The 392nd session: Brain and Cognitive Sciences

ABOUT THE XIANGSHAN-SCIENCE CONFERENCES:

The Xiangshan-Science Conferences (XSSC) was initiated by the former State Science and Technology Commission, now the Ministry of Science and Technology of (MOST). It was officially inaugurated in 1993 under the joint sponsorship of MOST and the Chinese Academy of Sciences (CAS). It also draws support from the National Natural Science Foundation of China, the Academic Divisions of CAS, the Chinese Academy of Engineering, the Ministry of Education of China, the State Commission of Science, Technology & Industry for National Defense, and the General Armament Department of the People's Liberation Army.

As a standing meeting series, which is usually held at the picturesque Xiangshan (Fragrant Hills), a famous scenic spot in the northwestern suburbs of , XSSC is mainly dedicated to symposia. We have brought it to the University of because this year is the University's 100th anniversary.

XSSC promotes multi- or inter-disciplinary research, overall comprehensive studies, innovative thinking and knowledge innovation by creating a relaxed environment for academic exchanges, upholding the spirit of free academic discussion, and giving priority to scientific frontiers and their future development. For more general information, please visit http://159.226.97.16/.

- 1 - The 392nd session: Brain and Cognitive Sciences

ABOUT XIANGSHAN-SCIENCE CONFERENCES: 392nd SESSION on Brain and Cognitive Sciences: New Frontiers and Strategic Development

The advent of neuroimaging (fMRI and PET) and single-cell recording techniques has enabled us, for the first time ever, to probe in vivo human brain under noninvasive or micro-invasive conditions to directly observe and study the cognitive functions of the human brain in its normal state, which breaks through the limitation of the traditional pathologic methods and undoubtedly represents a major milestone in the scientific endeavor of understanding brain functions. “Cognitive Neuroscience – with its concern about perception, action, memory, language and selective attention – will increasingly come to represent the central focus of all Neurosciences in the 21st century,” stresses Eric Kandel, the 2000 Nobel Prize Laureate in Physiology or Medicine. The past decade has witnessed tremendous advances in the understanding of the human brain and its functions. Emerging as a comprehensive and interdisciplinary issue, the characterization of brain structure, function and plasticity has been a major scientific problem and has attracted dramatic attention from the entire academic community. Investigation into cognitive functions will enormously increase our understanding of brain activity as well as promote the enlightened usage, development and protection of the human brain. Such research plays a pivotal role in a society with increasing emphasis on talented personnel and scientific and technological innovation. MOST, CAS and other central government institutions have given strong support to this genre of research; and scholars in Hong Kong and the mainland stay constantly informed of the rapid development of brain and cognitive sciences.

To keep fully acquainted with the latest research in brain and cognitive sciences, to broaden international academic cooperation and exchanges, to promote research in this subject in Hong Kong and the Mainland, to make the research direction in this subject clear yet inclusive, we host this Xiangshan Science Conference in the University of Hong Kong (HKU) on April 8-9, 2011, with the theme “Brain and Cognitive Sciences: New Frontiers and Strategic Development”. This conference will also be one of the activities commemorating the 100th anniversary of HKU. Experts from different disciplines have been invited to this conference to conduct in-depth discussions on the following central topics: (1) Perception and Attention; (2) Memory, Decision and Language; (3) Brain Dysfunction, Cognitive Deficits, and Neuroprotection.

Conference Theme and Executive Co-Chairs The XSSC exercises a chair responsibility system. After defining the theme for a session/ symposium, it will engage its executive chair(s).

- 2 - The 392nd session: Brain and Cognitive Sciences

Conference Theme: Brain and Cognitive Sciences: New Frontiers and Strategic Development Executive Co-Chairs: Kwok-Fai So (Chair of Anatomy and Co-Director, State Key Laboratory of Brain and Cognitive Sciences, HKU) Lin (Professor and Director, State Key Laboratory of Brain and Cognitive Sciences, Chinese Academy of Sciences) Robert Desimone (Professor and Director, McGovern Institute for Brain Research, MIT) Li-Hai Tan (Professor & Co-Director, State Key Laboratory of Brain & Cognitive Sciences, HKU)

Conference Venue: Council Chamber, 8/F Meng Wah Complex, The University of Hong Kong

Conference Secretariat State Key laboratory of Brain and Cognitive Sciences The University of Hong Kong Flat 3A, 2 University Drive Telephone: (852) 2241 5877 Fax: (852) 2549 6253 Email:[email protected] http://www.hku.hk/fmri/XSC/index.htm

- 3 - The 392nd session: Brain and Cognitive Sciences

SCIENTIFIC PROGRAMME

April 8, Friday 08:10-08:40 Registration

09:00-09:30 Opening Ceremony

Welcome Remarks by Xiangshan-Science Conferences Prof. Bingxin Yang Prof. Xian’en Zhang Welcome Remarks by University of Hong Kong Pro-Vice-Chancellor S.P. Chow Remarks by Executive Co-Chairs

Self-Introduction by Scientists Morning Session Executive C-Chairs: Kwok-Fai So, Lin Chen, Robert Desimone, Li-Hai Tan

Keynote Review Reports:

09:40-10:10 Research on Brain and Cognitive Sciences: New Frontiers and Challenges Robert Desimone (MIT) 10:10-10:50 Modeling Human Neural Systems: Insights from Brain Imaging Peter Fox (University of Texas Health Science Center) 10:50-11:10 Tea Break 11:10-12:30 Discussion

