What does the brain do when we breathe? Humans and other animals with lungs depend on breathing to supply their cells with oxygen for energy production. Neurons in the brain are supplied oxygen through an intricate system of blood vessels. When active, neurons consume a lot of energy and require a steady supply of oxygen-rich blood. In fact, this relationship between blood vessels and activity of neurons in the brain is so tightly linked that to study neuron activity researchers and clinicians often use an approach called functional magnetic resonance imaging (fMRI) to analyze the flow of oxygenated blood in the brain.

This imaging technique allows scientists to map how active different parts of the brain are at any given time without the need for an invasive medical procedure. Unfortunately, fMRI results are affected by the cycles of inhalation and exhalation that take place while breathing, even when an individual is at rest. This is because the rate and depth of respiration can vary, resulting in the body moving unpredictably and in CO₂ levels fluctuating in the brain, which can lead to changes in fMRI signals that do not correlate with neuron activity. Such misleading measurements are called ‘artifacts’. The assumption that these fMRI results do not represent real brain activity has meant that the effects of breathing on neuron activity in different parts of the brain is poorly understood.

To solve this issue, Tu and Zhang performed fMRI on rats and combined the results with measurements of the depth and rate of respiration, and with electrophysiology, an approach that allowed them to directly record the electrical properties of neurons. This allowed them to map out the network of neurons that become active in response to breathing.

The results show that breathing leads to a specific fMRI signal that can be distinguished from the artifacts introduced by fluctuating CO₂ levels and body movements. The signal correlates with the activity of neurons measured using electrophysiology and with breathing patterns, and it disappears when the electrical activity of neurons in the brain is suppressed, even if the rats are still breathing. This suggests that breathing affects brain activity that is independent of the previously described artifacts.

Future studies may focus on how the brain responds to breathing or how respiration itself is controlled by the brain, with the methods developed allowing researchers to explore regions of the brain that increase their activity while breathing. This clears the path towards investigating the neural mechanisms underlying therapies and exercises that focus on breathing.

Resting-state fMRI (rsfMRI), which measures spontaneous blood-oxygen-level-dependent (BOLD) signal, is a powerful tool for non-invasively investigating brain-wide functional connectivity (Biswal et al., 1995; Fox and Raichle, 2007; Smith et al., 2009). Due to its hemodynamic nature, the rsfMRI signal is susceptible to systemic physiological changes such as respiration and cardiac pulsations (Murphy et al., 2013; Liu, 2016; Caballero-Gaudes and Reynolds, 2017; Chen et al., 2020), and these effects are usually treated as non-neuronal artifacts in rsfMRI data.

Respiration is a major physiological process that drives fluctuations in cerebral blood flow and oxygenation (Liu, 2016). Respiration can affect the BOLD signal (Thomason et al., 2005; Abbott et al., 2005; Birn et al., 2006) with two types of effects resulting from slow respiration variations and fast cyclic changes, respectively. Slowly varying changes (typically below 0.15 Hz) in breathing rate and depth, which can be quantified as respiration volume per time (RVT) (Birn et al., 2006), covary with arterial CO₂ (Wise et al., 2004; Chang and Glover, 2009; Hoiland et al., 2016; Liu et al., 2017) and affect the BOLD signal via vasodilatory effects and/or autonomic influences of the vessel tone (Duyn et al., 2020; Picchioni et al., 2022). On the other hand, faster respiratory cycles (i.e., periodic inspiration and expiration), accompanied by the corresponding chest and neck movement, can induce changes in the static magnetic field, which in turn lead to BOLD signal changes (Glover et al., 2000; Raj et al., 2000; Windischberger et al., 2002). Both effects can be effectively mitigated/removed by standard rsfMRI preprocessing methods (Birn et al., 2006; Glover et al., 2000).

In addition to the well-characterized respiration-related non-neural artifacts, there is evidence hinting a possible neural component in the respiration–rsfMRI interaction (Yuan et al., 2013; Shams et al., 2021). First, respiration can directly drive brain-wide neuronal oscillations (Tort et al., 2018a; Kay et al., 2009) not only in the olfactory bulb (Adrian, 1942) and piriform cortex (Fontanini et al., 2003)—brain regions directly related to breathing—but also in the medial prefrontal cortex (mPFC), somatosensory cortex, and hippocampus, and this effect has been consistently found across species (Ito et al., 2014; Yanovsky et al., 2014; Biskamp et al., 2017; Zelano et al., 2016). In addition, respiration changes can be associated with arousal and/or emotion-related brain state changes, which covary with cortical activity (Yackle et al., 2017; Shea, 1996; Homma and Masaoka, 2008; Folschweiller and Sauer, 2021). Therefore, in addition to the artifactual effects aforementioned, respiration may affect the rsfMRI signal by directly modulating the neural activity. However, this potential neural component in the respiration–fMRI relationship is largely unexplored.

To gain a comprehensive understanding of the relationships between respiration, neuronal activity, and rsfMRI signal, here we simultaneously acquired rsfMRI, electrophysiology, and respiration data in anesthetized rats. Anesthesia was used to ensure our results are not confounded by the animal’s motion, which affects all three signals. Based on these measures, an RVT-correlated rsfMRI network was identified. Importantly, regressing out gamma activity or silencing neural activity across the brain disrupted this respiration-related network, suggesting that this respiration–rsfMRI relationship is mediated by neural activity.

To determine the potential role of neural activity in the respiration–rsfMRI relationship, we simultaneously recorded the electrophysiology and respiration signals along with rsfMRI data in rats (Figure 1A). The respiration signal was recorded by a respiration sensor placed under the animal’s chest (Figure 1A). Representative raw respiration signal, the respiration rate distribution across all scans, and the averaged respiration power are shown in Figure 1B, C. Slow respiration variations are quantified by RVT, calculated as the difference of consecutive peaks of inspiration and expiration divided by the time interval between the two adjacent signal maxima (or minima) (Figure 1D; Birn et al., 2006). The electrophysiology signal was recorded using an MR-compatible electrode implanted in the right side of the anterior cingulate cortex (ACC). The ACC was selected given its critical role in respiratory modulation, evidenced by early studies showing that ACC was activated during breathlessness (Liotti et al., 2001; Evans et al., 2002). More recent research further demonstrates the role of ACC in respiratory control (Evans et al., 2009; Holton et al., 2021; Tort et al., 2018b). For example, electrophysiology recordings in humans showed that ACC exhibited different neural responses to separate respirational tasks including breath-holding, voluntary deep breathing, and hypercapnia (Holton et al., 2021). In addition, the ACC is a key region in the rodent default-mode network (DMN) (Lu et al., 2012; Tu et al., 2021a), which has been linked to respiration-related fMRI signal changes (Birn et al., 2006). The location of the electrode was confirmed by T2-weighted structural images (Figure 1—figure supplement 1). MRI artifacts in the electrophysiology signal were removed using a template regression method (Logothetis et al., 2001; Pan et al., 2011) (see Materials and methods and Figure 1—figure supplement 2), and local field potential (LFP) and spectrogram were obtained using denoised electrophysiologic data (Figure 1E).

Figure 1 with 2 supplements see all

Download asset Open asset

Gamma-band neural oscillations are, respectively, associated with respiration and rsfMRI signals

We first asked whether the respiration is linked to neural activity changes (Figure 2A) in lightly sedated animals (combined low-dose dexmedetomidine and isoflurane, see Methods and materials). Prominent coherence between the respiration signal and LFP is observed with the dominant peak at ~1.1 Hz (Figure 2B, C), in line with the respiration frequency in animals (Figure 1C). This result is consistent with the finding of respiration-entrained LFP oscillations previously reported (Tort et al., 2018a; Kay et al., 2009).

Figure 2 with 2 supplements see all

Download asset Open asset

To further dissect the respiration–LFP relationship, the LFP was separated into five conventionally defined frequency bands including gamma (40–100 Hz), delta (1–4 Hz), theta (4–7 Hz), alpha (7–13 Hz), and beta band (13–30 Hz) (Lu et al., 2016; Zhang et al., 2020). Gamma-band power displays significant coherence with the RVT at ~0.035 Hz with a zero phase lag (Figure 2D, E, phase [mean ± ste] = −0.08 ± 0.5 pi, with the range of [−pi, pi]), suggesting a phase-locking relationship between the two measures. In contrast, this coherence is not observed in any other frequency bands (Figure 2F–I). These results remain the same when the LFP data were analyzed using the local subtraction method (Figure 2—figure supplement 1A–C), ruling out the potential artifact resulting from the volume conduction of signals between the cortex and olfactory bulb (Parabucki and Lampl, 2017). The RVT–gamma power coherence in individual lightly sedated animals is shown in Figure 2—figure supplement 2. Taken together, our data demonstrate that respiration is associated with gamma-band neural oscillations in the ACC.

We next examined the relationship between the gamma-band LFP and rsfMRI signals (Figure 3A; Winder et al., 2017). The ACC gamma power was first convolved with a hemodynamic response function (HRF, defined by a single gamma probability distribution function (a = 3, b = 0.8) Liang et al., 2017, Figure 3B), which was then voxel-wise correlated to brain-wide rsfMRI signals. The r value between the ACC rsfMRI signal and HRF-convolved gamma power was 0.086 (1200 time points, p = 0.003). Our data show that the gamma power-derived rsfMRI correlation map (Figure 3C) is highly consistent with the ACC resting-state functional connectivity (RSFC) seedmap with the right ACC (i.e., the electrode implanted side) defined as the seed (Figure 3D), evidenced by a strong voxel-to-voxel spatial correlation between these two maps (Figure 3E, R = 0.775, p < 10⁻¹⁵). Again, the same gamma power-derived correlation pattern is observed when the LFP data were analyzed using the local subtraction method (Figure 2—figure supplement 1D). The gamma power-derived rsfMRI correlation maps in individual animals are shown in Figure 3—figure supplement 1. These data demonstrate that the gamma-band LFP is tightly linked to the rsfMRI signal.

