Sleep and immune health: How dogs, goats and ‘factor S’ shaped a field

Abstract

Chronic insufficient sleep kills! Although this statement has high ‘face validity’, it is only recently that empirical evidence existed to support it. There are now sufficient data for numerous meta-analyses and systematic reviews to demonstrate that chronic insufficient sleep is associated with many inflammatory pathologies that are a public health burden. As a result, it is now well accepted that sleep is important for physical and mental health. This awareness derives from research that began in the late 19th and early 20th centuries and continues to the present day. In this narrative review we trace this rich history within the context of the research contributions of Professor James Krueger and his colleagues. The historic and current research by Professor Krueger and colleagues is fundamental to the many ongoing pre-clinical and clinical research programs focused on all aspects of sleep and immune health.

Graphical abstract

Highlights

• •Modern studies of humoral sleep regulation began in the 1960s.
• •Pioneering research in the 1970s and 1980s linked sleep and the innate immune system.
• •There is now public awareness that good sleep is essential for physical and mental health.

1. Introduction

Although there is now public awareness that good sleep is necessary for physical and mental health, empirical evidence supporting this assertion is relatively recent. It is now accepted there are bi-directional and reciprocal interactions between sleep and the immune system such that immune activation alters sleep, and insufficient sleep impairs immune function. In fact, the field has matured to the extent that reviews of various aspects of sleep – immune interactions are numerous. A PubMed search (February 2025) with keywords [sleep]+[immune function]+[review] resulted in 290 hits during the last five years alone! Given the amount of sleep – immune research being conducted in the 21st century, it is likely that many investigators are new to this field and not well-versed in the history that shaped our present-day understanding of the importance of sleep to immune health.

This narrative review recounts the history of the development of the field of sleep – immune interactions, which parallels the development of the broader field of neuroimmunology. This is not intended to be a comprehensive review of the vast sleep – immune literature or of specific mediators. Nor do we provide a detailed review of selected pathologies clearly linked to insufficient sleep and physical and mental health. For these aspects, the interested reader is referred to other comprehensive reviews, [e.g. Besedovsky et al., 2019; Imeri and Opp, 2009; Irwin and Opp, 2016; Irwin, 2019; Krueger and Opp, 2016; Zielinski and Krueger, 2011),]. Similarly, we do not attempt a comprehensive historical recounting of sleep research per se, which is fascinating and could fill volumes. Rather, we focus on a historical narrative about one aspect of sleep research, that which led to our current knowledge of sleep – immune interactions. This historical context begins about 140 years ago with pioneering studies that laid the foundation for those that followed and paved the way for the advances and knowledge that is prevalent today.

Although research in the latter part of the 19th and early 20th centuries set the stage, our current knowledge of sleep – immune interactions is firmly rooted in the many studies conducted during the past 50 years by Professor James M. Krueger. The contributions that Jim Krueger made early in his career were based on this history. As we review, Krueger's use of tools that were modern at the time he did his studies and not available earlier in the 20th century allowed Jim, his colleagues, and other investigators conducting this type of research to make discoveries that early pioneers could not. On a personal note, the authors were both postdoctoral fellows in Jim's laboratory. The training we received in his laboratory shaped our careers as scientists, and we are privileged to contribute this narrative review to a festschrift honoring his scientific achievements.

2. Historic Perspective

2.1. 19th and early 20th centuries

The ancients recognized the relationship of sleep to sickness and hypothesized that sleep onset was induced by ‘vapors’ [see (Krueger, 1989)]. The idea that sleep was induced by humoral mechanisms (vapors) persisted for millennia, and identification of ‘fatigue substances’ was the topic of research in the late 19th century [e.g. (Rosenbaum, 1892),]. However, it was the discovery of renin by Tigerstedt and Bergman in 1898 (Marks and Maxwell, 1979; Tigerstedt and Bergman, 1898) and secretin by Baylis and Starling in 1902 (Baylis and Starling, 1902) that laid the foundation for the discipline of endocrinology, from which the field of neuroimmunology would subsequently evolve.

One method by which the function for a naturally occurring process may be, at least potentially elucidated or inferred, is to restrict or deprive a subject from that process and determine consequences. Such an approach is a mainstay of sleep research, and hundreds of sleep deprivation studies are reported each year. The first formal studies of the impact of sleep loss were conducted by Maria Manàsseina (a.k.a. Marie Manacéïne) and her mentor Ivan Tarkhanov, physiologists in St. Petersburg, Russia. (Manàsseina was among the first women in Russia to graduate with a degree in medicine.) Tarkhanov had been interested in sleep problems, which led to early studies to determine effects of sleep loss. Manàsseina kept 10 puppies constantly active, and after four – five days without sleep, they died (Manàsseina, 1894). Observations during sleep deprivation and post-mortem analyses revealed that body temperatures dropped 4–6 °C even though the puppies were kept active, locomotor activity was impaired, red blood cell numbers were reduced, there were local brain hemorrhages, and cerebral ganglion were impaired. As a type of control experiment, Manàsseina also deprived puppies of food, and demonstrated that after 20–25 days they could be returned to health when food was once again provided. Based on these studies, Manàsseina concluded that sleeping is necessary for life and that death due to sleep loss is quicker than that by starvation. Manàsseina continued her sleep studies and in 1889 published the first comprehensive sleep textbook entitled Sleep as One Third of Human Life, or Physiology, Pathology, Hygiene and Psychology of Sleep, which was translated into English in 1897 (Manàsseina, 1897). As an aside, Manàsseina was (perhaps) the first scientist to state that the brain was active during sleep, a statement that contrasted with the idea that sleep was a passive process and preceded by more than 30 years the discoveries of Hans Berger and others who were developing tools to record brain activity (see later).

Manàsseina's research inspired others to conduct sleep deprivation studies. Shortly after Manàsseina published her work, Daddi (1898), Tarozzi (1899) and Agostini (1898) in Italy sleep deprived dogs, and published more detailed neuroanatomic histopathology results; collectively their results demonstrated changes in spinal ganglia and cerebral cortices due to sleep loss. Although not focused on health, Patrick and Gilbert's studies at the University of Iowa around this same time revealed that sleep deprivation impaired human performance (Patrick and Gilbert, 1896). For a more complete historical accounting of early sleep deprivation studies, we refer the interested reader to reviews by Kovalzon (2009) and Bentivoglio and Grassi-Zucconi (1997).

Although Manàsseina's research laid the groundwork, it was studies conducted more than 25 years later that are seminal to the sleep – immune field. Manàsseina did not postulate that sleep deprivation results in the release of a substance(s) associated with sleep. Rather it was a series of publications from two independent laboratories that provided empirical evidence that sleep deprivation induced the release of some unknown substance(s) that promoted subsequent sleep. In their sleep deprivation studies, Ishimori in Japan (Ishimori, 1909) and Piéron in France (Legendre and Piéron, 1910, 1913) took an approach that differed from that of Manàsseina. Both groups sleep deprived dogs, extracted cerebrospinal fluid, plasma or homogenized brain extracts, and observed effects of these substances when injected into rested, recipient dogs. Under these conditions, recipient dogs, when they received substances obtained from sleep-deprived dogs, fell into a deep, sleep-like state that resembled narcosis. When discussing results of his studies, Ishimori stated “that the cerebral matter of sleep-deprived animals contains a potent hypnogenic substance which cannot be detected in the brain of normally sleeping animals leads to clarify the cause of induction of our normal sleep … A special substance … is gradually produced [during waking] … and induces the so-called state of drowsiness, and finally causes natural sleep” (Ishimori, 1909). Piéron stated in his 1913 paper (Legendre and Piéron, 1913) that “… the accumulation of the hypnotoxin [during waking] produces an increasing need for sleep.” To our knowledge, these are the first explicit statements suggesting a biologic basis for humoral theories of sleep regulation. As we review later, there was renewed interest in the 1960s in the type of experiments conducted by Ishimori and Piéron, by which time this approach to elucidating mechanisms involved in sleep regulation was being referred to as ‘sleep factor research’.

2.2. 1920s – 1950s

Although some studies focused on humoral mechanisms (Schnedorf and Ivy, 1939), sleep research during the period of the 1920s–1950s was dominated by neuroanatomists, and neural theories of sleep regulation permeated neuroscience. A comprehensive review of neuroanatomy and neuroscience during this period is outside the scope of this current narrative, but two advances bear mentioning because of their impact on the developing field of sleep factor research. The first was the development of methods to record brain activity and the second was elucidation of two distinct brain regions involved in sleep – wake regulation.

Although Richard Caton reported in 1875 that he had used a galvanometer to observe electrical impulses on rabbit and monkey brains (Caton, 1875) it wasn't until 1924 that Hans Berger first recorded electrical activity of the human brain, although the paper was not published until 1929 (Berger, 1929). One of Berger's subjects lived with a partially exposed brain because there was a hole in his skull due to repeated surgeries to remove a tumor. This subject allowed Berger to place bare unipolar electrodes on the surface of his brain as Caton had done in his studies of rabbits and monkeys. The resulting recording was the first documented observation of human brain electrical activity, which Berger termed ‘elektrenkephalogramm’, and the EEG was born.

Following Berger's 1929 report, there were many studies of human brain electrical activity using EEG as a tool. During the 1930s there were rapid advances in describing changes in brain potentials during sleep. Notable works by Loomis, Harvey, Hobart, Blake, Gerard, Kleitman (Davis et al., 1938; Loomis et al., 1935a, 1935b, 1937, 1938) and others clearly demonstrated different patterns of brain electrical activity through the course of the night. Although all the hallmarks were there in these publications of the 1930s, it was more than 15 years later that Aserinsky and Kleitman (1953) would describe eye movements during sleep and the presence of a distinct sleep phase was reported. In this seminal report, Aserinsky and Kleitman recorded autonomic changes associated with periods of eye movements, which led them to conclude that they “… are very likely all manifestations of a particular level of cortical activity which is encountered normally during sleep” (Aserinsky and Kleitman, 1953). These observations in some ways recapitulated observations and statements made by Manàsseina more than 60 years before.

