SUBDIVISIONS OF THE SUPERIOR OLIVARY NUCLEAR COMPLEX: MARSUPIAL VERSIONS WITH PLACENTAL COMPARISONS Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY DANIEL DAVID LYDNS 1972 LA; - q--—o [his]. t;— 6 BY SDNS' IN NEW NE. \' '"RQ :3 ‘5 amB HDAB H Influx e ‘1' Bi. ‘2‘ II I. V m, SUBDIVISIONS OF THE SUPERIOR OLIVARY NUCLEAR COMPLEX: MARSUPIAL VERSIONS WITH PLACENTAL COMPARISONS BY Daniel David Lyons The superior olivary nuclear complex (SO) was studied to determine its differentiation into nuclear com- ponents in specimens representing five genera and five families of Australian marsupial mammals: Dasyurus viverrinus (the native cat, Dasyuridae), Isoodon obesulus (short-nosed bandicoot, Peramelidae), Phascolomis mitchelli (wombat, Phascolomidae), Trichosurus vulpecula (brush- tailed possum, Phalangeridae) and Thylogale billardierii (red-bellied wallaby, Macropodidae). In addition, for comparative purposes, the SO of an American marsupial, Didélphis marsupialis (the opossum, Didelphidae), and a cosmopolitan placental, Rattus norvegicus (albino rat, Muridae), were also examined. Cell body size (area), cellular density and cellular morphology were determined in each spatially distinct nu- cleus. Three brains (representing coronal, horizontal and sagittal planes of section) were examined in each species; the neural tissue had been stained for nissl bodies or Daniel David Lyons myelin sheaths. Cell body size distribution was determined for each component nucleus by sampling 150 cells (50 from each plane of section) and measuring each cell by planimetry. Density was measured for each component nucleus by counting the number of cells contained within 15 separately measured volumes, 5 volumes in each of three planes. Cellular mor- phology was determined by direct microscopic observation of Nissl stained material. Four distinct nuclei were recognized in every mammal studied: the lateral superior olivary nucleus (LSO), the superior paraolivary nucleus (PAR), the medial nucleus of the trapezoid body (MTB) and the lateral nucleus of the trapezoid body (LTB). In addition, all animals except the wombat possessed a distinct medial superior olivary nucleus (MSO); furthermore, the rat alone possessed a preolivary nucleus (PON). The results of this study agree with the literature on the anatomy of the rat SO but indicate two additional nuclei (LTB and M80) in the SO of Didelphis; the SO of Australian marsupials is presented for the first time. The organization of the SO in the rat (Eutheria) was found to be remarkably similar to the organization of the SO in both Didelphis and the Australian marsupials (Metatheria). From this, it is concluded that the SO complex found today in metatherian and eutherian mammals had a common evolu- tionary origin. SUBDIVISIONS OF THE SUPERIOR OLIVARY NUCLEAR COMPLEX: MARSUPIAL VERSIONS WITH PLACENTAL COMPARISONS BY Daniel David Lyons A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Zoology 1972 ACKNOWLE DGMENTS I wish to thank the following members of my com- mittee for their time and assistance: Dr. John I. Johnson (chairman), Dr. J. Alan Holman (both of the Department of Zoology, Michigan State University) and Dr. Glenn I. Hatton (Department of Psychology, Michigan State University). Special thanks go to Dr. Johnson for loan of the histologi— cal specimens utilized in this study. ii TABLE OF CONTENTS LI ST OF TABLES O O O O O 0 O O O O O 0 LIST OF FIGURES . . . . . . . . . . . INTRODUCTION 0 O O O O O O O O O O O O The superior olivary complex . . . . Functions of the superior olivary complex . METHOD 0 O O I I O I O O O O O O O O 0 Animals and histology . . . . . . . Cell body size measurements . . . . Cellular density . . . . . . . . . . Spatial relationships of component nuclei cellular morphology . . . . . . . RESULTS 0 O O O I O O I O O O I O O 0 Spatial relationships of component nuclei Cellular morphology . . . . . . . . Cell body size measurements . . . . Density measurements . . . . . . . . DISCUSSION 0 O O O O O O O O O O O O 0 Organization of the superior olivary Evidence for homology of the superior olivary complex . . . . . . . . . Morphology . . . . . . . . . . . . Functional morphology . . . . . . Phylogeny . . . . . . . . . . . . O ‘0 O O 0 complex Monotremata: an independent development of the mammalian auditory system? . . LIST OF REFERENCES . . . . . . . . . . iii Page iv H < 03.5 10 10 11 12 12 24 32 34 36 36 38 38 40 41 42 44 LIST OF TABLES Table Page 1. Results of cel ular density measurements (cells x 10‘ /u3) . . . . . . . . . . . . . . 35 iv Figure 1. 10. ll. 12. 13. 14. LIST OF FIGURES Spatial relationships of the component nuclei of the superior olivary complex . . . . . . . . Coronal section through approximately the middle of the SO in Thylogale. Thionin stain . . . . Coronal section through approximately the middle of the SO in Rattus. Thionin stain . . . . . . Coronal section through approximately the middle of the SO in Trichosurus. Thionin stain . . . Coronal section through approximately the middle of the SO in Phascolomis. Thionin stain . . . Coronal section through approximately the middle of the so in Isoodon. Thionin stain . . . . . Coronal section through approximately the middle of the SO in Dasyurus. Thionin stain . . . . . Coronal section through approximately the middle of the SO in Didelphis. Thionin stain . . . . Cells of the lateral trapezoid body nucleus Thionin stain . . . . . . . . . . . . . . . . . Cells of the medial trapezoid body nucleus Thionin Stain O O O O O O O O O O O O O O O O 0 Cells of the lateral superior olivary nucleus Thionin Stain O O O O O O O O O O O O O O O O 0 Cells of the medial superior olivary nucleus Thionin Stain O O O O O O O O O O O O O O O O 0 Cells of the superior paraolivary nucleus Thionin Stain O O O O O O O O O O O O O O O 0 0 Distributions of cell body area measurements . . Page 13 15 15 l7 17 19 19 21 26 26 28 28 30 33 INTRODUCTION Apparently, mammals have evolved several times independently from the therapsid reptiles. Some of those mammalian lines have since become extinct, but persisting today are the results of two presumably independent reptilian