Skip to main content
European Commission logo print header

The NGF system and its interplay with endocannabinoid signalling, from peripheral sensory terminals to the brain: new targets for the development of next generation drugs for neuropathic pain

Final Report Summary - PAINCAGE (The NGF system and its interplay with endocannabinoid signalling, from peripheral sensory terminals to the brain: new targets for the development of next generation drugs for neuropathic pain)

Executive Summary:
This final report is updated also with the results obtained by PAINCAGE in the extension period from 1/04/2017 to the 31/12/2017. The extension of the Project was granted and approved by the Project Officer, on the basis of the request by the Project Coordinator. For the sake of clarity, we are reporting in bold the WP4, WP5 and WP7 (as per the Extension request document) results obtained within the extension period.

The PAINCAGE Project investigated the role of the Nerve Growth Factor (NGF) system in pain mechanisms by comparing the effectiveness and/or safety of candidate drugs targeting different components of the NGF/TrkA/p75NTR axis and by investigating its interactions with the endocannabinoid (EC) system. PAINCAGE focussed on different levels of pain transmission and perception, from peripheral sensory terminals to the brain. The PAINCAGE project has satisfactorily addressed its objectives and provided the mechanistic basis to accelerate i) the development of a new generation of painkillers and ii) the characterization of new pain relevant targets.

The main results achieved by PAINCAGE have been:
The analgesic actions of anti-NGF and anti-TrkA antibodies in Neuropathic Pain (NP) model induce a long lasting analgesia by over three months which correlates to specific gene expression fingerprinting and epigenetic mechanisms occurring at different levels of the ascending pain pathways. The analysis of the pain-related DNA methylation and histone acetylation by anti-NGF and anti-TrkA treatments display different kinetics. Methylation results to be an earlier event linked to the analgesic anti-TrkA effect, while acetylation follows. The long term analgesic effect of the anti-TrkA treatment in the NP model might be driven, at earlier time point, by regulating the genome-wide methylation status and by modulating Histone acetylation mechanisms subserving a long lasting “analgesic” gene expression program. The underlying mechanisms have been characterized, including the identification of new targets from transcriptomic and epigenetics studies and the modulation of pain mediators by inflammatory cells. This provides a strong validation for anti- TrkA as a new analgesic drug for NP with long lasting analgesic properties based on epigenetic mechanisms. A new drug candidate, the p75NTR receptor scavenger (p75NTR-Fc, or LEVI-04) has been studied and developed by the SME Levicept within the PAINCAGE consortium. Concerning osteoarthritis (OA), the systemic administration of anti-TrkA, anti-NGF and of a p75NTR scavenger effectively reverted allodynia in this model. Interestingly, with reference to safety, it was found that the adverse event of "accelerated OA", previously reported in a subset of OA patients receiving anti-NGF antibodies, can be replicated in rodents receiving anti NGF (a very significant reverse clinical-to-preclinical translation), showing instead that the p75NTR-Fc is safer and does not induce this osteonecrosis. These data provided a strong basis for clinical testing p75NTR-Fc in OA patients (even if the clinical trial per se has not been funded by PAINCAGE): a first-in-human phase 1, study of LEVI-04, to evaluate the safety, tolerability and pharmacokinetics of p75NTR-Fc, has started in July 2017 in healthy volunteers and OA patients. Congenital pain insensitivity is the other side of the coin of chronic pain: mutations in TrkA or NGF genes determine congenital pain insensitivity (Hereditary Sensory Autonomic Neuropathy (HSAN) type IV and V, respectively. We have generated lines of knock-in mice harbouring TrkA and NGF HSAN mutations, demonstrating that they replicate the pain insensitivity clinical phenotype. These mouse models of painlessness genetic disorders in patients, together with mice in which a TrkA mutation exerts instead the opposite phenotype of pain sensitization, have been exploited to dissect mechanisms linking alterations in the NGF-TrkA signaling axis to the processing and central perception of pain. We have made the surprising discovery that growing without pain leads to deficits in pain-related fear responses but not in innate fear responses not related to pain.
Concerning the EC system, NP determines a decrease in cortical synaptic plasticity, via a reduction of EC receptor CBR1 expression. Consistently, we found that the selective deletion of CBR1 expression in brain neurons correlates with a reduced threshold for pain and CBR1 are necessary for the analgesic effect of electroacupunture. We further showed that mice with a specific deletion of CB1 receptor in both glutamatergic and GABAergic neurons in the forebrain (GABA/Glu-CB1KO) have a significant mechanical hypersensitivity, demonstrating the crucial role of CB1 expression in forebrain glutamatergic and GABAergic neurons for the encoding of sensory processing and of pain. Most notably, the reduction of mechanical sensitivity observed in GABA/Glu-CB1KO mice was reversed by sequestering NGF with the anti-NGF antibody alfaD11. This suggests that the analgesia induced by anti-NGF antibodies requires forebrain CB1 receptors to be maintained. Finally, a novel form of the CBR1, specifically expressed in mitochondria (mtCB1Rs), has been discovered to mediate the functional effects of EC in the central nervous system. The development of a technological platform based on activity reporters (TRAP/ permanent access to transiently active neurons) allowed mapping pain pathways under different pathological or pharmacological conditions. We focused on the interrogation of molecular and cellular determinants of pain sensation at the DRG and spinal levels and showed that the differential expression and sensory-specific contributions of the calcium binding proteins secretagogin and NECAB2 (-1) gate excitatory neurotransmission in inflammatory but not neuropathic pain. These data set the basis to dissect the molecular composition, differential responsiveness and synaptic wiring of pain circuits.
A lead small molecule peptide agonists of NGF was identified on the basis of structural data.
Finally, we have identified NGF- and EC-linked clinical biomarkers for NP, validated in animal models and in clinical samples, which may result in clinical benefits for the stratification of patients suffering from different neuropathies.
The NGF and the EC systems are crucial regulators of pain pathways. The results of the project provide a solid basis to accelerate the development of already identified next-generation therapeutics (anti-TrkA, p75NTR-FC and anti sortilin), based on the NGF and EC systems, as well as for the identification and validation of new druggable targets that were identified and validated in PAINCAGE.
Project Context and Objectives:
The PAINCAGE Project has been funded under the European Commission's 7th Framework Programme from April 2014 to March 2017, extended to December 2017, under the Theme: Health.2013.2.2.1-5 Understanding and controlling pain.

Chronic pain (CP) is a huge unmet medical need and more effective safe drugs are badly needed. Among the most common forms of CP are: i) neuropathic pain (NP), associated with nerve injury in the peripheral or central nervous system; and ii) osteoarthritis (OA), a progressive degeneration of cartilage of joints, characterized by inflammation and pain. Sofar there is no effective treatment for these highly prevalent disabling disease states and current treatments (NSAIDs and opioids) cause serious unwanted side-effects. The NGF ligand-receptor system has recently emerged as a novel target for NP and OA, with a great therapeutic potential. NGF and its receptors appear to be master regulators of these forms of pain, controlling both neuropathic and inflammatory components. Excessive levels of NGF, caused by tissue injury or other factors, result in the rapid sensitization of nociceptors, reducing the threshold for pain sensation and increasing the pain intensity experienced (e.g. allodynia and hyperalgesia). Conversely, human genetic studies further validated NGF as a target, since mutations in the genes for NGF receptor TrkA and for NGF itself cause congenital forms of insensitivity to pain (Hereditary Sensory Autonomic Neuropathy type IV, HSAN IV and HSAN type V, respectively). For these reasons, blocking the NGF signaling system is a rational and thoroughly validated approach to pain therapy. Extensive evidence for potent analgesic efficacy of anti-NGF monoclonal antibodies (MAbs) has been obtained in different preclinical models of CP, from NP to inflammatory models (Cancer pain and Osteoarthritis, Interstitial Cystitis, respectively), as well as in clinical trials, showing remarkable analgesic efficacy and creating great expectations for this new class of analgesic compounds. Therefore, targeting this system activates several different downstream cascades providing exciting therapeutical opportunities. Indeed, NGF modulates, by the PLC/PI3K/MAPK pathway, the TRPV1 channel (transient receptor potential, vanilloid subfamily type 1), a molecular integrator of painful stimuli. Molecules that modulate TRPV1 activity are currently evaluated as pain treatments. TRPV1 is also endogenously activated by the EC, anandamide. ECs display potent analgesic activity by interacting with cannabinoid receptors (CBR) type 1 and 2. Background data from Consortium partners showed that NGF regulates EC signaling, underlying a novel NGF/ECs interplay amenable for drug targeting.
Thus, blocking the NGF signaling system is a rational and well validated approach to pain therapy. However, safety concerns related to off-target side effects have been raised, after the results of initial clinical trials with the anti-NGF antibody tanezumab reported cases of suspected osteonecrosis that led to total joint replacements. This led to a clinical hold of the anti-NGF programs. Further investigation showed that most of these cases were due to rapidly progressing OA (RPOA) rather than osteonecrosis. After much consultations and data analysis, the FDA lifted eventually the clinical hold and called for more preclinical research. To fully exploit the huge therapeutic potential of NGF system, we built a consortium of leading researchers in the NGF, EC and pain scientific arena. This interdisciplinary approach was aimed at understanding the specific molecular mechanisms and cellular interactions linked to the regulation of pain processing at different levels of the nociceptive pathways triggered by NGF and dampened by the EC systems. The successful outcome of the project has been ensured not only by the carefully planned objectives but also by an interdisciplinary, complementary and well-integrated network, involving both academic groups and SME.
This innovative project investigated new strategies for the treatment of OA and different NP forms, based on the NGF system and its interplay with EC signaling. The key elements of the project have been:
i) a well validated target system, with a strong basis of lead therapeutic compounds;
ii) a strong mechanistic basis for the interaction between the NGF and the EC systems;
iii) a focus on different levels of the pain transmission systems, from peripheral sensory terminals to the brain. The input to the project was represented by:
i) a set of existing candidate targets (e.g. TrkA, p75NTR, sortilin, TRPV1, CBRs) belonging to the NGF and the EC system;
i) a set of existing lead biological drugs (alphaD11 anti-NGF antibody, MNAC13 anti-TrkA antibody, p75NTR-Fc (LEVI-04), anti-sortilin antibodies and small molecules); In particular, the anti-TrkA antibody is a first-in-class, currently in Phase I in its humanized form, and the p75-Fc is also a first in class, that
entered clinical trials at the end of the project.
iii) extensive, cutting edge knowledge, expertise, models and reagents on the NGF and EC systems, both of which are crucially involved in NP;
iv) a set of clinically and genetically characterized human amples.


