- Email: [email protected]
Changes of Nuclear PI-PLC γ1 During Rat Liver Regeneration
Cell. Signal. Vol. 9, No. 5, pp. 353–362, 1997 Copyright 1997 Elsevier Science Inc.
ISSN 0898-6568/97 $17.00 PII S0898-6568(96)00178-7
Changes of Nuclear PI-PLC g1 During Rat Liver Regeneration Luca M. Neri,†‡ Daniele Ricci,† Cinzia Carini,† Marco Marchisio,† Silvano Capitani† and Valeria Bertagnolo†* †Signal Transduction Unit/Laboratory of Cell Biology, Institute of Human Anatomy, University of Ferrara, Ferrara, Italy and ‡Institute of Normal and Pathological Cytomorphology C. N. R., c/o I. O. R., Bologna, Italy
ABSTRACT. We have previously demonstrated that rat liver nuclei contain PI-PLC b1 and g1 in the inner nuclear matrix and lamina associated with specific phosphodiesterase activity (Bertagnolo et al., 1995, Cell Signall. 7, 669–678). Since compensatory hepatic growth is an informative and well characterized model for natural cell proliferation, the presence of specific PI-PLC isoforms and their activity as well as PIP2 recovery were studied at various regenerating times, ranging from 3 to 22 h after partial hepatectomy. Three PI-PLC isoforms (b1, g1, d1) were examined in control and regenerating liver cells by using specific antibodies. By means of in situ immunocytochemistry and confocal microscopy, PI-PLC b1 was found mainly in the nucleoplasm and this pattern was not modified after hepatectomy. On the contrary, the nuclear g1 isoform showed a marked decrease at 3 and 16 h after hepatectomy, but a clear increase at 22 h covering with bright intensity the whole nucleus. The PI-PLC d1 isoform, which is exclusively cytoplasmic, was not altered during rat liver regeneration. By western blotting analysis on whole cell homogenates, none of the PI-PLC isozymes under study showed proliferation-linked modification. However, analyses of isolated nuclei identified changes in the nucleus associated PI-PLC g1 that paralleled the in situ observation whereas the b1 isoform was unmodified at all the times examined. Nuclear phosphodiesterase activity on PIP2 was lower at 3 and 16 h, in comparison with sham operated rats, increased at 6 h and reached the highest value after 22 h. Consistently, the recovery of PIP2, obtained in conditions that optimise PIP-kinase activity, showed a marked decrease at 3 h and an increase up to 16 h of liver regeneration, followed by a further decrease at 22 h. These data are consistent with a close relationship between cell proliferation and the nuclear inositide cycle, depending, in rat liver, predominantly on the modulation of the g1 isoform of PI-PLC. cell signal 9;5:353–362, 1997. 1997 Elsevier Science Inc. KEY WORDS. Nucleus, Inositide cycle, Phosphoinositidases, Rat liver, Liver regeneration, Confocal microscopy
INTRODUCTION The initial event of inositol lipid turnover for transmembrane signalling is the phosphodiesteratic cleavage of phosphatidylinositol 4,5-bisphosphate (PIP2). This reaction is catalysed by a phosphoinositide specific phospholipase C (PI-PLC) to yield inositol 3,4,5-trisphosphate (IP3) and diacylglycerol (DAG) [1]. The three major classes of PI-PLC identified (b, g, d) are tissue and cell type specific [2, 3], but all three subtypes require the presence of Ca11 for full activation and are capable of the hydrolysis of the three major inositol lipids with the selectivity for PIP2 over PIP and PI decreasing in the order of PI-PLC b, g and d [4]. *Address all correspondence to Valeria Bertagnolo, Institute of Human Anatomy, University of Ferrara, Ferrara, Italy. Abbreviations: PI-PLC—phosphoinositide-specific phospholipase C; PI—phosphatidylinositol; PIP—phosphatidylinositol 4-monophosphate; PIP2—phosphatidylinositol 4,5-bisphosphate; PKC—protein kinase C; DAG—diacylglycerol. Received 22 August 1996; and accepted 30 October 1996.
Distinct mechanisms of signal transduction are involved in the activation of PI-PLC following agonist occupancy of phosphoinositide linked receptors. Receptors with intrinsic tryosine kinase activity, such as those for PDGF (Platelet Derived Growth Factor) and EGF (Epidermal Growth Factor), activate PI-PLC g by phosphorylating its tyrosine residues [5], whereas bradykinin, vasopressin and other receptors activate PI-PLC b through an intervening guanine nucleotide binding protein (G protein) of the aq or bg class [6, 7, 8, 9]. The regulation of the d isoforms is still not clear. The signalling mechanism based on PI-PLC activity is not confined to the cytoplasm. Evidence for a nuclear inositol lipid cycle was first reported by Smith and Wells in nuclei purified from rat liver [10]. Subsequently it was demonstrated that in Friend erythroleukemia cell nuclei, PIP2 is formed by phosphorylation of PIP when cells undergo terminal differentiation [11]. In Swiss 3T3 cells, within minutes of insulin-like growth factor I (IGF-I) treatment, the nuclear mass of PIP and PIP2 halved, while that of diacyl-
354
glycerol (DAG) doubled by using picomole-sensitive mass assay [12]. Many other works pointed to a nuclear phosphoinositide cycle entirely separate from the known one in the plasma membrane (for review see [13, 14]. The decrease in nuclear PIP and PIP2 accompanied by an increase in DAG is coincident with the activation of phosphoinositidase C in the nucleus. Phosphoinositidase C activity in nuclei of Swiss 3T3 cells increases four fold after IGF-I treatment [15] and two fold in osterosarcoma SaOS-2 cells after Interleukin-I (IL-1) stimulation [16]. This increase of PLC activity was only dependent on the b1 isoform, which was confined to the nucleus [15, 16]. This observation was extended by other works that showed the nuclear localization and activity of PLCb1 in nuclei of rat pheochromocitoma PC12 and Friend erythroleukemia cells [17, 18]. Specific nuclear phosphoinositide cycle and PLC activity has been described also in rat liver. It was first demonstrated that the cell nucleus is a site of lipid phosphorylation, not involving enzymes and substrates located on the nuclear membrane [19]. The incorporation of PI into rat liver nuclei was greatly stimulated by phosphatidylinositol transfer protein (PI-TP) and nuclei were able to phosphorylate PI, with the production of PIP and PIP2 [20]. The first evidence of a PI-PLC isoform in rat liver nuclei was demonstrated by Divecha et al. [21], which showed that the only isoform specifically, but not exclusively, located in the cytoplasm was PI-PLCb1. Recently the subcellular distribution of PI-PLCb1, g1 and d1 has been investigated in the same model, indicating that the b1 isoform is predominantly nuclear, whereas the g1 isozyme is largely cytoplasmic and PLCd1 is restricted to the cytoplasm [22]. The assay of phosphodiesterase activity showed a discrete intracellular distribution and, by specific inhibition with neutralising anti PI-PLCb1 and g1 antibodies, this enzyme activity was confirmed at the nuclear level. Although some studies have been carried out, it has not yet been clearly identified which PI-PLC isoforms are involved in the nuclear inositol lipid cycle during rat liver regeneration nor their nuclear localisation at various times post hepatectomy. Therefore we have investigated the behaviour of these isozymes in the liver compensatory growth response after partial hepatectomy. The results indicate that the selective localization of PIPLC isozymes is accompanied, after partial hepatectomy, by changes of a specific isoform, PI-PLCg1. An immunocytochemical analysis in combination with confocal microscopy showed the increase of PI-PLCg1 nuclear content both on tissue sections and isolated nuclei. Immunochemical analysis demonstrated an identical increase in isolated nuclei of PI-PLCg1. These changes have been related to IP3 production due to phosphodiesterase activity and to PIP2 recovery in conditions that optimised PIP-kinase activity both before and during DNA synthesis period. MATERIALS AND METHODS Source of Materials Unless otherwise indicated, all reagents were obtained from the Sigma Chemical Co. (St. Louis, MO).
L. M. Neri et al.
Partial Hepatectomy Male adult Wistar rats (150–200 g body wt) were operated on according to Higgins and Anderson [23]. Approximately two thirds of the liver was surgically removed, and shamoperated rats were used as control for normal livers. The rats were sacrificed after 3, 6, 16, and 22 h, and livers were removed and used for preparation of total homogenate and isolation of nuclei. Preparation of Rat Liver Fractions Membrane-depleted nuclei were isolated from normal and regenerating rat livers as previously described [19], in the presence of 0.2% Triton X-100. All the purification buffers contained 10 mg/ml Aprotinin, 25 mg/ml Leupeptin, 50 mg/ ml STI, 1 mM PMSF, 0.5 mM DTT and 1 mM Na 3VO4, 15 mg/ml Calpain inhibitor I and 30 mg/ml Calpain inhibitor II (Calbiochem, La Jolla, CA). Total homogenates and cytoplasmic fractions were obtained as previously described [22]. Ultrastructural analysis and assays of cytoplasmic enzyme markers for assessment of nuclear purity were performed as previously reported [22]. Immunochemical Analysis Proteins from total homogenate and nuclei (50 mg), were separated on 7.5% polyacrylamide denaturing gels [24] and blotted to Trans-Blot nitrocellulose membrane (Bio Rad Laboratories, Richmond, CA). Monoclonal antibodies against the three isoforms b1, g1 and d1 of PI-PLC, originally developed by Suh et al., [25] employed as previously described [17] and a monoclonal antibody anti Phosphotyrosine (PTyr) were purchased from UBI (Lake Placid, NY). Peroxidase-conjugated anti-mouse IgG (Amersham International, Bucks U. K.) was used as secondary antibody, and the final detection was performed using the ECL system (Amersham), according to the manufacturer’s instructions. Densitometric analysis was done with an imaging densitometer BioRad PhosphorImager (Model GS 670, BioRad, U. K.) using the Molecular Analyst software. Each lane was analysed on the basis of specific signal i. e. bands with peaks of positivity and regions with low or negligible positivity. Regions were arbitrarily chosen for each sample to optimize the analysis of each immunoreactive band. Within each lane, data were averaged on the basis of 10 measurements, performed on the same band. For statistical evaluation the results were expressed as means 6 standard deviations of four different experiments. Standard deviation was in all curves less than 9%. As a further control we have compared different means from the same preparation, and we have calculated the p value. This was always . 0.05 and therefore not significant. Immunofluorescence and Confocal Microscopy Analysis Samples of rat liver, 4.5 mm thick, were quickly frozen in liquid nitrogen and sectioned in a Lauda 1720 cryostat (Leica, Germany). Cryo-sections were fixed in 4% paraformaldehyde/PBS followed by a mild permeabilisation with 0.2% Triton for 8 min. Per sample, 1 3 106 isolated nuclei
Nuclear PI-PLC c1 Increase after Hepatectomy
were dropped on glass microscope slides coated with polyL-lysine (Sigma, USA), allowed to stick for 30 min at room temperature and fixed in 4% paraformaldehyde/PBS. Tissue sections and isolated nuclei were then treated as follows: slides were incubated to block non specific binding with 2% BSA, 3% NGS in PBS (immunoreaction buffer) for 30 min at 378C and then reacted with primary antibodies anti PI-PLC isoforms diluted 1:50 in immunoreaction buffer for 3 h at 378C. After three washing steps, primary antibody binding was visualised by incubating the specimens with FITC-conjugated goat anti-mouse IgG (Sigma) diluted 1:100 for 1h at 378C. Slides were then incubated with PBS containing DAPI (Sigma) to assess the nuclear domains and embedded in glycerol containing 1,4-diazabicyclo [2.2.2] octane (DABCO) (Sigma) to retard fading, using additional coverslips as spacers to preserve the threedimensional structure of tissue sections or of nuclei. A variety of controls was performed to ensure the specificity of the antibody used: coverslips were treated with preimmune mouse or rabbit serum as well as with the secondary antibody alone. In all cases negligible levels of fluorescent signal were observed (not shown). All the experiments were done at least in triplicate. To define the nuclear region, the tissue sections were counterstained with 1 mg/ml Propidium Iodide (PI). The samples were analysed by a confocal laser scanning microscope Phoibos 1000 (Molecular Dynamics, USA) coupled with a Nikon Optiphot equipped with a 100 3 oil immersion lens (NA 5 1.4). The instrument was equipped with an argon laser 488/514 nm tuned at 10 mW and decreased by a neutral density filter to 30% of its power. On the illumination pathway a dichroic mirror (500 nm) was used, while on the detection pathway a 515 OG filter was employed. For double stained specimens a second dichroic mirror (550 nm) was placed in front of the two photomultipliers (PMT) in combination with a 530 6 15 nm or a 590 6 15 nm band pass filter for FITC or PI respectively. The PMT gains were set at 934 for FITC fluorescence and 786 for PI fluorescence and settings were rigorously maintained for all experiments. Image acquisition, recording, filtering and reconstruction were performed on z series of confocal data (stacks) by an Indy 4400 graphic workstation (Silicon Graphics, USA) as previously described [26, 27].
