Glucocorticoid modulatory element-binding protein 1 (GMEB1) interacts with the de-ubiquitinase USP40 to stabilize CFLARL and inhibit apoptosis in human non-small cell lung cancer cells

Background GMEB1 was originally identified via its interaction with GMEB2, which binds to the promoter region of the tyrosine aminotransferase (TAT) gene and modulates transactivation of the glucocorticoid receptor gene. In the cytosol, GMEB1 interacts with and inhibits CASP8, but the molecular mechanism is currently unknown. Methods Human non-small cell lung cancer cells and 293FT cells were used to investigate the function of GMEB1/USP40/CFLARL complex by WB, GST Pull-Down Assay, Immunoprecipitation, Immunofluorescence and Flow cytometry analysis. A549 cells overexpressing green fluorescent protein and GMEB1 shRNA were used for tumor xenograft using female athymic nu/nu 4-week-old mice. Results We found GMEB1 interacted with CFLARL (also known as c-FLIPL) in the cytosol and promoted its stability. USP40 targeted CFLARL for K48-linked de-ubiquitination. GMEB1 promoted the binding of USP40 to CFLARL. USP40 knockdown did not increase CFLARL protein level despite GMEB1 overexpression, suggesting GMEB1 promotes CFLARL stability via USP40. Additionally, GMEB1 inhibited the activation of pro-caspase 8 and apoptosis in non-small cell lung cancer (NSCLC) cell via CFLARL stabilization. Also, GMEB1 inhibited the formation of DISC upon TRAIL activation. CFLARL enhanced the binding of GMEB1 and CASP8. Downregulation of GMEB1 inhibited A549 xenograft tumor growth in vivo. Conclusions Our findings show the de-ubiquitinase USP40 regulates the ubiquitination and degradation of CFLARL; and GMEB1 acts as a bridge protein for USP40 and CFLARL. Mechanistically, we found GMEB1 inhibits the activation of CASP8 by modulating ubiquitination and degradation of CFLARL. These findings suggest a novel strategy to induce apoptosis through CFLARL targeting in human NSCLC cells. Electronic supplementary material The online version of this article (10.1186/s13046-019-1182-3) contains supplementary material, which is available to authorized users.


Background
Glucocorticoid modulatory element-binding protein 1 (GMEB1) was originally identified as a nuclear protein with a molecular weight of 88 kDa [1][2][3]. GMEB1 interacts with GMEB2 and binds to the promoter sequence glucocorticoid modulatory element (GME) of the tyrosine aminotransferase (TAT) gene to modulate glucocorticoid receptor transactivation [4]. GMEB1 also exists in the cytosol and functions at the protein level. It interacts with the heat shock protein HSP27, but the function is not well studied. GMEB1 also binds to pro-caspases and inhibits their activation and cell apoptosis. However, it is still unknown how GMEB1 does this [5][6][7][8].
CFLAR L , also known as c-FLIP L , plays an important role in extrinsic ligand-induced apoptosis. In this pathway, FasL/tumor necrosis factor-alpha (TNF-α)/TRAIL binds to cell surface receptors, forms Death-Inducing Signaling Complex (DISC) and triggers the caspasedependent apoptotic pathway. CFLAR L interacts with CASP8 via DED domains and inhibits the activation of CASP8, and thus apoptosis [9][10][11][12]. ITCH is an E3 ligase of CFLAR L , which enhances apoptosis by targeting CFLAR L through ubiquitin-proteasome pathway [13,14]. USP8 is a de-ubiquitination enzyme of CFLAR L that enhances stability and inhibits apoptosis induced by extrinsic ligands [15]. Therefore, approaches to promote the ubiquitination and degradation of CFLAR L are potential effective cancer therapies.
In the present study, we showed GMEB1 directly interacted with CFLAR L and increased its stability at the protein level. GMEB1 inhibited the activation of pro-caspase 8 via CFLAR L . We found the de-ubiquitination enzyme USP40 bound to CFLAR L and GMEB, which enhanced the interaction between USP40 and CFLAR L, resulting in reduced ubiquitination and degradation of CFLAR L . These findings suggest GMEB1 inhibits the activation of CASP8 and apoptosis via CFLAR L , which highlights potential implications for lung cancer therapy.

