Mol Cell Biochem DOI 10.1007/s11010-008-9824-3
Biochemical characterization of CK2a and a0 paralogues and their derived holoenzymes: evidence for the existence of a heterotrimeric CK2a0 -holoenzyme forming trimeric complexes Birgitte B. Olsen Æ Tine Rasmussen Æ Karsten Niefind Æ Olaf-Georg Issinger
Received: 19 May 2008 / Accepted: 29 May 2008 Ó Springer Science+Business Media, LLC. 2008
Abstract Altogether 2 holoenzymes and 4 catalytic CK2 constructs were expressed and characterized i.e. CK2a1-335 2 b2; CK2a0 -derived holoenzyme; CK2a1-335; MBP-CK2a0 ; His-tagged CK2a and His-tagged CK2a0 . The two Histagged catalytic subunits were expressed in insect cells, all others in Escherichia coli. IC50 studies involving the established CK2 inhibitors DMAT, TBBt, TBBz, apigenin and emodin were carried out and the Ki values calculated. Although the differences in the Ki values found were modest, there was a general tendency showing that the CK2 holoenzymes were more sensitive towards the inhibitors than the free catalytic subunits. Thermal inactivation experiments involving the individual catalytic subunits showed an almost complete loss of activity after only 2 min at 45°C. In the case of the two holoenzymes, the CK2a0 -derived holoenzyme lost ca. 90% of its activity after 14 min, whereas CK2a1-335 b2 only showed a loss of ca. 2 40% by this time of incubation. Gel filtration analyses were performed at high (500 mM) and low (150 mM) monovalent salt concentrations in the absence or presence of ATP. At 500 mM NaCl the CK2a0 -derived holoenzyme eluted at a position corresponding to a molecular mass of 105 kDa which is significantly below the elution of the CK2a12 335 b2 holoenzyme (145 kDa). Calmodulin was not phosphorylated by either CK2a1-335 b2 or the CK2a0 -derived 2 holoenzyme. However, in the presence of polylysine only the CK2a1-335 b2 holoenzyme could use calmodulin as a 2 B. B. Olsen T. Rasmussen O.-G. Issinger (&) Institut for Biokemi og Molekylær Biologi, Syddansk Universitet, Campusvej 55, 5230 Odense, Denmark e-mail:
[email protected] K. Niefind Department fu¨r Chemie, Institut fu¨r Biochemie, Universita¨t zu Ko¨ln, Zu¨lpicher Str. 47, 50674 Ko¨ln, Germany
substrate such as the catalytic subunits, in contrast to the CK2a0 -derived holoenzyme which only phosphorylated calmodulin weakly. This attenuation may be owing to a different structural interaction between the catalytic CK2a0 subunit and non-catalytic CK2b subunit. Keywords Substrate
Protein kinase CK2 Isozymes Inhibitors
Abbreviations AMPPNP Adenylyl imidodiphosphate DMAT 2-Dimethyl-4,5,6,7-tetrabromobenzimidazole MBP Maltose binding protein TBBt Tetrabromobenzotriazole TBBz Tetrabromobenzimidazole
Introduction Protein kinase CK2 is a tetrameric enzyme, consisting of 2 catalytic a-subunits and 2 non-catalytic b-subunits. The enzyme has been distantly assigned to the CMGC group of protein kinases. It is ubiquitously found from yeast to man. The kinase phosphorylates a serine or threonine when an acidic residue, either an acidic amino acid or a phosphorylated serine/threonine, is found in n + 3 position of the target hydroxyamino acid [1-7]. However, exceptions to this general rule have been described [8]. Owing to the simplicity of the consensus sequence it is not surprising that CK2 phosphorylates a plethora of substrates as many other protein kinases do. CK2 is a constitutively active protein kinase, obviously without a regulator. Moreover, it can use ATP and GTP as a phosphoryl donor, owing to its particular nucleotide binding site [9].
123
Mol Cell Biochem
In man, 2 catalytic isozymes and 1 non-catalytic subunit have been identified, i.e. CK2a, CK2a0 and CK2b all of which are located on different chromosomes. The two catalytic subunits are very related only showing a distinct sequence difference in the last 50 amino acids of the C-terminal part of the enzyme [10]. In mammals the CK2a subunit is found in all tissues, from head to toe, whereas the CK2a0 subunit is mainly found in brain and testes, supporting the notion that this subunit may have functions specific in these tissues [11]. Indeed, a knockout of the CK2a0 subunit in mice showed a phenotype similar to globozoospermia in man, whereas a knockout of the CK2a-subunit was lethal as was a knockout of the non-catalytic subunit [12–14]. CK2 has also been found altered in many diseases, especially cancer [15–17], making it an interesting target within the druggable family of protein kinases [18]. From the 3 CK2 subunits, so far only the CK2a and CK2b subunits and the derived holoenzyme were crystallized and the structures solved [19–21]. Corresponding data for the CK2a0 subunit and the derived holoenzyme are still lacking. However, preliminary biochemical and biophysical data obtained from experiments involving the CK2a0 subunit suggest that it is distinctively different from CK2a [22]. The data presented in this paper further support this view especially with respect to differences in the elution behaviour in the presence of high and low monovalent salt concentrations and whether ATP is absent or present.
