Resting state cerebral networks in mouse lemur primates: from multilevel validation to comparison with humans

Abstract

1Measures of resting-state functional connectivity allow the description of neuronal 2 networks in humans and provide a window on brain function in normal and 3 pathological conditions. Animal models are critical to further address experimentally 4 the function of brain networks and their roles in pathologies. Here we describe for the 5 first time brain network organization in the mouse lemur (Microcebus murinus), a 6 small primate attracting increased attention as a model for neuroscience. Resting-7 state functional MR images were recorded at 11.7 Tesla. Forty-eight functional 8 regions were identified and used to identify networks using graph theory, dictionary 9 learning and seed-based analyses. Comparison of results issued from these three 10 complementary methods allowed the description of the most robust networks from 11 mouse lemurs. Large scale networks were then identified from resting-state 12 functional MR images of humans using the same method as for lemurs. Strong 13 homologies were outlined between cerebral networks in mouse lemurs and humans. 14 15 Keywords 16 Brain function, Cerebral networks, Functional MRI, Graph theory, Human, Microcebus 17 murinus, Mouse lemur, Primate, Resting state 18Blood-oxygen level dependent (BOLD) functional magnetic resonance imaging 20 (fMRI) is largely used to investigate brain function in response to specific tasks. In the 21 absence of explicit tasks (i.e. in resting state conditions) patterns of oscillations of the 22 fMRI signal are similar in functionally connected brain structures (Biswal et al., 1995). 23 The detection of the synchronicity of BOLD signal in various brain regions in resting 24 state conditions can thus be used to describe cerebral network organization. In 25 particular this allows the characterization of i. local regions in which highly 26 coordinated neuronal activity occurs and ii. large scale networks composed of 27 widespread functional regions connected together (Biswal et al., 1995; Power et al., 28 2014). 29 Studies of brain networks have contributed to many breakthroughs in the 30 understanding of brain function in normal as well as in pathological conditions such 31 as Alzheimer's or Parkinson's diseases (Buckner et al., 2005; Gao and Wu, 2016). 32 However, many questions remain concerning both the technique and interpretation of 33 resting state fMRI. For example, both the role of resting state networks in cerebral 34 function, and the biological mechanisms underlying their activity, are still partly 35 unknown. Also, how their modulations impact behaviour and cognition in pathological 36 conditions is still debated (Mohan et al., 2016). 37Using animal models is critical to further address these questions. Indeed, in 38 animals it is possible to artificially stimulate neuronal activity to characterize biological 39 mechanisms underlying network function (Gerits et al., 2012). Another interest of 40 studying neuronal networks in animals is to evaluate how evolution has driven 41 network architecture and to assess to wha...

Animal Functional Magnetic Resonance Imaging: Trends and Path Toward Standardization

“…Concerning highly connected regions in human, macaque and mouse lemur, the posterior cingulate cortex was found to be critical in these three species with its major functional hubs located in the DMN (Garin et al, 2019). Interestingly, these areas seem to be instead shifted anteriorly in rodents, in which the anterior cingulate and prefrontal areas exhibit robust hub-like properties (Liska et al, 2015;Gozzi and Schwarz, 2016;Garin et al, 2019). This finding is consistent with rodent species lacking an evolutionary homolog of the primate posterior cingulate cortex (Vogt and Paxinos, 2014).…”

“…One of them is the smallworld feature which maximizes the efficiency of information transferred within a network. This network property has been found in multiple species including humans (Bullmore and Sporns, 2009), NHP (Barttfeld et al, 2015;Garin et al, 2019), rodents (Mechling et al, 2014), and ferrets (Zhou et al, 2016). Moreover, graph-based approaches have clearly revealed a modular nature of human (Sporns and Betzel, 2016), and rodent (Liska et al, 2015) rsfMRI networks, along with evidence of strongly functionally interconnected polymodal areas, exhibiting hub-like properties (Buckner et al, 2009;Liska et al, 2015).…”

“…Moreover, graph-based approaches have clearly revealed a modular nature of human (Sporns and Betzel, 2016), and rodent (Liska et al, 2015) rsfMRI networks, along with evidence of strongly functionally interconnected polymodal areas, exhibiting hub-like properties (Buckner et al, 2009;Liska et al, 2015). Concerning highly connected regions in human, macaque and mouse lemur, the posterior cingulate cortex was found to be critical in these three species with its major functional hubs located in the DMN (Garin et al, 2019). Interestingly, these areas seem to be instead shifted anteriorly in rodents, in which the anterior cingulate and prefrontal areas exhibit robust hub-like properties (Liska et al, 2015;Gozzi and Schwarz, 2016;Garin et al, 2019).…”

“…Using ICA, Jonckers et al found that the extracted components, i.e., functional network regions, were more unilateral in mice compared to rats (Jonckers et al, 2011), however, this effect failed to be replicated in numerous follow-up studies in mice (e.g., Grandjean et al, 2014;Sforazzini et al, 2014). In mouse lemur primates and humans, the cortical large-scale networks repertoire presents important similarities but the regional organization into networks highlighted compositional and structural divergences (Garin et al, 2019). Strong interhemispheric functional connectivity (FC) between homotopic regions has been consistently observed in humans and primates suggesting a phylogenetically preserved mammalian characteristic (Hutchison and Everling, 2012).…”

Sammba-MRI: A Library for Processing SmAll-MaMmal BrAin MRI Data in Python

“…Sammba-MRI proposes two ready-to-use pipelines to perform spatial registrations to a population or standard reference template as well as inter-modalities registration between anatomical, functional, or perfusion images. These pipelines have been tested throughout the different stages of their development process on various datasets from mice, rats and mouse lemurs and used in several publications from our lab (Garin et al, 2018(Garin et al, , 2019Nadkarni et al, 2019). The two pipelines are called Coregistrator and TemplateRegistrator.…”