- Email: [email protected]
E2a Double-Deleted Virus for Preparation of Helper-Dependent Adenovirus Vector
doi:10.1006/mthe.2001.0288, available online at http://www.idealibrary.com on IDEAL
METHOD
A Cre-Expressing Cell Line and an E1/E2a Double-Deleted Virus for Preparation of Helper-Dependent Adenovirus Vector Heshan Zhou,*,†,‡,1 Tiejun Zhao,*,‡ Lucio Pastore,† Maged Nageh,† Wendy Zheng,*,‡ X. Mei Rao,* and Arthur L. Beaudet†,‡ *Center for Cell and Gene Therapy, †Department of Molecular and Human Genetics, and ‡ Department of Pediatrics, Baylor College of Medicine, Houston, Texas 77030 Received for publication January 5, 2001; accepted in revised form February 9, 2001; published online April 16, 2001
Adenoviral vectors are attractive for the delivery of transgenes into mammalian cells because of their efficient transduction, high titer, and stability. The major concerns with using E1-deleted adenoviral vectors in gene therapy are the pathogenic potential of the virus backbone and the leaky viral protein synthesis that leads to host immune responses and a short duration of transgene expression. Helper-dependent (HD) adenoviral vectors that are devoid of all viral protein-coding sequences have significantly increased the safety and reduced the immunogenicity of these vectors. Currently available HD vectors depend on an E1-deleted adenovirus as a helper to provide viral proteins in trans. As a consequence, contamination with helper virus cannot be avoided in the HD vector preparation though it can be decreased to 0.01% using a Cre/loxP mechanism. Since the presence of E1-deleted helper virus may have substantial unwanted effects, we have developed a new Cre-expressing cell line based on an E1- and E2a-complementing cell. This new cell line can efficiently cleave the packaging region in the helper virus genome. We have also developed an E1 and E2a double-deleted helper virus. By using the CreE cell with the helper virus deleted in both the E1 and the E2a genes it may be possible to further improve the safety of the vectors.
INTRODUCTION The human adenovirus genome is about 36 kb of doublestranded linear DNA. Adenoviral (Ad) vectors are particularly attractive for in vivo and in vitro gene delivery, primarily because of their highly efficient gene transfer and relatively easy production of stable and high-titer vectors. Numerous studies using Ad vectors in a variety of animal models have demonstrated successful gene transfer to many tissues, with high levels of recombinant gene expression, making these vectors attractive delivery systems to treat a variety of human diseases (1–5). The major concerns with using the E1-deleted Ad vectors are the pathogenic potential of the adenovirus backbone and the transient expression of the therapeutic gene caused by immune responses against viral-transduced cells expressing low levels of Ad genes (1, 4, 6 –9). The E1 region encodes proteins for the regulation of other viral genes (10). Mutation of the E1 region does not completely eradicate the synthesis of viral proteins and DNA (11, 12). The host cells may provide cellular “E1-
1 To whom correspondence and reprint requests should be addressed at the Center for Cell and Gene Therapy, Baylor College of Medicine, One Baylor Plaza, Room N1010, Houston, TX 77030. Fax: (713) 7981230. E-mail: [email protected].
MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy 1525-0016/01 $35.00
like” factors, such as the 70-kDa human heat shock protein (13) or interleukin-6 (14, 15), to substitute E1 functions. When cells respond to “emergency” stimuli such as viral infection or heat, they may be more likely to produce or activate cellular E1-like factors. These cellular factors can cause unwanted synthesis of viral proteins and replication of the viral DNA. Recent gene therapy studies have indicated that leaky expression of viral genes from E1deleted vectors can occur in transduced cells (16, 17). In addition, replication-competent adenovirus (RCA) may emerge during Ad vector propagation. The E1-deleted vectors are usually produced in 293 cells generated by the transfection of human kidney cells with Ad serotype 5 (Ad5) DNA (18). Mapping of the Ad sequences in the 293 cell line shows the presence of contiguous Ad5 sequences from the left-hand end of the genome up to position 4344 (19). When typical E1-deleted vectors, which contain a deletion from 340 to 3500 of the Ad genome, are propagated in 293 cells, sequence overlap between the vector and the complementing cell frequently results in the generation of RCA (20 –22). Since adenoviruses can replicate very well even at extremely low levels of E1 (23, 24), even limited contamination with E1-positive RCA or a low level of cellular E1-like factors may have catastrophic consequences. Recent studies have also demonstrated that E1-deleted vectors could induce cell G2 growth arrest,
613
METHOD which may produce unwanted effects during human gene therapy (25). One strategy to overcome such limitation is to develop vectors with deletions of additional essential viral genes. Ad vectors with deleted E1/E2 or E1/E4 genes can be propagated in cell lines that complement both E1 and E2 or E1 and E4 (26 –30). Unlike the E1 products, the E2a encodes for a DNA binding protein directly involved in viral DNA replication (10). No cellular factor has been found that can functionally replace the E2a product. Ad vectors with deleted E2 do not express detectable adenoviral early or late proteins, do not synthesize viral DNA in noncomplementing cells, and demonstrate improved safety (27, 29, 31). However, the E1/E2-deleted vectors still stimulate host immune responses for unknown reasons, which may benefit vaccine development but limit the duration of transgene expression (32–34). Another approach to reduce the immunogenicity and improve the safety of Ad vectors is to delete all viral coding sequences so that leaky expression of viral proteins is completely eliminated. Helper-dependent (HD; also called “gutless”) vector systems have been developed in which all or most of the viral protein-encoding sequences have been deleted (35). An improved HD system involves a 293-derived cell line (293Cre4) that stably expresses the bacterial phage P1 Cre recombinase and an E1-deleted helper virus that provides the necessary viral proteins in trans for the replication and packaging of the HD vectors (36, 37). The helper virus contains two loxP sites flanking a viral region that is essential for packaging virus DNA into virions. Upon infection of 293Cre4, the packaging signal in the helper virus DNA is excised through the Cre/loxP interaction, so that the helper virus DNA cannot be packaged. Ad HD vectors are substantially less toxic than other Ad vectors studied to date (38 – 41). Long-term transgene expression of HD vectors in the liver and other organs has been reported with experimental animals (39, 40, 42– 45). Even though first-generation adenoviral vectors have lost their initial luster, HD vectors are emerging as one of the most promising vector systems for gene therapy (46). Clinical trials of HD vectors are dependent on the production of safe HD vectors with high yield. With the current HD Ad vector systems, contamination with helper virus in the final purified HD vector has been decreased to 0.01%, but cannot be avoided (37, 47). When HD vectors are used in clinical trials that may involve 1012 or even more functional units of the HD vector, 108 or more infectious helper virus units may thus be administered. Contamination with E1-deleted helper virus may cause complications in clinical trials. To further improve Ad HD vector production and safety, the E3 and the packaging region in the helper virus and the backbone structure of the HD vector have been modified (48, 49). In this study, we have addressed the current limits of the HD vector system by developing a new Cre-expressing cell line based on E2T, an E1- and E2a-complementary cell line (30). One of the cell lines can efficiently excise the packaging region of the helper virus. We also developed
614
an E1 and E2a double-deleted helper virus, which can be used with the new cell line to produce HD vector with low helper contamination. This system may further improve the HD vector safety.
MATERIALS
AND
METHODS
Cell lines and cell culture. 293 cells contain chromosome-integrated Ad5 E1 sequences complementing the functions of the Ad E1 genes (18). 293Cre4 is a 293-derived cell line that expresses Cre recombinase and is used for the production of HD vectors (36, 37). E2T is an E1 and E2a double-complementing cell line derived from 293 cells (30). Expression of the E2a gene in E2T is inducible upon withdrawal of tetracycline. Several Cre-expressing cell lines were developed in this study by transfection of Cre-expressing cassettes into 293 or E2T cells. All cells were grown in ␣MEM (GIBCO, Grand Island, NY) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), 100 units of penicillin, and 100 g of streptomycin per milliliter. E2T and E2T-derived cells were maintained in this medium containing also 10 g/ml tetracycline to suppress E2a expression. Plasmids. The plasmid pOG231 contains a Cre-expressing cassette and was successfully used in chromosome engineering (50). The plasmid pUHD10-3 contains an inducible promoter regulated by tetracycline (51). A DNA fragment encoding the Cre open reading frame (ORF) was cleaved from pOG231 with restriction enzymes BglII and SalI and inserted into pUHD10-3 to construct plasmid pBZ73. The Cre ORF in pBZ73 was under the regulation of the tetracycline-inducible promoter. Two plasmids with drug-resistant genes were used to select Cre-expressing cell colonies. The PGL-VP plasmid contains the hygromycin-resistance gene (52). The pPGKpurobpA plasmid, containing the puromycin-resistance gene, was kindly provided by Dr. Allan Bradley. The above Cre-expressing and selective plasmids were used to transfect 293 and E2T cells. The plasmid pLoxZ was constructed by separating the EF-1␣ promoter (53) and the Escherichia coli -galactosidase (lacZ) ORF with two loxP sites, which sandwich a segment containing a stop codon (Fig. 1). Expression of lacZ from this plasmid was inhibited because of premature termination of translation at the stop site. Deletion of the DNA segment between the loxP sites through Cre/loxP interaction activated the expression of lacZ (Fig. 1). This plasmid was used to screen the Cre-expressing cells. Plasmid pDelta28 is an HD backbone plasmid containing the Ad left inverted terminal repeat (ITR) and packaging signals (Ad5, 1– 440 bp) and the right ITR (Ad5, 35818 –35935 bp) (L. Pastore and A. Beaudet, unpublished data). A DNA “stuffer” was included in pDelta28 to achieve a final vector size close to the natural Ad5 size for increasing the vector stability (54). For this purpose, a 16,054-bp EclXI/PmeI fragment from the HPRT gene (GenBank Accession No. Humhprtb) and an 11,105-bp BamHI fragment from the cosmid C346 gene (GenBank Accession No. L31948) were inserted into the pDelta28 construct. A HD vector in plasmid format was constructed based on the plasmid pDelta28. The -geo gene (a fusion of -galactosidase and neomycin-resistance genes) with a strong mammalian promoter, SR␣, was cleaved from the plasmid pSR␣-geo (35) and subcloned into the unique AscI site of the pDelta28 plasmid, resulting in the plasmid p⌬-geo. Transfection and Cre cell selection. Transfection or cotransfection required calcium phosphate precipitation (55). For production of Cre-expressing cells, E2T cells were cotransfected with the Cre-expressing plasmid pBZ73 and the hygromycin-resistant plasmid PGL-VP. Two days after transfection, the cells were selected in medium containing 50 g/ml hygromycin. The medium was changed every 3 to 5 days. 293 cells were cotransfected with pOG23 and pPGKpurobpA and selected in 1 g/ml puromycin. To screen Cre-expressing cell clones, drug-resistant clones cultured in 96-well plates were transfected with pLoxZ plasmid DNA. Two days after transfection, the cells were stained with 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal). Only the cells that expressed Cre recombinase stained blue. One pair of PCR primers was designed to assay the excision of the packaging region of the Ad helper virus propagated in Cre cells. The primers, Ad121F (5⬘-GAT GTT GCA AGT GTG GC) and Ad3704R (5⬘-GGA GCC CAT CAC ATT CTG), flank the adenoviral packaging region in the MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
METHOD helper virus. Various cells were infected with the helper virus at a multiplicity of infection (m.o.i.) of 1. Two days after infection, the cells were collected and analyzed by PCR, using reaction conditions of 94°C for 2 min, followed by 30 cycles 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min. Helper virus and HD vector. The helper virus AdLC8cluc contains two loxP sites flanking the Ad packaging signal and the expression cassette of firefly luciferase inserted in the E3-deleted region of the Ad genome (37). This helper virus contains a deletion in the E1 region. The HD vector, Ad⌬-geo, was converted from plasmid p⌬-geo DNA in the CreE cells developed in this study. The plasmid p⌬-geo was first cleaved with PmeI to free the adenoviral ITRs and release the HD vector genome and then transfected into CreE cells. The second day after transfection, the CreE cells were infected with the helper virus AdLC8cluc that provides viral proteins in trans to rescue the HD Ad⌬-geo as previously described (37). The HD Ad⌬-geo was then amplified through a series of co-infections with the helper virus in CreE cells. The amplification of HD Ad⌬-geo could be easily monitored by assaying lacZ expression in the vector. The blue cells after each passage were directly counted in the plates after X-Gal staining as follows: the cells with 90% cytopathic effect (CPE) after co-infection with the helper virus and the HD vector were collected; the residual attached cells were fixed with 0.2% glutaraldehyde and 2% formaldehyde in PBS for 5 min at room temperature and then incubated at 37°C overnight with 0.1% X-Gal in a solution of 5 mM K4Fe(CN)6, 5 mM K3Fe(CN)6, and 2 mM MgCl2. Purification of the Ad⌬-geo vector was carried out by two rounds of CsCl centrifugation (55) and passage through desalting columns (EconoPac, 10 DG; Bio-Rad, Hercules, CA). The HD vector preparations were evaluated by total viral particle count, as determined by optical density measurements of DNA (56), and by measurement of blue forming units (BFU) of the HD Ad⌬-geo vector in the E2T cells. The quantity of contaminating helper viruses in the HD vector preparation was determined by measurement of infective units (IFU) in 96-well plates as described previously (49). Southern blot analysis of helper virus and HD vector. Various cells in 10-cm dishes were infected with the helper virus AdLC8cluc at an m.o.i. of 1. After 2 days, the cells were collected, washed with PBS, and digested overnight with 0.5 ml of proteinase K (10 mM Tris–HCl (pH 7.5)/10 mM EDTA/0.5% SDS/0.04% proteinase K) at 50 –C. The total DNA was then extracted once with buffer-saturated phenol:chloroform (Amresco, Solon, OH) and precipitated with ethanol. For isolation of the vector DNA from the CsCl-purified preparation, 1 l of the purified vector was mixed with 2 g of salmon DNA (as a carrier) and incubated in 100 l of the proteinase K solution before purification as described above. The isolated DNA samples were then cleaved with the restriction endonuclease PstI, separated on agarose gels, and transferred to Hybond-N positively charged nylon membranes (Amersham International plc, Amersham, UK) by capillary transfer (57) for hybridization with a radioactively labeled DNA probe.
