Deciphering Authentic Nociceptive Thalamic Responses in Rats

Abstract

The thalamus and its cortical connections play a pivotal role in pain information processing, yet the exploration of its electrophysiological responses to nociceptive stimuli has been limited. Here, in 2 experiments we recorded neural responses to nociceptive laser stimuli in the thalamic (ventral posterior lateral nucleus and medial dorsal nucleus) and cortical regions (primary somatosensory cortex [S1] and anterior cingulate cortex) within the lateral and medial pain pathways. We found remarkable similarities in laser-evoked brain responses that encoded pain intensity within thalamic and cortical regions. Contrary to the expected temporal sequence of ascending information flow, the recorded thalamic response (N1) was temporally later than its cortical counterparts, suggesting that it may not be a genuine thalamus-generated response. Importantly, we also identified a distinctive component in the thalamus, i.e., the early negativity (EN) occurring around 100 ms after the onset of nociceptive stimuli. This EN component represents an authentic nociceptive thalamic response and closely synchronizes with the directional information flow from the thalamus to the cortex. These findings underscore the importance of isolating genuine thalamic neural responses, thereby contributing to a more comprehensive understanding of the thalamic function in pain processing. Additionally, these findings hold potential clinical implications, particularly in the advancement of closed-loop neuromodulation treatments for neurological diseases targeting this vital brain region.

Fig. 1.

Experimental design and main analyses. (A) The same experimental setup was utilized for both experiments, with rats moving freely within the plastic chamber during recording sessions. Laser stimuli were delivered to the plantar surface of the left or right forepaw when the animals were spontaneously still. (B) Schematic diagram displaying electrode implantation positions for simultaneous recording of multiple brain regions. (C) Experiment 1 involved laser stimulation with 3 intensities (2.5, 3, and 3.5 J), comprising 10 trials for each stimulation site and intensity. Experiment 2 utilized laser stimulation with 2 intensities (2.5 and 3.5 J), with 15 trials for each stimulation site and intensity. All stimuli were pseudo-randomly applied to the rats' forepaws. (D) The study encompassed 4 primary analyses: LFP analysis, spike firing rate analysis, PCA of LFPs, and SFGC analysis.

Fig. 2.

Nocifensive behaviors. (A and B) Nocifensive behavioral scores were significantly influenced by stimulus intensity in both experiments, with notably higher scores observed at greater stimulus intensity. * P < 0.05, ** P < 0.01.

Fig. 3.

Laser-evoked field potentials. (A) In Experiment 1, N1 amplitudes were significantly influenced by stimulus intensity, hemisphere, and brain region, with no significant interactions. N1 latencies were significantly affected by stimulus intensity. The magnitudes of LEP and GBO were also significantly influenced by stimulus intensity. (B) In Experiment 2, N1 amplitudes were significantly influenced by stimulus intensity and brain region, with no significant interactions. Note that there are significant differences in N1 latencies among brain regions, with significantly shorter latency in S1 than that in VPL. The magnitudes of LEP and GBO were significantly influenced by stimulus intensity. Error bars represent SEM. Grand-averaged waveforms for each condition are presented.

Fig. 4.

Laser-induced spike activities. The firing rates of single units were normalized as Z-scores relative to the baseline (500 ms preceding laser stimuli). Units were arranged on the y-axis to visualize the relative changes in firing rates (left panel of [A] and [B]). In both experiments, spike firing rates were represented as the average curve with an error band (standard error), and the firing rates (gray shaded area) within the 0- to 500-ms time window post-stimulation were used for statistical analysis. (A) In Experiment 1, spike-firing rates increased with stimulus intensity in both hemispheres and brain regions. Notably, the firing rates in contralateral VPL were significantly higher than those in ipsilateral VPL. (B) In Experiment 2, spike-firing rates in the 4 brain regions were significantly affected by intensity and hemisphere. Error bars represent SEM. Grand-averaged spike firing waveforms for each condition are presented.

Fig. 5.

PCA to isolate the EN in the thalamus. (A and B) In both experiments, 4 principal components were extracted, with their explained variance shown in the eigenvalue graph (Experiment 1: [PC1] 30.58%, [PC2] 11.87%, [PC3] 8.56%, [PC4] 4.82%; Experiment 2: [PC1] 32.57%, [PC2] 7.49%, [PC3] 7.33%, [PC4] 7.03%). PC2 in Experiment 1 and PC3 in Experiment 2 closely resembled the original N1 wave but revealed a distinct EN at 108 and 100 ms, respectively. Reconstructed field potentials based on PC2 in Experiment 1 and PC3 in Experiment 2 consistently demonstrated the presence of EN in all rats. (C) In Experiment 1, the modulatory effect of stimulus intensity on thalamic EN amplitudes was significant, with the EN amplitudes elicited by middle-intensity stimuli being significantly higher than those induced by low-intensity stimuli. In Experiment 2, EN amplitudes were also significantly influenced by stimulus intensity. EN amplitudes elicited by high-intensity stimuli in both thalamic regions were significantly larger than those induced by low-intensity stimuli. Furthermore, in both experiments, thalamic EN latencies were significantly shorter than cortical N1 latencies in both pain pathways. * P < 0.05, ** P < 0.01, *** P < 0.001. Error bars represent SEM. Grand-averaged waveforms of the field potentials and single rat waveforms of the EN component are presented.

Fig. 6.

SFGC of brain responses in the thalamus and cortex. The SFGC time courses for both experiments are presented as average curves with standard error bands. (A and B) Dark gray shadows highlight significant time intervals, determined through point-by-point paired-sample t tests for information flow within each brain region. In Experiment 1, VPL exhibited significantly higher differential information flow (i.e., Spike-to-LFP minus LFP-to-Spike) than S1. Experiment 2 also showed higher differential information flow within the thalamus (i.e., VPL and MD) compared to the cortex (i.e., S1 and ACC). (C) In both experiments, information flow from the thalamus to the cortex peaked before 100 ms. The information flow from the thalamic spikes to the cortical LFPs at 0 to 100 ms was significantly stronger than that from the cortex to the thalamus. Moreover, significant differences in peak latency were observed between (1) the information flow from the thalamus to the cortex, (2) the EN in the thalamus, and (3) the N1 in the cortex. In Experiment 1, the peak latencies of the information flow from VPL to S1 and the EN in VPL occurred significantly earlier than the N1 in S1. In Experiment 2, the peak latency of the information flow from VPL to S1 was significantly earlier than the peak latencies of the EN in VPL and the N1 in S1. The peak latency of the EN in VPL was also significantly earlier than that of the N1 in S1. Furthermore, the peak latencies of the information flow from MD to ACC and the EN in MD also occurred significantly earlier than the peak latency of the N1 in ACC. * P < 0.05, ** P < 0.001, *** P < 0.001. Grand-averaged SFGC waveforms for each condition are presented.