Figure 3 with 2 supplements see all

Download asset Open asset

Respiration is associated with a characteristic rsfMRI network, mediated by gamma-band neural activity

Given that the gamma-band power is, respectively, associated with the RVT and rsfMRI signals, we specifically asked how RVT is related to the rsfMRI signal in lightly sedated animals by calculating voxel-wise correlations between the RVT and rsfMRI signals (Figure 4A). This analysis generates a respiration-related rsfMRI network (Figure 4B), involving key brain regions controlling respiration such as the piriform cortex. It is known that the piriform cortex receives inputs from the olfactory bulb, and can be directly activated when the animal breathes. In addition, this network includes regions involved in the rodent DMN such as the ACC, mPFC, orbital, retrosplenial, and primary somatosensory cortices, as well as the hippocampus (Lu et al., 2012; Figure 4C, one-sample t-test, p < 0.05, false discovery rate [FDR] corrected). The resemblance of this respiration-related network and DMN well agrees with the human literature that the physiological effects on rsfMRI data are colocalized with the DMN (Birn et al., 2006). The respiration-related rsfMRI network in individual animals is shown Figure 4—figure supplement 1. Taken together, slow respiration variations exhibit a characteristic correlation pattern with brain-wide rsfMRI signals, representing a respiration-related rsfMRI network.

Figure 4 with 3 supplements see all

Download asset Open asset

To test whether this respiration-related rsfMRI network is mediated by gamma-band activity, the gamma power was, respectively, regressed out from the RVT and voxel-wise rsfMRI signals, and the same correlational analysis was repeated on the residual signals. In this case, the respiration-related network is diminished (Figure 4D). This result is confirmed by the contrast of the correlation maps before (Figure 4B) and after (Figure 4D) gamma power regression, with essentially all brain regions involved in the respiration-related network exhibiting reduced RVT–rsfMRI correlation after the gamma power is regressed out (Figure 4E, paired t-test, p < 0.05, FDR corrected). These results demonstrate that the respiration-related rsfMRI network is mediated by gamma-band neural activity.

The respiration-related rsfMRI network is absent at the isoelectric state

To further confirm the necessary role of neural activity in the respiration-related rsfMRI network, we experimentally silenced the neural activity in the whole brain by inducing an isoelectric brain state using high-dose sodium pentobarbital, while maintained the respiration in the rat (Figure 5A; Du et al., 2008). At the isoelectric state, the LFP amplitude and power are close to zero, in remarkable contrast to the electrophysiology data recorded before the drug infusion (Figure 5B). In addition, rsfMRI data recorded at the isoelectric state exhibit a flat, noise-like power distribution, unlike the characteristic 1/f pattern during light sedation (Figure 5I; Zhang et al., 2021). Furthermore, there is no meaningful ACC RSFC in the ACC seedmap at the isoelectric state (Figure 5H), which is distinct from the ACC seedmap during light sedation (Figure 3D). The brain-wide ROI-based RSFC matrix also reveals a global suppression of RSFC at the isoelectric state (Figure 5K), compared to the RSFC matrix observed during light sedation. These data collectively confirm that brain-wide neural activity is silenced in the isoelectric state.

Figure 5 with 1 supplement see all

Download asset Open asset

To examine whether the temporal variation of respiration in the isoelectric state and light sedation is consistent, we calculated the respiration rate (Figure 5D) as well as the standard deviation (SD) and power spectrum of RVT across all scans. We did not observe any significant difference in the SD of the respiration rate (Figure 5E, two-sample t-test, t = −1.0, p = 0.32) or the SD of RVT (Figure 5F, two-sample t-test, t = −0.87, p = 0.39) between the light sedation and isoelectric states, suggesting that the temporal variance of respiration is similar between the two states. Interestingly, we observed a characteristic 1/f pattern in the RVT power during light sedation, but this pattern was not present at the isoelectric state (Figure 5G). This difference is consistent with the ACC BOLD signal power spectra in the light sedation and isoelectric states (Figure 5I).

After confirming the global neural silencing effect, we calculated voxel-wise correlations between the RVT and rsfMRI signals in the isoelectric state. The respiration-related network observed during light sedation is absent when brain-wide neural activity is silenced (Figure 5J), despite similar breathing patterns (Figure 5C–F) at both conditions. The RVT–rsfMRI correlation maps in individual animals in the isoelectric state are shown in Figure 5—figure supplement 1. These data demonstrate that the neural activity plays a necessary role in the respiration-related rsfMRI network, corroborating the notion that the respiration-related rsfMR network we observed is mediated by neural activity.

In this study, we discovered a slow respiration variation-related functional network in rodents. Importantly, we confirmed the neural underpinning of this network, indicated by the phase-locking relationship between the RVT and gamma-band power in the ACC—a node in this network. Regressing out the gamma-band power disrupted the respiration network, suggesting a mediating role of neural activity in this network. To further validate this finding, we experimentally silenced the neural activity across the brain but maintained the respiration in the animal. In this condition, we again failed to observe the respiration-related brain network, confirming the neural underpinning of the network. Lastly, we showed that the respiration-related neural network is distinct from respiration-related rsfMRI artifacts. Overall, our study maps brain-wide neural responses related to respiration. These data provide new insight into understanding the neural activity-mediated respiratory effects on resting-state functional networks. Importantly, this breathing-related brain network might be altered in brain disorders, and thus our findings might potentially provide important clinical value.

All data for this study have been deposited to NITRIC repository.

1. 1. Tu W
   2. Zhang N

   (2022) NeuroImaging Tools and Resources Collaboratory (NITRC)

   ID 1582. Electrophysiology, resting state fMRI and respiration in rats.

   https://www.nitrc.org/docman/?group_id=1582

1. 1. Abbott DF
   2. Opdam HI
   3. Briellmann RS
   4. Jackson GD

   (2005) Brief breath holding may confound functional magnetic resonance imaging studies

   Human Brain Mapping 24:284–290.

   https://doi.org/10.1002/hbm.20086

   • PubMed
   • Google Scholar
2. 1. Adrian ED

   (1942) Olfactory reactions in the brain of the hedgehog

   The Journal of Physiology 100:459–473.

   https://doi.org/10.1113/jphysiol.1942.sp003955

   • PubMed
   • Google Scholar
3. 1. Baertsch NA
   2. Bush NE
   3. Burgraff NJ
   4. Ramirez JM

   (2021) Dual mechanisms of opioid-induced respiratory depression in the inspiratory rhythm-generating network

   eLife 10:e67523.

   https://doi.org/10.7554/eLife.67523

   • PubMed
   • Google Scholar
4. 1. Bastos AM
   2. Donoghue JA
   3. Brincat SL
   4. Mahnke M
   5. Yanar J
   6. Correa J
   7. Waite AS
   8. Lundqvist M
   9. Roy J
   10. Brown EN
   11. Miller EK

   (2021) Neural effects of propofol-induced unconsciousness and its reversal using thalamic stimulation

   eLife 10:e60824.

   https://doi.org/10.7554/eLife.60824

   • PubMed
   • Google Scholar
5. 1. Benveniste H
   2. Lee H
   3. Ding F
   4. Sun Q
   5. Al-Bizri E
   6. Makaryus R
   7. Probst S
   8. Nedergaard M
   9. Stein EA
   10. Lu H

   (2017) Anesthesia with dexmedetomidine and low-dose isoflurane increases solute transport via the glymphatic pathway in rat brain when compared with high-dose isoflurane

   Anesthesiology 127:976–988.

   https://doi.org/10.1097/ALN.0000000000001888

   • PubMed
   • Google Scholar
6. 1. Birn RM
   2. Diamond JB
   3. Smith MA
   4. Bandettini PA

   (2006) Separating respiratory-variation-related fluctuations from neuronal-activity-related fluctuations in fmri

   NeuroImage 31:1536–1548.

   https://doi.org/10.1016/j.neuroimage.2006.02.048

   • PubMed
   • Google Scholar
7. 1. Birn RM
   2. Smith MA
   3. Jones TB
   4. Bandettini PA

   (2008) The respiration response function: the temporal dynamics of fMRI signal fluctuations related to changes in respiration

   NeuroImage 40:644–654.

   https://doi.org/10.1016/j.neuroimage.2007.11.059

   • PubMed
   • Google Scholar
8. 1. Biskamp J
   2. Bartos M
   3. Sauer JF

   (2017) Organization of prefrontal network activity by respiration-related oscillations

   Scientific Reports 7:45508.

   https://doi.org/10.1038/srep45508

   • PubMed
   • Google Scholar
9. 1. Biswal B
   2. Yetkin FZ
   3. Haughton VM
   4. Hyde JS

   (1995) Functional connectivity in the motor cortex of resting human brain using echo-planar MRI

   Magnetic Resonance in Medicine 34:537–541.

   https://doi.org/10.1002/mrm.1910340409

   • PubMed
   • Google Scholar
10. 1. Caballero-Gaudes C
    2. Reynolds RC

    (2017) Methods for cleaning the BOLD fMRI signal

    NeuroImage 154:128–149.

    https://doi.org/10.1016/j.neuroimage.2016.12.018

    • PubMed
    • Google Scholar
11. 1. Carmichael ST
    2. Clugnet MC
    3. Price JL

    (1994) Central olfactory connections in the macaque monkey

    The Journal of Comparative Neurology 346:403–434.

    https://doi.org/10.1002/cne.903460306

    • PubMed
    • Google Scholar
12. 1. Chang C
    2. Glover GH

    (2009) Relationship between respiration, end-tidal CO2, and BOLD signals in resting-state fMRI

    NeuroImage 47:1381–1393.

    https://doi.org/10.1016/j.neuroimage.2009.04.048

    • PubMed
    • Google Scholar
13. 1. Chen JE
    2. Lewis LD
    3. Chang C
    4. Tian Q
    5. Fultz NE
    6. Ohringer NA
    7. Rosen BR
    8. Polimeni JR

    (2020) Resting-state “physiological networks.”

