Thalamic connections of the caudal part of the posterior parietal cortex differ from the rostral part in galagos (Otolemur garnetti)

Abstract

In this study, thalamic connections of the caudal part of the posterior parietal cortex (PPCc) are described and compared to connections of the rostral part of PPC (PPCr) in strepsirrhine galagos. PPC of galagos is divided into two parts, PPCr and PPCc, based on the responsiveness to electrical stimulation. Stimulation of PPC with long trains of electrical pulses evokes different types of ethologically relevant movements from different subregions (“domains”) of PPCr, while it fails to evoke any movements from PPCc. Anatomical tracers were placed in both dorsal and ventral divisions of PPCc to reveal thalamic origins and targets of PPCc connections. We found major thalamic connections of PPCc with the lateral posterior and lateral pulvinar nuclei, distinct from those of PPCr which were mainly with the ventral lateral, anterior pulvinar, and posterior nuclei. The anterior, medial, and inferior pulvinar, ventral anterior, ventral lateral. and intralaminar nuclei had fewer connections with PPCc. Dominant connections of PPCc with lateral posterior and lateral pulvinar nuclei provide evidence that unlike the sensorimotor-orientated PPCr, PPCc is more involved in visual-related functions.

Introduction

The present investigation is part of a series of studies aimed at advancing our understanding of the posterior parietal cortex (PPC) in a variety of primate species from the aspect of anatomical connectivity. The cortical region surrounding the intraparietal sulcus (IPS) is designated as PPC, which resides between the rostrally located somatosensory cortex, the caudally located visual cortex, and the laterally located auditory cortex. In humans and non-human primates, the importance of PPC in the process of producing intended actions, from integrating multiple sensory inputs to generating goal-directed movements, has been supported by anatomical, functional, and lesion studies (Buneo and Andersen, 2006; Cavada and Goldman-Rakic, 1989; Kastner et al., 2017; Lamotte and Acun, 1978; Padberg et al., 2010; Rathelot et al., 2017; Strick and Kim, 1978; for reviews, see Andersen, 1995; Calton and Taube, 2009; Cohen and Andersen, 2002; Culham and Valyear, 2006; Darian-Smith et al., 1979). Multiple lines of research indicate that distinct parieto-frontal networks exist for different actions, such as reaching, grasping and manipulating objects, defensive gestures of forelimbs and face, or control of eye movements. Of each network, a certain small part of PPC is involved as a critical node (Cooke et al., 2015; Gharbawie, Stepniewska and Kaas, 2011; Graziano and Cooke, 2006; Johnson et al., 1996; Shields et al., 2016; Stepniewska et al., 2011, 2014, 2020; for reviews, see Kaas and Stepniewska, 2016; Wise et al., 1997). Thus, it has been proposed that PPC contains several functional zones associated with sensorimotor or visuomotor behaviors that are organized in the form of “action maps” (Stepniewska et al., 2005; Gharbawie, Stepniewska and Kaas, 2011).

Previous studies from our laboratory as well as others provided compelling evidence in support of this proposal. Using long-train intracortical microstimulation (LT-ICMS) technique, we identified functionally distinct zones, termed “domains”, in PPC of major branches of primates, including strepsirrhine galagos (Stepniewska, Fang et al., 2009; Wang et al., 2021), New World owl monkeys and squirrel monkeys (Gharbawie, Stepniewska and Kaas, 2011), and Old World macaque monkeys (Baldwin et al., 2018; Gharbawie, Stepniewska, Qi et al., 2011). From each of these PPC domains, LT-ICMS evoked a specific type of complex movement (e.g., grimacing, reaching, climbing, etc.). The concept of PPC domains was also corroborated by experiments with cortical deactivation and optical imaging methods (Stepniewska et al., 2011, 2014, 2023; Cooke et al., 2015). We further revealed distinct connectional patterns of different PPC domains by cortical injections of tracers, suggesting an anatomical basis for such functional distinctions in PPC (Gharbawie et al., 2010; Gharbawie, Stepniewska and Kaas, 2011; Stepniewska, Cerkevich et al., 2009; Wang et al., 2021, 2023).

In the systematically studied PPC of galagos, LT-ICMS evoked movements only from the rostral portion of PPC (PPCr), but failed to elicit any movements from the caudal part (PPCc). The non-responsive nature of PPCc makes its functional aspects more difficult to infer. It is also difficult to determine physiologically whether PPCc contains functional domains like PPCr, or if PPCc is rather homogeneous with minimal subdivisions. Though the functions of PPCc are less clear, anatomical studies revealed that PPCc has major connections with cortical extrastriate visual areas (e.g., V2, V3, DM, DL, etc.) and intrinsic connections within the PPC region (Beck and Kaas, 1998a; Krubitzer and Kaas, 1993; Stepniewska et al., 2016). This is different from PPCr that predominantly connected with frontal motor and higher-order somatosensory areas in addition to intrinsic PPC connections (Stepniewska, Cerkevich et al., 2009; Wang et al., 2021), implying a pivotal role of PPCc in providing highly processed visual information to PPCr domains to guide motor behaviors (Figure 1). Differences in connection patterns between PPCc and PPCr suggest they are differentially involved in the production of motor behaviors.

Figure 1.