RESULTS Isolation and Screening of Cre Cell Clones E2T cells were cotransfected with the plasmid pBZ73 containing the Cre expression cassette and the plasmid pGL-VP with the hygromycin-resistance marker. 293 cells were cotransfected with the Cre plasmid pOG231 and the plasmid pPGKpuropbA with the puromycin-resistance gene. The drug-resistant clones were isolated and then cultured in duplicated 96-well plates. One set of the plates was used for analysis of Cre activity, and the other set was saved for further analyses. To facilitate screening of Creexpressing cell clones, we designed a method based on the Cre/loxP-mediated deletion activating expression of the lacZ gene. The plasmid pLoxZ contains a loxP–stop–loxP segment that separates the promoter and the lacZ ORF. Two days after transfection with pLoxZ, the cells were MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
stained with X-Gal. Only the cells that express Cre recombinase can remove the loxP-flanked segment in the plasmid, resulting in expression of the lacZ gene (Fig. 1). Without overstaining, the number of blue-positive cells and the color density can partially reflect the levels of Cre expression in the cells. We found that this method can be easily and reliably used for initial screening of the Crepositive cells. Several hundred clones were checked using this method. We picked up several clones, most with high numbers of dense blue cells (293-Cre8, E2T-Cre2, E2TCre4, E2T-Cre5, and E2T-Cre6) and a few with low numbers of light blue cells (E2T-Cre7 and E2T-Cre9) for further analysis with more accurate methods.
Deletion of Helper Virus Packaging Region in Cre Cells Since deletion of the helper virus packaging region is critical for the preparation of highly purified HD vector, the efficiency of Cre/loxP-mediated deletion was further evaluated using the Southern blot hybridization method. The total DNA samples were isolated from the 293, E2T (both Cre negative controls), 293Cre4 (Cre positive), and Cre-positive clone cells infected with AdLC8cluc. The DNA was then digested with the restriction enzyme PstI, which cleaved a 0.53-kb fragment from the left terminus and a 2.06-kb fragment from the right end of the helper virus (Fig. 2). The hybridization probe was a 0.14-kb fragment generated from pXCJL.2 (55) with EcoRI and AflIII. This probe, covering the 103-bp ITRs, can hybridize to both the 2.06- and the 0.53-kb terminal fragments. Deletion of the packaging signal (⌿) in the helper virus through the Cre/loxP interaction will shorten the 0.53-kb fragment to a 0.26-kb fragment as shown in the positivecontrol 293Cre4 cells (Fig. 2). Hence, the ratio of the 0.26-kb band to the 0.53-kb band represents the excision efficiency of the packaging signal by Cre recombinase in the cells. The 2.06-kb band representing the right terminus of helper virus acts as an internal control for equal loading. For 293 and E2T cells without Cre expression, there was no deletion of the packaging signal in the helper virus (Fig. 2). In the Cre-positive control 293Cre4 cells, most of the packaging signal was deleted, with about 10% of 0.53-kb fragment remaining. The excision of the packaging region in the screened cells was compared with the control 293Cre4 samples. 293-Cre14, E2T-Cre2, E2TCre7, and E2T-Cre9 were less efficient than the control 293Cre4 in deleting the packaging region; 293-Cre8 and E2T-Cre4 were similar to 293Cre4 and E2T-Cre5 and E2TCre6 were more efficient than 293Cre4. These results generally agreed with the results obtained by using plasmid pLoxZ. In E2T-Cre9, there was an extra band about 0.8 kb (Fig. 2) that also existed in repeated Southern blot assays. We do not know where this band comes from. This clone with several others was not used in the HD vector preparation and not further studied. Excision of the packaging region from the helper virus in E2T-Cre6 cells was highly efficient, and the 0.53-kb fragment could hardly be detectable as most of the DNA converted to the 0.26-kb fragment (Fig. 2). This result suggested that the E2T-Cre6
615
METHOD
FIG. 1. (A) Screening of the Cre-expressing cell clones by Cre/loxP deletion activation. The Cre/loxP activation mechanism is shown at the top. Plasmid pLoxZ contains two loxP sites that flank a DNA segment with a stop codon. Expression from the lacZ gene in a Cre-negative cell (E2T) is blocked because of premature termination. Excision of the segment between the loxP sites in Cre-expressing cells (E2T-Cre6) activates the expression of the lacZ gene. (B) Resistance to helper virus infection. 293, 293Cre4, E2T-Cre4, and E2T-Cre6 cells were infected with the helper virus AdLC8cluc at an m.o.i. of 0.25. Efficient deletion of the viral packaging region in E2T-Cre6 cells prevented packaging of the viral DNA into infectious particles. The helper virus spread inefficiently in the E2T-Cre6 culture.
616
MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
METHOD
FIG. 2. Southern blot assay for deletion of the packaging region in the helper virus DNA. Total DNA was isolated from different cells infected with the Ad helper virus AdLC8cluc and cleaved with restriction enzyme PstI. (A) The probe containing the Ad ITR could hybridize with both Ad terminal fragments. Deletion of the packaging region in the helper virus decreased the left terminal fragment from 0.53 to 0.26 kb. The size of the right terminal fragment did not change. The DNA isolated from Cre-negative 293 and E2T cells showed undeleted fragments, while the packaging region was deleted in the 293Cre4 cells. (B) The efficiency of the Cre/loxP-mediated deletion in different cell lines is shown.
clone derived from E2T could efficiently excise the packaging region from the helper virus DNA. The E2T-Cre6 was renamed CreE.
CreE Resistance to Helper Virus Infection Since the helper virus DNA deleted in the packaging region could not be packaged into viral particles, we reasoned that infectious virions would not efficiently propagate in CreE cells. Based on this concept, when the CreE cells were infected with the helper virus at a low m.o.i., the uninfected cells would survive because new virus would not be efficiently produced from the infected cells. To evaluate this concept, 293, 293Cre4, E2T-Cre4, and CreE cells were infected with the helper virus at an m.o.i. of 0.25. All 293 cells, and most 293Cre4 and E2T-Cre4 MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
cells, showed CPE at 2 days postinfection, while only a small portion of E2T-Cre6 cells were killed (Fig. 1B). This result, combined with the data from the Southern blot analysis (Fig. 2), suggested that the infectious helper virus could be quickly cleaned out by passage in CreE cells.
Large-Scale Preparation of Highly Purified HD Vector in CreE Cells HD Ad⌬-geo was converted from plasmid p⌬-geo according to the method in (37). The plasmid p⌬-geo DNA was first linearized with restriction enzyme PmeI and then transfected into CreE cells. The helper virus AdLC8cluc was added to the cell culture to provide necessary viral proteins for rescuing the HD vector. A serial co-infection of the HD vector with the helper virus in CreE cells re-
617
METHOD sulted in an increase in the titer of the HD vector that was easily monitored by X-Gal staining. With the CreE cell line, rescuing and amplification of the HD vector were similar to those with the 293Cre4 cell line. To manufacture HD vectors on a large scale, 16 triplelayer flasks (500 cm2 per flask) of CreE cells were cultured until about 90% confluence. The cells were co-infected with both HD vector Ad⌬-geo, m.o.i. of 10 (BFU), and helper virus AdLC8cluc, m.o.i. of 5 (IFU). The infected cells were harvested when over 90% showed CPE (about 3 days). We made a total of six preparations by using AdLC8cluc. Each preparation yielded 3 ml of HD vector with 0.7 to 1.2 ⫻ 1012 particles/ml. The vector preparation was diluted and titered on E2T cells in 96-well plates for its infectious activity. The HD vector titer (BFU) counted after X-Gal staining was between 5 ⫻ 1010 to 1 ⫻ 1011 BFU/ml. The helper virus titer was determined as IFU causing CPE 14 days postinfection. The helper virus was 105/ml IFU or lower in the purified HD vector preparation. Therefore the helper:vector ratio is about 10⫺5 or lower (IFU:BFU). To further determine contamination of the helper virus in the HD vector preparations, the DNA was purified from the above vector preparations for Southern blot analysis. The DNA was digested with PstI and hybridized with the probe with Ad ITR as used in analysis of the deletion of the Ad packaging region. Since the helper virus DNA deleted of its packaging region cannot be packaged into virions, the DNA isolated from the purified HD vector preparation represented only the DNA from virions. The control having only the helper virus DNA was isolated from Cre-negative E2T cells infected with AdLC8cluc. The largest (2.06 kb) and the smallest (0.53 kb) fragments represent the helper virus DNA right and left terminal segments, which were also shown in Fig. 2. The helper virus contamination was barely detectable in the HD vector preparations purified from the CreE cells (Fig. 3).
