Cellular Proteins Homologous to the Viral yes Gene Productt

MOLECULAR AND CELLULAR BIOLOGY, Aug. 1986, p. 2839-2846 0270-7306/86/082839-08$02.00/0 Copyright © 1986, American Society for Microbiology

Vol. 6, No. 8

Cellular Proteins Homologous to the Viral yes Gene Productt MARIUS SUDOL* AND HIDESABURO HANAFUSA The Rockefeller University, New York, New York 10021 Received 3 March 1986/Accepted 26 April 1986

We raised antibodies in rabbits against the amino-terminal portion of the viral yes protein produced in bacteria with the use of an expression vector based on the lac operon. The anti-yes serum thus obtained precipitated p9Wgag-yes from Yamaguchi 73 virus-transformed chicken embryo fibroblasts, and this immunoprecipitation was blocked by the purified antigen. The anti-yes serum did not recognize viral src, fps, or fgr proteins. Affinity-purified anti-yes immunoglobulin G (IgG) precipitated two proteins of 59 and 62 kilodaltons from lysates of normal chicken embryo fibroblasts. Two-dimensional tryptic peptide mapping showed that these proteins are closely related to p9A09-yes and that they are different from pp60c -S. Similar to P90919-y", the 59- and 62-kilodalton proteins were phosphorylated exclusively on tyrosine in an in vitro kinase reaction, whereas in vivo they were phosphorylated on serine and, to a lesser extent, on tyrosine as well. Expression of the 59- and 62-kilodalton proteins, determined by the immune complex kinase assay, was relatively high in brain, retina, kidney, and liver. The presence in normal chicken embryo fibroblasts and in chicken kidney of two transcripts, 3.7 and 3.9 kilobases in length, that hybridize with a yes-specific DNA probe, as well as the two proteins recognized by anti-yes IgG, suggests either differential splicing of cellular yes gene transcripts or the existence of another yes-related gene.

Yamaguchi 73 (Y73) and Esh avian sarcoma viruses are replication-defective retroviruses that were independently isolated from spontaneous tumors in chickens (16, 42). Both viruses induce sarcomas in vivo and transform chicken embryo fibroblasts (CEF) in vitro (11, 18). The transforming properties of these viruses are directed by cell-derived sequences called yes (33, 43). The yes transforming genes of Y73 and Esh avian sarcoma virus encode pg9yag-yes and P80Iag-Yes phosphoproteins, respectively (10, 18). The amino acid sequence of the p90yag-yes fusion protein predicted from the DNA sequence of the Y73 genome (19) shows a high degree of homology with pp60v-rc, the transforming protein of Rous sarcoma virus (31, 40). The greatest extent of homology exists in the carboxy-terminal domains, whereas the amino-terminal portions of these two oncogene proteins are different (19). Our interest in cellular yes (c-yes) proteins(s) stems from two initial observations. Studies at the RNA level have shown that c-yes expression in various chicken tissues is one to two orders of magnitude higher than that of other known c-onc sequences from the avian sarcoma virus family, and that the pattern of tissue-specific expression of c-yes shows elevated levels of c-yes mRNA in kidney and brain (33). In this report, we describe the characterization of two cellular proteins of 59 and 62 kilodaltons (kDa) immunoprecipitated from normal chicken cells with antibodies raised against the amino-terminal portion of the viral yes (v-yes) protein expressed in bacteria. By a number of criteria we show that these proteins are closely related to the v-yes gene product and differ from pp60csrs. We describe the pattern of tissue-specific expression of these proteins and show the presence of two transcripts in CEF and chicken kidney which hybridize with a yes-specific DNA probe.

MATERIALS AND METHODS Cell culture and viruses. Primary and secondary CEF were prepared and maintained as described previously (15). A virus stock of Y73 was a gift from S. Kawai, University of Tokyo, Tokyo, Japan. The Schmidt-Ruppin strain of Rous sarcoma virus, Fujinami sarcoma virus, and UR2 virus were from laboratory stocks. Mink cells (252-M2) permanently transformed by Gardner-Rasheed feline sarcoma virus and expressing the v-fgr oncogene product were kindly provided by K. Robbins, National Cancer Institute, Bethesda, Md. Plasmid construction. All methods for DNA and RNA manipulations were as described previously (24, 34, 38). The expression vector pMR/YA (see Fig. 1) is a recombinant between the pMR100 plasmid ([13] provided by M. Rosbash, Brandeis University, Waltham, Mass.) and part of the pY73 plasmid containing Y73 proviral DNA subcloned into the pBR322 (8). To construct an insert coding for v-yes-specific sequences, a 2,009-base-pair (bp) DNA fragment was isolated from a HindIII digest of the pY73 plasmid and digested with Bal 31 nuclease to generate a population of fragments of 400 to 450 bp. These fragments were ligated into the SmaIcloning site of the pMR100 vector. P-Galactosidase-negative bacteria (Escherichia coli strain LG-90) transformed with the recombinant vectors were screened for 3-galactosidase activity with MacConkey agar plates as described previously (13). From 3 x 103 clones, 18 positive clones were isolated, one of which contained an insert of approximately 450 bp. Direct DNA sequence analysis was used to map the boundaries and to determine the length of the insert. Protein biochemistry. Labeling of cultures with [3H]leucine and 32Pi and immunoprecipitations of proteins were essentially as described previously (7). Cell extracts were prepared in RIPA buffer (3, 7) containing, in addition to 10-7 M Trasylol (Mobay Chemical Co., New York, N.Y.) (7), 10-6 M leupeptin and 10-6 M antipain (Sigma Chemical Co., St. Louis, Mo.). The immune complex kinase assay was performed as described previously (3, 7). Affinity reagents were made by coupling purified proteins to CNBr-activated

* Corresponding author. t Dedicated to Samuel Beckett.