    NeuroImage 213:116707.

    https://doi.org/10.1016/j.neuroimage.2020.116707

    • PubMed
    • Google Scholar
14. 1. Doll A
    2. Hölzel BK
    3. Mulej Bratec S
    4. Boucard CC
    5. Xie X
    6. Wohlschläger AM
    7. Sorg C

    (2016) Mindful attention to breath regulates emotions via increased amygdala-prefrontal cortex connectivity

    NeuroImage 134:305–313.

    https://doi.org/10.1016/j.neuroimage.2016.03.041

    • PubMed
    • Google Scholar
15. 1. Du F
    2. Zhu XH
    3. Zhang Y
    4. Friedman M
    5. Zhang N
    6. Ugurbil K
    7. Chen W

    (2008) Tightly coupled brain activity and cerebral ATP metabolic rate

    PNAS 105:6409–6414.

    https://doi.org/10.1073/pnas.0710766105

    • PubMed
    • Google Scholar
16. 1. Duyn JH
    2. Ozbay PS
    3. Chang C
    4. Picchioni D

    (2020) Physiological changes in sleep that affect fMRI inference

    Current Opinion in Behavioral Sciences 33:42–50.

    https://doi.org/10.1016/j.cobeha.2019.12.007

    • PubMed
    • Google Scholar
17. 1. Evans KC
    2. Banzett RB
    3. Adams L
    4. McKay L
    5. Frackowiak RSJ
    6. Corfield DR

    (2002) BOLD fmri identifies limbic, paralimbic, and cerebellar activation during air hunger

    Journal of Neurophysiology 88:1500–1511.

    https://doi.org/10.1152/jn.2002.88.3.1500

    • PubMed
    • Google Scholar
18. 1. Evans KC
    2. Dougherty DD
    3. Schmid AM
    4. Scannell E
    5. McCallister A
    6. Benson H
    7. Dusek JA
    8. Lazar SW

    (2009) Modulation of spontaneous breathing via limbic/paralimbic-bulbar circuitry: an event-related fMRI study

    NeuroImage 47:961–971.

    https://doi.org/10.1016/j.neuroimage.2009.05.025

    • PubMed
    • Google Scholar
19. 1. Fan J
    2. Xu P
    3. Van Dam NT
    4. Eilam-Stock T
    5. Gu X
    6. Luo Y
    7. Hof PR

    (2012) Spontaneous brain activity relates to autonomic arousal

    The Journal of Neuroscience 32:11176–11186.

    https://doi.org/10.1523/JNEUROSCI.1172-12.2012

    • PubMed
    • Google Scholar
20. 1. Folschweiller S
    2. Sauer JF

    (2021) Respiration-driven brain oscillations in emotional cognition

    Frontiers in Neural Circuits 15:761812.

    https://doi.org/10.3389/fncir.2021.761812

    • PubMed
    • Google Scholar
21. 1. Fontanini A
    2. Spano P
    3. Bower JM

    (2003)

    Ketamine-xylazine-induced slow (< 1.5 hz) oscillations in the rat piriform (olfactory) cortex are functionally correlated with respiration

    The Journal of Neuroscience 23:7993–8001.

    • PubMed
    • Google Scholar
22. 1. Fontanini A
    2. Bower JM

    (2005) Variable coupling between olfactory system activity and respiration in ketamine/xylazine anesthetized rats

    Journal of Neurophysiology 93:3573–3581.

    https://doi.org/10.1152/jn.01320.2004

    • PubMed
    • Google Scholar
23. 1. Fox MD
    2. Raichle ME

    (2007) Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging

    Nature Reviews. Neuroscience 8:700–711.

    https://doi.org/10.1038/nrn2201

    • PubMed
    • Google Scholar
24. 1. García-Cabezas MÁ
    2. Barbas H

    (2014) A direct anterior cingulate pathway to the primate primary olfactory cortex may control attention to olfaction

    Brain Structure & Function 219:1735–1754.

    https://doi.org/10.1007/s00429-013-0598-3

    • PubMed
    • Google Scholar
25. 1. Glover GH
    2. Li TQ
    3. Ress D

    (2000) Image-based method for retrospective correction of physiological motion effects in fmri: RETROICOR

    Magnetic Resonance in Medicine 44:162–167.

    https://doi.org/10.1002/1522-2594(200007)44:1<162::aid-mrm23>3.0.co;2-e

    • PubMed
    • Google Scholar
26. 1. Grandjean J
    2. Schroeter A
    3. Batata I
    4. Rudin M

    (2014) Optimization of anesthesia protocol for resting-state fMRI in mice based on differential effects of anesthetics on functional connectivity patterns

    NeuroImage 102 Pt 2:838–847.

    https://doi.org/10.1016/j.neuroimage.2014.08.043

    • PubMed
    • Google Scholar
27. 1. Herrero JL
    2. Khuvis S
    3. Yeagle E
    4. Cerf M
    5. Mehta AD

    (2018) Breathing above the brain stem: volitional control and attentional modulation in humans

    Journal of Neurophysiology 119:145–159.

    https://doi.org/10.1152/jn.00551.2017

    • PubMed
    • Google Scholar
28. 1. Hoiland RL
    2. Bain AR
    3. Rieger MG
    4. Bailey DM
    5. Ainslie PN

    (2016) Hypoxemia, oxygen content, and the regulation of cerebral blood flow

    American Journal of Physiology. Regulatory, Integrative and Comparative Physiology 310:R398–R413.

    https://doi.org/10.1152/ajpregu.00270.2015

    • PubMed
    • Google Scholar
29. 1. Holton P
    2. Huang Y
    3. Bahuri NFA
    4. Boccard S
    5. Hyam JA
    6. Paterson DJ
    7. Dorrington KL
    8. Aziz TZ
    9. Moosavi SH
    10. Green AL

    (2021) Differential responses to breath-holding, voluntary deep breathing and hypercapnia in left and right dorsal anterior cingulate

    Experimental Physiology 106:726–735.

    https://doi.org/10.1113/EP088961

    • PubMed
    • Google Scholar
30. 1. Homma I
    2. Masaoka Y

    (2008) Breathing rhythms and emotions

    Experimental Physiology 93:1011–1021.

    https://doi.org/10.1113/expphysiol.2008.042424

    • PubMed
    • Google Scholar
31. 1. Hyam JA
    2. Wang S
    3. Roy H
    4. Moosavi SH
    5. Martin SC
    6. Brittain JS
    7. Coyne T
    8. Silburn P
    9. Aziz TZ
    10. Green AL

    (2019) The pedunculopontine region and breathing in Parkinson’s disease

    Annals of Clinical and Translational Neurology 6:837–847.

    https://doi.org/10.1002/acn3.752

    • PubMed
    • Google Scholar
32. 1. Illig KR

    (2005) Projections from orbitofrontal cortex to anterior piriform cortex in the rat suggest a role in olfactory information processing

    The Journal of Comparative Neurology 488:224–231.

    https://doi.org/10.1002/cne.20595

    • PubMed
    • Google Scholar
33. 1. Ito J
    2. Roy S
    3. Liu Y
    4. Cao Y
    5. Fletcher M
    6. Lu L
    7. Boughter JD
    8. Grün S
    9. Heck DH

    (2014) Whisker barrel cortex delta oscillations and gamma power in the awake mouse are linked to respiration

    Nature Communications 5:3572.

    https://doi.org/10.1038/ncomms4572

    • PubMed
    • Google Scholar
34. 1. Kay LM
    2. Beshel J
    3. Brea J
    4. Martin C
    5. Rojas-Líbano D
    6. Kopell N

    (2009) Olfactory oscillations: the what, how and what for

    Trends in Neurosciences 32:207–214.

    https://doi.org/10.1016/j.tins.2008.11.008

    • PubMed
    • Google Scholar
35. 1. Kluger DS
    2. Balestrieri E
    3. Busch NA
    4. Gross J

    (2021) Respiration aligns perception with neural excitability

    eLife 10:e70907.

    https://doi.org/10.7554/eLife.70907

    • PubMed
    • Google Scholar
36. 1. Lecrux C
    2. Hamel E

    (2016) Neuronal networks and mediators of cortical neurovascular coupling responses in normal and altered brain states

    Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 371:20150350.

    https://doi.org/10.1098/rstb.2015.0350

    • PubMed
    • Google Scholar
37. 1. Liang Z
    2. Ma Y
    3. Watson GDR
    4. Zhang N

    (2017) Simultaneous gcamp6-based fiber photometry and fMRI in rats

    Journal of Neuroscience Methods 289:31–38.

    https://doi.org/10.1016/j.jneumeth.2017.07.002

    • PubMed
    • Google Scholar
38. 1. Liotti M
    2. Brannan S
    3. Egan G
    4. Shade R
    5. Madden L
    6. Abplanalp B
    7. Robillard R
    8. Lancaster J
    9. Zamarripa FE
    10. Fox PT
    11. Denton D

    (2001) Brain responses associated with consciousness of breathlessness (air hunger)

    PNAS 98:2035–2040.

    https://doi.org/10.1073/pnas.98.4.2035

    • PubMed
    • Google Scholar
39. 1. Liu TT

    (2016) Noise contributions to the fMRI signal: an overview

    NeuroImage 143:141–151.

    https://doi.org/10.1016/j.neuroimage.2016.09.008

    • PubMed
    • Google Scholar
40. 1. Liu P
    2. Li Y
    3. Pinho M
    4. Park DC
    5. Welch BG
    6. Lu H

    (2017) Cerebrovascular reactivity mapping without gas challenges

    NeuroImage 146:320–326.

    https://doi.org/10.1016/j.neuroimage.2016.11.054

    • PubMed
    • Google Scholar
41. 1. Lockmann ALV
    2. Laplagne DA
    3. Leão RN
    4. Tort ABL

    (2016) A respiration-coupled rhythm in the rat hippocampus independent of theta and slow oscillations

