Zn²⁺ transport from DRG to spinal dorsal horn increases during aging

To investigate the role of endogenous Zn²⁺ in itch sensation, we assessed its distribution in the DRG and spinal cord using zinc sulfide autometallography (ZnS^AMG) staining, which is used to specifically detect free Zn²⁺ [28,29]. The results showed that there was a large amount of Zn²⁺ selectively distributed in large-sized DRG neurons (cross-sectional area ≥600 μm²), the central branches of which mainly terminate in laminae III-V of the spinal dorsal horn (Fig 1A–1D). Notably, there was a significant increase in the intensity of ZnS^AMG staining specifically within large-sized DRG neurons in aged mice (≥18 months) (Fig 1B and 1C). In the spinal cord, Zn²⁺ was mostly distributed in the deep dorsal horn from laminae III to laminae V (Fig 1E); this area overlapped with the area where the central branches of large-sized DRG neurons terminated. Consistently, the intensity of ZnS^AMG staining in both laminae III-V of the spinal dorsal horn and the epidermis of nape region significantly increased in aged mice (Figs 1E and 1F and S1A and S1B). We measured the concentration of Zn²⁺ ([Zn²⁺]) in the cerebrospinal fluid (CSF), and the results showed that the [Zn²⁺] increased from 0.808 ± 0.054 ng/L in young mice (6 to 8 weeks) to 1.270 ± 0.080 ng/L in aged mice (Fig 1G). Consistently, the [Zn²⁺] in the CSF of elderly people (≥54 years) was significantly higher than that in the CSF of young people (≤38 years) (Fig 1H). Dry skin is a common skin condition in older adults [30,31]. Thus, we evaluated the distribution of vesicular Zn²⁺ in the spinal cord and the [Zn²⁺] in the CSF in a well-established acetone–ether–water (AEW) model that recapitulates dry skin observed in elderly patients [30]. The intensity of ZnS^AMG staining in the lateral part of dorsal horn, receiving sensory information from nape region, significantly increased in ipsilateral side, while that of medial part was not significantly changed (Fig 1I–1K). However, the [Zn²⁺] in the CSF of dry skin model mice was comparable to that of naïve control mice (Fig 1L), possibly because the pathological changes in the nape are insufficient to elicit the change of [Zn²⁺] in CSF. The selective distribution of Zn²⁺ in large-sized DRG neurons and laminae III-V of the spinal dorsal horn indicated that Zn²⁺ in the deep laminae of the dorsal horn may be transported from large-sized DRG neurons. To test this hypothesis, we performed a spinal dorsal root ligation assay (Fig 1A). Interestingly, our results demonstrated a substantial decrease in the intensity of ZnS^AMG staining specifically in laminae III-V on the ipsilateral side compared to the contralateral side 3 days after ligation (Fig 1M and 1N). These findings strongly suggest that Zn²⁺ in the deep laminae of the dorsal horn is indeed transported from DRG neurons and its level increased during aging and in dry skin.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. The distribution of vesicular Zn²⁺ in the DRG and spinal cord.

(A) Schematic diagram illustrating the projection of central branches arising from different DRG neuron subtypes within the spinal dorsal horn and the dorsal root ligation model. (B, C) ZnS^AMG staining in the DRG in young mice and aged mice. (B) Representative images displaying ZnS^AMG staining in DRG sections. Scale bar = 50 μm. (C) Quantitative analysis of the mean gray value of ZnS^AMG. Welch’s t test, n = 6/group. (D) Size distribution of ZnS^AMG-positive DRG neurons and total DRG neurons in naïve young mice. (E, F) ZnS^AMG staining in the spinal cords of young mice and aged mice. (E) Representative image of ZnS^AMG staining in the spinal cord. Scale bar = 50 μm. (F) Quantitative analysis of the mean intensity of ZnS^AMG staining; unpaired t test, n = 10/group. (G) The [Zn²⁺] in the CSF of young mice and aged mice; unpaired t test, n = 8 mice/group. (H) The [Zn²⁺] in the CSF of young patients and elderly patients; Welch’s t test, n = 5–6/group. (I–K) ZnS^AMG staining in the spinal cords of dry skin mice. (I) Representative image of ZnS^AMG staining in the spinal cord. Scale bar = 50 μm. (J, K) Quantitative analysis of the mean intensity of ZnS^AMG staining on the contralateral and ipsilateral sides; (J) lateral spinal cord, Mann–Whitney test, n = 11/group; (K) medial spinal cord, unpaired t test, n = 11/group. (L) The [Zn²⁺] in the CSF of naïve mice and dry skin model mice; unpaired t test, n = 6 mice/group. (M, N) ZnS^AMG staining in the spinal cords of dorsal root ligation model mice. (M) Representative image of ZnS^AMG staining in the spinal cord. Scale bar = 50 μm. (N) Quantitative analysis of the mean ZnS^AMG staining intensity in spinal cords of dorsal root ligation model mice; unpaired t test, n = 9/group. All data are expressed as the mean ± SEM. Significant differences were analyzed using unpaired t tests; *p < 0.05, **p < 0.01, and ***p < 0.001. The underlying data for Fig 1C, 1D, 1F, 1G, 1H, 1J, 1K, 1L, and 1N can be found in S1 Data. CSF, cerebrospinal fluid; DRG, dorsal root ganglia.

https://doi.org/10.1371/journal.pbio.3002888.g001

Zn²⁺ in spinal cord regulates aging-related itch

Next, we tested whether spinal Zn²⁺ mediates aging-related itch, including spontaneous itch and mechanical itch hypersensitivity. Mechanical itch hypersensitivity, known as alloknesis, represents an abnormal sensory state in which mechanical stimuli, typically non-itch-inducing, induce intense itchiness, a phenomenon commonly observed in aged skin [32,33]. Aged mice developed both spontaneous itch and mechanical itch hypersensitivity (S2A–S2D Fig), which was consistent with previous studies [3,30]. Intrathecal (i.t.) injection of the specific Zn²⁺ chelator TPEN (100 μm, 10 μl) or ZX1 (100 μm, 10 μl) [34,35] both can significantly reduce the intensity of ZnS^AMG staining in laminae III-V of spinal dorsal horn (Fig 2A–2D). Importantly, these 2 chelators effectively alleviated both spontaneous itch and touch-evoked itch in aged mice in the first hour after injection (Fig 2E–2G). Consistently, both spontaneous itch and touch-evoked itch in the AEW treatment-induced dry skin model mice were effectively alleviated after chelation of spinal Zn²⁺ through i.t. injection of TPEN (Fig 2H and 2I). However, neither histamine-dependent acute chemical itch (elicited by intradermal injection of compound 48/80 and histamine) nor histamine-independent acute chemical itch (elicited by intradermal injection of chloroquine, serotonin, β-alanine, and SLGRL) were affected by chelation of spinal Zn²⁺ with i.t. injection of TPEN (Fig 2J–2O). Next, we investigated whether Zn²⁺ can directly elicit itch. The results demonstrated that i.t. injection of ZnCl₂ increased the intensity of the Zn²⁺ signal in the spinal dorsal horn (Fig 2P and 2Q). Behavioral tests showed that i.t. injection of ZnCl₂ was insufficient to elicit spontaneous itch but significantly potentiated touch-evoked itch stimulated by application of a 0.07 g filament to both the nape and back of the ear (Fig 2R–2T). Collectively, these data indicate that endogenous Zn²⁺ in the spinal cord regulates aging-related itch.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Spinal Zn²⁺ regulates aging-related itch but not acute chemical itch.

