Maudsley Hospital, London, S.E. 5 (Received 15 June 1959)

Transcription

1 394 J. Physiol. (I959) I49, THE EFFECTS OF PARTIAL HYPOPHYSECTOMY IN THE RABBIT By H. J. CAMPBELL From the Department of Neuroendocrinology, Institute of Psychiatry, Maudsley Hospital, London, S.E. 5 (Received 15 June 1959) Section of the pituitary stalk in the adult female chinchilla rabbit is followed by a reduction in volume of the pars distalis (Campbell & Harris, 1957) and by some loss of pars distalis function (Fortier, Harris & McDonald, 1957; Brown-Grant, Harris & Reichlin, 1957). It has been suggested that the deficit in pars distalis function found in stalk-sectioned rats may be due to the reduced volume of the gland (Barrnett & Greep, 1951). Few data are recorded on the functional capacity of anterior pituitary remnants left after partial hypophysectomy, apart from the early studies of Smith (1932) in the rat, and the recent findings of Ganong & Hume (1956) in the dog. This paper deals with an investigation undertaken to determine the dependency of adrenocortical, thyroidal and ovarian function on pars distalis volume. The tests used were similar to those employed by Fortier et al. (1957) and Brown-Grant et al. (1957). METHODS Forty-nine adult female chinchilla rabbits of body weight ranging from 2-0 to 3-5 kg were used. They were kept under uniform conditions of light, temperature and diet (M.R.C. pellet diet 18, supplied by A. C. Taylor Ltd, London) and tap water ad libitum from the time they were introduced into the animal house until killed. The 49 animals consisted of 26 normal controls, 8 completely hypophysectomized and 15 subjected to varying degrees of partial hypophysectomy. No post-operative tests were done until at least 10 days from the operation of hypophysectomy. Tests were done in the following sequence: lymphopenic response to emotional stress, thyroidal release of radio-iodine in response to emotional stress, thyroidal radio-iodine release and lymphopenic response to physical stress (done simultaneously), ovulation response to coitus as a terminal procedure, thyroidal radio-iodine uptake and adrenal ascorbic acid depletion measured in response to physical stress. Hypophysectomy was performed by the parapharyngeal method of Jacobsohn & Westman (1940) with the modifications described by Fortier et al. (1957). Emotional stress was induced by placing the animals for 3 hr in the restraining clamp described by Harris (1948). Physical stress was produced by laparotomy. Under ether anaesthesia a mid-line abdominal incision was made, extending 10 cm caudally from a point just below the xiphisternum. The

2 PARTIAL H YPOPH YSECTOM Y 395 saline-moistened abdominal viscera were exposed to air for min, at the end of which time the incision was closed in layers with continuous and interrupted sutures. A single dose of Seclomycin (penicillin and streptomycin; Glaxo Laboratories Ltd.) was administered subcutaneously immediately after operation. Thyroidal release of radio-iodine. This was determined in a manner similar to that described by Brown-Grant, von Euler, Harris & Reichlin (1954), using 12-hourly neck counts after injection of 5,uc 131J. Adrenal ascorbic acid content and depletion, combined with thyroidal radio-iodine uptake. At approximately the same time of day each animal was given a subcutaneous injection of 5,tc 1311 and kept in a diuresis cage for a further 45 min. Left adrenalectomy was then performed as rapidly as possible under ether anaesthesia, and the thyroid gland removed; the time of removal of the thyroid varied from 58 to 62 min after the injection of radio-iodine. Both ureters were clamped and cut and a further clamp placed on the urethra-vaginal junction. After collecting a blood sample from the marginal vein of the ear, the abdominal and neck wounds were closed with clips. Anaesthesia was continued, the right adrenal removed 1I hr after the time of the left adrenalectomy and the animal was killed. During the waiting period the thyroid gland and left adrenal were dissected free from fat and weighed to the nearest 0-1 mg. The thyroid was cut into two pieces, one piece weighed separately and kept for radioactivity measurements, the other piece being reserved for histological examination. After killing the animal, the bladder was removed and washed out, the washings being added to any urine in the flask under the diuresis cage. The measurement of the radioactive content of the thyroid, urine and plasma was performed by methods similar to those described by Brown-Grant, von Euler et al. (1954). Determination of adrenal ascorbic acid content was done by the method described by Fortier et al. (1957). Blood lymphocytes. The validity of the lymphopenic response as an index of the release of adrenocorticotrophic hormone in the rabbit has been demonstrated by Colfer, de Groot & Harris (1950). The technique used in the present study was essentially similar to that of Fortier et al. (1957). Histological procedures. Immediately after death the rabbits were injected with 200 ml 1:3 Indian ink (Mandarin Black, Winsor & Newton) through the carotid arteries; the head was removed, trimmed and fixed in 10% formalin saturated with mercuric chloride. After complete fixation all heads were decalcified in a formic acid and sodium formate mixture, and dehydrated with iodized ethyl alcohol solution. A block of tissue containing the hypothalamus, base of skull and pituitary gland was embedded in low-viscosity nitrocellulose, serially sectioned at 200p and lightly stained with haematoxylin and eosin. The volume of the pituitary gland and its parts was determined in the manner described by Campbell & Harris (1957). Wherever it was difficult to be cer tain as to the part of the pituitary gland represented by a remnant, or where it was thought of special interest to know the cytologyof a fragment of pituitary tissue, the thick celloidin sections, after volume tracings had been made, were dismounted and de-stained. The whole section or the relevant part of it was then embedded in wax and sectioned at 5-10 u. These thin sections were stained trichromatically with a modification of the Mallory-Azan technique. The periodicacid-schiff method was also used. The ovaries, adrenal glands and thyroid lobes were fixed in 10% formalin, dehydrated, embedded in wax, sectioned at 10 tu and stained with haematoxylin and eosin, care being taken to ensure uniformity of treatment for all tissues. Measurement of thyroid epithelial cell height was carried out as follows: thyroid gland sections were examined under the oil immersion lens with a binocular microscope fitted with a micrometer eyepiece. A clearly defined cell from each of five distinct follicles from the widest part of each thyroid lobe was measured five times. The mean height of these five cells was taken as the 'epithelial cell height' for that animal. Measurements were made of ovarian follicular diameters. The sections of ovary were examined under a 33 mm objective, and the largest diameter of the five largest follicles in either of the ovaries of each animal was measured with a micrometer eyepiece. The mean of the five determinations for each animal was taken as the 'follicular diameter' for that animal.

