- Email: [email protected]
Phosphorylation of bovine prolactin eliminates luteotropic activity in the rat
Life Sciences, Vol. 63, No. 14, pp. 12El-117, 1998 COpyright* 1998 Elsetier Science Inc. Printed in the USA. All rights reserved ooz4-3205/98 $19.00 t .oo
PII SOO24-3205(98)00389-O
ELSEVIER
PHOSPHORYLATION
OF BOVINE PROLACTIN ELIMINATES ACTIVITY IN THE RAT’
LUTEOTROPIC
Charles L. Brooks2 and Syed Saiduddin Department of Veterinary Biosciences The Ohio State University 1925 Coffey Road, Columbus, OH 432 10 (Received
in final form July 23, 198)
Summary Phosphorylated and nonphosphorylated prolactins were isolated from bovine pituitaries and their luteotropic activity determined in female rats. Three groups of rats in day 1 of diestrus were treated ip twice daily for three days with 0.25 mg of either prolactin preparation or vehicle. Rats were sacrificed each day of treatment. Serum progesterone concentrations ofthe groups receiving vehicle or phosphorylated prolactin were similar and the vaginal cytology of these animals indicated that phosphorylated bovine prolactin (bPRL) treatment had not prolonged diestrus. Treatment with nonphosphorylated bPRL significantly increased serum progesterone concentration and the vaginal cytology indicated a dies&us prolonged for up to 4 days. Nonphosphorylated and phosphorylated bPRLs were cleared from the blood at similar rates after ip injection. In vitro receptor binding studies demonstrated that phosphorylated bPRL did not bind the ovarian prolactin receptor. Nonphosphorylated, but not phosphorylated, bPRL competed with radiolabeled bovine hormone for occupancy of rat ovarian prolactin receptors. These data are the first to test the activities of phosphorylated bPRL in vivo and indicate; 1) nonphosphorylated bPRL is luteotropic, 2) phosphorylated bPRL is neither luteotropic nor aprolactin receptor agonist or antagonist and 3) phosphorylated bPRL is not dephosphorylated in vivo rapidly enough to provide sufficient biologicallyactive bPRL to maintain luteal function. Key Words:
prolactin,
phosphorylation,
progesterone,
bovine
Prolactin (PRL) is modified by several processes including: glycosylation, deamination, cleavage and phosphorylation (1). Three groups have demonstrated the presence of phosphorylated PRL in the rat (2), chicken (3) and cow (4). The biological significance of prolactin phosphorylation is unclear, although several possibilities are proposed. First, pituitary PRL phosphorylation may be
‘Supported by the National Institutes of Health (DK42604) and a Research Career Development Award (DK01989) to CLB. 2Corresponding author: C. L. Brooks, Department of Veterinary Biosciences, The Ohio State University, 1925 Coffey Road, Columbus, OH 43210. Telephone: (614) 292-9641. FAX:(614) 292-6473. E-Mail: [email protected].
1282
Bovine PRL and Luteotropic Activity
Vol. 63, No. 14, 1998
associated with PRL storage in dense secretory granules. Our laboratory has shown that most pituitary bovine PRL (bPRL) is found in a phosphorylated form (Brooks and Isaccs, unpublished data), while in rat pituitaries PRL phosphorylation has been shown to vary with the estrous cycle (5). Second, phosphorylation might alter the biological activities of PRL. An in vitro assay system, rat lymphomaNB, cells (6), has been used to determine the activities of PRL variants. Phosphorylated bovine or rat PRLs showed reduced activities which were regained after enzymatic removal of phosphate (78). The reduced activity of phosphorylated bPRL was associated with a reduced affinity for the PRL receptor (7). Phosphorylated rat PRL is an antagonist based on a greater than expected increase in activity associated with dephosphorylation. Phosphorylated rat PRL also reduces PRL secretion from rat pituitary cells in vitro (9). The in vivo physiological significance of these variants is yet to be clarified. Finally, phosphorylated PRL may be secreted from the pituitary and activated by phosphatases in the blood or other tissues, thus functioning as a reserve in the blood. No data is currently available for either the blood concentration, perpheral conversion or clearance of phosphorylated PRL. We have developed a method to isolate both nonphosphorylated and phosphorylated bPRLs (10). The sites of bPRL phosphorylation have been identified as serines 26, 34 and 90 with a 1: 1: 10 stoichiometry (11). Glutamic acid mimicry of phosphorylation at serine 90 reduced the biological activity of bPRL in an in vitro bioassay (12) similarly to that observed in phosphorylated bPRL isolated from pituitaries. Following the introduction of the term luteotropin (13) which describes one of the functional properties of PRL in the rat, methods were developed to determine the bioactivity of PRL in an in vivo system using the female rat (14,15). It was shown that administration of PRL prevents the corpus luteum of the cycle from regression and promotes progesterone (P.J secretion. The resultant increase in P, prevents the occurrence of proestrus and estrus as determined by vaginal cytology. In this study we have tested the relative in vivo luteotropic activities of nonphosphorylated and phosphorylated bPRL in the rat. Two questions were addressed in these studies. First, what are the relative activities of nonphosphorylated and