The p38 mitogen‐activated protein kinase (MAPK) cascade as well as the enzyme monoamine oxidase‐A (MAO‐A) have both been associated with oxidative stress. We observed that the specific inhibition of the p38(MAPK) protein [using either a chemical inhibitor or a dominant‐negative p38(MAPK) clone] selectively induces MAO‐A activity and MAO‐A‐sensitive toxicity in several neuronal cell lines, including primary cortical neurons. Over‐expression of a constitutively active p38(MAPK) results in the phosphorylation of the MAO‐A protein and inhibition of MAO‐A activity. The MAO‐A(Ser209Glu) phosphomimic – bearing a targeted substitution within a putative p38(MAPK) consensus motif – is neither active nor neurotoxic. In contrast, the MAO‐A(Ser209Ala) variant (mimics dephosphorylation) does not associate with p38(MAPK), and is both very active and very toxic. Substitution of the homologous serine in the MAO‐B isoform, i.e. Ser200, with either Glu or Ala does not affect the catalytic activity of the corresponding over‐expressed proteins. These combined in vitro data strongly suggest a direct p38(MAPK)‐dependent inhibition of MAO‐A function. Based on published observations, this endogenous means of selectively regulating MAO‐A function could provide for an adaptive response to oxidative stress associated with disorders as diverse as depression, reperfusion/ischemia, and the early stages of Alzheimer’s disease.

The concentration of presenilin-1 (PS-1) protein at the mitochondrial-associated aspect of the endoplasmic reticulum supports the potential for a mitochondrial influence of PS-1. Given that carriers of certain Alzheimer's disease (AD)-related PS-1 variants are predisposed to clinical depression and that depression has been historically associated with the mitochondrial enzyme, monoamine oxidase-A (MAO-A), we investigated cortical MAO-A function in the AD-related PS-1(M146V) knock-in mouse. The MAO-A system was clearly altered in the PS-1(M146V) mouse as revealed by (a) a mismatch between MAO-A protein expression and MAO-A activity; (b) changes in MAO-A-mediated monoaminergic neurotransmitter metabolism; (c) changes in non-cognitive behavior following treatment with the irreversible MAO-A inhibitor clorgyline; and (d) an increase in the potency of clorgyline in these same mice. We next investigated whether PS-1(M146V) could be influencing MAO-A directly. We observed (a) an enhanced MAO-A activity in necropsied PS-1(M146V) mouse cortical extracts incubated with DAPT (a PS-1 substrate-competitor); (b) the proximity of PS-1 with MAO-A and mitochondrial markers in cortical sections and in primary cortical neurons; (c) the co-segregation and co-immunoprecipitation of PS-1 and MAO-A within the mitochondrial fraction; and (d) the co-immunoprecipitation of overexpressed PS-1(M146V) and MAO-A proteins from N2a lysates. The PS-1(ΔEx9) and PS-1(D257A) variants, known to have low substrate-binding capacity, co-immunoprecipitated weakly with MAO-A. These combined data support a physical interaction between PS-1 and MAO-A that could influence MAO-A activity and contribute to the monoaminergic disruptions common to disorders as seemingly diverse as depression and AD.

Monoamine oxidase-A (MAO-A) has been associated with both depression and Alzheimer disease (AD). Recently, carriers of AD-related presenilin-1 (PS-1) alleles have been found to be at higher risk for developing clinical depression. We chose to examine whether PS-1 could influence MAO-A function in vitro. Overexpression of selected AD-related PS-1 variants (wildtype, Y115H, ΔEx9 and M146V) in mouse hippocampal HT-22 cells affects MAO-A catalytic activity in a variant-specific manner. The ability of the PS-1 substrate-competitor DAPT to induce MAO-A activity in cells expressing either PS-1 wildtype or PS-1(M146V) suggests the potential for a direct influence of PS-1 on MAO-A function. In support of this, we were able to co-immunoprecipitate MAO-A with FLAG-tagged PS-1 wildtype and M146V proteins. This potential for a direct protein-protein interaction between PS-1 and MAO-A is not specific for HT-22 cells as we were also able to co-immunoprecipitate MAO-A with FLAG-PS-1 variants in N2a mouse neuroblastoma cells and in HEK293 human embryonic kidney cells. Finally, we demonstrate that the two PS-1 variants reported to be associated with an increased incidence of clinical depression [e.g., A431E and L235V] both induce MAO-A activity in HT-22 cells. A direct influence of PS-1 variants on MAO-A function could provide an explanation for the changes in monoaminergic tone observed in several neurodegenerative processes including AD. The ability to induce MAO-A catalytic activity with a PS-1/γ-secretase inhibitor should also be considered when designing secretase inhibitor-based therapeutics.