Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The functions and endocrine control of epithelial mucus secretion in the family Cottidae Marshall, William Smithson 1977

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1977_A1 M37_8.pdf [ 12.37MB ]
Metadata
JSON: 831-1.0094116.json
JSON-LD: 831-1.0094116-ld.json
RDF/XML (Pretty): 831-1.0094116-rdf.xml
RDF/JSON: 831-1.0094116-rdf.json
Turtle: 831-1.0094116-turtle.txt
N-Triples: 831-1.0094116-rdf-ntriples.txt
Original Record: 831-1.0094116-source.json
Full Text
831-1.0094116-fulltext.txt
Citation
831-1.0094116.ris

Full Text

THE FUNCTIONS AND ENDOCRINE CONTROL OF EPITHELIAL MUCUS SECRETION IN THE FAMILY COTTIDAE by WILLIAM SMITHSON MARSHALL B.Sc. (Hon.), Acadia University, 1973 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Zoology) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August, 1977 © William Smithson Marshall, 1977 In presenting th i s thes i s in pa r t i a l fu l f i lment of the r e q u i r e m e n t s f o r an advanced degree at the Univers i ty of B r i t i s h Co lumb ia , I ag ree that the L ibrary sha l l make it f ree ly ava i lab le for r e f e r e n c e and s tudy . I further agree that permission for extensive copying o f t h i s t h e s i s for scho lar ly purposes may be granted by the Head o f my Department o r by his representat ives. It is understood that c o p y i n g o r p u b l i c a t i o n o f th i s thes is for f inanc ia l gain sha l l not be allowed without my written permission. Department of The Univers i ty of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 D a t e ^ d ^ / /¥/77 i i ABSTRACT The mucus coat on the skin and g i l l of fishes serves a variety of protective functions including lubrication of the body surface, prevention of infection, and deterrence of parasites. Since mucus secretion in some teleosts appears to be controlled by the osmoregulatory hormones, prolac-t i n and Cortisol, i t was thought that the mucus coat may participate i n teleost hydromineral balance. The primary objective of this study was to -examine the possible role of mucus in-osmoregulation of Leptocottus arma- tus Girard 1854 (Teleostei, Cottidae) and to relate these findings to the -endocrine control of mucus secretion. •^ Whereas Leptocottus skin includes three types of secretory cells (eosino-p h i l i c granular, goblet, and cuticle-secreting cells) , the g i l l epithelium -has only goblet c e l l s . Of ten cottid species examined histochemically, only seawater sculpins produce neutral mucins, while fresh water, sea-water, and euryhaline species produce acidic mucins. Leptocottus primari-l y secretes a sialoglycoprotein, though sulphated mucins are present i n g i l l goblet cells and the cuticle-secreting c e l l s . The granular cells produce a tryptophan-rich basic protein. - Hypophysectomy or seawater adaptation reduce the number of g i l l gob-let c e l l s , compared to respective sham-operated or 5% seawater-acclimated •controls. Ovine prolactin treatment of hypophysectomized fish prevented this decrease. In seawater-acclimated Leptocottus prolactin injections increase the number of g i l l mucus c e l l s , while Cortisol injections of 5% "seawater-adapted fi s h had the opposite effect. The cuticle-secreting cells were most active in 5% seawater-adapted fi s h and this state was maintained by prolactin. A moderately active state was typical of seawater-adapted i i i f i s h and this condition could be maintained by C o r t i s o l or ovine growth hormone. The lowest activity of the cuticle-secreting cells occurred i n hypophysectomized fi s h . Though prolactin and C o r t i s o l appear to con-t r o l the g i l l goblet cells and the cuticle-secreting c e l l s , neither hy-pophysectomy nor treatment with prolactin or C o r t i s o l significantly af-fected the skin goblet c e l l s . The mucus coat appears incapable of reducing integumental permeabili-ty through steric interference of diffusion or through Donnan exclusion of ions from the mucus coat. Further, i t is argued that enhancement of unstirred layers by mucus could not significantly affect hydromineral bal-ance. Instead, alterations of the mucus coat with environmental s a l i n i t y may be associated with the lubricating properties of mucus. Leptocottus mucus is a viscous non-newtonian f l u i d when concentrated but is an e f f i -cient lubricant when dilute. Seawater-adapted Leptocottus produce a more efficient lubricating mucus than do 5% seawater-acclimated f i s h ; this effect appears to involve the cuticle-secreting c e l l s . In an associated study I showed that the skin of the goby Gillichthys  mirabilis participates actively in osmoregulation through extra-renal ion excretion. Thus the effects of prolactin and C o r t i s o l may be primarily to control active transport i n the skin. i v TABLE OF CONTENTS Page ABSTRACT i i LIST OF FIGURES v i i LIST OF TABLES x INTRODUCTION 1 MATERIALS AND METHODS 4 Animal Care, Collecting and Holding Facilities. .... 4 Histology and Histochemistry Methods .... 7 Electron Microscopy .... 15 C e l l Counts and Nuclear Diameter .... 15 Surgical Operations .... 17 Hormones and Injection Vehicles .... 19 Serum Ions .... 19 Chemical Assays .... 21 Mucus Shedding Rate .... 21 Viscosimetry and Diffusion Methods .... 24 Measurement of Diffusion Coefficients .... 24 Titration of Fixed Acidic Groups .... 26 Viscosity Measurements .... 27 St a t i s t i c a l Methods .... 30 RESULTS 31 Histology of the Skin .... 31 Filament-containing Cells .... 31 Goblet Cells .... 32 Eosinophilic Granular Cells .... 34 Histology of the G i l l Epithelium .... 34 V Table of Contents (cont'd) Page Ultrastructure of the Skin .... 35 Histochemistry and Chemistry of Cottid Mucus .... 37 Histochemistry of Leptocottus Mucus .... 37 Chemistry of Leptocottus mucus .... 45 Comparative Histochemistry of Mucus .... 47 Large Goblet Cells of the Skin .... 47 Minor Mucus-secreting Cells of the Skin .... 49 G i l l Mucus Cells 53 Endocrine Control of Mucus Secretion 55 Effects of Hypophysectomy and Prolactin Replacement .... 55 Effects of Environmental Salinity .... 60 Prolactin Injection of Seawater-acclimated Leptocottus .... 67 Cortisol Treatment of 5% Seawater-adapted Leptocottus .... 69 Hypophysectomy and Hormone Replacement in Seawater Sculpins .... 71 Summary .... 76 Physical Properties of Mucus .... 77 Titration of Fixed Acidic Groups 77 Determination of Diffusion Coefficient (D) Through Mucus .... 78 Lubricating Properties of Mucus .... 81 DISCUSSION 90 Histology and Histochemistry of Cottid Mucus Cells .... 90 Goblet Cells .... 91 Eosinophilic Granular Cells 92 Cuticle-Secreting Cells .... 93 Endocrine control of Mucus-secreting Cells .... > 94 vi Table of Contents (cont'd) Page Endocrine Control of Skin Goblet Cells .... 96 Endocrine Control of G i l l Goblet Cells .... 99 Endocrine Control of Cuticle-secreting Cells .... 102 Functional Significance of the Mucus Coat: Osmoregulation .... 106 Other Functions of the Mucus Coat .... 113 CONCLUSIONS 117 REFERENCES CITED 119 APPENDIX 1 135 APPENDIX 2 137 APPENDIX 3 148 v i i LIST OF FIGURES FIGURE P a g e 1 Seawater System .... 8 2 Mucus Cell Index .... 14 3 Holding Chamber for Small Fish .... 23 4 Apparatus for Measuring Diffusion Coefficients .... 25 5 High Pressure Viscosimeter .... 28 6 Low Pressure Viscosimeter .... 28 7 Leptocottus Skin Stained with Masson's Trichrome .... 33 8 Leptocottus G i l l Stained with Masson's Trichrome .... 33 9 Ultrastructure of Cuticle-Secreting Cell Form 5% Seawater Leptocottus: High Magnification .... 38 10 Ultrastructure of Cuticle-secreting Cell from 5% Seawater Leptocottus .... 38 11 Ultrastructure of Cuticle-secreting Cell and Malpighian Cell from Seawater Leptocottus .... 39 12 Ultrastructure of Cuticle-secreting Cell from Seawater Leptocottus: High Magnification .... 39 13 Goblet Cell Ultrastructure .... 40 14 Granular Cell Ultrastructure .... 40 15 Leptocottus Skin AB (pH 2.5)-PAS after Wild Methylation .... 44 16 Leptocottus Skin: DMAB-Nitrite reaction 44 17 Leptocottus G i l l : AF-AB .... 44 18a,b, Oligocottus maculosus Skin; AB (pH 2.5)-PAS 50 19a,b, Icelinus borealis Skin: AB (pH 2.5)-PAS 50 20a,b, Ascelichthys rhodorus Skin: AB (pH 2.5)-PAS .... 50 21 Cottus asper skin; AB (pH 2.5)-PAS .... 51 v i i i L i s t of Figures (cont'd) Page FIGURE 22 Cottus aleuticus Skin; AB (pH 2.5)-PAS 51 23a Skin of Seawater-adapted Leptocottus; AB (pH 2.5)-PAS... 51 23b Skin of 5% Seawater-adapted Leptocottus; AB (pH 2.5)-PAS .... 51 24 Rostral Pars Distalis of 5% Seawater-Acclimated Leptocottus; Masson's Trichrome .... 62 25 Rostral Pars Distalis of Seawater-acclimated Leptocottus; Masson's Trichrome .... 62 26 Rostral Pars Distalis of Fresh Water-Acclimated Cottus asper: Masson's Trichrome .... 65 27 Rostral Pars Distalis of Seawater-Adapted Cottus asper; Masson's Trichrome 65 28 Fresh Water-Acclimated Cottus asper G i l l ; AB (pH 2.5)-PAS 66 29 Seawater-adapted Cottus asper G i l l ; AB (pH 2.5)-PAS 66 30 Skin from Hypophysectomized, Seawater-acclimated Leptocottus (Vehicle-injected); AB (pH 2.5)-PAS 75 31 Skin from Hypophysectomized Seawater-acclimated Leptocottus, inj ected with Cortisol; AB (pH 2.5)— PAS 75 32 Skin from Hypophysectomized, Seawater-acclimated Leptocottus Injected with Prolactin; AB (pH 2.5)-PAS 75 33 Titration of Skin Mucus with KCl for Estimation of Fixed Charge Density .... 79 34 Viscosity of Leptocottus and Icelinus Mucus at Slow Flow Rates .... g2 35 Effect of Mucus Concentration on the Lubricating Properties of Leptocottus Mucus .... 83 36 Viscosity of Leptocottus Mucus at Different Flow Rates .... 86 List of Figures (cont'd) ix Page FIGURE 37 Effect of Salinity Acclimation on the Lubricating Properties of Leptocottus Mucus .... 87 38 Profile of Unstirred Layers Adjacent to a Membrane 110 X LIST OF TABLES TABLE P a S e . . . 9 1 Pretreatment and Experimental Holding Conditions 2 Histology and Histochemistry Techniques .... 12 3 Hormone and Vehicle Specifications 20 4 Chemical Assays for Mucins .... 22 5 Histochemistry of Leptocottus Mucus-Secreting Cells .... 41 6 Composition of Mucus Glycoproteins from Leptocottus .... 46 7 Comparative Histochemistry of Major Mucus-Secreting Cells of the Skin .... /+g 8 Comparative Histochemistry of the Minor Mucus-Secret-ing Cells of the Skin 52 54 56 9 Comparative Histochemistry of the Mucus-Secreting Cells of the G i l l 10 Hypophysectomized vs. Sham-Operated Leptocottus 11 Hypophysectomized vs. Sham-Operated vs. Intact Leptocottus .... 57 12 Prolactin Replacement of Hypophysectomized Leptocottus armatus .... 59 1 3 Effects of Environmental Salinity: I. Leptocottus armatus • 61 14 Effects of Environmental Salinity: II. Cottus asper .... 64 15 Effects of Prolactin on Seawater-acclimated Leptocottus armatus .... 68 16 Effects of Cortisol in 5% Seawater-acclimated Leptocottus .... 70 17 Hypophysectomy and Hormone Replacement of Sea-water-adapted Leptocottus .... 72 xi L i s t of Tables (cont'd) TABLE Page 18 Determination of Self-diffusion Coefficient of 22 N a and 36d through Skin Mucus from Leptocottus .... 80 19 Effect of High Ionic Strength on the Lubricating Properties of Leptocottus Mucus .... 85 x i i ACKNOWLEDGMENT I am most grateful to Dr. B i l l Hoar, my supervisor, for his support and guidance throughout this study and for his careful scrutiny of my thesis and associated manuscripts. Dr. Hoar's interest was a great moti-vation and i t made the thesis a joy to execute. I sincerely thank Dr. Yoshi Nagahama, postdoctoral fellow, for many illuminating discussions and helpful suggestions. Dr. Nagahama also generously supplied the electron microscope figures of sculpin integument. I thank Dr. Tom Lam, Dr. S. Pandey, and Dr. Jim Fenwick, v i s i t i n g professors, who broadened my knowledge and interest during many discus-sions over coffee. 1 also wish to thank Dr. Khay Khoo, Dr. Norman Stacey, and Dr. Pandey for teaching me the art of hypophysectomy. I am grateful to Mari Hurlburt and Daphne Hards for help in histology and to Dr. Mark Christie for an insight into biometrics. Special thanks are due to my fellow students in the lab., Dr. B i l l Driedzic, Dr. Khay Khoo, Dr. Ken Chan, Mari Hurlburt and Ross Newman, who shared the daily joys and sor-rows of research. I wish to express my gratitude to the members of my research commit-tee, Drs. D.J. Randall, V. Palaty, J. Gosline and J. P h i l l i p s for their help and advice on the practical and written aspects of the study. Many thanks to Dr. Howard Bern, Dr. Richard Nishioka, and Dr. Shelden Shen for their generous aid and advice during my v i s i t to U.C., Berkeley. Thank-you to a l l who braved winter darkness or summer heat to help me collect those scatophagous sculpins! This work was supported by the National Research Council of Canada through a scholarship to W.S.M. and a grant-in-aid to W.S. Hoar. INTRODUCTION The skin and i t s associated secretions are important for the sur-vival of fish through a multitude of functions. Fish are coated and lined with a thin layer of mucus which comprises the ultimate interface between the animal and i t s environment. This mucus coating acts as a deterrent against fungal and bacterial infections (Fletcher and Grant, 1969; Bradshaw et a l . , 1971; Willoughby, 1972) and as an efficient l u -bricant to aid swimming (Rosen and Cornford, 1972). In cyprinids, .the skin produces an alarm substance (Schreckstoff) which can e l i c i t fright responses in these schooling species (Von Frisch, 1941), while the skin of the ammocoete larva produces a distasteful 'predator-deterring' sub-stance (Pfeiffer and Pletcher, 1964) . Mucus secretion in the skin and g i l l of some species of teleosts appear to be controlled by the anterior pituitary hormone, prolactin. Hypophysectomy causes a drop in the number of mucus cells in Fundulus  heteroclitus, Betta splendens, Umbra l i m i , Carassius auratus, and Anguilla  anguilla, an effect which may be reversed by pituitary autotransplantation or injection of ovine prolactin (Burden, 1956; Schreibman and Kallman, 1965; Stanley and O'Connell, 1970; Ogawa, 1970; Olivereau and Lemoine, 1971a,b; Olivereau and Olivereau, 1972). However, hypophysectomy or pro-la c t i n injection failed to affect the mucus cells of Tilapia mossambica - now: Sarotherodon mossambicus (Bowman, 1966, cited by Bern, 1967). Injection of ovine prolactin (Egami and I s h i i , 1962) or Tilapia or Perca prolactin (Blum, 1973) causes a marked increase i n the mucus secretion of intact Symphysodon discus. Likewise, Leatherland and Lam (1969) found that ovine prolactin injection increased the number of mucus cells i n the g i l l s of intact freshwater but not seawater Gasterosteus aculeatus. The number of mucus cells in the buccal epithelium of Anoptichthys jor- dani and in the g i l l s of A. jbrdani, Anguilla anguilla, Mugil cephalus, and M. capito, i s reduced by acclimation of these species to sea water, and these decreases are correlated with reductions i n the apparent ac-t i v i t y of the prolactin-secreting cells of the rostral pars d i s t a l i s (Mattheij and Sprangers, 1969; Blanc-Livni and Abraham, 1970; Jozuka, 1966). However, few studies have monitored simultaneously the mucus secretion i n the skin and g i l l , and no study has been undertaken u t i l i -zing hypophysectomized fish in sea water or using members of the Cottidae, the family of fishes investigated in this study. Prolactin i s also intimately involved i n hydromineral balance of teleosts in hyposmotic environments (reviews: Johnson, 1973; Bern, 1975). Though many workers have alleged that some of the osmoregulatory effects of prolactin may be due to prolactin's stimulation of mucus secretion, this has not been rigorously tested. Other pituitary hormones, mineralocorticoids, and gonadal steroids may also be involved i n the control of mucus secretion in fishes. The number of skin mucus cells in Blennius sphinx is decreased by LH and FSH treatment (Blum, 1972), and i n Blennius pavo these cells increase i n num-ber after treatment with TSH or prolactin (Blum et a l , , 1972). LaRoche and Leblond (1952) showed that the skin of parr and of smolt Salmo salar was thickened after injections of bovine thyroid extract. Cortisol i n -jection of hypophysectomized eels reduces skin content of s i a l i c acid (a measure of mucus secretion), compared to saline-injected controls (Lemoine and Olivereau, 1974). Similar treatment also reduced the s i a l i c acid content of the g i l l (Lemoine, 1974b). Finally, treatment of salmonids with gonadal steroids appears to cause thickening of the skin and stimulation of cutaneous mucus secretion (McBride and van Overbeeke, 1971; Yamazaki, 1972). The primary purpose of this investigation was to examine the role of mucus secretion i n the osmoregulation of Leptocottus armatus and relate these results to the functions of prolactin. To this end the present study investigates the pituitary endocrine control of mucus secretion i n the skin and g i l l of the rarely-studied family Cottidae, using primarily the Pacific Staghorn Sculpin, Leptocottus armatus. Leptocottus was chosen because i t isa hardy euryhaline species with a slimy scaleless skin. This species is also well suited to studies of pituitary endocrinology since the fis h i s easily hypophysectomized. The histology of four types of mucus-secreting cells is described, and electron microscopy used to cl a r i f y the role of the 'cuticle-secreting' cells in the skin. The nature of the secretion i s examined using histochemical and chemical techniques. The histology and histochemistry of the mucus cells in the skin and g i l l of seawaterr-resident, fresh water-resident, and euryhaline cottids are compared to determine i f environmental salin i t y affects the nature of the mucus secreted. The viscosimetric and lubricating properties of mucus are further studied, including the effects of environmental salinity and of ionic strength. Finally, three mechanisms by which mucus may act in hydromineral balance are proposed and tested. 4 MATERIALS AND METHODS Animal Care, Collecting and Holding F a c i l i t i e s The experimental animals were collected from local waters. Pacific staghorn sculpin (Leptocottus armatus Girard 1854) and prickly sculpin (Cottus asper Richardson) were seined from the lower reaches of the L i t t l e Campbell River, Whiterock, British Columbia during low spring tides from June to September of 1974 to 1976. The seine net, intended for two-man operation, was a 3 by 1 meter nylon mesh, with floats and lead l i n e , and with 1.5 meter poles attached to each end. The net was dragged across the f u l l width of the stream and onto the river-bank where the catch was sorted. The f i s h . were immediately placed in large plastic bags, i n four to eight cm of water, and the bag f i l l e d with pure oxygen. If the fish were kept cool, mortality was low during transport. Upon arrival the fish were transferred to permanent holding f a c i l i t i e s . Rosylip sculpin (Ascelichthys rhodorus Jordan and Gilbert 1880) and f l u f f y sculpin (Oligocottus snyderl Greeley 1901) were trapped in tide pools at Botannical Beach, Port Renfrew, B.C. during a low tide in October, 1976. Galvanized wire minnow traps, baited with crushed mussels, were placed i n the pools as the tide ebbed, and were retrieved as the tide rose. The traps could not be l e f t overnight since surf action would have dis-lodged them. These fish were also transported in oxygen-filled bags. Tide pool sculpin (Oligocottus maculosus Girard 1856) , northern sculpin (Icelinus borealis Gilbert 1895) and great sculpin (Myoxephalus  polyacanthocephalus Pallas 1811) were collected by beach seine from Burrard Inlet at Cates Park, North Vancouver, B.C., during a low spring tide in December, 1976. Tissue samples of mottled sculpin (Cottus bairdi Girard), Aleutian sculpin (Cottus aleuticus Gilbert) and torrent sculpin (Cottus  rhotheus Rose Smith) were taken from formalin-fixed museum specimens that were generously supplied by the University of Bri t i s h Columbia Vertebrate Museum, through the courtesy of N.J. Wilimovsky curator. The reference numbers for the collections were BC-56-505, BC-59-475, and BC-58-241, for Cottus bairdi, C_. aleuticus, and C_. rhotheus, respec-tively. Laboratory sea water from Burrard Inlet had a s a l i n i t y of 30 o/oo and was available at room temperature and at 5-7 C as refrigerated sea water. Volume dilutions of this sea water were used to produce hypo-osmotic media and these solutions are expressed as 'percent sea water'. Hence, 100% ( f u l l strength) sea water was 30 o/oo salts; 25-35% sea water (approximately isosmotic to fis h blood) was 7-8.5 o/oo salts, and 5% sea water (the hyposmotic medium)^" was 1.5 o/oo salts. Salinity was measured using a hydrometer and density reduction tables for sea water (Zerbe and Taylor, 1953). The main outdoor holding f a c i l i t y , where most of the Leptocottus were kept, was a shallow rectangular tank containing 5% sea water. The water was constantly recirculated by an immersible pump ( L i t t l e Giant ^"Leptocottus does not survive i n fresh water, but may be kept in d e f i -nitely in 5% sea water. Pump Co.)» aerated by compressed a i r , and fi l t e r e d through 2 cm polyfoam and 5 cm of fiberglass batting. A thin layer (about 1 cm) of coarse sand was spread on the tank bottom. There was no photoperiod or tempera-ture control on this tank. Indoor holding f a c i l i t i e s included two 200 l i t r e oval fiberglass tanks with flowing dechlorinated tap water, and centrally-placed standpipe drains. Cottus asper was kept in flowing fresh water i n these tanks but when the tanks were equipped with immersible pumps and f i l t e r s , they could also be used for holding Leptocottus in full-strength or dilute sea water. Other holding f a c i l i t i e s included groups of AO-litre pl a s t i c tubs that were used with recirculating, f i l t e r e d seawater or dilute sea water. These systems were ideal for experiments with several control and treatment groups. This inexpensive and reliable system i s described i n Figure 1. A l l indoor holding and experimental tanks were equipped with photoperiod-controlled fluorescent lamps. The recirculating systems were cooled with running tap water, when necessary, so to maintain temperature between 10 and 15 C. Tap water temperature varied seasonally between 5 and 12 C, and these variations are reflected i n the holding conditions of the ex-periments (Table 1), The fish were fed on alternate days with brine shrimp. In most of the experiments, however, feeding was reduced or discontinued, as shown in Table 1. Disease was not considered a major problem in holding Leptocottus or Cottus asper. The Leptocottus had to be separated according to size, since cannibalism was common when the fis h were held at high densities. When Leptocottus were kept in dilute sea water, there was some danger of proto-7 zoan and fungal infection i f the s a l i n i t y f e l l below 1 o/oo for more than a day. Treatment of infected fish was generally unsuccessful, but the health of the remaining fi s h could be assured simply by increasing the salinity to 1.5 - 2.0 o/oo. No fatal infections were apparent in Leptocottus that were kept in sea water, though parasite infestations occurred occasionally. On one occasion a parasitic nematode attacked the g i l l s of Leptocottus in one of the seawater tank systems; the infes-tation was successfully treated by flushing the system and adding 1:10,000 phenoxethol, according to the method described by van Duijn (1973). Histology and Histochemistry Methods The skin, g i l l , and pituitary were histologically examined after each experiment. Pieces of skin from the dorsal side of the fi s h adjacent to the anterior dorsal f i n , and parts of the g i l l (I chose, a r b i t r a r i l y , the f i r s t and second g i l l bars) were fixed in Bouin's f l u i d overnight. The pituitary gland and brain were removed intact from the fi s h and fixed in Bouin-Hollande sublimate overnight. A H tissues were stored in 70% ethanol, when necessary. No decalcification was required since the skin i s scale-less, and the pituitary and brain were easily separated from the cranial bones. Paraffin embedding followed fixation and involved: (1) dehydration of the tissue, 30 minutes in 70, 90, and 95% ethanol followed by two 15-minute changes of absolute ethanol, and (2) i n f i l t r a t i o n with paraffin (Tissueprep, Fisher Scientific Co.), 30 minutes i n each of benzene and benzene:Tissueprep (50% v/v at 60 C) followed by two 20-minute changes of melted Tissueprep (60-63 C), The f i n a l change of paraffin was done under 8 Figure 1. Recirculating sea water system. The pump (P) was a non-submersible magnet-driven chemical pump (Model 1-MD, L i t t l e Giant Pump Co.). The sea water was distributed to 4-7 t h i r t y - l i t r e tanks by a manifold system, and flow was controlled by clamps on the individual delivery tubes. Water was drawn off the tank bottoms through a drain which was equipped with an aspirator (to prevent siphoning) and was delivered to the sump. The f i l t e r was 2 cm polyfoam over 5 cm of fiberglass batting (activated charcoal optional). Good clarity of the water could be 2 maintained for two-three months when 0.1m of f i l t e r area was allowed for each 125 l i t r e s of water. An over flow was supplied in case of f i l t e r clogging. The level in the sump was easily checked by an external glass tube. Keeping the tanks in a wet tray or using a stainless steel cooling c o i l in the sump aided in temperature con-t r o l . The whole system was placed in a controlled-photoperiod room. 8a Fig. 1 O u t l e t s J i Ji It Wet T r a y D ra in s W f t n — > Ii J . LI / f / m F i l te r_ S U M P 3. Table 1. Pretreatment and Experimental Holding Conditions Experiment # Days i n : Salinity Temp. °C Photoperiod Food Hypophysectomized vs, Sham-operated Hypophysectomized vs, Sham vs. Intacts Prolactin replacement of Hypophysectomized Leptocottus Effect of Environmental s a l i n i t y (Leptocottus) Effect of Environmental sa l i n i t y (Cottus asper) pretreatment: experimental (both groups) pretreatment: experimental ( a l l groups): pretreatment: experimental ( a l l groups) pretreatment: experimental 5% sea water: 100% sea water 100% sea water pretreatment: experimental SW group: FW group: >14 7 15 1 >14 7 4 2 6 >14 4 3 1 11 >14 21 21 71 >14 1-2 o/oo 8 8 <1 9 6 3 1.5 1-2 9 9 4.5 2 1-2 1.5 30 30 47 47 De-Cl FW 30 De-Cl FW ambient 12-15 12-15 12-15 ambient 15 + 1 15 + 1 15 + 1 15 + 1 ambient 10-12 10-12 10-12 10-12 ambient 10-12 10-12 10-12 7-9 15+1 10-12 natural 14L 10D 14L 10D 14L 10D natural 14L 10D 14L 10D 14L 10D 14L 10D natural 14L 10D 14L 10D 14L 10D 14L 10D natural natural natural natural natural 14L 10D 14L 10D yes no yes no yes no no no no yes yes no no no yes yes yes yes yes yes yes .cont'd vo Table 1 (cont'd) Experiment # Days in: Salinity Temp. °C Photoperiod Food Prolactin treatment of pretreatment: >14 1-2 o/oo ambient natural yes SW-acclimated fish experimental > 30 30 10-15 14L 10D yes (both groups): 24 30 10-12 14L 10D yes Cortisol treatment of pretreatment: >14 1-2 ambient natural yes 5% seawater-adapted Leptocottus experimental (both groups): 11 1.5 7-9 natural yes Hormone replacement of pretreatment: 50 30 10+1 14L 10D yes seawater-acclimated, hypophysectomized fi s h experimental ( a l l groups) : > 14 30 10+1 . 14L 10D no Physical properties of Sea water: > 14 30 10-12 natural yes mucus (viscosity and diffusion exper.) 