Spontaneous mutations in the human gene for p53 in recombinant adenovirus during multiple passages in human embryonic kidney 293 cells

BBRC Biochemical and Biophysical Research Communications 300 (2003) 448–456 www.elsevier.com/locate/ybbrc

Spontaneous mutations in the human gene for p53 in recombinant adenovirus during multiple passages in human embryonic kidney 293 cells Hideyo Ugai,a Erika Suzuki,a,1 Kumiko Inabe,a Takehide Murata,a Hirofumi Hamada,b and Kazunari K. Yokoyamaa,* a

Gene Engineering Division, Department of Biological Systems, BioResource Center, RIKEN (The Institute of Physical and Chemical Research), 3-1-1 Koyadai, Tsukuba, Ibaraki 305-0074, Japan b Department of Molecular Medicine, Sapporo Medical University, Sapporo, Hokkaido 060-8556, Japan Received 1 November 2002

Abstract Infectious recombinant adenovirus (rAd) is usually produced in human embryonic kidney 293 cells that harbor the E1 gene and rAd has been shown to be an efficient tool for gene transfer both in vivo and in vitro. It also has considerable potential in human gene therapy. However, rates of spontaneous mutations in genes introduced into host cells after multiple passages remain to be clarified. We have characterized the spontaneous mutation of genomes derived from human adenovirus type 5 (Ad5) and of human p53-rAd during multiple passages by two different methods, namely, a plaque assay and a molecular cloning assay, with subsequent direct nucleotide sequencing. Using the plaque assay, we found no mutations in the E1A and p53 genes derived from infectious Ad5 and p53-rAd, respectively. By contrast, we found spontaneous mutations in the E1A gene of Ad5, with a mutation rate of 9:28
108 per base pair per plaque, in the molecular cloning assay. The rate of mutation of the p53 gene of p53-rAd, as determined by the molecular cloning assay, ranged from 1:50
107 to 3:25
107 per base pair per passage. The mutations in the p53 gene of p53-rAd were localized mainly in the transcriptional activation domain, the SH3 domain, and the regulation domain and they were rarely found in the DNA-binding domain, which is a major site for mutations in human cancers. Our results indicate that multiple passages can generate a heterogeneous population of p53-rAd and that the molecular cloning assay is an efficient technique with which to search for mutations in the genome of p53-rAd that cannot be detected by a plaque assay. Ó 2002 Elsevier Science (USA). All rights reserved.

The gene encoding p53 is the gene that is most frequently mutated in human cancers and loss of p53 enhances the risk of developing malignancies [1–3] since p53 is a tumor-suppressor protein that is involved in arrest of the cell cycle or apoptosis in response to DNA damage [1,4]. These roles of p53 are based on its ability to function as a sequence-specific transcription factor [1]. Recent attempts at gene therapy for cancer have used the gene for p53 as a target [5], and the forced expression of wild-type p53 (p53WT), with or without * Corresponding author. Fax: +81-298-36-9120. E-mail address: [email protected] (K.K. Yokoyama). 1 Present address: Department of Pharmacology, Southern Illinois University School of Medicine, Springfield, IL 62794, USA.

DNA-damaging reagents and ionizing radiation, has been shown to induce the activation of apoptotic signals in cancer cells [1,4]. An adenovirus (Ad) serotype 5 vector, which corresponds to an Ad genome without the gene for E1, has often been used for delivery of exogenous genes [6,7]. Such Ad vectors can be prepared at much higher titer (1011 pfu/ml) than can retroviral vectors (106 –107 pfu/ ml) because they are non-enveloped viruses, remaining stable during concentration and purification [8]. Recombinant Ads (rAds) are derived from replicationdeficient Ads in which the gene for E1 is replaced by foreign genes [9–11]. These rAds can be propagated in human embryonic kidney (HEK) 293 cells, in which the left arm [nucleotides (nt) 1–4344; GenBank Accession

0006-291X/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. doi:10.1016/S0006-291X(02)02852-8