    The Journal of Neuroscience 36:5338–5352.

    https://doi.org/10.1523/JNEUROSCI.3452-15.2016

    • PubMed
    • Google Scholar
42. 1. Logothetis NK
    2. Pauls J
    3. Augath M
    4. Trinath T
    5. Oeltermann A

    (2001) Neurophysiological investigation of the basis of the fMRI signal

    Nature 412:150–157.

    https://doi.org/10.1038/35084005

    • PubMed
    • Google Scholar
43. 1. Lu H
    2. Zou Q
    3. Gu H
    4. Raichle ME
    5. Stein EA
    6. Yang Y

    (2012) Rat brains also have a default mode network

    PNAS 109:3979–3984.

    https://doi.org/10.1073/pnas.1200506109

    • PubMed
    • Google Scholar
44. 1. Lu H
    2. Wang L
    3. Rea WW
    4. Brynildsen JK
    5. Jaime S
    6. Zuo Y
    7. Stein EA
    8. Yang Y

    (2016) Low- but not high-frequency LFP correlates with spontaneous BOLD fluctuations in rat whisker barrel cortex

    Cerebral Cortex 26:683–694.

    https://doi.org/10.1093/cercor/bhu248

    • PubMed
    • Google Scholar
45. 1. Magri C
    2. Schridde U
    3. Murayama Y
    4. Panzeri S
    5. Logothetis NK

    (2012) The amplitude and timing of the BOLD signal reflects the relationship between local field potential power at different frequencies

    The Journal of Neuroscience 32:1395–1407.

    https://doi.org/10.1523/JNEUROSCI.3985-11.2012

    • PubMed
    • Google Scholar
46. 1. McKay LC
    2. Evans KC
    3. Frackowiak RSJ
    4. Corfield DR

    (2003) Neural correlates of voluntary breathing in humans

    Journal of Applied Physiology 95:1170–1178.

    https://doi.org/10.1152/japplphysiol.00641.2002

    • PubMed
    • Google Scholar
47. 1. McKay LC
    2. Janczewski WA
    3. Feldman JL

    (2005) Sleep-disordered breathing after targeted ablation of prebötzinger complex neurons

    Nature Neuroscience 8:1142–1144.

    https://doi.org/10.1038/nn1517

    • PubMed
    • Google Scholar
48. 1. Melnychuk MC
    2. Dockree PM
    3. O’Connell RG
    4. Murphy PR
    5. Balsters JH
    6. Robertson IH

    (2018) Coupling of respiration and attention via the locus coeruleus: effects of meditation and pranayama

    Psychophysiology 55:e13091.

    https://doi.org/10.1111/psyp.13091

    • PubMed
    • Google Scholar
49. 1. Murphy K
    2. Birn RM
    3. Bandettini PA

    (2013) Resting-State fMRI confounds and cleanup

    NeuroImage 80:349–359.

    https://doi.org/10.1016/j.neuroimage.2013.04.001

    • PubMed
    • Google Scholar
50. 1. Nguyen Chi V
    2. Müller C
    3. Wolfenstetter T
    4. Yanovsky Y
    5. Draguhn A
    6. Tort ABL
    7. Brankačk J

    (2016) Hippocampal respiration-driven rhythm distinct from theta oscillations in awake mice

    The Journal of Neuroscience 36:162–177.

    https://doi.org/10.1523/JNEUROSCI.2848-15.2016

    • PubMed
    • Google Scholar
51. 1. Pais-Roldán P
    2. Biswal B
    3. Scheffler K
    4. Yu X

    (2018) Identifying respiration-related aliasing artifacts in the rodent resting-state fMRI

    Frontiers in Neuroscience 12:788.

    https://doi.org/10.3389/fnins.2018.00788

    • PubMed
    • Google Scholar
52. 1. Pan W-J
    2. Thompson G
    3. Magnuson M
    4. Majeed W
    5. Jaeger D
    6. Keilholz S

    (2011) Broadband local field potentials correlate with spontaneous fluctuations in functional magnetic resonance imaging signals in the rat somatosensory cortex under isoflurane anesthesia

    Brain Connectivity 1:119–131.

    https://doi.org/10.1089/brain.2011.0014

    • PubMed
    • Google Scholar
53. 1. Parabucki A
    2. Lampl I

    (2017) Volume conduction coupling of whisker-evoked cortical LFP in the mouse olfactory bulb

    Cell Reports 21:919–925.

    https://doi.org/10.1016/j.celrep.2017.09.094

    • PubMed
    • Google Scholar
54. 1. Picchioni D
    2. Özbay PS
    3. Mandelkow H
    4. de Zwart JA
    5. Wang Y
    6. van Gelderen P
    7. Duyn JH

    (2022) Autonomic arousals contribute to brain fluid pulsations during sleep

    NeuroImage 249:118888.

    https://doi.org/10.1016/j.neuroimage.2022.118888

    • PubMed
    • Google Scholar
55. 1. Purdon PL
    2. Pierce ET
    3. Mukamel EA
    4. Prerau MJ
    5. Walsh JL
    6. Wong KFK
    7. Salazar-Gomez AF
    8. Harrell PG
    9. Sampson AL
    10. Cimenser A
    11. Ching S
    12. Kopell NJ
    13. Tavares-Stoeckel C
    14. Habeeb K
    15. Merhar R
    16. Brown EN

    (2013) Electroencephalogram signatures of loss and recovery of consciousness from propofol

    PNAS 110:E1142–E1151.

    https://doi.org/10.1073/pnas.1221180110

    • PubMed
    • Google Scholar
56. 1. Raj D
    2. Paley DP
    3. Anderson AW
    4. Kennan RP
    5. Gore JC

    (2000) A model for susceptibility artefacts from respiration in functional echo-planar magnetic resonance imaging

    Physics in Medicine and Biology 45:3809–3820.

    https://doi.org/10.1088/0031-9155/45/12/321

    • PubMed
    • Google Scholar
57. 1. Rojas-Líbano D
    2. Kay LM

    (2008) Olfactory system gamma oscillations: the physiological dissection of a cognitive neural system

    Cognitive Neurodynamics 2:179–194.

    https://doi.org/10.1007/s11571-008-9053-1

    • PubMed
    • Google Scholar
58. Book

    1. Schmahmann JD

    (2009) Fiber Pathways of the Brain

    Oxford University Press.

    https://doi.org/10.1093/acprof:oso/9780195104233.001.0001

    • Google Scholar
59. 1. Schwinn DA
    2. McIntyre RW
    3. Reves JG

    (1990) Isoflurane-induced vasodilation: role of the alpha-adrenergic nervous system

    Anesthesia and Analgesia 71:451–459.

    https://doi.org/10.1213/00000539-199011000-00001

    • PubMed
    • Google Scholar
60. 1. Shams S
    2. LeVan P
    3. Chen JJ

    (2021) The neuronal associations of respiratory-volume variability in the resting state

    NeuroImage 230:117783.

    https://doi.org/10.1016/j.neuroimage.2021.117783

    • PubMed
    • Google Scholar
61. 1. Shea SA

    (1996) Behavioural and arousal-related influences on breathing in humans

    Experimental Physiology 81:1–26.

    https://doi.org/10.1113/expphysiol.1996.sp003911

    • PubMed
    • Google Scholar
62. 1. Smith JC
    2. Ellenberger HH
    3. Ballanyi K
    4. Richter DW
    5. Feldman JL

    (1991) Pre-Bötzinger complex: a brainstem region that may generate respiratory rhythm in mammals

    Science 254:726–729.

    https://doi.org/10.1126/science.1683005

    • PubMed
    • Google Scholar
63. 1. Smith SM
    2. Fox PT
    3. Miller KL
    4. Glahn DC
    5. Fox PM
    6. Mackay CE
    7. Filippini N
    8. Watkins KE
    9. Toro R
    10. Laird AR
    11. Beckmann CF

    (2009) Correspondence of the brain’s functional architecture during activation and rest

    PNAS 106:13040–13045.

    https://doi.org/10.1073/pnas.0905267106

    • PubMed
    • Google Scholar
64. Book

    1. Swanson LW

    (2004)

    Brain Maps: Structure of the Rat Brain

    Elsevier.

    • Google Scholar
65. 1. Talke P
    2. Anderson BJ

    (2018) Pharmacokinetics and pharmacodynamics of dexmedetomidine-induced vasoconstriction in healthy volunteers

    British Journal of Clinical Pharmacology 84:1364–1372.

    https://doi.org/10.1111/bcp.13571

    • PubMed
    • Google Scholar
66. 1. Thomason ME
    2. Burrows BE
    3. Gabrieli JDE
    4. Glover GH

    (2005) Breath holding reveals differences in fMRI BOLD signal in children and adults

    NeuroImage 25:824–837.

    https://doi.org/10.1016/j.neuroimage.2004.12.026

    • PubMed
    • Google Scholar
67. 1. Tort ABL
    2. Brankačk J
    3. Draguhn A

    (2018a) Respiration-entrained brain rhythms are global but often overlooked

    Trends in Neurosciences 41:186–197.

    https://doi.org/10.1016/j.tins.2018.01.007

    • PubMed
    • Google Scholar
68. 1. Tort ABL
    2. Ponsel S
    3. Jessberger J
    4. Yanovsky Y
    5. Brankačk J
    6. Draguhn A

    (2018b) Parallel detection of theta and respiration-coupled oscillations throughout the mouse brain

    Scientific Reports 8:6432.

    https://doi.org/10.1038/s41598-018-24629-z

    • PubMed
    • Google Scholar
69. 1. Tu W
    2. Ma Z
    3. Ma Y
    4. Dopfel D
    5. Zhang N

    (2021a) Suppressing anterior cingulate cortex modulates default mode network and behavior in awake rats

    Cerebral Cortex 31:312–323.

    https://doi.org/10.1093/cercor/bhaa227

    • PubMed
    • Google Scholar
70. 1. Tu W
    2. Ma Z
    3. Zhang N

    (2021b) Brain network reorganization after targeted attack at a hub region

    NeuroImage 237:118219.

