• Sleep increases the sensory threshold reducing the information flow to the cortex (Edeline et al., 2000).

  • Visual and auditory stimulations reach the cortex and are processed during sleep (Nir et al., 2015, Sharon et al., 2018, Züst et al., 2019).

  • Thalamo-cortical (TC) oscillations during sleep contribute to the processing and learning of sensory inputs. This might optimize the memory consolidation of the sensory experience (Andrillon et al., 2016).

  • It remains to be elucidated:

    1. The processing of painful stimuli during different arousal states by the TC network.
    2. The influence of chronic pain on the thalamic gating of sensory stimuli.
  • Our objective:
    Understand the thalamo-cortical circuits involved in the thalamic gating of sensory stimuli in sleep and wake states in both healthy and pathological conditions.


Experimental Setup

Animals were implanted in both hemispheres in the ventroposterior lateral nucleus of the thalamus (VPL), the somatosensory cortex of the hind paw (S1HL) and the anterior cingulate cortex (ACC) (Figure 1). The choice of this areas was based on the spinothalamic tract that relies information from the periphery to the somatosensory cortex, through the VPL. The ACC was chosen for its known activation in acute stimulus and its hyperactivation in states of chronic pain as well as for its function as a hub for the integration of the experience of pain.

Figure 1. Experimental Setup. a Position of the ground and EEG screws as well as from the tetrodes. b Illustration of our custom electronic interface board and implanted brain areas. c Collected data types.

Experiment Timeline

Animals undergo stereotxic implantation during which tetrodes are implanted in the ACC, S1HL and VPL, screws for the EEG are positioned above the prefrontal cortex (frontal EEG) and above the hippocampus (parietal EEG) and the EMG wires are inserted. After one to two weeks of recovery, the recordings begin.
Recording sessions take place during the light phase.
In each recording session, animals are positioned over a grid and recevie only one type of stimulation. Each animal receives a total of 200 stimulations of each type before the SNI surgery. After the induction of chronic pain, animals are recorded twice per week and receive a total of 100 stimulations of each type of stimulation (Figure 2).
Recordings are performed with an Intan Technologies system.

Figure 2. Project timeline. Complete timeline of the project. The data shown in this poster is before performing the Spare Nerve Injury (SNI). This means that the animals have not developed chronic pain.

Evaluation of the Stimulation with the Pipette Tip

I evaluated the withdrawal threshold of each hind paw with a series of Semmes-Weinstein filaments ranging from 0.07g to 4g. Knowing the withdrawal threshold for each filament allows to evaluate the level of pain caused by the stimulations with the pipette tip.
The results show that the pipette tip elicits paw withdrawal in 46 +/- 6.9% of the stimulations. This equals a von Frey filament of 0.4-0.5g and therefore we considered the stimulations with the pipette tip as non-painful (Figure 3).

Figure 3. Evaluation of the Stimuli with Pipette Tip. Withdrawal threshold for each hind paw. 


Stimuli in REM do not elicit paw withdrawal but can change the arousal state

The percentage of stimuli in which the animals withdrew the paw is represented in the Withdrawal Response graphic. The Arousal State Change graphic shows the percentage of stimuli that evoked a change in the arousal state (Figure 4).

For both stimulation types, the percentage of stimulations in which the animals moved the paw away is higher in wake than in NREM sleep. And, as expected, the percent withdrawal is more pronnounced in the stimulations done with the needle than the ones done with the pipette tip.
Interestingly, stimulations performed during REM sleep do not evoke movement of the paw. Yet, the percentage of stimuli that awoke the animals is similar to the one of the stimuli applied during NREM.

Figure 4. Effect of the stimuli on the withdrawal response and the arousal state. a Withdrawal response for the pipette tip and the needle stimuli. b Arousal state change in percentage of the stimuli with pipette tip and the needle.

Stimuli are represented in both ipsi- and contra- lateral brain areas in wake and NREM sleep

The data is grouped by brain area and by stimulation site respective to the implanted brain area (i.e. Contralateral vs. Ipsilateral).
In each pannel, there is the heatmap of the stimulations representing the voltage deflections of the LFP (top), the evoked response potentials for each stimulation in grey and the mean of all the trials in black (middle), and the grand average spectrogram (bottom) (Figure 5).

A first observation of these results lets us conclude that:

  1. The stimuli are represented in both contralateral and ispsilateral brain areas.
  2. Stimuli applied during NREM sleep are relayed by the thalamus to the cortex

From the trial heatmap, there is no clear time-locked increase or decrease of the voltage in the LFP in neither of the brain areas nor the shown arousal states. However, there is a consistent activation of a wide range of frequencies upon stimulation onset. This activation seems to be faster in NREM stimuli compared to wake stimuli, especially in the contralateral side.

Figure 5. Comparison of ipsilateral and contralateral stimuli performed with the pipette tip. 

Trials without paw withdrawal are represented in all brain areas

The data shows contralateral stimuli and is grouped by brain area and by stimulation type.
In each pannel, there is the heatmap of the stimulations representing the voltage deflections of the LFP (top), the evoked response potentials for each stimulation in grey and the mean of all the trials in black (middle), and the grand average spectrogram (bottom) (Figure 6).

