Molecular and Cellular
139
Endocrinology, 12 (1978) 139-149 0 Elsevier/North-Holland Scientific Publishers, Ltd.
ROLES OF GROWTH HORMONE AND TESTOSTERONE OF MOUSE KIDNEY GLUCURONIDASE
IN THE SYNTHESIS
*
Richard T. SWANK Department of Molecular Biology, Roswell Park Memorial Institute, 660 Elm Street, Buffalo, New York 14263, U.S.A.
Received 7 April 1978; accepted 30 June 1978
Mouse kidney p-glucuronidase production is under multihormonal control. In normal mice, kidney glucuronidase is induced over lOO-fold by testosterone. However, hypophysectomy reduces this induction to about 5% of normal. This loss in inducibility was in part restored by growth hormone. Simultaneous administration to hypophysectomized female mice of growth hormone and testosterone, but not of prolactin and testosterone, restored kidney glucuronidase concentration to half that found in testosterone-treated normal female mice. Growth hormone alone had no effect in hypophysectomized females nor did it enhance glucuronidase activity in testosterone-treated normal females. Radiolabeling experiments demonstrated that the enhancement by growth hormone of glucuronidase activity was accompanied by a corresponding increase in its rate of synthesis. Kidney hypertrophy and kidney glucuronidase production may be under common hormonal regulation. Testosterone or growth hormone treatment alone of hypophysectomized mice had little or no effect on either process, but combined treatment with the two hormones significantly enhanced both. The rate of synthesis of kidney glucuronidase is controlled by the Cur gene. Relative differences in kidney glucuronidase synthesis in mice of different Cur genotype were maintained in testosterone-treated hypophysectomized mice. This suggests that control of glucuronidase synthesis by the Cur locus is exerted by interaction with androgens rather than pituitary products. Keywords:
multihormonal
control; kidney hypertrophy;
Cur gene; lysosome.
Mouse kidney glucuronidase is a model system for the study of the genetic regulation of enzyme synthesis by steroid hormones (Paigen et al., 1975; Swank et al., 1978). The Cur gene located on chromosome 5 in the region of the glucuronidase structural gene controls the testosterone-mediated rate of synthesis of glucuronidase by a cis-acting mechanism (Swank et al., 1973). Mice of homozygous Curala genotype synthesize the enzyme at 3 times the rate of Gz&lb mice. More
* Supported by USPHS Research Grant GM-19521.
R. T. Swank
140
recently, new phenotypes affecting glucuronidase inducibility have been described (Swank et al., 1978). The hormonal requirements for kidney glucuronidase induction have been partially defined. Administration of androgens to female mice greatly enhances the activity and rate of synthesis of kidney glucuronidase (Fishman, 1965; Swank et al., 1973, 1978). The cytosolic androgen receptor protein is necessary for glucuronidase induction, since testosterone does not induce the enzyme (Dofuku et al., 1971) in tfm (testicular feminization) mice which have depressed levels of this receptor (Bullock and Bardin, 1972; Gehring and Tomkins, 1974; Attardi and Ohno, 1974). Also, pituitary products are required for normal glucuronidase induction, since hypophysectomized mice synthesize the enzyme at only 5% of the normal rate after testosterone treatment (Swank et al., 1977). Glucuronidase induction is confined to the proximal tubule epithelial cells (Fishman et al., 1969). In males or testosterone-treated females these cells are larger, and there are significant increases in the activity of several other kidney enzymes (Swank et al., 1978). In these studies we present evidence that growth hormone is an important requirement for normal testosterone-mediated glucuronidase synthesis and proximal tubule hypertrophy. We also show that the disparate rates of testosteronemediated glucuronidase synthesis in Gz.@la and Gurbb mice are maintained in the absence of pituitary hormones.
MATERIALS
AND METHODS
Mice Mice were either female inbred Balb/cStCrlIBr or female outbred CD-l(ICR)BR from Charles River Mouse Farms, Inc. The Balb/c strain was used in most experiments since high rates of kidney glucuronidase synthesis were obtained after only 3 days of hormonal treatment. To obtain comparable rates in CD-1 mice required at least 8 days’ treatment. Mice were hypophysectomized by Charles River Mouse Farms, Inc. at 8-10 weeks of age and experiments were initiated at 7-10 days post-hypophysectomy. Drinking water of hypophysectomized mice was supplemented with 5% glucose and cotton was added to cages for warmth. All mice were fed solid pellets of Teklad Mouse Breeder Diet (Teklad Mills, Winfield, Iowa) containing 16.5% minimum crude protein, 10% minimum crude fat and 2.5% maximum crude fiber. Hormorte treatment and tissue preparation Testosterone pellets (35 mg), prepared by Dr. Bhogi B. Sheth of the Department of Pharmaceutics of the University of Tennessee, were implanted subcutaneously and simultaneously with initial growth hormone and/or prolactin injections. About 0.3 mg testosterone was absorbed each day to provide a saturating level of
Growth hormone and mouse kidney glucuronidase
141
hormone. Bovine growth hormone (NIH-GH-B18; 0.81 IU/mg) and bovine prolactin (NIHP-B4; 18.5 IU/mg) were obtained through the National Pituitary Agency, NIAMDD. Both hormones were reconstituted in saline at 1 mg/ml final concentration by adjusting to pH 9.0-9.5 with 0.1 M NaOH, and then readjusting to pH 7.4 with 0.5 M sodium phosphate buffer. Growth hormone and prolactin were injected daily intraperitoneally at 9 a.m. and 5 p.m. at a dose of 100 pg/20 g body weight for each injection unless otherwise indicated. All animals were sacrificed after 72 h of hormone or saline treatment unless otherwise indicated. Mice were killed by anoxia with carbon dioxide gas. Tissue homogenates were prepared and stored at -20°C. &Glucuronidase (EC 3.2.1.31) was assayed by a fluorometric method using 4-methylumbelliferyl-/3-D-glucuronide as substrate (Brandt et al., 1975). Rate of synthesis of glucuronidase Each mouse received 100-300 PCi L-[4,S-3H]leucine (Amersham/Searle, TRK, 170: 54 Ci/mmol) intraperitoneally and was sacrificed 1 h later. Rates of synthesis of glucuronidase were measured by specific antibody precipitation of enzyme from kidney homogenates which had been partially purified by heat treatment, as previously described (Cleveland and Swank, 1978). Counts incorporated into the 73,000 MW glucuronidase subunit were determined after electrophoresis of the immunoprecipitate on sodium dodecyl sulfate (SDS)-polyacrylamide gels. Samples were counted in a Nuclear Chicago/Isocap 300 scintillation counter with a 36% efficiency and a background of 17 cpm.
