¹Mary C Crowley Cancer Research Program, Baylor Research Institute, Baylor University Medical Center, Dallas, TX, USA

²US Oncology, Inc. Dallas, TX, USA

³Berlex Biosciences, Richmond, CA, USA

⁴Cancer Immunology Research Laboratory, Baylor-Sammons Cancer Center, Baylor University Medical Center, Dallas, TX, USA

Correspondence to: A W Tong, Cancer Immunology Research Laboratory, Baylor-Sammons Cancer Center, Baylor University Medical Center, 3500 Gaston Avenue, Dallas, TX 75246, USA

Approximately 15-30% of human non-small cell lung cancers (NSCLC) carry K-ras mutations, among which point mutations at codon 12 are the most common. This study characterizes the anti-tumor effect of an anti-K-ras ribozyme adenoviral vector (KRbz-ADV; replication-deficient, E1-deleted Ad5 backbone) against NSCLC lines that express the relevant mutation (K-ras codon 12 GGT GTT; H441 and H1725). KRbz-ADV significantly inhibited tumor cell growth (38-94% reduction by ³H-thymidine uptake) in a time- and dose-dependent manner, but produced minimal growth inhibition on normal epithelial cells, or NSCLC H1650 cells that lack the relevant mutation. The in vivo anti-tumorigenic effect of KRbz-ADV treatment was characterized with cell line xenografts in nu/nu mice. Pre-treatment with KRbz-ADV (10 or 20 p.f.u. per cell) completely abrogated subcutaneous engraftment of H441 (n = 13) or H1725 cells (n = 8), as compared with a 100% tumor take and progressive tumor growth in animals that received untreated tumor cells, or control vector (luciferase-adenovirus/Luc-ADV)-treated tumor cells. Pre-treatment with a mutant anti-K-ras ribozyme adenoviral vector (mutKRbz-ADV), which has the same specificity as KRbz but lacks ribozyme catalytic activity, did not produce an anti-tumorigenic effect. The in vivo effect of KRbz-ADV treatment was further examined by initiating injections (2 ´ 10⁹ p.f.u.) at 7 days after tumor induction. Pre-existing tumor growth was reduced by 39% by a single intratumoral injection. Repeat injections (three or five KRbz-ADV-intratumoral injections at 2 ´ 10⁹ p.f.u. every other day) resulted in complete tumor regression in five of seven mice. In contrast, single or multiple injections of control vector Luc-ADV did not significantly alter tumor xenograft outcome. Ribozyme expression was confirmed in H441 cells that demonstrated reduced growth after KRbz-ADV treatment. Reduced growth corresponded to significantly lowered levels of K-ras mRNA, as defined by RT-PCR (51% of untreated level, n = 3) and RNase protection assay (56% of untreated level, n = 4) analyses. Further, 37.5% of KRbz-ADV-treated cells underwent apoptosis, as compared with 11.7%, and 19.0% in untreated and Luc-ADV-treated cultures, respectively. A significantly higher proportion of KRbz-ADV-treated H441 cells (58.2%) underwent apoptosis when maintained under anchor-independent conditions that simulate in vivo tumorigenesis ('anoikis'). This is the first report that demonstrates that KRbz-ADV can effectively inhibit in vivo tumorigenesis, and produces regression of pre-existing human lung tumor xenografts having the relevant K-ras mutation. Gene Therapy (2000) 7, 2041-2050.

ribozyme; K-ras; non-small cell lung cancer; adenoviral vector, nude mouse xenograft

Introduction

The protooncogene ras encodes a guanine nucleotide-binding GTPase (p21^ras) that relays extracellular growth and/or differentiation signals to the nucleus, thereby regulating mammalian cell growth. Currently, four ras isoforms (H-ras, N-ras, K-ras4A, K-ras4B) and two ras-related proteins (R-ras and TC21) have been identified, all with similar functions. Each Ras protein (p21^ras) alternates between a GTP-bound active state and a GDP-bound inactive state. Activating ras point mutations, invariably found in the GTP binding regions, produce an activated, GTP-locked p21^ras that constitutively activates nuclear oncogene products. The Ras signaling process involves the activation of Raf/MAP kinase, as well as multiple effectors including phosphatidylinositiol-3 kinase and the Rho family proteins (Rac/Rho).^1,2 Mutant Ras may contribute to malignant transformation via its multiple roles in cell differentiation and cell cycle regulation, in apoptosis modulation, in reactive oxygen species superoxide generation, and in tumor angiogenesis.^3,4,5,6 Recently, Hahn and coworkers⁷ demonstrated that mutation in the H-ras gene is pivotal for tumorigenesis of normal human epithelial and mesenchymal cells. Ras mutation also appears essential for maintenance of the malignantly transformed state.⁸

Ras oncogene mutation is one of the most common oncogenetic defects in human cancers. Approximately 30% of lung adenocarcinomas, and 15-20% of all non-small cell lung cancers (NSCLC) carry ras mutations, which portend a poor prognosis for both early and late stage disease.⁹ The mutant ras gene is an attractive candidate for lung cancer gene therapy in view of its relatively high mutation rate and restricted localization of point mutations. Indeed, blockage of Ras function at the protein and RNA level produced an anti-tumor effect.^10,11,12 The down-regulation of mutant p21^ras correspondingly decreased lung tumor growth in vitro and in vivo.¹⁰ A K-ras antisense retroviral construct against the homozygous K-ras codon 61 mutated sequence similarly inhibited the growth of the human lung cancer line H460a in vitro and in vivo.^11,12

Ribozymes (catalytic RNAs, RNA enzymes) are RNAs with site-specific RNA cleavage or ligation activities.^13,14 Ribozymes directed at oncogenes (c-fos, BCR-ABL) or the human immunodeficiency virus type 1 were more effective than their antisense counterpart in terms of target mRNA reduction and biologic efficacy.¹⁵ In human tumor models, treatment with a hammerhead ribozyme against the H-ras codon 12 mutation (GGC GUC) inhibited tumor cell growth in vitro, and partially reverted the malignant phenotype of human bladder carcinoma and melanoma and murine NIH3T3 cells expressing the relevant H-ras mutation.^16,17,18 Similarly, anti-K-ras ribozyme targeting the codon 12 mutation (GGU GUU) inhibited human pancreatic cancer cell growth in vitro, and specifically cleaved mutant but not wild-type K-ras mRNA.^19,20 By comparison, the applicability of K-ras-specific ribozymes against human non-small cell lung cancers has not been characterized. We recently examined the NSCLC growth-modulating effect of an anti-K-ras codon 12 (GUU) hammerhead ribozyme, since K-ras mutations account for 90% of all NSCLC ras mutations, and 85% of K-ras mutations involve codon 12. We constructed a KRbz adenoviral vector (KRbz-ADV) by homologous recombination of the pACCMVpLpA-KRbz shuttle vector with the pJM17 adenoviral vector. Experiments were carried out using the human lung adenocarcinoma cell line H441 and H1725 that express a heterozygous K-ras codon 12 mutation of GGT GTT. This report characterizes the tumor growth inhibitory properties of the KRbz-ADV in vitro and in vivo.

