Fos expression in the vestibular brainstem: What one marker can tell us about the network

Brain Research Reviews 50 (2005) 200 – 211 www.elsevier.com/locate/brainresrev

Review

Fos expression in the vestibular brainstem: What one marker can tell us about the network Galen D. Kaufman* 7.102 Medical Research Building, University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77555-1063, USA Accepted 9 June 2005 Available online 22 July 2005

Abstract Fos inducible transcription factor expression in rodent brains (rats and gerbils) during manipulations of vestibular input is reviewed. Stimuli included centripetal hypergravity, unilateral labyrinth lesion or semicircular canal plugging, rotational axis cross-coupling (Coriolis forces), high and low rotational vestibulo-ocular reflex gain adaptation, translabyrinth galvanic stimulation, pharmacological manipulation, and combinations thereof. Each type of stimulation elicited unique but partially redundant response patterns in the vestibulo-olivo-cerebellar (VOC) network that reflect the origin and interaction of the labyrinth inputs. On the basis of these patterns, a trained observer can predict what the animal experienced during testing; the patterns of VOC Fos expression reveal a trace of recent genomic activity. Based on principal component analysis, VOC network modules associated with lesion recovery, spatial representation and the calibration of gravity, and optokinetic influences are proposed. Probable and possible gene targets of the Fos protein are also reviewed. D 2005 Elsevier B.V. All rights reserved. Theme: Sensory systems Topic: Auditory, vestibular, and lateral line Keywords: c-fos; Inducible transcription factor; Immediate early gene; Plasticity; Rodent; Rat; Gerbil; Neural network; Gravity; Coriolis; Ketamine

Contents 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Within a narrow non-lethal range of environmental changes or sensory pathology, the nervous system has a remarkable ability to adjust its network to retain function in a new situation. The cellular and molecular basis of this adaptive capacity is the subject of many neuroscience studies. The recognition of genomic mechanisms controlling these changes entices us with a molecule-to-behavior scaffold of understanding. Brainstem networks process * Fax: +1 409 772 5893. E-mail address: [email protected]. 0165-0173/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainresrev.2005.06.001

200 208 209

well-defined sensory modalities and are simpler compared to their cortical descendents. The geometrical nature of vestibular inputs in particular (head rotation, position, and translation) and an easily measured output (eye movement) make this system attractive for studying neural adaptation. The vestibulo-olivo-cerebellar (VOC) network is a well-studied circuitry of bilateral synaptically related neurons that influence each other and receive vestibular labyrinth and multiple sensory inputs [3]. This review will focus on the VOC and expression of the inducible gene regulatory transcription factor protein, Fos [37], during many types of challenging new environments that require network change. Although the VOC itself is reciprocally

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

connected to myriad other brain systems, it is these neurons that carry vestibular signals and are within three synapses of their source and can be shown to depend on a functional labyrinth for Fos expression [46]. While the vestibular system in general must coordinate a host of head inertia responsibilities, the vestibulo-ocular reflex (VOR) in particular is a highly conserved bilateral system for stabilizing visual input. In most animals, a high frequency rotational mechanism (semicircular canals, RVOR) contributes to retinal stability in conjunction with a low frequency translational (otolith, TVOR) and visual (optokinetic) assist. The basic anatomical pathways are known. The RVOR can be as simple as a three-neuron chain: sensory afferent, vestibular, and oculomotor neurons. The TVOR pathways are more polysynaptic for computational reasons [2]. Four basic functional cell types have been identified in the vestibular nuclei: position-vestibular-pause, burst-position, eye-head, and vestibular-only neurons, which differ in their responses and response polarity to head and eye movements [2,53,76]. Neuronal responses to vestibular inputs and training have been investigated, the role of each area has been explored, and some agreement has been reached as to their individual function. For example, the cerebellar flocculus is now well accepted as an important part of horizontal VOR gain adaptation [38,59]. Floccular target neurons (FTNs) are a subpopulation of vestibular nucleus neurons that receive direct inhibitory Purkinje cell projections from the cerebellum and are thought to be one physical site of adaptive changes. In contrast, floccular projecting neurons (FPNs) are a separate subpopulation of vestibular neurons that can influence floccular responses. In non-foveated animals, visual information reaches the vestibular system via retinal slip information, generating the optokinetic reflex (OKR), one type of slow eye movement [40]. Optokinetic signals passing through the accessory optic system reach the prepositus and vestibular nuclei, the cerebellum via mossy fibers, and the powerful climbing fibers of the inferior olivary dorsal cap which project contralaterally onto modular cerebellar networks of Purkinje cells. Therefore, another site of plasticity is the vestibulocerebellum, specifically Purkinje cells that can act as comparators of intended and actual movement. A long literature deals with the arguments for different sites of VOR plasticity (reviewed in [5]). Different ocular kinematics are required when a fovea is present. The optokinetic reflex [58] is supplemented by other visuomotor and target following reflexes (ocular following reflex, smooth pursuit) requiring cerebral cortical inputs. The VOR in three dimensions reflects both canal and otolith inputs and continues to be modeled [73]. A tightly coupled triad of sensory control and feedback therefore exists between the vestibular nuclei, inferior olive, and vestibulocerebellum. However, understanding the basic anatomy and function of vestibular neurons does not necessarily tell us very much about how the same system responds to functional

201

challenges. VOR reflexes are adaptable quickly with visual/head mismatch paradigms (minutes) and in both directions (high or low), but the mechanism of each gain direction apparently differs [25,79]. Adaptation depends on an intact inferior olivary nucleus, which provides corrective signals to cerebellar networks [83]. The sensorimotor brain is highly interconnected and can meld inputs of all kinds to extract behavioral responses. The brain is opportunistic and redundant in the sense of using whatever sensory or efference modality is present [5,7,78]. For example, optokinetic signals share at least some penultimate pathways prior to oculomotor neurons [58]. Recent models propose such combinations by using coordinate transformation of canal signals from a head-fixed to a spatial reference frame [33]. Vestibular cortical projections likely have important roles to play in perception and navigation [17,22,88]. Furthermore, there is good evidence that the brain can maintain multiple functional states or response patterns, as demonstrated by context-specific VOR gain adaptation [77] and yaw and pitch cross-axis adaptation [70]. Such adaptation appears to be constrained by the complexity of the adaptation task [86]; see also [51]. Many studies have now shown that there exists a complex interplay of stimulus direction, habituation, and adaptation in the VOR [12]. Recent evidence also supports the existence of multiple cellular mechanisms of VOR plasticity which operate over distinct time courses [27,79]. Fig. 1 outlines a three-dimensional dorsally oriented coronal view of the rodent VOC network in situ and serves as a guide for the simpler representations in Fig. 2. Furthermore, representative Fos immunolabeling in VOC neurons following hemilabyrinthectomy– a peripheral loss of vestibular input unilaterally – is shown within the regions [10,45,52,54,78]. Large (red) dots represent Purkinje cells in the ventral paraflocculus, and other dots represent other neurons of mixed types. Each letter in the abbreviation VOC represents a module of a triad: (1) vestibular nuclei complex, (2) inferior olivary complex, and (3) vestibular (posterior) cerebellum. These three modules are heavily interconnected with both excitatory and inhibitory projections in specific pathways, as well as commissural crossing. Each module has subregions and unique populations of cell types with special roles [89]. For example, vestibular primary afferent neurons are excitatory and reach, on the same side, both the vestibular nuclei and the vestibulocerebellum at granule cells. The cerebellar Purkinje cells form the only output of the cerebellum and send inhibitory projections ipsilaterally to the deep cerebellar nuclei and vestibular complex. The vestibular and inferior olivary complexes exchange reciprocal projections that are both crossed and uncrossed, excitatory and inhibitory [3]. And, the inferior olive projects slowly firing and powerful excitatory climbing fibers contralaterally to the Purkinje cells. These anatomical relationships only begin to frame the network. Both the instantaneous firing rate and firing

