Fig 1: Fbxo45 activates ERK by suppression of NP‐STEP46 in NSCLC cells. (A) HEK‐293T cells transiently transfected with or without HA‐Fbxo45 plasmid for 48 h were treated with the fresh complete cell culture medium containing DMSO, U0126 (20 μm), or LY294002 (50 μm) for 1 h. The extracted cell lysis was resolved in SDS/PAGE gel for western blotting with antibody pERK, pAKT, HA, and tubulin as the loading control. Two individual replications were set, and the representative images were reported. (B) HEK‐293T cells were transiently transfected with a gradient amount of HA‐Fbxo45 plasmid for 48 h, and the cell lysis was resolved in SDS/PAGE gel for immunoblotting with the indicated antibodies. Two individual replications were set, and the representative images were reported. (C) Three independent pathologists estimated and scored the IHC score of Fbxo45 and pERK staining from 71 pairs of samples. The expression relation of Fbxo45 and pERK were statistically analyzed with Pearson correlation, the 95% CI was dotted (right panel). The representative images for IHC detection of Fbxo45 and pERK, along with HE staining, were presented in the left panel. Scale bar: 100 μm. (D) The cultured A549 and H1299 cells were harvested and lysed in RIPA buffer (150 mm NaCl, 50 mm tris–HCl pH7.4, 1% NP‐40, 0.1% SDS, 1% sodium deoxycholate, 1 mm EDTA, 1× proteasome inhibitor cocktail) for immunoblotting with the indicated antibodies. Two individual replications were set, and the representative images were reported. (E) The constructed cell lines with or without Fbxo45 shRNA expression were harvested and lysed in RIPA buffer for western blotting with the indicated antibodies. Two individual replications were set, and the representative images were reported. (F) The A549 cell was used to generate the Fbxo45‐knockout cell line (A549/Fbxo45KO) by the CRISPR‐Cas9 system, and the wide‐type A549 (A549/WT) cell line was set as a negative control. 3xFlag tagged Fbxo45 was re‐introduced into A549/Fbxo45KO cells to generate A549/Fbxo45KO + flag‐Fbxo45 cell line by the lentiviral system, and the A549/Fbxo45KO and A549/WT infected with the control lentivirus were set as the system controls. A549/Fbxo45KO, A549/Fbxo45KO + flag‐Fbxo45, and A549/WT cells were harvested and lysed in RIPA buffer for immunoblotting with the indicated antibodies. Two individual replications were set, and the representative images were reported. (G) A549/Fbxo45KO, A549/Fbxo45KO + flag‐Fbxo45, and A549/WT cells were seeded in 96‐well plates (1000 cells per well) for the CCK8 assay. Data are presented as mean ± SEM with t‐test from three independent experiments. **P < 0.01. (H) A549/Fbxo45KO, A549/Fbxo45KO + flag‐Fbxo45, and A549/WT cells were seeded in 6‐well plates (2000 cells per well) for the anchorage‐independent growth assay, and cell colonies were numbered and statistically columned. Data are presented as mean ± SEM with t‐test from three independent experiments. **P < 0.01. (I) Fbxo45, NP‐STEP, and pERK were detected by H&E or immunohistochemistry staining in tumors from xenografted mice with A549 cells with or without Fbxo45 shRNA expression. Three individual replications were set, and the representative images were reported. Scale bar: 100 μm. [Colour figure can be viewed at wileyonlinelibrary.com]
Fig 2: Silencing Fbxo45 inhibits the proliferation of NSCLC cells and tumor growth in mice. (A) Fbxo45 expression was detected by western blotting in the indicated cell lines infected and selected using the shRNA lentiviral system with three independent replications. (B) The proliferation of cell lines with or without Fbxo45 shRNA expression was presented in columns according to the CCK8 assay and statistically analyzed by graphpad prism 8 software. Data are presented as mean ± SEM with t-test from three independent experiments. **P < 0.01. (C) The colony formation of cells was stained with 0.02% crystal violet in 12-well plates (left panel) and statistically analyzed according to the positive-stained cell number (right panel, data are presented as mean ± SEM with t-test from three independent experiments. **P < 0.01). (D) In the anchorage-independent growth assay, cells with or without Fbxo45 shRNA expression were stained with 0.05% crystal violet in 0.35% soft agarose (left panel) and statistically analyzed according to the positive-stained cell colonies (right panel, data are presented as mean ± SEM with t-test from three independent experiments. **P < 0.01). (E) The tumors were resected and compared from subcutaneous xenografted Balb/c nude mice at 27 days post-injection with A549 cell lines with or without Fbxo45 shRNA expression (left panel). The tumor size was collected and compared in different periods, and data are presented as mean ± SEM with t-test from three independent experiments. **P < 0.01 (right panel). (F) Fbxo45 and Ki67 were detected by H&E staining or immunohistochemistry detection in tumors from xenografted mice with A549 cells with or without Fbxo45 shRNA expression. Three individual replications were set, and the representative images were reported. Scale bar: 100 µm. [Colour figure can be viewed at wileyonlinelibrary.com]
