Eur. J. Biochem. 210,45-51 (1992)
0FEBS 1992
Nuclear Substrates of Protein Kinase C Roland BECKMANN', Klaus BUCHNER', Peter R. JUNGBLUT', Christoph ECKERSKORN3, Christoph WEISE and Ferdinand HUCHO'
', Ralf HILBERT'
' Institut fur Biochemie, Freie Universitat Berlin, Federal Republic of Germany Deutsches Hcrzzentrum Berlin, Berlin, Federal Republic of Germany MPI fur Biochemie, Genzentrum, Martinsried, Federal Republic of Germany (Received June 23,1992)
-
EJB 92 0872
Starting from the finding that, for neuronal cells, the nuclear-membrane-associated protein kinase C (PKC) is the so-called 'membrane inserted', constitutively active form, we attempted to identify substrates of this nuclear PKC. For this purpose, nuclear membranes and other subcellular fractions were prepared from bovine brain, and in-vitro phosphorylation was performed. Several nuclear membrane proteins were found, the phosphorylation of which was inhibited by specific PKC inhibitors and effectively catalyzed by added PKC. Combining the methods of two-dimensional gel electrophoresis, in-situ digestion, reverse-phase HPLC and microsequencing, two of these nuclear PKC substrates were identified; the known PKC substrate Lamin B2, which serves as a control of the approach and the nucleolar protein B23. Our data suggest, that, for B23, Ser225 is a site of phosphorylation by PKC.
The nuclear envelope is an efficient barrier between cytoplasmic and nuclear compartments, controlling trafficing of, among others, macromolecules, components for their synthesis, ions and signal molecules regulating transcription (Gerace and Burke, 1988). It is composed of two concentric bilayer membranes, separated by the perinuclear space. The outer nuclear membrane and the perinuclear space are continuous with the rough endoplasmatic reticulum and its lumen, respectively. Exchange between nuclear and cytoplasmic compartments is thought to take place via nuclear-pore complexes which span both membranes. Analogous to events taking place at the plasma membrane, one could envisage a second type of intercompartmental exchange. Signals regulating chromosomal functions could be transduced by components of the nuclear membranes outside the pore complexes. Since it is of central importance, much research is focussed on the molecular architecture of the pore complexes, and a wealth of biochemical and ultrastructural data is available (for review see Hanover, 1992). However, relatively little is known concerning the biochemistry of the rest of the nuclear membranes. Recently, various signal-transduction components have been detected which seem to be specific for the nuclear envelope. Various GTP-binding proteins were identiCorrespondence to F. Hucho, Institut fur Biochemie, Freie Universitat Berlin, Thielallee 63, W-1000 Berlin 33, Federal Republic of Germany Fax: + 49 30 838 3753. Abbreviations. PKC, protein kinase C; PKA, protein kinase A; LDH, lactate dchydrogenase; 2-D, two dimensional. Enzymes. Protein kinase C (EC 2.7.1. -); protein kinase A (EC 2.1.7.37); 5'-nucleotidase (EC 3.1.3.5); lactate dehydrogenase (EC 3 .I .1.27).
fied (Seydel and Gerace, 1991; Rubins et al., 1990; Otto et al., 1992), potassium channels (Mazzanti et al., 1990), chloride channels (Tabares et al., 1991) and even an InsP, response system (Nicotera et al., 1990). A further signal-transducing pathway, in addition to the nuclear-pore complexes, involve the protein kinase C (PKC). If located in the nuclear envelope, it could transmit cytoplasmic signals to the chromatin by being activated from the cytoplasmic side and by phosphorylating nucleoplasma pro-
Fig. 1. Neuronal and glial nuclei. Phase-contrast picture; calibration bar = 30 pm. Small and dense nuclei with adiameter of approximately 7 pm are reported to be glial nuclei, the others with a diameter of approximately 12 pm, to be neuronal nuclei (Thompson, 1987).