12:30- Lunch

Afternoon Session Executive C-Chairs: Li-Hai Tan, Kwok-Fai So, Lin Chen, Robert Desimone

Central Topic 1: Perception and Attention

14:00-14:30 Mechanisms of Visual Attention Sabine Kastner (Princeton University) 14:30-15:00 How Does the Brain Recognize Objects and Faces? Doris Tsao (California Institute of Technology) 15:00-16:00 Discussion

16:00-16:30 Tea Break

19:30- Dinner 9 April (Saturday) Morning Session Executive C-Chairs: Robert Desimone, Kwok-Fai So, Lin Chen, Li-Hai Tan

- 4 - The 392nd session: Brain and Cognitive Sciences

Central Topic 2: Memory, Decision and Language

09:00-09:30 Neural Decision Computation and Memory in Cortical Circuits Xiao-Jing Wang (Yale University) 09:30-10:00 Dopamine Reveals Neural Circuit Mechanisms of Decision-Making: From Fly to Human Aike Guo (Institute of Neuroscience, CAS, Shanghai) 10:00-10:30 Language, Culture and Brain William S-Y Wang (Chinese University of Hong Kong) 10:30-11:00 Brain Systems Responsible for Reading Li-Hai Tan (University of Hong Kong) 11:00-11:20 Tea Break

11:20-13:00 Discussion

13:00- Lunch Afternoon Session Executive C-Chairs: Lin Chen, Li-Hai Tan, Kwok-Fai So, Robert Desimone

Central Topic 3: Brain Dysfunction, Cognitive Deficits, and Neuroprotection

14:00-14:30 Promoting Plasticity and Functional Recovery in the Brain Sarah Dunlop (The University of Western Australia) 14:30-15:00 Brain Injuries: Cognitive Function Impairment and Its Management Prof. Wai Sang Poon (Chinese University of Hong Kong) 15:00-15:30 Neuroprotection and Neural Repairing Kwok-Fai So (University of Hong Kong)

15:30-15:50 Tea Break

15:50-17:00 Discussion

17:00-17:30 Summary of the Session & closing remarks

18:00- Dinner

- 5 - The 392nd session: Brain and Cognitive Sciences

INVITED KEYNOTE SPEECHES

Research on Brain and Cognitive Sciences: New Frontiers and Challenges Robert Desimone, Ph.D MIT, Massachusetts Institute of Technology

Understanding the human brain is a truly astronomical problem – the number of neurons in the brain is about the same as the number of stars in our galaxy. A better understanding of how these billions of neurons interact will not only help us understand the complexities of the human mind, but will also address the urgent health problems facing people around the world. These range from disorders of childhood, such as autism and dyslexia, to those of middle age, including depression and schizophrenia, to disorders of aging, including Alzheimer’s disease. Statistically, one in four families will be affected by a brain disorder of some type. These are global problems, and solving them will require a global effort, including the collaboration of scientists from many countries. Fortunately, we are aided by three recent scientific revolutions that have affected neuroscience. These are the revolutions in the fields of genetics, molecular biology, and systems neuroscience. In just the past 3-4 years, genetic studies have identified numerous genes that make us vulnerable to all types of brain disorders. Molecular studies are beginning to show how these genes affect neurons in specific brain circuits. And systems neuroscience studies are beginning to show us how specific neural circuits lead to both complex behavior in the normal brain and disrupted behavior in brain disease, in both animal models and human subjects.

One specific example of the kind of work that bridges a genetic mutation to a complex behavior comes from Guoping Feng at MIT, who has identified a gene for a synaptic protein in striatal synapses, which when mutated in mice causes obsessive grooming, which resembles obsessive compulsive disorder. This genetic mutation throws striatal circuits out of balance. These circuits could, in an oversimplified way, be described as “stop” and a “go” circuit. Restoring the normal gene puts the circuits back into balance, and restores normal behavior. Establishing circuit-based mechanisms for brain disorders such as autism, schizophrenia, and depression will be a key challenge for the upcoming decade.

The speakers at this meeting will provide numerous examples of cutting-edge research at the forefront of neuroscience, spanning molecular, systems, and quantitative modeling approaches. I will take some examples from recent work of my own and other labs working in systems neuroscience, including technological breakthroughs that are giving us unprecedented ability to both observe neural circuits in operation and perturb their function. In my lab, we are studying how the prefrontal cortex and the visual cortex interact with each other in order to produce a focus of visual attention. We have found that when animals focus their attention on an object, neurons in the frontal eye fields in prefrontal cortex synchronize their activity with cells in visual cortex, in the gamma frequency range. However, the phase of interaction between the two structures is shifted by the equivalent of 10 ms, which seems to be just the amount of time needed for action potentials to travel from one brain region to the other and cross synapses. Thus, as neurons fire in one structure, their action potentials will reach neurons in the coupled structure when these cells are maximally

- 6 - The 392nd session: Brain and Cognitive Sciences depolarized and thus prepared to accept new input. These timing effects may underlie not only attention, but also learning and synaptic plasticity, which also depend on synchronization of pre- and post-synaptic timing. Time-shifted synchrony may be a general principle of functional interactions among brain systems. Using fMRI, MEG, and EEG in humans, neuroscientists are also beginning to study how numerous brain structures interact in during rest and during many complex abilities, including memory and language, and computational modelers are making progress in developing models that explain the operations of these systems in the normal and diseased state.