2839

2840

SUDOL AND HANAFUSA

MOL. CELL. BIOL.

A

B C

D SRC FIG. 1. Vector expressing part of the v-yes gene and characterization of the insert sequences. (A) pMR/YA expression vector derived from the parental pMR100 plasmid (13) by inserting into its SmaI-cloning site a Bal 31-generated DNA fragment of the v-yes gene, as described in the text. The ampicillin resistance gene is indicated with amp. The two small boxes indicated with arrows indicate two lac promoters placed in tandem in the same orientation toward cI. cl and lacl indicate parts of the genes coding for lambda and lac represors, respectively. lacZ indicates part of the ,-galactosidase gene. Y73 indicates here part of the v-yes gene that corresponds to the amino-terminal portion of the protein. (B) DNA sequence in the vicinity of the cloning site was verified by direct sequence analysis. Numbers indicate the first and the last nucleotide of the 434-bp yes DNA insert. Numbering is as described previously (19). Solid triangles delineate codons of the reading frame set by the frame of the lambda cI gene. (C) Schematic comparison of the amino acid sequence of the expressed fragment of v-yes with the corresponding fragment of v-src. The sequences were aligned to give the maximum homology (19). The asterisk points to a bar with a width that corresponds to a single amino acid homology between the indicated fragments of p9gyag-yes and ppWc-src. (D) Schematic representation of v-yes and v-src gene products. Their relative positions reflect the maximum homology alignment. Broken lines show the position of the expressed fragment. H-H indicates a 251-bp HaeIII DNA fragment that is unique for v-yes, and P-P indicates a 1,149-bp PstI fragment of yes covering part of the gene coding for the protein kinase domain. Both DNA fragments were used as probes in the hybridization studies, results of which are shown in Fig. 10 and described in the text.

Sepharose 6MB as described previously (30, 39). Purification of fusion proteins from bacterial lysates was performed essentially by the procedure described previously (12), except that all of the buffers used were supplemented with the same mixture of protease inhibitors described above for RIPA buffer. Briefly, in the first step of the purification we used affinity chromatography with 3-galactosidase antibodies coupled to Sepharose. Proteins eluted from the column were concentrated and further fractionated on a preparative sodiumn dodecyl sulfate-polyacrylamide gel (20), followed by the elution of the separated proteins by the method of Hager and Burgess (14). The radioiodination and performic acid oxidation of proteins in single polyacrylamide gel slices, tryptic digestion, and peptide mapping were performed as described previously (6, 39). Two-dimensional analysis of phosphoamino acids was carried out as described previously (7), except that constantly boiling HCI was used and the time of hydrolysis was extended to 150 min. Animals, antisera, and antibodies. All rabbits used for producing antibodies were bled prior to injection of antigen

to obtain the preimmune sera used. Antisera against fusion proteins were obtained by injecting 5 to 15 jig of the purified antigen (diluted in 90 RI of phosphate-buffered saline mixed with 10 ,ul of complete Freund adjuvant) into each of the popliteal lymph nodes. Starting 3 weeks later the rabbits were injected intramuscularly each week for the next 5 weeks with 100 to 200 ,ug of the purified antigen in incomplete Freund adjuvant. Collection of sera began 6 weeks after the initial injection. Antibodies against the affinitypurified 3-galactosidase (Worthington Diagnostics, Freehold, N.J.) were raised in rabbits by a previously published procedure (39). Serum from a rabbit bearing a tumor induced by the Schmidt-Ruppin strain of Rous sarcoma virus (subgroup D) was obtained by the procedure of Brugge and Erickson (3), as described by Karess and Hanafusa (17). This serum (TBR D-7-9) had a high titer against gag proteins and was used to purify p9Ygag-yes from Y73-transformed CEF. The p9gyag-yes, as well as other viral proteins eluted from the TBR D-7-9 Sepharose column were concentrated and coupled to Sepharose resins as described previously (30). This affinity column was used to purify and concentrate

PROTEINS HOMOLOGOUS TO THE VIRAL yes GENE PRODUCT

VOL. 6, 1986

B

A ....

C

A B

D

C D E

F G

2841

H I J

-

..-

0

.. a"

1- 200

-200 -97

RNNMW

-

97 -68

- 68

:- 4

3

FIG. 2. Sodium dodecyl sulfate-polyacrylamide gel analysis of hybrid proteins. Proteins were analyzed on an 8% polyacrylamide gel (20) and then visualized by Coomassie blue staining. Lanes A and B, total proteins from LG-90 strain of Escherichia coli transformed by recombinant vectors pMR/YA (A) or pMR200 (B). Single-cell colonies selected on MacConkey agar plates were later grown in the presence of the lac inducer and lysed as described previously (12, 13). Upper arrow, cI-yes-lacIZ tetrabrid protein; lower arrow, either a degradation product of the cI-yes-lacIZ protein or a fusion protein initiated at one of the internal ATG sites of the yes insert; A, cI-lacIZ fusion protein. Indicated proteins are recognized by anti-p-galactosidase IgG and are absent in LG-90 bacteria grown in the absence of lac inducer (data not shown). Lanes C and D, hybrid proteins purified by immunoaffinity chromatography with antibodies against 3-galactosidase and by preparative gel electrophoresis (14). (C) Purified cI-yes-lacIZ fusion protein (10 ,ug); (D) the purified cI-lacIZ protein (10 ,ug). Numbers at the right side of the figure denote the molecular masses of marker proteins in kilodaltons.

