Fabp7, which encodes for brain lipid binding protein (BLBP), a fatty acid transporter, is one of the most highly expressed transcripts in mouse satellite glial cells (Avraham et al., 2020; Kurtz et al., 1994; Mapps et al., 2022). Using immunostaining for BLBP and tyrosine hydroxylase (TH), a marker for noradrenergic sympathetic neurons, we observed BLBP-positive satellite glia in the mouse sympathetic ganglia (superior cervical ganglia [SCG]) at both developmental and adult stages (Figure 1A). BLBP labeling was ‘patchy’ in the embryonic and neonatal ganglia, a period when satellite glia undergo migration into the ganglia and are also proliferating (Hall and Landis, 1992). However, by 2 weeks after birth, BLBP-positive satellite glial cells had expanded processes around neuronal cell bodies and formed thin ring-like glial sheaths around neuronal somas when observed at P31 (Figure 1A).

In the adult PNS, BLBP expression is restricted to satellite glia and is not detected in Schwann cells based on single-cell RNA-sequencing analysis and characterization of Fabp7-CreER2-driven reporter mice (Avraham et al., 2020). Using data sets from our recently published single-cell RNA-sequencing analysis of peripheral sympathetic and sensory ganglia from adult mice (postnatal days 30–45) (Mapps et al., 2022), we found that BLBP (Fabp7) is ~140-fold enriched in satellite glial cells compared to other ganglionic cell types (Figure 1—figure supplement 1A). Of note, Schwann cells are scarce in adult sympathetic ganglia compared to sensory ganglia based on single-cell RNA-sequencing analyses (Mapps et al., 2022). To further ensure the cellular specificity of BLBP, we generated genetic reporter mice by crossing Fabp7-CreER2 mice with ROSA26^mEGFP mice (Muzumdar et al., 2007), which drives expression of membrane-tagged EGFP (Figure 1—figure supplement 1B). We found that tamoxifen treatment (180 mg/kg body weight for five consecutive days) in Fabp7-CreER2;ROSA26^mEGFP mice resulted in m-EGFP reporter expression in satellite glia marked by Kir4.1 immunolabeling (Vit et al., 2006), but not in TH-positive sympathetic neurons, IBA1-labeled macrophages, or Pdgfrβ-labeled vascular mural cells in sympathetic ganglia (Figure 1—figure supplement 1C–F). Together, these results are consistent with previous studies (Avraham et al., 2022; Avraham et al., 2020; Avraham et al., 2021; Mapps et al., 2022) showing that BLBP is a specific marker for satellite glia in adult mice.

