Signalling between cells of the neurovascular unit, or neurovascular coupling, is essential to match local blood flow with neuronal activity. Pericytes interact with endothelial cells and extend processes that wrap capillaries, covering up to 90% of their surface area1,2. Pericytes are candidates to regulate microcirculatory blood flow because they are strategically positioned along capillaries, contain contractile proteins and respond rapidly to neuronal stimulation3,4, but whether they synchronize microvascular dynamics and neurovascular coupling within a capillary network was unknown. Here we identify nanotube-like processes that connect two bona fide pericytes on separate capillary systems, forming a functional network in the mouse retina, which we named interpericyte tunnelling nanotubes (IP-TNTs). We provide evidence that these (i) have an open-ended proximal side and a closed-ended terminal (end-foot) that connects with distal pericyte processes via gap junctions, (ii) carry organelles including mitochondria, which can travel along these processes, and (iii) serve as a conduit for intercellular Ca2+ waves, thus mediating communication between pericytes. Using two-photon microscope live imaging, we demonstrate that retinal pericytes rely on IP-TNTs to control local neurovascular coupling and coordinate light-evoked responses between adjacent capillaries. IP-TNT damage following ablation or ischaemia disrupts intercellular Ca2+ waves, impairing blood flow regulation and neurovascular coupling. Notably, pharmacological blockade of Ca2+ influx preserves IP-TNTs, rescues light-evoked capillary responses and restores blood flow after reperfusion. Our study thus defines IP-TNTs and characterizes their critical role in regulating neurovascular coupling in the living retina under both physiological and pathological conditions.
This is a preview of subscription content, access via your institution
Open Access articles citing this article.
Functional ultrasound localization microscopy reveals brain-wide neurovascular activity on a microscopic scale
Nature Methods Open Access 04 August 2022
Cell Death & Disease Open Access 05 July 2022
Cell & Bioscience Open Access 19 May 2022
Subscribe to Nature+
Get immediate online access to the entire Nature family of 50+ journals
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Get time limited or full article access on ReadCube.
All prices are NET prices.
All the data analysed in this study, including raw data, are provided in this published article and the supplementary information files. There are no restrictions on data availability. Source data are provided with this paper.
The R code used for data analysis is included in the Supplementary Information files.
Attwell, D., Mishra, A., Hall, C. N., O’Farrell, F. M. & Dalkara, T. What is a pericyte? J. Cereb. Blood Flow Metab. 36, 451–455 (2016).
Dalkara, T. & Alarcon-Martinez, L. Cerebral microvascular pericytes and neurogliovascular signaling in health and disease. Brain Res. 1623, 3–17 (2015).
Alarcon-Martinez, L. et al. Capillary pericytes express α-smooth muscle actin, which requires prevention of filamentous-actin depolymerization for detection. eLife 7, e34861 (2018).
Hall, C. N. et al. Capillary pericytes regulate cerebral blood flow in health and disease. Nature 508, 55–60 (2014).
Hamilton, N. B., Attwell, D. & Hall, C. N. Pericyte-mediated regulation of capillary diameter: a component of neurovascular coupling in health and disease. Front. Neuroenergetics 2, 5 (2010).
Zhu, X., Bergles, D. E. & Nishiyama, A. NG2 cells generate both oligodendrocytes and gray matter astrocytes. Development 135, 145–157 (2008).
Henle, F. Ueber die Ausbreitung des Epithelium im menschlichen Körper. Arch. Anat. Physiol. Wiss. Med. 1838, 103–128 (1838).
Cerletti, U. Die gefassvermehrung in zentralnervensystem. Histol. Histopathol. Arb. Grosshirnr. 4, 1–168 (1910).
Cammermeyer, J. A comparative study of intervascular connective tissue strands in the central nervous system. J. Comp. Neurol. 114, 189–208 (1960).
Reissenweber, N. J. & Pessacq, T. Intervascular strands in the central nervous system. A histochemical approach. Acta Anat. 78, 51–57 (1971).