Cell lines and cell culture
The human NSCLC cell lines A549, H1299, H1792. H157, H460, Calu-1 and HEK293FT cell lines were originally obtained from the American Type Culture Collection (ATCC). A549 and H1792 cell lines have been authenticated in Microread Gene Technology by STR analysis. The NSCLC cells were grown in monolayer culture in RPMI 1640 with 5% FBS (Gibco, US) at 37°C in a humidified atmosphere consisting of 5% CO 2 and 95% air. HEK293FT cells were grown in DMEM with 5% FBS (Gibco, US) at 37°C.
GMEB1 and USP40 coding regions were amplified by PCR from A549 genomic DNA using following primers: GMEB1 sense: CGGATCCGCCGCCACCATGGCTAA TGCAGAAGTGAG GMEB1 antisense: CCTCGAGTTAATCCTCTAAGAC CACAATC USP40 sense: CTAGCTAGCGCCGCCACCATGTCA CTTTTTTTAAGGGTAG USP40 antisense: CGCGGATCCTTATCTGAAGCTC CCCACG HA tag was cloned to the N-terminal of GMEB1, His tag was cloned to the C-terminal of USP40. His-tagged, FLAG-tagged and GST-tagged CFLAR L were cloned previously by our team.

Western blot analysis
Cells were harvested and rinsed with pre-chilled PBS on ice. They were lysed in lysis buffer on ice for 30 min and then purified via centrifugation for 15 min at 4°C. Samples of the whole-cell protein lysates (35 μg) were prepared for SDS-PAGE and transferred to a polyvinylidene fluoride (PVDF) membrane by electro blotting. The proteins were probed with the appropriate primary and secondary antibodies. Antibody binding was detected by an HRP system according to the manufacturer's protocol [16].

Immunoprecipitation
Cells were lysed in lysis buffer (20 mM Tris-HCl, pH 7.5; 150 mM NaCl; 1 mM Na 2 EDTA; 1 mM EGTA; 2.5 mM sodium pyrophosphate; 1 mM β-glycerophosphate; 1 mM Na 3 VO 4 ; 0.5% Triton) on ice for 30 min then purified via centrifugation for 15 min at 4°C. The supernatants were incubated with antibody at 4°C for 1 h. Then the mixture was incubated with protein A beads (ThermoFisher) at 4°C for 2 h. The beads were washed twice with 1 ml of lysis buffer. 20 μl 2 × SDS buffer were added for elution (100°C, 10 min). Samples were centrifuged for western blot analysis.

Immunofluorescence
Cells were fixed with PHEMO buffer (0.025 M HEPES, 0.068 M PIPES, 0.003 M MgCl 2 ·6H 2 O, 0.015 M EGTA·Na 2 , 10% DMSO, pH adjusted to 6.8. Additional reagents were added before use, with a final concentration as follows: 0.05% glutaraldehyde, 0.5% Triton X-100, 3.7% formaldehyde) for 10 min at room temperature before washing with PBS for 3 times. Then, cells were incubated with blocking buffer (3% BSA) for 30 min at room temperature. Afterward, cells were incubated with primary antibodies against CFLAR L (Santa Cruz, dilution at 1:500) for 1 h at room temperature. After washing with PBS for 3 times, cells were incubated with another primary antibodies GMEB1 (Santa Cruz, dilution at 1:500) or USP40 (Santa Cruz, dilution at 1:500) for 1 h. Alex Fluor 488 (Green) and Alex Fluor 568 (Red)-conjugated secondary antibodies were then applied and incubated at room temperature for 1 h. Cell nuclei were stained with DAPI. Images were captured using a confocal microscope (ZEISS, LSM700).

Flow cytometry analysis
Annexin V-FITC Apoptosis Detection Kit (Biobox Biotech, Nanjing, China) was used for cell apoptosis analysis according to the manufacture's protocol.

CASP8 activity detection
A549 cells were prepared for CASP8 activity according CASP8 Activity Apoptosis Assay Kit protocol (Sangon Biotech, Shanghai, China).

Tumor xenograft model
Fifteen female athymic nu/nu 4-week-old mice were purchased from Vital SPF Biotechnology (Beijing, China). For tumor xenograft establishments, A549 cells overexpressing green fluorescent protein or shGMEB1 RNA (1 × 10 6 cells/100 μL) were subcutaneously injected into the right side of the abdominal region of mice. Weight of mice and tumor size were detected every 2 days. The tumor volume was calculated as V = π × (length×width 2 )/6.

