Histological characterization of ChR2::CaMKIIα viral expression in the olfactory bulb (OB) excitatory neurons and OB afferents at the anterior olfactory nucleus (AON) and the piriform cortex (Pir), and optogenetic fMRI stimulation setup

(A) Confocal images, 10x magnification (Left) and 40x magnification (Right), of ChR2-mCherry expression in histological slices covering OB, AON and Pir. Overlay of images co-stained for the nuclear marker DAPI and mCherry in OB (Top) revealed ChR2 expression in the soma of the mitral cell layer & external plexiform layer in OB (indicated by green arrows). Meanwhile, ChR2 expressions were found in OB afferents at AON and Pir (Middle & Bottom), indicated by no colocalization between mCherry and DAPI. (B) Illustration of three optogenetic stimulation targets, namely the OB excitatory neurons, and OB afferents in AON and Pir (Left), which were conducted in separate animal groups. T2-weighted anatomical MRI image showing the location of the implanted optical fiber (asterisk: stimulation target). Optogenetic fMRI block-designed stimulation paradigm (Right). Five different frequencies (1, 5, 10, 20, and 40 Hz) were presented in a pseudorandomized order within each session. We repeated the stimulations for a total of three sessions in each animal fMRI experiment. All frequencies were presented at a 30% duty cycle except for 1 Hz, which was presented at a 10% duty cycle.

Optogenetic excitation of OB excitatory neurons and OB afferents at AON and Pir recruit distinct brain-wide long-range olfactory networks

Averaged BOLD activation maps and signal profiles (Left) upon 1 Hz stimulation of (A) OB excitatory neurons, (B) OB afferents in AON, and (C) OB afferents in Pir during the first fMRI session (n = 11 for fMRI experiments in A & B; n = 9 for C; t > 3.1 corresponding to P < 0.001; error bars indicate ± SEM), and the corresponding area under the BOLD signal profiles (AUC) for each atlas-defined region-of-interests (ROIs; Right). Note that the BOLD activation maps displayed in A-C were further corrected for multiple comparisons with threshold-free cluster enhancement with family-wise error rate (TFCE-FWE) at P < 0.05. (D) Summary of BOLD activation strength to quantify the extent of various brain networks engaged upon respective optogenetic stimulation of the three different primary olfactory stimulation targets. BOLD activation strength was computed as the ratio of AUC for each ROI over AUC of all ROIs (individual animal data points are shown; error bars indicate ± SEM; Tukey’s multiple comparisons test; **P < 0.01, and *P < 0.05). OB-driven neural activities mainly activated the primary olfactory network (AON, Pir, TT, Ent and Tu). AON-driven neural activities were found to preferentially recruit the striatal (NAc and vCPu) and hippocampal networks (vHP), while Pir-driven activities strongly targeted the limbic network (Cg, PrL, OFC, Ins and Amg). Abbreviations. Regions in primary olfactory network: olfactory bulb (OB), anterior olfactory nucleus (AON), piriform cortex (Pir), tenia tecta (TT), entorhinal cortex (Ent), olfactory tubercle (Tu); limbic network: cingulate (Cg), prelimbic cortex (PrL), orbitofrontal cortex (OFC), insular (Ins), amygdala (Amg); hippocampal network: ventral hippocampus (vHP); striatal network: nucleus accumbens (NAc), ventral caudate putamen (vCPu); sensorimotor network: motor (MC), somatosensory (S1), auditory (AC), visual (V1) cortex.

Repetitive optogenetic excitation of OB excitatory neurons and OB afferents at AON and Pir reveal varied neural activity adaptation properties in olfactory networks

Averaged BOLD activation maps (Top) upon 1 Hz stimulation of (A) OB excitatory neurons, (B) OB afferents in AON, and (C) OB afferents in Pir during the second (n = 11 for fMRI experiments in A & B; n = 9 for C; t > 3.1 corresponding to P < 0.001) and third (n = 8 for fMRI experiments in A; n = 9 for B; n = 6 for C; t > 3.1 corresponding to P < 0.001) fMRI sessions, respectively. We found a dramatic decrease or even absent bilateral activations in long-range olfactory networks upon stimulation of OB excitatory neurons or OB afferents at AON, but not Pir, indicating strong neural activity adaptation mediated by AON-driven neural activity to downstream targets. Note that similar to Fig. 2, the BOLD activation maps displayed in A-C were further corrected for multiple comparisons with threshold-free cluster enhancement with family-wise error rate (TFCE-FWE) at P < 0.05, except for that the maps during the third session in B and C were generated directly after one-sample t-test. The extent of neural activity adaptation in brain-wide olfactory networks was quantified with the true BOLD power (AUC of the BOLD signal profiles) for each atlas-defined ROI (Bottom; individual animal data points are shown; error bars indicate ± SEM; Dunnett’s multiple comparisons test of BOLD power during 2nd vs. 1st fMRI session, 3rd vs. 1st session; ***P < 0.001; **P < 0.01; *P < 0.05). Note that the AUC for each ROI was normalized to their respective counterpart in the first fMRI session for display.

