Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Anterior cingulate cortex connectivity is associated with suppression of behaviour in a rat model of chronic pain:

Anterior cingulate cortex connectivity is associated with suppression of behaviour in a rat model... A cardinal feature of persistent pain that follows injury is a general suppression of behaviour, in which motivation is inhibited in a way that promotes energy conservation and recuperation. Across species, the anterior cingulate cortex is associated with the motivational aspects of phasic pain, but whether it mediates motivational functions in persistent pain is less clear. Using burrowing behaviour as an marker of non-specific motivated behaviour in rodents, we studied the suppression of burrowing following painful confirmatory factor analysis or control injection into the right knee joint of 30 rats (14 with pain) and examined associated neural connectivity with ultra-high-field resting state functional magnetic resonance imaging. We found that connectivity between anterior cingulate cortex and subcortical structures including hypothalamic/preoptic nuclei and the bed nucleus of the stria terminalis correlated with the reduction in burrowing behaviour observed following the pain manipulation. In summary, the findings implicate anterior cingulate cortex connectivity as a correlate of the motivational aspect of persistent pain in rodents. Keywords Chronic pain, confirmatory factor analysis pain model, resting state functional magnetic resonance imaging, pain in rodents Received: 11 January 2018; accepted: 4 May 2018 Introduction One of the greatest challenges in the development of novel anal- thalamus, primary and secondary somatosensory cortex, cerebel- gesics is the difficulty in evaluating pain in animal models of lum, the insular cortex, and the anterior cingulate cortex (ACC) chronic pain. Pain is a uniquely subjective experience, and the mainstay of outcome measures in clinical trials remains subjec- Department of Psychology and Behavioral and Clinical Neuroscience tive ratings. Since these are not available in animals, evaluation Institute, University of Cambridge, Cambridge, UK of pain depends on surrogate measures of behaviour, primarily Center for Information and Neural Networks, National Institute of motor responses directly related to evoked or spontaneous pain. Information and Communications Technology, Osaka, Japan However, the difficulty in translating these behaviours to humans Computational and Biological Learning Laboratory, Department of is well recognised, and this greatly hampers the ability to predict Engineering, University of Cambridge, Cambridge, UK whether analgesics that are successful in animals will work in Pain & Neuroscience, Drug Discovery & Disease Research Laboratory, humans. This has led to a requirement for behavioural measures Shionogi & Co., Ltd., Osaka, Japan that better reflect persistent pain in animals, and which have Translational Research Unit, Biomarker R&D Department, Shionogi & translational validity to humans. Co., Ltd., Osaka, Japan Immunology Frontier Research Center, Osaka University, Osaka, Japan Functional neuroimaging offers a novel approach to the eval- Brain Information Communication Research Laboratory Group, Advanced uation of pain in rodents (Baliki et al., 2014; Thompson and Telecommunications Research Institute International, Kyoto, Japan Bushnell, 2012). In humans, a broad set of cortical and subcorti- cal regions are implicated in pain processing and the expression Corresponding author: of pain behaviour. This diversity reflects the fact that pain is a Christian Sprenger, Computational and Biological Learning Laboratory, multidimensional experience, engaging sensory, affective, and Department of Engineering, University of Cambridge, Trumpington cognitive processing. Human studies of phasic experimental Street, Cambridge CB2 1PZ, UK. pain implicate a network of pain-related regions that includes Email: cs910@cam.ac.uk Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Brain and Neuroscience Advances (Moisset and Bouhassira, 2007; Price, 2000). Although the pre- Preparation of pain model rats cise functions of each region remain unclear, there is a consen- Rats were anesthetised with isoflurane (Mylan, Canonsburg, PA, sus that some regions (thalamus, primary and secondary USA) and then intra-articular injected with 25 µL of either somatosensory cortex, cerebellum) better reflect sensori-motor Freund’s Complete Adjuvant (confirmatory factor analysis processes, and others (the anterior insular cortex and in particu- (CFA); Mycobacterium butyricum (BD DIFCO, Franklin Lakes, lar the ACC) better reflect affective processes related to pain NJ, USA); 2 mg/mL of liquid paraffin (Maruishi Pharmaceutical (Büchel et al., 2002; Morrison et al., 2004; Rainville et al., Co., Ltd., Osaka, Japan)), or vehicle (sham rats) into the right 1997). However, there remains uncertainty as to how well stud- knee joint of the hind leg. In both cases, the left hind knee joint ies of phasic experimental pain translate to the persistent pain was kept untreated. Both the pain model and the sham rats were seen in chronic pain models. Although phasic pain occurs as a 6 weeks old at the time of injection. part of many chronic pain conditions, as seen with spontaneous phasic and hyper-sensitive evoked pain, the nature of persistent underlying pain is quite different and serves a distinct behav- Pain behaviour ioural function to reduce activity and conserve energy. The ACC is a strong candidate to play a dominant role in Rats were tested for pain behaviour 18 and 22 days after CFA/ mediating the affective behavioural manifestations of persistent vehicle injection. First, the knee diameter (KD) was determined pain. It serves a primary function as a modulator of the affective using a digital caliper (CD-15CX; Mitutoyo Corporation, tone of visceral, motor and endocrine efferents to downstream Kanagawa, Japan). Second, a weight bearing difference (DWB) regions involved in responses to nociceptive stimuli, including test on each hind leg was performed (Bioseb, Boulogne, France). the periaqueductal grey (PAG), the amygdala (Calejesan et al., Values of WBD are obtained using the following formula, where 2000; Vogt et al., 1992; Zhuo, 2007), and the basal forebrain W corresponds to the amount of weight put on the right leg and (Kash et al., 2015). W to the amount of weight put on the left leg Recently, novel behavioural tasks such as burrowing behav- iour (BB) have been proposed as a measure of the affective %/ WW =× 100 WW + () RR RL impact of pain. BB is an innate, spontaneous behaviour that dem- This formula expresses the percentage of the rat’s body weight put onstrates the overall ‘wellbeing’ or affective-motivational tone of on the CFA/vehicle-injected right leg; thus, a 50% value means the rats (Deacon, 2006) and is not affected by limb hypersensitiv- equal weight distribution across both hind legs. ity (Andrews et al., 2012). Burrowing provides shelter and pro- Third, a grip strength (GS) test (San Diego Instruments, San tection from environmental predators, food storage, and foraging, Diego, CA, USA) was conducted to quantify the mechanical but also comes with energy costs (Reichman and Smith, 1990). strength of the hind legs and hence to control for the development This behaviour is reduced in rats following peripheral nerve of pain in model rats. Briefly, each rat was gently restrained and injury or pain (Andrews et al., 2012), and therefore may provide allowed to grasp the wire mesh frame with its hind limbs and was a useful index of the motivational component of persistent back- moved in a rostral-to-caudal direction until the grip released. ground pain. Finally, BB was investigated. For burrowing experiments, plastic Taken together, these considerations suggest that changes in tubes (32 cm in length and 10 cm in diameter) were filled with ACC connectivity may be related to the modulation of motiva- 3 kg of gravel (5–8 mm particle size) and placed in Plexiglas tional behaviours, such as burrowing, by persistent pain. To test cages (560 × 440 × 200 mm). The open-end of the tube was ele- this hypothesis, we studied BB in adult rats in a model of vated 6 cm from the floor of the cage. Rats were allowed to indi- inflammatory arthritis pain (intra-articular injection of com- vidually burrow during 30 min for 18 days after CFA or vehicle plete Freund’s adjuvant) and measured ACC functional connec- injections were performed and the