The PAINCAGE project has pursued the following specific objectives by a cross-disciplinary and modular work-program:
WP1 - Understand the role of the individual components of the NGF target system (NGF, proNGF, TrkA, p75NTR, sortilin) in the signaling and regulation of different forms of NP with the neurotrophic, neuropathic and inflammatory components, including OA.
WP2 - Dissect endocannabinoid signaling (ligands and receptor systems) in relation to the NGF pain- related signaling pathways in NP.
WP3 - Validate the NGF involvement in NP by the development and pharmacological/behavioral characterization of insensitivity to pain in new transgenic, pain insensitive, mouse models (HSAN IV and V).
WP4 - Analyze epigenetic mechanisms in NP and OA by focusing on the long-term gene expression changes during NP onset and NGF-targeting analgesic treatments in NP and OA. The objective
was to provide new targets for future development, possibly related to the epigenetics regulation in pain.
WP5 - Control NP and OA onset, progression and perception by pharmacological treatment with anti- NGF, anti-TrkA and p75-Fc (NGF scavenger). Transgenic models have been used to define specific roles of CB1R, TRPV1, CB2R and sortilin in NP.
WP6 - Analyze the safety issues related to the anti-NGF therapy, and its unwanted side effects, in comparison to the p75NTR-Fc (LEVI-04).
WP7 - Optimize the structure of lead NGF-based drug candidates for NP, based on biophysical and structural data.
WP8 - Identify and validate pain related biomarkers and druggable targets in humans, to improve patient stratification.
WP9 – Disseminate the results of the project.

PAINCAGE focuses on two molecular systems of emerging importance, the NGF and the EC system. Unifying the two systems and studying their functional interactions, at different levels of the pain transmission and perception pathways, is a highly innovative approach. By exploiting several integrated in vivo and in vitro techniques, the project investigated i) the role of the individual components of the NGF and endocannabinoid systems in the regulation of NP and OA, including its neuronal, inflammatory and neurotrophic mechanisms, ii) the clinically proven or potential liabilities and safety concerns of targeting the NGF system for the treatment of NP, iii) the regulation of new targets by the NGF system and related signalling by comparing NGF biology between NP and OA. A number of recombinant proteins and monoclonal antibodies were used within the Consortium, including anti-NGF mAb alfaD11, anti-TrkA mAb MNAC13, soluble p75NTR-Fc, anti-sortilin antibodies. The anti-NGF alfaD11, anti- TrkA MNAC13 mAbs are the murine counterpart of humanized antibodies that are currently being tested in clinical trials in CP patients. The p75-Fc NGF scavenger has entered Phase I clinical trials in OA patients, leveraging efficacy and safety results obtained in the PAINCAGE project. Compared to current pain research, the PAINCAGE Project was the first to tackle the NP and OA problem focusing on the NGF-mediated signaling through their interaction with TrkA/p75NTR and sortilin receptors and on the interaction of the NGF system with another crucial player in the pain mechanisms, the EC system. Pain perception and pain experience is not just about sensory neurons and nociceptors. A distinctive aspect of the PAINCAGE proposal was the analysis at different levels of the pain processing and perception cascade, from periphery to the central cortical areas, thereby studying also the emotional and cognitive aspects involved in NP.
The results obtained during the PAINCAGE project provided solid, mechanism-based grounds for the development of already identified second- generation therapeutics, based on the "NGF target" system, as well as for the identification and validation of new druggable targets emerging from the elucidated mechanisms. It also identified clinical biomarkers for NP, validated in animal models and clinical samples, that could result in future clinical benefits, for the stratification of patients suffering from different neuropathies and their treatment. In conclusion, the PAINCAGE project contributed to the understanding and controlling NP mechanisms, with an interdisciplinary approach leading to the development of next- generation NGF targeting drugs and to uncovering profound connections between the NGF and the EC systems in controlling pain.
Project Results:
Introduction

Blocking the NGF signaling system is a rational and now thoroughly validated approach to pain therapy. Extensive evidence for potent analgesic efficacy of anti-NGF mAbs has been obtained in preclinical models and in clinical trials, showing remarkable analgesic efficacy and creating great expectations for this new class of analgesic compounds. To fully exploit the huge therapeutic potential of NGF system, we built a consortium of leading researchers in the NGF, EC and pain scientific arena for understanding mechanisms of pain, with main interest for neuropathic pain (NP), and identifying innovative targets for better designed therapeutic drugs. NP is a form of neurogenic pain, arising from nerve injury in the peripheral or central nervous system. The input to the project was represented by a set of existing lead biological drugs (alphaD11 anti-NGF antibody, MNAC13 anti-TrkA antibody, p75NTR-Fc neurotrophin scavenger).
The anti-NGF alfaD11, anti-TrkA MNAC13 mAbs are the murine counterpart of humanized antibodies that are currently being tested in clinical trials in CP patients. The p75-Fc NGF scavenger is a new lead molecule at the preclinical development phase, that has entered Phase I clinical trials in OA patients, leveraging efficacy and safety results obtained in the PAINCAGE project. The anti-TrkA monoclonal antibody and the p75NTR-Fc scavenger are first in class molecules.
This section will report the main S&T results/foreground according to the following framework:
The NGF system and neuropathic pain (NP) Endocannabinoid (EC) signalling in NP: a tale of cell-specific, subcellular-specific and CB receptor- type-specific actions Clinical Biomarkers for NP The NGF system and Osteoarthritis (OA): efficacy and safety Other pain models and therapeutic approaches Suffering without pain (Hereditary sensory and autonomic neuropathies, HSAN IV and V) Towards structure-derived new compounds to modulate the NGF system General conclusion


The NGF system and neuropathic pain (NP)


Long-lasting analgesic effects of anti-NGF and anti-TrkA treatment in the CCI NP model: a transcriptomic and epigenomic study.

We tested in a side-by-side comparative study the analgesic efficacy of anti-NGF mAbs and anti-TrkA mAbs in the chronic constriction injury of the sciatic nerve (CCI), a NP murine model. We surprisingly observed that the anti-TrkA MNAC13 (100 micrograms/mouse/day) administered on the third day from the sciatic nerve ligature for 8 days, induces a long-lasting lasting anti-allodynic effect in mice with NP (CCI mice). The analgesic effect lasts longer than three months after the last dosing of the antibody. The nociceptive thresholds of treated mice were maximal until D90, after which the nociceptive threshold decreases, but remains significantly higher than controls until D130, before returning to the baseline level of control mice at D140 and D150. Similarly, anti-NGF alfaD11 antibody (100 micrograms/mouse/day) ameliorates neuropathic pain in the CCI model for at least 3 months after the last dosing. The pharmacokinetic profiles for both anti-TrkA and anti-NGF in serum of naive mice indicate that the antibody concentration at D45, D72 and D90 was significantly lower than that necessary to induce analgesia, demonstrating that the potent analgesic effect seen at these time points was not due to the persistence of pharmacologically active doses of anti-TrkA/anti-NGF antibodies. In parallel to the pain behavioral outcomes, we found that anti-TrkA mAb MNAC13 also rescues the maladaptive cortical synaptic plasticity, which is thought to contribute to the generation, development, and maintenance of NP. Since the antibody should not cross the blood brain barrier, these data suggest that inhibition of TrkA receptors at the peripheral level may block the spreading of central sensitization and thus the brain circuit alteration following nerve injury. The expression of proteins related to peripheral nerve de/regenerative processes and to spinal glial activation in neuropathic animals, was studied by immunohistochemistry, showing that the anti-NGF and anti-TrkA antibodies administration counteracted both the structural alterations and glia activation induced by CCI. These data suggest that the long lasting analgesic effects (at D90) of anti-NGF and anti-TrkA antibodies are independent from their continuing expression/presence and we postulated that analgesia may rely on the activation of transcriptional programs possibly involving epigenetic mechanisms. Identification of the differential profile of gene expression in the neuropathic pain model, as planned in the project, has been completed, providing a comprehensive set of transcriptomic data correlated to the long lasting analgesic effects of both the anti-NGF and anti-TrkA treatment in the CCI model. The differentially expressed genes (DEGs), modulated by anti- NGF and anti-TrkA antibodies in the dorsal root ganglia, show only a partial overlap. This suggests that the overlapping set of genes modulated by the two antibodies might be responsible for influencing a common signaling pathway, relevant for the long term analgesia. Instead, in the central nervous system (spinal cord and cortex) only the anti-NGF appears to modulate the expression profile, while the anti-TrkA, that have a major effect in the DRG, affect a very small amount of genes. The transcriptomic profile supports the pharmacological data, in that treatments with anti- NGF or anti-TrkA mAbs induce distinct time-dependent analgesic effects: the anti-NGF is active since the earliest post-CCI phase, while the anti-TrkA becomes effectively analgesic some days later. The overlap between the genes differentially expressed in the CCI model (pain sensitive genes) and those modulated by the anti-NGF or the anti-TrkA antibodies, at different time-points shows that the two antibody treatments restore a common set of genes genes to the naive expression level. Those are characterized by 26 genes differentially expressed, "cured" by both NGF and TrkA targeting treatments. Most of the identified genes were down-regulated by NP and inversely modulated (up-regulated) by the treatments, and include GABA or Glutamate receptor families and Collagen pathways. In conclusion, these data prove that anti-NGF and anti- TrkA antibodies effectively counteract NP in CCI model and identify differentially expressed genes reverted by the treatment, thus representing new pharmacological targets for pain treatment, whose functional validation is ongoing. Although the mechanisms underlining the acute-to-chronic transition in pain are still incompletely understood, epigenetic processes, including histone modifications, DNA methylation, and chromatin re modeling, have emerged as important determinants in the pathogenesis of chronic pain (Denk and McMahon 2012; Alvarado et al. 2013; Denk et al. 2013; Descalzi et al. 2015; Laumet et al. 2015). A study was undertaken to evaluate the contribution of epigenetic mechanisms to the long term analgesia (D90) induced by both anti-TrkA and anti- NGF treatments. Genome-wide methylation data show a strong impact of the pain injury on tissue- specific methylation, with an overall global hypomethylation in DRG and mainly a hypermethylation in the spinal cord. These pain-induced changes in the CCI model show a progressive, trend towards the physiological state. Data show a time-dependent re-methylation in the DRG, even with saline: the number of regions with a state of methylation like the naive increases progressively with time. 87% of the regions returned to the naive level at Day 90. The effect of the anti-TrkA treatment speeds up the re-methylation process towards the physiological state in the DRG. Also in the spinal cord we observe a return to the baseline physiological methylation state. At each time point, the anti-TrkA treatment significantly modifies the methylation level of several genome regions: these "cured" regions are therefore regions that had been methylated by the injury and that the anti-TrkA treatment demethylated, with respect to saline (spinal cord), or regions that had been demethylated by the injury and that the anti-TrkA treatment re- methylated (DRG). For DRG samples, where the number of differentially expressed genes is much greater, several genes have been identified that have modified both the methylation state and the expression level. In many of those genes a methylation increase corresponds to a reduction of expression, and vice versa. In these cases, the data therefore suggest a direct relationship between the action of anti-TrkA treatment on methylation and the gene expression fingerprint. We conclude that the long term analgesic effect of the anti-TrkA treatment in the NP model, might be driven by the genome-wide methylation status, subserving a long lasting gene expression program. This dataset provides valuable new information to understand epigenetics mechanisms in chronic pain and to identify new targets to be validated for further development. A number of epigenetically relevant targets are regulated by anti NGF and anti TrkA. Our data link the pharmacological actions of a new analgesic compound (anti trkA antibody MNAC13) to recent very hot targets for CP, acting at the epigenomic level (Denk and McMahon 2012; Alvarado et al. 2013; Denk et al. 2013; Descalzi et al. 2015; Laumet et al. 2015).
We further showed that global Histone 4 (H4) acetylation in DRG is increased by the anti TrkA MNAC13 treatment, with a peak at D24. H4 acts through inhibition of HDAC (Liang et al, 2015). The up-regulation of mGLuR2 in response to HDAC inhibitor treatment has been shown to be protective in pain conditions (Denk & McMahon, 2012) and HDAC inhibitors produce analgesic effects via the upregulation of metabotropic glutamate receptor type-2 (mGlu2), an inhibitory receptor whose activation mediates analgesic effects in different pain models (Chiechio et al, 2009). Similarly, our data show that the mGluR2 is up-regulated by MNAC13, allowing to postulate that anti TrkA analgesic effect is controlled by Histone acetylation mechanisms, through an HDAC inhibitory effect.
Our overarching question was to evaluate whether the anti-NGF and anti-TrkA long lasting analgesic effect acted through epigenetic mechanisms and whether such mechanisms might be proposed as new
pharmacological targets. We performed ChIP-Seq analysis, confirming the transcriptional dysregulation in NP. Furthermore, H3K4me3 and H3K27me3 show a higher number of differential results compared to MeCP2, which correlates well with CpG methylation. In order to provide a biological interpretation of DNA methylation-seq and ChIP-seq results, we identified those genes whose methylation level was affected by CCI and then reversed by anti-TrkA or anti-NGF compared to the saline group. These “cured” genes are either genes de-methylated by CCI and then re- methylated by the drug, or, vice versa, genes that increase their methylation level after CCI and then are demethylated by the antibody drug. The lists of cured genes have been functionally annotated and cross-checked with the transcriptomic profiles. The intersection of the epigenomic and transcriptomic data has allowed to identify a small set of validated targets potentially involved in the long lasting analgesic actions of the anti NGF and anti TrkA antibodies.


