Genomes of endogenous and exogenous avian retroviruses

VIROLOGY

126, 51-72 (1983)

Genomes

of Endogenous

JOHN M. COFFIN,‘,* ALAYNE

and Exogenous

Avian Retroviruses

PHILIP N. TSICHLIS,‘,* KATHLEEN F. CONKLIN,3** SENIOR,? AND HARRIET L. ROBINSON?

*Department of Molecular Biology and Microbiology and Cancer Research Center, Tufts University School of Medicine, 136 Harrison Avenue, Boston, Massachusetts 02111, and t Worcester Foundation for Experimental Biology, 222 Maple Avenue, Shrewsbuq, Massachusetts 01545 Received September 15, 1982; accepted November

18, 1982

The endogenous viruses of chickens are closely related to the exogenous avian leukosis viruses (ALV) yet as a group differ from these viruses in their host range, growth rate, and oncogenicity. The present study was undertaken to determine the patterns of relationship among the genomes of endogenous and exogenous ALVs. Complete or partial T, oligonucleotide maps were prepared from the genomes of endogenous viruses that reside at eight distinct loci in chickens. Selected endogenous viruses and recombinants of endogenous or endogenous and exogenous viruses were characterized for host range and growth rate. From these data we could infer the following: (1) Endogenous viruses form a distinct lineage of ALVs with the most distinctive differences occurring in the portion of env that encodes host range and the Us portion of the long terminal repeat; (2) The Ua sequences of endogenous ALVs determine the low growth rates of these viruses; and (3) Endogenous ALVs have distinctive oligonucleotide markers that allow them to be subclassified into distinct lineages. Our results suggest that endogenous viruses are derived from one another and not from exogenous field strains of ALV. This phenomenon may be related to the unique env encoded host range of endogenous ALVs, their unique U, encoded growth rates, or perhaps their unique access, as residents of germ line DNA, to germ line cells.

ruses share the same structure; i.e., LTRgag-pal-env-LTR, where gag, pal, and env encode virion proteins and the long terminal repeat (LTR) is formed during viral DNA synthesis from unique sequences near the 5’ (U,) and 3’ (U,) ends of the genome, flanking a repeated sequence (R) found at both ends (Hughes et al., 1978; Shank et al., 1978; Hsu et al., 1978; Coffin, 1979; for review see Varmus and Swanstrom, 1982). Proviruses at individual loci confer distinctive phenotypes in the cells that contain them. Such phenotypes include inducible expression of infectious virus (ev2, 10, 11, and 12); high level expression of biologically active virion envelope glycoprotein (chf), in the presence (ev-3) or absence (ev-6, 9) of viral gag and pol gene products (gs antigens) (Hanafusa et al., 1974; Astrin et al., 1979a, 1980a; Robinson et al., 1981); inducible expression of noninfectious virions (ev-1,7) (Robinson et al., 1979a; Groudine et al., 1981; Conklin et al.,

INTRODUCTION

The genomes of many species contain inherited endogenous proviral DNA sequences, often closely related to retroviruses that are endemic to the same or another species (for reviews see Aaronson and Stephenson, 19’78, Robinson, 1978; Pincus, 1980; Weinberg, 1980; Coffin, 1982). The best studied group of endogenous proviruses is found in white leghorn chickens. Proviruses have been found at at least 13 distinct loci designated ev loci, and numbered 1 through 13 (Astrin, 1978; Astrin et al., 1980a; Hughes et al., 1981). Avian leukosis viruses and endogenous provi’ To whom requests for reprints should be sent. a Present address: Laboratory of Tumor Virus Genetics, National Cancer Institute, National Institutes of Health, Bethesda, Md. 20205. “Present address: Fred Hutchinson Cancer Research Center, Division of Oncology, 1124 Columbia St., Seattle, Wash. 98104. 51

004%6822/83 $3.00 Copsrirht All rights

Z; 1983 hy Academic Press, Inc of reproduction in any form reserved.

52

COFFIN

1982); and no detectable expression of virus-related protein or RNA (ev-4, 5, and 8) (Astrin et al., 1980a; Hayward et al., 1980). The phenotypes of many defective and/or noninducible endogenous viruses (including ev-3, ev-4, ev-5, and ev-6) appear to be determined by deletions that often include the left LTR. Two (ev-1 and ev-7) encode noninfectious virions in the absence of apparent deletions (Conklin et al., 1982; Skalka et al., 1980; Robinson et al., 1979a; Hughes et al., 1981). Replication-competent endogenous viruses, such as RAV-0 (the product of ev2), differ from exogenous viruses in several respects. First, endogenous viruses do not grow as rapidly in culture or in animals as exogenous viruses (Hanafusa et al., 1975; Robinson, 1976; Linial and Neiman, 1976; Robinson et al., 1980). This difference is correlated with reduced levels of viral RNA in infected cells (Wang et al., 1977) and has been mapped to the different U3 regions of the viral genomes (Tsichlis and Coffin, 1980a, b). Second, endogenous viruses have a subgroup E host range, distinct from those of exogenous viruses (Weiss, 1969; Vogt and Friss, 1971) encoded by the gp85 portion of env (Tsichlis and Coffin, 1980a; Tsichlis et al., 1980). Third, endogenous viruses are nonpathogenic in chickens (Motta et al., 1975; Robinson et al., 1982) although a preliminary report has suggested that related species of fowl may be susceptible to pathogenesis by RAV-0 (Weiss and Frisby, 1982). By contrast, all known exogenous viruses induce one or more of a variety of diseases, such as B-cell lymphoma, or osteopetrosis, typically with a slow onset (for reviews- see Teich et al., 1982; Robinson, 1982). Correlation of differences in pathogenicity with genomic differences has revealed that the II3 region is a major determinant; however, other regions of the genome must also be important in determining the specificity of disease induced (Robinson et al., 1982). In a previous study (Coffin et al., 1978a), we used a combination of nucleic acid hybridization and oligonucleotide mapping to localize closely related and distantly related regions on the genomes of RAV-0 and an exogenous virus (Pr-RSV-B). This anal-

ET AL

ysis made it clear that these viruses are closely related and must have diverged from a common ancestor, with the greatest conservation being found in pol, portions of gag, and a portion of the gp85 and gp37 coding regions of env. More distant, but detectable sequence relationship was found in the subgroup-encoding region of env and the remainder of gag. No relationship could be detected by this analysis in U3 or a region extending from the 3’ end of env to Ua. Nucleotide sequence analysis (Hishinuma et al., 1981; Hughes, 1982) has shown short patches of homology in these regions interspersed with unrelated sequence. The present study was undertaken to extend these findings to a much wider sample of endogenous and exogenous viruses. We present here the results of oligonucleotide mapping of the products of a number of endogenous viruses derived from different loci, as well as some exogenous viruses. From this analysis, we conclude that the two groups are derived from a common ancestor yet form distinct lineages, and that the endogenous viruses, despite long residence at distinct chromosomal sites, are very closely related to one another, varying by less than 2% of random base changes. Furthermore, all endogenous viruses examined contain U3 regions closely related to that of RAV-0, and were found to have growth rates substantially lower than recombinants which had acquired U3 regions from several exogenous viruses. MATERIALS

AND

METHODS

Cells and Viruses The chicken cells used in this study were prepared from fertilized eggs of various chicken lines shown in Table 1 and propagated as described (Conklin et al., 1982). Turkey (T/BD) cells were prepared from fertilized Orlopp turkey eggs purchased from Orlopp Enterprises, Arosi, California. All cells were stored frozen in liquid nitrogen and were thawed and used as needed. All of the replication competent endogenous viruses were isolated from cells that had been induced with BrdU for virus production (Robinson et al., 1976).

ev-1, -7 ev-I, -7 ev-1, -3, -7 ev-1, -3, -6, -7, -8 ev-1, -7, -9

RPRL RPRL HRob HRob

15B(C/C)

15B(C/C) 15B X KlG(C/E)

15B X KlG(C/E)

none

” The pattern of resistance of each embryo to infection by ALV of different subgroups is indicated *RPRL; Regional Poultry Research Laboratory, East Lansing, Mich. (courtesy of L. Crittenden); Worcester Foundation for Experimental Biology, Shrewsbury, Mass. ‘H. Robinson and S. Astrin, unpublished.