developments: marsupials and placentals on the one hand, and monotremes on the other (Kermack, '67 and Clemens, '68). Based on fossil evidence and comparative anatomical studies of recent mammals, marsupials and placentals are believed to have had a common mammalian ancestry, but one perhaps as ancient as the earliest Cretaceous (Lillegraven, '69). In Australia, the native mammalian fauna is essen- tially marsupial; on this island continent marsupials have developed in isolation from the placentals since late Cretaceous or early Tertiary times (Clemens, '68). In isolation, they have evolved a structural and ecological parallel for many forms the placentals have produced on other continents (Keast, '68). Fossil evidence shows that this diversified marsupial fauna has developed from primi- tive opossum-like ancenstors (Clemens, '68), in much the same way the diversified placental fauna has developed from primitive insectivore-like ancestors (Lillegraven, '69). The fact that so many convergent forms have been developed in the two groups indicates the adaptive importance of very specific suites of characters. Placentals have received much anatomical study, but in comparison, much yet remains to be investigated in mar- supials. In particular, concerning the nervous system, only Didelphis among the marsupials has been investigated near to the extent that placental species have been in- vestigated. Thus it was decided to investigate aspects of the nervous system of Australian marsupials to see if their nervous system has converged with that of placentals as so many other structures have. The auditory system was well suited for this kind of investigation, as it seems obvious that this sensory system should have been under strong se- lective pressure throughout evolutionary history. As indi- cated by Masterton, Jane and Diamond ('67) and Webster, Ackerman and Longa ('68), the superior olivary nuclear complex (SO), because of its location not more than one synapse away from the cochlea, is behaviorly important to a mammal, both for reflex movement to sound stimulation and for auditory localization in space. This is given support by the works of Zvorykin ('64), Harrison and Irving ('66c), Sehmsdorff ('66), Irving and Harrison ('67), and Webster, et al ('68); these investigators not only demonstrated the great variety of forms the SO can take in various placental mammals but also related this variation (in part) to various behavioral specilizations. In studying the phenomenon of convergence, there must be an awareness for the possibility of homology. As stated before, marsupials and placentals are believed to have had a common mammalian ancestor. Structures as highly significant behaviorally as the SO may have reached their present stage of development early in mammalian history. As Tumarkin ('68) pointed out, a possible reason for the extinction of large therapsids (non-mammalian ancestors) and the perfection of middle ear structures (via smaller therapsid ancestors and early mammals) was the result of intense pressure exerted by the dominant Mesozoic archosaurs with their efficient, air sensitive hearing structures. Considering the diminutive ecological role of mammals in Mesozoic times, it is easy to picture an environment with strong selective pressure for early perfection of hearing structures; certainly it is likely that strong selective pressures operative then resulted in the attainment of the advanced features characteristic of mammals today. If, in the analysis of any structure found in both marsupials and placentals, the point is reached where it can be debated if the structure is of common or convergent origin, then that in itself is highly significant. For if the structure is truly homologous, then certainly strong selective pressure must have been active to keep it in near ancestral condition for over 100 million years of widespread radiation in other structures. But on the other hand, if the structures concerned have evolved in parallel, the belief occurs--earlier stated-—that that particular type of arrangement is functionally demanded by the environment. In either case, if the selective factors involved could be precisely elucidated, functional assignments could readily be made. The superior olivary complex The superior olivary nuclear complex (SO) is a bi- lateral aggregate of discrete nuclei located within the rostral medulla of amphibians, reptiles, birds and mammals. Mammals possess the most complex SO (Ariens Kappers et al, '36). These neurons are largely (if not completely) con- cerned with auditory functions; present belief (based on mammalian studies) is that they serve as a sorting and integrating center, mediating auditory reflexes and re- laying auditory input from the cochlear nuclei to different areas of the brain (for example, Goldberg, '68; Masterton et al, '67; Galambos et al, '59; Irving and Harrison, '67). Among marsupial mammals, the SO of only Didelphis has been investigated. Stokes ('12) was the first to de- scribe the SO of the opossum.\lHe recognized medial and lateral parts of a superior olivary nucleus and a trapezoid nucleus with "poor differentiation of parts". Subsequent investigations of Didelphis by Voris and Hoerr ('32) re- vealed essentially the same organization, except that they recognized three "more or less clearly defined areas" en- meshed within the fibers of the trapezoid body. An atlas of the opossum brain (Oswaldo-Cruz and Rocha Miranda, '68) indicates a medial and lateral superior olivary nucleus (probably equivalent to the lateral and medial parts or lobes of the superior olive in Stokes, '12 and Voris and Hoerr, '32) and a nucleus of the trapezoid body. They believed that subdivision of the nucleus of the trapezoid body (referring to Voris and Hoerr, '32) is hazardous. Thus, according to the literature, the SO of Didelphis is currently believed to be differentiated into only three parts. Contrary to the situation with regard to marsupials, investigations of the SO of placental mammals are numerous. Essentially, all these works substantiate the fundamental work of Ramon Y Cajal ('09). But one very