Neuroscience Letters, 146 (1992) 91-95 91 © 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00

NSL 09047

The primary auditory cortex in cetacean and human brain: a comparative analysis of neurofilament protein-containing pyramidal neurons

P a t r i c k R. H o f a'b, I l y a I. G l e z e r c, N a n c y A r c h i n a, W i l l i a m G . J a n s s e n ~, P e t e r J. M o r g a n e d a n d J o h n H . M o r r i s o n "'b

"Fishberg Research Center for Neurobiology and bDepartment of Geriatrics and Adult Development, Mount Sinai School of Medicine, New York, N Y 10029 (USA), CDepartment of Cell Biology and Anatomical Sciences, CUN Y Medical School, New York, N Y 10031 (USA) and dNeurobiology

Laboratory, Worcester Foundation for Experimental Biology, Shrewsbury, MA 01545 (USA)

(Received 20 May 1992; Accepted 23 July 1992)

Key words: Dolphin; Whale; Cetacean brain; Primate brain; Human brain; Brain evolution; Neurofilament protein; Quantitative neuroanatomy

To extend our investigation of the anatomy of sensory systems in highly adapted aquatic and terrestrial mammals, we have analyzed the distribu- tion of a particular population of efferent neurons in the cetacean and human primary auditory cortex using an antibody to non-phosphorylated neurofilament protein (SMI32). The neurofilament protein triplet is differentially distributed within neuronal subpopulations in the primate and cetacean neocortex. In primates, it appears that the somatodendritic domain of a subset of pyramidal neurons furnishing specific corticocortical connections contains high concentrations of neurofilament protein. In the human primary auditory cortex these neurons are located in layers III, V and VI, whereas in cetaceans they are concentrated almost exclusively in the cortical efferent layer IIIc/V. Previous analyses have shown that SMI32 immunoreactivity in the cetacean neocortex is uniformly distributed among functionally different areas, while in human neocortex, the distribution of SMI32-positive neurons exhibit a high degree of regional and laminar specialization that is correlated with the functional and anatomical diversity of the cortical areas. In addition, the overall distribution of SMI32°immunoreactive neurons in the cetacean neocortex is comparable to that observed in paralimbic areas of the human, suggesting that the cetacean neocortex has retained many features of phylogenetically older cortical regions.

Cetaceans are thought to have originated f rom terres- trial ca rn ivorous m a m m a l s tha t made a t ransi t ional ad- ap ta t ion f rom land to the mar ine milieu dur ing the ear ly Eocene [7]. Mode rn whales are a g roup o f aquat ic m a m - mals tha t have achieved an extremely high degree o f ad- ap ta t ion to their na tura l env i ronment [ t l , 16, 21-23]. However , as c o m p a r e d to mos t m o d e r n terrestrial m a m - mal ian species, the cetacean central nervous system dem- onstra tes striking organizat ional differences [1, 5, 6, 9 - 12, 16, 20-22]. In part icular , the s tructure o f the cerebral cor tex differs no tab ly f rom tha t o f highly evolved land m a m m a l s such as pr imates , in that it has retained some features tha t are observed only in non-progress ive terres- trial species such as bats and hedgehogs [4, 5, 8, 11, 19- 21, 28]. The cetacean neocor tex is thin and is charac- terized by a poor ly developed lamina t ion pa t te rn with an overall reduct ion in granular iza t ion and incipient layer IV, a very thick layer I, ext raver ted neurons in layer II , and a dense band o f large pyramida l neurons referred to

Correspondence: ER. Hof, Fishberg Research Center for Neurobiol- ogy, Mount Sinai School of Medicine, Box 1065, One Gustave L. Levy Place, New York, NY 10029, USA. Fax: (1) (212) 996-9785.

as layer I I Ic /V [1, 5, 6, 10, 11, 16, 17, 21, 22]. In addit ion, there are no definite ana tomica l boundar ies between funct ional ly separa te neocort ical areas, since these cy- toarchi tectural features are observed th roughou t the ne- ocor tex [1, 11, 16, 21, 23].

