to image brain plasticity in song birds

NMR IN BIOMEDICINE NMR Biomed. 2004;17:602–612 Published online in Wiley InterScience (www.interscience.wiley.com). DOI:10.1002/nbm.936

Applications of manganese-enhanced magnetic resonance imaging (MEMRI) to image brain plasticity in song birds Annemie Van der Linden,1* Vincent Van Meir,1 Ilse Tindemans,1 Marleen Verhoye1 and Jacques Balthazart2 1

Bio-Imaging Laboratory, Department of Biomedical Sciences, University of Antwerp, Campus Middelheim, Groenenborgerlaan 171, 2020 Antwerp, Belgium 2 Center for Cellular and Molecular Neurobiology, Research Group in Behavioural Neuroendocrinology, University of Lie`ge, B-4020 Lie`ge, Belgium Received 31 March 2004; Revised 28 September 2004; Accepted 28 September 2004

ABSTRACT: The song control system of song birds is an excellent model for studying brain plasticity and has thus far been extensively analyzed by histological and electrophysiological methods. However, these approaches do not provide a global view of the brain and/or do not allow repeated measures, which are necessary to establish correlations between alterations in neural substrate and behavior. Application of in vivo manganese-enhanced MRI enabled us for the first time to visualize the song control system repeatedly in the same bird, making it possible to quantify dynamically the volume changes in this circuit as a function of seasonal and hormonal influences. In this review, we introduce and explore the song control system of song birds as a natural model for brain plasticity to validate a new cutting edge technique, which we called ‘repeated dynamic manganese enhanced MRI’ or D-MEMRI. This technique is based on the use of implanted permanent cannulae— for accurate repeated manganese injections in a defined target area—and the subsequent MRI acquisition of the dynamics of the accumulation of manganese in projection brain targets. A compilation of the D-MEMRI data obtained thus far in this system demonstrates the usefulness of this new method for studying brain plasticity. In particular it is shown to be a perfect tool for long-term studies of morphological and functional responses of specific brain circuits to changes in endocrine conditions. The method was also successfully applied to obtain quantitative measures of changes in activity as a function of auditory stimuli in different neuronal populations of a same nucleus that project to different targets. D-MEMRI, combined with other MRI techniques, clearly harbors potential for unraveling seasonal, hormonal, pharmacological or even genetically driven changes in a neuronal circuit, by simultaneously measuring changes in morphology, activity and connectivity. Copyright # 2004 John Wiley & Sons, Ltd. KEYWORDS: brain plasticity; song control system; song bird brain; manganese-enhanced MRI, ME MRI

*Correspondence to: A. Van der Linden, Bio-Imaging Laboratory, Department of Biomedical Sciences, University of Antwerp, Campus Middelheim, Groenenborgerlaan 171, 2020 Antwerp, Belgium. E-mail: [email protected] Contract/grant sponsor: Flemisch National Science Foundation; contract/grant numbers: G.0075.98, G.0420.02. Contract/grant sponsors: Institute for the Promotion of Innovation by Science and Technology in Flanders; BOF-NOI project from the University of Antwerp; RAFO project from the University of Antwerp. Contract/grant sponsor: NINDS; contract/grant number: NS 35467. Contract/grant sponsor: FRFC Project; contract/grant number: 2.4555.01. Contract/grant sponsor: Government of French Community of Belgium Project; contract/grant number: ARC 99/04-241. Abbreviations used: (AIM-MRI), activation-induced manganeseenhanced magnetic resonance imaging; BBB, blood–brain barrier; DLM, dorsal part of the medial thalamus; DM, dorsomedial part of the intercollicular nucleus of the midbrain; D-MEMRI, dynamic manganese-enhanced magnetic resonance imaging; E2, æstradiol; FOV, field of view; HVC, used as proper name, previously known as high vocal center; lMAN, lateral magnocellular nucleus of the anterior nidopallium; MEMRI, manganese-enhanced magnetic resonance imaging; n, coefficient that describes the shape of the curve; nXIIts, tracheosyringial part of the hypoglossal motor nucleus in the brainstem; RA, robust nucleus of the arcopallium; Ram, nucleus retroambigualis; SCN, song control nuclei; SCS, song control system; SImax, maximal signal intensity; T, testosterone; T50, time required to reach 50% of the maximum SImax; X, area X; Zenk, an immediate early gene also known as zif-268, egr-1, NGFI-A and Krox-24 in the literature on mammals—ZENK is the acronym of these four names. Copyright # 2004 John Wiley & Sons, Ltd.

INTRODUCTION Song birds share with humans the capacity to produce learned vocalizations, i.e. songs.1 The acquisition and production of songs is controlled by steroid-sensitive song control nuclei (SCNs) that show a remarkable seasonal plasticity.2 The volume and cytoarchitectonic organization of the SCNs vary as a function of the season, which makes these nuclei an ideal model system to study brain plasticity. These structural changes are associated with seasonal changes in song production and learning, making the song bird an exquisite model to study the interactions between neuroplasticity and learning. Since the early discovery of the SCN by Nottebohm and colleagues in 1976,3 the analysis of the song control system has become a ‘high impact’ research area, but it remained until recently deprived of the latest in-vivo imaging tools used in mammalian and human brain research. So far, the song bird brain has been successfully approached by histological and electrophysiological techniques, mainly focusing the song control system. However, these methods either fail to provide a global NMR Biomed. 2004;17:602–612

MEMRI OF BRAIN PLASTICITY IN SONG BIRDS

view of the brain (since they only focus on particular brain areas/cells) or fail to allow repeated measures, which remains the prerequisite to allow the identification of correlations between alterations in neural substrate and behavior. For this reason, multiple questions remain unsolved, dealing with the relationship between brain and behavior, that require an ‘entire brain’ approach to study in a holistic manner brain circuits and their functional implications. In 1998 we published the first in-vivo MRI canary atlas providing detailed three-dimensional views of the brain that identified major fiber tracts and several brain nuclei.4,5 Unfortunately, the vocal control circuit could not be resolved in these images. It was not until Pautler et al.6 introduced manganese-enhanced MEMRI as an in-vivo tract-tracing method that we started using MnCl2 injections in the brain of song birds and successfully visualized the vocal system for the first time in a living song bird.7 Meanwhile it was also discovered that MEMRI can be used to trace neuronal connections activated following specific stimulation.8 Being able to visualize the vocal control system in song birds and intrigued by the neuronal plasticity taking place in this circuit, we then introduced repeated dynamic manganese-enhanced D-MEMRI as a novel tool to assess changes in brain activity and functional connectivity that are dominant features of the neuronal plasticity in the vocal system.9,10 This paper provides an overview of what has been established so far in the song control system of the song bird brain with the use of ME and D-MEMRI. In addition, it also aims to introduce the reader to the song bird brain as a natural model to validate new cutting edge imaging techniques which are necessary to relate changes in specific areas of the brain to changes in brain function.

