Short and long sympathetic-sensory feedback loops in white fat

Abstract

We previously demonstrated white adipose tissue (WAT) innervation using the established WAT retrograde sympathetic nervous system (SNS)-specific transneuronal viral tract tracer pseudorabies virus (PRV152) and showed its role in the control of lipolysis. Conversely, we demonstrated WAT sensory innervation using the established anterograde sensory system (SS)-specific transneuronal viral tracer, the H129 strain of herpes simplex virus-1, with sensory nerves showing responsiveness with increases in WAT SNS drive. Several brain areas were part of the SNS outflow to and SS inflow from WAT between these studies suggesting SNS-SS feedback loops. Therefore, we injected both PRV152 and H129 into inguinal WAT (IWAT) of Siberian hamsters. Animals were perfused on days 5 and 6 postinoculation after H129 and PRV152 injections, respectively, and brains, spinal cords, sympathetic, and dorsal root ganglia (DRG) were processed for immunohistochemical detection of each virus across the neuroaxis. The presence of H129+PRV152-colocalized neurons (∼50%) in the spinal segments innervating IWAT suggested short SNS-SS loops with significant coinfections (>60%) in discrete brain regions, signifying long SNS-SS loops. Notably, the most highly populated sites with the double-infected neurons were the medial part of medial preoptic nucleus, medial preoptic area, hypothalamic paraventricular nucleus, lateral hypothalamus, periaqueductal gray, oral part of the pontine reticular nucleus, and the nucleus of the solitary tract. Collectively, these results strongly indicate the neuroanatomical reality of the central SNS-SS feedback loops with short loops in the spinal cord and long loops in the brain, both likely involved in the control of lipolysis or other WAT pad-specific functions.

we previously demonstrated postganglionic sympathetic nervous system (SNS) innervation of inguinal and epididymal white adipose tissue [IWAT and EWAT, respectively; (69)] and together with others determined that the pattern of SNS drive is fat pad-specific (15, 29, 43). Collectively, these and other data such as the blockade of lipolysis with WAT SNS denervation (for review see Refs. 8 and 9) establish that activation of the SNS innervation of WAT is the principal mechanism underlying lipid mobilization. To unravel the central nervous system (CNS) origins controlling the sympathetic drive to WAT, we used the retrograde transneuronal tract tracer pseudorabies virus (PRV) Bartha's K strain to define the CNS origins of the sympathetic outflow to IWAT and EWAT in Siberian hamsters as well as in Sprague-Dawley rats (4). We and others expanded these studies not only for subcutaneous WAT and the retroperitoneal WAT (RWAT) depot (1, 13, 52, 57) but also for true visceral WAT (mesenteric WAT; see Ref. 46). It is noteworthy that the SNS components responsible for sympathetic outflow to the aforementioned fat pads share many central sites in the circuitries innervating these adipose tissues, but also have separate innervation circuitries. For example, microinjections of PRV152 into different WAT depots results in a general pattern of viral labeling including, for example, the nucleus of the solitary tract (NTS) in the hindbrain, the periaqueductal gray (PAG) in the midbrain, the hypothalamic paraventricular nucleus (PVH), dorsomedial hypothalamus (DMH), and medial preoptic area (MPA) in the forebrain, suggesting that divergence in the sympathetic drives to WAT that underlies the fat pad-specific patterns of SNS drive originates at the caudal orders of the sympathetic neuroaxis (1, 4, 13, 46, 52, 57, 61, 65). Conversely, we demonstrated WAT sensory innervation and inflow to the CNS using the established anterograde sensory system (SS)-specific transneuronal viral tract tracer, the H129 strain of herpes simplex virus-1 (60). In addition, the sensory nerves innervating WAT are responsive electrophysiologically to increases in the SNS drive to WAT (60). Additional supportive neuroanatomical evidence for WAT SS innervation is the presence of classical sensory nerve-associated peptides found in neural fibers innervating WAT such as calcitonin gene-related peptide (38) in laboratory rat (30) and in Siberian hamster WAT (26, 53, 54) as well as substance P (28, 30). In addition, others (25) and we (44) demonstrated the direct innervation of WAT by pseudounipolar dorsal root ganglia (DRG) neurons using conventional retrograde tract tracers administered to Sprague-Dawley rat and Siberian hamster WAT, respectively. Collectively, it is clear that WAT possesses sensory innervation (for review see Refs. 8 and 9).

We noticed several brain areas were part of both innervations suggesting possible SNS-SS feedback loops, between our viral tract tracing studies identifying the SNS outflow from brain to WAT and the sensory inflow from WAT to brain discussed above.

Therefore, the purpose of the present study was to test whether individual neurons that are part of the SNS outflow to IWAT also receive sensory inflow from the same tissue centrally and peripherally. Because WAT does not possess parasympathetic nervous system (PSNS) innervation (3, 11, 31, 32), this could be accomplished by injecting both the retrograde transneuronal tract tracer PRV152 to label only the SNS innervation and the anterograde transneuronal tract tracer H129 to label the SS into the same IWAT depot.

METHODS

Animals.

Adult male Siberian hamsters (Phodopus sungorus; ∼3–4 mo old) from our breeding colony were individually housed in a temperature-controlled vivarium (21 ± 2°C, ∼50 ± 10% humidity) with ad libitum access to food (LabDiet Rodent Chow 5001, St. Louis, MO) and water. Hamsters were maintained under 12-h light-dark schedule and habituated to vivarium conditions 1 wk before virus injections. All animal procedures were approved by the Georgia State University Institutional Animal Care and Use Committee and were in accordance with Public Health Service and United States Department of Agriculture guidelines.

Viral injections.

