Elevated TMEM120B expression in breast cancer correlated with advanced TNM stage, positive lymph node metastasis, and poor prognosis

First, we used The Cancer Genome Atlas (TCGA) database to explore the mRNA expression of TMEM120B in pan-cancer and normal tissues and found that TMEM120B was highly expressed in most cancerous tissues than in non-cancerous ones; however, breast cancer was excluded (Fig. 1A). Subsequent immunohistochemical (IHC) staining of 20 cases of lung cancer, 29 cases of breast cancer, 21 cases of gastric carcinoma, 24 cases of colon cancer, 20 cases of ovarian cancer, and paired normal tissues suggested that TMEM120B expression was elevated in all malignant cancerous tissues compared with that in normal tissues (Fig. 1B:a-j and Table 1). Moreover, we examined TMEM120B mRNA expression specifically in breast cancer based on the gene expression profiling interactive analysis (GEPIA) database and found that TMEM120B mRNA expression was significantly higher in breast cancer than in normal tissues, both in paired and unpaired breast cancer specimens (Fig. 1C and D). Next, we performed IHC staining of 140 breast cancer and 42 normal samples and found that the positive rate of TMEM120B in breast cancer tissues(50.7%,71/140) was significantly higher than that in the paired non-cancerous tissues (23.8%,10/42, P < 0.001,Fig. 1E: a-e), with a cytosolic positive expression rate of 47.1%(66/140) (Fig. 1E: a-e), and a nuclear expression rate of 3.5% (5/140; Additional file 3: Fig. S1A). Subsequent statistical analysis indicated that total and cytosolic TMEM120B expression positively correlated with advanced TNM stage (P = 0.011,P = 0.015) and lymph node metastasis (P = 0.006,P = 0.007) but not with age or triple-negative breast cancer, however, nuclear TMEM120B revealed no visible correlation with clinicopathologic factors(P > 0.05, Table 2). Bioinformatics analysis also suggested that TMEM120B expression was higher in patients with advanced SBR grade and positive distant metastasis than lower grade and negative groups (Additional file 3: Fig. S1B-C). Whereas it reveal no obvious differences for TMEM120B RNA among diverse subtypes of breast cancer(Additional file 3: Fig.S1D).Kaplan–Meier analysis revealed that both TMEM120B mRNA and protein levels were higher in patients with a poor prognosis than better ones (P = 0.095 and P = 0.079, Fig. 1F-G). However, Cox univariate analysis revealed that TMEM120B expression was not an independent prognostic factor in patients with breast cancer (Table 3). Additionally, western blotting was performed on fresh breast cancer and paired normal tissue samples. The results suggested that TMEM120B protein levels were higher in cancerous samples than in non-cancerous samples (Fig. 1H and Additional file 3: Fig. S1E), which was consistent with the IHC staining results.

Fig. 1

TMEM120B was highly expressed in breast cancer specimens and cell lines. (A) TCGA database was assessed to explore the mRNA expression of TMEM120B in pan-cancer and normal tissues, N for Normal, T for Tumor, Meta for metastasis. (B) Representative images of immunohistochemistry staining of TMEM120B in normal breast epithelial cells (a), normal intestinal epithelial cells (c), normal gastric epithelial cells (e), normal lung epithelial cells(g), normal ovarian epithelial cells (i) breast cancer epithelial cells (b), colon cancer epithelial cells (d), gastric carcinoma epithelial cells (f), lung cancer epithelial cells (h) and ovarian cancer epithelial cells (j), N for Normal, T for Tumor. (C-D) TMEM120B mRNA levels were identified between non-cancerous and cancerous tissues using the TCGA database (E) Representative images of immunohistochemistry staining of TMEM120B in (a) both normal and cancerous tissues in the same specimen, (b) adjacent normal tissue and breast cancer with diverse staining (c, weak, d, moderate, e, strong), N for Normal, T for Tumor. (F) Kaplan–Meier curves showed a correlation between mRNA expression of TMEM120B and overall survival in breast cancer patients. (G) Kaplan–Meier curves showing a correlation between TMEM120B protein expression and overall survival of patients with breast cancer. (H) TMEM120B protein level in 16 pairs of freshly isolated samples from patients with breast cancer was analyzed by western blotting (I) The protein expression of TMEM120B in breast cancer cell lines and normal breast cells. (J) Immunofluorescence assay was used to evaluate the subcellular localization of TMEM120B in breast cancer cells (scale bar = 20 μm). Quantification data are expressed as mean ± SD of three independent experiments (t-test, two-sided, *P < 0.05, **P < 0.01, ***P < 0.001)