Validating sortilin as a target for NP

In order to investigate the contribution of pro-neurotrophins/sortilin signaling in NP, we exploited sortilin deficient mice and observed that these animals were resistant to the development of NP. This result was confirmed by the delayed onset of allodynia in WT mice intrathecally injected with anti- sortilin polyclonal antibodies (abs/mg/kg) at the time of surgery. The potential to rescue established allodynia was tested by i.t. injections of anti-sortilin polyclonal antibodies (abs/mg/kg) 6 days post- surgery, at which time a stable mechanical sensitivity was manifested. This treatment partially rescued allodynia for 2-3 days. Altogether these data indicate that while proNGF-Sortilin pathway affects the induction and a very short phase of maintenance of NP, the NGF and its high affinity receptor, TrkA, are likely the key targets for the NP amelioration.


Endocannabinoid (EC) signalling in NP: a tale of cell-specific, subcellular-specific and CB receptor-type-specific actions

For a deeper understanding of the circuits involved in both the analgesia and sensory transmission by (endo)cannabinoids, we used cell-type specific mutants of cannabinoid receptor type 1 (CB1R), TRPV1 KO mice and CB2R ko mice.
The first are mice lacking CB1R in specific neurons (GABAergic and Glutamatergic cells) of the forebrain while the remaining part of the CNS and PNS still express the protein. By using these mice, we uncovered a crucial role of brain CB1R in the transmission of sensory signaling. Indeed, naive GABA/GLU-CB1-KO mice were significantly more sensitive to both mechanical and thermic stimuli than their CB1+/+, Glu- CB1KO and GABA-CB1KO naive littermates, suggesting an alteration of the sensory perception. In line with the behavioral results, ex vivo recordings revealed that also the cortical synaptic plasticity in the GABA/GLU-CB1-KO mice was strongly impaired. In addition, when these mice underwent to the ligature of sciatic nerve, both the allodynia and the alteration of cortical synaptic circuits were occluded. Thus, these data denote that requirement of CB1R expression in forebrain glutamatergic and GABAergic neurons for the encoding of sensory and pain processing. Notably, the NGF antibody (alfaD11) rescues sensory processing alteration in these mice for three weeks from the start of the treatment.
This finding provides a significant piece of evidence on NGF-EC systems interaction and suggests that the long lasting analgesic properties of the anti-NGF antibody require forebrain CB1 receptors to maintain its effects.
PAINCAGE also explored the role of the brain cannabinoid receptor type 2 (CB2R) in NP. This question is a fundamental one, since the lack of psychotropic activity ascribed to CB2 derived molecules depends on the sparse distribution of this receptor in the brain. So far, while the analgesic role of CB1R agonists is well established, little information on the role and mechanisms of CB2R on pain circuits is available. In this project we provided new evidence that indeed CB2Rs are present in the cortex of naive mice. We found that, surprisingly, these proteins are intracellularly expressed in the pyramidal neurons and, once activated, drastically increased inhibitory neurotransmission by enhancing nitric oxide retrograde signal on one side, and by stimulating the trafficking of GABA A receptors into the membrane of pyramidal neuron, on the other. Conversely, activation of CB2Rs onto microglia cells is responsible for the glutamatergic neurotransmission augmentation by CB2 agonists. Finally, we found that in mice suffering from neuropathic pain the cortical CB2 signalling was highly impaired. While the CB2R-controlled glutamatergic transmission significantly increased, the enhancement of inhibitory transmission by these receptors was reduced, suggesting that selective CB2-based pharmacological treatment aimed at neurons or glia may counteract cortical hyperexcitability, a key feature of chronic pain.
In line with the new functional evidence of the intracellular localization of CB2R, partners in the project showed that CB1R are also present in mitochondria (mtCB1). In particular, we showed that the activation of mtCB1 decreased protein kinase A (PKA) activity and ATP production through inhibition of the respiratory chain and, in turn, dysregulates memory processes. We also found that stimulation of mtCB1 receptors in astrocytes is responsible for the IP3-dependent Ca2+ increase in the mitochondria. This regulation was cell- type-specific since in similar experiments performed in primary hippocampal neurons we did not observe a modulation of the levels of mitochondrial Ca2+. To conclude, mtCB1 regulates Ca2+ levels in the mitochondria of astrocytes, that could be a novel mechanism by which astroglial CB1 receptors could control peripheral and central sensitization. In support of this claim, a growing body of literature indicates that the malfunction of mitochondria is among the cause of painful neuropathies. Therefore, mitochondrial cannabinoid system may represent a selective target to induce specific analgesia, thus reducing the broad action of this system at different cellular and subcellular levels.


Brain TRPV1 expression shifts from microglia to neurons in the course of NP states

In addition to CBRs, endocannabinoids also activate TRPV1. Peripheral activation of this receptor is involved in nociception, as it can produce either analgesia or pain, depending on the degree of TRPV1 desensitization. In the periphery, the activation of TRPV1 by NGF activated TrkA is an important point of NGF induced pain sensitization, thus providing a direct link between NGF and EC systems in pain. Importantly, TRPV1 are also expressed in the brain, where their functional role is less understood. Here we demonstrate that in the cingulate cortex of mice, as well as in other pain matrix brain areas, this channel is expressed and is functional mainly in the immunocompetent cells of the central nervous system, the microglia. Once activated, TRPV1 caused release of proinflammatory cytokines in parallel to a change of microglia morphology, induced microglial potassium currents, tuned the activity of inflammatory agents, i.e. the lipopolysaccharide and stimulated the shedding of microglia microvesicles which in turn increased the glutamatergic neurotransmission. A major finding of this study is that expression of TRPV1 shifts also to cortical pyramidal neurons of adult mice suffering from neuropathic pain. We provided evidence that, in mice with NP, TRPV1 activation in the anterior cingulate cortex (ACC) is responsible for the neuronal and cortical hyperexcitability, a distinct aspect of central sensitization. This new TRPV1 cellular distribution might be part of a homeostatic response that links physiology to pathology and is time-dependent from pain onset, being absent at early stages and gradually overspreading in all cortical layers of the contralateral hemisphere once the chronicization gets established.
Our background data suggested ECs to act as "signal output" effectors downstream of the neurotrophin systems in terms of enhancement of the endocannabinoid 2-AG production by NGF. Here, we demonstrated that the TRPV1/anandamide system interact with the TrkA/NGF signaling at the glial level in the cortex of mice. This evidence arises from the following results:
i) TRPV1 and TrkA are co-expressed in cortical microglial cells in resting conditions; ii) both capsaicin and NGF activate outward rectifying currents in microglia; iii) both capsaicin and NGF indirectly increase glutamatergic neurotransmission onto cortical pyramidal neurons by activation of microglia; iv) capsaicin and NGF mutually occlude their effects on neurotransmission and glial currents.
These multiple lines of evidence suggest that TRPV1 agonists and NGF (presumably by TrkA signaling) share the same mechanism. In addition, in AD11 anti-NGF mice, in which the NGF activity is chronically impaired, TRPV1 was expressed, in addition to microglial cells, also in thalamic neurons, and the capsaicin induced increase of excitatory neurotransmission was significantly higher than in wt. These data indicate that when the NGF system is altered, also the TRPV1 signaling is impaired.



Clinical biomarkers for NP

In pain, as in any other clinical field, there is a need to develop treatment strategies that allow stratification of therapies to optimize efficacy and minimize toxicity. Unfortunately, in the pain therapeutic space, a large number of studies have reported contradictory and inconclusive results. Thus, there exists an urgent need to establish new biomarkers that will select which patients will gain optimal benefit from the treatments, or those that will suffer from reduced side effects and monitor the response to the drugs. Up to now, stratification of patients upon pain symptoms evaluation is made difficult by the lack of adequate objective pain measurements. While efforts are made upon objective clinical pain measurements, any clinical trial to test drug efficacy on might pain provide clear criteria to classify patients.
The PAINCAGE project succeeded at subgrouping patients on the basis of individual measures (NGF levels) or genetic specific polymorphisms (EC SNPs) for improved patient stratification and drug design. With the objective of finding biomarkers and stratifying CP patients, ECs genetic polymorphisms and total NGF levels were measured in human samples of NP patients suffering from Multiple Sclerosis (MS). Measurement of NGF levels has been performed in CSF samples from MS patients with or without NP. We found that the total Nerve Growth Factor levels (mature NGF plus proNGF) are increased in the CSF of patients affected by MS with central NP, compared to MS patients without pain. CSF levels of NGF did not differ between patients with MS disease activity, compared to patients without it. Disease duration and EDSS (Expanded Disability Status Scale) did not differ between the two groups. This study revealed that the presence of central neuropathic pain is associated with higher CSF levels of total NGF (mature NGF plus proNGF) in newly diagnosed MS patients. This validates NGF as a biomarker for NP and confirms that the selective measures of mature NGF versus those of proNGF might provide additional stratification power of different pathological states.
Furthermore, the project explored the role of CB1R on the effects of anodal tDCS to treat chronic central neuropathic pain in patients with MS. Data showed that a genetic polymorphism of cannabinoid CB1 receptors influences the effects of transcranial direct current stimulation (tDCS) against central neuropathic pain in MS. The effectiveness of the treatment, however, showed inter-individual variability, raising the possibility that genetic variants affecting regulation of neuronal plasticity in response to non-invasive brain stimulation techniques may affect responsiveness to tDCS. tDCS is a noninvasive brain stimulation

technique that utilizes low amplitude direct currents applied via scalp electrodes to inject currents in the brain. It can produce analgesic effects on several painful conditions by modulating neuronal plasticity. The patients affected by MS with neuropathic pain, carrying the genetic variant CB1R, showed a reduction in the affective pain component after anodal tDCS.
Our results provide evidence for the first time that the CB1R genetic polymorphism can affect responsiveness to anodal tDCS on the affective dimension of pain and on anxiety in patients affected by MS with neuropathic pain. Moreover, data from the present work confirm the previous observation that anodal tDCS of the primary motor cortex ameliorates pain sensation in patients affected by MS with chronic, central, neuropathic pain. CB1R genetic polymorphism, implicated in the pathogenesis of the affective component of pain, may thus modulate tDCS effects on neuropathic pain. The identification of a SNP related to the development of NP in certain patients enable better patient stratification, and allow the designing of personalized treatments for high-risk patients carrying this SNP.
The identification of a SNP related to NP is a breakthrough in the pain field, enhancing the novelty of the PAINCAGE proposal.


