Myeloarchitecture, intra- and interhemispheric and thalamocortical connections of area 3b representing the teeth and tongue in galagos

Significance

This study characterizes the neuronal representations of the teeth and tongue, their relationship with myeloarchitecture, and their cortical and thalamic connections in area 3b of strepsirrhine galagos, with results closely resembling those of New World monkeys. The double representation of the teeth and tongue we describe may be unique to primates. Because galagos closely resemble the earliest primate ancestors, these characteristics may have evolved even earlier and may be conserved in humans. Thus, our results push back the timeline for the emergence of this neuroanatomical organization in evolution.

Abstract

The representations of the face and oral cavity occupy a large but understudied portion of primary somatosensory cortex (S1 or area 3b) in primates. These studies are clearly important in processing food, taste, and vocalizations. Little is known about the neuronal representations of the teeth and tongue, their relationship with myeloarchitecture, and their cortical and thalamic connections in area 3b of strepsirrhine galagos. In owl and squirrel monkeys, and likely in macaque monkeys, the most lateral portion of S1 represents the face, teeth, and tongue in a series of myelin-dense ovals, including ovals for the contralateral teeth and tongue and ipsilateral teeth and tongue. As this complex representation of the teeth and tongue may differ among mammals and even primates, we sought to evaluate these representations in the galagos, representing an early branch of primates. Using a combination of neuronal responses to touch, neuroanatomical tracer injections, and tissue processing to reveal anatomical area borders, we found that lateral S1 in galagos resembles those of studied monkeys by having a series of ovals that double the representation of the tongue and teeth, along with similar intracortical, callosal, and thalamic connections, suggesting that these features of S1 emerged in the first primates.

The face and oral cavity representations in primary somatosensory cortex (S1 or area 3b) occupy most of the lateral extent of this region in primates. A complex network of connections involving the tongue representation has been identified, but the specific functions of these connections remain unclear.

Early research relating myelin-stained sections to electrode mapping data from area 3b has demonstrated that myeloarchitecture helps identify areal borders and topographic modules including for orofacial regions, in brain sections cut parallel to the surface of flattened cortex in New World owl and squirrel monkeys (1–3), marmosets (4–6), and Old World macaque monkeys (7, 8). Dark myelin patches or ovals in layer 4 of area 3b represent the body parts.

Matching receptive fields with myelin architecture in the flattened cortex preparation correlates myelin-rich ovals lateral to the “hand–face septum,” with representations of specific parts of the face (ovals F1, F2, and F3) and oral cavity [ovals O1, O2, O3, and O4; (3)]. Placing neuronal tracers into the contralateral tongue representation revealed a complex network involving adjacent areas 1, 3a, insula, secondary somatosensory area (S2), parietal ventral (PV), proposed gustatory cortex “G” (see ref. 9 for review), and orbital frontal cortex in the lateral prefrontal cortex (10), as well as connections within 3b to representations of the contralateral face, lips, teeth, and tongue, and ipsilateral teeth and tongue.

Examining thalamocortical input can help identify the type of stimulus (mechanical, thermal, or chemical) encoded in the tongue representation of area 3b. In the somatosensory thalamic nucleus, the ventroposterior nucleus (VP), mechanical and thermal stimulation of the tongue are represented in the medial VP (VPM). Medial to VPM is a parvicellular nucleus (VPMpc), which has been shown to represent taste based on retrograde degeneration after lesioning the gustatory cortex “G” and S1 (11) and neuronal responses to tactile and taste stimuli (12). While no cortical area solely dedicated to taste has been identified, it has been suggested that, in addition to the proposed gustatory cortex “G,” somatosensory areas may participate in taste processing (9, 13–15). Connection data from New World monkeys support this idea (1), with VPMpc projecting to the tongue representation in area 3b. Thus, inputs from somatosensory and taste receptors in the tongue may be integrated at the earliest levels of cortical processing (10). However, data from Old World monkeys (16) remain inconclusive, as connections from the thalamus to the tongue representation of area 3b in macaques did not include VPMpc.

Monkeys are classified as haplorrhine (“simple nose”) and anthropoid primates, which diverged within the primate lineage about 47 million years ago. Here, we study galagos, which are strepsirrhine (“curved nose” or “wet nose”) primates, diverging approximately 66 million years ago (17, 18). Therefore, strepsirrhine galagos may share characteristics with other living primates. Alternatively, they may retain traits related to the well-developed chemosensory mucosa found in strepsirrhine primates, which is less developed in primates that rely more heavily on vision (17). Our study examined how closely the representation of the tongue in contralateral area 3b in a strepsirrhine primate resembles that of anthropoids. Specifically, we characterized the neuronal receptive fields of the teeth and tongue region of area 3b, the coregistration of somatotopic maps with myeloarchitecture, ipsilateral and contralateral cortical connections, and thalamocortical connections of oral cavity representations in area 3b of galagos. Our multifaceted approaches, including architectonic, electrophysiological, and connectivity patterns, offer insights into how the brain processes somatosensory information from the oral cavity in galagos, which may represent features common to primates.

