ClpP regulates breast cancer cell proliferation, invasion and apoptosis by modulating the Src/PI3K/Akt signaling pathway

Background Caseinolytic protease P (ClpP), which is located on the inner mitochondrial membrane, degrades mitochondrial proteins damaged by oxidative stress. The role of ClpP varies among tumor types. However, the expression pattern and biological functions of ClpP in breast cancer (BC) have not yet been investigated. Methods The Cancer Genome Atlas (TCGA) and Kaplan Meier-plotter database were used to analyze the expression level of ClpP in BC tissues, relationships with clinicopathological characteristics, and the influence on the prognosis of BC. Protein and mRNA expression levels of ClpP in BC cell lines and tissues were detected by quantitative real-time PCR, western blot and immunohistochemical (IHC) analyses. The colony formation assay, transwell assay and flow cytometric analysis were performed to assess various functions of ClpP. Western blot analysis was also conducted to determine the mechanism of ClpP. Results ClpP expression was markedly increased in BC cells and tissues. High expression of ClpP was significantly correlated with the T stage, estrogen receptor (ER) expression, and poor recurrence-free survival (RFS) in TCGA and Kaplan Meier-plotter database. ClpP silencing significantly inhibited proliferation, migration, invasion, and promoted apoptosis of BC cells, which resulted in suppression of the Src/PI3K/Akt signaling pathway. The gain-of-function assay confirmed partial these results.


INTRODUCTION
Breast cancer (BC) is the most common malignant tumor among women worldwide (Bray et al., 2018). At present, the main treatment strategy for BC is surgical resection of the breast and axillary lymph nodes, in combination with adjuvant therapies, which include chemotherapy, radiotherapy, endocrine and targeted therapy (National Comprehensive Cancer Network, 2019). The rapid development of deep sequencing and molecular biotechnologies have allowed for the identification of new cancer targets and targeted therapeutic drugs.
The mitochondria have been a focus of cancer research since the 1950s, when it was discovered that cancer cells produce adenosine triphosphate (ATP) and employ different mechanisms to support cell growth than the normal surrounding tissues. Therefore, it was suggested that a defect in the mitochondrial mechanism will not only lead to increased glycolysis, but also to the transformation of normal cells into cancer cells (Warburg, 1956). The maintenance of mitochondrial function requires strict control of protein homeostasis via independent mechanisms for protein synthesis and degradation. In addition to the cytoplasmic ubiquitin/proteasome and protein quality control systems, the mitochondria of mammalian cells have three ATP-dependent protease families: Lon (Wang et al., 1993), FtsH (Banfi et al., 1999;Casari et al., 1998) and ClpXP (Corydon et al., 1998;Kang et al., 2002). These proteases regulate protein degradation and maintain protein quality control (Goldberg, 2003;Sauer et al., 2004). There is also evidence that an elevation to the proteostatic threshold can lead to the onset of various diseases, especially cancer, and maintain the stability of mitochondrial proteins in tumor cells (Seo et al., 2016).
The ClpXP protease complex is composed of two proteins: hexamers of a AAA+ ATPase (ClpX) and the tetradecameric peptidase caseinolytic protease P (ClpP) (Bross et al., 1995;Corydon et al., 1998;Kang et al., 2002), which contribute to the pathogenesis of human disease (Gispert et al., 2013). A recent study reported that ClpXP is upregulated in primary and metastatic human tumors, necessary to support tumor cell proliferation, motility and heightened metastatic competence in vivo, and correlated with shortened survival. However, the ClpP and ClpX subunits may not have completely overlapping function(s) in the tumor mitochondria (Seo et al., 2016).
ClpP is encoded by nuclear genes in mammalian cells and plays a central role in the quality control of mitochondrial proteins via the degradation of misfolded proteins. ClpP was first identified in bacteria and has since aroused interest as a potential anti-microbial therapeutic target (Zeiler et al., 2012). Studies have shown that bacterial ClpP inhibitors with a beta-lactone structure have antibacterial activities (Bottcher & Sieber, 2008;Gersch et al., 2013;Szyk & Maurizi, 2006). It has also been reported that mutations in human ClpP were associated with Perrault syndrome (Jenkinson et al., 2013). However, few studies have examined the role of ClpP in tumorigenesis.
The expression level of ClpP is greater in acute myeloid leukemia (AML) cells than in normal hematopoietic cells (Cole et al., 2015). Further research found that ClpP hyperactivation can lead to the death of leukemia and lymphoma cells due to selective proteolysis of mitochondrial proteome subsets involved in mitochondrial respiration and oxidative phosphorylation (Ishizawa et al., 2019). Cancer stem cells and chemo-resistant cells are highly dependent on oxidative phosphorylation (Farge et al., 2017;Kuntz et al., 2017;Lagadinou et al., 2013;Marin-Valencia et al., 2012;Viale et al., 2014). Hence, ClpP could be exploited as a novel target in cancer treatment. Therefore, the aim of the present study was to explore the expression pattern, biological functions and underlying mechanisms of ClpP as a novel target for the treatment of BC.