The realization that distinct brain regions are involved in the regulation of sleep and wakefulness resulted from a global pandemic of an atypical form of encephalitis that came to be referred to as encephalitis lethargica. This encephalitis was first described in 1917 by the neurologist Constantin von Economo (von Economo, 1917) and the pathologist René Cruchet (Cruchet et al., 1917). It is estimated that during the encephalitis lethargica pandemic (1915–1927) more than one million people were infected and that more than 500,000 deaths were directly attributed to the disease (McCall et al., 2008). Von Economo described three forms of the disease, each with a distinct pathology and symptom manifestation. Von Economo first published his findings in 1917 (von Economo, 1917) and later in a much larger monograph (von Economo, 1929). Important to sleep research, von Economo correlated symptoms with the location of lesions in brains of patients who had died from the disease. Von Economo noted that lesions in the anterior hypothalamus were associated with insomnia, whereas those in the posterior hypothalamus were associated with somnolence. Based on these observations, he proposed that the anterior hypothalamus contained the ‘sleep center’ and the posterior hypothalamus a ‘wake center’ (von Economo, 1930). As we review later, knowledge of discrete brain regions involved in the regulation of sleep and wakefulness was critical to determining mechanisms of action and neuroanatomic substrates mediating sleep – immune interactions. For a detailed review of the encephalitis lethargica pandemic see (Hoffman and Vilensky, 2017).

Finally, we highlight one additional research focus during this period that is not directly sleep-related but is important to neuroimmunology. In 1927, the Nobel Prize in Physiology or Medicine was awarded to Julius Wagner-Jauregg for “his discovery of the therapeutic value of malaria inoculation in the treatment of dementia paralytica” (Nobel Prize Outreach, 2025). Dementia paralytica, a.k.a. general paralysis of the insane, was a fatal neural disorder caused by syphilis. Wagner-Jauregg, a psychiatrist and neurologist, observed that some patients with nervous disorders improved after febrile illnesses. His early systematic attempts to treat dementia paralytica with fever, which were not successful, used tuberculin inoculation. Wagner-Jauregg postulated that the fevers induced by tuberculin were not of large enough magnitude and decided to use malaria as an infectious agent instead. His first attempts using malaria inoculation to induce fever were successful as six of nine dementia paralytica patients in this small study showed remarkable improvement (Wagner-Jauregg and Bruetsch, 1946). Wagner-Jauregg and others conducted many more studies, and until the advent of antibiotics, ‘fever therapy’ was the treatment of choice for dementia paralytica. Although not directly sleep – related, the studies of Wagner-Jauregg demonstrated links between the immune system and mental health, the fundamental basis of neuroimmunology. The story of ‘fever therapy’ is a fascinating one, and the interested reader is referred to several reviews of this topic (Howes et al., 2009; Zuschlag et al., 2016). As an aside, von Economo was an assistant to Wagner-Jauregg when he presented his case studies of encephalitis lethargica to the Viennese Society of Psychiatry and Neurology in 1917.

2.3. 1960s – 1980s

The decades of the 1960s and 1970s saw a resurgence in research focused on humoral regulation of sleep. The major goal of sleep factor research was to identify, characterize, and eventually determine the structures and mechanisms of action of endogenous substances involved in sleep regulation. Multiple approaches were used to identify sleep factors, including total sleep deprivation, selective rapid-eye-movement (REM) sleep deprivation, and inducing sleep by electrical stimulation in brain.

In the early 1960s several investigators began lines of research aimed at identifying sleep factors. Kornmüller (Kornmüller et al., 1961) in Göttingen, Germany, and Monnier (Monnier and Hoesli, 1964) in Basel, Switzerland used electrical stimulation of the thalamus to induce sleep in experimental cats and rabbits, respectively, after which they transfused blood dialysates to rested recipient animals, which then slept more. Monnier (Monnier and Hoesli, 1964) reported that dialysate from sleeping rabbits differed in action from that obtained from control animals after sham stimulation of the thalamus, perhaps the first sleep factor study with an appropriate control. In addition, Monnier demonstrated in a contrasting experiment that injection of dialysate from alert donor rabbits induced arousal in recipient animals. Monnier published a series of papers in the 1960s and 1970s, and his line of research ultimately resulted in the identification and characterization of the first endogenous sleep factor, delta sleep-inducing peptide (DSIP), so named because it increased amplitudes of slow waves (delta frequency band) in the EEG (Schoenenberger et al., 1977). As briefly mentioned in the previous section, the ability to record brain activity was a major advance and observations that experimental manipulation changes not only its behavioral manifestations, but the electrical properties of the brain during sleep are now fundamental to sleep research. By the late 1970s, the amino acid sequence of DSIP was known (Schoenenberger et al., 1978), a synthetic peptide with sleep-inducing properties had been produced (Monnier, 1977), which was transported through the blood-brain barrier (Monnier et al., 1977b) and its effects on sleep of rabbits and rats had been established (Kafi et al., 1979; Monnier et al., 1977a). Since DSIP was obtained from peripheral blood samples, the demonstration that it was transported into brain was important. For historical context of DSIP transport into brain, see Banks, this issue (Banks, 2025) and papers by Banks and Kastin (Banks et al., 1982; Kastin et al., 1982). Others contributing to sleep factor research during this period included Uchizono and Inoué in Japan, who demonstrated that a sleep-promoting substance existed in brains of sleep-deprived rats (Nagasaki et al., 1974); Ringle and Herndon in the United States, although their study failed to detect a sleep-promoting effect in rats of cerebrospinal fluid obtained from sleep-deprived rabbits (Ringle and Herndon, 1969); and Borbély in Switzerland, who demonstrated that factors in cerebrospinal fluid of rats affected locomotor activity (Sachs et al., 1976).

Although multiple groups were by this time elucidating endogenous sleep-promoting substances, it was the sustained research program of John Pappenheimer that was most relevant to sleep and immunity. A respiratory physiologist at Harvard, Pappenheimer demonstrated that cerebrospinal fluid from sleep-deprived goats reduced locomotor activity and increased sleep of rats (Pappenheimer et al., 1967). Within a few years, this unknown substance was referred to as ‘Factor S’, for sleep (Fencl et al., 1971). As with others, Pappenheimer was to pursue this line of research for many years because the process of identifying and characterizing endogenous sleep-promoting substances was time consuming and labor intensive (see Box 1).

Box 1. Jim Krueger's Recollections of the Isolation and Characterization of Factor S.

How I Got Started in Sleep Research

After completing my PhD at the University of Pennsylvania/Monell Chemical Senses Center I applied for a position in the Harvard Medical School to join the John Pappenheimer/Manfred Karnovsky laboratories. The interview went fine, although I had to admit I knew almost nothing about the science of sleep. I was well received by John and Manfred, likely because the Penn faculty, many respiratory physiologists, gave strong recommendations to John, and Manfred was my academic grandfather (my mentor at Penn was Robert Cagan who had been Manfred's graduate student in biochemistry at Harvard Medical School). It was simultaneously an odd interview because neither John nor Manfred were ‘sleep researchers’, although they had worked for several years on the characterization of Factor S.

After the interview I went back to Philadelphia and in short order was sent an offer to join John and Manfred with the goal of isolating and characterizing Factor S. They specified the goal because it was clear to them that I thought I was interviewing for a position dealing either with control of respiration (John's area) or macrophage function (Manfred's focus). A few months later my wife and I were in Boston, and I started my sleep research work.

Scaling Up and Structural Characterization

For several years before I arrived at Harvard in 1974, John and Manfred had via ultra filtration and gel filtration techniques established that the sleep promoting material in cerebral spinal fluid (CSF) of sleep-deprived goats was of low molecular weight, likely less than 5000 Da, a peptide (determined using enzymatic digestion techniques), and promoted inactivity and sleep in rats and rabbits. Thus, when I arrived standard EEG sleep recordings were in place, and rats had implanted EEG electrodes plus an intracerebroventricular guide cannula for administering samples of partially purified Factor S. They had also come to the end of their sampling from goats as they had concluded it was too labor intensive, expensive, and calculations suggested they could never collect enough CSF to isolate then chemically characterize Factor S. So, I was dealt the most difficult project which would occupy the next six years of my life. After a few months, during which I learned techniques and much about sleep, John announced that he was going to Oxford for a sabbatical and I was left mostly on my own, except that I also had space in Manfred's lab and a group of young scientists there to learn from.

From a theoretical perspective, the search for Factor S was modeled on successful searches for brain hormones in the 1960s and 1970s. There was the assumption that there would be a ‘sleep hormone’ with much specificity of action for sleep promotion, that it was synthesized and released during waking, and acted in an endocrine- or paracrine-like manner. Lucky for me, one of the scientists leading the searches for brain hormones, Susan Leeman, was in the Harvard Medical School Physiology department. She had already isolated two brain hormones (substance P, neurotensin) and was most helpful to me and the project I was working on. Eventually, it was her amino acid analysis equipment that allowed us to establish the amino acid/amino sugar sequence of Factor S. She also helped me in my thinking and adaptation of brain extraction and isolation techniques.

