POSTER SESSION 1:THURSDAY, OCTOBER 11,2007

BASIC NEUROBIOLOGY (P1 – P10)

P1. Characterization of Basal Ganglia AxonPathfinding Using ZebrafishJL Bonkowsky, CB Chien; Salt Lake City, UT

Objectives: To characterize the development and geneticcontrol of basal ganglia connectivity, using zebrafish (Daniorerio). Zebrafish offers key advantages for this work, includ-ing rapid CNS development; straightforward visualization ofneurons and axons; and the ability to economically performlarge-scale genetic screens.Methods: We used in situ gene markers to label the devel-oping zebrafish CNS. These genes (dat, lmx1a, lmx1b, msxE,nurr1, pitx3) are necessary in mice for early specification ofthe mesodiencephalic dopaminergic (mesDA) neurons (thehomolog of the primate substantia nigra pars compacta-SNpc).

We used anti-sense morpholinos against pitx3 and nurr1to test whether these genes are required to generate mesDAneurons.

Finally, we generated transgenic zebrafish that expressGFP under control of basal ganglia-specific enhancers.Results: In situ analysis revealed a complex and dynamic pat-tern of genes that specify mesDA neurons. Morpholinosagainst nurr1 and pitx3 lead to decreased dat (adopaminergic-neuron specific marker) expression.

We generated several transgenic zebrafish lines which ex-press GFP in subsets of the basal ganglia neurons, includingdlx(mini):GFP and foxP2-enhancerB:GFP.Conclusions: We have identified the mesDA neurons inthe developing zebrafish and shown that they appear to begenetically equivalent to the SNpc. We have created basalganglia transgenic GFP lines, and have several other poten-tial enhancers pending characterization. Our research willallow characterization of basal ganglia pathfinding and itsgenetic control. Further, our zebrafish basal ganglia modeloffers a rapid and powerful means for understanding themechanisms of development, which may play a role in dis-eases involving the basal ganglia.

P2. Cerebral blood Flow Velocity and CognitiveFunction in Children with Mild Sleep-disorderedBreathing before and after AdenotonsillectomyA Hogan, C Hill, D Harrison, F Kirkham;Southampton, UK and London, UK

We have previously demonstrated that middle cerebral ar-tery velocity (MCAV) is raised in children with mild sleepdisordered breathing (SDB), alongside deficits in processingspeed, visual attention and executive function. There arefew data on the effect of treatment. We describe thefollow-up of this cohort after amelioration of SDB by ad-enotonsillectomy. MCAV and neurobehavioral assessmentswere repeated in 19 young children (mean age 6 years) witha history of mild SDB, an average of 12 months after ad-

enotonsillectomy, and 14 healthy ethnic and age-similarcontrols. MCAV significantly decreased in the SDB grouppost-operatively, whereas there was an increase in MCAVwith time in controls. This resulted in a significant inter-action between time and group (F(1,24)�35.7, p�0.001).Importantly, while there was a significant group differenceat Time 1 (pre-operative: T(22.1)�7.9, p�0.001), byTime 2 this difference had reduced (post-operative:T(25)�1.8, p�0.077), suggesting that CBFV in the chil-dren with mild SDB had normalised. There was a signifi-cant association between MCAV change (Time 2-Time1)and pre-operative apnea/hypopnea index (n�12, Rs -.576,one-tailed, p�0.025) i.e. the largest post-surgical reduc-tions in MCAV were found in those children with higherpre-operative apnea/hypopnea indices. Children who hadthe greatest increase in mean overnight oxyhemoglobin sat-uration following surgery also showed the greatest reduc-tion in MCAV (n�9, Rs -.916, one-tailed, p�.001). Therewere also improvements in visual attention and processingspeed in children with treated SDB. These data provideimportant evidence of potentially reversible cerebral hemo-dynamic and cognitive change in otherwise healthy youngchildren with apparently mild SDB.

