Molecular and Cellular Biochemistry 127/128: 103-112, 1993. 9 1993 Kluwer Academic Publishers. Printed in the Netherlands.

In vitro substrate specificity of protein tyrosine

kinases Heung-Chin Cheng, 1 Isao Matsuura 2 and Jerry H. Wang 2 2Department of Medical Biochemistry, MRC Group in Signal Transduction, The University of Calgary, Calgary, Alberta T2N 4NI, Canada; 1Russell Grimwade School of Biochemistry, University of Melbourne, Parkville, Victoria 3052, Australia

Abstract Synthetic peptides such as P60 ~rcautophosphorylation site peptides and angiotensin are indiscriminately phosphorylated by protein tyrosine kinases. The observation has led to the general belief that protein tyrosine kinases are highly promiscuous, displaying little in vitro site specificity. In recent years, evidence has been accumulating to indicate that such a belief requires close examination. Synthetic peptides showing high substrate activity for specific groups of protein tyrosine kinases have been obtained. Systematic modification of certain substrate peptides suggests that kinase substrate determinants reside with specific amino acid residues proximal to the target tyrosine. A number of protein kinases have been shown to be regulated by tyrosine phosphorylation at specific sites by highly specific protein tyrosine kinases. These and other selected biochemical studies that contribute to the evolving view of in vitro substrate specificity of protein tyrosine kinases are reviewed. (Mol Cell Biochem 127/128: 103-112, 1993)

Key words: protein tyrosine kinase, phosphorylation site sequence, autophosphorylation

Introduction As is the case with many other aspects of protein phosphorylation research, the central question in protein kinase specificity study is derived from one of Fischer and Krebs' classic papers, where they demonstrated that a single serine residue, ser-14, was phosphorylated in glycogen phosphorylase in the phosphorylase kinase-catalyzed phosphorylase b to a conversion [1]. How does phosphorylase kinase recognize this specific serine residue? If put into a more general form, how do protein kinases recognize their substrate proteins and target specific phosphorylation sites of the substrates? Thirty years later, this question is still a hotly pursued research topic. While a definitive answer to the question is not forthcoming, for many protein kinases residues of the

polypeptide chain proximal to the phosphorylation sites of the protein substrate appears to contain important structural determinants for substrate specificity. The importance of the primary structure of phosphorylation site in the determination of protein kinase substrate activity was originally suggested from the observation that a chymotryptic peptide of glycogen phosphorylase containing ser-14 was a specific substrate of phosphorylate kinase [2]. Peptides derived from protein substrates of other protein kinases have subsequently shown to serve as substrates of the respective protein kinases. Using synthetic peptides, the roles of individual amino acid residues surrounding the phosphorylation site of the peptide can be investigated. Early studies on

Address for offprints: J.H. Wang, University of Calgary School of Medicine, Department of Medical Biochemistry, 3330 Hospital Drive N.W., Calgary, AB T2N 4N1, Canada

104 substrate specificity of cAMP-dependent kinase [3] and phosphorylase kinase [4] using the synthetic peptide approach led to the identification of specific residues adjacent to the phosphorylation sites that are essential or important for the substrate activity of the peptide. Amino acid residues around the phosphorylation site essential for substrate activity were subsequently elucidated for a large number of other protein kinases. Comprehensive listing of the phosphorylation site motifs or consensus sequences, as these essential residues are called, for protein kinases can be found in a number of recent reviews [5-7]. Almost all the phosphorylation motifs determined to date are for protein serine/threonine kinases. In spite of serious attempts by many investigators, progress in elucidating phosphorylation site motifs for protein tyrosine kinases has been frustratingly slow. This is at least in part due to the paucity of information about natural protein substrates of protein tyrosine kinases. The best elucidated phosphorylation sites of protein tyrosine kinases are the autophosphorylation sites, but synthetic peptides derived from autophosphorylation sites are found to be indiscriminately phosphorylated by protein tyrosine kinases [8-12]. Biochemical studies of protein tyrosine kinases frequently use the so-called general tyrosine kinase substrate, such as angiotensin and poly (glu/tyr) copolymer [13, 14]. These practices are based on the assumption that protein tyrosine kinases are highly promiscuous, exhibiting little in vitro protein substrate specificity. During the last few years, the notion that protein tyrosine kinases display little individual specificities has undergone a slow change. In this article, we will briefly review the evolving view governing in vitro substrate specificity of protein tyrosine kinases, and present results supporting the notion that protein tyrosine kinases display individual phosphorylation site specificities. Examples will be provided to show how the new understanding of the kinase specificity facilitates biochemical and cell biological studies of protein tyrosine kinases.