anti-yes immunoglobulin G (IgG) from serum against the cI-yes-lacIZ fusion protein. Monoclonal antibody M327, which recognizes pp60-csrc (22), was obtained from J. Brugge, State University of New York, Stony Brook. RESULTS Tetrabrid fusion protein containing part of the v-yes oncogene product is expressed in bacteria. P-Galactosidasenegative bacteria transformed with the recombinant vector pMR/YA (Fig. 1) and with the plasmid pMR200 ([13] derivative of the parental pMR100 with corrected reading frame), and grown in the presence of the lac inducer, each produced a fusion protein of the expected molecular weight (Fig. 2, lanes A and B). The tetrabrid fusion protein cI-yes-lacIZ of pMR/YA was approximately 20 kDa larger than the initial cI-lacIZ fusion protein (Fig. 2), which is consistent with a DNA insert of 434 bp (Fig. 1). In the lysates of the pMRIYAtransformed bacteria, another protein of lower molecular weight was detected (Fig. 2, lane A). This protein may represent either a fusion product initiated at one of two internal ATG sites of the yes insert which are in-frame with the 13-galactosidase gene or it may be a degradation product of the cI-yes-lacIZ protein. Purified fusion proteins (Fig. 2, lanes C and D) were injected into popliteal lymph nodes of rabbits, and antisera were obtained as described above. For simplicity, antiserum against the cI-yes-lacIZ fusion protein is designated here as anti-yes serum. Antiserum against the amino-terminal portion of the v-yes protein specifically recognizes P9Otag-yes. A protein of the

-43

-26 FIG. 3. Immunoprecipitation of viral p90ag-yes fusion protein by anti-yes serum. Normal and virus-infected cells were labeled for 4 h with [3H]leucine and lysed in RIPA buffer, and lysates were incubated with an appropriate antiserum. Proteins from lysates of normal CEF (lane A), of CEF infected with Y73-associated virus (lane B), and of CEF transformed with Y73 virus (lanes C, D, and E) were immunoprecipitated with anti-yes serum (lanes A, B, and E), preimmune serum (lane C) or serum against cI-lacIZ protein (lane D). Reaction products of the immune complex kinase assay performed on parallel samples are shown in the same order as lanes A to E from lanes F to J. Immunoprecipitates were analyzed on a 10o polyacrylamide gel and visualized by fluorography (lanes A to E) (exposure time, 6 days) or by autoradiography (lanes F to J) (exposure time, 1 day). Expected mobility of the p9ggag-yes polyprotein is indicated by the arrow. Markers are as described in the legend to Fig. 2.

expected mobility characteristic for p9ggag-yes was immunoprecipitated with anti-yes serum from lysates of [3H]leucine-labeled CEF transformed with Y73 virus (Fig. 3, lane E). Equivalent amounts of preimmune serum or anti-

FIG. 4. Test for the specificity of anti-yes serum. Virus-infected cells were labeled and lysed as described in the legend to Fig. 3. Lane A, lysate from Y73-transformed CEF was incubated with anti-yes serum after preincubation for 30 min with 20 ,ug of the purified antigen cI-yes-lacIZ. Proteins from lysates of CEFtransformed with Y73 (lane B); with Schmidt-Ruppin strain of Rous sarcoma virus (lane C); with Fujinami sarcoma virus (lane D); with UR2 virus (lane E); and of the mink cell line 252-M2 (expressing the v-fgr oncogene product) (lane F) were immunoprecipitated with anti-yes serum. Immune complex kinase assay was performed on parallel samples shown in the same order as lanes A to F from lanes G to L. Other details are as described in the legend to Fig. 3.

2842

SUDOL AND HANAFUSA A B

C

D

E

MOL. CELL. BIOL.

F

G

A -68

82 2~~~~~9

29

8

D

o59

E 443

9 10

L~~-43 1"^l < FIG. 5. Immunoprecipitation of two cellular proteins from CEF with anti-yes IgG. Proteins from lysates of normal CEF (lanes B to G) and of Y73-transformed CEF (lane A) were first immunoprecipitated with the appropriate preimmune serum, and the supernatants were then immunoprecipitated with affinity-purified anti-yes IgG (lanes A, B, D, F, and G) or with the equivalent amount of purified IgG directed against cI-lacIZ polyprotein (lanes C and E). Lanes A and B, cells were labeled with 32pj; lanes C and D, cells were labeled with [3H]leucine, as described in the text; lanes E, F, and G, results of the immune complex kinase assay; lane G, to resolve the closely spaced doublet, a 50-cm-long polyacrylamide gel was used (only part of the gel is shown, as indicated with solid triangles. Exposure time for fluorography was 10 days, and for autoradiography exposure time was less than 2 days. Numbers to the sides of the gels are molecular masses in kilodaltons. serum

against cI-lacIZ protein did

not

recognize P90-4y-es.