ev-1

Orlopp HRob

(15B X K(-)

1384

x KZS)(C/O)

ev-1, -7, -9

HRob HRob

15B X K18(C/E)

15B X KlS(C/E) Orlopp Turkey (T/BD) Enterprises

ev-1, -7

RPRL

lBB(C/C)

in parentheses. HRob, from flocks maintained

none

none

15B X K18-E-2 15B X K18-E-1

15B X K16-E-2 15B X K16-E-1

15B-E-3 15B-E-11, 1 15B-E-12j

of ev-1,

of ev-1,

by H. Robinson

at the

of ev-1, ev-7, and ev-9

recombinants

ev-3, and ev-7

recombinants

and ev-7

recombinants

noninfectious product of ev-7

of ev-12

product

of ev-10 of ev-12

product product

ev-7-ILV

of ev-10

product

ev-12, 15,-ILV

RPRL RPRL

ev-10, 151,-ILV ev-12 15,-ILV

of ev-2 of ev-11

of virus

product product

Origin

ev-1, -7

ev-1, -12

RPRL

ev-11, 15II-ILV ev-10, C-ILV

ev-2 II*-ILV(RAV-0)

Induced virus

ev-1, -7, -12

m-1, -10

RPRL

(15B x 151) x 15B(C/O) (15B x 15i) x 15B(C/O) lBB(C/C)

w-1, -2 ev-1, -11 w-1, -10

ev proviruses’

(72 x 15IJ x 15B(C/O)

RPRL Reaseheath

RPRL

Source *

1188 11278 T109

1263 i 1264

1

1087 8715 I 8694

21907 922

16436 22245

(7* X 15IJ X 15B(C/O)

18154 983

C(C/AE)

7dCW

Pedigree”

1072

Embryo

1

CELLS AND VIRUSES

TABLE

Y

z E

22

K c

E

!z -~

5 m

2 ul

i2

8

54

COFFIN

The isolation of NTRE-7 has been described previously (Tsichlis and Coffin, 1980a). The RAV-60 isolates were originally obtained as a gift from Dr. H. Hanafusa (Rettenmeir and Hanafusa, 1977) and have been previously characterized (Coffin et al., 1978b; Robinson et al., 1980; Crittenden et al., 1980). Brd U Induction Virus production was induced from cell cultures by growth for one cell doubling in 0.1 mM bromodeoxyuridine (BrdU) and 0.01 mM deoxycytidine in Dulbecco’s modified Eagle’s medium supplemented with 5% calf serum. Following treatment, cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% tryptose phosphate broth and 5% calf serum. Stocks of induced virus were harvested 48 and 72 hr after the termination of BrdU treatment. Tl Oligonucleotide Fingerpkting Genome RNA was analyzed as described (Coffin et al., 1978a). Briefly, virus-infected cells grown in loo-mm culture dishes were labeled with 5 mCi 32P per dish and the radiolabeled virus was pelleted. The virus RNA was purified by sedimentation in 523% sucrose gradients. The purified 70 S genome RNA was digested with RNase Ti and the digestion products were separated by two-dimensional polyacrylamide gel electrophoresis. Some fingerprints were obtained using a tris-borate buffer in the second dimension (Pederson and Haseltine, 1980) rather than tris-citrate as previously described (Coffin et al., 1978a; Tsichlis and Coffin, 1980a). Fingerprints of the 3’ proximal sequences of viral RNA were obtained as follows. 32P-labeled 70 S viral RNA was hydrolyzed with 0.1 M Na&03 at 50’ for 15 min. The poly(A)-containing fragments were isolated in a poly(U) Sephadex column and were digested with RNase T1 in the same way as 70 S viral RNA. The digestion products were separated in twodimensional polyacrylamide gels with the second dimension run only 2/3 as far as

ET AL

with the fingerprints of 70 S RNA in order to retain smaller oligonucleotides. The oligonucleotide maps shown in Fig. 3 were derived from our previous mapping data (Coffin et al., 1978a, b) and, wherever possible, from nucleotide sequence information (D. Schwartz, R. Tizard, and W. Gilbert, personal communication; Hughes, 1982; Hishinuma et al., 1981). A complete argument for the assignment of specific oligonucleotides to their location in the genome is available from the authors on request. Virus Titrations The titer of virus in infectious units/ml was determined by endpoint dilutions of stocks on (15B X K(-)) X K28 cells that are permissive for growth of subgroup E viruses (Robinson, 1976; Robinson et al., 1979b). The amount of particulate RNAdirected DNA polymerase in stocks was determined using the template primer poly(rC) . oligo(dG) (Robinson, 1976). The ratio of infectious units to particulate RNA-directed DNA polymerase for nine different stocks of endogenous (U3”), exogenous (U,‘), and recombinant (U,‘) viruses was similar for all subgroup E viruses of similar passage history. Since all subgroup E viruses had comparable ratios of infectious units/particulate RNA-directed DNA polymerase, levels of particulate reverse transcriptase were used to estimate titers of infectious virus. RESULTS

Genomes of Nondefective Endogenous Vimses Endogenous proviruses at four loci (ev2, -10, -11, -12) are associated with low levels of spontaneous production of infectious virus (Astrin, 1978; Astrin et al., 1980b; Smith and Crittenden, 1981). Production of virus by all of these proviruses is increased approximately loo-fold by growth of cells in bromodeoxyuridine (Robinson et al., 1976). To characterize and compare the viruses encoded by ev-2, -10, -11, and -12, viruses were induced from cells that contained each of these loci. Induced viruses

ENDOGENOUS

AND

EXOGENOUS

ALV

GENOMES

55

FIG. 1. Genomes of viruses from cells with inducible nondefective a proviruses. Virus was induced with BrdU from line 7 embryo 1072 (m-2) (A); (72 X 1514) X 15B embryo 18154 (ev-11) (B); Reaseheath line C embryo 983 (ev-10) (C); (72 X 1514) X 15B embryo 16436 (~-10) (D); (15B X l&) X 15B embryo 22245 (ev-12) (E); and (15B X 15,) X 15B embryo 21907 (ev-12) (F). T1 oligonucleotide fingerprints were prepared from total genome RNA. The numbering of oligonucleotides in A conforms to our previous convention (Coffin et al., 1978a; Tsichlis et al., 1980). Only oligonucleotides not present in ev-2 virus (i.e., RAV-0) are numbered in B-F. Dashes indicate ev-2 oligonucleotides absent from the other viruses. Note that in this and other figure legends, only the ev proviruses contributing to the virus analyzed are indicated. For a complete list, see Table 1.

were grown to and analyzed fingerprinting amination of

high titer on permissive cells for genetic composition by genomic RNA (Fig. 1). Exthese fingerprints revealed

several points. First, all oligonucleotides were present in equimolar yield, indicating that each genome was the product of a single provirus. No contribution from ev-

FIG. 2. UB regions of the virus genomes. Short, poly(A)-containing fragments from the genome of Pr-RSV-B(A), RAV-0 (from line 100 cells (en-2) identical to the virus in Fig. 1A) (B); virus induced from embryo 983 (w-10, Fig. 1C); and infectious virus found after induction of cells from embryo 1087 (ev-1 X eu7; Fig. 5B) were analyzed as described (Coffin et ul., 1978a). All oligonucleotides shown, with the exception of 03, are within Ua. The numbering conforms to our orevious convention (Tsichlis and Coffin. 1980a).

ENDOGENOUS

AND

EXOGENOUS

1 which was also present in all the cells could be detected (see below). Second, the genome of virus induced from ev-2 cells (Fig. 1A) had a fingerprint identical to that previously described for RAV-0 which is spontaneously produced by ev-2 containing line 100 cells (Coffin et al., 1978a). Third, the products of ev-2 (Fig. 1A) and ev-11 (Fig. 1B) yielded identical fingerprints, indicative of a very close relationship between these proviruses. The genomes associated with ev-10 (Figs. lC, and D) and ev-12 (Figs. 1E and F) were distinct from that of ev-2 and from each other. The ev10 virus genome differed from that of ev2 by the presence of three ev-lo-specific oligonucleotides (602,611,614) and the absence of four ev-2-specific oligonucleotides (05,06,07,09). Similarly, the ev-12 genome differed from ev-2 by the absence of two and the presence of three oligonucleotides (616,619,620). That the differences are inherent in the proviruses associated with each locus is shown by the identical fingerprints obtained from virus isolated from different ev-12 containing embryos (Figs. 1E and F) and from virus isolated from two independent lines of chickens that contain ev-10 (Figs. 1C and D).