phosphorylated bPRL in prolonging P, secretion from the corpus luteum and maintaining diestrus vaginal cytology? Second, are biologically significant amounts of phosphorylated bPRL peripherally converted to the active nonphosphorylated form in the blood or other tissues? Methods Bovine Prolactin Preparations Bovine pituitaries were purchased from Pel-Freeze (Rogers, AK). Prolactins were prepared by sequential immunoaffnity chromatography and separation of the nonphosphorylated and phosphorylated hormones by preparative DEAE anion-exchange HPLC (10). Purity was confirmed by SDS-containing polyacrylamide gel electrophoresis. Nonphosphorylated and phosphorylated bPRL preparations were weighed, placed into 10 mM ammonium bicarbonate and lyophilized in amounts sufficient for one day’s treatments. The hormones were stored at -20°C. Nonphosphorylated and phosphorylated bPRLs were reconstituted in 10 mM NaPO, pH 7.4, 150 mM NaCl each morning and stored at 4°C until injected. Animals The use of rats in these studies was approved by the Institutional Laboratory Animal Care Committee at The Ohio State University. Female Sprague-Dawley rats (70-80 days old) were
Vol. 63, No. 14, 1998
Bovine PRL and Luteotropic
Activity
1283
purchased from Charles River (Portage, MI). They were housed in a temperature-regulated room illuminated between 0600-2000 h. Food and water were offered ad libitum. Rats showing two consecutive 4-day cycles as determined by daily vaginal cytology were used in these experiments. Treatments Determination of the Luteotropic Effects of Bovine Prolactin Variants Treatments were begun on the first day of dies&us (designated day 1) and continued daily thereafter. Treatments consisted of 0.2 ml ip injections at 0800 and 1800 hours of either vehicle (phosphatebuffered saline), phosphorylated bPRL (250 ug) or nonphosphorylated bPRL (250 pg). Groups containing between 4 and 6 rats from each treatment were sacrificed on day 2. 3 or 4 after the treatment commenced. On the day of sacrifice one injection was given at 0800 hours and the rats were killed between 1500 and 1600 hours. Blood was collected and serum separated and stored at -30°C. Ovaries and uteri were excised and weighed. Uteri were grossly examined for ballooning. Ovaries were fixed in buffered formalin for histological confirmation of corpora lutea. Determination of Serum Concentration of Bovine Prolactin The time course of serum prolactin concentrations in rats resulting from ip treatment with nonphosphorylated and phosphorylated bPRL was performed to determine the relative halflives of phosphoylated and nonphosphorylated bPRL in the blood. Groups of rats were injected ip with 200 ul of either vehicle, phosphorylated bPRL (250 ug) or nonphosphorylated bPRL (250 ug). Blood Serum was collected under ether anesthesia by cardiac puncture 2,6 and 12 hours post-injection. was separated and stored at -30°C. Radioimmunoassays Serum Progesterone Progesterone concentrations were determined by a coated-tube assay employing (Diagnostic Products, Los Angeles, CA, cat. # TKPGS).
[‘251]progesterone
Serum Bovine Prolactin Bovine PRL concentrations were determined by a double antibody method employing supplied by the National Hormone and Pituitary Program (Baltimore, MD). The B-l was used as the standard. [‘251]BPRL was prepared by the chloramine-T method (16). studies demonstrated that phosphorylated and nonphosphorylated bPRLs had similar the antiovine PRL antibody used in this assay.
the reagents preparation Preliminary avidities for
Prolactin Receptor-binding Studies Bovine PRL (National Hormone and Pituitary Program, preparation AFP4835B) was labelled with [iz51]iodine by the Method of Thorell and Johansson (17). Membranes were prepared from rat ovaries by differential centrifugation of tissue homogenates as previously described (18). Triplicate assays contained 625 ug membrane protein, 1 ng (18 1,000 cpm) [‘251]bPRL and either no hormone, or between 1 ng and 1 ug of phosphorylated or nonphosphorylated bPRL in 0.5 ml of 10 mM Tris, pH 7.4, 10 mM MgCl,, 0.1% bovine albumin. The samples were incubated overnight at 4°C. Two ml of ice cold 25 mM Tris, pH 7.4, 10 mM MgCl, were added to each tube and the membranes collected by centrifugation. The supematant was aspirated and the radioactivity associated with the membranous pellets determined. Statistical Analysis The data were analyzed by 2 way ANOVA and the significant tested by Tukey HSD test (SYSTAC Inc., Evanston, IL).
differences
between means were
1284
Bovine PRL and Luteotropic
Activity
Vol. 63, No. 14, 1998
Results Luteotropic effects of bPRL variants Serum P, concentrations of rats receiving nonphosphorylated bPRL were significantly higher (PcO.01; Table 1) than controls that were treated with vehicle on days 2, 3 and 4 of treatment. In rats receiving nonphosphorylated bPRL, serum P, increased between days 2 and 4. In contrast, phosphorylated bPRL failed to significantly increase serum P, levels when compared to the controls (Table 1)~
Mean serum progesterone
concentrations
TABLE I. (@ml) in rats treated with bPRL variants
Treatment”
Dav of sacrificeb
Control
Phosphorylated
bPRL
Nonphosphorylated
bPRL
7.1 f 1.7 (5)
12.9 f 4.1
9.9 f 1.1
(5)
(5)
17.0 zt 2.7
13.7 & 1.4
19.7 * 3.9
(4)
(5)
(5)
44.2 + 1.7*
48.2 *2.6*
58.8 *4.2*?