46 Table 1. Enzymatic characterization of the preparation. n.d., not determined. For protein, DNA, PKC and 5’-nucleotidase, n = 3. LDH shows a typical preparation.
Fraction
Homogenate Nuclei Nuclear membrane Cytosol Plasma membrane
Protein
DNA
PKC
5’ nucleotidase
c mg
ng/g
nmol/g
u is
8970 k2110 5.5 20.4+ 1.5 9.9f n.d. n.d.
11f 6 1777 324 3 2 + 12 n. d. n.d.
21.0 f 1.5 7.6 & 1.2 6.5 f 0.5 2.8 0.7 14.6 f 7.8
18.8 -t 2.5 7.7 1.9 5.2f 1.0 n.d. 153.6 f 32.6
4.5
I
pl
0.5
I
4.5
I
1469 33 0 3733 n. d.
+
PI
LDH
0.5
I
kDa
--
76.0 66.2 43.0 36.0
-
-
21.5 17.5 31.0
76.066.243.036.031.021.517.5-
Fig. 2.2-D electrophoresispatterns of subcellular fractions. 2-D gel electrophoresis was performed with 150 pg protein of (A) nuclear membranes (Coomassie-blue stain), (B) microsomes (Coomassie-blue stain), (C) plasma membranes (Coomassic-blue stain), (D) nuclear membranes (silver stain). The first dimension (non-equilibrium pH-gradient electrophoresis) consisted of a 4% polyacrylamide rod gel with a diameter of I .5 mm. The second dimension was a 15% polyacrylamide gel according to Laemmli (1970). Triangles indicate proteins specific for nuclear membranes.
teins. Various pieces of evidence have been presented that PKC is indeed located in the nucleus (Rogue et al., 1990; Misra and Sahyoun, 1987; Buchner et al., 1992); it may even be translocated to the nucleus as a consequence of various extracellular signals ranging from phorbol esters (Leach et al., 1989), bryostatin (Hocevar and Fields, 1991), to bradykinin (Hilbert, R. and Buchner, K., unpublished results). The PKC seems to shuttle between different cellular compartments, transmitting information by phosphorylating various proteins which have key functions in cellular activities. In the present investigation, we attempted to identify substrates of the PKC in brain nuclei. By applying two-dimensional (2-D) SDS/polyacrylamide gel electrophoresis to stateof-the-art nuclear preparations in combination with microsequencing techniques, we found one known (lamin B) and one unexpected nuclear PKC substrate (the nucleolar protein B23).
MATERIALS AND METHODS Materials DNAse I was from Boehringer and heparin from Sigma. P from [3H]Phorbol 12,13-dibutyrate and [ Y - ~ ~ P I A Twere NEN. The PKC inhibitors C2 (compound 2; Ro 31-7549) and C3 (compound 3; Ro 31-8220) were generous gifts from Dr P. D. Davis, Roche Products. Purified PKC was prepared from bovine brain as described previously (Kruger et al., 1990). Other chemicals were purchased from Sigma or BioRad in the highest quality available. Subcellular fractionation Nuclei, nuclear membranes, plasma membranes, microsomes and cytosol were prepared from bovine brain cortex as described previously (Otto et al., 1992).
47 Briefly, the nuclei were obtained from homogenized cortex by centrifugation at 170 g for 15 min, centrifugation of the resuspended pellet at 370 g for 15 min, followed by sedimentation through 2.0 M sucrose at 67000 g for 60 min. Finally, the resuspended nuclei were sedimented through a 2.1 5-M sucrose layer by centrifugation at 67000 g for 60 min. Nuclear membranes were prepared by incubating nuclei in 200 ml hypotonic buffer (10 mM Tris/HCl, pH 8.0, 10 mM Na2HP04)with heparin (50 mg) and DNAse 1(2 mg/approximately 6000 U), followed by centrifugation at 9250 g for 10 min (Otto et al., 1992). For preparing plasma membranes, microsomes and cytosol by differential-gradient and density-gradient centrifugation, the 170 g supernatant of the homogenate was used. Plasma membranes were collected at the interface between a 0.83-M and a 1.05-M sucrose layer after 40 rnin centrifugation at 105000 g. For preparing microsomes and the cytosolic fraction, a high-speed centrifugation at 105000 g in 0.25 M sucrose for 12 h was performed. The pellet was designated as the microsome fraction and the supernatant as the cytosolic fraction.
