Potential therapeutic use of deferoxamine and mesenchymal stem cells in type-1 diabetes: assembling another piece of the jigsaw, in what is a complex puzzle

Abstract

Mesenchymal stem cells (MSCs) are recognized for their potential in regenerative medicine. Due to long-term negative effects associated with insulin administration and difficulty with pancreas transplantation, patients with type-1 diabetes could significantly benefit from organ-targeted cell-based therapy. Although several pharmacological agents increase the homing capacity of MSCs, the mechanisms regulating this process are still poorly understood. In this issue, Najafi et al. have demonstrated that pre-treatment of bone marrow-derived MSCs with the iron chelator and hypoxia mimetic, deferoxamine, can increase homing to the pancreas in a rat model of diabetes. This effect appears to be driven through specific chemokines in addition to hypoxia-inducing factor 1-alpha. Results from this study provide important clues in our endeavour to solve a complex process. Further studies will help determine whether these findings may offer potential therapeutic benefit to patients with diabetes.

Keywords::

• CXCR4
• deferoxamine
• diabetes
• homing
• hypoxia-inducible factor 1α
• mesenchymal stem cells

1. Introduction

In this issue, Najafi et al. [1] have published a research article examining the homing capacity of mesenchymal stem cells (MSCs) in vitro and in a rat model of type-1 diabetes. This is a well designed study that yields important additional information to what is currently known on the topic. They demonstrated enhanced homing of MSCs to the pancreas, when these cells were subjected to hypoxic conditions, by pre-treating them with the iron chelating agent deferoxamine (DFO). Furthermore, the underlying mechanisms by which DFO elicited its favorable effects were shown to be mediated through specific chemokine receptors (CXCR4 and CCR2) and matrix metalloproteinases (MMP-3 and MMP-9), and the transcription factor HIF-1α. This study provides a snapshot of the various mechanisms involved during the homing process. To best understand how these various elements knit together in this study, we need to take a step back and examine each piece of the jigsaw closely.

Type 1 diabetes mellitus (T1DM) results from autoimmune destruction of insulin-producing pancreatic β-cells. As a consequence of beta cell destruction, patients require long-term insulin administration to control their blood glucose levels. Furthermore, difficulty in maintaining and controlling near physiological blood glucose levels results in a higher prevalence of both microvascular and macrovascular complications in these patients [2]. Transplantation of whole pancreas and/or pancreatic islet cells is another choice for the treatment of T1DM [3], although scarcity of organs and necessity of long-term immunosuppressant therapy has limited this therapeutic option. In light of these unsatisfactory therapeutic strategies, the search for and successful implementation of an alternative approach is imperative.

One such approach that seems to offer much potential is the use of MSCs [4]. MSCs not only possess immunomodulatory properties that could aid in interventions designed to reduce immune rejection, but there is also evidence that they can differentiate into glucose-responsive, insulin producing cells [5]. Furthermore, these cells can be readily isolated from a patient, expanded in culture and have shown limited tendency to form tumors.

MSCs also express various chemokine receptors, including CXCR4 and CCR2, both of which play a key role in the migration and homing of MSCs and other haematopoietic cells to ischemic, damaged or inflammatory tissue states [6]. Of particular interest is the CXCR4 receptor and its specific ligand, stromal cell-derived factor-1α (SDF-1α). The SDF-1α/CXCR4 signalling axis plays a central role in the homing of MSCs, haematopoietic stem cells (HSCs) and progenitor cells from the bone-marrow to distal sites. Moreover, patients with type-2 diabetes have been shown to have depleted levels of endothelial progenitor cells (EPCs), which has been linked to reduced expression and/or function of CXCR4 [7]. CXCR4 along with MMP-3 and -9, are critical to the migration of MSCs to injured tissues, such as the pancreas in the streptozotocin (STZ)-rat model of type-1 diabetes, chosen by Najafi et al. Since the induction of CXCR4 by hypoxia has previously been shown to be driven by the transcription factor HIF-1α in progenitor cells [8], Najafi et al. also examined how chemokine-driven cell homing was regulated by controlling the expression of HIF-1α through its selective inhibition.

The iron chelating agent DFO, a hypoxia mimetic, was used by Najafi et al. to pre-treat MSCs. Previous studies in other cell types have demonstrated protective effects afforded by DFO, through stabilization of HIF-1α. In a seminal study by Thangarajah et al. [9], it was demonstrated that impairment in neovascularization and wound healing from a glycemic induced defect in the transactivation of HIF-1α could be reversed by DFO administration in diabetic mice. Moreover, Thangarajah et al. demonstrated the importance of HIF-1α in the control of vascular endothelial growth factor (VEGF), a pivotal growth factor involved in neovascularisation, the effect of which could be augmented by DFO administration. The role of VEGF in angiogenesis is well established; however, the underlying mechanisms regulating VEGF expression by DFO in diabetes is a novel concept. Lesser understood are the signaling mechanisms driving homing of MSCs following exposure to DFO. A particularly novel aspect of the work by Najafi et al. is the possibility that DFO can be used to target MSCs through CXCR4/SDF-1α axis to injured pancreas and potentially other injured/ischemic sites. It will be important to establish the individual and combined role of VEGF and SDF-1α by DFO in future studies in animal models of diabetes.

