Enhanced Nuclear Diacylglycerol Kinase Activity in Response to a Mitogenic Stimulation of Quiescent Swiss 3T3 Cells with Insulin-like Growth Factor I¹ Free

Compelling evidence from several independent laboratories supports the notion that a polyphosphoinositide cycle distinct from the“classic” cycle located at the plasma membrane and under a separate control exists in the nucleus (1, 2, 3, 4). The presence of such a nuclear cycle has been demonstrated in several cell lines and tissues (5, 6, 7, 8, 9, 10, 11). It has been shown that the nuclear polyphosphoinositide cycle is involved in the control of both proliferative and differentiative events (12, 13, 14, 15). A key enzyme of this cycle is PI-PLC³β1, which, upon stimulation with agonists such as interleukin 1 and IGF-I (9, 12), hydrolyzes phosphatidylinositol 4,5-bisphosphate, yielding the two second messengers, DAG and inositol 1,4,5-trisphosphate. In mouse Swiss 3T3 cells, PI-PLC β1 is preferentially located in the nucleus and is activated in response to IGF-I stimulation (12). In these cells, overexpression of antisense mRNA for PI-PLC β1 blocks IGF-I-induced mitogenesis (14), whereas its inducible overexpression in the sense orientation enhances the mitogenic effect of IGF-I (16). We have recently demonstrated that in Swiss 3T3 cells stimulated with IGF-I, the rise in nuclear DAG mass is essential to attract the αisoform of PKC to the nucleus (17). The question then arises as to how to switch off the signal constituted of nuclear DAG. It is currently thought that the enzyme DAG kinase, which phosphorylates DAG to PA, plays an important role in terminating the PKC-mediated signals (18). Previous evidence has demonstrated the presence of DAG kinase in isolated nuclei and subnuclear fractions (19, 20, 21). In this context, Topham et al. (22) have shown that the ζ isoform of the enzyme plays a key role in the control of nuclear DAG mass.

It is worth emphasizing here that a growing body of evidence suggests that the IGF-I-elicited signal transduction pathways play an important role in several forms of human cancer (e.g., Refs. 23 and 24).

With the above rationale in mind, we sought to determine whether or not there is an increase in DAG kinase activity in isolated nuclei prepared from IGF-I-stimulated Swiss 3T3 cells. Here, we demonstrate the existence of an inverse relationship between nuclear DAG levels and DAG kinase activity. Evidence is also provided that enhanced nuclear DAG kinase activity is likely to be important for terminating the activation of PKC-α within the nuclear compartment. Moreover,treatment of Swiss 3T3 fibroblasts with inhibitors of DAG kinase led to an increase in cells that entered the S phase of the cell cycle in response to mitogenic stimulation with IGF-I.

DMEM, FCS, 1,2-dioleyl-3-palmitoyl-glycerol,3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid, PS,1,2-dioleoylglycerol, l-α-PA (dioleoyl), histone H1,leupeptin, aprotinin, PMSF, NGS, rabbit polyclonal antibody to PKC-α,BrdUrd, peroxidase-conjugated antimouse and antirabbit IgG,Cy3-conjugated antimouse and antirabbit IgG, and BSA were from Sigma Chemical Co. (St. Louis, MO). DAG kinase inhibitor I (R59022) and II(R59949) and the monoclonal antibody to lamin B1 were from Calbiochem(La Jolla, CA). NP40, IGF-I, RNase-free DNase I, RNase A, and the Lumi-Light^Plus enhanced chemiluminescence detection kit were from Roche Molecular Biochemicals (Milan, Italy). Monoclonal antibody to BrdUrd was from Becton Dickinson (Milan, Italy).[γ-³²P]ATP and carbamylated glyceraldehyde-3-phosphate dehydrogenase standards were from Amersham/Pharmacia Biotech (Uppsala, Sweden). Recombinant rabbit PKC-α was from Upstate Biotechnology Inc. (Lake Placid, NY). Monoclonal antibody to protein C23/nucleolin was a kind gift from Dr. R. L. Ochs (Precision Therapeutics, Pittsburgh, PA). The Protein Assay kit (detergent compatible) was from Bio-Rad (Hercules, CA). Silica Gel G-60 thin layer chromatography plates and p81 paper were from Whatman (Maidstone, United Kingdom).