Materials and methods Materials Apigenin and Emodin were purchased from Calbiochem, TBBt, TBBz and DMAT were a gift from Dr. Zygmunt Kazimierczuk (Department of Experimental Pharmacology, Polish Academy of Sciences Medical Research Center, Warsaw, Poland). Stock solutions of inhibitors were 10 mM in DMSO and these were diluted in assay buffer to the desired concentrations. The synthetic peptide substrate RRRDDDSDDD was from KinaseDetect. [c-32P]ATP with a specific activity of 3,000 Ci/mmol was from Hartmann Analytic. Polylysine and calmodulin used for phosphorylation assays were from Sigma. The Bac-toBac HT Vector kit and the Bac-to-Bac Baculoviral Expression System were purchased from Invitrogen. Cloning, expression and purification CK2a1-335 b2, CK2a1-335, CK2a1-335 bD1-4 , CK2a0 holo2 2 2 0 enzyme and MBP-CK2a were expressed and purified as
123
previously described [21–24]. The coding region of CK2a and CK2a0 was subcloned into pFastBac HTB and C, respectively, to introduce an N-terminal histidine-tag. Production of recombinant baculoviral particles and expression of His-CK2a and His-CK2a0 in Sf9 cells was performed as recommended by the manufacturer. 72 h post infection, Sf9 cells were lysed by sonication in 20 mM Na2HPO4 pH 7.4, 500 mM NaCl, 20 mM imidazole, protease inhibitors (Complete, Roche) and the cleared lysate loaded onto a HisTrap column (GE Healthcare). His-tagged proteins were eluted with an imidazole gradient up to 500 mM and peak fractions dialyzed against 25 mM TrisHCl pH 7.2, 300 mM NaCl, 1 mM DTT. Protein kinase assays CK2 activity was tested in a 40 ll assay mix (final concentration: 37.5 mM Tris-HCl pH 8.5, 25 mM MgCl2, 1.5 mM DTT, 125 lM [c-32P]ATP, 100 lM synthetic peptide RRRDDDSDDD, 0.25 mg/ml BSA and for the holoenzymes 150 mM NaCl). The reactions were incubated 10 min at 30°C and 20 ll were spotted on P81 paper (Whatman). The filter papers were washed in 0.75% phosphoric acid before counting in a scintillation counter (Canberra-Packard). For thermal inactivation, the enzymes were incubated at 45°C at the times indicated in the figures prior to activity testing as described above. For determination of IC50 values, the CK2 inhibitors were included in the CK2 activity tests at final concentrations ranging from 0 to 10 lM. For autophosphorylation experiments CK2a1-335 b2, 2 1-335 D1-4 0 CK2a2 b2 and CK2a holoenzyme were incubated at 30°C for 30 min in a total volume of 20 ll containing 25 mM Tris-HCl pH 8.0, 150 mM NaCl, 5 mM MgCl2, 1 mM DTT and 50 lM [c-32P]ATP. The reactions were terminated by addition of SDS-sample buffer and the proteins separated by SDS-PAGE. The gel was stained with Coomassie Blue and used for autoradiography. 2 lg calmodulin was incubated with 0.3 mU CK2a12 335 b2, CK2a0 holoenzyme, CK2a1-335 or His-CK2a0 in the absence or presence of 2 lM polylysine in a total volume of 20 ll as described above. Reactions were terminated by addition of SDS-sample buffer and resolved by 15% SDSPAGE. Analytical gelfiltration CK2a1-335 b2, CK2a1-335 bD1-4 , and CK2a0 holoenzyme 2 2 2 were loaded onto a Superdex200 column mounted on a SMART chromatography system (Amersham-Pharmacia) equilibrated in either 25 mM Tris–HCl pH 8.5, 1 mM DTT, 150 mM NaCl or 25 mM Tris–HCl pH 8.5, 1 mM DTT, 500 mM NaCl.
Mol Cell Biochem
Prior to analysis, CK2a1-335 b2, CK2a1-335 bD1-4 and 2 2 2 0 CK2a holoenzyme were incubated with 100 lM ATP (Sigma) for 10 min at 30°C.
Results and discussion Expression and purification of the CK2 constructs CK2a1-335 and CK2b were expressed separately in E. coli without tags. Bacterial pellets harbouring the expressed recombinant subunits were mixed and the self-assembled holoenzyme was purified by conventional chromatography as described earlier [23]. When the CK2a0 isozyme was expressed in E. coli it was insoluble, which could be
overcome by expressing the CK2a0 subunit as an MBPfusion protein. The odd behaviour of the CK2a0 isozyme with respect to its expression as a soluble protein in E. coli is a puzzle, especially since the closely related CK2a isozyme is readily soluble. The primary sequences of the two isozymes (Fig. 1) do not let us expect such a fundamental difference with respect to the results we have obtained. The major sequence deviation between the two catalytic isozymes is found in the C-terminal amino acids. Hence, one could speculate that this part of the CK2a0 molecule might be responsible for the lack of its solubility. However, a chimeric protein consisting of a CK2a molecule to which the last 25 C-terminal amino acids from CK2a0 were attached was soluble, supporting the notion that also other
Fig. 1 Multiple sequence alignment of the two CK2a isozymes CK2a and CK2a0 with maize CK2a. The alignment was calculated with PILEUP from the sequence analysis package GCG [52]. Positions identical in all three chains are printed with bold capital letters
123
Mol Cell Biochem
sequences within the CK2a0 molecule might be responsible for its lack of solubility upon expression in E. coli (Olsen et al., unpublished). When CK2a0 was expressed as a Histagged protein in insect cells the severe solubility problems encountered in E. coli were much less pronounced, albeit after some time also the insect cell-derived recombinant human CK2a0 showed faint precipitation. Hence, for the production of a CK2a0 -based holoenzyme, CK2b (no tag) and MBP-CK2a0 were separately expressed in E. coli. Both proteins were purified and used for the production of the holoenzyme by self-assembly followed by treatment with Factor Xa to cleave the MBP tag. Purification of the resulting holoenzyme was performed by gel filtration. The His-tagged CK2 subunits a and a0 were expressed in Sf9 insect cells and purified by Ni-NTA affinity chromatography. Figure 2a shows an SDS PAGE analysis of 6 different CK2 constructs after purification; the two holoenzymes, CK2a1-335 b2 (lane 1) and CK2a0 holoenzyme (lane 2), and 2 4 catalytic subunits CK2a1-335 (lane 3), MBP-CK2a0 (lane 4), His-tagged CK2a (lane 5) and His-tagged CK2a0 (lane 6). All recombinant CK2 constructs were[95% pure. Only in the case of the His-tagged CK2a a faint lower molecular mass band is visible, a well-known degradation product usually seen with all recombinant CK2a preparations which has led to the design of a CK2a deletion mutant