All virus injections were performed according to Biosafety Level 2 standards. Animals (n = 9) were anesthetized with isoflurane (2–3% in oxygen; Baxter Healthcare, Deerfield, IL) inhalation throughout the viral injection procedure. Hamsters were placed in dorsal recumbency, and the skin around the inguinal region was wiped with 10% povidone iodine (Ricca Chemical, Arlington, TX) and then 70% ethanol, finishing with povidone iodine. An incision was made to expose the right IWAT. Pilot studies using each viral tract tracer separately showed that the SNS and SS innervation of the right and left pads were not different from one another except for some increased infections ipsilateral to the injected WAT pad for both innervations (Song CK and Bartness TJ, unpublished observations). A series of PRV152 (gift from Dr. Enquist, Princeton University and Center for Neuroanatomy with Neurotropic Viruses) microinjections (4.5 × 10⁸ pfu/ml; 200 nl/locus) were made directly into seven loci across the IWAT pad followed 24 h later by H129 (gift from Dr. Richard Dix, Georgia State University) microinjections (2.0 × 10⁸ pfu/ml; 200 nl/locus) into the same fat pad. During each viral injection the syringe was held in place for 60 s to prevent efflux of virus. The skin was closed with sterile wound clips. After surgery, ketofen (5 mg/kg sc; Fort Dodge Animal Health, Fort Dodge, IA) was administered for 3 days postinjection to minimize postoperative discomfort.

As a control for possible viral diffusion, we previously demonstrated that the same virus titer of PRV152 or of H129 and volume placed on the surface of the exposed IWAT pad resulted in no infection in the sympathetic chain or DRG, respectively, spinal cord, and brain as opposed to intra-IWAT viral infections (Song CK and Bartness TJ, unpublished observations). Furthermore, surgical isolation of IWAT from the surrounding tissues before H129 or PRV injections resulted in a pattern of infection indistinguishable from those injections into the intact IWAT, suggesting the specific sensory and sympathetic routes of infections originating from the IWAT (Song CK and Bartness TJ, unpublished observations).

Histology.

Hamsters were euthanized 6 days after PRV152 and 5 days after H129 postinjection. The postinjection times differ because of differing rates of viral transit across the neuroaxis as determined in pilot studies (Ryu V and Bartness TJ, unpublished observations). Because one cannot adjust the timing such that all brain areas receive both viruses at approximately the same time, we used the PVH in particular but also a general survey across the neuroaxis to determine these times. Animals were given an overdose of pentobarbital sodium (300 mg/kg) and transcardially perfused with 0.9% heparinized saline followed by 4% paraformaldehyde in 0.1 M PBS. The brains and spinal cords were postfixed in the same fixative for 3–4 h and then transferred to a 30% sucrose solution in 0.1 M PBS containing 0.1% sodium azide at 4°C overnight until they were sectioned at 30 μm. As for the DRG and sympathetic ganglia (from T11 to L3), they were carefully extracted, peeled from the epineurium, and transferred to an 18% sucrose solution in 0.1 M PBS containing 0.1% sodium azide at 4°C overnight. All ganglia and spinal cords were sectioned longitudinally into 20- and 30-μm serial sections, respectively, and mounted in three series onto slides (Superfrost Plus; VWR International, West Chester, PA) with every forth section on the same slide. Generally, this procedure yielded 24–27 sections with each slide containing eight to nine ganglia and nine spinal cord sections. After being dried on the slides, the sections were rehydrated and processed for immunodetection of the selected antigen.

For immunohistochemistry, free-floating brain sections were subsequently rinsed in 0.1 M PBS (2 × 15 min) and 0.1% sodium borohydride in 0.1 M PBS to reduce autofluorescence (20) followed by 30 min incubation in a blocking solution of 10% normal goat serum (NGS) and 0.4% Triton X-100 in 0.1 M PBS. Sections were then incubated in the mixture of primary rabbit anti-herpes simplex virus-1 (HSV-1, 1:2,000; DakoCytomation, Carpinteria, CA) and mouse anti-GFP (1:700; Abcam, Cambridge, MA) antibodies with 2% NGS in 0.4% Triton X-100 in 0.1 M PBS for 18 h. Subsequently, sections were incubated in the appropriate mixture of the secondary goat anti-rabbit Cy3 (1:700; Jackson Immunoresearch, West Grove, PA) and goat anti-mouse Alexa 488 (1:700; Jackson Immunoresearch) antibodies with 2% NGS in 0.4% Triton X-100 in 0.1 M PBS for 2 h. All steps were performed at room temperature. For immunohistochemical controls, the primary antibody was either omitted or preadsorbed with the immunizing peptide overnight at 4°C resulting in abolished immunostaining (data not shown). Sections were mounted onto slides and cover slipped using ProLong Gold Antifade Reagent (Life Technologies, Grand Island, NY).

After rehydration, the DRG, sympathetic ganglia, as well as spinal cords, were processed for immunohistochemical detection of primary rabbit anti-HSV-1 (1:100; DakoCytomation) and mouse anti-GFP (1:500; Abcam) antigens directly on the slides using the same protocol as above.

Quantitative and statistical analysis.

Images were viewed and captured under ×100 and ×200 magnification with the Olympus DP73 imaging photomicroscope (Olympus, Tokyo, Japan), equipped with appropriate filters for Cy3 and Alexa 488. The captured images were evaluated with the aid of Olympus CellSens and the Adobe Photoshop CS5 (Adobe Systems, San Jose, CA) software. After two images were overlaid, exhaustive counts of H129- and PRV152-single neurons as well as H129+PRV152-colocalized neurons were performed by use of the manual tag feature of the Adobe Photoshop CS5 software in every sixth section of the brain and every forth section of the ganglia and the spinal cord to eliminate the likelihood of counting the same neuron twice. Only single labeling was considered for the number/percentage of each tracer labeling. The total number/percentage of neurons was considered as H129 single labeling + PRV152 single labeling + H129 and PRV152 double labeling. The neurons were considered positively labeled based on the fluorescent intensity, cell size, and shape. Both absolute values of those neurons and their corresponding percentages in the brain, ganglia, and spinal cord were averaged across each examined region/nucleus from all hamsters. A mouse brain atlas was used (2, 49) because it matches best with most Siberian hamster brain areas, doing so much better than Syrian hamster brain atlases, and because no Siberian hamster brain atlas is available. For preparation of the microscopic illustrations, the Adobe Photoshop CS5 was used to adjust only brightness, contrast, and to prepare the composite plates.

