In our efforts study the brain simultaneously at different spatiotemporal scales we have long been developing and optimizing two powerful methodologies. These are first, in vivo connectivity based on MR-detectable paramagnetic transsynaptic tracers, such as manganese, e.g.(Saleem et al., 2002, Murayama et al., 2006, Eschenko et al., 2010), and second, direct electrical stimulation and fMRI combined (DES-fMRI) (Tolias et al., 2005, Logothetis et al., 2010). The latter permits the investigation of many different large-scale interactions, including the organization of projective fields, large-scale effects of local synaptic plasticity changes, and the spatiotemporal profile of neuromodulatory effects induced by diffuse ascending systems. Yet, our most recent pilot physiology-MRI study examining global signal propagation following electrical stimulation has demonstrated that DES most likely silences all cortical areas receiving direct input from the fibers that they excite (Logothetis et al., 2010). Together with results from previous electrophysiology studies, this finding is thought to reflect the synaptic organization of microcircuits and their response to an unnatural spatiotemporal input organization. Specifically, optimal operation of recurrent microcircuits is only ensured with spatio-temporally structured input patterns that respect synaptic delays. ES violates this principle and consequently silences the output of neocortex. To better understand feedforward signal propagation for spatiotemporally blurred inputs, we have conducted two more series of experiments.
The first study aimed to examine whether unstructured whole-field visual stimulation may to some extent resemble the pattern of signal. We used whole-field luminance flicker as stimulus and recorded from areas V1 and V2 during EPI MR imaging. In the second study we have started a systematic mapping of thalamus by local microstimulation and imaging. The aim of this investigation is to map cortico-subcortico-cortical pathways by means of ortho- and antidromic stimulation and try to understand the logic of the replication principle, namely the existence of a cortico-thalamo-cortical connection for any given cortico-cortical connection of two points, each in one cortical area. We stimulate the pulvinar but we also record spontaneous activity and study locally generated characteristic events that may be related to thalamo-cortical communication.
Recording and microstimulation hardware, including electrodes and microdrives, were developed at the Max Planck Institute for Biological Cybernetics. Recording chambers and head holders were positioned stereotaxically on the basis of individual, high-resolution MRI. Stimulation sites were selected so as to ensure reliable electrically evoked BOLD activation on the operculum of the brain. Visual stimuli (structured stimuli or whole-field luminance flicker) were presented binocularly, initially with our own custom-made MR-compatible visual stimulator. Electrical microstimulation was performed with a custom-built constant current source. Current amplitude, pulse duration, train duration and stimulation frequency were controlled digitally using our own TCL-based software and a QNX (Canada) real time operating system.
Results and Initial Conclusions
The stimulation results were presented in detail in a recent publication (Logothetis et al., 2010). The response to luminance flicker was surprising indeed. Figure 1 shows the maps and Figure 2 the temporal dynamic of the BOLD and neural responses to two types of stimulation: structured (counterphase pinwheel) and full-field luminance flicker. Striking and consistence was the patent deactivation of V2. V1 showed PBR in some animal and NBR in others. Physiological responses often showed continuation of stimulus-induced modulation, albeit with a strongly reduced baseline activity. It follows that unstructured, full-field luminance stimulations, just like the electrical stimulation, blurs the normal spatiotemporal input patterns in cortex and strong disrupts signal propagation. Our current experiments in thalamus yielded the first patterns of cortical activation (work in progress). Characteristic is the widespread activation that is expected given the parallel projections of this nucleus to the entire cortex.
Figure 1: BOLD responses to counter-phase and Full-Field Luminance Flicker. The left plots shows response to counterphase flicker. As expected there is strong V1 and V2 activation accompanied by sustained negative BOLD (NBR) in non-stimulated areas.
Figure 2: Time course of BOLD and of neural responses to counter-phase and Full-Field Luminance Flicker. Top row shows responses to structured flicker. Red and Blue curves are BOLD responses in V1 and V2 respectively. Peristimulus histograms show multiunit responses in V1 (left) and V2 (right). Bottom row: same conventions but for full-field luminance flicker.
In the aforementioned studies I have been collaborating with Mark Augath, Josina Goense and Hellmut Merkle on MR in MR imaging, with Alexander Rauch and Nikos K. Logothetis in neuroscientific matters and data analysis, and with Axel Oeltermann in developing all the hardware required to perform the experiments (see details in publications).
- Mapping of functional brain activity in freely behaving rats during voluntary running using manganese-enhanced MRI: Implication for longitudinal studies. NeuroImage 49 2544-2555.
- The effects of electrical microstimulation on cortical signal propagation. Nature Neuroscience 13 1283-1291.
- Tracing neural circuits in vivo with Mn-enhanced MRI. Magnetic Resonance Imaging 24 349-358.
- Magnetic resonance imaging of neuronal connections in the macaque monkey. Neuron 34 685-700.
- Mapping cortical activity elicited with electrical microstimulation using fMRI in the macaque. Neuron 48 901-911.