Veins on the brain – how the vascular endothelium affects brain function

"I've never had an idea before where everything we've done to test it and every thing I've read, has kept pointing to the fact that it's right. That's why I'm so passionate about trying to tell the world."Elizabeth Hillman does not always speak like this. For the vast majority of a 40-minute interview, the Columbia University associate professor of biomedical engineering gives a suitably academic explanation of her latest paper, 'A Critical Role for the Vascular Endothelium in Functional Neurovascular Coupling in the Brain'.

It is only when asked if she has anything to add before the recorder is turned off that her strength of feeling about her work gets the better of her.

"We lived and breathed this for two years," she says. "I know I sound a little crazy when I start talking about it, but it was just such a difficult thing to get people to reconsider."

Brain fuel

Hillman has spent over ten years studying how bloodflow is controlled in the brain. Put simply, her latest break-through paper argues that the vascular endothelium, already known to regulate flow in other parts of the body, performs the same function in the brain.

This might sound uncontroversial, or even obvious in some respects, but it totally contradicts every existing theory. Even more importantly, it has potentially enormous consequences for how Alzheimer's, and other neural diseases, are understood and treated.

The relationship between the brain's blood flow and the organ's health is, to some extent, already understood. A 400-mile long network of vessels delivers the vital oxygen and glucose to neurons in order for them to fire.

This system is extremely efficient: when an area of the brain is working particularly hard, blood is quickly redirected to it to refuel the busy neurons. This redirection is controlled by vessel dilation, whereby a wider opening attracts more blood.

What has never been entirely clear is precisely how the blood vessels know when to dilate in the first place. How does the signal reach them?

Challenging the orthodoxy

Since 2005, the dominant theory has been that astrocytes - star-shaped glial cells that can be found clinging to the brain's blood vessels - controlled dilation. It is thought that these sense glutamate from the neurons, then release prostaglandins, causing the vessels to expand. In vivo experiments have also shown that activating an astrocyte can cause dilation.

"That idea went into the textbooks and people really loved it," says Hillman. "Before that, we hadn't really known what astrocytes did. So there was a sense of relief that we'd found a reason for them to be there."

Broken link

Over the last decade however, plenty of other reasons have been found to explain the astrocytes' presence. What's more, Hillman's team was unable to find any real evidence that the astrocytes were actually responsible for controlling the blood-vessel dilation.

"We built a really high-speed microscope, and the astrocytes just didn't seem to be responding quickly enough to be responsible for the change in blood flow," she says. "Then, last year, a very comprehensive paper came out concluding that a link between the astrocytes and blood vessels simply couldn't be proven."

It is also thought that pericytes - known to spiral around the brain's capillaries - are able to squeeze and relax and thereby manipulate blood vessel dilation. But while these cells certainly can exert some control, whether or not they respond directly to fuel-hungry neurons remains unclear.

Unconvinced by either existing explanation, Hillman continued to experiment. She noticed that pial arteries, which navigate the brain's surface and are not connected to either astrocytes or pericytes, still dilated when nearby neurons were stimulated.

Intrigued, she began to read around the subject, and happened upon a paper about endothelial hyperpolarisation: a mechanism causing dilation along the arterial endothelium.

This, she realised, could explain how the dilatory signal travelled along the vessels and made them expand. "It was being studied in the rest of the body by people who had nothing to do with brain research, and it just looked so perfect. The dynamics, the range, the speed - it all fitted with what we'd been observing," she says.

"I grabbed the paper, ran into my student's office, and said, 'Look, look, I think I've found it.' It was one of those amazing things where the more we read about it and the more questions we asked, the more it fitted in with everything we'd been seeing.

"So then all we had to do was design an experiment that would prove this was actually happening in the brain."

Speed camera

The experiment Hillman's team eventually came up with was unconventional, but highly effective. Rather than relying on fMRI (functional magnetic resonance imaging), a camera imaging at 75 frames per second was used to look directly at the blood vessels of rats' brains.