SONG BIRDS AS AN IDEAL MODEL TO STUDY NEURONAL PLASTICITY The song bird brain and the song control system Song birds are one of the few groups of vertebrates that have evolved a vocal communication system based on the production of songs of various complexities. These songs play a key role in social (sexual, aggressive) interactions and are learned during ontogeny. During the phylogenetic evolution that led to song birds, several telencephalic brain regions became part of a spatially organized neuronal circuitry along with the ability to learn and produce songs. These brain regions are called song control nuclei11 (Fig. 1). HVC (formerly called the high vocal center) is an important sensory motor region that serves as a relay within the vocal network connecting the brain areas involved in hearing, song production and vocal learning. It contains three distinct types of neurons from which two project respectively to the nucleus robustus Copyright # 2004 John Wiley & Sons, Ltd.

603

arcopallialis (RA, formerly nucleus robustus archistriatalis) or to area X of the medial striatum. The third type can be identified as interneurons. The two types of projection neurons are part of two brain circuits that play a critical role in the production of song (RA-projecting neurons) or its acquisition in juveniles and stability in adults (X-projecting neurons).3,12–17 RA-projecting neurons are part of a motor pathway that controls the activity of the syrinx, the sound-producing organ in birds, and of a medullary pathway that synchronizes respiration with song production.18 The X-projecting neurons, in contrast, belong to the so-called anterior forebrain pathway that is required for song learning but also plays a key role in maintaining song stability in adulthood.14,15 This pathway includes area X, located in the avian homologue of the mammalian basal ganglia (see Fig. 1) and it is therefore not surprising that this circuit plays a role in the coordination of fine motor actions. It is well know that in human neurodegenerative diseases such as Parkinson’s disease, alterations of the basal ganglia circuits, are associated with severe problems of motor coordination.

Neuroplasticity in the song control system It is only recently that scientists have recognized the existence of naturally occurring brain plasticity in adult homeothermic vertebrates. Most neuroscientists focus on laboratory species for obvious reasons, so historically there has been little opportunity to observe natural variation in the brain. The absence of substantial behavioral recovery after brain damage produced by accidental trauma or neurodegenerative diseases was, for many years, taken as evidence for the absence of marked plasticity in the adult central nervous system of many vertebrate species including humans. One major discovery that challenged this view was made in the brain of song birds (e.g. canaries, Serinus canaria), in which it was discovered that the volume of telencephalic brain nuclei specifically involved in the control of vocal behavior changes with season.19 This observation of substantial seasonal variation in the boundaries of a brain area based on Nissl-stained material raised a host of questions about the cellular basis of the phenomenon and what factors might regulate these changes. It was apparent from some of the initial investigations of volumetric changes in the song control system that testosterone promotes plasticity in this system.20,21 Continuous neurogenesis takes place in the ventricular zone of the lateral ventricles. Young neurons migrate and become incorporated into functional circuits throughout most of the telencephalon, but less so or not at all in other brain regions. This remarkable restriction of postnatal neurogenesis to the telencephalon suggests that these new cells may participate in higher associative brain functions such as perception and learning.22 The available data indicate NMR Biomed. 2004;17:602–612

604

A. VAN DER LINDEN ET AL.

that among song control nuclei, HVC is special in its ability to recruit, apart from ‘interneurons’, large numbers of projection neurons formed after hatching. This recruitment is also very specific: RA projection neurons in HVC are replaced at high rate during the annual cycle under the control of testosterone, while there is no replacement of area X projecting cells in HVC.23–25 Area X also incorporates large numbers of new neurons after hatching. However, contrary to what is observed in HVC, post-hatching area X neurogenesis is predominantly associated with the recruitement of interneurons.26 In RA only few neurons continue to be added after hatching.27 In this nucleus, volumetric changes are mainly associated with modifications of the neuronal size and extracellular space (neuropile) concomitant with dendritic growth.28 Neuronal recruitment occurs widely throughout the adult song bird brain,29 but has been studied best in the telencephalic HVC and adjacent mediocaudal nidopalium (formerly neostriatum30,31). In both cases the magnitude of the neuronal recruitment in adulthood is extraordinary: the adult HVC in canaries recruits over 1.4% of its neurons daily. Experiments also suggest that

gonadal androgens (testosterone) do not markedly affect the rate of neurogenesis but drastically increase the rate of post-mitotic neuronal survival.25,32,33 Even higher rates of neuronal production and recruitment have been inferred from ventricular zone labeling indices and neuronal recruitment patterns in the medioventral striatum.34–36 In contrast to the widespread occurrence of neurogenesis in the adult avian brain, the adult mammalian forebrain generates new neurons in small numbers only and in only a few regions, including the hippocampal dendate gyrus and olfactory bulb.37–39

THE SONG CONTROL SYSTEM OF SONG BIRDS AS AN IDEAL MODEL FOR IN VIVO MRI TRACT-TRACING STUDIES Manganese-enhanced magnetic resonance imaging (MEMRI) was used to trace neuronal connections in vivo in the olfactory and visual pathways of mice and rats6,40–42 and in several other neuronal circuits in monkeys43 and rats.44 The method also allows specific brain

———————————————————— Figure 1. (a) Schematic overview of the adult song bird brain. Birds have a very different pattern of telencephalic organization as compared with mammals. Although there are strong indications that all functional sensory, motor and cognitive regions found in the mammalian telencephalon are present in the telencephalon of birds, the avian gray matter is organized in conglomerates of cells interleaved with layers of white matter. This in contrast to the mammalian organization, which consists of two large concentrations of gray matter: the deep structures called basal ganglia, also composed of conglomerates of cells, and a layered cortex at the surface, separated by a vast mass of white matter. This fact originally led to the misunderstanding that the brain of birds only represents the basal ganglia and lacks the functional areas that are homologous to the mammalian cortex. Recent research in comparative neuroanatomy has proved this conclusion wrong and an international group of neuroanatomists has therefore decided recently to change the nomenclature of the telencephalic areas of birds. The terminology used in the present paper is based on the revised nomenclature that has recently been adopted.61 Additional information on the changes in neuroanatomical nomenclature for the avian brain can be found at the Avian Brain Nomenclature Exchange web site (http://avianbrain.org). This schematic drawing shows the major telencephalic subdivisions, the basal ganglia, the thalamus, the midbrain and the cerebellum. The song control system (SCS), which is present bilaterally in both hemispheres, comprises the following nuclei: HVC, RA, lMAN (lateral magnocellular nucleus of the anterior nidopallium, formerly neostriatum), DM (dorsomedial part of the intercollicular nucleus of the midbrain), area X, DLM (dorsal part of the medial thalamus), nXIIts (tracheosyringial part of the hypoglossal motor nucleus in the brainstem), RAm (nucleus retroambigualis). (b) In vivo midsagittal T2-weighted SE image of starling brain (TR/TE ¼ 2000/42 ms) Copyright # 2004 John Wiley & Sons, Ltd.