202

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

Fig. 1. A three-dimensional schematic view of the rodent vestibulo-olivo-cerebellar (VOC) network, with representative Fos expressing neurons following acute hemilabyrinthectomy. The inset shows the relative plane of the depicted coronal section in the whole brain. The Fos labeling can be directly compared to the all-or-none schematic in Fig. 2G. The interconnected triad of the VOC can be visualized with the (1) vestibulocerebellum (flocculus, paraflocculus, nodulus, uvula, and vermis) shown as a background profile and the (2) vestibular nuclei complex and (3) inferior olivary complex shown as if viewed from a caudodorsal central position. The locus coeruleus (LC) is also shown bilaterally, projecting rostrally from the vestibular nuclear complex. The drawing shows general relationships between major nuclei and is not meant to be anatomically precise. Abbreviations: floc, flocculus; g7, genu of the facial nerve; IOh, inferior olive beta subnucleus; IOC, C subnucleus; IOK, dorsal cap of Kooy; vlo, ventrolateral outgrowth; IODM, dorsomedial cell column; LC, locus coeruleus; LVe, lateral vestibular nucleus; MVe, medial vestibular nucleus; parafloc, paraflocculus; PrH, prepositus hypoglossus; SpVe, spinal (descending) vestibular nucleus; SuVe, superior vestibular nucleus.

pattern of the respective neurons and their neuromodulatory activity (e.g. peptide and amine transmitters and other messengers) can change the relationship of a particular cell to the network. For example, the timing of synaptic input (bursts versus slow continuous with the same total number of spikes) can be shown to influence Fos recruitment in dorsal root ganglion neurons by way

of tightly controlled temporal calcium networks [28,29]. Many investigators are now using genetically engineered mice to investigate the role of transcription factors in learning and memory [85]. Mice with a Fos gene knockout had impaired spatial and associative learning that could be correlated with reduced hippocampal longterm potentiation [30], although another group concluded

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

203

Fig. 2. Nine schematic binary Fos pattern representations of the bilateral inferior olivary (mainly medial accessory olive), prepositus hypoglossus (PrH), and medial vestibular (MVe) nuclei during nine vestibular-related stimuli or treatments. A filled region indicates significant Fos expression above controls in that brainstem nucleus, although gradation or intensity of the labeling is not shown. Animals tested in the dark are indicated by a shaded background. The afterimage shift at the edge of the IOK is to indicate retinal slip—the direction is arbitrary but consistent across panels (i.e. opposite for low versus high gain VOR adaptation trials). The vestibular inputs primarily active for each stimulus are denoted in the lower left of each box: C, canal; O, otolith. A ( ) indicates loss of input; an (*) denotes the side of a lesion or plug. (A – I) Abbreviations: CH, centripetal hypergravity; CC, cross-coupling; PO, pseudo-off axis vertical rotation; HCP, hemi-canal plugging; LGA, low gain horizontal VOR adaptation; HGA, high gain adaptation; HL, hemilabyrinthectomy; LGAHL, low gain adaptation following chronic hemilabyrinthectomy; KT, systemic ketamine (NMDA receptor noncompetitive antagonist). See text for details.

that hippocampal Fos was not essential for spatial learning [91]. It has also been shown that different members of the Fos gene family control different aspects of long-term potentiation, contextual learning, spatial reference memory formation [32], and chronic drug reactions [66] in mice. Many sensory systems reveal transduction-related Fos expression related to functional roles. Oder-reward learning results in Fos expression patterns that differentiate consolidation from re-consolidation [87]. Fos expression attenuation parallels auditory adaptation in the rat cochlear nuclei [43]. And, the level of Fos expressed in hypothalamic suprachiasmatic neurons controlling circadian rhythm is directly proportional to the number of photons collected by the retina in the gerbil [23]. It is important to consider that brain activity, in a combination of many mechanisms, is a smear in time of

very rapid and more gradual activity [1]. Why has Fos proven useful? The Fos protein plays a role in the intermediate to longer-term (minutes to hours) activities which require changes in gene expression in the cell nucleus. Thus, Fos expression represents an entirely different kind of cellular activity compared to a neuron’s typical role in millisecond input integration measured using methods that detect action potentials. While the Fos response reflects primarily new or increased input [37], in many sensory systems, it is further restricted functionally with unexpected or unusual input. That is, the mature neuron seems to recruit Fos when it receives inputs that are not typical for its synaptic environment and is therefore stimulated to readjust or when a state change becomes necessary. However, Fos expression is only one filter on the subcellular story. Although most neurons appear to be capable of expressing Fos– both excitatory

204

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

and inhibitory barrel cortex neurons expressed Fos after a novel environment [80] – a general background state of stimulation or attention is important. Tonic activity in locus coeruleus neurons might be a necessary prerequisite for Fos expression in many regions, and these neurons also produce Fos themselves in all of the conditions illustrated below. The norepinephrine secreted throughout the brain by these neurons appears to be permissive for the Fos cascade [72]. Serial training of a task or repeated exposure to a new environment decreases the amount of Fos expression in general, although certain regions, like the prepositus hypoglossi, can maintain Fos expression with experience and might be related to ongoing changes [51]. The relative lack of Fos expression in the deep cerebellar nuclei and the lateral vestibular nucleus following vestibular stimulation– both targets of inhibitory Purkinje output –could relate to its role as an excitatory marker, and this polarity should be kept in mind when reading the expression patterns below. That is, the lack of Fos expression might mask other genomic changes initiated by inhibitory inputs. Another result of this potential dependence on circuit polarity is that Fos expression patterns cannot necessarily be associated with vestibular afferent distribution patterns or spike activity if inhibitory inputs overwhelm a particular neuron. There is no clear correlation between afferent distribution and Fos labeling in these studies. While Fos labeling in the ipsilateral periventricular MVe after hemilabyrinthectomy generally occurs in non-second order neuron regions [8], this area has been shown to receive direct primary afferent input in the gerbil [68]. Periventricular MVe Fos labeling occurs during a period when neurons in the region are not firing [67]. On the other hand, Fos in the inferior olive beta subnucleus (IOh) during cross-axis rotation that includes a roll component agrees with spike activity recorded there during roll tilt (reviewed in [3]). Therefore, the precise triggers of Fos expression in individual neurons rely on many variables including the type of receptors present, the pattern of incoming spikes, and their polarity. The presence of complimentary circuit polarity in most brain systems means that the Fos responses outlined below could reflect any combination of excitation or disinhibition. For example, the microcircuitry of the inferior olive consists of glomerular spines with both inhibitory and excitatory inputs and electrotonic coupling [20]. How the inhibitory vestibular input from the parasolitary nucleus [4] and the PrH [18] to the inferior olive affects the local inter-olivary circuitry, for example, is not yet clear. There is nothing particularly special about Fos– other regulatory factors show similar (and unique) results [37] – but a large database has formed around its activity, and Fos has proven to be an interesting marker for transcriptional changes in vestibular-related neurons. There is widespread recognition of the ‘‘essential role of inducible

