Initial efforts of translational development of AAV-encoded Na_ViPA1 for peripherally targeted analgesia in neuropathic pain

Abstract

We previously reported the discovery of a multipronged sodium channel-inhibitory peptide aptamer (Na_ViPA1), derived from the intracellular disordered sequence conserved among tetrodotoxin-sensitive (TTXs) sodium channels (Na_Vs), for adeno-associated virus (AAV)-mediated sensory neuron-targeted analgesia in rat models of chronic pain. In this study, an initial translational effort was undertaken, including (1) removing GFP from the AAV transgene cassette and enhancing small peptide Na_ViPA1 molecular engagement by constructing a concatemeric Na_ViPA1 (CoNa_ViPA1) containing two copies of consensus Na_ViPA1 as a tandem repeat and (2) evaluating long-lasting preclinical analgesic effectiveness in rats by injecting AAV-CoNa_ViPA1 into the sciatic nerve in a rat model of tibial nerve injury-induced painful neuropathy. Results showed analgesic efficacy that persisted for 3–4 months of observation, with magnitudes and time courses comparable between male and female rats. AAV genome analysis confirmed the peripheral nerve-restricted biodistribution of AAV-CoNa_ViPA1 and long-term transgene expression. Inhibition of TTXs Na_Vs and suppression of action potential firing by CoNa_ViPA1 were demonstrated in human sensory ganglion neurons, underscoring the translational potential of AAV-mediated CoNa_ViPA1 expression in the peripheral sensory pathway for treating intractable chronic neuropathic pain conditions resistant to current therapies.

Graphical abstract

Delivery of AAV6.2FF-CoNa_ViPA1 into the sciatic nerve induces prolonged pain relief in rats with neuropathic pain. CoNa_ViPA1 inhibits TTXs I_Na and action potential firing in human DRG neurons, supporting the translational potential.

Introduction

Chronic pain is a devastating problem.¹ Peripheral sensory neuron (PSN) hyperexcitability, commonly induced by nerve injury in the peripheral nervous system (PNS), is sufficient and necessary to initiate and maintain neuropathic pain.² Clinical chronic pain in almost all cases is driven by ongoing afferent hyperactivity originating from peripheral pathological sources.^3,4 It often worsens with time, leading to intractable pain conditions that significantly diminish the quality of life.⁵ Therefore, treatments targeting PSNs for patients with intractable pain conditions that arise and are limited to these specific areas of the body can avoid central nervous system (CNS) and cardiac adverse effects and are most likely to succeed.⁶

Sodium channels (Na_Vs) play fundamental roles in sensory processing, particularly in neuronal excitability, sensory signal transmission, and pain regulation. Multiple Na_Vs in the PSNs of dorsal root ganglia (DRGs) contribute to pain electrogenesis after nerve injury.⁷ This suggests that developing an approach to block multiple Na_Vs restricted to PSNs would produce more robust therapeutic outcomes than blocking only a single Na_V subtype. We recently reported that targeting the intrinsically disordered regions of Na_Vs facilitates the identification of an inhibitory peptide aptamer (Na_ViPA1) that blocks multipronged tetrodotoxin-sensitive (TTXs) Na_Vs. Adeno-associated virus (AAV)-mediated PSN-targeted expression of Na_ViPA1 mitigates pain behaviors in rat chronic pain models.^8,9 Translation of this discovery into an effective analgesic approach could dramatically enhance addiction-free treatment options for chronic pain.

Many challenges, as well as opportunities, remain in advancing this project. Specifically, further research is needed to determine whether AAV-delivered Na_ViPA1 can be expressed persistently in the peripheral sensory pathway, thereby providing prolonged analgesia. Additionally, the AAV construct in our original studies^8,9 encodes a small peptide, Na_ViPA1, fused to scaffold GFP for in vivo tracking, which is not appropriate for human use. The AAV gene therapeutic product needs to be optimized to remove GFP for suitable clinical use. Importantly, biological differences in sensory processing between humans and model organisms may pose significant obstacles to translational approaches in clinical chronic pain treatment¹⁰; therefore, the inhibition of TTXs-Na_Vs by Na_ViPA1 needs to be confirmed in human sensory ganglion neurons (h-SNs). These challenges must be addressed as we pursue the opportunity to advance this peripherally targeted analgesic gene therapy approach for clinical pain therapy.

The sciatic nerve block is used clinically for analgesia.¹¹ AAV injection into the sciatic nerve is a promising strategy for delivering therapeutic biologics to PSNs.^12,13 Accumulation and altered expression of various Na_Vs in peripheral afferents and in painful neuromas facilitate nociceptive signal generation and propagation after neuropathic injury.^14,15 In this report, we generated an AAV-encoded CoNa_ViPA1 that combines two copies of consensus human Na_ViPA1 in tandem repeat, with GFP removed. We extended the treatment duration to 3–4 months after AAV-CoNa_ViPA1 delivery into the sciatic nerve in male and female rats with tibial nerve injury (TNI)-induced painful neuropathy and assessed analgesic effects on both stimulated and spontaneous pain behaviors. AAV-CoNa_ViPA1 biodistribution in vivo along the peripheral sensory pathway and in off-target tissues, as well as long-term transgene expression, was analyzed by quantitative PCR (qPCR) of the AAV genome. CoNa_ViPA1’s effects on selective inhibition of TTXs-Na_Vs and action potential (AP) generation were determined in h-SNs to assess translational potential.

