Therapeutic advantage of pro-electrophilic drugs to activate the Nrf2/ARE pathway in Alzheimer's disease models

Abstract

Alzheimer's disease (AD) is characterized by synaptic and neuronal loss, which occurs at least partially through oxidative stress induced by oligomeric amyloid-β (Aβ)-peptide. Carnosic acid (CA), a chemical found in rosemary and sage, is a pro-electrophilic compound that is converted to its active form by oxidative stress. The active form stimulates the Keap1/Nrf2 transcriptional pathway and thus production of phase 2 antioxidant enzymes. We used both in vitro and in vivo models. For in vitro studies, we evaluated protective effects of CA on primary neurons exposed to oligomeric Aβ. For in vivo studies, we used two transgenic mouse models of AD, human amyloid precursor protein (hAPP)-J20 mice and triple transgenic (3xTg AD) mice. We treated these mice trans-nasally with CA twice weekly for 3 months. Subsequently, we performed neurobehavioral tests and quantitative immunohistochemistry to assess effects on AD-related phenotypes, including learning and memory, and synaptic damage. In vitro, CA reduced dendritic spine loss in rat neurons exposed to oligomeric Aβ. In vivo, CA treatment of hAPP-J20 mice improved learning and memory in the Morris water maze test. Histologically, CA increased dendritic and synaptic markers, and decreased astrogliosis, Aβ plaque number, and phospho-tau staining in the hippocampus. We conclude that CA exhibits therapeutic benefits in rodent AD models and since the FDA has placed CA on the 'generally regarded as safe' (GRAS) list, thus obviating the need for safety studies, human clinical trials will be greatly expedited.

Figure 1

Carnosic acid (CA) treatment ameliorates Aβ-induced dendritic spine loss in cultured rat neurons. (a) Images of rat cortical neurons transfected with pmax-GFP and then exposed to synthetic Aβ₄₂ (250 nM oligomers) with or without CA (10 μM) for 4 days. After fixation, dendritic spines were visualized by GFP fluorescence. (b) Quantification of spine density per micrometer of dendritic length. Neurons exposed to oligomeric Aβ manifested significantly reduced dendritic spine density, and treatment with CA ameliorated this loss. Values are mean+S.E.M. (n>20 fields of dendritic spines for each condition; *P<0.05 by ANOVA compared with other conditions). Scale bar, 5 μm

Figure 2

CA treatment improves spatial learning in hAPP-J20 mice. (a) J20 and WT littermate controls were assessed for learning in the Morris water maze test after treatment with CA or vehicle (Control). Evaluation of time to find the hidden platform during training in the four groups of mice indicates that CA treatment improved spatial learning in J20 mice (n⩾6 mice/group; **P<0.03 by two-way ANOVA with Tukey's multiple comparisons test). (b) Following training, a probe test, in which the hidden platform was removed, was performed to assess spatial memory. Mice were scored for the time spent in the bottom left quadrant (prior location of the platform) versus the top right quadrant (n⩾5 mice/group; **P<0.01 by Student's t-test). The performance of J20-CA mice is not significantly different from those of WT-Control and WT-CA mice by two-way ANOVA with Tukey's multiple comparisons test. Dashed line indicates time spent in a quadrant by chance alone (15 s)

Figure 3

CA treatment reverses deficits in neuropil and synaptic density in hAPP-J20 mice. Quantitative immunohistochemistry of hippocampal sections prepared from wild-type (WT) and J20 mice treated with CA or vehicle (Control). (a) Immunohistochemistry with MAP2 antibody. CA treatment restored MAP2 staining/neuropil density to normal levels in the hippocampus of hAPP-J20 mice. (b) Immunohistochemistry with synaptophysin antibody. CA treatment restored synaptophysin staining/synapse density to normal levels in both cortex and hippocampus of hAPP-J20 mice (n=4–10 mice/group; **P<0.01 by ANOVA). Scale bar, 250 μm

Figure 4

CA treatment prevents reactive astrocytosis and blocks accumulation of Aβ protein aggregates in hAPP-J20 mice. Quantitative immunohistochemistry of hippocampal sections from wild-type (WT) and J20 mice treated with CA or vehicle (Control). (a) Immunohistochemistry with GFAP antibody. CA treatment decreased GFAP staining/astrocytosis in the hippocampus of hAPP-J20 mice. (b) Immunohistochemistry with Aβ antibody. Aβ protein aggregates were significantly decreased after treatment with CA (n=4–10 mice/group; **P<0.01 by ANOVA). Scale bar, 250 μm

Figure 5

CA treatment improves AD brain markers in 3xTg AD mice. Quantitative immunohistochemistry of hippocampal sections from 3xTg AD mice treated with CA or vehicle (Control). After CA treatment, 3xTg AD hippocampus displayed increased MAP2/neuropil and synaptophysin/synaptic densities, and reduced GFAP/astrocytosis and p-tau (n=8–9 mice/group; *P<0.05, **P<0.01 by t-test). Scale bar, 40 μm

Figure 6

Schematic model showing action of a pro-electrophilic drugs (PEDs) in activating the Nrf2 transcriptional pathway. In this case the PED Carnosic Acid (CA) is activated by reactive oxygen species (ROS) to the active quinone form. This activated form of the drug reacts with a critical thiol (-SH) group on the cytoplasmic protein Keap1, which releases the transcription factor Nrf2. Nrf2 then is free to enter the nucleus where it transcriptionally activates a series of endogenous, neuroprotective antioxidant and anti-inflammatory proteins known as phase 2 enzymes