3 396 H. J. CAMPBELL RESULTS Pituitary volume Pars distalis. The volume of the pars distalis determined in ten normal animals was F57 mm3. This figure is about 40% greater than that given by Campbell & Harris (1957). An increase of this order was found for the other parts of the pituitary gland in this present series. The explanation offered is that the pituitaries described by Campbell & Harris (1957) were fixed in 10% formalin, whereas those of the present study were fixed in formol-sublimate; it is well known that the latter fixative produces considerably less shrinkage of tissues than formalin alone. Table 1. Absolute and relative (normal = 100) volumes of pars distalis remnants in partially hypophysectomized rabbits Absolute Absolute No. (mm") Relative No. (mm3) Relative 10 0* * * '93 3*9 38 8* *0 4 8* * ' * '99 54* * * The volumes of the pars distalis remnants found in the partially hypophysectomized animals are shown in Table 1. In all except one animal the pars distalis remnant had excellent vascular connexion with the primary plexus of the median eminence, as demonstrated by ink-filled portal vessels. The exceptional animal was rabbit No. 24 where the remnant was located under the dorsum sellae and was well vascularized from surrounding cancellous bone. A portion of the infundibular process remained attached to the median eminence in this rabbit, but no, or practically no, vascular connexion existed between the median eminence and the pars distalis remnant. It will be seen from results presented that rabbit No. 24 consistently exhibited a functional depression similar to that of the completely hypophysectomized animal. Other rabbits with less pars distalis showed a higher level of function. Pars intermedia. Pars intermedia remnants occurred in eleven partially hypophysectomized rabbits. The volume of the remnant could not be correlated with any target gland function. Neural lobe. Neural lobe remnants occurred in twelve of the partially hypophysectomized animals. No correlation existed between the volume of the remnant and any target gland function. Thyroid Weight. The thyroid weights found in the several groups of animals, and the statistical data derived from them are given in Table 2.

