Cellular Signalling Vol. 4, No. 4, pp. 385-391, 1992. Printed in Great Britain.
CHANGES NUCLEI
0898-6568/92 $5.1)0 + 0.00 © 1992 Pergamon Press Ltd
IN POLYPHOSPHOINOSITIDE IN RESPONSE
TO PROLACTIN, MITOGEN
LEVELS
IN RAT LIVER
A KNOWN
HEPATIC
P. SANrI,* A. M. MARTELLI,*R. S. GILMOUR,t E. FALCIERI,*R. RANA,~:A. CAa'ALDI,:~ G. LATTANZI,§ R. BAREGGIII and L. C o c c o * ¶
*Istituto di Anatomia Umana Normale, Universita' di Bologna, Italy, IIIstituto di Anatomia Umana Normale, Universta' di Trieste, Italy, :~Istituto di Morfologia Umana Normale, Universita' di Chieti, Italy, §Istituto di Citomorfologia del CNR, Bologna, Italy and tlnstitute of Animal Physiology and Genetics Research, AFRC, Babraham, Cambridge, U.K. (Received 23 January 1992; and accepted 17 March 1992) Abstract--The effect of prolactin action on nuclear polyphosphoinositide synthesis was investigated in isolated rat liver nuclei. An increased uptake of phosphate from [~32p]adenosinetrisphosphate was observed in both phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate with a maximum response at 10-~M concentration of hormone. Pulse-chase experiments in isolated nuclei following prolactin treatment indicate that the observed increase in accumulation of phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate is mainly due to a decrease in their rate of turnover possibly induced by a change in activity of polyphosphoinositide-specific monoesterases. In vitro prolactin also reduces the activity of nuclear phospholipase C specific for phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate. Moreover, this feature is strongly supported by the concomitant decrease in nuclear diacylglycerol mass. Thus these data suggest that once prolactin reaches the nucleus an intranuclear signalling is evoked through inositol lipid metabolism. Key words: Prolactin, polyphosphoinositides, rat liver nuclei. tically activated by the addition of prolactin [3]. This latter finding is supported by the suggestion that prolactin receptors are present in the nucleus [4, 5]. For its optimal activity, PKC requires phosphatidylserine (PS) as well as diacylglycerol (DAG) and free calcium ions which are liberated from cellular stores by inositoi (l,4,5)trisphosphate (IP3) [6, 7]. Moreover, the presence of inositol lipids with the potential for PKC activation in the nucleus of different cell types clearly indicates that these molecules play a role in the intranuclear signalling during cell differentiation and cell proliferation induced by growth factors [8-12]. Indeed, changes in nuclear polyphosphoinositide metabolism occur only in mitogenresponsive Swiss 3T3 cells and PKC translocation to the nucleus of these cells follows mitogen-induced changes of nuclear polyphosphoinositides [13-15]. Moreover, IP 3 stimulates PKC activity in nuclei from mitogen-treated
INTRODUCTION PROLACTINis an adenohypophyseal polypeptide hormone which exerts a mitogenic effect in rat hepatocytes [1]. There is some evidence that suggests that prolactin-dependent intracellular signalling involves the activation of protein kinase C (PKC) and that a significant part of this response is due to the activation of nuclear PKC. Indeed, highly purified, membranedeprived nuclei, as well as nuclear matrix from rat liver contain PKC [2], which can be drama¶Author to whom correspondence should be addressed at: Istituto di Anatomia Umana Normale, Via Imerio 48, 40126 Bologna, Italy. Abbreviations: ATP--adenosinetrisphosphate; BSA-bovine serum albumin; DAG---diacylglycerol; HPLC--high pressure liquid chromatography; IGF-l--insulin-like growth factor I; IP2--inositol( 1,4)bisphosphate; IP3--inositol(l,4,5)trisphosphate; PA--phosphatidic acid; PI--phosphatidylinositol; PIP--phosphatidylinositol 4monophosphate; PIP2--phosphatidylinositol 4,5-bisphosphate; PKC--protein kinase C; PS--phosphatidylserine; PLC--phospholipase C. 385
