- Email: [email protected]
Three-dimensional analysis of avian song control nuclei
Journal of Neuroscience Methods 121 (2002) 75 /80 www.elsevier.com/locate/jneumeth
Three-dimensional analysis of avian song control nuclei George E. Bentley *, Eliot A. Brenowitz Departments of Psychology and Zoology, and Virginia Merrill Bloedel Hearing Research Center, Box 351525, University of Washington, Seattle, WA 98195-1525, USA Received 26 February 2002; received in revised form 9 August 2002; accepted 9 August 2002
Abstract Analysis of seasonal and developmental changes in the morphology of avian song control nuclei has traditionally been performed using two-dimensional (2-D) cross-sectional traces from brain sections. This method, although reliable, does not encompass the possibility that subdivisions of a nucleus might change in size to different degrees. Three-dimensional (3-D) analysis of song nuclei under different conditions could provide insight on this issue. This approach could also be of value in guiding and evaluating the use of lesions to study the functions of subdivisions of song nuclei. We used customized computer software to produce 3-D images of song nuclei from 2-D brain sections of spotted towhees (Pipilo maculatus ) in different hormonal status, and from Gambel’s whitecrowned sparrows (Zonotrichia leucophrys gambelii ) with unilateral lesions of the higher vocal center (HVc). 3-D images show that some sub-regions of song nuclei indeed change in size to a greater extent than others. 3-D analysis of HVc lesions provides a clearer view of the size and shape of the lesion site within the target nucleus and relative to the surrounding tissue. Used in conjunction with 2-D analysis, the 3-D method will aid investigations of the song system and contribute to the understanding of its regulation by hormones. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Birdsong; Bird; Volume reconstruction; Lesion
1. Introduction Songbirds possess a hormone-sensitive network of interconnected brain nuclei that regulates song learning, perception and production (Nottebohm et al., 1976; Nottebohm, 1980, 1981), as depicted in Fig. 1. All seasonally breeding songbirds examined exhibit dramatic seasonal changes in the size of these nuclei (e.g. Nottebohm 1981; Tramontin and Brenowitz, 2000; Ball, 2000). Increases in size of the song control nuclei during the breeding season (spring) are associated with increased song stereotypy and duration (Tramontin and Brenowitz, 2000). The telencephalic song regions HVc (also known as the higher vocal center) and the robust nucleus of the archistriatum (RA) increase in volume by as much as 188 and 135%, respectively, in the spring (Smith, 1996; Tramontin and Brenowitz, 2000).
* Corresponding author. Tel.: /1-206-543-3356; fax: /1-206-6853157 E-mail address: [email protected] (G.E. Bentley).
Study of seasonal plasticity in the song control system of songbirds has largely focused on changes in volumes of the song control nuclei (for review see Tramontin and Brenowitz, 2000). The volumetric changes of song control nuclei are caused by changes in neuron number, size, or spacing, depending upon the nucleus. In whitecrowned sparrows (Zonotrichia leucophrys ) and song sparrows (Melospiza melodia ), fewer neurons are present in HVc of fall birds as compared to spring birds (Smith et al., 1995, 1997; Tramontin and Brenowitz, 2000). Neurons in RA are smaller and more closely spaced in the fall as compared to the spring. Measuring the volume of a brain nucleus under different physiological or developmental conditions provides useful information on changes in overall size (see Fig. 2 for example). This approach, however, does not provide information on whether such changes occur equally throughout the nucleus, or whether there are differential changes in portions or subdivisions of that nucleus. In the song control system, for example, functional and morphological distinctions have been demonstrated between the lateral and medial portions of
0165-0270/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 0 2 7 0 ( 0 2 ) 0 0 2 4 4 - 3
76
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
structions to the contexts of seasonal plasticity and identification of lesion sites will be considered.
