Skip to main content
NCBI home page
Neuro-Oncology logoLink to Neuro-Oncology
. 2016 May 18;18(11):1508–1518. doi: 10.1093/neuonc/now097

Systemic AAV9-IFNβ gene delivery treats highly invasive glioblastoma

Dwijit GuhaSarkar 1, Qin Su 1, Guangping Gao 1, Miguel Sena-Esteves 1
PMCID: PMC5063516  PMID: 27194146

Abstract

Background

Complete surgical removal of all glioblastoma (GBM) cells is impossible due to extensive infiltration into brain parenchyma that ultimately leads to tumor recurrence. The current standard of care affords modest improvements in survival. New therapeutic interventions are needed to prevent recurrence. Local AAV-hIFNβ gene delivery to the brain was previously shown to eradicate noninvasive orthotopic U87 tumors in mice. However, widespread CNS gene delivery may be necessary to treat invasive GBMs. Here we investigated the therapeutic effectiveness of systemically infused AAV9-hIFNβ against an invasive orthotopic GBM8 model.

Methods

Mice bearing human GBM8 brain tumors expressing firefly luciferase (Fluc) were treated systemically with different doses of scAAV9-hIFNβ vector. Therapeutic efficacy was assessed by sequential bioluminescence imaging of tumor Fluc activity and animal survival. Brains were analyzed post mortem for the presence and appearance of tumors. Two transcriptionally restricted AAV vectors were used to assess the therapeutic contribution of peripheral hIFNβ.

Results

Systemic infusion of scAAV9-hIFNβ vector induced complete regression of established GBM8 tumors in a dose-dependent manner. The efficacy of this approach was also dependent on the stage of tumor growth at the time of treatment. We also showed that peripherally produced hIFNβ contributed considerably to the therapeutic effect of scAAV9-hIFNβ. A comparative study of systemic and unilateral intracranial delivery of scAAV9-hIFNβ in a bilateral GBM8 tumor model showed the systemic route to be the most effective approach for treating widely dispersed tumors.

Conclusions

Systemic delivery of AAV9-IFNβ is an attractive approach for invasive and multifocal GBM treatment.

Keywords: adeno associated virus (AAV), glioblastoma (GBM), interferon-beta (IFNβ), mouse orthotopic xenograft, systemic AAV gene therapy

Glioblastoma (GBM) is the most common1 and aggressive (WHO grade IV) primary brain tumor.2 These tumors are characterized by aggressive invasiveness, long-distance migration, and neovascularization. The standard-of-care treatment involves surgical resection, followed by radiation and chemotherapy with temozolomide. Complete surgical resection of tumor is almost impossible because of macroscopically indistinguishable, diffuse GBM cell infiltrates far from the main tumor mass. This results in tumor recurrence, which is often radio- or chemotherapy resistant, eventually causing death. Overall median survival of GBM patients post diagnosis is 14.6 months.3 Hence, new therapeutic approaches are needed to prevent tumor recurrence.

Interferon beta (IFNβ), a cytokine secreted from cells upon viral infection, activates several genes downstream through Janus kinases (JAKs) and signal transducers and activators of transcription (STATs).4 Although IFNβ is best known for its antiviral immune modulatory roles, it also has antitumor properties. In addition to direct cytostatic and proapoptotic effects on cancer cells,5,6 IFNβ also has antiangiogenic7,8 and immune-stimulatory9,10 properties. In addition, IFNβ has been shown to potentiate the effect of some chemotherapeutic drugs.11 Previously, a phase 1 clinical trial on newly diagnosed GBM patients showed improved survival in patients treated with IFNβ as an adjuvant to the standard of care treatment with minimal adverse effects.12 However, given the relatively short half-life of IFNβ,13 achieving greater therapeutic benefit is challenging.

Gene therapy approaches may bypass this limitation through continuous expression of INFβ in transduced cells. Moreover, as IFNβ is a secreted protein, transduction of a small number of endogenous cells may be sufficient to achieve a widespread therapeutic effect. Previously, we have shown that local expression of human IFNβ (hIFNβ) from a recombinant adeno-associated viral (AAV) vector successfully inhibits growth of intracranial human U87 tumors through gene delivery to normal cells in the brain.14 U87 brain tumors grow rapidly in the mouse brain, but mostly as large spheroids with distinct borders, and thus do not reproduce the invasiveness and migratory properties of GBM that are responsible for recurrence. To test the effectiveness of AAV-hIFNβ gene therapy in a more realistic model, we carried out studies using GBM8 cells, which upon transplantation into mouse brain produce diffuse tumors with extensive infiltration and long-distance migration.15

A potential approach to counteract the properties of GBM cells that limits the efficacy of local interventions is to generate a disperse network of endogenous cells expressing IFNβ to prevent single tumor cell infiltration and growth (ie, matching the delivery modality to the disease characteristics). Intravascular infusion of AAV9 vectors achieves widespread gene delivery to the CNS of mice16 and large animals.17,18 Therefore, systemic delivery of an AAV9-hIFNβ vector may be effective for generating a widely distributed CNS network to combat the invasive and migratory properties of GBM. Here we tested this concept in an orthotopic xenograft model of invasive human GBM.

