Molecular and Cellular Endocrinology, 41 (1986) 59-10 Elsevier Scientific Publishers Ireland, Ltd.

59

MCE 01512

Parap~ysectomy-induced stimulation of parathyroid glands in mature frogs ( Rana catesbeiana) : evidence for telencephalic regulation

of parathyroid gland function

Michael P. Sarras, Jr. *, Dennis R. Trune **, Elaine M. Merisko, Floyd M. Foltz and Stanley R. Nelson

Department of Anatomy/Division of Celt Biology, University of Kansas Medical Center, 39th and Rainbow Boulevard: Kansas City, KS 66103 (U.S.A.]

(Received 6 March 1986; accepted 6 May 1986)

Key words: amphibian endocrine; Ca*+ regulation.

Summary

The relationship of the paraphyseal-choroid plexus complex to parathyroid gland function was investigated in adult frogs. Light microscopy and mo~homet~c analysis indicated that total parathyroid gland volume, cell volume and vascular volume doubled by 7-28 days after surgical removal of the paraphyseal-choroid plexus complex (paraphysectomy). This increase correlated with the appearance of (1) large Golgi-associated vesicles, (2) an increase in the apparent number of cytoplasmic dense-core granules, and (3) PTH within the parenchymal cells as monitored by immunofluorescence. Twelve months after paraphys~tomy, parathyroid glands became cystic with a central ~uid-filled cavity surrounded by a stratified cuboidal cell layer. The parenchymal cells of cystic glands contained numerous cytoplasmic dense-core granules and were also positive for PTH. Radioimmunoassay of cystic parathyroid fluid indicated a PTH concentration of 2 pg/pl; however, analysis by SDS-PAGE indicated a wide range of proteins in cystic fluid. The results of this study indicate that paraphysectomy induces st~ulation of the parathyroid glands and suggest a role for the paraphyseal-choroid plexus complex in the regulation of amphibian parathyroid gland function.

Introduction

The paraphysis cerebri is a compound tubular gland of the posterior telencephalon which is fused

* Address correspondence and reprint requests to: Dr. Michael P. Sarras, Jr., Department of Anatomy/Division of Cell Biology, University of Kansas Medical Center, 39th and Rainbow Boulevard, Kansas City, KS 66103, U.S.A.

** Present address: Department of Cell Biology and Anat- omy, The Oregon Health Science University, Portland, OR 97201, U.S.A.

with the choroid plexus of the third ventricle (Selenka, 1891; Kappers, 1956). The paraphyseal choroid plexus complex is most developed in amphibians and reptiles, but is also briefly present during mammalian embryonic development (Kappers, 1955). In adult frogs the paraphyseal- ohoroid plexus complex extends from the roof of the third ventricle into the ventricular chamber. Although the precise function of the paraphysis is unknown, we have previously shown that excision of the paraphyseal-choroid plexus complex from adult frogs usually results in hyperextension of

0303-7207/84/$03.50 0 1986 Elsevier Scientific Publishers Ireland, Ltd.

60

limbs, tetanic convulsions, and death (Nelson and Foltz, 1983). Animals which survive paraphysec- tomy (excision of both the paraphysis and at- tached choroid plexus) develop cystic parathyroid glands within 4-12 months (Nelson and Foltz, 1983). Paraphysectomy of tadpoles does not result in death but does result in bone abnormalities and cystic parathyroid glands in froglets (Nelson et al., 1985). These observations suggest that the paraphyseal-choroid plexus complex may be in- volved in systemic Ca*’ regulation. To better evaluate the relations~p of the paraphyseal- choroid plexus complex to calcium regulating sys- tems, paraphysectomy-induced changes in the parathyroid glands of adult frogs were studied. A preliminary report of these studies has previously been published in abstract form (Sarras et al., 1984).

Materials and methods

Animals and surgical procedures Adult frogs (Rana catesbeiana) of both sexes

were purchased from William A. Lemberger As- soc. (Germantown, WI) and house in metal tanks containing sand and water maintained at a depth of 5-8 cm. Tanks were placed near windows to expose frogs to outdoor light. Frogs were fed l-2 newborn rats (5-7 g) every 10 days. Paraphysec- tomy was performed on 12 frogs. After 7-28 days (short-interval) or 12 months (long-interval) parathyroid glands were removed and utilized for morphological or biochemical analyses as de- scribed below.