P3. Synaptogenesis in Human Fetal Basal GangliaHB Sarnat, L Flores-Sarnat; Calgary, Alberta

Objective: to show the normal spatial and temporal se-quence of synaptogenesis in basal ganglia of human fetalbrain using a marker of a synaptic vesicle protein, synapto-physin.Methods: Over 30 months, 156 human fetal and neonatalbrains at 6-41 weeks gestation were studied for synaptophy-sin reactivity. Exclusion criteria: chromosomopathies, majormalformations, chronic metabolic or ischemic encephalopa-thies, massive hydrocephalus and ventriculitis. Maceratedbrains and chronic fetal ischemia were studied separately.Synaptophysin antibody was applied immunocytochemicallyto paraffin sections of brain using a modified technique forfetal brain (Sarnat and Born 1999).Results: The corpus striatum shows a patchy, striated pat-tern with zones of strong alternating with no reactivity, be-ginning at 14 weeks and incomplete until 36 weeks. Initialreactivity is around neurons between fascicles of the internalcapsule. The globus pallidus shows uniformly increasing syn-aptic reactivity from15 weeks to strong at near-term. Thesubstantia nigra shows the earliest synaptophysin, beginningat 9 weeks, strong by 15 weeks. Synaptophysin resists post-mortem autolysis, hence is reliable for stillborns.Conclusions: The patchy reactivity in the fetal corpus stria-tum correspond to the “striosomes of Graybiel”, neuro-transmitter-rich/neuropeptide-poor zones not evident bystandard histology. Athetoid movements and postures at25-32 weeks gestation in preterm neonates, described byPeiper in 1961, appear to correlate temporally with matura-tion of caudate and putamen synaptogenesis. These normaldata provide a basis for interpreting alterations in brains ofinfants dying with dyskinesias, including metabolic enceph-alopathies. Synaptogenesis cannot be demonstrated in theliving patient clinically, by imaging or neurophysiology. EEGreflects large fields of neocortical synaptic activity.

Program and Abstracts, Child Neurology Society S105

P4. Neural Stem Cell Treatment for Neonatal HypoxicIschemic Injury in the 10-day Rat PupS Ashwal, A Obenaus, R Hartman, B Tone, A Wang,HR Tian, N Dilmac, M Digicaylioglu, E Snyder;La Jolla, CA

Neural stem cells (NSCs) have attracted increasing attentionas a potential treatment for neonatal hypoxic ischemic injury(HII) and several studies have demonstrated a reparative ef-fect on brain injury in rats after unilateral carotid liga-tion/8% hypoxia (Rice-Vannucci model, RVM). We haveimplanted iron-labeled NSCs into the contralateral ventricleor brain parenchyma of 10-day RVM and control rat pupsand have performed serial neurobehavioral testing and MRI(11.7 T) studies to monitor NSC tracking and changes inbrain volume for up to 1 year post implantation. NSCs re-mained within the ventricles of control pups whereas inRVM pups, migration of NSCs towards the area of ischemicinjury was seen as early as 4 days and continued for 1 year.We also observed that the apparent initial volume of NSCschanged over time, suggesting that the NSCs migrated to-wards different regions of the ischemic injury, possibly inresponse to different homing signals released by injured tis-sue. Tract-tracing and immunohistochemistry studies ofthese animals are able to define the final differentiated phe-notype and if these NSCs make viable connections with thenon-injured tissues. Neurobehavioral testing showed a corre-lation between severity of ischemic injury and behavioral ab-normalities. Our findings demonstrate that iron-labeledNSCs can be non-invasively tracked for extended periods oftime in a rat pup RVM model and that such methods can beused to develop the optimal method for NSC implantationto maximize recovery after neonatal HII.

