[CANCER RESEARCH 52, 6431-6439, December 1, 1992]

Monoclonal Antibodies to the Myogenic Regulatory Protein MyoDl: Epitope Mapping and Diagnostic Utility 1 Peter Dias, David M. Parham, David N. Shapiro, Stephen J. Tapscott, and Peter J. Houghton 2 Departments of Biochemical and Clinical Pharmacology [P. D., P. J. H.], Pathology and Laboratory Medicine [D. M. P.], and Hematology-Oncology [D. N. S.], St. Jude Children's Research Hospital, Memphis, Tennessee 38101; Department of Genetics, Fred Hutchinson Cancer Research Center, Seattle, Washington 98104 [S. J. T.]; and Department of Pathology, University of Tennessee, Memphis, Tennessee 38103 [D. M. P.]

ABSTRACT

Monoclonal antibodies (MoAbs) were developed against recombinant wild-type murine MyoD1 protein. Each of 4 MoAbs was immunologi- cally reactive with recombinant MyoD1 protein by enzyme-linked im- munosorbent assay, and each specifically stained the nuclei of myogenic cells. Epitopes were mapped using fusion protein constructs with spe- cific deletions of defined regions of the MyoD1 molecule. MoAb 5.2F recognized an epitope in the amino terminal region between amino acid residues (AAR) 3 and 56, whereas epitopes for MoAbs 1.1A, 5.4G, and 5.8A were in the carboxyl terminus (AAR 167-318) of the MyoD1 protein. The epitope for MoAb 5.8A was further delineated to AAR 170-209 by Western analysis and immunoprecipitation of in vitro tran- scribed and translated MyoDl protein having specific deletions in the carboxyl terminus. The 5.8A epitope was ultimately localized to the region between AAR 180 and 189 of the protein by enzyme-linked immunosorbent assay using 10-amino acid residue synthetic peptides. This sequence is apparently unique to MyoDl and has little homology to other myogenic regulatory proteins (myogenin, Myf5, Myf6, and MRF4). Transfection of cDNA for murine MyoD1 into a nonmuscle cell line conferred 5.8A reactivity, confirming the specificity of this reagent.

MoAb 5.8A was then used to examine the expression of MyoD1 in normal and malignant human tissues. MyoD1 was not detected in any normal adult tissue but was detected in 25 of 25 histologically confirmed rhabdomyosarcomas. Staining was localized to the nucleus and showed marked heterogeneity between cells as well as differential staining within nuclei. Specific subcellular localization of 5.8A was further de- termined by immunoelectron microscopy, where antibody was found to localize to electron-dense areas, more frequently associated with the nuclear submembranous region. In addition to rhabdomyosarcomas, MoAb 5.8A stained 2 of 5 Wilms' tumors and one ectomesenchymoma, neoplasms known to contain myogenic elements. The 5.8A reagent was also of value in the accurate histopathological classification of 2 of 4 tumors previously diagnosed as extraosseous Ewing's sarcoma and 2 of 3 tumors diagnosed as undifferentiated sarcomas.

INTRODUCTION

MyoD1 encodes a Mr 45,000 nuclear phosphoprotein, the expression of which is restricted to skeletal muscle (1, 2). It is a member of a family of proteins which share a conserved sequence motif capable of forming two amphipathic a helices separated by a nonhelical sequence, termed helix-loop-helix proteins (3). In addition to the helix-loop-helix region involved in homo- or heterodimer formation, MyoD 1 is characterized by a basic domain that confers DNA binding (4, 5) and a sequence homology with c -myc (1). The basic helix-loop-helix motif in MyoD1 is shared by three other related myogenic proteins, myogenin (6, 7), myf5 (8), and MRF4 (9). Extensive studies have demonstrated formation of MyoD1 homodimers and

Received 6/8/92; accepted 9/18/92. The costs of publication of this article were defrayed in part by the payment of

page charges. This article must therefore be hereby marked advertisement in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported by USPHS awards CA23099, 5U10CA24507, and CA21765 (CORE) from the National Cancer Institute and by the American Lebanese Syrian Associated Charities.

2 To whom requests for reprints should be addressed, at St. Jude Children's Research Hospital, 332 N. Lauderdale, Memphis, TN 38101.

heterodimerization with the protein encoded by E2A, which may be the usual cellular partner of MyoD1. The dimers bind DNA in a highly sequence-specific manner as a transcriptional activator complex (10, l l). Transfection of MyoD1 cDNA 3 into a wide variety of normal and neoplastic nonmuscle cells either converts them into myogenic cells (12) or transcription- ally activates muscle-specific genes (13). Although the regula- tion of MyoD1 is not fully understood, this and other myogenic regulatory proteins appear to perform critical functions in the commitment, differentiation, and maintenance of the myogenic lineage.

We have been particularly interested in the expression of myogenic regulatory proteins in malignant tissues. The histo- logical distinction of rhabdomyosarcomas from other small round cell or spindle cell neoplasms is often difficult due to a lack of morphological features characteristic of myogenic dif- ferentiation (14). Recently, it was shown that expression of MyoD1 transcripts was restricted to tumors of the myogenic lineage and was therefore diagnostic for rhabdomyosarcomas (15). In a subsequent study, we extended the ability to diagnose rhabdomyosarcomas by immunohistochemistry using a rabbit polyclonal antibody to trypE-MyoD 1 fusion protein on surgical specimens of childhood tumors (2). We demonstrated that all rhabdomyosarcomas stained for MyoD1. Of particular interest was the finding that some undifferentiated tumors previously diagnosed as sarcoma-type indeterminate and extraosseous Ewing's sarcomas were positive for MyoD1, suggesting that they were also tumors of myogenic lineage. Although staining for MyoD1 was more sensitive than that for the intermediate filament protein desmin, the interpretation of the staining was somewhat difficult due to the high background often associated with the use of polyclonal antibodies.

In the present study, we describe the development, charac- terization, and clinical application of a panel of mouse mono- clonal antibodies reacting with different epitopes of the murine MyoD1 protein. By using one of the monoclonal antibody re- agents to MyoD1 (5.8A), we confirmed our previous results demonstrating the diagnostic utility of MyoD1 protein for dif- ferentiating rhabdomyosarcomas from other soft tissue malig- nancies. Our results also indicate the potential utility of these antibodies in the characterization of the role and distribution of the MyoD1 protein in in vitro cell systems and in normal and pathological muscle tissue.