In a previous s tudy o f p r imary visual cortex in a series o f cetaceans, we found tha t the typology, l aminar distri- but ion and cell density o f GABAerg ic in terneurons dif- fered significantly f rom tha t observed in p r imary visual areas o f p r imates and rodents [8, 9, 12]. In addit ion, we have shown tha t the cetacean p r imary visual cor tex con- tains a par t icular popu la t ion o f large pyramida l neurons in layer I I Ic /V tha t is s t rongly labeled by a monoc lona l an t ibody (SMI-32) to a non -phosphory l a t ed epi tope on the m e d i u m and heavy molecular weight subunits o f the neurof i lament prote in triplet [8, 9]. The neurof i lament prote in tr iplet has been shown to be present only in par - t icular neurona l subpopula t ions [29]. In fact, in the mon- key and h u m a n neocor tex this an t ibody labels a distinct popu la t ion o f pyramida l neurons that exhibits a specific regional and laminar dis tr ibut ion [2, 3, 13-15, 17, 25]. These neurons are also known to furnish specific subsets o f cort icocort ical connect ions between visual areas and

92

Fig. 1. Distribution of SMI32-ir neurons in the primary auditory cortex of the beluga whale Delphinapterus leucas (A,B) and human (C). Note the large SMI32-ir pyramidal cells sending apical dendrites to layer I in layers Illc/V of the cetacean species (A,B), and the contrasting laminar distribution of these neurons in the human auditory cortex (C). The arrow in A denotes a lightly stained isolated pyramidal neuron in layer l l lab of the beluga whale. B: a high magnification of two typical SMI32-ir pyramidal neurons from A. Similar staining patterns were observed in the

bottlenose dolphin. Layers are indicated by Roman numerals; WM, white matter. Bar = 100 pm (A,C) and 50 pm (B).

93

between high order association systems in the monkey [2, 17, 25]. In order to further characterize the differential organization and circuitry of sensory systems in aquatic mammals and primates, we have performed a compara- tive analysis of SMI-32-immunoreactive (SMI32-ir) py- ramidal neurons in the primary auditory area of two ce- tacean species and of humans.

The brains of a beluga whale (Delphinapterus leucas,

ca. 35 years old) and of a bottlenose dolphin (Tursiops

truncatus, ca. 15-17 years old) were obtained from the Naval Ocean Systems Center, San Diego (beluga whale), and from the Department of Zoology, Moscow State University, Russia (bottlenose dolphin). The terminally ill animals were deeply anaesthetized by an overdose of valium and demerol (beluga whale) or nembutal (bot- tlenose dolphin) and perfused with 4% paraformalde- hyde/0.5% glutaraldehyde through the abdominal aorta. This perfusion method is necessary since the entire blood supply to the cetacean brain comes through a spinal rete mirabile system fed directly from the descending aorta [19]. Human brains were obtained at autopsy from 4 pa- tients with no history of neurological or psychiatric dis- orders (49, 72, 79 and 85 years old, post mortem delay 2-6 h, all from the Institute of Biogerontology, Sun City, AZ). The human brains were perfused ex situ through the internal carotid and basilar arteries with cold 4% paraformaldehyde. The brains were subsequently sus- pended by the basilar artery in the same fixative for 12 h, cut into 1-cm-thick coronal blocks, and postfixed for 48 72 h. After postfixation, the blocks were washed in a se- ries of graded sucrose solutions in cold phosphate-buff- ered saline (PBS) [3, 13, 15]. Samples from the area deter- mined physiologically as the primary auditory cortex of the whale and dolphin (posterior suprasylvian gyrus, be- tween the suprasylvian and entolateral sulci [5, 18, 27]) and from the right Heschl gyrus in human brains were frozen and cut at 40 pm on a crytostat. The 40-pm-thick sections were incubated overnight at 4°C with the mono- clonal antibody SMI-32 (Sternberger Monoclonals) at working dilutions of 1:3,000 (cetaceans) to l:10,000 (human) in PBS containing 0.3% Triton X-100, and 0.5 mg/ml bovine serum albumin. Following incubation, the sections were processed by the avidin-biotin method using a Vectastain ABC kit (Vector Laboratories) and diaminobenzidine. Finally, the immunoreactivity was in- tensified with osmium. Adjacent sections were stained with Cresyl violet in order to clarify the cytoarchitecture. All the sections were systematically surveyed and ana- lyzed using a computer-assisted image analysis system consisting of a Zeiss Axiophot photomicroscope equipped with a motorized stage, a high sensitivity cam- era, a DEC 3100 workstation and Macintosh II micro- computer, and custom software. On each slide, SMI32-ir