P5. Repeated Amphetamine Produces LocomotorSensitization by Reversing Chronic CorticostriatalDepressionNS Bamford and CA Scarlis; Seattle, WA

Objective: Intermittent and repetitive treatment with psy-chostimulants results in a progressive increase in locomotoractivity. How drugs produce long-lasting neuroplasticchanges and why relapse provides compensation remains un-known, although a relationship between dopamine and cor-ticostriatal activity is strongly implicated.Methods: We treated mice with the dopamine releaser am-phetamine (2 mg/kg/day, i.p.) for 5 days and challengedwith amphetamine (2 mg/kg, i.p.) on withdrawal day 3 and21. Locomotor ambulations were recorded using activity-monitoring cages. On withdrawal day 50, corticostriatal re-lease (t1/2) was determined in striatal slices using the endo-cytic tracer FM1-43.Results: Chronic amphetamine produced an increase in lo-comotor ambulations during treatment that persisted inwithdrawal (p�0.001, repeated measures ANOVA). Insaline-treated controls, amphetamine in vitro depressed asubset of cortical terminals with the lowest probability of re-lease (10 �M; t1/2�289 sec vs. 219 sec for controls;p�0.001, Mann-Whitney). Chronic amphetamine however,produced a chronic depression in corticostriatal release (t1/2�290 sec; p�0.001, ANOVA) that was reversed by drugreinstatement in vitro (t1/2�217 sec; p�0.001, ANOVA).The increase in release from glutamatergic terminals on with-drawal day 50 correlated with the increase in locomotor re-sponses due to an amphetamine challenge on treatment day5 and also on withdrawal day 3 and 21 (p�0.001, repeatedmeasures ANOVA).

Conclusions: This suggests that locomotor sensitization maybe determined by a potentiation in glutamate release, whichwould excite the striatum and enhance locomotor activity.These observations in psychostimulant-induced synapticplasticity provide further insight into the mechanisms under-lying drug dependence and pharmacological treatment of at-tention deficit disorder.

P6. Glutamate Release in the Ventral Striatum isMediated by D1 and D2 Receptor ActivityD Dever and NS Bamford; Seattle, WA

Objective: Abnormal levels of dopamine in the striatumhave been implicated in motor disorders, drug addiction, andschizophrenia. We observed the effect of dopamine on glu-tamate release from frontal cortical projections to the nucleusaccumbens core. The effect of cortical frequency and dopa-mine receptor 1 (D1) and 2 (D2) manipulation was exam-ined.Methods: 300�m-thick saggital sections containing the cor-tex and striatum were prepared from adult C57Bl6 mice.Cortical terminals within the striatum were loaded with theendocytic tracer FM1-43 via bipolar stimulation over theprefrontal cortex. Florescent puncta in the ventral striatumwere visualized using a multiphoton confocal microscope.Re-stimulation of the cortex produced exocytosis that wasquantified by the release halftime (t1/2), defined as the timerequired for each individual puncta to reach 50% lumines-cence.Results: Corticoaccumbal release was significantly slower inthe presence of the dopamine-releaser amphetamine(t1/2�252 sec vs. 175 sec for controls; n�57-184 puncta;p�0.001,Mann-Whitney). This effect was seen at 20Hz cor-tical stimulation but not at 1Hz or 10Hz (p�0.5). Both D1(t1/2�256 sec; SKF38393, 10�M) and D2 (t1/2�267 sec;quinpirole, 10�M) agonists inhibited release (n�201-208puncta;p�0.001) by depressing activity from terminals withthe lowest probability of release.Conclusions: Excitation of the ventral striatum by glutamateis dependent on the frequency of cortical stimulation and ismodulated by both D1 and D2 receptor activation. Thus,dopamine acts a low-pass filter, with filtering applied to less-active cortical inputs. Disruption of this neural integratorwould likely disturb critical cortical- basal ganglionic net-works resulting in disorders of thought and movement.