amount of gravel burrowed tivity in the brain using functional magnetic resonance imaging was recorded. (fMRI). For the statistical comparison of the behavioural data from the CFA pain model and control animals, we employed a two-sample student’s t-test (one-tailed). Results were considered significant Materials and methods at p < 0.05. Experimental animals Seventy-two male LEW/CrlCrlj rats (Charles River Laboratories, Animal selection and transfer to magnetic Japan, Inc.) were used. Rats were housed in groups of three in resonance imaging facilities plastic cages under controlled temperature and humidity and pro- vided free access to food and water under a 12/12 h reversed Based on the results obtained from the WBD on the 18th day, rats light-dark cycle (lighting at 8:00 a.m.). To reduce transfer stress were selected out of the ones that displayed either moderate or (single housing, changing the cage-mates) and negative impact severe pain and were sent to the magnetic resonance imaging (active social interaction or attacks from non-injured rats) from (MRI) facility. Selection criteria for CFA rats were <35% of non-injured rats to injured rats, the rats were housed and trans- WBD value. Functional MRI scanning was conducted at the ported only with rats from their group. All procedures were Center for Information and Neural Networks (CiNet, Osaka approved by internal animal care and use committee of Shionogi University, Suita, Japan). After being transferred to the MRI Pharmaceutical Research Center (Osaka, Japan) instructed by facility, animals were kept for 4 days under standard laboratory Association for Assessment and Accreditation of Laboratory conditions (room temperature of 22°C–23°C and a 12 h light/ Animal Care International (AAALAC) guidelines. dark cycle) with free access to food and water. Rats were group Morris et al. 3 housed with previous cage mates during and after transportation. maximum = 0.5 mm, band pass filtering (0.01 < f < 0.1), and lin- MRI data were acquired on the third day (i.e. 21 days after CFA/ ear regression of motion parameters (3dTproject). vehicle injection). On the fourth day, behavioural data were obtained once more. Assessment of functional connectivity Functional connectivity (Friston, 1994) of the ACC was assessed Resting state functional MRI data acquisition with a seed-based functional connectivity approach. The ACC Data was acquired from a total of 37 rats (18 CFA pain model) seed region was determined as the anterior portion of Brodmann’s with an 11.7 Tesla Avance II vertical bore system (Bruker area 24 (Vogt and Peters, 1981), approximately 1 mm rostral to BioSpin, Ettlingen, Germany) and a home-made transmit/receive Bregma (Calejesan et al., 2000). The mean time series in this surface radio frequency (RF) coil. Rats were anesthetised with a region were extracted for each animal from a sphere of 0.5 mm mixture of air and 2.8% isoflurane (Wako Pure Chemical radius around this coordinate and functional connectivity was Industries Ltd., Osaka, Japan) and then placed in an MRI- determined by calculating brain-wide z-transformed correlation compatible animal cradle. The isoflurane concentration was maps based on the preprocessed functional time series. maintained at 2% ± 0.5%, adjusted to maintain the respiration rate Subsequently, functional connectivity maps were compared at 70 ± 10 breaths per min throughout the sessions. between groups employing a two-sample t-test. Finally, correla- An axial T2-weighted (T2W) imaging was performed using tions between the individual functional connectivity maps and a rapid acquisition of relaxation enhancement (RARE) the behavioural measures (KD, DWB, GS, BB) were calculated sequence (repetition time/echo time (TR/TE) = 6500/45 ms, across animals and compared between groups employing like- number of averages (NA) = 8, field of view (FOV) = 32 × 16 mm, wise t-statistics. matrix size = 256 × 256, slice thickness = 500 µm, acquisition Cluster correction with AFNI’s 3dClustSim (Cox et al., 2017) time = 14 min). was used to reduce the instance of false positives caused by spa- To acquire resting state functional magnetic resonance imag- tial autocorrelation suggested by Eklund et al. (2016) (p < 0.05 ing (rsfMRI) data, we performed gradient-echo echo-planar per voxel plus cluster size k = 15,131 voxel, corrected at imaging (TR/TE = 2500/7 ms, number of segments = 2, flip alpha < 0.01). angle = 60°, FOV = 51.2 × 51.2 mm, matrix size = 64 × 64, slice thickness = 1 mm, in-plane resolution = 800 × 800 µm , band- Results width = 300 kHz, acquisition time = 20 min). Behavioural and physiological measures Imaging data preprocessing KDs were measured to determine the amount of joint swelling as an index of inflammation. Ipsilateral KD in CFA pain model The imaging data preprocessing was performed with AFNI rats was significantly increased compared to vehicle-treated (https://afni.nimh.nih.gov) and FSL (FMRIB, University of rats at day 22 (mean KD CFA-treated rats 10.8 ± 0.15 (SEM), Oxford, UK; https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/MELODIC). mean KD control group 8.7 mm ± 0.03, t(28) = 14.2, p < 0.001; Data from seven animals could not be analysed due to technical Figure 1(a)). CFA pain model rats also showed decreased GS difficulties and resulting poor image quality leaving 30 animals (mean GS CFA animals 931.0 g/kg ± 30.1, mean GS control for the final rsfMRI analysis (14 animals with pain, 16 con- group 1259.0 g/kg ± 10.6, t(28) = 17.4, p < 0.001; Figure 1(b)), trols). Functional data were corrected for slice timing offsets, weight on ipsilateral paw (mean DWB CFA-treated rats with mean and linear detrending (3dTshift) and images were 27.7% ± 1.5, mean DWB control group 49.5% ± 0.7, t(28) = 13.9, realigned to a single middle slice (3dvolreg) for volume-to- p < 0.001, Figure 1(c)), and amount of burrowed gravels (mean volume rigid body correction. Resultant motion and displace- BB CFA-treated rats 444.5 g ± 121.5, mean BB control group ment parameters were visually inspected for outliers. 1238.2 g ± 72.2, t(28) = 5.8, p < 0.001, Figure 1(d)). The results Anatomical data were skull stripped with a standard spherical suggest that CFA pain model rats showed evoked and spontane- model (3dSkullStrip). In order to do this skull strip, anatomical ous inflammatory knee joint pain at the time of fMRI data were ‘shrunk’ in the anterior–posterior direction (3drefit) scanning. to make the brain more spherical, in a manner similar to (Kundu et al., 2014). After effective skull strip, anatomical data were returned to their original voxel dimensions. Functional and ana- Resting state functional connectivity tomical data were then roughly aligned in space (@align_cent- ers) and orientation when oblique (3drotate). Functional data The ACC seed region showed reduced functional connectivity in were up-sampled to match anatomical voxel dimensions (3dre- the CFA pain group compared to control animals with central fit). Anatomical and functional data were then coregistered parts of the contralateral somatosensory cortex and dorsal por- using a 12-paramter registration (3dAllineate). A study-specific tions of the cingulate cortex (Figure 2(a)). In addition, the ACC anatomical template was created by averaging four normal con- exhibited increased functional connectivity in the pain group trol rat anatomical datasets. Anatomical data for each rat were with rostral portions of the somatosensory cortex and in the sub- normalised to this study-specific template and resultant warp cortex with the bilateral striatum, structures of the basal fore- parameters were applied to their respective functional data brain region, hypothalamic region/preoptical area (POA), and (3dAllineate). Functional data were furthermore subjected to the area of the bed nucleus of the stria terminalis (BNST; see de-spiking (3dDespike), smoothing with full width at half Figure 2(a)). 