Field Observations on the Effect of a Mannan Oligosaccharide on Mortality and Intestinal Integrity of Sole (Solea senegalensis, Kaup) Infected by Photobacterium damselae subsp. piscicida

The commercial production of sole currently takes place mainly in Spain and Portugal with Solea senegalensis being the preferred species. Production levels of this species increased from 110 tons in 2008 to 500 tons per year in 2010 [1]. However disease control and feed related problems such as poor growth rates and feed conversion ratios (FCR’s) are the main factors affecting production performance and these are critical factors to the commercial success of any aquaculture business.


Introduction
The commercial production of sole currently takes place mainly in Spain and Portugal with Solea senegalensis being the preferred species. Production levels of this species increased from 110 tons in 2008 to 500 tons per year in 2010 [1]. However disease control and feed related problems such as poor growth rates and feed conversion ratios (FCR's) are the main factors affecting production performance and these are critical factors to the commercial success of any aquaculture business.
The gastrointestinal (GI) tract has been recognized as one of the major routes of infection in fish [2,3].Understanding of the GI structure and its function and how it interacts with infectious diseases, anti-nutritional dietary components or stress in fish is a major factor in improving feed utilisation and protecting fish health in applied aquaculture practices.
There have been several approaches to disease prevention in aquaculture. The development and use of vaccines has expanded over the years both with regard to fish species and microbial diseases and vaccination is considered an effective tool in a farms strategy for disease control. However in the culture of new species the availability of a functional vaccine for the disease encountered may not be a realistic option and vaccines developed for other species specific pathogens may not be applicable until further development work is undertaken.
In addition to vaccination there has been a major initiative towards the appraisal of functional dietary supplements in commercial feeds for fish and crustacean species not only for their role in improving disease resistance but for improving nutrient utilisation. Many of these types of products are extracted from yeast [4,5]. Mannan oligosaccharide (MOS) has been shown to be an effective tool and works in several different ways within the digestive tract of poultry and swine [6][7][8] and more recently in aquaculture [9][10][11][12].
In order for an infection to occur pathogenic bacteria have to adhere to the enterocyte cell wall in the intestine [13] and this is mediated by their interaction with carbohydrates present on the cell surface [14]. Dietary MOS has been used to bind pathogens with a mannose -specific Type 1 fimbrae therefore preventing the bound bacteria from colonizing the intestine and allowing them to pass through the intestine and be excreted [15,16]. This has resulted in the use of MOS to reduce the incidence and severity of infections of salmonella [17] and Pasteurella multocida in broilers [18].
MOS has also been shown to bind to a number of bacterial strains, including Aeromonas spp., Vibrio spp. and Photobacterium damselae subsp. piscicida (#17911), known to cause disease in fish and shrimp [Dr Newman, Venture Laboratories, personal communication]. Torrecillas et al. [11,19] demonstrated that dietary MOS supplementation reduced infectionby reducing pathogen translocation across the intestinal tract of Vibrio alginolyticus and L. anguilarum in European sea bass (Dicentrarchus labrax) when challenged via exposure by anal canalization or cohabitation.
In addition MOS has been shown to improve gut function and health by increasing the villi height, uniformity and integrity [9,10,[20][21][22][23][24]. As a result, the feed within the digestive tract can be more efficiently digested, leading to superior nutrient absorption and improved growth [25][26][27]. Furthermore it has been stated that MOS can modulate the intestinal and systemic immune systems, by acting as a non-pathogenic microbial type antigen with an adjuvant like effect [28].
This study took place at a commercial farm where sole have been susceptible to infection with pasteurellosis in the past and this has led to considerable mortality and reduced performance.The aim of the study therefore was to investigate the prophylactic effect of MOS on the intestinal morphologyand fish health status. The impact of the use of an experimental vaccine for pasteurellosis, a common Gram negative bacterial disease which is caused by Photobacterium damselae subsp. piscicida (former Pasteurella piscicida), on the control and MOS treated fish was also investigated.

