Neuronal Networks Laboratory Sorting the connections with multi-electrode

Neuronal Networks Laboratory Sorting the connections with multi-electrode neuronal ensemble recording techniques: Single-unit and local field potential activity from rat to man Dr Rob Mason Institute of Neuroscience School of Biomedical Sciences University of Nottingham Medical School and the LAB TEAM

Neuronal Networks Laboratory LAB TEAM epilepsy Ben Coomber Clare Roe [Dr Mike O’Donoghue ~ schizophrenia Dr Jill Suckling Dr Dissanayake [Prof CA Marsden] anxiety Dr Carl Stevenson [Prof CA Marsden] pain Dr Steve Elmes [Dr V Chapman] rodent USVs Beth Tunstall [Dr S Beckett] data analysis Margarita Zachariou [Prof S Coombes / Dr M Owen] Dept Neurology, QMC] [Mathematics Dept] Dr David Halliday (University of York) Prof D Auer (f. MRI / ph. MRI ~ QMC)

Neuronal Networks Laboratory Sorting the connections with multi-electrode neuronal ensemble recording techniques To & From ~ rat prefrontal cortex Seminar overview • Multichannel Electrophysiological Recording Technologies Illustrate with reference to 2 experimental rodent model projects: • Hippocampus-m. PFC ~ Epilepsy model – role of endocannabinoid system • Sensory gating in hippocampus & m. PFC ~ Schizophrenia model – PCP effects • Amygdala-m. PFC ~ Stress models – maternal separation & “drug stressors” • Left-Right m. PFC ~ • Rat language ~ ultrasound vocalisations (USVs) & affective state

Sorting signal from noise Ensemble Neuronal Unit activity LFP activity DATA Analysis & Interpretation

Distribution of neurones contributing to signals recorded by tetrode array after Busaki 2004

Examples of Electrode Arrays 4 -channel - independent manipulation Michigan MEAprobe 16 -channels Bionics 100 -channel (“hedgehog”) array NBLabs 8 -channel array

MULTIPLE ELECTRODE ARRAY RECORDING in vivo NBLabs 16 -channel micro-wire array • CA 1 & CA 3 • Single-site recording e. g. hippocampal sub regions (CA 1 & CA 3) • Multiple-site recording e. g. prefrontal cortex & hippocampus

Multiple Electrode Recordings in vivo simultaneous multichannel spike & field potential recording • 64 -channel MAP system • 32 -channel MAP system • 64 -chanel Recorder system • 16 -chanel Recorder system – with Cine. Plex for behavioural studies Ø used MAP & Recorder for in vitro MEA studies with brain slices & neuronal cultures

Extracellular Recording - signal filtering & discrimination Raw Signal ( V) 100 0 Units + EEG / LFP Filtered Signal ( V) 100 Units 0 Impulse Events 0 0. 10 0. 20 TIME (s) AP Spike discrimination: separate action potential (AP) signal from noise • AP amplitude detection / AP waveform shape recognition Signal filtering: separate unit activity and Local Field Potentials (LFPs)

Multiple Neuronal (spike) Recording - two electrodes Electrode 1 Electrode 2 micro-electrodes Amplifier Neurones nerve impulses (action potential “spikes”)

Off Line Sorting of spike waveforms - single unit isolation

Movie: Sorting of unit spike data using Principal Component Analysis • distinct AP spike waveforms represented as clusters in 3 D space • 7 units isolated – each unit colour-coded

Multiple (ensemble) neurone recording Advantages • Investigating neuronal ensemble/network function - closer to • Good experimental design - fewer animals required • Masses of data Disadvantages • Masses of data ? Data processing ? Data interpretation working “brain” (“ 3 Rs” ~ Home Office)

ENSEMBLE DATA - DISPLAY & ANALYSIS MASSIVE data sets Data Visualisation Emergent Properties Population Dynamics Unit activity Local Field Potentials Single units – spike rasters /FRH / ISIH / PSTHs burst analysis / Unit pairs – cross-correlation / coherence / Unit ensembles – PCA / ICA / synchrony index / PDC FFT / spectrograms LFP/EEG signal bands LFP-unit coherence LFP PDC

