Effect of Pediococcus sp . , Pediococcus pentosaceus , inulin and fulvic acid , added to the diet , on growth of Oreochromis niloticus

Lactic acid bacteria (LAB) were isolated from the intestine of Oreochromis niloticus and its effect, along with that of inulin and fulvic acid, on growth of the same species was determined. Characterization of the isolates was performed through assessing the Gram stain, morphology, cell arrangement, hemolytic activity, antagonism, growth, hydrophobicity, extracellular enzyme activity, and counting of colony forming units (CFU), to select those microorganisms with probiotic potential. The selected isolates were identified at the molecular level by the amplification of 23S and 5S ribosomal genes. Treatments were performed in triplicate in plastic tanks with 600 l of filtered fresh water and 12 fish per tank (1.3 ± 0.12 g):


INTRODUCTION
Aquaculture requires quality diet with high protein content and complementary additives to maintain healthy organisms and to favor the growth of the animals.Some of the most used additives to promote growth include hormones, antibiotics, ionophores, and some salts (Klaenhammer, 1988;Fuller, 1994;Mohapatra et al., 2012).The inappropriate use of these growth promoters can have adverse effects on the animal and on the end consumer, and can promote resistance to pathogenic bacteria, as occurs with the use of antibiotics (Lara-Flores et al., 2003;Mohapatra et al., 2012).
Based on the aforementioned, there is a great interest in the use of natural feed additives such as probiotics to improve the resistance to diseases, water quality and/or growth of cultivated organisms (Verschuere et al., 2000;Nayak, 2010;Mohapatra et al., 2012).Probiotics have beneficial effects on the host due to the changes in the microbial community related to the host and the environment through an improvement in the utilization of the food or in its nutritional value, or by improving the host's response to disease or the quality of its environment (Verschuere et al., 2000;Balcázar, 2002;Balcázar et al., 2008;Zhou et al., 2010a;Mohapatra et al., 2012).Among the organisms commonly used as probiotics in fish are lactic acid bacteria (LAB) like Lactobacillus plantarum and Lactobacillus fructivorans, which have been tested in Sparus aurata (Carnevali et al., 2004), Lactobacillus sp. in ornate fish Carassius auratus and Xiphophorus helleri (Abraham, 2008), Enterococcus durans, O. niloticus (Poot-Poot, 2001), Streptococcus faecium and L. acidophilus in O. niloticus (Lara-Flores et al., 2003), the commercial probiotic, Renascitur ® , in Ariopsis bonillai (Rodríguez-Méndez et al., 2006), and Bacillus amyloliquefaciens (Mohammad et al., 2012).
Fulvic acid is a derivative of humic acids, the latter being a complex mixture of organic material coming from leaves, branches, stems, etc., that decompose in soil.This decomposition process is accomplished by microorganisms and fungi, producing fulvic acid and involves formation of compounds with very low molecular weight and positively charged ions (chelation process).Mineral chelated compounds are highly absorbable by plants and animals (Osman et al., 2009).
In this work, we isolated and characterized LAB from the intestine of tilapia and we determined their effect together with inulin and fulvic acid on growth in weight of O. niloticus, cultured under laboratory conditions.

Isolation of presumptive LAB
Fish were collected (13.00 ± 12.84 g) from the Sinaloa River and from an agricultural irrigation channel (Guasave, Sinaloa, Mexico).
Wild fish came from artificial dams.Petri plates were prepared with (Man, Rogosa and Sharp) (MRS) and Rogosa agar (BD Difco, Sparks, MD, USA).Each fish was measured, weighed, and disinfected with alcohol (70%).The whole intestine was extracted with a scalpel and macerated in a mortar with 150 µl of distilled water.From the macerated product, 100 µl were taken and transferred in the Rogosa agar plates using a triangular plastic loop.Petri plates were incubated at 30°C, for 48 h.The presumptive LAB colonies were identified (small, convex, circular, with smooth border, and white in color) and subsequently isolated again in MRS medium through cross streaking.Petri plates were incubated as mentioned before.
Isolated colonies were transferred on Petri plates with MRS medium and incubated at 30°C, for 48 h.Bacteria were harvested and re-suspended in Eppendorf tubes with 1 ml of MRS medium and 15% glycerol (v/v).The bacterial suspension was preserved at -80ºC.

