大鼠脑图谱

Relaxing holiday at Club Med, Nea Makri, Greece (July 2004) George Paxinos Prince of Wales Medical Research Institute The University of New South Wales Sydney, Australia g.paxinos@unsw.edu.au www.powmri.edu.au/staff/paxinos.htm Charles Watson Division of Health Sciences Curtin University of Technology Perth, Australia c.watson@curtin.edu.au Relaxing holiday at Club Med, Nea Makri, Greece (July 2004) We dedicate this book to Kosta Theodore Paxinos and Anwen Angharad Williams. Preface In the first four editions of this atlas, we relied on a coronal section set that had some significant limitations; the section frequency proved over time to be too wide, sections did not always appear at regular intervals, and a few damaged sections had been replaced with sections from another brain. The fifth edition is based on a new coronal set which includes 161 sections from a single brain at regular 120µm intervals. This edition of the atlas is not simply an incremental improvement on the previous edition, but a completely new and far more comprehensive map of the rat brain. Although the fifth edition features a different coronal section set, readers can be assured that the stereotaxic coordinates in the new atlas match those in previous editions. We have increased the scope of the atlas by incorporating new anatomical concepts where appropriate, and have once again delineated and named some areas not previously recognised. Over the past decade our efforts at mapping the brain have been greatly enhanced by the availability of sections stained with a wide range of different chemical markers. A further contribution to the accuracy of our maps has been the knowledge we have we have gained from comparative neuroanatomical studies. One of us (GP) has published atlases of the human (Paxinos and Huang, 1995; Mai et al. 2004), monkey (Paxinos et al. 2000), and mouse brain (Paxinos and Franklin, 2004), and both of us are part of a team that is in the final stages of preparing an atlas of the chicken brain (Puelles et al. in press). Each of these atlas projects has provided us with new insights that have enhanced our ability to interpret the anatomy of the rat brain. Acknowledgements We thank Hongqin Wang for outstanding technical assistance from the inception to the completion of the project. We are indebted to Lewis Tsalis for speed, accuracy and brilliance in construction of diagrams and designing this book; to Paul Halasz for his imagination in construction of the CD-Rom; to Julia Tsalis for accurate labelling of diagrams; to Hongmei Liu for technical assistance, and to Yvette Paxinos for the cover design. We acknowledge with gratitude the intellectual contribution to delineations made by Yuri Koutcherov (hypothalamus), Konrad Talbot (hippocampus), Nicola Palomero-Gallagher and Karl Zilles (cortex), Brent Vogt (cingulate cortex), Jan Vogt (cerebellum, precerebellar nuclei and vestibular nuclei), George Alheid (basal forbrain), Pascal Carrive (periaqueductal gray and hypothalamus), Glenda Halliday (substantia nigra and VTA), Ellen Covey and Manolo Malmierca (auditory system), Joel Elmquist (hypothalamus), Ann Goodchild and David Hopkins (rostroventrolateral and caudoventrolateral medulla), Jose DeOlmos (amygdala), Henk Groenewegen (thalamus), Joseph Travers (orofacial motor nuclei), John Mitrofanis (zona incerta), Pierre-Yves Risold (septum), Miklos Palkovits (paralemniscal nuclei), Harvey Karten (pretectal area), Chip Gerfen (basal ganglia), Terry Furlong (hypothalamus), Jean Buettner-Ennever (oculomotor nuclei), Marina Bentivoglio (parafascicular nucleus). George Paxinos acknowledges the support he has received from the Australian National Health and Medical Research Council (he holds an NHMRC Principal Research Fellowship), as well as assistance from the Clive and Vera Ramaciotti Foundation, the Rebecca Cooper Foundation, and the Brennan Foundation. We have appreciated the intelligent and enthusiastic support from our Elsevier editor Johannes Menzel. His patience and consideration have made a real difference to the successful completion of this project. We also thank Maureen Twaig and other Elsevier staff for their willingness to help in solving production problems. Features of the Fifth Edition • 161 coronal diagrams based on a single brain • Diagrams spaced at constant 120 µm intervals giving scientists the most comprehensive and convenient atlas of the rat brain • The most accurate stereotaxic reference system available • Outlines of figures and brain structures in blue, but labels and leader lines in black for increased clarity of delineations • All delineations re-examined in the light of recent findings • Delineations of brain structures have been made with reference to sections stained for Nissl substance, AChE, parvalbumin, calbindin, calretinin, SMI-32, tyrosine hydroxylase, and NADPH diaphorase (Paxinos et al. 1999a,b) • Extensive