J. C:omp.


1992 Vol.

106. 255-265

Responses of Mucus-Producing Cells of Rainbow Trout (Oncorhynchus H. W. Ferguson

*, D. Morrison,

Fish Pathology Laboratory,

V. E. Ostland, P. Byrne

in Gill


mykiss) J. Lumsden


Department of PatholoQ, Ontario I’eterinary College. lInir:er.rilv Guelph, Ontario, Canada NIG 2M’l


Summary This paper documents the responsesof mucus-producing cells in the gills of rainbow trout (Oncorhynchus mykiss) throughout a naturally occurring outbreak

of bacterial gill disease (BGD) and following exposure to experimentally induced high concentrations of ammonia and suspendedsolids.The responses were examined at three sites on the gill filament with three histochemical stains selected to identify the main types of mucous glycoproteins; thesewere periodic acid-Schiff (PAS), alcian blue pH 2.5 (AB2) and alcian blue pH I.0 (ABl). In the BGD-infected fish, there was an increase in the numbers of PAS-positive and AB2-positive mucous cells and a corresponding decrease in ABl-positive cells. The greatest increase in mucus-producing cells occurred at the tips of the filaments, but the greatest relative change occurred at the midfilamental (inter-lamellar) position. Fish exposed to high ammonia concentrations also had elevated numbers of mucus-producing cells, but there was no statistically significant change in fish exposed to high amounts of kaolin. The

possibleimplications of thesefindings are discussed.

Introduction Teleost gills provide a mucous membrane responsible for gas exchange, as well as for ammonia excretion and ion regulation. Not all species have micro-ridges on the lamellar surfaces, but in those that do, the micro-ridges are generally thought to be covered by a layer of mucus (Randall, Lin and Wright, 1991). The existence of a mucous coating of the lamellar surface of normal fish had recently been challenged (Handy and Eddy, 1991), but the importance of the interface between the gill surface and the water, the “unstirred layer”, in helping to regulate carbon dioxide and ammonia excretion, is well accepted (Wright, Randall and Perry, 1989). Bacterial gill disease (BGD) is common in farmed salmonids and other intensively reared species of fish (Ostland, Stevenson and Ferguson, 1989) and is a major cause of loss in Ontario trout hatcheries (Daoust and Ferguson, 1983; Speare and Ferguson, 1989). Its infectious nature has recently been established by experimentally reproducing the disease with pure cultures of the bacterium Flavobacterium branchiophila (Kimura, Wakabayashi and Kudo, 1978; * To whom


002 I -9975/92/030255


+ 11 s03.00/0

be addressed C’I 1992 Academic

Press Limitrd


H. W. Ferguson

et al.

Ferguson, Ostland, Byrne and Lumsden, 199 1 1, but epidemiological data h;t\~t~ always strongly implicated unfavourable water quality as a predisposing C~;LIIS(~ (Bullock, 1972). Two factors commonly believed to be important arc high suspended solid and ammonia concentrations. One of the clinical signs of BGD and other forms of gill irritation and damage is mucus hypersecretion. Indeed, the quantities produced may be sufficient to create foaming of the water in which the fish are kept I,Ferguson, 1989). From a comparative standpoint, changes in the numbers of mucusproducing cells, and increased rates of production and enhanced viscosit) (If. mucus, are seen in pneumonic and enteric diseases of man and other mammals (Neutra and Forstner, 1987; Lundgren, 1990; Verdugo, 19901. Incrrasrcl production of mucus in fish may be a protective response of the mucosa that helps to clear, dilute or neutralize toxins or pathogens, and it may also reduce loss of ions (Lock and van Overbeeke, 1981). Serum ion depletion, notably of sodium and chloride, is one of the major pathogenetic mechanisms of BGD (Byrne, Ferguson, Lumsden and Ostland, 199 1) and of toxic exposure, as stv1 in “acid rain” (Wood and McDonald, 1987). Despite the probable protective function of branchial mucus, most attention has been focused on the skin and there is little detailed information on fish gills (Fletcher, -Jones and Reid, 1976; Lopez-Vidricro, Jones, Reid and Fletcher, 1980). There is a similar paucity of information about the extent and regional distribution of any increased secretion in diseased fish and on possible qualitative changes in the mucus. This study assesses the number and types of mucus-producing cells present in the gills of normal and diseased rainbow trout (Oncorhynchus mykiss). Fish were sampled throughout a naturally occurring outbreak of BGD and after experimental exposure to high concentrations of ammonia and suspended solids (kaolin).