RESULTS Control of glucuronidase activity and synthetic rate As previously reported (Swank et al., 1977) hypophysectomy does not affect basal kidney glucuronidase activity (saline-treated mice) but greatly depresses testosterone-mediated kidney glucuronidase induction (fig. 1). By day 3 of testosterone treatment there was a 6-fold increase in kidney glucuronidase activity in normal Balb/c mice but no significant increase in hypophysectomized mice. Growth hormone alone had no effect on enzyme activity in either normal or hypophysectomized mice. However, in hypophysectomized mice treated simultaneously with both growth hormone and testosterone, glucuronidase activity was about 3-fold higher than in hypophysectomized mice given testosterone only. Growth hormone treatment thus restored testosterone-mediated glucuronidase inducibility to about 50% of that observed in normal mice. In normal mice, treatment with both hormones did not further augment the potent induction resulting from treatment with testosterone alone. Simultaneous treatment of hypophysectomized mice with prolactin and testosterone did not significantly increase glucuronidase activity over that of testoster-
R. T. Swank
142
R=
HYPox
q =Normal
Testostemne
GH
Fig. 1. Kidney glucuronidase activity in hormone-treated mice. All assays were performed on individual Balb/c mice except in the case of saline-treated mice where assays were performed on two kidney extracts, each obtained by combining kidneys from 3 mice, and in testosterone f GH + PRL mice where kidney extracts from 3 mice were combined for assay. Bars indicate standard errors. The number of determinations on individual mice or groups of mice is indicated in parentheses. Hypox = hypophysectomized.
one-treated mice. Also, combined treatment with testosterone, growth hormone and prolactin did not amplify the induction caused by testosterone plus growth hormone. A possible mechanism for the increased kidney glucuronidase activity elicited by treatment of hypophysectomized mice with both growth hormone and testosterone is an increased rate of synthesis of the enzyme. There was in fact an increased rate of incorporation of radiolabeled leucine into the glucuronidase 73,000 MW subunit, purified from kidney by immunoprecipitation and SDSpolyacrylamide gel electrophoresis, in mice treated with both hormones as compared with testosterone treatment alone (fig. 2). Treatment of hypophysectomized mice with both hormones did not, however, fully restore the rate of synthesis of glucuronidase observed in normal mice treated with testosterone. Like kidney glucuronidase activity, the rates of glucuronidase synthesis measured by the imm~oprecipitation technique were similar in saline-treated normal and hypophysectomized mice (fig. 3). Likewise, the large (60-fold) increase in rate of glucuronidase synthesis after testosterone treatment observed in normal mice was depressed to a 5fold increase in hypophysectomized mice. Growth hormone injection of hypophysectomized mice did not increase the synthetic rate observed with saline. Most significantly, simultaneous treatment with growth hormone and testosterone increased the ~ucuronidase synthetic rate in hypo-
Growth hormone and mouse kidney glucuronidase 1
I
1
I
143
I
400
300
$200 u:
loo
Fig. 2. Synthesis of the glucuronidase subunit in hormone-treated normal and hypophysectomized mice. Each Balb/c mouse was injected intraperitoneally with 100 &i [3H]leucine at 3 days after the start of hormone treatments. 1 h after administration of radiolabeled leucine, kidney glucuronidase immunoprecipitates were prepared from individual mice and electrophoresed on SDS-polyacrylamide gels. Fractions of the gel in the region of the hypophysectomized mouse + testosterone; 73,000 MW subunit are displayed. l-, X-X, hypophysectomized mouse + testosterone + growth hormone; o----o, normal mouse + testosterone.
physectomized mice 5-6-fold (P < 0.001) as compared with testosterone treatment alone, so that it was restored to about 50% of that of testosterone-treated normal mice. Also, as in glucuronidase activity measurements, prolactin did not affect either the testosterone or testosterone plus growth hormone mediated increase in glucuronidase synthetic rate. Glucuronidase synthetic rate was dependent on growth hormone dose between 20 and 200 pg hormone per 20 g body weight, the amount used in these experiments (fig. 4). Doubling the dose to 400 pg caused no further stimulation. A very similar dose-response curve (not shown) was obtained for enhancement of glucuronidase activity. Control of kidney hypertrophy Another testosterone-mediated effect on mouse kidney that is greatly depressed by hypophysectomy is hypertrophy of proximal tubule cells, and this cellular process is likewise regulated by growth hormone (fig. 5).
R. T. Swank
m=Hypox 0
=Normol
(21
101.~ El
Saline
.1 I-@
GH
I Testospcme GH
+ PRL
Testosterone G+ A.
Fig. 3. Rates of kidney gfucuronidase synthesis in hormone-treated mice. All Balb/c mice were treated with saline or hormones for 3 days before determination of synthetic rate. Salinetreated mice and growth-hormone-treated mice received 300 PCi [3H]leucine/mouse and kidney extracts from 3 mice were combined for each immunoprecipitation. In all other treatments, 100 &i [3H]leucine/mouse was administered and rates of synthesis determined on individual mice except in the case of the testosterone + GH + PRL treatment where a single assay was performed on the combined kidney extracts of 3 mice, The relative rate of kidney glucuronidase synthesis is the ratio of counts incorporated into the 73,000 MW glucuronidase subunit peak on the polyacrylamide gel divided by the counts incorporated into total kidney TCA-precipitable protein. There was no significant difference in total kidney protein synthesis among these groups of mice. Bars indicate standard errors and the number of determinations on individual Balb/c mice or groups of mice is indicated in parentheses.
