Spinal cord injury (SCI) at high spinal levels (e.g., above thoracic level 5) causes systemic immune suppression; however, the underlying mechanisms are unknown. Here we show that profound plasticity develops within spinal autonomic circuitry below the injury, creating a sympathetic anti-inflammatory reflex, and that chemogenetic silencing of this reflex circuitry blocks post-SCI immune suppression. These data provide new insights and potential therapeutic options for limiting the devastating consequences of post-traumatic autonomic hyperreflexia and post-injury immune suppression.
Subscribe to Journal
Get full journal access for 1 year
only $4.92 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Blackmer, J. CMAJ 169, 931–935 (2003).
Inskip, J.A., Ramer, L.M., Ramer, M.S. & Krassioukov, A.V. Spinal Cord 47, 2–35 (2009).
Zhang, Y. et al. J. Neurosci. 33, 12970–12981 (2013).
Lucin, K.M., Sanders, V.M., Jones, T.B., Malarkey, W.B. & Popovich, P.G. Exp. Neurol. 207, 75–84 (2007).
Lucin, K.M., Sanders, V.M. & Popovich, P.G. J. Neurochem. 110, 1409–1421 (2009).
Meisel, C., Schwab, J.M., Prass, K., Meisel, A. & Dirnagl, U. Nat. Rev. Neurosci. 6, 775–786 (2005).
Smith, B.N. et al. Proc. Natl. Acad. Sci. USA 97, 9264–9269 (2000).
Cano, G., Sved, A.F., Rinaman, L., Rabin, B.S. & Card, J.P. J. Comp. Neurol. 439, 1–18 (2001).
Anderson, C.R. & Edwards, S.L. J. Neurosci. Methods 53, 137–141 (1994).
Vong, L. et al. Neuron 71, 142–154 (2011).
Nakamura, T., Colbert, M.C. & Robbins, J. Circ. Res. 98, 1547–1554 (2006).
Hou, S. et al. J. Comp. Neurol. 509, 382–399 (2008).
Duale, H., Lyttle, T.S., Smith, B.N. & Rabchevsky, A.G. J. Neurotrauma 27, 1369–1378 (2010).
Sternson, S.M. & Roth, B.L. Annu. Rev. Neurosci. 37, 387–407 (2014).
Brommer, B. et al. Brain 139, 692–707 (2016).
Iversen, P.O. et al. Blood 96, 2081–2083 (2000).
Rossi, J. et al. Cell Metab. 13, 195–204 (2011).
Banfield, B.W., Kaufman, J.D., Randall, J.A. & Pickard, G.E. J. Virol. 77, 10106–10112 (2003).
Paxinos, G. The Mouse Brain in Stereotaxic Coordinates (Academic, 2001).
Yang, B. et al. Cell 158, 945–958 (2014).
We would like to thank L. Enquist (Princeton University) and the Center for Neuroanatomy with Neurotropic Viruses (CNNV; NIH grant P40RR018604) for providing PRV; Y. Zhang for advice on SCI surgery; Z. Gu and S. Goyama for comments, P. Thanh, A. Epstein, M. Sandy and M. Maezawa for help with experiments, and M. Muntifering for the clearing technique. This work is supported by NINDS-NS083942 and the Ray W. Poppleton Endowment (P.G.P.); NINDS-NS093002 (Y.Y.); PRESTO (JST), JSPS Postdoctoral Fellowships for Research Abroad and KANAE Foundation for the Promotion of Medical Science (M.U.).
The authors declare no competing financial interests.
Integrated supplementary information
(a) Gross anatomy of representative spleens from control and T3 SCI mice at 28 days post-injury. (b and c) Hematoxylin and eosin staining (HE) of spleen tissue sections. (d and e) Immunostaining of CD45R (B220; green), a marker of B lymphocytes, with DAPI (blue). Scale bars, 2 mm (a); 500 μm (b, c); 100 μm (d, e).
Supplementary Figure 2 Distribution of PRV-positive cells throughout the central nervous system (CNS) that are directly or indirectly connected to the spleen, i.e., a neural-splenic circuit.