Swiss 3T3 mouse fibroblasts were cultured in DMEM containing 10% FCS. Before stimulation, cells were subcultured at a density of 10⁴ cells/cm² and incubated until they became confluent (6 days). They were then cultured for an additional 24 h in serum-free medium containing 0.5% BSA. Quiescent cultures were washed twice with serum-free medium containing 0.2% BSA and then incubated in the same medium for the indicated times in the presence of 50 ng/ml IGF-I. For the studies in the presence of DAG kinase inhibitors, the chemicals (stored as 5 mm stock solutions in DMSO) were added to the tissue culture medium 10 min before IGF-I. Addition of the same amount of solvent to the cultures had no effect (data not shown).

This was accomplished as reported previously (12). Briefly, cells (5 × 10⁶) were suspended in 500 μl of 10 mm Tris-Cl (pH 7.8), 1% NP40,10 mm β-mercaptoethanol, 0.5 mm PMSF, and 1μg/ml leupeptin and aprotinin for 2 min at 0°C. Double-distilled H₂0 (500 μl) was then added, and the cells were allowed to swell for 2 min. Cells were sheared by 10 passages through a 22-gauge needle. Nuclei were recovered by centrifugation at 400 × g for 6 min and washed once in 10 mm Tris-Cl (pH 7.4), 2 mmMgCl₂, and protease inhibitors as described above.

Intact nuclear matrices were isolated according to Payrastre et al. (19). Briefly, membrane-depleted nuclei were resuspended and incubated for 20 min at 37°C in CSK buffer [10 mm PIPES (pH 6.8), 300 mmsucrose, 50 mm NaCl, 3 mmMgCl₂, 0.5 mmCaCl₂, 1.2 mm PMSF, 10μg/ml soybean trypsin inhibitor, and 1 μg/ml of leupeptin and aprotinin]. They were then digested with 100 units/ml RNase-free DNase I and 20 units/ml RNase A for 60 min at 25°C. Subsequently, the chromatin-associated proteins were released by adding 2 m(NH₄)₂SO₄dropwise to a final concentration of 0.25 m(NH₄)₂SO₄. After 15 min of incubation on ice, the nuclear matrices were pelleted at 2000 × g for 10 min on a cushion containing 43% glycerol and 2 m sucrose in PBS.

For the isolation of the internal matrix, nuclei were first stabilized by incubation for 1 h with 0.5 mm sodium tetrathionate in CSK buffer on ice. Nuclei were washed twice with CSK buffer,digested with nucleases, and extracted with 0.25 m(NH₄)₂SO₄as described above. The nuclear matrices were then resuspended in CSK buffer and incubated for 20 min at 37°C in the presence of 0.25 m(NH₄)₂SO₄and 40 mm DTT. The solubilized internal matrix was cleared from the peripheral matrix by pelleting the peripheral matrix at 10,000 × g for 10 min. For isolation of the peripheral matrix, nuclear matrices were incubated for 20 min at 37°C in CSK buffer supplemented with 0.25 m(NH₄)₂SO₄and 40 mm DTT. The peripheral matrix was pelleted by centrifugation for 10 min at 10,000 × gon a cushion containing 43% glycerol and 2 msucrose in PBS (19).

This was performed according to the manufacturer’s instructions using the Bio-Rad Protein Assay (detergent compatible).

The assay was performed according to the method of Divecha et al. (5), using DAG kinase enzyme purified from rat brain. DAG was extracted from nuclei as reported previously (5), dissolved in 20 μl of 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid (9.2 mg/ml), and sonicated at room temperature for 15 s. After the addition of 80 μl of reaction buffer [50 mmTris acetate (pH 7.4), 80 mm KCl, 10 mm magnesium acetate, and 2 mm EGTA], the assay was started by adding 20μl of DAG kinase enzyme, followed by 80 μl of reaction buffer containing 5 μm ATP and 1 μCi of[γ-³²P]ATP. Incubation was for 1 h at room temperature; then PA was extracted, chromatographed,autoradiographed, and its radioactivity was counted in a liquid scintillation system. Standard curves were obtained as reported by Divecha et al. (5), using 1,2-dioleyl-3-palmitoyl-glycerol as substrate.

