}V1531_] BETURNING MATERIALS: P1ace in book drop to LJBRARJES remove this checkout from .‘nuuzs-nnL your record. FINES w111 be charged if book is returned after the date stamped be1ow. 1M- ———'._— A. ANALYSIS OF THE HUMAN C-SRC LOCUS B. EXAMINATION OF GENETIC HOMOLOGY BETWEEN MAREK'S DISEASE VIRUS AND HERPESVIRUS OF TURKEYS By Carol Patrice Gibbs A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1984 ABSTRACT A. ANALYSIS OF THE HUMAN C-SRC LOCUS B. EXAMINATION 0F GENETIC HOMOLOGY BETWEEN MAREK'S DISEASE VIRUS AND HERPESVIRUS 0F TURKEYS By Carol Patrice Gibbs Structural analyses of the human c-src locus are described in part A of the thesis. Both the endogenous chicken c-src and the retroviral v-src genes have been extensively characterized; the studies presented here were initiated to provide a detailed analysis of the human homolog of these genes. The human c-src locus was examined by detailed restriction mapping and hybridization to v-src probes. Nine exons were identified spanning a 20kb region. Strong homology to v-src was observed, although certain exons were found to be somewhat more divergent. Analysis of human c-src/chicken c-src heteroduplexes supported the hybridization results. In addition, six regions containing human Alu repeat family sequences were identified. The nucleotide sequence of exons 2 and 4-12 was obtained. Comparison with chicken c-src and retroviral v-src demonstrated a strong sequence conservation within exons 4-12, with the majority of base substitutions silent. The human c-src retains the chicken c-src 3' terminus, rather than the v-src divergent 3' terminus, and also Carol Patrice Gibbs retains the precise intron-exon junctions of chicken c-src. Exon 2 has undergone a rearrangement involving both insertions and deletions. These studies revealed a strong conservation of amino acid sequence in the carboxy three-fourths of the protein, the region containing enzymatic activity, with alterations in the amino-terminal portion that is involved in attachment to the plasma membrane. ’ The degree of genetic homology between Marek's disease virus (MDV) and herpesvirus of turkeys (HVT) is examined in part B of the thesis. Despite their strong antigenic relationship, previous studies suggested the two viruses share only limited (1-5%) homology. The goals of this study were two-fold: first, both plasmid and phage MDV clone banks were constructed to facilitate studies of the MDV genome; second, less stringent hybridization conditions combined with improved hybridization kinetics were used to reexamine the sequence homology between MDV and HVT. Cross-hybridizations using the total viral DNA indicated the two viruses are closely related: sequences sharing 70-80% homology were detected over 90-95% of the viral genomes. These results were confirmed using individual cloned MDV DNA fragments. ACKNOWLEDGEMENTS I particularly want to thank my two advisors, Dr. Hsing-Jien Kung and Dr. Leland F. Velicer for their guidance and encouragement. My other committee members, Dr. Michele Fluck and Dr. Harold Sadoff are also thanked for their support. I would like to acknowledge excellent collaboration with Dr. Donald J. Fujita, Dr. Akio Tanaka and Steven K. Anderson on the human c-src project and with Dr. Keyvan Nazerian on the MDV work. Finally, Doris Bauer and Hans Cheng are gratefully thanked for invaluable technical assistance. 11' List List List Part Part TABLE OF CONTENTS Page of Tables.......................................................v of Figures............ ....... ..................................vi of Abbreviations..............................................vii A. Analysis of the Human c-src Locus Chapter 1: Introduction and Literature Review (part A)..........1 Introduction........ ........... .... ....... ..................2 Literature Review...........................................4 Chapter II: Structural Analysis of the Human C'Src Locus ...... 0.00.0.0...OOOOOOOOOOOOOOOOOOO00.30 Materials and MethOdSOOOOOOOOOOOOOOOOOOOOOO0.0.00.00000000031 Resu‘tSOOOOOOOO 000000000000 O00.0.0.0...0.0.0.0000000000000037 DiSCUSSionOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.0.00.0...56 Chapter III: Nucleotide Sequence Analysis of the Human c-Src LOCUSOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.0.60 Materials and Methods......................................61 Results....................................................65 Discussion.................................................78 Summary (part A)...............................................83 References (part A)............................................85 8. Examination of Genetic Homology Between Marek's Disease Virus and Herpesvirus of Turkeys Chapter IV: HIntroduction and Literature Review (part B).......94 IntrOdUCtionoooooooooooooooooooooooo0000000000000000000000095 iii Page Literature Review..........................................97 Chapter V: Cross-Hybridization Between MDV and HVT...........108 Materials and Methods.....................................109 Results...................................................111 Discussion................................................13D Summary (part B)..............................................135 References (part B)...........................................137 General References (part B)...................................142 iv LIST OF TABLES Page Chapter 3 1. Homology between the human and chicken c-src and SR-A V'Src geneSIOOOIOOO0.00000000000000000.0.0.000000075 2. Comparison of intron-exon boundaries in the hman and ChiCken C-SY‘C 10Cioooooooooooooooooooo000.000.00.77 LIST OF FIGURES _Page Chapter 2 1. Clones spanning the human c-src locus......................39 2. Detailed restriction enzyme map of the human c-src locus...OD.0.0.......OOOOOOOOOOOOOO0.0.0.00.00000000000000042 3. Hybridizations using the v-src subclones...................45 4. Localization of human c-src exons..........................47 5. Differential hybridization under different stringencjeso.0.00.00.00.00.000000000000000000000000.0.0.0050 6. Heteroduplex between human c-src and chicken c-Src c1oneSOOIO....0.0000000000000000000.00......0.0.0.00053 7. Hybridization of v-src to human genomic DNA................55 Chapter 3 1. Nucelotide sequence of the human c-src coding regionOOOO0.0.00000000000000000000......0.0.0.0000...0.0.0.68 2. Nucleotide sequence of the rearranged exon 2...............72 Chapter 5 1. Hybridization of MDV viral DNA to DNA from infeCted ce11SOOOOOOOOOOOOOOOOOOOOOO0.0...0.00.00.00.00000114 2. Hybridization of HVT viral DNA to MDV viral DNA...........117 3. .15 situ colony hybridization analysis of MDV p1asmid clones...I.O...OO...OOOOOOOOOOOOOOOOOOOO0.0.......120 4. Southern hybridization analysis of MDV plasmid C10ne500000OOOOOOOOOOOOOO0.000000000000IOOOOOOOOO0.0.0.000124 5. Restriction enzyme maps of MDV phage clones...............127 6. Southern hybridization analysis of MDV phage clones.......129 vi APS BSA dNTP DNase ddNTP DTT EDTA kb uCT PEG RNase SDS TEMED TES Tris LIST OF ABBREVIATIONS mnnoniumpersulfate bovine serum albumin deoxyribonucleotide triphosphate deoxyribonuclease dideoxyribonucleotide triphosphate dithiothreitol (ethylenedinitriol)-tetraacetic acid, disodium kilobase micro Curie polyethylene glycol ribonuclease sodium dodecyl sulfate N,N,N',N'-tetramethylethylenediamine 2-([tris-(hydroxymethyl)methyl]amino)ethanesulfonic acid tris(hydroxymethyl)aminoethane vii CHAPTER I INTRODUCTION AND LITERATURE REVIEW (part A) INTRODUCTION One of the most extensively studied retroviral oncogenes is the transforming gene of Rous sarcoma virus (RSV), v-src. The v-src gene OV-SrC. In encodes a tyrosine phosphorylating protein kinase, pp6 normal cells, tyrosine phosphorylation is a relatively minor modification, accounting for only 0.01% of the phosphorylated amino acids. However, infection with RSV results in an increase in total cellular phosphotyrosine, as a number of specific cellular proteins, as well as pp60V'5rc itself, become phosphorylated. Although the mechanism by which v-src transforms cells has been the focus of numerous studies, the actual alterations induced by pp60V'5"'c which lead to transformation have yet to be elucidated. Normal cellular sequences related to v-src (termed c-src) have been found in a wide variety of distantly related species. DNA hybridizations in liquid demonstrated these c-src sequences share extensive nucleotide homology with the viral src gene, and antisera raised against pp60V'S"c is also capable of immunOprecipitating the normal cellular protein, pp60c'src. Like its viral counterpart, ppGOC'STC possesses tyrosine kinase activity. The function of the normal cellular src gene is not yet known. Elevated levels of c-src expression have been found in adult brain and neural tissue and the level of c-src message is particularly high in these tissues during embryogenesis, suggesting a possible role of c-src 3 in the development of nervous tissue. The high degree of homology between the c-src sequences of widely divergent species indicates a strong selection against mutation, which implies pp60c-src has an essential function for which substantial alterations within the protein can not be tolerated. Elucidation of the normal function of c-src should provide insight into the v-src induced alterations leading to transformation. 0f the endogenous cellular c-src genes, only the chicken locus has been extensively studied. The amino acid sequence of the chicken pp60c‘src protein is nearly identical to that of pp60v’5rc, with a few amino acid changes scattered throughout the protein. However, 19 different residues in pp60c"src have been substituted for the carboxy-terminal 12 amino acids of pp60V'src. Although . the data are not yet conclusive, evidence suggests this region plays a major role in determining the oncogenic potential of the src gene. The main focus of the studies presented here is the detailed analysis of another c-src gene, the human c-src locus. The structural studies are described in Chapter 2 and the nucleotide sequence is presented in Chapter 3. The human c-src locus is compared with both the chicken c-src and the retroviral v-src genes. It is hoped further analyses of both c-src and v-src gene expression will lead to an understanding of the function of pp60c-src in normal cells as well as an understanding of the changes elicited by pp60V'5rc that lead to transformation. LITERATURE REVIEW The protein product of the v-src gene was initially identified as a non-structural transformation-specific antigen present in avian sarcoma virus (ASV)-transformed cells (11,13,90): it is a 60,000 dalton phosphoprotein, pp60V'Src. pp60""src can be synthesized in .yitgg by translation of the 3' third of ASV viral RNA (90,91), corresponding to the region identified by genetic analyses as the viral transforming gene (9,60,75,135). This is the only viral gene required for transformation; direct injection of v-src DNA into chickens will induce tumors (44). A. Enzymatic Activity of pp60v'src A unique feature of pp60V"src is the ability to phosphorylate tyrosine. This protein kinase activity can be easily assayed in 11359; extracts of transformed cells will phosphorylate the immunoglobulin heavy chain, while extracts from uninfected cells can not (10,21). Analysis of temperature-sensitive transformation mutants has confirmed the existence of this enzymatic activity (70). Protein kinase activity can be detected in immunoprecipitates of cells infected with the mutant tsNY68 when the cells are grown at the permissive temperature but not in immunoprecipitates from cells grown at the non-permissive oV‘Src is similar at both temperature, although the level of pp6 temperatures and the protein stability is unaffected by growth temperature. 5 In vitro assays further demonstrated the thermolability of the OV-src obtained from protein kinase activity of pp6 temperature-sensitive mutants (71). pp60V'src was prepared by immunoadsorption from cells infected with either wild-type Schmidt-Rupin 0 strain of ASV (SR-D ASV) or tsNY68 that had been grown at the permissive temperature, the extracts were incubated at 41.5°C (non-permissive temperature) for various intervals, then assayed for the ability to phosphorylate immunoglobulin. Kinase activity from the tsNY68-infected cells was inactivated 15-fold more rapidly than was the activity from SR-D ASV-infected cells. Several other mutants (tsLA90, BKS and tsGIZSl) have been found to possess extremely labile kinase activity (105). The relationship between the kinase activity of pp60"'src and the transformed phenotype has been investigated using a number of partial transformation mutants (3). Cells infected with two of the three mutants assayed (C02 and tsCUll) contain levels of pp60""s"c similar to that found in SR-A (Schmidt-Rupin A strain of ASV)-infected cells. However, the specific activity of the kinase activity of these mutant pp50v-src proteins is less than half that of wild-type ppsoV'STC. Cells infected with the third mutant, cu12 have a much lower level of pp60v'src , yet the protein exhibits slightly higher specific activity. The close correlation between enzyme activity and transformation observed with both the temperature-sensitive mutants and the partial transformation mutants strongly suggests the kinase activity plays an essential role in transformation. 