Purification and Characterization of a Rat Liver Membrane Tyrosine-Protein Kinase, the Possible Protooncogene c-yes Product, p60c-yes *

A tyrosine-protein kinase was purified more than 270-fold from the rat liver plasma membrane fraction by successive column chromatographies on Sephacryl S-300, wheat germ agglutinin-agarose, casein-Sepha-rose, and hydroxylapatite, followed by isoelectrofocusing electrophoresis. The enzyme with PI of 6.2 was a 60-kDa single polypeptide which represented 42% of total protein. The enzyme reacted quantitatively with a monoclonal antibody to the amino-terminal sequence (Cys-3 to Ser-66) specific to the human c-yes protein, but not with antibodies to the specific amino-terminal sequences of the c-src, fyn, and lck proteins. The purified enzyme contained almost no phosphotyrosine residue but was autophosphorylated with Mg*ATP exclusively at tyrosine residues with concomitant increase in the kinase activity. The rates of autophosphorylation of the enzyme and phosphorylation of ty-rosine-glutamate (1:4) copolymers, catalyzed by the enzyme were proportional to the square of enzyme concentration, suggesting that p60e-yes undergoes autophosphorylation

A tyrosine-protein kinase was purified more than 270-fold from the rat liver plasma membrane fraction by successive column chromatographies on Sephacryl S-300, wheat germ agglutinin-agarose, casein-Sepharose, and hydroxylapatite, followed by isoelectrofocusing electrophoresis. The enzyme with PI of 6.2 was a 60-kDa single polypeptide which represented 42% of total protein. The enzyme reacted quantitatively with a monoclonal antibody to the amino-terminal sequence (Cys-3 to Ser-66) specific to the human c-yes protein, but not with antibodies to the specific amino-terminal sequences of the c-src, fyn, and lck proteins. The purified enzyme contained almost no phosphotyrosine residue but was autophosphorylated with Mg*ATP exclusively at tyrosine residues with concomitant increase in the kinase activity. The rates of autophosphorylation of the enzyme and phosphorylation of tyrosine-glutamate (1:4) copolymers, catalyzed by the enzyme were proportional to the square of enzyme concentration, suggesting that p60e-yes undergoes autophosphorylation through intermolecular catalysis, resulting in stimulation of the enzyme activity. Although the enzyme reaction showed an essential requirement for Mg2+ or Mn2+ with optimal concentrations of 20 and 3 mM, respectively, autophosphorylation significantly activated the enzyme only in the presence of M8+. Autophosphorylation of the enzyme reduced the K, for tyrosine-glutamate copolymers and tubulin, but not for ATP, and increased the V,,, of copolymer and tubulin phosphorylation.
Tyrosine-protein kinase activity was found to be associated with the products of several viral oncogenes and their cellular homologs, the protooncogenes (l), as well as with transmembrane receptors of some growth factors (2-4) or growth related hormones (5,6), and therefore was suggested to play important roles in the control of cell growth and differentiation. Among the products of cellular protooncogenes, tyrosineprotein kinases are a family of a 60-kDa phosphoprotein, *This work was supported in part by grants-in-aid for Cancer Research and Scientific Research from the Ministry of Education, Science, and Culture of Japan (1986)(1987)(1988)(1989). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ To whom correspondence should be addressed.
p60"-"" encoded by the protooncogene c-src. The src family tyrosine-protein kinases are encoded by very closely related genes (fgr (7), yes (a), fyn (9, IO), lyn ( l l ) , Ick (12), and hck (13,14)), which all lie in the size range of 505-543 amino acid residues, are highly conserved over a 460-residue contiguous region at the carboxyl terminus, and differ largely in an 80residue contiguous region at the amino terminus. They lack a transmembrane domain, but are nonetheless associated with membranes via their myristoyl group which is linked to aamino group of the amino-terminal glycine (15,16).
In spite of extensive studies especially with p60"~"" and p56Ick, the precise physiological functions of the src family tyrosine-protein kinases have not yet been clearly elucidated.
