One Nervous System: Critical Links Between Central and Peripheral Nervous System Health and Implications for Obesity and Diabetes

As someone trained with a PhD in Neuroscience who has taught undergraduate and graduate courses in this field, I know firsthand how singularly focused we (neurologists, neuroscientists, psychiatrists, and neurological surgeons) are on the brain and how little we acknowledge an integral role of the peripheral nervous system (PNS) in neurological health and disease. The seminal neuroscience textbook by Kandel, Schwartz, Jessell, and colleagues, Principles of Neural Science (1), covers the neuromuscular junction and somatosensation but devotes zero chapters to the rest of the PNS, as one key example of overlooking the PNS in the neuroscience field. Similarly, the largest scientific conference in the world, hosted by the Society for Neuroscience, features mostly research on the brain with relatively little representation of findings in the PNS. We have known about the existence of a PNS since the third century BCE, when physician-philosopher Herophilus of Chalcedon demonstrated that nerves connected the brain to the interior of organs, and not just the vasculature, as was the common thought at that time (2). He also used a holistic approach to the treatment of neurological disorders that included the whole nervous system and distinguished between sensory and motor nerves as subtypes in the PNS. As I will lay out in this article, emerging evidence increasingly indicates that neurological diseases and disorders that impact the central nervous system (CNS) (brain and spinal cord) tend to also have effects in the PNS and vice versa. We can no longer think of brain disorders as solely brain disorders when we have not investigated PNS involvement—because when we do look, both arms of the nervous system tend to be implicated (as outlined in the examples in Table 1). This is especially important for studies of neurometabolism, or the system- and cellular-level cross talk between the nervous system and metabolic regulation, especially since the orchestration of neurometabolism underlies much of the study of energy balance homeostasis and diseases like obesity and diabetes.

Essentially all the same cell types that are important for brain function (neurons; immune cells including brain macrophages, microglia, and more; blood and lymphatic/glymphatic vasculature; and glial cells like astrocytes and oligodendrocytes) are also cell types important for peripheral nerve function (where neurons and their long axons extend into tissues and organs, with support from vasculature, neuroimmune cells, and glial support cells, including numerous types of Schwann cells) (3) (Fig. 1). The same neuropeptides we investigate in the hypothalamic control of appetite and energy expenditure pathways (e.g., neuropeptide Y [NPY], pituitary adenylate cyclase–activating polypeptide [PACAP], calcitonin gene–related peptide [CGRP], and many more) are also expressed in the peripheral nerves that innervate our adipose tissues, but are vastly understudied there despite the likelihood that they exert metabolic effects across multiple tissues outside the brain (4,5) (Fig. 2). In support of this idea, numerous human genes have been implicated through genome-wide association studies for metabolic diseases like obesity and diabetes as clearly important for the nervous system and neuronal functions (6), implicating dysregulation of the nervous system’s energy balance homeostasis in the development of obesity. However, it is glaring that mechanistic investigations of these genes have almost solely been focused on their involvement in brain processes, with potential functions in peripheral nerves ignored, despite similar expression of these genes in the PNS. For example, brain-derived neurotrophic factor (BDNF), its receptor TrkB, Calca, and many more genes associated with human obesity are expressed in both the brain and in peripheral nerves, but have almost exclusively been studied in the brain. Relatedly, results of several studies with investigation of proteins that can act as axon outgrowth signals (Sema7a and Slit2 as prime examples) have implicated their importance in adipose tissue functions, but in these studies measurements of tissue innervation, neurite density, or nerve function, that likely were contributing to the adipose phenotypes, were completely ignored (7,8).