Neuronal Networks Laboratory • Dual/Triple site recordings ~ 64 channels simultaneous units & LFPs VTA - m. PFC hippocampus - m. PFC amygdala - m. PFC – m. PFC spinal cord – thalamus – cerebral cortex • Systemic pharmacological manipulation • Local pharmacological manipulation “injectrode” integrated microiontophoresis with recording array • Electrical stimulation – periphery / CNS structures • Independent electrode array microdrive – 8 channel drive • Anaesthetised & awake-behaving preparations

LAB EXPERIMENTAL DIRECTIONS m. PFCx interconnectivity & functional context – in vivo studies schizophrenia drugs of abuse medial Pre-Frontal Cortex Contralateral m. PFC VTA bladder PVt SCN circadian - affective states Hippocampus - Sensory Gating - Schizophrenia - Epilepsy nucleus Accumbens USVs - affective states Amygdala maternal separation - Depression - Stress

Neuronal Networks Laboratory University of Nottingham Medical School Neuronal Networks Laboratory Project #1: Epilepsy • kainate-induced epileptiform activity ~ TLE • functional network interactions ~ hippocampus m. PFC • role of endocannabinoid system ~ CB 1 R pharmacology • perforant path stimulation-evoked seizure activity

Endo-CANNABINOID System • Endogenous cannabinoids (e. CBs) identified e. g. anandamide (AEA) and 2 -arachidonylglycerol (2 -AG) • Act at cannabinoid G protein coupled-Receptors: CB 1 & CB 2 - ? CB 3 • e. CBs are synthesised postsynaptically on-demand. • e. CBs act at pre-synaptic CB 1 receptors. • e. CB reuptake occurs via a transporter. FAAH MGL Reuptake Figure taken from Wilson & Nicoll (2002) Science. • Metabolism: – AEA by fatty acid amide hydrolase (FAAH) – 2 -AG by monoacylglycerol lipase (MGL).

Epilepsy Study: AIMS • Kainic acid (KA): established convulsive agent producing seizures in awake rats ~ targeting temporal lobe Model uses KA, administered systemically (10 mg/kg, i. p. ) anaesthetized rats ~ ensemble neuronal unit and LFP activity • Study aims to establish whether URB 597, selective inhibitor of FAAH enzyme ( e. CB levels), attenuates KA-evoked neuronal activity • Role of CB 1 cannabinoid receptors assessed using selective CB 1 antagonist AM 251 URB 597 Anandamide Metabolism by FAAH Inhibits anandamide metabolism

Unit and LFP activity in m. PFC and hippocampus Basal: Rat #1 KA + Vehicle: Rat #1 Basal: Rat #2 KA + URB 597: Rat #2

• Effects on hippocampal neural firing rate (~40 cells; n=5 rats) • Effects on spike-triggered averaging of m. PFC LFPs Post-KA administration

Cross-correlation analysis ~ PFC – PFC / PFC- Hippocampal neuronal pairs # units n=2 Ref unit n=12 n=7

Cross-correlation analysis of PFC neuronal pairs • Effect of kainate (10 mg/kg; i. p. ) administration at epoch 4 Time-Series cross-correlograms – TF #1 -13 correlation strength Cross-correlogram TF #1 Time Frame #1 -13 Time (s) 0 Time (s) -1 s correlation strength +1 s Time Frames #1 -13

PDC Analysis – unit ensemble data • PDC was applied to identify the direction of activity between hippocampus and m. PFC - technique that has the potential to reveal the neuronal ensemble drives. Note: the magnitude of the classical coherence gives no information about directional connectivity - but its phase may do so.

