Eur. J. Biochem. 195, 377-383 (1991)

001429S6910004SD

CJ FEBS 1991

Kinetic characterisation of the enzymatic activity of the eEF-2-specific Ca2 + - and calmodulin-dependent proteinkinase I11 purified from rabbit reticulocytes Anders NILSSON, Ulf CARLBERG and Odd NYGARD Department of Cell Biology, Stockholm University, Sweden

(Received June 25/0ctober 11,1990) - EJB 90 0728

The Ca2+- and calmodulin-dependent protein kinase 111, which specifically phosphorylates the eukaryotic elongation factor 2 (eEF-2), has been purified to apparent homogeneity from the post-ribosomal fraction of rabbit reticulocytes by an efficient four-step method. The method results in a more than 4000-fold purification of the enzyme. SDS-gel electrophoresis showed that the purified kinase contained only one polypeptide with the apparent molecular mass of 90 kDa. The kinase activity was associated with the 90-kDa protein as shown by analyzing the phosphorylating activity of SDS gel electrophoretically purified protein electroblotted to nitrocellulose mem- branes. The purified kinase was dependent on Ca2+, Mg2+ and calmodulin for activity. Kinetic analysis of the phosphorylation reaction indicates that the turnover number of the kinase was approximately 1 s-'. The K, for the two substrates ATP and eEF-2 was calculated to be approximately 100 pM and 10 pM, respectively. The activity of the kinase was competitively inhibited by CAMP. The inhibition constant Ki (0.5 mM) was found to be in the same order of magnitude as that calculated for the competitive product inhibition caused by ADP. GTP was ten-times less efficient as competitor, indicating that the kinase had a preference for adenosine nucleotides. Phosphorylation of eEF-2 did not interfere with the diphtheria-toxin-catalysed ADP-ribosylation of the factor nor did ADP-ribosylation inhibit phosphorylation.

Reversible protein phosphorylation is a common and im- portant mechanism for regulation of biochemical processes such as cell motility, release of neurotransmitters, glycogen metabolism, gene activity and protein synthesis [l]. The ac- tivity of the kinases responsible for these protein phos- phorylations are regulated by different mechanisms. The best characterised protein kinases are regulated by cyclic nucleotides such as cAMP and cGMP, or by Ca2+ 12, 31. In the latter case enzyme activity is also dependent on a phospholipid or the calcium-binding protein calmodulin

Several Ca2+/CaM-dependent protein kinases are known [2]. This group of enzymes includes phosphorylase kinase, myosin light chain kinase, Ca2 +/CaM-dependent protein kinases I, I1 and 111. Ca2 +/CaM protein kinase I11 is a recently discovered kinase that specifically phosphorylates eEF-2 [4 - lo], the eukaryotic elongation factor catalyzing the translo- cation of peptidyl-tRNA from the ribosomal A-site to the P- site in the protein synthesis elongation cycle [l l] .

The extent of eEF-2 phosphorylation in vzvo is increased after treatment of cells with drugs that raise the intracellular level of Ca2+, such as thrombin, histamine and veratridine [lo, 121. Phosphorylation of eEF-2 both in vivo and in vitro is correlated with an inhibition of the protein synthesis as a result of a reduced affinity of the modified factor for ribosomes in the pre-translocation phase of the elongation cycle [6, 8, 13,

( c a w ~ 3 1 .

Correspondence to 0. Nygird, Department of Cell Biology, Bi- ology E5, Stockholm University, S-106 91 Stockholm, Sweden

Abbreviations. CaM, calmodulin; Ca2 +/CaM protein kinase 111, calcium- and calmodulin-dependent protein kinase 111 ; eEF-2, eukaryotic elongation factor 2.

141. Substances know to increase the intracellular level of cAMP such as forskolin and nerve growth factor (NGF) re- duce the phosphorylation of eEF-2 [15] and increase the rate of protein synthesis [16]. These results suggest that variations in the extent of phosphorylation of eEF-2 plays an important role in the regulation of gene expression in eukaryotes, by controlling the rate of elongation of nascent polypeptide chains. To provide a regulatory mechanism that could account for rapid changes in the translational efficiency, the phos- phorylation has to be reversible. This requirement is appar- ently fulfilled as recent observations show that eEF-2 is dephosphorylated in the presence of a type 2A phosphatase [17]. An inhibition of the phosphatase activity by okadaic acid results in an increased phosphorylation of eEF-2 and a reduced translational activity [13]. These observations may indicate that regulation of both the kinase and the phospha- tase activities could be involved in controlling the rate of protein synthesis.