NGF system and Osteoarthritis (OA): efficacy and safety issues


Efficacy issues related to NGF targeting therapies in OA
Osteoarthritis (OA) is a chronic, progressive disease of synovial joints that leads to pain and histological deterioration of cartilage and bone. The principal symptom of OA is joint pain. Anti-NGF antibodies provided a very strong clinical validation for the role of the NGF system in OA pain states, but safety concerns called for more preclinical research. Among PAINCAGE Aims, the identification and validation of NGF system therapeutic targets (NGF, TrkA, sortilin, p75NTR) for OA pain was investigated, in terms of comparative effectiveness and safety. To achieve this objective, we used the murine monoiodoacetate (MIA) model of osteoarthrosis (OA), that involves injections of MIA into the knee joint and induces rapid pain-like responses in the ipsilateral limb. The MIA model of arthropathy has proven useful for the pre-clinical profiling of putative osteoarthritis therapeutic candidates for pain and analgesia outcomes.
In a first study, OA was studied in TrkAP782S mice (TrkA Knock-in mice), which harbour a mutation in the TrkA NGF receptor gene that determines a reduced TrkA ubiquitination and an increase of NGF-mediated signaling. Ligand-dependent ubiquitination of TrkA is a key regulation point of NGF-TrkA signalling. To model human OA, the model of MIA-induced OA was further refined. The significant loss of structural architecture and fibroblastic replacement of cartilage, observed in response to standard intra-articular injection of 1.0 mg MIA, compared to 0.3 mg MIA, suggested that the lower dose of MIA provided a model of slow disease progression, with histological changes similar to those seen in human OA. The 0.3 mg MIA- induced OA model provides an experimental window that facilitates exploration of treatment-induced rapid progression of OA or its converse - inhibition of disease progression and joint repair. In our experimental conditions, MIA wild type mice developed allodynia (a painful sensation caused by innocuous stimuli) and exhibited i) significant cartilage degeneration scores, ii) increase in both number of CG PR positive neurons and activated neurons (p-ERK) iii) higher number of Iba-1 expressing microglia (immunocompetent cells of the central nervous system) in the lumbar ipsilateral dorsal horn of the spinal cord, when compared to Saline treated mice. When we induced osteoarthrosis in TrkAP782S mice (TrkA KI mice), which show a reduced TrkA ubiquitination and an increase of NGF-mediated signaling, we found that these mice were more sensitive to mechanical allodynia than wild type (WT) mice. MIA-induced allodynia developed faster in TrkA KI mice and also the on-going pain associated with MIA was exacerbated in these mice. 7 days after injection, MIA had no effect in the progression of knee pathology; however, MIA treated TrkA KI mice had a greater number of TrkA+ and CGRP+ neurons as well as activated neurons. Dorsal horn activation and microglial response were significantly higher in the spinal cord of TrkA KI mice. Moreover, TrkA KI ipsilateral knee exudates contained significant higher numbers of macrophages and mast cells close to CGRP fibers in synovia. The anti-NGF tanezumab prevents the development of allodynia in both TrkA KI and wt mice with OA. By using a mast cell line (RBL-2H3), we analyzed the NGF contribution to PGD2 (prostaglandin D2) production and observed that NGF potentiates IgE-induced PGD2 generation in a Cox-2 dependent manner. Thus, the ability of HQL-79, a PGD2 synthase inhibitor was tested to prevent MIA- induced hypersensitivity. Interestingly, at the lowest dose tested (3mg/kg) HQL-79 was more effective in TrkA KI mice. Microarray analyses followed by quantitative PCR on dorsal root ganglia (DRG) from saline- and MIA-injected WT and TrkA KI mice revealed significant differential gene expression between these

groups of mice. So far, we focused on 3 of them, USPI5, NDFP2 and Adra2b based on previous studies related with ubiquitination. The results indicated that the 3 genes showed increased expression in KI mice subjected to MIA injection. In addition, KI mice injected with saline showed reduced RNA amount of Ndfip2 and Adra2b compared with WT saline. The protein levels of different genes were also assessed in DRGs of TrkA KI mice and found that Adra2b levels besides increasing in response to MIA injection were regulated by both NGF and Nedd4-2, an ubiquitin ligase. In addition, we identified a regulation mechanism of TrkA levels that is dependent on the deubiquitinase USP36.
In another set of studies, we performed experiments to investigate key NGF system targets for OA amelioration and prevention. In particular we have tested the pharmacological effects of two anti-NGF antibodies (anti-NGF mAb alphaD11 and Tanezumab-like antibody) on nocifensive behavior in MIA mice. anti-NGF mAb alphaD11 (3.3 mg/Kg) i.p administered every 5 days starting one day prior to MIA injection, significantly prevented MIA-induced mechanical hypersensitivity. Likewise, a tanezumab-like antibody (5mg/kg s.c.) injected following the same scheme above, completely prevented the development of MIA-induced mechanical hypersensitivity and weight bearing changes. Regarding the efficacy of the anti-TrkA antibody MNACI3 (100 micrograms/mouse i.p.) in MIA-induced OA, a reduction in both weight bearing and hypersensitivity on the ipsilateral limb was observed when administered I0 days after MIA injection for 7 consecutive days.
Thus, the systemic administration of a TrkA antibody (MNAC13) reverted allodynia and weight bearing asymmetries in the MIA model of OA, providing the first evidence that targeting TrkA is as affective as anti NGF in this model.
Preventive treatment with the therapeutic candidate p75-Fc, a soluble form of the p75NTR neurotrophin receptor (LEVI-04), showed analgesia at higher doses of the compound (1 mg/kg), administered subcutaneously every five days from first injection.
The sortilin receptor appeared instead not to be relevant for OA. Indeed, sortilin-KO mice, did develop a similar allodynia as their littermates WT, thus suggesting that sortilin is not a drug target in OA mediated pain.


Safety issues related to NGF targeting therapies in OA

OA treatment aims to reduce symptoms and slow disease progression, thus impacting the patient's mobility and quality of life, reducing rescue analgesia use and delaying or preventing joint replacement surgery. Approved treatments include non-pharmacological therapies, analgesics to treat an acute flare, non-steroidal anti-inflammatory NSAIDs or intra-articular injection of corticosteroids for the treatment of OA pain, and slow acting drugs such as glucosamine and chondroitin, however the use of glucosamine and chondroitin is not recommended by the National Institute for Health and Care Excellence. The efficacy of NSAIDs, Cox-2 inhibitors and opioids for the treatment of OA pain is modest and all are associated with side-effects. There is thus a need for novel, safe therapies for OA pain and currently there are no treatment options to ameliorate OA progression.
Nerve growth factor is an important mediator of pain associated with OA. Monoclonal antibodies against NGF showed efficacy in treating chronic pain due to OA in phase III clinical trials. However, risk of rapid progression of OA (RPOA), which was exacerbated in the presence of NSAIDs and adverse effects on the sympathetic nervous system raise concerns regarding their safety. Nevertheless, inhibition of NGF function offers hope of treatment, if the unwanted adverse effects can be minimised. In PAINCAGE we investigated the safety of modulating NGF on rapid progression of OA (RPOA) and neuronal degeneration. The safety profile of the p75-Fc protein (LEVI-04), being developed by the SME Levicept, was compared to that of anti-NGF.
LEVI-04 supplements the endogenous circulating levels of p75NTR, a soluble neurotrophin (NT) binding protein, resulting in modulation of the excess NTs, including NGF present in OA. This modulation provides analgesia, and bone and cartilage repair in preclinical models of OA (see above). LEVI-04 is not an antibody and has a different antagonist profile compared with the anti-NGF antibodies. It inhibits the agonist function of the neurotrophins at the three tyrosine receptor kinase (Trk) receptors by binding all four neurotrophins, NGF, BDNF, NT-3 and NT-4 with nanomolar (nM) to picomolar (pM) affinity. It is an antagonist of neurotrophin-induced activation of tyrosine receptor kinase (Trk) receptors. Whereas anti-NGF antibody treatment requires plasma concentrations 1000-fold the primary affinity to demonstrate efficacy in clinical studies and preclinical models, LEVI-04 is effective in preclinical models of OA at concentrations similar to those required for inhibition of NGF activity. LEVI-04 has a more graded inhibition of NGF-induced activity compared with the steep all-or-nothing response seen with anti-NGF inhibition. LEVI-04 also inhibits NT-3- induced activity.
The PAINCAGE study demonstrated that, unlike anti-NGF antibodies, LEVI-04 does not induce RPOA in preclinical models of OA. Importantly, the exacerbation of RPOA with concomitant NSAID therapy seen in clinical studies with the anti-NGF antibody tanezumab was reproduced in the preclinical model, but no RPOA was seen with LEVI-04 in combination with NSAID treatment. Thus, MIA mice injected s.c. with anti-NGF or anti-NGF (5 mg/kg every five days) plus NSAID indomethacin (I mg/kg oral dose daily) underwent to a more rapid progression of histological changes compared with corresponding control
animals. MIA-injected knees from animals treated with control IgG1 showed mild chondrocyte degeneration. In contrast, animals treated with anti-NGF antibodies had significant degeneration of the cartilage and subchondral bone of the MIA-treated knee. Moreover, degeneration of cartilage and bone were further accelerated in combination with indomethacin. Although anti-NGF antibody treatment alone and in combination with indomethacin increased the rate of joint histopathology progression, both treatment regimens were analgesic. The observations of anti-NGF antibody-induced rapid progression of OA RPOA were further confirmed in a time course study. Animals treated with anti-NGF antibody (3 mg/kg SC every five days) developed significant OA-like pathology in the MIA-injected knee over the study duration. By Day 21 there was complete loss of the overall integrity of the knee, with associated loss of cartilage, subchondral bone and evidence of bone necrosis in animals treated with anti-NGF antibody.
The effect of p75NTR-Fc fusion protein was also assessed in a murine toxicology study. There was no difference in observations of spontaneous behaviour in the home cage, in the hand and in a novel arena between groups receiving LEVI-04 or anti-NGF or control. In a four 4 week toxicology safety studies investigating the effects of p75NTR-Fc fusion protein (LEVI-04) and anti-NGF following intravenous dosing up to 15 mg/kg every 7 days 4 weeks no adverse effects were noted with LEVI-04. However, a significant reduction in dorsal root ganglion greater than 20% was observed in rodents treated with anti-NGF, relative to corresponding control animals, at pharmacological relative doses to achieve analgesia in man. No shrinkage was observed in the rodents treated with LEVI-04. The toxicity of LEVI-04 was evaluated in pilot and formal toxicology studies. LEVI-04 was well tolerated in rats and cynomolgus monkeys at all doses tested.
In conclusion, LEVI-04 provides analgesia in preclinical OA and inflammatory pain models in a dose- dependent manner, but unlike anti-NGF antibodies, does not induce RPOA in preclinical models of OA. These preclinical findings support the investigation of LEVI-04 as a novel, disease-modifying drug for the treatment of OA. Based on these data collected in PAINCAGE, the p75NTR-Fc fusion protein LEVI-04 is currently being developed towards clinical studies in man for the treatment of OA. The first patient has been dosed in 3Q 2017.