Characterization of presumptive LAB
For the hemolysis test, Petri plates were prepared with blood agar with 5% human blood, making 6-mm diameter holes in the plate with a sterile holemaker.Isolates were cultured in MRS broth at 30°C, for 48 h, and centrifuged at 10,000 g, for 10 min.The LAB supernatant was adjusted to a pH of 6.5-7.0 with 1 M NaOH.Wells were inoculated with 50 µl of the supernatant and the plates were incubated at 30°C for 24 h.The lysis halo was observed to determine the type of hemolysis (alpha, beta, or gamma).Isolates with alpha or beta hemolysis were discarded because they can lysate cells.
Presumptive LAB isolates with gamma hemolysis were antagonized against Vibrio sp.They were transferred by distributing 5 × 10 4 CFU of vibrio in Petri dish containing trypticase soy agar (TS, BD Bioxon, Sparks, MD, USA), in which holes were made as mentioned before.Each well was inoculated with 50 µl of the supernatant from a 24 h liquid LAB and yeast cultures.The supernatant of yeast cultures was taken and the pH of the liquid LAB culture was adjusted as mentioned before.MRS broth was used as negative control.Plates were incubated for 24 h at 37C, and the inhibition halo was measured (Bauer et al., 1966).Growth kinetics (absorbance vs time) was applied to the selected LAB isolates LAB 35,and LAB 37) to know the adaptation, exponential, and stationary stages.MRS broth (50 ml) was prepared and this was inoculated with 20 µl of the stock of each LAB isolate, and incubated at 30ºC.Absorbance (580 nm) was determined at 6, 12, 24, 48, 72, and 96 h in a spectrophotometer Thermo Spectronic Genesys 2 (Thermo Scientific, Rochester, NY, USA).The hydrophobicity test was performed on the three selected isolates that were transferred through cross streaking in MRS agar medium supplemented with 0.03% Congo red (Sigma, St. Louis, MO, USA) and plates were incubated at 30°C, for 24-48 h.Positive results corresponded to those isolates with red coloring, and negative were those with white or translucent coloring (Sharma et al., 2006).
To obtain the LAB supernatant, the three selected isolates were cultured in MRS broth and incubated at 30°C, for 24-48 h.Samples were centrifuged at 10,000 g for 10 min to obtain the supernatant.The pH of the LAB supernatant was adjusted as mentioned above (León et al., 2000).
For the casein (proteases) degradation test, Petri plates were prepared with the basal medium (1.5% agar and 0.5% yeast extract) supplemented with 2% fat-reduced milk, once the medium had solidified, perforations were made as mentioned above.Each hole was inoculated with 50 µl of the culture supernatant, using as control the culture medium.Plates were incubated at 30°C, for 24-48 h, considering those isolates with a transparent halo around the holeas positive.
For the gelatin hydrolysis test (proteases), Petri plates were prepared with basal medium supplemented with gelatin (1%).Perforations were made as mentioned above.Wells were inoculated with 50 µl of the culture supernatant, using as control the culture medium.Plates were incubated at 30°C, for 24-48 h, considering as positive result those isolates with an opaque halo around the well.
For the Tween 80 hydrolysis test (lipases), Petri plates were prepared with basal medium supplemented with 1% Tween 80 (Sigma, St. Louis, MO, USA), once the medium had solidified perforations were made as described above.Wells were inoculated with 50 µl of the culture supernatant, using as control the culture medium.Plates were incubated at 30°C, for 24-48 h, considering as positive result those isolates with an opaque halo around the well.
To count the selected LAB isolates, we performed a 24-h culture at 30ºC in MRS broth to count the CFU.Bacterial cultures were centrifuged at 10,000 g for 10 min, the supernatant was discarded and the pellet was re-suspended in 1 ml of sterile distilled water.The bacterial solution was adjusted to an optical density (580 nm) of 1, in a Thermo Spectronic Genesys 2 spectrophotometer.We determined the CFU/ml for each isolate, using the serial dilution method.
The DNA was extracted following the instructions of the Bactozol kit (MRC, Cincinnati, OH, USA).The quantification of DNA was performed with the Quant-iT™ dsDNA HS kit (Invitrogen, Carlsbad, CA, USA) and fluorescence was measured in a Qubit Q32854 equipment (Invitrogen).
For the molecular identification of the three LAB isolates with probiotic potential, we amplified genes 23S and 5S of the ribosomal DNA.
PCR products were sent to be sequenced (CINVESTAV, Irapuato, Mexico) with internal oligonucleotides designed by Leyva-Madrigal et al. (2011), which amplify a fragment of 550 bp.
Similarity searched was performed against the GenBank database of the National Center for Biotechnology Information (NCBI), using the BLAST program (Basic Local Alignment Search Tool).The phylogenetic analyses were performed with the Molecular Evolutionary Genetics Analysis software (MEGA 5 Beta) (Tamura et al., 2011).Evolutionary relationships among sequences were inferred by using the neighbor-joining method (NJ) (Saitou and Nei, 1987).The robustness of the NJ topology was evaluated by bootstrap test using 1000 replicates.The Thermotoga maritima sequence was used as outgroup.