In this report, we described the purification of a 60-kDa tyrosine-protein kinase from the rat liver plasma membrane fraction. The 60-kDa tyrosine-protein kinase reacts quantitatively with a monoclonal antibody (17) generated against an amino-terminal 64 amino acid sequence specific to the human c-yes product. The purified enzyme seems to be mostly in a dephosphorylated form and undergoes autophosphorylation with ATP exclusively at tyrosine residues probably through intermolecular catalysis, resulting in stimulation of the tyrosine-protein kinase activity. This is the extension of our study (18) on tyrosine-protein kinases associated with mouse liver plasma membranes, and the first demonstration of kinetic properties and activation by autophosphorylation, of the possible protooncogene c-yes product.

4831
Identification of the 60-kDa Tyrosine-Protein Kinase as the Product of the c-yes Protooncogene-In order to study the relationship of the rat liver 60-kDa tyrosine-protein kinase to known membrane-associated tyrosine-protein kinases of about 60 kDa, immunoreactivity of the rat liver enzyme was investigated with antibodies to the amino-terminal sequences specific to p60c-yes, p59"", p6OC' "", and p56ICk. Fig. 8 shows an autoradiogram of a Western immunoblot of the rat liver tyrosine-protein kinase with antibodies against ~6 0 ' '~'~ (mAb 3H9), p60""" (mAb 327), and ~59"". After the double-antibody recognition procedure, radiolabeled bands of 60 kDa were observed with an antibody against p60'-ye" (Fig.  8 A ) , but not with an antibody against p60""" (Fig. 8B). No significant radioactive bands were detected with an antibody against ~5 9 '~" (Fig. 8C). The marginal immunoreactivity of the 60-kDa tyrosine-protein kinase with the anti-p5gry" antibody ( Fig. 8C) is considered further under "Discussion." The rat liver tyrosine-protein kinase was not immunoprecipitated with an antibody against p56Ick (data not shown).
The radioactivities detected with two different amounts of the liver tyrosine-protein kinase were proportional to the amount of the enzyme applied to the gel (Fig. 8A, lanes c and d). Furthermore, the radioactivities detected with three different enzyme samples from the final three steps of purification were also nearly proportional to the units of enzyme applied (Fig. 8A). The radioactivities were 253 cpm/unit of casein-Sepharose fraction (lane a), 341 cpm/unit of hydroxylapatite fraction (lane b), and 329 cpm/unit of Ampholine fraction (lane d).
These results suggest that the majority of tyrosine-protein kinase activity in the preparation is attributed to p60c-ye'.
Autophosphorylation of p60'-Yes-The initial rates of phosphorylation of 60-kDa tyrosine-protein kinase during the tyrosine-glutamate copolymer phosphorylation, increased with increasing enzyme concentration (Fig. 9A). In addition, when the initial rate of the reaction was plotted against the square of the protein concentration, a straight line was obtained. The observation therefore suggested that the autophosphorylation of the 60-kDa tyrosine-protein kinase proceeds by an intermolecular catalytic mechanism.
The direct quantitation of p60c-ye" in our purified tyrosineprotein kinase preparation by immunoprecipitation with a monoclonal antibody (mAb 3H9) specific for p60c-yes revealed that at least 94% of the tyrosine-protein kinase activity was attributable to p60'-ye8 (Table 11). However, it is still possible that another kinase is involved in the phosphorylation of p60c-ye8.
To eliminate the possible involvement of other tyrosineprotein kinases in the phosphorylation of p60c-ye", the immunoprecipitates of p60'-ye" with mAb 3H9 were employed to measure the rates of autophosphorylation and tyrosine-glutamate copolymer phosphorylation simultaneously in a same tube (Fig. 9B). The rates of these reactions changed with the increase in p60c-ye8 concentration parabolically and were proportional to the square of p60c-ye8 concentration (Fig. 9B), as observed with the final enzyme preparation (Fig. 9A), confirming the autophosphorylation of ~6 0 ' -~~" , which was suggested to take place by intermolecular catalysis. About 0.5 mol of phosphate/mol of p60"ye" was incorporated for 10 min at the highest concentration of p60c-ye8 (Fig. 9B).
