Effects of posterior hypothalmic lesions on formalin induced pain behaviours 1

EFFECTS OF POSTERIOR HYPOTHALAMIC LESIONS ON FORMALIN-INDUCED PAIN BEHAVIOURS LIU LIMENG NATIONAL UNIVERSITY OF SINGPAORE 2010 EFFECTS OF POSTERIOR HYPOTHALAMIC LESIONS ON FORMALIN-INDUCED PAIN BEHAVIOURS LIU LIMENG (B.SCI.(HONS), NUS) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF SCIENCE DEPARTMENT OF PHYSIOLOGY NATIONAL UNIVERSITY OF SINGPAORE 2010 ACKNOWLEGEMENT I would like to express my thanks to the following people whose help and support I greatly appreciate: A/P Sanjay Khanna for giving me the opportunity to work in his laboratory. For his patient guidance and assistance throughout the project and for the mind-opening discussions without which this thesis would not have been possible. Esther for her assistance and encouragement throughout the project. Zacky, Andy, Seok Ting, Jenn, and Brenda whom I have worked with, for their assistance throughout the project and also for their company which made the laboratory experience a truly enjoyable one. i TABLE OF CONTENTS TITLE PAGE ACKNOWLEDGEMENT i TABLE OF CONTENTS ii LIST OF FIGURES v LIST OF TABLES viii LIST OF ABBREVIATIONS ix LIST OF PUBLICATIONS xi SUMMARY xii INTRODUCTION 1 1.1 Functional organisation of the hippocampal formation 2 1.1.1 Hippocampus cytoarchitecture 3 1.1.2 Intrinsic connection within the hippocampus formation 6 1.1.2.1 Perforant pathway from the entorhinal cortex to the dentate gyrus 7 1.1.2.2 The mossy fibres projection from the dentate gyrus to CA3 7 1.1.2.3 The Schaffer collaterals to CA1 9 1.1.3 Extrinsic connection of hippocampus formation 11 1.1.3.1 Septo-hippocampal connections 11 1.1.3.2 Hippocampal-septal projection 12 ii 1.1.4 Hippocampal theta rhythm 1.2 Anatomy of supramammillary area (SuM) 13 15 1.2.1 Projections from SuM 16 1.2.2 Projections to SuM 18 1.2.3 SuM and hippocampal theta 19 1.2.4 Effects of manipulations of SuM on behaviour and hippocampal theta 21 1.2.4.1 Spatial memory 21 1.2.4.2 Exploratory and defensive behaviour 22 1.2.4.3 Passive avoidance and Fixed Interval schedule 22 1.3 Pain 24 1.3.1 General description of pain 24 1.3.2 Pain transmission pathway 27 1.3.3 Formalin model of persistent pain 32 1.4 c-Fos 37 1.4.1 General description of c-Fos 37 1.4.2 Fos a an index of spinal nociceptive information processing 39 1.5 AMPA 41 1.6 Rationale and objectives of the present study 42 MATERIALS AND METHODS 44 2.1 Animals 45 2.2 Surgery procedure 45 2.3 Behavioural tests 46 iii 2.3.1 Scheduling of the behavioural tests 46 2.3.2 Open field test 46 2.3.3 Formalin test 47 2.4 Immunocytochemistry 51 2.4.1 Tissue preparation 51 2.4.2 Staining procedures 51 2.5 Data analysis 53 2.5.1 Quanitification of the size of the lesion in the brain 53 2.5.2 Quantification of Fos positive cells in spinal cord and hippocampus 2.5.3 Statistical analysis 54 57 RESULTS 60 3.1 AMPA induced lesion in PH-SuM region 61 3.2 Effects of PH-SuM lesion on animal exploratory behaviour 76 3.3 Effects of PH-SuM lesion on animal behaviour and spinal c-Fos in the formalin test 3.4 Effects of PH-SuM lesion on hippocampal c-Fos in the formalin test 77 91 DISCUSSION 100 4.1 Lesions 101 4.2 Open field 102 4.3 Formalin test 104 4.4 Summary and implications 108 iv REFERENCES 109 LIST OF FIGURES Figure 1 A three-dimensional organisation hippocampal system in the rat brain. Figure 2 The trisynaptic loop of the hippocampus. Figure 3 Schematic diagram of projections from supramammillary nucleus (SuM) to medial septum (MS), lateral septum (LS) and hippocampus 17 Figure 4 Open field apparutus. 48 Figure 5 Formalin test apparatus. 50 Figure 6 Diagrammatic representation of lumbar L4 spinal cord coronal section with laminar subdivisions. 56 Figure 7 Diagrammatic representation of anterior dorsal field CA 1 (bregma -1.72 to bregma -4.65 mm Paxinos and Watson, 2007). 58 Figure 8 Diagrammatic representation of posterior dorsal field CA 1 and the ventral CA1 (bregma -4.65 to bregma 5.76 mm; Paxinos and Watson, 2007). 59 Figure 9 Digital images (500 dpi) of coronal sections of the supramammillary (SuM) region showing microgliosis following pretreatment with the immunotoxin AMPA. 62 Figure 10 Diagrammatic representation of the supramammillary (SuM) region. 64 Figure 11 Diagrammatic representation of the largest (hatched) and smallest (black and white checked) lesion at each plane of section in each of the groups formed for behavioural analysis. 65 Figure 12 Digitized images (500 dpi) of the lesion at each plane of section in an animal with mSuM lesion 67 Figure 13 Illustration of the calculations to determine the extent of the lesion. 70 v of the septo- 5 8 Figure 14 Illustration of the calculations to determine the extent of the lesion against total volume of the supramammillary region (SuM). 72 Figure 15 Illustration of the calculations to determine the extent of the lesion in the medial and lateral supramammillary regions (mSuM and SuML, respectively) against total volume of the lesion. 73 Figure 16 The time course of the distance traveled upon exposure to novel environment. 79 Figure 17 The total distance traveled over 20min in a novel environment. 80 Figure 18 Different ambulatory behaviors in animals with lesion in the medial and lateral supramammillary regions (mSuM and SuML, respectively). 81 Figure 19 The time course of formalin-induced nociceptive behaviors in animals with lesion of the medial or lateral supramammillary region (mSuM and SuML, respectively, on the figure) or the region dorsal to SuM (Dorsal on the figure). 85 Figure 20 The total number of flinches (A), total duration of lick (B), and total ambulation distance (C) in the second phase of formalin test (11-60 min). 86 Figure 21 Digitized section (500 dpi) of lumbar (L4) spinal cord showing expression of c-Fos positive cells (relatively dark compared to the background) following right hind paw injection of formalin (1.25%, 0.1ml). 87 Figure 22 The lack of effect of lesion of the medial and lateral supramammillary region (mSuM and SuML, respectively, on the figure) and of the region dorsal to SuM (Dorsal on the figure) on c-Fos positive cell counts from whole spinal cord (A) and the different spinal laminae (B-E) ipsilateral to formalin injection. 90 Figure 23 c-Fos expression in anterior dorsal field CA1 following injection of formalin (0.1 ml, 1.25%, subcutaneous into right hind paw). 92 Figure 24 c-Fos expression in posterior dorsal field CA1 and ventral CA1 following injection of formalin (0.1 ml, 1.25%, subcutaneous into right hind paw). 93 Figure 25 The lack of effect of lesion of the medial and lateral 95 vi supramammillary regions (mSuM and SuML, respectively, on the figure) and of the dorsal region (Dorsal on the figure) on c-Fos positive cell counts from hippocampal field CA1 to formalin injection (0.1ml, 1.25%, right hind paw). Figure 26 The lack of effect of lesion of the medial and lateral supramammillary regions (mSuM and SuML, respectively, on the figure) and of the dorsal region (Dorsal on the figure) on c-Fos positive cell counts from hippocampal field CA3 and dentate gyrus (DG) to formalin injection (0.1ml, 1.25%, right hind paw). 96 Figure 27 C-Fos expression in dorsal field of the hippocampus following subcutaneous injection of formalin (0.1 ml, 1.25%) into right hind paw. 97 Figure 28 The lack of lateralization of effect of lesion of the medial and lateral supramammillary regions (mSuM and SuML, respectively, on the figure) and of the dorsal region (Dorsal on the figure) on c-Fos positive cell counts from the hippocampus. 99 vii LIST OF TABLES Table 1 Extent of lesion within medial supramammillary and lateral supramammillary against whole mSuM and the whole unilateral SuML region, respectively. 68 Table 2 Extent of lesion of medial supramammillary and lateral supramammillary against whole supramammillary. 74 Table 3 Extent of lesion of medial supramammillary and lateral supramammillary against whole lesion. 75 Table 4 Summary of the groups of animals that were evaluated for the effects of lesion of the posterior hypothalamus (PH) - supramammillary (SuM) region on exploration. 78 Table 5 Summary of the groups of animals that were evaluated for the effects of lesion of the posterior hypothalamus (PH) - supramammillary (SuM) region on indices of formalin pain. 82 Table 6 Summary of the groups of animals that were evaluated for the effects of lesion of the posterior hypothalamus (PH) - supramammillary (SuM) region on spinal c-Fos in the formalin test. 88 Table 7 Summary of the groups of animals that were evaluated for the effects of lesion of the posterior hypothalamus (PH) - supramammillary (SuM) region on hippocampal c-Fos in the formalin test. 94 viii LIST OF ABBREVIATIONS ABC AMPA AP-1 AP-1RE BDNF CaMKs CHAT CNS CRE CREB DAB DG dpi EC ff FLI FI GABA HCN IASP IEG i.p. I/R LS MAPK mf mSuM MS MS-VLDBB avidin-biotin-peroxidase complex α-amino-3-hydroxyl-5-methyl-4isoxazole-propionate activator protein-1 activator protein-1 response element brain-derived neurotophic factor calmodulin-independent protein kinases choline acetyltransferase central nervous system cAMP response element cAMP-responsive element-binding protein diaminobenzidine treatment dentate gyrus dotes per square inch entorhinal cortex fimbria fornix Fos-like immunoreactivity fixed interval schedule test gamma-amino butyric acid hyperpolarization-induced cation channel International Association for the Study of Pain immediate early gene NMDA NK1 NS OF intraperitoneal infrared lateral septum mitogen activated protein kinase Mossy fibers medial supramammillary Medial septum medial septal nuclei and the vertical limb of diagonal band of Broca N-methyl-D-aspartate neurokinin 1 nociceptive specific open field O-LM PAG PDGF PET oriens-lacunosum moleculare periaqueductal gray platelet-derived growth factor positron emission tomography ix PHA-L PH-SuM PKA pp PV REM RPO sc SIE SuM SuML TRP WDR WGA-HRP Phaseolus vulgaris leuco-agglutinin Posterior hypothalamussupramammillary protein kinase A Perporant pathway parvalbumin rapid eye movement rostral periolivary region Shaffer collateral sis-inducible element supramammillary Lateral supramammillary transient receptor potential family of ion channels wide dynamic range wheat germ agglutinin conjugated to horseradish peroxidase x LIST OF PUBLICATIONS ABSTRACT Liu LM and Khanna S (2009) Effects of posterior hypothalamic lesions on formalin-induced pain behaviours. Proceedings of the International Australasian Winter Conference on Brain Research, 27: 4.43, Aug 29-Sep 2, Queenstown, New Zealand. xi SUMMARY The posterior hypothalamus (PH)-supramammillary (SuM) region connects to variety of regions in the CNS, some of which influence affective-motivational behaviors. For example, the region is reciprocally connected to the hippocampal formation. The hippocampal formation is crucial for adaptive behaviors and, importantly, contributes to the negative affective-motivational state during pain. In the present study we explored whether PH-SuM as well modulate animal pain behaviors in the formalin model of persistent inflammatory pain. Formalin (1.25%, 0.1ml) was injected subcutaneously into the plantar surface of the right hind paw of PH-SuM lesioned or control nonlesioned animals. The lesion was induced by microinjection of the glutamate receptor ligand, α-amino-3-hydroxyl-5-methyl-4-isoxazole-propionate (AMPA, 0.99ng in 0.3µl). The lesioned animals were subdivided into a ‘dorsal lesion group’ and a ventral lesion