Statistical analysis
GraghPad Prism version 5.00 was used for statistical analysis. All data are presented as the mean ± SD. Differences between groups were identified using Student's t-test. P < 0.05 was considered statistically significant.
To detect whether SAHA also decreases GMEB1, we treated NSCLC cells with SAHA at different concentrations and time points. The results show SAHA decreased GMEB1 in a dose-dependent ( Fig. 1b and Additional file 1: Figure S1D) and time-dependent manner ( Fig. 1c and Additional file 1: Figure S1E). This is consistent with the effect of SAHA on CFLAR L . These data indicate SAHA treatment affects GMEB1 and CFLAR L similarly in NSCLC cells.

GMEB1 enhanced the stability of CFLAR L
We then characterized the biological function of the interaction between GMEB1 and CFLAR L . GMEB1 was originally identified as a transcription factor, so we first determined if GMEB1 regulates CFLAR L at the transcriptional level. GMEB1 and CFLAR L relative mRNA levels were detected using q-PCR in A549 cells; GMEB1 siRNA was transfected for 24 h. Results show that GMEB1 knock-down did not affect the relative mRNA level of CFLAR L (Fig. 2a). NSCLC cells with GMEB1 knock-down were treated with CHX [10 μg/ml] for various time points. WB data show GMEB1 knockdown decreased the stability of CFLAR L (Fig. 2b), while overexpression of GMEB1 increased the stability of CFLAR L (Fig. 2c). This confirms GMEB1 enhances the stability of CFLAR L at post-translational level. Next, we knocked down GMEB1 by siRNA in A549, H1299 and Calu-1 cell lines and treated cells with SAHA [2.0 μM] for 6 h. Results show GMEB1 knockdown decreased CFLAR L protein level (Fig. 2d). Overexpression of GMEB1 upregulated CFLAR L protein (Fig. 2e). To confirm the effect of GMEB1 on CFLAR L , we knocked down GMEB1 using GMEB1 shRNA in A549 cell lines and overexpressed GMEB1 using plasmid. We found that GMEB1 overexpression rescued the reduced CFLAR L protein level caused by GMEB1 knockdown (Fig. 2f). These data indicate GMEB1 plays a role in maintaining the protein level of CFLAR L .
Most protein degradation occurs in either the ubiquitination-proteasome or ubiquitination-lysosome pathway. As such, de-ubiquitinases inhibit the effective degradation of proteins. Therefore, we focused how GMEB1 affects the ubiquitination of CFLAR L . First, we knocked down GMEB1 in H1299 cells and treated them with DMSO, MG132 [20 μM] and E64D [15 μM] for 6 h. MG132 inhibits the degradation of proteins by blocking proteasomes, and E64D inhibits the degradation of proteins via lysosomes. Western blot analysis shows MG132 treatment rescued the reduced CFLAR L protein level caused by GMEB1 knockdown. This indicates CFLAR L is degraded through the proteasome pathway when GMEB1 protein levels are decreased (Fig. 2g). In addition, we designed a co-IP assay to determine whether GMEB1 affects the ubiquitination  (Fig. 2h). Thus, we propose GMEB1 enhances the stability of CFALR L by modulating its ubiquitination level.

GMEB1 physically interacted with CFLAR L in NSCLC cells
GMEB1 interacts with CASP8 and inhibits its activation. CFLAR L gene has high homology with CASP8 gene, and the proteins display similar structures that may confer interaction with each other through DED domains. Thus, we determined if GMEB1 and CFLAR L bind one another via a co-immunoprecipitation (co-IP) assay in HEK293FT cells. The data show that HA-tagged GMEB1 interacted with FLAG-tagged CFLAR L ( Fig. 3a and b). After GST-tagged CFLAR L was pulled down with Glutathione Sepharose beads, GMEB1 was detected using WB assay, indicating GMEB1 physically interacted with CFLAR L (Fig. 3c). An additional IP assay using A549 and H1299 cells (Fig. 3d) shows that endogenous CFLAR L interacted with endogenous GMEB1. To further evaluate the interaction between GMEB1 and CFLAR L , immunofluorescence staining experiments were conducted in Calu-1 cells. Results show GMEB1 localized in the cytosol. GMEB1 and CFLAR L were co-localized in the cytosol (Fig. 3e). We determined which domains of CFLAR L are required for this binding. Our data indicated that DED domains of CFLAR L were not necessary for interaction with GMEB1. However, P20 and P12 fragments of CFLAR L interacted with GMEB1 (Additional file 1: Figure S2A, B and C). Additional results show the N-terminal of GMEB1 was essential for interaction with CFLAR L (Additional file 1: Figure S2D and E). And, the fragment 325-573 of GMEB1, which doesn't interact with CFLAR L , didn't increase the protein level of CFLAR L in A549 cell lines.