386
P. SANTIet al.
Swiss 3T3 fibroblasts [16]. On the other hand, phosphatidylinositol 4,5-bisphosphate (PIP2) itself can act as a regulatory molecule for PKC activity, since it may antecede D A G as activator of PKC interacting with the activatorreceiving region of the regulatory moiety o f this phosphorylative enzyme [17, 18]. Because of the evidence of the involvement of nuclear polyphosphoinositides in the mitogenic response we have sought to determine whether prolactin treatment could affect the in vitro synthesis of inositol lipids in rat liver nuclei. MATERIALS AND METHODS Preparation of rat liver nuclei
Rat liver nuclei preparation, in the presence of Triton X-100, was as previously described [8]. Phosphorylation of isolated nuclei and analysis of lipid extracts
All the procedures were as in Ref. [8]. Prolactin, solubilized in 10mM Tris-HC1, pHS.0, 1% BSA, was added directly to the phosphorylation mixture for the entire phosphorylation time (i.e. 4 min) at the concentrations indicated in the figure legends. Pulsechase experiments were carried out by incubating the nuclei in the presence of 2.5 /aCi [y32P]ATP (5000 Ci/mmol), final ATP concentration of 100/aM, for 4rain and then by adding a 10-fold excess of unlabelled ATP for 15 or 30 rain. Assay for PIP and PIP2 specific PLC
Phospholipase C (PLC) activity was assayed using 3 nmol of 3H-phosphatidylinositol 4-monophosphate (PIP) or 3H-PIP2 (90,000 d.p.m.) as exogenous substrates, 60/~g nuclear protein as enzyme source in the presence of 0.06% Taurodeoxicolate and by incubating for 30 min at 37°C in the presence or absence of prolactin. For PIP hydrolysis the buffer used was 100/~M Tris-HCl, pH 7.0, plus I mM CaCI2, whereas for PIP 2 the buffer used was 100/zM (2-[N-morpholino]ethanesulphonic acid) pH6.2, plus 100/~M CaCI2. Inositol phosphates liberated were recovered from the aqueous phase as in Ref. [8] and analysed by HPLC as described by Irvine et al. [19]. DAG assay
The total amount of nuclear DAG was measured in isolated nuclei after incubation in the same conditions as above as described by Divecha and Irvine [101.
Electron microscopy and cytoplasmic enzyme marker assay
Nuclear samples were fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, postfixed in 2% osmium tetroxide, dehydrated with acetone and embedded in Araldite. Thin sections stained with uranyl acetate and lead citrate were observed in a Zeiss El09 electron microscope to check the purity of the preparations. Nuclear preparations were also tested for glucose-6-phosphatase as described by Garland and Cori [21]. Routinely, this was less than 0.9% of the activity present in whole-cell homogenate. RESULTS Rat liver nuclei isolated by the procedure described are highly pure. There is no evidence of the outer nuclear envelope or of extranuclear debris when examined by electron microscopy (Fig. l), nor is there detectable glucose-6-phosphatase activity, a recognized cytoplasmic marker, confirming previous observations [8, 121. Incubation of purified rat liver nuclei shows the appearance of radiolabelled lipids identified, because of the exact coincidence of the radioactive spots with the internal standards, as phosphatidic acid (PA), PIP and PIP 2. The incorporation of radiolabelled phosphate came to a steady state after 4rain as previously shown in other nuclei [8, 12]. When prolactin was added to the incubation mixture the rate of synthesis of both PIP and PIP 2 was increased (Fig. 2). The dose-response from 10 _5 to 10 -~4 indicates that the effect is maximal at 10-~2M prolactin (Fig. 2). When pulse-chase experiments (Fig. 3) were carried out, the addition of excess unlabelled ATP chased out a large part of the incorporated 32p in both PIP and PIP z (83 and 61%, respectively). In the presence of 10-t2M prolactin the displacement of incorporated 3zp in PIP as well as in PIP 2 is less marked (i.e. 38 and 50%, respectively). In order to investigate whether PLC activities specific for PIP and PIP 2 are affected by prolactin we have set up an in vitro assay using 3H-labelled PIP and PIP 2 as specific substrates. Table 1 shows that these activities are present indeed in
FIG. 1. Electron microscopy of nuclear preparations. Rat liver nuclei isolated in the presence of Triton X-100. Both dispersed and condensed chromatin domains as well as interchromatin areas are visible. The inset shows at higher magnification the absence of the nuclear membrane (arrows). Scale bars = 1 pm.
387
Prolactin affects nuclear inositol lipid cycle Prolactin effect Percentage increase over the control 100
[] P,P []
Analysing the Ca 2+ dependency of 3H-inositol(1,4,5)trisphosphate (IP2) and 3H-inositol(1,4)bisphosphate (IP3) production one can see that also at 10/~M Ca :+ IP2 and IP 3 liberated/mg protein/30min incubation are 4 + 0 . 2 and l l _ 0 . 5 n m o l respectively, in control nuclei. The prolactin effect at this Ca 2÷ concentration is still evident (i.e. 1.4 nmol IP 2 and 9.1 nmol IP 3 liberated/mg protein). Taking into account the input of labelled PIP and PIP 2 the recoveries of liberated IP 2 and IP 3 are 42 and 35%, respectively. Moreover, Table 2 indicates that also the whole mass of D A G is affected by prolactin treatment since this hormone is capable of reducing to one-half the nuclear D A G content.
