Differential expression of neuregulin-1 isoforms and downregulation of erbin are associated with Erb B2 receptor activation in diabetic peripheral neuropathy

Abstract

Background: Aberrant neuron/glia interactions can contribute to a variety of neurodegenerative diseases and we have previously demonstrated that enhanced activation of Erb B2, which is a member of the epidermal growth factor receptor (EGFR) family, can contribute to the development of diabetic peripheral neuropathy (DPN). In peripheral nerves, Erb B receptors are activated by various members of the neuregulin-1 (NRG1) family including NRG1 Type I, NRG1 Type II and NRG1 Type III to regulate Schwann cell (SC) growth, migration, differentiation and dedifferentiation. Alternatively, Erb B2 activity can be negatively regulated by association with the Erb B2-interacting protein, erbin. Since the effect of diabetes on the expression of NRG1 isoforms and erbin in peripheral nerve are unknown, the current study determined whether changes in NRG1 isoforms and erbin may be associated with altered Erb B2 signaling in DPN.

Results: Swiss Webster mice were rendered diabetic with streptozotocin (STZ) and after 12 weeks of diabetes, treated with erlotinib, an inhibitor of Erb B2 activation. Inhibition of Erb B2 signaling partially reversed several pathophysiologic aspects of DPN including a pronounced sensory hypoalgesia, nerve conduction velocity deficits and the decrease in epidermal nerve fiber innervation. We also observed a decrease of NRG1 Type III but an increase of NRG1 Type I level in diabetic sural nerves at early stage of diabetes. With disease progression, we detected reduced erbin expression and enhanced MAPK pathway activity in diabetic mice. Inhibition of Erb B2 receptor suppressed MAPK pathway activity in treated-diabetic sural nerves.

Conclusions: These results support that hyperglycemia may impair NRG1/Erb B2 signaling by disrupting the balance between NRG1 isoforms, decreasing the expression of erbin and correspondingly activating the MAPK pathway. Together, imbalanced NRG1 isoforms and downregulated erbin may contribute to the dysregulation of Erb B2 signaling in the development of DPN.

Figure 1

Inhibition of Erb B2 attenuated NCV and sensory deficits in diabetic mice. Swiss Webster mice were rendered diabetic and after two weeks of diabetes, mechanical (A) and thermal sensitivity (B) were assessed weekly (*, p < 0.05 versus time-matched Veh + Veh; ^, p < 0.05 versus time-matched STZ + Veh). Assessment of MNCV (C) and SNCV (D) in a subgroup of animals (n = 4 per group) after 12 weeks of diabetes confirmed the onset of nerve dysfunction prior to drug treatment. At 13 weeks of diabetes, animals were given vehicle or 25 mg/kg erlotinib once per week for 4 weeks (solid arrow) and then twice per week (dashed arrow) for the final four weeks (n = 8-12 per group). Erlotinib significantly improved mechanical and thermal sensitivity and decreased the deficits in both MNCV and SNCV compared to vehicle treated diabetic mice (**, p < 0.01; ***, p < 0.001).

Figure 2

Inhibition of Erb B2 improved IENFD in DPN mice. Footpad samples were collected from the plantar surface of the hind paws after 12 (n = 4 per group) or 21 weeks (n = 6–8 per group) of treatment. (A): Representative images of IENFD in non-diabetic (Veh) and diabetic (STZ) mice treated with vehicle (Veh) or erlotinib (Erl). Nerve fibers immunopositive for PGP 9.5 are stained red by the chromagen (arrows) and epidermal cells were stained purple by hematoxylin. (B): Quantification demonstrated a significant loss of nerve innervation in diabetic mice prior to drug administration and erlotinib treatment induced a partial recovery in fiber loss (*, p < 0.05; **, p < 0.01; ***, p < 0.001).

Figure 3

Validation of specificity of the custom NRG1 Type III antibody. (A): Comparison of NRG1 Type III detection using the custom N-terminal NRG1 Type III polyclonal antibody (left) and a commercial C-terminal NRG1 Type III polyclonal antibody (SMDF, right). (B-D) Demonstration of the specificity of the N-terminal NRG1 Type III polyclonal antibody. The indicated amounts of the N-terminal NRG1 Type III antibody (B), a polyclonal antibody against NRG1 Type I (C) or the C-terminal NRG1 Type III antibody (SMDF, D) were preabsorbed with the immunizing peptide in a 1:50 to 1:200 ratio for 1 hr at 25°C and then used for immunoblot analysis of a sciatic nerve sample. The peptide specifically decreased the detection of the 75 kDa NRG1 Type III band but had no effect on detecting NRG1 Type I or the 65 kDa C-terminal NRG1 Type III fragment.

Figure 4

NRG1 Type III was decreased in sural nerve after 9 and 12 weeks of diabetes. Mice were treated with vehicle (Veh) or STZ and sural nerves were isolated at 9 (A) or 12 (B) weeks after the induction of diabetes (n = 6–7 per group). Protein lysates were prepared and NRG1 Type III level was determined by immunoblot using the two antibodies. (C): Bands were quantified, NRG1 Type III levels were normalized to β-actin and expressed as a fold of the levels in control sural nerves (*, p < 0.05 versus Veh; **, p < 0.01 versus Veh).

Figure 5

NRG1 Type III was decreased in sciatic nerve after 9 and 12 weeks of diabetes. Mice were treated with vehicle (Veh) or STZ and sciatic nerves were isolated at 9 (A) or 12 (B) weeks after the induction of diabetes (n = 6–7 per group). Protein lysates were prepared and NRG1 Type III level was determined by immunoblot using the two antibodies. (C): Bands were quantified, NRG1 Type III levels were normalized to β-actin and expressed as a fold of the levels in control sciatic nerves (*, p < 0.05 versus Veh; **, p < 0.01 versus Veh).

Figure 6

NRG1 Type I was increased in sural nerves after 9 and 12 weeks of diabetes. (A): Sural nerves were isolated from 9 and 12-week control and diabetic mice (n = 6–7 per group). Protein lysates were prepared and NRG1 Type I levels were determined by immunoblot using the Abcam NRG1 Type I antibody. (B): Bands were quantified, NRG1 Type I levels were normalized to β-actin and expressed as a fold of the levels in control sural nerves (*, p < 0.05 versus Veh).

Figure 7

Erbin was decreased and p42/p44 MAPK activity enhanced in diabetic sciatic nerve. Sciatic nerves were isolated from 9, 12 (A) and 16-week (C) control and diabetic mice (n = 8–9 per group). Protein lysates were prepared and Erbin levels and p42/p44 MAPK (pErk) levels were determined by immunoblot. Quantification demonstrated a significant decrease of Erbin levels (B) and an increase in p42/p44 MAPK activation at 16 weeks (D). (*, p < 0.05 versus Veh; **, p < 0.01 versus Veh).

Figure 8

Inhibition of Erb B2 with erlotinib suppressed diabetes-induced p42/p44 MAPK pathway activation in sural nerves. (A): Sural nerves were isolated from vehicle or erlotinib-treated control and diabetic mice (n = 3–4 per group) at week 21. Protein lysates were prepared and erbin levels and p42/p44 MAPK (pErk) levels were determined by immunoblot. (B): Quantification demonstrated a significant decrease of erbin and an increase in p42/p44 MAPK activation in vehicle-treated diabetic mice. Erlotinib treatment suppressed p42/p44 MAPK activation in erlotinib-treated diabetic mice (*, p < 0.05; **, p < 0.01).