Brain Research 1045 (2005) 164 – 174 www.elsevier.com/locate/brainres Research report Is humanlike cytoarchitectural asymmetry present in another species with complex social vocalization? A stereologic analysis of mustached auditory cortex

Chet C. Sherwooda,b,*, Mary Ann Raghantia,b, Jeffrey J. Wenstrupc

aDepartment of Anthropology, Kent State University, 226 Lowry Hall, Box 5190, Kent, OH 44242-0001, USA bSchool of Biomedical Sciences, Kent State University, Kent, OH 44242-0001, USA cDepartment of Neurobiology, Northeastern Ohio Universities College of Medicine, Rootstown, OH 44272, USA

Accepted 15 March 2005 Available online 19 April 2005

Abstract

Considerable evidence suggests that left hemispheric lateralization for language comprehension in humans is associated with cortical microstructural asymmetries. However, despite the fact that left hemispheric dominance for the analysis of species-specific social vocalizations has been reported in several other species, little is known concerning microstructural asymmetries in auditory cortex of nonhumans. To test whether such neuroanatomical lateralization characterizes another species with complex social vocalizations, we performed stereologic analyses of Nissl-stained cells in layer III of area DSCF in mustached (Pteronotus parnellii). Area DSCF was selected because it contains neurons which are sensitive to several temporal features of conspecific vocalizations. Primary visual cortex (V1) was also studied as a comparative reference. We measured neuron densities, glial densities, and neuronal volumes in both hemispheres of 10 adult male bats. Results indicate that these variables are not significantly lateralized in area DSCF or V1. Additionally, magnopyramidal cells (i.e., the largest 10% of neurons from both hemispheres) were not asymmetric in their frequency of distribution at the population level. Although several individual bats had asymmetric neuron distributions, consistent hemispheric bias was not evident. Absence of population- level microstructural asymmetry in area DSCF of mustached bats suggests alternative evolutionary scenarios including: (1) microstructural lateralization of auditory cortical circuitry may be a unique adaptation for human language, and (2) the specialized biosonar function of mustached bat auditory cortex may require symmetrical cytoarchitectural structure. Resolution of these alternatives will require further data on the microstructure of auditory cortex in species with lateralized perception of acoustic social communication. D 2005 Elsevier B.V. All rights reserved.

Theme: Neural basis of behavior Topic: Neuroethology

Keywords: Asymmetry; Mustached bat; Social vocalization; Stereology; DSCF; V1

1. Introduction majority of individuals, regions of the left cerebral hemi- sphere are specialized for the comprehension and production Neuroanatomical and behavioral lateralization are wide- of language [17,57]. Left hemispheric dominance for the spread across vertebrate species [47]. The link between perception and analysis of species-specific social voca- behavioral lateralization and neocortical asymmetry is lizations, however, is not unique to humans and has perhaps best established in humans where, in the vast been reported in several nonhuman species, especially among Old World monkeys. Evidence from behavior, experimental lesion studies, and functional imaging, for * Corresponding author. Fax: +1 330 672 2999. instance, indicates that macaques display a left hemisphere E-mail address: [email protected] (C.C. Sherwood). bias for discriminating acoustic features of species-specific

0006-8993/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.03.023 C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174 165 vocal calls [20,21,41,42,44]. In addition, house mouse To address the lack of data concerning microstructural mothers exhibit a right ear preference in their orientation asymmetries in nonhumans, the current study examines response to the ultrasonic distress calls of their pups [8]. whether cortical asymmetries occur in an echolocating bat. Finally, chick and starling individuals express unilateral The greater mustached bat, Pteronotus parnellii,isan dominance in their perception of conspecific vocal calls, insectivorous species of the neotropics that displays two however, they do not exhibit strong population-wide direc- highly developed audio–vocal behaviors. In echolocation, it tional bias to the left [15,35]. Taken together, these findings emits a multiharmonic sonar signal and analyzes the suggest that lateralized auditory processing of communica- resulting echoes to navigate and to catch flying insects. In tion signals is common across vertebrates, possibly due to social communication, it uses a complex repertoire of the constraint of reducing temporal delays associated with vocalizations, with well-defined rules for combining simple interhemispheric transfer in the analysis of complex, time- syllables into composites that can last a second or more [32]. dependent, serial streams of acoustic information as Mustached bats are highly social, often forming groups of represented by many species-specific vocal calls [46]. tens of thousands in which adults of both sexes live together In this context, numerous studies have documented for a part of the year but then segregate when females give microstructural asymmetries of auditory cortical areas in birth and nurse their pups [3,18,52]. Since these bats live in humans that putatively relate to perceptual lateralization. A large dark caves, they depend on acoustic communication principal focus of these investigations has been area Tpt (the for many of their social interactions. posterior portion of Brodmann’s area 22), a eulaminate cortex The auditory cortex of this displays highly at the core of Wernicke’s area, a polymodal association specialized response properties both to the echoes of sonar region important for language comprehension in humans [1]. signals [40,55] and to social vocalizations [9,39].For In an early study, left hemispheric dominance in the size of example, auditory cortical neurons show sensitivity to area Tpt was demonstrated in four human brain specimens several temporal features of social vocalizations, such as [11]. More recently, long-range intrinsic connections within temporal ordering and timing of syllables, analogous to area Tpt labeled in postmortem brains with lipophilic dyes temporal features that underlie syntax in human speech and revealed greater spacing between interconnected patches in other primate vocal communication systems [9]. Preliminary the left hemisphere compared to the right [12]. The layer III studies of mustached bat auditory cortex suggest that pyramidal cell population also exhibits left hemisphere functional asymmetries occur. In the Doppler-shifted con- dominant asymmetry within several different cortical areas stant frequency area (area DSCF) of primary auditory along the temporal auditory processing stream. For instance, cortex, neurons on the left side appear to respond better to in many auditory areas, including primary auditory cortex communication sounds than to sonar pulse–echo combina- and area Tpt, the left hemisphere has a greater number of the tions, while this was not the case for neurons on the right largest pyramidal cells in layer III, known as magnopyrami- side [30,31]. dal cells, that give rise to long corticocortical association We examined whether cytoarchitectural asymmetries projections [27]. Furthermore, acetylcholinesterase-rich pyr- similar to those reported for area Tpt in humans are present amidal cells display greater cell soma volumes in the left in the auditory cortex of mustached bats. We hypothesized hemisphere of secondary and language-associated areas that asymmetries would be present in area DSCF favoring despite lacking asymmetry in terms of number [28]. Lastly, the left hemisphere. For comparison, we also analyzed left area Tpt has a greater amount of neuropil and contains asymmetries in primary visual cortex (V1), under the axons with thicker myelin sheaths, possibly to facilitate hypothesis that lateralization would not be present, espe- specialized processing of signals with rapid temporal cially considering the extreme visual reduction of this variation, such as speech [2]. species. To test these predictions, we used design-based Unfortunately, there is a paucity of data on the presence stereologic techniques to characterize cellular densities and of analogous microstructural asymmetries in the auditory neuronal volume distributions in layer III and made cortex of nonhuman species. In the only study to directly comparisons between homotopic cortical areas. compare cytoarchitectural asymmetries in layer III of area Tpt in humans and other species, humans were found to have wider minicolumns and greater neuropil space on the 2. Materials and methods left, whereas such asymmetries were absent in chimpanzees and rhesus macaques [5]. Other data, however, indicate the 2.1. Subjects and sample preparation presence of asymmetries in volume [14] and the distribution of inhibitory microcircuitry in area Tpt of macaques [33]. The brains of ten adult male greater mustached bats Overall, these results support the idea that cortical asym- were used in this study. Greater mustached bats (P. parnellii metries comprise an important anatomical substrate for parnellii) were captured in Jamaica, West Indies. In lateralized auditory processing of language in humans and captivity, male and female bats were housed communally may, to some extent, serve a homologous function in other in a large room (8.5V Â 15V Â 8.5V) with incomplete species. partitions to allow the formation of social groups. The 166 C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174