    https://doi.org/10.1016/j.neuroimage.2021.118219

    • PubMed
    • Google Scholar
71. 1. Winder AT
    2. Echagarruga C
    3. Zhang Q
    4. Drew PJ

    (2017) Weak correlations between hemodynamic signals and ongoing neural activity during the resting state

    Nature Neuroscience 20:1761–1769.

    https://doi.org/10.1038/s41593-017-0007-y

    • PubMed
    • Google Scholar
72. 1. Windischberger C
    2. Langenberger H
    3. Sycha T
    4. Tschernko EM
    5. Fuchsjäger-Mayerl G
    6. Schmetterer L
    7. Moser E

    (2002) On the origin of respiratory artifacts in BOLD-EPI of the human brain

    Magnetic Resonance Imaging 20:575–582.

    https://doi.org/10.1016/s0730-725x(02)00563-5

    • PubMed
    • Google Scholar
73. 1. Wise RG
    2. Ide K
    3. Poulin MJ
    4. Tracey I

    (2004) Resting fluctuations in arterial carbon dioxide induce significant low frequency variations in BOLD signal

    NeuroImage 21:1652–1664.

    https://doi.org/10.1016/j.neuroimage.2003.11.025

    • PubMed
    • Google Scholar
74. 1. Yackle K
    2. Schwarz LA
    3. Kam K
    4. Sorokin JM
    5. Huguenard JR
    6. Feldman JL
    7. Luo L
    8. Krasnow MA

    (2017) Breathing control center neurons that promote arousal in mice

    Science 355:1411–1415.

    https://doi.org/10.1126/science.aai7984

    • PubMed
    • Google Scholar
75. 1. Yanovsky Y
    2. Ciatipis M
    3. Draguhn A
    4. Tort ABL
    5. Brankačk J

    (2014) Slow oscillations in the mouse hippocampus entrained by nasal respiration

    The Journal of Neuroscience 34:5949–5964.

    https://doi.org/10.1523/JNEUROSCI.5287-13.2014

    • PubMed
    • Google Scholar
76. 1. You T
    2. Im GH
    3. Kim SG

    (2021) Characterization of brain-wide somatosensory BOLD fmri in mice under dexmedetomidine/isoflurane and ketamine/xylazine

    Scientific Reports 11:13110.

    https://doi.org/10.1038/s41598-021-92582-5

    • PubMed
    • Google Scholar
77. 1. Yuan H
    2. Zotev V
    3. Phillips R
    4. Bodurka J

    (2013) Correlated slow fluctuations in respiration, EEG, and BOLD fMRI

    NeuroImage 79:81–93.

    https://doi.org/10.1016/j.neuroimage.2013.04.068

    • PubMed
    • Google Scholar
78. 1. Zelano C
    2. Jiang H
    3. Zhou G
    4. Arora N
    5. Schuele S
    6. Rosenow J
    7. Gottfried JA

    (2016) Nasal respiration entrains human limbic oscillations and modulates cognitive function

    The Journal of Neuroscience 36:12448–12467.

    https://doi.org/10.1523/JNEUROSCI.2586-16.2016

    • PubMed
    • Google Scholar
79. 1. Zhang X
    2. Pan WJ
    3. Keilholz SD

    (2020) The relationship between BOLD and neural activity arises from temporally sparse events

    NeuroImage 207:116390.

    https://doi.org/10.1016/j.neuroimage.2019.116390

    • PubMed
    • Google Scholar
80. 1. Zhang Q
    2. Gheres KW
    3. Drew PJ

    (2021) Origins of 1/f-like tissue oxygenation fluctuations in the murine cortex

    PLOS Biology 19:e3001298.

    https://doi.org/10.1371/journal.pbio.3001298

    • PubMed
    • Google Scholar
81. 1. Zhao X
    2. Bodurka J
    3. Jesmanowicz A
    4. Li SJ

    (2000) B (0) -fluctuation-induced temporal variation in EPI image series due to the disturbance of steady-state free precession

    Magnetic Resonance in Medicine 44:758–765.

    https://doi.org/10.1002/1522-2594(200011)44:5<758::aid-mrm14>3.0.co;2-g

    • PubMed
    • Google Scholar
82. 1. Zhong W
    2. Ciatipis M
    3. Wolfenstetter T
    4. Jessberger J
    5. Müller C
    6. Ponsel S
    7. Yanovsky Y
    8. Brankačk J
    9. Tort ABL
    10. Draguhn A

    (2017) Selective entrainment of gamma subbands by different slow network oscillations

    PNAS 114:4519–4524.

    https://doi.org/10.1073/pnas.1617249114

    • PubMed
    • Google Scholar

Decision letter after peer review:

Thank you for submitting your article "Respiration-driven brain network: neural underpinning of breathing correlates with resting-state fMRI signal" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Timothy Behrens as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Joanes Grandjean (Reviewer #1).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

1) Tempering of language purporting to establish a causal relationship between respiration and the observed γ oscillations.

2) Comment on the relatively small number of rodents used in the study. [In consultation between the reviewers, it has been pointed out that the reported effects are based on statistical tests driven by the number of timepoints, rather than the number of individual rodents.]

3) Discussion of the potential effects of anaesthesia on the study conclusions.

4) Placing results in the context of related human studies.

5) Analysis of RVT estimates (variance) in the isoelectric state.

Reviewer #1 (Recommendations for the authors):

As indicated in the public review, I don't find anything that is lacking in the study. I find it compelling, and I do not wish to burden the authors with additional work that may only remotely nuance their excellent work.

If anything, I am not sure the statement below (and other similar later in the discussion) are sufficiently substantiated. Γ oscillations are correlated with respiration in this study. Other tests to further establish causality would be required, such as comparing spontaneous vs mechanical respiration. Personally, I can think of other mechanisms to explain the results, such as predictive encoding of respiration-associated outputs in the cingulate area.

"These results suggest that respiration can drive γ oscillations, which further lead to BOLD signal changes in the respiration network"

Reviewer #2 (Recommendations for the authors):

1) The sample size is small -- the study hasn't been pre-registered nor has a formal sample size calculation been provided. This does mean that a number of analyses could have been tried before coming out with this result. With calls for ever increasing numbers of participants in human studies, a sample size of 7 seems difficult to justify.

2) Stronger justification for the use of the ACC as electrode location is essential. The Liotti paper gave CO2 inhalation to healthy humans and reported which areas of the brain showed increased cerebral blood flow, but interpreted the findings in the context of breathlessness. There is a lot more recent work from a range of laboratories done in a much more controlled manner that would justify better the use of the ACC as a component of the respiratory network. There are other areas consistently identified in respiration (e.g. anterior insula, prefrontal cortex etc) that could have been chosen so it's important here not to make the conclusions too ACC specific (in the most part the authors have done this)

3) The effects of anaesthesia are, in my view, a major potential confound. In addition to direct effects on BOLD responsiveness, isoflurane, pentobarbital and (particularly) buprenorphine will depress respiration, raising arterial CO2 and dampening BOLD responsiveness. Ketamine may stimulate respiration doing the opposite. Interaction effects between the different drugs are highly likely and difficult to predict how they would affect BOLD responses.

4) Why was RVT used as the measure of minute ventilation when it would have been possible to take direct measures of breathing via the endotracheal tube? This would be considerably more accurate.

5) Novelty – with such a small sample size it's difficult to know what is truly new. For example Alexander Green's group in Oxford report EEG changes (in α band) in patients with deep brain stimulator electrodes during various respiratory manoeuvres. Of course those studies are also limited by small sample size and diseased patients so comparison is difficult.

Reviewer #3 (Recommendations for the authors):

1) As mentioned in the public review, the statistics of RVT in the isoelectric condition are not reported (unless I have missed it), so it is not clear whether the corresponding decrease in correlation with fMRI could be due to reduced temporal variation in RVT. I would therefore recommend reporting the magnitude of respiratory volume and rate variation and the RVT power spectrum, comparing both the light anesthesia and isoelectric states, in order to determine if this is the case. If RVT is indeed found to be less variable in this condition, it would strengthen the study if the authors are able to perform additional experiments that induce variations in breathing (if that is a possibility).

2) Relatedly, further support could be provided if the authors observe that the isoelectric state selectively disrupts the potential neuronal component (instantaneous fMRI correlation with RVT), but preserves the non-neural CO2 effect on BOLD (assessed by convolving RVT with the respiration response function).

3) To my understanding, RVT was correlated with fMRI at zero time-lag. However, when γ power was correlated with fMRI, it was first convolved with a standard hemodynamic response function. It was not clear to me why RVT was also not convolved with a hemodynamic response function to assess potential neural correlations with fMRI, especially given that it has highest coherence with the γ power signal at zero phase lag.

4) It is shown that the fMRI spatial maps that correlate with ACC γ power resemble the maps of the seed-based correlation from the ACC fMRI signal. For completeness, it would be useful to also report the temporal correlation between the fMRI and γ power signals measured in the ACC.

5) It is mentioned on p.2, line 85, that anesthesia was used to ensure that results are not confounded by motion. Though animals were only lightly sedated, it would be helpful to explain in the discussion whether this level of anesthesia might also impact fMRI and/or electrophysiological signals, to help readers better interpret the findings.

6) One analysis supporting the possible neural origin of RVT correlation in fMRI data is that regressing out neural (γ-band LFP) signals reduced the correlation between RVT and fMRI. This is indeed promising, though I suppose that if there are any non-neural hemodynamic effects of respiration on brain vasculature closer to zero lag (and which therefore also correlate with γ LFP), these would also be reduced when regressing out γ LFP. It would be helpful to include a discussion of whether any non-neural hemodynamic effects in fMRI could occur at zero lag.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Neural underpinning of a respiration-associated resting-state fMRI network" for further consideration by eLife. Your revised article has been evaluated by Timothy Behrens (Senior Editor) and a Reviewing Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

In the Response to Reviewers, the authors present arguments that their anaesthesia protocol has minimal disruption on neurovascular coupling. Unfortunately, though, this does not come through in the discussion added to the paper itself, which focuses on respiration-LFP coupling. Given that the effects of anaesthesia on neurovascular coupling were very specifically raised by Reviewers 2 and 3 (and will be shared by many readers), please add a brief discussion along the lines of the response to Reviewer 2 Comment 3.