In this report, we describe the isolation of homogeneous Ca2 +/CaM protein kinase 111 from rabbit reticulocytes. The K,,, for the two substrates ATP and eEF-2 as well as the turnover number of the kinase was determined by kinetic methods. The inhibition of the kinase caused by ADP and cAMP was kinetically characterized.

MATERIALS AND METHODS

Chemical

[Y-~~P]ATP and [I4C]NAD+ were from Amersham Inter- national (UK). GTP, ADP, ATP, and cAMP were from Sigma Chemicals Co. (St Louis, MO, USA). CaM was from

378

Boehringer (Mannheim, FRG). Diphtheria toxin was a gift from Dr M. Tiru (National Bacteriological Laboratories, Solna, Sweden). DEAE-cellulose (DE-52) was from Whatman BioSystems (Maidstone, UK). Nitrocellulose membranes were from Sartorius (Gottingen, FRG). The Mono-Q column was from Pharmacia (Uppsala, Sweden). eEF-2 was purified from rat liver as previously described [18]. Protein concen- trations were determined according to the method of Bradford [19] using bovine serum albumin as standard.

Purification of' the kinusr

Rabbits reticulocytes were prepared as previously de- scribed [20]. The isolated cells were lysed by addition of an equal volume cold distilled water and centrifuged at 15 000 x g,, for 15 min. The supernatant was centrifuged for an additional 180 rnin at 150000 x g,".

The post-ribosomal supernatant was fractionated with ammonium sulphate between 40 - 60% saturation at 0 "C. The precipitate was collected by centrifugation for 10 rnin at 10000 x g,,, dissolved in buffer A (20 mM Tris/HCl pH 7.6, 15 mM 2-mercaptoethanol, 0.1 mM EDTA and lo%, by vol., glycerol) containing 100 mM KC1 and dialysed against the same buffer overnight. The dialysed material was applied onto a 40-ml DEAE-cellulose column equilibrated in buffer A con- taining 100 mM KC1. The column was washed with the same buffer containing 225 mM KC1 and the bound kinase was eluted from the column with buffer A containing 400 mM KCl. Fractions containing protein were pooled, concentrated by ultrafiltration and dialysed against buffer B (20 mM Tris/ HCl, pH 7.6, 0.1 niM EDTA, 5 mM thiodiethanol).

The kinase was further purified using an FPLC system equipped with a Mono-Q column equilibrated in buffer B. The partially purified kinase was applied to the column at a flow rate of 1 mlimin. The column was washed with a steep gradient of 0 - 350 mM KC1 in buffer B and the bound kinase eluted with a linear salt gradient of 350-550 mM KCl in buffer B.

The fractions containing the kinase activity were pooled, concentrated by ultrafiltration and dialysed against buffer A containing 225 mM KC1. The dialysed material was applied to a linear 10-25% (massivol.) sucrose gradient in buffer A containing 225 mM KC1. The gradients were centrifuged for 23 h at 256000 x g,,, fractionated [21] and the kinase activity was determined. The active fractions were frozen in liquid nitrogen and stored at -80°C.

Determination of kinuse activity

The phosphorylation reactions contained, if not otherwise indicated, 100 mM KCl, 20 mM Tris/HCl pH 7.6, 15 mM 2- mercaptoethanol, 10 mM MgC12, 1 .0 mM EGTA, 1.5 mM Ca2+, 0.07mM EDTA, 7% (by vol.) glycerol, 0.25 mM [y-32P]ATP (specific activity 280 Ci/mol), 120 pmol CaM, 50 pmol eEF-2 and 1 pmol kinase. The reaction mixtures, 25 pl, were incubated at 30 C for 10 rnin if not otherwise indicated. The reaction was stopped by addition of electro- phoresis sample buffer [22] and the samples were heated for 2 min at 90 'C and analyzed by gel electrophoresis.

Alternatively the phosphorylation reaction was stopped by addition of 1 ml cold 10% (massivol.) trichloroacetic acid. After 10 rnin at 0 C the precipitates were collected on nitrocel- lulose filters, washed with 5% (massivol) trichloroaccetic acid, dried and counted in a scintillation counter.