Tissue specimens
Human BC tissues and corresponding adjacent normal tissues were collected from patients who underwent surgery at the First Affiliated Hospital of Chongqing Medical University from 2014 to 2018, snap-frozen in liquid nitrogen, and then stored at −80 • C. This study was approved by the Institutional Ethics Committees of the First Affiliated Hospital of Chongqing Medical University (approval no. 2019-208) and conducted in accordance with the tenets of the Declaration of Helsinki. Each participant signed an informed consent form prior to study inclusion.

RNA extraction and quantitative real-time PCR (RT-qPCR)
Total RNA was extracted from cells and tissues with TRIzol reagent (Invitrogen) in accordance with the manufacturer's instructions. RT-qPCR was performed using an ABI 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with the SYBR Green kit (Invitrogen). β-actin was used as an internal control. Each sample was tested in triplicate. The followig primers were used for RT-qPCR analysis: ClpP, forward primer: 5 -GCC AAG CAC ACC AAA CAG A-3 , reverse primer: 5 -GGA CCA GAA CCT TGT CTA AG-3 ; β-actin, forward primer: 5 -CCT GTG GCA TCC ACG AAA CT-3 , reverse primer: 5 -GAA GCA TTT GCG GTG GAC GAT-3 .

Immunohistochemical (IHC) analysis
All specimens were formalin-fixed, paraffin-embedded and cut into 4 µm-thick sections, which were mounted onto glass slides. IHC analysis was conducted as described previously (Li et al., 2018). Briefly, the slides were incubated with a primary antibody against ClpP overnight at 4 • C. Images were obtained using a Leica microscope equipped with a digital camera (Leica Microsystems, Wetzlar, Germany) and the IHC results were analyzed using Image-Pro Plus 6.0 software (Media Cybernetics, Bethesda, MA, USA). The grayscale units were converted to optical density, units and the area and integrated optical density (IOD) of the sections were then measured to calculate the mean optical density (MOD) for semi-quantitative statistical analysis, where MOD = total IOD/total area. The mean density was calculated as IOD/area.

Cell apoptosis analysis
Cell apoptosis was detected with the Annexin V-FITC Apoptosis Detection Kit (BD Biosciences, Franklin Lakes, NJ, USA) in accordance with the manufacturer's protocol and a FACSCalibur flow cytometer (BD Biosciences).

Colony formation assay
MDA-MB-231 and ZR-75-1 cells were seeded into triplicate wells of 6-well plates at 1,000 cells/well, and cultured for 7 days. The cells were then fixed with 4% paraformaldehyde and stained with 0.1% crystal violet solution (C0121, Beyotime Institute of Biotechnology, Haimen, China).

Cell migration and invasion assay
Transwell chambers (8-µm pore size; Corning, NY, USA) were used to detect the migratory and invasive capabilities of BC cells. For the transwell migration assays, 200-µlL aliquots of transfected MDA-MB-231, ZR-75-1, MCF-7 and T47D cells (4 * 10 4 cells/well) were added into the upper transwell chamber, and 800 µL of medium containing 10% FBS were added to the lower chamber. After incubation at 37 • C under an atmosphere of 5% CO 2 /95% air for 24 h (MDA-MB-231 and ZR-75-1 cells), or 72 h (MCF-7 and T47D cells), cells that migrated through the membrane pores were fixed in 4% paraformaldehyde for 30 min and stained with 0.1% crystal violet (DC079; Genview, Beijing, China) for 15 min at room temperature. Matrigel TM -coated transwell filters (Matrigel TM : serum-free medium = 1:7, 70 µl/chamber) were used to evaluate the invasion capability of the cells. The subsequent procedures were the same as those for the cell migration assay. Cells from six random fields were counted under a microscope. All experiments were repeated three times.

