I:v.. "'.t. ‘ c .o. .‘_..... -. . - ‘ ‘ . ‘ N: - ‘ . I - ‘1 . '1’..""¢""‘.""E r~r" ‘-:\a.'£:fc:-:-:334o¢—Q.«_~.unfinoovnrvnuon-ch-gw'oy ‘ ‘ . ~ . l ‘ - é. - . ‘ o - . - . - - _ v " -’ . I o . ‘ A . — - . - . § " 1 - ~ \ ' Q ~\ . . . o . . o . . . . . ~ -‘- - - - . ‘ . . - Q .‘ - ‘ u C . C ‘ O C . - 9 . n - . - Q”- . . ‘. s , , .... R .l I 5- - r .- . - i g i A. g Q _ ‘ ~ . . . . - - - . \ ~ - - . . y . ' o - . - . . ‘ - . ‘ ‘ . “-.~. ,' - ‘ . - . . - ' ' ‘ ‘ ' - - 3 a"; -‘. . -. ‘ ° . 'v u p--.- - ' ' . c 7- . ‘ - . o. - ‘ . '. ‘ . ‘ . .... - . -- . -|- :_- . - A . » - v ' ‘ - ~~- ' . ~ . - . n‘v Q . h I -‘ 0 '.V . ' ' . , . _..- - .c - 0‘ '. ' ‘ ‘ ‘ ' ‘. o o s . u - _“ - ‘ ' o u - o R - . . u I ‘ . - -.- - ' . . ‘i ‘c 0.... I. O -... - . ' ' - ‘ ‘ o . . - ‘ - ._~ - . . ‘7 ¢ - . - . - 0 ~ O -. — -- - ..- . - . . , .- ‘ .._ . . ‘ . . ~ . § -- s a . .._ .' .‘ _ . . n o - ' ‘ . - -' I A - ‘ ‘ ° ‘ ' ‘.°-' - '.' ‘ ‘ °- ‘. .‘- ‘ V .- ‘. a, - - ‘ ' _ ,. '. . - . .‘.‘. ... . < I ' ’ I . . . . ~_ - ».-.-. -_ .1. .-_ ‘- - -...- - . . . . o o . - - - - . - .I ' ' . . s V _ - u - ‘ - q o ‘ -_ . ‘ _- '. 4 -...~ - _ .:._‘ j. ,:, -.-: _.‘- A . . AI. . . n . . . . . . . a . . . _‘ . s ‘ , , . _ . . ‘ -. ‘ ._. . ..--... '7‘,‘.’.“ , ., . . ' I ‘ ' ' ‘ - ' ' ' .‘ ' ~‘. . - u - . V ' ~ - . '. ... .' g .. ' a ,‘ A , .. ., - - . ' ~ ' . . . V ' ... 0‘. I“- ‘ .9 ‘ ' - u I . . I O - - 0 WI. - . I . ' ' R . ,. . a ‘ ".. -. ." g' . . . V -‘ ..- ' . o . g . V . . o - - a u . - - - _ - . -- >- . .‘ I ‘ - ‘ . I ..l .I ‘ ' . V..... O "-'c A“‘> . ‘ ‘ . . ' ., . ,. .I ."- - I‘u‘. C>~'- v .- ' R , . .‘ . o‘.-- .- ... . . ‘ '. . . . ~ ' o . I > ‘ ‘ n - c I C O . I g . ' ‘ V ‘ ' ‘ ' Q . , ’ ' , . . - - . . . ' _ q . ... A 0-0 ' . - f. .. .. ---..‘ _ ‘ ’ . . . ... ’.'- 0.7 o - . -. . ‘-l-. o 0‘ ‘ . I I ‘ . . -' _ . - u . v - _ . . . _ - . - I V ' o »- ~ - . ° '. o o o s — . - a o - . - - . ~ . - - ‘ - 1~ - - . . , . . , '~ _ ' . ' K n . ' ‘.. -' .-. - . T‘."'_‘_"_ ' _‘ ' . . .- A-o o o - — - o - . . . a - - - - I . ~ . . ' ' ‘ '- ". 'I' “' ’ ' '.' .n . "-’.--'n - - . .‘_.. , . r U - D- O --I . . ‘ . . . - I . _ C . o " - . ‘.'.'. . . - a D ' ' . ' n .‘q- . o . --. o. - ~ . . o a . , < c_~ - _q_ -<. ' a .. ‘ .._.. .... _ . V . - _ - . . I C v -. ... -"-~ - C. . o . _ . - ._4 - ~.. .-. . . - - . ‘ - o . - . .‘.'.‘. .... ... . ..._. ._» - . . - o - . . x . - n .0 - f‘ .. .-.> _.. ..., - ‘, . - . - ' . ‘ V... n. -.‘n . ‘I‘ ‘- . - . I . Q I ‘ p ' . C C - O \ Q to ‘ ’ . - - o - - a - ' ‘ ~ a . . - o o .O . . 9 ‘ U ' - . l ‘ . "O ' . . ' '~ .‘C'-‘—'. - - ‘ 7 . ‘ . ‘ ' I. ‘ . u 9 -_ - 0 9 - _..Q _ ‘ . . . _‘ :‘ ‘ — O . . - - - ~ . . 6'.‘ -. . _ _-.‘. 0 v : * -_ ' . ' ‘ c ' ‘ . . -.. ' —-. ‘ ‘ ~ . ‘ . ;' - 9 u i ‘ - . O . O C. - . ‘..I ~ .... C . g o - rttn "... o .- - _~ - . . -. ' . . g a ' " v n _ A... o A- . “ - — o . - ~ - , .' -' ‘- - - o . D ‘ d O O . u . c . a "' - ‘ ‘ - - . ‘ . . . . - o . . o . . - . _ _ . ‘ - . ‘ - U C ‘- . . - . ‘ . ‘ .0 4 ~ g . D g- ‘ .- Q I O - - 9. A '. - ‘ -' - . . ' . . . _ - u o _'_ o o .' ' . .- : O O - C C .V l o a . . . o v ' - . D , - n u .' . . - . . -. ' . . .‘_ . _ - . .0 . . . ¢ 0 . - - ‘ O - h - - o . K ‘ v ' ~ . . a ' - C a - - C - I I . . . , ' . .. , ‘ . '~5 0 ' _ ,, .7 . - . . . -‘ ‘ I. . - MICHIGAN STATE UNIVERSITY ‘ r ‘ _ _ ROBERT c.‘ SWITZER m _{ . LIBR:' RY Michigan 5"” U nivcfii‘Y ? smiths BY '7’ - i ., nous a my "' : {aux mm m. E IBRARY BINDER. i‘ sn:umn.mcm1 l i \E. .< ’— -- ABSTRACT QUANTITATIVE RELATIONSHIPS OF CELL POPULATIONS IN OLFACTORY BULBS OF MAMMALS BY Robert C. Switzer III Qualitative comparisons of olfactory bulb structure in many different mammals indicated a need to define and describe characteristics of cell pOpulations and corresponding layers quantitatively. Four species of the carnivorous family of marsupials, the Dasyuridae, were studied because of their prominence and good definition of particular features and the large size range of their bodies and olfactory bulbs. Selected sections from brains stained for Nissl substance were projected and olfactory bulb layers traced. A compensating polar planimeter was used to determine the areas of the traced layers. Layer volumes were calculated from areas plotted according to section position. Planimetry was performed on plots of numbers of cells versus section number to obtain the total number of cell types. Results indicate that each olfactory bulb layer volume remains a constant function of increasing bulb volume as do the numbers of each cell type. The density of each cell type diminishes with increasing bulb volume, while more numerous cell types increase with increasing bulb volume at a faster rate than do less numerous cell types. Internal granule cells were found to be collected into groups hereafter designated as "discoids? Robert C. Switzer III The discoids were found to exhibit at least two size populations. The smaller population of large discoids lies in the external region of the layer and the large pOpulation of small discoids occupy the inner part of the layer. QUANTITATIVE RELATIONSHIPS OF CELL POPULATIONS IN OLFACTORY BULBS 0F MAMMALS I Robert Ci Switzer III A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Biophysics 1972 To my family ii ACKNOWLEDGMENTS I wish to thank my parents for giving me the means to obtain my undergraduate degree which is the basis from which I launch my career, and to their unending support and encouragement. My thanks to Jean Toth Allen who introduced me to the Biophysics laboratory; to Dr. Barnett Rosenberg who fanned my spark of interest and gave me my start in Biophysics; and to Dr. John I. Johnson who, in a firm and tacit way, has guided me in the ways of producing meaningful scientific work of high excellence. A very special thanks to my wife, Julie, who, besides helping with the cell counts and typing of this manuscript, has supplied a home life conducive to productive work in the laboratory. This research was supported by NSF Research Grants GB 13854 and GB 30783 and NIH training grant GM 1422. Australian specimens were secured through the courtesy of the Animals and Birds Protection Board of Tasmania, the Feuna Protection Panel of New South Wales, Prof. P. 0. Bishop, J. A. W. Kirsch and C. G. WOod. Histological preparations were by G. C. Kim. Specimens of chimpanzee, baboon and echidna were provided by W. I. Welker. iii TABLE LIST OF TABLES . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . PART I. QUALITATIVE OBSERVATIONS . Methods . . . . . . . . . . . . . Results 0 O O O O O O O O O O O O Olfactory bulb anatomy . . . . . Comparisons . . . . . . . . . . PART II. QUANTITATIVE OBSERVATIONS Methods Volume determinations . . . . . Cell counts . . . . . . . . . . Number of IG cells . . . . . . Discoid parameters . . . . . . . Radial density gradients . . . Discoid size . . . . . . . . Number of discoids . . . . . . Results . . . . . . . . . . . . Layer volumes . . . . . . . . Number of cells . . . . . . . . Cell densities . . . . . . . . . Ratio of cell types . . . . . Discoid parameters . . . . . . . DISCUSSION . . . . . . . . . . . . . Discoids of IQ cells . . . . . . . Phylogeny of discoids . . . . . . Ontogeny of discoids . . . . . . . Tufted cells . . . . . . . . . . . BIBLIOGRAPHY 0 O O O O O O O O O O 0 OF CONTENTS iii Page iv 19 23 23 26 28 3O 31 32 33 33 33 36 36 39 39 47 47 48 51 51 55 LIST OF TABLES Page Table l. Qualitative comparisons of selected features of olfactory bulbs from animals of diverse taxa....ll,12 iv LIST OF FIGURES Figure Page 1. Cross section of an opossum head ........................ 4 2. Layers of the olfactory bulb............................. S 3. Tangential view of discoids of internal granule cells ... 9 4. Sections of olfactory bulbs from animals of diverse taxa.l3-l4 5. Section of an olfactory bulb demonstrating placement of sample areas......OOOOOCOOOOOOOOO......OOOIOOOOOOOOOOOOIS 6. Columnar stacking of internal granule cells within d18C°1d8 O0....O......0..........0.00.00.00.0000000000016 7. Nerve fiber flow around discoids ........................l7 8. Brains of the Dasyuridae ................................21 9. Section of olfactory bulbs of the Dasyuridae ............22 10. Layer volume plots ......................................34-35 11. Number of cells .........................................37 12. Graph of cell densities .................................38 13. Graph of cell ratios ....................................40 14. Histograms of chords and diameters of discoids ..........41 15. Histograms of chords of discoid by radial region ........43 16. a. Density of IG cells within discoids .................45 b. Numbers of discoids .................................45 17. Average discoid chord length with respect to radial pagition 0.00.00.00.00...0.0.0.0...0.0.0.0000000000000046 l8. Phylogeny of discoids ...................................49 Detailed studies on the olfactory bulb have been confined mostly to rodents and rabbits (Cajal, '11-'55; Valverde, '65; Loman and Mentink, '69;.Allison and Warwick, '49; Nicoll, '70; Price and Powell, '70) . The work of Crosby and Humphrey ('39) covers several animals among reptiles, birds and mammals. Their work has been