Retinoids, including vitamin A and its analogues, regulate the growth and differentiation of cells by modulating the expression of a wide variety of genes. Previous animal studies have demonstrated that vitamin A deficiency is associated with increased susceptibility to lung cancer and that retinoids suppress carcinogenesis in many epithelial tissues.1, 2, 3 In addition, differentiation-induced complete remissions have been achieved with all-trans-retinoic acid (ATRA), which is a commonly used retinoid derived from vitamin A and β-carotene metabolism, in patients with acute promyelocytic leukemia (APL).4, 5, 6 These observations suggest the potential of retinoids for understanding the biology of cancer as well as for developing new treatment strategies.
The effects of retinoids are mediated through binding to 2 types of nuclear retinoid receptors: retinoic acid (RA) receptors (RARs) and retinoid X receptors (RXRs), which differ in their retinoid-binding specificity. Three major forms of RAR (RARα, RARβ, and RARγ), which share substantial homology, have been identified and characterized.7 Each receptor subtype is differentially expressed in various adult tissues and therefore considered to regulate the expression of the distinct sets of genes. Several studies have demonstrated that the expression of RARβ is selectively lost in most of cultured cancer cell lines as well as premalignant and malignant lesions obtained from patients, whereas normal tissues express detectable levels of RARβ.8, 9, 10, 11 In addition, transgenic mice that express the antisense RARβ2 gene, which inactivates the endogenous RARβ function developed pulmonary tumors.12 These results suggest that a defect of RARβ expression is associated with human cancer development. The finding that transfectants expressing the RARβ gene grow at reduced rates in the presence of RA and are less tumorigenic than the parental cells in nude mice, however, provide evidence that the abnormalities are reversible.13
p21, also known as sdi1,14 Cip1,15 WAF1,16 or mda-6,17 was identified as a molecule that regulates the transition from the G1 phase to the S-phase of the cell cycle. During skeletal muscle differentiation, muscle-specific transcriptional regulator, MyoD, increases p21sdi1 expression, thereby inducing terminal cell cycle arrest,18 Moreover, overexpression of p21sdi1 has been shown to induce differentiation in monoblastic cell lines19 and human esophageal squamous cell carcinoma.20 These results indicate that p21sdi1 is involved in a terminal differentiation program in normal as well as cancer cells. We demonstrate that ectopic p21sdi1 gene transfer augments RARβ expression and induces the sensitivity to ATRA in human cancer cells. Our data could be potentially important for the developing novel differentiation-directed anti-cancer therapy, which can redirect cancer cells to the normal phenotype.
The human non-small cell lung cancer cell lines H1299, which has homozygously deleted p53 and the human colon carcinoma cell line DLD-1, which exhibits a homozygous p53 gene mutation, were routinely propagated in monolayer culture in RPMI 1640 medium supplemented with 10% FCS, 25 mM HEPES, 100 U/ml penicillin, and 100 mg/ml streptomycin. The transformed embryonic kidney cell line 293 was grown in DMEM with high glucose (4.5 g/l), supplemented with 10% FCS, 100 U/ml penicillin, and 100 mg/ml streptomycin. The 293 cells were used for the production of adenovirus vectors.
The recombinant adenovirus vector expressing human p21sdi1 cDNA was previously constructed and characterized.21 The resultant virus was named Ad5CMV-p21. Adenoviral vectors containing luciferase cDNA (Ad5RSVLuc) or no expression cassette (dl312) were used as a control vector. The viral stocks were quantified by a plaque-forming assay using 293 cells and stored at −80°C.
ATRA was purchased from Sigma Chemical Co. (St. Louis, MO), dissolved in dimethyl sulphoxide (DMSO) at 1 mM for in vitro studies and 1.5 mg/ml for in vivo studies, and used as RA. It was stored at −30°C and protected from light.