Paraphys~tomy was performed as described previously (Nelson and Foltz, 1983). Briefly, animals were anesthetized with a 50% solution of ethyl carbamate (urethane), 5 ml/kg, i.p. After application of be~~onium chloride (1: 750) to the skin, a midline scalp incision was made to expose the skull. Using a dental burr, the dura was exposed over the third ventricle and the paraphy- seal-choroid plexus complex was removed with fine-tipped forceps. Controls were similarly treated but the paraphyseal-choroid plexus complex was not removed.

Light mi&~oscopy and rn~~p~ornet~ Morphometry was used to compare the

parenchymal and stromal cell volumes of parathyroid glands of experimental (5 frogs: 7, 9, 10, 11 and 28 days after p~aphys~tomy) and control (4 sham-operated) frogs. Four parathyroid glands were removed from each frog and processed for either light microscopy (2 glands) or electron microscopy (2 glands) as described below. For mo~homet~ and light microscopy, glands were fixed by immersion in 3% paraformaldehyde con- taining 0.1 M phosphate buffer, pH 7.4 for 24 h at 4°C. The tissue was extensively washed in phos- phate-buffered amp~bi~‘s saline (PBS, contain- ing 111 mM NaCl, 1.88 mM KCl, 0.82 mM CaCl,, 2.38 mM NaHCO,, 0.07 mM NaHPO, and 11.1 mM glucose) and deionized-distilled water, and then directly processed into JB-4 embedding medium (Polysciences, War~ngton, PA). Morpho- metric analysis was carried out on glands which were serially sectioned (5.0 pm sections) on glass knives and stained with hematoxylin and eosin.

Camera lucida drawings of serial sections were made using a Nikon microscope. Drawings were made of 3 tissue compartments which included: (1) total gland area, (2) cellular areas, and (3) vascular structures. Using a digitizer and a Radio Shack TRS-80, Model III microcomputer, a mor- phometric program was used to determine the total volume of each of these compartments. The volumes of each compartment for experimental and control glands were statistically compared using a Wilcoxon Rank Sum test. Photomicro- graphs were obtained using a Leitz photomicro- scope.

Indirect immunofluorescent techniques were used to analyze changes in the cellular content of parathyroid hormone (PTH) of experimental and control tissue. Parathyroid glands from 2 animals of the short-interval group (7 days post- paraphysectomy), long-interval group (12 months post-paraphysectomy), and control group (sham- operated) were fixed and processed into JB-4 em- bedding medium as described above. Sections (2 pm) were quenched with 10 mM glycine, washed with PBS, and incubated at room temperature for 2 h with an antibody (diluted 1 : 10 in PBS) di- rected against the amino terminal fragment of human PTH (Hruska et al., 1975). After the in-

61

cubation, tissue sections were washed in PBS and incubated for 1 h with a secondary antibody con- jugated to fluorescein (Miles Laboratories, Naper- ville, IL). After a final wash in PBS, sections were mounted in a 10% glycerol-PBS solution, and viewed with a Leitz fluorescence photomicroscope. For controls, samples were incubated either in the

absence of the primary antibody or with the

primary antibody preabsorbed with purified

bovine PTH.

Electron microscopy

To study ultrastructural changes associated with

paraphysectomy, 2 parathyroid glands from the

short-interval group (same frogs used for the mor-

phometry study), long-interval group (12 months post-paraphysectomy) and control group were

placed in Kamovsky’s fixative (1965) for 24 h at 4°C and processed for electron microscopy. Ultra- thin sections were stained with uranyl acetate and lead citrate, and then viewed on a JOEL 100s electron microscope.

Radioimmunoassay (RIA) of PTH

PTH content of cystic parathyroid glands was determined by RIA analysis. Both cystic fluid and

acid-treated cellular homogenates of cystic glands were analyzed. RIA was performed using an anti- body directed against an amino terminal fragment

of human PTH (the same antibody used for the fluorescence studies). Purified bovine PTH was used as a standard and the 1-34 amino terminal

fragment of bovine PTH was used as the iodinated

tracer. The lower limit of sensitivity for this assay

was 0.2 ng of PTH.

SDS-polyacrylamide gel electrophoresis (SDS-

PAGE)

The spectrum of proteins and peptides present in cystic fluid was compared to that present in the

cells of cystic and normal glands using SDS-PAGE (Maizel, 1971). The samples (50 pg of protein) (Lowry et al., 1951) were boiled for 3 min in Tris-PO, buffer, pH 6.7, containing 2% SDS, 5% P-mercaptoethanol and 10% glycerol, and were electrophoresed on a gradient gel (5-20%) at 5 mA for 16 h. Samples and molecular weight stan-

dards (Bio-Rad Laboratories, Richmond, CA) were visualized by the silver-staining technique of Switzer et al. (1979) and Merril et al. (1979).