PIP2-Specific Phospholipase C Activity Total homogenate and nuclei (50 mg of protein) were incubated in a final volume of 100 ml for 15 min at 378C in the presence of 3 nmols of PIP2, 30,000 dpm [3H] PIP2, 0.06% tauro-deoxycholate, 0.6% NaCl, 0.1 M Mes, pH 6.2, 0.1 mM CaCl2. The reaction was stopped by addition of the lipid extraction mixture and the hydrosoluble products were counted as previously described [17]. For detection of IP3 the aqueous phases were loaded on a Partisil 10 SAX column, 25 cm length, from Whatman, then eluted with a linear gradient of ammonium formate (0–2 M, pH 3.7 with phosphoric acid) [28]. Elution peak of IP3 was identified by comparison with authentic standard. Radioactivity was determined by liquid scintillation
355
counting. Statistical analysis was performed using the two tailed Student’s t test for unpaired data. In Vitro PIP Phosphorylation Normal and regenerating rat liver nuclei (200 mg of protein) were preincubated for 10 min at 308C in the presence or absence of 0.05% Triton X-100, 100 mM PIP, then incubated for 15 min with 15 mM MgCl2, 10 mM b-mercaptoethanol, 1 mM Na3VO4, 100 mM ATP and 10 mCi [g-32P] ATP, in a final volume of 200 ml. The reaction was terminated by adding chloroform/methanol/HCl (200:100:0.75, v/v) and the phosphoinsitides were extracted according to Shaikh and Palmer [29]. After extraction, the organic phases were developed on TLC plates and the phosphoinositides were scraped off and counted by liquid scintillation. Statistical analysis was performed using the two tailed Student’s t test for unpaired data. RESULTS Immunochemical Analysis As demonstrated in our previous work concerning the distribution of PI-PLC in rat liver fractions [22], both b 1 and g1 isoforms were present in isolated nuclei. The two isozymes were investigated by Western Blot analysis in whole cell homogenates and nuclei obtained from normal and regenerating rat liver. In total homogenates no differences on the amount of both B1 and g1 isozymes were shown (Fig. 1A). In isolated nuclei the signal corresponding to the b1 isoform was similar at all the explored regeneration times, while the amount of the g1 isoenzyme changed. The latter isoform lowered after 3 h, raised to control levels at 6 h, lowered again 16 h from hepatectomy and showed the maximum increase after 22 h, compared with sham operated conditions (Fig. 1B). The amount of phosphorylated PI-PLCg1, as detected by anti-PTyr antibody, paralleled the decrease of the whole nuclear isozyme (Fig. 1C) at three and 16 h and the increase after 22 h from hepatectomy. As determined by densitometric scanning of fluorograms, we have analysed several experiments and additional timepoints. Table 1 shows that at 90 min., 2 and 3 h there is a relevant decrease of two folds of PI-PLCg1 at the nuclear level whereas both the cytoplasmic and the whole homogenates were unchanged. At 6 and 8 h time after partial liver resection no differences were appreciated in comparison with unoperated rats. At 14, 16 and 18 h after hepatectomy the level of intranuclear PI-PLCg1 were markedly reduced by two fold and the immunobands obtained at 22 and 26 hours were both increased by about 65%. The d1 isoform, present exclusively at cytoplasmic level, was also investigated but no differences associated with liver regeneration were shown (data not shown). Confocal Analysis on Isolated Nuclei To analyze in more details the distribution of the nuclear isoforms identified by Western Blot, the same time-course study was performed by immunocytochemical analysis on isolated
356
nuclei. Optical sections were chosen from confocal microscopy data sets about 3 mm from the bottom side of the nuclei attached to the glass surface, which corresponded to their equatorial plane.
FIGURE 1. Immunochemical analysis of PI-PLCb1 and g1 in total
homegenate (A) and isolated nuclei (B) at different hours from partial hepatectomy. In (C) P-Tyr labelling of the PI-PLCg1 bands. Molecular weights are indicated in KDa. See description in text.
L. M. Neri et al.
The staining of PI-PLCb1 was substantially unchanged at the various times after hepatectomy (Fig. 2A–E). The pattern was characterized by an intense labelling in all the nucleoplasm with very fine tiny rounded dots that allowed us to identify the nucleolar areas that appeared negative. Sometimes in the central part of the nucleus, especially at 6, 16 and 22 h (Fig. 2C–E), large areas of homogeneous labelling were observable. PI-PLCg1 immunofluorescence stained with a fine punctate labelling the whole nucleoplasm of control nuclei. Nucleoli could be observed as negative unstained areas. The labelling was present also at the nuclear periphery even if with a lower degree of intensity (Fig. 2F). A very consistent decrease of staining occurred after 3 h (Fig. 2G). In the nuclear interior few large fluorescent spots and some tiny dots were observable with an irregular distribution, showing many different nuclear regions almost negative. Some labelling was increased at the periphery, with a ring-like aspect that resembled the nuclear lamina and a fluorescent mass was located in contact with the peripheral staining. After 6 h the labelling resembled the levels of control nuclei (Fig. 2H). Fluorescence was regularly distributed in the nucleus and some addensation was seen at the periphery. A second very evident decrease was observable after 16 h from the hepatectomy (Fig. 2I): the immunofluorescence was organised in a few small dots present mainly in the central part of the nucleoplasm, except for the nucleolar area. A very weak labelling was observable at the nuclear periphery, except for a few irregular fluorescent areas, some of which were located at the nuclear edge. The 22 h sample was characterized by a strong increase in PI-PLCg1 immunoreactivity that acquired a new aspect of labelling (Fig. 2L): a large central area was characterized by a fairly homogeneous pattern of very brilliant fluorescence that that did not allow the recognition of single spots or dots. In fact, the fine granular pattern of control nuclei was replaced by large areas of labelling emerging from a brilliant fibrogranular meshwork. The nucleoli and the periphery of the nucleoplasm were labelled with lower intensity, but still higher than the control nuclei. In some cases at the nuclear periphery some very large fluorescent spots more brilliant than the surrounding labelling were seen. To verify the described PI-PLC distribution, confocal analysis on tissue sections was performed on two sample times (0 and 22 h after partial hepatectomy) as a further control. Control liver sections when probed with the anti PI-PLb1 antibodies displayed an intense immunofluorescence characterized by a fine puncatate labelling, that stained the whole nucleoplasm with negative nucleoli and also partially stained the cytoplasm (Fig. 3A). The partial hepatectomy did not modify intensity and distribution pattern of the labelling after 22 h (Fig. 3B). Cytoplasm and nucleus were both labelled by anti-PI-PLCg1 antibodies in liver cells and DNA was counterstained with PI. Fine little dots were distributed in the nucleoplasm and larger fluorescent areas were often observable both in the nuclear interior and at its periphery (Fig. 3C). The cytoplasmatic labelling was un-
Nuclear PI-PLC c1 Increase after Hepatectomy
357
TABLE 1. Densitometric analysis of immunoreactive PI-PLC g1 and P-Tyr in total homogenates (HT), cytoplasmic fration
(Cyt) and isolated nuclei (Nu) at different regenerating timepoints: data are expressed as percentage variations of control conditions, detected by densitometric scanning, as described in the Results section Hours Anti Anti Anti Anti
PI-PLC g1 HT PI-PLC g1 Cyt PI-PLC g1 Nu P-Tyr Nu
0
909
2
3
6
8
14
16
18
22
26
100 100 100 100
98 103 50* 37*
101 105 49* 40*
105 101 47* 35*
106 103 95 95
100 104 97 97
98 99 51* 50*
96 99 46* 43*
99 100 47* 40*
102 104 161* 150*
100 103 155* 156*
Data are the mean from four different experiments. For each of the antibodies, the staining patterns of total homogenates, cytoplasmic fraction and isolated nuclei were compared with the unoperated liver, that served as a control. * Highly significant differences (p , 0.001). All of the other differences were found to be not significant with p , 0.05.
changed after hepatectomy, whereas the nuclear staining showed a marked increase after 22 h of liver regeneration corresponding to that described in insolated nuclei: the whole nucleoplasm was labelled displaying large very brilliant areas including also its periphery (Fig. 3D). PI-PLCd1 showed a cytoplasmatic localization of a discrete and punctate immunofluorescence (Fig. 3E), that did not show appreciable changes 22 h after partial hepatectomy (Fig. 3F). Phosphodiesterase Activity
in Table 2, the recovery of PIP2 in conditions optimising PIP-kinase activity after 3 h was reduced to 25% of control value. Starting from 6 h of liver regeneration, the recovery of PIP2 increased progressively until 16 h post hepatectomy. The analysis during the S phase (22 h) showed a moderate increase of the activity in respect to sham operated samples, but a marked decrease (230%) when compared to the 16 h sample.