(a and b) Post-ganglionic adrenergic sympathetic nerves innervate the spleen. PACT clearing of spleen (a) and staining with anti-tyrosine hydroxylase (TH) antibody (b, red). TH+ fibers broadly innervate the spleen, which allows PRV to be transported retrogradely into sympathetic chain ganglia and then into the CNS. Scale bar, 2 mm. (c) Schema of retrograde and transsynaptic labeling of neural-splenic circuit by PRV. (d) Distribution summary of PRV-GFP positive cells in the CNS, 72–144 hours after PRV injection into the spleen (modified from8). Time points when PRV-GFP+ cells first appeared are indicated as black to white colors in each CNS region. (e–x) Representative images of PRV-GFP+ cells in different CNS regions and time points after PRV152 injection into the spleen. Mice were perfused at 48 (day 2, n = 2 mice), 72 (day 3, n = 2 mice), 96 (day 4, n = 2 mice), 108 (day 4.5, n = 1 mouse), 120 (day 5, n = 2 mice), 144 (day 6, n = 3 mice), and 168 hr (day 7, n = 2 mice) after injection. To verify that PRV labeling from the spleen occurs via the splenic nerve and is not the result of spurious labeling due to blood-borne virus, the splenic nerve was cut before PRV injection then analyzed at 120 hr (n = 2 mice). Thoracic cord (T4–8), e (72 hr), f (96 hr), g (120 hr); dorsal motor nucleus of vagus (10N), h (108 hr), i (120 hr), j (120 hr, splenic nerve cut); Reticular formation, k (120 hr), l (144 hr), m (120 hr, splenic nerve cut); paraventricular nucleus (PVN), n (120 hr), o (144 hr), p (120 hr, splenic nerve cut); PsTh, q (120 hr), r (144 hr); zona incerta (ZI), s (120 hr), t (144 hr); locus coeruleus (LC), u (144 hr); periaqueductal gray (PAG), v (144 hr); basolateral amygdaloid nucleus (BLP), w (144 hr); bed nucleus of the stria terminalis (BST), x (144 hr). Coronal sections. Scale bar, 200 μm.
Supplementary Figure 3 PRV+ spinal sympathetic neurons innervating the spleen in control and SCI mice.
(a–i) Horizontal sections of spinal cord in control (a–c), T9 SCI (d–f), and T3 SCI mice (g–i), 96 hrs after PRV injection into the spleen. Stained with anti-GFP antibody (green). T2–4 (a, d, g), T9–11 (b, e, h), and L5–S1 level (c, f, i). Arrowheads, lesion sites (e and g). Rostral is left. Scale bar, 200 μm. Images are representative of 4 (T9–11) and 5 (T2–4 and L5–S1) independent experiments. (j–m) Distribution of PRV+ cells in dorso-ventral axis of thoracic cord after T3 SCI. (j) Schematic figure of transverse section of thoracic cord. Dotted lines represent approximate dorsoventral positions of images shown in k, l, m, and Fig. 1c, f. (k–m) Horizontal images of PRV-GFP+ cells distributed in dorsoventral axis. Rostral is left. Scale bar, 100 μm. (n–q) Representative images of Vglut2-GFP+ (green) / PRV+ (RFP, red) neurons in the sacral spinal cord after T3 SCI. Right panels (o–q) are higher magnification of the neurons indicated by arrows in (n). 76.7 ± 2.86% of PRV+ were Vglut2-GFP+ neurons at the sacral level (n = 2 mice). Scale bars, 100 μm (n), 50 μm (o–q).
Supplementary Figure 4 Time-dependent increase in number of VGLUT2+ pre-synaptic contacts on SPNs after T3 SCI.
(a) Schematic illustrating intrasplenic injections of PRV and intraperitoneal injections of FG to distinguish SPNs (FG+/PRV+) from interneurons (FG–/PRV+) in spinal cord-splenic circuit. (b–g) Representative confocal images of FG+ (blue) / PRV+ (green) SPNs and VGLUT2+ puncta (red) in control (b, c) and at 10 (d, e) and 28 days post-injury (f, g). Images are representative of 2, 4, and 5 independent experiments for control, 7–10 days and 28 days. SPNs in (b), (d) and (f) (arrowheads) are reconstructed into 3D images using Imaris software are isolated in (c), (e) and (g), respectively. Scale bar, 40 μm. (h) Quantification of the density of VGLUT2+ synaptic puncta on FG+/PRV+SPNs at day 7–10 and day 28, the time points before and after onset of immune suppression in T3 SCI mice3. Dotted line indicates the density of VGLUT2+ synaptic puncta quantified from naïve/uninjured mice (n = 2 mice, 44 total neurons) and is provided as a reference for post-injury changes in puncta density. Data are represented as mean ± s.e.m (day 7, n = 3 mice; day 10, n = 1 mouse; day 28, n = 5 mice). Student’s t test, p = 0.0434, *p < 0.05.
Supplementary Figure 5 Activating visceral-sympathetic reflexes via CRD induces c-Fos expression in Vglut2+ excitatory spinal interneurons.
(a and b) CRD induced c-Fos expression (red) in the thoracic cord of T3 SCI mice. T6–8 level, horizontal sections. T3 SCI without (a) and with CRD (b) at day 28. (c) Quantitative analyses of c-Fos+ cells at T4–9 level of control or T3 SCI mice with or without CRD. Data are represented as mean ± s.e.m (T3 SCI with CRD, n = 5 mice, other groups, n = 3 mice). One way ANOVA followed by Tukey test, p = < 0.0001, < 0.0001, 0.0013 (left to right), **p < 0.01. (d–f) c-Fos expression (red) in GFP+ cells (green) in Vglut2-Cre; cc-GFP mice with T3 SCI and CRD at day 28 (arrows). Scale bars, 200 μm (a, b); 50 μm (d–f). (g) Quantification of Vglut2-GFP+ and Chat-GFP+ population in c-Fos-expressing cells with T3 SCI and CRD at day 28 (n = 3 mice). Images are representative of 3 (a, d–f) and 5 (b) independent experiments.