MATERIALS AND M E T H O D S

Northern Blot Analysis. For Northern analysis, total cellular RNA was isolated from 13 rhabdomyosarcoma cell lines (noted in Fig. 1 as cell lines A-M) by the guanidine hydrochloride method (16). RNA was

3 The abbreviations used are: cDNA, complementary DNA; MoAb, monoclonal antibody; ELISA, enzyme-linked immunosorbent assay; AAR, amino acid residues; RIPA buffer, radioimmunoprecipitation buffer (10 mM Tris, pH 7.4; 150 mM NaCl; 1% Nonidet P-40; 1% sodium deoxycholate; 0.1% sodium dodecyl sulfate); PBS, phosphate-buffered saline.

6431

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from

MONOCLONAL ANTIBODIES TO THE MYOGENIC REGULATORY PROTEIN MyoD1

separated through a 1% agarose-formaldehyde gel by electrophoresis and then transferred to Hybond-N nylon filters (Amersham, Arlington Heights, IL). The blots were hybridized with the indicated 32p-labeled cDNA probes at 42~ washed, and autoradiographed. The following probes were used for these studies: a 1.9-kilobase EcoRI fragment from murine MyoD1 cDNA clone pV2CI IB (12), a 1.5-kilobase EcoRI frag- ment from the murine myogenin cDNA clone pUC 65-2 (6), and a 1.26-kilobase EcoRI fragment from the human myf-5 cDNA clone myf-5/18-7 (8).

Immunization. An 8-week-old female BALB/c mouse received an i.p. injection of 25 ~g recombinant wild-type MyoDl protein (the kind gift of Dr. H. Weintraub, Fred Hutchinson Cancer Research Center, Seattle, WA) in 100 ul PBS emulsified with 100 ul Freund's complete adjuvant. At days 14 and 28 after the initial immunization, the mouse was boosted i.p. with 25 #g MyoD1 in 100 ~1 PBS, emulsified with 100 ~1 Freund's incomplete adjuvant. A final boost with 25 #g MyoD1 in 100 ul PBS was injected via the tail vein 3 days before fusion. During the immunization period, mouse serum was collected periodically and tested by ELISA against recombinant MyoDl protein as well as lysate from uninduced bacteria as a negative control and by immunohis- tochemistry against myogenic (Rh28 and Rh30 rhabdomyosarcoma) and nonmyogenic (SKNSH neuroblastoma, KB-V-1 cervical carci- noma, and GC3 colon carcinoma) cells.

Cell Fusion, Hybridoma Screening, and Cloning. Splenocytes were fused to SP2/0-Agl4 (SP2/0) mouse nonsecretory plasmacytoma cells (17) using standard procedures (18). Supernatant was harvested from wells with hybridoma growth when approximately 25% confluent. Su- pernatant from hybridomas which were positive against recombinant MyoDl protein by ELISA was then tested by indirect immunofluores- cence against Rh28, Rh30, SKNSH, KB-V-I, and GC3 cells. Positive cultures were expanded and cloned by limiting dilution. The resultant hybridoma clones were tested again by ELISA and indirect immuno- fluorescence. Hybridomas of interest were used to produce highly con- centrated monoclonal antibodies as ascites fluid in pristane-primed BALB/c mice. Monoclonal antibodies were affinity purified from as- cites fluid using a MabTrap G affinity purification column (Pharmacia LKB Biotechnology, Inc., Piscataway, N J) according to the manufac- turers' instructions.

Enzyme-linked Immunosorbent Assay. ELISAs were developed to screen hybridomas and for epitope mapping. Conditions for the differ- ent ELISAs were identical except that the proteins (recombinant wild- type MyoDl protein or mutant MyoD1 fusion proteins) or the 10- amino acid peptides were varied. Proteins or peptides diluted in 50 mM Na2CO3/NaHCO3 buffer (pH 9.5) were coated at 50 ng/100 ul/well and incubated overnight at 4~ As a negative control, lysate from the same strain of Escherichia coli which did not contain the MyoD1 expression vector was used. After washing, the plates were incubated for 2 h with hybridoma culture supernatant or control antibodies which included preimmunization or postimmunization mouse serum. An alkaline phosphatase-conjugated sheep anti-mouse Ig (1:1000 dilution, whole molecule; Sigma) was used as an enzyme-labeled secondary antibody, andp-nitrophenyl phosphate (Sigma) was used as the enzyme substrate. The color reaction was monitored at 405 nm by an ELISA plate reader (Bio-tek Instruments, Inc., Winooski, VT).

Expression and Purification of Wild-Type and Mutant MyoDl Fu- sion Proteins for MoAb Epitope Mapping. Alul fragments of MyoD 1 cDNA isolated from wild-type (12) and mutant cDNAs were ligated into glutathione S-transferase expression vector (pGEX) and used to produce wild-type and mutant MyoD1 fusion proteins. These fusion proteins lacked the acidic amino terminus [amino acids deleted (d) 3-56], the cysteine-histidine rich region (d63-99), the highly basic re- gion (d102-135), the region of myc homology (d143-162), and the carboxyl-terminal half of the protein (d167-318). A MyoDl eDNA fragment coding only the basic and myc homology regions (containing amino acids 102-166) was also ligated into the pGEX expression vec- tor. The construction of the above expression vectors and the expres- sion and purification of the fusion proteins have been as previously described (19).

In Vitro Transcription and Translation of Carboxyl-Terminus Mu- tant MyoDl. In vitro transcription of the carboxyl-terminus mutant MyoDl RNA was carried out in pEMSV scribe as previously described (20). Briefly, mutant MyoD1 plasmids were used to synthesize RNA using T3 RNA polymerase in preparative reactions (Stratagene, La Jolla, CA). Transcribed RNA from mutant MyoD1 plasmids was trans- lated in vitro to produce proteins deleted in various regions of the carboxyl-terminus of the MyoD1 protein. The amino acids deleted in these proteins were d170-209, d218-269, and d270-318.

For in vitro translation, 1 /zg of RNA was used in a 50-#1 reaction volume using a nuclease-treated rabbit reticulocyte in vitro translation kit (Promega, Madison, WI) and translation grade [a5S]methionine (DuPont Company, Boston, MA). The reaction was carried out at 30~ for 2 h according to the manufacturers' instructions. The translation reactions were stored at -70~ until processed for immunoprecipita- tion and for Western blotting with anti-MyoD1 monoclonal antibodies.

Immunoprecipitation. ELISA data obtained using MyoD1 fusion proteins indicated that most monoclonal antibodies reacted with epitopes lying toward the carboxyl terminus of the MyoD 1 protein. One of these MoAbs, 5.8A, was extensively characterized using normal and tumor tissue and on various cell lines and was found to be a useful reagent for the diagnosis and classification of myogenic neoplasms. For this reason, the epitope recognized by this reagent was further mapped to a smaller region of the protein. In vitro translated proteins deleted in various parts of the carboxyl terminus (d170-209, d218-269, and d270-318) were used in immunoprecipitation reactions to further map the epitope of the 5.8A MoAb to a smaller region.

