Discussion
Here we described the development and testing of the GPEx® Lightning
technology. Combining GPEx®, GS selection, and targeted recombination
technologies leverages the advantages of both semi-targeted integration
and site-specific recombination for cell line development. This
technology consistently results in stable cell pools containing
approximately 50+ copies of recombined Boat gene construct per cell on
average. Clones isolated from these pools can contain 150 to 200+ copies
of Boat expression construct. This represents a total of up to 1.4
million base pairs of DNA stably inserted into the CHO genome at
multiple chromosome sites from a single transfection.
Generic fed-batch production of the resulting pools and clones showed
that Boat gene copy number correlated with final titer. Titers from pool
populations typically reach 3-7 grams per liter in non-optimized
conditions and up to 12 g/L with minimal optimization. Clones typically
reach 4-12 g/L and up to 15 g/L. Both pools and clones reach high cell
density with prolonged cell viability. Recent efforts focusing on
process intensification (i.e. higher inoculation density, feed rates,
and temperature) have achieved similar titers in shorter production runs
(not shown). Additionally, these pools and clones show very high genetic
and production stability over greater than 40 generations. The N-glycan
distribution across 40+ generations also remains remarkably consistent,
especially the amount of high-mannose and total fucosylated species.
Clones show N-glycan profiles consistent with the pools from which they
were isolated mitigating the risk of clones having divergent protein
quality.
This technology also allows for shortened development timelines. Once
the Boat gene construct is cloned and purified, transfection and
selection typically take less than 25 days to produce pool research cell
banks. These banks can then be scaled for large scale production. When
combined with the Berkeley Lights Beacon instrument for cloning, clones
can be produced, expanded, banked, and characterized in an additional 12
weeks.
Many pharmaceutical/biotech companies and CDMOs have moved away from
random integration approaches to cell line development in favor of
semi-targeted or targeted approaches to improve process consistency,
increase yield, and reduce timelines (Rajendra, 2017; Rajendran, 2021;
Feary, 2020; Ng, 2021). To our knowledge this is the first example of a
technology that combines the two latter approaches, allowing for a very
high number of transgenes to be inserted at consistent locations in the
genome. The speed of development, high and stable expression with
consistent glycosylation, and ability to scale to bioreactors make GPEx
Lightning pool-derived material suitable for early phase clinical
studies. The use of pool-derived material should also allow for cGMP
production of diagnostic assay components, molecules that are part of
medical devices, or process reagents needed to produce cell/gene therapy
products to be produced not only in a more cost-effective manner but
also with greatly reduced timelines. As we have seen with the COVID
pandemic, reducing timelines and costs through approaches like this can
have significant impact for drug development, patients, and society.