Activation of p60c-yeS by Autophosphorylation-The initial rate of tyrosine-glutamate copolymer phosphorylation by p60c-yes also increased with increasing enzyme concentration ( Fig. 9B), suggesting that the p60"ye8 was activated by autophosphorylation.
Treatment of the enzyme preparation with Sepharose-conjugated acid phosphatase did not change the characteristic enzyme dose response curve for the rate of tyrosine-glutamate copolymer phosphorylation (data not shown), suggesting that p60C-ye8 may be partially active as a tyrosine-protein kinase even before autophosphorylation at tyrosine residues.
To confirm that the 60-kDa rat liver tyrosine-protein kinase was being activated by autophosphorylation, a high concentration (5.5 units, 137 ng of protein) of enzyme, which would be expected to be activated by autophosphorylation from the results in Fig. 9A, was preincubated with ATP for 30 min at 30 "C. Tyrosine-protein kinase activity toward tyrosine-glutamate copolymers was then measured at a low concentration (13.7 ng of protein) of the preincubated enzyme, which should be activated to a lesser extent by ATP (Fig. 9A 1. As shown in Table 111, enzyme activity was stimulated 2.6-fold by the preincubation with ATP compared with enzyme which was preincubated without ATP. ATP could not be substituted by AMP-PNP, a nonhydrolyzable analog of ATP (data not shown).
To further confirm the activation of the enzyme by autophosphorylation at tyrosine residues, immunoreactivity of the FIG. 9. Enzyme dose curves of the autophosphorylation and tyrosineglutamate copolymer phosphorylation. Indicated doses of rat liver tyrosine-protein kinase (Ampholine fraction) ( A ) and immunoprecipitated ~6 0 "~" ( B ) were employed. p60e-yes of different doses was immunoprecipitated with mAb 3H9 as described in the legend to Table 11. The amount of p60'~y"8 was estimated from the recovery of tyrosineprotein kinase activity in the immunoprecipitate and from the amount of 60-kDa protein (42% of total protein) in the immunoreaction. Rates of Pi incorporation into endogenous 60-kDa protein (autophosphorylation) (A, A) and HPLC-fractionated copolymers (0, 0) were measured at 30 "C for 10 min as described under "Experimental Procedures."  Zmmunoreaction of rat liver tyrosine-protein kinase with antibody against p60e-yea Indicated doses of tyrosine-protein kinase (Ampholine fraction, 29,900 units/mg) were incubated with mAb 3H9 (10.8 pg protein) in 80 pl of buffer C on ice for 2 h with tapping every 15 min. The reaction mixture was added by 40 pl of protein A-Sepharose beads (1:l suspension in buffer C) and further incubated on ice for 30 min with occasional mixing with a thin glass rod and then centrifuged at 16,000 X g for 2 min. The precipitate was separated from the supernatant and was washed once with 400 pl of buffer C. Tyrosine-protein kinase activity in the reaction mixture and the precipitate was measured under standard assay conditions with 0.056 mg/ml HPLCfractionated tvrosine-glutamate couolvmers as substrate. preincubated enzyme with a monoclonal antibody against phosphotyrosine was investigated (Table 111). Rat liver tyrosine-protein kinase was preincubated with or without ATP and then incubated with a Sepharose-conjugated monoclonal antibody against phosphotyrosine. After centrifugation, the tyrosine-protein kinase activity in the supernatant and the immunoprecipitate was measured with tyrosine-glutamate copolymers as substrate. Tyrosine-protein kinase activity associated with the immunoprecipitate was measured after elution of phosphorylated enzyme with p-nitrophenyl phosphate. As shown in Table 111, when rat liver tyrosine-protein kinase was preincubated without ATP and then incubated with an antiphosphotyrosine antibody, more than 95% of the recovered tyrosine-protein kinase activity was not immunoprecipitated and recovered in the supernatant. The same results were obtained without preincubation, indicating that