lacking the last 55 C-terminal amino acids that are cleaved by an ‘autoproteolytic’ process [25]. The deletion mutant CK2a1-335 is stable and undistinguishable from the full length CK2a molecule with respect to all known parameters inherent to the wildtype a-catalytic subunit. Yet, the role and function of the ‘autocleavage’ product is obscure. Figure 2b shows the theoretically calculated molecular masses of the recombinant subunits used throughout the experiments. The tagging led to additional sequences upstream and downstream of the His-tag and downstream of the MBP-tag, respectively. Hence, the recombinant proteins deviate in their molecular masses from the wildtype CK2 subunits, which have theoretical molecular masses of 45,141 (CK2a) and 41,400 daltons (CK2a0 ), respectively. Specific activity of the constructs Incorporation of phosphate into the synthetic peptide (RRRDDDSDDD) was measured in the presence of increasing amounts of enzyme, ranging from 2.5–40 ng. The enzyme activity curves for the 6 constructs are shown in Fig. 3a. The calculated specific activities for the 6 constructs (lmoles/min/mg) are shown in Fig. 3b. The Cterminal deletion mutant CK2a1-335 shows the highest specific activity, whereas the other constructs exhibit similar specific activities. Inhibitor studies
A 1-3
kDa
2α CK
35 2β
2
2α CK
o ’h
e ym nz 5 l oe 1-3
2α CK
3
α’ α α’ K2 K2 K2 -C P s-C B s-C i i M H H
220
In search for differences between the two isozymes, inhibitor studies were performed using 5 well-established
97 66
A
46 CPM
CK2α1-3352β2 CK2α’ holoenzyme CK2α1-335 MBP-CK2α’ His-CK2α His-CK2α’
30 20
B
Construct 1-335
2 β2
MW of catalytic subunit
Additional N-terminal aa
40,002
-
CK2α’ holoenzyme
41,834
ISEFGS
CK2α1-335
40,002
-
CK2α
MBP-CK2α’ His-CK2α His-CK2α’
84,298 48,515 44,748
MalE-TNSSSNNNNNNNNNNIEGRISEFGS MSYYHHHHHHDYDIPTTENLYFQGAMGSMS MSYYHHHHHHDYDIPTTENLYFQGAMGIH
Fig. 2 SDS-PAGE analysis of the two bacterially expressed holoenb2 and CK2a0 holoenzyme and catalytic subunits zymes CK2a1-335 2 CK2a1-335, MBP-CK2a0 , as well as the two catalytic subunits HisCK2a and His-CK2a0 expressed in Sf9 cells (a). The catalytic subunits were expressed in different vectors with different aminoterminal tags causing additional N-terminal amino acids as indicated giving rise to different molecular masses (in daltons) (b)
123
B
Construct
Specific activity [µmoles/min/mg]
CK2α1-3352β2
3,43
CK2α’ holoenzyme
2,26
CK2α1-335
7,88
MBP-CK2α’
2,14
His-CK2α
1,60
His-CK2α’
2,96
Fig. 3 Activity curves from the two CK2 holoenzymes CK2a1-335 b2 2 (r) and CK2a0 holoenzyme (h) and the four catalytic subunits CK2a1-335 (m), MBP-CK2a0 (9), His-CK2a ( ) and His-CK2a0 () (a). The linear part of the curves was used to calculate the specific activity of the enzymes expressed in lmoles/min/mg (b)
Mol Cell Biochem
subunits. Moreover, only DMAT showed a concentrationdependent inhibition of enzyme activity. The other inhibitors were mostly ineffective until a threshold [1 lM was reached (Fig. 4). Within the last few years the number of CK2-specific inhibitors has continuously increased [31–38], supporting the notion that CK2 is increasingly being recognized as a ‘druggable’ member of the protein kinase family, especially owing to its strong anti-apoptotic properties [18]. This development will require detailed knowledge about the best conditions for a high throughput screening in order to discover potential inhibitors. In order to achieve this goal, attention should be focused on: (i) origin of species from which the kinase is derived, (ii) subunit or holoenzyme and (iii) peptide or protein substrate. (i) Concerning the origin of species, a recent report from Raaf et al. [39] showed that significant structural differences between the CK2a subunit from Z. Mays and man
CK2-specific inhibitors, i.e. DMAT, TBBt, TBBz, apigenin and emodin [26–28]. From the obtained IC50 values (Fig. 4) the Ki values (Table 1) were calculated [29]. The data show that the CK2a1-335 b2 holoenzyme was the most sensitive CK2 2 construct towards all inhibitors employed. The CK2a0 holoenzyme was less sensitive but still showed a stronger inhibition than the free catalytic subunits. Hence, we conclude that in a holoenzyme conformation the binding of the ATP-mimetic inhibitors differs from what is observed in the case of the free catalytic subunits. In particular TBBz is more effective in inhibiting the holoenzymes than the catalytic subunits. This observation agrees with observations from others [30]. The strongest effect on all CK2 constructs was obtained with DMAT [27]. Remarkable is the observation that with almost all inhibitors the holoenzymes were more sensitive towards the employed inhibitors than the free catalytic Fig. 4 Dose-dependent b2 (r), inhibition of CK2a1-335 2 CK2a0 holoenzyme (h), CK2a1-335 (m), MBP-CK2a0 (9), His-CK2a ( ) and HisCK2a0 () by DMAT, TBBt, TBBz, apigenin and emodin
Table 1 Ki values for the five different inhibitors i.e. DMAT, TBBt, TBBz, apigenin and emodin tested on the different forms of CK2. Ki values (in lM) were calculated using the Cheng-Prussof equation [29]
b2 CK2a1-335 2
CK2a0 2b2
CK2a1-335
MBP-CK2a0
His-CK2a
His-CK2a0 0.3
DMAT
0.11
0.27
0.3
0.32
0.31
TBBt
0.25
0.7
0.68
0.64
0.54
0.43
TBBz Emodin
0.29 0.2
0.66 0.54
1.1 0.72
1.53 0.79
0.87 0.72
0.8 0.55
Apigenin
0.33
0.29
0.85
0.49
0.7
0.43
123
Mol Cell Biochem
exist especially with respect to inhibitor binding, supporting the notion that only human CK2 subunits should be employed, when searching for an inhibitor targeting human CK2. (ii) Concerning the question whether the holoenzyme or the catalytic subunit should be used for HTS the answer is less clear if we look at the modest differences in the Ki values (Table 1), however, we tend to suggest the use of the holoenzyme, especially with respect to inhibitors, which are not ATP mimetic but target sites which are involved in substrate binding. (iii) To answer the question whether a peptide or a protein substrate should be used we clearly favour the use of a protein substrate exhibiting higher KM and Kcat values than the currently used peptide substrates and besides owing to their size, corroborating the ‘real’ situation during enzyme/substrate interaction.