Results

Translational therapeutic AAV gene product design

In this study, a concatemeric Na_ViPA1 (CoNa_ViPA1) was constructed by combining two copies of the consensus human Na_ViPA1 (54-mer) sequence, which is expected to enhance target engagement since Na_ViPA1 directly binds to full-length Na_V1.7,⁸ and GFP was removed from the construct. This design of a consensus tandem repeat of a small peptide aptamer has achieved several successes in stabilizing the expression of small peptides without a scaffold protein while retaining biological activity.^16,17 Triple glycine (G) was inserted between the two Na_ViPA1 copies for stabilizing the complex,¹⁸ and the CoNa_ViPA1 was transcribed by a hybrid CMV enhancer/chicken β-actin (CBA) promoter. Immunoblot analysis of HEK293 cell lysates using an Myc antibody, 5 days after transfection of a plasmid expressing Myc-tagged CoNa_ViPA1 into HEK293 cells, revealed stable expression of MycCoNa_ViPA1. Two copies of consensus sequences of Na_V1.7NP (CoNP), which is a non-functional peptide from the Na_V1.7 N terminus,⁸ were constructed as the control. Whole-cell patch-clamp recording of human embryonic kidney (HEK) cells stably expressing human Na_V1.7 cells (HEK1.7 cells), 4–6 days after co-transfection of the plasmid encoding CoNa_ViPA1 or CoNP with a plasmid coding GFP, confirmed that CoNa_ViPA1 inhibited Na_V1.7 current by >50% relative to naive HEK1.7 cells, while no effect was observed in the presence of CoNP (Figure 1). AAVs were packaged into serotype 6.2FF, a triple AAV6 variant containing F129L, Y445F, and Y731F mutations, which exhibits more rapid and potent in vivo transduction and enhanced immune evasion than AAV6.^8,19 Hereafter, we used “CoNa_ViPA1” as the treatment vector and “CoNP” as the control, respectively.

Figure 1.

Therapeutic AAV gene product design and testing

(A) Alignment of human and rat Na_ViPA1 sequences. (B) A map illustrating AAV expression plasmid encoding CoNa_ViPA1, with the AAV transgene cassette shown at the top of the map. (C) Concatemeric Na_ViPA1 (CoNa_ViPA1) and concatemeric NP (CoNP), with triple glycine (G) inserted between each peptide. (D) Immunoblot (lysates) of Myc-tagged CoNa_ViPA1 using an Myc-Tag antibody (SCB sc-57592), 5 days after transfection of a plasmid expressing Myc-CoNa_ViPA1 into HEK239 cells. (E) I_Na1.7 traces recorded from HEK1.7 cells of naive, CoNP, and CoNa_ViPA1, respectively. Inset: current/timescales and pulse protocol. (F–I) Comparison of I/V curves (F), peak I_Na1.7 density (G), steady-state activation curves (inset: half activation) (H), and fast inactivation curves (inset: half inactivation) of naive, CoNP, and CoNa_ViPA1 (I). ∗∗∗p < 0.001, one-way ANOVA and Tukey’s post hoc.

Long-lasting analgesia after DRG treatment by AAV6.2FF-CoNa_ViPA1

Experimental design

Since sex differences exist in experimental and clinical pain, as well as in responsiveness to treatment,²⁰ this study examined pain hypersensitivity and its reversal response to treatment in both male and female TNI rats. In the behavioral experimental design, sensitivity to mechanical and thermal hindpaw cutaneous stimulation was assessed at baseline (BL) and weekly thereafter for 2 weeks after TNI. Rats were then randomized to receive intra-sciatic nerve injection of either AAV6.2FF-CoNa_ViPA1 or AAV6.2FF-CoNP (Figure 2A) into the sciatic nerve ipsilateral to TNI, after which evoked sensory behaviors were evaluated for 3 months (12 weeks) (Figure 2B). Behaviors measured before vector injection on the 14^th day after TNI, a time point at which hypersensitization is fully established,^8,21 were used as the treatment baseline (tBL) to evaluate the effectiveness and magnitude of hypersensitivity reduction after AAV treatment. As a terminal experiment, gabapentin (GBP, 100 mg/kg, intraperitoneal [i.p.])-induced conditioned place preference (CPP) was performed in both groups at 12–14 weeks to evaluate spontaneous pain, as previously described.^8,21 After the behavior evaluation, animals were euthanized, and multiple tissues were harvested for AAV genome analysis by quantitative polymerase chain reaction (qPCR).

Figure 2.

Treatment of established TNI pain by DRG delivery of AAV6.2FF-CoNa_ViPA (male rats)

(A and B) Silver stain of purified AAVs (A) used in the animal protocol, schematically outlined (B). (C–F) Time courses and (C1–F1) percentage of pain reduction for vF, pin, heat, and cold after DRG injection of either AAV6.2FF-CoNa_ViPA (n = 8) or AAV6.2FF-CoNP (control, n = 8). ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001 for comparisons to the tBL within the CoNa_ViPA1 group, and #p < 0.05, ##p < 0.01, and ###p < 0.001 between the CoNP and CoNa_ViPA1 groups. Repeated-measures two-way ANOVA for vF and heat tests, with Tukey’s (within Co3.2iPA group) and Bonferroni’s (between CoNP and CoNa_ViPA1 groups) post hoc; and nonparametric Friedman ANOVA for pin and cold tests, with Dunn’s post hoc. Dashed lines in (C1)–(F1) denote the absence of hypersensitivity. Average pain relief during a 3-month treatment course was 52%, 49%, 69%, and 67% reductions in vF-, pin-, cold-, and heat-stimulated mechanical and thermal pain behaviors, respectively. ∗∗p < 0.01, ∗∗∗p < 0.001, unpaired, two-tailed Student’s t test.

Analgesia of treatment in male TNI rats

Results showed that all tested rats developed multiple modalities of evoked pain behaviors 2 weeks after TNI, including hypersensitivity to mild mechanical stimuli (von Frey testing, vF), more frequent hyperalgesia-type responses (sustained lifting, shaking, and grooming) after noxious mechanical stimulation (pin testing), and hypersensitivity to heat and cooling (acetone) stimulation. These behaviors persisted after injection of the control AAV6.2FF-CoNP during the observation course (3 months). In contrast, rats injected with AAV6.2FF-CoNa_ViPA1 (4.00E+11 GC virions per rat) showed a reversal of these TNI-induced hypersensitivities, which persisted throughout the observation period and potentially lasted longer (Figures 2C–2F). A typical requirement for a “minimum clinically important difference” in analgesic trials is ≥30% pain reduction.^22,23,24,25 Since the pain hypersensitivity in multiple modalities was fully established 2 weeks after TNI as we reported in the previous reports,^8,21,26 we specified the measures on the 14^th day after TNI and before AAV treatment (tBL) as the peak pain intensity (100%), and the measures of each sensory modality after treatment were normalized to the measures at the tBL and presented as the percentage of pain relief for each modality at each time point thereafter. The average pain relief during the 3-month treatment course showed reductions of 46%, 46%, 79%, and 49% in vF-, pin-, cold-, and heat-stimulated mechanical and thermal pain behaviors, respectively (Figures 2C1–2F1).