(6)
(5)
(6)
Values are mean f SE; ( ) = n a Rats showing 4-day cycles were injected, starting on the first day of diestrus and thereafter, 0.2 ml ip twice daily 0800 and 1800 h (except on the day of sacrifice). Control rats received phosphate buffered saline, while the test groups received 250 ug of either nonphosphorylated or phosphorylated bPRL. b Rats were sacrificed 2,3 or 4 days following the commencement of treatment. On the day of sacrifice, rats received one ip injection at 0800 h and were sacrificed between 1500 and 1600 h. * Significantly different from controls (PcO.01) as determined by ANOVA and Tukey HSD test. t Significantly different from Day 2 (P < 0.05) as determined by ANOVA and Tukey HSD test.
Animals in control and phosphorylated bPRL treatment groups had nucleated cells on the 3rd day and comified epithelial cells on the 4th day which is consistent with proestrus and estrus phases of a 4-day cycle (Table II). Treatment with nonphosphorylated bPRL prevented proestrus and estrus as determined by the finding of leukocytes in vaginal smears on days 3 and 4.
Vol. 63, No. 14, 1998
Observations
Bovine PRL and Lutcatropic
of vaginal cvtolow
Day of sacrifice
1285
Activity
TABLE II. following treatment with bPRL variants Treatmenta and cvtologv
Control
Nonphosphoryl.
bPRL
Phosphorylated
bPRL
2
leukocytes
leukocytes
leukocytes
3
nucleated cells
leukocytes
nucleated cells
4
cornified cells
leukocytes
comified cells
aSee Table I for treatment details. Competition for Ovarian Prolactin Receptors Specific binding of [‘251]bPRL represented 19.3% of the total radioactive hormone. Doses of nonphosphorylated bPRL between 1 ng and 1 ug (final assay concentrations between 0.9 x 1O-‘Oand 0.9 x 1O-‘)effectively competed for rat ovarian PRL receptors (Figure 1). Similar concentrations of phosphorylated bPRL failed to effectively compete for receptors. Fig. 1 40000
:zz
-
z 4
25000
-
I
2oooo 15000
=
v1
10000
-
5000
-
O-
a
--c
nonphosphorylated
+
phosphorylated
1
10
bPRL
bPRL 100
1000
Bovine ProlacUn (ng)
Competitive binding of nonphosphorylated and phosphorylated bPRLs to rat ovarian receptors. Triplicate determinations contained 1 ng of [‘251]bPRL, 625 ug protein of rat ovarian membranes, and between 1 ng and 1 ug of either nonphosphorylated or phosphorylated bPRL in 0.5 ml of Tris-Mg buffer, pH 7.4. The samples were incubated at 4°C for 16 hours and the membranes collected by centrifugation. The membrane-associated radioactivity was measured. The calculation of nonspecific binding was based on the tubes containing 1 ug of nonphosphorylated bPFU. Standard errors for individual data points varied between 155 and 2367 cpm. Serum concentrations of injected bovine PRLs in rats The relative clearance from the blood of the two bPRL variants following a single ip injection are presented in Table 3. The phosphorylated and nonphosphorylated bPRL were significantly elevated (PcO.01) in the rat serum 2 h after injection when compared to the control group. No significantdifferences were observed between treatment groups at 6 and 12 h post-injection. No
1286
Bovine PRL and Luteotropic
Activity
Vol. 63, No. 14, 1998
significant differences were observed between the nonphosphorylated and phosphotylated bPRL concentrations at any of the sampling times. The control group did not receive exogenous bPRL but cross reactivity of the assay detected low concentrations of endogenous rat PRL. Morphological evaluation of uterus and ovary The uteri of rats in proestrus were ballooned due to accumulation of luminal fluid when compared to other stages of the cycle. Ovaries contained corpora lutea in all treatment groups when sections were examined under low power magnification. TABLE III. Serum
bPRL concentrations
(@ml)
in rats treated with bPFU variants
Treatmenta
Hours nost-injection 38* 10
Control
(4) Phosphorylated
bPRL
Nonphosphorylated
bPRL
1410* 50*
30& 5 (4) 46 + 4
26* 1 (4) 31+2
(4)
(4)
(3)
1755 + 358”
61 k 13
(3)
(4)
33 + 3 (4)
Values are mean & SE; ( ) = n “Female rats were injected ip with 0.2 ml quantity of vehicle only or one of the bPRL variants (250 pg) and killed at 2, 6 or 12 h. *Significantly different (PcO.01) from control as determined by ANOVA and Tukey HSD test. Discussion
