Project FER kinase
Triple Negative Breast Cancer
and
HPV negative Head & Neck Squamous Cell Carcinoma
The do-it-all FER (FES related) kinase.
Schematic representation of the protein domains of the 90 KDa non receptor tyrosine kinase FER. Shown are the position of the tyrosine phosphorylated residues (orange balls). As a comparison FES and FER-T as shown as well.
Below the cartoon is a hypothetical model for FER di- or oligomerization at the membrane, where FER kan directly target substrates such as MAPK1, Dynactin2, Tubulin, SKOR1 and Cortactin (indicates as an Orange star)
A bit of history on how we got started with FER
FER was initially picked up around 2007 as a candidate HIF target in breast cancer. While this turned out to be a false-positive finding, it became clear that high FER levels correlated strongly with breast cancer aggressiveness in clinical samples. In 2010 our lab started working on the mechanisms that underpin FER’s function. Iordanka Ivanova, a senior postdoc in the lab made the first observations that high FER levels correlated with the basal phenotype in breast cancer cell lines. The same was later found for the clinical breast cancer cohorts we looked at: FER was highly correlated with high grade and basal typer breast cancers. Subsequent loss of function experiments showed that FER inactivation using siRNA mediated knockdown led to a complete inhibition of cellular motility and a mild inhibition in growth. Next, Iordanka could show that FER acted upstream of the integrins α6 and β1 integrins. FER loss induces a retardation of integrin expression on the cell membrane and perinuclear (endosomal?) compartments. These findings were published in Oncogene in 2013. It remained unclear if the tyrosine kinase activity of FER was central to the observed phenotypes in cancer.
FER controls adhesion and migration through integrin recycling
(A) Cell spreading. Control and FER KD cells were plated on Collagen I and expression of F-actin (green) and phospho-paxillin (p-pax; red) were analyzed . (B) FER kinase downregulation promotes accumulation of α6-integrin in endosomes in MDA-MB-231 (MM231) on laminin-coated coverslips. Shown are α6 integrin (α6; green) and early endosome antigen 1 (EEA1; red) distribution. Localisation plasma membrane (arrow) and endosomes (arrow head).
Generation and validation of the Gatekeeper FER mutant
(A) Chemical genetics to identify FER substrates. Mutation of the gatekeeper residue (M637A/G) and inhibition by an analog of the Src inhibitor PP1 (NM-PP1). (B) In vitro kinase assays using recombinant kinase-dead GST-FER (K742R; left lane) and the gatekeeper GST-FER (M637A; middle and right lane) in the presence of ATPγS. (C) In vitro kinase assay with analog-sensitive M637A FER in the presence of ATPγS, N6-furfuryl (Fu)-ATPγS, N6-phenylethyl(Phe)-ATPγS, or N6-benzyl(Bn)-ATPγS. Recombinant Cortactin substrate was used in each assay as control.
FER Chemical Genetics in action
In 2016, Dr. Sandra Tavares started a postdoctoral fellowship in the lab, working on the FER project (sponsored by KWF). Based on preliminary work from Dr. Ron Schackmann (lab alumnus), she performed a chemical genetics screen, attempting to identify direct FER tyrosine kinase substrates using a synthetic “Shokat” allele. As one of the first labs in the world we succeeded in successfully performing this screen, which pinpointed approximately 8 novel and 2 suspected direct FER substrates. First, Sandra could show that MAPK1 is directly phosphorylated by FER on Tyr187, which turned out to induce 20% of the total MAPK1 (ERK2) activity in MDA-MB-231 cells after EGF activation. Second, she identified the Tyr234 of Dynectin-2 (DCTN2) as a direct FER site that appeared to mediate retrograde endosomal transport over the microtubule (MT) cytoskeleton. Other hits were SKOR1, Tubulin and Cortactin. See Sandra’s Cell Reports paper from 2022.
FER is prognostic and predictive in breast cancer
Given the links with MT mediated transport, we also investigated the clinical impact of high FER on Taxane responses in breast cancer patients. Using the MATADOR patient set, we could show that high FER mRNA levels were prognostic. More important, Sandra showed in a collaboration with Prof. Sabine Linn (NKI-AvL and UMC Utrecht) that High FER predicted Taxane benefit in high-risk breast cancers, especially those of the triple negative subtype (TNBC). Combined, these findings were published in Cell Reports in 2022. Other direct substrates we identified were Vimentin, Cortactin, Tubulin and SKOR1. The latter plays a role in the control of invasive breast cancer growth, findings that were published in the Journal of Cell Science in 2023.
A model how FER controls invasive growth in cancer cells
Shown is a cartoon with a model on how we think FER controls microtubule (MT) -dependent transport of endosomal cargo during invasive growth of cancer. Taxanes are beneficial for patients with HIGH-FER tumors, because these cells rely on MTs for effective transport of oncogenic receptors, being both adhesion and signaling receptors.
An example of a representative experiment showing the function of DCTN2 in cell adhesion, migration and invasion. We reconstituted the gatekeeper FER in FER deficient cells to demonstrate that FER’s functions are mostly kinase dependent (using NM-PP1, left images). On the top right we show the effect of DCTN2 loss and the impact of the FER site Y6 on DCTN2 on adhesion. The below panel (DIC images) shows the same for invasion in Collagen I gels (3D).
Now, we have extended our work on FER into the HPV negative Squamous Cell Head and Neck cancers (HNSCC), work done by Peter Haughton (PhD candidate)
Our group is performing in vitro (patient derived organoid models) and in vivo mouse studies to understand the cellular recycling mechanisms that underpin FER’s oncogenic roles and developing interventions therapies targeting FER and its effectors in TNBC and HNSCC.
Peter developed a spatial expression classifier together with Wisse Haakma (a fromer technician in the lab) based on cytokeratin expression in organoids and patient samples that is prognostic in HNSCC. The work was published in Oral Oncology Reports in 2024.
Shortly thereafter, Peter published a paper in Oncogene, where we identified a transcriptional signature based on collective invasion in 3D. This Collective Invasion signature showed prognostic value in clinical HNSCC cohorts. The paper was a fruitful collaboration with the group of Stefan Prekoviç at the next door Center for Molecular Medicine (CMM).
We are currently studying the role of FER kinase in the regulation of endosomal dependent activation of growth factor receptor signalling in HNSCC.
Definition of the prognostic 5-tier spatial expression system
Cartoon depicting the observed spatial expression patterns in HNSCC: Negative, Homogeneous (expressed by all cells), Mosaic (heterogeneous expression), Core (expression limited to a central tumor region) and Edge (first 2-3 cell layers adjacent to the ECM interface). See paper for details.