Magnetic nanoparticles enable magnetic resonance imaging of neuron
activity in brain
7 November 2006
Cambridge, USA. New chemical sensors that indicate the firing of neurons
in the brain and show up strongly in magnetic resonance imaging will open
the way for new research into the way the brain works.
Neuroscientists eagerly anticipate the day when they can use non-invasive
brain imaging to see precisely what the 10 billion neurons are doing
throughout a person’s brain. But a trio of limitations facing current
functional magnetic resonance imaging (fMRI) technology stands in the way of
that goal: time, space, and specificity.
At the McGovern Institute for Brain Research at the Massachusetts
Institute of Technology (MIT), Alan Jasanoff is developing new chemical
sensors that are detectable by MRI machines and will overcome these
limitations. The first of these tools, a nano-sized calcium contrast agent,
is reported in the October edition of the
Proceedings of the National Academy of Sciences.
|At top, a
schematic shows how superparamagnetic nanoparticles coated with two
proteins (red and green) form mixed aggregates due to
calcium-dependent protein-protein interactions.
The middle panel shows atomic force micrographs of
actual sensor particles in the absence (left) or presence (right) of
calcium. The bottom panel shows progressive brightening of MRI
images (green to red pseudocolour) as the calcium concentration is
increased from zero (left) to 5 micromolar (right).
“Using conventional fMRI to study the brain is like trying to understand
how a computer works by feeling which parts of it are hot because of energy
dissipation in different components of the machine,” explained Jasanoff.
“But chemical sensors for MRI could show what each individual element in
each integrated circuit is doing and how it performs the computations and
The analogy is apt, because fMRI indirectly measures neural activity by
detecting changes in blood flow to brain regions with increased energy
requirements. However, these homodynamic changes occur several seconds after
the neurons actually fired, too slow to study precise neural activity. The
spacing of the capillaries limits the spatial resolution of the technique to
volumes containing at least 1,000 neurons, too coarse for discrimination of
highly specialized functional areas within a brain region.
Calcium, however, provides a direct measure of neural activity because
calcium almost instantly flows into neurons when they fire, and the faster
the rate of firing, the higher the calcium level. Thus, tracking calcium
levels in the brain actually tracks information flow through the brain’s
“The changes are pretty dramatic,” Jasanoff said. “Calcium concentrations
can vary by more than an order of magnitude, and the fluctuations are
relatively long-lasting.” Jasanoff drew his inspiration for this contrast
agent from optical microscopy, which uses light to study cellular properties
and has long targeted calcium as a way to image neural signals. But optical
microscopy can only penetrate about 2 millimeters and cannot image the
deeper brain tissues, while MRI could detect contrast agents throughout the
To be visible to MRI, which detects changes in magnetic properties, a
contrast agent must include a magnetically-active “paramagnetic” component.
Previous approaches to designing MRI calcium sensors used only one or two
metal atoms, to create a relatively weak magnetic effect that required high
concentrations of the agent for detection. Such high concentrations of a
calcium-sensitive agent could potentially swallow up enough calcium to
interfere with neural circuitry in vivo.
To overcome this limitation, Jasanoff designed the sensor to incorporate
so-called “superparamagnetic nanoparticles” — extra strength molecular-sized
magnets that had previously been designed for ultrasensitive tumour imaging.
Because the new calcium sensor based on these particles produces large MRI
contrast changes, it may be used at low concentrations that will not perturb
Jasanoff’s sensor is actually made from two similar types of
superparamagnetic nanoparticles. One is attached to a short corkscrew-shaped
protein segment called M13, and the other is attached to another protein
called calmodulin, that binds to M13 in the presence of calcium. When
calcium levels rise, the two types of particle stick to each like
Velcro-coated balls. They form aggregates that in turn affect MRI contrast.
A characteristic of sensors that work this way is called “T2 relaxivity.”
This property makes the sensors suitable for use in powerful MRI scanners
capable of producing very high-resolution images. Also important to the new
calcium sensor’s function is the fact that its aggregation and readout are
reversible. This property allows it to indicate the temporal dynamics of
calcium-related neural activity, such as the sequence in which populations
of cells become active, or the synchronization of neurons during certain
Jasanoff is currently working on non-invasive methods to deliver the
calcium sensor to brain cells in vivo, focusing on small animals like flies
and rats. Graduate student Tatjana Atanasijevic, who is the lead author on
the PNAS paper, is optimizing properties of the sensor for applications in
animals. Meanwhile, others in Jasanoff’s group are changing the makeup of
the nanoparticle so that it can target specific genetic characteristics or
different populations of neurons, such as inhibitory or excitatory neurons
or those that produce specific neurotransmitters.
“These will be tools for making the shift from imaging gross functional
properties of the brain through its haemodynamic changes to a fine tuned
analysis based on information flow involving cells and circuits,” Jasanoff
said. “There are many potential applications for studying learning, memory,
and behaviour, and we need the new tools to get to the applications.”
In addition to his appointment as an associate member of the McGovern
Institute, Jasanoff is Assistant Professor in the Departments of Nuclear
Science & Engineering, Brain & Cognitive Sciences, and Biological
Engineering Division. This research is supported by grants from the Raymond
& Beverley Sackler Foundation and the NIH/NIBIB, and a McKnight Foundation
Technological Innovations in Neuroscience award.