The cytotoxicity of ATRA was determined by measurement of cell viability by using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Cells were seeded in 96-well tissue culture plates and treated the next day with ATRA at the indicated concentrations. After 3 days, the cells were washed once with PBS. A medium containing 0.5 mg/ml MTT (Sigma) was added to each well. The cells were incubated at 37°C for 4 hr and then an equal amount of solubilization solution (0.04 N HCl in isopropyl alcohol) was added to each well and mixed thoroughly to dissolve the crystals of MTT formazan. After all of the crystals were dissolved, the plates were read on a Labsystems Multiskan MS at 540 nm.
Total RNA was isolated from mock-, dl312-, and Ad5CMV-p21-infected cells using RNAzol (Cinna/BioTecx, Friendswood, TX) in a single-step phenol-extraction method and used as templates. Reverse transcription was carried out at 22°C for 10 min and then 42°C for 20 min using 1.0 μg of RNA per reaction to ensure that the amount of amplified DNA was proportional to that of specific mRNA in the original sample. PCR was carried out with specific primers in volumes of 50-μl according to the s protocol provided by the manufacturer (PCR kit; Perkin-Elmer/Cetus, Norwalk, CT). The specific primers used for p21 were p21-S (5′-GCG CCA TGT CAG AAC CGG CTG-3′) and p21-AS (5′-GCA GGC TTC CTG TGG GGG GAT-3′); for the RARβ were RARβ-S (5′-TGC TTC GTC TGC CAG GAC-3′) and RARβ-AS (5′-TTG CAC GAG TGG TGA CTG-3′); and for GAPDH were GAPDH-S (5′-CAG CCG AGC CAC ATC-3′) and GAPDH-AS (5′-TGA GGC TGT TGT CAT ACT TCT-3′). For p21sdi1, reactions were run according to the following cycle profile: denaturation at 95°C for 30 sec, annealing at 58°C for 30 sec, and extension at 72°C for 1 min using a thermal cycler (Perkin-Elmer, Foster City, CA) for 25 cycles. The amplification reaction for RARβ involved denaturation at 92°C for 1 min, annealing at 58°C for 1 min, and 72°C for 1 min using a thermal cycler for 35 cycles. The PCR products were resolved on 1% agarose gels and visualized by SYBR Gold Nucleic Acid Gel Stain (Molecular Probes, Inc., Eugene, OR).
Western blot analysis
Cells were collected by trypsinization and washed twice in cold phosphate-buffered saline (PBS). To detect p21sdi1 protein, cells were then lysed in SDS solubilization buffer (62.5 mM Tris-HCl, pH 6.8, containing 10% glycerol, 5% β-mercaptoethanol, and 2% SDS; Sigma). For detection of RARβ protein, the nuclear extracts were prepared. Briefly, cells were washed in PBS, pelleted, and resuspended in lysis buffer (10 mM Tris-HCl [pH 8.0], 60 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, protease inhibitors, and 0.1% NP-40). After 5 min on ice, the lysates were spun at 2,500 rpm in a microcentrifuge at 4°C for 4 min. The supernatant was then removed and spun at 14,000 rpm for 5 min. The supernatants were used as cytoplasmic extracts. The pelleted nuclei from the first spin were briefly washed in lysis buffer without NP-40. The nuclear pellet was then resuspended in an equal-volume nuclear extract buffer (20 mM Tris-Hcl [pH8.0], 420 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, and 25% glycerol). After a 10 min incubation at 4°C, the nuclei were briefly vortexed and spun at 14,000 rpm for 5 min. The supernatant was then removed and used as a nuclear extract. Protein concentration was determined using the Bio-Rad protein determination method (Bio-Rad, Hercules, CA). Equal amounts (20 μg) of proteins were boiled for 5 min and electrophoresed under reducing conditions on 12.5% (w/v) polyacrylamide gels. Proteins were electrophoretically transferred to a Hybond-polyvinylidene difluoride (PVDF) transfer membranes (Amersham, Arlington Heights, IL), and incubated with primary antibodies against p21 (EA10) (Oncogene Science, Manhasset, NY), RARβ (C-19; Santa Cruz Biotechnology, Santa Cruz, CA), or actin (AC-40; Sigma), followed by peroxidase-linked secondary antibody. An Amersham ECL chemiluminescent western system (Amersham, Japan) was used to detect secondary probes.