Animals and experimental design
The experiment was conducted in Spain under commercial conditions. Twelve round plastic tanks of 3m diameter and 2m 3 water volume, were supplied with ambient sea water, filtered to approximately 10 μm, and renewed two times per hour.
Fish for this trial were acclimatised in the tanks (850 fish per tank of average weight 45.0g) and the fish in half the tanks vaccinated using an experimental vaccine manufactured by HIPRA (AMER, Spain), which targets Pasteurella spp. and Vibrio spp. The dose was injected (endoperitoneal) according to manufacturer instructions. The fish in all tanks were fed the control diet prior to the start of the trial 14 days later.
Fish were subjected to one of four different treatment regimes, in triplicate, for a period of 10 weeks. During the trial, mortality was recorded daily and fish weighed at the start and end of the trial and weekly in-between. Mean water temperature was 18 0 C at the start of the trial and this gradually increased to 26 0C by the end of the trial. A natural photoperiod was used throughout the trial period (14 hr light: 10 dark hr at the start and 15:9 by the end of the trial).

Dietary treatments
All fish were fed using the same standard commercial diet (Skretting LE Europa 18%, 2mm) (Table1) with or without the addition of 0.4% MOS (Bio-Mos® Alltech Inc.).
A week after the start of the trial, as the water temperature increased to 19 -21 o C, an outbreak of pasteurellosis occurred. This was caused by Photobacterium damselae subsp. piscicida and was confirmed and verified in a representative number of moribund fish from all treatment groups by the veterinary laboratory of Skretting Spain (TrouwEspaña, Cojobar, 09620 Burgos). Further inlet water samples also confirmed the presence of this opportunistic pathogen in the water supply to the farm.
The feeding rate was reduced from 1.5% of body weight at the start of the trial to 0.8% during the disease outbreak due to the loss of fish appetite during this period.

Data collection
At the end of the trial period, intestinal samples (4 fish per experimental tank, 12 fish per treatment) were retained for histological examination by light and scanning electron microscopy (LM and SEM). Anterior and posterior intestinal sections were examined. Anterior intestinal samples were excised approximately 1 cm after the stomach and the posterior samples were excised approximately 1 cm before the anus. Material for LM was fixed in 4% saline formalin, embedded in paraffin wax and cut into 8 μm thickness sections. The sections were stained using Mallory's Trichrome staining technique as described by Handy et al. [29] and screened under low power magnification microscope (Olympus SZH-ILLD) with an attached digital camera (Nikon CoolPix 990).
Samples for the SEM were rinsed in 1% S-carboxymethyl-L-cysteine for 30 s in order to remove epithelial mucus and then fixed in 2.5% glutaraldehyde with 0.1 M cacodylate buffer and 1% NaCl, and adjusted to pH 7.2. The samples were treated using an Emitech K850 critical point drier, with ethanol as the intermediate fluid and CO 2 as the transition fluid. All samples were covered with gold in Emitech K550 sputter coater and were observed with a Jeol JSM 5600 LV electron microscope at 15 kV.

Analysis
Digital images were taken and analysed using appropriate software (ImageJ 1.34, National Institutes of Health, USA). Each tissue sample was analysed using 3 -5 images. Images from the light microscopy were analysed to determine the perimeter ratio (PR) between the internal perimeter (IP) of the intestinal lumen and the external perimeter (EP) of the intestine (PR = IP/EP, arbitrary units). A high PR value indicates high villi length, increased mucosal folding or both. SEM images were used to calculate the density of the microvilli structures (MD) of the enterocytes on top of the villi by calculating the ratio between the microvilli (foreground, FG) and the background (BG) (MD = FG/BG). Two-way ANOVA (SPSS 15.0; SPSS Inc., USA) was used for the statistical analysis of the data.Significance was accepted at the P<0.05 level.