Neuronal Networks Laboratory University of Nottingham Medical School Neuronal Networks Laboratory Project #2: Schizophrenia • auditory-evoked sensory gating ~ hippocampus & m. PFC • effects of PCP / ketamine ~ model • effects of social isolation ~ model

Neuronal Networks Laboratory Sensory gating in hippocampus: A model for schizophrenia ? • Sensory gating: mechanism(s) by which irrelevant sensory information is filtered ~ enables efficient information processing • Auditory Conditioning-Test paradigm: measures reduction in auditory-evoked response produced by Test stimulus following a Conditioning stimulus • Stimuli: 3 k. Hz sine-wave / 10 ms duration / presented 500 ms apart / 80 -90 d. B • human P 50 wave = rat N 40 component • Gating absent in - schizophrenic patients (& family) - normal volunteers given PCP / amphetamine - rats given PCP / amphetamine

Hippocampal CA 3 - auditory-evoked unit & LFP activity Units event tone LFPs 1 s LFP 2 averaged 128 trials -1 Trial # - 128 Averaged LFP m. V N 40 T/C ratio = 55% Cs Ts s

Single-unit PSTHs ‘ rasters histograms – “gating rats” dentate gyrus CA 3 region Spike raster Unit 1 PSTH Unit 2 Unit 3 Unit 4 LFP trial raster LFP

Hippocampal CA 3 auditory-evoked unit & LFP activity Effects of PCP (1 mg/kg i. p. ) attenuates / abolishes sensory gating Basal [128 trials] T/C ratio = 32% i. e. exhibits gating 45 mins after PCP T/C ratio = 66% gating attenuated 1 Trail # 128

SUMMARY III Sensory Gating Studies LFP studies: • Demonstrate sensory gating in isoflurane-anaesthetised rat : T/C ratio = 35 ± 15% • SG is abolished / attenuated following PCP : T/C ratio = 65 ± 5% • control rats SG is unaffected by clozapine : T/C ratio 40% • Clozapine (5 mg. kg-1) blocks action of PCP on SG : T/C ratio 35% Unit studies: • Similar observations

Neuronal Networks Laboratory Project #3: Rat ultrasound vocalisations (USVs) & affective state • Sorting the pips from the squeaks • Role of nuc Accumbens in USV-mediated Behaviours ~50 k. Hz (Reward) call (Brudzynski, 2001) • nucleus Accumbens - m. PFCx functional connectivity

Neuronal Networks Laboratory CURRENT APPROACHES & FUTURE DIRECTIONS Human Studies Neuro-Robotics & Neuro. Prosthetics Hybrid Brain-Machine interfaces (HBMIs) • Cochlear implants • Monitoring & control of epileptic seizures • Robotic limbs

(A) Seizure control (B) Robotic arm control

Chips in the Brain ~ Brain-Machine Interfaces (BMIs) ~ Brain-Computer Interfaces Current issue (May 2007) Scientific American MIND 18 (2): 65 -69

Brain-Computer Interfaces Signal choice & Algorithm Unit activity vs LFPs

Brain-Computer Interfaces

Electrode array Implant Recording Work. Station – Plexon Inc

(1) ENSEMBLE RECORDINGS OF HUMAN SUBCORTICAL NEURONS AS A SOURCE OF MOTOR CONTROL SIGNALS FOR A BRAIN-MACHINE INTERFACE Parag G. Patil et al - Neurosurgery (2004) – Duke University ~ 32 -channel Pt. Ir 40 m wire array ~ 4 Deep Brain Stimulation electrodes [Medtronics DBS] Unit recordings – 4 microwires Thalamic VOP/VIN STN [Plexon Inc MAP recording system]

(2) ENSEMBLE RECORDINGS OF HUMAN SUBCORTICAL NEURONS AS A SOURCE OF MOTOR CONTROL SIGNALS FOR A BRAIN-MACHINE INTERFACE Parag G. Patil et al - Neurosurgery (2004) – Duke University Patient performance motor task STN recording ~ 24 units

REFERENCES Human / Primate Recording Reviews • LEARNING TO CONTROL A BRAIN–MACHINE INTERFACE FOR REACHING AND GRASPING BY PRIMATES JM Carmena, et al PLo. S Biology ~ http: //biology. plosjournals. org 1(2): 193 -208 (2003) • ENSEMBLE RECORDINGS OF HUMAN SUBCORTICAL NEURONS AS A SOURCE OF MOTOR CONTROL SIGNALS FOR A BRAIN-MACHINE INTERFACE PG Patil, JM. Carmena, Miguel AL Nicolelis & DA Turner Neurosurgery 55(1): 27 -38 (2004) • ASSISTIVE TECHNOLOGY & ROBITC CONTROL USING MOTOR CORTEX ENSEMBLEBASED NEURAL INTERFACE SYSTEMS IN HUMANS WITH TETRAPLEGIA JP Donoghue et al J. Physiol 579(3) 603 -611 (2007)

