| BackgroundThe majority of patients with metastatic bone cancer will experience moderate tosevere pain. Bone pain is one of the most common types of chronic pain in thesepatients.Bone cancer pain is usually progressiveas the disease advances, and is particularlydifficult to treat. The mechanisms responsible for bone pain are poorly understood, butbone cancer pain seems to be enhanced by a state of spinal sensitization. Previous studiesindicate that spinal sensitization of bone cancer pain appears unique when compared tochanges that occur in persistent inflammatory or neuropathic pain states. It has beenhypothesized that one such sensitization mechanism is the regulation of gene expression.Thus, identification of unique gene expression profiles in the spinal cord from animals withbone cancer pain may provide insight into the widespread factors that drive spinal cordplasticity. This, in turn, will contribute to the development of more effective treatments forbone cancer pain.Genechip of full gene probe sets instead of traditional approach andcoulddescribe all of the potential genes that may contribute to generation and maintenanceof chronic pain.In the present study, whole-genome mRNA expression profiles in spinal cords of rats with bone cancer pain, inflammatory pain and neuropathic pain wereexamined to identify target genes that may contribute to spinal sensitization of chronic pain.The mRNA expression was examined using Affymetrix Rat Genome230.2microarrays thatinclude30,000probe sets. The function and cross-effect of the pain-related genes wereanalysised with cluster and molecule annotation system to reveal the molecule mechanismin development of chronic pain.Spinal sensitization is the main mechanism of chronic pain.Plasticity changes ofsynapseincluding modifications of existing synapses and formation or loss of synapticconnections contribute to sensitization. Previousstudies demonstrated that bone cancer painresulted from an augmentation of excitatory synaptic transmission between neurons over awide area of spinal lumbar segments, but whether the synaptic plasticity is due tomodifications of existing synapses or the formation of new synaptic connections is stillunknown. Here we have shown that sarcoma implantation induced excitabilitysynaptogenesis which drives the development of bone cancer pain. But the mechanisms ofsynapse formation induced by sarcoma inoculation are still unknown.New synapse establishment requires an interaction between axons and dendrites,accompanied by the appositional organization of pre-and postsynaptic specializations.Thus, the axon guidance cues are required to help establish specific connections between aneuron and its multiple cellular targets. Slit is a largeextracellular matrix protein that wasidentified and implicated in axonal guidance and branching during the development ofnervous system. Roundabout (Robo) was found to encode a protein which is atransmembrane receptor of slit and a member of the immunoglobulin superfamily. Slits actthrough receptors of the Robo in axon chemorepulsion and guidance cues during thedevelopment of the central nervous system. The small GTPase RhoA, a signaling proteinregulating neuronal morphogenesis, plays a critical role in regulationof axonalre-generation.In our chip array experiments, we have identified tSlit2upregulation andRobo1and RhoA downregulation in rats with bone cancer pain. Thus, it seems reasonable to hypothesize that Slit2/Robo1via RhoA mediate the synaptogenesis and contribute to thespinal sensitization of bone cancer pain. To test this notion, siRNA was used to knock downSlit2and Robo1in vivo and in vitro, a carcinoma tibia implantation rat model was used totest the bone cancer-related pain behaviors, and synaptogenesis was examined in vivo andvitro. We have shown thatcarcinoma inoculation induces excitatory synaptogenesis andbone cancer pain behaviors which are reversed by Slit2knockdown but aggravated byRobo1knockdown. In vitro experiments, neurite outgrowth and synaptogenesis of culturedneurons are inhibited by knockdown Slit2, but enhanced by Robo1knockdown. And thatcarcinoma implantation induces an increase of Slit2and decrease Robo1andRhoA.whileSlit2knockdown results in increase of Robo1and RhoA via N-terminal of Slit2directly bounding to Robo1and Robo1knockdown decreased RhoA with Slit2uneffected.These results indicate thatSlit2inhibiting Robo1and then RhoA promotessynaptogenesisand contributes to bone cancer pain. Our findingssuggest a new spinal sensitizationmechanism underlying the development of bone cancer pain.Methods and results1. Unique gene expression profiles in spinal cord of rats with bone cancer painMethods To identify altered genes that might contribute to spinal sensitization of chronicpain, we conducted