Preparation of experimental diets with additives
Inulin, LAB, and fulvic acid were sprayed on commercial diet (Camaronina ® , Purina, Mexico, 45% protein).Dry Oil ® (Innovaciones Acuícolas, S.A. de C.V., Culiacán, Mexico) was used as adhesive and feed attractant.Diet was dried at room temperature overnight and stored at 4°C in a refrigerator for 10 days (Apún-Molina et al., 2009).
The LAB doses managed in the bioassay were based on the works of Apún-Molina et al. (2009)

045
concentrations of fulvic acid were chosen empirically.The three selected isolates (LAB 1-6, LAB 35, and LAB 37) were mixed in the same proportions and then they were put in the diet.

Experimental design
Hormone-treated animals (males) weighing 1.3 ± 0.12 g were obtained from the farm DIBSA (Bacubirito, Sinaloa de Leyva, Sinaloa, Mexico) and acclimated for four days in two outdoor 1000-l plastic tanks with 800 l aerated fresh-water.To evaluate the effect of feed additives on growth performance of tilapia, the outdoor culture system was the same as for acclimation.The bioassay was conducted as a completely randomized design with four treatments in triplicate.The fish (12 organisms/tank) were fed ad libitum with floating pellets containing 45% protein.Water exchange (80%) and cleaning were performed weekly.Photoperiod was 12:12 h light: dark cycle.Values of temperature (HI 98127 pHep, Hanna Instruments, Woonsocket, RI, USA), pH (HI 98127 pHep, Hanna Instruments), and dissolved oxygen (YSI model 55 Oxygen meter, Yellow Spring Instruments, Yellow Springs, OH, USA) were determined weekly.Water samples for nitrite, nitrate, and ammonia determinations were analyzed at the beginning (first week of culture), middle, and end of each bioassay for all treatments following the Strickland and Parsons (1968) method.Growth in weight (g) was monitored twice a month by weighing each fish (12) of tanks in a balance (Scout Pro SP 601, Ohaus Corporation, Pine Brook, NJ, USA).Mortality was recorded daily.

Specific growth rate
The specific growth rate (SGR) was calculated based on the following formula (Ziaei-Nejad et al., 2006): Where, W 0 represents the initial body weight of tilapia fry, W represents the final body weight of tilapia and t represents time in days.

Statistical analysis
A one-way analysis of variance (ANOVA) was used to examine growth differences among treatments.When significant differences were found, Tukey's HSD test was used to identify the source of these differences (P < 0.05).