DISCUSSION
The percentage of tritiated PIP2 cleaved by the isozymes present in liver nuclei from sham operated and hepatectomized rats is shown in Table 2. In the presence of PIP2 as substrate, a significant decrease of hydrolysis was revealed at 3 h from hepatectomy and returned similar to control values at 6 h. Starting from the 16th hour of liver regeneration, PIP2 decreased 40% after 16 h and reached the highest value, 50% higher than in control nuclei, at 22 h (Table 2). To assess the fraction of hydrosoluble compounds generated by PIP2 hydrolysis, some aqueous phases were analysed by HPLC. Consistent with the whole hydrosoluble counts, the recovery of IP3 was decreased after 16 h from hepatectomy and at 22 h was significantly increased of 80% than sham operated nuclei (Table 3). Phosphoinositide Kinase Activity PIP-kinase activity was monitored by assaying the phosphorylation of exogenous PIP at the different regeneration times (0, 3, 6, 16, 22 h from partial hepatectomy). As shown
Regenerating rat liver is a well characterized and informative model of cell proliferation in vivo. By means of confocal microscopy we have shown for the first time in situ modifications of PI-PLCg1 isoforms in liver tissue sections and isolated nuclei induced by partial hepatectomy. The in situ analysis by confocal microscopy on membrane depleted isolated nuclei was chosen to evaluate the real presence of PIPLCb1 and PI-PLCg1, within the nucleoplasm without interfering with the contribution of cytoplasmic or membrane located enzyme. The changes of PI-PLCg1 isoform were found in isolated nuclei and in tissue sections only at the nuclear level, whereas the cytoplasm appeared unchanged. Of the other isoforms analysed in this study, PI-PLCd1, which resulted to be exclusively cytoplasmic, and PI-PLCb1, mostly located in the inner nucleoplasm, were unaffected during cell proliferation. Main features obtained from in situ analysis of PI-PLCg1 are the g1 immunostaining resembling a fibrogranular meshwork with both lamina- and nucleoskeletal-like structures [30, 27],
TABLE 2. Levels of specific PIP2 phosphodiesterase activity in conditions optimising PI-PLC activity (PIP2 hydrolysis)
and of PIP kinase activity (PIP2 recovery) in rat liver nuclei at different regeneration time H PIP2 hydrolysis (nmol/mg proteins) Sham operated Hepatectomized PIP2 recovery (pmol/mg proteins) Sham operated Hepatectomized
0
3
6
16
22
16.0 6 1.2 —
15.5 6 1.4 9.6 6 0.94*
16.5 6 1.1 19.7 6 1.5
16.2 6 1.4 10.5 6 0.83*
15.4 6 1.3 23.5 6 1.9*
1.8 6 0.22 3.7 6 0.32*
2.3 6 0.21 2.7 6 0.18
2.3 6 0.17 —
1.9 6 0.16 0.6 6 0.03*
2.2 6 0.19 2.9 6 0.23
The values 6 SD are the mean from three separate determinations on lipid fractions from isolated rat liver nuclei. * Significant differences (p , 0.005 Student’s t test) from control.
358
L. M. Neri et al.
FIGURE 2. Optical sections through isolated nuclei placed on glass slides and immunostained for PI-PLCb1 (A–E) or PI-PLCg1 (F–L)
at various times during rat liver regeneration: 3 h (B, G), 6 h (C, H), 16 h (D, I) and 22 h (E, L). PI-PLCb1 (A) control nucleus. A fine granular labelling can be easily appreciated as well as the nucleolus. The nuclear periphery shows a weaker staining than the nuclear interior. (B) nucleus 3 h after hepatectomy. PI-PLCB1 staining is similar to A. (C) nucleus 6 h after hepatectomy. In the central
Nuclear PI-PLC c1 Increase after Hepatectomy
359
TABLE 3. Recovery of IP3 separated by HPLC after PIP2 specific hydrolysis
in rat liver nuclei at different regeneration time H IP3 recovery (nmol/mg proteins) Sham operated Hepatectomized
0
16
22
11.5 6 1.2
12.1 6 0.9 7.1 6 0.7*
11.0 6 1.1 19.4 6 1.9*
The values 6 SD refer to the mean from three separate determinations on lipid fractions from isolated rat liver nuclei. * Significant differences (p , 0.005 Student’s t test) from control.
its decrease at 16 h, before the S phase, and its large increase in the nucleoplasm at 22 h, that strongly associates this variation with cell proliferation induced by partial hepatectomy. The PI-PLC isoforms’ association with nucleoskeletal structures has been detected at the same sites in different cell lines [19, 31, 32, 33]. Interestingly, all these reports support the notion that all inositide-metabolising enzymes are in the internal nuclear matrix, except for PI-kinase, which is in the peripheral matrix [31]. In rat liver nuclei the detection of the nuclear isoforms PI-PLCb1 and g1 was not affected by the removal of the nuclear membrane and the two enzymes showed a tight association with nuclear matrix and lamina [22]. Western blot analysis strongly supported the in situ evidence that only the nuclear PI-PLCg1 isoform changed, whereas the whole cells and the cytoplasmic fraction did not show variations. After 3 and 16 h a decreased nuclear presence is consistent with a reduced synthesis of the entire cellular enzyme, whereas after 22 h of liver regeneration this PI-PLC isozyme may be newly synthesized and translocated from cytosol to nucleus in response to mitogenic stimuli. From the analysis by anti P-Tyr antibody it emerged that there is a constitutively phosphorylated form of the enzyme inside the nucleus and that it is this phosphorylated enzyme that changes during the different regeneration timepoints. Interestingly, these oscillations are sustained for a certain period of time, which appears to characterize the proliferation event as forming different phases that requires the presence, or not, of phosphorylated PI-PLCg1 within the nucleus. An increase of phosphorylation, that occurs only for PI-PLCg1 but not for PI-PLCb1 or PI-PLCd1 is a well known system of PI-PLCg1 activation by growth factor receptor protein-tyrosine kinase and is correlated with the stimulation of PIP2 hydrolysis [34]. However, very little is known about the molecular events that occur after the tyrosine phosphorylation of PI-PLCg1.
A recent paper described a novel PI-PLC isozyme, the d4 isoform, that is expressed more remarkably in regenerating liver than in normal resting liver as well as in Swiss 3T3 fibroblasts nuclei during the S phase [35]. In agreement with our data, the same authors observed that PI-PLCb1 content did not vary. However, they did not detect PI-PLCg1 changes. The difference, in the latter isozyme may be due to the different antibodies used, cell type and model (in vitro vs. in vivo). A unique 88 kDa form of PI-PLC has been purified only in regenerating liver nuclei, but the authors failed to identify which PLC isozyme corresponded to this detected species [36]. A number of reports have addressed the issue of an isotype specific involvement of the PI-PLC family in the cell response to different stimuli. Resting NIH 3T3 fibroblasts were induced to synthesise DNA after PI-PLCb or g injection, and proliferation was inhibited when anti PI-PLC g antibodies were injected, showing that PI-PLC isoforms are related to growth arrest and differentiation [37]. Mouse erythroleukemia cells, when treated with DMSO, show a decrease in nuclear PI-PLC activity and in the levels of nuclear DAG associated with a decrease of the b1 isoform, as demonstrated by the use of isoform specific antibodies [18, 38]. Differentiation-linked decrease of PI-PLCs was also described on U937 human histiocytic leukemia cells in which a marked and selective decrease in enzyme activity and protein level of PI-PLCg1 was observed, while PI-PLCg2 was not affected [39]. The nuclear inositide cycle is specifically affected by stimuli that promote proliferation. Agonist stimulation of the IGF-I receptor tyrosine kinase activity does not lead to PIP2 breakdown or DAG production at the plasma membrane, but rather in the nucleus [12, 15], suggesting a possible link between nuclear signalling, PI-PLC isoforms and the onset of DNA synthesis [40]. In our system, the cellular proliferation induced by regen-
part of the nucleoplasm the spots are fused together to form large and brilliant fluorescent areas. The nucleoli are still observable. (D and E) nuclei at 16 or 22 h of liver regeneration, respectively. The fluorescent pattern is very similar to C. PI-PLCg1 (F) control nucleus. The whole nucleoplasm is labelled by a fine granular pattern more intense in the center than at the nuclear periphery. Nucleoli appear negative. (G) nucleus 3 h after hepatectomy. A scarce fluorescence can be observed except for few large and brilliant irregular areas (arrow). Note the presence of ring-like fluorescence at the periphery. (H) nucleus 6 h after hepatectomy. The original fluorescence intensity is restored with a pattern very similar to f, with a slight accumulation at the periphery. (I) nucleus at 16 h of liver regeneration. A strong decrease of the labelling is displayed. Some irregular faint spots are visible together with two larger areas (arrows), one of which is in close association with the nuclear periphery. (L) nucleus at 22 h of liver regeneration. An intense increase of the immunofluorescence is observable in the nuclear interior. The periphery of the nucleoplasm is less intense but some fluorescent areas are observable (arrows). Bar: 2 mm.
