graf

Future Projects
Conceptual Background

Control of gaze, stabilization of posture, and coordination of movements require coordinate transformations between sensory and motor systems of various degrees of freedom. These transformations imply complicated calculations for multisensory interaction, sensory-motor integration, and motor coordination. One way for the brain to deal with multi-dimensional systems is to employ constraints within reference frame systems, in order to simplify these calculations. Our line of research over the last years has been dealing with spatial aspects of postural control mechanisms as a vehicle to understand the operational functioning of the brain. The actual object of our studies is orienting behavior that is composed of eye and head movements in a temporally and spatially coordinated manner in most animals. It is a vital requirement for most animals to move about their habitat. Besides spontaneous impulses, this behavior results from different sensory inputs that produce a motor response involving a number of muscle systems. At issue is how these inputs are efficiently integrated and transformed to produce meaningful responses.

A number of conceptual and programmatic approaches to study brain function can be envisioned in this context in light of the almost unparalleled accuracy with which modern experimental technology allows us to describe the relevant input-output relationships. First of all, we will be able to uncover the mechanisms which underlie sensory-motor transformations, for instance, how visual images are stabilized during head movements by compensatory eye movements. Since several sensory systems usually interact to produce a co-contraction of several muscles resulting in a motor reaction that involve several joints, we can re-formulate the problem as the study of multi-sensory multi-motor transformations. Since all orienting behavior takes place in three-dimensional space, it necessitates studying how a three-dimensional movement is detected by the respective peripheral sensory systems, how it is represented in central neurons, and how the appropriate three-dimensional motor output is produced. Furthermore, we will also be able to address questions of cognition and perception of three-dimensional movement space, and of motor learning. Three major avenues of programmatic approaches will be outlined in this research proposal the study of 1) POSTURE and its underlying sensory and biomechanical parameters, 2) peripheral and central REPRESENTATION OF THREE-DIMENSIONAL SPACE, and 3) ADAPTIVE PLASTICITY following altered sensory input or motor performance and ADAPTATION to a different life situation.

Previously, we studied the geometry and central representation of vestibular and visual sensors regarding the three-dimensional coding of movement space, as well as related motor systems involved in orienting responses and postural control (in particular eye and head movement systems). From these studies we learned that significant constraints are imposed to reduce the degrees of freedom of these sensory and motor systems. For instance, the semicircular canals of the labyrinth have a three-dimensional orientation in the head that allows retrieval of a given movement vector in space with the least number of sensors possible. Multisensory integration is facilitated by organizing sensory systems of similar behavioral context in the same geometrical framework as the semicircular canal orientation. At the motor output side, the necessity for large-scale sensory-motor transformations is decreased by aligning the pulling direction of certain muscles with that of the sensory receptors. This is the case, for instance, with the semicircular canals and the extraocular muscle system. A similar reduction in degrees of freedom is present in the head movement system. In this case, the motor axes around which head-neck movements can be performed are also significantly constrained while preferred joints for particular movements in space become obvious.

The discovery and description of these preferred sensory and motor coordinates not only enlightened our understanding of postural control mechanisms. It also established a framework to study principles of brain function in a meaningful and biologically relevant way. As a logical consequence, this conceptual approach also involves questions of embryological development, in particular how sensory and motor components are put together, and of vertebrate evolution, specifically why and how biological systems employ preferred coordinates for sensory-motor control. We can summarized this concept in one sentence: We are interested in the question why we look how we look.

As a long-term goal, the proposed projects should demonstrate that the brain utilizes preferred coordinates to reduce the degrees of freedom of sensory-motor systems when executing complicated integrative tasks of potentially high degrees of freedom, in order to simplify neuronal operations and to economize brain function in light of a narrow window of optimal brain-to-body ratio.

I. Posture
Higher vertebrates adopt a stereotyped resting posture of the head-neck arrangement. The cervical vertebral column is oriented vertically, and the horizontal semicircular canals are tilted upward from earth-horizontal. Subsequent investigations quantified the range of motion of the different articulations of the head-neck ensemble in monkey, cat, rabbit and guinea-pig by X-ray photography and dissection. We made significant progress in regard to establishing the degrees of freedom of the system and recognizing the strategies employed by different animal species. Our data indicate that biomechanical constraints limit the number of possible solutions how an animal can perform a given orienting movement starting from the stereotyped resting position. For example, in the sagittal plane, the upper cervical vertebrae allow only flexion, whereas the lower cervical and upper thoracic vertebrae only permit extension of the vertebral column. Thus, mechanical constraints almost exclusively provide the biomechanical means of maintaining an upright cervical vertebral column with minimum energy expenditure. From resting position, lowering of the head is only possible by moving the C6-Th3 vertebrae out of their extreme extension into the flexion direction, and raising of gaze is only possible by extension of the atlanto-occipital articulation. Our cineradiographic data from freely moving animals indicate that in rabbits, as one extreme example, head posture is maintained entirely independent of cervical vertebral column orientation, whereas in the monkey, orientation of the head almost entirely rests on orientation of the cervical vertebral column in space. This strategy seems directly related to the limited range of motion of the atlanto-occipital articulation of approximately 20 deg, much in contrast to the rabbit where the range of motion for the same articulation averages to more than 110 deg.