6 In addition to possessing kinase activity, pp60V“src is itself phosphorylated at two residues, an amino-terminal serine, which can be phosphorylated by a cellular cyclic AMP-stimulated protein kinase (19), and a carboxy-terminal tyrosine residue that is phosphorylated in a cyclic AMP-independent reaction (19,22,43). pp60V”src phosphorylated in 1119 appears to contain twice as much, phosphoserine as phosphotyrosine (108). Additional minor sites of phosphorylation are also present (19,29,87,116). It has been suggested that pp60V'5rc is autOphosphorylated since incubation of highly purified pp60v’s"c with [32P-y]ATP results in phosphorylation of the protein in 11539 (22,89). The major phosphotyrosine of pp60v'src is residue 416 (87,116). Analysis of several other retroviral transforming proteins (p90 of Y73 and p105 of PRC 11, both strains of ASV; p80 of ESV, Esh sarcoma virus; p140 of FSV, Fujinami sarcoma virus; p85 of ST-FeSV, Snyder-Theilin strain of feline sarcoma virus; p120 of Ableson virus) (81,87) has revealed two common features: the phosphotyrosine is located 7 residues to the carboxy-terminal side of a basic amino acid and 4 or 5 residues to the carboxy-terminal side of a glutamic acid residue. Other proteins phosphorylated by pp60V'src (p36C911, p50cen and the immunoglobulin heavy chain) show similar structural features: a basic amino acid is located 7 residues to the amino-terminal side of the phosphotyrosine, with one or more acidic residues between. These features may therefore be important in substrate recognition of the pp60"'src protein kinase. The phosphotyrosine residue in pp50v-src was initially thought to be essential for transformation because a good correlation 7 was observed beween phOSphorylation at tyrosine 416 and transformation (19,70). When tsNY68-infected cells are grown at the non-permissive temperature, the level of phosphotyrosine in pp60v'5rc is greatly reduced while the level of phosphoserine remains unaffected. However, using site-specific mutagenesis at either the serine 17 (119) or the tyrosine 416 (29,119) it has been conclusively demonstrated that_ neither phosphorylation is required. Mutants lacking serine 17 are identical to wild-type SR-A in all characteristics assayed. The biochemical properties of mutants lacking tyrosine 416 are likewise similar to those of wild-type SR-A except the mutants are slower in inducing foci and tumor formation requires a longer latent period. These results lead to the conclusion that neither residue is required for transformation, but tyrosine 416 might facilitate transformation by affecting the kinetics. In support of these findings, pp60"'s"c synthesized in bacteria lacks phosphorylation yet retains kinase activity (44,77). B. Subcellular Localization of pp60V'src Immunofluorescence microscopy studies suggest pp60v-src is associated with the plasma membrane and is particularly concentrated near cell-cell junctions (64,95) and adhesion plaques (96) and may be associated with the cytoskeleton (14). Similarly, immunochemical electron microscopy techniques have been used to localize pp60V‘5r'c to the inner surface of the plasma membrane: the protein is concentrated at gap junctions (139). Analysis of in vitro phosphotransferase activity in cellular components obtained after cell fractionation (27,65) substantiates these findings: kinase activity is primarily found in the membrane fraction of SR-A infected cells. 8 Pulse-chase experiments demonstrated pp60V‘5rC is synthesized on soluble polyribosomes and rapidly becomes associated with the cell membrane (69). Immediately after synthesis, pp60"'src can be found in the soluble fraction, yet within 5 minutes virtually all of the protein has become membrane-associated. These analyses further demonstrated that pp60V'src is not a cleavage product of a larger precursor since both the soluble and membrane-bound forms have the same mobility in $05 gels. ppesoV"src can be removed from the membrane only by treatment with non-ionic detergents (27) suggesting the protein is inserted into the plasma membrane. In SR-D-infected rat (SR-RK and RR1022) and goat (Pcl) cells, pp60v'5rc appears to be associated with the nuclear, rather than cytoplasmic, membrane, and can also be found in the juxtanuclear reticular membranes (46,64,95). This altered distribution is stably inherited: these proteins have undergone alterations in the amino-terminus, which may be responsible for the different subcellular location (46). Limited proteolysis was performed to identify the membrane-bound portion of pp60V’5"c (69). Results indicate the amino-terminal region (approximately one-fourth of the molecule) is associated with the inner cell membrane, while the carboxy-terminal two-thirds extends into the cytoplasm. Kinase activity is retained in the cytoplasmic domain: the carboxy-terminal fragment obtained from chymotrypsin digestion (approximately half of the molecule) is capable of functioning in an in vitro phosphotransferase assay. At least two cellular proteins, pp89can and pp50°911, form a complex with cytoplasmic pp60v’5rc (12,26). The proteins are tightly bound to pp60v'5rc since the complex can be 9 immunoprecipitated with antisera specific for pp60V‘5'c. Sedimentation analysis suggests the complex contains a single molecule of ppeoV'Src, pp89can and ppsocen (12). The pp50v-src protein within the complex is phosphorylated only at serine and lacks tyrosine kinase activity (12,26). The two cellular proteins are not associated with the active membrane-bound form of pp60v‘src (12). The defect of one temperature-sensitive mutant, tsNY68, appears to OV'SrC with the plasma membrane involve the association of pp6 (12,26,83). At the non-permissive temperature 90-95% of the protein is found in the soluble fraction, associated with pp89can and pp50cen and is not phosphorylated at tyrosine (19). The minor fraction of membrane-bound pp60""5"c in these cells also lacks phosphotyrosine. After a shift down to the permissive temperature, cytoplasmic levels of pp60v'src decrease as the protein become associated with the membrane. In the converse experiment, shifting cells from the permissive to the non-permissive temperature, 70% of the ov-src becomes soluble and is found associated membrane-bound pp6 with the two cellular proteins pp89cen and ppSOCEII. These results indicate the association of pp60""5"c with the membrane is reversible, and the released protein is capable of re-forming complexes with pp89can and ppsoce“. The normal functions of pp89cen and pp50cen are unknown, however, pp89cen has been identified as a heat-shock protein (85): synthesis of pp89cen is greatly enchanced after treatment of normal chicken cells with a variety of chemical agents or by incubation at 45°C for 3 hours. Tryptic peptide maps of the IO heat-shock protein and the pp60V’src-associated protein are identical, strongly indicating they are the same protein. Nevertheless, antisera raised against the heat-shock protein is unable to precipitate the pp60v'src complex from ASV-infected cells, suggesting the protein conformations are different. C. Cellular Proteins Phosphorylated bypp60v‘src The ability of pp60V'5rc to phosphorylate tyrosine residues of normal cellular proteins may play a critical role in the transformation process. In normal cells, phosphotyrosine is present in minor amounts, constituting approximately 0.01-0.03% of the phosphorylated amino acid residues, with phosphoserine and phosphothreonine comprising the majority (>90% and 5-10%, respectively) (107). ASV infection results in a 6-10 fold increase in phosphotyrosine residues found in total cell protein (54,92,107). Phosphorylation of the pp60V'5rc protein itself accounts for approximately 20% of this increase (107), thus the increase in tyrosine phosphorylation of cellular proteins is around 5-6 fold. Tyrosine phosphorylation of cellular proteins has been found to correlate quite well with the transformed phenotype. Cells infected with the transformation-defective mutant tdSR-D do not possess elevated levels of phosphotyrosine (54). The correlation between transformation and the level of phosphotyrosine in total cell proteins has been further substantiated using the temperature-sensitive mutant tsLA29 (107). Cellular levels of phosphotyrosine are not elevated in cells infected with tsLA29 grown at the non-permissive temperature, but are greatly increased when the cells are grown at the permissive temperature. When the cells are initially grown at the non-permissive temperature, then shifted down to the permissive temperature, the level 11 of phosphotyrosine increases rapidly: within 60 minutes after the temperature shift, the phosphotyrosine level increases to 60% of that present in cells grown at the permissive temperature. The converse results are obtained when cells are initially grown at the permissive temperature and shifted up to the non-permissive temperature: the amount of phosphotyrosine decreases to the level found in uninfected cells within 60 minutes after the shift. These findings suggest phosphorylation of specific cellular proteins may be essential for transformation. To investigate the effect of pp60v'5rc on the synthesis and phosphorylation of individual cellular polypeptides, total cell proteins were analyzed on two-dimensional polyacrylamide gels (93). After labeling with 35s-methionine, only a low percentage (2.5%) of the normal cell proteins show increases in labeling intensity after Prague A (Pr-A) or SR-A infection. Similarly, only 1.7% show decreases in labeling intensity. Thus, viral infection appears to produce only minor change in the levels of the majority of cellular proteins. When the cells are labeled with 32P—orthophosphate, approximately 6% of the cellular proteins exhibit dectectable changes in relative labeling intensity. Phosphorylation of the pp60"'src protein is readily observed; in addition, phosphorylation of one cellular protein, p36CEI], is markedly increased. Since this protein appears to be a major substrate of pp60V'5rc in 1119 accounting for approximately 15% of the total phosphotyrosine (7), it has been studied in some detail. Kinetics of phosphorylation of p36cen were examined using the temperature-sensitive mutant tsLA29 (93). At the non-permissive 12 temperature phosphorylation of p36cen is barely detectable. Phosphorylation can be detected within 20 minutes after a shift down to the permissive temperature, and protein synthesis is not required for the phosphorylation to occur (94). The protein is phosphorylated at both serine and tyrosine (92), yet the charge difference between the phosphorylated and non-phosphorylated species suggests the phosphorylated p36cen consists of two distinct populations, one phosphorylated at a serine residue and the other at a tyrosine residue. p36cell isolated from normal cells is an efficient substrate for phosphorylation by pp60""src in 31359 (40), with tyrosine, but not serine, phosphorylated. Although the function of p36cen is unknown, the protein has been localized to the cytoskeletal framework in both normal and ASV-transformed cells (16.23.28), and both monomer and dimer forms are present (41). Analyses of the phosphorylation of p36cell in cells infected with temperature-sensitive mutants (RSV-B77 Rat 1 cells and tsLA24-infected rat cells) suggests a high correlation between phosphorylation of p36°911 and transformation (16). At the non-permissive temperature, p36cen is not phosphorylated; at the permissive temperature it is. Furthermore, an increase in p36cen phosphorylation can be detected within 1-2 hours after a shift from the non-permissive to the permissive temperature. Despite this good correlation observed with the temperature-sensitive mutants, analysis of partial revertant mutants indicates phosphorylation of p36cen is not sufficient for transformation. Phosphorylation of p36cell was examined in fully transformed (1T), partial revertant (866-R5C) and full revertant 13 (866-4) ASV-transformed vole cell lines (80); the partial revertant had regained normal cellular morphology yet was capable of growing in soft agar and inducing tumors in nude mice. p36cell in lines IT and 866-R5C is phosphorylated to a similar extent. Phosphorylation of p36cen in line 866-4 is greatly reduced. These results suggest phosphorylation of p36cen is not associated with morphological changes in transformed cells, yet might play a role in the ability to grow in soft agar and tumorigenicity. The correlation between phosphorylation of p36cell and a number of transformation-specific characteristics has been examined using a battery of 8 partial transformation mutants (24). Phosphorylation of p36cen correlates well only with the production of plasminogen activator, but not with changes in adhesiveness, hexose transport or colony formation in soft agar, nor is there a clear correlation between phosphorylation of paace“ and the ability of infected cells to form tumors in nude mice (57). In addition, the phosphorylation occurs only when total cellular phosphotyrosine levels are high, suggesting the protein may be a low-affinity substrate for pp60v'src. Thus, phosphorylation of p36cell could be non-essential for transformation. A number of other cellular proteins have been identified as substrates for pp60v'5rc in £119, although the increase in phosphotyrosine in these molecules is somewhat less than that observed in p36ce‘]. Vinculin, a protein associated with the cytoplasmic side of adhesion plaques, which anchor actin-containing microfilaments to the plasma membrane and attach cells to the substratum, was identified as an in vivo substrate of pp50v-src (104). The 14 levels of phosphotyrosine in vinculin are 8-fold higher in ASV-infected cells than in uninfected cells. Nevertheless, only 1% of the vinculin molecules in transformed cells actually contain phosphotyrosine, so the biological significance of this phosphorylation remains unclear. In addition, vinculin is not phosphorylated in PRCII-infected cells (104). Other cytoskeletal proteins (filamin, myosin heavy chains, a-actinin and vimentin) are not phosphorylated by pp60""src i vivo. ’ Two cytOplasmic proteins were identified as in vivo substrates of ppaoV'SrC, p45can and p28can (23,24). A series of partial transformation mutants were assayed for levels of phosphorylation of both p46can and ngcell (24). Phosphotyrosine levels of both molecules correlate well with increased 5cell also correlates with hexose transport. Phosphorylation of p4 the production of plasminogen activator. It was concluded that phosphorylation of either p46cen or p28cen is not sufficient for transformation but may be involved in expression of a particular transformation-specific phenotype. Alternatively, phosphorylation of either of these cellular proteins could be unrelated to transformation. These results would imply that phosphorylation of other cellular proteins, which appear as minor substrates by two-dimensional gel analysis, are the critical proteins necessary for transformation. A more quantitative, although less qualitative approach has been used to analyze phosphotyrosine content in total cellular proteins (7). Phosphoproteins from SR-A infected cells were compared to those isolated from uninfected cells by SOS-polyacrylamide gel electrophoresis. The gels were then sectioned into 10 slices and each slice subjected to phOSphoamino acid analysis. Phosphotyrosine is 15 found in all size classes in uninfected cells, ranging from 0.1 to 0.15% of the phosphoamino acids per fraction. In proteins from SR-A-infected cells, phosphotyrosine is likewise found in all fractions, but the levels are much higher than in uninfected cells, ranging from nearly 0.2% to 0.8%. Two broad peaks at approximately 35,000 daltons and 60,000 daltons are probably largely due to .p36ce‘] and pp60v'src. These data demonstrate that a relatively large number of cellular proteins are phosphorylated in infected cells, more than are discernable by two-dimensional gel electrophoresis. Cellular proteins critical for transformation remain to be identified: it is likely phosphorylation of more than one cellular protein is required for