action of polylysine in the case of the CK2a0 -based holoenzyme was less pronounced, supporting the notion that binding of polylysine does not target the CK2b in the same way as has been observed for the CK2a1-335 b2 holoen2 0 zyme. Since CK2a alone phosphorylates calmodulin to the same extent as the CK2a-subunit, a possible explanation could be a difference in the ‘shielding’ of the CK2a0 subunit by the CK2b as in the CK2a1-335 b2 holoenzyme. The 2 background for this shielding can be a positioning of the CK2b in such a way that the binding of polylysine does not effect its relation to the CK2a0 -subunit hence attenuating its positive effect it normally has on the CK2b subunit within the CK2a1-335 b2 holoenzyme with respect to calmodulin 2 phosphorylation. Thermal inactivation studies
Calmodulin phosphorylation A peculiarity of CK2 is the recognition of substrates. Calmodulin, which is not a substrate for the holoenzyme, is readily phosphorylated by the catalytic subunit alone. However, calmodulin becomes a substrate for the holoenzyme in the presence of polybasic compounds such as polylysine, spermine or spermidine supporting the notion that the non-catalytic CK2b subunit negatively affects the recognition of calmodulin as a substrate target for the catalytic CK2 subunit within the tetrameric holoenzyme complex [40–42]. Based on so far available evidence that the a0 /b subunit interaction within the holoenzyme complex is less tight than that of the a/b complex, we hypothesized that calmodulin could be a substrate for the a0 -based holoenzyme, also in the absence of polybasic compounds. Figure 5 shows an autoradiograph of calmodulin phosphorylation by the various CK2 constructs. Unexpectedly, the CK2a0 -based holoenzyme behaved similar to the CK2a1-335 b2 holoenzyme, inasmuch that calmodulin did 2 not serve as a substrate in the absence of polylysine. In the case of the CK2a1-335 b2 holoenzyme the presence of 2 polylysine abolished the negative action of the CK2bsubunit leading to a phosphorylation of calmodulin; the
Next we performed thermal inactivation studies. Figure 6a shows the thermal inactivation pattern of the 4 catalytic subunits. Loss of activation is already at its maximum ([90%) after 2 min. Similar results have been described [21, 43], strengthening the concept that the CK2b dimer exerts a powerful protection of the catalytic subunit when together in a tetrameric complex [44]. The MBP-tagged
e s in yl y l e o in ine +p ne l ys l ys ysi ol y aM aM ol y l yl p o + p e+ C + C + p aM M + aM CaM ym yme M M a ine + C β 2+ oenz oenz Ca Ca ’ + C + C lys y 5 2 + l l 5 β ’ 3 o 35 + 2α o 35 2 33 2 ol 1-3 2α 1-3 ’ h α’ h K + p α1-3 α1K α α 2 2α s-C -C 2 K2 K2 K2 M M C C CK CK Hi His Ca Ca CK C
- CaM
Fig. 5 Autoradiograph showing phosphorylation of calmodulin by b2, CK2a0 holoenzyme, CK2a1-335 and His-CK2a0 in the CK2a1-335 2 absence or presence of polylysine
123
Fig. 6 The influence of 45°C over a time of up to 14 min on enzyme b2 (u) and CK2a0 activity for the two holoenzymes CK2a1-335 2 holoenzyme (h) (a) and the catalytic subunits CK2a1-335 (m), MBPCK2a0 (9), His-CK2a ( ) and His-CK2a0 () (b)
Mol Cell Biochem
CK2a0 protein still retains ca. 40% of its activity after 2 min-incubation, but its activity also declines to almost zero after 6 min. The slightly more stable nature of MBP-tagged CK2a0 can be explained by the MBP-tag which may provide a protective effect, especially since its size is in the same molecular mass range as the associated CK2a0 subunit. Figure 6b shows the inactivation curves of CK2a1-335 b2 2 and CK2a0 holoenzyme after 2, 6, 10 and 14 min at 45°C. There is a significant difference between the two holoenzymes. Whereas the CK2a1-335 b2 holoenzyme loses ca. 2 about 30% of its activity after the first 2 min at 45°C, further incubation up to 14 min only results in an additional loss of activity of 10%, so that the total percentage of inactivation is ca. 40%. In the case of the CK2a0 holoenzyme the initial loss of activity after 2 min-incubation is 50% and after 14 min of incubation only 10% activity remains. Interestingly, a CK2 chimera consisting of CK2a from Z. Mays and human CK2b showed a similar loss of activity as it was observed for the CK2a0 holoenzyme [25]. Since all three catalytic subunits are closely related in their primary sequence (Fig. 1) and share similar molecular masses, the protective loss of the associated CK2b dimer in the holoenzyme complexes from CK2a0 and CK2a from maize cannot be readily explained. The significant difference between the CK2a1-335 b2 and 2 the CK2a0 -containing holoenzyme towards heat denaturation could be suggestive of the CK2a0 being in a less tight contact with the regulatory CK2b subunits than is the case of the CK2a. Gel filtration analyses Gel filtration analyses of CK2a1-335 b2 and the CK2a0 ho2 loenzymes in the absence and presence of ATP were conducted at either 500 mM NaCl or 150 mM NaCl. CK2a1-335 b2 in 500 mM NaCl 2 In the presence of high salt CK2a1-335 b2 elutes at a posi2 tion corresponding to a molecular mass of ca. 145 kDa, i.e. calculated under the assumption that the molecule elutes as a globular protein, which is in the molecular mass range for a heterotetrameric holoenzyme (132 kDa) consisting of two catalytic subunits (80 kDa) and the CK2b dimer (52 kDa) (Table 2, upper panel). Elution was unaffected by the presence of ATP. CK2a1-335 b2 in 150 mM NaCl 2 At low salt concentrations and in the absence of ATP the CK2a1-335 b2 holoenzyme elutes at a position correspond2 ing to ca. 290 kDa (Table 2, lower panel), which is double the value found at 500 mM NaCl. Hence, low salt
Table 2 Analytical gel filtration analysis on Superdex200 of CK2a12 335 b2 and CK2a0 holoenzyme 500 mM NaCl