Materials and Methods

Cell Culture

GBM8 cells were a kind gift from Dr. Samuel Rabkin (Massachusetts General Hospital, Boston, Massachusetts). These cells were transduced with the lentivirus vector CSCW2-Fluc-IRES-mCherry as described14 to generate GBM8-Fluc cells constitutively expressing firefly luciferase. Cells were grown as neurospheres in neurobasal media (21103-049, Gibco) supplemented with 3 mM L-glutamine (25-005-CI, Mediatech), 1x B27 supplement (17504-044, Gibco), 0.5x N2 supplement (17502-048, Gibco), 2 µg/mL heparin (H3400, Sigma-Aldrich), 1x antibiotic-antimycotic solution (30-004-CI, Mediatech) and 1x amphotericin B (30-003-CF, Mediatech), 20 ng/mL recombinant human bFGF (100-18B, PeproTech), and 20 ng/mL of recombinant human EGF (AF-100-15, PeproTech).

Animals

Six- to eight week-old male athymic nude mice were obtained from the National Cancer Institute for this study. All animal studies were approved by the University of Massachusetts Medical School Institutional Animal Care and Use Committee following guidelines set forth by the NIH's Guide for the Care and Use of Laboratory Animals (https://grants.nih.gov/grants/olaw/Guide-for-the-Care-and-use-of-laboratory-animals.pdf).19

Orthotopic Xenografting

Two days prior to implantation into the mice, the medium of GBM8-Fluc cells was replaced with fresh medium. On the day of injection, GBM8-Fluc cells were dissociated into a single cell suspension by pipetting. Cells were washed twice in Dulbecco′s phosphate-buffered saline (PBS; 14190-250, Gibco) and resuspended in the same to a concentration of 50 000 cells/µL. One µL of cell suspension was injected stereotaxically into the left striatum. The stereotaxic coordinates for tumor implantation from bregma were (in mm): AP: +0.5, ML: 2.0 (left) and DV from brain surface: −2.5. Bilateral tumors were generated by injecting 50 000 GBM8-Fluc cells into both striata.

AAV Vector Design, Production, and Delivery

All recombinant AAV9s used in the study were self-complementary (sc) vectors. scAAV9/CB-hIFNβ and scAAV9/CB-hIFNβ -miRBS-1-122 vectors encode human interferon-β under the chicken β-actin promoter and cytomegalovirus enhancer (CB promoter) and carry a rabbit beta-globin polyadenylation (RBGpA) signal. The scAAV9/CB-hIFNβ-miRBS-1-122 vector carries 3 copies of miR-1 and miR-122 binding sites (miRBS) in the 3′untranslated region as described.20 The scAAV9/TBG-hIFNβ vector carries a thyroxin-binding globulin (TBG) promoter to drive liver-specific gene expression.21 The scAAV9/CB-EGFP and scAAV9/TBG-EGFP vectors encode enhanced green fluorescence protein (EGFP).

AAV9 vectors were produced at the University of Massachusetts Medical School Gene Therapy Center Viral Vector Core as described.22 Vector titers were determined by quantitative PCR (qPCR) of vector genomes using the following primers and probe specific for RBGpA (Eurofins):

Primer1: 5′-GCCAAAAATTATGGGGACAT-3′;

Primer2: 5′-ATTCCAACACACTATTGCAATG-3′;

Probe: 6FAM-ATGAAGCCCCTTGAGCATCTGACTTCT-TAMRA

For systemic administration, AAV9 vectors were injected via the tail vein in a total volume of 200 µL in PBS. In the intracranial treatment paradigm, 7.6 × 109 genome copies (gc) of scAAV9/CB-hIFNβ vector were infused in 2 µL at 200 nL/min in the same stereotaxic coordinates used for tumor implantation. For control groups, an equal volume of PBS was injected into the mice for all of the experiments.

Live Bioluminescence Imaging

Imaging of tumor-associated bioluminescence signal (TABS) was performed using the Xenogen IVIS 100 imaging system (PerkinElmer) 3 minutes after intraperitoneal administration of D-luciferin (4.5 mg). Image analysis was performed using Living Image software (PerkinElmer).

Preparation of Tissue DNA and RNA and Quantification of Vector Genomes and hIFNβ Transcripts

Total DNA was extracted using the DNeasy Blood & Tissue kit (QIAGEN). DNA was diluted to a final concentration of 50–100 ng/µLfor vector genome quantification by qPCR using RBGpolyA specific primers and probe.

Tissue RNA was isolated using TRIzol (15596-018, Invitrogen) and Direct-zol RNA MiniPrep (R2052, Zymo Research Corporation). RNA was treated with TURBO DNase (AM1907, Ambion) for 30 minutes at 37°C prior to reverse transcription using High Capacity RNA-to-cDNA kit (4387406, Applied Biosystems). Quantitative PCR was performed with the following primers and probe for hIFNβ (IDT): Primer-1: 5′-GCAATTGAATGGGAGGCTTG-3′; Primer-2: 5′-TCATAGATGGTCAATGCGGC-3′; Probe: 5′-/6-FAM/TGTCAAAGT/ZEN/TCATCCTGTCCTTGAGGC/3IABkFQ. Mouse HPRT1 expression was used as an internal reference gene to normalize all values (Assay ID: Mm00446968_m1; Applied Biosystems). Mean expression values for PBS group animals were considered to be background noise and thus were subtracted from all values.

Quantification of Human IFNβ in Cell-conditioned Media and Mouse Plasma

hIFNβ in conditioned growth medium and mouse plasma was measured using ELISA assays (41 410 and 41 415, respectively; PBL Assay Science). Conditioned growth media were collected for ELISA 48 hours after transfection. Mouse plasma was collected 1 month after AAV infusion. Media and plasma were diluted 1:10 and 1:4,000 in PBS, respectively, for ELISA measurements.