Ten ~1 of in vitro-translated, 35S-labeled protein were diluted 1:100 in RIPA buffer. The samples were precleared by the addition of 50 #1 of 10% w/v Pansorbin (Calbiochem Corporation, San Diego, CA) and incubated on ice for 30 min followed by centrifugation at 10,000 x g for 20 min. The cleared supernatant was then incubated with the appro- priate primary antibody reagent overnight at 4~ For each in vitro translated mutant protein, four samples were prepared. One sample was incubated with the test antibody 5.8A and the other three samples were incubated with the following control reagents in equal amounts: (a) antibody against leucocyte common antigen (21) (negative control); (b) a cocktail comprising various monoclonal antibodies to MyoD1 (1.1A, 4.11 D, 5.2F, 5.4G, and 5.8A) (positive control); and (c) a control with no antibody but with an equivalent volume of RIPA buffer. After over- night incubation 5 ug of affinity purified rabbit anti-mouse Ig (Organon Teknika Corp., West Chester, PA) were added to each sample and incubated at 4~ for 2.5 h. Samples were immunoprecipitated by the addition of 70/~1 packed volume protein A-Sepharose CL-4B (Sigma), incubated on a shaker at 4~ for 1 h and then centrifuged at 10,000 x g for 3 min. The supernatant was aspirated, and the Sepharose pellet was resuspended with RIPA buffer and recentrifuged. After 3 fur- ther washes with RIPA buffer, sodium dodecyl sulfate gel sample buf- fer was added to each immunopreeipitate and heated at 90~ for 5 min. The samples were then centrifuged, and the supernatants were loaded onto 10% discontinuous sodium dodecyl sulfate-polyacrylamide gels. After resolving the proteins, the gels were fixed in methanol:acetic acid, prepared for fluorography (Enhance DuPont, Boston, MA), and autoradiographed.

Western Blotting. Western blotting was performed on nuclear ly- sates of tissue culture cells, on tumor tissue and in vitro translated proteins according to the method described by Towbin (22). Mono- clonal antibody culture supernatants or affinity-purified 5.8A MoAb to MyoD1 (0.6 mg/ml) were used at 1:5 or 1:300 dilutions, respectively, and 1:600 dilution of alkaline phosphatase-conjugated sheep anti- mouse IgG (Sigma) was used as the labeled secondary reagent.

Synthesis of 10 Amino Acid Peptide Sequences. For epitope map- ping of one of the anti-MyoD1 MoAbs (5.8A) four peptides, each 10 amino acids in length and corresponding to the residues 170-179, 180-189, 190-199, and 200-209 of the MyoDl protein (12), were synthesized by the solid phase method (23) on an Applied Biosystems 430A peptide synthesizer (Applied Biosystems, Inc., Foster City, CA) using standard N-methylpyrrolidone-hydroxybenzotriazole 9-fluorenyl- methoxycarbonyl chemistry.

6432

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from

MONOCLONAL ANTIBODIES TO THE MYOGENIC REGULATORY PROTEIN MyoD1

The sequences of the synthetic peptides were verified by phenylthio- hydantoin analysis on a Applied Biosystems 470A protein sequencer (Applied Biosystems, Inc.). The amino acid sequences of the four pep- tides synthesized were:

Peptide 1: H-A17~ A179_OH

Peptide 2: H-p18~

Peptide 3: H-E19~

Peptide 4: H-S2~176

DNA Transfection. The HTI080 fibrosarcoma cells were trans- fected with a MyoDl expression vector pEMCIls (derived from pEMSV scribe a2) or control expression vector pEMSV scribe a2 by standard calcium phosphate precipitation (24). Both vectors were cotransfected with a plasmid conferring neomycin resistance, at a molar ratio of 10:1. The MyoD1 expression vector and control vector were the kind gift of Dr. H. Weintraub. The transfectants were selected by growth in 800 #g/ml of Geneticin (Gibco BRL, Gaithersburg, MD), and isolated colonies were ring cloned and grown for further characteriza- tion.

Immunohistochemical Staining Techniques. lmmunohistochemical staining using anti-MyoDl reagents, anti-desmin MoAh (clone D33; DAKO, Carpenteria, CA), and anti-leukocyte common antigen (clones PD 7/26/16 and 2BI 1; DAKO) was carried out on tissue culture cells or cryostat sections of biopsy or autopsy tissue.

Tissue culture cells used for characterization of antibody staining included rhabdomyosarcoma cells (Rh28 and Rh30), colon carcinoma cells (GC3), cervical carcinoma cells (KB-V-1), neuroblastoma cells (SKNSH), and malignant rhabdoid tumor cells. The cells were seeded at a density of 30,000 cells/chamber in 4 chamber-well slides (Lab-Tek Nunc, Inc., Naperville, IL) 2 days before staining.

A large proportion of the tumors used in this study were previously acquired as fresh tissue within 24 h of surgery from collaborating in- stitutions of the Intergroup Rhabdomyosarcoma Study (25). The re- maining tumor tissues and normal tissues were obtained from the tu- mor bank of St. Jude Children's Research Hospital and were freshly frozen during either biopsy or autopsy. All fresh tissue specimens were rapidly frozen in isopentane which was precooled in liquid nitrogen. The specimens were then stored at -70~ until use. The frozen tissues were warmed from -70~ to -20~ before cutting 5-#m sections.

Tissue culture cells or cryosections were fixed and permeabilized for staining as previously described (2). The slides were then either incu- bated for 1 h at room temperature or overnight at 5~ with various dilutions of primary antibody. The anti-desmin monoclonal antibody was used at a dilution of 1:400, whereas the anti-leukocyte common antigen monoclonal antibody was used at a dilution of 1:100 in PBS. Each of the anti-MyoDl reagents (whole mouse serum, hybridoma culture supernatant, ascites fluid, or affinity-purified monoclonal anti- body) was titrated to give optimal staining. Concurrent staining for desmin was carried out for comparison and used as a positive control in myogenic tumors and cell lines. In desmin-negative tumors, staining of blood vessels served as an internal positive control. Staining for leuko- cyte common antigen was used as a negative control (with the exception of lymphomas); infiltrating leukocytes served as an internal control.

Immunoperoxidase staining was carried out by the avidin-biotin complex peroxidase method (26) using a Vectastain kit (Vector Labo- ratories, Burlington, CA). A 1:5 dilution of stock light green yellowish dye (0.2% w/v in distilled water with 0.2% v/v glacial acetic acid; Fisher- Scientific, Pittsburgh, PA) was used as a counterstain.

Indirect immunofluorescence staining was carried out as previously described (2), except that all test and control primary reagents were mouse monoclonal antibodies, and labeled secondary antibody reagent was fluorescein-conjugated F(ab')2 fragment rabbit anti-mouse Ig (DAKO) used at a dilution of 1:20 in PBS.