neurons were counted in each layer separately in the pri- mary auditory cortex in the cetacean and human materi- als in a series of 10-20 one-ram-wide cortical traverses parallel to the pial surface. Perikaryal size was measured at the same time [13]. Three to five slides were analyzed in the two cetacean species and in each human case.

A distinctive feature of the distribution of SMI32-ir neurons in the cetacean neocortex is their concentration in the efferent layer IIIc/V [8, 9]. This typical pattern is observed in the primary auditory cortex of two species reported in the present study (Fig. 1A). Layer IIIc/V is characterized by a unique row of intensely SMI32-1a- beled, large pyramidal neurons (Fig. 1A,B), ranging in size from 470 to 1,190 t i m E . These neurons have darkly stained apical dendrites that reach the upper one-third of layer I (Fig. 1A), and are frequently grouped in clusters of 5-10 cells. These clusters span over 500-800 ~tm of cortex and are spaced by 200-300/lm intervals. It should be noted that not all pyramidal neurons in layer IIIc/V are SMI32 positive and that the labeled neurons are thus likely to represent a distinct subset of efferent neurons in the cetacean neocortex. There is no difference in SMI32- ir neuron density between the two cetacean species (Table I). The other layers of the cetacean neocortex are characterized by much lower densities of SMI32-ir neu- rons. A few small and lightly labeled pyramidal and polymorphic neurons are encountered in layer VI, while there are very few immunoreactive neurons in layers II and IIIab (Table I, Fig. 1A). Layer I contains no SMI32- ir neurons. Finally, there is a denser plexus of im- munoreactive fibers in layers IIIc/V and VI as compared to layers II and IIIab.

The human primary auditory cortex displays a strik- ingly different distribution of SMI32-ir neurons as com- pared to that observed in cetaceans (Fig. 1C, Table I). SMI32-ir neurons are present throughout the entire thickness of layer III and a few labeled neurons are ob- served in layer II (Fig. 1C). A particular feature of the human primary auditory cortex is the presence of numer- ous SMI32-ir dendritic bundles in layer IIIa-c (Fig. 1C). The largest SMI32-ir neurons are found in layer IIIbc, with a perikaryon size ranging from 250 to 530/~m E. Layer IV contains rare labeled neurons. Layers V and VI are distinguished by the presence of high densities of small and lightly stained pyramidal neurons and very low counts of large pyramidal neurons (Fig. 1 C, Table I). Layer I has no SMI32-ir profiles, and there is an even staining of the neuropil in layers II to VI.

The main differences between the distribution of SMI32-ir neurons in the cetacean and human primary auditory cortex are the differential laminar localization of these cells (layer III in human and layer IIIc/V in the cetaceans), the overall smaller size of SMI32-ir neurons

94

TABLE I

SMI32-IMMUNOREACTIVE NEURON COUNTS IN THE CETA- CEAN AND HUMAN PRIMARY AUDITORY CORTEX

Results represent counts per mm of cortical traverse and are expressed as means _+ S.E.M. Counts from the 4 human brains were pooled as only marginal differences were observed between the 4 cases. Layers are indicated by Roman numerals in cetaceans and human separately. There are no SMI32-ir neurons in layer I in both cetaceans and human. The cetacean neocortex has no layer IV. See text for details.