(a) Design of hM4Di-mCherry-expressing AAV, using FLEX switch strategy to induce Cre-mediated transgene inversion and expression in the thoracic spinal cord. (b) hM4Di-mCherry expression in Vglut2-Cre mice. T6–8 level at day 28 after T3 SCI. (c–e) hM4Di-mCherry (red) is expressed in GFP+ neurons in Vglut2-Cre; cc-GFP mice (arrows). (f) hM4Di-mCherry expression in Chat-Cre mice. T6–8 level at day 28 after T3 SCI. (g–i) hM4Di-mCherry (red) is expressed in GFP+ neurons in Chat-Cre; cc-GFP mice (arrows). Note that most hM4Di-mCherry+ cells are located in the intermediate and medial zone. Few exist in the lateral zone where SPNs predominate. Images are representative of 2 independent experiments. Scale bars, 500 μm (b, f); 50 μm (c–e, g–i). (j) Percentage of hM4Di-mCherry+/GFP+ cells in total hM4Di-mCherry+ cells in Vglut2-Cre; cc-GFP and Chat-Cre; cc-GFP mice. Data are represented as mean ± s.e.m (n = 4 horizontal sections from 2 mice). (k) A representative image of brain stem region (reticular formation) in Vglut2-Cre mice at day 28 after T3 SCI, injected with AAV-hSyn-DIO-hM4Di-mCherry at thoracic level at P14. No hM4Di-mCherry was visible in any sections indicating that the AAV vector does not get retrogradely transported into the brain stem. Thus, no descending modulation by brainstem circuitry can contribute to hM4Di-mediated inhibition of spinal circuitry. The image is representative of 2 independent experiments. Scale bar, 200 μm.
Supplementary Figure 7 Chemogenetic silencing of spinal autonomic circuitry reduces c-Fos expression elicited by CRD, a potent visceral-sympathetic reflex stimulus.
(a–f) c-Fos expression in horizontal thoracic spinal cord sections of AAV-hSyn-DIO-hM4Di-mCherry-injected T3 SCI Vglut2-Cre mice injected with vehicle (a–c) or CNO and with CRD (d–f). Scale bar, 50 μm. Images are representative of 3 independent experiments. (g, h) Quantification of density of c-Fos+ cells and percentage of cFos+/ hM4Di-mCherry+ cells. Data are represented as mean ± s.e.m (n = 3 mice). Student’s t test; p = 0.0485 (g), 0.0187 (h); *p < 0.05.
Supplementary Figure 8 Representative flow cytometry scatter plots of splenocytes derived from control and SCI mice with or without chemogenetic manipulation of autonomic circuitry.
(a, c, e, g, i) Gating strategies using FSC-A vs SSC-A, FSC-H vs FSC-W and SSC-H vs SSC-W plots (to remove doublets). (b, d, f, h, j) Dot plots of B220, CD4, and CD8-labeled lymphocytes from WT control (b), WT + T3 SCI (d), WT + AAV-hSyn-DIO-hM4Di-mCherry + T3 SCI + CNO (f), Vglut2-Cre + AAV-hSyn-DIO-hM4Di-mCherry + T3 SCI + CNO (h), and Chat-Cre + AAV-hSyn-DIO-hM4Di-mCherry group (j). Mean percentage of B220+, CD4+ or CD8+ cells generated by gating for each cell population was multiplied by total number of splenocytes to generate graphs in Fig. 3e–g.
Supplementary Figure 9 Proposed model by which activation of a sympathetic anti-inflammatory reflex (SAR) causes immune suppression after high-level SCI.
(a) Normally, there is tonic brainstem control over sympathetic preganglionic neurons (SPNs). This tonic regulation is essential for normal neural-immune communication and maintenance of homeostasis. (b) After high-level T3 SCI (red dotted lines), supraspinal control over SPNs is lost allowing normally innocuous stimuli below the level of injury to elicit exaggerated and potentially life-threatening autonomic reflexes (gray dotted lines). New data show that profound plasticity develops within spinal autonomic circuitry below the level of SCI. Notably, new connections form between SPNs and spinal excitatory interneurons throughout and beyond the thoracic spinal cord (Supplementary Fig. 3i and n–q). This rewired circuit is the neural substrate underlying SAR-mediated suppression of immune function.
About this article
Cite this article
Ueno, M., Ueno-Nakamura, Y., Niehaus, J. et al. Silencing spinal interneurons inhibits immune suppressive autonomic reflexes caused by spinal cord injury. Nat Neurosci 19, 784–787 (2016). https://doi.org/10.1038/nn.4289
Pharmacological Inhibition of Soluble Tumor Necrosis Factor-Alpha Two Weeks after High Thoracic Spinal Cord Injury Does Not Affect Sympathetic Hyperreflexia
Journal of Neurotrauma (2021)
The Journal of Spinal Cord Medicine (2021)
Acute post-injury blockade of α2δ-1 calcium channel subunits prevents pathological autonomic plasticity after spinal cord injury
Cell Reports (2021)
Experimental Neurology (2021)