Tracing neuronal synchronized slow oscillations with simultaneous fMRI and optical imaging in mice

Poster No:

1581 

Submission Type:

Abstract Submission 

Authors:

Wen-Ju Pan1, Lauren Daley1, Harrison Watters1, Lisa Meyer-Baese1, Shella Keilholz1

Institutions:

1Emory University/Georgia Institute of Technology, Atlanta, GA

First Author:

Wen-Ju Pan  
Emory University/Georgia Institute of Technology
Atlanta, GA

Co-Author(s):

Lauren Daley  
Emory University/Georgia Institute of Technology
Atlanta, GA
Harrison Watters  
Emory University/Georgia Institute of Technology
Atlanta, GA
Lisa Meyer-Baese  
Emory University/Georgia Institute of Technology
Atlanta, GA
Shella Keilholz  
Emory University/Georgia Institute of Technology
Atlanta, GA

Introduction:

Resting state fMRI functional connectivity (FC) has been widely used to study the functional network organization of the brain with the assumption that it reflects underlying slow synchronized neuronal activity 1-3. However, little is known about the correspondence between fMRI FC and the spatial and temporal properties of the slow neuronal oscillations due to the challenges of obtaining of concurrent neuronal FC during fMRI4. In our studies, we combined a conventional wide-field cortical optical imaging method with ZTE fMRI in mice to investigate the concurrent neuronal fluorescent signals across cortical areas during whole brain fMRI. Using our setup, the FC of neuronal calcium signals (GCaMP) or membrane voltage signals (VSFPB or JEDI) from infraslow bands to 25Hz were examined from one side of the upper limb area during fMRI. The concurrent BOLD EPI signals or ZTE signals and FCs were evaluated from different cortical layers.

Methods:

Five mice were examined under two anesthesia conditions, (1% isoflurane (iso) and 0.5% iso mixed with 0.05mg/kg/hr Dexmedetomidine (iso-dex)) with simultaneous fMRI and optical imaging setup, 3 with neuronal calcium fluorescence (GCaMP 6f) and 2 with neuronal voltage indicators (VSFPB or JEDI). All mice were imaged on the recently upgraded Bruker BioSpin 9.4T scanner with AVANCE NEO console and Paravision360 v3.4. ZTE fMRI was set with ~ 345um isotropic voxels and whole brain coverage and temporal sampling rate of 2s per brain volume scan. TR: 0.673 ms, flip angle 3.7°, bandwidth 187.5 kHz, oversampling 4, matrix size 72 × 72 × 72, field-of-view 25 × 25 × 25 mm3, polar under sampling factor 5.64, and number of projections 2460. The 10 min resting state scan sessions were conducted in iso or iso-dex anesthesia. The optical imaging system was designed to fit inside the magnet with a long tube lens5. The wavelength was 466/40nm for fluorescence excitation, and the emitted 525/50nm were detected in a camera at 50Hz in 16bit with image matrix 110 by 110. Meanwhile, the green light pulses were illuminated alternatively at 50Hz and the reflection signals were detected with the same camera. The raw fluorescent signals were purified to represent neuronal signals by regression of simultaneous green reflection imaging of total hemoglobin absorption. All data were preprocessed and registered to Allen mouse atlas for FC analysis from left side upper limb area, Figure 1.
Supporting Image: fig1.PNG
 

Results:

Neuronal FCs across hemisphere were detected in slow frequency bands but not in high frequencies > 1Hz or >0.5Hz (example in Figure 2). The neuronal voltage FCs exhibited relatively narrow frequency bands, 0.1-0.5 Hz, comparing with calcium FCs, 0.01-1 Hz. The fMRI FCs were evaluated over the conventional 0.01-0.1Hz band for different layers of left upper limb area, layers 1, 2/3, 4, 5 and 6. The localized FCs observed mostly in top and middle layers across scans of ZTE fMRI with varied anesthesia conditions.
Supporting Image: fig2.PNG
 

Conclusions:

In this report, we demonstrated the concurrent measurement of neuronal and fMRI FCs in anesthetized mice. Under both anesthesia conditions, iso or iso-dex, the neuronal FCs came from slow oscillations but not high frequencies. The neuronal voltage FCs may be in a relatively narrow band of the slow oscillations. The neuronal calcium FCs of the slow oscillations included infraslow bands additionally. The corresponding fMRI FCs in a primary somatosensory cortex, were mostly from top and middle layers. The present methods and the initial examination may pave a way for a wide application in the future.

Modeling and Analysis Methods:

Connectivity (eg. functional, effective, structural) 1

Novel Imaging Acquisition Methods:

Non-BOLD fMRI 2

Physiology, Metabolism and Neurotransmission :

Neurophysiology of Imaging Signals

Keywords:

Cortical Layers
ELECTROPHYSIOLOGY
fMRI CONTRAST MECHANISMS
FUNCTIONAL MRI
Neuron
Optical Imaging Systems (OIS)
Other - ZTE

1|2Indicates the priority used for review

Provide references using author date format

1. Li, J. M., Bentley, W. J., Snyder, A. Z., Raichle, M. E. & Snyder, L. H. Functional connectivity arises from a slow rhythmic mechanism. Proceedings of the National Academy of Sciences of the United States of America 112, E2527–E2535 (2015).
2. Pan, W. J., Thompson, G. J., Magnuson, M. E., Jaeger, D. & Keilholz, S. Infraslow LFP correlates to resting-state fMRI BOLD signals. NeuroImage 74, 288–297 (2013).
3. Schwalm, M. Cortex-wide BOLD fMRI activity reflects locally-recorded slow oscillation-associated calcium waves. eLife 6, e27602 (2017).
4. Lake, E. M. R. et al. Simultaneous cortex-wide fluorescence Ca2+ imaging and whole-brain fMRI. Nature Methods 17, 1262–1271 (2020).
5. Pan, W.-J. et al. (ISMRM 2022) Optimization of wide-field optical imaging method towards fMRI integration in mice. https://archive.ismrm.org/2022/3331.html.