4 PARTIAL HYPOPHYSECTOMY 397 Histology. Visual comparison of thyroid sections from animals in the various groups gave a rough correlation between histological characters (size of follicles, amount of colloid, appearance of cells) and the pars distalis volume, but it was considered that more precise quantitative methods would be useful. Accordingly, the thyroid epithelial cell height was measured. There is much variability in the size of thyroid follicles in the rabbit and it is intended that the figures given in Table 2 for epithelial cell height should be taken as only approximate. TABLE 2. Thyroid: summary of data and statistical derivations Pars distalis Complete, Normal hypophysectomy <30 0/* >30%/. <10 k >10% Weight -. Absolute (mg) (10) 149-5±15t (3) 66-9= E12 Relative '17 Significance of - <OAD2 difference from normal (P) Significance of mtergroup difference (P) Cell height Absolute (in-8) (10) 592:0±29-7 (3) ±19.1 (9) 342-8±52-5 (6) 479-1±13-8 (6) 291-4±45-5 (9) 480-0±37-3 Relative * Significance of - <OADoi <0001 <0 05 <0-001 <0 05 difference from normnal (P) Significance of mtergroup difference (P) Iodine uptake (10) 4-85±0-65 (1) 0-68 % dose by whole gland % dose by i mg gland Relative uptake : I. By whole gland By 100 mg gland Significance of difference in uptake from normal (P) By whole gland By 100mg gland * Pars distalis remnant below 30% of normal volume. (9) 111-0±10 (6) 124-4±12 (6) 100-4±12 (9) 127-0± <0-01 >0-2 <0 05 >02 >04 >02 <0-05 <0-02 (6) 4-32±1-34 (5) 4-81±0-49 (5) 3-29 b 1-58 (6) 5-25± >0-7 > ± ± >0-9 _ > t ±standard error; number of animals in brackets. Rate ofrelease ofradio-iodine. The mean rate of release, determined in eleven experiments on eleven normal rabbits, was found to be % per day, a figure in close agreement with that reported by Brown-Grant et al. (1957). In six experiments on three completely hypophysectomized animals the corresponding mean was % per day, a reduction of 87-2 % from normal. Twenty-two thyroid radio-iodine release experiments were carried out on fifteen partially hypophysectomized animals. The release curves fell fairly sharply into two groups. Rabbits with pars distalis remnants greater than 10% of normal volume gave release curves of the normal or 'steep slope'

5 398 H. J. CAMPBELL type (Fig. 1 a). Animals with pars distalis remnants below 10% of normal volume gave release curves of the 'flat slope' type (Fig. 1 b) as obtained with the completelyhypophysectomized animal (Brown-Grant, von Euler et al. 1954). Two rabbits in this latter group (Nos. 9 and 14) were treated with thyrotrophic hormone (0.5 mg, Armour Laboratories Ltd, Lot No. R377157) during 4) _C v4) rjcc E =1 I 4o Cz 4) Rabbit 19 (a) E E S z 1-_ Rabbit 16 > ' 25I Hours 1001 I I I I I Hours C a _._ -a c Pars distalis volume (% of normal) Fig. 1. Partial hypophysectomy. (a) Typical release curve for radio-iodine of the thyroid of a normal rabbit or one with more than 10% (in this animal 12%) of normal pars distalis volume. (b) Typical release curve from a completely hypophysectomized animal or one with less than 10% normal pars distalis volume. (c) Relation between thyroidal rate of release of radio-iodine (as %/day) to volume of pars distalis remnant (as % of normal). a release curve determination. There was an immediate increased release in both animals similar to that described by Brow-n-Grant, von Euler et at. (1954). The relation between rate of release and pars distalis volume is shown graphically in Fig. 1 c. It may be seen that there was no appreciable change in rate of release in the series of animals with pars distalis remnants ranging from 10 to 100% of normal. Uptake of radio-iodine. The values found and statistical data derived from

6 PARTIAL HYPOPHYSECTOMY 399 them are given in Table 2. It was possible to carry out an uptake experiment on only one completely hypophysectomized rabbit, but in this animal a clearcut result was obtained. There is obviously no significant difference between the group of animals with less than 30 % of normal pars distalis volume and the group with more than this amount. In view of the fact that the uptake figures in the group of animals with less than 10% of normal pars distalis volume are not normally distributed, it is not possible to apply tests of statistical significance to the differences. Statistical methods are not necessary to see that the group with remnants greater than 10% of normal has much greater uptakes than the group with smaller remnants. It is of interest that the l3l uptake of rabbit 24, the animal with little or no portalvessel connexion, was similar to that found in the completely hypophysectomized animal (uptake by whole gland, 0-29 %; by 100 mg, 0 26 %). To see whether the difference in thyroidal uptake of 131J in the two operated groups could be explained by differences of urinary excretion of 131J, measurements of the radio-iodine content of the urine and plasma of the partially hypophysectomized animals were performed. These showed that no significant difference existed between the various groups in this respect. Stress responses. Brown-Grant, Harris & Reichlin (1954) have demonstrated that restraint diminishes the rate of radio-iodine release from the thyroid gland and that the normal rate is rapidly restored when the restraint is removed. This has been confirmed in eight experiments on eight animals. That complete hypophysectomy results in a curve of the 'flat slope' type has also been confirmed by five experiments on three rabbits. Those partially hypophysectomized animals (9 in number) which had pars distalis remnants larger than 10 % of the normal behaved like normal animals in that restraint slowed the release of radio-iodine (Fig. 2a). Those (6 in number) which had remnants of less than 10% behaved like the completely hypophysectomized animal: the release of radio-iodine was very slow (flat-type curve) and consequently no attempt was made to show any further slowing as a result of emotional stress. Likewise, the slowing effect of physical stress (laparotomy) on iodine release demonstrated by Brown-Grant et al. (1954) was observed with six animals which exhibited 'steep-slope' release curves (Fig. 2 b) and was not investigated on the 'flat-slope' type. One animal (No. 14) gave an atypical response in that the release curve first determined was of the flat type typical of an animal with less than 10% distalis volume. This release was increased by thyrotrophic hormone. The release, however, 20 days later was at the rapid rate of 8% per day and was not diminished by laparotomy. This animal had 5% of normal pars distalis volume.