4 Brain and Neuroscience Advances Figure 1. Behavioural and physiological changes of CFA pain model rats compared to controls at the time of fMRI scanning. (a) Changes in contralateral and ipsilateral knee diameter (KD) in vehicle and CFA rats at day 22. (b) Changes in hind limb grip strength (GS) in vehicle and CFA rats. Data are shown as an average value of days 18 and 22 expressed as grip strength/body weight. (c) Changes in dynamic weight bearing (WBD) on the ipsilateral paw in vehicle and CFA rats. Each data were shown as an average value of days 18 and 22. (d) Changes in burrowing behaviour (BB) in vehicle treated and CFA pain model rats at day 18. *** indicates p < 0.001. In CFA pain animals, we found a marked negative correlation ability, and reduced weight bearing. As hypothesised, we also between ACC functional connectivity and innate BB in the latter observed reduced BB accounting on average for less than 50% of regions (POA & BNST; Figure 2(b)), that is, higher functional the behaviour in control animals. Resting state fMRI revealed a connectivity between ACC and these regions was associated with pronounced negative correlation between the individual strength suppression of BB. of ACC functional connectivity and BB in structures of the basal For illustration purposes, Figure 2(c) shows the correlation of forebrain, comprising the hypothalamic region and the BNST. connectivity strength between ACC with the peak subcortical No correlations with the ACC functional connectivity were region of interest and BB. ACC connectivity shows a negative observed for motor responses directly associated with joint pain. correlation with burrowing in the CFA pain group (r = –0.63, The findings therefore indicate a relatively specific relationship p = 0.02) but not in the control group (r =−0.23, p = 0.39). The between ACC functional coupling and the suppression of BB by interaction failed, however, significance (z = 1.2, p = 0.11). We persistent pain and suggest that ACC connectivity might be a found no significant correlations between ACC connectivity with good marker for the affective-motivational component of pain in KD, weight bearing, and GS. rodents. Across mammalian species neuronal responses to noxious stimuli have been shown to involve a number of brain regions Discussion including the primary and secondary somatosensory cortex, the In the present study, we assessed spontaneous BB in the CFA ACC, and the insular cortex (Iadarola et al., 1998; Lenz et al., model of persistent inflammatory arthritis pain and investigated 1998; Peyron et al., 2000; Ploner et al., 2002; Pohlmann et al., ACC resting state functional connectivity compared to controls. 2016; Thompson and Bushnell, 2012). The behavioural measures consistently indicated inflammatory Although the precise contribution of each region to the gen- nociception related to the CFA-treated joint at the time of MRI eration of pain remains to be elucidated (Moayedi and Davis, 2012; Segerdahl et al., 2015), neuroimaging studies highlighted measurement by showing ipsilateral swelling, reduced motoric Morris et al. 5 Figure 2. Changes in resting state functional connectivity in CFA pain model rats compared to control and correlation with burrowing behaviour. (a) Changes in ACC resting state functional connectivity in CFA pain rats compared to vehicle-treated animals. The ACC showed increased functional connectivity in CFA pain rats with parts of the contralateral somatosensory cortex (upper and lower panels), the bilateral striatum (upper and middle panels) and with the basal forebrain region/preoptical area and the region of the bed nucleus of the stria terminalis (lower panel). ACC functional connectivity was decreased in parts of the cingulate cortex itself and in the somatosensory cortex (middle panel). The three white lines in the smaller panel next to the upper panel of (a) indicate the approximate rostro-caudal position of the three axial sections of (a) in the rat brain. The colour bar indicates t values. Hot colours represent increased functional connectivity and cold colours represent decreased connectivity. A schematic representation of the involved subcortical regions is shown next to the lower panel of (a). (1) Hypothalamic/preoptical area, (2) basal forebrain region, (3) bed nucleus of the stria terminalis, (4) globus pallidus, (5) striatum, (6) septal region. (b) Negative correlation of burrowing behaviour with the individual strength of ACC functional connectivity. The colour bar indicates t values. The visualisation threshold is set to p < 0.005 uncorrected in (a) and (b). Note the spacial overlap of increased ACC-subcortical connectivity in (a) with the negative connectivity–behaviour correlation in (b) and the slightly different positions of the axial sections in (b) compared to (a). (c) Illustration of the negative correlation between burrowing behaviour with the individual connectivity strength at the subcortical peak in the upper panel of (b). ACC connectivity shows a marked negative correlation with burrowing in the CFA pain group but not in the control group. some general characteristics of the involved regions. While the While the association between ACC-hypothalamic coupling somatosensory cortex responds faster to dimensional features of and BB might reflect a general homeostatic dimension related to noxious stimuli, and is more engaged when larger body surfaces the animal’s behavioural suppression, the correlation between are exposed to pain, the ACC as part of the limbic system is burrowing and coupling strength with the BNST is especially thought to be modulated by the affective relevance pertinent to a interesting as it has been linked to sustained vigilance associated change in motivational tone or response selection (Peyron et al., with ambiguous or distant threat cues. This potentially provides 2000; Ploner et al., 2002). a direct link to the affective dimension of pain and illustrates The ACC is often conceptualised as a nexus for the processing BB as an expression of the affective-motivational tone of the of external salient stimuli, autonomic response regulation, and animals.That is, when a limb is injured, the potential danger by subsequent affective learning (Gao et al., 2004; Vogt, 2005). For predators and accordingly monitoring of potential threats example, fear learning in mice through observing other mice becomes much more important, so that behaviours such as bur- receiving painful foot shocks has been demonstrated to involve rowing are no longer prioritised. In line with this, lesions of the the ACC (Jeon et al., 2010). In the context of pain, the ACC is BNST have been demonstrated to disrupt the individual variabil- thought to allow potentially harmful stimuli to engage appropri- ity in the rodent’s anxiety-like behaviour (Durvarci, 2009). ate affective and motivationally relevant behaviours, and as such Finally, the role of the BNST as a site of integration of limbic one would expect the strength of ACC-based connectivity to be forebrain information is also supported by tracing studies show- associated with elevated aversive behaviors and reduced behav- ing direct anatomical BNST connections with the ACC (area 32) iours of well-being. Spontaneous BB in rodents is considered as (Kash et al., 2015). The findings have implications for pain testing in animals. one of the latter (Deacon, 2006), and hence, the association of Most tests of pain in chronic pain models evaluate evoked pain, higher connectivity between ACC and subcortex with reduced in which enhanced defensive behaviours are observed in response burrowing is consistent with a possible functional inhibitory to stimulation of some sort. Such tests are likely to be highly pathway. 6 Brain and Neuroscience Advances sensitive to hyperalgesia and allodynia, but less so to the overall with concurrent neuroimaging could be highly informative with decrease in ‘wellbeing’ that accompanies it. Other tests, designed regard to the identification of potential causal relationships. to capture persistent pain better, such as dynamic weight bearing, may still involve modulation of motor responses related to the Acknowledgements pain associated with movement, but may be less likely to relate Masahide Fujita gave advice for animal transportation. L.S.M., C.S., and purely to affective-motivational component of persistent pain. In K.K. contributed equally to this study. contrast, burrowing is an innate motivated behaviour, and in the case of hind leg CFA injection, modulation by pain is more likely Declaration of conflicting interests to reflect the underlying affective suppression of behaviour in a The author(s) declared no potential conflicts of interest with respect to way not directly related to exacerbation of the pain-inducing the research, authorship, and/or publication of this article. lesion. Given accumulating evidence that the transition to chronic pain can be characterised as the formation of a pathological Funding affective state (Apkarian et al., 2013), pain tests related to the affective-motivational dimension of the sustained pain state This work was supported by Shionogi & Co. Ltd and the National