full expression of the transformed phenotype. Partial transformation mutants manifest different degrees of expression of transformation-specific characterisitics (3,24), and can not readily be arranged in a heirarchy of transforming potential. Furthermore, certain transformation-associated changes, for instance the loss of fibronectin, do not correlate with tunorigenicity (57). Analysis of temperature-sensitive mutants has also suggested that pp60V‘5rc must interact with a number of cellular proteins to express the fully transformed phenotype (138). When infected cells were assayed for the expression of several transformation parameters after growth at temperatures intermediate between the permissive and non-permissive temperatures, 4 of the 5 mutants analyzed could be roughly ranked with respect to the severity of the mutation. The fifth mutant examined, tsGIZSl, is highly abnormal; changes in biochemical parameters are similar to those induced by the other mutants, yet temperature has an opposite effect on expression of growth-related properties. In 16 contrast to infection with most temperature sensitive mutants, tsGIZSl-infected cells form colonies in soft agar more efficiently at higher, rather than lower temperatures. Biochemical parameters and growth characteristics can therefore be clearly distinguished. Thus, a number of studies suggest that the pleiotropic effects of pp60v-src may actually be brought about by the interaction of the protein with several normal cellular proteins, rather than a single substrate. Mutants have been described which elicit certain transformation-associated characteristics but not others; which of the many cellular changes are essential for tumorigenicity remains to be determined. It is clear that some of the cellular alterations associated with transformation are not required for tumor development; whether expression of these characteristics enchances the tumorigenic potential has not yet been elucidated. It has been proposed that the v-src gene actually encodes two proteins read in overlapping reading frames (74). Evidence for the existence of a second protein rests primarily on analysis of the ASV-transformed rat cell line 831, the revertant line 000 and the retransformed line 000*. A single base pair mutation in line 000 causes a frameshift, which is compensated for by a 242 base pair duplication in line 000*. Examination of aberrant proteins synthesized in these two cell lines suggests initiation can occur at an internal AUG. The nucleotide sequence implies initiation at this internal site should result in the synthesis of a 7,000 dalton protein in the parental 831 line. Nevertheless, this small protein has not yet been identified, nor has any function been assigned to it. If the 17 7,000 dalton protein does exist, it is possible that the protein is responsible for one or more of the many cellular changes associated with transformation. 0. Structural Features of the v-src and c-src Loci Complete nucleotide sequence of the v-src locus of both SR-A and Prague C (Pr-C) has been obtained (30,31,103). The v-src coding, sequence is nearly 1.6kb in length, encoding 526 amino acids; the sequence of Pr-C differs from that of SR-A at 42 nucleotides, resulting in 22 amino acid changes. Several interesting features have been observed in the regions flanking v-src (103). The v-src gene is flanked by direct repeats, slightly longer than 120 base pairs. Only one copy of the sequence is found in ALV, which lacks src sequences. A second sequence, termed E, is located upstream from v-src in SR-A, immediately after the direct repeat. In Pr-C, this sequence is located downstream fran v-src, before the second direct repeat. Short (11 nucleotide) direct repeats (boundary repeats) flank the E region, suggesting E may actually be an insertion sequence. E itself is capable of forming a large hairpin structure. Sequences homologous to the v-src coding region were detected in a variety of avian species using v-src specific cDNA as a probe under stringent conditions (125). Thermal stability of the v-src/avian DNA hybrids correlates strongly with phylogenetic distances between the species. Calculations indicate only 3-4% base mismatch between v-src and the endogenous chicken src (c-src) sequences. Endogenous src sequences from more distantly related avian species are only slightly less homologous. 18 Although under highly stringent conditions v-src related sequences cannot be detected in non-avian species (125), v-src will anneal to DNA from a variety of vertebrate species when the hybridization conditions are somewhat relaxed (124). Homology of these distantly related species to v-src is extremely high, with an average of only 8-10% base mismatch. Under relaxed conditions, v-src will also hybridize to Drosophila DNA (53,113). Hybridization kinetics suggest the c-src sequences in most species are present at a maximum of 5-10 copies per cell. The high degree of homology over wide phylogenetic distances suggests the c-src gene may have a critical, indispensable function, providing selective pressure for the conservation of nucleotide sequence. Genomic clones spanning the chicken c-src locus have been obtained (86,110,128, 129); v-src related sequences are distributed over a region approximately 8kb in length. Initially 6 introns could be observed in chicken c-src/v-src heteroduplexes (86,110,129), 5 smaller introns have been identified by direct sequence analysis (128). The clones appear to contain the major chicken c-src locus since restriction maps of the cloned regions precisely predict the sizes of fragments detected in Southern blots of chicken DNA when v-src is used as probe (110,129). In addition, a second locus bearing homology to the 5' portion of v-src may also be present (86). Direct comparison of the nucleotide sequence of chicken c-src and v-src (SR-A) reveals a number of interesting features (128). With the exception of the 3'-terminus, the sequences are highly homologous, with only 18 nucleotide changes producing 8 amino acid alterations. The major difference between the two proteins is found at the 3' terminus, 19 where 19 completely different amino acids in pp60c'src have been substituted for the the carboxy-terminal 12 amino acids of pp60V’5rc. A 39 base pair sequence corresponding to the 3' terminus of v-src is located approximately 900 base pairs downstream from the termination codon of chicken c-src. Sequence homology between chicken c-src and v-src extends approxi- mately 90 nucleotides upstream from the v-src initiation codon (128). The sequences diverge immediately upstream from a splice acceptor site present in the RSV genome. Neither the E sequence nor the direct repeats flanking v-src are found flanking c-src. Finally, all chicken c-src splice donor and acceptor sites retain the GT-AG consensus sequence. RNA hybridizing to v-src DNA has been found in uninfected chicken cells (123), with an estimated 1-4 copies per cell, suggesting the c-src protein is present at low levels in normal cells. The chicken c-src mRNA has a sedimentation coefficient of 28-305, thus is approxi- mately 6kb long; it is polyadenylated and associated with polyribosomes (122). Levels of c-src mRNA are identical in sparse and densely cul- tured cells, and are not affected by serum starvation (123), therefore c-src expression does not appear to correlate with cell proliferation. E. Analysis ofpp60c'src Serum obtained from tumor-bearing animals is able to precipitate a protein antigenically related to pp60V'src from uninfected avian cells (18,20,84,97,106) termed pp60c’5'c. Partial proteolysis of pp60c'5"c revealed the normal cellular protein is structurally very similar to pp60V‘5'c (ASV), but not identical. In particular, the carboxy-terminal fragment produced by V8 protease digestion is slightly larger in pp60c'5rc than in pp60v'5'C, 20 This fragment contains the divergent carboxy terminus, which the nucleotide sequence predicted should be 7 amino acids longer in pp60‘i'5"c (128). Several other differences between the two proteins have been identified by tryptic peptide analysis (106). Like the viral protein, pp60c'src is phosphorylated at both serine and tyrosine residues (20). However, the phosphopeptide maps are completely different (18), indicating the residues phosphorylated in OC-SY'C OV-SI‘C pp6 are not identical to those in pp6 The level of pp60c‘5rc in normal cells is estimated to be 30-100 fold lower than the levels of pp60V'srC in SR-A transformed cells (18,84). Although pp60‘:'5"‘c immunoprecipitated with certain sera does not exhibit protein kinase activity (18,97), phosphotransferase activity can be demonstrated when other antisera are used (20,84,97). Similar enzymatic activities suggest pp60c'src and pp60""5'“c are functionally similar. Sera from tumor-bearing animals can also precipitate a pp60v’5r°-related protein from a variety of avian cells (18), as well as mammalian (20.84.97.106) and frog cells (97), but not from Drosophila cells (97,84). Partial proteolytic digests (20.84.97) and tryptic peptide mapping (106) demonstrated the different src proteins are nearly identical to ppsoc-src. This high degree of amino acid conservation is in complete agreement with the strong nucleotide homology observed among the divergent species (124). As with chicken pp60c‘5rc, the src proteins from other species also exhibit protein kinase activity (20.97). Two forms of pp60c'src are present in human fibroblasts and human epidermal carcinoma cells, with molecular weights of 59,000 21 daltons and 60,000 daltons (111). The two forms are differentially phosphorylated: phosphorylation of the 59,000 dalton fonn is almost exclusively on the amino-terminal serine. while the 60,000 dalton form is phosphorylated predominantly on the carboxytenminal tyrosine, with minor serine phosphorylation. Two-dimensional chymotryptic analysis demonstrated the two human pp60‘:"S"c forms share several common ' peptides. However, significant differences are present. It has been suggested the two forms arise from two distinct human c-src loci or they are generated by alternate splicing of the same initial transcript. Although most sera are unable to precipitate a src-related protein from Drosophila or other lower organisms (84,97), one serum was obtained which could precipitate a tyrosine kinase activity from a wide variety of species, including mammals, birds, fish, insects and sponges (100). A 50,000 dalton phosphoprotein similar to ppooV'src could be detected by immunoprecipitation from all tissues that could be labeled vfith 32P-orthophosphate. The protein was capable of phosphorylating immunoglobulin jg_yjtrg, with phosphorylation specific for tyrosine. In all species examined, the highest level of phosphotransferase activity was detected in brain and nervous tissue. No kinase activity could be detected in hununoprecipitates of unicellular organisms or plants. Expression of c-src in normal brain and nervous tissue was confirmed by analysis of various tissues from chicken embryos (25). Lysates of brain and neural retina tissues have 8-10 fold higher levels of tyrosine kinase activity than do other tissues. pp60c'5rc concentration remains low until day 6 of embryogenesis, when it rapidly 22 reaching a plateau at day 9. Slightly before hatching, the level drops to that found in the adult tissue. In retinal neurons. pp60c'5rc is localized within cell bodies and processes of ganglion neurons, bodies of amacrine cells and processes of rods and cones (120). In these cells, levels of pp50c-src are particularly elevated during developmental stage 35, at the time when cell-cell contacts between neurons are established. The demonstration of a p60v'sr°-related tyrosine kinase activity in phylogenetically distant species and the strong conservation of amino acid sequence in organisms as different as birds and mammals implies pp60c"s"c serves an extremely important function in normal cells. Elevation of c-src expression in the brain and neural tissue during development suggests a role in the differentiation of nervous tissue. Nevertheless, the actual function of pp50c-src remains unknown. F. Other Members of the src Family Comparison of the nucleotide sequence of v-src with sequences of other retroviral oncogenes revealed v-src belongs to a large family. Related sequences include abl (from Ableson virus; 48.53), erbB (from avian erythroblastosis virus; 88, 140), fes (from feline sarcoma virus; 50), fps (from Fujinami sarcoma virus; 112), mos (from Moloney murine sarcoma virus; 134) and yes (from Y73 strain of ASV; 61). Homologous sequences are clustered in the 3' half of v-src, with homology particularly strong in the regions flanking the codon for Tyr 416. Most similar to src is yes (61), with 82% amino acid homology beginning around residue 80 of v-src. At the 3' terminus, the yes sequence shares homology with c-src, rather than v-src. In addition, a 23 Drosophila clone hybridizing to both v-src and v-abl has been obtained (53). Several members of the src family (abl, fes, fps and yes) exhibit tyrosine kinase activity. Also related to src is the catalytic chain of bovine cyclic AMP-dependent protein kinase (4). Homologous sequences are localized in the carboxy-terminal half of v-src, suggesting this protein is also a member of the src family. In addition, several peptides derived from the human epidermal growth factor (EGF) receptor are similar to regions of v-erbB (39). One function of the EGF receptor is the ability to phosphorylate tyrosine, thus the EGF receptor may be another member of the src family. G. v-src/c-src Hybrids The specific differences between v-src and c-src that determine tumorigenic potential are not clear. Transformation-defective mutants containing large deletions in the v-src gene have been observed to recombine with endogenous c-src sequences generating fully transforming revertants, rASVs (59,136,137) which contain a chimeric src gene. The ability to generate rASVs is directly proportional to the amount of v-src sequence remaining in the viral genome. Partial proteolytic digestion patterns confirm the pp60src proteins encoded by rASV isolates are hybrids of ppooc'src and ppooV‘SFC (59,126). Peptides common to either pp60V'src or pp60c'5rc have been observed by tryptic peptide mapping (58) and differences between RNase Tl-resistant oligonucleotides have been observed by fingerprint analysis (136,137). The extent of v-src or c-src derived sequence is different for each rASV isolate; the amino-terminal region is 24 particularly plastic, capable of tolerating large variations in size (58). Three rASV src genes analyzed in detail (157, 1441 and 1702) appear to have the same 3' terminus (less than 20% of the coding sequence) derived from