Other pain models, pain pathway mapping and other therapeutic approaches


Neuropeptides in orthopedic pain

We used orthopedic surgery (tibial fracture) as an alternative model of pain, to gain deeper insights in changes in neuropeptides and brain-derived trophic factor expression and roles at hierarchical levels from DRGs to limbic systems, with particular interest on the association of these systems to pain-related amnesia. We showed that tibial fracture with pinning triggers cold allodynia and up-regulates nerve injury and inflammatory markers in dorsal root ganglia (DRGs) and spinal cord up to 2 weeks after intervention. At 72 h after surgery, there was an increase in activating transcription factor 3 (ATF3), the neuropeptides galanin and neuropeptide Y (NPY), brain-derived neurotrophic factor (BDNF), as well as neuroinflammatory markers including Iba1, glial fibrillary acidic protein (GFAP), and the fractalkine receptor CX3CR1 in DRGs. We observed similar, but more pronounced, changes in these markers in an established model of complete transection of the sciatic nerve. However, protein levels of BDNF remained elevated for a longer period after fracture. In the hippocampus, BDNF protein levels were increased, yet there were no changes in Bdnf mRNA in the parent granule cell bodies. Further, c-Fos was down-regulated in the hippocampus, together with a reduction in neurogenesis in the subgranular zone. These results suggest that attenuated BDNF release and signaling in the dentate gyrus may account for cognitive and mental deficits sometimes observed after surgery.


Pain pathways mapping

Another significant aim of PAINCAGE was the identification of neuronal substrates of pain sensation in medically relevant models using advanced mouse genetics and /imaging/ network tracing methods.
A proof of concept of a reporter technology for the analyses of activity dependent network modalities ("TRAP/ permanent access to transiently active neurons") has been obtained, a breakthrough in pain research, which has allowed mapping pain pathways under different pathological or pharmacological conditions related to the EC and NGF systems. In the specific, we undertook the analysis of pathways underpinning pain sensitization in models of acute neuropathic and inflammatory pain and found that a copious number of cells were activated in the dorsal horn of the spinal cord only in the long lasting inflammatory pain. Thus, these data show that TRAP models are enabling to discriminate neuronal pathway activation in models of pain that associate with long-lasting sensitization. We exploited the TRAP technology also to resolve the conundrum of CBR (or CB1R and CB2R) localization in the spinal cord. By injecting TRAP mice with a selective CB1R agonist the cellular distribution of activated cells was different respect to that observed with a CB1/CB2 agonist. Therefore, we have reason to believe that the above genetic models are sufficiently sensitive to map cannabinoid-sensing neuronal circuits.

We have also characterized Ca2+-sensor proteins that are sensitive to or mediate pain sensation focussing on NECABs1-3 and NECAB2. In these studies, Ca2+-sensor proteins are investigated as activation markers for pain pathways, and also as effectors of pain transmission. Briefly, we found that neuronal calcium-binding proteins 1/2 localized to dorsal root ganglia and excitatory spinal neurons were regulated by nerve injury. Notably, subsequent analyses on human and rat tissue samples revealed that the association of secretagogin (NECAB2) to excitatory neuronal circuits in both DRG and spinal cord was evolutionarily conserved across the mammalian species. By using NECAB2 KO mice we also established that this calcium-sensor protein is dispensable for the induction of NP and that these mice show accelerated recovery after inflammatory pain, thus showing that NECAB2 in excitatory circuits in DRGs and spinal cord play a role in the maintenance of inflammatory pain.


Neuronal circuitries underpinning the analgesic effect of electrical stimuli during electro-acupuncture

As an essential part of traditional Chinese medicine, acupuncture has been used to alleviate pain in clinical practice for centuries. With standardized electrical pulses, electroacupuncture (EA) is nowadays successfully used worldwide. Previous studies have shown solid evidence that EA can significantly increase the tail-flick latency (TFL) in rodents, but the underlying mechanisms are only partially elucidated. The tail-flick response is induced by noxious stimuli and widely used as a first line choice model to identify the mechanisms of nociceptive reflexes and their modulation. Notably, evidence in the literature shows the involvement of the nGF system in EA analgesia. Thus, here we determined whether the endocannabinoid system is involved in EA-induced antinociception. EA treatment reduced TFL in mice, and this effect was abolished by systemic administration of CB1R antagonists, indicating the involvement of CB1Rs in this effect of EA. We also determined the neuronal types expressing CB1Rs responsible for this function, by testing conditional mutant mice lacking CB1R in cortical glutamatergic neurons (Glu-CB1-KO), in forebrain GABAergic neurons (GABA-CB1-KO), or both (Glu/GABA-CB1-KO). Interestingly, the EA effect was abolished in Glu-CB1- KO, indicating that excess glutamate release (due to the lack of CB1Rs) might be the cause of this phenotype. To test this hypothesis, we used an NMDA receptor antagonist (MK801) at a dose that was unable to alter TFL in wild-type mice (partial blockade). This treatment restored the EA effect on TFL in full CB1R-KO and conditional Glu- CB1R-KO mice. Thus, CB1Rs expressed in cortical glutamatergic neurons mediate EA effect on TFL. Cortical glutamatergic neurons, in particular in the prelimbic cortex (PrL) project to the periaqueductal grey (PAG), a region important for pain perception and reflexes. Anterograde tracing experiments using biotinylated dextran amine (BDA) revealed that these projections contain CB1Rs, and local injection of the CB1R antagonist AM251 blocked EA-induced reduction of TFL in wild-type mice indicating that CB1R-dependent control of cortical glutamatergic transmission to the PAG mediates EA- induced antinociception.



Suffering without pain (Hereditary sensory and autonomic neuropathies, HSAN IV and V)

Human genetic studies further validated NGF as a target, showing that mutations in the genes for NGF receptor TrkA and for NGF itself cause congenital forms of insensitivity to pain (Hereditary Sensory Autonomic Neuropathy type IV, HSAN IV and HSAN type V, respectively). The opposite pathological condition of NP is pain insensitivity, a congenital form of which can be caused by mutations in the TrkA/NGF system genes or excess level of NGF. This occurs in patients affected by the genetic HSAN disease type IV and type V. The latter was found to be associated to one mutation related to NGF signaling, in the coding region of the NGF gene (Arg100Tryp, the human HSAN V NGF mutant NGFR100). Contrary to HSAN V, HSAN IV patients exhibit, besides pain insensitivity, a severe cognitive and adaptive behavioral impairment.
The R100W mutation, in the mature moiety of NGF, provokes a reduced processing of proNGF to mature NGF in cultured cells and inhibition of the secretion of mature NGF, with a higher percentage of neurotrophin secreted in the proNGF form. On the other hand, the R100 mutation selectively disrupts binding of hNGF to p75 receptor, while the affinity of NGF R100 for TrkA receptor is not affected.
It is important to understand whether the major impact of the mutation is on the biological function of proNGF or of mature NGF and to what extent the effects of the R100W mutation on the HSAN V clinical phenotype are developmental, or whether they reflect an impaired effectiveness of NGF to regulate and mediate nociceptive transmission in adult sensory neurons. Also, we need to clarify the structural underpinnings of NGF R100 differential properties and processing (see below). Finally, we sought to understand why the HSAN IV mutations in TrkA determine clinically a more severe neurological picture (with mental retardation), but an identical pain phenotype, with respect to HSAN V patients. To this aim, we developed and characterized mouse models for HSAN IV and HSAN V, to study the opposite process of Chronic Pain and to elucidate mechanisms related to cognitive, emotional and behavioral effects of pain. The knock-in mice harbour the HSAN IV and V mutations in the context of the human TrkA and NGF locus respectively, that have been substituted to their mouse counterpart.
Both murine models for HSAN IV (hTrkAR649W) and HSAN V (hNGFR100W) were characterized and validated. The HSAN IV mice show an impairment of thermal sensitivity and a tendency to attention and memory deficits. HSAN V mice mainly recapitulate the lack of pain perception and the absence of memory deficits reported in HSAN V patients, mimicking the clinical diseases symptoms. Altogether, these mouse models of pain insensitivity have proved invaluable to study the role of NGF signaling in pain transmission and perception. We also demonstrated that they are an exquisite and unique model to explore the emotional consequences of not feeling pain, from the behavioral and molecular point of view.
The development of HSAN mouse models allowed: i) understanding differences between HSAN IV and HSAN V clinical phenotypes, ii) studying the cognitive and emotional consequences of insensitivity to pain and iii) validating mechanisms related to NP. These HSAN murine model represent a major tool to study the emotional and affective aspects of pain syndromes.
A first manuscript describing the combined results obtained from the mice, and from the HSAN V patients, is about to be submitted.