By the time John had returned from his sabbatical we had made good progress in using brain, instead of CSF, as the starting material from which to isolate Factor S. While I had already concluded that we would have to scale up and extract huge amounts of brain, or urine because it contains most small molecules found in the body, it would take me a while to convince John of this. Initially, we would take brains from sleep deprived animals, extract and purify Factor S using various chromatographic approaches, all at a relatively small scale, such as in university labs. In the end we never got very far because we ran out of chemically detectable Factor S, although we could detect purified samples by using sleep assays. I needed more starting material. Eventually, we scaled up to very large amounts of starting materials, e.g. brains from 5000 sleep deprived rabbits each batch, or 1500 L batches of human urine. [Author's note: for a more detailed description of the approach used to collect urine, see (Krueger et al., 1982b)]. To do this we were helped by Stirling Winthrop, a pharmaceutical company across the river from Albany NY. They had a pilot plant for scaling up manufacturing of new drugs, which they allowed me to use (in their presence, of course). As one example of the large-scale production, a single G10 column had $20,000 worth of Sephadex in it [Author's note: almost $70,000 in today's dollars]!

Purification, Sample Loss, and Increased Specific Activity

We developed an eight-step purification procedure to isolate Factor S from both urine and brain [see (Krueger et al., 1982b)]. After step two, samples were clean enough to inject into rabbit or rat brains so we could establish a sleep dose, which we expressed as the amount of product per mg dry weight needed to induce about a 10 % increase in sleep for about 6 h. So, after step two, we called that amount a dose of “1”. After the eighth step, the dose-specific activity was 71,000 doses per mg dry weight! When looking at the test tube there was nothing to see but Factor S was there. Later we characterized it by amino acid analyses and mass spectrometry as a muramyl peptide. We did the amino acid analyses on both the product from urine and separately on the brain product with similar results. Brain and urine samples both contained muramic acid and diaminopimelic acid, which were bacterial cell wall components. Although we were very successful in increasing the specific activity, which allowed chemical characterization, the losses during purification were great; we only recovered about 3 % of the active Factor S we started with (determined from the sleep bioassay). The effective sleep dose of Factor S turned out to be about 5–10 pmol injected into either rabbits or rats; it is a very effective material.

Scope of Work to the First Successful Characterization of Factor S

The purification protocol for Factor S was developed over a period of years using CSF, brain, and urine as starting materials. We tried multiple chromatographic systems. For example, high pressure liquid chromatography and high voltage paper electrophoresis were used in pilot studies on material purified through step four of our protocol. However, these methods were found to be unsatisfactory. I emphasize that a single biological assay for sleep required many hours of recording on each of at least two rabbits. In total, about 10,000 h of EEG recordings associated with about 1000 assays on more than 200 rabbits with chronically implanted EEG electrodes and cerebroventricular guide cannulae over a period of more than 5 years were carried out before reaching our goal of chemical characterization of Factor S.

It is safe to say, no sane scientist would repeat this in today's world. But remember, in those days, mammalian genomes were not sequenced, and nucleic acid and protein chemistry had not reached today's standards and sensitivity. Nevertheless, our work was on par with much of the endocrinology success stories developing in the 1960's, 70's and 80's.

Anecdotes: or Life is Great from Time to Time in the World of Biomedical Research

Goats: By the time I joined John's and Manfred's labs, they were down to one goat. While the goat was very friendly and because a group of people who thought it needed company and exercised him almost daily in the labs and halls of HMS Building C, he was no longer used for CSF donations. We (John) found a home for him in the Army Research Labs near Boston where for a few years we received periodic reports of his good health. I remember the goat clearly because during John's yearlong sabbatical at Oxford University we got the goat to sit in John's office chair with our lab members surrounding the goat/chair and took a picture. We sent the picture to John as a lab season's greeting card. John had a good sense of humor, saw ‘the old goat’ in his chair and got the picture.

Rabbits: Soon after we made the decision to scale up and use many rabbit brains for our starting material from which to isolate Factor S, we had a lab visitor who introduced himself as the son of the owner of the only rabbit abattoir in the USA. He was taking a summer course at MIT and had read about our work in John's article in Scientific American [Author's note: Volume 235, August 1976, pp 24–29]. He indicated two things; first his company could quickly isolate rabbit brains and rapidly freeze them and second, he said that one of their rabbit suppliers, who was in Texas, every week delivered rabbits to them after an overnight ride in an open flatbed truck (a way to keep temperatures low to prevent heat-stressing the rabbits). He noticed that as soon as the truck parked the rabbits were all asleep within minutes, likely because they stayed awake all night due to the vibration of the truck. (Rabbits are very sensitive to ground vibrations as a way to detect predators.) Regardless, we ordered, what was in our mind at the time a very large order, maybe 20–30 brains from truck ride-induced sleep deprived rabbits for our initial studies.

It worked, so within the year we started scaling up, but by that time the Texas farm had stopped delivering rabbits to the abattoir and the rabbit production was moved to Rogers, Arkansas. So, to sleep deprive the rabbits, the abattoir used their flatbed truck and had one of their employees drive the truck loaded with rabbits all night in and out of Rogers. After several passes in and out of town, the police got suspicious of such activity and pulled over the driver, rabbits and all, and started quizzing him on his very suspicious behavior! As told to me, he told the policeman “You won't believe it, but these crazy Harvard professors want very tired rabbits before they are killed for their studies”. Lucky for the driver, Rogers is a small place, he established his Rogers' bonafide's (good ol’ boy status) and they all had a good laugh and story to tell. We got the brains, they worked, and the method was repeated several times over the next year. The rest of the rabbits were sold as meat, so their deaths were not in vain for science. In the end, we used about 15,000 rabbits that were sleep deprived by truck!

Alt-text: Box 1

As a postdoctoral fellow in Pappenheimer's laboratory (Fig. 1), Jim Krueger collaborated with biochemist Manfred Karnovsky to extract Factor S from brains of sleep-deprived animals (Pappenheimer et al., 1975), purify it (Krueger et al., 1978), characterize it (Krueger et al., 1980, 1982a), and identify it as a muramyl peptide (Krueger et al., 1982a; Martin et al., 1984). Because Krueger had isolated it from human urine and from cerebral spinal fluid and brains of experimental animals, the identification of Factor S as a muramyl peptide was startling because there is no mammalian synthetic pathway for it, i.e., mammals cannot produce muramyl peptide. Muramyl peptide is a component of peptidoglycan, the structural building block of bacterial cell walls. The only source for muramyl peptide in animal brain or human urine is bacterial cell wall peptidoglycan! When bacteria are exposed to the immune system, enzymes break down peptidoglycan, releasing muramyl peptide. Some immune cell types have receptors that recognize muramyl peptide, and when binding and signal transduction occurs, an immune response is triggered. With the demonstration that muramyl peptide and a synthetic muramyl dipeptide induced sleep in rats, rabbits and cats (Krueger et al., 1982a), a direct connection between one endogenous sleep factor and the immune system had been made.

Fig. 1.

Left to right: John Pappenheimer, Manfred Karnovsky, Jim Krueger.

The unexpected finding that Factor S in brain was a bacterial product was more than just startling, it was met with general disbelief, resulting in the often-stated question “What the hell is a bacterial product doing in brain?” and the follow-up statement “There is no way a bacterial product can be naturally occurring in brain. You must have contaminated your samples”. In general, the idea that a bacterial cell wall component could somehow be involved in the regulation of behavior as complex as sleep was anathema to sleep researchers and neuroscientists at the time. Now of course, it is well accepted that the gut (bacterial) microbiome plays a role in sleep, circadian rhythms, and many other behavioral and physiological processes [see for example (Cryan and Clarke, 2016; Krueger and Opp, 2016; Withrow et al., 2021)], but in those early days statements that bacterial products were important drivers of physiological behavior in the absence of pathology were viewed as tantamount to heresy.

Despite skepticism, sleep research on muramyl peptide accelerated because immunologists at this time were using them, and their synthetic analogues, as immune adjuvants. It was already known that muramyl peptide induced the synthesis/secretion of lymphocyte activating factor, also known as endogenous pyrogen and now called interleukin-1β (hereafter, IL-1). Fontana in Switzerland demonstrated that IL-1 was constitutively expressed in brain astrocytes (Fontana et al., 1982), and because muramyl peptide induced sleep and the synthesis/secretion of IL-1, Krueger wanted to know if it could also induce sleep. IL-1 effects on sleep were first reported as an abstract and poster presentation at the 1983 annual meeting of the Federation of American Societies for Experimental Biology (FASEB) (Krueger et al., 1983) and subsequently in two full-length papers (Krueger et al., 1984; Tobler et al., 1984). Collectively, these papers demonstrated that IL-1 administered centrally into rabbits and rats increased slow-wave sleep (SWS), suppressed rapid-eye movement (REM) sleep, and increased the amplitude of the EEG in the low frequency (delta) band during SWS. Readers interested in additional information about sleep factor research from the period of the 1960s to the mid-1980s are referred to reviews by Krueger (Krueger and Majde, 1995; Krueger, 2023).

By the early 1980s, it had become much more prevalent, and many groups were focused on sleep factor research. The search for endogenous sleep factors played a central role in sleep research during this period as the mechanisms and substrates by which sleep and wakefulness were regulated were not well understood. In 1980, Alex Borbély and Irene Tobler published a short review of the state-of-the-art of sleep factor research at that time (Borbély and Tobler, 1980) and suggested that criteria were necessary by which to define a sleep factor. Six requirements were proposed that should be met for a substance to be considered a sleep factor. To paraphrase from this publication (Borbély and Tobler, 1980), to be considered a sleep factor: 1) the substance should induce and/or maintain sleep; 2) there should be dose-related effects on sleep; 3) its action should be similar across species with respect to time course and effects on sleep; 4) the substance should be present in the organism; 5) changes in the substance should be associated with changes in sleep; and 6) the substance should be chemically identified. These proposed criteria were refined during the next decade, but fundamentally remained as originally stated [see later and (Borbély and Tobler, 1989)].

