Eur. J. Biochem. IY8,25-30 (1991) C FEBS 1991 0014295691 003051
Isolation and characterization of recombinant human casein kinase I1 subunits a and /l from bacteria Nikodem GRANKOWSKI I , Brigitte BOLDYREFF’ and Olaf-Georg ISSINGER’ Department of Molecular Biology, Institute of Microbiology, Maria Curie-Sklodowska University, Lublin, Poland
’ Institut fur Humangenetik, Universitat des Saarlandes, Homburg, Federal Republic of Germany (Received January 7, 1991)
-
EJB 91 0054
cDNA encoding the casein kinase I1 (CKII) subunits a and /3 of human origin were expressed in Esclwriclzia coli using expression vector pT7-7. Significant expression was obtained with E. coli BL21(DE3). The CKII subunits accounted for approximately 30% of the bacterial protein; however, most of the expressed proteins were produced in an insoluble form. The recombinant CKII M subunit was purified by DEAE-cellulose chromatography, followed by phosphoceliulose and heparin-agarose chromatography. The recombinant CKII p subunit was extracted from the insoluble pellet and purified in a single step on phosphocellulose. From 10 g bacterial cells, the yield of soluble protein was 12 mg a subunit and 5 mg /3 subunit. SDSjPAGE analysis of the purified recombinant proteins indicated molecular masses of 42 kDa and 26 kDa for the a and p subunits, respectively, in agreement with the molecular masses determined for the subunits of the native enzyme. The recombinant c( subunit exhibited protein kinase activity which was greatest in the absence of monovalent ions. With increasing amounts of salt, a subunit kinase activity declined rapidly. Addition of the p subunit led to maximum stimulation at a 1 : 1 ratio of both subunits. Using a synthetic peptide (RRRDDDSDDD) as a substrate, the maximum protein kinase stimulation observed was fourfold under the conditions used. The K, of the reconstituted enzyme for the synthetic peptide (80 pM) was comparable to the mammalian enzyme (40-60 pM), whereas the a subunit alone had a K,,, of 240 pM. After sucrose density gradient analysis. the reconstituted holoenzyme sedimented at the same position as the mammalian CKII holoenzyme.
Casein kinase I1 (CKII) is a cyclic-nucleotide-independent serine/threonine protein kinase which has been identified in the cytoplasm and the nucleolus of numerous mammalian organisms and cell types (for review, see [l -41). Its ubiquitous distribution together with its conserved sequence imply an important function in the metabolism of eukaryotes. CKII is unique among the protein kinases in being able to use ATP as well as GTP as a phosphoryl donor. It is a protein kinase which preferentially phosphorylates serine and threonine residues surrounded by a cluster of acidic amino acid residues (XSXXEX) [5]. It has a heterotetrameric structure, either a2P2or aa‘pz. The a subunit contains the ATP-binding site and therefore represents the catalytic subunit of the enzyme [6, 71. The molecular mass of the holoenzyme is about 130 kDa (sZ0,,,= 6) [8]. The a and a’ subunits have molecular masses of 44 kDa and 42 kDa, respectively, whereas the /3 subunit has a molecular mass of 26 kDa (for review, see [l, 21). CKII is a pleiotropic enzyme which has been shown to phosphorylate numerous substrates in vitro and in vivo (for a complete list of all known substrates see [2]). However, recent data suggest a role of CKII during the processes of proliferation [9- 131 and differentiation [14]. Correspondence to 0 . - G . Issinger, Institut fur Humangenetik, Universitat des Saarlandes, W-6650 Homburg 3, Federal Republic of Germany Abbreviations. CKII, casein kinase I1 ; PCR, polymerase chain reaction.
cDNA from the CKII a and subunits from various organisms have been isolated and sequenced [15- 181. The a subunit displays the expected homology to the catalytic domain of other protein kinases, confirming its identification as the catalytic subunit. The fl subunit exhibits no marked similarity to the other sequenced proteins, but contains possible CKII recognition sites, one or more of which causes autophosphorylation of this subunit. CAMP-dependent protein kinase is dissociated into two subunits upon reaction with a specific modulator (CAMP). Protein kinase C or cGMP-dependent protein kinase are monomeric kinase molecules with different domains, e.g. binding and kinase domains on the same molecule. CKII is unique that it consists of two or three different subunits (a2p2, aa’p2) which form a complex that cannot be dissociated readily except under denaturing conditions; yet they are not covalently linked to form a single molecule with different domains, e.g. cGMP kinase and protein kinase C. So far, there is no modulator known which would dissociate a and p subunits as is the case for the R and C subunits from A-kinase which are dissociated by CAMP. This situation is unique among the known protein kinases, therefore the purpose of the p subunit is an interesting question. Are there external modulators which may regulate CKII activity? Some data suggest an epidermal-growth-factor-mediated pathway involving an unknown serine protein kinase which leads to phosphorylation of the p subunit [19]. The function of the p subunit is still a matter of speculation and is difficult to address
26 because the holoenzyme can only be separated into its subunits under denaturing conditions. Despite the denaturing conditions, the subunits were shown to reconstitute into a functional enzyme after considerable renaturation time, yet the yield was only 10% of the starting material [20]. The observed low yield of CKII activity seriously raises the question whether the native state of the enzyme was indeed fully retained. One way of circumventing this problem might be through the separate expression of recombinant subunits for CKII. Our efforts to characterize recombinant CKII a and /3 subunits were therefore directed toward characterizing the roles of both subunits separately. It is shown here that the CKII a subunit by itself contains a basal activity, but this is increased severalfold by the addition of equimolar amounts of the /3 subunit.
MATERIALS AND METHODS Cloning of the coding regionsfrom human CKII a and /3 subunits Total RNA from the human glioblastoma cell line HeRoSV was isolated according to [21]. Specific single-stranded cDNA were synthesized using primers complementary to the last 20 nucleotides of the coding regions of the human CKII a (AAGGATCCAAGCTTACTGCTGAGCGCCAGCGGC) 01 fi (AAGGATCCAAGCTTCAGCGAATCGTCTTGACTGG) subunits [16, 171. The primers contain at their 5' end additional nucleotides, i.e. restriction enzyme recognition sequences. The reaction was carried out with 5 pg RNA, 25 units (U) RNase inhibitor (Boehringer, Mannheim), 50 pmol primer, and 200 U murine Moloney leukaemia virus reverse transcriptase (BRL, Eggenstein) in a buffer containing 50 mM KC1, 10mM Tris/Cl, pH 8.3, 4 mM MgC12, 1 mM dithiothreitol and 1 mM dNTP. The reaction volume was 20 p1 and incubation was at 37°C for 1 h. 2 p1 of this reaction was subjected to the polymerase chain reaction (PCR) using the primers mentioned above and primers comprising the first 20 nucleotides of the coding region of CKII a (AATCTAGACATATGTCGGGACCCGTGCCAAGC) or P (CCTCTAGACATATGAGCAGCTCAGAGGAGGT) subunits. These primers contain an Ndel recognition sequence at their 5'end which covers the first ATG codon. Reaction was carried out in a final volume of 100 p1 containing 50 mM KC1,lO mM Tris/Cl, pH 8.3, 2.5 mM MgCI2, 200 nM dNTP, 50 pmol of each primer and 2.5 U Taq polymerase (BRL, Eggenstein). 30 cycles (1 min, 94°C; 1 min, 50°C; 3 min, 72°C) were run. In the first cycle, the 94 C incubation was for 5 min; in the last cycle the 72 'C incubation was for 10 min. PCR products were analyzed on a 1% agarose gel. Bands of the correct size (a subunit, 1.2 kb; j? subunit, 0.65 kb) were cut out of the gel and the DNA was isolated from the gel piece according to [22]. 200 ng NdeIIHindIII-digested PCR product was ligated with 100 ng NdeIIHindI11-digested pT7-7 vector. Ligation was carried out according to [23]. The pT7-7 vector was kindly provided by S. Tabor [24]. After transformation of the bacterial strain DH1, plasmids were isolated. The sequences of several cloned PCR products were determined using the T7 sequencing kit (LKB-Pharmacia, Freiburg) and sequencing primers according to the published sequences of human CKII a and subunits. The primers were synthesized on a Gene Assembler Plus DNA synthesizer from Pharmacia. Plasmids with the correct insert sequence were termed pBB-3 ( a subunit) and pBB-4 (fi subunit).