H. Ugai et al. / Biochemical and Biophysical Research Communications 300 (2003) 448–456

Nos. M73260 and M29978] of the Ad5 genome has been integrated into the genome of 293 cells. One of the features of rAd is that replication-competent adenovirus (RCA) is generated over the course of more than 10 passages in HEK 293 cells by heterogeneous infection as a result of homologous recombination between the E1 region in the host genome and a residual small region in the rAd genome [12–14]. It is necessary to monitor the emergence of RCA in preparations of infectious rAd derived from HEK 293 cells. It is also necessary to monitor mutations in the infectious and non-infectious rAds, such as single base substitutions, insertions, and deletions of nucleotides, to ensure the safety of laboratory and clinical studies. However, the rates of mutations in introduced genes and in the rAd genomes produced in HEK 293 cells after multiple passages have not been characterized. An understanding of mutations during the replication of infectious Ad and rAd is essential for studies of the safety, both in the laboratory and in clinical experiments, of Ad vectors. In the present study, we determined the sites and frequencies of mutations, such as substitutions, insertions, and deletions of nucleotides, in infectious and non-infectious populations of Ad5 and p53-rAd genomes prepared from single plaques and in p53-rAd genomes generated during multiple passages. We found an unexpectedly high rate of mutation of the p53 gene in the p53-rAd genome and in the vector sequences of rAd. Thus, this report describes our initial attempts at the molecular characterization of spontaneous mutations that occur during multiple passages of rAd so that it can be more safely used as a tool for gene transfer in vivo and in vitro.

Materials and methods Cells, cell culture, adenoviruses, and titration. Human embryonic kidney cells (HEK 293 cells; ATCC, Rockville, MD, USA) and human lung epithelial cells (A549 cells; ATCC) were cultured in DulbeccoÕs modified EagleÕs medium (DMEM; Nissui, Tokyo, Japan) and used for the production of Ad5 and rAd, respectively [15,16]. Human adenovirus type 5 (Ad5) was purchased from ATCC (Manassa, VA, USA) and a fiber-modified p53-rAd, AxCAhp53-F/K20 [the recombinant adenovirus was provided with a titer of 2:53
108 plaque-forming units (pfu)/ml], was prepared as described elsewhere [17]. The AxCAhp53-F/K20 encoded a gene for the fiber protein of Ad5 that is followed by alanine-serine (AS) linkers and 20 lysine residues at the carboxyl terminus. The rAd that included the gene from Herpes simplex virus for thymidine kinase (HSVtk) plus the gene for wild-type (WT) fiber protein, AxCAHSVtk (RDB No. 1429), was provided by the RIKEN DNA Bank (Tsukuba, Japan). A replication-competent adenovirus (RCA) was obtained by multiple passages of AxCAHSVtk by standard methods [14]. Titers of Ad5 and rAd were determined by a modified version of the end-point cytopathic-effect assay using HEK 293 cells [18]. Construction of a wild-type fiber-expressing p53-recombinant adenoviral vector. Human WT and mutant p53 genes were cloned at the

449

SwaI site of pAxCAwt ([9]; RDB No. 1678; RIKEN DNA Bank), which includes the genome of Ad5 and a CAG promoter, to generate pAxCAhp53-F/WT. We then transfected HEK 293 cells with pAxCAhp53-F/WT by the COS-TPC method [9] and WT fiber-expressing p53-rAd, AxCAhp53-F/WT, was generated by homologous recombination in the cells. It was isolated, after screening, from a single plaque. Production of high-titer preparations of adenoviruses by multiple passages. Fig. 1 shows the experimental strategies for preparation and assays of Ads. The Ads isolated from a single plaque were used to infect host cells at 80% confluency with a multiplicity of infection (MOI) of 10 pfu/cell. After the cytopathic effects (CPE) of the virus had become apparent, cells were harvested, with the culture medium, by pipetting. Suspensions of cells were sonicated with a Bioruptor (Cosmo Bio, Tokyo, Japan) and lysates were centrifuged at 9300g for 5 min at 4 °C. Supernatants were used as solutions of virus for the next infection and stored at )80 °C until the next passage. This procedure was repeated 10 times to increase the titer of Ads. After the 10th passage, the viral particles were purified by ultracentrifugation in CsCl as described elsewhere [18]. The purified Ads were used as working stocks for subsequent experiments. The titers of Ad5, AxCAhp53-F/WT, AxCAhp53-F/K20, and AxCAHSVtk were 1:48
108 , 6:40
109 , 4:24
109 , and 5:08
109 pfu/ml, respectively. Isolation of adenoviruses by the plaque assay. Adenoviruses were isolated from a single plaque as follows. A working stock (see above) after the 10th passage was diluted from 109 to 103 and diluted samples were used to infect host cells. After infection of cells for 1 h, a mixture of 3 ml DMEM (Nissui, Tokyo, Japan) that contained 5% FCS (Gibco-BRL, Life Technologies, Rockville, MD, USA) and 2% Seqplaque (BioWhittaker Molecular Applications, Rockland, ME, USA) was laid over the infected cells. Four days later, 3 ml DMEM that contained 5% FCS and 2% Seqplaque was laid over the cells [19]. When single plaques became visible, Ads were picked up from individual single plaques by pipetting and suspended in 200 ll PBS that contained 10% glycerol. These suspensions of Ads were used for detection of mutations in the Ad genome. Analysis of target DNAs by a molecular cloning assay. For the molecular cloning assay, we extracted the genomic DNA from Ads that had been prepared as described above. We amplified genes by PCR using Ad genomic DNA as template and then we cloned the amplified DNA fragments into the EcoRV site of pBluescript II KS()) (Stratagene, La Jolla, CA, USA) and isolated cloned DNAs from Escherichia coli strain XL1-Blue. We cloned the genomic DNA of AxCAhp53-F/K20 into the blunt-ended BamHI site of SuperCos1 (Stratagene) to generate pAxCAhp53-F/K20 for detection of mutations. Clones were sequenced to confirm the sites of mutations. We used KOD Plus DNA polymerase (Toyobo, Tokyo, Japan) rather than Taq DNA polymerase for the amplification of DNA because of the 82fold higher fidelity of this enzyme, as determined by the method of Mo et al. [20]. We amplified the target genes for DNA sequencing and cloning using Ad genomic DNA as template and the p53-primer pairs 50 -GGCTTCTGGCGTGTGACCGGC-30 (forward direction) and 50 -CAGAGGGAAAAAGATCTCAGTGG-30 (reverse direction), the E1A-primer pairs 50 -ATGAGACATATTATCTGCCACGGAGGTG TTATTAC-30 [forward direction, corresponding to nucleotide (nt) positions 560–595] and 50 -TTATGGCCTGGGGCGTTTACAGC TCAAGTCCAAAG-30 (reverse direction, nt 1545–1511), and the VAprimer pairs 50 -CGGGATCCGGTCGGGACGCTCTGGCCGGT CAGGCG-30 (forward direction, nt 10,545–10,571) and 50 -CGG GATCCTGGGAAAAGCAAAAAAGGGGCTCGTCCC-30 (reverse direction, nt 11,049–11,021; GenBank Accession Nos. M73260 and M29978). Amplification by PCR was allowed to proceed for 30 cycles of denaturation for 30 s at 96 °C, annealing for 15 s at 60 °C, and extension for 1 min/kb at 68 °C. All products of PCR were purified with a QIAquickTM spin column (Qiagen GmbH, Hilden, Germany) or a 96well filter system (Millipore, Bedford, MA, USA). Nucleotide sequences of the products of PCR were determined directly with the