E1 and E2a Complementation of CreE Cells To test the E1 and E2a complementation of CreE cells, we used our previously constructed Adgal⌬E1 (deleted E1) Adgal⌬E1E2 (deleted at both E1 and E2a) vectors (26) to infect 293, E2T, and CreE cells at an m.o.i. of 1. The total cells were collected 2 days after infection; lysates were prepared and serially diluted to determine the titers of BFU on E2T cells in 96-well plates. The titers of the Adgal⌬E1 containing wild-type E2a were similar regardless of the cell lines, because the Adgal⌬E1 did not depend on E2a complementation from the cells. The E2adeleted vector did not replicate in 293 cells, as the titer of Adgal⌬E1E2 vector in 293 cells was only 1% of the input virus. However, the Adgal⌬E1E2 vector titer in CreE6 cells was increased as it was in E2T cells (Fig. 4).
A Helper Virus Deleted at both E1 and E2a Two loxP sites were introduced into the shuttle plasmid p⌬E1sp1A (58) to flank the Ad package region. The two
618
loxP sites are at locations similar to those of the published helper virus AdLC8cluc (37). The resulting shuttle plasmid was cotransfected into E2T cells with plasmid pBZ71, which is an Ad backbone plasmid for the construction of E1- and E2a-deleted vectors (30). A helper virus with deletions of E1 and E2a was rescued and named Adhv⌬E1/E2 (Fig. 4). Two PCR primers (Ad121F and Ad3704R) that flank the packaging region of adenovirus were designed to analyze this region. Three viruses, Ad⌬E1, AdLC8cluc, and Adhv⌬E1/E2, were used to infect Cre-negative E2T cells and Cre-positive E2T-Cre6 cells. The E1-deleted Ad⌬E1 (unpublished work, H. Zhou), identical to Adhv⌬E1E2 in the deleted E1 region but without inserted loxP, was used as negative control. The helper virus AdLC8cluc, containing two inserted loxP sites, was used as positive control. PCRs with Ad⌬E1 DNAs isolated from both Cre-negative and Cre-positive cells produced a 0.4-kb fragment (Fig. 5). Insertion of two loxPs in AdLC8cluc and Adhv⌬E1E2 increased the size of the PCR fragments. When the two helpers were cultured in CreE cells, deletion of the packaging region through loxP and Cre interaction resulted in shorter fragments (Fig. 5). This suggested that the Ad packaging region Adhv⌬E1E2 could be efficiently deleted. With the new CreE cell line, the E1 and E2a doubledeleted helper Adhv⌬E1E2 could be used to replace the E1-deleted helper virus AdLC8cluc to produce the HD vector. The method for preparation of the HD vector with the Adhv⌬E1E2 was the same as when the AdLC8cluc was used. We made a total of three preparations with the helper Adhv⌬E1E2. Each preparation yielded 3 ml of HD vector with 1 to 5 ⫻ 1011 particles/ml. The HD vector titer (BFU) was between 0.8 and 2 ⫻ 1010 BFU/ml. The helper virus was 104/ml IFU in the purified HD vector preparation. Therefore the helper:vector ratio is about 10⫺6 (IFU: BFU). The structure of the HD vector prepared with the E1 and E2a double-deleted helper was also analyzed. We had a result similar to that shown in Fig. 3 when the helper ADLC8cluc was used. When the HD Ad⌬-geo vector prepared in this study was injected into rat brain, the vector resulted in prolonged transgene expression and lower immune response compared with the first-generation vector (41, 45).
DISCUSSION The HD Ad vector system is one of the most promising virus vector systems for gene delivery. However, largescale preparation of highly purified and safe HD vector has proved difficult, limiting vector use for clinical trials. The construction and production of HD vectors are more complex than those of the first-generation vectors. There are three key components in the HD vector system: the Cre-expressing cells, the helper virus, and the vector itself. The growth conditions of the cells, the structure of helper virus and HD vector, and their titer used in co-infection may all affect the final vector yield and quality. In efforts to improve the vector system, both the helper virus and the vector structure have been modified (48, 49). In all the MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
METHOD
FIG. 3. Southern blot assay for purified HD vector DNA. (A) The probe DNA containing Ad ITR can hybridize with the terminal fragments of helper virus and HD vector DNA digested with the restriction enzyme PstI. (B) When the helper virus DNA was present, hybridization resulted in 2.06- and 0.53-kb bands as labeled. The contamination of helper virus DNA was hardly detectable in the HD vector preparations.
current HD vector systems, an E1-deleted Ad vector is used as a helper virus for providing viral proteins in trans. Contamination with E1-deleted helper virus in the HD-Ad vector preparations is unavoidable (37, 47). Based on our recently reported E1 and E2a complementing cell line (30), we have developed a new Creexpressing cell line (CreE) that could efficiently excise the packaging region (Fig. 2) and complement both E1 and E2a genes (Fig. 4). We have also developed an E1 and E2a MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
double-deleted helper virus (Adhv⌬E1E2) for production of HD vectors. The E2a gene encodes the single-stranded DNA binding protein that is essential for viral DNA replication and for transcription from the major late promoter (10). Previous studies confirmed that, unlike the E1-deleted vectors, vectors with deletions in both E1 and E2a do not express detectable adenovirus early and late proteins and do not synthesize viral DNA in noncomplementing cells (27, 29, 31, 59). Hence using the new HD
619
METHOD
FIG. 4. Comparison of the titers of viruses with wild-type E2a or with deletion of E2a on 293, E2T, and E2T-Cre6 cells. Titer of the virus with E1 deleted, Adgal⌬E1, is shown on the left, and titer of the virus with E1 and E2a deleted, Adgal⌬E1E2, is shown on the right. Viruses were collected from medium after 2 days of infection with an m.o.i. of 1, and the titer was measured as blue forming units (BFU) on E2T cells.
vector system can further increase the safety of the HD vector. The CreE cell could efficiently delete the packaging signal in a helper virus (Fig. 2) and therefore produced the HD vector with low contamination of the helper virus. The decreased helper contamination is dependent on the cell line Cre activity. The helper virus-to-HD vector ratio in the HD vector preparations was about 10⫺5 for AdLC8cluc and 10⫺6 for Adhv⌬E1E2 (IFU/BFU). The smaller Adhv⌬E1E2 ratio may due to slightly lower level of replication with the E1/E2a double-deleted virus. The yield of the HD vector using E1 and E2a double-deleted helper virus was slightly lower than that using AdLC8cluc helper, so that it will be necessary to further systematically study and optimize production conditions. The HD vector Ad⌬-geo was rescued and amplified with this system. To achieve a stock with amplified HD vector from a plasmid DNA, the system required six to seven rounds of coinfection with a helper virus (data not shown). We made several preparations of the HD Ad⌬-geo vector using the CreE cell line with both the helper virus AdLC8cluc and the helper Adhv⌬E1E2. The particle yield is similar to that seen when 293Cre4 cells were used. The vector quality prepared in this study was confirmed in animal studies that resulted in a stable transgene expression with reduced immune responses (41, 45). In this study, we have developed a new method for efficiently screening Cre cell lines according to the Cre/ loxP-induced deletion. In plasmid pLoxZ, two loxP sites separate the promoter and the lacZ ORF. When the plasmid is transfected into Cre-expressing cells, deletion through Cre/loxP results in active expression of the lacZ
620
gene. This method is reliable and can be easily used to screen Cre-expressing cells. We have also established a Southern blot method for analysis of HD vector preparation. With this method, only 1 l of purified HD vector is required for analysis. The probe containing the adenovirus ITR can hybridize with the terminal segments of both the HD vector and the helper virus, allowing the ratio of helper virus to HD vector to be directly compared. There is no a standard method to determine the ratio of a helper virus to an HD vector in final vector preparation. Plaque-forming unit and infective unit have been used to determine the helper virus titer (37, 49). The blue forming unit (for the lacZ gene) and functional unit, by comparing with a first-generation vector carrying the same transgene, have been used to determine the HD vector titer (49). We determined the helper:HD vector ratio in IFU: BFU. We should point out that all these methods are not directly comparing the helper virion number with the vector virion number. Southern blot method can directly compare the helper DNA with the vector DNA. However, when the helper DNA is lower than 1% of the HD vector DNA, the helper DNA may be not easily detected. Extending exposure of the Southern blot could increase the density of the weak helper virus bands, but then the HD vector bands would be overexposed. Working out a testing standard for the HD vector system is important and obviously requires collaborations among academia, industry, and some regulation agents. In the CreE cell line, Cre expression is under the regulation of a tetracycline-inducible mechanism. However, we observed no difference in packaging deletion in the CreE cells under inducible (minus tetracycline) or uninMOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
METHOD
FIG. 5. PCR analysis of deletion in the packaging region in helper viruses. Ad⌬E1, AdLC8cluc, and Adhv⌬E1E2 are E1-deleted viruses, while AdLC8cluc and AdhvDE1E2 contain two loxPs flanking the packaging region. The Ad⌬E1 without the inserted loxP was used as negative control; the AdLC8cluc containing two inserted loxP sites was used as positive control. PCR with Ad⌬E1 produced a 0.40-kb fragment regardless of the cell line. Insertion of two loxPs in AdLC8cluc and Adhv⌬E1E2 increased the fragment size. However, deletion of the packaging region through loxP/Cre interaction in E2T-Cre6 cells produced shorter fragments.
