Gabapentin's effect on human dorsal root ganglia: Donor-specific electrophysiological and transcriptomic profiles

Abstract

Neuropathic pain affects approximately 10% of the adult population and is commonly treated with gabapentin (GBP), a repurposed anticonvulsant drug. Despite its widespread use, GBP's effectiveness varies significantly among patients, highlighting the need to better understand its functional and molecular impacts on human nociceptors. Here we characterized the electrophysiological and transcriptomic effects of GBP on primary neurons derived from the dorsal root ganglia (DRGs) of ethically consented human donors. Using patch-clamp electrophysiology, we demonstrated that GBP treatment reduced neuronal excitability, with more pronounced effects in multi-firing vs. single-firing neurons. Notably, significant donor-specific variability was observed in electrophysiological responsiveness to GBP treatment in vitro. RNA sequencing of DRG tissue from the donor that was more responsive to GBP revealed differences in transcriptome-wide expression of genes associated with ion transport, synaptic transmission, inflammation, and immune response. Cross-transcriptomic analyses further showed that GBP treatment counteracted these alterations, rescuing aberrant gene expression at the pathway level and for several key genes. This study provides a comprehensive electrophysiological and transcriptomic profile of the effects of GBP on human DRG neurons. These findings enhance our understanding of GBP's mechanistic actions on peripheral sensory neurons and could help optimize its use for managing neuropathic pain.

Keywords: Gabapentin; hDRG; ion channel; sensory neurons; α2δ1.

Figure 1.

Target of GBP, α2δ1, is highly expressed in hDRG neurons. (a) Uniform manifold approximation and projection (UMAP) plot depicting clustering of cross-species DRG neurons labeled by marker gene expression, identifying diverse neuronal subtypes.^, Represented species include humans, cynomolgus macaque, rhesus macaque, mouse, rat, and guinea pig. (b) UMAP overlay showing the cross-species expression (normalized and log-transformed counts) of CACNA2D1 (encoding α2δ1) across neuronal subtypes.^, (c) Percentage of hDRG-N expressing CACNA2D1 across neuron subtypes.^,,– (d) Immunohistochemical staining of hDRG tissue from donor H22 for α2δ1 (green), peripherin (red, peripheral neuron and nociceptor marker), and DAPI (blue, nuclear stain). Merged images confirm robust expression of α2δ1 in peripherin-positive neurons.

Figure 2.

GBP reduces neuronal excitability in hDRG neurons. (a) Percentage of multi- vs. single-firing hDRG-N treated with vehicle (n = 38) or GBP (n = 21). (b) Representative traces of a multi-firing vehicle- or GBP-treated cell. (c) Percentage of cells with rebound firing. (d) Percentage of cells with delayed firing: first-spike latency (FSL) > 100 ms. (e) Frequency-current (f–I) relationship in multi-firing cells. (f) Percentage of cells with ongoing activity (spontaneous activity (SA) when held at −45 mV). (g) Percentage of cells with SA at resting membrane potential (RMP). (h) Representative traces of a vehicle-treated control cell with SA (examples indicated by red arrows) and depolarizing spontaneous fluctuations (DSFs; purple arrows) and a GBP-treated cell with no SA or DSFs. (i) Percentage of cells with DSFs. (j) DSF frequency/30 s in cells with DSFs. (k) DSF amplitudes in cells with DSFs. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by Fisher’s exact test for (c), (d), (g), and (i) and by Kolmogorov-Smirnov test for (j).

Figure 3.

GBP reduces the excitability of multi-firing hDRG neurons but not single-firing hDRG neurons. Top half: single-firing hDRG-N treated with vehicle (n = 15) or GBP (n = 8). (a) Representative rheobase trace; black trace is vehicle-treated control while teal trace is GBP-treated. (b) Input resistance. (c) Resting membrane potential (RMP). (d) Rheobase. (e) Normalized rheobase. (f) Action potential (AP) peak. (g) AP rise time. (h) AP max rise slope. Bottom half: multi-firing hDRG-N treated with vehicle (n = 23) or GBP (n = 13). (i) Representative rheobase trace; black trace is vehicle-treated control while teal trace is GBP-treated. (j) Input resistance. (k) RMP. (l) Rheobase. (m) Normalized rheobase. (n) AP peak. (o) AP rise time. (p) AP max rise slope. *p < 0.05, **p < 0.01 by Mann–Whitney U test.

Figure 4.

Donor-specific electrophysiological effects of GBP. (a) Percentage of multi-firing cells in vehicle-treated H16/H17 (n = 26), GBP treated H16/H17 (n = 10), vehicle-treated H22 (n = 12), and GBP-treated H22 (n = 11) hDRG-N. (b) Percentage of cells with rebound firing. (c) Resting membrane potential (RMP). (d) Percentage of cells with spontaneous activity (SA) at rest. (e) Percentage of cells with ongoing activity (SA at −45 mV). (f) Rheobase. (g) Normalized rheobase. (h) Input resistance. *p < 0.05, **p < 0.01, ****p < 0.0001 by Fisher’s exact test for (a), (b), (d), and (e) and by two-way ANOVA test for (c), (f), (g), and (h).