450

H. Ugai et al. / Biochemical and Biophysical Research Communications 300 (2003) 448–456

Fig. 1. Schematic representation of the production of high-titer preparations of adenoviruses by multiple passages and strategies for detection of mutations in the genome. Recombinant adenoviruses and wild-type adenovirus type 5 were used at an MOI of 10 to infect 293 cells and A549 cells, respectively. Working stock solutions of viruses were prepared as described in Materials and methods.

primers used for PCR, a DYEnamic ET terminator kit (Amersham Biosciences, NJ, USA) and an automated sequencer (ABI PRISM 377; Applied Biosystems, Foster City, CA, USA; and MegaBase 1000, Amersham Biosciences). All primers were synthesized by Invitrogen (Carlsbad, CA, USA). The products of PCR were also cloned into pBluescript II KS ()) and sequenced. Calculation of frequencies of mutation and rates of mutation. We calculated the mutation frequency using the following formula: mutation frequency ¼ number of mutants per number of clones of which the nucleotide sequence was determined. We calculated the mutation rate using the following formula: mutation rate ¼ mutation frequency per total genome of adenovirus (35,935 bp) per passage.

Results Production of high-titer preparation of adenovirus by multiple passages Human adenovirus (Ad) and recombinant adenovirus (rAd) were isolated from individual plaques derived from single viral particles. The Ads were used at an MOI of 10 pfu/cell to infect human embryonic kidney (HEK) 293 cells. Then cells were harvested and progeny virus

H. Ugai et al. / Biochemical and Biophysical Research Communications 300 (2003) 448–456

was recovered from the infected cells. We repeated this process through 10 passages to increase the titer to a level sufficient for delivery of exogenous genes in vitro. After 10 passages we purified Ads by ultracentrifugation in CsCl. The titers of Ad5, AxCAhp53-F/WT, AxCAhp53-F/K20, and AxCAHSVtk were 1:48
108 , 6:40
109 , 4:24
109 , and 5:08
109 pfu/ml, respectively. The solutions of purified viral particles were used as working stocks for comparison of rates of spontaneous mutations in genomes of p53-rAd (Fig. 1). Determination of mutation rates by the plaque assay The working stocks of Ads were diluted and used to infect host cells under conditions for the plaque assay, as described above, for isolation of individual infectious clones. The plaques of Ad5 were approximately 2 mm in diameter and those of rAds were less than 1 mm in diameter. We picked up individual plaques and amplified the E1A gene of Ad5 and the p53 genes of rAds by PCR. Then we sequenced the products of PCR directly. The mutation frequencies that we determined after analyzing the nucleotide sequences of the target genes of Ads are summarized in Table 1. We detected no mutations in genes for E1A in Ad5 and in p53 genes in AxCAhp53-F/ WT and AxCAhp53-F/K20 by the plaque assay. Thus, we were unable to calculate mutation rates. The possibility remained that mutations might occur during the formation of single plaques derived from single viral particles. Therefore, we used a different approach to isolate the E1A gene of Ad5 E1A, namely, a molecular cloning assay, and then we determined nucleotide sequences (Fig. 1). We amplified the E1A gene by PCR and cloned the products into pBluescript II KS ()). We isolated 60 clones and sequenced them to identify sites of mutations. One of the 60 clones was