Fig. 1. Coronal Nissl-stained sections showing areas DSCF and V1 (A). Fiducial marks denote the right hemisphere. The boundaries of regions of interest used in stereologic sampling are shown with arrowheads. Scale bar = 1 mm. Typical cytoarchitecture of areas DSCF and V1 (B). Note that all layer boundaries are clearly evident, with the exception of the uppermost boundary of layer IV, which is indistinct from layer III. Thus, we performed stereologic analyses from the upper part of the region shown here as layers III/IV. Scale bar = 100 Am. room was heated (78 -F) and humidified (>75% relative the absence of these data, we defined V1 as the three humidity) to suit neotropical bats. All procedures on the caudalmost sections in our series that contain neocortex were approved by the Institutional Animal Care interposed between perirhinal and retrosplenial cortex. and Use Committee of Northeastern Ohio Universities These criteria conservatively estimate the location of V1 College of Medicine. After the bat was deeply anesthetized in light of strong evidence that its position is relatively with Nembutal (75 mg/kg, i.p.) and reflexes were invariant across diverse mammalian species [29,48,49].We eliminated, the chest was opened, and the bat was perfused restricted our stereologic sampling to the dorsal curvature of through the heart with phosphate buffer and 10% formalin. the occipital cortex (Fig. 1) because the extent of V1 in the The head was removed, blocked, and refrigerated over- mustached bat is likely to be small since the thalamic targets night in 30% sucrose. The brain was sectioned with a of retinal input are highly restricted [6]. freezing microtome in the coronal plane at 30 Am. During sectioning, a fiducial mark was consistently placed in the 2.3. Identification of cell types right hemisphere. The sections were collected in chilled 0.1 M phosphate buffer and rinsed in 0.05 M phosphate We quantified cellular densities and volumes of diffe- buffer before mounting on slides. A series of equidistant rent cortical cell types (neuron-, oligodendrocyte-, and sections spaced 120 Am apart was stained with cresyl astrocyte-like cells) on the basis of Nissl-staining charac- violet to reveal cytoarchitecture. teristics (Fig. 2). Neuron-like cells (henceforth called ‘‘neurons’’) were identified by the presence of dark course 2.2. Identification of cortical regions of interest clumps of stained Nissl substance in the perikaryon, a