In vitro substrate specificity of protein

tyrosine kinases: An evolving view Most protein tyrosine kinases undergo autophosphorylation, and autophosphorylation sites of several protein tyrosine kinases has been characterized (for review, see [15]). Peptides modelled after amino acid sequences of the autophosphorylation site of P60 Srchave been synthe-

sized and used as protein tyrosine kinase substrates. Since the autophosphorylation site region is highly conserved among src family members, these peptides are equally representative of autophosphorylation site of other src family kinases such as lck, lyn and fyn. Several groups of investigators have carried out kinetic characterizations of the phosphorylation of src autophosphorylation peptide by various protein tyrosine kinases, including many not belonging to the src family. Selected results from these studies are given in Table 1. Although the src autophosphorylation peptides used in different studies varied in length, and some were modified to facilitate the kinase assay or to maintain stable peptide charge (see footnotes to Table 1, for explanation), these structural differences did not appear to alter the kinase substrate activity of the peptides significantly. Thus, comparisons of results between different protein tyrosine kinases are assumed to be valid. From such comparisons, it may be concluded that the src kinase autophosphorylation site peptides, instead of being specific substrates of src family kinases, are indiscriminately phosphorylated by many tyrosine kinases. In fact, the peptides appeared to be better substrates for certain non-src family kinases such as EGF receptor, insulin receptor [16] and especially a spleen tyrosine kinase called PTK-IIB [17] than for src family tyrosine kinases (Table 1). In addition to autophosphorylation site peptides, a number of other synthetic peptides have been found to be phosphorylated indiscriminately by protein tyrosine kinases. Angiotensin and angiotensin II are phosphorylated by crude biological samples on tyrosine residues. The observation has led to the practice of using these Table 1. Phosphorylation of Src autophosphorylation peptides

Kinase

Peptides

Km(nM)

Reference

P60 ..... P56 ~ck EGF-R Ins-R TPK-1 (lyn) TPK-IIB

EDNEYTARQG IEDNEYTARQG RRa'bLIEDNEYTARQG RRbLIEDACEYAdAR-~G EDNEYTA EDNEYTA

6.25 5.0 0.28 1.2 3.8 0.058

[13] [11] [10] [16] [17] [17]

aResidue underlined indicates addition to or substitution of the autophosphorylation site sequence. bThe change was introduced to facilitate the binding of phosphopeptide to phosphocellulose paper in the kinase assay. ~ change was introduced to avoid charge change of the peptide due to deamidation. aThe change was made so that the tyrosine residue was the sole phosphorylatable residue.

105 peptides as general protein tyrosine kinase substrates in biochemical studies [13]. This is because they do not contain thr or ser residue in the sequences so that tyrosine phosphorylation activity can be measured in the presence of generally overwhelming protein serine/threonine kinase activity. Similarly, amino acid random copolymers of glu and tyr or ala, glu and tyr are widely used as protein tyrosine kinase substrates since many protein tyrosine kinases have been shown to phosphorylate these copolymers [14]. These practices of using so-called general protein tyrosine kinase substrates reinforced the belief that protein tyrosine kinases were highly promiscuous in vitro, a notion dominated much of the early phase of protein tyrosine kinase research (for review, see [18]). Studies on tyrosine phosphorylation of cellular proteins seem to support the notion that protein tyrosine kinases display little individual specificities. A number of cytoskeletal proteins including MAP 2, tubulin, tau proteins and fodrin can be phosphorylated in vitro by multiple protein tyrosine kinases, such as EGF receptor, insulin receptor and src family tyrosine kinases [19-23]. In more recent studies, it was frequently observed that tyrosine phosphorylation of a protein in the cells could be induced by signalling pathways involving different protein tyrosine kinases. Phospholipase C-7, originally found to be phosphorylated and activated in EGF-stimulated cells [24], was later shown to be tyrosyl phosphorylated also upon T-cell receptor stimulation, a pathway mediated by src family kinases [25, 26]. Similarly, vav proto-oncogene product, P95 vav, a protein tyrosine kinase substrate, was found to undergo phosphorylation in intact cells either by E G F treatment or upon stimulation of T-cell receptor [27, 28]. These are just a few examples from the vast literature. Although there are numerous studies demonstrating the phosphorylation of a single protein by multiple protein tyrosine kinases, few studies addressed the question whether the same or distinct sites of the protein were

phosphorylated by different kinases. It has been well established in the more fully developed protein serine/ threonine kinase field, different kinases may phosphorylate common protein substrates on distinct sets of sites (for review, see [29]). In one study, both EGF-R and the src family kinase P56 lyn were shown to phosphorylate clathrin light chain a, but only P56 lyncatalyzed the phosphorylation of clathrin light chain b [30]. Furthermore, the two kinases were found to phosphorylate distinct tyrosine residues in light chain a (Cheng et al., unpublished observation), suggesting that in vitro substrate specificities of the two kinases are different. Thus, protein tyrosine kinases may be revealed to possess in vitro specificity when the phosphorylation sites in the protein substrates are examined. Probably the most compelling evidence for distinct site specificity among protein tyrosine kinases is from the recent discoveries of novel protein tyrosine kinases which possess highly specialized regulatory functions and therefore target very specific cellular proteins. Some of these novel kinases are shown in Table 2. In all cases, the phosphorylated forms of the substrate proteins were identified prior to the discovery of the respective kinases. All these protein substrates are themselves protein kinases which are regulated by specific tyrosyl phosphorylation. In vitro reconstitution of the regulatory reaction can be achieved only when the specific protein tyrosine kinase is used. In the case of glycogen synthase kinase 3, the specific tyrosine kinase is not yet found. In all cases, the phosphorylation occurs at specific tyrosine residue of the protein substrate. Thus, like protein serine/threonine kinases, protein tyrosine kinases may display highly restricted phosphorylation site specificities.