group based on the affected area visualized using OX42 immunocytochemistry as a marker for microglia activation. The damaged area in the dorsal lesion group included the PH and adjacent areas, whereas the ventral lesion extended into the lateral SuM. The lesioned area was bilateral or either ipsilateral or contralateral to the injected paw. The lesion of SuM was incomplete. Neither dorsal nor ventral lesion affected animal behavior in open field. Interestingly, ventral lesion encompassing the lateral SuM attenuated animal behavior to formalin. The effect was more marked during the later part of the formalin response leading to truncated duration of formalin pain-induced behaviors. An effect was also seen with dorsal lesion but was not as marked. xii These trends raise the possibility that PH-SuM region, especially ventrally influence the affectivemotivational drive to pain. xiii INTRODUCTION 1 The supramammillary area (SuM) of the posterior hypothalamus has of late gained prominence because of its modulation of hippocampal neural activity (Pan and McNaughton, 2004). Indeed, it is part of an ascending relay from the brainstem that influences hippocampal synaptic excitability and neural synchronization via the medial septum. Further, much of the behavioural effects observed with manipulation of SuM mimic that seen with medial sepal and hippocampal manipulations. Thus, the SuM is partly viewed as a component of the neuraxis that also includes the septohippocampal formation. The present thesis is built from this perspective and thus first describes the septohippocampal region followed by a review of the SuM. 1.1 Functional organization of the hippocampal formation The term hippocampal formation generally applies to the dentate gyrus, the subicular complex (which is divided into three subdivisions: the subiculum, pre-subiculum and parasubiculum), the entorhinal cortex (which is divided into medial and lateral subdivisions) and the hippocampus proper (Amaral and Witter, 1989). Hippocampus proper can be divided into three sub-fields, the Cornu Ammonis fields CA1, CA2 and CA3. CA2 represents only a very small portion of the hippocampus and its presence is often ignored in accounts of hippocampal function (Johnston and Amaral, 1998). 2 The three-dimensional hippocampal formation in the rodent extends in a long axis bending in a C-shaped fashion from the septal nuclei of the basal forebrain rostrodorsally to the incipient temporal lobe caudoventrally (Amaral and Witter, 1989). The long axis and orthogonal axis of the hippocampal formation are referred as septotemporal axis and transverse axis, respectively (Fig 1). 1.1.1 Hippocampus cytoarchitecture There are two types of neurons in the hippocampal formation, namely principal and non-principal neurons. The principal neurons include pyramidal cells in CA fields and granule cells in the dentate gyrus. They form a narrow compact cell layer (3-8 somta in thickness) and comprise about 90% of the cell population in the hippocampal formation (Olbrich and Braak, 1985). CA fields are structured depth wise in clearly defined dorsal-ventral strata: 1) the alveus is the most superficial layer and contains the axons from pyramidal neurons; 2) stratum oriens is the next layer below the alveus. The cell bodies of inhibitory basket cells, horizontal trilaminar cells and the basal dendrites of pyramidal neurons are located in this stratum; 3) stratum pyramidale contains the cell bodies of the pyramidal neurons which are the principal excitatory neurons of the hippocampus; 4) stratum radiatum which, like stratum oriens, contains septal and commissural fibres. It also contains Schaffer collateral fibres which are the projection fibres from CA3 to CA1. Some interneurons including basket cells, bistratified cells, and radial trilaminar cells can also be 3 found here; 5) stratum lacunosum is a thin stratum that also contains Schaffer collateral fibres and perforant path fibres from the superficial layers of entorhinal cortex. Due to its small size, it is often grouped together with stratum moleculare into a single stratum called stratum lacunosum-moleculare; 4 Figure 1. A three-dimensional organisation of the septo-hippocampal system in the rat brain. The hippocampus is the C-shaped structure. Abbreviations: fx = fornix; fi = fimbria; HC = hippocampus; MS = medial septum (modified from Amaral and Witter 1995; picture adapted from (Ikonen, 2001). 5 6) stratum moleculare contains the perforant path fibres which form synapses onto the distal, apical dendrites of pyramidal cells (Johnston and Amaral, 1998). Granule cells in dentate gyrus have a cone-shaped dendritic tree towards the hippocampal fissure, with all branches confined to molecular layer (a relatively cell free layer). The three layers in dentate gyrus are as follow: 1) the polymorphic layer is the most superficial layer of the dentate gyrus and is often considered as a separate subfield; 2) stratum granulosum contains the cell bodies of the dentate granule cells; 3) stratum moleculare is where both commissural fibres from the contralateral dentate gyrus run and form synapses as well as where inputs from the medial septum terminate, both on the proximal dendrites of the granule cells (Johnston and Amaral, 1998). The non-principal cells mainly refer to local inhibitory interneurons scattered throughout the hippocampal formation (Olbrich and Braak, 1985). Examples of such interneurons include the basket cells, horizontal trilaminar cells, bistratified cells, and radial trilaminar cells. The interneurons play a critical role in modulation of principal cell excitability and activity via local circuits (Freund and Buzsaki, 1996). 1.1.2 Intrinsic connection within the hippocampus formation A unidirectional intrinsic connection which consists of a series of excitatory pathways enables a sequential flow of information through the hippocampal 6 formation. It is organized in a trisynaptic circuit, starting from the perforant pathway from the entorhinal cortex to the dentate gyrus, continues with the mossy fibres projection from the dentate gyrus to CA3 and ends with the Schaffer collaterals to CA1 (Fig 2). 1.1.2.1 Perforant pathway from the entorhinal cortex to the dentate gyrus The perforant pathway is the major input to the hippocampus. The axons of the perforant path arise principally in layers II and III of the entorhinal cortex, with minor contributions from the deeper layers IV and V. Axons from layers II/IV project to the granule cells of the dentate gyrus and pyramidal cells of the CA3 region, while those from layers III/V project to the pyramidal cells of the CA1 and the subiculum (Johnston and Amaral, 1998). 1.1.2.2 The mossy fibres projection from the dentate gyrus to CA3 The unmyelinated mossy fibres arise from granule cells and are organized in a lamellar fashion, with little divergence in septotemporal axis (Amaral and Witter, 1989). All mossy fibres extend throughout the full transverse extent of CA3 field despite the location of granule cells of origin. They are selective in their contacts in which each axon only contacts fourteen pyramidal cells 7 Figure 2. The trisynaptic loop of the hippocampus. The filled triangles represent the pyramidal cell layer (CA1 and CA3) and the filled circles represent the granular cell layer of the dentate gyrus. The trisynaptic circuit is composed of (1) Perforant pathway (pp) fibres which arise from EC and terminate on granule cells of DG, (2) mossy fibres (mf) of the granule cells which contact pyramidal cells of field CA3, and (3) Schaffer collaterals (sc) which synapse onto pyramidal cells in field CA1. Abbreviations: EC = entorhinal cortex; DG = dentate gyrus; pp = perforant pathway; mf = mossy fibres; sc = Schaffer collaterals; ff = fimbria fornix. (Picture adapted from Ikonen, 2001). 8 (Claiborne et al., 1986). Most of mossy fibres terminate in the proximal dendritic region of ipsilateral CA3 pyramidal cells. 1.1.2.3 The Schaffer collaterals to CA1 Schaffer collaterals arise from CA3 pyramidal neurons in ipsilateral hippocampus and projects both in the septotemporal axis and in the transverse axis topographically. For instance, CA3 pyramidal cells located close to dentate gyrus tend to project more heavily to the septal level of CA1 and near the subicular border (Amaral, 1993) and terminate mostly in the superficial stratum radiatum of CA1. On the other hand, those located close to CA2 tend to project to temporal CA1 and terminate heavily in both stratum radiatum and stratum oriens. The septotemporal divergence of Schaffer collateral from a CA3 pyramidal cell is as much as three-fourths of the entire septotemporal extent of CA1 (Ishizuka et al., 1990). In addition, CA3 pyramidal neurons also send projections to CA1 pyramidal cells via the Schaffer collaterals and join with the commissural fibres from the CA3 of the contralateral hippocampus (Amaral, 1993). The intrinsic connection between principal neurons and the interneurons is mainly inhibitory in nature (Glickfeld et al., 2009), and different interneurons exert their inhibitory effects over pyramidal cell excitability through different mechanisms. There are two types of interneurons of the hippocampus, namely, axo-axonic/somatic and axo-dendritic interneurons (Freund and Buzsaki, 1996). Chanderlier and basket cells belong to the first type while oriens- 9 lacunosum moleculare (O-LM), bistratified and trilaminar neurons are examples from the latter type. The main axons of axo-axonic interneurons are myelinated and run tangentially to their soma. They emit collaterals that arborize and form dense multiple synapses in the prisomatic region of the principal cells. Axons of axo-dendritic neurons, on the other hand, are unmyelinated and arborize upon reaching the target laminae. To discuss the main difference in controlling pyramidal cells, basket cells and O-LM are used as examples. For basket cells, the axons arborise only within the stratum pyramidale and do not penetrate the other layers of the hippocampus. The axons form symmetrical synapses with the soma and initial axon segments of the neighbouring pyramidal cell (Sik et al., 1995). The dendritic tree of basket cells arborises in the stratum oriens, pyramidale and lacunosum-moleculare and receives inputs from neighbouring pyramidal cells (Freund and Buzsaki, 1996). On the other hand, the dendritic tree of O-LM interneurons is confined to the stratum oriens and alveus, while the axons penetrate the hippocampus and arborise at the stratum lacunosum moleculare (Sik et al., 1995). The axons were shown to perfectly overlap with the terminals of entorhinal afferents in the stratum lacunosum moleculare, suggesting that O-LM interneurons could exert an inhibitory regulation of entorhinal synapses terminating on the distal dendrites of the pyramidal cells (Blasco-Ibanez and Freund, 1995; Katona et al., 1999). 10 1.1.3 Extrinsic connection of hippocampus formation Two types of inputs to hippocampal principle cells have been identified: 1) laminated afferents from cortical areas (such as entorhinal area and amygdala) terminating in the specific layers along the dendrites; 2) diffuse afferents from various subcortical regions (such as medial septal nuclei and brainstem monoaminergic nuclei) targeting all layers including perisomatic and dendritic regions (Paxinos, 1985). Given the focus of this study, only projections between hippocampal and septal region are described. 