In normal mice there was a 23% increase (P< 0.001) in relative kidney weight over saline controls after testosterone treatment, but no significant increase (P> 0.1) in mice treated with growth hormone alone. Also, combined treatment with testosterone and growth hormone did not significantly affect (P> 0.05) the hypertrophy observed with testosterone treatment alone. Relative kidney weight of saline-treated hypophysectomized mice was only 80% of that of normal mice and there was no hypertrophy after either testosterone or growth hormone treatment. The hypertrophy was, however, restored in hypophysectomized mice treated simultaneously with both testosterone and growth hormone (33% increase over saline treatment; P < 0.01). The increase in kidney weight/20 g body weight in mice treated with testosterone and growth hormone
145
Growth hormone and mouse kidney glucuronidase
0
KKJ Growth
Ho-
200
300
400
CwlZOqbodyW+)
Fig. 4. Effect of growth hormone dose on rate of synthesis of kidney glucuronidase. All Balb/c mice were treated with testosterone. Saline or growth hormone in 0.1 ml was injected in two equal doses per day to give the indicated total daily dose. Rates of synthesis were determined in individual mice 1 h after injection of 100 nCi [3H]leucine per mouse after 3 days of hormone treatment. Each point is the mean f SEM of 3-4 determinations.
(31
n
= HYPOX
El=Ncfmol .32 % 2
.30
8
.2e
it
Fig. 5. Kidney determinations
(6)
hypertrophy on individual
in hormone-treated mice. Balb/c mice are indicated.
Standard
errors
and
the
number
of
146
R.T. Swank I
,
1
I
T
I
4
6 DAYS
I
GurO’a-Normol
-/
8
.
IO
12
14
INDUCED
Fig. 6. Time course of induction of kidney glucuronidase synthesis by testosterone in normal and hypophysectomized Cur-a/a (Balb/c) and Curb/b (CD-l) mice. All mice were pelleted at day 0 with testosterone. Uninduced mice (day 0) were injected with 300 &i [3H]leucine/ mouse and induced mice received 100 pCi/mouse. After 1 h, kidney homogenates from 3 mice were combined and rates of synthesis determined as described in fig. 3.
compared to saline controls was, in fact, similar in hypophysectomized and normal mice. The apparent small increase in kidney weight of hypophysectomized mice treated with testosterone and prolactin was not significant (P> 0.1) nor did prolactin affect kidney hypertrophy mediated by combined treatment with testosterone and growth hormone. Hypophysectomy and the Gur gene After testosterone treatment normal GuPla mice such as Balb/c synthesize glucuronidase at 3-7 times the rate of normal Gurbb mice such as CD-1 resulting in proportionate increases in kidney glucuronidase concentrations in GuF+la mice (Swank et al., 1973) (fig. 6). Because the Gur gene regulates ghrcuronidase synthesis in kidney of testosterone-treated mice but not in kidney of saline-treated mice or in other organs, it has been postulated to be a site on chromosome 5 which interacts with the androgen-receptor complex. Since these experiments have demonstrated that both androgens and pituitary hormones are necessary for maximal glucuronidase synthesis, an important question in regard to the mechanism of action of the Gur locus is whether it might, in fact, be responding to pituitary products rather than to androgens. It was found that hypophysectomy did not affect glucuronidase synthesis in
Growth hormone and mouse kidney glucuronidase
147
mice not treated with testosterone, but greatly depressed synthesis in mice of both genotypes over the 14-day time course of testosterone treatment (fig. 6). However, the differential rate of glucuronidase synthesis in mice of the two Gur genotypes was maintained in hypophysectomized mice. That is, the ghrcuronidase synthetic rate in hypophysectomized Guiala mice was always greater (1 O-l 5 fold) than that of hypophysectomized Gz4rbD mice after testosterone treatment. Likewise, after hypophysectomy, kidney glucuronidase activity (not shown) was always 6-10 times greater in Balb/c than in CD-l mice. These results suggest that the genetically controlled difference in hormone-mediated glucuronidase synthesis in these mice is not mediated by pituitary products.
DISCUSSION Kidney glucuronidase synthesis is under multihormonal control with two of the major hormones being androgen and growth hormone. Testosterone administration to hypophysectomized mice yielded a rate of synthesis 5--10% of that of similarly treated normal mice. However, simultaneous administration of testosterone and growth hormone to hypophysectomized mice boosted the rate of synthesis to 50% of that in normal mice. On the other hand, the related pituitary hormone, prolactin, had no effect on kidney glucuronidase production. Thus growth hormone serves to facilitate the action of testosterone. It does not, however, act independently of testosterone. Growth hormone treatment of hypophysectomized mice did not increase kidney glucuronidase synthetic rate or activity over that in saline-treated controls in these short-term 3-day experiments with Balb/c mice. Similarly, growth hormone alone did not increase kidney glucuronidase activity in longer 7-day experiments with hypophysectomized CD-1 mice (Swank, unpublished). Also, growth hormone treatment did not further accelerate glucuronidase activity in testosterone-treated normal mice. Growth hormone also causes about 50% restoration of testosterone-mediated kidney glucuronidase synthesis in hypophysectomized CD-1 mice (Bullock, in discussion of Swank et al., 1978; Swank, unpublished). Additional evidence that growth hormone is involved in kidney glucuronidase production derives from the noninducibility (Swank et al., 1978) of Snell’s dwarf (dw/dw) mutant mice which lack several anterior pituitary hormones including growth hormone (Lewis, 1967). Also, testosterone-mediated kindey glucuronidase inducibility is reduced about 50% in mutant little @it/lit) mice (Swank et al., 1978) which have a relatively specific defect of serum growth hormone (Either and Beamer, 1976). Restoration of glucuronidase synthetic rate by injected growth hormone was incomplete even at double the dose of growth hormone routinely used in these experiments. While this could be due to the artificial nature of the injection protocol, a more likely explanation is that other pituitary products are required. Preliminary experiments suggest that thyroxine and glucocorticoids, administered
