www.elsevier.com/locate/ynimgNeuroImage 41 (2008) T208–T210

Keynote Speaker I

The Chemistry of Mental Activity

Henry N. Wagner, Jr.

Johns Hopkins University School Of Medicine

In 1983, positron emission tomography (PET) was usedfor the first time to image dopamine receptor binding in thebasal ganglia of a living human being. The enthusiasm ofscientists and the public that greeted this news equaled thatfollowing the invention of electroencephalography (EEG) byHans Berger, an Austrian psychiatrist, in 1924.

The use of EEG to study the mysterious mind was metwith skepticism. In 1940, Charles Sherrington wrote: bThemind is something with such manifold variety, such fleetingchanges, such countless nuances, such wealth of combina-tions, such heights and depths of mood, such sweeps ofpassion, such vistas of imagination, that the submission ofsome electrical potentials recognizable in nerve-centers ascorrelative to all these may seem to the special student of themind almost derisory.Q What would Sherrington say if hecould see how molecular imaging was being used today tostudy the relationship between mind and brain?

Molecular imaging uses the tracer principle to createquantifiable images of the distribution of a radioactive tracerwithin regions of the brain. Magnetic resonance imaging(MRI), computed axial tomography (CAT), and ultrasoundprimarily reveal structure and function, but are being ex-tended to apply the tracer principle as well.

Molecular imaging can be used to relate mental illness tospecific abnormalities of molecular processes in differentparts of the brain. Molecular phenotypes are being used moreand more in diagnosis, prognosis, and treatment planning, aswell as in the assessment of the results of treatment, providinglong-term follow up. If the treatment is going to be chemical,the diagnosis should be chemical, that is, the finding ofdeficient or excessive molecular processes in different partsof the brain.

Has the time come to create a molecular theory of mentaldisease? Instead of putting patients into diagnostic pigeonholes, such as depression, Parkinson's disease, or schizo-phrenia, that are poorly defined, heterogeneous, and non-specific, we can use molecular imaging to relate a patient'sproblems to regional brain chemistry.

doi:10.1016/j.neuroimage.2008.04.021

Hundreds of neurotransmitters have been identified thatinclude catecholamines (norepinephrine), biogenic amines(serotonin, acetylcholine), polypeptides/proteins (endor-phins), and amino acids/amino acid analogues (glutamate,gamma amino butyric acid).

Many pharmaceuticals act by inhibiting or enhancing theeffects of neurotransmitters, or the action of hormones.Measuring the effects of altering molecular processes in thebrain, in patients with mental disorders, need not provide abiological explanation of what causes these diseases, butrather provides modifiable molecular manifestations.

Keynote Speaker II

Mapping the Structure and Function of BrainAn Integrated Computational Approach

Arthur W. Toga

Laboratory of Neuro Imaging, Department of Neurology,UCLA School of Medicine

The ability to statistically and visually compare andcontrast brain image data from multiple subjects is essentialto understanding normal variability and differentiatingnormal from diseased populations. This talk describessome of these approaches and their application in basicand clinical neuroscience. There are numerous probabilisticatlases that describe specific subpopulations, measure theirvariability and characterize the structural differencesbetween them. Utilizing data from structural MRI, we havebuilt atlases with defined coordinate systems creating aframework for mapping data from functional, histologicaland other studies of the same population in several species.This talk describes the basic approach and some of theconstructs that enable the calculation of probabilistic atlasesand examples of their results from several different normaland diseased populations. Computational strategies thatintegrate across modalities and subjects also establish timeand data trends across a variety of observations. This talkwill also illustrate some approaches useful in understandingmultidimensional data and the relationships between themover time. Finally, there will be challenges identified forfuture mapping and modeling between modalities, time,subjects and species.

T209

Invited Thematic Speaker I

Pittsburgh Compound-B Six Years Later:What have we learned, what lies ahead?

William E. Klunk

Departments of Psychiatry and Neurology,University of Pittsburgh

Introduction: The first human amyloid-beta imaging studyusing Pittsburgh Compound-B (PiB) was performed inUppsala, Sweden, in February 2002. The preliminary reportin July 2002 suggested that amyloid-beta could besuccessfully imaged and quantified using PiB-PET.

Objective: Summarize key findings of amyloid-imagingstudies using PiB over the past six years.