NMR Biomed. 2004;17:602–612

MEMRI OF BRAIN PLASTICITY IN SONG BIRDS

areas to be highlighted that are active [activation-induced manganese-enhanced (AIM-MRI)]45–47 or activated fiber tracts and their targets to be traced specifically.8 Similarly, intravenous or subcutaneous injections of MnCl2 resulted in Mn2þ accumulation in a large number of activated brain areas of the mouse brain.48 Manganese, however, does not easily cross the blood– brain barrier (BBB), and these authors had to introduce MnCl2 into the nervous tissue either by circumventing the BBB, using the olfactory, intravitreal, retinal and circumventricular pathways6,8,40–42,44,48,49 or by transient pharmacological opening of the BBB.45–47 In our studies on song birds, we used a different approach and injected manganese (MnCl2) stereotaxically directly into the brain to reach the neurons of interest.7,9,10 Other authors used a similar administration route in monkeys43 and mice.49 In comparison with injections performed in the caudate

605

nucleus, putamen, rostral striatum and the orbitofrontal cortex in monkeys43 or in the dorsal striatum and amygdala of mice,49 injections in the song control system of song birds are clearly easier and less harmful due to the superficial localization (0.7 mm subdural) of the most interesting target, HVC [Fig. 2(a,c)]. We have in addition established and validated a procedure with permanent cannulation of the target to improve the reproducibility of repeated injections even when performed several weeks apart. The song control circuit in song birds is characterized by a high density of axons connecting HVC with its target nuclei RA and area X and by the relatively large volume and clearly defined boundaries of these targets [Fig. 2 (c–f)]. Average volumes of area X and RA, as measured in female starlings by MEMRI, are 2.53
0.33 and 1.13
0.38 mm3, respectively.7 These characteristics find no match in any mammalian brain circuit. In

Figure 2. (a) Schematic overview of the adult songbird brain showing the song control nuclei (SCN) and their connections in the telencephalon. The black arrows represent the anterior forebrain pathway that starts in HVC and projects to area X ! DLM ! lMAN. The gray arrows indicate the motor pathway that projects from HVC directly to RA and further down to nXIIts and the syrinx. The two pathways originate from distinct cell populations within HVC. (b) Sagittal in vivo T2-weighted SE image of a starling brain (TR/TE ¼ 2000/ 42 ms) at the position where the SCN should appear, but they cannot be discerned on plane T2-weighted MRI [for comparison see (c)]. (c) Sagittal in vivo manganese-enhanced MRI of a male starling brain obtained 6 h after MnCl2 injection into HVC. The injection area is indicated by the gray arrow in panel (a) and by the signal enhanced area on the corresponding sagittal MRI slice displayed in this panel and the coronal MRI slice in (f). (c) Illustration of the different planes of imaging for (d–f). On all the images the areas enhanced by a brighter signal correspond to the labeling by Mn2þ. They correspond exactly to RA and area X as identified by histological techniques. Image resolution in the coronal plane is 97 mm (pixel size). Magnification bar ¼ 1 cm. [Adapted and reprinted from Neuroscience, 112, Van der Linden A, Verhoye M, Van Meir V, Tindemans I, Eens M, Absil P, Balthazart J. In vivo manganese-enhanced magnetic resonance imaging reveals connections and functional properties of the songbird vocal control system. 467–474., Copyright # 2002, with permission from Elsevier.]7 Copyright # 2004 John Wiley & Sons, Ltd.

NMR Biomed. 2004;17:602–612

606

A. VAN DER LINDEN ET AL.

particular the existence of dense connections between HVC, on the one hand, and RA and area X, on another hand, results in axonal transport ‘highways’ from the injection site (HVC) to the projection areas (RA and area X). As a result MEMRI using three-dimensional (3D) MRI sequences enabled us to visualize, accurately delineate and quantify the volume of two song control nuclei, repeatedly in the same living subjects in several species of birds including starlings, Sturnus vulgaris,7,10 canaries, Serinus canaria9 and even zebra finches, Taeniopygia guttata (unpublished). Clearly, this has not been accomplished before by any other method. Moreover, as outlined in the previous chapter, the volume of the song control nuclei HVC, RA and area X and the density of the connection between HVC and RA change as a function of seasonal and gonadal hormone influences. This makes the song control system the bestdocumented brain circuit available to study seasonal neuroplasticity. Besides using MEMRI to delineate song control nuclei and quantify their volume changes, we also expanded the method in order to obtain a novel MEMRI-based tool for the study of seasonal and functional changes in the organization of connections between, and activity changes within, the nuclei.

MEMRI for repeated visualization of the song control system Until recently, studies of the song control system exclusively relied on histological techniques that require the sacrifice of the experimental subjects and therefore prevent the use of birds as their own controls in successive measurements. The first MRI investigations from our group succeeded in discerning several anatomical structures in the avian brain, but failed to discern the song control nuclei.4 We discovered later that the injection of MnCl2 into HVC of starlings labels within a few hours RA and area X in images collected by 3D MEMRI. For the first time ever, the volume of these two song control nuclei could be visualized, accurately delineated and measured in a living song bird.7 To study seasonal or hormonally induced volume changes in these nuclei, repeated visualization of RA and area X in the same subjects was required. To that end, a permanent non-magnetic plastic cannula (i.d. 0.39 mm, o.d. 0.69 mm; Plastics One Inc., Roanoke, VA, USA) was implanted in HVC to allow repeated injections of MnCl2 (typically 200 nl of a 10 mM solution) at exactly the same location.10) No obvious behavioral differences were discerned between cannulated and non-cannulated birds nor between manganese-injected and non-injected birds as they resumed singing and breeding, and they regained their previous social status. The potential impact of the cannula on brain morphology was also tested in starlings by histology10 and revealed no differences between the cannulated and non-cannulated hemisphere. Copyright # 2004 John Wiley & Sons, Ltd.