Analgesia of treatment in female TNI rats

A treatment protocol in female TNI rats was conducted similarly to the tests in male animals, as described in Figure 2. Results showed that female rats displayed phenotypic hypersensitivity development comparable to that of male rats after TNI induction and that the evoked mechanical/thermal hypersensitivity was normalized over 3 months of observation after AAV6.2FF-CoNa_ViPA1 treatment, showing analgesic effects comparable to those of male animals (Figures 3A–3D). The average reversal of TNI-induced hypersensitivity on vF-, pin-, cold-, and heat-stimulated mechanical and thermal pain behaviors in the 3-month treatment course was 37%, 42%, 71%, and 78%, respectively (Figures 3A1–3D1).

Figure 3.

Analgesia of DRG-AAV6.2FF-CoNa_ViPA1 treatment in female TNI rats

(A–D) Analogous to Figure 2, these images show the time course of analgesia (A–D) and the percentage of pain reduction (A1–D1) after DRG delivery of AAV6.2FF-CoNa_ViPA1 in TNI-induced hypersensitivity in female rats. ∗, ∗∗, and ∗∗∗ denote p < 0.05, 0.01, and 0.001 for comparisons to the treatment baseline (tBL) within the CoNa_ViPA1 group, and #p < 0.05, ##p < 0.01, and ###p < 0.001 for comparisons between CoNP and CoNa_ViPA1 groups. Repeated-measures parametric two-way ANOVA for vF and heat tests, followed by Tukey’s (within the CoNa_ViPA1 group) and Bonferroni’s (between groups) post hoc; and nonparametric Friedman ANOVA for pin and cold tests, with Dunn’s post hoc. Dashed lines in (A1)–(D1) denote the level of absence of hypersensitivity. Average pain relief for each modality during the 3-month treatment; ∗∗p < 0.01 for comparisons between groups (unpaired, two-tailed Student’s t test).

Treatment relieves ongoing spontaneous pain

Using a biased CPP paradigm,^21,27 the effect of AAV6.2FF-CoNa_ViPA1 treatment on the affective aspects of spontaneous pain was evaluated (at 12–14 weeks after evoked behavior tests) in both male and female rats. None of the animals in the groups was excluded from the study because of their baseline preference/avoidance for a chamber. A significant conditioning effect of GBP administration was observed in the TNI rats injected with AAV6.2FF-CoNP, but there was no significant difference in the time spent in the initially nonpreferred chamber during baseline vs. testing period in TNI animals subjected to AAV6.2FF-Co1.7iPA1, indicating that AAV6.2FF-CoNa_ViPA1 treatment relieves ongoing spontaneous pain in TNI rats (Figure 4). Together, the behavior tests reveal that, although not rigorously compared, a sexual dimorphism does not appear to exist for either pain behavior phenotype after TNI or in responsivity to DRG-AAV6.2FF-CoNa_ViPA1 treatment in this study.

Figure 4.

AAV6.2FF-CoNa_ViPA1 treatment on CPP

(A and B) CPP difference scores (s) for the preconditioning chamber and the GBP-paired chamber between AAV6.2FF-CoNa_ViPA1 (n = 8) and AAV6.2FF-CoNP (control, n = 8) in males (A) and females (B). ∗p < 0.05 and ∗∗∗p < 0.001 (unpaired, two-tailed Student’s t test).

qPCR quantification of AAV in vivo biodistribution

Determining AAV vector in vivo biodistribution via qPCR analysis of the AAV genome in the absence of a staining marker is a standard molecular approach. We performed qPCR of the AAV genome to determine the biodistribution and persistence of AAV6.2FF-CoNa_ViPA1 in vivo. Positive amplification of the AAV genome was defined as samples (naive rats) with copy numbers greater than the average (180 copies) plus 3 standard deviations (99 copies), as determined by comparison with a standard curve of known AAV6.2FF-CoNa_ViPA1 plasmid concentrations (Figure S1). Therefore, the limit of quantification (the lowest number of copies that can be reliably measured) was set at 500 copies as the cutoff value in this study, in the presence of 2 ng of total DNA. Samples of total DNA from tissues, cerebrospinal fluid (CSF), and sera from AAV6.2FF-CoNa_ViPA1-subjected rats (n = 6, tissues collected after 12–14 weeks after GBP tests) were evaluated for the presence of AAV genomes. qPCR analysis of the AAV6.2FF-CoNa_ViPA1 genome in DNA samples extracted from multiple tissues after behavior evaluation in male animals showed that high copies of AAV genomes were detected in the ipsilateral sciatic nerve and hindpaw glabrous skin, and positive AAV genomes were detected in the L3-L6 DRG and lumbar spinal cords, ipsilateral to AAV6.2FF-CoNa_ViPA1 treatment, 12–14 weeks after behavior evaluation and AAV6.2FF-CoNa_ViPA1 injection. Results confirmed that the TNI pathological peripheral nerve received the AAV, indicating persistent transgene expression. There were weak qPCR signals of the AAV genome below the cutoff of 500 copies in the liver, kidney, heart, blood, and biofluids (blood and CSF) (Figure 5). These results suggest that the AAV6.2FF-CoNa_ViPA1 persists and remains restricted in the injected peripheral nerves in the long term.

Figure 5.

qPCR quantification of AAV in vivo biodistribution

qPCR quantification of AAV6.2FF-CoNa_ViPA1 genome copy numbers and anatomical biodistribution in the multiple tissues (n = 6 rats), harvested 3 months after AAV6.2FF-CoNa_ViPA1 injection into RL4 and RL5 DRG, ipsilateral to TNI.