Results

Light microscopy and morphometric analysis

As shown in Fig. 1, parathyroid glands of con- trol adult frogs consist of a central parenchyma

surrounded by a highly vascularized stroma. The parenchyma of control frogs contains densely

packed cells of a fusiform shape (Fig. la).

Parathyroid glands from the short-interval group

(7-28 days post-paraphysectomy) showed signifi-

cant changes as compared to the control group. The most notable change was a dilation of the

stromal vasculature surrounding the parenchyma (Fig. lb). In comparison to control values, our

morphometric analysis indicated a 2-fold increase

in the experimental vascular compartment by 7-28 days post-paraphysectomy (0.0066 mm3 vs.

0.00325 mm3). While no change in cell mor-

phology was noted by light microscopy at this time, the total gland volume and volume of the

cellular compartment increased by a factor of 2.6 and 2.5, respectively (0.1598 mm3 vs. 0.0625 mm3

and 0.1462 mm3 vs. 0.0592 mm3). As previously shown by Nelson et al. (1985), cystic parathyroid

glands from the long-interval group (12 months

post-paraphysectomy) consisted of a fluid-filled central cavity surrounded by a stratified cuboidal layer of parenchymal cells (Fig. lc) and were approximately 8 times larger in diameter than normal glands.

Electron microscopy

At the ultrastructural level, a number of changes

were observed in the parenchymal cells after

paraphysectomy. For the short-term interval group (7-28 days post-paraphysectomy), the most sig- nificant alterations were noted in the Golgi re-

gions. In contrast to controls (Fig. 2a, 2b and 2c),

by 7 days post-paraphysectomy numerous Golgi- associated vesicles (0.8-1.6 pm diameter) contain- ing an amorphous material (Fig. 3a, 3b, 3c and

Fig. 4) were observed. Also observed throughout the cytoplasm was an apparent increase in the

number of dense-core granules with a diameter of 0.2-0.6 pm (Fig. 3b and Fig. 4). For the long-in- terval group (12 months post-paraphysectomy) cells contained dense-core granules which were more numerous than those observed in the short- interval group but lacked the Golgi-associated vesicles (Fig. 5a, 5b and 5~).

62

Fig. 1. Light microscopy of frog parathyroid glands at various times after paraphysectomy. Glands from control frogs contain a fusiform parenchyma surrotmded by a vascularized connective tissue (et) capsule (Fig. la). In Fig. la, arrows indicate blood vessels in the connectivel tissue capsule. By 7 days ~st-p~aphys~tomy, a dilation of blood vessels (bv) in the connective tissue (ct) capsule is observed (Fig. lb). As shown in Fig. lc, cystic parathyroid glands (12 months post-paraphysectomy) are composed of a stratified cuboidal epithelium which surrounds a fluid-filled central cavity (cc). Cystic glands also have a connective tissue (ct) capsule. All figures, X 500.

Fig. 2. Electron microscopy of control frog parathyroid glands. At low (Fig. 2~) and intermediate magnification (Fig. 26)

parenchymal cells are shown to have numerous mitochondria and RER distributed throughout the cytoplasm, as well as perinuclear Golgi (arrows, Fig. 2b). Perinuclear Golgi (go) profiles are better seen at higher magnification in Fig. 2~. Nuclei (n) appear spherical

or elongated depending on the plane of section. Fig. 2a, x 5000; Fig. 2b, X 15000; Fig. 2c, X 57 500.

Fig. 3. Electron microscopy of frog parathyroid glands at 7 days post-paraphysectomy. Paraphysectomy induces the appearance of

Go&i-associated vesicles which are indicated by arrows in Fig. 30 and b. Go& are indicated by double arrowheads in Fig. 36. In

addition, paraphysectomy also induces an apparent increase in the number of smaller dense-core granules which are shown in Fig. 3b by single arrowheads. A higher magnification of these granules is shown in Fig. 4. A higher magnification .of Golgi (go)-associated

vesicles (v) containing an amorphous material is shown in Fig. 3c. Fig. 3~2, X5000; Fig. 3b, X15000; Fig. 3c, X57500.

65

Fig. 4. Electron micrograph of a Golgi region at 7 days post-paraphysectomy. The Golgi stacks are demarcated by the arrows.