Fig 3: Aberrant Fbxo45 expression is probably associated with the development of NSCLC. (A) Expression distribution of the Fbxo45 gene in tumor and normal tissues was generated from the website (www.aclbi.com) based on TCGA and GTEx datasets. The significance of the two groups of samples passed the Wilcox test, and the asterisk represents the degree of importance (***P < 0.001). TPM, Transcripts per million; Null, Incomparable; T, Tumor; N, Normal; the case number was shown in the brackets. (B) The indirect immunofluorescence was used for Fbxo45 detection in the paraffin sections from NSCLC patients (right panel). The immunochemistry staining was used for TTF‐1 detection (middle panel), and HE staining was shown in the same scope (left panel). Three individual replications were set, and the representative images were reported. Scale bar: 20 μm. (C) The Fbxo45 expression was dot‐plotted and compared among normal lung, large cell lung carcinoma, lung adenocarcinoma, and squamous cell lung carcinoma according to the ONCOMINE TCGA dataset (https://www.oncomine.org). The case number was shown in the bracket in each group. The data are presented as mean ± SD with t‐test analysis, ***P < 0.001. (D) The GEPIA web tool (http://gepia.cancer‐pku.cn) was used to generate the survival curve for LUAD and LUSC patients based on Fbxo45 expression in TCGA datasets. The 95% confidence interval (CI) was set as dotted line. (E, F) the Fbxo45 protein level was detected by immunohistochemistry in the microarray tissue chip containing 75 pairs of LUAD patients, and the representative samples were presented with Fbxo45 detection and HE staining (E) scale bar: 100 μm. The positive incidence of Fbxo45 staining from 71 pairs of LUAD samples was statistically calculated using the software of aperio imagescope (Leica Biosystems) and dot‐plotted in pairs (G) or non‐pairs (F) with a t‐test by graphpad prism 8 (GraphPad Software Inc.). The data are presented as mean ± SD. (H) The fresh surgical resections from NSCLC patients were lysed in 2% SDS lysis buffer for immunoblotting. Two individual replications were set, and the representative images were reported.
Fig 4: Fbxo45 silencing increases the sensitivity to Afatinib on H1975 in vitro and in vivo. (A, B) The colony formation of cells harboring control shRNA or Fbxo45 shRNA with triplicates each was treated with or without EGFT-TKI Afatinib (1 µm) and stained with 0.02% crystal violet in 12-well plates (A). The status of EGFR and KRAS in each cell line was shown (A). The graphpad prism 8.0 software statistically columned the positive-stained cell number, and data are presented as mean ± SEM with t-test from three independent experiments. **P < 0.01 (B). (C–F) In four cell lines, cells with or without Fbxo45 shRNA expression were treated with different Afatinib concentrations with triplicates. The cell viability was detected using the CCK8 kit and calculated for IC50 curve making. The data was reported as mean ± SEM. Drug concentrations corresponding to IC50 were indicated with the dash lines. (G, H) H1975 cells harboring Fbxo45 shRNA were harvested individually for subcutaneous xenografted in Balb/c nude mice. After the tumor formation at 10 days, the mice with similar tumor sizes were divided into two groups for intragastric administration with DMSO or Afatinib (10 mg·kg-1). The tumor volume was measured every 4 days to generate the growth curve, and data are presented as mean ± SEM with t-test from five tumors of each group. ***P < 0.001 (G). Mice were executed at 27 days post-injection, and tumors were dissected for photograph (H). [Colour figure can be viewed at wileyonlinelibrary.com]
Fig 5: Fbxo45-mediated NP-STEP46 degradation via K6-linked ubiquitination sustains ERK activity and drug resistance in lung cancer. The dynamic balance of ERK1/2 activation and inactivation is essential for mammalian cells to maintain normal development, proliferation, differentiation, transformation, and apoptosis. The abnormal activation of ERK1/2 or failure to dephosphorylate pERK1/2 in time is the primary cause of drug resistance in lung cancer cells, a severe EGFR-targeted therapy issue. Aberrantly expressed Fbxo45 in NSCLC cells can recognize ERK-specific phosphatase NP-STEP46 (an isoform of activated STEP) and mediate proteasome degradation through conjugating the K6-ubiquitin linkage on it in the nucleus, resulting in the maintenance of abnormal phosphorylation levels of ERK1/2. On the contrary, inhibiting Fbxo45 prevents NP-STEP46 from degrading efficiently and suppresses the abnormal activation of ERK. Here, we reveal insight into pERK-maintenance for NSCLC cell survival, proliferation, and drug resistance. Targeting Fbxo45 or combining it with EGFR-TKIs but not a limit may be an attractive therapeutic strategy to combat the issue of drug resistance in NSCLC patients. [Colour figure can be viewed at wileyonlinelibrary.com]
Supplier Page from MilliporeSigma for Anti-FBXO45 antibody produced in rabbit