important point, that there is no such thing as a typical SO, has recently emerged. Certainly there is considerable interspecific variation in not only size of component nuclei but also in the presence or absence of component nuclei (e.g. Harrison and Irving, '66c). At least one nucleus not found in the cat has been indicated in some other placentals: the superior paraolivary nucleus. I follow Irving and Harrison ('67) in recognizing its distinctiveness based upon cellular morphology. The present work appears to be the first description of the $0 of Australian marsupials. Nomenclature and cri- teria of subdivision of the SO have been extracted from previous works on both placentals and Didelphis (especially from Oswaldo-Cruz and Rocha Miranda ('68) on Didelphis; Stotler ('30) and Rasmussen ('46) on the cat; Harrison and Irving ('66 a,b,c) on the rat; Irving and Harrison ('67) on various placentals; Goldberg and Brown ('68) on the dog; Meesen and Olszewski ('49) on the rabbit). In this study, however, quantitative neuroanatomical methods have been applied in addition to the previously utilized qualitative methods. The criterion of Irving and Harrison ('67), that of using internal morphological features rather that pri- marily geographical features to subdivide the complex, has been followed. The nomenclature, however, still consists of positional terms, reflecting the lack thus far of func- tional identification of the components of the SO. Functions of the superior olivary complex Behavioral functions have been suggested for only 2 of the component nuclei of the SO: M80 and LSO. Since the work of Stotler ('53), cells of the M50 have been re- peatedly shown to receive a bilateral input from the ventral cochlear nuclei. Cells of the M80 have been described as possessing 2 dendrites in bipolar fashion; one dendrite projecting laterally and receiving fibers from the ipsilateral cochlear nucleus and one dendrite projecting medially and receiving fibers from the contralateral cochlear nucleus. Numerous physiological investigators have since shown that MSO may be involved in the phenomenon of sound localization (e.g. Galambos et al, '59). This conclusion appears to correlate well with the anatomical observations. More recently, however, Harrison and Irving ('66c) and Irving and Harrison ('67) demonstrated that well known echolocators (bats and dolphins) lack a medial superior olivary nucleus. In animals that do possess a MSO, the number of component cells shows a linear correlation to the number of cells in the sixth cranial nucleus (abducens). From these observations, they postulate 2 auditory systems in mammals. One involving the M80 is concerned with reflex movements of the eyes to a sound source; the other involving LSO is concerned with the control of body movements to various properties of sounds. From the physiological evidence of other workers and lesion studies of their own, Masterton and Diamond ('67) conclude that M80 is involved in sound localization by responding to time differences in arrival of stimuli at each ear (AT) and LSO is involved in sound localization by responding to intensity differences at each ear as a func- tion of frequency (AFi). From this, they state that MSO (AT localization) is more important in animals with wide set ears and LSO (AFi localization) is more important in animals with close set ears, aquatic habits or those that can respond to higher frequency. These conclusions appear to agree with those of Irving and Harrison ('67); Masterton and Diamond ('67) also agree with Irving and Harrison ('67) on the correlation M80 and VI. Harrison and Beecher ('67) warn that structural interpretations of lesions to the trapezoid body (as per- formed by Masterton and Diamond) may be very difficult, as such a procedure abolishes a number of ipsilateral and contralateral tracts besides those to MSO. Nonetheless, Masterton and Diamond ('67) receive support from the work of Bordreau and Tsuchitani ('68), who arrive at the same conclusions for LSO function, based on their own physio- logical studies. And numerous investigators have shown the importance of AT in stimulating MSO cells (e.g. Galambos et al, '59; Goldberg and Brown, '69). METHOD Animals and Histology Brains of the following mammals were studied: (1) Dasyurus viverrinus Shaw (Eastern native cat, Dasyuridae), (2) Isoodon obesulus Shaw and Nodder (Short-nosed bandicoot, Peramelidae), (3) Trichosurus vulpecula Kerr (Brushtail possum, Phalangeridae), (4) Phascolomis mitchelli=Vombatus ursinus Shaw (Common wombat, Vombatidae), (5) Thylogale billardierii Desmarest (Red-bellied wallaby, Macropodidae), (6) Didelphis marsupialis Linnaeus (Virginia opossum, Didelphidae), and (7) Rattus norveqicus (Erxleben) (Albino rat, Muridae). The particular animals studied were selected due to availability and due to the desire to observe a variety of Australian marsupials and a relatively non- specialized placental SO. Three brains (representing coronal, horizontal and sagittal planes of section) were examined in each species. All animals had been perfused intracardially with 0.87% saline followed by a 0.87% saline plus 10% formalin mixture. The brains were removed, embedded in celloidin and sectioned at 25p (Didelphis and Rattus) or 30m (all 10 other brains). Alternate sections were stained with iron hematoxylin (Sanides Heidenhain or Weil method) for myelin sheaths and thionin (Nissl method) for cell bodies. All the brains used in this study are from the collection of Dr. John I. Johnson, Departments of Biophysics, Psychology and Zoology at Michigan State University. Col- lecting and histological processing of the mammals were performed by Dr. Johnson and his histological assistants. Cell body size measurements Using sections of brain tissue stained for cell bodies (Nissl stain), component nuclei were magnified and projected upon a drawing table at 2140K. From each com- ponent nucleus, samples of cells from several sections in each plane (coronal, sagittal and horizontal) were taken, such that a total of 150 cells were sampled from each component nucleus. Only those cell bodies with visible nucleoli