P7. Expression Patterns of the Organic Cation/Carnitine Transporter Family in Adult Murine Brain:Implications for Cerebral DysfunctionAM Lamhonwah, CE Hawkins, C. Tam, J Wong, L Mai,I Tein; Toronto, Ontario, Canada

Objective: To characterize expression patterns of mOctn1,-2 and -3 in murine brain.Methods: We applied our transporter-specific antibodies tomOctn1,-2 and -3, followed by 20 antibody and DAB per-oxidase detection to adult murine brain sections counter-stained with hematoxylin.Results: All 3 transporters showed strong expression in theexternal plexiform layer of olfactory bulb and in olfactorynerve, the molecular layer and neuronal processes of inputfibres extending vertically in motor cortex, in the dendriticarborization of cornu ammonis and dendate gyrus, neuronalprocesses in arcuate nucleus, choroid plexus cells, and neu-ronal cell bodies and dendrites of cranial nerve nuclei V andVII. In the cerebellum, all three were strongly expressed in

S106 Annals of Neurology Vol 62 (suppl 11) 2007

dendritic processes of Purkinje cells, but Octn1 and -2 wereexpressed more strongly than Octn 3 in Purkinje cell bodies.In spinal cord, Octn1,-2 and -3 were prominent in axonsand dendritic end-arborizations of spinal cord neurons in as-cending and descending white matter tracts, whereas Octn3was also strongly expressed in anterior horn cell bodies.Conclusions: hOCTN2 deficiency presents with carnitine-responsive cardiomyopathy, myopathy and hypoglycemic,hypoketotic coma with strokes, seizures and delays. Inmouse, Octn1,-2 and -3 are expressed in a CNS pattern sug-gestive of roles in modulating cerebral bioenergetics and inacetylcholine generation for neurotransmission in olfactory,satiety, limbic, memory, motor and sensory functions. Thisdistribution may play a role in the pattern of neurologicalinjury that occurs in hOCTN2 deficiency during catabolicepisodes of encephalopathy which may manifest with cogni-tive impairment, hypotonia and seizures.

P8. Minocycline Pretreatment Confers Relatively LongLasting Neuroprotection against Acute BilirubinEncephalopathy in Jaundiced Gunn Rat Pups withMinocycline PretreatmentJN Wu, AC Rice, SM Shapiro; Richmond, VA

Objective: Minocycline is a neuroprotective agent with anti-inflammatory, antibacterial and anti-apoptotic properties. Todetermine whether the previously demonstrated neuroprotec-tive effect of minocycline pretreatment against sulfadime-thoxine (sulfa) induced acute bilirubin encephalopathy(ABE) extended beyond 6 hours, we treated jaundiced (jj)Gunn rat pups with minocycline or saline prior to adminis-tering sulfa and recorded brainstem auditory evoked poten-tials (BAEPs) 48 hours later.Methods: Littermate 15-16 day old jj pups were pre-treatedwith either minocycline 50 mg/kg or an equal volume ofsaline 15 minutes before either sulfadimethoxine 200 mg/kg,to induce ABE by displacing bilirubin from blood into tis-sue, or an equal volume of saline as follows: 1) minocycline/sulfa (experimental, N�7); 2) saline/sulfa (positive control,N�7); and 3) saline/saline, (negative control, N�4). BAEPI latency, I-II and I-III interwave intervals, and amplitudes ofwaves I, II, and III were scored.Results: Four of seven in the saline/sulfa group died prior to48 hours vs. none in the minocycline/sulfa or saline/salinegroups. Statistically significant differences (p�0.05 by2-tailed t-tests) were found to exist between the latencies andamplitudes of BAEP waves II and III (equivalent to humanIII and V), and I-II and I-III interwave intervals in the mi-nocycline/sulfa versus saline/sulfa group; minocycline/sulfainjected animals were no different than saline/saline controls.Conclusion: Minocycline protection against sulfa-inducedBAEP abnormalities in jj Gunn rats is relatively long lasting.Future experiments are planned to determine the efficacy ofpost-treatment with minocycline.