v-src. the remainder from c-src (58). The pp60src isolated from rASV-infected cells retains tyrosine kinase activity (59) and is phosphorylated at both serine and tyrosine (58). The phosphotyrosine residue is identical to that found on pp60""5"c not ppGOC’src. Direct nucleotide sequencing of rASV1441 revealed the 5' recombination occurred upstream from the initiation codon and the second crossover site was slightly upstream from the pptsoV"src phosphotyrosine residue 416 (126,127). Despite the fact that src sequences in rASVs are primarily derived from endogenous c-src, the viruses resemble SR-A in transformation potential. Two possibilities have been suggested: either OV-SY‘C OC-SY‘C pp6 and pp6 are identical in function, but overexpression of pp50v-src is sufficient to cause transformation or the two proteins recognize different intracellular targets, with the src resulting in altered substrate specificity of pp60V' transformation. To elucidate the factors contributing to oncogenicity, chimeric v-src/c-src molecules have been constructed and assayed for the ability to transform fibroblasts in culture. Results from the hybrids in plasmid vectors (109) are somewhat different from those in vectors designed to allow for virus production (51). Constructions containing the chicken c-src gene in an expression vector were unable to transform NIH3T3 cells, while similar constructs containing v-src transformed with high efficiency (109). Src 25 expression and kinase activity in both cultures were identical, and cells transfected with either construct showed similar increases in total phosphotyrosine. The results suggest functional differences exist between pp60""s"'c and pp60c'5rc, overexpression of pp60c'5rc and the resultant increase in the level of phosphotyrosine is not sufficient for transformation. . Chimeric v-src/c-src hybrids were also constructed and assayed for the ability to transform NIH3T3 cells (109). A plasmid containing the 5' portion of v-src and the 3' terminus of c-src was able to induce only a small number of foci. The reverse construction. containing the 5' portion of c-src and the 3' terminus of v-src transformed efficiently. A final construct, containing v-src sequences with the carboxy-terminal 9 amino acids replaced by 9 random amino acids. also transformed efficiently. Results from these chimeric constructions indicate the carboxy terminus is responsible for the functional differences of pp60V’5rc and pp60c‘5rc and suggest the carboxy terminus of pp60c'5rc actually inhibits transformation. Chimeric constructs which will result in the production of infectious virus (51) produced somewhat different results upon their transfection into chicken embryo fibroblasts. Constructs containing sequences derived solely from c-src produced limited morphological transformation, with small. compact colonies that were highly sensitive to culture conditions. Virus recovered from the media was also weakly transforming; it was suggested these weakly transforming viruses may contain a mutated c-src gene. Both hybrid constructs containing c-src with the v-src 3' terminus and those containing v-src with the c-src 3' terminus transformed with 26 equal efficiency, and high-titer transforming virus could be obtained from the culture media. These results suggest the transforming potential is dependent upon interactions of the amino and carboxy portions of the protein, rather than residing solely in the carboxy terminus. These results are different from those obtained with the plasmid constructs in NIH3T3 cells (109), and may reflect differences between the primary chicken cells and the established mouse cell line or subtle differences in the constructs themselves, or could be a function of the ability to generate infectious virus. Thus, while the 3' terminus of the src gene appears to play a major role in determining the oncogenic potential. other factors may also be required. H. Involvement of Proto-oncogenes in Human Cancers A number of other human counterparts of retroviral oncogenes (proto-oncogenes) have been identified; most of the analyses have focused on the role of these genes in human cancers. At least two proto-oncogenes have been found at the breakpoint of chromosomal translocations associated with specific neoplasias, others have been identified by their ability to transform NIH3T3 cells. Proposed mechanisms of activation include both mutation and over-expression due to promotion. action of enhancers. amplification or deregulation. The involvement of c-myc, c-ras (Ha, Ki and N) and c-abl in tumorigenesis has been examined in some detail (myc is the transforming gene of MC29. Ha-ras of the Harvey murine sarcoma virus, Ki-ras of the Kirsten murine sarcoma virus and N-ras a human gene with sequences related to v-ras). In the majority of Burkitt's lymphoma cells, c-myc has undergone rearrangements (1,6,32,33,132). Three specific translocations to 27 immunoglobulin loci have been observed, translocation involving the 19K locus on chromosome 2, translocation to the IgH locus on chromosome 14, and translocation to the 191 locus on chromosome 22. In translocations involving either chromosome 14 or 22, the breakpoint occurs near or within the 5' end of the c-myc gene (6,132), thus the c-myc promoters may or may not be lost. The altered c-myc is usually overexpressed (42.49) and the ratio of utilization of the two promoters (if present) is frequently changed, with the Burkitt tumor cells favoring P1, normal cells favoring P2 (133). In most cases, the c-myc protein itself does not appear to be altered (6). It has been proposed that the translocation disrupts the 5' regulatory sequences (68.133), thus activation of c-myc involves deregulation. The role of c-myc in translocations involving chromosome 2 is not quite clear since the translocation actually occurs distal to the c-myc gene (33). Members of the human c-ras family have been implicated in several carcinomas of epithelial origin: activation of Ha-ras in bladder (37) carcinomas, activation of Ki-ras in lung (36.37) and colon (36,76) carcinomas and activation of N-ras in neuroblastomas (114.115). Elevated expression of N-ras has also been observed in bone marrow cells from a patient with acute myeloblastic leukemia (46). A single amino acid change has been found sufficient to activate both the Ha-ras gene (15.131) and the N-ras gene (130). Amplification of Ki-ras has been observed in several tumor cell lines (76). The c-abl gene has been implicated in the Philadelphia chromosome translocation associated with chronic myelogenous leukemia (5,35). C-abl is located at the breakpoint in at least one cell line (52), however, there is no evidence of increased expression. 28 Involvement of other proto-oncogenes in human neoplasia is not well docunented. Amplification of N-myc has been observed in neuroblastomas (2,63,101,102), and in small cell lung cancers (72), with the amplified sequences located in double minutes or homogeneously staining regions. The Blym gene isolated from Burkitt's lymphoma cells will transform NIH3T3 cells (38). At least one human T-cell leukemia virus (HTLV)-transformed cell line has elevated levels of c-sis expression (17; sis is the transforming gene of simian sarcoma virus). A cDNA clone containing the 3' portion of the c-sis message isolated from these cells was capable of transforming NIH3T3 cells. Elevated levels of c-src message have been found in a few sarcomas and mammary carcinomas (55). Evidence for the involvement of other c-onc genes in neoplasia is circumstantial. c-mos is located near the breakpoint of translocations common in the M2 type of acute nonlymphocytic leukemia (ANLL), c-fes near the breakpoint of translocations common in the M3 type of ANLL (141). The chromosomal region containing c-myb (myb is the transforming gene of avian myeloblastosis virus) is deleted in some cases of actue lymphocytic leukemia. A number of recent studies have suggested transformation is a multi-step process involving alteration of at least two or more cellular functions (66.82.98). Two distinct complementation groups have been identified. The first is involved in immortalization and can be assayed by the establishment of primary cells in tissue culture. This group includes myc, polyoma large T antigen and adenovirus E1a. The second complementation group is implicated in morphological changes and anchorage independence, and includes Ha-ras. N-ras and polyoma 29 middle T antigen. A third complementation group may be involved in determining the tumorigenic/metastatic potential of transformed cells. While the involvement of proto-oncogenes in neoplasia is becoming clearer, little is known of the roles of these genes in normal cells. Analysis of these proteins in transformed cells may shed light on their normal function. Likewise, elucidation of the normal role of proto-oncogenes should provide valuable insight into the biochemical changes leading to neoplasia. CHAPTER II STRUCTURAL ANALYSIS OF THE HUMAN C-SRC LOCUS 3O CHAPTER II MATERIALS AND METHODS A. Materials Restriction enzymes were purchased from Bethesda Research Laboratories, Boehringer Mannheim or Promega Biotec. EcoRI was a gift from Dr. A. Revzin. DNA polymerase I was from New England Nuclear; ligase from Promega Biotec; proteinase K and calf intestine alkaline phosphatase from Boehringer Mannheim; DNase, RNase and pronase were from Sigma. [o-3ZPJdNTPs were purchased from Amersham. Ampicillin, tetracycline chloramphenicol, lysozyme, salmon sperm DNA and yeast RNA were obtained from Sigma. All materials for electron microscopy were purchased from Pelco; cytochrome c was from Calbiochem. B. ‘Mggig BHI media is 37 9 brain heart infusion per liter. L8 is 10 g tryptone; 5 9 NaCl; 5 g yeast extract per liter. M9 is 6 g NazHP04; 3 g KH2P04; 0.5 9 NaCl; 1 g NH4Cl per liter, with 25 ml 20% glucose, 25 ml 20% casamino acids, 1 ml 2% thiamine HCl and 1 ml 1M MgClZ added after autoclaving. NZCYM is 10 g NZamine; 5 9 NaCl; 5 g yeast extract; 1 g casamino acids per liter, with 10 ml 1M M9504 added after autoclaving. All media was prepared in twice-distilled H20. Plates were made with 1-1.5% agar; tap agar contained 0.7% agar. 31 32 C. Isolation of Bacteriophage DNA Phage were grown in the host bacteria K802 or K803; phage were isolated by PEG precipitation followed by centrifugation through a glycerol step gradient (73). Liter cultures were grown in NZCYM media at 37°C until lysis was evident; 3 ml of chloroform was added and the cells incubated for an additional 15 min. The preparation was centrifuged for 10 min at 8K. RNase and DNase were added to the supernatent to a final concentration of 1 ug/ml each and the lysate incubated at 4°C for 60 min. 58.44 g of NaCl and 100 g of PEG were added and the preparation incubated at 4°C for another 60 min. The precipitate was spun at 8K for 10 min. The pellet was resuspended in 5 ml TM (50 mM Tris-HCl pH 8.0; 10 mM M9504) and extracted with an equal volume of chloroform. The phage solution was layered over a 5%/40% glycerol step gradient and spun at 35K for 60 min at 4°C. The pellet was resuspended in 2 ml SM (0.1 M NaCl; 10 mM M9504; 50 mM Tris-HCl pH 7.5; 0.01% gelatin), RNase was added to 50 ug/ml and DNase added to 1 ug/ml. The solution was incubated at 37°C for 30 min. 5X STEP buffer (0.5% SDS; 50 mM Tris-HCl pH 7.5; 0.4 M EDTA; 1 mg/ml proteinase K) was added to one-fifth the final volume and the solution incubated at 50°C for 15 min. The DNA was extracted once with phenol: chloroform (1:1) and once with chloroform and dialyzed against 20 mM Tris-HCl pH 8.1; 0.1 mM EDTA. 0. Isolation of Plasmid DNA Bacteria was grown at 37°C in M9 media in the presence of ampicillin (50 ug/ml) or tetracycline (15 ug/ml) until reaching an 00500 of 0.6. 150 mg dry chloramphenicol was added per liter and the culture incubated at 37°C for 16 hrs. Cells were chilled on ice 33 for 5 min, then centrifuged at 5K for 10 min. The pellet was resuspended in 40 ml 10 mM Tris-HCl; 1 mM EDTA pH 8.0 and recentrifuged at 7K for 10 min. The cell pellet was resuspended in lysozyme buffer (50 mM glucose; 10 mM EDTA; 25 mM Tris-HCl pH 8.0), 8 mg dry lysozyme was added and the suspension incubated for 30 min at 4°C. 8 ml of 0.2 N NaOH; 1% SDS was added, the sample was vortexed and incubated at 4°C for 5 min. 6 ml of 3 M NaAc pH 4.8 was added, mixed gently, and the suspension incubated at 4°C for 60 min. The sample was centrifuged at 15K for 20 min. Two volumes of ethanol were added to the supernatent and the nucleic acid precipitated at -20°C overnight. DNA (and RNA) was precipitated by centrifugation at 7K for 40 min. The pellet was resuspended in 5 ml 0.2 M NaCl and precipitated with ethanol a second time. The sample was again spun at 7K for 40 min. The pellet was resuspended in 15 ml 150 mM Tris-HCl; 1 mM EDTA pH 7.4, 16 g of cesium chloride and 0.75 ml of 10 mg/ml ethidium bromide were added and the sample centrifuged at 22°C, 37K for 48 hrs. The plasmid band was collected, extracted with butanol and dialyzed against 10 mM Tris-HCl; 0.1 mM EDTA pH 7.4. E. Isolation of Genomic DNA Cells were washed twice in cold phosphate buffered saline (PBS), then resuspended in 10 ml P85. 508 was added to 1% and pronase added to 0.5 mg/ml. The suspension was incubated at 37°C for 2 hrs. DNA was gently extracted twice with phenol, once with phenol: chloroform (1:1) and once with chloroform, then dialyzed against 10 mM Tris-HCl; 0.1 mM EDTA pH 7.4. 34 F. Restriction Enzyme Digestion Digestions were performed according to manufacturer's specifications. G. Agarose Gel Electrophoresis and Southern Transfer DNAs were electrophoresed through 0.5-2.0% agarose gels in TBE buffer (89 mM Tris-HCl; 89 mM boric acid; 2.5 mM EDTA pH 8.3). After electrophoresis, gels were stained with 0.5 ug/ml ethidum bromide for 20 min, then visualized and photographed under ultraviolet illumination. Gels were treated with 0.25 M HCl for 15 min, rinsed with distilled H20,incubated in denaturation buffer (1.5 M NaCl; 0.5 M NaOH) for 60 min, then in neutralization buffer (3M NaCl; 0.5 M Tris-HCl pH 7.4) for 120 min. DNA was transfered to nitrocellulose by the standard Southern blotting procedure (121), using 10 x SSC (20 x SSC is 3 M NaCl; 0.3 M sodium citrate pH 7.2). Transfer was carried out for 18 hrs for cloned DNA, 48 hrs for genomic DNA. After transfer, filters were air-dried and vacuum-baked for 2 hrs at 80°C. H. Isolation of DNA Fragments from Agarose Gels Large-scale digestions of DNA were electrophoresed through agarose as described above. The segment containing the desired DNA band was cut out of the gel and placed within dialysis tubing containing a minimal amount of TBE buffer. The DNA was electroeluted into the buffer. After electrophoresis, the agarose gel slice was gently removed from the dialysis tubing. The DNA was concentrated by passage over an Elutip-d column (Schleicher and Schuell). The DNA was then precipitated with ethanol. 