overlooked, probably because their descriptions considered.more the populations of cells rather than nerve fiber connectivity. References to Cajal attend more to his Golgi pictures than to his complete pictures; the papulations of cells, and some other features, in the olfactory bulb described by Cajal who looked at other than Golgi material ('11-'52, '11-'55) are often ignored. This and the confusion of some aspects of his description have led to misinterpretation in some studies. g For example, Valverde's ('65) labeling of the internal plexiform layer is confused with Cajal's two separate layers, the internal plexiform layer and the internal granule cell layer (see Fig. 2). This appears to be due to the young age of Valverde's animals. Andres ('70) seems to have made the same indistinction of these layers in.the part of his description dealing with mammalian olfactory'bulbs. Nicoll's ('70) interpretation of electrophysiological data involving tufted cells makes no distinction of the three populations of tufted cells which can be classified by their graded size and position. These classes would seem to be important for Nicoll's statements regarding cell size and action potential conduction velocities. The study of Rall and Shepherd ('68) on field potentials in the rabbit olfactory bulb bases relative resistance values and densities of internal granule cells on anatomical assump- tions derived from Golgi stain studies. No acknowledgment is made of the distinctive clustering of these cells (see Fig. 2) and what effect this may have on current flow. To resurrect, stabilize and add to Cajal's description of the olfactory bulb (Cajal, '11-'52, p.651, Fig. 411; '11-'55, p.4, Fig. 1) Part I of this study, Qualitative Observations is presented. Part I. Qualitative Observations Methods Subjects for the qualitative observations are listed in Table 1. They comprise a group of animals which exhibit a large range of variations of anatomical features of olfactory bulbs. These animals were also chosen to include those more commonly observed (i.e. rat, mouse, and cat), and representatives from diverse taxa (i.e. shrew, baboon, and opossum.) Brains of all animals had already been fixed in 10% formalin and embedded in celloidin, sectioned at 22, 25 or 30um and stained, and were ready for examination when this study began. Every other section was stained for Nissl substance with thionin, the others stained for myelin with iron hematoxylin, using a Hail or Sanides-Heidenhain method. Results Olfactory Bulb Anatomy The outermost layer is the fibrous layer of afferent olfactory cell axons which arrive at the bulb through the cribriform plate (Fig. 1). These axons terminate in the globular, cell-scarce glomeruli (Fig. 2). There the axons intermingle and synapse with the primary dendrites of the mitral and tufted cells and the axons and dendrites of the short axon periglomerular cells. The peripheral processes of the internal granule (16) cells and extrabulbal axon afferents also penetrate to the glomerular layer. The glomeruli them- selves are embedded in a dense matrix of granule, periglomerular and tufted cells. These cells form the external granule layer (EGL). The tufted cells of the EGL are typically small and were labeled by Cajal as peripheral or interstitial tufted cells. The tufted cell layer (TCL) internal to the glomerular layer, has been included by many workers as part of the external plexi- form layer. A fuller discussion of why this is probably so will be given in the Discussion. The tufted cells of the TCL are termed the middle tufted cells. How sharply defined this layer is depends on the number of middle tufted cells present. The tufted cells in this layer are distributed with the largest cells internal and the smallest external, nearest the glomerular layer. The Nissl substance of the tufted cell bodies has an Figure 1. This photograph of a cross section of an opossum head (from which the cerebellum has been removed for another experiment) reveals, among other things, the following: B, the olfactory bulb; C, the cribriform plate, MT, the maxilloturbinal; NT, the nasoturbinal and the ethmoturbinals l, 9, 5 and b. It is on ethmoturbinals and the posterior half of the nasoturbinal that the olfactory epithelium is found; easily identified by the light brown-yellow color of the epithelium. mu... has.» 0 ‘ o e L f ' -...“ or...” Figure 2. Layers of the olfactory bulb. This is a photomicrograph of a segment of a coronal section stained for Nissl substance with thionin from Dasyurus. Bulb layer boundaries are labelled and indicated along the right hand edge of the segment. obvious tufted appearance where it extends into the primary dendrite and axon from.the soma. Also quite evident in the larger cells of the middle tufted cell layer is a cruciform appearance of the Nissl substance extending into the horizontal dendrites. The external plexiform layer (EPL) is located between the layer of middle tufted cells and the mitral cell layer (MCL). The EPL has received much attention in recent years following the discovery of the reciprocal synapse between the horizontal (accessory or basal) dendrites of the mitral cells and the peripheral processes of the internal granule cells (Hirata, '64; Price and Powell, '70; Ball and Shepherd, '68). These two types of fibers comprise the bulk of the EPL. Most of the internal tufted cells occupy the inner region of the EPL while some are scattered in the outer EPLt The tufted cells nearest and in the mitral cell layer are easily confused with the mitral cells. The mitral €911 layer is made up of three cell types: mitral, tufted and granule cells. The arrangement of’mitral cells varies ranging from appearing to be placed one cell deep on the surface of a basement layer of granule cells to being more than one deep and embedded in a granule cell matrix. The mitral cells, so named because of their likeness to a'bishop's mitre, are the largest cells in the olfactory bulb. Tufted cells are interspersed among the mitral cells along the outer edge of the layer. There is also a range of sizes of mitral cells, some actually miniature and smaller than the larger tufted cells in some animals (e.g. opossum, Dasyurids). In these animals some of the middle and internal tufted cells appear larger than any mitral cells. If true, then it could be expected that the two papulations would exhibit appreciable overlap in their size distribution curves. The internal plexiform layer (IPL) is a cell-scarce layer containing fibers running radially and others running parallel to the MCL. This layer is said by Cajal ('11-'52, p. 665; '11-'55, p.11) to be the richest in nonmyelinated fibers in the bulb. With the Neil and Heidenhain stains, however, the IPL~shows a darker color than layers on either side implying the presence of myelin. This may not be unreasonable since afferent axons course through this layer without making contacts until terminating in the MCL or EPL. The work by Price and Powell ('70) reveals very few spines on the peripheral processes of the 16 cells where they pass through the IPL, implying very little synaptic activity on them and in this layer. The IPL appears to be a layer for transmission and not for terminal contacts. Tufted cell axons emit collaterals to the IPL. Valverde includes mitral axon collaterals in this layer but as was pointed out above, his IPL included the IG layer. Cajal specifically states that there are no mitral cell axon collaterals in the strictly defined IPL ('11152, p. 664, '11-'55, p.13). Afferent fibers from the anterior commissure are represented in the IPL with extensive bifurcation of fibers in the IPL region before penetrating to the EPL and to the MCL to form retes amongst those cells (Cajal '11-'52, p.667; '11-'55, p.13). The size of the IPL appears to range from barely recognizeable to something with quite respectable dimensions among different animal types. The next layer inward has been called the internal granule cell layer,but because of the character of the organi- zation of the internal granule cells this layer will be referred to as the discoid layer (DL). The IG cells have nearly round somas and are mostly nuclear showing only a bare rim of cytoplasm. Cajal referred to the aggregation of IG cells as clusters or islands, ovoid in shape. Examination of these ovoids in different planes in sections cutting per- pendicularly (see Fig. 2) to the olfactory bulb surface and in sections tangential to the olfactory bulb surface (Fig. 3), reveals that, three dimensionally, they are discoid in shape, that is, oblate spheroids. The term discoid will be used hereafter to refer to these IG cell aggregations as well as the layer in which they are found. Among different types of animals there is a wide range of variations in the following characteristics of discoids: density, thickness, (which can be expressed in number of cells thick) and radial gradients of length and thickness (Fig. 4 and 5). Sections from different planes reveal some degree of regional differences of these characteristics of IG discoids. Particularly in regions of high curvature of the bulb and in the lateral-anterior regions, fiber bundles perforate the DL resulting in distortion of discoid shape. MCL IPL DL Figure 3. thionin. Tangential View of discoids of internal granule cells. O.5 mm -r- Stain: uh— 10 The longest, thickest and most dense discoids are typically external to the smaller ones. In some animals, however, a few smaller discoids lie most externally. In the largest discoids a columnar stacking of cells is evident in varying degrees (Fig. 6). This feature was briefly noted by Cajal ('11-'52, p. 667; '11-'55, p.13) ". . . . these branchlets (of thick centrifugal fibers) prefer the vicinity of the granules and the intergranular septa or columns . . . Where the discoids are most dense they seem to con- form their shape in accordance with the presence of neigh- boring discoids. It is as if a "membrane" enveloped each discoid. Fibers coursing across the surfaces and in between the ends of the discoids also show a curious feature. The fibers, as revealed in myelin stains, on the ventricle side of the discoids leave vacant a space closest to the discoids while on the external surface of the discoids the fibers flow close to the discoid surface. This is evident throughout most of the DL with less prominence near the inner DL boundary (Fig. 7). 