were traced. Cell body areas were determined by planimetry with a Keuffel and Esser compensating polar planimeter. From these measures, distributions of cell body sizes for each component nucleus were constructed. Cellular density Again using sections stained for cell bodies, cellular densities were determined by counting the number of cells with nucleoli enclosed by the image of the grid 11 in a Whipple—Hauser iisk. The disk is placed within the microscope ocular, and projected with the section, the two appearing superposed upon the drawing table (X665). Five samples were taken from each of three planes of section for each nucleus. Density=number of cells/grid area x section thickness. Section thickness was measured by focusing up and down on each section of the SO, and taking readings from the vernier on the fine focusing knob at each end of the field of view. Spatial relationship of component nuclei; cellular morphology Spatial relationship of the component nuclei to each other and to the trapezoid body were determined by the use of the low power (50X) projections of the SO complex; both myelin-stained and Nissl-stained sections were ex- amined. Cellular morphology was determined by direct microscopic observation (300x) of Nissl-stained material. RESULTS Spatial relationships of component nuclei All animals studied exhibited the same general characteristics of SO differentiation. Figure 1 depicts the general arrangement of the component nuclei within the brainstem. This figure, while drawn from section of Trichosurus, is representative of the nuclear arrangements for all the animals of this study. As shown in figure 1, the SO is located in the ventral medulla, wedged in between the pontine nuclei and the ventral nucleus of the lateral lemniscus anteriorly and the motor nucleus of the VII nerve posteriorly. Figures 2—8 are photomicrographs of coronal sections through approximately the middle of the SO, equiv- alent to figure 1, part 3. With 2 exceptions, the SO consists of 5 spatially distinct nuclei: the lateral superior nucleus (LSO), the medial superior olivary nucleus (MSO), the superior parao- livary nucleus (PAR), the medial nucleus of the trapezoid body (MTB) and the lateral nucleus of the trapezoid body (LTB; medial or internal preolivary nucleus of most authors). The first exception occurs in the wombat (Phascolomis), 12 Figure l. 13 Spatial relationships of the component nuclei of the superior olivary complex. The figures illustrate actual tracings of the nuclei of Trichosurus, but are meant to represent in general the spatial relationships for all the animals of this study. Part 1 illustrates horizontal sections through (a.) ventral, (b.) intermediate and (c.) dorsal regions of the complex. Part 2 illustrates sagittal sections through (a.) lateral, (b.) intermediate and (c.) medial regions of the complex. Part 3 illustrates a coronal section through approx- imately the middle of the complex. LSO, lateral superior olivary nucleus; LTB, lateral nucleus of the trapezoid body; MSO, medial superior olivary nucleus; MTB, medial nucleus of the trapezoid body; NLL, ventral nucleus of the lateral lemniscus; PAR, superior paraolivary nucleus; PONTn, pontine nuclei; VIIn, motor nucleus of the VII nerve. I. HORIZONTAL I I I I 14 2. SAGITTAL IMM Figure 2. IMM SO Corona on through approximately the middle of the SO in Thylogale. Thionin stain. Root fibers of the VI nerve can be seen penetrating MTB dorso—ventrally. LSO, lateral superior olivary nucleus; LTB, lateral nucleus or the trape- zoid body; MSO, medial superior olivary nucleus; MTB, medial nucleus of the trapezoid body; PAR, superior paraolivary nucleus; PON, preolivary nucleus; PYR, pyramidal tract. Coronal section through approximately the middle of the SO in Rattus. Thionin stain. Root fibers of the VI nerve can be seen penetrating MTB dorso-ventrally. Abbreviations as in Figure 2. 1...! . o . .0 a .. U s a o ‘ .v .9. a O O .2 .0. .1. u . Figure 4. Figure 5. Coronal section through approximately the middle of the SO in Trichosurus. Thionin stain. Root fibers of the VI nerve can be seen penetrating MTB dorso- ventrally. Abbreviations as in Figure 2. L80 PYR Coronal section through approximately the middle of the SO in Phascolomis. Thionin stain. Root fibers of the VI nerve can be seen penetrating MTB dorso- ventrally. Abbreviations as in Figure 2. 19 /<:§1- PYR Figure 6. Coronal section through approximately the middle of the SO in Isoodon. Thionin stain. Root fibers of the VI nerve can be seen penetrating MTB dorso— ventrally. Abbreviations as in Figure 2. I I SO IMM LSO Figure 7. Coronal section through approximately the middle of the SO in Das urus. Thionin stain. Root fibers of the VI nerve can Be seen penetrating MTB dorso- ventrally. 'Abbreviations as in Figure 2. '3 a C, . .- ... .' s ',\~ ~ .. ',‘t..";‘-l “I. 9 . r :‘qu-‘N‘j‘ - ~ “-9” .54 x . I ‘e-‘. 3.2;" I I -.j;wr.:5‘ 21 L80 Figure 8. Coronal section through approximately the middle of the SO in Didelphis. Thionin stain. Abbreviations as in Figure 2. 22 (“W‘OIMW6 WW3; .. ...... .. (J... «Paw-1. .%. .rhfk...8. “WV . .....wnl. -—. 23 which lacks a distinct M80. The second exception is Rattus, which in addition to the "normal" five had an additional nucleus, the preolivary nucleus (PON; lateral or external preolivary nucleus of most authors). The names LTB and PON were chosen instead of the usual internal and external PON, respectively, for several reasons. First of all, LTB is actually enmeshed within the fibers of the trapezoid body. Secondly, among the 7 mammals studied, the nucleus here named PON is apparently present only in Rattus; and here it is definitely dorsal to the mass of trapezoid body fibers. My investigations of the cat (Felis catus) SO showed a similar arrangement of LTB and PON (my names). Furthermore, while poorly developed in Rattus, the PON is very well developed in Felig, and forms a distinct cell group running ventrally, laterally and rostrally around the SO, but always dorsal to trapezoid body fibers (see Berman, '68 for figures and descriptions of this relationship). Finally, Shaner ('34) in his studies of the