Viral DNA in transformed cells. II. A study of the sequences of adenovirus 2 DNA IN NINE LINES OF TRANSFORMED RAT CELLS USING SPECIFIC FRAGMENTS OF THE VIRAL GENOME;

Tn5 mutagenesis of the transforming genes of human adenovirus type 5

We have cloned the entire human adenovirus type 5 (Ad5) genome into the pBR322 plasmid in two segments: the BamHI-A fragment (21 kb) and the BamHI-B fragment (15 kb). We have also generated a series of clones with smaller Ad5 DNA inserts, all containing the left-end of the viral genome. One such clone, pXCl, containing the left 16% of the Ad5 DNA molecule, has been shown to transform rodent cells by DNA transfection. We have used the transposable element Tn5 as an insertion mutator to isolate pXCl mutants containing Tn5 inserted at a large number of sites. By assaying transforming activity of selected pXC::Tn5 plasmids we have identified Ad5 sequences which are essential for DNA-mediated transformation. Our results with these mutants and with a plasmid pCD1, containing a deletion within the Ad5-transforming region, indicate that sequences present in both early region 1a and the N-terminal region of early region 1b are essential for DNA-mediated transformation.

Analysis of type 5 adenovirus transformation with a cloned rat embryo cell line (CREF)

A recently isolated cloned cell line of Fischer rat embryo fibroblasts (CREF) can be transformed at high frequency by adenovirus serotype 5 (Ad5). The CREF cells display a near diploid karyotype, do not crisscross at confluency, can be maintained at confluency for greater than 7 wk at 36 degrees C, and do not form macroscopic colonies when seeded in agar. Transformed cells are identified by their altered morphology and the presence of adenovirus DNA sequences in the transformants, which can be demonstrated by lysing cells directly on nitrocellulose filters and hybridizing with 32P-labeled Ad5 DNA (spot hybridization). The frequency of transformation at 36 degrees C is approximately equal to 2 x 10(-4) with wild-type Ad5 and approximately equal to 2 x 10(-3) with the temperature-sensitive mutant H5ts125. Southern blot hybridization analysis indicates that five out of six wild-type Ad5- and six out of six H5ts125-transformed CREF clones isolated at 36 degrees C contain the complete integrated Ad5 genome. Preliminary characterization of four transformed clones (two wild-type and two H5ts125) indicates that, even though transformation was done in CREF cells (a clonal cell line), they differ in their biological properties, such as saturation density and anchorage dependence.

The relationship between region E1a and E1b of human adenoviruses in cell transformation

Baby rat kidney (BRK) cells were transfected either with intact region E1 DNA of adenovirus type 5 (Ad5) or with mixtures of DNA fragments containing the separated E1a and E1b regions. The results showed that mixtures of regions E1a and E1b transform with a similar efficiency as intact region E1. DNA fragments containing region E1b alone have no detectable transforming activity in primary BRK cells nor in established rat cell lines. When region E1a and Ad5 was combined with region E1b and Ad12 complete transformation was also obtained. Characterization of the cell lines transformed by separated E1a and E1b regions have led to the following conclusions: (1) Expression of region E1b is not dependent on specific linkage to region E1a as it occurs in the intact E1 region. (2) Region E1b is normally expressed into the corresponding major adenovirus T antigens (65,000 and 19,000 Mr with region E1b of Ad5; 60,000 and 19,000 Mr with E1b or AD12). (3) Region E1b of Ad12 can be activated by region E1a of Ad5 indicating that the Ela regions of both serotypes are functionally similar in transformation. (4) Cell lines containing region E1b of Ad5 are weakly oncogenic in nude mice whereas cells containing E1b of Ad12 are highly oncogenic in nude mice, indicating that the degree of oncogenicity is determined by region E1b.