Transfection procedure and CAT assay
To measure the transcriptional activation of RARβ promoter in human cancer cells, the RARβ promoter linked with chloramphenicol acetyltransferase (CAT) gene was kindly provided by Dr. Kazuo Umezono (Kyoto University) and used as a reporter gene.22, 23 Cells (2 × 105/well) were seeded on 6-well tissue culture plates and incubated until 90% confluent, followed by the addition of 500 μl of OPTI-MEM I Reduced Medium (Life Technologies, Inc.) containing 5 μg of RARβ promoter/CAT reporter plasmid and 15 μl of LIPOFECTAMINE 2000 Reagent (Life Technologies, Inc.). Cells were incubated for 24 hr, infected with Ad5CMV-p21 or control dl312 vector, and then grown for an additional 48 hr before harvest. For harvest, cells were washed twice in PBS and scraped into Reporter Gene Assay Lysis Buffer (Boehringer-Mannheim, Indianapolis, IN). The cell extracts were immediately assayed for CAT activity using the CAT ELISA kit (Roche Molecular Biochemicals, Mannheim, Germany).
Visualization of apoptotic nuclei
Tumor cells (1 × 106) were grown on plastic cover slips and treated as described. Cells were fixed with 1.0% glutaraldehyde and stained with the DNA intercalating dye Hoechst 33342 (1 mM in PBS). Cells were then viewed and photographed with a fluorescence microscope (Zeiss, Oberkochen, Germany).
In vivo evaluation of tumor growth
H1299 cells (1 × 107) resuspended in Hank's balanced salt solution at a final volume of 100 μl were injected subcutaneously in the flank of 4-week-old BALB/c female nu/nu mice that were kept in a temperature-controlled room on a 12/12-hr light/dark schedule with food and water. The mice were randomly divided into 4 groups of 5–6 mice each. ATRA (100 μl per injection) or control PBS was administered intraperitoneally and a 100-μl solution containing 1 × 109 PFU of Ad5CMV-p21 or dl312 was injected intratumorally. The perpendicular diameters of subcutaneous H1299 tumors were measured with calipers and tumor volume was calculated as length × width2 × 0.52. The experimental protocol was approved by the Ethics Review Committee for Animal Experimentation of Okayama University Graduate School of Medicine and Dentistry.
Differences in the tumor volumes for the treatment groups were compared using the Student's t-test. Statistical significance was defined as p < 0.05.
Effect of ATRA on growth of human cancer cell lines
Human cancer cell lines H1299 and DLD-1, which have a p53-null and missense mutant p53 status, respectively, were chosen for our study. We first tested the effect of RA on the in vitro growth of these cell lines. The growth rate of both cell lines were not affected in the presence of 1 μM of ATRA, although the addition of more than 1 μM of ATRA to the medium resulted in suppression of the growth of H1299 and DLD-1 cells (Fig. 1). We also studied alterations in cell morphology after ATRA treatment at a dose of 1 μM; there was, however, no apparent morphological changes, such as an enlarged, flattened shape, an increased cytoplasmic-to-nuclear ratio, or decreased cell density, consistent with the differentiation phenotype (data no shown). These results suggest that these cell lines are relatively resistant to this concentration of ATRA.
Adenovirally transduced p21sdi1 gene expression in human cancer cell lines
Monolayer cultures of H1299 and DLD-1 cell lines were infected with either a replication-deficient adenovirus vector carrying a p21sdi1 gene under the control of the cytomegalovirus (CMV) promoter (Ad5CMV-p21) or control dl312 at a multiplicity of infection (MOI) of 30 and 100, respectively, in the presence or absence of 0.5 μM of ATRA. Semi-quantitative RT-PCR assay demonstrated readily detectable p21sdi1 mRNA expression in both cell lines as early as 12 hr after infection, which was enhanced in the presence of ATRA, but not in parental and dl312-infected cells (Fig. 2a). In addition, endogenous p21sdi1 mRNA expression was apparently induced in both cell lines, when cells were treated with 0.5 μM of ATRA alone for 48 hr.