[] /I/ ,// // // f/ // /./
80 - -
60 /V .+-/ // .// ,..-/ // /-/ //
/i // II // //
40
// f/ //
20
,// /-/
// f/ //
/-/ i
i
10 -s
i0-10
//-j./J // // // // // /i /i // // // // //
'i
t 10-12
FIG. 2. Histogram showing the effect of prolactin at different concentrations (reported below the x-axis) on the in vitro synthesis of nuclear polyphosphoinositides. Prolactin at 10 -s and 10 -14 behaves as at the concentration 10-5 M. The data are the mean of five separate experiments ( _ S.D.). isolated rat liver nuclei and that prolactin dramatically reduces the amount of IP2 liberated and to a lesser extent that of IP 3. The rate of hydrolysis is poorly affected by exogenously added Ca 2+ even if the optimal concentrations are those reported in Materials and Methods. TABLE 1. HYDROLYSISOF 3H-PIP AND 3H-PIP2 IN ISOLATED
RAT
LIVER
NUCLEI
ABSENCE
Conditions Control Prolactin 10-12 M
IN
TIlE
PRESENCE
OR
OF PROLACTIN
IP 2
IP 3
5 + 0.3
13 + 0.9
2-t-0.1"
11 -t-0.7"*
389
DISCUSSION The time and the concentration for prolactin stimulation of polyphosphoinositide synthesis in isolated nuclei are identical to those reported for prolactin stimulation of nuclear P K C [3], suggesting a prolactin-dependent intranuclear signalling mechanism. Moreover, the prolactin effects at nuclear level could be the consequence of the presence of prolactin receptors as suggested by experiments employing antiprolactin and anti-prolactin receptor antibodies [3]. It is worth mentioning that more recently it has been shown that translocation of prolactin occurs into the nucleus during interleukin-2stimulated mitogenesis [22]. The combined analysis of both synthesis and hydrolysis of nuclear inositol lipids as well as of TABLE 2. DAG BATED
Values are expressed as nanomoles of IPz and IP 3 liberated/mg nuclear protein/30 min incubation and are the means of seven separate experiments+ S.D. Ca 2+ requirement and pH show a broad range from 1 mM to 10#M and from 6.0 to 8.0, respectively, in which the activities are quite similar. The conditions reported in Materials and Methods are those in which the PLC activity is higher; however, prolactin effect is also evident at lower Ca ~÷ concentrations. *P < 0.001; **P < 0.01.
IN
UNDER
THE
CONTENT PRESENCE
OPTIMAL
Conditions Control Prolactin 10-e M
OF RAT LIVER NUCLEI OR
CONDITIONS
ABSENCE FOR
OF
PLC
INCU-
PROLACTIN ACTIVITY
DAG (pmol) 98 + 8 41 + 5
Values are expressed as picomoles of DAG/mg nuclear protein and are the means of three separate determinations +_S.D.
390
P.
SAwrI et al.
Control pulse-chase c.p.m./mg protein x 10-3 ~5~-0.24
Pulse
-Y///I/III//////II/////////I/II/A-
6.s7
Ji-0.453
0.183
Chase 15'
2.866 ~00 0"3 .146
[ ] PIP2
Chase 30' - l ' / / / / / / / / . ~ ' 1 . 7 8 6
' ' -T°'28~1 "~' 0
1
I 2
I 3
[]
I
I
4
5
PiP
[]
PA
I 6
I 7
Prolactin pulse-chas9 c.p.m./mg protein x 10-0 Pulse
~0.913 -~//'/////////////////////,/~//i///,////~
- 10.39
Chase 15' - ~ / ' / / / / / / / / / / / / / / / / , / , / / , / / / 1 "
7.583
-~-O .663 .48
Chase 30' n v / / / / / / / / / / / / / / / , / I / ' . / ~ r ,JL ~ . 0.54 I 0 2
I 4
6.52 [ 6
I 8
I 10
I 12
FIG. 3. Incorporation of 3~p from [~,32]p ATP into nuclear polyphosphoinositidcs in control and prolactintreated rat liver nuclei. Chase at 30°C was carried out by adding unlabelled ATP at the final concentration of 1 raM. The data arc the mean of five separate experiments ( + S.D.).