2. Materials and methods 2.1. Tissue sections
Fig. 1. Schematic sagittal drawing of the songbird brain showing projections of the major nuclei in the song system. The motor pathway (solid black arrows) controls the production of song. The solid black arrows indicate the descending pathway from HVc in the neostriatum to RA in the archistriatum. From RA, the pathway projects to the vocal motor nucleus nXIIts. The open (white) arrows indicate the anterior forebrain pathway that is essential for song learning. This pathway indirectly links HVc to RA, via the basal ganglia homolog area X, DLM in the thalamus, and lMAN in the neostriatum. lMAN also projects to area X. DLM, medial portion of the dorsolateral nucleus of the thalamus; lMAN, lateral portion of the magnocellular nucleus of the anterior neostriatum; RA, robust nucleus of the archistriatum; V, ventricle; X, area X; nXIIts, tracheosyringeal part of the hypoglossal nucleus. Adapted from Brenowitz et al. (1997).
the nuclei MAN and area X. In area X, immediate-early gene expression differs between the lateral (l) and medial (m ) portions with social context (Jarvis et al., 1998), and the distribution of melatonin receptors differs between these portions according to photoperiodic status (Bentley and Ball, 2000). As social cues and melatonin are known to influence seasonal changes in volumes of song control nuclei (Tramontin et al., 1999; Bentley et al., 1999), it seems feasible that these environmental cues might affect the change in size and structure of one part of a song nucleus more than another. Analyzing the three-dimensional (3-D) structure of song nuclei under different conditions could provide insight on this issue. Such information, combined with the known connectivity of the song control pathway and hormone receptor distribution, would shed light on the pathways involved in developmental and seasonal plasticity in this hormone-sensitive model system. 3-D analysis could also be of value in guiding and evaluating the use of lesions to study the functions of song nuclei (e.g. Scharff et al., 1998; Brainard and Doupe, 2000; Brenowitz and Lent, 2001). Traditional 2D reconstructions of lesioned nuclei suffer from similar limitations as discussed above. It would be informative to reconstruct 3-D images of lesion sites within a target nucleus to determine the extent of lesion. Rotation of the 3-D image would allow one to see exactly where the lesion has been made relative to inputs/outputs and hormone receptor distribution. We describe here the use of customized computer software to produce 3-D images of song control nuclei. The application of 3-D recon-
Tissue sections from previously-published studies were used for the 3-D reconstruction (Smith, 1996; Brenowitz and Lent, 2001). Brains were collected after perfusion /fixation, sectioned coronally at 50 mm, and Nissl-stained. Although Nissl stain was used in this study, any histological marker that provides clearlydefined detail of the boundaries of the area(s) of interest would be suitable. Images of seasonal changes in song control nuclei were constructed from male brains of spotted towhees (Pipilo maculatus ). Brains from individuals in different reproductive conditions were used to illustrate the morphometric changes of song control nuclei in 3-D. Images of HVc lesions were constructed from brains of Gambel’s white-crowned sparrows (Zonotrichia leucophrys gambelii ). 2.2. 3-D image construction Images were captured using NIH Image version 1.62 (Wayne Rasband, NIH) on an Apple Macintosh computer linked to a video camera and light microscope. Sequential images through each nucleus of interest were captured and 3-D images were constructed using ‘MacSurfer’ software version 1.3.2 (S. Kurkin, Japan Science and Technology Corporation, Department of Physiology, Hokkaido University School of Medicine, Japan. E-mail: [email protected]). The following link may also be used to download the software free of charge: http://133.87.71.180/MacSurfer.sit. The most critical consideration in capturing 2-D images for use in 3-D image construction is the correct alignment of sequential images. The MacSurfer program interpolates between sequential images, so misalignment can result in distortion of the 3-D image as a whole. For this study, the first section containing a song control nucleus of interest was captured and the rest of the sections through that nucleus were aligned relative to the first section. Alignment was achieved by using the ‘paste control’ function in NIH Image, in conjunction with the ‘Xor’ transfer mode in either the ‘scale math’ or ‘live paste’ options. This procedure enabled us to lay a semi-transparent image of the second section over an opaque image of the first section. Alignment was based on several landmarks in each section, including the midline, the nucleus of interest and the outline of the telencephalon. Other landmarks that were used depended upon the area of the brain that was sectioned.
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
77
Fig. 2. Thionin-stained transverse sections of HVc (A1, A2); RA (B1, B2); and area X (C1, C2) in breeding (A1, B1, C1) and non-breeding (A2, B2, C2) male spotted towhees. All sections are from the central portion along the rostral /caudal axis of each nucleus. Arrowheads indicate borders of song nuclei. Bars /0.5 mm. Adapted from Smith (1996).
A similar alignment method can be employed using Adobe Photoshop† software. Once sections were aligned, an image for each section was saved in TIFF format. The TIFF files were imported into the MacSurfer program, and an outline of the area of interest was drawn using a computer mouse. A computer-generated 3-D image was created from the stack of outlines created with the mouse. The entire process of image capturing, alignment and 3-D construction was repeated for each nucleus shown.