35 I. Ligation All vectors with staggered ends were treated with phosphatase prior to ligation. After digestion, 1 M Tris-HCl pH 9.0 was added to a concentration of 0.1 M and the sample heated at 68°C for 10 min. Calf intestine alkaline phosphatase was added, followed by incubation at 37°C for 60 min. Vector DNA was extracted with phenol:chloroform (1:1) and precipitated with ethanol. Ligations were performed following manufacturer's Specifications. Ligations involving staggered ends were performed at 4°C, those involving blunt ends were performed at room temperature. All ligations were incubated overnight. J. Bacterial Transformation The host bacteria used was E. 2211 strain HB101. Cells were grown in 200 ml LB to an 00600 of 0.3. chilled on ice for 5 min and centrifuged at 5K for 10 min. The cell pellet was resuspended in 20 ml cold buffer A (1.4 g MnClz-4H20; 0.33 g NaAc; 0.44 g CaClZ-ZHZO per 100 ml). Cells were incubated fer 20 min on ice, centrifuged at 5K for 10 min, and resuspended in 6.7 ml cold buffer A. Ligated DNA was mixed with 0.2 ml competent cells and incubated on ice for 60 min. Cells were then heated at 37°C for 2 min and plated on BHI media containing ampicillin or tetracycline. K. Nick Translation Nick translations were performed following standard protocols (73), using all four radioactive dNTPs. 100 uCi (10 ul) [a-32p1dNTP, 2 ul 10x buffer (10x buffer is 0.5 M Tris-HCl pH 7.2; 0.2 M M9504; 1 mM DTT; 0.5 mg/ml BSA) and 0.5-1.0 ug DNA were mixed together and distilled H20 added to a final volume of 18.5 ul. A 1 mg/ml solution of DNase I was diluted 104-fold in cold distilled 36 water; 1 ul of the diluted DNase I was added to the nick translation mixture. One unit of E. ggli_DNA polymerase I was added and the sample incubated at 16°C for 60 min. Unincorporated nucleotides were removed by chromatography through a Sephadex G-50 column. The labeled DNA was precipitated wnth ethanol. Short DNA fragments (less than 500 base pairs) were ligated prior to nick translation to increase the efficiency of incorporation. L. Hybridization Hybridizations were performed at 42°C in 1 M NaCl; 10 mM Tris-HCl pH 7.4; 1 x Denhardts' solution (100x Denhardt's solution is 2% polyvinylpyrrolidone; 2% ficoll; 2% BSA); 45% formamide; 100 ug/ml sheared, single-stranded salmon sperm DNA; 100 ug/ml yeast RNA. After hybridization, filters were washed in 0.2 x SSC; 0.2% SDS at 50°C (conventional) or 64°C (high stringency). These conditions are described in detail in Chapter 5. M. Elution of Probes from Nitrocellulose Probes were eluted from nitrocellulose filters by washing the blots in 0.1 M NaOH, followed by neutralization with 3 M NaCl; 0.5 M Tris-HCl pH 7.4. All filters were then exposed to X-ray film to insure complete removal of probe. If treated gently, the filters could be subjected to 8 rounds of hybridization before finally disintegrating. Only slight loss of filter-bound DNA was observed. N. Heteroduplex Analysis A modified renaturation protocol was used (34). Briefly, 250 ng of each DNA species were gently mixed together in a final volume of 25 ul containing 0.1 N NaOH; 20 mM EDTA, and incubated at room temperature for 10 min. 5.5 ul of 1 M TES was added by gently mixing, then 2.5 ul 37 of distilled H20 and 22 ul (relaxed) or 27 ul (stringent) deionized formamide were added. DNAs were allowed to renature for 60 min at room temperature. The samples were dialyzed overnight in 10 mM Tris-HCl; 1 mM EDTA pH 7.8 at 4°C. The Kleinschmidt formamide protocol (62) was used to spread the samples. Several formamide concentrations were assayed for stability of the heteroduplexed regions. A formamide concentration of 40% in the hypophase, 10% in the hyperphase was found to be Optimal, although some variation was noted from experiment to experiment. Generally, the heteroduplexes were not stable at higher formamide concentrations, and tangling was a problem at lower concentrations. RESULTS A. Isolation and Preliminary Analysis of Human c-src Clones To obtain genomic clones spanning the human c-src locus, a human genomic library (67) was screened with the 3.1kb Eco RI fragment of cloned RSV-SRA-Z (D. Fujita and S. Anderson). This clone contains the entire 1.6kb v-src coding region, as well as flanking sequences, which do not hybridize to hunan DNA. Ten clones containing human c-src sequences were obtained; four of these clones, AS3H, AS4H, ASSH and ASIIH were examined in detail. A preliminary restriction enzyme map of the cloned region is shown in Figure 1A; the inserts from the four phage clones are illustrated in Figure 18. Initial restriction mapping of the lambda clones demonstrated the inserts overlap extensively, together spanning a region approximately 24kb in lenth. The genomic library was found to be severely biased: Figure 1. 38 Clones spanning the human c-src locus. A. Preliminary restriction enzyme map of the human c-src locus. B. Phage clones isolated from the human fetal liver library. C. Fragments subcloned into plasmid vectors. 0, Eco RI; 0, BamHI; V, HindIII; A, XhoI. 39 II odma Qdma memze maze Nazme «fine 93 me Qdma «sue :vm‘ :nmx :mm4 e... .m ~ ~ ~ + w » y ma « m .m 40 other clones isolated overlap with AS3H and AS4H, rather than with ASIIH. Southern hybridizations using the 3.1kb v-src probe indicated that the human c-src sequences are widely distributed within the cloned region. Sequences homologous to v-src were detected in the 7.2kb EcoRI fragment as well as the 7.8 , 2.8 and 0.6kb BamHI fragments. This distribution of v-src related sequences over a region larger than 10kb suggested the human c-src gene contained several introns. To localize the exons more precisely, the human c-src locus was extensively mapped and analyzed by hybridization to a battery of region-specific v-src probes. Restriction mapping was facilitated by subcloning specific fragments into plasmid vectors: the subclones that were constructed are shown in Figure 1C. Inserts from the subclones were gel-purified away from the plasmid vector and subjected to single and double restriction enzyme digestions; the DNAs were then electrophoresed through 1.0-1.8% agarose gels. After staining with ethidium bromide. bands as short as 50-75 base pairs in length could usually be detected. The restriction enzyme map was constructed by logical deduction from the fragment sizes generated by the various restriction enzyme combinations. The majority of sites were later confirmed by partial digestions of end-labeled DNA (118) (A. Tanaka, personal conmunication). A summary of the mapping data is presented in Figure 2. It should be noted the 5' 1.0kb and the 3' 0.8kb have not been subcloned and therefore certain enzyme sites have not been mapped within these regions. B. Localization of Exons Hybridizations with v-src subgenomic probes were performed to identify narrowly-defined regions homologous to v-src. Five 41 Figure 2. Detailed restriction enzyme map of the human c-src locus. The 5' 1.0kb and 3' 0.8kb have not been mapped to completion with PstI, PvuII, SacI and SmaI. e: 42 _ _ Sex _ _ fl _ _ :sx _ _ _ _ _ = A _ _ _ _ _4 .25 __E_ -__:_ q___ _= =__ a : n... 41 _ _ _ _ _ = _ = :_ _ _ a _ _ — "Home _ 2.3. _ _ =35: _ _ Eoom _ _ fl — __ _ _ _ = .28 o FL _ _ _ _ "Lemon 43 region-specific v-src probes were constructed by subcloning the v-src PstI fragments into the pBR322 plasmid vector (Figure 3). Southern blots containing digests of the various human c-src subclones were analyzed by successive hybridization to each of the 5 v-src subgenomic probes. Typical hybridization results are illustrated in Figure 3. In this example, a series of double digestions were perfonmed on the 14.6kb EcoRI fragment, which had been gel-purified from the pE14.6 plasmid (see Figure 1). This fragment extends from the second EcoRI site to an aritificial EcoRI site (generated during construction of the library by addition of synthetic EcoRI linkers) present at the 3' end of the xS3H insert. The ethidium bromide stained gel is shown in panel A. The blot was hybridized sequentially to v-src probes I-V. Probes I and V did not hybridize to this fragment. Hybridizations to probes II, III and IV are shown in panels B-D. Each probe detects a specific set of fragments, although some of the fragments hybridize to both 11 and III or to III and IV. Similar analyses were performed using a variety of restriction enzymes and other human c-src fragments. Specific fragments hybridizing to the v-src probes were aligned on the restriction enzyme map, allowing delineation of those regions homologous to v-src, presumably the human c-src exons. Figure 4 summarizes the exon-intron map obtained from the hybridization analyses.. A minimum of 9 distinct regions bearing homology to v-src can be identified, with the homologous sequences distributed over a region 20kb in length. The exons have been numbered by correspondence to the chicken c-src exons, based on heteroduplex and sequence analysis to be discussed in detail Figure 3. 44 Hybridizations using the v-src subclones. The 3.1kb EcoRI fragment of cloned RSV-SRA-Z is illustrated, the box depicts the v-src coding region. PstI sites are indicated by vertical lines; fragments used as probes are labeled I-V. A. Ethidium bromide stain of restriction digests of the 14.6kb EcoRI fragment containing the 3' portion of the human c-src locus. B. Hybridization to probe 11. C. Hybridization to probe III. 0. Hybridization to probe IV. E. Hybridization to the BLURB probe. Digestions are as follows: (1), SacI; (2), SacI + BamHI; (3), SacI + BglII; (4), SacI + HindIII; (5), SacI + KpnI; (6), SacI + SalI; (7). SacI + XbaI; (8), SacI + XhoI; (9), SacI + SmaI. Molecular weights are given in kb, sizes were calculated from a standard of HindIII-digested lambda DNA and HaeII-digested pBR322 DNA. 45 a: m6 05¢u> Figure 4. 46 Localization of human c-src exons. Regions hybridizing to v-src probes I-IV are depicted beneath the restriction map of the human c-src locus. The v-src map illustrates the PstI probes used. A smmnary of the hybridization data is shown in the exon map at the bottom of the figure. Exons are indicated by open boxes and are numbered according to correspondence to the chicken c-src exons. Regions hybridizing to the BLUR8 probe are also indicated. 0, Bam HI; 9, Eco RI; 'V, Hind III; A, XhoI. 47 DD _U D D D 25.5 SN Q. a >_ m a a :— m a m __ 4.4 fl fl _ .m H F p «F » a t. w .m > >_ =_: _ 93.0 _§fifl§ 48 below. Of particular interest is probe V. which contains the divergent 3' terminus of the v-src gene. This probe did not hybridize to the human c-src clones analyzed here. The placement of exon 3 remains tentative. With the exception of the 3' 17 base pairs, which correspond to the 5' end of probe II. the chicken c-src exon 3 is primarily homologous to probe 1. Probe I has a high GC content (66%) and will bind non-specifically to several fragments, complicating the hybridization analyses. The region labeled exon 3 in Figure 4 hybridizes somewhat more intensely than other fragments, but the bands are not stable under stringent washing conditions (see below). It is possible these bands are an aritifact due to the GC content; an alternative location of exon 3 is downstream from, but very close to, exon 2. In this location, exon 2 and exon 3 would not be distinguished by the restriction enzymes used. Somewhat weaker hybridization was observed for several of the exons (3,4,5 and 9); in these regions, hybridization to v-src is not stable under stringent conditions. For example, when a Southern blot containing restriction digests of the 7.8kb BamHI fragment (gel-purified from the pB7.8 plasmid, see Figure 1) is hybridized to probe 11, two bands can be detected in each lane (Figure 5). These bands correspond to exons 4 + 5 and 6. After washing under stringent conditions, only one band, corresponding to exon 6, remains in each lane. Similar results were obtained for exons 3 and 9. It can be estimated from the reaction conditions (see Chapter 5 for a detailed explanation) that these exons share less than 90% sequence homology with v-src. C. Human c-src and Chicken c-src Heteroduplex Analysis Figure 5. 49 Differential hybridization under different stringencies. Digests of the 7.8kb BamHI fragment was hybridized to probe 11 under conventional and stringent conditions. A. Ethidium bromide stain. B. Hybridization to probe II under conventional conditions. C. Hybridization to probe II under stringent conditions. Digestions are as follows: (1), PstI; (2), PstI + Hind III; (3), PstI + PvuII; (4), PstI + SacI; (5), PstI + SalI; (6), PstI + XhoI. Molecular weights are given in kb, sizes were calculated from a standard of HindIII-digested lambda DNA and HaeII-digested pBR322 DNA. 50 a. ENS-um mv 51 Heteroduplexes of human c-src and chicken c-src lambda clones were examined by electron microscopy (Figure 6). In heteroduplexes of 153M, which contains human c-src exons 4-12, and the chicken c-src clone 1460, which contains chicken c-src exons 1-8, four double-stranded regions are consistently observed. Size measurements indicate these regions correspond to exons 4 and 6-8 of chicken c-src. The large loop between exon 8 and the right arm of the lambda vector contains the 3' portion of the AS3H insert. including exons 9-12. which are not present in A460. Heteroduplex at exon 5 was found only in two of the molecules examined. Under the most stringent renaturation and spreading conditions, only heteroduplex at exon 8 remained stable. Attempts to form heteroduplexes at exons 9-12 were unsuccessful due to technical reasons. The inserts of the human c-src clones and the available chicken c-src clone containing the 3' portion of the gene (A54E) are in the opposite orientations. During the renaturation step. the vector arms rapidly anneal, leaving the inserts in reverse orientations and therefore unable to form heteroduplexes. D. Hybridization of v-src to Genomic DNA To verify the authenticity of the human c-src clones, Southern blots of normal human DNA (line NF812, provided by 0. Fry) were probed with the v-src 3.1kb EcoRI fragment (Figure 7). Based on the restriction enzyme map of the cloned region, BamHI digestion (lane A) should yield 7.8, 2.8 and 0.6kb fragments, as well as a 5'-specific fragment that extends upstream and is larger than 7.2kb. Bands corresponding to the 7.8. 2.8 and 0.6kb fragments can be detected. In addition, three other bands of 10.5, 9.2 and 6.7kb hybridize to the 52 Figure 6. Heteroduplex between human c-src and chicken c-src clones. A. Electron micrograph of a heteroduplex between ASBH and 1460. 