DAG mass indicate that prolactin is actually capable of inducing an increase of both PIP and PIP s. The pulse-chase experiments show that prolactin exerts its effect on nuclear polyphosphoinositide metabolism mainly by affecting the activity of inositol lipid monoesterases, which are less active under hormone treatment. Prolactin affects also the rate of hydrolysis of both PIP and PiPs inhibiting the PLC activity at the nucleus. Actually the PLC specific for PIP is inhibited to a higher extent and that gives rise to an increased level of PIP which is indeed available for the further synthesis of PIP2. Moreover, the results dealing with PLC activity are in good agreement with DAG mass measurement. In fact the reduced PLC-driven
hydrolysis of PIP and PIP 2 induced by prolactin faces the reduced amount of nuclear DAG. All in all, these data show that prolactin is capable of inducing at the nucleus an increase of both PIP and PIP s acting preferentially through an inhibition of their degradative enzymes, even if we cannot rule out that prolactin might also activate lipid kinases. It could seem contradictory that a decreased amount of inositol phosphates and DAG are concomitant to PKC activation by prolactin described by others [3]. However, our data, obtained by a combined analysis of synthetic and degradative pathways of nuclear polyphosphoinositides, clearly show that prolactin in isolated nuclei induc~ inhibition of both monoesterases and PLC specific for PiP and
Prolactin affects nuclear inositol lipid cycle PIP 2. Nevertheless a possible correlation stems from recent evidence showing that PIP2 activates P K C 50 times more efficiently than D A G competing with them for binding at the same activator region o f the regulatory domain of P K C [17, 18]. However, P K C activation during rat liver mitogenesis seems to be important since it has been reported also during regeneration after partial hepatectomy [23]. All in all, in our experimental conditions prolactin induces changes in nuclear polyphosphoinositide metabolism acting on both monoesterases and P L C specific for inositol lipids and this might represent a key step in the intranuclear signalling responsible for the onset of D N A synthesis during cell growth [13-15]. Acknowledgements--This work was supported by Italian CNR grants PF IG and PF BTBS. The authors wish to thank Dr G. PASQUINELLIfor his skilled assistance.
REFERENCES 1. Buckley A. R., Putman C. W., Montgomery D. W. and Russell D. H. (1986) Biochem. biophys. Res. Commun. 138, 1138-1145. 2. Capitani S., Girard P. R., Mazzei G. J., Kuo J. F., Berezney R. and Manzoli F. A. (1987) Biochem. biophys. Res. Commun. 142, 367-375. 3. Buckley A. R., Crowe P. D. and Russell D. H. (1988) Proc. natn. Acad. Sci. U.S.A. 85, 8649-8653. 4. Weiner R. I. and Bethea C. L. (1981) In Prolactin (Jaffee R. B., Ed.), pp. 19-55. Elsevier/ North-Holland, New York. 5. Nolin J. M. (1980) Peptides 1, 249-255. 6. Nishizuka Y. (1984) Nature 308, 693-698. 7. Berridge M. J. and Irvine R. F. (1989) Nature 341, 197-205.
391
8. Cocco L., Gilmour R. S., Ognibene A., Lctcher A. J., Manzoli F. A. and Irvine R. F. (1987) Biochem. J. 248, 765-770. 9. Capitani S., Cocco L., Maraldi N. M., Mazzotti G., Barnabei O. and Manzoli F. A. (1991) Adv. Enzyme Regul. 31, 103-124. 10. Cataldi A., Miscia S., Lisio R., Rana R. and Cocco L. (1990) FEBS Lett. 269, 465-468. 11. Manzoli F. A., Martelli A. M., Capitani S., Maraldi N. M., Rizzoli R., Barnabei O. and Cocco L. (1989) Adv. Enzyme Regul. 28, 25-34. 12. Cocco L., Martelli A. M., Gilmour R. S., Ognibene A., Manzoli F. A. and Irvine R. F. (1989) Biochem. biophys. Res. Commun. 159, 720-725. 13. Martelli A. M., Giimour R. S., Neff L. M., Manzoli L., Corps A. N. and Cocco L. (1991) FEBS Left. 283, 243-246. 14. Martelli A. M., Gilmour R. S., Falcieri E., Manzoli F. A. and Cocco L. (1989) Expl Cell Res. 185, 191-202. 15. Martelli A. M., Neff L. M., Gilmour R. S., Barker P. J., Huskisson N. J., Manzoli F. A. and Cocco L. (1991) Biochem. biophys. Res. Commun. 177, 480--487. 16. Martelli A. M., Gilmour R. S., Manzoli F. A. and Cocco L. (1990) Biochem. biophys. Res. Commun. 173, 149-155. 17. Chauhan V. P. S. and Brockerhoff H. (1988) Biochem. biophys. Res. Commun. 155, 18-23. 18. Chauhan A., Chauhan V. P. S., Deshmukh D. S. and Brockerhoff H. (1989) Biochemistry 28, 4952-4956. 19. Irvine R. F., Anggard E. A., Letcher A. J. and Dowries C. P. (1985) Biochem. J. 229, 505-511. 20. Divecha N. and lrvine R. F. (1990) In Methods in lnositide Research (Irvine R. F., Ed.), pp. 179-185. Raven Press, New York. 21. Garland R. C. and Cod C. F. (1972) Biochemistry 11, 4712-4718. 22. Clevenger C. V., Altman S. W. and Prystowsky M. B. (1991) Science 253, 77-79. 23. Martelli A. M., Carini C., Marmiroli S., Mazzoni M., Barker P. J., Gilmour R. S. and Capitani S. (1991) Expl Cell Res. 195, 255-262.
