Adenovirus-Mediated Delivery of Decoy Hyper Binding Sites Targeting Oncogenic HMGA1 Reduces Pancreatic and Liver Cancer Cell Viability

Abstract

High mobility group AT-hook 1 (HMGA1) protein is an oncogenic architectural transcription factor that plays an essential role in early development, but it is also implicated in many human cancers. Elevated levels of HMGA1 in cancer cells cause misregulation of gene expression and are associated with increased cancer cell proliferation and increased chemotherapy resistance. We have devised a strategy of using engineered viruses to deliver decoy hyper binding sites for HMGA1 to the nucleus of cancer cells with the goal of sequestering excess HMGA1 at the decoy hyper binding sites due to binding competition. Sequestration of excess HMGA1 at the decoy binding sites is intended to reduce HMGA1 binding at the naturally occurring genomic HMGA1 binding sites, which should result in normalized gene expression and restored sensitivity to chemotherapy. As proof of principle, we engineered the replication defective adenovirus serotype 5 genome to contain hyper binding sites for HMGA1 composed of six copies of an individual HMGA1 binding site, referred to as HMGA-6. A 70%-80% reduction in cell viability and increased sensitivity to gemcitabine was observed in five different pancreatic and liver cancer cell lines 72 hr after infection with replication defective engineered adenovirus serotype 5 virus containing the HMGA-6 decoy hyper binding sites. The decoy hyper binding site strategy should be general for targeting overexpression of any double-stranded DNA-binding oncogenic transcription factor responsible for cancer cell proliferation.

Keywords: HMGA1; adenovirus; cancer therapy; chemotherapy resistance; decoy binding site; high mobility group A protein; liver cancer; neoadjuvant therapy; oncogenic transcription factor; pancreatic cancer.

Graphical abstract

Figure 1

Schematic Depiction of the Design of the HMGA-6 Hyper Binding Site and Its Insertion into a Shuttle Vector Needed for Incorporation into the Virus Genome (A) The HMGA-6 hyper binding site is depicted by six consecutive boxes labeled A15 or T15. The site was integrated into the pShuttle CMV vector in preparation for homologous recombination with the pAdEasy vector (B). The regions of sequence homology are designated as the “Left Arm” and the “Right Arm” common to both vectors. Successful homologous recombination resulted in insertion of the HMGA-6 hyper binding site into the adenovirus genome as indicated in (C).

Figure 2

Cytotoxic Effects Caused by Viral Infection (i) Negative control, AD293 cells transfected with the pUC-GFP plasmid DNA. (ii) Infection with AdEasy DNA caused detachment and clumping of cells characteristic of cytotoxicity associated with viral replication. (iii) Infection with the AdEasy-HMGA-6 DNA also resulted in a cytotoxic effect. All images were taken with a 20× objective lens.

Figure 3

Immunocytofluorescence Assays for Viral Coat Proteins in Infected AD293 Cells (i) Fluorescence images of AD293 cells infected with linearized native AdEasy DNA. (ii) Fluorescence images of AD293 cells infected with linearized AdEasy-HMGA-6 DNA. Assays for uninfected cells exhibited no fluorescence (data not shown).

Figure 4

Western Blot Analysis of HMGA1 Expression Levels in Different Cancerous and Non-cancerous E6E7 Cell Lines 20 μg of total nuclear protein was loaded in each well, transferred to membrane and probed with rabbit anti HMGA1 antibody. TATA binding protein was used as loading control.

Figure 5

Cell Viability Assays following Treatment of Pancreatic and Liver Cancer Cells with AdEasy-HMGA-6 DNA (A–F) The effect of AdEasy-HMGA-6 infection on cell viability was determined by comparing virus treated and untreated cell lines in the presence or absence of chemotherapy drug gemcitabine (GEM). 3 × 103 cells were seeded in a 96-well plate 24 hr before infection for (A) MIA PaCa-2, (B) AsPC-1, (C) BxPC-3, (D) PANC-1 (E) HepG2, and (F) E6E7. Cells were infected at three different viral doses (0.33, 3.3, and 33.3 ppc) of AdEasy or AdEasy-HMGA-6 virus in the presence or absence of four different doses of GEM (0 nM, 1 nM, 10 nM, and 100 nM). Cell viability was measured 72 hr after infection and/or gemcitabine treatment using the CellTiter 96 Aqueous one solution cell proliferation assay kit (Promega). Results were averages of two experiments performed in triplicate (n = 6). The data points represent the mean of the measurement. The error bars represent ± SD.

Figure 6

Plaque Assay Confirming Adenovirus Infection of E6E7 Cells (A–F) Monolayers of 100% confluent non-cancerous E6E7 cell line (immortalized pancreas ductal epithelial cells) were infected with Ad5ΔΔ (A–C) or Ad5 (D and E) virus in semi-solid agarose media. The following dilutions of virus were used: (A) 10⁻¹, (B) 10⁻², (C) 10⁻³, (D) 10⁻⁵, (E) 10⁻⁶, and (F) E6E7 cell only.

Figure 7

Apoptosis and Necrosis Assays following Treatment of MIA PaCa-2 Cells with AdEasy-HMGA-6 DNA (A–D) MIA PaCa-2 cells were seeded at a density of 2000 cells/well 24 h before infection with four different dosages (10, 20, 40, and 80 ppc) of either AdEasy-HMGA-6 (A or C) or AdEasy (B or D). Apoptosis and necrosis were measured using the RealTime-Glo Annexin V Apoptosis and Necrosis Assay (Promega) kits. Apoptosis (A and B) and necrosis (C and D) assays were measured in every 6 hr following virus infection and continued for 120 h. The data is only shown to 72 hr for the apoptosis assays. The data points represent the mean of 4 replicates. The error bars represent ± the standard deviation.