transcription factors in the reprogramming of cells to a different functional state’’ [75]. In summary, the Fos data have provided useful information concerning: 1. Relative abundance (network magnitude or weight) of the necessary change to effect adaptation, i.e. how many and which neurons are recruited. 2. Symmetry of the network (e.g. following hemilabyrinthectomy). 3. Interrelationships and correlations between areas [78,79]. 4. Unique network Ffootprints_ (amenable to pattern recognition paradigms, this review). Genomic sequence data have now begun to provide more detail about what Fos and other transcription factors might control, based on the identification of DNA sequences with their promoter (the latch which fits their key) upstream to specific genes. The detected genes have broad functional implications. Table 1 describes a recent focused search (and literature review) of the mouse genome for the activator 1 (AP-1) promoter element where Fos binds along with the Jun protein to control transcription. The list in Table 1 was generated using positive probe hybridization targets from a microarray study of the gerbil brainstem after hemilabyrinthectomy. Recent evidence now suggests that Fos can in addition interact with GATA transcription proteins independent of the AP-1 site [62]. Fig. 2 provides nine examples of regional Fos expression patterns in the VOC consistently found in repeated experiments across two or more species during specific types of vestibular labyrinth stimuli. In these schematic panels, the representation of Fos expression is binary, either on or off for the entire nucleus of hundreds or thousands of neurons, which vastly oversimplifies the details of individual cell number and intensity and any relationship of these measures with the specific stimulus parameters. The percentage of Fos-positive neurons in a particular region or nucleus can vary from a fraction of 1% (e.g. in MVe during VOR adaptation) to almost 100% (e.g. in the IODM during hypergravity). Therefore, the threshold for the binary determination of each subnucleus was different across stimulus examples and represents a significance call based on the mean expression levels for that condition alone. For example, in the MVe, hemilabyrinthectomy had the strongest labeling in terms of percentage of neurons labeled, VOR gain adaptation had the least, and the other conditions fell in between. Nonetheless, these overall patterns can provide useful information, as described below. Vestibulocerebellar labeling is not included below due to incomplete data. In the cerebellum, highly differentiated connectivity further complicates Fos interpretation. To at least some degree, vestibulocerebellar granule cell Fos labeling was routinely present in all conditions, while Purkinje cell expression was found to be relatively rare and sensitive to antibody epitopes [78]. Different cerebellar cell types

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

205

Table 1 Describes a list of possible genes influenced by Fos Genbank ID

Title

Gene symbol

Celera (2002)

NCBI (2004)

Transcript

Polarity

M60655 M64780 D49836 AF055477 L27487 AF005720 X53944 X06656 AI137421 J00739 D13418 S45392 U70988 AF081366 S68944 Z11504 M15191

Adrenergic, alpha 1B receptor Agrin Thymoma viral proto-oncogene 3 Calcium channel, voltage-dep, L type, alpha 1D Calcitonin receptor-like receptor Chloride channel 2 Dopamine receptor 3 Gap junction protein, alpha 1 (connexin 43) Glutamate dehydrogenase Gonadotropin-releasing hormone receptor Transcription factor HES-3 Heat shock protein 90 Chemokine (C-X-C) receptor 2 Potassium inwardly-rectifying channel Na+/Cl dependent neurotransmitter transporter NPY-1 receptor Tachykinin (substance P, neurokinin A, neuropeptide K, gamma)

Adra1b Agrn Akt3 Cacna1d Calcrl Clcn2 Drd3 Gja1 Glud1 Gnrhr Hes3 Hspca Il8rb Kcnj1 – Npy1r Tac1

229 191 160 160 580 +20 324 +124 380 280 400 143 +61 200 32 264 400

99 +18 112 – 573 +71 – 66 345 291 376 – +40 – 32 – 353

NM007416 BC064021 NM011785 NM028981 NM018782 NM009900 NM007877 NM010288 NM008133 NM010323 NM008237 NM010480 NM009909 NM019659 NM172271 NM010934 NM009311

minus minus minus minus minus minus plus plus plus minus minus minus plus plus minus plus plus

Genes identified in gerbil brainstem tissue after hemilabyrinthectomy using Affymetrix (rat RN_U34) probes were used as a gene list to search the mouse genome databases for the Fos DNA promoter sequence (activator protein-1, AP-1, site). The approximate base pair position of the AP-1 sequence relative to the gene transcript’s exon starting position is given in the numerical columns (from separate Celera and NCBI data, respectively). From a starting pool of 339 genes, these 17 hits therefore predict an AP-1 promoter element in ¨5% of genes. Note the occurrence of many ligands, receptors, and enzymes in the listing, indicating redundant and multi-layered control. Previous literature supporting yet other Fos-related genes includes links between Fos and annexin II and V, calcitonin gene-related peptide, cholecystikinin, dopamine beta-hydroxylase, Fos-like 1 (Fra-1, Fosl1), galanin, glutamic acid decarboxylase, interleukins, e.g. Il1rl1, myc, neuropeptide Y, nitric oxide synthase, ornithine decarboxylase, prodynorphin, somatostatin, tumor necrosis factor A, and tyrosine kinase receptor B [6,35,36,65,82]. The role of metalloproteinases in shaping synaptic plasticity by modification of neurexins and dystroglycans also appears to involve the AP-1 site [42].

express Fos after different stimuli (harmaline or galvanic) in distinct spatiotemporal patterns [84]. Sustained (static) linear acceleration above 1 G (9.8 m/s/s) was hypothesized to activate otolith-dependent inputs. Fig. 2A demonstrates VOC Fos expression following head-fixed, horizontal, left ear out (i.e. away from the axis of rotation) centripetal acceleration (CH, centripetal hypergravity) in the dark at 2 G (19.6 m/s/s) for 90 min [46]. Intense inferior olivary dorsomedial cell column (IODM) labeling was observed, and there was bilateral vestibular complex labeling, but virtually no labeling in the prepositus hypoglossi (PrH). The vector of fixed linear acceleration across a remaining macula (after chronic vestibular lesion of the other side) could be shown to affect the resulting brainstem Fos expression [47]. Subsequent studies have shown that milder hypergravity conditions can cause IODM Fos expression, as well as certain labyrinth applications of galvanic current. Depending on the placement, polarity, intensity, and duration of the applied current, unique Fos expression patterns were induced [44]. For example, electrode placement near the posterior canal ampulla resulted in the induction of IODM labeling, physiologically observed only under hypergravity conditions [61]. Finally, results from orbital flight (microgravity) are becoming available [16], which reveal complex expression patterns in the rat linked to the dynamically changing gravity environment of space flight. In another study which exposed 1 G and 2 G hypergravity-raised animals to the contrary condition [34], both directions of change elicited widespread cortical (suprabulbar) Fos expression, but only a change from