ALV

GENOMES

57

RAV-0 has a U3 region that is distinct from that of exogenous viruses (Coffin et al., 1978a; Tsichlis and Coffin, 1980a; Hayward, 1977; Hughes et aZ., 1979). The U3 region comprises a 3’ noncoding sequence which forms part of the LTR of integrated proviruses. The 3’ proximal regions of the genomes of ev-2 and ev-10 are displayed in Figs. 2B and C. The identical composition of this portion of these two endogenous viruses indicates that the U3 regions of these viruses are closely related and distinct from that of exogenous viruses (Fig. 2A). The inferred oligonucleotide maps of these and other endogenous viruses discussed later are shown in Fig. 3A and the composition of RNase A digestion products of oligonucleotides that distinguish the various endogenous virus genomes from one another is shown in Table 2. Note that most or all of the differences are explicable as single base changes; for example, the base composition of 01 differs from that of 619 only by one additional U and one less C residue. Thus, although most of the endogenous proviruses are marked by distinctive oligonucleotides, they are very closely related to one another. Since

FIG. 3. Genomes of endogenous, exogenous, and recombinant viruses. The top line shows the oligonucleotide map of the genome of RAV-0 (Coffin et al., 1978a) modified according to available nucleotide sequence information (see Materials and Methods). Downward pointing arrows indicate the location of oligonucleotides whose exact position can be inferred with confidence; lines without arrowheads mark the approximate locations of other oligonucleotides. The remaining lines show the genomes of endogenous viruses (A), a selection of exogenous virus genomes (B), and some recombinants between endogenous and exogenous viruses (C). The oligonucleotide maps in A are inferred from the experiments shown in Figs. l-6 and other data (en-3 from Coffin et al., 1978b; ev-9 from Robinson et al., 1980; and ~-1 from Conklin et al., 1982). In all panels, oligonucleotides identical to those found in the RAV-0 genome are shown by a vertical line; the absence of a RAV0 specific oligonucleotide by a small box. An open box indicates oligonucleotides found in endogenous viruses other than RAV-9, a shaded box indicates oligonucleotides not detected in any endogenous virus examined. In some eases, allelic oligonucleotides (differing from prototype by one or a few base changes; see Table 2) are indicated by the number over a box; (-) indicates the absence of a detectable allelic oligonucleotide (i.e., the sequence must differ by at least a (C, A, or U) - G change). Dashed lines show regions not determined or deleted. B shows exogenous virus genomes: Pr-B (Coffin and Billeter, 1976), Pr-A and Pr-C (which although not identical, differ in the same way from RAV-0) (Joho et al., 1976, and SR-RSV-D (Tsichlis et al., 1980), as well as B-RSV(-), a replication-defective transforming virus (M. Champion, personal communication), and three lymphoid leukosis viruses: RAV-1 (Joho et al., 1975; M. Champion, personal communication), RAV-2 (Coffin et al., 1978b), and Cr-117, a fresh field isolate of ALV provided by L. Crittenden (C. Halpern, personal communication). C shows five recombinants between exogenous and endogenous viruses as indicated: four RAV-60’s (Rettenmier and Hanafusa, 1977; Robinson et al., 1980), and NTRE-7 (Tsichlis and Coffin, 1980a; Robinson et al., 1982).

-gag

A. ENDOGENOUS

5’ RU5L

0 I

I I

2 I

3 I

4 I

PO1

5 I

env-

6 I

7 I

u3

3’

ev-I

ev-3

ev-IO

ev-7

ev-I2

ev-9

ev-II

ev-2

RAV- 0

I

II

I I

I

I

I

I

I

I

I

II

II

II

II

II

f

I

II

I:

:

I

II

1:

I

II

C. RECOMBINANT

t

I

I

I

t

8. EXOGENOUS

II

I

I

I

I

I

II

II

II

II

II

II

I

I

II

II

I I

I I

II

I

I

I

I

I

I

II

II

II

II

II

II

II

I

I

I

I

I

I

I

I

1

I

I

I

I

I

I

I

I

I

I

I

I

I

II

II

II

II

II

II

II

II

II

III1

616

II

II

II

II

II

II

II

I I I I/.# I III I I IIIIII

II

III

607,-,602601

607,-)60260!

llll+j+Hj.#

I

II

I I

II

II

--------------

II

II

I

'1

'1

IM

'1

I

I

‘1

I

-I

-m

NTRE-7 (Pr -B x ev-2)

NY204 RAV-60 (RAV-2 x ev- 91

NY203 RAV-60 (RAV-I x ev-9)

NY 202 RAV-60 (RAV-2 x ev-3)

NY 201 RAV-60 (RAW x ev-3)

Cr-117

RAV2

RAV-I

RS V(-l

Pr A,C

Pr-B

5

60

COFFIN ET AL. TABLE 2 OLIGONUCLEOTIDES SPECIFIC FOR DIFFERENT ENDOCNEOUS VIRUSES~ Oligonucleotide

Composition

or Sequence

ev loci

1617 13

m’GpppGmCCAUUUUACCAUCCACCAUUG” m7GpppGmCCAUUUUACCAUUCACCAUUGb

2, 1 7, 10, 11, 12

i 014 C-1

AU, AC, 9C, 4U, G

71, 2, 3, 10, 11, 12

i 01 619

AU, 9C, lOC, 5U, 4U, GG

12 2, 3, 7, 10, 11 1,

i 05 614

A&, 3AU, A&, 2AU, 2C, 2C,4U, 4U, G G

10 1, 2, 7, 11, 12

1607 t-1

A&, 3AU, 3U, G

31, 2, 7, 9, 10, 11, 12

i 308 t-1

AU, 3AC, 6C, 3U, G

2, 1 7, 9, 10, 11, 12

i 07 l-1

A&J, 2AC, 7C, U, A2G

3, 1, 2, 7, 9, 10 11, 12

r(-) ( 602 .613 i 012 601 i 06 611 1402 616

2AzC, AU, 3AC, 4C, 2U, G AzU, A&, 2AU, 2AC, 7C, llU, AG AJJ, A&, 2AU, 2AC, 6C, lOU, AG A&, AU, AC, 2C, 5U, G A&, AU, 2AC, 2C, 5U, G A&, AC, AC, 4C, 5C, 2U, 3U, A3G ABG A&,

2, 9, 11, 12 3, 7, 10 1 1, 2, 7, 9, 10, 11, 12 3 1, 2, 3, 7, 9, 11, 12 10 9, 1, 2, 12 3, 7, 10, 11

i 09 t-1

A&, AU, 2AC, 4C, 2U, AG

7, 1, 2, 10 3, 11, 12

i 10 618 (08 ’ 612 ,615

A&, AU, 2AC, 4C, 2U, G

2, 3, 7, 10, 11, 12 1 2, 7, 10, 11, 12 1 3

A&J, 2AC, 8C, 3U, A2G

A&, 2AU, AC, 4C, 2U, G CCACCAUCAAAUAAACGd CCAUCAAAUAAACG” A&,

A&, AU, 2AC, 3C, G

” RNase digestion products shown in most cases were determined 1978a), taking into account the relative mobility in the two-dimensional of the numbers of C and U residues. b Schwartz et al., 1977; Hughes, 1982. ’ Hishinuma et al., 1981. d Hughes, 1982.

the fingerprint displays a total of about 550 nucleotides (Coffin et al., 19’78a), the five changes that distinguish ev-2 from ev10, for example, probably correspond to less than 1% sequence divergence.