3. Results 3.1. Photoperiod-induced changes in the song control system 3-D comparison of the song control system of spotted towhees on short day lengths (non-breeding) and long day lengths (breeding) depicts areas of song nuclei that undergo volumetric changes as a result of changes in
hormonal status associated with changing photoperiod (Fig. 3A). The caudal view is presented here instead of the rostral view to enable the caudomedial extension of HVc to be viewed more clearly. The dorsal view was created by the MacSurfer program using the same information as for the caudal view; the image was rotated through 908 using one of the program options. It is clear from these figures that the song control nuclei depicted do not change size uniformly throughout their structure. Looking at each nucleus that was reconstructed in turn, 3-D image construction allows us to describe morphological characteristics that are not easily discernible from 2-D images: Area X : In the transition from short to long days, area X grows in all planes, but more in the medial-lateral plane that in other planes. There is proportionally more growth in m area X and the ‘neck’ joining m area X and l area X than in l area X alone. HVc: In its most lateral portion (main corpus), HVc grows to some extent in the dorsal /ventral plane.
78
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
Fig. 3. (A) Song control nuclei of male spotted towhees represented in 3-D. Song nuclei from non-breeding (green) and breeding (red) individuals are included in a single image to facilitate comparison. Caudal and dorsal views allow for easy viewing of morphology. Medial and lateral area X are labeled m and l , respectively. (B) Rostral and caudal views of unilateral lesions directed at HVc in a Gambel’s white-crowned sparrow (Bird A). Green color depicts the target HVc, red depicts the intact HVc. Arrows indicate the injection sites, and the total lesioned area is represented in purple. (C) As B, but for a separate individual (Bird B).
Overall, it increases a great deal in its rostral /caudal plane. The caudomedial extension of HVc also projects further ventrally as it extends caudally, curling over on itself at its trailing edge. RA : Growth of RA occurs in all planes, but to a great extent in the medial/lateral plane. As it extends caudally, it also curves medially.
3.2. 3-D reconstruction of lesions directed at HVc Fig. 3B and C consist of rostral and caudal views of 3D reconstructions of lesions directed at HVc in Gambel’s white-crowned sparrows. HVc was lesioned unilaterally with two injections of 1% racemic N -methylaspartic acid (NMA) in phosphate-buffered saline.
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
Lesions were directed at both the rostral main corpus and caudomedial extension of HVc (see Brenowitz and Lent, 2001 for full details). The two sites of insertion of the injection micropipette can clearly be seen in Fig. 3B and C. It is apparent from this figure that in both birds the NMA spread from the two injection sites and joined them together, creating a single large lesion. The first lesion attempt (Bird A; Fig. 3B), destroyed part of the lateral portion of the main corpus of HVc. Lesioned HVc was overall much smaller than on the unlesioned side. In the second lesion attempt (Bird B; Fig. 3C), the rostral part of the caudomedial extension of HVc was destroyed. The size and shape of the main corpus of HVc appeared to be unaffected by the lesion.
