Electron microscopy: an important tool to assess the effects of dietary

components on the gastrointestinal tract of fish

G. M. Harper1, A. Dimitroglou, E. Ring2 and D. L. Merrifield3

1

2Norwegian

Electron Microscopy Centre

The University of Plymouth

College of Fishery Science, Faculty of Biosciences, Fisheries and Economics,

University of Troms, Norway.
3The

School of Biomedical and Biological Sciences

The University of Plymouth
Plymouth, UK
Daniel.merrifield@plymouth.ac.uk

The primary purpose of the alimentary tract is to digest foodstuffs into molecules suitable for absorption
via the various transport mechanisms of the epithelial border cells of the gastrointestinal (GI) sections.
In essence, the fish GI tract is a tube-like structure that varies in complexity depending on the various
feeding habits of the species. It can be subdivided into the mouth, oesophagus, stomach, pyloric ceaca,
mid intestine, distal intestine and the rectum; however, not all of these distinctive regions are present in
all fish. Some fish do not have a discernable stomach and those that do may have a U- shaped or a Yshaped stomach. Distal to the stomach, in the mid gut region, some fish have pyloric ceaca. These
finger-like pouches can range in number from a few to several hundred. Distal to the pyloric ceaca is
the intestine which also varies in physiology across fish species. Typically, the GI tract of carnivorous
fish species is short, often less than one body length long, whereas the GI tract of herbivorous fish
species is much longer and can be found to be greater than 20 times the body length. Distinctive
physiological differences are also observed between fresh water and marine species. The surface cells
exposed to the luminal contents are the enterocyte cells, which with their microvilli structures, comprise
the epithelial brush border. Between the epithelial cells goblet cells excrete a continuous layer of mucus
which forms an effective protective barrier. Beneath the inner epithelium we find the cellular connective
tissues which comprise the lamina propria and the submucosa. Beyond the submucosa lie a circular
muscular layer and a longitudinal muscular layer. Finally, the serosa forms the outer layer of the GI
tract.
It is thought that the GI tract of fish is a portal of entry for some fish pathogen sand indigenous gut
bacteria (Ring et al. 2007; Figure 1).Indeed, several studies have observed that exposure of the
epithelium to fish pathogens can result in severe tissue damage, characterised by necrotic enterocytes,
deteriorated tight junctions (proteinacious complexes which bind adjacent enterocytes), disorganised
and damaged microvilli and damage to the lamina propria (Ring et al. 2007). An example of the
damage that can be visualised at the ultra structural scale with transmission electron microscopy (TEM)
is shown in Figure 2, where the mid gut of rainbow trout has been exposed to Vibrio (Listonella)
aguillarum, the causative agent of vibriosis.
Fish have evolved an effective range of protective mechanisms to hinder pathogenic colonisation,
translocation and ultimately infection of GI mucosa. These mechanisms include gastric acidity, the
secretion of mucus, the acidic microenvironment of the apical brush border, cellular turnover, peristalsis
and the gut associated lymphoid tissues (GALT). In fish, the level of organisation of the GALT is lower

than that of mammals but the many diffuse lymphoid cells, macrophages, granulocytes and mucus IgM
constitute the intestinal immune function of fish (Dimitroglou et al. 2011). Examples of intraepithelial
immune cells are displayed in Figure3.
This mucosal layer, in conjunction with the associated organs (e.g. the pancreas, liver, and gall
bladder) and tissues, is central to the functions of digestion, osmoregulation, immunity, GI endocrine
regulation and maintaining an effective barrier against external pathogens and contaminants. This
complex system can be easily disrupted by a range of environmental factors and several studies have
shown that the gut of fish, in terms of structure and function, is sensitive to dietary changes. In recent
years we have used electron microscopy to observe the effects of various feed components on the GI
tract of fish.
Concerted efforts have been made to assess alternative protein sources, such as soybean meal (SBM)
and krill meal,to improve aquaculture sustainability and to meet the ever growing volume of protein
required for use in aquafeeds. We have observed that both of these alternative meal products can
affect the epithelial mucosa of fish in terms of structure and/or mucosal-microbial interactions (Ring et
al. 2006; Merrifield et al. 2009). Recently we observed the effect of replacing 50% of fish meal protein
with SBM on the intestinal epithelium and autochthonous (epithelium associated) microbial communities
of rainbow trout Oncorhynchus mykiss (Merrifield et al. 2009). After 16 weeks of feeding on diets
containing 46% SBM, electron microscopy revealed an enteritis-like effect associated with the fish fed
SBM. Specifically, SBM fed fish displayed missing, damaged, deformed, shorter and thicker microvilli.
Additionally, enterocytes appeared to be malformed and irregular (Figure 4a).The net result of this was
significantly shorter and less densely packed microvilli on the enterocyte surfaces of fish fed the SBM
diet than fish fed the fishmeal diet (Figure 4b). This reduction of microvilli density consequently led to
increased exposure of enterocyte tight junctions which (Figure 4a), combined with necrotic enterocytes,
is likely to diminish the protective barrier of the intestinal epithelium. Such enteritis effects have
routinely been described in fish with the use of light microscopy but to our knowledge this was the first
study to assess the impact using electron microscopy to visualise brush border damage. It is generally
accepted that this type of damage is due to the presence of various anti nutritional factors present in
SBM. For example, soy saponins, which disrupt the intestinal barrier by altering membrane
permeability. These histological and morphological changes which result in damaged enterocytes and
exposed intracellular tight junctions may help explain previous observation of increased susceptibility of
SBM fed fish to pathogenic infection (Krogdahl et al. 2000).
Furthermore, we observed that dietary SBM might also affect the microbial populations in close
association with the rainbow trout intestinal epithelium (Merrifield et al. 2009). For instance, the
proportion of yeast recovered from the anterior intestinal mucosa of the SBM fed fish was far greater
(39% of the cultured isolates identified) than that of the fishmeal fed fish (8.7%). We suggested that this
might a direct result of fermentable carbohydrates provided by SBM. Oligosaccharides, such
asstachyose and raffinose, typically constitute about 45% of SBM by dry weight. Raffinose and
stachyose consist of fructose, glucose and galactose. Many genera of yeast are able to ferment various
sugars, including glucose and galactose; hence, an increase in yeast numbers may be a result of
increased available sugars. The disruption of the complex microbe-host interactions within the GI tract
may have significant consequences to the host because not all microbes represent a pathogenic threat.
The resident commensal microbes help maintain efficient functioning of the gut by supporting gut
mucosal barrier function and maintaining mucosal tolerance (immuno non-responsiveness) to luminal
contents which allow for nutrient absorption (Merrifield et al. 2010a; Sanchez et al. 2010). These
microbial populations also provide a competitive barrier mechanism against transient and opportunistic
pathogens within the gut and also aid digestive function via the production of enzymes and vitamins.