Next, to accomplish ablation of satellite glia, we crossed Fabp7-CreER2 mice (Maruoka et al., 2011) with ROSA26^eGFP-DTA mice (Ivanova et al., 2005), where Cre drives expression of a copy of diphtheria toxin subunit A (DTA). At postnatal day 30, Fabp7-CreER2;ROSA26^eGFP-DTA mice were injected with either vehicle (corn oil) or tamoxifen (180 mg/kg body weight) for five consecutive days, and all analyses were performed at 5 or 14 days post injection. By 2 weeks, we observed a drastic loss of BLBP expression in sympathetic (superior cervical) ganglia from tamoxifen-treated mice (henceforth referred to as BLBP:iDTA mice) compared to vehicle-injected control (Ctrl) mice (Figure 1B). While satellite glia in control ganglia formed characteristic ring-like structures around neuronal soma, glial organization was disrupted in mutant mice with diminished BLBP staining detected within the ganglia. To determine whether DTA expression results in cell death, we assessed apoptosis using TUNEL labeling and observed a threefold increase in TUNEL-positive cells in BLBP:iDTA sympathetic ganglia compared to controls at 5 days post-tamoxifen injection (Figure 1C and D). To identify cell types undergoing apoptosis, we performed TUNEL labeling together with immunostaining for Sox10, a transcription factor expressed in satellite glia (Britsch et al., 2001), and TrkA, a sympathetic neuron marker (Fagan et al., 1996). We found a twofold increase in the number of TUNEL⁺;Sox10⁺ cells in BLBP:iDTA ganglia compared to controls at 5 days post-tamoxifen injection (5 dpi) (Figure 1E–G). Although there was a trend toward enhanced neuronal apoptosis, the number of TUNEL⁺;TrkA⁺ sympathetic neurons in BLBP:iDTA ganglia was not statistically different from that in controls at this stage (Figure 1F–H). Despite increased glial apoptosis at 5 dpi, there was no significant loss of satellite glial cells at this time as assessed by quantification of Sox2-immunoreactive cells (Figure 1—figure supplement 2A and B), where Sox2 is another transcription factor expressed in satellite glia (Koike et al., 2014). However, at 5 dpi, we did observe fewer Sox2-positive glial cells juxtaposed to cell bodies of individual sympathetic neurons in mutant ganglia (Figure 1—figure supplement 2 A and C), suggesting that unhealthy/dying satellite glia lose their contacts with neuronal soma. By 14 days post-tamoxifen injection (14 dpi), however, there was a pronounced loss of Sox2-labeled satellite glial cells (54.2% decrease) in sympathetic ganglia from BLBP:iDTA mice (Figure 1I and J). To confirm ablation of Sox2-positive cells, rather than downregulated Sox2 expression in mutant ganglia, we generated binary images of Sox2-labeled cells by filtering and thresholding using ImageJ. This method allowed us to simply record the presence or absence of cells in the images in a manner independent of pixel values. Quantification of binary images revealed a substantial decrease in the number of Sox2-labeled cells (33% decrease) in BLBP:iDTA ganglia compared to controls at 14 dpi (Figure 1—figure supplement 2D and F). Of note, this 33% decrease is lower than the 54% loss quantified using Sox2 immunofluorescence (Figure 1J). Assessing binary images from the 5 dpi time point indicated that there was no significant loss of Sox2-labeled cells at this stage (Figure 1—figure supplement 2E and G), consistent with quantifications based on Sox2-immunofluorescence (see Figure 1—figure supplement 2B). As an additional measure to visualize sympathetic neurons and associated satellite glial cells in a manner that does not rely on fluorescence, we labeled sympathetic neurons using TH DAB (3,3'-diaminobenzidine) immunohistochemistry and visualized satellite glia nuclei abutting TH-positive neuronal cell bodies based on their distinctive appearance and location using hematoxylin staining. We observed a significant loss of glial nuclei associated with soma of individual sympathetic neurons in BLBP:iDTA mice compared to control mice at 14 dpi (Figure 1K and L). Together, these results indicate that satellite glia are undergoing apoptosis by 5 dpi, and by 14 dpi, there is a substantial loss of glial cells in BLBP:iDTA ganglia. Further, consistent with cell loss at 14 dpi, we found a pronounced downregulation of several satellite glia-specific transcripts, including Fabp7, Fasn, Apoe, and Kcnj10 (Mapps et al., 2022), in BLBP:iDTA sympathetic ganglia at this time as revealed by quantitative PCR (qPCR) analyses (Figure 1—figure supplement 2H).

In BLBP:iDTA ganglia, we did not find increased macrophage infiltration as assessed by quantification of IBA-1-labeled cells (Figure 1—figure supplement 3A and B). It remains to be determined whether satellite glia loss triggers increased macrophage reactivity or phagocytosis. Further, EdU labeling indicated that satellite glial cells do not undergo increased proliferation as a compensatory or injury-induced response after DTA expression (Figure 1—figure supplement 3C and D). Together, these results suggest that DTA-induced ablation of satellite glia is not accompanied by increased macrophage density or glial proliferation in mutant sympathetic ganglia.

When performing TH immunostaining to visualize sympathetic neuron morphology, we noticed a striking downregulation of TH immunoreactivity in sympathetic neuronal cell bodies from BLBP:iDTA mice. TH is the rate-limiting enzyme in the biosynthesis of NE, the classical sympathetic neurotransmitter. qPCR analyses revealed a drastic downregulation in Th and Dopamine Beta-Hydroxylase (DBH) transcript levels, (88 and 99% decrease, respectively) (Figure 2A and B). DBH converts dopamine to NE in the NE biosynthetic pathway. Thus, satellite glia maintain noradrenergic enzymes in sympathetic neurons.

Using TH immunohistochemistry, we also observed pronounced atrophy of neuronal cell bodies in sympathetic ganglia after satellite glia loss (Figure 2A). Quantification of soma sizes revealed significantly reduced neuronal soma areas in satellite glia-depleted mice, with a greater distribution of smaller-sized cell bodies in mutant ganglia relative to controls (Figure 2C). Median values for soma areas are 162 ± 17 μm² for control neurons vs 118 ± 7.1 μm² for mutant neurons (Figure 2D). Since the PI3K/Akt/mTOR pathway is a known regulator of neuronal soma size (Kwon et al., 2006; van Diepen et al., 2009; Zhou et al., 2009), we assessed expression of key downstream effectors of the mTOR pathway, phosphorylated ribosomal protein 6 (p-S6) and phosphorylated eukaryotic translation initiation factor 4E (eIF4E)-binding protein 1 (p-4-EBP1) (Meyuhas, 2008; Saxton and Sabatini, 2017), using immunostaining. We observed a dramatic reduction in p-S6 and p-4-EBP1 immunoreactivity in satellite glia-depleted ganglia (Figure 2E, Figure 2—figure supplement 1A). These results suggest that satellite glia promote soma growth, in part, via regulating mTOR signaling in sympathetic neurons.