360
L. M. Neri et al.
Nuclear PI-PLC c1 Increase after Hepatectomy
erating rat liver caused an increase of nuclear PIP2 recovery before the S-phase as well as an increase of phosphodiesteratic activity during DNA synthesis. The data about PIP2 synthesis and hydrolysis point out the need to modulate the quantity of this phospholipid according to the phases of liver regeneration. In particular, 3 h after hepatectomy, PIP2 hydrolysis is reduced as well as PIP2 recovery, indicating that the small quantity of substrate is accompanied by a modest phosphodiesterase activity. At 16 h post hepatectomy, data on PIP2 synthesis showed an increase of the phospholipid production, accompanied by a reduction of IP3 generation, thus resulting in a clear accumulation of the PIP2 molecule. PIP2 recovery instead showed a modest decrease at 22 h in comparison with 16 h. The higher PIP2 recovery was detected coincidentally with a decreased phosphodiesterase activity and can be related with the cell need to accumulate PIP2 a few hours before the S-phase. In fact, the observation that PIP2 hydrolysis is largely increased at 22 h, as detected also by IP3 formation, can be explained by cells’ need for products of PIP2 hydrolysis. This indicates a close relationship of nuclear inositol cycle with DNA synthesis and mitosis, because all cells reach the peak of proliferation 22–26 h after hepatectomy [41]. The finding about change in IP3 levels is in agreement with the observation that rat liver nuclei possess the nuclear IP3 receptor, which showed a decline in binding sites following partial hepatectomy [42] and which is a nuclear substrate of PKC [43]. The PIP2 hydrolysing activity of rat liver nuclei has already been closely associated with cell proliferation. It was observed that at 22 h post-hepatectomy PI-PLC activity was 2.5 fold higher than in the control when measured with PIP but was only 30% higher when measured on PIP2 [44]. During rat liver regeneration there is also a marked increase in nuclear, but not in whole cell, DAG [45] (Banfic 1993). Nuclear DAG level starts to increase near 16 h, is maximal at 20–25 and is probably over by 40 h. It should be noted that these authors did not observe a PIP2 breakdown comparable with the DAG formation, since only a slight decrease occurred after hepatectomy. We were able to observe significant variations both for PIP2 recovery, increased at 16 h, and for IP3 production increased at 22 h. The activation of a specific nuclear PI turnover during DNA synthesis was also demonstrated in HeLa cells [46] in which the levels of nuclear PIP and PIP2 decreased by 66% during the S phase of the cell cycle whereas the mass of these inositol lipids did not vary throughout the cell cycle. Differentiation process in Friend erythroleukemia cells is characterized by an accumulation of PIP2 and by a decrease of DAG second messenger in the nucleus [47], indicating that the nuclear PI turnover is also related to growth arrest. Similar phenomenon was observed during rat spermatogen-
361
esis, when PIP2, present at the nuclear level in all cell types, accumulated in the nuclei of late spermatidis and spermatozoa, phosphoinositide specific phospholipase C activity was also markedly decreased [48]. A possible consequence of nuclear PI turnover during DNA synthesis is that the accumulation of DAG could attract protein kinase C (PKC) into the nucleus. Direct evidences showed that during rat liver regeneration the large rise in nuclear DAG levels is temporally related to the PKC translocation to the nucleus [44] and an increased phosphorylation of nuclear PKC substrates, such as histone H1, after partial hepatectomy [49, 50]. Taken together, these data described the fine tuning mechanism deriving from the relevant role of nuclear inositide cycle, in which generation of IP3 after hepatectomy is a signal arising from the activation of a specific nuclear PIPLC isozyme. In this context the g1 isoform of this enzyme is the most important candidate to regulate PIP2 hydrolysis during rat liver regeneration. We are grateful to Aurelio Valmori and Maurizio Stroscio for skillful photographic assistance, and to Dr. Carlo Mischiati for densitometric analysis. This work was supported by Italian C. N. R. Progetto Finalizzato A.C.R.O. and grant 94.00412.CT12, AIRC, 40% and 60% grant to Universita’ di Ferrara.
References 1. Berridge M. J. (1993) Nature 361, 315–325. 2. Rhee S. G., Kim H., Shu P. G. and Choi W. C. (1991) Biochem. Soc. Trans. 19, 337–344. 3. Cockroft S. and Thomas G. M. H. (1992) Biochem. J. 288, 1–14. 4. Fisher S. K. (1995) Eur. J. Pharmacol. 288, 231–250. 5. Hasegawa Sasaki H., Lutz F. and Sasaki T. (1988) J. Biol. Chem. 263, 12970–12976. 6. Smrcka A. V., Hepler J. R., Brown K. O. and Sternweis P. C. (1991) Science 251, 804–807. 7. Taylor S. J., Chae H. Z., Rhee S. G. and Exton J. H. (1991) Nature 350, 516–518. 8. Hepler J. R., Kozasa T., Smrcka A. V., Simon M. I., Rhee S. G., Sternweis P. C. and Gilman A. G. (1993) J. Biol. Chem. 268, 14367–14373. 9. Lee C. H., Park D., Wu D., Rhee S. G., and Simon M. I. (1992) J. Biol. Chem. 268, 7949–7954. 10. Smith C. D., and Wells W. W. (1983) J. Biol. Chem. 258, 9368–9373. 11. Cocco L., Gilmour R. S., Ognibene A., Letcher A. J., Manzoli F. A., Irvine R. F. (1987) Biochem. J. 248, 765–770. 12. Divecha N., Banfic H. and Irvine R. F. (1991) EMBO J. 10, 3207–3214. 13. Divecha N., Banfic H. and Irvine R. F. (1993) Cell 74, 405–407. 14. Cocco L., Martelli A. M. and Gilmour R. S. (1994) Cell. Signall. 6, 481–485. 15. Martelli A. M., Gilmour R. S., Bertagnolo V., Neri L. M., Manzoli L. and Cocco L. (1992) Nature 358, 242–245. 16. Marmiroli S., Ognibene A., Bavelloni A., Cinti C., Cocco L. and Maraldi N. M. (1994) J. Biol. Chem. 269, 1–4.
FIGURE 3. Immunofluorescence staining of PI-PLCb1 (A–B), g1 (C–D) and d1 (E–F) before (A, C, E) and during (B, D, F) rat liver
regeneration. PI-PLCb1 localized mainly in the nucleus, but some faint labelling is observable also in the cytoplasm of both control and (A) and regenerating rat liver (B). PI-PLCg1 stains the cytoplasm and the nucleus where the green FITC labelling can be observed in yellow when it colocalizes with red PI in the nucleus (C). A marked increase of nuclear PI-PLCg1 is seen in D. Note some large areas and peripheral nuclear labelling (arrows). Bar: 2 mm.
362 17. Mazzoni M., Bertagnolo V., Neri L. M., Carini C., Marchisio M., Milani D., Manzoli F. A. and Capitani S. (1992) Biochem. Biophys. Res. Commun. 187, 114–122. 18. Martelli A. M., Billi A. M., Gilmour R. S., Neri L. M., Manzoli L., Ognibene A. and Cocco L. (1994) Cancer Res. 54, 2536–2540. 19. Capitani S., Bertagnolo V., Mazzoni M., Santi P., Previati M., Antonucci A., and Manzoli F. A. (1989) FEBS Lett. 254, 194–198. 20. Capitani S., Helms B., Mazzoni M., Previati M., Bertagnolo V., Writz K. W. A. and Manzoli F. A. (1990) Biochim. Biophys. Acta 1044, 193–200. 21. Divecha N., Rhee S. G., Letcher A. J. and Irvine R. F. (1993) Biochem. J. 289, 617–620. 22. Bertagnolo V., Mazzoni M., Ricci D., Carini C., Neri L. M., Previati M. and Capitani S. (1995) Cell. Signall. 7, 669–678. 23. Higgins G. M. and Anderson R. M. (1931) Arch. Pathol., 12, 186–202. 24. Laemmli U. K. (1970) Nature 277, 680–685. 25. Suh P. G., Ryh S. H., Choi W. C., Lee K. Y. and Rhee S. G. (1988) J. Biol. Chem. 263, 14497–14504. 26. Martelli A. M., Neri L. M., Zamai L., Bareggi R., Manzoli L. and Cocco L. (1994) Histochem. J. 26, 179–188. 27. Neri L. M., Riederer B. M., Marugg R. A., Capitani S. and Martelli A. M. (1995) Histochem. Cell Biol. 104, 29–36. 28. Irvine R. F., Anggard E. E., Letcher A. J. and Downs P., (1985) Biochem. J., 229, 505–511. 29. Shaik H. A. and Palmer F. B. St. (1976) J. Neurochem. 26, 597–603. 30. Neri L. M., Marinelli F., Martelli A. M., Manzoli L., Maraldi N. M. and Mazzottti G. (1990) Transactions of the Royal Microscopical Society (Elder H. Y., Ed.), vol. 1, 437–440, IOP Publishing Ltd., London. 31. Payrastre B., Nievers M., Boonstra J., Breton M., Verkelij A. J. and van Bergen en Henegouwen P. M. P. (1992) J. Biol. Chem. 267, 5078–5084. 32. Zini N., Martelli A. M., Cocco L., Manzoli F. A. and Maraldi N. M. (1993) Exp. Cell Res. 280, 257–269. 33. Zini N., Mazzoni M., Neri L. M., Bavelloni A., Marmiroli S., Capitani S. and Maraldi N. M. (1994) Eur. J. Cell Biol. 65, 206–213.
L. M. Neri et al. 34. Noh D. Y., Shin S. H. and Rhee S. G. (1995) Biochem. Biophys. Acta 1242, 99–114. 35. Liu N., Fukami K., Yu H. and Takenawa T. (1996) J. Biol. Chem. 271, 355–360. 36. Asano M., Tamiya-Koizumi K., Homma T., Takenawa T., Nimura Y., Kojima K. and Yoshida S. (1994) J. Biol. Chem. 269, 12360–12366. 37. Smith M. R., Liu Y., Kim H., Rhee S. G. and Kung H. (1990) Science 247, 1074–1077. 38. Divecha N., Letcher A. J., Banfic H. H., Rhee S. G. and Irvine R. F. (1995) Biochem. J. 312, 63–67. 39. Lee H. L., Lee H. J., Lee S. J., Min D. S., Baek S. H., Kim Y. S., Ryu S. H. and Suh P. G. (1995) FEBS Lett. 358, 105–108. 40. Martelli A. M., Gilmour R. S., Neri L. M., Manzoli L., Corps A. N. and Cocco L. (1991) FEBS Lett. 283, 243–246. 41. Vitale M., Rizzoli R., Mazzotti G., Zamai L., Galanzi A., Rizzi E., Manzoli L., Matteucci A. and Papa S. (1991) Cell Prolif. 24, 331–338. 42. Feng L. and Kraus-Friedmann N. (1994) Biochem. Biophys. Res. Commun. 205, 291–297. 43. Matter N., Ritz M. F., Freyermuth S., Rogue P., Malviya A. N. (1993) J. Biol. Chem. 268, 732–736. 44. Kuriki H., Tamiya-Koizumi K., Asano M., Yoshida S., Kojima K. and Nimura Y. (1992) J. Biochem. 111, 283–286. 45. Banfic H., Zizak M., Divecha N. and Irvine R. F. (1993) Biochem. J. 290, 633–636. 46. York J. D. and Majerus P. W. (1994) J. Biol. Chem. 269, 7847–7850. 47. Martelli A. M., Cataldi A., Manzoli L., Billi A. M., Rubbini S., Gilmour R. S. and Cocco L. (1995) Cell. Signall. 7, 53–56. 48. Caramelli E., Matteucci A., Zini N., Carini C., Guidotti L., Ricci D., Maraldi N. M. and Capitani S. (1996) Eur. J. Cell Biol. (in press). 49. Martelli A. M., Carini C., Marmiroli S., Mazzoni M., Barker P. J., Gilmour R. S. and Capitani S. (1991) Exp. Cell Res. 195, 255–262. 50. Mazzoni M., Carini C., Matteucci A., Martelli A. M., Bertagnolo V., Previati M., Rana R., Cataldi A. and Capitani S. (1992) Cell. Signall. 3, 313–319.