8. Schematic representation of the introns and exons of the heteroduplex in A. 53 Figure 7. 54 Hybridization of v-src to human genomic DNA. (1) BamHI and (2) EcoRI digests of human genomic DNA were probed with the 3.1kb EcoRI fragment containing the entire v-src coding region. Sizes of the hybridizing fragments (difficult to see in the photographic reproduction) are indicated. Molecular weights were calculated from a standard of HindIII digested lambda DNA. 55 56 v-src probe; either the 10.5 or the 9.2kb fragment presumably contains the 5' end of the c-src locus present in the clones analyzed here. Digestion of human DNA with EcoRI (Figure 7. lane B) should produce a 7.2kb fragment and a fragment larger than 14.6kb that contains the 3' sequences and extends downstream. Both a large fragment of approximately 17kb and a 7.2kb band are detected with the v-src probe. In addition, an 11kb fragment also hybridizes. When the v-src PstI subclones (see Figure 3) are used to probe genomic DNA, faint bands can be seen in addition to the bands predicted by the retriction enzyme map (0. Fujita, personal communication). These additional bands suggest the presence of other v-src related sequences within the human genome. E. Localization of Human Alu Repeat Sequences Use of the human c-src clones to probe human genomic DNA results in a smear, strongly suggesting repeat sequences are present within the lambda inserts. To identify the location of these sequences, the BLURB plasmid (56), which contains one member of the human Alu repeat family was used as a probe. As an example, hybridization to the 14.6kb Eco RI fragment is shown in Figure 3E. Altogether, 6 regions containing human Alu repeat sequences were localized within the human c-src locus (Figure 4). DISCUSSION Four overlapping clones encompassing the human c-src locus have been examined in detail. Together the cloned inserts span a region 24kb in length (Figure 1); v-src related sequences have been localized 57 within these clones to a region 20kb in length. Thus the human c-src locus is far larger than the corresponding chicken c-src locus, which is approximately 8kb in length (86.110.128.129). By combining extensive restriction enzyme mapping with hybridization to region-specific v-src probes. 9 specific regions homologous to v-src have been identified (Figure 4). These regions have been numbered 2-12. based on correspondence with the chicken c-src exons observed in heteroduplex analysis as well as direct nucleotide sequence comparison, which will be presented in Chapter 3. In addition, use of the v-src probes has allowed the orientation to be determined as indicated in Figure 4. It is assumed the homologous regions contain exon sequences, while the regions that do not hybridize to v-src are introns, although this has not been conclusively demonstrated. The location of exon 3 has not yet been unequivocally determined. Assignment of this exon to the position indicated in Figure 4 is based solely on hybridization to probe I. However, due to a high GC content (66%), probe I hybridizes weakly to numerous fragments throughout the v-src locus, although hybridization is more intense at the region labeled exon 3 in Figure 4. Nevertheless, it is possible the hybridization attributed to exon 3 is an artifact of this non-specific binding. In addition, hybridization in this region is not stable under stringent washing conditions. This phenomenon has also been observed for other exons, however, notably 4, 5 and 9. and could be the result of nucleotide sequence divergence. An alternative location for exon 3 is close to the 3' end of exon 2, with no pertinent restriction enzyme sites between the exons. In this position, exons 2 and 3 would not be 58 distinguished by the restriction mapping and hybridization procedures utilized. A third possibility, the nucleotide sequence of exon 3 might be so divergent that it does not hybridize to v-src probe 1, is also consistent with the data. Both the heteroduplex and hybridization analyses suggest certain regions of the human c-src coding sequence are somewhat less homologous than others. In particular, hybridizations at exons 3 (if identified correctly), 4, 5 and 9 are not stable under stringent conditions (see Figure 5), with an estimated 10-20% base mismatch in these regions. Heteroduplex analysis indicates exon 5 is highly divergent; few hybrid molecules contained heteroduplex at exon 5. The inability of probe V to hybridize to the human c-src clones is particularly intriguing. This probe is specific for the divergent 3' terminus of v-src. It is not present at the 3' terminus of the chicken c-src gene, instead, this sequence is located approximately 900 base pairs downstream from the chicken c-src termination codon (128). Unfortunately, the human c-src clones only extend about 800 base pairs downstream from the termination site. It is possible sequences corresponding to the divergent terminus are located further downstream. Clearly, genomic clones extending further downstream should be examined for the presence of this sequence. The size fragments obtained in genomic blots of normal human DNA using the v-src probe suggests the lambda clones are indeed authentic and have not undergone any obvious gross rearrangements during the cloning procedure. All bands predicted by the restriction enzyme map can be detected. However, in addition to the predicted bands. other bands can also be seen. Although the intensity of these bands is simi- 59 lar to that of the predicted bands when the 3.1kb v-src fragment is used as probe (Figure 7), the extra bands only hybridize weakly when the v-src PstI subclones are used as probe (0. Fujita, personal commun- ication). The reason for the intensity difference is not clear, but may be a result of the different stringencies of the hybridization con- ditions used. Some of the extra bands are smaller than the predicted bands and therefore can not be explained by incomplete digestion of the DNA. In addition, the extra bands have been observed in DNA from three human cell lines (NF812, HT1080 and T044. provided by 0. Fry) as well as in DNA extracted from human placenta (D. Fujita, personal communication). The presence of additional bands suggests the existence of a second c-src locus or src-related sequence within the human genome. It may be that the two human c-src loci encode the two forms of pp60c‘s"c observed in human cells (111). Analysis of these src-related sequences should prove to be very interesting. Finally, six regions containing human Alu repeat family sequences have been identified within the human c-src locus. Generally, these sequences are spaced approximately 2-3kb apart (56), thus there is nothing abnormal about the distribution within the c-src locus. Although the function of the repeat sequences is unknown. it has been suggested they serve as origins of DNA replication (56). CHAPTER III NUCLEOTIDE SEQUENCE ANALYSIS OF THE HUMAN C-SRC LOCUS 6O CHAPTER III MATERIALS AND METHODS Digestions, isolation of DNA fragments from agarose gels, and ligations were described in Chapter 2. A. Materials Klenow polymerase was purchased from Boehringer Mannheim. Replicative forms of MP8 and MP9 (79) and dideoxyNTPs were obtained from P-L Biochemicals. The pentadecamer universal primer was from New England Biolabs. DeoxyNTPs were obtained from Sigma. ISOprOpylthio-e-D-galactoside (IPTG) and 5-bromo-4-chloro-3-indolyl-3-D-galactoside (X-Gal) were from Bethesda Research Laboratories. B. Media YT media is 8 g tyroptone; 5 9 NaCl; 5 g yeast extract per liter. 2xYT is double this concentration. C. M13 Transformation E, 5211 strain JM103 was transformed by standard protocols (78). Host cells were grown in 100 ml YT media to an 00500 of 0.3-0.4. Cells were centrifuged at 6K for 5 min, resuspended in 50 ml cold 50 mM CaClZ and chilled on ice for 20 min. Cells were recentrifuged and resuspended in 10 ml cold 50 mM CaClz. Ligated DNA was mixed with 0.3 ml competent cells and the mixture incubated on ice for 40 min. The cells were heat-shocked for 2 min at 42°C. 10 ul 0.1 M IPTG, 50 pl 61 62 2% X-Gal and 0.2 ml exponentially growing JM103 cells were added and the bacteria plated out on YT agar. D. In situ Screenipgyof M13 Plaques Plates containing 100-300 plaques were screened by 13 gigp hybridization. Nitrocellulose filters were gently placed on the plates and allowed to adsorb for 30 sec. Filters were then immersed in denaturation buffer (0.1 N NaOH; 1.5 M NaCl) for 5 min, then transferred to neutralization buffer (3 M NaCl; 0.5 M Tris-HCl pH 7.0) for 10 min. Filters were dried and baked in a vacuum oven at 80°C for 2 hrs. E. M13 Dot Blot Assay A saturated JM103 culture was diluted 1:100 into 2xYT media. Clear plaques were transferred into 2 ml of the diluted JM103 culture and incubated at 37°C for 6-8 hrs. 20 pl of the supernatent was spotted onto nitrocellulose. The filters were submerged in denaturation buffer (0.1 N NaOH; 1.5 M NaCl) for 5 min, then transferred to neutralization buffer (3 M NaCl; 0.5 M Tris-HCl pH 7.0) for 10 min. Filters were air-dried and vacuum-baked for 2 hrs at 80°C. F. Hybridization Hybridizations were performed as described in Chapter 2, except the hybridization buffer contained 25% formamide and the washings were conducted at 36°C. G. Template Isolation. 2 ml cultures were spun in a Brinkman microfuge for 10 min. 1.2 ml of the supernatent was transferred to an Eppendorf tube containing 0.3 ml 20% PEG; 2.5 M NaCl. The sample was mixed well and incubated at 63 room temperature for 30 min. The phage were pelleted by a 10 min centrifugation in the microfuge and resuSpended in 160 pl TES (20 mM Tris-HCl pH 7.5; 10 mM NaCl; 0.1 mM EDTA). The DNA was extracted with 160 pl phenol: chlorofonn (1:1); 140 pl of the aqueous phase was then reextracted with 140 pl phenol: chloroform. 120 pl was removed, 12.5 ul 3 M NaAc pH 5.2 and 340 pl 95% ethanol were added and the mixture left at -20°C overnight. Prior to the annealing reaction, the samples were spun 10 min in the microfuge and resuspended in 20 ul TES. H. Dideoxy Sequencing Reaction To anneal the primer to the template, 7.5 ul template, 1.5 ul 10X Hin (60 mM Tris-HCl pH 7.5; 60 mM MgClz; 10 mM DTT; 0.5 M NaCl) and 0.2 pmole primer were mixed and the volume brought to 13.5 ul with distilled H20. The mixture was heated at 98°C for 3 min, incubated to 65°C for 30 min and chilled on ice for 5 min. Samples were used immediately or frozen at this point. Several modifications of the dideoxy reaction procedure (78,99) were made in an attempt to develop a reliable protocol using all four labeled [a-32PJdNTPs. 1 ul [a-32PJdNTPs (1o uCi) and 0.5u Klenow polymerase were mixed with the primed template and the sample incubated at room temperature for 10 min. 3 nI was then transferred to each of 4 tubes (containing 1 ul of the N° and 1 ul of the ddN solutions described below) for the C, T, A and G reactions. Samples were incubated at 37°C for 30 min, or occasionally at higher temperatures when strong secondary structure was a problem. 1 ul of chase solution (0.125 mM each dCTP, dTTP, dATP and dGTP) was added and the solutions incubated at room temperature for 15 min. Finally. 10 ul of stop dye (0.3% xylene cyanol; 0.3% bromphenol blue; 99% formamide; 64 10 mM EDTA pH 7.5) was added. Prior to electrophoresis, the samples were heated at 98°C for 3 min and chilled on ice. Optimal concentration of dNTPs and ddNTPs in the reactions were determined in the following manner. Each sequencing reaction should contain an excess of the other three dNTPs. with the dNTP of the reaction in limiting quantities. The extent of elongation and the degree of termination are dependent upon the dNTP:ddNTP ratio. Initially, the dNTP present in the [0-32P]dNTPS was used as the limiting factor. These reactions proved to be fairly reproducible. but occasionally the concentration of one of the dNTPS in the isotope preparation was observed to be low, causing reactions to terminate prematurely due to the decrease in the dNTP:ddNTP ratio. Conditions were therefore developed that would not rely so heavily on the dNTP concentration in the [a-3ZPJdNTP preparation. Briefly, the N° solutions contained 1 ul 0.5 mM dNTP (limiting), 15 ul of each of the other three dNTPs (0.5 mM each, in excess), 20 ul 10X Hin buffer and 14 ul distilled H20. Concentrations of the ddNTPs were determined empirically and could be adjusted for early or late termination. These conditions, although not yet tested extensively, have thus far proven to yield consistent results: 400-500 bases can be read on a good gel. I. Sequencing Gel Electrophoresis Both 8% acrylamide gels and 6% buffer gradient gels were used. The stock solutions of the 8% gel consisted of 420 g urea; 200 ml 40% acrylamide (deionized); 4 g bis-acrylamide; 100 ml 10XTBE (10X TBE is 162 g Tris base; 27.5 g boric aid; 9.3 9 EDTA per liter) per liter. The stock solution was degassed for 30 min and stored at 4°C. Each 85 X 17.8 cm gel required 100 ml of stock solution, 1 ml 10% APS, and 30 ul TEMED. 65 The buffer gradient gel specifications were modified from the standard protocol (8) to decrease the spacing between bands. The top solution consisted of 6% acrylamide (38:2 acrylamide: bis-acrylamide); 0.5 X TBE; 8 M urea. The bottom solution contained 6% acrylamide (38:2 acrylamide: bis-acrylamide); 2.5 X TBE; 8 M urea; 10% sucrose and enough bromphenol blue to turn the solution dark blue. Both solutions were degassed 30 min and stored at 4°C. To pour the gels, 20 ml of bottom and 20 ml of top solution were placed in separate beakers. 200 pl of 10% APS was added to each solution. 5 ul TEMED was added to the bottom solution; 10 ul was added to the top. The solutions were then mixed well. 15 ml of the top solution was drawn up into a 25 ml pipet. 