Immunoelectron Microscopy. For immunoelectron microscopy, the Rh30 cells were grown in culture flasks (Costar) until confluency was reached. The attached cells were then washed, fixed, and embedded, and ultrathin sections were stained for MyoD1 using a 1:200 dilution of

5.8A monoclonal antibody and 5 nm colloidal gold-conjugated goat anti-mouse Ig (Amersham, Arlington Heights, IL) according to a mod- ified method described by Bendayan (27).

R E S U L T S

Expression of Myogenic Regulatory Genes in Rhabdomyo- sarcomas. Northern analysis with probes for the three myo- genic factors MyoD 1, myogenin, and Myf5 showed a detectable level of MyoD1 transcripts in all rhabdomyosarcoma cell lines tested (Fig. 1, cell lines A-M) , whereas some of these cell lines did not show detectable levels of myogenin or Myf-5 tran- scripts. Four of 13 cell lines and 6 of 13 cell lines did not show detectable levels of myogenin and Myf-5, respectively. Interest- ingly, the level of MyoD 1 transcripts varied markedly between cell lines. Expression of myogenin and Myf-5 also showed sim- ilar variations with no apparent consistency in the level of tran- scripts. However, since MyoD1 was expressed in all rhabdo- myosarcoma lines and in tumor tissues (15, 28), it was considered to offer the greatest potential for the development of diagnostic reagents.

Development of MoAbs and Epitope Assignment. After im- munizat ion of the mouse with recombinant mouse MyoD 1 pro- tein, serum was tested for antibodies against MyoD 1 by ELISA and by indirect immunofluorescence on rhabdomyosarcoma cells. Both techniques indicated high antibody titers against MyoD1. After fusion, hybridoma growth was observed in 400 of 576 microti ter wells. Only 11 of the 400 hybridoma super- natants reacted with recombinant MyoD1 protein by ELISA. All 11 culture supernatants stained nuclei of rhabdomyosar- coma cells but did not stain the negative control cell lines in- cluding neuroblastoma (SKNSH), cervical carcinoma (KB-V- 1), and colon carcinoma (GC3) cells. Only 6 of 11 hybridoma cultures were stable and subsequently cloned by limiting dilu- tion. Monoclonal antibodies from these clones were further

18 S - -

MyoD

A B C D E F G H I J K L M

i ..... : , >,

Myogenin

2.0 Kb

18 s -

Myf -5

N,

28 S --

Fig. 1. Expression of MyoD 1, myogenin, and Myf-5 transcripts in 13 rhabdo- myosarcoma cell lines (A-M) detected by Northern blot analysis. The blot was probed with 32p-labeled cDNA probe for MyoD1 and subsequently for myogenin and Myf-5. All rhabdomyosarcoma cell lines expressed MyoD1 transcripts, whereas some cell lines did not express myogenin or Myf-5 transcripts.

6433

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from

MONOCLONAL ANTIBODIES TO THE MYOGENIC REGULATORY PROTEIN MyoDl

characterized against a large panel of myogenic and nonmyo- genie cell lines and were found to be specific for myogenic cell nuclei (results not shown). The supernatants were then tested against recombinant MyoDl by Western blotting. Immuno- staining corresponding to the Mr 45,000 MyoD1 protein was observed for each reagent (results not shown). Epitope mapping by ELISA, which utilized the various MyoD1 fusion proteins, indicated that MoAb 5.2F reacted with an epitope between AAR 3 and 56. The epitopes for three other anti-MyoDl Mo- Abs, 1.1A, 5.4G, and 5.8A, were found to be between AAR 167 and 318 of the MyoDl protein.

The epitope recognized by MoAb 5.8A was further mapped to a smaller region by immunoprecipitation and Western blot- ting using in vitro translated mutant MyoD proteins having deletions in the carboxyl terminal of the molecule (d170-209, d218-269, and d270-318). The epitope which monoclonal an- tibody 5.8A recognizes was found to be between AAR 170 and 209 of the MyoD1 protein. Fig. 2 shows that 5.8A did not immunoprecipitate the mutant protein lacking AAR 170-209 (Fig. 2, D170-209, Lane C), whereas the positive control which comprised a cocktail of anti-MyoD1 monoclonal antibodies immunoprecipitated the same mutant protein (Fig. 2, D170- 209, Lane .4). The negative control antibody, a reagent of the same isotype as 5.8A (IgGl), did not immunoprecipitate the mutant protein as expected (Fig. 2, D170-209, Lane B). In contrast, 5.8A immunoprecipitated mutant proteins lacking AAR 218-269 (Fig. 2, D218-269, Lane C) and AAR 270-318 (Fig. 2, leu 270, Lane C). These data were further supported by the Western blotting data shown in Fig. 3, where 5.8A did not recognize the mutant protein deleted in AAR 170-209, whereas it specifically recognized the wild-type MyoD 1 protein and mu- tants d218-269 and d270-318. These results suggested that the 5.8A epitope lay between AAR 170 and 209 and that a sequence in this region was necessary for antibody binding.

To further map the 5.8A epitope to a smaller region within the AAR 170-209 fragment, four peptides, each consisting of 10-amino acid residues, were synthesized. The amino acid se- quence of these peptides corresponded with the sequences of the MyoD1 fragment spanning the region AAR 170-209. By ELISA, 5.8A monoclonal antibody reacted with the peptide containing AAR 180-189 and not with peptides containing AAR 170-179, 190-199, or 200-209, indicating that the 5.8A epitope lies between AAR 180 and 189.

205 kD "-~

116 kD "-~

&leu 2 7 0 D218-269 D170-209 ,~, ;~,,~ ~

66 kD "-~

45 kD "-~

29 kD "-~

14.3 kD

A B C A B C A B C Fig. 2. Immunoprecipitation of in vitro transcribed and translated mutant

MyoD1 proteins having deletions in the carboxyl terminus of the MyoDl protein with monoclonal antibody 5.8A. 5.8A immunoprecipitated leu 270 and D218- 219 mutant proteins (Lanes C, Aleu 270 and D218-269, respectively) but did not immunoprecipitate the D 170-209 mutant protein. The positive control, compris- ing a cocktail of anti-MyoD1 monoclonal antibodies (1.1A, 4.11D, 5.2F, 5.4G, and 5.8A), immunoprecipitated all 3 mutant proteins (Lanes A). The negative control antibody of the same isotype as 5.8A did not immunoprecipitate all three mutant proteins (Lanes B).