Results

Characterization of KRbz-ADV anti-tumor effect in vitro

Our KRbz-ADV construct contains a hammerhead ribozyme sequence insert that targets K-ras codon 12 with a GTT mutation.^19,21 Its antitumor efficacy was evaluated with the NSCLC cell lines H441 and H1725 cells. Both lines have a heterozygous K-ras codon 12 GGT/GTT phenotype as defined by DNA sequencing.²² KRbz-ADV at 25 p.f.u. per cell significantly inhibited H441 cell growth in vitro in a time-dependent manner, as defined by ³H-thymidine uptake analysis. There was no significant inhibition at 24 h, whereas H441 growth was reduced by 39 ± 7% (mean ± s.d.; P < 0.05) at 72 h (Figure 1), and corresponded to a 31% reduction in viable cell number as defined by trypan blue exclusion analysis. Growth inhibition at 96 h was more pronounced than the observed effect at 72 h (data not shown). KRbz-ADV treatment similarly reduced ³H-thymidine uptake of H1725 cells by 71 ± 3.1% at 72 h after infection (Figure 1). Growth inhibition was increased following treatment with a higher dose (100 p.f.u. per cell; reduction of 63 ± 5% and 94 ± 3% for H441, H1725 cells, respectively). To further correlate antitumor effect with K-ras ribozyme function, parallel studies were carried out with a mutant anti-K-ras ribozyme adenoviral vector (mutKRbz-ADV) that has the same specificity as KRbz-ADV, but lacks ribozyme catalytic activity. MutKRbz-ADV produced only moderate growth inhibition (15 ± 14% and 16 ± 5% for H441 and H1725 cells, respectively) that was significantly lower than that by KRbz-ADV (Figure 1). These findings indicate that KRbz-ADV inhibited the in vitro growth of NSCLC cells with the relevant K-ras mutation, and effective inhibition was dependent on the catalytic cleavage activities of the K-ras ribozyme.

For H441 cells, there was no significant growth reduction in H441 cultures following treatment with comparable doses of a control adenoviral vector containing the luciferase gene insert (Luc-ADV; 4 ± 14%; P < 0.05; Figure 1). At an MOI of 5 p.f.u. per cell, only KRbz-ADV significantly inhibited the growth of H1725 cells (24.6 ± 3.9%; versus 2.0 ± 2.4% by Luc-ADV, and -2.0 ± 10.5% by mutKRbz-ADV). At an MOI of 25 p.f.u. per cell, KRbz-ADV inhibited H1725 growth by 71 ± 3.1% (Figure 1), although this growth inhibitory effect may be partially attributed to a higher sensitivity of this cell line to the ADV infection process: Luc-ADV also inhibited H1725 growth, albeit at a significantly lower level (39 ± 2.8%, Figure 1).

The selective growth inhibitory effect of KRbz-ADV was further confirmed with the NSCLC cell line H1650 having a wild-type K-ras phenotype, as well as with a primary culture of normal human bronchial epithelium (NHBE) cells. KRbz-ADV did not significantly inhibit the growth inhibition of H1650 cells (12.5 ± 16%), according to ³H-thymidine uptake analysis. Cell viability analysis was carried out on NHBE primary cells (rather than the ³H-thymidine uptake) because of limited proliferative activity of the primary culture. KRbz treatment was minimally toxic (>90% viable) to NHBE cells at MOIs of 25-100 p.f.u. per cell. Thus KRbz-ADV appears to produce minimal toxicity on cells with a wild-type K-ras sequence, or normal bronchial epithelial cells that are expected to be most susceptible to ADV vector-related cytotoxic effects during intratumor injections.

Adenoviral transduction efficiency

To further characterize the varied growth inhibitory effect of KRbz-ADV on H1725 and H441 cells, we examined the ADV transduction efficiency on these lines with an adenoviral vector containing the E. coli -galactosidase transgene (LacZ-ADV). Between 5 and 90% of H441 cells and 35-100% of H1725 cells were positive with X-gal staining following LacZ-ADV infection (MOI of 1-100 p.f.u. per cell; Table 1). H1725 cells were consistently more susceptible to ADV transduction as compared with H441 cells in the range of MOIs tested. In vivo analysis indicated that H441 tumor xenografts displayed a focal pattern of X-gal staining following LacZ-ADV infection with approximately 50% of X-gal-positive tumor cells adjacent to the virus release site. Our findings indicate that the adenoviral vector can efficiently deliver the KRbz transgene into NSCLC cells in vitro and in vivo.

Effect of KRbz-ADV on K-ras expression

We examined K-ras mRNA levels to identify the contribution of K-ras cleavage in the KRbz-ADV-mediated anti-tumor effect. Two sets of RT-PCR reactions were used to establish the level of uncleaved K-ras mRNA (product I) and total of (uncleaved + cleaved) K-ras mRNA (product II). Product I spans the second nucleotide of exon 1 to the 62nd nucleotide of exon 2, lapses the potential KRbz cleavage site at codon 12 (GUU; Figure 2a), and corresponds to full length K-ras mRNA. Its level is expected to be decreased upon cleavage of mutant K-ras mRNA. Product II spans the 56th nucleotide of exon 1 to the 62nd nucleotide of exon 2, and reflects total (KRbz-cleaved and uncleaved) K-ras mRNA, since the 3' fragment of KRbz-cleaved K-ras mRNA still serves as a viable template for product II amplification. The molecular identity of product I and II was verified by automated DNA sequencing. There were approximately equal proportions of wild-type and mutant K-ras mRNA in untreated H441 cells, based on peak height analysis of product I DNA sequences.

KRbz-ADV treatment manifested as a reduction of full length K-ras mRNA (product I) (49 ± 3.2%, n = 3; P < 0.05; Figure 2), whereas Luc-ADV treatment did not significantly alter the level of intact K-ras mRNA. Total K-ras mRNA level was reduced moderately following KRbz-ADV treatment (19 ± 4.6% decrease in product II). The reduction in functional full length K-ras mRNA is consistent with the KRbz ribozyme cleavage effect at the post-transcriptional level.²³ Ribonuclease protection assay (RPA) analysis similarly confirmed that KRbz-ADV reduced K-ras mRNA expression (44 ± 13.3%, n = 4; P < 0.05) as compared with untreated cultures (Figure 3). By contrast, mutKRbz-ADV treatment did not significantly reduce K-ras mRNA levels (0 ± 8.8%, n = 4; P > 0.05). These findings support the thesis that growth inhibition by KRbz-ADV correspondingly lowers the level of K-ras mRNA, an outcome that is dependent on the catalytic cleavage activity of the K-ras ribozyme.