Genomes of Defective Endogenous Proviruses Many endogenous proviruses have defects that prevent their expression as complete infectious virus, We have therefore

as previously described (Coffin et al., gel to obtain a more accurate estimate

used other strategies to examine the genomes associated with such proviruses. Ev-3, -9. Portions of some defective endogenous virus genomes recombine with superinfecting exogenous virus to form subgroup E recombinants (Hanafusa et al., 1970). Such recombinants, generically called RAV-GOs, inherit the env* gene from the endogenous parent, the U3 region from the exogenous parent, and other regions of the genome from either parent (Hayward and Hanafusa, 1975; Coffin et al., 197813;

ENDOGENOUS AND EXOGENOUS ALV GENOMES

Robinson et al., 1980). Comparison of the genomes of these viruses with those of the parent exogenous virus permits the inference of a partial oligonucleotide map of the endogenous virus parent. We have previously published fingerprints of RAV-60s obtained from ev-3 and ev-9 (Coffin et al., 1978b; Robinson et al., 1980). Figure 3C shows schematic oligonucleotide maps of four of these viruses, and Fig. 3A shows inferred oligonucleotide maps of the genomes encoded by these loci. Ev-1. A different approach was taken with ev-1, which is nearly ubiquitous in white leghorn chickens (Tereba and Astrin, 1980). Ev-1 encodes no distinctive phenotype and only very low levels of virus-related RNA (Hayward et al., 1980). We have recently found a chicken embryo with a high spontaneous expression of ev1 (Conklin et al., 1982). We and others have also found that high levels of expression of ev-1 can be induced with 5-azacytidine (Groudine et al., 1981; Conklin et al., 1982). Cells that express ev-1 produce noninfectious, reverse transcriptase negative, envelope glycoprotein negative particles. These particles contain genomic RNA from which fingerprints have been obtained (Conklin et al., 1982). The map and composition of oligonucleotides contained in ev-1 particle RNA are shown in Fig. 3A and Table 2. The map is consistent with that inferred from RAV-60 type recombinants isolated from ev-1 containing cells (Conklin, 1982). The presence of ev-1 oligonucleotides in such recombinants indicates that ev-1 can donate sequences encoding subgroup E envelope determinants to viruses of other subgroups. Ev-1 can also undergo recombination with infecting subgroup E viruses. A preparation of ev-2 ILV was passaged 20 times in either turkey cells (Fig. 4A) or C/O ev1 K28 chicken cells (Fig. 4B). The unpassaged virus is shown in Fig. 1A. Although passage through the ev-negative turkey cells had no visible effect on the virus genome, several changes were apparent in the virus passed in m-1 cells, including a reduction in the amounts of ev-Zspecific oligonucleotides 13 and 308 and the appearance of less than molar amounts of

61

oligonucleotides characteristic of ev-l(613 and 617). We conclude that repeated passage of a subgroup E virus through ev-1 containing cells can lead to the accumulation of a significant fraction of recombinant genomes between the infecting virus and the ev-1 genome, even though the latter is ordinarily expressed at a low level and there is no apparent selective advantage of the recombinant, Ev-7. Cells that contain ev-‘7 and ev-1 can be induced with BrdU to yield noninfectious reverse transcriptase-containing virus at a frequency of approximately one particle per 10 cells per day in culture (Robinson et al., 1979a). These same cultures also produce infectious virus, although the frequency of occurrence of infectious virus is much lower than that for noninfectious virus and was estimated to be about one particle per 4 X 10’ cells per day (Robinson et al., 1979b). The occurrence of infectious virus is more frequent if cells contain, in addition to ev-1 and ev7, either ev-3 or ev-9, which are associated with a high level of env glycoprotein expression (Astrin and Robinson, 1979; Astrin et al., 1979a, 1980a). Analysis of the proteins of infectious virus produced by ev-7 containing cells reveals that each of these viruses contains unique p19 proteins (Robinson et al., 197913).These results were interpreted to indicate that the virus coded for by ev-p7 can, at a low frequency, recombine with information encoded by other loci to give infectious virus. Analysis of the genomes of viruses induced from various cells containing ev-7 supported this conclusion (Figs, 5 and 6) and allowed us to infer the structure of the ev-7 genome. The genome of the noninfectious virus released soon after BrdU treatment is shown in Fig. 5A. Although it was impossible to obtain sufficient material for a complete analysis (50 mCi of 32P were required to obtain the lo3 cpm fingerprinted), it is apparent that this “genome” was a mixture of species which were closely related to the products of other endogenous proviruses. Since only ev-1 and ev-7 were present in the cells, we assume that products of both were present in the mixture. Some of the low-yield oligonucle-

62

COFFIN

ET AL.

FIG. 4. T1 oligonucleotide fingerprints of genomes of ILV induced from embryo 1972 (w-2) and passaged 20 times in turkey (T/BD) ceils (A) and K28 (C/O en-l) cells (B).

otides (e.g., 613 and 617) are characteristic of ev-1 (Fig. 3); we therefore can assign allelic low-yield oligonucleotides (612, 13, and 308) to ev-7. By a similar argument, we can infer that ev-‘7 does not encode oligonucleotides 07 and 09. Most of the remaining oligonucleotides were apparently common to both proviruses, since they were present in high yield. A few additional oligonucleotides (arrows) seen in high yield in the preparation were not readily identifiable as characteristic of endogenous viruses, and may represent uncharacterized regions of ev-7 or contaminating RNA species.

cells

Several examples of the genomes of infectious virus appearing after a long latent period following BrdU treatment are shown in Figs. 5B-F. By comparison with the noninfectious 15B-ILV, these viruses were relatively pure and contained, in high yield, markers which were only fractionally present in the induced noninfectious virus, i.e., oligonucleotides 07, 08, 602, 612, 613, and 308. The content of these markers varied from isolate to isolate. Since all virus genomes contained at least one ev-1 specific marker, and none contained all, the viruses must have been recombinants of ev-1 and ev-‘7. This conclusion is consistent

ENDOGENOUS AND EXOGENOUS ALV GENOMES

FIG. 5. Genomes of the virus induced from ev-7-containing cells of line 15B. (A) shows the fingerprint of the genome of noninfectious virus particles produced early after induction of 1087 cells (eu-1, ev-7) and (B) the infectious virus produced by the same cells 2 weeks later. (C-E) show independently induced infectious viruses from embryo 8715 (en-l, eu-7). C and D were isolated by the same protocol as in B; E was isolated by passage of the supernatant on turkey cells for 2 weeks after BrdU treatment. F shows the virus from embryo 8694 (ev-1, ev-7). Only oligonucleotides which vary from one virus to another are labeled. The fingerprints in C-F were run using tris-borate rather than tris-citrate in the second dimension (see Materials and Methods).

with that reached by analysis of the proteins of a similar set of viruses (Robinson et al., 1979b). The U3 region of one of these viruses (shown in Fig. 2D) was indistinguishable from that of RAV-0 and contained oligonucleotide 08. Since the 3’ proximal oligonucleotide of ev-1 is 612,

these results indicate that the 3’ proximal oligonucleotide of ev-7 is 08. Introduction of a provirus encoding chf (ev-3 or ev-9) in combination with ev-1 and ev-7 by crossing 15B with K16 or K18 chickens leads to cells that can be induced to yield infectious virus at a much higher

64

COFFIN

ET AL.

lr.: ) *

FIG. 6. Genomes of viruses induced from 15B-related embryos. Virus was induced from (15B X K16) embryo 1263 (m-1, m-3, a-7) (A); (15B X K16) embryo 1264 (a-1, a-3, m-7) (B); (15B X K18) embryo 1188 (en-l, eu-7, m-9) (C); and (15B X KM) embryo 1278 (eu-1, e-u-7, m-9) (D). AD show fingerprints of the total genomes; fingerprints of the respective 3’ proximal sequences are shown in E-H.

frequency than those with only ev-1 and ev-7 (Robinson et al., 1979b). Figure 6 shows the genomes of uncloned infectious viruses induced from two 15B X K16 (Fig. 6A and B) and two 15B X KU3 (Fig. 6C and D) embryos. These viruses differed from the ev-

7 X ev-1 viruses (Fig. 5) in two ways. First, most were mixtures of viruses as was evident from the presence of oligonucleotides in less than molar amount. This result suggests that the virus arose from relatively frequent recombinational events.