Data were analyzed by one-way repeated measures analysis of variance (ANOVA) followed by the post hoc Bonferroni's and Holm-Sidak's tests using NCSS (version 2007, Kaysville, UT). Significance was set at P < 0.05. For simplicity and clarity, values with P < 0.05, P < 0.01, and P < 0.001 were all indicated with a single asterisk. All values are presented as means ± SE.

RESULTS

All hamsters infected with H129 and PRV152 remained asymptomatic until 5 days after PRV152 virus injections, whereupon many began to display some symptoms of infection including loss of body weight, some immobility, and an ungroomed coat. Hamsters (n = 9) were euthanized at 5–6 days postinjection when such symptoms became apparent. Five of these animals displayed very similar viral immunolabeling in the brain to the other hamsters and therefore were included in the study. Four other hamsters displayed overinfection of the CNS either by H129 and/or PRV152 indicated by widespread “cloudy plaques” around the overinfected neurons and therefore were excluded from the analyses.

Viral infections in the sympathetic ganglia and DRG.

Viral immunolabeling was distributed across T11-L3 levels of the sympathetic chain with the highest absolute numbers of labeled neurons at T13-L1 and L2 (Fig. 1, A and B). The absolute number as well as the percentage of PRV152-ir neurons was significantly higher (P < 0.05) compared with that of single H129-ir and double H129+PRV152-ir neurons at T13-L1 (Fig. 1B). The number of PRV152-ir neurons also was significantly higher (P < 0.05) than the number of single H129-ir or double H129+PRV152-ir cells at T11 level (Fig. 1B).

Fig. 1.

A: labeling of H129 and PRV152 in the T13-L2 sympathetic ganglia after viral injections into the inguinal white adipose tissue (IWAT). Colocalization of H129 (red) + PRV152 (green) is indicated by arrows. B: total number and percentage of H129-, PRV152-, and H129+PRV152-ir neurons in the T11-L3 sympathetic ganglia. *P < 0.05 vs. H129, #P < 0.05 vs. H129+PRV152. Scale bar = 200 μm.

Both H129 and PRV152 immunolabeling in the DRG pseudounipolar neurons innervating IWAT were associated with spinal cord vertebrae T11-L3 (Fig. 2, A and B). H129-ir DRG neurons were relatively evenly distributed across T11-L3 in accordance with our previous finding (44) when we used the retrograde tracer Fluorogold to label DRG neurons innervating IWAT. The absolute numbers of H129-ir cells were significantly increased (P < 0.05) compared with the numbers of PRV-152- and colocalized H129+PRV152-ir cells at T11, T13, L1, and L2 (Fig. 2B). The percentages of H129-ir cells versus single PRV152- or double H129+PRV152-ir cells also were significantly higher at T11, T13, L1, and L2 (P < 0.05; Fig. 2B). In addition, the percentage of single H129- and PRV152-ir neurons was significantly higher (P < 0.05) than that of double-labeled neurons at L3 vertebral level (Fig. 2B).

Fig. 2.

A: labeling of H129 and PRV152 in the dorsal root ganglia (DRG) (T12-L1) following viral injections into the IWAT. Double-labeled H129 (red) + PRV152 (green) pseudounipolar neurons are indicated by arrows. B: total number and percentage of H129-, PRV152-, and H129+PRV152-ir neurons in the T11-L3 DRG. *P < 0.05 vs. H129, #P < 0.05 vs. PRV152. Scale bar = 100 μm.

Viral infections in the spinal cord.

Both H129 and PRV152-ir cells in the spinal cord were more prevalent in the T13-L1 and L2 sympathetic preganglionic neurons of the IML ipsilateral to the injection site (Fig. 3A). The numbers of single-labeled PRV152 and double-labeled H129+PRV152 neurons were significantly increased compared with single-labeled H129 neurons both at T13-L1 and L2 (P < 0.05; Supplemental Table S1; Fig. 3B). Surprisingly, there was a substantial percentage of H129+PRV152 colocalized neurons (∼52%) at T13-L1 and 48% at L2 (Fig. 3B).

Fig. 3.

A: labeling of H129 and PRV152 in the spinal cord (T13-L2) after viral injections into the IWAT. Colocalization of H129 (red) + PRV152 (green) is indicated by arrows. B: total number and percentage of H129-, PRV152-, and H129+PRV152-ir neurons in the spinal cord. cc, central canal; IML, intermediolateral nucleus. *P < 0.05 vs. H129. Scale bar = 50 μm.

Viral infections in the brain.

The dual injections of the sensory nerve tract tracer H129 and the sympathetic nerve tract tracer PRV152 unilaterally into the same right IWAT appeared bilaterally in the brain with slightly ipsilateral domination to the side of the injection for both viruses. Injections of the two viruses revealed a marked degree of double-infected (labeled) neurons in the brain.