Table 1 Expression of TMEM120B in various epithelial malignanciesTable 2 Correlation of the expression of TMEM120B with clinicopathological features in 140 cases of breast cancerTable 3 Summary of Cox univariate and multivariate regression analysis of the association between clinicpathological features and overall survival in 140 cases of breast cancer

We examined the expression and subcellular localization of TMEM120B in seven breast cancer cell lines and one normal breast cancer cell line (MCF-10 A). Western blot analysis indicated that TMEM120B expression was higher in all detected breast cancer cell lines than in MCF-10 A cells (Fig. 1I). Immunofluorescence staining also indicated that TMEM120B exhibited both cytosolic and nuclear localization in breast cancer cell lines (Fig. 1J).

Overexpression of TMEM120B accelerated breast cancer proliferation and invasion both in vitro and in vivo

We first overexpressed TMEM120B in both MCF-7 and SK-BR-3 cells or deleted TMEM120B with two different sgRNAs using CRISPR-Cas9 in MDA-231 and MDA-453 cells (Additional file 3: Fig. S2A). The results of the MTT assay (Fig. 2A and Additional File 3: Fig. S2B), colony formation (Fig. 2B; Additional file 3: Fig. S2C), and EdU (Fig. 2C and Additional file 3: Fig. S2D) assays indicated that the overexpression or deletion of TMEM120B may promote or abrogate MCF-7, SK-BR-3,MDA-231 and MDA-453 cells proliferation, respectively. Transwell assay (Fig. 2D and Additional file 3: Fig. S2E) and wound healing (Fig. 2E and Additional file 3: Fig S2F) assays also revealed that migration and invasion were enhanced or suppressed by the overexpression or inhibition of TMEM120B. Additionally, xenograft assays revealed that tumor volumes significantly increased in SK-BR-3 cells overexpressing TMEM120B (Fig. 2F), whereas the number of lung metastases visibly increased in the ectopic TMEM120B group, however there were no obvious changes in metastasis of liver, brain, kidney and heart (Fig. 2G, Additional file 3: Fig. S2G).

Fig. 2

Overexpression of TMEM120B promoted breast cancer cell proliferation and invasion both in vitro and in vivo. The MTT assay (A), colony formation assay (B), and EdU assay (C, scale bar = 100 μm) were performed to examine the effects on the proliferation of after overexpressing or silencing TMEM120B in SK-BR-3 or MDA-231 cells. Transwell (D) and wound healing (E) assays were used to assess the effects of TMEM120B-myc, TMEM120B sgRNA, and the control on cell invasion and migration in SK-BR-3 and MDA-231 cells. Representative examples of explanted tumors (F) and lung metastases (G) in the negative SK-BR-3-control (NC) and SK-BR-3- overexpressing TMEM120B groups. Quantification data are expressed as mean ± SD of three independent experiments (t-test, two-sided, *P < 0.05, **P < 0.01, ***P < 0.001)