Neuronal Networks Laboratory University of Nottingham Medical School Neuronal Networks Laboratory That’s all folks Lab web site www. nottingham. ac. uk/neuronal-networks

in vivo Basal Bicuculline (7. 5 mg. kg-1 i. v. ) in vivo Basal Kainate (10 mg. kg- I i. v. )

Neuronal Networks Laboratory University of Nottingham Medical School Neuronal Networks Laboratory Project #4: Affective state Cortico-limibic network interactions • anxiety effects of maternal separation pharmacologically-induced (e. g. FG-7142) anxiety • behavioural sequalae rodent ultrasound vocalisations

Neuronal Networks Laboratory University of Nottingham Medical School Neuronal Networks Laboratory Project #5: Nociception & pain management • dual spinal cord / supraspinal recording Role of endocannabinoid system in normal physiology and pain (e. g. neuropathic) states

Mechanically-evoked response in somatosensory thalamus (VPM) innocuous (7 g) stimulation noxious (65 g) stimulation

The Role of the CB 2 Receptor in Nociceptive Processing: Neuronal Networks Laboratory An in vivo electrophysiological study • Cannabinoid receptor agonists are antinociceptive. • CB 1 predominantly expressed in the CNS but also present in the periphery. • CB 2 agonists inhibit: Acute pain [Zimmer et al. ] Inflammatory pain [Clayton et al. ] Neuropathic pain [Ibrahim et al. ] • CB 2 receptors located on: Immune cells Neuronal cells (? ) [Griffin et al; Ross et al; Patel et al. ] • CB 2 agonists lack CNS side effects. Development of potent selective CB 2 ligands: Agonist: JWH-133 Ki 3. 4 n. M with a 200 -fold selectivity over CB 1 receptors. Antagonist: SR 144528 Ki 0. 67 n. M with a 50 -fold selectivity over CB 1 receptors. Aim: To determine the involvement of the CB 2 receptor in nociceptive processing.

Combined unit / LFP with USV / behavioural recording unit activity - nuc accumbens USV call start time Behavioural data Local Field Potential (LFP) USV Recorder input Spectrogram of specific calls [Avi. Soft]

Behavioural Electrophysiology Neuronal Networks Laboratory Circular arena recording using Cine. Plex Movie - rat Hop. Scotch: 8 -channel array in nuc. accumbens - 6 weeks post implant Recorded video Neural data

DISCUSSION • Following KA, hippocampal units (~80%) show an increase in firing; while m. PFC units show either a decrease (~80%) or increase (~20%) in firing rate. m. PFC units lose their characteristic bursting pattern after KA administration. • CCH analysis shows that unit pair activity under basal conditions is more correlated within the m. PFC compared to intra-hippocampal; m. PFC appears to lead hippocampal firing. KA increased correlation within the hippocampus and m. PFC; but the m. PFC-hippocampal drive was lost. • PDC of unit population activity also shows basal m. PFC-hippocampal directionality (predominantly at low frequencies); this initially decreases after KA administration, then later increases at all frequencies. • In basal conditions, PDC analysis of LFPs revealed evidence of information flow from CA 3 to CA 1 and reciprocal hippocampal-m. PFC connectivity with predominant drive from m. PFC to hippocampus. Following KA, there was increased drive from m. PFC to hippocampus. v This alteration in functional connectivity in a seizure model has implications for memory and learning in epilepsy. Caveat(s): Need to consider possible influence of anaesthesia in directing “information flow” and/or (anaesthesia/KA-induced) short-term rewiring of neural circuitry. Other regions (not recorded) may be involved in communication between m. PFC and hippocampus.