5% Sea water: 14 1-2 10-15 natural yes O 11 14 psi vacuum. The tissue pieces were mounted in Labtek cups and frames (Labtek Plastics) and cooled rapidly on ice. Sections were cut on a rotary microtome (American Optical Co. model 820) at 5, 7, or 10 ym. G i l l and pituitary sections were normally cut at 5 ym, with a few sections from each pituitary at 10 ym thickness; while the thicker sections yielded better demonstration of the stains, the thin sections provided greater cellular detail. Skin sections were generally cut at 7 ym, since thinner sections were d i f f i c u l t to obtain consistently, due to hardening of the muscle tissue underlying the epidermis. A summary l i s t i n g of the staining techniques appears i n Table 2. Any changes or adaptations of the original methods are presented below. Most of the mucus c e l l counts used sections that were stained in the following manner. Alcian Blue 8GX (K&K Laboratories or Fisher Scientific Co.) was dissolved i n 3% acetic acid to a concentration of 0.1% w/v (Mowry and Winker, 1956); the f i n a l pH was adjusted to 2.5 with small amounts of 1 N HCl. The sections were de-paraffinized, hydrated, stained "in Alcian Blue, AB (pH 2.5), for 30 minutes, and washed in tap water. This stain was followed by Lillie^Mayer Haematoxylin ( L i l l i e , 1965) , six minutes, and 0.5% aqueous Eosin Y, 30 seconds. This procedure distinguished those cells of the skin and g i l l that contained acidic mucins (AB-posi-tive), and also allowed inspection of the normal histology of the other cells in the tissue. Lillie-Mayer haematoxylin was used because i t im-parts a reddish colour to the nuclei, a colour which contrasts the Alcian Blue stain. Table 2. Histology and Histochemistry Techniques Method Purpose and Result Reference Haematoxylin-Eosin Masson's Trichrome Alcian Blue 8GX (pH 2.5) AB (pH 2.5) AB (pH 2.5) -H & E . Periodic acid-Schiff AB (pH 2.5)-PAS AB (pH 1.0)-PAS Mild methylation (37 C) (0.1 N HC1 in abs. methanol for 5 hours, followed by AB (pH 2.5)-PAS) Active methylation (60 C) (as above, but 4 hours i n -cubation only) General histology. Skin & G i l l . General histology. Skin, G i l l , and pituitary Acidic mucopolysaccharides stain blue. Skin & g i l l . Demonstration of mucus-secreting c e l l s . Skin & g i l l . Periodate oxidizable carbohydrates stain magenta. Skin & g i l l . Neutral carbohydrates stain magenta. Acidic mucins stain blue. Skin & g i l l . Neutral and weakly-acidic ( s i a l i c acid-containing) mucins; magenta. Sulphated mucins: blue-purple. S i a l i c acid-containing mucins; magenta. Sulphated mucopolysaccharides: blue. Skin fie g i l l S i a l i c acid-containing and sulphated mucins stain magenta. Gurr, 1962 Foot, 1933 Mowry and Winkler, 1956 adapted, see text McManus, 1946 Mowry and Winkler, 1956 Mowry and Winker, 1956 Spicer and Duvenci, 1964 Spicer and Duvenci, 1964 .cont'd Table 2 (cont'd) Method Purpose and Result Reference Methanol Control (absolute methanol for 4-5 hours at 37 C, followed by AB (pH 2.5)-PAS) C r i t i c a l electrolyte concentration; AB (pH 5.7)-PAS Aldehyde fuchsin-AB (pH 2.5) Millon's test Control for the methylation methods, above. Same result as for AB (pH 2.5)-PAS, Skin & G i l l . Increase in ionic strength (MgCl^) abolishes alcianophilia of weakly acidic mucins. Skin & G i l l . Sulphated mucins stain purple. S i a l i c acid-containing mucins stain blue. Skin & g i l l . Tyrosine-containing protein and mucoproteins stain red-orange. Spicer and Duvenci, 1964 Pearse, 1968 Spicer and Meyer, 1960 Baker, 1956 Fast green FCF, pH 8 Basic proteins stain green. Alfert and Geschwind, 1953 De-amination, Fast Green FCF, pH 8 Sakaguchi reaction Basic proteins with guanidine groups stain green. Skin. Arginine-containing protein stains red-orange. Skin. Alfert and Geschwind, 1953 McLeish, et a l . , 1957 Sudan Black B Dimethylaminobenzaldehyde-nitrate reaction Lipids stain grey to black. Skin. 3-indole derivatives (primarily tryptophan) stain deep blue. Cahyen et a l . , 1973 Pearse, 1968 Neuraminidase incubation Alcianophilia of s i a l i c acid-containing mucins is blocked Spicer and Duvenci, 1964. Hyaluronidase incubation AB (pH 2.5) Alcianophilia of some sulphated mucins is blocked. Spicer and Duvenci, 1964 14 The periodic acid-Schiff's procedure for carbohydrate (Gurr,1962) was accompanied by a benzoylation control. De-paraffinized sections were given two three-minute changes of dry acetonitrile (in a dessicator) and transferred to the benzoylation solution for two hours (acetonitrile: 50 ml, benzoyl-chloride: 4.2 ml, and dry pyridine: 2.2 ml). The sections were then washed in 100% ethanol and the PAS procedure continued as usual. Neuraminidase from Clostridium perfringens (Sigma) was dissolved in citrate=phosphate buffered saline (pH 5.0) to a concentration of 2.0 units of activity per ml. This solution was flooded on the test slides, while control slides were flooded with saline only. Half of the slides were pre-treated with 1% KOH in 70% ethanol for five minutes, in order to saponify the tissue and enhance digestion by neuraminidase. Hyaluronidase from bovine testes (Sigma) was dissolved in 0.85% NaCl to a concentration of 1.0 mg/ml (470 units/ml) and was likewise flooded on test slides. Control, neuraminidase- and hyaluronidase-treated slides were incubated at 37 C. Hyaluronidase-treated slides were removed at 3 hours incubation; neuraminidase-treated slides were removed at 4, 5.5, and 7 hours (saponif-ied sections) or 8, 16, and 24 hours incubation (unsaponified sections). Control slides were removed at 7 and 24 hours for the saponified and un-saponified sections, respectively. A l l the slides were stained with Alcian Blue 8GX (0.1% in 3% acetic acid), followed by PAS. The mounting medium used on finished slides was De-Pe-X (xylene: distrene 80: dibutylpthalate, 7:2:1 v/w/v). The microscopes used were a Wild, model M20 with camera and exposure meter (MEL13), or a Zeiss, model 52182 with camera and exposure meter. The latter microscope was 15 available through the generosity of C.V. Finnegan. Electron Microscopy Skin samples of 5% seawater- and seawater-acclimated Leptocottus were fixed i n 2% glutaraldehyde and 3% paraformaldehyde i n 0.1 M-cacodylate buffer with 0.5% CaC^ added (pH 7.5), Details may be found in Nagahama et a l . (1976). Cell Counts and Nuclear Diameter The number of skin mucus cells was estimated from 7 ym paraffin sec-tions of skin taken from the dorsal side adjacent to the anterior dorsal f i n . A H the goblet cells (Alcian Blue-positive) that appeared along a 600 ym length of section were counted and the result was expressed as the number of cells per millimeter of section. Five such sections were counted for each animal. The number of g i l l mucus cells was estimated by an index method based on the relative area occupied by the goblet cells and spaces between cells (Figure 2). The index has two distinct advantages over c e l l counts: f i r s t many g i l l filaments may be indexed, thus making the sample more re-presentative, and second, since the index estimated the relative area occupied by the c e l l s , i t i s equally applicable to large and small fish i n the experimental groups. Ten sections of g i l l filament were indexed from each fi s h , and the mean scores were used in non-parametric s t a t i s -t i c s . The activity of the epsilon (ACTH-secreting) and the eta (prolactin-secreting) cells of the rostral pars di s t a l i s was estimated from the 16 Figure 2. Mucus c e l l index. The number of mucus cells on the leading or t r a i l i n g sides of the g i l l filaments was estimated by this index method. The scores were based on the relative area of the section which was occupied by mucus cells and spaces between cell s . Thus g i l l filaments from both large and small fish could be mea-sured with equal accuracy, The drawings represent sections of g i l l filaments with darkly-stained goblet c e l l s . \ 16a M U C U S C E L L I N D E X S C O R E D E S C R I P T I O N O All ce l l s t ouch ; no s p a c e s . F e w s p a c e s , s m a l l e r than one ce l l d i ame te r . M o r e spaces , s o m e l a r g e r than one ce l l d iameter, -m o s t c e l l s t o u c h . Many s p a c e s , al l o n e ce l l d i a m e t e r o r l a r g e r ; f e w c e l l s t o u c h . V e r y l a rge s p a c e s ; no c e l l s t o u c h ; m o r e than t w o c e l l s . V e r y l a r g e s p a c e s ; t w o o r f e w e r c e l l s . No c e l l s . 17 histological appearance of the cells and from measurements of the nu-cleus diameter of the ce l l s . The diameters of ten nuclei from a mid-saggital section of the pituitary were measured. Both the long and short diameters of the cells were recorded and the square root of the product of the two diameters was calculated. This square root transformation of the data allows direct and representative comparisons of spherical and elongated nuclei (Yamazaki and Donaldson, 1968), Surgical Operations Mucus samples, intended for chemical analysis and viscosimetric stu-dies, were removed from the fish by 'scraping' the skin in a controlled manner. The fish was f i r s t l i g h t l y anaesthetized in cold (0 C) 1:10,000 MS222 (Sandoz) and then placed on a large sheet of plexiglass. The f i s h was moved back and forth across the glass as the mucus-water mixture was collected with a squeegee and set aside in test tubes on ice. This pro-cess was repeated twice or three times, and d i s t i l l e d water was added to keep the fish moist. Seawater-acclimated fish were anaesthetized in sea water, and 5%-seawater fish likewise i n 5% sea water. Some fish were anaesthetized i n cold water only, to control for possible effects of MS222 on the mucins. The 'scraping' did not appear to harm the f i s h , but some fish suffered from cold-temperature shock. Skin samples were taken for histological evaluation of the method and to visualize the extent of damage to the skin. Hypophysectomy was performed routinely on Leptocottus of 5-30 g. Smaller fi s h were preferred because their pituitary would be easily seen 18 through the translucent bones of the sk u l l ; i n larger f i s h the cranium was thick and opaque. The hypophysectomies started after the fish were acclimated to the laboratory conditions (see table 1 for details of specific experiments). Each fish was anaesthetized in cold (0 C) 1:5,000 MS222 in either 5% sea water or f u l l strength sea water and then placed on an ice-packed sponge. The sponge, which served as the operat-ing surface, was placed under a dissecting microscope where the fish's branchiostegal membrane was s l i t anteriorly and the operculum retracted to expose the buccal cavity. After a 3-6 mm incision was made in the buccal skin, a de n t i s t ' s d r i l l and round 0.5 mm burr were used to d r i l l through the skull at the tip of the parasphenoid bone. The dura mater was then pierced to expose the brain and pituitary gland. The Leptocottus pituitary Is not encapsulated by bone, so the gland was easily removed by gentle suction through a fine glass pipette. The lobed, opaque p i -tuitary was easily recognized i n the pipette. As a further check of the success of the operation, a visual inspection was made during dissection at the end of the experiments. Sham hypophysectomy operations were per-formed identically, except that the pituitary was not removed, After the operation the wound was sealed with a drop of Eastman 910 acrylic adhesive (Eastman Chemical products) and the fish put in 25-30% sea water with 1:40,000 tetracycline. Once the method was mastered, the operation caused no major haemorrhages, and a l l operated fish recovered rapidly and survived for several weeks. Normally, hypophysectomized fish were not fed post-operatively, since food, i f lodged in the wound, promoted secondary infection. 19 Hormones and Injection Vehicles Ovine prolactin, ovine growth hormone and hydrocortisone (Cortisol) were dissolved i n specific vehicles and injected intraperitoneally into Leptocottus during the various experiments of this study. Care was taken in each case to inject as small a volume as possible, usually 0.1 ml. Where possible, the vehicle was isotonic and contained a minimum amount of carrier. The specific characteristics of the hormones and their vehi-cles are summarized i n Table 3. The doses given to the fish (usually 10 ug/g) were based on the mean weight of the group at the beginning of the injection schedule. How-ever, because the fish were usually starved during the experiments and thus lost weight, the dosages reported in the results (based on i n d i v i -dual fish weights at the end of the experiment) are 10-15% higher than the i n i t i a l dosage. This was true for a l l the experiments except C o r t i -sol injection of 5% seawater-acclimated Leptocottus where individual fish were identified and were given injection volumes appropriate to their known weight. The time for injection was ar b i t r a r i l y chosen to be 1600-1700 hours that i s , 4-5 hours before the onset of the dark phase of the 14L 10D photoperiod. Serum Ions At the end of each experiment the fish were anaesthetized in cold (0 C) 1:5,000 MS222, and blood samples were taken from the caudal vessels using non-heparinlzed microhaematocrit tubes. The tubes were centrifuged in a microhaematocrit centrifuge (Clay-Adams) and the haematocrit was recorded. Serum samples were diluted 1:3,000 in d i s t i l l e d water for Table 3. Hormone and Vehicle Specifications Hormone Supply Source Purity Vehicle Preparation Experiment(s) Ovine Prolactin NIH-P-S-11 26.4 IU/mg Prolactin 2.5 mg Prolactin replacement of hypo-0.1 N NaOH 50 Hi physectomized f i s h , and Prolac-0.6% NaCl 1.2 ml t i n injection of SW-adapted (0.9% NaCl for SW- f i s h acclimated fish) Ovine Prolactin NIH-P-S-12 35 IU/mg Prolactin 1,2 mg ' Prolactin replacement of hypo-0.1 N NaOH 50 ul physectomized, SW-adapted fis h 0.9% NaCl 1.1 ml Ovine Growth NIH-GH-S-11 0.56 IU/mg Growth hormone 1,2 mg Growth hormone replacement of Hormone 0.1 N NaOH 50 Ul hypophysectomized, SW-adapted 0.9% NaCl 1.2 ml fi s h . Hydrocortisone Sigma Cortisol 1.0 mg Cortisol injection of 5% SW-(Cortisol) 95% ethanol 75 y l adapted f i s h , and Cortisol re-0.9% NaCl 1.1 ml placement of hypophysectomized, (0.6% NaCl for 5% SW- SW-acclimated f i s h , acclimated fish) Hirano and Utida, 1968 SW: - sea water to o 21 determination of serum sodium, 1:1,000 in 500 ppm NaCl for measurement of serum potassium, and 1:100 i n 0.5% LaCl^ + 1% concentrated HCL for serum calcium determinations. Serum ion concentrations were measured with a Techtron AA120 atomic emission spectrophotometer. Lambda max was 589.4, 766.9, and 422.7 nm for sodium, potassium, and calcium determinations, respectively. Only serum sodium was measured routinely; potassium, c a l -cium, and chloride were measured in order to design a balanced salt solu-tion for Leptocottus. Chloride was measured using a Radiometer CMT 10 chloride t i t r a t o r . Chemical Assays The constituents of mucus were par t i a l l y analysed by chemical assays for protein, neutral hexose, s i a l i c acid (N-acetyl-neuraminic acid), and fucose. The terminal residues of the carbohydrate moieties are particular-ly Important since these groups dictate the ionic nature of the mucins. A l l the assays are spectrophotometry; readings were taken on a Unicam model 1800 spectrophotometer. Table 4"shows details of the assays and gives the appropriate standards and references for the methods. Mucus Shedding Rate The shedding rate of mucus from the body surface of Leptocottus was estimated i n order to confirm the findings of the mucus c e l l counts. The fis h were denied food for at least five days before testing to minimize interference of fecal material i n the assay. The fish (10-17 g) were kept overnight i n small volume of 5% sea water (60-80 ml per fish) inside small fis h chambers (Figure 3). During this Interval the fi s h shed measurable Table 4. Chemical Assays for Mucins A s s a y s - Specific Molecule(s) Assayed Standard Reference Lowry Phenol-HoS0. 2 4 Fucose Thiobarbituric acid assay^ Protein Neutral hexose Methyl pentose N-acetyl-neuraminic acid ( s i a l i c acids) Bovine serum albumin Galactose Fucose N-acetyl-neuraminic acid Lowry et al.,1951 Lo at a l . , 1970 Dische, 1962 Warren, 1959 1 Bovine submaxillary mucin (Sigma) was included in each assay as a control. "The assay was performed after acid hydrolysis, according to Warren (1959). t o u> Figure 3. Holding chamber for small f i s h . Acrylic plastic tubing was used to make this small fish chamber. The fi s h was introduced by removing" the stopper; bubbles were removed through the stopcock as the chamber was f i l l e d . The water was recirculated and aerated by an air l i f t pump and the volume could be adjusted (or samples taken) through the open-ended aeration chamber. 24 quantities of mucus into the surrounding water. The water + mucus was removed from the chambers after the allotted time and the mixture was assayed for neutral hexose by the phenol-sulphuric acid method. The results were expressed as uM.hexose per 100 g fish per day, Viscosimetry and Diffusion Methods Measurements of Diffusion Coefficients Self-diffusion of sodium or chloride ions through a mucus solution was measured by the capillary method (Wang, 1951, and Wang e_t al_., 1954; details in Appendix 1). Mucus was removed from small (10-15 g) Leptocottus and dialyzed against d i s t i l l e d water for three hours. Aliquots were taken to determine protein concentrations (by the Lowry method) and sodium concentration (by atomic emmission spectrophotometry). Sodium chloride was added to the sample to a concentration of 100 mM. Radioactive sodium ( 2 2NaCl, 200 uCi/ml as 1.0 ug NaCl/ml) or chloride ( 3 6C1, 77 yCi/ml as 1.3 mg NaCl/ml) was then added to the mucus solution and to 100 mM NaCl (control solution) in proportions of: 1 yl radionuclide solution to 199 Ul saline or mucus solution. The prepared solutions were put into 20 capillary tubes (2.0 x 0.076 cm), ten control tubes containing radionuclide + saline and ten experimental tubes with mucus + radionuclide. Five tubes from each group were immediately emptied onto sample pans and dried prior to counting of gamma emissions. The remaining ten tubes were placed in the diffusion chamber (Fig. 4). The saline in the chamber (also 100 mM NaCl) was slowly stirred by the magnet, and the revolutions per minute of the magnet were recorded using a stroboscope. The tubes were removed from fO 22 Figure 4, Estimation of the diffusion coefficients of Na and 36 Cl through saline or saline + mucus. Capillary tubes (2.0 x 0.076 cm) were f i l l e d with saline or mucus and radionuclide. Slow st i r r i n g of the saline bath main-tains the concentration of radionuclide at the open end of the tube at zero. Temperature was controlled +0.1 C. Decreases in radionuclide content in the tube over time reflects the rate of diffusion. E s t i m a t i o n of D by the C a p i l l a r y t u b e M e t h o d sa l ine+"Na o r J 6 C I m u c u s • "Na o r " C I 2.0 x 0.076 c m . t u b e s SAL INE A ) s t i r r i n g b a r 2 0 c m . d iam. g lass d i s h s t i r r i n g m o t o r c o n t r o l l e d t emp , bath IV) 26 the chamber after 20-24 hours (the diffusion time, t ), and were emptied onto sample pans, dried, and counted by a gamma emission counter. The CPM of the i n i t i a l and f i n a l sets of tubes represented the values of C° and C , respectively. The quantities C ( i n i t i a l concentration), C (final concentration), the tube height (1), and the diffusion time (t) gave the self-diffusion coefficient (D), according to equation (1). Dt 4 ln 2 „2 8 C° f2 „av (1) 2 In a l l the tests Dt/1 was greater than 0.2, so the assumptions inherent in the above equation were satisfied. Background information on diffusion coefficients was obtained from Gosting (1956); see appendix 1. Titration of Fixed Acidic Groups The concentration of fixed acidic groups in the mucus from Lepto- cottus skin was estimated in a manner similar to the pH t i t r a t i o n methods outlined in Helfferich (1962). Disodium EDTA was added to a mucus sample (from a 5% seawater-adapted sculpin) to a f i n a l concentration of 10 mM. The mucus was washed repeatedly with d i s t i l l e d water, centrifuged and the supernatant discarded u n t i l the conductivity of the supernatant approached d i s t i l l e d water (less than 2 ^ imho). The mucus sample, now free of EDTA, was acidified to pH 2.0 with added HCl; the acidification caused rapid precipitation of the mucus. After the mucus was forcibly re-suspended, i t was again washed un t i l the pH and conductivity of the supernatant were the same as d i s t i l l e d water (pH 6.0; conductivity 0.9 ^ imho). The mucus sample was then s p l i t into three portions; one for determination of the mucus 27 concentration as mg solids per ml lyophilized mucus, while the other portions were used for t i t r a t i o n of fixed acidic groups by added CaC^ or KCl. The pH of the sample was monitored continuously as small vol-umes of the salts were added. The increase i n H (drop in pH) represents the concentration of fixed acidic groups, according to the reaction: H:..Y + Me+ >».Me...Y + H + + + 2+ The addition of metallic cation, Me ,(in this case K or Ca ) to the acidified mucus, H...Y, liberates H + when the cation binds to the mucus to form an ionic complex. Me...Y. Viscosity Measurements Two simple viscosimeters were devised to measure the viscosity and lubricating properties of fis h slimes. Both instruments were tube-type viscosimeters; the tube was a 30.5 cm polyethylene tube with an inside diameter of 0.076 cm. This tube was sleeved in stainless steel (a 16 guage Hamilton needle) to keep the viscosimeter tube r i g i d and straight during the tests. One of the viscosimeters was designed for rapid move-ment of the sample at high pressures (0.5-4.5 atmospheres) (Figure 6), and the other was designed for relatively low velocities and pressures (0.0-0.5 atmospheres) (Figure 5). Mucus samples for viscosity measurements were collected from seawater-or 5% seawater-acclimated Leptocottus and from seawater-acclimated Icelinus borealis. The mucus was diluted to various v/v concentrations with either d i s t i l l e d water or a variety of prepared salines. When more Figure 5. Left, High pressure viscosimeter. Pressure (0.5- NJ oo 4.0 atm) was applied to the slime sample (S) using compressed air (CA). Opening the valve (V) allowed the sample to pass through the tube (T) ; as the sam-ple l e f t the tube the time was recorded (R) by the completion of the ci r c u i t . The samples (of fixed volume) were loaded through syringe (L) . Figure 6. Right, Low pressure viscosimeter. Pressure (0.0-0.5 atm) was applied by a water column (hatched) and recorded with a mercury manometer (M) . The slime sample (S), of known volume, flowed through the viscosimeter tube (T) when the valve (V) was open. Time was recorded with a stop watch. R^  and are the upper and lower reservoirs (respectively) of the water column. 28a 29 standardized conditions were required the mucus samples were dialyzed against d i s t i l l e d water for three hours at 4 C, and a 1.0 ml aliquot taken for freeze-drying. The dialyzed samples were diluted to 1:5,000 or 1:10,000 w/v so that the properties of the mucus from different i n -dividuals could be directly compared. The viscosity of the mucus or standard solutions was calculated using the Poiseuille equation (2) and the results were expressed in centipoise. 4 • • Hr P " = 81 V ( 2 ) Where viscosity (n ) is calculated from the tube inside radius (r) , the 2 force applied to the sample (P i n dynes/cm ), the length of the tube (1, 3 cm), and the resultant flow of fl u i d through the tube (V) in cm /sec. When flow through the viscosimeter tube was laminar equation (2) gives 'absolute viscosity' but when the flow was turbulent the result i s bet-ter described as 'apparent viscosity'. A measurement of the lubricating properties of fis h slimes was ob-tained when the velocity of the mucus was expressed as a percentage i n -crease over the velocity of the standard. The velocity of the sample through the tube i s : 0.01 V v = 2 (3) r t where V is the volume of the sample, r i s the inside radius of the tube, and t i s the time taken for the passage of the sample through the tube. When a l l measurements are i n cgs units, the velocity, v, i s in meters/second. 30 S t a t i s t i c a l Methods Fisher's exact probability test was used for nominal data from two-sample experiments or from experiments with larger numbers of treatments for which orthogonal sets of. comparisons had been planned. The Mann-Whitney U-test was used i n similar situations, but where the data were ordinal (for example, results from the mucus c e l l index). The Kruskal-Wallis one-way analysis, of variance was used with ordinal data in experi-ments that contained more than two groups. Student's t-test was used for two-sample cases with interval data, when the assumption of equality of variances was met. The null hypothesis was considered rejected when P <0.05. Sample means are expressed as the mean plus or minus one standard error. The methods used are presented in Sokal and Rolf (1969) or Siegel (1956). 31 RESULTS Histology of the Skin The histology of.Leptocottus skin is described f i r s t , and compari-sons among other species of Cottids follows. Leptocottus has a totally scaleless skin; there are no scales even along the late r a l line. The dermis is 100-150 urn thick and primarily consists of fibrous connective tissue. The connective tissue of the upper layers of the dermis is dense fibrous tissue with the fibers running parallel to the basement membrane of the epidermis (Fig. 7). However, in the lower layers of the dermis, adjacent to the musculature, the tissue i s diffuse (areolar) connective tissue. Most of the blood vessels of the dermis occur in the latter type of tissue, and appear primarily to serve the underlying mus-cles. Few blood vessels appear in the upper dermis, thus blood supply to the epidermis is poor. Skin from the dorsal side of the body has two distinct chromatophore layers, one at the junction of the dermis and mus-culature and the other adjacent to the basement membrane of the epidermis. The skin of the ventral side lacks chromatophores but otherwise i s similar to the dorsal skin. The epidermis of Leptocottus is 60-90 um thick, and consists of closely-packed cells in a typical s t r a t i f i e d epithelium (Fig. 7). Filament-containing Cells Three c e l l types are recognized i n the epidermis of Leptocottus; filament-containing cells., goblet c e l l s , and eosinophilic granular c e l l s . 32 The histology of these c e l l types are presented i n sequence. The pre-dominant cells of the epithelium are the filament-containing (also called Malpighian) c e l l s . These cells are small, 4-6 um; they are polygonal in the lower layers of the epidermis, but become cuboidal in the uppermost layers. The Malpighian cells exposed to the environment are responsible for the pattern of microridges on the skin surface (Fig. 11, 12). In seawater-adapted Leptocottus this distal layer of cells stains strongly with Alcian Blue and PAS, indicating the presence of mucins in the c e l l s , but i n 5% seawater-acclimated fish the cells are usually unstained. Whitear (1970) confirmed with electron microscopy that this c e l l layer i s responsible for secreting a mucus-like substance and called these spe-cialized Malpighian cells 'cuticle-secreting c e l l s ' . It is clear that the cuticle-secreting cells of seawater-adapted Leptocottus do contribute to the mucus coat (Fig. 10, 15) and that the type of mucin produced by these cells differs from that of the goblet c e l l s . Goblet Cells The mucus-secreting goblet cells stain darkly with Alcian Blue (pH 2.5) and are 25-30 ym wide, 35-45 ym t a l l , when mature (Fig. 7, 15). The common spherical or ovoid shape of developing goblet cells is changed as the cells mature and small bulges appear at the apical end of the c e l l s . These bulges appear to be weaknesses in the c e l l ; they appear just prior to c e l l rupture in the process of holocrine secretion. Nucleus and cytoplasm are restricted to the baso—lateral sides of the c e l l s , while most of the c e l l is f i l l e d with large membrane-bound vesicles con-taining the secretion product. 33 Figure 7. Leptocottus skin; Masson's trichrome. The eosinophilic granular cells (E) are darkly-stained with Biebrich Scarlet but the goblet cells (G) are unstained. The chromatophore layer (C) may be seen just beneath the basement membrane. Below are the dermis (D ) and the hypodermal musculature (HD). Figure 8, Leptocottus g i l l ; Masson's trichrome. This i s a section of g i l l filament showing the supporting cartilage (C) , the afferent (A) and efferent (E) blood vessels, and the g i l l lamellae (L). Goblet cells (G) are present on both the leading and the t r a i l i n g sides of the filament and a few cells may be seen on the g i l l lamellae (arrow). 