dephosphorylation of tyrosine in the enzyme did not take place during the preincubation. In contrast, when the enzyme was preincubated with ATP, more than 75% of the recovered tyrosineprotein kinase activity was immunoprecipitated with the antiphosphotyrosine antibody with a concomitant reduction of activity in the supernatant. Enzyme activity eluted from the   TABLE 111 Zmmunoreaction of rat liver ~60"" with antibody against phosphotyrosine The phosphorylated or unphosphorylated enzyme sample was prepared by incubating tyrosine-protein kinase (Ampholine fraction, 5.5 units, 137 ng of protein) with or without 30 p M ATP at 30 "C for 30 min in 60 p1 of a reaction mixture containing 20 mM Hepes-NaOH, pH 7.4, 20 mM MgC12, 30 p~ sodium vanadate, 0.5 mM dithiothreitol, and 0.2% Triton X-100. The reaction was stopped by adding 10 p1 of 200 mM EDTA. After diluting the mixture with 30 pl of buffer C, 50 p1 of the mixture was incubated at 0 "C for 2 h with a Sepharoseconjugated antibody against phosphotyrosine (8 p1 gel, 113 pg of antibody) with resuspension by gentle tapping every 15 min and centrifuged for 3 min at 16,000 X g. The supernatant was used as an enzyme sample (Supernatant). The precipitate was once washed with buffer C and associated tyrosine-protein kinase was extracted twice with 50 p1 of 27 mM p-nitrophenyl phosphate in buffer C at 0 "C by gentle mixing three times with a glass rod every 15 min. The combined extracts were used as an enzyme sample (Immunoprecipitate). Pi incorporation into copolymers was measured under the standard assay conditions with 10 pl of enzyme samples. Western immunoblotting was performed with 20 pl of enzyme samples as described under "Experimental Procedures." After the double-antibody recognition procedure, '251-labeled 60-kDa bands were excised from blotting membranes and their radioactivities were measured to represent the amount of p60c-ye*. immunoprecipitate with p-nitrophenyl phosphate was higher than the total activity recovered from the sample preincubated without ATP, indicating that autophosphorylation of the enzyme indeed stimulated tyrosine-protein kinase activity.
In order to know whether we were dealing with the phosphorylation and activation of p60c-y"", the enzyme preparation was preincubated with or without ATP and immunoreactivity of p60c-ye" with a monoclonal antibody against phosphotyrosine was analyzed by quantitative Western immunoblotting

Purification and
Characterization of Rat Liver p60c-yes with mAb 3H9 (Table 111). Reactivity of p60C-yeS with mAb 3H9 was not changed by incubation with ATP, under conditions where tyrosine-protein kinase activity was stimulated 2.6-fold (Table 111). When p60e-yes was incubated with ATP, after treatment of the reaction mixture with a Sepharoseconjugated anti-phosphotyrosine monoclonal antibody, 24% of ~60'.~'* remained in the supernatant and 43% was recovered from the immunoprecipitate. On the other hand, when p60e-yes was incubated without ATP, 90% of p60e-yeS was detected in the supernatant and only 3% was recovered from the immunoprecipitate (Table 111). These results confirm that p60"-ye8 in our purified preparation is mostly unphosphorylated a t tyrosine residues and is autophosphorylated at tyrosine residues by incubation with ATP. From the analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, 20% of "'P-labeled 60-kDa protein was supposed to be not immunoprecipitable with an anti-phosphotyrosine antibody under the conditions employed in Table I11 (data not shown). Taking this point into account, the distribution of p60c-yes in the supernatant and the immunoprecipitate was well correlated with the distribution of tyrosine-protein kinase activity, consistent with the notion that p60"-y"8 is the major tyrosineprotein kinase in the preparation.
These results confirmed that tyrosine-protein kinase activity of ~6 0 " -~" " was stimulated by the autophosphorylation a t the tyrosine residues.
