scholarly journals Learning Deep Features for Stain-free Live-dead Human Breast Cancer Cell Classification

2020 ◽  
Author(s):  
Gisela Pattarone ◽  
Laura Acion ◽  
Marina Simian ◽  
Emmanuel Iarussi
Keyword(s):  
Breast Cancer ◽  
Machine Learning ◽  
Drug Treatment ◽  
Cancer Cells ◽  
Cancer Cell ◽  
Breast Cancer Cell ◽  
Human Breast Cancer ◽  
Breast Cancer Cells ◽  
Human Breast ◽  
Bright Field

Abstract Automated cell classification in cancer biology is a challenging topic in computer vision and machine learning research. Breast cancer is the most common malignancy in women that usually involves phenotypically diverse populations of breast cancer cells and an heterogeneous stroma. In recent years, automated microscopy technologies are allowing the study of live cells over extended periods of time, simplifying the task of compiling large image databases. For instance, there have been several studies oriented towards building machine learning systems capable of automatically classifying images of different cell types (i.e. motor neurons, stem cells). In this work we were interested in classifying breast cancer cells as live or dead, based on a set of automatically retrieved morphological characteristics using image processing techniques. Our hypothesis is that live-dead classification can be performed without any staining and using only bright-field images as input. To our knowledge, there is no previous work attempting this task on in vitro studies of breast cancer cells, nor is there a dataset available to explore solutions related to this issue. We tackled this problem using the JIMT-1 breast cancer cell line that grows as an adherent monolayer. First, a vast image set composed by JIMT-1 human breast cancer cells that had been exposed to a chemotherapeutic drug treatment (doxorubicin and paclitaxel) or vehicle control was compiled. Next, several state-of-the-art classifiers were trained based on convolutional neural networks (CNN) to perform supervised classification using labels obtained from fluorescence microscopy images associated with each bright-field image. Model performances were evaluated and compared on a large number of bright-field images. The best model reached an AUC = 0.941 for classifying breast cancer cells without treatment. Furthermore, it reached AUC = 0.978 when classifying breast cancer cells under drug treatment. Our results highlight the potential of machine learning and computational image analysis to build new diagnosis tools that benefit the biomedical field by reducing cost, time, and stimulating work reproducibility. More importantly, we analyzed the way our classifiers clusterize bright-field images in the learned high-dimensional embedding and linked these groups to salient visual characteristics in live-dead cell biology observed by trained experts.

scholarly journals Learning deep features for dead and living breast cancer cell classification without staining

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Gisela Pattarone ◽  
Laura Acion ◽  
Marina Simian ◽  
Emmanuel Iarussi
Keyword(s):  
Breast Cancer ◽  
Machine Learning ◽  
Drug Treatment ◽  
Cancer Cells ◽  
Cancer Cell ◽  
Breast Cancer Cell ◽  
Cancer Biology ◽  
Breast Cancer Cells ◽  
Bright Field ◽  
Cell Classification

AbstractAutomated cell classification in cancer biology is a challenging topic in computer vision and machine learning research. Breast cancer is the most common malignancy in women that usually involves phenotypically diverse populations of breast cancer cells and an heterogeneous stroma. In recent years, automated microscopy technologies are allowing the study of live cells over extended periods of time, simplifying the task of compiling large image databases. For instance, there have been several studies oriented towards building machine learning systems capable of automatically classifying images of different cell types (i.e. motor neurons, stem cells). In this work we were interested in classifying breast cancer cells as live or dead, based on a set of automatically retrieved morphological characteristics using image processing techniques. Our hypothesis is that live-dead classification can be performed without any staining and using only bright-field images as input. We tackled this problem using the JIMT-1 breast cancer cell line that grows as an adherent monolayer. First, a vast image set composed by JIMT-1 human breast cancer cells that had been exposed to a chemotherapeutic drug treatment (doxorubicin and paclitaxel) or vehicle control was compiled. Next, several classifiers were trained based on well-known convolutional neural networks (CNN) backbones to perform supervised classification using labels obtained from fluorescence microscopy images associated with each bright-field image. Model performances were evaluated and compared on a large number of bright-field images. The best model reached an AUC = 0.941 for classifying breast cancer cells without treatment. Furthermore, it reached AUC = 0.978 when classifying breast cancer cells under drug treatment. Our results highlight the potential of machine learning and computational image analysis to build new diagnosis tools that benefit the biomedical field by reducing cost, time, and stimulating work reproducibility. More importantly, we analyzed the way our classifiers clusterize bright-field images in the learned high-dimensional embedding and linked these groups to salient visual characteristics in live-dead cell biology observed by trained experts.

