Structure and Function of the Pituitary Gland in Gonadal Tumor-Bearing and Normal Cyprinid Fish1

[CANCER RESEARCH 37, 3151-3168, September 1977]

Structure and Function of the Pituitary Gland in Gonadal Tumor-bearing and Normal Cyprinid Fish1

John F. Leatherland and R. A. Sonstegard2

Departments of Zoology [J. F. L.] and Microbiology Â¡R.A. S.Â¡,College of Biological Science, University of Guelph, Guelph, Ontario, Canada

SUMMARY Of particular concern with waterborne carcinogens are the reports of Harris and coworkers (12, 31), which present The F, generation hybrids of carp and goldfish exhibited a epidemiological evidence suggesting a significant relation marked hyperplasia of pars distalis basophil (gonado- ship between cancer mortality in white males and drinking trophic) cells, which first became evident in 3- to 4-year-old water that was obtained from the Mississippi River. Al fish, the age at which carp and goldfish normally become though the results of the New Orleans study cannot be sexually mature. Many of these hybrids also developed gon- considered conclusive evidence that cancer in man is adal tumors, the genesis of which also coincided with the caused by the consumption of contaminated water, these onset of gonadotroph proliferation. In these fish the baso findings must be taken into consideration. phil hyperplasia was more marked than in nontumorous hy Recently, the technology for the quantitation and qualita- brids. Tumor-bearing carp also exhibited gonadotroph hy tion of dilute organic and metallic agents in water has been perplasia similar to that in tumorous hybrids. Both tumor- rapidly improved (34). However, the problems are complex, bearing and nontumorous hybrids showed a marked reduc and the need for surveillance is acute. It is increasingly ap tion in gonadotroph cell number when fed a methallibure parent that investigations must be addressed to waterborne diet, although there was no apparent change in either the factor(s), the existence and effects of which on animal gross pathology or the histopathology of the tumors. These health are unknown. A sentinel system for the early detec findings suggest that the hyperplasia of the gonadotrophs tion and identification of waterborne environmental carcin in tumorous fish is not, ;'pso facto, responsible for the ogens is urgently needed. As an approach to this situation, maintenance of the tumors. On the contrary, since gonado we have made a survey monitoring the frequency of occur troph hyperplasia was evident in both tumorous and nontu rence of fish tumors in polluted and nonpolluted waters in morous hybrids (all of which were sterile), it may be sympto the Great Lakes to explore the potential utility of neoplasia matic of the sterile condition of the fish rather than of the in feral fish as an indicator of environmental carcinogens tumor per se. Evidence for the presence of a second gona- (36-38, 40). Epizootics of gonadal tumors in carp (Cyprinus dotrophic cell type in cyprinids is presented and discussed. carpio), goldfish (Carassius auratus), and goldfish-carp hy Other pituitary cell types were similar in structure and ultra- brids were found, and they appear to have environmental structure in carp, goldfish, and hybrids and were unaffected relevance (36-38, 40). The tumors appear to be of a Sertoli by methallibure treatment. cell origin, although some may also contain Leydig cell elements (40). The tumor incidence is as high as 100% in some age-sex groups (40). INTRODUCTION One of the characteristics of the gonadal tumor-bearing fish and also of nontumor-bearing (sterile) hybrids is a Our aquatic environments are known to be fouled with massive proliferation of pituitary basophils (40). In this pa chemicals and classes of chemicals that have carcinogenic per we report investigations undertaken to elucidate the potential (6, 24, 25, 35, 41). This area has been excellently pituitary-gonadal feedback axis in these fish. In addition, reviewed by Kraybill (18). Although there is little evidence the effect of methallibure on the number and apparent that waterborne carcinogens have produced widespread activity of the gonadotrophs was examined. In teleosts cancer problems in man, it is not difficult to envisage such a methallibure appears to act as a gonadal inhibitor by block possibility. At present, because of ground water depletion, ing gonadotropin synthesis or release (13, 19, 32), possibly many cities are processing for drinking purposes water that by inhibiting the synthesis or release of hypothalamic re may be in excess of 30% recycled (9). Many inorganic and leasing factors (26). Methallibure was given to tumor-bear organic compounds (some known carcinogens) are not re ing hybrids to determine whether the proliferation of gonad moved in current water treatment processes (5, 30). In fact, otrophs evident in these fish can be regressed in the same treatment may be producing carcinogens; i.e., chlorination way that it is in other methallibure-treated teleosts. It was may produce chloroform and carbon tetrachloride, both hoped that these experiments might provide an insight into known carcinogens (3, 15, 33). the cause(s) of the gonadotroph hypertrophy. 1 This work was supported by grants from the National Cancer Institute of Canada, Environment Canada, and the National Research Council of Can MATERIALS AND METHODS ada. 2 Research Scholar, National Cancer Institute of Canada. Received February 28, 1977; accepted June 7, 1977. Collection and Maintenance of Fish. One hundred and

