The genetic circuit designed to control chimeric antigen receptor (CAR) [link to page] expression in the tumor microenvironment relies on exclusive control of hypoxia-inducible factor 1 alpha (HIF1α) by the introduced system. Various strategies have been employed to obtain HIF1α deficient cells. CRISPR strategies were applied by transient transfection and knockdown was lentivirally transduced (link to method) followed by sorting (method link) of fluorescence positive cells. Insertion of mutations was assessed by PCR or T7E1 assay (link) and the absence of gene product was confirmed by Western Blot (link).

Approach 1: CRISPR/Cas9 with HDR

The CRISPR/Cas9 system (clustered regularly interspaced short palindromic repeats / CRISPR-associated protein 9) allows precise editing of genes. Cas9 forms a complex with guide RNA (gRNA) and introduces DNA double strand breaks at sites complementary to the gRNA. These are repaired by different endogenous repair mechanisms, dependent on the circumstances. One process called homology-directed repair (HDR) can be triggered by co-transformation with a repair template containing regions homologous to the target. A plasmid kit with Cas9 from S. pyogenes was used to eliminate HIF1α by HDR (sponsored by OriGene). This was conducted with two plasmids containing different gRNAs, both targeting the first exon of HIF1α, close to the transcription initiation site (Fig. X). The reporter genes GFP and puromycin resistance were stably integrated via a repair template.

Approach 2: CRISPR/Cas9 induced NHEJ

The CRISPR/Cas9 system can also be used without repair template. This forces cells to repair the double strand break by non-homologous end joining (NHEJ), an error-prone process. At the cleavage site nucleotides are randomly added and removed during repair resulting in deletions, substitutions and insertions. This stochastically produces a shift of the translational frame in the gene, ablating the gene product’s function. For this strategy gRNAs were designed using the Broad Institute’s Genetic Perturbation Platform (GPP) online tool. Off-target effects were predicted using the program CCTOP. The most promising candidates were inserted into Cas9 (S. pyogenes) plasmids either by class-II restriction enzyme cloning of dsDNA oligomers or by site-directed mutagenesis PCR. Two selection markers, GFP (Cas9-GFP provided by Reth Lab at BIOSS, originally Zhang Lab at MIT) and puromycin resistance (lenti-Cas-puro-v2, provided by Cathomen Lab at CTC, originally Zhang Lab at MIT) were tested.

Approach 3: RNA Interference

Another technique to inhibit gene expression is RNA interference, where translation is inhibited by antisense RNA binding to target mRNA. This prevents association of the translational machinery and therefore protein production. Our approach involved stable lentiviral transduction of such short hairpin RNA (shRNA) into cells to obtain continuous downregulation of HIF1α mRNA, as opposed to an only transiently decreased expression level caused by treatment with the RNA molecule itself. These shRNAs rely on the endogenous processing mechanisms of cellular regulatory RNA. The endo-ribonuclease Dicer cuts off the hairpin of the palindromic transcript and the dsRNA forms together with several proteins the so-called RNA-induced silencing complex (RISC). It unwinds the dsRNA producing active RNA complementary to the target mRNA. Since the eradication of HIF1A gene product in our project serves to allow control of this gene’s expression by a synthetic construct, it is crucial that the designed shRNA sequences are complementary to endogenous mRNA but not that of the introduced gene. Therefore, untranslated regions of this transcript were used as target for shRNA sequence design. Some sequences were obtained from research group Zuber in Vienna, others designed with GPP as for the CRISPR experiment. The lentiviral transfer plasmid into which shRNA sequences were cloned contained cyan fluorescent protein (CFP) and neomycin resistance markers.