scholarly journals MARCH8 Suppresses Tumor Metastasis and Mediates Degradation of STAT3 and CD44 in Breast Cancer Cells

Cancers ◽  
2021 ◽  
Vol 13 (11) ◽  
pp. 2550
Author(s):  
Wenjing Chen ◽  
Dhwani Patel ◽  
Yuzhi Jia ◽  
Zihao Yu ◽  
Xia Liu ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Cancer Cells ◽  
Cell Lines ◽  
Cancer Cell ◽  
Breast Cancer Cell ◽  
Human Breast Cancer ◽  
Breast Cancer Cells ◽  
Human Breast Cancer Cell ◽  
Breast Cancer Cell Lines ◽  
Human Breast

Protein stability is largely regulated by post-translational modifications, such as ubiquitination, which is mediated by ubiquitin-activating enzyme E1, ubiquitin-conjugating enzyme E2, and ubiquitin ligase E3 with substrate specificity. Membrane-associated RING-CH (MARCH) proteins represent one novel family of transmembrane E3 ligases which target glycoproteins for lysosomal destruction. While most of the MARCH family members are known to degrade membrane proteins in immune cells, their tumor-intrinsic role is largely unknown. In this study, we found that the expression of one MARCH family member, MARCH8, is specifically downregulated in breast cancer tissues and positively correlated with breast cancer survival rate according to bioinformatic analysis of The Cancer Genomic Atlas (TCGA) dataset. MARCH8 protein expression was also lower in a variety of human breast cancer cell lines in comparison to immortalized human mammary epithelial MCF-12A cells. Restoration of MARCH8 expression induced apoptosis in human breast cancer cell lines MDA-MB-231 and BT549. Stable expression of MARCH8 inhibited tumorigenesis and lung metastases of MDA-MB-231 cells in mice. Moreover, we discovered that the breast cancer stem-cell marker and metastasis driver CD44, a membrane protein, interacts with MARCH8 and is one of the glycoprotein targets subject to MARCH8-dependent lysosomal degradation. Unexpectedly, we identified a nonmembrane protein, signal transducer and transcription activator 3 (STAT3), as another essential ubiquitination target of MARCH8, whose degradation through the proteasome pathway is responsible for the proapoptotic changes mediated by MARCH8. These findings highlight a novel tumor-suppressing function of MARCH8 in targeting both membrane and nonmembrane protein targets required for the survival and metastasis of breast cancer cells.

scholarly journals Long non-coding RNA LINC00662 promotes proliferation and migration of breast cancer cells via regulating the miR-497-5p/EglN2 axis

2020 ◽  
Author(s):  
Wuqin Xu ◽  
Zihe Xing ◽  
Peng Zhang ◽  
Wuqin Xu
Keyword(s):  
Breast Cancer ◽  
Cancer Cells ◽  
Cell Lines ◽  
Cancer Cell ◽  
Breast Cancer Cell ◽  
Human Breast Cancer ◽  
Breast Cancer Cells ◽  
Cancer Cell Lines ◽  
Breast Cancer Cell Lines ◽  
Human Breast