Given the soma atrophy, we asked whether satellite glia ablation affects neuron survival. At 5 dpi, sympathetic neuron numbers in BLBP:iDTA mice were not significantly different from that in control littermates (12,240 ± 1514 in mutant vs. 14,200 ± 1005 in control mice) as revealed by Nissl staining and quantification of cell counts in tissue sections from SCG (Figure 2—figure supplement 1B). However, by 14 dpi, we found a marked loss of sympathetic neurons (24% decrease) in satellite glia-ablated ganglia (11,790 ± 706 mutant neurons vs. 15,460 ± 734 control neurons) (Figure 2F and G). Despite the decrease in neuronal numbers, sympathetic axon innervation was maintained in target organs in BLBP:iDTA mice, when assessed by wholemount TH immunostaining in iDISCO-cleared tissues and light sheet microscopy (Figure 2—figure supplement 1C–E). Intriguingly, TH levels in mutant axons appeared to be similar to that in controls (Figure 2—figure supplement 1C–E), despite the marked reduction in Th mRNA and protein in neuronal cell bodies residing in the ganglia (see Figure 2A and B), suggesting differential regulation of TH distribution in neuronal soma vs. axons.

Together, these results indicate that satellite glia provide trophic and metabolic support to adult sympathetic neurons and regulate noradrenergic biosynthetic machinery, specifically in neuronal cell bodies.

Given neuronal deficits with satellite glia depletion, we sought to determine whether autonomic responses were impacted in BLBP:iDTA mice. In mammals, pupil size can serve as a noninvasive and rapid readout for autonomic function (McDougal and Gamlin, 2015). Pupil size is modulated by a balance of sympathetic vs. parasympathetic activity, with the sympathetic component regulating pupil dilation while parasympathetic activity controls pupil constriction (McDougal and Gamlin, 2015). To measure basal pupil size, control and BLBP:iDTA mice were dark-adapted for 2 days, and pupil sizes recorded for 5–10 s in the dark in non-anesthetized mice (Keenan et al., 2016). Surprisingly, despite the neuronal loss and downregulated noradrenergic biosynthetic enzymes in neuronal soma, we observed increased basal pupil areas in BLBP:iDTA mice compared to controls (Figure 3A and B). To ask whether this phenotype is due to decreased parasympathetic activity, we measured pupil constriction in response to increasing light intensities, ranging from 0.01 to 1000 lux, administered for 30 s. Light onset at 0.1 lux or higher resulted in rapid constriction with greater constrictions at higher light intensities in both BLBP:iDTA and control mice (Figure 3—figure supplement 1A). The intensity responses were virtually identical for the two groups (Figure 3—figure supplement 1A). These results suggest that parasympathetic function is intact in BLBP:iDTA mice and that the enlarged pupil areas likely reflect an increase in sympathetic tone with the loss of satellite glia.

As a second and independent assessment of autonomic function, we measured heart rate and heart rate variability (HRV) using electrocardiogram (ECG) recordings in mice (Thireau et al., 2008). Increased sympathetic activity results in an accelerated heart rate and decreased HRV, defined as the variation in time intervals between consecutive heartbeats (Thireau et al., 2008). Strikingly, BLBP:iDTA mice exhibited increased heart rates (Figure 3C and D) and decreased HRV (Figure 3—figure supplement 1B) compared to controls. Tamoxifen injections had no effect on heart rate in wild-type C57BL/6J mice or control ROSA26^eGFP-DTA mice that do not express Cre (Figure 3—figure supplement 1C and D).

To directly test the involvement of sympathetic neurons in the heart defect observed in BLBP:iDTA mice, we chemically ablated sympathetic nerves using 6-hydroxydopamine (6-OHDA), which selectively destroys sympathetic, but not parasympathetic or sensory nerves (Borden et al., 2013; Kostrzewa and Jacobowitz, 1974). Further, 6-OHDA does not cross the blood–brain barrier after intraperitoneal injections in adult animals, restricting its actions to the periphery (Kostrzewa and Jacobowitz, 1974). Ablation of sympathetic nerves prevented the augmented heart rate in BLBP:iDTA mice (Figure 3E and F), suggesting the dysfunction of peripheral sympathetic neurons as the primary contributor to the cardiac defect in satellite glia-ablated mice.