ISSN 0898-6568/97 $17.00 PII S0898-6568(96)00178-7
Changes of Nuclear PI-PLC g1 During Rat Liver Regeneration Luca M. Neri,†‡ Daniele Ricci,† Cinzia Carini,† Marco Marchisio,† Silvano Capitani† and Valeria Bertagnolo†* †Signal Transduction Unit/Laboratory of Cell Biology, Institute of Human Anatomy, University of Ferrara, Ferrara, Italy and ‡Institute of Normal and Pathological Cytomorphology C. N. R., c/o I. O. R., Bologna, Italy
ABSTRACT. We have previously demonstrated that rat liver nuclei contain PI-PLC b1 and g1 in the inner nuclear matrix and lamina associated with specific phosphodiesterase activity (Bertagnolo et al., 1995, Cell Signall. 7, 669–678). Since compensatory hepatic growth is an informative and well characterized model for natural cell proliferation, the presence of specific PI-PLC isoforms and their activity as well as PIP2 recovery were studied at various regenerating times, ranging from 3 to 22 h after partial hepatectomy. Three PI-PLC isoforms (b1, g1, d1) were examined in control and regenerating liver cells by using specific antibodies. By means of in situ immunocytochemistry and confocal microscopy, PI-PLC b1 was found mainly in the nucleoplasm and this pattern was not modified after hepatectomy. On the contrary, the nuclear g1 isoform showed a marked decrease at 3 and 16 h after hepatectomy, but a clear increase at 22 h covering with bright intensity the whole nucleus. The PI-PLC d1 isoform, which is exclusively cytoplasmic, was not altered during rat liver regeneration. By western blotting analysis on whole cell homogenates, none of the PI-PLC isozymes under study showed proliferation-linked modification. However, analyses of isolated nuclei identified changes in the nucleus associated PI-PLC g1 that paralleled the in situ observation whereas the b1 isoform was unmodified at all the times examined. Nuclear phosphodiesterase activity on PIP2 was lower at 3 and 16 h, in comparison with sham operated rats, increased at 6 h and reached the highest value after 22 h. Consistently, the recovery of PIP2, obtained in conditions that optimise PIP-kinase activity, showed a marked decrease at 3 h and an increase up to 16 h of liver regeneration, followed by a further decrease at 22 h. These data are consistent with a close relationship between cell proliferation and the nuclear inositide cycle, depending, in rat liver, predominantly on the modulation of the g1 isoform of PI-PLC. cell signal 9;5:353–362, 1997. 1997 Elsevier Science Inc. KEY WORDS. Nucleus, Inositide cycle, Phosphoinositidases, Rat liver, Liver regeneration, Confocal microscopy
INTRODUCTION The initial event of inositol lipid turnover for transmembrane signalling is the phosphodiesteratic cleavage of phosphatidylinositol 4,5-bisphosphate (PIP2). This reaction is catalysed by a phosphoinositide specific phospholipase C (PI-PLC) to yield inositol 3,4,5-trisphosphate (IP3) and diacylglycerol (DAG) [1]. The three major classes of PI-PLC identified (b, g, d) are tissue and cell type specific [2, 3], but all three subtypes require the presence of Ca11 for full activation and are capable of the hydrolysis of the three major inositol lipids with the selectivity for PIP2 over PIP and PI decreasing in the order of PI-PLC b, g and d [4]. *Address all correspondence to Valeria Bertagnolo, Institute of Human Anatomy, University of Ferrara, Ferrara, Italy. Abbreviations: PI-PLC—phosphoinositide-specific phospholipase C; PI—phosphatidylinositol; PIP—phosphatidylinositol 4-monophosphate; PIP2—phosphatidylinositol 4,5-bisphosphate; PKC—protein kinase C; DAG—diacylglycerol. Received 22 August 1996; and accepted 30 October 1996.
Distinct mechanisms of signal transduction are involved in the activation of PI-PLC following agonist occupancy of phosphoinositide linked receptors. Receptors with intrinsic tryosine kinase activity, such as those for PDGF (Platelet Derived Growth Factor) and EGF (Epidermal Growth Factor), activate PI-PLC g by phosphorylating its tyrosine residues [5], whereas bradykinin, vasopressin and other receptors activate PI-PLC b through an intervening guanine nucleotide binding protein (G protein) of the aq or bg class [6, 7, 8, 9]. The regulation of the d isoforms is still not clear. The signalling mechanism based on PI-PLC activity is not confined to the cytoplasm. Evidence for a nuclear inositol lipid cycle was first reported by Smith and Wells in nuclei purified from rat liver [10]. Subsequently it was demonstrated that in Friend erythroleukemia cell nuclei, PIP2 is formed by phosphorylation of PIP when cells undergo terminal differentiation [11]. In Swiss 3T3 cells, within minutes of insulin-like growth factor I (IGF-I) treatment, the nuclear mass of PIP and PIP2 halved, while that of diacyl-
354
glycerol (DAG) doubled by using picomole-sensitive mass assay [12]. Many other works pointed to a nuclear phosphoinositide cycle entirely separate from the known one in the plasma membrane (for review see [13, 14]. The decrease in nuclear PIP and PIP2 accompanied by an increase in DAG is coincident with the activation of phosphoinositidase C in the nucleus. Phosphoinositidase C activity in nuclei of Swiss 3T3 cells increases four fold after IGF-I treatment [15] and two fold in osterosarcoma SaOS-2 cells after Interleukin-I (IL-1) stimulation [16]. This increase of PLC activity was only dependent on the b1 isoform, which was confined to the nucleus [15, 16]. This observation was extended by other works that showed the nuclear localization and activity of PLCb1 in nuclei of rat pheochromocitoma PC12 and Friend erythroleukemia cells [17, 18]. Specific nuclear phosphoinositide cycle and PLC activity has been described also in rat liver. It was first demonstrated that the cell nucleus is a site of lipid phosphorylation, not involving enzymes and substrates located on the nuclear membrane [19]. The incorporation of PI into rat liver nuclei was greatly stimulated by phosphatidylinositol transfer protein (PI-TP) and nuclei were able to phosphorylate PI, with the production of PIP and PIP2 [20]. The first evidence of a PI-PLC isoform in rat liver nuclei was demonstrated by Divecha et al. [21], which showed that the only isoform specifically, but not exclusively, located in the cytoplasm was PI-PLCb1. Recently the subcellular distribution of PI-PLCb1, g1 and d1 has been investigated in the same model, indicating that the b1 isoform is predominantly nuclear, whereas the g1 isozyme is largely cytoplasmic and PLCd1 is restricted to the cytoplasm [22]. The assay of phosphodiesterase activity showed a discrete intracellular distribution and, by specific inhibition with neutralising anti PI-PLCb1 and g1 antibodies, this enzyme activity was confirmed at the nuclear level. Although some studies have been carried out, it has not yet been clearly identified which PI-PLC isoforms are involved in the nuclear inositol lipid cycle during rat liver regeneration nor their nuclear localisation at various times post hepatectomy. Therefore we have investigated the behaviour of these isozymes in the liver compensatory growth response after partial hepatectomy. The results indicate that the selective localization of PIPLC isozymes is accompanied, after partial hepatectomy, by changes of a specific isoform, PI-PLCg1. An immunocytochemical analysis in combination with confocal microscopy showed the increase of PI-PLCg1 nuclear content both on tissue sections and isolated nuclei. Immunochemical analysis demonstrated an identical increase in isolated nuclei of PI-PLCg1. These changes have been related to IP3 production due to phosphodiesterase activity and to PIP2 recovery in conditions that optimised PIP-kinase activity both before and during DNA synthesis period. MATERIALS AND METHODS Source of Materials Unless otherwise indicated, all reagents were obtained from the Sigma Chemical Co. (St. Louis, MO).
L. M. Neri et al.
Partial Hepatectomy Male adult Wistar rats (150–200 g body wt) were operated on according to Higgins and Anderson [23]. Approximately two thirds of the liver was surgically removed, and shamoperated rats were used as control for normal livers. The rats were sacrificed after 3, 6, 16, and 22 h, and livers were removed and used for preparation of total homogenate and isolation of nuclei. Preparation of Rat Liver Fractions Membrane-depleted nuclei were isolated from normal and regenerating rat livers as previously described [19], in the presence of 0.2% Triton X-100. All the purification buffers contained 10 mg/ml Aprotinin, 25 mg/ml Leupeptin, 50 mg/ ml STI, 1 mM PMSF, 0.5 mM DTT and 1 mM Na 3VO4, 15 mg/ml Calpain inhibitor I and 30 mg/ml Calpain inhibitor II (Calbiochem, La Jolla, CA). Total homogenates and cytoplasmic fractions were obtained as previously described [22]. Ultrastructural analysis and assays of cytoplasmic enzyme markers for assessment of nuclear purity were performed as previously reported [22]. Immunochemical Analysis Proteins from total homogenate and nuclei (50 mg), were separated on 7.5% polyacrylamide denaturing gels [24] and blotted to Trans-Blot nitrocellulose membrane (Bio Rad Laboratories, Richmond, CA). Monoclonal antibodies against the three isoforms b1, g1 and d1 of PI-PLC, originally developed by Suh et al., [25] employed as previously described [17] and a monoclonal antibody anti Phosphotyrosine (PTyr) were purchased from UBI (Lake Placid, NY). Peroxidase-conjugated anti-mouse IgG (Amersham International, Bucks U. K.) was used as secondary antibody, and the final detection was performed using the ECL system (Amersham), according to the manufacturer’s instructions. Densitometric analysis was done with an imaging densitometer BioRad PhosphorImager (Model GS 670, BioRad, U. K.) using the Molecular Analyst software. Each lane was analysed on the basis of specific signal i. e. bands with peaks of positivity and regions with low or negligible positivity. Regions were arbitrarily chosen for each sample to optimize the analysis of each immunoreactive band. Within each lane, data were averaged on the basis of 10 measurements, performed on the same band. For statistical evaluation the results were expressed as means 6 standard deviations of four different experiments. Standard deviation was in all curves less than 9%. As a further control we have compared different means from the same preparation, and we have calculated the p value. This was always . 0.05 and therefore not significant. Immunofluorescence and Confocal Microscopy Analysis Samples of rat liver, 4.5 mm thick, were quickly frozen in liquid nitrogen and sectioned in a Lauda 1720 cryostat (Leica, Germany). Cryo-sections were fixed in 4% paraformaldehyde/PBS followed by a mild permeabilisation with 0.2% Triton for 8 min. Per sample, 1 3 106 isolated nuclei
Nuclear PI-PLC c1 Increase after Hepatectomy
were dropped on glass microscope slides coated with polyL-lysine (Sigma, USA), allowed to stick for 30 min at room temperature and fixed in 4% paraformaldehyde/PBS. Tissue sections and isolated nuclei were then treated as follows: slides were incubated to block non specific binding with 2% BSA, 3% NGS in PBS (immunoreaction buffer) for 30 min at 378C and then reacted with primary antibodies anti PI-PLC isoforms diluted 1:50 in immunoreaction buffer for 3 h at 378C. After three washing steps, primary antibody binding was visualised by incubating the specimens with FITC-conjugated goat anti-mouse IgG (Sigma) diluted 1:100 for 1h at 378C. Slides were then incubated with PBS containing DAPI (Sigma) to assess the nuclear domains and embedded in glycerol containing 1,4-diazabicyclo [2.2.2] octane (DABCO) (Sigma) to retard fading, using additional coverslips as spacers to preserve the threedimensional structure of tissue sections or of nuclei. A variety of controls was performed to ensure the specificity of the antibody used: coverslips were treated with preimmune mouse or rabbit serum as well as with the secondary antibody alone. In all cases negligible levels of fluorescent signal were observed (not shown). All the experiments were done at least in triplicate. To define the nuclear region, the tissue sections were counterstained with 1 mg/ml Propidium Iodide (PI). The samples were analysed by a confocal laser scanning microscope Phoibos 1000 (Molecular Dynamics, USA) coupled with a Nikon Optiphot equipped with a 100 3 oil immersion lens (NA 5 1.4). The instrument was equipped with an argon laser 488/514 nm tuned at 10 mW and decreased by a neutral density filter to 30% of its power. On the illumination pathway a dichroic mirror (500 nm) was used, while on the detection pathway a 515 OG filter was employed. For double stained specimens a second dichroic mirror (550 nm) was placed in front of the two photomultipliers (PMT) in combination with a 530 6 15 nm or a 590 6 15 nm band pass filter for FITC or PI respectively. The PMT gains were set at 934 for FITC fluorescence and 786 for PI fluorescence and settings were rigorously maintained for all experiments. Image acquisition, recording, filtering and reconstruction were performed on z series of confocal data (stacks) by an Indy 4400 graphic workstation (Silicon Graphics, USA) as previously described [26, 27].