451

mutated at position 1003 (G ! A) and one was mutated at position 1316 (T ! C; positions correspond to those in the Ad5 genome: GenBank Accession Nos. M73260 and M29978). We calculated the mutation frequency and the mutation rate for the E1A gene during the formation of a single plaque. The mutation frequency was 2/60 (3:33
102 ) and the mutation rate was 9:28
108 per base pair per plaque (Table 2). We examined the fidelity of KOD Plus DNA polymerase during amplification of E1A genes by PCR and found no mutations in the E1A gene (data not shown). We also failed to detect any cloning artifacts in a cloned E1A gene that had been amplified by PCR (data not shown). Thus, the mutations in the E1A gene would correspond to mutations in the individual plaques. Determination of mutation rates by the molecular cloning assay We next examined mutations in the p53-rAd genome by the molecular cloning assay (Fig. 1). After we had extracted p53-Ad genomic DNA from the above-described working stock solutions, we amplified the p53 gene by PCR using the genomic DNA from AxCAhp53F/WT and AxCAhp53-F/K20 as templates. The products of PCR were cloned in pBluescript II KS ()). We also cloned the genomes of AxCAhp53-F/K20 into SuperCos1. Then we sequenced all the clones to identify mutations. The mutation frequency and the mutation rate for the p53 gene of AxCAhp53-F/WT are shown in Table 3 and the sites of mutations in the p53 gene are summarized in Fig. 2 and Table 4. The p53 gene isolated from AxCAhp53-F/WT revealed the presence of seven mutants among the 60 clones examined. One of the seven mutants, p53P128T, was found to have the same point

Table 1 Frequencies of mutations in genes of wild-type adenovirus type 5 and recombinant adenoviruses isolated by the plaque assay Virus

Target gene

Target size (bp)

GC content (%)

Mutation frequency

Ad5 AxCAhp53-F/WT AxCAhp53-F/K20

E1A p53 p53

500 600 600

49.1 57.5 57.5

0/92 0/83 0/82

Table 2 Frequencies of mutation and rates of mutation in the E1A gene in Ad5 plaques, as determined by the molecular cloning assay Target genea

Nucleotide changeb

Mutationc

Mutation frequency

Mutation rated

E1A

1003G ! A 1316T ! C

L214P Silent

2/60 (3:33
102 )

9:28
108

a

The E1A gene in Ad5, isolated by the plaque assay, was amplified by PCR and the products of PCR were cloned in pBluescript II KS()) and sequenced. b Position 1 refers to the 50 end of the Ad5 genome. c Position 1 refers to the site of initiation of translation of the E1A gene. d The mutation rate was calculated as follows: mutation rate ¼ mutation frequency/total genome (bp)/plaque.

452

H. Ugai et al. / Biochemical and Biophysical Research Communications 300 (2003) 448–456

Table 3 Frequencies of mutations and rates of mutations for target genes of recombinant adenoviruses, as detected by the molecular cloning assay Virus

Target gene

Size of target gene (bp)

GC content (%)

Mutation frequency

Mutation ratea

AxCAhp53-F/WT AxCAhp53-F/K20

p53 p53 p53 VA E1A

1459 1459 1459 505 986

57.5 57.5 57.5 62.8 51.8

7/60 ð1:17
101 Þb 12/115 ð1:04
101 Þb 2/37 ð5:40
102 Þd 3/41 ð7:32
102 Þb 1/20 ð5:00
102 Þb

3:25
107 c 2:90
107 c 1:50
107 c 2:04
107 c 1:39
107 c

RCAe a

The mutation rate was calculated as follows: mutation rate ¼ mutation frequency/total genome (bp)/passage. The p53 gene was amplified by PCR with the rAd genome as template and cloned in pBluescript II KS()). c Statistical analysis was performed by ‘‘non-repeated measures ANOVA.’’ There were no significant differences among results (P < 0:01). d The rAd genomes were cloned into the cosmid SuperCos 1. e RCAs (replication-competent adenoviruses) were isolated from working solution of AxCAHSVtk. b

