Penguin. The Relation Between Structure and Function of Bile Ducts in Man, Some Laboratory Animals and the Adelie

Transcription

1 Quarterly Journal of Experimental Physiology (1979) 64, The Relation Between Structure and Function of Bile Ducts in Man, Some Laboratory Animals and the Adelie Penguin C. J. H. ANDREWS and the late W. H. H. ANDREWS From the Department of Anaesthetics, Royal Devon and Exeter Hospital (Wonford), Barrack Road, Exeter EX2 5DW, and Department of Physiology, Royal Free Hospital School of Medicine, Pond Street, London NW3 2QG. (RECEIVED FOR PUBLICATION 2 MAY 1977) The biliary trees of man, dog, cat, rabbit, rat, guinea pig and penguin were examined in histological sections and by latex casts. The trees of man, dog and cat were similar with only minor differences. Tubulo-alveolar glands were present in all three species around large intrahepatic ducts and in large portal tracts there were zones of ductules (areas with many small bile ducts), surrounded by small vessels with no apparent relation to hepatocytes. Both these features were present in the guinea pig and tubulo-alveolar glands were present in the penguin liver. The biliary epithelium of the rat was comparatively simple but that of the rabbit appeared to be highly specialized. An estimation of the complexity of the biliary tree was obtained in the mammals by comparing the circumference of small portal venous branches with the circumference of the accompanying bile ducts, and obtaining a ratio. Man, dog and cat had fewer and smaller bile ducts than the other species. The literature on the rate of formation and composition of bile in the species studied here was reviewed and it appears that the physiology of bile secretion can be related to the morphology of the biliary tree. Rous and McMaster [1921] demonstrated that the large intrahepatic bile ducts of dogs could secrete white bile. Andrews [1955] suggested, on indirect evidence, that the bile duct was a highly active structure and was involved in both secretion and reabsorption. Subsequently, it was shown that the bile duct itself may secrete bicarbonate in the dog [Wheeler and Mancusi-Ungaro, 1966] and various inorganic ions in the rabbit [Chenderovitch, 1972]. The bile duct is technically difficult to study and workers have sought a substance which will pass into the canaliculus, giving the rate of formation of canalicular bile flow in the same way as inulin may be used to determine the rate of glomerular filtration in the kidney. Claims have been made that erythritol and mannitol will diffuse freely into the canaliculus from blood but cannot diffuse into, or out of any other part of the biliary tree [Forker, 1967; Wheeler, Ross and Bradley, 1968]. This technique of estimating canalicular bile flow is now generally Correspondence to: Dr C. J. H. Andrews, Department of Anaesthetics, Royal Devon and Exeter Hospital (Wonford), Barrack Road, Exeter EX2 5DW, Devon. 61