Expression of CKII a and P subunits pBB-3 and pBB-4 were used to transform the bacterial expression strain BL21 (DE3), provided by F. W. Studier [25]. An overnight culture was inoculated with a single colony, and for expression, 2.5 ml overnight culture was added to 500 ml growth medium (10 g/1 tryptone, 5 g/1 yeast extract, 5 g/l NaCl and 75 pg/ml ampicillin). The culture was incubated at 37°C with vigorous shaking until an A,,, of 0.5 was reached, when 50 mM isopropyl-thio-P-D-galactoside was added. After 4 h, bacteria were harvested (3000 x g, 10 min, 4"C), washed twice with 0.9% NaCl and 0.2 mM phenylmethylsulfonyl fluoride, then frozen. Purijication qf the CKII a subunit 10.2 g E. coli corresponding to 6 1 bacterial culture were suspended in 160 m1 20 mM Tris/HCl, pH 7.5,l mM EDTA, 0.5 mM EGTA, 7 mM 2-mercaptoethanol, 50 mM NaCl, 0.2 mM phenylmethylsulfonyl fluoride, 0.5 pg/ml leupeptin and 0.7 pg/ml pepstatin (buffer A). Bacteria were disrupted by sonication (five 1-min cycles, with 5-min intervals on ice). Quantitative disruption of the bacteria was verified by phasecontrast microscopy. After sonication, the bacterial extract was treated with DNase (2 pg/ml) for 60 min at 4°C and centrifuged for 10 min at 30000 x g. The resulting supernatant and pellet are referred to as S-I and P-I fractions, respectively. The P-I fraction was discarded. Alternatively, the cells were disrupted using a French press. The S-I fraction was centrifuged at 150000 x g for 3 h to remove ribosomes. The obtained postribosomal supernatant, S-11, was used for the purification of the CKII a subunit (Table 1). The ribosomal pellet was discarded. The S-I1 fraction, 594mg protein, was loaded on to DEAE-cellulose (70 ml). A linear gradient of 2 x 250 ml50 - 600 mM NaCl in buffer A was applied. Active fractions (150- 300 mM NaC1) containing 435 mg protein were applied to a PI 1 phosphocellulose column (20 ml), equilibrated with 250 mM NaCl in buffer A. A linear gradient (2 x 3 50 ml) ranging over 250 1200 mM NaCl was applied. The enzyme activity eluted at approximately 700 mM NaCl. Active fractions containing 46 mg protein were diluted with buffer A to a final salt concentration of 500 mM NaCl and were then loaded on to heparinagarose (4 ml). The column was washed with 10 vol. 500 mM NaCl in buffer A. A linear gradient ranging over 5001000 mM (2 x 100 ml) was applied. The kinase activity eluted at approximately 600 mM salt. Active fractions were concentrated by Amicon ultrafiltration using a YM-30 membrane. The total yield of a subunit preparation was 12 mg. PAGE analysis shows that the three-step purification procedure led to an apparently homogenous CI subunit. The protein was stored in small aliquots at - 70 "C. Repeated freezing and thawing did not have any damaging effect on a subunit kinase activity. Purifccation ojthe CKII P subunit The sonication procedure was carried out as already described for the purification of the CKII a subunit. After centrifugation of the bacterial extract for 10 min at 30000 xg, the supernatant (S-I) was discarded and the pellet (P-I) was used for the isolation of the CKII p subunit. The extraction of the CKII /3 subunit from the pellet was performed with 100 mM NaCl in buffer A. After overnight extraction and stirring in the cold room, the suspension was centrifuged for
27
A
kDa
- 96 - 68 - 46
,-30
- 21
a
b
C
d
a
b
C
d
Fig. 1. Extraction qf CKII SI ( A ) and [j subunits (B),from the bacterial extracts. (A) CKII SI subunit: lane a, crude bacterial extract; lane b, PI fraction; lane c, S-I fraction; lane d, postribosomal supernatant (S-I1 fraction). (B) CKII J-subunit: lane a, crude bacterial extract; lane b, S-I fraction; lane c, P-I fraction; lane d, S-I1 fraction
10 min at 30000 x g. The resulting supernatant (S-11, 15 mg) was chromatographed on phosphocellulose (5 ml) which was equilibrated with 100 mM NaCl in buffer A. CKII p subunit eluted from the column at 1 M NaCl in buffer A as a homogenous protein. The yield of soluble CKII p subunit was about 5 mg from 10 g bacterial cells. Gel electrophoresis 12.5% SDSjPAGE was essentially as described by [26]. Gels were stained with 0.1 YOCoomassie brilliant blue R-250 in 45% methanol and 10% acetic acid, then destained in 45% methanol and 10% acetic acid. The insoluble material obtained after centrifugation of the crude bacterial extracts could not be used for protein determination. Consequently, it was not possible to apply equal amounts of material in terms of protein on to the gels shown in Fig. 1. However, in order to give an impression of the purification progress, we always adjusted the volumes of the various fractions obtained to that of the original starting material (i.e. 160 ml). 5-pl aliquots were removed and after mixing with an equal volume of SDS sample buffer were analysed on one-dimensional SDS/PAGE. Casein kinase II CKII from rat liver was a gift from Dr L. Pinna, Padua. CKII from Krebs I1 mouse ascites tumour cells was purified following a procedure previously described [26] with slight modifications. Preparation of polyclonal antibodies Antiserd directed against the holoenzyme from Krebs I1 mouse ascites tumour cells were prepared as described by Miinstermann et al. [13]. Antibodies to each of the subunits of CKII from human were prepared as follows. Female Chinchilla rabbits were injected intradermally with 500 pg purified polypeptide in an emulsion with an equal volume of complete Freund’s adjuvant. The rabbits were boosted with 500 pg protein emulsified with an equal volume of complete Freund’s adjuvant after 4 weeks and bled after 8 weeks. Protein transfer Protein was transferred after SDSjPAGE to nitrocellulose paper and reacted with antibodies and phosphatase-labelled goat anti-(rabbit IgG) antibody as described earlier [27].
