Identification of neuronal subpopulations that project from hypothalamus to both liver and adipose tissue polysynaptically Sarah Stanleya,1, Shirly Pintoa,b,1, Jeremy Segala,c, Cristian A. Péreza, Agnes Vialea,d, Jeff DeFalcoa,e, XiaoLi Caia, Lora K. Heislerf, and Jeffrey M. Friedmana,g,2 aLaboratory of Molecular Genetics, Rockefeller University, New York, NY 10065; bMerck Research Laboratories, Rahway, NJ 07065; cDepartment of Pathology, Weill Cornell Medical College, New York, NY 10065; dGenomics Core Laboratory, Memorial Sloan Kettering Hospital, New York, NY 10065; eRenovis, San Fransisco, CA 94080; fDepartment of Pharmacology, University of Cambridge, Cambridge, CB2 1PD United Kingdom; and gHoward Hughes Medical Institute, Rockefeller University, New York, NY 10065 Contributed by Jeffrey M. Friedman, March 4, 2010 (sent for review January 6, 2010) The autonomic nervous system regulates fuel availability and of the solitary tract (NTS) (6) all modulate metabolic activity in energy storage in the liver, adipose tissue, and other organs; how- liver and white fat. Neuronal tracing studies confirm these areas, ever, the molecular components of this neural circuit are poorly and other studies innervate peripheral organs involved in car- understood. We sought to identify neural populations that project bohydrate and lipid metabolism (7–9). However, little is known from the CNS indirectly through multisynaptic pathways to liver of the site, organization, or connectivity of the CNS neural pop- and epididymal white fat in mice using pseudorabies virus strains ulations that effect metabolic responses. expressing different reporters together with BAC transgenesis and To further delineate the neural networks integrating metab- immunohistochemistry. Neurons common to both circuits were olism in hepatic and adipose tissue, we sought to map the syn- identified in subpopulations of the paraventricular nucleus of the aptically linked multineuronal efferent pathways from the CNS hypothalamus (PVH) by double labeling with markers expressed in to liver and epididymal white fat (eWAT). We employed the viruses injected in both sites. The lateral hypothalamus and arcuate retrograde neuronal tracer pseudorabies virus (PRV) to identify nucleus of the hypothalamus and brainstem regions (nucleus of the CNS regions innervating liver or eWAT, and we showed sig- solitary tract and A5 region) also project to both tissues but are nificant overlap in specific molecularly defined neuronal pop- labeled at later times. Connections from these same sites to the PVH ulations that project indirectly to both organs. These data suggest were evident after direct injection of virus into the PVH, suggesting that specific hypothalamic and brainstem neural populations that these regions lie upstream of the PVH in a common pathway to contain potential command-like neurons that form a circuit to liver and adipose tissue (two metabolically active organs). These integrate hepatic and epididymal adipose sympathetic outflow. common populations of brainstem and hypothalamic neurons This is in contrast to both the classic theory of sympathetic express neuropeptide Y and proopiomelanocortin in the arcuate activation as a single unit, described by Cannon (1), and the nucleus, melanin-concentrating hormone, and orexin in the lateral more recent data proposing independent autonomic inputs to hypothalamus and in the corticotrophin-releasing hormone and individual organs, which suggests coordinated sympathetic acti- oxytocin in the PVH. The delineation of this circuitry will facilitate a vation of functionally related organs. functional analysis of the possible role of these potential command- fl like neurons to modulate autonomic out ow and coordinate met- Results abolic responses in liver and adipose tissue. Time Course of Multisynaptic CNS Projections to Hepatic Tissue and White Adipose Tissue. We first mapped the synaptically linked autonomic nervous system | neuronal tracing | pseudorabies virus multineuronal efferent pathways from the CNS to liver and eWAT. We established the hierarchy of the projections by fol- he autonomic nervous system plays a prominent role in lowing the sites of PRV infection over time after injection of Tmodulating carbohydrate and lipid metabolism. The original PRV152 into the liver or eWAT. The infection course was grou- theories of sympathetic activation (1) proposed a body-wide ped into five phases based on the pattern observed at different fl increase in sympathetic out ow. However, later studies suggested times postinfection; these phases trace the course of infection fi differential sympathetic activation on speci c organs through along chains of synaptically connected neurons. fi distinct autonomic projections, which permits more nely tuned