THERAPEUTIC EFFECT OF HUMAN ADIPOSE-DERIVED STEM CELLS AND THEIR SECRETOME IN EXPERIMENTAL DIABETES: FOCUS ON NEUROPATHIC PAIN

Diabetes mellitus is one of the most common and serious chronic disease in the world. Although the number of available agents to manage diabetes continues to rapidly expand, treatment of diabetes complications, such as neuropathy that is one of the most frequent complication of diabetes mellitus, remains a substantial challenge [Aring et al., 2005]. Pathophysiology of diabetic neuropathy is complex and not fully elucidated; it has multipathogenic mechanisms that cause a diversity of physical symptoms: allodynia, hyperalgesia, numbness and cutaneous ulceration [Vinik et al., 1995]. Persistent Neuropathic Pain (NP) interferes significantly with quality of life, impairing sleep, and emotional well-being, and is a significant causative factor for anxiety, loss of sleep, and non-compliance with treatment. Recent advances in the mechanisms involved in NP have demonstrated that pro- and anti-inflammatory cytokines produced by immune cells as well as by glia and microglia in nerve, dorsal root ganglia (DRG) and spinal cord are common denominators in neuropathic pain [Sacerdote et al., 2013; Old et al., 2015]. These start a cascade of neuroinflammation-related events that may maintain and worsen the original injury, participating in pain generation and chronicization [Valsecchi et al., 2011; Sommer and Kress, 2004; Austin and Moaelem-Taylor, 2010]. Activation of inflammatory cascade, pro- inflammatory cytokines upregulation, and neuroimmune communication pathways play a vital role in structural and functional damage of the peripheral nerves leading to the diabetic peripheral neuropathy. Unfortunately, most of the available analgesic drugs appear to be relatively ineffective in controlling diabetic neuropathic pain, both for insufficient efficacy and side effects [Galer et al., 2000; Kapur, 2003]. Thus, there is a clear need for new disease-modifying therapeutic approaches. Mesenchymal stem/stromal cells (MSCs) may offer a novel therapeutic option to treat diabetic neuropathy. MSCs modulate the nervous system injured environment and promote repair as they secrete anti-inflammatory, anti-apoptotic molecules, and trophic factors to support axonal growth, immunomodulation, angiogenesis, remyelination, and protection from apoptotic cell death [Ma et al., 2014]. Transplanted MSCs not only directly differentiate into endogenous cells on administration, but also secrete a broad range of biologically active factors, generally referred to as the MSCs secretome; in fact even if initially MSCs were proposed for cell therapy based on their differentiation potential, the lack of correlation between functional improvement and cell engraftment or differentiation at the site of injury has led to the proposal that MSCs exert their effects not through their differentiation potential but through their secreted products [Makridakis, 2016; Blaber et al., 2012]. For these reasons in the present study we analyze in a Streptozotocin mouse model of type 1 diabetes the therapeutic effect of hASC (human adipose stem/stromal cells) and their conditioned media (CM-hASC/ secretome) on allodynia and hyperalgesia, on pro- and anti- inflammatory cytokines expression in the main tissue stations involved in nociception transmission as well as in peripheral immune responses. Type 1 diabetes was induced in mice by intraperitoneal (i.p.) injection of moderate low doses of Streptozotocin (STZ, 80 mg/kg, daily for three consecutive days) while control mice were injected with vehicle (citrate buffer). In all groups, mechanical allodynia was evaluated by Von Frey test before diabetes induction and every week after STZ until the end of protocol (14 weeks after STZ). When allodynia was established (2 weeks after STZ) animals were treated with 106 hASC that have been mechanically dissociated to a single cell suspension in PBS solution with 2.5% heparin; CM-hASC from 2x106 cells was also re-suspended in PBS solution with 2.5% heparin and both hASC and CM-hASC were intravenously injected in the tail vein to mice. Animals injected with vehicle only were considered as controls. Our data demonstrated that hASC and CM-hASC treatments were able to reduce allodynia, although the effect of hASC was significantly higher than that elicited by CM-hASC. The effect of both hASC and their secretome was very fast, since a significant reduction of mechanical allodynia was evident already 3 hours after the injection. The antiallodynic effect was maximal beetwen 1 and 2 weeks after treatments and it was extremely long lasting: a significant reduction of allodynia was still present 12 weeks after a single hASC and CM-hASC treatment. Moreover, 4 weeks after the first hASC/CM-hASC treatment (6 weeks after STZ) we decided to treat again a group of diabetic animals with hASC or CM-hASC; repeated hASC treatment did not further ameliorate allodynia. On the other hand, already few hours after the second CM-hASC injection, the antiallodynic effect was significantly potentiated and it completely mimicked the effect evoked by hASC. In order to discover whether hASC and CM-hASC treatments were effective also in a more advanced stage of the disease, when a severe loss of nerve function is reported, we treated animals 6 weeks after diabetes induction. Also in this situation both treatments were efficacious in providing a fast and irreversible antiallodynic effect. Futhermore, in order to verify whether stemness is a fundamental prerequisite for obtaining pain relief a group of STZ-mice was treated with CM obtained from 2x106 human fibroblasts (CM-hF). CM-hF did not exert any effect on mechanical allodynia, demonstrating that only secretome from stem cell cultures is biologically active. It is very important also to consider preparation method of secretome, because lyophilized CM-hASC was unable to provide pain relief, suggesting that during the lyophilization