4. Discussion This study describes a method by which the 3-D characteristics of song control nuclei can be analyzed. To convey the same information using 2-D images would require multiple figures that are not as easily interpreted. It is clear from the results that the song control nuclei depicted do not change size uniformly throughout their whole structure. This observation could have implications for determining the relative contributions of different environmental cues to plasticity in the song control system. For example, changes in photoperiod alone might affect subdivisions of a nucleus that are not affected by social cues alone, and vice-versa. Importantly, failure to observe an overall difference in volume of a nucleus between two treatment groups might belie significant changes in subregions, possibly as a result of heterogeneous hormone receptor distribution or afferent/efferent projections. A similar principle is true for studies involving lesions of song control nuclei. The 3-D analysis provides a more informative picture of lesion location, both within a song control nucleus and relative to the surrounding neural tissue. The ability to pinpoint lesions in 3-D to exact locations within the target nucleus creates potential for greater understanding of the relationship between its topography and behavioral outputs. When embarking on a study of this type, the investigator must decide what section thickness to use. The optimal thickness will depend upon the system studied and clarity of staining. There tends to be a tradeoff between section thickness and clarity of delineation of the area in question. Ideally, thin sections allow the most accurate measurement of the volume occupied by a brain nucleus of interest. In practice, however, the borders of brain nuclei are more clearly delineated in thick sections. As a result, sampling intervals of between 40 and 120 mm tend to be used for measurement of the
79
song control nuclei. In the present study we use alternately-collected 50 mm sections (sampling interval of 100 mm). Seasonal changes within the song system can easily be detected with this sampling interval, as is shown in Fig. 3A. In summary, the 3-D reconstruction approach has several features that allow for finely-detailed investigation of manipulations of the song control system, and can complement 2-D methods of analysis. Although this 3-D method is not in itself quantitative, it is an excellent aid to lesion studies, and provides additional information for studies of seasonal regulation of the song control system. Increasingly, it is becoming apparent that subdivisions of a nucleus process environmental stimuli in different ways, e.g., m area X vs. l area X (Jarvis et al., 1998; Bentley and Ball, 2000). We propose that if adopted for routine use, this 3-D approach will provide new information about the contribution of environmental cues to morphometric changes in the subdivisions of song control nuclei. In addition, these methods are well-suited for studying developmental changes in the song system, as well as other forms of adult plasticity such as ongoing neuronal recruitment. This approach could also be applied to other vertebrate neural systems that undergo morphometric plasticity at different life stages, such as during development or senescence. The same approach would also be suitable for comparison of, for example, the sexually dimorphic nucleus in rats, the preoptic area in birds, rats and lizards and the medial amygdala in rats.
Acknowledgements We thank Glen MacDonald and Karin Lent for their technical assistance. Supported by NIH grant MH53032 (to E.A.B.) and the Virginia Merrill Bloedel Hearing Research Center.
References Ball GF. Neuroendocrine basis of seasonal changes in vocal behavior among songbirds. In: Hauser M, Konishi M, editors. Neural mechanisms of communication. Cambridge MA: MIT press, 2000:213 /53. Bentley GE, Ball GF. Photoperiod-dependent and -independent regulation of melatonin receptors in the forebrain of songbirds. J Neuroend 2000;12:745 /52. Bentley GE, Van’t Hof TJ, Ball GF. Seasonal neuroplasticity in the songbird telencephalon: a role for melatonin. Proc Natl Acad Sci USA 1999;96:4674 /9. Brainard MS, Doupe AJ. Interruption of a basal ganglia-forebrain circuit prevents plasticity of learned vocalizations. Nature 2000;404:762 /6. Brenowitz EA, Lent K. Afferent input is necessary for seasonal growth and maintenance of adult avian song control circuits. J Neurosci 2001;21:2320 /9.
80
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
Brenowitz EA, Margoliash D, Nordeen KW. An introduction to birdsong and the avian song system. J Neurobiol 1997;33:495 /500. Jarvis ED, Scharff C, Grossman MR, Ramos JA, Nottebohm F. For whom the bird sings: context-dependent gene expression. Neuron 1998;21:775 /88. Nottebohm F. Testosterone triggers growth of brain vocal control nuclei in adult female canaries. Brain Res 1980;189:429 /36. Nottebohm F. A brain for all seasons: cyclical anatomical changes in song control nuclei of the canary brain. Science 1981;214:1368 /70. Nottebohm F, Stokes TM, Leonard CM. Central control of song in the canary, Serinus canarius . J Comp Neurol 1976;165:457 /86. Scharff C, Nottebohm F, Cynx J. Conspecific and heterospecific song discrimination in male zebra finches with lesions in the anterior forebrain pathway. J Neurobiol 1998;36:81 /90.
Smith GT. Seasonal plasticity in the song nuclei of wild rufous-sided towhees. Brain Res 1996;734:79 /85. Smith GT, Brenowitz EA, Beecher MD, Wingfield JC. Seasonal changes in testosterone, neural attributes of song control nuclei, and song structure in wild songbirds. J Neurosci 1997;17:6001 /10. Smith GT, Brenowitz EA, Wingfield JC, Baptista LF. Seasonalchanges in song nuclei and song behavior in Gambel’s whitecrowned sparrows. J Neurobiol 1995;28:114 /25. Tramontin AD, Brenowitz EA. Seasonal plasticity in the adult brain. Trends Neurosci 2000;23:251 /8. Tramontin AD, Wingfield JC, Brenowitz EA. Contributions of social cues and photoperiod to seasonal plasticity in the adult avian song control system. J Neurosci 1999;19:476 /83.