This was accomplished essentially as described by Flores et al. (25), with some modifications (19, 21). Assays (150 μl) contained 50 mmTris acetate (pH 7.4), 10 mm magnesium acetate,80 mm KCl, 2 mm EGTA, 50μg of nuclear protein, 50 μg/ml 1,2-dioleoylglycerol, and 20μ m ATP. The stock solution of 1,2-dioleoylglycerol in chloroform was dried under a nitrogen stream;resuspended in a small volume of 50 mm Tris acetate (pH 7.4), 10 mm magnesium acetate, 80 mm KCl, and 2 mm EGTA; and water bath sonicated before being added to the reaction mix. The reaction was started by adding 10 μCi of[γ-³²P]ATP. Standard assays were performed for 30 min at room temperature. When inhibitions studies were performed, DAG kinase inhibitors were added to the reaction mixture 10 min before adding 1,2-dioleoylglycerol. Again, no effect was seen when the same volume of solvent was added to the assays (data not shown). Lipids were extracted by the subsequent addition of 600 μl of CHCl₃/methanol (2:1), 50 μl of CHCl₃, and 50 μl of 0.1 nHCl. After centrifugations at 500 × g, the organic layers were recovered, dried under a stream of nitrogen,dissolved in 20 μl of CHCl₃/methanol (2:1), and applied to Silica Gel G-60 plates along with the dioleoyl PA standard. Plates were developed with a solvent system of CHCl₃, methanol, and 4 mNa₄OH (9:7:2, v/v/v) and autoradiographed. Lipids were identified by iodine staining, labeled PA spots were scraped off,and radioactivity was determined by liquid scintillation.

Immunofluorescence staining for PKC-α was performed essentially as reported by Zini et al. (26), using a polyclonal antibody diluted 1:400, followed by a Cy3-conjugated secondary antibody diluted 1:500. Samples were photographed using a Zeiss Axiophot epifluorescence microscope.

Proteins separated on 7.5% polyacrylamide gels (27) were transferred to nitrocellulose sheets using a semidry blotting apparatus(Hoefer/Pharmacia Biotech, Uppsala, Sweden). Sheets were saturated in PBS containing 5% NGS and 4% BSA for 60 min at 37°C (blocking buffer) and then incubated overnight at 4°C in blocking buffer containing primary antibody to PKC-α. After four washes in PBS containing 0.1% Tween 20, sheets were incubated for 30 min at room temperature with peroxidase-conjugated antirabbit IgG diluted 1:3000 in PBS-Tween 20 and washed as described above. Bands were visualized by the enhanced chemiluminescence method. Densitometric analysis of Western blots was performed on a Molecular Analyst GS670(Bio-Rad). In some cases, blots were stripped and reprobed with a monoclonal antibody to lamin B1, which was then visualized by means of a peroxidase-conjugated antimouse IgG.

Nuclear protein (200 μg) was separated by means of two-dimensional gel electrophoresis as reported elsewhere (28) and transferred to nitrocellulose sheets as described above. Blocked sheets were incubated for 1 h at 30°C in 50 mm Tris-HCl (pH 7.4), 150 mm NaCl, 10 mg/ml BSA, 20 μg/ml PS, 1 mm EGTA, 1.2 mm CaCl₂, 10μg/ml leupeptin and aprotinin, and 10 μg/ml recombinant rabbit PKC-α, essentially as described by Hyatt et al.(29). After two washes in PBS containing PS and Ca²⁺, blotted proteins were fixed in 0.5%formaldehyde in PBS for 20 min at room temperature and then incubated in 2% glycine in PBS for 20 min to block reactive aldehyde groups. After three washes in 50 mm Tris-HCl (pH 7.4) and 150 mm NaCl, samples were processed with anti-PKC-α antibody and secondary antibody as described for Western blots.

Isolated nuclei (10 μg of protein) were incubated at 30°C for 10 min in 20 mm Tris-HCl (pH 7.4), 10 mmMgCl₂, 10 μm ATP, 0.4 μg/ml histone H1, and 5 μCi of [γ-³²P]ATP in the presence of 1.2 mm CaCl₂ and 40 μg/ml PS. The reactions were terminated with 15 μl of acetic acid and spotted onto Whatman p81 paper, followed by washing with 0.75 mm H₃PO₄. Radioactivity was measured by Cerenkov counting.

