skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Deep learning-based classification of breast cancer cells using transmembrane receptor dynamics
Abstract Motivation Motions of transmembrane receptors on cancer cell surfaces can reveal biophysical features of the cancer cells, thus providing a method for characterizing cancer cell phenotypes. While conventional analysis of receptor motions in the cell membrane mostly relies on the mean-squared displacement plots, much information is lost when producing these plots from the trajectories. Here we employ deep learning to classify breast cancer cell types based on the trajectories of epidermal growth factor receptor (EGFR). Our model is an artificial neural network trained on the EGFR motions acquired from six breast cancer cell lines of varying invasiveness and receptor status: MCF7 (hormone receptor positive), BT474 (HER2-positive), SKBR3 (HER2-positive), MDA-MB-468 (triple negative, TN), MDA-MB-231 (TN) and BT549 (TN). Results The model successfully classified the trajectories within individual cell lines with 83% accuracy and predicted receptor status with 85% accuracy. To further validate the method, epithelial–mesenchymal transition (EMT) was induced in benign MCF10A cells, noninvasive MCF7 cancer cells and highly invasive MDA-MB-231 cancer cells, and EGFR trajectories from these cells were tested. As expected, after EMT induction, both MCF10A and MCF7 cells showed higher rates of classification as TN cells, but not the MDA-MB-231 cells. Whereas deep learning-based cancer cell classifications are primarily based on the optical transmission images of cell morphology and the fluorescence images of cell organelles or cytoskeletal structures, here we demonstrated an alternative way to classify cancer cells using a dynamic, biophysical feature that is readily accessible. Availability and implementation A python implementation of deep learning-based classification can be found at https://github.com/soonwoohong/Deep-learning-for-EGFR-trajectory-classification. Supplementary information Supplementary data are available at Bioinformatics online.  more » « less
Award ID(s):
2029266
PAR ID:
10309984
Author(s) / Creator(s):
; ; ; ;
Editor(s):
Xu, Jinbo
Date Published:
Journal Name:
Bioinformatics
ISSN:
1367-4803
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; ; ; ; ; ; ; et al (, Molecular Omics)
    Triple negative breast cancer (TNBC) is an aggressive type of breast cancer with very little treatment options. TNBC is very heterogeneous with large alterations in the genomic, transcriptomic, and proteomic landscapes leading to various subtypes with differing responses to therapeutic treatments. We applied a multi-omics data integration method to evaluate the correlation of important regulatory features in TNBC BRCA1 wild-type MDA-MB-231 and TNBC BRCA1 5382insC mutated HCC1937 cells compared with non-tumorigenic epithelial breast MCF10A cells. The data includes DNA methylation, RNAseq, protein, phosphoproteomics, and histone post-translational modification. Data integration methods identified regulatory features from each omics method that had greater than 80% positive correlation within each TNBC subtype. Key regulatory features at each omics level were identified distinguishing the three cell lines and were involved in important cancer related pathways such as TGFβ signaling, PI3K/AKT/mTOR, and Wnt/beta-catenin signaling. We observed overexpression of PTEN, which antagonizes the PI3K/AKT/mTOR pathway, and MYC, which downregulates the same pathway in the HCC1937 cells relative to the MDA-MB-231 cells. The PI3K/AKT/mTOR and Wnt/beta-catenin pathways are both downregulated in HCC1937 cells relative to MDA-MB-231 cells, which likely explains the divergent sensitivities of these cell lines to inhibitors of downstream signaling pathways. The DNA methylation and RNAseq data is freely available via GEO GSE171958 and the proteomics data is available via the ProteomeXchange PXD025238. 
    more » « less
  2. ; ; ; ; ; ; (, Journal of Biomechanical Engineering)
    Abstract During the migration of cancer cells for metastasis, cancer cells can be exposed to fluid shear conditions. We examined two breast cancer cell lines, MDA-MB-468 (less metastatic) and MDA-MB-231 (more metastatic), and a benign MCF-10A epithelial cell line for their responsiveness in migration to fluid shear. We tested fluid shear at 15 dyne/cm2 that can be encountered during breast cancer cells traveling through blood vessels or metastasizing to mechanically active tissues such as bone. MCF-10A exhibited the least migration with a trend of migrating in the flow direction. Intriguingly, fluid shear played a potent role as a trigger for MDA-MB-231 cell migration, inducing directional migration along the flow with significantly increased displacement length and migration speed and decreased arrest coefficient relative to unflowed MDA-MB-231. In contrast, MDA-MB-468 cells were markedly less migratory than MDA-MB-231 cells, and responded very poorly to fluid shear. As a result, MDA-MB-468 cells did not exhibit noticeable difference in migration between static and flow conditions, as was distinct in root-mean-square (RMS) displacement—an ensemble average of all participating cells. These may suggest that the difference between more metastatic MDA-MB-231 and less metastatic MDA-MB-468 breast cancer cells could be at least partly involved with their differential responsiveness to fluid shear stimulatory cues. Our study provides new data in regard to potential crosstalk between fluid shear and metastatic potential in mediating breast cancer cell migration. 