Protein kinase assay CKII was assayed at 37°C in a reaction mixture containing 20 mM Mes, pH 6.9, 130 mM KCl, 10 mM MgCI2, 4.8 mM dithioerythritol, 50 pM ATP, [ Y - ~ ~ P I A T (specific P activity, 100 cpm/pmol) and 0.32 mM synthetic peptide substrate (RRRDDDSDDD) in a total volume of 50 pl. Reactions were started by the addition of 20 p1 column fraction or pure enzyme diluted into a 2 0 4 volume and terminated after 15 min by spotting 30 p1 of the reaction mixture on to P-81 phosphocellulose paper as described previously [28, 291. 1 U CKII activity is defined as the amount of activity required to transfer 1 pmol phosphate/min into substrate at 37°C. Determination of K,
K, of the CI subunit and the reconstituted enzyme for the synthetic peptide were determined from Lineweaver-Burke plots by linear-regression analysis. Data reflect the average of three assays for each concentration. Protein determination Protein concentration was estimated by the method of Bradford [30], using bovine serum albumin as the standard. Sucrose density gradient analysis CKII holoenzyme and CI and p subunits were analysed by centrifugation on a 15 - 35% (mass/vol.) sucrose gradient in 10 mM Tris/HCl, pH 7.5, 5 mM MgC12, 300 mNaC1. Centrifugation was carried out in a Kontron TST 60.4 rotor at 40000 rpm for 22 h at 4°C. 0.2-ml fractions were collected. 20p1 aliquots were assayed for kinase activity using the synthetic peptide as substrate. The remainder of the fraction was precipitated by the addition of an equal volume 20% trichloroacetic acid. 50 pg lysozyme was added as a carrier. The precipitated material was characterized by 12.5% SDS/ PAGE analysis and immunostaining. RESULTS AND DISCUSSION Cloning and expression of CKII subunits The coding region of the human CKII subunits were amplified from specific single-stranded cDNA and ligated into
28 CKII blot
Y
P
a stain
blot
--
stain
blot
kDa
Table 1. Purificution of recombinant C H I x subunit 1 U = 1 nmol phosphate transferred to the synthetic peptide/min at 37'C Step
Volume
Protein
Total activity
Specific activity
ml
mg
U
U/mg protein
160 290 414 1.5
594 435 46 12
188 2588 2668 850
0.3 5.9 58 70.8
- 46 s-I1 DEAE-cellulose P11 Heparin-agarose
Fig. 2. Purified CKII x and /3 subunits. 3 pg purified CKII c( subunit and 4 v g purified C K l l p subunit were analyzed on 12.5% PAGE by staining and immunoblotting. CKII c( subunit was incubated with an anti-(CK11 x subunit) antibody. CKII p subunit was incubated with an anti-(CKII p subunit) antibody. For comparison, an immunoblot with 4 pg CKII holocnzyme from Krebs I1 mouse ascites tumour cells is included. Detection of the subunits was carried out using an antibody against the CKII holoenzyme
the Ndel- Hind111 restriction site of pT7-7. An advantage of this procedure is the very easy creation of suitable restriction sites for cloning. Therefore no time-intensive site-directed mutagenesis is necessary. A disadvantage is the fact that one has to confirm the whole sequence of the cloned PCR products, since Taq polymerase used for amplification is known for misreading while synthesising D N A strands. In our case, we sequenced two clones encoding the a subunit and two clones encoding the j subunit. One of each pair of a and p clones contained at least one wrong nucleotide which led to changes in the amino acid sequences, the other ones (pBB-3 and pBB4) contained the correct sequences. In the pBB-3 and pBB-4 vectors, the CKII a and fl subunits, respectively, are under the control of the T7 promoter. In the expression strain BL21(DE3), the T7 RNA polymerase is produced after induction with isopropyl-thio-P-D-thiogalactoside, then leads to expression of CKII subunits. In total cell extracts, new proteins of 42 kDa and 26 kDa were observed when the bacteria were transformed with pBB-3 and pBB-4, respectively. These proteins were not present in cells transformed with the pT7-7 vector alone (not shown). The expressed proteins represented up to 30% of the total bacterial protein. Purification qf the CKII suhunits,from E. coli extracts The purification of the CKII subunits was described in detail in Materials and Methods. For the isolation of the M subunit, the bacterial cell extract (Fig. 1A, lane a) was centrifuged to remove bacterial debris. The resulting pellet (Fig. 1A, lane b) contained the bulk of the expressed a subunit, yet in an insoluble form and devoid of activity. The protein pattern of the corresponding supernatant (S-I) is shown in Fig. 1 A (lane c). This fraction contained less, but active, a subunit. Therefore, the S-I fraction was used in the course of the further purification of the M subunit. After removal of bacterial ribosomes, the postribosomal supernatant, shown in Fig. 1A (lane d), was subjected to chromatographic analysis similar to the published procedures for the isolation of the holoenzyme (results not shown). After the last purification step, affinity chromatography on heparin-agarose, the a subunit was virtually homogenous (Fig. 2). The whole purification scheme is shown in Table 1. The unusually high recovery of protein (i.e. 435 mg from 594 mg starting material; S-11) after the DEAEcellulose step can be partly explained by the broad elution
pattern obtained. Consequently, about two-thirds of the fractions were collected. However, CKII-a-subunit-suppressing factors must be removed by this chromatography step leading to the observed 20-fold increase in specific activity. The subsequent phosphocellulose step showed a narrow CKII CI subunit elution pattern which allowed collection of a small number of fractions. About 10% (i.e. 46 mg) of the original amount of protein applied was recovered. Specific activity increased about tenfold. After the phosphocellulose step, CKII a subunit was about 80 - 90% pure. The final chromatography step on heparin-agarose led to a homogenous CKII M subunit preparation, as judged by PAGE analysis. Although a significant portion of the expressed protein remained insoluble, up to 12 mg M subunit/lO g bacteria was soluble and enzymatically active. The fl subunit was isolated from the pellet in a one-step purification scheme, usually in less than 24 h. Although most of the P subunit was insoluble, we could extract 15 mg soluble protein by salt treatment for the subsequent phosphocellulose chromatography (Fig. 1 B). After this purification step, the P subunit was virtually homogenous (Fig. 2). Here we obtained 5 mg P subunit from 10 g bacteria. The yield of active a and /3 subunits could not be increased considerably by disintegration of the bacterial cells with a French press instead of using sonication and by carrying out induction at 30°C instead of 37°C [31]. The yield of CKII a subunit from the nematode Candida elegans which was expressed in almost the same expression system as the one used here, was only 10% of that obtained for the human enzyme [32]. A reason for the lower yield of the recombinant a subunit from the nematode [32] may be partly due to the different lysis protocol, which involved osmolytic lysis. The purity of CKII-a subunit was established by Coomassie blue staining and immunodetection using two different CKII-specific anti-CKII antibodies. One of the two antibodies was directed against the holoenzyme from Krebs I1 mouse ascites tumor cells; the other antibody was directed against the purified CKII a subunit. The purity of the CKII fl subunit was verified by Coomassie blue staining and immunodetection using an anti-(CKII [j subunit) antibody (Fig. 2). Injluence oj'salt concentration on CKII activity The influence of increasing amounts of KCl on the a subunit and reconstituted CKII ( a and P subunits) were tested (Fig. 3). Reconstitution was complete within the time needed to complete an enzyme assay. There was a little difference between the results obtained when NaCl was used instead of KC1 (results not shown). Since we d o not know for sure whether all subunits (here we used 7 p m o l a subunit and 7 pmol J/' subunit) d o indeed associate quantitatively to form
29
11 12 13 14 15 0
0 0 0 0 0 0 0 m g ~Co o o o U m T t i n
o m o o ~o o ~ o oo oo
N m P m
o
o
KCI (rnM)
ADH
I.
Fig. 3. Eflect of NaCl and KC1 on the activity of the CKII a subunit. The effect of different concentrations of KCI (0-500 mM) on the activity of the z subunit (66 pmol) was carried out in two sets of experiments. The first set was conducted with the a subunit alone (E!). The second one was supplemented with CKII subunit at a 1 : 1 molar ratio (8)
6o
I Bottom
5
15
10
TOP
Fraction number
Ratio a/P (rnol 'rnol-')
Fig. 4. Actiwtion of the CKII a subunit by CKII J/' subunit. Activity test of the CKII a subunit was performed under standard assay conditions as described under Materials and Methods using a constant amount of CKII z subunit (36 pmol) and increasing amounts of j subunit (7.8 pmol-93.6 pmol)
an a2P2 complex, we have related the activity values to the amount of protein subunit in the assay instead of relating it to the number of subunits used. In the absence of salt, the LX subunit kinase activity was about three times higher (specific activity, 154 U/mg protein) than at 300 m M KCI. The reconstituted recombinant human enzyme showed the opposite effect. Here, the activity was lowest in the absence of salt. The observed specific activity was highest at 300 m M KCI (155 U/mg protein). This observation is consistent with earlier observations with the native holoenzyme from different mammalian sources, which exhibit maximum activities at 200-300 m M salt [l]. Recombinant CKII tl subunit from C. &guns did not show such a n opposite salt effect when compared to the native holoenzyme. Here the x subunit and the holoenzyme were most active at 100 m M NaCl[32]. One of the reasons for this different behaviour may be due to the fact the ci subunit from C. elegans is 22 amino acids shorter than the human counterpart, and what seems to be even more crucial is that the last 27 carboxy-terminal amino
Fig. 5. Sucrose density gradient analysis of CKII. CKII a subunit (1 1 pg) and CKII p subunit (4.5 pg) were in 100 pI(l0 mM Tris/HCI, pH 7.5, 5 mM MgClz and 300 mM KCI) and analyzed by sucrose density gradient centrifugation. In separate tubes, CKII z subunit ( a ) , CKII p subunit (p), reconstituted CKII holoenzyme (a + p), CKII holoenzyme (H) from Krebs I1 mouse ascites tumour cells and alcohol dehydrogenase (ADH, 150 kDa) were subjected to sucrose density gradient analysis. The positions of these proteins are indicated by arrows. In all cases, the position was either established by immunostaining using a specific anti-(CKII p subunit) antibody or by both immunostaining and activity measurement (CKII a subunit, reconstituted CKII and CKII holoenzyme from Krebs I1 mouse ascites tumour cells). Alcohol dehydrogenase was identified by Coomassie blue staining. Activity measurements of rcconstitutcd CKII are shown here. The inset shows the corresponding immunostaining of fractions 1 1 - 15
acids differ completely from the human 331.
tl
subunit kinase [16.