To understand the molecular and/or cellular basis for enhanced sympathetic activity in satellite glia-depleted mice, we measured circulating NE levels and observed a significant 1.6-fold increase in BLBP:iDTA mice compared to controls (Figure 3G). These results suggest that, despite the downregulation of NE biosynthetic machinery in neuronal cell bodies (see Figure 2A and B), neurotransmitter secretion is augmented in sympathetic axons and/or its reuptake/degradation decreased, in satellite glia-depleted mice. NE acts through α- and β-adrenergic receptors in target tissues. Heightened sympathetic activity and increased circulating NE elicits the compensatory downregulation of adrenergic receptor levels and activities in target tissues (Eschenhagen, 2008). We found significantly decreased expression of Adrb1, the major adrenergic receptor for NE signaling in the heart (de Lucia et al., 2018), as well as Adrb2 and Adra2c, using qPCR analyses of cardiac tissue (Figure 3—figure supplement 1E). Lastly, immunostaining for c-Fos, an immediate early transcription factor that serves as a reporter of neuronal activity (Sheng and Greenberg, 1990), revealed a striking eightfold increase in the number of c-Fos-positive sympathetic neurons in BLBP:iDTA ganglia compared to controls (Figure 3H and I).

Together, these results indicate that satellite glia depletion results in elevated sympathetic neuron activity, impaired NE homeostasis, and autonomic behavioral defects in mice.

Kir4.1 is a glial-specific, inwardly rectifying K⁺ channel that shows the highest expression in astrocytes and satellite glia (Hanani and Spray, 2020; Olsen et al., 2015; Vit et al., 2008). Kcnj10 mRNA for Kir4.1 is highly enriched in satellite glia (approximately threefold enriched) compared to other ganglionic cell types based on our analysis of data sets obtained from single-cell RNA-sequencing of sympathetic and sensory ganglia from adult mice (postnatal days 30–45) (Mapps et al., 2022; Figure 4—figure supplement 1A). Co-immunolabeling for Kir4.1 and m-EGFP in Fabp7-CreER2;ROSA26^mEGFP reporter mice revealed co-localization of the two signals in sympathetic ganglia (Figure 1—figure supplement 1C). Further, in BLBP:iDTA sympathetic ganglia, we observed a pronounced decrease in Kcnj10 mRNA compared to controls as assessed by qPCR analysis (Figure 1—figure supplement 2H) and single-molecule fluorescence in situ hybridization (smFISH) (Figure 4—figure supplement 1B). Our results are consistent with previous studies that Kcnj10 is enriched in satellite glial cells (Avraham et al., 2020; Hanani and Spray, 2020; Vit et al., 2008).

In mice, global or tissue-specific Kir4.1 deletion results in impaired K⁺ and glutamate homeostasis, loss of glial K⁺ conductance, neuronal excitability, epileptic seizures, and pain-like behaviors (Cui et al., 2018; Djukic et al., 2007; Olsen et al., 2015; Tang et al., 2010; Vit et al., 2008). To ask whether impaired Kir4.1 function might contribute to the neuronal defects observed in satellite glia-ablated mice, we generated satellite glia-specific Kir4.1 knockout mice (Kir4.1 cKO) by crossing Fabp7-CreER2 mice to Kcnj10 floxed mice (Djukic et al., 2007). Fabp7-CreER2; Kcnj10^fl/fl (henceforth called Kir4.1 cKO) mice were treated with vehicle or tamoxifen for five consecutive days to conditionally delete Kir4.1 from satellite glial cells. Expression of Kir4.1, visualized by immunofluorescence, was significantly reduced in Kir4.1 cKO sympathetic ganglia (Figure 4A), and qPCR analysis indicated a 76% decrease in Kcnj10 transcript levels (Figure 4B).

Loss of Kir4.1 did not alter satellite glial numbers in sympathetic ganglia as assessed by quantification of Sox2-positive cells (Figure 4C and D, Figure 4—figure supplement 1C and D). Kir4.1 deletion also did not affect Fabp7 (BLBP) expression in satellite glia assessed by qPCR analysis and immunostaining (Figure 4—figure supplement 1E and F). BLBP immunostaining also indicated that the characteristic ring-like glial organization around neuronal soma was not perturbed in Kir4.1 cKO sympathetic ganglia (Figure 4—figure supplement 1F). However, we observed a marked increase in the number of c-Fos-positive cells in Kir4.1 cKO SCGs compared to controls (Figure 4E and F). The majority of c-Fos-positive cells in Kir4.1 cKO ganglia appeared to be neurons based on their morphology. Intriguingly, despite increased c-Fos-positive sympathetic neurons, Kir4.1 deletion resulted in downregulated expression of noradrenergic biosynthetic enzymes, TH and DBH, in neuronal cell bodies (Figure 4C, G, and H), similar to the phenotype observed with satellite glia depletion.