CHANGES NUCLEI
0898-6568/92 $5.1)0 + 0.00 © 1992 Pergamon Press Ltd
IN POLYPHOSPHOINOSITIDE IN RESPONSE
TO PROLACTIN, MITOGEN
LEVELS
IN RAT LIVER
A KNOWN
HEPATIC
P. SANrI,* A. M. MARTELLI,*R. S. GILMOUR,t E. FALCIERI,*R. RANA,~:A. CAa'ALDI,:~ G. LATTANZI,§ R. BAREGGIII and L. C o c c o * ¶
*Istituto di Anatomia Umana Normale, Universita' di Bologna, Italy, IIIstituto di Anatomia Umana Normale, Universta' di Trieste, Italy, :~Istituto di Morfologia Umana Normale, Universita' di Chieti, Italy, §Istituto di Citomorfologia del CNR, Bologna, Italy and tlnstitute of Animal Physiology and Genetics Research, AFRC, Babraham, Cambridge, U.K. (Received 23 January 1992; and accepted 17 March 1992) Abstract--The effect of prolactin action on nuclear polyphosphoinositide synthesis was investigated in isolated rat liver nuclei. An increased uptake of phosphate from [~32p]adenosinetrisphosphate was observed in both phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate with a maximum response at 10-~M concentration of hormone. Pulse-chase experiments in isolated nuclei following prolactin treatment indicate that the observed increase in accumulation of phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate is mainly due to a decrease in their rate of turnover possibly induced by a change in activity of polyphosphoinositide-specific monoesterases. In vitro prolactin also reduces the activity of nuclear phospholipase C specific for phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate. Moreover, this feature is strongly supported by the concomitant decrease in nuclear diacylglycerol mass. Thus these data suggest that once prolactin reaches the nucleus an intranuclear signalling is evoked through inositol lipid metabolism. Key words: Prolactin, polyphosphoinositides, rat liver nuclei. tically activated by the addition of prolactin [3]. This latter finding is supported by the suggestion that prolactin receptors are present in the nucleus [4, 5]. For its optimal activity, PKC requires phosphatidylserine (PS) as well as diacylglycerol (DAG) and free calcium ions which are liberated from cellular stores by inositoi (l,4,5)trisphosphate (IP3) [6, 7]. Moreover, the presence of inositol lipids with the potential for PKC activation in the nucleus of different cell types clearly indicates that these molecules play a role in the intranuclear signalling during cell differentiation and cell proliferation induced by growth factors [8-12]. Indeed, changes in nuclear polyphosphoinositide metabolism occur only in mitogenresponsive Swiss 3T3 cells and PKC translocation to the nucleus of these cells follows mitogen-induced changes of nuclear polyphosphoinositides [13-15]. Moreover, IP 3 stimulates PKC activity in nuclei from mitogen-treated
INTRODUCTION PROLACTINis an adenohypophyseal polypeptide hormone which exerts a mitogenic effect in rat hepatocytes [1]. There is some evidence that suggests that prolactin-dependent intracellular signalling involves the activation of protein kinase C (PKC) and that a significant part of this response is due to the activation of nuclear PKC. Indeed, highly purified, membranedeprived nuclei, as well as nuclear matrix from rat liver contain PKC [2], which can be drama¶Author to whom correspondence should be addressed at: Istituto di Anatomia Umana Normale, Via Imerio 48, 40126 Bologna, Italy. Abbreviations: ATP--adenosinetrisphosphate; BSA-bovine serum albumin; DAG---diacylglycerol; HPLC--high pressure liquid chromatography; IGF-l--insulin-like growth factor I; IP2--inositol( 1,4)bisphosphate; IP3--inositol(l,4,5)trisphosphate; PA--phosphatidic acid; PI--phosphatidylinositol; PIP--phosphatidylinositol 4monophosphate; PIP2--phosphatidylinositol 4,5-bisphosphate; PKC--protein kinase C; PS--phosphatidylserine; PLC--phospholipase C. 385