Results

As shown in Table 1, the oral cavity representation of area 3b was either briefly mapped to determine the sites of tracer injections or more thoroughly mapped to characterize somatotopy and neuronal receptive fields in six adult galagos. Cortical maps were coregistered with the myelin staining patterns of flattened cortex for all cases. Anatomical tracers were injected into the electrophysiologically determined tongue representation of the left area 3b for five cases. The results are described in five parts: 1) myelin architecture in the orofacial region of area 3b, 2) somatotopy of orofacial representation in area 3b and its relation to myeloarchitecture, 3) intracortical connections after tracer injections into the tongue representation in area 3b, 4) callosal connections, and 5) thalamocortical connections.

Table 1.

Case information

Case (Sex)	Age (years, months)	Tracer injected into Left 3b (transportation time)	Detailed mapping	Myeloarchitecture	Figures
04-19 (Female)	3y7m	Fluororuby, FR; biotinylated dextran amine, BDA (9 days)	Right hemisphere, RH	LH & RH	S2, S3, S4, S10
04-29 (Female)	5y6m	–	RH	RH	2, S3
04-42 (Female)	2y	cholera toxin subunit B, CTB (5 days)	Left hemisphere, LH	LH & RH	S1, S3, S5, S6, S8
05-44 (Female)	10y	CTB (1 day)	–	LH & RH	3, 5, 7
06-17 (Male)	3y3m	CTB (1 day)	–	LH & RH	4, 6, S9
06-31 (Male)	2y8m	CTB (1 day)	–	LH	S11

Part 1. Myeloarchitecture Reveals Modular Organizations in Orofacial Representations of Area 3b in Galagos.

Modules are characterized by a densely stained granular layer 4 in the primary sensory cortex in primates and other mammals. In the somatosensory cortex, a distinct myelin-poor septum between the myelin-dense representations of the hand and face has been reported in area 3b of both New World and Old World monkeys (2, 8), as well as between different body parts in the orofacial region of area 3b (3). Fig. 1 shows a photomicrograph of a myelin-stained section of flattened cortex from a nonexperimental galago, showing that there are roughly seven myelin-dense modules (ovals). Each oval (F1–F3, O1–O4) likely represents a specific body part (face, lips, contralateral teeth and tongue, ipsilateral teeth and tongue, respectively) as described by Jain et al. (3). Unlike anthropoid primates, galagos have a less pronounced hand–face septum in area 3b. Here, we determined the presumptive hand–face border from myelin-stained sections with a line between the caudal tip of the frontal sulcus posterior (FSp) and the rostral tip of the intraparietal sulcus (IPS), based on a previous study from Nissl preparations in galagos (19).

Fig. 1.

Myeloarchitecture of area 3b in strepsirrhine galago (Otolemur garnetti). (A) Photomicrograph of a myelin-stained section from a flattened cortex of the left hemisphere. The brain was sectioned parallel to the pial surface after unfolding the lateral sulcus. A presumptive hand–face border is indicated by a thick dashed line based on a previous study (19), which typically aligns with a horizontal line between the caudal tip of the frontal sulcus posterior and the rostral tip of the intraparietal sulcus. Lateral to the hand–face border, myelin-dense ovals are visible (indicated by circles); they represent different parts of the face and oral cavity. (B) Schematic showing myelin-dense regions of area 3b (shaded). We follow the same terminology as Jain et al. (3), where F1-F3 represent different parts of the face, and O1-O4 represent various parts of the oral cavity. Inset in B, a schematic drawing of a galago brain depicts the location of area 3b (shaded). Abbreviations: 1/2, 3a, 3b, somatosensory areas 1/2, 3a, 3b; FSa, frontal sulcus anterior; FSp, frontal sulcus posterior; IPS, intraparietal sulcus; LS, lateral sulcus; M, medial; M1, primary motor cortex; PV, parietal ventral somatosensory area; R, rostral; S2, secondary somatosensory area. The figure is modified from (10).

Part 2. Neuronal Responses and Somatotopy of Orofacial Representations in Area 3b and Correspondence With Myelin-Dense Modules.

Teeth and tongue representations in area 3b.

In this study, we focused on mapping the oral cavity region in area 3b using multiunit microelectrode recordings of neuronal responses while the body part was stimulated by touch. The detailed maps of teeth and tongue representations in area 3b were conducted in three cases (Fig. 2 and SI Appendix, Figs. S1 and S2). The teeth territory was roughly located lateral to the FSp, rostrolaterally adjacent to the lip representations. Neurons in this region responded to light touch on contralateral or bilateral incisors, as well as to contralateral canines, premolars, and molars (SI Appendix, Fig. S3). In the current study, we encountered more sites where neurons responded to light touch on the lower front teeth than on the upper teeth.

Fig. 2.