ducible (plus tetracycline) conditions (data not shown). One possibility is that the level of Cre even under uninduced conditions may already be at a biologically maximal level. In summary, a CreE cell line has been developed that can complement both E1 and E2a and efficiently excise the packaging signal in a helper virus. By using CreE cells, contamination of helper virus in HD vector preparations can be decreased. An E1 and E2a double-deleted helper virus has also been developed and can be used with the new cell line. Since the E2a gene is essential for adenovirus DNA replication and protein synthesis, a helper virus deleted for both E1 and E2a should further improve the safety of the HD vector system. ACKNOWLEDGMENTS We thank Dr. Malcolm Brenner for critically reading the manuscript and Thomas Gegeny of the Houston Academy of Medicine–Texas Medical Center Library’s publication service for editing the manuscript. The 293Cre4 cell line and helper virus AdLC8cluc (36, 37) were obtained from Merck & Co. This work was supported by grants from the NIH (HL 51751 and HL 59314), the Cystic Fibrosis Foundation (F984), the Texas Higher Education Coordination Board (ATP 004949-119), and the Shell Center for Gene Therapy at Baylor College of Medicine. MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
REFERENCES 1 Stratford-Perricaudet, L. D., et al. (1990). Evaluation of the transfer and expression in mice of an enzyme-encoding gene using a human adenovirus vector. Hum. Gene Ther. 1: 241–256. 2 Rosenfeld, M. A., et al. (1991). Adenovirus-mediated transfer of a recombinant alpha 1-antitrypsin gene to the lung epithelium in vivo. Science 252: 431– 434. 3 Jaffe, H. A., et al. (1992). Adenovirus-mediated in vivo gene transfer and expression in normal rat liver. Nat. Genet. 1: 372–378. 4 Quantin, B., et al. (1992). Adenovirus as an expression vector in muscle cells in vivo. Proc. Natl. Acad. Sci. USA 89: 2581–2584. 5 Hitt, M. M., Addison, C. L., and Graham, F. L. (1997). Human adenovirus vectors for gene transfer into mammalian cells. Adv. Pharmacol. 40: 137–206. 6 McElvaney, N. G. (1996). Is gene therapy in cystic fibrosis a realistic expectation? Curr. Opin. Pulmon. Med. 2: 466 – 471. 7 Marshall, E. (1999). Gene therapy death prompts review of adenovirus vector. Science 286: 2244 –2245. 8 Engelhardt, J. F., et al. (1993). Direct gene transfer of human CFTR into human bronchial epithelia of xenografts with E1-deleted adenoviruses. Nat. Genet. 4: 27–34. 9 van Ginkel, F. W., et al. (1997). Adenoviral gene delivery elicits distinct pulmonaryassociated T helper cell responses to the vector and to its transgene. J. Immunol. 159: 685– 693. 10 Shenk, T. (1996). Adenoviridae: The viruses and their replication. In Fields Virology (B. N. Fields, D. M. Knipe, and P. M. Howley, Eds.), pp. 2111–2148. Raven Press, Philadelphia. 11 Winberg, G., and Shenk, T. (1984). Dissection of overlapping functions within the adenovirus type 5 E1A gene. EMBO J. 3: 1907–1912. 12 Shenk, T., and Williams, J. (1984). Genetic analysis of adenoviruses. Curr. Top. Microbiol. Immunol. 111: 1–39. 13 Imperiale, M. J., et al. (1984). Common control of the heat shock gene and early adenovirus genes: Evidence for a cellular E1A-like activity. Mol. Cell. Biol. 4: 867– 874.
621
METHOD 14 Spergel, J. M., and Chen-Kiang, S. (1991). Interleukin 6 enhances a cellular activity that functionally substitutes for E1A protein in transactivation. Proc. Natl. Acad. Sci. USA 88: 6472– 6476. 15 Rancourt, C., et al. (1999). Interleukin-6 modulated conditionally replicative adenovirus as an antitumor/cytotoxic agent for cancer therapy. Clin. Cancer Res. 5: 43–50. 16 Mittereder, N., et al. (1994). Evaluation of the efficacy and safety of in vitro, adenovirus-mediated transfer of the human cystic fibrosis transmembrane conductance regulator cDNA. Hum. Gene Ther. 5: 717–729. 17 Marienfeld, U., et al. (1999). ‘Autoreplication’ of the vector genome in recombinant adenoviral vectors with different E1 region deletions and transgenes. Gene Ther. 6: 1101– 1113. 18 Graham, F. L., et al. (1977). Characteristics of a human cell line transformed by DNA from human adenovirus type 5. J. Gen. Virol. 36: 59 –72. 19 Louis, N., Evelegh, C., and Graham, F. (1997). Cloning and sequencing of the cellular– viral junctions from the human adenovirus type 5 transformed 293 cell line. Virology 233: 423– 429. 20 Lochmuller, H., et al. (1994). Emergence of early region 1-containing replicationcompetent adenovirus in stocks of replication-defective adenovirus recombinants (DE1⫹DE3) during multiple passages in 293 cells. Hum. Gene Ther. 5: 1485–1491. 21 Hehir, K. M., et al. (1996). Molecular characterization of replication-competent variants of adenovirus vectors and genome modifications to prevent their occurrence. J. Virol. 70: 8459 – 8467. 22 Zhu, J., et al. (1999). Characterization of replication-competent adenovirus isolates from large-scale production of a recombinant adenoviral vector. Hum. Gene Ther. 10: 113–121. 23 Hearing, P., and Shenk, T. (1983). Functional analysis of the nucleotide sequence surrounding the cap site for adenovirus type 5 region E1A messenger RNAs. J. Mol. Biol. 167: 809 – 822. 24 Hitt, M. M., and Graham, F. L. (1990). Adenovirus E1A under the control of heterologous promoters: Wide variation in E1A expression levels has little effect on virus replication. Virology 179: 667– 678. 25 Wersto, R. P., et al. (1998). Recombinant, replication-defective adenovirus gene transfer vectors induce cell cycle dysregulation and inappropriate expression of cyclin proteins. J. Virol. 72: 9491–9502. 26 Zhou, H., et al. (1996). Development of a complementing cell line and a system for construction of adenovirus vectors with E1 and E2a deleted. J. Virol. 70: 7030 –7038. 27 Gorziglia, M., et al. (1996). Elimination of both E1 and E2a from adenovirus vectors further improves prospects for in vivo human gene therapy. J. Virol. 70: 4173– 4178. 28 Amalfitano, A., et al. (1998). Production and characterization of improved adenovirus vectors with the E1, E2b, and E3 genes deleted. J. Virol. 72: 926 –933. 29 Lusky, M., et al. (1998). In vitro and in vivo biology of recombinant adenovirus vectors with E1, E1/E2A, or E1/E4 deleted. J. Virol. 72: 2022–2032. 30 Zhou, H., and Beaudet, A. (2000). A new vector system with inducible cell line E2T for production of safer and higher titer adenoviral vectors. Virology 275: 348 –357. 31 Rittner, K., et al. (1997). Conditional repression of the E2 transcription unit in E1–E3deleted adenovirus vectors is correlated with a strong reduction in viral DNA replication and late gene expression in vitro. J. Virol. 71: 3307–3311. 32 Fang, B., et al. (1996). Lack of persistence of E1⫺ recombinant adenoviral vectors containing a temperature-sensitive E2A mutation in immunocompetent mice and hemophilia B dogs. Gene Ther. 3: 217–222. 33 Morral, N., et al. (1997). Immune responses to reporter proteins and high viral dose limit duration of expression with adenoviral vectors: Comparison of E2a wild type and E2a deleted vectors. Hum. Gene Ther. 8: 1275–1286. 34 Amalfitano, A. (1999). Next-generation adenoviral vectors: New and improved. Gene Ther. 6: 1643–1645. 35 Mitani, K., et al. (1995). Rescue, propagation, and partial purification of a helper virus-dependent adenovirus vector. Proc. Natl. Acad. Sci. USA 92: 3854 –3858.
622
36 Chen, L., Anton, M., and Graham, F. L. (1996). Production and characterization of human 293 cell lines expressing the site-specific recombinase Cre. Somatic Cell Mol. Genet. 22: 477– 488. 37 Parks, R. J., et al. (1996). A helper-dependent adenovirus vector system: Removal of helper virus by Cre-mediated excision of the viral packaging signal. Proc. Natl. Acad. Sci. USA 93: 13565–13570. 38 Morral, N., et al. (1998). High doses of a helper-dependent adenovirus vector yield supraphysiological levels of a1-antitrypsin with negligible toxicity. Hum. Gene Ther. 9: 2709 –2716. 39 Schiedner, G., et al. (1998). Genomic DNA transfer with a high-capacity adenovirus vector results in improved in vivo gene expression and decreased toxicity. Nat. Genet. 18: 180 –183. 40 Morsy, M. A., et al. (1998). An adenoviral vector deleted for all viral coding sequences results in enhanced safety and extended expression of a leptin transgene. Proc. Natl. Acad. Sci. USA 95: 7866 –7871. 41 Zou, L., et al. (2001). Helper-dependent adenoviral vector-mediated gene transfer in aged rat brain. Hum. Gene Ther. 12: 181–196. 42 Chen, H. H., et al. (1997). Persistence in muscle of an adenoviral vector that lacks all viral genes. Proc. Natl. Acad. Sci. USA 94: 1645–1650. 43 Kumar-Singh, R., and Chamberlain, J. S. (1996). Encapsidated adenovirus minichromosomes allow delivery and expression of a 14 kb dystrophin cDNA to muscle cells. Hum. Mol. Genet. 5: 913–921. 44 Burcin, M. M., et al. (1999). Adenovirus-mediated regulable target gene expression in vivo. Proc. Natl. Acad. Sci. USA 96: 355–360. 45 Zou, L., et al. (2000). Prolonged transgene expression mediated by a helper-dependent adenoviral vector (hdAd) in the central nervous system. Mol. Ther. 2: 105–113. 46 Verma, I. M. (2000). Gene therapy: The need for basic science. Mol. Ther. 2: 531. 47 Parks, R., Evelegh, C., and Graham, F. (1999). Use of helper-dependent adenoviral vectors of alternative serotypes permits repeat vector administration. Gene Ther. 6: 1565– 1573. 48 Parks, R. J., et al. (1999). Effects of stuffer DNA on transgene expression from helperdependent adenovirus vectors. J. Virol. 73: 8027– 8034. 49 Sandig, V., et al. (2000). Optimization of the helper-dependent adenovirus system for production and potency in vivo. Proc. Natl. Acad. Sci. USA 97: 1002–1007. 50 Ramirez-Solis, R., Liu, P., and Bradley, A. (1995). Chromosome engineering in mice. Nature 378: 720 –724. 51 Resnitzky, D., et al. (1994). Acceleration of the G1/S phase transition by expression of cyclins D1 and E with an inducible system. Mol. Cell. Biol. 14: 1669 –1679. 52 Wang, Y., et al. (1994). A regulatory system for use in gene transfer. Proc. Natl. Acad. Sci. USA 91: 8180 – 8184. 53 Uetsuki, T., et al. (1989). Isolation and characterization of the human chromosomal gene for polypeptide chain elongation factor-1␣. J. Biol. Chem. 264: 5791–5798. 54 Parks, R., and Graham, F. L. (1997). A helper-dependent system for adenovirus vector production helps define a lower limit for efficient DNA packaging. J. Virol. 71: 3293–3298. 55 Graham, F. L. (1991). Manipulation of adenovirus vectors. In Methods in Molecular Biology (E. J. Murray, Ed.), pp. 109 –128. Humana Press, Clifton, NJ. 56 Maizel, J. V., Jr., White, D. O., and Scharff, M. D. (1968). The polypeptides of adenovirus. I. Evidence for multiple protein components in the virion and a comparison of types 2, 7A, and 12. Virology 36: 115–125. 57 Southern, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98: 503–517. 58 Bett, A. J., et al. (1994). 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: 8802– 8806. 59 Christ, M., et al. (1997). Gene therapy with recombinant adenovirus vectors: Evaluation of the host immune response. Immunol Lett. 57: 19 –25.
MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
METHOD
A Cre-Expressing Cell Line and an E1/E2a Double-Deleted Virus for Preparation of Helper-Dependent Adenovirus Vector Heshan Zhou,*,†,‡,1 Tiejun Zhao,*,‡ Lucio Pastore,† Maged Nageh,† Wendy Zheng,*,‡ X. Mei Rao,* and Arthur L. Beaudet†,‡ *Center for Cell and Gene Therapy, †Department of Molecular and Human Genetics, and ‡ Department of Pediatrics, Baylor College of Medicine, Houston, Texas 77030 Received for publication January 5, 2001; accepted in revised form February 9, 2001; published online April 16, 2001
Adenoviral vectors are attractive for the delivery of transgenes into mammalian cells because of their efficient transduction, high titer, and stability. The major concerns with using E1-deleted adenoviral vectors in gene therapy are the pathogenic potential of the virus backbone and the leaky viral protein synthesis that leads to host immune responses and a short duration of transgene expression. Helper-dependent (HD) adenoviral vectors that are devoid of all viral protein-coding sequences have significantly increased the safety and reduced the immunogenicity of these vectors. Currently available HD vectors depend on an E1-deleted adenovirus as a helper to provide viral proteins in trans. As a consequence, contamination with helper virus cannot be avoided in the HD vector preparation though it can be decreased to 0.01% using a Cre/loxP mechanism. Since the presence of E1-deleted helper virus may have substantial unwanted effects, we have developed a new Cre-expressing cell line based on an E1- and E2a-complementing cell. This new cell line can efficiently cleave the packaging region in the helper virus genome. We have also developed an E1 and E2a double-deleted helper virus. By using the CreE cell with the helper virus deleted in both the E1 and the E2a genes it may be possible to further improve the safety of the vectors.
INTRODUCTION The human adenovirus genome is about 36 kb of doublestranded linear DNA. Adenoviral (Ad) vectors are particularly attractive for in vivo and in vitro gene delivery, primarily because of their highly efficient gene transfer and relatively easy production of stable and high-titer vectors. Numerous studies using Ad vectors in a variety of animal models have demonstrated successful gene transfer to many tissues, with high levels of recombinant gene expression, making these vectors attractive delivery systems to treat a variety of human diseases (1–5). The major concerns with using the E1-deleted Ad vectors are the pathogenic potential of the adenovirus backbone and the transient expression of the therapeutic gene caused by immune responses against viral-transduced cells expressing low levels of Ad genes (1, 4, 6 –9). The E1 region encodes proteins for the regulation of other viral genes (10). Mutation of the E1 region does not completely eradicate the synthesis of viral proteins and DNA (11, 12). The host cells may provide cellular “E1-
1 To whom correspondence and reprint requests should be addressed at the Center for Cell and Gene Therapy, Baylor College of Medicine, One Baylor Plaza, Room N1010, Houston, TX 77030. Fax: (713) 7981230. E-mail: [email protected].
MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy 1525-0016/01 $35.00
like” factors, such as the 70-kDa human heat shock protein (13) or interleukin-6 (14, 15), to substitute E1 functions. When cells respond to “emergency” stimuli such as viral infection or heat, they may be more likely to produce or activate cellular E1-like factors. These cellular factors can cause unwanted synthesis of viral proteins and replication of the viral DNA. Recent gene therapy studies have indicated that leaky expression of viral genes from E1deleted vectors can occur in transduced cells (16, 17). In addition, replication-competent adenovirus (RCA) may emerge during Ad vector propagation. The E1-deleted vectors are usually produced in 293 cells generated by the transfection of human kidney cells with Ad serotype 5 (Ad5) DNA (18). Mapping of the Ad sequences in the 293 cell line shows the presence of contiguous Ad5 sequences from the left-hand end of the genome up to position 4344 (19). When typical E1-deleted vectors, which contain a deletion from 340 to 3500 of the Ad genome, are propagated in 293 cells, sequence overlap between the vector and the complementing cell frequently results in the generation of RCA (20 –22). Since adenoviruses can replicate very well even at extremely low levels of E1 (23, 24), even limited contamination with E1-positive RCA or a low level of cellular E1-like factors may have catastrophic consequences. Recent studies have also demonstrated that E1-deleted vectors could induce cell G2 growth arrest,
613
METHOD which may produce unwanted effects during human gene therapy (25). One strategy to overcome such limitation is to develop vectors with deletions of additional essential viral genes. Ad vectors with deleted E1/E2 or E1/E4 genes can be propagated in cell lines that complement both E1 and E2 or E1 and E4 (26 –30). Unlike the E1 products, the E2a encodes for a DNA binding protein directly involved in viral DNA replication (10). No cellular factor has been found that can functionally replace the E2a product. Ad vectors with deleted E2 do not express detectable adenoviral early or late proteins, do not synthesize viral DNA in noncomplementing cells, and demonstrate improved safety (27, 29, 31). However, the E1/E2-deleted vectors still stimulate host immune responses for unknown reasons, which may benefit vaccine development but limit the duration of transgene expression (32–34). Another approach to reduce the immunogenicity and improve the safety of Ad vectors is to delete all viral coding sequences so that leaky expression of viral proteins is completely eliminated. Helper-dependent (HD; also called “gutless”) vector systems have been developed in which all or most of the viral protein-encoding sequences have been deleted (35). An improved HD system involves a 293-derived cell line (293Cre4) that stably expresses the bacterial phage P1 Cre recombinase and an E1-deleted helper virus that provides the necessary viral proteins in trans for the replication and packaging of the HD vectors (36, 37). The helper virus contains two loxP sites flanking a viral region that is essential for packaging virus DNA into virions. Upon infection of 293Cre4, the packaging signal in the helper virus DNA is excised through the Cre/loxP interaction, so that the helper virus DNA cannot be packaged. Ad HD vectors are substantially less toxic than other Ad vectors studied to date (38 – 41). Long-term transgene expression of HD vectors in the liver and other organs has been reported with experimental animals (39, 40, 42– 45). Even though first-generation adenoviral vectors have lost their initial luster, HD vectors are emerging as one of the most promising vector systems for gene therapy (46). Clinical trials of HD vectors are dependent on the production of safe HD vectors with high yield. With the current HD Ad vector systems, contamination with helper virus in the final purified HD vector has been decreased to 0.01%, but cannot be avoided (37, 47). When HD vectors are used in clinical trials that may involve 1012 or even more functional units of the HD vector, 108 or more infectious helper virus units may thus be administered. Contamination with E1-deleted helper virus may cause complications in clinical trials. To further improve Ad HD vector production and safety, the E3 and the packaging region in the helper virus and the backbone structure of the HD vector have been modified (48, 49). In this study, we have addressed the current limits of the HD vector system by developing a new Cre-expressing cell line based on E2T, an E1- and E2a-complementary cell line (30). One of the cell lines can efficiently excise the packaging region of the helper virus. We also developed
614
an E1 and E2a double-deleted helper virus, which can be used with the new cell line to produce HD vector with low helper contamination. This system may further improve the HD vector safety.
MATERIALS
AND
METHODS
Cell lines and cell culture. 293 cells contain chromosome-integrated Ad5 E1 sequences complementing the functions of the Ad E1 genes (18). 293Cre4 is a 293-derived cell line that expresses Cre recombinase and is used for the production of HD vectors (36, 37). E2T is an E1 and E2a double-complementing cell line derived from 293 cells (30). Expression of the E2a gene in E2T is inducible upon withdrawal of tetracycline. Several Cre-expressing cell lines were developed in this study by transfection of Cre-expressing cassettes into 293 or E2T cells. All cells were grown in ␣MEM (GIBCO, Grand Island, NY) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), 100 units of penicillin, and 100 g of streptomycin per milliliter. E2T and E2T-derived cells were maintained in this medium containing also 10 g/ml tetracycline to suppress E2a expression. Plasmids. The plasmid pOG231 contains a Cre-expressing cassette and was successfully used in chromosome engineering (50). The plasmid pUHD10-3 contains an inducible promoter regulated by tetracycline (51). A DNA fragment encoding the Cre open reading frame (ORF) was cleaved from pOG231 with restriction enzymes BglII and SalI and inserted into pUHD10-3 to construct plasmid pBZ73. The Cre ORF in pBZ73 was under the regulation of the tetracycline-inducible promoter. Two plasmids with drug-resistant genes were used to select Cre-expressing cell colonies. The PGL-VP plasmid contains the hygromycin-resistance gene (52). The pPGKpurobpA plasmid, containing the puromycin-resistance gene, was kindly provided by Dr. Allan Bradley. The above Cre-expressing and selective plasmids were used to transfect 293 and E2T cells. The plasmid pLoxZ was constructed by separating the EF-1␣ promoter (53) and the Escherichia coli -galactosidase (lacZ) ORF with two loxP sites, which sandwich a segment containing a stop codon (Fig. 1). Expression of lacZ from this plasmid was inhibited because of premature termination of translation at the stop site. Deletion of the DNA segment between the loxP sites through Cre/loxP interaction activated the expression of lacZ (Fig. 1). This plasmid was used to screen the Cre-expressing cells. Plasmid pDelta28 is an HD backbone plasmid containing the Ad left inverted terminal repeat (ITR) and packaging signals (Ad5, 1– 440 bp) and the right ITR (Ad5, 35818 –35935 bp) (L. Pastore and A. Beaudet, unpublished data). A DNA “stuffer” was included in pDelta28 to achieve a final vector size close to the natural Ad5 size for increasing the vector stability (54). For this purpose, a 16,054-bp EclXI/PmeI fragment from the HPRT gene (GenBank Accession No. Humhprtb) and an 11,105-bp BamHI fragment from the cosmid C346 gene (GenBank Accession No. L31948) were inserted into the pDelta28 construct. A HD vector in plasmid format was constructed based on the plasmid pDelta28. The -geo gene (a fusion of -galactosidase and neomycin-resistance genes) with a strong mammalian promoter, SR␣, was cleaved from the plasmid pSR␣-geo (35) and subcloned into the unique AscI site of the pDelta28 plasmid, resulting in the plasmid p⌬-geo. Transfection and Cre cell selection. Transfection or cotransfection required calcium phosphate precipitation (55). For production of Cre-expressing cells, E2T cells were cotransfected with the Cre-expressing plasmid pBZ73 and the hygromycin-resistant plasmid PGL-VP. Two days after transfection, the cells were selected in medium containing 50 g/ml hygromycin. The medium was changed every 3 to 5 days. 293 cells were cotransfected with pOG23 and pPGKpurobpA and selected in 1 g/ml puromycin. To screen Cre-expressing cell clones, drug-resistant clones cultured in 96-well plates were transfected with pLoxZ plasmid DNA. Two days after transfection, the cells were stained with 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal). Only the cells that expressed Cre recombinase stained blue. One pair of PCR primers was designed to assay the excision of the packaging region of the Ad helper virus propagated in Cre cells. The primers, Ad121F (5⬘-GAT GTT GCA AGT GTG GC) and Ad3704R (5⬘-GGA GCC CAT CAC ATT CTG), flank the adenoviral packaging region in the MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
METHOD helper virus. Various cells were infected with the helper virus at a multiplicity of infection (m.o.i.) of 1. Two days after infection, the cells were collected and analyzed by PCR, using reaction conditions of 94°C for 2 min, followed by 30 cycles 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min. Helper virus and HD vector. The helper virus AdLC8cluc contains two loxP sites flanking the Ad packaging signal and the expression cassette of firefly luciferase inserted in the E3-deleted region of the Ad genome (37). This helper virus contains a deletion in the E1 region. The HD vector, Ad⌬-geo, was converted from plasmid p⌬-geo DNA in the CreE cells developed in this study. The plasmid p⌬-geo was first cleaved with PmeI to free the adenoviral ITRs and release the HD vector genome and then transfected into CreE cells. The second day after transfection, the CreE cells were infected with the helper virus AdLC8cluc that provides viral proteins in trans to rescue the HD Ad⌬-geo as previously described (37). The HD Ad⌬-geo was then amplified through a series of co-infections with the helper virus in CreE cells. The amplification of HD Ad⌬-geo could be easily monitored by assaying lacZ expression in the vector. The blue cells after each passage were directly counted in the plates after X-Gal staining as follows: the cells with 90% cytopathic effect (CPE) after co-infection with the helper virus and the HD vector were collected; the residual attached cells were fixed with 0.2% glutaraldehyde and 2% formaldehyde in PBS for 5 min at room temperature and then incubated at 37°C overnight with 0.1% X-Gal in a solution of 5 mM K4Fe(CN)6, 5 mM K3Fe(CN)6, and 2 mM MgCl2. Purification of the Ad⌬-geo vector was carried out by two rounds of CsCl centrifugation (55) and passage through desalting columns (EconoPac, 10 DG; Bio-Rad, Hercules, CA). The HD vector preparations were evaluated by total viral particle count, as determined by optical density measurements of DNA (56), and by measurement of blue forming units (BFU) of the HD Ad⌬-geo vector in the E2T cells. The quantity of contaminating helper viruses in the HD vector preparation was determined by measurement of infective units (IFU) in 96-well plates as described previously (49). Southern blot analysis of helper virus and HD vector. Various cells in 10-cm dishes were infected with the helper virus AdLC8cluc at an m.o.i. of 1. After 2 days, the cells were collected, washed with PBS, and digested overnight with 0.5 ml of proteinase K (10 mM Tris–HCl (pH 7.5)/10 mM EDTA/0.5% SDS/0.04% proteinase K) at 50 –C. The total DNA was then extracted once with buffer-saturated phenol:chloroform (Amresco, Solon, OH) and precipitated with ethanol. For isolation of the vector DNA from the CsCl-purified preparation, 1 l of the purified vector was mixed with 2 g of salmon DNA (as a carrier) and incubated in 100 l of the proteinase K solution before purification as described above. The isolated DNA samples were then cleaved with the restriction endonuclease PstI, separated on agarose gels, and transferred to Hybond-N positively charged nylon membranes (Amersham International plc, Amersham, UK) by capillary transfer (57) for hybridization with a radioactively labeled DNA probe.