Together, these results suggest that satellite glial cells control sympathetic neuron activity and noradrenergic enzyme expression via Kir4.1-dependent regulation of the neuronal microenvironment.

We next addressed the relevance of satellite glia Kir4.1 expression in sympathetic neuron viability. Quantification of soma sizes revealed that neurons undergo atrophy in Kir4.1 cKO sympathetic ganglia (Figure 5A), similar to the defect observed with DTA-induced loss of satellite glia. Soma areas from tissue sections, represented as median values, were 155 ± 13 μm² for control neurons vs. 118 ± 3.6 μm² for mutant neurons (Figure 5B). Kir4.1 loss in satellite glial cells also elicited a pronounced decrease in the levels of p-S6 and p-4-EBP1, two well-established downstream effectors of mTOR activity (Meyuhas, 2008; Saxton and Sabatini, 2017), in sympathetic neurons (Figure 5C, Figure 5—figure supplement 1A), similar to the effects of satellite glia ablation (Figure 2E, Figure 2—figure supplement 1A). Notably, we observed an eightfold increase in apoptotic cells in Kir4.1 cKO sympathetic ganglia using TUNEL labeling at 14 days post-tamoxifen injection (Figure 5D and E). The apoptotic cells were primarily sympathetic neurons based on co-labeling for TUNEL and TrkA (Figure 5—figure supplement 1B and C). We did not observe increased apoptosis of satellite glial cells in Kir4.1 cKO sympathetic ganglia compared to controls (Figure 5—figure supplement 1D and E), consistent with results that the number of Sox2-positive satellite glia was not altered by Kir4.1 deletion (see Figure 4C and D). Consistent with enhanced neuronal apoptosis, we observed a 22% decrease in sympathetic neuron numbers in Kir4.1 cKO sympathetic ganglia by 2 weeks after tamoxifen injections (Figure 5F and G). Similar to the findings in BLBP:iDTA mice, sympathetic axon innervation and axonal TH expression were similar between Kir4.1 cKO mice and controls (Figure 5—figure supplement 1F and G), despite the decrease in sympathetic neuron numbers and downregulated TH expression in neuronal cell bodies. Together, these results highlight that Kir4.1 is necessary for the survival of adult sympathetic neurons.

We next asked whether satellite glia-specific Kir4.1 deletion recapitulates the autonomic defects observed in BLBP:iDTA mice. Measurements of basal pupil size, heart rate, and circulating NE levels indicated that these parameters were not significantly altered in Kir4.1 cKO mice (Figure 5H and I, Figure 5—figure supplement 1H–J), despite increased neuronal activity as assessed by c-Fos immunoreactivity (see Figure 4E and F). Parasympathetic activity assessed by pupil constriction in response to different light intensities was also normal in Kir4.1 cKO mice (Figure 5—figure supplement 1K).

Together, these results suggest that Kir4.1 deletion recapitulates the cellular phenotypes observed in BLBP:iDTA mice, notably, increased neuron activity, impaired metabolic signaling, and enhanced apoptosis. However, the deletion of a single gene, Kcnj10 encoding for Kir4.1, from satellite glial cells is not sufficient to drive behavioral changes at the whole animal level as seen with genetic ablation of satellite glial cells. These results suggest that there are other satellite glia-dependent mechanisms, in addition to Kir4.1 activity, that contribute to regulation of neuronal excitability to drive circuit-level changes.

Information for statistical analyses for all experiments is provided in ‘Materials and methods’ and figure legends.

All procedures relating to animal care and treatment conformed to The Johns Hopkins University Animal Care and Use Committee (ACUC) and NIH guidelines. Animals were group housed in a standard 12:12 light–dark cycle, except for the pupil analyses. Adult mice between 1 and 1.5 months of age, and of both sexes, were used for analyses. The following mouse lines were used in this study: C57Bl/6J (JAX:000664) mice, Fabp7-CreER2 mice (Maruoka et al., 2011) were generously provided by Dr. Toshihiko Hosoya (RIKEN Brain Science Institute), Kcnj10^fl/fl mice (Djukic et al., 2007) (The Jackson Laboratory, stock no: 026826) by Dr. Dwight Bergles (Johns Hopkins School of Medicine), and ROSA26^mEGFP mice (Muzumdar et al., 2007) (The Jackson Laboratory, stock no: 007576) by Dr. David Linden (Johns Hopkins School of Medicine). ROSA26^eGFP-DTA mice were obtained from The Jackson Laboratory (stock no: 006331).