Myeloarchitectural correlation with oral cavity representations in the Right area 3b for Galago 04-29. (A) A photomicrograph of a myelin-stained section (F17) from a flattened cortex of the right hemisphere showing myelin-rich regions and ovals. White arrows point to electrolytic lesion sites. (B) Same image from A showing outlines of myelin-dense ovals located lateral to the hand–face border. The boundaries were drawn from a series of adjacent sections. Note that some ovals (dashed ovals) are not clearly apparent in the section shown. The horizontal dashed line marks the hand–face border. (C) Neuronal responses and receptive fields to tactile stimulation in the face and oral cavity region of 3b and surrounding regions are overlaid on the myelin-dense regions (shaded in gray) from B. Filled circles show excellent or good responses (ER/GR), unfilled circles show weak responses (WR), and diamonds indicate very weak responses to taps (vWR). Stars mark lesion sites. “x” indicates no response was elicited. Colors denote neuronal receptive fields from different parts of orofacial locations. Note that the most rostrolateral two ovals likely represent the ipsilateral teeth (Te) and the ipsilateral tongue (Tng) in area 3b. Abbreviations: Aud, auditory cortex; CC, corpus callosum; Cg, cingulate gyrus; CgS, cingulate sulcus. For other abbreviations, see Fig. 1. (Scale bar, 1 mm.)

More rostral and lateral to the contralateral teeth representation was the contralateral tongue representation. Neurons in this region generally responded to light touch on the contralateral tongue. With limited mapping sites, we found that neuronal receptive fields for the anterior part of the tongue were relatively smaller than receptive fields for the middle or posterior tongue.

Relationship between somatotopy and myeloarchitecture.

Toward the end of the mapping experiments, small electrolytic lesions were made at selected recording sites to relate the mapping data to the myelin-stained sections. When these maps were aligned with the myelin sections, some correspondence was observed between clusters of neuronal responses and the heavily myelinated ovals. It is worth mentioning that the brain section cut parallel to the surface of the flattened cortex was difficult to achieve perfectly. As sections may contain regions involving different layers, the myelin-stained sections were typically reconstructed across several brain sections based on blood vessels, sulci, and electrolytic lesion markers.

Fig. 2 shows an example of the flattened right cortex stained for myelin from Galago 04-29. All four lesion markers (arrows) are visible on the myelin-stained section, and three lesion markers correspond to three myelin-dense ovals, which match the microelectrode map representing contralateral lip, teeth, and tongue in caudal-to-rostral order. This progression from contralateral lips and chin, teeth, and tongue matches ovals F3, O1, and O2, respectively, and was consistent across the three galago cases studied (Fig. 2 and SI Appendix, Figs. S1 and S2).

SI Appendix, Fig. S3 illustrates the somatotopic organization of neuronal receptive fields in relation to myeloarchitecture. Overall, we observed trends in somatotopy where neuronal receptive fields for the anterior teeth are generally rostral and/or lateral to those for the posterior teeth, and receptive fields for the anterior tongue are roughly caudal and/or lateral to those for the posterior tongue, aligning with the literature (e.g., (15, 20) for review). Although we could not map the more rostral or lateral parts of area 3b corresponding to O3 and O4 territories, our findings from myeloarchitecture and tracer injection results (see below) support the idea that the galago has a similar modular organization in the orofacial region of area 3b as reported previously in New World monkeys (3).

Part 3. Intracortical (Ipsilateral) Cortical Connections of the Tongue Representation of Area 3b.

The intracortical connections were investigated by injecting neuroanatomical tracers into the electrophysiologically identified tongue and teeth regions of area 3b. Cholera toxin subunit B (CTB) was injected in four cases, and biotinylated dextran amine (BDA) and Fluororuby (FR) were injected in one case (Table 1). Labeled neurons were plotted in sections of flattened cortex cut parallel to the cortical surface, which allowed us to superimpose the anatomical connection results, electrophysiological maps, and myeloarchitectural features in the same plane. Although all three tracers (BDA, CTB, FR) labeled neurons that projected to and from the injection sites, retrogradely labeled cell bodies were more readily visualized than anterogradely labeled terminals in most cases. Thus, we focused on the distribution patterns of retrogradely labeled neurons.

Injection cores mainly restricted to the tongue oval of area 3b.

The tracer injection in Galago 05-44 targeted the tongue representation of area 3b (Fig. 3A) in the caudolateral part of module O2. The dense tracer uptake zone may have extended slightly beyond the tongue territory caudolaterally into area 1 (Fig. 3B). Labeled neurons were densely clustered around the injection site within the same oval (O2), and in the adjacent ovals presumably in the ipsilateral teeth and tongue territories of area 3b (O3, O4; Fig. 3C). Labeled neurons were also found in the neighboring areas 3a and 1, as well as in the teeth representation (O1) of area 3b. Scattered neurons were observed in S2/PV regions (Fig. 3C). A similar pattern was seen in Galago 04-19, where a smaller injection of FR labeled neurons densely around the site and in adjacent regions of areas 3a, 1, and PV (SI Appendix, Fig. S4).

Fig. 3.

Distribution of retrogradely labeled neurons from CTB injection into the electrophysiologically defined tongue representation in area 3b in Galago 05-44. (A) Photomicrograph of a myelin-stained section from a flattened cortex of the left hemisphere showing a few microelectrode mapping sites with responses to tactile stimulations, the location of injection (green star), areal borders (dashed lines), and myelin-defined orofacial ovals (circles). (B) Photomicrograph showing the CTB injection site. (C) The illustration is a composite of multiple flattened sections aligned based on anatomical landmarks such as sulci, tissue artifacts, and blood vessels. The myelin-defined anatomical landmarks, including areal borders and oval circles, are superimposed with retrogradely labeled neurons (dark green dots). Abbreviations: CTB, cholera toxin subunit B; LH, left hemisphere. For other abbreviations and conventions, see Figs. 1 and 2. (Scale bar, 1 mm.)