Williamson, J. R., Tilton, R. G., Kilo, C. & Yu, S. Immunofluorescent imaging of capillaries and pericytes in human skeletal muscle and retina. Microvasc. Res. 20, 233–241 (1980).
Leibnitz, L. & Bär, B. A blood capillaries-bridging cell type in adult mammalian brains. J. Hirnforsch. 29, 367–375 (1988).
Mendes-Jorge, L. et al. Intercapillary bridging cells: immunocytochemical characteristics of cells that connect blood vessels in the retina. Exp. Eye Res. 98, 79–87 (2012).
Korenkova, O., Pepe, A. & Zurzolo, C. Fine intercellular connections in development: TNTs, cytonemes, or intercellular bridges? Cell Stress 4, 30–43 (2020).
Harris, A. L. Connexin channel permeability to cytoplasmic molecules. Prog. Biophys. Mol. Biol. 94, 120–143 (2007).
Sartori-Rupp, A. et al. Correlative cryo-electron microscopy reveals the structure of TNTs in neuronal cells. Nat. Commun. 10, 342 (2019).
Theofilas, P., Steinhäuser, C., Theis, M. & Derouiche, A. Morphological study of a connexin 43-GFP reporter mouse highlights glial heterogeneity, amacrine cells, and olfactory ensheathing cells. J. Neurosci. Res. 95, 2182–2194 (2017).
Rustom, A., Saffrich, R., Markovic, I., Walther, P. & Gerdes, H.-H. Nanotubular highways for intercellular organelle transport. Science 303, 1007–1010 (2004).
Pham, A. H., McCaffery, J. M. & Chan, D. C. Mouse lines with photo-activatable mitochondria to study mitochondrial dynamics. Genesis 50, 833–843 (2012).
Osswald, M. et al. Brain tumour cells interconnect to a functional and resistant network. Nature 528, 93–98 (2015).
Wang, X., Veruki, M. L., Bukoreshtliev, N. V., Hartveit, E. & Gerdes, H.-H. Animal cells connected by nanotubes can be electrically coupled through interposed gap-junction channels. Proc. Natl Acad. Sci. USA 107, 17194–17199 (2010).
Longden, T. A. et al. Capillary K+-sensing initiates retrograde hyperpolarization to increase local cerebral blood flow. Nat. Neurosci. 20, 717–726 (2017).
Nilsson, S. F. & Alm, A. in Ocular Blood Flow (eds Kiel, J. W. & Schmetterer, L.) 25–47 (Springer-Verlag, 2012).
Alarcon-Martinez, L. et al. Retinal ischemia induces α-SMA-mediated capillary pericyte contraction coincident with perivascular glycogen depletion. Acta Neuropathol. Commun. 7, 134 (2019).
Underly, R. G. & Shih, A. Y. Photothrombotic induction of capillary ischemia in the mouse cortex during in vivo two-photon imaging. Bio Protoc. 7, e2378 (2017).
Borysova, L., Wray, S., Eisner, D. A. & Burdyga, T. How calcium signals in myocytes and pericytes are integrated across in situ microvascular networks and control microvascular tone. Cell Calcium 54, 163–174 (2013).
Franco, C. A. et al. Dynamic endothelial cell rearrangements drive developmental vessel regression. PLoS Biol. 13, e1002125 (2015).
Yu, D.-Y. et al. Retinal capillary perfusion: spatial and temporal heterogeneity. Prog. Retin. Eye Res. 70, 23–54 (2019).
Mouton, P. R. Principles and Practices of Unbiased Stereology: An Introduction for Bioscientists (Johns Hopkins Univ. Press, Baltimore, 2002).
Takihara, Y. et al. In vivo imaging of axonal transport of mitochondria in the diseased and aged mammalian CNS. Proc. Natl Acad. Sci. USA 112, 10515–10520 (2015).
We thank T. Kennedy, N. Belforte, P. D’Onofrio, Y. Shiga, and R. Araya for comments on the manuscript; J. Mui and W. Leelapornpisit for assistance with sample preparation, microscope operation and data collection; and G. Seifert and C. Steinhäuser for providing eyes from CX43–ECFP mice. Supported by grants from the Canadian Institutes of Health Research (A.D.P.).