11 Table l. Qualitative Comparisons of Selected Features of Olfactory Bulbs from Animals of Diverse Taxa This table is a compilation of qualitative descriptions of some features of the olfactory bulbs of several different animals. A section from the olfactory bulb of each animal is shown in Figure 4. This table and Figure 5 show that certain features of olfactory bulb architecture are more accentuated in some animals than in others, eg. the tufted cell layer. The term "502 to the MCL" means that the inner edge of the TCL is about half way between the glomerular layer and the mitral cell layer (MCL). ”Thin" is defined as three cells or less in thickness while "thick" as greater than three cells in thickness. 12 Table l Mitral Tufted Cell Tufted Cell IPL Cell Discoid Density Sizes Breadth Density Character Rat Moderate, Obvious Wide High Long, thin, Rattus TCL-EPL line heterogeneity closely norvggicus less than range spread Rodentia 502 to MCL Deer mouse High, well Medium Wide High Thick, graded Peromyscus defined, heterogeneity size, but californicus TCL-EPL line not obvious longest, Rodentia about 502 to thickest and MCL most dense near IPL Cat Low, Medium, almost Wide High Gradations in Felis indefinite, homogeneous size, catus TCL-EPL line within TCL, variations in Carnivora less than 502 large near thickness to MCL, barely TCL moderately evident. spaced. Shrew High, well Even Narrow High Gradations in Blarina defined, heterogeneity, size, tightly brevicauda TCL-EPL line largest nearly packed, thick Insectivora about 502 to the size of MCL mitrals Chimpanzee Very low, Medium to Wide Low Diffuse, §§p_ layer not well small but not clumps barely troglodytes defined,mostly well defined evident Primates near glomerular layer Baboon Low, TCL-EFL Medium low Medium— Medium Thick, large, Chaeropithecus line not dispersion, wide clumping low number of papio defineable almost thin smaller size Primates homogeneous layer of small granule gradation cells in size Opossum High, very Many large, Very High Even size Didelphis distinct, heterogeneous, wide gradations, marsupialis TCL-EPL line large range long, thin virginiana greater than Marsupialia 502 to MCL Sugar Glider Moderate, Obvious Medium High Thickness and Petaurus TCL-EFL line heterogeneity, length graded, breviceps about 50% to large cells moderately Marsupialia MCL mostly near spaced;longest TCL-EPL line. and thickest in a band nearest IPL. 13 ,‘l, .‘ '.-.:' .\ 'l ‘u . 8;;" H," ...‘a .‘:'~. — R01 {’13}. ,5:.';;"‘3,:’“~27-'1fi;{m Deer mouse 1!- ..L Figure 4. shown in these photomicrographs are those compared qualitatively in Sections of olfactory bulbs from animals of diverse taxa Table 1. Take particular note of: tufted cell density, breadth of the IPL, mitral cell density, and length, thickness and distribution of discoids. 14 Bobbon Opossum l j fllfll 7igure 5. A micrograph of a parasagittal section from the olfactory bulb of Sarcophilus demonstrates some of the conventions for: A - A , the caudal boundary for parasagittal and horizontal sections of the olfactory bulb: B, the sample area for point count volumetry to determine the fractional volume of the discoids in the DL; C&C: radial sample lines for discoid size sampling and radial position of the discoids; D - an example of a region of a section where the discoids were not sectioned parallel to their minor axes. Such a region was not used in point count volumetry or radial line sampling. Such criteria rendered the very small olfactory bulb of Sminthopsis very difficult to extract discoid data from. A section from Sarcophilus, the cells were stained with thionin. Figure 6. The columnar stacking of internal granule cells within discoids is seen in this photomicrograph from Antechinus. Section stained with thionin. l I Figure 7. Nerve fiber flow around discoids is shown in a section from Petaurus breviceps stained for myelin. Note the interesting feature of fibers leaving vacant the region immediately below the larger discoids. Section stain: Sanides-Heidenhain. 18 Sections stained fOr myelin exhibit a gradation of staining density in the periventricular layer. The density of stain diminishes as the discoid layer is approached. It seems that myelination begins where the discoid layer ends. That is, the radial gradients of discoid size and degree of’myelination are oppositely directed. The various short axon cells described by Cajal are evident in varying degrees. Often these cells are somewhat concealed within the dense discoids and are apparent only with higher magnifications (600-800x). The periventricular layer is barely existent in smaller olfactory bulbs but attains appreciable volume in larger bulbs. The ventricle varies greatly in size; it is cavernous in some speCies (e.g. Sarcophilus) and barely evident in others (e.g. Petaurus). ngparisons Figs. 4 and 5 show sections of olfactory bulbs from various animals which.point up similarities and differences seen over a number of animals. Table l compiles qualitative descriptions of scme of the morphological features of these olfactory bulbs. The distinctness of the tufted cell layer can be seen to vary considerably. Density of cells would be the best measure to quantify this apparent variation. The line of demarcation between.TCL and EPL varies from‘being closer to the glomerular layer to being closer to the MCL. The degree tO‘WhiCh the middle tufted cells follow the size gradation described by Cajal ('11-'52, p. 659; '11-'55, p. 7) ranges from.barely evident to very obvious. Measurements of cell body areas plotted into histograms should resolve this apparent degree of size dispersion. 19 The density of mitral cells is quite high in the smaller animals and seemingly quite low in the larger animals. Chimpanzee and baboon have small olfactory bulbs but are low in mitral cell density. Determination of the MIL area and the actual number of mitral cells would put these variations in better perspective. The IPL in cat seems to be larger than that seen in the shrew. A volume determination of the IPL instead of a canparison by its breadth would be a better basis for comparison. The discoids in rat are relatively long, thin, and closely spaced. In contrast to chimpanzee, the discoids in baboon are very well developed. The difference between the discoids of rat and baboon is characterized by the discoids being thicker, more widely spaced and of lower size dispersion in baboon. In all of the animals observed, differences in size distributions, density and thickness of the discoids are quite evident. To determine whether or not some of these features are due to bulb size and/or shape the discoids must be characterized quantitatively. It is rather evident that the olfactory bulb of the rat is not the best for studies of olfactory bulb anatonv. Other mammals, such as shrew, mouse and many marsupials, show accentuation of the discoid formations and a distinct layer of middle tufted cells. Part II. Quantitative Observations Few quantitative studies have been done on the anatomy of the olfactory bulb. Allison and Warwick ('49) determined values for the number of olfactory receptor cells, glomeruli, tufted cells and mitral cells. Which population of tufted cells was counted is not evident. The IG cells were ignored. The earlier studies by Holt ('17) and Smith ('28) were concerned with the number of mitral and IG cells only. A quantitative study involving more than one animal type was done by Mann ('63) on several species of Chiroptera and on the shrew (Blarina brevicauda). No cell counts were made but volumes were determined for certain layers within the olfactory bulbs from macroslices of bulb tissue. To maintain a truly synergistic relationship between structure and function it is appropriate to expand and reinforce the knowledge of olfactory bulb structure at not Just the fiber and synaptic level but in terms of cell populations. This is important in functional considerations since confidence