At postnatal day 30, C57Bl/6J, Fabp7-CreER2;ROSA26^mEGFP, Fabp7-CreER2;ROSA26^eGFP-DTA, or Fabp7-CreER2;Kcnj10^fl/fl mice were injected subcutaneously with either vehicle corn oil (Sigma) or tamoxifen (180 mg/kg body weight) dissolved in corn oil for five consecutive days. All analyses were performed at 5 or 14 days after the last injection.

RNA was isolated from dissected SCGs or cardiac tissue using Absolutely RNA Nanoprep Kit (Agilent) or Trizol-chloroform extraction. cDNA was prepared using Superscript IV First Strand Synthesis System. Real-time qPCR analysis was performed using Maxima SYBER Green/Rox Q-PCR Master Mix (Thermo Fisher) and gene-specific primers for SCG tissues or TaqMan probes for adrenergic receptors in heart tissue, in a 7300 Real time PCR System (Applied Biosystems). Each sample was analyzed in triplicate reactions. Fold change was calculated using the 2(ΔΔCt) method, normalizing to 18S transcript.

SCGs were dissected from P48 mice, cryoprotected in 30% sucrose in PBS for 1 hr and embedded in OCT and frozen at –80°C. Ganglia were cryosectioned at 14 µm and kept at –80°C until smFISH was performed. Target mRNA was probed using RNAscope Multiplex Fluorescent Reagent Kit v2 Assay. Tissues were incubated in fresh 4% paraformaldehyde for 5 min, washed twice in 1× PBS, and dehydrated with increasing concentrations of ethanol. Subsequently, tissues were treated with hydrogen peroxide for 10 min and protease treatment for 15 min. RNAscope assays were performed following the manufacturer’s instructions. Tissues were mounted with Prolong Anti-fade Mounting Media and imaged using a Zeiss LSM 800 confocal microscope.

SCGs were harvested and incubated in 4% paraformaldehyde for 4 hr at room temperature, cryoprotected in 30% sucrose in 1× PBS for 3 days, embedded in OCT (Sakura Finetek) and stored at –80°C. Ganglia were cryosectioned at 12–30 μm sections for immunohistochemistry. For paraffin embedding, SCGs were fixed in Bouin’s solution for 1 hr at room temperature, left in 70% ethanol overnight at room temperature, followed by consecutive washes in 70, 80, 95, and 100% ethanol, and xylene. Tissues were embedded in paraffin wax and sectioned at 6 µm using a microtome, and rehydrated using a series of xylene and ethanol washes at room temperature. Tissue sections were permeabilized in 0.1% Triton X-100 in 1× PBS at room temperature 3× for 5 min each, followed by incubation in 10 mM sodium citrate buffer (pH 6) at 95°C for 10 min. Tissues were blocked in 5% goat serum/3% bovine serum albumin in 0.1% or 0.3% Triton X-100 in 1× PBS (blocking solution) for 1 hr at room temperature. Primary antibodies used were mouse anti-BLBP (1:500), rabbit anti-BLBP (1:200), mouse anti-TH (1:300), rabbit anti-TH (1:300), rabbit anti-TrkA (1:200), rabbit anti-IBA1 (1:200), rabbit anti-pS6 (1:200), rabbit anti-p4E-BP1 (1:200), rabbit anti-c-Fos (1:1000), goat anti-Sox10 (1:50), or rabbit anti-Kir4.1 (1:100) with incubations performed overnight at 4°C. Slides were washed in blocking solution and then incubated in Alexa-488 ,-546, or-647 conjugated anti-rabbit or anti-mouse secondary antibodies (1:200 dilution) and DAPI (1:1000) in blocking buffer. Tissues were then mounted in Aqueous Mounting Medium and imaged at 1 µm optical slices using a Zeiss LSM 700 or LSM 800 confocal microscope. Maximum intensity projections and maximum intensity 3D projections were generated using ImageJ.

For Sox-2 immunostaining, SCGs were embedded in 3% agarose (Sigma) and sectioned at 100-µm-thick sections using a vibratome. Tissues were permeabilized/blocked in 10% goat serum/3% tween in 1× PBS for 2 hr at room temperature, incubated with rabbit anti-Sox2 (1:500) and mouse anti-TH (1:300) antibodies in blocking solution for 2 days at 4°C, followed by Alexa-546 or -647 conjugated anti-rabbit or anti-mouse secondary antibodies (1:400 dilution) and DAPI (1:1000). Maximum intensity 3D projections were generated using ImageJ.

Binary images of Sox2⁺ cells were generated using Fourier Bandpass Filter plug-in on ImageJ to first reduce edge artifacts, and then adjusting the threshold to the maximum fluorescence intensity of 255 nm. Particles of sizes between 3 and 50 μm were quantified using the ‘Analyze Particle’ plug-in on ImageJ.