1 G to higher G elicited strong brainstem Fos labeling, in patterns similar to other centrifuge studies. This result might relate to the nature of Fos as a marker for positive transduction stimuli only. Another example of novel vestibular input occurs during prolonged rotations around mixed axes. Fig. 2B reveals the VOC Fos expression following rotational cross-coupling (CC) consisting of a sinusoidal roll (e.g. 0.5 Hz, 40 d/s roll) and simultaneous leftward earth horizontal rotation (200-/s constant velocity rotation) for 60 min. Bilateral Fos expression could be observed in the inferior olive beta subnucleus (IOh) and ventrolateral outgrowth (vlo) and the spinal vestibular nuclei (not shown), while the PrH expressed Fos on the left side, and the C subnucleus (IOC) expressed Fos on the right side. Novel patterns of canal input recruit specific vestibular networks, and these patterns could be shown to be dependent on rotational direction, total input symmetry, frequency and magnitude of the stimulus, and the presence of additional stimuli, like linear acceleration [51]. A dynamic otolith-only stimulus occurs when constant velocity rotation is maintained about a tilted or off-center axis. Fig. 2C represents Fos expression following various levels of eccentric horizontal counter-rotation, or pseudo-off axis vertical rotation, (pseudo-OVAR, PO) where a sweeping (slightly tilted from vertical) gravito-inertial vector is rotating through the animal’s head in a clockwise fashion. At moderate velocities and frequency (~140-/s or 0.4 Hz), the vlo and IOC revealed Fos on one side, while the PrH

206

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

expressed Fos on the other. One study in the rat of the simpler (tilted) off-vertical axis rotation (OVAR), which also generates a sweeping otolith vector, resulted in strong vestibular Fos expression dependent on intact labyrinths and showed a high percentage of presumably otolith-receiving neurons with Fos and NMDA receptor double-labeling [9]; Fos expression has been linked to NMDA receptor activation [90]. However, the OVAR stimulus was very fast (1.5 Hz, 540-/s, 10- tilt for 90 min), and there was no mention of vestibular nuclei asymmetry or inferior olivary labeling. It is likely that the frequency of the rotating vector has an effect on the intensity and the symmetry of the Fos response in the brainstem [51]. Fig. 2D summarizes VOC Fos expression following canal plugging of all three semicircular canals on one side (HCP, hemi-canal plugging). The pattern was similar to that following hemilabyrinthectomy (Fig. 2G) but more sparse and with no inferior olivary expression. The shifted edge of the IOK nucleus in the figure represents retinal slip information—with reduced canal input from one side, poor vestibulo-ocular reflex (VOR) performance would simulate a high gain visual adaptation environment. Fos expression results can separate dynamic from static oculomotor compensation to unilateral lesion, mainly by the lack of labeling in the inferior olive IOK in the latter [69]. Fig. 2E illustrates VOC Fos expression following low gain VOR adaptation (LGA) in the horizontal plane (e.g. 1 h of 0.5 Hz, 60-/s sinusoidal rotation with an optokinetic spotted drum moving with the animal but at half the amplitude). Bilateral IOK and weak vestibular/prepositus labeling was observed. Floccular Purkinje cells also labeled sparsely [79]. Fig. 2F shows the resulting pattern when VOR gain was driven higher by reversing the optokinetic visual environment at half the velocity of head motion (HGA, high gain adaptation). The IOK failed to express Fos in this situation that was otherwise identical to 2E [79]. In Fig. 2G, the well-characterized VOC Fos expression following unilateral surgical lesion (HL, hemilabyrinthectomy) is shown [10,45,52,54]. This condition is a powerful stimulus for adaptation as the subject is virtually unable to move until recovery occurs (called vestibular compensation), and this lesion strikingly illustrates the bilateral symmetry of the vestibular network via the Fos expression. Contralateral expression was seen in the PrH and inferior olivary IOK, IOh, and IODM, while ipsilateral expression predominates in the vestibular complex, where neurons are known to be silent as a result of lack of input and commissural inhibition early in the process (see [69] for review). A ventral tier of contralateral parafloccular Purkinje cells also labels in the gerbil [78]. Note that this panel corresponds to the more detailed illustration in Fig. 1. Recovery from unilateral vestibular lesion is remarkably similar across species but exhibits different time scales for static and dynamic reflexes [69]. Many well-characterized

neurotransmitter systems have been shown to change after vestibular lesion [15]. There is widespread activity-dependent reorganization of the excitatory and inhibitory network [21]. The commissural system between the vestibular nuclei in mammals is predominantly inhibitory and is clearly an important part of the solution to rebalancing vestibular tone bilaterally. Plasticity in the commissural system [26] could act to restore the lesion-induced imbalance across the midline of the MVe [14,67], and many Fos expressing MVe neurons after lesion were commissural [78] or flocculus-projecting (unpublished data). Both ionotropic GABAA and metabotropic GABAB receptors decrease functionally on the ipsilateral side of the MVe post-lesion, but GABAB receptors remain low indefinitely [41]. However, plasticity across the commissures is likely just part of the recovery mechanism. Impairing or damaging one early response (e.g. Fos expression) in any one of the regions of the vestibulo-olivary-cerebellar (VOC) pathways during vestibular compensation produces a more distributed response-manifested as an alteration or increase in Fos expression throughout the rest of the VOC [48,56]. This distributed plasticity [60] can explain why commissural lesions sometimes fail to produce predicted effects [80]. Interregional correlations in vestibular brainstem Fos labeling were increased during compensation in the dark compared to a normal visual environment [78]. Visual feedback can stabilize synaptic plasticity [92], and without vision, the brain may require more motor repetition and feedback from other systems to generate a new stable network. Fig. 2H demonstrates VOC Fos expression following low gain VOR adaptation identical to Fig. 2E, but in an animal that has compensated 4 weeks or more from a previous hemilabyrinthectomy (LGAHL, low gain adaptation after hemilabyrinthectomy). Only contralateral IOK, PrH, and MVe Fos expression was seen. The pattern symmetry is clearly different than that observed after acute labyrinthectomy alone [79]. The final panel (Fig. 2I) is not a motion stimulus but rather shows how pharmacological manipulation can also be used to probe the system. Fig. 2I represents immunolabeling observed after an anesthetic dose (e.g. 20 mg/kg) of intramuscular ketamine (KT), an NMDA receptor noncompetitive antagonist. Diffuse to heavy immunolabeling was seen in the IOh and vestibular/prepositus complex. This pattern of expression can be used to infer certain relationships between regions, as well as how the NMDA receptor distribution in the system is played out functionally. For example, there are strong GABAergic vestibulo-olivary connections to the IOh [3,4], implying that a glutaminergic disinhibition releases IOh neurons under ketamine. Other investigations using similar NMDA antagonists have helped to define the system. MK-801 results in decompensation and HL patterns opposite in symmetry to those observed after the initial lesion at early time points [11,55,57], as does PCP [64].