development of the pig SO showed that developmentally the "PON nucleus is a fragment of the trapezoid nucleus and should be grouped with it." He suggests using the name "lateral trapezoid nucleus" for his PON. His figures label this "lateral trapezoid nucleus" as PON; this nucleus corresponds in position to my LTB. I do not agree with Shaner ('34) that these different nuclei (his trapezoid nucleus and his PON, respectively corresponding to my MTB and LTB) are cytologically 24 similar, at least in the animals reported in this study (see next section on cellular morphology). In studying spatial relationships, I found it neces- sary to view the complex from coronal, sagittal and hori- zontal planes. The MSO of both Didelphis and Dasyurus are very difficult to see in coronal section; however, in transverse and horizontal sections, this nucleus becomes very evident. Perhaps the reason earlier investigators missed M80 in Didelphis was because they failed to look at the SO in planes other than coronal. In some sections, these spatially separate nuclei blended at their borders. This occurred most frequently between M80 and LTB, and less frequently between LSO and LTB, and MTB and LTB. In this study, the ventral nucleus of the lateral limniscus (NLL) was considered spatially distinct from the SO (as Stokes, '12 but contrary to Barnes et al, '43 and Papez, '30). As shown in Figure 1, part 2a., in a few sections through the lateral part of the SO, NLL comes in contact with both LTB and LSO, but the 3 nuclei could be separated. NLL, however, was not examined in detail in this study. Cellular morphology Cellular morphology was distinctive for each sub- nucleus, and consistent within all the mammals studied. Descriptions of cellular morphology are given with each 25 figure (figures 9+l3); general comments are listed below. a. Lateral nucleus of the trapezoid body (figure 9). Frequently (especially anteriorly), a spatially sepa- rate component can be observed in the most ventral brain- stem, below the lateral edge of MTB. As this component contains the same cell body types and often blends with the main part of LTB, both are considered one nucleus. b. Medial nucleus of the trapezoid body (figure 10). MTB can be immediately recognized from any plane, or any section, due to its characteristic darkly staining, densely packed and very smoothly contoured cell bodies. c. Lateral superior olivary nucleus (figure 10). In many--but not all-~coronal sections of the rat, LSO cells form a curved column. No marsupial LSO forms such a column, but then neither do all placental LSO's form such a column (e.g., microchiropterans--Irving and Harrison, '67). The apparent internal subdivision of LSO in figure 2 is not consistent throughout LSO, nor does it provide a reliable subdivision on the basis of cellular morphology, cellular density or cell body size. I believe the subdi— vision is due to entering trapezoid body fibers. Nonethe- less this apparent morphological subdivision may indicate a functional subdivision. d. Medial superior olivary nucleus (figure 12). In all the animals studied (except Phascolomis) MSO cells form a distinct column extending nearly the length of the Figure 9. Figure 10. 26 Cells of the lateral trapezoid body nucleus (Trichosurus). LTB cell bodies are generally multipolar, with polyhedral outlines. Several cell processes are often visible on one cell. Thionin stain. Length of line equals lOOu. Cells of the medial trapezoid body nucleus (Trichosurus). MTB cell bodies are circular to oval shaped, with very rounded, smooth outlines. Cellular processes are rarely visible. Nuclei are often eccentric. Thionin stain. Length of line equals lOOu. 27 Figure 11. Figure 12. 28 Cells of the lateral superior olivary nucleus (Trichosurus). LSO cell bodies are circular or oval to fusiform in shape. Oval cell bodies often have a process extending from their more slender pole. Thionin stain. Length of line equals lOOu. Cells of the medial superior olivary nucleus (Trichosurus). MSO cell bodies are generally fusiform. Cellular processes are rarely seen. Thionin stain. Length of line equals lOOu. Figure 13. 30 Cells of the superior paraolivary nucleus (Trichosurus). PAR cell bodies are rounded, yet multipBIar. Several (3 or more) processes are often visible on one cell. Thionin stain. Length of line equals 100u. 32 SO. Degree of development (expressed in dorso-ventral heighth) varied considerably. Contrary to Harrison and Warr ('62), the accessory superior olivary nucleus (M80) is definitely present in "animals lower than the rabbit" (reference to Didelphis). e. Superior paraolivary nucleus (figure 13). After the criterion of Irving and Harrison ('67), the presence of large, rounded, multipolar cell bodies identi- fied this nucleus. f. Preolivary nucleus. PON cells were found only in the rat, and for this reason they were not studied in detail. Some study showed these cell bodies to morpho— logically closely resemble those of LSO and certainly not those of LTB at all. Cell body size measurements The results of cell body size measurements, per- formed on the spatially separate nuclei, are presented in figure 14. Each distribution represents 150 measured bodies, 50 cells from each of 3 planes (coronal, horizontal and sagittal). (Distributions drawn from separate planes for each nucleus were not significantly different, so they were presented in composite of 3 planes.) These measure- ments give support to the differentiation of nuclei based upon spatial relationships. In all the animals LTB con- tained the smallest cells, while PAR contained the largest. FREQUENCY Figure 14. mmspmms: THYLOGALE TRICHOSURUS ISOODON DASYURUS DIDELPHIS RATTUS in DZ. 3 . I LIB - 3)— fl - i-rgi'I-r 20 ii” 111] 8 111 l 1111] ~ O 111 J 114 L l l LLJJ 1L] 1 l_1 O 2l8 l l 3 1 l_lJ _LJ1 MTB LLIJ 20 Ill l 1111] N O l l l 1 S? g“ R? llllllJ N O l f.“ rr ill 0 r.— ZIB 437 1 i? llllLlllJ J J l Lillllj ff? 1 20 l l l 0 RAR 2|8 43? 