As similar as the structure-function relationships are between the CNS and PNS, important differences remain as well. One difference is the relative lack of protection in the PNS, without the blood-brain and blood–cerebrospinal fluid interfaces/barriers in the CNS, leaving the PNS with less protection from peripheral toxins, inflammatory signals, and damaging metabolites that can lead to glucotoxicity or lipotoxicity with diabetes or obesity. Similarly, the PNS is also directly exposed to the circulating hormones, nutrients, immune cells, and many more blood and lymphatic components that allow interorgan communication and a more diverse fuel supply. Another distinction is that the neuronal cell bodies in the PNS are housed in specialized structures called ganglia that are synaptically connected to the CNS, where nearby Schwann cells, other glial cells, and immune cells reside. Some are “intrinsic” ganglia that sit directly on top of organs like the intestine, heart, and pancreas (Fig. 2), while other neuronal cell bodies are in spinal ganglia (e.g., dorsal root ganglia or sympathetic chain ganglia), which are sometimes at long distances from their axon terminals in tissues like skin and muscle (of note, the longest axon in the human body is the sciatic nerve, which can measure >1 m in length) (9). This creates long distances for cell survival signals that originate at the cell body to reach the termini of PNS axons, a challenge not faced by neurons in the brain. In the brain, axons, dendrites, and cell bodies are housed in relatively close proximity, forming shorter neural circuits that are nearby to support cells like glia. Furthermore, myelin chemistry is largely similar between the CNS and PNS, but there are differences in composition (10).

Finally, the brain’s synapses are nearly all classical synapses (presynaptic neuron to postsynaptic neuron), whereas in the PNS we know too little about specialized nerve terminals and junctions, aside from the well-characterized skin nerve ending structures and the neuromuscular junction (Fig. 2). Only recently was a synaptic structure first described in adipose tissues, the neuro-adipose nexus (NAN) (11–13) for example, despite the critical role of adipose nerves in the control of metabolic health and adipose tissue functions (14,15). Nerve terminal structures are far better characterized in muscle and skin (Fig. 2). The NAN features axon terminals with presynaptic protein markers and synaptic vesicle markers and is comprised of both tyrosine hydroxylase–expressing fibers and calcitonin gene–related peptide (CGRP) fibers, indicating that NANs may be a mix of sensory and sympathetic nerve endings. Similar to the neuromuscular junction, the NANs are myelinated leading up to the start of the terminal structure. NANs are relatively infrequent in the tissue, and typically appear at the tissue surface, but are found in nearly all mouse adipose depots. They increase in number with obesity and aging, as we demonstrated recently, potentially in response to inflammation in the tissue (14,15).

Many who conduct biomedical research focused on peripheral tissues and organs tend to ignore the presence of a peripheral nerve supply that may impact tissue functions. For example, individuals in my laboratory continue to meet colleagues who are surprised that adipose tissues are innervated, and not just clusters of lipid-laden adipocytes that communicate with the brain solely through the hormone leptin or other adipokines, when we know from surgical, chemical, genetic, and viral denervation experiments that adipose nerves are important for the tissue’s function and provide bidirectional neural communication with the brain. Unlike circulating factors, a peripheral nerve–mediated communication route is swift and directed and can impact a single tissue or organ without systemic effects.

Now that advanced tissue clearing and microscopy techniques have evolved, antibodies have improved, and more nerve reporter mouse lines are available for experiments, investigation of tissue and organ innervation patterns has revealed a robust network of vagal and spinal afferents and efferents that enable bidirectional neural communication between the brain and peripheral tissues, including those important for metabolic health (Fig. 2). The broad categorization of nerve subtypes (e.g., vagal and spinal sensory, motor, autonomic/sympathetic) likely underrepresents even more subdivisions of axonal subtypes, with their own orchestration of neurotransmitter and neuropeptide signals, or individualized responses to the plasticity-inducing cues from axon outgrowth signals and neurotrophic factors. Newer single-cell -omics studies of ganglia are revealing some of this neuronal diversity in the PNS (16–18). With emphasis more recently, the function of sensory nerves is an important new horizon for neuroscience research because we know that the body’s interoceptive and exteroceptive sensations of chemical, mechanical, heat, and other signals are integral for our cognitive experience of the world, and there is strong evidence that PNS function impacts brain health, including mental health (19,20). For example, patients with peripheral neuropathy report cognitive impairment and depression (Table 1).

