A new biosensor combining several technologies is taking a significant step forward in the diagnosis of neurodegenerative diseases such as Parkinson’s and Alzheimer’s disease. Named ImmunoSEIRA by its creators, researchers at the École Polytechnique Fédérale de Lausanne (EPFL) in Switzerland, the device makes it possible to detect and identify protein biomarkers associated with these diseases. The treatment of neurodegenerative diseases faces a major challenge due to the lack of effective diagnostic methods for early detection and monitoring of disease progression. Protein misfolding, a common mechanism in neurodegeneration, has been identified as an important event in disease progression. Healthy proteins are assumed to degenerate first into oligomers in the early stages and into fibrils in later stages of the disease. These aggregates of misfolded proteins circulate in the brain and biofluids and also accumulate as deposits in the brains of deceased people with neurodegenerative diseases. However, the development of tools to detect these telltale signs of disease, known as biomarkers, has so far remained tentative. Obstacles to accurate detection are varied, among them technological limitations to accurately separate and quantify different protein aggregates. The EPFL research, published in Science Advances, relied on Artificial Intelligence (AI), employing neural networks to quantify the stages and progression of diseases. According to the researchers, this technological advance is promising not only for early detection and monitoring, but also for evaluating treatment options at various stages of neurodegenerative disease progression. To create this biosensor, researchers from the Bionanophotonic Systems Laboratory and the Molecular Neurobiology and Neuroproteomics Laboratory combined several fields of science, from protein biochemistry, optofluidics to nanotechnology and AI. “Unlike current biochemical approaches that rely on measuring the levels of these molecules, our approach focuses on detecting abnormal structures. This technology also allows us to differentiate the levels of the two main abnormal forms implicated in the development and progression of neurodegenerative diseases, oligomers and fibres,” explains Professor Hilal Lashuel, from the second lab involved in the study. The ImmunoSEIRA biosensor employs a technology called surface-enhanced infrared absorption spectroscopy (SEIRA). This method allows detecting and analysing the forms of biomarkers associated with neurodegenerative diseases. The sensor uses a unique immunoassay that acts as a molecular detective, identifying and capturing these biomarkers with high accuracy. The ImmunoSEIRA sensor features gold arrangements with antibodies for detection of specific proteins. It enables real-time specific capture and structural analysis of target biomarkers in extremely small samples. Neural networks are then employed to identify the presence of specific misfolded protein forms, the oligomeric and fibrillar aggregates, achieving an unprecedented level of detection accuracy as diseases progress. Lashuel believes this is a significant advance in the detection of these diseases. “Since the disease process is strongly associated with changes in protein structure, we believe that structural biomarkers, especially when integrated with other biochemical and neurodegeneration biomarkers, can pave the way for more accurate diagnosis and monitoring of disease progression,” Lashuel emphasises. The research team went a step further and showed that the ImmunoSEIRA biosensor can be used in real clinical settings, i.e. in biofluids. It was possible to accurately identify the specific signature of abnormal fibrils, an important indicator of neurodegenerative diseases, even in complex fluids such as human cerebrospinal fluid. The next phase of this new technology will be to continue expanding its capabilities and evaluate its diagnostic potential in Parkinson’s disease and the growing number of diseases caused by protein folding and aggregation. Multipurpose biosensor On the other hand, scientists from UC Santa Cruz presented chip-based biosensors that allow a single device to perform multiple medical tests simultaneously on different biomolecules, even at very different concentrations. The solution uses machine learning to recognise particles with high accuracy, enabling data analysis in real time. New signal processing techniques have been applied to an optofluidic chip-based biosensor to perform continuous fluorescence detection of a mixture of nanospheres at concentrations over eight orders of magnitude (atmolar to nanomolar). This extends the concentration range in which these sensors can work by more than 10,000 times, explains a story on the SciTechDaily website. The researchers also applied an extremely fast algorithm to identify signals from single particles at low concentrations in real time. Machine learning helped recognise signal patterns so that different types of particles could be distinguished with high accuracy.