Overexpression of TMEM120B enhanced stemness of breast cancer cells

Next, we explored the mechanism by which TMEM120B promoted breast cancer cell proliferation, invasion, and metastasis. RNA-sequencing was performed to identify the differentially expressed genes (DEGs) after deleting TMEM120B in MDA-453 cells, revealing 793 DEGs (fold change > 1.5), among which 687 genes were downregulated, and 106 genes were upregulated (PRJNA938979, https://www.ncbi.nlm.nih.gov/bioproject/PRJNA938979/; Additioanl file 4). Gene ontology (GO) analysis indicated that the DEGs were enriched in the following processes: signaling pathways regulating the pluripotency of stem cells, cell cycle, regulation of actin cytoskeleton, and focal adhesion (Fig. 3A). Previous studies have demonstrated that enhancing the stemness of breast cancer cells may accelerate their proliferation, invasion, and metastasis [47]. Interestingly, bioinformatics analysis indicated that TMEM120B expression was significantly positively correlated with breast cancer stemness (Fig. 3B). Western blotting indicated that the expression of pivotal breast CSC markers, such as ALDH1, OCT4, Nanog, and Sox2, was increased or decreased upon TMEM120B overexpression or inhibition (Fig. 3C and Additional file 3: Fig. S3A). Both the first and second rounds of sphere formation assays suggested that the stemness of breast cancer cells was enhanced or abrogated after the overexpression or depletion of TMEM120B in MCF-7 and SK-BR-3 or MDA-231 and MDA-453 cells (Fig. 3D-E and Additional file 3: Fig. S3B-C), respectively. Moreover, we isolated both sphere and adhesive MDA-231 cells and detected the expression of stem cell markers, which indicated that TMEM120B and stem cell markers were elevated in sphere cells compared with those in adhesive cells (Fig. 3F). Flow cytometry also indicated that the ratio of ALDH1-positive cells increased or decreased upon the overexpression or deletion of TMEM120B in SK-BR-3 or MDA-231 cells, respectively (Fig. 3G). Further, we assessed the effects of limiting dilution xenograft on breast cancer stemness upon TMEM120B depletion in MDA-MB-231 cells in vivo, which also indicated that overexpression of TMEM120B may strengthen breast cancer stemness (Fig. 3H).

Fig. 3

Overexpression of TMEM120B enhanced stemness of breast cancer cells. (A) GO analysis was performed to detect the biological process significantly correlated with the deletion of TMEM120B in MDA-453 cells. (B) Bioinformatics analysis for the mRNAsi Stemness score of TMEM120B (C) Immunoblotting of Myc-tag, ALDH1, OCT4, NANOG, SOX2, and GAPDH after overexpressing or deleting TMEM120B in SK-BR-3 and MDA-231 cells. Both the first (D, scale bar = 250 μm) and second round of sphere formation assays (E) were performed to examine the effects on stemness of cells after overexpressing or knocking out TMEM120B in SK-BR-3 or MDA-231 cells. (F) Immunoblotting of Myc-tag, OCT4, NANOG, ALDH1, SOX2, and GAPDH in MDA-231 cells in both non-sphere and sphere groups. (G) Flow cytometry assay detected the ratio of ALDH1+ cells upon overexpression or deletion of TMEM120B in SK-BR-3 or MDA-231 cells (H) Diluted injection xenograft assays to explore the effects on stemness of breast cancer cells upon depleting TMEM120B in MDA-231 cells in vivo. Quantification data are expressed as mean ± SD of three independent experiments (t-test, two-sided, *P < 0.05, **P < 0.01, ***P < 0.001)

Overexpression of TMEM120B strengthened breast cancer stemness via TAZ-mTOR signaling axis