Neuropeptides are likely understudied due to their small size and short half-life and because of the difficulty in capturing them with use of traditional methods like Western blots or proteomics. Other challenges include the persistent lack of reliable antibodies for G-protein–coupled receptors, such as the adrenergic receptors that respond to catecholamines like norepinephrine (NE). Lastly, an important research and gene therapy goal is the targeting of viral vectors to the PNS while avoiding off target transduction in the brain and spinal cord, which our group has been optimizing for adipose nerves and others have optimized for gut and pancreatic nerves.

Diabetes and obesity are prime examples of diseases with whole nervous system involvement. For example, body weight homeostasis is regulated by a complex interplay among nutrients, circulating neuroendocrine hormones that are released from peripheral tissues to act in the brain and other tissues, and coordination by the peripheral nerves innervating the gut, adipose, pancreas, and more. Even the release of hormonal signals from endocrine tissues is largely under PNS control; as a recent example, the release of intestinal incretin hormones was found to be under sympathetic nerve control (21), and the question of whether obesity-associated peripheral neuropathy may impede these neuroendocrine signals in the gut remains unexplored. Glucose regulation is also not a purely cell-autonomous process of glucose-stimulated insulin release from β-cells in the pancreas. It involves neuronal sensors of nutrient status, pancreatic nerves that control glucose homeostasis, and the action of intestinal incretin hormones that improve diabetes and whole-body metabolic health.

The most effective treatments for human obesity have consistently impacted the nervous system, from bariatric surgery to appetite suppressants and the newer broad-acting glucagon-like peptide 1 (GLP-1) and GIP dual agonists. Indeed, GLP-1 and GIP receptors are expressed both in the brain and in peripheral nerves (22,23). By contrast, treatments for metabolic disease that prompt rapid weight loss can have unintended negative consequences on the peripheral nerves, causing neuropathies instead of preventing them (24).

For decades investigators have worked to capitalize on the antiobesity and antidiabetes functions of thermogenic brown adipocytes, e.g., increasing their number or activity to drive more energy expenditure and clearance of glucose and lipids from the circulation. To maintain this caloric drive, brown adipocytes expressing uncoupling protein 1 (UCP1) need neuronal stimulation (primarily via NE), and without these nerve signals they can revert back to a less energetically active state (“whitening”). Recent work focused on human-derived brown adipocytes as a cell-based therapy underscored the importance of peripheral nerves for metabolic health. The researchers demonstrated that human preadipocytes treated with forskolin (a sympathomimetic) increased UCP1 expression and maintained this UCP1 thermogenic activity when transplanted into the body because the stimulated cells released neurotrophic factors like BDNF to promote the cell transplant’s neural innervation (25–28). Brown adipose therapies are often critiqued because of this need for sympathetic drive, which can be detrimental if stimulated at a whole-body level (29) but targeted to a tissue transplant is likely to be less problematic, since this mimics physiological sympathetic drive to brown fat. The next nervous system challenge will be overcoming the compensatory appetite increase that can result when the body goes into a state of negative energy balance, regardless of the body’s need to dispense of excess adipose for health reasons. Again, this cross talk between the brain and peripheral tissues and organs is bidirectional: changes in the periphery can drive afferent neural activity to the brain, which can then trigger stimulation of efferent nerve signals back out to the periphery. Losing weight drives increased appetite, at any body weight, potentially via these afferent signals. This neural cross talk has also repeatedly been observed in studies with denervation of adipose depots, for example. Denervation of one depot tends to lead to compensatory changes to innervation of another depot in the body (or increased activity of remaining nerves in the denervated tissue), underscoring the redundant coordination that ensures our brain can continuously communicate with our fat storage organs through peripheral nerves, and utilize those stored lipids (14).