USP40 interacted with GMEB1 and CFLAR L
We next aimed to identify the de-ubiquitination enzyme that targets CFLAR L . We examined protein level of The band intensity of CFLAR L was quantified by Photoshop CS6 and plotted. Experiments were repeated three times, a representative experiment is presented. Every experimental group was compared with the negative control group. Error bars represent s.d. *P < 0.05. d, e A549, H1299, Calu-1 and H1792 cells were seeded in 6-well plates. GMEB1 siRNA or plasmid were transfected for 24 h. Cells were treated with SAHA [2.0 μM] for 6 h and harvested for western blot analysis. f A549-LUC, A549-shGMEB1-1# and A549-shGMEB1-2# were seeded in 6-well plates. GMEB1 plasmid was transfected for 24 h. Cells were harvested for western blot analysis. g H1299 cells were seeded in 6-well plates. GMEB1 siRNA was transfected for 24 h. Cells were treated with DMSO, MG132 [20 μM] and E64D [15 μM] for 6 h. Cells were harvested for western blot analysis. h HEK293FT cells were prepared for GST pull down assay using GST-CFLAR L , FLAG-GMEB1 and HA-Ub plasmids to detect the function of GMEB1 on the ubiquitination of CFLAR L several de-ubiquitinases (USP4, USP40, USP7, USP8 and USP22) in six NSCLC cell lines and found USP40 was positively correlated with CFLAR L (Fig. 1a and Additional file 1: Figure S1B) and GMEB1 ( Fig. 1a and Additional file 1: Figure S1C). We proposed that USP40 modulated the ubiquitination of CFLAR L via its de-ubiquitinase activity. We then examined the interaction between GMEB1 and USP40 using Co-IP assay and found that tagged GMEB1 directly interacted with tagged USP40, which was confirmed by a reverse experiment (Additional file 1: Figure S3A and B). Given that de-ubiquitination enzymes interact with substrates, we then detected if USP40 directly interacted with CFLAR L using Co-IP assay in 293FT cells and found that tagged USP40 did in fact interact with tagged CFLAR L . This was also confirmed by a reverse experiment (Additional file 1: Figure S3C and D). IP assay using anti-FLIP L showed that endogenous CFLAR L interacted with endogenous USP40 in A549 and H1299 cells (Fig. 2d). Immunofluorescence staining results indicate that USP40 also co-localized with CFLAR L in cytosol (Additional file 1: Figure S3E). We next aimed to identify the domains of CFLAR L that were required for binding with USP40. Our data show that DED domains of CFLAR L did not interact with USP40. P20 and P12 fragments of CFLAR L interacted with USP40 (Additional file 1: Figures S2A, S3F and S3G).