The effect of 6 weeks of testosterone treatment on the volumes of RA and area X was investigated using repeated 3D MEMRI (coronal 3D, spin echo, T1weighted, FOV 25 mm in all directions, image matrix 265  256  256, acquisition time 90 min, 7 T system) in cannulated female starlings. A clear but non-significant volume increase (50.2 and 19.6% for RA and area X respectively) was observed in testosterone-treated birds. It was, however, paralleled by a volume decrease in controls (25.6 and 5.7% for RA and area X respectively; Fig. 4), possibly related to spontaneous seasonal changes or to the mild but chronic stress induced by captivity. Independent of the origin of this decrease in controls, these results illustrate the power of repeated imaging by MEMRI to analyze brain plasticity. Effects of testosterone were indeed superimposed to the effects of a nonidentified phenomenon leading to a substantial although not significant volume decrease in controls. As a consequence, the two-way ANOVA analyzing these results detected no effect of the two main factors (the endocrine condition of the birds and the two repeated measures before and after treatment), but identified a significant interaction between these two variables, indicating that testosterone indeed had a significant effect. The identification of this interaction was only made possible by the use of birds as their own controls and a parallel control group so that we could analyze volumes before and after a period of treatment with or without the steroid. This procedure detected effects of testosterone that would have gone unnoticed in a single histological measure. This clearly illustrates the value of repeated MEMRI as a tool for the study of brain plasticity over time.

Dynamic MEMRI vs MEMRI The song control circuit represents an excellent model of a brain circuit that changes in volume, connectivity and activity over time. MEMRI provides, as illustrated above, sensitive measures of volumetric changes in these nuclei but it can additionally be used to assess functional connectivity. We extended the MRI experiment described above by quantifying the dynamics of the Mn2þ uptake in RA and area X by an approach that we called ‘repeated dynamic manganese enhanced MRI’ or ‘D-MEMRI’. The rationale behind this approach is that, since RA and area X are monosynaptically connected to HVC, changes in the dynamics of Mn2þ accumulation in these nuclei are directly related to the electrical activity of the HVC neurons projecting to these targets and/or to the density of these projections. Indeed, as a Ca2þ analog, Mn2þ enters neurons via voltage-gated calcium channels.50 Once taken up by a projection neuron in HVC, Mn2þ is anterogradely transported along its axon probably via transport vesicles towards synapses located in RA or area X. To follow the dynamics of Mn2þ uptake, we developed a number of significant modifications to our imaging NMR Biomed. 2004;17:602–612

MEMRI OF BRAIN PLASTICITY IN SONG BIRDS

607

Figure 3. (a) Manganese accumulation in two HVC targets, RA and area X, as observed by manganeseenhanced MRI of starling brains, displayed as a function of time (up to 6 h) after MnCl2 injection in HVC. The figure shows changes in manganese-enhanced signal intensity in the area X and RA of a male and a female starling in coronal brain sections. The manganese-enhanced areas clearly illustrate the sex differences in shape and size of RA and area X. Magnification bar ¼ 5 mm. (b) Schematic representation of the changes in relative manganese-enhanced signal intensity observed in one target of HVC (RA or area X) as a function of time (up to 6 h) after MnCl2 injection in HVC (0 point). Data are plotted as a function of time and fitted by nonlinear regression to a sigmoid curve. This curve fully defines the kinetics of manganese accumulation in the target using three parameters: the maximal signal intensity (SImax), the time required to reach 50% of this maximum (T50) and the n coefficient that describes the shape of the curve. [Adapted and reprinted from Neuroscience, 112, Van der Linden A, Verhoye M, Van Meir V, Tindemans I, Eens M, Absil P, Balthazart J. In vivo manganese-enhanced magnetic resonance imaging reveals connections and functional properties of the songbird vocal control system. 467–474., Copyright # 2002, with permission from Elsevier.]7

protocol. Mn2þ had to be injected while the bird was already positioned in the magnet so that images could be acquired before, during and after Mn2þ injection. Therefore we used plastic cannullae chronically implanted in the HVC (as described in the previous chapter), but this time they were connected to a very long and narrow tubing (2 m long PE 50 tubing, i.d. 58 mm, o.d. 965 mm, Intramedic, Sparks, MD, USA) that allowed stereotaxic injection of very small volumes to be made at a distance. Subsequently, images (Coronal Multislice, Spin Echo, T1-weighted, FOV 30 mm in all directions, acquisition matrix 256  128, slice thickness 0.8 mm, 7 T system) were collected every 15 min, starting before the injection and for up to 6–7 h after the injection. Changes in the mean signal intensity were determined within the two song control nuclei (RA and area X) and in adjacent control areas in all images collected sequentially as follows (Fig. 3). RA and area X and adjacent control areas were first segmented on the T1-weighted multislice images obtained 6 h after injection and this area was then transposed to all other earlier images in the sequence as an area of interest where signal intensity was measured. Changes in relative signal intensity were then plotted as a function of time and fitted by nonlinear regression to a sigmoid curve to describe the kinetics of Mn2þ accumulation in the areas of interest. These curves were defined by the maximal signal intensity (SImax), the time required to reach 50% of this maximum (T50) and the n coefficient that describes the shape of the curve.7 When a sigmoid curve is used to describe complex (allosteric) kinetics in enzymology, n reflects the index of cooperativity and provides a measure of the numbers of active sites in an Copyright # 2004 John Wiley & Sons, Ltd.

enzyme that participate to the catalysis of the reaction (number of substrate binding sites); n therefore provides by analogy a measure of the complexity of the processes involved in the uptake, transport and accumulation of Mn2þ in the targets even if the cellular nature of these processes cannot be inferred from the value of this coefficient. With this experimental approach, the speed of axonal transport was estimated in starlings to be at least equal to 1.25 mm/h,7 which corresponds to the ‘fast axonal transport’ by which calcium is transported towards a synapse.51 Similar estimates of transport rate were obtained during MEMRI studies in rodents,6 despite the fact that birds have a higher body temperature than mammals (41–42 vs 36–37  C). This experimental approach was first validated by using it to compare male and female brains in European starlings (Sturnus vulgaris). Besides confirming previously established volumetric sex differences, the analysis of the dynamics of manganese uptake from HVC to RA and area X revealed new functional sex differences affecting manganese transport. A faster transport was observed in males than in females and different relative amounts of Mn2þ were transported to RA and area X in males as compared with females (see Van der Linden et al.7 for details). The cellular mechanisms that control the uptake and transport of calcium (and therefore Mn2þ) are still not completely understood but, based on available knowledge, it appears likely that the differences in Mn2þ accumulation that were observed here in RA and area X of males and females reflect mainly differences in the NMR Biomed. 2004;17:602–612

608

A. VAN DER LINDEN ET AL.

electrical activity of the corresponding HVC neurons. The rate of fast axonal transport is indeed supposed to be fairly stable in neurons in physiological conditions and can only be affected in pathological states (anoxia, temperature changes) or following very deep anesthesia.51,52 Although the velocity of fast axonal transport is not linked to the electrical activity of the axon, there are, however, some indications that this activity may influence the amount of transported material51 so that transport could further enhance changes in uptake driven by increased electrical activity in the neuron perikaryon itself. It is of course also evident that not only alterations in neuronal electrical activity, but also differences in the number of axons, a feature that is affected by a high degree of plasticity, will have an impact on the amount of transported manganese. The sex differences in the rate and amount of uptake detected in this study could thus reflect functional as well as morphological differences in connectivity. When uptake rates are compared on a shorter term basis, however, it becomes unlikely that the numbers of connections between nuclei could change and differences in uptake can more safely be attributed to functional alterations (see below).

