Network dynamics in nociceptive pathways assessed by the neuronal avalanche model
© Wu et al.; licensee BioMed Central Ltd. 2012
Received: 1 February 2012
Accepted: 26 April 2012
Published: 26 April 2012
Traditional electroencephalography provides a critical assessment of pain responses. The perception of pain, however, may involve a series of signal transmission pathways in higher cortical function. Recent studies have shown that a mathematical method, the neuronal avalanche model, may be applied to evaluate higher-order network dynamics. The neuronal avalanche is a cascade of neuronal activity, the size distribution of which can be approximated by a power law relationship manifested by the slope of a straight line (i.e., the α value). We investigated whether the neuronal avalanche could be a useful index for nociceptive assessment.
Neuronal activity was recorded with a 4 × 8 multichannel electrode array in the primary somatosensory cortex (S1) and anterior cingulate cortex (ACC). Under light anesthesia, peripheral pinch stimulation increased the slope of the α value in both the ACC and S1, whereas brush stimulation increased the α value only in the S1. The increase in α values was blocked in both regions under deep anesthesia. The increase in α values in the ACC induced by peripheral pinch stimulation was blocked by medial thalamic lesion, but the increase in α values in the S1 induced by brush and pinch stimulation was not affected.
The neuronal avalanche model shows a critical state in the cortical network for noxious-related signal processing. The α value may provide an index of brain network activity that distinguishes the responses to somatic stimuli from the control state. These network dynamics may be valuable for the evaluation of acute nociceptive processes and may be applied to chronic pathological pain conditions.
KeywordsPain assessment Neuronal avalanche Network dynamics
The critical evaluation of the pain response, especially in chronic, spontaneous pain, is valuable for clinical treatment. Traditional pain assessment measures the patient’s waveform pattern, frequency domain, and pair-wise cross-correlations of electroencephalography (EEG) and magnetoencephalography (MEG) recordings [1, 2] to characterize the pain response [3, 4]. The perception of pain, however, may involve a series of signal transmission pathways in higher cortical function [5, 6]. Recent studies have shown that a mathematical model, the neuronal avalanche model, can estimate higher-order cortical network dynamics . This model shows a cascade of neuronal activity in neuronal networks, the neuronal avalanche event size distribution of which can be approximated by a power law distribution manifested by the slope of the α value . This distribution can be taken as an index of network dynamics, with an α value in the range of −1 to −2 in vitro and in vivo [8–11]. This phenomenon is robust and lasts for many hours in continuous recordings in rats , cats , and monkeys . Calculating the power law exponent could offer quantitative means to evaluate the efficacy and the state of cortical networks for information transmission [7–12], which could not be provided by conventional multi-electrode recording methods. Recent studies have shown that the neuronal avalanche exists under many physiological conditions, such as wakefulness, slow-wave sleep, and rapid eye movement sleep, and may be involved in larger amounts of information processing . However, direct evidence that correlates network dynamics with functional tasks associated with specific signal processing, such as nociceptive responses, is still lacking. Nociceptive information was conveyed in lateral and medial pain pathways which project to the primary somatosensory cortex (S1) and anterior cingulate cortex (ACC) respectively [13–15]. The ACC receives mainly inputs from the medial thalamus and play a different pain-related functional role from that of the S1 [16, 17]. Thus the medial and lateral pain systems provide two distinct nociceptive pathways for testing the differential nociceptive responses to peripheral noxious stimuli. Thus the aim of the present study was to examine whether the neuronal avalanche in nociceptive pathways can objectively indicate network activity in brain regions that process nociceptive information.