Beginning in the early 1980s, symposia on the topic of sleep factors were proposed and accepted, and oral presentations were included as part of the scientific programs at conferences. Interest had grown such that dedicated workshops on endogenous sleep promoting substances were organized. The first workshop of which we are aware that was specifically focused on sleep factors was organized by Borbély (Switzerland) and Reidun Ursin (Norway) in conjunction with the 6th European Congress of Sleep Research held in Zurich in 1982. Reflecting the growing international interest in sleep factor research, presentations at this workshop were given by Krueger (USA); Hondo and Komoda (Japan); Cespuglio and Jouvet (France); Scherschlicht (Switzerland); Kovalzon (Russia); Ursin (Norway); Graf, Christen, Tobler, Baumann and Schoenenberger (Switzerland); Shiyi (China); Schneider-Helmert (Switzerland); Mendelson (USA); and Pavel (Romania).

Two years later (1984), the Taniguchi Foundation, Division of Brain Sciences sponsored a week-long symposium in Kyoto, Japan entitled ‘Humoral Control of Sleep and Its Evolution’. The Taniguchi Foundation, Division of Brain Sciences was chaired by Osamu Hayaishi, at the time a professor at Osaka Medical College, whereas the program committee for the symposium was chaired by Shojiro Inoué of the Tokyo Medical and Dental University. Although the symposium also included presentations on other aspects of sleep and circadian rhythms, most of the scientific program was devoted to endogenous sleep substances. Attendees included Inoué, Borbély, Horne (UK), Liu (China), Tobler (Switzerland), Mukhametov (Russia), Honma (Japan), Honda (Japan), Nakagawa (Japan), Kastin (USA), Schoenenberger (Switzerland), Krueger, Oomura (Japan), Adrien (France), Dugovic (Belgium), Yamamoto (Japan) and others. A book of proceedings from the symposium, edited by Inoué and Borbély, was published the following year (Inoué and Borbély, 1985).

A larger workshop organized by Krueger and Inoué was held in Honolulu, Hawaii in 1988. This workshop was funded by the National Science Foundation of the United States under the auspices of the U.S. – Japan Joint Seminar program. This U.S. – Japan Joint Seminar on Endogenous Sleep Factors was attended by approximately 40 investigators (Fig. 2), notably including Krueger, Inoue, Borbély, Karnovsky, Pappenheimer, Uchizono, Hayaishi, Kluger (USA), Kilduff (USA), Kimura (Japan, Krueger postdoctoral fellow at the time), Dement (USA), Turek (USA), Blatteis (USA), Stenberg (Finland), Honda (Japan), Danguir (France), Oomura (Japan), Van Cauter (USA), Obál (Hungary, but a visiting professor in Krueger's laboratory at the time), Johanssen (Germany, Krueger postdoctoral fellow at the time), Mignot (USA), E. Smith (USA), Onoe (Japan), Cespuglio, and others.

Fig. 2.

Attendees at the US – Japan joint seminar on endogenous sleep factors. November 1988, Honolulu, HI.

By the late 1980s additional volumes and monographs focused on sleep factors had been published [see (Borbély and Tobler, 1989; Inoue, 1989; Inoué and Borbély, 1985)]. As previously mentioned, with the increasing number of research groups working on sleep factors, and the identification of many endogenous substances potentially involved in sleep regulation, there was further refinement to criteria by which a substance could be considered a ‘true’ sleep factor. Although the criteria proposed earlier by Borbély and Tobler (1980) formed the basis, by the time of their comprehensive review less than 10 years later (Borbély and Tobler, 1989), they had been distilled to two fundamental criteria, each with multiple components: 1) the sleep promoted by the substance must be physiologic, and 2) changes in the endogenous substance should be associated with changes in sleep of the organism. These criteria helped to further focus the field, which was necessary because sleep factor research had exploded, and more and more substances were demonstrated to be involved in sleep regulation. As an example of the growth in sleep factor research during this decade, in their 1980 review, Borbély and Tobler listed only four ‘sleep factors’: DSIP, arginine vasotocin (AVT), Factor S, and an unidentified substance that Koji Uchizono's group in Japan was working to identify and characterize. In their publication nine years later (Borbély and Tobler, 1989), they listed more than 20 substances as sleep factors, or potential sleep factors, and Inoue in his book The Biology of Sleep Factors (Inoue, 1989) listed approximately 30 putative sleep factors. Substances on these lists included ions, neurotransmitters, neuropeptides, gut peptides, hormones, hypothalamic and pituitary releasing and inhibiting factors, and immunomodulators, including bacterial products. Few of these substances had been characterized well enough to meet all proposed sleep factor criteria. Several substances fulfilled the criterion of inducing physiologic sleep, and some observations indicated there were diurnal differences in responsiveness to sleep factors [see (Inoue et al., 1984)]. Although none had been demonstrated, the fluctuation in sleep factors with natural sleep was anticipated as by this time it was well known that transitions from wakefulness to sleep and from sleep to wakefulness involved changes in most, if not all, physiological measures, ranging from hormones to blood gasses. However, observation that endogenous substances did indeed fluctuate in relationship to sleep – wake states was on the horizon.

2.4. mid-1980s – early 2000s

The period of the early 1960s to the mid-1980s saw sleep factor research evolve from a few studies by a handful of laboratories to become mainstream sleep research. By this time, dozens of laboratories around the world were engaged in sleep factor research and the list of purported endogenous sleep-inducing substances had increased dramatically. For the remainder of this narrative review we focus specifically on the immunomodulators demonstrated to modulate and/or regulate sleep, and primarily on the work of Krueger and his colleagues. Even though ‘endogenous pyrogen’ (a.k.a. IL-1) was described in brain, ‘critics’ remained who refused to believe that molecules of the peripheral immune system were present in normal brain. However, by this time enough groups in the sleep and neuroimmunology fields were reporting findings of immunomodulators in brain that these observations could no longer be dismissed as ‘artifacts of contaminated samples’.

The authors joined the Krueger laboratory as postdoctoral fellows during this period of expanding focus on immunomodulators and sleep. Opp (Krueger laboratory from 1987 to 1993; Fig. 3) first met Jim when, as a graduate student he attended the 1986 European Sleep Research Society meeting held in Szeged, Hungary. Many sleep researchers who would influence Mark's career were at that meeting, including Ferenc Obál, Jr., Irene Tobler, Alex Borbély and others who would become colleagues and friends during the next four decades. But it was Jim Krueger who had the greatest impact on Mark's career, in no small part because he offered him a postdoctoral position. This was somewhat surprising given that the behavioral ecology field research conducted with Charles Amlaner, Nigel Ball, and Don Miller as part of his PhD in zoology were far removed, at least in Mark's mind, from laboratory-based studies of neuroscience and the immune system. Nevertheless, this training period in the Krueger laboratory formed the basis of what was to become a friendship and productive collaboration for the next 40 years.

Fig. 3.

Members of the Krueger laboratory, Department of Physiology, University of Tennessee Memphis (1992). The laboratory was multi-national with members from the United States, Hungary, Italy, Germany, Japan, China and Russia. Not pictured: Luca Imeri, who had returned to Italy shortly before this picture was taken. Back row, left to right: Mark Opp; Levente Kapas; Jim Krueger; Edward Fincher, 4th; Jidong Fang; Sebastian Bredow Middle row, left to right: Lielie Hong; Linda Toth; Sandi Johnson; Linda Payne; Ferenc Obál, Jr. Front row, left to right (seated): Sema Geller; Gail Richmond; Mayumi Kimura-Takeuchi.

After reading some of Krueger's reviews, particularly a lay piece published in The Scientist (Krueger, 1989), Imeri became intrigued by the ’story’ of the relationship between sleep and the immune system. Luca thought the study of sleep – immune interactions could reveal something about the mind – body problem, because, as he naïvely thought “mind is different from brain, but there is no mind without brain”. Although Alan Hobson infamously wrote that “sleep is of the brain, by the brain, for the brain” (Hobson, 2005), sleep is also for the body, and what is more ‘bodily’ than an immune reaction? Luca's choice to join Jim's laboratory as a postdoctoral fellow was not without some controversy. Luca's mentor, Mauro Mancia, the famous Italian psychoanalyst and neurophysiologist at the University of Milan who was among the early founders of sleep science, wanted to send him to Mircea Steriade's laboratory at the University of Laval in Quebec, Canada. It was Maria Grazia De Simoni, his supervisor and mentor during his first postdoctoral fellowship at the Mario Negri Institute in Milan, who convinced Mancia that Luca should further his training in Krueger's laboratory, an intervention for which he will always be grateful. Luca was a postdoctoral fellow in the Krueger laboratory from 1991 to 1992, after which he returned to a faculty appointment at the University of Milan (Fig. 4).

Fig. 4.

Maria Grazia de Simoni, Mario Negri Institute for Pharmacological Research in Milan, Italy and Luca Imeri, Institute of Human Physiology II, University of Milan, Italy, 1992.

The overlapping period that Opp and Imeri spent in Krueger's laboratory was productive. As a result, they established a long-standing collaboration that resulted in multiple NIH awards, numerous publications, and many reciprocal trips between the US and Italy as part of their joint research efforts.

During the period of the mid-1980s to 2000, research in the Krueger laboratory focused on two parallel areas of investigation: determining effects on sleep of infection with replicating pathogens and elucidating the role of immunomodulators and their interactions as part of sleep-regulatory systems. We first briefly review research from the Krueger laboratory and others that demonstrated the impact of infections on sleep before shifting focus for the remainder of this narrative review to the role of immunomodulators in sleep regulation.