in interpretations of electroplwsiological data is highly dependent on the knowledge of structure. Study of olfactory bulbs of differing size and shape on a qualitative level make it quite evident that meaningful canparisons can only be obtained by quantifying the features in question. The animals chosen for this study were from a single family which showed large variations in bulb size (Fig. 8). These were four species from the carnivorus family of Australian marsupials, the Dasyuridae: Sarcophilus harrisii (Tasmanian devil), Dasmg viverrinus (native cat or quell), Antechinus f lavipgs (cannon du-nnart) , and aninthopsis “253.9% (marsupial mouse). The olfactory bulbs of these animals were outstanding with respect to the accentuation of particular features, especially distinct layers of tufted cells and size gradation of IG cell discoids (Fig. 9). («Jay/undue .. a Dosyurus , ' . Antechinus , * . ‘ . ‘I ‘ ‘ - Sminthopsis Figure 8. Brains of the Dasyuridae are shown in this photograph of casts of the four species used in the quantitative part of this study. ?? :53 3‘3 " '73 ' A . . . ' ‘. . ntechinus 3; ‘ Figure 9. Sections from the olfactory bulbs of each of the four species of Dasyurids used in the quantitative part of this study are shown in this photomicrograph. Note the accentuated layer of middle tufted cells with the largest of the tufted cell perikarya predominantly occupying the inner boundary of the TCL. The discoids exhibited by the Dasyurids are large and well defined. Thionin stained sections. 23 There were two specimens of Sarcophilus; in one the olfactory bulbs were sectioned coronally, in the other the right bulb was sectioned parasagittally and the left bulb horizontally. Of three specimens of Daszgrus, the brains including olfactory bulbs were sectioned in different planes, one in coronal, one in parasagittal, and one in horizontal plane. The three specimens of Antechinus were sectioned as for Dasygrus. The single specimen of SminthoPsis had the left olfactory bulb sectioned parasagittally and the right coronally., For this study one bulb from each plane of section available for each species was used. For this study answers to the following questions were sought. -Does a given cell layer remain the same function of the bulb volume in animals of different sizes? - How does the number of a given cell type vary with bulb size? - Are the ratios between cell types dependent on bulb size? -Is the density of a given cell type a function of bulb size? — How are the discoids of internal granule cells organized? Methods Histological procedures were as described in Part 1. Volume Determinations Selected thionin sections were projected at magnifications of X30, 50 or 70. Outlines were traced of the olfactory bulb perimeter including the accessory bulb and the various layers to determine volumes. The external boundary of the bulb was taken as the outer outline of the glomeruli; this excluded olfactory nerve fibers. Any ventricle was excluded from the bulb volume. 24 The cell layer boundaries were defined as follows (see Fig. 2): external granule layer (EGL) Outer: the inner outline of the glomeruli. Inner: a line skirting the inner edge of the cell dense region internal to the glomeruli. Tufted cell layer (TCL) Outer: internal outline of the EGL. Inner: a line skirting the innermost extension of the middle tufted cells (very obvious in these animals). external plexiform layer (EFL) Outer: the inner boundary of the TCL. Inner: the mitral cell layer. mitral cell layer (MCL) Considered as a surface papulated with cells. internal plexiform layer (IPL) Outer: mitral cell layer. Inner: the internal granule cell layer. discoid layer (9L) Outer: the inner edge of the IPL. Inner: a line within which no discoid-like groupings of 16 cells are seen. periventricular layer Outer: inner DL. Inner: the ventricle or if none there is no inner boundary. 25 In coronal sections boundary definition was quite evident but some conventions had to be adapted in the horizontal and sagittal planes. The posterior aspect of the bulb in some sections ended adjacent to a heavy nucleus (AON) or even caudal to it. Whatever the case a line was drawn skirting the edge of the innermost extension of the olfactory bulb cells including the accessory bulb, if present, then internally along the caudal aspect of the DL. At this juncture a line was drOpped to the surface of the ventricle such that it was perpendicular to the ventricular boundary (Fig. 5). When no ventricle was present a smooth line was drawn to the "corresponding" opposite point of the bulb section. The AON was always excluded. With a set of about 12 tracings per bulb, approximately evenly distributed through the bulb, areas of the traced figures were determined with a compensating polar planimeter. Each area was traced three times and the average taken. These data yield each section's volume when multiplied by the section thickness and divided by the areal magnification due to projection (Dornfield, et al '42). Each of these section areas were plotted on a graph where section area showed the ordinate and the abscissa depicted the distance between sections. With the sections containing the beginning and termination 26 points of each olfactory bulb and bulb layers plotted as the zero points, a smooth line was drawn connecting all the plotted points. By planimetry the underlying area was measured from which bulb and layer volumes were determined. The mitral cell surface area was determined similarly except that the perimeter of the layer in each section was measured with a plan measure (Keuffel and Esser #62-0310, which measures line length). These perimeters were then plotted against section distance, the area beneath the resulting curve yielding the layer's surface area. Cell Counts Mitral cells were counted in every other section for Which an area had been determined. Plotting the obtained numbers versus section number and using the same zero points for beginning and end, as in the volume determinations, the smooth line Joining these points was planimetered, yielding the total number of cells. That this was a reasonable means of sampling was determined by selecting two sets of sections to be counted in the coronally sectioned olfactory bulb of Dasyurus. The two sets together comprised the group of the sections in which volumes had been determined, each set containing every other section of the group. Each set was plotted and a line drawn for each set of points. The number of cells determined from the planimetry of the curve was compared. The two counts were within 3% of their average. All counts were made in collaboration with another person. Counts were compared and accepted if they were within 10% of each other for counts greater than 100 and within 10 for counts less than 100. The internal tufted cells fall close to and mingle with the mitral cells, therefore leading to confusion as to which are which. Some of the tufted cells are clearly not in the mitral cell layer while others are close. The convention adopted in counting the mitral cells was to include all cells having their nucleoli on or within an imaginary line skirting the external edge of the bulk of the mitral cell.bodies. Even so, it is felt that the estimates of numbers of mitral cells may be too high, as was judged by.Allison and‘Warwick (f49) in their counts of mitral cells. There is prdbably further inaccuracy in counts in sections cut Obliquely to the EL since the external edge of the FCL was at an angle and the imaginary boundary was more difficult to determine. The middle tufted cells lie within the boundaries described above for the tufted cell layer volume determinations. Counting the larger tufted cells near the layer's inner edge was no prdblem. But the smaller tufted cells approach the size of the granule-like cells in this layer and unless they obviously bore the "tufted" appearance they were not counted. The tufted cells falling close to the glomerular layer were counted only if their nucleoli fell inside an imaginary line skirting the edge of the dense region of periglomerular cells, the EGL. Upon finishing the same procedure fer the tufted cells in the coronal sections for each animal it was found that a single section sampling could be employed. The section in which counts were to be made was selected by referring to the volume curves and choosing a section from a region of the curve where the volume was changing slowly, i.e. a.point where the volume profile line has a nearly zero slope. This point was usually near the center of the bulb. 28 This single-section sampling was tested in the coronally sectioned bulbs by selecting the section in which counts had already been made and which fit the above criteria. The cell density for that section was then determined by dividing the number of cells by the layer volume. This density was then multiplied by the total TCL volume for that bulb. The resulting cell number was compared with the cell number determined graphically and was found to be within 6%. This amount of variation was deemed satisfactory. This sampling method could not be employed for mitral cells since the values obtainedhad