1.1.3.1 Septo-hippocampal connections Projections from neurons in the medial septum are mainly from the medial septal nuclei and the vertical limb of diagonal band of Broca (MS-VLDBB). Using anterograde tracer Phaseolus vulgaris leuco-agglutinin (PHA-L), Nyakas et al. (1987) reported that the medial septal projections to the hippocampus formation are organized in a topographic manner. For instance, the dorsal field CA1 and dorsal dentate gyrus receive dense afferent input from VLDBB and relatively less dense afferents from MS while the ventral hippocampal subfields receive massive input from both MS and VLDBB. In addition, the septo-hippocampal projections are dominantly ipsilateral in nature (Monmaur and Thomson, 1983). Moreover, the septohippocampal efferents are identified as type 1 that are thick and coarse axons with large terminal boutons, and type 2 that are thin axons with varicosities (Nyakas et al., 1987). 11 There are at least three types of neurons identified in the medial septal region by immunohistology and biochemical assays, namely cholinergic (Brashear et al., 1986), GABAergic (Barry et al., 1985) and glutamatergic (Colom et al., 2005) neurons. A significant proportion of medial septal neurons are cholinergic (30%-70%, depending on the rostrocaudal level; (Senut et al., 1989). Indeed, many neurons in the medial septum that were labelled following microinjection of retrograde tracer into the dorsal hippocampus were immunoreactive for the acetylcholine synthesizing enzyme choline acetyltransferase (CHAT), indicating that these were cholinergic in nature (Kiss et al., 1990). The septal cholinergic projection innervates both pyramidal cells and interneurons with both symmetrical and asymmetrical synapse (Frotscher and Leranth, 1985). The septal GABAergic projection, in contrast, mainly target cell bodies or dendrites of hippocampal interneurons (Acsady et al., 1993). Interestingly, the type 2 fibres of Nyakas et al. (1987) correspond to septo-hippocampal GABAergic afferent (Freund and Antal, 1988). Glutamatergic neurons concentrate in the medial septal regions and are distributed among the cholinergic and GABAergic neurons (Colom et al., 2005) 1.1.3.2 Hippocampal-septal projection The hippocampal-septal projection is the most prominent efferent projections from the hippocampus among all the efferent projections from the region (Leranth et al., 1992). The main subcortical target of this efferent is lateral septum (Tamamaki et al., 1984). The hippocampal projections to the medial 12 septal region are relatively sparse and believed to be GABAergic (Toth and Freund, 1992). It is notable that the projection from hippocampus to medial septal region is from field CA2 and CA3, with less contribution from field CA1 and the dentate gyrus (Gaykema et al., 1991). In turn, evidence from tract tracing and double labelling suggests that GABAergic and glutamatergic neurons at the border of the medial septum and lateral septum project to SuM (Borhegyi and Freund, 1998; Leranth et al., 1999; Kiss et al., 2002). 1.1.4 Hippocampal theta rhythm Hippocampal theta rhythm is one of the several types of the extracellular field potentials observed in field CA1 that reflect activity of network of local neurons. It is a large amplitude sinusoidal rhythmic EEG pattern. Two types of theta have been characterised according to the different behavioural and pharmacological properties. Type 1 theta (~8 Hz) occurs during locomotion and other types of voluntary behaviour and during rapid eye movement (REM) sleep, and is unaffected by the anticholinergic drug atropine. Type 2 theta (4~6 Hz) occurs during immobility and during anaesthesia induced by urethane, and is abolished by administration of atropine (Kramis et al., 1975). The extracellular theta waves recorded from CA1 pyramidal cell layer are mirror images of the rhythmic intracellular oscillations of the pyramidal cell somatic membrane potential (Leung and Yim, 1986; Soltesz and Deschenes, 13 1993; Ylinen et al., 1995). Here it is notable that the intracellular membrane potential oscillations recorded from pyramidal cell soma are induced, at least in part, by rhythmic inhibition impinging upon these neurons. Consistent with this, the putative inhibitory interneurons in CA1 increase their discharge around the positive phase of the local extracellular theta that temporally corresponds to the somatic hyperpolarisation of the pyramidal cells (Buzsaki and Eidelberg, 1983; Fox et al., 1986; Ylinen et al., 1995; Klausberger et al., 2003) Conversely, the probability of action potential discharge of majority of pyramidal cells is high around the negative phase of the local theta (Buzsaki and Eidelberg, 1983; Fox et al., 1986; Ylinen et al., 1995; Csicsvari et al., 1998). The negative phase corresponds to the release of pyramidal cells from rhythmic inhibition. Furthermore, the phase-locked pattern of firing indicates that the neurons of the hippocampus exhibit synchronized phasic firing during theta. The MS-VLDBB region is believed to be the pacemaker of the hippocampal theta activity (Buzsaki and Eidelberg, 1983; Bland, 1986; Vinogradova, 1995). The crucial role that MS-VLDBB plays in generation of hippocampal theta is supported by a large body of evidence. These included (1) permanent lesions or reversible blockade of the MS-VLDBB region eliminated hippocampal theta in behaving rat (Buzsaki and Eidelberg, 1983); (2) inactivation of the MS-VLDBB by local anaesthetic procaine or the anticholinergic drug atropine attenuated reticularly-induced hippocampal theta (Jiang and Khanna, 2004; Li et al., 2007). (3) lesion of either septal cholinergic neurons with the immunotoxin 192 IgG-saporin or the parvalbumin (PV) positive septal 14 GABAergic neurons with the glutamate agonist kainic acid attenuated the power of sensory- and exploration-induced theta in anaesthetized and behaving animals, respectively (Lee et al., 1994; Zheng and Khanna, 2001; Yoder and Pang, 2005). However, a recent study by Simon et al (2006) demonstrated that, in urethane anesthetized rats and unanaesthetized restrained rats, cholinergic neurons were slow-firing and did not display theta-related bursting. This may suggest that cholinergic neurons play a more permissive and modulatory role other than acting as pacemakers (Simon et al., 2006). Indeed, Hangya et al (2009) argued that PV positive GABAeric neurons expressing hyperpolarization-induced cation channel (HCN) may be the putative pacemaker neurons leading the hippocampal network activity (Hangya et al., 2009). In this study, the extracellular discharge of putative PV positive GABAergic septohippocampal neurons that express HCN were observed to precede the transition of hippocampal EEG from irregular activity to theta rhythm in anaesthetized rat. 1.2 Anatomy of Supramammillary area (SuM) SuM as a whole is found to connect to many forebrain areas, most of which are the structures of limbic system such as amygdala, entorhinal cortex, dentate gyrus, septum, preoptic area, anterior hypothalamus, locus coeruleus (Saper et al., 1976; Ottersen, 1980; Swanson, 1982; Haglund et al., 1984; Gonzalo-Ruiz et al., 1992b; Hayakawa and Zyo, 1994; Leranth and Kiss, 1996; 15 Borhegyi et al., 1998; Vertes and McKenna, 2000; Kiss et al., 2002). In addition, the projections from SuM and to SuM are chemically very diverse. In the following section, more focus will be given to connections pertinent to hippocampus and septal area. 1.2.1 Projections from SUM SuM as a whole contains calretinin-containing cells (Kiss et al., 2000), dopaminergic cells (Swanson, 1982), and substance P containing cells (Borhegyi and Leranth, 1997; Fig 3). The most medial portion of SuM (mSuM) contains small packed cells and is distinguished by having substance P containing cells. Cells in this region have weak connections to hippocampus and these connections originate mostly from substance P containing cells (Borhegyi and Leranth, 1997; Fig 3). There is a high density of dopaminergic cells in the mSuM (Swanson, 1982; Fig 3). Retrograde tracers and immunochemical techniques showed that these dopamininergic cells project to lateral septum (Swanson, 1982; Shepard et al., 1988) and mammillary bodies (Gonzalo-Ruiz et al., 1992a; Fig 3). 16 Hippocampus MS Calretinin LS LS Substance P Dopamine SuML mSuM SuML MM Figure 3. Schematic diagram of projections from supramammillary nucleus (SuM) to medial septum (MS), lateral septum (LS) and hippocampus. Symbols , , ,represent cells with specific histochemical markers such as calretinin, substance P and dopamine respectively, but with no indication of co-localization of multiple markers. SuM as a whole contains calretinin-containing cells, dopaminergic cells and substance P containing cells. Medial SuM (mSuM) is distinguished by having substance P containing cells which send projections to hippocampus. There is high density of dopaminergic cells in mSuM and they project to LS and mammillary bodies (MM). Cells in lateral portion of SuM (SuML) stain for calretinin and project to hippocampus and MS/LS region. 17 The lateral portion of SuM (SuML) around principal mammillary tract contains large cells. The cells become more loosely packed when moving more laterally (Leranth and Kiss, 1996). Cells in this region stain for calretinin but not substance P. The great majority of the calretinin containing cells were back-labelled with [3H]D-aspartate microinjected into the dentate gyrus or the MS-VLDBB/LS region, suggesting that they project to dentate gyrus and MS-VLDBB/LS and utilize aspartate/glutamate as neurotransmitters (Kiss et al., 2000). In addition, the terminals of supramammillary axons, labelled with the anterograde tracer PHA-L, were observed to innervate the principal neurons in DG (Magloczky et al., 1994). Apart from projecting to septohippocampus, by anterograde and retrograde tracers and transmitter immunocytochemistry, SuM was also found to send projections to LS, especially the dorsal and intermediate LS nuclei, mammillary nuclei (especially lateral mammillary nucleus), thalamus, posterior hypothalamus, amygdale, medial prefrontal and the entorhinal cortices (Swanson, 1982; Shibata, 1987; Vertes, 1988; Hayakawa et al., 1993; Risold and Swanson, 1997) 1.2.2 Projections to SuM Injections of wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP) into the supramammillary nucleus resulted in retrogradely labelled neurons in MS-VLDBB region and other regions including infralimbic cortex, medial and lateral preoptic nucleus, subiculum, laterodorsal 18 tegmental nucleus, compact subnucleus of the central superior nucleus and the dorsal raphe nucleus (Gonzalo-Ruiz et al., 1999). In the MS-VLDBB region, most of the SuM projectiong neurons (80-85%) were found to be cholinergic neurons whereas small numbers of them were identified as GABA-containing neurons. In addition, GABAergic neurons which contain calretinin and/or calbindin and located at the border between LS and MS region were found to project to SuM (see section 1.1.3.2). Because these GABAergic neurons receive input from the entorhinal cortex, the authors proposed a negative feedback mechanism whereby entorhinal cortex influenced SuM neurons through these neurons to modulate the SuM mediation of hippocampal theta activity (Borhegyi and Freund, 1998; Leranth et al., 1999) 1.2.3 SuM and hippocampal theta The SuM is implicated in modulation of the hippocampal synaptic activity and theta activation. Indeed, in context of the latter, it has been proposed that the SuM