4.5
I
pl
8.5
I
kDa
43.0 36.0 31.0 21.5 17.5 -
76.0 66.2
--
76.0
66.2
43.036.031.0-
Characterization of the preparation 21.5-
Integrity of the nuclei was judged using phase-contrast microscopy. Protein determination was performed according to Bradford (1976) using bovine serum albumin as standard. The DNA content was determined according to Chandra and Appel (1970). The purity of the nuclei and nuclear membranes was judged by assaying 5'-nucleotidase (Kai et al., 1966) as a plasma membrane marker and lactate dehydrogenase (Storrie and Madden, 1990) as a cytosolic marker. For determination of PKC, phorbol-ester-binding assays were performed as described by Uchida and Filburn (1 984).
17.5-
76.066.2-
31.O-
43.0 36.0
21.5-
Phosphorylation of nuclear membrane proteins Membranes were phosphorylated in reaction mixtures containing 20 mM Hepes, pH 7.4, 1 mM dithioerytritol, 5 mM magnesium acetate and 0.1 mM calcium chloride and additions listed in the figure legends. Activators or inhibitors of PKC or protein kinase A (PKA) were added 10 min before the start of the reaction by ATP (final concentration 20 pM, 2 pCi [Y-~~PI-ATP). After 2 rnin incubation at 30"C, the reaction was stopped by addition of ice-cold 60% trichloroacetic acid. Precipitated proteins were redissolved in 20 p1 sample buffer for 2-D gel electrophoresis. After 2-D PAGE, gels were stained with Coomassie blue, dried and autoradiographed using Kodak X-OMAT AR films. 2-D gel electrophoresis
2-D gel electrophoresis and silver staining were performed exactly as described by Jungblut and Seifert (1990).
-
17.5
Fig. 3. Kinase activity of nuclear membranes and inhibition by the specific PKC inhibitor C2. Nuclear membranes (70 pg protein) were phosphorylated by endogenous kinase activity as described in Materials and Methods, separated by 2-D PAGE and autoradiographed. Autoradiographs of phosphorylated nuclear membranes; (A) control, (B) 1 pM PKC-inhibitor C2 (Davis et al., 1989), (C) 10 pM PKC inhibitor C2. Arrows indicate proteins, the phosphorylation of which is inhibited.
D, 2 mm x 100 mm, Merck). The solvent system used was 0.1% (by vol.) trifluoroacetic acid in water (solvent A) and 0.1 YO(by vol.) trifluoroacetic acid in acetonitrile (solvent B). A gradient of 0-60% B was run over 90 rnin using a flow rate of 0.3 ml/min. Fractions were collected manually. Amino-acid-sequence analysis of the purified peptides was performed using a gas-phase sequencer 473 A (Applied Biosystems) or the Berlin sequencer (Knauer GmbH).
Identification of proteins from 2-D gels Protein spots were excised from 24 Coomassie-bluestained gels. The proteins were digested in the gel matrix with trypsin and the resulting peptides eluted as described by Eckerskorn and Lottspeich (1989). The eluted peptides were separated by reverse-phase chromatography using a CI8 column (LiChro CART CHR
RESULTS Preparation of nuclei and nuclear membrane Pure and reproducible preparations of nuclei and nuclear membranes were prerequisites for this investigation. Fig. 1 shows a typical preparation obtained from bovine cortex, as
48 4.5
I
PI
8.5
I
4.5
1
Pi
8.5
I
kDa
43.076.0 66.2
36.0
-
-
31.0
21.517.5
-
43.0 36.076.0 66.2
31.021.5
-
17.5-
Fig. 4. Phosphorylation of nuclear membranes by added PKC or activated PKA. Autoradiographs of phosphorylated nuclear membranes (70 pg protein) after 2-D PAGE; (A) control, (B) 44 ng PKC (purified from bovine brain as described by Kriiger et al., 1990) and 1.6 pM phorbol 12-myristate 13-acetate. Arrows indicate proteins, the phosphorylation of which was efficiently catalyzed by added purified PKC. (C) Control; (D) 2.5 pM cAMP and 1 mM 3-isobutyl-I-methylxanthine. Proteins, the phosphorylation of which was stimulated by CAMP, are indicated by arrows. The triangle indicates a protein, the phosphorylation of which is inhibited by PKA stimulation. The otherwise identical controls A and C look different because of different exposure times of the autoradiographs (one and three days, respectively).