The location of area DSCF was identified anatomically by reference to published cortical maps [10] and the results of previous tract-tracing experiments in a subset of the study animals [58,59]. Stereologic sampling of area DSCF was performed in coronal sections at the level distinguished by the first rostral appearance of the Sylvian fossa, where the hippocampal formation is bordered subjacent by the caudate nucleus laterally and the dorsal lateral geniculate nucleus medially. The specific region of the cortical mantle analyzed was located immediately lateral to the position of the hippocampal formation (Fig. 1). Fig. 2. Cytomorphologic appearance of different cell types identified from Correlative physiological and cytoarchitectural mapping Nissl-stained materials. Astrocytes are indicated by arrowheads and studies of visual cortex in the mustached bat do not exist. In oligodendrocytes are indicated by arrows. Scale bar = 25 Am. C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174 167 large lightly stained nucleus, a distinct nucleolus, and the frequent appearance of lightly stained proximal segments of dendritic processes. In contrast, glial cells lack a conspicuous nucleolus and contain significantly less endoplasmic reticulum in the cytoplasm, so only the nucleus is stained. Oligodendrocyte- and astrocyte-like cells (henceforth called ‘‘oligodendrocytes’’ and ‘‘astro- cytes’’) were distinguished from each other based on cytomorphologic characteristics. Oligodendrocytes appear as round hyperchromatic nuclei usually surrounded by a pale halo and tend to be aggregated around neuronal cell Fig. 3. The stereologic sampling approach for area DSCF is shown. In each section, the borders of the region of interest are drawn at low magnification bodies. In contrast, astrocytes typically appear as a larger (A). Optical disector frames are distributed to cover 9% of the section and paler nucleus with variable irregular shapes. It should surface area. At high magnification, different cell types are counted and be noted that, because immature oligodendrocytes are more volumes of neurons are measured using the nucleator probe (B). In this lightly stained in Nissl preparations, they may be mistaken image, neurons are marked by yellow squares and astrocytes are marked by for astrocytes. However, in general, both oligodendrocytes orange diamonds. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) and astrocytes can readily be distinguished from microglia which are relatively sparse in distribution and have darkly stained, small, elongated nuclei. To account for possible each hemisphere, and 88.8 T 9.3 disectors were placed in inaccuracies in our identification of glial cell subtypes, we DSCF, resulting in samples of 581.3 T 180.3 cells in V1 and performed statistical analyses on the population of pooled 637.3 T 115.6 in DSCF. oligodendrocytes and astrocytes (named ‘‘glial cells’’) as Cellular density (Nv) was calculated as the sum of cells well as each subtype separately. counted with the optical disectors (~QÀ) divided by the product of the sum of the disectors and the volume of the 2.4. Stereology disector [26]. Tissue shrinkage in the z dimension was corrected for density measurements by multiplying the height Quantification of cellular densities and volumes within of the disector by the ratio of the sectioned thickness to the layer III was performed using a computerized stereology actual number weighted mean section thickness after mount- system consisting of a Zeiss Axioplan 2 photomicroscope ing and dehydration [43]. The actual mean mounted section equipped with a Ludl XY motorized stage, Heidenhain z axis thickness was optically measured at every 8th disector encoder, an Optronics MicroFire color videocamera, a Dell sampling location to obtain a number weighted correction PC workstation, and StereoInvestigator software (Micro- factor for each cortical region. Since shrinkage in the surface BrightField Inc., Wiliston, VT). Beginning at a random area of mounted sections is minimal, no corrections were starting point, three sections equidistantly spaced 120 Am made for the x and y dimensions. apart were selected for analysis from each cortical region. The cellular volumes of neurons were calculated using Densities of neurons, astrocytes, and oligodendrocytes were the nucleator probe [19]. In each cortical region, neuronal obtained using the optical disector combined with a volumes were measured simultaneously during optical fractionator sampling scheme [36,53,60]. After outlining disector sampling from 2/3 of the sections. Nucleoli of the boundaries of layer III within the cortical region of neurons included in optical disectors were marked and interest at low magnification (4 Zeiss Achroplan, N.A. two isotropic lines from randomly selected directions 0.10), a set of optical disector frames measuring 30 Am  30 were centered at the nucleolus and superimposed over the Am were placed in a systematic random fashion to cover 9% neuron. The intersection of each line with the outer of the sampled area. Because there is not a sharp border surface of the neuronal soma was marked and cellular between layers III and IV, as has also been observed in volume was measured based on the nucleator principle Rhinolophus rouxi [45], we drew a conservative boundary [19]. This sampling scheme resulted in the measurement that did not extend to the deepest portion of layer III (Fig. of cellular volumes in 282 T 63 neurons (range = 177– 3). Disector analysis was performed under Koehler illumi- 432) for each cortical area from each hemisphere in each nation using a 63 objective (Zeiss Plan-Apochromat, N.A. bat, yielding a total of 11,292 neuron volumes measured 1.4). Each cell type was counted with a different marker for this study. when its nucleolus was encountered within the optical disector frame according to standard stereologic principles 2.5. Data analysis [26,36]. The thickness of optical disectors was consistently set to 6 Am to allow for a minimum 2 Am guard zone on Lateral asymmetries in cellular densities and volumes either side of the section after z axis collapse from were quantified by calculating an asymmetry coefficient histological processing. On average, 76.0 T 22.8 (mean T (AC) using the formula: (R À L) / [(R + L)  0.5]. The standard deviation) disector frames were placed in V1 for sign of the AC indicates the direction of asymmetry such 168 C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174

Table 1 Results of stereologic estimates of cellular densities (cells per mm3) Individual Neuron density Astrocyte density Oligodendrocyte density Glial density DSCF V1 DSCF V1 DSCF V1 DSCF V1 Left Right Left Right Left Right Left Right Left Right Left Right Left Right Left Right 1 209,935 188,426 292,459 292,614 34,459 44,410 88,356 97,008 35,519 45,045 37,110 35,782 69,978 89,455 125,466 132,789 2 243,557 242,024 258,436 275,415 69,040 60,506 68,724 71,988 57,054 56,381 55,556 52,980 126,094 116,887 124,280 124,968 3 151,771 168,981 144,823 160,957 47,101 44,444 36,642 37,809 41,432 42,593 28,354 21,065 88,533 87,037 64,996 58,873 4 202,105 211,856 212,638 230,137 61,287 56,864 60,835 44,444 32,749 48,309 30,702 38,965 94,035 105,173 91,537 83,409 5 90,294 97,511 183,073 133,206 24,436 28,220 45,582 34,568 15,496 21,793 25,780 20,860 39,932 50,013 71,361 55,428 6 209,839 245,977 216,650 227,709 64,191 59,195 53,412 58,711 42,499 52,299 49,211 54,870 106,690 111,494 102,623 113,580 7 218,642 253,411 222,481 232,644 48,889 53,281 42,377 52,397 36,667 42,885 31,783 39,822 85,556 96,166 74,160 92,219 8 300,545 263,695 318,563 319,616 94,365 58,226 85,450 71,331 46,536 35,271 53,219 42,524 140,901 93,497 138,668 113,855 9 332,647 364,364 348,212 346,516 87,791 79,705 80,792 76,271 74,074 74,391 47,667 43,315 161,866 154,096 128,459 119,586 10 309,467 334,134 290,123 240,404 87,829 62,612 62,678 55,387 53,731 54,184 37,987 45,370 141,559 116,796 100,665 100,758