Ribozyme expression in cells that exhibited the relevant growth inhibitory effect was verified by RPA (Figure 3) and RT-PCR (data not shown) analyses. The relative molar ratio of ribozyme versus K-ras mRNA was determined to be 1.55:1, based on gel band densitometric analysis of ribonuclease-protected products in KRbz-ADV infected H441 cells (100 p.f.u. per cell). These findings are consistent with observations that ribozyme:targeted RNA ratios of >1 are needed for effective substrate cleavage.^24,25

Antitumorigenic effect of KRbz-ADV on NSCLC heterotransplants

NSCLC lines H1725 and H441 were successfully engrafted in Swiss nude mice by s.c. injection of 1 ´ 10⁷ cells (H441, seven of seven tested, Figure 4; H1725, 11 of 11 tested, data not shown). KRbz-ADV treatment (10 p.f.u. per cell) completely abrogated H441 tumor xenograft growth in nu/nu mice (n = 13; Figure 4). The lack of tumor growth at the site of injection was confirmed by histological evaluations at the end of observation period. By contrast, pre-treatment with mutKRbz-ADV did not affect H441 tumor take or growth (n = 4), as compared with untreated mice (Figure 4).

Similarly, mice inoculated with same number of KRbz-ADV-pretreated H1725 cells did not produce tumor xenografts for up to 90 days (n = 8), whereas untreated H1725 tumors grew exponentially within 2-3 weeks with a 100% tumor take. These findings indicate that KRbz-ADV inhibited the in vivo tumorigenesis of NSCLC H441 and H1725 cell xenografts, and this effect was dependent on the catalytic cleavage activity of KRbz-ADV.

Parallel Luc-ADV treatments were carried out to characterize specificity of the KRbz-ADV growth inhibitory effect. Luc-ADV at 10 p.f.u. per cell did not inhibit H441 heterotransplants, which grew at the same rate as untreated tumor xenografts (Figure 4). Treatment with a higher dose of Luc-ADV (20 or 50 p.f.u. per cell) retarded growth and delayed tumor emergence (palpable at day 35, as compared with 14-21 days for untreated inoculants; Figure 4), although its growth inhibitory effect was significantly lower than the corresponding dose of KRbz-ADV. No overt metastatic growth was detected in the major organs (lung, liver, spleen, kidney, heart and peritoneal membranes) of mice that received untreated, or KRbz-ADV pre-treated H1725 or H441 cells at day 60-90 after inoculation. Gross toxicity (loss of activities, body weight) was not evident for either KRbz-ADV or Luc-ADV-treated mice.

To characterize better the clinical applicability of KRbz-ADV treatment on pre-existing tumors, intratumoral injections were initiated at day 7 after tumor induction. For H441 xenografts, a single intratumoral injection (2 ´ 10⁹ p.f.u.) significantly reduced the primary tumor size by 39% (6.04 ± 1.09 mm at day 21, n = 7; P = 0.0061) as compared with untreated mice (9.86 ± 2.84 mm, n = 7), but did not completely inhibit tumor xenograft growth as observed in pretreatment studies (Figure 4). One reason may be that the viral transduction process in vivo is less efficient than that in vitro. By comparison, multiple KRbz-ADV intratumoral injections (2 ´ 10⁹ p.f.u. every other day) can completely inhibit pre-existing tumor growth (Table 2). Only two of seven animals that received three or five KRbz-ADV injections continued to have tumor growth. This represents a significantly decreased incidence (P < 0.05) as compared with untreated or mock treated groups (Table 2). Further, tumor size of the two remaining tumor-bearing mice was reduced markedly (tumor diameter of 9.15 mm in one mouse after three KRbz-ADV injections; 7.60 mm in one mouse after five KRbz-ADV injections, versus 13.19 ± 1.0 mm in six of six untreated mice at day 42 after tumor inoculation). One animal had complete tumor regression after five repeat Luc-ADV injections, which may be attributable to the non-specific growth inhibitory effect of ADV infection.²⁶ However, single or multiple control Luc-ADV injections did not significantly alter tumor xenograft outcome (P > 0.05). Our findings demonstrate that multiple treatments with KRbz-ADV are effective in inducing NSCLC tumor regression in vivo.

Induction of apoptosis on NSCLC cells by KRbz-ADV

To explore cellular mechanisms that contribute to the KRbz-ADV growth inhibitory effect, we examined the level of apoptosis in KRbz-ADV-treated H441 cells by vital dye (7-AAD) uptake.^27,28 The percentage apoptotic cells in KRbz-ADV-treated H441 cultures was significantly elevated (37.5%, P < 0.01) as compared with untreated (11.7%) and control-vector treated (19.0%) cultures (Table 3). KRbz-induced apoptosis was further increased (58.2%) under anchor-independent growth conditions, and was significantly higher than that in KRbz-treated adherent cultures (P < 0.05, Table 3). By comparison, apoptotic activities did not differ significantly among adherent and non-adherent cultures for untreated (P > 0.05) or Luc-ADV-treated cells (P > 0.05; Table 3). These observations were confirmed by flow cytometric analysis of Annexin V (ANN)- and/or propidium iodide (PI)-stained H441 cells. KRbz-ADV treatment (50 p.f.u. per cell) induced apoptosis in 45% of cells (ANN⁺ or ANN⁺PI⁺) under anchor-independent culture conditions and 33% in adherent cultures. Our findings demonstrate that KRbz-ADV treatment significantly up-regulated apoptotic activity, an effect that was augmented under anchor-independent culture conditions that simulate in vivo anti-tumorigenesis.⁵

Discussion

In vitro and in vivo antitumor outcome through inhibition of mutant K-ras oncogene activity has been demonstrated with Ras neutralizing antibodies and anti-K-ras antisense-plasmids or retroviral-constructs.^11,12,29,30 Another alternative for targeting the ras mutant gene is by the anti-ras ribozyme construct, which provides theoretical stoichiometric advantage over its antisense counterpart and may be delivered as a viral construct for improved pharmacokinetics.¹⁴ Rbzs targeting mutant H-ras and K-ras sequence have been shown to inhibit the growth of human bladder, melanoma and pancreatic carcinomas.^16,17,19 However, there is limited information regarding the efficacy of KRbz on NSCLC tumors expressing K-ras codon 12 mutations, despite prevailing observations that <85% of NSCLC ras mutations reside at K-ras codon 12.^9,31

In this study, we have examined the in vitro and in vivo growth inhibitory effects on NSCLCs by KRbz, and its antitumor mechanism of action. H441 and H1725 cells, both having a single point mutation (GGT GTT) in one of the K-ras codon 12 alleles, are expected to reflect the heterozygous ras status in most NSCLC patient tumors. These cell lines were used to examine the antitumor effect of a hammerhead ribozyme designed to cleave the GUU sequence in codon 12 of the activated K-ras mRNA transcript, one of most prevalent ras mutations in human lung cancer.^9,31 A KRbz-ADV viral construct was generated, since ADV vectors are capable of infecting most dividing and non-dividing cells with a high efficiency of gene transfer, and are obtainable in high titers.³² In light of its natural tropism for respiratory tissues, the adenovirus vector is particularly appropriate for anti-tumor gene delivery to lung cancer. The efficacy of this KRbz-ADV construct was confirmed by its capacity to inhibit H441 and H1725 cell growth by up to 94% in vitro. Therapeutic efficacy of KRbz-ADV was supported by our in vivo findings that KRbz-ADV abrogated tumorigenesis, reduced pre-existing tumor xenograft growth, and induced complete tumor regression in up to 75% of mice after repeat intratumoral KRbz-ADV injections. These findings are consistent with an essential role of mutant Ras in tumorigenesis and lung tumor maintenance.⁸ Certain anti-ras approaches such as farnesyltransferase inhibitors are more effective against tumors with H-ras mutations.^33,34 Our study indicates that the appropriate Rbz appears to be equally effective against tumors with H-ras or K-ras mutations. This consideration is particularly significant, in view of the prevalence of K-ras mutant human solid cancers.³⁵ Our report represents the first documentation of in vivo antitumor efficacy by a K-ras ribozyme against human lung tumors with the relevant mutation.

