The catalytic component of human telomerase reverse transcriptase (hTERT) is not expressed in most primary somatic human cells, whereas the majority of cancer cells reactivate telomerase by transcriptional up-regulation of hTERT. Several studies demonstrated that the hTERT promoter can be used to restrict gene expression of E1-deleted replication defective adenoviral vectors to telomerase-positive cancer cells. In this study, a conditionally replicating adenovirus (hTERT-Ad) expressing E1A genes under control of a 255-bp hTERT-promoter was constructed. Additionally, an internal ribosomal entry site-enhanced green fluorescent protein cassette was inserted downstream of the E1B locus to monitor viral replication in vivo. Adenoviral replication of hTERT-Ad and enhancement of enhanced green fluorescent protein expression could be observed in all investigated telomerase-positive tumor cell lines. In contrast, hTERT-Ad infection of telomerase-negative primary human hepatocytes did not result in significant replication. The capability of hTERT-Ad to induce cytopathic effects in tumor cells was comparable with that of adenovirus wild type and significantly higher compared with ONYX-015, regardless of the p53 status of the tumor cells. Single application of low-dose hTERT-Ad to tumor xenografts led to significant inhibition of tumor growth, confirming the potential therapeutic value of conditionally replicative adenoviral vectors. These in vivo experiments also revealed that hTERT-Ad-mediated oncolysis was more efficient than ONYX-015 treatment. These results demonstrate that expression of E1A under transcriptional control of the hTERT promoter is sufficient for effective telomerase-dependent adenovirus replication as a promising perspective for the treatment of the majority of epithelial tumors.
Telomeres are essential elements at chromosome termini that preserve chromosomal integrity by preventing DNA degradation, end-to-end fusions, rearrangements, and chromosome loss (1, 2, 3). Each cell replication is associated with the loss of 30–150 bp of telomeric DNA that can be compensated by telomerase, an RNA-dependent DNA polymerase (4). Ectopic expression of hTERT,3 the catalytic subunit of telomerase, is capable to reconstitute telomerase activity in telomerase-negative cells, indicating that hTERT is the major determinant of telomerase activity in mammalian cells (5, 6, 7). Most human somatic cells exhibit neither hTERT expression nor telomerase activity, whereby the number of cell divisions is limited because of the reduction of telomeres to a critical length. In contrast to quiescent somatic cells, in highly proliferative cells, such as germ-line, hematopoetic stem, or transformed cancer cells, diverse molecular mechanisms are necessary to maintain telomere length. Although some tumors activate a yet unknown alternative mechanism of telomere extension, the majority (>90%) of human cancer cells acquire immortality by expression of the hTERT (8).
It has been shown that hTERT expression is regulated at the transcriptional level (9). The hTERT promoter is highly G/C rich and lacks both TATA and CAAT boxes but contains several putative binding sites for transcription factors involved in cellular proliferation and tumorigenesis. The hTERT core promoter contains two E boxes, and previous studies have shown that myc-max heterodimers bind to these motives and activate hTERT promoter transcription (10, 11). In contrast to the oncogene c-myc, overexpression of the p53 tumor suppressor protein results in significant repression of hTERT-promoter activity, which appears to be dependent on SP1-binding sites (12).
Previous studies demonstrated that the hTERT promoter is inactive in normal cells but is activated during carcinogenesis, thereby providing a promising tool for tumor-specific gene expression (13, 14, 15, 16, 17, 18, 19). By using hTERT core promoters, several studies demonstrated tumor restricted adenoviral expression of suicide genes or proapoptotic genes (20, 21, 22). However, these gene therapy strategies using nonreplicating adenoviral vectors have some limitations for clinical application because of the low transduction rate of cancer cells. In contrast to nonreplicating vectors, replicating viruses have the potential to overcome the hurdles of ineffective tumor cell transduction. Designed to specifically and effectively replicate in tumor cells, these vectors allow rapid lysis, spreading of infection, and oncolysis throughout the whole tumor mass (23).