180 kD

116 kD

84 kD-->

58 kD

48.5 kD

36.5 kD

26.5 kD

Fig. 3. Western blot analysis of in vitro transcribed and translated mutant MyoDl proteins having deletions in the carboxyl terminus of the MyoD1 protein with monocional antibody 5.8A. 5.8A formed a band with mutant proteins D218- 269 and leu 270 and wild-type recombinant MyoD1 protein but not mutant protein D170-209. The negative control (TC), which was an in vitro translation negative control (comprising all components of a translation mixture but without RNA), did not form a band with 5.8A.

Specificity and Diagnostic Value of MyoD1 MoAbs. Each of the MoAbs raised against recombinant MyoD1 distinctly stained myogenic cell nuclei, the 1.1A, 5.2F, and 5.8A being specific only for cell nuclei, whereas the 5.4G reagent also weakly stained cell cytoplasm. The weak reactivity of cell cyto- plasm by the IgM antibody 5.4G was considered to be nonspe- cific binding sometimes observed with antibodies of this iso- type. The 5.8A reagent, which was the most extensively characterized, was then used to examine the distribution of MyoDl in cell lines, normal tissues, and pediatric neoplasms. Fig. 4 shows 5.8A staining of cultured rhabdomyosarcoma cells and HT1080 cells transfected with a MyoD1 expression vector pEMC11 s. Strong nuclear staining of rhabdomyosarcoma cells and MyoDl-transfected HT1080 cells was seen, whereas the 1080 cells transfected with the vector control did not show 5.8A staining of nuclei. Of note was the speckled or punctate staining of Rh30 nuclei (Fig. 4D).

Results of 5.8A and desmin immunostaining of normal tissue are listed in Table 1. Antidesmin staining gave the anticipated results, with staining of smooth and skeletal muscle as well as ovarian stroma. Conversely, 5.8A did not stain any of the nor- mal tissues sampled, including fetal muscle (16 weeks of gesta- tion) and adult skeletal muscle.

Results of 5.8A and desmin immunostaining of pediatric neo- plasms is listed in Table 2. Desmin staining was again as ex- pected, staining neoplasms composed of myogenic and myofi- broblastic tissue, and generally paralleled our previous results (2). In addition, 25 of 25 rhabdomyosarcomas were 5.8A-pos- itive, including one desmin-negative case. On the other hand, 5.8A staining was negative in several desmin-positive, nonrhab- domyosarcomatous lesions, including 2 fibromas (one ovarian and one soft tissue), one peripheral neuroepithelioma, one neu- roblastoma, and one embryonal sarcoma of the liver. Three

6434

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from

A MONOCLONAL ANTIBODIES TO THE MYOGENIC RECd~IATORY PROTEIN MyoD1

B

wl;" " O "r4 , -*

Fig. 4. 5.8A staining of cultured cells. The HTI080 vector control cells (A) show a lack of nuclei staining, whereas the nuclei of HT1080 cells transfected with MyoD1 expression vector stain strongly (B). In C, 5.8A staining of cell nuclei of Rh30 rhabdomyosarcoma cells was observed at a low dilution and appeared punctate at a higher antibody dilution (D). E, immunoelectron micrograph showing the punctate nuclear distribution of the antigen. Colloidal gold grains are discretely localized at small patches which were more frequent but not restricted to the submembranous region. Protrusion into the cytoplasm, possibly through a nuclear pore, is evident in one focus (arrow). A-D, avidin-biotin complex staining with 5.8A and light green counterstain: .4, B, and D, x 800; C, • 400; E, Rh30 rhabdomyosarcoma cell labeled with 5.8A and colloidal gold-conjugated second antibody, • 48,000.

Wilms ' tumors with myomatous s t roma were desmin-positive, whereas only one was 5.8A-positive. One ectomesenchymoma, a rare soft tissue neoplasm with combined neural and myogenic elements, stained positively with both 5.8A and antidesmin. Positive staining for 5.8A and desmin was also found in 2 extraosseous Ewing's sarcomas, 2 sarcomas of indeterminate classification, and one sarcoma with poor tissue preservation. All other pediatric neoplasms tested, as listed in Table 2, were negative for both 5.8A and desmin staining. Examples of 5.8A staining and the utility of this reagent are shown in Fig. 5, where staining was highly specific (Fig. 5.4 versus Fig. 5B). In some tumors strong homogeneous nuclear staining of all cells was observed, whereas in other specimens heterogeneous stain- ing was observed between cells (Fig. 5, C and D). As with some of the cell lines, there was a distinctly punctate distribution of MyoD1 in the nuclei of some tumor cells (Fig. 5E). Most

striking, however, was the distinction between rhabdomyosar- coma cells invading skeletal muscle, where staining was de- tected only in the neoplastic populat ion (Fig. 5F).

Nuclear Distribution of MyoD1 Staining. Immunoelec t ron microscopy confirmed the punctate nature of MyoD 1 distribu- tion within the nucleus. Aggregates of colloidal gold grains were generally associated with electron-dense structures which were prominent in the periphery of the nucleus. Occasional colloidal gold grain aggregates appeared to protrude through the nuclear membrane into the cytoplasm, indicating a location in nuclear pores (Fig. 4E, arrow).

D I S C U S S I O N

We initially determined the distribution of transcripts for three myogenic regulatory genes, MyoD1, myogenin, and Myf5,

6435

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from

MONOCLONAL ANTIBODIES TO THE MYOGENIC REGULATORY PROTEIN MyoDl

Table 1 Immunostaining of normal tissue

Tissue type a 5.8A Desmin

Adrenal gland _ d _ Brain - - Breast - - Fetal skeletal muscle b _ + Heart - - Liver - - Lung - - Lymph node - - Ovary - + Pancreas - - Skin - - Skeletal muscle c _ + Small intestine - + Spleen - - Thymus - - Thyroid - -

a Desmin immunostaining was present in the vascular smooth muscle of all organs, whereas 5.8A staining was negative. Only one sample of each tissue type was tested, with the exception of skeletal muscle.

b Fetal muscle was from an autopsy of a fetus of approximately 16 weeks' gestational age.

c Only one of 6 adult skeletal muscle biopsy contained foci which were very weakly stained with 5.8A.

d-, absent; +, present.

in 13 rhabdomyosarcoma cell lines. Our results confirm previous studies (15, 28) that demonstrated consistent expression of MyoD1 in all lines but more variation in the expression of myo- genin and myf5. The biological and clinical significance of the variation in expression is currently unknown. It is notable, how- ever, that MyoD 1 is expressed in all cell lines, whereas Myf5, a gene expressed prior to MyoD1 in fetal muscle development (29), is expressed at low or undetectable levels in several of these cell lines. From this and other data (15, 28) it would appear that expression of MyoD 1 is at least one consistent molecular marker of rhabdomyosarcoma and a reasonable target for the develop- ment of monoclonal antibody reagents.