Digital images of the lateral (a) and dorsal (b) view of the left hemisphere of a galago brain. Intraparietal sulcus (IPS) is marked with black line. The caudal portion of posterior parietal cortex (PPCc) is shaded in gray. A schematic diagram of the left hemisphere of the galago cortex in dorsolateral view (c). PPC is shaded in pink-sage gradient to reflect proposed different functional relevance between rostral and caudal PPC. Related cortical areas are marked with corresponding colors. 1–2, somatosensory areas 1 and 2; 3a, somatosensory area 3a; 3b, primary somatosensory area; DM, dorsomedial visual area; DL, dorsolateral visual area; M1, primary motor cortex; MT, middle temporal cortex; PMd, dorsal premotor area; PMv, ventral premotor area; PPCc, caudal posterior parietal cortex; PPCr, rostral posterior parietal cortex; PV, ventral parietal area; S2, secondary somatosensory area; SMA, supplementary motor area; V1, primary visual area; V2, secondary visual area; V3, third visual area. CgS, cingulate sulcus; FS, frontal sulcus; IPS, intraparietal sulcus; LS, lateral sulcus; STS, superior temporal sulcus.

We conducted a collection of studies to elaborate the functional organization of PPC in galagos, a representative of early strepsirrhine primates with many of the proposed cortical areas of other primates (e.g., Kaas, 2013; Wong and Kaas, 2010). Previous studies concerned cortical and thalamic connections of PPCr, as well as cortical connections of PPCc (Stepniewska et al., 2016; Stepniewska, Cerkevich et al., 2009; Wang et al., 2021, 2023). To further reveal the global connections of PPCc, this follow-up study focused on its thalamic connections, as thalamic inputs provide and modulate information flow in the cortex (Sherman and Guillery, 2002). The main goals of this study were, first, to determine the thalamic sources and targets of projections to and from PPCc, and second, to compare and contrast thalamic connection patterns of PPCr and PPCc. We employed classic tract-tracing methods combined with histochemical methods to localize thalamic projections with high accuracy. Our results reveal that PPCc has major connections with the lateral posterior nucleus (LP) and the visual pulvinar (mostly lateral pulvinar, LPul), and relatively fewer connections with the motor thalamus (ventral anterior, VA, and ventral lateral nuclei, VL), in agreement with the visually dominant pattern of cortical connections of PPCc. By synthesizing current and previous results, we provide more insights into the functional organization of the whole PPC region in primates.

Materials and Methods

Six galagos (Otolemur garnetti) of both sexes aging from 12 to 45 months were examined in the present study. All animals have been reported previously in related studies of cortical connections of PPC (Stepniewska, Cerkevich et al., 2009; Stepniewska et al., 2016). Thus, multiple anatomical tracers were injected into various locations throughout PPC. Injections placed in the caudal portion of PPC, where electrical stimulations failed to evoke body movements, were of particular interest in the present study. All experimental procedures followed the Guide for the Care and Use of Laboratory Animals established by the National Institutes of Health and were approved by the Animal Care and Use Committee of Vanderbilt University.

Surgery and tracer injection

Our procedures have been described in detail previously (Stepniewska, Cerkevich et al., 2009; Stepniewska et al., 2016). Briefly, galagos were given an intramuscular injection of ketamine hydrochloride (20–40 mg/kg) initially and were maintained at a surgical level of anesthesia with isoflurane (1–2%) delivered through a tracheal tube during the procedure. After fixing the animal on the stereotaxic apparatus, a craniotomy and a durotomy were performed on one hemisphere under aseptic conditions to expose the PPC region surrounding IPS. Borders between the rostral and caudal portion of PPC were established upon the responsiveness of PPCr, but not PPCc, to long-train intracortical microstimulation (LT-ICMS). During the ICMS mapping stage, isoflurane was replaced with ketamine hydrochloride delivered intravenously with an infusion pump to maintain the animal at a proper level of anesthesia while not suppressing the cortical responses to ICMS. Different anatomical tracers [cholera toxin subunit B, CTB (1% in distilled water); biotinylated dextran amines, BDA (MW 3k and 10k 1:1 mixed, 10% in 10mM phosphate buffer); fast blue, FB (3% in distilled water); fluoroemerald, FE (10% in distilled water); fluororuby, FR (10% in distilled water)] were used. At all chosen locations in electrophysiologically defined PPCc and PPCr, tracers were pressure injected into the cortex at two depths, 1.6–2 mm and 1 mm, below the pial surface with glass pipettes attached to Hamilton syringes. Information for all injections is summarized in Table 1. After completing injections, the exposed cortex was covered with Gelfilm and the skull opening was sealed with an artificial bone cap made of dental cement before the skin sutured. Galagos were closely monitored during recovery from anesthesia and received antibiotics and analgesics in the following days. 5 to 9 days after injections, allowing for tracer transport, a terminal LT-ICMS session for additional mapping of the PPC region was performed on some of the galagos.

Table 1.

Experimental and histological information of cases

Case	Sex	Age (months)	Injected region	Tracer	Section plane	Section thickness (μm)	Injection volume (μL)	Post-injection period (days)
04–04	female	32	dorsal PPCc	FE	horizontal	50	1.8	7
			dorsal PPCr	FR			1.8
			ventral PPCr	BDA			1
04–07	male	45	dorsal PPCc	CTB	coronal	50	1.8	9
			dorsal PPCr	FR			1.4
			dorsal PPCr	FE			2
			ventral PPCr	BDA			1.5
04–39	female	21	dorsal PPCc	BDA	sagittal	40	2	7
			ventral PPCc	FB			1
			dorsal PPCr	FR			1.2
			ventral PPCr	FE			0.8
04–47	male	14	ventral PPCc	BDA	coronal	40	1	7
			dorsal PPCc	FR			0.7
			ventral PPCr	FE			0.5
05–15	female	14	ventral PPCc	BDA	coronal	40	2	5
			dorsal PPCc	FR			1
			dorsal PPCr	FB			0.4
05–20	male	12	ventral PPCc	BDA	coronal	50	3.2	7
			dorsal PPCc	FR			1.9