Lack of apparent pathology in DRG and sciatic nerve after intraneural delivery of AAV6.2FF-CoNa_ViPA1

After 12–14 weeks of intraneuronal delivery of AAV6.2FF-CoNa_ViPA1, the ipsilateral DRG and sciatic nerves of TNI rats were examined microscopically using conventional hematoxylin and eosin (H&E) staining, which showed no observable microscopic pathology compared to tissues from naive animals. Immunohistochemistry (IHC) revealed mild proliferation of GFAP-positive satellite glial cells and Iba1-positive microglia in the DRG from TNI rats compared with naive rats, indicating low-grade DRG neuroinflammation after treatment. No significant changes in myelination, as shown by myelin basic protein (MBP) staining, were observed in the sciatic nerve of TNI rats after treatment compared to naive animals (Figure 6). We further performed IHC for activating transcription factor 3 (ATF3) to assess neurotoxicity, CD6 and CD8 to evaluate T cell immunity, and caspase-3 to assess neuronal apoptosis. Results showed that the neuronal injury marker ATF-3 was comparably positive in the DRG from TNI only, TNI injected with control AAV, and TNI after AAV6.2FF-CoNa_ViPA1, compared with the DRG from naive rats. We did not observe any positive cells labeled with CD6, a pan-T cell marker, or CD8, a cytotoxic T cell marker, compared with lymph node tissue used as a positive control. Caspase-3 immunostaining was negative in the DRG and sciatic nerve after AAV6.2FF-CoNa_ViPA1 (Figure S2). Overall, the IHC data show a mild inflammatory response without apparent IHC evidence of neurotoxicity or cellular immune activation 3 months after intraneuronal delivery of AAV6.2FF-CoNa_ViPA1. We will continue to investigate the safety of long-term treatment, e.g., by determining innate immunity responses that lack these surface markers after AAV in vivo administration,²⁸ which is crucial for strengthening the translational relevance of this approach and supporting future clinical development for chronic neuropathic pain.

Figure 6.

Lack of apparent pathology in the DRG and sciatic nerve

(A–J) Illustration of the representative H&E stain and IHC immunostaining images of the DRG and sciatic nerve sections from naive rat and TNI rat 3 months after intrasciatic nerve delivery of AAV6.2FF-CoNa_ViPA1. H&E stain images of the DRG and sciatic nerve sections from naive rat (A and B) and TNI rat (C and D). IHC for GFAP and costaining with Tubb3 in DRG sections from naive rat (E and E1) and TNI rat (F and F1), and arrowheads point to GFAP-positive satellite glial cells. IHC of Iba1 and costaining with Tubb3 of the DRG sections from naive rat (G and G1) and TNI rat (H and H1), and arrowheads point to Iba1-positive microglia. IHC of MBP and costaining with Tubb3 of the sciatic nerve sections from naive rat (I and I1) and TNI rat (J and J1). Scale bars: 200 μm for (A)–(D), 100 μm for (E), (F), (I), and (J), and 50 μm for (G) and (H).

CoNa_ViPA1 blocks TTXs I_Na in human DRG neurons

We previously showed that Na_ViPA1 is efficacious in human iPSC-derived sensory neurons (hiPSC-SNs)⁸; however, hiPSC-SNs represent a single population of small-sized sensory neurons. In contrast, DRG neurons comprise a variety of sensory neuronal subtypes, including low-threshold mechanoreceptors (LTMRs) and high-threshold mechanoreceptors (HTMRs), which exhibit distinct biological and electrophysiological (EP) properties.²⁹ We therefore further evaluated CoNa_ViPA1 in blocking Na_Vs and suppressing excitability in h-SNs. We first produced a high-titer lentiviral vector (LV)-encoded CoNa_ViPA1 or CoNP fused to the C terminus of GFP (Figure S3). The dissociated h-SNs from three male donors (17, 37, and 35 years old, respectively, without chronic pain and drug use, deceased due to head trauma) were utilized (Anabios, San Diego, CA). Whole-cell voltage- and current-clamp recordings were performed on the h-SNs of naive, GFPCoNP, and GFPCoNa_ViPA1, 3–5 days in vitro (DIV) after LV transduction. Results showed that total and TTXs I_Nas, but not TTXr I_Nas, in the h-SNs expressing GFPCoNa_ViPA1 were significantly reduced, and AP firing rates were suppressed, compared with the h-SNs of naive and GFPCoNP expression (Figure 7). The results are comparable to those from our previously collected data using hiPSC-SNs.⁸ Further studies will investigate the sex-dimorphic effects of CoNa_ViPA1 expression on the biophysical features of h-SNs.

Figure 7.

CoNa_ViPA1 blocks TTXs I_Nas in human DRG neurons

(A–C) Images from top to bottom illustrate representative traces and averaged peak I_Na densities of total I_Na (A), TTXs I_Na (B), and TTXr I_Na (C) recorded from h-SNs (males) of naive, CoNP, and CoNa_ViPA1, respectively. Insets: representative images of h-SNs expressing CoNP and CoNa_ViPA1, current/timescales, and pulse protocol. (D) Representative AP traces elicited by 500 ms depolarizing current of 1,000 pA from RMP were recorded from h-SNs of naive, CoNP, and CoNa_ViPA1, as indicated. (E) Comparison of responses (number of APs evoked by a 500 ms stimulus) for the populations of h-SNs in different groups across a range of step current injections to 1,000 pA; ∗∗∗p < 0.001, two-way ANOVA of main effects of groups with Bonferroni’s post hoc. (F–H) Scatterplots with bars show analyses of the RMP (F), rheobases (G), and AP numbers evoked by a 1,000 pA input current (H). The number in each group is the number of analyzed neurons. ∗∗p < 0.01 and ∗∗∗p < 0.001, one-way ANOVA and Tukey’s post hoc.

Discussion

Na_Vs in PSNs are crucial mediators of pain and have been targeted for analgesic development. An effective, peripherally targeted, safe analgesic approach targeting multiple PSN-Na_Vs without addiction risk for treating chronic pain will be valuable. The new findings of this study supplement our previous reports8^8,9 and include the following: (1) sciatic nerve delivery of AAV6.2FF-encoded CoNa_ViPA1 elicits persistent expression of CoNa_ViPA1 in sensory neuronal skin terminals, afferents, somata at multiple segmental levels, and spinal cord along the peripheral sensory pathway, which provides prolonged analgesia in relief of neuropathic pain behaviors. The durability of AAV6.2FF-CoNa_ViPA1 in vivo encourages its clinical development.^30,31 (2) CoNa_ViPA1 expression results in TTXs Na_Vs inhibition and suppression of AP firing of the human DRG neurons, underscoring translational potential. (3) Although the AAV capsid-specific humoral immunity was not examined in this study, persistent CoNa_ViPA1 expression and restrictive AAV genome in the peripheral sensory pathway suggest tolerable immunogenicity against the gene therapy product of AAV6.2FF-CoNa_ViPA1. It is reported that the engineered AAV6.2FF vector exhibits relative immune “silence” with reduced immune recognition compared to its parent AAV6, primarily by bypassing specific intracellular degradation pathways. AAV6.2FF displays greater resistance to pooled immunoglobulin neutralization and mediates long-term transgene expression in vivo.^19,32 Prolonged AAV-mediated transgene expression is observed in quiescent neurons,³³ including those involved in AAV pain therapy.³⁴ Most clinical immunotoxicities observed with gene therapies have emerged as higher doses of AAV vectors have been used.^28,35 These findings suggest that AAV6.2FF-CoNa_ViPA1 doses and peripherally restricted expression in this study would not elicit a significant immune response, thereby preventing loss of transgene expression.^30,36