We explored the signaling pathway transduction upon TMEM120B overexpression to maintain the stemness of breast cancer cells. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of the differential genes from RNA-sequencing after deleting TMEM120B in MDA-453 cells indicated that Hippo, Wnt, AKT/mTOR signaling pathways were involved (Fig. 4A). Gene set enrichment analysis (GSEA) also indicated that mTOR signaling was positively correlated with higher TMEM120B expression (Additional file 3: Fig. S4A). We assessed the phosphorylation antibody array to screen the signaling pathways involved in TMEM120B overexpression, which indicated that the phosphorylation of AKT at serine 473 was increased; however, the phosphorylation of β-catenin, which is crucial for Wnt signal transduction, and other signaling pathways, was not altered (Fig. 4B). Hippo and PI3K-AKT signaling were chosen for further studies as they are both proven to be crucial in maintaining stemness of breast cancer cells [48,49,50].Subsequent western blotting indicated that levels of phosphorylated AKT, phosphorylated mTOR, and TAZ were significantly increased or decreased, and YAP level was slightly increased or decreased after the overexpression or depletion of TMEM120B in SK-BR-3 or MDA-231 cells, respectively (Fig. 4C). Additionally, qPCR was used to examine the expression of CYR61 and CTGF, the downstream target genes of YAP and TAZ, and we found that CYR61 and CTGF were upregulated or downregulated upon TMEM120B overexpression or deletion in SK-BR-3 or MDA-231 cells, respectively (Fig. 4D). Previous studies have demonstrated that both Hippo and AKT/mTOR signaling pathways are involved in maintaining breast cancer stemness [51,52,53]. Next, we assessed the specific mTOR inhibitor rapamycin or silenced TAZ with siRNA to clarify whether the enhancement of breast cancer stemness induced by overexpressing TMEM120B was dependent on the activation of both signaling pathways.Western blotting results revealed that adding 2 nmol rapamycin for 48 h or inhibiting TAZ may counteract the increase in the proliferation, invasion, and stemness of SK-BR-3 cells induced by TMEM120B overexpression (Additional file 3: Fig. S4B-G). Moreover, inhibiting mTOR signaling did not block the increase in TAZ level, whereas silencing TAZ using siRNA neutralized the elevation of phosphorylated mTOR (Fig. 4E-F), indicating that TAZ may act upstream of mTOR upon overexpression of TMEM120B.

Fig. 4

Overexpressing TMEM120 accelerated breast cancer stemness by activating TAZ-mTOR signaling axis. (A) KEGG analysis was conducted to detect the signaling pathway significantly correlated with deletion of TMEM120B in MDA-453 cells. (B) Phosphorylation antibodies array kit was used to explore the key signaling pathway involved in TMEM120B overexpression in SK-BR-3 cells. (C) Immunoblotting of Myc-tag, TMEM120B, AKT, p-AKT, mTOR, p-mTOR, YAP, TAZ, and GAPDH after overexpressing or silencing TMEM120B in SK-BR-3 or MDA-231 cells. (D) qPCR assay was used to investigate the alteration of the target genes of YAP/TAZ within ectopic or deleted TMEM120B in SK-BR-3 or MDA-231 cells. (E) Immunoblotting of Myc-tag, mTOR, p-mTOR, TAZ, and GAPDH after overexpressing TMEM120 with or without mTOR signaling pathway inhibitor rapamycin in SK-BR-3 cells. (F) Immunoblotting of Myc-tag, mTOR, p-mTOR, TAZ, and GAPDH after overexpressing TMEM120 with or without knocking down TAZ by siRNA in SK-BR-3 cells. Subcellular localization of TAZ was evaluated by immunofluorescence assay (G) or western blotting assay (H) within ectopic TMEM120B in SK-BR-3 cells. Scale bar = 20 μm. (I) After being treated with CHX at indicated time points, the expression of TAZ was evaluated by western blotting after overexpressing or silencing TMEM120B in SK-BR-3 or MDA-231 cells. Quantification data are expressed as mean ± SD of three independent experiments (t-test, two-sided, ***, P < 0.001)

We investigated the mechanism by which TMEM120B overexpression activates TAZ. The qPCR results indicated that overexpression or silencing of TMEM120B in SK-BR-3 or MDA-231 cells did not alter the mRNA expression of TAZ (Additional file 3: Fig. S4H). However, both immunofluorescence staining and western blotting indicated that nuclear TAZ expression increased or decreased after the overexpression or deletion of TMEM120B, respectively (Fig. 4G-H). Translocation of TAZ controls its stability, which is crucial for Hippo signaling pathway transduction [18]. We used cycloheximide (CHX) to block de novo protein synthesis and investigated the effect of TMEM120B overexpression on the stability of TAZ. The results revealed that TAZ degradation was blocked or accelerated upon TMEM120B overexpression or deletion in SK-BR-3 or MDA-231 cells, respectively (Fig. 4I).