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

The VOC system is also sensitive to non-vestibular environmental variables. Scattered Fos expression was observed in the MVe, for example, after simply moving an animal from the colony facility to the laboratory in the same building. This expression can be significant (labeled neuronal counts of 30 or more per tissue section) and gradually goes away over a period of approximately 2 days. Such expression is likely related to spatial challenges shown to be represented in hippocampal neurons [81]. An animal sacrificed after three or more days of laboratory acclimatization does not reveal this expression. Growing evidence in mutant mouse strains shows that cage size, environmental cues like light and complexity, isolation, and the presence of human observers (i.e. big predators) can have significant effects on behavior [31]. Merely walking on a treadmill can activate Fos in the IOh in rats [74]. Table 2 is a crude attempt to build a digital meta-analysis of the Fos patterns observed in past studies. On or off Fos expression for an entire nuclear or subnuclear group is

207

represented in binary form (1 = expression, 0 = none) following the patterns in Fig. 2, while the basic transducer activity component (end-organ or optokinetic input) during the applied stimulus is coded in trinary format (0 indicates normal or mild inputs, 1 indicates lack of input, and +1 indicates excessive or novel input). Principal component analysis (PCA) – a method of linearly grouping total variability onto three axes to find trends– of the data in Table 2 clusters known functional links between transduction sources and the resulting brainstem Fos expression, as modeled here (Fig. 3). The individual treatment values (eigenvectors) from this analysis (Table 2) show that principal component 1 (PC 1) most strongly identifies the lesion conditions (HL, LGAHL), PC 2 the hypergravity and cross-coupling conditions (CH, CC), and PC 3 primarily the conditions where optokinetic slip was present. In other words, PC 1 identifies the lesion conditions and correlates strongly with MVe and PrH activity. PC 2 isolates otolith activity with the dorsomedial cell column (IODM) of the

Table 2 Binary representation of Fos expression in each nucleus and condition from Fig. 2, with a trinary representation of the corresponding end-organ and optokinetic stimulus factors that would have been present in each of the nine treatment conditions

HC, horizontal canal; PC, posterior canal; AC, anterior canal; U, utricle; S, saccule; OKR, optokinetic flow; i, ipsi (left in Fig. 2); c, contra (right in Fig. 2). For stimuli, a 0 indicates normal or mild inputs, 1 indicates lack of input, and +1 indicates excessive or novel input. In the optokinetic column, the sign refers to the relative retinal slip direction. Principal component analysis (PCA) values on all data shown are displayed below each column. PCA isolates lesional, otolith, and optokinetic sources of variance (bold numbers) based on this analysis and represented 71.5% of the total variability with three dimensions.

208

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

(2) Lesion (loss of normal input) (a) Hemilabyrinthectomy (canal, otolith, and afferent activity loss) (b) Canal plugging (only dynamic low frequency canal loss) (3) Pharmacologic manipulation (e.g. ketamine)

Fig. 3. Principal component analysis (PCA) of Fos expression variation in the seven nuclei described in Fig. 2 (IOC, IOK, IOh, vlo, IODM, PrH, and MVe), based on the model data in Table 2. While the end-organ and optokinetic stimulus values in Table 2 influence the PCA values of binary Fos expression in each brainstem nucleus (the box in Table 2), the stimulus PCA values are not shown in this figure. Principal component 1 (PC 1) isolates most strongly the nuclei involved in vestibular lesion, PC 2 appears to be related to otolith inputs, and PC 3 has indications of an optokinetic influence.

inferior olive. And, PC 3 partially highlights retinal slip inputs via dorsal cap (IOK) labeling, an area of the inferior olive known to receive retinal information [63]. There was also a significant correlation (Spearman’s rho, P  0.052) among binary Fos expression representation in the IOC, IOh, and vlo, and between the PrH, IODM, and macular (utricular and saccular) stimulus sources. While this analysis is somewhat contrived, by backing away from a cellular to a nuclear regional pattern view, the exercise reveals relationships that fit well with what is known about the functional divisions of the VOC. 2. Discussion Clearly, many different kinds of stimuli elicit a Fos response in the vestibular system. The stimuli described here can be categorized by the following classification schema: (1) Physiologic (a) Novel combinations (multiple sensory error signals) (i) Low gain VOR adaptation (ii) High gain VOR adaptation (iii) Cross-coupling (iv) Off-vertical axis rotation (v) New spatial environment (b) Novel levels or patterns (i) Hypergravity (centripetal acceleration) (ii) Galvanic labyrinth stimulation (direct current)

Fos data can be used to probe the organization and mechanisms of the vestibular system. For example, how do separate sensory modalities interact? Significant differences in Fos expression patterns can depend on the available input modalities. The Fos patterns resulting from the applied motion of the conditions in Figs. 2A, B, and C took place in complete darkness. The remaining conditions illustrate expression occurring during a normal or modified visual surround. As discussed above, the presence of visual input during the period of compensation to vestibular lesion can significantly shift the patterns and amount of Fos expression in these regions following hemilabyrinthectomy [78]. Focusing on vestibular signals alone, specific vectors, frequency, and magnitude of motion influence the input topology of the network and the resulting Fos expression [47,51]. And, Fos expression, along with behavioral measures, can help to illustrate interactions between otolith and canal inputs [51,69]. There are growing indications that canal-otolith convergence is necessarily fused at certain levels of the network, as described for visual cues [39]. The brain appears to require both canal and otolith inputs in its solution to balancing the bilateral network during vestibular compensation. The observation that compensation to unilateral canal plugging alone can actually protect otolithocular responses briefly after a subsequent complete hemilabyrinthectomy [69] is evidence for the necessary fusion of canal and otolith representation centrally. Thirty-eight percent of neurons in the MVe and PrH region were found to exhibit convergent otolith and canal responses in the gerbil [49]. During compensation, the system must adjust canal and otolith function in a fashion that best serves global functional recovery but might attenuate specific reflexes. How many neurons are taking part in the process, or how much activity is necessary? The number of Fos immunolabeled neurons during vestibular compensation (hundreds per section in the response to unilateral lesion, Fig. 2G) is much greater than the relatively few cells (tens) that respond with Fos to a VOR gain adaptation (Figs. 2E, F [79]) or a canal plugging (Fig. 2D) paradigm. Principal component 1 (PC 1) of the 3-dimension analysis included the greatest percentage of variability in the model (40%) and reflected the lesion conditions, correlating highly with the PrH/MVe complex, a region with many cells and strong commissural inputs important for restoring bilateral balance. In which situation is subjective vertical likely to be challenged? It is now clear that the brain uses the gravity vector for egocentric orientation and perceptual transformations that rely on multiple sensory inputs [13,50,71]. The

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

omnipresent signal of gravity is perceptually changed (the subject ‘‘feels tilted’’) during centripetal hypergravity and following unilateral lesion (Figs. 2A, G). The IODM appears to play a role in this calibration. The IODM’s projection to the nodulus (lobule X) of the posterior cerebellum is consistent with functional studies that show this region affecting gravity sensing [3]. IODM appears to be a bilateral nexus for inferior olivary connectivity, collecting widespread olivary afferent connections and then projecting to discrete zones in the cerebellar vermis [19]. Could the IODM be a spatial orientation pathway based on gravity [13]? PC 2 revealed greater changes in the otolith organ elements and the IODM. In which situation is the VOR adjusted with visual input? (Figs. 2D –H) The inferior olive dorsal cap (IOK) plays a role in low gain horizontal adaptation, but apparently not, as defined by Fos, in high gain adaptation or canal plugging. PC 3 had power in the conditions where optokinetic input was a factor and showed some distinction in the dorsal cap (IOK). Our data demonstrate a neural network with functional adaptation or homeostasis as a goal. Just as subsets of neurons become synaptically related during experience with specific stimuli in cortical population coding, e.g. ‘‘across neuron response patterns’’ [24], but can also contribute to the combined representation of other modes or characteristics of stimuli, these data show that similar redundancy can exist in the brainstem VOC. That is, a particular neuronal population can be active (as defined by Fos) for several different adaptive stimuli, but the overall network pattern in which that population takes part might be different for each stimulus. The specific role of the Fos expression in each condition might also be different in the same neurons [32,66]. One theoretical framework predicts that vestibular compensation and motor learning share limited mechanisms among individual cells, but also likely differ, within cells and between pathways. At the vestibular nuclei and downstream, the combination of optokinetic and vestibular (both rotational and linear) signals generate behavior that can be explained by attributes of population coding, whereby synaptic input from convergent sources contributes to vestibular tone or drive, but with the output heavily eighted by gravitational and rotational inertia. Accessory optic system and vestibular signals, along with reticular/raphe and autonomic inputs, converge on the vestibular/prepositus nuclei. The network is characterized by distributed plasticity, partially overlapping circuitry, adaptability, and redundancy. A highly simplified but useful analogy is to water pressure: each sensory modality contributes (functional, synaptic) pressure, and system inputs have a threshold. All neurons can adapt. It is their phenotype and location in the transduction network architecture that determines their shaping. The binary on and off representation of Fos expression in Fig. 2 neglects the subtle variation of both individual cell expression intensity and the percentage of

