SUM-52 cells are one of four cell lines that we developed that have a receptor tyrosine kinase (RTK) gene as one of the driving oncogenes in the oncogene signature. The other three cell lines are SUM-185 (FGFR3), SUM-225 and SUM-190 both of which have HER2 amplification and overexpression. In SUM-52 cells, the FGFR2 gene is amplified and highly overexpressed, and these cells are very sensitive to targeted FGFR inhibitors. As we and others have shown, the SUM-52 cells not only overexpress FGFR2 at the message and protein level, the cells express many FGFR2 isoforms some of which are associated with expression of transformed phenotypes. The FGFR family is noteworthy by the way alternative splicing of the four FGFR genes is used to generate receptors with specificity for the more than 20 members of the fibroblast growth factor (FGF) family of ligands. In normal cells, the specific splice isoform generated is cell type and developmentally specific. However, SUM-52 cells express over 10 different FGFR isoforms, some of which express the C3 carboxy terminus. This type of FGFR2 is not expressed in normal mammary epithelial cells, and has been associated with expression of transformed phenotypes. As a result, FGFR2 gene amplification provides for a potent transforming stimulus to these cells, but also renders them highly sensitive to FGFR specific kinase inhibitors.
In addition to FGFR2, SUM-52 cells overexpress two other amplified oncogenes that we have characterized: CDK6 and KATA6A, otherwise known as MYST3 (the oncogene signature contains other candidate oncogenes that we have not yet studied). Indeed, both cyclin D1 and CDK6 were hits on our functional shRNA screen. The KATA6A story is of particular interest to our lab as this gene has been predicted to be a human oncogene based on genomic and bioinformatic analysis of human cancers, but has, thus far, not been described as a candidate oncogene from the 8p11 amplicon, which is also present in SUM-52 cells. In fact, it was the presence of the 8p11-p12 amplicon that we were initially interested in because this region also contains the FGFR1 candidate oncogene, however, we found out early on that FGFR1 is not expressed in SUM-52 cells. Rather, the NSD3 (WHSC1L1) and KAT6A oncogene are the more important drivers from the 8p11 region in this cell line.
SUM-52, like the other cell lines with RTK oncogenes, exhibits a large number of hit genes that map to the PI3’K/AKT pathway. Examination of the hit genes that map to this pathway shows that several other important kinases and signal transducers were also strong screen hits including PDK1, AKT, mTOR, IRS1 and others. And, we showed that FGFR signaling is directly linked to AKT phosphorylation in these cells. In addition, it can be seen that the cell survival gene BCL2L1 (Bcl-xL) was also a screen hit that maps to this pathway. Indeed, we showed that inhibition of FGFR2 coupled with inhibition of BCL2L1 resulted in synergistic cell death of SUM-52 cells, providing a good example of how interpretation of screen hit data in terms of canonical pathways can provide insights into potentially effective targeted drug strategies. Interestingly, combining PI3’K/AKT inhibitors in combination with BCL2L1 targeted drugs were as effective as combining the latter with FGFR2 inhibitors.
Once again we see that in SUM-52 cells, like the other SUM lines, WNT signaling, and cytokine signaling appears to be important for the growth and survival of these cells. WNT8a was a hit in the screen, as was XCL1, CCL20, CCL3, IL9, and a number of cytokine receptors such as FLT4, which was a very strong hit in the screen. Thus, even though these cells appear to be driven by classical oncogenes like FGFR2 and CDK6, the cells seem also to synthesize some of their own growth factors and receptors that play a role in their proliferative capacity. With respect to WNT signaling, we showed many years ago that FGFR2 activity in SUM-52 cells has a dramatic effect on the expression of the TC1 gene, which is also amplified in these cells, and TC1 has been linked to WNT pathway activity.
Finally, examination of the Signaling Pathways Regulating Pluripotency of Stem Cells provides some insights as to how multiple oncogenes active in SUM-52 cells may interact to influence this important phenotype of metastatic cancer cells. It is well known that FGFs and FGFRs play a role in the biology of stemness, as does WNT signaling, and this pathway shows how these regulators could be interacting with the presence of the KAT6A (Myst3) oncogene to regulate expression of genes associated with stemness. Indeed, several members of the Notch signaling pathway were also strong hits in the shRNA screen in SUM-52 cells (check it out in the Pathway Engine).
Thus, the SUM-52 cell line is an outstanding model for luminal metastatic breast cancer that exhibits the gene expression profiling of a luminal B breast cancer, and in which expression of the estrogen receptor has been lost. As a result, the growth and survival of these cells has been taken over by classical oncogenes and genes that block programmed cell death.
Finally, examination of the peer-reviewed publications in which data derived from SUM-52 cells were published revolves around a few specific themes including:
-
The mapping of the break points in the 8p11 amplicon
-
The sensitivity of FGFR2 amplified cells to clinical drugs that target FGFRs
-
The biology of splice variants of FGFR2
-
The EMSY oncogene, which was initially cloned and identified in SUM-52 cells
-
Cooperation between oncogenes on the 8p11 and 11q13 regions which occurs commonly in breast cancer
-
Identification of candidate oncogenes from the 8p11 amplicon including LSM1, ERLIN2, and NSD3, in addition to KAT6A
Please refer to the SUM-52 Bibliography page for key papers from the Ethier lab on the SUM-52 cells