Inhibitory effect of AON outputs and excitatory effect of Pir outputs to downstream targets in long-range olfactory networks

(A) Graphical representation of matrix A which defines the a priori anatomical connections of the olfactory network model utilized in dynamic causal modeling (DCM). (B) Effective connectivity matrices of the olfactory network model upon stimulation of OB excitatory neurons and their afferents at AON and Pir. (C) Quantitative summary of the significant differences in connectivity strength of downstream connections identified from the statistical comparison of matrices in B (Left; n = 11 for OB and AON stimulations; n = 9 for Pir stimulation; individual animal data points are shown; error bars indicate ± SEM; paired t-test with FDR correction; **P < 0.01, and *P < 0.05) and the corresponding graphical representation of inhibitory effect of AON and excitatory effect of Pir outputs to downstream targets (Right). We identified robust AON negative vs. Pir positive connectivity to downstream targets such as Ent, vCPu and Amg, indicating distinct inhibitory vs. excitatory effect of the two primary olfactory cortical outputs that can shape the dynamic properties (i.e., BOLD activations across repeated stimulation sessions) of long-range olfactory networks.

Local field potential (LFP) recordings confirm neuronal activity underlying BOLD activations and reveal orthodromic neural activity propagation to downstream targets in primary olfactory, striatal, visual, limbic and hippocampal networks

Illustration of electrophysiological recording regions (Top-Left), averaged evoked LFPs from the seven recorded regions (Bottom-Left & Middle), and summary of orthodromic neural activity propagation latencies measured from the stimulation onset over mono- and poly-synaptic projections (Right) upon 1 Hz optogenetic stimulation of (A) OB excitatory neurons, (B) OB afferents at AON and (C) OB afferents at Pir (n = 7, n = 7 and n = 5, respectively; 10% duty cycle; error bars indicate ± SEM; crosshairs denote LFP peaks that represent initial neuronal population activity for latency measurements). Significant delay between AON/Pir and OB neuronal population activity with appreciable jitter showed that neural activity propagation through the excitations of OB axonal terminals at AON or Pir, respectively, were largely orthodromic in our experiment.

LFP recordings confirm varied neural activity adaptation in olfactory networks upon repetitive optogenetic excitations of OB excitatory neurons and OB afferents at AON and Pir

LFP power of evoked neural activities upon 1 Hz optogenetic stimulation of (A) OB excitatory neurons (n = 7; 10% duty cycle; error bars indicate ± SEM), (B) OB afferents at AON (n = 7; 10% duty cycle; error bars indicate ± SEM) and (C) OB afferents at Pir (n = 5; 10% duty cycle; error bars indicate ± SEM) across three electrophysiological recordings sessions. (D) Summary of the area under LFP power to quantify the extent of neural activity adaptation in the recorded regions of olfactory networks (individual animal data points are shown). In accord with our fMRI findings, the strength of neural activities, which were quantified through LFP power analysis here (i.e., area under absolute LFP trace normalized to 30-s baseline LFP power), was also decreased under repetitive 1 Hz stimulation of OB and OB afferents at AON but was relatively unchanged under Pir stimulation (1st vs 2nd and 3rd recording sessions). The LFP recordings demonstrate strong neural activity adaptation mediated by AON-driven neural activities to downstream targets.

Decreased BOLD activations in primary olfactory and limbic networks, and impaired AON to Pir connectivity in aged animals

(A) Averaged BOLD activation maps and signal profiles from the 1st fMRI session upon 1 Hz stimulation of OB excitatory neurons in D-galactose aged rat model (n = 13; t > 3.1 corresponding to P < 0.001; error bars indicate ± SEM). Note that the BOLD activation maps displayed in A were further corrected for multiple comparisons with TFCE-FWE at P < 0.05. (B) Comparison of the corresponding area under the BOLD signal profiles for each atlas-defined region-of-interests between healthy and aged animals (individual animal data points are shown; Fisher’s LSD test; **P < 0.01, and *P < 0.05; error bars indicate ± SEM). (C) Effective connectivity matrices of the olfactory network model upon stimulation of OB excitatory neurons in aged brains. (D) Quantitative summary of the significant differences in connectivity strength of downstream connections identified from the statistical comparison of effective connectivity matrices between healthy and aged brains (individual animal data points are shown; unpaired t-test; *P < 0.05; error bars indicate ± SEM). We uncovered a switch in the polarity from positive to negative in the connectivity of AON neural activity outputs to Pir in aged animals, indicating the dysfunction of a key primary olfactory cortical circuit that likely caused the significantly decreased BOLD activations in the primary olfactory and limbic networks.