In the present study, an adenoviral vector (hTERT-Ad) was constructed by replacing the internal adenoviral E1A promoter by a 255-bp hTERT promoter fragment. This altered transcriptional control of E1A expression was hypothesized to restrict adenoviral replication to telomerase-positive tumor cells. In addition, an IRES-EGFP cassette was inserted downstream of E1B to monitor viral replication in vivo. Replication of hTERT-Ad as well as enhancement of EGFP expression in various telomerase-positive tumor cell lines derived from different tissue origin could be demonstrated. In contrast, in telomerase-negative primary human hepatocytes, only replication of adenovirus wild type, but not of hTERT-Ad, could be observed. The ability of hTERT-Ad to lyse tumor cells was comparable with adenovirus wild type. Single administration of low-dose hTERT-Ad into rapidly growing s.c. tumor xenografts on nude mice led to more efficacious inhibition of tumor growth than ONYX-015, whose oncolytic properties have been intensively studied during the recent years. The results demonstrate that hTERT-Ad is a new conditionally replicating adenoviral vector with promising antitumor activity and a wide applicability in gene therapy of cancer.
MATERIALS AND METHODS
The human tumor cell lines CaCo2, HeLa, AGS, Huh7, and Hep3B and the human embryonal kidney cell line 293 were obtained from the American Type Culture Collection. All cells were maintained in growth medium (DMEM plus Glutamax; Life Technologies, Inc.) supplemented with 10% heat-inactivated fetal bovine serum (Life Technologies, Inc.), 100 units/ml penicillin, and 100 μg/ml streptomycin (Seromed) at 37°C in 5% CO2.
Primary human hepatocytes embedded in collagen layers were purchased from Cytonet (Hannover, Germany) and maintained in growth medium provided by the manufacturer.
Genetic Construction and Plasmids.
The plasmids pGL2-Basic and pGL2-Prom were purchased from Promega. Luciferase reporter plasmids hTERT255-T1 to hTERT255-T5 were constructed on the base of pGL2-Basic (Promega). The hTERT promoter DNA was excised from the plasmid pGL3B-TRTP (kindly provided by Horikawa; Ref. 24) with PvuII and XhoI and was cloned into the SmaI/XhoI sites of pGL2, resulting in the plasmid hTERT255. Double-stranded oligonucleotides 5′-ACGTGTAGTGTATTTATACCCGGTGA-GT-3′ were then inserted in excess into the Klenow-filled XhoI site of hTERT255. The resulting plasmids hTERT255-T1–T5 were sequenced for the number and orientation of inserted TATA motives.
The shuttle vector for hTERT-Ad was constructed by digestion of pHM3 with HindIII followed by Klenow fill in and religation, thus yielding an NheI site (pHM3Nhe I). Subsequently, the original MCS was replaced by an oligonucleotide containing restriction sites for NheI-HindIII-AseI-SalI-BamHI-XhoI-EcoRI-NotI-MluI-EcoRV. An expression cassette containing the adenoviral E1 region from bases 458 to 3533 sequentially followed by an IRES element and the cDNA for EGFP was cloned into the XhoI/NotI sites of the modified pHM3NheI, resulting in the plasmid pE1/IRES/EGFP. A poly(A) signal was PCR generated using pHOOK (Invitrogen) as template, 5′-TTAGGCGGGCGG-CCGCGATCAGCCTCGACTGTGCCTTCTAGTTGC-3′ as sense primer, and 5′-GTTAC-GCGTTCCCCAGCATGCCTGCTATTGTCTTCC-3′ as antisense primer. The resulting PCR fragment was MluI/NotI digested and inserted into the plasmid pE1/IRES/EGFP downstream of the EGFP sequence. The final shuttle plasmid pTERT-1 was cloned by inserting the PvuII/XhoI fragment of pGL3B-TRTP containing the hTERT core promoter into the NheI and XhoI sites of pE1/IRES/EGFP.
Replication and Cytolysis Assays.