Expression of the gene encoding the adenovirus DNA terminal protein precursor in productively infected and transformed cells

The major product of in vitro translation of early RNA prepared from H5ts125-infected cells and selected by hybridization to adenoviral DNA fragments spanning the region from 14.7 to 31.5 map units had been shown to be identical to the 87-kilodalton terminal protein precursor. A 72- to 75-kilodalton polypeptide whose rRNA can be selected by DNA from this same region and made in the presence of anisomycin was indistinguishable from the 72-kilodalton single-stranded DNA-binding protein encoded by the region from 60.1 to 66.6 map units. The accumulation of cytoplasmic RNA sequences complementary to these l-strand genes under various conditions of infection and in certain lines of transformed cells has been investigated by solution hybridization of cytoplasmic RNA to the separated strands of restriction endonuclease fragments of adenoviral DNA. During the early phase, RNA sequences complementary to the region from 11.6 to 36.7 map units were present at a concentration of 10 to 60 copies per cell, regardless of the nature of the block used to inhibit viral DNA synthesis. By 24 h after infection in the absence of any such block, sequences complementary to the regions from 11.6 to 18.2 map units (IVa2) and from 18.6 to 36.7 map units (E2B) accumulated to concentrations of 4,800 and 280 copies per cell, respectively. The ratio of cytoplasmic E2A RNA sequences to E2B RNA sequences remained close to 10:1 throughout the time period investigated. Of the transformed cell lines which retained E2B DNA sequences that were examined, only the T2C4 line expressed these sequences in cytoplasmic RNA. The implications of these observations for regulation of expression of the adenoviral early l-strand genes are discussed.

Viral gene products in adenovirus type-2 transformed hamster cells

I have analyzed viral gene products expressed in five adenovirus type 2 (Ad2)- cytoplasmic, viral RNA which was selected by hybridization to cloned restriction endonuclease fragments of Ad2 DNA. Proteins synthesized in vitro were analyzed by electrophoresis in sodium dodecyl sulfate-polyacrylamide gels and compared with those directed by RNAs prepared from productively infected cells. The early regions E1 and E4 of adenovirus type 2 (Ad2) were found to be expressed in all of five Ad2-transformed hamster embryo cells lines. RNA transcribed from early region E2, which codes for the 72,000-molecular-weight (72K) DNA-binding protein was detected in cell line HE1 only, and early region E3 was expressed exclusively in cell line HE4. RNA transcribed from the region between approximately 12 and 35 map units, coding for immediate early (13.5K, 52/53K) and immediate early proteins (13.6K, 16K, 17K, 87K), as well as RNA from late genes, was not found in any of the cell lines HE1 to HE5 had electrophoretic mobilities similar to those programmed by RNA from productively infected cells.

Sequence organization of a viral DNA insertion present in the adenovirus-type-5-transformed hamster line BHK268-C31

The hamster cell line BHK268-C31 contains two large viral inserts which both include sequences from the left-hand end of adenovirus type 5 (Ad5) DNA. One of these viral inserts has been cloned in the lambda vector Charon 4A. Electron microscopic analysis and restriction enzyme mapping shows that the recombinant carries a 4.4-kb-long colinear segment of viral DNA, which is located between map positions 1.5 and 14.2 in the Ad5 genome. The junctions between viral DNA and flanking sequences have been sequenced and found not to show any specific features. One of the junctions is located in the E1a coding region, 573 bp from the left-hand end of the Ad5 genome, whereas the other junction is situated in the coding region for polypeptide IVa2. The promoter region as well as the cap site for the mRNAs from region E1a are thus missing from this insert and its role in viral transformation is unclear.