0.3 c A 4 0.6

h 1 '

Fraction number

B Fraction number

M, x10-3 5 10 I I

15 I

92.5-

66.2-

45-

31 -

21.5- 14.4-

Fig. 1. Purification ofthe eEF-2-specific kinase by chromutography on Mono-Q columns. Partly purified kinase obtained from the DEAE- cellulose column was applied to a Mono-Q column as described in Materials and Methods. The bound protein was eluted using a linear KCI gradient (-----) and consecutive I-ml fractions were collected. (A) Chromatogram. (B) the kinase activity of the collected fractions were determined as described in Materials and Methods and the phos- phorylation products were analyzed by SDS gel electrophoresis and autoradiography

Gel electrophoresis

The phosphorylation products were analyzed by SDS gel electrophoresis using a linear 7 - 15% (mass/vol.) polyacryl- amide gradient gel 1221. The gels were stained with Coomassie brilliant blue. destaincd. dried and exposed at -80°C to X-ray films using an intensifying screen.

For analysis of kinase activity associated with the purified kinase isolated by SDS-gel electrophoresis, the kinase was electro-transferred from the gels to nitrocellulose membranes [23]. The membranes were stained using Ponceau S in 10% (by vol.) acetic acid and the membrane containing the individual blotted protein bands excised, cut into small fragments, trans- ferred to incubation tubes and incubated in a buffer containing 100 mM KC1, 20 mM Tris/HCl (pH 7.6), 15 mM 2- mercaptoethanol, 10 mg MgC12, 1.0 mM EGTA, 1.5 mM Ca2+, 0.07 mM EDTA, 7% (by vol.) glycerol, 0.1% (by vol.) Triton X-100 and 120pmol CaM. After 2 h at 0"C, [y- 32P]ATP and eEF-2 was added as described above and the incubations (final volumes 150 pl) continued for 30 min at

379

A B C D E ~ ~ ~ 1 0 - 3 1 2 1 2 3 4 1 2 3 4

1 1 2

92.5-

66.2-

45-

31 -

21.5- 14.4-

Fig. 2. Effects ofCaZ+, CaM and Mg2+ on the eEF-2-specific kinase. (A) Effect of Ca2+ on the electrophoretic mobility of the kinase obtained from the Mono-Q column. The material was analyzed by gel electrophoresis and stained with Coomassie brilliant blue as described in Materials and Methods. Migration in the presence of 1.5 mM CaCI2 (lane 1) or in the absence of added Ca2+ (lane 2). The protcin band coinciding with the 92.5-kDa marker corresponds to added eEF-2. (B) Effect of CaM and Ca2+ on the catalytic activity of the purified kinase. Phosphorylation and the analysis by SDS gel electrophoresis and autoradiography was as described in Materials and Methods. Control without CaM and CaZ+ (lane I), 30 pmol CaM without Ca2+ added (lane 2), addition of both 1.5 mM CaClz and 30 pmol CaM (lane 3) or 1.5 mM Ca2+ alone (lane 4). (C) Mg2+ requirement for kinase activity. The kinase activity at various concentrations of Mg2+ was analyzed as described above. The concentrations used were, 0 (lane I), 0.5 (lane 2), 1.0 (lane 3), and 10 mM (lane 4). (D) Kinase activity of the 90-kDa polypeptide. The protein was isolated by SDS gel electrophoresis and electroblotted to nitrocellulose membranes as described in Materials and Methods. The kinase activity of the membrane-bound 90-kDa polypeptide was analyzed as described in Materials and Methods and the phosphorylation reaction analyzed by SDS gel electrophoresis and autoradiography. (E) Purity of the kinase preparation. Purified kinase, 1 pg (lane 1) and 5 pg (lane 2), was analyzed by SDS gel electrophoresis and stained with Coomassie brilliant blue as described in Materials and Methods