USP40 targets CFLAR L for de-ubiquitination
Previous work showed ITCH targets CFLAR L as an E3 ligase, and USP8 is one of the de-ubiquitination enzymes of CFLAR L . Here, we found USP40 was positively correlated with and interacted with CFLAR L in NSCLC cells. Therefore, USP40 may target CFLAR L for de-ubiquitination. To confirm, we first evaluated the function of USP40 on CFLAR L in NSCLC cell lines treated with CHX. Results show that knocking down USP40 decreased the stability of CFLAR L (Fig. 4a). Furthermore, we knocked down USP40 in H1299, A549 and H157 cell lines using siRNA and treated cells with SAHA [2.0 μM] for 6 h. Data show protein level of CFLAR L decreased after knockdown of USP40 (Fig. 4b).
Conversely, overexpression of USP40 in H1299 and A549 cell lines increased the protein level of CFLAR L (Fig. 4c). These findings demonstrate USP40 promotes the stability of CFLAR L at the protein level. We used GST pull-down assays to determine the role of USP40 during CFALR L ubiquitination. Results show that overexpression of USP40 decreased the ubiquitination of CFLAR L (Fig. 4d), and knockdown USP40 increased the ubiquitination of CFLAR L (Fig. 4e). Additional results show that USP40 targeted CFLAR L for K48-linked de-ubiquitination (Fig. 4f). To confirm our proposal, we cloned mutant USP40 (C62A) (Additional file 1: Figure S4A) without enzyme activity [20]. Data show mutant USP40 (C62A) did not increase the protein level of CFLAR L in A549 cell lines (Additional file 1: Figure S4B), but it still interacted with CFLAR L (Additional file 1: Figure S4C). GST pull-down assay results show mutant USP40 (C62A) did not de-ubiquitinate CFLAR L (Additional file 1: Figure S4D). Co-IP assays were conducted in HEK293FT cells using FLAG-CFLAR L and HA-GMEB1 plasmids. c GST-pull down assay was conducted in A549 cells using GST-CFLAR L plasmids. d IP assays was conducted in A549 and H1299 cells using anti-FLIP L (Santa Cruz, US) antibody. e Calu-1 cells were fixed and subjected to indirect immunofluorescence staining with anti-FLIP L (Santa Cruz, US) and GMEB1 (Santa Cruz, US). The red signal (CFLAR L ) was obtained with anti-rabbit IgG Alexa 568-conjugated secondary Ab, and the green signals (GMEB1) were obtained with anti-mouse IgG Alexa 488-conjugated secondary Ab. Nuclei were stained with DAPI Taken together, these data suggest that USP40 is a de-ubiquitinase of CFLAR L .

GMEB1 promoted the interaction between USP40 and CFLAR L
Our findings show that GMEB1, USP40 and CFLAR L interact to form a complex. We looked into the function of GMEB1 in this complex. Data from co-IP assay in 293FT cells show that knock down of GMEB1 weakened the interaction between CFLAR L and USP40; overexpression of GMEB1 enhanced the interaction ( Fig. 5a  and b, respectively). In addition, we found that overexpression of HA-GMEB1 (325-573), which did not interact with CFLAR L , had no significant impact on the interaction between CFLAR L and USP40 (Fig. 5c). Our data indicate that GMEB1 acts as an adaptor protein in the complex, and GMEB1 is essential for the interaction between CFLAR L and GMEB1.
Further, we questioned whether GMEB1 affects CFLAR L via USP40. We knocked down USP40 and overexpressed GMEB1 in H460 and A549 cell lines and found no impact on CFLAR L (Fig. 5d). Using a reverse experiment in A549 cell lines, we showed that USP40 overexpression slightly increased CFLAR L protein when GMEB1 was knocked down (Fig. 5e). These data indicate GMEB1 confers stability to CFLAR L via USP40.
GMEB1 inhibited apoptosis via CFLAR L and DISC formation upon TRAIL exposure GMEB1 inhibits the activation of pro-caspases, but the molecular mechanism is still unclear. Our study demonstrates that GMEB1 interacts with CFLAR L and promotes The band intensity of CFLAR L was quantified by Photoshop CS6 and plotted. Experiments were repeated three times, a representative experiment is presented. Every experimental group was compared with the negative control group. Error bars represent s.d. *P < 0.05. b, c H1299, A549 and H157 cells were seeded in 6-well plates. USP40 siRNA or plasmid were transfected for 24 h. Cells were treated with SAHA [2.0 μM] for 6 h and harvested for western blot analysis. d, e HEK293FT cells were prepared for GST pull down assays using GST-CFLAR L , HIS-USP40 and HA-Ub plasmids or USP40 siRNA to detect the de-ubiquitination of USP40 targeted on CFLAR L . f HA-Ub-K48WT and HA-Ub-K63WT were cloned from HA-Ub wild type plasmid. Only the 48th or 63rd lysine was natural while all other lysine amino acids were mutated to alanine. GST pull down assay was conducted in HEK293FT cells and the de-ubiquitination model of USP40 targeted on CFLAR L was detected CFLAR L stability via the de-ubiquitination enzyme USP40. We asked if GMEB1 inhibits apoptosis via CFLAR L . GMEB1 siRNA was transfected in A549 and H1299 cell lines for 48 h, and then treated with TRAIL for 6 or 24 h. Western blot results show that GMEB1 knockdown increased the level of cleaved CASP8, CASP3 and PARP (Fig. 6a). Flow Cytometry analysis shows that GMEB1 knockdown enhanced apoptosis of A549 induced by TRAIL (Additional file 1: Figure S5B and C). Further, we transfected plasmid FLAG-CFLAR L while knocking down GMEB1 in A549 cells. Western blot results show that overexpression of CFLAR L decreased protein levels of cleaved CASP8, CASP3 and PARP that were induced by GMEB1 knockdown and TRAIL treatment (Additional file 1: Figure S5A). Flow Cytometry analysis show that CFLAR L overexpression partially attenuated apoptosis induced by GMEB1 knockdown, indicating GMEB1 inhibited apoptosis through regulating CFLAR L (Additional file 1: Figure S5D and E). GMEB1 knockdown increased the activation of CASP8 upon TRAIL treatment in A549 cells (Fig. 6b). And another similar experiment showed that CFLAR L overexpression partially attenuated the activation of CASP8 upon knocking down GMEB1 and TRAIL treatment (Fig. 6c). We then evaluated the formation of DISC induced by TRAIL treatment. Data show GMEB1 knockdown promoted the interaction of FADD and CASP8 (Fig. 6d and e).