Most recently, we have started to describe the biomechanical parameters that underlie human head-neck movements using cineradiography. In human as in other vertebrates the elements of the cervical column move as an ensemble to provide head orientation in space. However, when moving the head in a forward/backward (flexion/extension) direction, humans display only one functional joint as in monkey, in contrast to two functional joints in quadrupedal vertebrates. When comparing our more general data from the monkey to those obtained in humans, we observe striking similarities. When the analysis of monkey cervical column biomechanics has been brought to level as demonstrated in humans, we will be able to study head-neck movements in a biologically relevant model including neck muscle kinematics and neuronal substrates. Given the large similarities between human and primate in this particular case, we will be able to develop a model of human head movement control that will have widespread relevance for clinical and other scientific applications. One of the more important questions regarding the head-neck system is the sensory basis of posture. At issue is the implementation of the subjective vertical to maintain head posture. In order to assume a vertically oriented cervical column, the animal has to have a notion about the direction of gravity. Thus, we are planning a project on the role of otolithic and visual cues for establishing the multi-sensory end-product of the subjective vertical.

All projects taken together will provide an overall description of the head-neck movement system and how it is embedded in the behavior of orienting responses in three-dimensional space including its biomechanical, sensory and motor components, and the neuronal structures that underlies it.

II. Central Representation of Three-Dimensional Movement Space
Our previous investigations on visual representation of movement space focused on the rabbit, a lateral-eyed animal, and several promising experimental avenues will be pursued in the future in this animal. However, given the uniform structure of the peripheral, sensory and motor apparatus in regard to compensatory eye movements, it is now important to determine the extent to which that same central representation is realized in a frontal-eyed animal, preferably in primates. There is ample evidence that cortical factors play a more important role in higher animals of the vertebrate hierarchy, and the monkey preparation will give us the ultimate model to describe the end product of a long evolutionary development. Furthermore, because of the training aspects of the primate preparation all these systems can be tested in a meaningful behavioral context.

1) Neuronal Substrates of Reflex and Voluntary Eye Movements
One of these behaviors to be analyzed are visuo-vestibulo-ocular reflexes and the reference frames they operate in. This particular approach reveals organizational principles of brain operation at the systems-level and will ultimately contribute to an understanding of sensory-motor behavior. The vestibulo-ocular reflex (VOR) is a particularly valid tool to achieve this objective, because of the unparalleled accuracy with which its temporal and spatial parameters can be determined. Specifically, we will focus on the neuronal signals carried by second-order vestibular neurons in alert behaving monkeys and will describe their structure-function relationship to compensatory and voluntary eye movements in all three dimensions of physical space. This distinction is important because compensatory eye movements are three-dimensional, whereas voluntary eye movements that obey Listing's law are two-dimensional.

2) Transneuronal labeling of eye mouvement circuits with rabies virus
The functioning of entire neuronal assemblies is one of the foremost interests of current research that forms the leading edge in today's neuroscience, imaging technique, computer sciences and robotics. Although neuronal network theories introduce algorithms that are thought to resemble actual nervous system operations, functional mechanisms of neuronal nets are still elusive, largely because of a lack of knowledge about connectivity of all nervous structures concerned. Such knowledge can be provided by a new transneuronal tracing technology that is based on the use of an attenuated strain of rabies virus as marker. While a number of transneuronal tracers are available, none has the specificity and potency of rabies virus. This viral tracer propagates exclusively via transneuronal transfer (specific cell-to-cell communication) between synaptically connected neurons. Sequential visualization of serially connected neurons of an entire networks occurs in the asymptomatic period of infection. The number of neuronal relays is basically unlimited, because the virus functions as a self-amplifying marker, by replicating in each relay, and can be easily detected immunohistochemically. This technology offers for the first time the opportunity to describe functionally related neuronal networks in their entirety, allowing assessment of the hierarchy of their building blocks, their accessory elements and their parallel processing characteristics.