But how to go beyond studies of correlation and actually study causality between different neural signals? In humans, transmagnetic stimulation remains a useful tool for testing critical function. However, one of the greatest technological innovations of the past decade in neuroscience has been the development of optogenetic tools for manipulating neural circuits. Karl Deisseroth, at Stanford, and Ed Boyden, now at MIT, discovered that light sensitive channels and pumps, when placed into neurons, allow neural activity to be turned on and off with millisecond precision. Thus, neuronal activity of specific cell types can now be controlled by light! Not only are these very powerful research tools, but working with Ed, we have found that these molecules can also be used safely in monkeys, which is a step towards human applications. There are potential clinical applications in several disorders, including Parkinson’s disease, but perhaps the most likely application in the near term the restoration of vision in forms of blindness such as macular degeneration and retinitis pigmentosa.

These are just a very few examples of the opportunities and challenges ahead. In sum, some of the key challenges for neuroscience in the coming decade will be to (1) complete the genetic analysis of human brain diseases, (2) develop and expand the application of powerful new technologies for measuring and controlling neuronal activity, (3) understand the operation of the neural circuits important for complex behaviors such as memory, attention, perception, and language in both healthy and ill individuals, and (4) devise all new treatments for brain disorders that are guided by our new knowledge of genetics and neural circuits.

- 7 - The 392nd session: Brain and Cognitive Sciences

Modeling Human Neural Systems: Insights from Brain Imaging Peter T. Fox, M.D. Research Imaging Institute, University of Texas Health Science Center at San Antonio

Inter-regional connectivity is a fundamental aspect of brain organization for which powerful imaging probes have been developed, the most widely utilized being resting-state network (RSN) analysis of BOLD fMRI (a functional connectivity measure) and diffusion tensor imaging (DTI) tractography (a structural connectivity measure). A problem facing both types of connectivity metric is that they are capable (at least in principle) of computing connection strength between any two points in the brain, generating extraordinarily large data sets and potentially sacrificing statistical power by requiring correction for multiple comparisons. A second problem facing these forms of connectivity analysis is that they do not carry information about function, being either purely structural (DTI) or acquired at rest (RSN). One strategy for overcoming these two limitations is to model the neural circuits of interest meta-analytically. To this end, we have developed meta-analytic approaches to connectivity mapping which have the distinct advantages of: 1) using very large, pre-existing, published datasets; 2) using behavioral meta-data to characterize the behaviors and mental operations supported by individual networks; and, 3) generating output very similar in format and results to per-subject connectivity mapping methods. The data input for meta-analytic connectivity modeling (MACM) are center-of-mass stereotactic location coordinates from published brain-activation or voxel-based morphometry studies and the study-associated meta-data, as stored in the BrainMap database (www.brainmap.org).1 For MACM analyses, all data sets within BrainMap can be included, unlike “traditional” activation likelihood estimation (ALE) analyses, in which only datasets from similar tasks are grouped.2 Similar to connectivity analyses of resting-state fMRI, MACM analyses can be region-seeded3 (e.g., assessing amygdala connectivity) or be performed via independent components analysis4 (ICA). In either application, behavioral meta-data can be used to characterize and filter connectivity results. For example, the functional role of specific pathways within the default mode network (DMN) can be discriminated meta-analytically, a type of analysis not possible using structural connectivity data or resting-state functional connectivity data.5 A variety of MACM validations have been performed, including comparisons to resting-state fMRI analyses, to DTI tractography, and to primate tract tracing literature.3,4,6 MACM can readily model connectivity patterns and behavioral functions throughout the human brain, providing a framework within which per-subject data can be analyzed in a more directed and statistically powerful manner. For example, we have used MACM to select the optimal imaging endophenotypes to assess genetic influences on working memory; prior modeling “limited the search space” and increased statistical power.7 MACM can also be used to construct fully data driven starting models for causal modeling (e.g., structural equation modeling [SEM] and dynamic causal modeling [DCM]) and graph analytic modeling8, a general requirement for using these methods. Collectively, MACM connectivity modeling approaches compliment per-subject approaches both by “limiting the search space” and by providing task-based behavioral information. We suggest that virtually all connectivity studies performed on per-subject data can be enhanced by doing prior modeling using MACM.

1. Fox PT, Lancaster JL. Mapping context and content: The BrainMap model. Nature Rev

- 8 - The 392nd session: Brain and Cognitive Sciences

Neurosci 3, 319-321, 2002. 2. Turkeltaub PE, Eden GF, Jones KM, Zeffiro TA. Meta-analysis of the functional neuroanatomy of single-word reading: Method and validation. NeuroImage 16, 765-780, 2002. 3. Robinson JL, Laird AR, Glahn DC, Lovallo WR, Fox PT. Meta-analytic connectivity modeling: Delineating the functional connectivity of the human amygdala. Hum Brain Mapp 31, 173-184, 2010. 4. Smith SM, Fox PT, Miller KL, Glahn DC, Fox PM, Mackay CE, Filippini N, Watkins KE, Toro R, Laird AR, Beckmann CF. Correspondence of the brain's functional architecture during activation and rest. Proc Natl Acad Sci USA 106, 13040-13045, 2009. 5. Laird AR, Eickhoff SB, Li K, Robin DA, Glahn DC, Fox PT. Investigating the functional heterogeneity of the default mode network using coordinate-based meta-analytic modeling. J Neurosci 29, 14496-14505, 2009. 6. Eickhoff S, Jbabdi S, Caspers S, Laird AR, Fox PT, Zilles K, Behrens T. Anatomical and functional connectivity of cytoarchitectonic areas within the human parietal operculum. J Neurosci 30, 6409-6421, 2010. 7. Karlsgodt KH, Kochunov P, Winkler AM, Laird AR, Almasy L, Duggirala R, Olvera RL, Fox PT, Blangero J, Glahn DC. A multimodal assessment of the genetic control over working memory. J Neurosci 30, 8197-8202, 2010. 8. Neumann J, Fox PT, Turner R, Lohmann G. Learning partially directed functional networks from meta-analysis imaging data. Neuroimage 49, 1372-1384, 2010.

