High glucose up-regulates Semaphorin 3A expression via the mTOR signaling pathway in keratinocytes: A potential mechanism and therapeutic target for diabetic small fiber neuropathy

Highlights

• Higher Sema3A expression, over-activation of mTOR signaling, and reduced IENFD are observed in diabetic patients with SFN.

• HG or rSema3A up-regulates Sema3A expression via the mTOR signaling in HaCaT cells.

• Increased Sema3A expression, reduced IENFD and hyperalgesia are ameliorated by rapamycin in diabetic rats.

Abstract

Small fiber neuropathy (SFN) is a common complication in diabetes, and is characterized by decreased intraepidermal nerve fiber density (IENFD). Semaphorin 3A (Sema3A), which is produced by keratinocytes, has a chemorepulsive effect on intraepidermal nerve fibers. mTOR signaling can mediate local protein synthesis that is critical for growth of axons and dendrites. Therefore, this study aimed to investigate whether Sema3A is up-regulated in diabetic keratinocytes via the mTOR-mediated p70 S6K and 4E-BP1 signaling pathways, and furthermore whether it is involved in the pathogenesis of diabetic SFN. IENFD, expression of Sema3A, and mTOR signaling, were evaluated in the skin of diabetic patients with SFN as well as control subjects. Sema3A and mTOR signaling were also assessed in HaCaT cells which had been treated with high glucose (HG) or recombinant Sema3A (rSema3A) in the presence or absence of rapamycin. Small fiber dysfunction was evaluated by examining IENFD and using behavioral tests in control and streptozotocin-induced diabetic rats treated with or without rapamycin. We found that higher Sema3A expression and over-activation of mTOR signaling, was accompanied by reduced IENFD in the skin of diabetic patients compared with control subjects. The expression of Sema3A, and mTOR signaling were up-regulated in HaCaT cells incubated with HG or rSema3A, and this could be attenuated by rapamycin. Hyperalgesia, reduced IENFD, and up-regulated Sema3A and mTOR signaling were also detected in diabetic rats. These effects were ameliorated by rapamycin treatment. Our data indicate that HG up-regulates Sema3A expression by activating mTOR signaling in diabetic keratinocytes. This pathway may therefore play a critical role in diabetic SFN.

Introduction

With the high prevalence of diabetes in China (Yang et al., 2010), diabetic peripheral neuropathy (DPN) has become a major cause of foot ulceration and amputation, affecting up to 50% of diabetic patients (Tesfaye et al., 2010). DPN can involve both small and large fibers, and small fibers are affected earlier and are more prone to be damaged than large fibers (Smith et al., 2001, Sumner et al., 2003, Tesfaye et al., 2010). Small fibers (thinly myelinated Aδ and unmyelinated C fibers) originated in the dorsal nerve root ganglia (DRG), mediate pain, temperature sensation, and autonomic functions (Hoeijmakers et al., 2012), and terminate in the skin as intraepidermal nerve fibers (IENFs) (Novotny and Gommert-Novotny, 1988). Therefore, numerous investigators have suggested that assessment of intraepidermal nerve fiber density (IENFD) during a skin biopsy is a reliable and standardized technique for diagnosing small fiber neuropathy (SFN) (Chen et al., 2015, Loseth et al., 2008, Tesfaye et al., 2010, Ziegler et al., 2014).

