Pegging Pathology on Mitochondrial Dysfunction

MALFUNCTIONING MITOCHONDRIA

Recently, however, high-throughput microarray technologies have revealed evidence of striking mitochondrial-function deficits from the postmortem brains of people with bipolar disorder and schizophrenia. These results are backed up by brain-imaging studies. But causal relationships between mitochondria and dysfunction are far from established; differences in experimental design confound direct comparisons among studies. Nevertheless, current clinical treatments appear to boost mitochondrial function. If drug efficacy depends on this effect, then directly targeting the mitochondrion could provide the basis for a generation of more effective treatments.

Three recently published papers implicate mitochondrial abnormalities in diseased brains. Though some dissimilarities emerge, they support a mitochondrial dysfunction hypothesis. Christine Konradi and collaborators at McLean Hospital, Belmont, Mass., and Harvard Medical School detected a striking decrease in the expression of genes associated with mitochondrial energy metabolism in postmortem bipolar brains.²

Sabine Bahn and colleagues at the University of Cambridge studied brains from patients with schizophrenia and found evidence for mitochondrial dysfunction and oxidative stress; her group confirmed gene-expression results by measuring protein and metabolite changes within the same brains.³ Stephen Dager and collaborators at the University of Washington, Seattle, and Harvard Medical School detected brain chemical elevations in medication-free patients with bipolar disorder, consistent with a shift in mitochondrial function away from oxidative phosphorylation and toward glycolyis.⁴ This, Dager says, suggests an underlying mitochondrial abnormality.

"It's wonderful news to people in the neuroimaging field that some of our results are starting to converge with other modalities such as postmortem studies," he adds, referring specifically to Konradi's work. Bahn says that she also has unpublished findings of mitochondrial gene-expression changes in bipolar brains consistent with Konradi's results, but she adds that she found more striking changes in the expression of other genes.

AGREEING TO DISAGREE

The mitochondrial dysfunction and oxidative stress Bahn sees in schizophrenia brains are consistent with long-standing observations of abnormal glucose metabolism in the prefrontal cortex, she says. "We are quite confident in our results," says Bahn, "because all three [methodologies we used] actually point in the same direction."

Unlike Bahn's study, Konradi's reported no significant mitochondria-related changes in schizophrenia. But Bahn points out that the two studies examined distinct brain regions, a difference Konradi agrees is important. Bahn's group used a chip carrying twice as many genes and sampled five times as many brains, which might have resulted in differential representation of schizophrenia subtypes between the studies.

Konradi agrees that sample size matters, and she points out that different brain banks may use slightly different diagnostic criteria. Rather than being a weakness, she says this may precipitate the identification of biological markers by which to identify distinct subtypes and more appropriately target an individual's treatment.

Other difficulties nevertheless plague the postmortem approach. Vahram Haroutunian, at Mount Sinai School of Medicine and Bronx Veterans Affairs Medical Center, was among the first (another being Middleton) to apply microarray analysis to postmortem schizophrenia brains. Both say they're concerned that the circumstances associated with death – sudden or prolonged, or occurring in the middle of winter versus the height of summer, for example – confound analysis of brain energy metabolism. But Konradi's and Bahn's studies, says Haroutunian, "have been well controlled, and they're obviously attentive to these issues."

Dissecting cause from effect remains problematic. "We're looking always at the end stage of a very chronic, remitting, relapsing disorder," says Bahn. "It's quite difficult to say what exactly led to the situation that we find ... Is it the mitochondria being damaged first, and then oxidative stress, or the other way around?" Mitochondria generate oxidative stress, and are susceptible to its damaging consequences.

Moreover, mitochondrial density is highest at the synapse, says Middleton, and schizophrenia is often associated with decreased synaptic density for certain neuron classes in the prefrontal cortex. "The reason we may see decreased expression of certain transcripts related to those energy pathways is that there's fewer synapses to begin with," says Middleton. But brain architecture, Middleton adds, depends on normal metabolic activity.

Furthermore, while convergence of ideas may strengthen an argument, commonality can also hurt the case for causality. Arturas Petronis at the Center for Addiction and Mental Health and the University of Toronto points out that mitochondrial dysfunction has been detected in many other pathological conditions, including specific brain disorders as well as nonspecific processes such as inflammation and aging. "It would be interesting," he says, "to find out to what extent those findings are specific to major psychosis ... to some cells, to some neurons, to some brain regions."

Haroutunian and Bahn are particularly interested in whether expression changes in the brain occur within a particular cell type. To this end both say they plan to use laser-capture microscopy to profile individual cells and cell types, a technique that bridges the historically disparate fields of neuroanatomy and molecular biology. Bahn and Haroutunian both note one attractive possibility: a link between mitochondrial dysfunction and findings of oligodendrocyte dysfunction in schizophrenia and bipolar disorder. "Oligodendrocytes are the most vulnerable cells to oxidative stress," says Bahn.

SEEKING THE ROOT CAUSE

The highlighted studies do not attempt to address the primary defects leading to pathology or mitochondrial dysfunction. Bahn says she has some indication that analyzing cerebrospinal fluid from new patients who have never been treated may reflect early events in pathology.

Middleton and his collaborators are looking to linkage studies. They have recently published a genome-wide linkage analysis of bipolar disorder indicating that a small spot on chromosome 6 shows a strong linkage signal.⁵ He says they are following up this study by looking for expression abnormalities in the same population; if the two sets of findings converge, it will certainly strengthen the argument for a specific chromosomal region.

Regardless of the primary defect, intriguing indications suggest that targeting the mitochondrion may have clinical benefits. Husseini Manji at the National Institute of Mental Health says that he and others, including RIKEN's Kato, are proposing mitochondria as integral to the pathophysiology and treatment of bipolar disorder. Manji says the most effective current treatments for bipolar disorder and schizophrenia influence genes and proteins that regulate mitochondrial function.

These current treatments work through a long cascade of processes, adds Manji, and he says that something more directly targeting the mitochondrion may work better. One candidate is a dietary supplement, coenzyme Q10, a component of the mitochondrial respiratory chain. Manji says he hopes that CoQ10 can be used to benefit both symptoms and long-term disease progression in patients with bipolar disorder.

Bahn, meanwhile, is taking advantage of her large sample size, provided by the Stanley Medical Research Institute, Bethesda, Md., and using the meticulous details concerning patient history, to dissect the influence of drug treatment on gene expression. She says she will publish these findings separately. Dager says he is following up his imaging study of medication-free patients by looking at brain-chemical changes during treatment.

Whatever the ultimate answers, Haroutunian says that these studies are having what he thinks is a huge beneficial effect on mental illness research by broadening horizons. " [They] add to the neurotransmitter focus that has been so prevalent for the last 30 years," he says. " [They] say that there is something more to brain function than just dopamine, or just serotonin."

A. Nicola Schweitzer nschweitzer@mac.com is a freelance writer in Boston.