Perfusion and tissue processing

Animals were administered a lethal dose of sodium pentobarbital (120 mg/kg) at the end of terminal experiments, and were subsequently perfused transcardially with 0.9% phosphate-buffered saline (PBS), and then fixed with 2–4% paraformaldehyde in phosphate buffer (fixative) followed by 10% sucrose in fixative. The thalamus was separated from the cortex (except cases 05–15 and 05–20) and stored in 30% sucrose solution at 4°C for 1–2 days before frozen microtomy. In four cases, the thalamus was sectioned coronally, while a horizontal and a sagittal plane of cutting were used in two additional cases to show connections in different views of thalamic nuclei. Tissue blocks were cut at a thickness of 40–50 μm and saved in several series. Two to three series were either mounted, air-dried, and coverslipped for fluorescence microscopy, or processed to reveal CTB (Angelucci et al., 1996) and BDA (Veenman et al., 1992) labeled neurons. The remaining series were stained for Nissl substance and acetylcholinesterase (AChE; Geneser-Jensen and Blackstad, 1971) to reveal thalamic architecture.

Data analysis and anatomical reconstruction

In the series of sections devoted to visualizing neurons labeled with fluorescent tracers, labeled neurons were plotted using Igor Pro 3.14 (WaveMetrics Inc., Portland, OR) coupled with a Microcode II digital readout (Boeckeler Instruments Inc., Tucson, AZ) and a Leitz Orthoplan 2 microscope (Leica Microsystems Inc., Deerfield, IL) under fluorescence illumination. Stereo Investigator 2019 (MBF Bioscience, Williston, VT) coupled with a Zeiss Axio Imager 2 microscope (Carl Zeiss AG, Oberkochen, Germany) was used to plot CTB and BDA labels, as well as to image Nissl- and AChE-stained sections. For each case, all sections within the extent of the thalamus were examined and plotted. Labeled neurons and axon terminals were marked and outlined at a lower magnification (16x for Leitz microscope, 10x for Zeiss microscope) and verified at a higher magnification (40x for Leitz microscope, 20x for Zeiss microscope). Section plots and architectural images were then imported into Adobe Illustrator CC2014 (Adobe Inc., San Jose, CA) for alignment. Locations of labeled cells and terminals were related to thalamic structures revealed in adjacent sections stained for architecture by matching blood vessels and common anatomical features. Images were adjusted only for contrast and brightness with Adobe Photoshop CC2014 (Adobe Inc., San Jose, CA) but otherwise not altered. Adjacent sections from the immunohistochemically processed series (for CTB and BDA label visualization) as well as the unprocessed series (for FR, FE, and FB label visualization) were aligned together for better appreciation of joint distributions of different labels. Parcellation of the thalamus and determination of borders of thalamic nuclei followed previous studies on galagos (Wong et al., 2009; Baldwin et al., 2013; Li et al., 2013), while referencing studies on other primate species as well (Lanciego and Vázquez, 2012; Paxinos et al., 2012). In transitional regions where borders of nuclei were uncertain and hard to demarcate, slashes were used for designations (e.g., VA/Vla, LP/LPul). Subnuclei or subdivisions of major thalamic structures were not necessarily marked unless they were confidently distinguished in selected sections (e.g., IL instead of CL and PC; Pul instead of APul, MPul, LPul, and IPul).

Results

Thalamic connections of different movement domains identified in the rostral portion of PPC of galagos have been published previously (Wang et al., 2023). The present study continues to look into patterns of thalamic connections of PPC in galagos, with a focus on its caudal part that is unresponsive to long-train intracortical microstimulation (LT-ICMS). Nine injections of tracers were placed in the dorsal or ventral division of PPCc, namely the surrounding PPC regions above and below the caudal half of the intraparietal sulcus (IPS), of six galagos. All galagos received injections in dorsal PPCc, whereas four of these galagos were also injected in ventral PPCc. Results from all cases collectively revealed that major thalamic connections of PPCc were with the lateral posterior nucleus (LP) and pulvinar (Pul) in general. Five studied cases with multiple tracers injected in both PPCr and PPCc are particularly informative, as they allow the comprehensive comparison of thalamocortical connection patterns between PPCr and PPCc regions. In these cases, label profiles resulting from individual PPCc injections are presented first (Figure 2), followed by distributions of labeled neurons produced by different tracers placed across PPC. For example, the consecutive Figures 3 and 4 both illustrate the left thalamus of case 04–07, with Figure 3 focusing on the label distribution after the PPCc injection, whereas Figure 4 compares distribution patterns of labeled neurons that resulted from injections in different PPC locations of the same hemisphere.

Figure 2.

Photomicrographs of thalamic neurons labeled by biotinylated dextran amines (BDA) (a) and cholera toxin subunit B (CTB) (b), and corresponding injections sites (c, d) in PPCc, captured with a light microscope at 20x and 5x magnification, respectively. (a): BDA-labeled neurons and axon terminals in pulvinar (Pul) in case 04–47. (b): CTB-labeled neurons in Pul in case 04–07. Arrowheads mark labeled neurons and dashed line circles labeled terminals. BDA injection site (c) and CTB injection site (d) in case 04–47 are circled by dashed line. Scale bar in panels (a) and (b) = 100 μm, and in panels (c) and (d) = 500 μm.

Figure 3.