The authors declare no competing interests.
Peer review information Nature thanks Mark Nelson, Frank Winkler, Chiara Zurzolo and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
a–d, Representative examples of NG2–DsRed retinas labelled with DAPI (blue) and lectin (green) in combination with 3D reconstruction (right panel for each set) showing a pericyte and its IP-TNT (dotted line), termed here the proximal pericyte, connecting with the processes of a pericyte located on a distal capillary (distal pericyte). IP-TNTs connected a proximal pericyte with a distal one in all retinas analysed. Pericytes and their processes are shown in red. e, Single-pericyte electroporation of fluorescein-coupled dextran (3,000 Da) into a DsRed-positive pericyte with an IP-TNT (dotted line). Dextran entered the proximal pericyte (green) and IP-TNT but did not diffuse into the distal pericyte (red), allowing clear identification of the two connected pericytes. e′, Higher-magnification inset shows the interface between the two IP-TNT-coupled pericytes at the level of the IP-TNT end-foot (ef, green) and the distal pericyte process (dpp, red). f, 3D reconstruction unambiguously shows that the two pericytes are in contact at the level of the IP-TNT end-foot and distal pericyte process. Experiments were replicated 4 times.
Extended Data Fig. 2 IP-TNT characterization: localization of molecular markers and presence in the brain.
a, b, NG2–DsRed and lectin-labelled IP-TNTs (dotted line) show absence of immunolabelling for the endothelial cell markers VE-cadherin and CD31. c, Labelling for the basal lamina component laminin confirmed that IP-TNTs are enclosed by a basement membrane. d–f, Representative IP-TNT from CX43–ECFP reporter mouse retina showing endogenous CX43 expression in the IP-TNT end-foot (ef; d). IP-TNTs contain the contractile protein α-SMA (e), but not α-tubulin (f). g, h, Immunolabelling and 3D reconstruction demonstrated that FAK and S6 proteins are localized in IP-TNTs, notably at end-feet (arrowheads). i–l, IP-TNTs do not contain either GAP43 or the mesenchymal stem cell markers CD90, CD44 or CD45. Positive controls for each antibody were postnatal day 5 retinal ganglion cell axons for GAP43 (i′), retinal ganglion cells for CD90 (j′) and peripheral leukocytes for CD44 (k′) and CD45 (l′). m, n, IP-TNTs detected in the visual cortex shown in fixed tissue (m) and by TPLSM in vivo (n). A minimum of 100 IP-TNT-connected pericytes from 5 different animals were analysed for each marker.
Extended Data Fig. 3 IP-TNTs exist in all retinal vasculature plexuses and connect distal capillaries of the same and different orders.
a, Schematic of the systematic uniform random sampling method used for histological quantification in whole-mounted retinas. b, Quantitative analysis demonstrated that IP-TNTs are abundant in all plexuses of the retinal vasculature, with a higher density in the intermediate and deep plexuses than in the superficial plexus (n = 3 mice, two-tailed ANOVA Tukey’s test, ***P < 0.001). c, 3D reconstruction of a representative vascular network in which each branch order has been pseudocoloured. The first-order vessel was defined as the first branch division (arteriole) from the central retinal artery, and subsequent branches were assigned consecutively higher orders (n = 3 mice). d, Most IP-TNT-linked capillaries (87%) originate from different branches (87%), while the remainder derive from the same branch (n = 94 IP-TNTs/capillaries, n = 3 mice). e, IP-TNTs connected pericytes located either on same-order vessels (values on diagonal line) or different-order vessels (values above or below diagonal line) (n = 94 IP-TNTs/capillaries, n = 3 mice). f, Analysis of the number of IP-TNTs per vessel order shows that these processes are more prevalent in fourth- to eighth-order vessels (n = 3 mice). Total sampled area = 560 μm × 451 μm × 10 disectors = 2,525,600 μm2. Data are presented as mean values ± s.e.m.
a, b, Time-lapse recordings of ICWs propagation between a proximal and a distal pericyte, visualized with lectin, connected by an IP-TNT (dotted line) in NG2–GCaMP6 retinas. Panels in b show the propagation of ICWs over time from the proximal pericyte (1), through the IP-TNT, onto the distal pericyte (2). See Supplementary Video 3. Experiments were replicated 5 times.
