NaV1.8 voltage-gated sodium channels are involved in the mechanisms of inflammatory and chronic pain perception. The expression of these channels is restricted to subsets of neurons in dorsal root ganglia (DRG). We isolated mRNA from human and rat DRG tissue samples to screen for alternatively spliced channel isoforms. Human SCN10A, the gene encoding for hNaV1.8, was completely covered with RT-PCR amplicons and the resulting PCR products were analyzed by 454 sequencing. This technology gives access to a high-coverage analysis of low abundance transcripts. We identified a splice event, taking place at a so-called NAGNAG tandem splice acceptor site. Alternative 3'-splice site selection results in cDNA in which the codon for glutamine 1032 (hNaV1.8) within the cytoplasmic loop between domains II and III is missing. A paralogous splicing event was previously reported by Kerr et al. (J Biol Chem 279: 24826) for rat and mouse.

Mutant hNaV1.8_dQ1032 representing the gene product of the alternatively spliced sequence was analyzed in whole-cell patch-clamp experiments after heterologous expression in the neuroblastoma cell line Neuro-2A. Experiments were performed under application of 100 nM TTX in the bath solution to suppress TTX-sensitive currents endogenous to Neuro-2A cells. Mutant and wild-type channels reached similar current densities, indicating that the splice event has no effect on the channels functional efficiency. Voltage dependence and kinetics of channel opening and inactivation was assayed by tailored pulse protocols but in none of the measured parameters the alternatively spliced isoform deviated from the control channel (n=20 for both channels, t-test, no p-values <0.05). Use-dependent current block under application of 10 mM lidocaine was measured with depolarizing pulse trains to 0 mV with frequencies of 1 Hz and 20 Hz. Lidocaine induced the expected tonic block behavior and a shift in steady-state inactivation for both channels. The splice event leading to NaV1.8_dQ1032 channels thus has no obvious impact on channel function. This result is compatible with an analogous splice event in the cardiac sodium channel hNaV1.5 (human SCN5A, leading to dQ1077): the NAGNAG splicing in NaV1.5 channels has no apparent phenotype but may be clinically important when combined with other disease-related mutations (Tan et al., 2006, Am J Physiol Heart Circ Physiol 291: H1822).

The abundance of the splice event was independently analyzed by capillary electrophoresis of fluorescence-labeled RT-PCR products. The relative amount of NaV1.8_dQ1032 compared to NaV1.8 wild-type mRNA in adult DRGs was 141% in humans and 112% in rats. Interestingly, this is in contrast to an abundance of 70% in mouse DRG as previously estimated based on 20 transcripts by Kerr et al. (2004). Thus, the NAGNAG splicing in the SCN10A gene is conserved among rodents and human but apparently occurs with species-specific abundance.