might be more closely related to relevant pathophysiological Institute of Information and Communications Technology. B.S. is also supported by the Wellcome Trust and Arthritis research UK. C.S. is sup- changes associated with disease progression. That suppression of ported by the German Research Foundation. BB relates to ACC connectivity, which is strongly implicated in affective components of pain, further supports this notion, and ORCID iD adds to evidence that burrowing may provide a valuable comple- ment to conventional measures in the evaluation of pain in rodent Christian Sprenger https://orcid.org/0000-0002-0307-7383 models of chronic pain. Assessing motivational parameters in the context of sustained pain in animals also captures information References that has greater relevance to general well-being and higher-order Andrews N, Legg E, Lisak D, et al. (2012) Spontaneous burrowing biological goals compared to purely local measures, which is in behaviour in the rat is reduced by peripheral nerve injury or inflam- line with the recent notion that more natural animal models might mation associated pain. European Journal of Pain 16(4): 485–495. have a better translational validity (Klinck et al., 2017) and like- Apkarian AV, Baliki MN and Farmer MA (2013) Predicting transition to wise with the finding that also in humans local parameters have chronic pain. Current Opinion in Neurology 26(4): 360–367. only limited prognostic value. Baliki MN, Chang PC, Baria AT, et al. (2014) Resting-state functional Several limitations of the study should be noted. Experiments reorganization of the rat limbic system following neuropathic injury. Scientific Reports 4: 6186. were conducted in anesthetised animals and we cannot exclude Büchel C, Bornhovd K, Quante M, et al. (2002) Dissociable neural that the narcotic agent interfered to a certain extent with the responses related to pain intensity, stimulus intensity, and stimulus involved nociceptive mechanisms (Tsurugizawa et al., 2016) or awareness within the anterior cingulate cortex: A parametric single- the observed association between resting state connectivity and trial laser functional magnetic resonance imaging study. The Journal suppression of behaviour. However, even high concentrations of of Neuroscience 22(3): 970–976. isoflurane exert only mild antinociceptive effects and reexamina- Calejesan AA, Kim SJ and Zhuo M (2000) Descending facilitatory mod- tion of the behavioural measures after fMRI measurements ulation of a behavioral nociceptive response by stimulation in the showed no significant changes. Anaesthetics naturally interfere adult rat anterior cingulate cortex. European Journal of Pain 4(1): with the brain-wide spatiotemporal organisation of functional 83–96. networks, but it appears very unlikely that a narcotic agent selec- Cox RW, Chen G, Glen DR, et al. (2017) FMRI clustering in AFNI: False-positive rates redux. Brain Connectivity 7(3): 152–171. tively induced associations between spontaneous fluctuations of Deacon RMJ (2006) Burrowing in rodents: A sensitive method for resting state brain activity and behavioural measures in one group detecting behavioral dysfunction. Nature Protocols 1(1): 118–121. of animals. Durvarci S, Bauer EP, Paré D (2009) Journal of Neuroscience 29(33): Notwithstanding, neuroimaging experiments alone are inher- 10357–10361. ently correlational, and it cannot be assumed that the ACC connec- Eklund A, Nichols TE and Knutsson H (2016) Cluster failure: Why fMRI tions mediate a direct causal suppression of affective-motivational inferences for spatial extent have inflated false-positive rates. Pro- behaviour on the basis of the evidence provided. Even if the con- ceedings of the National Academy of Sciences 113(28): 7900–7905. nectivity is indeed causal, one needs to remain cautious about gen- Friston KJ (1994) Functional and effective connectivity in neuroimaging: eralising to other motivated behaviours, in case the pathway was A synthesis. Human Brain Mapping 2(1–2): 56–78. specific to BB. With this in mind, it is difficult to find an exact Gao Y-J, Ren W-H, Zhang Y-Q, et al. (2004) Contributions of the ante- rior cingulate cortex and amygdala to pain- and fear-conditioned parallel to BB in humans. However, these issues can be potentially place avoidance in rats. Pain 110(1): 343–353. addressed. For instance, the suppression of affective and motivated Iadarola MJ, Berman KF, Zeffiro TA, et al. (1998) Neural activation dur- behaviours could be studied in human pain patients with regard to ing acute capsaicin-evoked pain and allodynia assessed with PET. ACC connectivity. Brain 121(5): 931–947. The employed experimental paradigm also provides interest- Jeon D, Kim S, Chetana M, et al. (2010) Observational fear learning ing perspectives for future studies in rodents. Pharmacological 2+ involves affective pain system and Ca 1.2 Ca channels in ACC. studies using systemic analgesics could investigate correlations Nature Neuroscience 13(4): 482–488. between suppression/reinstatement of BB, ACC connectivity, Kash TL, Pleil KE, Marcinkiewcz CA, et al. (2015) Neuropeptide regu- and the development of the painful condition. And targeted lation of signaling and behavior in the BNST. Molecules and Cells manipulations of the ACC (local pharmacology, optogenetics) 38(1): 1–13. Morris et al. 7 Klinck MP, Mogil JS, Moreau M, et al. (2017) Translational pain assess- Price DD (2000) Psychological and neural mechanisms of the affective ment: Could natural animal models be the missing link? Pain 158(9): dimension of pain. Science 288(5472): 1769–1772. 1633–1646. Rainville P, Duncan GH, Price DD, et al. (1997) Pain affect encoded Kundu P, Santin MD, Bandettini PA, et al. (2014) Differentiating in human anterior cingulate but not somatosensory cortex. Science BOLD and non-BOLD signals in fMRI time series from anesthe- 277(5328): 968–971. tized rats using multi-echo EPI at 11.7 T. NeuroImage 102(Pt 2): Reichman OJ and Smith SC (1990) Burrows and burrowing behavior by 861–874. mammals. Current Mammalogy 2: 197–244. Lenz FA, Rios M, Zirh A, et al. (1998) Painful stimuli evoke potentials Segerdahl AR, Mezue M, Okell TW, et al. (2015) The dorsal posterior recorded over the human anterior cingulate gyrus. Journal of Neuro- insula is not an island in pain but subserves a fundamental role – physiology 79(4): 2231–2234. Response to: ‘Evidence against pain specificity in the dorsal poste- Moayedi M and Davis KD (2012) Theories of pain: From specificity to rior insula’ by Davis et al. F1000Research 4: 1207. gate control. Journal of Neurophysiology 109(1): 5–12. Thompson SJ and Bushnell MC (2012) Rodent functional and anatomical Moisset X and Bouhassira D (2007) Brain imaging of neuropathic pain. imaging of pain. Neuroscience Letters 520(2): 131–139. NeuroImage 37(Suppl 1): S80–S88. Tsurugizawa T, Takahashi Y and Kato F (2016) Distinct effects of iso- Morrison I, Lloyd D, di Pellegrino G, et al. (2004) Vicarious responses flurane on basal BOLD signals in tissue/vascular microstructures in to pain in anterior cingulate cortex: Is empathy a multisensory issue? rats. Scientific Reports 6: 38977. Cognitive, Affective & Behavioral Neuroscience 4(2): 270–278. Vogt BA (2005) Pain and emotion interactions in subregions of the cin- Peyron R, Laurent B and García-Larrea L (2000) Functional imaging of gulate gyrus. Nature Reviews Neuroscience 6(7): 533–544. brain responses to pain. A review and meta-analysis (2000). Neuro- Vogt BA and Peters A (1981) Form and distribution of neurons in rat physiologie Clinique/Clinical Neurophysiology 30(5): 263–288. cingulate cortex: Areas 32, 24, and 29. The Journal of Comparative Ploner M, Gross J, Timmermann L, et al. (2002) Cortical representation Neurology 195(4): 603–625. of first and second pain sensation in humans. Proceedings of the Vogt BA, Finch DM and Olson CR (1992) Functional heterogeneity in National Academy of Sciences 99(19): 12444–12448. cingulate cortex: The anterior executive and posterior evaluative Pohlmann A, Reimann HM, Hentschel J, et al. (2016) Normothermic regions. Cerebral Cortex 2(6): 435–443. mouse functional MRI of acute focal thermostimulation for probing Zhuo M (2007) Neuronal mechanism for neuropathic pain. Molecular nociception. Scientific Reports 6: 17230. Pain 3: 14. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Brain and Neuroscience Advances SAGE

Anterior cingulate cortex connectivity is associated with suppression of behaviour in a rat model of chronic pain:

Download PDF
7 pages

Loading next page...