use was made of reference works, including the third edition of The Rat Nervous System (Paxinos, 2004) and other recent neuroanatomical literature • Spinal cord drawings from the atlas of Molander and Grant(1995) • Diagrams available on CD-ROM for printing. Introduction There are many reasons why the rat is the most commonly selected subject for research in mammalian neuroscience. First, rats are the right size: neither too small for accurate stereotaxic localization of discrete brain areas, nor too large for cost-effective laboratory management. Second, rats are generally hardy animals and are resistant to infections. Third, a number of inbred strains are available commercially, so that animals of consistent size can be used for stereotaxic procedures. When the first edition of The Rat Brain in Stereotaxic Coordinates was published in 1982, it was the first atlas to be based on the flat skull position. It offered a choice of bregma, lambda, or the midpoint of the interaural line as the reference point. Although the coordinates were developed from study of adult male Wistar rats with weights ranging from 270 to 310 g, the atlas can be successfully used with male or female rats, with weights ranging from 250 to 350 g (Paxinos et al., 1985). With each new edition of the atlas, we have attempted to improve the accuracy of our delineations and have incorporated new findings on brain anatomy. However, our work has been hampered by the fact that our original series of coronal sections suffered from a number of limitations. First of all, our primary sections series showed sections at 0.5 mm intervals, which is insufficient to adequately represent all major structures in the brain for modern research purposes. Although we later attempted to better illustrate some areas by using some intervening sections, we could never fully compensate for the wide section interval in the primary series. In addition, we lost some sections in some areas of the brain and were forced to interpolate sections from another brain to compensate for the missing sections. We were aware that the only real solution to these problems was to replace the coronal section series with a new section set based on shorter intervals and with all sections taken from the one brain. The new coronal section series is presented in the present (fifth) edition of the atlas. It shows diagrams of sections taken at regular intervals of 0.12 mm. Having constant intervals between the sections shown in the atlas diagrams eliminates one of the annoying features of many brain atlases – the fact that when the reader turns a page they do not know how far they have advanced along the prime axis. All sections are from the one brain. The sections in our new coronal series were stained with cresyl violet or with methods to demonstrate AChE or NADPH diaphorase because we found that these three methods were compatible with using fresh (unfixed) tissue, a requirement for deriving an accurate stereotaxic grid. However, we consistently used other markers to confirm our delineations (Paxinos et al., 1999a, Paxinos et al., 1999b). We have once again been greatly assisted by the suggestions of many colleagues in the delineation of structures. We welcome further advice that might improve the accuracy of our diagrams in the future. Please email us on g.paxinos@unsw.edu.au or c.watson@curtin.edu.au. The present book will be followed by a comprehensive publication, which will include accompanying photographs and revised diagrams of sagittal and horizontal sections. Methods A fresh brain from a male 290 g Wistar rat was frozen, and coronal sections were cut at 40 mm thickness. The sections were cut at right angles to the horizontal plane joining bregma and lambda. Stereotaxic surgery We placed an anesthetized rats in a Kopf small-animal stereotaxic instrument, and the incisor bar was adjusted until the heights of lambda and bregma were equal. This flat-skull position was achieved when the incisor bar was lowered 3.3 ± 0.4 mm below horizontal zero (Table 1). Because the point of intersection of the lambdoid and sagittal sutures is variable, we have chosen to define lambda as the midpoint of the curve of best fit along the lambdoid suture (see skull diagram). This redefined reference point is considerably more reliable than the true lambda (the point of intersection of the sagittal and lambdoid sutures), and it is located 0.3 ± 0.3 mm anterior to the interaural line. We defined bregma as the point of intersection of the sagittal suture with the curve of best fit along the coronal suture. When the two sides of the coronal suture meet the sagittal suture at different points, bregma usually falls midway between the two junctions. The anteroposterior position of bregma was 9.1 ± 0.3 mm anterior to the coronal plane passing through the interaural line, but for the brain represented in this atlas bregma is deemed