Network-driven plasma proteomics expose molecular changes in the Alzheimer’s brain

Philipp A. Jaeger, Kurt M. Lucin, Markus Britschgi, Badri Vardarajan, Ruo-Pan Huang, Elizabeth D. Kirby, Rachelle Abbey, Bradley F. Boeve, Adam L. Boxer, Lindsay A. Farrer, NiCole Finch, Neill R. Graff-Radford, Elizabeth Head, Matan Hofree, Ruochun Huang, Hudson Johns, Anna Karydas, David S. Knopman, Andrey Loboda, Eliezer Masliah, Ramya Narasimhan, Ronald C. Petersen, Alexei Podtelezhnikov, Suraj Pradhan, Rosa Rademakers, Chung-Huan Sun, Steven G. Younkin, Bruce L. Miller, Trey Ideker, Tony Wyss-Coray

Biological pathways that significantly contribute to sporadic Alzheimer's disease are largely unknown and cannot be observed directly. Cognitive symptoms appear only decades after the molecular disease onset, further complicating analyses. As a consequence, molecular research is often restricted to late-stage post-mortem studies of brain tissue. However, the disease process is expected to trigger numerous cellular signaling pathways and modulate the local and systemic environment, and resulting changes in secreted signaling molecules carry information about otherwise inaccessible pathological processes. To access this information we probed relative levels of close to 600 secreted signaling proteins from patients' blood samples using antibody microarrays and mapped disease-specific molecular networks. Using these networks as seeds we then employed independent genome and transcriptome data sets to corroborate potential pathogenic pathways. We identified Growth-Differentiation Factor (GDF) signaling as a novel Alzheimer's disease-relevant pathway supported by in vivo and in vitro follow-up experiments, demonstrating the existence of a highly informative link between cellular pathology and changes in circulatory signaling proteins.