Injection of nonphosphorylated bPRL increased serum P, concentrations above those of the control or phosphorylated bPRL groups for up to 4 days from the first day of diestrus (Table 1). In contrast, treatment with phosphorylated bPRL failed to increase serum P4 concentrations above those of the controls. According to the vaginal cytology, the rats in control and phosphorylated bPRL treatment groups maintained normal cycles, consistent with the lack of rise in P, beyond the 2nd day of the cycle (Table 2). In the nonphosphorylated bPRL group diestrus smear was extended up to the 4th day after treatment began which reflects the rise in P, levels. Therefore, the nonphosphorylated bPRL is a luteotropin in the rat while its phosphorylated variant is not a luteotropin. Previously ovine PRL has been shown to be luteotropic (15); to our knowledge this is the first report that bPRL also possesses this activity. The clearance of ip injections of immunoreactive bPRLs from the blood of the rats showed no differences based on the state of bPRL phosphorylation (Table 3). Therefore, large differences in the release from the site of injection and clearance of bPRL from the blood are unlikely to be responsible for their differences in luteotropic actions. We have previously demonstrated that dephosphorylation of phosphorylated bPRL can be enzymatically performed in vitro with a concomitant increase in biological activity (7). The lack of a sustained increase in P, or retention of a diestrus vaginal cytology demonstrates that biologically
Vol. 63, No. 14, 1998
Bovine PRL and Luteotropic
Aclivily
1287
significant amounts ofphosphorylated bPRL are not converted to the active nonphosphorylated form in either the blood or peripheral tissues. Thus, pituitary secretion of phosphorylated bPRL coupled with peripheral conversion to the active nonphosphorylated form does not appear to be a biologically important mechanism to maintain a steady-state blood concentration ofbiologically active hormone. Wang and Walker (5) presented data suggesting that rat PRL obtained from the National Hormone and Pituitary Program contained a small portion of phosphorylated PRL and when treated with acid phosphatase the biological activity is increased disproportionally due to the conversion of the phosphorylated form when measured by the Nb, rat lymphoma cell assay. From this data they suggest that the phosphorylated rat PRL is a receptor antagonist. We have directly studied this issue using phosphorylated and nonphosphorylated bPRL isolated from pituitaries (10) in in vitro studies. We previously demonstrated in Nb2 rat lymphoma cells that phosphorylated bPRL is not a receptor antagonist (7). In the current study, binding of [“‘I]bPRL was reduced by competition with nonphosphorylated bPRL, while phosphorylated bPRL was not a competitor for PRL receptors in the rat ovary. These results provide additional evidence that phosphorylated bPRL is not a receptor antagonist and that the lack of PRL activity in vivo is explained by the loss of receptor binding capacity observed in vitro. Such a loss of receptor binding is consistent with our observations (11,12) indicating a phosphorylation-induced change in the conformation of this protein. References
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
Y.N. SINHA, Endocrine
Reviews 16 354-369 (1995). W.S. OETTING, P.T. TUAZON, J.A. TRAUGH and A.M. WALKER, J. Biol Chem. 261 1649-1652 (1986). C. ARAMBURO, J.L. MONTEIL, J.A. PROUDMAN, L.R. BERGHMAN and C.G. SCANES, J. Mol. Endocrinol. 8 183-191 (1992). C.L. BROOKS, B.G. KIM, P. APHALE, B.E. KLEEMAN and G.C. JOHNSON, Mol. Cell. Endocrinol. 71 117-123 (1990). T.C. W. HO and A.M. Walker, Neuroendocrinol. 58 160- 165 (1993). T. TANAKA, R.P.C. SHIU, P.W. GOUT, C.T. BEER, T.L. NOBLE and H.G. FRIESEN, J. Clin. Endocrinol. Metab. 51 1058-1063 (1980). J.R. WICKS and C.L. BROOKS, Mol. Cell. Endocrinol. 112 223-229 (1995). Y.-F. WANG and A.M. WALKER, Endocrinol. 133 2156-2160 (1993). T.C.W. HO, J.R. GREENAN and A.M. WALKER, Endocrinol. 124 1507-1514 (1989). C.L. BROOKS, L.A. ISAACS and J.R. WICKS, Mol. Cell.Endocrinol. 99 301-305 (1994). B.G. KIM and C.L. BROOKS, Biochem. J. 296 41-47 (1993). P.M. MACIEJEWSKI, F.C. PETERSON, P.J. ANDERSON and C.L. BROOKS, J. Biol. Chem. 270 23660-23665 (1995). E.B. ASTWOOD, Endocrinol. 28 309-320 (1941). M.S. SMITH, M.E. FREEMAN and J.D. NEILL, Endocrinol. 96 219-226 (1975). M.S. SMITH, B.K. MCLEAN and J.D. NEILL, Endocrinol. 98 1370-137 (1976). F.C. GREENWOOD, W.M. HUNTERand J.S. GLOVER, Biochem. J. 89 114-123 (1963). J.I. THORELL and B.G. JOHANSSON, Biochem. Biophys. Acta. 251363-369 (1971). C.L. BROOKS, J.G. LEINEN, E.R DeSOMBRE and E.V. JENSEN, Mol. Cell. Endocrinol. 26 81-94 (1982).