Reconstitution of the holoenzyme The isolation of the tl and p subunits from CKII made it possible to provide the first answer to the question whether the p subunit is a vital component of the holoenzyme; either influencing the activity of the catalytic tl subunit or being responsible for the specificity of the holoenzyme, or maybe both. Previously, the question could not be satisfactorily answered because the ci and p subunits of the holoenzyme can only be separated under denaturing conditions and activity only partially (10%) restored [20]. In order to determine the optimum subunit composition which would yield maximum enzyme activity, we designed an experiment where increasing amounts of p subunit were added to a constant amount of the tl subunit. With increasing amounts of p subunit added, a stimulation of the kinase activity of the tl subunit was obtained which was highest at a 1 : l clip subunit ratio. Additional
amounts of B subunit did not lead to a further increase in CKII activity. The stimulation was highest (fourfold) in the presence of 300 mM KCl (Fig. 4), but was heavily dependent upon the salt concentration present, especially since the a subunit kinase activity is high in the absence and low in the presence of high salt (Fig. 3). In order to find out whether the reconstitution of a and jsubunits also leads to an enzyme with biophysical properties similar to the native holoenzyme, we carried out sucrose density gradient analysis. We analyzed the holoenzyme from Krebs I1 mouse ascites tumour cells and the purified a and p subunits and the reconstituted enzyme, using alcohol dehydrogenase and other proteins as reference standards. The positions of the individual proteins after centrifugation are indicated by arrows and were established by staining with Coomassie blue (protein markers) or by immunodetection with specific antibodies directed against the purified a and jsubunits (Fig. 5). An activity profile is only shown for the reconstituted kinase. The inset shows the immunostaining obtained from the reconstituted enzyme using an anti-CKII antibody. It is clear that the reconstituted enzyme is virtually identical to the native holoenzyme [6]. The CI subunit was detected around 4.5 S, where one would expect it to sediment according to its molecular mass assuming a ?lobular structure. The same is true for the position of the B \Libunit. For reasons of clarity only the position of alcohol dehydrogenase is indicated. The K, of the reconstituted enzyme was found to be 80 pM for the synthetic peptide. This is comparable to the published K, of 60 pM for the bovine holoenzyme [29]. Using rat liver CKII, we obtained a K, of 40 pM for the synthetic peptide. The recombinant a subunit alone had a K, of 240 pM. This observation is in agreement with the data shown in Fig. 5, and suggests that association with the p subunit may also alter the substrate-binding specificity of the a subunit. This work was supported by Sonderfor.sc/zungsbereich grant 2461 B3 to 0 . G . I . N.G. was supported by a stipend from SFB 246.
REFERENCES 1. Hathaway, G. M. & Traugh, J. A. (1982) Curr. Top. Cell Regul. 21, 101 -127. 2. Pinna, L. A. (1990) Biochim. Biophys. Acta 1054, 267 -284. 3. Schncider, H. R. & Issinger, 0.G.(1989) Biotechnol. Eur. 6, 8288. 4. Boldyreff, B., Schncider, H. R., Fritz, G., Traffa, K., Seitz, G. & Issinger, 0.G.(1990) in BioTechForum, advances in molecular genetics (Kompf, J., Henke, J., Driesel, A. J. eds) vol. 3, pp. 91 -98.
5. Pinna, L. A., Meggio, F., Marchiori, F. & Borin, G. (1984) FEBS Lett. 171, 211 -214. 6. Hathaway, G. M., Zoller, M. J. & Traugh, J. A. (1981) J . Biol. Chem. 256, 11442 - 11446. 7. Feige, J.-J., Cochet, C., Pirollet, F. & Chambaz, E. M. (1983) Biochemistry 22, 1452- 3459. 8. Glover, C.V. C. (1986) J . Biol. Chem. 261, 14349-14354. 9. Prowald, K., Fischer, H. & Issinger, 0 . G . (1984) FEBSLett. 176, 479-483. 10. Schneider, H. R., Reichert, G. H. & Issinger, 0.-G. (1986) Eur. J . Biochem. 181, 733-738. 11. Perez, M., Grande, J. & Itarte, E. (1988) FEBS Lett. 238, 273276. 12. Seitz, G., Miinstermann, U., Schneider, H. R. & Issinger, 0 . G . (1989) Biochem. Biophys. Res. Commun. 163, 635-641. 13. Miinstermann, U., Fritz, G., Seitz, G., Lu, Y., Schneider, H. R. & Issinger, 0.43.(1990) Eur. J . Biochem. 189, 251 -257. 14. Sommercorn, J. & Krebs, E. G. (1987) J . Biol. Chem. 262,38393843. 15. Saxena, A., Padmanabha, R. &Glover, C. V. C. (1987) Mol. Cell. Biol. 7, 3409 - 341 7. 16. Meisner, H., Heller-Harrison, R., Buxton, J. & Czech, M. P. (1989) Biochemistry 28, 4072 -4076. 17. Heller-Harrison, R. A., Meisner, H. & Czech, M. P. (1989) Biochemistry 28, 9053 -9058. 18. Jakobi, R., Voss, H. & Pyerin, W. (1989) Eur. J . Biochem. 183, 227-233. 19. Ackerman, P., Glover, C. V. C. & Osheroff, N. (1990) Proc. Nut1 Acad. Sci. U S A 87, 821 - 825. 20. Cochet, C. & Chambaz, E. M. (1983) J . Biol. Chem. 258, 14031406. 21. Reichert, G. H. & Issinger, 0.-G. (1985) Biochimie (Paris) 67, 657 - 661. 22. Tautz, D. & Renz, M. (1983) Anal. Biochem. 132,14- 19. 23. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular cloning, a laboratory manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 24. Tabor, S. & Richardson, C. C. (1985) Proc. Nail Acad. Sci.U S A 82, 1074-1078. 25. Studier, F. W. & Moffat, 9 . A. (1986) J . Mol. Biol. 189, 113130. 26. Laemmli, U. K . (1970) Nature 227, 680-685. 26. Schneider, H. R. (1988) PhD Thesis, University Kaiserslautern Germany. 27. Issinger, 0.-G., Martin, T., Richter, W. W., Olson, M. & Fujiki, H. (1988) E M B O J . 7,1621-1626. 28. Kuenzel, E. A. & Krebs, E. G. (1985) Proc. Nail Acad. Sci. USA 82, 737 - 741. 29. Kuenzel, E. A., Mulligan, J. A,, Sommercorn, J. & Krebs, E. G . (1987) J . Biol. Chem. 262, 9136-9140. 30. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254. 31. Slice, L. W. & Taylor, S. S. (1989) J . Biol. Chem. 264, 2094020946. 32. Hu, E. & Rubin, C. S. (1990) J . Biol. Chem. 265, 20609-20615. 33. Hu, E. & Rubin, C. S. (1990) J . Biol. Chem. 265, 5072-5080.
