Reversal of Neuropathic Pain in Diabetes by Targeting Glycosylation of Ca_v3.2 T-Type Calcium Channels

Molecular mechanisms of glycosylation of Ca_V3.2 channels. A: Schematic diagram of Ca_v3.2 showing the position of conserved putative N-glycosylation sites in the extracellular face of the channel in domains I and III. Designated asparagine residues (in red bold fonts) were mutated to alanine residues. B: Representative traces in gray show nickel inhibition of T-current (black traces) in HEK-293 cell transiently transfected with WT Ca_V3.2 (top) and N192Q Ca_V3.2 (bottom) channels. In both experiments, 100 μmol/L NiCl₂ was applied in the bath. On average, nickel blocked 97 ± 2% of inward currents of N192Q Ca_V3.2 (n = 5) and WT Ca_V3.2 (n = 6) channels. Bars indicate calibration. C: Bar graph represents the average effect of N192Q Ca_V3.2 mutation compared with WT Ca_V3.2 T-current kinetics (V_h = −90 mV, V_t = −30 mV) in HEK-293 cells. On average, N192Q mutant has slower 10–90% rise times by ~60% (8.2 ± 0.8 s) compared with WT Ca_V3.2 currents (5.1 ± 0.2 s). Similarly, on average N192Q mutant has slower inactivation τ-values by ~50% (31.4 ± 2.2 s) compared with WT Ca_V3.2 currents (21.8 ± 0.6 s). Vertical lines are ±SEM of multiple determinations. Number of cells in each experiment is indicated in parentheses. *Significance of P < 0.001.

Altered membrane expression of putative glycosylation sites in Ca_V3.2 channels. A: Representative confocal images in the left panel show HEK-293 cells transiently transfected with WT Ca_V3.2 channels where concanavalin A immunofluorescence is represented in red color. Middle panel shows green immunofluorescence representing EGFP-tagged Ca_V3.2 transfection in the same cells. Right panel shows merged images where overlay is presented in yellow color. B: Representative confocal images in the left panel show HEK-293 cells transiently transfected with N192Q Ca_V3.2 channels where concanavalin A immunofluorescence is represented in red color. Middle panel shows green immunofluorescence representing EGFP-tagged N192Q Ca_V3.2 transfection in the same cells. Right panel shows merged images where overlay is presented in yellow color. C: Representative confocal images in the left panel show HEK-293 cells transiently transfected with N271Q Ca_V3.2 channels where concanavalin A immunofluorescence was represented in red color. Middle panel shows green immunofluorescence representing EGFP-tagged N271Q Ca_V3.2 transfection in the same cells. Right panel shows merged images where overlay is presented in yellow color. D: Representative confocal images in the left panel show HEK-293 cells transiently transfected with N1466Q Ca_V3.2 channels where concanavalin A immunofluorescence was represented in red color. Middle panel shows green immunofluorescence representing EGFP-tagged N1466Q Ca_V3.2 transfection in the same cells. Right panel shows merged images where overlay is presented in yellow color. Note that EGFP and concanavalin A show very little overlap in their subcellular distributions. E: Bar graphs represent average values from multiple experiments (those depicted in A–D). Colocalization values for EGFP and concanavalin epifluorescence were quantified and compared between mutants and WT Ca_V3.2 channels. Only N1466Q mutants (dark-gray column, 9.9 ± 3.7, n = 4) displayed ~73% decrease in colocalization value compared with WT channels (black column, 36.2 ± 3.8, n = 4). In contrast, colocalization values of N192Q (open column, n = 4) and N271Q mutants (light gray bar, n = 3) were not significantly different from WT Ca_V3.2. **Statistically significant difference from Ca_V3.2 WT channels (P < 0.01). NS (not significant), P > 0.05 compared with Ca_V3.2 WT channels. Calibration bars are marked on all panels.

Biochemical evidence of glycosylation of Ca_V3.2 channels. Immunoblotting with FLAG antibody reveals a shift in apparent molecular weight of NH₂-terminal fragment of Ca_V3.2 channel (bottom arrow) but not full channel (top arrow). Note that treatment with PNGase-F and Endo F1 but not Endo F2 and Endo F3 caused an obvious change in mobility of FLAG-labeled NH₂-terminal fragment of Ca_V3.2 from ~60 to 50 KDa. Using mass spectrometry, we confirmed that the deglycosylated NH₂-terminal fragment of Ca_V3.2 channels recognized by FLAG antibodies contains most of the repeat I of Ca_V3.2 channel (data not shown). N-term., NH₂-terminal.