2.4.1. Infection-induced alterations in sleep

Although recognized for millennia, it wasn't until the late 1980s that systematic studies demonstrated the extent to which sleep is altered during infection. This early work on infection and sleep was pioneered by Linda Toth and Jim Krueger. Research on infection and sleep at this time was an important parallel to other projects on immunomodulators in the Krueger laboratory (see next section) because replicating pathogens are a ‘real world’ immune challenge with clinical consequences. Importantly, the question of whether sleep facilitates recovery from sickness could not be addressed until the precise nature, and mechanisms mediating infection-induced alterations in sleep were known.

Toth, DVM, PhD, was a postdoctoral fellow in the Department of Comparative Medicine at the University of Tennessee College of Medicine in Memphis (now part of the University of Tennessee Health Science Center) when she began collaborating with Krueger. Toth's graduate training was in pharmacology, but her clinical duties were in the Laboratory Animal Care Unit, and as a veterinarian interested in infectious disease her research interests were a good fit with Krueger's. Toth's first study of infection and sleep was published in 1988 (Toth and Krueger, 1988) and demonstrated that sleep of rabbits inoculated with Staphylococcus aureus, a Gram-positive bacterium that can cause serious, even fatal infections in humans, was dramatically altered. Inoculation with S. aureus resulted in a rapid increase in SWS, in SWS bout length, and in the amplitude of EEG delta waves during SWS. Subsequently, these sleep parameters were reduced to values that were less than corresponding control values. That is, there was a bi-phasic response in which SWS was initially increased and subsequently suppressed. After about 6–10 h post-inoculation REM sleep was suppressed and remained so for the duration of the recording period. Linda extended these studies in 1989 when she published a comparative paper that reported results from three groups of rabbits, each inoculated with a different pathogen: Streptococcus pyogenes (Gram-positive bacteria), Escherichia coli (Gram-negative bacteria), and Candida albicans (Gram-positive fungus) (Toth and Krueger, 1989). Although the exact time course and magnitude of responses differed with pathogen, changes in sleep followed a pattern like that previously reported in rabbits inoculated with S. aureus; SWS sleep duration, bout length and EEG delta waves during SWS initially increased and were subsequently suppressed, while REM sleep was suppressed. In 1990, Toth and Krueger published yet another paper demonstrating infection-induced alterations in sleep of rabbits, this time using Pastorella multocida, a Gram-negative bacterium that is a natural pathogen of rabbits. As with previous studies, a bi-phasic sleep response ensued with an initial increase and subsequent suppression of SWS parameters, and REM sleep suppression (Toth and Krueger, 1990). These three papers were important because they demonstrated a generalized sleep response of rabbits to infection with five different pathogens, including one that is common to the species.

We stated previously that the question of whether sleep during infectious disease facilitates recovery could not be addressed until the nature of infection-induced alterations in sleep was better understood. In 1993, Toth and Krueger published an important retrospective study that did not answer this question directly, but did support the hypothesis that good sleep facilitates recovery from infectious disease. Linda used data from her previous studies of rabbits inoculated with E. coli, S. aureus, and C. albicans to develop a sleep quality score based on the duration of SWS and the increase in EEG delta power during SWS. Stratification of data revealed that rabbits with enhanced sleep quality scores (i.e., good sleep) survived, whereas those that had reduced sleep quality scores (i.e., poor sleep) were more likely to die (Toth et al., 1993).

Bacteria are not the only infectious pathogens, of course, and Krueger and Toth conducted a series of experiments to determine the impact of viral products and viral infections on sleep. Krueger's early studies used polyriboinosinic:polyribocytidylic acid (poly I:C) as an immune stimulant. Poly I:C is a synthetic double-stranded RNA (dsRNA) analog that mimics viral dsRNA, and when administered ICV into rabbits it increases SWS, EEG delta power during SWS, and suppresses REM sleep (Krueger et al., 1988). In addition, Krueger showed that total RNA extracted from lung tissue of mice infected with influenza A/Puerto Rico/8/1934 (H1N1) (PR8), an attenuated, mouse-adapted strain that had been isolated from a human in Puerto Rico, also increases SWS, EEG delta power during SWS, and suppresses REM sleep when administered into rabbits (Majde et al., 1991). Krueger continued studies on various aspects of viral components as stimulators of innate immune responses and their impact on sleep of rabbits (Kimura-Takeuchi et al., 1992a, 1992b; Kimura et al., 1994). In the meantime, Toth began studying effects of viral infection on sleep of mice, mostly using the PR8 mouse-adapted influenza strain (Toth, 1994; Toth et al., 1995; Toth and Hughes, 2004), and continued this research focus for some time (Jhaveri et al., 2007; Trammell and Toth, 2008, 2014, 2015). We also demonstrated that sleep of mice infected with murine gammaherpes virus 68, a model of chronic viral-induced fatigue (Olivadoti et al., 2007), is altered (Olivadoti et al., 2011). Collectively, these studies demonstrated that, as with bacterial infections, sleep is altered through the course of viral infections. For more detailed reviews of infection-induced alterations in sleep see (Krueger et al., 1994; Krueger and Opp, 2016; Opp and Toth, 2003; Toth, 1995; Zielinski et al., 2025)

At about the same time as Krueger, Toth and colleagues were publishing their studies, Susan Norman at Mount Sinai Medical Center in Miami Beach, Florida reported that sleep was altered in individuals infected with human immunodeficiency virus (HIV) who were seropositive, but otherwise healthy and before they were symptomatic with AIDS (Norman et al., 1988, 1990, 1992). These studies revealed increases in SWS and a particular change in the timing of sleep stages; there was more SWS during the second half of the night than during normal sleep of healthy subjects. We subsequently reported that HIV envelope glycoproteins and co-receptors alter sleep of rats (Gemma and Opp, 1999; Gemma et al., 2000; Hogan et al., 2001; Opp et al., 1996), suggesting mechanisms by which sleep of patients is altered by the virus before their immune systems are compromised and AIDS develops.

During the intervening years since the seminal papers of Toth and Krueger, many pre-clinical, laboratory-based clinical and prospective clinical studies have been published demonstrating infection-induced alterations in sleep in multiple model systems. Although differences exist depending on host species, pathogen and many other factors, one common observation is that infectious agents alter sleep. Although mechanisms focused on the relationship between sleep, thermoregulation and febrile responses to infection have been proposed as a functional basis for survival (Imeri and Opp, 2009), the question of the extent to which good sleep facilitates recovery from infectious disease remains to be fully investigated and is a challenge for the field.

2.4.2. Immunomodulators and their interactions

While the studies on infection-induced alterations in sleep continued, much of the research effort in the Krueger laboratory during this period focused on the role of immunomodulators in sleep regulation. As previously reviewed, by the mid-to late 1980s many groups were actively working on sleep factors, and many endogenous substances had been identified that seemed to play some role in regulating or modulating sleep. For the remainder of this review we focus primarily on one class of immunomodulators that were the core of research efforts in the Krueger laboratory, cytokines.

Cytokines are low molecular weight proteins that play a critical role in signaling between and among cell types in the peripheral immune system. We now know that they, and their receptors, are present in normal (i.e., non-pathologic) brain, where they function in many physiological and behavioral processes, including sleep, thermoregulation, appetite and feeding, sexual behavior, social exploration, mood and others. At the time Krueger started his studies on cytokines and sleep, the only other physiological process in which a cytokine had been implicated was fever generated by endogenous pyrogen (IL-1).

After the initial publications of IL-1 effects on sleep in 1984 (Krueger et al., 1984; Tobler et al., 1984), research on cytokines progressed rapidly. In 1986, Harvey Moldofsky at the University of Toronto demonstrated that IL-1 concentrations in human plasma peak at sleep onset (Moldofsky et al., 1986) and a few years later he reported that concentrations in cat cerebrospinal fluid vary in phase with sleep-wake behavior (Lue et al., 1988). Other papers of the impact of IL-1 on sleep of laboratory animals followed in rapid succession, including those that demonstrated dose-related responses and importantly, time of day effects (Imeri et al., 2004; Opp and Krueger, 1988, 1991; Opp et al., 1991; Opp and Imeri, 2001). Krueger published the first paper describing effects of tumor necrosis factor-α (TNF) on sleep in 1987 (Shoham et al., 1987) and later demonstrated that IL-1 and TNF mRNA and protein exhibit diurnal variation in brain (Bredow et al., 1997; Floyd and Krueger, 1997; Taishi et al., 1997, 1998). One final note on TNF, in 1999 Moldofsky demonstrated that TNF administered ICV into sheep increases SWS, the first demonstration that a cytokine alters sleep in that species (Dickstein et al., 1999). In addition, perhaps the most important aspect of this study was the use of radio-labeled TNF, which was detected in cervical lymphatics within 30-min after ICV administration! The demonstration that CNS TNF directly drains into the lymphatic system provided evidence that a central cytokine could alter sleep and the peripheral immune system by this mechanism. We now know that there are multiple mechanisms mediating bi-directional communication between the CNS and the peripheral immune system, but this seminal paper provided context by which brain-to-immune communication could occur.