Mean 226,880 237,038 248,746 245,922 61,939 54,746 62,485 59,991 43,576 47,315 39,737 39,555 105,514 102,061 102,222 99,547 SEM 23,397 24,454 20,172 20,900 7452 4312 5787 6118 4998 4387 3438 3617 11,801 8462 8357 8479 that positive values specify right hemisphere dominance, were calculated to reflect a possible correlate of metabolic negative values specify left hemisphere dominance, and and supportive requirements for greater neuronal activity. zero denotes perfect symmetry. The absolute value of the Increased glial–neuron ratios, for instance, are associated AC indicates the magnitude of asymmetry. One sample t with environmental enrichment in rats [7]. Asymmetry tests were performed on AC values to examine popula- coefficients were calculated from the data for each tion-level deviations from perfect symmetry. Additional individual, and one sample t tests were performed to analyses of neuron volume distributions used Kolmo- evaluate whether significant deviations from symmetry gorov–Smirnov and binomial tests. All statistical analyses were present in this population of bats. The results of this were performed using Statistica software version 6.0 analysis, shown in Table 3, indicate that none of these (StatSoft, Inc., Tulsa, OK). Statistical significance was set variables are significantly lateralized at the population level. at a = 0.05 (two-tailed). Next, we examined whether the same morphometric variables express a greater degree of laterality, irregard- less of direction, in area DSCF as compared with V1. To 3. Results test this hypothesis, we took the absolute value of each of the variables in Tables 1 and 2 and performed 3.1. Cell densities and glial–neuron ratios dependent samples t tests with cortical area as the grouping factor. None of the variables showed a Table 1 shows cell densities and Table 2 shows glial– significant difference in the degree of laterality between neuron ratios from both hemispheres. Glial–neuron ratios area DSCF and V1.

Table 2 Glial–neuron ratios Individual Astrocyte–neuron ratio Oligodendrocyte–neuron ratio Glial–neuron ratio DSCF V1 DSCF V1 DSCF V1 Left Right Left Right Left Right Left Right Left Right Left Right 1 0.16 0.24 0.30 0.33 0.17 0.24 0.13 0.12 0.33 0.47 0.43 0.45 2 0.28 0.25 0.27 0.26 0.23 0.23 0.21 0.19 0.52 0.48 0.48 0.45 3 0.31 0.26 0.25 0.23 0.27 0.25 0.20 0.13 0.58 0.52 0.45 0.37 4 0.30 0.27 0.29 0.19 0.16 0.23 0.14 0.17 0.47 0.50 0.43 0.36 5 0.27 0.29 0.25 0.26 0.17 0.22 0.14 0.16 0.44 0.51 0.39 0.42 6 0.31 0.24 0.25 0.26 0.20 0.21 0.23 0.24 0.51 0.45 0.47 0.50 7 0.22 0.21 0.19 0.23 0.17 0.17 0.14 0.17 0.39 0.38 0.33 0.40 8 0.31 0.22 0.27 0.22 0.15 0.13 0.17 0.13 0.47 0.35 0.44 0.36 9 0.26 0.22 0.23 0.22 0.22 0.20 0.14 0.13 0.49 0.42 0.37 0.35 10 0.28 0.19 0.22 0.23 0.17 0.16 0.13 0.19 0.46 0.35 0.35 0.42

Mean 0.27 0.24 0.25 0.24 0.19 0.21 0.16 0.16 0.47 0.44 0.41 0.41 SEM 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.02 C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174 169

Table 3 Asymmetry coefficient (AC) statistics and results of one sample t tests (N = 10) Variable Region Mean AC Standard deviation Standard error tP Neuron density DSCF 0.046 0.099 0.031 1.469 0.176 V1 À0.016 0.132 0.042 À0.385 0.709 Astrocyte density DSCF À0.077 0.214 0.068 À1.137 0.285 V1 À0.047 0.172 0.054 À0.864 0.410 Oligodendrocyte density DSCF 0.107 0.196 0.062 1.733 0.117 V1 À0.016 0.195 0.062 À0.262 0.799 Glial density DSCF 0.000 0.196 0.062 0.005 0.996 V1 À0.033 0.139 0.044 À0.747 0.474 Astrocyte–neuron ratio DSCF À0.123 0.217 0.069 À1.798 0.106 V1 À0.031 0.159 0.050 À0.606 0.559 Oligodendrocyte–neuron ratio DSCF 0.061 0.184 0.058 1.055 0.319 V1 0.000 0.220 0.069 0.005 0.996 Glial–neuron ratio DSCF À0.046 0.191 0.061 À0.762 0.465 V1 À0.017 0.146 0.046 À0.359 0.728 Mean neuronal volume DSCF À0.029 0.157 0.050 À0.587 0.572 V1 0.007 0.141 0.045 0.168 0.871