Results: Approximately 40 institutions worldwide areperforming PiB studies. We estimate that well over 2000PiB scans have been performed, without adverse effects. Over300 PiB studies have been performed in Pittsburgh with MR-based co-registration and early studies included arterialsampling and sequential [O-15]water and FDG studies. PiBdata is amenable to pharmacokinetic analysis by standardmethods, in particular the Logan graphical method. Arterialinput data are advantageous, but not absolutely required. Inputcan be derived from cerebellar or carotid reference regions.Late-summed standardized uptake value (SUV) methods alsocan be employed. Voxel-based (SPM) methods producerobust results as well. PiB retention correlates inversely tometabolism (FDG) in precuneus and frontal cortex: howeverFDG has a lower dynamic range. [O-15]Water studies showPiB retention is not strongly influenced by blood flow. Inclinical studies, most normal controls show no evidence ofcortical PiB retention, but ~25k of those over age 65 showevidence of PiB retention. Neurocognitive testing has notdemonstrated a clear relationship between cognitiveperformance and amyloid deposition in this small series, butwork is ongoing. MCI subjects do not cluster between controland AD ranges. About ~25k of MCI subjects showed no PiBretention (control-like) and ~60k had AD-like PiB retention;the remainder are intermediate. Asymptomatic FAD subjectsharboring PS-1 mutations show focal amyloid deposition inthe striatum, suggesting this is a site of initiation for amyloiddeposition in these Ab-overproduction forms of AD. Incontrast, initiation of amyloid deposition in late-onset ADmay begin in the frontal or precuneus cortex. PiB retention hasnow been shown to correlate with the post-mortem Ab load intwo subjects.

Conclusions: Early studies continue to suggest that amyloidimaging with PiB-PET is a safe and effective way to studyAb amyloid deposits in the brain.

Invited Thematic Speaker II

Tonic–Phasic Regulation of Dopamine NeuronActivity: Relationship to the Pathophysiology

of Schizophrenia

Anthony A. Grace

Departments of Neuroscience, Psychiatry, and Psychology,University of Pittsburgh

Introduction: The dopamine (DA) system exhibits functionalcompartmentalization into tonic and phasic response systems.Tonic activity is defined as the spontaneous discharge ofpopulations of DA neurons in the midbrain, whereas the phasicresponse is related to the behaviorally-driven burst discharge ofthese neurons. It is the phasic response system that is proposedto be reflected by raclopride displacement in imaging studies.Tonic and phasic DA neuron firing are furthermore regulatedby independent mechanisms, with tonic activity being drivenby the ventral hippocampus and phasic bursts driven bybrainstem glutamatergic afferents. Only DA neurons that aretonically activewill be driven into phasic burst firing; therefore,by controlling the number of DA neurons active, the ventralhippocampus determines the bgainQ of the phasic response,with more hippocampal activity causing more DA neurons tobe tonically active and capable of being driven into burst firing.

Methods: Single unit extracellular recordings were performedfrom neurons in the ventral striatum, the ventral hippocampus,and from mesolimbic DA neurons. DA neuron activity wasclassified in terms of the number of neurons firing, their firingrate, and firing pattern. A rat model of schizophrenia wasgenerated by administering themitotoxinmethylazoxymethanolacetate (MAM) to pregnant dams at gestational day 17 andexamining the offspring as adults.

Results: MAM-treated rats exhibited anatomical, behavioral,and pharmacological responses consistent with an animalmodel of schizophrenia. These include limbic corticalthinning without decreases in neuron number, deficits inprepulse inhibition and executive function, and abnormallyheightened responses to PCP and amphetamine. The DAsystem in schizophrenia subjects has also been shown inimaging studies to be hyper-responsive to amphetamine. InMAM treated rats, there was a substantial increase in thenumber of DA neurons that were tonically active, and this wasapparently driven by hyperactivity within the ventralhippocampus. When the ventral hippocampus wasinactivated, DA neuron firing returned to baseline levels,and the behavioral hyper-responsivity to amphetamine wasrestored to that observed in non-treated rats.

Conclusion: Evidence from rat studies demonstrates that theDA system is functionally compartmentalized into a baseline

T210

tonic DA neuron activity and behaviorally salient burstfiring. By controlling the number of DA neurons that can bephasically activated by a behavioral stimulus, the ventralhippocampus controls the amplitude of the phasic response.In conditions such as schizophrenia, hyperactivity within thehippocampal complex leads to increases in the proportion ofDA neurons firing spontaneously, and this mediates theamphetamine hyper-responsivity in the system. A similar

system also appears to account for the amphetamine hyper-responsivity observed following amphetamine sensitization.Pharmacological intervention that targets normalization ofhippocampal activity may thus be an effective treatmentstrategy for disorders characterized by a hyper-responsiveDA system.

Acknowledgments: USPHS MH57440 and DA15408.