Tumor cells were seeded at 1 × 105 cells/cm2 density and infected with adenoviral vectors at indicated MOI. At the time points indicated, EGFP expression was determined by fluorescence microscopy of living cell layers. Adenoviral replication was monitored by counting viral particles in freeze/thaw lysates of infected cells using the TCID method. Permissive 293 cells were seeded in 96-well plates (104 cells/well). The next day, eight serial 10-fold dilutions of freeze/thaw lysates from infected cells were subjected to the 96-well plate (100 μl/well). After 10 days of incubation, the observable CPE per dilution was counted, and the ratio between infected and unaffected wells was determined. The titer was calculated using the equation T (infective particles/milliliter) = 10sum of ratios + 0.5. Viral titers were then further confirmed by standard plaque assays and Rapid Titer Kit (Clontech), according to the manufacturer’s protocol.
Cristal violet staining was used to demonstrate adenoviral-associated CPE and subsequent destruction of the cell layer. Infected cell layers were rinsed briefly with PBS and then fixed for 10 min with 10% formalin in PBS. The cells were washed with distilled water and stained for 30 min with 0.1% crystal violet in 10% ethanol.
Tumor cells (3 × 105) were seeded in 60-mm dishes 16 h before transfection. Cells were transfected using calcium phosphate or LipofectAMINE (Life Technologies, Inc.) with a total amount of 5 μg of DNA per dish (filled up with pBluescript KS; Stratagene). After transfection (48 h), whole cell extracts were obtained according to standard procedures and measured for luciferase activity using a Berthold Lumat LB9501. Luciferase activity was normalized by cotransfection of 0.5 μg of CMV-β-gal.
Adenoviral Vector Construction and Preparation.
The recombinant adenovirus was constructed using the in vitro ligation system of Mizuguchi and Kay (25). The viral vector was generated by ligating the PI-Sce/I-CeuI fragment of the p-TERT-1 shuttle vector (see above) into pAdHM4. To generate infective particles, the resulting plasmid was digested with PacI and transfected into 293 cells. Transfected cells were incubated until a CPE became visible. The cells were collected and lysed in PBS by repeated freezing and thawing for virus release.
For preparation of high titer viral stocks, 2 × 108 293 cells at 90% confluency were infected at an MOI of 10. The infected cells were incubated until a strong CPE could be observed and ∼50% cells were detached. The cells were then collected, and viral particles were released by freezing and thawing. For further purification, the virus preparation was purified twice by CsCl density gradient centrifugation and stored according to standard protocols. Infectivity was determined by plaque assay or Rapid Titer Kit (Clontech).
Western Blot Analysis.
Infected hepatocytes and Huh7 cells were harvested and lysed with radioimmunoprecipitation assay buffer. Protein concentration of the obtained cell extracts was measured by Bio-Rad microassay (München, Germany). Protein (10 μg) was separated on a 10% SDS polyacrylamide gel and blotted onto HyBond N membrane (Millipore). Equal loading and extract quality were controlled by Coomassie staining. Adenoviral E1A expression was detected using rabbit anti-Ad2 E1A (sc-430; Santa Cruz Biotechnology) as primary antibody and HP-conjugated donkey antirabbit IgG as secondary antibody (711–035-152; PharMingen). The antigen-antibody complexes were visualized using the enhanced chemiluminescence detection system as recommended by the manufacturer (Amersham).
Animal Experiments and in Vivo Analysis of Viral Spread.
Pathogen-free male NMRI-nu/nu mice (age 6–8 weeks) were obtained from the Animal Research Institute of the Medizinische Hochschule Hannover. All animal experiments were performed according to German legal requirements.
Tumor xenografts were established by s.c. inoculation of 106 Huh7 cells into the flanks of nude mice. Infection experiments were carried out when the resulting tumor nodules reached a volume of ∼300 mm3. Adenoviral preparations used in these infection trials were prepared, purified, and titered as described above. Before infection, the virus was dialyzed twice against a solution containing 10 mm Tris-HCl (pH 8.0), 1 mm MgCl2, and 140 mm NaCl at 4°C. Tumor nodules were infected once with 1 × 109 infective particles in a total volume of 100 μl via local infiltration. For assessment of viral spread, tumors were harvested at the time points indicated. Thin slices of native tumor samples were screened for EGFP expression under a fluorescence microscope. Furthermore, tumor samples were embedded in TissueTec (Sakura) and shock frozen in liquid nitrogen. Slices (7 μm) were prepared from cryopreserved tumor specimen and fixed with 4% paraformaldehyde in PBS.