3.2. Neuronal volume distributions frequently than expected in the left hemisphere in two cases. In V1, magnopyramidal cells were symmetrically distributed The cellular volumes of layer III neurons encountered in eight cases, more frequent than expected in the right during fractionator sampling were measured using the hemisphere in one case, and more frequently than expected in nucleator probe, yielding a systematic random sample of the left hemisphere in one case. Finally, data were pooled the layer III neuronal population. Asymmetry coefficients across cases with the largest 10% of cells in each case were calculated from mean neuronal volumes in each separately coded. Binomial analysis of the pooled data found region. One sample t tests (Table 3) indicate that mean symmetric distributions of magnopyramidal cells in both area neuronal volumes are not significantly asymmetric at the DSCF and V1 (Fig. 6). population level in either area DSCF or V1. Next, we examined asymmetry in neuron volume distributions using Kolmogorov–Smirnov tests for each cortical area in each 4. Discussion individual. Overall, this analysis found a largely symmetric pattern of cell volume distributions in our sample. For both Our findings demonstrate that interhemispheric asymme- area DSCF (Fig. 4) and V1 (Fig. 5), four individuals try of layer III cytoarchitectural organization is not present showed no significant differences between hemispheres and in area DSCF or V1 of male mustached bats. In particular, six individuals displayed asymmetries in cell volume we did not find population-level asymmetries of neuron distributions. For both cortical areas, among the individuals density, glial density, glial–neuron ratio, or mean neuron with significant differences between hemispheres, three had volume. Although several individuals had neuron volume a larger mean neuron volume in the left hemisphere and distributions that deviated from symmetry in areas DSCF three had a larger mean neuron volume in the right and V1, there was interindividual variation in the direction hemisphere. of hemispheric bias. Importantly, magnopyramidal cells Our next analysis focused on a subset of neurons with the were not consistently encountered more frequently in area largest cell volumes (i.e., magnopyramidal cells). To evaluate DSCF of the left hemisphere. Taken together, these results laterality in the distribution of these magnopyramidal cells in indicate that area DSCF does not exhibit consistent each individual, we followed the method of Hutsler [27]. population-wide microstructural asymmetries in mustached First, neurons from both hemispheres were pooled and ranked bats and, moreover, does not express a degree of lateraliza- according to cellular volume. The largest 10% of these tion beyond what is observed for area V1. These findings neurons were then coded and regrouped with the other stand in contrast with reports of hemispheric differences in neurons sampled in their respective hemisphere of origin. the cytoarchitecture of area Tpt in humans, including lower Within each region and each hemisphere, binomial analyses cell density [5,51] and a greater number of magnopyramidal were used to test whether the proportion (i.e., 10%) of cells [27] on the left. magnopyramidal cells differed from the number expected for In the context of the known phylogenetic distribution of the total number of neurons sampled. Using this approach, left hemisphere dominance for the perception of species- each individual was first analyzed separately (Fig. 6). Results specific vocalizations, it is interesting that we did not indicate that magnopyramidal cells in area DSCF were observe cytoarchitectural asymmetry within area DSCF of distributed symmetrically in five cases, more frequent than mustached bats. Evidence of functional lateralization is most expected in the right hemisphere in three cases and more abundant for catarrhine primates. For example, macaques 170 C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174

Fig. 4. Histograms of cellular volume distributions of layer III in area DSCF in both hemispheres for each individual. The values presented in each graph are mean T standard deviation and the P statistic of the Kolmogorov–Smirnov test. ‘‘Left’’ or ‘‘right’’ is underlined to indicate the hemisphere with the largest mean neuronal volumes only in cases where the Kolmogorov–Smirnov test revealed significant differences in cell volume distributions. C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174 171

Fig. 5. Histograms of cellular volume distributions of layer III in area V1 in both hemispheres for each individual. The values presented in each graph are mean T standard deviation and the P statistic of the Kolmogorov–Smirnov test. ‘‘Left’’ or ‘‘right’’ is underlined to indicate the hemisphere with the largest mean neuronal volumes only in cases where the Kolmogorov–Smirnov test revealed significant differences in cell volume distributions. 172 C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174