TMEM120B directly bound to MYH9 through their coil-coil domains and maintained breast cancer stemness

An endogenous co-IP assay indicated that TMEM120B did not interact with TAZ, which revealed that TAZ stabilization by overexpression may not result from direct interactions (Additional file 3: Fig. S5A). Mass spectrometry (MS) was performed to identify candidates for interaction with TMEM120B. We found 316 potential binding proteins, as shown in Fig. 5A and Additional file 5. MYH9 and ACTN4 were used to test their interactions with TMEM120B, as they were in the top 10. Moreover, both are involved in modulating the cell actin cytoskeleton, which tightly modulates TAZ by affecting mechanical force [54, 55]. An endogenous co-IP assay indicated that MYH9, rather than ACTN4, interacts with TMEM120B in MDA-231 cells (Fig. 5B and Additional file 3: Fig. S5B). Subsequent co-IP revealed that exogenous TMEM120B interacted with MYH9 in SK-BR-3 cells (Fig. 5C). The GST pull-down assay showed that TMEM120B directly binds to MYH9 (Fig. 5D). Immunofluorescence assays suggested that TMEM120B co-localized with MYH9 in the cytoplasm of MDA-231 cells, which was quantified using the ImageJ software (Fig. 5E, R = 0.77). Next, we tested whether enhanced stemness induced by TMEM120B overexpression was dependent on MYH9. We transfected TMEM120B plasmids and MYH9-siRNA in SK-BR-3 cells; western blotting results indicated that elevated TAZ, phosphorylated mTOR, and ALDH1 were blocked by MYH9 silencing (Additional file 3: Fig. S5C). Additionally, MYH9 inhibition suppressed SK-BR-3 cells proliferation, invasion, and stemness(Fig. S5D-F). Divergent TMEM120B and MYH9 splice-mutant plasmids were designed to determine the exact domain responsible for the interaction between TMEM120B and MYH9 (Fig. 5F). The co-IP assay results suggested that the coil-coil domains in both TMEM120B and MYH9 dominated their binding in SK-BR-3 cells (Fig. 5G-H). GST pull-down assays also revealed that the coil-coil domain was crucial for TMEM120B–MYH9 binding (Fig. 5I). Overexpression of TMEM120B-∆CCD may neutralize the elevation of TAZ, phosphorylated mTOR, and the enhancement of proliferation, invasion, and stemness of SK-BR-3 cells both in vitro and in vivo (Fig. 5J-Q).

Fig. 5

TMEM120B promoted breast cancer cell stemness by binding with MYH9 via their coil-coil domains. (A) Mass spectrometry (MS) analysis was performed to identify candidates for interaction with TMEM120B after overexpressing TMEM120B in SK-BR-3 cells. Endogenous (B) and exogenous co-IP assay (C) were assessed to detect the interaction between MYH9 and TMEM120B in MDA-231 and SK-BR-3 cells. (D) GST pull-down assay was used to confirm the direct binding between MYH9 and TMEM120B in SK-BR-3 cells. (E) Immunofluorescence assay was used to show the co-localization of TMEM120B and MYH9 in SK-BR-3 cells, subcellular location coefficient of TMEM120B–MYH9 interaction was quantified by Fiji software (scale bar = 50 μm) (F) Divergent TMEM120B and MYH9 splicing mutant plasmids were designed to examine the domain responsible for the interaction between TMEM120B and MYH9.(I) GST pull-down assay was performed to confirm the direct interaction between MYH9 and TMEM120B after overexpressing TMEM120B-WT or TMEM120B-∆CCD plasmids in SK-BR-3 cells. Colony formation assay (J), Transwell assay (K), and sphere formation assay (L) were performed to detect the effects on the proliferation, invasion, and stemness of breast cancer cells after transfecting TMEM120B, TMEM120B-∆CCD, and control plasmid in SK-BR-3 cells. (M) Immunoblotting of Myc-tag, mTOR, p-mTOR, TAZ, and GAPDH after overexpressing TMEM120B or TMEM120B-∆CCD in SK-BR-3 cells. (N) Immunoblotting was used to evaluate the expression of cytosolic or nuclear TAZ after overexpressing TMEM120B or TMEM120B-∆CCD in SK-BR-3 cells. The effects on proliferation, metastasis, and stemness were verified in nude mice by subcutaneous tumorigenesis (O), tail vein injection (P), and diluted injection xenografts assays (Q) by overexpressing TMEM120B and TMEM120B-∆CCD in SK-BR-3 cells. Quantification data are expressed as mean ± SD of three independent experiments (t-test, two-sided, **, P < 0.01, ***, P < 0.001)