2 62 C. J. H. ANDREWS AND W. H. H. ANDREWS accepted, though small bile ducts appear to be permeable to erythritol, mannitol and even larger molecules [Peterson and Fujimoto, 1973]. The use of the erythritol clearance test has diverted attention away from bile ducts and it is possible that some functions ascribed to hepatocytes may be carried out by small bile ducts. In this study we attempt to relate the structure of the bile duct in various species with published data on the composition and rate of formation of bile. A preliminary report has been published [Andrews and Andrews, 1976a]. Methods Morphological. Two methods were used. Casts of the biliary tree and hepatic blood vessels were made with neoprene and examined under a stereoscopic microscope. Histological sections of livers were examined which had been fixed in neutral four per cent formol saline and embedded in either paraffin wax or ethyl methacrylate. The species examined were: man (6), dog (4), cat (3), rat (5), guinea pig (5), rabbit (6) and penguin (4). Several blocks were taken from many livers and serial sections were examined from each species. In bile ducts of all sizes, including the smallest (ductules), the relative thickness of the epithelial layer was compared to the lumen. Areas of sections were found where the portal tract had been cut at right angles to its main axis and the diameter of the main portal venous branch was loo1um or less. In such tracts the perimeter of the lumen of all portal veins and all bile ducts was measured. Measurements were made on between 45 and 89 tracts in each species except the guinea pig, where only 32 were measured. The large intrahepatic ducts were examined for the presence of tubulo-alveolar glands [McMinn and Kugler, 1961]. Results The smallest ducts (ductules) of all species were essentially similar in appearance by light microscopy, with low cuboidal cells surrounding a small lumen; nuclei were relatively large, occupying most of the cell, and dark staining. The histological appearance of the large bile ducts was, however, highly characteristic and it was usually possible to make a positive identification of the species without difficulty. Portal tracts in some species were noted to contain areas of an apparently glandular tissue, containing crypts and acini. Cells in these areas stained as did bile duct cells, and were microscopically indistinguishable from ductular. cells. These areas we have called zones of ductules. Casts of small portal veins were very similar among the different mammals studied and for this reason the portal vein was used as a standard against which the biliary tree was compared. In the guinea pig, however, some of the portal veins are very muscular and it was thought the relationship between portal and biliary trees might consequently be distorted. Parallel studies were therefore made comparing the circumference of the portal vein with that of the hepatic artery and then the hepatic artery with the bile duct.

3 z~~ s;/*sv ir XL FIG.~~~~~~~: Secto 1. hog a:lag bile duc nea th hilum Tbo-veoa glans ar rrwd Manfcto 200.:#:: ::* ; iio ;; arounth larget ducts bt abrupl ceseasvath ducssbeom smalr Maniictsioni;; e~~~~~~~~~~~~~~t fae pags7 :e 62_ s

4 FIG. 3. Rabbit. Bilary epithelium typical of large and medium bile ducts. The cells appear to be highly specialized. Magnification x 800.

5 BILIARY FUNCTION 63 There was a well-marked peribiliary plexus of blood vessels in all species, being least extensive in the rat. Man, dog and cat. The appearance of bile ducts in these species was very similar. Distal to the ductules, there is a gradual increase in the thickness of the epithelium as the lumen of the duct becomes greater. The increase in cell size is due chiefly to a proportionately larger amount of cytoplasm. The largest cells in the dog are about 30 gm high, those in the cat about 20 4um and the cells of man are intermediate. Through most of their course the bile ducts enlarge smoothly and regularly. The largest intrahepatic ducts, especially of man and cat (Fig. 1) contain tubulo-alveolar glands, similar to those described by McMinn and Kugler [1961] in extrahepatic and pancreatic ducts. The blood supply to these glands is profuse. One new feature was seen in large portal tracts near the hilum in man, dog, cat and guinea pig: large numbers of ductules and glands were present which apparently originated in circumscribed areas and which, while not-part of the overall tributary system of bile ducts, were associated with it. Many small vessels were present between these small ducts and the general impression was that of a secretory tissue. The ratio of the internal circumferences of portal venous branches to that of accompanying bile ducts was: man, 5.1 to 1; dog, 5.3 to 1 and cat, 3.4 to 1. Rat. The smallest ducts resembled those of other species and the largest ducts were lined with cuboidal cells. Neither tubulo-alveolar glands nor ductular zones were present. The ratio of the circumferences of small portal veins to bile ducts was 1 8 to 1, and the number of small ducts accompanying portal venous branches was greater than in the species described above. Guinea pig. The number of small bile ducts accompanying portal venous branches was much greater than in man, cat or dog. The ratio of portal vein to bile duct circumference was 0 9 to 1. Goblet cells were present in ducts near the hilum. Tubuloalveolar glands were present, but not profuse, around the largest intrahepatic bile ducts. Zones of ductules, with accompanying blood vessels, were present in portal tracts near the hilum. Rabbit. The large bile ducts were lined with a high columnar epithelium with large, elongated nuclei (Fig. 3). Inclusion bodies were present and in places the epithelium was several layers thick. Folds and crypts were present in large ducts, but tubulo-alveolar glands and zones of ductules were not seen. The portal vein: bile duct circumference ratio was 2-1 to 1. Penguin. Large ducts were studied, and their epithelium,was columnar. Tubuloalveolar glands are particularly well marked (Fig. 2), the point at which they begin is sharply defined and is apparently related to the diameter of the duct. The portal vein tapers more rapidly in the penguin [Andrews and Andrews, 1976b], than in the mammals investigated and as a consequence the portal vein to bile duct ratio is not constant and is not given in Table II. Discussion It would appear from our observations that the morphology of the biliary system is correlated with the rate of formation and the composition of bile. The