 
/lp/sage/anterior-cingulate-cortex-connectivity-is-associated-with-suppression-hAV27Jw02m

References (65)

Publisher
SAGE
Copyright
Copyright © 2022 by SAGE Publications Ltd and British Neuroscience Association, unless otherwise noted. Manuscript content on this site is licensed under Creative Commons Licenses
ISSN
2398-2128
eISSN
2398-2128
DOI
10.1177/2398212818779646
Publisher site
See Article on Publisher Site

Abstract

A cardinal feature of persistent pain that follows injury is a general suppression of behaviour, in which motivation is inhibited in a way that promotes energy conservation and recuperation. Across species, the anterior cingulate cortex is associated with the motivational aspects of phasic pain, but whether it mediates motivational functions in persistent pain is less clear. Using burrowing behaviour as an marker of non-specific motivated behaviour in rodents, we studied the suppression of burrowing following painful confirmatory factor analysis or control injection into the right knee joint of 30 rats (14 with pain) and examined associated neural connectivity with ultra-high-field resting state functional magnetic resonance imaging. We found that connectivity between anterior cingulate cortex and subcortical structures including hypothalamic/preoptic nuclei and the bed nucleus of the stria terminalis correlated with the reduction in burrowing behaviour observed following the pain manipulation. In summary, the findings implicate anterior cingulate cortex connectivity as a correlate of the motivational aspect of persistent pain in rodents. Keywords Chronic pain, confirmatory factor analysis pain model, resting state functional magnetic resonance imaging, pain in rodents Received: 11 January 2018; accepted: 4 May 2018 Introduction One of the greatest challenges in the development of novel anal- thalamus, primary and secondary somatosensory cortex, cerebel- gesics is the difficulty in evaluating pain in animal models of lum, the insular cortex, and the anterior cingulate cortex (ACC) chronic pain. Pain is a uniquely subjective experience, and the mainstay of outcome measures in clinical trials remains subjec- Department of Psychology and Behavioral and Clinical Neuroscience tive ratings. Since these are not available in animals, evaluation Institute, University of Cambridge, Cambridge, UK of pain depends on surrogate measures of behaviour, primarily Center for Information and Neural Networks, National Institute of motor responses directly related to evoked or spontaneous pain. Information and Communications Technology, Osaka, Japan However, the difficulty in translating these behaviours to humans Computational and Biological Learning Laboratory, Department of is well recognised, and this greatly hampers the ability to predict Engineering, University of Cambridge, Cambridge, UK whether analgesics that are successful in animals will work in Pain & Neuroscience, Drug Discovery & Disease Research Laboratory, humans. This has led to a requirement for behavioural measures Shionogi & Co., Ltd., Osaka, Japan that better reflect persistent pain in animals, and which have Translational Research Unit, Biomarker R&D Department, Shionogi & translational validity to humans. Co., Ltd., Osaka, Japan Immunology Frontier Research Center, Osaka University, Osaka, Japan Functional neuroimaging offers a novel approach to the eval- Brain Information Communication Research Laboratory Group, Advanced uation of pain in rodents (Baliki et al., 2014; Thompson and Telecommunications Research Institute International, Kyoto, Japan Bushnell, 2012). In humans, a broad set of cortical and subcorti- cal regions are implicated in pain processing and the expression Corresponding author: of pain behaviour. This diversity reflects the fact that pain is a Christian Sprenger, Computational and Biological Learning Laboratory, multidimensional experience, engaging sensory, affective, and Department of Engineering, University of Cambridge, Trumpington cognitive processing. Human studies of phasic experimental Street, Cambridge CB2 1PZ, UK. pain implicate a network of pain-related regions that includes Email: cs910@cam.ac.uk Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Brain and Neuroscience Advances (Moisset and Bouhassira, 2007; Price, 2000). Although the pre- Preparation of pain model rats cise functions of each region remain unclear, there is a consen- Rats were anesthetised with isoflurane (Mylan, Canonsburg, PA, sus that some regions (thalamus, primary and secondary USA) and then intra-articular injected with 25 µL of either somatosensory cortex, cerebellum) better reflect sensori-motor Freund’s Complete Adjuvant (confirmatory factor analysis processes, and others (the anterior insular cortex and in particu- (CFA); Mycobacterium butyricum (BD DIFCO, Franklin Lakes, lar the ACC) better reflect affective processes related to pain NJ, USA); 2 mg/mL of liquid paraffin (Maruishi Pharmaceutical (Büchel et al., 2002; Morrison et al., 2004; Rainville et al., Co., Ltd., Osaka, Japan)), or vehicle (sham rats) into the right 1997). However, there remains uncertainty as to how well stud- knee joint of the hind leg. In both cases, the left hind knee joint ies of phasic experimental pain translate to the persistent pain was kept untreated. Both the pain model and the sham rats were seen in chronic pain models. Although phasic pain occurs as a 6 weeks old at the time of injection. part of many chronic pain conditions, as seen with spontaneous phasic and hyper-sensitive evoked pain, the nature of persistent underlying pain is quite different and serves a distinct behav- Pain behaviour ioural function to reduce activity and conserve energy. The ACC is a strong candidate to play a dominant role in Rats were tested for pain behaviour 18 and 22 days after CFA/ mediating the affective behavioural manifestations of persistent vehicle injection. First, the knee diameter (KD) was determined pain. It serves a primary function as a modulator of the affective using a digital caliper (CD-15CX; Mitutoyo Corporation, tone of visceral, motor and endocrine efferents to downstream Kanagawa, Japan). Second, a weight bearing difference (DWB) regions involved in responses to nociceptive stimuli, including test on each hind leg was performed (Bioseb, Boulogne, France). the periaqueductal grey (PAG), the amygdala (Calejesan et al., Values of WBD are obtained using the following formula, where 2000; Vogt et al., 1992; Zhuo, 2007), and the basal forebrain W corresponds to the amount of weight put on the right leg and (Kash et al., 2015). W to the amount of weight put on the left leg Recently, novel behavioural tasks such as burrowing behav- iour (BB) have been proposed as a measure of the affective %/ WW =× 100 WW + () RR RL impact of pain. BB is an innate, spontaneous behaviour that dem- This formula expresses the percentage of the rat’s body weight put onstrates the overall ‘wellbeing’ or affective-motivational tone of on the CFA/vehicle-injected right leg; thus, a 50% value means the rats (Deacon, 2006) and is not affected by limb hypersensitiv- equal weight distribution across both hind legs. ity (Andrews et al., 2012). Burrowing provides shelter and pro- Third, a grip strength (GS) test (San Diego Instruments, San tection from environmental predators, food storage, and foraging, Diego, CA, USA) was conducted to quantify the mechanical but also comes with energy costs (Reichman and Smith, 1990). strength of the hind legs and hence to control for the development This behaviour is reduced in rats following peripheral nerve of pain in model rats. Briefly, each rat was gently restrained and injury or pain (Andrews et al., 2012), and therefore may provide allowed to grasp the wire mesh frame with its hind limbs and was a useful index of the motivational component of persistent back- moved in a rostral-to-caudal direction until the grip released. ground pain. Finally, BB was investigated. For burrowing experiments, plastic Taken together, these considerations suggest that changes in tubes (32 cm in length and 10 cm in diameter) were filled with ACC connectivity may be related to the modulation of motiva- 3 kg of gravel (5–8 mm particle size) and placed in Plexiglas tional behaviours, such as burrowing, by persistent pain. To test cages (560 × 440 × 200 mm). The open-end of the tube was ele- this hypothesis, we studied BB in adult rats in a model of vated 6 cm from the floor of the cage. Rats were allowed to indi- inflammatory arthritis pain (intra-articular injection of com- vidually burrow during 30 min for 18 days