In the hindbrain, the numbers of H129+PRV152 colocalized neurons were significantly increased (P < 0.05) in almost all regions compared with single-labeled H129 or PRV152 neurons (Supplemental Table S1; Fig. 4B). Among the regions with markedly higher absolute numbers of PRV152-ir versus H129-ir neurons were the medial septal nucleus (MS) in the forebrain, the dorsomedial periaqueductal gray (DMPAG) in the midbrain, the prepositus nucleus (Pr), the raphe obscurus nucleus (ROb), and the spinal trigeminal nucleus, caudal part (Sp5C) in the medulla. Regions such as the NTS (Supplemental Table S1; Fig. 4A), the intermediate reticular nucleus (IRt), the lateral reticular nucleus (LRt), the ventral spinocerebellar tract (vsc), and the medullary reticular nucleus, dorsal part (MdD) contained the highest absolute numbers of both single- or double-labeled viruses (P < 0.05 vs. other regions in the hindbrain; Supplemental Table S1). The only region represented mostly by H129-labeled neurons was the lateral paragigantocellular nucleus (LPGi) (72.00 ± 23.12 H129-ir neurons vs. 27.20 ± 2.40 PRV152-ir neurons; Supplemental Table S1 and Fig. 5).

Fig. 4.

A: low and high (inset) magnification of the photomicrograph illustrating single H129 (red), single PRV152 (green), and double H129+PRV152 (arrows) immunolabeling in the nucleus of the solitary tract (NTS) and intermediate reticular nucleus (IRt) following viral injections into the IWAT. B: percentile distribution of H129- and PRV152-infected neurons in the medulla. 4V, forth ventricle; 10N, dorsal motor nucleus of vagus; 12N, hypoglossal nucleus; Gi, gigantocellular reticular nucleus; PCRt, parvicellular reticular nucleus; PMn, paramedian reticular nucleus; py, pyramidal tract; ROb, raphe obscurus nucleus; RPa, raphe pallidus nucleus; sol, solitary tract; Sp5I, spinal trigeminal nucleus, interpolar part. *P < 0.05 vs. H129, #P < 0.05 vs. PRV152. Scale bar = 200 μm.

Fig. 5.

High magnification image of the lateral paragigantocellular nucleus (LPGi) with predominantly H129 labeling after viral injections into the IWAT. GiV, gigantocellular reticular nucleus, ventral part; IOPr, inferior olive, principal nucleus; Li, linear nucleus of the medulla; py, pyramidal tract; RVL, rostroventrolateral reticular nucleus. Scale bar = 100 μm.

The percentages of H129+PRV152 double-labeled neurons in the midbrain and pons were significantly higher (P < 0.05) than those of single-labeled H129 or PRV152 neurons in most regions examined with the exception of the deep mesencephalic nucleus (DpMe), the dorsal raphe nucleus, dorsal part (DRD), the Kölliker-Fuse nucleus (KF), the medial lemniscus (ml), the paranigral nucleus (PN), and the lateral parabrachial nucleus, dorsal part (LPBD) (Supplemental Table S1; Fig. 6B). The number of colocalized neurons was significantly higher (P < 0.05) than the number of single-labeled H129-ir neurons in the dorsal raphe nucleus, interfascicular part (DRI). The number of PRV152-labeled neurons in the DMPAG was significantly increased (P < 0.05) compared with the number of H129-labeled neurons (Supplemental Table S1; Fig. 6A). Among the regions with the heaviest absolute numbers of infected neurons in the midbrain and pons were the PAG, DpMe, the pontine reticular nucleus, oral part (PnO), and the locus coeruleus (LC) (P < 0.05 vs. other regions in the midbrain and pons; Supplemental Table S1).

Fig. 6.

A: low and high (inset) magnification of the photomicrograph showing single H129 (red), single PRV152 (green) and double H129+PRV152 (arrows), immunolabeling in the periaqueductal gray (PAG) and deep mesencephalic nucleus (DpMe) after viral injections into the IWAT. B: percentile distribution of H129- and PRV152-infected neurons in the midbrain and pons. 3N, oculomotor nucleus; Aq, aqueduct; DLPAG, dorsolateral periaqueductal gray; DMPAG, dorsomedial periaqueductal gray; DR, dorsal raphe nucleus; LPAG, lateral periaqueductal gray; mlf, medial longitudinal fasciculus; VLPAG, ventrolateral periaqueductal gray. *P < 0.05 vs. H129, #P < 0.05 vs. PRV152. Scale bar = 200 μm.

In the hypothalamus and thalamus, the numbers of H129+PRV152-colocalized neurons were significantly increased (P < 0.05) compared with the corresponding single-labeled H129 or PRV152 neurons in most regions examined, except for the arcuate hypothalamic nucleus, lateral part (ArcL), the ventromedial hypothalamic nucleus, central part (VMHC), the ventromedial hypothalamic nucleus, dorsomedial part (VMHDM), and the ventromedial hypothalamic nucleus, ventrolateral part (VMHVL) regions that we scarcely infected by either virus (Supplemental Table S1, Fig. 7B) corresponding to our previous finding when only H129 (60) or PRV (4) was injected intra-IWAT. The hypothalamic and thalamic regions possessing the highest absolute numbers of both viruses were the lateral hypothalamic area (LH) (Supplemental Table S1; Fig. 7A), PVH (Fig. 8; Supplemental Table S1) and MPA (Fig. 9A; Supplemental Table S1) (P < 0.05 vs. other regions in the hypothalamus and thalamus; Supplemental Table S1).

Fig. 7.

A: low and high (inset) magnification of the photomicrograph illustrating single H129 (red), single PRV152 (green), and double H129+PRV152 (arrows) immunolabeling in the lateral hypothalamic area (LH) after viral injections into the IWAT. B: percentile distribution of H129- and PRV152-infected neurons in the forebrain (hypothalamus and thalamus). 3V, third ventricle; Arc, arcuate hypothalamic nucleus; DM, dorsomedial hypothalamic nucleus; VMH, ventromedial hypothalamic nucleus. *P < 0.05 vs. H129, #P < 0.05 vs. PRV152. Scale bar = 200 μm.

Fig. 8.