b2 CK2a1-335 2 0
CK2a holoenzyme
-ATP (kDa)
+ATP (kDa)
145
145
95
145 (15%) 95 (85%)
150 mM NaCl -ATP (kDa)
+ATP (kDa)
CK2a1-335 b2 2
290
[670 (80%)
CK2a0 holoenzyme
290
290
CK2a1-335 bD1-4 2 2
290
[670 (85%)
290 (20%)
290 (15%) The molecular weight indicates the elution of the enzymes in 500 mM NaCl (upper panel) or 150 mM NaCl (lower panel). In cases where more than one peak were observed the relative amount of enzyme in each peak is indicated in percentage. Prior to analysis the enzymes were incubated with ATP as described in materials and methods
concentrations initiate either a structural change altering the overall shape of the molecule in a way that it elutes faster, or alternatively low salt concentrations lead to dimerization/trimerization. However, in the presence of ATP the elution pattern of the CK2a1-335 b2 holoenzyme drastically changes. Ca. 80% 2 of the molecules are eluting at a position corresponding to [670 kDa, i.e. at the exclusion volume of the column. Only ca. 20% of the molecules elute at 290 kDa. High molecular mass complexes of CK2a holoenzyme depending on the salt conditions have been described previously [45, 46] and intermolecular phosphorylation of the CK2b subunit was shown to accompany this aggregation process [21, 47]. The target site within the CK2b subunit was identified as Ser2 and Ser3 [48, 49]. CK2 holoenzyme crystal contacts between different CK2 tetramers have been demonstrated and these mainly ionic interactions lead to trimeric rings of CK2 holoenzymes in the crystal. In these rings each CK2 tetramer possesses one CK2a subunit open for substrate binding and another one whose active site is blocked by a secondary contact with CK2b from a neighbouring tetramer. This observation fits to previous findings that salt-sensitive ring-like aggregates of CK2 holoenzymes can exist which possess significant catalytic activity [50]. Interestingly, these experimental findings were postulated by modelling studies [51]. CK2a0 holoenzyme in 500 mM NaCl The theoretical molecular mass of a tetrameric CK2a0 holoenzyme (consisting of 2 a0 molecules and 1 CK2b
123
Mol Cell Biochem
dimer) is 136 kDa and thereby very similar to the molecuIar mass of the CK2a1-335 b2 holoenzyme (132 kDa). 2 Hence, one would expect the CK2a0 holoenzyme to elute at the same position as the CK2a1-335 b2 holoenzyme. 2 However, to our surprise the CK2a0 holoenzyme eluted at a position corresponding to a calculated molecular mass of 95 kDa (Table 2, upper panel). In order to explain the difference in the elution pattern, one can either speculate that the shape of the CK2a0 holoenzyme differs significantly from the CK2a holoenzyme or postulate a heterotrimeric complex consisting of 1 CK2a0 subunit in association with 1 CK2b dimer. Such a complex has a molecular mass of 94 kDa and fits the experimental data obtained from the elution pattern. To further analyse the presence of trimers in the case of the CK2a0 holoenzyme, the peak fractions from SMART analyses of CK2a0 holoenzyme and CK2a1-335 b2 at 2 500 mM NaCl were subjected to SDS-PAGE and the gel was coomassie stained. As can be seen from Fig. 7, although the same intensity is seen in the CK2b bands, there is a difference with respect to staining intensities of the catalytic subunits, with the intensity of CK2a1-335 being obviously stronger than the intensity of the CK2a0 band. The amino acid composition of the catalytic subunits with respect to amino acids able to bind coomassie stain is negligible and cannot account for the difference in staining intensity. This finding supports the idea that the CK2a0 holoenzyme is a heterotrimeric complex (a0 b2) at 500 mM. In the presence of ATP, the majority (85%) of the CK2a0 holoenzyme molecules eluted at a position corresponding to a molecular mass of about 95 kDa; however, ca. 15% of the molecules eluted at a position corresponding to ca. 145 kDa. The reason for the appearance of a CK2a0 holoenzyme population that elutes at a position corresponding to that of the CK2a holoenzyme is not clear. Yet one could speculate that the binding of ATP causes a structural change in the trimeric complex such that it can bind another a0 subunit. Hence, a minority of the molecules e
35 2β 2
ym
1-3
2α
CK
lo
ho
’ 2α
z en
CK
-CK2α’ -CK2α1-335 -CK2β
Fig. 7 Coomassie stained SDS-PAGE of the peak fractions from b2 and CK2a0 SMART analyses at 500 mM NaCl of the CK2a1-335 2 holoenzyme. The loading was adjusted so that the intensity of the CK2b bands were the same
123
found after incubation with ATP would be tetrameric complexes, consisting of 2 a0 subunits and 1 CK2b dimer. Although very speculative at the moment this assumption cannot completely be ruled out. The sparsity of the presumed a0 heterotetramers may be owing to technical and methodological reasons such as protein concentration, suboptimal pre-incubation times and temperatures. What is lending support to our speculations is the elution pattern of the CK2a1-335 b2 holoenzyme, which is unaf2 fected by the presence of ATP, a result one would expect if a stable tetrameric structure is prevailing and which is already in a ‘subunit-saturated complex’. CK2a0 holoenzyme in 150 mM NaCl When the CK2a0 holoenzyme was subjected to gel filtration analyses in the presence of low salt, either in the absence or presence of ATP, the enzyme eluted at 290 kDa (Table 2, lower panel) giving rise to the speculation that in the presence of low salt the heterotrimeric CK2a0 holoenzyme can associate into ‘trimers’. This association would be independent of an intermolecular or intramolecular autophosphorylation event, which is in agreement with the trimers