Adenovirus E1b-58kd tumor antigen and SV40 large tumor antigen are physically associated with the same 54 kd cellular protein in transformed cells

The adenovirus E1b-58kd tumor antigen has been detected in a physical association with a 54 kilodalton cellular protein in adenovirus-transformed mouse cells. Antibody specific for the E1b-58kd protein coimmunoprecipitates a 54 kd protein from transformed, but not from productively infected, cells. Monoclonal antibody specific for the cellular 54 kd protein coimmunoprecipitates the adenovirus E1b-58kd protein from transformed cell extracts. The same or closely related cellular 54 kd protein, associated with the adenovirus E1b-58kd protein, was present in the SV40 large T antigen-54 kd complex previously detected in SV40-transformed mouse cells. The identity of the 54 kd protein is based on the immunological specificities of the anti-54 kd monoclonal antibodies and partial peptide maps of the 54 kd protein associated with the adenovirus and SV40 tumor antigens. The adenovirus E1b-58kd-54 kd complex, like the SV40 large T antigen-54 kd complex, is heterogeneous in size or mass. While all of the cellular 54 kd protein in the adenovirus-transformed cell extract is found in a complex with the E1b-58kd protein, some of the viral 58 kd antigen is detected in a form not associated with the 54 kd protein. The fact that the adenovirus and Sv40 tumor antigens, both required for transformation, can be found in physical association with the same cellular protein in a transformed cell is a good indication that these two diverse viral proteins share some common mechanisms or functions.

Integration of foreign genome fragments into cells transformed or cotransformed with fragmented adenoviral DNA

The integration of DNA of highly oncogenic simian adenovirus type 7 (SA7) and non-oncogenic human adenovirus type 6 (Ad6) into the genome of newborn rat kidney cells transformed by fragmented DNA preparations was studied using reassociation kinetics and spot hybridization. Transforming DNA was fragmented with the specific endonuclease SalI (SA7) and BglII (Ad6). In contrast to the cell transformation by intact viral DNA, transformation by fragmented DNA resulted in integration into the cellular genome of not only the lefthand fragment with the oncogene but also of other regions of the viral genome. Additionally integrated fragments were stable and preserved during numerous passages of cells lines, although they were no expressed, at least in the case of the Ad6-transformed cell line. The integration of the fragments of SA7 DNA was accompanied by loss of 25-50% of the mass of each fragment. Adding the linear form of the pBR322 plasmid to the preparation of transforming Ad6 DNA also contributed to its cointegration into the genome of the transformed cell. This technique of cell cotransformation with any foreign DNAs together with the viral oncogens may be used as an equivalent of an integration vector for eukaryotic cells.

Control of adenovirus early gene expression: accumulation of viral mRNA after infection of transformed cells

We studied the accumulation of viral mRNA in the presence of inhibitors of protein synthesis in an adenovirus type 5-transformed cell line (line 293 cells). An analysis of the endogenous viral mRNA's and proteins revealed that only early regions 1A and 1B were expressed in uninfected 293 cells. However, viral mRNA's from early regions 2, 3, and 4, as well as mRNA's from early regions 1A and 1B, accumulated in 293 cells after infection with adenovirus type 2. Cells treated with anisomycin before infection showed a drastic enhancement of mRNA from early region 4 compared with drug-free controls, This increase in viral mRNA was detected by using filter hybridization, S1 endonuclease mapping, and in vitro translation. The rate of transcription of early region 4 nuclear RNA also increased significantly in the presence of anisomycin. In contrast to the results with early region 4, the levels of cytoplasmic mRNA's from early regions 2 and 3 did not increase in cells treated with inhibitors. These results suggested that multiple virus encoded controls operate on the early regions of the adenovirus genome.