Low and high (inset) magnification of the photomicrograph illustrating single H129 (red), single PRV152 (green), and double H129+PRV152 (arrows) immunolabeling in the paraventricular hypothalamic nucleus (PVH) after viral injections into the IWAT. 3V, third ventricle; AHP, anterior hypothalamic area, posterior part; LH, lateral hypothalamic area; opt, optic tract; VMH, ventromedial hypothalamic nucleus. Scale bar = 200 μm.

Fig. 9.

A: low and high (inset) magnification of the photomicrograph demonstrating single H129 (red), single PRV152 (green), and double H129+PRV152 (arrows) immunolabeling in the medial preoptic area (MPA) after viral injections into the IWAT. B: percentile distribution of H129- and PRV152-infected neurons in the forebrain (preoptic area). aca, anterior commissure; MnPO, median preoptic nucleus; VMPO, ventromedial preoptic nucleus. *P < 0.05 vs. H129, #P < 0.05 vs. PRV152. Scale bar = 200 μm.

In the preoptic area, the numbers of double-labeled neurons were significantly (P < 0.05) higher in the bed nucleus of the stria terminalis, medial division, posterolateral part (BSTMPL), the anteroventral periventricular nucleus (AVPe), the lateral preoptic area (LPO), the lateral septal nucleus, intermediate part (LSI), the median preoptic nucleus (MnPO), the medial preoptic nucleus, lateral part (MPOL), the medial preoptic nucleus, medial part (MPOM), the ventromedial preoptic nucleus (VMPO), and the vascular organ of the lamina terminalis (VOLT) compared with single H129- or PRV152-labeled neurons in each region examined (Supplemental Table S1; Fig. 9, A and B). Among these regions, the highest absolute numbers of the H129, PRV152, and dually infected neurons were detected in the MPOM and LSI regions (P < 0.05 vs. other regions in the preoptic area; Supplemental Table S1). The number of PRV152 singly labeled neurons was markedly increased in the MS region (P < 0.05) compared with H129 single- and H129+PRV152 double-labeled neurons (Supplemental Table S1; Fig. 9B). In the nucleus of the horizontal limb of the diagonal band (HDB) region, the number of H129+PRV152-colocalized neurons was statistically significant (P < 0.05) only when compared with that of the H129-ir neurons (Supplemental Table S1; Fig. 9B). There were no significant differences in numbers of single- or double-labeled neurons in remaining regions of the preoptic area (Supplemental Table S1; Fig. 9B).

DISCUSSION

Previously, we (44, 69) and others (25) demonstrated the distribution of the postganglionic SNS and SS innervation of IWAT using conventional retrograde tract tracers, as well as previously defining the SNS outflow from the brain to WAT (e.g., 4, 13, 31, 52, 57, 59) and sensory inflow to the brain from WAT (60) using transneuronal tract tracers. In these separate viral tract tracing studies of the central SNS and SS circuitries, many common areas were revealed suggesting the possibility of neuroanatomical SNS-SS crosstalk across the neuroaxis. Therefore, here we performed dual microinjections of both the WAT SNS-specific retrograde transneuronal tract tracer PRV152 [given there is no PSNS innervation of WAT (3, 11, 31, 32)] and SS-specific anterograde transneuronal tract tracer H129 into the IWAT and, for the first time, revealed a surprising coincidence of individual neurons participating in both the central SNS outflow from brain to IWAT, but also the central SS inflow from IWAT to brain, suggesting SNS-SS feedback loops across the neuroaxis.

In the present study, we performed exhaustive histological analysis across the hierarchical arrangement of the IWAT-associated SNS outflow from brain to IWAT (i.e., the sympathetic ganglia, the IML horn of the spinal cord, and the brain) as well as sensory inflow from IWAT to brain (i.e., the DRG, the spinal cord and the brain). The results of the present study expand our knowledge of the neuroanatomical organization of WAT SNS-SS circuitries showing a surprising and striking similarity in the significant number of individual neurons participating in both the sympathetic and sensory circuitries across the brain, including more peripheral nervous systems sites such as the sympathetic chain and the IML of the spinal cord. Specifically, we found a remarkable overlap of SNS-SS IWAT-innervating circuits (∼27–75%), as evidenced by colocalization of PRV152 and H129 immunolabeling, collectively suggesting a vast integration of the WAT SNS-SS feedback loops across the neural axis. For example, among the brain regions with the highest overlap of dually infected neurons were the MPOL (∼71%) in the preoptic area, the supramammillary nucleus, medial part (SuMM) (∼73%) in the hypothalamus, the lateral habenula nucleus, lateral part (LHbL) (∼70%) in the thalamus, DpMe (∼66%) and the pedunculopontine tegmental nucleus (PPTg) (∼71%) in the midbrain and pons, respectively, and sp5 (∼75%) in the medulla. The brain areas represented by the highest absolute numbers of the double-infected neurons included the MPOM in the preoptic area; the MPA, PVH, and LH in the hypothalamus; the DpMe and PAG in the midbrain; the LC and PnO in the pons; and the IRt, LRt, MdD, the raphe pallidus nucleus (RPa), NTS, and vsc in the medulla. In preliminary studies that were not exhaustive as the present study (Song CK and Bartness TJ, unpublished observations), the existence of these individual SNS-SS feedback neurons also was seen after injection of both viruses into EWAT in Siberian hamsters suggesting that such feedback is not unique to IWAT.

The function of these SNS-SS feedback circuits is not presently known. Because activation of the SNS innervation of WAT is the principal initiator of lipolysis in mammals, including of course humans (for review see Refs. 7–9) and because of our previous studies demonstrating increases in the SNS drive to WAT in response to glucoprivation-induced increase in WAT sensory nerve electrophysiological multiunit activity (60), a well-known WAT lipolytic stimulus (15, 21, 33, 34, 63), it appears that WAT sensory nerves are sensitive to some aspect of lipolysis (e.g., free fatty acids or glycerol) or an associated factor of SNS activation in WAT [e.g., prostaglandin E₂ (51)]. In addition, the sensory nerves innervating WAT are sensitive to intra-WAT leptin administration (44, 47, 56) suggesting a role of these feedback circuits in the control of leptin synthesis/secretion.