Previous reports indicated that long noncoding RNA 662 (LINC00662) plays a crucial role in several human cancers. Here, we studied the expression pattern of LINC00662 and explored its function in human breast cancer. The expression level of LINC00662 was determined in human breast cancer cell lines and tissues by real-time quantitative polymerase chain reaction (RT-qPCR). Cytoplasmic and nuclear RNA from MDA-MB-157 cells were extracted to analyze the subcellular location of LINC00662. Moreover, the MTT assay, wound-healing assay, colony-forming assay and transwell assay were employed in MDA-MB-157 cells to detect the effect of LINC00662 on cell apoptosis, invasion, migration and proliferation, respectively. LINC00662-specific miRNA and miRNA-gene axis were examined in a dual-luciferase reporter assay and Western blot. We found that LINC00662 was overexpressed in both breast cancer cell lines and tissue compared to normal breast cell lines and healthy breast tissue. Analysis of subcellular localization revealed that LINC00662 was mainly found in the cytoplasm. Furthermore, LINC00662 silencing reduced cell viability and inhibited the proliferation, migration and invasion of MDA-MB-157 cells. Bioinformatics analysis predicted that LNC00662 binds to miR-497-5p. A series of studies confirmed that LINC00662 directly interacted with miR-497-5p and downregulated its expression in MDA-MB-157 cells. MiR-497-5p knockdown significantly reversed the inhibitory effect of shLINC00662. Moreover, egl-9 family hypoxia inducible factor 2 (EglN2) was verified as a target of miR-497-5p. Overall, our results demonstrated that overexpression of LINC00662 accelerated the malignant growth of breast cancer cells via sponging miR-497-5p and upregulating EglN2 expression, and indicate that targeting LINC00662 may represent a novel strategy for breast cancer therapy.

scholarly journals Macrophage conditioned media promotes adipocyte cancer-association, which in turn stimulates breast cancer proliferation and migration

2020 ◽  
Author(s):  
Karin A. Vallega ◽  
Dale B. Bosco ◽  
Yi Ren ◽  
Qing-Xiang Sang
Keyword(s):  
Breast Cancer ◽  
Cancer Cells ◽  
Cancer Cell ◽  
Breast Cancer Cell ◽  
Human Breast Cancer ◽  
Breast Cancer Cells ◽  
Conditioned Media ◽  
Human Breast ◽  
And Migration ◽  
Proliferation And Migration

Abstract Background Breast cancer is the most common cancer in women and the leading cause of female cancer deaths worldwide. Obesity causes chronic inflammation and is a risk factor for post-menopausal breast cancer and poor prognosis. Obesity is known to trigger increased infiltration of macrophages into adipose tissue, yet little research has focused on the effects of macrophages in the early stages of breast tumor development in obese patients. In this study, the effects of pro-inflammatory macrophages on breast cancer-adipocyte crosstalk were investigated. Methods An innovative human cell co-culture system was used to model the paracrine interactions among adipocytes, macrophages, and breast cancer cells, and how they can facilitate tumor progression. The effects on human breast cancer cells were examined using cell counts and migration assays. Quantitative reverse-transcription polymerase chain reaction (qRT-PCR) was used to measure the expression levels of several cytokines and proteases to analyze adipocyte cancer-association. Results Macrophage conditioned media intensified the effects of breast cancer-adipocyte crosstalk. More specifically, adipocytes became delipidated and increased production of pro-inflammatory cytokines, even in the absence of breast cancer cells, although the expression levels were highest with all three cell components. As a result, co-cultured breast cancer cells became more aggressive, with increased proliferation and migration potential when compared to adipocyte-breast cancer cell co-cultures treated with unconditioned media. Conclusions Macrophage conditioned media promotes adipocyte cancer-association and production of pro-inflammatory factors. These macrophage-adipocyte paracrine interactions promote human breast cancer cell proliferation and migration. Thus, macrophages may contribute to adipocyte inflammation and cancer-association and promote breast cancer progression.