7 400 H. J. CAMPBELL Restraint Hours 3- c 4-0, C E 0) C 0. S 0 4, z o Hours Fig. 2. Partial hypophysectomy. (a)typical release curve for radio-iodine of the thyroid during the stress of restraint in a normal rabbit or one with more than 10% normal pars distalis volume (in this animal 70%). (b) Typical release curve response to laparotomy in a normal rabbit or one with more than 10% (in this animal 12%) of normal pars distalis volume.

8 PARTIAL HYPOPHYSECTOMY 401 Adrenal Weight. The values found and the statistical data derived from them are given in Table 3. Histology. Very marked asymmetry of the cortex in the normal animals precluded the use of exact measurements such as of the cortex: medulla ratio. The most obvious feature in the completely hypophysectomized animals was a marked reduction in the width of the fasciculate zone. The reticulate zone exhibited a much smaller reduction in width. There appeared to be a definite increase in the glomerular zone width. The adrenal cortices of the partially hypophysectomized animals fell into two fairly sharply defined groups. TABLE 3. Adrenal: summary of data and statistical derivations Complete Pars distalis hypo- A Normal physectomy <30%* >30% <10% >10% Weight Absolute (mg) (10) 441-2±32t (6) 160-5±17 (9) 27609±24 (6) 436-1±49 (6) 24406±26 (9) 404*5±35 Relative * * Significance of difference - <0 001 <001 >0 9 <0.01 >0.4 from normal (P) Significance of intergroup <0-01 <0 01 difference (P) Ascorbic acid content Absolute (jg) By whole gland (22) 516±50 (4) 326±62 (4) 331±56 (5) 465±72 (3) 352±40 (6) 433±67 By 100mg 225±16 355± ±15 268±54 209±15 Relative By whole gland By loomg Lymphopenic response to emotional stress Absolute drop (8) 51±3 (2) 7±7 - (6) 35+5 (9) 44±5 Relative drop Significance of difference - - <005 >0 2 from normal (P) No. of experiments Conventions as for Table 2. Rabbits with less than 5% of normal pars distalis volume had adrenals which were close to the appearance in the completely hypophysectomized with enlarged glomerular zone and markedly atrophic fasciculate zone. Animals with more than 5 % of normal pars distalis volume showed a very slight atrophy of the fasciculate zone but no other abnormal features. Ascorbic acid content. There is obviously (see Table 3) no significant difference in concentration of ascorbic acid, but there is a difference in whole gland content of ascorbic acid between the groups with more or less than 30 %, and the groups with more or less than 10 % of normal pars distalis volume. Lymphopenic response to emotional stress. The effect of emotional stress in the form of restraint was tested in twenty-nine experiments on eight normal rabbits, three experiments on two completely hypophysectomized animals and twenty-two experiments on fifteen partially hypophysectomized rabbits (Table 3). In the individual rabbit (No. 24) with little or no portal vessel

9 402 H. J. CAMPBELL connexion with the pars distalis remnant, which measured 6 % of normal volume, there was no significant fall in blood lymphocytes. Rabbits with less than this amount of pars distalis tissue gave a positive response to the test. Lymphopenic response to physical stress. The effect of physical stress in the form of laparotomy was determined in nine normal rabbits. The mean drop in absolute number oflymphocytes was % /mm3. In four completely hypophysectomized animals laparotomy produced an increase of %/mm3. This difference from normal is evidently significant; lymphocytosis had occurred. Laparotomy was performed on five partially hypophysectomized animals with pars distalis remnants ranging from 5 to 70 %. Each animal gave a significant lymphopenic response. Ascorbic acid depletion withphysical stress. Laparotomyin twenty-two normal rabbits produced a mean fall in ascorbic acid concentration of %. In four completely hypophysectomized animals the corresponding figure was / ; this is 28-3% of normal. Three animals with pars distalis remnants below 10% ofnormalvolume gave a mean ascorbic acid depletion of %/". The corresponding figure in six animals with pars distalis remnants greater than 10% of normal volume was %. These figures are not significantly different from each other or from normal, whereas both are significantly different at the P = 041 level from the value obtained in completely hypophysectomized animals. Ovary Weight. One normal animal was found to be pseudopregnant when killed; its ovarian weight is not included in the data given in Table 4. There was no significant difference between the ovarian weight of the normal group and the ovarian weight ( mg) of the animals which ovulated (see below). TABLE 4. Ovary: summary of data and statistical derivations Pars distalis Complete, Normal hypophysectomy <30%* >30% <10% >10% Weight Absolute (mg) (9) 648-0±112t (7) 113-9±11 (9) 181-8±20 (6) 425-7±91 (6) 186-3±28 (9) 341-4±73 Relative Follicular diameter Absolute (mm) (8) 1-320±0-16 (3) 0-411±0-03 (9) ±10 (6) 1-292±0-14 (6) 0-645±0-13 (9) 1-100±0-14 Relative Conventions as for Table 2. Histology. On visual comparison of ovarian sections a rough correlation between histological features and pars distalis volume was observed, but it was considered that measurements of follicular diameters should be made. The values obtained are given in Table 4. Post-coital ovulation. The ovulation response to mating was tested in ten normal, two completely hypophysectomized and thirteen partially hypophysectomized rabbits. The procedure was as follows. The female experimental