Injection cores involving the surrounding regions of the teeth and tongue ovals of area 3b.

Some injections extended beyond their intended targets; however, the results are still useful for comparison purposes. In Galago 04-19, a BDA injection at the teeth–tongue border of area 3b, likely including area 3a (SI Appendix, Fig. S4A), labeled neurons in both teeth and tongue regions of 3b, as well as in adjacent areas 3a and 1. Sparse labeling appeared in S2, PV, and parietal rostral area PR (SI Appendix, Fig. S4C). Anterograde BDA-labeled axons showed reciprocal connections with these somatosensory areas and projected to the gustatory cortex, anterior insula, and prefrontal cortex (SI Appendix, Fig. S4C).

In Galago 06-17, two CTB injections were placed in the oral cavity region (Fig. 4): one in the tongue representation of area 3b, similar to Galago 05-44, and another near the 3a/3b border. Dense CTB-labeled neurons surrounded the injection sites in O2 and extended into areas 3a, 1, and possibly the orofacial region of M1 (Fig. 4C). Less dense but abundant labeling appeared medially in the teeth representation of O1 and laterally in the proposed ipsilateral tongue territory. Scattered neurons were found in the orofacial regions of somatosensory areas 3b, S2, PV, PR, gustatory cortex, and anterior insula. Notably, this case showed widespread labeling in the frontal cortex, including dense clusters in the forelimb region of M1 and sparse neurons in premotor, prefrontal, orbitofrontal, cingulate, and temporal cortices.

Fig. 4.

Distribution of retrogradely labeled neurons from two CTB injections into the electrophysiologically defined tongue representation and myelin-defined tongue oval in area 3b for Galago 06-17. (A) Photomicrograph showing two CTB injection sites in area 3b. (B) Photomicrograph of a myelin-stained section from a flattened cortex of the left hemisphere showing microelectrode mapping sites with responses to tactile stimulations, two CTB injection sites (stars), areal borders (dashed lines), and myelin-defined orofacial ovals. Note that some ovals (dashed ovals) are not clearly apparent in the section shown. (C) Schematic of myelin-defined anatomical landmarks with superimposed retrogradely labeled neurons (dark green dots). Abbreviations: G, gustatory cortex; OFC, orbitofrontal cortex; PMd, dorsal premotor cortex; PMv, ventral premotor cortex; PR, parietal rostral somatosensory area. For other abbreviations and conventions, see Fig. 2. (Scale bar, 1 mm.)

In Galago 04-42, two CTB injections were placed in different myelin-stained ovals (O2, O4) within area 3b (SI Appendix, Fig. S5). Dense labeling surrounded both sites and extended into adjacent areas 3a and 1. This case also showed widespread cortical connections, with strong labeling in frontal regions, including the orofacial region of M1, premotor, prefrontal, and orbitofrontal cortices. Scattered neurons were also found in the retroinsula (Ri), temporal cortex, and possibly auditory cortex (Aud).

In summary, tongue representations of 3b are densely connected within the same myelin-stained oval and, to a lesser extent, to neighboring ovals representing teeth and the ipsilateral tongue in area 3b, as well as in nearby regions of areas 3a and 1. This suggests that different parts of the oral cavity are highly interconnected. Large tracer injections involving the orofacial region of area 3a resulted in widespread labeled neurons in M1, premotor, orbitofrontal cortices, somatosensory areas S2 and PV, and insular areas.

Part 4. Interhemispheric (Callosal or Contralateral) Cortical Connections of the Tongue Representations of Area 3b.

The callosal connections of the 3b oral cavity representation were examined in three cases (Galagos 05-44, 06-17, and 04-42). As described above, labeled neurons were plotted from sections cut parallel to the brain surface, and the results were carefully superimposed on adjacent brain sections processed for myelin. The myelin-stained sections allowed the teeth and tongue representations of area 3b of the right hemisphere to be identified (Figs. 5B and 6B and SI Appendix, Fig. S6B). In Galago 04-42, the teeth region of right area 3b was also identified by electrophysiological responses (green dots in SI Appendix, Fig. S6B).

Fig. 5.

Interhemispheric cortical connections for Galago 05-44. (A) Photomicrograph of myelin-stained section of flattened cortex from the left hemisphere showing CTB injection site (star) in the tongue representation in area 3b (filled and unfilled red circles). (B) Photomicrograph of a myelin-stained section from a flattened cortex of the right hemisphere showing the location of area 3b with myelin-dense regions and ovals. (C) Schematic of myelin-defined anatomical landmarks with superimposed callosally connected retrogradely labeled neurons (green dots). Abbreviations: RH, right hemisphere; Te, teeth. For other abbreviations and conventions, see Figs. 1 and 2. (Scale bar, 1 mm.)

Fig. 6.