RESULTS Isolation and Screening of Cre Cell Clones E2T cells were cotransfected with the plasmid pBZ73 containing the Cre expression cassette and the plasmid pGL-VP with the hygromycin-resistance marker. 293 cells were cotransfected with the Cre plasmid pOG231 and the plasmid pPGKpuropbA with the puromycin-resistance gene. The drug-resistant clones were isolated and then cultured in duplicated 96-well plates. One set of the plates was used for analysis of Cre activity, and the other set was saved for further analyses. To facilitate screening of Creexpressing cell clones, we designed a method based on the Cre/loxP-mediated deletion activating expression of the lacZ gene. The plasmid pLoxZ contains a loxP–stop–loxP segment that separates the promoter and the lacZ ORF. Two days after transfection with pLoxZ, the cells were MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
stained with X-Gal. Only the cells that express Cre recombinase can remove the loxP-flanked segment in the plasmid, resulting in expression of the lacZ gene (Fig. 1). Without overstaining, the number of blue-positive cells and the color density can partially reflect the levels of Cre expression in the cells. We found that this method can be easily and reliably used for initial screening of the Crepositive cells. Several hundred clones were checked using this method. We picked up several clones, most with high numbers of dense blue cells (293-Cre8, E2T-Cre2, E2TCre4, E2T-Cre5, and E2T-Cre6) and a few with low numbers of light blue cells (E2T-Cre7 and E2T-Cre9) for further analysis with more accurate methods.
Deletion of Helper Virus Packaging Region in Cre Cells Since deletion of the helper virus packaging region is critical for the preparation of highly purified HD vector, the efficiency of Cre/loxP-mediated deletion was further evaluated using the Southern blot hybridization method. The total DNA samples were isolated from the 293, E2T (both Cre negative controls), 293Cre4 (Cre positive), and Cre-positive clone cells infected with AdLC8cluc. The DNA was then digested with the restriction enzyme PstI, which cleaved a 0.53-kb fragment from the left terminus and a 2.06-kb fragment from the right end of the helper virus (Fig. 2). The hybridization probe was a 0.14-kb fragment generated from pXCJL.2 (55) with EcoRI and AflIII. This probe, covering the 103-bp ITRs, can hybridize to both the 2.06- and the 0.53-kb terminal fragments. Deletion of the packaging signal (⌿) in the helper virus through the Cre/loxP interaction will shorten the 0.53-kb fragment to a 0.26-kb fragment as shown in the positivecontrol 293Cre4 cells (Fig. 2). Hence, the ratio of the 0.26-kb band to the 0.53-kb band represents the excision efficiency of the packaging signal by Cre recombinase in the cells. The 2.06-kb band representing the right terminus of helper virus acts as an internal control for equal loading. For 293 and E2T cells without Cre expression, there was no deletion of the packaging signal in the helper virus (Fig. 2). In the Cre-positive control 293Cre4 cells, most of the packaging signal was deleted, with about 10% of 0.53-kb fragment remaining. The excision of the packaging region in the screened cells was compared with the control 293Cre4 samples. 293-Cre14, E2T-Cre2, E2TCre7, and E2T-Cre9 were less efficient than the control 293Cre4 in deleting the packaging region; 293-Cre8 and E2T-Cre4 were similar to 293Cre4 and E2T-Cre5 and E2TCre6 were more efficient than 293Cre4. These results generally agreed with the results obtained by using plasmid pLoxZ. In E2T-Cre9, there was an extra band about 0.8 kb (Fig. 2) that also existed in repeated Southern blot assays. We do not know where this band comes from. This clone with several others was not used in the HD vector preparation and not further studied. Excision of the packaging region from the helper virus in E2T-Cre6 cells was highly efficient, and the 0.53-kb fragment could hardly be detectable as most of the DNA converted to the 0.26-kb fragment (Fig. 2). This result suggested that the E2T-Cre6
615
METHOD
FIG. 1. (A) Screening of the Cre-expressing cell clones by Cre/loxP deletion activation. The Cre/loxP activation mechanism is shown at the top. Plasmid pLoxZ contains two loxP sites that flank a DNA segment with a stop codon. Expression from the lacZ gene in a Cre-negative cell (E2T) is blocked because of premature termination. Excision of the segment between the loxP sites in Cre-expressing cells (E2T-Cre6) activates the expression of the lacZ gene. (B) Resistance to helper virus infection. 293, 293Cre4, E2T-Cre4, and E2T-Cre6 cells were infected with the helper virus AdLC8cluc at an m.o.i. of 0.25. Efficient deletion of the viral packaging region in E2T-Cre6 cells prevented packaging of the viral DNA into infectious particles. The helper virus spread inefficiently in the E2T-Cre6 culture.
616
MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
METHOD
FIG. 2. Southern blot assay for deletion of the packaging region in the helper virus DNA. Total DNA was isolated from different cells infected with the Ad helper virus AdLC8cluc and cleaved with restriction enzyme PstI. (A) The probe containing the Ad ITR could hybridize with both Ad terminal fragments. Deletion of the packaging region in the helper virus decreased the left terminal fragment from 0.53 to 0.26 kb. The size of the right terminal fragment did not change. The DNA isolated from Cre-negative 293 and E2T cells showed undeleted fragments, while the packaging region was deleted in the 293Cre4 cells. (B) The efficiency of the Cre/loxP-mediated deletion in different cell lines is shown.
clone derived from E2T could efficiently excise the packaging region from the helper virus DNA. The E2T-Cre6 was renamed CreE.
CreE Resistance to Helper Virus Infection Since the helper virus DNA deleted in the packaging region could not be packaged into viral particles, we reasoned that infectious virions would not efficiently propagate in CreE cells. Based on this concept, when the CreE cells were infected with the helper virus at a low m.o.i., the uninfected cells would survive because new virus would not be efficiently produced from the infected cells. To evaluate this concept, 293, 293Cre4, E2T-Cre4, and CreE cells were infected with the helper virus at an m.o.i. of 0.25. All 293 cells, and most 293Cre4 and E2T-Cre4 MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
cells, showed CPE at 2 days postinfection, while only a small portion of E2T-Cre6 cells were killed (Fig. 1B). This result, combined with the data from the Southern blot analysis (Fig. 2), suggested that the infectious helper virus could be quickly cleaned out by passage in CreE cells.
Large-Scale Preparation of Highly Purified HD Vector in CreE Cells HD Ad⌬-geo was converted from plasmid p⌬-geo according to the method in (37). The plasmid p⌬-geo DNA was first linearized with restriction enzyme PmeI and then transfected into CreE cells. The helper virus AdLC8cluc was added to the cell culture to provide necessary viral proteins for rescuing the HD vector. A serial co-infection of the HD vector with the helper virus in CreE cells re-
617
METHOD sulted in an increase in the titer of the HD vector that was easily monitored by X-Gal staining. With the CreE cell line, rescuing and amplification of the HD vector were similar to those with the 293Cre4 cell line. To manufacture HD vectors on a large scale, 16 triplelayer flasks (500 cm2 per flask) of CreE cells were cultured until about 90% confluence. The cells were co-infected with both HD vector Ad⌬-geo, m.o.i. of 10 (BFU), and helper virus AdLC8cluc, m.o.i. of 5 (IFU). The infected cells were harvested when over 90% showed CPE (about 3 days). We made a total of six preparations by using AdLC8cluc. Each preparation yielded 3 ml of HD vector with 0.7 to 1.2 ⫻ 1012 particles/ml. The vector preparation was diluted and titered on E2T cells in 96-well plates for its infectious activity. The HD vector titer (BFU) counted after X-Gal staining was between 5 ⫻ 1010 to 1 ⫻ 1011 BFU/ml. The helper virus titer was determined as IFU causing CPE 14 days postinfection. The helper virus was 105/ml IFU or lower in the purified HD vector preparation. Therefore the helper:vector ratio is about 10⫺5 or lower (IFU:BFU). To further determine contamination of the helper virus in the HD vector preparations, the DNA was purified from the above vector preparations for Southern blot analysis. The DNA was digested with PstI and hybridized with the probe with Ad ITR as used in analysis of the deletion of the Ad packaging region. Since the helper virus DNA deleted of its packaging region cannot be packaged into virions, the DNA isolated from the purified HD vector preparation represented only the DNA from virions. The control having only the helper virus DNA was isolated from Cre-negative E2T cells infected with AdLC8cluc. The largest (2.06 kb) and the smallest (0.53 kb) fragments represent the helper virus DNA right and left terminal segments, which were also shown in Fig. 2. The helper virus contamination was barely detectable in the HD vector preparations purified from the CreE cells (Fig. 3).