TMEM120B stabilized MYH9 by preventing its ubiquitination to CUL9 in a competitive manner

MYH9 protein levels, rather than mRNA levels, were significantly increased or decreased when TMEM120B was overexpressed or deleted in SK-BR-3 or MDA-231 cells, respectively (Fig. 6A-B). However, the mRNA and protein levels of TMEM120B were not altered by MYH9 overexpression or silencing in SK-BR-3 or MDA-231 cells (Fig. 6C-D). CHX was used to block the de novo protein synthesis of MYH9, which revealed that overexpression or deletion of TMEM120B may abolish or enhance the degradation of MYH9 in SK-BR-3 cells or MDA-231 cells (Fig. 6E). MYH9 is unstable and can be degraded via proteasome-dependent ubiquitination [35]. Our results, which are consistent with those of previous studies, indicated that MYH9 ubiquitination was abrogated or enhanced after the overexpression or depletion of TMEM120B in SK-BR-3 or MDA-231 cells, respectively (Fig. 6F and Additional file 3: Fig. S6A-B). Intriguingly, overexpressing TMEM120B-∆CCD may counteract the decreased ubiquitination induced by overexpressing TMEM120B in SK-BR-3 cells (Fig. 6F). TMEM120B has not been reported to function as an E3-ubiquitin ligase, which indicated that the TMEM120B–MYH9 interaction may prevent the degradation of MYH9 by a certain E3-ubiquitin ligase. MS analysis was performed to identify the candidate E3-ubiquitin ligases after overexpressing TMEM120B and TMEM120B-∆CCD in SK-BR-3 cells (Fig. 6G). Among the 141 proteins screened in the TMEM120B group (but not in the TMEM120B-∆CCD group), we identified CUL9, an E3-ubiquitin ligase and E3-adaptor proteins RACK1 and UBD protein BRSK2 (Additional file 6). Interestingly, MYH9 remained on the candidates list, confirming our previous co-IP results. Co-IP assays indicated that TMEM120B, MYH9, and CUL9 formed ternary complexes in MDA-231 cells (Fig. 6H). Overexpression of TMEM120B, together with CUL9, abolished the elevation of MYH9 and the reduction in MYH9 ubiquitination (Fig. 6I and J). Overexpression of TMEM120B and CUL9 may counteract the elevated proliferation, invasion, and stemness induced by TMEM120B overexpression in SK-BR-3 cells (Additional file 3: Fig. S6C-E). Further, the interaction between TMEM120B and MYH9 was blocked by CUL9 overexpression in a dose-dependent manner, indicating that TMEM120B competitively binds to MYH9 from CUL9, thus preventing the degradation of MYH9 (Fig. 6K).