PII SOO24-3205(98)00389-O
ELSEVIER
PHOSPHORYLATION
OF BOVINE PROLACTIN ELIMINATES ACTIVITY IN THE RAT’
LUTEOTROPIC
Charles L. Brooks2 and Syed Saiduddin Department of Veterinary Biosciences The Ohio State University 1925 Coffey Road, Columbus, OH 432 10 (Received
in final form July 23, 198)
Summary Phosphorylated and nonphosphorylated prolactins were isolated from bovine pituitaries and their luteotropic activity determined in female rats. Three groups of rats in day 1 of diestrus were treated ip twice daily for three days with 0.25 mg of either prolactin preparation or vehicle. Rats were sacrificed each day of treatment. Serum progesterone concentrations ofthe groups receiving vehicle or phosphorylated prolactin were similar and the vaginal cytology of these animals indicated that phosphorylated bovine prolactin (bPRL) treatment had not prolonged diestrus. Treatment with nonphosphorylated bPRL significantly increased serum progesterone concentration and the vaginal cytology indicated a dies&us prolonged for up to 4 days. Nonphosphorylated and phosphorylated bPRLs were cleared from the blood at similar rates after ip injection. In vitro receptor binding studies demonstrated that phosphorylated bPRL did not bind the ovarian prolactin receptor. Nonphosphorylated, but not phosphorylated, bPRL competed with radiolabeled bovine hormone for occupancy of rat ovarian prolactin receptors. These data are the first to test the activities of phosphorylated bPRL in vivo and indicate; 1) nonphosphorylated bPRL is luteotropic, 2) phosphorylated bPRL is neither luteotropic nor aprolactin receptor agonist or antagonist and 3) phosphorylated bPRL is not dephosphorylated in vivo rapidly enough to provide sufficient biologicallyactive bPRL to maintain luteal function. Key Words:
prolactin,
phosphorylation,
progesterone,
bovine
Prolactin (PRL) is modified by several processes including: glycosylation, deamination, cleavage and phosphorylation (1). Three groups have demonstrated the presence of phosphorylated PRL in the rat (2), chicken (3) and cow (4). The biological significance of prolactin phosphorylation is unclear, although several possibilities are proposed. First, pituitary PRL phosphorylation may be
‘Supported by the National Institutes of Health (DK42604) and a Research Career Development Award (DK01989) to CLB. 2Corresponding author: C. L. Brooks, Department of Veterinary Biosciences, The Ohio State University, 1925 Coffey Road, Columbus, OH 43210. Telephone: (614) 292-9641. FAX:(614) 292-6473. E-Mail: [email protected].
1282
Bovine PRL and Luteotropic Activity
Vol. 63, No. 14, 1998
associated with PRL storage in dense secretory granules. Our laboratory has shown that most pituitary bovine PRL (bPRL) is found in a phosphorylated form (Brooks and Isaccs, unpublished data), while in rat pituitaries PRL phosphorylation has been shown to vary with the estrous cycle (5). Second, phosphorylation might alter the biological activities of PRL. An in vitro assay system, rat lymphomaNB, cells (6), has been used to determine the activities of PRL variants. Phosphorylated bovine or rat PRLs showed reduced activities which were regained after enzymatic removal of phosphate (78). The reduced activity of phosphorylated bPRL was associated with a reduced affinity for the PRL receptor (7). Phosphorylated rat PRL is an antagonist based on a greater than expected increase in activity associated with dephosphorylation. Phosphorylated rat PRL also reduces PRL secretion from rat pituitary cells in vitro (9). The in vivo physiological significance of these variants is yet to be clarified. Finally, phosphorylated PRL may be secreted from the pituitary and activated by phosphatases in the blood or other tissues, thus functioning as a reserve in the blood. No data is currently available for either the blood concentration, perpheral conversion or clearance of phosphorylated PRL. We have developed a method to isolate both nonphosphorylated and phosphorylated bPRLs (10). The sites of bPRL phosphorylation have been identified as serines 26, 34 and 90 with a 1: 1: 10 stoichiometry (11). Glutamic acid mimicry of phosphorylation at serine 90 reduced the biological activity of bPRL in an in vitro bioassay (12) similarly to that observed in phosphorylated bPRL isolated from pituitaries. Following the introduction of the term luteotropin (13) which describes one of the functional properties of PRL in the rat, methods were developed to determine the bioactivity of PRL in an in vivo system using the female rat (14,15). It was shown that administration of PRL prevents the corpus luteum of the cycle from regression and promotes progesterone (P.J secretion. The resultant increase in P, prevents the occurrence of proestrus and estrus as determined by vaginal cytology. In this study we have tested the relative in vivo luteotropic activities of nonphosphorylated and phosphorylated bPRL in the rat. Two questions were addressed in these studies. First, what are the relative activities of nonphosphorylated and phosphorylated bPRL in prolonging P, secretion from the corpus luteum and maintaining diestrus vaginal cytology? Second, are biologically