Specificity of KRbz-ADV mediated tumor-growth inhibition is validated by our current findings of its selective toxicity on H441 cells. H1725 cells appeared to be sensitive to ADV infection. Nevertheless, KRbz-ADV produced significantly higher growth inhibition, as compared with the control Luc-ADV vector. We previously showed that this KRbz, when delivered as a plasmid, inhibited the in vitro growth of H1725 cells by up to 82% without any cytotoxic effect on NSCLC H460 cells lacking the relevant K-ras mutation.^22,36 The same KRbz-plasmid was found to reduce selectively pancreatic cancer K-ras mRNA level without affecting control cells having an irrelevant mutation.²⁰

Conversely, our KRbz-ADV produced a minimal growth inhibitory effect on NSCLC H1650 cells and normal human bronchial epithelial cells with a wild-type K-ras phenotype. In our xenograft model, intratumoral KRbz-ADV injections produced no gross toxicity (loss of activities, body weight) or overt toxicity in major organs of mice. These findings indicate that such an adenoviral construct at <1 ´ 10¹⁰ p.f.u. may be safely administered in vivo^37,38,39 Similarly, the Ad5CMV-p53 adenovirus, which shares a similar adenoviral construct as KRbz-ADV, produced only mild to moderate acute toxicity when given intratracheally at 10¹⁰ p.f.u.³⁹

The in vitro KRbz-ADV anti-tumor effect correlated with an approximately 50% decrease in full length K-ras mRNA transcripts, suggesting that the down-regulation of ras expression may be involved in tumor growth inhibition. Previously, we have reported that KRbz-ADV collaterally decreased mutant p21 Ras protein levels.³⁶ For this study, the significant down-regulation of full length K-ras (product I) is consistent with the expected cleavage activity of KRbz.²³ The approximate 50% reduction in product I was confirmed by RT-PCR and subsequent RPA analyses. A moderate decrease in the total (cleaved and intact) K-ras mRNA (product II) may be attributable to subsequent degradation of KRbz-cleaved mRNA by RNase L.⁴⁰ For H441 cells with a heterozygous ras phenotype, we found that both the wild-type and mutant K-ras genes were co-expressed at a comparable level. An incomplete reduction in ras mRNA expression is consistent with the premise that KRbz-ADV does not act on wild-type K-ras. Our findings support prior observations that tumor cells appear to be more dependent on mutant ras function,⁴¹ such that a less than complete reduction of activated ras expression could nevertheless significantly affect tumor cell growth.

The experimental characterization of K-ras mRNA levels required RNA extraction from KRbz-ADV-infected cells, RNA reverse transcription and RNA hybridization processes during RT-PCR and RPA analyses. Although theoretically possible, it is unlikely that the observed down-regulation of K-ras mRNA represented the outcome of cell-free cleavage by KRbz during these procedures. RNA isolation was performed primarily at 4°C in less than 2 h. In vitro cell-free ribozyme-dependent substrate cleavage requires stringent conditions, including high pH (7.5), optimized temperature (37°C), extended incubation (usually 18 h) and the presence of magnesium (1 mM).^24,25,42 These requirements were inconsistent with our conditions for RT-PCR. RPA did not require any divalent metal ions during RNA hybridization reaction or a RNA transcription step but nevertheless demonstrated that KRbz-ADV reduced K-ras mRNA expression by 50% (Figure 3). MutKRbz-ADV did not produce significantly altered K-ras mRNA under the same RPA conditions. Thus it is unlikely that the decreased level of K-ras mRNA represented experimental artifacts.

Activated Ras has been shown to play an important role in tumor cell proliferation, differentiation, in vivo cell attachment/survival and angiogensis. Our thesis of a KRbz-dependent growth inhibitory phenomenon is supported by the findings that only KRbz-ADV was effective in inhibiting xenograft growth, whereas comparable doses of Luc-ADV did not. It is not fully understood as to why KRbz-mediated growth inhibition was more pronounced in vivo. Recent studies suggest that antisense-suppression of activated ras increases apoptosis.⁴³ Mutant ras is particularly effective in nullifying 'anoikis', an apoptotic event that comes into play when cells become detached from an underlying extracellular matrix.^3,5 Similarly, we found that KRbz-ADV treatment was 50% more effective in inducing apoptosis (from 37.5% to 58.2%) under anchor-independent culture conditions that simulate in vivo tumorigenesis. This may explain, at least in part, the more dramatic growth inhibitory effect of KRbz-ADV in vivo. Future studies are needed to evaluate other potential contributing mechanisms, such as anti-angiogenic and cell cycle regulatory events that may also manifest as 'bystander effect' on neighboring untransfected tumor cells.^6,44,45

Recent clinical trials have utilized the selective adenoviral cytopathic effect by targeting p53 deficient cancer cells with E1B defective adenoviruses, since cells with normal p53 are more resistant (100-fold) to viral lysis.⁴⁶ NSCLC H1725 cells have a wt p53 gene, whereas H441 cells have a point mutation in exon 5 of p53 gene.⁴⁷ Nevertheless, we observed significantly higher growth inhibition in wt p53 H1725 cells by KRbz-ADV. Thus the selective adenoviral oncolytic effect on p53-deficient cells appears not to be the dominant factor for the antitumor outcome of KRbz-ADV.

Our studies demonstrated that KRbz-ADV is effective for in vitro and in vivo growth inhibition of NSCLCs. The potential clinical applicability of KRbz-ADV treatment is supported by our findings that a single intratumoral injection was effective in reducing tumor growth by >35%, and that multiple intratumoral injections of KRbz-ADV produced complete xenograft regression in 65-75% of animals. Further analyses are planned to determine the anti-tumor effect of systemic KRbz-ADV treatment. Despite the limited proportion of lung cancers eligible for treatment by each Rbz, the ribozyme adenoviral vector approach is nevertheless potentially applicable because of the considerable incidence of human lung cancers. Upon demonstration of clinical efficacy, it is not inconceivable that multiple individual ribozymes can be generated, each targeting a specific K-ras mutation. In parallel with current clinical trials on p53-ADV or E1B defective ADV, the use of KRbz-ADV represents another gene therapy approach potentially applicable to the treatment of lung cancers.