ENDOGENOUS

AND

EXOGENOUS

Second, they contained additional markers not found in the ev-7 X ev-I viruses. As shown in Fig. 3, many of these new markers were also found in RAV-60s derived from ev-3 or ev-9 embryos. For example, 601 and 607 are characteristic of RAV-60s from g.s+chf+(ev-3) embryos (Fig. 3C), and the absence of both the 602 and 613 alleles is characteristic of RAV-60s from gs-chf + (m-9) embryos (Fig. 3C). We conclude that, in these embryos, the recombinational events also involved the ev-3 or ev-9 encoded genome. Gmwth Rates of End,ogenms and Recombinant Viruses The comparison of endogenous and exogenous virus genomes, shown in Figs. 3A and B, shows that the most consistent dif-

ALV

GENOMES

ferences between endogenous and exogenous virus genomes are found in a portion of env and the U3 region. The endogenous virus-specific region of env is virtually identical to that which encodes subgroup E host range (Tsichlis and Coffin, 1980a). Analysis of a set of recombinants between RAV-0 and exogenous viruses (Tsichlis and Coffin, 1980a, b) implied that the U3 region difference contained the major determinant encoding the difference in growth rate between those two groups of viruses. Figure 7 shows an experiment designed to test this hypothesis more directly and to determine whether these growth differences are consistent among II3 regions derived from different endogenous and exogenous viruses. In the growth test shown, all endogenous (i.e., Us”) viruses (open symbols) had indistinguishable growth rates in spite

FIG. ‘7. Growth potential of viruses with endogenous and exogenous Us regions. Cells from embryo 1334 were seeded as subconfluent monolayers and infected with the indicated dilutions of each virus. Three days after infection, cultures were transferred and challenged for superinfection resistance to subgroup E pseudotype of BH-RSV; BH-RSV (RAV-60). Levels of interference are given as the ratio of the titer (in focus forming units/ml) of challenge virus on infected cultures relative to the titer on parallel uninfected control cultures. Panels A and B contain data from independent experiments. IJ,’ viruses are represented by closed symbols and Uan viruses by open symbols: 0 NTRE-7; Uax of Pr-RSV-B A NY201 RAV-6% II,’ of RAV-1 0 NY202 RAV-6% U,” of RAV-2 n NY293 RAV-60, U,’ of RAV-1

65

0 RAV-0; A 15B X 0 15B X 0 15B X @ 15B X 0 c-ILV; T.Z15B -

U,” of ev-2 K16-E-1; U3” of K16-E-2; II,” of K18-E-1; Us” of KlS-E-2; Ua” of II,” of a-10 E-3; Us” of m-7

eu-3 m-3 m-1 and -7 or -9 m-1

66

COFFIN

of the different origins of these viruses and of small allelic differences in U3 detectable by fingerprinting (Fig. 3A). Similarly, all viruses with US’ regions (closed symbols) had identical growth rates regardless of additional regions of the genome derived from exogenous viruses. Note particularly that RAV-0 (open circles) and NTRE-7 (closed circles),viruses that apparently differ only within and near the USregion (Fig. 3; Tsichlis and Coffin, 1980a; Robinson et al., 1982) had growth rates identical to all other Us” and US’ viruses respectively. DISCUSSION

Relationships among the ev Proviruses In this paper we have presented a survey of the genomes encoded by eight of the ev loci commonly encountered in laboratory white leghorn chickens. From analysis of the genomes of viruses induced or rescued by recombination from cells containing proviruses at defined ev loci, we were able to infer complete oligonucleotide maps for ev-1, -2, -10, -11, and -12, and partial oligonucleotide maps for ev-3, -7, and -9. Pro-

ET AL.

viruses which we did not characterize include three, ev-4, ev-5, and ev-8, that do not appear to be expressed (Hayward et al., 1980), and one, ev-6, that is not efficiently rescued into virus and does not seem to yield recombinants (Conklin, 1982). Examination of the oligonucleotide maps of the various endogenous viruses shows that all of the proviruses studied are very closely related to one another. We estimate less than 2% sequence divergence between the most distantly related pair (ev1 and ev-7). There is, nevertheless, sufficient divergence to provide markers that distinguish all of the endogenous viruses from one another, with the exception of ev-2 and ev-11. Based on the oligonucleotide maps in Fig. 3, we have derived the relationship scheme shown in Fig. 8. While this scheme must be considered as tentative and represents only a subset of all endogenous ALV-related viruses (S. Astrin and A. Tereba, personal communication), several interesting features emerge. (1) ev-2 and ev-11 are relatively central to the scheme since they alone contain no oligonucleotides not shared with at least one other provirus. This implies possible di-

-* (4 * 62

* + _

09 - (4

,T t

* _ 0

*

. , .!$-3

07 * C-1

1

Ev-7

*

d

_ f

f

*

2

3

4

OLlGONUtLEOTIE

4 EV-l0 Cd-1 5

CMAIEES

FIG. 8. Divergence of eight endogenous viruses of white leghorn chickens. The minimum changes relating one genome to another are inferred from the data tabulated in Fig. 3 and Table 2 as oligonucleotide changes. The numbers on the bottom indicate the number of such differences relating each virus to ew-2 and ~-11. The numbers above each branch indicate probable point mutations in the indicated oligonucleotides, with the RAV-0 oligonucleotide shown first. Note that this is the simplest pattern possible for these data.

ENDOGENOUS AND EXOGENOUS ALV GENOMES vergence (of these eight proviruses at least) from a virus closely resembling RAV-0. The ancestral virus cannot be identical to RAV0 since Hughes et al. (1981) have found that ev-2 differs by a Sac1 restriction endonuclease cleavage site from several of the other proviruses. (2) Three proviruses (ev3, -7, and -10) form a distinct branch, characterized by the presence of oligonucleotide 602. In spite of this relationship, the phenotypes of cells containing these proviruses are very different; ev-10 is inducible to give infectious virus (Astrin et al., 1980a), ev-7 to give noninfectious virus (Robinson et al., 1979a; Smith and Crittenden, 1981), and ev-3 is defective due to a deletion in the gag-pal junction (Hayward et al., 1980; Hughes et al., 1981) yet expresses a high amount of env glycoprotein (Astrin and Robinson, 1979). Quite interestingly, however, none of these proviruses, despite an apparent common ancestor, occur in the same line of chicken (Tereba and Astrin, 1980). (3) ev-1 is not closely related to the other genomes. This result is interesting, since ev-1 is by far the most common of the loci studied in white leghorn chickens (Astrin, 1978; Hughes et al., 1979, 1981; Tereba and Astrin, 1980). However, its oligonucleotide composition and sequence data suggest that it is not ancestral to tl,e other proviruses. The ev-1 provirus has been molecularly cloned (Skalka et al., 1980) and the nucleotide sequence of each copy of its LTR has been determined (Hishinuma et al., 1981). Two of the oligonucleotides that distinguish ev-1 from the proviruses at other loci (617 and 612) lie within both LTRs. It seems most likely that the identical sequences arose as single mutations in the U3 and R portion of a viral genome which gave rise to duplicated mutations in the proviral form. An alternative (but less likely) explanation is that single mutations in a proviral LTR were duplicated by gene conversion with the other LTR (Scherer and Davis, 1980).

Recombinaticm between Endogenous Proviruses We have presented evidence for the formation of recombinants between the prog-

67

eny of proviruses at different ev loci. Such recombination was detected both between defective and nondefective endogenous viruses, for example, by passaging RAV-0 repeatedly in ev-1 containing cells (Fig. 4), and between the two defective endogenous proviruses at ev-7 and ev-1 (Fig. 5). Induction of ev-1 and ev-7 cells leads to a small “burst” of polymerase containing particles (Robinson et al., 1979a, b) with oligonucleotide markers characteristic of each provirus (Fig. 5A), followed much later by the clonal or nearly clonal outgrowth of replication-competent recombinants containing various combinations of markers from ev-1 and ev-7 (Figs. 5BF). This process is greatly accelerated in the presence of a defective provirus (ev-3 or ev-9) encoding high levels of chf. Induction of ev-1 and ev-7 cells also containing ev-3 or ev-9 often leads to the production of complex mixtures of recombinants, in which markers associated with all three loci can be detected (Fig. 6). It seems probable that the generation of these recombinants occurs by mechanisms similar to those involved in recombination between exogenous viruses (Wyke et al., 1974; Hunter, 1978; Coffin, 1979), i.e., formation of virions whose genomes are heterozygous for the parent virus, followed by recombination during infection. The role of ev-3 and ev-9 in this process may involve either provision of a relatively high level of RNA for the virions, or enhanced infectivity of the initial progeny due to the additional env glycoprotein.