Table 2. Novel protein tyrosine kinases and their physiological substrates

Protein substrate

Protein kinase

Phosphorylation site sequence

Reference

P34 c~c2 Glycogen synthase kinase MAP kinase P60 ..... P56 lck

Wee 1 kinase Unknown MAP kinase kinase b C-src kinase C-src kinase

K6VEKIGEGT~__aGVVYK V208RGEPNNSY__ICSRYY PlvgEHDETGFLTEYVATR T523EQPYQPGENL Ts0IEGQY_QPQP

[33-36] [37] [38, 39] [40] [41]

aThe residues phosphorylated by the respective kinases are underlined. bMAP kinase kinase may phosphorylate a threonine, thr-188, and a tyrosine, tyr-190 of MAP kinase [39].

106

Substrate specificity determinants surrounding target tyrosine Autophosphorylation site sequences have been elucidated for many protein tyrosine kinases, some of which are listed in a recent review [7]. A common feature of these sequences is the presence of multiple acidic residues on the amino terminal side of the tyrosine residue. These observations led to the initial suggestion that local negative charges around the tyrosine target provided the basis for tyrosine kinase specificity. In general, acidic residues proximal to the tyrosine residue do appear to facilitate tyrosine phosphorylation. The peptide gastrin, which contains a string of 5 consecutive glutamate at the aminoterminal side of a tyrosine residue is phosphorylated by EGF receptor, this phosphorylation has a Km value lower than tyrosine kinase-catalyzed peptide phosphorylations [42]. On the other hand, the src family autophosphorylation peptides contain three acidic residues upstream of tyrosine residue, however, substitution of any individual acidic residues by alanine has little effect on the kinetic property of the EGF receptor [10] or P561Ck-catalyzedpeptide phosphorylation [12]. Thus, it was believed for a while, that factors determining substrate specificity of protein tyrosine kinases might be fundamentally different from those governing protein serine/threonine kinase specificity. Evidence has begun to accumulate, albeit slowly, to suggest that important substrate determinants for certain protein tyrosine kinases do exist locally around the tyrosine target. Insulin receptor, a tetrameric protein of two distinct subunits, ch~2, is a protein tyrosine kinase. Upon insulin binding, the receptor undergoes autophosphorylation on multiple tyrosine residues on the ~ subunits with accompanying increase in tyrosine kinase activity (for review, see [43]). A synthetic peptide derived from the three dominant autophosphorylation sites, RRDIYETDYYRK, was found to be phosphorylated by a 48 kDa recombinant insulin receptor protein kinase domain. Proton NMR studies of the peptide phosphorylation indicated that the three tyrosine residues were phosphorylated in an ordered manner, with tyr-9 phosphorylated first, followed by tyr-10, then tyr-5 [44], and that tyr-9 phosphorylation rendered tyr-10 a much better phosphate acceptor than tyr-5. The study elegantly demonstrated the importance of local environment on tyrosine phosphorylation reactions. Insulin stimulation of Chinese hamster ovary cells expressing insulin receptor results in rapid tyrosine phos-

phorylation of a 165-185kDa protein. This protein, called insulin receptor substrate 1 (IRS-I) has been cloned and sequenced [45]. Synthetic peptides corresponding to the sequences surrounding the various phosphotyrosine residues were synthesized and tested as insulin receptor substrates [46]. Six of these peptides, containing a common sequence motif: Tyr-Met-X-Met were phosphorylated by insulin receptor with good (low) Km values, ranging from 24 to 90gM. Kinetic analysis of the substitution analogues of one of these peptides showed that the two methionine residues were essential for efficient phosphorylation of the peptide by insulin receptor. The protein tyrosine kinases which possess highly specialized functions and therefore more strict protein substrate specificities (Table 2), are particularly useful for the study of the nature of substrate specificity determinants of protein tyrosine kinases. All proto-oncogenic forms of src family protein tyrosine kinases contain a carboxyterminal tyrosine residue, corresponding to tyr-527 in c-src, that can be phosphorylated to result in the inactivation of the enzyme (see [47], for review). The enzyme catalyzing this inhibitory phosphorylation, called c-src kinase (CSK) has been identified [40]. In our laboratory, a 35 amino acid peptide corresponding to the carboxyterminal region of proto-oncogene fyn product (Fyn [503-537]: WKKDPEERPTFEYLQSFEDYFTATEPQYQPGENL) was synthesized and found to be efficiently phosphorylated by a CSK sample partially purified from bovine thymus cytosol. A sample of the peptide phosphorylated by CSK in the presence of r[32p]ATP was found to contain phosphotyrosine as the only phosphoamino acid. The phosphopeptide was subjected to exhaustive digestion by chymotrypsin and analyzed by two-dimensional thin layer electrophoresis/ chromatography and autoradiographed. Although three tyrosine containing peptides were expected, each containing one tyrosine residue (EYLQSF, LEDYF and TATEPQYQPGEND), only one phosphopeptide was obtained (Fig. 1). The phosphopeptide was isolated and subjected to radiosequencing, the radioactivity appeared at the 7th cycle of Edman degradation. The result suggested that the tyrosine residue in this peptide corresponds to the negative regulatory tyrosine of c-fyn product. The few studies described above have provided support to the thesis that protein tyrosine kinases, like protein serine and threonine kinases, have diverse protein substrate specificities. For some of these, the substrate specificity determinants are partly localized with the

