Subtle mitochondrial stress can act as an early warning biomarker for drug candidate toxicity, appearing long before changes in conventional cell viability or cytotoxicity results. This newsletter edition includes case study data demonstrating how mitochondrial stress can be used as a sensitive early warning biomarker for cytotoxicity prediction, as well as the latest publications featuring label-free live cell imaging and AI analysis in cytotoxicity, immunotherapy, and phenotypic profiling.
Using mitochondrial stress as an early warning biomarker to de-risk preclinical toxicity screens
AI-powered imaging and analysis of mitochondria is a powerful noninvasive tool for accelerating drug discovery and enhancing safety assessment. Incorporating live, label-free mitochondrial analysis into in vitro workflows provides a practical way to de-risk compounds earlier, improve confidence in go/no-go decisions, and reduce the likelihood of late-stage attrition.
In preclinical toxicity screening, early mitochondrial stress is a critical but often overlooked indicator of compound toxicity. Monitoring mitochondria at high resolution is essential for measuring mitochondrial dynamics and identifying subtle phenotypic shifts in response to drug-induced stress, however conventional imaging methods rely on chemical stains or fluorescent dyes that can induce DNA damage and phototoxicity. To maintain
high-resolution imaging that doesn’t damage mitochondria, Nanolive uses proprietary interferometric holotomography and AI-powered mitochondrial segmentation instead of dyes.
Mitochondrial structures in a preadipocyte imaged and processed with the Smart Mitochondrial AssayLIVE under noninvasive conditions (top) and exposed to high-energy light (bottom).White arrows indicate swollen mitochondria.
For the Smart Mitochondrial AssayLIVE flyer, click here.
Early mitochondrial stress is detectable before cellular toxicity
We carried out a case study to analyze the effects of damaging high-energy light at the cellular and mitochondrial-levels over 60 minutes, using Nanolive’s AI-powered assay to digitally segment and quantify mitochondria from label-free images. By comparing mitochondrial readouts to cell morphology readouts like cell death, or area, we confirmed that mitochondrial stress was detectable before cellular stress. As shown in the figure, while the cell began to contract and exhibited membrane blebbing after 55 minutes of exposure and imaging, its mitochondria had already begun to fragment and swell much earlier, detected here by measuring the mean mitochondrial length.
To download a poster of this figure, click here.
A tool for studying mitochondrial morphology at high-resolution
Nanolive’s AI mitochondrial analysis solution enables label-free quantification of mitochondrial morphology without user training or input. Some of the automated outputs used to assess the impact of stressors on mitochondrial morphology include mitochondrial count, length, area, and radius. Our full application note contains more detail on how these metrics can be used to assess changes in mitochondrial network fragmentation, swelling, mitochondrial fission, and mitophagy: Application Note.
High-content data from a single experiment
- Non-invasive imaging preserves biological relevance: Nanolive’s standard low-power illumination and label-free approach sets the standard for live cell imaging, maintaining cellular and mitochondrial health over time.
- Quantification of mitochondrial phenotype reveals early stress responses: easily monitor and quantify mitochondrial fragmentation over time, without the limits of staining and phototoxicity.
- Mitochondrial changes precede detectable cellular toxicity: automatically detect fine shifts in mitochondrial morphology that precede signs of overall cellular toxicity.
Want to experience Nanolive’s analysis software?
Get a first-hand look into Nanolive’s workflow in this Online Demo, as our Field Application Specialist Matthias guides us through the insights gained from a high-content phenotypic screen.
Join us to discover how a single experiment provides:
- More insights into drug mechanism of action and side effects
- More biologically relevant conclusions
- Dynamic, real-time data, to move beyond the limits of endpoint assays
- Noninvasive monitoring of sensitive cells to preserve their natural behavior
Please click here to register.
Latest publication highlights with Nanolive imaging:
- Phenotypic profiling: Zhang, Y. et al. (2026) ‘Schisandrin B Targets PXR to Enhance Bile Acid Metabolism and Alleviate ANIT-Induced Cholestatic Liver Injury via Dual Pathways’, International Journal of Biological Sciences, https://www.ijbs.com/v22p0387.htm
- Immunotherapy: Zhang, Y. et al (2026) ‘Biomimetic SIRPα–CAR Engineering for In Situ Macrophage Reprogramming and Potent Solid Tumor Immunotherapy’, Advanced Functional Materials, https://doi.org/10.1002/adfm.202527483
- Nucleocytosis: Negishi, H. et al. (2026) ‘cGAS-IFN-I responses by extracting nuclear DNA from dying cells via nucleocytosis’, Nature Communications, https://doi.org/10.1038/s41467-026-68839-w
- Cell death: Kawamura, K. et al (2026) ‘Cellular Mg2+ decrease causes a distinctive NF-κB-dependent form of cell death’, Cell Reports, https://doi.org/10.1016/j.celrep.2026.116964
- Immunotherapy: Wu, L. et al (2026) ‘Controlled pyroptosis of engineered macrophages enables biphasic antitumor via the release of oncolytic bacteria and inflammatory signals’, Cell Reports, https://doi.org/10.1016/j.celrep.2025.116918
- Protein aggregation: Mallon, S. et al (2026) ‘Polyalanine Expansion in PABPN1 Alters the Structure and Dynamics of Its Nuclear Aggregates in Differentiated Muscle Cells’, FASEB Journal, https://doi.org/10.1096/fj.202501097R
Find over 400 publications featuring Nanolive imaging here.
Subscribe to this monthly newsletter to stay up to date with AI applications in cell biology by clicking the button at the bottom of the page.
Read our latest news
Seminar – Label-Free Live Cell Analysis: Applications in Cancer Cell Death and Immune Engineering
Measuring CAR T potency with label-free imaging