Database analysis TCGA dataset analyses
Gene expression data of BC tissues were downloaded from the TCGA database (https://tcga-data.nci.nih.gov/tcga/). The study cohort consisted of a total of 1010 BC patients. ClpP expression data of 112 normal breast samples were included to compare differences in ClpP expression levels in BC tissues. Overall survival (OS) and complete clinicopathological data of 990 BC patients, and RFS data of 792 BC patients were screened. ClpP expression levels were ranked from low to high based on median values. The first 50% of patients were considered as the low-expression group and the second 50% as the high-expression group.

Kaplan Meier-plotter dataset analyses
Prognosis based on ClpP levels in BC patients was analyzed using the Kaplan Meier-plotter database (http://kmplot.com/analysis/).

Statistical analysis
Statistical analyses were performed using IBM SPSS Statistics for Windows, version 22.0 (Armonk, NY, USA) and GraphPad Prism 7.0 software (San Diego, CA, USA). The two-tailed Student's t -test was used to compare two groups of independent samples. The chi-square test was used to assess the correlation between ClpP expression and the clinicopathological characteristics of BC patients. Kaplan-Meier analysis was performed to plot survival curves, which were compared with the log-rank test. A p-value < 0.05 was regarded as statistically significant.

ClpP is overexpressed in human BC tissues and associated with clinical outcomes
ClpP expression levels in BC tissues were evaluated using the TCGA dataset. The results showed that ClpP expression was significantly up-regulated in BC tissues, as compared with normal tissues (p < 0.001) (Fig. 1A). Receiver operating characteristic (ROC) curve analysis was conducted to estimate the diagnostic value of ClpP. The area under the ROC curve (AUC) was 0.829 (95% confidence interval [CI] [0.79-0.869], p < 0.0001) (Fig. 1B). To verify the results, 18 pairs of BC and adjacent normal tissues were used for RT-qPCR analysis of ClpP expression levels. Consistently, ClpP expression was significantly increased in BC tissues, as compared with normal tissues (p < 0.01) (Fig. 1C). The AUC was 0.713 (95% CI [0.543-0.883], p = 0.029) (Fig. 1D). ClpP protein expression was also assessed by IHC analysis. As compared with normal tissues, cytoplasmic ClpP immunoreactivity was markedly higher in tumor tissues (p < 0.001) (Figs. 1E-1G). These results showed that ClpP expression was significantly increased in BC tissues, suggesting high diagnostic potential. The clinical significance of ClpP in BC was assessed in tissues from 990 BC patients with complete clinicopathological characteristics (Table 1). High ClpP expression was found to be significantly associated with the T stage (p = 0.0154) and ER expression (p = 0.0164). Kaplan-Meier survival curves from TCGA indicated that ClpP was not associated with RFS (p = 0.506) or OS (p = 0.619) ( Figs. 2A and 2B). However, within the Kaplan Meier-plotter database, high ClpP expression was associated with poor RFS (p = 0.00071) (Figs. 2C and 2D). These results suggest that upregulation of ClpP was correlated with the T stage, ER expression, and poor RFS, suggesting a potential essential role in BC tumorigenesis. The transwell migration and invasion assay was performed to quantitatively assess cell metastasis and invasiveness. The results showed that cell migration and invasion were significantly reduced in cells transfected with si-ClpP-B, as compared with control cells, due to the down-regulation of MMP7 and vimentin, and the up-regulation of E-cadherin . Conversely, the numbers of migratory and invasive MCF-7 and T47D cells were markedly increased by ClpP overexpression (Figs. S1B-S1E).
To confirm the specificity of ClpP siRNAs, the same functional experiments were performed with a second siRNA (si-ClpP-A). The results of the loss-of-function studies were consistent (Fig. S2).
These data revealed the anti-proliferative, anti-migration, anti-invasion and proapoptotic roles of silencing ClpP in BC cells. Notes. BC, breast cancer; TCGA, The Cancer Genome Atlas; ER, estrogen receptor; PR, progesterone receptor; Her-2, human epidermal growth factor receptor-2. *p < 0.05 indicates statistical significance.