For immunohistochemical identification of virally infected cells, the slices were treated with a goat anti-Ad-hexon antibody (20-AG02; Fitzgerald), diluted 1:100 in 20% FCS/PBS. As secondary antibody, a FITC-conjugated antigoat IgG antibody (81-1611, diluted 1:100; Zymed) was used, and FITC-positive cells were monitored by fluorescence microscopy.
For statistical tumor growth, evaluation infected tumors were monitored by weekly measurement of the tumor size. Tumor volume was calculated using the formula Vtumor = (length × width2)/2. The experiment was aborted after 4 weeks because of the extensive tumor burden of the control group.
Transcriptional Activity of a 255-bp hTERT Core Promoter Is Increased by Combination with an E1A TATA Box Motive.
hTERT-luciferase reporter studies with series of unidirectionally deleted fragments derived from the hTERT promoter identified a 208-bp core promoter responsible for maximal activity (24). In this study, a 255-bp hTERT-promoter fragment was used. To investigate the effect of the E1A TATA box on hTERT-promoter activity, several promoter luciferase reporter plasmids were subcloned as indicated in Fig. 1,A. These hTERT-promoter subclones contained different numbers of E1A TATA box motives in orthograde or retrograde orientation. To assess transgene expression of these hTERT promoters in different tumor cells, telomerase-positive cancer cell lines CaCo2, HeLa, Huh7, Hep3B, and AGS were transfected, and luciferase measurements were performed. The transcriptional activity of hTERT255 in each cell line was considered as 100%. As shown in Fig. 1,B, in the investigated cancer cell lines, the hTERT-promoter constructs showed transcriptional activity comparable with the truncated SV40 promoter present in the control vector pGL2-Prom. In the majority of cell lines, the hTERT-promoter activity could be significantly enhanced by inserting the E1A TATA-box element downstream of the 255-bp hTERT core promoter (Fig. 1,B). The retrograde insertion of this TATA element did not alter the transcriptional activity of the hTERT promoter, whereas double or triple insertion led to a considerable reduction of transcriptional activation. Reflecting the chimeric promoter construct of hTERT255-T1, the E1A promoter upstream of the TATA box was replaced by the 255-bp hTERT core promoter for construction of the conditionally replicating adenovirus hTERT-Ad, as illustrated in Fig. 2.
Selective Replication, Oncolysis, and Spreading of hTERT-Ad in Telomerase-positive Cancer Cells.
Human primary hepatocytes are highly permissive for adenoviral replication, and liver toxicity is an important issue of adenoviral gene therapy in vivo (26, 27). Primary hepatocytes in vivo and in vitro exhibit neither hTERT expression nor telomerase activity. Therefore, primary adult hepatocytes were used to assess the selectivity of hTERT-Ad viral replication compared with replication of wild-type adenovirus. In addition, we investigated replication of ONYX-015, an E1B-55k deleted adenovirus mutant reported to replicate selectively in p53 mutated cells. As shown in Fig. 3,A, hTERT-Ad was not capable to replicate in primary human hepatocytes. In contrast, wild-type adenovirus and ONYX-015 both showed continuous replication after infection with low doses of viral particles. After an initial increase of hTERT-Ad particles in hepatocytes, viral titers dropped to background levels at later stages of the monitored time course. This limited replication competency might be explained by transient up-regulation of hTERT-promoter activity attributable to a transcriptional response after the adenoviral infection process. Monitoring EGFP fluorescence as an indirect marker for replication and viral spread, hTERT-Ad-dependent EGFP expression was only detectable during the initial phase of the trial in accordance with the viral titer results above. However, as hTERT-Ad-dependent fluorescence faded, Ad-GFP infection resulted in a strong and persistent fluorescence signal (Fig. 3 B). Interestingly, only few hTERT-Ad-infected hepatocytes seemed to express EGFP at levels resulting in visible fluorescence. This discrepancy may be attributable to the E1B promoter strength compared with the strong cytomegalovirus promoter of the Ad-GFP control vector.