Fig. 2. Schematic representation of sites of mutations in the human p53 gene. (A) Sites of mutations in p53 genes isolated from the p53-recombinant adenoviral vector AxCAhp53-F/WT and AxCAhp53-F/K20. The numbers refer to positions in the nucleotide sequence. Mutations are indicated by black arrows (AxCAhp53-F/K20) and red arrowheads (AxCAhp53-F/WT). UTR indicates an untranslated region. Position 1 refers to the A of the ATG triplet that is the site of initiation of translation. (B) Functional domains of human p53. Exons and introns are indicated as thick and thin lines, respectively. NES, Nuclear export signal; and NLS1, NLS2, and NLS3, nuclear localization signals 1, 2, and 3, respectively. (C) The numbers refer to amino acid residues. FS indicates a frame-shift mutation and SM indicates a silent mutation.

mutation as that detected by the cloning assay of AxCAhp53-F/K20 genomes (Table 4). All the mutations that we identified were single-base substitutions; five of them were missense mutations and three of them were silent mutations. Six mutations were

localized in the 50 region of the p53 gene (see Fig. 2 and Table 4). The mutation frequency of the p53 gene was calculated as 7/60 (1:17
101 ) and the mutation rate was 3:25
107 per base pair per passage (Table 3). We identified 12 mutants among the 115 cloned p53 genes

H. Ugai et al. / Biochemical and Biophysical Research Communications 300 (2003) 448–456

453

Table 4 Sites of mutations in the p53 gene of recombinant adenoviruses, as detected by the molecular cloning assay Virus

Assay

Name of clone

Nucleotide change

Mutation

Domain

AxCAhp53-F/WT

PCR cloning

p53M1I

3G ! T

M1I

Transcriptional activation domain

p53L14P

41C ! T

L14P

Transcriptional activation domain

p53-68SM

204G ! A

Silent

SH3 domain

p53P87L

260C ! T

P87L

SH3 domain

p53T102I

305C ! T

T102I

DNA-binding domain

p53P128T

382C ! A

P128T

DNA-binding domain

1095C ! T

Silent

Regulatory domain DNA-binding domain

AxCAhp53-F/K20

Cloning PCR cloning

p53-250SM

750C ! A

Silent

p53-42SM

126T ! C

Silent

DNA-binding domain

p53del1-293

nt-93 (del978)

Multiple deletions

Amino terminal deletion

p53-30FS

90Cdel

Frame shift

Transcriptional activation domain

p53P89H

266C ! A

P89H

SH3 domain

p53S96P

286T ! C

S96P

SH3 domain

p53G105D

314G ! A

G105D

DNA-binding domain

p53P128T

382C ! A

P128T

DNA-binding domain

1095C ! T

Silent

Regulatory domain

484A ! G

I162V

DNA-binding domain

797G ! A

G266E

DNA-binding domain

1246T ! A

—

Untranslated region

p53S240N

719G ! A

S240N

DNA-binding domain

p53-316SM

948C ! T

Silent

Tetramerization domain

p53-360FS

1078Gdel

Frame shift

Regulatory domain

p53G374C

1120G ! T

G374C

Tetramerization domain

p53-1244FS

1244Tdel

Frame shift

Untranslated region

p53del1-293

nt )93 (del978)

Multiple deletions

Amino terminal deletion

p53I162V, G266E

from AxCAhp53-F/K20 by PCR cloning (Table 3). Seven of the 12 mutations were single-base substitutions but one of the 12 mutants had nucleotide changes at positions 484 (A ! G), 797 (G ! A), and 1246 (T ! A; in the 30 -untranslated region), which resulted in two mutated amino acids, namely, I162V and G266E. One of the 12 mutants had mutations at positions 382 (C ! A) and 1095 (C ! T), as also detected in the analysis of the AxCAhp53-F/WT genome. Two of the 12 mutants (p5330FS and p53-360FS) had frame-shift mutations in the coding region and one of the 12 mutants (p53-1244FS) had a frame-shift mutation in the 30 -untranslated region. The sites of mutations were localized predominantly in the 50 - and 30 -end regions of the p53 gene. One of the 12 mutations was found to be a large deletion from position )93 to position +885, which resulted in p53del1-293 (Fig. 2). The mutation frequency of the p53 gene was 12/ 115 (1:04
101 ) and the mutation rate was calculated to be 2:90
107 per base pair per passage (Table 3). We also identified two mutated clones of the AxCAhp53-F/K20 genome after cloning into the cosmid vector (Fig. 2 and Table 4). One of these mutants had a

