1. Inhibitory Cerebello-Olivary Projections and Blocking Effect in Classical Conditioning J J Kim, D J Krupa, R F Thompson Science, vol. 279, 570-573 (1998)

2. Blocking: Observed data CS A US UR (A-US) First CS (A) paired with US –A-US First CS (A) and second CS (B) paired with US –Compound conditioning –AB-US UR (AB-US) CS A CS B US CS B CRCS A Test A alone –Normal CR Test B alone –Very little or no CR

3. Blocking: Interpretation If US is fully predicted by A (A-US), then adding B does not provide new information –Save on unnecessary computation Weaker pre-conditioning of A-US causes a stronger effect of B-US –Inverse proportionality of A-US and B-US CS B CR CS A CS B CRCS A Redundant

4. NMR = nictitating membrane response Eyeblink Conditioning Speaker (CS A) Light (CS B) Air nozzle (US) Thread to eyelid Eyelid movement measurement device Eyelid closes (UR, CR)

5. Postulated eyeblink conditioning circuit CS-US association (Purkinje cells in HVI spike when CR is learned) Blocking inhibition GABA antagonists (eg. Picrotoxin [PTX]) prevent blocking Interpositus nucleus The question: Is this circuit correct?

6. Experiment 1: Procedure Standard classical conditioning –Tone CS –Airpuff US –Eye closing UR becomes CR during training 54 Purkinje cells recorded during conditioning –31 in lobule HVI <--- most likely activity site for eyeblink conditioning –12 in anterior lobe HV, 6 in HVIIA, 5 in paramedian lobule

7. Experiment 1: Eyelid & Cell Responses Naïve animals (5 cells)Trained animals (11 cells) CS-US trials US only trials CR No purkinje cell response Purkinje cell spikes UR

8. Experiment 1: Control, Conclusions, Comments Control case: strictly unpaired tone and airpuff trials –20 out of 45 cells responded to the airpuff with complex spikes –Indicates that tone and airpuff must be paired for spike suppression Conclusion: as eyeblink conditioning occurs the inferior olive’s ability to convey US information to the cerebellum is suppressed –This is not really shown - just the involvement of Purkinje cells Comments –UR amplitude in response to US-only trials is higher for trained animals: why?

9. Trained eyelid response to picrotoxin (PTX) US only CS + US before PTX infusion CS + US after PTX infusion Purkinje cell spikes No purkinje cell spikes Purkinje cell spikes CR UR 3 cells recorded Well-trained rabbits –how many? PTX injected into inferior olive

10. Experiment 2: Procedure Preparation –Rabbits implanted with guide cannulae above contralateral inferior olive CS A US UR (A-US) Phase I: Tone-airpuff conditioning –7 sessions, 1 per day (10 blocks x 10 trials) UR (AB-US) CS A CS B US Phase II: Tone-light-airpuff conditioning –Simultaneously introduce one of two fluids: GABA antagonist: picrotoxin (PTX) Placebo: artificial cerebrospinal fluid (ACSF) –5 sessions, 1 per day (10 blocks x 10 trials) CS B ??? US Light-airpuff test –Light CS + airpuff US testing (B-US) –5 sessions, 1 per day (10 blocks x 10 trials)

11. Experiment 2: Test groups Main group –ACSF: 6 rabbits –PTX: 12 rabbits Control group –5 rabbits Phase IPhase IILight-airpuff Phase IILight-airpuff

12. Experiment 2: Results Phase I Tone CS Light-airpuff test Light CS Phase II Tone + light CS Normal acquisition ACSF/PTX maintain response Control case similar to PTX ACSF shows blocking, then re-learning Blocking Control case acquisition PTX does not affect UR amplitude US-only Partial response

13. Is the circuit correct? Experiment 1 --> something stops Purkinje cell spiking –Purkinje cell spiking correlated with CR PTX infusion in inferior olive restores Purkinje cell spiking –Inferior olive and GABA are involved Experiment 2 --> PTX infusion prevents blocking –PTX seems to prevent GABA inhibition of inferior olive Blocking inhibition Interpositus nucleus

14. Specific Comments Mechanism for inverse relationship between strength of A-US and B-US is not explained Slow acquisition during Experiment 2 light-airpuff test (compared to Phase I) not explained Decrease of blocking over time is not explained –Due to simultaneous extinction of A-US and acquisition of B-US?

15. The End

16. Cerebellar Cortex (Ghez & Tach, 2000) Molecular layer Granule-cell layer Purkinje cell layer Parallel fibres Climbing fibre Purkinje cell Purkinje cell axon Mossy fibre Golgi cell Basket cell Stellate cell Granule cell