MIT researchers say they have built an AI-powered \u201cbarcode\u201d that can identify aging \u201czombie\u201d cells \u2014 senescent cells that stop dividing but don\u2019t die \u2014 by pairing Raman microscopy with spatial RNA sequencing, MIT News reported Monday. The paper appears in Nature Aging.<\/p>\n
Senescent cells accumulate with age and are linked to inflammation, tissue degeneration, cancer, and other age-related disease, though they also play beneficial roles in development and regeneration. Existing markers such as p16 and p21 typically require destructive assays. Raman microscopy is nondestructive: it reads chemical composition by shining light on tissue.<\/p>\n
The team studied mouse skin and lung from 2-month-old versus 26-month-old animals and is adapting the method for human tissue. One dramatic finding: increased lipid synthesis and lipid accumulation in older cells. Effects were tissue-specific \u2014 skin showed shifts in muscle-contraction, collagen, and extracellular-matrix remodeling pathways; lung showed immune-activation and inflammation genes.<\/p>\n
Current scans take about 30 hours for roughly one square millimeter of sample; researchers are building a higher-speed system. Senior authors include Jeon Woong Kang and Peter So of MIT\u2019s Laser Biomedical Research Center, and Jian Shu of MGH\/Harvard Medical School (Broad and Ragon associate). Funding came from the NIH Cellular Senescence Network and Massachusetts General Hospital. Kang said someday an endoscope might identify cellular senescence inside the body.<\/p>\n