Distribution of labeled neurons in the left thalamus (LT) after a CTB injection in the dorsal division of PPCc of the left hemisphere in case 04–07, shown in a series of coronal sections arranged from rostral (17) to caudal (72). Each dot represents one labeled neuron. Dashed lines drawn from adjacent sections for architectonic staining mark borders of thalamic nuclei. Notable fiber bundles are shaded in gray. Labeled neurons are mostly found in VL, LP and pulvinar nuclei. R, rostral; C, caudal; D, dorsal; M, medial. Top left: drawing of PPC and adjacent regions with the injection core reconstructed from flattened cortical sections (Stepniewska, Cerkevich et al., 2009). Dashed line marks the border between PPCr and PPCc established by electrical stimulation. Opened sulci with buried cortex are shaded in gray. Top right: a favorable section stained for acetylcholinesterase (AChE), with approximate borders of thalamic nuclei marked with white dashed lines. For abbreviations see list.

Figure 4.

Comparison of distributions of thalamic neurons labeled by fluororuby (FR), fluoroemerald (FE), and BDA injected in rostral PPC (PPCr), and CTB injected in PPCc in case 04–07, shown in a series of coronal sections arranged from rostral (21/22) to caudal (61/62). Sections with fluorescent tracers (FR and FE) are aligned to adjacent sections from the BDA/CTB series, with both section numbers tagged. CTB-labeled neurons are distributed more dorsally than FR- and FE-labeled neurons in the thalamus. Top right: drawing of PPC and adjacent areas with all injection cores reconstructed from flattened cortical sections. Bottom left: schematic 3-dementional diagram illustrating a thalamus cut in coronal plane. Adapted from the graph generated by a toolbox (Treu et al., 2020) with the dataset (Ilinsky et al., 2018). Major thalamic nuclei are color-coded. Other conventions as in Figure 3.

Examples of labeled neurons and axon terminals as well as injection sites are shown in Figure 2. Serial drawings of selected sections are shown in horizontal, coronal, or sagittal views. The use of different sectioning planes allowed a more intuitive appreciation of the spatial distribution of labeled neurons, but also made comparisons across cases more difficult.

Case 04–07

A relatively large CTB injection was placed just caudal to the electrophysiologically defined border between PPCr and PPCc, presumably in the region corresponding to the caudal portion of area 7dm of Preuss and Goldman-Rakic (1991). The injection core was restricted to dorsal PPCc and did not involve the cortex buried in IPS. Neurons were mostly labeled in dorsal thalamic nuclei, as shown in a series of coronal sections (Figure 3). At the anterior level, patches of labeled neurons were primarily identified in the ventral anterior (VA), the ventral lateral nuclei (VL), as well as the transitional zone between VL and the lateral posterior nucleus (LP). These neurons were clustered in two major groups, a more lateral region bordering immediately the reticular nucleus (Rt), and a more medial region adjacent to the central lateral nucleus (CL) of intralaminar nuclei (IL). Proceeding to the more posterior level, labeled neurons were concentrated in the anterior and medial subdivisions of the pulvinar complex (APul, MPul), while scattering in the lateral and inferior subdivisions of pulvinar (LPul, IPul) as well. Except for IPul, borders among other pulvinar subdivisions in galagos were sometimes ill-defined, leaving open the possibility of mischaracterized labeled neurons in particularly APul and MPul in this case. Other neurons were labeled in the lateral dorsal (LD), medial dorsal (MD), posterior nuclei (Po), as well as CL and the paracentral nucleus (PC) of IL.

In addition to the dorsal PPCc injection, this case also received injections of two tracers (FR and FE) in the forelimb domain of dorsal PPCr, and one tracer (BDA) in the face domain of ventral PPCr (Stepniewska, Cerkevich et al., 2009). Thalamic origins of PPC connections exhibited a ventral-to-dorsal shift as injection sites progressed from rostral to caudal PPC (Figure 4). This was best demonstrated with the three dorsal PPC injections, where distributions of CTB, FE, and FR labeled neurons in VL, LD, LP, and Pul followed a dorsoventral sequence. Such tendency was less evident at the posterior end of the thalamus (IPul, APul, and Po), where foci of neurons labeled by different tracers overlapped to some extent (sections not shown). Comparing regions within PPCr, the face domain of ventral PPCr received major projections from the posterior half of the thalamus, especially APul, whereas the forelimb domain of dorsal PPCr also had substantial connections with the anterior part of VL in addition to APul.

Case 04–04

FE was injected in two adjacent sites in dorsal PPCc. Both injection cores were slightly caudal in location compared to the CTB injection in case 04–07 (Figure 5). Labeled neurons appeared only at the top level of this series of horizontal sections, confirming the restricted connections of dorsal PPCc with the dorsal part of the thalamus. Major foci of labeled neurons were identified in LP and APul/MPul with a separation between the laterally and centrally located groups, similar to the observations in case 04–07 (Figure 3). A few cells were also labeled caudally in Pul, compared to the clusters of cells labeled in APul/MPul of rostral Pul. Some small foci and scattered neurons with label were distributed in VL, as well as in regions close to the medial portion of IL. In general, there was consistency of the thalamic connection patterns of dorsal PPCc between this and case 04–07.

Figure 5.

Distribution of labeled neurons in LT after two close FE injections in the dorsal division of PPCc of the left hemisphere (top left panel) in case 04–04, shown in a series of horizontal sections arranged from dorsal (12) to ventral (42). Clusters of labeled neurons are found in VL, LP and pulvinar nuclei except the inferior pulvinar (IPul). D, dorsal; V, ventral; A, anterior; M, medial. Other conventions as in Figure 3.