Interhemispheric cortical connections for Galago 06-17. (A) Photomicrograph of myelin-stained section of flattened cortex from the left hemisphere showing two CTB injection sites (stars) in the tongue representation region in area 3b (filled and unfilled red circles). (B) Photomicrograph of a myelin-stained section of flattened cortex from the right hemisphere showing myelin-dense regions and ovals. (C) Schematic of myelin-defined anatomical landmarks with superimposed callosally connected retrogradely labeled neurons (green dots). (D) Photomicrograph showing an example of CTB-labeled neurons (arrowheads) in the right temporal lobe. Abbreviation: Tng, tongue. For other abbreviations and conventions, see Figs. 2 and 5. (Scale bar, 1 mm.)

The main findings include: 1) Callosal connections in the right hemispheres were observed in all three cases (Figs. 5C and 6C and SI Appendix, Fig. S6C). 2) After CTB injection into the left 3b tongue representation, labeled neurons in the right hemisphere were mainly in the oral cavity regions of areas 3b, 3a, and 1, and they were mostly homotopic, with scattered labeling in the ipsilateral tongue oval (O4) in areas 3b and 3a (Fig. 5C). 3) When CTB was injected into the border zone or area 3a, dense callosal connections extended into oral cavity regions of areas 3a and 1 (Fig. 6C and SI Appendix, Fig. S6C). 4) Interhemispheric labeled neurons were also found in a more rostrolateral patch, likely representing the ipsilateral tongue (O4) of right 3b in all cases. 5) Homotopic connections were seen for both tongue and teeth representations of area 3b. In Galago 04-42, CTB injection into the ipsilateral teeth representation (O3) of left 3b led to labeled neurons in the ipsilateral teeth territory (O3) of right 3b (SI Appendix, Fig. S6C). 6) An unexpected finding in Galago 06-17 showed that larger CTB injections into the left tongue representation (O2) of 3b and adjacent area 3a led to callosal labeling in the temporal cortex caudolateral to the right auditory cortex (Fig. 6 C and D).

Taken together, unlike the lack of callosal connection in the hand region of area 3b in primates (21), neurons in the oral cavity regions representing the teeth and tongue of area 3b in both hemispheres are directly connected and are generally homotopic.

Part 5. Thalamocortical Connections of the Teeth and Tongue Representation of Area 3b.

Accurately defining the subnucleus boundaries of the thalamus is crucial for understanding thalamocortical connections, especially when these vary across species. We analyzed these connections for the teeth and tongue representations by overlaying serial section plots on CO-stained sections for five cases. See SI Appendix, Fig. S7 for examples of architectural subdivisions.

Fig. 7 shows thalamocortical connection patterns in Galago 05-44, where a CTB injection was in the caudolateral tongue representation of area 3b (Fig. 7A). Retrogradely labeled neurons were densely packed in the medial subdivision of the ventroposterior nucleus (VPM), and in the parvicellular subdivision (VMPpc), with additional labeling in the superior subnucleus of VP (VPS) and regions outside VP, including the anterior pulvinar (PA). In Galagos 04-42 (SI Appendix, Fig. S8) and 06-17 (SI Appendix, Fig. S9), similar thalamocortical patterns were observed, with extra labeling in the central median (CM), intralaminar (IL), and medial dorsal (MD) nuclei, likely due to injections near area 3a. In Galago 04-19 (SI Appendix, Fig. S10), where a small injection of FR was placed in the center of the tongue representation of area 3b, and in Galago 06-31 where the CTB injection most likely involved the tongue representation of area 1 (SI Appendix, Fig. S11), labeling was confined to VPM, excluding VMPpc. Overall, the tongue representation in area 3b primarily receives projections from the VPM, with fewer labeled neurons in VPMpc and sparse labeling in other thalamic somatosensory and sensorimotor nuclei.

Fig. 7.

Distributions of CTB-labeled neurons in the left thalamus for Galago 05-44. (A) Photomicrographs of two coronal sections (69 and 79) showing examples of immunoreacted CTB-labeled neurons and anterogradely labeled axon terminal fields in the ventroposterior medial nucleus (VPM). Note that immunostaining for revealing the CTB tracer also shows dense patches of cells separated by cell-poor septa, which resemble patterns seen in Nissl and CO staining. This feature allows for precise coregistration of the labeled neuron locations with the subdivisions of the thalamus. (B) Schematic showing the CTB injection site in the tongue representation of the left area 3b. Areal boundaries and myelin-dense ovals are based on myelin outlines from Fig. 6. (C) A rostrocaudal series of coronal sections showing the locations of retrogradely labeled neurons (green dots). Subdivisions of the ventroposterior nucleus of the thalamus are defined based on alternate sections stained for cytochrome oxidase (CO). Abbreviations: CM, centromedian nucleus of the thalamus; D, dorsal; IL, intralaminar nucleus of the thalamus; L, lateral; M, medial; PA, anterior pulvinar; V, ventral; VPI, inferior subdivision of ventroposterior nucleus of the thalamus; VPL, lateral subdivision of ventroposterior nucleus of the thalamus; VPM, medial subdivision of ventroposterior nucleus of the thalamus; VPMpc, parvicellular subdivision of the VPM; VPS, superior subdivision of ventroposterior nucleus of the thalamus. For other abbreviations and conventions, see Figs. 1, 2, and 5. (Scale bar, 1 mm.)