P9. Preventing an Identity Crisis: Unexpected Co-Expression of Class III Beta-tubulin and Glial FibrillaryAcidic Protein in Human Fetal Astrocytes in CultureCD Katsetos, E Draberova, L del Valle, L Bertrand,DP Agamanolis, JP de Chadarevian, A Legido, P Draber;Philadelphia, PA, Prague, Czech Republic, Akron, OH

The class III Beta-tubulin isotype (Beta-III-tubulin) is widelyregarded as a neuronal marker in developmental neurobiol-ogy. The localization of this cytoskeletal protein after differ-entiation induction of embryonic stem cells has been used to

confirm the identity of neurons. Similarly, the detection ofBeta-III-tubulin staining in immature cells of the rostral mi-gratory stream has provided the basis for postnatal neurogen-esis. We have hypothesized that during development Beta-III-tubulin is not neuronal-specific but is also transientlyexpressed in immature glia, in addition to being present inpost-mitotic neurons. To test this hypothesis we determinedthe expression of Beta-III-tubulin in glial fibrillary acidicprotein (GFAP)-expressing, early passaged, fetal astrocytesisolated from the cerebral hemisphere of a male human fetus20 weeks of gestation (Clonetics®, Cambrex). Cells weremaintained as monolayer cultures for 2, 6 and 9 days understandard conditions and without exposure to differentiation-inducing agents. After extraction with Triton-X100 and fix-ation in formaldehyde immunofluorescence microscopy wascarried out using a panel of polyclonal and monoclonal anti-peptide antibodies recognizing distinct epitopes of Beta-III-tubulin and GFAP. Co-expression of Beta-III-tubulin andGFAP was detected in 100% of cells from all time pointsbut in spatially distinct patterns. Also, by confocal micros-copy extensive co-localization of GFAP and Beta-III-tubulinin cells of the telencephalic germinal matrix and the neocor-tex of human fetuses 16-20 weeks of gestation was demon-strated. Taken together, our findings have potential implica-tions in stem cell research as they raise doubt about theassumed neuronal identity of Beta-III-tubulin-expressing cellsin vitro and in situ.

P10. Neuroprotective Effect of Human PlacentalExtract on Hypoxic-Ischemic Brain Injury in NeonatalRatsWS Choi, GY Chae, HC Choi, SW Park, HS Lee, DH Shin,JY Lee, SH Eun, BL Eun; Seoul, South Korea

Objective: The variety of biological actions of human pla-cental extract (HPE) is a matter of increasing interest. Recentresearch studies reveal that HPE is the rich resources of var-ious bio-active substances like polydeoxyribonucleotides(PDRN), RNA, DNA, peptides, amino acids, enzymes, traceelements, etc. The purpose of this study was to evaluate theneuroprotective effect of HPE on hypoxic-ischemic (H-I)brain damage in the newborn rat.Methods: Rt. common carotid arteries of 7-day-old rats werecoagulated and then the rats were exposed to 8% oxygen.Rats received HPE (5, 10, 25, 50 mL/kg/dose) by intraperi-toneal injection immediately before and again after ischemia,or saline for control. On postnatal day 12, brains of the ratwere removed and gross morphological damage was evalu-ated. The brains were placed in a deep freezer, sectionedcoronally into 20 �m thick slices and then stained. For quan-tification of the severity of brain injury, the bilateral cross sec-tional areas of anterior commissural and posterior hippocam-pal level were measured with NIH image analysis system.Results: HPE pretreatment resulted in a decreased incidenceof liquifactive cerebral infraction. Quantitation of hemisphericareas in rats receiving HPE and control littermates confirmedthe results of initial inspection.Conclusions: Pretreatment with HPE decreases the incidenceand severity of ischemic brain damage in this animal model ofperinatal cerebral hypoxia-ischemia. These data indicate thatHPE plays a role in the evolution of HI injury to the devel-oping brain, and that HPE pretreatment may offer an effectivemeans to decrease the incidence and severity of perinatal HIbrain injury.

DOI: 10.1002/ana.11631

Program and Abstracts, Child Neurology Society S107