Additionally, infected hepatocyte layers were screened for the occurrence of CPEs by light microscopy (Fig. 3,C). In contrast to common cell lines, collagen-fixed hepatocytes do not detach but round up after viral replication. Adenoviral replication of ONYX-015 and wild-type adenovirus resulted in obvious CPE in primary human hepatocytes. However, the morphology of hepatocytes after hTERT-Ad infection remained unchanged and was comparable with cells infected by the replication defective Ad-GFP. To test whether insertion of a TATA box motive affects the specificity of hTERT promoter-driven E1A expression, Western blot analysis of infected hepatocytes and telomerase-positive Huh7 hepatoma cells was performed. As shown in Fig. 3 D, E1A expression levels in hTERT-Ad-infected Huh7 cells were comparable with Ad-wt. In contrast, in hTERT-Ad-infected hepatocytes, E1A expression was completely absent, thus demonstrating the specificity of the hTERT promoter. In accordance with the results of the replication analysis, E1A protein was detectable in both ONYX-015-infected Huh7 cells and hepatocytes, despite the different p53 status of the target cells.
In contrast to telomerase-negative hepatocytes, infection of tumor cells with hTERT-Ad resulted in continuous viral replication as confirmed by viral titering (Fig. 4,A). Measurements revealed similar replication levels for hTERT-Ad and ONYX-015, almost reaching wild-type levels. Consistent with these findings, hTERT-Ad-mediated EGFP expression spread rapidly throughout tumor cell layers, whereas the number of Ad-GFP-infected cells remained constant (Fig. 4 B).
To test whether viral replication results in effective oncolysis, tumor cell layers were infected with different MOIs of hTERT-Ad, ONYX-015, and wild type. Target cell lysis was then visualized by crystal violet staining. Tumor cell lines with different p53 status were used to analyze the influence of p53 on viral replication. HeLa cells express wild-type p53, whereas CaCo2 and Huh7 cells have an altered p53 status. Independent of the cellular p53 status, all three replicative virus types showed effective lysis compared with the replication-deficient control. Although hTERT-Ad showed less rapid replication in tumor cells than adenovirus wild type (Fig. 4,A), the capability to lyse tumor cells was comparable (Fig. 4 C). In all tumor cell lines tested, hTERT-Ad was more effective than ONYX-015 in terms of cancer cell lysis, despite the similar replicative potential.
hTERT-Ad Inhibits Growth of s.c. Xenograft Tumors in Vivo.
To evaluate the efficacy of viral spreading and therapeutic potential of hTERT-Ad in vivo, Huh7-derived s.c. tumor xenografts on nude mice were infected with a single dose of hTERT-Ad, ONYX-015, and Ad-GFP, respectively. Monitoring EGFP expression in native tumor samples, as well as immunohistochemical staining of adenoviral hexon protein, revealed increasing spread throughout the tumor tissue by hTERT-Ad (Fig. 5,A). Although Huh7 cells represent a very fast growing tumor cell line, infection with hTERT-Ad led to significant retardation of tumor growth, resulting in complete growth arrest after 4 weeks of observation (Fig. 5 B). This inhibitory effect of hTERT-Ad appeared to be more distinct compared with ONYX-015.
An obstacle for successful gene therapy of cancer is the difficulty to achieve complete tumor cell transduction throughout the whole tumor mass by the vectors available to date. Tumor-specific, conditionally replicating viruses are promising tools to overcome low tumor cell transduction (23, 28, 29). Because of extensive knowledge about the regulatory mechanisms involved in adenoviral replication, adenoviruses are most frequently used to design tumor-specific replicative vectors. Four strategies have been exploited to achieve tumor-specific replication of recombinant adenoviruses: (a) deletion or mutation of the adenoviral genome to restrict viral replication to tumor cells functionally altered at crucial cell cycle or apoptosis checkpoints (30, 31); (b) tumor-specific recombination of adenoviruses (32); (c) selective transcriptional repression of cellular genes in normal cells essential for viral replication (33); and (d) transcriptional control of essential adenovirus early genes by tumor-specific promoters (34, 35, 36).