30°C. The incubation mixtures were analyzed by SDS gel Fraction number

I I electrophoresis and autoradiography as described above. ~ ~ ~ 1 0 - 3 5 10

RESULTS AND DISCUSSION

The eukaryotic elongation factor eEF-2 is specifically phosphorylated by the Ca2 + /CaM-dependent protein kinase I11 [4 - lo]. For detailed studies of the phosphorylation reaction and for characterization of the structural and func- tional properties of the enzyme, it is necessary to obtain the kinase free from contaminating proteins. For these purposes

the post-ribosomal supernatant of rabbit reticulocytes. The supernatants were fractioned with ammonium sulphate be- tween 40-60% saturation at 0°C and the precipitated ma- terial applied to a DEAE-cellulose column. The bound kinase was eluted from the column by increasing the KC1 concen- tration from 0.25 M to 0.40 M. After concentration and dialy- sis the partly purified kinase was applied to a Mono-Q column

and with a gradient from 0'35 to 0'55 KC' Fig. 3. The use of sucrosegradient centrifugation in thepurification of (Fig. 1 A). AS Seen in Fig. 1 B, the kinase eluted from the the eEF-2-specific kinase. Partly purified kinase was applied to 10- column at approximately 0.44 M KCl. This is in agreement 25% (mass/vol.) sucrose gradients, centrifuged and fractionated as with previous observations [8]. As gel electrophoresis showed dcscribed in Materials and Methods. The kinase activity was deter- that the pooled active fractions contained several protein mined and analyzed by gel electrophoresis and autoradiography as bands (Fig. 2A, lane 2), the kinase was further purified by described in Materials and Methods sucrose gradient centrifugation. Under the separation con- ditions used the kinase sedimented surprisingly close to the bottom of the gradients (Fig. 3). Calculation of the sedimen- tation coefficient [24] indicated that the enzyme sedimented as an approximately 8s particle, suggesting a mass of 0.3 MDa

92.5-

66.2-

we have purified the kinase to apparent homogeneity from 45-

31 -

21.5- 14.4-

[25]. Previous determinations of the size of the kinase using gel filtration has suggested that the kinase activity was associated with a 140-kDa molecule [ 5 , 261. Analysis of the protein com-

3 80

Table 1. Purificution uf the eEF-2-specific protein kinasc One unit of enzyme activity corresponds to the amount of enzyme required to phosphorylate 50 pmol eEF-2 in 30 min at 30°C

Purification step Total protein Specific activity x Total activity Purification Yield ~~~ ~

m& units/pg units

Ammonium sulphate 40 -60% 2300 0.05 115 DEAE-cellulose 21.4 3 64

Gradient centrifugation 0.2 200 40 Mono-Q 2.4 18.5 44

-fold Yo 1 100

60 56 370 39

4000 35

N r; 0.8

P

0 I I

0 50 100 150 200 250 300 I I I

0 50 100 150 200 250 300 Calmodulin (pmol)

7'

I

10 20 30 40 50 0"

0 Incubation time (min )

4

L I 1 1

0 0 0.2 0.4 0.6 0.8

Kinase (pmol)

Fig. 4. 0ptirni:ation of' the phosphorylation reaction. Phosphorylation of eEF-2 was determined by incubation for 30 min in the presence of 0.5 mM [y-32P]ATP as describcd in Materials and Methods. The ex- tent of phosphorylation was analyzed by precipitation with trichloroacetic acid was as described in Materials and Methods. Influ- ence of calmodulin (A), incubation time (B) and kinase concentration (C) on the extent of eEF-2 phosphorylation

position of the gradient purified kinase by SDS gel electrophoresis showed that the active fractions contained one single polypeptide with an apparent molecular mass of 90 kDa (Fig. 2E).

To verify that the kinase activity was associated with the 90-kDa component and not with a minor non-detectable con- taminating protein having a high catalytic activity, SDS gels containing 1 pg of the purified material was blotted to nitro- cellulose membranes. After rapid staining, the membrane area containing the 90-kDa protein was excised and analyzed for kinase activity. As seen in Fig. 2D, the membrane-bound 90- kDa polypeptide was able to phosphorylate eEF-2. The result shows that the kinase activity was associated with the 90- kDa protein. Apparently a fraction of the kinase was able to recover the active conformation despite SDS denaturation and electrophoretic immobilisation on nitrocellulose membranes. The large particle size of the kinase observed by sucrose gradi- ent centrifugation suggests that, in the sucrose gradients, the enzyme migrated as a particle composed of three or four copies of the 90-kDa protein. Such aggregation may result from hydrophobic interactions induced by the high salt concen- trations used to remove contaminating proteins. The purifi- cation method resulted in a 4000-fold purification of the kinase with recovery of approximately 35% of the initial ac- tivity (Table 1).