to lie at 9.0 mm. The top of the skull at bregma and lambda was 10.0 ± 0.2 mm dorsal to the interaural zero plane. To confirm the stereotaxic orientation of sections in the brain used for this atlas, reference needle tracks were made perpendicular to the horizontal and coronal planes. One horizontal needle insertions perpendicular to the coronal plane were made from the posterior of the brain at 4.0 mm above the interaural line and was 2.0 mm lateral to the midline. The reference track from the horizontal needle appears as a small hole in coronal sections. Following surgery, the rat was decapitated and the whole head frozen on dry ice. The frozen skull was then prised off the frozen brain, and the brain was carefully mounted on the stage of microtome so that the sections would be cut in the coronal stereotaxic plane. Every third section was used for preparation of the atlas diagrams, so that the interval between atlas diagrams is 0.12 mm. Exceptions to this rule are found in the region rostral to the rostrum of the corpus collosum (Interaural AP 11.28) and in the region of the medulla caudal to the inferior olive (Interaural AP -5.76 mm). In these two regions, sections were selected for presentation in the atlas at 0.24 mm intervals. Finally, the olfactory bulb is depicted at only three representative levels. Histological methods The ‘atlas’ sections were stained with either cresyl violet or for the demonstration of AChE on an alternate basis, so that cresyl violet sections are 0.24 mm apart and AChE sections are also 0.24 mm apart. The two sections that intervene between each ‘atlas’ section were stained with cresyl violet or for the presence or AChE or NADPH diphorase, according to the following sequence: 1. Cresyl violet ‘atlas’ section 2. AChE intervening section 3. NADPH intervening section 4. AChE ‘atlas’ section 5. Cresyl violet intervening section 6. NADPH diaphorase intervening sections This sequence was repeated throughout the series of coronal section. This arrangement ensures that every ‘atlas’ section is accompanied by two adjacent sections, each of a different stain. For example, the ‘atlas’ AChE section described as number four above is preceded by an NADPH diaphorase section and followed by a cresyl violet section. This arrangement gave us maximum information for each ‘atlas’ section from the three stains. Staining was carried out on the same day as section cutting. All sections, whether ‘atlas’ or intervening, were photographed on 4"x5" black and white negatives and printed on 36"x24" photographic paper. Each ‘atlas’ section was then covered with a sheet of ‘Mylar’ tracing film and outlines of structures were drawn in pencil. The final pencil drawings were scanned and then digitized using Adobe Illustrator. Quality of Sections In some cases, the sections were slightly stretched or compressed in the process of cutting and mounting on slides. We have compensated for this by constructing diagrams which represent, as best we can judge from the study of adjacent sections, the original shape of the brain section. In the worst cases, the ‘atlas’ section was so badly damaged that we have taken our drawing from an adjacent section. Cresyl Violet Staining Slides were immersed for 5 min in each of the following: xylene, xylene, 100% alcohol, 100% alcohol, 95% alcohol, and 70% alcohol. They were dipped in distilled water and stained in 0.5% cresyl violet for 15-30 min. They were differentiated in water for 3-5 min and then dehydrated through 70% alcohol, 95% alcohol, 100% alcohol, and 100% alcohol. They were then put in xylene and coverslipped. To make 500 mL of 0.5% cresyl violet of about pH 3.9, mix 2.5 g of cresylecht violet (Chroma Gesellschaft, Postfach 11 10, D-73257, Kongen, Germany, Fax number: 49-7024-82660), 300 mL of water, 30 mL of 1.0 M sodium acetate (13.6 g of granular sodium acetate in 92 mL of water), and 170 mL of 1.0 M acetic acid (29 mL of glacial acetic acid added to 471 mL of water). Mix this solution for at least 7 days on a magnetic stirrer, then filter. AChE Histochemistry The method for the demonstration of AChE followed the procedures of Koelle and Friedenwald (1949) and Lewis (1961). Slides were incubated for 15 h in 100 mL of stock solution (see below) to which had been added 116 mg of S-acetylthiocholine iodide and 3.0 mg ethopropazine (May & Baker). The slides were rinsed with tap water and developed for 10 min in 1% sodium sulphide (1.0 g in 100 mL of water) at pH 7.5. They were then rinsed with water and immersed in 4% paraformaldehyde in phosphate buffer for 8 h, and then allowed to dry. Subsequently, they were dehydrated for 5 min in 100% alcohol, then immersed in xylene and coverslipped with Permount. The stock solution was a 50 mM sodium acetate buffer at pH 5.0 which was made 4.0 mM with respect to copper sulphate and 16 mM with respect to glycine. This was done by adding 6.8 g of sodium