Isolation and characterization of recombinant human casein kinase I1 subunits a and /l from bacteria Nikodem GRANKOWSKI I , Brigitte BOLDYREFF’ and Olaf-Georg ISSINGER’ Department of Molecular Biology, Institute of Microbiology, Maria Curie-Sklodowska University, Lublin, Poland
’ Institut fur Humangenetik, Universitat des Saarlandes, Homburg, Federal Republic of Germany (Received January 7, 1991)
-
EJB 91 0054
cDNA encoding the casein kinase I1 (CKII) subunits a and /3 of human origin were expressed in Esclwriclzia coli using expression vector pT7-7. Significant expression was obtained with E. coli BL21(DE3). The CKII subunits accounted for approximately 30% of the bacterial protein; however, most of the expressed proteins were produced in an insoluble form. The recombinant CKII M subunit was purified by DEAE-cellulose chromatography, followed by phosphoceliulose and heparin-agarose chromatography. The recombinant CKII p subunit was extracted from the insoluble pellet and purified in a single step on phosphocellulose. From 10 g bacterial cells, the yield of soluble protein was 12 mg a subunit and 5 mg /3 subunit. SDSjPAGE analysis of the purified recombinant proteins indicated molecular masses of 42 kDa and 26 kDa for the a and p subunits, respectively, in agreement with the molecular masses determined for the subunits of the native enzyme. The recombinant c( subunit exhibited protein kinase activity which was greatest in the absence of monovalent ions. With increasing amounts of salt, a subunit kinase activity declined rapidly. Addition of the p subunit led to maximum stimulation at a 1 : 1 ratio of both subunits. Using a synthetic peptide (RRRDDDSDDD) as a substrate, the maximum protein kinase stimulation observed was fourfold under the conditions used. The K, of the reconstituted enzyme for the synthetic peptide (80 pM) was comparable to the mammalian enzyme (40-60 pM), whereas the a subunit alone had a K,,, of 240 pM. After sucrose density gradient analysis. the reconstituted holoenzyme sedimented at the same position as the mammalian CKII holoenzyme.
Casein kinase I1 (CKII) is a cyclic-nucleotide-independent serine/threonine protein kinase which has been identified in the cytoplasm and the nucleolus of numerous mammalian organisms and cell types (for review, see [l -41). Its ubiquitous distribution together with its conserved sequence imply an important function in the metabolism of eukaryotes. CKII is unique among the protein kinases in being able to use ATP as well as GTP as a phosphoryl donor. It is a protein kinase which preferentially phosphorylates serine and threonine residues surrounded by a cluster of acidic amino acid residues (XSXXEX) [5]. It has a heterotetrameric structure, either a2P2or aa‘pz. The a subunit contains the ATP-binding site and therefore represents the catalytic subunit of the enzyme [6, 71. The molecular mass of the holoenzyme is about 130 kDa (sZ0,,,= 6) [8]. The a and a’ subunits have molecular masses of 44 kDa and 42 kDa, respectively, whereas the /3 subunit has a molecular mass of 26 kDa (for review, see [l, 21). CKII is a pleiotropic enzyme which has been shown to phosphorylate numerous substrates in vitro and in vivo (for a complete list of all known substrates see [2]). However, recent data suggest a role of CKII during the processes of proliferation [9- 131 and differentiation [14]. Correspondence to 0 . - G . Issinger, Institut fur Humangenetik, Universitat des Saarlandes, W-6650 Homburg 3, Federal Republic of Germany Abbreviations. CKII, casein kinase I1 ; PCR, polymerase chain reaction.
cDNA from the CKII a and subunits from various organisms have been isolated and sequenced [15- 181. The a subunit displays the expected homology to the catalytic domain of other protein kinases, confirming its identification as the catalytic subunit. The fl subunit exhibits no marked similarity to the other sequenced proteins, but contains possible CKII recognition sites, one or more of which causes autophosphorylation of this subunit. CAMP-dependent protein kinase is dissociated into two subunits upon reaction with a specific modulator (CAMP). Protein kinase C or cGMP-dependent protein kinase are monomeric kinase molecules with different domains, e.g. binding and kinase domains on the same molecule. CKII is unique that it consists of two or three different subunits (a2p2, aa’p2) which form a complex that cannot be dissociated readily except under denaturing conditions; yet they are not covalently linked to form a single molecule with different domains, e.g. cGMP kinase and protein kinase C. So far, there is no modulator known which would dissociate a and p subunits as is the case for the R and C subunits from A-kinase which are dissociated by CAMP. This situation is unique among the known protein kinases, therefore the purpose of the p subunit is an interesting question. Are there external modulators which may regulate CKII activity? Some data suggest an epidermal-growth-factor-mediated pathway involving an unknown serine protein kinase which leads to phosphorylation of the p subunit [19]. The function of the p subunit is still a matter of speculation and is difficult to address
26 because the holoenzyme can only be separated into its subunits under denaturing conditions. Despite the denaturing conditions, the subunits were shown to reconstitute into a functional enzyme after considerable renaturation time, yet the yield was only 10% of the starting material [20]. The observed low yield of CKII activity seriously raises the question whether the native state of the enzyme was indeed fully retained. One way of circumventing this problem might be through the separate expression of recombinant subunits for CKII. Our efforts to characterize recombinant CKII a and /3 subunits were therefore directed toward characterizing the roles of both subunits separately. It is shown here that the CKII a subunit by itself contains a basal activity, but this is increased severalfold by the addition of equimolar amounts of the /3 subunit.
MATERIALS AND METHODS Cloning of the coding regionsfrom human CKII a and /3 subunits Total RNA from the human glioblastoma cell line HeRoSV was isolated according to [21]. Specific single-stranded cDNA were synthesized using primers complementary to the last 20 nucleotides of the coding regions of the human CKII a (AAGGATCCAAGCTTACTGCTGAGCGCCAGCGGC) 01 fi (AAGGATCCAAGCTTCAGCGAATCGTCTTGACTGG) subunits [16, 171. The primers contain at their 5' end additional nucleotides, i.e. restriction enzyme recognition sequences. The reaction was carried out with 5 pg RNA, 25 units (U) RNase inhibitor (Boehringer, Mannheim), 50 pmol primer, and 200 U murine Moloney leukaemia virus reverse transcriptase (BRL, Eggenstein) in a buffer containing 50 mM KC1, 10mM Tris/Cl, pH 8.3, 4 mM MgC12, 1 mM dithiothreitol and 1 mM dNTP. The reaction volume was 20 p1 and incubation was at 37°C for 1 h. 2 p1 of this reaction was subjected to the polymerase chain reaction (PCR) using the primers mentioned above and primers comprising the first 20 nucleotides of the coding region of CKII a (AATCTAGACATATGTCGGGACCCGTGCCAAGC) or P (CCTCTAGACATATGAGCAGCTCAGAGGAGGT) subunits. These primers contain an Ndel recognition sequence at their 5'end which covers the first ATG codon. Reaction was carried out in a final volume of 100 p1 containing 50 mM KC1,lO mM Tris/Cl, pH 8.3, 2.5 mM MgCI2, 200 nM dNTP, 50 pmol of each primer and 2.5 U Taq polymerase (BRL, Eggenstein). 30 cycles (1 min, 94°C; 1 min, 50°C; 3 min, 72°C) were run. In the first cycle, the 94 C incubation was for 5 min; in the last cycle the 72 'C incubation was for 10 min. PCR products were analyzed on a 1% agarose gel. Bands of the correct size (a subunit, 1.2 kb; j? subunit, 0.65 kb) were cut out of the gel and the DNA was isolated from the gel piece according to [22]. 200 ng NdeIIHindIII-digested PCR product was ligated with 100 ng NdeIIHindI11-digested pT7-7 vector. Ligation was carried out according to [23]. The pT7-7 vector was kindly provided by S. Tabor [24]. After transformation of the bacterial strain DH1, plasmids were isolated. The sequences of several cloned PCR products were determined using the T7 sequencing kit (LKB-Pharmacia, Freiburg) and sequencing primers according to the published sequences of human CKII a and subunits. The primers were synthesized on a Gene Assembler Plus DNA synthesizer from Pharmacia. Plasmids with the correct insert sequence were termed pBB-3 ( a subunit) and pBB-4 (fi subunit).