386
P. SANTIet al.
Swiss 3T3 fibroblasts [16]. On the other hand, phosphatidylinositol 4,5-bisphosphate (PIP2) itself can act as a regulatory molecule for PKC activity, since it may antecede D A G as activator of PKC interacting with the activatorreceiving region of the regulatory moiety o f this phosphorylative enzyme [17, 18]. Because of the evidence of the involvement of nuclear polyphosphoinositides in the mitogenic response we have sought to determine whether prolactin treatment could affect the in vitro synthesis of inositol lipids in rat liver nuclei. MATERIALS AND METHODS Preparation of rat liver nuclei
Rat liver nuclei preparation, in the presence of Triton X-100, was as previously described [8]. Phosphorylation of isolated nuclei and analysis of lipid extracts
All the procedures were as in Ref. [8]. Prolactin, solubilized in 10mM Tris-HC1, pHS.0, 1% BSA, was added directly to the phosphorylation mixture for the entire phosphorylation time (i.e. 4 min) at the concentrations indicated in the figure legends. Pulsechase experiments were carried out by incubating the nuclei in the presence of 2.5 /aCi [y32P]ATP (5000 Ci/mmol), final ATP concentration of 100/aM, for 4rain and then by adding a 10-fold excess of unlabelled ATP for 15 or 30 rain. Assay for PIP and PIP2 specific PLC
Phospholipase C (PLC) activity was assayed using 3 nmol of 3H-phosphatidylinositol 4-monophosphate (PIP) or 3H-PIP2 (90,000 d.p.m.) as exogenous substrates, 60/~g nuclear protein as enzyme source in the presence of 0.06% Taurodeoxicolate and by incubating for 30 min at 37°C in the presence or absence of prolactin. For PIP hydrolysis the buffer used was 100/~M Tris-HCl, pH 7.0, plus I mM CaCI2, whereas for PIP 2 the buffer used was 100/zM (2-[N-morpholino]ethanesulphonic acid) pH6.2, plus 100/~M CaCI2. Inositol phosphates liberated were recovered from the aqueous phase as in Ref. [8] and analysed by HPLC as described by Irvine et al. [19]. DAG assay
The total amount of nuclear DAG was measured in isolated nuclei after incubation in the same conditions as above as described by Divecha and Irvine [101.
Electron microscopy and cytoplasmic enzyme marker assay
Nuclear samples were fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, postfixed in 2% osmium tetroxide, dehydrated with acetone and embedded in Araldite. Thin sections stained with uranyl acetate and lead citrate were observed in a Zeiss El09 electron microscope to check the purity of the preparations. Nuclear preparations were also tested for glucose-6-phosphatase as described by Garland and Cori [21]. Routinely, this was less than 0.9% of the activity present in whole-cell homogenate. RESULTS Rat liver nuclei isolated by the procedure described are highly pure. There is no evidence of the outer nuclear envelope or of extranuclear debris when examined by electron microscopy (Fig. l), nor is there detectable glucose-6-phosphatase activity, a recognized cytoplasmic marker, confirming previous observations [8, 121. Incubation of purified rat liver nuclei shows the appearance of radiolabelled lipids identified, because of the exact coincidence of the radioactive spots with the internal standards, as phosphatidic acid (PA), PIP and PIP 2. The incorporation of radiolabelled phosphate came to a steady state after 4rain as previously shown in other nuclei [8, 12]. When prolactin was added to the incubation mixture the rate of synthesis of both PIP and PIP 2 was increased (Fig. 2). The dose-response from 10 _5 to 10 -~4 indicates that the effect is maximal at 10-~2M prolactin (Fig. 2). When pulse-chase experiments (Fig. 3) were carried out, the addition of excess unlabelled ATP chased out a large part of the incorporated 32p in both PIP and PIP z (83 and 61%, respectively). In the presence of 10-t2M prolactin the displacement of incorporated 3zp in PIP as well as in PIP 2 is less marked (i.e. 38 and 50%, respectively). In order to investigate whether PLC activities specific for PIP and PIP 2 are affected by prolactin we have set up an in vitro assay using 3H-labelled PIP and PIP 2 as specific substrates. Table 1 shows that these activities are present indeed in
FIG. 1. Electron microscopy of nuclear preparations. Rat liver nuclei isolated in the presence of Triton X-100. Both dispersed and condensed chromatin domains as well as interchromatin areas are visible. The inset shows at higher magnification the absence of the nuclear membrane (arrows). Scale bars = 1 pm.
387
Prolactin affects nuclear inositol lipid cycle Prolactin effect Percentage increase over the control 100
[] P,P []
Analysing the Ca 2+ dependency of 3H-inositol(1,4,5)trisphosphate (IP2) and 3H-inositol(1,4)bisphosphate (IP3) production one can see that also at 10/~M Ca :+ IP2 and IP 3 liberated/mg protein/30min incubation are 4 + 0 . 2 and l l _ 0 . 5 n m o l respectively, in control nuclei. The prolactin effect at this Ca 2÷ concentration is still evident (i.e. 1.4 nmol IP 2 and 9.1 nmol IP 3 liberated/mg protein). Taking into account the input of labelled PIP and PIP 2 the recoveries of liberated IP 2 and IP 3 are 42 and 35%, respectively. Moreover, Table 2 indicates that also the whole mass of D A G is affected by prolactin treatment since this hormone is capable of reducing to one-half the nuclear D A G content.