seen in the phase-contrast microscope. Nuclei of two distinct sizes can be seen, the larger ones representing neuronal, the smaller ones glial nuclei (Thompson, 1987). Very few impurities, e. g. other organelles, membrane vesicles or cellular ‘debris’, are detectable. Purity was further checked by measuring marker enzymes, especially lactate dehydrogenase for cytosolic, and 5’-nucleotidase for plasma membrane contaminants (Table 1). As described elsewere, the nuclear and nuclear-membrane fractions contained significant amounts of PKC. This was previously identified as a ‘membrane-inserted’, i. e. the constitutively active form of the a- and y-isozymes (Buchner et al., 1992). PKC substrates in the nuclear-membrane fraction Chromatin and nucleoplasma were removed from nuclei using a combined DNAse and heparin treatment. This procedure, avoiding high-salt washing steps as applied in standard membrane purifications, preserved the PKC activity to a large extent (see Table 1). Fig. 2 shows the protein composition analyzed by 2-D gel electrophoresis (Coomassie blue stained) of (A) nuclear membranes, as compared to (B) microsomal, and (C) plasma membranes. The results of this experiment are as follows. (a) The protein patterns are different. This is further proof for the purity of the preparations. In Fig. 2A, two proteins occuring exclusively in the nuclear-membrane preparation are marked. (b) The protein patterns are surprisingly complex. The complexity is even more evident in Fig. 2D, showing a silverstained 2-D electrophoresis gel. Besides several proteins of the
nuclear-pore complex, and some of the membrane-associated matrix proteins, none of these proteins has been yet characterized. Several of the proteins appearing in Fig. 2 are phosphorylated by protein kinases (Fig. 3A). Among these, we concentrated on possible PKC substrates. Since the PKC in nuclear membranes is the ‘membrane inserted’ permanently active form (Buchner et al., 1992),differentiation against other protein kinases could not be achieved through specific activators. However, there are kinase substrates, the phosphorylation of which was especially susceptible to inhibition by low concentrations of the PKC inhibitors C2 (Fig. 3) and C3 (data not shown) and was most efficiently catalyzed by added purified PKC (Fig. 4A, B). The significance of these criteria, which define a protein as a PKC substrate, is supported by Fig. 4C, D ; PKA of the nuclear membrane, activated by cAMP in the presence of the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine, stimulated phosphorylation of a different set of proteins (indicated by arrows). Interestingly, stimulation of PKA-inhibited phosphorylation of some proteins, assumed to be catalyzed by PKC (in Fig. 4 D one example is indicated by the triangle). Comparing the autoradiographs of 2-D gels of phosphorylated microsomal and plasma membranes with the autoradiograph of nuclear membranes (Fig. 5 ) provides further evidence for the nuclear location of the kinase substrates we are interested in. The distinctly different pattern is obvious. This is also further proof for the purity of the membrane preparations; very little cross-contamination can be detected.