- 9 - The 392nd session: Brain and Cognitive Sciences

INVITED SPEECHES

Neural Basis of Visual Attention in the Human Brain Sabine Kastner, Ph.D Princeton University

Natural scenes are cluttered and contain a multitude of objects that cannot be processed at the same time due to limited processing capacity of the visual system. Mechanisms of attention help to overcome limited processing capacity and to effectively select the information that is most relevant to ongoing behavior from the environment. In the first part of my talk, I will introduce the problem of limited processing capacity and the large-scale network of brain areas that mediate the selection of behaviorally relevant information, as measured with functional magnetic resonance imaging (fMRI) in the human brain. Neural activity in the visual system is modulated by selective attention in line with a neural gain control mechanism that facilitates visual processing. This mechanism appears to be controlled by a distributed network of areas in frontal and parietal cortex. Damage, or dysfunction of this network results in attention deficits. In the second part of my talk, I will discuss recent studies that have begun to provide a neural basis for attentional selection from natural scenes. In natural scenes, objects appear in many different contexts and under variable viewing conditions, which presents a challenge for the visual system in representing information. I will show that internal search templates bias neural processing in the visual system in favor of the stimuli that motivate ongoing behavior (e.g. cars when crossing a street), thereby effectively filtering out irrelevant information and rendering us “blind” towards distracter information. Studying attention in more naturalistic conditions that embrace the complexity of natural vision is an important step to fully understand the neural machinery that mediates the selection of information from the environment.

- 10 - The 392nd session: Brain and Cognitive Sciences

How Does the Brain Recognize Objects and Faces? Doris Tsao, Ph.D California Institute of Technology

Recognition is half of vision (“Vision is knowing what is where by looking.” – David Marr). Tackling the problem of how the brain attaches a particular visual form to a set of spots, curves, and shadings is incredibly difficult due to the almost infinite number of possible forms, and the huge cortical territory dedicated to encoding visual form. For making headway into this problem, it would be ideal if there were a small piece of brain specialized to encode a single visual form. This situation, surprisingly, exists: functional magnetic resonance imaging (fMRI) reveals six small regions in the macaque temporal lobe and three regions in the frontal lobe that show strongly increased blood flow in response to faces compared to images of other objects. In my talk, I will discuss out experiment to analyze in detail the neural mechanisms underlying face processing. We are combining electrophysiological, fMRI, and behavioral techniques in macaques to address how faces are detected and recognized. Face detection: It is widely assumed that the result of early visual processing is a contour representation of object shape. Our experiments indicate that local contrast features also play a critical role in shape coding. Using fMRI-guided electrophysiology, we targeted single-unit recordings to the macaque middle face patch. We presented a face-like collage of 11 luminance regions and modulated the contrast between regions while keeping all contours constant. This stimulus drove about half the cells from no response to a response greater than that to a real face. The critical factors determining response were the sign and magnitude of contrast between pairs of regions. Preferred contrast polarities were strikingly consistent across cells, suggesting a common computational strategy. Face identification: Primates can recognize faces across a range of viewing conditions. Representations of individual identity should thus exist that are invariant to accidental image transformations like view direction. We targeted the two middle patches (ML, middle lateral, and MF, middle fundus) and two anterior patches (AL, anterior lateral, and AM, anterior medial). We found that the anatomical position of a face patch was associated with a unique functional identity: Face patches differed qualitatively in how they represented identity across head orientations. Neurons in ML and MF were view-specific; neurons in AL were tuned to identity mirror-symmetrically across views, thus achieving partial view invariance; and neurons in AM, the most anterior face patch, achieved almost full view invariance.

- 11 - The 392nd session: Brain and Cognitive Sciences

Neural decision computation and memory in cortical circuits Xiao-Jing Wang, Ph.D Yale University

What is a “cognitive-type” neural circuit? That is the question I would like to address at this meeting. In mammalian neocortex, it is believed that there is a canonical microcircuit, as advocated by Kevin Martin and Rodney Douglas. If so, does it mean that any microcircuit is the same regardless of whether it is dedicated to early sensory processing, integration of information across modalities, decision-making, rule-based behavior, etc? The key to answer this question rests on the idea that quantitative differences give rise to qualitatively different behavior, therefore graded differences in microcircuit properties can lead to novel processes and functions. I will illustrate this perspective by discussing a biophysically-based neural circuit model for working memory, our ability to internally and actively hold information in the absence of external stimulation. As it turns, the circuit mechanism we identified requires low reverberation? that depends on the NMDA receptors, and competition that depends on feedback synaptic inhibition. This model prediction is now confirmed by iontophoresis experiment in behaving monkeys. And the same model has been successfully applied to various types of decision-making behaviors, selective attention, inhibitory control of action, suggesting a canonical “cognitive-type” microcircuit. I will discuss my perspectives on where we could go from here, in order to build a circuit neurobiology of cognition.