Expression of CKII a and P subunits pBB-3 and pBB-4 were used to transform the bacterial expression strain BL21 (DE3), provided by F. W. Studier [25]. An overnight culture was inoculated with a single colony, and for expression, 2.5 ml overnight culture was added to 500 ml growth medium (10 g/1 tryptone, 5 g/1 yeast extract, 5 g/l NaCl and 75 pg/ml ampicillin). The culture was incubated at 37°C with vigorous shaking until an A,,, of 0.5 was reached, when 50 mM isopropyl-thio-P-D-galactoside was added. After 4 h, bacteria were harvested (3000 x g, 10 min, 4"C), washed twice with 0.9% NaCl and 0.2 mM phenylmethylsulfonyl fluoride, then frozen. Purijication qf the CKII a subunit 10.2 g E. coli corresponding to 6 1 bacterial culture were suspended in 160 m1 20 mM Tris/HCl, pH 7.5,l mM EDTA, 0.5 mM EGTA, 7 mM 2-mercaptoethanol, 50 mM NaCl, 0.2 mM phenylmethylsulfonyl fluoride, 0.5 pg/ml leupeptin and 0.7 pg/ml pepstatin (buffer A). Bacteria were disrupted by sonication (five 1-min cycles, with 5-min intervals on ice). Quantitative disruption of the bacteria was verified by phasecontrast microscopy. After sonication, the bacterial extract was treated with DNase (2 pg/ml) for 60 min at 4°C and centrifuged for 10 min at 30000 x g. The resulting supernatant and pellet are referred to as S-I and P-I fractions, respectively. The P-I fraction was discarded. Alternatively, the cells were disrupted using a French press. The S-I fraction was centrifuged at 150000 x g for 3 h to remove ribosomes. The obtained postribosomal supernatant, S-11, was used for the purification of the CKII a subunit (Table 1). The ribosomal pellet was discarded. The S-I1 fraction, 594mg protein, was loaded on to DEAE-cellulose (70 ml). A linear gradient of 2 x 250 ml50 - 600 mM NaCl in buffer A was applied. Active fractions (150- 300 mM NaC1) containing 435 mg protein were applied to a PI 1 phosphocellulose column (20 ml), equilibrated with 250 mM NaCl in buffer A. A linear gradient (2 x 3 50 ml) ranging over 250 1200 mM NaCl was applied. The enzyme activity eluted at approximately 700 mM NaCl. Active fractions containing 46 mg protein were diluted with buffer A to a final salt concentration of 500 mM NaCl and were then loaded on to heparinagarose (4 ml). The column was washed with 10 vol. 500 mM NaCl in buffer A. A linear gradient ranging over 5001000 mM (2 x 100 ml) was applied. The kinase activity eluted at approximately 600 mM salt. Active fractions were concentrated by Amicon ultrafiltration using a YM-30 membrane. The total yield of a subunit preparation was 12 mg. PAGE analysis shows that the three-step purification procedure led to an apparently homogenous CI subunit. The protein was stored in small aliquots at - 70 "C. Repeated freezing and thawing did not have any damaging effect on a subunit kinase activity. Purifccation ojthe CKII P subunit The sonication procedure was carried out as already described for the purification of the CKII a subunit. After centrifugation of the bacterial extract for 10 min at 30000 xg, the supernatant (S-I) was discarded and the pellet (P-I) was used for the isolation of the CKII p subunit. The extraction of the CKII /3 subunit from the pellet was performed with 100 mM NaCl in buffer A. After overnight extraction and stirring in the cold room, the suspension was centrifuged for
27
A
kDa
- 96 - 68 - 46
,-30
- 21
a
b
C
d
a
b
C
d
Fig. 1. Extraction qf CKII SI ( A ) and [j subunits (B),from the bacterial extracts. (A) CKII SI subunit: lane a, crude bacterial extract; lane b, PI fraction; lane c, S-I fraction; lane d, postribosomal supernatant (S-I1 fraction). (B) CKII J-subunit: lane a, crude bacterial extract; lane b, S-I fraction; lane c, P-I fraction; lane d, S-I1 fraction
10 min at 30000 x g. The resulting supernatant (S-11, 15 mg) was chromatographed on phosphocellulose (5 ml) which was equilibrated with 100 mM NaCl in buffer A. CKII p subunit eluted from the column at 1 M NaCl in buffer A as a homogenous protein. The yield of soluble CKII p subunit was about 5 mg from 10 g bacterial cells. Gel electrophoresis 12.5% SDSjPAGE was essentially as described by [26]. Gels were stained with 0.1 YOCoomassie brilliant blue R-250 in 45% methanol and 10% acetic acid, then destained in 45% methanol and 10% acetic acid. The insoluble material obtained after centrifugation of the crude bacterial extracts could not be used for protein determination. Consequently, it was not possible to apply equal amounts of material in terms of protein on to the gels shown in Fig. 1. However, in order to give an impression of the purification progress, we always adjusted the volumes of the various fractions obtained to that of the original starting material (i.e. 160 ml). 5-pl aliquots were removed and after mixing with an equal volume of SDS sample buffer were analysed on one-dimensional SDS/PAGE. Casein kinase II CKII from rat liver was a gift from Dr L. Pinna, Padua. CKII from Krebs I1 mouse ascites tumour cells was purified following a procedure previously described [26] with slight modifications. Preparation of polyclonal antibodies Antiserd directed against the holoenzyme from Krebs I1 mouse ascites tumour cells were prepared as described by Miinstermann et al. [13]. Antibodies to each of the subunits of CKII from human were prepared as follows. Female Chinchilla rabbits were injected intradermally with 500 pg purified polypeptide in an emulsion with an equal volume of complete Freund’s adjuvant. The rabbits were boosted with 500 pg protein emulsified with an equal volume of complete Freund’s adjuvant after 4 weeks and bled after 8 weeks. Protein transfer Protein was transferred after SDSjPAGE to nitrocellulose paper and reacted with antibodies and phosphatase-labelled goat anti-(rabbit IgG) antibody as described earlier [27].