Histological Analysis

Brain cryosections (20 µm) and paraffin sections (5 µm) were used for histological studies. The primary antibodies used for immunohistochemistry were rabbit monoclonal anti-GFP (1:1000, G10362, Invitrogen) and rabbit monoclonal anti-Olig2 (1:250, ab109186, Abcam). Biotinylated goat anti-rabbit IgG (1:1000, BA-1000, Vector Laboratories) was used as secondary antibody. VECTASTAIN Elite ABC Kit (PK-6100, Vector Laboratories) and DAB substrate kit (SK-4100, Vector Laboratories) were used for immunohistochemical detection and Mayer′s hematoxylin (Sigma) as counterstain. Cryosections (20 µm) were used to detect human GBM8-Fluc cells in the brain by immunofluorescence using mouse anti-human nuclei (1:150, MAB4383, EMD Millipore) and detection with Alexa-fluor 594 goat anti-mouse (1: 2000, A-11020, Invitrogen). Sections were also stained with 4′,6-diamidino-2-phenylindole (DAPI) (PI-62247, Thermo Fisher Scientific).

Graphs and Statistical Analysis

Kaplan-Meier survival curves were plotted in Prism 6 (GraphPad Software), and the Mantel-Cox log-rank test was used for statistical analysis. Bar graphs were also plotted using the same software. And an unpaired 2-tailed t test was used for statistical analysis. Calculated P values were defined as the probability of null hypothesis being true; *P < .05; **P < .01; ***P < .001; ****P < .0001; ns, not significant (P > .05).

Results

Dose-dependent Therapeutic Response With Systemic Delivery of scAAV9/CB-hIFNβ

To assess the therapeutic effectiveness of systemically delivered scAAV9/CB-hIFNβ, athymic nude mice were treated with different vector doses (1 × 1011, 3 × 1011, and 1 × 1012 gc) or PBS 2 weeks after GBM8-Fluc tumor implantation. We measured TABS by live bioluminescence imaging to gain some insight into changes in tumor growth kinetics in the mouse brain after AAV treatment. In PBS-injected control mice, TABS increased exponentially until the animals reached the humane endpoint defined by >15% loss in maximum body weight. In the 2 top doses of scAAV9/CB-hIFNβ, TABS decreased over time and was indistinguishable from baseline (week-1 signal) by 2 weeks after treatment (4 weeks after tumor implantation). This remained unchanged over time (Fig. 1A and B). Treatment with 1 × 1011 gc showed partial response. TABS decreased to baseline levels in 2 mice while increasing over time in 3 mice, albeit with different kinetics than in the PBS control group (Fig. 1B).

Fig. 1.

Fig. 1.

Open in a new tab

Systemic delivery of scAAV9/CB-hIFNβ produces dose-dependent therapeutic response in an orthotopic GBM8 xenograft mouse model. (A) Bioluminescence imaging of tumor burden in representative mice from each group for weeks 2–6 after tumor implantation. Regions of interest (ROI) used for signal quantification are shown as red circles. (B) Tumor-associated bioluminescence signal (TABS) was assessed weekly up to 16 weeks and represented as fold change over signal at 1 week after tumor injection. Data are shown as mean ± SD. Black arrow on the x-axis indicates time of treatment. Treatment groups are represented as follows: 1 × 1012 gc (red), 3 × 1011 gc (blue), 1 × 1011 gc (green), and phosphate-buffered saline (PBS) control group (black); “s1” and “s2,” nonrespondent and respondent subgroups, respectively. For the 1 × 1011 gc group, the TABS curves for complete (dark green) and partial (light green) responders are plotted separately. (C) Kaplan-Meier survival curves for different treatment groups show dose-dependent survival benefit. Survival curves are identified as follows: 1 × 1012 gc (red), 3 × 1011 gc (blue), 1 × 1011 gc (green), and PBS control group (black); log-rank P values: ****P < .0001; ***P < .001. (D) Representative images of hematoxylin-and-eosin staining of mouse brain sections at the endpoints. Top row depicts representative brain sections from different dose groups. Bottom row shows the magnified views of the corresponding areas indicated as “a,” “b,” “c,” and “d” in the top row. Black arrows point to tumorlets with distinct borders. The brains were collected at the humane endpoints on day 44 (PBS control) and day-111 (1 × 1011 gc), or at the study endpoint on day-244 (3 × 1011 gc and 1 × 1012 gc). Scale bars represent 100 µm. Abbreviation: n, number of animals in the group.

The scAAV9/CB-hIFNβ dose response was also reflected in long-term survival with 100%, 88.89%, and 20% of animals alive at 244 days post tumor implantation in 1 × 1012, 3 × 1011, and 1 × 1011 gc dose groups, respectively (Fig. 1C). All animals from the PBS control group reached the humane endpoint between 42 and 49 days with a median survival of 46 days after tumor implantation.

Histological analysis of the brains from animals that survived to 244 days revealed normal histology with no microscopic evidence of tumors (Fig. 1D, Supplementary Fig. S1). Interestingly, in animals that succumbed to disease progression in the 1 × 1011 gc group, there were large tumors with sharply defined borders to normal tissue that were composed of numerous well-defined tumorlets (arrows in Fig. 1D). This is in contrast to the diffuse nature of GBM8 tumors in mouse brain found in the PBS control group (Fig. 1D).

