Prevention and prediction of production instability of CHO-K1 cell lines by the examination of epigenetic mechanisms

The CHO-K1 cell line is the most common expression system for
therapeutic proteins in the pharmaceutical industry. Due to the
nature of economics, the cell lines and the vector design are
subject to constant change to increase product quality and
quantity. During the cultivation, the production cell lines are
susceptible to decreasing productivity over time. Often the loss of
production can be associated with a reduction of copy number and
the silencing of transgenes. During cell line development, the most
promising cell lines are cultivated in large batch culture.
Consequently, the loss of a stable production cell line can be very
cost-intensive. For this reason I developed different strategies to
avoid a reduced productivity. Instability of production cell lines
can be predicted by the degree of CpG methylation of the driving
promoter. Considering that the DNA methylation is at the end of an
epigenetic cascade and associated with the maintenance of the
repressive state, I investigated the upstream signals of histone
modifications with the assumption to obtain a higher predictive
power of production instability. For this reason I performed a
chromatin immunoprecipitation of the histone modifications H3K9me3
and H3K27me3 as repressive signals and H3ac as well as H3K4me3 as
active marks. The accumulations of those signals were measured
close to the hCMV-MIE at the beginning of the cultivation and were
then compared with the loss of productivity over two month. I found
that the degree of the H3 acetylation (H3ac) correlated best with
the production stability. Furthermore I was able to identify an
H3ac threshold to exclude most of the unstable producers. In the
second project I aimed to improve the vector design by considering
epigenetic mechanisms. To this end I designed on the one hand a
target-oriented histone acetyltransferase to enforce an open and
active chromatin status at the transgene. On the other hand I
point-mutated methylation-susceptible CpGs within the hCMV-MIE to
impede the maintenance of inactive heterochromatin formation.
Remarkably, the C to G mutation located 179 bp upstream of
transcription start site resulted in very stable antibody producing
cell lines. In addition, the examination of cell pools expressing
eGFP showed that G-179 promoter variants were less prone to a
general methylation and gene amplification, which illustrates the
dominating effect in epigenetic mechanisms of one single CpG. The
last project was performed to localize stable integration sites
within the CHO-K1 genome. In so doing I could show that the
transfection leads predominantly to integration into inactive
regions. Furthermore I identified promising integration sites with
a high potential to induce stable expression. However, those
results are preliminary and must be viewed with caution. Further
examination needs to be done to confirm these results. Considering
the results of all three projects, I propose that the interplay of
metabolic burden and selection pressure at an early time point of
cultivation plays an important role in cell line development. Small
alterations of selection pressure can lead to a decisive change of
cell properties. Therefore, stable cells are less susceptible than
weak producers. The increase of selection pressure leads to
compensatory effect by gene amplification in the instable cell
lines. The resulting adjustment of productivity masks the truly
stable cells, which precludes the selection of the right cell
lines. For this reason the selection pressure, the copy number as
well as the growth rate should be considered to minimize repressive
effects.