PIP2-Specific Phospholipase C Activity Total homogenate and nuclei (50 mg of protein) were incubated in a final volume of 100 ml for 15 min at 378C in the presence of 3 nmols of PIP2, 30,000 dpm [3H] PIP2, 0.06% tauro-deoxycholate, 0.6% NaCl, 0.1 M Mes, pH 6.2, 0.1 mM CaCl2. The reaction was stopped by addition of the lipid extraction mixture and the hydrosoluble products were counted as previously described [17]. For detection of IP3 the aqueous phases were loaded on a Partisil 10 SAX column, 25 cm length, from Whatman, then eluted with a linear gradient of ammonium formate (0–2 M, pH 3.7 with phosphoric acid) [28]. Elution peak of IP3 was identified by comparison with authentic standard. Radioactivity was determined by liquid scintillation
355
counting. Statistical analysis was performed using the two tailed Student’s t test for unpaired data. In Vitro PIP Phosphorylation Normal and regenerating rat liver nuclei (200 mg of protein) were preincubated for 10 min at 308C in the presence or absence of 0.05% Triton X-100, 100 mM PIP, then incubated for 15 min with 15 mM MgCl2, 10 mM b-mercaptoethanol, 1 mM Na3VO4, 100 mM ATP and 10 mCi [g-32P] ATP, in a final volume of 200 ml. The reaction was terminated by adding chloroform/methanol/HCl (200:100:0.75, v/v) and the phosphoinsitides were extracted according to Shaikh and Palmer [29]. After extraction, the organic phases were developed on TLC plates and the phosphoinositides were scraped off and counted by liquid scintillation. Statistical analysis was performed using the two tailed Student’s t test for unpaired data. RESULTS Immunochemical Analysis As demonstrated in our previous work concerning the distribution of PI-PLC in rat liver fractions [22], both b 1 and g1 isoforms were present in isolated nuclei. The two isozymes were investigated by Western Blot analysis in whole cell homogenates and nuclei obtained from normal and regenerating rat liver. In total homogenates no differences on the amount of both B1 and g1 isozymes were shown (Fig. 1A). In isolated nuclei the signal corresponding to the b1 isoform was similar at all the explored regeneration times, while the amount of the g1 isoenzyme changed. The latter isoform lowered after 3 h, raised to control levels at 6 h, lowered again 16 h from hepatectomy and showed the maximum increase after 22 h, compared with sham operated conditions (Fig. 1B). The amount of phosphorylated PI-PLCg1, as detected by anti-PTyr antibody, paralleled the decrease of the whole nuclear isozyme (Fig. 1C) at three and 16 h and the increase after 22 h from hepatectomy. As determined by densitometric scanning of fluorograms, we have analysed several experiments and additional timepoints. Table 1 shows that at 90 min., 2 and 3 h there is a relevant decrease of two folds of PI-PLCg1 at the nuclear level whereas both the cytoplasmic and the whole homogenates were unchanged. At 6 and 8 h time after partial liver resection no differences were appreciated in comparison with unoperated rats. At 14, 16 and 18 h after hepatectomy the level of intranuclear PI-PLCg1 were markedly reduced by two fold and the immunobands obtained at 22 and 26 hours were both increased by about 65%. The d1 isoform, present exclusively at cytoplasmic level, was also investigated but no differences associated with liver regeneration were shown (data not shown). Confocal Analysis on Isolated Nuclei To analyze in more details the distribution of the nuclear isoforms identified by Western Blot, the same time-course study was performed by immunocytochemical analysis on isolated
356
nuclei. Optical sections were chosen from confocal microscopy data sets about 3 mm from the bottom side of the nuclei attached to the glass surface, which corresponded to their equatorial plane.
FIGURE 1. Immunochemical analysis of PI-PLCb1 and g1 in total
homegenate (A) and isolated nuclei (B) at different hours from partial hepatectomy. In (C) P-Tyr labelling of the PI-PLCg1 bands. Molecular weights are indicated in KDa. See description in text.
L. M. Neri et al.
The staining of PI-PLCb1 was substantially unchanged at the various times after hepatectomy (Fig. 2A–E). The pattern was characterized by an intense labelling in all the nucleoplasm with very fine tiny rounded dots that allowed us to identify the nucleolar areas that appeared negative. Sometimes in the central part of the nucleus, especially at 6, 16 and 22 h (Fig. 2C–E), large areas of homogeneous labelling were observable. PI-PLCg1 immunofluorescence stained with a fine punctate labelling the whole nucleoplasm of control nuclei. Nucleoli could be observed as negative unstained areas. The labelling was present also at the nuclear periphery even if with a lower degree of intensity (Fig. 2F). A very consistent decrease of staining occurred after 3 h (Fig. 2G). In the nuclear interior few large fluorescent spots and some tiny dots were observable with an irregular distribution, showing many different nuclear regions almost negative. Some labelling was increased at the periphery, with a ring-like aspect that resembled the nuclear lamina and a fluorescent mass was located in contact with the peripheral staining. After 6 h the labelling resembled the levels of control nuclei (Fig. 2H). Fluorescence was regularly distributed in the nucleus and some addensation was seen at the periphery. A second very evident decrease was observable after 16 h from the hepatectomy (Fig. 2I): the immunofluorescence was organised in a few small dots present mainly in the central part of the nucleoplasm, except for the nucleolar area. A very weak labelling was observable at the nuclear periphery, except for a few irregular fluorescent areas, some of which were located at the nuclear edge. The 22 h sample was characterized by a strong increase in PI-PLCg1 immunoreactivity that acquired a new aspect of labelling (Fig. 2L): a large central area was characterized by a fairly homogeneous pattern of very brilliant fluorescence that that did not allow the recognition of single spots or dots. In fact, the fine granular pattern of control nuclei was replaced by large areas of labelling emerging from a brilliant fibrogranular meshwork. The nucleoli and the periphery of the nucleoplasm were labelled with lower intensity, but still higher than the control nuclei. In some cases at the nuclear periphery some very large fluorescent spots more brilliant than the surrounding labelling were seen. To verify the described PI-PLC distribution, confocal analysis on tissue sections was performed on two sample times (0 and 22 h after partial hepatectomy) as a further control. Control liver sections when probed with the anti PI-PLb1 antibodies displayed an intense immunofluorescence characterized by a fine puncatate labelling, that stained the whole nucleoplasm with negative nucleoli and also partially stained the cytoplasm (Fig. 3A). The partial hepatectomy did not modify intensity and distribution pattern of the labelling after 22 h (Fig. 3B). Cytoplasm and nucleus were both labelled by anti-PI-PLCg1 antibodies in liver cells and DNA was counterstained with PI. Fine little dots were distributed in the nucleoplasm and larger fluorescent areas were often observable both in the nuclear interior and at its periphery (Fig. 3C). The cytoplasmatic labelling was un-
Nuclear PI-PLC c1 Increase after Hepatectomy
357
TABLE 1. Densitometric analysis of immunoreactive PI-PLC g1 and P-Tyr in total homogenates (HT), cytoplasmic fration
(Cyt) and isolated nuclei (Nu) at different regenerating timepoints: data are expressed as percentage variations of control conditions, detected by densitometric scanning, as described in the Results section Hours Anti Anti Anti Anti
PI-PLC g1 HT PI-PLC g1 Cyt PI-PLC g1 Nu P-Tyr Nu
0
909
2
3
6
8
14
16
18
22
26
100 100 100 100
98 103 50* 37*
101 105 49* 40*
105 101 47* 35*
106 103 95 95
100 104 97 97
98 99 51* 50*
96 99 46* 43*
99 100 47* 40*
102 104 161* 150*
100 103 155* 156*
Data are the mean from four different experiments. For each of the antibodies, the staining patterns of total homogenates, cytoplasmic fraction and isolated nuclei were compared with the unoperated liver, that served as a control. * Highly significant differences (p , 0.001). All of the other differences were found to be not significant with p , 0.05.
changed after hepatectomy, whereas the nuclear staining showed a marked increase after 22 h of liver regeneration corresponding to that described in insolated nuclei: the whole nucleoplasm was labelled displaying large very brilliant areas including also its periphery (Fig. 3D). PI-PLCd1 showed a cytoplasmatic localization of a discrete and punctate immunofluorescence (Fig. 3E), that did not show appreciable changes 22 h after partial hepatectomy (Fig. 3F). Phosphodiesterase Activity
in Table 2, the recovery of PIP2 in conditions optimising PIP-kinase activity after 3 h was reduced to 25% of control value. Starting from 6 h of liver regeneration, the recovery of PIP2 increased progressively until 16 h post hepatectomy. The analysis during the S phase (22 h) showed a moderate increase of the activity in respect to sham operated samples, but a marked decrease (230%) when compared to the 16 h sample.