noting, however, that the experimental design of some of these behavioral studies has been questioned [25]. Nevertheless, current evidence suggests that hemispheric dominance for vocal perception may extend to other vertebrate taxa besides primates. Elecrophysiological map- ping of the house mouse brain revealed a greater extent of auditory cortex surface area in the left hemisphere compared to the right [54] which may be associated with lateralized responses to conspecific communication [8]. In non-mam- malian vertebrates, the presence of unilateral specialized processing of communication signals has been reported in avians, however, there does not appear to be the population- wide bias towards the left that has been reported in . Data from multiunit recordings in starlings, for instance, show lateralization in the strength of neuronal activation in response to the presentation of species-specific songs, but not other artificial sounds, with interindividual variation in the dominant hemisphere [15,16]. The comparative data concerning lateralized processing of social vocalizations, reviewed above, pertain to a small number of species of distant phylogenetic affinities, con- sequently limiting our ability to provide a robust reconstruc- tion for the evolutionary emergence of this neural specialization. Nonetheless, considering the evidence that left hemisphere dominance for the analysis of conspecific acoustic communication is present in rodents and primates, it is most parsimonious to infer that this trait was present in the last common ancestor of these lineages, at the origin of the Fig. 6. Bar graphs illustrating the results of binomial tests for the lateralized supraordinal group Euarchontoglires approximately 85–95 distribution of magnopyramidal cells. In each cortical area, neuron volumes million years ago [37]. Furthermore, if preliminary elecro- derived from both hemispheres were pooled and rank ordered to identify the physiological studies showing more selective responses to largest 10% across both hemispheres. These largest neurons, called social vocalization in the left hemisphere in mustached bats magnopyramidal cells, were then regrouped with other neurons sampled [30,31] are borne out, then left dominant lateralization for the from their hemisphere of origin and a binomial test was performed to determine whether the number of magnopyramidal cells differed signifi- perception of social vocalization may date back as far as the cantly from the expected frequency of 10%. A significant difference split between Euarchontoglires and and hence (denoted by an asterisk with P statistics shown above each bar) indicates be a conservative trait of Boreoeutheria. cases where magnopyramidal cells were either more or less frequent than Although there is apparent incongruence between evi- expected. dence for functional lateralization in area DSCF and our findings of a lack of microstructural asymmetry, interpreta- have been reported to direct the right ear towards the source tion of these results is limited by the absence of comparable of species-specific vocal calls [20] and to display a right ear data on microstructural asymmetries in auditory areas of advantage in discriminating acoustic features of communi- other nonhuman species. In particular, similar rigorous cation-related sounds [41,42]. Experimental ablation of the stereologic methods have not yet been applied to the left superior temporal gyrus of Japanese macaques results in investigation of asymmetries in communication-related initial impairment of the ability to discriminate coo vocali- cortical areas of any species, including humans. Furthermore, zations, whereas right hemisphere lesion has no effect [21]. to our knowledge there are no other comparative data on In addition, recent data from positron emission tomography asymmetries in glial–neuron ratios of auditory cortex. Thus, indicate that the left temporal pole of rhesus macaques is in the absence of these crucial comparative data, evolutionary activated during the perception of social vocalizations [44]. scenarios to explain our findings depend on alternative Asymmetric activity is abolished in forebrain commissuro- assumptions regarding the association between functional tomized monkeys, indicating that social vocalizations are perceptual lateralization and neuroanatomical asymmetry. initially analyzed bilaterally in the temporal cortex and On the one hand, if we assume that functional and subsequent specialized processing is directed towards the microstructural asymmetries are correlated, then left domi- left hemisphere. As a whole, these data suggest that the nant microstructural asymmetry of auditory cortex may be a auditory cortex of the left hemisphere is involved in the symplesiomorphy among boreoeutherian mammals. If this is analysis of social vocalizations in primates. It is worth true, then our observation of a lack of microstructural C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174 173 asymmetry in area DSCF of mustached bats would represent Foundation for Anthropological Research, Kent State Uni- a derived condition. In this case, we speculate that the versity, and research grant R01 DC 00937 from the National absence of detectable cytoarchitectural asymmetries in DSCF Institute on Deafness and Other Communication Disorders. of this species may relate to this cortical area’s dual function in the analysis of social vocalization and biosonar. The important biosonar function of the DSCF neurons may References require bilateral symmetry which masks the more evolu- tionary conservative condition of asymmetry for the process- [1] F. Aboitiz, R. Garcia, The evolutionary origin of the language areas in the human brain. A neuroanatomical perspective, Brain Res. Rev. 25 ing of social communication. It is possible that, for the spatial (1997) 381–396. orientation function of the DSCF area, structural symmetry is [2] B. Anderson, B.D. Southern, R.E. Powers, Anatomic asymmetries of necessary to permit the equal analysis and representation of the posterior superior temporal lobes: a postmortem study, Neuro- both left and right sound fields. psychiatry Neuropsychol. Behav. Neurol. 