Materials and methods

Cells

NSCLC cell lines NCI H441 (adenocarcinoma) and H1725 (large cell carcinoma) were kindly provided by Dr Herbert Oie (National Cancer Institute, Bethesda, MD, USA). The NSCLC cell line H1650 (adenocarcinoma with wt K-ras) was obtained from the American Type Culture Collection, Rockville, MD, USA.⁴⁸ These cell lines were cultured in T75 flasks in RPMI 1640 (Gibco/BRL) and 10% fetal bovine serum (FBS; Atlanta Biological, Atlanta, GA, USA). The low-passage transformed human embryonic kidney cell line 293 was obtained from American Type Culture Collection, and cultured in Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated horse serum and penicillin-streptomycin (Gibco, Grand Island, NY, USA). Primary cultures of normal human bronchial epithelium (NHBE) cells were purchased from BioWhittaker/Clonetics (Walkersville, MD, USA) and maintained according to the manufacturer's recommendations in optimized growth medium (SABM and SAGM, BioWhittaker/Clonetics).

Generation of the anti-K-ras ribozyme adenoviral vector (KRbz-ADV)

The K-ras ribozyme was prepared as described previously.²¹ An E1-deleted, replication defective adenoviral vector containing anti-K-ras ribozyme insert (KRbz-ADV) was constructed by homologous recombination of the adenoviral shuttle vector pACCMVpLpA-KRbz and the packaging plasmid pJM17. pACCMVpLpA plasmids with the anti-K-ras ribozyme insert were co-transfected with pJM17 into the E1A trans-complementing cell line 293, using the commercial cationic liposome vector DOTAP (Boehringer-Mannheim, Mannheim, Germany).²¹ A mutant anti-K-ras ribozyme-adenoviral vector (mutKRbz-ADV), which contains a single residue replacement (G to A) in the catalytic core with respect to KRbz-ADV construct, shares the same specificity as KRbz-ADV but lacks ribozyme function.⁴⁹ Adenoviral vector that contained the Escherichia coli -galactosidase gene (LacZ-ADV) was also constructed in a similar manner.²⁶ The primary recombinant adenovirus stock was clonally selected three times by standard 293 cell plaque-forming assays.^50,51 Recombinant adenoviruses contain the correct ribozyme insert and are free of detectable wild-type adenovirus, as defined by appropriate PCR assays.^36,52 The recombinant adenoviruses were prepared by two rounds of CsCl gradient centrifugation and titered in 293 cells.^50,51 The corresponding ADV vector with a luciferase gene insert (Luc-ADV) was obtained from Dr Stephen A Johnston, University of Texas Southwestern Medical Center, Dallas, TX, USA. This control vector has a similar structural configuration as the KRbz-ADV vector except for the luciferase gene insert at multiple cloning sites.⁵³

KRbz-ADV treatment

³H-thymidine uptake of KRbz-ADV-infected NSCLC cells was quantified in 96-well cultures. Subconfluent NSCLC cells (0.5 or 1 ´ 10⁴ per well) were washed once with phosphate buffered saline (PBS, 200 l), infected with recombinant adenoviruses (suspended in 50 l of RPMI1640 and 2% FBS, 37°C, 90 min), followed by the addition of 150 l of fresh RPMI1640 with 10% FBS. ³H-thymidine (DuPont, Wilmington, DE, USA; 0.15 Ci per well) was added to the culture at 24 to 96 h after infection. The culture was harvested after an additional 16 h. Radiolabel uptake was determined by liquid scintillation (Betaplate 1205; Wallac Inc, Gaithersburg, MD, USA). Percentage of inhibition was determined by the formula: (1 - (mean c.p.m._treatment - mean c.p.m._background)/(mean c.p.m._control - mean c.p.m._background)) ´ 100; where c.p.m._control represents the endogenous ³H-thymidine uptake of untreated culture, and c.p.m._background, non-specific uptake by culture wells with no cells. The effect of ADV vector infection on cell viability and total cell numbers was carried out in 24-well plates with subconfluent H441 cultures and NHBE cultures by trypan blue exclusion and Coulter Counter (Coulter, Hialeah, FL, USA) enumeration, respectively.

Gene transduction by adenoviral vectors

In vitro transduction efficiency of adenoviral vector on NSCLC lines and NHBE cells was determined with the LacZ-ADV vector under the same experimental conditions as KRbz-ADV treatment. For characterization of transduction efficiency in vivo, H441 s.c. xenograft was injected intratumorally with LacZ-ADV (1 ´ 10¹⁰ p.f.u.). The tumor was excised at 72 h after injection. Frozen 8-m sections were analyzed by the X-gal staining assay^26,54 after glutaraldehyde fixation (0.24% in PBS, 10 min, 4°C). The biopsy sections were treated with 0.1% X-gal solution (5 mM K₃Fe(CN)₆, 5 mM K₄Fe(CN)₆^.3H₂O, 2 mM MgCl₂, PBS; 37°C, 14-16 h). The 0.1% X-gal solution was made fresh and filtered through a 0.2 m syringe-top filter before use. The overall percentage of X-gal stained cells was estimated, based on light microscopy enumeration of 200-400 total cells.

Reverse transcription (RT)-PCR assay and ribonuclease protection assay (RPA)

Subconfluent H441 cells (4.5 ´ 10⁶) were infected with the appropriate ADV vector at 100 p.f.u. per cell, then cultured for 72 h. Total RNA was isolated with TRIZOL reagents according to manufacturer's protocol (Gibco). RT-PCR assays were carried out using GeneAmp PCR System 2400 thermal cycler and GeneAmp RNA PCR kit (Perkin Elmer, Norwalk, CT, USA) as described previously with modifications.⁵⁵ To quantify the level of K-ras mRNA, co-amplification of K-ras mRNA and an internal control message (-actin) was carried out in the same reaction mixture for both RT and PCR steps. Three g of total RNA template were co-annealed with the K-rasP2 primer and/or -actinP2 primer (50 ng each, 45 min 55°C). Reverse transcription (20 l) was carried out with 6 l of the annealing mixture (1 hr, 42°C). The cDNA products were amplified by standard PCR (94°C, 45 s; 55°C, 30 s; 72°C, 45 s; 32 cycles). The use of K-rasP1 primer and K-rasP2 primer generates a cDNA product only from the uncleaved K-ras mRNA (RT-PCR product I, 172 bp). The K-rasP3 primer and K-rasP2 primer were used to generate a cDNA product corresponding to the KRbz-cleaved sequence, thereby identifying total (both uncleaved and cleaved) K-ras mRNA (RT-PCR product II, 118 bp). The -actinP1 and P2 primers were added into each PCR reaction mixture after seven PCR cycles in order to balance the final yields of -actin (249 bp) and K-ras product. The K-ras RT-PCR products were confirmed by DNA sequencing (Big Dye Terminator method, ABI 310 Automated DNA Sequencer; Perkin Elmer). Changes in the levels of K-ras intact and cleaved mRNA (product I and product II) were determined by densitometric analysis following normalization to the corresponding -actin value, which serves as the internal standard in semi-quantitative RT-PCR analysis (AlphaImager 2000 D, Alpha Innotech, San Leandro, CA, USA; and Scion Image Software, Scion, Frederick, MD, USA). Mean ± s.e.m. were determined by averaging the data from a minmum of three separate experiments.