10 PARTIAL HYPOPHYSECTOMY 403 animal was placed successively in the cages of two normal male rabbits until mating occurred, or for two 10 min periods. In the absence of mating the test was repeated three times at weekly (or longer) intervals. If copulation occurred on any occasion, the animal was killed about 48 hr later and the ovaries were examined for ovulation points. If copulation did not occur, the animal was killed after the third test and its ovaries examined in the usual way. Nine of the ten normal animals accepted both males within 5 min on the first occasion that they were put together. The other normal rabbit accepted the male on the third occasion. In all ten normal animals the ovaries showed multiple ovulation points when examined 48 hr after copulation. Neither of the two completely hypophysectomized rabbits copulated with the males, and neither showed ovulation points. Of the thirteen partially hypophysectomized animals, two accepted both males on the first occasion, and one copulated with both males on the second occasion. These three (Nos. 38, 26, 28) showed multiple ovulation points 48 hr later. They had pars distalis remnants above 30 % of normal volume. None of the other animals in this group copulated and none of them showed any ovulation points. All except two of the ten rabbits which did not copulate had pars distalis remnants below 30 % of normal volume. Body weight Graphical examination of the body weight and pars distalis volume in ten normal rabbits revealed no correlation between these two parameters. This was also the case with the partially hypophysectomized animals. Over the period of the experiments, the normal animals showed a mean increase in body weight of 27 %. Three completely hypophysectomized animals showed a mean decrease in body weight of 14 % between operation and death. The partially hypophysectomized animals showed a mean increase in body weight of 5% between operation and death. DISCUSSION Results have been presented in which morphological, histological and physiological criteria of the target glands are related to the pars distalis volume in partially hypophysectomized animals. It is necessary to examine the degree of reliance that may be placed upon the pars distalis relative volume figures assigned to each animal in the series. The normal range of pars distalis volume was found to be mm3. This would suggest that the absolute volume determined in the partially hypophysectomized series might be difficult to relate to the individual animal's preoperative pars distalis volume. The relative volume figures given in the text and tables are based upon the normal mean. This is considered acceptable since (a) the normal mean had a low standard

11 404 H. J. CAMPBELL error, indicating that most normal animals had a pars distalis volume close to the normal mean, and (b) it is unlikely that any systematic results would have been obtained if the experimental relative volume figures were subject to a twofold variation. Another factor to be considered is the possibility of regeneration of pars distalis tissue. Mitotic figures in the intact pars distalis of several species have been described (Cameron, 1952). More recently, Daniel & Prichard (1958) have observed mitoses in the pars distalis region a few days after pituitary stalk section in the goat. In this species and in the sheep (Daniel & Prichard, 1957) pituitary stalk section appears to be followed in a short time (24-48 hr) by almost total necrosis of the pars distalis, only a thin rim of surviving tissue remaining. Animals which survived for about a month were found to have distinctly thicker rims of living pars distalis tissue, and the possibility is raised that regeneration may have occurred during this month. However, in the present series, animals with pars distalis volumes at death of 1, 2, 3 and 5% of normal, had lived postoperatively 165, 101, 97 and 191 days, respectively. In view of these small volume figures and long survival times, it is felt that regeneration of pars distalis tissues did not occur to any relevant extent in the present series of animals. If regeneration did occur, then the figures here quoted for size of anterior pituitary remnant during the tests are higher than they should be. It will be convenient to discuss the results for each target gland separately before considering the over-all findings. Removal of 99% of the pars distalis did not abolish the response of the adrenal cortex to stress; removal of 90 % did not reduce its ability to maintain a normal ascorbic acid concentration; removal of 90% did not disturb the normal histological appearance of the adrenal cortex; but removal of 70% of the pars distalis produced a highly significant reduction in adrenal weight. Since many reports (for example, White, 1933; McPhail, 1935; Cutuly, 1936) have established the fact that adrenal atrophy after hypophysectomy is limited to the cortex, it seems that partial hypophysectomy has a differential effect upon the size and function of the adrenal cortex. Aschner (1912) and Smith (1932) found that 30% or more of the pars distalis was necessary to maintain normal adrenal weight in the rat. Ganong & Hume (1956) report no loss of adrenal weight in the dog after removal of 95 % on the anterior pituitary. However, Ganong & Hume (1956) appear to have used only one animal for the determination of normal pars distalis volume, and estimated their volumes by a method that is not clear from their account. The present findings with regard to the histology and functional capacity of the adrenals are essentially in accord with the observations of Ganong & Hume (1956). The most likely explanation of the differential effect of partial hypophysectomy upon adrenocortical weight and function is felt to reside in a greater sensitivity of the weight-maintaining mechanism to reduction in blood levels of ACTH. It is improbable that the