Fig. 6

TMEM120B stabilized MYH9 by preventing its ubiquitin-mediated degradation from CUL9. (A and B) Western blotting and qPCR assays were performed to determine the protein and mRNA expression of MYH9, respectively, after overexpressing or deleting TMEM120B in SK-BR-3 or MDA-231 cells. (C-D) Western blotting and qPCR assays were performed to assess the protein and mRNA expression of TMEM120B, respectively, after overexpressing or deleting MYH9 in SK-BR-3 or MDA-231 cells. (E) After being treated with CHX at indicated time points, the expression of MYH9 was evaluated by western blotting after overexpressing or silencing TMEM120B in SK-BR-3 or MDA-231 cells. (F) The ubiquitination level of MYH9 was detected using western blotting after being transfected with TMEM120B, TMEM120B-∆CCD, and control plasmids in SK-BR-3 cells. (G) Mass spectrometry (MS) analysis was performed to identify candidates for interaction with ectopic TMEM120B or TMEM120B-∆CCD in SK-BR-3 cells, respectively. (H) Endogenous co-IP assay was performed to detect the interaction between MYH9, CUL9, and TMEM120B in MDA-231 cells. (I and J) Protein levels of MYH9 and the ubiquitination level were assessed using western blotting after transfecting TMEM120B alone or co-transfecting both TMEM120B and CUL9 in SK-BR-3 cells. (K) Co-IP assay was used to evaluate the interaction among TMEM120B, MYH9, and CUL9 after overexpressing TMEM120B and CUL9 in different doses in SK-BR-3 cells. Quantification data are expressed as mean ± SD of three independent experiments (t-test, two-sided)

TMEM120B–MYH9 interaction enhanced breast cancer stemness via β1-integrin-FAK-TAZ-mTOR axis

MYH9 is responsible for the formation of FAs, thereby promoting colon cancer progression [35]. We performed GO analysis of TMEM120B interaction candidates from the MS analysis in SK-BR-3 cells and found that they were enriched in the modulation of the cell skeleton and focal adhesion (Fig. 7A), which was consistent with the RNA-seq results (Fig. 7B). Tang et al. showed that the integrin-FAK signaling axis is involved in mechanical force transduction and accelerates the nuclear translocation of YAP/TAZ [26]. First, a 3D collagen gel invasion assay was performed to evaluate whether TMEM120B is involved in mechanical force transduction. The results showed that overexpression of TMEM120B, rather than TMEM120B-∆CCD, enhanced 3D invasion in MDA-231 cells (Fig. 7C). Western blotting assay indicated that overexpression of TMEM120B increased phosphorylated FAK levels in Tyr 397 and active-β1-integrin; however, ectopic TMEM120B-∆CCD did not (Fig. 7D). PF562271, a FAK signaling-specific inhibitor, was applied after overexpressing TMEM120B in SK-BR-3 cells, and western blotting results revealed that the elevation of TAZ and phosphorylated mTOR was blocked (Fig. 7E). SK-BR-3 cells proliferation, invasion, and stemness were abrogated by FAK inhibition (Fig. 7F-H). FA assembly can be modulated by recycling integrins from the membrane to the cytosol upon activation of mechanical force signaling [56]. FA assembly was evaluated following NZ treatment. After treatment with NZ for 30 min, no obvious membranous integrin or phosphorylated FAK was observed (Fig. 7I and J). However, overexpression of TMEM120B, rather than TMEM120B-∆CCD in MDA-231 cells, accelerated membrane-expressed β1-integrin and phosphorylated FAK after treatment with NZ for 60 min (Fig. 7I and J). Finally, we overexpressed TMEM120B, TMEM120B-∆CCD, MYH9, MYH9-∆CCD alone or co-transfected TMEM120B + MYH9, TMEM120B-∆CCD + MYH9, and TMEM120B + MYH9-∆CCD into breast cancer cells. Western blotting results indicated that co-transfected TMEM120B + MYH9, rather than TMEM120B-∆CCD + MYH9 or TMEM120B + MYH9-∆CCD, significantly increased the expression of phosphorylated FAK and mTOR, TAZ, and ALDH1 (Fig. 7K). Flow cytometry also indicated that the ratio of ALDH1-positive cells elevated more than the other groups upon co-transfecting TMEM120B + MYH9 (Additional file 3: Fig. S6F).Our results indicated that the TMEM120B–MYH9 interaction may enhance breast cancer stemness by activating the β1-integrin-FAK-TAZ-mTOR signaling axis.

