Page 87 - Read Online

Michaelis et al. Cancer Drug Resist 2019;2:447-56 I http://dx.doi.org/10.20517/cdr.2019.005 Page 451

also shown that acquired resistance to cytotoxic drugs can be associated with decreased sensitivity to kinase
inhibitors [77,78] . The clinical impact of this is difficult to determine, however, because the baseline sensitivity
of tumours to different anti-cancer therapies prior to the first-line treatment is not typically known.

MULTIPLE RESISTANCE MODELS ARE NEEDED TO REFLECT THE HETEROGENEITY OF THE
PROCESSES ASSOCIATED WITH RESISTANCE FORMATION
It is now generally accepted that cancer diseases are associated with tremendous intra-tumour
heterogeneity [79-81] . Although therapy-induced heterogeneity has not been investigated to the same extent,
there are indications that the processes underlying resistance formation are likely to be as complex [52,82-88] .

The advantage of cancer cell lines as models is that they are relatively easy to handle and enable high
throughput analysis at relatively low cost and in a timely fashion. Although they do not reflect the original
heterogeneity of the tumour they have been derived from, they are not as homogenous or clonal as
previously believed [34-37,89] . Resistance can occur by selection of pre-existing drug-resistant subpopulations
or by adaptation of originally drug-sensitive cells to anti-cancer therapies. Both mechanisms have been
shown to be represented in drug-adapted cancer cell lines [52,66-70,90-101] .

In this context, we have adapted the TP53 wild-type acute myeloid leukaemia (AML) cell lines MV4-11, OCI-
[102]
AML-2, OCI-AML-3, and SIG-M5 to the MDM2 inhibitor nutlin-3 in multiple independent experiments .
Nutlin-3-adapted sublines of the same AML cell lines displayed a substantial heterogeneity in the response
to other anti-cancer drugs. Notably, the biggest fold change (11.4) was detected in the response of two nutlin-
3-adapted MV4-11 sublines to doxorubicin, although nutlin-3 treatment selected a pre-existing TP53 mutant
subpopulation in this cell line. This indicates that even the drug-induced selection of a defined pre-existing
[102]
subpopulation in a cell line can result in phenotypically different sublines . New technologies including
single cell approaches will enable the elucidation of selection and adaptation processes during resistance
formation in more detail [94,103,104] .

Since many models will be needed to cover the complexity associated with acquired resistance formation,
we have established the Resistant Cancer Cell Line collection by adapting initially chemosensitive cancer
cell lines to clinical concentrations of targeted and cytotoxic anti-cancer drugs to enable the systematic
investigation of acquired drug resistance mechanisms. It currently contains 1300 cancer cell lines based on
125 parental cell lines from 16 cancer entities and reflects acquired resistance to 67 drugs (https://research.
kent.ac.uk/ibc/the-resistant-cancer-cell-line-rccl-collection). The DEN50-R platform is another project
dedicated to the generation of drug-adapted cancer cell line panels (http://www.den50-r.org).

CONCLUSION
This perspective is focused on the use of drug-adapted cancer cell lines as models of acquired drug
resistance in cancer. Drug-adapted cancer cell lines are, like every model system, associated with specific
advantages and limitations. Models including primary cancer cell cultures, three-dimensional cell (co-)
culture systems, tumour-derived organoids, and animal models better reflect certain aspects of tumour
growth such as intra-tumour heterogeneity, three-dimensional architecture, cancer cell interaction with the
cancer microenvironment, and/ or metastatic behaviour [105-114] . Such models can be used to study processes
that cannot be studied in cell lines. In this context, acquired resistance models have been established based
on cell line- and patient-derived xenografts, organoids, and transgenic tumour models [115-125] . However, cell
lines enable the establishment of a substantially larger number of models within a given timeframe and
at a given cost, which is critical for studying the drug-induced heterogeneity. Notably, data so far suggest
that the drug adaptation of cancer cell lines reveals similar resistance mechanisms as cell line-derived
xenografts and transgenic mouse models [116,118,123,125] .