We optimized the double-viral technique by matching viral rates of progression through the neuroaxis as closely as possible to create approximately equal percentages for single PRV152- and H129-labeled neurons across the brain. Because of the virology “principle of exclusion” (41) whereby infection of one cell by a virus can preclude its infection by a second different virus (58), this makes it likely that the number of double-labeled neurons actually was underestimated in the present study despite the surprisingly large number of double-infected neurons that were revealed.

Regardless of mostly similar SNS and SS locations in the brain regions examined, several brain sites appeared to have divergent sympathetic output/sensory input. Among the brain regions with a predominantly sympathetic efferent output to IWAT compared with a sensory afferent input from IWAT were the MS in the forebrain; the DMPAG in the midbrain; and the Pr, ROb, and Sp5C in the medulla. We previously noted participation of the lateral septal nucleus in the SNS outflow to WAT using PRV in Siberian hamsters and laboratory rats (1, 4, 46, 52, 57, 59) as well as others in the domestic pig (23, 24), but no functional role for this nucleus in WAT lipid mobilization has been identified to date. We reported that the PAG was among other indisputable brain regions sending SNS output to WAT in Siberian hamsters (4, 46, 59) as we (4) and others did in laboratory rats (1) and midbrain lesions encroaching on the PAG result in obesity (14). The role of the Pr and the Sp5C in relation with WAT remains elusive. Finally, we found previously (4, 31, 57, 59), as in the present report, that the ROb contains neurons that are part of the SNS outflow to WAT, but functions for the control of WAT are unknown. Collectively, based on the predominate PRV versus H129 infection, it appears that these neuronal populations may not receive the sensory neural input from WAT, although the absence of proof is not the proof of absence. If the viral tracing is as it appears, then these areas/neurons may not receive sensory feedback from WAT or obtain WAT feedback via circulating factors that can cross the blood-brain barrier such as the products of lipolysis (free fatty acids and glycerol).

By contrast to the brain regions with a predominantly sympathetic efferent output to WAT, the only region that received mostly SS input from IWAT was the LPGi. The sparse PRV152 immunolabeling within LPGi was unexpected given that this nucleus, along with the MPA, the ventromedial PAG, and the KF, were reported to be involved in respiratory and/or sympathetic control and send direct afferent projections to the RPa (6, 18, 37, 39, 64). The RPa, in turn, is part of the SNS outflow to WAT as we have shown (4, 59) and BAT as we (5) and others have shown (e.g., 12, 16, 45).

Of somewhat a surprise was that the dual microinjections of the SNS-specific tract tracer PRV152 and SS-specific tract tracer H129 into the same IWAT yielded a substantial number of double-infected sympathetic preganglionic neurons of the IML (∼52% at T13-L1 and ∼48% at L2) suggesting an short SNS-SS neural feedback loop from IWAT. There is some functional precedence for this. Spinal cord-severed patients that have decentralized sympathetics and afferents below the level of the severed cord or spinal cord lesion show increases in SNS activity with bladder depression (as measured by norepinephrine spillover which does have many caveats; see Ref. 8), but not above the lesion level intimating the possibility of a spinal reflex arc producing this response that could affect lipolysis (40), perhaps having such short SNS-SS neural feedback loops as their underlying neurology. In addition, although not implicating only the spinal cord, we previously demonstrated SNS-stimulated lipid mobilization when chronic decerebrate rats that clearly have no forebrain involvement but instead are functioning with only a truncated neuroaxis from the mesodiencephalic juncture [i.e., an abbreviated neuroaxis consisting of the hindbrain, spinal cord, and pre- and postganglionic sympathetic nerves (36)] were food deprived. Finally, of additional potential importance in the present study was that both sympathetic and sensory ganglia had a cross representation of roughly equal (∼20%) numbers of neurons. That is, some sensory-labeled (H129-infected) neurons were in the sympathetic chain and some sympathetically labeled (PRV-infected) neurons resided in the DRG. In support of these findings that may seem at first to be nonspecific viral labeling, sensory afferent somata are found within the superior cervical sympathetic ganglia (68) and sympathetic afferent somata are found within thoracolumbar DRG, the latter considered to be non-nociceptive visceral sensory neurons involved in the reflexive control and/or homeostasis (17). These SNS and SS ganglia findings suggest peripheral interaction between the SNS and SS and support the notion of the neuroanatomical reality of WAT SNS-SS short neural feedback loops.

Collectively, the results of the present neuroanatomical study suggest functionally for the effect known as the “adipose afferent reflex” (AAR) promoted by Niijima with his findings that intra-epididymal WAT leptin injections increase the sensory electrophysiological activity of WAT afferents [a finding we have confirmed recently (44)] that, in turn stimulate the sympathetic neural electrophysiological activities to the same WAT depots (47, 48). This AAR has been extended recently to other WAT pads, but with renal sympathetic nerve activity increases (adipose afferent renal reflex, AARR) with related changes in blood pressure and other cardiovascular responses (19, 22, 35, 42, 55, 66, 67).

