Towards frailty biomarkers: Candidates from genes and pathways regulated in aging and age-related diseases

Objective: Use of the frailty index to measure an accumulation of deficits has been proven a valuable method for identifying elderly people at risk for increased vulnerability, disease, injury, and mortality. However, complementary molecular frailty biomarkers or ideally biomarker panels have not yet been identified. We conducted a systematic search to identify biomarker candidates for a frailty biomarker panel. Methods: Gene expression databases were searched (http://genomics.senescence.info/genes including GenAge, AnAge, LongevityMap, CellAge, DrugAge, Digital Aging Atlas) to identify genes regulated in aging, longevity, and age-related diseases with a focus on secreted factors or molecules detectable in body fluids as potential frailty biomarkers. Factors broadly expressed, related to several “hallmark of aging” pathways as well as used or predicted as biomarkers in other disease settings, particularly age-related pathologies, were identified. This set of biomarkers was further expanded according to the expertise and experience of the authors. In the next step, biomarkers were assigned to six “hallmark of aging” pathways, namely (1) inflammation, (2) mitochondria and apoptosis, (3) calcium homeostasis, (4) fibrosis, (5) NMJ (neuromuscular junction) and neurons, (6) cytoskeleton and hormones, or (7) other principles and an extensive literature search was performed for each candidate to explore their potential and priority as frailty biomarkers. Results: A total of 44 markers were evaluated in the seven categories listed above, and 19 were awarded a high priority score, 22 identified as medium priority and three were low priority. In each category high and medium priority markers were identified. Conclusion: Biomarker panels for frailty would be of high value and better than single markers. Based on our search we would propose a core panel of frailty biomarkers consisting of (1) CXCL10 (C-X-C motif chemokine ligand 10), IL-6 (interleukin 6), CX3CL1 (C-X3-C motif chemokine ligand 1), (2) GDF15 (growth differentiation factor 15), FNDC5 (fibronectin type III domain containing 5), vimentin (VIM), (3) regucalcin (RGN/SMP30), calreticulin, (4) PLAU (plasminogen activator, urokinase), AGT (angiotensinogen), (5) BDNF (brain derived neurotrophic factor), progranulin (PGRN), (6) α-klotho (KL), FGF23 (fibroblast growth factor 23), FGF21, leptin (LEP), (7) miRNA (micro Ribonucleic acid) panel (to be further defined), AHCY (adenosylhomocysteinase) and KRT18 (keratin 18). An expanded panel would also include (1) pentraxin (PTX3), sVCAM/ICAM (soluble vascular cell adhesion molecule 1/Intercellular adhesion molecule 1), defensin α, (2) APP (amyloid beta precursor protein), LDH (lactate dehydrogenase), (3) S100B (S100 calcium binding protein B), (4) TGFβ (transforming growth factor beta), PAI-1 (plasminogen activator inhibitor 1), TGM2 (transglutaminase 2), (5) sRAGE (soluble receptor for advanced glycosylation end products), HMGB1 (high mobility group box 1), C3/C1Q (complement factor 3/1Q), ST2 (Interleukin 1 receptor like 1), agrin (AGRN), (6) IGF-1 (insulin-like growth factor 1), resistin (RETN), adiponectin (ADIPOQ), ghrelin (GHRL), growth hormone (GH), (7) microparticle panel (to be further defined), GpnmB (glycoprotein nonmetastatic melanoma protein B) and lactoferrin (LTF). We believe that these predicted panels need to be experimentally explored in animal models and frail cohorts in order to ascertain their diagnostic, prognostic and therapeutic potential