34 Eosinophilic Granular Cells The third type of c e l l that contributes to the mucus coat i s the eosinophilic granular c e l l , named for i t s appearance in formalin-fixed histological sections. These cells are 15-20 um wide and 35-40 um t a l l ; they are approximately pear-shaped when mature. Though the apical mem-brane of these cells often appears to be ruptured, I have not witnessed the secretion of the c e l l contents in any of the skin sections I have examined. The apical portion of many of these cells appears more dense than lower portions (Fig. 7, 16). This dense apical 'plug' may be the cause of the apparent reluctance of the cells to release their contents. The actual secretion of the c e l l contents may require physical distur-bance of the skin (Pfeiffer and Pletcher, 1964). Though the eosinophilic granular cells are as common as the goblet cells in the skin of the body surface, there are no granular cells i n the skin of the g i l l or buccal cavity. The existence of three distinct types of secretory cells (goblet c e l l s , granular c e l l s , and cuticle-secreting cells) suggests that the cuticle or mucus coat i s a heterogeneous mixture of secretion products. In contrast, the mucus coating of the g i l l and buccal epithelium i s the product of a single type of mucus c e l l . Histology of the G i l l Epithelium The epithelium of the g i l l (Fig. 8) i s 40-60 uci thick over the g i l l filaments in the interlamellar spaces, but only 3-6 ym thick on the 35 secondary lamellae. The mucus-secreting cells are much smaller than the goblet cells of the skin, only 6-9 ym wide and 12-15 ym t a l l . These ovoid cells are located almost exclusively in the uppermost layers of cells i n the thicker parts of the g i l l epithelium, but some mucus ce l l s are seen i n the thin epithelium of the secondary lamellae. The mucus cells of the latter region are small (5 ym) and spherical. The nucleus and cytoplasm of the g i l l mucus cells are restricted to the basal side of the c e l l s ; this gives the cells their typical 'signet ring' appear-ance. The secretion product, contained in membrane-bound vesicles, ap-pears to be a mucin because of i t s strong staining with Alcian Blue (pH 2.5) or PAS (Fig. 17). Typical chloride-secreting cells appear i n the interlamellar spaces near the afferent side of the g i l l of seawater-adapted Leptocottus. The chloride cells stain li g h t l y with Biebrich Scar-let in Masson's Trichrome; the granular cytoplasm, large basally-located nucleus, and apical p i t are easily seen. The skin and g i l l epithelia of Leptocottus are highly specialized tissues. The skin produces a heterogenous mixture of secretions that combine to form the mucus coat, while the g i l l epithelium produces a mucus coat from a single c e l l type. A later section concerning histo-chemistry presents further differences between the skin and g i l l , based on the chemical nature of the secretion products of the various c e l l types. Ultrastructure of the Skin Though the ultrastructure of teleost skin has been extensively stu-died by other workers, the present study required c l a r i f i c a t i o n of one 36 feature of sculpin skin, the structure of the cuticle-secreting c e l l s . It i s not clear from histology and histochemistry whether the Alcian Blue-staining of these cells in the seawater fi s h reflects active se-cretion of mucus or simply an accumulation of mucins in the c e l l s . The cuticle-secreting cells of both 5% sea water- and 100% seawater-acclimated Leptocottus contain bundles of filaments, suggesting that these cells originate directly from the filament-containing (also called Malpighian) cells that make up the bulk of the epidermal tissue of tele-osts. In the lower layers, these cells do not appear to accumulate secretory granules (Fig. 11) and do not stain with Alcian Blue (Fig. 15). Here the filament-containing cells are tightly-bound to adjacent cells through numerous desmosome junctions (Fig. 11). In contrast, the cuticle-secreting cells at the distal surface of the epithelium appear to pro-duce secretory granules and are only loosely-bound to the underlying c e l l layers. The cuticle-secreting cells of seawater-adapted Leptocottus (Fig. 11, 12) have pycnotic nuclei and few mitochondria, suggesting a low rate of activity and degeneration of the c e l l . These cells contain many electron-lucent granules in the perinuclear area, and a few of these may be seen in the process of exocytosis at the apex of the c e l l . A skin sample from a similar seawater-adapted fish shows that this ultrastructur-al configuration corresponds to strong staining by Alcian Blue (Fig. 23a) . In contrast, the cuticle-secreting cells of 5% seawater-acclimated Leptocottus have more spherical nuclei, numerous mitochondira and rough 37 endoplasmic reticulum, suggesting a higher.rate of metabolic activity. Small numbers of electron-lucent and electron-dense secretory granules were seen throughout the cytoplasm and near Golgi apparatus (Fig. 9, 10). These granules are also seen i n the process of exocytosis. Histological-l y , cells in this condition are either unstained or only l i g h t l y stained in the Alcian Blue procedure, as shown in Figure 23b. In summary, i t appears that the strong Alcian Blue staining of the cuticle-secreting cells in the skin of seawater-acclimated sculpins re-present the'accumulation of mucins in dying c e l l s , and the lightly-stained or unstained cuticle-secreting cells of 5% seawater-adapted fish are more actively producing and secreting mucus. Electron micrographs of the eosinophilic granular cells (Fig. 14) and goblet cells (Fig. 13) show the fundamental differences between these two c e l l types. The eosinophilic granular c e l l i s not 'granular', as in formalin-fixed histological preparations, but rather appears to contain an amorphous electron-dense secretory product. The secretion appears to be produced i n Golgi complex associated with rough endoplasmic reticulum. The secretion vesicles seem to coalesce into one very large vesicle that occupies most of the c e l l . In contrast, the goblet c e l l s produce an elec-tron-lucent secretion i n large vesicles that do not appear to coalesce be-fore secretion. Histochemistry and Chemistry of Cottid Mucus  Histochemistry of Leptocottus Mucus The histochemistry of the mucus-secreting cells of the skin and g i l l r of Leptocottus i s summarized i n Table 5. The mucus cells of Leptocottus 38 Figure 9. Ultrastructure of cuticle-secreting c e l l of a 5% seawater-acclimated Leptocottus. This figure shows the extensive Golgi apparatus (G) associated with the cuticle-secreting cells of 5% seawater fi s h . Some secretion granules (arrows) are present with mitochondria (M) near the Golgi apparatus in the peri-nuclear area. Figure 10. Ultrastructure of cuticle-secreting c e l l of a 5% seawater-acclimated Leptocottus. This lower magnification figure also shows numerous mitochondria (M), rough ER and Golgi apparatus (G). The nucleus of the c e l l (N) i s large and rounded, and the cytoplasm is ligh t l y stained indicating a c e l l i n a 'highly active' state. Compare this figure with Fig. 11, 12 that represent the 'moderately active' form of the cuticle-secreting c e l l s . A few secretory granules may be seen in the cytoplasm (arrows) and one appears to be undergoing exocytosis. Figure 11, l e f t . Ultrastructure of cuticle-secreting and Malpighian c e l l s in UJ sea water Leptocottus. The filament-containing or Malpighian cells (FC) have mitochondria (M) in the perinuclear area and large bundles of filaments (F) that take up much of the cyto-plasm and are continuous with desmosomal int e r c e l l u l a r connnec-tions (D). The cuticle-secreting c e l l (CS) have a darkly-stained cytoplasm, a pycnotic nucleus and few cellular organ-elles except for numerous secretion vesicles (V) . These fea-tures indicate a c e l l in poor condition, but some of the vesi-cles appear to be undergoing exocytosis (Arrow). Figure 12, right. Ultrastructure of cuticle-secreting c e l l i n sea water Leptocottus. Like the cells of Fig. 11, these cells are i n an apparently weakened condition, and are defined as 'moderately active'. This condition corresponds to the strongly staining cuticle-secreting cells in Alcian Blue procedures (Fig. 23a). Compare this figure with the 'highly active' cuticle-secreting cells in Fig. 10. 40 Figure 13. Ultrastructure of a Goblet C e l l . The secretory product of goblet c e l l s i s electron-lucent and i s contained i n membrane-bound vesicles (v) that remain intact u n t i l secretion. The cytoplasm (C) of this mature c e l l i s restricted to the periphery. Figure 14. Ultrastructure of an eosinophilic granular c e l l . Here the secretory product i s electron-dense and, though i t Is produced i n membrane-bound vesicles (V), these vesi-cles appear to coalesce during' the maturation of the c e l l . This c e l l i s not entirely mature and retains mito-chondria (M) and rough ER, apparently for the production of the secretory material. 41 Table 5. Histochemistry of Mucus-secreting Cells in the Pacific Staghorn Sculpin, Leptocottus armatus. M, magenta (PAS-positive); B, blue (AB-positive; P, purple (both AB and PAS-positive); G green; RO, red-orange; DB, deep blue; (-) unstained; 1,2,3,4 reflect stain intensity. Skin Cells G i l l Cells Histochemical Test Goblet Granular C.S.* Goblet Cart.** AB (pH 2.5)-PAS AB (pH 1.0)-PAS Mild Methylation-AB (pH 2.5)-PAS Methanol Control-AB (pH 2.5)-PAS Active Methylation-AB (pH 2.5)-PAS C r i t i c a l Electrolyte Cone. (M MgCl„) in AB (pH 5.7) Aldehyde Fuchsin-AB (pH 2.5) Fast Green FCF (pH 8) De-amination, Fast Green FCF (pH 8.0) Sakaguchi Reaction DMAB-Nitrite Reaction Hyaluronidase (3 Hr. at 37 C)-AB (pH 2.5) 4 B 3 M 3 M 4 B 1 M (0.1 M) 2 B 3 B Neuraminidase (4-24 Hr. at 37 C)-AB (pH 2.5)-PAS (-) (some 3M) 3 P 2 G 1 G 1 RO 3 DB 4 B 3 P 3 B 4 B + M 4 B 2 M 3 B-P 4 B 2 M (0.2 M) 3 B-P 4 B 3-4 B 3 B 2 B 3 B 3 B + M (0.6 M) 3 P + B 3 B * Cuticle-secreting cells from a seawater-acclimated f i s h . **Hyaline cartilage in g i l l sections, a sulphated mucopolysaccharide. 42 stain strongly with PAS (Table 5) indicating the presence of carbohy-drate. Benzoylation prior to PAS eliminated this staining; thus PAS reactivity i s not an arti f a c t . Further/Sudan Black staining of the mucus cells i s slight, so periodate-oxidizable saturated fatty acids are probably not responsible for the PAS reactivity. Millon's test was positive for the mucus c e l l , indicating the presence of protein. Thus the substance in the mucus cells i s a combination of protein and poly-saccharide , The goblet cells of the skin and g i l l , and the cuticle-secreting cells a l l stain strongly with Alcian Blue (AB) i n the combined Alcian Blue, pH 2.5, Periodic acid-Schiff (PAS) procedure, AB (pH 2.5)-PAS, (Fig. 23a,b). This demonstrates that the mucins secreted by these cells are acidic mucopolysaccharides, since neutral mucins would stain magenta (PAS-positive) in this procedure. To distinguish these mucins from the more tenacious sulphated muco-polysaccharides , the Alcianophilia of weakly acidic ( s i a l i c acid con-taining glycoproteins) can be blocked by three methods. F i r s t , i f the pH of the AB stain is lowered to pH 1.0, most weakly-acidic mucins f a i l to stain with AB and instead stain with PAS in the combined procedure. Second, i f the sections are pretreated (with appropriate controls), in 0.1 N HC1 in absolute methanol, the weakly acidic groups are methylated and the alcianophilia of these mucins i s blocked. Third, i f the ionic strength of the AB stain i s increased (0.6-0.8 M MgCl2) the alcianophilia of weakly acidic mucins i s once again blocked. The supporting bar of hyaline cartilage in the g i l l serves as an inherent control for a l l of the above methods, since i t is composed of chondroitin sulphate, a strongly 43 acidic (sulphated) mucopolysaccharide. According to the above c r i t e r i a i t appears that the skin goblet cells secrete a weakly acidic mucin (Table 5) and that the acidic groups are s i a l i c acid, since neuraminidase digestion blocks alcianophilia of these ce l l s . Though i t is unusual that these cells are not noticeably PAS-positive in the combined AB(pH 2.5)-PAS procedure, Harris _e_t al_. (1973) have reported such variable staining of trout goblet cells with this pro-cedure^ The cuticle-secreting c e l l s , unlike the skin goblet c e l l s , seem to secrete a sulphated mucin, since these cells react similarly to hyaline cartilage (Table 5). The g i l l goblet cells clearly secrete an acidic mucin, but i t is not certain whether these cells primarily contain a sialoglycoprotein, as is suggested by the results of AB(pH 1.0)-PAS,' AP-AB(pH 2.5), and CEC-AB(pH 5.7) procedures, or a sulphated mucin, as i t would appear from the AB(pH 2.5)-PAS after mild methylation and these c e l l s ' apparent resistance to d i -gestion by neuraminidase. It may be that the g i l l goblet cells secrete a mixture of these two mucins, but certainly more definitive tests are indicated. These comments also apply to the mucus histochemistry of other species in this study. The eosinophilic granular cells are not mucus cells in the s t r i c t definition since they do not contain PAS-reactive carbohydrate (Table 5). Instead the secretion appears to be a basic protein, in that these cells stain with Fast Green at pH 8.0. Further, and most dramatic, the protein appears to contain large amounts of tryptophan, or some other 3-indole derivative, because of i t s strong reaction to the dimethylamino benzoate-n i t r i t e reaction (DMAB-nitrite), Table 5 and Figure 16. ^Fletcher, Jones, and Reid (Histochera. J. 8^:597-608)have also reported variable staining of f i s h goblet cells after AB(pH 2.6)-PAS. 44 Figure 15. Leptocottus armatus skin stained with AB (pH 2.5)-PAS after mild methylation. Note that the goblet cells (G) stain magenta (PAS +) indicating weakly acidic mucins, but the cuticle-secreting cells(C) retain alcianophilia, indicating strongly acidic (sulphated) mucins. Note also that the secreted mucus (M) appears to be a mixture of the two types of mucins, x 625. Figure 16. Leptocottus armatus skin stained by the DMAB-nitrite method for tryptophan. The eosinophilic granular cells (E) stain dark blue, but the goblet cells (G) are unstained. Note the more dense 'apical plug' (arrow), x 625. Figure 17. Leptocottus g i l l ; AF-AB. While the goblet cells (G) stain darkly with Alcian Blue, indicating a weakly acidic mucin, the supporting cartilage (C) stains purple, suggesting the presence of strongly acidic mucins. Note the difference in the size of the goblet cells of the g i l l here and the skin, above, x 625. 45 In summary, the skin of Leptocottus produces three types of mucins in three distinct c e l l populations; a s i a l i c acid-containing glycopro-tein from the goblet cells, a sulphated mucopolysaccharide from the cuticle-secreting c e l l s , and a basic protein from the eosinophilic granu-lar c e l l s . In contrast, the goblet cells of the g i l l , a single c e l l type, appear to produce a mixture of weakly and strongly acidic mucins. Chemistry of Leptocottus mucus The chemical analysis of mucus from Leptocottus skin (Table 6) sup-ports the conclusions of the histochemical study of the mucus-secreting ce l l s . Like bovine submaxillary mucin, a commercially-available glyco-protein, Leptocottus mucin contains protein and s i a l i c acid as i t s major components, and lower proportions of neutral hexose and fucose. Because the Leptocottus mucus i s a crude preparation and may contain some cellular contaminants, the estimate of s i a l i c acid may be overstated since 2-deoxy-ribose also reacts in the thiobarbituric acid assay. Suspecting this con-taminent I employed the following equation (Warren, 1959) which compensates for the contribution of 2-deoxyribose to the absorption maximum for s i a l i c acid at 549 nm. uMoles N-acetylneuraminic acid = 0.090 A,.^ - 0.033 Knowing the nature of mucus secretion in one euryhaline Cottid, Leptocottus, we are now prepared to compare the mucus secretion of this species with those of stenohaline freshwater and stenohaline seawater spe-cies from the same family. This was done to gain insight into a possible correlation of mucus secretion and environmental s a l i n i t y . 46 Table 6. Composition of Mucus, percent dry weight Glycoprotein as Compound BSM-P* Kent (1967) BSM-P* Present Study Mucus** S i a l i c Acid 30.8 11.0 17.3 Fucose 1.8 0.7 1.1 Neutral Hexose 3.6 6.2 10.3 Protein 36.2 68.0 46.0 % accounted for 72.4 85.9 74.7 Bovine submaxillary mucin as isolated by the Pigman method. ** Dialysed mucus from 5% seawater-adapted Leptocottus. 47 Comparative Histochemistry of Mucus The terminal carbohydrate residues of the carbohydrate moieties in mucopolysaccharides dictate the ionic characteristics of these macro-molecules. The terminal sugars have no ionic charge (e.g., Fucose), weak negative charge ( s i a l i c acid), or strong negative charge (sulphated sugars). The three corresponding protein + polysaccharide macromole-cules are: neutral mucopolysaccharide, s i a l i c acid-containing glycopro-tein, and sulphated mucopolysaccharide, respectively. The histochemical c r i t e r i a for distinguishing among these mucins are shown in Table 2. The methods employed here were AB (pH 2.5)-PAS, AB (pH 1.0)-PAS, mild methyla-tion-AB (pH 2.5)-PAS, and active methylation-AB (pH 2.5)-PAS. Ten species of the family Cottidae were examined histochemically to determine the nature of the skin and g i l l mucus secretions. Five of the species are seawater-resident, four species are fresh water-resident, and one species, Leptocottus armatus, is estuarine and a normal resident of both habitats. Three c e l l populations were considered in the compari-son; large goblet cells of the skin, small goblet (or cuticle-secreting) cells in the skin, and the goblet cells of the g i l l epithelium. Two or more fish from each species were examined except for the species Ascelichthys rhodorus, Oligocottus snyderi, and Myxocephalus polyacan- thocephalus for which only one individual of each was available. Large Goblet Cells of the Skin The histochemistry of the major mucus-secreting cells of the skin is summarized in Table 7. A l l the seawater-resident species had neutral Table 7. Comparative Histochemistry of Major Mucus-secreting Cells of the Skin M, magenta (PAS-positive); B, blue (AB-positive); P, purple (both AB and PAS-positive); 1, 2, 3, and 4 reflect the intensity of staining AB-PAS AB-PAS Methylation-AB (pH 2.5)-PAS Type of Habitat & Species Cell Size # Cells pH 2 .5 pH 1.0 Mild Active Mucin Seawater: Ascelichthys rhodorus 65-90 ym 16/mm 4 M 4 M 3 M 3 M neutral Oligocottus maculosus 60-100 ym 28/mm 4 M 4 M 3 M 3 M neutral Oligocottus snyderi 50-85 ym 20/mm 3 M 3 M 3 M 3 M neutral Icelinus borealis 50-80 ym 13/mm 1 M 1 M 2 M 2 M neutral Myxocephalus polyacanthocephalus 10-18 ym 30/mm 4 M 4 M 3 M 2 M neutral Estuarine: Leptocottus armatus 25-30 ym 30/mm 4 B 4 M 4 M-P 4 M acid (COOH) Fresh water: Cottus asper 30-40 ym 25/mm 4 B 4 B 4 B 3 M acid (so 4) Cottus bairdi 40-60 ym 15/mm 4 B 2 P 3 B 2 M acid (so 4) Cottus aleuticus 70-100 ym 22/mm 4 B 2 P 3 B-P 3 M acid (so 4) Cottus rhotheus 40-60 ym 20/mm 4 B 2 P 4 B 2 M acid (SO ) 4> oo 49 mucopolysaccharide i n the large mucus-secreting cells (Fig. 18, 19, 20; Table 7) while the corresponding cells of the fresh water, resident species contained sulphated mucopolysaccharide (Fig. 21, 22; Table 7). The estuarine species, Leptocottus armatus, had s i a l i c acid-containing glyco-protein i n these cells (Fig. 23; Tables 5 and 7), a type of mucin that i s intermediate between the neutral and sulphated mucopolysaccharides of the other two groups. The histochemistry of the large goblet cells of Leptocottus skin did not appear to vary with acclimation to sea water or 5% sea water (Table 7). The size and number of large goblet cells in the skin of the ten species showed no clear trend according to habitat s a l i n i -ty, but the small sample size did not allow proper s t a t i s t i c a l analysis. Though the museum specimens of Cottus bairdi, C_. aleuticus, and C_. rhotheus were as much as twenty years old, the integrity of the skin was well pre-served (Fig. 22), and did not markedly di f f e r from the freshly prepared tissues of Cottus asper (Fig. 21). Minor Mucus-secreting Cells of the Skin The histochemistry of the 'minor' mucus-secreting cells in the skin (small goblet cells and cuticle-secreting cells i n the upper layers of the epidermis) is summarized in Table 8. While the seawater-resident species had various forms of these cells (Figures 18-20; Table 8) the fresh water-resident species lacked this c e l l type. Thus the fresh water-resident species have only one form of mucus-secreting c e l l in the skin. Seawater-resident Leptocottus armatus had prominent cuticle-secreting cells (Fig. 23a) but these cells were not stained i n Leptocottus that were captured in fresh water (Fig. 23b). Other than Leptocottus the only species that 50 Skin of Seawater-Resident Cottlds Figure 18a. Oligocottus maculosus skin stained with AB (pH - 2.5)-PAS, showing large goblet cells (C) containing neutral mucins (AB-negative) that secrete through common pores (P). xl60. Figure 18b. Oligocottus maculosus skin. AB (pH 2.5)-PAS, A few small goblet cells (g) appear to secrete acidic mucins. The eosinophilic granular cells (E) , are unstained, x 625 Figure 19a. Icelinus borealis skin. AB (pH 2.5)-PAS, showing large goblet cells (G), containing neutral mucins, and eosinophilic granular cells (E), unstained. xl60. Figure 19b. Icelinus borealis skin. AB (pH 2.5)-PAS, note the numerous small goblet cells (g) that stain blue, indicat-ing the presence of acidic mucins. x625. Figure 20a. Ascelichthys rhodorus skin, AB (pH 2.5)-PAS, showing the extremely large goblet cells (G), containing neutral mucins, and the few small goblet cells (g), and the un-stained granular cells (E). xl60. Figure 20b. Ascelichthys rhodorus skin. AB (pH 2.5)-PAS, note the extremely thick mucus coat (C) being secreted from the large goblet' cells (G), and the unusual tubule (T) through which the small goblet cells (g) secrete. x625. 50a, 51 Skin of Fresh water-Resident/Euryhaline Cottids Figure 21. Cottus asper skin stained with AB (pH 2.5)-PAS. Note that the goblet cells (G) stain with Alcian Blue, i n d i -cating the presence of acidic mucins, and that minor mucus-secreting cells are lacking, x 625. Figure 22. Cottus aleuticus (U.B.C. Vertebrate Museum #BC-59-495) skin stained with AB (pH 2.5)-PAS. The large goblet cells (G) of this species also appear to contain acidic mucin, and the skin also lacks minor mucus ce l l s , x 625. Figure 23a. Leptocottus armatus (seawater-acclimated) skin stained with AB (pH 2.5)-PAS. The goblet cells (G) and cuticle-secreting cells (C) both appear to contain acidic mucins since both c e l l types stain with Alcian Blue, x 625. Figure 23b. Leptocottus armatus (5% seawater-acclimated) skin stained with AB (pH 2.5)-PAS. Note that the goblet cells (G) stain similarly to those in the seawater-acclimated Leptocottus, but the cuticle-secreting cells (C) are not stained with Alcian Blue in this 5% seawater-acclimated fish, x 625, 5/cx Table 8. Comparative Histochemistry of the Minor Mucus-secreting Cells of the Skin. B, blue (AB+); M, magenta (PAS+); p, purple (AB & PAS+) Methylation-Habitat and Species Cell size # Cells* AB-PAS pH 2.5 AB-PAS pH 1.0 AB (pH 2.5)-PAS Mild Active Type of Mucin Sea water: Ascelichthys rhodorus 12 .um 25/mm 4 B 4 B 3 B 2 P acid (S0 4) Oligocottus maculosus 9 ym 9/mm 4 M 4 M 4 M 4 M neutral 9 ym few 4 B 4 B 3 B 3 P acid (S0 4) Oligocottus snyderi 9-12 ym 12/mm 4 M 4 M 4 M 4 M neutral 6x9 ym ** 3 B - - - acid (COOH) Icelinus borealis 9-15 ym 40/mm 4 B 4 B 4 B 4 M acid (S0^) Myxocephalus polyacanthocephalus none Leptocottus armatus 6x9 ym A* 4 B 3 P 3 B 3 P acid (S0 4) Fresh water: Leptocottus armatus - ** Cottus asper - none Cottus bairdi - none Cottus aleuticus - none Cottus rhotheus - none * number of cells per mm of 7 ym section. ** Cuticle-secreting cells (see Fig.23a ). The cuticle-secreting cells of Leptocottus i n fresh water do not normally stain with Alcian blue. 53 had cuticle-secreting cells that stained with Alcian Blue was Oligocottus  snyderi. The other cottids instead had small goblet cells i n the upper layers of the epithelium; these cells showed a wide variation of structure and histochemistry (Fig. 18, 19, 20). A l l of the cottids examined, save Myxocephalus, produced acidic mucins in at least some of the mucus cells of the skin. G i l l Mucus Cells The histochemistry of the mucus-secreting cells of the g i l l epithe-lium (Table 9) showed that both seawater- and fresh water-resident cottids produce acidic mucins, but only the sea water species secrete neutral mucins. The estuarine Leptocottus armatus produces large quantities of acidic mucins in the g i l l , while the fresh water-resident species secrete only small amounts of mucus in the g i l l . The only species that had num-bers of g i l l mucus cells comparable to Leptocottus were Ascelichthys and Myxocephalus (Table 9); the latter two species are also the only species examined that had neutral mucopolysaccharides i n their g i l l mucus c e l l s . In summary, no neutral mucopolysaccharides are produced in the mucus cells of fresh water-resident cottids, while a l l of the seawater-resident species, save Leptocottus, produce this type of mucin in at least one c e l l type i n the skin. A l l of the cottids examined had large numbers of goblet cells in the skin. Most of the stenohaline (fresh water or sea water) Cottids had very few mucus cells i n the g i l l , compared to the estuarine Leptocottus armatus. The latter species i s the only cottid that i s commonly found in both sea water and fresh water environments during i t s normal l i f e cycle, and thus i t is the obvious choice in the investigation of the Table 9. Comparative Histochemistry of the Mucus-secreting cells of the G i l l M, magenta (PAS +); B, blue (AB +); P, purple (AB & PAS +) Habitat and Species # Cells* Methylation-AB-PAS AB-PAS AB (pH 2.5)-PAS pH 2.5 pH 1.0 Mild Active Type of mucin Sea water: Ascelichthys rhodorus Oligocottus maculosus Oligocottus snyderi Icelinus borealis Myxocephalus  polyacanthocephalus Estuarine Leptocottus armatus Fresh water: Cottus asper Cottus bairdi Cottus aleuticus Cottus rhotheus 4- 5 aff 2-3 eff 2-4 aff 1- 3 eff 0-2 aff 0-2 eff 2- 4 aff 0-1 eff 6 aff 5- 6 eff 4-6 aff 4-6 eff 1- 2 aff 0-1 eff 0-2 aff 0 eff 0-2 aff 0-1 eff 2- 4 aff 0-1 eff 2 M 4 B 3 B 4 B 4 M 4 B 3 P 3 B 3 B 3 B + M 4 P 2 P 4 B (3 M) 3-4 M 2 M 2 M 3 M 3 P + M 1 M 4 B 3 P 3-4 B 4 M 3 B 2 M 3 BP 3 BP + P + M 4 M 3 M 2 P 3-4 M 2 M 1 M 2 M 2 M + M neutral acid (SO^) acid (S0 4) acid (S04) neutral acid (mixed?) acid (COOH) acid (mixed?) acid (S0 4) acid (COOH) * as estimated using the Mucus Cell Index (Fig. 1 ). 55 effects of salin i t y on mucus secretion and i t s control. Endocrine Control of Mucus Secretion Effects of Hypophysectomy and Prolactin Replacement The following three experiments describe the effects of hypophy-sectomy of 5% seawater-acclimated Leptocottus. Immediately post-opera-tive the fish were placed in 25-30% sea water but during the course of each experiment the salin i t y was decreased to 5% sea water or fresh water; the schedules of these sa l i n i t y changes appear in Table 1. The f i r s t experiment included hypophysectomized and sham-operated (control) fish (Table 10). The number of mucus cells on the efferent (leading) edge of the g i l l filaments was significantly lower i n the hy-pophysectomized group, indicating that the pituitary i s required for maintenance of mucus cells in this part of the g i l l . However, no signi-ficant difference was found in the number of goblet cells in the skin or on the afferent (t r a i l i n g edge of the g i l l ) . The haematocrit of the hypophysectomized group did not diff e r significantly from the control group, but serum sodium concentration was significantly lower i n the hy-pophysectomized group, compared to controls. The second experiment included hypophysectomized and sham-operated groups, as before, and an additional intact control group (Table 11). No difference was observed in the number of mucus-secreting cells of the skin. There were, however, significantly fewer mucus cells on both the afferent and efferent sides of the g i l l in the hypophysectomized group, compared to the pooled values for the other two groups. Table 10. Hypophysectomized vs. Sham-operated Leptocottus Group: Hypophysectomized Sham-operated • n = 8 n = 8  Weight (g): 10.1+0.4 12.3+2.6 Haematocrit (%): 35.4+1.7 33.4+0.9 Serum sodium (mM/1): 129 +5.0 165 +9.7 ''"Gill Mucus Cells * Efferent side: 2.21+0.46 3.36+0.43 Afferent Side: 3.81 + 0.24 4.24 + 0.15 Skin Mucus Cells Goblet c e l l s : 2 25.6 +1.74 27.7+1.45 Cuticle c e l l s : 3 0/8 0/8 As measured with the mucus c e l l index (Fig. 2). 2 Number of cells per mm of 7 ym section. 3 Number of fish with Alcian Blue-Positive cuticle-secreting c e l l s . * P <0.05 ** p <0.001 57 Table 11. Hypophysectomized vs. Sham-operated vs. Intact Leptocottus armatus Group: Hypox. Sham-Hypox. Intact n • = 11 n = 10 n = 10 Weight (g) : 12.7 + 1.1 14.5 + 1 .2 12.4 +2 .6 Haematocrit (%): 36.0*+ 1.5 28.9 + 1 .3 31.8 + 1 .5 Serum sodium (mM/1): 157 + 6.8 166 + 1 .9 166 + 4 .1 G i l l Mucus cells"*" Efferent side: Afferant side: ** 3.16 + 0.37 3.66* + 0.31 3.57 +0 4.24 + 0 .28 .30 3.56 + 0 4.57 + 0 .47 .21 Skin Mucus Cells 2 Goblet Cells: 3 Cuticle Cells: 37.4 + 1.95 ** 2/11 36.3 + 2 4/10 .63 38.1 + 1 6/10 .77 """As measured using the mucus c e l l index (Fig. 2). 2 Number of cells per mm of 7 ym section. 