- 12 - The 392nd session: Brain and Cognitive Sciences

Dopamine Reveals Neural Circuit Mechanisms of Decision Making: From Fruit Fly to Human Beings Aike Guo, Ph.D Institute of Neuroscience, SIBS, CAS and Institute of Biophysics, CAS

The neurotransmitter dopamine (DA) plays a crucial role in motivational control: rewarding, aversive, and alerting (Bromberg-Martin et al, 2010).

It should be not surprisingly that dopamine is also involved in the control of reinforcement, motivation, arousal and stimulus saliency in the fruit fly Drosophila (Scott Waddell, 2010).

“Life is endless series of decision, regardless of Drosophila or human beings”. It has been established that DA signals from the mammalian VTA (ventral tegmental area) implement the gating mechanism by controlling the ‘‘filter’’ of afferent information into PFC (Montague et al., 2004). Here we show that even Drosophila can make clear-cut, salience-based decisions when faced with competing alternatives. In our ‘color/position’ dilemma, individual flies were conditioned to choose a flight direction in accordance with the color (green and blue) and par position (Upper and lower) in the flight simulator, and decision tests with the color and position cues mismatched after the training. The trained flies resolve the ‘position/color’ dilemma by taking into account the reliability of the current information during retrieval. The decision curve of wild-types (WTB and CS) exhibited a sharp and complete transition as a function of relative salience between two cues, which could be fitted to a sigmoid function. It could be interpreted as a form of switch between two stable ‘memory attractor’ states. The switch point between the two attractors was approximately at the point of symmetric salience between the two conflicting visual cues. Using binary expression system: “GAL4-Enhancer Trapping & Region-Specific Gene Expression”, we have demonstrated that DA plays important role in saliency-based decision making in fruit fly. The nonlinear decision behavior became linear performance when DA activity in the entire brain or synaptic activity specifically in mushroom bodies (MBs, a specific brain region in fly’s central brain) was blocked. By similar genetic manipulation, we revealed that the decision making in Drosophila consists of two phases: an early phase that requiring DA and MB activities and a late phase independent of these activities. Our immunohistochemical studies showed that the MB vertical lobe is densely innervated by the DA axons (Zhang et al., 2007). Thus DA-MB circuit regulates salience-based decision-making in flies. We have also found that flies can make dynamic balance between maintaining the existing choice and switching to a new decision which requires DA-MB activities again. Thus, we suggested that the DA-MB circuit regulates salience-based decision making in Drosophila by both gating inhibition and gain-control mechanisms (Zhang et al., 2007; Guo et al., 2009; Guo et al., 2010; Wu and Guo, 2011). Taken together, the DA-MB circuit in fly may operate in the same or similar way as the large VTA-PFC circuit in human. Thus we have demonstrated that “even animals with very small brains can behave in a surprisingly rational manner under a broad range of conditions.”(Glimcher et al., 2005). Therefore, from an evolutionary viewpoint, the general computational rule underlying salience-based decision-making processes seems to be conserved.

- 13 - The 392nd session: Brain and Cognitive Sciences

References： 1. Elhan S. Bromberg-Martin, Masayuki Matsumoto and Okihide Hikosaka (2010), Dopamine in Motivational Control: Rewarding, Aversive, and Alerting, Neuron 68:819-834. 2. Paul W. Glimcher, Michael C. Dorris, and Hannah M. Bayer (2005), Physiological utility theory and the neuroeconomics of choice, Games Econ Behav 52(2):213-256. 3. Aike Guo, Ke Zhang, Yueqing Peng and Wang Xi (2009), Heisenberg’s Roadmap Guides our Journey to the Small Cognitive World of Drosophila, J Neurogenetics 23:100-103. 4. Aike Guo, Ke Zhang, Yueqing Peng and Wang Xi (2010), Research Progress on Drosophila Visual Cognition in China, Science China (Life Sciences) 53(3):375-384. 5. P. Read Montague, Steven E.Hyman and Jonathan D.Cohen (2004), Computational roles for dopamine in behavioral control, Nature 431:760-767. 6. Scott Waddell, (2010), Dopamine reveals neural circuit mechanisms of memory, Trends in Neuroscience 33:457-464. 7. Zhihua Wu and Aike Guo (2011), A model study on the circuit mechanism underlying decision-making in Drosophila , Neural Networks, 24，333-344. 8. Ke Zhang, Jianzen Guo, Yueqing Peng, Wang Xi and Aike Guo (2007), Dopamine-Mushroom Body Circuit Regulates Saliency-Based Decision-Making in Drosophila. Science 316:1901-1904.

- 14 - The 392nd session: Brain and Cognitive Sciences

Language Complexification and Ambiguityi. William S-Y. Wang, Ph.D Chinese University of Hong Kong

Language, with its combination of design features1, emerged perhaps 100 kya, inferring from the archeological and fossil data currently available. A major challenge in the life sciences is to determine how language is built upon the various biological and cognitive systems that pre-date its emergence, and how language in turn influences the development of these systems. With the rapid advances in neuroscience in recent decades, we are now in a good position to take on this challenge. Among the many issues of deep cognitive interest, a particularly intriguing one is how the brain deals with the abundant ambiguity found in every language.

Language is a complex adaptive system2, which consists of several interacting components that jointly define its lexicon, including phonology, morphology, syntax, semantics and pragmatics. Each of these components is constantly adapting to its changing environments, both physical and social. These components are also sensitive to biological constraints, which are quite different for the speaker and for the hearer.