CFLAR L is critical for the interaction of GMEB1 and CASP8
GMEB1 interacts with pro-caspase 8 and inhibits its activation, but the mechanism is not clear. Our experiments indicated that GMEB1 also interacted with CFLAR L , which interacted with pro-caspase 8 and inhibited its activity. We questioned which protein plays the dominant role in the interaction with GMEB1, CASP8 or CFLAR L . Co-IP assays were conducted to evaluate the interaction between GMEB1 and CFLAR L while CASP8 was knocked down using siRNA. Results show no significant change (Fig. 7a). We conducted another co-IP assay to evaluate the interaction between GMEB1 and CASP8 (using a plasmid HIS-CASP8M that 374 and 384 Aspartic acids mutated to Alanine acids) while  USP40 and CFLAR L . a, b GMEB1 siRNA or plasmid were co-transfected with FLAG-CFLAR L and HIS-USP40 in HEK293FT cells. Cells were harvested after 24 h for co-IP assay analysis. c HA-GMEB1-325-573 plasmid were co-transfected with FLAG-CFLAR L and HIS-USP40 in HEK293FT cells. Cells were harvested after 24 h for co-IP assay analysis. d H460 and A549 cells were seeded in 6-well plates. GMEB1 plasmid and USP40 siRNA were co-transfected for 24 h; non-sense siRNA and pcDNA3.1 were transfected as control. Cells were harvested for western blot analysis. e A549 cells were seeded in 6-well plates. USP40 plasmid and GMEB1 siRNA were co-transfected for 24 h; non-sense siRNA and pcDNA3.1 were transfected as control. Cells were harvested for western blot analysis CFLAR L was knocked down using siRNA. Results show CFLAR L knockdown decreased the interaction between GMEB1 and CASP8 (Fig. 7b). These results confirm CFLAR L directly interacts with GMEB1, and CFLAR L plays a critical role in the interaction of CASP8 and GMEB1.

Downregulation of GMEB1 inhibited A549 xenograft tumor growth in vivo
To evaluate whether the tumor growth of A549 is regulated by GMEB1 in vivo, A549-LUC, A549-shGMEB1-1# and A549-shGMEB1-2# were subcutaneously injected into the right side of the abdominal region of athymic nu/nu mice. GMEB1 and CFLAR L protein levels were detected in A549 cell lines (Fig. 8a). Mice weights were measured using an electronic balance, results show no significant difference among the three groups (Fig. 8b). Tumor sizes were measured with calipers (Fig. 8c). Results show tumor growth was inhibited after GMEB1 knockdown compared with the control group (Fig. 8d). And, tumor weights also support this finding (Fig. 8e). These results suggest GMEB1 plays a key role in cellular mechanisms related to apoptosis and cancer progress.

