Purpose: Alzheimer’s disease (AD) is a major cause of dementia in the elderly. Excessive deposition of amyloid beta (AB) in the brain and peripheral circulation represents one of the major pathological hallmarks of AD. Type II diabetes has recently emerged as an important risk factor for AD. Brain insulin levels in AD patients at progressive stages of the disease are significantly reduced relative to that of age-matched controls. Insulin present in brain is predominantly derived from the peripheral circulation via transport at the blood brain barrier (BBB). Lower insulin levels and insulin signaling defects in AD brain are claimed to accelerate disease progression by contributing to impaired neuronal function and brain energy metabolism. We hypothesize that AB exposure to the BBB endothelium reduces insulin delivery to brain and perturbs downstream insulin signaling in the neurovascular unit. In order to fully resolve the mechanisms of brain insulin resistance, it is critical to investigate impaired insulin trafficking at the BBB and altered brain disposition due to AB peptide exposure. The objective of this work was to investigate the transport/elimination kinetics of insulin in WT vs. AD transgenic mice, which express excess AB, and identify rate-limiting steps by constructing a multi-compartment PK/PD model to describe insulin kinetics following an intravenous bolus injection.
Methods: A multi-compartment PK/PD model incorporating central, peripheral, BBB, and brain compartments was constructed using Stella dynamic modeling software. Model input parameters were derived from in vitro studies using human BBB endothelial cell (hCMEC) monolayers, as well as from plasma and brain pharmacokinetic studies conducted in WT and AD transgenic mice. Brain insulin levels simulated by the pharmacokinetic model were linked to a pharmacodynamic model to evaluate AMPA and NMDA receptor recycling dynamics in response to insulin appearance in the brain. Sensitivity analyses were also conducted on the model parameters for WT mice to assess the impact of individual components of AD pathology on insulin brain uptake/elimination.
Results: Based on the simulation output, a small portion of the insulin dose administered intravenously reaches the brain. Simulated brain insulin levels were found to be highly sensitive to changes in the Michaelis-Menten parameters for BBB uptake. Notably, BBB uptake was found to be saturated at physiological doses. Increased central AUC and decreased brain AUC were predicted for AD mice compared to the WT littermates. This can be explained by the decreased clearance from central compartment and lower brain influx clearance predicted for AD mice. A slight decrease in the intensity of the pharmacodynamic response of AMPAR and NMDAR recycling was predicted for the AD mice. Overall, the simulation output was found to be in agreement with experimental and literature data.
Conclusion: Reduced brain insulin delivery due to exposure to toxic AB peptides likely contributes to insulin signaling defects and diminished neuronal function in AD mice.
Suresh Kumar Swaminathan– Research Associate, University of Minnesota, MN 55455
Vidur Sarma– University of Minnesota
Kristen Ahlschwede– University of Minnesota, Minnesota
Geoffry Curran– Mayo Clinic
Teresa Deckler– Mayo Clinic, Minnesota
Krishna Rani Kalari– Mayo Clinic
Val Lowe– Mayo Clinic
Karunya Kandimalla– University of Minnesota