Focal adhesion kinase promotes ribosome biogenesis to drive advanced thyroid cancer cell growth and survival

Abstract

Introduction: Advanced thyroid cancer, including papillary (PTC) and anaplastic thyroid cancer (ATC), are the leading causes of endocrine cancer deaths. Thus, there is a critical need to identify novel therapeutic targets to improve standard of care. Focal Adhesion Kinase (FAK) is overexpressed and phosphorylated in thyroid cancer and drives thyroid cancer growth, invasion, and metastasis. FAK is a nonreceptor tyrosine kinase that is autophosphorylated at tyrosine 397 (Y397) in response to integrin or growth factor receptor signaling, resulting in the recruitment of SRC proto-oncogene and downstream signaling pathways. FAK is predominately localized at the plasma membrane but has recently been shown to accumulate in the nucleus as well as the nucleolus to drive tumor growth. The nucleolus is a membraneless subnuclear organelle that is involved in ribosomal biogenesis through the transcription, processing, and assembly of ribosomal RNA (rRNA). The role of FAK in ribosome biogenesis is currently unknown.

Methods: Nuclear/nucleolar FAK localization and function were studied using genetic and pharmacological approaches. High resolution microscopy was used to study the subcellular localization of FAK. Functional and biochemical assays including transformation and clonogenic assays, polysome profiling, and nascent protein synthesis assays were utilized to assess cell growth and survival. Protein-protein interactions of FAK were determined using a proximity dependent biotinylation (BioID) proteomics approach.

Results: We have found that pY397 FAK accumulates in the nucleolus of advanced thyroid cancer cells and that autophosphorylation of FAK at pY397 and FAK kinase activity are important for nucleolar accumulation of FAK. Furthermore, knockdown of nucleophosmin 1 (NPM1), an important structural component of the nucleolus, reduced pY397 FAK nucleolar accumulation. Functionally, we showed that nuclear FAK and FAK kinase activity are necessary for anchorage independent growth. We demonstrated that targeted degradation of FAK results in decreased protein synthesis with a specific decrease in the 60S ribosomal subunit. Using a BioID proteomics approach, we showed that autophosphorylated FAK interacts with a network of nucleolar proteins including nucleolar protein 56 (NOP56) which is a core small ribonucleoprotein (snoRNP) important for 60S ribosome biogenesis. Finally, we found that pY397 FAK co-localizes with NOP56 and that knockdown of NOP56 phenocopies FAK depletion.

Conclusions: Overall, these findings highlight a novel function for FAK in promoting ribosome biogenesis and suggest that nucleolar FAK represents a promising therapeutic target.

Keywords: NOP56; focal adhesion kinase; nucleolus; ribosomal biogenesis; thyroid cancer.

Figure 1

Excluding FAK from the nucleus decreases anchorage independent growth. (A) Parental and FAK Knockout CRISPR cells (FAK-KO) were immunoblotted with antibodies against pY397 FAK, pS910, pY925, total FAK, and β-actin. Molecular weight (MW) marker is shown on the left of each blot in kiloDaltons (kDa). (B) Parental and FAK-KO cells were plated for a 6 Day Vi-cell and total number of cells per mL was recorded. (C) Soft agar transformation assays were performed for parental and FAK-KO cells with the number of colonies counted after three weeks. (D) FAK-KO cells transduced with EV, WT FAK, NLM FAK, Y397F FAK, and KD FAK were immunoblotted with antibodies against pY397 FAK, total FAK, and α-tubulin. MW marker is shown on the left of each blot in kDa. (E) FAK-KO cells transduced with EV, WT FAK, and NLM FAK were stained with antibodies against pY397 FAK, Phalloidin, and DAPI prior to imaging at 60X with scale bar at 20μm. Inserts of zoomed in images have a scale bar of 5μm. At least 50 cells were imaged in three independent experiments. (F) FAK-KO CRISPR cells transduced with EV, WT FAK, NLM FAK, Y397F, and K454R were grown in soft agar assays, and the number of colonies were counted after three weeks. Experiments were performed in biological triplicates. Results are displayed as mean +/- SEM. **p < 0.01; ***p < 0.001.

Figure 2

pY397 FAK localizes to the nucleolus and influences nucleolar FAK accumulation (A) BCPAP, 8505C, and KTC2 cells were stained with antibodies against pY397 FAK, FBL, and DAPI prior to confocal imaging at 100X. (B) BCPAP, 8505C cells, and KTC2 cells were probed with antibodies against total FAK and DAPI prior to imaging at 100X confocal. Results shown are representative images from 20–30 images taken in at least two independent experiments. Scale bar represents 20μM. (C) FAK-KO cells transduced with EV, N-terminal NLS FAK, and N-terminal NLS-Y397F were probed with antibodies against V5, Fibrillarin, and DAPI prior to imaging at 100X on Olympus Confocal. At least 50 cells were imaged in three independent experiments. Scale bars represent 5μm. (D) Percent of co-localization between V5 and FBL in NLS versus NLS-Y397F KO-FAK transduced cells using Integrated Density Analysis by Image J. Results displayed as mean +/- SEM. *, p<0.05.