Relationship between Endogenous and Exogenous Viruses Comparison of the genomes of the eight endogenous viruses to a collection of exogenous viruses shows that these two groups are related. There are several regions of the genome in which the exact nucleotide sequence must be conserved, since all viruses studied to date have some identical oligonucleotides. Highly conserved oligonucleotides are found in pal and in a portion of env not involved in subgroup specificity, encoding the N- and C-terminal parts of gp85 and the N-terminal part of gp37 (Coffin et al., 1978a; Tsichlis and Cof-

68

COFFIN

fin, 1980a). Less highly conserved regions are found in gag. In spite of these close relationships, endogenous viruses form a lineage distinct from exogenous viruses. This is evidenced by unique oligonucleotides common to all endogenous viruses in the region of env which encodes subgroup specificity and in Us. Endogenous viruses are therefore derived from an ancestor that is distinct from known exogenous viruses and exogenous viruses are descended from an ancestor that is distinct from that of the endogenous viruses. The conservation of endogenous virus genomes relative to exogenous viruses argues that endogenous viruses did not usually arise from rare germline integrations of exogenous infections, but rather from proviruses already transmitted in the germline. The absence of related endogenous proviruses in the germline of other species suggests that these viruses first gained access following divergence of the species gallus (Frisby et al., 1979). The generation of new endogenous proviruses seems most probably due to reinfection of the germ line. However, possibilities such as direct transposition of proviral DNA cannot be firmly excluded although there is no experimental evidence that such processes occur. In the case of some laboratory strains of mice (such as AKR and C,H), recent evidence implies that the introduction of new endogenous ecotropic proviruses into the germline is correlated with viremia arising from expression and replication of preexisting endogenous proviruses (Quint et al., 1981; Herr and Gilbert, 1982; Steffen et al., 1982). However, most chickens are resistant to infection by the subgroup E endogenous proviruses that they carry, and only a few lines (such as line 100) contain the appropriate combination of nondefective endogenous proviruses (ev-2 ) and susceptibility to subgroup E virus to undergo endogenous virus infection at an early age. By contrast, many chickens become infected with exogenous viruses of subgroup A and B, yet no related endogenous proviruses have been detected. There are several possible reasons for this apparent selectivity. First, endogenous viruses may have special features which permit them to infect

ET AL.

germline cells. For example, it is conceivable (in the complete absence of experimental evidence) that germline cells are permissive only for subgroup E virus infection but not for infection by virus of other subgroups. Second, it is possible that the germline represents a compartment of the host that is relatively protected from virus infection. However, once a virus has infected some cells within this compartment, it might have unique accessto germline cells. Thus, introduction of an endogenous provirus into the germline would be an extremely rare event, but once one was introduced, it would generate additional endogenous proviruses relatively rapidly. Much of the uniqueness of endogenous relative to exogenous viruses might therefore be due to their being descended from a single ancestor which happened, by chance, to gain entry into the germline. Third, the endogenous lifestyle must impose certain selective constraints on the genomes of the virus. The finding and breeding of chickens with no ALV-related endogenous proviruses (Astrin et al., 197913) demonstrates that these elements are not essential for the individual. It is therefore possible that endogenous proviruses are maintained in the germline not by selective forces acting on their hosts, but rather by virtue of their ability to replicate into new integration sites at a rate which balances their tendency to be lost by random deletion, and are therefore a form of “selfish DNA” element as proposed in another context by Doolittle and Sapienza (1980) and Orgel and Crick (1980). Although some proviruses (such as ev-3 and ev-6) which express high levels of env glycoprotein may confer subtle selective advantage on their host in the form of reduced susceptibility to ALV-induced pathogenesis (L. Crittenden, personal communication; Weiss and Frisby, 1982), there is no evidence for such a role of the majority of endogenous proviruses which are expressed at much lower levels (Hayward et al., 1980; Baker et al., 1981). It is thus important that endogenous viruses do no significant harm to their host, and it is likely that the failure to observe other types of ALV as germline proviruses reflect the pathogenic effects

ENDOGENOUS

AND

EXOGENOUS

such proviruses might have (for an analogous situation in mice, see Jaenisch, 1980). We consider it likely that the relatively low replication rate of endogenous viruses is related to their nonpathogenicity. It has previously been shown that the II3 region and adjacent sequences differ significantly in the genomes of RAV-0 and exogenous ALV (Coffin et al., 1978a; Neiman, 1978; Hayward, 1977). We have proposed that these differences are responsible for the difference in growth rate between these viruses (Tsichlis and Coffin, 1980a, b). In this report we have shown that they confer 30- to 50-fold differences in growth rate and that all of the endogenous viruses examined have U3 regions and growth rates characteristic of RAV-0. Based on its position in the LTR, the U3 region could be expected to contain several sequences important in different parts of the virus life cycle. Recent evidence implies a crucial role of the promoter activity of this region in lymphoma induction (Neel et al., 1981; Payne et al., 1981; Hayward et al., 1981). It is likely therefore that the difference in U3 is related to the difference in pathogenicity of endogenous and exogenous viruses (Tsichlis and Coffin, 1980b). Consistent with this idea, all viruses with U3 regions characteristic of exogenous viruses have been found to cause disease, although the pathogenic spectra of these viruses may vary from one isolate to another. All the RAV-60s shown in Fig. 3C have been found to induce a high incidence of lymphoma similar to that of standard lymphoid leukosis virus (Robinson et al., 1980; Crittenden et al., 1980) and the transformationdefective derivative of Pr-B induces a high incidence of osteopetrosis (Robinson et al., 1982). Interestingly, however, the recombinant NTRE-7 which apparently has only the U3 portion and a small amount of additional information from exogenous virus (Fig. 3C; Tsichlis and Coffin, 1980a; Tsichlis et al., 1982) and a replication rate identical to the RAV-60 strains shown (Fig. 7) does not induce a high incidence of lymphoma or osteopetrosis but rather induces a low incidence of a variety of neoplasms (Robinson et al., 1982). Thus, there appear to be sequences outside of U3 which target avian leukosis viruses for the induction of

ALV

GENOMES

69

lymphoma or osteopetrosis. These sequences may represent additional regions of the ALV that have been selected for nonpathogenicity in endogenous viruses. ACKNOWLEDGMENTS We thank L. Crittenden and H. Hanafusa for providing some of the cells and viruses used in this study, and M. Champion and 0. Turetsky for skilled teehnical assistance. This work was supported by Grants CA 17659 (to J.C.) and CA 23806 and P30 CA 127708 (to H.R.) from the National Cancer Institute, and a grant from the Leukemia Association of America to P.N.T. J.M.C. is a recipient of a Faculty Research Award from the American Cancer Society; P.N.T. was a postdoctoral fellow of the National Cancer Institute; and K.F.C. was supported in part by Grant GM 07310 from the National Institutes of Health. REFERENCES AARONSON, S. A., and STEPHENSON, J. R. (1978). Independent segregation of loci for activation of biologically distinguishable RNA C-type viruses in mouse cells. Proc. Nat. Acad. Sci. USA 70, 20552058. ASTRIN, S. (1978). Endogenous viral genes of white leghorn chickens: A common site of residence as well as sites associated with specific phenotypes of viral gene expression. Proc. Nat. Acad Sci. USA 75, 5941-5945. ASTRIN, S. M., and ROBINSON, H. L. (1979). Gs, an allele of chickens for endogenous avian leukosis viral antigens, segregates with ev 3, a genetic locus that contains structural genes for virus. J. viral. 31,420-425. ASTRIN, S. M., CRITTENDEN, L. B., and Buss, E. G. (1979a). eu-3, a structural gene locus for endogenous virus, segregates with the gs+chf+ phenotype in matings of line 6a chickens. Virology 99, 1-9. ASTRIN, S. M., Buss, E. G., and HAYWARD, W. S. (1979b). Endogenous viral genes are non-essential in the chicken. Nature (londtrn) 282, 339-341. ASTRIN, S. M., ROBINSON, H. L., CRIITENDEN, L. B., Buss, E. G., WYBAN, J., and HAYWARD, W. S. (198Oa). Ten genetic loci in the chicken that contain structural genes for endogenous avian leukosis viruses. Cold Spring Harbor Symp. &ant. BioL 44. 11051109. ASTRIN, S. M., CRITTENDEN, L. B., and Buss, E. G. (1980b). ev-2, a genetic locus containing structural genes for endogenous virus, codes for Rous-associated virus type 0 produced by line 7* chickens. J. Viral 33, 250-255. BAKER, B., ROBINSON, H., VARMUS, H. E., and BISHOP, J. M. (1981). Analysis of endogenous avian retrovirus DNA and RNA: Viral and cellular determinants of retrovirus gene expression. Virology 114, 8-22.