The overall injection arrangement of this case was comparable to that of case 04–07, except only one tracer was placed in dorsal PPCr (Figure 6). In the dorsal PPC region, FE injections in PPCc labeled neurons primarily in dorsal thalamic sections, while more ventral sections were almost void of neurons. In contrast, neurons labeled with FR that was injected in PPCr were rarely identified in the most dorsal sections, but were distributed heavily in ventral sections. Such observations were consistent with those of case 04–07. With regard to PPCr, injections that involved the forelimb and face domain of dorsal and ventral PPCr, respectively, both resulted in labels in LP, APul, VP, and VL. The majority of thalamic projections to ventral PPCr was posterior to that of dorsal PPCr, except scattered neurons in the more anterior part of the thalamus (VA) also projected to ventral PPCr.

Figure 6.

Comparison of distributions of thalamic neurons labeled by FR and BDA injected in PPCr, and FE injected in PPCc (top right panel) in case 04–04, shown in a series of horizontal sections arranged from dorsal (26/27) to ventral (51/52). FE-labeled neurons are distributed more dorsally and caudally than FR- and BDA-labeled neurons in the thalamus. Bottom left: schematic 3-dementional diagram illustrating a thalamus cut in horizontal plane. Other conventions as in Figure 4.

Case 04–39

The small BDA injection was located at the very caudal part of dorsal PPCc, with its core encroaching on the upper bank of the posterior tip of IPS (Figure 7). Densely labeled neurons were present in VL, Pul, and somewhat unexpectedly, in the lateral geniculate nucleus (LGN), as displayed in sagittal sections. Neurons labeled in Pul were concentrated in the anterior portion, while almost void in the posterior portion, particularly in IPul. Moreover, VA, LP, LD, MD, zona incerta (ZI), and limitans nucleus (Li) contained a few labeled neurons. Anterograde labeling of this case was also available, as BDA worked well as a bidirectional tracer. Two foci of dense-to-moderately labeled terminals were evident, one at the border between VL and Rt, and one close to the anterior border of Pul. Labeled terminals and neurons mostly overlapped, except in the pretectum (PT), which contained only labeled terminals but not neurons.

Figure 7.

Distribution of labeled neurons and axon terminals in LT after a BDA injection in the dorsal division of PPCc of the left hemisphere (top left panel) in case 04–39, shown in a series of sagittal sections arranged from lateral (14) to medial (80). Spots of red and orange indicate dense and moderate labeling of axon terminals, respectively. Labeled neurons and axon terminals are also concentrated in VL, LP and pulvinar nuclei except IPul. Note that some neurons are labeled in LGN in this case. L, lateral; M, medial; D, dorsal; P, posterior. Other conventions as in Figure 3.

A comprehensive arrangement of injections, with four different tracers injected into each quarter of PPC (dorsal PPCr, ventral PPCr, dorsal PPCc, and ventral PPCc), was implemented in this case. All injections involved more or less the bank of IPS (Figure 8). Series sectioned in the sagittal plane provided another view of label distributions in the thalamus. PPC, as a whole, received most of the thalamic projections from VL and Pul, and scattered projections from VA, LD, MD, Li, and centre median nucleus (CM). In dorsal PPC, thalamic projections to PPCc (BDA labels) were generally dorsal to projections to PPCr (FR labels), similar to the overall pattern shared by cases 04–07 and 04–04 (Figures 4 and 6). Similar topographic organization of thalamic connections was also observed in ventral PPC of this case, and though less complete, in case 04–47 as well (see below Figure 10 for case 04–47).

Figure 8.

Comparison of distributions of thalamic neurons labeled by FR and FE injected in PPCr, and BDA and fast blue (FB) injected in PPCc in case 04–39, shown in a series of sagittal sections arranged from lateral (31/32) to medial (79/80). BDA- and FB-labeled neurons are distributed more dorsally than FR- and FE-labeled neurons in the thalamus. Bottom left: schematic 3-dementional diagram illustrating a thalamus cut in sagittal plane. Other conventions as in Figure 4.

Figure 10.

Comparison of distributions of thalamic neurons labeled by FE injected in PPCr, and FR and BDA injected in PPCc in case 04–47, shown in a series of coronal sections arranged from rostral (51/52) to caudal (105/106). Other conventions as in Figure 4.

Case 04–47

The core of the BDA injection was in the middle of the ventral division of PPCc, putatively corresponding to area 7a-m of Preuss and Goldman-Rakic (1991) (Figure 9). The majority of labeled neurons were in Pul, spanning across APul, LPul and MPul, but almost void in IPul. LP was another major source of projections to ventral PPCc. Remaining labeled neurons were sporadically observed in VL, CL, PC, Li, PT, and LGN. Anterograde labeling was also prominent in this case. Dense foci of labeled terminals were distributed in two separate territories at more rostral levels, with a lateral patch in LP involving LPul, and a medial patch at the edge of CL and VL. Foci of terminals shifted mainly to APul and the lateral part of MPul as progressing caudally, with the remainder in the lateral part of LPul. Density of labeled terminals appeared to be lower at the rostral than the caudal level, where sparse to moderate terminals were detected in IPul and Li. Moreover, the superior colliculus (SC) contained some dense projections.

Figure 9.

Distribution of labeled neurons and axon terminals in LT after a BDA injections in the ventral division of PPCc of the left hemisphere (top left panel) in case 04–47, shown in a series of coronal sections arranged from rostral (27) to caudal (117). Spots of yellow indicate sparse labels of axon terminals. Dense labeled terminals are present in VL, LP, CL, and pulvinar nuclei except IPul. Labeled neurons overlap labeled terminals mostly in LP and pulvinar. For abbreviations see list. Other conventions as in Figures 3 and 7.