Extended Data Fig. 5 Live TPLSM imaging of light-evoked haemodynamic responses of retinal capillaries.
a, TPLSM recordings of capillary diameter changes at pericyte soma in NG2–DsRed mice subjected to a flash light stimulus (102 cd m–2, 5 ms). A linear probe perpendicular to the capillary at a pericyte location denotes the site where capillary diameter was measured (dotted line). Experiments were replicated in 211 capillaries, n = 4 mice. b, Longitudinal analysis of diameter changes by TPLSM in vivo shows little change in capillary diameter before light stimulation (dilation: n = 20 capillaries, n = 4 mice; constriction: n = 18 capillaries, n = 4 mice). c, By contrast, light provoked coordinated capillary dilation (blue: n = 46 capillaries, n = 4 mice) and constriction (red: n = 60 capillaries, n = 4 mice). d, Maximum response graph confirms substantial light-evoked capillary dilation and constriction after a flash light stimulus (pre-stimulus: n = 105 capillaries, n = 4 mice; dilation: n = 46 capillaries, n = 4 mice; constriction: n = 60 capillaries, n = 4 mice; two-tailed ANOVA Tukey’s test, ***P < 0.001). e–g, The overall response of arterioles (labelled with FITC in e), identified by a diameter larger than 9 μm, was substantially slower than that of capillaries (arterioles: n = 15 vessels, n = 3 mice; capillaries: n = 18 vessels, n = 3 mice; two-tailed Mann–Whitney U test, *P = 0.026). Data are presented as mean values ± s.e.m.
a, b, The laser parameters used to ablate IP-TNTs did not damage endothelial cells, even when aimed directly at capillaries, or alter blood retinal barrier integrity, as there was no leakage of FITC-coupled dextran (70 kDa, a) or fluorescein (332 Da, b). Experiments were replicated 3 times. c, Longer laser exposure time directed at endothelial cells in a single capillary (white circle) was required to cause damage (>4 s), as assessed by leakage of FITC-coupled dextran. Experiments were replicated 3 times. d, e, Ca2+ signals in endothelial cells after laser ablation, visualized with Fluo-4-AM, were indistinguishable from those in non-laser-ablated controls, suggesting that this procedure caused no major alterations (n = 9 capillaries per group, n = 2 mice per group; two-tailed Student’s t-test, n.s., not significant). Data are presented as mean values ± s.e.m.
a–c, The same laser parameters used to ablate IP-TNTs were applied to adjacent retinal ganglion cells (white circle). Experiments were replicated in 18 laser spots, n = 3 mice. d, e, Longitudinal analysis of light-evoked capillary dilation (blue; n = 15 capillaries, n = 3 mice) and constriction (red; n = 21 capillaries, n = 3 mice) demonstrated that microvessel dynamics were preserved under these conditions (pre-stimulus: n = 79 capillaries, n = 3 mice; dilation: n = 15 capillaries, n = 3 mice; constriction: n = 21 capillaries, n = 3 mice; two-tailed ANOVA Tukey’s test, *P < 0.05, ***P < 0.001). f–h, An AAV was used to selectively express GCaMP6 in retinal ganglion cells and monitor their ability to respond to light following laser ablation. Identical light-evoked Ca2+ responses were recorded before and after laser exposure, confirming the functional integrity of these neurons (n = 11 retinal ganglion cells per group, n = 3 mice per group; two-tailed Student’s t-test, n.s., not significant). i, j, Retinas were extracted after the experiment and immunostained with an antibody against RBPMS (specific for retinal ganglion cells), and neuronal density was analysed at the site of laser application (white circle within square). Experiments were replicated in 39 laser spots, n = 4 mice. k, Quantification of retinal ganglion cell numbers across a 2,000-μm2 disector area centred at the laser site revealed no significant difference in neuronal density relative to equivalent non-lasered areas (n = 43 control disectors, n = 39 laser-treated disectors, n = 4 mice per group; two-tailed Student’s t-test, n.s., not significant). Data are presented as mean values ± s.e.m.