148
R. T. Swank
together with growth hormone and testosterone, restore glucuronidase production to near-normal levels (Swank, unpublished). It is unlikely that levels of kidney cytosolic testosterone receptor protein or of active testosterone metabolite(s) are being modulated by pituitary removal since several other kidney enzymes are equally inducible by testosterone in normal and hypophysectomized mice (Swank et al., 1977). We cannot, of course, exclude the possibility that there are specific requirements for testosterone metabolites among the kidney enzymes normally inducible by testosterone. Kidney hypertrophy appears to be under similar hormonal control as kidney glucuronidase synthesis. That is, neither testosterone nor growth hormone alone affected relative kidney weight in hypophysectomized mice. However, simultaneous administration of testosterone and growth hormone caused as large an increase in kidney weight as that observed in testosterone treatment of normal mice. As in the case of glucuronidase production, prolactin had no effect on kidney hyper trophy. An apparent dissimilarity in glucuronidase synthesis and kidney hypertrophy is that while testosterone treatment of hypophysectomized mice caused a 6-fold increase in rate of glucuronidase synthesis, it had no measurable effect on kidney hypertrophy. There was likewise little or no increase in kidneyglucuronidase activity in these same mice. These apparent differences may, however, simply reflect the sensitivity of measurement of the various responses. For example, 3-day testosterone treatment of normal mice causes a SO-fold increase in rate of synthesis but only a 6-fold and 1.3-fold increase in activity and hypertrophy, respectively. Mouse kidney hypertrophy is known to include significant increases in the rates of protein and RNA synthesis (Kochakian, 1969). While not directly measured in these experiments, it seems likely that these metabolic processes also require the combined action of testosterone and growth hormone. These experiments suggest that the Gur locus does not regulate glucurinodase production by interaction with pituitary products since the regulatory difference is maintained in the absence of these products. Final proof of this proposal will require comparison of the rate of synthesis of glucuronidase in congenic hypophysectomized mice which are known to differ only at the Gur locus. However, these results are consistent with the hypothesis that the Gur locus on chromosome 5 may be a genomic site of interaction with the testosterone-receptor complex (Swank et al., 1978). Growth hormone has also been implicated in the induction of other specific gene products such as rat liver, kidney and adrenal ornithine decarboxylase (Janne and Raina, 1969; Nicholson et al., 1977) and in maintaining the normal rate of synthesis of rat liver serum albumin (Feldhoff et al., 1977). It likewise acts in concert with androgens, glucocorticoids and thyroid hormones in the induction of rat liver a,,-globulin (Roy, 1973) where it increases levels of translatable mRNA (Roy and Dowbenko, 1977). Widnell and Tata (1966) observed that growth hormone and testosterone stimulate deoxyribonucleic acid-dependent ribonucleic acid polymerase of rat liver nuclei in an additive manner. How several hormones
Growth hormone and mouse kidney glucuronidase
149
interact to regulate specific gene expression in these and other systems (Martial et al., 1977; Kwitz and Feigelson, 1977; Oka and Perry, 1974) is an important but still incompletely understood problem.
ACKNOWLEDGEMENTS The author thanks Gerald Jahreis for exellent technical assistance. Purified growth hormone and prolactin preparations were generously supplied by the National Pituitary Agency, NIAMDD.
REFERENCES Attardi, B. and Ohno, S. (1974) Cell 2, 205-212. Brandt, E.J., Elliott, R.E. and Swank, R.T. (1975) J. Cell Biol. 67, 774-788. Bullock, L.P. and Bardin, C.W. (1972) 3. Clin. Endocrinol. Metab. 35,935-937. Cleveland, C.E. and Swank, R.T. (1978) Biochem. J. 170,249-256. Dofuku, R., Tettenborn, U. and Ohno, S. (1971) Nature New Biol. 232,5-7. Either, E.M. and Beamer, W.G. (1976) J. Hered. 67,87-91. Ernest, M.J., Chen, C.-L. and Feigelson, P. (1977) J. Biol. Chem. 252,6783-6791. Feldhoff, R.C., Taylor, J.M. and Jefferson, L.S. (1977) J. Biol. Chem. 252, 3611-3616. Fishman, W.H. (1965) In: Methods in Hormone Research, Vol. IV, Part B, Ed.: R.I. Dorfman (Academic Press, New York) pp. 2733326. Fishman, W.H., Ide, H. and Rufo, R. (1969) Histochemie 20, 287-299. Gehring, U. and Tomkins, G.M. (1974) Cell 3,59-64. Janne, J. and Raina, A. (1969) Biochim. Biophys. Acta 174,769-772. Kochakian, C.D. (1969) Steroids 14,77-90. Kurtz, D.T. and Feigelson, P. (1977) Proc. Natl. Acad. Sci. (U.S.A.) 74,4791-4795. Lewis, U.J. (1967) Mem. Sot. Endocrinol. 15, 179-191. Martial, J.A., Seeburg, P.H., Guenzi, D., Goodman, H.M. and Baxter, J.D. (1977) Proc. Natl. Acad. Sci. U.S.A. 74,4293-4295. Nicholson, W.E., Barton, R.N., Holladay, L.A., Orth, D.N. and Puett, D. (1977) Endocrinology 100,459-467. Oka, T. and Perry, J.W. (1974) J. Biol. Chem. 249,7647-7652. Paigen, K., Swank, R.T., Tomino, S. and Ganschow, R.E. (1975) J. Cell. Physiol. 85, 379-392. Roy, A.K. (1973) J. Endocrinol. 56,295-301. Roy, A.R. and Dowbenko, D.J. (1977) Biochemistry 16,3918-3922. Swank, R.T., Paigen, K. and Ganschow, R.E. (1973) J. Mol. Biol. 81,225-243. Swank, R.T., Davey, R., Joyce, L., Reid, P. and Macey, M.R. (1977) Endocrinology 100, 473-480. Swank, R.T., Paigen, K., Davey, R., Chapman, V., Labarca, C., Watson, G., Ganschow, R., Brandt, E.J. and Novak, E. (1978) Recent Prog. Horm. Res. 34, in Press. Widnell, CC. and Tata, J.R. (1966) Biochem. J. 98,621-628.