Besides the BDA injection, a FR injection in dorsal PPCc and a FE injection in ventral PPCr were placed in this case as well (Figure 10). Distributions of FR- and BDA-labeled neurons overlapped somewhat in the middle portion of the thalamus, though both injections labeled relatively small amounts of neurons. In the ventral division of PPC, fair comparison of connection patterns between PPCr and PPCc was difficult to make, as the BDA injection in PPCc labeled much fewer neurons than the FE injection in PPCr did. Several caudal sections implied that thalamic projections to PPCc originated more dorsally than those to PPCr in the caudal pulvinar.

Cases 05–15 and 05–20

The two cases were both injected with BDA in the caudal part of ventral PPCc. Since in these cases whole brain hemispheres were cut coronally without separation and flattening of cortices, instead of illustrating injection cores on the flattened cortex, only the injection sites are shown on the brain surface for the sake of accuracy. In case 05–15, two BDA injections were placed very close to the caudal tip of IPS, resulting in substantial distributions of labeled neurons and axon terminals in the posterior thalamus, primarily throughout LPul, MPul, and IPul (Figure 11a). Densely labeled terminals overlapped labeled neurons in the ventromedial portion of LPul and MPul, as well as the medial portion of IPul, while labeled neurons were more widespread throughout the entire Pul. Several cells were labeled in VL, LP, Li, and PT, and a patch of terminals was spotted in CL. Labeled neurons and terminals were also identified in LGN. Case 05–15 were also injected with FB in dorsal PPCr, and FR in dorsal PPCc (Figure 11b). Topographically, FR-labeled neurons were distributed rostral and dorsal to FB-labeled neurons in the thalamus, conforming to thalamic connection patterns of the dorsal division of PPC in cases 04–07, 04–04 and 04–39 (Figures 4, 6 and 8). Within PPCc region, a weak topography that thalamocortical projections to dorsal PPCc (FR labels) originated more dorsally than those to ventral PPCc (BDA labels) in Pul appeared to exist.

Figure 11.

Distributions of labeled neurons and axon terminals in LT after two BDA injections in the ventral division of PPCc of the left hemisphere (a), and comparison of distributions of thalamic neurons labeled by FB injected in PPCr, and FR and BDA injected in PPCc (b) in case 05–15. Coronal sections are arranged from rostral (102) to caudal (42). Top right: photomicrographs of the PPCc region around injection sites. White strokes mark IPS; Magenta, blue and orange symbols mark injection sites. Note that labeled neurons and terminals are identified in LGN. Other conventions as in Figures 3 and 7.

In case 05–20 (not illustrated), BDA and FR injections were somewhat rostral compared to those of case 05–15, while their overall distributions of labeled neurons and terminals were similar. Dense patches of BDA terminals were found in LP and LPul/MPul, though limited, comparable to the pattern of terminal distributions of case 04–47 (Figure 9). BDA-labeled neurons were few. Both tracers labeled neurons in LP, VL, MPul, LPul, and Li, with FR labeling an additional patch of neurons in CL.

Discussion

The current report is a continuation of our previous study on the thalamic connections of the rostral portion of PPC (PPCr) in galagos (Wang et al., 2023). Here we focused on the thalamic connections of the caudal part of galago PPC (PPCc), and compared them with thalamic connections of PPCr. It was advantageous and informative to conduct connection studies on galagos, as they have relatively small and less complex brains compared to anthropoid primates, and yet they share basic features of brain structures. A summary of these connections is illustrated in Figure 12. We have the following main findings: first, PPCc receives major thalamic projections from the posterior part of thalamus, especially visual-related LP and the lateral pulvinar complex, and receives less projections from VL of the motor thalamus than does PPCr; second, the main targets of PPCc projections overlap the origins of PPCc projections in VL, LP, and Pul, demonstrating reciprocal connections between PPCc and these nuclei; third, a coarse topography of connections between PPC and thalamus appears to exist such that PPCr connects to more ventral parts of these thalamic nuclei, and PPCc to more dorsal parts. Below we compare the PPC connection patterns in galagos with those of other primates, and also relate the thalamic connections of PPC to its corresponding cortical connections, both in an effort to better understand the overall connectivity of PPC in primates.

Figure 12.

A summary of thalamic connections of the entire PPC of galagos, divided into PPCr and PPCc by the responsiveness to long-train intracortical microstimulation (LT-ICMS). PPCr contains small zones, termed “domains”, where specific types of complex movements resembling natural behaviors are evoked by LT-ICMS. PPCc fails to respond to LT-ICMS. Different functional domains have distinct pattern of thalamic connections. Connections of four domains (i.e., climbing, reaching, hand-to-mouth, grimacing; bottom box) revealed by tract tracing methods have been reported previously in Wang et al., 2023. Connection patterns of dorsal and ventral divisions of PPCc are somewhat different, but overall similar. Thalamic nuclei associated with motor, somatosensory and visual functions are arranged roughly from left to right. The thickness of lines represent the strength of the connectivity. Note that connections between LGN and PPCc (asterisks) were not consistently identified across cases. Dense connections of PPCc are with LP and LPul, and dense connections of PPCr are with VL and APul. For abbreviations see list.

It is worth pointing out that there are obstacles in making direct comparisons between thalamic connection patterns of PPC, especially PPCc, in galagos and that in other New World and Old World monkeys. The action domain map, as well as the PPCr/PPCc border have not been fully established by electrical stimulation in monkeys as they have been in galagos. In addition, the counterclockwise tilted IPS of monkeys compared to the horizontal IPS of galagos may reflect uneven cortical expansions within and around PPC regions, which may cause some PPC areas to be displaced. Thus, the equivalents to galago PPCc in monkeys are difficult to determine.