Golgi-associated dense-core granules (arrowhead) and vesicles (v) containing an amorphous material are observed following

paraphysectomy. X 46000.

Immunofluorescence and RIA of cellular and cystic ated vesicles and the apparent increase in the fluid PTH number of dense-core granules correlated with the

To determine if the appearance of Golgi-associ- presence of FTH, indirect immunofluorescence was

66

Fig. 5. Electron microscopy of cystic parathyroid glands (12 months post-paraphysectomy). Long-term paraphysectomy correlates

with the presence of numerous dense-core granules (arrowheads) which are dispersed throughout the cytoplasm. These granules vary

in size between 0.2 and 0.6 pm and were often Go&j-associated (arrow, Fig. Sb). These dense-core granules are better seen in Fig. 5~.

Both small (0.2 Pm, arrowhead) and large (0.6 pm. arrow) granules are shown in Fig. 5~. Fig. 5a, x 5000; Fig. 56, ~1.5000; Fig. SC, x57500.

performed. As shown in Fig. 6a and 6b, PTH could not be detected in control glands. By 7 days post-paraphysectomy, however, PTH could be de- tected in parathyroid parenchymal cells (Fig. 6c and 6d). Compared to the short-interval group, a relatively higher level of PTH was present in cells of the cystic parathyroid glands, 12 months post- paraphysectomy (Fig. 6e and 6f). A fluorescent signal appearing as scattered aggregates of amorphous material was observed in the cystic fluid cavity (data not shown). This apparent ag- gregation of PTH may have occurred during tissue fixation. The presence of PTH was confirmed by preabso~tion of the ~tibody with purified bovine PTH (Fig. 6g and 6h).

Although the levels of PTH in cystic parathyroid tissue were below the range of sensitivity for our RIA, analysis of cystic fluid indicated the pres- ence of PTH at a concentration of 2 pg/pl. Based on cystic fluid volume this represented 200 pg of PTH per gland.

SDS-PAGE analysis of normal and cystic parathyroid giands

The proteins present in the fluid of cystic parathyroid glands were analyzed by SDS-PAGE. As shown in Fig. 7, electrophoretic analysis dem- onstrated that a wide range of proteins were pres- ent in cystic fluid. The most prominent proteins in the cystic fluid were detected in the high molecu- lar weight range (M, 70000-200000), with the most concentrated protein having an apparent MW of 70000 (indicated by an arrowhead in Fig. 6, lane 3). A lightly stained protein was detected in the lower molecular weight range (M, 14000) which had a mobility similar to mammalian PTH (electrophoretic mobility of purified bovine PTH indicated by the arrow in Fig. 6, lane 3). Many proteins appearing in cystic cells (Fig. 7, lane 2) were not detected in the cystic fluid (Fig. 7, lane 3) as mo~tored by the silver-staining technique used in this study.

Discussion

The observations reported in this paper indi- cate that paraphysectomy of adult frogs results in stimulation of the parathyroid glands. As shown by light microscopy and morphometry, a doubling

in the volume of the cellular compartment was observed by 7-28 days post-paraphysectomy. Taken in concert with previous studies which re- port an increase in parathyroid dry weight within 6-12 months post-p~aphysectomy (Nelson et al., 1985), the present data are consistent with the occurrence of cellular hyperplasia as well as cellu- lar hypertrophy. Paraphysectomy-induced stimu- lation correlated not only with an increase in total cell volume but also with an increase in vascular volume.

The ultrastructural and immunofluorescent studies in paraphysectomized frogs indicate in- creased p~ath~oid cell activity which is con- sistent with an increase in the cellular synthesis of PTH. As early as 7 days post-paraphysectomy, the appearance of Golgi-associated vesicles and an apparent increase in the number of cytoplasmic dense-core granules was observed. Immunofluo- rescent studies demonstrated that these morpho- logical changes correlated with the detection of PTH within the parenchymal cell. Due to inter- ference with serum proteins, we were unable to measure PTH in the blood of post-paraphysecto- mized frogs using our RIA procedure. Immuno- fluorescent and RIA studies did demonstrate the presence of PTH in cystic parathyroid glands (long-interval group, 12 months post-paraphys~- tomy). The dense-core granules observed in amphibian cells resembled those observed in mammalian parathyroid glands. Similar to mammals (Nunex et al., 1974), granules were in the range of 0.2-0.6 pm in diameter and pleio- morphic in appearance. There was an apparent increase in the number of dense-core granules observed in cells of the long-interval group (12 months,post-p~aphysectomy) and this correlated with an apparent higher content of PTH based on the immunofluorescent studies. The appearance of smaller secretory granules with a diameter of ap- pro~mately 0.1 pm has been reported to accu- mulate in activated parathyroid glands of mammals exposed to low levels of Ca2’ (Roth and Raisz, 1966). It should be noted, however, that the presence of secretory granules is species depen- dent and their absence may not indicate cellular inactivity (see review by Cohn and MacGregor, 1981). Our data does not directly address the question as to whether systemic PTH levels in-