D-MEMRI to discern changes in the connectivity of a brain circuit Effects of testosterone on the song control system can be best studied either in castrated males or in females. We used repeated dynamic MEMRI to study in individual female starlings (Sturnus vulgaris,
75 g) the long-term effects (6 weeks) of testosterone on the song control system. It could be expected, based on previous work on the effects of testosterone in male songbirds, that testosterone-treated female starlings would experience an increase in the number of RA-projecting HVC neurons while the number of area X-projecting HVC neurons should remain unchanged. The dynamics of Mn2þ accumulation in RA and area X was studied in two groups of female starlings that were treated with testosterone or kept as controls and two sets of measures were collected, just before the initiation of the treatment and then 6 weeks later. Testosterone did not affect the Mn2þ accumulation kinetics in RA (SImax, T50 and n), suggesting that the overall activity of RA-projecting HVC neurons was not affected by the steroid (Fig. 4). However testosterone at the same time increased RA volumes. If no change in the activity or density of HVC neurons and their projections had really occurred, given the increased RA size (see Fig. 4), the same amount of Mn2þ should have accumulated in a larger volume thus resulting in lower SImax values. It was therefore concluded that larger amounts of Mn2þ had been transported in testosteronetreated females, which presumably reflects the positive Copyright # 2004 John Wiley & Sons, Ltd.

effects of testosterone on the density of neuronal projections from HVC to RA. This increased number of projecting axons allowed more Mn2þ to be transported to a larger RA. This increase was directly proportional to the increase in RA volume so that no change in SImax was detected in this nucleus. The functional properties of these more numerous HVC neurons projecting to RA were also unchanged following testosterone treatment, so that the dynamics of transport (T50 and n parameters) were not altered. Testosterone also increased the total amount of Mn2þ transported to area X. In contrast to what had been observed in HVC, this increased Mn2þ accumulation was, however, associated with an increase in the maximal signal intensity (SImax) reached in this nucleus (Fig. 4). The shape of the curve (n parameter) describing this accumulation was also affected.10 These testosteroneinduced changes in area X cannot, as in RA, be explained by a change in the numbers of HVC neurons projecting to area X, since these neurons are known to be present in stable numbers in adult song birds. They presumably reflect functional changes in a stable number of neurons and projections. A variety of mechanisms could be involved, including an increased activity of the projection neurons, an increased recruitment of interneurons associated with area X-projecting HVC neurons or an increase in the numbers of tight junctions between these two types of neurons (see Van Meir et al.10 for more detailed explanations). In conclusion, the repeated Dynamic MEMRI experiments analyzing the song control system of female starlings demonstrated the presence of testosterone-dependent regionally specific changes in brain activity and functional connectivity. Because different cell populations within HVC project to RA and area X, the changes observed can also be univocally attributed to different sub-populations of neurons within a same area. The slow time scales investigated by this technique compared with electrophysiology or even functional MRI appear ideally suited for characterizing slow processes such as those involved in brain plasticity.

D-MEMRI to reveal changes in activity of specific neuronal populations D-MEMRI can, however, also be used to investigate faster changes in neuronal activity that have been traditionally studied by electrophysiology. It has been known for almost two decades that exposure of song birds to their own or to conspecific song causes a marked increase in neuronal firing within HVC.53 HVC contains both interneurons and neurons that project to area X or to RA and these different classes of HVC neurons react differentially to these auditory stimuli.54,55 Recently Mooney56 demonstrated that HVC neurons that project to RA or area X (identified based on morphological features) ‘generate remarkably NMR Biomed. 2004;17:602–612

MEMRI OF BRAIN PLASTICITY IN SONG BIRDS

609

Figure 4. Schematic overview of the known effects of testosterone (T) in the song bird brain that could be confirmed and extended in female starlings by dynamic manganese-enhanced MRI. Comparison of the bottom right drawing (brain of a control bird) with the upper left drawing (brain of a T-treated bird) reveals that the song control nuclei, HVC, RA and area X (see Fig. 2 for location and names of the nuclei) increase in volume following treatment with T. HVC recruits new interneurons (represented by an increased number of open circle ‘cells’) and large numbers of new RA-projecting neurons (represented by gray ‘cells’ with gray ‘axon arrows’). Neurogenesis in area X is predominantly associated with the formation of interneurons (increased number of open circle ‘cells’). In RA volumetric changes are mainly associated with dynamics in cellular size and extracellular space concomitant with dendritic growth and branching (same amount of open circle and gray ‘cells’ but increased volume of these cells and of the space between them). Testosterone exerts a direct influence on the brain through its binding to testosterone (T) receptors. However, significant indirect effects on brain activity can also be produced via binding to estrogen (E2) receptors of estrogens, such as estradiol (E2), produced by aromatization of T. The enzyme aromatase which catalyzes this enzymatic reaction is present in large amounts in the song birds’ telencephalon. Repeated 3D manganese-enhanced MRI confirmed the increased volume of RA and area X following T treatment, as shown in the inserts (volumetric changes) of the figures displayed upper right (area X) and lower left (RA). Repeated MRI also demonstrated changes in the dynamics of manganese accumulation in RA and area X following T treatment. These figures are reconstructions of the mean sigmoid curves based on the mean parameter values in each group from data obtained in control sham-implanted (gray curves) and testosterone-treated birds (black curves). The dashed curves represent pooled data for all birds before the beginning of the endocrine treatments. For interpretation of the data see text and also Van Meir et al.10 [Adapted and reprinted from Neuroscience, 112, Van der Linden A, Verhoye M, Van Meir V, Tindemans I, Eens M, Absil P, Balthazart J. In vivo manganese-enhanced magnetic resonance imaging reveals connections and functional properties of the songbird vocal control system. 467–474., Copyright # 2002, with permission from Elsevier.]7

similar patterns of song selective firing via different subthreshold mechanisms’. Owing to inherent limitations of electrophysiology, these studies were, however, limited to only one or a few neurons per bird. In addition, although large numbers of HVC neurons express the immediate early genes c-Fos and Zenk ( ¼ egr1) in a bird that is singing, no increase in immediate early gene expression Copyright # 2004 John Wiley & Sons, Ltd.