49 4.5
I
pl
RO-SP66
8.5
I
70
kDa
I!
i
-
76.0 66.2 43.0 36.0
-
21.5 31.0
-
17.5
10
20
40
30 (min
T i m e
50
1
RO-5P38
:Ij
76.066.243.036.031.0-
60
0FEBS 1992
Nuclear Substrates of Protein Kinase C Roland BECKMANN', Klaus BUCHNER', Peter R. JUNGBLUT', Christoph ECKERSKORN3, Christoph WEISE and Ferdinand HUCHO'
', Ralf HILBERT'
' Institut fur Biochemie, Freie Universitat Berlin, Federal Republic of Germany Deutsches Hcrzzentrum Berlin, Berlin, Federal Republic of Germany MPI fur Biochemie, Genzentrum, Martinsried, Federal Republic of Germany (Received June 23,1992)
-
EJB 92 0872
Starting from the finding that, for neuronal cells, the nuclear-membrane-associated protein kinase C (PKC) is the so-called 'membrane inserted', constitutively active form, we attempted to identify substrates of this nuclear PKC. For this purpose, nuclear membranes and other subcellular fractions were prepared from bovine brain, and in-vitro phosphorylation was performed. Several nuclear membrane proteins were found, the phosphorylation of which was inhibited by specific PKC inhibitors and effectively catalyzed by added PKC. Combining the methods of two-dimensional gel electrophoresis, in-situ digestion, reverse-phase HPLC and microsequencing, two of these nuclear PKC substrates were identified; the known PKC substrate Lamin B2, which serves as a control of the approach and the nucleolar protein B23. Our data suggest, that, for B23, Ser225 is a site of phosphorylation by PKC.
The nuclear envelope is an efficient barrier between cytoplasmic and nuclear compartments, controlling trafficing of, among others, macromolecules, components for their synthesis, ions and signal molecules regulating transcription (Gerace and Burke, 1988). It is composed of two concentric bilayer membranes, separated by the perinuclear space. The outer nuclear membrane and the perinuclear space are continuous with the rough endoplasmatic reticulum and its lumen, respectively. Exchange between nuclear and cytoplasmic compartments is thought to take place via nuclear-pore complexes which span both membranes. Analogous to events taking place at the plasma membrane, one could envisage a second type of intercompartmental exchange. Signals regulating chromosomal functions could be transduced by components of the nuclear membranes outside the pore complexes. Since it is of central importance, much research is focussed on the molecular architecture of the pore complexes, and a wealth of biochemical and ultrastructural data is available (for review see Hanover, 1992). However, relatively little is known concerning the biochemistry of the rest of the nuclear membranes. Recently, various signal-transduction components have been detected which seem to be specific for the nuclear envelope. Various GTP-binding proteins were identiCorrespondence to F. Hucho, Institut fur Biochemie, Freie Universitat Berlin, Thielallee 63, W-1000 Berlin 33, Federal Republic of Germany Fax: + 49 30 838 3753. Abbreviations. PKC, protein kinase C; PKA, protein kinase A; LDH, lactate dchydrogenase; 2-D, two dimensional. Enzymes. Protein kinase C (EC 2.7.1. -); protein kinase A (EC 2.1.7.37); 5'-nucleotidase (EC 3.1.3.5); lactate dehydrogenase (EC 3 .I .1.27).
fied (Seydel and Gerace, 1991; Rubins et al., 1990; Otto et al., 1992), potassium channels (Mazzanti et al., 1990), chloride channels (Tabares et al., 1991) and even an InsP, response system (Nicotera et al., 1990). A further signal-transducing pathway, in addition to the nuclear-pore complexes, involve the protein kinase C (PKC). If located in the nuclear envelope, it could transmit cytoplasmic signals to the chromatin by being activated from the cytoplasmic side and by phosphorylating nucleoplasma pro-
Fig. 1. Neuronal and glial nuclei. Phase-contrast picture; calibration bar = 30 pm. Small and dense nuclei with adiameter of approximately 7 pm are reported to be glial nuclei, the others with a diameter of approximately 12 pm, to be neuronal nuclei (Thompson, 1987).
46 Table 1. Enzymatic characterization of the preparation. n.d., not determined. For protein, DNA, PKC and 5’-nucleotidase, n = 3. LDH shows a typical preparation.