mediates the generation of hippocampal theta rhythm, at least partly. Thus, Kirk and McNaughton (1991) found that mSuM cells discharge rhythmically which is in phase with concurrent hippocampal theta activity. Further, microinjections of the local anaesthetic procaine into posterior hypothalamus blocked the reticularly-elicited rhythmic discharge of MS neurons, but inactivation of the MS did not block the rhythmic bursting of hypothalamic neurons (Bland et al., 1994; Kirk et al., 1996). Moreover, in urethane anaesthetised rats, procaine injections into mSuM decreased both the frequency and amplitude of reticularly-elicited theta (Kirk and McNaughton, 19 1993; Khanna et al., 2004). On the other hand, procaine microinjected more rostrally into the medial forebrain bundle (that carries afferents from posterior hypothalamus to medial septal region) or the medial septal region decreased the amplitude only. As regards SuML, Jiang and Khanna (2004) found that GABA, the ubiquitous inhibitory ligand, microinjected into SuML attenuated the power of reticularly-elicited theta without having effects on frequency. Similarly, microinjection of the nicotinic receptor antagonist, mecamylamine, into the SuML reduced the power, but not frequency, of both reticular stimulation- and sensory stimulation- induced theta (Ariffin et al., 2010). However, this effect was not seen with mSuM microinjection of the antagonist (Ariffin et al., 2010). The SuM is also implicated in modulation of the hippocampal synaptic activity. Glutamate stimulation of SuML produced a substantial increase in the perforant path stimulation-induced population spike in the dentate gyrus (Carre and Harley, 1991). Inactivation of the mSuM with GABA abolished reticularly-elicited suppression of CA1 population spike, thus indicating that SuM mediated reticularly-elicited suppression of CA1 pyramidal cell synaptic excitability to CA3 stimulation (Jiang and Khanna, 2004). Moreover, microinjection of mecamylamine into the SuML, but not mSuM, attenuated both reticularly-elicited and noxious stimulation induced population spikes in CA1 suggesting that nicotinic cholinoceptive neurons in SuML is a possible element of the neural drive to the hippocampus on rostral periolivary region (RPO) activation and noxious stimulation (Ariffin et al., 2010). Taken together, the above data showed that SuM as a whole modulates hippocampal theta and 20 population spike suppression. In addition, there may be functional and neurochemical heterogeneity along the mediolateral axis of SuM neurons in controlling different aspects of hippocampal theta such as frequency and amplitude. 1.2.4 Effects of manipulations of SuM on behaviour and hippocampal theta The effect of manipulations of SuM has been examined on a variety of animal behaviors. Especially the focus of such investigations has been to determine whether SuM manipulations affect hippocampal-mediated behaviors (see sections 1.2.4.1 to 1.2.4.3) since SuM and the septohippocampal regions are interlinked. 1.2.4.1 Spatial memory Large lesions produced by microinjection of α-amino-3-hydroxyl-5-methyl-4isoxazole-propionate (AMPA) into mSuM (made prior to training) resulted in only modest decrease in theta frequency (by about 0.4 Hz) without affecting spatial reference memory (as assessed by averaging the distance swum to reach the target on each of the 4 days of training) or swimming speed in the Morris water maze (Pan and McNaughton, 2002). Consistently, infusion of a benzodiazepine, that acts on the GABA-benzodiazepine-chloride-ionophore complex to potentiate the effects of endogenous GABA (McNaughton and Sedgwick, 1978), into SuM produced no obvious impairment of learning and 21 theta activity (Pan and McNaughton, 1997). Thus, SuM, unlike hippocampus, seems not to play important roles in either spatial working memory or reference memory. The conclusion is further supported by Sziklas and Petride who reported that large lesions including destruction of medial mammillary bodies and SuM, do not affect radial arm maze learning (Sziklas and Petrides, 1993). 1.2.4.2 Exploratory and defensive behaviour Pan and McNaughton reported that neurotoxic lesions of mSuM increased exploratory ambulation in open field test but did not reduce the average frequency of theta, whereas lesions in SuML had no effect on exploratory behaviour and theta (Pan and McNaughton, 2002). In fear conditioning test, neurotoxic lesions of mSuM, but not unilateral SuML, decreased contextual fear conditioning (i.e. decreased freezing and so increase movement) without affecting conditioning to the discrete stimulus (a 10-KHz, 75-dB tone which was present for 20s; Pan and McNaughton, 2002). Similarly, hippocampal lesions impaired conditioning of fear to the context in which conditioning is carried out but did not impair conditioning to the discrete fear stimulus itself. (Phillips and LeDoux, 1992, 1994) 1.2.4.3 Passive avoidance and Fixed Interval schedule According to Gray and McNaughton, passive avoidance is a test not only of defense but also of behavioural inhibition - a postulated fundamental function 22 of the hippocampus (Gray, 1982; Gray and McNaughton, 2000). Generally, the rat is allowed to move from the one compartment (e.g. white compartment) to the other compartment (e.g. black). In the conditioning session the entry into the black compartment is punished with a mild inescapable electrical shock (Barrionuevo et al., 2000). The typical behavior with passive avoidance on the test day is that the rat tends to delay its time in entering the black compartment and spend more time in the white compartment. Neurotoxic lesions of either mSuM or unilateral SuML produced an increase in responding (i.e. reduced latency and increased time in black compartment) and thus impaired inhibition (Pan and McNaughton, 2002). Another test of behavioral inhibition is fixed interval schedule test (FI). In this test, the rat is rewarded for the first desired response (i.e. bar press), after which a set interval of time must pass before any responses will be rewarded. For example, if a rat is on a fixed interval of 60 s, it will be rewarded for its first bar press, but not again until after 60 s has passed - no matter how many time it presses the bar. The typical pattern of responding with a fixed interval schedule is to see most of the responding around the time at which the reward is due. Then once the rat has received its reward for an interval, it usually stops responding for most of the remainder of the interval (Ellen E Pastorino, 2008; Ellen and Susann, 2008). With FI, injections of a benzodiazepaine into mSUM produced increased responding and a concomitant decrease in theta frequency. The effects were even comparable to the effects of i.p. injection of the benzodiazepine (Woodnorth and McNaughton, 2002b, a). Moreover, neurotoxic lesions of either mSuM or unilateral SuML increased responding 23 and so impaired inhibition. The change in behaviour produced by lesions of mSuM was accompanied by a decrease in theta frequency of about 0.4 Hz whereas lesions of SuML did not reduce theta frequency (Pan and McNaughton, 2002). The behavioural effects were similar with hippocampal lesions but greater when compared to either mSuM or SuML lesion (Manning and McDonough, 1974). Overall, the behavioural data discussed above indicated that SuM lesions affected a number of hippocampal-dependent behaviours, except spatial learning in Morris water maze. The effect on some behaviours, such as tests of behavioural inhibition were accompanied by a decrease in theta frequency. On the other hand, SuM lesion did not change theta frequency although it increased animal ambulation in open field test. Similarly, large electrolytic lesion of the SuM did not affect theta in freely moving animals (Thinschmidt et al., 1995). Likely, therefore, exploration-induced theta may be mediated via other pathways to the hippocampus. 1.3 Pain 1.3.1 General description of pain Pain has been defined by the International Association for the Study of Pain (IASP) as “an unpleasant sensory and emotional experience associated with actual or potential tissue damage or described in terms of such damage 24 (Merskey, 1979).” Thus, pain is a multifaceted and highly subjective experience that is unique to each individual. Indeed, McCaffrey and Beebe described pain as “whatever the experiencing person says it is, existing whenever the experiencing person says it does (McCaffrey and Beebe, 1989; McCaffrey M., 1989)”. Due to its subjective nature, the word pain is not used with reference to animals as they are unable to communicate verbally the pain experience. Nonetheless, animals display overt and covert changes similar to that in human upon exposure to noxious stimuli which induce (1) behavioural responses, (e.g. flexion reflex) and voluntary reactions (e.g. escape and avoidance), and (2) physiological responses, (e.g. change in heart rate and release of stress hormones). It should be noted that nociception and pain are somewhat different. Nociception is a process in which nociceptors are activated by a noxious stimulus. Nociceptors are free nerve endings found in most organs and tissues in the body and activated by either a noxious mechanical (touch or pressure), thermal (hot or cold), or chemical (endogenous or exogenous) stimulus (Millan, 1999). Pain refers to a feeling or a perception of irritating, miserable or unbearable sensations. Pain can be caused without nociception. For instance, in the case of allodynia, a pathological condition of a painful response to typically non-painful stimulus, patients can experience pain by a light touch without nociceptors being activated (Merskey and Bogduk, 1994). In the clinical setting, there are two broad categories of pain, acute and chronic. Under these broad categories fall the subtypes of pain, which include 25 inflammatory, neuropathic, cancer etc. Acute pain is characterized by its short duration, identifiable cause and self-limitedness. It functions as an endogenous protective mechanism that signals the brain of the occurrence of real or potential tissue injury, thus prompting a protective response (Henry, 1962). The commonly used models to study acute pain include hot-plate, tailflick and tail-pinch tests, in which the subject is allowed to remove the affected region from noxious stimuli by making a response (D'Amour and Smith, 1941; Woolfe and MacDonald, 1944). The foregoing are models of brief pain. One model used to study acute persistent pain is formalin model, in which the subject must cope with pain/nociception due to hind paw injection of the dilute formaldehyde (Albe-Fessard, 1990). Chronic pain is, on the other hand, unrelenting, has no identifiable cause, spreads beyond the original site of injury, and serves no biological function. Three interactive components of pain have been identified by Melzack and Casey (1968): sensory-discriminative, affective-motivational and cognitiveevaluative. The sensory discriminative component describes the intensity, location, quality and duration of a noxious stimulus and these features are encoded by the ascending pain pathways that transmits nociceptive information from peripheral tissues to the cerebral cortex for interpretation as pain. The affective component of pain refers to the unpleasantness or emotional distress that invariably attends the sensation of physical pain. It encompasses a range of negative emotions such as anger, fear, sadness and suffering and triggers the subjects to seek pain relief. Further, the affective component is closely linked to the autonomic nervous system and the 26 associated cardiovascular, respiratory and gastrointestinal responses. The cognitive-evaluative components reflects the evaluation of the meaning and possible consequences of pain based on prior experiences. This component interacts intimately with the affective-motivational aspect of pain. The neural systems of the sensory-discriminative of pain have been extensively studied. However, relatively less attention has been paid to the study of affective-motivational and cognitive characteristics of pain. It’s important to note that sensory-discriminative component of pain may utilize different neural pathways, at least in part, from affective-motivational and cognitive component. For example, in the case of demented patients with impairment of affective and cognitive functions, such as Alzheimer’s disease, the emotional response to painful stimuli was significantly reduced, though the sensory-discriminative component of pain was maintained intact (Benedetti et al., 1999). Furthermore, Rainville et al demonstrated using positron emission tomography (PET; an imaging technique measuring cerebral blood flow) that hypnotic suggestion that altered the unpleasantness but not the perceived intensity of noxious heat stimulus resulted in significant change in painevoked activity in cingulated cortex but not primary somatosensory cortex (Rainville et al., 1997). 