9 December 2007. Could oxidative stress underlie the pathology of schizophrenia? A paper in the December 7 issue of Science reports that reactive oxygen species mediate some of the effects of the NMDA receptor antagonist ketamine on inhibitory interneurons. Researchers led by Margarita Behrens and Laura Dugan at the University of San Diego, La Jolla, California, demonstrate that activation of NADPH oxidase, an enzyme that generates the toxic reactive oxygen species superoxide, is crucial for ketamine-induced disruption of parvalbumin-expressing interneurons. These inhibitory neurons play a crucial role in regulating excitatory neural networks in the brain, and there is evidence that their activity is compromised in schizophrenia patients. As ketamine and other NMDA-type glutamate receptor blockers induce psychotic behaviors indistinguishable from those seen in schizophrenia patients, the finding raises the possibility that in schizophrenia, like many other nervous system disorders, neurons may be particularly susceptible to oxidative stress.
Linking neurotransmitter hypotheses of schizophrenia
Fast-spiking interneurons expressing the calcium-binding protein parvalbumin (PV) modulate cortical networks via their release of γ-aminobutyric acid (GABA), the major inhibitory neurotransmitter. Postmortem analysis shows that this neuronal phenotype is depleted in the cortex of patients with schizophrenia (see, for example, Beasley et al., 2002; Hashimoto et al., 2003). There is evidence that these particular interneurons have not been lost, but are merely producing less parvalbumin, along with producing less GAD67, a GABA-synthesizing enzyme that is also depleted in schizophrenia (see SRF related news story).
Linking these GABAergic phenomena with glutamatergic signaling via the NMDA receptor, Behrens and colleagues have previously shown that ketamine exposure also depletes interneurons in culture of parvalbumin and GAD67 (Kinney at al., 2006). This suggests that blocking NMDA signaling in GABAergic neurons causes a loss in GABA transmission, which in turn frees glutamatergic neurons from inhibitory innervation and leads to heightened glutamate release. By a hypothetical feedback mechanism, cortical neuronal networks would respond to the excess glutamatergic signaling by reducing the GABAergic signaling through parvalbumin expressing interneurons.
In support of this model, Behrens and colleagues now report that adding the GABA agonist muscimol to mouse neuronal cultures treated with ketamine—in effect, fooling the system into thinking that the GABAergic interneurons are performing normally—protects GABAergic neurons from parvalbumin and GAD67 loss.
Oxidative stress as an underlying mechanism?
But what is it about NMDA antagonists that could cause loss of PV and GABA in the first place? Researchers have recognized that NMDA antagonists cause an increase in reactive oxygen species both in vivo and in vitro (Xia et al., 2002; Zuo et al., 2007), though it is not clear why. But Behrens and colleagues, noting recent reports that NADPH oxidase is found in the brain, wondered if this enzyme may be involved. Recent evidence suggests that the enzyme may have unappreciated roles in cellular communication, but its primary role appears to be generation of highly toxic superoxide for the destruction of engulfed bacteria in phagocytes.
Several lines of evidence suggest that Behrens and colleagues may have identified a pathway crucial to the loss of parvalbumin and GAD67 expression in GABAergic interneurons. They report that prolonged exposure to low levels of ketamine drives superoxide production in primary neuronal cultures, and that both a superoxide scavenger (C3) and an NADPH oxidase inhibitor (apocynin) protect the PV/GAD67 phenotype from ketamine.
In the bottom left panel of the figure above (figure 4B from the article), ketamine produces a red cloud of superoxide in mouse prelimbic cortex, with an accompanying loss of parvalbumin reactivity (green). But pretreatment to reduce superoxide production preserves parvalbumin reactivity as seen in the panels on the right. [Image credit: Behrens et al., Science. 2007 December 7;318:1645-1647. Reprinted with permission from AAAS.]
More significant, perhaps, is that pretreatment with both the scavenger and inhibitor also prevent the effects of ketamine in live animals. Analysis of mouse prefrontal cortex, an area of the brain thought to be particularly important for the psychopathology of schizophrenia, showed that ketamine induces a widespread increase in superoxide, which can be abrogated by C3 or apocynin. The effects extend to the PV-positive interneurons, which retain their parvalbumin and GAD67 when animals are prophylactically protected.
All told, the authors hypothesize that, "NADPH oxidase may be a contributor to oxidative mechanisms involved not only in the psychotomimetic effects of NMDA-R antagonists, but also in schizophrenia and other processes involving increased oxidative stress in the brain.” The work also contributes to the ongoing debate about the induction of schizophrenia-like symptoms in healthy people who take NMDAR antagonists “recreationally” (see comment from John Krystal below).
It is also worth noting that other researchers have recently made a connection between schizophrenia and oxidative stress, namely, a genetic polymorphism in a gene needed for synthesis of the antioxidant peptide glutathione (see related SRF news). It could be that oxidative stress may play a much greater role in the disease than previously anticipated.—Tom Fagan.
Behrens MM, Ali SS, Dao DN, Lucero J, Shekhtman G, Quick KL, Dugan L. Ketamine-induced loss of phenotype of fast-spiking interneurons is mediated by NADPH-oxidase. Specific developmental disruption of disrupted-in-schizophrenia-1 function results in schizophrenia-related phenotypes in mice. Science. 2007 December 7;318:1645-1647. Abstract