Optimisation of molecular assays for clinical trial of GL67A/pGM169 delivery to nose and lung of CF patients.

Sumner-Jones SG, McCormick D, Coles RL, Pringle IA, McLachlan G, Collie DD, Vrettou C, Davies JC, Alton EW, Gill DR, Hyde SC

Pediatric Pulmonology, 44 S32 Abstract 269


The North American Cystic Fibrosis Conference, Minneapolis, 2009

We have identified the non-viral gene transfer formulation pGM169/GL67A for the next phase of the UK Cystic Fibrosis Gene Therapy consortium clinical programme. Quantification of pGM169-derived vector specific mRNA by real-time quantitative TaqMan RT-PCR is a key outcome measure for our clinical programme, additionally we also determine endogenous CFTR mRNA levels allowing us to determine %vector mRNA/endogenous mRNA ratio (%VE). Cell mixing, transgenic animal and mutational analysis suggest that a gene transfer formulation that achieves >5 %VE might be expected to be therapeutic. Simplistically, the determination of %VE in clinical samples is straightforward - standard methods exist for bronchial and nasal cell sampling, mRNA preparation and TaqMan RT-PCR. However, when we evaluated the performance of these standard methods in bronchial brushing samples from volunteer CF subjects it became apparent that pGM169 expression levels at 5 %VE would be required in order to allow detection and quantification of successful gene transfer. It was thus essential to refine our method by systematic optimisation of every individual step in the assay with the intention of reducing the Residual Sensitivity (RS) of the assay (that value of %VE required for detection and quantification) to levels well below that required for therapeutic benefit. During initial sample collection we evaluated the impact of bronchial and nasal brushings methods, airway cell retrieval strategies and time-allowed from sample removal to cellular preservation. We found that pooling bronchial brushing samples dramatically improved assay RS by minimising the fixed losses associated with mRNA purification. Furthermore, the speed of initial sample processing was found to be a key factor, and now require that it is completed with 15 minutes. Nevertheless, mRNA yields from bronchial brushing samples remain the key limiting step in the assay. Consequently, in the TaqMan RT-PCR stages, we evaluated the impact on RS and assay range and precision of alternative sample division strategies - higher RNA usage per replicate with low replicate numbers and vice versa along with the impact of multiplexing vector and endogenous CFTR assay within the same replicate. While each one of these steps in isolation had a modest impact on assay RS, the cumulative effect was highly significant. Using the optimised sample handling and assay conditions RS in CF bronchial brushing samples improved dramatically to 0.09% (p<0.05) and RS in CF nasal brushings improved from 0.9% to 0.08% (p<0.05). In parallel, we also optimised a purification method that allowed efficient DNA extraction from the cell lysates discarded at the end of the mRNA sample purification process. This co-purification strategy has allowed us to additionally determine plasmid DNA delivery in clinical sample without affecting mRNA assay RS. Together, these improvements in the processing of clinical samples and detection of vector RNA and DNA will permit the best opportunity to evaluate the delivery and expression of our gene therapy formulations in CF patients.