J. Biol. Chem.-1992-Yeung-23447-50

8/8/2019 J. Biol. Chem.-1992-Yeung-23447-50

http://slidepdf.com/reader/full/j-biol-chem-1992-yeung-23447-50 1/4

Communication Vol. 267, No . 33, Issue of November 25, pp. 23447-23450,19920 1992 by Th e American Society for Biochemistry and Molecular Biology, Ine.

Printed in U.S.A.

THEJOURNALOF BIOLOGICALCHEMISTRY

Protein Tyrosine  Phosphatase-1 CIs Rapidly  Phosphorylatedin

Tyrosine in Macrophages inResponse  to  Colony  StimulatingFactor- 1"

(Received for publication, August 27,  1992)

Yee-GuideYeung, Karen L . Berg$,FionaJ. Pixleys, Ruth Hogue Angeletti, andE. Richard  StanleyT

From  the  Departm ent  ofDeuelopmental Biology and

Cancer, Albert  E instein College of Medicine,Bronx,  New  York10461

An -64-kDa cytoplasmic  protein is rapidly  phos-phorylated in tyrosine in the  response  of  macrophages

to colony  stimulatingfactor- 1. To identify this protein,BAC1.2F6 macrophages were incubated with or with-out  colony  stimulating factor- 1, the  phosphotyrosine-containing portion   oftheir cytosolic fractions subjectedto size exclusion  chromatography,  and  the46-70-kDafraction further  fractionated by reverse phase highpressure liquid  chromatography  (RP-HPLC).  Trypticpeptides of pooled  RP-HPLCfractions from  stimulatedcells (containing  the-64-kDa protein  and  an-64-kDaprotein) and  from  unstimulated cells (containing the-64-kDa protein alone), were sequenced directly. Allseven readable  sequences  of8 sequenceable  peptidespresent  uniquely  in  the  stimulatedfraction were pres-ent in the sequence of the 81% homology 2 domain-containing protein tyrosine phosphatase-lC (PTP- 1C).

The  identityof the -64-kDa protein was confirmed  byWestern  blottingwith an antibody  raised  to  aPTP-1Cpeptide. The rapid, growth factor-induced tyrosinephosphorylation  of PTP-1C suggests that it may  beinvolved in very early events in growth factor signaltransduction.

The action of the mononuclear  phagocyte  growth  factor,colony stimulating factor-1 (CSF-1)l is mediated  by  aspecific

* This work  was supported by National Institutes of Health GrantCA  26504, Albert Einstein Core Cancer Grant P3O-CA 1330, and agrant from the Lucille P. Markey Charitable  Trust. The costs ofpublication of this article were defrayed in part by the payment of

page charges. This article must therefore be hereby marked "aduer-tisernent" in accordance with 18U.S.C. Section 1734  solelyto indicatethis fact.

2 T32 CA09173.$Supported by National Institutes of Health Training Grant

Fellow  of the Leukemia Society of America.

II To whom correspondence should be addressed Dept. of Devel-opmental Biology and Cancer, Albert Einstein College of Medicine,1300 Morris Park Ave., Bronx, New York  10461.

The abbreviations used are: CSF-1, colony stimulating factor-1;PY, phosphotyrosine; aPY, anti-phosphotyrosine; CSF-lR, colonystimulating factor-1 receptor; FPLC,  fast  protein liquid chromatog-raphy; PBS, phosphate-buffered saline; PTPase, protein tyrosinephosphatase; PT P- lC , protein tyrosine phosphatase 1C; RP-HPLC,reverse phase high pressure liquid chromatography; PAGE, polyacryl-amide gel electrophoresis; SH2, src homology 2;TBS, Tris-bufferedsaline.

receptor (I), the c-fms proto-oncogene product (2) , whichpossesses tyrosine  kinase activity (3). Incubation  of rnacro-phages with CSF-1 causes non-covalent CSF-1 receptor (CSF-

1R) dimerization, activation, and tyrosine phosphorylationfollowed by the tyrosine  phosphorylationof several  primarilycytoplasmic proteins (4-7).  Theidentity of these proteins hasnot  been established. Their appearance is maximally stimu-lated by 60 s at 37 "C,but  differences in the kinetics of theirappearance  can be resolved at4 "C (6). They may  be phos-phorylated  directly  bythe activated CSF-IR or indirectly bynon-receptor tyrosine kinases that are activated as part  of  asignal transduction  process,  ortheir tyrosine phosphorylationmay  increase  due togrowth factor-induced inhibition of aprotein tyrosine phosphatase.

