Research Projects

Below is a description of both active and future research projects at ICNDE. The scientific process is not linear. Promising projects are always being considered, while less successful ideas are often altered or put on hold. The description below is meant to paint a picture of the type of work ICNDE is passionate about. However, ICNDE is not restricted to the research described below.

Optical Measurement and Control of Neuronal Activity
Dynamics of Seizure

Optical Measurement and Control of Neuronal Activity
 
Perhaps the major obstacle to studying collective excitations in the brain has not been our lack of well-developed theories of collective phenomena, but rather our paucity of experimental techniques for measuring the activity of large numbers of individual neurons. Developing new optical techniques to both control and record neuronal activity is one of the major initiatives at ICNDE. This approach can be used as an alternative to, or in conjunction with electronic techniques to study spatiotemporal patterns in neuronal networks during both physiological and seizure activity. Optical based techniques have the following advantages:

i) Light is minimally invasive
In contrast to electrodes, the interaction of light with biological tissue is typically limited, particularly at near infrared wavelengths where the absorption of light and heat dissipation are minimal. The majority of tissue damage incurred from these techniques comes from exciting the light active molecules used in these experiments. This toxicity varies with the molecule utilized, and with the exception of voltages sensitive dyes discussed below is often not problematic.

ii) Experiments using light can be done in a highly parallel fashion
Light can be delivered and recorded from thousands of individual cells or several brain areas simultaneously. This is essential for studying neuronal activity that not only involves large numbers of cells, but also exhibits evolving behavior that must be captured in a parallel fashion. In contrast, electrode arrays are limited from a handful to at most a few hundred electrodes.

iii) The spatial resolution of light spans many orders of magnitude
Optical resolutions from the single molecule scale to the size of the brain are achievable. Although challenging, a wide range of scales can be probed in the same experiment. Electronic techniques generally require utilizing different electrodes to probe the neuronal activity at different spatial scales.

iv) Light can probe cellular and molecular processes other than electrical activity
Light active molecules can be engineered to measure and control gene expression, protein function, cellular pH, as well as ion and neurotransmitter concentrations. In addition, fluorescent retrograde viral markers have been used to map out synaptic connections in the brain. One can imagine using a variety of light sensitive probes to simultaneously study the relationship between multiple neuronal processes including electrical activity.

Initially, ICNDE plans to focus its efforts on three projects in this area.

1) Designing new light activated molecular sensors for neuronal activity
While many optical probes exist for neuronal activity, all of them exhibit features that severely limit their utility. Ca sensitive fluorescent probes measure the influx of calcium into the cell that follows action potential generation, and report that information by emitting light. While this signal is relatively strong, it is only an indirect measure of neuronal activity, as many signaling processes in the cell can alter Ca concentration. Also, Ca influx is slow and additive, making it difficult to resolve neuronal spiking and postsynaptic potentials. Fast voltage sensitive dyes have been developed that probe electrical activity with speeds commensurate with neuronal spiking. However, in practice these sensors suffer from poor signal to noise ratios, and experiments utilizing them rely heavily on spatial or temporal averaging. In addition, these fluorescent probes often exhibit significant cell toxicity. ICNDE hopes to develop new optical sensors for neuronal activity that overcome these limitations.

2) Developing novel optical techniques for observing neuronal activity
In addition to developing optically sensors, one must have a way to efficiently excite these molecules and collect their fluorescence emission. Typically an image of the fluorescence emanating from a 2D section or slice of neuronal tissue labeled with optical probes is captured onto a 2D sensor, such as a digital camera. However, optically exciting a thin slice of tissue is difficult, and often there is significant fluorescence contribution coming from tissue outside the region of interest that contaminates the signal. Researchers have developed techniques such as confocal microscopy and two-photon microscopy to eliminate this problem. However, these techniques measure the fluorescence at one small focused spot at a time. To measure the fluorescence from the slice, one must move or scan a focused laser beam throughout the entire sample of interest. The faster or larger the spatiotemporal pattern one wishes to observe the faster one must scan the laser beam. However, scanning the beam with increasing speed results in spending less time at, and collecting less light from each spot in the sample. Unfortunately photon emission is inherently noisy, and the less photons one collects the noisier the signal becomes. In this photon-limiting situation, one has to trade temporal resolution or sampling area for signal to noise. Clearly, neither compromise is ideal in measurements of extended patterns of neuronal activity that change with time. ICNDE hopes to develop new technologies and explore novel geometries for measuring optical signals from large numbers of neurons simultaneously, hoping to overcome the above limitations.

3) Developing optical techniques for treating seizures
Several chemically caged compounds have been developed that become activated with light. In addition, researchers have engineered proteins and genes that become turned on upon exposure to light. Since the brain is enclosed in a dark space, it is an ideal environment for selective manipulation with light sensitive agents.

Seizures often begin in a focused brain region, and can spread to involve the entire cortex. It is thought that these foci are pathological and likely the cause of the abnormal activity. Thus, these areas are often targeted by surgical interventions. In contrast, when treating seizures pharmacologically, systemic agents are given that bath the entire brain, regardless of the seizure origin. We hope by utilizing light activated agents we can take advantage of the efficacy and noninvasive nature of pharmacological agents while exceeding the anatomical specificity of surgical treatments. By focusing light and activating molecules in small anatomical areas and sparing the rest of the brain, we hope to deliver high doses of anticonvulsants with a fraction of the side effects. In addition, with novel protein based agents it may be possible to target specific cell populations that may be responsible for seizure activity. Lastly, in combination with future seizure detection technology, this approach would be able to deliver anticonvulsant therapy as the seizure is starting, or even prior to seizure onset. Due to the minimization of side effects, focal delivery should be more forgiving of the false positives that are inevitable in any seizure detection algorithm. We believe that given the current progress in caged compounds, light activated anticonvulsant therapy is not only feasible, but can have a significant impact on seizure therapy in the near future.

Dynamics of Seizure

What initiates seizure activity? Can we learn about the cause of seizure from its dynamics and evolution? What causes seizures to terminate? Through both animal models and EEG data from humans, we hope to address these questions.

1) Seizure activity in humans
EEG records the projection of the brain's electrical activity on the surface of the head. Although this technique has several weaknesses including poor spatial resolution, it allows for noninvasive measurement of neuronal activity in humans.
As described in the introductory material, seizure activity on EEG displays wide spread synchronized oscillatory activity. In addition to this feature, seizures typically have a well-defined lifetime, lasting on the order of a minute. What causes this apparently sudden onset of synchrony, and why it abruptly stops within this characteristic time scale is not known. Interestingly, our preliminary work on the EEG of patients with seizures reveal that both the synchrony and frequency of oscillations evolve during the seizure in manner that appears to be universal. We hope to construct simple models that will not only give us insight into these universal features and the dynamics of seizures, but also help us understand the mechanisms responsible for this pathological neuronal activity.

2) Animal models of seizures
Using a variety of techniques, including those described earlier, we hope to understand the microscopic basis for the seizure dynamics observed in humans. Here, more detailed and invasive measurements in animal models can be used to test mechanisms of seizure formation and the dynamics predicted by our theoretical models. In addition, we hope that these techniques and models will provide a foundation for similar experiments involving the study of spatiotemporal patterns in physiological activity.