Fig 1.
Following a first fMRI scan at day 0 (baseline), animals were injected s.c. with increasing doses (30, 70, 100 μg/kg) of RTX on three consecutive days. Eight days following the last injection, a second fMRI scan (RTX) was performed, to assess the effects of desensitization of TRPV1-expressing neurons on central temperature processing. Each fMRI scan included two resting state sessions, one before (RSpre) and one after (RSpost) the thermal stimulation. During the stimulation period, the contact heat stimuli sequence was presented at the dorsal side of the left hind paw of the animal. Temperatures ranging from 40°C to 54°C were arranged in a pseudo-randomized fashion per set, and were held for 20 sec each followed by 3 min 40 sec rest. Three different sets were presented during the 100 min scan. This protocol was re-used for every animal.
Fig 2.
Whole brain BOLD response amplitude and activated brain volume evoked by different temperatures.
Depicted are single animal values (dots) as well as mean (dotted lines) BOLD response amplitude (a) and activated volume (b) over all brain areas evoked by innocuous (light grey box), noxious (middle grey box) and highly noxious (dark grey box) temperatures for both groups. Whereas the untreated animals (baseline, red) showed increasing brain responses well correlated to the linear temperature increase (rAmplitude = 0.89, rVolume = 0.88; also see S1 Fig, correlation coefficient), RTX-treated rats (blue) failed to show increased responses to temperatures between 48°C and 52°C (rAmplitude = -0.04, rVolume = 0.03). With 54°C, a slight but insignificant (p52vs54°C = 0.067, paired Student’s t-test) increase could be noted for the response amplitude. Activated brain volume is represented as percentage of total brain volume per animal. Statistical significance between groups was calculated using homoscedastic Student’s t-test and corrected for multiple comparisons by FDR q = 0.05. Boxplots span the 25% to the 75% quartile (median included). A horizontal line marks the median. The whiskers represent minimum and maximum values. Values below 1.5 times the interquartile range from minimum and above 1.5 times the interquartile range from maximum are considered as outliers (triangles). Means of each group are connected by a dotted line. * p≤0.05; ** p≤ 0.01; *** p≤ 0.001. (nBaseline = 18; nRTX = 9).
Fig 3.
Significant differences in the whole brain maximal BOLD response amplitude evoked by different temperatures before and after RTX-treatment.
The voxels in red-yellow showed greater activation at baseline, whereas the blue-green voxels showed greater activation after RTX treatment. Only very small differences (raw data showed comparable activation for both groups, data not shown) could be shown for temperatures ≤ 48°C, indicating that processing of innocuous temperatures remained unaffected by RTX and the resulting desensitization of TRPV1-expressing neurons. However, baseline showed significantly higher activation for noxious temperature above 48°C in wide areas of the brain (subcortical and cortical). Data are represented as percentage change of maximal BOLD response amplitude. Only voxels with p-values ≤ 0.05 (FWE permutation corrected) are shown. (nBaseline = 18; nRTX = 9).
Fig 4.
Mean BOLD response amplitudes for representative functional groups.
Desensitization of TRPV1-expressing neurons (RTX) had a widespread effect on both sensory-discriminative and emotional-affective brain regions, as well as on some output structures. No or only minor differences between groups were found for brain structures not involved in stimulus-driven tasks, such as the pons, the tegmentum and the output structure hypothalamus, which instead showed significant differences for the activated brain volume, indicating that many voxels were activated at baseline, but with low amplitude (S2 Fig). Statistical significance between groups was calculated using homoscedastic Student’s t-test and corrected for multiple comparisons by FDR q = 0.05. Data are represented as mean ± standard error (SEM). * p≤0.05; ** p≤0.01; *** p≤0.001. (nBaseline = 18; nRTX = 9).
Fig 5.
Administration of RTX led to a widespread reduction of stimulus-driven functional connectivity in two distinct brain networks serving for noxious temperatures.