139
Endocrinology, 12 (1978) 139-149 0 Elsevier/North-Holland Scientific Publishers, Ltd.
ROLES OF GROWTH HORMONE AND TESTOSTERONE OF MOUSE KIDNEY GLUCURONIDASE
IN THE SYNTHESIS
*
Richard T. SWANK Department of Molecular Biology, Roswell Park Memorial Institute, 660 Elm Street, Buffalo, New York 14263, U.S.A.
Received 7 April 1978; accepted 30 June 1978
Mouse kidney p-glucuronidase production is under multihormonal control. In normal mice, kidney glucuronidase is induced over lOO-fold by testosterone. However, hypophysectomy reduces this induction to about 5% of normal. This loss in inducibility was in part restored by growth hormone. Simultaneous administration to hypophysectomized female mice of growth hormone and testosterone, but not of prolactin and testosterone, restored kidney glucuronidase concentration to half that found in testosterone-treated normal female mice. Growth hormone alone had no effect in hypophysectomized females nor did it enhance glucuronidase activity in testosterone-treated normal females. Radiolabeling experiments demonstrated that the enhancement by growth hormone of glucuronidase activity was accompanied by a corresponding increase in its rate of synthesis. Kidney hypertrophy and kidney glucuronidase production may be under common hormonal regulation. Testosterone or growth hormone treatment alone of hypophysectomized mice had little or no effect on either process, but combined treatment with the two hormones significantly enhanced both. The rate of synthesis of kidney glucuronidase is controlled by the Cur gene. Relative differences in kidney glucuronidase synthesis in mice of different Cur genotype were maintained in testosterone-treated hypophysectomized mice. This suggests that control of glucuronidase synthesis by the Cur locus is exerted by interaction with androgens rather than pituitary products. Keywords:
multihormonal
control; kidney hypertrophy;
Cur gene; lysosome.
Mouse kidney glucuronidase is a model system for the study of the genetic regulation of enzyme synthesis by steroid hormones (Paigen et al., 1975; Swank et al., 1978). The Cur gene located on chromosome 5 in the region of the glucuronidase structural gene controls the testosterone-mediated rate of synthesis of glucuronidase by a cis-acting mechanism (Swank et al., 1973). Mice of homozygous Curala genotype synthesize the enzyme at 3 times the rate of Gz&lb mice. More
* Supported by USPHS Research Grant GM-19521.
R. T. Swank
140
recently, new phenotypes affecting glucuronidase inducibility have been described (Swank et al., 1978). The hormonal requirements for kidney glucuronidase induction have been partially defined. Administration of androgens to female mice greatly enhances the activity and rate of synthesis of kidney glucuronidase (Fishman, 1965; Swank et al., 1973, 1978). The cytosolic androgen receptor protein is necessary for glucuronidase induction, since testosterone does not induce the enzyme (Dofuku et al., 1971) in tfm (testicular feminization) mice which have depressed levels of this receptor (Bullock and Bardin, 1972; Gehring and Tomkins, 1974; Attardi and Ohno, 1974). Also, pituitary products are required for normal glucuronidase induction, since hypophysectomized mice synthesize the enzyme at only 5% of the normal rate after testosterone treatment (Swank et al., 1977). Glucuronidase induction is confined to the proximal tubule epithelial cells (Fishman et al., 1969). In males or testosterone-treated females these cells are larger, and there are significant increases in the activity of several other kidney enzymes (Swank et al., 1978). In these studies we present evidence that growth hormone is an important requirement for normal testosterone-mediated glucuronidase synthesis and proximal tubule hypertrophy. We also show that the disparate rates of testosteronemediated glucuronidase synthesis in Gz.@la and Gurbb mice are maintained in the absence of pituitary hormones.
MATERIALS
AND METHODS
Mice Mice were either female inbred Balb/cStCrlIBr or female outbred CD-l(ICR)BR from Charles River Mouse Farms, Inc. The Balb/c strain was used in most experiments since high rates of kidney glucuronidase synthesis were obtained after only 3 days of hormonal treatment. To obtain comparable rates in CD-1 mice required at least 8 days’ treatment. Mice were hypophysectomized by Charles River Mouse Farms, Inc. at 8-10 weeks of age and experiments were initiated at 7-10 days post-hypophysectomy. Drinking water of hypophysectomized mice was supplemented with 5% glucose and cotton was added to cages for warmth. All mice were fed solid pellets of Teklad Mouse Breeder Diet (Teklad Mills, Winfield, Iowa) containing 16.5% minimum crude protein, 10% minimum crude fat and 2.5% maximum crude fiber. Hormorte treatment and tissue preparation Testosterone pellets (35 mg), prepared by Dr. Bhogi B. Sheth of the Department of Pharmaceutics of the University of Tennessee, were implanted subcutaneously and simultaneously with initial growth hormone and/or prolactin injections. About 0.3 mg testosterone was absorbed each day to provide a saturating level of
Growth hormone and mouse kidney glucuronidase
141
hormone. Bovine growth hormone (NIH-GH-B18; 0.81 IU/mg) and bovine prolactin (NIHP-B4; 18.5 IU/mg) were obtained through the National Pituitary Agency, NIAMDD. Both hormones were reconstituted in saline at 1 mg/ml final concentration by adjusting to pH 9.0-9.5 with 0.1 M NaOH, and then readjusting to pH 7.4 with 0.5 M sodium phosphate buffer. Growth hormone and prolactin were injected daily intraperitoneally at 9 a.m. and 5 p.m. at a dose of 100 pg/20 g body weight for each injection unless otherwise indicated. All animals were sacrificed after 72 h of hormone or saline treatment unless otherwise indicated. Mice were killed by anoxia with carbon dioxide gas. Tissue homogenates were prepared and stored at -20°C. &Glucuronidase (EC 3.2.1.31) was assayed by a fluorometric method using 4-methylumbelliferyl-/3-D-glucuronide as substrate (Brandt et al., 1975). Rate of synthesis of glucuronidase Each mouse received 100-300 PCi L-[4,S-3H]leucine (Amersham/Searle, TRK, 170: 54 Ci/mmol) intraperitoneally and was sacrificed 1 h later. Rates of synthesis of glucuronidase were measured by specific antibody precipitation of enzyme from kidney homogenates which had been partially purified by heat treatment, as previously described (Cleveland and Swank, 1978). Counts incorporated into the 73,000 MW glucuronidase subunit were determined after electrophoresis of the immunoprecipitate on sodium dodecyl sulfate (SDS)-polyacrylamide gels. Samples were counted in a Nuclear Chicago/Isocap 300 scintillation counter with a 36% efficiency and a background of 17 cpm.