acetate, 1.0 g of copper sulphate crystals, and 1.2 g of glycine to 1.0 L of water and lowering the pH to 5.0 with HCl. We found that fresh, unfixed tissue from the frozen brains showed a substantially stronger reaction for both stains than tissue fixed with formalin, paraformaldehyde, glutaraldehyde, or alcohol. NADPH diaphorase The sections were washed in phosphate buffer for 10 minutes and incubated in 10 ml of a phosphate buffer solution containing 0.0125% nitroblue tetrazolium, 0.05% NADPH, 0.5% Triton X-100, and 1 mM magnesium chloride. The pH of the solution was adjusted to 7.6. The sections were incubated at 4ºC for 48 hours. The incubation was stopped with a wash in phosphate buffer. Photography and drawings Photography The photographs of stained brain sections were taken with a Nikon Multiplot macrophotographic apparatus using 4"x5" Kodax Plus X film. High contrast paper was used to print the photographs of Nissl sections, whereas lower contrast paper was used to print the photographs of AChE and NADPH sections. Drawings Drawings, which later formed the basis of the figures, were made by tracing the photographs of sections. We drew only the right side of each section and derived the outline of structures on the left side by mirror image construction using Adobe Illustrator. Fiber tracts in the drawings are outlined by solid lines, and nuclei and cell groups are outlined by broken lines. In general, each abbreviation is placed in the center of the structure to which it relates; where this is not possible, the abbreviation is placed alongside the structure and a leader line is used. The abbreviations for fiber tracts and fissures are almost always positioned on the left side of the figure, and the abbreviations for nuclei and other cell groups are generally positioned on the right side. The outlines of the ventricles and aqueduct are filled in with solid color. Stereotaxic Reference System The stereotaxic reference system is based on the flat skull position, in which bregma and lambda lie in the same coronal plane. Two coronal and two horizontal zero-reference planes are referred to in these drawings. One reference coronal plane cuts through bregma and the other cuts through the interaural line. Similarly, one horizontal plane is at the level of bregma on the top of the skull and the other is at the level of the interaural line. Lambda is usually located 0.3 mm anterior to the interaural line, and it can be used as an alternative reference point in conjunction with the dorsoventral coordinate of bregma. The position of the stereotaxic reference points and planes are indicated on the skull diagram. The stereotaxic reference grid shows 0.2 mm intervals. Drawings of coronal brain sections In each of the coronal drawings, the large number at the bottom left shows the anteroposterior distance of the section from the vertical coronal plane passing through the interaural line. The large number at bottom right shows the anteroposterior distance of the plate from a vertical coronal plane passing through bregma. Note that these two coronal planes are 10 mm apart, so the two numbers on any one plate add up to 10 mm. The small numbers on the left margin show the dorsoventral distance from the horizontal plane passing through the interaural line. The numbers on the right margin show the dorsoventral distance from the horizontal plane passing through bregma and lambda on the surface of the skull. The numbers on the top and bottom margins show the distance of structures from the midline sagittal plane. Skull Diagram Dorsal and lateral views of the skull of a 290 g Wistar rat. The positions of bregma, lambda and the plane of the interaural line are shown above the lateral view. The distance between the horizontal plane passing through the interaural line is shown on the right of the lateral view. The distance between the incisor bar and the horizontal plane passing through the interaural line is shown on the left of the lateral view. Lambda (midpoint of the curve of best fit along the lambdoid suture) is 0.3 mm anterior to the coronal plane passing through the interaural line. 290 g Male Wistar Interaural Line Incisor Bar Lambda Lambda Bregma Bregma 10.0 mm 3.3 mm 9.0 mm Accuracy of the stereotaxic coordinates In almost all cases, the potential error in defining the position of any point in the brain is less than 0.5 mm. Although we used medium-sized (average 290 g) male Wistar rats in the construction of this atlas, we recognize that researchers often use animals of different sex, strain, and weight. Because of this, we have estimated the error that may occur if this atlas is used with female Wistar rats, male hooded (Long Evans) rats, male Sprague Dawley rats of 300-g weight, juvenile (180 g) Wistar rats, and mature (436 g) Wistar rats. The results of these estimations are shown in Table 1 (reproduced from Paxinos et al., 1985). I