Effect of Mg2' and Mn2+ on the Activation of Tyrosine-Protein Kinase-The effect of various concentrations of M$+ and Mn2+ on the tyrosine-glutamate copolymer phosphorylation catalyzed by rat liver tyrosine-protein kinase was investigated at high and low concentrations of enzyme (Fig. 10, A  and B ) . The level of autophosphorylation was expected to be higher at 2.0 units (34.0 ng) of the enzyme but lower a t 0.28 unit (4.75 ng) during the phosphorylation of exogenous sub-

80
Metal Concentration (mM) strate (Fig. 9A). Either M$+ or Mn2+ was necessary for the tyrosine-protein kinase activity. Optimal concentrations of M$' and Mn2+ were 20 and 3 mM, respectively. The same optimal concentrations were observed with high or low concentrations of the enzyme. However, the maximal activity at optimal Mg2+ concentration relative to the maximal activity at optimal Mn2+ concentration was more than three times higher with high concentrations of the enzyme than with low concentrations of the enzyme (Fig. 10, A and B ) , suggesting that the autophosphorylated enzyme might only be activated in the presence of M$+. T o justify this supposition, the effect of Mg2+ and Mn2+ on the rate of tyrosine-glutamate copolymer phosphorylation at a low concentration (0.33 unit) of autophosphorylated enzyme were investigated (Fig. lOC). The autophosphorylated enzyme was prepared by incubation of a high concentration of enzyme with Mg . ATP, followed by immunoprecipitation with an antiphosphotyrosine antibody and by elution with p-nitrophenyl phosphate from the immunoprecipitate. Low concentrations of phosphorylated enzyme showed an absolute requirement for Mg2+ or Mn2+ and the same optimal concentration of M$+ or Mn2+ observed with low and high concentrations of unphosphorylated enzyme. Actually, the maximal activity of the phosphorylated enzyme a t optimal Mg2+ concentration relative to the maximal activity a t optimal Mn2+ concentration, was nearly five times higher than that of low concentrations of the unphosphorylated enzyme (Fig. 10, B and C), supporting the notion that the phosphorylated enzyme was significantly activated only with Mg2+.
Indeed, the initial velocity of tyrosine-glutamate copolymer phosphorylation at optimal Mn2+ concentration showed a first-order dependence on the enzyme concentration, in contrast to a parabolic enzyme dose curve of the reaction at optimal M e concentration (Fig. 11).
Additionally, in time courses of tyrosine-glutamate copolymer phosphorylation by the unphosphorylated enzyme, a slight lag was observed within 1 min with 20 mM MgC12, but not with 3 mM MnC12 (data not shown). Such a lag was also not observed with the phosphorylated enzyme even in the presence of 20 mM MgC12 (data not shown).
These results indicate that the autophosphorylation of ~6 0 " -~' " significantly stimulates the tyrosine-protein kinase  Kinetic Constants of Phosphorylated and Unphosphorylated Tyrosine-Protein Kinase-To clarify the mechanism of activation of rat liver tyrosine-protein kinase by phosphorylation, the kinetic constants of the reaction were measured at high (1.1 units, 25 ng) and low (0.34 unit, 7.6 ng) concentrations of enzyme (Table IV). Since autophosphorylation of rat liver tyrosine-protein kinase increased with increasing enzyme concentration (Fig. 9A), the ratio of phosphorylated enzyme to unphosphorylated enzyme during the enzyme reaction was expected to be higher in the reaction mixture with 25 ng of the enzyme than with 7.6 ng of the enzyme. An increase in the enzyme concentration decreased the K, for tyrosineglutamate copolymers and tubulin, but not for ATP, and increased the V,,, of copolymer and tubulin phosphorylation (Table IV).
Using a low concentration (0.34 unit) of phosphorylated enzyme prepared as described in Table 111, K,,, values for tyrosine-glutamate copolymers, tubulin, and ATP were determined to be 1.06 ? 0.08, 8.81 .t 0.88, and 29.6 ? 2.4 p M , respectively. The results indicate that activation of rat liver p6W-yes by autophosphorylation is due to a decrease in the K,,, values for copolymers and tubulin with a concomitant increase in the V,,,.