Contrary to the observations in rainbow trout fed SBM rich diets, electron microscopical analysis of the
intestinal tract of gilthead sea bream (Sparus aurata) revealed that high dietary SBM did not have
detrimental effects on epithelial morphology and integrity (Dimitroglou et al. 2010a). These observations
are in agreement with light microscopy studies which indicate a more SBM-tolerant intestinal response
to dietary SBM in omnivorous and herbivorous fish species. Likewise, the gut microbial populations of
sea bream also appeared to be less sensitive to dietary SBM.
Some feed components on the other hand may be able to beneficially affect the morphology of the fish
GI tract. Our recent studies have shown that some dietary feed additives can improve brush border
integrity by enhancing microvilli morphology (Dimitroglou et al. 2009; Dimitroglou et al. 2010a; Merrifield
et al. 2010b). For instance, we have observed that dietary mannan oligosaccharides (MOS) can
improve sub adult rainbow trout (ca. 120 g) intestinal microvilli morphology when fed dietary levels of
0.2% for 58 days (Dimitroglou et al. 2009). Scanning electron microscopy (SEM) images showed that
the microvilli density of the posterior intestine in the sub adult MOS fed fish was significantly increased
compared with the control fed fish. Additionally, TEM revealed that MOS increased the microvilli length
in both the anterior and posterior intestinal regions (Figure 5). Similar beneficial observations have also
been seen in seen in gilthead sea bream (Dimitroglou et al. 2010a) and white sea bream Diplodus
sargus (Dimitroglou et al. 2010b). Such benefits are not only confined to prebiotics such as MOS. In a
recent study we observed that some commercially available probiotics could improve rainbow trout
microvilli morphology and enterocyte endocytic activity while some others could not (Merrifield et al.
2010b; Figure 6).The difference observed in the microvilli morphology was linked to differences
observed in the levels of close association of the various probiotics with the mucosal brush border.
SEM revealed that only the Pediococcus acidilactici probiotic (Bactocell) was observed to be in close
association with the mucosa, the other probiotics tested appeared to populate the mucus layer without
populating the epithelium itself. Subsequently only P. acidilactici fed fish displayed improved brush
border morphology.
These studies highlight the sensitivity of the GI tract of fish to various feed components. Given the
multiple important roles that this organ and its associated tissues play in digestive function, disease
resistance and immunity, it is important that future studies assessing novel feed ingredients consider
assessing gut morphology. Electron microscopy is an important technique which should be utilised in
such studies in order to assess the impacts at the brush border and ultra structural levels.

References
Dimitroglou, A., Merrifield, D. L.,Moate, R., Davies, S.J., Spring, P., Sweetman, J., Bradley, G., (2009)
Dietary mannan oligosaccharide supplementation modulates intestinal microbial ecology and improves
gut morphology of rainbow trout, Oncorhynchus mykiss (Walbaum). Journal ofAnimal Science 87,
3226-3234.
Dimitroglou, A., Merrifield, D. L., Spring, P., Sweetman, J., Moate, R. & Davies, S. J., (2010) Effects of
dietary mannan oligosaccharides (MOS) and soybean meal on growth performance, feed utilisation,
intestinal histology and gut microbiota of gilthead sea bream (Sparus aurata). Aquaculture, 300, 182188.
Dimitroglou, A., Davies, S. J., Sweetman, J., Divanach, P., Chatzifotis, S. (2010b)Dietary
supplementation of mannan oligosaccharide on white sea bream (Diplodus sargus L.) larvae: effects on
development, gut morphology and salinity tolerance. AquacultureResearch, 41, e245-e251.