too great a variation from those values obtained graphically. Consequently, all of the mitral cell totals were determined graphically. Number of IG Cells Determination of the number of IG cells required different methods than the above since the cells were so numerous and dense within discoids that they were uncountable. Therefore the following scheme was employed: 1. Determine the fractional volume of the discoids in the discoid layer: VDL 2. Determine the fractional volume of the cells within discoids, i.e. packing density: VP 3. Determine average cell diameter, then calculate average cell volume: Vc 4. Volume of the discoid layer: VI. Fitting into the equation: VL x VDL x VP Vc IG cells = The fractional volume of the discoid profile was determined with modified point count volumetry. Selected sections meeting the same criteria set forth above for tufted cell counts were projected directly through a Zeiss microscope equipped with a photo changer, projection tube and deflecting prism onto a rectangular grid of dots. The dot spacing was 2.87mm or 12 dots per cm2. The projected image on the grid of dots was 238K. The method for volume fraction analysis was that set forth by Hilliard and Cahn ('61) using a dense grid of dots. A less dense grid of dots can be employed only if the particles are nearly spherical, randomly oriented and distributed and of low dispersion in size. Since the discoids did not meet any of these criteria, a dense grid of dots was required. The concentric symmetry of the discoid layer and anisotrOpic dis- tribution of size of discoids called for a sampling area that spanned the region between the external and internal DL boundaries (see Fig. 5). The width of this radial sector was arbitrary. The sample area was drawn with the radially directed sides parallel so that the total number of dots in the area could be easily calculated. Sample areas were taken in regions such that equal representation of all regions of each section would be obtained. For each olfactory bulb the discoid volume fraction was an average of all sample areas. The packing density of cells within discoid structures was estimated as 522 based on the following observations. The cells are not closely 3O packed, i.e. not hexagonally packed. This is revealed by tangential sections through discoids at different levels. The largest discoids exhibit columnar stacking of cells such that the cells are in close contact but the columns themselves(Fig. 6) are not in close or frequent contact. In smaller discoids where columnar stacking is not evident the cells are in contact but "loosely packed" using the term of Dallavalle ('43), i.e. every element is not in contact with all its neighbors. In analyses of porosity in geological considerations the porosity is 482 for a loosely packed state or a 522 occupation of total volume by the packed particles. Cell size was sampled within discoids along a given radial line through the DL. The average cell size was used to compute the cell volume. Determining the number of 16 cells was then just a matter of substituting values into the equation. Discoid Parameters The discoids when viewed along a line parallel to their minor axis have a nearly circular profile. Slicing the discoids along this same line results in only one cut that is along the diameter while the rest are chord lengths. Viewed from the side or perpendicular to the minor axis, each of the sections of the discoid have an elliptical profile. The major axis of each of the discoid section profiles is identical to a chord length of the discoid when viewed along a line parallel to the minor axis, i.e. perpendicular to the major axis. In the rest of this text the major axis of the elliptically shaped discoid section profiles will be termed "discoid chord lengths.” 31 The discoid section profiles were counted by using a rectangular grid reticle in the eyepiece of the microscope and scanning across rows of the grid. In the course of the scan only the leading ends of discoid profiles which appeared in each square of the grid were counted. For the discoids a sampling procedure similar to that used fOr the tufted cells was feund practicable except that two sections in each bulb were counted to obtain an average discoid profile.density between them. From this density the total number of discoid section profiles were determined. Two sections rather than one were counted since regions of slow section to section volume change in the DL were not as extensive as for other subvolumes: DL volumes varied more from section to section: the section-to-section volume slopes were higher fOr the DL. Radial Density Gradients Since there is a high degree of concentric symmetry to the arrangement of the discoids, sampling lines placed radially on a given section (perpendicular to the maJor axes of the discoid profiles) were used to ascertain radial density gradients and length versus radial position of the discoids. Placement of the radial sample line was accomplished by a line reticle placed on the top of the radiant field diaphragm. This could be brought into the same focus as the section with the stage condenser and still have the section properly illuminated. Criteria were set for beginning and end points along the radial line, and in what sections and their regions were to be measured as follows (see Fig. 5): 32 «Beginning or zero point: innermost part of the mitral cell layer, i.e. boundary between IPL and MCL. «End Point: the internal boundary of the discoid layer. --The length, zero to end point, was established as the normal length. The position of discoids along this line as converted into percentages of this normal length. --Radial lines were placed to give as equal representation as possible of anterior, ventral, dorsal, lateral and medial regions. Only those sections or regions within a section where the discoids had been out nearly perpendicularly were sampled. Otherwise the position and dimensions of a discoid could not be easily determined. Discoid Size The discoid size class analysis was carried out by first compiling a histogram of discoid profile lengths measured in the radial line sampling. Since these profiles are from discoids cut parallel to their minor axes these profile lengths are chord lengths of the discoids of which only a small percentage are diametric lengths. Because the discoids' minor axes are radially aligned and in the same plane as the sections chosen for measurement, Hennig's graphical method (Elias and Hennig, '67) can be employed for extraction of diameter size distributions from chord size distributions. The chords are from sections of discoids whose diameters fall into a larger size class. Naturally, the largest size class in a histogram of chord sizes is comprised only of diameters. The numbers of diameters in smaller size classes are obtained by progressively subtracting the number of chords that are fran discoids whose diameters have already been counted. 33 Number of Discoids To extract the number of discoids from the number of discoid section profiles the following formula was employed: Define: D - as the number of whole discoids per olfactory bulb. P - as the total number of discoid section profiles per olfactory bulb. p - as the number of discoid section profiles in the sample. d - as the number of discoid profiles in the sample whose major axes were actually diameters of discoids rather than chords. Then for each animal: D-%XP Results Layer Volumes Layer volumes, numbers of cells and ratios between them are all referenced to the olfactory bulb volume. A plot of layer volumes is shown in Fig. 10. In a log-log plot the data points for each layer can be joined by a straight line. This indicates that the volume of each layer is a constant function of the bulb volume. Each layer has a rate of increase in volume nearly that of the bulb itself indicated by the nearness of 1.0 of the slopes. LAYER VOLUME v: OLFACTORY BULB VOLUME 9° IO3 F 6 LAYER! t° b «'0 Sarcophilus\‘ / /fl ‘° oscmo . '02 ._ Dosyurus \ / . TUFTEO \, cau. 63 A 9 If! E EXTERNAL .. GRANULE w CELL 3 ..l o > £5 4 .J a 65 9°“? / .0 v / 5° , 4/ -I 1 I 1 I L l 0 IO ‘ Id) IO' :02 '03 , OLFACTORY BULB VOLUME (mm3) 1 .o > Figure 10. Layer volume plots. A log—log plot of each layer of the olfactory bulbs of the Dasyurids against the total volume of the olfactory bulbs is shown in graphs (3) and (b). Grapha (a) is a plot of the cell layer and (b) a plot of fiber layers. The data points for the olfactory bulbs of any given species is enclosed by a line. Individual data points are indicated by a letter apprOpriate to the layer. The olfactory bulb volume is plotted against itself to serve as a reference line for a slope of 1.0 and as an aid in determining what fraction a particular point is of the olfactory bult volume of interst. The slope of each line is given near the center of each line. The straight line character of each set of data points indicates that the particular layer remains a constant function of the bulb volume with respect to size. If the 510pes were exactly 1.0 then the layer would also be a constant fraction of the olfactory bulb volume as the total volume changed. Note that in these graphs and those to follow which are log—log plots the equations of the straight lines are of the form log y=m log x + 102 b- 35 LAYER VOLUME vs OLFACTORY BULB VOLUME l03r v- /V\00 1 Sarcophilus \ 1 ”0 fl, / d ,-_ - .. , Dosyums / ° l0 r- \ J P ,3" GL 4' / i e e/ ”' _ /' EPL “A J ’ A h' a 1 IO. L Mec mus y ‘ IPL 1. /P A Smintho s' 5 n p l$\ E - y r‘ e / / v { Ae.