appearing independently of the presence of ATP. In earlier studies it was indirectly postulated that autophosphorylation of CK2b is a pre-requisite for the aggregate formation (higher molecular mass complexes [670 kDa). As seen in Fig. 8a only the CK2a1-335 b2 2 holoenzyme shows autophoshorylation of the CK2b subunit, whereas the CK2a0 holoenzyme does not. To investigate further whether the lack of high molecular mass aggregation in the case of the CK2a0 holoenzyme is not due to a lack of CK2b phosphorylation but due to the trimeric state, we used a CK2a1-335 bD1-4 holoenzyme, 2 2 where the CK2b phosphorylation sites were deleted. Figure 8b shows, as expected, that there is no autophosphorylation of the CK2b subunit. More interesting, however, is the fact that CK2a1-335 bD1-4 behaves identical 2 2 to what is observed with the CK2a1-335 b 2 2 holoenzyme in gel filtration experiments (Table 2, lower panel) i.e. it forms higher molecular mass aggregates in the presence of ATP. Hence, phosphorylation of the CK2b subunit is not a prerequisite for high molecular mass aggregation. A convincing model of high molecular mass aggregation was published by Poole et al. [51]. According to this model tetrameric CK2 holoenzymes can interdigitate in a linear fashion to form long linear filaments. This type of aggregation requires a holoenzyme complex with two CK2a chains; it is, however, not compatible with a trimeric ab2 or a0 b2 complex. This is quite different from another type of aggregation, namely trimeric rings as they occur in the CK2a2b2
Mol Cell Biochem
e
A 35
1-3
α K2
C
ym
β2
2
α’ K2
z en
lo
ho
C
-CK2β
B
35 β 2 1-3 2
2α CK
5
3 1-3
2α CK
-4 2
∆1
β
2
-CK2β
Fig. 8 Autoradiograph showing the autophosphorylated b-subunit b2 and CK2a0 holoenzyme (a). Autoradiograph from CK2a1-335 2 b2 showing the autophosphorylated b-subunit from CK2a1-335 2 and CK2a1-335 bD1-4 (b) 2 2
holoenzyme [50]. These rings can be formed by heterotetramers, however, only with a conspicuous break of symmetry; only one of the two catalytic chains is required for ring closure and only this one harbours a bound AMPPNP molecule [20]. Accordingly heterotrimeric holoenzymes that miss the superfluous catalytic chain are even more unconstrainedly compatible with trimeric rings. This is illustrated in Fig. 9, which shows a model of such a ‘trimer of heterotrimers’. The model in Fig. 9 integrates various experimental facts: (i) It shows how heterotrimers can reversibly stabilize to ring-like 290 kDa aggregates as they are found in gelfiltration experiments. (ii) It explains how autophosphorylation can happen, but only as a by-product while the really important step is capturing of a cosubstrate molecule. (iii) The rings are compatible with a dissociation/ association scenario in the presence of polybasic effectors.
Conclusions and perspectives The results obtained within these investigations show for the first time that the intermolecular phosphorylation of the CK2a1-335 b2 holoenzyme is not responsible for the high 2
Fig. 9 Possible arrangement of CK2a0 -based hetero trimers to a ring like structure of about 290 kDa. The model was generated from the ‘trimers of heterotetramers’ found in the crystals of the CK2a-based holoenzyme. In those trimeric rings one of the catalytic subunits is required for the ring closure while the other one is merely attached to the periphery and thus disposable. In this way heterotrimers composed of one CK2a0 chain (shown in blue, grey and red) and a CK2b dimer (shown in black, yellow and magenta) can stabilize by formation of ‘trimers of heterotrimers’
molecular mass aggregation observed at physiological salt concentrations in the presence of ATP. It is rather the binding of ATP that leads to structural changes within the b2 holoenzyme, which allows heterotetrameric CK2a1-335 2 the formation of high molecular mass complexes. In the case of the heterotrimeric CK2a0 -based holoenzyme the high molecular mass aggregation is not observed; hence the trimeric state must be made responsible for the lack of high molecular mass aggregation, i.e. the binding of ATP within one catalytic subunit is not sufficient to induce this high molecular mass aggregation. If the aggregation is of physiological relevance it should be reversible; indeed Olsen et al. [21] have shown that CK2 aggregates can be dissociated in vitro in the presence of polybasic molecules such as polylysine. Therefore, we hypothesize that also under in vivo conditions dissociation of the aggregates might be possible. Taken together, the results of the thermal inactivation studies and the gelfiltration observations indicate (i) that the affinity between human CK2a0 and CK2b is significantly lower than that of human CK2a and CK2b and (ii) that this difference—since it also refers to the artificial variant CK2a1-335—has nothing to do with the unrelated C-terminal segments of CK2a and CK2a0 . The structural
123
Mol Cell Biochem
basis of the affinity difference is unclear at the moment. However, the finding fits nicely to observations accumulated in recent years that the classical heterotetrameric holoenzyme complex is not the only form for CK2 activity. Acknowledgments We thank Drs. B. Boldyreff and B. Guerra for critically reading the manuscript and Hans H. Jensen for expert technical assistance. B.B.O is supported by grant no. DP06083 from the Danish Cancer Society, K.N from the Deutsche Forschungsgemeinschaft (DFG) grant no. NI 643/1-3 and O.G.I by grant no. 002521109210 from the Danish Cancer Society; grant no. 21-04-0517 from the Danish Research Agency, the NOVO Nordisk Foundation and from a donation dedicated to Cancer Research from Karen Marie Maaløe.
15.
16.
17.
18.
19.