107

Fig. 1. Analysis of CSK-phosphorylatedpeptide Fyn [503-537]. The peptide, Fyn [503-537] was phosphorylated by a partially purified CSKfrombovinethymusin the presenceof r[32p-ATP].The radioactive phosphopeptide was subjectedto partial acid hydrolysisfollowed by thin-layer electrophoresis and autoradiograph for phosphoamine acid analysis(A), or the phosphopeptidewas exhaustivelyhydrolyzed by chymotrypsin,followedby two-dimensionalthin-layerelectrophoresis/chromatography and autoradiography (B).

polypeptide chain proximal to the tyrosine target. Synthetic peptide approach will be useful in elucidating substrate specificity determinants for these protein tyrosine kinases.

Searching for substrate specificity determinants for src family protein kinases: the use of cdc2 peptides P34 c~c2kinase, a cell division cycle regulatory protein serine/threonine kinase is partly regulated by phosphorylation-dephosphorylation of a specific tyrosine residue, tyr-15 [31]. The active form of the enzyme requires the complexion of the catalytic subunit P34 cd~2with a regulatory protein cyclin B. The activity of the complex, however, is inhibited during interphase of the cell cycle due partly to the phosphorylation of tyr 15 (for review, see [32]). During purification of a src family protein tyrosine kinase, P56 lyn,we made the fortuitous observation that a synthetic peptide derived from aminoterminal region of P34 ~dr cdc2 [8-20] was an exceptionally good peptide substrate of the protein kinase [48]. Under standard conditions, 100gM ATP and 300gM peptide substrate, the rate of P561yn-catalyzed cdc2 [8-20] phosphorylation

was at least 20-fold higher than that of angiotensin phosphorylation. Kinetic characterization of the peptide phosphorylation by purified P56 lynshowed that the Km value, about 0.35gM was comparable to those of the best tyrosine kinase substrate peptides, and the Vmax, about 5 gM/min/mg was by far the highest among those reported for peptide tyrosine phosphorylation [48, 49]. The peptide contained the important regulatory tyrosine residue of P34 cdc2kinase, tyr-15 [31]. Using an extended cdc2 peptide, cdc2 [1-24] which contained two tyrosine residues, tyr-14 and tyr-19 in addition to tyr-15, it was found that only tyr-15 was phosphorylated by P56 lyn. The observation suggested that the peptide contained structural elements rendering tyr 15 an efficient and specific phosphate acceptor. Since the kinetic properties of cdc2 [1-24] and cdc2 [8-20] phosphorylations were similar (Table 3A), we concluded that the substrate specificity determinants for P56 lyn resided in the sequence of amino acid residues 8 to 20 of P34 c~c2[49]. While other peptide substrates such as src autophosphorylation site peptides and angiotensins are indiscriminately phosphorylated by protein tyrosine kinases (see above), the aminoterminal cdc2 peptides are phosphorylated with high efficiency by specific protein tyrosine kinases. Several protein tyrosine kinases: a few src family members, P60 P56 lck, P55 lyn and P60 yes and a number of others, including EGF receptor, insulin receptor, CSK and P43 ab~ were tested for the ability to phosphorylate cdc2 peptides. All the src family members displayed high activity towards cdc2 peptides, with rates of cdc2 peptide phosphorylation 10- to 90-fold higher than the corresponding angiotensin phosphorylation rates. In contrast, the kinases not belonging to src family showed little or no preferential phosphorylation of cdc2 peptides relative to angiotensin [50]. The results led us to suggest that cdc2 peptides are specific substrates for src family tyrosine kinases. Paradoxically, although P56 ly~showed high efficiency phosphorylation of cdc2 peptides, it could not phosphorylate P34 ~dc2kinase to high stoichiometry [51]. The observation, along with the recent identification of human Wee 1, and the demonstration of P34 ~d~2kinase inactivation by Wee 1 [35, 36] argue against that a src family kinase is the kinase phosphorylating tyr-15 of P34 cdr kinase in vivo. In addition, Wee lkinase has been shown to also phosphorylate cdc2 aminoterminal peptides [35]. Irrespective of the physiological significance, the high efficiency and specific phosphorylation of cdc2 peptides by src family kinases is among the first evidence for unique in vitro substrate specificity of a tyrosine kinase .....

,

108

or a tyrosine kinase family. In addition, an opportunity had presented itself to elucidate substrate specificity determinants for a protein tyrosine kinase, since structural feature(s) important for the high efficiency and specific phosphorylation by src family kinases appear to exist locally in the sequence of amino acid residue 8 to 20 of P34 ~dc2. We used a synthetic peptide-based approach to search for the substrate specificity determinants for src family kinases. Key results of the study [49] are summarized in Table 3. Computer search for protein sequence homologous to cdc2 [8-20] uncovered several protein kinase sequences containing a tyrosine residue at a position corresponding to tyr-15 of cdc2. One of these, the sequence from ribosome $6 protein kinase (RSK), RSK [436-456] was synthesized and found to be a poor substrate of P56 tyn, with a rate of phosphorylation by the enzyme only about 2% that of cdc2 [8-20] phosphorylation. Comparison of the sequence of cdc2 [8-20] with the corresponding sequence in RSK peptide (see Table 4 for alignment) has revealed that the two peptides differ in