DISCUSSION
The percentage of tritiated PIP2 cleaved by the isozymes present in liver nuclei from sham operated and hepatectomized rats is shown in Table 2. In the presence of PIP2 as substrate, a significant decrease of hydrolysis was revealed at 3 h from hepatectomy and returned similar to control values at 6 h. Starting from the 16th hour of liver regeneration, PIP2 decreased 40% after 16 h and reached the highest value, 50% higher than in control nuclei, at 22 h (Table 2). To assess the fraction of hydrosoluble compounds generated by PIP2 hydrolysis, some aqueous phases were analysed by HPLC. Consistent with the whole hydrosoluble counts, the recovery of IP3 was decreased after 16 h from hepatectomy and at 22 h was significantly increased of 80% than sham operated nuclei (Table 3). Phosphoinositide Kinase Activity PIP-kinase activity was monitored by assaying the phosphorylation of exogenous PIP at the different regeneration times (0, 3, 6, 16, 22 h from partial hepatectomy). As shown
Regenerating rat liver is a well characterized and informative model of cell proliferation in vivo. By means of confocal microscopy we have shown for the first time in situ modifications of PI-PLCg1 isoforms in liver tissue sections and isolated nuclei induced by partial hepatectomy. The in situ analysis by confocal microscopy on membrane depleted isolated nuclei was chosen to evaluate the real presence of PIPLCb1 and PI-PLCg1, within the nucleoplasm without interfering with the contribution of cytoplasmic or membrane located enzyme. The changes of PI-PLCg1 isoform were found in isolated nuclei and in tissue sections only at the nuclear level, whereas the cytoplasm appeared unchanged. Of the other isoforms analysed in this study, PI-PLCd1, which resulted to be exclusively cytoplasmic, and PI-PLCb1, mostly located in the inner nucleoplasm, were unaffected during cell proliferation. Main features obtained from in situ analysis of PI-PLCg1 are the g1 immunostaining resembling a fibrogranular meshwork with both lamina- and nucleoskeletal-like structures [30, 27],
TABLE 2. Levels of specific PIP2 phosphodiesterase activity in conditions optimising PI-PLC activity (PIP2 hydrolysis)
and of PIP kinase activity (PIP2 recovery) in rat liver nuclei at different regeneration time H PIP2 hydrolysis (nmol/mg proteins) Sham operated Hepatectomized PIP2 recovery (pmol/mg proteins) Sham operated Hepatectomized
0
3
6
16
22
16.0 6 1.2 —
15.5 6 1.4 9.6 6 0.94*
16.5 6 1.1 19.7 6 1.5
16.2 6 1.4 10.5 6 0.83*
15.4 6 1.3 23.5 6 1.9*
1.8 6 0.22 3.7 6 0.32*
2.3 6 0.21 2.7 6 0.18
2.3 6 0.17 —
1.9 6 0.16 0.6 6 0.03*
2.2 6 0.19 2.9 6 0.23
The values 6 SD are the mean from three separate determinations on lipid fractions from isolated rat liver nuclei. * Significant differences (p , 0.005 Student’s t test) from control.
358
L. M. Neri et al.
FIGURE 2. Optical sections through isolated nuclei placed on glass slides and immunostained for PI-PLCb1 (A–E) or PI-PLCg1 (F–L)
at various times during rat liver regeneration: 3 h (B, G), 6 h (C, H), 16 h (D, I) and 22 h (E, L). PI-PLCb1 (A) control nucleus. A fine granular labelling can be easily appreciated as well as the nucleolus. The nuclear periphery shows a weaker staining than the nuclear interior. (B) nucleus 3 h after hepatectomy. PI-PLCB1 staining is similar to A. (C) nucleus 6 h after hepatectomy. In the central
Nuclear PI-PLC c1 Increase after Hepatectomy
359
TABLE 3. Recovery of IP3 separated by HPLC after PIP2 specific hydrolysis
in rat liver nuclei at different regeneration time H IP3 recovery (nmol/mg proteins) Sham operated Hepatectomized
0
16
22
11.5 6 1.2
12.1 6 0.9 7.1 6 0.7*
11.0 6 1.1 19.4 6 1.9*
The values 6 SD refer to the mean from three separate determinations on lipid fractions from isolated rat liver nuclei. * Significant differences (p , 0.005 Student’s t test) from control.
its decrease at 16 h, before the S phase, and its large increase in the nucleoplasm at 22 h, that strongly associates this variation with cell proliferation induced by partial hepatectomy. The PI-PLC isoforms’ association with nucleoskeletal structures has been detected at the same sites in different cell lines [19, 31, 32, 33]. Interestingly, all these reports support the notion that all inositide-metabolising enzymes are in the internal nuclear matrix, except for PI-kinase, which is in the peripheral matrix [31]. In rat liver nuclei the detection of the nuclear isoforms PI-PLCb1 and g1 was not affected by the removal of the nuclear membrane and the two enzymes showed a tight association with nuclear matrix and lamina [22]. Western blot analysis strongly supported the in situ evidence that only the nuclear PI-PLCg1 isoform changed, whereas the whole cells and the cytoplasmic fraction did not show variations. After 3 and 16 h a decreased nuclear presence is consistent with a reduced synthesis of the entire cellular enzyme, whereas after 22 h of liver regeneration this PI-PLC isozyme may be newly synthesized and translocated from cytosol to nucleus in response to mitogenic stimuli. From the analysis by anti P-Tyr antibody it emerged that there is a constitutively phosphorylated form of the enzyme inside the nucleus and that it is this phosphorylated enzyme that changes during the different regeneration timepoints. Interestingly, these oscillations are sustained for a certain period of time, which appears to characterize the proliferation event as forming different phases that requires the presence, or not, of phosphorylated PI-PLCg1 within the nucleus. An increase of phosphorylation, that occurs only for PI-PLCg1 but not for PI-PLCb1 or PI-PLCd1 is a well known system of PI-PLCg1 activation by growth factor receptor protein-tyrosine kinase and is correlated with the stimulation of PIP2 hydrolysis [34]. However, very little is known about the molecular events that occur after the tyrosine phosphorylation of PI-PLCg1.
A recent paper described a novel PI-PLC isozyme, the d4 isoform, that is expressed more remarkably in regenerating liver than in normal resting liver as well as in Swiss 3T3 fibroblasts nuclei during the S phase [35]. In agreement with our data, the same authors observed that PI-PLCb1 content did not vary. However, they did not detect PI-PLCg1 changes. The difference, in the latter isozyme may be due to the different antibodies used, cell type and model (in vitro vs. in vivo). A unique 88 kDa form of PI-PLC has been purified only in regenerating liver nuclei, but the authors failed to identify which PLC isozyme corresponded to this detected species [36]. A number of reports have addressed the issue of an isotype specific involvement of the PI-PLC family in the cell response to different stimuli. Resting NIH 3T3 fibroblasts were induced to synthesise DNA after PI-PLCb or g injection, and proliferation was inhibited when anti PI-PLC g antibodies were injected, showing that PI-PLC isoforms are related to growth arrest and differentiation [37]. Mouse erythroleukemia cells, when treated with DMSO, show a decrease in nuclear PI-PLC activity and in the levels of nuclear DAG associated with a decrease of the b1 isoform, as demonstrated by the use of isoform specific antibodies [18, 38]. Differentiation-linked decrease of PI-PLCs was also described on U937 human histiocytic leukemia cells in which a marked and selective decrease in enzyme activity and protein level of PI-PLCg1 was observed, while PI-PLCg2 was not affected [39]. The nuclear inositide cycle is specifically affected by stimuli that promote proliferation. Agonist stimulation of the IGF-I receptor tyrosine kinase activity does not lead to PIP2 breakdown or DAG production at the plasma membrane, but rather in the nucleus [12, 15], suggesting a possible link between nuclear signalling, PI-PLC isoforms and the onset of DNA synthesis [40]. In our system, the cellular proliferation induced by regen-
part of the nucleoplasm the spots are fused together to form large and brilliant fluorescent areas. The nucleoli are still observable. (D and E) nuclei at 16 or 22 h of liver regeneration, respectively. The fluorescent pattern is very similar to C. PI-PLCg1 (F) control nucleus. The whole nucleoplasm is labelled by a fine granular pattern more intense in the center than at the nuclear periphery. Nucleoli appear negative. (G) nucleus 3 h after hepatectomy. A scarce fluorescence can be observed except for few large and brilliant irregular areas (arrow). Note the presence of ring-like fluorescence at the periphery. (H) nucleus 6 h after hepatectomy. The original fluorescence intensity is restored with a pattern very similar to f, with a slight accumulation at the periphery. (I) nucleus at 16 h of liver regeneration. A strong decrease of the labelling is displayed. Some irregular faint spots are visible together with two larger areas (arrows), one of which is in close association with the nuclear periphery. (L) nucleus at 22 h of liver regeneration. An intense increase of the immunofluorescence is observable in the nuclear interior. The periphery of the nucleoplasm is less intense but some fluorescent areas are observable (arrows). Bar: 2 mm.