and Methods


Fingerling (5 to 10cm) rainbow trout were held in 60 I wash-tub tanks containing flow-through aerated water at pH 7.8 and a temperature of 10°Cf 1°C. They were fed twice a day on a commercial pelleted diet. Fish in the ammonia experiment were from the same stock as those in the kaolin experiment, but the BGD fish were from a different stock. Sampling Prolocol

The BGD outbreak occurred spontaneously in our laboratory holding facilities. Diagnosiswas basedon clinical appearance, whole-mount examination and culture of’ affected gills, routine histopathology and electron microscopy of moribund animals (Speare, Ferguson, Beamish, Yager and Yamashiro, 1991). The baselinequality of the water used for these fish and the controls was the same as for the two experimental groups. Untreated water contained unionized ammonia levels of 0.001 mg per 1, and no suspended solids detectable by our methods (photometric methodology with Hach Kit DR-EL/4, Hach Co., Loveland, CO, U.S.A.). Fish were monitored and sampled histologically before any signs of clinical BGD were detected (day zero) and thereafter



in Gill



on days 3, 7, 9 and 11 throughout the entire course of the disease, with examination of 6 to 10 fish per sample. Morbidity rose from zero to over 80 per cent within 48 h. Mortality rates were 10 to 20 per cent by 48 h, but had declined by day 7 to 10 per cent. Healthy control fish from the same stock were sampled from a different tank in which BGD never occurred. In the ammonia experiment, fish were continuously exposed to unionized ammonia levels of 0.4 mg per 1 (in the form of ammonium chloride) for a period of 90 days :Daoust and Ferguson, 1984). F‘tve Iish were sampled at this time. Control fish were held in identical tanks but with no ammonia added. In the kaolin experiment, fish were exposed continuously to kaolin, approximately 7000 mg per 1, over a period of 64 days (Goldes, Ferguson, Moccia and Daoust, 1988). Six fish were sampled at this time. Control fish were held in identical tanks but with no kaolin.

Ti \.\ UP Pmming

and Staining

All fish were killed with tricaine methanesulfonate i MS-222) and the four gill arches from the left side removed and fixed for 24 h in Bouin’s fixative (Harleco Diagnostics. B.11.H. Chemicals, Toronto, Ontario, Canada). After routine processing and emhedding in paraffin wax, 4 to 6-pm-thick serial sections from each fish were stained with ( 1 ) periodic acid-Schiff (PAS), (2) alcian blue pH 2.5 (AB2) or (3) alcian blur pH 1.0 (ABl 1. These were selected to identify the glycoproteins as neutral. carboxvlatrd or sulphated mucins (Jones and Reid. 1978).



Ttte mucous cell counts were performed on stained sections examined with the MOP-3 machine (Carl Zeiss Inc., Ontario, Canada) and camera lucida. Three filaments were exitmined on each of the four arches from every fish; those selected were the crntrc filament on the arch, and a ventral and dorsal filament at a distance of 1.5 mm from the centre filament. Three sites on each filament were examined, namely (1) the distal end of‘the filament, (2) half-way down the filament, for examination of the cells of the inter-lamellar space and (3) the proximal inter-filamental spaces. For tht inter-lamellar and inter-filamental space sites, both sides of the filament were examined and the results averaged. An area of0.01 mm was measured at each site and the number of positively stained mucous cells recorded. .\\erages from the four arches for each fish were made for each site. Statistical anal.ysis of the data was carried out using SAS programs (Statistical Analysis Systems Institute’ Inc., Version 5.18, North Carolina, U.S.A.). ANOVA (analysis of variance, values wet-c computed on the data. An alpha \ralue of 0.05 was used for all statistical tests. \Wien significant, Tukey’s studentized range test was usrd to make paired c.ornparisons of means. \Yhen estimating the relative change in the number of mucous cells at each site, the numbers were expressed as a percentage increase or decrease compared with the t~)~rtr.ols ihr each of the three stains: for BGD, the estimate was made only on dav 1 1.