Dimitroglou, A., Merrifield, D. L., Carnevali, O., Picchietti, S., Avella, M., Daniels, C. L., Groy, D.,
Davies, S. J. (2011) Microbial manipulations to improve fish health and production - a Mediterranean
perspective. Fish and Shellfish Immunology, 30, 1-16.
Ring, E., Myklebust, R., Mayhew, T.M., Olsen, R.E. (2007). Bacterial translocation and
pathogenesis in the digestive tract of larvae and fry. Aquaculture 268, 251264.
Ring, E., Sperstad, S., Myklebust, R., Mayhew, T. M., Mjelde, A., Melle, W., Olsen, R. E. (2006) The
effect of dietary krill supplementation on epithelium-associated bacteria in the hindgut of Atlantic
salmon (Salmo salar L.): a microbial and electron microscopical study. AquacultureResearch 37, 16441653.
Krogdahl, A., Bakke-McKellep, A. M., Red, K. H., Baeverfjord, G. (2000) Feeding Atlantic salmon
Salmo salar L. soybean products: effects on disease resistance (furunculosis), and lysozyme and IgM
levels in the intestinal mucosa. Aquaculture Nutrition 6, 7784.
Merrifield, D. L., Bradley, G., Baker, R. T. M., Dimitroglou, A. & Davies, S. J. (2009) Soybean meal
alters autochthonous microbial populations, microvilli morphology and compromises intestinal
enterocyte integrity of rainbow trout (Oncorhynchus mykiss Walbaum). Journal Fish Diseases 32, 755766.
Merrifield, D. L., Dimitroglou, A., Foey, A., Davies, S. J., Baker, R. T. M., Bgwald, J., Castex, M. &
Ring, E., (2010a) The current status and future focus of probiotic and prebiotic applications for
salmonids. Aquaculture, 302, 1-18.
Merrifield, D. L., Harper, G. M., Dimitroglou, A., Ring, E. & Davies, S. J. (2010b) Possible influence of
probiotic adhesion to intestinal mucosa on the activity and morphology of rainbow trout (Oncorhynchus
mykiss) enterocytes. Aquaculture Research, 41, 1268-1272.
Prez, T., Balczar, J. L., Ruiz-Zarzuela, I., Halaihel, N., Vendrell, D., de Blas, I., Mzquiz, J. L.(2010)
Host-microbiota interactions within the fish intestinal ecosystem. Mucosal Immunoligy 3:355-360.
Figure legends
Figure 1. Transmission electron micrograph of enterocytes in the foregut of spotted wolffish
(Anarhichas minor Olafsen) fry. Bacterial profiles (arrows) are seen at various depths within
enterocytes. Intraepithelial lymphocytes (IEL) are seen between enterocytes near the basal lamina. mv
microvilli, N nuclei, LP lamina propria, V vacuole. Bar 2 m. From Ring et al. (2007).

Figure 2. Transmission electron micrograph of the anterior intestine (mid gut) of rainbow trout exposed
to Vibrio (Listonella) anguillarum (VA). Clear signs of tissue damage are characterised by necrotic
enterocytes (NE), disorganised microvilli and irregular tight junctional complexes (TJ) resulting in
intracellular spaces (*). E = enterocyte. Scale bar = 5 m.

Figure 3. Transmission electron micrographs of immunological cells present in the epithelium of the
tilapia intestine. M = macrophage; L = lymphocyte; LP = lamina propria; = Ep epithelium. (Harper and
Merrifield unpublished data). The authors thank Prof. S. Picchietti for aiding with identification of cells.
Scale bars = 5 m.
Figure 4. Scanning electron microscopy of the epithelial brushborder of rainbow trout fed SBM rich
diets for 12 weeks. (Merrifield et al. 2009).SBM fed fish display irregular enterocyte formations with
sparse and irregular microvilli which consequently exposed tight junctions between the enterocytes (*)
to luminal contents. Scale bars = 5 m (A) and 1 m (B).
Figure 5.Comparative transmission electron micrographs of the posterior region of sub adult rainbow
trout fed diets containing 0.2% mannan oligosaccharide (MOS; A) or diets void of MOS(B). Although
microvilli appear healthy in both treatments, microvilli are more regular and significantly longer in MOS
fed fish (Dimitroglou et al. 2009). Bars = 1 m.
Figure 6. TEM micrographs reveal that although microvilli appear healthy in both groups they are
significantly longer in trout fed probiotic P. acidilactici (B) than the control (A). Additional, elevated
endocytic activity (*) was observed in P. acidilactici fed fish. Scale bars = 1 m (A) and 2 m (B).
(Merrifield et al. 2010b).