Machine Learning Classifies Ferroptosis and Apoptosis Cell Death Modalities with TfR1 Immunostaining

Abstract

Determining cell death mechanisms occurring in patient and animal tissues is a longstanding goal that requires suitable biomarkers and accurate quantification. However, effective methods remain elusive. To develop more powerful and unbiased analytic frameworks, we developed a machine learning approach for automated cell death classification. Image sets were collected of HT-1080 fibrosarcoma cells undergoing ferroptosis or apoptosis and stained with an anti-transferrin receptor 1 (TfR1) antibody, together with nuclear and F-actin staining. Features were extracted using high-content-analysis software, and a classifier was constructed by fitting a multinomial logistic lasso regression model to the data. The prediction accuracy of the classifier within three classes (control, ferroptosis, apoptosis) was 93%. Thus, TfR1 staining, combined with nuclear and F-actin staining, can reliably detect both apoptotic and ferroptotis cells when cell features are analyzed in an unbiased manner using machine learning, providing a method for unbiased analysis of modes of cell death.

Ferroptosis turns 10: Emerging mechanisms, physiological functions, and therapeutic applications

“…My lab recently reported that increased abundance and plasma localization of the Tf receptor TfR1 is a marker for ferroptosis, which can be detected with the newly identified 3F3-FMA antibody, as well as with additional anti-TfR1 antibodies (Feng et al, 2020). We then found that TfR1 staining could be coupled with machine learning to distinguish cells undergoing ferroptosis from those undergoing apoptosis (Jin et al, 2022). Several genes, such as CHAC1, PTGS2, SLC7A11, and ACSL4, are induced during ferroptosis, and RGS4 is downregulated during ferroptosis; altered expression of these genes can be detected by qPCR as indicators of ferroptosis (Figure 5).…”

Suppression of ferroptosis by vitamin A or antioxidants is essential for neuronal development

“…3e). We recently demonstrated that the transferrin receptor (TfR1) can be utilized to visualize ferroptotic events 22,23 . Therefore, we sectioned organoids generated in +AO, -AO+vA lo and -AO+vA lo +Fer-1 conditions at day 60 and stained these with antibodies against TfR1.…”

Application of Machine Learning in Spatial Proteomics

“…122 In addition, the subcellular distribution pattern of proteins can be used to train a ML to automatically distinguish between ferroptosis and apoptosis cells. 197 The investigation of organelle proteome heterogeneity in neighboring cells can reveal cell interactions and disease-related microenvironment features in diseases, such as systemic autoimmune disease 114 and breast cancer. 198,199 Cell segmentation is an important step in imaging-based spatial proteomic data analysis and application, and various advanced CNN-based cell segmentation methods have been developed.…”

“…The ensLOC model has been applied to detect changes in yeast protein abundance and localization under environmental and genetic perturbations . In addition, the subcellular distribution pattern of proteins can be used to train a ML to automatically distinguish between ferroptosis and apoptosis cells . The investigation of organelle proteome heterogeneity in neighboring cells can reveal cell interactions and disease-related microenvironment features in diseases, such as systemic autoimmune disease and breast cancer. , …”