To ascertain that the purified kinase was identical to the Ca2+/CaM-dependent protein kinase Ill, the effect of Ca2+ and CaM on the eEF-2 phosphorylation was analyzed. As seen in Fig. 2 B, the kinase was absolutely dependent on both Ca2+ and CaM for activity. In the presence of Ca2+, the kinase activity was almost linearly dependent on CaM up to a concentration of 2.4 pM (Fig. 4A). However, the CaM dependence was only observed with the purified kinase (not illustrated). Analysis of the protein composition of the par- tially purified kinase isolated from the Mono-Q column using SDS gel electrophoresis showed that these kinase preparations contained additional low-molecular-mass components of 16 - 28 kDa (Fig. 2A). The fastest migrating band showed a Ca2+- induced switch in the apparent mass (Fig. 2A). As this behav- iour is typical for calmodulin [27], the result suggests that this band was derived from calmodulin. Hence, the lack of CaM- dependence was due to the presence of CaM in the cruder preparations.

Many kinases require divalent cations such as Mn2 + or Mg2+ for binding of the nucleotide and thus for activity [28]. It was therefore not surprising that no Ca2+/CaM protein kinase 111 activity could be detected in the absence of MgZt (Fig. 2C). Furthermore, Mg2'. could not replace Ca2+ in the phosphorylation reaction as Ca2+ was required also in the presence of Mg2+. The concentration of Mg2+ required for activity was dependent on the ATP concentration used (not illustrated), indicating that the Mg2 + was needed for proper binding of the nucleotide to the kinase.

The optimal conditions for phosphorylation of eEF-2 were established with respect to incubation time and the concen- tration of kinase. As seen in Fig. 4B, the extent of eEF-2 phosphorylation was proportional to the incubation time up

381

150 - A = - E, ; 100 -

6 U

C P

0

0 50 100 150 200 250 eEF-2 added (pmol)

0.2

0.0

-

I a 50.0 100.0 150.0 0.0

V/[S] (s-1 M.110-3,

Fig. 5. Kinetic determination of the enzyme activity. Phosphorylation of eEF-2 in the presence of 0.5 mM [Y-~’P]ATP was as described in Materials and Methods. The extent of phosphorylation was deter- mined by precipitation with trichloroacetic acid as described in Ma- terials and Methods. (A) Phosphorylation as a function of increasing concentrations of eEF-2. (B) Linearization of the data in (A) accord- ing to Eadie and Hofstee [40, 411

Table 2. Kinetic constants of the phosphorylation reaction Kapp is the apparent K,,, in the presence of inhibitor

Variable Inhibitor I [I] k,,, Kapp K m Ki substrate

mM s - l mM

1.0 - 0.030 -

1.4 - 0.010 - 0.9 - 0.076 -

- - CaM eEF-2 - -

ATP ATP CAMP 1 0.8 0.25 - 0.43 ATP CAMP 2.5 0.9 0.42 - 0.55 ATP ADP 1 0.8 0.22 - 0.54 ATP ADP 2.5 0.8 0.30 - 0.86 ATP GTP 1 0.9 0.09 - 5.1 ATP GTP 2.5 0.8 0.17 - 2.0

- -

to 10 min. After a 30-min incubation the reaction was almost complete, resulting in a stoichiometric phosphorylation of eEF-2. The extent of phosphorylation was also proportional to the kinase concentration up to 26 nM (Fig. 4C). At higher concentrations, the phosphorylation was complete resulting in an incorporation of approximately one mole phosphate/ mole eEF-2. Previous observations using two-dimensional isoelectric focusing/SDS gel electrophoresis have shown that eEF-2 phosphorylated under these conditions is homogeneous [14]. These results suggest that eEF-2 contains one primary site accessible for phosphorylation by Ca2+/CaM protein kinase 111. This site is equivalent to the site previously located

1.0 t n

- o’8A\ 0.6

’0’4i:._ 0.2 0.0

0.0 2.0 4.0 6.0 8.0 10.0 12.0 v/[S] (s-l M-llO-3)