As few  of the tyrosine-phosphorylated proteins appear tostably associate with the CSF-1R: receptor-based  purifica-tion/cloning methods are not generally  applicable.  Thereforewe  have adopted the approach  of  identifyingthem directly bypurification  and  microsequencing.  Because  a-64-kDa proteinis markedly  tyrosine-phosphorylated  in  macrophages  in  re-sponse to CSF-1, we have focused our initial purificationefforts on the CSF-1-induced tyrosine-phosphorylated  pro-teins of  45-70 kDa. In this communication, we briefly  describemethods for the purification  ofthese proteins in  sequenceableamounts and show that the -64-kDa cytoplasmic proteinwhich is rapidly  tyrosine-phosphorylated  in  responseto CSF-1 is protein  tyrosine  phosphatase 1C (PTP-IC).

EXPERIMENTAL PROCEDURES

Cell Culture ,  Protein  Purifica tion,and Sequencing-Cells of theBAC1.2F5 macrophage line were cultured in 100-mm tissue culturedishes and stimulated with CSF-1  (human recombinant macrophage

colony stimulating factor, a gift from Chiron Corp.) at 4 "C in thepresence of 2 mM iodoacetic acid to increase the yield of phosphoty-rosine-containing proteins as described previously (4, 8). The cellswere then washed once with ice-cold phosphate-buffered saline (136mM NaCI, 3 mM KCI, 8 mM NaZHPO,, 1.5 mM KH zP 04 ,pH 7.4)(PBS), scraped in cold PBS containing 100pM sodium orthovanadate(Fisher Scientific), 100 p~ phenylmethylsulfonyl fluoride (Sigma)and 2 mM iodoacetic acid (Fluka), collected in centrifuge bottles, andpelleted at 400 X g for 4 min at 4 "C. Homogenization and thesubcellular fractionation were performed as described by Yeung andStanley  (9). Phosphotyrosine (PY)-containing  proteins from the cy-tosol of two thousand subconfluent cultures were prepared by affinitycolumn chromatography using anti-phosphotyrosine (aPY)antibody(10) coupled to Sepharose 4B (Oncogene Science). The affinity chro-matography was carried out with the buffer system of  Li et al. (8)except that 0.8% octyl glucoside (Boehringer Mannheim) was usedinstead of 0.5% Nonidet P-40  inthe last 3 washes of the column prior

to elution and in the elution buffer. Proteins eluted with 5 mM phenylphosphate  (Sigma) were concentrated to approximately 500pl byCentriprep 30 (Amicon) and  then   to100pl by Centricon 30 (Amicon).Crystalline guanidine-HCI (110 mg) (Pierce Chemical Co.), 94 p1 of 2M Tris-HC1 (Sigma),  pH8.5, and 1.3 pl  of P-mercaptoethanol (Pierce)

were added to yield  190p l of a  6M guanidine-HC1,lOO mM Tris-HCI,100 mM mercaptoethanol solution, p H 8.5. The proteins were  reducedand denatured by incubation for 2 h at 20 "C and then overnight at4 "C. The denatured phosphotyrosyl proteins were separated by sizeexclusion chromatography on a Superose S-6 column (10 X 300  mm,Pharmacia) in an FPLC system (Pharmacia)  in 6M guanidine-HC1,50 mM Tris-HC1,0.5%dodecyltrimethylammonium bromide (Sigma),and 100 mM mercaptoethanol, pH 6.5, a t room temperature with aflow rate of 0.25 ml/min. Fractions (0.3 ml) were collected and  theprotein in each fraction detected by silver staining (11) of gradient