Thalamic connections of PPCc in primates

Early studies have only occasionally revealed, but not systematically investigated, thalamic connections of PPCc in galagos and New World monkeys. Therefore, there are only a few reports on such connections. In galagos, tracer injections at the border between ventral PPCc and the middle temporal area labeled neurons mainly in the inferior pulvinar, with some groups of neurons in the superior pulvinar (Raczkowski and Diamond, 1980). Note that the abovementioned superior pulvinar is deemed to correspond to MPul and LPul of the current paper (also see Glendenning et al., 1975). In another study, the injection involving the caudal edge of PPCc labeled neurons primarily in LPul and MPul (Beck and Kaas, 1998b). These observations, though concerning solely the pulvinar, are consistent with our results, confirming the essential connections between PPCc and the pulvinar complex. Similarly in marmosets, injections placed in the border zone between area 7 and area 19, putatively the posterior end of PPCc, primarily labeled neurons in LP, LPul and MPul, with some neurons labeled in IL, VA, and LD (Brysch et al., 1990). In owl monkeys, injections located at the rostral edge of the dorsomedial area (DM) that encroached PPCc produced labeled neurons mainly in IPul, LPul, and the ventral posterior nucleus (VP) (Beck and Kaas, 1998b).

Thalamic connections of PPCc in Old World macaque monkeys have been studied more extensively. PPC of macaques consists of the superior parietal lobule (SPL) and the inferior parietal lobule (IPL). In SPL, areas PEc, PGm, and V6A have been usually considered as caudal PPC (Gamberini, Passarelli, Fattori et al., 2020). PEc on the lateral surface and PGm on the medial aspect both receive major projections from LP, APul, MPul, and LPul. Additionally, PEc has strong connections with VL and ventral posterolateral nucleus (VPL), whereas PGm has moderate connections with VL, MD, and IL (Yeterian and Pandya, 1985; Schmahmann and Pandya, 1990; Cappe et al., 2007; Impieri et al., 2018; Gamberini, Passarelli, Impieri et al., 2020; Gamberini et al., 2021). The visuomotor area V6A, rostrally adjacent to the parieto-occipital sulcus, receives prominent projections from LP and MPul, and sparse projections from VL, MD, PC, and Li (Gamberini et al., 2016). Tracer injections in IPL that covered the caudal areas PG and Opt generally labeled a large portion of neurons and terminals in MPul, as well as in LP, LPul, MD, and IL (Mesulam et al., 1977; Kasdon and Jacobson, 1978; Weber and Yin, 1984; Asanuma et al., 1985; Yeterian and Pandya, 1985; Schmahmann and Pandya, 1990; Baleydier and Morel, 1992). The caudalmost part of IPL, Opt, had additional substantial connections with LD and anterior nuclei that were barely detected in either adjacent area V6A or PG (Schmahmann and Pandya, 1990). The posterior bank of IPS, likely corresponding to caudal parts of areas LIP, MIP, and VIP, was connected with LPul, MPul, VA, VL, LP, CL, and Li (Baizer et al., 1993). Taken together, the caudal sector of PPC of macaques, though not exactly defined, had predominant connections with Pul and LP, and other connections with VL, VPL, VA, IL, MD, and Li. These results suggest that thalamic connections of PPCc are somewhat similar in galagos and monkeys in general, both having major connections with LP and Pul. Nevertheless, more comparative studies are needed for a thorough understanding of similarities and differences of PPCc connections among different primate species.

Our injections in PPCc of galagos labeled some neurons in LGN (cases 04–39 and 05–15), and these connections have not been reported elsewhere in galagos or other primates. Although the possibility that labeled neurons in LGN resulting from injected tracer uptake by fibers of passage terminating in the striate cortex (V1) and traveling through PPCc cannot be entirely ruled out, these connections extend previous evidence for widespread projections of some LGN neurons to extrastriate visual areas such as V4, MT, and DM (Lysakowski et al.,1988; Beck and Kaas, 1998b; Sincich et al., 2004). It will be important to look for further evidence of connections between PPC and LGN in future studies in all kinds of primates.

Topographic patterns of PPC thalamic connections in galagos and macaques

Results from our previous (Wang et al., 2023) and current studies suggest certain topographical arrangements of thalamocortical connections of PPC in galagos. A major topographic pattern in the rostro-caudal direction of PPC is such that PPCr has stronger connections with the anterior half of the thalamus, particularly VL and VA of the motor thalamus, as well as somatosensory-related APul, whereas PPCc connects heavily with the posterior half of the thalamus, namely LP and Pul, which are largely involved in visual processing (Asanuma et al., 1983; Cusick and Gould, 1990; Grieve et al., 2000; Kaas and Baldwin, 2019; Krubitzer and Kaas, 1992; Sommer, 2003). Such rostro-caudal arrangement of thalamocortical connections appears to be common in both dorsal and ventral divisions of PPC (Figures 6 and 10). Less evidently, with gradual progression from PPCr to PPCc, distributions of the thalamic projecting neurons tend to shift from ventral to dorsal (Figures 4, 6,8 and 11). Within PPCr where somatotopically organized movement-specific domains have been identified by electrical stimulation, thalamic connections to these domains have different preferences, with VL and VA predominating the inputs to domains dorsal to IPS (i.e., climbing, hand-to-mouth and reaching domains), while APul, LP and the posterior group of nuclei predominating that of the grimacing domain ventral to IPS (Wang et al., 2023; also see Figures 4 and 6). The connection patterns of dorsal and ventral regions within PPCc, on the other hand, are less distinctive, suggesting that PPCc of galagos is more likely to function as a single area rather than separate domains as does PPCr.