Kaposi's sarcoma. IV. Detection of CMV DNA, CMV RNA and CMNA in tumor biopsies

In order to determine whether human cytomegalovirus- (CMV) DNA homologous sequences as well as CMV-specific RNA(s) and antigen(s) exist in tumor biopsies of Kaposi's sarcoma (KS) DNA-DNA reassociation, RNA-DNA in situ cytohybridization and anticomplement immunofluorescence test (ACIF) tests were applied. Three of 10 DNAs extracted from Kaposi sarcoma biopsies contained DNA sequences homologous to radioactively labelled human CMV DNA probe. The amount of CMV DNA in these sarcoma tissues was calculated to range from 0.7 to 1 genome equivalent per diploid cell. The presence of virus-specific RNA was also demonstrated in section from five of 10 tumor biopsies. CMV-determined nuclear antigen(s) CMNA) present in variable degrees were also demonstrated. In contrast, we could not detect any herpes simple virus type II (HSV-2) or Epstein-Barr virus (EBV) DNA sequences in DNA of these tumor biopsies. Furthermore, there were no detectable HSV-2 or EBV-specific RNA or virus-specific antigens in sections of these biopsies. These results provide new lines of evidence for the relationship between CMV and Kaposi's sarcoma.

Mechanism of activation of early viral transcription by the adenovirus E1A gene product

Results obtained with certain adenovirus host range mutants (dl312 and hr1) have demonstrated that a functional viral E1A gene is required for the accumulation of early viral mRNA. In this report, it is demonstrated that the failure to accumulate early viral mRNA in E1A mutant-infected cells is due to a lack of transcription of the appropriate transcription units. The E1A gene product appears to enhance transcription of the early transcription units since there is not an absolute requirement for its function. At low multiplicities of infection, transcription from the early transcription units can be detected but at a later time. By increasing the multiplicity of infection, this show activation of early transcription can be enhanced, even though at this multiplicity all input viruses are still mutant. It was also found that early transcription could be detected in mutant-infected cells if the virus was added to cells in which protein synthesis was inhibited prior to infection. Thus inhibition of cellular protein synthesis can mimic the function of the E1A gene. It is proposed that the role of the E1A gene product in the activation of early viral transcription is to inactivate a cellular factor that prevents transcription from the early viral promoters.

Arrangement of integrated viral DNA sequences in cells transformed by adenovirus types 2 and 5

The organization of the viral DNA sequences in 15 adenovirus-transformed cell lines was analyzed by the Southern blotting procedure. The site of adenovirus integration in the cellular genome was found not to be unique, and the viral DNA sequences involved in integration were not confined to a specific region of the adenovirus genome. Several cell lines showed simple integration patterns that demonstrated the presence of large continuous stretches of viral DNA. In four cell lines, containing sequences from both molecular ends of the viral genome, the left- and right-hand-terminal sequences appeared to be linked to each other.

Complementation of adenovirus type 5 host range mutants by adenovirus type 12 in coinfected HeLa and BHK-21 cells

We have studied the ability of adenovirus type 12 (Ad12) to complement the Ad5 transformation-defective host rang (hr) mutants during infection of human cells (HeLa) or hamster cells (BHK-21). The group I mutant hr3 (mapped within 1.3 to 3.7 map units), which is incapable of synthesizing viral DNA, was complemented for both DNA synthesis and infectious virus production in nonpermissive HeLa cells during coinfection with Ad12. Similarly, the group II mutant hr6 (6.1 to 9.4 map units), which does synthesize DNA, was also shown to be complemented for virus production. When the host cells were BHK-21, an established hamster cell line that is permissive for Ad5 but nonpermissive for Ad12 DNA synthesis and virus production, coinfection with Ad5 and Ad12 did not overcome the block to Ad12 DNA synthesis. Coinfection of BHK-21 cells with Ad12 and either hr3 or hr6 leads to the complementation of only the group I mutant (hr3). The inability of Ad12 to complement hr6 in BHK-21 cells may be due to the failure of Ad12 to express an early gene product from the region corresponding to early region 1B (4.5 to 11 map units) Ad5 where hr6 and the other group II mutations are located.