The present study showed that the neuronal avalanche could be detected and modulated in nociceptive-reactive brain regions. The power law distribution indicated that the cortical network for pain signaling was in a critical state and optimal for a large amount of signal processing [7–10, 19]. The neuronal avalanche, which has been noted as a state of self-organized criticality (SOC), may be an intrinsic property of cortical networks that results from the interactions between neurons in local circuits . Modeling studies have suggested that SOC networks are optimized for input processing, information storage, and transfer [20–22]. Thus, the neuronal networks in the ACC and S1 operate in a critical state and maintain a moderate level of interactions that can satisfy the competing demands of nociceptive information capacity and transmission [8, 12, 19]. The present study had three major findings. First, the neuronal avalanche revealed large changes in network dynamic, reflected by changing the noxious inputs. The slope of the power law distribution revealed the state of the network dynamics in which increasing α values indicated the increasing possibility of larger activity events and an expansion of the field of excitatory networks . Noxious inputs may alter the excitation-to-inhibition ratio in the cortical network and thus change the network dynamics profile and modulate information processing . Second, network dynamics may be modulated by the level of anesthesia. Isoflurane affects many types of neuronal transmitters in the brain, such as acetylcholine and 5-hydroxytryptamine in the prefrontal cortex and glutamate and γ-aminobutyric acid in the somatosensory cortex . Thus, deep anesthesia may downregulate the excitatory inputs of cortical dynamics and causes a shift of α values. Third, the modulation of the neuronal avalanche is input-specific and pathway-dependent. The α value increased in the ACC in response to pinch stimulation and was affected by MT lesion, but not in the S1. Ascending nociceptive responses were blocked by MT lesion, and the network dynamics in the ACC reverted to control values. The noxious stimulation-induced increase in α values in the ACC was blocked by MT lesion, whereas the brush and pinch stimulation-induced increases in α values in the S1 were not affected by MT lesion. These results indicate that the network dynamics that are modulated by noxious inputs are pathway-specific.
The sensitivity of this analytical method can distinguish changes in cortical network dynamics following peripheral somatic stimulation. We recently used an animal model of central pain and found hyperexcitability and allodynia  that may change cortical dynamics by inducing thalamocortical dysrhythmia. The thalamic rhythm is an essential element in the generation and perception of neuropathic pain, and dysrhythmia of these activities can disturb the synchronicity of pain processing . Abnormal neuronal activities could be subtracted from network dynamics by calculating the neuronal avalanche in the altered network operations. We anticipate that the neuronal avalanche model may provide an objective index of network dynamics for the evaluation the abnormal cortical rhythmicity and occurrence of spontaneous pain in chronic pathological conditions, such as neuropathic and central pain.
Material and Methods
Preparation of animals and electrodes
Male Sprague–Dawley rats (300–400 g) were housed in an air-conditioned room with free access to food and water. All of the experiments were performed in accordance with the guidelines of the Academia Sinica Institutional Animal Care and Utilization Committee. The rats were initially anesthetized with 4 % isoflurane (in pure O2) in an acrylic box. The animals were then placed in a stereotaxic apparatus and maintained under anesthesia with 2 % isoflurane during surgery. One Michigan probe with 32 contact points (150 μm-lead interval, eight leads on one shank, and four parallel shanks) was used to record the extracellular field potentials in the right ACC. Another Michigan probe was used to record extracellular field potentials in the hindpaw projection area in the right S1. DiI was dissolved in isopropanol at a saturated concentration and coated on the Michigan probe three times to ensure successful coating. The animals were subsequently maintained under anesthesia with 1.25-2.5 % isoflurane during the recording session. The depth of anesthesia was continuously monitored by an anesthesia monitor (Capnomac Ultima, General Electric Company, Fairfield, CT, USA) that monitored the minimum alveolar concentration (MAC) of isoflurane. Two depths of anesthesia, light and deep, were maintained, based on three criteria: (i) isoflurane MAC value for light anesthesia and 2×MAC for deep anesthesia; (ii) the presence (light anesthesia) or absence (deep anesthesia) of a withdrawal response to pinch stimulation of the paw; (iii) an EEG waveform and FFT peak value that shifted from 5 Hz (light anesthesia) to 2 Hz (deep anesthesia). To deactivate the MT, a tungsten electrode was inserted into the MT for electrolytic lesion, which was performed with a direct current of 100 μA for 100 s by a constant current pulse generator (Model 2100, A-M Systems, WA, USA).