Fraction
Homogenate Nuclei Nuclear membrane Cytosol Plasma membrane
Protein
DNA
PKC
5’ nucleotidase
c mg
ng/g
nmol/g
u is
8970 k2110 5.5 20.4+ 1.5 9.9f n.d. n.d.
11f 6 1777 324 3 2 + 12 n. d. n.d.
21.0 f 1.5 7.6 & 1.2 6.5 f 0.5 2.8 0.7 14.6 f 7.8
18.8 -t 2.5 7.7 1.9 5.2f 1.0 n.d. 153.6 f 32.6
4.5
I
pl
0.5
I
4.5
I
1469 33 0 3733 n. d.
+
PI
LDH
0.5
I
kDa
--
76.0 66.2 43.0 36.0
-
-
21.5 17.5 31.0
76.066.243.036.031.021.517.5-
Fig. 2.2-D electrophoresispatterns of subcellular fractions. 2-D gel electrophoresis was performed with 150 pg protein of (A) nuclear membranes (Coomassie-blue stain), (B) microsomes (Coomassie-blue stain), (C) plasma membranes (Coomassic-blue stain), (D) nuclear membranes (silver stain). The first dimension (non-equilibrium pH-gradient electrophoresis) consisted of a 4% polyacrylamide rod gel with a diameter of I .5 mm. The second dimension was a 15% polyacrylamide gel according to Laemmli (1970). Triangles indicate proteins specific for nuclear membranes.
teins. Various pieces of evidence have been presented that PKC is indeed located in the nucleus (Rogue et al., 1990; Misra and Sahyoun, 1987; Buchner et al., 1992); it may even be translocated to the nucleus as a consequence of various extracellular signals ranging from phorbol esters (Leach et al., 1989), bryostatin (Hocevar and Fields, 1991), to bradykinin (Hilbert, R. and Buchner, K., unpublished results). The PKC seems to shuttle between different cellular compartments, transmitting information by phosphorylating various proteins which have key functions in cellular activities. In the present investigation, we attempted to identify substrates of the PKC in brain nuclei. By applying two-dimensional (2-D) SDS/polyacrylamide gel electrophoresis to stateof-the-art nuclear preparations in combination with microsequencing techniques, we found one known (lamin B) and one unexpected nuclear PKC substrate (the nucleolar protein B23).
MATERIALS AND METHODS Materials DNAse I was from Boehringer and heparin from Sigma. P from [3H]Phorbol 12,13-dibutyrate and [ Y - ~ ~ P I A Twere NEN. The PKC inhibitors C2 (compound 2; Ro 31-7549) and C3 (compound 3; Ro 31-8220) were generous gifts from Dr P. D. Davis, Roche Products. Purified PKC was prepared from bovine brain as described previously (Kruger et al., 1990). Other chemicals were purchased from Sigma or BioRad in the highest quality available. Subcellular fractionation Nuclei, nuclear membranes, plasma membranes, microsomes and cytosol were prepared from bovine brain cortex as described previously (Otto et al., 1992).
47 Briefly, the nuclei were obtained from homogenized cortex by centrifugation at 170 g for 15 min, centrifugation of the resuspended pellet at 370 g for 15 min, followed by sedimentation through 2.0 M sucrose at 67000 g for 60 min. Finally, the resuspended nuclei were sedimented through a 2.1 5-M sucrose layer by centrifugation at 67000 g for 60 min. Nuclear membranes were prepared by incubating nuclei in 200 ml hypotonic buffer (10 mM Tris/HCl, pH 8.0, 10 mM Na2HP04)with heparin (50 mg) and DNAse 1(2 mg/approximately 6000 U), followed by centrifugation at 9250 g for 10 min (Otto et al., 1992). For preparing plasma membranes, microsomes and cytosol by differential-gradient and density-gradient centrifugation, the 170 g supernatant of the homogenate was used. Plasma membranes were collected at the interface between a 0.83-M and a 1.05-M sucrose layer after 40 rnin centrifugation at 105000 g. For preparing microsomes and the cytosolic fraction, a high-speed centrifugation at 105000 g in 0.25 M sucrose for 12 h was performed. The pellet was designated as the microsome fraction and the supernatant as the cytosolic fraction.