70

COFFIN

COFFIN, J. M. (1979). Structure, replication, and recombination of tumor virus genomes: Some unifying hypothesis. J. Gen Viral 42, l-46. COFFIN, J. M. (1982). Endogenous Viruses. Zn “Molecular Biology of Tumor Viruses. Part III. RNA Tumor Viruses.” (R. Weiss, N. Teich, H. Varmus, and J. M. Coffin, eds.), Chap. 10, pp. 1109-1203. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. COFFIN, J. M., and BILLETER, M. A. (1976). A physical map of the Rous sarcoma virus genome. J. Mel BioL 100, 293-318. COFFIN, J. M., CHAMPION, M. A., and CHABOT, F. (1978a). Nucleotide sequence relationships between the genomes of an endogenous and an exogenous avian tumor virus. J. ViroL 28, 972-991. COFFIN, J. M., CHAMPION, M. A., and CHABOT, F. (1978b). Genome structure of avian RNA tumor viruses: Relationships between exogenous and endogenous viruses. In “Avian RNA Tumor Viruses.” (S. Barlatti and C. deGiuli-Morghen, eds.), pp. 68 78. Piccin Medical Books, Padua, Italy. CONKLIN, K. F. (1982) Doctoral thesis. Tufts University, Boston, Massachusetts. CONKLIN,K. F., COFFIN,J. M., ROBINSON,H. L., GROUDINE, M., and EISENMAN,R. (1982). Role of methylation in the induced and spontaneous expression of the avian endogenous virus ev-1: DNA structure and gene products. MoL Cell. BioL 2, 638-652. CRITTENDEN, L. B., HAYWARD, W. S., HANAFUSA, H., and FADLEY, A. M. (1980). Induction of neoplasms by subgroup E recombinants of exogenous and endogenous avian retroviruses (Rous-associated virus type 60). J. ViroL 33, 915-919.

DOOLITTLE,W. F., and SAPIENZA, C. (1980). Selfish

ET AL.

viral RNAs in avian oncovirus infected cells. J. ViroL 24, 47-63.

HAYWARD,W. S., and HANAFUSA, H. (1975). Recombination between endogenous and exogenous RNA tumor virus genes as analyzed by nucleic acid hybridization.

J. ViroL 15, 1367-1377.

HAYWARD, W. S., BRAVERMAN,S. B., and ASTRIN, S. M. (1980). Transcriptional

products

and DNA

structure of endogenous avian proviruses. Cold Spring Harbor

Symp. Quant. Bio. 44, 1111-1121.

HAYWARD,W. S., NEEL, B. G., and ASTRIN,S. M. (1981). Activation

of a cellular

one gene. Nature

(London)

290, 475-480. HERR, W., and GILBERG,W. 1982. Germ-line MuLV reintegrations 865-868.

in AKR/J

mice. Nature (London) 296,

HISHINUMA,F., DEBONA,P. J., ASTRIN,S., and SKALKA, A. M. (1981). Nucleotide sequence of the acceptor site and termini of integrated avian endogenous provirus ev-1: Integration creates a 6bp repeat of host DNA. Cell 1, 155-164.

Hsu, J., SABRAN,J., MARK, G., GUNTAKA,R., and TAYLOR, J. (1978). Analysis of unintegrated avian RNA tumor virus double-stranded DNA intermediates. J. ViroL 28, 810-818.

HUGHES,S. (1982). Sequence of the LTR and adjacent segments of the endogenous avian virus RAV-0. J. ViroL 43, 191-200.

HUGHES, S., VOGT, P. K., SHANK, P. R., SPECTOR, P. H., KUNG, H. J., BREITMAN,M. L., BISHOP,J. M., and VARMUS,H. E. (1978). Proviruses of avian sarcoma viruses are terminally redundant, coextensive with unintegrated linear DNA, and integrated at many sites in rat cell DNA. Cell 15, 1397-1410.

genes, the phenotype paradigm and genome evolution. Nature (Lo&o@ 284, 601-603.

HUGHES, S. H., PAYVAR, F., SPECTOR,D., SHINKE, R. T., ROBINSON,H. L., PAYNE, G., BISHOP,J. M.,

FRISBY,D. P., WEISS, R. A., ROUSSEL,M., and STEHELIN, D. (1979). The distribution of endogenous

and VARMUS, H. E. (1979). Heterogeneity of genetic loci in chickens: Analysis of endogenous viral and nonviral genes by cleavage of DNA with restriction endonucleases. Cell 18, 347-359.

retrovirus sequences in the DNA of galliform does not coincide with avian phylogenetic tionships. Cell 17, 623-634.

birds rela-

GROUDINE,M., EISENMAN, R., and WEINTRAUB, H. (1981). Chromatin structure of endogenous retroviral genes and activation by an inhibitor of DNA methylation. Nature (Lo&m) 292, 311-317.

HANAFIJSA, T., HANAFUSA, H., and MIYAMOTO, T. (1970). Recovery of a new virus from apparently normal cells by infection with avian tumor viruses. Proc. Nut. Acad. Sci. USA 67, 1797-1803.

HANAFUSA, H., HANAFUSA, T., KAWAI, S., and LuGINBUHL,R. E. (1974). Genetic control of expression of endogenous virus genes in chicken cells. Virology

58, 439-448. HANAFUSA, H., HAYWARD, W. HANAFUSA, T. (1975). Control

S., CHEN, J. H., and

of expression of tumor virus genes in uninfected chicken cells. Cold Spring Harbor Symp. Quant. BioL 39, 1139-1144.

HAYWARD,W. S. (1977). Size and genetic content of

HUGHES, S. H., TOYOSHIMA,K., BISHOP, J. M., and VARMUS, H. E. (1981). Organization of the endogenous proviruses of chickens: Implications for origin and expression. Virology 108, 189-207. HUNTER, E. (1978). The mechanism for genetic recombination in the avian retroviruses. Cum. Top Microbial ZmmunoL 79, 295-309. JAENISCH, R. (1980). Retroviruses and embryogenesis: Microinjection of Moloney leukemia virus into midgestation mouse embryos. Cell 19.181-188. JOHO, R. H., STOLL, E., FRIIS, R. R., BILLETER, M. A., and WEISSMANN, C. (1976). A partial genetic map of Rous sarcoma virus RNA: Location of polymerase, envelope and transformation markers. In “Animal Virology.” (D. Baltimore, A. S. Huang, and C. F. Fox, eds.), pp. 127-145. Academic Press, New York. LINIAL, M., and NEIMAN, P. E. (1976). Infection of