For TH DAB immunohistochemistry, paraffin sections (6 μm) of SCGs were prepared as described above. After tissues were rehydrated, peroxidase quenching was done by incubation in a tris buffer solution (TBS) containing 3% hydrogen peroxide/5% methanol for 15 min at room temperature. Tissues were then permeabilized/blocked in 10% donkey serum/1% glycine/2% bovine serum albumin/0.4% Triton X-100 in 1× TBS for 2 hr at room temperature, incubated with rabbit anti-TH (1:200) in blocking solution overnight at 4°C. Slides were washed three times with TBS, incubated with rabbit anti-HRP in blocking solution for 1 hr at room temperature, washed with TBS before being incubated with SIGMAFAST 3,3'-diaminobenzidine for 10 min. Tissues were washed once with TBS and counterstained with hematoxylin for 1 min prior to dehydrating in increasing concentrations of ethanol and xylene. Tissues were mounted with Permount Mounting Media. Glial nuclei associated with TH-positive neurons in SCGs were counted using ImageJ.

iDISCO-based tissue clearing for wholemount immunostaining of organs from P48 mice was performed as previously described (Renier et al., 2014). Briefly, hearts were fixed in 4% PFA/PBS, then dehydrated by methanol series (20–80%) and incubated overnight in 66% dichloromethane (DCM)/33% methanol. Samples were then bleached with 5% H₂O₂ in methanol at 4°C overnight, then re-hydrated and permeabilized first with 0.2% Triton X-100 followed by overnight permeabilization with 0.16% Triton X-100/20% DMSO/0.3 M glycine in PBS. Samples were incubated in blocking solution (0.17% Triton X-100/10% DMSO/6% Normal Goat Serum in PBS) for 8 hr, and then incubated with rabbit anti-TH (1:400) in 0.2% Tween-20/0.001% heparin/5% DMSO/3% Normal Goat Serum in PBS at 37°C for 96 hr. Samples were then washed with 0.2% Tween-20/0.001% heparin in PBS and incubated with anti-rabbit Alexa-546 secondary antibody (1:400) in 0.2% Tween-20/0.001% heparin/3% Normal Goat Serum in PBS. After 96 hr, organs were extensively washed with 0.2% Tween-20/0.001% heparin in PBS and dehydrated in methanol. Samples were cleared by successive washes in 66% DCM/33% methanol, 100% DCM, and 100% dibenzyl ether. Organs were imaged on a lightsheet microscope (LaVision BioTec Ultra Microscope II). Imaris was used for 3D manipulations. Total axon lengths and number of branch points were quantified using Imaris Filament Tracer and normalized to total organ volume.

SCG tissue sections were labeled with hematoxylin and eosin. Cells with characteristic neuronal morphology and visible nucleoli were identified, soma were traced and areas (µm²) quantified using ImageJ (Fiji).

Neuronal counts were performed as previously described (Scott-Solomon and Kuruvilla, 2020). In brief, torsos of P39-48 mice were fixed in 4% PFA/PBS overnight and cryoprotected in 30% sucrose/PBS for 48 hr. Torsos were then mounted in OCT and serially sectioned (12 μm). Next, every fifth section was stained with solution containing 0.5% cresyl violet (Nissl). Cells in both SCGs with characteristic neuronal morphology and visible nucleoli were counted using ImageJ.

Apoptotic cells were identified in every fifth section of ganglia. Tissues were first incubated with primary and secondary antibodies as described above, followed by detection of cell death using TUNEL staining (Roche) according to the manufacturer’s protocol. Cells that were double positive for DAPI and TUNEL were counted as dying cells. To quantify neuronal or glial apoptosis, TUNEL labeling was done together with TrkA or Sox10 immunostaining as described above. TUNEL⁺; TrkA⁺ and TUNEL⁺; Sox10⁺ cells were expressed as percentages of the total number of TrkA⁺ and Sox 10⁺ cells, respectively.

Tamoxifen- or corn oil-injected adult mice were injected intraperitoneally with EdU (Invitrogen, 100 µg/ml in 3:1 PBS/DMSO) for five consecutive days before harvesting at 5 days post-tamoxifen injections. SCG tissue sections (12 μm) were processed for 30 min at room temperature in EdU reaction cocktail (Thermo Fisher EdU kit C10337; Click-iT buffer, Buffer additive, CuSO₄ solution, and Alexa Fluor 488). Sections were then washed in PBS +0.1% Triton X-100 and mounted with Fluoromount + DAPI. Images were collected using a Zeiss LSM 700 confocal microscope. The total number of cells that incorporated EdU in each section was counted and summed for an entire SCG.