One of the limitations for studies of cytokines and sleep in those early days was the lack of specific reagents. Without knowing the structure and/or sequence of a cytokine, species-specific reagents could not be developed. For example, human IL-1 was cloned and sequenced before mouse or rat IL-1, and as such for many years human recombinant IL-1 was administered into laboratory rabbits and rodents. Lack of reagents was problematic in other ways as well. For most early studies, antibodies for rat, mouse, or rabbit cytokines were not available, which meant it was difficult to assay them in samples from experimental animals. When antibodies against these cytokines did become available, they were often not very specific, and we had to state that it was ‘IL-1-like activity’ or ‘TNF-like activity’ we were detecting in samples. Finally, and perhaps most importantly, without specific antagonists it was difficult to implement experiments in which endogenous cytokines were targeted. Although papers were frequently published demonstrating the effects on sleep of administering cytokines, few papers at the time were published that attempted to inhibit actions of endogenous cytokines. This was a critical limitation early on because administering a cytokine initiates a cascade of responses, any of which could be the ‘proximal inducer’ of sleep changes. Whether a particular cytokine was a ‘regulator’ of sleep, rather than just a ‘modulator’ of sleep could be demonstrated if a specific antagonist altered sleep. As one example, IL-6 is considered a modulator of sleep because under conditions in which it is elevated NREM sleep increases, but antagonizing it with antibodies does not alter sleep, and IL-6 knockout mice sleep normally (Hogan et al., 2003; Morrow and Opp, 2005a, 2005b). In contrast, IL-1 and TNF are cytokines considered to be regulators of sleep because specifically antagonizing them with antibodies, receptor antagonists, synthesis inhibitors, or by genetic ablation reduces sleep, and as we review later, they act directly on sleep- and wake-related neurons in brain regions and nuclei known to be involved in sleep regulation. For more comprehensive reviews of cytokines and distinctions between their roles as sleep regulators or sleep modulators, see (Krueger, 2008; Opp, 2005).

Despite these limitations, research in the Krueger laboratory on cytokines and sleep accelerated. Following publication of the impact of TNF on sleep in 1987, many additional studies on this cytokine followed [reviewed (Jewett and Krueger, 2012; Krueger, 2008; Krueger et al., 2011; Opp, 2005)]. At this point there was much interest in structure/function relationships, i.e., specific molecular sequences or minimal components of the molecule capable of altering sleep. Papers demonstrating that fragments of IL-1 and TNF and their receptors also alter sleep followed [e.g. (Kapás et al., 1992; Kushikata et al., 1998; Obál et al., 1990; Takahashi et al., 1996)]. As species-specific recombinant proteins and better reagents became more readily available, many experiments demonstrated that targeting IL-1 or TNF reduced spontaneous, physiological sleep and that increases in sleep after sleep deprivation or other manipulations (ambient warming, high fat diets, infection) could be attenuated or blocked by pre-treatment with specific cytokine antagonists [reviewed (Jewett and Krueger, 2012; Opp, 2005)].

It became apparent early on that cytokines function as components of complex networks of interactions in the peripheral immune system and in brain. Immunologists had described two general classes of cytokines as pro-inflammatory (such as IL-1 and TNF) and anti-inflammatory, endogenous cytokines that either directly or indirectly antagonize actions of pro-inflammatory cytokines. Examples of anti-inflammatory cytokines are IL-4 and IL-10, and there are others. It followed that if receptor antagonists and antibodies directed against IL-1 or TNF reduced sleep, then a negative-feedback (anti-inflammatory) cytokine such as IL-4 or IL-10 should as well. Studies demonstrated that this was indeed the case, and now it is possible to generalize that pro-inflammatory cytokines, or stimuli that induce their synthesis/secretion, increase NREM sleep whereas anti-inflammatory cytokines, or cytokine antagonists, reduce it [reviewed (Jewett and Krueger, 2012; Kapsimalis et al., 2007; Opp, 2005)].

The complex networks by which cytokines regulate sleep are not limited to interactions between/among components of the immune system. By the late 1990s – early 2000s it was clear that mechanisms by which IL-1 and TNF regulate physiological NREM sleep included, in part, interactions with many other well-defined neuromodulatory systems. For example, IL-1 stimulates the synthesis and/or release of growth hormone-releasing hormone, prostaglandin D2, adenosine and nitric oxide. Each of these substances is implicated in regulating or modulating NREM sleep, and antagonizing these systems attenuates or blocks IL-1-induced increases in NREM sleep. By the late 1990s, these interactions had been well-described (Fig. 5), but mechanisms by which these humoral networks regulated sleep were not well-understood. The interested reader is referred to (Krueger et al., 2001; Obal and Krueger, 2003) for state-of-the-art reviews of the role of neuromodulatory systems in mediating cytokine effects on sleep as understood at the time.

Fig. 5.

Sleep regulatory system [as published in 2001]. Extensive evidence exists showing that the sub-stances illustrated are involved in sleep regulation (see text). These substances affect the pro-duction or action of one another. Collectively, these actions form the biochemical cascade illustrated. Arrows (→) indicate activation or enhanced production; (–) indicates an inhibitory effect. Substances in boxes inhibit NREMS, while other substances enhance NREMS. Many of those substances that promote sleep are growth factors (e.g., IL1, TNF, BDNF, NGF, IL-2, NO, PGs, adenosine, NFκB) and are upregulated by neuronal activity (e.g., IL1, TNF, BDNF, NGF, adenosine, NO). We have hypothesized that sleep mechanisms involve these neuron-activity-dependent substances and that sleep function is related to synaptic efficacy (see text). Abbreviations: LPs, lipopolysaccharides; MPs, muramyl peptides; dsRNA, double-stranded RNA; IL1RA, interleukin-1 receptor antagonist; sIL1R, soluble IL1 receptor; anti-IL1, anti-IL1 antibodies; CRH, corticotropin-releasing hormone; PGE 2, prostaglandin E 2 ; αMSH, αmelanocyte-stimulating hormone; IL1, interleukin-1; TNF, tumor necrosis factor; sTNFR, soluble TNF receptor; anti-TNF, anti-TNF antibodies; BDNF, brain-derived neurotrophic factor; NGF, nerve growth factor; NFκB, nuclear factor kappa B; IL4, interleukin-4; IL10, inter-leukin-10; IL13, interleukin-13; TGFβ, transforming growth factor β; L-NAME, an arginine analog; NOS, nitric oxide synthase; COX-2, cyclooxygenase-2; IL2, interleukin-2; IGF-1,insulin-like growth factor 1; anti-GHRH, anti-GHRH antibodies; GHRH, growth hormone releasing hormone; NO, nitric oxide; PGD2, prostaglandin D 2; A1R, the adenosine A1 receptor; GABA, γ-aminobutyric acid; glu, glutamic acid. Reprinted with permission from (Krueger et al., 2001).

Studies of the role of cytokines in sleep regulation were not limited to pre-clinical research using animal models. David Dinges, Bill Shearer, Janet Mullington, Hans van Dongen and colleagues at the University of Pennsylvania demonstrated that sleep deprivation alters immune parameters in human volunteers (Dinges et al., 1994), and that total sleep deprivation or sleep restriction increase circulating soluble TNF receptor 1 and IL-6 (Shearer et al., 2001). Also notable during this period were seminal studies by Alex Vgontzas, Ed Bixler and their colleagues at Pennsylvania State University. Among their studies, Vgontzas and Bixler demonstrated that excessive daytime sleepiness in patients suffering from narcolepsy and sleep apnea is associated with elevated TNF and IL-6 (Vgontzas et al., 1997); that secretory patterns of circulating IL-6 are altered in human volunteers by sleep deprivation, which coincide with post-deprivation daytime sleepiness and fatigue (Vgontzas et al., 1999); and that insomnia is associated with a shift in the normal secretory patterns of IL-6 and TNF from nighttime to daytime (Vgontzas et al., 2002). Two additional studies by this group bear mentioning. In a 2004 publication, Vgontzas and Bixler reported that restricting sleep by only 2 h per night (from 8 h to 6 h) for one week induced daytime sleepiness, impaired performance and increased circulating IL-6 and TNF (Vgontzas et al., 2004a). Interestingly, sex differences were revealed in this study as IL-6 and TNF are elevated in male subjects, but in female subjects only IL-6 is elevated. Finally and importantly, Vgontzas and Bixler reported that an intervention using a TNF antagonist, etanercept, decreases sleepiness in sleep apnea patients (Vgontzas et al., 2004b), demonstrating that targeting cytokines represents a therapeutic strategy for alleviating daytime sleepiness associated with some sleep disorders. There were many other studies of sleep – immune function in human volunteers during this period, including those of Michael Irwin at UCLA (Irwin et al., 2006)] and Ken Wright at the University of Colorado Boulder, whose studies focused on circadian misalignment in addition to sleep deprivation (Frey et al., 2007; Wright et al., 2015). Interested readers are referred to (Besedovsky et al., 2019) for a more comprehensive review.

2.5. 2000 – 2010

As briefly reviewed in the previous section, by the early 2000s much was known about cytokines and the interactions between/among them and other transmitters, peptides and hormones implicated in sleep regulation. What was not known at this time were the neuroanatomic substrates and neurochemical cell types on which cytokines acted to modulate/regulate sleep. This was an important knowledge gap because up to this point immunomodulators had been administered either intracerebroventricularly, intravenously or intraperitoneally. If a substance of interest was administered and sleep was altered, the tacit assumption was that it had reached an effective site of action. However, no systematic research had been conducted to determine where brain cytokines acted to modulate/regulate sleep.

Historic studies by neuroanatomists in the 1940s and 50s, including those of Hess, Bremer, Moruzzi, Magoun, and Nauta provided a foundation for later research by Jouvet, Pompeiano, Steriade, Sterman, Chase, McCarley, Saper, Jones, Baghdoyan, Luppi and others to determine regions of brain and the neural circuitry involved in sleep regulation. Although many details remained unknown, by the early 2000s fundamental brain regions and nuclei had been well-documented as being involved in the regulation of NREM sleep, REM sleep and wakefulness. A comprehensive review of the neuroanatomy of sleep – wake regulation is beyond the scope of this paper and the interested reader is referred to many excellent reviews (Gompf and Anaclet, 2020; Jones, 2005; Saper et al., 2001; Scammell et al., 2017; Schwartz and Kilduff, 2015).