Discussion

The present study is part of ongoing efforts to learn how primate brains are similar or vary across branches of primate evolution. Primates provide many opportunities for comparative research, as this highly diverse order includes 14 families and over 350 species (22). Archaic primates appeared around 82 million years ago (mya) and formed the strepsirrhine (lemurs, lorises, and galagos) and haplorrhine branches approximately 66 mya (17, 18). What characteristics are likely shared in extant primates, including humans? Answers come from comparative studies of the brains, especially those of other primates.

Brain organization has been studied extensively in strepsirrhine galagos, as well as in New World and Old World monkeys. Although galago and monkey brains differ considerably, they also share features that distinguish them from nonprimates.

Here, we specifically focused on the teeth and tongue representation in the primary somatosensory area 3b in galagos. Together with early evidence obtained from New World monkeys, we found that the teeth and tongue are represented twice in area 3b of each hemisphere in studied primates. However, the two representations are not simply repeated. Neurons in the myelin-dense module O1 predominantly respond to touch on contralateral teeth, and neurons in module O3 predominantly respond to touch on ipsilateral teeth. Likewise, the tongue representations in area 3b are predominantly contralateral in module O2 and predominantly ipsilateral in O4.

This finding is important in several aspects. First, it pushes back the emergence of this neuroanatomical organization in the evolutionary timeline. Strepsirrhine primates emerged around 66 mya, and strepsirrhine galagos are closely related to ancestral primates. This suggests that all primates, including humans, likely have double representations of the teeth and tongue in area 3b. Second, as galagos appear to closely resemble the earliest ancestors of primates in many features, these features of S1 may have evolved even earlier. Third, this report raises questions about the double representations of these features in mammals. Is the organization of the double representation of the teeth and tongue unique to primates? Further study is needed.

Myeloarchitecture and Somatotopy of the Oral Cavity Representation in Area 3b of Galagos Resemble Those of Monkeys.

Our summary of how the face and mouth are represented in the primary somatosensory cortex across these major branches of primate evaluation is illustrated in Fig. 8. The brains of galagos and smaller New World monkeys offer some advantages for microelectrode recordings and tracer injections, as much of area 3b is accessible on the brain’s surface, except for the rostral tip that extends into the ventral cortex of the frontal lobe in some monkeys (as in owl monkeys).

Fig. 8.

Schematic drawings showing the locations of area 3b (shaded) in five primate species. The face and oral cavity representations of area 3b are divided into seven myelin-rich ovals: F1 (Upper face), F2 (Upper lip), and F3 (chin and Lower lip). Rostral to F1-F3 are four more ovals: O1 (contralateral teeth), O2 (contralateral tongue), O3 (ipsilateral teeth), and O4 (mainly ipsilateral tongue) (3). Note that in galago and marmoset, the labels O3-O4 are used instead of “O3” and “O4” because detailed electrophysiological mapping data to confirm them are not available, but these are present in the myelin preparations. For other abbreviations, see Figs. 1 and 2.

In galagos, myelin modules are less distinct compared to anthropoid primates. Specifically, we did not detect a highly distinct hand–face border in myelin preparations in galagos, in strong contrast with anthropoids such as owl and squirrel monkeys (3) and marmosets (4, 6), and macaques (2, 8) Despite this, the ovals revealed by myelin stains in galagos allow for oval-by-oval comparisons of related features, such as connections, between galagos and other primates.

In owl and squirrel monkeys (3), distinct myelin-rich ovals in area 3b represent orofacial regions, such as the upper face (F1), upper lip (F2), lower lip and chin (F3), contralateral teeth (O1), contralateral tongue (O2), ipsilateral teeth (O3), and ipsilateral tongue (O4), arranged in caudal-to-rostral sequence. This pattern, or something similar, also seems to occur in smaller New World tamarins and marmosets (1, 4, 6, 23).

Old World macaque monkeys, with larger brains and bodies, are commonly used in studies of cortical organization because their brains are more like the human brain. However, studying area 3b has been more challenging since much of it is hidden in the caudal bank of the central sulcus, similar to humans. The hand–face junction in lateral 3b is well defined, with the cortex just lateral to it representing the lower face rostrally in the unfolded map (deep in the sulcus), and the upper face caudally. Teeth are also represented nearby (see refs. 24, 25 for review). However, the exact relationship between orofacial representations and the myelin-dense ovals has not been fully clarified, especially for the ipsilateral teeth and the most ventrolateral representation of the ipsilateral tongue in macaque monkeys.

Human studies of face and mouth representation in area 3b are incomplete, partly due to overlap with neighboring areas 3a and 1. However, fMRI studies show a similar somatotopic arrangement in humans as found in other primates, particularly in the representation of the lip, teeth, and tongue in area 3b (26). The ipsilateral responses to tactile stimulation on the tongue in areas 3b and 1 are also found in humans (27). Early reports from cortical microstimulation in neurosurgery patients found that when ipsilateral, bilateral, or midline sensations of the tongue were reported, they tended to be grouped separately from the contralateral responses (28), further suggesting that the organization in nonhuman primates is preserved in humans.

