|Home | About | Journals | Submit | Contact Us | Français|
The treatment of malignant gliomas with current therapies remains a challenge in neurooncology. Our recent work showed that embryonic stem cell (ESC)-derived astrocytes conditionally expressing genes can be used to induce apoptosis in malignant glioma cells in vitro. The tumor necrosis factor–related apoptosis-inducing ligand (TRAIL) gene has been shown to induce apoptosis in a variety of tumor cells, including gliomas. The aim of this study was to assess the proapoptotic effects of transgenic TRAIL delivered by ESC-derived astrocytes on malignant gliomas in vivo. Malignant glioma A172 cells were used to induce heterotopic xenografts in nude mice. ESC-derived astrocytes conditionally expressing TRAIL were injected into the xenografts. TRAIL expression was documented in the malignant glioma xenografts by reverse transcription PCR and immunohistochemistry after external gene induction. A significant reduction in tumor volume occurred 48 h after a single injection (14%) and double injections (31%) in the experimental groups. Terminal dUTP nick end labeling (TUNEL) revealed abundant apoptotic tumor cells in the experimental groups. Seven days after injection, the tumor had undergone severe necrosis, with only scattered residual tumor cells at the periphery. Death receptor DR4 expression increased significantly in the experimental groups compared with controls. Our data suggest that ESC-derived astrocytes conditionally expressing TRAIL should be considered as vectors to deliver gene therapy for malignant gliomas.
Currently, there is no optimal treatment for malignant gliomas, the most common malignant brain tumors in adults. Patients with these tumors have a dismal prognosis, with a median survival of less than 1 year.1 Although chemotherapy has been incorporated into the treatment of various intracranial neoplasms, the resistance and invasive nature of these tumors call for new treatment strategies. These include targeting brain tumor cells while sparing normal tissue and the use of gene therapy approaches. For brain tumors, the success of gene therapy depends on the ability to deliver genes to the infiltrating tumor cells interspersed within the normal cells.
Viral vectors have been considered the most effective in vivo gene delivery reagents. Retrovirus, adenovirus (Adv), and herpes simplex virus-1 (HSV-1) are the best-studied viral brain tumor therapy vectors.2 Replication-competent retroviruses have infection rates of 97%, with tumor specificity and lack of spread to nontumor tissue. Potential disadvantages, however, include insertional mutagenesis.3 Adv and HSV-1, used in both replicating and replication-deficient forms, offer numerous advantages, including high transgenic capacity (Adv) and persistent gene expression (HSV-1).4,5 Potential disadvantages, however, include short-term transgene expression and potential toxicity for Adv, and neuro-virulence and recombination with wild type for HSV-1. Oncolytic viruses have also been used successfully for clinical trials,6 because they replicate selectively within tumor cells, leading to increased intratumoral viral titers. To date, however, their clinical effectiveness for brain tumors remains to be substantiated.7,8
Recent evidence suggests that stem cells can be used as delivery vehicles for brain tumor therapy.9,10 In vitro migration assays confirm the ability of isolated stem cells to migrate toward factors produced by glioma cells.11 There is, however, concern that these undifferentiated stem cells can interfere with baseline functions.12 Thus, ideally, differentiated cells such as astrocytes should be considered as vectors. Astrocytes are native to the CNS, which should maximize their survival and function after transplantation. Astrocytes are highly secretory cells and can generate large amounts of proteins.13 Therefore, transplanted astrocytes should be protective within the host brain. Finally, normal astrocytes can migrate along white matter tracts after transplantation into the brain.14,15 This migratory capacity may be useful for delivery of gene therapy to infiltrative tumors such as malignant gliomas.
Autologous astrocytes can be derived from the pluripotent neural precursor cells in the periventricular area or from fetal tissue.16,17 These approaches are hazardous, with limited capability of generating large numbers of cells. Our recent work shows that we can generate a pure population of astrocytes from mouse embryonic stem cells (ESCs).18 Thus, it is our hypothesis that ESC-derived astrocytes conditionally expressing transgenes can be used as an alternative source for gene delivery in malignant glioma therapy.