All the sera (including the preimmune serum) recognized two proteins of 200 and 40 kDa in Y73-transformed CEF. Antiyes serum did not precipitate those proteins from normal CEF or from CEF infected with Y73-associated virus. Specificity of the p9ygag-yes recognition by anti-yes serum was

A

5-P T-P

D

TP

E

4: S-P

FIG. 6. Identification of phosphoamino acids in phosphorylated proteins. Partial acid hydrolysates of 32P-labeled proteins were separated in two dimensions: electrophoresis at pH 1.9 from left to right was the first dimension and electrophoresis at pH 3.5 from bottom to top was the second dimension (7). Positions of the internal phosphoamino acid markers, indicated by ninhydrin staining, are indicated with dotted circles. Standard phosphoaminoacids: S-P, phosphoserine; T-P, phosphothreonine; Y-P, phosphotyrosine. Stars indicate positions of free phosphate. (A) 59-kDa protein; (B) 62-kDa protein; (C) p9(yag-y,Y,s protein phosphorylated in vitro and analyzed as described in the text. (D) 59-kDa protein; (E) 62-kDa protein; (F) p9Q(ag-yes protein phosphorylated in vivo and analyzed under identical conditions as gels in panels A to C.

F

6

6~~~~~~ 2

9

10

8

.162

C

2

4 0050~~

8

2

191

10

FIG. 7. Tryptic peptide maps of P90,ag-yes, pp60c-src, and the 59and 62-kDa proteins. [3H]leucine-labeled proteins were immunoprecipitated from Y73-transformed CEF (p90g'g-yes) and from normal CEF, separated on a long, 7.5% polyacrylamide gel and located by fluorography. The proteins in the single gel slices were oxidized with performic acid, labeled with radioactive iodine, and digested with L-(tOsylamido 2-phenyl)ethyl chloromethyl ketonetreated trypsin, as described in the text. All of the proteins were processed under the same conditions. Tyrosine-containing peptides labeled with 1251 were separated in the first dimension by electrophoresis at pH 4.5 on a thin-layer chromatography plate (from left to right) and by ascending chromatography in the second dimension (from bottom to top). (A) p9(,yg-yes; (B) 62-kDa protein; (C) p9Wyag-yes and 62-kDa protein; (D) pp60-src (E) 59-kDa protein; (F) schematic summary of the results. All circles (open and closed) show major tryptic peptides common for p90gag-yes and the 62-kDa protein. Longer exposure times show the presence of two additional peptides (labeled 5 and 6) in the digest of pg(yag-yes. Closed circles show peptides common to p9g,ga-yes and the 62- and 59-kDa proteins. Solid triangles point to peptides (labeled 1 and 2) common to pp60sc-rc, p9ryag-yes, and the 62-kDa protein. Data from other maps not shown here, with mixed samples (pp60`icrc + 62-kDA protein; 59-kDa protein + p90gag-yes) were used to confirm the comigration of the analyzed peptides.

demonstrated by competing the immunoprecipitation of the metabolically labeled v-yes protein with the purified antigen (Fig. 4, lane A). In addition, anti-yes serum did not recognize viral src, fps, and fgr proteins (Fig. 4). However, when the gel showing the results of the immune complex kinase assay was overexposed, a band migrating at the position of 68 kDa was detected in the immunoprecipitate from lysates of UR2-transformed CEF. One-dimensional V8 protease peptide mapping of this phosphoprotein showed an identical pattern of peptides to that of the p685a5ros (data not shown). This finding is surprising to us because no significant homology exists between the v-ros protein and the expressed fragment of cI-yes-lacIZ. Anti-yes IgG precipitates from normal cells two phosphoproteins that are structurally related to the v-yes protein. We isolated from anti-yes serum the anti-yes IgG fraction using an affinity column constructed by coupling to Sepharose resins viral proteins, including p90oa5-yes, immunoprecipitated from lysates of Y73-transformed CEF with TBR D-7-9 serum (see above). Anti-yes IgG precipitated p90gag-Yes from lysates of Y73-transformed CEF labeled with 32Pi and two proteins of 59 and 62 kDa from lysates of 32P-labeled normal CEF (Fig. 5, lanes A and B). Overexposed autoradiograms showed weak bands of the 59 and 62-kDa doublet in Y73-transformed cells. Apparently, higher affinity of the anti-yes IgG toward the viral protein than toward the cellular antigens, the relatively high concentration of the v-yes protein in transformed cells, and also conditions of immu-

PROTEINS HOMOLOGOUS TO THE VIRAL yes GENE PRODUCT

VOL. 6, 1986

noprecipitation in which the antibody was not in excess are responsible for the absence of the 59- and 62-kDa doublet in the immunoprecipitate from Y73-transformed cells. The 59and 62-kDa closely spaced doublet was also detected in immunoprecipitates from normal cells labeled with [3H]leucine and in the immune complex kinase assay (Fig. 5, lanes D and F). In control experiments we used equivalent amounts of IgG directed against the cI-lacIZ fusion protein (Fig. 5, lanes C and E). Viral p9,yag-yes and both cellular proteins of 59 and 62 kDa were phosphorylated in vitro exclusively on tyrosine. These proteins were phosphorylated in vivo on serine and to a lesser extent on tyrosine as well (Fig. 6). Two-dimensional tryptic peptide mapping revealed that the 62-kDa protein is homologous to p90gag-yes, whereas the 59-kDa protein shares only partial homology (Fig. 7). All of the proteins described above have peptide maps that are different from that of pp6csrc. However, in the pp60c-src tryptic digests we detected two peptides with the same migration as in the digest of the pg9yag-yes and 62-kDa B