Regarding the “long SNS-SS neural feedback loops,” the PVH appears strongly implicated in the AARR and AAR and is one of the many areas that we demonstrated to have double-infected PRV152+H129 neurons in the present study. We previously defined the central origins of the SNS outflow from the brain to WAT using PRV152 showing predominantly unilateral sympathetic outflow from each half of the PVH to ipsilaterally located IWAT depots (4). Attempts to block “basal lipolysis” by unilateral PVH electrolytic lesions (PVHx) did not increase lipid accumulation in WAT pads ipsilateral to the side of the PVHx nor were there laterality effects on food deprivation-induced WAT lipid mobilization in hamsters bearing unilateral or bilateral PVHx. Indeed, unilateral or bilateral PVHx did not attenuate or abolish food deprivation-induced lipid mobilization strongly suggesting that an intact PVH is not necessary for an apparent full lipolytic response to this energy availability challenge (27). Therefore, we postulated that the distributed system of nodes in the central SNS outflow circuitry may compensate for the loss of the PVH in the case of food deprivation-induced lipid mobilization (27). This lack of necessity of the PVH for food deprivation-induced lipid mobilization is not seen for the AAR and AARR. That is the AAR can be triggered by intra-IWAT injection of capsaicin (at doses that stimulate the sensory nerves, but are not toxic to them) in otherwise normal laboratory rats. PVH microinjection of the glutamate NMDA receptor antagonist AP5 or MK-801, or CNQX the non-NMDA glutamate receptor (AMPA/kainate receptor) antagonist diminishes the intra-IWAT capsaicin-induced AAR, as well as the increases in renal sympathetic nerve activity and blood pressure effects to this response [the AARR (22)]. In addition, the combination of AP5 and CNQX PVH microinjections interact to nearly block the AAR induced by intra-IWAT capsaicin and by intra-IWAT leptin (22). Furthermore, kainic acid-induced chemical PVH lesions also block the AAR in response to intra-IWAT capsaicin (55). These AAR findings suggest that, despite the distributed system of SNS-SS feedback neurons in the present study, the PVH, although not necessary for food deprivation-induced lipolysis, appears necessary for the AAR suggesting these responses may involve separate neurologies. Other areas also may be involved in the AAR SNS-SS long neural feedback loops, but they have not been explored to date; however, they may have been identified by the dual labeling with PRV152 and H129 infections in the present study.

Our data imply that WAT pads have the potential to control their SNS drives via sensory feedback and/or help to explain the differential sympathetic drives that are fat-pad specific depending on the lipolytic stimulus (for review see Refs. 7, 8, 10). Furthermore, the present data suggest, at the neuroanatomical level of analysis, important integration of both the sympathetic and the sensory systems thereby providing a valuable and expanded view on the congruent diversity of the structural organization of the SNS and SS. In addition, the present data show the neuroanatomical reality of individual neurons across the neuroaxis participating in both the SNS outflow from the brain to WAT and the sensory inflow from WAT to the brain, perhaps demonstrating the neuroanatomical basis of the so-called AAR and/or the neurologies underpinning some of the fat pad-specific responses involved in adipose functions.

Perspectives and Significance

Understanding the functions of these short- and long-neural feedback loops seems critical to reverse the obese condition or inhibit its development. That is, it is well known in the scientific literature, as well as from many personal experiences, that once lipid mass decreases, potent compensatory responses are initiated to promote increasing adiposity (for review see Ref. 50). Thus, for example, knowing the stimuli sensed by WAT afferent nerves that are associated with increases in adipose tissue lipid stores [for example, leptin as a paracrine factor given its ability to stimulate WAT afferents locally (e.g., 44, 48, 62, 66)], then perhaps a “false signal” could be generated pharmacologically via this sensory neural conduit to the brain to indicate a high degree of adiposity despite the actual decreases in adiposity and thereby prevent the compensatory adiposity-promoting responses. At present, this appears fanciful, but perhaps not impossible.

GRANTS

This work was funded by National Institutes of Health R37 DK-35254.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: V.R. and T.J.B. conception and design of research; V.R. performed experiments; V.R. analyzed data; V.R. and T.J.B. interpreted results of experiments; V.R. prepared figures; V.R. drafted manuscript; V.R. and T.J.B. edited and revised manuscript; V.R. and T.J.B. approved final version of manuscript.