Tumor necrosis factor (TNF)–related apoptosis-inducing ligand (TRAIL), a member of the TNF family, has been shown to induce apoptosis in a variety of transformed cell lines but to spare most normal cells.19,20 Systemic administration of recombinant TRAIL protein was associated with minimal toxicity in recent clinical trials with cancer patients.21 Recombinant TRAIL, however, does not cross the blood–brain barrier and has a short half-life.22 To obviate these limitations, convection-enhanced delivery (CED) of TRAIL has been used in a malignant glioma xenograft model with good results.23 Although CED is a promising technique to bypass the blood–brain barrier, side effects in the clinical setting include invasive implant of multiple catheters and difficult calculation of drug delivery.24 To circumvent these side effects, alternative delivery methods have been sought. Cellular gene delivery methods are gaining importance as stem cells have shown to have migratory and homing capacity.25 Our recent work shows that we can generate a pure population of transgenic ESC-derived astrocytes with a highly regulated, robust conditional expression of TRAIL under the control of a tetracycline (tet)-inducible promoter.26 Furthermore, we have shown that transgenic delivery of TRAIL by ESC-derived astrocytes induced apoptosis in malignant glioma cells in vitro.26 Based on our recent work, we hypothesize that ESC-derived astrocytes conditionally expressing TRAIL may represent an effective and novel approach for treatment of human malignant gliomas in vivo. Using a heterotopic xenograft model in the mouse, we have found that ESC-derived astrocytes conditionally expressing TRAIL can be used as a vector for gene therapy in malignant gliomas in vivo.
Human malignant glioma A172 cells were used for the in vivo heterotopic malignant glioma tumor xenografts. Cells were cultured in standard condition. In brief, human malignant glioma A172 cell lines were obtained from the Neurosurgery Tissue Bank (San Francisco, CA, USA). Cells were cultured in Iscove’s modified Dulbecco’s medium (Cellgro, CA, USA) supplemented with 10% fetal calf serum at 37ºC in an incubator supplemented with 5% CO2 Mouse ESCs were passaged and maintained according to previously published protocols.18 Mouse Ainv-18 ESCs (a gift of Michael Kyba) constitutively express a doxycycline (Dox)-binding transcriptional activator fusion protein and include a Dox-responsive promoter upstream of a Lox (locus of X-over P1) cloning site.27 Genes cloned into the Lox site using Cyclization Recombination (Cre)-assisted recombination27 are then expressed when the cells are exposed to Dox. ESCs were directed to differentiate into astrocytes using protocols recently developed in our lab.26
The experimental protocol was approved by the Institute of Animal Care and Use Committee of Mount Sinai School of Medicine and met all federal guidelines. A172 cells (3 × 106 cells/200 μl phosphate-buffered saline [PBS]) were injected subcutaneously in the right flank of athymic NCr-nu/nu nude mice (5-week-old females, 30 g body weight; Taconic Farms, Germantown, NY, USA). Visible tumors (~7.5 mm3) occurred in 7–14 days. Intratumoral injections of ESC-derived astrocytes conditionally expressing TRAIL (103 cells/20 μl PBS) were administered to the right flank when the xenograft reached a volume of 7.5 mm3. Prior to intratumoral injection, ESC-derived astrocytes conditionally expressing TRAIL were labeled with the vital red dye PKH2628 using the Cell Linker Kit (Sigma, St. Louis, MO, USA) following the manufacturer’s protocol.
Experimental animals (n = 6 per group) received ESC-derived astrocytes conditionally expressing TRAIL once (group 1) or twice for 2 consecutive days (group 2). These animals were sacrificed 48 h after the last astrocyte injection. Other animals (group 3) underwent daily intratumoral injections of ESC-derived astrocytes conditionally expressing TRAIL for 7 days and were then sacrificed. To study “homing,” group 4 mice received systemic delivery of ESC-derived astrocytes conditionally expressing TRAIL in the tail vein and were sacrificed after 7 days.