Therapeutic Outcome Is Dependent on Tumor Growth Phase at the Time of Treatment

To assess the effect of tumor growth phase on the therapeutic efficacy 3 × 1011 gc (minimum effective dose) of scAAV9/CB-hIFNβ was infused at 2, 3, 3.5 (∼84 hours after week-3 treatment), and 4 weeks after GBM8-Fluc tumor implantation. Treatment at week 2 and week 3 prevented tumor growth as suggested by TABS assessment over time (Fig. 2A). More importantly, this ultimately resulted in long-term survival of 100% and 88.9% of mice to 244 days after tumor implantation (Fig. 2B). Treatment at 3.5 and 4 weeks had a modest impact on TABS increase over time (Fig. 2A). Nonetheless, it resulted in significant increases in median survival to 60 and 61 days, respectively, compared with 46 days for PBS control mice. The maximum survivals were increased to 150 and 80 days, respectively, compared with 54 days for PBS controls (Fig. 2B).

Fig. 2.

Fig. 2.

Open in a new tab

Treatment by systemic delivery of scAAV9/CB-hIFNβ before the onset of rapid tumor growth is critical for long-term survival benefit. (A) Tumor-associated bioluminescence signal (TABS) is represented as fold change over signal at 1 week after tumor implantation. The timing of treatment (with 3 × 1011 gc scAAV9/CB-IFNβ) after tumor implantation is identified as follows: second week=red, third week=purple, three-and-a-half weeks=orange; fourth week=green (treated), and PBS control group=black (injected at second week). Data are shown as mean ± SD. Arrows of corresponding colors indicate the times of treatment for each group. (B) Kaplan-Meier survival curves showing effect of time of treatment on survival. Survival curves are shown for the same groups (with the same color coding) as for TABS change over time. Log-rank P values: ****P < .0001; ***P < .001. (C) Images of TABS in representative mice for each treatment time-point. Red circles are the regions of interest (ROI) for signal quantification. (D) Quantification TABS from imaging of live untreated GBM8-Fluc-bearing mice at treatment time points after tumor implantation. Each data point on the plot is the TABS value from one mouse. Values are shown with group mean (black horizontal bar) and range (bars with different colors). Unpaired 2-tailed t test results: ** P < .01. (E) Representative images of untreated GBM8-Fluc-bearing mouse brains immunostained for human nuclei at treatment time points. Abbreviations: n, number of animals in the group; ns, not significant; Str, left striatum (where tumor was implanted); W, weeks after tumor implantation.

Analysis of tumor burden by TABS at the treatment time points revealed that tumors remained largely unchanged between weeks 2 and 3 but increased thereafter (Fig. 2C and D). Immunostaining of tumor sections for human nuclear antigen and GBM-specific OLIG2 marker at the time of treatment corroborated this observation (Fig. 2E, Supplementary Fig S2). These results suggest a correlation between therapeutic efficacy of scAAV9/CB-hIFNβ and tumor growth phase.

Systemically Delivered AAV9 Primarily Transduces Astrocytes and Endothelial Cells in the Glioblastoma-bearing Mouse Brain

Tumor-bearing mice were injected systemically with 3 × 1011 gc of scAAV9/CB-EGFP vector to determine the cell types transduced and thus were likely mediators of the therapeutic effect. Based on morphology (Fig. 3A and B), it appeared that most EGFP-positive cells were astrocytes and endothelial cells found primarily in cortex and periventricular regions (Fig. 3B), with fewer transduced cells in the striatum or tumor. These results suggest that, at the intermediate dose, the therapeutic effect of scAAV9/CB-hIFNβ is very likely achieved through expression of hIFNβ in normal astrocytes and vascular endothelia.

Fig. 3.

Fig. 3.

Open in a new tab

Astrocytes and vascular endothelial cells are the predominant brain cells transduced upon vascular delivery of scAAV9-EGFP vector in glioblastoma (GBM) tumor-bearing mice (A) Representative pictures of immunohistochemical staining for EGFP-expressing cells in brain sections from GBM8-Fluc-bearing mice 3 weeks after tail vein delivery of 3 × 1011 gc/mouse of scAAV9/CB-EGFP and scAAV9/TBG-EGFP. Control mice were infused with PBS. 3,3′-diaminobenzidine (DAB) was used for detection of EGFP-positive cells (brown) in the brain and hematoxylin was used for counterstaining. (B) High magnification pictures of 4 different brain areas (identified in A by boxed small case letters) from mice injected with scAAV9/CB-EGFP vector. Large arrows in Ba and Bc indicate EGFP-expressing cells with astrocyte morphology. Small arrows in Bb and Bd indicate EGFP-positive cells with vasculature morphology. Images from one animal representative of each group (n = 3).