Keratinocytes, the major epidermal cell type, can modulate epidermal innervation by synthesizing and secreting numerous nerve repulsion factors and nerve elongation factors (Kumamoto et al., 2014, Roggenkamp et al., 2012, Tominaga et al., 2009, Tominaga and Takamori, 2014). Semaphorin 3A (Sema3A), a member of the semaphorin family, acts as the first secreted repulsion molecule in axonal formation during neuronal development (Messersmith et al., 1995), and in regeneration after peripheral nervous system injury (Pasterkamp et al., 1998). Sema3A which is also produced by keratinocytes (Fukamachi et al., 2011, Tominaga and Takamori, 2014) is one of the most important nerve repulsion factors involved in regulating epidermal innervation in skin diseases (Tominaga et al., 2009, Tominaga and Takamori, 2014, Yamaguchi et al., 2008). However, whether Sema3A affects IENFs in diabetes is unknown. Studies have demonstrated that Sema3A is elevated in the vitreous of diabetic patients with retinopathy and in the podocytes of diabetic patients with nephropathy (Aggarwal et al., 2015, Cerani et al., 2013). Hyperglycemic conditions can also lead to the dysregulation of keratinocyte morphology, differentiation, mobility and proliferation (Spravchikov et al., 2001, Usui et al., 2008). An important question is therefore whether high glucose (HG) induces Sema3A expression in keratinocytes. Many reports have suggested that the severity of diabetic SFN is closely related to the degree of hyperglycemia (Liu et al., 2015, Sveen et al., 2013). Sema3A, depending on its concentration, has been reported to inhibit the growth of DRG neuronal axons in vitro (Ben-Zvi et al., 2008, Kumamoto et al., 2014). Therefore, in this study we assessed whether HG up-regulated Sema3A expression in keratinocytes, which might then damage IENFs and contribute to diabetic SFN.

The mammalian target of rapamycin (mTOR) stimulates protein synthesis and regulates glucose metabolism (Duvel et al., 2010, Krebs et al., 2007, Ma and Blenis, 2009) via phosphorylation and subsequent activation of its downstream effectors. These downstream effectors are the p70 ribosomal protein S6 kinase (p70 S6K) and the eukaryotic initiation factor 4E-binding protein 1 (4E-BP1) (Hay and Sonenberg, 2004, Ma and Blenis, 2009), activation of which can be inhibited by the mTOR inhibitor rapamycin (Hay and Sonenberg, 2004). Studies have shown that mTOR mediated-local protein synthesis can participate in the development of axons and dendrites (Asante et al., 2010, Jimenez-Diaz et al., 2008, Raab-Graham et al., 2006) and can modulate transmission of pain messages (Asante et al., 2010, Obara and Hunt, 2014). Because the mTOR inhibitor attenuates streptozotocin (STZ)-induced hyperalgesia (He et al., 2016)and nerve injury-induced hypersensitivity (Jimenez-Diaz et al., 2008), the use of rapamycin is therefore a potential therapeutic strategy for the treatment of peripheral neuropathy. Several experiments have shown that HG activates mTOR signaling in many different types of cells, such as HT22 cells and hippocampal neurons (Liu et al., 2016a, Wu et al., 2017). mTOR signaling has been shown to regulate Sema3A mRNA expression in osteoblast-like MC3T3-E1 cells (Yamada et al., 2016a). Based on these published data, we wished to investigate whether HG regulates the expression of Sema3A via mTOR signaling. A previous study has indicated that mTOR/4E-BP1 signaling can be activated in growth cones by recombinant Sema3A (rSema3A) (Manns et al., 2012). Additionally, Sema3A can regulate mTOR/p70 S6K signaling in Lewis lung carcinoma cells (Yamada et al., 2016b, Yamada et al., 2016c). Therefore, we hypothesized that a positive feedback loop exists between the Sema3A and mTOR signaling pathways that might aggravate the pathology caused by HG up-regulation of Sema3A expression.

In this study, we used the STZ-induced diabetic rat model to mimic some of the functional and structural features of diabetic SFN. We evaluated the Sema3A and mTOR-mediated p70 S6K and 4E-BP1 signaling pathways, as well as IENFD in diabetic patients and STZ-induced diabetic rats. We assessed Sema3A and mTOR signaling in HaCaT cells which had been treated with HG or rSema3A in the presence or absence of rapamycin. Rapamycin was also used to elucidate the roles of mTOR signaling in diabetic SFN. Our data offer new a new perspective in the potential molecular mechanisms underlying diabetic SFN and provide a novel therapeutic target for the treatment of DPN.