Results Control


Most of the mucous cells were found at the distal end of the filament. Numbers of mucous cells at the other sites (inter-lamellar spaces and inter-filamental spaces) were fewer but similar to each other. PAS gave the highest number of positive cells at all three sites. with AB2 next, followed bv ABl.


H. W. Ferguson


et al.

Infected Fish

Trends seen during the course of the disease included increases in the numbers of PAS-positive and AB2-positive mucous cells and a corresponding decrease in the number of ABl-positive cells. This indicated an increase in the number of cells containing neutral and carboxylated mucins, and a decrease in the number containing sulphated mucins. The greatest change in numbers of mucous cells occurred at the distal end of the filaments. Numbers of PAS-positive mucous cells here were significantly greater in fish at day 11 for all three sites (Fig. 1) than they were in the controls, indicating an increase in cells containing neutral mucins. AB at pH 2.5 will stain both carboxylated and sulphated mucins. Although not always significant, by day 11 there were increases in the number of AB2positive mucous cells at all three sites (Fig. 2). Comparison of the AB2-positive and ABl-positive mucous cells showed an increase in the number of cells containing carboxylated mucins and a decrease in the number containing sulphated mucins on days 7, 9 and 11 (Figs 2 and 3). Ammonia-exposed and Kaolin-exposed Fish In the ammonia-exposed fish, the greatest numbers of mucous cells were again found at the distal end of the filament. They had greater numbers of PASpositive mucous cells than the controls at all three sites (Fig. 4) and, although not always statistically significant, there was a trend towards greater numbers of AB2-positive mucous cells at all three sites (Fig. 5). This indicates an increase of cells containing neutral and carboxylated mucins in fish exposed to ammonia. The ANOVA values for ABI were not significant. No statistically significant differences were found between the controls and the fish exposed to kaolin.




Effect of time on numbers gill disease. •I Distal end

of periodic filament,


of days

acid-Schick-positive inter-lamellar



mucous cells in gills H inter-filament

of fish with space.




in Gill



Although the greatest increase was often at the tips of the filaments, the greatest relative increase (i.e. relative to the numbers of cells at time zero or in the controls) was at the inter-lamellar (mid-filament) location (Table 1). This was especially striking in the BGD fish, which showed a four-fold increase in numbers of mucous cells for all three stains.





Effect of time on numbers gill diseasc. W Distal end




Effect of time on number gill disease-. I@ Distal end


3 Number

7 of days

9 infected

of alcian blue pH 2.5-positivr of filament, 0 inter-lamellar


3 Number

7 of days

9 Infected

of alcian blue pH I,O-positive of filament, 0 inter-lamellar



mucous cells in gills spacr. q inter-flament


of fish with space.


of fish with spar?.



mucous cells in gills space. q intrr-filament


Fig. 4.