RESULTS Control of glucuronidase activity and synthetic rate As previously reported (Swank et al., 1977) hypophysectomy does not affect basal kidney glucuronidase activity (saline-treated mice) but greatly depresses testosterone-mediated kidney glucuronidase induction (fig. 1). By day 3 of testosterone treatment there was a 6-fold increase in kidney glucuronidase activity in normal Balb/c mice but no significant increase in hypophysectomized mice. Growth hormone alone had no effect on enzyme activity in either normal or hypophysectomized mice. However, in hypophysectomized mice treated simultaneously with both growth hormone and testosterone, glucuronidase activity was about 3-fold higher than in hypophysectomized mice given testosterone only. Growth hormone treatment thus restored testosterone-mediated glucuronidase inducibility to about 50% of that observed in normal mice. In normal mice, treatment with both hormones did not further augment the potent induction resulting from treatment with testosterone alone. Simultaneous treatment of hypophysectomized mice with prolactin and testosterone did not significantly increase glucuronidase activity over that of testoster-
R. T. Swank
142
R=
HYPox
q =Normal
Testostemne
GH
Fig. 1. Kidney glucuronidase activity in hormone-treated mice. All assays were performed on individual Balb/c mice except in the case of saline-treated mice where assays were performed on two kidney extracts, each obtained by combining kidneys from 3 mice, and in testosterone f GH + PRL mice where kidney extracts from 3 mice were combined for assay. Bars indicate standard errors. The number of determinations on individual mice or groups of mice is indicated in parentheses. Hypox = hypophysectomized.
one-treated mice. Also, combined treatment with testosterone, growth hormone and prolactin did not amplify the induction caused by testosterone plus growth hormone. A possible mechanism for the increased kidney glucuronidase activity elicited by treatment of hypophysectomized mice with both growth hormone and testosterone is an increased rate of synthesis of the enzyme. There was in fact an increased rate of incorporation of radiolabeled leucine into the glucuronidase 73,000 MW subunit, purified from kidney by immunoprecipitation and SDSpolyacrylamide gel electrophoresis, in mice treated with both hormones as compared with testosterone treatment alone (fig. 2). Treatment of hypophysectomized mice with both hormones did not, however, fully restore the rate of synthesis of glucuronidase observed in normal mice treated with testosterone. Like kidney glucuronidase activity, the rates of glucuronidase synthesis measured by the imm~oprecipitation technique were similar in saline-treated normal and hypophysectomized mice (fig. 3). Likewise, the large (60-fold) increase in rate of glucuronidase synthesis after testosterone treatment observed in normal mice was depressed to a 5fold increase in hypophysectomized mice. Growth hormone injection of hypophysectomized mice did not increase the synthetic rate observed with saline. Most significantly, simultaneous treatment with growth hormone and testosterone increased the ~ucuronidase synthetic rate in hypo-
Growth hormone and mouse kidney glucuronidase 1
I
1
I
143
I
400
300
$200 u:
loo
Fig. 2. Synthesis of the glucuronidase subunit in hormone-treated normal and hypophysectomized mice. Each Balb/c mouse was injected intraperitoneally with 100 &i [3H]leucine at 3 days after the start of hormone treatments. 1 h after administration of radiolabeled leucine, kidney glucuronidase immunoprecipitates were prepared from individual mice and electrophoresed on SDS-polyacrylamide gels. Fractions of the gel in the region of the hypophysectomized mouse + testosterone; 73,000 MW subunit are displayed. l-, X-X, hypophysectomized mouse + testosterone + growth hormone; o----o, normal mouse + testosterone.
physectomized mice 5-6-fold (P < 0.001) as compared with testosterone treatment alone, so that it was restored to about 50% of that of testosterone-treated normal mice. Also, as in glucuronidase activity measurements, prolactin did not affect either the testosterone or testosterone plus growth hormone mediated increase in glucuronidase synthetic rate. Glucuronidase synthetic rate was dependent on growth hormone dose between 20 and 200 pg hormone per 20 g body weight, the amount used in these experiments (fig. 4). Doubling the dose to 400 pg caused no further stimulation. A very similar dose-response curve (not shown) was obtained for enhancement of glucuronidase activity. Control of kidney hypertrophy Another testosterone-mediated effect on mouse kidney that is greatly depressed by hypophysectomy is hypertrophy of proximal tubule cells, and this cellular process is likewise regulated by growth hormone (fig. 5).
R. T. Swank
m=Hypox 0
=Normol
(21
101.~ El
Saline
.1 I-@
GH
I Testospcme GH
+ PRL
Testosterone G+ A.
Fig. 3. Rates of kidney gfucuronidase synthesis in hormone-treated mice. All Balb/c mice were treated with saline or hormones for 3 days before determination of synthetic rate. Salinetreated mice and growth-hormone-treated mice received 300 PCi [3H]leucine/mouse and kidney extracts from 3 mice were combined for each immunoprecipitation. In all other treatments, 100 &i [3H]leucine/mouse was administered and rates of synthesis determined on individual mice except in the case of the testosterone + GH + PRL treatment where a single assay was performed on the combined kidney extracts of 3 mice, The relative rate of kidney glucuronidase synthesis is the ratio of counts incorporated into the 73,000 MW glucuronidase subunit peak on the polyacrylamide gel divided by the counts incorporated into total kidney TCA-precipitable protein. There was no significant difference in total kidney protein synthesis among these groups of mice. Bars indicate standard errors and the number of determinations on individual Balb/c mice or groups of mice is indicated in parentheses.