68

69

crease following paraphysectomy. The occurrence of bone resorption following paraphysectomy (Nelson et al., 1985) is consistent, however, with an increase in systemic PTH levels. Unique to this study was the appearance of Golgi-associated vesicles filled with an amorphous material in cells of the short-interval group (7-28 days post- paraphysectomy). These vesicles were much larger than the dense-core granules (OS-l.6 pm vs. 0.2-0.6 pm) and were almost always associated with the Golgi. It is possible that these vesicles are involved in PTH processing and storage since

Fig. 6. Immunofluorescence microscopy of PTH in parathyroid

glands of control and paraphysectomized frogs. Fluorescence

micrographs are shown on the left (Fig. 6a, c, e, g) and corre-

sponding phase micrographs are shown on the right (Fig.

66, d, f, h). Arrows indicate corresponding regions in fluores-

cence and phase micrographs. PTH cannot be detected in

parenchymal cells of control glands (Fig. 6a and 6). By 7 days

post-paraphysectomy, however, PTH can be localized within

parenchymal cells (Fig. 6c and d). In Fig. 6c and d parenchymal cells surrounding a blood vessel (bv) are shown.

A relatively higher content of PTH is indicated for parenchymal

cells of cystic parathyroid glands (12 months post-paraphysec-

tomy) as shown in Fig. 6e and f (compare to Fig. 6c and d).

Although not shown in Fig. 6f, fluorescent signals appearing

as scattered aggregates of amorphous material were observed

in the cystic fluid cavity. This was likely due to aggregation of

PTH during tissue fixation and processing. Localization of

PTH is abolished in parenchymal cells of cystic glands follow-

ing preabsorption of the primary antibody with purified bovine

PTH (Fig. 6g and h). As shown in Fig. 6e, f, g and h, parenchymal cells of cystic parathyroid glands surround a

fluid-filled central cavity (cc). All figures, x 2000.

Fig. 7. SDS-PAGE of control and cystic parathyroid glands.

Total cell homogenate of control glands is shown in lane 1,

while total cell homogenate of cystic glands (12 months post-

paraphysectomy) is shown in lane 2. Cystic gland fluid is

shown in lane 3. The cellular protein pattern of control and

cystic glands is similar. The most prominent proteins present

in cystic parathyroid gland fluid were detected in the high

molecular weight range (Mr 70000-200000). The arrow indi-

cates the position of purified bovine PTH when run as a molecular weight standard. The arrowhead indicates the posi-

tion of the most concentrated protein of cystic fluid which had

a M, 70000. Molecular weight markers are indicated on the

left and include: myosin, 200000; P-galactosidase, 116250;

phosphorylase B, 92 500; bovine serum albumin, 66200;

ovalbumin, 45000; carbonic anbydrase, 31000; soybean tryp-

sin inhibitor, 21500; lysozyme, 14400.

2(

1

30k-

16k-

92.5k-

66.2k-

45k-

31k-

14.4k-

70

similar structures have been reported in mam- malian parathyroid glands treated so as to disrupt the Gob&related packaging process (Chu et al., 1974; MacGregor et al., 1977). The relationship of these vesicles to PTH synthesis awaits EM im- munocytochemical and autoradiographic studies.