six positions. Thus, it may be suggested that the difference in substrate activity of the two peptides might be due to the amino acid difference at one or more of the six positions. Single substitution analogues of cdc2 [6-20] peptides, each with one of these amino acids substituted by the corresponding residue in the RSK peptide were synthesized and tested as P 5 6 tyn substrates. As shown in Table 4, only when glu-12 or thr-14 were substituted by val or ser, respectively, the resulting peptide showed a significant decrease in substrate activity. A peptide with residues at both positions 12 and 14 of cdc2 [6-20] substituted was found to be as poor a substrate for P56 lyn as the RSK peptide (Table 3). Thus, the difference in amino acid residue between cdc2 and RSK peptides at these two positions could fully account for the drastic disparity in substrate activity. While the results of Table 3 suggest that glutamate and threonine at -3 and -1 positions of the target tyrosine residue serve as substrate specificity determinants, the possibility that valine at -3 or serine at -1 position exerts negative influence on the substrate activity of the

Table 3. Phosphorylation of aminoterminal cdc2 peptides and peptide analogues by P56 ~yn"

A: cdc2 peptides of varying length Peptide

Sequence

Km (mM)

cdc2[1-24] cdc214-24] cdc216-20] cdc2[8-20]-NH 2

MENYQKVEKIGEGTYGVVYKARHK YQKVEKIGEGTYGVVYKARHK KVEKIGEGTYGVVYK-NH2 EKIGEGTYGVVYK-NH4

0.16 0.30 0.33 Same as cdc2 [6-20] b

B: cdc2 and RSK peptides Peptide

Sequence

Km (mM)

cdc214-24] RSK[436~-56]

YQKVEKIGEGTYGVVYKARHK YVVKETIGVGSYSVCKRCVHK

0.30 Poor substrate c

C: Substitution analogues of cdc216-20] Peptide Sequence

Km (raM)

cdc2[6-20]NH2 [Thrg]cdc2[6-20]NH 2 [ValIZ]cdc2[6-20]NHz [Ser14]cdc216-20]NH2 [Ser16]cdc216-20]NH2 [CyslS]cdc2[6-20]NH2 [Lys~9]cdc2[6-20]NH2 [Valn,Ser~4]cdc2[6-20]NH2

0.33 Same as cdc216-20] b 0.47 1.0 0.37 0.25 0.3 Poor substrate c

KVEKIGEGTYGVVYK-NHz KVETIGEGTYGVVYK-NH z KVEKIGVGTYGVVYK-NH2 KVEKIGEGSYGVVYK-NH2 KVEKIGEGTYSVVYK-NH2 KVEKIGEGTYGVCYK-NH2 KVEKIGEGTYGVVKK-NHz KVEKIGVGSYGVVYK-NHz

Vm (gM/min/mg) 4.0 4.0 6.1

Vm (gM/min/mg) 4.0

Vm (gM/min/mg) 6.1 2.2 1.25 6.7 6.7 10

"Results summarized from [49]. bThe rate of phosphorylation at a peptide concentration of 0.3 gM as determined by HPLC analysis of the phosphopeptide was found to be same as that of cdc[6-20] peptide phosphorylation. CThe peptide phosphorylation was too low for kinetic analysis.

109

I

:r "5

E E

/d 5

3C

[ n 0

E

~8

~4

30

E

LU r

>, I-

'o x

I> I-,

20

0.4 j

5 o

1C

J

J

oz O

20

4O

60

80

FRACTION NUMBER

Fig. 2. Separationof protein tyrosinekinasesin membraneextractof

bovinethymusby hydroxyapatidechromatography.Columnfractions were assayed for both poly (Glu/Tyr) phosphorylation activity (0--0) and cdc2 [6-20] peptide phosphorylationactivity(O--O). The figureis modifiedfromFig.2 of [50]. peptide cannot be excluded. It is also unclear whether or not these amino acids have an effect on the substrate activity only within the particular sequence of cdc2 [820]. Questions about the nature of the effects of the amino acid determinants also need to be addressed, e.g., does the effect of glutamate arise from the negative charge of the amino acid? These, and other ambiguities indicate that further studies are required for the elucidation of specificity determinants of src kinases.

Synthetic peptides as substrates of specific protein tyrosine kinases The discovery of protein tyrosine kinases preceded the discovery of physiologicallyrelevant protein tyrosine kinase substrates by several yars. The lack of readily available and specific substrates for the enzyme assay has greatly hindered the progress of biochemical studies of protein tyrosine kinases (for review, see [17]). In addition, cellular concentrations of protein tyrosine kinases are orders of magnitude lower than these of protein serine/threonine kinases. Any procedures for the assay of protein tyrosine kinase in crude biological samples have to take into consideration the overwhelming protein serine/threonine phosphorylation. To overcome these difficulties, synthetic peptides were used as protein tyrosine kinase substrates early on. For example, angiotensin peptides have been used as protein tyrosine kinase substrates because they contain tyrosine residue as the only phosphorylatable residue. These peptides, as well as other commonly used substrates of protein tyrosine

O

/"