655 AREA (112) Distributions of cell body area measurements. ordinate plots the frequency of occurrence of a given cell size while the abscissa plots the cell body area 11] N O l ‘(t-see below) 141] I N O 1111 l e; Ill] 20- l l J 11111 N 0 Ill 1 l 20" llllll] N O l 3'"??? i- I l 1 O 2|8 437 L J l N 111411? N O l l l 1 L80 JJIIIIII f? f?! llIlJ 20 111 l l l l m O I III ~ 0 lLlllllJ l l I O rrrr 28 437 The Each distribution is based on a sample of 150 cells, 50 cells from each of 3 planes (coronal, hori- zontal, and sagittal). 34 In the entire series of seven animals each nucleus appears to present a similarly shaped distribution (e.g. the broad abscissa spread of PAR); this is possibly a reflection of their homology. In general, however, a combination of this measure with density and qualitative features is necessary to arrive at a satisfactory distinction of component nuclei. Density measurements r7; The results of density measurements are presented in Table l. (Densities drawn from separate planes for each subnucleus were not significantly different, so they were presented in Table 1 only in composite of three planes.) As can be seen in the Table, M80 is by far the most dense nucleus; LTB and PAR are the least dense nuclei. From this, note that density is not at all related to cell size (cf. figure 14). Densities for each nucleus are similar through— out the animals investigated with several exceptions. MTB is generally variable; Rattus LTB and M80 are much denser than the corresponding nuclei in marsupials. Although M50 in Rattus forms only a very short column (dorso-ventrally) in coronal section, its density indicates a large total number of cells (contrary to Harrison and Warr, '62). 35 Results of cellular densit Table l. y measurements (cells x 10‘5/u3). MSO LSO PAR MTB LTB 085 51 813 21 454 PHASCOLOMIS 957 82 006 861 003 THYLOGALE 903 51 875 444 31 157 TRICHOSURUS 874 074 703 31 vAcoS ISOODON 174 61 680 531 123 31 795 923 31 vAcoS DASYURUS 854 61 611 441 734 21 526 02 923 21 FHS.S DIDELPHIS 744 81 284 454 41 229 226 92 vAcoS RATTUS DISCUSSION Organization of the superior olivarygcomplex The SO of all mammals in this study was found to exhibit the same general features of differentiation. A lateral superior olivary nucleus (LSO), a superior para- olivary nucleus (PAR), a medial nucleus of the trapezoid body (MTB) and a lateral nucleus of the trapezoid body (LTB) were present in all animals. In addition, all animals except the wombat possessed a distinct medial superior olivary nucleus (MSO); furthermore, the rat alone possessed a preolivary nucleus (PON). The original criterion of subdivision by spatial separation was found to be supported by measures of cellular density and cell body size and ob- servations of cellular morphology. The absence of M80 in the wombat may not be real, but at least no spatially separate equivalent of M80 was observable. PON observed in Rattus was very poorly developed as compared to Eglig; no equivalent of PON was observed in the marsupials. The subdivision of the SO in Rattus was entirely consistent with that of earlier workers (for example, Harrison and Warr, '62). However, the names LTB and PON have been applied instead of the usual internal and external 36 37 PON, respectively, for reasons given in the results section. It is suggested, whatever terms are used for these partic- ular nuclear groups, that workers indicate whether or not their PON or LTB is within or outside of the trapezoid body fibers. Otherwise confusion results. For example, Webster et a1 ('68) label a nucleus lateral preolivary that by their figure and description appears to be enmeshed within trapezoid body fibers. Therefore their PON may be equiva- lent to the nuclei identified in this paper as LTB; on the contrary, it could be equivalent to the preolivary nucleus of Stotler ('53) located outside trapezoid body fibers. Two nuclear groups were added to those usually given for Didelphis: M80 and LTB (cf. Oswaldo-Cruz and Rocha-Miranda, '68; Stokes, '12; Voris and Hoerr, '32). This further subdivision is supported by all the observations reported in this paper. SO differentiation may be even more complex than the present study indicates. For example, as noted in the results, LSO in some sections appears to be non-uniform in morphology (see figure 2). But even nuclei that appear uniform in morphology when stained for cell bodies (as in this study) may be subdivided with regard to other struc- tures (Golgi studies of Morest, '64). Furthermore, struc- tures obviously important but not even considered in this study include dendritic and axonal processes (including their destinations); neuroglial investment; type, number 38 and location of synaptic terminals; and even subtle dif- ferences within the cell body itself. In this study the component nuclei were found to exhibit interspecific variation in extent of develOpment. For example, LTB was large in Thylogale and Rattus, but small in Trichosurus; MSO probably exhibited the greatest variation, from possibly nothing in Phascolomis through small in Rattus and Didelphis to large in Thylogale. It could be interesting to carefully study this variation in component nuclei, patterned after the successful investi— gations of Irving and Harrison ('67) on the M80 in placentals. Evidence for homology of the superior olivary complex The organization of the SO in Australian marsupials and Didelphis was found to be strikingly similar to that reported for other mammals (that is, placentals). In fact structurally, the 80's in the 2 groups are so similar, it seems very likely that they are homologous. As Edinger ('49) illustrated in her studies of the evolution of the horse brain, speculations about evolution of characters derived from only living mammals are full of dangers. Therefore additional evidence is offered here to support the statement of homology. Morphology.