Fig. 6. Effect of increusing ATP concentration on the rate of eEF-2 phosphorylation. Phosphorylation in the presence of 9.2 pM eEF-2 was as described in Materials and Methods. The extent of eEF-2 phosphorylation was determined by precipitation with trichloroacetic acid as described in Materials and Methods. The experimental data were linearized according t? Eadie and Hofstee [40,41]. (A) Phos- phorylation as a function of increasing ATP concentration without (0 ) or in the presence of 1 mM (A) and 2.5 mM (B) CAMP. (B) Phosphorylation at increasing ATP concentrations in the presence of 1 mM (A) and 2.5 mM (B) ADP. (C) Phosphorylation in the pres- ence of 1 mM (A) and 2.5 niM ( H ) GDP

to the N-terminal amino acid residues 51 -60 in eEF-2 [6] (not illustrated). Additional phosphorylation sites on eEF-2 have been reported [26]. However, under the assay conditions established here, the kinase was apparently less effective in modifying these putative additional sites.

The turnover number of the kinase was analyzed by in- creasing the substrate concentrations, i. e. the concentrations of eEF-2 and ATP. In the experiments with increasing concen- trations of eEF-2 shown in Fig. 5A, the concentration of ATP was kept high and saturating. From the linearized data in Fig. 5 B the K , and the turnover number (kcat) of the reaction could be calculated. As seen in Table 2, the turnover number

382

A B

~ ~ ~ 1 0 - 3 1 2 3 4 1 2 3 4

92.5-

66.2-

45-

31-

21.5-

Fig. 7. Effect of’ ADP-rihusylution and c A M P on the efficiency of the phosphorylation reaction. eEF-2. was phosphorylated and analyzed by gel electrophoresis and autoradiography as described in Materials and Methods. (A) Effect of ADP-ribosylation on the phosphorylation of eEF-2. ADP-ribosylation [42] with unlabelled NAD+ followed by phosphorylation in the presence of [y-”P]ATP (lane 1). Phosphoryla- tion using unlabelled ATP followed by ADP-ribosylation in the pres- ence of [14C]NAD (lane 2). Phosphorylation with [y-3zP]ATP (lane 3). ADP-ribosylation with [14C]NADt (lane 4). (B) Inhibition of the kinase activity at increasing concentrations of CAMP. The concentrations werc 0 (lanc I) , 0.05 (lane 2), 0.5 (lane 3) and 5 mM (lane 4)

corresponded to slightly more than 1 mol eEF-2 phosphor- ylated . s- ’ . mol kinase- ’. The K, for eEF-2 was calculated to be approximately 10 pM (Table 2). The influence of increasing ATP concentrations on the eEF-2 phosphorylation was analyzed in the presence of high concentrations of eEF-2. From the linearized experiments in Fig. 6A, the K, for ATP could be calculated to approximately 0.2 mM (Table 2). This experiment also provided a possibility for an independent calculation of the turnover number of the kinase. Also in this case the k,,, was found to be approximately one, thus confirming the previous estimation. The relatively high K, values for the two substrates, eEF-2 and ATP, indicate that the enzyme had moderate affinity for both substrates. However, as the intracellular concentration of ATP is two orders of magnitude higher that the K,, the kinase is likely to be satu- rated with ATP in vivo [29]. Calculations of the intracellular concentration of eEF-2 based on previous observations [30] indicate that the concentration of eEF-2 is in the same order of magnitude as the K,,, for eEF-2. Hence, the rate-limiting step for the phosphorylation of eEF-2 in vivo is the formation of the kinase . ATP . eEF-2 transition complex from eEF-2 and the binary kinase . ATP complex.

EF-2 contains a modified histidine, called diphthamide, in position 715 [31]. This amino acid is ADP-ribosylated in the presence of NAD ’ and an endogenous ADP-ribosyl transfer- ase or diphtheria toxin [32, 331. This modification reduces the affinity of the factor for the ribosome [34]. Similar effects has also been found after phosphorylation and after tryptic cleavage of the N-terminal region of eEF-2 [14, 351, suggesting that both the N-terminus and the diphthamide region are important for the attachment of the factor to the ribosome. A direct influence of the diphthamide region on the N-terminal

domain is also indicated by the observed reduction in the rate of guanosine nucleotide exchange after ADP-ribosylation [36]. To see if ADP-ribosylation interfered with the phosphoryla- tion reaction, ADP-ribosylated eEF-2 was used as a substrate for the kinase. As seen in Fig. 7 A, the extent of phosphoryla- tion was not reduced by the ADP-ribosylation. Moreover, phosphorylation of eEF-2 did not reduce the diphtheria-toxin- catalysed ADP-ribosylation of the factor. These results show that the two types of modifications were not mutually exclud- ing. Obviously, the factor can contain both the phosphate and the ADP-ribosyl group in vituo, although eEF-2 with these double modifications has not been observed in vivo [37].