In normal mice there was a 23% increase (P< 0.001) in relative kidney weight over saline controls after testosterone treatment, but no significant increase (P> 0.1) in mice treated with growth hormone alone. Also, combined treatment with testosterone and growth hormone did not significantly affect (P> 0.05) the hypertrophy observed with testosterone treatment alone. Relative kidney weight of saline-treated hypophysectomized mice was only 80% of that of normal mice and there was no hypertrophy after either testosterone or growth hormone treatment. The hypertrophy was, however, restored in hypophysectomized mice treated simultaneously with both testosterone and growth hormone (33% increase over saline treatment; P < 0.01). The increase in kidney weight/20 g body weight in mice treated with testosterone and growth hormone
145
Growth hormone and mouse kidney glucuronidase
0
KKJ Growth
Ho-
200
300
400
CwlZOqbodyW+)
Fig. 4. Effect of growth hormone dose on rate of synthesis of kidney glucuronidase. All Balb/c mice were treated with testosterone. Saline or growth hormone in 0.1 ml was injected in two equal doses per day to give the indicated total daily dose. Rates of synthesis were determined in individual mice 1 h after injection of 100 nCi [3H]leucine per mouse after 3 days of hormone treatment. Each point is the mean f SEM of 3-4 determinations.
(31
n
= HYPOX
El=Ncfmol .32 % 2
.30
8
.2e
it
Fig. 5. Kidney determinations
(6)
hypertrophy on individual
in hormone-treated mice. Balb/c mice are indicated.
Standard
errors
and
the
number
of
146
R.T. Swank I
,
1
I
T
I
4
6 DAYS
I
GurO’a-Normol
-/
8
.
IO
12
14
INDUCED
Fig. 6. Time course of induction of kidney glucuronidase synthesis by testosterone in normal and hypophysectomized Cur-a/a (Balb/c) and Curb/b (CD-l) mice. All mice were pelleted at day 0 with testosterone. Uninduced mice (day 0) were injected with 300 &i [3H]leucine/ mouse and induced mice received 100 pCi/mouse. After 1 h, kidney homogenates from 3 mice were combined and rates of synthesis determined as described in fig. 3.
compared to saline controls was, in fact, similar in hypophysectomized and normal mice. The apparent small increase in kidney weight of hypophysectomized mice treated with testosterone and prolactin was not significant (P> 0.1) nor did prolactin affect kidney hypertrophy mediated by combined treatment with testosterone and growth hormone. Hypophysectomy and the Gur gene After testosterone treatment normal GuPla mice such as Balb/c synthesize glucuronidase at 3-7 times the rate of normal Gurbb mice such as CD-1 resulting in proportionate increases in kidney glucuronidase concentrations in GuF+la mice (Swank et al., 1973) (fig. 6). Because the Gur gene regulates ghrcuronidase synthesis in kidney of testosterone-treated mice but not in kidney of saline-treated mice or in other organs, it has been postulated to be a site on chromosome 5 which interacts with the androgen-receptor complex. Since these experiments have demonstrated that both androgens and pituitary hormones are necessary for maximal glucuronidase synthesis, an important question in regard to the mechanism of action of the Gur locus is whether it might, in fact, be responding to pituitary products rather than to androgens. It was found that hypophysectomy did not affect glucuronidase synthesis in
Growth hormone and mouse kidney glucuronidase
147
mice not treated with testosterone, but greatly depressed synthesis in mice of both genotypes over the 14-day time course of testosterone treatment (fig. 6). However, the differential rate of glucuronidase synthesis in mice of the two Gur genotypes was maintained in hypophysectomized mice. That is, the ghrcuronidase synthetic rate in hypophysectomized Guiala mice was always greater (1 O-l 5 fold) than that of hypophysectomized Gz4rbD mice after testosterone treatment. Likewise, after hypophysectomy, kidney glucuronidase activity (not shown) was always 6-10 times greater in Balb/c than in CD-l mice. These results suggest that the genetically controlled difference in hormone-mediated glucuronidase synthesis in these mice is not mediated by pituitary products.
DISCUSSION Kidney glucuronidase synthesis is under multihormonal control with two of the major hormones being androgen and growth hormone. Testosterone administration to hypophysectomized mice yielded a rate of synthesis 5--10% of that of similarly treated normal mice. However, simultaneous administration of testosterone and growth hormone to hypophysectomized mice boosted the rate of synthesis to 50% of that in normal mice. On the other hand, the related pituitary hormone, prolactin, had no effect on kidney glucuronidase production. Thus growth hormone serves to facilitate the action of testosterone. It does not, however, act independently of testosterone. Growth hormone treatment of hypophysectomized mice did not increase kidney glucuronidase synthetic rate or activity over that in saline-treated controls in these short-term 3-day experiments with Balb/c mice. Similarly, growth hormone alone did not increase kidney glucuronidase activity in longer 7-day experiments with hypophysectomized CD-1 mice (Swank, unpublished). Also, growth hormone treatment did not further accelerate glucuronidase activity in testosterone-treated normal mice. Growth hormone also causes about 50% restoration of testosterone-mediated kidney glucuronidase synthesis in hypophysectomized CD-1 mice (Bullock, in discussion of Swank et al., 1978; Swank, unpublished). Additional evidence that growth hormone is involved in kidney glucuronidase production derives from the noninducibility (Swank et al., 1978) of Snell’s dwarf (dw/dw) mutant mice which lack several anterior pituitary hormones including growth hormone (Lewis, 1967). Also, testosterone-mediated kindey glucuronidase inducibility is reduced about 50% in mutant little @it/lit) mice (Swank et al., 1978) which have a relatively specific defect of serum growth hormone (Either and Beamer, 1976). Restoration of glucuronidase synthetic rate by injected growth hormone was incomplete even at double the dose of growth hormone routinely used in these experiments. While this could be due to the artificial nature of the injection protocol, a more likely explanation is that other pituitary products are required. Preliminary experiments suggest that thyroxine and glucocorticoids, administered