Three-dimensional analysis of avian song control nuclei George E. Bentley *, Eliot A. Brenowitz Departments of Psychology and Zoology, and Virginia Merrill Bloedel Hearing Research Center, Box 351525, University of Washington, Seattle, WA 98195-1525, USA Received 26 February 2002; received in revised form 9 August 2002; accepted 9 August 2002
Abstract Analysis of seasonal and developmental changes in the morphology of avian song control nuclei has traditionally been performed using two-dimensional (2-D) cross-sectional traces from brain sections. This method, although reliable, does not encompass the possibility that subdivisions of a nucleus might change in size to different degrees. Three-dimensional (3-D) analysis of song nuclei under different conditions could provide insight on this issue. This approach could also be of value in guiding and evaluating the use of lesions to study the functions of subdivisions of song nuclei. We used customized computer software to produce 3-D images of song nuclei from 2-D brain sections of spotted towhees (Pipilo maculatus ) in different hormonal status, and from Gambel’s whitecrowned sparrows (Zonotrichia leucophrys gambelii ) with unilateral lesions of the higher vocal center (HVc). 3-D images show that some sub-regions of song nuclei indeed change in size to a greater extent than others. 3-D analysis of HVc lesions provides a clearer view of the size and shape of the lesion site within the target nucleus and relative to the surrounding tissue. Used in conjunction with 2-D analysis, the 3-D method will aid investigations of the song system and contribute to the understanding of its regulation by hormones. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Birdsong; Bird; Volume reconstruction; Lesion
1. Introduction Songbirds possess a hormone-sensitive network of interconnected brain nuclei that regulates song learning, perception and production (Nottebohm et al., 1976; Nottebohm, 1980, 1981), as depicted in Fig. 1. All seasonally breeding songbirds examined exhibit dramatic seasonal changes in the size of these nuclei (e.g. Nottebohm 1981; Tramontin and Brenowitz, 2000; Ball, 2000). Increases in size of the song control nuclei during the breeding season (spring) are associated with increased song stereotypy and duration (Tramontin and Brenowitz, 2000). The telencephalic song regions HVc (also known as the higher vocal center) and the robust nucleus of the archistriatum (RA) increase in volume by as much as 188 and 135%, respectively, in the spring (Smith, 1996; Tramontin and Brenowitz, 2000).
* Corresponding author. Tel.: /1-206-543-3356; fax: /1-206-6853157 E-mail address: [email protected] (G.E. Bentley).
Study of seasonal plasticity in the song control system of songbirds has largely focused on changes in volumes of the song control nuclei (for review see Tramontin and Brenowitz, 2000). The volumetric changes of song control nuclei are caused by changes in neuron number, size, or spacing, depending upon the nucleus. In whitecrowned sparrows (Zonotrichia leucophrys ) and song sparrows (Melospiza melodia ), fewer neurons are present in HVc of fall birds as compared to spring birds (Smith et al., 1995, 1997; Tramontin and Brenowitz, 2000). Neurons in RA are smaller and more closely spaced in the fall as compared to the spring. Measuring the volume of a brain nucleus under different physiological or developmental conditions provides useful information on changes in overall size (see Fig. 2 for example). This approach, however, does not provide information on whether such changes occur equally throughout the nucleus, or whether there are differential changes in portions or subdivisions of that nucleus. In the song control system, for example, functional and morphological distinctions have been demonstrated between the lateral and medial portions of
0165-0270/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 0 2 7 0 ( 0 2 ) 0 0 2 4 4 - 3
76
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
structions to the contexts of seasonal plasticity and identification of lesion sites will be considered.
2. Materials and methods 2.1. Tissue sections
Fig. 1. Schematic sagittal drawing of the songbird brain showing projections of the major nuclei in the song system. The motor pathway (solid black arrows) controls the production of song. The solid black arrows indicate the descending pathway from HVc in the neostriatum to RA in the archistriatum. From RA, the pathway projects to the vocal motor nucleus nXIIts. The open (white) arrows indicate the anterior forebrain pathway that is essential for song learning. This pathway indirectly links HVc to RA, via the basal ganglia homolog area X, DLM in the thalamus, and lMAN in the neostriatum. lMAN also projects to area X. DLM, medial portion of the dorsolateral nucleus of the thalamus; lMAN, lateral portion of the magnocellular nucleus of the anterior neostriatum; RA, robust nucleus of the archistriatum; V, ventricle; X, area X; nXIIts, tracheosyringeal part of the hypoglossal nucleus. Adapted from Brenowitz et al. (1997).