H. W. Ferguson

Numbers of periodic (mg per I) ammonia filament space.

acid-Schiff-positive 0.4 and kaolin



et al.

mucous cells in gills of control fish and l@l Distal end filament, 0 inter-lamellar


in fish exposed to space, q inter-





Numbers of alcian (mg per 1) ammonia filament space.

blue pH 2.5-positive mucws cells in gills of control fish and 0.4 and kaolin 7000. RI Distal end filament, 0 inter-lamellar

in fish exposed to spare, q inter-

Discussion This paper records hyperplasia of branchial mucous cells during the course of an outbreak of BGD, and after exposure to raised ammonia concentrations. It also demonstrates that there was a change in the type of mucus produced, but only in fish with BGD, not, for the most part, in ammonia-exposed fish. By contrast, high kaolin concentrations appeared to have no statistically significant effect on numbers or types of branchial mucous cells. Similar changes have been noted in catfish (Zctalurus nebulosus LeSueur) subjected to acid stress, but in this very different non-scaled species, only the skin was examined (Zuchelkowski, Pinkstaff and Hinton, 1985).







6 ; cl


5 ”


2 6 a z

0.8 0.6

4 0.4 E .+ 0.2



Number ammonia spacr.


of al&n blue pH l,O-positive cells in gills of control fish and Distal end filament, Cl intrr-lamellar 0.4 and kaolin 7000.

of three Change





seen in numbers

.,VH< PAS AB2 AB 1 BGD. 1.0.

57 72 69 bacterial



Table BGD)


of mucous

in fish exposed to (mg ptr space, q inter-filamtnt


1 on


of brancbial

CVILJ in 3 gill@ment



zn,bJh Jubjected


to thr .\tnkd,fnr


RGD - 30.4 -29 17.6 disease;


69.1 41.9 -71.6 periodic-acid

145 138 - 38.2 Schiff;

80.9 2.0 -91 AB?.

418 4’1-2 4.57 alcian


276 100 180

“51 218 -40 at pH




291 25f) 9.1 hluc

‘It pH

There is probably a poor correlation between the numbers of mucous cells seen in a histological preparation and the amount of mucus being produced. Indeed, many cells that have recently released their mucus may fail to stain until they have synthesized more. Moreover, this study ignores the question of’ the rate of mucus production and it is not unreasonable to suppose that the rate would increase in an irritated and inflamed tissue or in one that is subject to pronounced environmental pertubation, as happens in mammalian mucosae (Florey, 1970; Lundgren, 1990). If anything, therefore, the present values probably underestimate the overall tissue response. During the course of the BGD outbreak there was an increase, at all three sites examined, in the numbers of PAS-positive and AB2-positive mucous cells. Although the greatest increase was at the tips of the filaments, the greatest percentage increase (i.e. relative to the numbers of cells in the controls) was at the inter-lamellar (mid-filament) location. This corresponds with an overall increase in the thickness of the filament and probably has implications from the viewpoint of resistance to water flow. Moreover, hyperplasia of mucous