3 Number of fish with Alcian Blue-positive cuticle-secreting c e l l s . * P <0.05 ** P <0.01 58 Though no difference was found among the means of the serum sodium data, the modes of the groups (141, 162, and 162 nM Na +, for the hypophysec-tomized, sham-operated and intact groups, respectively) suggest that the hypophysectomized group had lower serum sodium values than the con-trols. The haematocrit of the hypophysectomized group was higher than that of the control groups. The two control groups did not diffe r i n any of the parameters measured, so i t i s unlikely that the stress of anaesthesia and the operation significantly affected long term mucus se-cretion or hydromineral balance. An estimated of the shedding rate of mucus from the fish showed that the hypophysectomized fish shed significantly less mucus (36 + 3.4 uM hexose/100 g day, n = 6) than the pooled values for the other two groups (47 +3.0 yM hexose/100 g day, n = 12); a one-tailed t-test showed that p <0.025. The fish had been starved for nineteen days, so the contribu-tion of the gut to assayable hexose i n the surrounding water was con-sidered negligible. There was no difference between the mucus shedding estimates of the intact and sham-operated control groups. Thus i t ap-pears that a reduction i n the number of mucus cells reflects a reduction in the amount of mucus shed by the fis h . The third experiment (Table 12) included hypophysectomized and sham-operated fi s h . The hypophysectomized. group was divided into prolactin-treated (13.4 yg/g I.P. Ovine Prolactin, NIH-P-S-11) and vehicle-injec-ted controls; the shams received injections of the vehicle alone. There was a significant drop in the number of mucus cells on the efferent side of the g i l l i n the hypophysectomized control f i s h , compared to the pooled 59 Table 12. Prolactin replacement of hypophysectomized Leptocottus armatus Sham + Hypox. + Hypox. + Treatment: Vehicle Vehicle Prolactin n = 9 n = 11 n = 9  Dosage (yg/g): 1 - - 13.4+1.1 Weight (g): 17.4+1.4 17.1+1.7 17.9+1.4 Haematocrit (%): 33.3+1.5 29.5+1.4 29.4+1.8 Serum sodium (mM): 152 +1.2 134 +5.6 150 +4.9 2 G i l l Mucus Cells Efferent side: 3.94 + 0.36 3.17*+ 0.24 3.84 + 0.31 Afferent side: 3.72+0.14 4.20+0.64 3.78+0.96 Skin Mucus Cells Goblet C e l l s : 3 30.4+0.36 26.9+1.58 28.8+1.58 Cuticle C e l l s : 4 5/9 8/11 1/10** '''Ovine prolactin (NIH-P-S-11) . Dose based on weights at the end of the experiment. 2 As measured using the mucus c e l l index (Fig. 2). 3 Number of cells per mm of 7 ym section. Number of fish with Alcian Blue-positive cuticle-secreting c e l l s . * F<0.05 ** P<0.01 *** P<0.01 60 values of the other groups, but the number of mucus cells i n the skin and the afferent side of the g i l l did not d i f f e r among the three groups (Table 12). Serum sodium concentration was significantly lower in the hypophysectomized fi s h without prolactin, compared to the pooled values for the other two groups. The sham-operated and hypophysectomized fish receiving prolactin had similar serum sodium values, demonstrating the well-recognized 'sodium-retaining' effect of prolactin in teleosts. There were no apparent differences in haematocrit among the three groups. The above (three) experiments show that the pituitary gland is re-quired to maintain serum sodium and the number of mucus-secreting c e l l s in the g i l l s of Leptocottus, and that replacement therapy with ovine prolactin can prevent the decreases i n serum sodium and the number of g i l l mucus cells of hypophysectomized Leptocottus in hyposmotic media. Effects of Environmental Salinity. Intact Leptocottus that were acclimated to full-strength sea water for 21 or 71 days had significantly fewer mucus-secreting cells on the efferent (leading) edge of the g i l l filaments, compared to controls that were acclimated to 5% sea water (Table 13). The number of goblet cells in the skin and on the afferent side of the g i l l did not change with sea water adaptation. The reduction in the number of mucus-secret-ing cells on the efferent side of the g i l l of seawater-acclimated fish was accompanied by a decrease in the apparent activity of the prolactin-secreting eta cells of the rostral pars d i s t a l i s , as determined by mea-surements of the nuclear diameter of these cells (Table 13, Fig. 24, 25). However, there was no apparent difference in the ACTH-secreting epsilon cells (Table 13, Fig. 24, 25). Significantly more of the seawater-61 Table 13. Effects of Environmental Salinity: I Leptocottus armatus 5% SW 100% SW 100% SW Salinity; n = fi n = 8 n = 6 Days Acclimation: 21+ 21 71 Weight (g): 21.1+2.1 30.2+2.5 26.3+2.0 Haematocrit (%): 38.8+1.8 35.6+2.9 42.7+3.6 Serum sodium (mM/1): G i l l Mucus C e l l s 1 Efferent Side: 4.9*+ 0.3 3,7+0.4 2.7+0.6 Afferent Side: 5.2+0.2 5.2+0.3 4.7+0.5 Skin Mucus Cells Goblet Cells: 28.8 + 2.0 28.6 + 1.6 24.0 + 1.6 (// cells/mm) Cuticle Cells: 2/8 14/14 Prolactin Cell Nucleus diam. (ym): 3.8+0.05 2.9 +0.08 (n = 5) (n = 5) ACTH Cells Nucleus Diam. (ym): 4.3 + 0.1 4.2 + 0.09 (n = 6) (n = 6) As measured using the Mucus Cell Index, Fig. 2. 2 Number of fish with Alcian Blue-positive cuticle-secreting c e l l s . * P <0.01 ** P <0.005 *** P <0.001 62 Figure 24. Rostral pars d i s t a l i s of 5% seawater-acclimated Leptocottus armatus, showing prolactin-secreting cells (P) , ACTH cells (A) , and neurohypophysis (N). Note the large nuclei of the prolactin cells and the mitotic figure (arrow) among these cel l s , x 625. Figure 25. Rostral pars d i s t a l i s of seawater-acclimated Leptocottus armatus. Note that while the ACTH cells (A) appear similar to the figure above, the prolactin cells (P) have smaller, less distinct nuclei, x 625. c 62 a 63 acclimated fish showed prominant staining of the epidermal cuticle-secreting cells (Table 13) . While Leptocottus can survive for months in 5% sea water or f u l l -strength sea water, this species does not tolerate very low s a l i n i t i e s (less than 1 o/oo) for more than two days. Unfed Leptocottus die in fresh water in 24 to 40 hours (LT 50 was 35 hours for 10 small sculpins), but dissolved calcium (100 ppm CaC^) appeared to improve survival (only one of ten fish died within 45 hours). Prior to death (by osmore-gulatory failure) the fish ceased active swimming and rested on the tank bottom with a l l fins folded close to the body. L i t t l e movement was apparent in these stressed fi s h except for slight and slow opercular movements during ventilation of the g i l l s . This typical behaviour of stressed sculpins may serve to decrease salt loss through increasing "unstirred" layers adjacent to the body. Cottus asper, a fresh water-resident cottid, had significantly higher serum sodium concentration after 47 days of adaptation to f u l l -strength sea water, compared to controls that were held in dechlorinated fresh water (Table 14). The sea water group also had significantly fewer mucus-secreting goblet cells in the skin, but neither group had alciano-p h i l i c cuticle-secreting cells i n the skin. Both groups had very few goblet cells in the g i l l epithelium; the mean mucus c e l l index for the g i l l areas was less than one (Table 14). The g i l l epithelium of the sea water group was grossly hypertrophied and some of the g i l l filaments had epidermis so thick as to entirely obscure the thin membranes of the g i l l lamellae (Fig. 28, 29). Five fish that were l e f t in sea water at 64 Table 14. Effects of Environmental Salinity: II Cottus asper Salinity: De-Cl FW n = 13 Sea water n = 14 Days Acclimation: 47 47 Weight (g): 17.3+1 .3 24.3 + 2.7 Haematocrit (%): 36.7 + 2 .4 (10) 39.4 + 3.5 (7) Serum Sodium (mM/1): 162 + 4 .2 (10) 181* +3.8 (12) G i l l Mucus C e l l s 1 Efferent Side: Afferent Side: none 0.7 + 0 .2 none 0.5 + 0.2 Skin Mucus Cells 2 Goblet Cells: 3 Cuticle Cells: 27.4 + 0 none .9 * * 11.0 + 2.0 none Prolactin Cells Nucleus Diam. (ym): 4.2 + 0 .1 (11) ** 3.5 + 0.1 (13) As measured using the Mucus Cell Index, Fig. 2. 2 Number of cells per mm of 7 ym section. 3 Number of fish with Alcian Blue-positive cuticle-secreting c e l l s . * P <0.01 ** P <0.001 65 Figure 26. Rostral pars d i s t a l i s of fresh water-adapted Cottus  asper; Masson's Trichrome. This section shows the (presumed) prolactin-secreting (P) eta cells and the (presumed) ACTH-secreting (A) epsilon c e l l s . Note the large nuclei and prominent nucleoli of the prolactin c e l l s , compared to those in Fig. 27, below, x 625. Figure 27. Rostral pars d i s t a l i s of seawater-adapted Cottus asper; Masson's trichrome. Here the prolactin cells (P) and their nuclei are smaller and the nucleoli less distinct, indicating a low activity of the prolactin cells and presumed hyposecretion of prolactin, x 625. 65a. 66 Figure 28. Cottus asper, fresh water-acclimated, showing g i l l filament stained with AB (pH 2.5)-PAS. Note the small number of mucus cells (M) , g i l l lamellae (L) , -afferent and efferent (A,E) blood vessels and sup-porting cartilage (C). x 160. Figure 29. Cottus asper, seawater-adapted, showing a cross-sec-tion of g i l l filament with grossly hypertrophied epi-thelium. The lamellae (L) are seen embedded in the thick epidermis; this situation would certainly result in a reduction of surface area for ion and gas exchange across the g i l l , x 160. 67 the end of the above experiment were found dead on days 53 to 55. It appears that the hypertrophy of the g i l l epithelium in response to high salinity ultimately contributed to the deaths of those fish that were l e f t in sea water for more than seven weeks. The prolactin-secreting eta cells of the sea water group had significantly smaller nuclear diameter, compared to fresh water controls (Table 14, Fig. 26, 27). Thus the reduction i n the number of goblet cells in the skin of Cottus asper during sea water adaptation is accompanied by a decrease in the apparent activity of the prolactin c e l l s . Prolactin Injection of Seawater-acclimated Leptocottus Prolonged treatment of seawater-acclimated Leptocottus armatus with Ovine prolactin (NIH-P-S11; 11.7 ug/g in each of 16 injections I.P. over 25 days) caused a significant rise in serum sodium over vehi-cle-injected controls (Table 15). There was no difference in haematocrit between the two groups. The increase i n serum sodium was accompanied by a significant increase in the number of goblet cells on both the afferent and efferent sides of the g i l l filaments (Table 15) but no difference was seen in the number of mucus cells in the skin. The cuticle-secret-ing cells in the skin of seawater-acclimated Leptocottus stain strongly with Alcian Blue, but prolactin injection abolished this feature of these cells (Table 15). The nuclei of the prolactin c e l l s of the rostral pars d i s t a l i s were small and pycnotic in both groups. The nuclei of the ACTH cells were more round ed and significantly larger than the pro-lactin c e l l nuclei, though there was no apparent difference between the ACTH. cells of the control and prolactin-injected groups. 68 Table 15. Effects of Prolactin on Seawater-acclimated Leptocottus armatus Treatment: Prolactin injected Vehicle injected n = 12 n = 12 Dosage (yg/g): 11.7 + 0.65 -Weight (g): 17.2 + 0.83 15.9 + 0.83 Haematocrit (%): 29.0 + 1.22 32.1+1.1 Serum Sodium (mM): 188 + 2.0 167 +4.5 G i l l Mucus C e l l s 1 Efferent Side: ** 4.9 + 0.2 3.9 + 0.3 Afferent Side: 4.3*+ 0.2 3.7 + 0.2 Skin Mucus Cells Goblet C e l l s : 2 27.2 + 1.3 23.5 + 1.8 3 Cuticle Cells: *** 1/12 12/12 ACTH Cell Nucleus Diameter (ym): 3.4 + 0.06 (10) 3.4 + 0.05 (10) Prolactin Cell (Prolactin vs . ACTH c e l l s : P <0.001) Nucleus Diam. (ym): 4.0 + 0.06 (10) 4.1 + 0.08 (10) As measured using the Mucus Cell Index, Fig. 2. 2 Number of goblet cells per mm of 7 ym section. 3 Number of fish with Alcian Blue-positive cuticle-secreting c e l l s . * P <0.05 ** P <0.01 ***P <0.001 69 While prolactin treatment of seawater-acclimated sculpins apparent-ly stimulates the mucus-secreting cells of the g i l l , this does not cor-relate with an osmoregulatory function for the mucus coat since this treatment also resulted in a detrimental increase i n serum ion concentra-tion. Cortisol Treatment of 5% Seawater-adapted Leptocottus Long term treatment with C o r t i s o l (ten injections I.P. of 10 'ug/g each over ten days) caused a significant decrease in the number of mucus-secreting cells on the efferent side of the g i l l , compared to vehicle-injected controls (Table 16). There was no significant difference in haematocrit or serum sodium between the two groups. The serum sodium val-ues for the control group were particularly high—compared to previous experiments u t i l i z i n g 5% seawater-acclimated Leptocottus (Tables 11 and 12); this was probably caused by dietary salt intake, since these fish were fed regularly with brine shrimp. Though the mucus cells on the ef-ferent side of the g i l l were apparently inhibited by C o r t i s o l treatment, the mucus cells on the afferent side of the g i l l and the skin mucus cells were not affected (Table 16). However, the cuticle-secreting cells ap-pear to be stimulated by the treatment, since significantly more of the cortisol-injected fish showed the characteristic strong staining with Alcian Blue. The diameter of the ACTH c e l l nuclei was significantly re-duced in the cortisol-treated group, compared to controls, while the pro-la c t i n cells were not apparently affected (Table 16). Cortisol treatment of 5% seawater-acclimated Leptocottus results in an apparent mimicking of 70 Table 16. Effects of Cortisol in 5% Seawater-acclimated Leptocottus Cortisol-injected n = 6 Vehicle-inj ected n = 7 Dosage (yg/g):^ Weight (g): Haematocrit (%): Serum Sodium (mM) : G i l l Mucus C e l l s 2 Efferent Side: Afferent side: 10 yg/g 86 + 11 33.5 + 1.7 163 + 8.0 AA 3.0 + 0.4 4.7 + 0.6 82 + 17 34.6 + 2.8 180 + 13.1 4.3 + 0.4 4.9 +0.3 Skin Mucus Cells Goblet C e l l s : 3 4 Cuticle Cells: 25 + 1.5 , * 5/6 26 + 1.5 3/7 ACTH Cell nucleus diameter (ym): Prolactin Cell Nucleus diam. (ym) 4.15*+*0.08 4.75 + 0.08 (prolactin vs. ACTH c e l l s : P <0.001) 4.12 + 0.04 4.11 + 0.06 The fish were injected different volumes according to their known weights. 2 As measured using the Mucus Cell Index (Fig. 2). 3 Number of goblet cells per mm of 7 ym section. 4 Number of fis h with Alcian Blue-positive cuticle-secreting c e l l s , divided by n. * P <0.05 ** P <0.025 AAA p <0.001 71 the effects of seawater acclimation, at least in the number of g i l l mucus cells and the staining of the cuticle-secreting cells (compare Tables 13 and 16). However, unlike seawater-acclimation, C o r t i s o l treatment does not apparently inhibit the prolactin-secreting cells in the pituitary. Hypophysectomy and Hormone Replacement in Seawater Sculpins This experiment tested the role of the pituitary gland i n seawater osmoregulation and the maintenance of the mucus cells of the skin and g i l l . Hormone replacement therapy used C o r t i s o l , ovine growth hormone (NIH-GH-S11), and ovine prolactin (NIH-P-S12); the mean doses, given by intraperitoneal injections, were 12.3+1.0, 11.6+1.0, and 11.3 + 0.6 Vg/g, respectively. The control fish i n the sham-operated and hypophysec-tomized groups received saline (0.9% NaCl) injections. The serum sodium concentration of the prolactin injected group was significantly higher than any of the other groups (Table 17) ; SS-STP showed that P <0.01 when the prolactin-treated group was compared to the control groups. This result was expected, since prolactin i s well known for i t s sodium-retaining action in teleost osmoregulation, but here the result should be considered pathological, since there was no significant d i f f e r -ence between the sham-operated and hypophysectomized groups. There was significant v a r i a b i l i t y among the five groups i n the number of mucus-secreting cells on the efferent side of the g i l l (P <0.01) as shown by the Kruskal-Wallis one-way analysis of variance (Table 17). The following set of orthogonal comparisons was established on the basis of previous experiments concerning the effects of hypophysectomy on g i l l Table 17. Hypophysectomy and Hormone Replacement of Seawater-adapted Leptocottus Treatment: Sham + Saline Hypox. + Saline Hypox. + Cortisol Hypox. + STH Hypox. + Prolactin n=8 n=8 n=9 n=9 n=9 Weight (g): 9.8 + 1.1 10.0 + 0.05 7.5 + 0.5 8.3 + 0.7 9.2 + 0.5 Haematocrit (%): (**) 27.5 + 1.2 29.1 + 1.6 41.9 +1.6 36.9 + 1.4 31.4 + 1.3 Serum Sodium (mM): (**) 158 +2.9 163 +4.8 163 + 1.3 165 + 3.1 185 +4.6 G i l l Mucus C e l l s 1 Efferent Side): (***) 3.4 + 0.6 2.2 + 0.3 2.8 + 0.6 3.3 + 0.4 3.2 + 0.6 Afferent Side: (***) 4.8 + 0.2 4.1 + 0.3 4.2 + 0.3 4.3 + 0.2 4.3 + 0.3 Skin Mucus Cells Goblet C e l l s : 2 (*) 18.5 + 2.6 21.8 +2.6 14.7 + 2.0 22.4 + 2.1 16.0 + 0.9 Cuticle C e l l s : 3 (++) 8/8 1/8 7/9 4/9 0/9 ^ s measured by the Mucus Ce l l Index, Fig . 2. l Number of goblet cells per mm of 7 ym section. 3 Number of fish with Alcian Blue-positive cuticle-secreting c e l l s . * P <0.05, Analysis of Variance, one way. ** P <0.01, One way analysis of variance. *** P <0.001, Kruskal-Wallls one-way analysis of variance (nonparametric). + P <0.001, see text for details. 73 mucus c e l l s , and on expectations that C o r t i s o l would in h i b i t , but growth hormone would stimulate these cel l s . This set of comparisons tests whether (1) hypophysectomy affects the mucus c e l l s , (2) prolactin and growth hormone have a similar action in maintaining mucus c e l l s , and (3) C o r t i s o l has an inhibitory action, compared to growth hormone and prolactin. The results of Mann-Whitney U-tests on the above comparisons were: (1) Hypophysectomy significantly reduces the number of mucus cells on the efferent side of the g i l l (P <0.025) (2) Prolactin and growth hor-mone have a similar stimulatory effect on these c e l l s . (3) The effect of C o r t i s o l is not significantly different from that of the pooled results of the two pituitary hormones. Thus i t appears that prolactin, growth hormone, and C o r t i s o l can maintain at a high level the number of mucus cells on the efferent side of the g i l l , but hypophysectomy causes a s i g -nificant reduction in the number of these cells (Table 17). The same set of comparisons was used on the data regarding the afferent side of the g i l l , and the results were: (1) Hypophysectomy significantly reduces the number of mucus cells on the afferent side of the g i l l , (P <0.025). How-ever, i t appears that none of the hormone treatments were effective i n r maintaining the number of mucus cells in this area of the g i l l at levels near that of the sham-operated controls, The staining of the. cuticle-secreting cells was examined using the same set of orthogonal comparisons that was used for the g i l l mucus cells (above). The Fisher exact probability test was used with this nominal data. Fewer of the hypophysectomized control f i s h showed Alcian Blue 74 'Staining of the cuticle secreting cells (P <0.001, Table 17), Growth -hormone^ replacement appeared to restore the staining of these c e l l s , compared to the prolactin-injected hypophysectomized f i s h . However, C o r t i s o l treatment of the hypophysectomized fish restored the staining of these cells i n significantly more f i s h than growth hormone and pro-l a c t i n (P <0.01). Though the cuticle-secreting cells of both the hypophysectomized control fish and the prolactin-replaced group did not stain with Alcian Blue, the cells of the prolactin injected group appeared to be larger and more tightly attached to the underlying layers of the epithelium. This implies also that the unstained cuticle-secreting cells of the hypophysec tomized group are less active than the corresponding cells of the c o r t i -«ol-injected hypophysectomized f i s h (Fig. 30, 31, 32). One-way analysis of variance revealed a marginally significant d i f -ference i n the number of skin goblet c e l l s among the treatments (Table 17) ; SS-STP showed that only when a l l four of the hypophysectomized group were considered, was the n u l l hypothesis rejected. Hypophysectomy had -no apparent effect on the skin mucus c e l l s , so i t i s l i k e l y that the vari Aance inherent, i n the hormone-treated groups was due to pathological ac-t i o n s of these substances. An estimate of the growth rate of the skin was based on the number o 2 -mitotic figures per 8000 ym of 7 ym sections of skin. The sham-operated and hypophysectomized control groups had low values (6.1 + 0.5 and 8.4 + 0.8, respectively) that were not significantly different when examined by the SS-STP method after analysis of variance. The prolactin-jcortisol-, *Sum of squares Simultaneous Test Procedure, an a posteriori test (Sokal and Rolf, 1969). 75 Figure 30. Figure 31. Figure 32. Skin from hypophysectomized, seawater-acclimated Leptocottus (vehicle-injected); AB (pH 2.5)-PAS. In these fish the cuticle-secreting cells are small, some-times sickle-shaped, and are loosely attached to the un-derlying cells (arrows). This figure represents the 'inactive' state of the cuticle-secreting c e l l s . Skin from hypophysectomized, seawater-adapted Leptocottus injected with C o r t i s o l ; AB (pH 2.5)-PAS. Here the cuticle-secreting cells are larger, stain darkly with A l -cian Blue, but are again only loosely attached to the underlying cells (arrows). This figure represents the 'moderately active' state of the cuticle-secreting c e l l s . Skin from hypophysectomized, seawater-adapted Leptocottus Injected with prolactin; AB (pH 2.5)-PAS. Note that the cuticle-secreting cells here are larger than those in the above two figures and that the cells do not stain with Alcian Blue, as in Fig. 31. These cuticle-secreting cells are in the 'highly active' state. 7Stx v . " , 76 and growth hormone-replaced hypophysectomized fish a l l had significant-ly higher (P <0.05) number of mitotic figures (12.4 + 1.0, 12.7 + 1.0, 2 and 10.7 +0.9 mitoses/8000 ym section, respectively), compared to the control groups. Analysis of variance showed that there was significant variance among the five groups (P <0.01), but the apparent increases in the number of mitotic figures in the hormone-replaced fi s h was probably a pathological effect, since there was no significant differences between the hypophysectomized and the sham-operated controls. One-way analysis of variance of the haematocrit data showed that there was significant variation among the groups (Table 17). An a posteriori test (SS-STP, Sokal and Rolf, 1969) revealed that the two control groups and the prolactin-treated group contributed l i t t l e to the variance among the five groups, but both the growth hormone-treated and the cortisol-treated groups had higher haematocrit (P <0.01) than the control groups. These effects of growth hormone and C o r t i s o l are probably pathological, since hypophysectomy did not affect haematocrit, compared to the sham-operated controls. Summary The pituitary is not required for the survival of Leptocottus i n sea water, nor for the maintenance of normal serum sodium. The goblet cells of the skin are also unaffected by hypophysectomy, but the Alcian Blue, staining of the cuticle-secreting cells i s inhibited. Prolactin re-placement increases the number of mucus cells in the g i l l s of hypophysec-tomized fis h , but also causes a (detrimental) increase in serum sodium. Prolactin treatment does not apparently affect the goblet cells of the 77 skin nor does this treatment restore Alcian Blue staining of the cuticle c e l l s . Growth hormone and Cortisol injection increase the number of g i l l mucus cells without affecting serum sodium, and also restore the alcianophilia of the cuticle-secreting c e l l s . Physical Properties of Mucus There are three ways by which mucus may act to reduce the permeabili-ty of the integument to ions. F i r s t , the ionic groups, primarily acidic sugars ( s i a l i c acid), of the glycoprotein may act as ion exchangers. The negative charges of these acidic sugars may tend to exclude from the mucus coat other negatively charged particles, while allowing positively charged ions (the counter-ions) to pass freely through the mucus layer. Second, mucus may reduce the diffusion coefficient of water and/or ions passing through the mucus coat through steric interference of molecular movement. Third, the mucus coat may reduce the concentration gradient across the integument through enhancement of 'unstirred' layers adjacent to the outer surface. A reduction in the concentration gradient would reduce the driving force for ions passing through the membrane, and thus decrease the observed permeability of the membrane. The f i r s t two mechan-isms of an action for mucus i n osmoregulation are examined here, while the third mechanism pertaining to the role of unstirred layers appears as a theoretical section i n the discussion. Titration of Fixed Acidic Groups The concentration of ionic groups that can freely exchange counter-ions represents the ion-exchange capability of the substance, i n this 78 case, mucin. The addition of 0.5 M calcium chloride to a small v o l -ume of acidified, de-ionized mucus caused a drop i n pH from 4.5 to 3.84, reflecting the release of hydronium ions from fixed charges in the mucus in exchange for ionic calcium. A total of 113 p.Eq/1 of H + was released, or expressed per gram of mucus solids after freeze-drying, 80 ;iEq/g mucus. A similar t i t r a t i o n used 3.0 M KCl on the re--maining half of the mucus sample (Figure 33). The concentration of fixed charges appeared to be higher after t i t r a t i o n with CaCl^ than with KCl, probably reflecting a higher a f f i n i t y of the mucus for d i -valent cations. In either case, the concentration of fixed acidic groups in mucus is lower than that of polyanionic ion-exchange resins (for ex-ample, 3500 pEq/g in DEAE-sephadex; Pharmacia Co.). Thus i t is unlikely that a thin layer of mucus with a limited ion-exchange capacity could significantly affect the passive movement of ions across the integument.. However, there remains the possibility that the mucus, through i t s ion exchange properties, might act in conjunction with active ion uptake mechanisms in freshwater f i s h . This p o s s i b i l i t y i s discussed further on page 111. Determination of Diffusion Coefficient (D) Through Mucus The amount of steric interference due to the mucus coat was tested by estimating the self-diffusion coefficient of ions through a column of mucus. The diffusion coefficients of Na and cl through Leptocottus mucus are the same (+ 10%) as those through saline (Table 18). This was true for both high (100 mM) and low (10 mM) NaCl salines. The F ig . 33 Ti t rat ion of sk in mucus w i th KCl f o r es t imat ion -of f ixed c h a r g e d e n s i t y 4-5-8 32 4-3 PH 4-1-39-sample v o l . : 3-9 ml sample w t . : 5-5 mg 97 x 3-9 5-5 = 6 9 j j E q / g O h 5 0 97LjEq/ [ H + ] -o (x10 6 M) •100 •129 3-7- H200 I—r -0 10 35 8 5 )j| 3-0M KCl 185 Table 18. Determination of Self-Diffusion Coefficient of Na and Cl through a skin mucus solution from Leptocottus armatus Temperature of bath = 15 + 0.1 C Length and inside diameter of capillary tubes - 2,0 + 0.1 cm x 0.076 cm Saline for experiments 1-3; 100 mM NaCl, and for experiment 4;10.0 mM NaCl. Expt. # 1. 2 3 4 Radionuclide 22. Na 22. Na 36 Cl 36 Cl S t i r Rate 225 RPM 150 RPM 225 RPM 225 RPM Diffusion Time 8.58 x 10 sec. 6.33 x 10 sec, 8.01 x 10 sec. 8.70 x 10 sec. saline 1.84 1.86 1.48 1.85 ft* 0 mucus 1.87 1.51 1.91 Protein (mg/ml) 0.93 2.4 1.1 * Lowry Protein assay of mucus sample -5 2 ** Self-Diffusion coefficient (x 10 cm. /sec.) oo o 81 diffusion of sodium or chloride ions i s not apparently Impaired by the presence of low concentrations of mucus. Thus i t i s unlikely that the mucus coat could act by this mechanism to decrease significantly the per-meability of the Integument to these ions. Lubricating Properties of Mucus Rosen and Cornford (1971) have demonstrated that dilute slimes from the skin of fast-swimming species of teleosts are eff i c i e n t f r i c t i o n re-ducers. The lubricating properties of fis h mucus are further investiga-ted here to include effects of mucus concentrations, of ionic strength, and of acclimation of the fish to 5% or full-strength sea water. Finally, a comparison is made of the properties of mucus taken from the seawater-resident cottid, Icelinus borealis, and the euryhaline Leptocottus armatus. The viscosity (cp) of mucus from Leptocottus armatus increases linearly with concentration, and is always greater than that of water (Fig. 34) when the flow rates of the solution through the viscosimeter are such that water would flow in a laminar fashion. However, at faster rates of flow where water flow i s turbulent (Reynolds number exceeding 2000) , the viscosity of dilute mucus solutions i s less than that of water (Fig. 35), and i s not linear with respect to concentration. Instead, there appears to be an optimum range of mucus concentration (1.5-5% v/v) 1 with the least viscosity, relative to water and thus the best lubricating properties. The higher concentrations (10-20% v/v) lubricate less e f f i -Mucus samples scraped from the fish are 1-4 mg/ml solids. The samples are further diluted for use in the viscosimetry. 