single-base substitution at position 126 (T ! C), which resulted in p53-42SM, and one had a deletion from position )93 to position +885, which resulted in p53 del1-293. The mutation in p53-42SM was a silent transition mutation. The gene for p53del1-293 lacked most of the coding region of p53, including the site of initiation of translation and it was identified by the two different screening methods. The mutation frequency of the p53 gene was 2/37 (5:40
102 ) and the mutation rate was calculated to be 1:50
107 per base pair per passage (Table 3). The VA region of the Ad genome is very similar in terms of GC content to the p53 gene [21,22]. Thus, we examined whether or not the VA region was mutated at a rate similar to the mutation rate of p53 gene. The mutation frequency of the VA region in the AxCAhp53-F/K20 genome was 3/41 (7:32
102 ) and the mutation rate was calculated to be 2:04
107 per base pair per passage. Thus, mutations were introduced not only in an inserted gene but also within the vector sequence of the p53-rAd genome. The RCAs were generated from AxCAHSVtk during the multiple passages. Thus, we also examined muta-

454

H. Ugai et al. / Biochemical and Biophysical Research Communications 300 (2003) 448–456

tions in the E1A gene of RCA genomes by the molecular cloning assay (Table 4). We isolated two mutants of the E1A gene among 60 clones examined and the mutation frequency of the E1A region was 1/20 (5:00
102 ) and the mutation rate was calculated to be 1:39
107 per base pair per passages (Table 3). Therefore we found no significant differences among the mutation rates of the respective genes in the genomes isolated from clones of AxCAhp53-F/WT, AxCAhp53-F/K20, and RCAs by the molecular cloning assay (p < 0:01, non-repeated measures ANOVA). Moreover, KOD Plus DNA polymerase, used for amplifications by PCR, did not introduce mutations as a result of errors in the amplification of the target genes (data not shown). Thus, our data suggest that mutations were most likely not generated by the polymerase during amplification by PCR.

Discussion The E1-deleted Ad has been widely used for delivery of exogenous genes to living cells in the laboratory and in preclinical and clinical studies. It has been reported that most preparations of purified Ads include viral particles that appear to be non-functional and non-infectious [23– 25]. We performed the present study to determine the precise rates of mutations of Ads and to characterize mutations in their genomes during multiple passages of rAds. We focused here on p53-rAd and examined the Ad5 and p53-rAd genomes produced in individual plaques. Then we calculated mutation rates by plaque assays and molecular cloning assays, respectively (Fig. 1). In the past, the plaque assay had been used most frequently for the isolation of infectious Ads [23,24]. In our analysis, we failed to detect any mutations in the gene for E1A and the gene for p53 of infectious rAds by the plaque assay (Table 1). However, these results did not exclude the possibility that heterogeneous populations of Ads, including non-infectious Ads, non-functional Ads and Ads with reduced infectivity, might be present within the host 293 cells, as indicated by the previous reports [23–26]. The sensitivity of the plaque assay is limited by the size of individual single plaques. Therefore, the possibility remained that mutations might occur during the replication of Ads that results in formation of individual plaques. Indeed, we isolated two mutated E1A genes from the population of virions in a single plaque of Ad5 by the molecular cloning assay (Table 2). Our results suggest that mutants might be generated during the replication of infectious Ad5 that results in formation of single plaques. The E1A gene is known as an immediate early gene, and the gene products are a family of proteins that act as positive and negative regulators of viral and cellular genes, which include the E1A gene itself [27–31]. Mutations in the E1A gene might result in loss of function such that an

E1A-mutated Ad5 genome is unable to replicate in host cells and is unable to form a plaque. This phenomenon might explain why mutations could not be identified by the plaque assay. We also identified infectious mutants in working viral stock solutions of virus when we used different methods for isolating the virus, such as the limiting-dilution method with 96-well plates ([18]; data not shown). Infectious mutants were found in the population of viruses produced by multiple passages. These mutations became fixed in the rAd genomes and accumulated during subsequent passages (Table 3). We isolated the gene for the p53P128T mutant protein in populations of AxCAhp53F/K20 and AxCAhp53-F/WT by the molecular cloning assay (Table 4) and by the limiting-dilution assay (data not shown). Multiple passages should produce populations of heterogeneous Ads and mutations should be detected during plaque formation (Table 2). In general, the ratio of particles to plaque-forming units of Ad5 is approximately 20:1 [24] since preparations of purified Ad5 include non-infectious viruses [23,24]. In other words, most Ads or rAds cannot form complete virus particles. Infectious and non-infectious Ad or rAd particles might include not only wild-type but also mutated Ad genomes, as indicated by this study. Although the plaque assay did allow us to isolate infectious Ads, it might be difficult to identify minimally infectious and non-infectious mutants by this method. Thus, it appears that the molecular cloning assay provides a more efficient method for the detection of mutations in Ad genomes during multiple passages. This assay also allows detection of mutations not only in infectious viruses but also in minimally and non-infectious viruses. Most of the mutations that we identified in p53 genes were localized in the transcriptional activation domain, the SH3 domain, and the regulatory domain (Fig. 2 and Table 4). Although the human p53 gene is often mutated in the DNA-binding domain in human cancers, spontaneous mutations in the p53-Ad genome after multiple passages were rarely detected in this domain (Fig. 2 and Table 4). The single-base substitutions, frame shifts as well as mutations of VA region on the Ad genome were probably the result of errors introduced by the DNA polymerase of Ad5 and the large deletion of nt )93 (del978) was generated by excision of part of the p53 gene. The molecular cloning assay allows detection not only of fixed mutations but also of mutations that reflect genome instability, which were not found in p53-rAd by the plaque assay. We were able to determine the precise rate of mutation during multiple passages from nucleotide sequences. We showed that mutations were generated and accumulated during the replication of Ad5 and p53-rAd genomes and that mutants were present in the working stock solutions of viruses at significant