(A, B) ZnS^AMG staining in the spinal cords after I.T. injection ZX1. (A) Representative image of ZnS^AMG staining in the spinal cord. Scale bar = 50 μm. (B) Quantitative analysis of the mean ZnS^AMG staining intensity in spinal cords after I.T. injection ZX1; unpaired t test, n = 26-21/group. (C, D) ZnS^AMG staining in the spinal cords after I.T. injection TPEN. (C) Representative image of ZnS^AMG staining in the spinal cord. Scale bar = 50 μm. (D) Quantitative analysis of the mean ZnS^AMG staining intensity in spinal cords after I.T. injection TPEN; unpaired t test, n = 21~19/group. (E) Spontaneous itch response of aged mice after intrathecal injection of the Zn²⁺ chelators TPEN and ZX1. Two-way ANOVA followed by Sidak’s multiple comparisons test; n = 6 mice/group. (F, G) Touch-evoked itch response of aged mice elicited with a 0.07 g von Frey filament after intrathecal injection of vehicle, the Zn²⁺ chelator TPEN (100 μm, 10 μl) (F) or ZX1 (100 μm, 10 μl) (G). Two-way ANOVA followed by Sidak’s multiple comparisons test; n = 5–6 mice/group. (H) Spontaneous itch response of dry skin model mice after intrathecal injection of vehicle or the Zn²⁺ chelator TPEN. Two-way ANOVA followed by Sidak’s multiple comparisons test; n = 6 mice/group. (I) Touch-evoked itch response of dry skin model mice after intrathecal injection of vehicle or the Zn²⁺ chelator TPEN. Two-way ANOVA followed by Sidak’s multiple comparisons test; n = 5–6 mice/group. (J–O) Effect of intrathecal injection of vehicle or the Zn²⁺ chelator TPEN on acute chemical itch elicited by intradermal injection of 48/80 (2 μg/μl, 50 μl), histamine (10 μg/μl, 50 μl), CQ (4 μg/μl, 50 μl), 5-HT (6 μg/μl, 50 μl), β-alanine (6 μg/μl, 50 μl), and SLGRL (2 μg/μl, 50 μl). Unpaired t test; n = 5–6 mice/group. (P, Q) ZnS^AMG staining in the spinal cords after I.T. injection Vehicle and ZnCl₂. (P) Representative image of ZnS^AMG staining in the spinal cord. Scale bar = 50 μm. (Q) Quantitative analysis of the mean ZnS^AMG staining intensity in spinal cords after I.T. injection Vehicle and ZnCl₂; unpaired t test, n = 19/group. (R) Spontaneous itch response of WT mice after intrathecal injection Vehicle and ZnCl₂. Unpaired t test, n = 5 mice/group. (S, T) Itch response evoked by touch stimulation of the nape (S) and back of the ear (T) after intrathecal injection of 2 × NaCl and ZnCl₂. Two-way ANOVA followed by Sidak’s multiple comparisons test, n = 6–12 mice/group. All data are expressed as the mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001. The underlying data for Fig 2B, 2D, 2E, 2F, 2G, 2H, 2I, 2J, 2K, 2L, 2M, 2N, 2O, 2Q, 2R, 2S, and 2T can be found in S1 Data. SEM, standard error of the mean.

https://doi.org/10.1371/journal.pbio.3002888.g002

Previous studies have highlighted the essential role of skin Zn²⁺ in both acute itch and chronic itch [36]. We then investigated whether skin Zn²⁺ is involved in aging-associated itch. Our results showed that chelating skin Zn²⁺ via intradermal injection of ZX1 at the ear back significantly reduced both spontaneous and touch-evoked itch in aged mice (S2E and S2F Fig), suggesting that skin Zn²⁺ plays a role in age-associated itch. I.t. injection is a well-established method for delivering drugs, siRNAs, or AAVs to DRG neurons and SC, as previously described [37–40]. Our findings demonstrated that I.t. injection of TPEN elicited an obvious tendency of reduction of [Zn²⁺], although this difference was not statistically significant (S2G and S2H Fig), indicating that Zn²⁺ in DRG neurons may be released into the skin and contribute to Zn²⁺ levels. However, I.t. injection of ZnCl₂ had no effect on the [Zn²⁺] in the skin (S2I and S2J Fig). These data suggest that I.t. injection of TPEN-elicited reduction of age-associated itch is partially due to the chelation of skin Zn²⁺.

TMEM163 mediates Zn²⁺ accumulation in spinal dorsal horn and aging-related itch

To elucidate the molecular mechanism underlying the accumulation of spinal Zn²⁺ in deep laminae of spinal dorsal horn and the selective regulatory effect of spinal Zn²⁺ on aging-related itch, we conducted a comprehensive analysis of Zn²⁺ transporters (ZnTs) responsible for transporting Zn²⁺ from the cytosol into organelles [8] within the DRG. Our results revealed that the expression level of TMEM163 (ZnT11) in the DRG was markedly higher than that of other ZnTs (Fig 3A). Furthermore, we examined the expression profiles of ZnTs in a single-cell sequencing database [41,42], and the results showed that TMEM163, but not ZnT1-10, was selectively highly expressed in large-sized DRG neurons (neurofilament (NF)-containing DRG neurons) (Figs 3B and S3A), which was consistent with the observed distribution of Zn²⁺ in DRG (Fig 1B and 1D). Consistently, ISH combined with IHC showed that Tmem163 was mainly expressed in large-sized DRG neurons, with 79.24 ± 2.86% of Tmem163⁺ DRG neurons expressing NF200 and 87.53 ± 1.56% of Tmem163⁺ DRG neurons being positive for Slc17a7 (VGluT1) (Figs 3C, 3D, S3B, and S3C). In mice with dry skin and aged mice, the expression of Tmem163 increased significantly (Fig 3E and 3F), indicating the potential role of this transporter in aging-related itch. To test the role of Tmem163 in aging-related Zn²⁺ accumulation in the spinal dorsal horn and aging-related itch, we generated Tmem163 conditional knockout (cKO) mice in which Tmem163 was selectively deleted in VGluT1⁺ DRG neurons (Fig 3G and 3H). In the cKO mice, the intensity of ZnS^AMG staining in the DRG was significantly reduced (Fig 3I and 3J). Furthermore, the concentration of Zn²⁺ in the CSF was markedly decreased from 0.927 ± 0.177 ng/L in WT mice to 0.332 ± 0.035 ng/L in cKO mice (Fig 3K). We also examined the distribution of Zn²⁺ in the deep laminae of the spinal dorsal horn in cKO mice and found a sharp reduction in the intensity of ZnS^AMG staining (Fig 3L and 3M). These findings collectively indicate the critical role of TMEM163 in the accumulation of Zn²⁺ in large-sized DRG neurons and the deep laminae of the spinal dorsal horn. We also assessed the itch behavior of the cKO mice under naïve conditions and in the context of dry skin. Touch-evoked itch was induced by stimulation of the ear region with graded von Frey filaments, and the results showed that itch evoked by the filaments with bending forces of 0.07 g and 0.16 g was significantly decreased in the cKO mice under naïve conditions (Fig 3N). Similarly, touch-evoked itch elicited by application of Von Frey filaments to the nape region was also reduced (S3D Fig). Under dry skin conditions, both spontaneous itch and touch-evoked itch were decreased dramatically in cKO mice (Figs 3O–3Q and S3E). Importantly, the reduction in touch-evoked itch can be reversed by i.t. injection of ZnCl₂ under dry skin conditions (Fig 3R). Collectively, these findings provide compelling evidence that the Zn²⁺ transporter TMEM163 mediates the accumulation of Zn²⁺ in large-sized DRG neurons and the deep laminae of the spinal dorsal horn, thereby playing a vital role in the pathogenesis of aging-related itch.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. TMEM163 mediates vesicular Zn²⁺ transport and aging-related itch.