1.3.2 Pain transmission pathway This section briefly describes the transmission of nociceptive signal in the central nervous system (CNS). The signal, which is action potential, is 27 generated via the nociceptors or free nerve endings as a result of application of noxious stimuli in their receptive fields. The signal is then conveyed along the associated axons, namely myelinated Aδ axons, which conduct at 5-30 m/s, or unmyelinated C fiber axons, which conduct at less than 2 m/s. In general, the Aδ- and C-nociceptors supplying the skin are functionally classified as: Aδmechanosentive nociceptor (respond to intense mechanical stimuli); Aδmechanothermal nociceptor (respond to both intense mechanical and thermal stimuli) and C-polymodal nociceptors (respond to thermal, mechanical and chemical stimuli; Millan, 1999). However, more recent findings suggest a greater complexity and phenotypical diversity, especially in relation to behavior. Thus, for example, the C-fibers are subdivided based on presence or absence of peptides and functionally destruction of the peptidergic C-fibers with the neurotoxin capsaicin in mice attenuated animal behavioral response to noxious heat, but not to noxious mechanical stimuli (Scherrer et al., 2009). Conversely, genetic deletion of Nav1.8 expressing neurons, including the destruction of non-peptidergic C-fibers but with sparing of population of peptidergic C-fibers, strongly attenuated the animal response to noxious mechanical stimulation, but not to noxious heat and non-noxious mechanical stimulation (Abrahamsen et al., 2008). Nociceptive afferents then enter the dorsal horn of the spinal cord where they give off branches that contact second-order neurons. Aδ-fibers terminate in laminas I and V while C-fibers terminate in lamina I and II (Besson and Chaouch, 1987; Millan, 1999). More recently, Braz et al (2005) identified the termination pattern of non-peptidergic C-fibers (which stain for isolectin IB4) 28 by way of genetic expression of wheat germ agglutinin (WGA) in these nociceptors. These primary afferents mostly terminated in inner lamina II above the layer of spinal interneurons positive for γ isoform of protein kinase C. In contrast, the peptidergic C-fibers terminate in lamina I, directly contacting projection neurons that transmit nociceptive messages to brainstem and/or thalamus as well as the interneurons in outer lamina II (Todd et al., 2000; Scherrer et al., 2009). Two distinct types of spinal second-order neurons are identified (Besson and Chaouch, 1987; Millan, 1999): nociceptive specific (NS) neurons or high threshold neurons, which respond only by nociceptive inputs, distributing primarily in lamina I, and (2) wide dynamic range (WDR) neurons, which respond to a broad range of stimulus, including both nociceptive and innocuous low-threshold mechanical activation, distributing in deeper laminae of dorsal horn. There is a visceral input to these WDR neurons which results in convergence of somatic and visceral inputs in the phenomenon of referred pain (Scherrer et al., 2009). Projection neurons within laminar I constitute the major output from the dorsal horn to the brain (Basbaum and Jessell, 2000). Neuroanatomical studies including anterograde tracing with PHA-L and retrograde tracer studies provided evidence that lamina I neurons project heavily to thalamus especially the ventroposterolateral nucleus (Gauriau and Bernard, 2004). Retrograde labeling indicated that greater than 80% of the lamina I spinothalamic neurons projected to the parabrachial area (Hylden et al., 1989). The nociceptive 29 responses of lamina I neurons projecting to the parabrachial area have been characterized in halothane anaesthetized rat (Bester et al., 2000). The spinoparabrachial neurons were mostly nociceptive specific (~75%) to noxious stimuli, including mechanical and thermal stimuli. The rest were wide-dynamic range type that responded to both noxious and non-noxious stimuli. The receptive field of these neurons was small, covering one to two toes and their firing rate increased with the intensity of noxious stimuli indicating that these played a role in intensity-encoding (Bester et al., 2000). Moreover, NS and WDR neurons are identified in VPL/VPM and are topographically organised (Albe-Fessard et al., 1985). Both anterograde and retrograde tracing techniques have identified a number of other ascending nociceptive pathways from the spinal neurons which are believed to involve in affective-motivational component of pain. These include the spinoreticular (Fields et al., 1977; Peschanski and Besson, 1984; Bester et al., 1995) and spinohypothalamic (Burstein et al., 1991; Giesler et al., 1994) pathways, the spino-parabrachio-amygdaloid (Schaible and Grubb, 1993; Millan, 1999) and other ascending pathways to the brainstem regions such as to ventrolateral medulla, the periaqueductal gray, and also further rostrally to the septum (Fields et al., 1977; Menetrey et al., 1982; Peschanski and Besson, 1984; Bester et al., 1995; Millan, 1999). Consistent with this, transneuronal transfer of WGA expressed genetically in non-peptidergic Cfibers indicated that these afferents influenced a variety of brain regions via spinal neurons (Braz et al., 2005). These brain regions include the amygdala, 30 ventromedial hypothalamus, bed nucleus of stria terminalis, globus palidus and periaqueductal gray. The nociceptive information eventually reaches a widespread cortical mantle. Consistent to this, both NS and WDR nociceptive neurons are identified in deep layers of the primary somatosensory cortex (Lamour et al., 1982; Kenshalo and Isensee, 1983). These cortical nociceptive neurons have small receptive fields that are arranged in a somatotopic pattern and code for intensity and localization of noxious stimuli (Lamour et al., 1983). Indeed, lesions in the sensory cortex and/or lateral thalamus impaired sensory attributes, leading to poor intensity encoding and spatial localization of noxious stimulus (Marshall, 1951; Treede et al., 1999). More recently, imaging techniques indicated a large distributed brain networks that was activated during painful stimulation. For example, the cortical and subcortical brain regions found to be commonly activated by nociceptive stimulation included: anterior cingulated cortex, insula, frontal cortices, and S1, S2 (Peyron et al., 2004). Nociceptive neurons have also been electrophysiologically recorded in limbic structures, such as anterior cingulated cortex (Vogt et al., 1992), hippocampus (Khanna, 1997) and amygdala (Bernard et al., 1992). Generally, nociceptive neurons in the limbic system have little or no somatotopic organization with large receptive fields. These limbic regions may receive spinal nociceptive inputs via the ascending diffuse system polysynaptically, though some other limbic structures and may also be directly accessed by spinal nociceptive 31 neurons (Burstein and Giesler, 1989; Burstein and Potrebic, 1993). For example, the hippocampus receives nociceptive inputs, at least in part, from the medial septum such that lesion of the region attenuates the noxious stimulus-induced theta and pyramidal cell suppression in hippocampus. 1.3.3 Formalin model of persistent pain Most of the traditional tests of nociception, such as tail-flick and hot-plate tests which are based on a phasic stimulus of high intensity, produce nociceptive responses which are short-lasting. It is therefore not suitable for the study of continuous pain. On the other hand, formalin test generates nociceptive responses which are long-lasting, making it a good model to study inflammatory persistent pain. Moreover, because formalin-induced pain is produced by injured tissue, the test provides a more valid model for clinical pain than the tests with phasic mechanical or thermal stimuli (Dubuisson and Dennis, 1977; Abbott et al., 1981; Alreja et al., 1984). Formalin is an aqueous solution of 37% (weight/weight) formaldehyde. In formalin test, diluted formalin (0.5-5%, 20-100µl) is injected subcutaneously into one of the paws (either the dorsal surface or the plantar; Dubuisson and Dennis, 1977). The hind paw is usually chosen for injection because the formalin-elicited response of the hind paw, such as licking, rarely occur during normal grooming behaviour (Tjolsen et al., 1992). Injection of formalin is generally made by lightly restraining the animal. Immediately after the injection, several reproducible and easy identifiable behavioural responses are 32 observed, which include lifting, licking, and flinching the affected paw (Dubuisson and Dennis, 1977). In addition, autonomic changes are observed that include an increase in blood pressure and heart rate following hind paw injection of formalin (Taylor et al., 1995). Both the nociceptive behaviours and the cardiovascular changes observed following injection of formalin are expressed in a biphasic fashion with an acute phase (phase 1, 0 to 5 min) that is followed by a tonic phase (phase 2, generally from 11 min to 60 min after injection) with a quiescent period (interphase) of about 5 min in between the two phases. Injection of formalin evokes a concentration-dependent increase in nociceptive behaviour with formalin concentration ranging from 0.625-5% (Khanna et al., 2004; Taylor et al., 1995). The autonomic changes such as blood pressure and heart rate (Taylor et al., 1995) as well as release of spinal nitric oxide metabolites (nitrite/nitrate) and glutamate (Okuda et al., 2001) are also found correlated highly with behaviours and are dependent on formalin concentration in those studies. However, at a higher concentration (10%) of formalin, nociceptive behavioural responses were decreased and failed to produce a clear-cut release of spinal nitrite/nitrate and glutamate (Okuda et al., 2001). The nociceptive responses in the formalin test can be suppressed by nonsteroidal anti-inflammatory drugs (Hunskaar et al., 1986; Drower et al., 1987), systemically