[] /I/ ,// // // f/ // /./
80 - -
60 /V .+-/ // .// ,..-/ // /-/ //
/i // II // //
40
// f/ //
20
,// /-/
// f/ //
/-/ i
i
10 -s
i0-10
//-j./J // // // // // /i /i // // // // //
'i
t 10-12
FIG. 2. Histogram showing the effect of prolactin at different concentrations (reported below the x-axis) on the in vitro synthesis of nuclear polyphosphoinositides. Prolactin at 10 -s and 10 -14 behaves as at the concentration 10-5 M. The data are the mean of five separate experiments ( _ S.D.). isolated rat liver nuclei and that prolactin dramatically reduces the amount of IP2 liberated and to a lesser extent that of IP 3. The rate of hydrolysis is poorly affected by exogenously added Ca 2+ even if the optimal concentrations are those reported in Materials and Methods. TABLE 1. HYDROLYSISOF 3H-PIP AND 3H-PIP2 IN ISOLATED
RAT
LIVER
NUCLEI
ABSENCE
Conditions Control Prolactin 10-12 M
IN
TIlE
PRESENCE
OR
OF PROLACTIN
IP 2
IP 3
5 + 0.3
13 + 0.9
2-t-0.1"
11 -t-0.7"*
389
DISCUSSION The time and the concentration for prolactin stimulation of polyphosphoinositide synthesis in isolated nuclei are identical to those reported for prolactin stimulation of nuclear P K C [3], suggesting a prolactin-dependent intranuclear signalling mechanism. Moreover, the prolactin effects at nuclear level could be the consequence of the presence of prolactin receptors as suggested by experiments employing antiprolactin and anti-prolactin receptor antibodies [3]. It is worth mentioning that more recently it has been shown that translocation of prolactin occurs into the nucleus during interleukin-2stimulated mitogenesis [22]. The combined analysis of both synthesis and hydrolysis of nuclear inositol lipids as well as of TABLE 2. DAG BATED
Values are expressed as nanomoles of IPz and IP 3 liberated/mg nuclear protein/30 min incubation and are the means of seven separate experiments+ S.D. Ca 2+ requirement and pH show a broad range from 1 mM to 10#M and from 6.0 to 8.0, respectively, in which the activities are quite similar. The conditions reported in Materials and Methods are those in which the PLC activity is higher; however, prolactin effect is also evident at lower Ca ~÷ concentrations. *P < 0.001; **P < 0.01.
IN
UNDER
THE
CONTENT PRESENCE
OPTIMAL
Conditions Control Prolactin 10-e M
OF RAT LIVER NUCLEI OR
CONDITIONS
ABSENCE FOR
OF
PLC
INCU-
PROLACTIN ACTIVITY
DAG (pmol) 98 + 8 41 + 5
Values are expressed as picomoles of DAG/mg nuclear protein and are the means of three separate determinations +_S.D.
390
P.
SAwrI et al.
Control pulse-chase c.p.m./mg protein x 10-3 ~5~-0.24
Pulse
-Y///I/III//////II/////////I/II/A-
6.s7
Ji-0.453
0.183
Chase 15'
2.866 ~00 0"3 .146
[ ] PIP2
Chase 30' - l ' / / / / / / / / . ~ ' 1 . 7 8 6
' ' -T°'28~1 "~' 0
1
I 2
I 3
[]
I
I
4
5
PiP
[]
PA
I 6
I 7
Prolactin pulse-chas9 c.p.m./mg protein x 10-0 Pulse
~0.913 -~//'/////////////////////,/~//i///,////~
- 10.39
Chase 15' - ~ / ' / / / / / / / / / / / / / / / / , / , / / , / / / 1 "
7.583
-~-O .663 .48
Chase 30' n v / / / / / / / / / / / / / / / , / I / ' . / ~ r ,JL ~ . 0.54 I 0 2
I 4
6.52 [ 6
I 8
I 10
I 12
FIG. 3. Incorporation of 3~p from [~,32]p ATP into nuclear polyphosphoinositidcs in control and prolactintreated rat liver nuclei. Chase at 30°C was carried out by adding unlabelled ATP at the final concentration of 1 raM. The data arc the mean of five separate experiments ( + S.D.).
