Comparative neuroanatomy: Integrating classic and modern methods to understand association fibers connecting dorsal and ventral visual cortex

Abstract

Comparative neuroanatomy studies improve understanding of brain structure and function and provide insight regarding brain development, evolution, and also what features of the brain are uniquely human. With modern methods such as diffusion MRI (dMRI) and quantitative MRI (qMRI), we are able to measure structural features of the brain with the same methods across human and non-human primates. In this review article, we discuss how recent dMRI measurements of vertical occipital connections in humans and macaques can be compared with previous findings from invasive anatomical studies that examined connectivity, including relatively forgotten classic strychnine neuronography studies. We then review recent progress in understanding the neuroanatomy of vertical connections within the occipitotemporal cortex by combining modern quantitative MRI and classical histological measurements in human and macaque. Finally, we a) discuss current limitations of dMRI and tractography and b) consider potential paths for future investigations using dMRI and tractography for comparative neuroanatomical studies of white matter tracts between species. While we focus on vertical association connections in visual cortex in the present paper, this same approach can be applied to other white matter tracts. Similar efforts are likely to continue to advance our understanding of the neuroanatomical features of the brain that are shared across species, as well as to distinguish the features that are uniquely human.

Keywords: Comparative anatomy; Diffusion MRI; Vertical occipital fasciculus; Visual cortex; White matter.

Figure 1.. Evidence supporting the identification of the human and macaque VOF with diffusion MRI.

The macaque VOF is visible in PDD map obtained by fitting the diffusion-tensor model to dMRI data. The VOF (dotted yellow outline) depicted using the PDD map in a coronal section, from one representative human and two representative macaque dMRI datasets (left: in vivo human dMRI data at 1.25 mm isotropic resolution, measured by WU-Minn Human Connectome Project; Van Essen et al., 2013; middle: ex vivo macaque dMRI data at 0.25 mm isotropic resolution, measured at National Institute of Health and provided by F. Q. Ye and D. A. Leopold; right: in vivo macaque dMRI data at 0.75 mm isotropic resolution, measured at Max Planck Institute and provided by G. A. Keliris and N. K. Logothetis). The color (see inset, upper left) indicates the principal diffusion direction (red: left-right; green: anterior-posterior; blue: superior-inferior). STS: Superior Temporal Sulcus; IPS: Intraparietal Sulcus, Calc: Calcarine Sulcus, POS: Parieto-occipital sulcus, OTS: Occipitotemporal sulcus; IOS: Inferior Occipital Sulcus. (B) Macaque and human VOF reconstructed using ensemble tractography (Takemura et al. 2016a). VOF reconstructions are shown for in vivo human (left), ex vivo macaque (middle) and in vivo macaque (right) data. Figures are reproduced from Takemura et al. (2017) with permission.

Figure 2.. From strychnine neuronography to dMRI: The identification of vertical connections in macaque visual cortex across methods.

(A) Top: Figure 1 from Bailey et al. (1944). Schematic illustration of long fiber tracts of Macaca mulatta identified with strychnine neuronography. Note the vertical connections depicted in the occipital lobe, which Bailey and colleagues (1944) described as connecting “the posterosuperior part of the inferior parietal lobule (area 39) to the posterior part of the temporal lobe (area 37)” (pg. 414). Bottom: Figure 4 from Bailey et al. (1943). Fiber tracts in chimpanzee as determined by strychnization and electrical recording. Again, note the vertical connections depicted in the occipital lobe, which Bailey and colleagues describe in the context of the VOF. (B) Figure 1 from Petr et al. (1949). Efferent connections from the lateral surface of the temporal lobe. +, connections found; o, areas explored, but no connections found.Heaviness of crosses indicates intensity of firing. Note that strychnization ventrally induced intense firing dorsally, which contributed to Petr and colleagues to differentiate TEO from TE. Figure is reproduced from Petr et al. (1949) with permission. (C) An example inflated cortical surface reconstruction of a right hemisphere in macaque. For comparison, the orientation of lateral (top) and ventral (bottom) views of the brain are similar to those in (B). Dark gray pixels depict sulci, while light gray pixels depict gyri. Estimated VOF streamline endpoints from high-resolution dMRI data acquired from a macaque brain (subject M1 in Figure 1A-B; Top, dorsal VOF endpoint; Bottom, ventral VOF endpoint; Takemura et al., 2017).The hot color map indicates the normalized number of VOF streamlines having endpoints close to gray matter voxels.

Figure 3.. Anatomical studies of association fibers in visual cortex using the Nauta method or chemical tracers.

(A) Clarke (1994) examined the association fibers in human visual cortex using the Nauta method. The figure depicts a coronal slice of a human brain that includes a lesion site in dorsal occipital cortex (hatched area) that is adjacent to area V3A. Arrows indicate the border of visual areas based on cyto-or myeloarchitectonic criteria. Clarke reported a number of connections by inspecting terminals of degenerated axons in other areas. There are a number of axon terminals in area V4, consistent with the dMRI-based observations of the human VOF (Takemura et al., 2016b). Reproduced from Clarke (1994) with permission. (B) After anterograde and retrograde tracer injections to macaque area TEO, Webster and colleagues (1994) identified connections between V3A dorsally and TEO ventrally. Reproduced from Webster et al. (1994) with permission. (C) Schmahmann and Pandya (2006) injected anterograde tracer with radiolabeled isotope into macaque dorsal occipital cortex (case 19 in Schmahmann and Pandya, 2006). The cortical area marked in black is the injection site (V4 dorsal and V4t) of anterograde tracer. Areas highlighted with dotted red lines in the white and gray matter indicate white matter tracts between the injection site and ventral occipital cortex around the OTS, which is consistent with VOF endpoints identified with dMRI and tractography (Figure 1). We note that Schmahmann and Pandya (2006) labelled connections as portions of the Inferior Longitudinal Fasciculus (ILF), likely because it is difficult to distinguish the VOF from the ILF in the macaque brain (Takemura et al., 2017). Reproduced from Schmahmann and Pandya (2006).

Figure 4.. Modern and classic microstructural tissue properties of the VOF in human and macaque.

(A) In vivo quantitative T1 measurements (qMRI) of an example human brain. The expanded T1 scale of the inset shows that T1 in the VOF (dotted white line) is higher than in the adjacent medial tracts. Higher T1 is associated with less myelination. Image reproduced from Yeatman et al. (2014). (B) Postmortem section in human from Vogt (1904) stained for myelin. The VOF stains lighter than nearby occipital lobe tracts, which captures the same differences in occipital lobe white matter tissue properties as the quantitative T1 measures in A. (C) Postmortem section in macaque from Bonin et al. (1942). Original caption for their Figure 8B reads: “Transverse section through brain of macaque, stained after Weigert. C, stratum calcinarum; S, fasciculus transversus cunei; S. e., stratum sagittale externum; S. i., stratum sagittale internum; T, tapetum; V, fasciculus transversus lingualis; W, vertical occipital fasciculus.” The authors refer to the latter two tracts with a V and W to reflect the anatomists (Vialet and Wernicke, respectively) whose names are most often associated with those tracts throughout history. Dotted white lines have been added to the Vogt and Bonin images to highlight the location of the VOF across images included in A-C. These data suggest that microstructural differences between the VOF and nearby white matter tracts in macaque are likely measurable with qMRI.