15 ml of the bottom solution was carefully drawn up into the same pipet. The solutions were gently mixed by allowing a few air bubbles to pass up the pipet. The mixture was then poured between the glass plates. The gradient was immediately overlaid with top acrylamide (60 ml top solution; 600 pl 10% APS; 15 ul TEMED). Gels were run with 0.5 X TBE in the upper reservoir and 2.5 X TBE in the lower reservoir. RESULTS A. Isolation of M13 Clones The human c-src exons were sequenced by the dideoxy method using an M13 selection protocol. Briefly, a restriction fragment containing the exon of interest was gel-purified, digested with an enzyme compatible with the M13 vectors (AluI, HaeIII, RsaI and Sau3A were used), and the resulting fragments subcloned into the MP8 or MP9 vectors (79). Phage from clear plaques were grown in 2 ml cultures for 6-8 hrs. M13 clones harboring inserts containing v-src related 66 sequences were identified by dot blot hybridization using the v-src PstI subclones as probe. Alternatively, a modified in gltp hybridization procedure was used. However, this technique was found to be less sensitive and could not distinguish clones containing only a short stretch (less than 30 nucleotides) of v-src related sequence. All hybridizations were performed under relaxed conditions in order that short v-src related sequences could be detected. The M13 clones were sequenced by a modified dideoxy protocol (see materials and methods). Initially, clones positive in the dot blot assay were screened by performing only the G reaction. This rapid . screen allowed the identification of identical inserts and provided an estimate of the amount of template DNA recovered as well as anl indication of the background level. B. Comparison of the Human c-src Sequence with v-src and Chicken c-src Sequences. All sequences were obtained by reading both strands and, if possible, by analysis of overlapping clones. Src-related sequences were identified by visual analysis or with computer assistance. The human c-src sequence has been determined for exons 2 and 4-12. exon 3 has not been sequenced. A comparison of the human c-src, chicken c-src and SR-A v-src nucleotide and deduced amino acid sequences is presented in Figure 1. Although a number of nucleotide changes are present, the major portion of the amino acid sequence is very similar, with only a few scattered substitutions. The human c-src retains the chicken c-src 3' terminus, rather than the divergent v-src terminus (128). The sequence of exon 2 appears to have undergone a rearrangement in the human locus (Figure 2). The 5' one-third and 3' one-third are Figure 1. 67 Nucleotide sequence of the human c-src coding region. The nucleotide sequence and deduced amino acid sequence of exons 4-12 are compared between the human and chicken c-src and SR-A v-src genes. The chicken c-src and SR-A v-src sequences (128) are given only where different from the human c-src sequence. Exons are denoted by arrows, divergent amino acids are indicated by bars. Numbering of nucleotides and amino acids follow reference 128. Portions of the human c-src sequence were obtained by S. Anderson and A. Tanaka. human 'chiok virus human chick virus human chick virus human chick virus human chick virus human chick virus human chick virus 683 4 AGAGGGAGACTGGTGGCTGGCCCACTCGCTCAGCACAGGACAGACAGGCTACATCCCCAGCAAC GluGlyAspTrpTrpLeuAlaHisSerLeuSerThrGlyGlnThrGlyTyrIleProSerAsn G A T T T C CT G T GluGly AlaHisSer Thr Thr Ser G A TA A T CG G C G T GluGlyAsn AlaHisSerValThr Thr Ser +5 TACGTGGCGCCCTCCGACTCCATCCAGGCTGAGGAGTGGTATTTTGGCAAGATCACCAGACGGGAG TeralAlaProSerAspSerIleGlnAlaGluGluTrpTerheGlyLysIleThrArgArgGlu T C A A C 6 TC T Teral Ser G1u Tyr 01y ThrArg T C A A C G TC T Teral Ser Glu Tyr Gly ThrArg TCAGAGCGGTTACTGCTCAATEEAGAGAACCCGAGAGGAECCTTCCTCGTGCGAGAAAGTGAGACC SerGluArgLeuLeuLeuAsnAlaGluAsnProArgGlyProPheLeuValArgGluSerGluThr C C G CC C A CC G A T G C G G C G Ser Leu AsnProGlu ProArg Thr LeuValArgGIuSer Thr T C G CC C A CC G A T G C G G C G Ser Leu AsnProGlu ProArg Thr LeuValArgGluSer Thr 6 ACGAAAGGTGCCTACTGCCTCTCAGTGTCTGACTTCGACAACGCCAAGGGCCTCAACGTGAAGCAC ThrLysGlyAlaTerysLeuSerValSerAspPheAspAsnAlaLysGlyLeuAanalLysHis A T C T T G T Thr Tyr SerVal Phe Gly Asn A T C T T G T Thr Tyr SerVal Phe Gly Asn TACAAGATCCGCAAGCTGGACAGCGGCGGCTTCTACATCACCTCCCGCACCCAGTTCAACAGCCTG TerysIleArgLychuAspSerGlyGlyPheTyrIleThrSerArgThrGluPheAsnSerLeu A A G Ser Thr Ser A A G Ser Thr Ser 7 CAGCAGCTGGTGGCCTACTACTCCNAACACGCCGATGGCTCGTGCCACCGCCTCACCACCGTGTGC GlnGlnLeuValAlaTeryrSerLysHi3A1aAspGlySerCysHisArgLeuThrThrVales T T T G A C HisAla Leu Leu Aanal T T T G A C HisAla Leu Leu Aanal CCCACGTCCAAGCCGCAGACTCAGGGCCTGGCCAAGGATGCCTGGGAGATCCCTCGGGAGTCGCTG ProThrSerLysProGlnThrGlnGlyLeuAlaLysAspAlaTrpGluIleProArgGluSerLeu C C A C C G A C Pro Thr GlyLeu AspAla Glu Pro C C A C C G A C Pro Thr GlyLeu AspAla Glu Pro 405 135 471 157 537 179 603 201 669 223 735 2H5 801 267 human chick virus human chick virus human chick virus human chick virus human chick virus human chick virus human chick virus 69 8 CGGCTGGAGGTCAACCTGGGGCAGGGCTGCTTTGGCGAGGTGTGGATGJLGACCTGGAACGGTACC ArgLeuGluVa1AsnLeuGlyGlnGlyCysPheGlyGIuValTrpHetGlyThrTrpAsnGlyThr G G A C C ValLeu Gly Val Gly G G A C C ValLeu Gly Val Gly ACCAGGGTGGCCATCAAAACCCTAAAGCCTGGCAACATGTCTEAEGAGGCCTTCCTGCAGGAAGCC ThrArgValAlaIleLysThrLeuLysProGlyAsnMetSerGlnGluAlaPheLcuGlnGluAla A A G T G C C C Arg IleLysThrLeu Pro SerPro A A G T G C C C Arg IleLysThrLeu Pro Schro CAGGTCATGAAGAAGCTGAGGCATGAGAAGCTGGTGCAGTTGTATGCTGTGGTTTCAGAGGAGCCC GanalMetLysLychuArgHisGluLysLeuValGlnLeuTyrAlaValValSerGluGluPro A G CC T C C A G G A Ganal LeuArg Val LeuTyrAla ValSerGlu A 6 CC T AC C A G G A Ganal LeuArg ValGlnLeuTyrAla ValSerGlu p9 ATTTACATCGTCACGGAGTACATGAGCAAGGGGAGTTTGCTGGACTTTCTCAAGGGGGAGACTGGC 11cTyrIleValThrGluTeretSerLysGlySerLeuLcuAspPheLeuLysGlyGluThrGly C T CC C T C G A TG Ile Thr SerLeu AspPheLeu Gly Met C TT CC C T C G A TO Ile Ile SerLeu AspPheLeu Gly Met +10 AAGTACCTGCGGCTGCCTCAGCTGGTGGACATGGCTGCTCAGATCGCCTCAGGCATGGCGTACGTG LysTereuArgLeuProGlnLeuValAspMetAlaAlaGInlleAlaSerGlyMetAlaTeral A C C T T A C C T Pro LeuValAsp IleAlaSer AlaTyr A C T T T A C C T Pro LeuValAsp IleAlaSer AlaTyr GAGCGGATGAACTACGTCCACCGGGACCTTCGTGCAGCCAACATCCTGGTGGGAGAGAACCTGGTG GluArgMetAsnTeralHisArgAschuArgAlaAlaAsnIleLeuValGlyGluAsnLeuVal A G A G G G G Arg Val Ara LeuArgAla Gly A G A G G G G Arg Val Arg LcuArgAla Gly 11 TGCAAAGTGGCCGACTTTGGGCTGGCTCGGCTCATTGAAGACfi‘I-‘GAGTACACGGCGCGGCAAGIGT CysLysValAlaAspPheGlyLeuAlaArgLeuIleGluAsleeGluTerhrAlaArgGlnGIy G T A C C G AC A A Lys Ala AlaArg IleGlu Asn ThrAla G T A C C G AC A A Lys Ala AlaArg IleGlu Asn ThrAla 867 289 933 311 999 333 1065 355 1131 377 1197 399 1263 421 human chick virus human chick virus human chick virus human chick virus human chick virus human chick 7O GCCAAATTCCCCATCAAGTGGACGGCTCCAGAAGCTGCCCTCTATGGCCGCTTCACCATCAAGTCG 1329 A1aLysPheProIleLysTrpThrAlaProGluAlaAlaLeuTyrGlyArgPheThrIleLysSer 443 G A C C G A G Lys ThrAlaProGluAla Ara G A C C G A G Lys ThrAlaProGluAla Arg 12 GACGTGTGGTCCTTCGGGATCCTGCTGACTGAGCTCACCACAAAGGGACGGGTGCCCTACCCTGGG 1395 AspValTrpSerPheGlyIIeLeuLeuThrGIuLeuThrThrLysGlyArgValProTerroGly 465 T C C G C C A A AspVal Gly Leu Thr Gly Pro Pro T C C G C C A A AspVal G1y Leu Thr Gly Pro Pro ATGGTGAACCGCGAGGTGCTGGACCAGGTGGAGCGGGGCTACCGGATGCCCTGCCCGCCGGAGTGT 1461 MetValAsnArgGluValLeuAspGlnValGluArgGlyTyrArgMetProCysProProGluCys 487 C A G A C C C Val Ar; Ar; Ar; Pro Cys GC G G G A C C C Gly Gly Arg Arg Ara Pro Cys CCCGAGTCCCTGCACGACCTCATGTGCCAGTGCTGGCGGAAGGAGCCTGAGGAGCGGCCCACCTTC 1527 ProGluSerLeuHisAspLeuMetCysGlnCysTrpArgLysGluProGlucluArgProThrPhe 509 G T G C T T Bar His ArgAsp ThrPhe G T T G C T T Ser His Leu ArgAsp ThrPhc GAGTACCTGCAGGCCTTCCTGGAGGACTACTTCACGTCCACCGAGCCCCAGTACCAGCCCGGGGAG 1593 GluTercuGlnAlaPheLeuGluAspTyrPheThrSerThrGluProGlnTyrGInProGlyGlu 531 C G A T A ThrSerThr ProGly CAGCTGCTTCCTGCTTGTGTGTTGGAGGTCGCTGAGTAG GlnLeuLeuProAlaCysValLeuGluValAlaGlu AACCTCTAG 1599 AsnLeu 533 A Leu Figure 2. 71 Nucleotide sequence of the rearranged exon 2. The nucleotide and deduced amino acid sequence of exon 2 is compared between the human and chicken c-src (128) genes. Sequences are aligned to maximize nucleotide homology. Divergent amino acids are underlined, nucleotide changes are indicated by'O. Numbering of nucleotides and amino acids follow reference 128. The sequence of the human c-src exon 2 was determined by A. Tanaka. human chick human chick human chick human chick human chick human chick 72 GACCATGGGTAGCAACAAGAGCAAGCCCAAGGATéCCAGC MetGlySerAsnLysSerLysProLysASpAlaSer CCCACCACCATGGGGAGCAGCAAGAGCAAGCCCAAGGACCCCAGC MecGlySeréggtysSerLys?roLysA5p3£35er O O I. O O 0.. O. O O CAGCGGCGCCGCAGCCTGGAGCCCGCCGAGAACGTGCACGGCSCT GlnArgArgArgSerLeuGluProAlaGluAana1H1sGlyAla CAGCGCCGGCGCAGCCTGGAGCCACCCGACAGCACCCAC CAC GlnArgArgArgSerLeuGluProProAspSerThrHis His 0. O O O O O O O O. GCGGGGCGCTTTCCCCGCGTCGCAGACCCCCAGCAAGCCAGC C AlaGIyArgPheProArgValAlaAsPProGInGInAlaSe: L 6666 GATT CCCAGCCTCGCAGACCCCCAACAAGACAGCAGC GlyG lyPh eProAlaSerGlnThrFroAsnLysThrAlan: O I .0. O. O. O. O. O. TCGCGACGGCCACCGCGGCCG AGCCGC TT GCCCCGTGGC G o euAlaThrAlaThrAlaAla luPr P roC ysFrovalAl CCCCGACACGCACCGCACCCCCAGCCGCTCCTTTGGGACCGTGGC aProASpThrHisArgThrFroSerArgSerPheGlyThrVa1A1 CECCGAGCCCAAGCTETTCGGAGGCTTeAACTCETCEGACACCGT aProGIuProLysLeuPheGIyGlyCysAsnSerSerASpThrVa CACCGAGCCCAAGCTCTTCGGGGGCTTCAACACTTCTGACACCGT aThrGIuProLysLeuPheGlyGlxgsgAanQESerASPThrVa O O O O O O. .0. O CACCTCCCCGCAGAGGGGGGGGCGGCTGGCCG lThrSerProGlnArgGlyGlyArgLerAla TACGTCGCCGCAGCGTGCCGGGGCACTGGCTG lThrSerProGlnArgélEGIyAlaLeuAla 1“») I.) 0‘ 209 241 73 closely homologous to the corresponding chicken c-src and SR-A v-src sequences, but in the central portion, the human c-src sequence has shifted, due to insertions and deletions. Nevertheless, the v-src initiation site has been conserved and the length of the coding sequence in exon 2 is the same in both the human and chicken loci. A comparison of the homology between the three src genes is presented in Table I. For the majority of exons, the nucleotide homology is fairly high, however, the homology within exon 5 is somewhat lower (76.9% between the human and chicken c-src genes), and exon 2 is quite divergent, as a result of the rearrangement. The deduced amino acid sequences (with the exception of exon 2) demonstrate a striking conservation. Excluding exon 2, the overall amino acid homology between human and chicken c-src is 97.1%. Similarly, amino acid homology between human c-src and v-src is quite high, 93.0%: nearly half of the different residues fall within the divergent carboxy terminus. The sequence analysis has also demonstrated that the chicken c-src exon structure is conserved in the human c-src locus. The intron-exon boundaries are identical in the two genes (Table 2), however, the intron sequences are completely divergent and the intron sizes are not conserved. One difference can be noted at the 5' boundary of exon 2, where the chicken c-src exon is 5 base pairs longer than the corresponding human c-src exon. This difference is located in the non-coding region upstream from the initiation codon. Finally, all of the human c-src splice sites are consistent with the AG/GT consensus rule. Table 1. 74 Homology between the human and chicken c-src and SR-A v-src genes. Nucleotide and amino acid homology are calculated for each exon. with the homology at the 3' terminus of exon 12 calculated separately. Overall homology was calculated from exons 4-12, and does not include exon 2. 75 Am_e\mmmv so.ma Akmmp\mmo_v a_.mw AN_\FV em.m AaM\oPV am.NN Aae\eev sw.mm Aae_\mm~v em.am Ame\mev eo.oo_ ANm_\N_PV aa.mm A_m\omv ao.ma Aem.\mmpv ae.mm Amm\emv so.aa ANN\mav sw..m Aam\va aa.aa aomp\om_v $5.0m Apm\mev &_.am Aomp\wm_v am.wm Ame\mev ao.ma Aom_\Fapv no.4a AeM\~mV s_.¢a A¢o_\omv aa.oe ANM\aNV aa.oa Aaa\mmv am.mw Aom\mav aw.mm A_am\mmpv em.oe 20m oEEm 02320:: o..m-> <45 Aa_e\Koav a_.ka AmeP\NQFFV aa.km Aap\a_v so.oo_ Anm\_mv am.aw Aae\kav aa.ma Ama_\ampv sm._a Ama\mav so.oop Amm_\mp_v sa.mm apm\omv ao.ma Aem_\mmpv s~.mm Amm\ewv eo.aa Akk\aav e_.ma Aam\wmv sm.ma Aomp\mmpv sm.nw A_m\mev &_.ea Aamp\mmpv am.mm Aaa\mev so.ma Aom_\,e_v ao.ea AeM\va N_.ea Aaop\owv am.e~ Amm\~mv sa.oa Aaa\mmv ea.mm Aom\eav so.mm AFeN\mm_v ao.m~ Rom 2.25 25020:: 05-0 zm¥0=._0 AN_-avpeoao Ammm-m_mvm_ Ae_m-oaevm_ PF op COXG 76 Table 2. Comparison of the intron-exon boundaries in the human and chicken c-src loci. Nucleotide sequences surrounding the putative Splice sites are given. 77 O .. Molecular weight sizes are in kb. 124 125 lymphoblastoid cell line derived from a splenic lymphoma of an MDV-infected chicken (2). and contains an estimated 15-90 copies of the MDV genome per cell (22,40). The library was screened with two probes: the first was nick-translated MDV viral DNA and the second consisted of a pool of 8 MDV-positive plasmid clones. A total of 14 plaques were detected with the MDV viral DNA probe, 7 of these also hybridized to the plasmid pool. Preliminary restriction enzyme maps of 7 of the clones have been constructed (Figure 5). Initial restriction mapping suggests only two of the inserts (2 and 10) overlap extensively; the analyses are not yet complete enough for other overlaps to be identified. However, hybridization of individual clones to DNA from MDV-infected cells suggests additional overlaps do exist (Y. Naidu, personal communication). It is estimated that 50% of the MDV genome is represented in these clones. The low nunber of clones obtained is a direct result of the inherent difficulties involved when using total viral DNA as a probe. The complexity of the viral DNA (180kb) combined with the large volume required during hybridization resulted in unfavorable hybridization kinetics: hybridizations could not be carried out to completion. Thus, the signals obtained with the total MDV viral DNA probe were extremely weak. However, rescreening with other plasmid clones or with Specific fragments from the existing lambda inserts should produce additional phage clones containing the remainder of the MDV genome. 4. Cross-hybridization utilizing MDV phage clones Cross-hybridization of the MDV lambda clones to HVT viral DNA was examined by Southern blot analysis of the phage DNAS. Figure 6A shows 126 Fig. 5 Restriction enz aimaps of MDV phage clones. Symbols are as folldws: ofiBamHI; x, BglI; o, EcoRI; v, HindIII; 0, KpnI; A. SaeI; o,‘ SalI. i 9:. l27 . .. e S a a a“ e 2 a fin. a a a. 2 S a A u a a. a m z. a A z a s a a. .— ~ hm 128 Fig. 6 Southern hybridization analysis of MDV phage clones. Phage DNA containing MDV inserts were digested with EcoRI and electrophoresed through 0.7% agarose. A. Ethidum bromide staining pattern. 