Peripherally produced hIFNβ contributes to the therapeutic effect of systemically delivered scAAV9/CB-IFNβ

Because AAV9 transduces peripheral organs at high efficiency after vascular delivery20,23 and the ubiquitous nature of the CB promoter used in scAAV9/CB-INFβ, it is possible that peripherally expressed hIFNβ contributes to the therapeutic effect against intracranial GBM8-Fluc tumors. Two AAV vectors were used to investigate this contribution: (1) scAAV9/CB-hIFNβ-miRBS-1-122 (de-targeted) vector where CB-driven transgene expression was de-targeted from liver, muscle, and heart by insertion of miR-1 and miR-122 binding sites into the transgene cassette as described20 and (2) scAAV9/TBG-hIFNβ (TBG) vector where transgene expression was driven by the liver-specific thyroxin binding globulin (TBG) promoter.21 The therapeutic efficacy of these AAV9 vectors was compared with scAAV9/CB-hIFNβ (CB vector) in GBM8-Fluc tumor-bearing mice treated by systemic delivery. The miR-1 and miR-122 regulation of AAV-CB-hIFNβ-miRBS-1-122 transgene expression was validated first in cell culture (Supplementary Fig. S3). Accordingly, plasma concentration of IFNβ was significantly lower in mice treated with scAAV9/CB-hIFNβ-miRBS-1-122 compared with scAAV9/CB-hIFNβ (Fig. 4A). Also, as expected, vascular infusion of 3 × 1011 gc of scAAV9/TBG-EGFP vector did not result in appreciable EGFP expression in brain (Fig. 3A). Both bioluminescence imaging and survival (Fig. 4B and C) showed that treatment with de-targeted scAAV9/CB-hIFNβ-miRBS-1-122 vector was less effective than scAAV9/CB-hIFNβ as it required 1 × 1012 gc/mouse to achieve comparable suppression of tumor growth and 100% long-term survival. Vector genome copy analysis in brain, liver, and skeletal muscle showed the 2 vectors to have comparable biodistribution/transduction profiles (Supplementary Fig. S2). Therefore the ∼3-fold lower therapeutic efficacy of de-targeted vector could be due to differences in peripheral tissue expression of hIFNβ or a difference in brain expression because of nonspecific miRNA interaction or both. Quantitative RT-PCR analysis of hIFNβ mRNA levels in brain showed comparable levels in mice treated with scAAV9/CB-hIFNβ and scAAV9/CB-hIFNβ-miRBS-1-122 vectors (Fig. 4D). However hIFNβ mRNA levels were significantly reduced in liver and skeletal muscle of mice treated with scAAV9/CB-hIFNβ-miRBS-1-122 compared with those infused with the same dose of scAAV9/CB-hIFNβ (Fig. 4E and F). These results suggest that peripheral hIFNβ expression contributed substantially to the therapeutic efficacy of scAAV9/CB-hIFNβ. This conclusion is supported by the therapeutic results with liver-specific scAAV9/TBG-IFNβ in which both doses tested changed the kinetics of GBM8-Fluc tumor growth (Fig. 4B) and improved survival significantly (Fig. 4C), albeit less effectively than scAAV9/CB-IFNβ. The biodistribution of scAAV9/TBG-IFNβ was comparable to the other vectors (Supplementary Fig. S4), but increased hIFNβ mRNA levels were found primarily in liver (Fig. 4D and F). Nonetheless, there was detectable hIFNβ mRNA present in the brains of mice treated with 1 × 1012 gc of scAAV9/TBG-IFNβ, albeit at >2-log lower levels compared with the other 2 vectors (Fig. 4D).

Fig. 4.

Fig. 4.

Open in a new tab

Peripheral expression of hIFNβ contributes to the therapeutic effect of systemically infused scAAV9/CB-hIFNβ in the GBM8 mouse model. (A) scAAV9/CB-hIFNβ (CB) or scAAV9/CB-hIFNβ-miRBS-1-122 (de-targeted) vectors were infused at the doses shown or phosphate-buffered saline (PBS) at 2 weeks after tumor implantation and plasma hIFNβ levels measured 1 month later by ELISA (n = 4 per group). Columns represent mean values, and error bars indicate SD. ****P < .0001 in unpaired 2-tailed t test. (B) Tumor-associated bioluminescence signal (TABS) kinetics are represented as fold change over signal at 1 week after tumor implantation for all treatment groups. CB vector, de-targeted vector, or scAAV9/TBG-hIFNβ (TBG vector) were injected 2 weeks after tumor implantation at doses shown. Data are shown as mean ± SD. (C) Kaplan-Meier survival curves for different treatment groups. In log-rank test, **P < .01; ns, not significant (P > .05). (D–F) Comparison of hIFNβ mRNA levels in brain, liver, and skeletal muscle in mice from different treatment groups at the humane or study endpoints by RT-qPCR (n = 3 per group). Top panels show the comparative delta CT values, and bottom panels show relative expression. Data are shown as mean ± SD. In unpaired 2 tailed t test, *P < .05; **P < .01; ***P < .001; Abbreviation: ns, not significant (P > .05).

Systemic Delivery of scAAV9-hIFNβ Is More Effective Than Local Administration for Treatment of Multifocal GBM

We have previously shown that intracranial injection of AAVrh8-IFNβ vectors leads to complete regression of established nonmigratory orthotopic human U87 brain tumors.14 Here we sought to compare the effectiveness of local versus systemic delivery of scAAV9/CB-hIFNβ in mice with bilateral GBM8-Fluc tumors to mimic multifocal recurrence. Mice were treated either by unilateral intracranial injection or systemic administration of scAAV9/CB-IFNβ.

Bioluminescence imaging showed (Fig. 5A and B) that tumors grew uninhibited on both sides of the brain in the untreated control group (Fig. 5A). The average right-to-left (R:L) TABS ratio for this group remained at 1.05 ± 0.11 over time (Fig. 5B). In the group treated with unilateral intracranial injection, TABS decreased over time in the left (treated) hemisphere but increased rapidly in the untreated right hemisphere (Fig. 5A). This was reflected in the increase in the right-to-left TABS ratio over time (Fig. 5B). In contrast, TABS decreased over time in both hemispheres and eventually became undetectable in mice treated systemically (Fig. 5A), and the average right-to-left TABS ratio remained unchanged at 1.04 ± 0.11. The impact on survival was consistent with the tumor-growth imaging data. The median survival of mice treated by unilateral intracranial injection of scAAV9/CB-IFNβ was 56 days compared with 43 days for untreated animals, but nonetheless all animals in this cohort succumbed to tumor growth. In contrast, all animals in the systemic treatment group survived until the 8-month experimental endpoint (Fig. 5C).