a time-course mRNA profiling experiment on carcinoma tibiaimplantation, L5spinal nerve ligation (SNL), and intraplantar complete Freund’s adjuvant(CFA) injection rat models. The mRNA expression in lumbar spinal cord of respectiveanimals was examined using Affymetrix Rat Genome230.2microarrays that include30,000probe sets.The gene expression profiles in the chronic pain rats were compared withthe corresponding data of sham.Results We identified2245,1989and1679probe sets that were upregulated respectivelyon days3,7,14post-CFA injection,1570,1560and1494probe sets that were upregulated respectively on days7,14,21post-SNL, and1381,1740and1908probe sets that wereupregulated respectively on days7,14,21post-carcinoma implantation.These genes mayinvolve the onset or maintenance of chronic pain. By comparing the microarray resultsamong the three rat models, we were able to identify a large set of genes whose expressionswere upregulated at each time point as a unique result of CFA injection, SNL andcarcinoma implantation (352,196and216probe sets identified respectively). These genesmay be involved in the maintenance of inflammatory, neuropathic and bone cancer pain,respectively. When the genes whose expressions were upregulated were pooled together,1093,884and344genes overlapped among three time points and344gene sets wereoverlap among three models,.This suggests that these three types of pain on the one handshared a similar mechanisms, but on the other hand have a special gene expression profilethemselves.2. Slit2/Robo1promotes synaptogenesis and contributes to bone cancer painMethods A Local injection of carcinoma cells directly into rat tibia were used to mimicclinical bone cancer pain and the siRNA lentivirus were injected into the ipsilateral dorsalhorn to specially knock down Slit2or Robo1expression. Coimmunoprecipitation assayusing L3-5spinal cord of carcinoma implantation rats and cultured neurons was used to testwhether functions of Slit2in bone cancer pain was mediated through direct interaction withRobo1. Immunoprecipitation was performed using antibody to Slit2and Robo1, Slit2weredetectable by western blotting using antibodies to Robo1, Slit2. Paw withdrawal thresholdwas measurement of mechanical allodynia.Tthe subcellular distribution and expression ofSlit2, Robo1and RhoA were tested by mmunofluorescent stain, western blot and RT-PCR.Synapse was examined with Transmission Electron Microscopy and excitatory synapse wasexamined using double immunofluorescent labeling with the anti-Synaptophysin (Syn), apresynaptic vesicle protein and anti-PSD95antibody, a major scaffolding protein in theexcitatory postsynaptic density (PSD).Results Immunofluorescence stain indicated that Slit2, Robo1and RhoA were colocalizedin neuron of dorsal spinal cord. Carcinoma implantation resulted in an increaseof Slit2, but decrease of Robo1and RhoA in ipsilateral dorsal horn. Transfection of dorsalhorn neurons with siRNAto slit2reduced the level of endogenous slit2protein, butincreased Robo1and RhoA, while knockdown of Robo1decreased RhoA and had no effecton Slit2. These Immunohistochemical studies were confirmed by RT-PCR and immunoblotanalysis. The results of immunoprecipitation suggested that Slit2could bind to Robo1directly in vivo and vitro.Taken together, carcinoma implantation upregulated Slit2whichinhibited the expression of Robo1and subsequent RhoA via Slit2binding to Robo1. Wehad showed that carcinoma inoculation resulted in significant bone cancer-related painbehaviors on the ipsilateral paw, but not on the contralateral side. Bone cancer pain wereattenuated by Slit2knockdown but aggravated by Robo1knockdown. These resultsindicated that upregulation of Slit2and thereby downregulation of Robo1and then RhoAwere necessary, but not sufficient for the development of bone cancer pain. Quantitativeanalysis synapse by Transmission Electron Microscopyshowed that synapse numberbroadly increased in cancer bearing rats and knockdown of slit2resulted in a significantreduction in the synapse number.Moreover, the opposite effect was observed upon Robo1knockdown, which resulted in a significant increase in the synapse number. Together withthe quantitative experiments of slit2, Robo1and RhoA expression, the results suggestedthat upregulation of slit2induced by carcinoma implantation resulted in an increase ofsynapse in dorsal horn by inhibition of Robo1and RhoA. Using confocal microscopy, weshowed that both the PSD95and Syn expression increased in ipsilateral dorsal hornbut notin contralateral of cancer bearing rats. Quantified analysis of the magnified images showedthat synapse number was significantly higher in cancer bearing rats compared with sham.Moreover, injection of siSlit2-LV into dorsal horn resulted in decreasing synaptic