Twenty-four h after beginning IGF-I stimulation, cells were labeled with 100 μm BrdUrd for 10 min, as described previously (30). Samples were fixed in freshly prepared 4% paraformaldehyde in PBS for 30 min at room temperature, and then DNA was denatured in 4 n HCl for 30 min and fixed in a−20°C graded ethanol series to prevent reannealing. Slides were then air dried and incubated for 3 h at 37°C with a monoclonal antibody to BrdUrd diluted 1:25 in PBS, 2% BSA, and 3% NGS. Slides were then washed three times in PBS and reacted with a Cy3-conjugated antimouse IgG diluted 1:400 in PBS, 2% BSA, and 5% NGS for 1 h at 37°C. Samples were subsequently washed three times in PBS and mounted as described above.

Data are the mean from three different experiments and are expressed as mean ± SD. In the tables, an asteriskindicates significant differences (P < 0.01)in a Student’s paired t test. All of the other differences were not significant (P > 0.01).

We first measured nuclear DAG levels and DAG kinase activity in response to IGF-I stimulation of quiescent Swiss 3T3 cells. As shown in Fig. 1, IGF-I exposure resulted in a 4-fold increase in nuclear DAG mass,which rose from 69 ± 9 to 271 ± 33 pmol/mg nuclear protein 30 min after the beginning of stimulation. However, at 60 min, the values had returned to basal levels(74 ± 10 pmol/mg nuclear protein). Afterward, no significant changes in the nuclear levels of DAG were measured up to 180 min after the beginning of stimulation. As far as nuclear DAG kinase activity was concerned, there was no significant increase after 30 min of exposure to IGF-I (125 ± 15 pmol PA produced/mg nuclear protein; basal level, 118 ± 16 pmol PA produced/mg nuclear protein). DAG kinase activity started to rise at 45 min, when the intranuclear levels of DAG showed a slight drop. At 60 min, we were able to detect a striking increase in nuclear DAG kinase activity, which rose to 366 ± 44 pmol PA produced/mg nuclear protein, i.e., a 3.1-fold increase. Nuclear DAG kinase activity remained high up to 90 min, and then it started to decrease, reaching basal levels at 180 min. When DAG kinase activity was measured in whole cell homogenates, no significant changes were seen at any time of stimulation (data not shown).

As shown in Table 1, the increased nuclear DAG activity was sensitive in vitroto two well-established inhibitors of the enzyme, R59022 and R59949. Indeed, the two chemicals were able to suppress about 75% of the nucleus-associated DAG kinase activity.

In Table 2 we demonstrate that when Swiss 3T3 cells were treated in vivo with either of the two DAG kinase inhibitors, the nuclear DAG mass remained elevated up to 120 min after the beginning of stimulation with IGF-I. We did not investigate longer times of stimulation.

Because a previous investigation showed DAG kinase activity to be preferentially associated with the internal nuclear matrix (19), we sought to determine whether or not the same holds true for IGF-I-stimulated Swiss 3T3 cells. To this end, we used the well-established nuclear subfractionation technique used by Payrastre et al. (19) in their studies with NIH 3T3 cells. This technique allows the separation of the internal matrix from the external matrix. The external matrix is mainly constituted of lamin proteins, whereas the internal matrix is composed of proteins of the ill-defined fibrogranular network (31, 32). As shown in Table 3, at time 0 (no stimulation), the specific activity of DAG kinase bound to intact nuclear matrix was similar to that of nuclei. However, at 90 min of IGF-I stimulation, it was markedly higher. It was also clear that the specific activity measured in the internal matrix fraction was strikingly higher than that in the external matrix, both in unstimulated cells (internal matrix:external matrix, 5.15) and in IGF-exposed cells (internal matrix:external matrix, 20.44). Therefore,the internal matrix was the nuclear domain where the highest induction of DAG kinase activity could be measured after stimulation of quiescent cells with IGF-I.

We have previously demonstrated that PKC-α translocates to the nucleus after exposure of quiescent Swiss 3T3 fibroblasts to IGF-I. The phenomenon starts after 5 min of stimulation (17, 33). However, in those studies, we did not investigate the behavior of intranuclear PKC-α at stimulation times longer than 45 min. As shown in Fig. 2,A, immunofluorescence staining for PKC-α in quiescent Swiss 3T3 cells revealed the enzyme to be mainly located in the cytoplasm. However, some intranuclear staining was present, in agreement with our own previous data (17, 33). At 30 min of stimulation,PKC-α was concentrated mainly in the nuclear region, with a corresponding decrease in the cytoplasm (Fig. 2,B). When cells stimulated for longer times with IGF-I (up to 180 min) were immunostained for PKC-α, it was evident that the enzyme was still located mainly within the nucleus (Fig. 2, C–F).