Essential revisions:

1) Tempering of language purporting to establish a causal relationship between respiration and the observed γ oscillations.

We appreciate and agree with this suggestion. In the revised manuscript, we modified all related statements, including the title, from establishing a causal relationship to an associative relationship between respiration and neural activity.

2) Comment on the relatively small number of rodents used in the study. [In consultation between the reviewers, it has been pointed out that the reported effects are based on statistical tests driven by the number of timepoints, rather than the number of individual rodents.]

The reviewer’s comment is well accepted. We completely acknowledge that the number of animals in the present study is limited, and the reported effects are dominated by the number of scans/imaging sessions, rather than the number of individual rats. In our revised manuscript, we added statements to discuss this issue.

We would like to further clarify that the number of animals used was based on the conventional sample size of electrophysiology-fMRI experiments, typically in the range of 5-10 animals ^1-5 (7 animals in our study). In electrophysiology recordings, a large number of data points are collected and the results are generally reproducible across animals. To compensate for lower sensitivity/higher variability of fMRI data, unlike conventional fMRI studies, here in each animal we acquired a great number of scans on separate days (total scan number = 99, ~14 scans per animal, 1200 fMRI volumes per scan, 51 scans at the light sedation state and 48 scans at the isoelectric state). This method can significantly reduce the variability of results obtained from individual animals. Below are four figures we added to the revised manuscript demonstrating major results from individual animals. Indeed, all major findings are reproducible across animals, and therefore, we believe this method can significantly mitigate the issue of the small sample size of animals in our study.

3) Discussion of the potential effects of anaesthesia on the study conclusions.

Thanks for the comment. In the revised manuscript, we added a paragraph to discuss the potential effects of anesthesia on our results, quoted below.

“In the current study, we performed the experiment in lightly sedated animals to ensure the results were not confounded by the animal’s motion. Animals’ motion, particularly respiration-correlated motion, can induce systematic variation in the rsfMRI signal that potentially leads to an artefactual correlation pattern. Such a correlation pattern is difficult to be parceled out in our respiration-associated brain network. However, it has to be recognized that anesthesia might be a potential confounder by itself, as it may affect the respiration, as well as respiration-induced neural and vascular activities. For instance, anesthetics can reduce the respiration rate and affect LFP by increasing the low frequency power and decreasing the high frequency power ^6-8. However, previous studies have demonstrated that similar respiration-coupled LFP oscillations were observed in both anesthetized and awake states ^9-11. As the relationship between respiration and neural activity is preserved under anesthesia, our results obtained in anesthetized rats should remain valid. It has to be noted that the neural contributions in the respiration-related network may be different in signal amplitude and/or spatial pattern between the awake and anesthetized states, and other potential neural contributions, such as those mediated by emotional or arousal changes, may only be present in the awake state. Such factors warrant further investigation in awake animal studies.”

4) Placing results in the context of related human studies.

We added more discussions of our results in the context of related human studies, quoted below.

“The findings in the present study to a large extent echo the results reported in human studies. Previous research demonstrated bidirectional interplay between respiration and neural activity in humans. Neural activity in brainstem regions such as the preBötzinger complex directly controls the respiration^12,13. In addition, respiration-induced LFP oscillations were observed in multiple human brain regions such as the piriform cortex and hippocampus^14,15, consistent with our data. McKay and colleagues also found that voluntary hyperpnea was related to the neural activity of various brain regions including the primary sensory and motor cortices, supplementary motor area, caudate nucleus, and globus pallidum, which largely agree with the respiration-related brain network we identified in rats¹⁶. Importantly, respiration-mediated neural activity was found to modulate brain function. For example, respiration is shown to modulate visual perceptual sensitivity, mediated by α power¹⁷. Respiration is also the key component in mindfulness meditation, which has been repeatedly shown to have modulatory effects on brain function¹⁸. Furthermore, the coupling between neural activity and respiration may facilitate our understanding of brain disease. Hyam et al. stimulated the pedunculopontine region (PPNr) and observed improvement in upper airway function in Parkinson’s patients¹⁹. Loss of neurons in the preBötzinger complex was found to cause sleep-disordered breathing²⁰. Taken together, those findings indicate that the respiration-related brain network might be conserved across species and represent a general phenomenon in the mammalian brain, and this brain network might play an important role in normal and diseased brain function.”

5) Analysis of RVT estimates (variance) in the isoelectric state.

We greatly appreciate the suggestion. We calculated the respiration rate (Figure 5D) as well as the standard deviation (SD) and power spectrum of RVT across animals. We did not observe any significant difference in the SD of the respiration rate (Figure 5E, two-sample t-test, t = -1.0, p = 0.32) or the SD of RVT (Figure 5F, two-sample t-test, t = -0.87, p = 0.39) between the light-sedation and isoelectric states, suggesting that the temporal variance of respiration is similar between the two states. Interestingly, we observed a characteristic 1/f pattern in the RVT power during light sedation, but this pattern was not present at the isoelectric state (Figure 5G). This difference is consistent to the ACC BOLD signal power spectra in the light-sedation and isoelectric states (Figure 5I). The new results and figure are added to the revised manuscript.

Reviewer #1 (Recommendations for the authors):

As indicated in the public review, I don't find anything that is lacking in the study. I find it compelling, and I do not wish to burden the authors with additional work that may only remotely nuance their excellent work.

If anything, I am not sure the statement below (and other similar later in the discussion) are sufficiently substantiated. Γ oscillations are correlated with respiration in this study. Other tests to further establish causality would be required, such as comparing spontaneous vs mechanical respiration. Personally, I can think of other mechanisms to explain the results, such as predictive encoding of respiration-associated outputs in the cingulate area.

"These results suggest that respiration can drive γ oscillations, which further lead to BOLD signal changes in the respiration network"

We really appreciate the encouraging comment from the reviewer. The reviewer’s suggestion is well accepted. In the revised manuscript, we modified all the related statements, including the title, from establishing a causal relationship to an associative relationship between respiration and neural activity. Below is the revised statement for the one the reviewer quoted. Other revised statements are highlighted in red in the revised manuscript.

“These results suggest that respiration is associated with γ oscillations, which is necessary for BOLD signal changes in the respiration network.”

Reviewer #2 (Recommendations for the authors):

1) The sample size is small -- the study hasn't been pre-registered nor has a formal sample size calculation been provided. This does mean that a number of analyses could have been tried before coming out with this result. With calls for ever increasing numbers of participants in human studies, a sample size of 7 seems difficult to justify.

Please refer to the response to Comment 2 in the section of Essential revisions from Reviewing Editor.

2) Stronger justification for the use of the ACC as electrode location is essential. The Liotti paper gave CO2 inhalation to healthy humans and reported which areas of the brain showed increased cerebral blood flow, but interpreted the findings in the context of breathlessness. There is a lot more recent work from a range of laboratories done in a much more controlled manner that would justify better the use of the ACC as a component of the respiratory network. There are other areas consistently identified in respiration (e.g. anterior insula, prefrontal cortex etc) that could have been chosen so it's important here not to make the conclusions too ACC specific (in the most part the authors have done this)

We appreciate the reviewer’s suggestion. In the revised manuscript, we added more rationale and references from recent studies supporting the vital role of the ACC in the respiration network, quoted below.

“The ACC was selected given its critical role in respiratory modulation, evidenced by early studies showing that ACC was activated during breathlessness^21,22. More recent research further demonstrates the role of ACC in respiratory control^23-25. For example, electrophysiology recordings in humans showed that ACC exhibited different neural responses to separate respirational tasks including breath-holding, voluntary deep breathing and hypercapnia²⁵. In addition, the ACC is a key region in the rodent default-mode network (DMN) ^26,27, which has been linked to respiration-related fMRI signal changes ²⁸.”

We completely agree with the reviewer that a number of other regions are also involved in the respiration network, and our results are indeed not ACC specific. In fact, the main finding of our paper is that we identified a network of brain regions whose neural activities are associated with respiration (also evident from the title), and ACC is one node of this network, through which the network is identified.

3) The effects of anaesthesia are, in my view, a major potential confound. In addition to direct effects on BOLD responsiveness, isoflurane, pentobarbital and (particularly) buprenorphine will depress respiration, raising arterial CO2 and dampening BOLD responsiveness. Ketamine may stimulate respiration doing the opposite. Interaction effects between the different drugs are highly likely and difficult to predict how they would affect BOLD responses.

We appreciate this valuable comment. Please refer to the response to Comment 3 in the section of Essential revisions from Reviewing Editor for our general response to the potential impact of anesthesia. Below are the more specific responses to the potential effects of different anesthetic drugs the reviewer mentioned.

First, we would like to clarify that we only used the combination of low-dose isoflurane and low-dose dexmedetomidine to induce the light sedation state in animals during simultaneous fMRI, electrophysiology and respiration recordings. Our rationale for choosing this anesthesia is as follows: isoflurane is a vasodilator, which can attenuate the BOLD signal ²⁹, whereas the dexmedetomidine is a vasoconstrictor, which can counteract the vasodilatory effect from isoflurane³⁰. This type of anesthesia has been shown to minimize the confounding effects in the BOLD response and maintain the functional connectivity across both cortical and subcortical regions^31,32. Pentobarbital was used in separate experiments to induce the isoelectric state. This control experiment allows us to determine the necessary role of neural activity in the respiration-related brain network identified in our study.

Ketamine and buprenorphine were not used in any experiment, but only used during the surgery for electrode implantation. The animals were allowed to recover for at least one week after surgery. The drugs should have been fully metabolized by the time of fMRI imaging, so effects from those drugs would not affect our results.