SEPTEMBER 1977 3151

two fish were used in the study of pituitary histology, includ fluid, transferred to 70% alcohol for 48 hr, and weighed. The ing 32 carp (19 males and 13 females), 63 hybrids (38 males, data were compared by 1-way analysis of variance. Individ 23 females, and 2 hermaphrodites), and 7 goldfish (4 males, ual means were compared by least significant difference. 2 females, a 1 hermaphrodite). A further 41 fish (19 males, 19 females, and 3 of indeterminate sex) were used to deter mine the effect of methallibure on pituitary histology, and RESULTS 39 5- to 6-year-old fish were used for measurements of pituitary weight. Fish were collected by gill nets from Hamil Adult (4 Years Old) Nontumorous Carp and Goldfish. ton Harbour, Lake Ontario. Some were biopsied as they The structure of the pituitary gland was essentially similar in were removed from the nets, some were returned to Guelph nontumorous carp and goldfish (Figs. 1 and 4). The organi for biopsy, and others (to be treated with methallibure) were zation and cellular characteristics of the cell types closely returned to Guelph and acclimated for 4 to 10 weeks in 150- resembled the descriptions of goldfish pituitary given by gallon aquaria of running well water. The age of each fish other authors (2, 7, 16, 20, 21, 23, 28, 35). The rostral pars was determined at biopsy by scale ring counts. distalis (Figs. 1, 2, and 11) formed a thin anterior cap, which Methallibuie Treatment. Methallibure was administered extended dorsally from the dorsoanterior region of the pi to hybrids in 2 separate experiments. In the 1st experiment tuitary stalk and ventrally to meet the most anterior regions hybrids that displayed a distinct tumorous condition, evi of the pars intermedia. The rostral pars distalis region was denced by a distended abdomen, were chosen. They were composed of 4 cell types, an acidophil, an amphiphil (possi bly a chromophobe), and 2 basophils. The acidophils [pro- maintained in the laboratory in continuously running and aerated well water. They were fed daily with a pelleted fish lactin cells (11)] were spherical or elongated cells with diet to which had been added methallibure (I.C.I. 33,838). A spherical nuclei and prominent nucleoli. The cytoplasm was known amount of methallibure-treated diet was given so poorly granulated; the granules stained deep red with ery- throsin but only pale orange with OG (Fig. 2). The amphi- that, assuming equal food consumption, each fish received 1 mg of methallibure per kg of body weight each day. In the philic cell type [probably ACTH (16, 29)] was interspersed 1st experiment a group of 12 fish was administered the diet with the acidophils and. together with the acidophils, com for 6 weeks, and the pituitaries were fixed for light micros posed the major component of this region of the gland. The copy. Pituitaries from 5 control fish, which had been fed a amphiphils were conversely elongated cells with oval or diet identical with that of the methallibure-treated fish ex spherical nuclei but with no prominent nucleoli. Their cyto cept that methallibure had been omitted, were also fixed for plasm contained little stainable material and appeared light microscopy. chromophobic with the stains used here. In the 2nd experiment sterile and tumor-bearing hybrids In electron micrographs the prolactin (Figs. 14 and 16) were used. A group of 5 fish was randomly selected as the and ACTH cells (Figs. 15 and 16) were similar in appear ance. Both contained numerous membrane-bound elec initial control group, and the remainder of the fish were tron-dense cytoplasmic granules, and both showed signs of administered the methallibure diet. Random selections of 3 to 4 fish were made 3, 5, 10, 15, 20, and 35 days after first moderate secretory activity, as evidenced by their content giving the methallibure diet. The pituitary from each fish of mitochondria, endoplasmic reticulum, and Golgi bodies. was preserved for light microscopy. The cells differed in the diameter of the cytoplasmic gran Histology. Each fish was killed by transection of the ules, those in the prolactin cells being larger than those in spinal cord, and the pituitary was rapidly dissected from the the ACTH cells (Figs. 14 to 16) (16, 20). Interspersed among skull, cut into halves in the sagittal plane, and placed into the prolactin and ACTH cells were nongranulated cells with Bouin's Hollande sublimate. The tissues were dehydrated in moderately electron-dense cytoplasm (Fig. 17). alcohol, embedded in Paraplast. and sectioned at 5 /Â¿m.The The 2 basophil cell types were present in small numbers. One type (Fig. 2), the presumptive thyroid-stimulating hor sections were stained with 1 of the following procedures: AB-PAS-OG,3 Herlant's tetrachrome, Cleveland-Wolf tri- mone cells (type 2 basophils) (16, 20), were small angular chrome, Mallory's trichrome, or GÃ¤be'saldehyde fuchsin. Pieces of partes intermedia and distalis were taken from 5 nontumorous carp, 5 nontumorous goldfish, 5 tumorous hybrids, and 2 tumorous carp for examination by means of an electron microscope. The pieces were fixed overnight in ice-cold 5% cacodylate-buffered (pH 7.4) glutaraldehyde. They were then postfixed in cacodylate-buffered (pH 7.4) 1% osmium tetroxide, dehydrated in acetone, and embed ded in Epon. Sections were mounted on uncoated copper grids, stained with uranyl acetate and lead citrate, and examined on a Philips 200 electron microscope. Pituitary Weights. Pituitary glands were dissected from CD rostral pars distalis 39 freshly killed carp and hybrids, fixed for 48 hr in Bouin's O proximal pars distalis O pars intermedia Hi pars nervosa 3The abbreviations used are: AB. Alcian blue; PAS. periodic acid-Schiff; Fig. 1. Diagram of a sagittal section through the pituitary gland of an adult OG, orange G; ACTH, adrenocorticotrophic hormone; GTH, gonadotrophic nontumorous cyprinid. showing the arrangement of rostral and proximal hormone. pars distalis. pars intermedia, and pars nervosa.