The results obtained by immunofluorescence staining were substantiated by Western blot analysis, which revealed a low amount of PKC-α in the nucleus of quiescent cells (Fig. 3,A). This amount increased at 30 min of stimulation and then remained constant up to 180 min. As a control, we probed the same blots with a monoclonal antibody to lamin B1, and this showed no changes in the amount of this nuclear protein proceeding from unstimulated cells to cells stimulated with IGF-I for various times (Fig. 3,B). Moreover, we checked to see that in vivo treatment of cells with the two DAG kinase inhibitors did not influence the intranuclear amount of PKC-α. As presented in Fig. 3 C,nuclei were prepared from cells stimulated for 30 min with IGF-I, both in the presence and in the absence of DAG kinase inhibitors. When the intranuclear amount of PKC-α was investigated by Western blot, it was evident that it was not dependent on the presence of inhibitors.

PKC-α remained within the nucleus even when the DAG mass had returned to basal levels, and we reasoned that this could be due to the presence of nuclear PKC-binding proteins. Therefore, we investigated the presence of some of these proteins using the well-established blot overlay assay technique described by Hyatt et al.(29). Nuclear protein was separated by means of two-dimensional gel electrophoresis, blotted to nitrocellulose paper,and incubated with recombinant PKC-α, which was then revealed by a specific polyclonal antibody. We saw a strong reaction with a group of three acidic proteins with a molecular weight of approximately 73,000/66,000 (Fig. 4,a). This reminded us of nuclear lamins A, B, and C, which are also substrates of PKC (34). Indeed, when duplicate blots were reacted with a monoclonal antibody recognizing all three nuclear lamins (J16; see Ref. 35), we obtained the same type of staining (Fig. 4,b). Another spot that gave a strong reaction exhibited a molecular weight of approximately 106,000 (Fig. 4,c). We reasoned that it could be protein C23/nucleolin,another nuclear substrate of PKC (36). In fact, a duplicate blot reacted with a monoclonal antibody (MS3) to C23/nucleolin confirmed our hypothesis (Fig. 4 d).

Previous findings by Sun et al. (10) have demonstrated that the increased DAG levels found in G₂-phase nuclei of HL-60 cells are able to support PKC-β_II-mediated lamin B phosphorylation in the absence of exogenously added DAG. Therefore, we sought to reproduce these results in our system. To this end, we always checked the amount of PKC-α present in nuclei prepared from cells treated for 30 min with IGF-I alone and in nuclei obtained from cells exposed to IGF-I and either of the two DAG kinase inhibitors for 60 min. As illustrated above, these nuclei have similar levels of PKC-α, but the amount of nuclear DAG is markedly different.

Therefore, nuclear preparations were aliquoted for Western blot analysis (5/10 of the final volume) and catalytic assays (5/10 of the final volume). Only nuclear preparations showing similar amounts of PKC-α protein (as determined by Western blot and densitometric scanning) were used for activity assays in the presence of PS and Ca²⁺ but in the absence of any exogenous DAG. Histone H1 was used as substrate. It is worth remembering that of the four PKC isoforms (PKC-α, -β_I, -ε, and-ζ) detected in Swiss 3T3 cells, only PKC-α is present in the nucleus (33). Thus, PKC activity assayed in isolated nuclei is due exclusively to this isozyme.

In Table 4 we show the results of densitometric scanning of the immunoblots and of the activity assays. It is evident that under all of the conditions tested, densitometric scanning of Western blots gave similar results,indicating the presence of an equivalent amount of PKC-α protein in our samples. The endogenous DAG was able to support PKC-α activity on exogenous histone H1 in all types of nuclei prepared from cells stimulated with IGF-I for 30 min (in the absence or presence of DAG kinase inhibitors). At 60 min of IGF-I stimulation, PKC-α activity decreased dramatically (more than 70%), whereas it did not change significantly in nuclei prepared from cells stimulated for 60 min with IGF-I and also exposed to DAG kinase inhibitors. It should also be underscored that similar results were obtained if no exogenous substrate was used, indicating that endogenous nuclear DAG was capable of supporting the PKC-dependent phosphorylation of endogenous nuclear proteins (data not shown; Refs. 37 and 38).