In the experimental groups, TRAIL was induced by giving animals free access to water containing Dox (20 μg/ml). Water bottles were wrapped with aluminum foil to protect Dox from light degradation. Control groups (n = 6 in each group) received no treatment (control 1), ESC astrocytes with Dox treatment (control 2), or ESC astrocytes engineered with TRAIL without Dox induction (control 3). Tumor volume (V) was calculated as V = π/6 × largest diameter (D) × (small D)2.29
Gene expression was quantified by reverse transcription (RT)-PCR. Total RNA was extracted and RT-PCR performed as previously described. In brief, tumor tissue was harvested for RT-PCR. Total cellular RNA from the tumor tissue was purified using the RNeasy kit (Qiagen, Valencia, CA, USA), with on-column DNase treatment. All RNA preparations were spectrophotometrically quantified and examined for degradation using gel electrophoresis prior to RT. RT was performed on 1 μg of total RNA using the Omniscript Reverse Transcriptase Kit (Promega, Madison, WI, USA) and PDN6 random hexamer primers (Pharmacia, Piscataway, NJ, USA) in a total volume of 20 μl. A total volume of PCR mixture was 12.5 μl; each contained 0.5 μl cDNA template. All PCR amplifications were carried out for 35 cycles. PCR products were examined on 1.8% agarose gels and photographed using an Eagle Eye II imager (Stratagene, La Jolla, CA, USA).
Forty-eight hours after the last ESC-derived astrocyte injection, animals were anesthetized with intraperitoneal ketamine/xylazine, and subcutaneous tumors were excised and snap-frozen into liquid nitrogen. Tissue blocks were embedded in Tissue Tak and sliced in the coronal plane in 10-μm-thick frozen sections. Serial sections were stained with hematoxylin and eosin, or left unstained for immunofluorescent labeling and terminal dUTP nick end labeling (TUNEL) detection. Slides were washed with PBS, fixed in 4% paraformaldehyde in PBS for 30 min at room temperature, washed, and preincubated in 10% normal goat serum (Sigma) for 30 min at room temperature and then in primary antibody for 1 h at room temperature (monoclonal anti-TRAIL antibody, 1:100; Chemicon International, Temecula, CA, USA). Alexa 488 antimouse antibody (1:400) was used as a secondary antibody. Cells were then washed and mounted in Vectashield with added 4',6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame, CA, USA), and examined under fluorescence microscopy using the Openlab imaging system (Improvision, Lexington, MA, USA). Omission of the primary antibody served as negative control. TUNEL staining was performed using the ApopTag Plus Fluorescein In Situ Apoptosis Detection kit (Chemicon International). Experimental sections were then mounted on slides with Vectashield mounting medium (Vector Laboratories), and images were taken with a Leica microscope (Leica, Bannockburn, IL, USA).
Briefly, tumor tissue was harvested in RIPA buffer (1% Nonidet P-40, 0.5% deoxycholate, 0.1% sodium dodecyl sulfate [SDS] in PBS) containing protease inhibitor cocktail (Protease Arrest™; Calbiochem, San Diego, CA, USA; 50 mM NaF and 1 mM Na3VO4), homogenized, and centrifuged at 12,000 rpm for 10 min at 4°C. The supernatant was used as total cell lysate. Protein lysates (20 μg) were denatured in 2% SDS, 10 mM dithiothreitol, 60 mM Tris-HCl (pH 6.8), and 0.1% bromophenol blue and loaded onto a 15% polyacrylamide/SDS gel (Bio-Rad). The separated proteins were then transferred by electroblot (80 mA, 1 h) onto a nitrocellulose membrane (Bio-Rad). The membrane was blocked for 1 h in room temperature in Tris-buffered saline/Tween 20 (TBS-T; 0.1 M Tris-HCl [pH 7.6], 1.37 M NaCl, 0.1% Tween 20) containing 5% nonfat dry milk and incubated overnight at 4°C in TBS-T containing the primary antibody. The membrane was washed in TBS-T, incubated with the secondary antibody conjugated to horseradish peroxidase for 1 h at room temperature, and then washed in TBS-T. The ECL nonradioactive detection system (Amersham Pharmacia Biotech, Buck-inghamshire, UK) was utilized to detect the antibody-protein complexes by exposure of the membrane to an X-ray film.