As detailed in the previous section, IL-1 administered to laboratory animals consistently increases NREM sleep and suppresses REM sleep [see (Imeri and Opp, 2009; Opp, 2005)]. By the late 1990s, Imeri and Opp had started a research program that aimed to address the knowledge gap of where in brain and on what neurochemical cell types IL-1 acted to regulate/modulate sleep. For several reasons, we began our studies by focusing on interactions between IL-1 and the serotonergic system. First, there was a previous literature demonstrating functional interactions between muramyl peptides, which as previously reviewed are bacterial products/immune adjuvants that alter sleep and induce IL-1, and the serotonergic system. Karnovsky had demonstrated binding sites for muramyl peptide in brain and that a glial cell line released IL-1 in response to muramyl peptide or serotonin (5-HT) (Silverman et al., 1989). He had also reported that 5-HT mediated actions of muramyl peptides in regulating macrophage function (Polanski et al., 1995). Karnovsky's work linked interactions between a bacterial product (muramyl peptide) and a neurotransmitter (5-HT) that had been extensively studied and implicated in sleep regulation [reviewed (Adrien, 2002; Imeri and Opp, 2009; Jouvet, 1999; Monti, 2011)].

Our early studies demonstrated there were interactions between the IL-1 and serotonergic systems with functional consequences for sleep. For example, by the late 1990s, in part by collaborating with the Maria Grazia De Simoni at the Mario Negri Institute for Pharmacological Research in Milan, Italy, we had demonstrated that IL-1 enhances 5-HT release in the hypothalamic medial preoptic area and that this increase correlates with the initial phase of NREM sleep enhancement that it induces (Gemma et al., 1997). In addition, our early studies revealed that the initial increase in NREM sleep induced by IL-1 (and MDP) is abolished when brain 5-HT is depleted (Imeri et al., 1997), and that serotonin may partially mediate IL-1 effects on NREM sleep via 5-HT₂ receptors (Imeri et al., 1999). However, the question of the neuroanatomic substrate(s) where these interactions occurred remained.

Evidence suggested that IL-1-induced increases in NREM sleep were mediated, in part, by opposing yet complementary actions on the serotonergic cell bodies and axon terminals in two distinct neuroanatomical regions, namely the brainstem dorsal raphe nucleus (DRN) and the preoptic area of the hypothalamus and the adjacent basal forebrain (POA/BF). Most serotonergic cell bodies are in the DRN, and the POA/BF, a brain region crucial for NREM sleep regulation (Jones, 2005; Saper et al., 2001), receives serotonergic innervation from the DRN. We demonstrated that microinjection of IL-1 directly into the DRN increases NREM sleep and that in a slice preparation IL-1 inhibits firing rates of physiologically and pharmacologically identified serotonergic neurons (Manfridi et al., 2003) by enhancing γ-aminobutyric acid (GABA)-induced inhibitory postsynaptic potentials (Brambilla et al., 2007; Manfridi et al., 2003). Shintani had shown that IL-1 also increases 5-HT release from axon terminals in the POA/BF (Shintani et al., 1993), and we demonstrated that IL-1 exerts this action without altering state-dependent 5-HT release. That is, the phasic increases in 5-HT release during wakefulness and the decreases during sleep are superimposed on the tonic IL-1-induced 5-HT increase (Gemma et al., 1997). Importantly, Shintani also demonstrated that 5-HT release from axon terminals in the POA/BF can result from local actions that are independent of the serotonergic cell bodies in the DRN because its release is elicited by application of IL-1 directly into the hypothalamus (Shintani et al., 1993). Not only does IL-1 induce 5-HT release in the hypothalamus but conversely increasing activation of the serotonergic system using 5-hydroxytryptophan (5-HTP), the direct 5-HT precursor, upregulates IL-1 mRNA in the hypothalamus (Gemma et al., 2003) during periods when NREM sleep is enhanced by the same treatment (Imeri et al., 2000).

Although Krueger had microinjected TNF into the preoptic area (Kubota et al., 2002), no studies had been conducted to determine effects on sleep of IL-1 in this brain region. As mentioned, the POA/BF is a primary NREM sleep regulatory center that receives innervation from the DRN. UCLA professors Dennis McGinty and Ron Szymusiak had extensively studied this brain region and the cell types it contains [see (McGinty and Szymusiak, 2001, 2003; Szymusiak and McGinty, 2008) for reviews], and the four of us (Opp, Imeri, McGinty, and Szymusiak) met during the 3^rd Congress of the World Sleep Federation over drinks in the Kempinski Hotel, Dresden, Germany (1999) to discuss a joint project aimed at elucidating a role for the POA/BF in mediating IL-1 effects on NREM sleep. As a result of that meeting, a grant application was written and funded, and we began the project.

It was known that the ventrolateral preoptic area (VLPO) and median preoptic nucleus (MnPN) contained large numbers of neurons that expressed c-Fos, a not-so-perfect indicator of neuronal activity, after sleep but not after waking (Gong et al., 2004; Sherin et al., 1996). In these early studies, we hypothesized that IL-1 increases NREM sleep by actions on sleep-active sites in the VLPO and MnPN. Experiments in this brain region revealed that ICV administration of IL-1 into rats increases the number of c-Fos reactive neurons in the MnPN, but not VLPO, which correlate with the amount NREM sleep (Baker et al., 2005). Dennis and Ron had also demonstrated that in the POA/BF there were sleep-active, wake-active, and state-indifferent neurons (Suntsova et al., 2002; Szymusiak et al., 1998). These designations were given based on the arousal state in which firing rates were the highest. For example, sleep-active neurons discharge at a higher frequency during NREM sleep than during wakefulness. They had perfected a technique by which substances could be administered through a microdialysis probe positioned adjacent to microwires that recorded unit activity within the area being perfused. Using this approach in a different set of studies, we hypothesized that IL-1 increases NREM sleep by increasing discharge rates of sleep-active neurons in the POA/BF. As is often the case, the data did not support our hypothesis. Instead, this study demonstrated that IL-1 perfused into the POA/BF of freely behaving rats primarily reduces firing rates of wake-active putative cholinergic neurons rather than by increasing discharge rates of sleep-active putative GABA-ergic neurons (Alam et al., 2004).

Collectively, results of studies we conducted on the DRN and the POA/BF demonstrated that IL-1-induced increases in NREM sleep are mediated, in part, by inhibitory actions on wake-promoting DRN neurons and inhibition of wake-active neurons in the POA/BF. These observations suggest a potential resolution to what for a long time had been, and in part still is (Oikonomou et al., 2019), a paradox in the field, i.e. whether 5-HT is a sleep- or wake-promoting transmitter. Although a comprehensive discussion of the role of the serotonergic system in sleep regulation is beyond the scope of this review [see (Adrien, 2002; Jones, 2005; Monti, 2011)], this paradox exists because data from many studies support both roles for this transmitter; 5-HT is maximally released during wakefulness, but induces subsequent sleep [for a review see (Imeri and Opp, 2009)]. To resolve this paradox, Jouvet proposed (Jouvet, 1999) that 5-HT promotes wakefulness through direct actions (i.e. the transmitter is wake-promoting) and also stimulates the synthesis and/or release of sleep promoting factors. As we reviewed earlier, results of our studies suggest that IL-1 is one of the sleep promoting factors through which 5-HT induces the physiological sleep that follows wakefulness. Results from our studies demonstrate that the opposing actions of 5-HT (wake-promoting, sleep-stimulating) are only an ‘apparent paradox’ as these effects are temporally related, i.e. they simply follow each other; 5-HT first directly promotes wakefulness by actions in the DRN and later, induces the synthesis and release of IL-1 in the POA/BF to promote NREM sleep (Imeri et al., 2000).

Finally, because IL-1 suppresses REM sleep, we conducted a study in which we targeted the cholinergic laterodorsal tegmental nucleus (LDT) in the brainstem. The LDT is part of the cholinergic circuitry that generates REM sleep (Baghdoyan, 2002; Jones, 2005; McCarley, 2007; Saper et al., 2001). When IL-1 is microinjected into the LDT, REM sleep is suppressed, and data from slice electrophysiology demonstrate that IL-1 inhibits firing of identified cholinergic neurons (Brambilla et al., 2010).

3. Current perspective on chronic insufficient sleep and public health

The last 15 years is characterized by continued, and in some ways explosive growth of the field of sleep – immune interactions, much of which is attributed to new tools and methodologies. For example, developing genetically-manipulated mice, once an expensive and complex process [e.g. (Ingiosi et al., 2015; Ingiosi and Opp, 2016)], is now common using gene editing tools and there are many commercially-available strains in which multiple components of cytokine systems have been targeted. A search on the Jax® Mice website using [cytokine] as the keyword results in more than 80 readily-available mouse strains (i.e. live) and another ∼200 cryopreserved strains. Not only is it much easier to obtain genetically-modified mice, but new technologies to manipulate systems using genetic engineering, such as opto- and chemo-genetics now make it possible to selectively activate or inhibit specific cell types in discrete brain regions/nuclei [see (Deisseroth, 2015; Rajasethupathy et al., 2016)]. Furthermore, using two-photon calcium imaging, it is possible to view cellular activity in vivo from freely-behaving mice in real time [see (Ma et al., 2024; Mizuta and Sato, 2024; Niethard et al., 2021)].

Awareness of the public health impact of chronic insufficient sleep has motivated pre-clinical research focused on sleep disruption rather than total sleep loss and numerous studies demonstrate the impact of short- and long-term sleep disruption on inflammatory processes that contribute to pathology [e.g. (Carreras et al., 2014; Poroyko et al., 2016; Ramesh et al., 2012; Ringgold et al., 2013; Trammell et al., 2014; Trammell and Toth, 2015; Wang et al., 2014)]. Conceptually, the sleep – immune field has moved beyond a neuron-centric view and much of current pre-clinical research focuses on the role of non-neuronal cell types (i.e. microglia, astrocytes) in health and disease (Halassa et al., 2009; Rowe et al., 2022; Wisor and Clegern, 2011; Wisor et al., 2011a, 2011b). As a consequence, gliotransmission, communication between/among neurons, astrocytes, microglia and oligodendrocytes, is now recognized as critical to physiological sleep regulation and a contributor to sleep disturbances during pathology. Experimental studies of gliotransmission are fundamentally neuroinflammation research, in which cytokines are critical mediators [see (Green et al., 2020; Ingiosi and Frank, 2023; Ingiosi et al., 2023)].