H. Ugai et al. / Biochemical and Biophysical Research Communications 300 (2003) 448–456

levels. When we infected target cells with some rAds that contained mutated p53 genes, excluding those with silent and frame-shift mutations, we found that the viruses expressed the mutated p53 proteins in host cells and were infectious (unpublished data). When these mutant proteins were expressed in host cells in conjunction with the wild-type proteins, it might be possible that the mutated proteins affect or modulate the normal function of the wild-type proteins. To maintain the high quality and ensure the safety of Ads and p53-rAds in laboratory experiments and in the delivery of genes for human gene therapy, the presence of mutations should be monitored during the production of viruses. Studies of the mechanisms that generate such mutations and measurements of rates of mutation of rAd vectors are necessary to prevent the spontaneous mutation of p53-rAd and other rAds during multiple passages.

Acknowledgments We thank Drs. G. Gachelin, K. Itakura, and R. Chiu for valuable comments and discussions. We also thank Ms. S. Watanabe, Y. Kujime, and M. Hirose for technical support. This work was supported by grants from the BioResource Research Projects of RIKEN (to K.K.Y.), the Uehara Memorial Foundation (to K.K.Y.), and the Special Coordination Funds of the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of Japan (to T.M. and K.K.Y.).

References [1] B. Vogelstein, D. Lane, A.J. Levine, Surfing the p53 network, Nature 408 (2000) 307–310. [2] L.D. Attardi, T. Jacks, The role of p53 in tumor suppression: lessons from mouse models, Cell. Mol. Life Sci. 55 (1999) 48–63. [3] S.G. Evans, G. Lozano, The Li–Fraumeni syndrome: an inherited susceptibility to cancer, Mol. Med. Today 3 (1997) 390–395. [4] J. Bargonetti, J.J. Manfredi, Multiple roles of the tumor suppressor p53, Curr. Opin. Oncol. 14 (2002) 86–91. [5] J.A. Roth, S.W. Swisher, M.F. Meyn, p53 tumor suppressor gene therapy for cancer, Oncology 13 (1999) 148–154. [6] M.M. Hitt, R.J. Parks, F.L. Graham, Structure and genetic organization of adenovirus vectors, in: T. Friedmann (Ed.), The Development of Human Gene Therapy, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1999, pp. 61–68. [7] H. Mizuguchi, K.A. Mark, T. Hayakawa, Approaches for generating recombinant adenovirus vectors, Adv. Drug Delivery Rev. 52 (2001) 165–176. [8] N.A. Wivel, G.-P. Gao, J.M. Wilson, Adenovirus vectors, in: T. Friedmann (Ed.), The Development of Human Gene Therapy, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1999, pp. 87–110. [9] S. Miyake, M. Makimura, Y. Kanegae, S. Harada, Y. Sato, K. Takamori, C. Tokuda, I. Saito, Efficient generation of recombinant adenoviruses using adenovirus DNA-terminal protein complex and a cosmid bearing the full-length virus genome, Proc. Natl. Acad. Sci. USA 93 (1996) 1320–1324.