‘ / “a I; ‘00 : Q g f/ > 0 9° ' IO *- , é // s .. é, / v—v: PV ‘9 «r e 0: GL : 0—0! EFL . «f/ P—r: IPL -' ‘/ l0 - h L / - 4» J/ \ d}. a e '62 0 1 I. 1 12 1 I IO no lo :03 OLFACTORY sULs VOLUME (mm31 Figure 10(b). 36 Number of Cells In a similar plot the numbers of the three cells types, middle tufted, mitral and internal granule are shown in Fig. 11. The lines do not have slopes as close to 1.0 nor are they as parallel as those in the plots of layer volumes. Since the lines connecting the data points are straight the number of each cell type remains a constant function of bulb volume. The lines do not have nearly identical slopes nor are any as close to 1.0 as was the case with the plots of layer volumes. Evidently in a given layer the layer volume increases at a faster rate than the number of cells. This is reflected in a plot of cell densities against bulb volume shown in Fig. 12. Cell Densities The density for each cell type declines with increasing bulb volume. From Sminthopsis to Sarcophilus the 16 cells decrease in density by about one half, the tufted cell density decreases by a factor of about one third and the density of mitral cells decreases by one fifth. (The mitral cell layer is here considered as a surface so that cell density is a surface density.) This trend of decreasing nerve cell density 37 CELL NUMBER V5 OLFACTORY BULB VOLUME / Dasyurus @ Sarcophilus SLOPE= 0.824 I07 ~— / CELL NUMBER '05 _ 0.626 / . SminthopsisW ::. 0.380 A I04 Io' :02 103 OLFACTORY BULB VOLUME (mm Figure 11. Number of cells. Log-log plots of each of the three cell types, internal granule, tufted (middle) and mitral cells against olfactory bulb volume are shown. The straight lines through the data points indicate that the number of each cell type remains a constant function of the total bulb volume regardless of size. (Only one data point exists for the number of 16 cells in Sminthopsis ysince the volume fraction of the discoids used in the cell number computation could not be determined in the olfactory bulb sectioned parasagittally.) 38 CELL DENSITIES vs OLFACTORY BULB VOLUME INTERNAL GRANULE CELLS "E 3'” s ‘5 ‘ e g xlO :3 I - 'oJ > 0 1 L 1 | 1 a: w )- fi 5C)F \ Sminthopsis TUFTED CELLS E 40 - Antechinus m :23 330 _ Dasyurus z KIO SOrCOphiius 3. 2° " . f “‘ ® ° I0 - o l l l J I l '5; 4- MITRAL CELLS g g 3 3 r— 2 g xIO D a: 2 I. . z 4 .J (r . .J m I .. 3 i b .1 O I l l I j l '00 IO' :02 I03 OLFACTORY BULB VOLUME (mm3) Figure 12. Graph of cell densities. The number of each cell type divided by the volume of the layer each occupies yields the cell density. Plotted against the volume of the olfactory bulb the trend of the cell density is to decrease with increasing bulb volume. Note that the mitral cell density is a surface density and not a volume density since the mitral cell layer is more of a surface populated with cells that a volume containing cells. 39 with increasing brain size follows that of cortical nerve cell density over the range of brain size of mouse to whale (Tower '54). Ratio of Cell Types The trends between the numbers of each cell type are shown in the plots of ratio of cell types against bulb volume (Fig. 13). In each case, the more numerous cell type was made the numerator. All of these ratios increase with increasing bulb volume, IG/mitrals the most and mitrals/tufted, the least. Allison and Warwick's ('49) ratio for tufted to mitral (assuming they counted only the middle tufted cells) was 2.5 which is close to the ratios exhibited by Dasyurus. The ratios ranged from 1.8:1 for Sminthopsis to 5.9:1 for Sarcophilus. In general it is observed that while the numbers of cells increase with bulb volume and the densities of each type de- crease, the ratios between cell types increase: the more numerous cells increase with bulb volume relatively more than do the less numerous cells. Discoid Parameters The histograms of discoid diameters in Fig. 14 reveal at least two peaks of size classes of discoids in all four animals. The primary peak is comprised of small discoids. Each animal exhibits a secondary peak of large discoids indicated by the arrows. 1&0 RATIOS OF CELL TYPES vs OLFACTORY BULB VOLUME '22.” 3 20 '- . SorCOphIlus 2 l8 - \ XIO 3 l6 " l4 " l2 " IO ' ' ' Dosyurus .3. 8 \‘ 5)."- Antechinus\ /. :1 6 L33??? m 4 - Sminthopsis\,~ ./ Q S“ A _§____RANULE _____ D ('50 ——- u. 2 ~ 5 ”"' TUFTED O 0 J I . 9. Io° 10' '02 ... <1 0: 6 p .- 4 ' / 3 2 TUFTED / - ' @W—w—v MITRAL O L l .LL I so° IO' I02 OLFACTORY sULs VOLUME (mm3) Figure 13. Graph of cell ratios. These plots show the ratios between cell types with the more numerous cell type of each possible pair as the numerator. 41 HISTOGRAMS 0F DISCOID: CHORDS DIAMETERS pm ' pm 0 I98 98.8 I728 0 I .8 98.8 I718 59.3 I38.3 2|7.4 59.3 l38.3 2l7.4 3° ’ r 20 . Sarcophflus 0 iifiriiT‘ 30' 20 . Dasyurus WM IO" Antechinus NORMALIZED ’ FREQUENCY on O 0 i‘riii‘fi‘iiv 4O ’ 4O ’rfi 30 - Sminthopsis 3o - ZOI 20 ' Oiiiw‘iiuw‘iif Oifii+1u1iii 0.2 0.6 I.0 L4 I8 22 0.2 0.6 LO L4 l.8 2.2 EYEPIECE MICROMETER EYEPIECE MICROMETER UNITS UNITS (a) (b) Figure 14. Histograms of the chord lengths of discoids are shown in (a). From these histograms are extracted the histograms for diameters of discoids, (b),by the method of Hennig (see text). Since the small discoids dominate the histogram profiles the overall profile of the chord histograms is not altered greatly in obtaining the diameter histograms except that secondary peaks (arrows) not present or evident in the histograms of chords appear in the diameter histograms. 42 In Sarcophilus and Sminthopsis, the largest and smallest animals of the group, this second peak is the lOOum range. The secondary peak in Dasyurus is the next largest size class (118um) and in Antechinus it is the next lowest (79mm). From largest to smallest olfactory bulbs, the major peak frequency of small discoids increases in proportion to the whole pOpulation of discoids, while the large discoids in the secondary peak follow no trend and remain near 62. The size class centered about 79um (0.8 units) stays very constant, deviating only in Antechinus where this is also the size class exhibiting the secondary peak. To demonstrate that the large discoids are predominantly found in the outer DL and the small discoids are preferentially located in the inner DL, the data for the discoid chord length histograms was divided so that histograms could be drawn showing the chord size distribution for the outer and inner half of the DL. Fig. 15 shows a pair of histograms for each of the four species. If "large" discoids are defined as any profile greater than 0.7 units than it is quite evident that relatively few discoids of this profile size range occupy the inner half of the DL. 1+3 INSTOGRAMS OFIMSCOH) CHORD LENGTHS pm 0 59.3 138.3 217.4 0 59.3 138.3 L A I A #1 a +1_ .— PLA A I A A v F' ' fi' ' ' "T f ' I ' . u r I 8 98.8 177.8 19.8 988 177.8 20 ’ I Sarcophilus 2° ’ I Dosyurus I l r ' ' .0 . I OUTER HALF OF OL l I l 0 I I I I I’ I j I I W l I 20 ~ I - 20 I— | a I l I '1 I A . I INNER HALF OF DL _ l INNER HALF OF DL g lo I 10 >. ‘5 L5,] 0 I IJ'IJTJFI I Tj I I I O I I ILI ILTl-lI-fI—‘T‘I—I o -.....__ [LI . . . E 20 ’ l Antechrnus 2° ’ I SmmthopSIS l I O 1,3 I 77 I 2 ,0 J“ . OUTER HALF OF DL ,0 . | OUTER HALF OF DL 2 I o: O z TLD 0 I I I I r I I I I I I_I' 0 I I T’ I I T I I I I r I —- — I | | I 20 ~ I 20 . I | I ' 7 I ,0. I INNER HALF OF DL .0 , | INNER HALF OF DL : I o T I I I f't-‘r fiW—T 0 1 I F1 I I I I I I o 0.4 0.8 12 I5 2.0 2.4 o 0.4 0.8 1.2 I.6 2.0 CHORD LENGTHS (EYEPIECE MICROMETER UNITS) Figure 15. Histograms of chord lengths of discoids, reduced to show radial regional differences in the discoid layer, are shown here. For each species of the Dasyurids, a pair of histograms shows the chord lengths of discoids found in the outer and inner half of the discoid layer. These histograms reveal that the large discoids (discoids greater than 0.7 units designated by the dotted line) are preferentially located in the outer half of the discoid layer. 411 Note that since these are discoid chord histograms and not discoid diameters histogram , some of the small profiles found in the outer half of the DL are due to chord sections of the larger discoids. The densities of 16 cells within discoids as a function of bulb volume are shown in Fig. 16A. The increase in density with increasing bulb size can be attributed to the increasing proportion of large discoids in bulbs of greater size(See Fig. 14). The number of discoids per bulb increases with increasing bulb volume as shown in Fig. 168. The rate of increase is not as great as that for cells; the number of discoids per bulb is closer to being proportional to the square root (0.526 - m) of the bulb volume. Another way of analyzing the discoid data is to determine the average discoid chord length as a function of radial position. This is done in Fig. 17. The most striking feature of these histograms is that the maximum average chord length of the discoids is very nearly the same for each species: regardless of olfactory bulb size the maximum average chord length of the discoids is almost the same. This could be interpreted in terms of there being some functional importance to the size of the discoids. 