6 64 C. J. H. ANDREWS AND W. H. H. ANDREWS biliary systems of man, dog and cat [Group 1 of Andrews and Andrews, 1976a] are similar in structure and function: resting secretion is low, bile salt concentration is high, whilst bicarbonate concentration is low (Table I and II). The bile flow of the other mammals (Group 2) is much higher with a lower bile salt and a higher bicarbonate concentration, excepting in the rat [Klaassen 1971]. There is some similarity in the morphology of the large bile ducts in the guinea pig and group 1 mammals, but small ducts (ductules) are far more profuse in the guinea pig. Ductules are about equally profuse in rabbit and rat, but morphologically the medium and large ducts of the rabbit appear to be more active than the equivalent ducts of the rat, and the rate of secretion of bile is slightly higher in rabbits [Pugh and Stone 1968] than rats [Klaassen 1971]. Table 1. Comparison of bile electrolytes in various species. The sources of information are: 1. Waitman et al. [1969]. 2. Pugh and Stone [1968]. 3. Kaminski et al. [1975]. 4. Reingold and Wilson [1934]. 5. Klaassen [1971]. 6. Rutishauser and Stone [1975]. 7. Scratcherd [1965]. 8. Chenderovitch and Phocas [1965]. 9. Andrews [1978]. P=pentobarbitone; U=urethane; b.a.=bile acids. *Four observations only. bile electrolytes: mmol.l- 1 Species No. subjects Na+ K+ Cl- HCO- b.a. anaesthetic source man None 1 dog P none none 4 cat P 2 rat P 5 rabbit P U U 2 guinea pig * U U 8 penguin P 9 Table II. Comparison of rates of flow of 'resting' bile with the morphology of the bile duct. The sources of data bear the same number as in Table I. perimeter of bile flow portal vein/ zones of tubulo-alveolar response to Species ul.g-. minm bile duct ductules glands secretin source man dog , 3, 4 cat rat rabbit , 6, 7 guinea pig