The planned experiments will represent the first large-scale application of this powerful technology to the study of the primate brain. The eye movement has the advantage that some of its elements are already well described, and constitute established sub-systems. At the same time, clarification is needed concerning higher-order processing (e.g., cortical areas) and communication within the various sub-systems. The oculomotor system is also an excellent example of the modern concept of modular organization of central nervous operations. Thus, spatial modules exist in the three-dimensional orientation of the three pairs of muscles that move the eye, forming a natural reference frame that is reflected in the structure and function of the underlying neuronal circuits. Viral transneuronal tracing from prototypical eye muscles, horizontal (lateral rectus, medial rectus) and vertical (superior rectus, superior oblique will clarify the modular system organization of the different oculomotor networks in experimental primates and provide a clear understanding of their functional role in eye movement control, vision, visuo-motor coordination and cognitive and attentional processes. Specific subsystems and their role in three-dimensional coordination of eye movements will be studied: 1) the vestibulo-ocular system producing reflex eye movements during passive motion; 2) the velocity-to-position integration network involved in generation of eye position signals; 3) the saccadic network producing rapid eye movements for orienting and reflex functions; 4) the cerebellar system allowing adaptive plasticity and motor learning; and 5) the cortical and subcortical networks involved in motion perception, attention processes and voluntary control of saccadic and smooth pursuit eye movements. The results will be correlated with findings from human brain material of clinical cases with well-documented oculomotor disorders. The results will illustrate the intrinsic architecture of an entire contextually related neuronal network, i.e. from simple reflex mechanisms to perceptual functions, thus giving incentives about clinical-medical and industrial applications regarding oculomotor deficits and rehabilitation of human patients, as well as for computer science in the field of neurocomputers, artificial intelligence, and robotics.

3) Neuronal Correlates of 3-D Movement Perception
One critical aspect of the analysis of movement is the differentiation between objects moving through the extrapersonal space and self-motion. The latter involves rotational as well as translational displacements of eyes, head and body. Unequivocal interpretation of self-motion by the nervous system requires converging multisensory input taking into account the six degrees of freedom associated with rotation and translation in space. While there are three-dimensional motion sensors available in the labyrinth for detection of rotations (the semicircular canals) and translations (the otoliths) in physical space, visual information about a three-dimensional movement has to be extracted from a basically two-dimensional receptor cell surface (the retina). Visual information processing, however, constructs centrally a three-dimensional representation of this particular movement space. At the level of the brainstem, this visual movement space has been shown to share in the reference frame of the spatial orientation of the semicircular canals (66). In the neocortex, the issue of three-dimensional space coding has been mostly studied from the perspective of the optic flow field representation. Several anatomically and functionally distinct areas in the neocortex participate in motion analysis. Within the parietal lobes, specialized centers appear to be involved in higher-order processing of motion, e.g., areas MT, MST, 7a, VIP. However, knowledge of the multisensory processing performed in these different areas and their specific behavioral role is still lacking.
Recent neurophysiological evidence indicates that in these areas, a single sensory quality is no longer of perceptual importance, and efferent signals carrying meaningful messages contain multisensory information. Anatomically and functionally, the motion areas in the parieto-temporal regions are situated at an intermediate stage between lower-level motion analysis and motor output, and are involved in multisensory integration and perceptual and motor decision-stage operations. Our principal objective is to understand how various sensory inputs are combined at the level of single neurons, or networks of neurons in the primate neocortex in order to generate the three-dimensional reference frames for perception and movement. The interpretation of motion by the nervous system requires a comparison of congruent and conflicting input originating in different sensors, that takes into account the rotational and translational displacements of eyes, head and body in extra-personal space. Such interactions, and especially those occurring during navigation in the environment, are believed to be particularly important but have not yet been investigated systematically at the single cell level.

Our recordings in awake monkeys will later focus on neuronal responses in the parietal cortical area VIP (ventral intraparietal area) during optic flow, and rotational optokinetic and vestibular stimulation in head-free animals. The latter aspect is of particular importance, since up to now, only head-fixed preparations have been employed in comparable experimental set-ups. The need to employ the head-free paradigm arose because most recent results in central vestibular neurons involving reflex motor circuits have shown that neuronal responses in animals undergoing passive versus active head rotations are fundamentally different. Furthermore, our own results have demonstrated that VIP neurons have conflicting, i.e., contradictory on-directions for vestibular and visual direction-selective responses. These results were interpreted to allow a distinction between object-movement in extrapersonal space, i.e., a passive movement with regard to the observer, and active self-motion. The expected data will be incorporated in a model of the cognitive processes involved in active visual exploration of the environment and perception of self-motion.