K. L. Berg, unpublished results.

23447

8/8/2019 J. Biol. Chem.-1992-Yeung-23447-50

http://slidepdf.com/reader/full/j-biol-chem-1992-yeung-23447-50 2/4

23448 CSF-1 Induces P TP -I C Tyrosine  Phosphorylation in Macrophages

(7.5-17.5% acrylamide) sodium dodecyl sulfate-polyacrylamide gelelectrophoresis (SDS-PAGE) gels (12). Fractions  containing  proteinsof interest were  pooled and concentrated to100 pl (Centricon 10,Amicon) at 18"C. The concentrate was diluted with 1 ml of 6 M

guanidine-HCI, containing 0.8% octyl glucoside and concentratedagain to 100 pl twice more at 18"C. It was then acidified with 20%trifluoroacetic acid (Pierce) to <pH 2, diluted with an equal volume

of  0.1% trifluoroacetic acid, centrifuged to removed debris, and  in-jected into a C, reverse phase high performance liquid chromatogra-

phy (RP-HPLC) column (2.1X

30 mm, BU300, Brownlee). Thecolumn was washed and developed with a  gradient of 35-50% aceto-nitrile (Burdick and Jackson)  in 0.1% trifluoroacetic acid at roomtemperature for 30 min with a flow rate of  0.2 ml/min  in  aHP1090M

system (Hewlett Packard) equipped with a diode array detector.

Fractions (100 pl) were collected, and 2 p1 from each fraction were

used for SDS-PAGE and silver staining.  Fractions  containing theproteins of interest were pooled and either digested with trypsin

(sequencing-grade, Boehringer Mannheim) immediately after dilu-

tion or pyridylethylated and purified once more on the same column

(13). The tryptic digest was separated  on  aCS RP-HPLC column (2.1X 30  mm, RP300, Brownlee) using a  gradient of5 4 0 % acetonitrile

in 0.1% trifluoroacetic acid in 100 min with a flow rate of 0.05 ml/min a t room temperature. Absorption peaks a t 214 nm were collected

manually, and sequences of the peptides in these fractions weredetermined using an Applied Biosystems 477A protein sequenator. Adata base search was carried out  to compare these sequences with

those stored  in the GenEMBL data  bank using the Wisconsin Ge-

netics Computer Group (WGCG) sequence analysis package (14).

Western Blotting-Western blotting was performed as described(X) , with the following modifications: Immobilon membrane (poly-vinylidene difluoride, Millipore) was used and  the  transfer buffercontained 0.05% SDS  and10% methanol. The  transfer was carried

out  at30 V for 2  h, then overnight a t 80V a t 5 "C. For the detection

of PY-containing proteins the blot was blocked with 5%bovine serumalbumin  (ICN), in 10 mM Tris-HCI, 140 mM NaCI, pH 7.2 (Tris-

buffered saline, TBS)  at4 "C overnight. After washing the blot with

TB S containing 0.05% Nonidet P-40 (Sigma), it was incubated with

the peroxidase-conjugated anti- PY antibody  (ICN)  in T BS contain-ing 0.05% Nonidet P-40 for 2  h a t room temperature and developedwith enhanced chemiluminescence (ECL)  reagents(Amersham). The

tyPTP-1C Western  blotting was carried out  as above except t ha t5%nonfat dry milk replaced bovine serumalbumin, and peroxidase-

conjugated goat anti-rabbit antibody (Amersham) was used to detectth e bound rabbit LuPTP-1C antibody. The LuPTP-IC antiserum (a giftfrom Drs. Jiao-Ti Huang and Joseph Schlessinger, Department of

Pharmacology, New  York University Medical Center, New York) wasraised to a unique carboxyl-terminal peptide (KREEKVKKQRS-

ADKEKS).

RESULTS

To identify the  proteinswhich are tyrosine-phosphorylated

in the response of  BAC1.2F5 macrophages to CSF-1, affinity-

purified aPY-reactive proteins from CSF-1-stimulated (Fig.