Discussion
GMEB1 is found throughout many cell types with multiple functions that are now being uncovered. GMEB1 and GMEB2 bind to the promoter sequence of the TAT gene and modulate GR transactivation. In the cytosol, GMEB1 interacts with HSP27, a protein chaperone with Fig. 6 GMEB1 inhibited the formation of DISC upon TRAIL activation. a A549 cells and H1299 cells were seeded in 12-well plates. GMEB1 siRNA was transfected for 24 h and a non-sense siRNA was transfected as negative control. Cells were then treated with TRAIL with different concentrations for 6 h and harvested for western blot analysis. b A549 cells were seeded in 96-well plates. GMEB1 siRNA was transfected for 48 h and a non-sense siRNA was transfected as negative control. Cells were then treated with TRAIL [40 ng/ml] for 6 h and prepared for CASP8 activity detection. c A549 cells were seeded in 6-well plates. GMEB1 siRNA and FLAG-CFLAR L plasmid were co-transfected for 48 h. Cells were then treated with TRAIL [40 ng/ml] for 6 h and prepared for CASP8 activity detection. d, e A549 cells were seeded in 96-well plates. GMEB1 siRNA was transfected for 48 h and a non-sense siRNA was transfected as negative control. Cells were then treated with TRAIL at different concentrations for 4 h and harvested for IP assay using FADD or CASP8 antibody Fig. 7 CFLAR L is critical for the interaction of GMEB1 and CASP8. a CASP8 siRNA were transfected in HEK293FT cells with FLAG-CFLAR L and HA-GMEB1 plasmids for a co-IP assay. b CFLAR L siRNA were transfected in HEK293FT cells with HIS-CASP8M and FLAG-GMEB1 plasmids for a co-IP assay Fig. 8 Downregulation of GMEB1 inhibited A549 xenograft tumor growth in vivo. a A549 cells stably overexpressing shGMEB1 RNA were seeded in 6-well plates and harvested after 24 h for western blot analysis. b A549 cells stably overexpressing shGMEB1 RNA were subcutaneously injected into the right side of the abdominal region of athymic nu/nu mice. Mice weights were measured using electronic balance every 2 days. c Tumor sizes were measured with calipers. The error bars represent the SD, ** P < 0.01, and *** P < 0.001. d Tumors were dissected and pictures taken. e Tumor weights were measured using electronic balance. The error bars represent the SD, ** P < 0.01, and *** P < 0.001 many critical functions in cancer invasion, metastasis, proliferation and apoptosis [21][22][23][24][25][26][27][28][29]. HSP27 is related to CFLAR L and the activation of caspases [30,31]. However, the relationship between GMEB1 and HSP27 has not been well studied. Recently, GMEB1 was reported to inhibit cell apoptosis by binding to pro-caspases and inhibiting their activation in cytosol [6]. GMEB1 interacts with pro-caspases via DED domains, but the molecular mechanism is unknown.
CFLAR L is an apoptotic inhibitor protein that interacts with pro-caspase 8 and inhibits its activation; DED domains play important roles in this interaction. The interaction between CFLAR L and pro-caspase 8 inhibits the formation of DISC [15]. As an inhibitor of histone deacetylase, SAHA is effective in treating skin T cell lymphoma clinically. It enhances the acetylation of Ku70 and disrupts the CFLAR/Ku70 complex and then triggers CFLAR poly-ubiquitination and degradation by the proteasome [32,33]. SAHA treatment in NSCLC cells shows that SAHA decreased CFLAR L and GMEB1 protein levels in a dose-dependent and a time-dependent manner. The results indicate GMEB1 and CFLAR L are positively correlated, which was confirmed by analyzing protein levels in six NSCLC cell lines and a HEK293FT cell line. In addition, GMEB1 protein reduction by SAHA suggests a relationship similar to SAHA and CFLAR L .
Given that GMEB1 acts in the nucleus as a transcription factor, we first conducted q-PCR experiments to check the function of GMEB1. We found GMEB1 affected the transcription of CFLAR L . We then turned our attention to the cytosol. Knock-down and overexpression experiments showed that GMEB1 promoted the stability and directly regulated the protein level of CFLAR L . Our in vitro studies indicated that GMEB1 interacts with CFLAR L outside the nucleus. DED domains of CFLAR L were not necessary for this process; P20 and P12 domains of CFLAR L accounted for the interaction with GMEB1. Our results also show that N-terminal of GMEB1 interacted with CFLAR L , and GMEB1 (325-573) fragments, which did not interact with CFLAR L , couldn't increase the protein level of CFLAR L . These findings suggest that the function of GMEB1 is dependent on the interaction with CFLAR L . Our experiments also indicate GMEB1 affects the degradation of CFLAR L through proteasome pathway. GMEB1 doesn't have ubiquitin-related enzymatic activity that directly modulates the ubiquitination of CFLAR L . This suggests that another enzyme is needed in this process. Therefore, USP40 was emphasized because its expression was positively correlated with GMEB1 and CFLAR L in NSCLC cells.