is observed in HVC following exposure to song.57 It is therefore impossible to obtain by these techniques a global view of the activity changes in all neurons that are affected by auditory stimuli and to assess simultaneously their hodology. We recently used D-MEMRI to achieve this goal. We investigated in canaries (Serinus canaria,
25 g) the effects of auditory stimulation by canary songs on the NMR Biomed. 2004;17:602–612

610

A. VAN DER LINDEN ET AL.

activity of the two types of HVC projection neurons. Permanent stereotaxic cannulae were used to allow repeated injection of MnCl2 at the exact same location within the HVC so that birds could be used repeatedly as their own control in the presence or absence of conspecific songs. MR imaging requires that birds remain immobile and therefore they are anesthetized for extended periods of time. A reasonable concern was therefore whether responses to song would be evoked under anesthetized conditions and whether these responses would not be confounded by the noise of the MR scanner. Fortunately, in anesthetized song birds, HVC and RA neurons display song-specific auditory responses that are sometimes stronger than in the awake state.58–60 Moreover, since we are not dealing with primary responses of the auditory system but with responses of the song nuclei that are triggered exclusively by very specific aspects of song, the noise of the MR scanner is of no concern in these experiments. Data summarized in Figure 5 show that exposure to songs markedly increased the slope of the sigmoid curve describing the accumulation of Mn2þ in RA (n coefficient) without affecting this value in area X. The maximal intensity of signal (SImax) reached in area X was, however, increased by exposure to song while it remained unaffected in RA. Based on the electrophysiological data of Mooney,56 it is not surprising that the two HVC cell populations projecting respectively to RA and area X responded differentially to a same set of auditory stimuli. The cellular/molecular mechanism(s) that mediate(s) the specific effects of song exposure on the dynamic of the

Mn2þ accumulation in RA but not in area X remain partly unexplained at present. It is clear, however, that the difference observed in Mn2þ accumulation in these two nuclei reflects a difference in the processing of Mn2þ ions in the corresponding circuits. Since Mn2þ acts in the brain as an analog of Ca2þ, it is likely that the effects observed here reflect differences in the Ca2þ physiology in the two types of HVC neurons that project to RA and area X. Although this represents the most likely explanation of the observed results, available data do not allow us to date to fully discard alternative explanations including effects of song exposure on the axonal rate of Mn2þ transport, on its synaptic release at the targets (RA or area X), its re-uptake by post-synaptic neurons in these targets or even its transport to other targets located further downstream in the song control circuit. It is indeed known, based on studies in mammals, that Mn2þ can be transported trans-synaptically49 and recent work in our laboratory confirms that this is also the case in song birds (unpublished data). In this particular experiment, changes in Mn2þ accumulation could not result from a change in HVC neurons numbers or in the density of their projections toward RA or area X since birds were kept on constant day/night cycle, the time between measurements with and without songs was too short and the order in which the two conditions were tested was selected at random for all subjects. The present study provides the first evidence that global differences in responding exist between the two populations of neurons projecting to RA or to area X. Previous suggestion that this is indeed the case was based on the electrophysiological analysis of at best a few dozen neurons whose connectivity was not always formally confirmed. Repeated D-MEMRI allowed for the first time to provide quantitative estimates of the global activity in the two neuronal populations with complete confidence about their target in a manner reminiscent of functional MRI or, on a slightly longer time scale, of 2-deoxyglucose autoradiography.

CONCLUSIONS Figure 5. Relative Manganese Enhanced Signal Intensity (relative SI) changes in RA (left) and area X (right) as a function of time after MnCl2 injection in the HVC in a representative canary. Thick lines and full squares represent data collected when the canary was allowed to listen to canary songs and thin lines and open squares indicate the control situation, without song stimulation, obtained in the same bird. The individual data acquired every 15 min are shown by squares, and the lines through these data points correspond to the sigmoid curves fitted by nonlinear regression as outlined in the text. For interpretation of the data see text and also Tindemans et al. [Reproduced from Tindemans I, Verhoye M, Balthazart J, Van der Linden A. In vivo dynamic ME-MRI reveals differential functional responses of RA and area X-projecting neurons in the HVC of canaries exposed to conspecific song. Eur. J. Neurosci. 2003; 18: 3352–3360. Reprinted with permission of Blackwell Publishing.]9 Copyright # 2004 John Wiley & Sons, Ltd.

A few years ago, studies of song birds based on MEMRI enabled us for the first time to visualize the song control system repeatedly in the same subjects, thereby making it possible to quantify longitudinally the volume changes that are known to occur under seasonal and hormonal influences in this circuit. Since then, our aim has been to explore, in a quantitative manner, the potentials of MEMRI for studying other features, besides volume changes, associated with neuronal plasticity in this circuit. To that end, we introduced the use of permanent non-magnetic cannulae chronically implanted in the injection target. This allowed accurate and repeated manganese injections as well as injections while the bird was already immobilized and oriented in the bore of the magnet. This allowed NMR Biomed. 2004;17:602–612

MEMRI OF BRAIN PLASTICITY IN SONG BIRDS

continuous MRI acquisition of the dynamic accumulation of manganese in the projection brain areas. In these studies, the rate of manganese accumulation in the projected areas was used as a measure of both the electrical activity of the projecting neurons at the injection site (HVC) and of the density of the connections between the injection site and its targets. We named this novel approach repeated dynamic manganese-enhanced D-MEMRI. The D-MEMRI data obtained so far in the song control system of song birds that were reviewed in the present paper provide evidence that this is a successful method for studying brain plasticity. In particular, it makes it feasible for the first time to perform long-term studies of morphological and functional responses of specific brain circuits to changes in endocrine conditions. The method was also demonstrated to be successful in obtaining quantitative estimates of activity in distinct neuronal populations within a same brain area provided they project to distinct brain targets. D-MEMRI obviously harbors an enormous potential for unraveling seasonal, hormonal, pharmacological or even genetically driven changes in a neuronal circuit, focussing on changes in morphology, activity and connectivity. The trans-synaptic nature of the manganese transport could even be used to expand our knowledge of the connections within the song control system and associated brain circuits. However this property also creates limitations because it complicates the interpretation of the images obtained by D-MEMRI. We recently started to use other MRI techniques to collect additional converging information on the song system, such as functional MRI to study the activity of the vocal circuit and diffusion tensor imaging to unravel changes in brain connections under changing circumstances. The synthesis of the data obtained with these different MRI methods will lead, without doubt, to a better understanding of the images obtained by D-MEMRI and a more exhaustive analysis of the seasonal plasticity and physiology of the song control system.