For screening hybridomas, we used methods which would select clones secreting MoAb for use in immunohis tochemistry . All hybridomas selected distinctly stained nuclei of only myo- genic cells, indicating that they were possibly reacting with MyoD1. The fact that the reagents stained HT-1080 fibrosar- coma cells only after transfection with a MyoD1 expression vector and not the vector control cells confirmed that these reagents were highly specific.

The results o f the epitope mapping ELISA using pGEX mu- tant MyoD1 fusion proteins indicated that hybridomas secret- ing antibodies which reacted with different epitopes of the MyoD1 protein were selected. The 5.2F reagent reacted with an epitope in the acidic amino terminus of the M y o D l protein, whereas reagents 1.1A, 5.4G, and 5.8A reacted with epitopes in the carboxyl terminal half of the M y o D l protein. The epitope of the 5.8A reagent was further mapped to a region between AAR 170 and 209 by Western blotting and immunoprecipi ta- t ion and then was confined to a region between AAR 180 and 189 of the MyoD1 protein by ELISA using synthetic peptides. This sequence is distal to the myc homology region (AAR 141- 162) (1,12). Therefore, it is unlikely that 5.8A cross-reacts with myc family proteins. The sequence is also distal to the 60-amino acid basic helix-loop-helix domain which spans AAR 102-162 (20) and is therefore unlikely to react with other proteins con- taining the helix-loop-helix motif.

Pairwise comparison of the MyoD1 peptide sequence which contains the 5.8A epitope (AAR 180-189) with protein se- quences in the Genbank showed that there was no homology between this region and other myogenic regulatory proteins (myf5, myf6, MRF4, and myogenin). Thus the staining of

HTI080-MyoD1 cells with 5.8A is not a consequence of cross- reactivity with other myogenic regulatory factors that may have been expressed as a result of transactivation by MyoD1. Other

proteins of potential relevance with similar but not identical amino acid sequences in this region are the human transcription factor E2-a (E2A), class 1 histocompatibility antigen, human t ranscription factor ITF-1, and DNA-binding protein E 12 (Fig. 6). Prel iminary observations suggest that the 5.8A reagent does not cross-react with the class 1 histocompatibil i ty antigen, since it stained only cell nuclei and not the cell surface. Also, if the 5.8A reagent cross-reacted with ITF-1, E l2 , or E2A in tissue, then staining of most cells in diverse tissue types would have been anticipated, since these factors are widely distributed and ubiquitously expressed. However, the staining data show that only myogenic cell nuclei reacted with 5.8A.

In our previous study using rabbit polyclonal antisera and indirect immunofluorescence staining technique, we reported cytoplasmic staining of some myogenic cells in addit ion to staining of the cell nuclei. In the present study using 5.8A reagent we observe strictly nuclear staining of myogenic cells. It is possible that what was previously interpreted as cytoplasmic staining may have actually been background staining associated with the use of polyclonal antisera. The immunofluorescence technique does not allow the study of cell and tissue morphol- ogy in conjunct ion with immunostaining. In the present study,

Table 2 Immunostaining of pediatric neoplasms

5.8A Desmin Total Diagnosis (no. positive) (no. positive) no.

Rhabdomyosarcoma Alveolar 8 7 8 Botryoid 1 1 1 Embryonal 16 15 16

Ectomesenchymoma l 1 1

Ewing's sarcoma Bone 0 0 4 Soft tissue a 2 2 3

Embryonal sarcoma of liver 0 1 1

Fibroma 0 2 2

Ganglioneuroma 0 0 1

Ganglioneuroblastoma 0 0 3

Melanoma 0 0 1

Malignant fibrous histiocytoma 0 0 1

Neuroblastoma 0 ! 6

Hodgkin's disease b 0 0 2

Non-Hodgkin's lymphoma c 0 0 3

Peripheral neuroepithelioma 0 1 5

Sarcoma, poor preservation 1 l 1

Sarcoma, type indeterminate 2 2 4

Wilms' tumor l 3 6 a One tumor, which has been previously reported (25), is also positive for a

variety of muscle markers by immunohistochemistry, using monoclonal and poly- cional antisera against desmin, polyclonal antisera against creatine, kinase M and myogiobin, and monoclonal antisera (HHF35) against muscle-specific actin. This tumor is interpreted as being a primitive rhabdomyosarcoma. The other tumor, originally diagnosed as rhabdomyosarcoma, was reinterpreted as being an ex- traosseous Ewing's sarcoma following histological reexamination and immunohis- tochemical negativity for other muscle markers (25).

Hodgkin's disease, nodular sclerosing pattern. c Large cell immunoblastic type (2 cases); small non-cleaved type (one case).

6436

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from

MONOCLONAL ANTIBODIES TO THE MYOGENIC REGULATORY PROTEIN MyoD1

lit

Fig. 5. 5.8A staining of surgical tumor specimens. Nuclear staining of representative rhabdomyosarcomas is seen (A and C-F). Lymphoma cells in B lack nuclear staining. Punctate staining of nuclei in an alveolar rhabdomyosarcoma is prominent at higher antibody dilution in E, whereas adjacent stromal cells are negative. Similarly, F shows nuclear staining of rhabdomyosarcoma cells infiltrating normal muscle with negative nuclei. Avidin-biotin complex staining with 5.8A and light green counterstain. A, B, and F, x 160; C and D, x 400; E, • 800.

however, we have slightly modified the immunoperoxidase technique to enhance visualization of MyoD1 staining, since our previous attempts with this technique were unsuccessful. Rather than use hematoxylin as a counterstain, we have used light green. The former dye stains cell nuclei and hence masks the immunostaining of nuclear antigens. The latter dye is more suitable for MyoD1 staining, since the staining intensity is lighter and it counterstains cell cytoplasm. This modified im- munoperoxidase staining also appears to be better for desmin staining.

As with rhabdomyosarcoma surgical specimens, MyoD1 staining of nuclei of rhabdomyosarcoma cells in vitro was het- erogeneous, being weakly expressed in some cells and strongly in other cells. In general, bi- and multinucleated cells appeared to stain more strongly than mononucleated cells.

The staining pattern observed by immunofluorescence (re- sults not shown) and by immunoperoxidase was punctate, and

the speckled staining was especially more prominent at higher antibody dilutions. The "punctate" staining pattern was also seen with immunoelectron microscopy whereby aggregates of colloidal gold grains tended to be discretely localized at elec- tron-dense structures which were more lYequent but not re- stricted to the submembranous region. The pattern of distribu- tion of the gold grains for MyoD1 staining is typical of that observed for other transcription factors such as c-jun (results not shown). Whether the 5.8A antibod,,: !s staining MyoD1 protein localized at transcription complext:s can only be spec- ulated on at this time.