82 Figure 34. Comparison of mucus from Leptocottus armatus (open c i r -cles) diluted with d i s t i l l e d water, and mucus from Ice- linus boreaiis diluted with d i s t i l l e d water (open triangles) or sea water (closed triangles). The viscosity increases linearly with concentration, within the range tested, and Leptocottus appears to have the more viscous mucus of the two species. A l l readings'were taken at slow (laminar) flow rates. 83 Figure 35 Effect of mucus concentration on the lubricating property of Leptocottus mucus at high flow rates. Percent f r i c t i o n reduction (F.R.) and viscosity (n) are relative to water (water = 0.0% F.R. = 1.0 cp). F.R. (%) Lubr icat ing property of Leptocottus mucus: Ef fect of concentrat ion. R (cp) 40 Pres su re =2-04 atm Reynolds number >2000 0.01 0.1 0.5 1 2 3 4 5 mucus concentrat ion (% v/v) 10 20 00 to 84 ciently, apparently because these solutions have much higher viscosi-ties (Fig. 35). The effects of increasing the flow rate of a mucus solution (0.05 mg/ml; a concentration that yields good lubricating properties) are shown in Figure 36. As described above, the mucus solution was more viscous than water at slow flow rates (Reynolds number less than 2000). As the pressure was increased so that the Reynolds number would exceed 2000, the mucus solution flowed more rapidly through the tube than did water, became effectively less viscous than water, and thus showed lubricating properties, represented as a positive 'friction reduction'. The point at which the 'fr i c t i o n reduction' becomes positive f a l l s almost exactly at the flow rate corresponding to Reynolds number = 2000, thus while water be-gins to flow in a turbulent fashion at this point, the solution which was lubricated by mucus continues to flow in the (more efficient) laminar fashion. Mucus diluted with high ionic strength media (2.5-5.0 M NaCl or 1.0-2.5 M CaC^) loses i t s lubricating property and i s apparently salted out of solution (Table 19). If, however, the mucus is diluted in more 'physio-logical' solutions (sea water or 20 mM CaC^) the lubricating properties are not impaired (Figures 36, 37). While mucus taken from any one fish has similar lubricating proper-ties when i t is diluted with d i s t i l l e d water, 20 mM CaC^ or sea water, i t is apparent that seawater-acclimated sculpins produce mucins that are significantly better lubricants, compared to those of 5% seawater-accli-mated fish (Fig. 37), Since the number of goblet cells i s similar i n the 85 Table 19. Effect of high ionic strength on the lubricating properties of Leptocottus mucus 2.5% (v/v) mucus diluted i n : Ionic Strength F.R.*(%) Dist. water - +19.5 NaCl 0.02 M + 19.7 NaCl 1.0 M + 20.0 NaCl 5.0 M + 14.5 NaCl 10.0 M + 3.1 CaCl 2 CaCl 2 CaCl 2 CaCl 2 0.04 M 2.0 M 4,0 M 10.0 M + 15.2 + 13.3 + 8.1 - 4.0 Percent f r i c t i o n reduction: flow of mucus solution minus flow rate of corresponding salt solution, divided by the latter. The pressure applied in the viscosimeter was 2.04 atm, yield-ing a Reynolds number greater than 2000. 86 Figure 36. Viscosity of Leptocottus armatus mucus (0,05 mg/ml) at different flow rates i n the high pressure viscosimeter Below flow rates equivalent to a Reynolds number of 2000 (arrow) mucus is more viscous than water, and fr i c t i o n reduction (F.R.) is negative. However, at higher pressures (Reynolds number >2000) mucus becomes less viscous than water, demonstrating the lubricating property of mucus. Mucus in d i s t i l l e d water, t^ ) , mucus in sea water, » d i s t i l l e d water alone, A , sea water alone, A . 86a 87 Figure 37. Effect of salinity acclimation on the lubricating property of Leptocottus mucus. The mucus concen-tration was 1:5000 w/v diluted i n sea water (SW), d i s t i l l e d water (DW), or 20 mM CaCl 2. Readings were taken using the high pressure viscosimeter run at 2.7 atmospheres pressure at room temperature (21 + 1 C). Regardless of the solvent, the mucus taken from seawater-acclimated sculpins showed significant-ly (P <0.005; Mann-Whitney U test) higher f r i c t i o n reduction than mucus from the 5% seawater-adapted individuals. F i g . 37. E f f e c t of Salinity Acc l ima t i on on the L u b r i c a t i n g P r o p e r t y of L e p t o c o t t u s Mucus 50 O f— ( J z> Q LU cn z: O LL 40 30 20 y 10 mean + 9 5 % C. N = 6 6 0 L mucus in: SW DW 2 0 m M C a C ^ fish adapted to: 5% S W 6 6 6 SW DW 2 0 m M C a C ^ 100 % SW 88 skin of fish from the two sa l i n i t i e s (Table 13), these c e l l s are pro-bably not responsible for the observed difference in the lubricating properties of the mucus unless the composition of the mucus i n these cells has changed during salin i t y adaptation. Though the goblet cells of Leptocottus from either s a l i n i t y appear similar histochemically (Fig. 23) the possibility of a change in composition (especially in molecular weight of the mucins) can not be entirely excluded. Instead i t may be recalled that the skin of seawater-acclimated Leptocottus only differs from that of the 5% seawater-adapted fish in the Alcian Blue-staining of the cuticle-secreting cells (Table 13). The presence of Alcian Blue-staining cuticle-secreting cells is thus correlated with the superior lubricating properties of slime from the seawater-adapted f i s h . The ultrastructure of the cuticle-secreting cells shows that these cells apparently accumulate acidic mucins, but also show distinct signs of de-gradation (Fig. 11, 12). Thus the superior lubricating property of mucus from seawater-adapted sculpins may be due to the mechanical removal of these cells (and their accumulated secretory material) during the collec-tion of mucus samples. The mucus scraped from the seawater-resident cottid, Icelinus borealis, appears to be less viscous" than Leptocottus armatus mucus (Fig. 34). This difference may be related to the differences in the acidity of the mucins produced by the two species. It would be expected that the more acidic mucins of Leptocottus would have a larger apparent molecular weight than the neutral mucins of Icelinus (Table 7) , since the presence of ionic groups would tend to maintain a straight-chain conformation on the acidic mucins through ionic repulsion while the neutral mucins would tend to collapse and precipitate. 90 DISCUSSION The present study, concerning the skin and epithelial mucus secre-tion i n the family Cottidae, demonstrates that while the production of the mucus coat i s under pituitary endocrine control, the mucus coat i s incapable, under normal circumstances, of affecting the ionic permeabili-ty of the integument. The structural and chemical features of the mucus-secreting c e l l s , upon which the rest of the study is based, are quite com-plex and hence these characteristics are considered f i r s t . The endocrine control of three different populations of mucus cells follows the review of mucus c e l l structure and relates the effects of osmoregulatory hor-mones and environmental salinity to mucus secretion. Discussion of the functional significance of the mucus coat includes the possible involve-ment of mucus in teleost osmoregulation, followed by other functions of the mucus coat that may also explain the importance of the mucus and i t s endocrine control. Histology and Histochemistry of Cottid Mucus Cells Three fundamental types of mucus-secreting cells have been recog-nized i n the skin and g i l l of cottids, F i r s t , the mucus-secreting goblet cells produce mucins in membrane-bound vesicles and release the mucus by holocrine secretion. The goblet cells of the skin and g i l l are similar in structure and mode of secretion, but differ in size. Second, the eosinophilic granular cells also appear to secrete as holocrine glands, but the secretion product i s not held in vesicles at the time of secretion since during the development of these cells the vesicles coalesce into 91 one large vesicle that occupies the greater part of the c e l l (Fig. 14). The third type i s the so-called 'cuticle-secreting' c e l l (Whitear, 1971). These c e l l s , like the goblet c e l l s , produce mucins in membrane-bound vesicles, but here the mode of secretion i s exocytotic rather than holocrine; the vesicles may be seen releasing their contents at the apex of the c e l l . A l l three c e l l types have been identified i n a variety of fishes (Appendix 2); however, the cottids lack club c e l l s , a common type of epithelial c e l l in some teleosts. Goblet c e l l s . The goblet cells may be further divided into two types according to the chemical properties of the secretory product. Euryhaline and freshwater-resident cottids produce only acidic mucins. This mucus may be a sialoglycoprotein, as in the skin goblet cells of Leptocottus, a sulphated mucin, as i n the skin goblet cells of the freshwater-resident Cottus, or an apparent mixture of these two acidic mucins—for example, in the goblet cells of Leptocottus g i l l s . The presence of s i a l i c acid-containing glycoproteins and sulphated mucopolysaccharides in the goblet cells and secreted mucus of fishes has been widely reported (Appendix 2). The second type of mucin produced in the goblet c e l l s , found only in the seawater cottids of this study, is a neutral mucopolysaccharide. This type of mucin has been reported in only four other species of fish , Zeus  faber, Cottus scorpio, Tetraodon laevis, and Platichthys flesus (Bremer, 1972; Fletcher et^ al_., 1976). A l l of these species are normal inhabitants of marine environments, though at least two (Platichthys flesus and Oligocottus maculosus) show euryhaline a b i l i t y (Potts and Eddy, 1973; 92 Morris, 1960). The possible importance of the neutral mucins is dis-cussed in a later section. Eosinophilic Granular Cells The eosinophilic granular cells are present in many teleosts and in one species of agnathans (Appendix 2). These cells appear to secrete a protein (Mittal and Munshi, 1969 and 1971), and the present study has characterized the secretion product further, showing that the granular cells contain a tryptophan-rich basic protein. The secretion product of these cells lacks carbohydrate, so i t may be distinguished from the glyco-proteins of the mucus ce l l s . Other proteinaceous secretions of fish skin include the club cel l s , that are associated with the production of and alarm substance in cyprinids (Pfeiffer, 1963; Carmignani and Zaccone, 1974a), and the thread cells of the hagfish (Ferry,' 1941; Wessler and Werner, 1957; Blackstad, 1963). Scanning Appendix 2, i t i s interesting to note that those species which have eosinophilic granular cells lack club c e l l s , and vice versa. This may suggest that the two c e l l types are merely different modifications of a single ancestral protein-secreting c e l l . Though the alarm substance of fis h has attracted much interest, the 'predator-deterring' substance of the granular cells (Pfeiffer and Pletcher, 1964) has not been effectively studied. If an appropriate beha-vi o r a l bioassay for the substance were developed, then the nature and im-portance of the granular cells could be determined. It would be interest-ing to compare the nature and actions of the alarm and 'predator-deterring' substances in studying the role of the fish skin in the production of pheromones. 93 Cuticle-Secreting Cells The cuticle-secreting cells of the skin may represent an ancestral form of mucus secretion, since similar cells are found i n the sim-ple integument of the cephalochordate Amphioxus lanceolatus (Olsson, 1961). Though these cells have not been widely reported i n the skins of f i s h , they have been found i n a l l of the species that have been examined by electron microscopy (Appendix 2). Since the bulk of literature on f i s h skin involves only histology and histochemistry and these small cells often do not stain with Alcian Blue, i t i s l i k e l y that this c e l l type i s more common than the literature indicates. The cuticle-secreting cells of Leptocottus appear i n three forms which reflect the state of activity of the c e l l s . The f i r s t and least •.active state i s represented by the condition of these c e l l s i n hypophysec-tomized and seawater-adapted f i s h , where the cells appear flattened, have small pycnotic nuclei and do not stain with Alcian Blue (Fig. 30). Mo-derately active cuticle-secreting cells typically appear in intact sea-water-adapted Leptocottus and are roughly cuboidal i n shape, with many membrane-bound vesicles which stain strongly with Alcian Blue. Cells i n this state have few mitochondria and a nucleus which appears pycnotic (Fig. 11, 31). The third and apparently most active state appears i n i n -tact Leptocottus that have been acclimated to 5% sea water (Fig. 10, 32). Here the cells are cuboidal, have rounded nuclei, many mitochondria, Golgi -complex, and small numbers of membrane-bound vesicles containing the electron-dense secretion product. These ce l l s are either"1 unstained or only weakly stained with Alcian Blue. 94 Histochemical examination of the cuticle-secreting cells has shown that the secretion product i s a sulphated mucopolysaccharide, in contrast to the sialoglycoprotein secreted by the goblet cells of Leptocottus skin. To my knowledge, this is the f i r s t attempt to charac-terize the contents of cuticle-secreting c e l l s . Because the cuticle-secreting cells differ from goblet cells in structure, mode of secretion, and type of mucin secreted, i t is possible that these two major c e l l types may impart distinct functional characteristics to the mucus coat. Cuticle-secreting cells have not been reported previously in the family Cottidae and to date no studies have attempted to monitor the ac-t i v i t y of these cells in response to environmental conditions, hypophy-sectomy or hormone treatment. Knowledge of the endocrine control of the cuticle-secreting cells may provide information regarding the function of these c e l l s . Endocrine control of Mucus-secreting Cells It i s convenient to partition the discussion of the endocrine con-t r o l of mucus cells into three categories according to the populations of skin goblet cel l s , g i l l goblet c e l l s , and the cuticle-secreting cells of the skin. Each c e l l type apparently reacted independently of the others, and each type represents a different cytological and/or cytochemical spe-cialization of the basic phenomenon of mucus secretion. The endocrine control of the eosinophilic granular cells was not r i -gorously examined and thus these cells are not discussed in detail. How-ever, cursory observation indicates that these cells are present in large numbers in fresh water- and seawater-acclimated Leptocottus skin and are 95 not apparently affected by hypophysectomy. The major criterion of these experiments was the change in number of mucus-secreting cells or, i n the case of the cuticle-secreting c e l l s , the number of fish that showed Alcian Blue staining of these c e l l s . Counts of the mucus cells merely reflect the standing crop of c e l l s , and the number of cells may be determined only once for a single animal. This severely limits the type of conclusions that may be validly drawn from the data. Changes in the turnover rate of the cells are not determined, except for gross alterations that would result in the disappearance of a c e l l type from the epithelium. The s i a l i c acid assay for epithelial mucus using whole skin (Lemoine and Olivereau, 1971) unfortunately suffers the same drawbacks. However, a possible method for estimating the actual 3 rate of mucus secretion might involve the incorporation of H-fucose into the mucus (Bennett, et a l . , 1974; Dermer and Sherwin, 1975) followed by autoradiography. Unfortunately, both developing and mature cells show 3 a f f i n i t y for H-fucose (Bennett et^ a l . , 1974) and i t is not possible to 3 apply a 'pulse' of H-fucose in vivo to trace the succession of one gen-eration of mucus ce l l s . Thus determination of the turnover rate of the cells would probably require in vitro organ culture where pulses of ra-dioactive metabolites may be traced through the development of the c e l l s . Finally, turnover rates determined in this manner may not reflect the in vivo condition. The only attempt to measure directly the rate of mucus secretion in this study was the determination of the 'mucus shedding rate' of fish 96 that were held in small chambers for one day. The shed mucus In the chamber water was then assayed by the phenol-sulphuric acid method. This procedure offers a direct measurement of mucus secretion and also allows repeated estimates from a single animal. Results using this method supported the assumption that a decrease in the number of mucus-secreting cells reflects a decline In the amount of mucus secreted (see page 58 ). In retrospect, this investigation would have been consider-ably strengthened had this method been used routinely. Endocrine Control of Skin Goblet Cells The number of goblet cells i n the skin of Leptocottus i s normally 20-30 cells per mm of 7 ym histological section, or roughly 3000-4000 cells per square millimeter of skin. Hypophysectomy of 5% seawater-acclimated or 100% seawater-adapted Leptocottus did not significantly af feet the number of these cells (Tables 10-12, 17); thus, the pituitary i apparently not required to maintain the goblet cells of the skin. Fur-ther, injection of prolactin or C o r t i s o l had no effect on the number of these cells (Tables 12, 15, 17). Finally, there was no effect of a c c l i -mation to 5% or 100% sea water (Table 13). In contrast, Cottus asper that were adapted to full-strength sea water had significantly fewer skin goblet c e l l s , compared to fresh water controls (Table 14). Cottus asper does not normally encounter sea water and long-term exposure to this salinity ultimately caused hypertrophy of the g i l l epithelium and death. It i s possible that the fresh water-resi dent Cottus asper relies more heavily on pituitary control of skin mucus secretion than the estuarine species, Leptocottus armatus. This possi-97 b i l i t y i s examined in the lig h t of previous studies on the topic. Hypophysectomy reduces the number of goblet cells i n the skin of the following fresh water-resident teleosts; Carassius auratus (Ogawa and Johanssen, 1967; Ogawa, 1970), Betta splendens (Schreibman and Kallman, 1965), and Umbra limi (Stanley and O'Connell, 1970). However, hypophysectomy was ineffective in Tilapia mossambica (now: Sarotherodon  mossambicus) (Bowman, 1966, cited by Bern, 1967) and Poecilia latipinna (Ball, 1969). Prolactin inject ion or pituitary autotransplantation re-versed the effect of hypophysectomy in Carassius and Umbra (Ogawa, 1970; Stanley and O'Connell, 1970). Prolactin injections of intact c i c h l i d (Symphysodon discus, Symphysodon aequifasciata axelrodi, and Aequidens  latifrons) induced increases i n the number of skin goblet cells (Egami and I s h i i , 1962; Blum and Fiedler, 1965), but similar treatment of Tilapia, Poecilia, or the c i c h l i d Pterophyllum was ineffective (Bern, 1967; B a l l , 1969; Blum and Fiedler, 1965). Adaptation to 25% sea water caused a decrease in the number of skin goblet cells i n Cichlasoma biocellatum (stenohaline, fresh water- resident) and this effect could be reversed by injections of ovine prolactin (Mattheij and Stroband, 1971; Mattheij et a l . , 1972); however, hyperosmotic salin i t y did not affect the goblet ce l l s in the skin of Anoptichthys jordani (Mattheij and Sprangers, 1969). There appears to be no general trend indicating pituitary control of the skin goblet cells of fresh water teleosts but in certain species, no-tably some of the cichlids, prolactin has been repeatedly shown to stimu-late skin mucus secretion. 9 8 The endocrine control of mucus secretion i n the skin of euryhaline or anadromous teleosts has been studied i n only two species, Anguilla  anguilla and Leptocottus armatus. No studies have been attempted on stenohaline seawater fish. The latter point i s interesting because i f the pituitary does control mucus secretion in seawater f i s h , i t would probably not involve prolactin, since this hormone i s generally associated with functions in the fresh water environment. N-acetyl neuraminic acid (NANA, a s i a l i c acid) concentration i n the skin (an estimate of mucus as assayed by the thiobarbituric acid assay, Lemoine and Olivereau, 1971; Pickering, 1974) of Anguilla anguilla is i n -creased by short-term adaptation to sea water (Olivereau and Lemoine, 1972) but decreased by long-term acclimation to sea water (Lemoine and Olivereau, 1973a). Prolactin injection of intact eels i s ineffective in fresh water-acclimated fish or after transfer to sea water (Olivereau and Lemoine, 1971a; Lemoine and Olivereau, 1973a), and C o r t i s o l injection of intact eels i s also ineffective (Lemo ine and Olivereau, 1974). Hypophy-sectomy reduces the NANA in fresh water-adapted eels (Olivex-eau and Lemoine, 1971a,b) or after transfer to sea water (Lemoine and Olivereau, 1973a), but not in seawater-acclimated eels (Lemoine and Olivereau, 1974). Pituitary autotransplants or prolactin injection of hypophysectomized eels maintains the NANA at levels near those of intact controls (Olivereau and Lemoine, 1971a,b; Lemoine and Olivereau, 1973a). Finally, C o r t i s o l i n -jections into hypophysectomized fresh water or sea water eels reduces the skin NANA, compared to uninjected hypophysectomized controls (Lemoine and Olivereau, 1974). 99 While prolactin stimulates and C o r t i s o l inhibits skin mucus secre-tion i n the eel, the pituitary does not seem to be important in control-ling the skin goblet cells of Leptocottus. It i s impossible at this time to generalize these results to encompass a l l euryhaline teleosts. Fur-ther, since no studies have been done on stenohaline sea water teleosts, correlations of the endocrine control of skin mucus secretion with habi-tat sa l i n i t y are incomplete. Endocrine Control of G i l l Goblet Cells The goblet cells in the g i l l epithelium of Leptocottus appear to be under pituitary endocrine control by prolactin and ACTH-cortisol. Hypo-physectomy of seawater- or 5% seawater-acclimated fish reduced the number of goblet cells on the efferent (leading) side of the g i l l (Tables 10, 11, 12, 17) and ovine prolactin injections, starting a few days after hypophy-sectomy, prevented this decrease (Tables 12, 17). However, the afferent (trailing) side of the g i l l did not consistently respond to hypophysectomy or prolactin replacement (Tables 10, 11, 12, 17). Sea water adaptation caused a reduction i n the number of mucus cells i n the efferent side of the g i l l , compared to controls i n 5% sea water, but again the cells of the afferent side of the g i l l failed to respond (Table 13). This decrease was accompanied by a decrease in the apparent activity of the prolactin cells of the rostral pars d i s t a l i s (Table 13; Figures 24, 25). Prolac-tin injection of intact seawater-adapted fish increased the number of mucus cells on the efferent side of the g i l l , and C o r t i s o l injection of 5% seawater-adapted fi s h caused a reduction in the numbers of these cells (Tables 15, 16). Thus prolactin appears to stimulate the mucus cells in 100 specific areas of the g i l l , associated with adaptation to a hyposmotic environment, and C o r t i s o l inhibits these c e l l s , associated with acclima-t i o n to sea water. While the mucus cells of the efferent side of the g i l l filaments appear to be controlled by the pituitary, those of the afferent side of the g i l l are apparently not under pituitary endocrine control. The involvement of pituitary hormones in the control of g i l l mucus was f i r s t observed by Burden (1956) who found that hypophysectomized and seawater-adapted Fundulus heteroclitus had fewer g i l l mucus cells than did intact and fresh water-adapted controls. He also found that injec-tions of pituitary brei could reverse the effect of hypophysectomy. Jozuka (1966) found a similar decline in the number of g i l l mucus cells 30 days after transfer of fresh water eels to sea water, especially on the afferent side of the g i l l where chloride cells appeared to displace the mucus ce l l s . Lemoine and Olivereau (1973b) found that transfer of sil v e r eels to sea water resulted in a transient increase in g i l l mucus cells followed by a gradual decline which was significant 160 days after transfer. Similar results were found in Anoptichthys jordani, Mugil cepha- lus, and M. capito; a l l three species had fewer g i l l mucus cells in hy-perosmotic media than i n fresh water, paralleled by decreased prolactin c e l l activity (Mattheij and Sprangers, 1969; Blanc-Livni and Abraham, 1970). Hypophysectomy of fresh water eels caused an increase in the number of g i l l mucus c e l l s , compared to intact controls and prolactin injection caused a further rise (Olivereau and Olivereau, 1971). Lemoine (1974a,b) 101 showed that both intact and hypophysectomized silver eels have higher g i l l NANA after transfer to sea water than intact controls in fresh water. While prolactin treatment of the hypophysectomized group had no effect, C o r t i s o l injections produced a drop in g i l l NANA, compared to uninjected hypophysectomized controls. Leatherland and Lam (1969) also found no effect of prolactin injections on the g i l l mucus cells of seawater-accli-mated Gasterosteus aculeatus, but i n fresh water, prolactin treatment i n -creased the number of g i l l mucus ce l l s . The g i l l mucus cells of Cichlasoma biocellatum are not affected by acclimation to fresh water or 25% sea water, and prolactin injections also had no effect (Matthij and Stroband, 1971). Though there i s at least preliminary evidence i n favour of endocrine control of g i l l mucus cells among the teleosts, there are several studies that directly contradict the hypothesis. As in the case of the skin gob-let c e l l s , no studies have included experiments on stenohaline seawater fi s h . If mucus does have a function in teleost osmoregulation, i t would be expected to act at the g i l l , where the majority of ion transfer takes place. If this hypothesis were correct, one might expect that the goblet cells of the g i l l may be controlled by osmoregulatory hormones, while the goblet cells of the skin may not. Several investigations have shown d i f -ferential endocrine control of the goblet c e l l s in the skin and the g i l l . This situation i s apparently true for Anoptichthys jordani (Mattheij and Sprangers, 1969) and Leptocottus armatus (present study) but Cichlasoma  biocellatum and Anguilla anguilla have similar endocrine control for both 102 skin and g i l l goblet cells (Mattheij and Stroband, 1971; Lemoine and Olivereau, 1973a; Lemoine, 1974a). Again, i f mucus were important in osmoregulation, one might expect that.euryhaline fishes that li v e in estuaries and face regular changes i n environmental salinity might have more mucus cells in the g i l l region than fishes (euryhaline or not) that li v e in stable environments. This appears to be generally true for the ten species of cottids examined i n this study (Table 9). Only two marine species have numbers of mucus cells i n the g i l l comparable to the estuarine Leptocottus armatus, and a l l of the fresh water-resident cottids studied had very few g i l l mucus c e l l s . Endocrine Control of Cuticle-secreting Cells The cuticle-secreting cells are highly active in 5% seawater-accli-mated Leptocottus but only moderately active in seawater-adapted fi s h . This effect i s correlated with a significantly higher activity of the pro-lactin cells i n the 5% seawater-adapted f i s h ; thus i t appears that pro-lact i n maintains these cells in a highly active state. Further evidence for the conclusion may be seen in the effect of ovine prolactin injections on seawater-adapted fi s h . Here prolactin abolishes the normal Alcian Blue staining of the c e l l s ; the effect is interpreted as an increase in activity to the highly active state, according to the c r i t e r i a set out previously (Page 75). On the other hand, C o r t i s o l injections of 5% sea-water-acclimated Leptocottus increases the Alcian Blue staining of the ce l l s , implying that this hormone reduces the cuticle-secreting cells to the 'moderately active' state, similar to the. normal sea water condition. Thus prolactin appears to increase the activity of the cuticle-secreting 103 c e l l s , while Cortisol appears to decrease the activity of the c e l l s . Hypophysectomy of seawater-adapted f i s h resulted i n a reduction of the staining of the cuticle-secreting cells with Alcian Blue. Rather than a stimulation of secretion, this result i s interpreted as a