The data are expressed as mean ± SD. Statistical analysis was performed by using Student’s t-test. Intergroup analysis was performed with analysis of variance and post hoc Bonferroni test. The criterion for statistical significance was taken as p < 0.05.
To study the in vivo effects of transgene expression, we used ESC-derived astrocytes carrying TRAIL in the context of a “tet-on” system (reverse tet transactivator [rtTA]), upstream of Lox, activated by Dox. A schematic representation of the insertional cassettes is shown in Fig. 1A. ESC-derived astrocytes conditionally expressing TRAIL were injected in malignant glioma xenograft in a mouse subcutaneous model. Transgenic TRAIL expression was detected in the explanted heterotopic xenograft by RT-PCR analysis 48 h and 7 days after Dox induction in all experimental animals (Fig. 1B). TRAIL was not detected in the control groups. Immunocytochemistry for the TRAIL protein showed in situ expression in the experimental animals (Fig. 2A). TRAIL expression was colocalized with the ESC-derived astrocytes conditionally expressing TRAIL labeled with PKH26 dye (Fig. 2A). At higher magnification, membrane localization of TRAIL protein was evident (Fig. 2B). Systemic administration of ESC-derived astrocytes (tail vein, experimental group 4) resulted in “homing” within the tumor nodule (Fig. 2C).
A 14% reduction in tumor size was found 48 h after single injections of ESC-derived astrocytes conditionally expressing TRAIL after Dox induction (experimental group 1) compared with controls: 6.25 ± 0.4 mm3 versus 7.6 ± 1.2 mm3 in controls (p < 0.05). Two injections (experimental group 2) resulted in 31% reduction in tumor size compared with controls: 4.50 ± 0.2 mm3 versus 7.6 ± 1.2 mm3 in controls (p < 0.01). Long-term treated animals (experimental group 3) showed a 40% reduction in tumor volume (Fig. 3A, B). Histology performed at this time point showed severe central necrosis with only scattered residual tumor cells at the periphery (Fig. 3C, D).
Apoptosis was detected by TUNEL in A172 cell xenografts in all experimental animals (Fig. 4A). Apoptosis was not observed in control animals (Fig. 4B). TUNEL-positive cells (apoptotic tumor cells) did not colocalize with ESC-derived astrocytes tracked by PKH26 red dye (Fig. 4C).
To further study the molecular events underlying the observed enhanced apoptosis caused by ESC-derived astrocytes conditionally expressing TRAIL after Dox induction in in vivo human gliomas, we assessed the expression of the death receptor TRAIL-R1 (DR4) by Western blotting. A 2.5-fold increase in DR4 expression was noted in the experimental groups compared with controls (Fig. 5A). Immunohistochemistry showed marked expression of DR4 in malignant glioma xenografts after Dox induction (Fig. 5B) but not in control groups. DR5 expression was not significantly changed (data not shown).
In this study, we investigated the application of in vivo gene delivery using ESC-derived astrocytes conditionally expressing TRAIL. ESC-derived astrocytes conditionally expressing TRAIL previously tested in vitro26 were used to induce apoptosis in vivo in a heterotopic brain tumor xenograft model. The finding of tumor reduction and apoptotic tumor cells after administration of ESC-mediated gene delivery supports the hypothesis that ESC-derived astrocytes should be considered as vectors for gene therapy of malignant astrocytomas.