DISCUSSION
In this paper, a 60-kDa tyrosine-protein kinase was solubilized from the rat liver plasma membrane fraction and purified more than 270-fold to about 42% purity. The tyrosine-protein kinase reacted with a monoclonal antibody (mAb 3H9) (17) which was raised with an amino-terminal sequence of 64 amino acid residues (Cys-3 to Ser-66) specific to human c-yes gene product as antigen. mAb 3H9 has been shown to react with human, mouse, and rat c-yes gene products, but not with human c-src, c-fgr, and fyn gene products (17). Expression of the c-yes protein in liver from chicken (36) and human (17) has been reported.
The polyclonal antibody (23) against ~5 9 '~" was raised with a human p5gfY" amino-terminal sequence of 117 amino acid residues (Ser-25 to Val-141) which contains not only a specific amino-terminal sequence (Ser-25 to Gly-80) but also a sequence (Gly-83 to Pro-140) similar to a sequence (Gly-92 to Pro-149) of human p60C-yeS. Consequently, the marginal immunoreactivity of the 60-kDa tyrosine-protein kinase with the polyclonal antibody against ~59"" (Fig. 8C) is probably due to cross reactivity of the antibody to p60'~ye". Indeed, the faint 60-kDa band showed slightly slower mobility than rat brain p5gfY" on sodium dodecyl sulfate-polyacrylamide gel TABLE IV Kinetic constants of rat liver tyrosine-protein kinase The rate of phosphorylation of tyrosine-glutamate copolymers (0.5-3 FM) and tubulin  was measured with indicated doses of rat liver tyrosine-protein kinase (Ampholine fraction, 0.045 unit/ng protein) under standard assay conditions. Tyrosine-glutamate copolymers were used as substrate for determining K, values for ATP (9-45 FM). K , values were estimated by fitting the data to Michaelis-Menten equations using the method of least squares. Values are averages & S.D. of more than three separate analyses.  (Fig. 8C). Moreover, rat brain ~59"" could be completely separated from rat brain ~6 0 " -~'~ by column chromatographies on casein-Sepharose and h~droxylapatite,~ which were employed in this paper for the purification of the rat liver p60C-yeS. Therefore, the possibility of slight contamination of ~5 9 '~" in the rat liver p60r-yeS preparation appears remote.

Substrate
The membrane tyrosine-protein kinase was found to be mainly p60"~ye" from the following evidence: 1) at least 94% of tyrosine-protein kinase activity in the final enzyme preparation was immunoprecipitated with mAb 3H9 (Table 11); 2) the enzyme reacted quantitatively with mAb 3H9 as a 60-kDa band during Western immunoblotting, but not with antibodies against p60"-"'" and p5gfY" (Fig. 8); 3) no tyrosine-protein kinase activity in the final enzyme preparation was reacted with an antibody against ~5 6 ' '~ (data not shown); 4) the amount of PGOC.~~" detected by quantitative Western immunoblot analysis with mAb 3H9 at the final three purification steps was nearly proportional to the units of enzyme in each fraction (Fig. 8); 5 ) when the purified enzyme preparation was incubated with or without ATP, and then with a Sepharoseconjugated anti-phosphotyrosine monoclonal antibody, the distribution of p60C"'efi in the supernants and the immunoprecipitates showed a good correlation with the distribution of tyrosine-protein kinase activity (Table 11).
The possibility that phosphorylation of p60c-ye" was due to trace contamination by a distinct tyrosine-protein kinase in the enzyme preparation appears unlikely, since p60c~y'" in the immunocomplex with mAb 3H9 was phosphorylated by incubation with ATP at the rate comparable to that observed with the final enzyme preparation (Fig. 9, A and B ) . Furthermore, the immunocomplexed p60c-ye* dose curve (Fig. 9B) of the autophosphorylation and the phosphorylation of tyrosineglutamate copolymers showed a good consistency with the corresponding dose curve (Fig. 9A) obtained with the final enzyme preparation.
The specific activity (kinase acti~ity/p6OC.~~') of autophosphorylated p60"yefi which was eluted from the immunoprecipitate with an anti-phosphotyrosine antibody was 5.5-fold higher than that of unphosphorylatedp60c~yeS, which remained in the supernatant after the antibody treatment (Table 111), indicating that p60"ye" is stimulated at least 5.5-fold by autophosphorylation.