3152 CANCER RESEARCH VOL. 37

largest component of the pars distalis. The region was composed of 4 cell types, an acidophil and 3 basophils. The acidophils of the proximal pars distalis (Fig. 6), probably somatotrophs (16, 20), were also found as small groups of cells. They were large cells with irregularly shaped cyto plasm. The cytoplasmic granules were only weakly erythro- sinophilic but tended to be more orangeophilic than the prolactin cells. The nuclei were round or oval, with moder ately prominent nucleoli. In electron micrographs the soma totrophs (Fig. 20) were seen to contain numerous rod- shaped, membrane-bound, electron-dense cytoplasmic granules, together with moderate numbers of mitochondria and Golgi bodies. Certain regions of cytoplasm were com posed almost wholly of whorls of endoplasmic reticulum (Fig. 19). Basophil cell types 1 and 3 (Figs. 6 and 7) were found in the form of lobes, with the type 3 cells in the center of a type 1 cell group. Type 1 cells were large spherical or oval cells, which in many fish contained a prominent, large vacuole occupying approximately 50% of the cell cytoplasm. The remainder of the cytoplasm contained coarse granules, which stained intensely with AB, aniline blue, and alizarin blue. There was a marked variability in the vacuolation of these cells in different fish, which could not be correlated with age, sex, or season of capture. The type 1 basophils are very similar in appearance to the cells in young goldfish that were previously termed gonadotrophic (16, 20, 22, 27). In electron micrographs these cells contained a number of types of cytoplasmic inclusions (Figs. 21 to 25). The most numerous were spherical electron-dense, membrane- bound granules (diameter, 3000 to 5000 A), which were scattered throughout the cytoplasm. Some of these gran ules were less electron dense than others and may be un dergoing structural changes (Fig. 24). The cells also con tained larger (diameter, 1 to 2 /urn) membrane-bound spheri cal droplets of varying electron density, from very electron dense to virtually electron translucent (Fig. 22). The "paler" droplets commonly contained fibrillar inclusions (Figs. 22 and 23). The presence of droplets containing an electron- dense core with peripheral fibrillar organelles suggests that 1 form of the droplet is derived from the other. In addition to the granules and droplets, these cells also contained cyto plasmic vacuoles, which formed the large organelles evi dent in light micrographs; they contained an amorphous, granular matrix (Figs. 21, 22, 24, and 25). Some of these Fig. 2. a, part of the rostral pars distalis Â¡nahybrid, showing a group of type 1 basophils had a cytoplasmic matrix that was more type 2 basophils (thyrotrophs) (arrows) among the pale prolactin and ACTH cells; these 2 cell types cannot be distinguished in the preparation. Paraffin electron dense than other cells; these may represent degen section, AB-PAS-OG, x 270. b, cyst, probably granuloma reaction to para erating cells (Fig. 25). sitic or bacterial infection, in the proximal pars distalis of a hybrid. Paraffin section, Cleveland-Wolfe, x 270. Type 3 basophils were smaller cells than the type 1 cells, usually grouped together in small clusters (Figs. 6 and 7) or, less commonly, scattered as single cells throughout the cells, mostly situated toward the periphery of the gland or in proximal pars distalis. Their cytoplasm was more finely the median part of the rostral pars distalis. The heavily granulated and stained less intensely than that of type 1 granulated cytoplasm stained moderately with AB, aniline cells; the cytoplasm commonly appeared flocculated due to blue, and alizarin blue. In electron micrographs (Fig. 18), its content of small vacuoles. In some fish the vacuoles the cells were identical in appearance with the thyrotrophs contained intensely PAS- or aniline blue-stainable rod- described in goldfish previously (16, 20). The 2nd basophil shaped bodies, but these were not present in all fish. In type found in the rostral pars distalis was identical with the electron micrographs these cells were similar in general type 1 basophil cells of the proximal pars distalis (see be appearance to the type 1 basophils but differed in a number low). of ways. Firstly, the electron-dense granules were smaller The proximal pars distalis (Figs. 1 and 6) formed the (diameter, 1500 to 2000 A, compared with 3000 to 5000 A in