scholarly journals Disruption of pH Dynamics Suppresses Proliferation and Potentiates Doxorubicin Cytotoxicity in Breast Cancer Cells

Pharmaceutics ◽  
2021 ◽  
Vol 13 (2) ◽  
pp. 242
Author(s):  
Diana Tavares-Valente ◽  
Bárbara Sousa ◽  
Fernando Schmitt ◽  
Fátima Baltazar ◽  
Odília Queirós
Keyword(s):  
Breast Cancer ◽  
Cancer Cells ◽  
Cell Lines ◽  
Cancer Cell ◽  
Breast Cancer Cell ◽  
Human Breast Cancer ◽  
Breast Cancer Cells ◽  
Human Breast ◽  
Cancer Tissues ◽  
Mcf 7

The reverse pH gradient is a major feature associated with cancer cell reprogrammed metabolism. This phenotype is supported by increased activity of pH regulators like ATPases, carbonic anhydrases (CAs), monocarboxylate transporters (MCTs) and sodium–proton exchangers (NHEs) that induce an acidic tumor microenvironment, responsible for the cancer acid-resistant phenotype. In this work, we analyzed the expression of these pH regulators and explored their inhibition in breast cancer cells as a strategy to enhance the sensitivity to chemotherapy. Expression of the different pH regulators was evaluated by immunofluorescence and Western blot in two breast cancer cell lines (MDA-MB-231 and MCF-7) and by immunohistochemistry in human breast cancer tissues. Cell viability, migration and invasion were evaluated upon exposure to the pH regulator inhibitors (PRIs) concanamycin-A, cariporide, acetazolamide and cyano-4-hydroxycinnamate. Additionally, PRIs were combined with doxorubicin to analyze the effect of cell pH dynamic disruption on doxorubicin sensitivity. Both cancer cell lines expressed all pH regulators, except for MCT1 and CAXII, only expressed in MCF-7 cells. There was higher plasma membrane expression of the pH regulators in human breast cancer tissues than in normal breast epithelium. Additionally, pH regulator expression was significantly associated with different molecular subtypes of breast cancer. pH regulator inhibition decreased cancer cell aggressiveness, with a higher effect in MDA-MB-231. A synergistic inhibitory effect was observed when PRIs were combined with doxorubicin in the breast cancer cell line viability. Our results support proton dynamic disruption as a breast cancer antitumor strategy and the use of PRIs to boost the activity of conventional therapy.

scholarly journals ERβ-induced MFN2 inhibits the migration and invasiveness of breast cancer cells by inhibiting the P-AKT signaling pathway

2020 ◽  
Author(s):  
Mengyu Wei ◽  
Jun Hao ◽  
Xiaomei Liao ◽  
Yinfeng Liu ◽  
Ruihuan Fu ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Cancer Cells ◽  
Cancer Cell ◽  
Breast Cancer Cell ◽  
Breast Cancer Cells ◽  
Akt Pathway ◽  
Cancer Cell Migration ◽  
Breast Cancer Patients ◽  
Human Breast ◽  
Mcf 7