25

50

I 75

100

125

FRACT~N NUMBER

Fig. 3. Separationofproteintyrosinekinasesin bovinethymuscytosol

by DEAE-chromatography.Columnfractionswere assayedfor poly (Glu/Tyr) phosphorylationactivity(O--O) and cdc2 [6-20] peptide phosphorylationactivity(O--9 kinases, such as the src autophosphorylation peptides and poly (Glu/Tyr) co-polymer are general protein tyrosine kinase substrates rather than substrates for specific kinases. Their usefulness in biochemical studies is therefore somewhat limited. The observation that peptides derived from aminoterminal region of P34 cat2 are specific and efficient substrates of src family protein kinases [48] has made it possible for the first time, to assay a family of protein tyrosine kinases in crude biological samples. When used judiciously, such specific assay will help in identifying and assaying for specific protein tyrosine kinases. In our laboratory, a cdc2 peptide-based assay has been used to identify and to monitor the purification of members of src family kinases, lck and fyn from the membrane extract of bovine thymus [48]. A bovine thymus membrane extract, upon chromatography on a hydroxyapatide column, was resolved into four peaks of general protein tyrosine kinase activity (measured by using poly (Glu/ Tyr) as the substrate). When the column fractions were assayed for cdc2 [6-20] phosphorylation activity, the first two peaks of poly (Glu/Tyr) kinase activity were found also to contain kinase activity towards the cdc2 peptide (Fig. 2). Subsequent purification and immunological characterization indicated that peaks 1 and 2 tyrosine kinase activities were from lck and fyn, respectively. Both are members of src family. The kinases in the other activity peak fractions which showed little or no activity toward cdc2 [6-20] did not belong to src family [48].

110

100 -

E

=o E

Z

g

o

=;

LL

o Z

_

,o_

t

I

/ /

-

0.4

I

f

/ /

0.3

/

50

/

o I,,-

p _ _

/

o

0.2

o

8

o O -- 0.1

z

,,=,

Z

P,

J

0

0 5

10 F R A C ~ O N NUMBER

Fig. 4. FPLC on Mono-S column of a partially purified sample of CSK from bovine thymus cytosol. Column fractions were assayed for Fyn [503-537] peptide phosphorylation activity ( 0 - - 0 ) and for P56 ~y~ inactivation activity (O--O). The procedure of P56 ly" inactivation assay is described in text.

More than 90% of the cdc2 [6-20] peptide phosphorylation activity in the extract of bovine thymus was found to be associated with the membrane fraction. The cytosolic protein tyrosine kinase activity appeared to be represented by distinct enzymes. A sample of cytosolic fraction (100,000g supernatant) of bovine thymus extract was subjected to DEAE-cellulose chromatography, and the column fractions were assayed for kinase activity using both poly (Glu/Tyr) copolymer and cdc2 [6-20] peptide as substrates. Figure 3 shows that two peaks of poly (Glu/Tyr) phosphorylation activity and one distinct cdc2 [6-20] phosphorylation activity peak were resolved on the column. Further purification and characterization of the peak fraction of cdc2 [6-20] phosphorylation activity revealed that it contained an active tyrosine kinase which catalyzed the phosphorylation of P34 cdc2kinase partially purified from Hela cells. The phosphorylation of P34 ~c2 was accompanied with an inactivation of the enzyme. The observation suggests that the protein tyrosine kinase is the bovine homolog of Wee 1 kinase. Although Wee i kinase may be assayed by its ability to inactivate P34 ~dc2kinase, the use of peptide phosphorylation assay is, however, much simpler and more practical than the functional assay, especially during enzyme purification when a large number of fractions may have to be assayed. As indicated in an earlier section (Table 2), C-src kinase (CSK), the protein tyrosine kinase exerting an inhibitory phosphorylation of C-src and of other protooncogenic forms of src family members, displays activity towards a 35 amino acid peptide derived from C-termi-

nal region of fyn. Thus, this peptide can be used to develop a relatively specific assay for CSK. As is shown in Fig. 4, Mono-S column chromatography profile of a partially purified CSK fraction showed a single peak of the peptide phosphorylation activity. The peptide phosphorylation activity correlated closely with the functional activity of CSK, i.e. the ability of the enzyme to inactivate a src family kinase, P56 tyn. The functional assay of CSK illustrates another application of the synthetic peptide substrate, cdc2 [6-20]. The procedure consists of preincubating P56 ly" with CSK sample in the presence of MgATP, followed by the addition of the cdc2 peptide to initiate the assay for P56 ]y~ activity. The change in P56 jyn activity as compared to a control sample of P56 ]yn preincubated with MgATP in the absence of CSK is a measure of the functional activity of CSK. Such a functional assay is possible only because cdc2 peptide is a specific substrate for P56 ]yn.

Acknowledgements The authors would like to thank Ms. Lenore Doell for skilled secretarial assistance.