Various routes of AAV delivery are used in preclinical studies to target PSNs and their afferents, including direct DRG microinjection (intraganglionic, IG), via the spinal fluid (intrathecal, IT), and intrasciatic injection. An increasing number of small- and large-animal preclinical studies are exploring the IG, IT, and sciatic nerve as the delivery routes of AAV biologics for pain control. We and others report that IG delivery of AAV using glass micropipettes provides the most efficient in vivo PSN transduction and induces expression of analgesic biologics in PSNs and their central and terminal processes restricted to the injected DRG, without significant DRG toxicity.^8,21,37 Delivery of AAV-encoded analgesic biologics to target and modulate PSN hyperexcitability for sustained pain relief is a promising approach to pain gene therapy. Currently, AAV-IG therapies are in early clinical stages or preclinical development, focusing on precise, localized gene expression to manage severe pain. A computed tomography (CT)-guided IG injection of an analgesic agent has been reported to be successful in treating patients with intractable phantom limb pain³⁸ and postherpetic neuralgia.³⁹ We expect that the standard procedure for clinical direct DRG injection will be fully established for PSN-targeted pain gene therapy. IT injection of AAV via an IT pump or lumbar puncture is an alternative route that enables effective targeting of AAV to PSNs across multiple segmental levels bilaterally. However, a higher AAV dose is required than that for the IG injection.⁴⁰ IT injection is a more suitable method for labeling neurons in the PNS and CNS of widespread neurodegenerative models.⁴¹ When targeting DRG innervation, IT injection reduces specificity due to AAV’s broad transduction and its potential to spread into the CNS, potentially causing off-target adverse effects.⁴¹ Additionally, retrograde axonal transport of AAV to DRG via vector delivery into peripheral terminal sites in adult rodents has been tested, but AAVs struggle to efficiently reach DRG neurons via peripheral terminal delivery.^42,43,44

The ultrasound-guided sciatic nerve block for analgesia is an established approach used in clinical practice, and the procedure can be repeated.^45,46 We and others have optimized a subepineural sciatic nerve injection technique in rodents that induces minor, transient hypersensitization comparable to that induced by DRG injection.^12,47 The minimally traumatic sciatic nerve injection implies that an intrafascicular injection that induces sciatic nerve injection injury (SNII)⁴⁸ is a preventable event in experimental rats using our microinjection technique. Although the efficiency of AAV-encoded transgene expression in DRG neurons is lower than that of AAV-DRG injection,^9,37 one-time intrasciatic AAV injection induces transgene expression in the sciatic nerve, multiple lumbar DRG neurons, their afferents, and central and peripheral terminals,⁴⁹ providing clear analgesic effects in mitigating established pain behaviors.

Animal and human studies found that multiple Na_V subtypes accumulated in the painful neuroma following peripheral nerve injury, causing ectopic AP firing and refractory, debilitating chronic pain.^14,50 Thus, AAV-mediated sciatic nerve blockade of multiple Na_Vs is well applicable to nerve injury-induced neuroma pain by blocking ectopic nociceptive input emanating from injured sensory afferent axons that causes or maintains the chronic pain state. Our results indicate that sciatic nerve delivery of AAV6.2FF-encoded CoNa_ViPA1 to inhibit multiple TTXs-Na_Vs in peripheral sensory pathways could be developed as a clinical approach to control intractable pain disorders. The wide PNS-restricted distribution of AAV6.2FF-CoNa_ViPA1 may be a therapeutic advantage because injury-induced peripheral hypersensitization associated with Na_V malfunction affects multiple sites of the peripheral sensory nervous system, including augmented pain perception in the peripheral terminals, enhanced nociceptive signal transduction in PSN somata and T-junction, and increased neurotransmission in the spinal dorsal horn.

There are limitations to the work presented. First, although we posited that the long-term persistence of the in vivo AAV genome suggests reduced immunogenicity of intrasciatic nerve injection with AAV6.2FF-CoNa_ViPA1, this speculation needs to be tested experimentally in future investigations. Second, intrasciatic delivery of AAV can be retrogradely transported to spinal cord motor neurons^9,51 that express Na_Vs, such as Nav1.7 and Nav1.6,⁵² and intrasciatic delivery of an AAV incorporating the CBA promoter will transduce Schwann cells that weakly express various Nav subtypes.⁵³ These findings suggest that the block of TTXs Na_Vs in spinal cord motor neurons and Schwann cells may partially contribute to pain reduction by intrasciatic AAV6.2FF-CoNa_ViPA1. Third, the AAV genome was detected at marginal levels in the liver, kidney, heart, and biofluids (blood and CSF [cerebrospinal fluid]) in several animals. Whether this was due to minimal leakage of virions into the bloodstream after intraneural delivery and subsequent sequestration in remote tissues remains to be investigated. Based on previous biophysical studies demonstrating the “fuzzy” binding properties of intrinsically disordered peptides with their partner proteins, we hypothesize that tandem repeats may enhance target engagement. This hypothesis remains speculative at present and will be experimentally addressed in future investigations. Furthermore, the biophysical features, key amino acids associated with phosphoinositides (PIP) binding and the biological properties of CoNa_ViPA1, and the CoNa_ViPA1 binding site in full-length Nav1.7 need to be delineated and compared with those of single Na_ViPA1. Finally, the suitable AAV dose for in vivo delivery in this study was determined using a dose-response test (see materials and methods); however, intrasciatic delivery of 20 μL of AAV6.2FF-CoNa_ViPA1 can be reduced by preparing high-titer (1 × 10¹⁴–1 × 10¹⁵ GC/mL) fully packed AAV in our future studies.