after CFA or vehicle plete Freund’s adjuvant) and measured ACC functional connec- injections were performed and the amount of gravel burrowed tivity in the brain using functional magnetic resonance imaging was recorded. (fMRI). For the statistical comparison of the behavioural data from the CFA pain model and control animals, we employed a two-sample student’s t-test (one-tailed). Results were considered significant Materials and methods at p < 0.05. Experimental animals Seventy-two male LEW/CrlCrlj rats (Charles River Laboratories, Animal selection and transfer to magnetic Japan, Inc.) were used. Rats were housed in groups of three in resonance imaging facilities plastic cages under controlled temperature and humidity and pro- vided free access to food and water under a 12/12 h reversed Based on the results obtained from the WBD on the 18th day, rats light-dark cycle (lighting at 8:00 a.m.). To reduce transfer stress were selected out of the ones that displayed either moderate or (single housing, changing the cage-mates) and negative impact severe pain and were sent to the magnetic resonance imaging (active social interaction or attacks from non-injured rats) from (MRI) facility. Selection criteria for CFA rats were <35% of non-injured rats to injured rats, the rats were housed and trans- WBD value. Functional MRI scanning was conducted at the ported only with rats from their group. All procedures were Center for Information and Neural Networks (CiNet, Osaka approved by internal animal care and use committee of Shionogi University, Suita, Japan). After being transferred to the MRI Pharmaceutical Research Center (Osaka, Japan) instructed by facility, animals were kept for 4 days under standard laboratory Association for Assessment and Accreditation of Laboratory conditions (room temperature of 22°C–23°C and a 12 h light/ Animal Care International (AAALAC) guidelines. dark cycle) with free access to food and water. Rats were group Morris et al. 3 housed with previous cage mates during and after transportation. maximum = 0.5 mm, band pass filtering (0.01 < f < 0.1), and lin- MRI data were acquired on the third day (i.e. 21 days after CFA/ ear regression of motion parameters (3dTproject). vehicle injection). On the fourth day, behavioural data were obtained once more. Assessment of functional connectivity Functional connectivity (Friston, 1994) of the ACC was assessed Resting state functional MRI data acquisition with a seed-based functional connectivity approach. The ACC Data was acquired from a total of 37 rats (18 CFA pain model) seed region was determined as the anterior portion of Brodmann’s with an 11.7 Tesla Avance II vertical bore system (Bruker area 24 (Vogt and Peters, 1981), approximately 1 mm rostral to BioSpin, Ettlingen, Germany) and a home-made transmit/receive Bregma (Calejesan et al., 2000). The mean time series in this surface radio frequency (RF) coil. Rats were anesthetised with a region were extracted for each animal from a sphere of 0.5 mm mixture of air and 2.8% isoflurane (Wako Pure Chemical radius around this coordinate and functional connectivity was Industries Ltd., Osaka, Japan) and then placed in an MRI- determined by calculating brain-wide z-transformed correlation compatible animal cradle. The isoflurane concentration was maps based on the preprocessed functional time series. maintained at 2% ± 0.5%, adjusted to maintain the respiration rate Subsequently, functional connectivity maps were compared at 70 ± 10 breaths per min throughout the sessions. between groups employing a two-sample t-test. Finally, correla- An axial T2-weighted (T2W) imaging was performed using tions between the individual functional connectivity maps and a rapid acquisition of relaxation enhancement (RARE) the behavioural measures (KD, DWB, GS, BB) were calculated sequence (repetition time/echo time (TR/TE) = 6500/45 ms, across animals and compared between groups employing like- number of averages (NA) = 8, field of view (FOV) = 32 × 16 mm, wise t-statistics. matrix size = 256 × 256, slice thickness = 500 µm, acquisition Cluster correction with AFNI’s 3dClustSim (Cox et al., 2017) time = 14 min). was used to reduce the instance of false positives caused by spa- To acquire resting state functional magnetic resonance imag- tial autocorrelation suggested by Eklund et al. (2016) (p < 0.05 ing (rsfMRI) data, we performed gradient-echo echo-planar per voxel plus cluster size k = 15,131 voxel, corrected at imaging (TR/TE = 2500/7 ms, number of segments = 2, flip alpha < 0.01). angle = 60°, FOV = 51.2 × 51.2 mm, matrix size = 64 × 64, slice thickness = 1 mm, in-plane resolution = 800 × 800 µm , band- Results width = 300 kHz, acquisition time = 20 min). Behavioural and physiological measures Imaging data preprocessing KDs were measured to determine the amount of joint swelling as an index of inflammation. Ipsilateral KD in CFA pain model The imaging data preprocessing was performed with AFNI rats was significantly increased compared to vehicle-treated (https://afni.nimh.nih.gov) and FSL (FMRIB, University of rats at day 22 (mean KD CFA-treated rats 10.8 ± 0.15 (SEM), Oxford, UK; https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/MELODIC). mean KD control group 8.7 mm ± 0.03, t(28) = 14.2, p < 0.001; Data from seven animals could not be analysed due to technical Figure 1(a)). CFA pain model rats also showed decreased GS difficulties and resulting poor image quality leaving 30 animals (mean GS CFA animals 931.0 g/kg ± 30.1, mean GS control for the final rsfMRI analysis (14 animals with pain, 16 con- group 1259.0 g/kg ± 10.6, t(28) = 17.4, p < 0.001; Figure 1(b)), trols). Functional data were corrected for slice timing offsets, weight on ipsilateral paw (mean DWB CFA-treated rats with mean and linear detrending (3dTshift) and images were 27.7% ± 1.5, mean DWB control group 49.5% ± 0.7, t(28) = 13.9, realigned to a single middle slice (3dvolreg) for volume-to- p < 0.001, Figure 1(c)), and amount of burrowed gravels (mean volume rigid body correction. Resultant motion and displace- BB CFA-treated rats 444.5 g ± 121.5, mean BB control group ment parameters were visually inspected for outliers. 1238.2 g ± 72.2, t(28) = 5.8, p < 0.001, Figure 1(d)). The results Anatomical data were skull stripped with a standard spherical suggest that CFA pain model rats showed evoked and spontane- model (3dSkullStrip). In order to do this skull strip, anatomical ous inflammatory knee joint pain at the time of fMRI data were ‘shrunk’ in the anterior–posterior direction (3drefit) scanning. to make the brain more spherical, in a manner similar to (Kundu et al., 2014). After effective skull strip, anatomical data were returned to their original voxel dimensions. Functional and ana- Resting state functional connectivity tomical data were then roughly aligned in space (@align_cent- ers) and orientation when oblique (3drotate). Functional data The ACC seed region showed reduced functional connectivity in were up-sampled to match anatomical voxel dimensions (3dre- the CFA pain group compared to control animals with central fit). Anatomical and functional data were then coregistered parts of the contralateral somatosensory cortex and dorsal por- using a 12-paramter registration (3dAllineate). A study-specific tions of the cingulate cortex (Figure 2(a)). In addition, the ACC anatomical template was created by averaging four normal con- exhibited increased functional connectivity in the pain group trol rat anatomical datasets. Anatomical data for each rat were with rostral portions of the somatosensory cortex and in the sub- normalised to this study-specific template and resultant warp cortex with the bilateral striatum, structures of the basal fore- parameters were applied to their respective functional data brain region, hypothalamic region/preoptical area (POA), and (3dAllineate). Functional data were furthermore subjected to the area of the bed nucleus of the stria terminalis (BNST; see de-spiking (3dDespike), smoothing with full width at half Figure 2(a)). 