Beluga Bottlenose Human whale dolphin

1 - - l

II 1.0+0.3 II 1.6+0.3 IIIab 1.4+0.3 1.0_+0.3 llIa~c 43. l _+2.4 -- -- -- IV 1.1+_0.1 Illc/V 21.3_+1.0 19.0_+0.9 V 33.7+1.7 VI 6.4_+0.4 5.3_+0.5 VI 28.7+2.4

in the human (303.1 + 12.6/,/m 2 in human vs. 779.7 + 34.2 ],/m 2 in cetacean), and the near complete absence of labeled cells in the other cortical layers of the cetacean brains. It is also interesting to note that a previous analy- sis of the cetacean sensory neocortex has revealed that the global pattern of SMI32 immunoreactivity is similar across functionally different neocortical areas [8, 9]. For instance, the laminar distribution of SMI32-ir neurons in the primary auditory cortex is similar to that observed in the primary visual cortex in several cetacean species in- cluding beluga whale (Delphinapterus leucas), pilot whale (Globicephala melaena), bottlenose dolphin (Tursiops truncatus), striped dolphin (Stenella coeruleoalba) and common porpoise (Phocoena phocoena) [8, 9]. In con- trast, the distribution and density of SMI32-ir neurons in the monkey and human neocortex reveals clear patterns of regional and laminar specialization [2, 3, 13-15, 17, 25]. This histochemical pattern in the cetacean neocortex correlates with the fact that conservative evolutionary features dominate the cytoarchitectural organization of toothed whales [8, 9, 21, 23]. This can be interpreted as a biochemical and morphological evidence of the proto- typical characteristics of the cetacean neocortex as com- pared to evolutionarily progressive land mammals. Thus, although primary visual and auditory areas have distinctly different laminar patterns of SMI32 staining in the primate [3], SMI32-ir neurons are present in all layers that contain pyramidal neurons [3, 17], whereas in the cetacean they are restricted to a subpopulation of very large pyramidal neurons of layer IIIc/V. A comparable distribution of SMI32-ir pyramidal cell in the primate cortex is found only in paralimbic transition areas (i.e. between phylogenetically older archicortex and neocorti- cal regions), such as anterior cingulate cortex, posterior

orbitofrontal and parainsular cortex [3, 14], where these cells are predominantly located within layer V. This sug- gests that the chemoarchitectural organization of the ne- ocortex in cetacean species has retained some of the fea- tures that are found only in proisocortical regions in pro- gressive mammalian species [3, 8].

It is worth noting that both primate and cetacean cor- tical efferent neurons seem to be strongly labeled by the SMI-32 antibody [2, 3, 8, 9, 17, 25]. Previous tract-trac- ing studies in the monkey have demonstrated that there is a high correlation between the distribution of these neurons and that of specific corticocortical projections [2, 17, 25]. The preferential localization of SMI32-ir neu- rons in layer IIIc/V suggests that these neurons may sub- serve a comparable role in furnishing corticocortical connections in sensory areas of the cetacean brain, in that their distribution correspond to what is considered to be the main neocortical efferent layer in these species.

A precise cellular role for the neurofilament protein triplet remains elusive but it has been claimed to play a major role in regulating axonal diameter and in the stabi- lization of the cytoskeleton [2, 3, 25, 26, 29]. Thus, in both primates and cetaceans, the functional integrity of neocortical efferent neurons may require a high degree of stabilization of their somatic and dendritic domains. Also, this biochemical specialization of neocortical ef- ferent neurons is likely to represent an early feature in mammalian phylogeny. Although cetaceans and pri- mates may use different information processing strate- gies, it is relevant that the cellular functions subserved by the neurofilament protein triplet may constitute a bio- chemical attribute that was conserved in the evolution of neocortical efferent projection systems. These data sug- gest that subsets of neurons providing crucial functions in neocortical circuitry share certain aspects of their bio- chemical phenotype in species that have evolved to a high level of adaptation to their particular environment, despite the contrast between cytoarchitectonic features and cellular morphology of sensory areas in cetaceans and primates including human.