It is interesting that although staining for MyoD1 appeared stronger in multinucleated myotubes in vitro, nuclei of normal adult skeletal muscle were almost always negative for MyoD1 (only one of 6 normal muscle biopsies very weakly stained for MyoD 1). In fact, in some rhabdomyosarcoma biopsies in which tumor cells infiltrated normal muscle, tumor cell nuclei stained

6437

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from

~yoDl sequence query

MONOCLONAL ANTIBODIES TO THE MYOGENIC REGULATORY PROTEIN MyoDl

1BO lBg

PGPLPPGRGS

Z~6 255 E2- klpha PLPLPPGSGP

MyoDl sequence query 180 189

PGPLP- PGRGS

13 23 ARPLPRPGRGS

Sequence similarity

"/07.

72Z

Major hlstocompatability antigen HLA ~ chain

addition, two neural tumors ~n'leh ~e~e ~esm'ln-Ds'lglvr vr162162 M~oDI-negative. This phenomenon iS ~I~SS~ ~ ~ S~])~N~r mide 04).

In conclusion, we have developed monoclonal antibodies to various epitopes of MyoD1 and by using one of these reagents we have demonstrated the potential application of such re- agents not only in the diagnostic histopathology of soft tissue tumors but also for studies investigating the cellular and mo- lecular basis of myogenesis. Such reagents may be useful in improving our understanding and classification of malignancies associated with the development of the soft tissue and disorders of the skeletal muscle.

180 189 MyoDl sequence query PGPLPPGRGS

33 42 El2 -Human PLPLPPGSGP

70Z

180 189 PGPLPPGRGS MyoDI sequence query

70Z

178 187 Human transcription PLPLPPGSGP factor ITF-I

Fig. 6. Proteins that contain sequence with some similarity to the AAR 180- 189 MyoDl sequence recognized by monoclonal antibody 5.8A.

strongly for MyoD1 whereas normal muscle nuclei were nega- tive (see Fig. 5F). Thus, MyoD positivity may be of consider- able utility in determining the presence and extent of neoplastic cell infiltrates into muscle and other tissues adjacent to tumor as well as assessing the regenerative capacity of skeletal muscle following injury. It is interesting that transcripts of MyoD1 have been detected in normal mouse muscle tissue by Northern blotting (1). One may speculate that this is a reflection of either heightened sensitivity of Northern analysis over immuno- histochemistry or posttranscriptional MyoD1 regulation. De- 5. finitive data are not currently available to support either of these hypotheses. It is, however, of interest that MyoDl is

6. expressed in mononuclear muscle precursor cells of mouse muscle following crush injury (30) and in injured regenerating 7. fibers (3 l).

The results of our immunohistochemical staining of tumor tissue with 5.8A extends our previous findings using indirect 8. immunofluorescence with polyclonal antisera (2). Both stud- ies demonstrate the heightened sensitivity and specificity of 9. MyoD1 antibodies in the immunostaining of rhabdomyosar- 10. coma as compared to antidesmin staining. In addition, MyoD1 staining furnished additional evidence of myogenic differentia- 11. tion in five undifferentiated tumors (2 originally diagnosed as extraosseous Ewing's sarcomas, 2 indeterminate sarcomas, and one poorly preserved sarcoma) which were also desmin-posi- 12. tive. These results support our previous contention that some 13. primitive neoplasms which could not be classified by routine morphological technique probably represent primitive rhab- domyosarcomas. The use of 5.8A permitted us to perform im- 14. munoperoxidase staining, which thus allowed visualization of

15. tissue by routine light microscopy, although unfortunately use of 5.8A requires frozen material rather than paraffin sections.

MyoD 1 appears to be a very specific marker of skeletal mus- 16.

cle differentiation, and the anti-MyoD1 antibody 5.8A does not stain myofibroblasts or smooth muscle as does antidesmin and 17. antiactin antibodies (32,33). Thus, no staining was obtained on samples of embryonal sarcoma of the liver and fibromas. In 18.

6438

ACKNOWLEDGMENTS

We thank Hallie Holt and Meredith White for immunohistochem- istry, Donna Davis for immunoelectron microscopy, Kent Williams for tissue processing, Christopher Payne and Alice J. Bell for technical assistance with the peptide synthesis, John Zacker and John Eakin for photomicroscopy, Dolores Anderson and Fabrienne Holloway for typ- ing the manuscript, and all of the collaborating investigators of the Intergroup Rhabdomyosarcoma Study for providing fresh tumor tissue (25).

REFERENCES

1. Tapscott, S. J., Davis, R. L., Thayer, M. J., Cheng, P-F., Weintraub, H., and Lassar, A. B. MyoDl: A nuclear phosphoprotein requiring a myc homology region to convert fibroblasts to myoblasts. Science (Washington DC), 242: 405-411, 1988.

2. Dins, P., Parham, D. M., Shapiro, D. N., Webber, B. L., and Houghton, P. J. Myogenic regulatory protein (MyoD1) expression in childhood solid tu- mors. Am. J. Pathol., 137: 1283-1291, 1990.

3. Murre, C., McCaw, P. S., and Baltimore, D. (1989) A new DNA binding and dimerization motif in immunoglobulin enhancer binding, daughterless, MyoD, and myc proteins. Cell, 56: 777-783.