Liver. At the earliest time point after hepatic PRV152 infection control of metabolism (2, 3). (3–4 days postinjection or phase I), GFP was detected in the The liver and adipose tissue play important roles in fuel storage spinal cord only. At 4–5 days postinjection (phase II), GFP was and release. These organs respond to altered energy availability detected in the brainstem [motor nucleus of the vagus (10N) and with a set of homeostatic responses mediated by humoral factors pontine reticular nucleus] with limited expression in the para- fl and autonomic out ow. For example, activation of hepatic sym- ventricular nucleus of the hypothalamus (PVH), zona incerta, pathetic aminergic and peptidergic innervation increases glucose and parasubthalamic nucleus (SI Methods and Table S1). Areas output and modulates fatty acid transport. Conversely, para- sympathetic activity decreases hepatic glucose output and increases carbohydrate storage (4). Likewise, the sympathetic innervation of Author contributions: S.S., S.P., J.S., and J.M.F. designed research; S.S., S.P., J.S., A.V., J.D., white adipose tissue induces lipolysis (5) and alters glucose uptake. X.C., and L.K.H. performed research; C.A.P. contributed new reagents/analytic tools; S.S., Thus, in general, parasympathetic activity favors fuel storage, S.P., and J.S. analyzed data; and S.S., S.P., J.S., L.K.H., and J.M.F. wrote the paper. whereas sympathetic activity increases the fuel available for im- The authors declare no conflict of interest. mediate use. Freely available online through the PNAS open access option. Many CNS regions regulate autonomic outflow to hepatic or 1S.S. and S.P. contributed equally to this work. adipose tissue to influence peripheral energy homeostasis. In 2To whom correspondence should be addressed. E-mail: [email protected]. particular, the ventromedial hypothalamus (VMH), dorsomedial This article contains supporting information online at www.pnas.org/cgi/content/full/ hypothalamus (DMH), lateral hypothalamus (LH), and nucleus 1002790107/DCSupplemental. 7024–7029 | PNAS | April 13, 2010 | vol. 107 | no. 15 www.pnas.org/cgi/doi/10.1073/pnas.1002790107 Downloaded by guest on September 25, 2021 with GFP staining during phase III (5–6 days postinjection) LH, and ARC (Fig. 2 A–E, respectively). The time course of the included the NTS, intermediate and medullary reticular nuclei, dual infection (Table S3) revealed that the PVH was the first CNS dorsal periolivary region, and raphe magnus. Phase IV (6–7 days site in which dual-labeled neurons could be identified. postinjection) revealed GFP expression in the majority of In summary, simultaneous injection of liver and eWAT with hypothalamic sites such as DMH, VMH, arcuate nucleus of the isogenic PRV strains expressing different reporters identified hypothalamus (ARC), and suprachiasmatic nucleus as well as dual-labeled neurons in the hypothalamic PVH followed at later central amygdaloid nucleus, whereas M1 cortical infection times by additional areas, predominantly the LH, ARC, NTS, occurred only at the latest stage infections (phase V; 7–8 days and A5 region. Therefore, common neurons are involved in the postinjection). The predominant sites of infection appearing at outflow circuits to both liver and eWAT. each stage are illustrated in Fig. 1. Neurons in the parvocellular PVH, particularly in the posterior part (PaPo), are infected, Characterization of Common Neuronal Populations. We next sought which is in keeping with known projections to the spinal cord and to chemically define the common neural populations innervating brainstem regions. both liver and eWAT in areas with significant numbers of dual- White adipose tissue. Phase V PRV infection after eWAT injection labeled neurons: PVH, LH, ARC, and NTS. showed a qualitatively similar pattern to that after liver injection Common neuronal populations in the PVH. In the PVH, PRV152 (Table S2). Also, the time course of infection after eWAT injec- infection after eWAT injection colocalized with neurons expressing tion showed early involvement of the PVH (SI Methods and Table AVP, oxytocin (OT), corticotrophin-releasing hormone (CRF), S2). However, unlike the hierarchy of infection after liver PRV and pro-TRH immunoreactivity. However, PRVBaBlu infection injection, eWAT injection resulted in early involvement of LH after liver injection colocalized with neurons expressing OT and and cortical regions. CRF but not AVP or pro-TRH immunoreactivity (Fig. 3 A–F). Dual These data indicated that the central multisynaptic outflow PRV injection (PRVBaBlu into liver and PRV152 into eWAT) with pathway to liver comprises many brainstem and forebrain regions subsequent immunohistochemistry (IHC) for OT or CRH identi- with proximal involvement of the NTS, 10N, and PVH.