We thank Drs. S. Ridgeway and T.F. Ladygina for providing the cetacean brains, Dr. J. Rogers for provid- ing the human brains, Dr. W.G. Young for software de- velopment, Drs. M.J. Campbell, J.C. Vickers and E.A. Nimchinsky for helpful discussion, and R. Woolley for expert technical assistance. Supported by The Brookdale Foundation, the American Health Assistance Founda- tion, and NIH AG06647, NIH Grant for MBRS-CRS of CUNY, PSC-CUNY-662232, and NSF BNS-89-03717.

1 Breathnach, A.S,, The cetacean nervous system, Biol. Rev., 35 (1960) 187 230.

2 Campbell, M.J., Hof, P.R. and Morrison, J.H., A subpopulation of primate corticocortical neurons is distinguished by somatodendritic distribution of neurofilament protein, Brain Res., 539 (1991) 133- 136.

3 Campbell, M.J. and Morrison, J.H., Monoclonal antibody to neu- rofilament protein (SMI-32) labels a subpopulation of pyramidal neurons in the human and monkey neocortex, J. Comp. Neurol., 282 (1989) 191-205.

4 Ferrer, I., The basic structure of the neocortex in insectivorous bats (Miniopterus sthreibersi and Pipistrellus pipistrellus). A Golgi study, J. Hirnforsch., 28 (1987) 237-243.

5 Ferret, I. and Perera, M., Structure and nerve cell organisation in the cerebral cortex of the dolphin Stenella coeruleoalba: a Golgi study - with special attention to the primary auditory area, Anat. Embryol., 178 (1988) 161-173.

6 Garey, L.J., Winkelmann, E. and Brauer, K., Golgi and Nissl stud- ies of the visual cortex of the bottlenose dolphin, J. Comp. Neurol., 240 (1985) 305-321.

7 Gingerich, P.D., Wells, N.A., Russell, D.E. and Shah, S.M.I., Ori- gin of whales in epicontinental remnant seas: new evidence from early Eocene of Pakistan, Science, 220 (1983) 403-406.

8 Glezer, I.I., Hof, P.R., Janssen, W., Morrison, J.H., Leranth, C. and Morgane, P.J., Comparative immunohistochemistry of cy- toskeletal proteins in visual cortex of cetaceans and primates, Soc. Neurosci. Abstr., 17 (1991) 846.

9 Glezer, I.I., Hof, P.R., Leranth, C. and Morgane, P.J., Morpholog- ical and histochemical features of cetacean sensory neocortex: im- munocytochemical analysis of pyramidal and non-pyramidal popu- lations of neurons. In J. Thomas, R. Kastelein and A. Supin (Eds.), Sensory Abilities in Aquatic Mammals, Plenum, New York, in press.

10 Glezer, I.I., Hof, P.R. and Morgane, P.J., Calretinin-immunoreac- tive neurons in the dolphin and human primary visual cortex, Brain Res., in press.

11 Glezer, I.I., Jacobs, M.S. and Morgane, P.J., The 'initial' brain con- cept and its implications for brain evolution in cetacea, Behav. Brain Sci., 11 (1988) 75-116.

12 Glezer, I.I., Morgane, P.J. and Leranth, C., Immunohistochemistry of neurotransmitters in visual cortex of several toothed whales: light and electron microscopic study. In J. Thomas and R. Kastelein (Eds.), Sensory Abilities of Cetaceans: Laboratory and Field Evi- dence, Plenum, New York, 1990, pp. 39-66.

13 Hof, P.R., Cox, K. and Morrison, J.H., Quantitative analysis of a vulnerable subset of pyramidal neurons in Alzheimer's disease: I. Superior frontal and inferior temporal cortex, J. Comp. Neurol., 301 (1990) 44-54.