SEPTEMBER 1977 3153

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1977 American Association for Cancer Research. J. F. Leather/and and R. A. Sonstegard the type 1 cells) and tended to be grouped together into Immature (4 Years Old) Nontumorous Carp and Gold clumps of granules rather than scattered, as they were in fish (Figs. 11 to 13). Only preparations preserved for light the type 1 cells (Fig. 24). Secondly, the droplets were fewer microscopy were studied. The arrangement and types of in number in the type 3 cells. Thirdly, the type 3 cells did not cells in the rostral pars distalis and pars intermedia were have the large vacuoles evident in the type 1 components; in essentially similar to those in the older fish, the exception the former the vacuoles took the form of scattered, smaller, being the PAS-negative cells of the pars intermedia, which multiple vacuoles (Fig. 24). The rod-shaped organelles were small cells and contained AB-stainable cytoplasmic present in light micrographs were not identified in electron granules in the young fish (Fig. 12), but which appeared micrographs. In some fish several of the globules (and pos chromophobic in the older animals (Fig. 3a). The bulk of the sibly the granules also) had fused to form large electron- proximal pars distalis (Fig. 13) was composed principally of dense masses (Fig. 26), which may account for the light 3 cell types, the somatotrophic acidophils and 2 basophils, microscope observations. the thyrotrophs and type 1 gonadotrophs (see above). A The 3rd basophil cell type was similar in every way to the small number of type 3 gonadotrophs were present in 3- putative thyroid-stimulating hormone cells also found in the and 4-year-old fish; in some 3- and 4-year-old fish, both rostral pars distalis. In electron micrographs a further cell type 1 and type 3 basophils were present in large numbers. type, a nongranulated cell, was evident in all regions of the The somatotrophs appeared moderately more orangeo- proximal pars distalis (Fig. 25). Cysts of a granuloma type philic in the young fish than in the older animals but were (Fig. 20) were found in the proximal pars distalis region of otherwise similar in appearance. The population of type 1 some of the older carp. basophils was considerably smaller than in the older fish. The pars intermedia appear to be composed of 2 cell The large vacuoles, which were characteristic of type 1 types (Figs. 1 and 3a). One was chromophobic with the basophils in larger animals, were only rarely found in the staining methods used here. The cells varied considerably younger specimens. in size and apparent activity between animals. In some fish Nontumorous Hybrids. In hybrid fish between 1 and 4 they were columnar cells with round or oval nuclei (presum years old that showed no evidence of gonadal tumors, the ably in an active phase), whereas in others they were small pituitary and gonad were identical in morphology with those with irregularly shaped nuclei (presumably the inactive in carp and goldfish of similar ages [see above; also see phase). The cytoplasm of these cells contained membrane- Leatherland (20) for literature review]. dense granules of widely different electron densities, to In nontumorous hybrids older than 4 years (these fish gether with marked Golgi areas and large formations of were sterile and the gonads were small), there was a hyper- endoplasmic reticulum (Figs. 27 and 28). A smaller number plasia of the basophil population of the proximal pars dis of feebly PAS-stainable cells (Fig. 3a) was also found in talis that was similar to, but less extensive than, that found regions immediately adjacent to the neurohypophysis. In in tumor-bearing hybrids (see below). The pituitary weight these cells the cytoplasmic granules were more electron of nontumorous hybrids was greater than that of nontumor dense, and the endoplasmic reticulum was scattered ous carp of a similar age, but only when the nontumorous throughout the cytoplasm (Fig. 28). Colloid-filled vacuoles males were considered separately (mean, 24.8 Â±3.3 mg) were evident in the pars intermedia of some older (10 + was the difference statistically significant (p < 0.01) (Table years) carp (Fig. 3b). The basophilic colloid and the ar 1). rangement of the PAS-negative cells around the colloid Tumor-bearing Hybrids, Goldfish, and Carp. The pitui were similar to that found in a number of mammals [e.g., tary in tumor-bearing fish was commonly 2 or 3 times larger humans (14) and seals (22)]. than that in nontumorous goldfish and carp (compare Fig. 5 The neurohypophysis (Figs. 1 and 3a) interpenetrated with Fig. 4). Pituitary weights in tumorous carp and hybrids with all regions of the adenohypophysis. Small bundles of were significantly greater than in nontumorous carp (p < axons extended deeply into the rostral and proximal pars 0.01 ) (Table 1). The increase in volume was so marked that, distalis (Figs. 1, 2a, 11, and 15). This anterior region of the on dissection, the gland was forced through the dorsal neurohypophysis contained few AB- and aldehyde fuchsin- membranous covering of the sella turcica and lacked the stainable axons. The posterior region of the neurohypophy firm texture of the adenohypophysis of nontumorous fish. sis was extensively interdigitated with the pars intermedia. Thus, dissection of complete and intact glands was some The largest accumulation of AB- and aldehyde fuchsin- times difficult. The rostral pars distalis, pars intermedia, stainable material was found in regions adjacent to the and pars nervosa were similar in size and cell complement blood vessels. The neurohypophysis and adenohypophysis to those in nontumorous carp and goldfish of comparable were separated by a PAS-stained "basement membrane" Table 1 similar in structure and fine structure (Figs. 15 and 16) to Pituitary weights in 5- to 6-year-old carp and hybrids that described previously in goldfish (20). Pituitary wt (mg)Â° A complex arrangement of capillaries was evident in the Group Number area of the pituitary stalk. Numerous blood vessels were Nontumorous carp 8 11.6 Â±3.4Â» also seen among the neurohypophysial tissue. Remarkably Tumorous carp 6 28.1 Â±3.1 Nontumorous carp 6 19.4 Â±3.0 few blood vessels were evident among the adenohypo Tumorous hybrids 19 27.8 Â±2.1 physis itself, which argues very strongly that, at least in fish, 0 The gland was dissected from each fish and fixed in Bouin's the basement membrane complexes, rather than blood cap fluid for 48 hr, transferred to 70% alcohol for 48 hr, blotted dry, illaries or sinusoids, are probably the 1st receptors of some and weighed. of the released hormone material (20). 6 Mean Â±S.E.