formation,but siRobo1-LV led to an opposite effects.RT-PCR and immunoblot analysis revealed thatcarcinoma implantation induced a prominent increase both in the levels of PSD95andSynwhich were reversed by knockdown of Slit2, but further increased by Robo1 knockdown. These results indicated that upregulation of Slit2inhibited Robo1andpromoted the excitatory synaptogenesis in dorsal horn post-carcinoma implantation.3. Slit2/Robo1promotesneurite outgrowth and contributes to synaptogenesisMethods To test whether upregulation of Slit2promote excitability synapse formationthrough inhibiting Robo1and subsequent removing the RhoA-inhibiting effect on theprocesses of axon enlongation and arborization, we first investigate the subcellulardistribution and interaction among Slit2, Robo1and RhoA. Cortial neurons from embryos(E17–E19) of pregnant rats were cultured, transfectedat5day in vitro (DIV)with a plasmidencoding green fluorescent protein(GFP) together with anRNAi to Slit2or Robo1or acontrol RNAi, and examined at10DIV. To unambiguously visualize multiple branchesfrom a single neuronal cell body, neurons were cultured at low density. We usedimmunocytochemistry to investigate the subcellular distribution of Slit2, Robo1and RhoA.Neuronal branches were examined using immunohistochemistrical staining ofmicrotubule-associated protein2(MAP2), a neuronal marker. We next examinedtheexcitability synapse formation of cortial neurons cocultured with astroglia from embryosof pregnant rats. The siRNA lentivires transfected cells at5DIV and and fixed10days laterfor staining with antibodies that recognize Synand PSD95and Syn.To quantify the numberof synapses formed on the normal or transfected neuron, we counted the number of apposedSyn/PSD95puncta along dendrites of normal or GFP-expressing neurons.Results We used immunocytochemistry to investigate the subcellular distribution of Slit2,Robo1and RhoA and found that they were broadly co-expressed onneuronal soma anddendritesas well as axons. Knockdown Slit2inhibited Robo1and RhoA, while knockdownRobo1resulted in decrease of RhoA but not Slit2. The results were further confirmed byRT-PCR and western blot. Together with immunoprecipitation study descripted above usingcultured-neurons, these results indicated that Slit2bound to and inhibited Robo1, which inturn inhibited RhoA in vitro. These results were agreed with those of animal experimentupon. Using immunohistochemistrical staining of microtubule-associated protein2(MAP2), we showed that Slit2knockdown strongly reduced neurite number and length, whileknockdown of Robo1enhanced neurite outgrowth. Together with quantitative analysis ofSlit2, Robo1and RhoA expression in neuron, the results indicated that Slit2promoted axonelongationand branching probably due to inhibiting Robo1and sequent blockingRhoA-inhibition on neurite extension. Thus we next examined the synapse formation ofneurons cocultured with astroglia from embryos of pregnant rats. Quantify the number ofsynapses showed that knockdown of Slit2resulted in a significant decrease inexcitatorysynaptic numbe. Conversely, knockdown of Robo1induced an increase of excitatorysynapse.4. Statistical analysisAll data are presented as means±SD. The statistical significance of difference betweenvalues was determined by analysis of variance (ANOVA). For all analyses, significancewas set at P <0.05.Conclusion1. We have identified hundreds of altered genes that may mediate the onset ormaintenance of inflammatory, neuropathy and bone cancer pain respectively. When thegenes whose expressions were upregulated were pooled together, lots of gene sets wereoverlap among three models.Thissuggests that these three types of pain on the one handshared a similarmechanisms, but on the other hand have a special gene expression profilethemselves.2. Slit2mediates axon guidance and branch and promotes excitatory synaptogenesis viainhibition of Robo1and subsequent RhoA–inhibition of axon remodeling and contributesbone cancer pain. SignificanceIn the present study, we have identified the unique gene expression profiles in the spinalcord from animals with bone cancer pain may provide insight into the widespread factorsthat drive spinal cord sensitivity of pain. This, in turn, will contribute to the development ofmore effective treatments for bone cancer pain.Synaptic plasticity is fundamental to spinal sensitivity of bone cancer pain. Here we haveshown that excitatory synaptogenesis contributes to bone cancer pain and described thatSlit2, Robo1and RhoA act as such cues that promote neurite outgrowth and guides theaxon for synapse formation. These results havedemonstrated a molecular mechanism ofsynaptogenesis in bone cancer pain.This maybe lead to a new treatment target for bonecancer pain. |
PDF Full Text Request