1,lane 2 )and unstimulated (Fig.1,lane 1) cells  wereseparated

under reducing and  denaturing conditions by size exclusionchromatography.The fractions containing a highly stimulated

-64-kDa protein of interest (Fig. 1, lane 13 )and correspond-ing fractions from the unstimulated cells  (Fig.1,lane 12 )were

separately pooled and  further fractionated by RP-HPLC.Asshown in Fig. 1, the -64-kDa protein was  recovered  exclu-

sively  inthe fractions from stimulated cells (lanes20-23) and

was almost completely  resolved  from a  protein of -54-kDa

found in the fractions from unstimulated cells (lanes 4-9).Fractions run in lanes 20-23 (Fig. l),containing the -64 kDa

protein, were  pooled and pyridylethylated prior to rechroma-

tography on RP-HPLC. While relatively pure when examined

by SDS-PAGE and silver staining, the  amountof pyridyleth-

ylated protein recovered (approximately 7 pmol) was not

sufficient to obtain a definitive amino acid  sequence  from

tryptic peptides. A second preparation was obtained in which

th e separation of the -64-kDa  from the -54-kDa protein(approximately 30 pmol of each) was not achieved.  However,

as the major  difference  betweenthe stimulated and  unstimu-

1

10-

5 - ,

0 J J

.,

L -

1 3 5 7 9 11 I 3 IS 17 19 21 23

205 ' a

116 4

98

m

4!

31

,+-64

"54

-+""+-+-

FIG.1. Cc RP-HPLC fractionation of  the 45-70-kDapro-teins from the aPY-reactive fraction of cytosols fromBAC1.2F5 cells. Cells were incubated for 2  h at 4"C with (+) orwithout (-) 4.4  nM human recombinant CSF-1 priorto fractionation

as described under "Experimental Procedures." The upper panelshows portions of the  RP-HPLC absorption profiles together withthe acetonitrile  gradient(- - -) in the range 37.25-39.50%. The lowerpanel shows the silver-stained  SDS-PAGEof the  RP-HPLCfractionsaligned with the absorption profiles shown in the upper panel. Lanes1 and 2, LuPY-reactive fractions; lanes 12and 13,45-70-kDa proteinsfrom the Superose 6 FPLC separations of the aPY-reactive  fractions

that were applied to the  RP-HPLCcolumn.

lated cell fractions was the presence of the -64-kDa protein

in the fraction from stimulated cells, we sequenced the tryptic

fragments that were unique to  the  latterfraction. Both stim-ulated (Fig. 2 A , inset) and unstimulated (Fig.  2B,inset) cellfractions were individually digested with trypsin and the

resulting tryptic peptides separated by RP-HPLC (Fig.  2).

Eleven peaks, apparently unique to  the stimulated cell frac-

tion (Fig. 2 A , 1-11) were sequenced. Peaks 1-4 and 6-8

contained sequences that were highly homologous to se-quences within PTP-1C  (TableI), peaks 9-11 failed to yield

a sequence, and   thesequence from peak 5 was uninterpretable.Three peaks present in both unstimulated and stimulated

fractions (peaks A , B, and C,Fig.  2,panels A and B ) were

sequenced from each fraction. Sequences obtained from peaks

A, B, and  C from the stimulated fraction were identical tothose from the unstimulated fraction. Peak B from the  stim-

ulated fraction also contained a   PTP-1Csequence not presentin the unstimulated fraction (data not shown).

To confirm the identity of the -64-kDa protein, the aPY-

8/8/2019 J. Biol. Chem.-1992-Yeung-23447-50

http://slidepdf.com/reader/full/j-biol-chem-1992-yeung-23447-50 3/4

CSF-1 Induces PTP -1C Tyrosine

A 907 li”3

zc5:

38 4

28 30 40 50 60 70Tlmc ( m l n . )

B I

20 30 48 5 6 60 7 0 8 0

T lm e (min. 1

FIG. 2. C. RP-HPLC  chromatography  of  tryptic digests ofpooled  fractions from  theC4 RP-HPLC purification step. A,CSF-1-stimulated;€3, unstimulated. Insets show silver-stained SD S-PAG E lan es for sam ples that were subjected to tryptic digestion.