References 1. Issinger OG (1993) Casein kinases: pleiotropic mediators of cellular regulation. Pharmacol Ther 59:1–30. doi:10.1016/01637258(93)90039-G 2. Guerra B, Issinger OG (1999) Protein kinase CK2 and its role in cellular proliferation, development and pathology. Electrophoresis 20:391–408. doi:10.1002/(SICI)1522-2683(19990201)20:2& lt;391::AID-ELPS391>3.0.CO;2-N 3. Guerra B, Boldyreff B, Sarno S, Cesaro L, Issinger OG, Pinna LA (1999) CK2: a protein kinase in need of control. Pharmacol Ther 82:303–313. doi:10.1016/S0163-7258(98)00064-3 4. Allende CC, Allende JE (1999) Promiscuous subunit interaction: a possible mechanism for the regulation of protein kinase CK2. J Cell Biochem 30:129–136 5. Pinna LA (2002) Protein kinase CK2: a challenge to canons. J Cell Sci 115:32873–33878. doi:10.1242/jcs.00074 6. Litchfield DW (2003) Protein kinase CK2: structure, regulation and role in cellular decisions of life and death. Biochem J 369: 1–15. doi:10.1042/BJ20021469 7. Olsen BB, Niefind K, Issinger OG (2005) Inter-and supramolecular interactions of protein kinase CK2 and their relevance for genome integrity. Genome integrity: Facets and perspectives, vol 1. Springer-Verlag Berl Heidelberg, pp 315–342 8. Meggio F, Pinna LA (2003) One-thousand-and-one substrates for protein kinase CK2? FASEB J 173:349–368. doi:10.1096/ fj.02-0473rev 9. Niefind K, Pu¨tter B, Guerra B, Issinger OG, Schomburg D (1999) GTP plus water mimic ATP in the active site of protein kinase CK2. Nat Struct Biol 6:1100–1103. doi:10.1038/70033 10. Lozeman FJ, Litchfield DW, Pienning C, Takio K, Walsh KA, Krebs ED (1990) Isolation and characterization of human clones encoding the a or a0 subunits of casein kinase II. Biochemistry 29:8436–8447. doi:10.1021/bi00488a034 11. Guerra B, Siemer S, Boldyreff B, Issinger OG (1999) Protein kinase CK2: evidence for a protein kinase CK2b subunit fraction, devoid of the catalytic CK2a subunit, in mouse brain and testicles. FEBS Lett 462:353–357. doi:10.1016/S0014-5793(99)01553-7 12. Xu X, Toselli PA, Russell LD, Seldin DC (1999) Globozoospermia in mice lacking the casein kinase II a0 subunit. Nat Genet 23:118–121. doi:10.1038/12729 13. Lou DY, Dominguez I, Toselli P, Landesman-Bollag E, O’Brien C, Seldin DC (2008) The alpha catalytic subunit of protein kinase CK2 is required for mouse embryonic development. Mol Cell Biol 28:131–139. doi:10.1128/MCB.01119-07 14. Buchou T, Vernet M, Blond O, Jensen HH, Pointu H, Olsen BB et al (2003) Disruption of the regulatory beta subunit of protein kinase CK2 in mice leads to a cell-autonomous defect and early
123
20.
21.
22.
23.
24.
25.
26.
27.
28.
29. 30.
31.
embryonic lethality. Mol Cell Biol 23:908–915. doi:10.1128/ MCB.23.3.908-915.2003 Mu¨nstermann U, Fritz G, Seitz G, Lu YP, Schneider HR, Issinger OG (1990) Casein kinase II is elevated in solid human tumors and rapidly proliferating non-neoplastic tissue. Eur J Biochem 189:251–251. doi:10.1111/j.1432-1033.1990.tb15484.x Ahmed K, Davis AT, Wang H, Faust RA, Yu S, Tawfic S (2000) Significance of protein kinase CK2 nuclear signaling in neoplasia. J Cell Biochem Suppl 35:130–135. doi:10.1002/10974644(2000)79:35+3.0.CO;2-N Tawfic S, Yu S, Wang H, Faust RA, Davis A, Ahmed K (2001) Protein kinase CK2 signal in neoplasia. Histol Histopathol 16:573–582 Pagano MA, Cesaro L, Meggio F, Pinna LA (2006) Protein kinase CK2: a newcomer in the ‘druggable kinome’. Biochem Soc Trans 34:1303–1306. doi:10.1042/BST0341303 Niefind K, Guerra B, Pinna LA, Issinger OG, Schomburg D (1998) Crystal structure of the catalytic subunit of protein kinase ˚ resolution. EMBO J 17:2451–2462. CK2 from Zea mays at 2.1 A doi:10.1093/emboj/17.9.2451 Niefind K, Guerra B, Ermakowa I, Issinger OG (2001) Crystal structure of human protein kinase CK2: insights into basic properties of the CK2 holoenzyme. EMBO J 20:5320–5331. doi: 10.1093/emboj/20.19.5320 Olsen BB, Boldyreff B, Niefind K, Issinger OG (2006) Purification and characterization of the CK2a0 -based holoenzyme, an isozyme of CK2a: a comparative analysis. Protein Expr Purif 47:651–661. doi:10.1016/j.pep. 2005.12.001 Ermakova I, Boldyreff B, Issinger OG, Niefind K (2003) Crystal structure of a C-terminal deletion mutant of human protein kinase CK2 catalytic subunit. J Mol Biol 330:925–934. doi:10.1016/ S0022-2836(03)00638-7 Grankowski N, Boldyreff B, Issinger OG (1991) Isolation and characterization of recombinant human casein kinase II subunits a and b from bacteria. Eur J Biochem 198:25–30. doi:10.1111/ j.1432-1033.1991.tb15982.x Boldyreff B, Meggio F, Pinna LA, Issinger OG (1993) Reconstitution of normal and hyperactivated forms of casein kinase-2 by variably mutated beta-subunits. Biochemistry 32:12672– 12677. doi:10.1021/bi00210a016 Guerra B, Niefind K, Ermarkova I, Issinger OG (2001) Characterization of CK2 holoenzyme variants with regard to crystallization. Mol Cell Biochem 227:3–11. doi:10.1023/A: 1013184000557 Pagano MA, Meggio F, Ruzzene M, Andrzejewska M, Kazimierczuk Z, Pinna LA (2004) 2-Dimethylamino-4,5,6,7tetrabromo-1H-benzimidazole: a novel powerful and selective inhibitor of protein kinase CK2. Biochem Biophys Res Commun 321:1040–1044. doi:10.1016/j.bbrc.2004.07.067 Pagano MA, Andrzejewska M, Ruzzene M, Sarno S, Cesar L, Bain J et al (2004) Optimization of protein kinase CK2 inhibitors derived from 4,5,6,7-tetrabromobenzimidazole. J Med Chem 47:6239–6247. doi:10.1021/jm049854a Meggio F, Pagano MA, Moro S, Zagotto G, Ruzzene M, Sarno S et al (2004) Inhibition of protein kinase CK2 by condensed polyphenolic derivatives. An in vitro and in vivo study. Biochemistry 43:12931–12936. doi:10.1021/bi048999g Knight ZA, Shokat KM (2005) Features of selective kinase inhibitors. Chem Biol 12:621–637. doi:10.1016/j.chembiol.2005.04.011 Zien P, Duncan JS, Skierski J, Bretner M, Litchfield DW, Shugar D (2005) Tetrabromobenzotriazole (TBBt) and tetrabromobenzimidazole (TBBz) as selective inhibitors of protein kinase CK2: evaluation of their effects on cells and different molecular forms of CK2. Biochim Biophys Acta 1754:271–280 Sarno S, de Moliner E, Ruzzene M, Pagano MA, Battistutta R, Bain J et al (2003) Biochemical and three-dimensional-structural