Abstract Background Mitofusin 2 (MFN2) is localized on the outer membrane of mitochondria and is closely related to the migration of malignant tumor cells. Estrogen receptor β (ERβ) plays an anticancer role in breast cancer. Our previous experiments showed that ERβ can induce MFN2 expression, which then inhibits breast cancer cell migration. However, the exact mechanism by which ERβ-induced MFN2 inhibits breast cancer cell migration is unknown. Methods In this study, immunohistochemistry was first used to detect the expression of MFN2 in breast cancer tissues, and its relationship with the clinicopathological characteristics and prognosis of breast cancer patients was analyzed. MCF-7 and MDA-MB-231 cells were transfected with ERβ and MFN2 knockdown or expression plasmids. Western blot was used to detect the effects of ERβ on MFN2 and MFN2 on P-AKT473 and MMP2; the P-AKT pathway inhibitor LY294002 was administered to cells transfected with MFN2 knockdown plasmids, Western blot, immunocytofluorescence, and a wound healing assay revealed the effect of MFN2 on its downstream signaling pathway and the migration of breast cancer cells. Results This study found that the expression of MFN2 is related to the molecular type and prognosis of breast cancer patients ( P <0.05). The positive expression rate of MFN2 in triple-negative breast cancer was significantly lower than that in the HER2 + and luminal types. However, MFN2 expression was unrelated to age, tumor size, lymph node metastasis, TNM stage, histological type and grade ( P >0.05); ERβ positively regulated MFN2 expression and reduced the migration of both MCF-7 and MDA-MB-231 cells, while MFN2 knockdown increased the expression of P-AKT473 and MMP2. In contrast, the overexpression of MFN2 inhibited the expression of P-AKT473 and MMP2. These results showed that in MFN2 knockdown cells treated with LY294002, P-AKT473 and MMP2 expression levels were reversed. The reversal of P-AKT473 and MMP2 expression levels inhibits the invasiveness of human breast cancer cells. Conclusion MFN2 is related to the molecular subtype and prognosis of breast cancer. In human breast cancer MCF-7 and MDA-MB-231 cells, ERβ-induced MFN2 can inhibit the P-AKT pathway, which inhibits the invasiveness and migration of both breast cancer cell lines.

scholarly journals Curcumin Down-Regulates Visfatin Expression and Inhibits Breast Cancer Cell Invasion

Endocrinology ◽  
2012 ◽  
Vol 153 (2) ◽  
pp. 554-563 ◽  
Author(s):  
Su-Ryun Kim ◽  
Hyun-Joo Park ◽  
Yun-Hee Bae ◽  
Soon-Cheol Ahn ◽  
Hee-Jun Wee ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Gene Expression ◽  
Cancer Cells ◽  
Breast Cancer Cell ◽  
Cell Invasion ◽  
Human Breast Cancer ◽  
Breast Cancer Cells ◽  
Human Breast Cancer Cells ◽  
Cancer Cell Invasion ◽  
Human Breast

Obesity is frequently associated with breast cancer. Such associations are possibly mediated by adipokines. Visfatin, an adipokine, has recently been shown to be related to the development and progression of breast cancer. Therefore, the down-regulation of visfatin may be a novel strategy for breast cancer therapy. Curcumin has anticancer activities by modulating multiple signaling pathways and genes. The purpose of this study was to investigate whether visfatin gene expression is affected by curcumin in human breast cancer cells and to characterize the functional role of visfatin in breast cancer. We found that the mRNA and protein levels of visfatin were down-regulated by curcumin in MDA-MB-231, MDA-MB-468, and MCF-7 breast cancer cells, along with decreased activity of constitutive nuclear factor (NF)-κB. We confirmed the repressive effect of curcumin on visfatin transcription by performing a visfatin promoter-driven reporter assay and identified two putative NF-κB-binding sites on visfatin promoter that are important for this effect. EMSA and chromatin immunoprecipitation analysis indicated the binding of p65 to the visfatin promoter, which was effectively blocked by curcumin. Enforced expression of p65 protein increased visfatin promoter activity, whereas blocking NF-κB signaling suppressed visfatin gene expression. Visfatin could enhance the invasion of MDA-MB-231 cells and also attenuate curcumin-induced inhibition of cell invasion; on the other hand, visfatin knockdown by small interfering RNA led to the reduction of cell invasion. Our data demonstrate, for the first time, that curcumin down-regulates visfatin gene expression in human breast cancer cells by a mechanism that is, at least in part, NF-κB dependent and suggest that visfatin may contribute to breast cancer cell invasion and link obesity to breast cancer development and progression.

Sign in / Sign up

Export Citation Format

Share Document