3154 CANCER RESEARCH VOL. 37

body weight. However, in tumorous fish, the rostral pars carp-goldfish hybrids is concomitant with the normal onset distalis cells formed a narrow zone of cells bordering the of maturity in carp and goldfish. This also coincides with the periphery of the proximal pars distalis (Fig. 5). The struc age at which the tumor frequency begins to increase mark tural and ultrastructural characteristics of the various cell edly, which suggests that the gonadal tumors may be in types were indistinguishable from those in nontumorous duced by hypersÃ©crÃ©tionof the hypothalamus-pituitary axis. carp and goldfish. The proximal pars distalis was markedly However, since the proliferation of gonadotrophs does not increased in volume, apparently due to an increase in the result in an elevation of plasma GTH (see above), the effect number of type 1 and type 3 basophils (Figs. 5 and 8). There of hypothalamic stimulation in the hybrids may be to induce appeared to be no difference in the cytological or ultra- the differentiation of gonadotroph cells rather than of GTH structural characteristics of the somatotrophs and thyro- release. Conversely, if the hypothalamic factor stimulates trophs, nor of the type 1 or type 3 basophils, but both both gonadotroph cell differentiation and GTH release in basophil cell types were vastly increased in number. carp and goldfish, the gonadotrophs in hybrids may be Whereas, in the nontumorous carp and goldfish, the type 1 unable to release GTH. This may well give rise to a contin cells surrounded the type 3 cells, in the tumorous fish there ued production of the hypothalamic factor(s), since there is were areas of the proximal pars distalis that contained only no negative feedback inhibition. either type 1 or type 3 basophilic cells. In addition central Methallibure caused a marked reduction of pituitary ba areas of the pituitaries in tumorous fish were necrotic. sophils in hybrid fish. The drug inhibits gonadal develop Methallibure-treated Hybrids. In the hybrids in both Ex ment in cyprinids (13), probably by inhibiting gonadotro- periments 1 and 2, fed a methallibure diet for periods phin secretion at the pituitary (19) or hypothalamic level greater than 25 days, type 1 and 3 basophils were both (26). That methallibure caused such a marked regression in greatly reduced in number (Fig. 9). In light and electron the gonadotrophs without apparently affecting the histopa- micrographs, the basophil type 1 and 3 cells were similar in thology of the tumor suggests that the onset of tumor prolif general appearance in methallibure- and nonmethallibure- eration, although coincident with the onset of gonadotroph treated fish, except for the extent of vacuolation of the type proliferation, is not brought about by the increased gonado 1 basophils, which was less in the former group. At the end troph activity. Further evidence for this comes from the of Experiment 2,1 fish was found to be a carp; the reduction nontumorous hybrids. The latter have a gonadotroph popu in number of basophils was more marked in this fish (Fig. lation similar to that of the tumorous fish but are, in fact, 10) than it was in the hybrids. Other pituitary cell types did sterile. It is probable that the gonadotroph proliferation in not seem to differ, either in number or appearance, from both tumor-bearing and nontumorous hybrids is a response comparable cells in untreated animals. The pituitary gland to the sterile condition of the fish rather than to the tumor, in treated hybrids was generally smaller and firmer to the per se. touch than in untreated animals. In hybrids given the meth The structure of the rostral pars distalis, pars intermedia, allibure diet for periods less than 25 days, the response of and pars nervosa in carp, goldfish, and hybrids was essen type 1 and 3 basophils was variable. Some fish demon tially similar to the descriptions given for goldfish previously strated a reduction in cell number, whereas others did not. (see "Results" for references). There was a marked varia In Experiment 1, in which only markedly tumorous fish were tion in the appearance of the pars intermedia cells between used, there was no evidence of gross or histopathological groups, which could not be correlated with species or tu changes in the gonads of methallibure-treated fish com mors but which may well be due to the variation in pigmen pared with control hybrids. tation of the specimens used, since this was found to affect pituitary cytology in goldfish (21). The significance of the basophilic pars intermedia (PAS-negative) cells in young DISCUSSION hybrids is not known. The formation of follicles of pars intermedia cells sur This study shows a marked hypertrophy of proximal pars rounding a colloid-filled lumen, which is found in older distalis basophils in tumor-bearing cyprinids and also in cyprinids, has not, to our knowledge, been reported previ nontumorous goldfish hybrids, resulting in a marked in ously in fish, although these structures are commonly found crease in pituitary weight and volume. Since, in teleosts, in mammals [e.g., humans (14) and seals (22)]. Their signifi gonadotrophic cells appear to be under stimulatory hypo- cance and function is not known. thalamic control (1), the proliferation of the gonadotrophs In this study of carp, goldfish, and their hybrids, as in the in tumorous fish is probably due to continued hypothalamic earlier study of the pituitary in goldfish (20), 3 basophil cell stimulation as a result of the low sex steroid levels in these types were found in the proximal pars distalis. One cell type fish (J. Sumpter, J. F. Leatherland, and R. A. Sonstegard, probably represents a thyrotrophic cell (basophil type 2), unpublished data). However, paradoxically, the gonado- whereas the other 2 are probably gonadotrophic (types 1 troph hyperplasia does not result in an increase in plasma and 3). Evidence for gonadotrophic roles for both these GTH levels (E. Tan, J. F. Leatherland, and R. A. Sonstegard, cells is provided by methallibure-treated fish in which both unpublished data). Thus, in hybrids older than 4 years (both cell types were partly diminished in a manner similar to that tumorous and nontumorous) and in tumorous carp and found in other species treated with methallibure (19, 32). goldfish, the negative feedback system, which is thought to Kaul and Vollrath (16) were able to identify only 1 putative regulate the pituitary-gonad axis in teleosts, is either im gonadotrophic cell type, the description of which corre paired or inoperative. sponds to the type 1 basophil reported here. The differ The commencement of gonadotroph proliferation in the ences in observations are difficult to explain, since in both