E1 and E2a Complementation of CreE Cells To test the E1 and E2a complementation of CreE cells, we used our previously constructed Adgal⌬E1 (deleted E1) Adgal⌬E1E2 (deleted at both E1 and E2a) vectors (26) to infect 293, E2T, and CreE cells at an m.o.i. of 1. The total cells were collected 2 days after infection; lysates were prepared and serially diluted to determine the titers of BFU on E2T cells in 96-well plates. The titers of the Adgal⌬E1 containing wild-type E2a were similar regardless of the cell lines, because the Adgal⌬E1 did not depend on E2a complementation from the cells. The E2adeleted vector did not replicate in 293 cells, as the titer of Adgal⌬E1E2 vector in 293 cells was only 1% of the input virus. However, the Adgal⌬E1E2 vector titer in CreE6 cells was increased as it was in E2T cells (Fig. 4).
A Helper Virus Deleted at both E1 and E2a Two loxP sites were introduced into the shuttle plasmid p⌬E1sp1A (58) to flank the Ad package region. The two
618
loxP sites are at locations similar to those of the published helper virus AdLC8cluc (37). The resulting shuttle plasmid was cotransfected into E2T cells with plasmid pBZ71, which is an Ad backbone plasmid for the construction of E1- and E2a-deleted vectors (30). A helper virus with deletions of E1 and E2a was rescued and named Adhv⌬E1/E2 (Fig. 4). Two PCR primers (Ad121F and Ad3704R) that flank the packaging region of adenovirus were designed to analyze this region. Three viruses, Ad⌬E1, AdLC8cluc, and Adhv⌬E1/E2, were used to infect Cre-negative E2T cells and Cre-positive E2T-Cre6 cells. The E1-deleted Ad⌬E1 (unpublished work, H. Zhou), identical to Adhv⌬E1E2 in the deleted E1 region but without inserted loxP, was used as negative control. The helper virus AdLC8cluc, containing two inserted loxP sites, was used as positive control. PCRs with Ad⌬E1 DNAs isolated from both Cre-negative and Cre-positive cells produced a 0.4-kb fragment (Fig. 5). Insertion of two loxPs in AdLC8cluc and Adhv⌬E1E2 increased the size of the PCR fragments. When the two helpers were cultured in CreE cells, deletion of the packaging region through loxP and Cre interaction resulted in shorter fragments (Fig. 5). This suggested that the Ad packaging region Adhv⌬E1E2 could be efficiently deleted. With the new CreE cell line, the E1 and E2a doubledeleted helper Adhv⌬E1E2 could be used to replace the E1-deleted helper virus AdLC8cluc to produce the HD vector. The method for preparation of the HD vector with the Adhv⌬E1E2 was the same as when the AdLC8cluc was used. We made a total of three preparations with the helper Adhv⌬E1E2. Each preparation yielded 3 ml of HD vector with 1 to 5 ⫻ 1011 particles/ml. The HD vector titer (BFU) was between 0.8 and 2 ⫻ 1010 BFU/ml. The helper virus was 104/ml IFU in the purified HD vector preparation. Therefore the helper:vector ratio is about 10⫺6 (IFU: BFU). The structure of the HD vector prepared with the E1 and E2a double-deleted helper was also analyzed. We had a result similar to that shown in Fig. 3 when the helper ADLC8cluc was used. When the HD Ad⌬-geo vector prepared in this study was injected into rat brain, the vector resulted in prolonged transgene expression and lower immune response compared with the first-generation vector (41, 45).
DISCUSSION The HD Ad vector system is one of the most promising virus vector systems for gene delivery. However, largescale preparation of highly purified and safe HD vector has proved difficult, limiting vector use for clinical trials. The construction and production of HD vectors are more complex than those of the first-generation vectors. There are three key components in the HD vector system: the Cre-expressing cells, the helper virus, and the vector itself. The growth conditions of the cells, the structure of helper virus and HD vector, and their titer used in co-infection may all affect the final vector yield and quality. In efforts to improve the vector system, both the helper virus and the vector structure have been modified (48, 49). In all the MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
METHOD
FIG. 3. Southern blot assay for purified HD vector DNA. (A) The probe DNA containing Ad ITR can hybridize with the terminal fragments of helper virus and HD vector DNA digested with the restriction enzyme PstI. (B) When the helper virus DNA was present, hybridization resulted in 2.06- and 0.53-kb bands as labeled. The contamination of helper virus DNA was hardly detectable in the HD vector preparations.
current HD vector systems, an E1-deleted Ad vector is used as a helper virus for providing viral proteins in trans. Contamination with E1-deleted helper virus in the HD-Ad vector preparations is unavoidable (37, 47). Based on our recently reported E1 and E2a complementing cell line (30), we have developed a new Creexpressing cell line (CreE) that could efficiently excise the packaging region (Fig. 2) and complement both E1 and E2a genes (Fig. 4). We have also developed an E1 and E2a MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
double-deleted helper virus (Adhv⌬E1E2) for production of HD vectors. The E2a gene encodes the single-stranded DNA binding protein that is essential for viral DNA replication and for transcription from the major late promoter (10). Previous studies confirmed that, unlike the E1-deleted vectors, vectors with deletions in both E1 and E2a do not express detectable adenovirus early and late proteins and do not synthesize viral DNA in noncomplementing cells (27, 29, 31, 59). Hence using the new HD
619
METHOD
FIG. 4. Comparison of the titers of viruses with wild-type E2a or with deletion of E2a on 293, E2T, and E2T-Cre6 cells. Titer of the virus with E1 deleted, Adgal⌬E1, is shown on the left, and titer of the virus with E1 and E2a deleted, Adgal⌬E1E2, is shown on the right. Viruses were collected from medium after 2 days of infection with an m.o.i. of 1, and the titer was measured as blue forming units (BFU) on E2T cells.
vector system can further increase the safety of the HD vector. The CreE cell could efficiently delete the packaging signal in a helper virus (Fig. 2) and therefore produced the HD vector with low contamination of the helper virus. The decreased helper contamination is dependent on the cell line Cre activity. The helper virus-to-HD vector ratio in the HD vector preparations was about 10⫺5 for AdLC8cluc and 10⫺6 for Adhv⌬E1E2 (IFU/BFU). The smaller Adhv⌬E1E2 ratio may due to slightly lower level of replication with the E1/E2a double-deleted virus. The yield of the HD vector using E1 and E2a double-deleted helper virus was slightly lower than that using AdLC8cluc helper, so that it will be necessary to further systematically study and optimize production conditions. The HD vector Ad⌬-geo was rescued and amplified with this system. To achieve a stock with amplified HD vector from a plasmid DNA, the system required six to seven rounds of coinfection with a helper virus (data not shown). We made several preparations of the HD Ad⌬-geo vector using the CreE cell line with both the helper virus AdLC8cluc and the helper Adhv⌬E1E2. The particle yield is similar to that seen when 293Cre4 cells were used. The vector quality prepared in this study was confirmed in animal studies that resulted in a stable transgene expression with reduced immune responses (41, 45). In this study, we have developed a new method for efficiently screening Cre cell lines according to the Cre/ loxP-induced deletion. In plasmid pLoxZ, two loxP sites separate the promoter and the lacZ ORF. When the plasmid is transfected into Cre-expressing cells, deletion through Cre/loxP results in active expression of the lacZ
620
gene. This method is reliable and can be easily used to screen Cre-expressing cells. We have also established a Southern blot method for analysis of HD vector preparation. With this method, only 1 l of purified HD vector is required for analysis. The probe containing the adenovirus ITR can hybridize with the terminal segments of both the HD vector and the helper virus, allowing the ratio of helper virus to HD vector to be directly compared. There is no a standard method to determine the ratio of a helper virus to an HD vector in final vector preparation. Plaque-forming unit and infective unit have been used to determine the helper virus titer (37, 49). The blue forming unit (for the lacZ gene) and functional unit, by comparing with a first-generation vector carrying the same transgene, have been used to determine the HD vector titer (49). We determined the helper:HD vector ratio in IFU: BFU. We should point out that all these methods are not directly comparing the helper virion number with the vector virion number. Southern blot method can directly compare the helper DNA with the vector DNA. However, when the helper DNA is lower than 1% of the HD vector DNA, the helper DNA may be not easily detected. Extending exposure of the Southern blot could increase the density of the weak helper virus bands, but then the HD vector bands would be overexposed. Working out a testing standard for the HD vector system is important and obviously requires collaborations among academia, industry, and some regulation agents. In the CreE cell line, Cre expression is under the regulation of a tetracycline-inducible mechanism. However, we observed no difference in packaging deletion in the CreE cells under inducible (minus tetracycline) or uninMOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
METHOD
FIG. 5. PCR analysis of deletion in the packaging region in helper viruses. Ad⌬E1, AdLC8cluc, and Adhv⌬E1E2 are E1-deleted viruses, while AdLC8cluc and AdhvDE1E2 contain two loxPs flanking the packaging region. The Ad⌬E1 without the inserted loxP was used as negative control; the AdLC8cluc containing two inserted loxP sites was used as positive control. PCR with Ad⌬E1 produced a 0.40-kb fragment regardless of the cell line. Insertion of two loxPs in AdLC8cluc and Adhv⌬E1E2 increased the fragment size. However, deletion of the packaging region through loxP/Cre interaction in E2T-Cre6 cells produced shorter fragments.