(A) Relative mRNA levels of ZnT1-ZnT10 and TMEM163 in the mouse DRG; n = 3/group. The underlying primer data for Fig 3A can be found in S1 Table. (B) Profiling of the cell type expressing TMEM163 in the mouse DRG based on a previously published single-cell RNA-seq database (GSE59739). (C) Representative images of ISH combined with IHC using a probe targeting mouse Tmem163 (red) and a large DRG neuron marker (NF3200). Scale bar = 100 μm. (D) Profiling of the cell type expressing TMEM163 in the DRG. Immunofluorescence staining using markers for different DRG neuron subtypes (large-sized DRG neurons (NF200), peptidergic DRG neurons (SP and CGRP), nonpeptidergic DRG neurons (IB4), and C-LTMRs (TH)) was performed; n = 17 /group. (E) Relative mRNA level of Tmem163 in the DRG of naïve control mice and dry skin model mice; unpaired t test, n = 4 /group. (F) Relative mRNA level of Tmem163 in the mouse DRG at 2 months (2 m), 4 months (4 m), 6 months (6 m), 12 months (12 m), 18 months (18 m), and 24 months (24 m). Unpaired t test; n = 3–7/group. (G) Schematic diagram of the strategy employed to generate VGlut1::cre;Tmem163^f/f mice. (H) Genotyping of Tmem163 cKO mice using the primer listed in the S1 Table. The gel image shows the bands corresponding to WT mice (132 bp), Tmem163^f/f mice (202 bp), and Cre recombinase (100 bp). M indicates molecular weight marker bands for reference. (I, J) ZnS^AMG staining in the DRG of WT mice and Tmem163 cKO mice. (I) Representative images of ZnS^AMG staining in DRG sections. Scale bar = 20 μm. (J) Quantitative analysis of the mean intensity of ZnS^AMG staining, Mann–Whitney test, n = 13–14/group. (K) Quantification of the [Zn²⁺] in the CSF of WT mice and Tmem163 cKO mice; n = 3/group. (L, M) ZnS^AMG staining in the spinal cords of WT mice and Tmem163 cKO mice. (L) Representative images of ZnS^AMG staining in spinal cord sections. Scale bar = 50 μm. (M) Quantitative analysis of the mean intensity of ZnS^AMG staining, Mann–Whitney test, n = 10/group. (N) Touch-evoked itch response elicited by application with different von Frey filaments in Tmem163^f/f mice and Tmem163 cKO mice; two-way ANOVA followed by Sidak’s multiple comparisons test; n = 6 /group. (O) Spontaneous itch response of Tmem163^f/f and Tmem163 cKO mice under dry skin conditions. Unpaired t test; n = 6/group. (P) Correlation analysis of Tmem163 mRNA level with spontaneous itch number of Tmem163^f/f mice and Tmem163 cKO mice. Pearson correlation analyze; n = 6/group. (Q) Touch-evoked itch response of Tmem163^f/f and Tmem163 cKO mice under dry skin conditions. Welch’s t test; n = 6/group. (R) Touch-evoked itch response of Tmem163^f/f and Tmem163 cKO mice after intrathecal injection of vehicle or ZnCl₂. Two-way ANOVA followed by Sidak’s multiple comparisons test; n = 5–6/group. All data are expressed as the mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001. The underlying data for Fig 3A, 3D, 3E, 3F, 3J, 3K, 3M, 3N, 3O, 3P, 3Q, and 3R can be found in S1 Data. cKO, conditional knockout; CSF, cerebrospinal fluid; DRG, dorsal root ganglia; ISH, in situ hybridization; SEM, standard error of the mean; WT, wild-type.

https://doi.org/10.1371/journal.pbio.3002888.g003

TMEM163 and VGluT1 are colocalized in the same presynaptic vesicles

Next, we further investigated how vesicular Zn²⁺ is transported from the DRG to the spinal dorsal horn by TMEM163 and released. A previous study showed that Zn²⁺ is mostly stored in the same synaptic vesicles as Glu and released together with Glu into the synaptic cleft during neuronal firing, leading to a transient increase in the free Zn²⁺ concentration in the synaptic cleft, which further modulates synaptic transmission in the hippocampus [8,14,43,44]. VGluTs localized on the membrane of synaptic vesicles play a critical role in incorporating Glu into synaptic vesicles. Because VGluT1 is selectively expressed in large DRG neurons [45], similar to the expression pattern of TMEM163, we examined the colocalization of TMEM163 and VGluT1 at the subcellular level. Data from VGluT1::cre; Ai14 mice showed that the central branches of VGluT1⁺ DRG neurons terminated in the deep laminae of the spinal dorsal horn, overlapping with the ZnS^AMG signal (Fig 4A). Then, we induced overexpression of TMEM163-EGFP and VGluT1-mCherry in the ND7/23 cell line, a hybridized cell line consisting of mouse neuroblastoma and rat DRG neurons. TMEM163 colocalized with VGluT1 in vesicle-like punctate structures (Fig 4B). Then, we employed electroporation to induce overexpression of TMEM163-EGFP and VGluT1-mCherry in cultured DRG neurons. Remarkably, we observed distinct colocalization of TMEM163-EGFP and VGluT1-mCherry within the same vesicle-like puncta in the cell body, neurites, and nerve terminals (Fig 4C and 4D). Quantitative analysis showed that 42.10% of TMEM163-EGFP⁺ vesicle-like puncta were vGluT1-mCherry-positive and 45.62% of vGluT1-mCherry⁺ vesicle-like puncta were TMEM163-EGFP-positive (Fig 4E). Live cell imaging was also used to trace the cotransport of TMEM163-EGFP and vGluT1-mCherry. The results showed that 35.51% of double-positive vesicle-like puncta were anterogradely transported toward the nerve terminal, 16.19% of them were retrogradely transported toward the cell body, and 48.30% of them did not show any movement (Fig 4D–4F).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. TMEM163 colocalizes with VGluT1 on synaptic vesicles.