The first languages which arose during the Late Pleistocene within the various prehistoric communities3 must have been simple utterances, based largely on suprasegmental features, used to supplement actions and gestures. As the number of distinct utterances expanded with the growth of culture, the words became more numerous (lexicon), and conventions arose for putting the words together (morphology and syntax) to represent novel thoughts and to convey increasingly complex meanings.

A lexicon can be expanded various ways as new needs arise. The most productive way is to assign new senses to existing words, such as ‘mouse’ for the computer accessory, based on resemblance in shape to the animal. There are subsets of the lexicon which are semantically extended whole-sale, such as the sharing of adjectives and prepositions to refer to space and to refer to time. This use of polysemy for expanding the lexicon is of course language specific. For example, English ‘long’ and Chinese 長 can refer to both space and time; but these concepts would require different words in Russian, namely, длиный and дорогий, respectively.

Polesemy is one major source for ambiguity at the lexical level, where a word has several related senses. The other major source for lexical ambiguity is sound change, creating homophony on a large scale. For instance, the words ‘reel’ and ‘real’ were pronounced with different vowels in an earlier stage of English, as revealed by their spelling. (The older

1 Hockett, C.F. & S.A. Altmann. 1968. A note on design features. Animal Communication, ed. by T.A. Sebeok, 61-72. Bloomington: Indiana University Press. 2 王士元. 2006. 語言是一個複雜適應系統 [Language is a complex adaptive system.]. 清華大 學學報 (哲學社會學版) 21.5-13. 3 Freedman, D.A. & W.S-Y. Wang. 1996. Language polygenesis: a probabilistic model. Anthropological Science 104.2.131-8.

- 15 - The 392nd session: Brain and Cognitive Sciences pronunciation of ‘real’ is better preserved in the word ‘reality’, where the suffix ‘-ity’ blocked the change.) As an example from Chinese, 涩，瑟，and 塞 are pronounced saap3, sat1, and sak1 respectively in . In Mandarin, however, there has been a whole sale loss of syllable final consonants, similar to what happened in French, and these three words are all pronounced se4.

When an ambiguous word, due either to polysemy or homophony, occurs in a sentence, the linguistic context usually disambiguates it. Presumably this is why languages tolerate large amounts of polysemy and homophony, rather than create new words to forestall miscommunication. For example, the English feminine pronoun her is polysemous, corresponding to two masculine counterparts, his and him, which are in the possessive and objective case respectively. Thus in the sentences I saw her book and I saw her jump, the her is not ambiguous. On the other hand, words like duck and fly are polysemous in being able to function as nouns or as verbs. Therefore sentences like I saw her duck and I saw her fly retain their ambiguity, though again these are usually resolved by larger contexts, linguistic or extralinguistic. There are many other types of ambiguity which arise at the sentential level4.

Another interesting large class of ambiguity has to do with figurative, as opposed to literal, language, such as in metaphors and idioms. As examples, the metaphor he is a mouse can be used to mean someone lacks courage, and the idiom he kicked the bucket means ‘he died’, although both constructions can be interpreted literally as well. Figurative use is abundant in all languages and highly culture specific though there must be universal elements as well common to all human cognition.

It seems that some aphasic patients have trouble retrieving the figurative meaning of such constructions. For instance, in Italian the construction Vuotare il sacco has the literal meaning of to empty the sack, but the figurative meaning of to confess something. When this construction was presented to aphasic patients, many produced a response triggered by the literal interpretation, i.e., Zaino (rucksack)5 . There is an increasing literature on the use of figurative language for investigating various brain disorders, including Alzheimer’s disease6 and schizophrenia7.

A plausible scenario for processing ambiguous utterances is that the brain must hold in store all possible interpretations of an utterance until a disambiguating context becomes available to determine the appropriate interpretation. From the viewpoint of cognitive neuroscience the challenge is to understand the exact mechanisms used by the brain in such scenarios; and whether these mechanisms are generally used in other forms of perception that evolved earlier, and only recently adopted in language processing.

4 Wang, W.S-Y. 2011. Ambiguity in language. Korea Journal of Chinese Language and Literature. 5 Cacciari, C., F. Reati, M. R. Colombo, R. Padovani, S. Rizzo & C. Papagno. 2006. The comprehension of ambiguous idioms in aphasic patients. Neuropsychologia 44.1305-14. 6 Papagno, Costanza. 2001. Comprehension of metaphors and idioms in patients with Alzheimer's disease. Brain 124.1450-60. 7 McKenna, Peter & Tomasina Oh. 2005. Schizophrenic Speech: Making Sense of Bathroots and Ponds that Fall in Doorways: Cambridge University Press.

- 16 - The 392nd session: Brain and Cognitive Sciences

Unfortunately our knowledge on these topics acquired over recent years is based on a very small sample of the world’s languages and cultures. This point has been highlighted by Henrich et al who refer to this sample as WEIRD8. To achieve a balanced perspective on the neuroscience of cognition for our species, the diversity of cultures, especially spoken and written languages, must be fully taken into account in future work.

8 Henrich, Joseph, Steven J. Heine & Ara Norenzayan. 2010. Most people are not WEIRD. Nature 466.29. Henrich, Joseph, Steven J. Heine & Ara Norenzayan. 2010. The Weirdest People in the World? Behavioral and Brain Sciences 33.61-135.