148
R. T. Swank
together with growth hormone and testosterone, restore glucuronidase production to near-normal levels (Swank, unpublished). It is unlikely that levels of kidney cytosolic testosterone receptor protein or of active testosterone metabolite(s) are being modulated by pituitary removal since several other kidney enzymes are equally inducible by testosterone in normal and hypophysectomized mice (Swank et al., 1977). We cannot, of course, exclude the possibility that there are specific requirements for testosterone metabolites among the kidney enzymes normally inducible by testosterone. Kidney hypertrophy appears to be under similar hormonal control as kidney glucuronidase synthesis. That is, neither testosterone nor growth hormone alone affected relative kidney weight in hypophysectomized mice. However, simultaneous administration of testosterone and growth hormone caused as large an increase in kidney weight as that observed in testosterone treatment of normal mice. As in the case of glucuronidase production, prolactin had no effect on kidney hyper trophy. An apparent dissimilarity in glucuronidase synthesis and kidney hypertrophy is that while testosterone treatment of hypophysectomized mice caused a 6-fold increase in rate of glucuronidase synthesis, it had no measurable effect on kidney hypertrophy. There was likewise little or no increase in kidneyglucuronidase activity in these same mice. These apparent differences may, however, simply reflect the sensitivity of measurement of the various responses. For example, 3-day testosterone treatment of normal mice causes a SO-fold increase in rate of synthesis but only a 6-fold and 1.3-fold increase in activity and hypertrophy, respectively. Mouse kidney hypertrophy is known to include significant increases in the rates of protein and RNA synthesis (Kochakian, 1969). While not directly measured in these experiments, it seems likely that these metabolic processes also require the combined action of testosterone and growth hormone. These experiments suggest that the Gur locus does not regulate glucurinodase production by interaction with pituitary products since the regulatory difference is maintained in the absence of these products. Final proof of this proposal will require comparison of the rate of synthesis of glucuronidase in congenic hypophysectomized mice which are known to differ only at the Gur locus. However, these results are consistent with the hypothesis that the Gur locus on chromosome 5 may be a genomic site of interaction with the testosterone-receptor complex (Swank et al., 1978). Growth hormone has also been implicated in the induction of other specific gene products such as rat liver, kidney and adrenal ornithine decarboxylase (Janne and Raina, 1969; Nicholson et al., 1977) and in maintaining the normal rate of synthesis of rat liver serum albumin (Feldhoff et al., 1977). It likewise acts in concert with androgens, glucocorticoids and thyroid hormones in the induction of rat liver a,,-globulin (Roy, 1973) where it increases levels of translatable mRNA (Roy and Dowbenko, 1977). Widnell and Tata (1966) observed that growth hormone and testosterone stimulate deoxyribonucleic acid-dependent ribonucleic acid polymerase of rat liver nuclei in an additive manner. How several hormones
Growth hormone and mouse kidney glucuronidase
149
interact to regulate specific gene expression in these and other systems (Martial et al., 1977; Kwitz and Feigelson, 1977; Oka and Perry, 1974) is an important but still incompletely understood problem.
ACKNOWLEDGEMENTS The author thanks Gerald Jahreis for exellent technical assistance. Purified growth hormone and prolactin preparations were generously supplied by the National Pituitary Agency, NIAMDD.
REFERENCES Attardi, B. and Ohno, S. (1974) Cell 2, 205-212. Brandt, E.J., Elliott, R.E. and Swank, R.T. (1975) J. Cell Biol. 67, 774-788. Bullock, L.P. and Bardin, C.W. (1972) 3. Clin. Endocrinol. Metab. 35,935-937. Cleveland, C.E. and Swank, R.T. (1978) Biochem. J. 170,249-256. Dofuku, R., Tettenborn, U. and Ohno, S. (1971) Nature New Biol. 232,5-7. Either, E.M. and Beamer, W.G. (1976) J. Hered. 67,87-91. Ernest, M.J., Chen, C.-L. and Feigelson, P. (1977) J. Biol. Chem. 252,6783-6791. Feldhoff, R.C., Taylor, J.M. and Jefferson, L.S. (1977) J. Biol. Chem. 252, 3611-3616. Fishman, W.H. (1965) In: Methods in Hormone Research, Vol. IV, Part B, Ed.: R.I. Dorfman (Academic Press, New York) pp. 2733326. Fishman, W.H., Ide, H. and Rufo, R. (1969) Histochemie 20, 287-299. Gehring, U. and Tomkins, G.M. (1974) Cell 3,59-64. Janne, J. and Raina, A. (1969) Biochim. Biophys. Acta 174,769-772. Kochakian, C.D. (1969) Steroids 14,77-90. Kurtz, D.T. and Feigelson, P. (1977) Proc. Natl. Acad. Sci. (U.S.A.) 74,4791-4795. Lewis, U.J. (1967) Mem. Sot. Endocrinol. 15, 179-191. Martial, J.A., Seeburg, P.H., Guenzi, D., Goodman, H.M. and Baxter, J.D. (1977) Proc. Natl. Acad. Sci. U.S.A. 74,4293-4295. Nicholson, W.E., Barton, R.N., Holladay, L.A., Orth, D.N. and Puett, D. (1977) Endocrinology 100,459-467. Oka, T. and Perry, J.W. (1974) J. Biol. Chem. 249,7647-7652. Paigen, K., Swank, R.T., Tomino, S. and Ganschow, R.E. (1975) J. Cell. Physiol. 85, 379-392. Roy, A.K. (1973) J. Endocrinol. 56,295-301. Roy, A.R. and Dowbenko, D.J. (1977) Biochemistry 16,3918-3922. Swank, R.T., Paigen, K. and Ganschow, R.E. (1973) J. Mol. Biol. 81,225-243. Swank, R.T., Davey, R., Joyce, L., Reid, P. and Macey, M.R. (1977) Endocrinology 100, 473-480. Swank, R.T., Paigen, K., Davey, R., Chapman, V., Labarca, C., Watson, G., Ganschow, R., Brandt, E.J. and Novak, E. (1978) Recent Prog. Horm. Res. 34, in Press. Widnell, CC. and Tata, J.R. (1966) Biochem. J. 98,621-628.