A

y S

y

C ys

D

97-

F

G

ys

ys

E

yE

yS

-

6843-

A

2843

B

200-

9768.W.-

43-

FIG. 9. Immune blot analysis of the relative amounts of pp6O-src and yes-related proteins in brain and retina. Equal amounts of protein (200 ,ug) extracted from the chicken brain (telencephalon) or from chicken embryonic retina were immunoprecipitated with antiyes IgG (lanes C and D) or with monoclonal anti-src IgG (lanes A and B), and the immunoprecipitates were fractionated on a 10% polyacrylamide gel. Proteins were electrophoretically transferred from the gel to a nitrocellulose filter and probed with 1251I-labeled antibodies as described previously (2, 4). Lanes A and C, retina; lanes B and D, brain. Arrows indicate expected migration of the 59and 62-kDa protein doublet. Numbers to the left to the gel are molecular masses in kilodaltons.

26-

H I y s y 5

J

y U-w

s

K y s

i

r

k s

y

s

-68

9768-

S.k

43-

26-

43

FIG. 8. Expression of the 59- and 62-kDa proteins in various chicken tissues; comparison with the expression of pp60`rc. One hundred micrograms of protein extracted from heart (lane A), lung (lane B), thymus (lane C), kidney (lane D), liver (lane E), bursa (lane F), spleen (lane G), muscle (lane H), bone marrow (lane I), telencephalon of the brain (lane J), and retina (lane K) was incubated with an excess of anti-yes IgG (y) or monoclonal antibody against src (s); and the immune complex kinase assay was performed. Immunoprecipitations from brain (lane j) and retina (lane k) were analyzed also on a long gel (as in Fig. 5) to improve separation of the protein doublet. An indistinct band of approximately 55 kDa was reproducibly immunoprecipitated from liver and kidney and to a lesser extent from other tissues. It is possible that this protein(s) represents a degradation product(s) of the immunoprecipitated protein(s). With the exception of retina, which was isolated from 14-day-old embryos, all other tissues were from 2-week-old animals. Two organs from two individual animals were taken for the assay. All gels shown were from one experiment in which samples were processed under identical conditions. Gels were exposed for 3 days. Numbers to the sides of the gels are molecular masses in kilodaltons.

proteins. In addition, pp60c-src seems to differ from the two cellular proteins immunologically. When normal CEF lysates were precipitated several times with monoclonal antibody against src until antigen was no longer detectable, we could easily precipitate 59- and 62-kDa proteins from the precleared lysates (data not shown). Expression of yes-related proteins is relatively high in brain, retina, kidney, and liver. We immunoprecipitated proteins from lysates of various chicken tissues (12-day-old chickens) and from embryonic retina (14-day-old embryos) with antiyes and anti-src IgG. Equal amounts of protein (100 ,ug) were used in the precipitations, and the immune complex kinase assay was performed (Fig. 8). High kinase activity was detected in immunoprecipitates with anti-yes IgG from brain, retina, kidney, and liver (Fig. 8, lanes D, E, J, and K); whereas low levels of kinase activity were detected in muscle, heart, bone marrow, and spleen (Fig. 8, lanes A, G, H, and I). The highest level of kinase activity detected for c-src protein was in retina. In the cerebral part of the brain (telencephalon) the level of pp6c-Jsrc kinase activity was moderate (Fig. 8, land J). The levels of expression of the c-src and c-yes protein in brain and retina were confirmed by immune blot analysis (Fig. 9). Two transcripts are detected by a yes-specific DNA probe in chicken kidney and CEF. Northern blot analysis of poly(A)+ RNA from chicken kidney and normal CEF revealed two transcripts of 3.7 and 3.9 kilobases hybridizing with a probe (a 251-bp HaeIII fragment; Fig. 1) specific for the v-yes sequence (Fig. 10, lanes B and C). Under the same conditions of hybridization stringency (37, 41), a probe corresponding to the v-yes kinase region (P-P; see Fig. 1) hybridized to two transcripts of the same size (data not shown). Using the probes described above, as well as a probe specific for v-src (40) and v-ros (27), we examined their hybridization

SUDOL AND HANAFUSA

2844

A B C D

r

MOL. CELL. BIOL.

A B C

D

_

-28S

-18 S

FIG. 10. Northern blot analysis of mRNA from normal and Y73 virus-transformed cells. Total RNA was isolated from tissues and

cultured cells by the procedure of Lizardi and Engelberg (23). Poly(A)+ selection

was as

described previously (1). Lanes B, 15 ,ug

of poly(A)+ RNA from chicken kidney (2-week-old chicken), lanes C, 20 ,ug of poly(A)+ RNA from CEF; lanes D, 20 ,ug of poly(A)+ RNA from Y73-transformed CEF, Lanes A, control, 20 ,ug of total RNA from E. coli. A 251-bp HaeIII fragment (Fig. 1) was labeled with