8. Hybridization to MDV viral DNA. C. Hybridization to HVT viral DNA under relaxed conditions. The lambda arms are indicated by <. Molecular weights sizes are in kb. 129 130 the ethidium bromide staining pattern of EcoRI digests of 8 of the phage clones; the two largest bands in each lane contain the lambda arms. All fragments derived from the inserts hybridized to the MDV viral DNA probe (Figure 68), while the lambda arms did not. This hybridization confirms the inserts are MDV-Specific. Under the low stringency conditions. a duplicate blot was hybridized the HVT viral DNA probe (Figue 6C). In agreement with the results obtained with the plasmid clones, the majority of insert fragments cross-hybridized with the HVT probe. Again, relative intensity differences among the fragments suggests homology is more extensive in certain regions. DISCUSSION A herpesvirus-induced malignant lymphoma of chickens, MD is effectively prevented by vaccination with the non-oncogenic virus, HVT. Antigenically, the two viruses appear to be closely related. Over 40 MDV-specific polypeptides can be detected in productively infected cells; the majority of these proteins can be immunoprecipitated with antisera raised against HVT (18.23.24.64). In addition, a number of monoclonal antibodies developed against various MDV antigens will cross-react with HVT proteins (25,55). Despite this close immunological relationship between the two viruses, cross-hybridization analyses suggested the viral genomes Share only 1-5% homology (19.26.31.50). This conflict has been addressed in these studies. The difficulties involved in isolation of viral DNA have proven to be a major deterrent in analysis of the MDV and HVT genomes, thus the degree of homology between the viral DNAS has not been examined in 131 detail. To avoid the problems inherent with the use of purified viral DNA, two MDV clone banks have been constructed, the first in a plasmid vector and the second in a lambda phage vector. Altogether, it is estimated over 90% of the MDV genome is represented in those two libraries. These cloned fragments have greatly facilitated cross-hybridization analyses and should be extremely useful for a variety of other studies. The homology between the MDV and HVT genomes has been reexamined using highly sensitive hybridization conditions: contrary to earlier reports, results indicate the two viruses share extensive homology. Under conditions allowing 30% base mismatch, virtually all of the HVT viral DNA restriction fragments can be detected with total MDV viral DNA as probe (Figure 1). The converse is also true: the majority of MDV restriction fragments hybridize to the HVT viral DNA probe (Figure 2). This extensive cross-hybridization was confirmed using cloned MDV fragments. 90% of the MDV fragments in the plasmid library hybridized to the HVT probe (Figure 4). Likewise, the majority of inserts in the MDV phage library could be detected with the HVT probe (Figure 6). This data unequivocally demonstrate the two viruses share extensive homology. The lack of cross-hybridization between MDV and FRV, an antigenically unrelated herpesvirus with a GC content identical to that of MDV and HVT, clearly indicates the hybridization conditions are highly specific. Since no cross-hybridization between MDV and FRV could be detected under the low stringency conditions, it is unlikely the bands observed in the MDV/HVT cross-hybridizations are an artifactual result of non-specific trapping of probe which might occur under relaxed conditions. In addition, viral DNAS did not hybridize to DEF DNA, nor to the lambda or plasmid vector DNAS. 132 Under conditions allowing 30% base mismatch, homology extending over 90-95% of the MDV genome could readily be detected with the HVT probe. These MDV/HVT hybrids are less stable under conditions of 20% base mismatch. and most have dissociated under conditions allowing only 10% base mismatch. Thus a 20-30% sequence divergence between the two viral genomes is indicated, with homologous sequences extending throughout the MDV and HVT genomes. These results readily explain the lack of homology reported by other investigators (19.26.31.50). Previous cross-hybridizations had been conducted under conditions approximating 15-20% base mismatch, and as shown here, most MDV/HVT hybrids are unstable below 20% base mismatch, although readily detected under less stringent conditions. In addition, it is likely hybridizations conducted by earlier investigators were not carried out to completion. Due to the complexity of the viral genomes (150-180kb), the DNAS follow relatively slow reassociation kinetics; under conventional hybridization conditions a lengthy incubation period is usually required to complete such a reaction (36). To overcome this problem, in these experiments dextran sulfate was included in all hybridization buffers when total viral DNAS were used as probe. This compound has been shown to increase the rate of reassociation ten-fold (65), thus allowing hybridizations to proceed to completion within 2-3 days. Direct comparison of the hybridization conditions used here with those used previously (31) demonstrates the striking difference in sensitivity between the two methods (Figure 3, B and C). Thus the quite different conclusion reached by earlier investigators (19.26.31.50) can be accounted for by the sensitivities of the 133 conditions used. Of primary importance is the lowered stringency used here; improved reassociation kinetics probably play a minor role. While these studies were in progress, two other groups initiated projects similar to those presented here (20,59); their results confirm the extensive homology described in this report. Substantial genetic homology is consistent with the strong imnunological relationship between MDV and HVT. Homologous sequences located throughout most of the viral genomes could easily accomodate the number of virus-specific polypeptides in MDV- or HVT-infected cells which can be immunoprecipitated with heterologous antisera (18.23.24.64). The estimated degree of mismatch, 20-30%. is sufficient to explain the minor differences observed between a number of the corresponding viral proteins (18.61), as well as the observation that certain antigens do not cross-react as strongly as others (23) and monoclonal antibodies have been obtained that are specific for either MDV or HVT (25,55). This level of homology also explains the different restriction enzyme patterns of the two viruses (19,26). Finally, the presence of regions that do not cross-hybridize suggests unique or highly divergent proteins may also exist, again consistent with the finding that a few viral-specific polypeptides do not cross-react with the heterologous antisera (23). Thus the previous conflicting data obtained from protein and DNA analyses of MDV and HVT have been clarified. The estimate of 70-80% sequence homology extending throughout 90-95% of the viral genomes is in complete agreement with the strong antigenic similarities between the viruses, yet would not have been detected under stringent hybridization conditions. These results suggest the efficacy of the 134 HVT vaccine could involve a number of viral antigens, although certain proteins are much more abundant and may play the major role in eliciting a protective immune response. SUMMARY (part 8) MD is the only neoplastic disease which can be effectively controlled by routine vaccination; the comonly used HVT vaccine elicits a strong immune response that protects chickens against MDV-induced lymphoma. Elucidation of the mechanism of this protection will aid in the understanding of the nature of defense mechanisms against viral-induced malignancies. Such information will be valuable for the deveIOpment of vaccines against other viral-induced neoplasias as well as vaccines effective against other herpesviruses. Despite this potential as a model system, little progress has been made toward understanding the nature of protection afforded by the HVT vaccine, primarily because of a number of inherent difficulties within the MDV system. Results presented here offer significant technical and conceptual contributions to the field. The availability of MDV genomic libraries, both the phage and plasmid clone banks, provides a source for obtaining unlimited amounts of viral DNA, thereby circumventing the problems encountered during purification of total viral DNA. which have hampered analyses of the MDV genome until now. These genomic clones will greatly facilitate the construction of both physical and genetic maps, and in conjunction with protein analyses Should provide a foundation for detailed examination of antigens involved in the protective immune response as well as those involved in 135 136 transformation. Such analyses have already been initiated. One plasmid clone, M, is capable of selecting message for A antigen (R. Isfort, personal communication), and therefore the insert contains part or all of the A antigen coding sequence. The finding of substantial homology between MDV and HVT DNA, 70-80% homology extending over 90-95% of the viral genomes, suggests the viruses share a number of antigenetically related, although not identical proteins. These results are in complete agreement with the close immunological relationship exhibited by the two viruses and clarify earlier reports that seemed to indicate MDV and HVT Share little genetic homology. Immunity to MDV is likely to involve a number of viral antigens, and it is probable that protection entails responses directed against the various stages of viral infection. LIST OF REFERENCES (part B) 137 11. 12. 13. 14. 15. 16. 17. 18. LIST OF REFERENCES Ahmed, M. and Schidlovsky, G. (1972) Cancer Res. 32:187-192.' Akiyama, Y., and Kato, S. (1974) Biken J. 17:105-116. Burgoyne, G.H. and Witter, R.L. (1973) Avian Dis. 15:662-671. Burgoyne, G.H. and Witter, R.L. (1973) Avian Dis. 17:824-837. Calnek, 8.W. (1972) Infect. Immun. 6:193-198. Calnek, B.W., Murthy, K.K. and Schat, K.A. (1978) Int. J. Cancer 21:100-107. Cebrian, J., Kaschka-Dierich, C., Berthelot, N. and Sheldrick, P. (1982) Proc. Natl. Acad. Sci. 79:555-558. Chen, J.H., Lee, L.F., Nazerian, K. and Burmester, B.R. (I972) Virology 47:434-443. Chen, J.H. and Purchase, H.G. (1970) Virology 40:410-412. Chubb, R.C. and Churchill, A.E. (1968) Vet. Rev. 83:1-7. Churchill, A.E. and Biggs, P.M. (1967) Nature 215:528-530. Churchill, A.E., Chubb, R.C. and Baxendale, W. (1969) J. Gen. Cole, R.K. (1968) Avian Dis. 12:9-28. Edison, C.S., Ellis, M.N. and Kleven, S.H. (1981) Poult. Sci. 60:317-322. Edison, C.S., Page, R.K. and Kleven, S.H. (1978) Avian Dis. 22:583-597. Else, R.W. (1974) Vet. Rec. 95:182-187. Fukuchi, K., Sudo, M., Lee, Y.-S., Tanaka, A. and Nonoyama, M. (1984) in press. Glaubiger, C., Nazerian, K. and Velicer, L.F. (1983) J. Virol. 45:1228-1234. 138 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 139 Hirai, K., Ikuta, K. and Kato, S. (1979) J. Gen. Virol. 45: 119-131. Hirai, K., Ikuta, K., Maotami, K. and Kato, S. (1984) J. Biochem. in press. Howley, P.M., Israel, M.A., Law, M.-F. and Martin, M.A. (1979) J. Biol. Chem. 254:4876-4883. Huges, S., Subblefield, E., Nazerian, K. and Varmus, H.E. (1980) Virology 105:234-240. Ikuta, K., Nishi, Y., Kato, S. and Hirai, K. (1981) Virology 114:277-281. Ikuta, K., Ueda, S., Kato, S. and Hirai, K. (1983) J. Gen. Virol. 64:961-965. Ikuta, K., Ueda, S., Kato, S. and Hirai, K. (1983) J. Gen. Virol. 64:2597-2610. Kaschka-Dierich, C., Bornkamm, G.H. and Thomssen, R. (1979) Med. Microbiol. Immunol. 165:223-239. Kaschka-Dierich, C. Nazerian, K. and Thomssen, R. (1979) J. Gen. Virol. 44:271-281. Lau, R.Y. and Nonoyma, M. (1980) J. Virol. 33:912-914. Lee, L.F., Nazerian, K., Leinbach, S.S., Reno, J.M. and Boezi, J.A. (1976) J. Natl. Cancer Inst. 56:823-827. Lee, Y.S., Tanaka, A. and Nonoyama, M. (1982) Gene 19:185-190. Lee, Y.S, Tanka, A., Silver, 5., Smith, M. and Nonoyma, M. (1979) Virology 93:277-280. Long, P.A., Clark, J.L. and Velicer, L.F. (1975) J. Virol. 15:1192-1201. Long, P.A., Kaveh-Yamini, P. and Velicer, L.F. (1975) J.Virol. 15:1182-1191. Longenecker, B.M. and Gallatin, W.M. (1980) In: Resistance and Immunit tg_Marek's Disease, ed. Biggs, P.M. (Commission of the European COmmunity, Luxembourg). Longenecker, B.M., Pazderka, F, Law, G.R.J. and Ruth, R.F. (1973) Fed. Proc. 32:966. Maniatis, T., Fritsch, E. and Sambrook, J. (1982) Molecular Cloning: A_Laborato:yManual (Cold Spring HarBor, NY). 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 140 Marek, J. (1907) Dtsch. Tieraerztl. HOchenschr. 15:417-421. Nazerian, K. (1980) In: Virol Oncology, ed. Klein, G. (Raven press, NY) pp. 665-682. Nazerian, K. and Chen, J.H. (1973) Arch. Gesamte Virustorsch. 41:59-65. Nazerian, K. and Lee, L.F. (1974) J. Gen. Virol. 25:317-321. Nazerian, K. and Lee, L.F. (1976) Virology 74:188-193. Nazerian, K., Payne, N., Lee, L.F. and Hitter, R.L. (1978) In: Proc. A.S.M. p. 274. Nazerian, K., Solomon, J.J., Hitter, R.L. and Burmester, B.R. (1968) Proc. Soc. Exp. Biol. Med. 127:177-182. Okazaki, N., Purchase, H.G. and Burmester, B.R. (1970) Avian Dis. 14:413-429. Powell, P.C., Payne, L.N., Frazier, J.A. and Rennie, M. (1974) Nature 251:79-80. Purchase, H.G. (1969) J. Virol. 5:79-90. Rispens, B.H., VanVloten, H., Maas, H.J.L. and Hendrick, J.L. (1972) Avian Dis. 16:108-125. Ross, L.J.N. (1977) Nature 268:644-646. Ross, L.J.N., DeLorbe, w, Varmus, H.E., Bishop, J.M., Brahic, M. and Haase, A. (1981) J. Gen. Virol. 57:285-296. Ross, L.J.N., Milne, B. and Biggs, P.M. (1983) J. Gen. Virol. 64:2785-2790. Rziha, H.-J. and Bauer, B. (1982) Arch. Virol. 72:211-216. Schat, K.A. Calnek, B.M. and Fabricant, J. (1982) Avain Path. 11:593-606. Sharma, J.M. and Coulson, 8.0. (1977) J. Natl. Cancer Inst. 58:1647-1651. Sharma, J.M., Hitter, R.L. and Coulson, 8.0. (1978) J. Natl. Cancer Inst. 61:1273-1280. Silva, R.F. and Lee, L.F. (1984) in press. Silver, S., Tanaka, A. and Nonyama, M. (1979) Virology. 93:127-133. Smith, M.H. and Calnek, B.M. (1973) Avian Dis. 17:727-736. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 141 Spencer, J.L. and Calnek, 8.". (1970) Avian Dis. 11:274-287. Suto, M., Donovan, J., Eklund, L., Jessip, J., Fukuchi, K., Tanaka, A. and Nonoyama, M. (1984) in press. Tanaka, A., Joel, M., Silver, S. and Nonoyama, M. (1978) Virology 8:19-24. VanZaane, D., Brinkhoff, J.M.A. and Gielkens, A.L.J. (1982) Virology 121:133-146. VanZaane, D., Brinkhoff, J.M.A., Hestenbrink, F. and Gielkens, A.L.J. (1982) Virology 121:116-132. VanZaane, D. and Gielkens, A.L.J. (1980) In: Resistance and Imunit _i_:_o Marek's Disease, ed. Biggs, P.M., (Comission of the Europeaanmmunity,Luxembourg) P. 62. Velicer, L.F., Yager, D.R. and Clark, J.L. (1978) J. Virol. 27:205-217. Nahl, G.M., Stern, M. and Stark, G.R. (1979) Proc. Natl. Acad. Sci. 76:3683-3687. Nildy, P., Field, H.J. and Nash, A.A. (1982) In: The Society of General Microbiology, Symposium 33: Viral Persistence, eds. Mahy,*B.w., Minson, A.C. and DarEy, G.K. (Cambridge University Press, UK) pp 136-137. Hitter, R.L. (1983) Avian Dis. 27:113-132. Hitter, R.L., Nazerian, K., Purchase, H.G. and Burgoyne, G.H. (1970) AM. J. Vet. Res. 31:525-538. Witter, R.L., Sharma, J.M. and Fadly, A.M. (1980) Avian Dis. 24:210-232. Hitter, R.L., Stephens, E.A., Sharma, J.M. and Nazerian, K. (1975) Jo Immun010 115:177'1830 Zander, D.V., Hill, R.N., Raymond, R.G., Balch, R.K., Mitchell, R.N. and Dunsing, J.H. (1972) Avian Dis. 16:163-178. 4. GENERAL REFERENCES Calnek, B.R. and Hitter, R.L. (1978) In: Diseases 2: Poultry 6:385-418. Nazerian, K. (1979) Biochimica et Biophysica Acta. 560:375-395. Nazerian, K. and Kaschka-Dierich, C. (1980) In: Cold S rin Harbor Conference on Cell Proliferation, Vol. 7. (Cold Spr ng Harbor Laboraotry) pp. 171-183. Nonoyama, M. (1982) In: The Herpesviruses, Vol. 1 142 "liliillllilillllllll“