From rat liver cytosol, Wong and Goldberg (37) purified a tyrosine-protein kinase to near homogeneity by using angiotensin I1 as substrate. The cytosolic tyrosine-protein kinase is clearly distinct from the membrane tyrosine-protein kinase reported in this paper, because the cytosolic enzyme has a molecular size of about 75 kDa, exhibits stringent dependence on Mn2+, and does not contain any of the major antigenic determinants found in retroviral tyrosine-protein kinases.
Swarup et al. (38) purified a 60-kDa tyrosine-protein kinase to near homogeneity from the particulate fraction of rat spleen. The rat spleen enzyme prefers Mg2+ to Mn2+ for its kinase activity and is suggested to be activated by autophosphorylation at the tyrosine residues.
The autophosphorylation of is suggested to be an intermolecular reaction since the initial rate of the reaction increase with increasing p60C~yes concentration and is proportional to the square of the concentration (Fig. 9B). Similar enzyme dose curves of autophosphorylation of a bovine spleen 50-kDa tyrosine-protein kinase, which relies on autophosphorylation for activity, have been reported (39). The autophosphorylation of solubilized epidermal growth factor receptor is also mediated by intermolecular cross-phosphorylation T. Miyauchi, unpublished data.

Purification and
Characterization of Rat Liver p60C-yeS (40), probably facilitated by receptor oligomerization, induced by the binding of epidermal growth factor to the receptor (41). Another src family tyrosine-protein kinase, p56Ick, has been known to associate with CD4 T-cell surface antigen which can transduce an independent signal during T-cell activation (42). The physical cross-linking of CD4 molecules with CD4 antibodies, but not the binding with monovalent fragments of the antibodies, stimulated rapidly the phosphorylation of tyrosine residues of p56Ick, with concomitant activation of the tyrosine-protein kinase activity, supporting the view that the C D 4 .~5 6 '~~ complex functionally mimics growth factor receptors with associated tyrosine-protein kinase function (42). Therefore, the intermolecular autophosphorylation of p60e-yes with the concomitant increase in tyrosine-protein kinase activity is speculated to be facilitated by p60c~ye" dimerization or oligomerization on an unidentified signaling pathway.

Purification and
Characterization of Rat Liver p60c-yes

Membranes.
Solubilization and Psrlisl Furificsfmn of Tyrosine-Protein Kinuse from Hat L
The aalubilirad tyrosine-protein kinase activity wlm resolved into two peaks by geifiltration on Sephscryi S-300 in the presence 01 0.2% Triton X-IO0 and 0 . 2 M NaCi (Fig. i l . About 80% of the tatsl tyrosine-protein kinase activity WNQ eluted us s major peak with B Stokes radius of 6.1-8.3 nnl, and about is% of tho activity v a s eluted B J 8 mmor peak with B Stokes radius of 2 . 7 -2 . 8 nm. The msjmr peak indicated by B solid bar in Fig.  1 vas pooled and further purified by WGA-egaro)be column chromatography under tho condition in which the EGF receptor and the insulin receptor would be expcctcd to be retained completely by adsorb to the WGA-agarose column. The flow-through fractions W C~C further purified by the column l i s , 35). More than 87% of the rceavcrrd tyrosine-profcln kinase sctivily did not succe99iYe column chromatographics on casein-Srpharosr (Fig.  2 ) and On hydroxylapatite ( Fig. 31, foilowed by isoelectrofacuatng (Fig. 4 1 . During eiectrophorcsis, tyrosine-protein protein kinase vas purified more than 270-fold from the rut liver plasm8 membrana froctlan kinase activity migrated as B sharp peak of pi 6 . 2 With a recovery of 768. The tyrosinowith the reCOVely af 7% ( T s b i e i l . msidues, each active fraction from 28 to 35 (Fig. 4 ) was incubated with 17-32P lATP for 10 Since most tyrosine-protein kinsscs are knawn to sutophosphorylote at tyrosine min at 30*C, and the phosphorylsted proteins were snniyacd by SDS-poiyscryisrnide gel CieCtrophoresis followed by sutoradiogruphy ( Fig.   51.