process some essential bioactive factors may be lost. Moreover, since in patients sensory alterations associated to diabetic neuropathy are often diverse in order to ascertain whether the effects of hASC and CM-hASC were limited only to mechanical allodynia, we evaluated thermal hyperalgesia (hot stimuli) and thermal allodynia (cold stimuli) by plantar test and acetone test, respectively. In STZ-mice cold allodynia was present and both treatments were able to significantly reduce it. As regards to heat hyperalgesia, it was present in diabetic mice until 3 weeks from STZ administration, but subsequently we observed hypoalgesia appearance and both treatments were able to avoid hypoalgesia development; these results demostrate the ability of stem cells and their secretome to relieve and prevent the typical diabetic hypersensitivity in response to different types of stimuli. In order to evaluate the impact of treatments on pro- and anti- inflammatory cytokines, animals were sacrificed at different time points: 2 weeks after STZ, i.e. 3 hours after hASC/CM-hASC treatment; 3 weeks after STZ, i.e. 1 week from treatments and 14 weeks after STZ, i.e. 12 or 8 weeks from treatments. From each animal, sciatic nerves, dorsal root ganglia, spinal cord and spleens were collected. IL-1β, TNF-α, IL-6 and IL-10, were evaluated as protein in nervous tissues by ELISA assay. Three weeks after neuropathy induction pro-inflammatory cytokines IL-1β, TNFα and IL-6 resulted overexpressed in peripheral (sciatic nerve and DRG) and central (spinal cord) nervous system of diabetic mice, both hASC and CM-hASC were similarly able to restore pro-inflammatory cytokine levels that 1 week from treatments were back to basal levels; while in all nervous tissues IL-10 levels appeared instead significantly reduced in diabetic animals and both hASC and CM-hASC significantly increased IL-10 concentrations, reaching physiological levels in DRG and spinal cord, while it exceeded basal levels in the sciatic nerve, indicating a switch towards an anti-inflammatory environment in all these tissues. Fourteen weeks after STZ, spinal cord IL-1β, TNF-α and IL-6 levels were still significantly elevated and IL-10 levels reduced in comparison to non diabetic mice, indicating the persistence of neuroinflammation. As observed for the antiallodynic effect, also cytokine modulation induced by hASC and CM-hASC was long lasting. Twelve weeks after treatments performed 2 weeks from STZ, IL-1β, TNF-α and IL-6 levels were still significantly reduced by hASC and CM-hASC treatments, while hASC-treated mice showed a significant normalization of IL-10 levels. Similar effects were observed also in double treatments (2 and 6 weeks after STZ) and both treatments were effective in modulating cytokine levels also when they were administered in an advanced pathological state (6 weeks after STZ). Moreover, to investigate the timing of cytokines modulation exerted by both treatments IL-1 and IL-10 levels in scatic nerves, DRG and spinal cord were measured. Two weeks after diabetic induction, STZ mice were characterized by pro- inflammatory profile and only 3 hours after hASC and CM-hASC administration , both treatments were able to modulate cytokines levels. To further demonstrate the modulation of treatments on pain-related mediators we demonstrated the ability of hASC and CM-hASC to normalize calcitonin gene related peptide level (CGRP), that was elevated in DRG from diabetic animals. Moreover, we evaluated loss of nerve fibers and skin thickness 1 and 12 weeks after a single hASC/CM-hASC administration at 2 weeks after STZ. Both treatments were able to contrast loss of nerve fibers and skin thickness, although hASC treatment was more effective. Since STZ multiple low-doses protocol that we utilized is able to develop an autoimmune response against pancreatic tissue sustained by a T-helper 1 pattern of activation, we studied whether a T-helper polarization was present in splenocytes from diabetic mice and whether hASC or their secretome did exert any immunomodulatory activity. Two weeks after STZ, Con-A stimulated splenocytes released higher levels of IFN-γ, while IL-10 release was significant reduced; both hASC and CM-hASC treatments 3 hours after administration were already able to augment IL-10 levels. Th1/Th2 cytokines unbalance was more evident 3 weeks after STZ and both tretaments appeared able to restablish a correct IFNγ/IL-10 balance. When cytokine levels were measured at longer time from diabetes induction, i.e. 14 weeks after STZ, a clear shift toward a Th1 pattern, characterized by higher IFN-γ and IL-2 secretion and lower levels of IL-4 and IL-10, was present and both hASC/CM-hASC treatments were able to normalize cytokine levels. In the whole, the data indicate that both hASC and CM-hASC treatments are able to block Th1 polarization that develops in this experimental model of diabetes. Moreover, throughout the experiment, blood glucose levels and weight were monitored. In respect to non-diabetic control animals, a significant body weight loss was observed in diabetic mice, that started to be significant 3 weeks after STZ. In STZ-mice the administration of hASC or CM-hASC, 2 weeks after diabetes induction significantly prevented the loss of body weight. Neither treatments did modify blood glucose levels that were elevated in STZ-mice nor glucose tolerance test response. Moreover both hASC and CM-hASC did ameliorate nephropathy that was present in diabetic animals, indicating that the treatments may be useful for treating also other diabetes complications. Our results demonstrated that hASC can control diabetic complications such as neuropathic pain, acting on several peripheral and central mechanisms involved in development and maintenance of this condition, such as neural and immune elements. Moreover the significant new positive results observed also with hASC conditioned medium strongly suggest that their effect is likely to be mediated by their secreted products.