Fig 1: GRB2 regulates miRNA expression in HEK293T cells. (a) Western blot of GRB2 expression in wild type (293 T) and depleted (G1) HEK293T clones 1 (G1.1) and 2 (G1.2). While G1.1 is a complete knockout, G1.2 contains a deletion and large insertion in the N-terminal SH3 domain. GRB2 was blotted with an antibody which recognised the C-terminal SH3 domain. Both long and short exposures were used to capture the GRB2 bands, whereas the GAPDH image was captured using a short exposure only. All images are taken from the same western blot. N = 3. (b) Heat plot highlighting miRNAs which show significant log2(fold changes) in expression (p < 0.05) between wild type HEK293T and G1 cells, measured by small RNA sequencing. Cells were deprived of growth factor. miRNAs demonstrated positive (red) and negative (blue) expression changes. N = 2. (c, d) RT-qPCR analysis of fold-change in mean expression of precursor miRNA transcripts (precursor and primary, pre-mir-, hashed bars) and mature miRNA (miR-, plain bars) derived from serum-starved G1 or wild type HEK293T cells. Two groups of miRNAs were observed: (c) miRNAs which diminished at both the level of the precursor and mature transcripts and, (d) miRNAs which were enhanced as mature transcripts but not as precursors. Comparisons were made using a two-tailed Student’s t-test and error bars show standard error of mean. N = 4. ns = not significant.
Fig 2: The GRB2-let-7 axis regulates oncogene expression. (a) RT-qPCR measurement of fold change in mean expression of let-7 g-5p miRNA and five target mRNAs in serum-starved GRB2 knockout cells (G1) compared to wild type HEK293T (293 T). Comparisons were made using a two-tailed Student’s t-test and error bars show standard error of mean. N = 3. (b) Western blot and (c) quantification of mean protein expression of let-7 targets in growth-factor-deprived G1 and HEK293T cells. The higher molecular band detected by the GRB2 antibody in G1 corresponds to an NSH3-mutated GRB2 polypeptide. For blot 1, a longer exposure was used to capture the DICER1 and GRB2 bands than was used for LIN28B and α-Tubulin. For blot 2, HMGA2 bands were captured using a longer exposure than that required for GRB2 and GAPDH. (d) Quantification of the area covered by migration of HEK293T cells expressing GFP-tagged wild type AGO2 (WT) or an AGO2 mutant which is incapable of binding GRB2 (4A), under conditions of reduced growth factor. Comparisons were made using a two-tailed Student’s t-test and error bars show standard error of mean. N = 3.
Fig 3: Impact of GRB2-AGO2 complex on interaction with DICER1 and miRNA. (a) Western blot of AGO2 and DICER1 pulldown by GST-GRB2 in HEK293T cells. HEK293T cells were serum-starved before lysis. Bands captured with both a long and short exposure are shown for DICER1, whereas only the image captured with a short exposure is shown for AGO2. GST proteins were detected by ponceau stain. All images are taken from the same western blot. N = 3. (b–d) MST of AGO2 binding to DICER1 C-terminal region, upon pre-incubation of AGO2 with increasing concentrations of GRB2. The difference in binding affinity was negligible. (e) MST of GRB2 with DICER1 C-terminal region. No binding is observed within a physiologically relevant range hence the two do not interact directly. (f) Expanded ribbon model of molecular docking of GRB2 (green; PDB: 1GRI77) to AGO2 PAZ domain (cyan; red and blue indicate positive and negative charges respectively; PDB: 6RA478). The 323PHLP326 sequence is shown (yellow). GRB2 W36 (magenta) interacts with AGO2 P249 (red). Other residues in GRB2 which may contribute towards the interaction are shown in orange. Also shown is space-filling representation of AGO2 PAZ domain with PRM shown (below); and ribbon model of PAZ domain rotated by 90° to highlight juxtaposition of GRB2 binding site PRM and docking site for miRNA (right). Figures generated using PyMOL.
Fig 4: Binding of GRB2 to AGO2 is mediated by GRB2 NSH3 and a PXXP motif in AGO2 PAZ domain. (a) Isothermal titration calorimetry (ITC) of a peptide spanning the proline-rich motif 323PHLP326 in AGO2 PAZ domain. (KD = 4.27 ± 1.17 µM). (b, c) ITC of MBP-tagged AGO2 PAZ domain titrated into GRB2. (b) PAZ WT (KD = 585 ± 61 nM). (c) No binding observed for mutation of PXXP (MBP-PAZ 4A). N = 2. (d) Fluorescence resonance energy transfer (FRET) between wild type (WT) and 323AAAA326 (4A) mutant GFP-tagged AGO2 and RFP-tagged GRB2 in HEK293T cells under conditions of serum starvation. White arrows indicate intracellular puncta which show increased FRET when WT AGO2 is expressed. N = 2. Scale bars are 10 μm. (e) Fluorescence lifetime imaging microscopy of RFP-tagged GRB2 proteins and GFP-AGO2 overexpressed in serum-starved HEK293T cells. The formation of a protein complex results in a reduction in fluorescent lifetime represented by a shift to the left of the population of fluorophores (measured in number of pixels). Lifetime population distribution shown by red line on graphs. x = Lifetime (ns), y = number of pixels. Solid black line corresponds to average fluorescent lifetime for GFP, 2.1 ns. Scale bars 25 μm. (f) Expanded region of interest (ROI) further exemplifying left-shift for AGO2/NSH3-SH2 interaction.
Fig 5: GRB2 complexes with AGO2 under non-stimulated conditions. Schematic diagram of, (a) AGO2 and, (b) GRB2 domain structures. Domains are named and colour coded and attributed amino acid sequence number. Red arrows indicate positions of PXXP motifs investigated in this work. (c) Western blot of AGO2 co-immunoprecipitated with GRB2 in serum starved HEK293T, A498 and PC3 cells. A longer exposure was used to capture AGO2 bands than for GRB2 and GAPDH. All images are taken from the same western blot. (d) Fluorescence and fluorescence resonance energy transfer signals of RFP-tagged GRB2 and GFP-tagged AGO2. HEK293T cells overexpressing fluorescent proteins were serum-starved before imaging. N = 3. Scale bars are 10 μm.
Supplier Page from OriGene Technologies for GRB2 Human Gene Knockout Kit (CRISPR)