Intrinsic functional clustering of anterior cingulate cortex in the common marmoset
Introduction
The common marmoset (Callithrix jacchus) is a New World primate that has garnered recent attention as a powerful complement to canonical Old World primate (e.g., macaques) and rodent models (e.g., rats, mice) for preclinical modelling of the human brain in healthy and diseased states. With a granular frontal cortex (Reser et al., 2017) and the advent of transgenic modification techniques (Park et al., 2016; Sasaki et al., 2009; Tomioka et al., 2017), marmosets are well positioned to serve as neuropsychiatric models of prefrontal cortex dysfunction (Okano et al., 2016). A critical step in the development of marmosets for such models is to characterize functional network topologies of frontal cortex in healthy, normally functioning marmosets – i.e., how these circuitries are functionally divided, and how those topologies compare to human circuitry. With dense connections from motor, limbic, and prefrontal regions (Barbas and Pandya, 1989), the primate anterior cingulate cortex (ACC) is a critical hub for integrating neuronal processes that result in intelligent behavior (e.g., task switching; Barbas and Pandya, 1989; Brown and Braver, 2005; Johnston et al., 2007). Although marmoset ACC shows architectonic and connectional features homologous to humans and macaques (Burman and Rosa, 2009; Paxinos et al., 2012; Reser et al., 2017; Roberts et al., 2007), the understanding of the extent of functional diversity of the ACC of marmosets is still in the nascent stage of investigation. Here, we sought to characterize the intrinsic functional organization of ACC using resting state functional magnetic resonance imaging (RS-fMRI) in lightly anesthetized marmosets.
Like in Old World primates, marmoset ACC is cytoarchitecturally subdivided into dorsal (Brodmann's area (BA) 24) and ventral (BA25 and BA32) components (Brodmann, 1909; Paus, 2001; Vogt et al., 1995) – the distinction between disgranular (i.e., BA32) and agranular (i.e., BA24) cortex is relatively clear in marmosets, demarcating the boundary between medial prefrontal and anterior cingulate areas (Reser et al., 2017). Importantly, rodents do not share this same homology in cytoarchitecture, making comparisons of ACC connectivity with primates difficult (Fillinger et al., 2018; Ongur, 2000; Vogt, 2009). As such, marmosets may be a more appropriate comparator for human ACC connectivity than rodents. With a recent push toward mapping a comprehensive atlas of cortical connections in marmosets (Majka et al., 2016), there is mounting evidence that marmosets share the same general patterns of structural connectivity with Old World primates, with the ventral regions of ACC more densely connected to limbic structures, whereas the dorsal regions have a higher density of connections to motor and prefrontal regions (Barbas and Pandya, 1989; Majka et al., 2016; Morecraft and van Hoesen, 1992).
Functional MRI allows for the substantiation (or challenge) of structural delineations within ACC, as it has the distinct advantage of being agnostic to cytoarchitectonic features. RS-fMRI has allowed for identification of functional subdivisions of ACC in both humans (Margulies et al., 2007) and macaques (Hutchison and Everling, 2012). Margulies et al. (2007) demonstrated the functional heterogeneity of ACC in humans, with a clear caudal division that was highly connected with sensorimotor regions and, in contrast, a rostral division that was highly connected to prefrontal regions. Using a similar approach, Hutchison et al. (2012) demonstrated homologous divisions in the macaque, with four separate functional subfields identified across the rostral-caudal axis which were differentially connected to motor, prefrontal, and limbic networks. Corroborating evidence for connectivity-based parcellations has also come from other fMRI studies (Johansen-Berg et al., 2004; Palomero-Gallagher et al., 2015; Yu et al., 2011), diffusion tensor imaging (Beckmann et al., 2009), positron emission tomography (PET; Paus et al., 1995), and lesion studies (Hashimoto and Tanaka, 1998) in humans, as well as fMRI and electrophysiological studies in macaques (Loh et al., 2018; Procyk et al., 2016).
In marmosets, documentation of the functional organization of the ACC with functional neuroimaging is less extensive, albeit RS-fMRI has shown that ACC (e.g., 24a) is a major connectivity hub (Belcher et al., 2016) and lesion data has implicated ACC in contingency learning (Jackson et al., 2016). Because RS-fMRI is task-independent, it is well suited for identifying similarities or differences in integrated networks across species (e.g., between New and Old World primates), although caution should always be taken when interpreting these comparisons. Given that the functional organization of the marmoset brain is relatively unknown, we sought to apply a hypothesis-free clustering approach previously applied in macaques (Hutchison et al., 2012; Hutchison and Everling, 2014) to gain insight into the functional organization of the marmoset ACC. Data-driven clustering (e.g., hierarchical clustering) has the distinct advantage of being free of preconceived notions of ACC functional organization. As such, the optimal number of discrete clusters is determined by the data, rather than a priori assumptions that may not necessarily correspond to functional delineations (e.g., cytoarchitectural delineations). These regions can then be vetted in a second-level assessment, where the functional connectivity of these clusters with the rest of the brain is determined – if these clusters show differential connectivity across the brain, then they indeed likely correspond to differential functional subregions of the ACC.