These ipsilateral orofacial representations in area 3b of monkeys (O4) and humans are not fully elucidated, but are likely to arise from cranial nerve projections that are contralateral only or bilateral, depending on the innervation (24), and connections to, the brainstem (29), subdivisions of the VPM of the thalamus (30) and cortical callosal projections, as discussed below. The ipsilateral representations of the oral cavity in somatosensory area 3b and the VP of thalamus have been reported in primates (1, 3, 25, 31) and other animal species such as sheep (32, 33) and cats (34). But the existence among other animal species has been variably described (see ref. 35 for review). There are several possible sources of ipsilateral receptive fields in the primary somatosensory cortex oral cavity region, such as through the ipsilateral trigeminothalamic pathway (30, 36), and direct callosal [(21, 37), see ref. 38 for review] or indirect callosal activations through higher-order somatosensory cortices areas 2, 5, S2 (see refs. 39–41 for review).

Connectivity of Tongue Representations in Area 3b of Galagos Resembles That of Monkeys.

As shown in a summary figure (SI Appendix, Fig. S12), galagos share corticocortical and thalamocortical connectivity patterns with other primates (1, 7, 16). The tongue representation in area 3b connects to surrounding orofacial representations within area 3b, as well as to other somatosensory areas 3a and 1. With larger injections that likely included adjacent parts of areas 3a and 1, labeled neurons were found in the primary motor cortex, premotor cortex, parietal cortex, somatosensory areas S2 and PV, orbitofrontal cortex, and the anterior insula that may include the primary gustatory area G (see ref. 9 for review). The forelimb connections seen in M1 after two CTB injections into the tongue representations in areas 3b and 3a in galago 06-17 are most likely due to the injection including area 3a. Area 3a is involved in sensorimotor integration, and it has much denser connections with motor and posterior parietal cortex (42). It also has very broad, topographically mismatched connections with the forelimb and face representations in primates [(43); see ref. 44 for review]. Understanding the functional implications of such connections and any special role of the regions that appear as dense myelin ovals will require further study. These widespread connections suggest that different parts of the area 3b representation of the oral cavity are highly interconnected in both anthropoid monkeys and strepsirrhine galagos, thus facilitating the integration of orofacial cutaneous sensation with tongue and jaw movements (45, 46) for mastication, recognizing and discriminating food during eating (47), and coordinating motor control of chewing and swallowing (see ref. 48 for review). The direct connections between the 3b tongue region and the primary gustatory cortex suggest that the primary somatosensory cortex is involved in taste processing in primates.

Callosal connections are broadly, but unevenly distributed, involving all or nearly all cortical areas (49). The representations of midline structures of the body in area 3b (such as face and trunk) are densely connected, but the distal extremities of the two sides of the body (such as fingers and toes) have few callosal connections. This phenomenon is called “the midline rule of interhemispheric interaction” or “midline fusion” (see refs. 41, 50 for review). Here, we found that callosal connections from the teeth and tongue region in the noninjected hemisphere of galagos were homotopically connected with those of representations in the area 3b of the injected hemisphere. Moreover, clusters of callosally labeled neurons were also observed in the contralateral cortex at more anterior and lateral locations (Figs. 5C and 6C and SI Appendix, Figs. S6C), which likely represent the ipsilateral teeth (O3) or tongue (O4) in area 3b. These findings align with those observed in New World monkeys (1). Additionally, the direct callosal connections between the tongue representation in the left somatosensory cortex and the right temporal cortex have not been documented in detail in the literature. Interestingly, in one galago with larger CTB injection sites that likely extended to areas 3a and 1, more labeled neurons appeared in the temporal cortex of the noninjected hemisphere than in the injected side of the temporal cortex (Fig. 6C). We speculate that our anatomical findings are reflective of human functional connectivity between the tongue representation in sensorimotor areas and the temporal cortex, which may facilitate vocalization processing and perception (28, 51–53). Our findings support the suggestion that there is topography and lateralization of the contralateral and ipsilateral connections related to the tongue representation in primates that will require more study.

The thalamocortical projections from the VPM and VPMpc to area 3b in galagos align with those found in New World monkeys (1), while a study in Old World macaques (16) reported that projections from the VPMpc are less prominent. On the other hand, other studies in macaque monkeys have found that VPMpc projects to two cortical areas: one to the ipsilateral insular-opercular cortex and the other to the somatosensory area 3b, located near the anterior precentral sulcus (54) (see ref. 15 for a review). The inconsistency in the thalamocortical projections from different studies with primates is likely due to the precise location and extent of the uptake zone of tracer injection, suggesting that neurons related to taste processing in area 3b may be scattered and modulatory. Indeed, in the present study, we did not detect the direct projections from VPMpc to tongue representations in area 3b in 2 out of 5 galagos. One (Galago 04-19) case had a focal injection of FR, which is a tracer known to have a restricted uptake zone (4); the other (Galago 06-31) had a CTB injection with tracer uptake that most likely occurred in the tongue region at the border of areas 3b and 1, rather than more centrally in the area 3b tongue oval. Since VPMpc is the thalamic relay nucleus for taste processing [(55); see ref. 56 for review], it is probable that in humans, collaterals from VPMpc terminate in primary somatosensory cortex regions representing oral and perioral structures (9).