Phosphorylation of eEF-2 in vivo is reduced by growth factors and forskolin [15]. As these two substances are known to increase the intracellular level of CAMP, these observations indicate that the extent of phosphorylation of eEF-2 and hence the rate of elongation is under the control of CAMP [15]. High concentrations of CAMP have also been shown to stimulate protein synthesis and reduce the extent of eEF-2 phosphoryla- tion in rabbit reticulocyte lysates, thus supporting the regulat- ory role of cAMP [16]. We were interested to see if this effect could be reproduced with the purified kinase or if the effect was due to additional components in the crude lysate. As seen in Fig. 7B, cAMP was able to inhibit the activity of the purified kinase. The inhibition was even slightly higher than that previously reported for the lysate [16], indicating that the effect was not dependent on additional factors. The inhibition must therefore be due to a direct modulation of the kinase activity or a competition between ATP and cAMP for the nucleotide binding site on the kinase. The latter possibility was investigated using kinetic analysis of the phosphorylation reaction. As seen in Fig. 6A, the effect of CAMP was counter- acted by increasing concentrations of ATP resulting in an unaffected k,,,. The lack of effect on the turnover number suggests that cAMP acted as a competitive inhibitor of the kinase. Due to the competitive nature of the inhibition, high intracellular concentrations of cAMP would be needed for inhibition of the kinase. Thus, it is likely that the competitive inhibition caused by cAMP has little if any significance for the regulation of the kinase activity in vivo. The reported in vivo effects of substances that increase the intracellular level of cAMP [15] must, therefore, be due to an indirect effect on the kinase.

The inhibitory effect of CAMP was compared to the prod- uct inhibition caused by ADP. As seen in Fig. 6B, ADP also caused a competitive product inhibition of the kinase activity. Surprisingly, the inhibitory constant was calculated to be slightly higher than that for CAMP, indicating that cAMP was a more potent inhibitor of the kinase than ADP.

EF-2 is a guanosine-nucleotide-binding protein in which the nucleotide binding domain seems to be localised in the N- terminal part of the factor, in close proximity to the suggested phosphorylation site [6, 13, 381. We have previously shown that guanosine nucleotides induce a conformational change in eEF-2 [39]. When analyzing the influence of GTP on the extent of eEF-2 phosphorylation using autoradiography, we observed a slight reduction in the phosphorylation of eEF-2 in the presence of high concentrations of GTP (not illustrated). This observation prompted us to analyze the effect kinetically. As seen in Fig. 6C, even GTP acted as a competitive inhibitor of the kinase. However, GTP was less efficient than the adenosine nucleotides, suggesting that the kinase had higher affinity for adenosine nucleotides than for guanosine nucleotides. The results also indicate that the conformational effect of GTP on eEF-2 had little, if any, effect on the accessi-

383

bility of the factor for modification by Ca2+/CaM protein kinase 111.

We are indebted to Mrs B. Lundberg for skilful technical assist- ance and to Dr M. Tiru for supplying purified diphtheria toxin. This work was supported by a grant (B-Bu-8463-301) from the Swedish Natural Science Research Council.

REFERENCES

1 . Shenolikar, S. (1988) FASBB J. 2, 2753-2764. 2. Blackshear, P. J., Nairn, A. C. & Kuo, J. F. (1988) FASEB J . 2,

3. Nairn, A. C., Hemmings, H. C. Jr & Greengard, P. (1985) Annu

4. Palfrey, €1. C. (1983) FEBS Lett. 157, 183- 190. 5. Nairn, A. C., Bhagat, B. &Palfrey, H. C. (1985) Proc. Nut1 Acad.

6. Nairn, A. C. & Palfrey, H. C. (1 987) J . Biol. Chem. 626, 17 299 -

7. Ryazanov, A. G. (1987) FEBS Lett. 214, 331 -334. 8. Ryazonov, A. G., Shestakova, E. A. & Natapov, P. G. (1988)

9. Kigoshi, T., Uchida, K. & Morimoto, S. (1989) J . Steroid Bio-

10. Cahill, A. L., Applebaum, R. & Perlman, R. L. (1988) Neurosci.

11. Moldave, K. (1985) Annu. Rev. Biochem. 54, 1109-1149. 12. Mackie, K. P., Nairn, A. C., Hampel, G., Lam, G. & Jaffe, E. A.