Protein kinase assay CKII was assayed at 37°C in a reaction mixture containing 20 mM Mes, pH 6.9, 130 mM KCl, 10 mM MgCI2, 4.8 mM dithioerythritol, 50 pM ATP, [ Y - ~ ~ P I A T (specific P activity, 100 cpm/pmol) and 0.32 mM synthetic peptide substrate (RRRDDDSDDD) in a total volume of 50 pl. Reactions were started by the addition of 20 p1 column fraction or pure enzyme diluted into a 2 0 4 volume and terminated after 15 min by spotting 30 p1 of the reaction mixture on to P-81 phosphocellulose paper as described previously [28, 291. 1 U CKII activity is defined as the amount of activity required to transfer 1 pmol phosphate/min into substrate at 37°C. Determination of K,
K, of the CI subunit and the reconstituted enzyme for the synthetic peptide were determined from Lineweaver-Burke plots by linear-regression analysis. Data reflect the average of three assays for each concentration. Protein determination Protein concentration was estimated by the method of Bradford [30], using bovine serum albumin as the standard. Sucrose density gradient analysis CKII holoenzyme and CI and p subunits were analysed by centrifugation on a 15 - 35% (mass/vol.) sucrose gradient in 10 mM Tris/HCl, pH 7.5, 5 mM MgC12, 300 mNaC1. Centrifugation was carried out in a Kontron TST 60.4 rotor at 40000 rpm for 22 h at 4°C. 0.2-ml fractions were collected. 20p1 aliquots were assayed for kinase activity using the synthetic peptide as substrate. The remainder of the fraction was precipitated by the addition of an equal volume 20% trichloroacetic acid. 50 pg lysozyme was added as a carrier. The precipitated material was characterized by 12.5% SDS/ PAGE analysis and immunostaining. RESULTS AND DISCUSSION Cloning and expression of CKII subunits The coding region of the human CKII subunits were amplified from specific single-stranded cDNA and ligated into
28 CKII blot
Y
P
a stain
blot
--
stain
blot
kDa
Table 1. Purificution of recombinant C H I x subunit 1 U = 1 nmol phosphate transferred to the synthetic peptide/min at 37'C Step
Volume
Protein
Total activity
Specific activity
ml
mg
U
U/mg protein
160 290 414 1.5
594 435 46 12
188 2588 2668 850
0.3 5.9 58 70.8
- 46 s-I1 DEAE-cellulose P11 Heparin-agarose
Fig. 2. Purified CKII x and /3 subunits. 3 pg purified CKII c( subunit and 4 v g purified C K l l p subunit were analyzed on 12.5% PAGE by staining and immunoblotting. CKII c( subunit was incubated with an anti-(CK11 x subunit) antibody. CKII p subunit was incubated with an anti-(CKII p subunit) antibody. For comparison, an immunoblot with 4 pg CKII holocnzyme from Krebs I1 mouse ascites tumour cells is included. Detection of the subunits was carried out using an antibody against the CKII holoenzyme
the Ndel- Hind111 restriction site of pT7-7. An advantage of this procedure is the very easy creation of suitable restriction sites for cloning. Therefore no time-intensive site-directed mutagenesis is necessary. A disadvantage is the fact that one has to confirm the whole sequence of the cloned PCR products, since Taq polymerase used for amplification is known for misreading while synthesising D N A strands. In our case, we sequenced two clones encoding the a subunit and two clones encoding the j subunit. One of each pair of a and p clones contained at least one wrong nucleotide which led to changes in the amino acid sequences, the other ones (pBB-3 and pBB4) contained the correct sequences. In the pBB-3 and pBB-4 vectors, the CKII a and fl subunits, respectively, are under the control of the T7 promoter. In the expression strain BL21(DE3), the T7 RNA polymerase is produced after induction with isopropyl-thio-P-D-thiogalactoside, then leads to expression of CKII subunits. In total cell extracts, new proteins of 42 kDa and 26 kDa were observed when the bacteria were transformed with pBB-3 and pBB-4, respectively. These proteins were not present in cells transformed with the pT7-7 vector alone (not shown). The expressed proteins represented up to 30% of the total bacterial protein. Purification qf the CKII suhunits,from E. coli extracts The purification of the CKII subunits was described in detail in Materials and Methods. For the isolation of the M subunit, the bacterial cell extract (Fig. 1A, lane a) was centrifuged to remove bacterial debris. The resulting pellet (Fig. 1A, lane b) contained the bulk of the expressed a subunit, yet in an insoluble form and devoid of activity. The protein pattern of the corresponding supernatant (S-I) is shown in Fig. 1 A (lane c). This fraction contained less, but active, a subunit. Therefore, the S-I fraction was used in the course of the further purification of the M subunit. After removal of bacterial ribosomes, the postribosomal supernatant, shown in Fig. 1A (lane d), was subjected to chromatographic analysis similar to the published procedures for the isolation of the holoenzyme (results not shown). After the last purification step, affinity chromatography on heparin-agarose, the a subunit was virtually homogenous (Fig. 2). The whole purification scheme is shown in Table 1. The unusually high recovery of protein (i.e. 435 mg from 594 mg starting material; S-11) after the DEAEcellulose step can be partly explained by the broad elution
pattern obtained. Consequently, about two-thirds of the fractions were collected. However, CKII-a-subunit-suppressing factors must be removed by this chromatography step leading to the observed 20-fold increase in specific activity. The subsequent phosphocellulose step showed a narrow CKII CI subunit elution pattern which allowed collection of a small number of fractions. About 10% (i.e. 46 mg) of the original amount of protein applied was recovered. Specific activity increased about tenfold. After the phosphocellulose step, CKII a subunit was about 80 - 90% pure. The final chromatography step on heparin-agarose led to a homogenous CKII M subunit preparation, as judged by PAGE analysis. Although a significant portion of the expressed protein remained insoluble, up to 12 mg M subunit/lO g bacteria was soluble and enzymatically active. The fl subunit was isolated from the pellet in a one-step purification scheme, usually in less than 24 h. Although most of the P subunit was insoluble, we could extract 15 mg soluble protein by salt treatment for the subsequent phosphocellulose chromatography (Fig. 1 B). After this purification step, the P subunit was virtually homogenous (Fig. 2). Here we obtained 5 mg P subunit from 10 g bacteria. The yield of active a and /3 subunits could not be increased considerably by disintegration of the bacterial cells with a French press instead of using sonication and by carrying out induction at 30°C instead of 37°C [31]. The yield of CKII a subunit from the nematode Candida elegans which was expressed in almost the same expression system as the one used here, was only 10% of that obtained for the human enzyme [32]. A reason for the lower yield of the recombinant a subunit from the nematode [32] may be partly due to the different lysis protocol, which involved osmolytic lysis. The purity of CKII-a subunit was established by Coomassie blue staining and immunodetection using two different CKII-specific anti-CKII antibodies. One of the two antibodies was directed against the holoenzyme from Krebs I1 mouse ascites tumor cells; the other antibody was directed against the purified CKII a subunit. The purity of the CKII fl subunit was verified by Coomassie blue staining and immunodetection using an anti-(CKII [j subunit) antibody (Fig. 2). Injluence oj'salt concentration on CKII activity The influence of increasing amounts of KCl on the a subunit and reconstituted CKII ( a and P subunits) were tested (Fig. 3). Reconstitution was complete within the time needed to complete an enzyme assay. There was a little difference between the results obtained when NaCl was used instead of KC1 (results not shown). Since we d o not know for sure whether all subunits (here we used 7 p m o l a subunit and 7 pmol J/' subunit) d o indeed associate quantitatively to form
29
11 12 13 14 15 0
0 0 0 0 0 0 0 m g ~Co o o o U m T t i n
o m o o ~o o ~ o oo oo
N m P m
o
o
KCI (rnM)
ADH
I.
Fig. 3. Eflect of NaCl and KC1 on the activity of the CKII a subunit. The effect of different concentrations of KCI (0-500 mM) on the activity of the z subunit (66 pmol) was carried out in two sets of experiments. The first set was conducted with the a subunit alone (E!). The second one was supplemented with CKII subunit at a 1 : 1 molar ratio (8)
6o
I Bottom
5
15
10
TOP
Fraction number
Ratio a/P (rnol 'rnol-')
Fig. 4. Actiwtion of the CKII a subunit by CKII J/' subunit. Activity test of the CKII a subunit was performed under standard assay conditions as described under Materials and Methods using a constant amount of CKII z subunit (36 pmol) and increasing amounts of j subunit (7.8 pmol-93.6 pmol)
an a2P2 complex, we have related the activity values to the amount of protein subunit in the assay instead of relating it to the number of subunits used. In the absence of salt, the LX subunit kinase activity was about three times higher (specific activity, 154 U/mg protein) than at 300 m M KCI. The reconstituted recombinant human enzyme showed the opposite effect. Here, the activity was lowest in the absence of salt. The observed specific activity was highest at 300 m M KCI (155 U/mg protein). This observation is consistent with earlier observations with the native holoenzyme from different mammalian sources, which exhibit maximum activities at 200-300 m M salt [l]. Recombinant CKII tl subunit from C. &guns did not show such a n opposite salt effect when compared to the native holoenzyme. Here the x subunit and the holoenzyme were most active at 100 m M NaCl[32]. One of the reasons for this different behaviour may be due to the fact the ci subunit from C. elegans is 22 amino acids shorter than the human counterpart, and what seems to be even more crucial is that the last 27 carboxy-terminal amino
Fig. 5. Sucrose density gradient analysis of CKII. CKII a subunit (1 1 pg) and CKII p subunit (4.5 pg) were in 100 pI(l0 mM Tris/HCI, pH 7.5, 5 mM MgClz and 300 mM KCI) and analyzed by sucrose density gradient centrifugation. In separate tubes, CKII z subunit ( a ) , CKII p subunit (p), reconstituted CKII holoenzyme (a + p), CKII holoenzyme (H) from Krebs I1 mouse ascites tumour cells and alcohol dehydrogenase (ADH, 150 kDa) were subjected to sucrose density gradient analysis. The positions of these proteins are indicated by arrows. In all cases, the position was either established by immunostaining using a specific anti-(CKII p subunit) antibody or by both immunostaining and activity measurement (CKII a subunit, reconstituted CKII and CKII holoenzyme from Krebs I1 mouse ascites tumour cells). Alcohol dehydrogenase was identified by Coomassie blue staining. Activity measurements of rcconstitutcd CKII are shown here. The inset shows the corresponding immunostaining of fractions 1 1 - 15
acids differ completely from the human 331.
tl
subunit kinase [16.