By shining different colours of light, it was possible to identify changes in vessel dilation and blood oxygen dynamics. The pictures produced were far more detailed than fMRI techniques (which are often limited in their effectiveness by their sensitivity to deoxygenated haemoglobin, making it difficult to see changes in overall levels) would have allowed.

Proof positive

The new method offered clear readings for total haemoglobin levels, letting Hillman's team see exactly when blood vessels were dilating. Next, the team had to prove that it was indeed the endothelium that was conducting the vessel dilation signal during neuronal stimulation. Somehow, it had to be knocked out, which would reveal whether or not it genuinely was part of the process.

"We found this really old technique from about 50 years ago called 'light dye' to help with that," says Hillman. "You inject a fluorescent dye into the vasculature, then use a very bright blue light to illuminate it at its excitation wavelength. This releases oxygen free radicals, which effectively frazzle the endothelium.

"If you do it for exactly the right amount of time, the effect is limited to just that endothelial layer, so you don't have to worry about damaging anything else in the system."

Paws for thought

Now ready to conduct the experiment, Hillman's team stimulated the rat's paw and observed the haemodynamic response in action. The light-dye technique was then used to knock out the vascular endothelium in one of the responding arteries, and the rat was stimulated again. As expected, the vessel failed to dilate past the damaged area.

"That demonstrated that the signal was coming along the endothelium," says Hillman. "Then we illuminated the whole surface of the responding region's cortex to damage the entire top layer of endothelium, and that entirely eradicated all the dilation. It caused a really significant decrease in blood flow change in response to the stimulus."

Hillman's research suggests that the signal for blood-vessel dilation during neuronal stimulation travels along the vascular endothelium; within the vessel itself, rather than through any external cells.

"This changes the picture, it gives another way for vessels to dilate, even when there doesn't seem to be a local trigger," says Hillman.

These results also caused Hillman and her team to explore conditions in which the vascular endothelium might be damaged, or unable to function normally. The team realised that a wide range of systemic conditions often associated with cognitive decline, from caridovascular disease to diabetes, could lead to endothelial dysfunction.

The reasoning is that if blood vessels are damaged and no longer able to do as they are told, neurons may fail to function properly and could ultimately become starved of nutrients. For some brain diseases, the suspicion is that problems previously attributed to faulty neurons could in fact be due to bad blood vessels.

Vein hopes

If this is indeed the case, then we might be much closer to treating such disorders than previously thought. The existing battery of drugs for blood vessel malfunction, the effects of which are already understood, could also be used here.

What's more, it may not be necessary for drugs to cross the blood-brain barrier in order to save neurons; restoring vascular function could prevent neuronal damage in some cases.

Hillman's work provides an alternative explanation to fMRI studies that identify altered BOLD responses or resting state functional connectivity networks in different brain disorders - that the vascular system may not always faithfully report underlying neuronal activity. Hillman hopes that models based on endothelial mechanisms could ultimately make fMRI data significantly easier to interpret.

Logical leap

One important question that the team is yet to answer, though, is how the dilatory signal jumps from the firing neuron to the vascular endothelium in the first place.

"It could go directly from neuron to endothelial cell... they have all kinds of receptors. Or it could go via the astrocytes or pericytes," says Hillman.

"The honest answer is that we don't know at the moment, and we're trying to get to the bottom of it. But we can do a lot with these animal models.

"And we're doing a ton of studies now that let us look at the brain in both its resting state and in stimulus evoked conditions. We're also doing lots of work on brain development."

A growing number of scientists are paying attention to Hillman's work. But, as with any new discovery, there remain plenty still wedded to the older theories.

In her own words, Hillman has yet to be accepted by "the neurovascular coupling people: most of them still think it's the astrocytes".

Further evidence is undoubtedly needed before Hillman and her team's research enters the mainstream, and still more before it leads to changes in brain treatment. But if this does eventually happen, the results could be mindblowing.