--Fernandez and Schmidt ('63) conclude after investigations of Didelphis and several placental species, that the gross anatomy, histology and physiology 39 of the inner ear of the opossum is "essentially the same as that of placental mammals. Those differences that have been found are smaller than such differences between vari- ous species of placentals." These authors also add that the tympanic membrane and ossicles exhibit no fundamental differences between the 2 groups of mammals. Also impor- tant, they point out, is the lack of a lagena in the coiled cochlea of Didelphis. (See Larsell et al ('35), p. 102 for a figure of the coiled cochlea of Didelphis.) Fernandez and Schmidt ('63) conclude: "The primitive nature of the Opossum and the numerous close similarities of its inner ear to that of placental mammals make it improbable that the coiled cochlea of these 2 groups evolved independently." The same authors do indicate that the middle ear of the opossum "differs considerably from that of placental species." They were referring to the lack of a bulla in Didelphis. Ride ('62) reviews the problem of marsupial and placental middle ears, and concludes that while there are basic differences, "a structural ancestor common to both insectivores and marsupials is not unlikely; such an ancestor would most likely have occurred soon after the evolution of the mammalian auditory system." Certainly, the absence of a bulla is no indication of differences between the 2 groups, for a bulla is absent in many species of placentals as well (e.g. Anderson and Jones, '67). 40 Harrison and Warr ('62) stated that the cochlear nuclei of Didelphis and placentals are quite different. Casual examination of the cat (Felis), rat and marsupial cochlear nuclei indicated to the author of this paper that the statement of Harrison and Warr is unjustified. M. Merzenich (U. of Wisconsin, Madison) has a paper in press concerning the cochlear nuclei of a series of marsupials and placentals, which should help resolve this problem. The nuclei of the SO are related insofar as they are all likely concerned with audition. When considered in detail, however, it appears that functionally (as well as structurally), separate structures are involved (e.g. Irving and Harrison, '67). Therefore when the phylogenesis of the SO is being considered, each component nucleus should be considered separately; if the superior olivary complexes of marsupials on the one hand and placentals on the other are believed to be derived independently, then it must also be accepted that not one, but possibly five structures (nuclei) have been developed independently of one another. Functional morphology.--As stated in the introduc- tion, the SO from a behavioral standpoint is very important. This point of View is substantiated by the auditory system of the kangaroo rat, as described by Webster et a1 ('68). Here the greatest morphological change in the central auditory system was found in the nuclei of the medulla (including SO). Webster et al ('68) conclude that this is 41 to be expected in an animal in which auditory reflexes are important in predator avoidance, since the SO is in close synaptic relationship with the external environment. The study of Webster et al ('68) in conjunction with the work of Irving and Harrison ('67) indicate that the SO has been under intense selective pressure in mammalian evolution. Phylogeny.--When studying the marsupial-placental dichotomy, it must be kept in mind that the "American group of marsupials has probably been genetically isolated from the Australian group since at least the early Paleocene, and quite possibly since the end of the Cretaceous. Thus any character found in common between the two groups has either been independently derived, or is truly ancient, probably dating back to the Mesozoic." (Lillegraven, '69, p. 90). From this, if the superior olivary complexes of only Didelphis and the Australian marsupials are considered to be homologous, one is forced back to at least Mesozoic times for its first appearance within the marsupials. If the SO was an independent development in the Eutheria, a Mesozoic origin for the SO would also have to be postulated for this group. This conclusion is based on two facts. One, placentals underwent an initial radia- tion in the Late Cretaceous and had by that time differen- tiated into 2 distinct lines (Lillegraven, '69). Secondly, in living placentals, the SO is present in characteristic 42 form in all species examined. Thus there has either been profound parallelism or far more likely, a common origin for the SO in the Mesozoic. Thus after considering the structures of the ear, including the inner ear, the cochlear nuclei, and the su- perior olivary complex of 5 separate nuclei and finding them to be structurally very' much alike in Didelphis, Australian marsupials and placentals, it is concluded that the auditory system characteristic of these mammals has likely evolved from a common ancestor. In addition, be- cause of its general predominance, the superior olivary complex of 5 component nuclei is believed the primitive condition; therefore animals such as microchiropterans and cetaceans with less than 5 nuclei would have lost com- ponents in their evolutionary past. Considering their ecological role in Mesozoic times, it is believed that the auditory system characteristic of mammals, behaviorally highly significant, evolved early in mammalian history as a prerequisite to survival. Monotremata: an independent development of the mammalian auditory system? In connection with hearing both in Mesozoic mammals and their therapsid ancestors, it could prove interesting to thoroughly study the auditory system of monotremes. Monotremes are now believed to have had a derivation from 43 therapsids independent of marsupials and placentals (e.g. Kermack, '67). The studies of Abbie ('34) on the echidna and Hines ('29) on the platypus are not detailed enough to be used in comparison to therian mammals. Monotremes do possess a lagena in their uncoiled cochlea, contrary to all therian mammals studied (Fernandez and Schmidt, '63). SUMMARY The superior olivary complex of Australian marsupials was found to exhibit the same characteristics of differen- tiation as that of Didelphis and Rattus. Close similarity in morphology of the SO between eutherian and metatherian mammals was taken as evidence for its common evolutionary origin. In addition, from the work of other authors, evidence is presented for a common evolutionary origin of other structures of the mammalian auditory system: cochlear