4.5
I
pl
8.5
I
kDa
43.0 36.0 31.0 21.5 17.5 -
76.0 66.2
--
76.0
66.2
43.036.031.0-
Characterization of the preparation 21.5-
Integrity of the nuclei was judged using phase-contrast microscopy. Protein determination was performed according to Bradford (1976) using bovine serum albumin as standard. The DNA content was determined according to Chandra and Appel (1970). The purity of the nuclei and nuclear membranes was judged by assaying 5'-nucleotidase (Kai et al., 1966) as a plasma membrane marker and lactate dehydrogenase (Storrie and Madden, 1990) as a cytosolic marker. For determination of PKC, phorbol-ester-binding assays were performed as described by Uchida and Filburn (1 984).
17.5-
76.066.2-
31.O-
43.0 36.0
21.5-
Phosphorylation of nuclear membrane proteins Membranes were phosphorylated in reaction mixtures containing 20 mM Hepes, pH 7.4, 1 mM dithioerytritol, 5 mM magnesium acetate and 0.1 mM calcium chloride and additions listed in the figure legends. Activators or inhibitors of PKC or protein kinase A (PKA) were added 10 min before the start of the reaction by ATP (final concentration 20 pM, 2 pCi [Y-~~PI-ATP). After 2 rnin incubation at 30"C, the reaction was stopped by addition of ice-cold 60% trichloroacetic acid. Precipitated proteins were redissolved in 20 p1 sample buffer for 2-D gel electrophoresis. After 2-D PAGE, gels were stained with Coomassie blue, dried and autoradiographed using Kodak X-OMAT AR films. 2-D gel electrophoresis
2-D gel electrophoresis and silver staining were performed exactly as described by Jungblut and Seifert (1990).
-
17.5
Fig. 3. Kinase activity of nuclear membranes and inhibition by the specific PKC inhibitor C2. Nuclear membranes (70 pg protein) were phosphorylated by endogenous kinase activity as described in Materials and Methods, separated by 2-D PAGE and autoradiographed. Autoradiographs of phosphorylated nuclear membranes; (A) control, (B) 1 pM PKC-inhibitor C2 (Davis et al., 1989), (C) 10 pM PKC inhibitor C2. Arrows indicate proteins, the phosphorylation of which is inhibited.
D, 2 mm x 100 mm, Merck). The solvent system used was 0.1% (by vol.) trifluoroacetic acid in water (solvent A) and 0.1 YO(by vol.) trifluoroacetic acid in acetonitrile (solvent B). A gradient of 0-60% B was run over 90 rnin using a flow rate of 0.3 ml/min. Fractions were collected manually. Amino-acid-sequence analysis of the purified peptides was performed using a gas-phase sequencer 473 A (Applied Biosystems) or the Berlin sequencer (Knauer GmbH).
Identification of proteins from 2-D gels Protein spots were excised from 24 Coomassie-bluestained gels. The proteins were digested in the gel matrix with trypsin and the resulting peptides eluted as described by Eckerskorn and Lottspeich (1989). The eluted peptides were separated by reverse-phase chromatography using a CI8 column (LiChro CART CHR
RESULTS Preparation of nuclei and nuclear membrane Pure and reproducible preparations of nuclei and nuclear membranes were prerequisites for this investigation. Fig. 1 shows a typical preparation obtained from bovine cortex, as
48 4.5
I
PI
8.5
I
4.5
1
Pi
8.5
I
kDa
43.076.0 66.2
36.0
-
-
31.0
21.517.5
-
43.0 36.076.0 66.2
31.021.5
-
17.5-
Fig. 4. Phosphorylation of nuclear membranes by added PKC or activated PKA. Autoradiographs of phosphorylated nuclear membranes (70 pg protein) after 2-D PAGE; (A) control, (B) 44 ng PKC (purified from bovine brain as described by Kriiger et al., 1990) and 1.6 pM phorbol 12-myristate 13-acetate. Arrows indicate proteins, the phosphorylation of which was efficiently catalyzed by added purified PKC. (C) Control; (D) 2.5 pM cAMP and 1 mM 3-isobutyl-I-methylxanthine. Proteins, the phosphorylation of which was stimulated by CAMP, are indicated by arrows. The triangle indicates a protein, the phosphorylation of which is inhibited by PKA stimulation. The otherwise identical controls A and C look different because of different exposure times of the autoradiographs (one and three days, respectively).