the nuclei MAN and area X. In area X, immediate-early gene expression differs between the lateral (l) and medial (m ) portions with social context (Jarvis et al., 1998), and the distribution of melatonin receptors differs between these portions according to photoperiodic status (Bentley and Ball, 2000). As social cues and melatonin are known to influence seasonal changes in volumes of song control nuclei (Tramontin et al., 1999; Bentley et al., 1999), it seems feasible that these environmental cues might affect the change in size and structure of one part of a song nucleus more than another. Analyzing the three-dimensional (3-D) structure of song nuclei under different conditions could provide insight on this issue. Such information, combined with the known connectivity of the song control pathway and hormone receptor distribution, would shed light on the pathways involved in developmental and seasonal plasticity in this hormone-sensitive model system. 3-D analysis could also be of value in guiding and evaluating the use of lesions to study the functions of song nuclei (e.g. Scharff et al., 1998; Brainard and Doupe, 2000; Brenowitz and Lent, 2001). Traditional 2D reconstructions of lesioned nuclei suffer from similar limitations as discussed above. It would be informative to reconstruct 3-D images of lesion sites within a target nucleus to determine the extent of lesion. Rotation of the 3-D image would allow one to see exactly where the lesion has been made relative to inputs/outputs and hormone receptor distribution. We describe here the use of customized computer software to produce 3-D images of song control nuclei. The application of 3-D recon-
Tissue sections from previously-published studies were used for the 3-D reconstruction (Smith, 1996; Brenowitz and Lent, 2001). Brains were collected after perfusion /fixation, sectioned coronally at 50 mm, and Nissl-stained. Although Nissl stain was used in this study, any histological marker that provides clearlydefined detail of the boundaries of the area(s) of interest would be suitable. Images of seasonal changes in song control nuclei were constructed from male brains of spotted towhees (Pipilo maculatus ). Brains from individuals in different reproductive conditions were used to illustrate the morphometric changes of song control nuclei in 3-D. Images of HVc lesions were constructed from brains of Gambel’s white-crowned sparrows (Zonotrichia leucophrys gambelii ). 2.2. 3-D image construction Images were captured using NIH Image version 1.62 (Wayne Rasband, NIH) on an Apple Macintosh computer linked to a video camera and light microscope. Sequential images through each nucleus of interest were captured and 3-D images were constructed using ‘MacSurfer’ software version 1.3.2 (S. Kurkin, Japan Science and Technology Corporation, Department of Physiology, Hokkaido University School of Medicine, Japan. E-mail: [email protected]). The following link may also be used to download the software free of charge: http://133.87.71.180/MacSurfer.sit. The most critical consideration in capturing 2-D images for use in 3-D image construction is the correct alignment of sequential images. The MacSurfer program interpolates between sequential images, so misalignment can result in distortion of the 3-D image as a whole. For this study, the first section containing a song control nucleus of interest was captured and the rest of the sections through that nucleus were aligned relative to the first section. Alignment was achieved by using the ‘paste control’ function in NIH Image, in conjunction with the ‘Xor’ transfer mode in either the ‘scale math’ or ‘live paste’ options. This procedure enabled us to lay a semi-transparent image of the second section over an opaque image of the first section. Alignment was based on several landmarks in each section, including the midline, the nucleus of interest and the outline of the telencephalon. Other landmarks that were used depended upon the area of the brain that was sectioned.
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
77
Fig. 2. Thionin-stained transverse sections of HVc (A1, A2); RA (B1, B2); and area X (C1, C2) in breeding (A1, B1, C1) and non-breeding (A2, B2, C2) male spotted towhees. All sections are from the central portion along the rostral /caudal axis of each nucleus. Arrowheads indicate borders of song nuclei. Bars /0.5 mm. Adapted from Smith (1996).
A similar alignment method can be employed using Adobe Photoshop† software. Once sections were aligned, an image for each section was saved in TIFF format. The TIFF files were imported into the MacSurfer program, and an outline of the area of interest was drawn using a computer mouse. A computer-generated 3-D image was created from the stack of outlines created with the mouse. The entire process of image capturing, alignment and 3-D construction was repeated for each nucleus shown.