RESULTS
From 37 presumptive LAB isolates, 26 presented betahemolysis and 11 presented gamma hemolysis.Gamma hemolysis is displayed by bacteria that do not induce hemolysis of the blood cells.Isolates with beta hemolysis were not characterized because of their capacity to lyse human and tilapia erythrocytes.All presumptive LAB isolates were Gram (+), 11 were cocci, and four were bacilli.The cellular arrangement of LAB varied, some isolates presented an arrangement in pairs, others in chains, and others had a grape-like arrangement.LAB isolates with the highest probiotic potential were chosen according to the following criteria: gamma hemolysis, hydrophobicity, Gram (+), and to be derived from fish that presented the largest number of bacteria in the intestine at the time of isolation.
The selected LAB isolates (LAB 1.6, LAB 35, and LAB 37) did not inhibit the Vibrio sp.isolate; however, LAB has other beneficial characteristics.
Isolate LAB 35 had an exponential phase of approximately 66 h and for the rest it was of 46 to 48 h.Microorganisms are applied to the diet when the culture is in its exponential phase.
None of the selected isolates presented extracellular enzymatic activity (proteases and lipases); however, all LAB isolates were hydrophobic, which is an important feature as it indicates that the isolates had the capacity to adhere to the tilapia's intestine.
The number of CFU/ml of LAB ranged between 17.2 and 24.7 millions.This value served to add the microorganisms to the diet in adequate amounts.
The isolates LAB 1-6 and LAB 37 depicted 100% similitude (genetic similarity) with P. pentosaceus, reported in the GenBank database, and group together to form a monophyletic group.The isolate LAB 35 depicted 99.6% similitude with P. pentosaceus but did not form a monophyletic group.The isolate LAB 35 was identified only at genus (Pediococcus) level (Figure 1).
The physicochemical water parameters, recorded during the study, were within the recommended tolerance range for Nile tilapia (Jiazhao, 1991;Popma and Lovshin, 1996) with exception of ammonia (0.30 ± 0.3 -0.481 ± 0.4 mg/l) that was above the critical value (0.1).The final survival was 100% in all treatments.Fish fed with LAB (2.5 × 10 5 CFU/g of diet) plus inulin and fulvic acid (treatment III) grew significantly (P = 0.03) better than the control (Figure 2).Results demonstrate that additives influenced positively the growth of organisms in the concentrations of the treatment III.