Fig. 5.

Fig. 5.

Open in a new tab

Systemic delivery of scAAV9-hIFNβ vector is more effective than intraparenchymal administration for treatment of multifocal glioblastoma. (A) Images of tumor-associated bioluminescence signal (TABS) over time (weeks 2–6) in representative mice with bilateral GBM8-Fluc tumors treated by systemic (3 × 1011gc) or unilateral (left tumor) intracranial (7.6 × 109gc) administration of scAAV9/CB-hIFNβ vector. Separate regions of interest (ROIs; red boxes) over the left and right side of the head were used to quantify TABS for each bilateral tumor. (B) Change in right-to-left hemisphere (R:/L) TABS ratio from week 1 to week 6 after tumor implantation in mice treated by systemic (green line) or unilateral intracranial injection (red line), and untreated controls (black line). All values are normalized to the week-1 signal (n = 5 mice per group) (C) Kaplan-Meier survival curves for mice treated by systemic (green line) or unilateral (left tumor) intracranial (red line) delivery of scAAV9/CB-hIFNβ vector and untreated control group (black line) (n = 5 mice per group). (D) Representative pictures of brain sections stained with hematoxylin and eosin from different groups at the humane (untreated: day 43 and unilateral intracranial treatment: day 56) or study endpoint (day 244). On the left, most panel whole brain sections are shown at low magnification. On the right 2 panels, high magnification images are shown of the corresponding areas indicated in the whole brain sections by L1, L2, or L3 and R1, R2, or R3. Scale bars represent 100 µm.

Histological analysis of brains at endpoint (Fig. 5D) showed no evidence of residual tumors in mice treated systemically. In mice treated with unilateral intracranial injection, there were large tumor masses in the right (untreated) hemisphere, but these had sharply defined tumor-brain parenchyma borders unlike the untreated tumors. This indicates that scAAV9/CB-IFNβ injection in the left hemisphere effectively prevented ipsilateral tumor growth, but it was insufficient to prevent distal tumor growth in the contralateral hemisphere. Nonetheless it appears that local treatment changed the phenotype of the untreated tumor in the contralateral hemisphere.

Finally, the presence of remnant GBM8-FLuc tumor cells was assessed in brain sections by immunostaining for oligodendrocyte transcription factor-2 (OLIG2) as this is a protein highly expressed in glioblastoma stem cells.24,25 No OLIG2-positive cells were apparent in the brains of systemically treated mice, while the OLIG2 immunostaining patterns in the other 2 groups were consistent with the histological findings (Supplementary Fig. S5).

The present study shows that systemic scAAV9/CB-hIFNβ gene therapy is an effective approach to overcoming the invasiveness and long-distance migratory properties of glioblastoma.

Discussion

Our group has previously shown that local AAV-IFNβ gene therapy is an effective approach for treating orthotopic human U87 GBM tumors in mice.14 However, in patients, unlike the noninvasive U87 model, GBM cells migrate extensively in the brain. Thus, reaching widely distributed cells may be challenging using intraparenchymal injections of AAV vectors even with convection-enhanced delivery. Here we have successfully demonstrated that systemic delivery of AAV9-IFNβ can engineer brain cells to form a global tumor inhibitory network, which can effectively treat an invasive GBM model. Besides, while our earlier study14 showed prevention of new tumor establishment in the contralateral hemisphere in a pretreatment model, it was unsuccessful for treating established contralateral tumors. In the current study, we show for the first time that systemic delivery can be therapeutically effective for such distant or multifocal tumors. The limitation of intraparenchymal infusion of AAV vectors for treating widespread GBM tumors was also apparent here, albeit in the context of an artificial model of concurrent bilateral tumors. These results are consistent with a recent study in which intracranial injection of an AAVrh8-sTRAIL vector extended survival of mice with GBM8 tumors, but all animals ultimately succumbed to tumor growth in <100 days.26

Interestingly, although distal intracranial injection was insufficient to achieve regression of established GBM8 tumors, it modified the histological phenotype of the invasive GBM to one of a noninvasive tumor (Fig. 5D and E). This is evident from the distinct border noticeable between tumor and nontumor tissues in contrast to the diffuse pattern seen in the untreated control group. A similar effect on tumor phenotype was documented in mice treated systemically with the low dose (1 × 1011 gc) of scAAV9/CB-hIFNβ (Fig. 1D). Presently the mechanism of low-level IFNβ-induced switch from a migratory to a local rapidly dividing phenotype is unknown, but further studies may reveal new targets in the molecular networks that regulate GBM migration and invasion.27–29

A previous study reported some therapeutic effect of a systemically delivered AAV8 vector encoding IFNβ in U87 tumors grown in the flank of mice;30 however, flank GBM xenografts do not reproduce the challenges of brain gene delivery systemically or otherwise. To our knowledge, this is the first study showing that systemic delivery of AAV9-IFNβ induces complete regression of tumors in a highly invasive orthotopic glioblastoma model.