decrease i n the activity of these c e l l s , because the cuticle-secreting cells of the hypophysectomized group had very small nuclei and the cells were small and flattened (Fig. 30). In contrast, the cells of the prolactin-injected hypophysectomized f i s h , while not staining with Alcian Blue, were cuboidal and had relatively larger nuclei than the hypophysectomized vehicle-injected group (Fig. 32). Thus prolactin again appears to i n -duce a high activity of the cuticle-secreting c e l l s similar to the 5% seawater condition. Cortisol and growth hormone, on the other hand, re-store the activity of the cells to the normal seawater (moderately active) condition (Fig. 31). Hypophysectomy of 5% seawater-acclimated fish seemed to produce i n -consistent effects on the cuticle—secreting c e l l . However, in the three experiments reported here, three different schedules were used to decrease the s a l i n i t y from the i n i t i a l (isosmotic) 25% sea water (see Table 1). In the f i r s t experiment the f i s h were held for 22 days post-operatively in 25% sea water and then transferred directly to fresh water for one day. This treatment i s rather harsh since even intact Leptocottus only survive for a few days i n fresh water. Both the intact and hypophysectomized f i s h showed no staining of the cuticle-secreting cells with Alcian Blue, and i n both groups the upper layers of the epithelium appeared to be loosened from the underlying epidermal c e l l s . It i s possible that the 104 rapid transfer to fresh water caused a release of mucus from the cuticle-secreting cells or a shedding of the upper layer of the epidermis. In either case, hypophysectomized fish appeared to react similarly to the sham-operated controls. In the second hypophysectomy experiment, significantly fewer of the hypophysectomized fi s h showed Alcian Blue staining of the cuticle-secret-ing c e l l s , compared to the pooled intact and sham-operated control f i s h . The hypophysectomized group had more flattened cuticle-secreting cells that often appeared to be loosened from the underlying layers of the epi-dermis. Thus the reduction in Alcian Blue staining is interpreted as a decrease in the activity of these c e l l s . The third hypophysectomy experiment differed from the f i r s t in that the salinity was decreased from 25% sea water to 5% sea water in two stages, and that these fish were not exposed to fresh water. Alcian Blue staining of the cuticle-secreting cells was weak in the sham-operated con-trols, strong in the hypophysectomized control fish, but absent from the hypophysectomized fish that received injections of ovine prolactin. Though the sham-operated and hypophysectomized control groups had similar numbers of fish with Alcian Blue-positive c e l l s , the cells of the hypophy-sectomized group were more darkly stained and were often loosened from the underlying layers of the epithelium. Thus the cuticle-secreting cells of the hypophysectomized fish were in the 'moderately active' state and were considered to be less active than the cells of the sham-operated fi s h . Significantly fewer of the prolactin-treated group showed Alcian Blue staining of the cuticle-secreting c e l l s , compared to the control groups. This result i s interpreted as an increase in the activity of the 105 cuticle-secreting cells to the 'highly active' state, since the stain-ing and histological appearance of these cells was similar to those of 5% seawater-acclimated f i s h . Maintenance of the cuticle-secreting cells appears to require pro-la c t i n in 5% seawater-acclimated f i s h and C o r t i s o l i n seawater-adapted f i s h . Prolactin appears to induce a highly active state in these c e l l s , while Cortisol appears to maintain the cells i n the moderately active state. The lowest or inactive state was seen only i n hypophysectomized f i s h . Bern (1975) maintains that prolactin's stimulation of the mammary gland alveolar cells may be analogous to prolactins actions on teleost g i l l epithelium. He suggests that "...there may yet emerge a commonality of action on secretory organs involved i n synthesis and release of a variety of secretory products." I consider the similarities between the cuticle-secreting cells of f i s h and the alveolar cells of the mammary gland to be significant i n examining possible homologous actions of pro-l a c t i n . Like the cuticle-secreting c e l l s , the alveolar cells of the mammary gland secrete a proteinaceous material by exocytosis (the l i p i d portion of milk on the other hand i s an apocrine secretion, Bargmann and Welsch, 1969). The secretion products of the two epithelia are similar in several ways. Both milk and mucus contain immunoglobulins, s i a l i c acid-containing glycoproteins^ saturated fatty acids, phospholipids, and cholesterol (milk: Patton and Hood, 1969; White et a l . , 1973; mucus: Lewis, 1970; Bradshaw et a l . , 1971; Harris and Hunt, 1973; Wold and Selset, 1977). However, fi s h mucus does lack triglycerides, lipoproteins, and lactose a l l of which 106 are present in milk. It appears that prolactin has quite similar ef-fects on the mammary gland by stimulating lactogenesis, and on the cu-ticle-secreting cells of cottid skin by stimulating mucogenesis. In addition to the apparent homology of prolactin action on epithelial se-cretions (above), at least one group of fishes u t i l i z e s epithelial mucus in a manner analogous to milk. Discus f i s h secrete mucus under the i n -fluence of prolactin and use the secretion to feed newly-hatched larvae (Egami and I s h i i , 1962; Blum and Fiedler, 1965; Blum, 1973); unfortu-nately i t is not possible to compare the chemical properties of this mucus with that of normal mucus, since no work has been done on the his-tochemistry or chemistry of the 'discus milk'. I believe that not only is there a 'commonality of action on secre-tory organs' but also the secretion products may, in some cases, be similar. Thus the apparent variety of prolactin's actions may be in part due to a diversification of the functions of similar secretions accord-ing to the particular 'needs' of the animal in adapting to i t s environ-ment . Functional Significance of the Mucus Coat: Osmoregulation We have seen that mucus secretion i n some teleosts is stimulated by prolactin and inhibited by C o r t i s o l . Though considerable contradictory evidence is recognized for other species, the estuarine cottid Leptocottus  armatus does increase the number or activity of the g i l l mucus cells and the skin cuticle-secreting cells in response to prolactin. Among other functions of this hormone, prolactin i s known to decrease the osmotic 107 and Ionic permeability of the g i l l (reviews: Bern and N i c o l l , 1968; Lam, 1972; Johnson, 1973; Maetz, 1974; Bern, 1975), an action which favours acclimation to hyposmotic environments. Pickford and asso-ciates (1966) and Potts and Evans (1966) have postulated that prolactin' action in reducing permeability of the g i l l and/or skin may be due to prolactin's stimulation of mucus secretion and thickening of the mucus coat. To date this hypothesis has not been rigorously tested. Wittouck (1975) and Marshall (1976) have further suggested that the mucus coat may act to decrease the permeability of the underlying epi-thelium by ionic interactions (assuming that the mucus coat i s a bed of negative charges which would tend to repel negatively charged ions and accumulate positively-charged ions) or by increasing 'unstirred' layers adjacent to the membrane. Attempting to test these possible mechanisms I have found that mucus scraped from the skin of Leptocottus does not significantly alter the diffusion coefficient of sodium or chloride ions passing through the mucus (Table 18), and though the mucus does have some ion exchange capability (Fig. 33), the concentration of ion-exchang able groups i s quite low. The involvement of the mucus coat in osmore-gulation thus may be limited to the enhancement of unstirred layers. Unstirred layers are areas of slow (laminar) flow adjacent to a solid surface, where particles approaching or leaving the surface must pass through the layers by diffusion alone. In steady-state systems concentration gradients are established i n the inner and outer unstirred layers, and this causes the concentration gradient across the membrane to be smaller than the concentration difference of the bulk solutions (Fig. 38 ). Where solutes pass rapidly across the membrane, unstirred 108 layers may become rate-limiting (Dainty and House, 1966). The skin of fis h i s generally considered to be practically impermeable to passive movement of water and ions, compared to the much thinner and well-vascu-larized g i l l epithelium (Motais et a l . , 1969; Kirsch, 1972). Thus across the skin solutes do not 'pass rapidly across the membrane' and hence unstirred layers are probably not rate-limiting. As such we are principally concerned with unstirred layers in the region of the (more permeable) g i l l epithelium. Mucus would tend to enhance unstirred layers because of i t s high viscosity, relative^to water, and i t s tendency to promote laminar flow— hence i t s lubricating properties (Figures 3 4 , 3 5 ) . To assess the contribution of the mucus coat i n reducing the permea-b i l i t y of the integument, a convenient measure is 1/P, where P is the membrane permeability and l/P i s a measure of the'impermeability' of the 2 -5 system. The apparent permeability of the system ( g i l l ) i s 10 to 10 ^ cm/sec. (Isaia, 1972; Motais and Isaia, 1972; Motais e_t al. , 1969), including the true permeability of the membrane plus the permeability of the unstirred layers on both sides of the membrane, according to equa-tion (4) taken from Dainty and House (1966). 1 1 + 1 (4) P P t P 6 a 2 1/P i s convenient since i t has been shown that these 'impermeabili-ties' are additive (Dainty and House, 1966). 109 Here the total apparent 'impermeability' of the system, —^— , i s ex-a 1 pressed as the sum of the true 'impermeability' of the membrane, , t and the 'impermeabilities' of the unstirred layers on the two sides of the membrane, ~ — and —jjj— . The'impermeability' of an unstirred layer 51 a2 varies directly with the thickness of the layer (6, in cm) and inversely 2 with the diffusion coefficient of the solute (D, in cm /s), according to (5) D (5) 1 have shown that the diffusion coefficient of Na or Cl through mucus is -5 2 on the order of 10 cm /s (Table 18). The maximum thickness of an un-stirred layer i n the g i l l is approximately 50 ym, half the distance be-tween two adjacent g i l l lamellae. Thus contribution of such an unstirred layer to the 'impermeability' of the system i s 0.5-5% of the total. Hence the contribution of unstirred layers due to the mucus coat i s small, rela-tive to the overall 'impermeability' of the g i l l . On the other hand, i f the fish remains s t i l l and reduces i t s i r r i g a -tion of the g i l l , as i s apparent i n osmotically stressed Leptocottus, the unstirred layers may become much thicker than the above estimates. Under these conditions the unstirred layers may contribute to the 'imper-meabilit} 7' of the g i l l membranes, and prolong survival under the adverse conditions of osmotic stress. Though mucus does not appear to impede the passage of small molecules (such as ions), this coating may prove to be a considerable barrier to larger molecules or small particles such as bacteria and viruses. Kent (1967) also holds this view regarding gastric mucus secretions and states 110 Figure 38. Profile of unstirred layers adjacent to a membrane. This diagram represents the effect of unstirred layers on the concentration gradient across a membrane of thickness Ax. Due to the slow (diffusional) movement of solute through the unstirred layers (thickness 6^ ) concentration gradients are established i n these layers such that the membrane gradient (AC ) becomes much smaller than the concentration d i f f e r -mem ence of the bulk solutions (AC, •,,)• Thus the gradient DVlXiC 'recognized' by the membrane and the movement of solutes are reduced. With increased s i t r r i n g of the bulk solutions, the unstirred layers are decreased (6^  goes to 6^) a n ^ the membrane gradient increases (AC' goes to AC" ). With mem mem ideal mixing of the bulk solutions, AC tends to AC, .,, . mem bulk 110a D I S T A N C E (x) I l l "Possibly, mucin barriers function as g e l - f i l t r a t i o n devices, r e s t r i c t -ing the approach of external proteins to the epithelial surface." While the concentration of ionic groups i n mucus i s apparently too low to expect that the mucus coat is selectively permeable to ions, Komnick and Bierther (1969) have demonstrated that the mucosubstances covering the chloride cells of fresh water-adapted Gasterosteus aculeatus concentrate both sodium and chloride ions. They postulate that such 'ion capture' may serve to increase the avail a b i l i t y of substrate for the active uptake of ions. This may be particularly true for the calcium ion, since fish mucus has a higher a f f i n i t y for calcium than for sodium or potassium (Maetz, personal communication) and because calcium i s re-versibly bound to 'weak' sites of mucins (Forstner, 1973). Komnick and Bierther (1969) found that sodium and chloride are pre-sent i n the. apical crypts of chloride cells at concentrations of 1.1 and 0.9 mg %, respectively. The figures compare favourably with my estima-tion of fixed charge density i n Leptocottus mucus (1.1 mg % corresponds to mucus solution of 0.8 mg/ml). For mucus to be effective in concentrat-ing ions, the environmental salin i t y would have to be less than 0.011 o/oo = 1.1 mg %. Because Leptocottus cannot survive in very low sa l i n i t i e s (less than 1 o/oo) the mucus i s probably not important i n concentrating ions for this species. It remains to be demonstrated whether the ion-concentrating property of mucus is important in fresh water f i s h . In conclusion, mucus probably does not function i n osmoregulation by decreasing the permeability of the general body surface, and i s pro-bably not important in connection with the transport of ions across 112 chloride cells in Leptocottus. Reductions in permeability due to prolactin thus do not appear to involve the mucus coat, but i t may be remembered that prolactin maintains the cuticle-secreting cells of the skin in a 'highly active' state. While such maintenance of these cells probably increases the amount of mucus produced by the c e l l s , i t may also promote intercellu-lar connections, particularly 'tight junctions' which do not allow pas-sage of low-molecular-weight solutes through the intercellular space (Stahelin, 1974). This may in part be the mechanism by which prolactin decreases the permeability of the integument. This conclusion awaits supportive evidence from quantitative examination of skin and g i l l u l -trastructure . Though there is considerable doubt that the mucus coat of fishes i s directly involved in osmoregulation, as discussed previously, the skin may participate actively in salt balance as a source of extra-renal ion in fish. The skin of Gillichthys mirabilis (lower jaw), Fundulus  heteroclitus (opercular epithelium), Carassius, Gasterosteus aculeatus , Pomatoschistus, and Blennius pholis contain typical mitochondria-rich chloride cells (Nishioka, personal communication; Karnaky and Kinter, 1977; Hendrikson and Matoltsy, 1968; Whitear, 1971). Chloride cells have also been found in the skin of eel and sardine larvae, but the skin of the adult eel lacks these cells (Lasker and Threadgold, 1968; Leonard and Summers, 1976). Thus the cytological basis for extra-renal salt se-cretion has been demonstrated in several species, but i t appears that the skin may be of greatest importance to the smaller larval fish. Here the high surface area-to-volume ratio may 'require' more extensive u t i l i z a t i o n 113 of the general body surface for osmoregulation. Indeed, newly hatched larvae may depend entirely on skin ion transport, since during these early stages the larvae lack g i l l filaments, the kidney is present only as a pronephros, and the gut may not be open (Holliday, 1969). Early workers in electrophysiology reported a 'current of rest' across fish skin (Hermann, 1882; Reid, 1893). This current was similar to but smaller than that reported across frog skin. Recent experiments on the opercular epithelium of Fundulus heteroclitus (Karnaky e_t a l . , 1977) and the skin of the lower jaw of Gillichthys mirabilis (Marshall, 1977, see Appendix 3) show that chloride i s actively transported from serosa to mucosa i n seawater-acclimated fi s h . The active transport i s electrogenic, and thus exhibits a spontaneous transepithelial potential (serosa-positive) and a short-circuit current. The results imply that the skin contributes to extra-renal salt extrusion, but the contribution in adults may be small, relative to fluxes across the much larger surface area of the g i l l epithelium. Other Functions of the Mucus Coat The mucus coat of fishes serves a variety of functions including the prevention of disease and lubrication of the body surface. These func-tions are br i e f l y reviewed i n the introduction. The proper functioning of the mucus coat relies on the property of the mucus to adhere to the body so that a protective coating may be established, but also the mucins must be soluble enough so that the mucus does not accumulate excessively and obstruct water flow over the g i l l . This semi-solubility of the mucus 114 i s reflected in i t s high viscosity relative to water (Fig. 34), and i t s lubricating properties (Fig. 35, 36). An increase in the ionic strength of the solvent results i n a loss of the lubricating property of the mucus (Table 19) and precipitation of the mucus. Alkaline water apparently also affects the mucus coat and renders fish more susceptible to disease (van Diujn, 1973, p. 83-84). Thus radical changes in the salini t y or pH of the environment may disrupt the semi-soluble state of the mucus and result in the loss of the lubricating and disease-preventing functions of the mucus coat. Seawater-acclimated Leptocottus produce mucins that show significant-ly better lubrication than mucus taken from 5% seawater-acclimated fish (Fig. 37). Thus these fish appear to adjust the composition of the mucus coat in response to salinity change. However, mucus samples from animals in either salinity have approximately equal lubricating properties when the mucus is diluted with seawater or d i s t i l l e d water (Fig. 37). The difference i n the lubricating property of mucus from the hyper- and hypo-osmotic media seems to be related to the activity of the cuticle-secret-ing c e l l s , since these cells are the only mucus-secreting cells that ap-pear to respond to salinity change, Rosen and Cornford (1971) found that fast-swimming predator species of fish (fresh water or sea water) tend to secrete mucus that is highly efficient in reducing f r i c t i o n , while seden-tary species, for example the hagfish, secrete mucins that have poor l u -bricating properties. A notable exception to the rule i s the mollusc Cypraea (Zonaria) spadicea which secretes an efficient lubricating mucus, apparently not for speed, but for lubricating i t s tissues inside i t s shell. 115 The secretion of a semi-soluble mucus with associated lubricating properties seems fundamental for efficient maintenance of a mucus coat. As such, the efficiency of those functions of the mucus which depend on a stable layer of semi-soluble material may be reflected i n the lubricat-ing properties of the mucus. We shall now examine a few cases where the secreted mucus is either lacking or radically altered from the semi-solu-ble 'norm'. These variations appear to reflect highly specialized func-tions for mucus secretion. In semi-aquatic habitats several species of fish show structural specialization of the skin associated with the problem of desiccation. Bagarius bagarius has a keratinized epithelium which apparently serves to reduce water loss through drying (Mittal and Munshi, 1970; Appendix 2). The lungfish, Protopterus annectens,avoids desiccation by aestivating in burrows and encasing i t s body i n a thick coating of dried mucus (Smith and Coates, 1936; Kitzan and Sweeny, 1968; Appendix 2). Secretion of coagulated mucus i s a specialization present in some parrotfishes that nightly secrete a large envelope of insoluble mucus, apparently as a protection from their abrasive coral reef habitats (Winn, 1955; Byrne, 1970; Appendix 2). The gelatinous mucus secreted by the hagfish i s likewise insoluble (Rosen and Cornford, 1970) but here this type of mucus appears to be secreted as a protection from predators (Blackstad, 1963). The neutral mucins produced by the sea water-resident cottids of this study are slightly less viscous than the acidic mucins of Leptocottus and the former are distinctly less slimy to the touch. For these reasons, the neutral mucins are probably not efficient l u b r i -cants, but alternative explanations regarding the functions of this type 116 of mucin are not available. Among the structural specializations of fish skin, one group stands out as an example of secondary loss of structure, or "reversion to the archaetype". These fish lack mucus-secreting goblet cells and the pro-tein-secreting club and granular cells but retain only the cuticle-secret-ing c e l l s . Included i n this diverse group are one agnathan, one chondrich-thian, one holocephalan, one chondrostean, and among the teleosts, two percifonnes and four gasterosteiformes (Appendix 2). This specialization does not appear to involve osmoregulation since both sea water and fresh water fishes are present in the group, however, a l l the members of the group are slow swimmers and thus may not 'require' large amounts of a l u -bricating mucus, Among the great diversity of fishes the skin also shows a diversity of form and function. The various forms of the skin are l i k e l y associated with one or more factors of the complex environments in which fish l i v e . Salinity i s only one of the many characteristics of an aquatic habitat and thus structural changes correlating with changes in environmental s a l i n i -ty may, in some cases, be only t r i v i a l l y associated with osmoregulatory mechanisms. Instead, salinity may only be a cue bringing about a change i n mucus secretion, while the primary function of the mucus might be rela-ted to some other parameter of the environment. More work i s certainly required before the diversity of structures may be explained in functional terms; this is even true for the most superficial organ, the skin. 117 CONCLUSIONS 1. The skin of the sculpin, Leptocottus armatus,includes three types of secretory cells; the eosinophilic granular c e l l s , goblet c e l l s , and cuticle-secreting c e l l s . The cottids lack club c e l l s . The g i l l s have but one type of mucus-secreting goblet c e l l and no cuticle-secret-ing c e l l s . 2. While sea water, estuarine and fresh water cottids produce a c i -dic mucins, only sea water sculpins have neutral mucins in the skin gob-let c e l l s . 3. The major type of mucin produced by the skin and g i l l of Leptocottus i s a sialoglycoprotein, though sulphated mucins are present in the g i l l goblet cells and the cuticle-secreting cells of the skin. The eosinophilic granular cells produce a tryptophan-rich basic protein. 4. While the skin goblet cells o f Leptocottus are not under p i t u i -tary endocrine control, the g i l l goblet cells increase in number in hypo-osmotic media and are stimulated by prolactin but inhibited by C o r t i s o l . 5. The cuticle-secreting cells of Leptocottus exist in three states of activity. The highest activity i s recognized i n 5% seawater-adapted fish and this state i s maintained by prolactin. A moderately active state i s associated with adaptation t o sea water, a n d appears to be main-tained by C o r t i s o l or growth hormone. The lowest activity appears i n hypophysectomized f i s h . 6. Mucus may be important i n increasing unstirred layers, thus re-ducing the permeability of the g i l l , but this may only occur during times of inactivity or acute osmotic stress. 118 7. Under normal circumstances the mucus coat appears incapable of reducing the permeability of the integument through steric interference of ion movement or through Donnan exclusion. 8. Leptocottus mucus is a viscous non-newtonian f l u i d in i t s con-centrated form, but when diluted i t becomes an efficient lubricant. Sea-water-acclimated Leptocottus produce a more efficient lubricating mucus than do 5% seawater-adapted f i s h , and this effect i s probably associated with the cuticle-secreting c e l l s . 9. Salinity adaptation in Leptocottus thus appears to involve alter-ations of the mucus coat under the influence of prolactin or C o r t i s o l , and these alterations may have significance i n osmoregulation or i n ad-justing the lubricating properties of the mucus. 10. In an associated study I have shown that the skin of sea water Gillichthys mirabilis actively excretes ions (most l i k e l y Cl ). Thus the skin of some species may actively contributed to hydromineral balance through extra-renal ion excretion. 119 REFERENCES CITED Aida, K., H. Takashi, 0. Yasukatsu, Y. Hashimoto, and J.E. Randall. 1973. Structure of the skin of the soapfish, Pogonoperca punctata. Bull. Jap. Soc. Sci. Fish. 39: 1351. Asakawa, M. 1970. Histochemical studies of the mucous on the epidermis of eel, Anguilla japonica. Bull. Jap. Soc. Sci. Fish. 36_: 83-87. Baker, J.R. 1956. The histochemical recognition of certain phenols, es-pecially tyrosine. Quart. J. Microsc. Sci. 9_7_: 161-164. Ba l l , J.N. 1969. Prolactin (fish prolactin or paralactiin) and growth hormone. In: Fish Physiology; W.S. Hoar and D.J. Randall, eds. Vol. 2: 207-240. Academic Press, New York. Bargmann, W., and U. Welsch. 1969. On the ultrastructure of the mammary gland. In: Lactogenesis; M. Reynolds and S.J. Folley, eds. U. Pennsylvania Press, Philadelphia, 269 pp. Bennett, G,, CP. Leblond and A. Haddad. 1974. Migration of glycopro-tein from the golgi apparatus to the surface of various c e l l types as shown by autoradiography after labelled fucose injection into rats. J. Cell Biol. 60: 258-284. Bern, H.A. 1967. Hormones and endocrine glands of fishes. Science 158: 455-462. Bern, H.A. 1975, Prolactin and osmoregulation. Amer. Zool. 15_: 937-948. Bern, H.A., and C.S. N i c o l l . 1968. The comparative endocrinology of prolactin. Rec. Prog. Horm. Res. 2k\ 681-720. Bhatti, H.K, 1938. The integument and dermal skeleton of Siluroidea. Trans. Zool. Soc. Lond. 24: 1-102. 120 Blackstad, T. 1963. The skin and i t s derivatives. In: Biology of myxine. A. Brodal and R. FHnge, eds., Universitets-forlaget, Oslo, 588 pp. Blanc-Livni, N., and M. Abraham. 1970. The influence of environmental sal i n i t y on the prolactin- and gonadotropin-secreting regions in the pituitary of Mugil (Teleostei). Gen. Comp. Endocrinol. 14: • 184-197. Blum, von V. 1972. The influence of follicle-stimulating hormone (FSH) and luteinizing hormone (LH) on the male reproductive system and the skin of the Mediterranean Bleniid fish Blennius sphinx (Valenciennes). J. Exp. Zool. 181: 203-215. Blum, von V. 1973. Experiments mit Teleosteer-Prolactin. Zool. Jb. Physiol. 77: 335-347. Blum, von V., L. Machemer, and W.D. Bartmann. 1972. The action of TSH and LtH on the behaviour and on the histology of the skin, the thyroid gland and the gonads of the Pfauenscheimfisch, Blennius pavo. (in Ger.) Zool. Jb. Physiol. 7§_: 340-355. Blum, von V., and K. Fiedler. 1965. Hormonal control of reproductive behaviour in some cichlid f i s h . Gen. Comp. Endocrinol. 5_: 186-196. Bock, F. 1930. Die Epidermis des Stichlings. Z. Zellforsch. Mikrosk. Anat, 10: 394-400. Bolognani-Fantin, A.M. and L. Bolognani. 1964. Observations on the histochemistry of the cutaneous mucus of Anguilla vulgaris. Int. Kongr. Hist. Cytochem. 1964: 183-184, 121 Bolognani, L. and A.M. Bolognani-Fantin. 1963. Prime richerche sulla composizions del muco cutaneo di anguilla. Arch. Sci. Bi o l . 47: 331-335. Bone, Q. and C.E.R. Brook. 1973. On Schedophilus medusophagus (Pisces: Stromatoidei). J. Mar. Biol. Ass. U.K. 53: 753-761. Bradshaw, CM., A.S. Richard, and M.M, Siegel. 1971. IgM antibodies in fish mucus. Proc. Soc. Exp. Biol. Med. 136_: 1122-1124. Bremer, von H. 1972. Einige Untersuchung zur Histochemie der sezernier-enden Elemente der Teleostier-Epidermis. Acta Histochem. 43_: 28-40. Brown, G.A. and S.R. Wellings. 1970. Electron microscopy of the skin of the teleost Hippoglossoides elassodon. Z. Zellforsch. 103: 149-169. Burden, C E. 1956. The failure of hypophysectomized Fundulus heterocli- tus to survive in fresh water. Biol. Bull. 110: 8-28. Byrne, J.E. 1970. Mucous envelope formation in two species of Hawaiian parrotfishes (Genus Scarus) . Pac. Sci. 24_: 490-493. Carmignani, M.P.A. and G, Zaccone. 