Overall, we agree with the reviewer that anesthesia is a potential confound. In the revised manuscript we have discussed this potential pitfall, quoted below.

“In the current study, we performed the experiment in lightly sedated animals to ensure the results were not confounded by the animal’s motion. Indeed, animals’ motion, particularly respiration-correlated motion, can induce systematic variation in the rsfMRI signal that potentially leads to an artefactual correlation pattern. Such a correlation pattern is difficult to be parceled out in our respiration-associated brain network. However, it has to be recognized that anesthesia might be a potential confounder by itself, as it may affect the respiration, as well as respiration-induced neural and vascular activities. For instance, anesthetics can reduce the respiration rate and affect local field potential by increasing the low-frequency power and decreasing the high-frequency power^6-8. However, previous studies demonstrated that similar respiration-coupled LFP oscillations were observed in both anesthetized and awake states^9-11. As the relationship between respiration and neural activity is preserved under anesthesia, our results obtained in anesthetized rats should remain valid. Notably, the neural contributions in the respiration-related network may be different in signal amplitude and/or spatial pattern between the awake and anesthetized states, and other potential neural contributions, mediated by emotional or arousal changes, are only present in awake state. Such factors warrant further investigation in future awake animal studies.”

4) Why was RVT used as the measure of minute ventilation when it would have been possible to take direct measures of breathing via the endotracheal tube? This would be considerably more accurate.

We agree with the reviewer that respiration can also be measured via the endotracheal tube in animals that were ventilated in the isoelectric state. However, the respiration in animals during light sedation, which were not intubated, can only be measured using the respiration pads. As we need to maintain a consistent way to measure respiration across conditions for comparison, we used the respiration pad in our study. This method also provides an accurate measurement of respiration and RVT ^33,34.

5) Novelty – with such a small sample size it's difficult to know what is truly new. For example Alexander Green's group in Oxford report EEG changes (in α band) in patients with deep brain stimulator electrodes during various respiratory manoeuvres. Of course those studies are also limited by small sample size and diseased patients so comparison is difficult.

The major novelty of the present study is the finding of a network of brain regions whose neural activities are associated with respiration. We achieved this goal by simultaneously recording electrophysiology, whole-brain fMRI, and respiration signals. This is distinct from most other human (or animal) studies that only focused on the role of a (or a few) specific brain region(s). For instance, the study by Alexander Green's group mentioned by reviewer ¹⁹ found that stimulation of the pedunculopontine region (PPNr) improved upper airway function in parkinson’s disease patients. The α power in PPNr increased during forced respiratory maneuvers. Their study is important for understanding the potential role of neural activity of one region (PPNr) in respiration control, which is very different from our study.

In the revised manuscript, we added the discussion for our results in relation to human studies, quoted below:

“The findings in the present study to a large extent echo the results reported in human studies. Previous research demonstrated bidirectional interplay between respiration and neural activity in humans. Neural activity in brainstem regions such as the preBötzinger complex directly controls the respiration^12,13. In addition, respiration-induced LFP oscillation was observed in multiple human brain regions such as the piriform cortex and hippocampus^14,15, which is consistent with our data. McKay and colleagues also found that voluntary hyperpnea was related to the neural activity of various brain regions including the primary sensory and motor cortices, supplementary motor area, caudate nucleus, and globus pallidum, which largely agreeing with the respiration-related brain network we identified in rats¹⁶. Importantly, respiration-mediated neural activity was found to modulate brain function. For example, respiration is shown to modulate visual perceptual sensitivity, mediated by α power¹⁷. Respiration is also the key component in mindfulness meditation, which has been repeatedly shown to have modulatory effects on brain function ¹⁸. Furthermore, the coupling between neural activity and respiration may facilitate our understanding of brain disease. Hyam et al. stimulated the pedunculopontine region (PPNr) and observed improvement in upper airway function in Parkinson’s patients ¹⁹. Loss of neurons in the preBötzinger complex was found to cause sleep-disordered breathing ²⁰. Taken together, those findings indicate that the respiration-related brain network might be conserved across species and represent a general phenomenon in the mammalian brain, and this brain network might play an important role in normal and diseased brain function.”

Reviewer #3 (Recommendations for the authors):

1) As mentioned in the public review, the statistics of RVT in the isoelectric condition are not reported (unless I have missed it), so it is not clear whether the corresponding decrease in correlation with fMRI could be due to reduced temporal variation in RVT. I would therefore recommend reporting the magnitude of respiratory volume and rate variation and the RVT power spectrum, comparing both the light anesthesia and isoelectric states, in order to determine if this is the case. If RVT is indeed found to be less variable in this condition, it would strengthen the study if the authors are able to perform additional experiments that induce variations in breathing (if that is a possibility).

Please refer to the response to Comment 5 in the section of Essential Revisions from Reviewing Editor.

2) Relatedly, further support could be provided if the authors observe that the isoelectric state selectively disrupts the potential neuronal component (instantaneous fMRI correlation with RVT), but preserves the non-neural CO2 effect on BOLD (assessed by convolving RVT with the respiration response function).

This is a very interesting point. The non-neural CO2 effect on the BOLD signal in the isoelectric state, estimated by convolving RVT with the respiration response function and then voxel-wise correlating to the BOLD signal, appeared to be weaker than the light-sedation state (Figure 4 —figure supplement 3), as shown below (Author response image 1, left). This is likely attributed to the fact that the metabolic rate was much lower in the isoelectric state (approximately half) ³⁵, resulting in weaker fluctuations in the CO2 level during respiration, which can explain less appreciable CO2 effect on the BOLD signal in the isoelectric state. However, statistical comparison of the respiration artifactual pattern between the light-sedation and isoelectric states did not reveal any significant difference (Author response image 1, middle, t-test, threshold p = 0.05, FDR corrected), in sharp contrast to the statistical comparison of the respiration network pattern between the light-sedation and isoelectric states (Author response image 1, right, t-test, p < 0.05, FDR corrected). These data indicate that the artefactual effects of respiration were more or less similar between the two states, but the respiration-related brain network activity was substantially suppressed at the isoelectric state.

Author response image 1

Download asset Open asset

3) To my understanding, RVT was correlated with fMRI at zero time-lag. However, when γ power was correlated with fMRI, it was first convolved with a standard hemodynamic response function. It was not clear to me why RVT was also not convolved with a hemodynamic response function to assess potential neural correlations with fMRI, especially given that it has highest coherence with the γ power signal at zero phase lag.

The hemodynamic response function (HRF) referred herein by definition represents the transfer function between neural activity and fMRI signal. It is effectively a low-pass filter that transfers fast neural activity to slower fMRI signal. Because HRF is not the transfer function between RVT and fMRI signal, it would not be appropriate to convolve RVT with HRF. In fact, as the RVT itself represents low-frequency fluctuations of respiration (the peak coherence between RVT and γ power is at a low frequency (0.035Hz)), convolving RVT with HRF should not significantly alter the RVT waveform. Indeed, in Author response image 2 we show the voxel-wise correlation pattern between rsfMRI signal and the time course after convolving RVT with HRF. The pattern is very similar to the RVT-rsfMRI correlation pattern we obtained (Figure 4B).

Author response image 2

Download asset Open asset

4) It is shown that the fMRI spatial maps that correlate with ACC γ power resemble the maps of the seed-based correlation from the ACC fMRI signal. For completeness, it would be useful to also report the temporal correlation between the fMRI and γ power signals measured in the ACC.

Thanks for your suggestion, we have reported the correlation between fMRI signal of ACC and γ power (r = 0.086, p = 0.003) in the revised manuscript.

5) It is mentioned on p.2, line 85, that anesthesia was used to ensure that results are not confounded by motion. Though animals were only lightly sedated, it would be helpful to explain in the discussion whether this level of anesthesia might also impact fMRI and/or electrophysiological signals, to help readers better interpret the findings.

Please refer to our response to Comment 3 in the section of Essential Revisions from Reviewing Editor, as well as our response to Comment 3 for Reviewer #2 for the issue of the potential impact of anesthesia.

6) One analysis supporting the possible neural origin of RVT correlation in fMRI data is that regressing out neural (γ-band LFP) signals reduced the correlation between RVT and fMRI. This is indeed promising, though I suppose that if there are any non-neural hemodynamic effects of respiration on brain vasculature closer to zero lag (and which therefore also correlate with γ LFP), these would also be reduced when regressing out γ LFP. It would be helpful to include a discussion of whether any non-neural hemodynamic effects in fMRI could occur at zero lag.

This is an interesting point. Although we cannot think of any non-neural hemodynamic effects of respiration on the rsfMRI signal that could occur at zero lag, we do agree with the reviewer that regressing out γ LFP will reduce these effects. We have discussed this possibility in our revised manuscript, quoted below.

“This notion is further supported by our data that regressing out γ-band LFP signals tremendously reduced the correlation between RVT and fMRI (Figure 4D-E), although it is still possible that if there are non-neural hemodynamic effects of respiration on the rsfMRI signal that occur at zero lag (i.e. a non-neural hemodynamic factor that are synchronized with γ LFP), regressing out γ LFP could also reduce these effects.”

References

1. Angenstein, F. The role of ongoing neuronal activity for baseline and stimulus-induced BOLD signals in the rat hippocampus. Neuroimage 202, 116082, doi:10.1016/j.neuroimage.2019.116082 (2019).

2. Baek, K. et al. Layer-specific interhemispheric functional connectivity in the somatosensory cortex of rats: resting state electrophysiology and fMRI studies. Brain Struct Funct 221, 2801-2815, doi:10.1007/s00429-015-1073-0 (2016).

3. Bovet-Carmona, M. et al. Disentangling the role of TRPM4 in hippocampus-dependent plasticity and learning: an electrophysiological, behavioral and FMRI approach. Brain Struct Funct 223, 3557-3576, doi:10.1007/s00429-018-1706-1 (2018).