Towards structure-derived new compounds to modulate the NGF system

In PAINCAGE we carried out a significant biophysical and structural activity, whose aim was to provide a structural basis for the design of a new generation of lead compounds to act on the NGF system. Also, we aimed at clarifying the structural underpinnings of NGF R100 differential properties.
First, during the Project we made a considerable progress into the biophysical characterization of complexes between two therapeutic monoclonal antibodies, the anti-TrkA MNAC13 Fab and the anti-NGF alfaD11 Fab and NGF and their corresponding TrkA and NGF antigens, respectively. The data collected provide the structural basis for the development of next generation new lead drug candidates, based on the carefully characterized antibody epitopes. Thus, the detailed structural information on NGF recognition by neutralizing anti-NGF antibodies will be of great aid in the development of new antagonists or agonists of neurotrophins.
Second, we provided a thorough biochemical and biophysical characterization of different variants of human NGF, inspired by the HSAN V NGF R100 mutation. The impact of the R100 mutation was investigated by Structure Activity Relationship (SAR) analysis, using a variety of human NGF variants, in which the R100 mutation was changed to different receiving residues and in the context of wild type hNGF, or of the P61S tagged variant of human NGF. This led to the identification of the hNGF P61S R100E molecule (also called painless NGF) as a new lead investigational compound, that holds therapeutic potential due to its unchanged neurotrophic potency, in face of a tenfold lower pain sensitization activity, with respect to human NGF. This new painless NGF recombinant protein, that received its full validation in the PAINCAGE project, will be tested in diabetic polyneuropathy, based on previous clinical studies with wild type human NGF in humans, where pain was the only side effect, limiting its therapeutic outcome. Painless NGF promises to be a molecule that fully exploits the broad neuroprotective therapeutic potential of human NGF. It is therefore of great interest to understand the mechanisms that underlie the reduced pain sensitizing activity of painless NGF and the reduced pain in HSAN V patients.
The position R100 in mature NGF was found to crucially affect the interactions of NGF with p75 receptors, as well as to be a key determinant for the interactions of mature NGF with its pro-domain in cis. This is relevant, because the HSAN V NGF is abnormally processed and leads to reduced levels of secreted mature NGF and increased amounts of proNGF.
In order to understand the impact of the R100W HSAN V NGF mutation, we undertook a structural study aimed at characterizing the structure of NGF in solution, and its interactions with the pro domain.
We have carried out a study to investigate the dynamic properties and flexibility of the long loops of NGF and the hitherto unknown N-terminus intrinsic conformational tendency of the unbound NGF molecule. Using a combination of nuclear magnetic resonance (NMR) and molecular dynamics (MD), we obtained, for the first time, the solution structure of the NGF homodimer. This was a remarkable achievement because only crystallographic structural data were so far available for NGF, which are not suited to extract the dynamical properties and to map interactions. We achieved a full assignment of the spectrum, and conversion of the NOE information into a structural model led eventually to a well compact bundle with a root mean square deviation (r.m.s.d.) of 1.3 Angstrom from the structure with minimal global energy as calculated on 236 residues. The structure closely resembles the available X-ray structures, especially in the Cys-knot, while exhibiting a much larger variability in the loop regions. The structure in solution allowed us to address a number of crucial aspects.
We described the N-terminus intrinsic conformational preferences of unbound NGF in solution. We showed that also in the absence of partners the previously unexplored NGF N-terminus has a strong tendency to fold into a helix, challenging the view that this region is unstructured. We capitalized on the structure of NGF in solution to identify new lead drug candidates (in the peptide form) for interfering with the NGF system. The structural information of NGF in solution, together with molecular dynamics simulations, was also used to gain insights into the structural impact of the NGF R100 mutations.
Our study also sets a definitive word on the structural plasticity of NGF loops II and V and provides a structural explanation for the large differential affinity of the alfaD11 anti-NGF therapeutic antibody for NGF vs proNGF. The study unveils the conformational versatility of the relatively rigid NGF loops, upon functional ligand binding and fills a gap in our structural understanding of NGF inter- and intra-molecular interactions, providing a strong basis for the design of more selective NGF antagonists.
Finally, we undertook a study on the biophysical and structural characterization of proNGF, whose structure is only known at a low resolution. We have exploited a new strategy based on nuclear magnetic resonance and modelling validated by X-ray small angle scattering to obtain novel insights on the pro-domain, both in isolation and in the context of proNGF. We showed that the isolated pro-domain is intrinsically unstructured and is able to interact with its mature counterpart, both in cis and in trans. We also demonstrate that the isolated pro-domain has per se the ability to induce growth cone collapse, indicating its functional independence. Our data represent an important step towards the structural and functional characterization of the properties of proNGF and the NGF pro-domain. This will allow clarifying at the structural level the reasons for the altered proNGF processing in the proNGF HSAN V R100W mutants. We are currently describing these results in a manuscript intended for the journal Structure.


Towards small molecule agonists of NGF: structure-inspired N-terminal peptides

The long term objective of the structural WP7 in PAINCAGE was to exploit structural information to develop new NGF agonists or antagonists.
NGF N-terminus (residues 1-14) plays a key role in the interaction with TrkA. An N-terminal truncated form of NGF has a significant drop in affinity for TrkA and in the ability of eliciting TrkA phosphorylation (Kahle et al., 1992; Hughes et al., 2001) which could not be ascribed to differences in folding or stability (Woo et
al., 1995).
In the hNGF/TrkA complex (Wehrman et al., 2007) and in a complex with a DNA aptamer (Jarvis et al.,
2015), the N-terminus is folded in an α-helix. However, the generally accepted view is that the N-terminus is disordered in the NGF unbound state (Settanni et al., 2003; Berrera et al., 2006).
Using a combination of NMR and MD simulations, in PAINCAGE we unveiled the N-terminus intrinsic conformational preferences of NGF in solution (Paoletti et al., 2016). We showed that even in the absence of the TrkA partner, the NGF N-terminus has a strong tendency to fold into a helix, challenging the current view that this region is unstructured.
These findings provided the structural rationale for the design of biologically active synthetic peptides endowed with agonist or antagonist NGF activity, based on the NGF N-terminal residues that could exert: (i) therapeutic beneficial effects; (ii) suitable pharmacokinetics and CNS penetration for drug development, without unwanted systemic effects by the full-length protein.
We designed and synthesized a number of chemically modified NGF N-terminal peptides, and tested their biological activity with a quantitative in vitro proliferation bio-assay on TF1 erythroleukemic cells, (Malerba et al. 2015). All the peptides are active and able to induce proliferation in TF1 cells All the peptides are quite similar in C50, and exhibit about only 50 folds less potency with respect to WT human NGF. To identify the shortest pharmacophore peptide that still preserves an acceptable bioactivity, the synthesis of shorter peptides, designed by trimming the residues at the C-term of the above four peptides, has been performed and their biological activity in NGF bioassay confirmed.
Extracting the essential pharmacophore features should pave the way to a virtual screening of chemical libraries resulting in the identification of small molecules endowed with agonist or antagonist NGF activity.


Conclusion

The results achieved by PAINCAGE are highly relevant towards the development of a new generation of painkiller analgesic drugs based on these target systems. The strength of the Consortium and of the results obtained has been the combination of: i) a strong focus on basic mechanisms of pain transmission and modulation, as a discovery engine for new targets and new mechanisms, ii) a unique platform of mouse models in which pain is positively or negatively modulated, at different levels and by different mechanisms, iii) a pipeline of identified therapeutic candidates (recombinant proteins or small molecules), iv) a focus on new technologies for the mapping of pain pathways, v) a diverse expertise and experimental approaches. Based on these strength points, PAINCAGE generated results that will provide a robust and solid basis to accelerate the development of already identified second-generation therapeutics, based on the "NGF target" system, as well as for the identification and validation of new druggable targets emerging from the elucidated mechanisms.
PAINCAGE has also identified clinical biomarkers for NP, validated in animal models and clinical samples, that could result in future clinical benefits, for the stratification of patients suffering from different neuropathies.
Pain starts at the periphery, but involves in a crucial way many central aspects. A fundamental point of the results obtained by PAINCAGE dwells with central aspects modulated by pain states. For instance, studying the emotional consequences of pain insensitivity, in HSAN V mice, has taught us the consequences of
growing without feeling pain, the converse of the maladaptive plasticity processes that we have also studied
in other mouse models, suffering from chronic NP. In addition, by using an animal model in which cannabinoid receptors are absent selectively in specific neuronal subtype and only in the forebrain, we learned that these receptors are crucial for the encoding of pain signals. These are crucial questions in basic neuroscience that PAINCAGE has set the grounds to approach in an original and comprehensive manner.
Potential Impact:
Main outputs of the Project
The main outputs of the project have been:
i) elucidating new mechanisms concerning neurotrophin and EC signalling and actions, relevant for Chronic Pain, at the peripheral and the central levels.
ii) new mouse models for CP-related mechanisms and validation of new technologies for the mapping of pain pathways (TRAP technology)
iii) validation of targets identified within the project for chronic pain (sortilin, p75NTR, brain TRPV1, mitochondrial CB1Rs and brain CB2R, that will provide the basis for specific drug development programs, tailored to these targets.
iv) advancing lead compounds (p75-Fc (LEVI-04)) to clinical testing in humans for CP indications, and providing new strategic data for proteins that are already in clinical trial (the humanized version of the mouse monoclonal anti-TrkA mAb MNAC13 used in this study).
v) new lead compounds based on all above points (structure-based NGF small molecule agonists).
vi) identification of new targets, from transcriptomic and epigenomic studies, that will provide the basis for new target validation and drug discovery programs

Impact on European research

This rich output of results will have an impact on the European economy and society at several different levels.
To fully exploit the huge therapeutic potential of NGF and EC systems in pain, we built a consortium of leading researchers in the field, from 6 European Countries, including UK. The output of the PAINCAGE project was a success, in terms of scientific results and industrially relevant results achieved. The main objectives of the collaborative project have been reached, as testified by the scientific publications during the project, and by the rich pipeline of manuscripts in preparation, capitalizing on the results obtained in the final part of the project. The results provide a huge contribution to the basic understanding of the role of the NGF and EC systems in NP. The consortium was also diverse in terms of expertise and main fields of interest, and the project was a success in fostering the collaborations and strengthening the bonds between the different partners, including two SMEs. First of all, we wish to emphasize that throughout the project we improved the collaborations between partners, and build new ones, involving researcher mobility and upgrade of skills within the EU Community. Despite the fact that the publication output is already very rich and already involves joint authorships between partners, a rich pipeline of joint papers, concerning the main outputs of the project, is being prepared and submitted. Also, a pipeline of new patent applications is foreseen, to capitalize on the results obtained in PAINCAGE. This consortium is now a well established network, that will have an impact in present and future European research in the fields of pain, of neurotrophins and of endocannabinoids.
The successful outcome of the project has been ensured by the clear objectives and the strength of the network, that is highly interdisciplinary, diverse and well integrated, both academic and SME involved. This interdisciplinary approach allowed us understanding the specific molecular mechanisms and cellular interactions linked to the regulation of pain processing at different levels of the nociceptive pathways triggered by NGF and dampened by the EC systems.
Within the neurotrophin system, the TrkA and Sortilin receptors have been validated as targets for NP and antibodies or small molecules targeting these receptors have been characterized.
Concerning the EC system, the PAINCAGE findings identify brain TRPV1 as a detector of harmful stimuli and a key player of microglia-to-neuron communication. PAINCAGE has exploited a large diversity of experimental strategies, with a strong focus on new technologies, which has allowed mapping pain pathways under different pathological or pharmacological conditions related to the EC and NGF systems and their interplay. It is noteworthy that a proof of concept of the non pharmacological therapy such as Electroacupuncture was obtained in this project, with the demonstration of the central involvement of the EC system (in addition to the demonstration (from the literature) of the NGF system). PAINCAGE also derived new mouse models, in which the NGF and the EC pain pathways have been specifically disrupted at different points and in different cells (CB1 receptor specific ko mice), or in which pathology-relevant (HSAN IV or HSAN V mice), or functionally relevant (TrkA ubiquitination defective mutant) single residue mutations have been introduced. All these advancements have been reached exclusively by European labs and companies, with great potentials of economic fall-out. The project has laid strong grounds for a new generation of analgesic drugs of new conception.