Extended Data Fig. 8 Ischaemia leads to pericyte-mediated capillary constriction and IP-TNT rupture.
a, In vivo TPLSM imaging of TRITC-lectin-labelled ischaemic retina, showing constricted capillary sites at pericyte locations (arrowheads). Experiments were replicated 3 times. b, c, Quantification of capillary diameter (b; n = 6 mice per group; two-tailed Student’s t-test, *P = 0.049) and number of constrictions (c; sham: n = 4 mice; ischaemia: n = 6; two-tailed Student’s t-test, *P = 0.023) in whole-mounted ischaemic and sham-operated control retinas. d, Frequency distribution of the number of constrictions relative to the degree of decrease in capillary diameter (%) in sham-operated controls (blue trace) and ischaemic retinas (red trace) (sham: n = 4 mice; ischaemia: n = 6; two-tailed Student’s t-test, –70% *P = 0.013, –50% *P = 0.019, –40% **P = 0.004, –20% **P = 0.005, –10% *P = 0.048). e, Changes in diameter of large vessels in ischaemic and sham-operated whole-mounted retinas (n = 6 mice per group, two-tailed Student’s t-test, n.s., not significant). f–h, Analysis of IP-TNTs in NG2–DsRed retinas before and after oxygen–glucose deprivation (OGD), to mimic ischaemia ex vivo. Arrowheads point at an intact IP-TNT before damage (f) and to the rupture of the process after 20 and 60 min of ischaemia with extravasation of DsRed into the extracellular space (g, h). Experiments were replicated 5 times. i, j, Quantification of the numbers of pericytes, visualized in Ds-Red retinas (i; sham: n = 4 mice; ischaemia: n = 5; two-tailed Student’s t-test, n.s., not significant), and retinal ganglion cells, visualized with the retinal ganglion cell marker RBPMS (j; n = 4 mice per group; two-tailed Student’s t-test, n.s., not significant), showed that these cells do not die in ischaemic retinas (1 h ischaemia) relative to sham-operated controls. k, Induction of ischaemic retinal microregions after tail vein injection of Rose bengal by application of a laser to a single capillary (arrow in inset), which causes blood cells to stall in photocoagulated capillaries (arrowheads in inset). Experiments were replicated in 17 laser spots; n = 7 mice. l, Light-evoked capillary responses were substantially reduced in ischaemic retinal microregions induced by Rose bengal photocoagulation (dilation: sham, n = 16 capillaries, n = 4 mice, reperfusion, n = 17 capillaries, n = 4 mice; constriction: sham, n = 13 capillaries, n = 4 mice, reperfusion, n = 5 capillaries, n = 4 mice; two-tailed Student’s t-test, **P = 0.009, ***P < 0.001). m, Quantitative analysis confirmed a substantial loss of IP-TNTs in ischaemic versus non-ischaemic microregions within the same retina (n = 9 microregions per group, n = 3 mice per group; two-tailed Student’s t-test, **P = 0.008). Data are presented as mean values ± s.e.m.