360
L. M. Neri et al.
Nuclear PI-PLC c1 Increase after Hepatectomy
erating rat liver caused an increase of nuclear PIP2 recovery before the S-phase as well as an increase of phosphodiesteratic activity during DNA synthesis. The data about PIP2 synthesis and hydrolysis point out the need to modulate the quantity of this phospholipid according to the phases of liver regeneration. In particular, 3 h after hepatectomy, PIP2 hydrolysis is reduced as well as PIP2 recovery, indicating that the small quantity of substrate is accompanied by a modest phosphodiesterase activity. At 16 h post hepatectomy, data on PIP2 synthesis showed an increase of the phospholipid production, accompanied by a reduction of IP3 generation, thus resulting in a clear accumulation of the PIP2 molecule. PIP2 recovery instead showed a modest decrease at 22 h in comparison with 16 h. The higher PIP2 recovery was detected coincidentally with a decreased phosphodiesterase activity and can be related with the cell need to accumulate PIP2 a few hours before the S-phase. In fact, the observation that PIP2 hydrolysis is largely increased at 22 h, as detected also by IP3 formation, can be explained by cells’ need for products of PIP2 hydrolysis. This indicates a close relationship of nuclear inositol cycle with DNA synthesis and mitosis, because all cells reach the peak of proliferation 22–26 h after hepatectomy [41]. The finding about change in IP3 levels is in agreement with the observation that rat liver nuclei possess the nuclear IP3 receptor, which showed a decline in binding sites following partial hepatectomy [42] and which is a nuclear substrate of PKC [43]. The PIP2 hydrolysing activity of rat liver nuclei has already been closely associated with cell proliferation. It was observed that at 22 h post-hepatectomy PI-PLC activity was 2.5 fold higher than in the control when measured with PIP but was only 30% higher when measured on PIP2 [44]. During rat liver regeneration there is also a marked increase in nuclear, but not in whole cell, DAG [45] (Banfic 1993). Nuclear DAG level starts to increase near 16 h, is maximal at 20–25 and is probably over by 40 h. It should be noted that these authors did not observe a PIP2 breakdown comparable with the DAG formation, since only a slight decrease occurred after hepatectomy. We were able to observe significant variations both for PIP2 recovery, increased at 16 h, and for IP3 production increased at 22 h. The activation of a specific nuclear PI turnover during DNA synthesis was also demonstrated in HeLa cells [46] in which the levels of nuclear PIP and PIP2 decreased by 66% during the S phase of the cell cycle whereas the mass of these inositol lipids did not vary throughout the cell cycle. Differentiation process in Friend erythroleukemia cells is characterized by an accumulation of PIP2 and by a decrease of DAG second messenger in the nucleus [47], indicating that the nuclear PI turnover is also related to growth arrest. Similar phenomenon was observed during rat spermatogen-
361
esis, when PIP2, present at the nuclear level in all cell types, accumulated in the nuclei of late spermatidis and spermatozoa, phosphoinositide specific phospholipase C activity was also markedly decreased [48]. A possible consequence of nuclear PI turnover during DNA synthesis is that the accumulation of DAG could attract protein kinase C (PKC) into the nucleus. Direct evidences showed that during rat liver regeneration the large rise in nuclear DAG levels is temporally related to the PKC translocation to the nucleus [44] and an increased phosphorylation of nuclear PKC substrates, such as histone H1, after partial hepatectomy [49, 50]. Taken together, these data described the fine tuning mechanism deriving from the relevant role of nuclear inositide cycle, in which generation of IP3 after hepatectomy is a signal arising from the activation of a specific nuclear PIPLC isozyme. In this context the g1 isoform of this enzyme is the most important candidate to regulate PIP2 hydrolysis during rat liver regeneration. We are grateful to Aurelio Valmori and Maurizio Stroscio for skillful photographic assistance, and to Dr. Carlo Mischiati for densitometric analysis. This work was supported by Italian C. N. R. Progetto Finalizzato A.C.R.O. and grant 94.00412.CT12, AIRC, 40% and 60% grant to Universita’ di Ferrara.
References 1. Berridge M. J. (1993) Nature 361, 315–325. 2. Rhee S. G., Kim H., Shu P. G. and Choi W. C. (1991) Biochem. Soc. Trans. 19, 337–344. 3. Cockroft S. and Thomas G. M. H. (1992) Biochem. J. 288, 1–14. 4. Fisher S. K. (1995) Eur. J. Pharmacol. 288, 231–250. 5. Hasegawa Sasaki H., Lutz F. and Sasaki T. (1988) J. Biol. Chem. 263, 12970–12976. 6. Smrcka A. V., Hepler J. R., Brown K. O. and Sternweis P. C. (1991) Science 251, 804–807. 7. Taylor S. J., Chae H. Z., Rhee S. G. and Exton J. H. (1991) Nature 350, 516–518. 8. Hepler J. R., Kozasa T., Smrcka A. V., Simon M. I., Rhee S. G., Sternweis P. C. and Gilman A. G. (1993) J. Biol. Chem. 268, 14367–14373. 9. Lee C. H., Park D., Wu D., Rhee S. G., and Simon M. I. (1992) J. Biol. Chem. 268, 7949–7954. 10. Smith C. D., and Wells W. W. (1983) J. Biol. Chem. 258, 9368–9373. 11. Cocco L., Gilmour R. S., Ognibene A., Letcher A. J., Manzoli F. A., Irvine R. F. (1987) Biochem. J. 248, 765–770. 12. Divecha N., Banfic H. and Irvine R. F. (1991) EMBO J. 10, 3207–3214. 13. Divecha N., Banfic H. and Irvine R. F. (1993) Cell 74, 405–407. 14. Cocco L., Martelli A. M. and Gilmour R. S. (1994) Cell. Signall. 6, 481–485. 15. Martelli A. M., Gilmour R. S., Bertagnolo V., Neri L. M., Manzoli L. and Cocco L. (1992) Nature 358, 242–245. 16. Marmiroli S., Ognibene A., Bavelloni A., Cinti C., Cocco L. and Maraldi N. M. (1994) J. Biol. Chem. 269, 1–4.
FIGURE 3. Immunofluorescence staining of PI-PLCb1 (A–B), g1 (C–D) and d1 (E–F) before (A, C, E) and during (B, D, F) rat liver
regeneration. PI-PLCb1 localized mainly in the nucleus, but some faint labelling is observable also in the cytoplasm of both control and (A) and regenerating rat liver (B). PI-PLCg1 stains the cytoplasm and the nucleus where the green FITC labelling can be observed in yellow when it colocalizes with red PI in the nucleus (C). A marked increase of nuclear PI-PLCg1 is seen in D. Note some large areas and peripheral nuclear labelling (arrows). Bar: 2 mm.
362 17. Mazzoni M., Bertagnolo V., Neri L. M., Carini C., Marchisio M., Milani D., Manzoli F. A. and Capitani S. (1992) Biochem. Biophys. Res. Commun. 187, 114–122. 18. Martelli A. M., Billi A. M., Gilmour R. S., Neri L. M., Manzoli L., Ognibene A. and Cocco L. (1994) Cancer Res. 54, 2536–2540. 19. Capitani S., Bertagnolo V., Mazzoni M., Santi P., Previati M., Antonucci A., and Manzoli F. A. (1989) FEBS Lett. 254, 194–198. 20. Capitani S., Helms B., Mazzoni M., Previati M., Bertagnolo V., Writz K. W. A. and Manzoli F. A. (1990) Biochim. Biophys. Acta 1044, 193–200. 21. Divecha N., Rhee S. G., Letcher A. J. and Irvine R. F. (1993) Biochem. J. 289, 617–620. 22. Bertagnolo V., Mazzoni M., Ricci D., Carini C., Neri L. M., Previati M. and Capitani S. (1995) Cell. Signall. 7, 669–678. 23. Higgins G. M. and Anderson R. M. (1931) Arch. Pathol., 12, 186–202. 24. Laemmli U. K. (1970) Nature 277, 680–685. 25. Suh P. G., Ryh S. H., Choi W. C., Lee K. Y. and Rhee S. G. (1988) J. Biol. Chem. 263, 14497–14504. 26. Martelli A. M., Neri L. M., Zamai L., Bareggi R., Manzoli L. and Cocco L. (1994) Histochem. J. 26, 179–188. 27. Neri L. M., Riederer B. M., Marugg R. A., Capitani S. and Martelli A. M. (1995) Histochem. Cell Biol. 104, 29–36. 28. Irvine R. F., Anggard E. E., Letcher A. J. and Downs P., (1985) Biochem. J., 229, 505–511. 29. Shaik H. A. and Palmer F. B. St. (1976) J. Neurochem. 26, 597–603. 30. Neri L. M., Marinelli F., Martelli A. M., Manzoli L., Maraldi N. M. and Mazzottti G. (1990) Transactions of the Royal Microscopical Society (Elder H. Y., Ed.), vol. 1, 437–440, IOP Publishing Ltd., London. 31. Payrastre B., Nievers M., Boonstra J., Breton M., Verkelij A. J. and van Bergen en Henegouwen P. M. P. (1992) J. Biol. Chem. 267, 5078–5084. 32. Zini N., Martelli A. M., Cocco L., Manzoli F. A. and Maraldi N. M. (1993) Exp. Cell Res. 280, 257–269. 33. Zini N., Mazzoni M., Neri L. M., Bavelloni A., Marmiroli S., Capitani S. and Maraldi N. M. (1994) Eur. J. Cell Biol. 65, 206–213.
L. M. Neri et al. 34. Noh D. Y., Shin S. H. and Rhee S. G. (1995) Biochem. Biophys. Acta 1242, 99–114. 35. Liu N., Fukami K., Yu H. and Takenawa T. (1996) J. Biol. Chem. 271, 355–360. 36. Asano M., Tamiya-Koizumi K., Homma T., Takenawa T., Nimura Y., Kojima K. and Yoshida S. (1994) J. Biol. Chem. 269, 12360–12366. 37. Smith M. R., Liu Y., Kim H., Rhee S. G. and Kung H. (1990) Science 247, 1074–1077. 38. Divecha N., Letcher A. J., Banfic H. H., Rhee S. G. and Irvine R. F. (1995) Biochem. J. 312, 63–67. 39. Lee H. L., Lee H. J., Lee S. J., Min D. S., Baek S. H., Kim Y. S., Ryu S. H. and Suh P. G. (1995) FEBS Lett. 358, 105–108. 40. Martelli A. M., Gilmour R. S., Neri L. M., Manzoli L., Corps A. N. and Cocco L. (1991) FEBS Lett. 283, 243–246. 41. Vitale M., Rizzoli R., Mazzotti G., Zamai L., Galanzi A., Rizzi E., Manzoli L., Matteucci A. and Papa S. (1991) Cell Prolif. 24, 331–338. 42. Feng L. and Kraus-Friedmann N. (1994) Biochem. Biophys. Res. Commun. 205, 291–297. 43. Matter N., Ritz M. F., Freyermuth S., Rogue P., Malviya A. N. (1993) J. Biol. Chem. 268, 732–736. 44. Kuriki H., Tamiya-Koizumi K., Asano M., Yoshida S., Kojima K. and Nimura Y. (1992) J. Biochem. 111, 283–286. 45. Banfic H., Zizak M., Divecha N. and Irvine R. F. (1993) Biochem. J. 290, 633–636. 46. York J. D. and Majerus P. W. (1994) J. Biol. Chem. 269, 7847–7850. 47. Martelli A. M., Cataldi A., Manzoli L., Billi A. M., Rubbini S., Gilmour R. S. and Cocco L. (1995) Cell. Signall. 7, 53–56. 48. Caramelli E., Matteucci A., Zini N., Carini C., Guidotti L., Ricci D., Maraldi N. M. and Capitani S. (1996) Eur. J. Cell Biol. (in press). 49. Martelli A. M., Carini C., Marmiroli S., Mazzoni M., Barker P. J., Gilmour R. S. and Capitani S. (1991) Exp. Cell Res. 195, 255–262. 50. Mazzoni M., Carini C., Matteucci A., Martelli A. M., Bertagnolo V., Previati M., Rana R., Cataldi A. and Capitani S. (1992) Cell. Signall. 3, 313–319.