(A) ZnS^AMG staining in the spinal cords of VGluT1::Cre;Ai14 mice. Scale bar = 100 μm. (B) Overexpression of mouse TMEM163-EGFP (green) and mouse VGluT1-mCherry (red) in ND7/23 cells. Scale bar = 10 μm. (C, D) TMEM163-EGFP (green) and VGluT1-mCherry (red) were overexpressed in cultured DRG neurons by electroporation. (C) VGluT1 (red) colocalized with TMEM163 (green) in the axons of cultured DRG neurons. (D) Live-cell imaging reveals the cotransport of TMEM163-EGFP (green) and VGluT1-mCherry (red) in the same vesicle. Scale bar = 5 μm. (E, F) Quantitative analysis of VGluT1 (red) cotransport with TMEM163 (green) in the axons of cultured DRG neurons. (E) The percentage of TMEM163-positive puncta containing VGluT1 and the percentage of VGluT1-positive puncta containing TMEM163, n = 11. (F) The percentage of VGluT1 and TMEM163 double-positive puncta exhibiting anterograde transport, retrograde transport, and no movement; n = 11. (G) Immunofluorescence staining of the spinal cord following intrathecal injection of AAV-TMEM163-EGFP into 6- to 8-week-old mice. Spinal cord tissue was harvested and subjected to immunostaining using primary antibodies against VGluT1 (red) and Nissl staining (purple). Scale bars = 100 μm (left), 20 μm (right). (H) Immunogold labeling of EGFP (10 nm) and VGluT1 (6 nm) in spinal cord slices from mice infected with AAV-TMEM163-EGFP. The dashed line circle indicates a vesicle containing both TMEM163-EGFP and VGluT1. The arrowhead indicates a TMEM163-EGFP-positive vesicle, and the arrow indicates a VGluT1-positive vesicle. Scale bars = 200 nm (left) and 50 nm (right). The underlying data for Fig 4E and 4F can be found in S1 Data. AAV, adeno-associated virus; DRG, dorsal root ganglia.

https://doi.org/10.1371/journal.pbio.3002888.g004

To investigate the potential colocalization of TMEM163 and VGluT1 on presynaptic vesicles, we employed immunoelectron microscopy. Because specific antibodies against TMEM163 are not available, we utilized an adeno-associated virus (AAV) vector expressing TMEM163 fused with an EGFP reporter at its C-terminus (AAV-TMEM163-EGFP). Subsequently, we administered AAV-TMEM163-EGFP (titer: 1.045E+12 vg/ml, 10 μl) by i.t. injection to specifically infect large DRG neurons, whose central nerve branch fibers predominantly terminated in the deep laminae of the spinal dorsal horn (Fig 4G). Then, we labeled EGFP with 10 nm gold particles and VGluT1 with 6 nm gold particles separately. The results demonstrated the presence of TMEM163-EGFP and VGluT1 either in the same presynaptic vesicles or distributed across separate presynaptic vesicles (Figs 4H, S4A, and S4B). Quantitative analysis showed that 20.24 ± 1.711 (%) 10 nm gold particles colocalized with 6 nm gold particles and 11.33 ± 1.624 (%) 10 nm gold particles colocalized with 10 nm gold particles in the same presynaptic vesicles (S4C Fig). These findings are consistent with the observed localization of TMEM163 and VGluT1 in ND7/23 cells and cultured DRG neurons (Fig 4B–4F). Collectively, these data strongly indicate that Zn²⁺ and Glu may be incorporated into the same vesicles, cotransported to nerve terminals, and subsequently coreleased upon stimulation and depolarization of TMEM163-positive primary afferents.

TMEM163⁺ primary afferents form direct synaptic connections with NPY⁺ INs

To unravel the specific involvement of vesicular Zn²⁺ stored in the central terminals of TMEM163⁺ primary afferents in mediating aging-related itch, we investigated the postsynaptic neurons of these primary afferents within the spinal cord. Recognizing the critical role of inhibitory INs derived from the NPY⁺ lineage in itch [3], we aimed to determine whether Tmem163⁺ primary afferents directly synapse with NPY⁺ spinal inhibitory INs. In the spinal cord of NPY::cre;Ai3 mice, NPY⁺-derived inhibitory INs, distributed in laminae I-V, were labeled with EGFP (S5A Fig). Given that Slc17a7 (which encodes VGluT1) mostly colocalized with Tmem163 in the DRG (Figs 3D and S3C), we assumed that the primary afferents of VGluT1⁺ neurons and Tmem163⁺ neurons in the spinal cord have the same projections and make the same synaptic connections. There were abundant VGluT1⁺ (Tmem163⁺) nerve terminals surrounding NPY⁺-derived INs in laminae III-V of the spinal cord (S5B Fig), indicating that there was a possible direct synaptic connection between VGluT1⁺ nerve terminals and NPY⁺-derived INs. To identify the presynaptic neurons in the DRG that formed direct synapses with NPY⁺ INs, we performed retrograde trans-monosynaptic tracing by injecting an AAV vector encoding a Cre-dependent RVG (AAV2/5-hEF1a-DIO-RVG-WPRE-pA) and TVA fused with mCherry (AAV2/9-CAG-DIO-TVA-mCherry-WPRE-pA) into laminae III-V of NPY-Cre mice and then injected EnvA-pseudotyped rabies virus (RV-ENVA-DG-EGFP) into the mice 3 weeks later (Fig 5A). The results showed that 38.2% of mCherry⁺ neurons expressing EGFP in the spinal dorsal horn (S6A Fig), and there were EGFP⁺ neurons in the DRG, with 80.09 ± 2.209 (%) of these neurons expressing Tmem163, 81.90 ± 4.448 (%) of them expressing the A-fiber-generating DRG neuron marker NF200, only 2.50 ± 2.007 (%) of them expressing the Aδ-LTMR marker Ntrk2 (TrkB), and none of them expressing the C-LTMR marker tyrosine hydroxylase (TH) (Figs 5B, 5C, S6B, and S6C). The same strategy was used to trace the presynaptic neurons of Ucn3::Cre INs in the DRG. Ucn3::Cre INs are mainly distributed in laminae II-III of spinal dorsal horn (S6D Fig). In contrast, only 27.74 ± 5.007 (%) of EGFP⁺ DRG neurons traced in Ucn3::Cre mice coexpressed Tmem163 (S6E and S6F Fig). Collectively, these data strongly indicate that NPY⁺-derived INs receive monosynaptic inputs from Tmem163⁺ primary afferents, whereas Ucn3::Cre neurons exhibit minimal monosynaptic connections with Tmem163⁺ primary afferents.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. TMEM163⁺ primary afferents form monosynaptic connections with NPY⁺ INs.