7 BILIARY FUNCTION 65 In general, the more complex the system of small ducts and ductules, the greater is the rate of flow, and therefore it would seem that the role of the small ducts is primarily that of secretion, especially as the morphology of the bile canaliculi is relatively constant in different species [Steiner and Carruthers, 1961]. The hypothesis that small ducts can secrete has experimental support. Goldfarb, Singer and Popper [1963] found that with hyperplasia of small ducts, such as that induced by chronic administration of alpha naphthyl isocyanate there is a faster flow of 'resting' bile. The chief evidence against the hypothesis that small ducts can secrete lies in the interpretation of erythritol clearance (E.C.), which is believed to be identical with canalicular flow. When canalicular flow is increased by administration of bile salts, the E.C. rises; whereas when there is an increased flow from the duct, induced by secretin, the E.C. does not increase [Forker, 1967]. Secretin acts only in animals in which tubulo-alveolar glands, similar to those of the pancreas, are present, and there is no evidence to support the idea that the physiological role of large and small bile ducts is the same. Other substances besides erythritol pass from plasma into bile, including plasma albumin, tagged with radio-active iodine [Rosenthal, Kubo, Dolenski, Marino, Mersheimer and Glass, 1965]; sucrose and inulin [Forker, 1970]; and insulin [Daniel and Henderson, 1967]. It is unlikely that these compounds are secreted by the hepatocyte. Sternleib [1972] has described a fenestrated endothelium in capillaries around small bile ducts and has suggested that substances can diffuse into bile through this endothelium. It seems possible that erythritol equilibrates at this site. If such is the case then E.C. is related to ductular and not canalicular flow. All published data fit the concept of the small bile ducts being relatively permeable to erythritol. The hypothesis that small bile ducts may be able to secrete bile is supported by data as presented in Tables I and II. In man, dog and cat, where the distal part of the biliary tract is simple (few ducts and hence a high portal vein/bile duct ratio), the resting bile flow is low and the bile salt concentration is high. This would appear to indicate that a large proportion of resting flow is canalicular in origin, in these species. Conversely, rat, rabbit and guinea pig have a greater number of small ducts (a smaller portal vein/bile duct ratio), a higher resting flow rate and a low bile salt concentration. We propose that a high proportion of bile secreted by these latter species is produced by the small bile ducts. Bicarbonate is known to be excreted by the larger bile ducts [Wheeler and Mancusi-Ungaro 1966]. However, part of bicarbonate excretion is dependent on bile salt excretion [Strasberg, Ilson, Siminovitch, Brenner and Palaheimo, 1975], and this fraction seems unlikely to be ductular. With the exception of the rat, species with high basal flow rates have high bicarbonate concentrations and those with low bile flow rates have low bicarbonate concentrations. Secretin has been shown to stimulate a bicarbonate rich secretion in all those species in which we observed tubulo-alveolar glands and zones of ductules; in man [Razin, Feldman and Dreiling, 1965], in dog [Wheeler and Mancusi-Ungaro, 1966], in cat [Scratcherd, 1965], and guinea pig [Rutishauser, personal com-

8 66 C. J. H. ANDREWS AND W. H. H. ANDREWS munication]. Secretin has no effect in the rabbit [Scratcherd, 1965] and although an increase has been reported in the rat [Forker, Hicklin and Sornson, 1967], Rutishauser [1976] ascribed the choleresis to impurities in the secretin. It would appear therefore that species with tubulo-alveolar glands respond to pure secretin, the association may be fortuitous, but similar glands in the pancreatic duct are generally believed to secrete bicarbonate. A study which is primarily morphological cannot, by itself, challenge modern concepts of biliary secretion, but it does support strongly workers who do not agree that the active functions of the bile duct are limited to absorption of fluid and the secretion of bicarbonate in response to secretin. Acknowledgments Thanks are due to Dr. D. Gall for supplying some of the histological material, and to Miss Goddard for technical assistance. Dr. Rutishauser has kindly allowed us to cite her unpublished experiments and has helped with constructive cr-iticism. References ANDREWS, C. J. H. (1978). Spontaneous and bile salt stimulated bile secretion in the Adelie penguin (Pygoscelis adeliae). Quarterly Journal of Experimental Physiology, 63, ANDREWS, C. J. H. and ANDREWS, W. H. H. (1976a). The site of tormation of 'bile salt-independent' bile. Journal of Physiology, 258, 11-12P. ANDREWS, C. J. H. and ANDREWS, W. H. H. (1976b). Casts of hepatic blood vessels: a comparison of the microcirculation of the penguin Pygoscelis adeliae, with some common laboratory animals. Journal of'anatomy, 122, ANDREWS, W. H. H. (1955). Excretory function of the liver; a re-assessment. Lancet, 1, CHENDEROVITCH, J. (1972). Secretory function of the rabbit common bile duct. American Journal of Physiology, 223, CHENDEROVITCH, J. and PHOCAs, E. (1965). The influence of intravenous hypertonic solutions on bile formation. In: The Biliary System. Ed. W Taylor. Blackwell, OxJord. pp DANIEL, P. M. and HENDERSON, J. R. (1967). Insulin in bile and other fluids. Lancet, 1, FORKER, E. L. (1967). Two sites of bile formation as determined by mannitol and erythritol clearances. Journal of Clinical Investigation, 46, FORKER, E. L. (1970). Hepatocellular uptake of inulin, sucrose and mannitol in rats. American Journal of Physiology, 219, FORKER, E. L., HICKLIN, T. and SORNSON, H. (1967). The clearance of mannitol and erythritol in rat bile. Proceedings of the Society for Experimental Biology and Medicine, 126, GOLDFARB, S., SINGER, E. J. and POPPER, H. (1963). Biliary ductules and bile secretion. Journal of Laboratory antd Clinical Medicine, 62, KAMINSKI, D. L., RUWART, M. and WILLMAN, V. L. (1975). The effect of prostaglandin A, and El on canine hepatic bile flow. Journal of Surgical Research, 18, KLAASSEN, C. D. (1971). Does bile acid secretion determine canalicular bile production in rats? American Journal of Physiology, 220, MCMINN, R. M. H. and KUGLER, J. H. (1961). The glands of the bile and pancreatic ducts; autoradiographic and histochemical studies. Journal of Anatomy, 95, PETERSON, R. E. and FUJIMOTO, J. M. (1973). Retrograde biliary injection; absorption of water and other compounds from the rat biliary tree. Journal of Pharmacology and Experimental Therapeutics, 185, PUGH, P. H. and STONE, S. L. (1968). The ionic composition of bile. Journal of'physiology, 201, 50-51P. RAZIN, E., FELDMAN, M. G. and DREILING, D. A. (1965). Studies on biliary flow and composition in man and dog. Journal oj'the Mount Sinai Hospital, 32, REINGOLD, J. G. and WILSON, D. W. (1934). The acid-base composition of human bile. 1. American Journal of Physiology, 107,