32P by nick-translation and used

as

a

probe. Right panel,

time, 1 day; left panel, the same gel exposed for 7 days. Arrows indicate two transcripts of approximately 3.9 kilobases (upper) and 3.7 kilobases (lower). 28S and 18S rRNA markers were run in a parallel lane. Formaldehyde-formamide gel (run in a vertical position), as described previously (24), was used. exposure

to digested chicken genomic DNA. The complexity of the pattern on Southern blots (data not shown) does not allow a conclusion to be drawn as to whether one or two genes code

for the observed yes-related transcripts. DISCUSSION

We described the preparation of anti-yes antibodies and the identification of the product of the c-yes gene in various tissues of normal chickens. The following aspects of the work deserve brief comment: (i) the choice of DNA fragment of v-yes for insertion into the pMR100 vector and its expression in bacteria; (ii) the presence of two proteins related to p90gag-yes and the existence of two transcripts that hybridize with the yes-specific DNA probe; (iii) evaluation of the relative levels of yes-related proteins in various chicken tissues. Three parameters were imposed on the DNA chosen to be expressed in the pMR100 vector. We wanted the expressed sequence to be yes-specific at the protein level and not to share homology with other known onc proteins, to recognize in normal cells the same transcript(s) that are recognized by other DNA fragments derived from different parts of the v-yes gene, and to be no longer than 500 bp. The DNA fragment of the v-yes gene used for expression in the pMR100 vector codes for a polypeptide that maps within the amino-terminal domain of the yes region of p9ygag-Yes (Fig. 1). The amino acid sequence of the expressed polypeptide shares some short fragments of homology, the longest one being of 10 amino acids, with pp60'csrc, but overall is significantly different from the corresponding sequence of the src protein even when these sequences are aligned for maximum homology (19). The extent of homology for these amino-terminal fragments is 36%, whereas the percentage of identical amino acids in the carboxy-terminal domains of

pp60c-src and p9Voag-yes is above 80% (19). Although the v-fgr protein from mink 252-M2 cells was not immunoprecipitated by the anti-yes serum, we cannot exclude the possibility that the expressed fragment of the v-yes protein is homologous to the chicken c-fgr protein. This is primarily because v-fgr is linked to the actin sequence at its amino terminus in the feline sarcoma virus genome (25, 26, 28), whereas c-fgr is not fused with actin sequences and may share homology with the amino-terminal region of c-yes. A published sequence of the human c-fgr protein in this region (29) indicates an homology of 20 of 24 amino acids in the carboxyl end of the v-yes polypeptide expressed as cI-yes-lacIZ. A probe derived from the 5' end region of the inserted v-yes DNA recognized the same RNA species in normal and Y73-transformed CEF, as did the DNA probe corresponding to the protein kinase domain of the v-yes protein. Thus, the yes DNA fragment expressed in the pMR/YA plasmid does not represent sequences derived from another cellular gene and fused to the 5' end of the yes sequence. In the selection of the yes-specific DNA to be expressed in bacteria we avoided sequences that map upstream of the expressed fragment because these contain long stretches of pyrimidines that resemble some of the intronic or noncoding sequences. We also deliberately chose the length of the expressed DNA to be shorter than 500 bp to avoid comigration of the fusion protein with the beta and beta' subunits of the bacterial RNA polymerase (34). The presence in normal cells of two protein species recognized by anti-yes IgG, and of two transcripts hybridizing with yes-specific DNA probes, could reflect the existence of two yes genes with slightly different structures. Alternatively, this could reflect the expression of one gene, which involves differential initiation or splicing of its RNA transcript or alternate polyadenylation sites. Other possibilities explaining the doublets include limited proteolysis of the 62-kDa c-yes protein which generates a discrete 59-kDa degradation product, or the cross-reactivity of anti-yes IgG with the product of another proto-oncogene, for example, c-fgr. When feline v-fgr DNA (26) was used to probe chicken DNA on a Southern blot under stringent conditions of hybridization, no signal was detected; and relaxed conditions of hybridization resulted in unreadable patterns (data not shown). The complexity of the pattern on the Southern blots revealed by hybridizations with viral yes, src, and ros DNAs (of chicken origin) confirmed previous results of separate loci for the corresponding c-onc genes (21, 33, 43) but could not provide any clues that would differentiate between the presence of one or two copies for the c-yes gene. It is of interest that two human c-yes loci have been located, one on chromosome 18 and another, which may be a pseudogene, on chromosome 6 (32, 44). The possibility that the protein doublet is due to proteolytic degradation of the higher molecular weight form cannot be excluded. Consistent with that speculation are data from peptide mapping showing that the set of peptides for the 59-kDa protein is a subset of the peptides shown for the 62-kDa protein (Fig. 7). However, in all protein manipulations we used a mixture of inhibitors covering a wide spectrum of protease acitivities, and variations in the duration of lysate preparations or immunoprecipitation procedures did not affect in any reproducible way the relative ratio of the higher and lower Mr forms. It is likely that the 62-kDa protein is the product of the c-yes gene because a high degree of structural similarity to the v-yes protein is shown by the peptide mapping. At present we cannot draw any definitive conclusions on the possibilities presented above. Analysis of