209

total cells in a given nucleus during a particular experiment. Seen in this way, however, the global patterns of activity can be differentiated. How can these patterns be reconciled based on the known anatomy? It is tempting to try and formulate a mathematical model that would take into account the known excitatory and inhibitory connections and general input and output behavior of these regions under each of the input conditions. It soon becomes apparent that the complexity of even this simple group of nuclei remains beyond a rigorous definition without acquiring much greater detail. Even if general electrophysiological classes of behavior among excitable cells in these regions can be summarized, the type of activity represented by transcription factors such as Fos quickly interdigitates with hundreds of metabolic and genomic pathways opening thousands of possible states. In the face of this richness, we must let the system first teach us about its behavior using simple probes.

References [1] L.F. Abbott, S.B. Nelson, Synaptic plasticity: taming the beast, Nat. Neurosci. 3 (2000) 1178 – 1183 (Suppl.). [2] D.E. Angelaki, Eyes on target: what neurons must do for the vestibuloocular reflex during linear motion, J. Neurophysiol. 92 (2004) 20 – 35. [3] N.H. Barmack, Central vestibular system: vestibular nuclei and posterior cerebellum, Brain Res. Bull. 60 (2003) 511 – 541. [4] N.H. Barmack, B.J. Fredette, E. Mugnaini, Parasolitary nucleus: a source of GABAergic vestibular information to the inferior olive of rat and rabbit, J. Comp. Neurol. 392 (1998) 352 – 372. [5] E.S. Boyden, A. Katoh, J.L. Raymond, Cerebellum-dependent learning: the role of multiple plasticity mechanisms, Annu. Rev. Neurosci. 27 (2004) 581 – 609. [6] S. Braselmann, G. Bergers, C. Wrighton, P. Graninger, G. Supertifurga, M. Busslinger, Identification of fos target genes by the use of selective induction systems, J. Cell Sci. 16 (1992) 97 – 109. [7] D.M. Broussard, C.D. Kassardjian, Learning in a simple motor system, Learn Mem. 11 (2004) 127 – 136. [8] J.A. Buttner-Ennever, Chapter I: overview of the vestibular system: anatomy, in: A.J. Beitz, J.H. Anderson (Eds.), Neurochemistry of the Vestibular System, vol. 1, CRC Press, 2000, pp. 3 – 24. [9] L.W. Chen, C.H. Lai, H.Y. Law, K.K. Yung, Y.S. Chan, Quantitative study of the coexpression of Fos and N-methyl-d aspartate (NMDA) receptor subunits in otolith-related vestibular nuclear neurons of rats, J. Comp. Neurol. 460 (2003) 292 – 301. [10] C. Cirelli, M. Pompeiano, P. D’Ascanio, O. Pompeiano, Early c-fos expression in the rat vestibular and olivocerebellar systems after unilateral labyrinthectomy, Arch. Ital. Biol. 131 (1993) 71 – 74. [11] C. Cirelli, M. Pompeiano, P. D’Ascanio, P. Arrighi, O. Pompeiano, cfos expression in the rat brain after unilateral labyrinthectomy and its relation to the uncompensated and compensated stages, Neuroscience 70 (1996) 515 – 546. [12] G. Clement, J.H. Courjon, M. Jeannerod, R. Schmid, Unidirectional habituation of vestibulo-ocular responses by repeated rotational or optokinetic stimulations in the cat, Exp. Brain Res. 42 (1981) 34 – 42. [13] B. Cohen, J. Maruta, T. Raphan, Orientation of the eyes to gravitoinertial acceleration, Ann. N. Y. Acad. Sci. 942 (2001) 241 – 258.

210

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211

[14] A.C. Cova, H.L. Galiana, A bilateral model integrating vergence and the vestibulo-ocular reflex, Exp. Brain Res. 107 (1996) 435 – 452. [15] C.L. Darlington, H. Flohr, P.F. Smith, Molecular mechanisms of brainstem plasticity—The vestibular compensation model, Mol. Neurobiol. 5 (1991) 355 – 368. [16] P. d’Ascanio, E. Balaban, M. Pompeiano, C. Centini, O. Pompeiano, Fos and FRA protein expression in rat precerebellar structures during the Neurolab Space Mission, Brain Res. Bull. 62 (2003) 203 – 221. [17] A. Deutschlander, S. Bense, T. Stephan, M. Schwaiger, T. Brandt, M. Dieterich, Sensory system interactions during simultaneous vestibular and visual stimulation in PET, Hum. Brain Mapp. 16 (2002) 92 – 103. [18] C.I. DeZeeuw, P. Wentzel, E. Mugnaini, Fine structure of the dorsal cap of the inferior olive and its GABAergic and non-GABAergic input from the nucleus prepositus hypoglossi in rat and rabbit, J. Comp. Neurol. 327 (1993) 63 – 82. [19] C.I. DeZeeuw, E.J. Lang, I. Sugihara, T.J.H. Ruigrok, L.M. Eisenman, E. Mugnaini, R. Llinas, Morphological correlates of bilateral synchrony in the rat cerebellar cortex, J. Neurosci. 16 (1996) 3412 – 3426. [20] C.I. De Zeeuw, J.I. Simpson, C.C. Hoogenraad, N. Galjart, S.K. Koekkoek, T.J. Ruigrok, Microcircuitry and function of the inferior olive, Trends Neurosci. 21 (1998) 391 – 400. [21] N. Dieringer, H. Straka, Steps toward recovery of function after hemilabyrinthectomy in frogs, Otolaryngol.-Head Neck Surg. 119 (1998) 27 – 33. [22] M. Dieterich, T. Brandt, Vestibular system: anatomy and functional magnetic resonance imaging, Neuroimaging Clin. N. Am. 11 (2001) 263 – 273 (iX). [23] O. Dkhissi-Benyahya, B. Sicard, H. Cooper, Effects of irradiance and stimulus duration on early gene expression (Fos) in the suprachiasmatic nucleus: temporal summation and reciprocity, J. Neurosci. 20 (2000) 7790 – 7797. [24] G.S. Doetsch, Patterns in the brain. Neuronal population coding in the somatosensory system, Physiol. Behav. 69 (2000) 187 – 201. [25] S. du Lac, J.L. Raymond, T.J. Sejnowski, S.G. Lisberger, Learning and memory in the vestibulo-ocular reflex, Annu. Rev. Neurosci. 18 (1995) 409 – 441. [26] K. Farrow, D.M. Broussard, Commissural inputs to secondary vestibular neurons in alert cats after canal plugs, J. Neurophysiol. 89 (2003) 3351 – 3353. [27] B.M. Faulstich, K.A. Onori, S. du Lac, Comparison of plasticity and development of mouse optokinetic and vestibulo-ocular reflexes suggests differential gain control mechanisms, Vision Res. 44 (2004) 3419 – 3427. [28] R.D. Fields, P.G. Nelson, Resonant activation of calcium signal transduction in neurons, J. Neurobiol. 25 (1994) 281 – 293 (Review). [29] R.D. Fields, P.R. Lee, J.E. Cohen, Temporal integration of intracellular Ca(2+) signaling networks in regulating gene expression by action potentials, Cell Calcium 37 (2005) 433 – 442. [30] A. Fleischmann, O. Hvalby, V. Jensen, T. Strekalova, C. Zacher, L.E. Layer, A. Kvello, M. Reschke, R. Spanagel, R. Sprengel, E.F. Wagner, P. Gass, Impaired long-term memory and NR2A-type NMDA receptor-dependent synaptic plasticity in mice lacking c-Fos in the CNS, J. Neurosci. 23 (2003) 9116 – 9122. [31] J.P. Garner, G.J. Mason, Evidence for a relationship between cage stereotypes and behavioural disinhibition in laboratory rodents, Behav. Brain Res. 136 (2002) 83 – 92. [32] P. Gass, A. Fleischmann, O. Hvalby, V. Jensen, C. Zacher, T. Strekalova, A. Kvello, E.F. Wagner, R. Sprengel, Mice with a fra-1 knock-in into the c-fos locus show impaired spatial but regular contextual learning and normal LTP, Brain Res. Mol. Brain Res. 130 (2004) 16 – 22. [33] A.M. Green, D.E. Angelaki, An integrative neural network for detecting inertial motion and head orientation, J. Neurophysiol. 92 (2004) 905 – 925.