1974a. Histochemical investigations of the epidermis of the electric fish Malopterus electricus Lace"p (Teleostei: Malopteruridae). Acta Histochem. 4_8_: 51-61. Carmignani, M.P.A. and G, Zaccone. 1974b. Histochemistry of the acid glycosaminoglycans (mucopolysaccharides) in the branchial epithe-lium of selachians. Acta Histochem, ^8: 326-335. Chayen, J., L. Bitensky and R. Butcher, 1973. Practical Histochemistry. J. Wiley & Sons, Toronto, 271 pp. Dainty, J. and C.R. House, 1966. Unstirred layers in frog skin. J. Physiol. 182: 66-78. 122 Dermer, G.B. and Sherwin, R.P. 1975. Autoradiographic localization of glycoprotein in human breast cancer cells maintained i n organ cul-3 3 ture after incubation with H-fucose or H-glucosamine. Cancer Res. 35: 63-67. Dische, Z. 1962. Methods Carbohydrate Chem. 1: 477. Diujn, van C. 1973. Diseases of fishes. 3rd edit. I l i f f e Books, Lon-. don. 372 pp. Downing, S.W. and R.R. Novales. 1971. The fine structure of lamprey epidermis I. Introduction and mucous cell s . II. Club c e l l s , III. Granular cel l s . J. Ultrastruct. Res. 35_: 282-294; 295-303; 304-313. Egami, N. and S. I s h i i , 1962. Hypophyseal control of reproductive func-tions i n teleost f i s h . Gen. Comp. Endocrinol. Suppl. 1_: 248-253. Ferry, J.D. 1941. A fibrous protein from the slime of the hagfish. J. Bio l . Chem. 138: 263-268. Finch, E.D. and A.S. Schneider. 1975. Mobility of water bound to biolo-gical membranes: A proton NMR study. Biochim Biophys. Acta 406: 146-154. Fishelson, L. 1972. Histology and ultrastructure of the skin of Lepadichthys lineatus (Gobisocidae: Teleostei). Mar. Biol. 17: 357-364. Fletcher, T.C. and P.T. Grant. 1969, Immunoglobulins in the serum and mucus of the plaice, Pleuronectes platessa. Biochem. J, 115: 65 P. Fletcher, T.C,, R. Jones, and L. Reid. 1976, Identification of glyco-proteins in the goblet cells of epidermis and g i l l of plaice (Pleuronectes platessa L.), flounder (Platichthys flesus L.) , and rainbow trout (Salmo gairdneri Richardson). Histochem. J. 8_: 597-608. 123 Foot, N.C. 1933. The Masson trichrome staining methods in routine laboratory use. Stain Technol. 8_: 101-110. Forstner, J.F., et a l . , 1973. Interaction of calcium ions with intestinal goblet c e l l mucus. Clin. Res. 21_: 1033. II Frisch, von H. 1941. Uber einen Schreckstoff der Fischhaut und seine biologische Bedeutung. Z. Vergl. Physiol. 29_: 46-145. Gosting, L.J. 1956. Measurement and interpretation of diffusion c o e f f i -cients of proteins. Adv. Prot. Chem. 11_: 429-548. Graupner, H. and I. Fischer. 1933. Beitrage zur Kenntnis der Goldfisch-haut T e i l I: Allgemeine und spezielle Histologie. Z. Mikrosk. Anat. Forsch. 33; 91-142. Gurr, E. 1962. Staining animal tissues. A practical theory. Leonard H i l l , London. Harris, J.E., A. Watson, and S. Hunt, 1973. Histochemical analysis of mucous cells in the epidermis of brown trout, Salmo trutta L. J. Fish Biol. 5_: 345-351. Harris, J.E., and S. Hunt. 1973. Epithelial mucins of the Atlantic s a l -mon, Salmo salar L. Trans. Biochem. Soc, 1_: 153-155. Helfferich, F. 1962. Ion exchange. McGraw-Hill, New York, pp. 81-84. Hendrikson, R.C. and A.G. Matoltsy. 1968. The fine structure of teleost epidermis. I. Introduction and filament-containing c e l l s . II. Mucous ce l l s . J. Ultrastruct. Res. 21: 194-212, 213-221. Hermann, L. 1882. Neue Untersuchung uber Hautstryme.• Pflllgers Arch. 27: 280. Holliday, F.G.T. 1969. The effects of salinity on the eggs and larvae of teleosts. In: Fish Physiology Vol. 1, W.S. Hoar and D.J. Randall eds. Academic Press, New York, pp. 293-311. 124 Isaia, J. 1972. Comparative effects of temperature on the sodium and water permeabilities of the g i l l s of a stenohaline fresh water fish (Carassius auratus) and a stenohaline marine fish (Serranus  scriba, Serranus cabrilla) . J. Exp. Biol. 5_7: 359-366. Jakubowski, M. 1963. The structure and vascularization of the skin of the river bullhead (Cottus gobio L.) and Black Sea turbot (Rhombis  maeotlcus Pall..). Acta B i o l . Cracov. Ser. Zool. 6_: 159-175. Johnson, D.W. 1973. Endocrine control of hydromineral balance in tele-osts. Amer. Zool. 13: 799-818. Jones, M.P., F.G.T. Holliday, and A.E.G. Dunn. 1966. The ultrastructure of the epidermis of the herring (Clupea hargangus) in relation to the rearing salinity. J. Mar. Biol. Ass. U.K. 46_: 235-239. Jozuka, K. 1966. Chloride-excreting cells and mucus-secreting cells in the g i l l s of the Japanese common eel, Anguilla japonica. Annot. Zool. Jap. 39_: 202-210. Kann, S. ' 1926. Die Histologie der Fischhaut von Biologischen Gesichts-punkten Betrachtet. Z. Zellforsch. Hikrosk. Anat. 4_: 482-514. Kapoor, B.G. 1965. Histological observations on the skin of the head of a Siluroid fish, Wallago attu (BI. & Schn.) Mikroskopie 20: 123-128. Karnaky, K.J. Jr. and W.B. Kinter. 1977. K i l l i f i s h opercular e p i t h e l i -um: a f l a t epithelium with a high density of chloride c e l l s . J. Exp. Zool. 199_: 355-364. Karnaky, K.J. Jr., K.J. Degnan, and J.A. Zadunaisky. 1977. Chloride transport across isolated opercular epithelium of k i l l i f i s h : a mem-brane rich in chloride c e l l s . Science 195: 203-205. 125 Kent, P.W. 1967. Structure and function of glycoproteins. Essays Biochem. 3: 105-151. Kirsch, R. 1972. The kinetics of peripheral exchanges of water and electrolytes in the silver eel (Anguilla anguilla) in fresh water and sea water. J. Exp. Bio l . 5_7: 489-512. Kitzan, S.M., and P.R. Sweeny. 1968. A light and electron microscope study of the structure of Protopterus annectens epidermis. I. Mucus production. Can. J. Zool. 46_: 767-772. Komnick, H. and M. Bierther. 1969. Zur histochemischen Ionenlokalisa-tion mit Hilfe der Elektonenmikroskopie unter besonderer Berucksich-tigung der Chloridreaktion. Histochemie 18_: 337-362. Lam, T.J. 1972. Prolactin and hydromineral regulation in fishes. Gen. Comp. Endocrinol. Suppl. _3: 328-338. La Roche, G. and CP. Leblond. 1952. Effect of thyroid preparations and iodide on Salmonidae. Endocrinology 5_1: 524-545. Lasker, R. and L.T. Threadgold. 1968. "Chloride c e l l s " in the skin of the larval sardine. Exp. Cell Res. 52: 582-590. Leatherland, J.F. and T.J. Lam. 1969. Effect of prolactin on the den-sity of mucous cells on the g i l l filaments of the marine form of the threespine stickleback. Can. J. Zool. 4_7: 787-792. Lemoine, A.M. 1974a. Action de l a prolactine chez l'anguille hypophy-sectomise"e en eau de mer. Effet sur l'acide N-ac£tyl-neuaminique branchial. C.R. Soc. Bio l . 168: 398-402. Lemoine, A.M. 1974b. Action du C o r t i s o l chez l'anguille hypophysec-tomise"e en eau de mer. Effet sur l'acide N-ace"tylneuraminique branchial. CR. Soc. Bio l . 168: 402-405. 126 Lemoine, A.M., and M. Olivereau. 1971. Presence d'acide N-ace"tylneur-aminique dans l a peau d'Anguilla anguilla L. Z. vergl. Physiol. 73: 22-33. Lemoine, A.M., and M. Olivereau. 1973a. Action de l a prolactine chez l'anguille intacte et hypophysectomisee IX. Effet sur l a teneur en acide N-ace*tylneuraminique de l a peau en eau de mer. Acta Zool. 54: 223-228. Lemoine, A.M. and M, Olivereau. 1973b, Variations de l a teneur en acide N-acetylneuraminique de l a branchie de l'anguille au cours des changements de sa l i n i t e . C.R. Soc. Biol, 167: 411-416. Lemoine, A.M. and M. Olivereau. 1974, Action du C o r t i s o l sur l'acide N-ac6tyl-neuraminique cutanee chez l'anguille intacte et hypophysec-tomisee en eau douce et apres transfert en eau de mer. Acta Zool. 55: 255-261. Leonard, J.B. and R.G. Summers. 1976. The ultrastructure of the integu-ment of the American eel, Anguilla rostrata. Cell Tiss. Res. 171: 1-30. Leppi, T.J. 1967, Histochemical studies on mucous cells in the skin and slime glands of hagfish. Anat. Rec. 157: 278. Lerche, D. 1976. Temporal and local concentration changes in diffusion layers at cellulose membranes due to concentration differences be-tween the solutions on both sides of the membrane. J. Mem. Biol. 27: 193-205. Lewis, R.W. 1970. Fish cutaneous mucus: a new source of skin surface l i p i d . Lipids 5_: 947-949. L i l l i e , R.D. 1965. Histopathologic technique and practical histochemis-try. McGraw-Hill Book Co., New York. 715 pp. 127 Lo, S., J.C. Russel, and A.W. Taylor. 1970. Determination of glycogen in small tissue samples. J. Appl. Physiol. 28_: 234-236. Logan, V.H. and P.H. Odense. 1974. The integument of the ocean sunfish (Mola mola L.) (Plectognathi) with observations on the lesions from two ectoparasites, Capsala martinerei (Trematoda) and Philortha- goriscus serratus (Copepoda). Can. J. Zool. 52_: 1039-1045. Lowry, O.H., N.J. Rosebrough, A.L. Farr and R.J. Randall. 1951. Protein measurement with the fo l i n phenol reagent. J. Bio l . Chem. 193: 265. Maetz, J. 1974. Aspects of adaptation to hypo-osmotic and hyper-osmotic environments. Biochem. Biophys. Perspect. Mar. Biol. 1_: 1-167. Maetz, J. 1976. Personal communication. (J. Maetz, Groupe de Biologie Marine, Dept. de Biologie du Commissariat a l'Energie Atomique, Station Zoologique, 06230 Villefranche-sur-mer, France). Marshall, W.S. 1976. Effects of hypophysectomy and ovine prolactin on the epithelial mucus-secreting cells of the Pacific staghorn sculpin, Leptocottus armatus (Teleostei: Cottidae) . Can. J. Zool. 54_: 1604-1609. Marshall, W.S. 1977. Transepithelial potential and short-circuit cur-rent across the isolated skin of Gillichthys mirabilis (Teleostei: Gobiidae), acclimated to 5% and 100% sea water. J. Comp. Physiol. 114: 157-165. Mattheij, J.A.M. and J.A.P. Sprangers. 1969. The site of prolactin se-cretion in the adenohypophysis of the stenohaline teleost, Anoptich- thys jordani, and the effects of this hormone on mucous c e l l s . Z. Zellforsch. 99: 411-419. 128 Mattheij, J.A.M. and H.W.J. Stroband, 1971. The effects of osmotic experiments and prolactin on the mucous cells in the skin and the ionocytes in the g i l l s of the teleost, Cichlasoma biocellatum. Z. Zellforsch. 121: 93-101. Mattheij, J.A.M., H.W.J. Stroband, F.J. Kingma, and P.G. vanOordt. 1972. Prolactin and osmoregulation in cichlid f i s h . Cichlasoma biocella-tum, and the effect of the hormone on thyroid, g i l l s and skin. Gen. Comp. Endocrinol. 18_: 607. Merrilees, M.J, 1974. Epidermal fine structure of the teleost Esox  americanus (Esqcidae, Salmoniformes). J. Ultrastruct. Res. 47: 272-283. Mittal, A,K. and T.K. Banerjee. 1974a. Structure and keratinization of the skin of a fresh water teleost, Notopterus notopterus (Notopteri-dae, Pisces). J. Zool. Lond. 174: 341-355. Mittal, A.K. and T.K. Banerjee. 1974b. A histochemical study of the epi-dermal keratinization i n the skin of a fresh water teleost Bagarius  bagarius (Ham.) (Sisoridae, Pisces). Mikroscopie 30_: 337-348. Mittal, A.K. and J.S.D. Munshi. 1969. Histochemical and autoradiographic studies of skin glands of Rita r i t a (Ham.) Bagridae, Pisces). Mikroscopie 24_: 193-205. Mittal, A.K. and J.S.D. Munshi. 1970. Structure of the integument of a resh water teleost, Bagarius bagarius (Ham.) (Sisoridae, Pisces). J. Morph. 130: 3-10. Mittal, A.K. and J.S.D. Munshi. 1971. A comparative study of the structure of the skin of certain air-breathing fresh-water teleosts. J. Zool. Lond. 163: 515-532. 129 Morris, R.W. 1960. Temperature, s a l i n i t y , and southern limits of three species of cottid fishes. Limnol. Oceanogr. 5_: 175-179. Motais, R., J. Isaia, J.C. Rankin, and J. Maetz. 1969. Adaptive changes of the water permeability of the teleostean g i l l epithelium in re-lation to external sa l i n i t y . J. Exp. Biol. 5_1: 529-546. Motais, R. and J. Isaia. 1972. Temperature dependence of permeability to water and to sodium of the g i l l epithelium of the eel, Anguilla  anguilla. J. Exp. Bi o l . 56_: 587-600. Mowry, R.W. and C.H. Winkler. 1956. The coloration of acidic carbohy-drates of bacteria and fungi in tissue sections with special refer-ence to capsules of Cryptococcus neoformans, Pneumococci, and Staphlococci. Amer. J. Path. 32.: 628-629. McBride, J.R. and A.P. van Overbeeke. 1971. Effects of androgens, es-trogens, and C o r t i s o l on the skin, stomach, l i v e r , pancreas, and kidney in gonadectomized adult sockeye salmon (Oncorhynchus nerka). J. Fish. Res. Bd. Can. 28: 485-490. McManus, J.F.A. 1946. Histological demonstration'of mucin after periodic acid. Nature 158: 202. Nagahama, Y., K. Chan, and W.S. Hoar. 1976. Histochemistry and ultra-structure of pre- and post-ovulatory f o l l i c l e s in the ovary of the goldfish, Carassius auratus. Can. J. Zool. 54: 1128-11.39. Nishioka, R. 1977. Personal communication. Dept. Zoology, U. Califor-nia, Berkeley, California, U.S.A. Ogawa, M. 1970. Effects of prolactin on the epidermal mucous cells of the goldfish, Carassius auratus L. , Can. J. Zool. 48^ : 501-503. 130 Ogawa, M. and P.H. Johansen. 1967. A note on the effects of hypophy-sectomy on the mucous cells of the goldfish, Carassius auratus L. Can. J. Zool. 45: 885-886. Olivereau, M. and A.M. Lemoine. 1971a, Action de l a prolactine chez l'anguille intacte et hypophysectomisee. VII. Effet sur l a teneur en acide sialique (E-acetyl-neuraminique) de l a peau. Z. Vergl. Physiol. 73: 34-43. Olivereau, M. and A.M. Lemoine. 1971b, Teneur en acide N-acetyl-neura-minique de l a peau chez l'anguille apres autotransplantation de 1-hypophyse, Z. Vergl. Physiol. 73_: 44-52. Olivereau, M. and A.M. Lemoine. 1972. Effects of external salinity changes on N-acetylneuraminique acid content of the skin in eels. J. Comp. Physiol. 79: 411-422. Olivereau, M. and J. Olivereau, 1971. Effect of hypophysectomy and pro-lactin treatment on g i l l mucous cells of eels. (French). C.R. Sco. Biol. 165: 2267-2271. Olsson, R. 1961. The skin of Amphioxus. Z. Zellforsch. _54: 90-104. Patton, S. and L.F. Hood. 1969, Biogenesis of milk l i p i d s . In: Lacto-genesis. M, Reynolds and S.J. Folley (eds.) U. Pa. Press, Philadel-phia. 269 pp. Pearse, A.G.E. 1968. Histochemistry: theoretical and applied, 3rd ed. Vol. 1. J. & A. Churchill Ltd. 759 pp. Pfeiffer, W, 1963. Alarm substances. Experimentia 19_: 113:1-11. Pfeiffer, W. and T.F. Pletcher. 1964. Club cells and granular cells in the skin of lamprey, J. Fish. Res. Bd. Can. 21: 1083-1088. 131 I I Pfeiffer, W. 1968. Uber die Epidermis von Latimeria chalumnae J. L.B. Smith 1939 (Crossopterygii, Pisces). Z. Morph. Tiere 63: 419-427. Pfeiffer, W. 1969. Kolbenzellen in der Epidermis der Galaxiidae, Galaxis attenuatus (Salmoniformes). Experimentia 25_: 432-433. Pickering, A.D. 1974. The distribution of mucous cells in the epider-mis of the brown trout Salmo trutta L. and the char Salvelinus albinus. J. Fish. Biol. §_: 111-118. Pickford, G.E., P.K.T. Pang, and W.H. Sawyer. 1966. Prolactin and • serum osmolality of hypophysectomized Fundulus heteroclitus in fresh water. Nature 209_: 1040-1041. Potts, W.T.W. and D.H. Evans. 1966. The effects of hypophysectomy and bovine prolactin on salt fluxes i n freshwater-adapted Fundulus heteroclitus. Biol. Bull. 131: 362-368. Potts, W.T.W. and F.B. Eddy. 1973. G i l l potentials and sodium fluxes in the flounder Platichthys flesus. J. Comp. Physiol. 87_: 29-48. Reid, E.W. 1893. The electromotive properties of the skin of the common eel. P h i l . Trans. Roy. Soc. Lond. 184: 335-365. Rosen, M.W. and N.E. Cornford. 1971. Fluid frictions of fis h slimes. Nature 234: 49-51. Roberts, R.J., H. Young, and J.A. Milne. 1971. Studies on the skin of plaice (Pleuronectes platessa L. ) I. The structure of normal plaice skin. J. Fish. Biol. 4_: 87-98. Schreibman, M.P. and K.D. Kallman, 1965. The effect of hypophysectomy on fresh water survival and scale mucous cells in two species of fresh water fishes. Amer. Zool. 5: 728. 132 Siegel, S. 1956. Non-parametric s t a t i s t i c s for the behavioral sciences. McGraw-Hill, Toronto. 312 pp. Smith, G.M. and C.W. Coates. 1937. On the histology of the skin of the . lungfish Protopterus annectens after experimentally induced aesti-vation. Quart. J. Microsc. Sci. 79_: 487-491. Sokal, R.R. and F.J. Rolf. 1969. Biometry. The principles and practice of sta t i s t i c s in biological research. W.H. Freeman & Co. San Fran-cisco. 776 p. Spicer, S.S. and J. Duvenci. 1964. Histological characteristics of muco-polysaccharides i n salivary and exorbital lacrimal glands. Anat. Res. 149: 333-357. Stahelin, L.A. 1974. Structure and function of intercellular junctions. Internat. Rev. Cytol. 39_: 191-284. Stanley, J.G. and J.K. O'Connell. 1970. Effects of hypophysectomy on plasma electrolytes and epidermal mucous cells in the mud minnow Umbra limi. Amer. Zool. 10_: 298. Walvig, F. 1960. The integument of the icef i s h (Chaenocephalus aceratus Lonnberg) . Nytt. Mag. Zool. 9_: 31-36. Wang, J.H. 1951. Self-diffusion and structure of liquid water. I. Measurement of self-diffusion of liquid water with deuterium as tracer. J. Amer. Chem. Soc. 7_3: 510-513. Wang, J.H., CB. Anfinsen, and F.M. Polestra. 1954. The self-diffusion coefficients of water and albumin in aqueous ovalbumin solutions at 10°. J. Amer. Chem. Soc. 76: 4763-4765. Warren, L. 1959. The thiobarbituric acid assay of s i a l i c acids. J. Biol. Chem. 234: 1971-1975. 133 Weisel, G.F. 1975. The integument of the paddlefish, Polyodon spathula. J. Morph. 145: 143-150. Wessler, E. and I. Werner. 1957. On the chemical composition of some mucous substances of fi s h . Acta Chem. Scand. 1_1: 1240-1247. White, A., P. Handler, and E.L. Smith. 1973. Principles of biochemistry. 5th ed. McGraw-Hill, New York. 1296 pp. Whitear, M. 1970. The skin surface of bony fishes. J. Zool. Lond. 160: 437-454. Whitear, M. 1971. Cell specialization and sensory function i n fi s h epi-dermis. J. Zool. Lond. 163: 237-264. Willoughby, L.G. 1971. Observations on fungal parasites of lake dis-t r i c t salmonids. Salm. Trout Mag. 192: 152-158. Willoughby, L.G. 1972. U.D.N, of lake d i s t r i c t trout and char: outward signs of infection and defense barriers examined further. Salm. Trout Mag. 195_: 149-158. Winn, H.E. 1955. Formation of a mucous envelope at night by parrot fis h . Zoologica 40: 145-147. Wittouck, P.J. 1975. Action de l a prolactine sur les cellules a mucus epidermiques chez l'axolotl intacte et hypophysectomisee. Gen. Comp. Endocrinol. 2_7: 254-261. Wold, J.K. and R. Selset. 1977. Glycoproteins in the skin mucus of the char (Salmo alpinus L.). Comp. Biochem. Physiol. 56B: 215-219. Wright, D.E. 1974. Morphology of the g i l l epithelium of the lungfish Lepidosiren paradoxa. Cell Tiss. Res. 153: 365-381. 134 Yamazaki, F. and E.M. Donaldson. 1968, The effects of part i a l l y puri-fied salmon pituitary gonadotropin on spermiogenesis, vitellogene-s i s , and ovulation in hypophysectomized goldfish (Carassius auratus). Gen. Comp. Endocrinol. 11: 292-299. Yamazaki, F. 1972. Effects of methyltestosterone on the skin and the gonad of salmonids. Gen. Comp. Endocrinol. Suppl. 3_: 741-750. Zerbe, W.B. and C.B. Taylor. 1953. Sea water temperature and density reduction tables. U.S. Dept. Commerce, Coast and Geodetic Survey Spec. Pub. No. 298. U.S. Gov't. Printing Office, Washington, D.C. 21 pp. OTHER REFERENCES Hart, J.L. 1973. Pacific fishes of Canada. Fish. Res. Bd. Can. Bull. 180, 740 pp. Goldstein, D.J. and R.W. Horobin 1974. Rate factors in staining by Alcian Blue. Histochem, J. 6_: 157-174. Jenkins, J.T. 1936. The fishes of the British Isles. F. Warne & Co., London. 408 pp. Katchalsky, A. and P.F. Curran. 1967. Nonequilibrium thermodynamics in biophysics. Harvard U. Press, Cambridge, Mass. Leim, A.H. and W.B. Scott. 1966. Fishes of the Atlantic coast of Canada. Fish. Res. Bd. Can. Bull, 155: 485 pp. Scott, W.B. and E.J. Crossman. 1973. Fresh water fishes of Canada. Fish. Res. Bd. Can, Bull. 184. 966 pp. Sterba, G. 1962, Fresh water fishes of the world. Vista Books, London. 878 pp. (Translated & revised by D.W. Tucker). 134a APPENDIX 1 135 Appendix 1. Measurement of Diffusion: The Capillary Method The open-ended capillary method i s in a sense intermediate between free and restricted diffusion. It is well suited for studies of d i f f u -sion using radioactive tracers. Capillary tubes are f i l l e d with a solution of the radioactively tagged solute and stoppered at one end. The tubes are then immersed in a well-stirred solvent, and l e f t for a diffusion time, t. The solute con-centration at the open end of the tube is maintained at near-zero by the sti r r i n g of the solvent. The diffusion distance i s the height of the capillary, h, and the i n i t i a l concentrations of solute is designated C^. The solute concentration, C, may be expressed as an i n f i n i t e series (1). The diffusion coefficient for the solute i s 'D'. . 4C° °? (-1)1 -(21 + l ) 2 H2 Dt/(4h 2) (21 + l)TTx (1) C = T~ . E n 2i + 1 e c o s 2h i=0 After diffusion time, t, the average concentration, C , which is obtained by mixing and analysing the contents of the tube, is given by (2) C a V 1 -h C dx = 8 » 1 -(21 + 1 ) V Dt/(4h 2) (2) C° " hC° 0 ~^ i-0 (21 + I ) 2 6 2 For relatively long times (Dt/h >0.2), a l l but the f i r s t term of the equation (2) may be neglected, giving (3). (3) Dt 4 In 2 2 h IT , 2 ~ c a V x * This information is taken from Gosting (1956) 136 A more complete proof of the method appears in Jost, 1952 (cited in Gosting, 1956). In favourable cases the open-ended capillary method is accurate + 1% and is incapable of yielding accuracy comparable to the diaphragm c e l l method, free diffusion methods using interferometric optical sys-tems, or to Harned's conductance method. With suitable rates of s t i r -ring and longer capillaries the error due to convection at the open end of the tubes may be minimized. 136a APPENDIX 2 137 Appendix 2 Comparative Histology of the Cutaneous Epithelia of Pisces. Abbreviations used: Sal. : Environmental sa l i n i t y ; (SW) sea water or (FW) fresh water. If euryhaline capability is recognized, this i s represented by (E). C.S. : Cuticle-secreting c e l l s . Easily recognized in electron microscopy, but may not be seen in histological prepara-tions. Gob. : Mucus-secreting goblet cel l s . EGC : Eosinophilic granular cells (also called Kornerdrusen, sacciform granular c e l l s , albuminous c e l l s , and serous c e l l s ) . Club : Club celis (also called Kolbenzellen, Schreckstoffzellen, and alarm substance cells) . Other : Includes specialized glands or other structures that have been shown in but a few species. E/D : Thickness of the epidermis (E) and dermis (D) in microns. Type of Mucous: As established using histochemical or chemical tech-niques; s i a l i c acid-containing glycoprotein (COOH), sul-phated mucins (SO^). Answers which appear in parentheses concern structures that were recognized i n figures or drawings, but were not elaborated by the authors. Question marks imply insufficient data. Species C e l l Types  S a l . C.S. Gob. EGC Club Other Scales Type Mucous E/D Reference CEPHALOCHORDATA Anphioxus lanceolatus (SW) yes no no no no no a c i d i c 7-8/0 Olsson, 1961. VERTEBRATA AGNATHA Myxiniformes Blackstad,1963. Myxine g l u t i n o s a (SW) yes yes no no thread c e l l no a c i d i c (S0 4 + COOH) 90/90 F e r r y , 1941. Leppi, 1967. Wessler & Werner, 1963. Petromyzoniformes Ichthyomyzon unicusDus (FW) yes yes yes yes no no _ 300/ Downing & Lampetra planera (FW) ? no yes yes no no - 160/ Novales, 1971. P f e i f f e r & Pletcher, 1964. CHONDRICHTHYS Raja S D D . (SW) ? yes no no no yes a c i d i c 45/70 Kann, 1926. Wessler & Torpedo marmorata (SW) ? yes no no no yes (COOH) a c i d i c (S0 4 + COOH) 80/ Werner, 1963. Kann, 1926. Carmignani & Scy l l i u m canicula (SW) ? yes no no no yes _ 30/130 Zaccone, 1974b Kann, 1926. S c y l l i o r h i n u s c a n i c u l a (SW) a c i d i c (S0 4 + COOH) Carmignani & Zaccone, 1974b Acanthias vu l g a r i s (SW) ? no no no no yes - 160/ Kann, 1926. S_pinax niger (SW) ? yes no no no yes - 120/280 Kann, 1926. Sauatina angelus (SW) ? yes no no no yes - 80/60 Kann, 1926. HOLOCEPKALI Chimaera raonstrosa (SW) ? • no no no no no - 330/150 Kann, 1926. CO oo C e l l Types Species S a l . C.S. Gob. EGC Club Other Scales Type Mucous E/D Reference OSTEICHTHYS; CROSSOPTERYGII Coelocanthiformes Latimeria chalumnae (SW) no yes no no serous' c e l l •* a c i d i c 430/ P f e i f f e r , 1968. Dipnoi Protopterus annectens (FW) yes yes no no 3 t y p e s ° f ve » ' • * • » gob. c e l l s * Lepidosiren sp. (FW) yes yes no no no yes 80/ Smith & Coates, 1936. K i t z a n & Sweeny, 1968. 20/ Wright, 1974. OSTEICHTHYS; CHONDROSTEI Acipenseriformes Polvodon SDathula (FW) ? no no ? 150/500 Weisel, 1975. OSTEICHTHYS; H0L0STEI Amia calva (FW) ? yes no no no yes 100/ Kann, 1926. C e l l Types Species S a l . C.S. Gob. EGC Club Other Scales Type Mucous E/D Reference OSTEICHTHYS; TELEOSTEI A r . g u i l l i f ormes A n g u i l l a japonica A n g u i l l a a n g u i l l a Conger v u l g a r i s Muraena helena (FW) (E) (FW) (E) (SW) (SW) ? yes no yes ? yes no yes ? yes no yes ? yes yes no no (yes) (yes) (yes) a c i d i c (COOH) a c i d i c (COOH) 200/ Asakawa, 1970. 200/220 Kann,- 1926. Bremer, 1972. Hendrikson & Matoltsy. 1968. 180/110 Kann, 1926. 90/400 Kann, 1926. Salmoniformes Esox americanus Esox l u c i u s Salmo g a i r d n e r i  Salmo t r u t t a Salmo f a r i o  S a l v e l i n u s alpinus Galaxias attenuatus (FW) yes yes no yes C l ~ c e l l s (FW) (FW) (FW) (FW) (FW) yes yes no (FW) no yes no no yes ? yes ? yes ? no no yes no no yes yes yes yes yes yes yes a c i d i c a c i d i c (COOH) a c i d i c (S0 4 + COOH) a c i d i c (COOH) a c i d i c (COOH) 50/ M e r i l l e s , 1974. 100/40 Bremer, 1972. Kann, 1926. Wessler & Werner, 1957. / F l e t c h e r et a l . . 1976. 100/ P i c k e r i n g , 1974. H a r r i s e t a l . , 1973. 140/ Kann, 1926. / P i c k e r i n g , 1974. P f e i f f e r , 1969. Clupeiformes Clupea hargenus (larvae) (SW) yes no no no Notopterus notopterus (FW) yes yes no no yes yes a c i d i c (so 4 ) / Jones et a l . , 1966. 60/500 M i t t a l 6. Banerjee, 1974a C e l l Types Species S a l . C.S. Gob. EGC Club Other Scales Type Mucous E/D Reference OSTEICHTHYS; TELEOSTEI (con't) Cypriniformes Carassius auratus (FW) ? yes no yes no yes 80/ Graupner & F i s c h e r . 1933 Hendrikson & Matoltsy,1968 Phoxinus phoxinus (FW) yes yes no yes no yes - 75/40 Whitear,1970. Phoxinus l a e v i s (FW) ? yes ? yes no yes - 75/40 Kann, 1926. I c t a l u r u s sn. Heteropneustes f o s s i l i s (FW) (FW) yes (yes) yes yes 7 no yes yes no no no a c i d i c <so4) 100/ Whitear, 1970. Pfeiffer,1 9 6 3 . M i t t a l & Munshi, 1971. CyDrinus carpio (FW) 7 yes ? yes no yes a c i d i c 160/3 0 Bremer, 1972. Kann, 1926. Saualius ceohalus (FW) ? yes ? ? yes - 100/ Kann, 1926. Leuciscus r u t i l u s (FW) ? yes 7 7 yes - 43/160 Kann, 1926. Abra.mis brama (FW) 7 yes ? no yes - 70/90 Kann, 1926. Gobio f l u v i a t i l i s (FW) ? yes ? no yes - 50/40 Kann, 1926. Alburr.us lucidus (FW) 7 yes ? no yes - 170/ Kann, 1926. Cyprinodontiformes P o e c i l i a l a t i p i n n a (FW) ye 3 yes no yes yes - 10/20 Hendrikson & . Matoltsy,1968 S i l u r o i d e a Wallago attu (FW) ? yes yes yes no no - 40/ Kapoor, 1965. I - 1 C e l l Types Species S a l . C.S. Gob. EGC Club Other Scales Type Mucous E/D Reference OSTEICHTHYS; TELEOSTEI, COn't S i l u r o i d e a , con't S i l u r i d a e S i l u r u s qlandis (FW) ? yes ? yes - no - 80/20 Kann, 1926. Doradidae Doras d o r s a l i s (FW) 7 yes (no) yes - no •- 80/ B h a t t i , 1938. Bagridae R i t a r i t a (FW) ? yes (no) yes - no a c i d i c ( s o 4 ) 170/ B h a t t i . 1938 M i t t a l & Munshi, 1969. Malopteruridae Malopterus e l e c t r i c u s (FW) ? yes no yes - no a c i d i c ( s o 4 ) 240/130 Carmignani & Zaccone, 1974a Kann, 1926. Callichthydae C a l l i c h t h y s p e c t o r a l i s (FW) ? yes ? yes - no - 40/ B h a t t i , 1938. Corvdoras s c h u l t z e i (FW) ? yes ? (yes) - no a c i d i c Bremer, 1972. Corvdoras aenaus (FW) ? yes Hendrikson & Matoltsy, 1968 L o r i c a r i idae P l e c t o s t o m u s commersonii A n c i s t r u s k n e r L o r i c a r i a cataohracta Otoclinus nigricauda (FW) (FW) (FW) (FW) ? ? ? •7 yes yes yes yes yes no no yes no no no no -no no no no -30/ 11/ 25/ 20/ B h a t t i , 1938. B h a t t i , 1938. B h a t t i , 1938. B h a t t i , 1938. S i s o r i d a e Bagarius bagarius Zeiformes (FW) ? yes (no) no k e r a t i n i z e d no 10-50/ M i t t a l & Banerjee,1974b M i t t a l & Munshi, 1970. Zeus faber (SW) ? yes ? - yes n e u t r a l Bremer, 1972. C e l l Types Species S a l . C.S. Gob. EGC Club- Other Scales Type Mucous E/D Reference TELEOSTEI, con't Synbranchia Amphipnous cuchia (FW) Opisthomi Mastacembelus pancalus (FW) Echeneiformes Echeneis naucrates Gadiformes Gadus c a l l a r i a s Lota l o t a  Zoarces viviparus Gasterosteiformes (SW) (SW) (FW) (SW) Sygnathus typhle yes yes yes yes yes yes yes yes yes ? (no) (no) yes yes yes yes no Gasterosteus aculeatus (SW) yes yes no (E) Hippocampus cuda (SW) yes no Hippocampus b r e v i r o s t i r i s ( S W ) ? no Kerophis ophidion (SW) yes no no no no no (SW) yes no C l c e l l s yes yes yes yes yes a c i d i c <so 4) a c i d i c ( so 4 ) a c i d i c a c i d i c (COOH) a c i d i c (COOH) a c i d i c (COOH) a c i d i c (COOH) a c i d i c a c i d i c (COOH) 120/ M i t t a l & Munshi, 1971. 