14 Hof, P.R., Hsu, P. and Morrison, J.H., Quantitative che- moarchitectonic analysis of the cingulate cortex in Alzheimer's dis- ease, Soc. Neurosci. Abstr., 17 (1991) 693.

15 Hof, P.R. and Morrison, J.H., Quantitative analysis of a vulnerable subset of pyramidal neurons in Alzheimer's disease: II. Primary and secondary visual cortex, J. Comp. Neurol., 301 (1990) 55~4.

95

16 Kesarev, V.S., Malofeyeva, L.I. and Trykova, O.V., Ecological specificity of cetacean neocortex, J. Hirnforsch., 18 (1977) 447-460.

17 Kupferschmidt, S.B., Hof, ER. and Morrison, J.H., Corticocortical connections in macaque visual cortex exhibit differential patterns of neurofilament protein distribution, Soc. Neurosci. Abstr., 17 (1991) 845.

18 Ladygina, T.F., Mass, A.M. and Supin, A., Multiple sensory pro- jections in the dolphin cerebral cortex, Zh. Vyssh. Nervn. Deyat. im. I.E Pavlova, 28 (1978) 1047-1058.

19 McFarland, W.L., Jacobs, M.S. and Morgane, EJ., Blood supply to the brain of the dolphin Tursiops truncatus, with comparative observations on special aspects of the cerebrovascular supply of other vertebrates, Neurosci. Biobehav. Rev., 3, Suppl. l (1979) 1 93.

20 Morgane, P.J. and Jacobs, M.S., Comparative anatomy of the ceta- cean nervous system. In R.J. Harrison (Ed.), Functional Anatomy of Marine Mammals, Academic Press, London, 1972, pp. 117-244.

21 Morgane, EJ., Glezer, I.I. and Jacobs, M.S., Comparative and evo- lutionary anatomy of the visual cortex of the dolphin. In E.G. Jones and A. Peters (Eds.), Cerebral Cortex, Vol. 8B, Comparative Struc- ture and Evolution of Cerebral Cortex, Part II, Plenum, New York, 1990, pp. 215-258.

22 Morgane, EJ. and Glezer, I.I., Sensory neocortex in dolphin brain. In J. Thomas and R. Kastelein (Eds.), Sensory Abilities of Ceta- ceans: Laboratory and Field Evidence, Plenum, New York, 1990, pp. 107-136.

23 Morgane, EJ., Jacobs, M.S. and Galaburda, A.M., Conservative features of neocortical evolution, Brain Behav. Evol., 26 (1985) 176 184.

24 Morris, J.R. and Lasek, R.J., Stable polymers of the axonal cy- toskeleton: the axoplasmic ghost, J. Cell Biol., 92 (1982) 192 198.

25 Morrison, J.H., Hof, ER., Kupferschmid, S.B., Good, EF., Jans- sen, W. and Archin, N., Relationship between connectivity and cy- toskeletal profile of corticocortically-projecting neurons, Soc. Neu- rosci. Abstr., 17 (1991) 1021.

26 Smith, D.E., The location of neurofilaments and microtubules dur- ing the postnatal development of Clarke's nucleus in the kitten, Brain Res., 55 (1973) 41-53.

27 Sokolov, V.E., Ladygina, T.F. and Supin, A., Localization of the sensory zones in the cerebral cortex of the dolphin, Dokl. Akad. Nauk SSSR, 202 (1972) 490-493.

28 Valverde, F., A comparative approach to neocortical organization based on the study of the brain of the hedgehog (Erinaceus eu- ropaeus). In S. Grisolia, C. Guerri, F. Samson, S. Norton and E Reinoso-Suarez (Eds.), Ram6n y Cajal's Contribution to the Neu- rosciences, Elsevier, Amsterdam, 1983, pp. 149-170.

29 Vickers, J.C. and Costa, M., The neurofilament triplet is present in distinct subpopulations of neurons in the central nervous system of the guinea-pig, Neuroscience, 49 (1992) 73-100.