4) Ontogeny of 3-D Space Representation
The third and last project in context of three-dimensional representation of movement space will determine the factors necessary for establishing the optokinetic reference frame within the retino-cerebello-vestibulo-oculomotor loop during ontogeny. This project will be conducted in rabbits. Specifically, we will study the temporal and spatial relationship of complex and simple spike discharges of floccular Purkinje cells by extracellular recordings during three-dimensional optokinetic stimulation in dark-reared rabbits. The long term goal of this particular project is to evaluate further the putative role of the climbing fibers in the general context of cerebellar function and theory.

III. Adaptive Plasticity and Adaptation
The mechanisms of motor learning and adaptation to a new life situation following altered sensory input or motor behavior is still one major focus of neuroscience research. Although lately this field has seen the implementation of neuropharmacological and molecular methods on a large scale, we nevertheless feel that ultimately these pressing questions have to be answered in a whole animal paradigm. In this context we propose two major lines of research of proven relevance regarding adaptive behavior:

1) Vestibulo-oculomotor Plasticity
The study of motor learning within the vestibulo-oculomotor system (vestibulo-oculomotor plasticity) has been the focus of controversy for a number of years since Ito developed his vestibulo-oculomotor plasticity paradigm following the publication of the cerebellar theories of Marr and Albus. It stated that the site of motor learning for the compensatory eye movement system is located in the flocculus, a part of the archicerebellum. Thus far, no conclusive evidence for a site of motor learning in the cerebellum has been provided. The monkey paradigm will allow also to study motor learning and its underlying principles in the specific example of vestibulo-oculomotor plasticity and the role the cerebellum plays therein. We believe that now the necessary theoretical and experimental background material is available to approach this problem in an experimentally sound way since only recently the most important building blocks have become available. The first important aspect is merely anatomical. The monkey "flocculus" that has been studied thus far as such is not the flocculus but the paraflocculus. In that respect it is imperative to identify the proper floccular location site of neurons that have relevance for the adaptation the eye movement reflex. Secondly, from our physiological data in the rabbit and prior anatomical studies, we know about the zonal structure of the cerebellum, and the flocculus in particular. Thus, it is necessary to study the neuronal responses in particular zones that are related to eye movements in particular planes of space. Third, only recently have neurons been identified that are output neurons of the cerebellar flocculus. Thus, the study of any vestibulo-oculomotor neuron will not yield an answer, but only the so-called floccular target neurons have to be studied and may be related to an altered behavioral context. With our methodology at hand we will be able to study all these neurons in both a stimulus related context as well as in a structure function context by combining extracellular recordings with intracellular recording methods and staining methods in a behaving animal.

2) Neuronal Adaptation Accompanying Metamorphosis in Flatfish
Flatfish are a natural paradigm to study general principles of adaptive changes (embryonic and genetic plasticity) in the vertebrate nervous system, because of a 90 deg relative rotation of the oculomotor versus vestibular reference frames during metamorphosis. Specifically, the bilaterally asymmetric placement of the eyes in the adult animals offers a unique opportunity to investigate the site(s), mode and time course of adaptation to altered conditions in the vestibulo-ocular reflex circuitry of vertebrates. In postmetamorphic animals, vestibulo-oculomotor pathways form new connections between the horizontal canals and vertical extraocular motoneuron pools on both sides of the brain to allow compensatory eye movements during swimming. After characterizing the neuronal structures responsible for adaptation of compensatory eye movement reflexes (VOR) in adults, we are now in a position to uncover the embryological and genetic mechanisms for this adaptation, in particular the nature of the neurons and their origin, and the time course of their development during embryogenesis. To that effect flounders will be raised from egg state through metamorphosis and cell birth dates and cell lineages of vestibulo-oculomotor neurons responsible for the VOR adaptation will be determined. In situ hybridization of GAP 43 will relate the axonal growth of oculomotor afferents to behavioral and previously obtained cell birth date data. Furthermore, the adaptive changes in the flatfish open an opportunity to address larger issues of triggering metamorphic changes, including the genetic, epigenetic and environmental factors involved. Control mechanisms for cell identity, cell to cell interaction and flatfish phenotype determination will establish the genetic basis of flatfish metamorphosis and related behavioral and neuronal changes using gene expression (i.e., a possible flatfish homeobox) and molecular biology technology.