    more » « less
  3. ; ; ; ; ; ; ; (, Biofabrication)
    Abstract Immunotherapy has revolutionized cancer treatment with the advent of advanced cell engineering techniques aimed at targeted therapy with reduced systemic toxicity. However, understanding the underlying immune–cancer interactions require development of advanced three-dimensional (3D) models of human tissues. In this study, we fabricated 3D tumor models with increasing complexity to study the cytotoxic responses of CD8 + T cells, genetically engineered to express mucosal-associated invariant T (MAIT) cell receptors, towards MDA-MB-231 breast cancer cells. Homotypic MDA-MB-231 and heterotypic MDA-MB-231/human dermal fibroblast tumor spheroids were primed with precursor MAIT cell ligand 5-amino-6-D-ribitylaminouracil (5-ARU). Engineered T cells effectively eliminated tumors after a 3 d culture period, demonstrating that the engineered T cell receptor recognized major histocompatibility complex class I-related (MR1) protein expressing tumor cells in the presence of 5-ARU. Tumor cell killing efficiency of engineered T cells were also assessed by encapsulating these cells in fibrin, mimicking a tumor extracellular matrix microenvironment. Expression of proinflammatory cytokines such as interferon gamma, interleukin-13, CCL-3 indicated immune cell activation in all tumor models, post immunotherapy. Further, in corroborating the cytotoxic activity, we found that granzymes A and B were also upregulated, in homotypic as well as heterotypic tumors. Finally, a 3D bioprinted tumor model was employed to study the effect of localization of T cells with respect to tumors. T cells bioprinted proximal to the tumor had reduced invasion index and increased cytokine secretion, which indicated a paracrine mode of immune–cancer interaction. Development of 3D tumor-T cell platforms may enable studying the complex immune–cancer interactions and engineering MAIT cells for cell-based cancer immunotherapies. 
    more » « less
  4. ; ; ; ; (, Photochemistry and Photobiology)
    Abstract P‐glycoprotein (P‐gp, ABCB1) is a well‐researched ATP‐binding cassette (ABC) drug efflux transporter linked to the development of cancer multidrug resistance (MDR). Despite extensive studies, approved therapies to safely inhibit P‐gp in clinical settings are lacking, necessitating innovative strategies beyond conventional inhibitors or antibodies to reverse MDR. Photodynamic therapy is a globally approved cancer treatment that uses targeted, harmless red light to activate non‐toxic photosensitizers, confining its cytotoxic photochemical effects to disease sites while sparing healthy tissues. This study demonstrates that photodynamic priming (PDP), a sub‐cytotoxic photodynamic therapy process, can inhibit P‐gp function by modulating cellular respiration and ATP levels in light accessible regions. Using chemoresistant (VBL‐MDA‐MB‐231) and chemosensitive (MDA‐MB‐231) triple‐negative breast cancer cell lines, we showed that PDP decreases mitochondrial membrane potential by 54.4% ± 30.4 and reduces mitochondrial ATP production rates by 94.9% ± 3.46. Flow cytometry studies showed PDP can effectively improve the retention of P‐gp substrates (calcein) by up to 228.4% ± 156.3 in chemoresistant VBL‐MDA‐MB‐231 cells, but not in chemosensitive MDA‐MB‐231 cells. Further analysis revealed that PDP did not alter the cell surface expression level of P‐gp in VBL‐MDA‐MB‐231 cells. These findings indicate that PDP can reduce cellular ATP below the levels that is required for the function of P‐gp and improve intracellular substrate retention. We propose that PDP in combination with chemotherapy drugs, might improve the efficacy of chemotherapy and overcome cancer MDR. 
    more » « less
  5. ; ; (, Cancer research)
    Senescence is a potent tumor-suppressive mechanism that irreversibly arrests the growth of damaged cells. However, senescent cells that accumulate in tissues eventually develop a senescence-associated secretory phenotype (SASP) that alters the microenvironment to promote cancer. Paracrine factors in the SASP may also contribute to the formation of rare giant polyploidal cancer cells (GPCCs). A single-cell mechanical approach was used to profile cytoskeletal and nuclear mechanics, morphology, motility, and adhesion for breast cancer cells treated with conditioned media from senescent fibroblasts. Our study showed that a small but significant population of MDA-MB-231 breast cancer cells (less than 5%) treated with conditioned media from senescent LF-1 fibroblasts develop an enlarged morphology, chromosomal instability, and polyploidy, a phenotype associated with GPCCs. Although GPCCs are highly invasive and chemoresistant, little is known about their biophysical properties. First, we developed a method for identifying the small subpopulation of GPCCs in a heterogeneous population of cancer cells based on increased nuclear area and confirmed that GPCCs are more resistant to paclitaxel than normal-size MDA-MB-231 cells (NCCs). We then compared critical biophysical properties of NCCs and GPCCs, including cytoskeletal and nuclear mechanics, cell and nuclear morphology, motility, and adhesion. Cells were stained for cytoskeletal proteins actin, tubulin, and vinculin. Cytoskeletal organization was dramatically altered in GPCCs compared to NCCs. GPCCs displayed more disorganized microtubule structure, dense actin stress fibers, and mature focal adhesions. Intracellular particle tracking microrheology was used to measure cytoskeletal and nuclear mechanics. These studies demonstrated that although GPCCs are thought to be highly invasive cancer cells, they are inherently stiffer than NCCs, in terms of both their cytoskeletal and nuclear mechanics. This was surprising since more invasive cancer cells are often more compliant than less invasive cancer cells. This result may be in part to the ability for GPCCs to behave like activated stromal cells that stiffen in the tumor; we confirmed that GPCCs display similar adhesive behavior as activated stromal cells. To determine how mechanics correlates with cell migration, we used time-lapse nuclear tracking to measure cell motility. The average cell speed was higher for NCCs than for GPCCs; however, GPCCs moved longer distances over time because their motion was more directional. These findings highlight the unusual biophysical behavior of GPCCs. To develop pharmacologic tools that target GPCCs, it is imperative to understand their biophysical properties. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email