Although RS-fMRI based clustering is of utility for quantifying the intrinsic functional subdivisions of ACC at large, regionally-specific connectivity may be lost amid broad intrinsic clusters. For example, the frontal eye fields (FEF) and primary motor cortex (M1) have been shown to be differentially connected across subregions of ACC (Amiez and Petrides, 2014; Procyk et al., 2016; Wang et al., 2004). As such, explicit hypothesis-driven seed analyses may also be valuable for delineating region-specific connectivity with ACC. Here, we leveraged these analyses in order to differentiate M1 and FEF connectivity with ACC, as previously shown by a tracing study in macaques (Wang et al., 2004). Additionally, we sought to differentiate cingulate connectivity between areas 8aD and 8aV, two putative homologs of FEF in marmoset (Blum et al., 1982; Burman et al., 2006; Ghahremani et al., 2017; Reser et al., 2013).
Section snippets
Image acquisition
Data were collected from seven adult marmosets (Callithrix jacchus) aged 1–6 years weighing from 300 to 500 g. Prior to each imaging session, anesthesia was induced with ketamine at 20 mg/kg. During scanning, marmosets were lightly anesthetized with isoflurane and maintained at a level of 1.5% throughout the scan by means of inhalation. Oxygen flow rate was kept between 1.75 and 2.25 l/min throughout the scan. Respiration, SpO2, and heart rate were continuously monitored via pulse oximeter and
Hierarchical clustering
Hierarchical clustering was run on the functional connectivity values between all of the voxels within the ACC (areas 24, 35, and 32). As shown in Fig. 1, the clustering analysis yielded solutions for 2 through 7 clusters based on the functional connectivity data. To determine the optimal number of clusters, we conducted a silhouette analysis which demonstrated that two or three clusters were optimal, given the input data (Fig. 1b). As such, we chose three clusters (in part because whole-brain
Discussion
Marmosets have garnered recent attention as a preclinical model for human brain diseases – with several major neuropsychiatric disorders characterized by ACC dysfunction proposed to be modeled in marmosets, understanding the functional organization of ACC in marmosets as it relates to Old World primates will be critical in the development of models of such intractable diseases. In this study, we applied a data-driven hierarchical clustering approach to marmoset RS-fMRI data to functionally
Author notes
We wish to thank Drs. Cirong Liu and Afonso Silva of the National Institutes of Health for providing the surface models of the marmoset brain. We also wish to thank Nicole Hague, Miranda Bellyou, and Lauren Schaeffer for animal preparation and care and Alex Li for scanning assistance. Support was provided by the Canadian Institutes of Health Research (FRN 148365) and the Canada First Research Excellence Fund to BrainsCAN.
References (57)
- et al.
A reproducible evaluation of ANTs similarity metric performance in brain image registration
Neuroimage
(2011) AFNI: software for analysis and visualization of functional magnetic resonance neuroimages
Comput. Biomed. Res.
(1996)- et al.
Functional MRI of visual responses in the awake, behaving marmoset
Neuroimage
(2015) - et al.
Broad intrinsic functional connectivity boundaries of the macaque prefrontal cortex
Neuroimage
(2014) - et al.
Top-down control-signal dynamics in anterior cingulate and prefrontal cortex neurons following task switching
Neuron
(2007) - et al.
The first step for neuroimaging data analysis: DICOM to NIfTI conversion
J. Neurosci. Methods
(2016) - et al.
A digital 3D atlas of the marmoset brain based on multi-modal MRI
Neuroimage
(2018) - et al.
Mapping the functional connectivity of anterior cingulate cortex
Neuroimage
(2007) - et al.
Marmosets: a neuroscientific model of human social behavior
Neuron
(2016) - et al.
Brain-mapping projects using the common marmoset
Neurosci. Res.