TABLEI

Comparison  of the  sequencesof trypticpeptidesfrom  the -64-kDaprotein  with  those  predicted  for  murinePTP-IC

SourceAmino acid

SeauencePTP-1C-64-kDa,  peak  1”

-64-kDa, peak 2

-64-kDa, peak 3

-64-kDa, peak  4

-64-kDa,  peak  6

-64-kDa, peak I

-64-kDa. Deak 8

PTP-1C

PTP-IC

PTP-1C

PTP-1C

PTP-1C

PTP-1C

WGLYSVTNSR

WGLYSVTNS

NILPFDHSR

NILPFXXS~

DSNIPGSDYINANYW

XSNIPGXDYXNANYW

IQNSGDFYDLYGGEK

IQNXXFYXLYGGEK

GEPWTFLVR

GXPXIFLV

AGFWEEFESLQK

AGFXEXF’XXLQ

ESLSQPGDFVLSVLNDQPK

ESLXIPGXFVLSVLN

a Peak  number,  Fig.lA.* Amino  acids  whoseidentities were  ambiguous  are  m arkedX.

reactive protein fraction of whole cell lysates, pooled fractionsfrom the size  exclusion chromatography in the region con-taining the -64-kDa protein, and the pyridylethylated,HPLC-purified, -64-kDa protein were Western-blotted andprobed with aPY and  aPTP-1C antisera (Fig.3 ) . The pyri-dylethylated purified -64-kDa protein (Fig. 3B, lane 11)was

shown to be immunoreactive to  aPTP-1C(Fig. 3A, lane 10).

Consistent with our observations that  the -64-kDa protein

can exist as a doublet during purification (Fig. 1,lanes 13 and

20-23) (possibly due to lengthy exposure to high concentra-tions of mercaptoethanol), the  aPTP-1Cantibody detected adoublet in that fraction (Fig. 3A, lane 9) , the faster moving

Phosphorylation in M acrophages 23449

A1 2 3 4 5 6 7 8 9 1 0

45-

15

- + - + - + - +Lrz PY J h ITP-ICJ

FIG. 3. aPY and aPTP-1C Western blots of fractions fromBAC1.2F5 cells. PanelA: lanes I , 2,6, and 7,aPY-reactive fractionsof whole cell lysates; lanes 3 ,4 ,8 , and 9, pooled 45-70-kDa fractionsfrom the Superose  6  chromatographystep; lanes 5 and 10,pyridyleth-ylated purified -64-kDa protein (pool of fractions from  lanes20-23

in Fig. 1). Panel B: lane 11,silver-stained SDS-P AGE of the pyridyl-ethylated purified -64-kDa protein. Cells were incubated for 2 h at4 “C with (+) or without (-) 4.4 nM human  recombinantCSF-1 prior

to fractionation as described  under  “Experimental  Procedures.”

component of which  was lost during reduction and pyridyle-thylation. In  theaPY-reactive fraction of whole  cell lysate, asingle aPTP-1C-reactive band (Fig. 3A, lane 7 ) co-migratedwith the  faster moving component of the doublet. Consistentwith the failure to detect the -64-kDa protein or  PTP-1C  inaPY-reactive fractions from unstimulated cells, the -64-kDabands were only observed in aPY Western blots of fractionsfrom stimulated cells (lanes 1-5). The -54-kDa band wasobserved in fractions from unstimulated cells but was alsogreatly increased in fractions from stimulated cells.  All bandsshown in the aPY Western blot were eliminated when theaPY used was pre-incubated with 1 mM phosphotyrosine

(data  not shown). These results confirm that  the -64-kDaprotein is PTP-1C  andindicate that   itis tyrosine-phosphoryl-ated in macrophages in response to CSF-1.