SEPTEMBER 1977 3155

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1977 American Association for Cancer Research. J. F. Leather/and and R. A. Sonstegard this study and that of Kaul and Vollrath (16) mature fish were 5. Bornett, J., and Fisher. R. Investigations of Filter Activated Carbon after Utilization in Water Plant. Arch Hyg. Bakteriol., 146: 1-16, 1962. used. It is possible that the discrepancies between the ob 6. Brown, E., Hazdra, J., Keith, L., Greenspan, I., Knapinsi. J., and Beamer. servations described here and those of Kaul and Vollrath P. Frequency of Fish Tumors Found in a Polluted Watershed as Com pared to Nonpolluted Canadian Waters. Cancer Res.,33: 189-198, 1973. (16) were due to the different sizes of fish used in the 2 7. Chavin, W. Pituitary-Adrenal Control of Melanization in Xanthic Goldfish studies. Kaul and Vollrath used specimens of unknown age, Carassius auratus L. J. Exptl. Zool.. 733: 1-45, 1956. 18 to 20 cm in total length, whereas specimens up to 12 8. Chestnut, C. W. The Pituitary Gland of Coho Salmon (Oncorhyncrtus kisutch Walbaum) and Its Function in Gonad Maturation and Thyroid years old and 60 to 70 cm in total length were used in this Activity. Ph.D. Dissertation, Simon Fraser University. 1970. study. Thus, the type 3 gonadotrophs, which are sparse in 9. Cillie, G. G., Van Varren, L. R. J.. Stander, G. J., and Kolbe. F. F. The smaller fish, may not have been identified as a separate cell Reclamation of Sewage Effluents for Domestic Use. In: S. H. Jenkins and L. Mendia (eds.), Proceedings of the Third International Conference type by the other authors. Another possible explanation is on Water Pollution Research, Munich, Vol. 2, pp. 1-36. Washington. that the 2 cell "types" represent different stages in the D.C.: Water Pollution Control Federation, 1966. 10. Cook. H., and Overbeeke, A. P. van. Ultrastructure of the Pituitary Gland development of the same cell type. However, the type 1 and (Pars Distalis) in Sockeye Salmon (Oncorhynchus nerka) during Gonad type 3 basophils differ in size, position within the gland, and Maturation. Z. Zellforsch. Mikroscop. Anat.. Ã•30338-350, 1972. 11. Emmart, E. W. The Localization of Endogenous "Prolactin" in the Pitui several cytoplasmic characteristics, including granule di tary Gland of the Goldfish, Carassius auratus, Linnaeus. Gen. Comp. ameter, vacuolation, and the presence of the PAS- or aniline Endocrinol., 12: 519-525, 1969. blue-stained rod-shaped bodies. These differences provide 12. Harris, R. H. The Implications of Cancer Causing Substances in Missis strong evidence for separate cell types. This hypothesis may sippi River Water. Washington, D. C.: Environmental Defense Fund, 1974. be supported by findings in salmonid species, in which only 13. Hoar, W. S.. Wiebe, J. P., and Wai, E. H. Inhibition of the Pituitary 1 gonadotroph has been described during the major portion Gonadotropic Activity of Fishes by a Dithiocarbamoylhydrazine Deriva tive (I.C.I. 33,828). Gen. Comp. Endocrinol., 8: 101-109, 1967. of the life cycle, but where a 2nd gonadotroph type be 14. Holmes, R. L., and Ball, J. N. The Pituitary Gland, a Comparative Ac comes evident in the sexually mature fish (4, 8, 10, 29, 39). count. London: Cambridge University Press. 1974. The ultrastructural appearance of the somatotrophs in 15. Jolly, R. L. Chlorination Effects on Organic Constituents in Effluents from Domestic Sanitary Sewage Treatment Plants, Publication 55. Oak these large cyprinids gave every impression of a high secre Ridge: Environmental Science Division, Oak Ridge National Laboratory, tory activity. This is surprising in light of the slow growth 1973. 16. Kaul. S., and Vollrath, L. The Goldfish Pituitary. I. Cytology. Cell Tissue rate of the older fish. These observations strongly suggest Res., 154: 211-230, 1974. that the somatotrophs are involved in processes other than 17. Komourdjian. M. P., Saunders, R. L., and Fenwick, J. C. Evidence for the growth regulation throughout the animal's whole life his Role of Growth Hormone as a Part of a Light-Pituitary Axis' in Growth and Smoltification of Atlantic Salmon (Salmo salar). Can. J. Zool.. 54: tory; a similar idea was recently proposed by Komourdjian 544-570, 1976. era/. (17). 18. Kraybill. H. F. Distribution of Chemical Carcinogens in Aquatic Environ Methallibure acts as a goiterogen in some fish (13, 32). No ments. Progr. Exptl. Tumor Res.. 20: 3-34, 1976. 19. Leatherland, J. F. Studies on the Structure and Ultrastructure of the marked effects of methallibure were found in the thyro- Intact and "Methallibure"-treated Meso-adenohypophysis of the Vivipa trophs. This finding agrees with that of others (13), who rous Teleost Cymatogaster aggregata Gibbons. Z. Zellforsch. Mikros showed that cyprinids are less susceptible to the goitero- cop. Anat., 98: 122-134, 1969. 20. Leatherland, J. F. Histophysiology and Innervation of the Pituitary Gland genic action of methallibure than other species. Certain of the Goldfish. Carassius auratus L.: a Light and Electron Microscope teleost fishes may be particularly suited for monitoring lev Investigation. Can. J. Zool., 50 835-844, 1972. 21. Leatherland, J. F.. Bekolay, R., McLean, R. S., and Brown, D. Pituitary els and effects of environmental carcinogens, particularly of Control of Pigmentation in Xanthic. Moor and White Carassius auratus L. carcinogens that cause endocrine dysfunction. The separa Can. J. Zool.. 55: 456-464, 1977. tion of the teleost pars distalis into 2 distinct zones (each of 22. Leatherland, J. F., and Ronald. K. Structure of the Adenohypophysis in Juvenile Harp Seal. Pagophilus groenlandicus. Cell Tissue Res., 773: which contains functionally different cell types) permits a 367-382, 1976. relatively easy identification of adenohypophysial cell types, 23. Levenstein, I. The Cytology of the Pituitary Gland of Two Varieties of compared with other vertebrates. Moreover, because of the Goldfish (Carassius auratus L.) with Some Reference to Variable Factors in the Gland Which May Possibly Be Related to the Different Morphologi process of biological magnification, fishes are often ex cal Types. Zoologica, 24: 47-60. 1939. posed to waterborne environmental insults of an order 24. Maugh, T. G. Trace Elements. A Growing Appreciation of Their Effects on Man. Science. 181: 253-254, 1973. higher than those to which other animals are exposed and 25 Morrison, A. B. The Role of the Food and Drug Directorate in Protecting to which they may be unusually responsive. the Canadian Public against Environmental Carcinogens. In: P. G. Scholefield (ed.). Proceedings of the Ninth Canadian Cancer Research Conference, pp. 37-49. Toronto: University of Toronto Press, 1971. 26. Mueh. J., Shane. J. M., and Jaftolin, F. The Effects of Methallibure (I.C.I. ACKNOWLEDGMENTS 33,828) on LH Release in Castrate Male Rats Challenged with LHRH. Endocrinology. 97: 493-495, 1975. We are indebted to L. Lin, B. Hicks, and M. Burke for their technical help. 27. Nagahama, Y., and Yamamoto. K. Basophils in the Adenohypophysis of the Goldfish (Carassius auratus). Gunma Symp. Endocrinol., 6: 39-55, 1969. 28. Olivereau, M. Cytologie de l'Hypophyse du Cyprin (Carassius auratus REFERENCES L.). Compt. Rend.,255: 2007-2009, 1962. 29. Olivereau, M. Les Cellules Gonadotropes Hypophysaires du Saumon de 1. Ball. J. N., Baker, B. I.. Olivereau, M.. and Peter, R. E. Investigations on l'Atlantique: UnicitÃ©ou DualitÃ©?Gen. Comp. Endocrinol.. 28: 82-95. Hypothalamic Control of Adenohypophysial Functions in Teleost Fishes. 1976. Gen. Comp. Endocrinol. Suppl..3: 11-21. 1972. 30. Ongerth. H. J., Spath, D. P., Crook, J.. and Greenberg, A. E. Public 2. Bell, W. R. Morphology of the Hypophysis of the Common Goldfish Health Aspects of Organics in Water. J. Am. Water Works Assoc., 65: (Carassius auratus L.). Zoologica, 23: 219-236, 1938. 495-497, 1973. 3. Sellar, T. A., Lichenberg, J. J., and Kroner, R. C. The Occurrence of 31. Page, T., Harris. R. H., and Epstein, S. S. Drinking Water and Cancer Organohalides in Chlorinated Drinking Waters. J. Am. Water Works Mortality in Louisiana. Science. 194: 55-56, 1976. Assoc.,66 703-706. 1974. 32. Pandey, S.. and Leatherland. J. F. Comparison of the Effects of Methalli 4. Boddingus. J. The Influence of Social Rank on Adenohypophysial Cell bure and Thiourea on the Testis, Thyroid and Adenohypophysis of the Activity in Sa/mo irideus. With Special Reference to Basophil ACTH-Cells Adult and Juvenile Guppy, Poecilia reticulata Peters. Can. J. Zool., 48: of the Proximal Pars Distalis. Cell Tissue Res., 170: 383-414, 1976. 445-450, 1970.