mRNA were extracted from total RNA using the Oligotex mRNA Mini-agents (Qiagen, Valencia, CA, USA) according to manufacturer's protocol. 0.1-0.3 g mRNA was screened for anti-K-ras ribozyme expression by standard RT-PCR with KRbzP1 primer (corresponding to CMV promoter) and KRbzP2 (corresponding to catalytic core of ribozyme insert) (130 bp; 94°C, 45 s; 56°C, 45 s; 72°C, 30 s; 25 cycles), or KRbz-P1 primer and KRbzP3 primer (corresponding to SV40 poly A signal of the ribozyme expression cassette) (301 bp; 94°C, 45 s; 54°C, 45 s; 72°C, 45 s; 32 cycles).²¹

For analysis by RPA, cDNA for K-ras (product I), KRbz and -actin was first prepared from untreated H441 cells, or KRbz-ADV treated H441 cells by RT-PCR assays and the appropriate primers. To add T7 promoter sequence into cDNA fragments (T7-DNA), each RT-PCR product (1 l, 1:20 dilution with H₂O) was re-amplified by PCR⁵⁶ using the following primers: KRbz T7-DNA (343 bp): K-ras T7-DNA (214 bp): with K-rasP4 and P5 primers (94°C, 1 min; 60°C, 45 s; 72°C, 45 s; 30 cycles); KRbzP4 and P5 primers (94°C, 1 min; 65°C, 45 s; 72°C, 45 s; 30 cycles); and -actin T7-DNA (291 bp) with -actin P3 and P4 primers (94°C, 1 min; 65°C, 45 sec; 72°C, 45 s; 30 cycles). Each T7 promoter-containing DNA template was resolved by agarose (2%) gel electrophoresis and extracted with QIAquick Gel Extraction agents (Qiagen).

Specific RNA probe for K-ras, KRbz and -actin (16 bases shorter than its corresponding template) was prepared from the T7-DNA template with MAXIscript in vitro Transcription (T7) agents (Ambion, Austin, TX, USA) in presence of UTP -³³P (NEN Life Science Products, Boston, MA, USA), purified by gel electrophoresis (6% TBE-urea gel: 6% polyacrylamide plus 7 M urea; Novex, San Diego, CA, USA), and eluted with RPA III agents (Ambion). The amount of radioactive label was determined by liquid scintillation (Betaplate 1205; Wallac). For the RPA reaction, 16 g of total RNA were hybridized with the appropriate amount of RNA probe (K-ras: 4 ´ 10⁴ c.p.m.; KRbz: 1 ´ 10⁵ c.p.m.; -actin: 1 ´ 10⁴ c.p.m. per reaction) at 52°C overnight, and digested by RNaseA/T1 mix. The protected RNA probe fragments (20 base shorter than the corresponding full-length probe) were resolved by 6% polyacrylamide gel electrophoresis (QuickPoint Pre-Cast gels, Novex, San Diego, CA, USA) alongside in vitro transcribed RNA size markers (RNA Century Marker Template; Ambion). Autoradiographic gel imaging was performed with Kodak BioMax film (Rochester, NY, USA) (14-16 h). Gel band densitometric analysis was performed as described above for RT-PCR assays. The level of ribozyme expression relative to K-ras mRNA was expressed as a molar ratio by comparing the optical density for their representative gel bands normalized to housekeeping -actin mRNA (an internal standard), and adjusted for molecular sizes.

The sequence for the primers used in RT-PCR and RPA studies were the following:

NSCLC cell line heterotransplantation

To determine optimal engraftment conditions, 1-10 ´ 10⁶ H1725 cells were injected subcutaneously (s.c.) into 4-6-week-old female NIH Swiss nude mice (NIHS-nufDF, Taconic Farm, Gaithersberg, MD, USA). Tumor engraftment was not evident in mice that received s.c. inoculation of 1 ´ 10⁶ cells for a period of 90 days after-injection. In contrast, all the mice that received 1 ´ 10⁷ H1725 cells developed palpable tumor xenografts that grew progressively at 2-4 weeks after inoculation. H441 xenograft nodules (inoculation of 1 ´ 10⁷ cells) were measurable at day 7, and these tumor xenografts grew progressively from 2 weeks after tumor inoculation.

DNA from tumor xenograft (but not mouse lung tissue) expressed the human K-ras sequence including the K-ras codon 12 GTT/GGT heterozygous mutation of the parental cell line, as defined by PCR. There was >95% tumor take in mice that received 1 ´ 10⁷ H1725 or H441 cells under optimized conditions. To study the antitumorigenic effect of KRbz treatment, subconfluent H1725 cells or H441 cells were either untreated, or infected with KRbz-ADV, mutKRbz-ADV or Luc-ADV (10 to 20 p.f.u. per cell, 20-24 h) before injecting into mice (s.c., 1 ´ 10⁷ cells in 0.1 ml RPMI 1640).⁵¹ The mice were observed twice weekly for tumor emergence. Palpable xenografts were measured by a vernier caliper twice a week for up to 42 days. Tumor size was expressed as the mean tumor diameter, based on the geometric mean of two perpendicular measurements. Animals were killed when the tumor xenograft exceeded 3 cm in diameter or at the end of experimental period of up to 60 days, whichever comes first. For treatment of pre-existing s.c. H441 xenografts, animals received single or multiple (three or five) intratumoral injections of KRbz-ADV or control Luc-ADV (2 ´ 10⁹ p.f.u. in 0.1 ml), or medium only, every other day starting at day 7 after tumor inoculation. All procedures involving animal use were approved by the Institutional Animal Care and Use Committee at Baylor University Medical Center. Differences in primary tumor xenograft size were compared statistically using the two-tailed Student's t test. Frequency of animals with post-treatment tumor regression was compared by the Fisher's exact two-sided test.

Analysis of apoptotic activity

The level of apoptosis induced by KRbz-ADV was determined by flow cytometric analysis following 7-amino actinomycin D (7-AAD) or annexin V and propidium iodide (PI) staining (FACScan, Becton Dickinson). Subconfluent cultures were infected with either KRbz-ADV or Luc-ADV vector (50 p.f.u. per cell, 20-24 h). The cells were trypsinized, then replated for 48 h (1 ´ 10⁶ cells per dish) in 60-mm standard culture dishes or polyHEMA-treated culture dishes (Sigma) that prevent cell adhesion.⁵⁷ The culture was harvested, washed, pelleted (1 ´ 10⁶ cells) and stained with 7-AAD (50 l, 400 M, 30 min, on ice) after washing in PBS and 1% FBS, then fixed with paraformaldehyde (0.5% in PBS). Analysis of 7-AAD uptake was carried out at the emitted wavelength of >650 nm following laser excitation at 488 nm, based on 1 ´ 10⁴ acquired events (FACScan, Becton Dickinson). The R&D apoptosis detection kit (R&D Systems, Minneapolis, MN, USA) was used to quantify the level of Annexin V-FITC- and propidium iodide-positive cells. Two color immunoflorescence analysis was performed using the FACScan CELLQUEST software (Becton Dickinson).

Acknowledgements

This work was supported in part by the Mary Crowley Research Foundation, the Robert A Schanbaum Memorial Fund for Cancer Research, and by a Berlex Biosciences Cancer Research grant. The authors thank Guido A Ordonez of the Immunology Laboratory, Baylor University Medical Center, for his assistance on flow cytometric analysis of apoptosis. We thank Joseph M Lawson and Beverly Peters, Baylor University Medical Center, for their assistance with preparation of the figures.