(2015)
Brain/minds: a Japanese national brain project for marmoset neuroscience
Neuron
Functional organization of human subgenual cortical areas: relationship between architectonical segregation and connectional heterogeneity
Neuroimage
Advances in functional and structural MR image analysis and implementation as FSL
Neuroimage
Midcingulate cortex: structure, connections, homologies, functions and diseases
J. Chem. Neuroanat.
Functional segregation of the human cingulate cortex is confirmed by functional connectivity based neuroanatomical parcellation
Neuroimage
Neuroimaging evidence of the anatomo-functional organization of the human cingulate motor areas
Cerebr. Cortex
Architecture and intrinsic connections of the prefrontal cortex in the rhesus monkey
J. Comp. Neurol.
Connectivity-based parcellation of human cingulate cortex and its relation to functional specialization
J. Neurosci.
Functional connectivity hubs and networks in the awake marmoset brain
Front. Integr. Neurosci.
Eye movements induced by electrical stimulation of the frontal eye fields of marmosets and squirrel monkeys
Brain Behav. Evol.
Learned predictions of error likelihood in the anterior cingulate cortex
Science
Cytoarchitectonic subdivisions of the dorsolateral frontal cortex of the marmoset monkey (Callithrix jacchus), and their projections to dorsal visual areas
J. Comp. Neurol.
Architectural subdivisions of medial and orbital frontal cortices in the marmoset monkey (Callithrix jacchus)
J. Comp. Neurol.
Microstimulation and architectonics of frontoparietal cortex in common marmosets (Callithrix Jaccus)
J. Comp. Neurol.
Efferents of anterior cingulate areas 24a and 24b and midcingulate areas 24aʹ and 24bʹ in the mouse
Brain Struct. Funct.
Common marmosets: a potential translational animal model of juvenile depression
Front. Psychiatr.
Frontoparietal functional connectivity in the common marmoset
Cerebr. Cortex
Cited by (21)
Noninvasive focal transgene delivery with viral neuronal tracers in the marmoset monkey
2024, Cell Reports MethodsThe marmoset as a model for investigating the neural basis of social cognition in health and disease
2022, Neuroscience and Biobehavioral ReviewsCitation Excerpt :However, while humans and non-human primates have similar prefrontal and whole-brain functional connectivity, rodents were different in this regard. This is also in line with Schaeffer et al. (2019), who reported that the intrinsic functional connectivity of the marmoset anterior cingulate cortex ACC; associated with several social-cognitive functions, and along with some other brain regions referred to as the “social brain” (Blakemore, 2008), closely resembles that of both humans and Old World primates. Similarly, (Roberts and Clarke, 2019) further highlighted that marmosets’ dlPFC, vlPFC (another prefrontal region associated with complex cognitive and social behaviours (Chaplin et al., 2013)), and mPFC have granular and agranular regions (Fig. 1).
Comparative connectomics of the primate social brain
2021, NeuroImageCitation Excerpt :These results suggest a close link between the serotonergic system, midline cortical functional connectivity, and social behavior. The marmoset PCC was identified as a core region of the DMN by resting-state fMRI (Liu et al., 2019) and the ACC as a critical region for emotion and motivation regulation (Alexander et al., 2019; Wallis et al., 2017), resembling its human homolog (Schaeffer et al., 2019). As for the neurobiology of human personality, scientists have been attempting to use population datasets such as the YA-HCP to relate scores based on the five-factor model of human personality (Costa Jr. and McCrae, 2008) to cortical structures (Lewis et al., 2018; Privado et al., 2017; Riccelli et al., 2017; Toschi and Passamonti, 2019) and the functional connectome (Cai et al., 2020; Dubois et al., 2018; Nostro et al., 2018; Passamonti et al., 2019; Toschi et al., 2018).
Intrinsic functional clustering of ventral premotor F5 in the macaque brain
2021, NeuroImageCitation Excerpt :The locations of the six clusters delineated in the current resting-state clustering study (Fig. S2A, colored regions) showed a great degree of overlap with the location of the seeds from a previous seed-based resting-state investigation (Sharma et al., 2019) that examined functional connectivity of seeds placed in the inferior arcuate and which locations were based upon anatomical and functional criteria (Fig. S2A, colored circles). Previous studies in primates have shown the promising application of using non-invasive resting-state fMRI data to map several organizational principles of different portions of the cortex (Goulas et al., 2017; Hutchison et al., 2012b; Hutchison and Everling, 2014; Schaeffer et al., 2019a, 2019b; Vijayakumar et al., 2018). In this study, we examined to what extent partitioning schemes of macaque inferior arcuate sulcus derived from data-driven hierarchical clustering analysis of resting-state fMRI data correspond to organizational schemes based upon functional/anatomical investigations.