We have used an immune-compromised athymic mouse model, which lacks T cells, for this study. Moreover, as human IFNβ does not interact with the mouse IFNβ receptors,31 it is unlikely that mouse innate immune cells (such as dendritic cells, macrophages, or NK cells) would be induced by human IFNβ as was also suggested in a previous study.32 Therefore, the therapeutic effect documented here is likely due to a direct effect of IFNβ on GBM8 tumor cells. In an immune-competent organism, the immune stimulatory effect4,33,34 of species-matched IFNβ would likely potentiate its antitumor role further. In addition, one possibility is that IFNβ gene therapy could be used as a concomitant therapy with temozolomide because it may help to increase the sensitivity of chemo-resistant tumor cells as shown in anaplastic astrocytoma patients35 and glioma stem cells.36 IFNβ gene therapy may also enhance the effect of other therapeutic approaches such as antibodies targeting oncogenic receptors, similar to what was shown in a recent study with an antibody-IFNβ conjugate.10

In this study, we observed that gene transfer efficiency in the tumor or surrounding striatum was noticeably low compared with cortex or periventricular region. Nonetheless, it appears that expression of IFNβ from normal cells in the brain, and to some extent from peripheral tissues, contributed to the therapeutic benefit. The therapeutic effect documented in this study lends support to the notion of exploring systemic AAV gene therapy encoding secretory proteins with antitumor properties (eg, interferon alpha, tumor necrosis factor alpha, TNF-related apoptosis-inducing ligand) given that widespread direct tumor cell transduction may be challenging.

Supplementary Material

Supplementary material is available online at Neuro-Oncology (http://neuro-oncology.oxfordjournals.org/).

Funding

This work was supported in part by grant R01NS066310 (M.S-E.) from the NIH.

Conflict of interest statement. G.G. is a founder of Voyager Therapeutics and holds equity in the company. G.G. is an inventor on patents licensed to Voyager Therapeutics.

Supplementary Material

Supplementary Data
supp_18_11_1508__index.html (1.2KB, html)

