Blood racing through a brain region's web of vessels is a sign that nerve cells in that locale have kicked into action. The blood rushes to active areas to supply firing neurons with the oxygen and glucose they need for energy.

It is this blood flow, which can last up to a minute, that scientists track in functional magnetic resonance imaging (fMRI) to determine which brain areas are responding to different stimuli. But a new theory could pave the way for a reinterpretation of fMRI images, elevating their measurements to the evaluation of actual neuronal processing rather than the subsequent blood flow that indirectly indicates it, and thereby enhancing the fMRI's usefulness in diagnosing neurological problems.

Christopher Moore, an assistant neuroscience professor at the Massachusetts Institute of Technology's McGovern Institute for Brain Research, detailed his hypothesis in a recent article published in the Journal of Neurophysiology. In essence, it suggests blood's role in the cortex (a key brain processing center), specifically, is more than just bringing nutrients to the cell, it can also alter the activity of local neuronal circuits. For instance, in experiments in his lab, Moore has seen that there is more blood flow can arrive in an area that processes information from a presented stimulus to a certain sense (e.g. touch, visual, auditory) prior to the appearance of the stimulus, implying that the flow can prime a circuit for activity, as well.

Researchers estimate that blood flow in areas of the brain increases by 40 percent when neurons start to fire (or send out electronic impulses), whereas the corresponding metabolic rate of the cells only increases by 4 percent, meaning the cell only needs a tenth of the blood it is supplied to reenergize. "[Neuroscientists] call this discrepancy an 'uncoupling' between flow and metabolism," says Kenneth Kwong, an associate professor in radiology at Harvard Medical School and the researcher, along with Seiji Ogawa, who is generally credited with developing fMRI.

Moore believes that the reason for the discrepancy could be that blood not only nourishes cells but may be intimately involved in the information processing.

If true, Moore says, blood should be factored into any model of neuronal processing—how nerve cells in the brain are activated, how impulses are transmitted between them, how long activity lasts, and how it is terminated. In addition to changing what fMRI is actually measuring, such models could potentially provide new clues to causes of enigmatic disorders such as Alzheimer's disease, multiple sclerosis and schziophrenia—potentially paving the way for treatments that involve correcting blood flow as well as (or rather than) chemical deficiencies.

"Historically, fMRI researchers have to be a little apologetic because they're not looking [directly] at the neuron," Moore says. "fMRI would stop being a second-class citizen; instead it would make fMRI a much more interesting tool…a Heisenberg sort of thing [referring to how the act of observing a quantum state changes it], where what you're looking at is actually a part of the computation going on." Further, scans taken over a number of years could help predict neurodegeneration, if vasculature in a particular brain region begins to weaken. Preliminary data already suggest this is the case for many neurological disorders such as schizophrenia.

The so-called hemo-neural hypothesis plays out in three tissue types: neurons, the blood vessels that feed them, and astrocytes, the star-shaped nerve cells that support and maintain neurons. (Astrocytes, the feet of which are splayed on blood vessels, also help maintain the endothelial cells that line the vessels as well as make up the semipermeable blood–brain barrier responsible for keeping chemicals in the blood from seeping into the brain unless they are needed for metabolism or some other function.)

According to Moore, the vasculature thus directly or indirectly (via astrocytes) influences neurons. He notes that substances in blood may modulate neuron activity. The most likely candidate, he says, is nitric oxide (NO), which easily crosses the blood–brain barrier and has been shown both in brain slices and in animal models to excite (and in some cases dampen) neuronal action. Blood vessels also affect neurons via thermal and mechanical stress. Increased blood flow can alter the local temperature in a brain region. For instance, a decrease of just one degree Celsius can lead to suppressed firing rates, in some circumstances. As a rule, blood flow changes increase the temperature in outer brain areas, while decreasing the temperature of more central regions. Pressure and volume, meanwhile, within the blood vessels can change the amount that the vessels physically impact the membranes of brain cells. If pressure or volume were to increase, a vessel could bulge, blocking receptors or ion channels and thereby causing a decrease in a neuron's electrical activity.

A change in blood flow could also trigger astrocytes to release certain hormones or neurotransmitters. "If anything is going on in the blood vessel," Moore says, "the glia (astrocytes and other nonneuronal nerve cells) is in a great position to sense it." For instance, astrocytes might secrete the excitatory neurotransmitter glutamate, which binds to neurons and allows ion exchanges that cause cells to fire.

Preliminary data from Moore's lab, involving a drug that selectively binds to receptors on blood vessels (and can open or close them), has shown neurons may become more active when blood flow increases. The M.I.T. group is now trying to develop light-activated ion channels on muscle cells, which they could then selectively control to induce changes in blood flow.

The theory "brings up something that a lot of people have been ignoring—thinking that blood vessels are tubes," says Edith Hamel, a professor of neurology at McGill University in Montreal, who believes that Moore's theory will one day prove true. "But, they are live cells…like neurons." She likens the vasculature interaction within the nervous system to that of the infrastructure of a highway system. "We have always been looking at the highway going out of the city," she says. "We need to look at the one coming into the city, as well."

Rick Buxton, a radiology professor at the University of California, San Diego, finds the idea intriguing, but he is "skeptical that blood flow is really an important modulator." In his interpretation, the rush of blood is necessary to maintain oxygen levels in the tissue, because neurons may take in oxygen at a slower rate than normal when blood is gushing by; therefore, more is needed to properly nourish the cells. Another possible way to account for the excess blood flow, according to some researchers, is that it may help carry away some of the heat generated by neuronal firing. "If there's some low level of neuromodulation in there, [as well], that's good," Buxton adds.

Moore notes that if the blood is responsible for a relatively low level of neuromodulation, it could still be significant. "Let's say that blood flow accounts for 5 percent of the variance of activity in cortical neurons," he supposes. "Five percent of the neuron's work—that's huge, if it's pushing around excitability by that scale."

Moore's theory is supported by research into neurodegenerative and mental disorders. Constantino Iadecola, a professor of neurology and neuroscience at Weill Cornell Medical College in New York City, for instance, has found a link between blood vessels and neurons in his work on Alzheimer's disease.

"We have provided evidence that the vasculature is the first thing that goes," he says, noting that Alzheimer's-associated dementia was previously split into two groups: vascular-induced (in which neurons die due to improper blood flow) and neurodegeneration-induced (with vasculature collapse following nerve cell death). "What's emerging from the literature now is that the [vessel changes occur] at least as early or earlier than the neuronal changes."

Abnormal blood flow has also been linked to epilepsy, which is caused by overactive neurons. And, according to Moore, an impoverished blood supply is usually noted in the areas of schizophrenia sufferers' brains that go awry in the disorder.

Moore envisions that in the future research on treatments for mental disorders will focus on potential drugs designed to maintain proper neuronal function by targeting vasculature. "It would be beneficial to upregulate and downregulate blood flow," he says, "the same way as it's beneficial to upregulate or downregulate [the neurotransmitter] dopamine in schizophrenics."