significant amounts of phosphorylated bPRL peripherally converted to the active nonphosphorylated form in the blood or other tissues? Methods Bovine Prolactin Preparations Bovine pituitaries were purchased from Pel-Freeze (Rogers, AK). Prolactins were prepared by sequential immunoaffnity chromatography and separation of the nonphosphorylated and phosphorylated hormones by preparative DEAE anion-exchange HPLC (10). Purity was confirmed by SDS-containing polyacrylamide gel electrophoresis. Nonphosphorylated and phosphorylated bPRL preparations were weighed, placed into 10 mM ammonium bicarbonate and lyophilized in amounts sufficient for one day’s treatments. The hormones were stored at -20°C. Nonphosphorylated and phosphorylated bPRLs were reconstituted in 10 mM NaPO, pH 7.4, 150 mM NaCl each morning and stored at 4°C until injected. Animals The use of rats in these studies was approved by the Institutional Laboratory Animal Care Committee at The Ohio State University. Female Sprague-Dawley rats (70-80 days old) were
Vol. 63, No. 14, 1998
Bovine PRL and Luteotropic
Activity
1283
purchased from Charles River (Portage, MI). They were housed in a temperature-regulated room illuminated between 0600-2000 h. Food and water were offered ad libitum. Rats showing two consecutive 4-day cycles as determined by daily vaginal cytology were used in these experiments. Treatments Determination of the Luteotropic Effects of Bovine Prolactin Variants Treatments were begun on the first day of dies&us (designated day 1) and continued daily thereafter. Treatments consisted of 0.2 ml ip injections at 0800 and 1800 hours of either vehicle (phosphatebuffered saline), phosphorylated bPRL (250 ug) or nonphosphorylated bPRL (250 pg). Groups containing between 4 and 6 rats from each treatment were sacrificed on day 2. 3 or 4 after the treatment commenced. On the day of sacrifice one injection was given at 0800 hours and the rats were killed between 1500 and 1600 hours. Blood was collected and serum separated and stored at -30°C. Ovaries and uteri were excised and weighed. Uteri were grossly examined for ballooning. Ovaries were fixed in buffered formalin for histological confirmation of corpora lutea. Determination of Serum Concentration of Bovine Prolactin The time course of serum prolactin concentrations in rats resulting from ip treatment with nonphosphorylated and phosphorylated bPRL was performed to determine the relative halflives of phosphoylated and nonphosphorylated bPRL in the blood. Groups of rats were injected ip with 200 ul of either vehicle, phosphorylated bPRL (250 ug) or nonphosphorylated bPRL (250 ug). Blood Serum was collected under ether anesthesia by cardiac puncture 2,6 and 12 hours post-injection. was separated and stored at -30°C. Radioimmunoassays Serum Progesterone Progesterone concentrations were determined by a coated-tube assay employing (Diagnostic Products, Los Angeles, CA, cat. # TKPGS).
[‘251]progesterone
Serum Bovine Prolactin Bovine PRL concentrations were determined by a double antibody method employing supplied by the National Hormone and Pituitary Program (Baltimore, MD). The B-l was used as the standard. [‘251]BPRL was prepared by the chloramine-T method (16). studies demonstrated that phosphorylated and nonphosphorylated bPRLs had similar the antiovine PRL antibody used in this assay.
the reagents preparation Preliminary avidities for
Prolactin Receptor-binding Studies Bovine PRL (National Hormone and Pituitary Program, preparation AFP4835B) was labelled with [iz51]iodine by the Method of Thorell and Johansson (17). Membranes were prepared from rat ovaries by differential centrifugation of tissue homogenates as previously described (18). Triplicate assays contained 625 ug membrane protein, 1 ng (18 1,000 cpm) [‘251]bPRL and either no hormone, or between 1 ng and 1 ug of phosphorylated or nonphosphorylated bPRL in 0.5 ml of 10 mM Tris, pH 7.4, 10 mM MgCl,, 0.1% bovine albumin. The samples were incubated overnight at 4°C. Two ml of ice cold 25 mM Tris, pH 7.4, 10 mM MgCl, were added to each tube and the membranes collected by centrifugation. The supematant was aspirated and the radioactivity associated with the membranous pellets determined. Statistical Analysis The data were analyzed by 2 way ANOVA and the significant tested by Tukey HSD test (SYSTAC Inc., Evanston, IL).
differences
between means were
1284
Bovine PRL and Luteotropic
Activity
Vol. 63, No. 14, 1998
Results Luteotropic effects of bPRL variants Serum P, concentrations of rats receiving nonphosphorylated bPRL were significantly higher (PcO.01; Table 1) than controls that were treated with vehicle on days 2, 3 and 4 of treatment. In rats receiving nonphosphorylated bPRL, serum P, increased between days 2 and 4. In contrast, phosphorylated bPRL failed to significantly increase serum P, levels when compared to the controls (Table 1)~
Mean serum progesterone
concentrations
TABLE I. (@ml) in rats treated with bPRL variants
Treatment”
Dav of sacrificeb
Control
Phosphorylated
bPRL
Nonphosphorylated
bPRL
7.1 f 1.7 (5)
12.9 f 4.1
9.9 f 1.1
(5)
(5)
17.0 zt 2.7
13.7 & 1.4
19.7 * 3.9
(4)
(5)
(5)
44.2 + 1.7*
48.2 *2.6*
58.8 *4.2*?