Figure 3

NPM1 is necessary for pY397 FAK localization to the nucleolus in adherent cells (A) BCPAP cells were probed with antibodies against NPM1, FBL, and UBTF and imaged using STED microscopy. At least 50 cells were visualized in three independent experiments. Scale bar represent 1μm. (B) Percent of co-localization between pY397 FAK and NPM1, UBTF, and FBL respectively using Integrated Density Analysis by Image J.(C) 8505C cells were knocked down with either a non-targeting control or three shRNAs targeting NPM1 prior to immunoblotting with antibodies against NPM1, pY397 FAK, total FAK, or α-tubulin. Quantification of NPM1 expression is normalized to α-tubulin and sh-COO2 control and is listed under the NPM1 blot. MW marker is shown on the left of each blot in kDa. Two independent experiments were performed. (D) Immunofluorescence was performed in 8505C shCOO2 or shNPM1 cells probed with antibodies against pY397 FAK, FBL, and DAPI. Cells were imaged at 60X confocal imaging in three independent experiments with visualization of at least 50 cells. Scale bar represent 5 μm. (E) Percent of co-localization between pY397 FAK and FBL in our non-targeting controls and shNPM1 knockdown cells using Integrated Density Analysis by Image J. Results displayed as mean +/- SEM. *, p<0.05; **p < 0.01; ***p < 0.001.

Figure 4

FAK regulates 60S biogenesis, nascent protein synthesis, and thyroid cancer growth (A) KTC1 and KTC2 cells were treated with 0, 4, or 24 hours of 1 μM PROTAC and immunoblotted with antibodies against pY397 FAK, total FAK, or α-tubulin. MW marker is shown on the left of each blot in kDa. (B) Polysome profiling of KTC1 and KTC2 cells treated with DMSO or 1 μM PROTAC for 24 hours was performed with absorbance measured at 260nm. Polysome profiling of KTC2 cells were performed in biological triplicate and KTC1 cells were performed in two biological replicates. (C) KTC1 and KTC2 cells were treated with DMSO or 1 μM PROTAC for 24 hours prior to methionine starvation, pre-treatment with water or CHX, and 1 hour treatment with methionine analog HPG. Lysates were immunoblotted with antibodies against TAMRA and vinculin. MW marker is shown on the left of each blot in kDa. At least two independent replicates were performed for all cell lines. (D) KTC1 and KTC2 cells were grown in clonogenic assays, treated with DMSO, 100 nM, or 1μM PROTAC, and stained with crystal violet after 14 days. Quantification of clongenic growth was performed by integrated density analysis using LICOR Odyssey. Experiments were performed in biological triplicate. Results are displayed as mean +/- SEM. *, p < 0.05; **, p < 0.01.

Figure 5

NOP56 co-localizes with pY397 FAK and promotes thyroid cancer growth (A) BioID workflow representing FAK fused to a promiscuous biotin ligase and interacting partners of FAK being biotinylated, pulled down, and identified by mass spectrometry.(B) Immunoblotting of BCPAP cells transduced with EV, WT, and Y397F FAK BioID2 and probed with antibodies against pY397 FAK, total FAK, and α-tubulin. The higher MW band represents the 27kDA BioID2 construct fused to WT or Y397 FAK and lower MW band is the endogenous FAK band. (C) STRING analysis of significantly enriched proteins reveals known interacting partners of FAK. Blue connections signify known interactions from database. Red connections signify experimentally determined known interactions. (D) FAK interacting partners were identified by analyzing proteins that bind to WT FAK in both datasets and enriched at least 2-fold change compared to EV. Gene Ontology Pathway Analysis was conducted for molecular function pathways with a false discovery rate (FDR) <0.05. (E) Gene Ontology Pathway Analysis was conducted for cellular compartment pathways with FDR < 0.05. (F) STRING analysis of nucleolar proteins identified by Gene Ontology Pathway Analysis. Blue connections signify known interactions from database. Red connections signify experimentally determined known interactions. (G) Confocal microscopy of KTC2 cells was performed with antibodies against pY397 FAK and NOP56. At least 50 cells were imaged at 100X with the Abberior Stedycon microscope in two independent experiments. Scale bar represent 2 μm. (H) Immunoblotting of KTC2 cells transduced with a non-targeting control or three NOP56 shRNAs was performed with antibodies against NOP56 and α-tubulin. MW marker is shown on the left of each blot in kDa .(I) Clonogenic assays of KTC2 cells transduced with non-targeting control or NOP56 shRNA were performed. Colonies were stained with crystal violet and imaged after 10 days. Integrated Density was utilized to quantify clonogenic growth using LICOR Odyssey software. Experiments were performed in biological triplicate experiments. Results are displayed as mean +/- SEM. *, p < 0.05.