4 Brain and Neuroscience Advances Figure 1. Behavioural and physiological changes of CFA pain model rats compared to controls at the time of fMRI scanning. (a) Changes in contralateral and ipsilateral knee diameter (KD) in vehicle and CFA rats at day 22. (b) Changes in hind limb grip strength (GS) in vehicle and CFA rats. Data are shown as an average value of days 18 and 22 expressed as grip strength/body weight. (c) Changes in dynamic weight bearing (WBD) on the ipsilateral paw in vehicle and CFA rats. Each data were shown as an average value of days 18 and 22. (d) Changes in burrowing behaviour (BB) in vehicle treated and CFA pain model rats at day 18. *** indicates p < 0.001. In CFA pain animals, we found a marked negative correlation ability, and reduced weight bearing. As hypothesised, we also between ACC functional connectivity and innate BB in the latter observed reduced BB accounting on average for less than 50% of regions (POA & BNST; Figure 2(b)), that is, higher functional the behaviour in control animals. Resting state fMRI revealed a connectivity between ACC and these regions was associated with pronounced negative correlation between the individual strength suppression of BB. of ACC functional connectivity and BB in structures of the basal For illustration purposes, Figure 2(c) shows the correlation of forebrain, comprising the hypothalamic region and the BNST. connectivity strength between ACC with the peak subcortical No correlations with the ACC functional connectivity were region of interest and BB. ACC connectivity shows a negative observed for motor responses directly associated with joint pain. correlation with burrowing in the CFA pain group (r = –0.63, The findings therefore indicate a relatively specific relationship p = 0.02) but not in the control group (r =−0.23, p = 0.39). The between ACC functional coupling and the suppression of BB by interaction failed, however, significance (z = 1.2, p = 0.11). We persistent pain and suggest that ACC connectivity might be a found no significant correlations between ACC connectivity with good marker for the affective-motivational component of pain in KD, weight bearing, and GS. rodents. Across mammalian species neuronal responses to noxious stimuli have been shown to involve a number of brain regions Discussion including the primary and secondary somatosensory cortex, the In the present study, we assessed spontaneous BB in the CFA ACC, and the insular cortex (Iadarola et al., 1998; Lenz et al., model of persistent inflammatory arthritis pain and investigated 1998; Peyron et al., 2000; Ploner et al., 2002; Pohlmann et al., ACC resting state functional connectivity compared to controls. 2016; Thompson and Bushnell, 2012). The behavioural measures consistently indicated inflammatory Although the precise contribution of each region to the gen- nociception related to the CFA-treated joint at the time of MRI eration of pain remains to be elucidated (Moayedi and Davis, 2012; Segerdahl et al., 2015), neuroimaging studies highlighted measurement by showing ipsilateral swelling, reduced motoric Morris et al. 5 Figure 2. Changes in resting state functional connectivity in CFA pain model rats compared to control and correlation with burrowing behaviour. (a) Changes in ACC resting state functional connectivity in CFA pain rats compared to vehicle-treated animals. The ACC showed increased functional connectivity in CFA pain rats with parts of the contralateral somatosensory cortex (upper and lower panels), the bilateral striatum (upper and middle panels) and with the basal forebrain region/preoptical area and the region of the bed nucleus of the stria terminalis (lower panel). ACC functional connectivity was decreased in parts of the cingulate cortex itself and in the somatosensory cortex (middle panel). The three white lines in the smaller panel next to the upper panel of (a) indicate the approximate rostro-caudal position of the three axial sections of (a) in the rat brain. The colour bar indicates t values. Hot colours represent increased functional connectivity and cold colours represent decreased connectivity. A schematic representation of the involved subcortical regions is shown next to the lower panel of (a). (1) Hypothalamic/preoptical area, (2) basal forebrain region, (3) bed nucleus of the stria terminalis, (4) globus pallidus, (5) striatum, (6) septal region. (b) Negative correlation of burrowing behaviour with the individual strength of ACC functional connectivity. The colour bar indicates t values. The visualisation threshold is set to p < 0.005 uncorrected in (a) and (b). Note the spacial overlap of increased ACC-subcortical connectivity in (a) with the negative connectivity–behaviour correlation in (b) and the slightly different positions of the axial sections in (b) compared to (a). (c) Illustration of the negative correlation between burrowing behaviour with the individual connectivity strength at the subcortical peak in the upper panel of (b). ACC connectivity shows a marked negative correlation with burrowing in the CFA pain group but not in the control group. some general characteristics of the involved regions. While the While the association between ACC-hypothalamic coupling somatosensory cortex responds faster to dimensional features of and BB might reflect a general homeostatic dimension related to noxious stimuli, and is more engaged when larger body surfaces the animal’s behavioural suppression, the correlation between are exposed to pain, the ACC as part of the limbic system is burrowing and coupling strength with the BNST is especially thought to be modulated by the affective relevance pertinent to a interesting as it has been linked to sustained vigilance associated change in motivational tone or response selection (Peyron et al., with ambiguous or distant threat cues. This potentially provides 2000; Ploner et al., 2002). a direct link to the affective dimension of pain and illustrates The ACC is often conceptualised as a nexus for the processing BB as an expression of the affective-motivational tone of the of external salient stimuli, autonomic response regulation, and animals.That is, when a limb is injured, the potential danger by subsequent affective learning (Gao et al., 2004; Vogt, 2005). For predators and accordingly monitoring of potential threats example, fear learning in mice through observing other mice becomes much more important, so that behaviours such as bur- receiving painful foot shocks has been demonstrated to involve rowing are no longer prioritised. In line with this, lesions of the the ACC (Jeon et al., 2010). In the context of pain, the ACC is BNST have been demonstrated to disrupt the individual variabil- thought to allow potentially harmful stimuli to engage appropri- ity in the rodent’s anxiety-like behaviour (Durvarci, 2009). ate affective and motivationally relevant behaviours, and as such Finally, the role of the BNST as a site of integration of limbic one would expect the strength of ACC-based connectivity to be forebrain information is also supported by tracing studies show- associated with elevated aversive behaviors and reduced behav- ing direct anatomical BNST connections with the ACC (area 32) iours of well-being. Spontaneous BB in rodents is considered as (Kash et al., 2015). The findings have implications for pain testing in animals. one of the latter (Deacon, 2006), and hence, the association of Most tests of pain in chronic pain models evaluate evoked pain, higher connectivity between ACC and subcortex with reduced in which enhanced defensive behaviours are observed in response burrowing is consistent with a possible functional inhibitory to stimulation of some sort. Such tests are likely to be highly pathway. 6 Brain and Neuroscience Advances sensitive to hyperalgesia and allodynia, but less so to the overall with concurrent neuroimaging could be highly informative with decrease in ‘wellbeing’ that accompanies it. Other tests, designed regard to the identification of potential causal relationships. to capture persistent pain better, such as dynamic weight bearing, may still involve modulation of motor responses related to the Acknowledgements pain associated with movement, but may be less likely to relate Masahide Fujita gave advice for animal transportation. L.S.M., C.S., and purely to affective-motivational component of persistent pain. In K.K. contributed equally to this study. contrast, burrowing is an innate motivated behaviour, and in the case of hind leg CFA injection, modulation by pain is more likely Declaration of conflicting interests to reflect the underlying affective suppression of behaviour in a The author(s) declared no potential conflicts of interest with respect to way not directly related to exacerbation of the pain-inducing the research, authorship, and/or publication of this article. lesion. Given accumulating evidence that the transition to chronic pain can be characterised as the formation of a pathological Funding affective state (Apkarian et al., 2013), pain tests related to the affective-motivational dimension of the sustained pain state This work was supported