Yesterday our group discussed the recent Nature paper from Lynda Chin’s lab that identified GOLPH3 as a “first-in-class” Golgi oncogene.  The study began where most cancer genomics efforts end up: with the identification of a genomic region (5p13) that’s amplified in numerous solid tumours.  The authors reasoned that the amplified region likely contains a gene whose over-expression drives tumor development and/or growth.  The minimal common region of amplification spanned six annotated protein-coding genes, yet only two of them (GOLPH3 and SUB1) showed significantly increased gene expression associated with increased copy number.

As John Wallis in our group pointed out, it’s the ensuing wet lab experiments that made this a Nature paper.  Using a series of cancer cell lines (melanoma, prostate, pancreatic, NSC lung, and others), the authors show that overexpression of GOLPH3, but not SUB1, significantly increased colony formation and cell proliferation in vitro.  Next, they performed a series of experiments to assess the function of GOLPH3 in cancer:

• siRNA knockout, to demonstrate the effect of GOLPH3 on colony formation and cell proliferation.
• Co-transformation assays, to show that GOLPH3 cooperates with HRAS in mouse embryonic fibroblasts and transforms human cell lines that constitutively express BRAF.
• Mouse xenografts, to assess the effect of over-expressing GOLPH3 on tumor size.

Next, it was on to characterizing the mechanisms of GOLPH3’s oncogenic activity.  Using a yeast two-hybrid screen, the authors found that GOLPH3 interacts with VPS35, a protein that’s part of the “retromer” – a protein complex involved in retrograde transport between endosomes and the trans-Golgi network.  Though the nature of the interaction between VPS35 and GOLPH3 was not explored in-depth, this interaction (supported by staining/microscopy in cancer cell lines) implicated GOLPH3 as a Golgi protein and got the authors excited because, to their knowledge, it would be the first “Golgi oncoprotein.”

Linking GOLPH3 to mTOR

Pathway analysis is one of Lynda Chin’s specialties.  Based upon their findings and evidence in the literature, the authors posited that GOLPH3 might be involved in the mTOR signaling pathway, which (as described in our TSP lung adenocarcinoma paper) plays a central role in many solid tumors.

In a compelling series of experiments, the authors demonstrated that overexpression of GOLPH3 correlates with mTOR expression and increased phosphorylation of S6K, a substrate of the mTOR complex. This link suggests that GOLPH3’s regulates mTOR activity in mammalian cells, and because mTOR is the mammalian Target Of Rapamycin, it seemed reasonable to assess the influence of 5p13 copy number (and GOLPH3 expression) in the presence of rapamycin.  As you could guess from the paper’s title, GOLPH3 overexpression increased sensitivity to rapamycin treatment.  In fact, treating tumor cell lines with rapamycin had a similar effect on cell growth/proliferation as siRNA knockout of GOLPH3.  For the life of me, I couldn’t understand the “peak-forward scatter height flow histograms” in figure 4A, but I got the message: same pathway inhibited, same effect.

Clinical Implications: How Will This Help Treat Cancer?

The authors are optimistic but careful in their discussion of the clinical implications of these findings.  Obviously, GOLPH3’s role in tumor growth makes it an attractive target for rational drug design.  Thanks to the key pathway analyses in this study, it is clear that GOLPH3 confers its oncogenic activity through the mTOR pathway, which can be inhibited by rapamycin treatment.  Thus, the authors suggest that 5p13 copy number or GOLPH3 expression level might serve as a biomarker for sensitivity to mTOR inhibitors.  I wonder if this possibility might be taken a step further – could induction of GOLPH3 overexpression in the tumors of cancer patients render them sensitive to rapamycin in a sort of cancer one-two punch?  Regardless, if GOLPH3 expression or 5p13 copy number yields an informative biomarker for such treatment, it’s another step along the road towards personalized cancer therapy.

References
Scott, K., Kabbarah, O., Liang, M., Ivanova, E., Anagnostou, V., Wu, J., Dhakal, S., Wu, M., Chen, S., Feinberg, T., Huang, J., Saci, A., Widlund, H., Fisher, D., Xiao, Y., Rimm, D., Protopopov, A., Wong, K., & Chin, L. (2009). GOLPH3 modulates mTOR signalling and rapamycin sensitivity in cancer Nature, 459 (7250), 1085-1090 DOI: 10.1038/nature08109