administrated opiate analgesic such as morphine (Abbott et al., 1995), intraplantar injection of lidocaine derivative (Taylor et al., 1995) and 33 electrical stimulation of various regions of the brain including the brain stem (Dubuisson and Dennis, 1977). Of the two behaviours that are formalin concentration-dependent, namely licking and flinching, the latter is regarded as a more robust parameter as compared to licking which is more readily influenced by other factors such as taste aversion of the formalin (Wheeler-Aceto and Cowan, 1991). Moreover, implantation of chronic intrathecal cannula has been reported to completely suppress licking but not flinching (Sawynok and Reid, 2001, 2003). Administration of pentobarbital, a non-analgesic sedative, is found to affect licking while sparing flinching (Abbott and Guy, 1995). In addition, Sawynok showed that flinching and licking could be differentially modulated, indicating that they may involve distinct mechanisms (Sawynok and Liu, 2003; Sawynok and Reid, 2003). For example, systemic administration of amitriptyline, an antidepressant, simultaneously increased flinching while suppressing licking (Sawynok and Reid, 2001). At the neural level, hind paw injection of formalin also elicited a biphasic increase in activity of peripheral nociceptors (McCall et al., 1996; Puig and Sorkin, 1996). More recent evidence suggested that formalin excited nociceptors by directly activating TRPA1, a member of the Transient Receptor Potential family of ion channels. Formalin induced robust calcium influx in cells expressing cloned or native TRPA1 channels, and these responses were suppressed by TRPA1 antagonist (McNamara et al., 2007). Moreover, behavioural studies showed that pharmacologic blockade or genetic ablation 34 of TRPA1 resulted in attenuated nociceptive responses in formalin pain model (Macpherson et al., 2007; McNamara et al., 2007). Consistent to a role of peripheral nociceptors in formalin pain, both the formalin-induced behaviours and cardiovascular changes are blocked by peripheral administration of local anaesthetic that presumably blocked afferent input to the central nervous system (Coderre et al., 1990; Taylor et al., 1995). More recent studies also provided evidence for role of peripheral nociceptors in formalin nociception. Mice with their sensory neurons expressing the sodium channel Nav1.8 destroyed showed suppressed pain responses in the formalin test (Abrahamsen et al., 2008). Similarly, a reduction in formalin induced pain behaviours was also observed in sodium channel Nav1.7 knockout mice (Nassar et al., 2004). As in the periphery, hind paw injection of formalin evoked a biphasic increase in discharge of dorsal horn neurons (Dickenson and Sullivan, 1987; Chapman and Dickenson, 1995; Pitcher and Henry, 2002). Evidence suggested that spinal excitation involved excitatory synaptic transmission that mediated, at least in part, via the alpha-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-D-aspartate (NMDA) glutamate receptors. In this regard, formalin injection elicited an increase in the spinal level of the excitatory amino acid neurotransmitter, glutamate (Kantner et al., 1985; Malmberg and Yaksh, 1995) while, conversely, the formalin-induced discharge of spinal neurons was antagonized by intrathecal administration of antagonist at the AMPA and NMDA glutamate receptors (Chapman and 35 Dickenson, 1995). Interestingly, formalin injection also evoked a persistent increase in the size of the spinal field potential (i.e. potentiation) evoked on Cfibre stimulation in anaesthetized rat (Ikeda et al., 2008). This effect was reproduced by low frequency stimulation of nociceptors that mimicked the persistent activation of primary afferents following injection of formalin. The potentiation was blocked by systemic administration of MK-801, an NMDA receptor antagonist. The authors proposed that the formalin-induced potentiation in the spinal cord was akin to long-term potentiation of synaptic transmission and might act as a synaptic pain amplifier mechanism (Ikeda et al., 2008). Indeed, and consistent with a key role of spinal glutamatergic transmission, intrathecal administration of antagonist at NMDA and AMPA receptors prior to formalin injection eliminated or strongly reduced the phase 2 response (Coderre and Melzack, 1992; Malmberg and Yaksh, 1992). At supraspinal level, injection of formalin activated a number of brain structures. Functional imaging studies using PET techniques suggested that formalin injected in left hind paw of rats activated hindlimb cortex, retrosplenial portion of the cingulated cortex and midbrain periaqueductal gray (PAG; Casey, 1999). Further, an increase in immediate early gene (IEG) (eg. c-fos) expression was also observed in prefrontal cortex, locus coeruleus, PAG and ventromedial nuclei following formalin injection into the hind paw of rats (Aloisi, 1997; Aloisi et al., 2000). In general, formalin test involves integration of many brain regions which subserve sensory and emotional processes. 36 1.4 c-Fos 1.4.1 General description of c-Fos The term immediate-early gene (IEG) was first used for viral genes (Kovacs, 2008). These genes were also activated by the eukaryotic cells upon viral invasion and were responsible for transcriptional reprogramming of the host to promote virus replication. Examples of IEGs are c-fos, c-jun and c-myc (Kovacs, 2008). Evidence suggested that c-fos was involved in regulating neuronal cell excitability, at least in part. For example, Zhang et al (2002) generated a mutant mouse in which c-fos expression was largely eliminated in the hippocampus. They found that these mutant mice displayed more severe kainic acid-induced seizures and increased neuronal excitability and neuronal cell death. In vivo and in vitro experiment showed that c-fos regulated the expression of the kainic acid receptor GluR6 and brain-derived neurotophic factor (BDNF), indicating that c-fos played a role in regulating cellular mechanisms that mediate neuronal excitability and survival. Moreover, the expression of c-fos and its protein product, c-Fos, is used as an index of neuronal excitation. As such, c-fos is the most widely used functional anatomical marker of activated neurons. There are several features that render this IEG as an excellent mapping tool: 1) it is expressed at low levels under basal conditions; 2) the response is transient; 3) it is induced upon wide range of extracellular stimuli (Kovacs, 1998, 2008). Indeed, multiple cis-acting 37 elements (e.g. Ca2+/CRE, SRE, SIS and AP-1RE) are present in the 5’flanking regulatory region of the c-fos gene. Thus, c-fos is induced in response to neurotransmitters and polypeptide hormones via cAMP response element (CRE) in a process that involves activation of protein kinase A (PKA) or calmodulin-independent protein kinases (CaMKs) which in turn phosphorylate cAMP-responsive element-binding protein (CREB); the phosphorylated CREB then binds to CRE and subsequently activates transcription (Gonzalez and Montminy, 1989; Bito et al., 1996). Whereas, c-fos expression is induced via serum response element (SRE) by growth factors, cytokines, calcium and other stimuli through the action of mitogen activated protein kinase (MAPK; Treisman, 1992). The sis-inducible element (SIE) has been proposed to contribute to c-fos induction by plateletderived growth factor (PDGF) and gamma-interferon (Wagner et al., 1990). The activator protein-1 response element (AP-1RE) is believed to be the site where AP-1 might mediate negative auto-feedback for c-Fos transcription (Sassone-Corsi et al., 1988). At the protein level, c-Fos, product of c-fos, is a member of a family that includes three other protein members, namely, FosB, Fra-1 and Fra-2 (Greenberg and Ziff, 1984; Cohen and Curran, 1988; Zerial et al., 1989; Nishina et al., 1990). These proteins possess a leucin zipper motif that forms heterodimers with Jun family proteins, and the resulting activator protein-1 (AP-1) complexes regulates subsequent gene expression by binding to the AP1 sequence found in many cellular genes (Chiu et al., 1988; Halazonetis et al., 38 1988; Kouzarides and Ziff, 1988). The heterodimer AP-1 complexes interact with a consensus sequence TGACTCA in the regulatory regions of target genes and, depending on which member of the Jun family Fos dimerize with, stimulate or repress transcription (Kovacs, 1998). For example, c-Fos/c-Jun complexes were shown to exert stimulatory effects on target gene expression, whereas c-Fos/JunB complexes were mostly inhibitory (Schutte et al., 1989). 1.4.2 Fos as an index of spinal nociceptive information processing The expression of c-Fos in spinal cord following noxious stimulation was initially reported by Hunt et al. (1987). As such it is a much used as a way to visualize the spinal neurons involved in integration of noxious input. Indeed, c-Fos is induced robustly following nociceptive stimuli in nuclei of spinal neurons that are localized in the laminae I-II, the deep dorsal horn and the ventral gray which is consistent with the distribution of nociceptive neurons recorded using electrophysiological techniques (Millan, 1999; Coggeshall, 2005). Whereas, non-noxious stimuli used as controls are generally ineffective in expressing c-Fos in the neuronal population within these spinal laminae. A variety of noxious stimuli induce c-Fos expression in the spinal dorsal horn. The most used noxious somatic stimuli include thermal (focused hot spots, dipping the lower extremity into hot water, the hot plate test, etc), mechanical (pinch, incision, pinprick), chemical (intraplantar injection of formalin and 39 carageenan, cutaneous injection of mustard or turpentine oils, etc.; Coggeshall, 2005). Interestingly, the spatial pattern of c-Fos expression in spinal cord may vary with time after peripheral application of various noxious stimuli including thermal (Herdegen et al., 1991; Schadrack et al., 1998), mechanical (Leah et al., 1992) and chemicals such as formalin (Presley et al., 1990a; Li et al., 2004), carrageenan (Draisci and Iadarola, 1989) and mustard oil (Hunt et al., 1987). For example, Presley et al (1990) reported that following injection of formalin (5%, 150µl), robust c-Fos expression was detected at 1 hr in laminae I and II after injection. Numbers of c-Fos neuronal profiles rose to a maximum in superficial laminae within 2-4 hr, and later declined and returned to basal levels 8-24 hr following stimulation. At 2 hr after injection, labeling was also observed in laminae III-IV, the deep dorsal horn (laminae V-VI) and in the ventral gray (laminae VII, VIII and X). A parsimonious interpretation of the pattern of temporal change is that the first wave of c-Fos induction in the superficial laminae partly represents direct activation of spinal neurons receiving nociceptive input while the second wave of c-Fos induction reflects transsynaptic changes. Indeed, in context of the former, injection of formalin induced c-Fos expression in ~80% of supraspinally projecting and neurokinin 1 (NK1) receptor-positive lamina 1 neurons, many of which received synapses from substance P containing, presumably nociceptive primary afferents (Naim et al., 1997; Todd et al., 2002). Conversely, deletion of NK1 receptor significantly reduced formalin-induced nociceptive behaviour that was accompanied by a decrease in number of c-Fos positive neurons in laminae III of the spinal cord (DeFelipe and Gonzalez-Albo, 1998) 40 Khanna et al (2004) reported that formalin-induced increase in spinal c-Fos was formalin concentration-dependent and paralleled the increase in nociceptive behavior. The concentration-dependent increase in c-Fos expression was observed in all laminae, especially in laminae I-II, V-VI and VII-X. Conversely, c-Fos expression decreased in concert with nociceptive behaviors in the formalin test with the following manipulations: 1) local plantar injection of alpha-2 adrenergic receptor agonist, medetomidine (Pertovaara et al., 1993); 2) systemic or local plantar injection of opiates, including morphine (Presley et al., 1990b; Barr et al., 2003); 3) fear conditioning (Abbadie et al., 1994); 4) pre-treatment of capsaicin that destroyed C-fibers, presumably including C-fiber nociceptors (Borsani et al., 2007). Put together, the above-mentioned studies suggests that spinal c-Fos expression following formalin injection is well correlated with nociceptive behaviours and can be used as a marker of nociception. 