12 PARTIAL H YPOPH YSECTOM Y 405 phenomenon is due to the existence of more than one kind of ACTH (Sulman, 1952), since graded partial hypophysectomy would not be expected to reduce selectively the output of such multiple hormones. Removal of 90 % or more of the pars distalis results in a profound reduction in the release and uptake of radio-iodine by the thyroid, and in the height of thyroid epithelial cells. These parameters and also the thyroidal responses to stress remained normal with pars distalis volumes greater than 10 % of normal. However, there was no significant difference in thyroid weight both in the animals with more and in those with less than 10% of normal pars distalis volume. Thus partial hypophysectomy seems to have a differential effect upon thyroid weight and function, the latter being more sensitive. This effect is the reverse of that reported by Greer (1952, 1957) who found that rats bearing ocular pituitary transplants exhibited thyroid atrophy in the presence of approximately normal thyroidal radio-iodine metabolism. These findings led Greer to postulate the existence of two kinds of TSH. This concept and the finding on which it is based have been criticized by Halmi, Spirtos, Bogdanove & Lipner (1953), Vanderlaan & Caplan (1954) and Bogdanove & Bogdanove (1957). The present results are in agreement with the views of these dissenters. In the present study it was found that the slope of the release curve did not appreciably alter with pars distalis volumes ranging from about 10 to 100 % of normal. There seems no reason to assume that significant variation in hormone content occurred in the thyroids of these animals, since (a) they were all kept under the same controlled conditions of diet, room temperature and light, (b) their thyroid weights were distributed over a small range, and (c) their radio-iodine uptakes also cover a small range. It might therefore reasonably be concluded that graded reduction of pars distalis volume from 100 to 10 % does not result in graded reduction in thyroid activity. It evidently does not cause graded reduction in radio-iodine release, or in the height of the epithelial cells of the thyroid. This lack of graded response is thought to be most likely explained by an amount of TSH-producing tissue in the normal pituitary about 10 times in excess of that required for basal function. Removal of 70% of the pars distalis resulted in abolition of the post-coital ovulation response, marked reduction in ovarian follicular diameter and a significant loss of ovarian weight. These parameters are normal with pars distalis volumes greater than 30 % of normal. Also, in confirmation of the observation of Smith (1932) in the rat, 10% of normal pars distalis volume maintained ovarian weight above the level of hypophysectomized controls, but produced disturbance of the histological characters. The present findings also confirm the morphological and histological data presented on the ovary by Ganong & Hume (1956). The present results indicate that as pars distalis volume is gradually reduced