All of the raw LFP data were recorded for 22 min and filtered with a preamplifier with a 0.1 Hz to 20 KHz band pass. The sampling rate of the recorded analog signals was 40 kHz, and the data were processed using a multichannel data acquisition system (ME, Multi Channel Systems, Reutlingen, Germany) on a computer. The nLFP data were further processed by filtering at 10–200 Hz. The time-point was selected as the nLFP that exceeded four-times the standard deviation of the threshold of basal activity and was marked as the digital unit for further neuronal avalanche calculation. The filtered traces were searched for time points at which they reach the threshold. The processed data thus contained a serial time point of nLFP and could be framed by selected time-bin. The time-bin was set to 4 ms according to our previous study . Each avalanche size is defined as number of digitized unit in each frame and the frame numbers of various neuronal avalanche sizes were counted for the avalanche size distribution. The avalanche size distribution was calculated using Matlab software (The MathWorks, Natick, MA, USA) and plotted using SigmaPlot software (Systat Software, Chicago, IL, USA). The slope of the power law distribution was selected from 1 to 10 in avalanche size, with a fitting index, R2, lager than 0.9. To verify that the power law distribution of unit activity is interdependent events, the original events were randomized to produce shuffled data. A fitting index of the shuffled data less than 0.9 indicated a change from a power law distribution to a Poisson distribution. The comparison of α values between groups was made using Student’s t-test and STSS software (IBM SPSS, IL, USA). Values of p < 0.05 were considered statistically significant.
JJSW and HCS participated in the design of the study, conducted the experiments, analyzed the data, and drafted the manuscript. CTY participated in the discussion of the experimental results and made experimental suggestions. BCS conceived of the study, participated in its design and coordination, and participated in the writing of the manuscript. All of the authors read and approved the final manuscript.
We thank Dr. Chi Keung Chan for his insightful comments on the experimental design. The present study was supported by grants from the National Science Council (99-2320-B-001-016-MY3) and Academia Sinica, Taiwan.
- Raij TT, Vartiainen NV, Jousmäki V, Hari R: Effects of interstimulus interval on cortical responses to painful laser stimulation. J Clin Neurophysiol 2003, 20: 73–79. 10.1097/00004691-200302000-00010View ArticlePubMedGoogle Scholar
- Lorenz J, Garcia-Larrea L: Contribution of attentional and cognitive factors to laser evoked brain potentials. Neurophysiol Clin 2003, 33: 293–301. 10.1016/j.neucli.2003.10.004View ArticlePubMedGoogle Scholar
- Kakigi R, Inui K, Tamura Y: Electrophysiological studies on human pain perception. Clin Neurophysiol 2005, 116: 743–763. 10.1016/j.clinph.2004.11.016View ArticlePubMedGoogle Scholar
- Baliki MN, Baria AT, Apkarian AV: The cortical rhythms of chronic back pain. J Neurosci 2011, 31: 13981–13990. 10.1523/JNEUROSCI.1984-11.2011PubMed CentralView ArticlePubMedGoogle Scholar
- Roy S, Llinás RR: Dynamic geometry, brain function modeling, and consciousness. Prog Brain Res 2008, 168: 133–144.View ArticlePubMedGoogle Scholar
- Hauck M, Lorenz J, Engel AK: Role of synchronized oscillatory brain activity for human pain perception. Rev Neurosci 2008, 19: 441–450.View ArticlePubMedGoogle Scholar
- Yu S, Yang H, Nakahara H, Santos GS, Nikolic D, Plenz D: Higher-order interactions characterized in cortical activity. J Neurosci 2011, 31: 17514–17526. 10.1523/JNEUROSCI.3127-11.2011View ArticlePubMedGoogle Scholar
- Beggs JM, Plenz D: Neuronal avalanches in neocortical circuits. J Neurosci 2003, 23: 11167–11177.PubMedGoogle Scholar
- Gireesh ED, Plenz D: Neuronal avalanches organize as nested theta- and beta/gamma-oscillations during development of cortical layer 2/3. Proc Natl Acad Sci U S A 2008, 105: 7576–7581. 10.1073/pnas.0800537105PubMed CentralView ArticlePubMedGoogle Scholar