1 Martinez AC, Rebollo A. Ras proteins: recent advances and new functions. Blood 1999; 94: 2971-2980, MEDLINE

2 Zohn IM et al. Rho family proteins and Ras transformation: the RHOad less traveled gets congested. Oncogene 1998; 17: 1415-1438, MEDLINE

3 Downward J. Ras signaling and apoptosis. Curr Opin Genet Dev 1998; 8: 49-54, MEDLINE

4 Iran K et al. Mitogenic signaling mediated by oxidants in ras-transformed fibroblasts. Science 1997; 275: 1649-1652, Article MEDLINE

5 Rak J et al. Massive programmed cell death in intestinal epithelial cells induced by three-dimensional growth conditions: suppression by mutant c-H-ras oncogene expression. J Cell Biol 1995; 131: 1587-1598, MEDLINE

6 Okada F et al. Impact of oncogenes in tumor angiogenesis: mutant K-ras up-regulation of vascular endothelial growth factor/vascular permeability factor is necessary, but not sufficient for tumorigenicity of human colorectal carcinoma cells. Proc Natl Acad Sci USA 1998; 95: 3609-3614, Article MEDLINE

7 Hahn WC et al. Creation of human tumour cells with defined genetic elements. Nature 1999; 400: 464-468, Article MEDLINE

8 Chin L et al. Essential role for oncogenic Ras in tumour maintenance. Nature 1999; 400: 468-472, Article MEDLINE

9 Rodenhuis S. RAS oncogenes and human lung cancer. In: Pass HI, Mitchell JB, Johnson DH, Turrisi AT (eds). Lung Cancer: Principles and Practice. Lippincott-Raven: Philadelphia, 1996, pp ;73-82.

10 Sun J, Qian Y, Hamilton AD, Sebti SM. Ras CAAX peptidomimetic FTI 276 selectively blocks tumor growth in nude mice of a human lung carcinoma with K-ras mutation and p53 deletion. Cancer Res 1995; 55: 4243-4247, MEDLINE

11 Zhang WW et al. Retroviral vector-mediated transduction of K-ras antisense RNA into human lung cancer cells inhibits expression of the malignant phenotype. Hum Gene Ther 1993; 4: 451-460, MEDLINE

12 Georges RN et al. Prevention of orthotopic human lung cancer growth by intratracheal instillation of a retroviral antisense K-ras construct. Cancer Res 1993; 53: 1743-1746, MEDLINE

13 Cech TR. Self-splicing of group I introns. Ann Rev Biochem 1990; 59: 543-568, MEDLINE

14 Kijima H et al. Therapeutic applications of ribozymes. Pharmacol Ther 1995; 68: 247-267, MEDLINE

15 Kashani-Sabet M, Scanlon KJ. Application of ribozymes to cancer gene therapy. Cancer Gene Ther 1995; 2: 213-223, MEDLINE

16 Ohta Y et al. H-ras ribozyme-mediated alteration of the human melanoma phenotype. Ann NY Acad Sci 1994; 716: 243-256,

17 Tone T et al. Suppression of EJ cells tumorigenicity. In Vivo 1993; 7: 471-476, MEDLINE

18 Kashani-Sabet M et al. Suppression of the neoplastic phenotype in vivo by an anti-ras ribozyme. Cancer Res 1994; 54: 900-902, MEDLINE

19 Kijima H, Bouffard DY, Scanlon KJ. Ribozyme-mediated reversal of human pancreatic carcinoma phenotype. In: Ikehara S, Takahu F, Good R (eds). Proceedings of International Symposium on Bone Marrow Transplantation. Springer-Verlag: Tokyo, 1996, 153-163.

20 Tsuchida T et al. Hammerhead ribozyme specifically inhibits mutant K-ras mRNA of human pancreatic cancer cells. Biochem Biophys Res Commun 1998; 253: 368-373, MEDLINE

21 Tong AW, Zhang YA, Nemunaitis J, Mues G. K-ras ribozyme for lung cancer. In: Scanlon KJ (ed). Therapeutic Application of Ribozymes: Methods in Molecular Medicine. vol. 11: Humana Press: Clifton, NJ, 1998, pp 209-222.

22 Zhang YA et al. In vitro suppression of human non-small cell lung cancer (NSCLC) cell growth by an anti-K-ras ribozyme. FASEB J 1996; 10: 6 (Abstr. 1409),

23 Cantor GH et al. Ribozyme cleaves rex/tax mRNA and inhibits bovine leukemia virus expression. Proc Natl Acad Sci USA 1993; 90: 10932-10936, MEDLINE

24 Kashani-Sabet M et al. Reversal of the malignant phenotype by an anti-ras ribozyme. Antisense Res Dev 1992; 2: 3-15, MEDLINE

25 Yamazaki H et al. Inhibition of tumor growth by ribozyme-mediated suppression of aberrant epidermal growth factor receptor gene expression. J Natl Cancer Inst 1998; 90: 581-587, MEDLINE

26 Irie A et al. Therapeutic efficacy of an adenovirus-mediated anti-H-ras ribozyme in experimental bladder cancer. Antisense Nucleic Acid Drug Dev 1999; 9: 341-349, MEDLINE

27 Marches R et al. Tumor dormancy and cell signalling III: role of hypercrosslinking of IgM and CD40 on the induction of cell cycle arrest and apoptosis in B lymphoma cells. Ther Immunol 1995; 2: 125-136, MEDLINE

28 Lecoeur H, Ledru E, Prevost MC, Gougeon ML. Strategies for phenotyping apoptotic peripheral human lymphocytes comparing ISNT, annexin-V and 7-AAD cytofluorometric staining methods. J Immunol Methods 1997; 209: 111-123, MEDLINE

29 Kung HF et al. Reversal of transformed phenotype by monoclonal antibodies against Ha-ras p21 proteins. Exp Cell Res 1986; 162: 363-371, MEDLINE

30 Cochet O et al. Intracellular expression of an antibody fragment-neutralizing p21 Ras promotes tumor regression. Cancer Res 1998; 58: 1170-1176, MEDLINE

31 Nemunaitis J et al. Prognostic value of K-ras mutations, Ras oncoprotein and c-erb B-2 oncoprotein expression in adenocarcinoma of the lung. Am J Clin Oncol 1998; 21: 155-160, MEDLINE

32 Kozarsky KF, Wilson JM. Gene therapy: adenovirus vectors. Curr Opin Genet Dev 1993; 3: 499-503, MEDLINE

33 Mahgoub N et al. In vitro and in vivo effects of a farnesyltransferase inhibitor on Nf1-deficient hematopoietic cells. Blood 1999; 94: 2469-2476, MEDLINE

34 Sun J, Qian Y, Hamilton AD, Sebti SM. Both farnesyltransferase and geranylgeranyltransferase I inhibitors are required for inhibition of oncogenic K-ras prenylation but each alone is sufficient to suppress human tumor growth in nude mouse xenografts. Oncogene 1998; 16: 1467-1473, MEDLINE

35 Rodenhuis S. Ras and human tumors. Semin Cancer Bio 1992; 3: 241-247,

36 Tong AW, Zhang YA, Bouffard DY, Nemunaitis J. The use of ribozymes for gene therapy of lung cancer. In: Kashani-Sabet M, Scanlon KJ (eds). Ribozyme in Gene Therapy of Cancer. RG Landes: Georgetown, TX, 1998, pp 151-164.