30/ M i t t a l & Munshi, 1971. Bremer, 1972. Wessler & Werner, 1957. 360/180 Bremer, 1972. Kann, 1926. 60/60 Bremer, 1972. a c i d i c 10-30/ Kann, 1926. Whitear, 1971. Bock, 1930. Bremer, 1972. Bremer, 1972. 30/270 Kann, 1926. 50/30 10/ Kann, 1926. Bremer, 1972. Bremer, 1972. C e l l Types Species S a l . C.S. Gob. EGC Club Other Scales Type Mucous E/D Reference TELEOSTEI, con't Perciformes Gobiidae G i l l i c h t h y s m i r a b i l i s Gobisocidae Lepodichthy3 l i n e a t u s  Lepadogaster sp. Blennidae Blennius pholis ' Trachinidae Trachinus dracho Grammistidae Pogonoperca punctata Anabantidae Betta splcndens  Ar.abas scandens (SW) (E) (SW) (SW) (FW) (FW) (no) yes yes no C l ~ c e l l s (SW) yes (SW) ? Monodactylidae Pteroshyllum sp. (FW) ? Stromate idae Schedophilus medusophagus(SW) ? Ophidiidae Ophidium barbatum yes yes (SW) yes yes yes yes yes no ? yes yes yes yes no no C l ~ c e l l s no no yes yes no dermal p o i - yes son glands yes yes (SW) yes yes k e r a t i n i z e d yes yes a c i d i c (COOH) a c i d i c a c i d i c (COOH) 180/120 Personal obs. 200+/ F i s h e l s o n . 1972. 70/20 Kann, 1926. Whitear, 1970. Whitear, 1971. 30/ Bremer, 1972. 100/ Aida et a l . , 1973. a c i d i c 23/ Bremer, 1972. 30/24 Kann, 1926. 10/ Bremer, 1972. Bone & Brook, 1973. 170/90 Kann, 1926. C e l l Types Species S a l . C.S. Gob. EGC Club Other Scales Type Mucous E/D Reference TELEOSTEI, con't Perciformes, con't Scaridae Scarus dubius Scarus perspici11atus Percidae Serranus s c r i b a Labridae (SW) (SW) Coris g e o f f r e d i i (SW) B a l i s t i d a e B a l i s t e s capriscus (SW) Family? Chaenocephalus aceratus (SW) Scorpaeiformes T r i g l i d a e T r i q l a lucerna T r i g l a corax Cyclopteridae Cyclopterus lumpus Cottidae Cottus scorpio  Cottus qobio (SW) (SW) (SW) (SW) (FW) yes yes yes yes no yes yes ? yes yes yes ? m u l t i c e l l u l a r yes glands poster, to g i l l yes yes yes yes (yes) no C l c e l l s yes yes yes no no no yes yes no no a c i d i c n e u t r a l Byrne, 1970. 440/35 Kann, 1926. 45/30 Kann, 1926. 20/ Kann, 1926. 150/300 Walvig, 1960. / 40/ / Whitear, 1970. Whitear, 1971. Kann, 1926. Bremer, 1972. 30/ Bremer, 1972. 110/90 Jakubowski,1963. C e l l Type Species S a l . C.S. Gob. EGC Club TELEOSTEI, con 11 Scorpaeniformes, con't Cottidae, con't Cottus asper (FW) ? yes yes no Cottus b a i r d i (FW) ? yes yes no Cottus a l e u t i c u s (FW) ? yes no no Cottus rhotheus (FW) ? yes no no I c e l i n u s b o r e a l i s (SW) ? yes yes no Myxocephalus sp. (SW) ? yes yes no Oligocottus snyderi (SW) yes yes yes no Oligocottus maculosis (SW) ? yes yes no A s c e l i c h t h y s rhodorus (SW) 7 yes yes no Leptocottus armatus (SW) (E) yes yes yes no suronectiformes Pleuronectes platessa (SW) •> yes yes no P l a t i c h t h y s flesus (SW) (E) 7 yes ? no HiDDoqlossoides elassodon(SW) yes yes ? no Rhombus maeoticus (SW) ? yes ? no Other Scales Type Mucous E/D Reference yes no no no yes yes no no a c i d i c (so4) a c i d i c ( s o 4 ) n e u t r a l + a c i d i c n e u t r a l n e u t r a l + a c i d i c (so4) n e u t r a l + a c i d i c ( S 0 4 ) n e u t r a l + a c i d i c (SO,) 4 a c i d i c (COOH) 90/110 t h i s study 90/500 t h i s study 100/120 t h i s study 100/120 t h i s study 160/180 t h i s study 120/250 t h i s study 180/140 t h i s study 200/180 th i s study 220/180 t h i s study 70/120 t h i s study yes yes yes yes a c i d i c (COOH) neut r a l + acidic(COOH) Roberts et a l . , 30/100 1971. Fl e t c h e r et a l . . 1976. Wessler & Werner, 1957. F l e t c h e r et a l . . 1976. 30/ Brown & Wellings,1970. 35/90 Jakubowski,1963. C e l l Types Species S a l . C S . Gob. EGC Club Other Scales Type Mucous E/D Reference TELEOSTEI, con't Tetraodontiformes Tetraodon l a e v i s (SW) Mola mola (SW) yes ? no yes (yes) no yes yes neutral + acidic(COOH) 210/ Bremer, 1972. Logan & Odense, 1974. S a l . : s a l i n i t y (FW) freshwater (SW) seawater (E) euryhaline. C.S.: C u t i c l e - s e c r e t i n g c e l l s . Gob.: Mucus-secreting goblet c e l l s . EGC : E o s i n o p h i l i c granular c e l l s . Club: Club c e l l s . E/D: Thickness of epidermis (E) and dermis (D) i n microns. 147a APPENDIX 3 148 TRANSEPITHELIAL POTENTIAL AND SHORT-CIRCUIT CURRENT ACROSS THE ISOLATED SKIN OF GILLICHTHYS MIRABILIS (TELEOSTEI:GOBIIDAE), ACCLIMATED TO 5% AND 100% SEA WATER* by William S. Marshall as the work appeared in Journal of Comparative Physiology • B 114: 157-165 (1977) With financial assistance from the National Research Council of Canada through a scholarship to W.S.M., a grant-in-aid of research to W.S. Hoar and from the National Science Foundation through a grant (BMS-16345) to H.A. Bern. 149 SUMMARY Skin samples taken from the scaleless and well-vascularized area of the lower jaw were used in a modified Ussing chamber to test for electrogenic ion transport. In the absence of electrochemical gradients the skin developed a transepithelial potential (TEP) of 10-30 mV, serosa-positive, with a short-circuit current (SCC) of 13.7 + 3 or 20.8 + 7 2 uamp/cm for fish acclimated to 5% or 100% sea water, respectively. Iodoacetamide + 2,4-dinitrophenol, ouabain, or acetazolamide rapidly Inhibited the TEP and SCC when perfused on the serosal side, but had l i t t l e effect when added to the mucosal side. Sodium-, potassium- or chloride-free Ringer, on both sides, reversibly reduced the TEP and SCC to near zero. The results indicate active ion transport across Gillichthys skin and suggest a functional chloride excreting pump in the skin of seawater-adapted fi s h . 150 INTRODUCTION Ion transport across the g i l l and gut of teleosts has been well characterized, and the importance of these organs in osmoregulation is generally accepted (Maetz, 1970, 1971; Maetz and Pic. 1975). The kidney and urinary bladder are also active osmoregulatory organs, but they are considered to play lesser roles (Maetz, 1971). To date the skin of teleosts has received l i t t l e attention since previous studies have failed to demonstrate an osmoregulatory function. Motais e_t a l . (1969) demonstrated the impermeability of eel skin to water. Kirsch (1972) showed that there i s no passive movement of chlor-ide across the general body surface of the silver eel, but suggested that the exceptional thickness of the skin might have contributed to this im-permeability. The transepithelial potential and short-circuit current across trout skin are zero (Fromm, 1968), indicating a lack of electro-genic ion transport, but the skin proved d i f f i c u l t to remove and the mucus and scales may have been disturbed during dissection. Both Bern (1975) and Johnson (1973) mention that the problem of ion transport across fish skin has not been properly considered, but state that osmoregulatory hormones, prolactin and ACTH/Cortisol, may act on the skin by altering mucus secretion. In the present study the skin of the Goby, Gillichthys mirabilis, was examined for electrogenic ion transport in an attempt to confirm the findings of Fromm (1968). To date no attempt has been made to examine the el e c t r i c a l properties os skin from seawater-acclimated f i s h . The Goby was chosen because i t is a euryhaline marine fish and because i t s skin i s thin and devoid of scales on certain areas of the body. 151 MATERIALS AND METHODS Fish were collected from the southern shallows of San Francisco Bay and transferred to indoor tanks at the University of California, Berkeley. The fish (length: 10-15 cm) were kept for at least 10 days in 100% a r t i -f i c i a l sea water or 5% sea water (salinity: 32 o/oo and 1.6 o/oo, respec-tively) at 12-14°C under a controlled photoperiod of 12L/12D and fed daily with live brine shrimp. Fish were k i l l e d by decapitation and the skin samples were carefully dissected from the ventral side of the lower jaw. In this area the skin i s scaleless and easily dissected without damage. The membrane chamber (Fig, 1) i s composed of two plexiglass half-chambers, each with a volume of 0.25 ml. The membrane aperture is a 1.0 cm diameter cir c l e (0.78 cm ), Ringer solutions were kept at 14 + 1 C in a controlled temperature bath and were perfused through the half-chambers using a p e r i s t a l t i c pump (Buchler Instruments). At normal speeds the pump exchanged the solutions i n the half-chambers once every five seconds, yielding an excellent s t i r r i n g of the membrane surface and minimizing effects of unstirred layers. The transepithelial potential (TEP) was measured via 1% agar-3 M KCl and calomel electrodes connected to an electrometer (Keithley Instruments) and chart recorder (Houston Instruments). The short-circuit current (SCC) was applied via platinum wire loops cemented to the inside of each half-chamber. Power for the SCC was supplied by a 9V c e l l and regulated through a series of potentio-meters (5000, 500, 100, and 10 kohm) and readings were taken with a micro-ammeter (range: 0-50 uamps). The system has a resistance of 35-40 ohm and a potential of 0.0 mV with Ringer solution (no membrane) in the chamber. 152 Figure 1. The apparatus used in the experiments. The circulating , system for perfusing Ringers solutions through the half-chambers is shown above the chamber, and the elec-t r i c a l circuits for applying the short-circuit cur-rent, and for measuring the transepithelial potential are shown at the bottom. The chamber (shaded) is des-cribed in the text. 152a 3M KCi 3M KCl 153 A l l readings were taken after a steady TEP was reached. A l l TEP read-ings are corrected for electrode asymmetry (usually less than 1 mV). Membranes that had high conductance and developed low TEP were considered to have extensive edge damage and were rejected. A r b i t r a r i l y , membranes which failed to develop a TEP at least 8 mV were rejected. Conductance was calculated from the change in TEP in response to small currents (2, 4, 6 and 10 uamp.). Rectification of the conductance was checked by revers-ing current flow. The composition of Gillichthys Ringer (hereafter termed "Ringer") was: Na + 160, C l " 143, HC0~ 25, C a 2 + 4,8, K + 4.3, and Mg 2 + 2.5 mEq/1 with 0.5 g/1 -lucose added. The Gillichthys Ringer had a total osmolarity of 340 milliosmols and a f i n a l pH of 7.5 +0.2, after bubbling with air. Potassium-free Ringer was made by replacing KCl with NaCl. Sodium-free Ringer was obtained by replacing NaCl with choline chloride and replacing NaHCOg with KHCO^. The sodium-free Ringer thus had an elevated potassium concentration (20 mEq/1) compared to normal Ringer (4.3 mEq/1). Chloride-free Ringer was made by replacing NaCl and KCl with Na2S0^ with K^SO^, respectively, and replacing CaC^ with calcium acetate. The modified Ringer solutions were checked and corrected for pH. The metabolic inhibitors, 0.25 mM 2,4-dinitrophenol, 0.25 mM iodo-acetamide, and 0.5 mM ouabain, were dissolved directly i n Ringer. Aceta-zolamide (Sigma), 5.0 mM, was dissolved in 0.1 N NaOH, neutralized with 0,1 N HC1, and added to Ringer solution, allowing for the additional salts from the neutralization. Samples of intact skin, dissected skin, and skin after use in the chamber were fixed i n Bouin's solution, embedded in paraffin, sectioned 154 at 5 ym and stained with haematoxylin-eosin to examine the skin and to visualize the extent of membrane damage during the experiments. The Mann-Whitney U test (Siegel, 1956) was used for comparisons be-tween means, as sample sizes were small. A l l means are expressed as the mean Plus or minus standard error. RESULTS The skin from scaled parts of the body is not suitable for in vitro studies since membranes dissected from these areas failed to develop a TEP. The lack of a TEP was attributed to the tearing of the epidermis around the scales during dissection, as could be seen with a dissecting microscope. On the other hand, when skin samples were dissected from the scaleless ventral side of the lower jaw the epidermis was not notice ably damaged. The only apparent effect of dissection was the discharge of some of the mucus-secreting c e l l s . The dissected membrane included the epidermis, dermis, and some of the hypodermis, including chromato-phores. The thickness of dissected skin was 0.1-0.3 mm. As with a l l membrane chambers that seal by applying pressure to the membrane, this chamber caused considerable "edge damage", but this was restricted to a 50-100 ym wide strip around the perimeter of the exposed membrane. A steady-state TEP was reached in 10-60 min after the start of per-fusion with Ringer on both sides of the membrane. The TEP and SCC were stable for several hours. In both 5% sea water and 100% sea water group the TEP was 10-30 mV, serosa-positive, and there was no significant d i f -ference between the two groups (Table 1). The SCC and conductance were significantly lower in the 5% sea water group. There was no apparent 155 Table 1. Effect of salinity acclimation on the transepithelial potential, short-circuit current, and conductance of the isolated skin of Gillichthys mirabilis. Acclimated n TEP SCC 2 Conductance salinit y (mV) (uamp/cm ) (x 10~3 mho/cm^ ) 5% sea water 8 +16.9+5.8 13.7+3.0* 0.87+0.19a 100% sea water 11 +15.2+3.6 20.8+7.2 1.40+0.42 a P<0.025 (Mann-Whitney U Test) 156 rectification of the conductance i n the few (6) membranes examined for this property. The metabolic inhibitors, 2,4-dinitrophenol and iodoacetamide, in combination, markedly reduced the TEP and SCC of skins from both 5% sea water and 100% sea water groups, when these inhibitors were perfused on the serosal side of the membrane (Table 2). The TEP started to decrease in 2-6 min after the start of the perfusion and a steady-state TEP was reached in 35-60 min. The inhibition of the TEP and SCC was accompanied by an apparent decrease in conductance (Table 2). However, there was no marked change in the TEP when the dinitrophenol-iodoacetamide combination was perfused on the mucosal side. Ouabain or acetazolamide, when per-fused on the serosal side, also reduced the TEP and SCC across skins from both groups (Table 2), and these decreases were also accompanied by de-creases i n the conductance. Ouabain and acetazolamide appeared to be faster acting than dinitrophenol plus iodoacetamide, because with the for-mer the TEP started to decrease i n 1-3 min and a steady-state TEP was reached in 12-26 min. Ten-minute exposures of the mucosal side to ouabain or acetazolamide had no effect on the TEP (see sample tracing, Fig. 2). When sodium-, potassium-, or chloride-free Ringer was perfused on both sides of the skin, the TEP and SCC decreased sharply (steady-state TEP reached in 8-17 min) to near zero (Tables 3,4). The TEP was also markedly reduced when only the serosal side of the membrane was exposed to these solutions (Table 3). The reduction in TEP and SCC from these treatments was freely reversible, as both values recovered rapidly when normal Ringer was perfused on both sides of the membrane (Fig. 2). Sodi-um- or chloride-free Ringer, on the mucosal side, caused a net increase Table 2. Effect of metabolic inhibitors (DNP-1AA = 2,4-dinitrophenol + iodoacetamide) on the transepithelial potential, short-circuit current, and conductance of the isolated skin of Gillichthys mirabilis Acclimated s a l i n i t y (% sea water) Inhibitor TEP 3 l TEPa % change sccj SCCa % change % change 5% DNP-1AA +19.6 +6.4 -67 5% DNP-1AA +30.4 +2.8 -94 100% DNP-1AA +13.6 +1.6 -88 100% DNP-1AA +14.6 +3.6 -73 16.0 3.6 -78 1.18 1.01 -14 100% DNP-1AA + 9.6 +1.0 -90 12.4 1.0 -92 1.29 0.97 -25 5% Ouabain +14.8 +0.8 -95 13.3 0.5 -96 0.90 0.61 -32 100% Ouabain +12.0 +0.5 -96 9.0 0.2 -98 0.75 0.41 -45 100% Ouabain +10.8 0.0 -100 30.8 0.0 -100 2.60 2.30 -11 5% Acetazol + 8.0 +0.6 -93 100% Acetazol + 9.2 +0.8 -91 13.0 0.9 -93 1.41 1.12 -21 TEP i and TEP f = i n i t i a l (Ringers on both sides) and fi n a l (after inhibitor) values of TEP in m i l l i v o l t s i 2 SCC^ and SCC^ = i n i t i a l and fi n a l short-circuit currents in uamp/cm . -3 2 C. and C,. = i n i t i a l and fi n a l conductance as 10 mho/cm membrane 158 Figure 2. This sample tracing shows the results from a 100% seawater-adapted fi s h . Readings of the TEP and SCC were taken from the chard record. The membrane was exposed to Na-free Ringer f i r s t on the mucosal side, then on the serosal side, followed by Cl-free Ringer on the serosal side; 0,5 mM ouabain on the mucosal side; and ouabain on the serosal side. R signifies rinsing periods when the membrane was exposed to Ringers on both sides. SCC readings were taken only when the solutions on both sides of the membrane were identical. 158a T E P ( mV, s e r o s a + ) S C C (/jamp/cm ) T ime (20 minute i n t e r v a l s ) 159 Table 3. Effect of sodium-, potassium- and chloride-free Ringer on the transepithelial potential across Gillichthys mirabilis isolated skin Acclimated Salinity (% sea water) n Perfused Serosa solution on Mucosa TEP^ TEP^ % change 5% 1 Na-free Na-free +40 0.0 -100 100% 2 Na-free Na-free +24 0.0 -100 5% 4 Na-free Ringer +24.3 +0.3 - 99 (+4.5) (+0.2) 100% 1 Na-free Ringer +15.2 +3.6 - 76 5% 3 Ringer Na-free +21.7 +35.3 + 39 (+2.7) (+3.2) 100% 2 Rlnger Na-free +13.5 +24.0 + 78 5% 1 Cl-free Cl-free +33.5 +0.8 - 98 100% 1 Cl-free Cl-free +18.4 +0.8 - 96 5% 1 Cl-free Ringer +33 -3.2 -109 100% 2 Cl-free Ringer +17 -0.8 -105 5% 1 Ringer Cl-free +28 +36 + 29 100% 1 Ringer Cl-free +15 +19 + 27 5% 2 K-free K-free +16 + 3 - 81 100% 2 K-free K-free +16 + 2 - 87 5% 1 K-free Ringer .+19 + 7 - 64 100% 1 K-free Ringer +17 +12 - 29 5% 2 Ringer K-free +23 +24 + 4 100% 2 Ringer K-free +14 +16 + 14 with ringer on both sides 'after steady-state TEP reached (usually within 10 min) 160 Table 4. Effect of sodium-, potassium-, and chloride-free Ringer on the short-circuit current across the isolated skin of Gillichthys mirabilis Acclimated Salinity (% sea water) Solution Serosa Mucosa s e c * SCC* % chan; 5% Na-free Na-free 21.2 1.0 -95 100% Na-free Na-free 17.8 0.0 -100 5% Cl-free Cl-free 21.6 0.4 -98 100% Cl-free Cl-free 18.4 0.9 -95 5% K-free K-free 12.3 2.0 -83 100% K-free K-free 18.1 1.9 -90 ge SCC. and SCC,. = short ci r c u i t current in uamp/cm membrane. SCC. = i t I i n i t i a l value with Ringer on both sides of the membrane. SCC^ = fi n a l value after steady-state TEP reached with Na-, C1-, or K-free Ringer 161 In the TEP (Table 3) after an i n i t i a l "overshoot" (Fig. 2). Potassium-free Ringer produced no consistent change when perfused on the mucosal side (Table 3). There appeared to be no difference between 5% sea water and 100% sea water fish in the responses of the skin to the modified Ringer solutions. Two membranes were exposed to phosphate-buffered Ringer, instead of the usual bicarbonate-Ringer. The TEP was slightly reduced from +11 and +10 mV to +7 and +9 mV, respectively), as was the SCC (from 3) and 14 2 2 uamp/cm to 21 and 13 uamp/cm , respectively) after replacement with phos-phate Ringer. DISCUSSION The skin of Gillichthys is apparently well supplied with blood; large blood vessels are present in the hypodermis and many capillaries are pre-sent in the hypodermis and dermis. Gillichthys is well known for i t s a b i l i t y to survive out of water, i f kept moist. Thus the skin, with i t s generous blood supply, may serve as an accessory respiratory surface. The blood supply to the skin could also supply energy and substrate for a highly aerobic ion transport system. Although the epidermis was not noticeably damaged by dissection, there was edge damage where the chamber halves sealed. The values of the TEP, SCC, and conductance are probably not optimal, since the damaged areas would allow back-fluxes of ions which would tend to increase conductance and decrease the TEP and SCC. In an attempt to minimize these errors, the chamber was designed with as large an aperture as possible since edge dam-age i s a function of perimeter, whereas conductance and SCC are functions 162 of the area of membrane exposed. It is assumed that edge damage i s not excessive since the conductance of the skins from the 5% sea water group -3 2 (0.87 +0.19 x 10 mho/cm ) compares favorably to previous studies on -3 2 (fresh water) frog skin (0.4-0.7 x 10 mho/cm ) (Ussing and Zerahn, 1951). The results indicate that there is electrogenic ion transport across Gillichthys skin. In the absence of electrochemical gradients (with Ringer on both sides of the membrane), Gillichthys skin develops a serosa-positive transepithelial potential and a short-circuit current associated with the TEP (Table 1). The SCC is commonly used to demonstrate electro-genic ion transport, as originally shown by Ussing and Zerahn (1951) on their work with frog skin. The active transport in Gillichthys skin ap-pears to be a sea water adaptation, since the SCC is significantly lower in the 5% sea water-adapted fish (Table 1). The conductance, as a rough estimate of ionic permeability, is significantly lower in the 5% sea water group (Table 1), agreeing with previous studies which demonstrated that the overall permeability of the integument is lower in euryhaline fish acclimated to hypotonic media (Motais ejt a l . , 1969; Ogawa, 1974). Further, the ion transport (as SCC) is inhibited by 2,4-dinitrophenol + Iodoaceta-mide (Table 2) indicating that ATP is required and that the SCC does re-f l e c t an active process. Though electrical measurements only show the net movement of charges, there are indications that point to transepithelial pumping of chloride. F i r s t , since the SCC is higher in the seawater-adapted f i s h , i t may be as-sumed that (like the g i l l ) ions are pumped from serosa to mucosa, as pump-ing of ions in the opposite direction would be detrimental to the fi s h . 163 Second, the TEP is serosa-positlve, indicating either a net movement of cations inward or of anions outward. Third, acetazolamide, an inhibitor of carbonic anhydrase and chloride-bicarbonate exchange, also inhibits the ion transport (as SCC) across Gillichthys skin. Finally, the TEP and SCC are reduced to near zero by the removal of chloride from both sides or only the serosal side of the membrane, and these decreases were reversible. This reduction in TEP and SCC in response to chloride-free Ringer differs from the situation with frog skin, where sodium sulphate (chloride-free) Ringer does not inhibit the pumping of sodium from mucosa to serosa (Loefoed-Johnsen and Ussing, 1958). Thus, in the skin of sea water-adap-ted Gillichthys, the serosa-positive TEP appears to reflect active trans-port of chloride from serosa to mucosa, rather than the pumping of sodium from mucosa to serosa. The Ringer solutions regularly used in the experi-ments contained appreciable amounts of bicarbonate, which could affect chloride-bicarbonate exchange. When phosphate-buffered Ringer was substi-tuted, the TEP and SCC decreased slightly suggesting that the bicarbonate in the normal Ringer may have enhanced this exchange. Both sodium and potassium appear to be involved with active transport in Gillichthys skin. The TEP and SCC are reduced to near zero by the re-moval of either of these ions from both sides or only from the serosal side of the membrane (Tables 3,4), The requirement for these ions and Na-K-ATPase is demonstrated by the inhibition of the TEP and SCC by ouabain (Table 2). Thus sodium, potassium and chloride are a l l required for the normal operation of the active pump(s) in Gillichthys skin, and the move-ments of these ions appear to be interdependent. Recently Maetz and Pic 164 (1975) demonstrated that the sodium and chloride extrusion pumps in the g i l l s of the seawater mullet are both activated by external potassium and concluded that the branchial chloride pump is associated with sodium and potassium exchanges across the g i l l . Chloride extrusion across the g i l l of the mullet appears to be electrogenic (Pic and Maetz, 1975). Further, Shuttleworth et a l . (1974) found that the in vitro g i l l of the flounder (sea water) develops a serosa-positive TEP of 6.9 mV with Ringer on both sides of the preparation. This TEP (like Gillichthys skin) is inhibited by ouabain and the authors believe that the excretion of sodium and chlor-ide are in some way linked. Whereas the SCC was rapidly reduced by exposure of the serosal side of the skin to metabolic inhibitors (Table 2), prolonged exposure of the mucosal side to inhibitors had l i t t l e effect. These results suggest a short diffusion distance between the serosal medium and the ion-transport-ing enzymes. Thus these enzymes are probably situated in the basal cells of the epidermis, and do not appear to be exposed to the mucosal side of the skin. This is consistent with the findings o f Ussing and Zerahn (1951) , who demonstrated that the enzyme responsible for sodium transport in frog skin i s located i n the basal cells of the epidermis. Exposure of the mucosal side of the skin to sodium- or chloride-free Ringer caused an increase in the TEP (Table 3). These results are d i f f i -cult to interpret since i t is unclear whether the response i s due to d i f -fusion potentials or a stimulation of the active transport. Information on the permeability of the skin to these ions using inhibitor-treated skins is necessary to analyse these results properly. 165 The importance of the skin i n osmoregulation of teleosts is suggested by the results of this study, but the extent of i t s contribution, relative to the g i l l s , i s unknown. The surface area of the skin, relative to the g i l l s , Is 0.05-0.5, according to a survey of 31 species of larger marine teleosts (Grey, 1954). Thus the skin's contribution i s probably less than that of the g i l l . However, the skin may prove to be a valuable accessory surface for ion transport i n fish. Because the skin i n vitro has a dis-tinct advantage over the g i l l in vitr o , in that the TEP and SCC can be monitored simultaneously with tracer flux studies, the skin may provide important information on teleost osmoregulation not otherwise available using the g i l l . ACKNOWLEDGMENT Dr. H.A. Bern kindly offered the use of his laboratory f a c i l i t i e s and provided the animals for this study. The author sincerely thanks Dr. Bern, Dr. R. Nishioka and Dr. S. Shen for their help during the experi-ments and in preparation of the manuscript. Thanks are also due to Dr. W.S. Hoar and Dr. D.J. Randall for their sound advice and c r i t i c a l comments on the manuscript. 166 LITERATURE CITED Bern, H.A. : Prolactin and osmoregulation. Amer. Zool. 15_, 937-948 (1975). Fromm, P.O.: Some quantitative aspects of ion regulation in teleosts. Comp. Biochem. Physiol. 27: 865-869 (1968). Grey, I.E.: Comparative study of the g i l l area of marine fishes. B i o l . Bull. 107_: 219-225 (1954). Johnson, D.W.: Endocrine control of hydromineral balance i n teleosts. Amer. Zool. 13: 799-818 (1973). Kirsch, R.: The kinetics of peripheral exchanges of water and electro-lytes i n the silver eel (Anguilla anguilla) in fresh water and i n seawater. J. exp. Bio l . 5_7 489-512 (1972). Koefoed-Johnsen, V., Ussing, H.H.: The nature of the frog skin potential. Acta physiol. scand. 42: 298-308 (1958). Maetz, J.: Mechanisms of salt and water transfer across membranes in tele-osts in relation to the aquatic environment. Mem. Soc. Endo. 18: 3-29 (1970). Maetz, J.: Fish g i l l s : mechanisms of salt transfer in fresh water and seawater. P h i l . Trans. B. 262: 209-249 (1971). Maetz, J,, Pic, P.: New evidence for a Na/K and Na/Na exchange carrier linked with the Cl pump, in the g i l l of Mugil capito in sea water. J. comp. Physiol. 102: 85-100 (1975). Motais, R. , Isaia, J., Rankin, J.C, Maetz, J. : Adaptive changes of the water permeability of the teleostean g i l l epithelium in relation to external s a l i n i t y . J. exp. Bi o l . 51_: 529-545 (1969). 167 Ogawa, M.: The effects of bovine prolactin, sea water, and environmental calcium on water influx in isolated g i l l s of the euryhaline teleost Anguilla japonica and Salmo gairdneri. Comp. Biochem. Physiol. 49A 545-554 (1974). Pic, P., Maetz, J.: Differences de potentiel trans-branchial et flux ioniques chez Mugil capito adapte a l'eau de mer. Importance de l'ion Ca 2 +. C.R. Acad. Sci. (Paris) 280: 983-985 (1975). Shuttleworth, T.J., Potts, W.T.W., Harris, J.N.: Bio-electric potentials in the g i l l s of the flounder Platichthys flesus, J. comp. Physiol. 94: 321-329 (1974). Siegel, S.: Nonparametric stati s t i c s for the behavioral sciences. New York: McGraw-Hill 1956. Ussing, H.H., Zerahn, K.: Active transport of sodium as the source of el< t r i c a l current in the short-circuited isolated frog skin. Acta physiol. scand. 23: 110-127 (1951). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0094116/manifest

Comment

Related Items