An evaluation of distinct adeno-associated virus vector strategies for driving transgene expression in spinal inhibitory neurons of the rat

Abstract

The spinal cord dorsal horn (DH) is essential for processing and transmitting nociceptive information. Its neuronal subpopulations exhibit significant heterogeneity in morphology and intrinsic properties, forming complex circuits that remain only partially understood. Under physiological and pathological conditions, inhibitory interneurons in the DH are of particular interest. These neurons modulate and refine pain-related signals entering the central nervous system. The ability to selectively target these inhibitory interneurons is key to investigating the underlying circuitry and mechanisms of pain processing, as well as to understand the specific role of inhibitory signaling within these processes. We employed a viral vector approach to deliver a fluorescent reporter protein specifically to inhibitory interneurons in the rat spinal cord. Using adeno-associated virus (AAV) vectors designed to express enhanced green fluorescent protein (EGFP) under the control of various promoters, we targeted distinct subtypes of spinal inhibitory interneurons. Through immunostaining, in situ hybridization, and confocal imaging, we evaluated the specificity and efficacy of these promoters. Our findings revealed that the promoter/vector combinations used did not achieve the desired specificity for targeting distinct interneuron populations in the DH. Despite these limitations, this work provides valuable insights into the potential and challenges of designing AAV-based approaches for selective neuronal targeting. These results emphasize the need for further refinement of promoter designs to achieve precise and reliable expression in specific spinal interneuron subtypes. Addressing these challenges will be crucial for advancing our understanding of spinal nociceptive circuits and developing targeted therapeutic approaches for pain syndromes.

Keywords: adeno-associated virus vectors; dorsal horn; inhibitory interneurons; intraparenchymal injection; nociception; rat; spinal cord.

Figure 1

Laminar structure and distribution of inhibitory neurons identified by specific markers in the rat spinal cord DH. (A) A schematic (left) and confocal image (right) illustrate the laminar organization of the DH. Laminae I and II (LI/II) are identified by CGRP⁺ fibers (green), IB4⁺ fibers (magenta), and PKCγ⁺ interneurons (cyan). Lamina III (LIII) is defined by specific anatomical landmarks, while Laminae IV and V (LIV/V) are located caudal to LIII and delimited by anatomical landmarks. The dotted line delineates the boundaries of the spinal DH laminae. (B) Left: Slices were stained with antibodies against NeuN (cyan) and Pax2 (magenta); right: FISH was performed with probes for GlyT2 (cyan) and combined with immunostaining for Pax2 (magenta). (C) Percentage of Pax2⁺ cells among all NeuN⁺ cells across distinct laminae; n = 26 slices from a total of 9 animals. (D) Percentage of GlyT2⁺ cells among all Pax2-expressing cells across distinct laminae; n = 6 slices from 3 animals. All data are shown as mean ± SEM.

Figure 2

AAV-GlyT2-mediated transgene expression was not restricted to inhibitory neurons. (A) Schematic of the AAV constructs showing the insertion of hSLC6A5_2kb (i), or hSLC6A5_3kb (ii), downstream of the AAV 5′ inverted terminal repeat (ITR). Vectors were injected into the spinal cord DH (iii), and viral expression was assessed 15–21 days post-injection. (B) Representative confocal images displaying an overview of AAV-GlyT2-2 (right) and AAV-GlyT2-3 (left), with magnified views of an area in LI/II marked by the white box; EGFP is shown in yellow, NeuN in cyan, and Pax2 in magenta. Arrows indicate co-localized cells. (C, D) Proportion of AAV⁺ cells in each lamina co-expressing either NeuN or Pax2 for AAV-GlyT2-2 (C) and AAV-GlyT2-3 (D). (E,F) Viral transduction efficiency, shown as the percentage of EGFP-expressing cells per mm² within NeuN⁺ (E) or Pax2⁺ (F) cell populations for each AAV-vector. (G) Viral spread across specific laminae, quantified in mm² for AAV-GlyT2-2 and AAV-GlyT2-3. (H) Viral spread expressed as a percentage of the total area within each respective region of interest. Data show mean ± SEM. AAV-GlyT2-2: n = 9 slices from 3 animals; AAVGlyT2-3: n = 8 slices from 3 animals; data were compared using unpaired t-Tests. No significant differences were detected between both AAV-vectors. (I) GlyT2 (cyan) and VGlut2 (red) were detected by FISH. (J) Pie charts represent the percentage of AAV⁺ cells co-localized with either GlyT2 or VGlut2: n = 8 slices from 3 animals.

Figure 3

The AAV-GAD67 construct used in this study transfected neuronal and non-neuronal cells. (A) Diagram of the AAV construct used. (B) Representative confocal images illustrating AAV-GAD67 expression (EGFP, yellow) in combination with NeuN (cyan) and Pax2 (magenta). Enlarged views of the region marked by the white box in the main image are shown in the middle and lower panel, with arrows highlighting co-localized cells. (C) Distribution of viral expression across individual laminae, represented in mm² (left) and as a percentage of the total area of each defined region (right). (D) Fraction of AAV⁺ cells within each lamina co-expressing NeuN or Pax2. (E) Representative image of AAV-GAD67 transfected cells (yellow) without NeuN (cyan) co-labeling (arrow). (F,G) Transduction efficiency, presented as the percentage of EGFP⁺ cells per mm² relative to the total NeuN⁺ (F) or Pax2⁺ (G) populations. n = 7 to 8 slices from 3 animals. Data show mean ± SEM.

Figure 4

The AAV-Pax2 vector used in the present study did not specifically transduce spinal inhibitory neurons. (A) Scheme of the AAV construct used. (B) Representative confocal images showing AAV-Pax2 expression (EGFP, yellow) alongside NeuN (cyan) and Pax2 (magenta). Enlarged views of the area outlined by the white box in the first image are presented in the middle and lower panels, with arrows indicating co-localized cells. (C) Distribution of viral expression across individual laminae, displayed in mm² (left) and as a percentage of the total area in each specified region (right). (D) Proportion of AAV + cells within each lamina co-expressing NeuN or Pax2. (E) Transduction efficiency, shown as the percentage of EGFP⁺ cells per mm² relative to the total NeuN⁺ (left) or Pax2⁺ (right) populations. n = 8 to 9 slices from 3 animals. Data show mean ± SEM.