Materials and methods

Animals

Adult male Sprague-Dawley (SD) rats weighing 100–125 g (Charles River Laboratories, Wilmington, MA) were used. Animals were housed individually in a room maintained at constant temperature (22°C ± 0.5°C) and relative humidity (60% ± 15%) with an alternating 12 h light-dark cycle. Animals had access to water and food ad libitum throughout the experiment, and all efforts were made to minimize suffering. All survival surgeries were completed in a sterile environment under a surgical microscope in animals anesthetized with isoflurane (2%–5%). For tissue harvest and euthanasia, animals were deeply anesthetized with isoflurane followed by decapitation with a well-maintained guillotine. The estimated number of animals needed was derived from our previous experience with similar experiments,^54,55 and no power analysis was performed. The number of rats used was detailed in the relevant sections or figure legends of the experiments.

Molecular cloning and viral vector constructs

AAV construct: an AAV vector encoding a CoNa_ViPA1 that combines two copies of consensus Na_ViPA1 (54-mer) sequences with triple glycine inserted between each Na_ViPA1 as a spacer,⁵⁶ transcribed by a hybrid cytomegalovirus (CMV) enhancer/CBA promoter, was constructed, and AAV-CoNP was constructed as a control. Specifically, the DNA sequences of CoNa_ViPA1 and CoNP were synthesized and subcloned into a single-strand AAV expression plasmid pAAV-CBA-GFP (GFP was removed) by GenScript (Piscataway, NJ). This generated pAAV-CBA-CoNa_ViPA1 and pAAV-CBA-CoNP. The aforementioned plasmids were subsequently used to produce AAVs packaged in the AAV6.2FF capsid by Packgene (Worcester, MA) to generate purified AAV-CBA-CoNa_ViPA1 (2 × 10¹³ GC/mL) and AAV-CBA-CoNP (2.2 × 10¹³ GC/mL).

Generation of lentivectors expressing GFPCoNa_ViPA1 and GFPNP for h-SN transduction and an AAV plasmid expressing Myc-tagged Na_ViPA1: lentiviral vector (LV) expression plasmids of pWPT-GFPCoNa_ViPA1 and pWPT-GFPCoNP (control) were generated by GenScript (Figure S3). LVs were packaged using pWPT-GFPCoNa_ViPA1 and pWPT-GFPCoNP with packaging plasmid pCMVdR8.74 and envelope plasmid pVSV-g, concentrated, and were titrated in the range of 1 × 10⁸ to 2 × 10⁸ transduction units/mL, as described previously.²¹ An AAV expression plasmid encoding Myc-tagged Na_ViPA, driven by a CBA promoter, was constructed (Figure S4).

Cell culture and transfection/transduction

HEK cells stably expressing human Na_V1.7 (HEK1.7 cells) were cultured by a standard protocol using a Dulbecco’s modified Eagle’s medium (DMED) supplemented with 10% fetal bovine serum (FBS) and antibiotics (Thermo Fisher, Milwaukee, WI) and were grown at 37°C and in 5% CO₂ in a humidified incubator.⁸ Polyethyleneimine 40 (PEI40)-mediated co-transfection of the plasmids pAAV-CBA-CoNa_ViPA1 or pAAV-CBA-NP with a plasmid encoding GFP (pAAV-GFP) into HEK1.7 cells was performed as described previously.^8,21

The dissociated h-SNs from three donors (17-, 37-, and 35-year-old males without chronic pain or a history of drug use who died from head trauma) were obtained from Anabios (San Diego, CA) and cultured in the vendor-defined growth-primed medium.^57,58 LV-GFPCoNa_ViPA1 or LV-GFPCoNP was transduced at an optimized multiplicity of infection MOI ≈ 5, immediately after plating h-SNs in culture. GFP expression was normally visualized 48–72 h after transduction.

LV-CoNa_ViPA1-transduced h-SNs with a diameter of 50–100 μm and clear long neurite outgrowth were selected to record I_Nas and AP generation at DIV5, and the peak current densities and AP spike frequency were compared.

EP recordings

EP recordings were performed, as we described previously, at room temperature (22°C–25°C), in a blinded manner in which the electrophysiologist was not aware of the treatment. Patch pipettes with <2 MΩ resistance were made from borosilicate glass (King Precision Glass Co., Claremont, CA) and fire-polished. Recordings were made with an Axopatch 700B amplifier (Molecular Devices, Downingtown, PA). For I_Na recordings, signals were filtered at 5 kHz and sampled at 20 or 50 kHz^{59,60,61,62,63,64,65} with a Digidata 1440A digitizer and pClamp10 software (Molecular Devices, San Jose, CA). After achieving whole-cell recording, capacitance (Cm) and series resistance (Rs) were compensated. Voltage errors defined by Rs × Imax were minimized by using 80%–85% series resistance compensation, and the compensation was readjusted before each voltage-clamp protocol. Cells were excluded if the voltage error exceeded 5 mV.

I_Na1.7 recording on HEK1.7 cells: HEK1.7 cells at 3–4 DIV after co-transfection of pAAV-CBA-CoNa_ViPA1 or pAAV-CBA-NP with pAAV-GFP were tested. Whole-cell voltage-clamp recordings of Nav1.7 current (I_Na1.7) were performed using a similar pulse protocol, external/internal pipette solution, and recording conditions, as previously described.⁸ In brief, to determine the current-voltage (I-V) relationships of voltage-dependent activation, the peak current densities during each voltage command step were fitted to a smooth curve with a Boltzmann equation G/G_max = 1/(1+exp(V₅₀−Vm)/k), where G was calculated as follows: G = I/(Vm−Erev). The steady-state inactivation curves were fitted with I/Imax = 1/(1+exp−(V₅₀-Vm)/k). In all the equations, V₅₀ denotes the half-activation and half-inactivation potentials, Vm is the membrane potential, Erev is the reversal potential, k is the slope factor, G is the conductance, and I is the current at a given Vm; G_max and I_max are the maximum conductance and current, respectively. I_Na1.7 density was calculated by normalizing maximal peak currents with cell capacitance.

