Faculty Bibliography

Intracellular recording and in vitro high-speed voltage-sensitive dye imaging were combined to investigate the role of electrotonic coupling in the generation and distribution of subthreshold oscillations in the inferior olivary (IO) nucleus. Brainstem slices from 2-3 week old rats were stained with a voltage-sensitive fluorescence dye RH-414 (Molecular Probes) and imaged with a fast CCD camera (Fujix HRDeltaron 1700; 128 x 128 pixels of spatial resolution and 4.8 ms of temporal resolution). Comparison of spatio-temporal profiles of subthreshold IO oscillations in control conditions and in the presence a specific gap junctional blocker (18-beta glycyrrhetinic acid) was implemented. In control conditions, spontaneous IO oscillations emanated from multiple clusters of coherently oscillating neurons. Addition of 18-beta glycyrrhetinic acid had a blocking effect such clusters. However, single-cell oscillations were still observed in concurrently obtained intracellular recordings from individual IO cells. Small but significant differences in the amplitude and frequency of the 'uncoupled' and control neuronal oscillations were detected. In addition, the cells had higher resistance (10-20% increase) and lower capacitance (20-30% decrease) in the presence of the blocker. We conclude that gap junctions are not necessary for generating subthreshold IO oscillations but are required for oscillatory synchronization and the clustering of coherent oscillatory activity in the inferior olivary nucleus

T-588 is a promising drug in neurodegenerative disorders treatment. We investigated the neuroprotective effect of T-588 and a newly synthesized compound, T-817MA on Purkinje cell (PC) damage produced by high-potassium concentration (HK) in cerebellar organotypic slice culture. Cerebellar slices were prepared from 10-13 day old rats and cultured on microporous membrane in horse serum medium for 12-day period. These slice cultures were then pre-treated with several concentrations of T-588 or T-817MA for 24 hours, and then exposed to 30 mM potassium medium for 24 hours, in the continuous presence of the protective compounds. Following these procedures slices were maintained overnight in normal medium containing 20 muM Propidium iodide (PI) to detect damaged cells. In the absence of protective compounds, HK produced PI-positive cell damage mostly in the Purkinje layer. At ultrastructural level HK exposed PCs demonstrated swollen cell nuclei without chromatin condensation, varying degrees of swollen ER, damaged mitocondria and stacked ER at the dendrites. Swirling myelin figures were observed in the molecular layer. The neuroprotective effect of T-588 and T-817MA was detected using total fluorescence intensity. Both T-588 and T-817MA significantly reduced PI fluorescence intensity at 1-10 muM and 0.01-1 muM, respectively. In addition T-588 also prevented deterioration of general Purkinje morphology, including organelle ultrastructural characteristic known to be sensitive to cellular damage. These findings indicate that HK treatment induces necrotic damage to PCs in cerebellar organotypic slice cultures and that both T-588 and T-817MA are significant neuroprotective compounds that may be used as therapeutic agents in the treatment of neurodegenerative disorders

Cerebellar slices from 11-13 day old rat brains were cultured for 2 weeks in Minimum Essential Medium with standard 5 mM KCL. At 14 day in vitro (DIV14), calbindin immunoreactivity and Nissl staining revealed many surviving Purkinje cells (PCs) but few granule cells. Ultrastructually, the neuropil was composed mostly of PC dendrites with small and long thin spines but few presynaptic boutons, since granule cell axons (GC parallel fibers) and boutons were negligible. Three experimental groups, exposed to increased K+ levels (10, 20 and 30 mM KCL) between DIV 7-14, had increased number of GCs but fewer PCs with shortened dendritic arbors. EM analysis revealed prominent long segments of parallel fibers and enlarged boutons with abundant synaptic vesicles. Bouton size increased consecutively with each higher concentration of KCL. Some large boutons were opposed to PC plasma membrane and formed multiple synaptic contacts with spines emerging from PC somata and various sized dendrites. Occasionally, 20 or more spine heads contacted each giant bouton. On the other hand, PCs had atrophy with cytological changes in mitochondria and swelling of PC dendrites with stacks of ER. This study shows that above normal K+ enhances GC axonal growth to form giant boutons and supports previous findings that GCs in dissociated cell cultures survive best with 25-30 mM KCL

The sources of the neuromagnetic fields recorded in Magnetoencephalography (MEG) are distributed compactly as sparse clusters of high current density in functionally interactive cortical and subcortical neural networks. Recursive minimum norm algorithms (e.g., FOCUSS) iteratively sharpen and eventually prune an initial distributed solution by down-weighting the contribution of low-amplitude dipoles. Most nonzero variables are not needed to explain the data and can be eliminated by iteratively adding a null space vector to the previous solution that minimizes a diversity measure and guarantees feasibility. Computer simulations with randomly selected source configurations of increasing order and with sparse extended sources were carried out to quantify the spatial resolution of recursive minimum-norm algorithms with several spatial error metrics. Different a priori weights and diversity measures were tested. Results demonstrate that stable super-resolution for high-order source models can only be achieved by using an optimal a priori weight matrix (full column normalization), followed by a sequence of iterations which minimize an optimal p-norm-like diversity measure (p=0.9-1). Further improvement is achieved by directly minimizing a cost of diversity and discrepancy at each iteration with a nonlinear 1-dimensional search of the optimal null vector length, and by simultaneously solving multiple measurement vectors from a time interval. Simulations with non-gaussian activation functions show that Independent Component Analysis (ICA) and the new FOCUSS-learning algorithms can dramatically increase the number of recovered sources. Finally, visually evoked and spontaneous signals are analyzed with these algorithms. Results demonstrate the importance of cortico-thalamo-cortical loops in global brain function