4. Jaime, S., Cavazos, J. E., Yang, Y. and Lu, H. Longitudinal observations using simultaneous fMRI, multiple channel electrophysiology recording, and chemical microiontophoresis in the rat brain. J Neurosci Methods 306, 68-76, doi:10.1016/j.jneumeth.2018.05.010 (2018).

5. Thompson, G. J., Pan, W. J. and Keilholz, S. D. Different dynamic resting state fMRI patterns are linked to different frequencies of neural activity. J Neurophysiol 114, 114-124, doi:10.1152/jn.00235.2015 (2015).

6. Purdon, P. L. et al. Electroencephalogram signatures of loss and recovery of consciousness from propofol. Proc Natl Acad Sci U S A 110, E1142-1151, doi:10.1073/pnas.1221180110 (2013).

7. You, T., Im, G. H. and Kim, S. G. Characterization of brain-wide somatosensory BOLD fMRI in mice under dexmedetomidine/isoflurane and ketamine/xylazine. Sci Rep 11, 13110, doi:10.1038/s41598-021-92582-5 (2021).

8. Bastos, A. M. et al. Neural effects of propofol-induced unconsciousness and its reversal using thalamic stimulation. ELife 10, doi:10.7554/eLife.60824 (2021).

9. Lockmann, A. L., Laplagne, D. A., Leao, R. N. and Tort, A. B. A Respiration-Coupled Rhythm in the Rat Hippocampus Independent of Theta and Slow Oscillations. J Neurosci 36, 5338-5352, doi:10.1523/JNEUROSCI.3452-15.2016 (2016).

10. Nguyen Chi, V. et al. Hippocampal Respiration-Driven Rhythm Distinct from Theta Oscillations in Awake Mice. J Neurosci 36, 162-177, doi:10.1523/JNEUROSCI.2848-15.2016 (2016).

11. Yanovsky, Y., Ciatipis, M., Draguhn, A., Tort, A. B. and Brankack, J. Slow oscillations in the mouse hippocampus entrained by nasal respiration. J Neurosci 34, 5949-5964, doi:10.1523/JNEUROSCI.5287-13.2014 (2014).

12. Baertsch, N. A., Bush, N. E., Burgraff, N. J. and Ramirez, J. M. Dual mechanisms of opioid-induced respiratory depression in the inspiratory rhythm-generating network. ELife 10, doi:10.7554/eLife.67523 (2021).

13. Smith, J. C., Ellenberger, H. H., Ballanyi, K., Richter, D. W. and Feldman, J. L. Pre-Botzinger complex: a brainstem region that may generate respiratory rhythm in mammals. Science 254, 726-729, doi:10.1126/science.1683005 (1991).

14. Herrero, J. L., Khuvis, S., Yeagle, E., Cerf, M. and Mehta, A. D. Breathing above the brain stem: volitional control and attentional modulation in humans. J Neurophysiol 119, 145-159, doi:10.1152/jn.00551.2017 (2018).

15. Zelano, C. et al. Nasal Respiration Entrains Human Limbic Oscillations and Modulates Cognitive Function. J Neurosci 36, 12448-12467, doi:10.1523/JNEUROSCI.2586-16.2016 (2016).

16. McKay, L. C., Evans, K. C., Frackowiak, R. S. and Corfield, D. R. Neural correlates of voluntary breathing in humans. J Appl Physiol (1985) 95, 1170-1178, doi:10.1152/japplphysiol.00641.2002 (2003).

17. Kluger, D. S., Balestrieri, E., Busch, N. A. and Gross, J. Respiration aligns perception with neural excitability. ELife 10, doi:10.7554/eLife.70907 (2021).

18. Doll, A. et al. Mindful attention to breath regulates emotions via increased amygdala-prefrontal cortex connectivity. Neuroimage 134, 305-313, doi:10.1016/j.neuroimage.2016.03.041 (2016).

19. Hyam, J. A. et al. The pedunculopontine region and breathing in Parkinson's disease. Ann Clin Transl Neurol 6, 837-847, doi:10.1002/acn3.752 (2019).

20. McKay, L. C., Janczewski, W. A. and Feldman, J. L. Sleep-disordered breathing after targeted ablation of preBotzinger complex neurons. Nat Neurosci 8, 1142-1144, doi:10.1038/nn1517 (2005).

21. Liotti, M. et al. Brain responses associated with consciousness of breathlessness (air hunger). Proc Natl Acad Sci U S A 98, 2035-2040, doi:10.1073/pnas.98.4.2035 (2001).

22. Evans, K. C. et al. BOLD fMRI identifies limbic, paralimbic, and cerebellar activation during air hunger. J Neurophysiol 88, 1500-1511, doi:10.1152/jn.2002.88.3.1500 (2002).

23. Evans, K. C. et al. Modulation of spontaneous breathing via limbic/paralimbic-bulbar circuitry: an event-related fMRI study. Neuroimage 47, 961-971, doi:10.1016/j.neuroimage.2009.05.025 (2009).

24. Tort, A. B. L. et al. Parallel detection of theta and respiration-coupled oscillations throughout the mouse brain. Sci Rep 8, 6432, doi:10.1038/s41598-018-24629-z (2018).

25. Holton, P. et al. Differential responses to breath-holding, voluntary deep breathing and hypercapnia in left and right dorsal anterior cingulate. Exp Physiol 106, 726-735, doi:10.1113/EP088961 (2021).

26. Lu, H. et al. Rat brains also have a default mode network. Proc Natl Acad Sci U S A 109, 3979-3984, doi:1200506109 [pii]10.1073/pnas.1200506109 (2012).

27. Tu, W., Ma, Z., Ma, Y., Dopfel, D. and Zhang, N. Suppressing Anterior Cingulate Cortex Modulates Default Mode Network and Behavior in Awake Rats. Cereb Cortex 31, 312-323, doi:10.1093/cercor/bhaa227 (2021).

28. Birn, R. M., Diamond, J. B., Smith, M. A. and Bandettini, P. A. Separating respiratory-variation-related fluctuations from neuronal-activity-related fluctuations in fMRI. Neuroimage 31, 1536-1548, doi:10.1016/j.neuroimage.2006.02.048 (2006).

29. Schwinn, D. A., McIntyre, R. W. and Reves, J. G. Isoflurane-induced vasodilation: role of the α-adrenergic nervous system. Anesth Analg 71, 451-459, doi:10.1213/00000539-199011000-00001 (1990).

30. Talke, P. and Anderson, B. J. Pharmacokinetics and pharmacodynamics of dexmedetomidine-induced vasoconstriction in healthy volunteers. Br J Clin Pharmacol 84, 1364-1372, doi:10.1111/bcp.13571 (2018).

31. Grandjean, J., Schroeter, A., Batata, I. and Rudin, M. Optimization of anesthesia protocol for resting-state fMRI in mice based on differential effects of anesthetics on functional connectivity patterns. Neuroimage 102 Pt 2, 838-847, doi:10.1016/j.neuroimage.2014.08.043 (2014).

32. Benveniste, H. et al. Anesthesia with Dexmedetomidine and Low-dose Isoflurane Increases Solute Transport via the Glymphatic Pathway in Rat Brain When Compared with High-dose Isoflurane. Anesthesiology 127, 976-988, doi:10.1097/ALN.0000000000001888 (2017).

33. Gass, N. et al. Brain network reorganization differs in response to stress in rats genetically predisposed to depression and stress-resilient rats. Transl Psychiatry 6, e970, doi:10.1038/tp.2016.233 (2016).

34. Baria, A. T. et al. BOLD temporal variability differentiates wakefulness from anesthesia-induced unconsciousness. J Neurophysiol 119, 834-848, doi:10.1152/jn.00714.2017 (2018).

35. Du, F. et al. Tightly coupled brain activity and cerebral ATP metabolic rate. Proc Natl Acad Sci U S A 105, 6409-6414, doi:10.1073/pnas.0710766105 (2008).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

In the Response to Reviewers, the authors present arguments that their anaesthesia protocol has minimal disruption on neurovascular coupling. Unfortunately, though, this does not come through in the discussion added to the paper itself, which focuses on respiration-LFP coupling. Given that the effects of anaesthesia on neurovascular coupling were very specifically raised by Reviewers 2 and 3 (and will be shared by many readers), please add a brief discussion along the lines of the response to Reviewer 2 Comment 3.

We apologize for missing the discussion of the anesthesia protocol in the main text. The following statements in our original response to Reviewer 2 Comment 3 have been added to the revised manuscript.

“Our rationale for choosing combined low-dose isoflurane and dexmedetomidine as our anesthesia protocol is as follows: Isoflurane is a vasodilator, which can attenuate the BOLD signal⁷⁵, whereas the dexmedetomidine is a vasoconstrictor, which can counteract the vasodilatory effect from isoflurane⁷⁶. This anesthesia protocol has been shown to minimize the confounding effects on the BOLD response and maintain the functional connectivity across both cortical and subcortical regions^77,78.”

Author details

1. Wenyu Tu

   1. The Neuroscience Graduate Program, The Huck Institutes of the Life Sciences, The Pennsylvania State University, University Park, United States
   2. Center for Neurotechnology in Mental Health Research, The Pennsylvania State University, University Park, United States

   Contribution

   Data curation, Formal analysis, Validation, Investigation, Methodology, Writing – original draft

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-3480-2098

2. Nanyin Zhang

   1. The Neuroscience Graduate Program, The Huck Institutes of the Life Sciences, The Pennsylvania State University, University Park, United States
   2. Center for Neurotechnology in Mental Health Research, The Pennsylvania State University, University Park, United States
   3. Department of Biomedical Engineering, The Pennsylvania State University, University Park, United States

   Contribution

   Conceptualization, Supervision, Funding acquisition, Validation, Investigation, Writing – original draft, Project administration, Writing – review and editing

   For correspondence

   nuz2@psu.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-5824-9058

• 2,421

  Page views

• 378

  Downloads

• 5

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.