for 93% in 16S rRNA gene sequencing [36]. This is in line with our findings as 89.7% were members of the Proteobacteria (Table 1). Considering that most bacteria in the intestine may not be cultivable with the media and under the conditions used here, there will be a considerable discrepancy to the actual microbiom.

Among the isolates, Psychrobacter candidates were particularly interesting due to the fact that Psychrobacter species are commonly observed in the aquatic environment as well as commercially important freshwater [54] and marine teleost species, e.g., Atlantic cod Gadus morhua [55], mackerel Scomber scombus [56], fine flounder Paralichthys adspersus [57] or grouper Epinephelus coioides [35, 58]. In our study, by far the most frequent genus was Psychrobacter accounting for 70 out of 195 (35.9%) isolates based on 16S rRNA information but unfortunately this does not allow species-specific distinction. For future work on specific candidates of this genus molecular characterization will target the gyrB gene [59].

Moreover, Psychrobacter has already been evaluated as a probiotic additive in fish nutrition. Lazado et al. reported on antagonistic activity of Psychrobacter sp. isolated from the intestine of Atlantic cod against two important fish pathogens Listonella anguillarum and Aeromonas salmonicida subsp. salmonicida [60]. Also, increased activity of digestive enzymes was reported by the same authors, indicating the potential for improved digestion upon feeding probiotic supplements [61]. Similar to the observations here, differential effects on three pathogens were reported. Surprisingly, in all our isolates, no antagonistic effects against two other pathogens frequently reported from turbot culture, Edwardsiella tarda and Listonella anguillarum were observed. Lazardo et al. demonstrated inhibitory activity against the two pathogens, even when pure supernatant was used [60]. This indicates that Psychrobacter strains might be capable of producing and secreting effective antimicrobial substances.

The use as probiotic in fish nutrition was also addressed by a classical feeding trial in the orange-spotted grouper Epinephelus coioides [58]. Here, Psychrobacter sp. as an autochthonous intestinal isolate induced a significantly elevated feed conversion and growth performance, correlated to increased activities of digestive enzymes (e.g., hepatopancreatic protease & lipase, intestinal amylase). Moreover, oxidative stress related parameters like superoxide dismutase were slightly elevated [58]. Congruently, Makled et al. observed increased growth rates and feed utilization as well as immune stimulation (e.g., significantly increased IgM, phagocytic and lysozyme activity) after feeding Psychrobacter to Nile tilapia Orechromis niloticus [62]. As conclusion, Psychrobacter isolates should be investigated in more depth and might be sufficient candidates for future in vivo evaluation in turbot.

One of the most promising isolates of the present culture collection is Acinetobacter haemolyticus, revealing strong antagonism towards T. maritimum, an interesting profile of essential fatty acids and the ability to metabolize saponin (Table 2). Interestingly, beneficial Acinetobacter species have been used as probiotics in shrimp aquaculture [63] but to our knowledge no Acinetobacter species have been evaluated as probiotic supplement in fish farming, probably due to increasing reports on opportunistic fish pathogens within this genus [64–67]. However, our isolate assigned to A. haemolyticus seems to be taxonomically not related to pathogenic Acinetobacter species.

In this study, we found 15.4% (n = 30) opportunistic or pathogenic species, which is in line with earlier findings. An uncontrolled proliferation of such species illustrates the outbreak of diseases [68]. Several isolates closely related to common pathogens such as Vibrio alginolyticus and Aeromonas hydrophila [69] have been suggested as probiotics in former studies [70–72]. Here, Staphylococcus saprophyticus 060 revealed saponin metabolization but has previously been suggested as potential pathogen by Yang et al. [35]. Similarly, evidence for the species-specific pathogenicity is often limited and argumentation, thus, strongly based on the phylogenetic relationships.