12 (1999) 247–254. Alternatively, perceptual lateralization in other nonhu- [3] G.C. Bateman, T.A. Vaughan, Nightly activities of mormoopid bats, J. . 55 (1974) 45–65. man taxa may not correlate with microstructural neuro- [4] P. Belin, M. Zilbovicius, S. Crozier, L. Thivard, A. Fontaine, M.C. anatomic asymmetry and such specializations of circuitry Masure, Y. Samson, Lateralization of speech and auditory temporal in the dominant hemisphere may represent an evolutionary processing, J. Cogn. Neurosci. 10 (1998) 536–540. novelty in humans uniquely related to language. Thus, [5] D.P. Buxhoeveden, A.E. Switala, M. Litaker, E. Roy, M.F. despite the presence of gross anatomic asymmetries of Casanova, Lateralization of minicolumns in human planum temporale is absent in nonhuman primate cortex, Brain Behav. temporal lobe structures in nonhuman primates such as Evol. 57 (2001) 349–358. leftward dominance of the planum temporale in great apes [6] E. Covey, W.C. Hall, J.B. Kobler, Subcortical connections of the [13,23] and a longer left Sylvian fissure in many superior colliculus in the mustache bat, Pteronotus parnellii, J. Comp. anthropoid primate species (e.g., [22,24,34,61]), interhemi- Neurol. 263 (1987) 179–197. spheric differences in microanatomical wiring may not be [7] M.C. Diamond, D. Krech, M.R. Rosenzweig, The effects of an enriched environment on the histology of the rat cerebral cortex, present. In support of this interpretation, it is noteworthy J. Comp. Neurol. 123 (1964) 111–120. that Buxhoeveden and colleagues [5] failed to find [8] G. Ehret, Left hemisphere advantage in the mouse brain for recognizing humanlike asymmetry of minicolumn architecture within ultrasonic communication calls, Nature 325 (1987) 249–251. layer III of area Tpt in macaques and chimpanzees. In any [9] K.H. Esser, C.J. Condon, N. Suga, J.S. Kanwal, Syntax processing by event, it is impossible at the present time to distinguish auditory cortical neurons in the FM–FM area of the mustached bat Pteronotus parnellii, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) among these alternative evolutionary scenarios. Future 14019–14024. studies directed at obtaining rigorous microstructural data [10] D.C. Fitzpatrick, J.F. Olsen, N. Suga, Connections among functional within species for which perceptual asymmetries have been areas in the mustached bat auditory cortex, J. Comp. Neurol. 391 documented should help to resolve these issues. (1998) 366–396. Whatever the evolutionary history of cortical lateralization [11] A.M. Galaburda, F. Sanides, N. Geschwind, Human brain. Cytoarch- itectonic left–right asymmetries in the temporal speech region, Arch. may be, our data have interesting implications for under- Neurol. 35 (1978) 812–817. standing the neuroanatomical conditions for processing [12] R.A. Galuske, W. Schlote, H. Bratzke, W. Singer, Interhemispheric temporally complex acoustic information such as syntactic asymmetries of the modular structure in human temporal cortex, social communication. Functional neuroanatomical asym- Science 289 (2000) 1946–1949. metry is frequently claimed to be a requisite feature under- [13] P.J. Gannon, R.L. Holloway, D.C. Broadfield, A.R. Braun, Asymme- try of chimpanzee planum temporale: humanlike pattern of Wernicke’s lying such complex temporal analysis due to the conduction brain language area homolog, Science 279 (1998) 220–222. delay of interhemispheric transfer and related unihemispheric [14] P.J. Gannon, P.R. Hof, N.M. Kheck, Brain language area evolution: specialization for processing signals with rapid acoustic human-like pattern of cytoarchitectonic, but not gross anatomic, L > R transitions [4,50,56]. However, our data suggest that this may hemispheric asymmetry of planum temporale homolog in Macaca not always be the case. Sequences of echolocation pulse– fascicularis, Am. J. Phys. Anthropol. 30 (2000) 155, (Suppl.). [15] I. George, H. Cousillas, J.P. Richard, M. Hausberger, Song perception echo combinations and social vocalizations are temporally in the European starling: hemispheric specialisation and individual complex acoustic signals, composed of multiple individual variations, C. R., Biol. 325 (2002) 197–204. elements and lasting over 1 s [32,38]. However, in the [16] I. George, B. Vernier, J.P. Richard, M. Hausberger, H. Cousillas, mustached bat, analysis of both of these types of sequences Hemispheric specialization in the primary auditory area of awake and does not appear to require structural asymmetries comparable anesthetized starlings (Sturnus vulgaris), Behav. Neurosci. 118 (2004) 597–610. to those that occur in areas of human cerebral cortex [17] N. Geschwind, W. Levitsky, Human brain: left–right asymmetries in associated with speech perception. temporal speech region, Science 161 (1968) 186–187. [18] R.E. Goodwin, The ecology of Jamaican bats, J. Mammal. 51 (1970) 571–579. Acknowledgments [19] H.J. Gundersen, The nucleator, J. Microsc. 151 (1988) 3–21. [20] M.D. Hauser, K. Andersson, Left hemisphere dominance for processing vocalizations in adult, but not infant, rhesus monkeys: We thank Carol Grose for providing expert technical field experiments, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) assistance. This work was supported by the Wenner-Gren 3946–3948. 174 C.C. Sherwood et al. / Brain Research 1045 (2005) 164–174