Figure 5.

Donor-specific transcriptomic profile of hDRG tissue associated with in vitro GBP responsiveness. (a) Principal component analysis (PCA) of gene expression in hDRG tissue from the donor that had a strong electrophysiological response to in vitro GBP treatment (H22, the “more-responsive donor”) and the donors that did not respond very strongly (H16/H17, the “less-responsive donors”). (b) Volcano plot representing the differential gene expression analysis of hDRG tissue from H22 vs. H16/H17. Significant differentially expressed genes (DEGs) (|log2FC| ≥ 0.4, p ≤ 0.05) in H22 vs. H16/H17 are colored red (upregulated) or blue (downregulated). DEGs that are core genes contributing to the enrichment of the gene set enrichment analysis (GSEA) terms shown in (c) are labeled. Labeled upregulated genes contribute to at least three of the positively enriched GSEA terms in (c) while downregulated genes contribute to at least one negatively enriched term in (c). The thicker central portions of the axes are expanded relative to the peripheral thinner portions. (c) GSEA of Gene Ontology (GO) terms comparing gene expression in hDRG tissue from H22 vs. H16/H17. Significant GO terms (p ≤ 0.05) related to ion transport, synaptic transmission, inflammation, and immune response are shown. The x-axis represents the set size (number of expressed genes in the term). The dot color represents the normalized enrichment score while the dot size represents -log10(p-value). (d) GSEA of GO terms related to ions (their transport, voltage-gated channels, and regulation), specifically focusing on calcium, sodium, and potassium. The dot plot on the left represents the statistics associated with the terms; the dot color represents the normalized enrichment score while the dot size represents -log10(p-value). The heatmap on the right represents the DEGs that constitute part of the core enrichment of the terms, with tile color indicative of log2FC of the DEGs.

Figure 6.

Transcriptomic response of cultured hDRG to GBP treatment. (a) Volcano plot representing the differential gene expression analysis of hDRG from the more-responsive donor cultured and treated with GBP vs. vehicle. Significant differentially expressed genes (DEGs) (|log2FC| ≥ 0.4, p ≤ 0.05) in GBP-treated vs. vehicle-treated control hDRG are colored red (upregulated) or blue (downregulated). DEGs that are core genes contributing to the enrichment of at least one gene set enrichment analysis (GSEA) term shown in (b) are labeled. The thicker central portions of the axes are expanded relative to the peripheral thinner portions. (b) GSEA of Gene Ontology (GO) terms comparing GBP-treated vs. control hDRG culture. Significant GO terms (p ≤ 0.05) related to ion transport, synaptic transmission, inflammation, and immune response are shown. The x-axis represents the set size (number of expressed genes in the term). The dot color represents the normalized enrichment score while the dot size represents -log10(p-value). (c) Heatmap depicting the relative expression of strong DEGs (|log2FC| ≥ 0.4, FDR ≤ 0.1) among the three control and three GBP-treated hDRG samples. The color represents relative expression within each gene (Z-score). Gene biotype and treatment are annotated by color, and hierarchical clustering of samples and genes is shown in their respective dendrograms. (d) Expression of strong DEGs (|log2FC| ≥ 0.4, FDR ≤ 0.1) in hDRG neuron subtypes as quantified and characterized by previous studies.^,– The dot color represents average gene expression (normalized and log-transformed counts) while the dot size represents the percentage of cells expressing the gene. The text color of hDRG neuron subtype labels indicates their fiber type.

Figure 7.

Cross-transcriptomic analysis comparing GBP-responsive and GBP-treatment signatures in hDRG. (a) Gene Ontology (GO) terms enriched by gene set enrichment analysis (GSEA) in opposite directions in the dataset of hDRG tissue from the more-responsive donor vs. less-responsive donors and the dataset of cultured hDRG from the more-responsive donor treated with GBP vs. vehicle. Significant GO terms (p ≤ 0.05) related to ion transport, synaptic transmission, inflammation, and immune response are shown. The x-axis represents −log10(p-value) bidirectionally, and the bar color represents the normalized enrichment score. (b) Common DEGs (|log2FC| ≥ 0.4, p ≤ 0.05) among the two datasets (hDRG tissue and hDRG treated in vitro). The dot color represents log2FC while the dot size represents -log10(p-value). (c) Expression of these common DEGs (|log2FC| ≥ 0.4, p ≤ 0.05) in hDRG neuron subtypes as quantified and characterized by previous studies.^,– The dot color represents average gene expression (normalized and log-transformed counts) while the dot size represents the percentage of cells expressing the gene. The text color of hDRG neuron subtype labels indicates their fiber type.