PROTEINS HOMOLOGOUS TO THE VIRAL yes GENE PRODUCT

VOL. 6, 1986

the c-yes genomic and cDNA clones, as well as partial protein sequencing of the purified proteins, should provide the needed answers. Results of the screening of various chicken tissues by the immune complex kinase assay with anti-yes IgG show a high level of expression of yes-related proteins in brain (telencephalon), retina, kidney, and liver and a very low level of expression in muscle and heart. The profile of expression of the c-yes protein in various tissues was clearly different from that of the c-src protein (5; see above). Using the immune complex kinase assay, we screened different parts of the chicken brain from 2-week-old animals and found that the cerebellum and optic lobes contain high levels of yes-related proteins (data not shown). Telencephalon, diencephalon, medulla, and spinal cord showed relatively low levels of kinase activity. A similar profile of expression in brain was observed for pp60-csrc. Relatively high expression of pp6fJ-src in neural tissues has been shown (5, 9, 35, 36). Histochemical staining is under way to map with higher resolution the areas of the brain that express these proteins. In general, the results obtained confirm the results of the previous analysis at the RNA level (33); however, more precise measurements are necessary to quantitate the expression at the protein level. ACKNOWLEDGMENTS We thank M. Young and G. Gasic for their suggestion on the choice of the expression vector, M. Rosbash for the pMR100 and pMR200 vectors as well as for the LG-90 strain of bacteria, S. Kawai for the Y73 virus, M. Yoshida for lambda-Y73-lla DNA, K. Robbins for Gardner-Rasheed feline sarcoma virus DNA and mink 252-M2 cells, J. Brugge for monoclonal antibody M327 and C. Grandori for purifying that antibody, E. Kolb for carrying out lymph node injections, T. Hanafusa for help in preparation of retina, A. Alvarez-Buylla for dissecting chicken brains, W. Neckameyer and J. Smith for help with dissections, L.-H. Wang and W. Neckameyer for the Southern blot, and E. Garber for advice on protein work. Our thanks are extended to R. Jove and E. Garber for valuable comments on this manuscript. This study was supported by Public Health Service grant CA 14935 from the National Cancer Institute. M.S. is supported by Damon Runyon-Walter Winchell Cancer Fund Fellowship DRG 737. LITERATURE CITED 1. Aviv, H., and P. Leder. 1972. Purification of biologically active globin messenger RNA by chromatography on oligothymidylic acid-cellulose. Proc. Natl. Acad. Sci. USA 69:1408-1412. 2. Bolton, A. E., and W. M. Hunter. 1973. The labelling of protein to high-specific radioactivities by conjugation to a '251containing acylating reagent. Biochem. J. 133:529-539. 3. Brugge, J. S., and R. L. Erickson. 1977. Identification of a transformation-specific antigen by an avian sarcoma virus. Nature (London) 269:346-348. 4. Burnette, W. N. 1981. "Western blotting": electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Anal. Biochem. 112:195-203. 5. Cotton, P. C., and J. S. Brugge. 1983. Neural tissues express high levels of the cellular src gene product pp6Ocsr. Mol. Cell. Biol. 3:1157-1162. 6. Elder, J. H., R. A. Pickett H, J. Hampton, and R. A. Lerner. 1977. Radioiodination of proteins in single polyacrylamide gel slices. J. Biol. Chem. 252:6510-6515. 7. Feldman, R. A., T. Hanafusa, and H. Hanafusa. 1980. Characterization of protein kinase activity associated with the transforming gene product of Fujinami sarcoma virus. Cell

2845

22:757-765. 8. Foster, D., J. B. Levy, G. Q. Daley, M. C. Simon, and H. Hanafusa. 1986. Isolation of chicken cellular DNA sequences with homology to the region of viral oncogenes that encodes the tyrosine kinase domain. Mol. Cell. Biol. 6:325-331. 9. Fults, D. W., A. C. Towle, J. M. Lauder, and P. T. Maness. 1985. pp60O-src in the developing cerebellum. Mol. Cell. Biol. 5:27-32. 10. Ghysdael, J., J. C. Neil, and P. K. Vogt. 1981. A third class of avian sarcoma viruses, defined by related transformationspecific protein of Yamaguchi 73 and Esh sarcoma viruses. Proc. Natl. Acad. Sci. USA 78:2611-2615. 11. Ghysdael, J., J. C. Neil, A. M. Wallbank, and P. K. Vogt. 1981. Esh avian sarcoma virus codes for a gag-linked transformation specific protein with an associated protein kinase activity. Virology 111:386-400. 12. Gilmer, T. M., J. T. Parsons, and R. L. Erikson. 1982. Construction of plasmids for expression of Rous sarcoma virus transforming protein, p6(src, in Escherichia coli. Proc. Natl. Acad. Sci. USA 79:2152-2156. 13. Gray, M. R., H. V. Colot, L. Guarente, and M. Rosbash. 1982. Open reading frame cloning: identification, cloning, and expression of open reading frame DNA. Proc. Natl. Acad. Sci. USA 79:6598-6602. 14. Hager, D. H., and R. R. Burgess. 1980. Elution of proteins from sodium dodecyl sulfate-polyacrylamide gels, removal of sodium dodecyl sulfate, and renaturation of enzymatic activity: results with sigma subunit of Escherichia coli RNA polymerase, wheat germ DNA topoisomerase, and other enzymes. Anal. Biochem. 109:76-86. 15. Hanafusa, H. 1969. Rapid transformation of cells by Rous sarcoma virus. Proc. Natl. Acad. Sci. USA 63:318-325. 16. Itohara, S., K. Hirata, M. Inoue, M. Hatsuoka, and A. Sato. 1978. Isolation of a sarcoma virus from a spontaneous chicken tumor. Jpn. J. Cancer Res. (Gann) 69:825-830. 17. Karess, R. E., and H. Hanafusa. 1981. Viral and cellular src genes contribute to the structure of recovered avian sarcoma virus transforming protein. Cell 24:155-164. I8. Kawai, S., M. Yoshida, K. Segawa, H. Sugiyama, R. Ishizaki, and K. Toyoshima. 1980. Characterization of Y73, an avian sarcoma virus: a unique transforming gene and its product, a phosphoprotein with protein kinase activity. Proc. Natl. Acad. Sci. USA 77:6199-6203. 19. Kitamura, N., A. Kitamura, K. Toyoshima, Y. Hirayama, and M. Yoshida. 1982. Avian sarcoma virus Y73 genome sequence and structural similarity of its transforming gene product to that of Rous sarcoma virus. Nature (London) 297:205-208. 20. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685. 21. Le Beau, M. M., C. A. Westbrook, M. 0. Diaz, and J. D. Rowley. 1984. Evidence for two distinct c-src loci on human chromosome 1 and 20. Nature (London) 312:70-71. 22. Lipsich, L. A., A. J. Lewis, and J. S. Brugge. 1983. Isolation of monoclonal antibodies that recognize the transforming proteins of avian sarcoma viruses. J. Virol. 48:352-360. 23. Lizardi, P. M., and A. Engelberg. 1979. Rapid isolation of RNA using proteinase K and sodium perchlorate. Anal. Biochem. 98:116-122. 24. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 25. Naharro, G., K. C. Robbins, and E. P. Reddy. 1984. Gene product of v-fgr onc: hybrid protein containing portion of actin and a tyrosine-specific protein kinase. Science. 223:63-66. 26. Naharro, G., S. R. Tronick, S. Rasheed, M, B. Gardner, S. A. Aaronson, and K. C. Robbins. 1983. Molecular cloning of integrated Gardner-Rasheed feline sarcoma virus: genetic structure of its cell-derived sequence differs from that of other tyrosine kinase-coding onc genes. J. Virol. 47:611-619. 27. Neckameyer, W. S., and L.-H. Wang. 1985. Nucleotide sequence of avian sarcoma virus UR2 and comparison of its transforming gene with other members of the tyrosine protein