(A) Schematic diagram illustrating the strategy for retrograde transsynaptic tracing of the upstream neurons of NPY::Cre INs or UCN3::Cre INs in the DRG by intraspinal injection of viruses. NPY⁺ mice or UCN3::Cre mice received intraspinal injection of an AAV vector encoding Cre-dependent RVG (AAV2/5-hEF1a-DIO-RVG-WPRE-pA) and TVA fused with mCherry (AAV2/9-CAG-DIO-TVA-mCherry-WPRE-pA) into laminae III-V. After 3 weeks, EnvA-pseudotyped rabies virus (RV-ENVA-DG-EGFP) was injected for retrograde tracing. (B) Representative images showing infected presynaptic neurons (EGFP⁺) of NPY⁺ INs in the DRG and their coexpression of Tmem163 using ISH. Scale bar = 100 μm. (C) Quantitative analysis of the data in (B). (D) Representative eEPSCs in laminae III-V NPY⁺ INs before and after ZX1 application (hold at −70 mV). (E) Time course of the change in eEPSC amplitude before and after ZX1 application; n = 9 cells. (F) Quantitative analysis of the eEPSC amplitude; paired t test; n = 9 cells/group. (G) Representative eEPSCs in laminae III-V NPY⁺ INs before and after ZX1 application (hold at −40 mV). (H) Time course of the change in eEPSC amplitude before and after ZX1 application; n = 10 cells. (I) Quantitative analysis of the eEPSC amplitude; paired t test; n = 10 cells/group. All data are expressed as the mean ± SEM; *p < 0.05, **p < 0.01. The underlying data for Fig 5C, 5E, 5F, 5H, and 5I can be found in S1 Data. AAV, adeno-associated virus; DRG, dorsal root ganglia; IN, interneuron; ISH, in situ hybridization; SEM, standard error of the mean.

https://doi.org/10.1371/journal.pbio.3002888.g005

Spinal Zn²⁺ modulates the neuronal activity of NPY⁺ neurons via AMPARs and NMDARs

Next, we investigated whether spinal Zn²⁺ stored in the central terminals of Tmem163⁺ primary afferents modulate the neuronal activity of NPY⁺ INs. Neither the amplitude nor the frequency of mEPSCs was affected by the chelation of Zn²⁺ with TPEN (100 μm) or ZX1 (100 μm) (S7A–S7K Fig). Next, we sought to investigate the impact of spinal Zn²⁺ chelation on the synaptic responses of NPY⁺ inhibitory INs mediated by AMPA receptors (AMPARs) and NMDA receptors (NMDARs). The amplitude of AMPAR-mediated EPSCs (AMPAR-EPSCs) was significantly larger in the presence of ZX1 (Fig 5D–5F). We further examined focally evoked NMDAR-mediated EPSCs (NMDAR-EPSCs) in the presence of NBQX (10 μm), bicuculline (10 μm), and strychnine (2 μm) at a holding potential of −40 mV in these 2 groups. The amplitude of NMDAR-EPSCs was also significantly increased in the presence of ZX1 (Fig 5G–5I). To investigate the potential involvement of presynaptic mechanisms in the effects of spinal Zn²⁺ on evoked excitatory postsynaptic currents (EPSCs), we employed paired-pulse ratio (PPR) analysis, a sensitive measure of presynaptic neurotransmitter release probability [46]. Chelation of spinal Zn²⁺ using ZX1 did not induce any significant changes in the PPR of EPSCs in NPY⁺ INs, indicating that presynaptic mechanisms do not contribute to the Zn²⁺-mediated attenuation of the EPSC amplitude (S7L–S7N Fig).

In the aged mice, both the amplitude of AMPAR-mediated and NMDAR-mediated EPSCs of NPY⁺ INs significantly reduced (S8A–S8D Fig). In dry skin model mice, the frequency of mEPSCs in NPY⁺ INs were dramatically reduced, while the amplitude did not change (S8E–S8I Fig). Chelation of spinal Zn²⁺ with ZX1 increased the amplitudes of both AMPAR-EPSCs and NMDAR-EPSCs (S8J–S8M Fig). These results are consistent with the notion that spinal Zn²⁺ inhibits both AMPAR- and NMDAR-mediated EPSCs and modulates the neuronal activity of NPY⁺ neurons via a postsynaptic mechanism. In sharp contrast, neither AMPAR-mediated EPSCs nor NMDAR-mediated EPSCs in Ucn3::Cre neurons were affected by chelation of endogenous Zn²⁺ with ZX1, indicating that Ucn3::Cre neurons was not directly inhibited by vesicular Zn²⁺ or that the inhibitory effect was not strong enough to affect the synaptic responses of Ucn3::Cre neurons (S9 Fig), which is also in line with the retrograde trans-monosynaptic tracing results (S6E and S6F Fig). Thus, these data provide evidence that Zn²⁺ released from presynaptic vesicles exerts a modulatory effect on the neuronal activity of NPY⁺ neurons through the inhibition of AMPARs and NMDARs.

Animals

Tmem163^f/f mice were procured from Cyagen Biosciences. The NPY::Cre mice (Stock No: 027851), VGluT1::Cre mice (Stock No: 023527), Ai3 mice (Stock No: 007903), and Ai14 mice (Stock No: 007908) were obtained from The Jackson Laboratory. Ucn3::Cre mice were generously provided by Dr. Haili Pan’s lab at Jiangxi Provincial People’s Hospital. C57BL/6 WT mice were purchased from Shanghai Lingchang Biological Technology Company. C57BL/6 aged mice were purchased from Jiangsu ALF Biotechnology Company.

All mice were group-housed and bred in the animal facilities of Fudan University, following a 12-h light/12-h dark cycle at a temperature of 22 ± 1°C. All behavioral tests were conducted by observers blind to the treatments and genotypes of animals. All mice used for behavior tests were genotyped and assigned to either experimental groups or control groups. All animal procedures were conducted in accordance with the guidelines outlined in the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee of the Institutes of Brain Science at Fudan University.