- 17 - The 392nd session: Brain and Cognitive Sciences

Brain Systems Responsible for Reading Li-Hai Tan, Ph.D State Key Laboratory of Brain and Cognitive Sciences University of Hong Kong

Reading is an important means to acquire world knowledge. Without proper skills in reading and writing, children will be hampered in their educational outcomes, social-emotional development, and eventual quality of life. According to recent estimates, approximately 7-15% of children can be classified as dyslexic readers who exhibit severe problems with reading that are unaccountable by any kind of deficit in general intelligence, sensory acuity, educational opportunity, and motivational factors. For readers of alphabetic (e.g. English) languages, reading impairment is critically associated with a core phonological processing deficit which does not co-exist with a general visual processing dysfunction in the majority of dyslexics. Recent neuroimaging studies have demonstrated that the phonological deficit is associated with weak reading-related activity in left temporoparietal and occipitotemporal regions, and this activity difference may reflect reductions in gray matter volume in these areas.

What is the brain mechanism underlying Chinese reading and reading difficulty? Because the Chinese writing system presents sharp contrast with alphabetic writing systems, research with Chinese is important to advance our understanding of the universality and particularity of the dysfunction of reading systems in the brain. Recent functional magnetic resonance imaging (fMRI) studies show that different brain systems are recruited when children read in Chinese and English. Specifically, the left middle frontal gyrus responsible for verbal working memory critically mediates Chinese reading, whereas the left posterior temporoparietal regions critical for English reading are less involved in Chinese word processing. The evidence also suggests an asymmetry: English readers have a neural network that accommodates the demands of Chinese by recruiting neural structures less needed for English reading. Chinese readers have a neural network that partly assimilates English into the Chinese system, especially in the visual stages of word identification.

- 18 - The 392nd session: Brain and Cognitive Sciences

Promoting Plasticity and Functional Recovery in the Brain Sarah Dunlop, Ph.D Experimental and Regenerative Neuroscience, School of Animal Biology, The University of Western Australia

Central nervous system (CNS) trauma is intractable and, in the peripheral nervous system (PNS), outcomes are less than optimal. Despite significant advances in cellular and drug-based therapies, there is no cure for either type of neurotrauma. Two animal models, one involving the CNS (lizard visual system) and the other the PNS (rat facial nerves) show striking parallels with robust regeneration. However, regeneration is highly inaccurate with axons neither finding appropriately located target cells nor restoring fast, secure synaptic transmission. Consequently, regeneration is dysfunctional with lizards remaining blind and rats unable to “whisk” or blink. Task-specific training in both models restores correctly located connections as well as normal behaviour. Together with other mounting evidence, these clear mechanistic examples of the effects of training on nerve cells have implications for translation into rehabilitation for neurotrauma. Such task-specific training, which appears to promote neurological recovery, is distinct from general exercise that has been shown to benefit physical and mental health and is also important. To examine the effects of task-specific training, the Victorian Neurotrauma Initiative and others have provided funding for “SCIPA” (Spinal Cord Injury and Physical Activity) to undertake 3 randomised controlled clinical trials for patients with spinal cord injury. SCIPA involves all spinal units in Australia and New Zealand and is using novel exercise programs to move the paralysed limbs. The 3 trials involve different types of intensive exercise for patients close to the time of their injury while they are in intensive care, for in-patients and for out-patients. SCIPA is also developing a community-based program to help ensure continued access to exercise programs for this life-time injury.

- 19 - The 392nd session: Brain and Cognitive Sciences

Brain Injuries: Cognitive Function Impairment and Its Management

Wai Sang Poon, Ph.D Division of Neurosurgery, Department of Surgery, Prince of Wales Hospital The Chinese University of Hong Kong

Advances in the understanding and treatments of brain injuries in the last thirty years, both experimental and clinical, have not lead to the emergence of any management strategy that resulted in a favourable clinical outcome. The widely accepted explanation for the failure is the heterogeneity of brain injuries [1]. In this presentation, a critical review of current classification, clinical assessments [2] and management strategies of brain injuries with emphasis on cognitive function impairment will be discussed for the benefits of designing future studies.

References:

1. Saatman KE et al 2008. The Classification of Traumatic Brain Injury for Targeted Therapies. J Neurotrauma 25: 719-38.

2. Truelle JL et al 2010. Quality of life after TBI: the clinical use of the QOLIBRI a novel disease-specific instrument. Brain Inj 24(11):1272-91.

- 20 - The 392nd session: Brain and Cognitive Sciences

Promotion of Neurogenesis and Preservation of Dendrites by Anti-Aging Lycium Barbarum (Wolfberry) in Experimental Depression Kwok Fai So, Ph.D Department of Anatomy, University of Hong Kong

Lycium barbarum (Wolfberry) is a traditional Chinese medicine herb which can be used for both disease treatment and as a functional food. For a long time, it is believed that Wolfberry has anti-aging properties but there is little scientific evidence to support this concept. Depression is a severe mental disorder related to aging and neurodegeneration. The current view on depression is related to the monoamine hypothesis. Depression is due to a deficiency in one or another of three monoamines: serotonin (5-HT), noradrenalin and /or dopamine. More recently, two new hypotheses are being proposed. 1. Neurogenesis hypothesis: decrease in the synthesis of new neurons in the adult hippocampus might be linked to depressive disorders (Kempermann and Kronenberg, 2003: Biol. Psychiatry). 2. Neuroplasticity hypothesis: Morphologic changes in the hippocampus of depressed patients reflecting alterations at the level of granule cells and the complexity of dendritic processes. We will provide evidence on how Wolfberry can be effective in reducing depressive-like behaviour in rats via the neurogenesis and dendritic plasticity mechanisms.