Because a previous investigation by Morris et al.(39) showed that the DAG kinase inhibitor R59022 potentiates bombesin-dependent DNA synthesis in Swiss 3T3 cells, we investigated whether or not this also happened with the IGF-I-stimulated DNA replication. As shown in Table 5, at 24 h after beginning stimulation, a marked increase in BrdUrd-positive cells was seen in samples treated for 24 h with IGF-I plus either of the two inhibitors. On the other hand, no stimulating effect was seen when the inhibitors were used without IGF-I. These findings strongly support the contention that DAG generated in the nucleus on IGF-I stimulation of quiescent cells is involved in control of the DNA synthetic machinery.

At present, great interest surrounds DAG kinase, the enzyme that phosphorylates DAG to PA, a molecule that also has signaling functions (18). Although it is commonly thought that the bulk of the signaling pool of PA derives from the activation of phospholipase D,DAG kinase certainly also contributes to it. Therefore, DAG kinase catalyzes a reaction that removes DAG, resulting in the termination of PKC-mediated cell signaling events, but it also yields a product, PA,with other signaling functions. For this reason, the net result of DAG kinase activation on cellular responses is quite difficult to predict,and contradictory results have been reported in the literature (18).

Nine isoforms of mammalian DAG kinase have been cloned thus far, and they are expressed differently in multiple cell types and in a wide range of tissues, an indication of their distinct functional significance (18).

Several precedent investigations have indicated the presence of DAG kinase activity in the nucleus of both hepatocytes and NIH 3T3 cells (19, 20, 21). Moreover, by immunocytochemical staining, Goto and Kondo (40) found that DAG kinase IV (the rat homologue of the recently cloned human DAG kinase-ζ) localizes in the nucleus of COS-7 cells when overexpressed as an epitope-tagged hybrid. These authors were also able to identify a nuclear targeting motif in the amino acid sequence of DAG kinase IV.

DAG kinase-ζ also possesses a nuclear localization sequence in a motif that corresponds to the PKC phosphorylation site of the myristylated alanine-rich C kinase substrate protein(MARCKS). Data by Topham et al. (22) have shown that phosphorylation of this site by either PKC-α or -γ determines whether or not the protein is localized to the nucleus, and also that conditional overexpression of DAG kinase -ζ in the nucleus can attenuate epidermal growth factor-dependent proliferation in A172 glioblastoma cells, conceivably because the intranuclear DAG mass is decreased.

When quiescent Swiss 3T3 cells are stimulated with a mitogenic concentration of IGF-I, there is a rapid and sustained increase in the nuclear DAG levels. However, at 60 min after beginning stimulation, the nuclear DAG mass returns to basal levels (5, 17). We have provided evidence that this rise in the nuclear DAG mass is dependent on activation of a PI-PLC and that it is essential to attract PKC-αto the nucleus (17).

In this study, we have shown that an inverse relationship exists in the nucleus of IGF-I-stimulated Swiss 3T3 cells between the levels of DAG and DAG kinase activity. Indeed, nuclear DAG kinase activity is unchanged after up to 30 min of IGF-I stimulation, and then it starts to rise, reaching the maximum at 90 min after the beginning of treatment. Afterward, it declines, returning to basal levels at 180 min. Treatment of intact cells with two well-established DAG kinase inhibitors blocked the IGF-I-dependent stimulation of nuclear DAG kinase and also maintained elevated DAG levels in the nucleus for up to 120 min of stimulation. These findings demonstrate that the drop in nuclear DAG mass that occurs in the nucleus at around 45 min of stimulation (5, 17) is related to the activation of nuclear DAG kinase.

After IGF-I stimulation, we found that DAG kinase activity is highly enriched in the insoluble nuclear matrix and especially in the internal nuclear matrix fraction, where the maximal induction is seen in response to the mitogenic stimulus. This is a further indication that some aspects of intranuclear cell signaling take place in close association with the insoluble inner nuclear skeleton (3, 4, 19, 33). Furthermore, it should also be remembered that PKC-α has been shown to preferentially associate with the internal nuclear matrix in response to IGF-I stimulation of quiescent 3T3 cells (41). These findings hint at an important role for compartmentalization of signaling pathways elicited by IGF-I within the cells.