13 406 H. J. CAMPBELL ovarian dysfunction first appears (30 % remnant), then there is loss of both ovarian and thyroidal function (10% remnant), and finally, with exceedingly small (less than 1 %) remnants there is also loss of adrenal function. Similar results with regard to morphological and histological characters were obtained by Smith (1932) in rats, and by Breckenridge & Keller (1948) and Keller, Lawrence & Blair (1945) in dogs. Progressive pituitary disease in humans is frequently first manifested by a depression in gonadal function, followed by symptoms of thyroid and adrenal insufficiency (Peters, German, Mann & Welt, 1954; Sheehan & Summers, 1949). The functional tests carried out by Ganong & Hume (1956) in the partially hypophysectomized dog yielded a similar spectrum of endocrine dysfunction. One explanation of this spectrum involves anterior pituitary architectonics. It is well established in several species that the pars distalis is not homogeneous; concentrations of the main cell-types are found in various regions. It is possible that the peripheral-lying 'gonadotrophs' are removed first at parapharyngeal hypophysectomy, and that the more centrally placed 'thyrotrophs' are next withdrawn, so that with small remnants there may be no basophilic elements present and therefore no gonadal or ovarian function. Ganong & Hume (1956) claim that cytological examination of the small remnants revealed the presence of all three types of pars distalis cell. In the present work, trichromatic staining of 10,u sections, cut from the haematoxylin and eosin de-stained 200,u sections, produced results that were far from satisfactory. But it did appear that all three cell types were present in even the smallest remnant. It seems, then, that the spectrum of endocrine dysfunction seen after graded partial hypophysectomy is not to be explained on the basis of pituitary cell disposition-though this conclusion must be regarded as tentative pending proper cytological studies. Another possible explanation is that the parameters used in evaluating a particular target gland are more sensitive than those used in evaluating another gland. It is difficult to believe that the release and uptake of radio-iodine by the thyroid gland is less sensitive than the criteria used here to assess adrenocortical activity. With regard to the ovary, it has to be admitted that post-coital ovulation is only one aspect of the ovarian functional complex. But the ovulation results are supported by the morphological and histological findings in the ovary. In general the spectrum of malfunction revealed in the present series, which is a pattern found by other workers under somewhat similar conditions, is supported in the case of each target gland by several criteria. It is therefore believed that the spectrum is not to be explained on the basis of choice of parameters. A third possibility is that there exists a difference in the sensitivity of the target glands to their trophic hormones. This possibility can be tested adequately only when pure preparations of the trophic hormones are available and when their activities can be expressed in terms of weight rather than arbitrary 'units'.

14 PARTIAL H YPOPH YSECTOM Y 407 A fourth possibility, and the one considered most likely, is that there are quantitative differences in the various hormones released from the pars distalis. This might be due to (a) varying proportions of the cell types, each type having roughly equal activity, or (b) differing resting levels of activity among the cell types. No evidence appears to have been recorded with regard to the second possibility. Rasmussen (1933) found that in the pars distalis of normal man the proportions of chromophobes, acidophiles and basophiles were 52, 37 and 11, respectively. Nelson (1935) in the normal rat found the series to be 42, 50 and 8, respectively. Similar findings are reported in the papers by Wolfe and his colleagues (see Wolfe & Brown, 1942). These several results suggest that the basophilic elements in most species are in the great minority. Since 'gonadotrophs' and 'thyrotrophs' are found among the basophiles, it seems likely that gonadal and thyroidal functions would be lost before adrenocortical function after reduction of pars distalis volume. No evidence of this kind is yet available to suggest that gonadal function would be lost before thyroidal. SUMMARY 1. Adult female chinchilla rabbits were subjected to graded partial hypophysectomy. Morphological, histological and physiological criteria of ovarian, thyroidal and adrenocortical activity were correlated with the volume of the pars distalis remnant. 2. The percentages of normal pars distalis volume required to maintain normal target gland weight were these: ovary and adrenal, 30 %; thyroid, 10 %. The ovary and adrenal weights were maintained above the level of completely hypophysectomized controls by 10% of normal pars distalis volume. 3. Normal histological features of the target glands were maintained by the following percentages of normal pars distalis volume: ovary, 30 %; thyroid, 1O%; adrenal, 5 %. 4. The post-coital ovulation response required 30 % of normal pars distalis volume. Normal uptake and release of radio-iodine and normal responses to emotional and physical stresses by the thyroid were all maintained by 10% of normal pars distalis volume. The adrenocortical lymphopenic and ascorbic acid depletion responses to emotional and physical stresses were maintained by 1% of normal pars distalis volume. 5. As pars distalis volume was reduced, the ovary first showed loss of function. With greater reduction, both ovary and thyroid were malfunctional. Reduction down to 1% of normal did not abolish adrenocortical function. This spectrum of malfunction may be due to different proportions of the various cell types of the pars distalis. 6. Graded partial hypophysectomy had no significant effect on body weight. 27 PHYSIO. CXLIX