The tumor suppressive effect of silencing ClpP is mediated by the Src/PI3K/Akt signaling pathway
Previous studies have reported that the PI3K/Akt signaling pathway is involved in cellular transformation, tumorigenesis, cancer progression, and proliferation of BC cells (Guerrero-Zotano, Mayer & Arteaga, 2016;Sharma et al., 2017). Further, ClpXP reported to mediate cell migration, invasion, and metastasis in vivo by increasing phosphorylation of the key cellular kinases Akt and Src (Seo et al., 2016). Here, western blot analysis was conducted to confirm whether the Src/PI3K/Akt pathway is involved in ClpP-induced BC progression. As shown in Fig. 6, silencing of ClpP inhibited the activation of Src, resulting in the inhibition of PI3K phosphorylation and the downstream signaling molecule Akt, which led to a series of changes in the activities of related molecules associated with cell proliferation, apoptosis, migration and invasion. These data confirmed the effect of ClpP on the proliferation and invasion capabilities of MDA-MB-231 and ZR-75-1 BC cells, as well as the induction of apoptosis.

DISCUSSION
A 2016 study conducted by Seo et al. (2016) found that ClpP was overexpressed in almost all human malignancies, as determined by immunohistochemical staining of a universal cancer tissue microarray. In the study, ClpP expression was significantly upregulated in both BC cell lines and tissues, which was consistent with previous findings. More, Ishizawa et al. (2019) reported that hyperactivation of ClpP could be a therapeutic strategy for patients with high ClpP expression, and that lower expression of ClpP was associated with reduced sensitivity to ClpP hyperactivation in ALM. Therefore, we can infer that high expression of ClpP presents not only a new therapeutic target for BC, but is also predictive of sensitivity to treatment. Seo et al. (2016) reported that ClpP expression is related to the histotype of breast adenocarcinoma. Likewise, the results of the present study showed that ClpP expression was closely correlated with the T stage and ER expression in BC, thereby demonstrating a link between ClpP expression and the clinical characteristics of BC patients. The results of bioinformatics and a meta-analysis revealed that ClpP expression was associated with a poorer outcome in 9 (64.3%) of 14 analyzed datasets. Importantly, high ClpP expression was correlated with shorter metastasis-free survival in BC patients and reduced RFS in those with lung adenocarcinoma (Seo et al., 2016). Consistent with previous studies, the results of the present study showed that ClpP was not significantly correlated with OS and RFS using the TCGA dataset. However, in the Kaplan Meier-plotter database, which contains a larger sample size, high ClpP expression was associated with poor RFS in BC patients. To the best of our knowledge, the number of clinical samples determines the possible relationship between ClpP expression and survival of BC patients. The Kaplan Meier-plotter database includes data from TCGA and GEO chips, which overlap with but greatly exceed the TCGA database. Further, survival analysis of 537 French BC patients from the E-MTAB-365 database with the Kaplan Meier-plotter showed that differences in diagnostic criteria, treatment regimens, and regions will result in differences in survival benefits of the same disease, suggesting that more comprehensive information must be obtained from diverse databases in order to address these issues in future studies.
ClpP is a subunit of the ClpXP complex. Seo et al. (2016) found that ClpXP was associated with tumor cell migration, invasion and metastasis. Consistent with these findings, the results of the present study suggest that silencing of ClpP inhibited BC cell proliferation, migration and invasion, and induced apoptosis.
The Akt signaling pathway is important in the regulation of various cellular functions, including metabolism, growth, proliferation, survival, transcription, protein synthesis and tumorigenesis (Aoki & Fujishita, 2017). Mutations in the PI3K/Akt pathway can reportedly mediate development, progression and drug resistance of BC (Guerrero-Zotano, Mayer & Arteaga, 2016;Sharma et al., 2017). Src can bind to different subtypes of the integrin family, which affects the motility and metastasis of tumor cells. Picon-Ruiz et al. (2016) found that the invasion of localized fat by cancer cells will activate Src, maintain the production of pro-inflammatory cytokines, and promote metastasis of BC cells. The results of the present study demonstrated that phosphorylation of Src, PI3K and Akt was markedly