Protein kinase activity immunoprecipitated from adenovirus infected cells by sera from tumor-bearing hamsters

Earlier, we reported (N. J. Lassam, S. T. Bayley, F. L. Graham, and P. E. Branton, Nature (London) 277:241-243, 1979) detecting protein kinase activity when cytoplasmic extracts of human adenovirus type 5 (Ad5)-infected KB cells immunoprecipitated with 14b antitumor serum directed against the transforming proteins of Ad5, were incubated with [gamma-32P]ATP. Here we show that in the in vitro assay this kinase phosphorylated both the heavy chain of immunoglobulin G and polypeptide than comigrated on sodium dodecyl sulfate gels with the 58,000-dalton Ad5 antigen. It also phosphorylated added histone H3. Evidence is presented that the protein kinase activity found with extracts from Ad5-infected cells is not due to nonspecific trapping of cellular enzymes in immune complexes, but to an enzyme which is distinct from kinases detected at background levels in controls. Serine and threonine were the major phosphorylated amino acids, and essentially no phosphotyrosine was detected. Protein kinase activity detected in Ad12-infected cells immunoprecipitated by an antiserum derived from hamsters bearing Ad12-induced tumors appeared to be immunologically distinct from that immunoprecipitated from Ad5-infected cells by 14b serum.

Regulation of integrated adenovirus sequences during adenovirus infection of transformed cells

A human cell line (293) transformed by adenovirus type 5 encodes mRNA's and proteins from the early region 1 (E1) of the viral genome. These products correspond to those synthesized early after adenovirus infection of normal cells. This pattern of expression is different from that observed at later times in the lytic cycle. We have determined whether integrated sequences can undergo the early-late transition during infection of transformed cells. Cultures of 293 cells were infected with mutants of adenovirus type 5 that have deletions in EI genes. In such infections, the integrated sequence complements the deletion mutants so that viral DNA replication, late mRNA and protein synthesis, and viral assembly occur. Because the infecting genomes lack EI sequences, the products synthesized from the integrated DNA could be analyzed. In contrast to the early-late transition that occurs with EI DNA in free viral genomes, the pattern of mRNAs and proteins made from the integrated sequences was restricted to the early pattern. Assuming that the viral sequences in 293 cells have not become altered during the history of the cells, our results suggest that regulation of integrated adenovirus genes may not be determined exclusively by nucleotide sequence recognition. Apparently, during infection certain factors prevent the integrated viral genes from responding to the regulatory signals which control late expression from free EI DNA. The distinction between integrated and free viral sequences might reflect the different fates of viral and host transcripts during the lytic cycle of adenovirus.

Transcription and RNA processing by the DNA tumour viruses

Messenger RNA synthesis by the DNA tumour viruses proceeds by a complex but versatile series of transcription and RNA processing steps. The major mechanistic features of this pathway are probably very similar to those used by the animal cell host itself. The viruses have, however, evolved intricate arrangements of protein coding sequences and sites for RNA initiation, polyadenylation and splicing which allow them to use their genetic information to maximum advantage.

Control of adenovirus early gene expression: a class of immediate early products

We have identified the viral mRNAs present in cells in which protein synthesis has been stringently inhibited prior to infection with adenovirus type 2. These species presumably represent the subset of viral mRNAs that are "immediate early" products, requiring only host cell genes of their expression, and they do not include any of the conventionally recognized early mRNAs. Treatment of cells with 100 microM anisomycin inhibits 99.6% of protein synthesis and substantially depresses (by 20--200 fold) the levels of the conventional early mRNAs from regions E1A, E1B, E2, E3 and E4. Also depressed are species encoding an 87K protein (11.6--31.5 map units) and a 13.6K protein (encoded a short distance to the right of 21.5 map units). The only mRNAs not depressed by this treatment are an mRNA for a 13.5K protein encoded between 17.0 and 21.5 map units, and the mRNA for the late 52,55K protein encoded between 29 and 34 map units, which is also present in small amounts at early times. Further proof that production of the mRNA for the immediate early 13.5K protein is independent of E1A gene function is provided by the observation that it can be detected in cells infected with the E1A deletion mutant dl312.