ZnS^AMG staining

Mice intended for sampling were deeply anesthetized with 0.75% pentobarbital sodium at 10 μl/g body weight. After deep anesthesia, the mice were subjected to transcardiac perfusion with PBS followed by Na₂S (FD Rapid TimmStain Kit, Cat: Pk701), followed by perfusion with 4% paraformaldehyde (PFA) supplemented with 0.1% picric acid in PBS as separate solutions. The DRG and spinal cord were subsequently excised, postfixed overnight at 4°C, and subjected to dehydration through a graded series of sucrose solutions. Serial transverse sections of the spinal cord (30 μm, free-floating) skin and DRG (14 μm) were prepared using a freezing microtome. The sections were thoroughly rinsed with 0.1 M phosphate buffer solution for 10 min and then immersed in an AMG incubation solution (containing 50% Arabic gelatin, sodium citrate buffer, p-diphenol, and silver lactate solution) for 35 min at 32°C. To ensure consistency of zinc reactions, tissue samples from different ages or groups were processed together in the same batch. For the dry skin model and dorsal root ligation experiments, treatments were applied to both sides of the same mice. This approach minimized variability caused by differences in processing times and conditions. All tissue samples were fixed, stained, and analyzed under identical conditions. Observers maintained blinding throughout the entire experimental and analysis processes to ensure the objectivity and consistency of data interpretation. Following dehydration and clearing procedures, the sections were sealed with neutral gum. Subsequently, brightfield microscopy images from ZnS^AMG staining section series were acquired using a Nikon microscope through a 20× objective. The white balance was adjusted prior to image capture. ZnS^AMG staining images were quantitatively analyzed using Image-Pro Plus 6.0 software. Before analysis, the optical density of the images was corrected, a white reference point was established, and the background was subtracted. Measurement regions were delineated using the irregular tool within the HSI color system. Subsequently, the total integrated optical density (IOD) was measured, and the mean density was calculated.

Mice and human cerebrospinal fluid (CSF) collection and [Zn²⁺] test

Mice were deeply anesthetized with isoflurane, and the cisterna magna was surgically exposed. A sharpened glass electrode was affixed to a capillary holder, which was then connected to a 1-ml syringe using a plastic tube. Negative pressure was created in the glass electrode by pulling the syringe plunger back by 100~200 μl. Subsequently, the electrode was carefully inserted into the cisterna magna to collect CSF. The collected CSF samples were either stored in a −80°C freezer for future use or subjected to [Zn²⁺] tests using the Zinc Assay Kit (Sigma Aldrich Cat: MAK032) following the recommended procedure. After the CSF collection, mice were anesthetized with 0.75% pentobarbital sodium at 10 μl/g body weight and sacrificed by decapitation.

For human CSF collection, the study was approved by the Ethics Committee of Huashan Hospital of Fudan University with certificate No. KY2020-1218. The CSF collection was a voluntary CSF donation by the patients develop a single meningioma who consented to their CSF being used for research purpose in anonymized way. The written informed consent was received from participants prior to inclusion in the study. Eleven patients were divided into 2 groups based on their age: young group with age ≤38 years old and the old group with age ≥54 years old. The CSF samples were obtained through standard lumbar puncture and centrifuged at 4°C, and the supernatants were transferred to fresh tubes and stored at −80°C.

In situ hybridization (ISH) and immunohistochemistry (IHC)

Mice were deeply anesthetized with isoflurane and subjected to transcardiac perfusion with PBS, followed by 4% PFA in 0.01 M PBS. The spinal cord was sectioned at a thickness of 30 μm, and the DRG was sectioned at a thickness of 14 μm using a cryostat (Leica). In situ hybridization (ISH) was performed using the RNAscope Fluorescent Multiplex Assay kit and probes specifically targeting mouse Tmem163 (Cat No. 854531-C1), Slc17a7(Cat NO.416631-C2), and Ntrk2 (Cat No.423611-C1) (Advanced Cell Diagnostics) in strict accordance with the recommended protocol. Subsequently, the sections were incubated with primary antibodies anti-SP (Guinea pig, 1:1,000; Neuromics; GP14103), anti-CGRP (goat, 1:1,000; Bio-Rad; AB_2290729), anti-VGluT1 (Guinea pig, 1:1,000; Millipore; AB_5905), anti-TH (rabbit, 1:1,000, Millipore; AB152), anti-NF200 (mouse, 1:1,000, Millipore; AB_477257), or Anti-NeuN (mouse, 1:1,000, Millipore; Cat No. MAB377) overnight at 4°C, followed by Cy3- (1:400, Jackson ImmunoResearch Laboratories Inc. Cat No.112-165-003), Cy5- (1:400, Jackson ImmunoResearch Laboratories Inc. Cat No.715-175-151), or FITC-conjugated secondary antibodies (1:400; Jackson ImmunoResearch Laboratories Inc. Cat No.115-095-003) or FITC-conjugated IB4 (10 μg/ml; Invitrogen. Cat No. I21411) or DAPI (1:1,000; Invitrogen; 62248), or Nissl (1:1,000; Invitrogen; N21483). Sections were mounted and examined under Olympus confocal laser scanning microscopy.

Itch behavioral test

All mice were habituated to being placed in a testing box (8 × 8 × 20 cm^3) positioned on a metal mesh over 2 days in advance, with each habituation session lasting for 2 h. Subsequently, scratching behaviors were recorded using video surveillance for a duration of 1 h, in the absence of any observer. A scratch event was defined when a mouse lifted its hind paw to scratch the shaved region and subsequently returned the paw to either the floor or the mouth for licking. The recording of spontaneous itch behaviors was performed blindly and evaluated without knowledge of the experimental conditions.

For touch-evoked itch testing, the fur behind the ear or on the nape was shaved 2 days before the experiments. Mice were habituated for 30 min on 2 consecutive days in the behavioral testing apparatus. On the testing day, mice were placed in plastic enclosures and further habituated for 60 min. To further enhance data reproducibility and evaluate the effect of a single stimulation on subsequent scratch responses in mice, we administered 5 stimulations, with 5 s interval between each stimulation. We then recorded the total number of scratch bouts induced by these 5 stimulations as the score for touch-evoked itch in each mouse behind the ears and nape. Mechanical stimuli were applied using 0.07 g Von Frey filaments (North Coast Medical, Morgan Hill, California, United States of America). The total number of positive responses across the 5 stimuli was recorded to indicate the mechanical itch score. For chemical itch testing, mice were shaved on the nape of the neck and acclimated in the recording chambers for 2 consecutive days before behavioral testing. On the testing day, different chemicals were injected intradermally into the middle of the nape separately. Compound 48/80 (2 μg/μl, 50 μl, Sigma, C2313), chloroquine (4 μg/μl, 50 μl, Sigma, C6628), β-alanine (60 μg/μl, 50 μl, Macklin, B874460), and SLGRL (2 μg/μl, 50 μl, tgpeptide) in a volume of 50 μl saline were injected intradermally. The number of scratching events was counted for 30 min after the injection. All behavioral tests were done by observers blind to the treatments or genotypes of animals.

Dry skin model

Fur on the nape was removed 2 days before treatment. To generate a dry skin model, the nape was subjected to a treatment involving a mixture of acetone/ether (1:1, Sigma, St. Louis, Missouri, USA) for 20 s, followed by distilled water for 40 s (referred to as AEW), twice daily, for 7 consecutive days. The spontaneous itch was recorded using a camera for 1 h following the protocol for itch behavioral testing. Mechanical itch hypersensitivity was assessed using a 0.07 g Von Frey filament, as previously described.