2846

SUDOL AND HANAFUSA

kinase oncogene family. J. Virol. 53:879-884. 28. Nishizawa, M., K. Semba, T. Yamamoto, and K. Toyoshima. 1985. Human c-fgr gene does not contain coding sequences for actin-like protein. Jpn. J. Cancer Res. (Gann) 76:155-159. 29. Nishizawa, M., K. Semba, M. C. Yoshida, T. Yamamoto, M. Sasaki, and K. Toyoshima. 1986. Structure, expression, and chromosomal location of the human c-fgr gene. Mol. Cell. Biol. 6:511-517. 30. Schleuning, W.-D., M. Sudol, and E. Reich. 1983. A proenzyme from chicken plasma similar to human plasma prekallikrein. J. Biol. Chem. 258:14106-14115. 31. Schwartz, D. E., R. Tizard, and W. Gilbert. 1983. Nucleotide sequence of Rous sarcoma virus. Cell 32:853-869. 32. Semba, K., Y. Yamanashi, M. Nishizawa, J. Sukegawa, M. Yoshida, M. Sasaki, T. Yamamoto, and K. Toyoshima. 1985. Location of the c-yes gene on the human chromosome and its expression in various tissues. Science 227:1038-1040. 33. Shibuya, M., H. Hanafusa, and P. C. Balduzzi. 1982. Cellular sequences related to three new onc-genes of avian sarcoma virus (fps, yes, and ros) and their expression in normal and transformed cells. J. Virol. 42:143-152. 34. Silhavy, J. T., M. L. Berman, and L. W. Enquist. 1984. Experiments with gene fusions. Cold Spring Harbor Laboratory. Cold Spring Harbor, N.Y. 35. Sorge, L. K., B. T. Levy, T. M. Gilmer and P. F. Maness. 1984. pp6OC-src expression in embryonic nervous tissue: immunocytochemical localization in the developing chick retina, p. 117-122. In G. F. Vande Woude, A. J. Levine, W. C. Topp, and J. D. Watson (ed.), Cancer cells, vol. 1. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

MOL. CELL. BIOL. 36. Sorge, L. K., B. T. Levy, and P. F. Maness. 1984. pp6Oc-src is developmentally regulated in neural retina. Cell 36:249-257. 37. Southern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517. 38. Sudol, M., T. L. Lerner, and H. Hanafusa. 1986. Polymerasedefective mutant of the Bryan high-titer strain of Rous sarcoma virus. Nucleic Acids Res. 14:2391-2405. 39. Sudol, M., and E. Reich. 1984. Purification and characterization of plasminogen activator secreted by a pig kidney cell line. Biochem. J. 219:971-978. 40. Takeya, T., R. A. Feldman, and H. Hanafusa. 1982. DNA sequence of the viral and cellular src gene in chickens. I. Complete nucleotide sequence of an EcoRI fragment of recovered avian sarcoma virus which codes for gp37 and pp6jsrc. J. Virol. 44:1-11. 41. Thomas, P. S. 1980. Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Proc. Natl. Acad. Sci. USA 77:5201-5205. 42. Wallbank, A. M., F. G. Sperling, K. Hubben, and E. L. Stubbs. 1966. Isolation of a tumor virus from a chicken submitted to a poultry diagnostic laboratory: Esh sarcoma virus. Nature (London) 209:1265. 43. Yoshida, M., S. Kawai, and K. Toyoshima. 1980. Uninfected avian cells contain structurally unrelated progenitors of viral sarcoma genes. Nature (London) 287:653-654. 44. Yoshida, M. C., M. Sasaki, K. Mise, K. Semba, M. Nishizawa, T. Yamamoto, and K. Toyoshima. 1985. Regional mapping of the human proto-oncogene c-yes-1 to chromosome 18 at band q21.3. Jpn. J. Cancer Res. (Gann) 76:559-562.