Development of a novel in vitro 3D model to identify cancer genes via insertional mutagenesis

Abstract

The aim of the presented research project was to develop a personalised in vitro based human model to test the genotoxicity of gene therapy viral vectors. For this purpose, human induced pluripotent stem cells (hiPSCs) and human induced pluripotent stem cell-derived 3D heps were used in combination with a number of lentiviral and adeno-associated viral vectors and molecular analysis was performed to detect insertion sites (IS) to assess the predictive power of 3D heps model to predict insertional mutagenesis. In this study, two previously designed plasmid were used, one of which contained the U3 region of the LTR from the wild type (pHV) and considered as “Unsafe” vector, and the other was a self-inactivating (SIN) lentiviral vector with no U3 region of the LTR (pHR) and used as a “Safe” vector. Initially, high-titre production of both lentiviral vectors was achieved in the HEK293 cell line. The titre was calculated using flow cytometry analysis. The recombinant AAV serotype-2 vectors were kindly provided by our collaborators in Australia. These vectors constructed with “Clean ITR” with strong and weak promoters, driving the green fluorescent protein (GFP) expression. In parallel, P106i, as an integration-free hiPSCs, was expanded and fully characterised, using flow cytometry, immunostaining and qPCR analysis. Expression of pluripotency cell surface markers such as SSEA4, TRA-1-81 and TRA-1-60 was detected in more than 90% of the cells using flow cytometry. The expression of POU domain class 5 transcription factor 1 (POU5F1 or OCT4) and SRY-Box Transcription Factor 2 (SOX2) as two major transcription factors regulating pluripotency were also confirmed at gene and protein levels by quantitative polymerase chain reaction (qPCR) and immunostaining, respectively. The high level of pluripotency markers confirmed the pluripotent state and the identity of P106i, which was used in this study. Then, a recently published protocol to generate phenotypically stable 3D heps from human embryonic stem cells (hESCs) were amended and optimised to generate 3D heps from hiPSC efficiently. This protocol addresses issues surrounding previous 3D protocols, such as scalability and long-term in vitro phenotypic stability. Notably, hiPSC-derived 3D heps displayed liver functions for an extended period. Standard characterisation tests were performed on day 20 of differentiation using a range of molecular and cell biology techniques, including immunofluorescence analysis of liver-specific markers such as HNF4-alpha and secretion of serum albumin (ALB) and alpha-fetoprotein (AFP) using ELISA assay. The result revealed a high level of ALB and a low level of AFP in hiPSC-derived 3D heps compared to conventional 2D hepatocyte-like cells. Following characterisation, hiPSCs and 3D heps were transduced with recombinant AAV serotype-2 vectors (at MOI 1E5) and lentiviral vectors (at MOI 20). Modifications were made to enhance transduction efficiency in 3D heps. The outer layer of the 3D heps exhibited the highest level of GFP expression, which transduced with rAVV serotype-2 vectors. In 3D heps transduced with pHR and pHV lentiviral vectors, the same pattern was observed; however, a higher level of GFP expression was observed in cells transduced with the pHR lentiviral vector. The result of successful transduction was confirmed by PCR analysis. hiPSCs were also transduced with pHR and pHV lentiviral vectors at two different time points, and the result was confirmed via PCR analysis. In order to see the effect of gene expression in the proximity of viral insertion at the individual cell level, single-cell cloning (SCC) was performed on hiPSCs which were transduced with pHR and pHV lentiviral vectors. The results showed positive GFP expression in hiPSCs, confirming transduction. The DNA and RNA of the single-cell clones, hiPSCs and 3D heps were collected and sent for downstream analysis of the insertion site (IS). Upon viral integration, it is crucial to detect IS in the setting of clonal dominance. For bioinformatic analysis, lentiviral IS was retrieved by EPTS/LM-PCR and CIS were detected in P106i and 3D heps samples. This result revealed an overall decrease in IS overtime in transduced cells for both lentiviral vectors by comparison of the days 3 and 30 time points. The number of identified IS in 3D heps transduced with pHR and pHV lentiviral was low compared to P106i cells. There was also an apparent reduction in IS in P106i cells transduced with pHR and pHV lentiviral vectors over time, suggesting possible cell death during propagation and culture expansion. Following EPTS/LM-PCR, identification of cancer-related genes IS, and CIS were performed in P106i cells and 3D heps. The result indicates that in P106i, the number of proto-oncogenes was higher in samples transduced with pHV lentiviral vectors, suggesting the possible effect of the full LTR vector on the number of cancer-related genes. In addition, polyclonality of pHV and pHR was observed in analysed single-cell clones of P106i. However, in clone G transduced with pHV lentiviral vector, the gene LINC01249 exhibited 93.4% of the viral sequence count. In addition, despite the polyclonality of the samples, the qPCR result revealed that the genes near the IS have increased in level of relative gene expression. In conclusion, the 3D heps provide a valuable in vitro tool to assess genotoxicity associated with viral vectors.