13. Redpath, N. T. & Proud, C. G. (1989) Biochem. J . 262, 69-75. 14. Carlberg, U., Nilsson, A. & Nyglrd, 0. (1990) Eur. J . Biochem.

15. Nairn, A. C., Nichols, R. A,, Brady, M. J . &Palfrey, H. C. (1987) J. Biol. Chem. 262,14265 - 14272.

16. Sitikov, A. S., Simonenko, E. A,, Shestakova, E. A,. Ryazanov, A. G. & Ovchinnikov, L. P. (1988) FEBS Lett. 228,327-331.

2957 - 2969.

Rev. Biochem. 54, 931 -976.

Sci. USA 82, 7939 - 7943.

17303.

Nature 334, 170- 173.

chem. 32, 381 -385.

Lett. 84, 345 - 350.

(1989) J . Biol. Chem. 264, 1748-1753.

191,639-645.

17. Gschwendt, M., Kittstein, W., Mieskes, G. & Marks, F. (1989)

18. Nilsson, L. & Nyg%rd, 0. (1984) Biochim. Biophys. Acta 782,

19. Bradford, M. M. (1976) Anal. Biochem. 127, 248-2254, 20. Nyglrd, 0. & Hultin, T. (1975) Chem.-Biol. Interact. 21, 589-

21. Lake, M., Nygbrd, 0. & Naslund, P. H. (1982) Anal. Biochem.

22. Laemmli, N. (1970) Nature 227, 680-685. 23. Bjerrum, 0. J. & Schafer-Nielsen, C. (1986) in Electrophoresis 86

24. McEwen, C. R. (1967) Anal. Biochem. 20, 114-149. 25. Bowen, T. J. (1971) An introduction to ultracentrifugation, p 116,

26. Ryazanov, A. G., Natapov, P. G., Shestakova, E. A., Severin, F.

27. Cohen, P. & Klee, C. B. (eds) (1988) Mol. Aspects Cell Regul5,

28. Mildvan, A. S. (1979) Adv. Enzymol. 49, 103-126. 29. Pettersson, B. & Vallin, I. (1976) Eur. J . Biochem. 62, 383 -390. 30. Nygird, 0. & Nilsson, L. (1984) Bur. J. Biochem. 145, 345-350. 31. Van Ness, B. G., Howard, J. B. & Bodley, J. W. (1978) J . Biol.

32. Pappenheimer, A. M. (1977) Annu. Rev. Biochem. 46, 69 -94. 33. Fenderick, J. L. & Iglewski, W. J. (1989) Proc. Nut1 Acad Sci.

34. Nyglrd, 0. & Nilsson, L. (1990) J . Biol. Chem. 265,6030-6034. 35. Nilsson, L. & Nygbrd, 0. (1988) Eur. J. Biochem. 171,293-299. 36. Burns, G., Abraham, K. A. & Vedeler, A. (1986) FEBS Lett. 208,

37. Marzouki, A., Lavergne, J.-P., Reboud, J.-P. & Reboud, A.-M. (1989) FEBS Lett. 255, 72-76.

38. Kohno, K., Uchida, T., Ohkubo, H., Nakanishi, S., Nakanishi, T., Fukui, T., Ohtsuka, E., Ikehara, M. & Okada, Y . (1986) Proc. Not1 Acad. Sci. USA 83,4978 - 49 82.

39. Nilsson, L. & Nygird, 0. (1985) Eur. J . Biochem. 148,299-304. 40. Eadie, G. S. (1942) J . Biol. Chem. 146, 85-93. 41. Hofstee, B. H. J. (1959) Nature 184, 1296-1298. 42. Nyglrd, 0. & Nilsson, L. (1985) Biochim. Biophys. Actu 824,

FEBS Lett. 257, 357- 360.

49 - 54.

598.

I20,208-210.

(Dunn, M. J., ed.) p 315, Verlag Chemie, Weinheim.

Unwin Brothers, London.

F. & Spirin, A. S. (1988) Biochimie 70, 619 -626.

1-371.

Chem. 253, 8687 -8690.

USA 86, 554- 557.

21 7 - 220.

152- 162.