Our SDS-PAGE studies indicated that a broad range of proteins were present in cystic parathyroid fluid. These proteins ranged in molecular weight from less than 14400 to greater than 200000, with the most relatively concentrated proteins in the 70000-200000 range. The lightly stained protein shown in Fig. 6, lane 3 which had a mobility similar to the bovine PTH standard may represent amphibian PTH; however, confirmation of this awaits Western blot analysis. While the RIA data indicated a high concentration of PTH in the cystic fluid (2.0 pg/pl), the SDS-PAGE studies indicated that a protein with an apparent molecu- lar weight of 70000 was the most concentrated, relative to all other cystic fluid proteins. The iden- tity of this protein is not known, but its mobility corresponds to mammalian secretory protein I (SP-I) glycoprotein (M, 70000) which has been co-localized with PTH within secretory granules (Cohn and MacGregor, 1981). These relative con- centration differences may simply be due to selec- tive proteolysis within the cystic fluid since PTH fragments would be detected by RIA but would run with the dye front during SDS-PAGE. On the other hand, it is interesting to speculate that hy- perstimulation of the cystic parathyroid gland cells has caused a biosynthetic shift in favor of SP-I glycoprotein. Analysis of this problem will require radiotracer pulse-chase immunoprecipitation stud- ies. Nevertheless, the fact that many proteins pres- ent in cystic parathyroid cells and normal cells were absent in cystic fluid, indicates that the pres- ence of these proteins in fluid is not solely the result of cell lysis.

Although our data suggests a functional re- lationship between the paraphyseal-choroid plexus and the parathyroid gland, the mechanism(s) of this relationship is not understood. The control of mammalian parathyroid function is believed to be a direct function of circulating Ca*’ levels (see review by Cohn and MacGregor, 1981). Although amphibian parathyroid glands respond to low levels of systemic Ca*+ (Contellyou and McWhin- nie, 1967) our data suggests that additional fac-

tar(s) from the paraphyseal-choroid plexus com- plex are involved in the regulation of this endo- crine gland. Studies are now underway to identify secretory products of the paraphyseal-choroid plexus complex and determine their possible role in Ca*+ regulation and metabolism.

Acknowledgements

The authors wish to thank Dr. Ronal R. Mac- gregor of the Veterans Administration Medical Center, Kansas City, MO, for kindly providing the antibody used in this study and for conducting the PTH radioimmunoassays. In addition, the authors also wish to thank Larry Nolar, Jacquelyn Huff, Lisa Brown, Airlean Bowls and Rosetta Barkley for their assistance with light microscopy. The authors also thank Ms. Brenda Tennyson-Jackson for her help in the preparation of this manuscript. This research was supported in part from funds from the Amyotrophic Lateral Sclerosis Society of America (S.R.N.) and from the USPHS, NIH Grant AM33362 (M.P.S.).

References

Chu, L.L.H., MacGregor, R.R., Hamilton, J.W. and Cohn,

D.V. (1974) Endocrinology 95, 1431-1438.

Cohn, D.V. and MacGregor, R.R. (1981) Endocrine Rev. 2,

l-26.

Cortelyou, J.R. and McWhinme, D.J. (1967) Am. Zool. 7,

843-855.

Hruska, K.A., Kopelman, R., Rutherford, W.E., KIahr, S. and

Slatopolsky, E. (1975) J. Clin. Invest. 56, 39-58.

Kappers, J.A. (1955) J. Comp. Neurol. 102, 425-510.

Kappers, J.A. (1956) Experientia 12, 187-188

Kamovsky, M.J. (1965) J. Cell Biol. 27, 137A.

Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J.

(1951) J. Biol. Chem. 193, 265-275.

Maizel, J.V. (1971) In: Methods in Virology, Vol. 5, Eds.: K.

Maromorosh and H. Koprowski (Academic Press, New

York) pp. 179-246.

McGregor, D.H., Chu, L.L.H., MacGregor, R.R. and Cohn,

D.V. (1977) Am. J. Pathol. 87, 553-568.

Merril, C.R., Switzer, R.C. and VanKeuren, M.L. (1979) Proc.

Natl. Acad. Sci. U.S.A. 76, 4335-4339.

Nelson, S.R. and Foltz, F.M. (1983) Exp. Neurol. 79, 763-772.

Nelson, S.R., Foltz, F.M., Camarata, P. and Sarras, Jr., M.P.

(1985) Anat. Rec. 211, 311-317.

Nunez, E.A., Hedhammar, A., Wu, F.M., Whalen, J.C. and

Krook, L. (1974) Lab. Invest. 31, 96-108.

Roth, S.I. and Raisz, L.G. (1966) Lab. Invest. 15, 1187-1211.

Sarras, Jr., M.P., Trune, D.R., Foltz, F.M. and Nelson, S.R. (1984) Fed. Proc. 43, 943.

Selenka, E. (1891) Biol. Zbl. 10, 323-326.

Switzer, R.C., Merril, C.R. and Shifrin, S. (1979) Anal. Bio- them. 98, 231-237.