Extended Data Fig. 9 Ca2+ influx during ischaemia disrupts IP-TNT-mediated ICWs, and the Ca2+ blocker nifedipine protects IP-TNTs.
a–c, Imaging of NG2–GCaMP6 mouse retinas showed that transient retinal ischaemia triggered a significant increase in intracellular Ca2+ in pericytes and their IP-TNTs relative to those in sham-operated controls (sham: n = 114 pericytes/IP-TNTs, n = 3 mice; ischaemia: n = 70 pericytes/IP-TNTs, n = 3 mice, two-tailed Mann–Whitney U test, ***P < 0.001). d–f, The frequency of ICWs between IP-TNT-coupled pericytes was markedly reduced in ischaemic retinas relative to sham controls (in f, sham: n = 24 capillaries, n = 4 mice; ischaemia: n = 20 capillaries, n = 4 mice; two-tailed Mann–Whitney U test, **P = 0.007). g–i, A single intraocular injection of the Ca2+ channel blocker nifedipine (30 μM), before ischaemia, was sufficient to lower intrapericyte Ca2+ (g, h; in h, vehicle: n = 90 pericytes, n = 4 mice; nifedipine: n = 49 pericytes, n = 4 mice; two-tailed Mann–Whitney U test, ***P < 0.001) and reduce the number of ruptured IP-TNTs (vehicle: n = 5 mice; nifedipine: n = 4 mice; two-tailed Student’s t-test, *P = 0.039; i). Data are presented as mean values ± s.e.m.
Pericyte processes connecting distal capillaries are tubular nanotube-like structures. 3D reconstruction of nanotube-like processes linking pericytes on distal capillaries are visualized in a DsRed mice (red) using lectin (green).
IP-TNT ultrastructural characterization by FIB-SEM. A representative IP-TNT visualized with correlative FIB-SEM showing capillaries (green), nucleus of the proximal pericyte (purple), IP-TNT (yellow), and distal pericyte processes (dpp, red). Scale bar = 1 µm.
ICWs in IP-TNT-coupled pericytes. NG2-GCaMP6 retinal explant showing the soma of two IP-TNT-coupled pericytes labeled with lectin (red). ICWs, visualized in green, propagate between connected proximal (1) and distal (2) pericytes, providing further evidence that two bona fide pericytes communicate through IP-TNTs. Scale bar: 5 µm.
Live imaging of ICWs in IP-TNT-coupled pericytes. TPLSM imaging of NG2-GCaMP6 retinas show ICWs flowing through an IP-TNT linking pericytes on distal capillaries. ICWs were measured as spontaneous synchronous Ca2+ increases in pericyte pairs. Scale bar = 5 µm.
IP-TNT and capillary dynamics ex vivo. An IP-TNT, visualized with lectin in a retinal explant, links two capillaries undergoing coordinated diameter changes. The graph shows diameter changes (%) for both IP-TNT-linked capillaries: one dilates (blue) while the other constricts (red). Scale bar = 5 µm.
Blood flow quantification. Blood flow in live retinas visualized by TPLSM and measured by quantification of red blood cells (RBC) crossing a defined location per time unit. RBC do not take up tail vein-injected FITC-dextran, hence are visualized as shadows against the fluorescent plasma. Scale bar = 5 µm.
In vivo disintegration of an IP-TNT during ischaemia. Time-lapse live imaging of a lectin-labeled IP-TNT undergoing damage during ischaemia. Arrowheads point at sites of breakage. IP-TNT collapse was not due to an out-of-focus movement artifact because each frame is the projection of an entire z-stack (depth: 30 µm). Time after ischaemia induction is shown. Scale bar = 2.5 µm.
Ischaemia increases Ca2+ influx in IP-TNT-coupled pericytes. Live imaging of a lectin-labeled IP-TNT and associated pericyte in a NG2-GCaMP6 retina. Time after ischaemia induction is shown. A robust increase in global Ca2+ signal is observed in the pericyte/IP-TNT, which persisted for at least 1 hr of ischaemia. Scale bar = 5 µm.
About this article
Cite this article
Alarcon-Martinez, L., Villafranca-Baughman, D., Quintero, H. et al. Interpericyte tunnelling nanotubes regulate neurovascular coupling. Nature 585, 91–95 (2020). https://doi.org/10.1038/s41586-020-2589-x
Cell & Bioscience (2022)
Molecular Neurodegeneration (2022)
Nature Reviews Methods Primers (2022)
Functional ultrasound localization microscopy reveals brain-wide neurovascular activity on a microscopic scale
Nature Methods (2022)
Cell Death & Disease (2022)