DISCUSSION
The use of probiotics in aquaculture has intensified in the last years (Verschuere et al., 2000;Balcazar et al., 2008).
To select the microorganisms with probiotic potential, these must be isolated preferably from the same organism in which they are going to be tested and should be from the same region (Verschuere et al., 2000), since there are many commercial probiotics that are less suitable because they were isolated in other regions or countries.In this work, we obtained 37 presumptive LAB isolates from the intestine of O. niloticus for their character-rization as potential probiotics; the same species.
A first criterion to select isolates with probiotic potential was the analysis of hemolysis as a virulence factor, since many organisms are able to synthesize exotoxins that induce partial or total lysis of human or animal erythrocytes (Gildberg et al., 1995;Zamora-Rodríguez, 2003).In this study, LAB with alpha or beta hemolysis (hemolytic activity) was discarded and only those with gamma hemolysis (lacking hemolytic activity) were selected. 4CFU/g of diet) + 2.5 g inulin + 2.5 g fulvic acid/kg diet; III) LAB (2.5 × 10 5 CFU/g diet) + 2.5 g de inulin + 2.5 g fulvic acid/kg diet; IV) LAB (5 × 10 5 CFU/g of diet) + 2.5 g de inulin + 2.5 g fulvic acid/kg diet.Data are expressed as average ± SD (standard deviation).Different letters indicate significant differences (P < 0.05).
The presumptive LAB isolates with gamma hemolysis were Gram (+), mostly cocci, and very few were bacilli; their arrangement was in pairs, some in chains and others presented a grape-like arrangement.According to Axelsson (1998), in general, LABs can be characterized as Gram (+), anaerobic or facultative anaerobes, and rods or cocci that do not form spores.Besides they are oxidase negative, catalase negative, and benzidinenegative, lack cytochromes, do not reduce nitrites to nitrates, are gelatinase-negative, and are unable to use lactate.
The three selected LAB isolates with the best probiotic potential did not antagonize the used Vibrio isolate, which disagrees with one of the characteristics recommended for the selection of probiotics, that is competitive exclusion of noxious bacteria due to antibiotic production (bacteriocins, lysozymes, proteases, and H 2 O 2 ) (Bruno and Montville, 1993;Naidu et al.,1999;Bjorn et al., 2003).However, the probiotic effect can be provided by the pH decrease due to the production of lactic acid, acetic acid (Aguirre-Gúzmán, 1994), fatty acids, and other molecules (Midolo et al., 1995;Kao and Frazier, 1996).In addition, according to Bruno and Montville (1993), it is assumed that LAB could have other action mechanisms, such as improving water quality by degrading organic matter.
Establishing growth kinetics of the isolates is considered useful since it allows for strategies for the manipulation.It is important to know the exponential phase of the isolates because if Gram stain is performed on a bacterium in stationary phase or declining stage an erroneous result can be obtained (Prescott et al., 1999).This is because old cultures can lose their ability to retain the crystal violet complex characteristic of Gram (+) staining.
The criteria to select bacteria with probiotic potential vary depending on the study.Some authors state that the extracellular production of enzymes like proteases and lipases provide positive additives to the nutrition of the host (Moriarty, 1999;Balcazar et al., 2008), whereas other authors consider that the excessive production of these enzymes is a pathogenicity factor because pathogenic strains depict a high extracellular proteolytic, lipolytic, and hemolytic activity (Quesada-Herrera et al., 2004).Therefore, we performed the tests to determine these activities.However, none of the selected LAB presented proteases and lipases activity.The three LAB isolates showed a positive hydrophobicity, which means that the organisms are able to bind non-specifically to the intestinal epithelium through hydrophobic interactions.According to Rinkinen (2004), if these hydrophobic molecules did not exist on the surface of microorganisms, they could not attach to the epithelium, since both are negatively charged.However, it is important to mention that the hydrophobic interactions also favor adhesion and colonization of some pathogens.
The bioassay showed significant differences in growth between treatment III and the control group.There are no reports about the effect of LAB, inulin, and fulvic acid on growth of O. niloticus.However, in terms of probiotics, our results agree with those of Apún-Molina et al. (2009), who added lactococci to the diet, achieving a better growth than in the control without bacteria and similar survivals to those found in the control without probiotics.Likewise, a positive effect on growth was found in O. niloticus fed with diets supplemented with Lactobacillus acidophillus, Saccharomyces cerevisiae, Streptococcus faecium, Bacillus sp., and Lactobacillus sp.(Lara-Flores et al., 2003).Also, a positive effect on growth was observed in tilapia fed with a diet supplemented with Bacillus sp. and Lactobacillus sp.(El-Haroun et al., 2006) or with Bacillus sp., Lactobacillus sp., and Saccharomyces sp.(Guevara et al., 2003).Characterization and selection of isolates was adequate and the combination of Pediococcus sp., P. pentosaceus, fulvic acid, and inulin can be added to the diet to improve growth of O. niloticus.However, it is necessary to investigate the effect of the tested additives on the immune system of O. niloticus.
and Luna-González et al. (2013), whereas those of inulin were based on reports by Zhou et al. (2007) and Li et al. (2007) for Litopenaeus vannamei.The tested Cota-Gastélum et al.

Figure 1 .
Figure 1.Phylogenetic tree (neighbor-joining) for LABs of O. niloticus and different LAB sequences (GenBank Access numbers are indicated) derived from partial sequences of the gene 23S rRNA.Thermotoga maritima was used as outgroup.Number on nodes indicates the bootstrap levels based on 1000 repeats.Bar = Sequence divergence.