115 I6 CELL DENSITY 0F DISCOIDS vs OLFACTORY BULB VOLUME a) (I) '03 F" j 9 m U (2 .- D 0/ _ osyurus ‘2 o 0/ SorOOphilus LL ll. ANIeChInUS / . O O _ g] S /o Sminthopsis m z z 10 e 10 - 10 10 NUMBER OF 01300103 vs OLFACTORY BULB VOLUME 106- Sorcophilus o/ Dosyurus / Sminthopsis 0 Antechinus NUMBER OF DISCOIDS 0| '04 I J I I l I 10° 10' 102 103 OLFACTORY BULB VOLUME 1mm3) Figure 16. The density of IG cells within discoids (a) (top) and the number of discoids (b) (bottom) are plotted against olfactory bulb volume in these graphs. For each species only one data point is shown in contrast to preceeding plots since the derivation of the number of discoids came from an average from the specimens of each Species. AVERAGE CHORD LENGTH OF DISCOIDS I/S RADIAL POSITION Sarcophilus Dosyurus 1.0 - 1.0 r r:— ~98.8 :3 _. I .E __ 3 'I_. ..— § 0.5 - ——._ 0.5 ~ ~49.4 (D E O .9 E L V O . . o . o :1: 0 20 40 60 80 10 0 20 40 60 80 100 5 11m 5 Antechinus Sminthopsis ....I IO .. [O r ”-98.8 o _ F‘— E I a .7 I) HF—I 0.5 ~ 0.5 :___ .1 ~49.4 =1 . I” 0 20 40 60 80 IOO AVE RAGE O 1 I I l I l 0 20 40 60 80 I00 RADIAL POSITION IN DISCOID LAYER (°/o) Figure 17. The average discoid chord length as a function of radial position is shown in these histograms. Of particular note is that the maximum average discoid chord length for each species is nearly the same: 1.06 units for Dasyurus; 0.94 units for Sarcophilus; 0.90 units for Antechinus and 0.88 units for Sminthopsis. ‘17 DISCUSSION Discoids of IG Cells The internal granule cells are organized populations of cells as manifested by their discrete grouping in the discoids. 'Where the internal granule cells have previously been treated as single interneurons, the possibility must now be considered that the discoids are discrete populations of interacting internal granule cells. The close Juxtaposition of cells within the discoids invites consideration of possibilities for weak electrical coupling. Price and Powell. ('70) Observed that where the cell membranes of internal granule cells oppose each other thickenings on.the membrance exist. .Although the extracellular space is maintained with no indication of a tight Junction, the thickenings of the cell membrances where cells are juxtaposed may be some structural specialization related to electrical coupling'between the cells. 'With such a situation excitation of a few granule cells in a given discoid could influence the whole population of cells within that discoid which could bring about synchronous activity. ‘What has been revealed quantitatively about the population of discoids was not at all expected from.qualitative Observations. Regardless of olfactory bulb size the distribution of size classes of discoids exhibits two peaks. Furthermore, each of those peak size classes is preferentially distributed in the radial dimension. Small discoids are congregated internally, large discoids are restricted to external locations. A possible significance of the two size populations being differentially distributed spatially may be f0und in interpreting the findings of Price and Powell ('70). They revealed that axons afferent to the olfactory bulb of rat, from the cerebral hemisphere, terminate on the individual internal granule cells in a laminar fashion. For instance, terminals firm the anterior commissure teminated on the internal part of the 16 cells. In rat the discoids are smaller and more densely spaced and their size distribution less evident than in larger animals. One may then ask the questions: 1. Does the compactness of the discoid layer in the rat olfactory bulb prevent one from determining whether the termination of axons on the internal granule cells is laminar with the respect to the individual IG cells or to the whole layer? 2. In larger olfactory bulbs do the deep dendrites of the outermost 16 cells extend all the way to the inner fringe of the discoid layer? If the axons afferent to the bulb teminate in a laminar way with respect to the discoid layer and the deep dendrites of the IG cells whose perikarya are situated externally toward the IPL do not extend the full length of the layer, then the population of small discoids would be receiving fibers fran the anterior commissure and the population of larger discoids would not. Phylpgeny of Discoids Examination of the IG cells in olfactory bulbs of animals from a wide range of taxa suggests that the organization of IG cells into discoids is a specialization of structure. The degree of developnent of the IG cells into discoids across several taxa of animals is schematized in Fig. 18. Varying degrees of developnent exist not only among taxa but within. Amphibians exhibiting rudiments of discoids are observed in some CHARACTER OF DKKXNDS ACROSS SEVERAL TAXA Arrangement of . Proximal to IPL= Lamina Of anino of thick “Vernal granule Ddiuse Fhuyon of IIUn concentnc Ith discmds; chscmds; kugesI “”5 high density Iomino largest ncor near IPL IPL CLASS tOOd Amphibians frog _ IurIIe Rephies OIIIqoior duck Birds sparrow kiwi dove chicken Mommab: Monotremes echidna giant anteater PIacentols mon pig shrew Morsupiois Opossum Dosyurids Figure 18. Phylogeny of discoids. The character of discoids across several taxa is depicted here with respect to the degree of aggregation of internal granule cells in terms of distinctness and size. ”Thin" discoids are defined as discoids three or less cells in thickness and ”thick" discoids are discoids greater than 3 cells in thickness. Descriptions of the layer containing the internal granule cells for some of the animals was obtained from the literature as follows: toad and frog, hoffman '63; alligator, Crosby ‘17, '39; turtle, Crosby and Humphrey '39; sparrow, dove, duck and chicken, huber and Crosby '29, kiwi, Craigie '30. Specimens of echidna and giant anteater were made available by w. l. Welker and shrew by W. L. Weller. 50 toads (Bufo marinus) and frogs (Rana pipiens) (Hoffman '63). The alligator'(§lligator mississipiensis) (Crosby, '17, '39) seems to exhibit the most well developed discoid layer among reptiles. Alligator surpasses most birds in terms of discoid develogment. The sparrow (Passer domesticus) is well down this scale as is kiwi. Kiwi (apteryx australis, Craigie, '30) seemingly demonstrates an intermediate stage in the devel- opment of discoids. No discoids are apparent in kiwi but the distribution of IG cells nearest the IPL is much denser than those toward the ventricle. From the description of Huber and Crosby ('29), dove (Columba domestics) and duck (Anas domestics) and chicken (Callus domesticus) show, in order, increasingly more developed discoid layers of birds. Man seems to be better than alligator and more like chicken. Opossum extends the lower range of the marsupials to overlap the placentals so that shrew and pig exhibit more well developed discoid structure according to the headings on the chart in Fig. 18. The general pattern among the placentals is the for orders, Insectivora, Dermoptera,Chiroptera, Primates, Edentata, Lagomorpha and Rodentia to have more distinctly developed discoids than the orders of Carnivora, Hyracoidea, Perisso- dactyla and Artiodactyla. 51 Ontgggny of Discoids Schultze-Westrum ('69) studied the increasing ability of pouch young of sugar gliders to discriminate among community members by smell. In opossum pouch young, (which spend about the same time in the pouch as sugar gliders), the time course of discoid develOpment appears to parallel the stages of greater discriminatory powers of olfaction in young sugar gliders. This parallel would be well worth detailed study. Tufted Cells It is evident from the layer volume determinations that the region occupied by the middle tufted cells is a well defined layer remaining a constant function of the whole bulb volume from Sminthopsis to Sarcophilus. This layer is very distinct among the Dasyurids but not so evident in many other animals. Although very poorly defined in the rat, Valverde ('65) recognized and labeled the region occupied by the muddle tufted cells as the layer of tufted cells. The region between the mitral cell layer and the glomerular layer, previously called the external plexiform layer, is divided into two layers: the tufted cell layer and, internal to it, the external plexiform layer. In some of the animals considered in this study, the middle tufted cells populate their layer so distinctly that previous notions (Crosby and Humphrey, '39; Valverde, '65) that these cells are displaced cells of any type are difficult to maintain. 