ENDOGENOUS AND EXOGENOUS ALV GENOMES

71

TSICHLIS,P. N., and COFFIN,J. M. (1980). Subgroup chick cells by subgroup E virus. fir73, 508E avian leukosis virus-associated disease in chick520. ens. Cold Spring Harbor Symp. Quant Bid 49,1133MOTTA, J. V., CRITTENDEN,L. B., PURCHASE,H. G., 1142. STONE,H. A., and WIRER, R. L. (1975). Low onROBINSON,H. L., ASTRIN, S. M., SENIOR,A. M., and cogenic potential of avian endogenous RNA tumor SALAZAR,F. H. (1981). Host susceptibility to envirus infection or expression. .Z. Nat. Cancer Inst. dogenous viruses: Defective, glycoprotein-express55, 685-689. ing proviruses interfere with infection. J. Viral. 40, NEEL, B. G., HAYWARD,W. S., ROBINSON,H. L., FANG, 745-751. J., and ASTRIN, S. M. (1981). Avian leukosis virusROBINSON,H. L., BLAIS, B. M., TSICHLIS,P. N., and induced tumors have common proviral integration COFFIN,J. M. (1982). At least two regions of the sites and synthesize discrete new RNAs: Oncogenviral genome determine the oncogenic potential of esis by promoter insertion. Cell 23, 323-334. NEIMAN, P. E. (1978). Mapping by competitive hyavian leukosis viruses. Proc. Nat. Acad Sci. USA 79, 1225-1229. bridization of sequences which differ between enROBINSON,H. L. (1982). Retroviruses in Cancer. Redogenous and exogenous chicken leukosis virus. Viviews of Infectious Disease. (In press). rology 85, 9-16. ORGEL,L. E., and CRICK,F. H. C. (1980). Selfish DNA: SCHWARTZ,D. E., ZAMECNIK,P. C., and WEITH, H. J. (1977). Rous sarcoma virus genome is terminally The ultimate parasite. Nature (London) 284, 604607. redundant: The 3’ Sequence. Proc. Nat. Acad Sci PAYNE,G. S., COURTNEIDGE, S. A., CRITTENDEN,L. B., USA 74,994-998. FADLY, A. M., BISHOP,J. M., and VARMUS,H. E. SHANK, P. R., HUGHES, S., KUNG, H. J., MAJORS, J. E., QUINTRELL, N., GUNTAKA, R. V., BISHOP, (1981). Analysis of avian leukosis virus DNA and RNA in bursal tumors: Viral gene expression is not J. M., and VARMUS,H. E. (1978). Mapping uninterequired for maintenance of the tumor state. Cell grated forms of avian sarcoma virus DNA: Both 23,311-322. termini of linear DNA bear a 300 nucleotide sePEDERSON,F. S., and HASELTINE,W. A. (1980). A miquence present once or twice in two species of circromethod for detailed characterization of high cular DNA. Cell 15, 1383-1395. molecular weight RNA. Methods Enz?/mol 65,680SCHERER,S., and DAVIS, R. W. (1979). Replacement 687. of chromosome segments with altered DNA sePINCUS,T. (1980). The endogenous murine type C viquences constructed in vitro. Proc. Nat. Acad. Sci. ruses. In “Molecular Biology of RNA Tumor ViUSA 76, 4951-4955. ruses.” (J. Stephenson, ed.). pp. 77-130. SKALKA, A., DEBONA,P., HISHINUMA,F., and MCCLEQUINT, W., QUAX,W., VANDERP~EN, H., and BERNS, MENTS,W. (1980). Avian endogenous proviral DNA: A. (1981). Characterization of AKR murine leuAnalysis of integrated ew-1 and a related gs-chfkemia sequences in AKR mouse substrains and provirus purified by molecular cloning. Cold Spring structure of integrated recombinant genomes in Harbor Symp. Quad. Biol. 44, 1097-1104. tumor tissues. J. ViroL 39, l-10. SMITH, E. J., and CRITTENDEN,L. B. (1981). SegreRE~ENMEIR, C. W., and HANAFUSA,H. (1977). Strucgation of chicken endogenous viral loci ev-7 and eutural protein markers in the avian oncoviruses. J. 12 with the expression of infectious subgroup E Viral 24, 850-864. avian leukosis viruses. Virology 112, 370-374. ROBINSON,H. L. (1976). Intracellular restriction on STEFFEN,D., TAYLOR, B. A., and WEINBERG,R. A. the growth of induced subgroup E avian type C (1982). Continuing germ line integration of AKV viruses in chicken cells. J. Viral 18, 856-866. proviruses during the breeding of AKR mice..Z. ViROBINSON,H. L., SWANSON,C. A., HRUSKA,J. F., and rol 42, 165-175. CRI’ITENDEN,L. B. (1976). The production of unique TEICH, N., MAK, T., WYKE, J., BOSE,H., and HARDY, C type viruses by chicken cells grown in bromoW. (1982). Pathogenesis. In “Molecular Biology of deoxyuridine. Virology 69, 63-74. Tumor Viruses. Part III. RNA Tumor Viruses.” ROBINSON,H. (1978). Inheritance and expression of (R. A. Weiss, N. Teich, H. E. Varmus, and J. M. chicken genes which are related to avian leukosisCoffin, eds.), Chap. 8, pp. 785-998. Cold Spring Harsarcoma viruses. Curr. Top. Micr&ioL Zmmunol. 83, bor Laboratory, Cold Spring Harbor, N. Y. 1-36. TEREBA,A., and ASTRIN, S. M. (1980). Chromosomal ROBINSON,H. L., ASTRIN, S. M., and SALAZAR,F. H. localization of m-1, a frequently occurring endog(1979a). V-15B, an allele of chickens for the proenous retrovirus locus in white leghorn chickens, duction of a noninfectious avian leukosis virus. Viby in situ hybridization. J. ViroL 35, 888-894. rology 99, 10-20. ROBINSON,H. L., EISENMAN,R. E., SENIOR,A., and TSICHLIS,P. N., and COFFIN,J. M. (1980a). Recombinants between endogenous and exogenous avian RIPLEY, S., (1979b). Low frequency production of recombinant subgroup E Ieukosis virus by unintumor viruses: Role of the c region and other porfected V-l& chicken cells. virology 99, 21-30. tions of the genome in the control of replication ROBINSON,H. L., PEARSON,M. N., DESIMONE,D. W., and transformation. J. ViroL 33, 238-249.

72

COFFIN

TSICHLIS, P. N., and COFFIN, J. M. (1986b). Role of the c region in relative growth rates of endogenous and exogenous avian oncoviruses. Cold Spring Harbor Symp. C&ant. Biol 44,1123-1132. TSICNLIS, P. N., CONKLIN, K. F., and COFFIN, J. M. (1980). Mutant and recombinant avian retroviruses with extended host ranges. Proc. Nat Acad Sci USA 77, 536-540. TSICHLIS, P. N., DONEHOWER, L., HAGER, G., ZELLER, N., MALAVARCA, R., ASTRIN, S., and SKALKA, A. M. (1982). Sequence comparison in the crossover region of an oncogenic avian retrovirus recombinants and its nononcogenic parent: Genetic regions that control growth rate and oncogenic potential. J. Mol. and Cell BioL 2, 1331-1338. VARMUS, H. E., and SWANSTROM, R. (1982). Replication. In “Molecular Biology of Tumor Viruses: RNA Tumor Viruses” (R. Weiss, N. Teich, H. Varmus, and J. Coffin, eds.), Chap. 5, pp. 369-512. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y.

ET AL. VOGT, P. K., and FRIIS, R. R. (1971). An avian leukosis virus related to RSV(0): Properties and evidence for helper activity. Virology 43, 223-234. WANG, S. Y., HAYWARD, W. S., and HANAFUSA, H. (1977). Genetic variation in the RNA transcripts of endogenous virus genes in uninfected chicken cells. J. Vird 24, 64-73. WEINBERG, R. A. (1980). Origins and roles of endogenous retroviruses. Cell 22, 643-644. WEISS, R. A. (1969). The host range of Bryan strain Rous sarcoma virus synthesized in the absence of helper virus. J. Ga. ViroL 5, 511-528. WEISS, R. A., and FRISBY, D. P. (1982). Are avian endogenous viruses pathogenic? Proc. Xth International Symposium for Comparative Research on Leukemia and Other Diseases (D. Yohn, ed.). In press. WYKE, J. A., BELL, J. G., and BEAMAND, J. A. (1974). Genetic recombination among temperature-sensitive mutants of Rous sarcoma virus. Co/d Spring Harbor Symp. Quad. BioL 39, 897-905.