Predicted structure of two adenovirus tumor antigens

Early adenovirus type 2(Ad2) mRNA sequences have been cloned by using the pBR322 plasmid as a vector. Two clones that include sequences from region E1B were identified and their DNAs were characterized by hybridization, restriction enzyme cleavage, and DNA sequence analysis. The results showed that the clones were derived from two different spliced mRNAs. By combining our results with the established DNA sequence for region E1B of the closely related adenovirus type 5[Maat, J., van Beveren, C.P. & van Ormondt, H. (1980) Gene, in press] it was possible to deduce the structure of a 13S and a 22S mRNA. The two mRNAs differ from each other by the size of their intervening sequences. If translation starts at the first AUG following the cap, the 22S mRNA encodes a Mr 67,000 polypeptide that is terminated by a UGA stop codon located immediately before the splice, whereas the 13S mRNA encodes a Mr 20,000 polypeptide that is translated in different reading frames before and after the splice. The Mr 20,000 and 67,000 polypeptides correspond in molecular weight to two proteins that invariably are precipitated from infected cell extracts by antisera from animals carrying adenovirus-induced tumors.

Structure and gene organization in the transformed Hind III-G fragment of Ad12

The nucleotide sequence of the transforming Hind III-G fragment of Ad12 DNA which encompasses the left 6.8% of the genome has been determined. The fragment was 2320 nucleotides long, and contained a GC cluster at positions 126-155 and a region extremely rich in AT at positions 1098-1142 (number from the leftmost end). Possible coding regions for the two transforming gene products were assigned. The predicted coding region for T antigen g is positions 502-1069 and positions 1144-1373, which are joined by splicing (266 amino acid residues, 30 kd), and that for T antigen f is positions 1845-2126 (94 amino acid residues, 10 kd). The sequence of the Hind III-G fragment was compared with that of the transforming DNA fragment of Ad5 which encompasses the left 8.0% of the genome (2809 nucleotides). There are several discrete regions with significant sequence homology. The comparison suggests that the regions in the left two thirds of the Ad5 and Ad12 transforming DNA fragments (map units 0-4.7% in Ad5 and 0-4.4% in Ad12) bear some resemblance in their gene organizations, and code for proteins containing structurally homologous regions.

Two tumor antigens and their polypeptides in adenovirus type 12-infected and transformed cells

A tumor (T) antigen, designated T antigen g, was visualized as fine fluorescent granules in nuclei of adenovirus type 12 (Ad12)-infected cells by immunofluorescence with sera from rats bearing HY cell tumors (H sera). HY cells are rat cells incompletely transformed by the Acc I-H endonuclease fragment (0-4.7 map units) of Ad12 DNA. The antigen is different from the usually described T antigen, designated T antigen f, which is visualized as fluorescent flecks or filaments in both nucleus and cytoplasm of Ad12-infected cells when tested with narrowly reacting T sera. Extracts of [(35)S]methioninelabeled infected cells were immunoprecipitated with H sera, and the resultant precipitate was analyzed by the two-dimensional gel electrophoresis technique of O'Farrell. The autoradiogram showed the presence of a cluster of several polypeptides (M(r) 35,000-40,000, pI 5.0-5.5) that was absent in extracts of mock-infected cells. A similar autoradiogram of infected cells analyzed with narrowly reacting T sera showed the presence of a small polypeptide (M(r) 10,000, pI 6.4), that was absent in extracts of mock-infected cells. The results show that M(r) 35,000-40,000 polypeptides are components of T antigen g and a M(r) 10,000 polypeptide is a component of T antigen f. Ad12-transformed cells showed a similar result. T antigen g was present and T antigen f was absent in HY cells. Both T antigen g and T antigen f were present in CY cells, which are rat cells completely transformed by the EcoRI-C endonuclease fragment (0-16 map units) of Ad12 DNA. The possible functions of these proteins are discussed.