37 Zabner J et al. Safety and efficacy of repetitive adenovirus-mediated transfer of CFTR cDNA to airway epithelia of primates and cotton rats. Nat Genet 1994; 6: 75-83, MEDLINE

38 Dong JY et al. Systematic analysis of repeated gene delivery into animal lungs with a recombinant adenovirus vector. Hum Gene Ther 1996; 7: 319-331, MEDLINE

39 Zhang WW et al. Safety evaluation of Ad5CMV-p53 in vitro and in vivo. Hum Gene Ther 1995; 6: 155-164, MEDLINE

40 Nechers L, Whitesell L, Rosolen A, Geselowitz DA. Antisense inhibition of oncogene expression. Crit Rev Oncogen 1992; 3: 175-231,

41 Aoki K et al. Suppression of Ki-ras p21 levels leading to growth inhibition of pancreatic cancer cell lines with Ki-ras mutation but not those without Ki-ras mutation. Mol Carcinogen 1997; 20: 251-258,

42 Leonidas A et al. Hammerhead ribozymes targeted to the FBN1 mRNA can discriminate a single base mismatch between ribozyme and target. Biochem Biophys Res Commun 1998; 249: 804-810, MEDLINE

43 Ward RL et al. Activation of the K-ras oncogene in colorectal neoplasms is associated with decreased apoptosis. Cancer 1997; 79: 1106-1113, MEDLINE

44 Kerbel RS, Viloria-Petit A, Okada FTI, Rak J. Establishing a link between oncogenes and tumor angiogenesis. Mol Med 1998; 4: 286-295, MEDLINE

45 Robles A et al. Reduced skin tumor development in cyclin D1-deficient mice highlights the oncogenic Ras pathway in vivo. Genes Dev 1998; 12: 2469-2474, MEDLINE

46 Bishoff JR et al. An adenovirus mutant that replicates selectively in p53-deficient human tumor cells. Science 1996; 274: 373-376, Article MEDLINE

47 Mtistudomi T et al. p53 gene mutations in non-small-cell lung cancer cell lines and their correlation with the presence of ras mutations and clinical features. Oncogene 1992; 7: 171-180, MEDLINE

48 Phelps RM et al. NCI-Navy medical oncology branch cell line data base. J Cell Biochem 1996; 24: (Suppl.) 32-91,

49 Tsuchida T et al. Adenovirus-mediated anti-K-ras ribozyme induces apoptosis and growth suppression of human pancreatic carcinoma. Cancer Gene Ther 2000; 7: 373-383, MEDLINE

50 Becker TC et al. Use of recombinant adenovirus for metabolic engineering of mammalian cells. Meth Cell Biol 1994; 43: 161-189,

51 Zhang YA, Nemunaitis J, Tong AW. Anti-K-ras ribozyme adenoviral vector for gene therapy of non-small cell lung cancer. In: Walther W, Stein U (eds). Gene Therapy of Cancer: Methods and Protocols: Methods in Molecular Medicine. vol. 35: Humana Press: Totowa, NJ, 1999, pp 257-267.

52 Feng M et al. Neoplastic reversion accomplished by high efficiency adenoviral-mediated delivery of an anti-ras ribozyme. Cancer Res 1995; 55: 2024-2028, MEDLINE

53 Toloza EM et al. In vivo cancer gene therapy with a recombinant interleukin-2 adenovirus vector. Cancer Gene Ther 1996; 3: 11-17, MEDLINE

54 Heike Y et al. Interleukin-2 gene transduction into freshly isolated lung adenocarcinoma cells with adenoviral vectors. Hum Gene Ther 1997; 8: 1-14, MEDLINE

55 Tong AW et al. Anti-CD40 antibody binding modulates human multiple myeloma clonogenicity in vitro. Blood 1994; 84: 3026-3033, MEDLINE

56 Logel J et al. Synthesis of cRNA probes from PCR-generated DNA. BioTechniques 1992; 13: 604-610, MEDLINE

57 Lebowitz PF et al. Farnesyl transferase inhibitors induce apoptosis of ras-transformed cells denied substratum attachment. Cancer Res 1997; 57: 708-713, MEDLINE

Figure 1 Growth inhibition by KRbz-ADV and Luc-ADV. H441 or H1725 cells (5 ´ 10³ per well) were infected by KRbz-ADV, control Luc-ADV, or mutKRbz-ADV vectors at 25 p.f.u. per cell, and cultured for 72 h (see Materials and methods). Percentage inhibition was determined by comparison of ³H-thymidine uptake with the untreated control (medium only). Value represents percentage mean inhibition (±s.d.; n = 6-9).

Figure 2 RT-PCR analysis of K-ras mRNA expression. RT-PCR analysis of uncleaved (product I) and total (cleaved + uncleaved; product II) K-ras mRNA was carried out using K-ras specific primers with concomitant amplification of -actin mRNA. (a) Schematic representation of K-ras specific primer locations (P1, P2 and P3) with respect to K-ras mRNA. (b) RT-PCR amplification products (K-ras 32 cycles; -actin: 25 cycles) as revealed by electrophoresis on 2% agarose gel with ethidium bromide staining. Lanes 1 and 4, untreated H441 cells; lanes 2 and 5, KRbz-ADV-treated H441 cells; lanes 3 and 6, Luc-ADV-treated H441 cells. (c) Relative levels (mean % ± s.e.m.) of K-ras mRNA as compared with untreated cultures, as determined by densitometric analysis of RJ-PCR products (see Materials and methods).

Figure 3 RPA analysis of K-ras mRNA expression. RPA was performed with radiolabeled RNA probes specific for KRbz (a), K-ras (b) and -actin (c), and RNA extracts from untreated H441 cells (lane 2) or cells treated by KRbz-ADV (lane 3), control Luc-ADV (lane 4) or mutKRbz-ADV (lane 5) for 72 h (100 p.f.u. per cell). Lane 1, corresponding full-length probes for KRbz (327 bases), K-ras (198 bases) and -actin (275 bases). Arrows indicate the expected positions for the intact full-length probes and their RNA-protected fragments (20 bases shorter), based on migration distance of concurrently resolved RNA size markers (not shown).

Figure 4 Antitumorigenic effect of KRbz-ADV in vivo. H441 cells was either untreated, or treated with KRbz-ADV, mutKRbz-ADV or Luc-ADV for 24 h before subcutaneous inoculation into athymic mice (1 ´ 10⁷ cells per mouse). Tumor growth was followed for up to 40 days. Value represents mean tumor diameter (s.d. 20% of mean).

Table 1 Adenoviral transduction efficiency as determined by the X-gal staining assay

Table 2 Increased growth inhibition of NSCLC xenografts by repeated injections of KRbz-ADV

Table 3 Quantification of apoptotic cells