To further confirm the functional p21sdi1 gene transduction, we conducted Western blot analysis using antibody against human p21sdi1. As shown in Figure 2b, high levels of p21sdi1 protein expression was achieved after Ad5CMV-p21 infection in both cell lines, and further increase in p21sdi1 protein expression occurred after treatment of cells with 0.5 μM ATRA. Consistent with the results of the RT-PCR analysis, ATRA alone induced endogenous p21 protein expression. These results confirmed our previous observations that the adenovirus-mediated gene transfer and expression were highly efficient on a variety of human cancer cell lines. There was no apparent toxicity of adenovirus infection up to 100 MOI (data not shown).
Effect of p21sdi1 gene transfer on RARβ expression in H1299 and DLD-1 cells
To examine whether overexpression of p21sdi1 protein specifically affects the expression of RARβ, we first analyzed RARβ expression by semi-quantitative RT-PCR and Western blotting. As illustrated in Figure 3a, both cell lines did not expressed RARβ transcripts, whereas Ad5CMV-p21 upregulated RARβ mRNA expression 48 hr after infection. Moreover, when cells were treated with Ad5CMV-p21 in combination with 0.5 μM ATRA, increased RARβ expression was detectable within 12 hr after treatment, which was markedly enhanced at 48 hr post-infection. Retinoids are known to increase RARβ transcription; our data, however, demonstrated that RARβ expression was weakly induced by ATRA treatment alone in the cell lines used in our study. Control dl312 infection at 100 MOI had no effects on the expression of RARβ mRNA in both cell lines.
We then investigated whether the ectopic p21sdi1 gene transfer changes RARβ protein expression on these cell lines using Western blot analysis (Fig. 3b). Ad5CMV-p21 infection alone and in the combination with 0.5 μM of RA resulted in an apparent increase in the levels of RARβ protein in H1299 and DLD-1 cells within 24 hr after infection compared to the mock-infected cells. Moreover, RARβ protein expression was not detectable after RA treatment alone.
To determine whether RARβ expression was dependent on transcriptional regulation by p21sdi1, we transiently transfected H1299 and DLD-1 cells with a CAT reporter construct containing an upstream promoter sequence of the RARβ gene. As can been seen in Figure 3c, Ad5CMV-p21 infection induced the RARβ promoter activity in a dose-dependent manner in both cell lines. Treatment of transfected cells with Ad5CMV-p21 at an MOI of 30 or 100 led to an approximately 2.7-fold or 1.3-fold induction of the RARβ promoter activity above untreated levels by 24 hr after infection in H1299 and DLD-1 cells, respectively. In contrast, the promoter activity was unaffected by infection with control dl312.
Effect of Ad5CMV-p21 and ATRA on growth of human cancer cells
To determine whether RARβ up-regulation can restore RA responsiveness, we tested the effects of ATRA on the in vitro growth of human cancer cells in the presence or absence of Ad5CMV-p21 infection. H1299 and DLD-1 cells were infected with 30 and 100 MOI of Ad5CMV-p21, respectively, and then exposed to 1 μM of ATRA for 72 hr and 96 hr, respectively. Because p21sdi1 is a major negative regulator of the cell cycle, overexpression of p21sdi1 alone decreased the rate of proliferation in both cell lines. The responsiveness of Ad5CMV-p21-infected cells to ATRA was significantly higher than that of parental cells, however, although ATRA alone did not affect the proliferation of parental cells (Fig. 4a).
To further characterize the effect of Ad5CMV-p21 infection on RA sensitivity, treated cells were analyzed for the induction of apoptosis by Hoechst staining. As shown in Figure 4b, apoptotic changes such as apoptotic bodies, chromatin condensation, and fragmented nuclei were observed in both cells treated with a combination of Ad5CMV-p21 and ATRA, whereas neither ATRA treatment alone nor Ad5CMV-p21 infection alone caused apoptotic cell death in both cell lines.