The boundaries and copy numbers of Ti plasmid T-DNA vary in crown gall tumors

The Ti plasmid DNA maintained in octopine-type crown gall tumor lines is variable, but always includes at least part of the Ti plasmid that maps over the region of Hind III fragment 1 of pTiB6-806. The right-hand boundary of transferred DNA (T-DNA) varies considerably among the three independent tumor lines examined; the left boundary was not located definitively. The T-DNA of two sibling clones of the same tumor line, E1 and E9, appears identical. The copy number of T-DNA in E9 tumor DNA appears higher for the right end (about 30 copies) than for the left end (approximately 1 copy).

Analysis of viral DNA sequences in hamster cells transformed by herpes simplex virus type 2

Herpes simplex virus type 2 (HSV-2) DNA was treated with four restriction endonucleases (EcoRI, HincIII, Bgl II, and Xba I) and eight fragments were purified and labeled with 32P in vitro. The kinetics of renaturation of each of the fragments was measured in the presence of DNA extracted from 333-8-9, a hamster cell line transformed by UV light-inactivated HSV-2 strain 333, and from a series of cloned derivatives and their tumor lines. All of the lines examined contained a partial set of viral sequences present at only a few copies per cell. Passage of the cell lines in tissue culture or in animals resulted in partial loss of viral DNA. Two blocks of sequences were present in most of the lines examined; those mapping at positions 21--33 of the HSV-2 genome were detected in seven of seven cell lines tested and those at positions 60--65 were detected in six of eight. Other sequences from the L component can also be present in the DNA of HSV-2-transformed hamster cells.

Different nucleosome structures on transcribing and nontranscribing ribosomal gene sequences

Monomeric DNA lengths from Physarum nuclear chromatin occur in two subunit forms which differ from each other and from higher oligomers of nucleosomes in content of transcribed ribosomal DNA sequences. Labeled DNA restriction fragments from ribosomal RNA coding regions reanneal most rapidly with DNA from a monomeric subunit fraction. A particles, isolated from growing plasmodia and containing 144 base pairs of DNA in an extended conformation. Higher oligomers of nucleosomes are depleted in sequences from transcribing gene regions but are enriched in sequences from the nontranscribed central spacer of the ribosomal DNA palindrome. Nucleosome configuration on two 26S gene intervening sequences resembles that on adjacent coding regions.

Tumor antigens of human Ad5 in transformed cells and in cells infected with transformation-defective host-range mutants

We have studied the polypeptides associated with the expression of the transforming region of the Ad5 genome by immunoprecipitating antigens (using the double antibody and protein A-Sepharose techniques) from cells infected with wild-type (wt) Ad5 or transformation-defective host range (hr) mutants and from cells transformed by Ad5. Three different antisera were used: P antiserum specific for early viral products (Russell et al., 1967) and two different hamster tumor antisera. Immunoprecipitation of antigens from wt-infected KB cells followed by SDS-polyacrylamide gel electrophoresis of precipitated proteins revealed that a major polypeptide having a molecular weight of approximately 58,000 was detected with all three antisera and with both the double antibody and the protein A-Sepharose techniques, while P antiserum also precipitated polypeptides of molecular weights 72,000, 67,000 and 44,000, which probably represent the DNA binding protein and related polypeptides, respectively. With the double antibody technique, in addition to the proteins mentioned above, P antiserum and the hamster tumor antisera precipitated a 10,500 dalton polypeptide which was not detected when the protein A-Sepharose procedure was used. Using either the double antibody or the protein A-Sepharose technique, we found that hr mutants from complementation group II failed to induce the synthesis of the 58,000 dalton protein, whereas mutants from complementation group I produced normal or near normal amounts. Using the double antibody technique, we found that the 10,500 dalton protein was absent or made in reduced amounts by group I mutants. A 58,000 dalton protein was detected in a number of different Ad5-transformed cell lines, including the 293 human line, the 14b hamster line and several transformed rat cell lines. This observation and the fact that transformation negative group II mutants fail to induce the synthesis of a 58,000 dalton polypeptide suggest that this protein is one of the Ad5-specific products necessary for cell transformation.