From a first glance of these results, we can conclude that stimuli without a clear withdrawal response of the stimulated paw evoke a clear activation of a wide range of frequencies. While this activation is particularly reliable in NREM sleep, in wake the activation is more diffuse and does not seem to be happen in all brain areas.

Figure 6. Comparison of the two stimulation types for contralateral stimuli without behavioral response.

Stimuli-evoked frequency changes

The coefficients resulting from the spectrograms for each trial were binned in time bins of 250 ms and in different frequency ranges. Data shown for the stimuli that evoked a behavioral response with the needle (Figure 7) and with the pipette tip (Figure 8).

Figure 7. Mean power for different frequency ranges for contralateral stimuli with the needle 

From these results, we can draw the following observations for the stimulation with the needle:

  1. There is a generalized decrease of the power of all frequencies upon stimulation in NREM sleep.
  2. In the somatosensory cortex, stimulation onset exerts a different effect on trials with a behavioral response from those without in NREM sleep. We can observe that the trials without behavioral response do not decrese the power upon stimulation as much as those in which there was a behavioral response.
  3. Stimuli performed during wake seem to have the strongest effect on the gamma range, between 30 and 100 Hz. This is only true for the recorded cortices, in which the stimuli with paw withdrawal have higher gamma power than those stimulations that did not evoke paw withdrawal.
  4. In the ACC, there are no apparent differences between those trials with and without paw withdrawal, except for the gamma range in wake.
Figure 8 (extra). Mean power for different frequency ranges for contralateral stimuli with the pipette tip.

On the other hand, for the stimulations with the pipette tip, we can make the following observations:

  1. As well as with painful stimuli, there is a transient increase and an important decrease in the power of the frequencies ranging from 2 Hz to 30 Hz in all brain areas.
  2. Stimuli in which the animal did not withdraw the paw the power decrease upon stimulus onset is not as pronounced as in the stimuli with behavioral response in the frequencies from 2 Hz to 30 Hz in the S1HL, from 2Hz to 16 Hz in the ACC and from 2Hz to 12 Hz in the VPL.
  3. Stimuli that felt into wake periods do not seem to evoke a change in the mean power of the studied frequency bands.

However, these are the raw data and a statistical analysis is still needed to withdraw more conclusive results.

NOTE: Please be aware that I have not done a statistical analysis of these results yet. Thus, the conclusions I have withdrawn from the figures are qualitative and not quantitative.


In this poster I have displayed the data obtained from stimuli done in wake and NREM sleep because the number of stimuli that felt into REM sleep episodes was very low (n = 10, per stimulation type).

From these results, we can observe that, as well as auditory and visual stimulation, painful and non-painful mechanosensation is relayed to the cortex in NREM sleep. Because this relay by the thalamus may be due to the arousal of the animal, in a next step I will separate those stimuli in which the animals woke up from those in which the animals continue sleep. From this division I expect to have a better differentiation from the trials that evoked the withdrawal of the paw from those which did not.

Additionally, it seems that the processing of mechanical stimulation, independently of its noxious value, takes place in both the contralateral and the ipsilateral hemispheres. It remains to be investigated whether there is one side that has a heavier weight on the processing of this type of information and I aim to get an answer to this question with the quantification of the changes evoked by the stimuli.

Unexpectedly, stimuli in which no withdrawal of the paw was observed, evoked activation of several frequencies. Here it remains the question whether awakening of the animals may have played a role in these observations. Thus, by separating the trials in which the animals awoke from those in which the animals continue sleeping, I expect to disentangle the difference between stimuli that do evoke a response from those that do not.

In short, I have shown that painful and non-painful mechanical stimuli do reach the cortex during sleep and are processed by the contralateral and ipsilateral hemispheres.


  1. Andrillon, T., Poulsen, A.T., Hansen, L.K., Léger, D., Kouider, S., 2016. Neural Markers of Responsiveness to the Environment in Human Sleep. J. Neurosci. 36, 6583–96.
  2. Edeline, J.M., Manunta, Y., Hennevin, E., 2000. Auditory thalamus neurons during sleep: changes in frequency selectivity, threshold, and receptive field size. J. Neurophysiol. 84, 934–952.
  3. Nir, Y., Vyazovskiy, V. V., Cirelli, C., Banks, M.I., Tononi, G., 2015. Auditory Responses and Stimulus-Specific Adaptation in Rat Auditory Cortex are Preserved Across NREM and REM Sleep. Cereb. Cortex (New York, NY) 25, 1362.
  4. Sharon, O., Nir, Y., 2018. Attenuated Fast Steady-State Visual Evoked Potentials During Human Sleep. Cereb. Cortex 28, 1297–1311.
  5. Züst, M.A., Ruch, S., Wiest, R., Henke, K., 2019. Implicit Vocabulary Learning during Sleep Is Bound to Slow-Wave Peaks. Curr. Biol. 29, 541–553.e7.1