Glossary

AAR

Adipose afferent reflex

AARR

Adipose afferent renal reflex

AMPA

α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

AP5

(2R)-amino-5-phosphonovaleric acid

cc

Central canal

CNQX

6-Cyano-7-nitroquinoxaline-2,3-dione

CNS

Central nervous system

DRG

Dorsal root ganglia

EWAT

Epididymal white adipose tissue

GFP

Green fluorescent protein

HSV-1

Herpes simplex virus-1

IML

Intermediolateral nucleus

IWAT

Inguinal white adipose tissue

MK-801

Dizocilpine

NGS

Normal goat serum

NMDA

N-methyl-d-aspartate

PBS

Phosphate-buffered saline

PRV

Pseudorabies virus

PSNS

Parasympathetic nervous system

RWAT

Retroperitoneal white adipose tissue

SNS

Sympathetic nervous system

SS

Sensory system

Brain

10N

Dorsal motor nucleus of vagus

12n

Root of hypoglossal nerve

3PC

Oculomotor nucleus

7N

Facial nucleus

A5

Noradrenaline cells

AHA

Anterior hypothalamic area, anterior part

AHC

Anterior hypothalamic area, central part

AHP

Anterior hypothalamic area, posterior part

AP

Area postrema

Arc

Arcuate hypothalamic nucleus

ArcL

Arcuate hypothalamic nucleus, lateral part

ArcLP

Arcuate hypothalamic nucleus, lateroposterior part

ArcMP

Arcuate hypothalamic nucleus, medial posterior part

AVPe

Anteroventral periventricular nucleus

BST

Bed nucleus of the stria terminalis

BSTMA

Bed nucleus of the stria terminalis, medial division, anterior part

BSTMPL

Bed nucleus of the stria terminalis, medial division, posterolateral part

CGPn

Central gray of the pons

CnF

Cuneiform nucleus

DLPAG

Dorsolateral periaqueductal gray

DM

Dorsomedial hypothalamic nucleus

DMC

Dorsomedial hypothalamic nucleus, compact part

DMD

Dorsomedial hypothalamic nucleus, dorsal part

DMPAG

Dorsomedial periaqueductal gray

DMTg

Dorsomedial tegmental area

DMV

Dorsomedial hypothalamic nucleus, ventral part

DPGi

Dorsal paragigantocellular nucleus

DpMe

Deep mesencephalic nucleus

DRD

Dorsal raphe nucleus, dorsal part

DRI

Dorsal raphe nucleus, interfascicular part

DRVL

Dorsal raphe nucleus, ventrolateral part

DTM

Dorsal tuberomammillary nucleus

Fu

Bed nucleus of the stria terminalis, fusiform part

Gi

Gigantocellular reticular nucleus

GiA

Gigantocellular reticular nucleus, α part

HDB

Nucleus of the horizontal limb of the diagonal band

IOBe

Inferior olive, β subnucleus

IOC

Inferior olive, subnucleus C of medial nucleus

IOD

Inferior olive, dorsal nucleus

IODM

Inferior olive, dorsomedial cell group

IOM

Inferior olive, medial nucleus

IOPr

Inferior olive, principal nucleus

IRt

Intermediate reticular nucleus

KF

Kölliker-Fuse nucleus

LA

Lateroanterior hypothalamic nucleus

LC

Locus coeruleus

LDTg

Laterodorsal tegmental nucleus

LH

Lateral hypothalamic area

LHbL

Lateral habenular nucleus, lateral part

LHbM

Lateral habenular nucleus, medial part

LPAG

Lateral periaqueductal gray

LPBC

Lateral parabrachial nucleus, central part

LPBD

Lateral parabrachial nucleus, dorsal part

LPBV

Lateral parabrachial nucleus, ventral part

LPGi

Lateral paragigantocellular nucleus

LPO

Lateral preoptic area

LRt

Lateral reticular nucleus

LRtPC

Lateral reticular nucleus, parvicellular part

LSI

Lateral septal nucleus, intermediate part

LSV

Lateral septal nucleus, ventral part

MdD

Medullary reticular nucleus, dorsal part

MdV

Medullary reticular nucleus, ventral part

ml

Medial lemniscus

mlf

Medial longitudinal fasciculus

MnPO

Median preoptic nucleus

Mo5

Motor trigeminal nucleus

MPA

Medial preoptic area

MPB

Medial parabrachial nucleus

MPOL

Medial preoptic nucleus, lateral part

MPOM

Medial preoptic nucleus, medial part

MS

Medial septal nucleus

MVeMC

Medial vestibular nucleus, magnocellular part

MVePC

Medial vestibular nucleus, parvicellular part

NTS

Nucleus of the solitary tract

PaAP

Paraventricular hypothalamic nucleus, anterior parvicellular part

PAG

Periaqueductal gray

PaLM

Paraventricular hypothalamic nucleus, lateral magnocellular part

PaMM

Paraventricular hypothalamic nucleus, medial magnocellular part

PaMP

Paraventricular hypothalamic nucleus, medial parvicellular part

PaPo

Paraventricular hypothalamic nucleus, posterior part

PaV

Paraventricular hypothalamic nucleus, ventral part

PCRt

Parvicellular reticular nucleus

PCRtA

Parvicellular reticular nucleus, α part

PF

Parafascicular thalamic nucleus

PH

Posterior hypothalamic area

PMD

Premammillary nucleus, dorsal part

PMn

Paramedian reticular nucleus

PMV

Premammillary nucleus, ventral part

PN

Paranigral nucleus

PnC

Pontine reticular nucleus, caudal part

PnO

Pontine reticular nucleus, oral part

PnV

Pontine reticular nucleus, ventral part

PPTg

Pedunculopontine tegmental nucleus

Pr

Prepositus nucleus

PrC

Precommissural nucleus

PS

Parastrial nucleus

PSol

Parasolitary nucleus

pv

Periventricular fiber system

PVH

Paraventricular hypothalamic nucleus

RC

Raphe cap

RMg

Raphe magnus nucleus

ROb

Raphe obscurus nucleus

RPa

Raphe pallidus nucleus

RVL

Rostroventrolateral reticular nucleus

SCh

Suprachiasmatic nucleus

SChDM

Suprachiasmatic nucleus, dorsomedial part

SChVL

Suprachiasmatic nucleus, ventrolateral part

SI

Substantia innominata

SolC

Nucleus of the solitary tract, commissural part

SolCe

Nucleus of the solitary tract, central part

SolDL

Solitary nucleus, dorsolateral part

SolG

Nucleus of the solitary tract, gelatinous part

SolI

Nucleus of the solitary tract, interstitial part

SolIM

Nucleus of the solitary tract, intermediate part

SolM

Nucleus of the solitary tract, medial part

SolV

Solitary nucleus, ventral part

SolVL

Nucleus of the solitary tract, ventrolateral part

sp5

Spinal trigeminal tract

Sp5C

Spinal trigeminal nucleus, caudal part

Sp5I

Spinal trigeminal nucleus, interpolar part

Sp5O

Spinal trigeminal nucleus, oral part

SpVe

Spinal vestibular nucleus

Su3

Supraoculomotor periaqueductal gray

Su3C

Supraoculomotor cap

SubCD

Subcoeruleus nucleus, dorsal part

SubCV

Subcoeruleus nucleus, ventral part

SuMM

Supramammillary nucleus, medial part

TC

Tuber cinereum area

VLPAG

Ventrolateral periaqueductal gray

VMHC

Ventromedial hypothalamic nucleus, central part

VMHDM

Ventromedial hypothalamic nucleus, dorsomedial part

VMHVL

Ventromedial hypothalamic nucleus, ventrolateral part

VMPO

Ventromedial preoptic nucleus

VOLT

Vascular organ of the lamina terminalis

VP

Ventral pallidum

vsc

Ventral spinocerebellar tract

VTA

Ventral tegmental area