Dorsal root ligation model

Mice were anesthetized with pentobarbital sodium (75 mg/kg i.p.). The hair on the back, ranging from lumbar (L) 3 to lumbar (L) 6, was removed using a hair remover. Mice were then fixed in a stereotaxic instrument (RWD Lift Science Co.), L4-L5 region was shaved and wiped with iodophor, and the skin and muscle were cut and removed with a scissor to expose the vertebral plate, zygapophysis, and transverse process of spine column. Using a cranial drill (RWD Lift Science Co.) carefully removed the left part of the L4 and L5 vertebral plate, zygapophysis, and transverse process to expose the L4 and L5 DRG. Tying the dorsal roots of L4-L5 with fine thread, disinfecting and suturing the skin, and collecting samples 3 days later.

Intrathecal injection of drugs

ZX1 (Cat. No: 07–0350) was purchased from Strem chemicals, TPEN (Cat. No: P4413), ZnCl₂ (Cat. No: 746355) were ordered from Sigma. TPEN were dissolved in 5% dimethyl sulfoxide (DMSO). Unless otherwise specified, other reagents are dissolved in sterile saline. Intrathecal injection of drugs or siRNA was performed as described previously (Wang and colleagues). For i.t. injection TPEN (100 μm) and ZnCl₂ (100 μm), a spinal cord puncture was made with a 30-gauge needle between L5 and L6 levels to deliver drugs (10 μl) to the CSF, and then performed behavioral tests after 1 h.

qRT-PCR

The mice were anesthetized and decapitated; DRG tissue was harvested and ground into powder using zirconia beads. Total RNAs were extracted using Trizol reagent. Reverse transcription of RNAs (0.5 to 2 μg) was performed using the 5 × All in One RT MasterMix (abm, G492). Power SYBR Green PCR Master Mix (Thermo 4367659) was used for the qRT-PCR experiment. For testing the expression of Tmem163 in the dorsal root ganglia (DRG) of wild-type (WT) mice and Tmem163 cKO mice, utilizing the following primer sequences: Forward primer: 5′-actgtctccgtcatgagg-3′, Reverse primer: 5′-agagtgcacagcagccgc-3′. The Real-time PCR was conducted following the protocol for Power SYBR Green Master Mix kit with primers presented in S1 Table. The expression levels of ZnTs were normalized to GAPDH using the 2^-ΔCt method.

Genotyping

To purify genomic DNA, mouse ear tissue was collected and digested in a solution containing 25 mM NaOH and 0.2 mM EDTA at 95°C for 1 h, followed by neutralization with 40 mM Tris-HCl (pH 5.5). Genotyping of Tmem163 flox mice was performed using the primer pair 5′-agggcccttatatatctcactgtaa-3′ and 5′-aaaaccacctgatgttgaactgtg-3′, resulting in a WT band size of 132 bp and a mutant band size of 202 bp. The primers for genotyping NPY::cre mice were 5′-acaccggccttattccaag-3′, 5′-tccatgatttgcctcttgtg-3′, and 5′-ctgagacatagattcgtccaagg-3′ with band size 100 bp. The primers for genotyping VGluT1::cre mice were 5′-ccctaggaatgctcgtcaag-3′, 5′-atgagcgag gagaagtgtgg-3′, and 5′-gtggaagtcctggaaactgc-3′ with band size 344 bp.

Plasmid construction and AAV construction

The full-length sequence of VGluT1 was amplified from mouse cDNA and cloned into pcDNA3.1, resulting in the generation of QE15-pcDNA3.1-VGluT1-mCherry. Truncations of VGluT1 were created using the KOD-Plus-Mutagenesis Kit (TOYOBO) and In-Fusion HD cloning kit (Takara). The pAKD-CMV-bGlobin-Tmem163-eGFP plasmid was obtained from Genechem company. Plasmids for Glycerol bacterium were provided by Genechem company for AAV construction. The coding sequence of Tmem163 was amplified from mouse cDNA and cloned into CMV-bGlobin-MCS-eGFP-3flag-Wpre-hGH-polyA. The synthesis and packaging of the AAV9-CMV-bGlobin-mTMEM163-EGFP-3FLAG-WPRE-hGH poly A virus were performed by Genechem company. AAV infection was achieved by intrathecal injection into the spinal cord.

Cell culture and plasmid transfection

Primary cultures of DRG neurons were prepared following the established protocol described previously. DRGs were dissected from postnatal day 0 to 10 (P0-P10) mice and subjected to digestion using collagenase (1 mg/ml, Sigma-Aldrich), trypsin (0.4 mg/ml, Sigma-Aldrich), and deoxyribonuclease I (0.1 mg/ml) for 45 min. Subsequently, non-neuronal cells were removed by centrifugation with 15% Percoll, and the purified DRG neurons were obtained. The neurons were then cultured in neurobasal medium supplemented with 2% B27 supplement, 2 mM L-glutamine (Invitrogen), and 100 ng/ml nerve growth factor (NGF; Invitrogen). The cells were seeded onto poly-D-lysine-coated glass cover slips and maintained in a neurobasal defined medium at 37°C with 5% CO₂ for 48 h before conducting experiments. The ND7/23 cell line was obtained from the Cell Bank of the Chinese Academy of Sciences. ND7/23 cells were cultured in DMEM media (without L-glutamine) supplemented with 10% FBS, 1% L-glutamine, and 1% PS at 37°C under 5% CO₂. Plasmid transfection was performed using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions. For electroporation of DRG neurons, cells were suspended in electroporation buffer and then electroporated using the Nucleofector II (Amaxa) with the O-003 program.

Live-cell imaging and data analysis

P10 DRG neurons that underwent electroporation were cultured on glass-bottom dishes placed on a Nikon temperature-controlled workstation maintained at 37°C. Live imaging of the neurons was performed using NIS-Elements viewer software. Specifically, axon segments located 100 to 150 μm from the cell body were carefully selected as regions of interest, and a total of 360 frames were acquired over a 6-min period. Subsequently, videos and individual frames were generated using Imaries software. The analysis of puncta was conducted using Image-Pro Plus 6.0 software.

Immunoelectron microscopy

Recombinant adenovirus AAV9-Tmem163-EGFP was generated and packaged for subsequent intrathecal injection. Following the injection, mice were perfused with a solution containing 4% PFA and 0.2% glutaraldehyde to facilitate post-embedding immunogold labeling. Vibratome sections from the cervical spinal cord segments were then post-fixed in 0.5% osmium tetroxide and processed for embedding in resin. Ultrathin sections of the dorsal horn were prepared and subjected to staining with specific antibodies. Specifically, mouse anti-GFP antibodies (1:200, Earthox, E22030-03) and guinea pig anti-VGluT1 antibodies (1:200, Millipore, Ab5905) were used. Immunogold particles measuring 10 nm (anti-mouse, 1:40; Amersham) and 6 nm (anti-guinea pig, 1:40; Jackson) were applied. The selection of different diameters of colloidal gold particles depended on the content of the molecules in the tissue and their activity after embedding. Subsequently, uranium staining was performed to visualize the localization of Tmem163 and VGluT1 molecules expressed on the vesicles under an electron microscope. To exclude nonspecific signals, we conducted 2 control experiments: a negative control and a no primary antibody control. For the negative control, naïve C57 mice were sampled and subjected to the same immunoelectron microscopy procedure described above. Ultrathin sections of the dorsal horn were stained with mouse anti-GFP and guinea pig anti-VGluT1 antibodies. Immunogold particles measuring 10 nm and 6 nm were used to label GFP and VGluT1, respectively. For the no primary antibody control, AAV9-Tmem163-EGFP infected mice were sampled and subjected to the same immunoelectron microscopy procedure described above. Ultrathin sections of the dorsal horn were incubated with primary antibody dilution buffer, followed by 10 nm and 6 nm immunogold particles.