TTXs/TTXr I_Nas and AP recordings on h-SNs: EP recordings with identical bath and internal pipette solution as previously described⁸ were performed on h-SNs with a diameter of 50–100 μm and clear neurite outgrowth at 4–7 DIV in naive h-SNs and h-SNs transduced with LV-GFPCoNa_ViPA1 or LV-GFPCoNP. A voltage protocol was adopted to separate TTXr I_Nas and TTXs I_Na. In brief, a 500 ms prepulse to −120 or −50 mV was applied before a 50 ms test pulse from −100 to 40 mV, with steps of 5 mV, followed by test pulses from −50 to 0 mV. To isolate somatic I_Na, a brief prepulse to voltage (−40 mV) near spikes inactivating h-SN axonal spikes but not somatic spikes was performed, as described previously. APs were generated by injection of a series of current pulses (0–1,000 pA in steps of 100 pA, 500 ms). The baseline potentials were recorded for 20 ms before the stimulus pulses were injected into the neurons. Resting membrane potential (RMP) was defined as the mean value of the 20 ms pre-stimulus potential in the first trial and the AP rheobase as the minimum current required to evoke the first AP. Neurons with stable RMPs more negative than −40 mV and overshooting APs (>80 mV from RMP to peak) were used for additional data collection. AP frequency was determined by quantifying the number of APs elicited in response to depolarizing current injections (500 ms).

Animal pain model and behavior testing

TNI: animals were anesthetized using isoflurane at 4% for induction and 2% for maintenance. Under anesthesia, the right sciatic nerve was exposed under aseptic surgical conditions by blunt dissection of the femoral biceps muscle. The sciatic nerve and its three branches (sural, common peroneal, and tibial nerves) were isolated. For TNI, the tibial nerve was then tightly ligated and transected distal to the ligation.⁵⁵ The overlying muscle and skin were then sutured following surgery. Sham-operated rats were subjected to all preceding procedures without nerve ligation or transection.

Evoked sensory behavior testing: behavioral tests were conducted between 9:00 a.m. and 12:00 p.m. Animals were habituated in individual test compartments for at least 1 h before each test. Stimuli, including nonnoxious mechanical allodynia (vF), noxious mechanical hyperalgesia (pin test), cooling stimulation (cold), and the heating plantar test (heat, temperature-regulated glass platform heated to 30°C), were applied to the hindpaw plantar skin in the tibial nerve innervating area and were carried out by personnel blind to treatments, as previously described.⁶⁶ CPP: gabapentin (GBP, 100 mg/kg, i.p.)-paired CPP (GBP-CPP) was performed after 12 weeks of evoked behavior evaluation after DRG injection of either AAV6.2FF-CoNa_ViPA1 or AAV6.2FF-CoNP in a 3-chamber CPP apparatus (Med Associates, St. Albans, VT) for both male and female rats, as previously described.⁶⁷ The CPP procedure consisted of the following phases. (1) On the preconditioning day, rats were allowed to explore both sides of chambers for 15 min and the time spent in each side was recorded, and the preferred and nonpreferred chambers were identified. Animals that showed a predetermined level of preference for one chamber (≥70% of total time) at this stage were excluded for further study. (2) On the conditioning days, place conditioning was conducted using a biased assignment approach to drug pairing: saline was paired with the preferred chamber in the morning, and GBP was paired with the non-preferred chamber in the afternoon (injections were never paired with the middle gray chamber). Conditioning consisted of the following sequential steps: i.p. injection and restriction of the animal within the non-preferred chamber for 45 min. We used a 45 min conditioning time based on tests showing that GBP maximally reduced mechanical hypersensitivity at 30–60 min after injection. Animals were conditioned for 2 days because 2-day GBP has been reported sufficient to produce CPP in rodent pain models.^68,69 (3) For postconditioning testing, the animals were placed back into the middle gray chamber of the CPP device with free access to all chambers for 15 min. The difference score for each animal was calculated by subtracting the time spent in the saline-paired or GBP-paired chamber before pairing (during preconditioning) from the time spent in each chamber after pairing (postconditioning) and then averaged within each group. Each rat had only a single CPP test 6 weeks after AAV injection. Preference was indicated by a significant increase in the total time spent in the conditioning chamber relative to baseline, as previously described.^8,9,21,27

AAV sciatic nerve injection

AAV was injected into the sciatic nerve using a procedure similar to that described previously, with a microprocessor-controlled injector (Nanoliter 2000, World Precision Instruments, Sarasota, FL, USA),⁴⁷ and comparable to a protocol for sciatic nerve injection in mice.¹² Briefly, after appropriate anesthesia was obtained by inhalation of 2% isoflurane, the right sciatic nerve was exposed through a lateral incision of the middle thigh and division of the superficial fascia and muscle, and the sciatic nerve was exposed at a point proximal to the bifurcation. Rats received a sciatic nerve injection of therapeutic or control AAV at a dose of 4 × 10¹¹ GC (20 μL) containing 0.1% Fast Green (0.1 μL) in the viral vector solution to visualize the injected solution. The dose was determined according to the results of a brief dose-ranging test using 1 × 10¹¹ GC/5 μL, 2 × 10¹¹ GC/10 μL, and 4 × 10¹¹ GC/20 μL of AAV6.2FF-CoNa_ViPA (two rats for each dose) in the pilot study, for which we found less efficacy of transgene expression at a lower dose, determined by qPCR of the AAV genome from DNA extracted from the ipsilateral sciatic nerve 3 weeks post-injection (not shown). AAV was injected directly into the subepineural space (beneath the clear fascia surrounding the nerve but outside the perineurium) with a pulled glass capillary tip (40–60 μm diameter) inserted into the sciatic nerve (∼10 mm), forming an angle with the longitudinal axis of the sciatic nerve over a 5-min period using a microprocessor-controlled injection system employing direct piston displacement mounted on a micromanipulator. Once penetration was achieved, the injector was backed off until tissue compression was no longer evident, reducing tissue pressure on the pipette aperture. Removal of the glass pipette was delayed for an additional 5 min at a rate of 0.5 μL/min to minimize extrusion of the injectate. Following injection and closure of the overlying muscle and skin, the animals were returned to their housing, where they remained for the duration of the designed experiments.