Alterations of macroscopic T-current kinetics in acutely dissociated small DRG cells from diabetic ob/ob mice. The data show original T-current traces (V_h −90 mV, V_t −80 mV through −30 mV) from representative DRG cells from a healthy WT mouse (A) and a diabetic ob/ob mouse (B). The averaged data show marked acceleration in T-current inactivation (C) and activation (D) kinetics in ob/ob mice compared with age-matched WT mice. Data are averages of multiple cells (WT n = 27, ob/ob n = 20) ±SEM. *P < 0.05. Solid lines on C and D are single exponential fits to experimental data points. E: Normalized peak T-current activation curves from similar experiments shown in A and B. Number of cells is indicated in the parentheses. Solid black lines are fitted using equation 1, giving half-maximal activation (V₅₀), which occurred at −46.6 ± 0.6 mV with a k of 6.3 ± 0.5 mV in WT mice. Similarly, V₅₀ was –47.6 ± 0.7 mV with a k of 7.4 ± 0.6 mV in the DRG cells from ob/ob mice. F: Normalized peak T-current steady-state inactivation curves. T-currents are evoked by test steps to −30 mV after 3.5-s prepulses to potentials ranging from −110 mV to −45 mV in 5-mV increments. Number of cells is indicated in the parentheses. Solid black lines are fitted using equation 2, giving V₅₀, which occurred at −75.0 ± 0.4 mV with a k of 8.9 ± 0.4 mV in WT mice. Similarly V₅₀ was –75.8 ± 0.4 mV with a k of 8.2 ± 0.3 mV in the DRG cells from ob/ob mice. The voltage dependencies of activation and steady-state inactivation were described with single Boltzmann distributions of the following forms where I_max is the maximal activatable current, V₅₀ is the voltage where half the current is activated or inactivated, and k is the voltage dependence (slope) of the distribution. (1) (2), where exp = e^x.

NEU treatment in vitro reversed kinetic alterations and normalized T-current density in small DRG cells from diabetic ob/ob mice. A: Traces represent families of T-currents evoked in representative DRG cells in a WT mouse (top panel) and a diabetic ob/ob mouse after incubation of 1.5 units/mL of NEU at 37°C for 3 h (lower panel) by voltage steps from V_h = −90 mV to V_t from −80 through −25 mV in 5-mV increments. Bars indicate calibration. Bar graphs with the averaged data show that NEU treatments completely reversed DRG T-current density (V_h = −90 mV, V_t = −30 mV) (B), T-current inactivation measured by inactivation τ (C), and activation kinetics measured by 10–90% rise time (D) in ob/ob mice compared with healthy WT mice. Control was compared with post-NEU treatments. DRG cells were freshly dissociated as noted in Fig. 5. Recordings were performed at room temperature while NEU was incubated for 1–3 h at 37°C. Control cells were treated with saline. Data are averages of multiple cells (as indicated in parenthesis) ±SEM. *P < 0.001; n.s., not significant, P > 0.05.

NEU treatment in vivo reversed mechanical hyperalgesia in diabetic ob/ob mice. A: The graph shows average data points indicating that i.pl. injections of NEU but not saline into right paws completely reversed mechanical hyperalgesia in diabetic ob/ob mice. Arrow indicates time point of i.pl. injections. *Significant change of PWRs with P < 0.05 compared with baseline PWR prior to i.pl. injections. B: The graph shows average data points indicating that i.pl. injections of NEU and saline into right paws did not affect PWRs in the left paws of diabetic ob/ob mice. Arrow indicates time point of i.pl. injections. C: The graph shows average data points indicating that i.pl. injections of NEU but not saline into right paws caused transient hyperalgesia in healthy WT mice at the time point of 30 min. Arrow indicates time point of i.pl. injections. *Significant change of PWRs with P < 0.001 compared with baseline PWR prior to i.pl. injections. D: The graph shows average data points indicating that i.pl. injections of NEU and saline into right paws did not affect PWRs in the left paws of WT mice. Arrow indicates time point of i.pl. injections. E: The graph with averaged data from eight experiments shows that intraperitoneal injections (thin arrow) of selective T-channel blocker ECN (25 mg/kg i.p.) completely reversed mechanical hyperalgesia in diabetic ob/ob mice as evidenced by decreased PWRs in both right and left paws (#P < 0.001). Subsequent i.pl. injections (thick arrow) of 1.5 units/mL NEU in the same animals did not significantly influence new baseline values of mechanical PWRs at a time point of 90 min, but it significantly increased PWRs at time points of 120, 150, and 180 min (*P < 0.05). F: The graph with averaged data from eight experiments shows that intraperitoneal injections (thin arrow) of vehicle used to dissolve ECN (Cyc) did not affect mechanical hyperalgesia in diabetic ob/ob mice as evidenced by stable PWRs in both right and left paws at a time point of 60 min. Subsequent i.pl. injections (thick arrow) of 1.5 units/mL NEU in the same animals effectively reversed diabetic hyperalgesia by significantly decreasing PWRs at time points of 120, 150, and 180 min (*P < 0.001).