3156 CANCER RESEARCH VOL. 37

33. Rook, N. J. Formation of Haloforms during Chlorination of Natural 37 Sonstegard, R. A. Environmental Carcinogenesis Studies of Fishes in the Waters. J. Sue. Water Treat. Exam., 23. 234-243, 1974. Great Lakes of North America. Ann. N. Y. Acad. Sci., in press. 34. Saffiotti, U. The Laboratory Approach to the Identification of Environ- 38 Sonstegard, R. A. The Potential Utility of Fishes as Indicator Organisms for Environmental Carcinogens. In: F. M. D'ltrie (ed.), Wastewater Reno mental Carcinogens. In: P.G. Scholefield (ed.), Proceedings of the Ninth Canadian Cancer Research Conference. Toronto: University of Toronto vation and Reuse, pp. 561-577. New York: Marcel Dekkar Inc., 1976. Press, pp. 22-36. 1974. 39 Sonstegard, R., and Leatherland, J. F. Studies of the Epizootiology and 35. Scruggs, W. A. The Epithelial Components of the Teleost Pituitary Gland Pathogenesis of Thyroid Hyperplasia in Coho Salmon. Cancer Res., 36. as Identified by a Standardized Method of Selective Staining. J. Mor- 4467-4475, 1976. phol.,65: 187-212, 1939. 40 Sonstegard, R. A., Leatherland, J. F., and Dawe, C. J. Effects of Gonadal 36. Sonstegard, R. A. Neoplasia Incidence Studies in Fishes Inhabiting Tumours on the Pituitary-Gonadal Axis in Cyprinids from the Great Polluted and Non-polluted Waters in the Great Lakes of North America. Lakes. Gen. Comp. Endocrinol.. 29; 269, 1976. In: P. G. Scholefield (ed.), Eleventh International Cancer Congress 41 Wogan, G. N. Naturally Occurring Carcinogens in Foods. Progr. Exptl. Panel, Vol. 17, pp. 172-173. Toronto: University of Toronto, 1974. Tumor Res. 11: 134-162, 1969.