[34] S. Gustave Dit Duflo, C. Gestreau, M. Lacour, Fos expression in the rat brain after exposure to gravito-inertial force changes, Brain Res. 861 (2000) 333 – 344. [35] N. Hay, M. Takimoto, J.M. Bishop, A FOS protein is present in a complex that binds a negative regulator of MYC, Genes Dev. 3 (1989) 293 – 303. [36] T. Herdegen, Jun, Fos, and CREB/ATF transcription factors in the brain: control of gene expression under normal and pathophysiological conditions, The Neuroscientist 2 (1996) 153 – 161. [37] T. Herdegen, J.D. Leah, Inducible and constitutive transcription factors in the mammalian nervous system: control of gene expression by Jun, Fos and Krox, and CREB/ATF proteins, Brain Res. Brain Res. Rev. 28 (1998) 370 – 490. [38] S.M. Highstein, Role of the flocculus of the cerebellum in motor learning of the vestibulo-ocular reflex, Otolaryngol.-Head Neck Surg. 119 (1998) 212 – 220. [39] J.M. Hillis, M.O. Ernst, M.S. Banks, M.S. Landy, Combining sensory information: mandatory fusion within, but not between, senses, Science 298 (2002) 1627 – 1630. [40] U.J. Ilg, Slow eye movements, Prog. Neurobiol. 53 (1997) 293 – 329. [41] A.R. Johnston, A. Him, M.B. Dutia, Differential regulation of GABA(A) and GABA(B) receptors during vestibular compensation, NeuroReport 12 (2001) 597 – 600. [42] L. Kaczmarek, J. Lapinska-Dzwonek, S. Szymczak, Matrix metalloproteinases in the adult brain physiology: a link between c-Fos, AP1 and remodeling of neuronal connections? EMBO J. 21 (2002) 6643 – 6648. [43] A. Kandiel, S. Chen, D.E. Hillman, c-fos gene expression parallels auditory adaptation in the adult rat, Brain Res. 839 (1999) 292 – 297. [44] G.D. Kaufman, A.A. Perachio, Translabyrinth electrical stimulation for the induction of immediate-early genes in the gerbil brainstem, Brain Res. 646 (1994) 345 – 350. [45] G.D. Kaufman, J.H. Anderson, A.J. Beitz, Brainstem Fos expression following acute unilateral labyrinthectomy in the rat, NeuroReport 3 (1992) 829 – 832. [46] G.D. Kaufman, J.H. Anderson, A.J. Beitz, Fos-defined activity in rat brainstem following centripetal acceleration, J. Neurosci. 12 (1992) 4489 – 4500. [47] G.D. Kaufman, J.H. Anderson, A.J. Beitz, Otolith-brain stem connectivity: evidence for differential neural activation by vestibular hair cells based on quantification of FOS expression in unilateral labyrinthectomized rats, J. Neurophysiol. 70 (1993) 117 – 127. [48] G.D. Kaufman, M.E. Shinder, A.A. Perachio, Correlation of Fos expression and circling asymmetry during gerbil vestibular compensation, Brain Res. 817 (1999) 246 – 255. [49] G.D. Kaufman, M.E. Shinder, A.A. Perachio, Convergent properties of vestibular-related brain stem neurons in the gerbil, J. Neurophysiol. 83 (2000) 1958 – 1971. [50] G.D. Kaufman, S.J. Wood, C.C. Gianna, F.O. Black, W.H. Paloski, Spatial orientation and balance control changes induced by altered gravitoinertial force vectors, Exp. Brain Res. 137 (2001) 397 – 410. [51] G.D. Kaufman, T. Weng, T. Ruttley, VOR adaptation and CNS Fos expression following centripetal Coriolis stimuli in the gerbil, J. Vestib. Res. (in press). [52] M.S. Kim, B.K. Jin, S.W. Chun, M.Y. Lee, S.H. Lee, J.H. Kim, B.R. Park, Effect of MK801 on cFos-like protein expression in the medial vestibular nucleus at early stage of vestibular compensation in uvulonodullectomized rats, Neurosci. Lett. 231 (1997) 147 – 150. [53] W.M. King, W. Zhou, R.D. Tomlinson, K.M. McConville, W.K. Page, G.D. Paige, J.S. Maxwell, Eye position signals in the abducens and oculomotor nuclei of monkeys during ocular convergence, J. Vestib. Res. 4 (1994) 401 – 408. [54] T. Kitahara, T. Saika, N. Takeda, H. Kiyama, T. Kubo, Changes in Fos and Jun expression in the rat brainstem in the process of vestibular compensation, Acta Oto-Laryngol. (1995) 401 – 404.

