Gabapentin Disrupts Binding of Perlecan to the α2δ1 Voltage Sensitive Calcium Channel Subunit and Impairs Skeletal Mechanosensation

Abstract

Our understanding of how osteocytes, the principal mechanosensors within bone, sense and perceive force remains unclear. Previous work identified "tethering elements" (TEs) spanning the pericellular space of osteocytes and transmitting mechanical information into biochemical signals. While we identified the heparan sulfate proteoglycan perlecan (PLN) as a component of these TEs, PLN must attach to the cell surface to induce biochemical responses. As voltage-sensitive calcium channels (VSCCs) are critical for bone mechanotransduction, we hypothesized that PLN binds the extracellular α₂δ₁ subunit of VSCCs to couple the bone matrix to the osteocyte membrane. Here, we showed co-localization of PLN and α₂δ₁ along osteocyte dendritic processes. Additionally, we quantified the molecular interactions between α₂δ₁ and PLN domains and demonstrated for the first time that α₂δ₁ strongly associates with PLN via its domain III. Furthermore, α₂δ₁ is the binding site for the commonly used pain drug, gabapentin (GBP), which is associated with adverse skeletal effects when used chronically. We found that GBP disrupts PLN::α₂δ₁ binding in vitro, and GBP treatment in vivo results in impaired bone mechanosensation. Our work identified a novel mechanosensory complex within osteocytes composed of PLN and α₂δ₁, necessary for bone force transmission and sensitive to the drug GBP.

Keywords: bone; gabapentin; mechanosensation; osteocytes; perlecan; voltage-sensitive calcium channels; α2δ1.

Figure 1

Structure of voltage sensitive calcium channels. The channel complex is composed of the α₁ pore-forming subunit with auxiliary β, γ, and α₂δ subunits bound to the pore, positioned to alter gating kinetics of the channel [16,17,18]. The α₂δ subunit is anchored in the membrane via the δ portion, with the α₂ region positioned extracellularly. In the extracellular portion (α₂) of the α₂δ subunit, the von Willebrand Factor A domain (vWFA) sequence and the Arg-Arg-Arg (RRR) motif for Gabapentin binding are indicated. Reprinted/adapted with permission from Ref. [19], 2022, Springer Nature.

Figure 2

PLN colocalizes with WGA and α₂δ₁ in osteocyte-like cells. MLO-Y4 cells stained with (a) wheat germ agglutinin (WGA)-FITC (green) and (b) perlecan (PLN) (red), (c) merge PLN and WGA. On the bottom panels, cells were stained for (d) α₂δ₁ (red) and (e) PLN (green), (f) merge PLN and α₂δ₁. White arrows indicate overlapping fluorescent signal. (g) Co-immunoprecipitation assays from MLO-Y4 lysates show that PLN and α₂δ₁ associate. IgG was used as a negative control. Blots were probed for β-actin antibody as a loading control.

Figure 3

Docking models of vWFA domain of α₂δ₁ and PLN Dm III-2. In silico protein–protein functional interactions and 3D docking models between the von Willebrand Factor A (vWFA) domain of α₂δ₁ and perlecan (PLN) domain (Dm) III-2 were generated with the free web server HDOCK. 4FX5 (brown) is the vWFA domain and 4YEP (yellow) is the L4b domain of human Laminin α₂ (PLN Dm III-2). (Top), cartoon ribbon-style 3D representations of receptor and ligand. (Bottom), surface style representation of the proteins. Each image is rotated 90° clockwise from the previous one.

Figure 4

Schematic representation of drug-binding experiments. (a) When we first added full-length PLN to α₂-nanoplasmonic sensors, PLN binds to α₂. However, further addition of GBP disrupts PLN::α₂ binding. (b) When GBP is added first to α₂ sensors, further addition of PLN does not disrupt the GBP::α₂ interaction. (c) When addition of PLN and GBP occur simultaneously, α₂ has greater affinity for GBP.

Figure 5

Gabapentin treatment impairs bone mechanosensitivity and load-induced bone formation. Male C57BL/6J mice were injected daily with saline (vehicle, VEH) (n = 9) or Gabapentin (GBP, 300 mg/kg BW) (n = 7) for 4 weeks while undergoing axial ulnar loading. (a) Representative images of control (non-loaded) and loaded ulnas from VEH and GBP treated mice. To monitor load-induced bone formation, calcein (green) and alizarin (red) fluorochromes were injected intraperitoneally 8 and 20 days after the first loading day, respectively. Changes in (b) mineralizing surface (MS/BS), mineral apposition rate (MAR), and bone formation rate (BFR/BS) in response to mechanical loading were assessed in VEH and GBP treated mice were analyzed by ANOVA and Fisher’s LSD test for pairwise comparisons. (c) The percent difference between a loaded and non-loaded limbs within each animal [%Δ  =  ((loaded limb − control limb)/control limb) × 100%)] for MS/BS, MAR and BFR/BS in GBP and VEH groups was calculated analyzed separately by unpaired Student’s t-tests. Each dot represents one mouse. Values are shown as Mean ± SEM; (*) p ≤ 0.05.

Figure 6

Summary of results. In this work we found that (a) the α₂δ₁ subunit of voltage sensitive Ca²⁺ channels binds perlecan (PLN) creating a mechanosensory complex that enables connection between the mineralized matrix and the osteocyte cell membrane. (b) We also demonstrated that gabapentin (GBP) interferes with binding of PLN and α₂δ₁ in vitro. As the PLN/α₂δ₁ complex is necessary for mechanotransduction, GBP uncoupling of the complex results in impaired osteocyte mechanosensation in vivo, which may account for the deleterious skeletal effects observed with chronic use of this drug.