References

  • 1.Preusser M, de Ribaupierre S, Wohrer A et al. Current concepts and management of glioblastoma. Ann Neurol. 2011;70(1):9–21. [DOI] [PubMed] [Google Scholar]
  • 2.Louis DN, Ohgaki H, Wiestler OD et al. The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol. 2007;114(2):97–109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Stupp R, Hegi ME, Mason WP et al. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. Lancet Oncol. 2009;10(5):459–466. [DOI] [PubMed] [Google Scholar]
  • 4.Schneider WM, Chevillotte MD, Rice CM. Interferon-stimulated genes: a complex web of host defenses. Annu Rev Immunol. 2014;32:513–545. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Kaynor C, Xin M, Wakefield J, Barsoum J, Qin XQ. Direct evidence that IFN-beta functions as a tumor-suppressor protein. J Interferon Cytokine Res. 2002;22(11):1089–1098. [DOI] [PubMed] [Google Scholar]
  • 6.Juang SH, Wei SJ, Hung YM et al. IFN-beta induces caspase-mediated apoptosis by disrupting mitochondria in human advanced stage colon cancer cell lines. J Interferon Cytokine Res. 2004;24(4):231–243. [DOI] [PubMed] [Google Scholar]
  • 7.Streck CJ, Ng CY, Zhang Y, Zhou J, Nathwani AC, Davidoff AM. Interferon-mediated anti-angiogenic therapy for neuroblastoma. Cancer Lett. 2005;228(1–2):163–170. [DOI] [PubMed] [Google Scholar]
  • 8.Xiao HB, Zhou WY, Chen XF et al. Interferon-beta efficiently inhibited endothelial progenitor cell-induced tumor angiogenesis. Gene Ther. 2012;19(10):1030–1034. [DOI] [PubMed] [Google Scholar]
  • 9.Wolpert F, Happold C, Reifenberger G et al. Interferon-beta modulates the innate immune response against glioblastoma initiating cells. PLoS One. 2015;10(10):e0139603. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Yang X, Zhang X, Fu ML et al. Targeting the tumor microenvironment with interferon-beta bridges innate and adaptive immune responses. Cancer Cell. 2014;25(1):37–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Kase S, Kubota T, Watanabe M et al. Interferon beta increases antitumor activity of 5-fluorouracil against human colon carcinoma cells in vitro and in vivo. Anticancer Res. 1993;13(2):369–373. [PubMed] [Google Scholar]
  • 12.Motomura K, Natsume A, Kishida Y et al. Benefits of interferon-beta and temozolomide combination therapy for newly diagnosed primary glioblastoma with the unmethylated MGMT promoter: A multicenter study. Cancer. 2011;117(8):1721–1730. [DOI] [PubMed] [Google Scholar]
  • 13.Buchwalder PA, Buclin T, Trinchard I, Munafo A, Biollaz J. Pharmacokinetics and pharmacodynamics of IFN-beta 1a in healthy volunteers. J Interferon Cytokine Res. 2000;20(10):857–866. [DOI] [PubMed] [Google Scholar]
  • 14.Maguire CA, Meijer DH, LeRoy SG et al. Preventing growth of brain tumors by creating a zone of resistance. Mol Ther. 2008;16(10):1695–1702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Wakimoto H, Kesari S, Farrell CJ et al. Human glioblastoma-derived cancer stem cells: establishment of invasive glioma models and treatment with oncolytic herpes simplex virus vectors. Cancer Res. 2009;69(8):3472–3481. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Foust KD, Nurre E, Montgomery CL, Hernandez A, Chan CM, Kaspar BK. Intravascular AAV9 preferentially targets neonatal neurons and adult astrocytes. Nat Biotechnol. 2009;27(1):59–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Duque S, Joussemet B, Riviere C et al. Intravenous administration of self-complementary AAV9 enables transgene delivery to adult motor neurons. Mol Ther. 2009;17(7):1187–1196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Bevan AK, Duque S, Foust KD et al. Systemic gene delivery in large species for targeting spinal cord, brain, and peripheral tissues for pediatric disorders. Mol Ther. 2011;19(11):1971–1980. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Committee for the Update of the Guide for the Care and Use of Laboratory Animals. Guide for the Care and Use of Laboratory Animals. 8th ed. Washington, DC: National Academies Press; 2011. [Google Scholar]
  • 20.Xie J, Xie Q, Zhang H et al. MicroRNA-regulated, systemically delivered rAAV9: a step closer to CNS-restricted transgene expression. Mol Ther. 2011;19(3):526–535. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Yan Z, Yan H, Ou H. Human thyroxine binding globulin (TBG) promoter directs efficient and sustaining transgene expression in liver-specific pattern. Gene. 2012;506(2):289–294. [DOI] [PubMed] [Google Scholar]
  • 22.Ayuso E, Mingozzi F, Montane J et al. High AAV vector purity results in serotype- and tissue-independent enhancement of transduction efficiency. Gene Ther. 2010;17(4):503–510. [DOI] [PubMed] [Google Scholar]
  • 23.Chen BD, He CH, Chen XC et al. Targeting transgene to the heart and liver with AAV9 by different promoters. Clin Exp Pharmacol Physiol. 2015;42(10):1108–1117. [DOI] [PubMed] [Google Scholar]
  • 24.Ligon KL, Alberta JA, Kho AT et al. The oligodendroglial lineage marker OLIG2 is universally expressed in diffuse gliomas. J Neuropathol Exp Neurol. 2004;63(5):499–509. [DOI] [PubMed] [Google Scholar]
  • 25.Trepant AL, Bouchart C, Rorive S et al. Identification of OLIG2 as the most specific glioblastoma stem cell marker starting from comparative analysis of data from similar DNA chip microarray platforms. Tumour Biol. 2015;36(3):1943–1953. [DOI] [PubMed] [Google Scholar]
  • 26.Crommentuijn MH, Maguire CA, Niers JM et al. Intracranial AAV-sTRAIL combined with lanatoside C prolongs survival in an orthotopic xenograft mouse model of invasive glioblastoma. Mol Oncol. 2016;10(4):625–634. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Godlewski J, Bronisz A, Nowicki MO, Chiocca EA, Lawler S. microRNA-451: A conditional switch controlling glioma cell proliferation and migration. Cell Cycle. 2010;9(14):2742–2748. [PubMed] [Google Scholar]
  • 28.Williams SP, Nowicki MO, Liu F et al. Indirubins decrease glioma invasion by blocking migratory phenotypes in both the tumor and stromal endothelial cell compartments. Cancer Res. 2011;71(16):5374–5380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Nevo I, Woolard K, Cam M et al. Identification of molecular pathways facilitating glioma cell invasion in situ. PLoS One. 2014;9(11):e111783. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Streck CJ, Dickson PV, Ng CY et al. Antitumor efficacy of AAV-mediated systemic delivery of interferon-beta. Cancer Gene Ther. 2006;13(1):99–106. [DOI] [PubMed] [Google Scholar]
  • 31.Harari D, Abramovich R, Zozulya A et al. Bridging the species divide: transgenic mice humanized for type-I interferon response. PLoS One. 2014;9(1):e84259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Qin XQ, Beckham C, Brown JL, Lukashev M, Barsoum J. Human and mouse IFN-beta gene therapy exhibits different anti-tumor mechanisms in mouse models. Mol Ther. 2001;4(4):356–364. [DOI] [PubMed] [Google Scholar]
  • 33.Goodbourn S, Didcock L, Randall RE. Interferons: cell signalling, immune modulation, antiviral response and virus countermeasures. J Gen Virol. 2000;81(Pt 10):2341–2364. [DOI] [PubMed] [Google Scholar]
  • 34.Stark GR, Kerr IM, Williams BR, Silverman RH, Schreiber RD. How cells respond to interferons. Annu Rev Biochem. 1998;67:227–264. [DOI] [PubMed] [Google Scholar]
  • 35.Fujimaki T, Ishii H, Matsuno A, Arai H, Nakagomi T. Effectiveness of interferon-beta and temozolomide combination therapy against temozolomide-refractory recurrent anaplastic astrocytoma. World J Surg Oncol. 2007;5:89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Shen D, Guo CC, Wang J et al. Interferon-alpha/beta enhances temozolomide activity against MGMT-positive glioma stem-like cells. Oncol Rep. 2015;34(5):2715–2721. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_18_11_1508__index.html (1.2KB, html)
supp_now097_now097supp.docx (17.7KB, docx)
supp_now097_now097supp_fig1.tif (34MB, tif)
supp_now097_now097supp_fig2.tif (33.2MB, tif)
supp_now097_now097supp_fig3.tif (24.9MB, tif)
supp_now097_now097supp_fig4.tif (25.1MB, tif)
supp_now097_now097supp_fig5.tif (30MB, tif)

Articles from Neuro-Oncology are provided here courtesy of Society for Neuro-Oncology and Oxford University Press

RESOURCES