455

[10] A.J. Bett, W. Haddara, L. Prevec, F.L. Graham, An efficient and flexible system for construction of adenovirus vectors with insertions or deletions in early regions 1 and 3, Proc. Natl. Acad. Sci. USA 91 (1994) 8802–8806. [11] Y. Sato, K. Tanaka, G. Lee, Y. Kanegae, Y. Sakai, S. Kaneko, H. Nakabayashi, T. Tamaoki, I. Saito, Enhanced and specific gene expression via tissue-specific production of Cre recombinase using adenovirus vector, Biochem. Biophys. Res. Commun. 244 (1998) 455–462. [12] K.M. Hehir, D. Armentano, L.M. Cardoza, T.L. Choquette, P.B. Berthelette, G.A. White, L.A. Couture, M.B. Everton, J. Keegan, J.M. Martin, et al., Molecular characterization of replication-competent variants of adenovirus vectors and genome modifications to prevent their occurrence, J. Virol. 70 (1996) 8459–8467. [13] J. Zhu, M. Grace, J. Casale, A.T. Chang, M.L. Musco, R. Bordens, R. Greenberg, E. Schaefer, S.R. Indelicato, Characterization of replication-competent adenovirus isolates from largescale production of a recombinant adenoviral vector, Hum. Gene Ther. 10 (1999) 113–121. [14] H. Lochm€ uller, A. Jani, J. Huard, S. Prescott, M. Simoneau, B. Massie, G. Karpati, G. Acsadi, Emergence of early region 1containing replication-competent adenovirus in stocks of replication-defective adenovirus recombinants (delta E1 + delta E3) during multiple passages in 293 cells, Hum. Gene Ther. 5 (1994) 1485–1491. [15] F.L. Graham, J. Smiley, W.C. Russell, R. Nairn, Characteristics of a human cell line transformed by DNA from human adenovirus type 5, J. Gen. Virol. 36 (1977) 59–74. [16] N. Louis, C. Evelegh, F.L. Graham, Cloning and sequencing of the cellular-viral junctions from the human adenovirus type 5transformed 293 cell line, Virology 233 (1997) 423–429. [17] Y. Yoshida, A. Sadata, W. Zhang, K. Saito, N. Shinoura, H. Hamada, Generation of fiber-mutant recombinant adenoviruses for gene therapy of malignant glioma, Hum. Gene Ther. 9 (1998) 2503–2515. [18] Y. Kanegae, M. Makimura, I. Saito, A simple and efficient method for purification of infectious recombinant adenovirus, Jpn. J. Med. Sci. Biol. 47 (1994) 157–166. [19] S.I. Schmid, P. Hearing, Adenovirus DNA packaging; construction and analysis of viral mutants, in: W.S.M. Wold (Ed.), Adenovirus; Methods and Protocols, Humana Press Inc., Totowa, NJ, 1999, pp. 47–59. [20] J.Y. Mo, H. Maki, M. Sekiguchi, Mutational specificity of the dnaE173 mutator associated with a defect in the catalytic subunit of DNA polymerase III of Escherichia coli, J. Mol. Biol. 222 (1991) 925–936. [21] M.B. Mathews, T. Shenk, Adenovirus-associated RNA and translational control, J. Virol. 65 (1991) 5657–5662. [22] Y. Ma, M.B. Mathews, Structure, function, and evolution of adenovirus-associated RNA: a phylogenetic approach, J. Virol. 70 (1996) 5083–5099. [23] C.E. Schwerdt, J. Fogh, The ratio of physical particles per infectious unit observed for poliomyelitis viruses, Virology 4 (1957) 41–52. [24] M. Green, M. Pina, R.C. Kimes, Biochemical studies on adenovirus multiplication. XII. Plaqueing efficiencies of purified human adenoviruses. Discussion and preliminary reports, Virology 31 (1967) 562–565. [25] W. Lawrence, H. Ginsberg, Intracellular uncoating of type 5 adenovirus deoxyribonucleic acid, J. Virol. 1 (1967) 851–867. [26] N. Mitterder, K.L. March, B.C. Trapnell, Evaluation of the concentration and bioactivity of adenovirus vectors for gene therapy, J. Virol. 70 (1996) 7498–7509. [27] A.J. Berk, F. Lee, T. Harrison, J. Williams, P.A. Sharp, Pre-early adenovirus 5 gene product regulates synthesis of early viral messenger RNAs, Cell 17 (1979) 935–944.

456

H. Ugai et al. / Biochemical and Biophysical Research Communications 300 (2003) 448–456

[28] C. Tibbetts, P.L. Larsen, S.N. Jones, Autoregulation of adenovirus E1A gene expression, J. Virol. 57 (1986) 1055– 1064. [29] P.L. Larsen, C. Tibbetts, Adenovirus E1A gene autorepression: revertants of an E1A promoter mutation encode altered E1A proteins, Proc. Natl. Acad. Sci. USA 84 (1987) 8185– 8189.

[30] J.D. Cogan, S.N. Jones, R.K. Hall, C. Tibbetts, Functional diversity of E1A gene autoregulation among human adenoviruses, J. Virol. 66 (1992) 3833–3845. [31] R. Hamid, J.D. Cogan, S.N. Jones, C. Tibbetts, Phenotypic determinants of adenovirus E1A gene autoregulation: variable region between conserved coding domains 2 and 3, Virology 213 (1995) 666–670.