References 1. Fischer EH, Graves D J, Crittenden ERS, Krebs EG: Structure of the site phosphorylated in phosphorylase b to a reaction. J Biol Chem 234: 1698-1704, 1959 2. Nolan C, Novoa WB, Krebs EC, Fischer EH: Further studies on the site phosphorylated in the phosphorylase b to a reaction. Biochemistry 3: 542-551, 1964 3. Kemp BE, Bylund DB, Huang T, Krebs EG: Substrate specificity of the cyclic AMP-dependent protein kinase. Proc Natl Acad Sci USA 72: 3448~452, 1975 4. Graves D J: Use of peptide substrate to study the specificity of phosphorylase kinase phosphorylation. Method Enzymol 99: 268-278, 1983 5. Kemp BE, Pearson RB: Protein kinase recognition sequence motifs. Trends Biochem Sci 15: 342-346, 1990 6. Kennelly PJ, Krebs EG: Consensus sequences as substrate specificity determinants for protein kinases and protein phosphfitases. J Biol Chem 266: 15555-15558, 1991 7. Pearson RB, Kemp BE: Protein kinase phosphorylation site sequences and consensus specificity motifs: tabulations. Method Enzymo1200: 62-81,1991 8. Hunter T: Synthetic peptide substrates for tyrosine kinase. J Biol Chem 257: 4843-4848, t982 9. House C, Baldwin GS, Kemp BE: Synthetic peptide substrates for the membrane tyrosine protein kinase stimulated by epidermal growth factor. Eur J Biochem 140: 363-369, 1984 10. Pike LJ, Gallis B, Casnellie JE, Bornstein P, Krebs EG: Epidermal

111

11.

12.

13.

14.

15. 16.

17.

18.

19.

20.

21. 22.

23.

24.

25.

26.

27.

28.

growth factor stimulates the phosphorylation of synthetic tyrosine-containing peptides by A431 cell membranes. Proc Natl Acad Sci USA 79:1443-1447,1982 Baldwin GS, Stanley I J, Nice EC: A synthetic peptide containing the autophosphorylation site of the transforming proteins of Harvey sarcoma virus is phosphorylated by EGF-stimulated tyrosine kinase. FEBS Lett 153: 257-260, 1983 Casnellie JE, Krebs EG: The use of synthetic peptides for defining the specifcity of tyrosine protein kinases. Adv Enzyme Regu122: 501-515,1984 Wong TW, Goldberg AR: In vitro phosphorylation of angiotensin analogs by tyrosyl protein kinases. J Biol Chem 258: 1022-1025, 1983 Braun S, Raymond WE, Racker E: Synthetic tyrosine polymer as substrates and inhibitors of tyrosine specific protein kinases. J Biol Chem 259: 2051-2054, 1984 Hunter T, Cooper JA: Protein tyrosine kinases. Ann Rev Biochem 54: 897-930, 1985 Kasuga M, Fujita-Yamaguchi Y, Blithe DL, White ME Kahn CR: Characterization of the insulin receptor kinase purified from human placental membranes. J Biol Chem 258: 10973-10979, 1983 Marin O, Donella-Deana A, Brunati AM, Fischer S, Pinna LA: Peptides reproducing the phosphoreceptor sites of PP60 .... as substrates for TPK-IIB, a splenic tyrosine kinase devoid of autophosphorylation activity. J Biol Chem 266: 17798-17803, 1991 Geahlen RL, Harrison ML: Protein tyrosine kinases, In: BE Kemp (ed.) Peptides and Proteins Phosphorylation. Uniscience CRC Press, Boca Raton, Florida, 1990, pp239-253 Akiyama T, Kadowaki T, Nishida E, Kedooka T, Egawala H, Fukani Y, Sakai H, Takaku F, Kasuga M: Substrate specificities of tyrosine-specific protein kinases toward cytoskeletal proteins in vitro. J Biol Chem 262: 14797-14803, 1986 Kadowaki T, Koyasu S, Nishida E, Tobe K, Iqumi T, Takaku E Sakai H, Yahara I, Kasuga M: Tyrosine phosphorylation of common and specific sets of cellular protein rapidly induced by insulin, insulin-like growth factor 1 and epidermal growth factor in intact cell. J Biol Chem 262: 7342-7350, 1987 Cheng N, Sahyoun M: Neuronal tyrosine phosphorylation in growth cone glycoproteins. J Biol Chem 263: 3935-3942, 1988 Wang CY, Kong SK, Wang JH: Characterization of fodrin phosphorylation by a spleen protein tyrosine kinase. Biochemistry 27: 1254-1260, 1988 Matten WT, Aubry M, West J, Maness PI: Tubulin is phosphorylated at tyrosine by PP60 c-srcin nerve growth core membrane. J Cell Bio1111: 1959-1970, 1990 Margolis B, Rhee SG, Felder S, Mervic M, Lyall R, Levitzki A, Ullrich A, Zilberstein A, Schlessinger J: EGF induces tyrosine phosphorylation of phospholipase C-II: A potential mechanism for EGF receptor signalling. Cell 57: 1101-1107, 1989 Park D J, Rho HW, Rhee SG: CD3 stimulation causes phosphorylation of phospholipase C-rl on serine and tyrosine residues in human T-cell. Proc Natl Acad Sci USA 88: 5453-5456,1991 Weiss A, Koretzki G, Schatzwan R, Kadlecek T: Functional activation of the T-cell antigen receptor induces tyrosine phosphorylation of phospholipase C-rl. Proc Natl Acad Sci USA 88: 54845488, 1991 Bustelo XR, Ledbetter JA, Barbacid M: Product of vav proto-oncogene defines a new class of tyrosine potein kinase substrates. Nature 356: 68-71, 1992 Margolis B, Hu P, Katzav S, Li W, Oliver JM, Ullrich A, Weiss A,

29. 30.