H. W. Ferguson et al.

cells at the lamellar base may alter ionoregulatory or gas exchange f’unctions owing to a space-occupying effect. This would be similar to the situation described by Thomas, Fievet, Claireaux and Motais (1988) for rainbow trout low in plasma Na+ and Cl-, in which there were increased numbers ofchloridc cells; these fish had reduced abilities to accommodate increased oxygen demand, or reduced availability of oxygen. The findings of filamental tip hyperplasia endorses the common gross and histopathological observation of’ gill “clubbing”; hyperplasia at this location in chronic branchitis may be due to increases in numbers of several ceil types, including eosinophilic granular cells, lymphocytes and structural epithelial cells, as well as mucus-producing cells. AB2 will distinguish between neutral and acid glycoproteins. Conversely, sialic acid-rich glycoproteins and sulphated sialic acid-rich glycoproteins will both stain positively with AB2. ABl, however, is specific for sulphated sialic acid-rich glycoproteins (Jones and Reid, 1978). During the course of BGD there was a decrease in the number of ABI-positive mucous cells. The use of both ABI and AB2 allows the conclusion that the increase in AB2-positive cells was due to an increased number of cells containing sialic acid-rich glycoproteins, since the reduction in ABl-positive cells rules out the possibility that the increase was due to sulphated sialic acid-rich glycoproteins. The quantitative variation in overall numbers of mucus-producing cells, and the shifts in histochemical staining probably represent two very important changes, but the biological significance of both must remain speculative, A shift from one type of glycoprotein to another may merely be a reflection of a change in the time of release or rate of production, some glycoproteins being easier to produce than others; any possible change in the viscosity of the mucus, which in mammals is normally associated with shifts in types of glycoproteins, may have different biological importance in an aqueous environment. Other components of the mucus, such as immunoglobulins and lysozyme, are probably more important from a protective point of view, and this study made no attempt to assessthem. One of the early effects of Flavobacterium branchiophila is degeneration and necrosis of the lamellar epithelium (Speare et al., 199 1) . An inevitable consequence of this would be leakage into the mucous layer of intracellular material and possibly also of blood components such as albumin. These would probably have a major effect on the homeostasis of the mucous layer, notably on its degree of hydration (Forstner, Jabbal, Findlay and Forstner, 1977). It has been suggested that mucus hypersecretion may be beneficial in slowing water and ion loss from the gills, but that it may also increase the diffusional resistance for gas transfer, thereby pre-disposing to hypoxia (Ultsch and Gros, 1979; Playle and Wood, 1989). However, an increased rate of mucus production would not necessarily lead to an increased depth of mucus at the lamellar surface, particularly if there were some change in the viscosity; indeed, the layer might be thinner, especially if an accompanying tachybranchia moved the mucus over the gill surface at a faster rate. We recently showed (Byrne et al., 1991) that fish with acute BGD have electrolyte imbalances and haemoconcentration, and we suggested that these are more



in Gill



critical in the pathogenesis of the disease than is hypoxia. By day 11 in the present study we saw increased numbers of PAS-positive and AB2-positive mucous cells. This fell within the chronic proliferative phase of BGD, in which hyperplasia, fusion and metaplastic changes would have replaced the more acute necrotic changes (Speare et al., 199 1). There is no information on the relative rates of ion and water loss, nor on the efficiency of oxygen uptake, in the acute and chronic phases of BGD, but it seems reasonable to suggest that ion loss would be reduced in the proliferative phase. Increased mucus secretion at this time, therefore, may or may not help to reduce any remaining tendency to lose electrolytes and water, depending on the actual thickness and composition of the mucous layer, but it may also alter the efficiency of oxygen uptake. The consequences of mucus hypersecretion in fish gills remain to be properly investigated. Compared with the controls, the fish exposed to high concentrations of unionized ammonia had increased numbers of mucous cells with all three stains and at the majority of sites. It used to be accepted that high concentrations of ammonia would, per se, produce hyperplastic branchial lesions (Smith and Piper, 1975). Several recent reports, however, which examined ammonia in isolation from other factors that normally accompany high ammonia in situations of heavy stocking, failed to confirm this (Mitchell and Cech, 1983; Daoust and Ferguson, 1984). The present findings of a relatively even increase in all types of glycoprotein production from ammonia-exposed gills provide some basis for linking high ammonia concentrations to enhanced susceptibility to gill disease. Aside from any physiological consequences of an altered mucous layer, which may indeed be profound (Wright et al., 1989), one possible consequence is an enhanced susceptibility to pathogens that have specific receptors for mucus. Hence, high ammonia concentrations, seen with high stocking densities, might predispose the gills to pathogens such as Flavobacterium b~~?~c~~~~~i~~ merely by altering the mucous coat. Acknowledgments The Fish Pathology Ministry of Agriculture Research Council.