All of these concepts hold for promising therapeutic potential demonstrated by the achievement of tumor regression in animal models (34, 37). However, each of them inheres concept-specific limitations. ONYX-015, e.g., a putative p53-dependent, conditionally replicative adenovirus, is supposed to be attenuated in somatic cells because of a partial deletion in the E1B-55k gene. Although early clinical phase studies demonstrated therapeutic efficacy of ONYX-015 (38, 39), some studies revealed that ONYX-015 replication appears to depend not only on p53 status but also on individual cell line characteristics concerning cell cycle regulation, induction of apoptosis, and the applied MOI (40, 41, 42, 43, 44). In contrast to ONYX-015, adenoviruses with tumor-specific transcriptional control of essential adenoviral genes (i.e., E1A or E1A/E1B) seem to replicate more selectively. In this context, various tumor-specific promoters have been used, such as AFP-PSE, OC, or Muc1 promoter (34, 35, 36), but the activity of these promoters is limited to certain tumor histologies. Consequently, a more promising strategy appears to be the linkage of viral replication to defined pathways that are altered at early stages of tumor development and maintained during carcinogenesis.
Meeting the requirements for tumor-restricted adenoviral replication and extensive applicability, the hTERT promoter represents an attractive tool for transcriptional control of replication regulatory elements. Several studies demonstrated telomerase-specific adenoviral gene expression by using 204- or 378-bp fragments of the hTERT promoter in E1-deleted vectors (20, 21, 22). In this study, a conditionally replicative adenovirus (hTERT-Ad) was constructed, expressing the E1A region under control of a 255-bp hTERT promoter upstream of the E1A TATA motive. In the majority of cell lines tested, this arrangement enhanced transcriptional activity of the hTERT promoter without altering its specificity. Consistently, we observed neither viral replication nor E1A expression in hTERT-Ad-infected, telomerase-negative primary human hepatocytes. Thus, the lack of hTERT-promoter activity observed in human hepatocytes was in accordance with recent findings that hTERT-controlled transgene expression could not be detected in livers of mice after adenoviral infection (22). In contrast, infection of telomerase-positive tumor cells by hTERT-Ad led to rapid viral replication and tumor cell lysis, and infection spread throughout the cell layer. The capability of hTERT-Ad to induce CPEs in tumor cells was comparable with that of adenovirus wild type and significantly higher compared with ONYX-015, regardless of the p53 status of the tumor cells. Additionally, ONYX-015 seemed to replicate in p53-positive hepatocytes to a certain extent. Therefore, hTERT-Ad appears to possess higher potential in terms of specificity and efficacy than ONYX-015.
In vivo studies revealed effective viral spreading and significant inhibition of tumor growth after a single application of hTERT-Ad to tumor xenografts, confirming the potential therapeutic value of hTERT-Ad as a conditionally replicative adenoviral vector. This effect, in agreement with the in vitro findings, was more distinct compared with ONYX-015.
In humans, telomerase activity is found in only few cell types like stem and germ-line cells. Thus, they might also be permissive for hTERT-Ad replication. However, these cells are not of epithelial origin and therefore extremely difficult to infect with adenoviral vectors after systemic vector application.
The transgene arrangement of hTERT-Ad broadens the perspectives for additional applications. In this study, an EGFP reporter gene was linked to the E1B transcription via an internal ribosome entry site to monitor viral replication and spreading of adenoviral infection. In additional studies, EGFP in hTERT-Ad may be replaced by genes coding for tumoricidal cytokines or suicide genes to enhance the therapeutic efficacy of hTERT-Ad.
Our results demonstrate that transcriptional control of E1A by the hTERT promoter is sufficient for effective telomerase-dependent adenovirus replication. Because >90% of human tumors exhibit telomerase activity and hTERT expression, hTERT-Ad promises to be applicable for many human epithelial tumors.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Supported by Grant 98.046.2 of the Wilhelm Sander Stiftung and Grant KU 12/13 3/1 from Deutsche Forschungsgemeinschaft.
The abbreviations used are: hTERT, human telomerase reverse transcriptase; IRES, internal ribosomal entry site; MOI, multiplicity of infection; CPE, cytopathic effect; TCID, tissue culture infectious dose 50; EGFP, enhanced green fluorescent protein; MCS, multiple cloning site; HP, horseradish peroxidase.