Only a few agents are truly cancer-cell specific in terms of efficacy and cell death induction. TRAIL is a rare example because it kills cancer cells but not normal cells.19,20 Recent studies have also demonstrated that targeting TNF death receptors is a promising strategy for cancer treatment because it induces powerful apoptosis.30 In this study, we used engineered mouse ESC-derived astrocytes to conditionally express TRAIL after Dox induction, as previously published for in vitro experiments.26 We hypothesized that the ESC-derived astrocytes conditionally expressing TRAIL in vivo would retain tet control. Our results showed that transgenic expression of TRAIL after Dox induction was maintained, as demonstrated by RT-PCR results and in situ immunohistochemistry. These data show that the “tet-on” system used to engineer our ESC-derived astrocytes maintains in vivo transcription Dox inducibility. These results, taken together with previously published data,26 support the safety and feasibility of using the “tet-on” system in vivo and suggest that this system can be considered for clinical trials.
Gene therapy provides promising hope for treatment of malignant gliomas. Encouraging results in vitro and in vivo have stemmed from clinical trials using viral vectors.31,32 However, these have had less therapeutic effects than desired, probably due to low levels of gene transduction.2 Alternatively, CED can be used, but this requires the invasive implantation of catheters and precise calculations of drug delivery.24 Improving gene delivery methods and increasing transgene potency are currently being explored to overcome these limitations.31 Alternatively, the use of genetically modified cells to deliver gene therapy to the CNS may avoid some of the limitations of conventional viral therapy. ESCs have tremendous potential in the field of tissue engineering and regenerative medicine because they have both unlimited proliferative ability and the capacity to produce every type of cell and tissue in the body.33 ESCs can propagate indefinitely but retain the capacity to differentiate into all mature somatic phenotypes when induced by appropriate signals.33 Moreover, with a Dox-responsive promoter, the gene can be kept silent when it is no longer required by omitting administration of Dox.18,26
In our study, we showed a significant decrease in tumor burden 48 h after a single injection of ESC-derived astrocytes conditionally expressing TRAIL. The increased reduction in tumor burden seen after multiple injections is most likely secondary to a better distribution of ESC within the tumor itself. Because TRAIL is a cell-membrane-bound protein,34 cell-to-cell interaction is important to increase its proapoptotic effects. The histological finding of minimal residual tumor cells 7 days after transgene administration supports the concept that transgenic delivery by ESC-derived astrocytes should be considered for malignant gliomas. For a clinical trial, human ESC-derived cells should be used to avoid a potential severe immunogenic response with allotransplant. In our laboratory, we are currently working with a human ESC line to replicate our published work with murine ESC.26,27,34
Results from the in vivo investigations reported here suggest two modes of actions for the transgenic TRAIL. First, the increased apoptotic rate using transgenic TRAIL in vivo seems to be independent of p53 status. The malignant glioma cells used for our xenograft model, A172, have mutant p53.34,35 Second, our data suggest that the increased apoptosis might be linked to an increase in DR4 receptor expression. The mechanisms of upregulation of DR4 receptors after TRAIL exposure in vitro are well documented.36–38 However, the mechanisms of TRAIL-R2 (DR4) upregulation and of apoptosis activation pathways (extrinsic vs. intrinsic) are not yet fully elucidated.39 In our study, we found no significant change in DR5 expression, which has been shown to be a p53-inducible gene.40
In conclusion, as the field of gene delivery for malignant gliomas advances, vectors other than viral might be considered. Our study is the first to report in vivo conditional transgenic delivery using ESC-derived astrocytes for malignant gliomas. Conditionally transgenically expressed TRAIL remained under external control of antibiotic administration (“tet-on” system) and resulted in reduced tumor burden, necrosis, and apoptosis. Based on our results, we suggest that ESC-derived astrocyte-mediated gene delivery should be pursued as a new anti-cancer therapeutic approach for patients with malignant gliomas.
The work presented in this article was funded in part by a Goldhirsh Foundation research award to I.M.G.