CFLAR L is an effective target for cancer therapy [9][10][11][12]. ITCH, which has important roles in cell immune regulation [34], was identified as an E3 ligase targeting CFLAR L . It induces apoptosis by degrading CFLAR L and activating pro-caspase 8. De-ubiquitination is an important protein modification that reverses the ubiquitination of proteins via E1/E2/E3 ligase. De-ubiquitination inhibits the degradation of proteins and drives the fate of substrate proteins [35]. USP8 regulates the morphology of the endosome by ubiquitinating proteins and is also involved in cargo sorting and membrane trafficking at the early endosome stage [36,37]. USP8 is a de-ubiquitination enzyme of CFLAR L that promotes the stability of CFLAR L and inhibits the activation of pro-caspase 8. Changes in CFLAR L protein levels also affect the formation of DISC and apoptosis induced by extrinsic ligands.
To find the de-ubiquitinase that regulates CFLAR L by GMEB1, we focused on USP8 that was identified as a de-ubiquitinase of CFLAR L . But our experiments showed that GMEB1 did not interact with USP8 (Additional file 1: Figure S4E). Western blot data from six NSCLC lines suggested USP40 protein levels are positively correlated with GMEB1 and CFLAR L . This underscores the importance of USP40.
Several reports show that USP40 is correlated with late-onset Parkinson's disease and USP24 [38,39]. In addition, USP40 affects glomerular permeability in zebrafish [40]. Our results suggest USP40 interacts with both GMEB1 and CFLAR L . GMEB1 promoted the binding of USP40 with CFLAR L , conferring a role as an adaptor protein. Consequently, USP40 augmented the stability of CFLAR L via its de-ubiquitinase activity. USP40 targeted CFLAR L for K48-linked de-ubiquintion. USP40 knockdown did not increase the protein level of CFLAR L even though GMEB1 protein level was overexpressed. And, overexpression of USP40 increased the protein level of CFLAR L though GMEB1 was knocked down. Our data indicate that USP40 is the key protein affecting GMEB1 on CFLAR L .
In addition, we also found that GMEB1 knockdown promoted the activation of pro-caspase 8 and apoptosis induced by TRAIL. CFLAR L attenuated apoptosis that was induced by GMEB1 knockdown, which highlights the function of GMEB1 in inhibiting apoptosis via CFLAR L . GMEB1 inhibited the formation of DISC upon TRAIL activation.
Previous studies show GMEB1 interacts with the DED domain of CASP8 and inhibits its activation. Our results show GMEB1 did not interact with DED domain of CFLAR L which has a similar structure with CASP8. However, GMEB1 interacted with CFLAR L via the P20 and P12 domains. Co-IP results showed that CASP8 knockdown didn't affect the interaction between GMEB1 and CFLAR L , while CFLAR L knockdown decreased the interaction between GMEB1 and CASP8. These findings suggest that CFLAR L is crucial for the interaction between GMEB1 and CASP8. Thus, GMEB1 interacts with CFLAR L , and CFLAR L interacts with CASP8 via DED domains. GMEB1 inhibits the activation of CASP8 via the function of CFLAR L .
In vivo data showed GMEB1 knockdown inhibited the A549 xenograft tumor growth, which also confirmed our results.

Conclusions
We described the interaction among GMEB1, USP40 and CFLAR L (Fig. 9). We found that GMEB1 promoted the stability of CFLAR L by de-ubiquitinase USP40. USP40 targeted CFLAR L for K48-linked de-ubiquitination. GMEB1 inhibited the activation of CASP8 and apoptosis in NSCLC via CFLAR L . CFLAR L promoted the interaction between GMEB1 and CASP8. GMEB1 knockdown inhibited tumor growth in vivo. These findings provide more in-depth knowledge that serves as potential therapies for cancer.

Additional file
Additional file 1: Figure S1. CFLAR L protein level positively correlated with GMEB1 and USP40 protein levels in NSCLC cell lines. Figure S2. CFLAR L directly interact with GMEB1. Figure S3. USP40 interacted with GMEB1 and CFLAR L . Figure S4. USP40 targeted CFLAR L for de-ubiquitination.