Fig. 7

TMEM120B–MYH9 interaction activated the TAZ-mTOR axis by accelerating FAK assembly. (A) GO analysis for TMEM120B interaction candidates from MS analysis after overexpressing TMEM120B in SK-BR-3 cells. (B) Venn analysis for the overlap between RNA-seq and MS analysis. (C) 3D collagen gel invasion assay was performed after overexpressing TMEM120B or TMEM120B-∆CCD and control in MDA-231 cells. (D) Immunoblotting assay was performed to evaluate the expression of Myc-tag, FAK, p-FAK (Tyr397), β1-integrin, active-β1-integrin, and LaminB1 after transfecting TMEM120B-myc, TMEM120B-∆CCD-myc, and control plasmids in SK-BR-3 cells. (E) Immunoblotting of Myc-tag, p-mTOR, TAZ, and GAPDH after overexpressing TMEM120B with or without FAK signaling pathway inhibitor PF562271 in SK-BR-3 cells. Transwell assay (F), sphere formation assay (G), and colony formation assay (H) were performed to detect the effect on the invasion, stemness, and proliferation of breast cancer cells upon ectopic TMEM120B or ∆TMEM120B-CCD in SK-BR-3 cells. Representative immunofluorescence images of p-FAK (I) and β1-integrin (J) after treatment with nocodazole (NZ), followed by washout for 0, 30, and 60 min. Scale bar = 10 μm. (K) Immunoblotting of Myc-tag, Flag-tag, FAK, p-FAK(Tyr397), mTOR, p-mTOR, TAZ, ALDH1, and GAPDH after transfected with TMEM120B-myc, TMEM120B-∆CCD-myc, MYH9-flag, MYH9-delCCD-flag alone, or TMEM120B-myc + MYH9-flag, TMEM120B-∆CCD-myc + MYH9-flag, or TMEM120B-myc + MYH9-delCCD-flag in SK-BR-3 cells, respectively. Quantification data are expressed as mean ± SD of three independent experiments (t-test, two-sided, *P < 0.05, **P < 0.01, ***P < 0.001)

Overexpression of TMEM120B promoted docetaxel and doxorubicin therapy resistance

CSC expansion is a major cause of therapeutic resistance [5, 57]. Based on the TCGA database, we found that TMEM120B mRNA levels were significantly higher in the docetaxel- and doxorubicin-resistant groups than in the docetaxel- and doxorubicin-sensitive groups (Fig. 8A-B). Both GSEA and RNA-seq results revealed that TMEM120B may be involved in homologous recombination (HR) (Fig. 8C-D). Subsequent western blotting indicated that the level of RAD51, an HR marker, was increased and that of γ-H2AX, a DNA damage marker, was suppressed upon TMEM120B overexpression rather than TMEM120B-∆CCD overexpression (Fig. 8E). The immunofluorescence assay also indicated that the nuclear foci of γ-H2AX were significantly abrogated within ectopic TMEM120B, but not in TMEM120B-∆CCD in SK-BR-3 cells (Fig. 8F). We treated MDA-231 cells with varying concentrations of docetaxel and doxorubicin and measured IC50. The results suggested that overexpressing TMEM120B increased IC50 values substantially (docetaxel, 25.27 ng/mL; doxorubicin, 28.23 ng/mL); however, there was no considerable alteration between SK-BR-3 cells overexpressing TMEM120B-∆CCD (docetaxel, 17.52 ng/mL, doxorubicin, 20.15 ng/mL) and the negative control (docetaxel, 14.06 ng/mL; doxorubicin, 20.07 ng/mL; Fig. 8G-H). Additionally, xenograft assays suggested that TMEM120B overexpression, rather than TMEM120B-∆CCD overexpression in SK-BR-3 cells, accelerated docetaxel and doxorubicin treatment resistance in vivo (Fig. 8I).

Fig. 8