Results and Discussion
The initial stocking number, final number and cumulative mortality figures are shown in table 2. The cumulative mortality figures were high due to the natural disease outbreak of pasteurellosis, one week after the start of the trial. MOS treated fish without vaccination (MOS) had the lowest mortality at 47.78%, followed by 53.22% in the MOS+ vaccination (MOS-V) group, 55.952% in the control fed (CON) group and 61.58% in the control feed + vaccination (CON-V) group. Statistical analysis showed that there was a significant effect (P = 0.050) of MOS supplementation on the total mortality which tended to be lower (by ca. 8%) in MOS treated fish compared to the control. The effect of vaccination   The increased mortalities observed in the vaccinated fish both from the control (CON -V) and the MOS group (MOS-V) is possibly due to the stress of vaccination combined with exposure to a disease challenge from the natural environment within three weeks of vaccination before optimal immunity could be developed. Endoperitoneal vaccination has been reported to be a stressful procedure and should be avoided in cases of diseased fish [30]. This is especially true with juvenile flatfish vaccination, such as sole, as it is an exceptionally difficult husbandry procedure due to fish morphology. Previous studies have reported that MOS enhances the survival of shrimp and trout respectively under non-challenged conditions [27,31]. In the present study it was observed that MOS reduced the negative effect that vaccination had on the mortality.
In the anterior and the posterior intestinal region MOS supplementation displayed a significant increase of PR (P = 0.016 and P = 0.002, respectively; Table 3 and Figures 1 and 2     More extensive and longer villi structures of the intestine would support the hypothesis of better nutrient uptake and utilization [32,33]. SEM was used to determine the density of the microvilli. The results, in Table  3 and Figure 3, showed that in the anterior region of the intestineof fish fed MOS supplementation alone and vaccination had a significant effect on the microvilli density (P = 0.028 and P = 0.001, respectively). There was an insignificant interaction effect between the MOS supplementation and the vaccination on the anterior intestinal region (P = 0.960). The microvilli density increased from 2.11 ± 0.60 AU in the CON group to 2.71 ± 0.57 AU in the MOS group, 3.03 ± 0.69 AU in the CON-V group and 3.66 ± 1.11 AU in the MOS-V group. In the posterior intestinal region neither MOS supplementation nor vaccination significantly (P = 0.188 and P = 0.371 respectively) affected the condition of the microvilli under these conditions.
Fish internal organs such as liver, kidney and spleen in sole (Solea senegallensis), sea bream (Sparus aurata) and turbot (Scophthalmus maximus) have been examined histologically to investigate the effect of pasteurellosis [34,35] however there is limited information on the effect of pasteurellosis on the GI tract.
The virulence of Gram negative pathogenic bacteria is linked to its serotype and this is related to the surface components of the bacteria such as lipopolysaccharides [43] and may reflect the ability of the bacterial surface antigens to interact with the host's tissues.The main molecule involved in antigen recognition and the binding process in antigen presenting cells is the mannose receptor [44]. Indeed, Torrecillas et al. [19] demonstrated that MOS can reduce pathogen translocation across the GI tract. The improved intestinal condition demonstrated in this study, supported by evidence from other work [9,10,12],indicates that MOS hinders the development of pathogenic infection. Therefore the reduced mortality observed in the MOS fed groups may have been induced by a reduction of infection (via the GI tract) rather than a reduction in mortality of infected fish.
Bacterial diseases often affect the morphology of the posterior intestine by altering its physiological condition, i.e. necrosis and sloughing of the intestinal mucosa [45,46]. However, during the healing process, the intestine will return to its normal condition and structure. In the posterior intestine, all treatments produced significantly higher PR compared to the control group, indicating an improved overall intestine condition. Qualitative analysis of the LM sections of the posterior intestine indicated that the MOS and MOS-V treated fish exhibited an elevated number of supranuclear vacuoles in the enterocytes in comparison to the CON and CON-V treated fish. The presence of such vacuoles has been related with the absorption of lipids, proteins and polypeptides in the intestine, leading to a superior nutrient utilization [47]. The observed histological changes suggest that MOS may work positively to affect the recovery process in the posterior intestine however further quantitative analysis would be required.
Overall, the results from this study indicate that MOS affects both the anterior and posterior intestinal regions of sole grown under commercial production conditions. It appears that the anterior part of the intestine is more responsive than the posterior region with respect to morphological features. MOS appears to confer a positive modulatory effect on stressed and diseased sole and possesses recovery properties against Gram negative bacterial infections encountered under culture conditions. The use of prebiotics, such as MOS, offer the commercial farmer nutritional strategies, in addition to vaccination, that can help recovery from disease and improve and promote better fish health. However the effect of MOS on disease pathogens needs to be further assessed under controlled challenge conditions to confirm the findings here at a commercial farm. Photobacterium damselae subsp. piscicida has been shown to infect yellowtail (Seriola quinqueradiata) and snake-head fish (Channa maculata Lacepede), through the oral route, via the GI tract [36][37][38]. There is evidence that Photobacterium damselae subsp. piscicida strongly binds to the intestinal mucosa [39]. Naka et al. [40] demonstrated that Photobacterium damselae subsp. piscicida has strong adherence to the fish intestine possibly due to the mannose sensitive hemaglutinin (mshA) on the bacterial cell. Infection with Photobacterium damselae subsp. piscicida has been shown to cause damage in the GI tract of European sea bass [41] and GI inflammation as evidenced by elevated expression of pro-inflammatory (TNF-α and IL-1β) genes [42].