seen in the phase-contrast microscope. Nuclei of two distinct sizes can be seen, the larger ones representing neuronal, the smaller ones glial nuclei (Thompson, 1987). Very few impurities, e. g. other organelles, membrane vesicles or cellular ‘debris’, are detectable. Purity was further checked by measuring marker enzymes, especially lactate dehydrogenase for cytosolic, and 5’-nucleotidase for plasma membrane contaminants (Table 1). As described elsewere, the nuclear and nuclear-membrane fractions contained significant amounts of PKC. This was previously identified as a ‘membrane-inserted’, i. e. the constitutively active form of the a- and y-isozymes (Buchner et al., 1992). PKC substrates in the nuclear-membrane fraction Chromatin and nucleoplasma were removed from nuclei using a combined DNAse and heparin treatment. This procedure, avoiding high-salt washing steps as applied in standard membrane purifications, preserved the PKC activity to a large extent (see Table 1). Fig. 2 shows the protein composition analyzed by 2-D gel electrophoresis (Coomassie blue stained) of (A) nuclear membranes, as compared to (B) microsomal, and (C) plasma membranes. The results of this experiment are as follows. (a) The protein patterns are different. This is further proof for the purity of the preparations. In Fig. 2A, two proteins occuring exclusively in the nuclear-membrane preparation are marked. (b) The protein patterns are surprisingly complex. The complexity is even more evident in Fig. 2D, showing a silverstained 2-D electrophoresis gel. Besides several proteins of the
nuclear-pore complex, and some of the membrane-associated matrix proteins, none of these proteins has been yet characterized. Several of the proteins appearing in Fig. 2 are phosphorylated by protein kinases (Fig. 3A). Among these, we concentrated on possible PKC substrates. Since the PKC in nuclear membranes is the ‘membrane inserted’ permanently active form (Buchner et al., 1992),differentiation against other protein kinases could not be achieved through specific activators. However, there are kinase substrates, the phosphorylation of which was especially susceptible to inhibition by low concentrations of the PKC inhibitors C2 (Fig. 3) and C3 (data not shown) and was most efficiently catalyzed by added purified PKC (Fig. 4A, B). The significance of these criteria, which define a protein as a PKC substrate, is supported by Fig. 4C, D ; PKA of the nuclear membrane, activated by cAMP in the presence of the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine, stimulated phosphorylation of a different set of proteins (indicated by arrows). Interestingly, stimulation of PKA-inhibited phosphorylation of some proteins, assumed to be catalyzed by PKC (in Fig. 4 D one example is indicated by the triangle). Comparing the autoradiographs of 2-D gels of phosphorylated microsomal and plasma membranes with the autoradiograph of nuclear membranes (Fig. 5 ) provides further evidence for the nuclear location of the kinase substrates we are interested in. The distinctly different pattern is obvious. This is also further proof for the purity of the membrane preparations; very little cross-contamination can be detected.
49 4.5
I
pl
RO-SP66
8.5
I
70
kDa
I!
i
-
76.0 66.2 43.0 36.0
-
21.5 31.0
-
17.5
10
20
40
30 (min
T i m e
50
1
RO-5P38
:Ij
76.066.243.036.031.0-
60