15 408 H. J. CAMPBELL The author, and the paper, owe much to the constant help and advice provided so willingly by Professor G. W. Harris, F.R.S. Technical assistance was expertly rendered by Messrs W. F. Piper, S. A. Williams and F. Smith. This work was accepted by the University of London as part of the requirements for the Ph.D. degree. REFERENCES AscHNma, B. (1912). t'ber die Funktion der Hypophyse. PflUg. Arch. gem. Physiol. 65, 341. BARRNETT, R. J. & GREEP, R. 0. (1951). Pituitary gonadotrophic activity of stalk-sectioned male rats. Endocrinology, 49, 337. BOGDANOVE, E. M. & BOGDANOvE, J. M. (1957). Thyroid weight, microscopic structure and radioiodine concentration at various levels of thyrotrophin secretion in the goitrogen-treated rat. Endocrinology, 61, BRECKENRIDGE, C. G. & KELLER, D. (1948). Retention of sex functions after isolation of the pars anterior by extirpation of the hypophysial stalk. Amer. J. Physiol. 152, BRowN-GRANT, K., EULER, C., VON HARRIs, G. W. & REICHLIN, S. (1954). Measurement and experimental modification of thyroid activity in the rabbit. J. Phy8iol. 126, BRowN-GRANT, K., HARRIS, G. W. & REICHLcN, S. (1954). The effect of emotional and physical stress on thyroid activity in the rabbit. J. Physiol. 126, BROWN-GRANT, K., HARRIS, G. W. & REIcISLN, S. (1957). The effect of pituitary stalk section on thyroid function in the rabbit. J. Physiol. 136, CAMPBELL, H. J. & HARRS, G. W. (1957). The volume of the pituitary and median eminence in stalk-sectioned rabbits. J. Physiol. 136, CAMERON, G.R. (1952). The Pathology of the Cell. London: Oliver and Boyd. COLFER, H. F., DE GROOT, J. & HARRIS, G. W. (1950). Pituitary gland and blood lymphocytes. J. Physiol.111, CUTULY, E. (1936). Quantitative study on the adrenals of hypophysectomized rats. Anat.Rec. 66, DANIEL, P. M. & PRICHARD, M. M. L. (1957). Anterior pituitary necrosis in the sheep produced by section of the pituitary stalk. Quart. J. exp. Physiol. 42, DANIEL, P. M. & PRICHARD, M. M. L. (1958). The effects of pituitary stalk section in the goat. Amer. J. Path. 34, FORTIEr, C., HARRs, G. W. & McDONALD, I.R. (1957). The effect of pituitary stalk section on the adrenocortical response to stress in the rabbit. J. Physiol. 136, GANONG, W. F. & HUME, D. M. (1956). The effect of graded hypophysectomy on thyroidal, gonadal and adrenocortical function in the dog. Endocrinology, 59, GREER, M. A. (1952). The role of the hypothalamus in the control of thyroid function. J.clin. Endocrin. 12, GREER, M. A. (1957). The influence of the central nervous system on the control of thyrotrophin secretion. Ciba Colloquium on Endocrinology, 10, HALMI, N. S., SPIRTOS, B. N., BOGDANOvE, E. M. & LIPNER, H. J. (1953). A study of various influences on the iodide concentrating mechanisms of the rat thyroid. Endocrinology, 52, HAIS, G. W. (1948). Electrical stimulation of the hypothalamus and the mechanism of neural control of the adenohypophysis. J. Physiol. 107, JAcOBSOIIN, D. & WESTMAN, A. (1940). A parapharyngeal method of hypophysectomy in the rabbit. Acta physiol. 8cand. 1, KELLER, A. D., LAWRENCE, W. E. & BTLAR, C. B. (1945). Effects of varying degrees of hypophysectomy in the dog. Arch. Path. (Lab. Med.), 40, McPHAL, M. K. (1935). Hypophysectomy of the cat. Proc. Roy. Soc. B, 117, NELSON, W. 0. (1935). Concerning the anterior pituitary-gonadal interrelations. Endocrinology, 19, PETERS, J. F., GERMAN, W. J., MANN, E. B. & WELT, L. G. (1954). Functions of gonads, thyroids and adrenals in hypopituitarism. Metaboli8m, 3, RASMUSSEN, A. T. (1933). The percentage of different types of cells in the anterior lobe of the hypophysis in the adult human female. Amer. J. Path. 9, SHEEHAN, H. L. & SUMMERS, L. K. (1949). The syndrome of hypopituitarism. Quart. J. Med. 18,

16 PARTIAL HYPOPHYSECTOMY 409 SMrrH, P. E. (1932). The secretory capacity of the anterior hypophysis as evidenced by the effect of partial hypophysectomy in rats. Anat. Rec. 52, SuIJT1, F. G. (1952). Mechanism of the chromatophorotrophic effect of ACTH. J. Endocrin. 8, VANDERLAAN, W. B. & CAPLAN, R. (1954). Observations on a relationship between total thyroid iodine content and the iodide-concentrating mechanisms of the thyroid gland of the rat. Endocrinology, 54, WHITE, W. E. (1933). The effect of hypophysectomy of the rabbit. Proc. Roy. Soc. B, 114, WOLFE, J. M. & BROWN, A. D. (1942). Action of diethyl-stilboestrol on the cytological characteristics of female rats, together with certain observations on the effect of castration. Endocrinology, 31,