31.

32. 33. 34. 35.

36.

37.

38.

39.

40.

41.

42.

43. 44.

45.

46.

47.

48.

Schlessinger J: Tyrosine phosphorylation of vav proto-oncogene product containing SH2 domain and transcription factor motifs. Nature 356: 71-74, 1992 Edelman AM, Blumenthal DK, Krebs EG: Protein serine/threonine kinases. Ann Rev Biochem 56: 567-613,1987 Mooibroek MJ, Cheng H-C, Wang JH: Differential in vitro phosphorylation of clathrin light chains by epidermal growth factor receptor-associated protein tyrosine kinase and a PP60C-~rC-related spleen tyrosine kinase. Arch Biochem Biophys 292: 448-455,1992 Gould KL, Nurse P: Tyrosine phosphorylation of the fission yeast cdc2 protein kinase regulates entry into mitosis. Nature 339: 626629, 1989 Norbury C, Nurse P: Animal cell cycle. Ann Rev Biochem 61: 441470, 1992 Featherstone C, Russell P: Fission yeast P107wee1mitotic inhibitor is a tyrosine/serine kinase. Nature 349: 808-811, 1991 Igarashi M, Nagata A, Jinno S, Suto K, Okayawa H: Weel-like gene in human cells. Nature 353: 80-83, 1991 McGowen CH, Russell P: Human Wee1 kinase inhibits cell division by phosphorylating P34 cdc2exclusively on Tyr-15. EMBO J 12: 75-85, 1993 Parker LL, Atherton-Fessler S, Piwnica-Worms H: P107w~elis a dual-specificity kinase that phosphorylates P34 cdc2on tyrosine 15. Proc Nail Acad Sci USA 89: 2917-2921, 1992 Hughes K, Nikalakaki E, Plyte SE, Totty NE Woodgett JR: Modulation of the glycogen synthase kinase-3 family by tyrosine phosphorylation. EMBO J 12: 803-808, 1993 Seger R, Ahn NG, Posoda J, Munar ES, Jensen AM, Cooper JA, Cobb MH, Krebs EG: Purification and characterization of mitogen-activated protein kinase activator(s) from epidermal growth factor-stimulated A431 cells. J Biol Chem 267: 14373-14381, 1992 Payne DM, Rossowando SJ, Martino R Erickson AK, Her JH, Shabbuowitz J, Hunt DE Weber M J, Sturgill JW: Identification of the regulatory phosphorylation sites in PP42/mitogen-activated protein kinase (MAP-kinase). EMBO J 10: 885-892, 1991 Nada S, Okada S, Yamanushi Y, Yamamato T, Nakagawa H: 'CSK': a protein tyrosine kinase involved in regulation of src family kinases. J Biol Chem 266: 24249-24252, 1991 Bergman M, Mustelin T, Oetken C, Partanen J, Hlint NK, Aurein KE, Autero M, Buru R Alitelo K: The human P50~k tyrosine kinase phosphorylates P56 lckat tyr505 and down regulates its catalytic activity. EMBO J 11: 2919-2924, 1992 Baldwin GS, Burgess AW, Kemp BE: Phosphorylation of a synthetic gastrin peptide by the tyrosine kinase of A431 cell membranes. Biochem Biophys Res Commun 109: 656-660, 1982 Rosen OM: After insulin binds. Science 237: 1452-1458, 1987 Levine BA, Clark B, Ellis L: A soluble insulin receptor kinase catalyzes ordered phosphorylation at multiple tyrosines of a decapeptide snbstrate. J Biol Chem 266: 3565-3570, 1991 Sun XJ, Rottenberg R Kahn CR, Bacher JM, Araki E, Wilden PA, Cahill DA, Goldstein B J, White MF: Structure of the insulin receptor substrate IRS-1 defines a unique signal transduction protein. Nature 352: 73-77, 1991 Shoelson SE, Chatterjee S, Chaudhuri M, White MF: YMXM motifs of IRS-1 define substrate specificity of the insulin receptor kinase. Proc Natl Acad Sci USA 89: 2027-2031, 1992 Cooper JA: The SRC family of protein tyrosine kinases. In: BE Kemp (ed.) Peptides and Protein Phosphorylation. Uniscience CRC Press, Boca Raton, Florida, 1990, pp 85-113 Litwin C, Cheng H-C, Wang JH: Purification and characterization

112 of a PP60C-~r~ tyrosine kinase that efficiently phosphorylates a synthetic peptide derived from P34 ~162J Biol Chem 266: 2557-2566, 1991 49. Cheng H-C, Litwin C, Hwang DM, Wang JH: Structural basis of specific and efficient phosphorylation of peptides derived from P34 cdc2by a PP60c-~rc-related tyrosine kinase. J Biol Chem 266: 17919-17925,1991

50. Cheng H-C, Nishio H, Hatase O, Ralph S, Wang JH: A synthetic peptide derived from P34 cdczis a specific and efficient substrate of src family tyrosine kinases. J Biol Chem 267: 9248-9256, 1992 51. Litwin C: Purification and characterization of P56 Iy"from bovine spleen. Ph.D. Thesis, University of Calgary.