- Petermann T, Thiagarajan TC, Lebedev MA, Nicolelis MAL, Chialvo DR, Plenz D: Spontaneous cortical activity in awake monkeys composed of neuronal avalanches. Proc Natl Acad Sci U S A 2009, 106: 15921–15926. 10.1073/pnas.0904089106PubMed CentralView ArticlePubMedGoogle Scholar
- Hahn G, Petermann T, Havenith MN, Yu S, Singer W, Plenz D, Nikolic D: Neuronal avalanches in spontaneous activity in vivo. J Neurophysiol 2010, 104: 3312–3322. 10.1152/jn.00953.2009PubMed CentralView ArticlePubMedGoogle Scholar
- Shew WL, Yang H, Petermann T, Roy R, Plenz D: Neuronal avalanches imply maximum dynamic range in cortical networks at criticality. J Neurosci 2009, 29: 15595–15600. 10.1523/JNEUROSCI.3864-09.2009PubMed CentralView ArticlePubMedGoogle Scholar
- Kuo C-C, Yen C-T: Comparison of anterior cingulate and primary somatosensory neuronal responses to noxious laser-heat stimuli in conscious, behaving rats. J Neurophysiol 2005, 94: 1825–1836. 10.1152/jn.00294.2005View ArticlePubMedGoogle Scholar
- Wang J-Y, Huang J, Chang J-Y, Woodward DJ, Luo F: Morphine modulation of pain processing in medial and lateral pain pathways. Mol Pain 2009, 5: 60. 10.1186/1744-8069-5-60PubMed CentralView ArticlePubMedGoogle Scholar
- Apkarian AV, Bushnell MC, Treede RD, Zubieta JK: Human brain mechanisms of pain perception and regulation in health and disease. Eur J Pain 2005, 9: 463–484. 10.1016/j.ejpain.2004.11.001View ArticlePubMedGoogle Scholar
- Sun J-J, Kung J-C, Wang C-C, Chen S-L, Shyu B-C: Short-term facilitation in the anterior cingulate cortex following stimulation of the medial thalamus in the rat. Brain Res 2006, 1097: 101–115. 10.1016/j.brainres.2006.04.065View ArticlePubMedGoogle Scholar
- Yang J-W, Shih H-C, Shyu B-C: Intracortical circuits in rat anterior cingulate cortex are activated by nociceptive inputs mediated by medial thalamus. J Neurophysiol 2006, 96: 3409–3422. 10.1152/jn.00623.2006View ArticlePubMedGoogle Scholar
- Shyu B-C, Chen W-F, Shih H-C: Electrically and mechanically evoked nociceptive neuronal responses in the rat anterior cingulate cortex. Acta Neurochir Suppl 2008, 101: 23–25. 10.1007/978-3-211-78205-7_4View ArticlePubMedGoogle Scholar
- Shew WL, Yang H, Yu S, Roy R, Plenz D: Information capacity and transmission are maximized in balanced cortical networks with neuronal avalanches. J Neurosci 2011, 31: 55–63. 10.1523/JNEUROSCI.4637-10.2011PubMed CentralView ArticlePubMedGoogle Scholar
- Werner G: Brain dynamics across levels of organization. J Physiol Paris 2007, 101: 273–279. 10.1016/j.jphysparis.2007.12.001View ArticlePubMedGoogle Scholar
- Chen W, Hobbs JP, Tang A, Beggs JM: A few strong connections: optimizing information retention in neuronal avalanches. BMC Neurosci 2010, 11: 3. 10.1186/1471-2202-11-3PubMed CentralView ArticlePubMedGoogle Scholar
- Wang SJ, Hilgetag CC, Zhou C: Sustained activity in hierarchical modular neural networks: self-organized criticality and oscillations. Front Comput Neurosci 2011, 5: 30.PubMed CentralPubMedGoogle Scholar
- Müller CP, Pum ME, Amato D, Schüttler J, Huston JP, Silva MADS: The in vivo neurochemistry of the brain during general anesthesia. J Neurochem 2011, 119: 419–446. 10.1111/j.1471-4159.2011.07445.xView ArticlePubMedGoogle Scholar
- Shih HC, Shyu BC: Enhancement of nociceptive thalamic activities developed in central post stroke pain. Taipei: The 7th Federation of the Asian and Oceanian Physiological Societies Congress: 27–30 Sep, 2011; 2011:P-E01.Google Scholar
- Walton KD, Llinas RR: Central pain as a thalamocortical dysrhythmia: a thalamic efference disconnection? In Translational Pain Research. Edited by: Kruger L, Light AR, Walton K. Boca Raton: CRC Press; 2010:Chap.13.Google Scholar
- Wu JSJ, Chang WP, Shih HC, Shyu BC: Anterior cingulate cortex exhibits the neuronal avalanche that could be modulated from medial thalamic nucleus input. The 9th International Association for the Study of Pain Research Symposium: 16–17 Oct, 2011, ShangHai; 2011:P-3.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.