Laboratory receives much of its funding and Food. J. Lumsden has a Fellowship

from the Ontario from the Medical

References Bullock, G. L. ( 1972). Studies of selected myxobacteria pathogenic for fishes and on bacterial gill disease in hatchery-reared salmonids. ‘Technical Papers of the Bureau of Sport Fisheries and WildltjL pp. l-30. Byrne, P., Ferguson, H. W., Lumsden, J. S. and Ostland, V. E. ( 1991). Blood chemistry of bacterial gill disease in brook trout Salvelinus fontinalis. Diseases q/’ Aquatic Organisms, 10, l-6. Daoust, P. Y. and Ferguson, H. W. (1983). Gill diseases of cultured salmonids in Ontario. Canadian Journal of Comparatir!e Medicine, 47, 358-362. Daoust, P. Y. and Ferguson, H. W. (1984). The pathology of chronic ammonia toxicity in rainbow trout. Journal of Fish Diseases, 7, 199-206. Ferguson, H. b%‘. (1989). Systemic Pathology of Fish: A text and atlas of comparative tissue responses in diseases of teleosts. Iowa State University Press, Ames.


H. W. Ferguson



Ferguson, H. W., Ostland, V. E., Byrne, P. and Lumsden, J. S. (1991 I, Experimental production ofbacterial gill disease in trout by horizontal transmission and by bath challenge. Journal of Aquatic Animal Health, 3, I 18-l 23. Fletcher, T. C., Jones, R. and Reid, L. (1976). Identification ofglycoproteins in goblet cells of epidermis and gill of plaice (Pleuronectes platessa I~.), flounder (Platichthw Jesus L.) and rainbow trout (S’almo gairdneri Richardson). Historhemical jrournal, 8,

587-608. Florey, H. W’. (197Oj. The secretion of mucus and inflammation of mucous membranes. III General Pathology, 4th Edit. H. W. Florey, Ed., W. B. Saunders, London, pp. 195-225. Forstner, .J. F., Jahbal, I., Findlay, B. and Forstner, G. G. (1977). Interactions of mucins with Ca, H’ ions and albumin. In Mucus Secretions and Cystic Fibrosis (Mod. Probl. Pediatr., 19). Karger, Basle, pp. 4-65. Goldes, S. A., Ferguson, H. W., Moccia, R. D. and Daoust, P. Y. ( 1988). Histological effects of the inert suspended clay kaolin on the gills of juvenile rainbow trout. Salmo gairdneri Richardson. Journal of Fish Diseases, 11, 23-33. Handy, R. D. and Eddy, R. B. (1991). Th e absence of mucus on the secondary lamellae of unstressed rainbow trout, Oncor-hynchus n~ykiss (Walbaum). journal of Fish Biology, 38, 153-155. in health and Jones, R. and Reid, L. (1978). S ecretory cells and their glycoproteins disease. British Medical Bulletin, 34, 9916. Kimura, N., Wakabayashi, H. and Kudo, S. (1978). Studies on bacterial gill disease in salmonids-I. Selection of bacterium transmitting gill disease. Fish Pathology. 12,

233-242. Lock,

R. A. C. and van Overbeeke, A. P. ( 1981). Effects of mercuric chloride and methylmercuric chloride on mucus secretion in rainbow trout, Salmo gairdneri Richardson. Journal of Comparative Biochemis@ and Physiology, 69C, 67-73. Lopez-Vidriero, M. T., Jones, R., Reid, L. and Fletcher, T. C. (1980). Analysis of skin mucus of plaice (Pleuronectes platessa L.). Journal of Comparative Pathology, 90, 415-

420. Lundgren, J. D. (1990). Pathogenesis of airway mucus hypersecretion. Journal o/ Allergy and Clinical Immunology, 85, 3999419. Mitchell, S. J. and Cech Jr, J. J. (1983). Ammonia-caused gill damage in channel catfish (Ictaluruspunctatus): Confounding effects of residual chlorine. Aquatic Science,

40,242-247. Neutra, M. R. and Forstner, J. F. (1987). Gastrointestinal mucus: synthesis, secretion. and function. In Physiology of the Gastrointestinal Tract, 2nd Edit. L. R.

Responses of mucus-producing cells in gill disease of rainbow trout (Oncorhynchus mykiss).

This paper documents the responses of mucus-producing cells in the gills of rainbow trout (Oncorhynchus mykiss) throughout a naturally occurring outbr...
1MB Sizes 0 Downloads 0 Views