G.D. Kaufman / Brain Research Reviews 50 (2005) 200 – 211 [55] T. Kitahara, N. Takeda, T. Saika, T. Kubo, H. Kiyama, Effects of Mk801 on Fos expression in the rat brainstem after unilateral labyrinthectomy, Brain Res. 700 (1995) 182 – 190. [56] T. Kitahara, M. Fukushima, N. Takeda, T. Saika, T. Kubo, Effects of pre-flocculectomy on Fos expression and NMDA receptor-mediated neural circuits in the central vestibular system after unilateral labyrinthectomy, Acta Oto-laryngol. 120 (2000) 866 – 871. [57] T. Kitahara, A. Nakagawa, M. Fukushima, A. Horii, N. Takeda, T. Kubo, Changes in fos expression in the rat brainstem after bilateral labyrinthectomy, Acta Oto-laryngol. 122 (2002) 620 – 626. [58] S.G. Lisberger, F.A. Miles, L.M. Optican, B.B. Eighmy, Optokinetic response in monkey: underlying mechanisms and their sensitivity to long-term adaptive changes in vestibuloocular reflex, J. Neurophysiol. 45 (1981) 869 – 890. [59] S.G. Lisberger, T.A. Pavelko, D.M. Broussard, Neural basis for motor learning in the vestibuloocular reflex of primates: I. Changes in the response of brain stem neurons, J. Neurophysiol. 72 (1994) 928 – 953. [60] R.R. Llinas, K. Walton, Vestibular compensation: a distributed property of the central nervous system, in: H. Asanuma (Ed.), Integration in the Nervous System, Igaku-Shon, Tokyo, 1979, pp. 145 – 166. [61] T.H. Marshburn, G.D. Kaufman, I.M. Purcell, A.A. Perachio, Saccule contribution to immediate early gene induction in the gerbil brainstem with posterior canal galvanic or hypergravity stimulation, Brain Res. 761 (1997) 51 – 58. [62] K. McBride, F. Charron, C. Lefebvre, M. Nemer, Interaction with GATA transcription factors provides a mechanism for cell-specific effects of c-Fos, Oncogene 22 (2003) 8403 – 8412. [63] M.J. Mustari, A.F. Fuchs, C.R. Kaneko, F.R. Robinson, Anatomical connections of the primate pretectal nucleus of the optic tract, J. Comp. Neurol. 349 (1994) 111 – 128. [64] R. Nakki, F.R. Sharp, S.M. Sagar, Fos expression in the brainstem and cerebellum following phencyclidine and Mk801, J. Neurosci. Res. 43 (1996) 203 – 212. [65] J.R. Naranjo, B. Mellstrom, M. Achaval, P. Sassone-Corsi, Molecular pathways of pain: Fos/Jun-mediated activation of a noncanonical AP-1 site in the prodynorphin gene, Neuron 6 (1991) 607 – 617. [66] E.J. Nestler, M.B. Kelz, J. Chen, DeltaFosB: a molecular mediator of long-term neural and behavioral plasticity, Brain Res. 835 (1999) 10 – 17. [67] S.D. Newlands, A.A. Perachio, Compensation of horizontal canal related activity in the medial vestibular nucleus following unilateral labyrinth ablation in the decerebrate gerbil: I. Type I neurons, Exp. Brain Res. 82 (1990) 359 – 372. [68] S.D. Newlands, J.T. Vrabec, I.M. Purcell, C.M. Stewart, B.E. Zimmerman, A.A. Perachio, Central projections of the saccular and utricular nerves in macaques, J. Comp. Neurol. 466 (2003) 31 – 47. [69] S.D. Newlands, S. Dara, G.D. Kaufman, Relationship of static and dynamic mechanisms in VOR compensation, The Laryngoscope 115 (2004) 191 – 204. [70] A.E. Petropoulos, C.r. Wall, C.M. Oman, Yaw sensory rearrangement alters pitch vestibulo-ocular reflex responses, Acta Oto-laryngol. 117 (1997) 647 – 656. [71] V.E. Pettorossi, P. Errico, A. Ferraresi, N.H. Barmack, Optokinetic and vestibular stimulation determines the spatial orientation of negative optokinetic afternystagmus in the rabbit, J. Neurosci. 19 (1999) 1524 – 1531. [72] O. Pompeiano, Noradrenergic control of cerebello-vestibular functions: modulation, adaptation, compensation, Neurosci. Mol. Cogn. 100 (1994) 105 – 114.

211

[73] T. Raphan, B. Cohen, The vestibulo-ocular reflex in three dimensions, Exp. Brain Res. 145 (2002) 1 – 27. [74] T.J.H. Ruigrok, H.v.d. Burg, E. Sabel-Goedknegt, Locomotion coincides with c-Fos expression in related areas of inferior olive and cerebellar nuclei in the rat, Neurosci. Lett. 214 (1996) 119 – 122. [75] K.H. Schlingensiepen, F. Wollnik, M. Kunst, R. Schlingensiepen, T. Herdegen, W. Brysch, The role of Jun transcription factor expression and phosphorylation in neuronal differentiation, neuronal cell death, and plastic adaptations in vivo, Cell. Mol. Neurobiol. 14 (1994) 487 – 505. [76] C.A. Scudder, A.F. Fuchs, Physiological and behavioral identification of vestibular nucleus neurons mediating the horizontal vestibuloocular reflex in trained rhesus monkeys, J. Neurophysiol. 68 (1992) 244 – 264. [77] M. Shelhamer, D.A. Robinson, H.S. Tan, Context-specific adaptation of the gain of the vestibulo-ocular reflex in humans, J. Vestib. Res. 2 (1992) 89 – 96. [78] M.E. Shinder, A.A. Perachio, G.D. Kaufman, VOR and Fos response during acute vestibular compensation in the gerbil in darkness and in light, Brain Res. 1038 (2005) 183 – 197. [79] M.E. Shinder, A.A. Perachio, G.D. Kaufman, Adaptation of the gerbil VOR and the Fos response, Brain Res. (in press). [80] P.F. Smith, C.L. Darlington, I.S. Curthoys, Vestibular compensation without brainstem commissures in the guinea pig, Neurosci. Lett. 65 (1986) 209 – 213. [81] R.W. Stackman, A.S. Clark, J.S. Taube, Hippocampal spatial representations require vestibular input, Hippocampus 12 (2002) 291 – 303. [82] D.J. Swanson, E. Zellmer, E.J. Lewis, AP1 proteins mediate the cAMP response of the dopamine beta-hydroxylase gene, J. Biol. Chem. 273 (1998) 24065 – 24074. [83] F. Tempia, N. Dieringer, P. Strata, Adaptation and habituation of the vestibulo-ocular reflex in intact and inferior olive-lesioned rats, Exp. Brain Res. 86 (1991) 568 – 578. [84] J.B. Tian, G.A. Bishop, Stimulus-dependent activation of c-Fos in neurons and glia in the rat cerebellum, J. Chem. Neuroanat. 23 (2002) 157 – 170. [85] W. Tischmeyer, R. Grimm, Activation of immediate early genes and memory formation, Cell. Mol. Life Sci. 55 (1999) 564 – 574. [86] P. Trillenberg, M. Shelhamer, D.C. Roberts, D.S. Zee, Cross-axis adaptation of torsional components in the yaw-axis vestibulo-ocular reflex, Exp. Brain Res. 148 (2003) 158 – 165. [87] S. Tronel, S.J. Sara, Mapping of olfactory memory circuits: regionspecific c-fos activation after odor-reward associative learning or after its retrieval, Learn Mem. 9 (2002) 105 – 111. [88] P. Vidal, C. DeWaele, P. Baudonniere, J. Lepecq, P.T. Ba Huy, Vestibular projections in the human cortex, Ann. N. Y. Acad. Sci. 871 (1999) 455 – 457. [89] J. Voogd, N. Gerrits, T. Ruigrok (Eds.), Organization of the Vestibulocerebellum, vol. 781, The New York Academy of Sciences, New York, 1996, 739 pp. [90] Z.G. Xia, H. Dudek, C.K. Miranti, M.E. Greenberg, Calcium influx via the Nmda receptor induces immediate early gene transcription by a map kinase/erk-dependent mechanism, J. Neurosci. 16 (1996) 5425 – 5436. [91] J. Zhang, J.M. McQuade, C.V. Vorhees, M. Xu, Hippocampal expression of c-fos is not essential for spatial learning, Synapse 46 (2002) 91 – 99. [92] Q. Zhou, H.W. Tao, M.M. Poo, Reversal and stabilization of synaptic modifications in a developing visual system, Science 300 (2003) 1953 – 1957.