Reconstitution of the holoenzyme The isolation of the tl and p subunits from CKII made it possible to provide the first answer to the question whether the p subunit is a vital component of the holoenzyme; either influencing the activity of the catalytic tl subunit or being responsible for the specificity of the holoenzyme, or maybe both. Previously, the question could not be satisfactorily answered because the ci and p subunits of the holoenzyme can only be separated under denaturing conditions and activity only partially (10%) restored [20]. In order to determine the optimum subunit composition which would yield maximum enzyme activity, we designed an experiment where increasing amounts of p subunit were added to a constant amount of the tl subunit. With increasing amounts of p subunit added, a stimulation of the kinase activity of the tl subunit was obtained which was highest at a 1 : l clip subunit ratio. Additional
amounts of B subunit did not lead to a further increase in CKII activity. The stimulation was highest (fourfold) in the presence of 300 mM KCl (Fig. 4), but was heavily dependent upon the salt concentration present, especially since the a subunit kinase activity is high in the absence and low in the presence of high salt (Fig. 3). In order to find out whether the reconstitution of a and jsubunits also leads to an enzyme with biophysical properties similar to the native holoenzyme, we carried out sucrose density gradient analysis. We analyzed the holoenzyme from Krebs I1 mouse ascites tumour cells and the purified a and p subunits and the reconstituted enzyme, using alcohol dehydrogenase and other proteins as reference standards. The positions of the individual proteins after centrifugation are indicated by arrows and were established by staining with Coomassie blue (protein markers) or by immunodetection with specific antibodies directed against the purified a and jsubunits (Fig. 5). An activity profile is only shown for the reconstituted kinase. The inset shows the immunostaining obtained from the reconstituted enzyme using an anti-CKII antibody. It is clear that the reconstituted enzyme is virtually identical to the native holoenzyme [6]. The CI subunit was detected around 4.5 S, where one would expect it to sediment according to its molecular mass assuming a ?lobular structure. The same is true for the position of the B \Libunit. For reasons of clarity only the position of alcohol dehydrogenase is indicated. The K, of the reconstituted enzyme was found to be 80 pM for the synthetic peptide. This is comparable to the published K, of 60 pM for the bovine holoenzyme [29]. Using rat liver CKII, we obtained a K, of 40 pM for the synthetic peptide. The recombinant a subunit alone had a K, of 240 pM. This observation is in agreement with the data shown in Fig. 5, and suggests that association with the p subunit may also alter the substrate-binding specificity of the a subunit. This work was supported by Sonderfor.sc/zungsbereich grant 2461 B3 to 0 . G . I . N.G. was supported by a stipend from SFB 246.
REFERENCES 1. Hathaway, G. M. & Traugh, J. A. (1982) Curr. Top. Cell Regul. 21, 101 -127. 2. Pinna, L. A. (1990) Biochim. Biophys. Acta 1054, 267 -284. 3. Schncider, H. R. & Issinger, 0.G.(1989) Biotechnol. Eur. 6, 8288. 4. Boldyreff, B., Schncider, H. R., Fritz, G., Traffa, K., Seitz, G. & Issinger, 0.G.(1990) in BioTechForum, advances in molecular genetics (Kompf, J., Henke, J., Driesel, A. J. eds) vol. 3, pp. 91 -98.
5. Pinna, L. A., Meggio, F., Marchiori, F. & Borin, G. (1984) FEBS Lett. 171, 211 -214. 6. Hathaway, G. M., Zoller, M. J. & Traugh, J. A. (1981) J . Biol. Chem. 256, 11442 - 11446. 7. Feige, J.-J., Cochet, C., Pirollet, F. & Chambaz, E. M. (1983) Biochemistry 22, 1452- 3459. 8. Glover, C.V. C. (1986) J . Biol. Chem. 261, 14349-14354. 9. Prowald, K., Fischer, H. & Issinger, 0 . G . (1984) FEBSLett. 176, 479-483. 10. Schneider, H. R., Reichert, G. H. & Issinger, 0.-G. (1986) Eur. J . Biochem. 181, 733-738. 11. Perez, M., Grande, J. & Itarte, E. (1988) FEBS Lett. 238, 273276. 12. Seitz, G., Miinstermann, U., Schneider, H. R. & Issinger, 0 . G . (1989) Biochem. Biophys. Res. Commun. 163, 635-641. 13. Miinstermann, U., Fritz, G., Seitz, G., Lu, Y., Schneider, H. R. & Issinger, 0.43.(1990) Eur. J . Biochem. 189, 251 -257. 14. Sommercorn, J. & Krebs, E. G. (1987) J . Biol. Chem. 262,38393843. 15. Saxena, A., Padmanabha, R. &Glover, C. V. C. (1987) Mol. Cell. Biol. 7, 3409 - 341 7. 16. Meisner, H., Heller-Harrison, R., Buxton, J. & Czech, M. P. (1989) Biochemistry 28, 4072 -4076. 17. Heller-Harrison, R. A., Meisner, H. & Czech, M. P. (1989) Biochemistry 28, 9053 -9058. 18. Jakobi, R., Voss, H. & Pyerin, W. (1989) Eur. J . Biochem. 183, 227-233. 19. Ackerman, P., Glover, C. V. C. & Osheroff, N. (1990) Proc. Nut1 Acad. Sci. U S A 87, 821 - 825. 20. Cochet, C. & Chambaz, E. M. (1983) J . Biol. Chem. 258, 14031406. 21. Reichert, G. H. & Issinger, 0.-G. (1985) Biochimie (Paris) 67, 657 - 661. 22. Tautz, D. & Renz, M. (1983) Anal. Biochem. 132,14- 19. 23. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular cloning, a laboratory manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 24. Tabor, S. & Richardson, C. C. (1985) Proc. Nail Acad. Sci.U S A 82, 1074-1078. 25. Studier, F. W. & Moffat, 9 . A. (1986) J . Mol. Biol. 189, 113130. 26. Laemmli, U. K . (1970) Nature 227, 680-685. 26. Schneider, H. R. (1988) PhD Thesis, University Kaiserslautern Germany. 27. Issinger, 0.-G., Martin, T., Richter, W. W., Olson, M. & Fujiki, H. (1988) E M B O J . 7,1621-1626. 28. Kuenzel, E. A. & Krebs, E. G. (1985) Proc. Nail Acad. Sci. USA 82, 737 - 741. 29. Kuenzel, E. A., Mulligan, J. A,, Sommercorn, J. & Krebs, E. G . (1987) J . Biol. Chem. 262, 9136-9140. 30. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254. 31. Slice, L. W. & Taylor, S. S. (1989) J . Biol. Chem. 264, 2094020946. 32. Hu, E. & Rubin, C. S. (1990) J . Biol. Chem. 265, 20609-20615. 33. Hu, E. & Rubin, C. S. (1990) J . Biol. Chem. 265, 5072-5080.