9 BILIARY FUNCTION 67 ROSENTHAL, W. S., KUBO, K., DOLINSKI, M., MARINO, J., MERSHEIMER, W. L. and GLASS, G. B. J. (1965). The passage of serum albumin into bile in man. American Journal of Digestive Diseases, 10, Rous, P. and MCMASTER, P. D. (1921). Physiological causes for the varied character of stasis bile. Journal of Experimental Medicine, 34, RUTISHAUSER, S. C. B. (1976). An analysis of the reported choleretic effect of secretin in the rat. Journal of Physiology, 257, 59P. RUTISHAUSER, S. C. B. and STONE, S. L. (1975). Aspects of bile secretion in the rabbit. Journal of Physiology, 245, SCRATCHERD, T. (1965). Electrolyte composition and control of biliary secretion in the cat and rabbit. In: The Biliary System. Ed. W Taylor. Blackwell, Oxford. Pp. 522 et seq. STEINER, J. W. and CARRUTHERS, J. S. (1961). Studies on the fine structure of the terminal branches of the biliary tree. 1. The morphology of normal bile canaliculi, bile pre-ductules (ducts of Hering) and bile ductules. American Journal of Pathology, 38, STERNLEIB, L. (1972). Special article: functional implications of human portal and bile ductule ultrastructure. Gastroenterology, 63, STRASBERG, S. M., ILSON, R. G., SIMINOVITCH, K. A., BRENNER, D. and PALAHEIMO, J. E. (1975). Analysis of the components of bile flow in the rhesus monkey. American Journal of Physiology, 228, WAITMAN, A. M., DYCK, W. P. and JANOWITz, H. D. (1969). Effect of secretin and acetazolamide on the volume and electrolyte composition of hepatic bile in man. Gastroenterology, 56, WHEELER, H. 0. and MANCUSI-UNGARO, P. L. (1966). Role of bile ducts during secretin choleresis in dogs. American Journal of Physiology, 210, WHEELER, H. O., Ross, E. D. and BRADLEY, S. E. (1968). Canalicular bile production in dogs. American Journal of Physiology, 214,