Intra-spinal injection

Mice were anesthetized with pentobarbital sodium (75 mg/kg i.p.), and their eyes were covered with erythromycin eye ointment to prevent ophthalmic drying. The hair on the back, ranging from thoracic (T) 10 to lumbar (L) 2, was removed using a hair remover. Mice were then fixed in a stereotaxic instrument (RWD Lift Science Co.), and the spinal cord between T11-T12, T12-T13, or T13-L1 was exposed. A glass micropipette connected to Nanoject III (was purchased from Drummond Scientific Company) was utilized for injection. The injection site was tilted at a 30-degree angle and positioned 500 μm away from the midline, which could be identified by the blood vessel running along the longitudinal axis of the spinal cord. The surface of the spinal cord served as the starting point for the Z-axis, and the micropipette was pulled back 200 μm, resulting in the settlement of the micropipette tip 300 μm from the dorsal surface. A total volume of 300 nl of viral particles was injected at a rate of 60 nl/min. Following the injection, the micropipette was held in place for 5 min to prevent virus leakage. To identify the presynaptic neurons that directly project to NPY⁺ INs, a viral mixture consisting of 300 nl, containing equal volumes of AAV2/5-hEF1a-DIO-RVG-WPRE-pA (Taitool, S0325-5, 2.20E+12 vg/ml) and AAV2/9-CAG-DIO-TVA-mCherry-WPRE-pA (Taitool, S0683-9, 4.50E+11 vg/ml) was injected. The incision was closed using 4–0 sutures. Three weeks later, 300 nl of RV-ENVA-DG-EGFP (BrainVTA, R01002, 2.00E+08 IFU/ml) was injected at the same site. After 1 week, mice were sacrificed for histochemical analysis. To label spinal NPY⁺ interneurons in aged mice or young control mice, rAAV2/9-NPY-mCherry-WPRE-bGH polyA (BrainVTA, PT1216, titer: 6.05E+12 vg/ml) was delivered via intraspinal injection into the spinal L2-L4 regions of both young and aged C57/BL6 mice. After 4 weeks, the mice were sacrificed for whole-cell patch clamp recordings.

Spinal cord slice preparation and patch-clamp recordings

Mice were deeply anesthetized with pentobarbital sodium (75 mg/kg, i.p.). Transcardiac perfusion of ice-cold modified artificial cerebrospinal fluid (ACSF) containing (in mM): KCl 2, NaH₂PO₄ 1.3, CaCl₂ 0.2, MgSO₄ 12, NaHCO₃ 26, D-glucose 10, sucrose 220, sodium ascorbate 1.3, sodium pyruvate 3.5 (oxygenated with 95% O₂ and 5% CO₂, pH 7.4). The spinal cord slices from lumbar segments 4 to 5 were prepared in the transverse plane by quickly dissecting the lumbar spinal cord, placing it into ice-cold modified ACSF, and cutting it at 300 μm using a vibratome (VT1000S, Leica, Germany). Slices were incubated in a solution containing (in mM): NaCl 126, KCl 2.5, CaCl₂ 2, MgCl₂ 1, NaH₂PO₄ 1.25, NaHCO₃ 26, D-glucose 10, sodium ascorbate 1.3, sodium pyruvate 3.5 (oxygenated with 95% O₂ and 5% CO₂) for 30 min and at RT for 1 h. The slices were transferred into a recording chamber and perfused with oxygenated recording solution at a rate of 5 ml/min prior to electrophysiological recordings at RT.

Whole-cell patch clamp recordings were performed on laminae III-V NPY⁺ neurons in voltage-clamp mode. Patch pipettes (5 to 10 M) from borosilicate glass (Sutter, USA) were made on a horizontal micropipette puller (P-1000, Sutter, USA). Data were acquired with pClamp10.0 software (Molecular Devices, USA) using MultiClamp700B patch-clamp amplifier and Digidata1550B (Molecular Devices, USA). For minimal excitatory postsynaptic current (mEPSC) recording, pipette solution contained (in mM): Cs-methanesulfonate 120, CsCl 15, Na2 creatine pgosphate 10, HEPES 10, Mg-ATP 2, Na-GTP 0.3, QX-314 2.5, EGTA 1, with pH at 7.3 and measured osmolarity at 300 to 310 mOsm. After establishing the whole-cell configuration, neurons were holding at −70 mV to record mEPSCs in the presence of bicuculline (100 μm), strychnine (2 μm), and TTX (0.5 μm). AMPA receptor-mediated EPSCs were elicited by electrical stimulation in the presence of bicuculline (10 mM) and strychnine (2 mM). NMDA receptor-mediated EPSCs were recorded using a low Mg²⁺ recording ACSF (2.5 mM Ca²⁺, 0.25 mM Mg²⁺) supplemented with NBQX (10 μm), bicuculline (10 μm), and strychnine (2 μm). The holding potential was set at −40 mV, as specified. EPSCs were evoked by applying a constant current pulse at 0.05 Hz. The signals were filtered at 3 kHz and digitized at 10 kHz.

Ethics statement

All experiments were approved by the Experimental Animal Ethics Committee of Shanghai Medical College and the Committee on the Use of Animal Experiments of Fudan University (approval number: SYXK2020-0032). For human CSF collection, the study was approved by the Ethics Committee of Huashan Hospital, Fudan University (approval number: KY2020-1218). The samples were voluntarily donated by patients diagnosed with solitary meningioma, who provided written informed consent for the anonymous use of their cerebrospinal fluid in research. The study was conducted in compliance with the principles outlined in the Declaration of Helsinki.

Quantification and statistical analysis

All data are presented as bar graphs indicating mean ± SEM (standard error of the mean). Dots in bar graphs and boxplots represent individual values per mouse or per image, horizontal line indicates average. GraphPad Prism scientific software version 8.4.0 was used for statistical analysis. For normally distributed, related samples with equal variances, unpaired t test and paired t test was used to compare 2 groups. Conversely, samples with unequal variances via Welch’s t test. Statistical significance was determined using both nonparametric (Mann–Whitney, Kruskal–Wallis test with Dunn’s multiple-comparisons post hoc test) and parametric methods (one-way ANOVA followed by Dunnett’s multiple comparisons, two-way ANOVA with Sidak’s multiple comparisons Bonferroni post hoc analysis) due to differences in the normality and sample size of the data. The specific tests used to analyze each data set is indicated within the individual figure legends, with results expressed as mean ± SEM. Significance levels are indicated as *p < 0.05, **p < 0.01, ***p < 0.001.