3. Results 3.1. Photoperiod-induced changes in the song control system 3-D comparison of the song control system of spotted towhees on short day lengths (non-breeding) and long day lengths (breeding) depicts areas of song nuclei that undergo volumetric changes as a result of changes in
hormonal status associated with changing photoperiod (Fig. 3A). The caudal view is presented here instead of the rostral view to enable the caudomedial extension of HVc to be viewed more clearly. The dorsal view was created by the MacSurfer program using the same information as for the caudal view; the image was rotated through 908 using one of the program options. It is clear from these figures that the song control nuclei depicted do not change size uniformly throughout their structure. Looking at each nucleus that was reconstructed in turn, 3-D image construction allows us to describe morphological characteristics that are not easily discernible from 2-D images: Area X : In the transition from short to long days, area X grows in all planes, but more in the medial-lateral plane that in other planes. There is proportionally more growth in m area X and the ‘neck’ joining m area X and l area X than in l area X alone. HVc: In its most lateral portion (main corpus), HVc grows to some extent in the dorsal /ventral plane.
78
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
Fig. 3. (A) Song control nuclei of male spotted towhees represented in 3-D. Song nuclei from non-breeding (green) and breeding (red) individuals are included in a single image to facilitate comparison. Caudal and dorsal views allow for easy viewing of morphology. Medial and lateral area X are labeled m and l , respectively. (B) Rostral and caudal views of unilateral lesions directed at HVc in a Gambel’s white-crowned sparrow (Bird A). Green color depicts the target HVc, red depicts the intact HVc. Arrows indicate the injection sites, and the total lesioned area is represented in purple. (C) As B, but for a separate individual (Bird B).
Overall, it increases a great deal in its rostral /caudal plane. The caudomedial extension of HVc also projects further ventrally as it extends caudally, curling over on itself at its trailing edge. RA : Growth of RA occurs in all planes, but to a great extent in the medial/lateral plane. As it extends caudally, it also curves medially.
3.2. 3-D reconstruction of lesions directed at HVc Fig. 3B and C consist of rostral and caudal views of 3D reconstructions of lesions directed at HVc in Gambel’s white-crowned sparrows. HVc was lesioned unilaterally with two injections of 1% racemic N -methylaspartic acid (NMA) in phosphate-buffered saline.
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
Lesions were directed at both the rostral main corpus and caudomedial extension of HVc (see Brenowitz and Lent, 2001 for full details). The two sites of insertion of the injection micropipette can clearly be seen in Fig. 3B and C. It is apparent from this figure that in both birds the NMA spread from the two injection sites and joined them together, creating a single large lesion. The first lesion attempt (Bird A; Fig. 3B), destroyed part of the lateral portion of the main corpus of HVc. Lesioned HVc was overall much smaller than on the unlesioned side. In the second lesion attempt (Bird B; Fig. 3C), the rostral part of the caudomedial extension of HVc was destroyed. The size and shape of the main corpus of HVc appeared to be unaffected by the lesion.
4. Discussion This study describes a method by which the 3-D characteristics of song control nuclei can be analyzed. To convey the same information using 2-D images would require multiple figures that are not as easily interpreted. It is clear from the results that the song control nuclei depicted do not change size uniformly throughout their whole structure. This observation could have implications for determining the relative contributions of different environmental cues to plasticity in the song control system. For example, changes in photoperiod alone might affect subdivisions of a nucleus that are not affected by social cues alone, and vice-versa. Importantly, failure to observe an overall difference in volume of a nucleus between two treatment groups might belie significant changes in subregions, possibly as a result of heterogeneous hormone receptor distribution or afferent/efferent projections. A similar principle is true for studies involving lesions of song control nuclei. The 3-D analysis provides a more informative picture of lesion location, both within a song control nucleus and relative to the surrounding neural tissue. The ability to pinpoint lesions in 3-D to exact locations within the target nucleus creates potential for greater understanding of the relationship between its topography and behavioral outputs. When embarking on a study of this type, the investigator must decide what section thickness to use. The optimal thickness will depend upon the system studied and clarity of staining. There tends to be a tradeoff between section thickness and clarity of delineation of the area in question. Ideally, thin sections allow the most accurate measurement of the volume occupied by a brain nucleus of interest. In practice, however, the borders of brain nuclei are more clearly delineated in thick sections. As a result, sampling intervals of between 40 and 120 mm tend to be used for measurement of the
79
song control nuclei. In the present study we use alternately-collected 50 mm sections (sampling interval of 100 mm). Seasonal changes within the song system can easily be detected with this sampling interval, as is shown in Fig. 3A. In summary, the 3-D reconstruction approach has several features that allow for finely-detailed investigation of manipulations of the song control system, and can complement 2-D methods of analysis. Although this 3-D method is not in itself quantitative, it is an excellent aid to lesion studies, and provides additional information for studies of seasonal regulation of the song control system. Increasingly, it is becoming apparent that subdivisions of a nucleus process environmental stimuli in different ways, e.g., m area X vs. l area X (Jarvis et al., 1998; Bentley and Ball, 2000). We propose that if adopted for routine use, this 3-D approach will provide new information about the contribution of environmental cues to morphometric changes in the subdivisions of song control nuclei. In addition, these methods are well-suited for studying developmental changes in the song system, as well as other forms of adult plasticity such as ongoing neuronal recruitment. This approach could also be applied to other vertebrate neural systems that undergo morphometric plasticity at different life stages, such as during development or senescence. The same approach would also be suitable for comparison of, for example, the sexually dimorphic nucleus in rats, the preoptic area in birds, rats and lizards and the medial amygdala in rats.