In aquaculture, particularly in a commercial context, probiotics mostly originate from terrestrial livestock and are frequently members of lactic acid bacteria [46, 73]. However, a successful and thus persistent colonization of these probiotics has never been demonstrated. Native probiotic bacteria are adapted to the ambient environment of the target species and thus, considered more effective, exerting longer lasting beneficial effects once applied [60, 74]. Screening for native isolates should consequently be prioritized in the future. Also, native bacteria can a priori be considered adapted to the environmental conditions in the gut, eliminating the need of further testing as required in non-native probiotics. In vivo screening strategies including assays as the ones used here will support the selection of potential candidates but evaluation in feeding trials limits the number of candidates drastically. The ultimate need for such an experimental evaluation requires any easy and cheap way to formulate experimental diets [74]. Here, we were able to demonstrate viability of probiotic candidates over 54 d upon dietary supplementation.

It is surprising that information on the probiotic diet preparation is often insufficient, including the probiotic cultivation and dosage, details on drying process and shelf life assessment. Here, we demonstrate comparable viability between different probiotic species upon top coating of any formulated diet. Expensive and elaborate probiotic incorporation processes using technologies like drum or vacuum coaters are not required for the preparation and extended storage assures high survival of the probiotic supporting experimental evaluation. After almost eight weeks the viable colony forming counts did not even decrease by a tenth power. When supplementing experimental diets with probiotics for a feeding trial, the need arises to up-scale cell cultivation. Here, bacterial cell should be harvested at the end of the logarithmic growth phase, because a delayed harvest could cause a self-inhibition of bacteria or the production of unwanted secondary metabolites. The potential loss of live and active probionts will undermine evaluation. Therefore, for any evaluation, viability should be assessed experimentally.

Plant derived proteins have been increasingly used as an alternative to fishmeal in aquafeeds [7, 75] but often exhibit deleterious effects on the health and performance of fish [12]. Among those, soybean protein represents a major source of protein for animal nutrition but is particularly challenging for carnivorous finfish due to the adverse effects observed. Although soybean meal has a favorable amino acid composition, substitution rates are kept relatively low to avoid negative effects on growth performance, feed conversion or even severe enteritis, mostly assigned to the secondary metabolite saponin [76]. Therefore, as a model substance to focus on here, we addressed the ability of bacteria to metabolize saponin in a screening test. We identified seven isolates that were able to metabolize saponin as only carbon and energy source at usual concentrations of synthetic nutrient solutions (see SMA). Thereby, it seems plausible that the respective isolates may improve degradation of saponin in the intestine of the host, counteracting detrimental consequences of soybean meal in fish nutrition. In the near future, respective isolates characterized here will be evaluated in a feeding trial.

Next to fishmeal, and currently an even more pressing problem in fish nutrition, is the limitation of fish oil in the context of rapidly increasing aquaculture activities worldwide. Here, its replacement by plant ingredients is more problematic, particularly from a consumer’s perspective, since fish oil is the main source of essential FA such as DHA or EPA. Considering a negligible endogenous synthesis in vertebrates and the outstanding role of these FA for human nutrition, reduced utilization of fish oil severely impacts product quality unless other sources (e.g., genetically modified organisms) assure the supply to the farmed fish. As an alternative, we tried to identify probiotic candidates characterized by high levels of DHA, EPA and the precursors LA, ALA and ETA. Such probiotics upon digestion could provide these essential FA to the farmed fish. The FA profile revealed a remarkable variability between isolates. Highest DHA and EPA levels were observed in the isolates 012 Psychrobacter sp. and 047 Paracoccus sp. Moreover, both isolates were particularly rich in essential FA precursor LA and ALA (Fig. 1). We consider these candidates highly promising with regard to the supply of essential FA, aiming at an in vivo evaluation.

When searching for probiotic candidates effective in vitro screening will focus on quantifiable traits that will be translated into beneficial effects in the performance of the farmed species, thereby reducing the number of candidates to be tested. We report the screening of native, autochthonous candidates based on a molecular characterization and subsequent screening for interesting features related to pathogen antagonism, SEFA and metabolization of feed-derived anti-nutrients. The preparation of experimental diets for in vivo evaluation assuring high viability of the candidate is furthermore presented.