ducible (plus tetracycline) conditions (data not shown). One possibility is that the level of Cre even under uninduced conditions may already be at a biologically maximal level. In summary, a CreE cell line has been developed that can complement both E1 and E2a and efficiently excise the packaging signal in a helper virus. By using CreE cells, contamination of helper virus in HD vector preparations can be decreased. An E1 and E2a double-deleted helper virus has also been developed and can be used with the new cell line. Since the E2a gene is essential for adenovirus DNA replication and protein synthesis, a helper virus deleted for both E1 and E2a should further improve the safety of the HD vector system. ACKNOWLEDGMENTS We thank Dr. Malcolm Brenner for critically reading the manuscript and Thomas Gegeny of the Houston Academy of Medicine–Texas Medical Center Library’s publication service for editing the manuscript. The 293Cre4 cell line and helper virus AdLC8cluc (36, 37) were obtained from Merck & Co. This work was supported by grants from the NIH (HL 51751 and HL 59314), the Cystic Fibrosis Foundation (F984), the Texas Higher Education Coordination Board (ATP 004949-119), and the Shell Center for Gene Therapy at Baylor College of Medicine. MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy
REFERENCES 1 Stratford-Perricaudet, L. D., et al. (1990). Evaluation of the transfer and expression in mice of an enzyme-encoding gene using a human adenovirus vector. Hum. Gene Ther. 1: 241–256. 2 Rosenfeld, M. A., et al. (1991). Adenovirus-mediated transfer of a recombinant alpha 1-antitrypsin gene to the lung epithelium in vivo. Science 252: 431– 434. 3 Jaffe, H. A., et al. (1992). Adenovirus-mediated in vivo gene transfer and expression in normal rat liver. Nat. Genet. 1: 372–378. 4 Quantin, B., et al. (1992). Adenovirus as an expression vector in muscle cells in vivo. Proc. Natl. Acad. Sci. USA 89: 2581–2584. 5 Hitt, M. M., Addison, C. L., and Graham, F. L. (1997). Human adenovirus vectors for gene transfer into mammalian cells. Adv. Pharmacol. 40: 137–206. 6 McElvaney, N. G. (1996). Is gene therapy in cystic fibrosis a realistic expectation? Curr. Opin. Pulmon. Med. 2: 466 – 471. 7 Marshall, E. (1999). Gene therapy death prompts review of adenovirus vector. Science 286: 2244 –2245. 8 Engelhardt, J. F., et al. (1993). Direct gene transfer of human CFTR into human bronchial epithelia of xenografts with E1-deleted adenoviruses. Nat. Genet. 4: 27–34. 9 van Ginkel, F. W., et al. (1997). Adenoviral gene delivery elicits distinct pulmonaryassociated T helper cell responses to the vector and to its transgene. J. Immunol. 159: 685– 693. 10 Shenk, T. (1996). Adenoviridae: The viruses and their replication. In Fields Virology (B. N. Fields, D. M. Knipe, and P. M. Howley, Eds.), pp. 2111–2148. Raven Press, Philadelphia. 11 Winberg, G., and Shenk, T. (1984). Dissection of overlapping functions within the adenovirus type 5 E1A gene. EMBO J. 3: 1907–1912. 12 Shenk, T., and Williams, J. (1984). Genetic analysis of adenoviruses. Curr. Top. Microbiol. Immunol. 111: 1–39. 13 Imperiale, M. J., et al. (1984). Common control of the heat shock gene and early adenovirus genes: Evidence for a cellular E1A-like activity. Mol. Cell. Biol. 4: 867– 874.
621
METHOD 14 Spergel, J. M., and Chen-Kiang, S. (1991). Interleukin 6 enhances a cellular activity that functionally substitutes for E1A protein in transactivation. Proc. Natl. Acad. Sci. USA 88: 6472– 6476. 15 Rancourt, C., et al. (1999). Interleukin-6 modulated conditionally replicative adenovirus as an antitumor/cytotoxic agent for cancer therapy. Clin. Cancer Res. 5: 43–50. 16 Mittereder, N., et al. (1994). Evaluation of the efficacy and safety of in vitro, adenovirus-mediated transfer of the human cystic fibrosis transmembrane conductance regulator cDNA. Hum. Gene Ther. 5: 717–729. 17 Marienfeld, U., et al. (1999). ‘Autoreplication’ of the vector genome in recombinant adenoviral vectors with different E1 region deletions and transgenes. Gene Ther. 6: 1101– 1113. 18 Graham, F. L., et al. (1977). Characteristics of a human cell line transformed by DNA from human adenovirus type 5. J. Gen. Virol. 36: 59 –72. 19 Louis, N., Evelegh, C., and Graham, F. (1997). Cloning and sequencing of the cellular– viral junctions from the human adenovirus type 5 transformed 293 cell line. Virology 233: 423– 429. 20 Lochmuller, H., et al. (1994). Emergence of early region 1-containing replicationcompetent adenovirus in stocks of replication-defective adenovirus recombinants (DE1⫹DE3) during multiple passages in 293 cells. Hum. Gene Ther. 5: 1485–1491. 21 Hehir, K. M., et al. (1996). Molecular characterization of replication-competent variants of adenovirus vectors and genome modifications to prevent their occurrence. J. Virol. 70: 8459 – 8467. 22 Zhu, J., et al. (1999). Characterization of replication-competent adenovirus isolates from large-scale production of a recombinant adenoviral vector. Hum. Gene Ther. 10: 113–121. 23 Hearing, P., and Shenk, T. (1983). Functional analysis of the nucleotide sequence surrounding the cap site for adenovirus type 5 region E1A messenger RNAs. J. Mol. Biol. 167: 809 – 822. 24 Hitt, M. M., and Graham, F. L. (1990). Adenovirus E1A under the control of heterologous promoters: Wide variation in E1A expression levels has little effect on virus replication. Virology 179: 667– 678. 25 Wersto, R. P., et al. (1998). Recombinant, replication-defective adenovirus gene transfer vectors induce cell cycle dysregulation and inappropriate expression of cyclin proteins. J. Virol. 72: 9491–9502. 26 Zhou, H., et al. (1996). Development of a complementing cell line and a system for construction of adenovirus vectors with E1 and E2a deleted. J. Virol. 70: 7030 –7038. 27 Gorziglia, M., et al. (1996). Elimination of both E1 and E2a from adenovirus vectors further improves prospects for in vivo human gene therapy. J. Virol. 70: 4173– 4178. 28 Amalfitano, A., et al. (1998). Production and characterization of improved adenovirus vectors with the E1, E2b, and E3 genes deleted. J. Virol. 72: 926 –933. 29 Lusky, M., et al. (1998). In vitro and in vivo biology of recombinant adenovirus vectors with E1, E1/E2A, or E1/E4 deleted. J. Virol. 72: 2022–2032. 30 Zhou, H., and Beaudet, A. (2000). A new vector system with inducible cell line E2T for production of safer and higher titer adenoviral vectors. Virology 275: 348 –357. 31 Rittner, K., et al. (1997). Conditional repression of the E2 transcription unit in E1–E3deleted adenovirus vectors is correlated with a strong reduction in viral DNA replication and late gene expression in vitro. J. Virol. 71: 3307–3311. 32 Fang, B., et al. (1996). Lack of persistence of E1⫺ recombinant adenoviral vectors containing a temperature-sensitive E2A mutation in immunocompetent mice and hemophilia B dogs. Gene Ther. 3: 217–222. 33 Morral, N., et al. (1997). Immune responses to reporter proteins and high viral dose limit duration of expression with adenoviral vectors: Comparison of E2a wild type and E2a deleted vectors. Hum. Gene Ther. 8: 1275–1286. 34 Amalfitano, A. (1999). Next-generation adenoviral vectors: New and improved. Gene Ther. 6: 1643–1645. 35 Mitani, K., et al. (1995). Rescue, propagation, and partial purification of a helper virus-dependent adenovirus vector. Proc. Natl. Acad. Sci. USA 92: 3854 –3858.
622
36 Chen, L., Anton, M., and Graham, F. L. (1996). Production and characterization of human 293 cell lines expressing the site-specific recombinase Cre. Somatic Cell Mol. Genet. 22: 477– 488. 37 Parks, R. J., et al. (1996). A helper-dependent adenovirus vector system: Removal of helper virus by Cre-mediated excision of the viral packaging signal. Proc. Natl. Acad. Sci. USA 93: 13565–13570. 38 Morral, N., et al. (1998). High doses of a helper-dependent adenovirus vector yield supraphysiological levels of a1-antitrypsin with negligible toxicity. Hum. Gene Ther. 9: 2709 –2716. 39 Schiedner, G., et al. (1998). Genomic DNA transfer with a high-capacity adenovirus vector results in improved in vivo gene expression and decreased toxicity. Nat. Genet. 18: 180 –183. 40 Morsy, M. A., et al. (1998). An adenoviral vector deleted for all viral coding sequences results in enhanced safety and extended expression of a leptin transgene. Proc. Natl. Acad. Sci. USA 95: 7866 –7871. 41 Zou, L., et al. (2001). Helper-dependent adenoviral vector-mediated gene transfer in aged rat brain. Hum. Gene Ther. 12: 181–196. 42 Chen, H. H., et al. (1997). Persistence in muscle of an adenoviral vector that lacks all viral genes. Proc. Natl. Acad. Sci. USA 94: 1645–1650. 43 Kumar-Singh, R., and Chamberlain, J. S. (1996). Encapsidated adenovirus minichromosomes allow delivery and expression of a 14 kb dystrophin cDNA to muscle cells. Hum. Mol. Genet. 5: 913–921. 44 Burcin, M. M., et al. (1999). Adenovirus-mediated regulable target gene expression in vivo. Proc. Natl. Acad. Sci. USA 96: 355–360. 45 Zou, L., et al. (2000). Prolonged transgene expression mediated by a helper-dependent adenoviral vector (hdAd) in the central nervous system. Mol. Ther. 2: 105–113. 46 Verma, I. M. (2000). Gene therapy: The need for basic science. Mol. Ther. 2: 531. 47 Parks, R., Evelegh, C., and Graham, F. (1999). Use of helper-dependent adenoviral vectors of alternative serotypes permits repeat vector administration. Gene Ther. 6: 1565– 1573. 48 Parks, R. J., et al. (1999). Effects of stuffer DNA on transgene expression from helperdependent adenovirus vectors. J. Virol. 73: 8027– 8034. 49 Sandig, V., et al. (2000). Optimization of the helper-dependent adenovirus system for production and potency in vivo. Proc. Natl. Acad. Sci. USA 97: 1002–1007. 50 Ramirez-Solis, R., Liu, P., and Bradley, A. (1995). Chromosome engineering in mice. Nature 378: 720 –724. 51 Resnitzky, D., et al. (1994). Acceleration of the G1/S phase transition by expression of cyclins D1 and E with an inducible system. Mol. Cell. Biol. 14: 1669 –1679. 52 Wang, Y., et al. (1994). A regulatory system for use in gene transfer. Proc. Natl. Acad. Sci. USA 91: 8180 – 8184. 53 Uetsuki, T., et al. (1989). Isolation and characterization of the human chromosomal gene for polypeptide chain elongation factor-1␣. J. Biol. Chem. 264: 5791–5798. 54 Parks, R., and Graham, F. L. (1997). A helper-dependent system for adenovirus vector production helps define a lower limit for efficient DNA packaging. J. Virol. 71: 3293–3298. 55 Graham, F. L. (1991). Manipulation of adenovirus vectors. In Methods in Molecular Biology (E. J. Murray, Ed.), pp. 109 –128. Humana Press, Clifton, NJ. 56 Maizel, J. V., Jr., White, D. O., and Scharff, M. D. (1968). The polypeptides of adenovirus. I. Evidence for multiple protein components in the virion and a comparison of types 2, 7A, and 12. Virology 36: 115–125. 57 Southern, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98: 503–517. 58 Bett, A. J., et al. (1994). 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: 8802– 8806. 59 Christ, M., et al. (1997). Gene therapy with recombinant adenovirus vectors: Evaluation of the host immune response. Immunol Lett. 57: 19 –25.
MOLECULAR THERAPY Vol. 3, No. 4, April 2001 Copyright © The American Society of Gene Therapy