Economic Impact
OA is a clinical situation, in the field of chronic pain, for which the NGF system has a strong rationale, and many efforts are being pursued in OA with anti NGF antibodies. A current conservative estimate is about 42m OA patients in EU. There will be a likely 6-fold increase in knee replacement surgery for OA patients within 2030. This suggests that a proper drug pricing is crucial to make a new compound cost-effective. Whilst the current NSAIDs / COXD2s set a relatively low price benchmark for one-year treatment (in the range 100-1000 Euro), and the opioids are in the range 1000-4000 Euro/year, the costs associated with knee surgery can be much more significant, about 40000 Euro. Currently marketed rheumatoid arthritis biologics have higher cost in the range 15000-30000 Euro/year, comparable to surgery. NSAIDs and COXD2
Inhibitors generates about 1050m Euro revenues in EU in 2015 for the treatment of OA. About 65m patients use NSAIDs every year in EU, with Cox-2 inhibitors prescriptions being about 10% of NSAIDs. Cox-2 inhibitors sales were > 7bn $ in US in 2003 but then gradually dropped up to 1bn $ in 2014 due to safety concerns, indicating an unmet need in this area.
While European payers (both governments and citizens) are expected to maintain tight constraints on drug budgets, on the contrary emerging markets (India, China,...) see a continuous growth, though slowed down by the global economic framework. In particular, there is a huge potential market for anti-NGF, anti-TrkA or p75-NTR targeting biologics for OA, as witnessed by the pipelines and trials of big pharma companies. There are a number of anti-NGF antibodies in advanced clinical trials. Fulranumab, an antibody currently in phase III trial for OA, has expected revenues exceeding1 billion Euro, to be used either as an adjunctive therapy or a monotherapy for OA. Sales forecast for Tanezumab, one of key product of Pfizer portfolio also in Phase III, is really promising and the forecast by Pfizer increases from 20-100m Euro in 2018 to 100-800m Euro in 2022. A conservative EU target population estimate would be around 1-2.5 million potential OA patients.
Concerning anti-TrkA antagonists, there are compounds in Phase I-II for OA and chronic pain. Also the MNAC13 murine monoclonal antibody investigated in this project, in its humanized version GBR-900, is currently in Phase I trial for chronic pain by Glenmark Pharmaceuticals. The anti-NGF biologic therapy is going to be an effective alternative both to traditional treatments for OA, including surgery, and specially to opioids for the well known side-effect and heavy addiction problems. Furthermore, it is easy to prescribe an injection every two months, the typical dosing of biologics. In this respect, the long lasting analgesic activity discovered and characterized in this project is likely to determine a terrific competitive advantage for this type of analgesic drug. This will also contribute to reduce the cost of the treatment, thus obviating to the otherwise higher cost of the biologic drug with respect to a small molecule drug. If a product (such as for instance the p75-Fc LEVI-04) achieved a penetration of 10% in the population, it could in theory generate large peak sales of 4bn - 9bn Euro, assuming an annual cost of treatment of 15000 Euro. Estimating, from Pfizer figures, about 95000 employees for 47bn Euro of revenues, 4bn - 9bn Euro may correspond 8000-18000 new workplaces within few years.

Clinical impact
The Management goals when treating pain are to achieve maximal reduction in pain intensity as rapidly as possible and lead to long term pain relief, to restore the individual's ability to function in everyday activities, to help the patient cope with residual pain, to assess for side effects of therapy, and to facilitate the patients passage through the legal and socioeconomic impediments to recovery. Also, for some people the goal is to use non-pharmacological therapies to manage the pain. Analgesic choice is also determined by the type of pain: for neuropathic pain (NP), traditional analgesics are less effective, and there is often benefit from classes of drugs that are not normally considered analgesics, such as tricyclic antidepressants and anticonvulsants with collateral effects.
The main expected impact of our project is the development of novel therapies for NP that modulate the neurotrophin and the EC pathways. PAINCAGE project contributed significantly to the development of innovative therapies, for controlling NP and OA onset, progression and perception by pharmacological treatment with anti-NGF, anti-TrkA and p75-Fc (NGF scavenger), as well as painless NGFR100.
A recent study on cost-effectiveness of anti-NGF in OA [Losina et al, 2016], a careful modelling approach shows the biologic drug Tanezumab could significantly improve symptoms and reduce the utilization of knee surgery (unit cost around 50000$ and performed over 600,000 times/year only in the US), delaying the age of surgery of >3 years, further reducing the much worse and less successful revision surgery. Rapid progression of OA remains a key concern surrounding NGF inhibitors, (raised in clinical trials especially in the combination with NSAIDs), nevertheless the authors show that, even in the case of an accelerated OA progression rate beyond the range observed in large Tanezumab trials, there is essentially no decrease in the Quality Adjusted Life Years (QALYs). Thus anti-NGF biologics remain a very promising treatment for OA, alone or in combination with other standard treatments, provided a proper selection of patient and drug pricing, that should make them a more cost-effective than surgical replacement therapy. In a reverse translational study in this project, we were able to reproduce in mice treated with anti NGF (and NSAIDS) the RPOA observed in patients, also showing that the 75-Fc dose does not show such a safety issue.
The knowledge gained on the interaction between the NGF and the EC system provided a proof-of-concept for the use of combined NGF- and EC-targeting approaches, that may allow even better treatment or prevention opportunities, including the non pharmacological Electroacupuncture.
Concerning the biomarkers identification for NP patients stratification, PAINCAGE has identified biomarkers for NP, validated in animal models and clinical samples, that could result in future clinical benefits, for the stratification of patients suffering from different neuropathies. Thus, there exists an urgent need to establish biomarkers that help to select which patients might gain optimal benefit from the treatments, or those that will suffer from reduced side effects.
The Consortium discovered the following NGF and EC-related biomarkers:
i) NGF levels are higher in the CSF of MS patients that experience NP than in those that do not, validating NGF as a biomarker for NP sensitivity, allowing stratification of patients with MS based on pain.
ii) A genetic polymorphism of EC CB1Rs influences the effects of tDCS against central NP in MS patients. The obtained data provided further perspective to translate pre-clinical and clinical results into solutions for the benefit of the patients.
To gain knowledge on the mechanisms of different pain syndromes whose treatments are inadequate, such as NP, the new genetic pain models generated in this project, of pain insensitivity (HSAN) and pain hyper- sensitivity (TrkA knock-in mice), provided unique insights in understanding pain processing and perception, including its cognitive and emotional aspects. The emotional consequences of pain insensitivity, in HSAN V mice, are showing the consequences of growing without feeling pain, the converse of the maladaptive plasticity processes studied in other mouse models. In addition, by using an animal model in which cannabinoid receptors are absent selectively in specific neuronal subtype and only in the forebrain, we learned that these receptors are crucial for the encoding of pain signals.
NGF is also known to modulate the TRPV1 channel (transient receptor potential, vanilloid subfamily type 1) a molecular integrator of painful stimuli. Molecules that modulate TRPV1 activity are currently evaluated as pain treatment. TRPV1 is also endogenously activated by EC. Our background data show that NGF regulates EC signalling, underlying a novel NGF/ECs interplay amenable for drug targeting. PAINCAGE findings identify brain TRPV1 as potential detector of harmful/in flammatory stimuli and showed that TRPV1- mediated microglia-to-neuron communication is down stream regulated by TrkA receptors, suggesting NGF as mediator of microglia TRPV1 signalling. In the healthy brain, TRPV1 can function as detector of inflammatory stimuli and have a key role in the communication between immunocompetent cell and neurons. Upon an injury, this channel receptor can behave as a brain biomarker of inflammation. This lays solid foundations for future investigation on a central-neuron selective analgesia by exploiting the TRPV1 brain expression in neurons involved in the supraspinal nociceptive pathway. Indeed, targeting specific neurons by stimulating TRPV1 will result in a pain pathway-specific local analgesia.
Transcriptomic and epigenomic analyses uncovering common- and specific- pathways involved in this long lasting analgesia, identified new candidates that might lead to a new generation of epigenetic-based analgesic drugs. The TrkA and Sortilin receptors have been validated as targets for NP and antibodies (or small molecules) targeting these receptors have been characterized.
This might improve the efficacy, reducing the drug doses and minimizing the side effects. Furthermore, the investigation of epigenetic changes (DNA methylation and histone methylation/acetylation) mediated by the NGF-related antibodies, will allow a more careful design of epigenetic modulators of N, which are currently acting only at a very broad-spectrum basis, likely prone to unpredictable side-effects.

Impact on public health system

The long-lasting analgesic effect of alfaD11 anti-NGF and MNAC13 anti-TrkA antibodies (> 3 months since last dosing for anti-NGF and > 4.5 months for anti-TrkA ), and demonstrated in this project, goes well beyond the typical injections every 8 weeks prescribed for biologics. A future clinical use of these compounds will provide a significant cost-saving for public health, about 30-50% compared to standard biologics. The effectiveness of these new treatments, if provided at a reasonable price, have the potential to reduce the use of surgery in OA, thus providing an addition cost reduction. This is a key advantage, since there is a serious concern about hospital-acquired post-surgery bacterial infections, which are often intractable by antibiotics and represent a heavy socio-economic burden.

