An Arrayed CRISPR Screen Identifies Knockout Combinations Improving Antibody Productivity in HEK293 Cells

Mammalian cells are used to express complex biologics,
such as
multispecific antibodies. While multispecifics enable promising new
strategies for treating human disease, their production at high expression
titer and purity can be challenging. To understand how cells respond
to antibody and multispecific expression, five molecules were selected
for bulk RNA sequencing (RNA-seq) early after the transfection of
a human embryonic kidney 293 (HEK293) host. All five molecules shared
a differential expression signature of secretory and protein folding
stresses, but this signature was stronger for molecules with low titer.
We then designed an arrayed CRISPR knockout screen of 206 differentially
expressed target genes and 223 literature-motivated targets to identify
knockouts that affect antibody productivity. Eight novel knockout
targets were identified that increased expression titers by 20–80%.
Notably, seven of these top eight hits were from the differentially
expressed set of candidate-gene knockouts. The top knockout target,
HIST2H3C, showed evidence for additivity with five other hits, including
a knockout combination that increased the titer of a difficult-to-express
antibody by up to 100%. Findings for both HIST2H3C and INHBE knockout
targets generalized to an alternate HEK293 host expressing an additional
antibody and a multispecific host with no meaningful impact on product
purity. Thus, we propose HIST2H3C and INHBE disruption as a promising
and novel strategy for host-cell engineering to improve antibody and
multispecific productivity.