nuclei, auditory ossicles and skull bones of the otic region. Thus a final conclusion is that an auditory system typical of the Mammalia developed early in evolution, at least within the therian ancestors of marsupials and placentals. In order to shed further light on this problem, studies are suggested on the Monotremata, mammals believed to have de- veloped independently from reptiles, and mammals already known to have peculiarities in the structure of their skull bones and cochlea. LIST OF REFERENCES Abbie, A.A. 1934 The brain stem and cerebellum of Echidna aculeata. Philos. Transact. Roy. Soc. Lond., BioI. ser., :24: 1-740 Anderson, S., and J. K. Jones, Jr. (ed.) 1967 Recent Mammals of the World. A Synopsis of Families. Ronald Press Co., New York, pp. 453. Ariens Kappers, C.V., G.C. Huber and E.C. Crosby 1936 The Comparative Anatomy of the Nervous System of Verte- brates, Including Man. Reprinted in 1960 by Hafner Pub. Co., New York. Barnes, W.T., H.W. Magoun and S.W. Ranson 1943 Ascending auditory pathways in the brainstem of the monkey. J. Comp. Neur., 79: 129-152. Berman, A.L. 1968 The Brain Stem of the Cat. University of Wisconsin Press, Madison, pp. 175. Boudreau, J.C., and C. Tsuchitani 1968 Binaural interaction in the cat superior olive S. segment. J. Neuro- physiol., 31: 442-454. Clemens, W.A. 1968 Origin and early evolution of marsupials. EVOlo' 22: 1-180 Edinger, T. 1949 Paleoneurology versus comparative brain anatomy. Confinia Neurologica, 9: 5-24. Fernandez, C., and R.S. Schmidt 1963 The opossum ear and evolution of the coiled cochlea. J. Comp. Neur., 121: 151-159. Galambos, R., J. Schwartzkopff and A. Rupert 1959 Microelectrode study of superior olivary nuclei. Am J. Physiol., 197: 527-536. Goldberg, J.M., and P.B. Brown 1968 Functional organization of the dog superior olivary complex: an anatomical and electrophysiological study. J. of Neurophysiol., 31: 639-656. 44 45 1969 Response of binaural neurons of dog superior olivary complex to dichotic tonal stimuli: some physiological mechanisms of sound localization. J. Neurophysiol., 32: 613-636. Harrison, J.M., and M.D. Beecher 1967 Medial superior olive and sound localization. Science, 155: 1697. Harrison, J.M., and R. Irving 1964 Nucleus of the trapezoid body: dual afferent innervation. Science, 143: 473-474. Harrison, J.M., and R. Irving 1965 The anterior ventral cochlear nucleus. J. Comp. Neur., 124: 15-42. 1966a Ascending connections of the anterior ventral cochlear nucleus in the rat. J. Comp. Neur., 126: 51-64. 1966b The organization of the posterior ventral cochlear nucleus in the rat. J. Comp. Neur., 126: 391-403. 1966c Visual and non-visual auditory systems in mammals. Science, 154: 738-743. Hines, M. 1929 The brain of Ornithorhynchus Anatinus. Philos. Transact. Roy. Soc. Lond.,fiIol. Ser., 217: 155-207. Irving, R., and J.M. Harrison 1967 The superior olivary complex and audition: a comparative study. J. Comp. Neur., 130: 77-86. Keast, A. 1968 Evolution of mammals on southern continents. IV. Australian mammals: zoogeography and evolution. Quart. Rev. Biol., 43: 373-408. Kermack, K.A. 1967 The interrelations of early mammals. J. Linn. Soc. (Zool.), 47: 241-249. Larsell, O., E. McCrady and A. Zimmerman 1935 Morphology and functional development of the membranous labyrinth in the opossum. J. Comp. Neur., 63: 95-118. Lillegraven, J.A. 1969 Latest Cretaceous mammals of upper part of Edmonton Formation of Alberta, Canada, and review of marsupial-placental dichotomy in mammalian evolution. U. of Kansas Paleo. Cont., Article 50 (Vertebrata 12), pp. 122. 46 Masterton, B.T., and I.T. Diamond 1967 Medial superior olive and sound localization. Science, 155: 1696- 1697. Masterton, B.T., J.A. Jane and I.T. Diamond 1967 Role of brainstem auditory structures in sound localization. I. Trapezoid body, superior olive and lateral lemniscus. J. Neurophysiol., 30: 341-359. Meesen, H., and J. Olszewski 1949 A cytoarchitectonic atlas of the rhombencephalon of the rabbit. S. Karger, Basel and New York, pp. 52. Morest, D.K. 1964 The laminar structure of the inferior colliculus of the cat. Anat. Rec., 148: 314. Olson, E.C. 1959 The evolution of mammalian characters. Evol., 13: 344-353. Oswaldo-Cruz, E., and C.E. Rocha-Miranda 1968 The Brain of the opossum (Didelphis marsupialis). Instituto de Biofisica, Universidade-Federal d6 Rio de Janeiro, Rio de Janeiro, Brazil, pp. 99. Papez, J.W. 1930 Superior olivary nucleus. Its fiber connections. Arch. of Neurol. and Psychiatry, 24: 1-200 Ramon y Cajal, S. 1909 Histologie du systeme nerveux de l'homme et des vertebres, vol. 1. Paul Masson Co., Paris, pp. 795-816. Rasmussen, G.L. 1946 The olivary peduncle and other fiber projections of the superior olivary complex. J. Comp. Neur., 84: 141-219. Ride, W.D.L. 1962 On the evolution of Australian marsupials. pp. 281-306 in The Evolution of Living Organisms. G.W. Leeper TEd.), Melbourne U. Press, Victoria. Sehmsdorff, J. 1966 The primate superior olivary complex. Anat. Rec., 154: 421. Shaner, R.F. 1934 The development of the nuclei and tracts related to the accoustic nerve in the pig. J. Comp. Neur., 60: 5-19. Simpson, G.G. 1959 Mesozoic mammals and the polyphyletic origin of mammals. Evol., 13: 405-414. 47 Stokes, J.H. 1912 The acoustic complex and its relation in the brain of the opossum (Didelphys virginiana). Amer. J. Anat. I 12: 401-4450 Stotler, W.A. 1953 An experimental study of the cells and connections of the superior olivary complex of the cat. J. Comp. Neur., 98: 401-431. Tumarkin, A. 1968 Evolution of the auditory conducting apparatus in terrestrial vertebrates. pp. 18-40 in Hearing Mechanisms in Terrestrial Vertebrates, CIba Foundation Symposium. A.V.S. De Reuck and J. Knight (ed.), Little, Brown & Co., Boston. Voris, H.C., and N.L. Hoerr 1932 The hindbrain of the opossum, Didelphis virginiana. J. Comp. Neur., 54: 277-355. Webster, D.B., R.F. Ackermann and G.C. Longa 1968 Central auditory system of the kangaroo rat, Dipodomys merriami. J. Comp. Neur., 133: 477-493. Zvorykin, V.P. 1964 Morphological substrate of ultrasonic and locational capacities in the dolphin. Trans. suppl. of Fed. Proc., 23: 647-654.