Our immunocytochemical and immunochemical investigations demonstrated that elevated intranuclear levels of PKC-α protein are detectable after up to 180 min of IGF-I stimulation. Sun et al. (10) were able to correlate the elevated intranuclear levels of DAG with their ability to support the phosphorylation of lamin B by PKC-β_II. We have been able to reproduce their results in our experimental model because,by using nuclear preparations with similar levels of PKC-α protein but markedly different amounts of DAG, we demonstrated that elevated endogenous DAG levels were able to support PKC-α-dependent histone H1 phosphorylation. This leads us to consider that the increase in nuclear DAG mass that follows stimulation with IGF-I is also important in controlling the activity of PKC-α within the nucleus, and not only for its attraction to this organelle. Thus, it is conceivable that when the intranuclear DAG mass diminishes, PKC-α activity is reduced in vivo. However, PKC-α remains inside the nucleus also when the nuclear levels of DAG are reduced to those of unstimulated cells. To explain this apparently contradictory finding, we reasoned that once in the nucleus, PKC-α might interact with PKC-binding proteins, which are then responsible for anchoring the protein and maintaining it within the nuclear compartment. Indeed, our results,obtained by means of blot overlay assays, indicated the existence of multiple PKC-binding proteins, including lamins A, B, and C and C23/nucleolin, in the nuclei of Swiss 3T3 cells. To our knowledge, this is the first report addressing the issue of the presence of PKC-binding proteins in the nucleus.

We have also demonstrated that when the increase in DAG kinase activity was inhibited by two chemicals, there was a marked increase in the number of cells that entered S phase in response to stimulation with IGF-I. This finding is in total agreement with Topham et al. (22), who, in their investigations using A172 cells, demonstrated that the overexpression of wild-type DAG kinase-ζ, but not of a kinase-dead mutant, induced an accumulation of cells in the G₀-G₁ phases of the cell cycle.

Somewhat opposite results have been reported by Flores et al. (25), who showed that in resting T lymphocytes,interleukin 2 induced a translocation of DAG kinase α from the nucleus to the perinuclear region, and this was accompanied by an enhanced production of PA. Nevertheless, treatment of cells with the same DAG kinase inhibitors we have used in this study induced an arrest in late G₁ in the interleukin-dependent cells. However, in their study, Flores et al. (25) did not investigate whether or not changes in nuclear DAG and PA levels also occurred.

An important issue that still awaits resolution is whether or not the marked intranuclear increase of DAG kinase activity that follows IGF-I stimulation of quiescent 3T3 cells depends on the activation of a resident enzyme or on translocation to the nucleus of a cytoplasmic protein. Despite the fact that there are several articles in the literature showing the existence of DAG kinase in Swiss 3T3 mouse fibroblasts (e.g., Refs. 42, 43, 44, 45), we have been unable to find a report in which there is an indication of the isoform(s) expressed by these cells. This fact, coupled to the lack of antibodies, prevented us from performing the experiments required to clarify this fundamental question. Topham et al.(22) demonstrated that phosphorylation of DAG kinase-ζby PKC-α blocked intranuclear localization of DAG kinase. However, we feel that activation of nuclear DAG kinase in IGF-I-exposed Swiss 3T3 cells is conceivably not related to the intranuclear migration of PKC-α. Indeed, phosphorylation of DAG kinase-ζ by PKC-αinterferes with the nuclear localization of DAG kinase in the nucleus,but there is no evidence that such a posttranslational modification enhances enzyme activity. Secondly, activation of DAG kinase follows intranuclear translocation of PKC-α. If DAG kinase were phosphorylated by PKC-α within the nucleus, and this resulted in migration of DAG kinase outside the nucleus, then a lowering of DAG kinase activity would ensue, not an increase. There is also evidence that DAG kinase activity might be regulated by Ca²⁺, arachidonic acid, sphingosine, ceramide,and several types of fatty acids (for an updated review, see Ref. 18). Because these molecules have also been found within the nucleus (3, 4), they might be involved in the regulation of nuclear DAG kinase activity, and this issue should be clarified in the future.

To our knowledge, this is the first report showing activation of an endogenous nuclear DAG kinase activity after agonist stimulation at the plasma membrane. Indeed, the results of Topham et al.(22) have been obtained mainly by using cell lines overexpressing DAG kinase. Although there are several issues that still await resolution, we feel that our findings strongly substantiate the notion that DAG and DAG kinase present in the nucleus play a prominent role in IGF-I-elicited nuclear signaling events that lead to cell proliferation and possibly to neoplastic transformation.

We thank Dr. A. M. Billi for valuable technical support and Dr. R. L. Ochs for the gift of the antibody.