Direct comparison of baseline and RTX revealed that two distinct subnetworks (encircled) of brain structures described below showed significantly decreased (blue edges) FC after RTX administration. While thalamo-cortical and hippocampal-cortical connections remained intact (no change), a great reduction was found between thalamus, hippocampus and brainstem such as the superior colliculi and, independently, within cortical structures (primary and sensory somatosensory cortex, association cortex and cingulate cortex). Different functional groups are distinguished by the node colors. Node size represents number of significantly different connections each node has to the others. Thickness of edges represents strength of connection (sum of Pearson r of all significantly different connections). Statistical significance was calculated using NBS, thereby correcting the p-value. (nBaseline = 18; nRTX = 9). Abbreviations: Am: amygdala; Ass: association cortex; BG: basalganglia (esp. striatum); Bs: brainstem structures; Cg: cingulate cortex; HC: hippocampus; Ic: inferior colliculi; Ins: insular cortex; Lo: limbic output/hypothalamus; M: motor cortex; S1/S2: primary/secondary somatosensory cortex; Sc: superior colliculi; Teg: Tegmentum; thM/L: medial/lateral thalamus; ThR: other thalamic nuclei.
Fig 6.
Stimulus-driven Blondel communities proofed that both affected subnetworks were an interconnected system.
(a) RTX administration repressed mainly intra-thalamic connections (size of light blue nodes massively decreased, RTX) within the thalamo-hippocampal community (blue arrows) but enhanced the number of cortical connections (size of pink nodes increased, RTX) within the sensorimotor-association cortex community (pink arrows). Different functional groups are distinguished by the node colors (see legend in Fig 5), the size of the node represents the node degree (number of connections one functional group has to each other and also within the group). Edge thickness represents FC between two functional groups. Black lines encircle the Blondel communities. (b) RTX (solid) compared to baseline (shaded) reduced effectively the number of intra-thalamic (also see (c)) but increased all higher order connections, i.e. thalamo-cortical, within cortex as well as cortical-subcortical (excluding thalamus) connections. These can be observed for all stimulation temperatures, shown in the graphic are data from 50°C to 54°C. (c) Especially the number of connections between thalamus and cingulate cortex was increased for temperatures below 54°C whereas above 54°C no difference was found for thalamic-cingulate connectivity but instead connectivity between thalamus and sensory cortex increased. Shown is the difference between RTX and baseline for the number of connections between thalamus and the denoted regions. Marked in red are connections that were increased by RTX, blue are connections that were decreased by RTX. (d) Example of connections that are independent of RTX-administration: motoric output regions (comprising motor cortex and cerebellum) displayed stable connections within as well as to somatosensory cortex. (nBaseline = 18; nRTX = 9).
Fig 7.
Impact of thermal stimuli on RS functional connectivity with and without peripheral TRPV1-signaling.
No significant differences could be found between baseline and RTX RS-networks comparing the RS measurement that took place before the stimulus-driven stimulation ((a), RSpre, p = 0.298). In contrast, the stimulation left significant imprints on RS-networks (RSpre compared to RSpost) for both groups (b). At baseline, animals showed distinct changes in RS FC after being subjected to the nociceptive stimulus-based fMRI session: increased FC (red edges) could be shown for thalamus, association cortex, amygdala and hippocampus, whereas reduced FC (blue edges) was found mainly in cortical and amygdala-hypothalamic regions. After RTX-treatment, enhanced FC could be found only in limbic regions (Lo, Am, BG, Ins). Reduced FC was now more prominent specifically in thalamus, but also found, alike baseline, in cortical and hypothalamic areas. Different functional groups are distinguished by the node colors. Node size represents number of significantly different connections each node has to others. Thickness of edges represents strength of connection (sum of Pearson r of all significantly differing connections). Statistical significance was calculated using NBS (a) and paired t-Test, permutation corrected p≤0.05 (b). (nBaseline = 18; nRTX = 9). Abbreviations: Am: amygdala; Ass: association cortex; BG: basalganglia; Bs: brainstem structures; Cg: cingulate cortex; HC: hippocampus; Ic: inferior colliculi; Ins: insular cortex; Lo: limbic output/hypothalamus; M: motor cortex; S1/S2: primary/secondary somatosensory cortex; Sc: superior colliculi; Teg: Tegmentum; thM/L: medial/lateral thalamus; ThR: other thalamic nuclei.