52 The work of Price and Powell ('70) on synaptology of olfactory bulb cells indicates that the deep half of what they define as the external plexiform.layer is a region of prime synaptic activity between horizontal dendrites of’mitral and internal tufted cells and the peripheral processes of the IG cells. The outer half of these external plexifbrm layer does not get as mmch.mention, undoubtedly because it is not part of’thg external plexiform layer but the (middle) tufted cell layer. Hence, the external plexifonm layer should be defined as the layer where the peripheral processes of the internal granule cells synapse with the horizontal dendrites of'the mitral and internal tufted cells. This definition of the TCL raises the question about where the horizontal dendrites of the middle tufted cells synapse with the peripheral processes of the internal granule cells. Do these dendrites remain confined to the TCL or do they mingle with the horizontal dendrites of the internal tufted cells and mitral cells? Possibly the horizontal dendrites of the large middle tufted cells do mingle while the smaller'middle tufted cells confine their horizontal dendrites to the tufted cell layer. Such an arrangement implies a division of function of the tufted cells, possibly related to the projections of the axons of the various tufted cells. Iohman and Mentink ('69) feund that axons of some of the tufted cells projected to the lateral olfactory tract while others did not. The idea that the large cells either of the internal or middle tufted cell population, or both, may project to the lateral olfactory tract receives support from the elecrophysiological experiments of Nicoll ('70). 53 Most of the tufted cells from which Nicoll recorded after antidromic stimulation in the lateral olfactory tract of rabbit were within a region about 200 um external to the mitral cell layer. This places these cells within the internal tufted cell population zone or external plexiform layer as redefined above. He also indicates that similar responses were obtained from a few cells nearer the glomerular layer. These probably were the large cells of the middle tufted cell population. Both studies of Nicoll and Lohman were done in rabbits which do not have a well developed middle tufted cell population. This also means there are not likely to be many large cells of that population. It would seem that Nicoll recorded from the large internal tufted cells and, more peripherally, from some of the few large cells of the middle tufted cell population. One would certainly expect to find more clear-cut results regarding tufted cell size and axon projections if studies similar to those of Nicoll, and Lehman and Mentink were carried out on animals with denser tufted cell layers such as opossum or any of the Dasyurids. Conclusion One of the most important things to be gleaned from the quantitative results is that within this family of animals, which covers a large range of olfactory bulb size, certain relationships between olfactory bulb components remain constant. For animals of other families these relationships may not be the same. As was pointed out before, the tufted cell layer seems to vary greatly in the volume it contributes to the bulb. The internal plexiform layer is also seen to vary considerably among other animals, being quite small in some to very wide in others . 5h The establishment of constancy in the character of structural features permits questions to be posed and hypotheses made about their functional significance. By observing in other groups of animals these same features, in quantified terms, variations or similarities in structure can be established which.may reflect differences in function. Net only does knowing the structure enable the electrophysiological data to be interpreted but it enables one to formulate the questions necessary to design the experiments which are to probe the functioning structure. Some of the questions which have arisen as a result of this study and can form.the basis for future work are: --Do the individual discoids project some sort of discrete inhibitory field to the mitral and tufted cells? --Given that a discoid inhibitory field exists what is the relationship of the axon collaterals of the mitral and tufted cells to it? -qAre the closely packed internal granule cells of the discoids weakly coupled so as to affect the excitability of neighboring cells? --Do the short axon cells in the discoid layer have inhibitory control over the cell population within a discoid? and do these short axon cells communicate between or couple discoids? --Is there a unique area to which the tufted cell axons project and is this area accentuated in size or in definition in annuals with high densities of middle tufted cells? 55 It is evident from this study that future experiments on the olfactory bulb can benefit by using animals in which features of interest in the olfactory bulbs are accentuated, as is the case in the marsupials studied here . B IBLIOGRAPHY 56 IITERATURE CI'IED Allison, A. C., and Warwick, R. T. Turner l9’+9 Quantitative observations on the olfactory system in the rabbit . Brain 72: 186-197. Andres, K. H. 1970 Anatomy and ultrastructure of the olfactory bulb in fish, amphibia, reptiles, birds and manuals. In Taste and Smell in Vertebrates, A Ciba Foundation Symposium. Ed. by G. Wolstenholme and J. Knight. Jae. Churchill, London, pp. 177-196. Cajal, S. R. 1911-1952 Histologie du Systeme Nerveux de PM et des Vertebres 2, Paris: Maloine: 6h7-672 Reprinted 1952 Institute Ramon y Cajal, Madrid. Two volumes 1-986, 1-993 pp. Cajal, S. R. 1911-1955 Studies on the cerebral cortex (Idmbic structures) trans. by Kraft, L. 14.; Lloyd-luke, London. 1-179 + xi pp. Craigie, E. H. 1930 Studies on the brain of the kiwi (m australis) J. Comp. Heur, M9: 223. Crosby, E. C. 1917 The forebrain of alligator mississippiensis. J. Comp. Neur. 27: 325-h02. Crosby, E. C. and Humphrey 1939 Studies of the vertebrate telencephalon. I. The nuclear configuration of the olfactory and accessory olfactory formations and of the nucleus olfactorius anterior of certain reptiles, birds and mammals. J. Comp. Neur. 71: 121-213. Dallavalle, J. M. 198 Micromeritics: Technology of fine particles. Pittman, N. Y. Dornfield, E. J., D. w. Slater and H. Scherre’ 191:2 A method for accurate determination of volume and cell numbers in small organs . Anat . Rec . 82: 255-259, 57 Elias, H. and A. Hennig 1967 Stereology of the human renal glomerulus. In: Quantitative Methods in Morphology, Ed. by E. R. Weibel and H. Elias. Springer-Verlag, Berlin, Heidelberg, N. Y. pp. 131-166. Hilliard, J. E. and J. W. Cahn 1961 An evaluation of procedures in quan- titative metallography for volume-fraction analysis. Trans. of Metallurgical Soc. of Amer. Inst. Metallurgical Engineers 221: 3hh-352. Hirata, Y. 196M Sane observations on the fine structure of the synapses in the olfactory bulb of the mouse, with particular reference to the atypical synaptic configurations. Arch. Histol. Jap. 2h: 293-302. Hoffman, H. H. 1963 The olfactory bulb, accessory bulb and hemispheres of some anurans. J. Comp. Neur. 120: 317-368. Holt, C. M. 1917 Studies on the olfactory bulbs of the albino rat in two parts. I. Effects of a defective diet and of exercise. II. Nmnber of cells in bulb. J. Cusp. Neur. 27: 201-261. Huber, G. C. and E. C. Crosby 1929 The nuclei and fiber paths of the avian diencephalon with consideration of the telencephslic and certain mesencephalic centers and connections. J. Comp. Neur. #8: 1-225. Lehman, A. H. M. and G. M. Mentink 1969 The lateral olfactory tract, the anterior conmissure and the cells of the olfactory bulb. Brain Research 12: 396-h13. Mann, G. 1963 The Rhinencephalon of Chiroptera. Investigaciones Zoolo’gicas Chilenas Vol. IX. , Centre de Investigaciones Zmlogicas, Universidad de Chile, Santiago, pp. 1-93. 58 Nicoll, R. A. 1970 Identification of tufted cells in the olfactory bulb. Nature 227: 623-625 Price, J. L. 1969 The structure and connexiona of the granule cells of the olfactory bulb. Proc. Physiol. Soc. 204: 77-78. Price, J. L. and T. P. S. Powell. 1968 The synaptic vesicles of the reciprocal synapse of the olfactory bulb: an electron microscope study. Brain Research 11: 697-700 Price, J. L. and T. P. S. Powell. 1970 The morphology of the granule cells of the olfactory bulb. J. Cell Science 7: 91-123. Price, J. L. and T. P. 8. Powell 1970 The synaptology of the granule cells of the olfactory bulb. J. Cell Science 7: 125-187. Price, J. L. and T. P. S. Powell 1970 The mitral and short axon cells of the olfactory bulb. J. Cell Science 7: 631-651. Rall, w. and G. M. Shepherd 1968 Theoretical reconstruction of field potentials and dendrodendritic interactions in olfactory bulbs. J. Neurophysiol. 31: 884-915. Schultze-Westrum, Th. G. 1969 Social communication by chemical signals in flying phalangers (Petaurus breviceps papanus) In: Olfaction and Taste 111. Ed. by C. Pfaffmann Rockefeller Univ. Press, 268-277. Smith, L. A. 1927 A comparison of the number of nerve cells in the olfactory bulbs of domesticated albino and wild Norway rats. J. Comp. Neur. 45: 483-501. Tower, D. B. 1954 Structural and functional organization of mamalian cerebral cortex: the correlation of neurone density with brain size. J. Comp. Neur. 101: 19-52. Valverde, F. 1965 Studies on the piriform lobe. Harvard University press, Cambridge, pp. 1-131 + i-vii. u”tummy”:MW!nymzlmusmum