CSF, blood, and tissue harvest

Only terminal blood, CSF, and tissues were taken from the animals at the endpoint for analysis. CSF was collected from the animals, as we described previously.⁷⁰ In brief, under isoflurane anesthesia, an incision was made on the back of the neck, and the muscles were separated by blunt dissection. The spinal dura at the atlanto-occipital level was exposed, the cisterna magna was entered with a 27-G needle, and an average of 20 μL of CSF was collected. Animals were euthanized by decapitation for whole blood and tissue harvest after CSF procedure. Tissues harvested included ipsilateral L3-L6 DRG, sciatic nerve, lumbar spinal cord, liver, kidney, and heart. After collecting the whole blood, it was allowed to clot by leaving it undisturbed at room temperature for 10 min. Serum was collected after clot removal by centrifuging at 2,000 × g for 10 min in a refrigerated centrifuge. CSF, serum, and tissue samples were frozen and stored at −80°C until use.

qPCR quantification of AAV biodistribution in vivo

Genomic DNA (gDNA) from multiple tissues was extracted using QIAamp DNA Kits (QIAGEN, Germantown, MD). Because the yield of gDNA from biofluids was very low, serum (2 μL) and CSF (5 μL) were used directly for qPCR. The primers spanning the CBA promoter and chimeric intron sequence within the pAAV6.2FF-CoNa_ViPA1 plasmid (forward: 5′-GGCGCGCTCCGAAAGTTTCC-3′ and reverse: 5′-CCGCTCACCTGTGGGAGTAACG-3′) were designed using MacVector (MacVector version 17.0.8, Apex NC) and synthesized by Integrated DNA Technologies (idtdna, Coralville, Iowa). A standard curve was generated for qPCR run by serial 10-fold dilutions of the pAAV6.2FF-CoNa_ViPA1 plasmid in TE buffer from 2.4 × 10⁹ to 240 copies/μL, and plasmid standards were mixed with 10 ng/μL gDNA extracted from DRG of naive rats. The number of copies per plasmid standard was calculated using the online resource (https://www.technologynetworks.com/tn). The thermal cycling conditions were 1 cycle at 95°C for 3 min, followed by 40 cycles at 95°C for 10 s, 62°C for 30 s, and 1 cycle at 72°C for 5 min, using 200 nM of each primer in a total volume of 20 μL using IQ SYBR Green supermix (Bio-Rad, Hercules, CA) on a Bio-Rad CFX96 Real-time PCR Machine. Triplicate wells were used for qPCR. The presence of only one product was confirmed by performing a melt cycle. A negative control using double-distilled water was included in each qPCR experiment. The limit of quantitation (LOQ) for each PCR run was set at 240 copies per reaction, the lowest number reliably measurable. Positive amplification of the AAV genome was defined as samples (naive rats) with copy numbers greater than the average (240 copies) plus 2 standard deviations (130 copies). Therefore, the limit of quantification⁷¹ (the lowest number of copies that can be reliably measured) was set at 500 copies as the cutoff value for this study, in the presence of 2 ng of total DNA. Samples of total DNA from tissues, CSF, and sera from AAV6.2FF-CoNa_ViPA-subjected rats (n = 7–8) were evaluated for the presence of AAV genomes.

Histology and IHC

After a 3-month behavior test, DRG and sciatic nerve ipsilateral to TNI and AAV6.2FF-CoNa_ViPA1 or AAV6.2FF-CoNP treatment, as well as those from age-matched naive rats, were harvested (two females). Formalin-fixed, paraffin-embedded tissue sections were prepared, and H&E staining and IHC with antibodies of GFAP (1:500, Dako Z0334), Iba1 (1:500, Wako 019–19741), MBP (1:600, Santa Cruz biotechnology, SCB, sc13912), ATF3 (1:200, SCB, sc188), pan T cell marker CD6 (1:100, Millipore, CBL554), cytotoxic T cell marker CD8 (1:100, Millipore, 217580), caspase-3 (1:100, Proteintech, 82707-13-RR), and Tubb3 (1:1,000, SCB, sc80016; 1:1,000, Proteintech, 66375-1-Ig) were performed, as described previously.⁵⁵ Histological images were captured using a Keyence BZ-X800 microscope (Keyence Corporation, Itasca, IL). IHC images were captured using a Nikon TE2000-S fluorescence microscope (El Segundo, CA) with filters for selective detection of green and red fluorescence and a QuantiFire digital camera (Optronics, Ontario, NY).

Statistical analysis

Statistical analysis was performed with GraphPad PRISM 10 (GraphPad Software, San Diego, CA). Results are reported as means and standard deviations. Differences were significant for values at p < 0.05. The number of biological replicates is provided in the corresponding figures and legends. No data points were excluded. Mechanical allodynia (vF), hyperalgesia (pin), and thermal (heat and cold) changes after sciatic nerve injection were compared to the measures immediately before AAV injection at the 14^th day post-TNI (tBL) with repeated measures of two-way ANOVA and Tukey’s post hoc for vF and heat, and Friedman’s tests and Dunn’s post hoc for pin and cold. AAV treatment data were also presented as a behavior reversal index, a derived value that quantifies the magnitude of treatment-induced reduction by comparing the percentage of pain behavior for each modality to the tBL hypersensitization value (100%) at each test point and over the 3-month treatment. CPP difference scores were compared between vectors by Student’s t test. The differences in the EP experiments were compared using one-way ANOVA and Tukey’s post hoc tests.

Data and code availability

Data for the manuscript materials, including plasmid nucleotide sequences (text) of pAAV-CBA-CoNa_ViPA1 and pWPT-GFPCoNa_ViPA1 with annotations of key components of the constructs, the value data from whole-cell patch-clamp recordings, behavior tests, and qPCR AAV in vivo genome copy, will be deposited and made available to the research community through the public repository Dataverse after the manuscript is accepted for publication.

Acknowledgments

This research was supported by the 2022 and 2024 awards from Dr. Ralph and Marian Falk Medical Research Trust, Bank of America, Private Bank (H.Y.); the National Institutes of Health grants 1R21NS137014-01 (H.Y.), 1R21NS138704-01 (H.Y.), and R33NS116203 (H.Y.); and 2025 MCW Anesthesiology Seed Funding (S.M.S.). The authors thank Dr. Theodore R. Cummins (Indiana University) for providing HEK cells stably expressing human Na_V1.7.

Author contributions

H.Y. and S.M.S. conceived and designed the study and wrote the manuscript. S.M.S., B.I.-Z., S.-K.H., F.F., and H.Y. performed the experiments, analyzed the data, and organized all figures. H.Y. and S.M.S. obtained funding.

Declaration of interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.