4. Murre, C., McCaw, P. S., Vaessin, H., Caudy, M., Jan, L. Y., Jan, Y. N., Cabrera, C. V., Buskin, J. N, Hauska, S. D., Lassar, A. B., Weintraub, H., and Baltimore, D. Interactions between heterologous helix-loop-helix pro- teins generate complexes that bind specifically to a common DNA sequence. Cell, 58.' 537-544, 1989. Voronova, A., and Baltimore, D. Mutations that disrupt DNA binding and dimer formation in the E47 helix-loop-helix protein map to distinct domains. Proc. Natl. Acad. Sci. USA, 87: 4722-4726, 1990. Wright, W. E., Sassoon, D. A., and Lin, V. K. Myogenin, a factor regulating myogenesis, has a domain homologous to MyoD. Cell, 56: 607-617, 1989. Edmondson, D. G., and Olson, E. N. A gene with homology to the myc similarity region of MyoDl is expressed during myogenesis and is sufficient to activate the muscle differentiation program. Genes Dev., 3: 628-640, 1989. Braun, T., Buschhausen-Denker, G., Bober, E., Tannich, E., and Arnold, H. H. A novel human muscle factor related to but distinct from MyoDl induces myogenic conversion of 10TI/2 fibroblasts. EMBO J., 8." 701-709, 1989. Rhodes, S. J., and Konieczny, S. F. Identification of MRF4: a new member of the muscle regulatory factor gene family. Genes Dev., 3:2050-2061, 1989. Tapscott, S. J., and Weintrauh, H. MyoD and the regulation of myogenesis by helix-loop-helix proteins. J. Clin. Invest., 87:1133-1138, 1991. Weintraub, H., Davis, R., Tapscott, S., Thayer, M., Krause, M., Benezra, R., Biackwell, T. K., Turner, D., Rupp, R., Hollenberg, S., Zhuang, Y., and Lassar, A. The MyoD gene family: nodal point during specification of the muscle cell lineage. Science (Washington DC), 251: 761-766, 1991. Davis, R. L., Weintraub, H., and Lassar, A. B. Expression of a single trans- fected eDNA converts fibroblasts to myoblasts. Cell, 51: 987-1000, 1987. Weintraub, H., Tapscott, S. J., Davis, R. L., Thayer, M. J., Adam, M. A., Lassar, A. B., and Miller, D. Activation of muscle-specific genes in pigment, nerve, fat, liver, and fibroblast cell lines by forced expression of MyoD. Proc. Natl. Acad. Sci. USA, 86: 5434-5438, 1989. Enzinger, F. M., and Weiss, S. W. Soft Tissue Tumors, Ed. 2, pp. 448-488. St. Louis: C. V. Mosby, 1988. Scrable, H., Witte, D., Shimada, H., Seemayer, T., Wang-Wu, S., Soukoup, S., Koufos, A., Houghton, P., Lampkin, B., and Cavanee, W. Molecular differential pathology of rhabdomyosarcoma. Genes Chromosomes Cancer, 1: 23-35, 1989. MacDonald, R. J., Swift, G. H., Przybyla, A. E., and Chirgwin, J. Isolation of RNA using guanidium salts. Methods Enzymol., 152: 219-227, 1987. Shulman, M., Wilde, C. D., and Kohler, G. A better cell line for making hybridomas secreting specific antibodies. Nature (Lond.), 276: 269-270, 1978. Oi, V. T., and Herzenberg, L. L. Immunoglobulin-producing hybrid cell lines.

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from

MONOCLONAL ANTIBODIES TO THE MYOGENIC REGULATORY PROTEIN MyoD1

In: B. B. Mishell and S. M. Shiigi (eds.), Selected Methods in Cellular Immunology, pp. 351-372. San Francisco: Freeman and Company, 1980.

19. Lassar, A. B., Buskin, J. N., Lockshon, D., Davis, R. L., Apone, S., Haus- chka, S. D., and Weintraub, H. MyoD is a sequence-specific DNA binding protein requiring a region of myc homology to bind to the muscle. Cell, 58: 823-831, 1989.

20. Davis, R. L., Cheng, P-F., Lassar, A. B., and Weintraub, H. The MyoD1 DNA binding domain contains a recognition code for muscle-specific gene activation. Cell, 60: 733-746, 1990.

21. Warnke, R. A., Gatter, K. C., Falini, B., Hildreth, P., Woolston, R. E., Pulford, K., Cordell, J. L., Cohen, B., deWolf-Peeters, C., and Mason, D. Y. Diagnosis of human lymphoma with monoclonal anti-leukocyte antibodies. N. Engl. J. Med., 309: 1275-1281, 1983.

22. Towbin, H., Staehelin, T., and Godron, J. Eiectrophoretic transfer of pro- teins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA, 76: 4350-4354, 1976.

23. Merrifield, R. B. Solid phase peptide synthesis. I. The synthesis ofa tetrapep- tide. J. Am. Chem. Soc., 85: 2149-2154, 1963.

24. Wigler, M., Pellicer, A., Silverstein, S., Axel, R., Urlaub, G., and Chasin, L. DNA-mediated transfer of the adenine phosphoribosyltransferase locus into mammalian cells. Proc. Natl. Acad. Sci. USA, 76: 1373-1376, 1979.

25. Parham, D. M., Webber, B. L., Holt, H., Williams, W. K., and Maurer, H. Immunohistochemistry of childhood rhabdomyosarcomas and related neo- plasms: results of an Intergroup Rhabdomyosarcoma Study project. Cancer (Phila.), 67: 3072-3080, 1990.

26. Hsu, S-M., Raine, L., and Fanger, H. A. A comparative study of the perox-

idase-antiperoxidase method and an avidin-biotin complex method for study- ing polypeptide hormones with radioimmunoassay antibodies. Am. J. Clin. Pathol., 75: 734-738, 1981.

27. Bendayan, M. Double immunochemical labelling applying the protein A-gold technique. J. Histochem. Cytochem., 30: 81-85, 1982.

28. Tonin, P. N., Scrable, H., Shimada, H., and Cavenee, W. K. Muscle-specific gene expression in rhabdomyosarcomas and stages of human fetal skeletal muscle development. Cancer Res., 51: 5100-5106, 1991.

29. Montarras, D., Chelly, J., Bober, E., Arnold, H., Ott, M-O., Gros, F., and Pinset, C. Developmental patterns in the expression of Myf5, MyoD, myo- genin, and MRF4 during myogenesis. New Biologist, 3: 592-600, 1991.

30. Grounds, M. D., Garrett, K. L., Lai, M. C., Woodring, E. W., and Beilharz, M. W. Identification of skeletal muscle precursor cells in vitro by use of MyoD1 and myogenin probes. Cell Tissue Res., 267: 99-104, 1992.

31. Fuchtbauer, E-M., and Westphal, H. MyoD and myogenin are coexpressed in regenerating skeletal muscle of the mouse. Dev. Dynamics, 193: 34-39, 1992.

32. Truong, L. D., Rangdaeng, S., Cagle, P., Ro, J. Y., Hawkins, H., and Font, R. L. The diagnostic utility of desmin: a study of 584 cases and review of the literature. Am. J. Ciin. Pathol., 93: 305-314, 1990.

33. Skalli, O., Schurch, W., Seemayer, T., Lagace, R., Montandon, D., Pittet, B., and Gabbiani, G. Myofibroblasts from diverse pathologic settings are heter- ogeneous in their content of actin isoforms and intermediate filament pro- teins. Lab. Invest., 60: 275-285, 1989.

34. Parham, D. M., Dias, P., Kelly, D. R., Rutledge, J. C., and Houghton, P. Desmin positivity in primitive neuroectodermal tumors of childhood. Am. J. Surg. Pathol., 16: 483-492, 1992.

6439

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from

1992;52:6431-6439. Cancer Res   Peter Dias, David M. Parham, David N. Shapiro, et al.   MyoD1: Epitope Mapping and Diagnostic UtilityMonoclonal Antibodies to the Myogenic Regulatory Protein

  Updated version

  http://cancerres.aacrjournals.org/content/52/23/6431

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/52/23/6431To request permission to re-use all or part of this article, use this link

Research. on May 25, 2020. © 1992 American Association for Cancercancerres.aacrjournals.org Downloaded from