Fig. 3. a, part of the neurointermediate lobe in a hybrid, showing 2 pars intermedia cell types, 1 of which is stained with PAS (small arrow}. Note the penetration of pars nervosa tissue (large arrow) among the strands of intermedia tissue. Paraffin section, AB-PAS-OG, x 270. b, colloid-filled follicles (F) in the pars intermedia tissue of a 9-year-old hybrid. Paraffin section, Cleveland-Wolfe, x 270. Fig. 4. Parasagittal section (10 to 20 Mto 1 side of the midline) of a pituitary gland in a 6-year-old nontumorous carp. Note the rostral pars distalis (R) and neurointermediate lobe iNi on either side of the proximal pars distalis (Pi. The latter is made up of more or less equal numbers of basophils (dark cells) and somatotrophic acidophils (pale cells). Paraffin section, AB-PAS-OG, x 23. Fig. 5. Section in a region of the pituitary similar to that in Fig. 4, showing pituitary gland in a 6-year-old tumor-bearing hybrid. The pale rostral pars distalis cells (arrows) now form a thin layer of the periphery of the greatly enlarged proximal pars distalis (P). The latter is composed, almost in its entirety, of gonadotroph cells. N, neurointermediate lobe. Paraffin section, AB-PAS-OG, x 23. Fig. 6. Part of the proximal pars distalis in a nontumorous carp, showing groups of pale acidophils (somatotrophs) interspersed with darker basophils (gonadotrophs). The clumps of basophils are divided into a central region of smaller cells (small arrows) surrounded by a ring of larger, more coarsely granulated cells (large arrows). Paraffin section, AB-PAS-OG, x 245. Fig. 7. Enlarged area of a clump of proximal pars distalis basophils, showing PAS-stained rod-shaped bodies (arrows) in the small basophils. Paraffin section, AB-PAS-OG, x 290. Fig. 8. Part of the proximal pars distalis in a tumorous carp. The larger gonadotrophs (which are vacuolated in this specimen) occupy much of the region: the clumps of acidophils (A} occupy only small areas compared with nontumorous fish (Fig. 6). Paraffin section, Cleveland-Wolfe, x 290. Fig. 9. Part of the proximal pars distalis in a methallibure-treated hybrid. Note the marked reduction in the population of gonadotrophs compared with Fig. 8. A, acidophils (somatotrophs). Paraffin section, AB-PAS-OG, x 230. Fig. 10. Region similar to Fig. 9 in a methallibure-treated carp, showing the marked reduction in gonadotroph population compared with Fig. 6. A, acidophils (somatotrophs). Paraffin section, AB-PAS-OG, x 230. Fig. 11. Sagittal section of the pituitary in 2-year-old hybrid. The section shows the separation of the zones of the pituitary, fl, rostral pars distalis: P, proximal pars distalis; PI, pars intermedia; NH, neurohypophysis (pars nervosa). There is no enlargement of the proximal pars distalis at this stage. Paraffin section, AB-PAS-OG, x 100. Fig. 12. Enlarged area of the neurointermediate lobe in Fig. 11. The darkly stained cells (arrows) are stained with AB and represent the non-PAS-stained cells in Fig. 3. The PAS-stained cells appear light grey in the micrograph. Paraffin section, AB-PAS-OG, x 190. Fig. 13. Enlarged area of proximal pars distalis in Fig. 11. Dark-staining basophils and clear acidophils (A) are apparent, but there is no sign of the proliferation of the gonadotrophs found in older hybrids. Paraffin section, AB-PAS-OG, x 190. Fig. 14. Prolactin cell of the rostral pars distalis of a nontumorous carp. Note the content of electron-dense cytoplasmic granules and active Golgi complex, x 10,580. Fig. 15. ACTH cells in the same animal as in Fig. 14. Note the smaller granules in this cell compared with the prolactin cell in Fig. 14. The ACTH cells commonly abut the basement membrane (M), which separates the rostral pars distalis and neurohypophysis (NH). x 10,580. Fig. 16. Area of rostral pars distalis in nontumorous carp, showing ACTH cell (Ai and prolactin cells (P) adjacent to the electron-dense basement membrane complex, which separates the neurohypophysis (NH) from the rostral pars distalis. x 10,580. Fig. 17. Nongranulated cell in nontumorous carp among a group of prolactin cells, x 10,580. Fig. 18. Rostral pars distalis cells in nontumorous carp, showing a type 2 basophil (thyrotroph) (large arrow) and 2 type 1 basophils (small arrows) adjacent to a somatotroph. x 10,735. Fig. 19. Similar to Fig. 20, showing the extensive endoplasmic reticulum in somatotrophs. x 10,175. Fig. 20. Part of 2 somatotrophs, showing the rod-shaped electron-dense granules of these cells, x 10,175. Fig. 21. Part of a type 1 basophil in a nontumorous carp, showing the large vacuole (V) in these cells. Note also the electron-dense cytoplasmic inclusions and the larger, less electron-dense droplets, x 13,500. Fig. 22. Similar to Fig. 21, showing the small electron-dense granules (which vary in electron density) and a larger droplet adjacent to a vacuole. In addition, membrane-bound organelles filled with a fibrous matrix are present, x 17,730. Fig. 23. Similar to Figs. 21 and 22, showing "fibrous" organelles, which also contain a core of material similar to that found in the droplets, x 13,575. Fig. 24. Adjacent basophil cell types 1 and 3. Note that the electron-dense granules in type 1 cells (/) are markedly larger than in type 3 (///), and also that the cell vacuolation in type 3 cells takes the form of small clumps of vacuoles rather than the single large vacuole (V] of the type 1 cells, x 13,470. Fig. 25. Apparently degenerating vacuolated type 1 basophil cell (V) adjacent to a nongranulated cell (WG) in the proximal pars distalis of a nontumorous carp, x 10,971. Fig. 26. Globular inclusion in the cytoplasm of a basophil type 3 cell of a tumorous hybrid, x 17,440. Fig. 27. Pars intermedia cell corresponding to the non-PAS-staining cell in light micrographs. Note the variation in electron density of the cytoplasmic granules and the marked formations of endoplasmic reticulum. x 10,580. Fig. 28. Similar to Fig. 27, showing cells corresponding to PAS-staining pars intermedia (PI) cells. NH, neurohypophysis. x 10,580.

SEPTEMBER 1977 3157

3158 CANCER RESEARCH VOL. 37

SEPTEMBER 1977 3159

3160 CANCER RESEARCH VOL. 37

SEPTEMBER 1977 3161

OF'-"mÃ â€¢A'*'**

* *â€¢ â€¢ â€¢â€¢ â€¢> 'â€¢>vvsÃteoft.>. \ V9f* **Â£&â€¢ Ã * â€¢â€¢â€¢ 'â€¢'.

â€¢ ^ft^k.4vS

'â€¢-^*^ ^*'â€¢:-;14NH0 _ â€¢^ " *^

.â€¢^*gvV:

3162 CÃ•NCER RESEARCH VOL. 37

SEPTEMBER 1977 3163

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1977 American Association for Cancer Research. *W3 19 *â€¢>Â«?**â€¢vfÃÃ¡^^ . smv '. â€¢*''^^^Mfe / $Â£iU&*& A> -:â€¢ Vtei^V * V .".â€¢>.4 â€¢Â»â€¢.^

3164 CANCER RESEARCH VOL. 37

SEPTEMBER 1977 3165

3166 CANCER RESEARCH VOL. 37

SEPTEMBER 1977 3167

V-

Â» *%- '"^-

3168 CANCER RESEARCH VOL. 37

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1977 American Association for Cancer Research. Structure and Function of the Pituitary Gland in Gonadal Tumor-bearing and Normal Cyprinid Fish

John F. Leatherland and R. A. Sonstegard

Cancer Res 1977;37:3151-3168.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/37/9/3151

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/37/9/3151. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.