Blood samples drawn from anesthetized mice were centrifuged for 15 min, 3000 rpm in 0.5 M EDTA-coated tubes at 4°C. NE levels in plasma (300 μl) was assessed using a ELISA kit (Abnova) according to the manufacturer’s protocol.

Pupil size measurements were performed as reported previously (Keenan et al., 2016). Briefly, all mice were dark-adapted and housed in single cages for 2 days and analyzed in the evenings. For all experiments, mice were unanesthetized and restrained by hand. To mitigate stress, which can affect pupil size, researchers handled mice for several days prior to the measurements. Videos of the eye were recorded for 5–10 s in the dark using a Sony Handycam (DCR-HC96) mounted on a tripod at a fixed distance from the mouse. Manual focus was maintained on the camera to ensure that only one focal plane existed for each mouse. Pupil size was recorded under dim red light and the endogenous infrared light source of the camera to capture the basal pupil size.

To examine parasympathetic activity, mice were dark-adapted for 2 days. Unanesthetized mice were restrained by hand. Pupil size was recorded first for 5–10 s in the dark followed by a 30 s exposure to a light step stimulus. The light stimulus was provided by 10 W or 14 W LED bulbs (Sunlite A19/or Sunlite 80599-SU LED A19 Super Bright Light Bulb, Daylight). Light intensity was measured using a light meter (EXTECH Foot Candle/Lux Light Meter, 401025) at the surface on which the mouse was held. Light intensity was adjusted by a combination of altering the distance of the light bulb from the mouse and/or applying neutral density filters (Roscolux). The light meter is unreliable at detecting light intensities below 1 lux, so one neutral density filter cutting the light intensity by 12.5% was applied to the bulb to estimate 1-log unit decreases in illumination below 1 lux. Light intensities above 500 lux required the use of multiple light bulbs.

ECG recordings were performed on adult mice as previously described (Ståhlberg et al., 2019). Briefly, adult mice (P39-48) were anesthetized with 4% isoflurane, intubated, and placed on ventilator support (settings 1.2 ml/g/min at 80 breaths/min). The animal’s dorsum was shaved, scrubbed with betadine and alcohol, and draped with a sterile barrier with the surgery site exposed. A small 0.5 cm midline incision was performed and ECG leads were implanted subcutaneously and sutured over the trapezius muscle on both sides. Body temperature was maintained at 37°C. Immediately following implantation, the wound was closed with a 3-0 silk suture. Anesthesia was turned off and the animal was monitored for spontaneous breathing and was given a subcutaneous buprenorphine injection (0.01–0.05 mg/kg buprenorphine, IM) to alleviate pain. ECGs were subsequently recorded continuously in conscious animals for approximately 7 days for each mouse using the Powerlab data acquisition device and LabChart 8 software (ADInstruments). Mice were kept at a stable temperature with regular 12 hr light/dark cycle. To exclude the effects of pain and anesthesia, continuous ECG recordings between day 4 and 7 post-lead implantation were only included in the analysis of mean heart rates. HRV was analyzed using LabChart 8 and at 10–12 hr intervals as previously described (Thireau et al., 2008).

To assess effects of tamoxifen alone, adult (P39-48) C57Bl/6J or ROSA26^eGFP-DTA mice that did not express Cre mice were injected with corn oil or tamoxifen (subcutaneous injections, 180 mg/kg body weight for five consecutive days) and ECGs recorded as described above.

For chemical ablation of sympathetic nerves, Fabp7-CreER2;ROSA26^eGFP-DTA mice were first injected with corn oil or tamoxifen. After 18–20 days, control and BLBP:iDTA mice received two intraperitoneal injections, 3 days apart, of 6-OHDA (150 mg/kg) dissolved in 0.1% ascorbic acid. ECGs were subsequently recorded continuously in conscious animals for approximately 4 days for each mouse as described above.

Sample sizes in this study were calculated using power analyses using RStudio statistical software. For practical reasons, analyses of innervation, cell death assays, proliferation assays, soma size measurements, and NE ELISA’s were performed in a semi-blinded manner. The experimenter was aware of the genotypes but performed each immunostaining and measurements without knowing the genotypes. All physiological experiments (pupil dilation and HRV) were performed in a blinded manner; the experimenter was only aware of the ear tag numbers. All t-tests were performed assuming Gaussian distribution, two-tailed, unpaired, and a confidence interval of 95%. One-way or two-way ANOVA with Bonferroni’s correction were performed when more than two groups were compared. All error bars represent the standard error of the mean (SEM).