Conclusions

We examined galagos as a representative of a branch of primates that diverged early in primate evolution, and we found similarities to other studied monkeys. Myelin-stained brain sections cut parallel to the pia surface of flattened cortex in galagos reveal a series of modular, organized ovals representing the face, lips, contralateral teeth and tongue, and ipsilateral teeth and tongue in area 3b (S1). The connection patterns of the orofacial representations in area 3b of galagos are similar to those described in New World monkeys. These cortical regions connect with surrounding orofacial regions of somatosensory areas 3a, 1, and S2/PV in the same hemisphere, and connect with their counterparts in somatosensory areas 3a, 3b, and 1 in the opposite hemisphere. These regions of 3b receive most of their inputs from the somatosensory thalamus (VPM), and fewer projections from the thalamic taste nucleus, VPMpc. In addition to those connection patterns, large injections involving areas 3a and 3b also labeled neurons in M1, orbitofrontal cortex, anterior insula, and gustatory cortex. These connections likely play roles in integrating touch and taste sensations, coordinating tongue and jaw movements for activities like chewing and swallowing, and supporting vocalization. The organization of the teeth and tongue representations for contralateral and ipsilateral parts in each hemisphere may be a common feature in primates yet to be revealed in humans.

Materials and Methods

Six adult galagos (Otolemur garnetti, 2 males, 4 females) were included in this study (Table 1) for mapping neuronal receptive fields, injecting neuronal tracers, and preparing tissue for neuroanatomical analysis. In three cases, more detailed mappings of neuronal receptive fields in area 3b were conducted to reveal representations for the contralateral teeth and tongue. All experimental procedures adhered to NIH Guidelines for the Use of Laboratory Animals and were approved by the Vanderbilt University Institutional Animal Care and Use Committee. For detailed methods, see SI Appendix.

Surgical Procedures.

The surgical procedures followed those described in previous studies [e.g., (57) and SI Appendix]. Briefly, each galago was initially sedated with an intramuscular (IM) injection of ketamine hydrochloride (10 to 25 mg/kg) and kept at a surgical anesthesia level with inhaled isoflurane during preparation and surgical procedures. After securing the galago’s head with a stereotaxic frame, under aseptic conditions, a craniotomy and durotomy were performed over the fronto-parietal cortex of the left hemisphere. Brief receptive field mapping was performed to identify the area 3b teeth and tongue representations after anesthesia was transitioned to ketamine infusion with isoflurane or xylazine administered to maintain anesthetic levels. Neuroantomical tracers of Fluororuby (FR; Invitrogen, Carlsbad, CA), biotinylated dextran amine (BDA; Invitrogen), and cholera toxin subunit B (CTB; Sigma-Aldrich, St. Louis, MO) were among the tracers used in cases to target the tongue and teeth representations of area 3b, as previously described (58). See SI Appendix for details of the surgery, mapping, tracer injections, closure, and postoperative care for cases with tracer transport times exceeding one day. Following tracer injections (short tracer transportation cases) or 5 to 9 d (longer tracer transportation cases), terminal somatosensory mapping procedures were conducted. Standard multiunit microelectrode mapping was used to determine the somatotopic organization of somatosensory area 3b (59) under general anesthesia with a mixture of ketamine and xylazine as previously described (57, 60). At the end of the final recording session, electrolytic lesions were created by passing anodal direct current (10 µA for 10 seconds) through the recording electrode at strategic locations within the teeth and tongue representation to align the recording results with cortical architecture after histological tissue processing.

Tissue Processing.

At the end of each experiment, the galago was deeply anesthetized with a lethal dose of pentobarbital (120 mg/kg, IV) and perfused as described previously (61). The brain was removed, and the cortex was separated from the underlying tissue and flattened by placing the cortical sheet between glass plates (62). The flattened cortex was cut parallel to the pial surface on a freezing sliding microtome at 40 μm, and the thalamus was cut coronally at 40 or 50 µm.

For the cortex and thalamus, alternate series of sections were processed for visualizing tracers and architecture. The BDA series was visualized by an avidin biotin-peroxidase reaction, and the CTB series was visualized with immunohistochemistry (57, 63). For histological reconstructions, a series of cortical sections was reacted for myelinated fibers (64); modified by (2) and for the thalamus, a series of sections was processed for cytochrome oxidase (CO, (65). For more details, see SI Appendix.

Data Analysis.

Following injections of BDA, CTB, and FR in area 3b, the distributions of labeled neurons in cortex and thalamus were systematically plotted using the Neurolucida system (MBF Bioscience, Williston, VT). To evaluate cortical connection patterns in relation to somatotopy and myelin architecture, we superimposed the plots of BDA, CTB, and FR-labeled neurons and the somatosensory maps onto images of myelin-stained sections based on landmarks and lesion sites for ten hemispheres of six galagos. For the alignment of the thalamus, plots of all serial sections were superimposed over the outlines of CO-dense modules of the ventroposterior nucleus of the thalamus based on landmarks of contours and blood vessels (SI Appendix).

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.