Hypercotisolemia is a common pathological situation which can lead to deficits in cognition and emotion, thus many scientists adopt injection of corticosterone (CORT) to induce experimental depression. Previous data in our lab showed that different doses of corticosterone injection lead to differential outcome of behavioral performance. Exercise, an effective strategy known to rescue cognitive and emotional deficits, was able to restore cognitive and emotional deficits in an experimental hypercortisolemia induced by 40 mg/kg, but not 50 mg/kg corticosterone (CORT) injection. At present study, we wanted to investigate whether polysaccharides of Wolfberry (also known as Lycium Barbarum) (LBP) could rescue those behavior impairments in the experimental hypercortisolemia induced by 40 mg/kg and 50 mg/kg corticosterone injection. All experimental animals were divided into four groups: PBS feeding group (PBS group), PBS feeding plus CORT injection group (PBS+C group), LBP feeding group (LBP group), and LBP feeding plus CORT injection group (LBP+C group). Our results showed that LBP feeding can restore depression-like behavior caused by 40mg/kg and 50 mg/kg CORT injection. Additionally, escape latency in the Morris Water Maze on day 2 were significantly decreased in the LBP+C group, when compared to those in PBS+C group. Following 40mg/kg CORT injection, BrdU positive cell number in the hippocampus significantly increased in the LBP+C group when compared to that in the PBS+C group. In 50mg/kg CORT injection, the BrdU positive cell number significantly decreased to 44.4% and 55.8% in the PBS+C and the LBP+C group, when compared to the PBS group, respectively. No difference were observed between PBS+C and LBP+C group, but the ratio of BrdU/Doublecortin double positive cell number to BrdU positive cell number in hippocampus was significantly increased in the LBP+C group, when compared to those in the PBS+C group. Furthermore, Western bloting data showed that the amount of postsynaptic density protein-95 significantly increased in the LBP+C group, when compared to those in the PBS+C group. In conclusion, our data indicate that LBP treatment is beneficial to those cognitive impairment and depressive behavior in an experimental hypercortisolemia induced by 40 mg/kg and 50 mg/kg

- 21 - The 392nd session: Brain and Cognitive Sciences

CORT. Wolfberry might be considered as a novel treatment approach for depression in clinical situation.

- 22 - The 392nd session: Brain and Cognitive Sciences

LIST OF PARTICIPANTS/ SPEAKERS/ ORGANIZERS

Albert, Yu Cheung-Hoi（于常海） Professor Peking University Alice, Chan Hiu Dan（陳曉丹） Assistant Professor Nanyang Technological University Guoying Cao（曹國英） Liaison Office of the Central People’s Government Chun-Qi Chang（常春起） Assistant Professor University of Hong Kong Yi-Zhang Chen（陳宜張） Professor The Second Military Medical University Chetwyn Chan（陳智軒） Professor Hong Kong Polytechnic University Yu-Pang Cho（左雨鵬） Professor Chinese University of Hong Kong Doris Tsao Assistant Professor California Institute of Technology Aike Guo（郭愛克） Professor Institute of Neuroscience, CAS, Shanghai Jun Han（韓軍） Xiangshan-Science Conferences Shing-Yan Huen（禤承恩） Assistant Professor University of Hong Kong John A. Spinks Professor University of Hong Kong Ken Yung（翁建霖） Professor Chinese University of Hong Kong Ying Li Professor City University of Hong Kong Zhendong Niu（牛振東） Professor Beijing Institute of Technology Feng Pan（潘鋒） Science Times Gang Peng（彭剛） Associate Professor Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Peter Fox Professor University of Texas Health Science Center Wai-Sang, Poon（潘偉生） Professor Chinese University of Hong Kong Raymond, Chang C.C.（鄭傳忠） Assistant Professor University of Hong Kong Robert Desimone Professor Massachusetts Institute of Technology Sabine Kastner Professor Princeton University Samson Tam, Wai-Ho（譚偉豪） Legislative Council (Information Technology) Sarah Dunlop Professor The University of Western Australia Kwok-Fai, So（蘇國輝） Professor University of Hong Kong Li-Hai, Tan（譚力海） Professor University of Hong Kong Tatia Lee, M. C.（李湄珍） Professor University of Hong Kong Xiao-Jing, Wang（汪小京） Professor Yale University Zhishen, Wen（溫植勝） Assistant Professor Hong Kong Shue Yan University William Wang, S-Y（王士元） Professor Chinese University of Hong Kong Wutian, Wu（吳武田） Professor University of Hong Kong Guoxiang, Xiong（熊國祥） Professor Xiangshan-Science Conferences

- 23 - The 392nd session: Brain and Cognitive Sciences

Bingxin Yang（楊炳忻） Professor Xiangshan-Science Conferences Xueqing You（游雪晴） Science and Technology Daily Xian’en Zhang（張先恩） Professor Ministry of Science and Technology of China Yan Zhao（趙彦） Science Times Ji-Zong Zhao（趙繼宗） Professor Beijing Tiantan Hospital

Local Working Group：

Yuxiang Liang（梁玉香） University of Hong Kong Min Xu（徐敏） University of Hong Kong Veronica Kwok（郭沛殷） University of Hong Kong Vivien Lok（駱穎君） University of Hong Kong

- 24 -