Introduction. Polypharmacy is common in old age and is associated with adverse geriatric outcomes. Deprescribing medications may alleviate some outcomes. The liver is a key metabolic organ and is affected by drugs and ageing. However, the molecular effects of chronic monotherapy, polypharmacy, and deprescribing in the ageing liver remains uncharacterised.
Aims. In aged mice, to investigate (i) the effect of chronic monotherapy, polypharmacy, and deprescribing on the hepatic proteome; and (ii) the relationship between the hepatic proteome and geriatric outcomes.
Methods. Healthy C57BL/6J mice aged 12 months commenced either (i) control, (ii) chronic monotherapy (oxybutynin, oxycodone, citalopram, simvastatin, or metoprolol at therapeutic doses), or (iii) chronic polypharmacy (all 5 monotherapies). At 21 months, treated mice were re-randomised to continue or deprescribe medications. Livers and serum were collected at age 26 months. Proteomics analysis was performed by data-independent acquisition (DIA) using a Q-Exactive (Hfx) orbitrap mass spectrometer (n=5-8/group). Serum drug levels were measured. Differential expression and pathway analyses were applied. To assess the relationship between the drug-induced proteome reorganisation and geriatric outcomes, we applied Weighted Gene Co-expression Network Analysis (WGCNA) to model the changes of the proteome as a co-expression network and assess systems-level association.
Results. Compared to control, several proteins were differentially expressed with monotherapy (oxybutynin: 26, oxycodone: 38, citalopram: 74, simvastatin: 36, metoprolol: 55; p < 0.05, fold-change threshold: ±1.50). With polypharmacy, 243 proteins were differentially expressed, of which 166 (68%) were unique to polypharmacy. The polypharmacy proteome signature is characterised by proteins involved in drug metabolism, immune function, and cholesterol biosynthesis. Serum levels of citalopram, noroxycodone, and metoprolol were 2-fold higher in polypharmacy than monotherapy (p < 0.05). Complete deprescribing reverses up to 50% of changes in the hepatic proteome but also introduce novel effects. The proteome was partitioned into 22 co-expressed protein clusters. Distance travelled (r = 0.22, p < 0.05) and rotarod endurance time (r = 0.38, p < 0.05) correlated with amino acid metabolic proteins.
Discussion. Polypharmacy altered expression of hepatic proteins, to a greater extent than all monotherapy, with partial reversal after deprescribing. Drug metabolism, immune function and cholesterol biosynthesis may drive effects of this polypharmacy regimen. Amino acid metabolic proteins were associated with adverse geriatric outcomes.