[21] H.E. Heffner, R.S. Heffner, Temporal lobe lesions and perception [41] M.R. Petersen, M.D. Beecher, S.R. Zoloth, D.B. Moody, W.C. of species-specific vocalizations by macaques, Science 226 (1984) Stebbins, Neural lateralization of species-specific vocalizations 75–76. by Japanese macaques (Macaca fuscata), Science 202 (1978) [22] P.L. Heilbroner, R.L. Holloway, Anatomical brain asymmetries in 324–327. New World and Old World monkeys: stages of temporal lobe [42] M.R. Petersen, M.D. Beecher, S.R. Zoloth, S. Green, P.R. Marler, D.B. development in primate evolution, Am. J. Phys. Anthropol. 76 Moody, W.C. Stebbins, Neural lateralization of vocalizations by (1988) 39–48. Japanese macaques: communicative significance is more important [23] W.D. Hopkins, L. Marino, J.K. Rilling, L.A. MacGregor, Planum than acoustic structure, Behav. Neurosci. 98 (1984) 779–790. temporale asymmetries in great apes as revealed by magnetic [43] D.A. Peterson, H.A. Dickinson-Anson, J.T. Leppert, K.F. Lee, F.H. resonance imaging (MRI), NeuroReport 9 (1998) 2913–2918. Gage, Central neuronal loss and behavioral impairment in mice lacking [24] W.D. Hopkins, D.L. Pilcher, L. MacGregor, Sylvian fissure asymme- neurotrophin receptor p75, J. Comp. Neurol. 404 (1999) 1–20. tries in nonhuman primates revisited: a comparative MRI study, Brain [44] A. Poremba, M. Malloy, R.C. Saunders, R.E. Carson, P. Herscovitch, Behav. Evol. 56 (2000) 293–299. M. Mishkin, Species-specific calls evoke asymmetric activity in the [25] W.D. Hopkins, D.L. Pilcher, C. Cantalupo, Neural basis of cognition, monkey’s temporal poles, Nature 427 (2004) 448–451. communication and handedness in nonhuman primates, in: D. [45] S. Radtke-Schuller, Neuroarchitecture of the auditory cortex in the Maiestripieri (Ed.), Primate Psychology, Harvard Univ. Press, Cam- rufous (Rhinolophus rouxi), Anat. Embryol. 204 (2001) bridge, 2003. 81–100. [26] C.V. Howard, M.G. Reed, Unbiased stereology—Three-dimensional [46] J.L. Ringo, R.W. Doty, S. Demeter, P.Y. Simard, Time is of measurement in microscopy, Springer-Verlag, New York, 1998. the essence: a conjecture that hemispheric specialization arises [27] J.J. Hutsler, The specialized structure of human language cortex: from interhemispheric conduction delay, Cereb. Cortex 4 (1994) pyramidal cell size asymmetries within auditory and language- 331–343. associated regions of the temporal lobes, Brain Lang. 86 (2003) [47] L.J. Rogers, R.J. Andrew, Comparative vertebrate lateralization, 226–242. Cambridge Univ. Press, Cambridge, 2002, 660 pp. [28] J.J. Hutsler, M.S. Gazzaniga, Acetylcholinesterase staining in human [48] M.P.G. Rosa, Visual maps in the adult primate cerebral cortex: some auditory and language cortices: regional variation of structural implications for brain development and evolution, Braz. J. Med. Biol. features, Cereb. Cortex 6 (1996) 260–270. Res. 35 (2002) 1485–1498. [29] J.H. Kaas, D.C. Lyon, Visual cortex organization in primates: [49] M.G. Rosa, L.A. Krubitzer, The evolution of visual cortex: where is theories of V3 and adjoining visual areas, Prog. Brain Res. 134 V2? Trends Neurosci. 22 (1999) 242–248. (2001) 285–295. [50] J. Schwartz, P. Tallal, Rate of acoustic change may underlie [30] J. Kanwal, Interhemispheric interactions suppress responses to com- hemispheric specialization for speech perception, Science 207 (1980) munication call but not tonal stimuli in the primary auditory cortex of 1380–1381. the mustached bat, Nineteenth Midwinter Meeting of the Association [51] H.L. Seldon, Structure of human auditory cortex. I. Cytoarchitectonics for Research in Otolaryngology, 1996, (abstract #624). and dendritic distributions, Brain Res. 229 (1981) 277–294. [31] J.S. Kanwal, N. Suga, Hemispheric asymmetry in the processing of [52] T.G. Silva, Los Murcielagos de Cuba, Editorial Academia, La Habana, calls in the auditory cortex of the mustached bat, Eighteenth Cuba, 1979. Midwinter Meeting of the Association for Research in Otolaryngo- [53] D.C. Sterio, The unbiased estimation of number and sizes of arbitrary logy, 1995, (abstract #413). particles using the disector, J. Microsc. 134 (1984) 127–136. [32] J.S. Kanwal, S. Matsumura, K. Ohlemiller, N. Suga, Analysis of [54] I. Stiebler, R. Neulist, I. Fichtel, G. Ehret, The auditory cortex of the acoustic elements and syntax in communication sounds emitted by house mouse: left–right differences, tonotopic organization and mustached bats, J. Acoust. Soc. Am. 96 (1994) 1229–1254. quantitative analysis of frequency representation, J. Comp. Physiol., [33] N.M. Kheck, P.R. Hof, M. Deftereos, T. Lo, P.J. Gannon, Brain A 181 (1997) 559–571. language area evolution IV: chemoarchitectonic interhemispheric [55] N. Suga, W.E. O’Neill, K. Kujirai, T. Manabe, Specificity of asymmetries of the planum temporale (PT) homolog in Old World combination-sensitive neurons for processing of complex biosonar monkeys, Abstr.-Soc. Neurosci. 25 (1999) 105. signals in auditory cortex of the mustached bat, J. Neurophysiol. 49 [34] M. LeMay, N. Geschwind, Hemispheric differences in the brains of (1983) 1573–1626. great apes, Brain Behav. Evol. 11 (1975) 48–52. [56] P. Tallal, S. Miller, R.H. Fitch, Neurobiological basis of speech: a case [35] A´ . Miklo´si, R.J. Andrew, M. Dharmaretnam, Auditory lateralization: for the preeminence of temporal processing, Ann. N. Y. Acad. Sci. 682 shifts in ear use during attachment in the domestic chick, Laterality 1 (1993) 27–47. (1996) 215–224. [57] J. Wada, T. Rasmussen, Intracarotid injection of sodium amytal for the [36] P.R. Mouton, Principles and practices of unbiased stereology: an lateralization of cerebral speech dominance. Experimental and clinical introduction for bioscientists, The Johns Hopkins University Press, observations, J. Neurosurg. 17 (1960) 266–282. Baltimore, 2002, 214 pp. [58] J.J. Wenstrup, C.D. Grose, Inputs to combination-sensitive neurons in [37] W.J. Murphy, E. Eizirik, S.J. O’Brien, O. Madsen, M. Scally, C.J. the medial geniculate body of the mustached bat: the missing Douady, E. Teeling, O.A. Ryder, M.J. Stanhope, W.W. de Jong, M.S. fundamental, J. Neurosci. 15 (1995) 4693–4711. Springer, Resolution of the early placental mammal radiation using [59] J.J. Wenstrup, D.H. Mittmann, C.D. Grose, Inputs to combination- Bayesian phylogenetics, Science 294 (2001) 2348–2351. sensitive neurons of the inferior colliculus, J. Comp. Neurol. 409 [38] A. Novick, J.R. Vaisnys, Echolocation of flying insects by the bat (1999) 509–528. Chilonycteris parnellii, Biol. Bull. 127 (1964) 478–488. [60] M.J. West, L. Slomianka, H.J.G. Gundersen, Unbiased stereological [39] K.K. Ohlemiller, J.S. Kanwal, N. Suga, Facilitative responses to estimation of the total number of neurons in the subdivisions of the rat species-specific calls in cortical FM–FM neurons of the mustached hippocampus using the optical fractionator, Anat. Rec. 231 (1991) bat, NeuroReport 7 (1996) 1749–1755. 482–497. [40] W.E. O’Neill, N. Suga, Encoding of target range and its representa- [61] G.H. Yeni-Komshian, D.A. Benson, Anatomical study of cerebral tion in the auditory cortex of the mustached bat, J. Neurosci. 2 asymmetry in the temporal lobe of humans, chimpanzees, and rhesus (1982) 17–31. monkeys, Science 192 (1976) 387–389.