DISCUSSION

An -64-kDa protein that is one of the macrophage proteinsexhibiting marked CSF-1-induced tyrosine phosphorylationhas been identified as PTP-1C. The isolation of cDNA clonesencoding PTP-1C was first reported by Shen et al. (16) forhuman PTP-1C  and subsequently by other workers for thehuman (17, 18) and mouse  (18, 19) proteins. These cDNAsexhibit a high  degree of homology across species and encodea soluble tyrosine phosphatase of -68 kDa that contains two

highly conserved SI % homology 2 (SH2) domains. PTP-1CmRNA  is expressed in hematopoietic cells and  in some epi-thelial cell lines but  not in fibroblasts (17-20). Expressedrecombinant forms of the protein have protein tyrosine phos-phatase  (PTPase) activity (16-19) and  the capacity to bindtyrosine-phosphorylated proteins in vitro (16, 19). As thetyrosine-phosphorylated CSF-1R could not be detected inaPTP-1C immunoprecipitates of lysates from stimulatedcells: it is likely that  the-64-kDa protein we have identifiedas  PTP-1C is associated with other tyrosine-phosphorylatedproteins in the cytoplasm.

In a recent study, Perkins et al. (21) have reported that   theDrosophila gene corkscrew (csw) (22) encodes a putative non-

receptor protein tyrosine phosphatase containing two NH2-

terminal  SH2domains that shares -52% amino acid  sequencesimilarity with human PTP-1C. csw is maternally requiredfor the normal determination of cell fates at  the termini of

8/8/2019 J. Biol. Chem.-1992-Yeung-23447-50

http://slidepdf.com/reader/full/j-biol-chem-1992-yeung-23447-50 4/4

23450 CSF-1 Induces PT P-1 C Tyrosine Phosphorylation in Macrophages

the embryo. The determination of terminal cell fates is regu-lated by a receptor tyrosine kinase, torso. Genetic evidence

indicates that csw acts downstream of torso, in  concert with

the serine/threonine kinase D- raf to regulate the downstream

terminal gene encoding the transcription factor, tailless (re-

viewed in Ref. 23) .

Analogous to  the regulation of D-raf by torso, the CSF-1R

has been shown to regulate RAF-1 in BAC1.2F5 cells (24).

CSF-1  stimulation causes maximum protein  tyrosine phos-phorylation of proteins, including the -64-kDa protein (PTP-lC), within 30 s at 37 "C (6, 8),' followed  by an increase in

the serine phosphorylation of RAF-1 that  is  apparent  at 1.5

min but of maximum intensity by 5-15 min after CSF-1

addition (24). RAF-1 serine phosphorylation is temporally

correlated with activation of a  RAF-1associated serine kinase

activity (24). These  experiments  indicate tha t  the tyrosine

phosphorylation of PTP-1C precedes the serine phosphoryl-

ation  andactivation of the RAF-1 kinase. We have also shown

that epidermal growth factor receptor phosphotyrosine de-

phosphorylation does not affect the increase in RAF-1 serine

phosphorylation during the response to epidermal growth

factor (25), suggesting that  there  is  atleast one step between

growth factor receptor activation and phosphorylation andactivation of RAF-1. If tyrosine phosphorylation of PTP-1C

has a role in regulating its enzyme activity, either by intra-

molecular interaction of the SH 2 domain with a  tyrosine

phosphorylation site, as suggested for the src-related family

of nonreceptor tyrosine kinases (reviewed in Ref.  26) or by

modulating the interaction of PTP-1C with other  proteins,

then these observations would favor the  firstof three models

suggested by Perkins  et al.( X ) , in which csw acts  upstream

of D-rat. Future  studies on the associationof PTP-1C with

other tyrosine-phosphorylated proteins,  its role in mitogenic

signaling, and  the effect of tyrosine phosphorylation on i ts

catalytic activity should greatly increase our understanding

of  how this enzyme is involved in the early events regulated

by growth factors.This  report also briefly describes methods that can be used

to identify other  proteinsthat  aretyrosine-phosphorylated in

the growth factor response. The procedures described are

suitable for proteins possessing a molecular mass of less tha n

70 kDa. For proteins with molecular masses significantly

above 70 kDa, significant losses are encountered in the HPLCstep.

Acknowledgments-We thank Drs. Jiao-Ti Huang and JosephSchlessinger for kindly providing the  aPTP-1C antiserum, Dr.R a y -mond Frackleton for advice, and  the various members of the labora-tory (past  andpresent) who assisted in the preparation of cell extracts.The sequencing was performed at   theLaboratory for Macromolecular

Analysis of the Albert Einstein College of Medicine. We thank YuanShi for careful handling of the samples and interpretation of the  data.