Social Impact

Neuropathic pain represents a major health problem and a sizeable economic and social burden, affecting 10% of the population, with total costs (health care and loss of work productivity) estimated to be around € 500 billion in EU annually. Pain may influence mood, sleep, activities of daily life and social activities, thus reducing the patient quality of life (QoL), increasing disability and worsens self-reported mental health status. Damage to different components of the peripheral nervous system (PNS) regardless of the etiology, may cause NP that results in disability, reduced participation in daily life activities, impaired QoL, and may limit the efficacy of neurorehabilitation procedures. Sex, gender, psycho-social variables, anthropological and cultural features may influence pain expression, and its pharmacological and non-pharmacological outcome, but the role of these factors has not been consistently explored in neurorehabilitation. There is a number of psychological factors that can be correlated with or represent a predictor for pain, or may influence the treatment and outcome of neurorehabilitation programs. All these factors should be considered when designing these programs, and future studies should incorporate them as potential covariates that may influence outcome.
The peripheral neuropathy might present of a "double disability" resulting from pain itself and the underlying disease, where the specific impact of the two conditions on physical and psychosocial function cannot be separated. The large number of underlying etiologies, the variety of clinical features and neurophysiological changes, as well as the biopsychosocial variables, may cause a delay in diagnosing pain in PNS diseases, thus facilitating pain chronicisation, which in turn may reduce the therapeutic response.
The recent Neuropathic Pain Society International Group (NeuPSIG) guideline, which addresses some important methodological issues, indicate that tricyclic antidepressants, serotonin norepinephrine re uptake inhibitors, a2-o ligands (i.e. pregabalin and gabapentin) should be first-line treatment, 5% lidocaine patch, 8% capsaicin patch, and tramadol represent second-line, and strong opioids and botulinum neurotoxin should be used as third-line for NP, with lidocaine, 8% capsaicin patch and botulinum neurotoxin type A to be used for peripheral NP only. The choice among first line drugs should take into account comorbidities, and regular follow-up is needed to monitor response to treatment and adverse events. Because of the large number of non-responders to drugs for NP, including first-line ones, some drug combinations can be considered.
What is the impact of pain treatment on functional recovery and neurorehabilitation outcome in patients with plexopathy, radiculopathy, mononeuropathy, peripheral neuropathy, and phantom limb pain?
Pain is a disabling symptom and is often the foremost symptom of conditions for which patients undergo neurorehabilitation. This is further confirmed by the International Classification of Functioning, Disability and Health (ICF) that includes pain in some categories as to Body Functions. Data on the influence of pain treatment on functional recovery and rehabilitation outcome in these conditions are scanty. NP worsens gait parameters and increases the risk of falling in patients with polyneuropathy, but whether treating pain may significantly ameliorate walking abilities is unclear. Pain is one of the main symptoms to be addressed with a multidimensional approach in the rehabilitation of brachial plexus lesions, because it may influence the rehabilitation outcome. It is hypothesized that effective treatment of pain may have benefits for neurorehabilitation outcomes, but future studies should further address this point.
It is therefore evident that the social impact of the PAINCAGE project is extremely high. The development of a new generation of analgesic drugs, based on the improved understanding of the NGF and EC involvement in NP, represents a major outcome of this project to control NP, will therefore improve the quality of life of millions of patients across Europe and improve the neurorehabilitation outcomes.
Most current therapeutic approaches are associated with unwanted effects and unsatisfactory pain relief, for this reason a number of clinical trials testing monoclonal antibodies acting on NGF system are ongoing. Thus the new results issued from this project will improve the combined NGF- and endocannabinoid-targeting approaches, addressing the two systems that represent the cutting edge of current pharmacological approaches.
The information gathered in this project and the fulfillment of its objectives provides major advances in understanding and controlling the mechanisms of Neuropathic Pain, laying the ground for the development of next-generation NGF-targeting drugs as effective NP treatments.
Also, the project provides information to achieve better patient stratification, according to biomarkers related to the mechanisms studied. Up to now, stratification of patients upon pain symptom evaluation was made difficult by the lack of adequate objective pain measurements. Visual Analog Scale (VAS, Wewers & Lowe, 1990) and Numeric Rating Scale (NRS, reviewed in Ferreira-Valente et al., 2011) are well credited scales to quantify pain, but they are not an objective scales, being pain perception qualified on personal experience. Furthermore, such a subjective measurement is useless in patients with cognitive impairment. While efforts are made upon objective clinical pain measurements, any clinical trial to test drug efficacy on pain should provide clear criteria to classify patients. According to our data, patients with Multiple Sclerosis might be phenotyped, for their pain sensitivity, on the basis the Nerve growth factor levels (mature NGF plus proNGF) which have been shown to be higher in the CSF of patients with multiple sclerosis AND central neuropathic pain with respect to those without neuropathic pain. This validates NGF as a biomarker for neuropathic pain and suggests that the selective measures of mature NGF versus those of proNGF could provide additional stratification power of different pathological states.
A new long lasting analgesic mechanism has been discovered for two anti-NGF and anti-TrkA antibodies of therapeutic interest: their analgesic actions in a NP model were found to outlast the last dosing by more than three months. The project provides a unique and unprecedented opportunity and summarizes the best that the industrial pipeline displays currently, world-wide. In conclusion, the information provided by the PAINCAGE project laid solid grounds for the development of safer and more effective new generation analgesic drugs. This will have a major social benefit, providing solutions to a big social and personal problem, namely chronic pain.
The results obtained strongly impact quality of life of patients suffering from NP, with great reduction of social costs due to inadequate pain treatment. The long lasting analgesic effects of this class of drugs will reduce daily dose intake, with welfare benefit and better compliance of patients to treatment.


Non-pharmacological analgesic therapies

The excessive use of analgesic drugs is never a good therapeutic option for a patient and sometimes they cannot be prescribed. There is knowledge available on acupuncture used clinically to treat acute and chronic pain conditions, but the mechanisms underlying its effect are not fully understood. Concerning the EC system, PAINCAGE scientists have found that NP decreases cortical synaptic plasticity, via a reduction of EC CBR1 receptor expression, necessary for the analgesic effect of electroacupunture, leading new opportunities for non pharmacological analgesic drugs to be used, when and if indicated. Furthermore, the Consortium has generated useful tools and models for the study of inducible systems, whose dysfunction in pain and recovery upon electroacupuncture can be a reliable measure of the degree of recovery experienced. They will be helpful in understanding mechanisms-based non pharmacological approach.

Exploitation of results

It is a priority of the PAINCAGE Project to promote in any possible way the exploitation of the results obtained by each of the Partners. The PAINCAGE Project has been very active in pursuing the exploitation of the results obtained in the project and efforts of favoring and accelerating the industrial exploitation of results of the project have been a constant priority of the paincage project. These efforts were based on the Exploitation Plan that was discussed by the Partners from the very beginning of the Project, and constantly confirmed and updated with new opportunities, in the context of the General Assembly and of the Steering Committee, and also, specifically, in the context of the Industrial Property Committee. This IPC was formed by the Coordinator and by the partners Malcangio and Westbrook, and appointed Simon Westbrook (Levicept) as the Exploitation Manager. In the final reporting Period, the Coordinator took a direct responsibility in pursuing the exploitation of the project results, on behalf of the Project, identifying the areas of major impact of the project and starting to present them to potentially interested industrial partners. It was decided that the preparation of the Final Report, and in particular of the sections 4.1 and 4.2 would give the best opportunity to focus on the identification of the knowledge gained in the Project, that could be the subject matter of IP protection, of clinical use or of dissemination based on contemplated publications and activity reports. Thus, as discussed in this section “Potential impact and main dissemination activities and exploitation results”, Paincage has achieved very significant results on improving our knowledge on NP and on the involvement of NGF and EC systems in NP, that are going to have a profound impact on research, on economy, on clinical practice, on the public health systems and on the clinical practices and finally on the society. The results obtained provide a very solid knowledge basis that will be strengthened by a rich pipeline of new IP (see below), that will ensure that the results of Paincage will be exploitable not only in terms of increase of scientific knowledge but also in terms of leveraging this scientific knowledge and innovation to explore opportunities for industrial pharmaceutical developments (both therapeutic, diagnostic and patient stratification) and for the improvement of clinical practices. Thus, Paincage explored in depth the emerging relationships between the NGF and the EC systems, particularly in the regulation of NP, based on interactions at many different intracellular and intercellular levels. Thus, the NGF and the EC system comprise many targets of potential therapeutic interest that need, however, to be carefully understood and validated. A big step in this direction has been made by PAINCAGE. The exploitation of this target system for drug development is emerging as an exciting opportunity to generate a new class of painkillers. Therapeutic anti-NGF antibodies have been clinically tested in patients, demonstrating a remarkably effective analgesia in patients suffering from OA and low back pain. However, safety concerns, stemming from those trials, need to be addressed and understood, to fully exploit the therapeutic analgesic potential of this target. The identification and validation provided by the project of new targets for chronic pain (sortilin, p75NTR, brain TRPV1, mitochondrial CB1Rs and brain CB2R), provide a basis for specific drug development programs, tailored to these targets.
Advancing lead compounds (p75-Fc (LEVI-04)) to clinical testing in humans for CP indications, provides new strategic data for proteins already in clinical trial (the humanized version of the mouse monoclonal anti-TrkA mAb MNAC13 used in this study).
The data gathered in Paincage have been instrumental in advancing these programs into the clinics, in identifying new biomarkers, in validating new targets, including targets involved in the long term epigenetic regulation of pain responses at the level of DRG neurons and nociceptors and in generating new lead compounds for the treatment of NP. Moreover, new technologies for the mapping of pain pathways and new mouse models harbouring pain related mutations have been established in the Consortium. To leverage these scientific findings, a pipeline of new patent applications is being generated by different Paincage partners, to exploit the results obtained within the project.
Once filed, the planned patent applications will be offered for licensing to pharmaceutical companies developing pain killers acting through the NGF and EC system, as appropriate.
To this aim, during the third reporting period the Coordinator started consultations and presentations of Paincage results to different pharmaceutical industries, with the aim of exploring their interest to enter into collaborations and agreements.
Moreover, Levicept has a strong patent position and exploitation plan for the industrial development of LEVI-04 (p75-Fc NGF scavenger), a molecule that entered Phase I clinical trials also leveraging data collected in Paincage.
In addition to the industrial exploitation of results, the project activities performed during the third reporting period led to identifying a number of results that can lead to improved clinical practices. These include biomarkers or genetic SNPs according to which patients affected by MS could be stratified, according to their pain sensitivity and non-pharmacological therapeutic approaches to the treatment of pain, that have received in Paincage a strong mechanistic validation.
The exploitation plan foresees that dissemination in the scientific peer reviewed literature plays an important role to facilitate the commercial and industrial exploitation. To this aim, in addition to the long pipeline of papers published and in preparation, after the special issue in the European Journal of Neuroscience on the Paincage topics, published around mid-term Project, a new special issue will be prepared for publication in Frontiers in Neurology, upon an invitation by the Editor, dedicated to the PAINCAGE main results and contribution to the EU community.

List of Websites:
WEB Site Address: www.paincage.it
Since January 2015 the PAINCAGE website is an important dissemination channel describing project activities and outcomes, such as latest news, articles, presentations, quarterly periodic reports, meeting's agenda and minutes, templates, reporting instructions and media repository and the project's internal documents. The website will be maintained at least 5 years after the project ends, even if with a little transition period (few days) due to the change of the provider just at the end of the project period.
Particular attention in creating the website was given to the Communications Best practices (see at http://cordis.europa.eu/fp7/ict/participating/communicationbest-practices_en.html#website) of the EC FP7 Project. We also took care to clearly acknowledge the EU as a source of funding and to use the EU emblem (i.e. the blue EU flag).

The statistics of the PAINCAGE web site follow:

Traffic from 1st Jan 2015 to 31st Dec 2017: Pageviews: 3143; Days in Range: 630; Average Daily
Pageviews: 6; From Search Results: 673; Unique IPs: 750

Audience Overview from 1st Jan 2015 to 31st Dec 2017: Visit: 1058; Unique IPs: 750; Bounce rate: 15.21; New Visitors: 482; Pageviews per visit: 3.12; Longest visit: 38 hits

Top 10 countries accessing the PAINCAGE web site, from 1st Jan 2015 to 31st Dec 2017:
Italy: 47.38%; Spain: 10.08%; United States: 7.44%; United Kingdom: 5.86%; Germany: 4.02%; Austria:
3.92%; France: 2.96%; Sweden: 2.17%; Denmark: 1.65%; Belgium: 1.32% Relevant Contact Details:
Three email accounts were available during the life time of the project:
for general information: info@paincage.it
to contact all consortium members: consortium@paincage.it to contact the Project Coordinator: paincage@sns.it
In addition email addresses of the Partners Principal Investigators are available on the beneficiaries pages on the PAINCAGE website under the Consortium section.
At the end of the project the website has been updated with an email address still available dedicated to the Project, that is paincage@sns.it.