1.5 AMPA RS-α-Amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) is a specific agonist for the AMPA receptor where it mimics the effects of the neurotransmitter glutamate (Purves et al., 2008). The neurotoxic effect of AMPA in the supramammillary region was first reported by Pan and MacNaughton (2002). They found that AMPA was able to produce specific lesions within a small region such as mSuM. More importantly, there was no lesion in any rat extended into the mammillary bodies (Pan and MacNaughton, 2002). Therefore, AMPA was chosen in the current study to damage SuM 41 while sparing the adjacent mammillary bodies. As such, it rules out the possibility that the behavioural effects of SuM lesions made by AMPA resulted from lesions of the mammillary bodies. 1.6 Rationale and objectives of the present study There are physiological (Khanna and Sinclair, 1989; Khanna, 1997; Khanna et al., 2004), pharmacological (Rodgers and Brown, 1976) and behavioral (McKenna and Melzack, 2001) data which suggest that the hippocampal formation is recruited by noxious stimuli and contributes to the negative affective-motional state of the animal during pain. Indeed, inactivation of the hippocampus decreased formalin-pain induced behaviors in rat (McKenna and Melzack, 2001). The posterior hypothalamus (PH)-supramammillary (SuM) region is reciprocally connected to the hippocampal formation. Furthermore, the region modulates hippocampal plasticity and neuronal synchronization, in form of hippocampal theta wave activity, through the medial septum (Ariffin et al.; Jiang and Khanna, 2004; Pan and McNaughton, 2004; Jiang and Khanna, 2006). Interestingly, the behavioral effects of SuM lesion overlap with the effects reported with lesion of the septum and the hippocampus (Pan and McNaughton, 2002, 2004). This strengthens the postulate that the SuM 42 interaction with the septohippocampal region is important for affectivemotivational behaviors. Most of investigations with SuM lesion have focused to date on the role of region in anxiety, and learning and memory (e.g. Pan and McNaughton, 2002; See review by Pan and McNaughton, 2004). Whereas, a recent report from the laboratory indicated that SuM also acted as relay of nociceptive information to the hippocampus (Ariffin et al., 2010). To investigate whether SuM affects nociception, the present study evaluated the effects of PH-SuM lesion on formalin-induced behaviours in rat. Given the binding of the PH-SuM region with the septohippocampal region, it was anticipated that PH-SuM lesion would attenuate formalin-induced nociception. 43 MATERIALS AND METHODS 44 2.1 Animals 42 adult male Sprague-Dawley rats weighing 270-300g at the time of surgery were used. Following the surgery, the animals were housed individually in polycarbonate cages and maintained on a 12-h light/dark cycle (on at 0700h, off at 1900 h). Water and food were available ad libitum. All efforts were made to minimize animal suffering. The current study has been approved by Institutional Animal Care and Use Committee (IACUC). 2.2 Surgery procedure Rats were first injected atropine sulphate (2mg/kg, Sigma, St louis, USA) i.p. to reduce bronchial and salivary secretions for improving respiration during surgery. After 10min, the rats were anaesthetized by i.p. injection of sodium pentobarbitone (60mg/kg; Abbot Laboratories, Australia). The head was then secured in a stereotaxic apparatus (Stereotaxic frame assembly, Schueler, USA). An incision was made into the scalp to expose the skull. The position of the incisor bar was then adjusted so that bregma and lambda were in the same horizontal plane. A hole was drilled into the skull for microinjection of 0.3 ml of 0.015M RS-α-Amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA; Tocris Cookson Inc., USA) made up in 0.01M phosphate buffer (pH 7.4) at the rate of 0.1 ml/min via a 33G stainless steel needle. Vehicle animals received similar injections of the phosphate buffer. The co-ordinates were AP 0.38, ML 0.07 and DV 0.80 mm (at an angle of 7° from the vertical) for medial SuM and AP 0.38, ML 0.07 and DV 0.82 for lateral SuM. The needle 45 was left in situ for 10 minutes before retraction to prevent back flow of the liquid and to allow diffusion into the surrounding tissue area. Nylon thread was used to stitch up the incision wound. They were treated with diazepam (2.5 mg/kg i.p; Mayne Pharma Pty Ltd; Australia) following surgery so as to control seizures that may otherwise be observed due to microinjection of AMPA. The rats were allowed to recover for about 10 days in animal holding unit (AHU) in National University of Singapore before behavioural tests. 2.3 Behavioural tests 2.3.1 Scheduling of the behavioural tests 15 animals were tested only in open field after which they were sacrificed. Another 27 animals that were also tested in open field were subsequently subject to the formalin test of persistent inflammatory pain. The formalin test was performed on the 4th day following the open field experiment. In the intervening period the animals were habituated to the experimental apparatus for at least 45 min every day. The animals were sacrificed after formalin test. 2.3.2 Open field test A 43.2 x 43.2 x 30.5cm (L X WQ X H) Open Field (OF) Activity chamber with transparent walls (OFA, Med Associates Inc., USA) was used for the open field test. The apparatus was elevated 70 cm above the floor. 46 Environmental temperature was kept at 22±1˚C. The animal was transferred from the AHU using a 30cm x 20cm clean plastic carrier box to the laboratory located within 10 min walking time. The animal was placed at the right lower corner of the chamber immediately after the observer reached laboratory and allowed to explore for 90 min in the absence of the observer (Fig 4). The OF apparatus was equipped with infrared (I/R) photo beams aimed at I/R detectors that were used to detect animal movement along the horizontal (# I/R beams) and vertical (# I/R beams) directions. Upon entry of the animal, signals from the I/R beams were fed into a computer and the number of IR beams broken by the animal provided measures of ambulation distance, velocity, vertical counts, stereotypic count and etc. The exploratory behaviours were scored as the ambulatory distance which was obtained from the computer system. An ambulatory episode was signalled when the rat moved beyond 4 IR beams (i.e. Box Size) in 500 ms (i.e. Resting Delay). After 90 min, the animal was moved to the carrier box brought into the behavioural room and habituated in that for another 30 min. Food and water were available to the animal in that period. Subsequently, the animals were sent back to AHU for formalin test 4 days later. 2.3.3 Formalin test 47 Open Field Apparatus I/R dischargers Computer Signal processing system I/R detectors Figure 4: Open field apparutus. The Open Field Activity (OFA) chamber (43.2 X 43.2 X 30.5 cm) was used. The photo beams from the I/R dischargers were received by the I/R detectors along both the horizontal and vertical directions. Upon entry of the animals, signals from the I/R beams were fed into a computer and the number of I/R beams broken by an animal provided measure of ambulation distance. 48 One half (21.7 x 43.2 x 30.5 cm) of the ‘two-compartment insert’ (Med Associates Inc., USA) was used for formalin test (Fig 5). The insert was placed inside the OFA and the half selected for the test had grid flooring. The formalin test involved injecting 0.1ml of dilute formalin (1.25%; made from 37% formaldehyde, Merck) subcutaneously into the plantar surface of the right hind paw of the rat with a 27G needle (Becton Dickinson & Co.).The animal was restrained manually during the hind paw injection. Following the injection, the animal was placed in the selected half of the ‘two-compartment insert’. The nociceptive behaviours were scored for 90min starting immediately after the injection of formalin. Parameters such as number of flinches and duration of licks were recorded manually by the observer while ambulatory distance was recorded by the computer. Flinching comprises shaking of the injected hindpaw, flinching of the hindquarters and the reflexive withdrawal of the injected hind paw. Each occurrence of shaking, flinching of hindquarters and reflexive withdrawal of the paw was considered as one count of flinching. The duration of licks was recorded as the time of licking of both the dorsal plantar surface of the injected paw. Biting of the nails was not included in scoring the nociceptive behaviour. After 90 min, the animal was placed back and habituated in the carrier box with water and food in the behavioural room for another 30 min. 2 hours after the injection of formalin, the animal was perfused (see section 2.4.1) and the spinal cord take out for c-Fos immunocytochemistry. 49 Formalin Test Apparatus I/R dischargers Computer Signal processing system I/R detectors Figure 5: Formalin test apparatus. For the purpose of the formalin test a two-compartment insert was placed into the Open Field Activity chamber. The two compartments of the insert were of similar size (21.7 X 43.2 X 30.5 cm each) but each had different flooring, namely grid or rod flooring. The rat was tested in the half with the grid flooring. 50 2.4 Immunocytochemistry 2.4.1 Tissue preparation The animal was deeply anaesthetized with urethane (1.5g/kg, i.p; Sigma, St Louis, MO, USA). Heparin (0.1ml, 1000IU/ml; Sigma, St Louis, MO, USA) was injected into the left ventricle of the heart followed by transcardial perfusion of 400ml sodium nitrite solution (1%, Merck, Germany), and later of 400ml of paraformaldehyde solution (4%, Merck, Germany). The two solutions were made in 0.5M and 0.1M sodium phosphate (Merck, Germany) buffer, respectively. The brain and the spinal cord were removed, blocked, post fixed in the above fixative for approximately 24 h and 4 hr at 4 ˚C, respectively. Alternate coronal 60 µm sections were taken through the brain and the lumbar spinal cord on a microtome with vibrating blade (Leica, Germany) and collected in 0.05M Tris-buffered saline (Fisher Scientific). 