Consciousness is hypothesized to be a process of sensorimotor transformations generated by coherent thalamocortical activity clusters oscillating at gamma-band (30-50 Hz) frequencies. This hypothesis predicts that higher amplitude gamma-band thalamocortical currents will correlate with shorter reaction times (RTs). We tested this prediction using whole-head 148-sensor MEG recordings during a simple RT task. Methods: 1 tactile stimulator under each middle finger, & 1 flat microswitch button under each index finger. Each trial had 1 finger tap, randomized across hands. Same-side responses for 400 trials, then opposite-side for 400 trials. 6 healthy subjects (4 age 20-40, 2 age 75-85). Results: In all subjects, the highest signal/noise ratio (SNR) independent component (IC_1) in somatosensory cortex indicated a large amplitude gamma-band current 30-80 ms post-stimulus. For each millisecond, we calculated correlations across trials between (a) IC_1 amplitude at that millisecond and (b) RT in that trial. Correlation minima ranged down to -0.6 and were synchronous with gamma-band amplitude maxima. Gamma-peaks and synchronous correlation minima occurred in both cortical hemispheres apprx50 ms in both response conditions, indicating that inter-hemispheric gamma-band currents correlate with faster RTs within and across hemispheres. Gamma-peaks and synchronous RT correlation-minima occurred in young and elderly subjects, but all elderly subjects exhibited lower gamma-band amplitudes, lower correlations, and longer RTs than young subjects. Conclusion: Early (apprx50) evoked 30-50 Hz activity within & across somatosensory cortices exhibits strong correlations with sensorimotor speed in individual reaction-time trials

Ongoing studies indicate that the conjunction of spontaneous thalamocortical activity, at low-(theta; 4-8Hz) and high-(gamma; 25-50Hz) frequencies serves as the physiological basis for a set of disorders we have termed the thalamocortical dysrhythmia syndromes (TCD). Elements in this set are proposed to include Parkinson's disease, tinnitus, central pain, OCD, depression and schizoaffective (SA) disorder. The common denominator is a prominent theta-range oscillation underlying negative symptoms, in temporal coherence with gamma band activity relating to positive symptoms. Results demonstrate that localization of TCD activity is possible and add a more direct functional dimension to results obtained with other imaging techniques.Continuous neuromagnetic activity was recorded with whole-head MEG (4D Neuroimaging) from 6 subjects with SA disorder and 8 healthy controls. Multitaper spectral estimation was used to calculate frequency spectra, and independent components (ICs) were derived with EEGlab software. Selected ICs were localized in a probabilistic sourcespace. A recursive weighted minimum norm algorithm was used to calculate solutions for current density localization.Power spectra from controls demonstrated typical occipital alpha rhythm, while spectra from SA subjects showed an increase in theta power localized in mediofrontal supraorbital and temporal areas. These results support a model in which deinactivation of thalamic T-type Ca++ currents leads to localized oscillatory changes. The presence of both frontal and temporal activity in individual ICs suggests functional synchronization between these areas in SA disorder and corroborates findings of low-frequency oscillation with metabolic hypofrontality in PET studies.It is concluded that ICA may identify and localize abnormal TC dynamics in SA disorder and that MEG represents an important tool in the investigation of TCD patients

Calcium mobilization in cerebellar Purkinje cell (PC) of CD38 knockout mouse was imaged using two photon microscopy. CD38 is a homolog of ADP-ribosyl cyclase which activates Cyclic ADP-ribose, an endogenous Calcium Induced Calcium Release (CICR)regulator (Lee, 2001 ). In the CD38 knockout mouse (CD38(-/-)), which showed no phenotypic motor abnormalities, whole cerebellar cyclic ADP-ribose (cADPR) was reduced to a quarter of that in a wild control (CD38(+/+)). In addition caffeine treated CD38(-/-) PC dendrite showed a markedly reduced calcium response to a test depolarizing pulse most probably related to the reduced CICR. Cerebellar long-term depression (LTD) induced by the parallel fiber (PF) -climbing fiber (CF) conjunctional stimulation was absent in CD38(-/-) PC. However, during very prolonged direct depolarization of Purkinje cell somata, or when the test (PF) stimulation was quite close to the cell body during direct somatic depolarization, LTD could be induced in CD38(-/-) PC by the PF/depolarizing pulse conjunction. Overall, the CD38/cADPR system seems to be sensing/regulating intradendritic calcium levels. This seems to be an important modulator of the integrative properties of Purkinje cell activity. It seems clear that if (Ca)i becomes high enough to be deleterious to neuronal viability such system must be down regulated via LTD as a neuroprotective response to reduce further calcium increase. Immuno-staining study with a polyclonal CD38 antibody is presently in progress which is expected to find different levels of CD38 on the Purkinje cell soma, dendrites and spines of wild type mouse (CD38(+/+)) vs CD38 knockout cerebella

The olivo-cerebellar network plays a key role in the organization of vertebrate motor control. The oscillatory properties of inferior olive (IO) neurons have been shown to provide timing signals for motor coordination in which spatio-temporal coherent oscillatory neuronal clusters control movement dynamics. Based on the neuronal connectivity and electrophysiology of the olivo-cerebellar network we have developed a general-purpose control approach, which we refer to as a universal control system (UCS), capable of dealing with a large number of actuator parameters in real time. In this UCS, the imposed goal and the resultant feedback from the actuators specify system properties. The goal is realized through implementing an architecture that can regulate a large number of parameters simultaneously by providing stimuli-modulated spatio-temporal cluster dynamics