Mol Cell Biochem
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
study of the specific inhibition of protein kinase CK2 by [5-oxo– 5, 6-dihydroindolo-(1, 2-a)quinazolin-7-yl]acetic acid (IQA). Biochem J 374:639–646. doi:10.1042/BJ20030674 Sarno S, Salvi M, Battistutta R, Zanotti G, Pinna LA (2005) Features and potentials of ATP-site directed CK2 inhibitors. Biochim Biophys Acta 1754:263–270 Sarno S, Ruzzene M, Frascella P, Pagano MA, Meggio F, Zambon A et al (2005) Development and exploitation of CK2 inhibitors. Mol Cell Biochem 274:69–76. doi:10.1007/s11010-005-3079-z Nie Z, Perretta C, Erickson P, Margosiak S, Lu J, Averill A et al (2008) Structure-based design and synthesis of novel macrocyclic pyrazolo[1, 5a][1, 3, 5]triazine compounds as potent inhibitors of protein kinase CK2 and their anticancer activities. Bioorg Med Chem Lett 18:619–623. doi:10.1016/j.bmcl.2007.11.074 Persson T, Yde CW, Rasmussen JE, Rasmussen TL, Guerra B, Issinger OG et al (2007) Pyrazole carboxamides and carboxylic acids as protein kinase inhibitors in aberrant eukaryotic signal transduction: induction of growth arrest in MCF-7 cancer cells. Org Biomol Chem 5:3963–3970. doi:10.1039/b711279c Golub AG, Yakovenko OY, Prykhod’ko AO, Lukashov SS, Bdzhola VG, Yarmoluk SM (2007) Evaluation of 4,5,6,7-tetrahalogeno-1H-isoindole-1, 3(2H)-diones as inhibitors of human protein kinase CK2. Biochim Biophys Acta 1784:143–149 Filhol O, Cochet C (2007) Structure-based design of small peptide inhibitors of protein kinase CK2 subunit interaction. Biochem J 408:363–373. doi:10.1042/BJ20070825 Nie Z, Perreta C, Erickson P, Margosia S, Almassy R, Lu J et al (2007) Structure-based design, synthesis, and study of pyrazolo[1, 5-a][1, 3, 5]triazine derivatives as potent inhibitors of protein kinase CK2. Bioorg Med Chem Lett 17:4191–4195. doi: 10.1016/j.bmcl.2007.05.041 Raaf J, Klopffleisch K, Issinger OG, Niefind K (2008) The catalytic subunit of human protein kinase CK2 structurally deviates from its maize homologue in complex with the nucleotide competitive inhibitor emodin. J Mol Biol. doi:10.1016/j.jmb. 2008.01.008 Meggio F, Boldyreff B, Marin O, Marchiori F, Perich JW, Issinger OG et al (1992) The effect of polylysine on casein kinase-2 activity is influenced by both the structure of the protein/peptide substrates and the subunit composition of the enzyme. Eur J Biochem 205:939–945. doi:10.1111/j.1432-1033.1992.tb16860.x Romero-Oliva F, Jacob G, Allende JE (2003) Dual effect of lysine-rich polypeptides on the activity of protein kinase CK2. J Cell Biochem 89:348–355. doi:10.1002/jcb.10493
42. Arrigoni G, Marin O, Pagano MA, Settimo L, Paolin B, Meggio F et al (2004) Phosphorylation of calmodulin fragments by protein kinase CK2. Mechanistic aspects and structural consequences. Biochemistry 43:12788–12798. doi:10.1021/bi049365c 43. Dobrowolska G, Lozeman FJ, Donxia L, Krebs EG (1999) CK2, a protein kinase of the next millenium. Mol Cell Biochem 191:3– 12. doi:10.1023/A:1006882910351 44. Issinger OG, Brockel C, Boldyreff B, Pelton JT (1992) Characterization of the a and b subunits of casein kinase 2 by Far-UV CD spectroscopy. Biochemistry 31:6098–6103. doi:10.1021/ bi00141a020 45. Glover CV (1986) A filamentous form of drosophila casein kinase II. J Biol Chem 261:14349–14354 46. Valero E, DeBonis S, Filhol O, Wade H, Langowski J, Chambaz EM et al (1995) Quartenary structure of casein kinase 2: Characterization of multiple oligomeric states and relation with its catalytic activity. J Biol Chem 270:8345–8352. doi:10.1074/ jbc.270.14.8345 47. Pagano MA, Sarno S, Poletto G, Cozza G, Pinna LA, Meggio F (2005) Autophosphorylation at the regulatory b-subunit reflects the supramolecular organization of protein kinase CK2. Mol Cell Biochem 274:23–29. doi:10.1007/s11010-005-3116-y 48. Litchfield DW, Lozeman FJ, Cicirelle MF, Harrylock M, Ericsson LH, Piening CJ et al (1991) Phosphorylation of the b subunit of casein kinase II in human A431 cells. Identification of the autophosphorylation site and a site phosphorylated by p34cdc2. J Biol Chem 266:20380–20389 49. Boldyreff B, James P, Staudenmann W, Issinger OG (1993) Ser-2 is the autophosphorylation site in the b subunit from bicistronically expressed human casein kinase-2 and from native rat liver casein kinase-2 b. Eur J Biochem 218:515–521. doi:10.1111/ j.1432-1033.1993.tb18404.x 50. Niefind K, Issinger OG (2005) Primary and secondary interactions between CK2a and CK2b lead to ring-like structures in the crystals of the CK2 holoenzyme. Mol Cell Biochem 274:3–14. doi:10.1007/s11010-005-3114-0 51. Poole A, Poore T, Bandhakavi S, McCann RO, Hanna DE, Glover CV (2005) A global view of CK2 function and regulation. Mol Cell Biochem 274:163–170. doi:10.1007/s11010-005-2945-z 52. Wisconsin Package Version 10.3, Accelrys Inc., San Diego, CA, USA
123