DAG mass indicate that prolactin is actually capable of inducing an increase of both PIP and PIP s. The pulse-chase experiments show that prolactin exerts its effect on nuclear polyphosphoinositide metabolism mainly by affecting the activity of inositol lipid monoesterases, which are less active under hormone treatment. Prolactin affects also the rate of hydrolysis of both PIP and PiPs inhibiting the PLC activity at the nucleus. Actually the PLC specific for PIP is inhibited to a higher extent and that gives rise to an increased level of PIP which is indeed available for the further synthesis of PIP2. Moreover, the results dealing with PLC activity are in good agreement with DAG mass measurement. In fact the reduced PLC-driven
hydrolysis of PIP and PIP 2 induced by prolactin faces the reduced amount of nuclear DAG. All in all, these data show that prolactin is capable of inducing at the nucleus an increase of both PIP and PIP s acting preferentially through an inhibition of their degradative enzymes, even if we cannot rule out that prolactin might also activate lipid kinases. It could seem contradictory that a decreased amount of inositol phosphates and DAG are concomitant to PKC activation by prolactin described by others [3]. However, our data, obtained by a combined analysis of synthetic and degradative pathways of nuclear polyphosphoinositides, clearly show that prolactin in isolated nuclei induc~ inhibition of both monoesterases and PLC specific for PiP and
Prolactin affects nuclear inositol lipid cycle PIP 2. Nevertheless a possible correlation stems from recent evidence showing that PIP2 activates P K C 50 times more efficiently than D A G competing with them for binding at the same activator region o f the regulatory domain of P K C [17, 18]. However, P K C activation during rat liver mitogenesis seems to be important since it has been reported also during regeneration after partial hepatectomy [23]. All in all, in our experimental conditions prolactin induces changes in nuclear polyphosphoinositide metabolism acting on both monoesterases and P L C specific for inositol lipids and this might represent a key step in the intranuclear signalling responsible for the onset of D N A synthesis during cell growth [13-15]. Acknowledgements--This work was supported by Italian CNR grants PF IG and PF BTBS. The authors wish to thank Dr G. PASQUINELLIfor his skilled assistance.
REFERENCES 1. Buckley A. R., Putman C. W., Montgomery D. W. and Russell D. H. (1986) Biochem. biophys. Res. Commun. 138, 1138-1145. 2. Capitani S., Girard P. R., Mazzei G. J., Kuo J. F., Berezney R. and Manzoli F. A. (1987) Biochem. biophys. Res. Commun. 142, 367-375. 3. Buckley A. R., Crowe P. D. and Russell D. H. (1988) Proc. natn. Acad. Sci. U.S.A. 85, 8649-8653. 4. Weiner R. I. and Bethea C. L. (1981) In Prolactin (Jaffee R. B., Ed.), pp. 19-55. Elsevier/ North-Holland, New York. 5. Nolin J. M. (1980) Peptides 1, 249-255. 6. Nishizuka Y. (1984) Nature 308, 693-698. 7. Berridge M. J. and Irvine R. F. (1989) Nature 341, 197-205.
391
8. Cocco L., Gilmour R. S., Ognibene A., Lctcher A. J., Manzoli F. A. and Irvine R. F. (1987) Biochem. J. 248, 765-770. 9. Capitani S., Cocco L., Maraldi N. M., Mazzotti G., Barnabei O. and Manzoli F. A. (1991) Adv. Enzyme Regul. 31, 103-124. 10. Cataldi A., Miscia S., Lisio R., Rana R. and Cocco L. (1990) FEBS Lett. 269, 465-468. 11. Manzoli F. A., Martelli A. M., Capitani S., Maraldi N. M., Rizzoli R., Barnabei O. and Cocco L. (1989) Adv. Enzyme Regul. 28, 25-34. 12. Cocco L., Martelli A. M., Gilmour R. S., Ognibene A., Manzoli F. A. and Irvine R. F. (1989) Biochem. biophys. Res. Commun. 159, 720-725. 13. Martelli A. M., Giimour R. S., Neff L. M., Manzoli L., Corps A. N. and Cocco L. (1991) FEBS Left. 283, 243-246. 14. Martelli A. M., Gilmour R. S., Falcieri E., Manzoli F. A. and Cocco L. (1989) Expl Cell Res. 185, 191-202. 15. Martelli A. M., Neff L. M., Gilmour R. S., Barker P. J., Huskisson N. J., Manzoli F. A. and Cocco L. (1991) Biochem. biophys. Res. Commun. 177, 480--487. 16. Martelli A. M., Gilmour R. S., Manzoli F. A. and Cocco L. (1990) Biochem. biophys. Res. Commun. 173, 149-155. 17. Chauhan V. P. S. and Brockerhoff H. (1988) Biochem. biophys. Res. Commun. 155, 18-23. 18. Chauhan A., Chauhan V. P. S., Deshmukh D. S. and Brockerhoff H. (1989) Biochemistry 28, 4952-4956. 19. Irvine R. F., Anggard E. A., Letcher A. J. and Dowries C. P. (1985) Biochem. J. 229, 505-511. 20. Divecha N. and lrvine R. F. (1990) In Methods in lnositide Research (Irvine R. F., Ed.), pp. 179-185. Raven Press, New York. 21. Garland R. C. and Cod C. F. (1972) Biochemistry 11, 4712-4718. 22. Clevenger C. V., Altman S. W. and Prystowsky M. B. (1991) Science 253, 77-79. 23. Martelli A. M., Carini C., Marmiroli S., Mazzoni M., Barker P. J., Gilmour R. S. and Capitani S. (1991) Expl Cell Res. 195, 255-262.