2.4.2 Staining procedures Previously described method (Zheng and Khanna, 2001; Khanna et al., 2004) were adapted for OX42 labeling and Fos immunocytochemistry of the brain sections. Alternate sections of the L4 lumbar spinal cord and the brain sections taken through the hippocampus were immunolabeled for c-Fos. In addition, alternate brain sections through the SuM were immunolabeled for OX42. The 51 Avidin-biotin-peroxidase complex (ABC) technique was applied to detect the antigen. Anitibody OX42 recognizes the rat equivalent of human CD11b, the receptor for the iC3b component of complement. The antigen is expressed on most macrophages, including resident and activated peritoneal macrophages and Kupffer cells and around 35% of alveolar macrophages. The antibody also labels dendritic cells, granulocytes and microglial cells in the brain (Whiteland et al., 1995). All procedures were carried out at 4 ˚C unless otherwise specified. Sections were incubated in wells of a 24-well cell culture plate (1 section/well, 300µl /well for brain; 2 sections/well, 250µl/well for spinal cord). Briefly, the sections were rinsed with 0.3% hydrogen peroxide to quench endogenous peroxidase for 20 min, and were incubated for 2 h in room temperature in 3% bovine serumalbumin (Sigma, St Louis, MO, USA) in 0.05 M Tris-buffered saline with 0.3%Triton X-100 (Bio-Rad Labs, USA) to block non-specific protein interactions. Subsequently, the sections were incubated for 70 h with the primary antibody (1:1000 mouse anti-rat anti-OX42 monoclonal antibody, AbD Serotec, Germany or 1:2000 rabbit anti-Fos (Ab-5) polyclonal antibody, Calbiochem, Germany espectively), followed by 3 h incubation with the secondary antibody (1:1000 biotinylated anti-mouse antibody for OX42, Vector Labs,USA or 1:1000 biotinylated goat anti-rabbit antibody for c-Fos, Calbiochem, Germany, respectively). All antibodies were diluted in 0.05M TBS containing BSA and 0.3% Trition® X-100. Sections were then treated 52 with the ABC (Vectastain Elite ABC Kit, Vector Labs, USA) for 3 h with constant shaking at room temperature on an orbital shaker (N-biotek, Korea) followed by diaminobenzidine treatment (DAB in 0.05% in 0.05 MTBS; Sigma, USA). Once the brown immunolabel had developed, the reaction was stopped. The sections were mounted on chrome alum gelatine-coated slides, air dried, dehydrated via ethanol (BDH Laboratory Supplies, UK) and ethanolxylene (J.T.Baker, Canada) cleared with xylene and coverslipped with DePeX (BDH Laboratory Supplies, UK). The Fos-like immunoreactivity (FLI) was visualised as brown reaction product. 2.5 Data analysis 2.5.1 Quantification of the size of the lesion in the brain Sections through the SuM region in the brain were digitized at 500 dpi using Montage Explorer (Synoptics Ltd.) using a Nikon Eclipse E408 microscope (Nikon Corp., Japan) at 2x. The lesion area in each section was clearly observed as dark brown cluster of activated microglia that was immunostained with OX-42 antibody. The quantification of the lesion size was done using commercially available software (AISTM Image Analysis Software; Imaging Research Inc., Canada). Briefly, the outer margin of the immunolabeled region was taken as the boundary of the lesioned area (see Fig 9). The software was programmed to detect and highlight the demarcated region which was at least three times more intense than the background. The condition was set based on comparisons with manual selections of the lesioned area. Subsequently, the 53 software calculated the area of the highlighted region. Given that the thickness of the section was 0.06 mm, the volume of the lesion region was obtained by multiplying the area with 0.06 for each section. 2.5.2 Quantification of Fos positive cells in spinal cord and hippocampus Alternate sections of the lumbar L4 spinal cord were digitized at 500dpi using Montage Explorer (Synoptics Ltd.) using a Nikon Eclipse E408 microscope (Nikon Corp., Japan) at 4x and counted using a commercially available software (AISTM Image Analysis Software; Imaging Research Inc., Canada). Briefly, the average intensity of each spinal region was determined and Fospositive cells in that region were detected as stained nuclei that were 300% intense as compared with the average intensity of the region (Lee et al., 2008). Besides the intensity difference, another two criteria were applied in identifying c-Fos positive neurons: 1) an area of greater than or equal to 5 units, and 2) a form factor of greater than 0.8. The form factor is a measure of the degree of roundness and a form factor of 1 is perfectly round. The c-Fos positive cells were counted for the spinal cord on the right side (ipsilateral to the formalin injection) of the lumbar L4 spinal segments. Laminar specific counts were made for four regions as described (Khanna et al., 2004). The spinal segmental level and the laminar organization were identified based on the configuration of the gray matter (Molander et al., 1984). The regions with corresponding laminae were: (A) superficial dorsal horn (laminae I-II), (B) nucleus proprius (laminae III-IV), (C) neck of the dorsal 54 horn (laminae V-VI) and (D) the ventral gray (laminae VII-X) (Fig 6). In addition, the total count was averaged for all L4 sections of each animal and then for the experimental group. The number of spinal sections through L4 that were analyzed for FLI cells ranged from 15-20. 55 Figure 6. Diagrammatic representation of lumbar L4 spinal cord coronal section with laminar subdivisions. (A) Superficial dorsal horn (laminae I-II); (B) nucleus proprius (laminae III-IV); (C) neck of the dorsal horn (laminae VVI); (D) ventral gray (laminae VII-X). (Adapted from Molander et al., 1984; Khanna et al., 2004) 56 The hippocampal c-Fos-positive cells counted manually using an Olympus microscope at 40x. A grid was inserted into the eyepiece that facilitated counting in non-overlapping squares over the region of interest. The cells were counted bilaterally in the pyramidal cell layer of CA1 and CA3 and dentate granule cell layer from alternate sections taken through the anterior-posterior extent of the hippocampus corresponding to bregma -1.72 to bregma -5.76 mm (Paxinos and Watson, 2007). Generally, 30-35 hippocampal sections were analysed per animal. As with spinal cord, the total count for each hippocampal region was averaged for the sections of each animal, and then for the experimental group. Such averages were also calculated for dorsal and ventral CA1. The dorsal CA1 included the anterior region of CA1 (bregma -1.72 to bregma -4.65 mm; Fig.7) and the dorsal CA1 in the posterior sections of the hippocampus (bregma -4.65 to bregma -5.76 mm; Fig.8). The dorsal CA1 in posterior sections was demarcated from ventral CA1 by the intervening CA2. 2.5.3 Statistical Analysis The results are expressed as mean ± S.E.M. One-way ANOVA with post hoc Newman-Keul’s test for multiple comparisons was performed for statistical comparisons of data from more than two groups. Comparison of data between two groups was performed using two-tailed unpaired t-test. Statistical significance was accepted at p[...]... population spike suppression In addition, there may be functional and neurochemical heterogeneity along the mediolateral axis of SuM neurons in controlling different aspects of hippocampal theta such as frequency and amplitude 1. 2.4 Effects of manipulations of SuM on behaviour and hippocampal theta The effect of manipulations of SuM has been examined on a variety of animal behaviors Especially the focus of. .. contributes to the negative affective-motivational state during pain In the present study we explored whether PH-SuM as well modulate animal pain behaviors in the formalin model of persistent inflammatory pain Formalin (1. 25%, 0.1ml) was injected subcutaneously into the plantar surface of the right hind paw of PH-SuM lesioned or control nonlesioned animals The lesion was induced by microinjection of. .. al., 19 76; Ottersen, 19 80; Swanson, 19 82; Haglund et al., 19 84; Gonzalo-Ruiz et al., 19 92b; Hayakawa and Zyo, 19 94; Leranth and Kiss, 19 96; 15 Borhegyi et al., 19 98; Vertes and McKenna, 2000; Kiss et al., 2002) In addition, the projections from SuM and to SuM are chemically very diverse In the following section, more focus will be given to connections pertinent to hippocampus and septal area 1. 2 .1 Projections... periolivary region Shaffer collateral sis-inducible element supramammillary Lateral supramammillary transient receptor potential family of ion channels wide dynamic range wheat germ agglutinin conjugated to horseradish peroxidase x LIST OF PUBLICATIONS ABSTRACT Liu LM and Khanna S (2009) Effects of posterior hypothalamic lesions on formalin- induced pain behaviours Proceedings of the International Australasian... Buzsaki, 19 96) 1. 1.2 Intrinsic connection within the hippocampus formation A unidirectional intrinsic connection which consists of a series of excitatory pathways enables a sequential flow of information through the hippocampal 6 formation It is organized in a trisynaptic circuit, starting from the perforant pathway from the entorhinal cortex to the dentate gyrus, continues with the mossy fibres projection... thalamus, posterior hypothalamus, amygdale, medial prefrontal and the entorhinal cortices (Swanson, 19 82; Shibata, 19 87; Vertes, 19 88; Hayakawa et al., 19 93; Risold and Swanson, 19 97) 1. 2.2 Projections to SuM Injections of wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP) into the supramammillary nucleus resulted in retrogradely labelled neurons in MS-VLDBB region and other regions including... Glutamate stimulation of SuML produced a substantial increase in the perforant path stimulation -induced population spike in the dentate gyrus (Carre and Harley, 19 91) Inactivation of the mSuM with GABA abolished reticularly-elicited suppression of CA1 population spike, thus indicating that SuM mediated reticularly-elicited suppression of CA1 pyramidal cell synaptic excitability to CA3 stimulation (Jiang and... regulation of entorhinal synapses terminating on the distal dendrites of the pyramidal cells (Blasco-Ibanez and Freund, 19 95; Katona et al., 19 99) 10 1. 1.3 Extrinsic connection of hippocampus formation Two types of inputs to hippocampal principle cells have been identified: 1) laminated afferents from cortical areas (such as entorhinal area and amygdala) terminating in the specific layers along the... Projections from SUM SuM as a whole contains calretinin-containing cells (Kiss et al., 2000), dopaminergic cells (Swanson, 19 82), and substance P containing cells (Borhegyi and Leranth, 19 97; Fig 3) The most medial portion of SuM (mSuM) contains small packed cells and is distinguished by having substance P containing cells Cells in this region have weak connections to hippocampus and these connections... Antal, 19 88) Glutamatergic neurons concentrate in the medial septal regions and are distributed among the cholinergic and GABAergic neurons (Colom et al., 2005) 1. 1.3.2 Hippocampal-septal projection The hippocampal-septal projection is the most prominent efferent projections from the hippocampus among all the efferent projections from the region (Leranth et al., 19 92) The main subcortical target of this ... connections 11 1. 1.3.2 Hippocampal-septal projection 12 ii 1. 1.4 Hippocampal theta rhythm 1. 2 Anatomy of supramammillary area (SuM) 13 15 1. 2 .1 Projections from SuM 16 1. 2.2 Projections to SuM 18 1. 2.3... schedule 22 1. 3 Pain 24 1. 3 .1 General description of pain 24 1. 3.2 Pain transmission pathway 27 1. 3.3 Formalin model of persistent pain 32 1. 4 c-Fos 37 1. 4 .1 General description of c-Fos 37 1. 4.2 Fos... gyrus 1. 1.2.2 The mossy fibres projection from the dentate gyrus to CA3 1. 1.2.3 The Schaffer collaterals to CA1 1. 1.3 Extrinsic connection of hippocampus formation 11 1. 1.3 .1 Septo-hippocampal connections