Acknowledgements We thank Glen MacDonald and Karin Lent for their technical assistance. Supported by NIH grant MH53032 (to E.A.B.) and the Virginia Merrill Bloedel Hearing Research Center.
References Ball GF. Neuroendocrine basis of seasonal changes in vocal behavior among songbirds. In: Hauser M, Konishi M, editors. Neural mechanisms of communication. Cambridge MA: MIT press, 2000:213 /53. Bentley GE, Ball GF. Photoperiod-dependent and -independent regulation of melatonin receptors in the forebrain of songbirds. J Neuroend 2000;12:745 /52. Bentley GE, Van’t Hof TJ, Ball GF. Seasonal neuroplasticity in the songbird telencephalon: a role for melatonin. Proc Natl Acad Sci USA 1999;96:4674 /9. Brainard MS, Doupe AJ. Interruption of a basal ganglia-forebrain circuit prevents plasticity of learned vocalizations. Nature 2000;404:762 /6. Brenowitz EA, Lent K. Afferent input is necessary for seasonal growth and maintenance of adult avian song control circuits. J Neurosci 2001;21:2320 /9.
80
G.E. Bentley, E.A. Brenowitz / Journal of Neuroscience Methods 121 (2002) 75 /80
Brenowitz EA, Margoliash D, Nordeen KW. An introduction to birdsong and the avian song system. J Neurobiol 1997;33:495 /500. Jarvis ED, Scharff C, Grossman MR, Ramos JA, Nottebohm F. For whom the bird sings: context-dependent gene expression. Neuron 1998;21:775 /88. Nottebohm F. Testosterone triggers growth of brain vocal control nuclei in adult female canaries. Brain Res 1980;189:429 /36. Nottebohm F. A brain for all seasons: cyclical anatomical changes in song control nuclei of the canary brain. Science 1981;214:1368 /70. Nottebohm F, Stokes TM, Leonard CM. Central control of song in the canary, Serinus canarius . J Comp Neurol 1976;165:457 /86. Scharff C, Nottebohm F, Cynx J. Conspecific and heterospecific song discrimination in male zebra finches with lesions in the anterior forebrain pathway. J Neurobiol 1998;36:81 /90.
Smith GT. Seasonal plasticity in the song nuclei of wild rufous-sided towhees. Brain Res 1996;734:79 /85. Smith GT, Brenowitz EA, Beecher MD, Wingfield JC. Seasonal changes in testosterone, neural attributes of song control nuclei, and song structure in wild songbirds. J Neurosci 1997;17:6001 /10. Smith GT, Brenowitz EA, Wingfield JC, Baptista LF. Seasonalchanges in song nuclei and song behavior in Gambel’s whitecrowned sparrows. J Neurobiol 1995;28:114 /25. Tramontin AD, Brenowitz EA. Seasonal plasticity in the adult brain. Trends Neurosci 2000;23:251 /8. Tramontin AD, Wingfield JC, Brenowitz EA. Contributions of social cues and photoperiod to seasonal plasticity in the adult avian song control system. J Neurosci 1999;19:476 /83.