REFERENCES 1. Doupe AJ, Kuhl PK. Birdsong and human speech: common themes and mechanisms. A. Rev. Neurosci. 1999; 22: 567–631. 2. Tramontin AD, Brenowitz EA. Seasonal plasticity in the adult brain. Trends Neurosci. 2000; 23: 251–258. 3. Nottebohm F, Stokes TM, Leonard CM. Central control of song in the canary, Serinus canarius. J. Comp. Neurol. 1976; 165: 457– 486. 4. Van der Linden A, Verhoye M, Peeters R, Eens M, Newman SW, Smulders T, Balthazart J, DeVoogd TJ. Non invasive in vivo anatomical studies of the oscine brain by high resolution MRI microscopy. J. Neurosci. Meth. 1998; 81: 45–52. 5. Verhoye M, Van der Linden A, Van Audekerke J, Sijbers J, Eens M, Balthazart J. Imaging birds in a bird cage: in-vivo FSE 3D MRI of bird brain. MAGMA 1998; 6: 22–27. 6. Pautler RG, Silva AC, Koretsky AP. In vivo neuronal tract tracing using manganese-enhanced magnetic resonance imaging. Magn. Reson. Med. 1998; 40: 740–748. Copyright # 2004 John Wiley & Sons, Ltd.

611

7. Van der Linden A, Verhoye M, Van Meir V, Tindemans I, Eens M, Absil P, Balthazart J. In vivo manganese-enhanced magnetic resonance imaging reveals connections and functional properties of the songbird vocal control system. Neuroscience 2002; 112: 467–474. 8. Pautler RG, Koretsky AP. Tracing odor-induced activation in the olfactory bulbs of mice using manganese-enhanced magnetic resonance imaging. Neuroimage 2002; 16: 441–448. 9. Tindemans I, Verhoye M, Balthazart J, Van der Linden A. In vivo dynamic ME-MRI reveals differential functional responses of RAand area X-projecting neurons in the HVC of canaries exposed to conspecific song. Eur. J. Neurosci. 2003; 18: 3352–3360. 10. Van Meir V, Verhoye M, Absil P, Eens M, Balthazar J, Van der Linden A. Differential effects of testosterone on neuronal populations and their connections in a sensorimotor brain nucleus controlling song production in songbirds: a manganese enhanced-magnetic resonance imaging study. Neuroimage 2004; 21: 914–923. 11. Brenowitz EA. Comparative approaches to the avian song system. J. Neurobiol. 1997; 33: 517–531. 12. Yu AC, Margoliash D. Temporal hierarchical control of singing in birds. Science 1996; 273: 1871–1875. 13. Scharff C, Nottebohm F. A comparative study of the behavioral deficits following lesions of various parts of the zebra finch song system: implications for vocal learning. J. Neurosci. 1991; 11: 2896–2913. 14. Brainard MS, Doupe AJ. Auditory feedback in learning and maintenance of vocal behaviour. Nat. Rev. Neurosci. 2000; 1: 31–40. 15. Brainard MS, Doupe AJ. Interruption of a basal ganglia-forebrain circuit prevents plasticity of learned vocalizations. Nature 2000; 404: 762–766. 16. Hessler NA, Doupe AJ. Singing-related neural activity in a dorsal forebrain-basal ganglia circuit of adult zebra finches. J. Neurosci. 1999; 19: 10461–10481. 17. Jarvis ED, Scharff C, Grossman MR, Ramos JA, Nottebohm F. For whom the bird sings: context-dependent gene expression. Neuron 1998; 21: 775–788. 18. Wild JM. Neural pathways for the control of birdsong production. J. Neurobiol. 1997; 33: 653–670. 19. Nottebohm F. A brain for all seasons: cyclical anatomical changes in song control nuclei of the canary brain. Science 1981; 214: 1368–1370. 20. DeVoogd T, Nottebohm F. Gonadal hormones induce dendritic growth in the adult avian brain. Science 1981; 214: 202–204. 21. Nottebohm F. Testosterone triggers growth of brain vocal control nuclei in adult female canaries. Brain Res. 1980; 189: 429–436. 22. Alvarez-Buylla A, Kirn JR, Nottebohm F. Birth of projection neurons in adult avian brain may be related to perceptual or motor learning. Science 1990; 249: 1444–1446. 23. Nordeen EJ, Nordeen KW. Sex and regional differences in the incorporation of neurons born during song learning in zebra finches. J. Neurosci. 1988; 8: 2869–2874. 24. Nordeen KW, Nordeen EJ. Projection neurons within a vocal motor pathway are born during song learning in zebra finches. Nature 1988; 334: 149–151. 25. Rasika S, Nottebohm F, Alvarez-Buylla A. Testosterone increases the recruitment and/or survival of new high vocal center neurons in adult female canaries. Proc. Natl Acad. Sci. USA 1994; 91: 7854–7858. 26. Sohrabji F, Nordeen EJ, Nordeen KW. Characterization of neurons born and incorporated into a vocal control nucleus during avian song learning. Brain Res. 1993; 620: 335–338. 27. Kirn JR, DeVoogd TJ. Genesis and death of vocal control neurons during sexual differentiation in the zebra finch. J. Neurosci. 1989; 9: 3176–3187. 28. Schlinger BA, Brenowitz EA. Neural and hormonal control of birdsong. In Hormones, Brain and Behavior, vol. 2, Pfaff DW, Arnold AP, Etgen AM, Fahrbach SE, Rubin RT (eds). Academic Press: San Diego, CA, 2002; 799–839. 29. Alvarez-Buylla A, Nottebohm F. Migration of young neurons in adult avian brain. Nature 1988; 335: 353–354. 30. Goldman SA, Nottebohm F. Neuronal production, migration, and differentiation in a vocal control nucleus of the adult female canary brain. Proc. Natl Acad. Sci. USA 1983; 80: 2390–2394. NMR Biomed. 2004;17:602–612