(6)
(5)
(6)
Values are mean f SE; ( ) = n a Rats showing 4-day cycles were injected, starting on the first day of diestrus and thereafter, 0.2 ml ip twice daily 0800 and 1800 h (except on the day of sacrifice). Control rats received phosphate buffered saline, while the test groups received 250 ug of either nonphosphorylated or phosphorylated bPRL. b Rats were sacrificed 2,3 or 4 days following the commencement of treatment. On the day of sacrifice, rats received one ip injection at 0800 h and were sacrificed between 1500 and 1600 h. * Significantly different from controls (PcO.01) as determined by ANOVA and Tukey HSD test. t Significantly different from Day 2 (P < 0.05) as determined by ANOVA and Tukey HSD test.
Animals in control and phosphorylated bPRL treatment groups had nucleated cells on the 3rd day and comified epithelial cells on the 4th day which is consistent with proestrus and estrus phases of a 4-day cycle (Table II). Treatment with nonphosphorylated bPRL prevented proestrus and estrus as determined by the finding of leukocytes in vaginal smears on days 3 and 4.
Vol. 63, No. 14, 1998
Observations
Bovine PRL and Lutcatropic
of vaginal cvtolow
Day of sacrifice
1285
Activity
TABLE II. following treatment with bPRL variants Treatmenta and cvtologv
Control
Nonphosphoryl.
bPRL
Phosphorylated
bPRL
2
leukocytes
leukocytes
leukocytes
3
nucleated cells
leukocytes
nucleated cells
4
cornified cells
leukocytes
comified cells
aSee Table I for treatment details. Competition for Ovarian Prolactin Receptors Specific binding of [‘251]bPRL represented 19.3% of the total radioactive hormone. Doses of nonphosphorylated bPRL between 1 ng and 1 ug (final assay concentrations between 0.9 x 1O-‘Oand 0.9 x 1O-‘)effectively competed for rat ovarian PRL receptors (Figure 1). Similar concentrations of phosphorylated bPRL failed to effectively compete for receptors. Fig. 1 40000
:zz
-
z 4
25000
-
I
2oooo 15000
=
v1
10000
-
5000
-
O-
a
--c
nonphosphorylated
+
phosphorylated
1
10
bPRL
bPRL 100
1000
Bovine ProlacUn (ng)
Competitive binding of nonphosphorylated and phosphorylated bPRLs to rat ovarian receptors. Triplicate determinations contained 1 ng of [‘251]bPRL, 625 ug protein of rat ovarian membranes, and between 1 ng and 1 ug of either nonphosphorylated or phosphorylated bPRL in 0.5 ml of Tris-Mg buffer, pH 7.4. The samples were incubated at 4°C for 16 hours and the membranes collected by centrifugation. The membrane-associated radioactivity was measured. The calculation of nonspecific binding was based on the tubes containing 1 ug of nonphosphorylated bPFU. Standard errors for individual data points varied between 155 and 2367 cpm. Serum concentrations of injected bovine PRLs in rats The relative clearance from the blood of the two bPRL variants following a single ip injection are presented in Table 3. The phosphorylated and nonphosphorylated bPRL were significantly elevated (PcO.01) in the rat serum 2 h after injection when compared to the control group. No significantdifferences were observed between treatment groups at 6 and 12 h post-injection. No
1286
Bovine PRL and Luteotropic
Activity
Vol. 63, No. 14, 1998
significant differences were observed between the nonphosphorylated and phosphotylated bPRL concentrations at any of the sampling times. The control group did not receive exogenous bPRL but cross reactivity of the assay detected low concentrations of endogenous rat PRL. Morphological evaluation of uterus and ovary The uteri of rats in proestrus were ballooned due to accumulation of luminal fluid when compared to other stages of the cycle. Ovaries contained corpora lutea in all treatment groups when sections were examined under low power magnification. TABLE III. Serum
bPRL concentrations
(@ml)
in rats treated with bPFU variants
Treatmenta
Hours nost-injection 38* 10
Control
(4) Phosphorylated
bPRL
Nonphosphorylated
bPRL
1410* 50*
30& 5 (4) 46 + 4
26* 1 (4) 31+2
(4)
(4)
(3)
1755 + 358”
61 k 13
(3)
(4)
33 + 3 (4)
Values are mean & SE; ( ) = n “Female rats were injected ip with 0.2 ml quantity of vehicle only or one of the bPRL variants (250 pg) and killed at 2, 6 or 12 h. *Significantly different (PcO.01) from control as determined by ANOVA and Tukey HSD test. Discussion