Taken together, the PNS is an integral, yet often overlooked, component of metabolically relevant organs including the heart, intestine, liver, adipose, pancreas, muscle, and many more (Fig. 2). While autonomic and motor nerves in the PNS are better understood, as evidenced by our deep understanding of NE-releasing sympathetic nerves in adipose tissue, we know far less about the afferent sensory nerves—those involved in interoception, somatosensation, and nociception in the body. It is laudable that the National Institutes of Health (NIH) has prioritized work in pain conditions (30) and interoception (as part of the Blueprint for Neuroscience Research [31]) to help close this gap in knowledge about peripheral nerve functions, but more still needs to be done. As a country we have devoted enormous resources to brain research (e.g., the NIH Brain Research through Advancing Innovative Neurotechnologies [BRAIN] Initiative [32]) and Alzheimer disease (at NIH alone, more than $3.7 billion in expenditures for research on this disease annually, according to the Alzheimer’s Association [www.alz.org/]). However, peripheral neuropathy is the number one manifestation of neurodegenerative disease in humans (33), and despite impacting 30 million Americans or more, work in peripheral neuropathy is relatively underfunded across all agencies and foundations. Since neuropathy overall has been understudied, we are still learning which tissues are impacted in the number one cause (diabetes/obesity) (Fig. 2)—not to mention the dozens of other neuropathy causes, including long COVID, aging, chemotherapy, and more.

When you are attuned to it, there is clear evidence underscoring the interconnectedness in our one nervous system: the brain, brainstem, spinal cord, ganglia, cranial nerves, peripheral nerves, and tissue nerve terminals that work in a single coordinated physiological system. Data on neurological diseases and disorders are revealing that those conditions formerly considered “brain only” also involve the PNS and vice versa (Table 1). Similarly, have we overlooked treatments and therapies that target the spinal cord, such as neuromodulation with spinal cord stimulation, that may also impact the brain and peripheral nerves? Or have we neglected investigating drugs for brain disorders (antidepressants, seizure medications) that may also improve PNS function and neuropathy? My guess is, yes, we have—if we are not concertedly looking for PNS effects, we certainly will not find them.

As covered in this article, mounting evidence now encourages neuroscientists to consider both arms of the nervous system in investigations of diseases and treatments. The nervous system is a single, interconnected physiological system that responds to internal and external signals to regulate numerous metabolically relevant processes in many tissues and organs that are important for obesity, diabetes, aging, and cardiometabolic health. Myriad areas of biomedical investigation remain to be fully explored related to the PNS and the interconnectedness of the nervous system in health and disease, including the nerve subtypes present in tissues/organs and their responses to local neurotrophic factors, the actions of various nerve products on different cell types in the tissue, the cross talk between arms of the PNS at the tissue or ganglia/spinal levels, the impact of peripheral neuropathy on tissues beyond the skin, and the structure/function relationships for nerve terminal structures in different tissues/organs.

Acknowledgments. The author thanks Gargi Mishra (The Ohio State University, Columbus, OH) for creating BioRender versions of the author’s schematic sketches, Gilian Gunsch (The Ohio State University, Columbus, OH) for reference assistance and edits, and Drs. Magdalena Blaszkiewicz (The Ohio State University, Columbus, OH), Erica De Sousa (The Ohio State University, Columbus, OH), and Michael Czech (UMass Chan Medical School, Worcester, MA) for a critical read of the draft manuscript. The author also thanks the Department of Neurological Surgery at The Ohio State University for their strong support of the research program of the author and colleagues, including support of the development of the Translational Research in NeuroMetabolism Symposia (TRINS), held each spring at The Ohio State University. (See ktownsendlab.com or @neuroadipo to learn more.)

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Prior Presentation. Parts of this study were presented in abstract form at the 84th Scientific Sessions of the American Diabetes Association, Orlando, FL, 21–24 June 2024. A video presentation can be found in the online version of the article at https://doi.org/10.2337/dbi24-0004.