REFERENCES

2. Sherr , C. J.,  Rhtenmier, C. W., Sacca, R., Roussel, M.  F., Look, A.T.,and1. Guilbert, L. J. and  Stanley, E.R. (1980)J.Cell Biol. 85 , 153-159

3. Rettenmier, C.  W., Chen,J. H., Roussel, M. F., and  Sherr, C.J. (1985)Stanley,  E.R. (1985) Cell 41 , 665-676

Science 2 2 8 . 3 2 0 - 3 2 24. Li,  W., a nds tan ley ,E.R. (1991) E M B O J. 10 , 277-2885. Bacearini, M., Li, W., Dello Sbarba , P., and  Stanley,E. R. (1991) Receptor

I ~~~ ~~~

6. Sengupta, A,, Liu, W.-K., Yeung,  Y.-G.,  Yeung,D. C.-Y., Frackelton, A. R. ,

7. Downing,J. R., Rettenmier, C. W. , and  Sherr, C.J. (1988) Mol.  Cell. Biol.

8. Li, W., Yeung, Y. G., and  Stanley,  E. R.(1991)J.Biol. Chem. 2 6 6 , 6 8 0 8 -

1,243-259

and  Stanley,  E.R. (1988) Proc. Natl.  Acad.  Sei.U .S. A . 85,8062-8066

8,1795-1799

&A14

9.

10 .

11 .12 .13 .

14 .

15 .

16.

Ye&-&  Y.G., and  Stanley,  E.R. (1990) in Receptor  Purification(Litwack,

Huhn, R. D., Posner, M. R., Rayten, S. I., Foulkes, J. G., and Frackelton,

Morrissey, J. H. (1981) Anal. Biochem. 117 , 307-310Laemmli, U . K. (1970) Nature 227 , 680-685Renlund, S., Klintrot, I. , Nunn, M., Schrimsher, J. L., Wernstedt, C., and

Devereux, J., Herbert, M., and  Smithies,0.(1984) Nucleic  Acids  Res.1 2 ,

Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci.

Shen, S.-H., Bastien, L., Posner, B. I., and Chrhtien, P. (1991) Nature 362,

G. , ed) pp. 315-328, Humana  Press, Clifton, NJ

A. R., Jr. (1987) Proc. Natl. Acad. Sei. U.S. A . 84,4408-4412

Hellmann, U. (1990)J.C h r o m t o g r . 512 , 325-335

387-395

U.S. A. 76 , 4350-4354

7 R f i " n Q

17 . Plutzky, J., Neel, B. G., and Rosenberg, R. D. (1992) Proc.  Natl.Acad. Sei.

18 . Yi, T.,Cleveland, J. L., and Ihle, J. N. (1992) Mol.  Cell.  Biol.12 , 836-84619 . Matthews, R. J., Bowne,  D.  B., Flores, E., and  Thomas,M.L. (1992) Mol.

20 . Yi, T.,Cleveland, J. L., and Ihle, J. N. (1992) Blood 78,2222-222821 . Perkins, L. A,, Larsen, I., and  Perrimon, N.(1992) Cell 70 , 225-23622 . Perrimon, N., Engstrom, L., and Mahowald, A. P. (1985) Genetics 1 1 1 ,

."" .""U.S.A. 89 , 1123-1127

Cell. Bid . 12 , 2396-2405

23-41

23 . St. Johnston,D. , and Nusslein-Volhard, C. (1992) Cell 68 , 201-21924 . Baccarini, M., Sabat ini, D.  M.,  App, H., Rapp,U . R., and  Stanley, E. R.

25 . Baccarini, M. , Gill, G. N., and  Stanley,  E. R.(1991) J. Biol. Chem. 2 6 6 ,(1990) EM BO J.9,3649-3657

26. Cantley, L. C., Auger, K. R., Carpenter, C., Duckworth, B., Graziani, A.,10941-10945

Kappeller, R., and Soltoff, S. (1991) Cell 6 4 , 281-302