Similar topographical patterns of thalamic connections of PPC appear to be preserved to some extent in macaque monkeys, although the designated PPC region expands significantly in anthropoids compared to that of prosimians. Multiple tracer injections that systematically covered the entire PPC region in macaques have provided evidence of the rostral-to-caudal progression of cortical connectivity paralleled by the rostral-to-caudal flow of thalamic connectivity (Yeterian and Pandya, 1985; Schmahmann and Pandya, 1990). As retrograde tracers were injected more caudally in PPC, loci of labeled neurons tended to occupy farther caudal portions of the thalamus, shifting distributions from VP, VL, LP, and APul in rostralmost injections, towards LP, APul, LPul, and MPul in caudalmost injections (Schmahmann and Pandya, 1990). A comparable rostral-to-caudal shift of labeled terminals was also observed in rostrocaudally placed serial injections of anterograde tracers in PPC (Weber and Yin, 1984; Yeterian and Pandya, 1985). The overall topographical arrangements were shared in both SPL and IPL. In addition to the rostro-caudal topography of the thalamic connections, a ventrodorsal topography occurs locally in regions of LP, APul and MPul, with more rostral levels of PPC receiving thalamic projections from relatively ventral positions and caudal levels of PPC receiving projections from dorsal positions (Schmahmann and Pandya, 1990). The consistent presence of the rostro-caudal and ventro-dorsal topographies of thalamocortical connections in both galagos and macaques suggests that such connection patterns of PPC are likely concerned with the basic functions PPC serves that have been preserved over the course of primate evolution.

Putative functions of PPCc with consideration of its thalamic and cortical connectivity

Previous study reveals that PPCc of galagos receives major cortical projections from the second visual (V2), third visual (V3), dorsal lateral (DL), dorsal medial visual areas (DM), the cingulate (CgC), and caudal inferotemporal cortex (ITc), and additional projections from the frontal region housing the frontal eye field (FEF) (Stepniewska et al., 2016). These are all notable as visual-related areas involved in a wide range of visual processing including object perception and motion detection (Allman et al., 1973; Rosa et al., 1997; Lyon and Kaas, 2002; Stepniewska et al., 2018). This visual-dominant pattern of cortical connections of PPCc echoes that of thalamic connections, a large proportion of which originates from LP and LPul that are critically components in the visual pathways (Jones, 2012). Thus, anatomical evidence from thalamic and cortical levels converge to lead to the implication that PPCc is engaged in further integrations of processed visual information of various forms and sources. PPCc also has substantial connections with PPCr, which is organized in a fashion of movement-specific domains, particularly with the caudally located reaching and defensive domains right next to the PPCr/PPCc border (Stepniewska, Cerkevich, et al., 2009; Wang et al., 2021). Such connections suggest that PPCc provides highly integrated visual information to PPCr domains to facilitate visually guided behaviors. Patterns of cortical projections to dorsal and ventral PPCc vary, with stronger projections to dorsal PPCc from PPCr, V2, V3, and CgC, whereas stronger projections to ventral PPCc are from PPCr, DL, and ITc (Stepniewska et al., 2016). Distinctions of the thalamic connections between dorsal and ventral PPCc, on the other hand, are less significant, suggesting information conveyed to PPCc remains largely unsorted at the level of thalamus. Taken together, PPCc is likely to function as a central hub for higher-order integration and processing of visual information, and for further distribution of such information accordingly to assist in the formulation of specific behaviors.

Key points.

1. PPCc of galagos receives major thalamic projections from the lateral posterior and lateral pulvinar nuclei, and additional projections from the ventral anterior, ventral lateral, intralaminar, as well as anterior, medial, and inferior pulvinar nuclei.

2. The thalamic connections of PPCc are dominated by visual-related thalamic nuclei, whereas connections of PPCr are dominated by motor- and somatosensory-related nuclei.

3. A coarse topography of thalamocortical connections exists such that PPCr connected to more ventral parts of the thalamic nuclei, and PPCc to more dorsal parts.

Abbreviations

Cortical and thalamic structures

1--2

somatosensory areas 1 and 2

AM

anterior medial nucleus

AnT

anterior nuclei

APul

anterior pulvinar

AV

anterior ventral nucleus

CL

central lateral nucleus

CM

centre median nucleus

fr

fasciculus retroflexus

Hb

habenula

IL

intralaminar nuclei

IPS

intraparietal sulcus

IPul

inferior pulvinar

LD

lateral dorsal nucleus

LGN

lateral geniculate nucleus

Li

limitans nucleus

LP

lateral posterior nucleus

LPul

lateral pulvinar

MD

medial dorsal nucleus

MGN

medial geniculate nucleus

MPul

medial pulvinar

mtt

mammillothalamic tract

PC

paracentral nucleus

PF

parafascicular nucleus

Po

posterior nuclei

PPC

posterior parietal cortex

PT

pretectum

Pul

pulvinar

Rt

reticular nucleus

SC

superior colliculus

VA

ventral anterior nucleus

VL

ventral lateral nucleus

VLa

ventral lateral nucleus anterior subdivision

VLp

ventral lateral nucleus posterior subdivision

VM

ventral medial nucleus

VP

ventral posterior nucleus

VPL

ventral posterolateral nucleus

VPM

ventral posteromedial nucleus

Anatomical tracers and molecular markers

BDA

biotinylated dextran amines

CTB

cholera toxin subunit B

FB

fast blue

FE

fluoroemerald

FR

fluororuby

AChE

acetylcholinesterase

Data availability

The data supporting this study’s findings are available from the corresponding author upon reasonable request.