2. Expert opinion

Collectively, findings from Najafi et al. and other studies provide us with a clue as to the various pieces of this complex jigsaw that may well interlock momentarily to provide a clear view of the complete picture. However, it is important to bear in mind that under hypoxic conditions (e.g., diabetic foot ulcers and diabetic retinopathy), not only HIF-1α will be activated, but also potentially a whole myriad of other signaling pathways (e.g., PI3K/AKT) in addition to other transcription factors (e.g., NF-kappa-B and CREB) [10], which in turn can lead to up-regulation of genes coding for production of other potent cell mediators (e.g., nitric oxide, growth factors, cytokines and chemokines) that may or may not be specific for MSCs. Therefore, although DFO may normalize hypoxic conditions through stabilization of HIF-1α, leading to favorable effects in specific conditions, activation of other signaling pathways and their downstream effects also need to be considered.

In this respect, therapeutic use of DFO in patients with T1DM may well facilitate transplantation of pancreatic β-cells either, through increased homing of MSCs directly from the bone-marrow reserve pool or from MSCs (previously primed with DFO) injected systemically. However, unlike small animal models of diabetes where no concomitant medication is administered, the majority of patients with diabetes have comorbid conditions and as such receive other medication (e.g., anti-hypertensives and statins, etc.) on top of anti-diabetic drugs, which are known to stimulate the mobilization of HSCs, therefore, potentially compromising any effect of a specific agent such as DFO.

The isolation and characterisation of MSCs is another important aspect that should not be underestimated. While Najafi et al. fulfilled two of the criteria for the characterization of MSCs (adherence to plastic under standard culture conditions and surface expression of the markers CD73, CD90 and CD105 and absence or low expression of CD14, CD34 and CD45), they did not demonstrate that these cells had the capacity to differentiate into adipocytes, chondrocytes and osteoblasts [11]. It is important that this and other studies alike adhere to strict and standard identification methods, so as to facilitate the rapid passage of regenerative medicine from the bench to the bedside.

Najafi et al. also demonstrated that 60% more grafted MSCs migrated to the pancreas after a period of 24 h compared to MSCs that were not pre-treated with DFO. Although beyond the scope of their study, which was mainly focused on enhancing migration/homing of MSCs, it would have been interesting and important clinically to know whether this improvement in migration equated to a significant impact in terms of pancreatic function (i.e., insulin production) over the days and weeks after, in STZ-diabetic rats. A recent report has confirmed the use of DFO in improving islet function, following transplantation in a murine model of diabetes. Stokes et al. [12]. directly examined the role of HIF-1α and its stability following islet transplantation in STZ-diabetic mice. They transplanted human or murine b-cell-specific HIF-1a-null (b-HIF-1a-null) islets with or without DFO treatment. b-HIF-1a-null transplants had poor outcomes, demonstrating that lack of HIF-1a impaired transplant efficiency. Increasing HIF-1a with DFO treatment improved outcomes for both mouse and human islets. They concluded that HIF-1α is a protective factor required for successful outcome following islet transplantation. Furthermore, iron chelation with DFO significantly improved transplant success in a HIF-1α-dependent manner. They also suggest that since DFO is already approved for human use, it may therefore have a therapeutic role in the setting of human islet transplantation. Since they estimated that the doses of DFO used led to an increase of between 50 and 60% pancreatic β-cell mass after 1 month in STZ-diabetic mice were in the clinical range for therapeutic use, DFO may well be considered for use in human islet transplantation procedures.

Actually, it is worth noting that increased tissue iron levels have previously been linked with diabetes, both in hereditary haemochromatosis and in dietary iron overload [13]. Furthermore, in animal models, excess iron has been shown to have adverse effects on the endothelium [14] and diabetes-induced endothelial dysfunction can be prevented by long-term treatment with hydroxyethyl starch-conjugated DFO [15]. In patients with coronary artery disease, DFO was also shown to improve nitric oxide-mediated vasodilation [16]. More recently, DFO has been shown to be able to ameliorate adipocyte hypertrophy through suppression of oxidative stress, inflammatory cytokines and macrophage infiltration in obese diabetic mice [17]. Collectively, these studies point towards potential cardioprotective and anti-inflammatory effects by DFO that may be linked towards its role in neovascularisation [9]. Although DFO is approved for human use in cases of acute iron overload, it is not indicated for the treatment of primary haemochromatosis, since phlebotomy is the method of choice in these patients. However, DFO is the second line treatment in individuals where venesection cannot be performed. On this basis, the use of DFO for the moment should be restricted to topical application in patients with diabetic foot ulcers and other ischemic lesions may be encouraged, but systemic use of DFO for the above described reasons should be performed with caution.

3. Conclusion

In this issue, Najafi et al. have made an important contribution to our understanding of the underlying mechanisms regulating homing of MSCs in a rat model of diabetes. Given the advantages that MSCs have over other cell types (potential to differentiate and regulate an immune response); they remain promising therapeutic candidates in diabetic complications. Previous evidence and findings in the present issue on the effect of DFO to drive the expression of VEGF and CXCR4/SDF-1α axis (via HIF-1α) on MSCs (and other HSCs) in animal models of diabetes warrants further investigation. Particularly, studies examining the effects of continuous DFO administration over the long-term (several weeks) on well-defined endpoints are lacking. Whether these pre-clinical findings can be translated to the clinic still remains to be established.

Declaration of interest

The author states no conflict of interest and has received no payment in preparation of this manuscript.