Effect of intratumoral injection of Ad5CMV-p21 and systemic administration of ATRA on growth of H1299 tumors transplanted in nu/nu mice
Results of in vitro experiments led us to investigate the effect of in vivo administration of RA combined with Ad5CMV-p21 injection. We have chosen H1299 cells for in vivo study, because they most rapidly responded to Ad5CMV-p21 in combination with ATRA in vitro (Fig. 4a). When 1 × 107 H1299 cells were subcutaneously inoculated into nu/nu mice, palpable tumors appeared in almost 100% of mice 7 days after injection. Fourteen days after tumor inoculation, animals bearing tumors with a diameter of 5–7 mm were treated with a single direct intratumoral injection of 1 × 109 PFU of Ad5CMV-p21 and intraperitoneal administration of 7.5 mg/kg of ATRA over three consecutive days. As shown in Figure 5a, tumor growth was significantly inhibited by the combination of Ad5CMV-p21 injection with intraperitoneal administration of ATRA compared to untreated tumors (p < 0.05), although the effect of Ad5CMV-p21 or ATRA alone was not statistically significant. Preliminary experiments demonstrated that the in vivo growth of control dl312-injected H1299 tumors was similar to that of mock-treated tumors (data not shown).
Finally, to further assess the efficacy of locoregional p21sdi1 gene transfer combined with ATRA, we examined the antitumor effect of repeated intratumoral injection of Ad5CMV-p21. Treatment with Ad5CMV-p21 plus ATRA over three consecutive days resulted in significant inhibition of tumor growth compared to tumors treated with Ad5CMV-p21 alone (p < 0.05) (Fig. 5b). The results indicate that adenovirus-mediated p21sdi1 gene transfer acts synergistically with systemically administered ATRA to suppress the in vivo growth of human lung tumors.
Many of cell lines derived form human solid tumors are resistant to the growth inhibitory effects of ATRA and other retinoids. It has been suggested that aberration of RARβ expression may be one of the reasons for their retinoid resistance.8, 9 Changes in the expression of RARβ could abrogate the retinoid signaling pathway and result in enhanced carcinogenesis, whereas increased RARβ expression could restore sensitivity to growth inhibition by retinoids such as ATRA.13, 24, 25 We have demonstrated that adenovirus-mediated p21sdi1 gene transfer enhances sensitivity to growth inhibition by ATRA through induction of RARβ expression. We showed previously that adenovirus-mediated transfer of p21sdi1 gene induces terminal differentiation in human esophageal squamous cell carcinoma cells, leading to a sequential apoptotic cell death.20 Our results provide evidence for the potential of p21sdi1 combined with ATRA in suppressing human cancer cell growth in vitro and in vivo, as well as in inducing apoptotic cell death.
We used H1299 human lung cancer cells and DLD-1 human colorectal cancer cells, both of which do not express RARβ mRNA constitutively (Fig. 3a). These cell lines are relatively resistant to 1 μM of ATRA, although higher concentrations of ATRA could suppress their growth in vitro (Fig. 1). Therefore, we considered that these cell lines might be a suitable system to explore the role of p21sdi1 gene transfer in the induction of RARβ expression and in the growth-inhibitory effects of ATRA. Because adenovirus-mediated gene transfer is a highly efficient gene delivery system, we used a recombinant adenovirus to overexpress p21sdi1 in H1299 and DLD-1 cells (Fig. 2). Several previous studies have demonstrated that RA treatment leads to the transcriptional activation of the p21sdi1 gene, which might be essential in facilitating the differentiation pathway.26, 27 Consistent with this finding, ATRA alone upregulated endogenous p21sdi1 expression 48 hr after treatment and, moreover, Ad5CMV-p21 infection in combination with ATRA treatment induced further increase in p21sdi1 expression as early as 12 hr after treatment (Fig. 2). A rapid p21sdi1 upregulation by Ad5CMV-p21 plus ATRA is most likely mediated through a positive feedback regulation of exogenous p21sdi1 expression by ATRA, although further studies are required to determine the precise mechanism.
