In contrast to many other mammals, cetaceans have relatively small hippocampi that appear to lack adult neurogenesis

Abstract

The hippocampus is essential for the formation and retrieval of memories and is a crucial neural structure sub-serving complex cognition. Adult hippocampal neurogenesis, the birth, migration and integration of new neurons, is thought to contribute to hippocampal circuit plasticity to augment function. We evaluated hippocampal volume in relation to brain volume in 375 mammal species and examined 71 mammal species for the presence of adult hippocampal neurogenesis using immunohistochemistry for doublecortin, an endogenous marker of immature neurons that can be used as a proxy marker for the presence of adult neurogenesis. We identified that the hippocampus in cetaceans (whales, dolphins and porpoises) is both absolutely and relatively small for their overall brain size, and found that the mammalian hippocampus scaled as an exponential function in relation to brain volume. In contrast, the amygdala was found to scale as a linear function of brain volume, but again, the relative size of the amygdala in cetaceans was small. The cetacean hippocampus lacks staining for doublecortin in the dentate gyrus and thus shows no clear signs of adult hippocampal neurogenesis. This lack of evidence of adult hippocampal neurogenesis, along with the small hippocampus, questions current assumptions regarding cognitive abilities associated with hippocampal function in the cetaceans. These anatomical features of the cetacean hippocampus may be related to the lack of postnatal sleep, causing a postnatal cessation of hippocampal neurogenesis.

Fig. 1

Graphical representation of the relationship between brain volume and hippocampal volume (a) and brain volume minus hippocampal volume and hippocampal volume (b) across 367 mammalian species. Note, in contrast to previous studies (Finlay and Darlington 1995; Reep et al. 2007), a function that approximates an exponential curve describes the data most efficiently and potentially reflects the presence of adult hippocampal neurogenesis in most mammalian species. Note that the hippocampal volumes of the West Indian manatee (Trichechus manatus), river hippopotamus (Hippopotamus amphibius) and African elephant (Loxodonta africana), which were not used in the determination of the descriptive function, fall within either the 95 % confidence intervals (dark grey shading) or the 95 % prediction intervals (light grey shading) determined from the data. In all cases, the cetaceans examined, harbour porpoise (Phocoena phocoena), bottlenose dolphin (Tursiops truncatus), Atlantic white-sided dolphin (Lagenorhynchus acutus) and minke whale (Balaenoptera acutorostratus), have hippocampal volumes substantially smaller than what would be predicted based on brain volume. AICC Akaike’s information criteria, Bv brain volume, Bv - Hv brain volume minus hippocampal volume, DOF degrees of freedom, Hv hippocampal volume

Fig. 2

Graphical representation of the relationship between brain volume and amygdala volume (a) and brain volume minus amygdala volume and amygdala volume (b) across 364 mammalian species. Note that similar to previous studies (Finlay and Darlington 1995; Reep et al. 2007), a linear function describes the data most efficiently. Note that the amygdala volumes of the West Indian manatee (Trichechus manatus), river hippopotamus (Hippopotamus amphibius) and African elephant (Loxodonta africana), which were not used in the determination of the linear function, fall within either the 95 % confidence intervals (dark grey shading) or the 95 % prediction intervals (light grey shading) determined from the data. In all cases the cetaceans examined, harbour porpoise (Phocoena phocoena), bottlenose dolphin (Tursiops truncatus) and minke whale (Balaenoptera acutorostratus), have amygdala volumes substantially smaller than what would be predicted based on brain volume, reflecting the loss, or near loss, of the olfactory system in cetaceans. AICC Akaike’s information criteria, Av amygdala volume, Bv brain volume, Bv - Av brain volume minus amygdala volume, DOF degrees of freedom

Fig. 3

Graphical representation of the relationship between brain volume minus hippocampal volume and hippocampal volume (a) and brain volume minus amygdala volume and amygdala volume (b) across mammalian species showing the contrast between the exponential function (pink shading) and the linear function (blue shading) describing these relationships. Note that the exponential function provides a more appropriate fit of the data for the hippocampus (a), while the linear function provides a more appropriate fit of the data for the amygdala (b). AICC Akaike’s information criteria, DOF degrees of freedom

Fig. 4

Plots of the residuals obtained using both linear and non-linear regression functions to describe the relationship between brain minus hippocampal volume and hippocampal volume (upper two plots) and between brain minus amygdala volume and amygdala volume (lower two plots). The residuals as based on the linear model for the hippocampus are not randomly scattered about zero as confirmed by a runs test, while both visual and statistical comparison of the nonlinear model for the hippocampus confirms its appropriateness for this data. While both linear and exponential models describe the amygdala volume well, the less scatter observed in the linear model indicates the appropriateness of this model for the amygdala data

Fig. 5

Graphical representation of the phylogenetically correct least-square regression and associated confidence and prediction intervals for the brain volume compared to hippocampal volume (a) and brain volume minus hippocampal volume compared to hippocampal volume (b). The resultant coefficient of determination for these models is r² = 0.85/0.83 with slopes of 0.77/0.75. These plots indicate that even after phylogenetic correction, the cetaceans lie well below the confidence and prediction intervals of the mammalian regression, underscoring the small size of the cetacean hippocampus. In addition, the non-linearity of the mammalian hippocampal data is also evident in these plots despite correction for phylogenetic relationships

Fig. 6

Higher-power photomicrographs of portions of the dentate gyrus immunohistochemically stained for doublecortin in a range of mammalian species. Upper two rows show artiodactyls, third row shows Afrotherians, fourth row shows rodents, and the bottom row shows Microchiropterans and Megachiropterans. Note the presence of immature neurons in all these species. Scale bar in the bottom right image 100 μm and applies to all

Fig. 7

Low-power photomicrographs of the hippocampus in certain key species investigated in the current study. a African lion (Panthera leo), Nissl stain; b African lion, immunohistochemical staining for doublecortin; c Northern fur seal (Callorhinus ursinus), Nissl stain; d river hippopotamus (Hippopotamus amphibius), Nissl stain; e West Indian manatee (Trichechus manatus), Nissl stain; f harp seal (Pagophilus groenlandicus), Nissl stain. Scale bar in each low-power image 1 mm. Insets in b–e are higher-power photomicrographs of immunohistochemical staining for doublecortin in each species. Scale bar in inset e 50 μm, and applies to all insets. CA cornu ammonis, DG dentate gyrus

Fig. 8

Low-power photomicrographs of the hippocampus in the harbour porpoise (Phocoena phocoena, a, b) and minke whale (Balaenoptera acutorostrata, c, d) stained for Nissl substance (a, c) or immunohistochemical staining for doublecortin (b, d). Note the loose organization of the dentate gyrus in both cetacean species (a, c) as well as the total lack of immunohistochemical staining for doublecortin in both species (b, d). Scale bar in b 1 mm and applies to a and b, scale bar in d 1 mm and applies to c and d. Insets in b and d are higher-power photomicrographs of immunohistochemical staining for doublecortin in the remnant of piriform cortex in the harbour porpoise (b) and the piriform cortex of the minke whale (d). The staining of neurons in the piriform cortex of both cetacean species acts as an internal control for the methods used and confirms the lack of adult neurogenesis in the cetacean dentate gyrus. Scale bar in inset d 50 μm, and applies to both insets. CA cornu ammonis, DG dentate gyrus, PIR piriform cortex