612

A. VAN DER LINDEN ET AL.

31. Paton JA, O’Loughlin BE, Nottebohm F. Cells born in adult canary forebrain are local interneurons. J. Neurosci. 1985; 5: 3088–3093. 32. Hidalgo A, Barami K, Iversen K, Goldman SA. Estrogens and non-estrogenic ovarian influences combine to promote the recruitment and decrease the turnover of new neurons in the adult female canary brain. J. Neurobiol. 1995; 27: 470–487. 33. Alvarez-Buylla A, Kirn JR. Birth, migration, incorporation, and death of vocal control neurons in adult songbirds. J. Neurobiol. 1997; 33: 585–601. 34. Alvarez-Buylla A, Garcia-Verdugo JM, Mateo AS, MerchantLarios H. Primary neural precursors and intermitotic nuclear migration in the ventricular zone of adult canaries. J. Neurosci. 1998; 18: 1020–1037. 35. Alvarez-Buylla A, Ling CY, Yu WS. Contribution of neurons born during embryonic, juvenile, and adult life to the brain of adult canaries: regional specificity and delayed birth of neurons in the song-control nuclei. J. Comp. Neurol. 1994; 347: 233–248. 36. Alvarez-Buylla A, Theelen M, Nottebohm F. Proliferation ‘hot spots’ in adult avian ventricular zone reveal radial cell division. Neuron 1990; 5: 101–109. 37. Altman J, Das GD. Autoradiographic and histological evidence of postnatal hippocampal neurogenesis in rats. Comp. Neurol. 1965; 124: 319–335. 38. Bayer SA, Yackel JW, Puri PS. Neurons in the rat dentate gyrus granular layer substantially increase during juvenile and adult life. Science 1982; 216: 890–892. 39. Kaplan MS, Hinds JW. Neurogenesis in the adult rat: electron microscopic analysis of light radioautographs. Science 1977; 197: 1092–1094. 40. Watanabe T, Michaelis T, Frahm J. Mapping of retinal projections in the living rat using high-resolution 3D gradientþ echo MRI with Mn2 -induced contrast. Magn. Reson. Med. 2001; 46: 424–429. 41. Lin CP, Tseng WY, Cheng HC, Chen JH. Validation of diffusion tensor magnetic resonance axonal fiber imaging with registered manganese-enhanced optic tracts. Neuroimage 2001; 14: 1035– 1047. 42. Natt O, Watanabe T, Boretius S, Radulovic J, Frahm J, Michaelis T. High-resolution 3D MRI of mouse brain reveals small cerebral structures in vivo. J. Neurosci. Meth. 2002; 120: 203–209. 43. Saleem KS, Pauls JM, Augath M, Trinath T, Prause BA, Hashikawa T, Logothetis NK. Magnetic resonance imaging of neuronal connections in the macaque monkey. Neuron 2002; 34: 685–700. 44. Leergaard TB, Bjaalie JG, Devor A, Wald LL, Dale AM. In vivo tracing of major rat brain pathways using manganese-enhanced magnetic resonance imaging and three-dimensional digital atlassing. Neuroimage 2003; 20: 1591–1600. 45. Lin YJ, Koretsky AP. Manganese ion enhances T1-weighted MRI during brain activation: an approach to direct imaging of brain function. Magn. Reson. Med. 1997; 38: 378–388.

Copyright # 2004 John Wiley & Sons, Ltd.

46. Duong TQ, Silva AC, Lee SP, Kim SG. Functional MRI of calcium-dependent synaptic activity: cross correlation with CBF and BOLD measurements. Magn. Reson. Med. 2000; 43: 383–392. 47. Aoki I, Tanaka C, Takegami T, Ebisu T, Umeda M, Fukunaga M, Fukuda K, Silva AC, Koretsky AP, Naruse S. Dynamic activityinduced manganese-dependent contrast magnetic resonance imaging (DAIM MRI). Magn. Reson. Med. 2002; 48: 927–933. 48. Watanabe T, Natt O, Boretius S, Frahm J, Michaelis T. In vivo 3D MRI staining of mouse brain after subcutaneous application of MnCl2. Magn. Reson. Med. 2002; 48: 852–859. 49. Pautler RG, Mongeau R, Jacobs RE. In vivo trans-synaptic tract tracing from the murine striatum and amygdala utilizing manganese enhanced MRI (MEMRI) 1. Magn. Reson. Med. 2003; 50: 33–39. 50. Narita K, Kawasaki F, Kita H. Mn and Mg influxes through Ca channels of motor nerve terminals are prevented by verapamil in frogs. Brain Res. 1990; 510: 289–295. 51. Grafstein B, Forman DS. Intracellular transport in neurons. Physiol. Rev. 1980; 60: 1167–1283. 52. Vallee RB, Bloom GS. Mechanisms of fast and slow axonal transport. A. Rev. Neurosci. 1991; 14: 59–92. 53. Margoliash D. Preference for autogenous song by auditory neurons in a song system nucleus of the white-crowned sparrow. J. Neurosci. 1986; 6: 1643–1661. 54. Lewicki MS. Intracellular characterization of song-specific neurons in the zebra finch auditory forebrain. J. Neurosci. 1996; 16: 5855–5863. 55. Dutar P, Vu HM, Perkel DJ. Multiple cell types distinguished by physiological, pharmacological, and anatomic properties in nucleus HVc of the adult zebra finch. J. Neurophysiol. 1998; 80: 1828–1838. 56. Mooney R. Different subthreshold mechanisms underlie song selectivity in identified HVc neurons of the zebra finch. J. Neurosci. 2000; 20: 5420–5436. 57. Jarvis ED, Nottebohm F. Motor-driven gene expression. Proc. Natl Acad. Sci. USA 1997; 94: 4097–4102. 58. Dave AS, Yu AC, Margoliash D. Behavioral state modulation of auditory activity in a vocal motor system. Science 1998; 282: 2250–2254. 59. Schmidt MF, Konishi M. Gating of auditory responses in the vocal control system of awake songbirds. Nat. Neurosci. 1998; 1: 513–518. 60. Nick TA, Konishi M. Dynamic control of auditory activity during sleep: correlation between song response and EEG. Proc. Natl Acad. Sci. USA 2001; 98: 14012–14016. 61. Reiner AD, Perkel J, Bruce L, Butler A, Csillag A, Kuenzel W, Medina L, Paxinos G, Shimizu T, Striedter G, Wild M, Ball GF, Durand S, Gu¨ntu¨rku¨n O, Lee DW, Mello CV, White SA, Hough G, Kubikova L, Smulders TV, Wada K, Dugas-Ford J, Husband S, Yamamoto K, Yu J, Siang C, Jarvis ED. Revised nomenclature for avian telencephalon and some related brainstem nuclei. J. Comp. Neurol 2004; 473(3): 377–414.

NMR Biomed. 2004;17:602–612