Studies with different human cancer cell lines have demonstrated that RARβ was upregulated by RA treatment when constitutively expressed,25, 28 but it has been also reported that RARβ was not detected in RARβ-negative cell lines after ATRA treatment even when RT-PCR was used.23 In fact, our data showed that ATRA alone failed to induce an apparent RARβ mRNA expression in H1299 and DLD-1 cell lines, which do not express RARβ. Exogenous p21sdi1 expression, however, was apparently associated with an induction of RARβ 48 hr after infection. In addition, the combination of Ad5CMV-p21 infection and ATRA treatment showed a rapid and enhanced induction of RARβ expression (Fig. 3a), which is in agreement with the induction pattern of p21sdi1 expression (Fig. 2). Recent studies have shown that aberrant methylation of the RARβ promoter is one mechanism that silences RARβ expression in many human lung cancer cells.29, 30 Moreover, the promoter region of RARβ in DLD-1 cells has been reported to contain 16 methylated CpG islands that are associated with transcriptional inactivation.31 Treatment with 5-aza-2′-deoxycytidine (5-Aza-CdR) restored RARβ transcript expression, however, indicating that methylation-related gene silencing may not be irreversible.32.32
It is known that ATRA-induced RARβ expression is mediated by the retinoic acid response element (RARE) present in the RARβ promoter. The CAT reporter assay demonstrated that exogenous p21sdi1 expression could induce the RARβ promoter activity (Fig. 3c); however, the mechanism by which RARE can be transactivated by Ad5CMV-p21 infection is not well understood. Chen et al.33 have demonstrated that overexpression of p21sdi1 mediates the activation of the estrogen-signaling pathway in human breast cancer cells through transcriptional induction of the estrogen receptor (ER) expression. ER belongs to a superfamily of ligand-activated nuclear transcriptional factors, which includes receptors for steroids, thyroid hormones, and RA, suggesting that a common mechanism(s) might be involved in p21sdi1-induced activation of nuclear receptors in human cancer cells. Previous studies showed that p21sdi1 enhances transcriptional activation by nuclear factor κB (NF-κB) through interactions with the coactivator p300.34 Others have reported that transcriptional activation by coactivators p300 and CREB binding protein (CBP) could be stimulated by co-expression of p21sdi1.35 In addition, Kawasaki et al.23 have demonstrated that ribozyme-mediated disruption of p300 expression impaired RARβ promoter activity. Considered together, we speculate that p21sdi1 stimulates p300 or CBP activity through specific interaction, which in turn transactivates target gene expression such as ER and RARβ. Studies are underway currently in our laboratories to examine this possibility.
Transcriptional activation of RARβ gene by p21sdi1 has important clinical implications. The combination of 13-cis-RA with paclitaxel or cisplatin or interferon-β has entered Phase I/II trials for treatment of head and neck squamous cell carcinomas, non-small cell lung carcinomas and prostate cancer.36, 37 More recently, RA combined with a site-selective cAMP analogue, 8-Cl-cAMP, have been shown to act synergistically to kill human breast cancer cells through induction of RARβ gene.38 We demonstrated here that Ad5CMV-p21 infection and ATRA treatment cooperatively exhibited antitumor effects by inducing apoptotic cell death in human cancer cells in vitro and in vivo (Figs. 4,5). The benefit of intratumoral injection of Ad5CMV-p21 include the locoregional sensitization of cancer cells to ATRA. It has been reported that repeated intratumoral injections of high doses of adenovirus vectors is often well tolerated and result in increased transgene expression.39, 40 Thus, intratumoral administration of Ad5CMV-p21 may offer the potential advantage of being devoid of systemic adverse side effects and be sufficient for inducing site-specific apoptosis in the presence of ATRA.
In conclusion, we have demonstrated that adenovirus-mediated p21sdi1 gene transfer can sensitize human cancer cells to retinoids such as ATRA through transcriptional activation of RARβ gene. Chemoprevention is an attractive approach in the treatment of human solid tumors; the obstacle of this therapeutic strategy, however, is their retinoid resistance. Our data suggest a novel approach using p21sdi1 gene to enhance the retinoid sensitivity or to overcome retinoid resistance.
We are grateful to Y. Takata for the excellent technical support and Dr. K. Umezono for providing us with the RARβ promoter CAT plasmid.