Evolutionarily conserved fMRI network dynamics in the mouse, macaque, and human brain

Daniel Gutierrez‐Barragan, Julian S.b. Ramirez, Stefano Panzeri, Ting Xu, Alessandro Gozzi

TL;DR

Researchers have discovered that the way the brain’s activity fluctuates over time—its "dynamic choreography"—is evolutionarily conserved across humans, macaques, and mice. By identifying shared "coactivation modes" (C-modes) that cycle at infraslow frequencies, the study reveals that mammalian brain dynamics follow a universal set of principles, regardless of brain size or complexity.

What Problem Does This Paper Address?

For years, neuroscientists have known that the "static" architecture of the brain—the maps of which areas tend to work together—is remarkably similar across mammalian species. We all have versions of the Default Mode Network (DMN), for instance. However, brain activity isn't static; it shifts and evolves every few seconds.

The big question was: Are these second-to-second dynamics also conserved by evolution? Or does a human brain "flicker" through states in a fundamentally different way than a mouse brain? Comparing these dynamics is difficult because of the massive differences in scale and the confounding effects of anesthesia often used in animal studies. This paper seeks to find the "species-invariant" principles that govern how mammalian brains transition between different functional states.

Methods

The authors employed a frame-wise clustering approach on fMRI time-series data. Instead of averaging activity over several minutes, they looked at individual "snapshots" (frames) of brain activity.

  1. CAPs to C-modes: They identified Coactivation Patterns (CAPs)—recurring spatial maps of activity. They found that these CAPs often come in "mirrored" pairs (one area is "on" while another is "off," and then they flip). They merged these pairs into Coactivation Modes (C-modes), representing the peaks and troughs of a single fluctuating rhythm.
  2. Cross-Species Matching: To compare humans, monkeys, and mice, they used the Hungarian Algorithm to find the best spatial matches between the C-modes of each species based on evolutionarily conserved brain landmarks.
  3. Dynamic Modeling: They used Markov processes to analyze the transitions between these modes and Rayleigh tests to see if these modes were "phase-locked" to the brain's global signal (the overall rise and fall of activity across the entire cortex).

Datasets

The study utilized high-quality, "awake" rsfMRI datasets to avoid the dampening effects of anesthesia:

  • Humans: Two independent cohorts—the Hangzhou Normal University (HNU) dataset (30 subjects) and the Midnight Scan Club (MSC) dataset (10 subjects).
  • Macaques: The Newcastle (NC) cohort (8 awake macaques).
  • Mice: A dataset of 44 head-fixed, awake male mice.

Key Insights and Ideas

  • The "Universal Four": Humans and macaques share four dominant C-modes. Three of these are also clearly present in mice. This suggests a fundamental "alphabet" of brain states that has persisted for millions of years.
  • Coupled Oscillators: The transitions between brain states aren't random. They behave like a system of coupled oscillators moving at "infraslow" frequencies (0.01–0.03 Hz). The brain doesn't just jump between states; it cycles through them in a predictable, wave-like fashion.
  • The Global Pulse: C-modes are timed to the "breath" of the brain. They occur at specific phases of the Global Signal. For example, certain networks always activate during the "rising" phase of the global pulse, while others activate during the "falling" phase.
  • Evolutionary Re-prioritization: While the modes are the same, the frequency with which we use them differs. Humans spend significantly more time in "associative" modes (linked to complex cognition and the DMN), while mice spend more time in "sensory-motor" modes. This reflects the different ecological and cognitive needs of each species.

Key Results

  • Spatial Conservation: The C-modes explained a large portion of variance in fMRI data across all species ($R^2 > 0.58$ in mice, $> 0.75$ in humans).
  • Dynamics Build the Static: The researchers showed that the most frequent dynamic C-modes actually "reconstitute" the static functional connectome. In other words, the long-term map of the brain is just the sum of these recurring dynamic flickers.
  • Predictive Power: The occurrence rate of a C-mode was highly predictive of its corresponding "connectivity gradient" ($R^2 > 0.78$ for all species), providing a direct link between momentary transitions and the brain's overarching hierarchical organization.

Limitations and Future Work

  • The Primate Gap: One specific human/macaque mode (C-mode 1, involving advanced attention networks) wasn't reliably found in mice. This is likely because rodents lack the highly developed prefrontal and parietal structures found in primates.
  • Physiological Origins: While the study confirms these patterns exist in the BOLD (blood-oxygen-level-dependent) signal, it remains debated how much of this is purely "neural" versus influenced by non-neural factors like cerebrospinal fluid flow or vascular rhythms.
  • Phylogenetic Sampling: The authors suggest that future work should include "intermediate" species, such as New World monkeys (marmosets), to better trace how these dynamic modes evolved and branched across the tree of life.

Summary updated

Abstract

Evolutionarily relevant networks have been previously described in several mammalian species using time-averaged analyses of fMRI time-series. However, fMRI network activity is highly dynamic and continually evolves over timescales of seconds. Whether the dynamic organization of resting-state fMRI network activity is conserved across mammalian species remains unclear. Using frame-wise clustering of fMRI time-series, we find that intrinsic fMRI network dynamics in awake male macaques and humans is characterized by recurrent transitions between a set of 4 dominant, neuroanatomically homologous fMRI coactivation modes (C-modes), three of which are also plausibly represented in the male rodent brain. Importantly, in all species C-modes exhibit species-invariant dynamic features, including preferred occurrence at specific phases of fMRI global signal fluctuations, and a state transition structure compatible with infraslow coupled oscillator dynamics. Moreover, dominant C-mode occurrence reconstitutes the static organization of the fMRI connectome in all species, and is predictive of ranking of corresponding fMRI connectivity gradients. These results reveal a set of species-invariant principles underlying the dynamic organization of fMRI networks in mammalian species, and offer novel opportunities to relate fMRI network findings across the phylogenetic tree.
BibTeX
@inproceedings{gutierrezbarragan2024evolutionarily,
  author = {Daniel Gutierrez‐Barragan and Julian S.B. Ramirez and Stefano Panzeri and Ting Xu and Alessandro Gozzi},
  title = {{Evolutionarily conserved fMRI network dynamics in the mouse, macaque, and human brain}},
  booktitle = {Nature Communications},
  year = {2024},
  doi = {10.1038/s41467-024-52721-8},
}
Figure