We stated at the outset of this narrative review that there is now public awareness that good sleep is necessary for physical and mental health. The extensive body of work we reviewed, produced during approximately the last 140 years, has resulted in the biomedical research community, from bench to bed side, understanding that strong bi-directional and reciprocal interactions between sleep and the immune system exist. This awareness has manifest in sleep – immune studies focused on many different public health issues. Some representative examples of meta-analyses and systematic reviews that demonstrate links between chronic insufficient sleep and pathology include, but certainly are not limited to obesity (Bacaro et al., 2020), diabetes (Gangwisch et al., 2007), breast cancer (Kakizaki et al., 2008b), prostate cancer (Kakizaki et al., 2008a); stroke (Wang et al., 2022), pain (Irwin et al., 2023) and importantly, cardiovascular disease (Cappuccio et al., 2011; Meisinger et al., 2007). Each of these pathologies is an inflammatory disease characterized by altered cytokine and immunomodulator profiles associated with poor sleep. In addition to these specific pathologies, chronic insufficient sleep is associated with all-cause mortality and overall health outcomes [e.g. (Garcia-Perdomo et al., 2019; Itani et al., 2017)], suggesting its impact on physical health is ‘global’ rather than restricted to one or a few systems or pathologies.

We briefly highlight two of the many public health problems associated with chronic insufficient sleep, cardiovascular disease and vaccine effectiveness. Among the many, we selected these two examples because cardiovascular disease is the number one cause of death in the United States and Europe (Ahmad and Didia, 2020; Townsend et al., 2022) and there is currently much public health concern with vaccine effectiveness (Kennedy, 2020; Siani, 2023).

The relationship between chronic insufficient sleep and cardiovascular disease is well established [see (Cappuccio et al., 2011; Meisinger et al., 2007)]. In fact, the American Heart Association now includes sleep in its Life's Essential 8. Although the negative consequences of chronic insufficient sleep on cardiovascular health are well known, new data suggest that good sleep after a cardiac event facilitates recovery. A recent study demonstrates in mice that after myocardial infarct monocytes are recruited to brain, where they release TNF. TNF subsequently acts on sleep-regulatory neurons to promote NREM/SWS, which reduces sympathetic outflow and inflammation and contributes to healing (Huynh et al., 2024; Rowe, 2024). Conversely, disrupting sleep of mice after myocardial infarct increases susceptibility to secondary cardiac events and impairs the heart's functional recovery (Huynh et al., 2024). The same authors prospectively analyzed outcomes of patients with acute coronary syndrome during a two-year follow up; the incidence of major adverse cardiovascular events was two-fold higher in poor sleepers as compared to good sleepers (Huynh et al., 2024). Although many questions remain (Rowe, 2024), this study suggests a bi-directional relationship between sleep and myocardial infarct such that good sleep facilitates, whereas poor sleep impedes recovery.

Multiple components of the innate immune system are frequently outcome measures of sleep disruption studies. Although fewer studies focus on sleep loss and the adaptive immune response, evidence suggests these interactions are important within the context of vaccine effectiveness. Vaccines are the antigenic challenge that stimulates the adaptive immune response, i.e. antibody production that is necessary for protection against disease. A recent meta-analysis (Spiegel et al., 2023) reports that antibody responses to vaccination are affected by sleep history. This analysis reveals that short sleep duration around the time of vaccination is associated with impaired antibody production and contributes to ‘non-responders’, i.e., individuals who are vaccinated yet not protected from the disease for which they got the shot (Opp, 2023). Although this is an important analysis, questions remain about the role of chronic insufficient sleep in vaccine effectiveness. For example, subjects in most studies of sleep and vaccine effectiveness are young, healthy volunteers. Does the conclusion that sleep restriction impairs immune responses to vaccination hold true in the (heterogenous) general population, characterized by co-morbidities and comprising, by definition, different age groups? For example, Faioni et al. (2024), found no differences in antibody responses to an mRNA SARS-CoV-2 vaccination between shift workers and non-shift workers, suggesting that disrupted sleep and daily schedules associate with shift work did not impact vaccine effectiveness. Finally, although we are beginning to understand the effects of sleep on responses to vaccination, we still do not know if vaccination, which is an antigenic (immune) challenge affects sleep and EEG activity. Since sleep and immune function are altered in different psychiatric disorders, for instance major depressive disorders (Irwin and Miller, 2007), is there any difference in response to vaccination between subjects suffering from this condition and control subjects? Although much remains to be learned about mechanisms, extant data strongly suggest that chronic insufficient sleep around the time of vaccination impacts antibody responses and likely contributes to ‘non-responders’ and reduces efficacy.

As reviewed, extensive data accumulated during the past 50 years link chronic insufficient sleep with disruption of the innate immune system and inflammation, which are associated with many chronic diseases. These studies raise important questions that remain unanswered. If chronic insufficient sleep induces inflammation, which in turn drives pathology, is it better to target inflammation as an intervention or to improve sleep? The obvious answer is ‘both’, but there are barriers to each approach. Cardiovascular disease serves as an example of the barriers to overcome when translating findings of sleep – immune interactions into actionable interventions for the benefit of public health. Lifestyle changes are among the most effective interventions because they not only contribute to prevention but also are effective in managing and treating cardiovascular disease after it develops. Incorporating lifestyle choices known to be effective seems simple: eat a heart-healthy diet; get regular exercise; quit smoking; manage stress; manage weight; and get good sleep. Although apparently simple, implementing these lifestyle choices to prevent or manage cardiovascular disease is challenging for many reasons, and the American Heart Association estimates that only 5 % of individuals follow practices that would result in achieving ‘ideal’ cardiovascular health (Rippe, 2019). Similarly, pharmacological interventions that target inflammation exist and are effective. In addition to lowering cholesterol, statins are anti-inflammatory; they reduce/modulate production of IL-1, IL-6, TNF and cyclooxygenase-2 (COX-2) among others [reviewed (Koushki et al., 2021; Maierean et al., 2022)]. Although effective, statins are not without risk; most common side effects are not life-threatening, but some are serious, such as liver and kidney damage, rhabdomyolisis, neuropathy, and increased risk of type 2 diabetes (McPherson et al., 2024). Within the context of this narrative review, some statins may disrupt sleep (Golomb et al., 2004; Szmyd et al., 2021), which potentially exacerbates the inflammatory condition under treatment. Finally, in addition to the ‘biologic’ factors, there are also systemic barriers to cardiovascular health, including poor access to healthcare and policy gaps. In the current climate, these systemic barriers are likely to worsen, much to the detriment to public health.

4. Conclusions

Although the ancients from at least the time of Hippocrates and Aristotle recognized that sleep is essential for health, empirical evidence supporting this assertion is relatively recent. In this narrative review we have summarized the development of the field of sleep – immune interactions during approximately the past 140 years, from the earliest formal studies of sleep deprivation to present day when it is accepted that good sleep is essential for physical and mental health. In keeping with the spirit of a festschrift, we use a passage from the 16th century English poet, scholar, soldier and cleric John Donne to underscore a point. Meditation XVII of his prose work “Devotions Upon Emergent Occasions” (1624), contains Donne's famous quote “no man is an island, entire of itself; everyman is a piece of the continent, a part of the main”. Of interest within the context of this narrative, the subtitle of Donne's Devotions is “and several steps in my sickness” because he penned them as he was recovering from spotted fever (caused by Rickettsia bacteria).

The interconnectedness of which Donne writes is a theme in the history of science that is not often considered, at least explicitly. Using our own experiences as an example, the author's careers would have been very different had they not been postdoctoral fellows in Krueger's laboratory. Furthermore, had they not overlapped during their time in Memphis, it is not likely Opp and Imeri would have had the successful collaboration that developed out of that experience, indicating that timing is indeed everything. Similarly, Krueger's career and contributions to the field would have differed dramatically had he not been offered a postdoctoral position in Pappenheimer's laboratory, or if he had been directed to study respiratory physiology. Of course, one cannot state there would be no public health awareness of the relationship of good sleep to immune health had Krueger done a postdoctoral fellowship elsewhere or was not involved in studies aimed at elucidating properties of Factor S, but our current knowledge would likely be much different and would have developed from a different trajectory. This interconnectedness is true for all of us in the field within the context of our respective contributions. Professionally we are each the product of our scientific forebearers, just as Ishimori and Piéron reexamined postulates stated millennia before, and Kornmüller, Monnier, Uchizono and Pappenheimer extended their studies, and so on. Within the context of this current narrative, the interconnectedness of the aforementioned individuals with many others (neuroanatomists, physiologists, physicians, etc.), resulted in a tremendous body of evidence that, if acted upon, would save lives. Given this knowledge base, it is quite simple, “sleep better and live longer and healthier”. Why is it so difficult to translate such a fundamental statement into actionable interventions to save lives? This is a problem that current and future generations of sleep researchers, physicians, and multidisciplinary implementation science teams will have to solve. They now have a good scientific basis from which to do so.

CRediT authorship contribution statement

Mark R. Opp: Writing – review & editing, Writing – original draft, Conceptualization. Luca Imeri: Writing – review & editing, Writing – original draft, Conceptualization.

Declaration of competing interest

The authors declare they have no competing interests.