Injection of nonphosphorylated bPRL increased serum P, concentrations above those of the control or phosphorylated bPRL groups for up to 4 days from the first day of diestrus (Table 1). In contrast, treatment with phosphorylated bPRL failed to increase serum P4 concentrations above those of the controls. According to the vaginal cytology, the rats in control and phosphorylated bPRL treatment groups maintained normal cycles, consistent with the lack of rise in P, beyond the 2nd day of the cycle (Table 2). In the nonphosphorylated bPRL group diestrus smear was extended up to the 4th day after treatment began which reflects the rise in P, levels. Therefore, the nonphosphorylated bPRL is a luteotropin in the rat while its phosphorylated variant is not a luteotropin. Previously ovine PRL has been shown to be luteotropic (15); to our knowledge this is the first report that bPRL also possesses this activity. The clearance of ip injections of immunoreactive bPRLs from the blood of the rats showed no differences based on the state of bPRL phosphorylation (Table 3). Therefore, large differences in the release from the site of injection and clearance of bPRL from the blood are unlikely to be responsible for their differences in luteotropic actions. We have previously demonstrated that dephosphorylation of phosphorylated bPRL can be enzymatically performed in vitro with a concomitant increase in biological activity (7). The lack of a sustained increase in P, or retention of a diestrus vaginal cytology demonstrates that biologically
Vol. 63, No. 14, 1998
Bovine PRL and Luteotropic
Aclivily
1287
significant amounts ofphosphorylated bPRL are not converted to the active nonphosphorylated form in either the blood or peripheral tissues. Thus, pituitary secretion of phosphorylated bPRL coupled with peripheral conversion to the active nonphosphorylated form does not appear to be a biologically important mechanism to maintain a steady-state blood concentration ofbiologically active hormone. Wang and Walker (5) presented data suggesting that rat PRL obtained from the National Hormone and Pituitary Program contained a small portion of phosphorylated PRL and when treated with acid phosphatase the biological activity is increased disproportionally due to the conversion of the phosphorylated form when measured by the Nb, rat lymphoma cell assay. From this data they suggest that the phosphorylated rat PRL is a receptor antagonist. We have directly studied this issue using phosphorylated and nonphosphorylated bPRL isolated from pituitaries (10) in in vitro studies. We previously demonstrated in Nb2 rat lymphoma cells that phosphorylated bPRL is not a receptor antagonist (7). In the current study, binding of [“‘I]bPRL was reduced by competition with nonphosphorylated bPRL, while phosphorylated bPRL was not a competitor for PRL receptors in the rat ovary. These results provide additional evidence that phosphorylated bPRL is not a receptor antagonist and that the lack of PRL activity in vivo is explained by the loss of receptor binding capacity observed in vitro. Such a loss of receptor binding is consistent with our observations (11,12) indicating a phosphorylation-induced change in the conformation of this protein. References
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
Y.N. SINHA, Endocrine
Reviews 16 354-369 (1995). W.S. OETTING, P.T. TUAZON, J.A. TRAUGH and A.M. WALKER, J. Biol Chem. 261 1649-1652 (1986). C. ARAMBURO, J.L. MONTEIL, J.A. PROUDMAN, L.R. BERGHMAN and C.G. SCANES, J. Mol. Endocrinol. 8 183-191 (1992). C.L. BROOKS, B.G. KIM, P. APHALE, B.E. KLEEMAN and G.C. JOHNSON, Mol. Cell. Endocrinol. 71 117-123 (1990). T.C. W. HO and A.M. Walker, Neuroendocrinol. 58 160- 165 (1993). T. TANAKA, R.P.C. SHIU, P.W. GOUT, C.T. BEER, T.L. NOBLE and H.G. FRIESEN, J. Clin. Endocrinol. Metab. 51 1058-1063 (1980). J.R. WICKS and C.L. BROOKS, Mol. Cell. Endocrinol. 112 223-229 (1995). Y.-F. WANG and A.M. WALKER, Endocrinol. 133 2156-2160 (1993). T.C.W. HO, J.R. GREENAN and A.M. WALKER, Endocrinol. 124 1507-1514 (1989). C.L. BROOKS, L.A. ISAACS and J.R. WICKS, Mol. Cell.Endocrinol. 99 301-305 (1994). B.G. KIM and C.L. BROOKS, Biochem. J. 296 41-47 (1993). P.M. MACIEJEWSKI, F.C. PETERSON, P.J. ANDERSON and C.L. BROOKS, J. Biol. Chem. 270 23660-23665 (1995). E.B. ASTWOOD, Endocrinol. 28 309-320 (1941). M.S. SMITH, M.E. FREEMAN and J.D. NEILL, Endocrinol. 96 219-226 (1975). M.S. SMITH, B.K. MCLEAN and J.D. NEILL, Endocrinol. 98 1370-137 (1976). F.C. GREENWOOD, W.M. HUNTERand J.S. GLOVER, Biochem. J. 89 114-123 (1963). J.I. THORELL and B.G. JOHANSSON, Biochem. Biophys. Acta. 251363-369 (1971). C.L. BROOKS, J.G. LEINEN, E.R DeSOMBRE and E.V. JENSEN, Mol. Cell. Endocrinol. 26 81-94 (1982).