by Shionogi & Co. Ltd and the National might be more closely related to relevant pathophysiological Institute of Information and Communications Technology. B.S. is also supported by the Wellcome Trust and Arthritis research UK. C.S. is sup- changes associated with disease progression. That suppression of ported by the German Research Foundation. BB relates to ACC connectivity, which is strongly implicated in affective components of pain, further supports this notion, and ORCID iD adds to evidence that burrowing may provide a valuable comple- ment to conventional measures in the evaluation of pain in rodent Christian Sprenger https://orcid.org/0000-0002-0307-7383 models of chronic pain. Assessing motivational parameters in the context of sustained pain in animals also captures information References that has greater relevance to general well-being and higher-order Andrews N, Legg E, Lisak D, et al. (2012) Spontaneous burrowing biological goals compared to purely local measures, which is in behaviour in the rat is reduced by peripheral nerve injury or inflam- line with the recent notion that more natural animal models might mation associated pain. European Journal of Pain 16(4): 485–495. have a better translational validity (Klinck et al., 2017) and like- Apkarian AV, Baliki MN and Farmer MA (2013) Predicting transition to wise with the finding that also in humans local parameters have chronic pain. Current Opinion in Neurology 26(4): 360–367. only limited prognostic value. Baliki MN, Chang PC, Baria AT, et al. (2014) Resting-state functional Several limitations of the study should be noted. Experiments reorganization of the rat limbic system following neuropathic injury. Scientific Reports 4: 6186. were conducted in anesthetised animals and we cannot exclude Büchel C, Bornhovd K, Quante M, et al. (2002) Dissociable neural that the narcotic agent interfered to a certain extent with the responses related to pain intensity, stimulus intensity, and stimulus involved nociceptive mechanisms (Tsurugizawa et al., 2016) or awareness within the anterior cingulate cortex: A parametric single- the observed association between resting state connectivity and trial laser functional magnetic resonance imaging study. The Journal suppression of behaviour. However, even high concentrations of of Neuroscience 22(3): 970–976. isoflurane exert only mild antinociceptive effects and reexamina- Calejesan AA, Kim SJ and Zhuo M (2000) Descending facilitatory mod- tion of the behavioural measures after fMRI measurements ulation of a behavioral nociceptive response by stimulation in the showed no significant changes. Anaesthetics naturally interfere adult rat anterior cingulate cortex. European Journal of Pain 4(1): with the brain-wide spatiotemporal organisation of functional 83–96. networks, but it appears very unlikely that a narcotic agent selec- Cox RW, Chen G, Glen DR, et al. (2017) FMRI clustering in AFNI: False-positive rates redux. Brain Connectivity 7(3): 152–171. tively induced associations between spontaneous fluctuations of Deacon RMJ (2006) Burrowing in rodents: A sensitive method for resting state brain activity and behavioural measures in one group detecting behavioral dysfunction. Nature Protocols 1(1): 118–121. of animals. Durvarci S, Bauer EP, Paré D (2009) Journal of Neuroscience 29(33): Notwithstanding, neuroimaging experiments alone are inher- 10357–10361. ently correlational, and it cannot be assumed that the ACC connec- Eklund A, Nichols TE and Knutsson H (2016) Cluster failure: Why fMRI tions mediate a direct causal suppression of affective-motivational inferences for spatial extent have inflated false-positive rates. Pro- behaviour on the basis of the evidence provided. Even if the con- ceedings of the National Academy of Sciences 113(28): 7900–7905. nectivity is indeed causal, one needs to remain cautious about gen- Friston KJ (1994) Functional and effective connectivity in neuroimaging: eralising to other motivated behaviours, in case the pathway was A synthesis. Human Brain Mapping 2(1–2): 56–78. specific to BB. With this in mind, it is difficult to find an exact Gao Y-J, Ren W-H, Zhang Y-Q, et al. (2004) Contributions of the ante- rior cingulate cortex and amygdala to pain- and fear-conditioned parallel to BB in humans. However, these issues can be potentially place avoidance in rats. Pain 110(1): 343–353. addressed. For instance, the suppression of affective and motivated Iadarola MJ, Berman KF, Zeffiro TA, et al. (1998) Neural activation dur- behaviours could be studied in human pain patients with regard to ing acute capsaicin-evoked pain and allodynia assessed with PET. ACC connectivity. Brain 121(5): 931–947. The employed experimental paradigm also provides interest- Jeon D, Kim S, Chetana M, et al. (2010) Observational fear learning ing perspectives for future studies in rodents. Pharmacological 2+ involves affective pain system and Ca 1.2 Ca channels in ACC. studies using systemic analgesics could investigate correlations Nature Neuroscience 13(4): 482–488. between suppression/reinstatement of BB, ACC connectivity, Kash TL, Pleil KE, Marcinkiewcz CA, et al. (2015) Neuropeptide regu- and the development of the painful condition. And targeted lation of signaling and behavior in the BNST. Molecules and Cells manipulations of the ACC (local pharmacology, optogenetics) 38(1): 1–13. Morris et al. 7 Klinck MP, Mogil JS, Moreau M, et al. (2017) Translational pain assess- Price DD (2000) Psychological and neural mechanisms of the affective ment: Could natural animal models be the missing link? Pain 158(9): dimension of pain. Science 288(5472): 1769–1772. 1633–1646. Rainville P, Duncan GH, Price DD, et al. (1997) Pain affect encoded Kundu P, Santin MD, Bandettini PA, et al. (2014) Differentiating in human anterior cingulate but not somatosensory cortex. Science BOLD and non-BOLD signals in fMRI time series from anesthe- 277(5328): 968–971. tized rats using multi-echo EPI at 11.7 T. NeuroImage 102(Pt 2): Reichman OJ and Smith SC (1990) Burrows and burrowing behavior by 861–874. mammals. Current Mammalogy 2: 197–244. Lenz FA, Rios M, Zirh A, et al. (1998) Painful stimuli evoke potentials Segerdahl AR, Mezue M, Okell TW, et al. (2015) The dorsal posterior recorded over the human anterior cingulate gyrus. Journal of Neuro- insula is not an island in pain but subserves a fundamental role – physiology 79(4): 2231–2234. Response to: ‘Evidence against pain specificity in the dorsal poste- Moayedi M and Davis KD (2012) Theories of pain: From specificity to rior insula’ by Davis et al. F1000Research 4: 1207. gate control. Journal of Neurophysiology 109(1): 5–12. Thompson SJ and Bushnell MC (2012) Rodent functional and anatomical Moisset X and Bouhassira D (2007) Brain imaging of neuropathic pain. imaging of pain. Neuroscience Letters 520(2): 131–139. NeuroImage 37(Suppl 1): S80–S88. Tsurugizawa T, Takahashi Y and Kato F (2016) Distinct effects of iso- Morrison I, Lloyd D, di Pellegrino G, et al. (2004) Vicarious responses flurane on basal BOLD signals in tissue/vascular microstructures in to pain in anterior cingulate cortex: Is empathy a multisensory issue? rats. Scientific Reports 6: 38977. Cognitive, Affective & Behavioral Neuroscience 4(2): 270–278. Vogt BA (2005) Pain and emotion interactions in subregions of the cin- Peyron R, Laurent B and García-Larrea L (2000) Functional imaging of gulate gyrus. Nature Reviews Neuroscience 6(7): 533–544. brain responses to pain. A review and meta-analysis (2000). Neuro- Vogt BA and Peters A (1981) Form and distribution of neurons in rat physiologie Clinique/Clinical Neurophysiology 30(5): 263–288. cingulate cortex: Areas 32, 24, and 29. The Journal of Comparative Ploner M, Gross J, Timmermann L, et al. (2002) Cortical representation Neurology 195(4): 603–625. of first and second pain sensation in humans. Proceedings of the Vogt BA, Finch DM and Olson CR (1992) Functional heterogeneity in National Academy of Sciences 99(19): 12444–12448. cingulate cortex: The anterior executive and posterior evaluative Pohlmann A, Reimann HM, Hentschel J, et al. (2016) Normothermic regions. Cerebral Cortex 2(6): 435–443. mouse functional MRI of acute focal thermostimulation for probing Zhuo M (2007) Neuronal mechanism for neuropathic pain. Molecular nociception. Scientific Reports 6: 17230. Pain 3: 14.

Journal

Brain and Neuroscience Advances – SAGE

Published: Jun 5, 2018

Keywords: Chronic pain; confirmatory factor analysis pain model; resting state functional magnetic resonance imaging; pain in rodents

There are no references for this article.