The sleep-deprived blood-brain barrier: Does the Brain's gatekeeper adapt or merely fail slowly?

Abstract

Chronic insufficient sleep is highly prevalent and is linked to an elevated risk of adverse health outcomes. Animal studies show that chronic sleep restriction, inducing a cumulative sleep debt, promotes blood-brain barrier (BBB) dysfunction, neuroinflammation, and can insidiously lead to neuropathology. Available studies primarily show diurnal or time-of-day variation in BBB permeability rather than direct manipulation of sleep state. Therefore, our fundamental understanding of BBB dynamics during the sleep-wake cycle remains limited. Investigating dynamic changes in the BBB induced by prolonged wakefulness is critical for understanding the heightened vulnerability of the brain to neurodegenerative diseases. In this review, we first discuss the potential role of sleep to modulate BBB dynamics. We then examine the known mechanisms driving BBB changes during sleep loss across different experimental paradigms and critically discuss the bidirectional interplay between sleep and the BBB, emphasizing how disruption of one system affects the other. This narrative-based review challenges the prevailing view of the BBB as a static, whole-brain barrier, proposing instead that it is a dynamic and adaptive interface whose regulation is tightly coupled to neuronal demands across the sleep-wake cycle.

Highlights

• •Sleep loss disrupts blood-brain barrier integrity and increases permeability.
• •Blood-brain barrier function varies dynamically across the sleep–wake cycle.
• •P-glycoprotein-mediated efflux is enhanced during sleep, reduced in wake.
• •Adenosine and tumor necrosis factor-α link sleep pressure to blood brain barrier dysfunction.
• •Blood-brain barrier adaptations to sleep loss are protective but can become maladaptive.

Abbreviation	Definition	Abbreviation	Definition	Abbreviation	Definition
Aβ	Amyloid Beta	CSF	Cerebrospinal fluid	NVU	Neurovascular unit
AD	Alzheimer's disease	EC(s)	Endothelial cell(s)	NREM	Non-rapid eye movement
ADORA2A	Adenosine receptor 2A	GLUT1	Glucose Transporter 1	Ocldn	Occludin
ATP	Adenosine triphosphate	IL-1β	Interleukin 1 beta	Pgp	P-glycoprotein
AQP4	Aquaporin-4	IL-6	Interleukin 6	PD	Parkinson's disease
BBB	Blood–brain barrier	IL-17	Interleukin 17	TJ(s)	Tight junction(s)
Cldn-1	Claudin-1	LAT1	L-type amino acid transporter 1	TNF-α	Tumor necrosis factor alpha
Cldn-5	Claudin-5	LRP1	Low-density lipoprotein receptor-related protein 1	ZO-1	Zonula occludens-1
CNS	Central nervous system	MCT1	Monocarboxylate transporter 1
CRP	C-reactive protein	MDR1	Multidrug resistance protein 1

1. Introduction

Chronic sleep loss is increasingly prevalent in modern societies characterized by around-the-clock operations, extended work hours, and insufficient opportunities for recovery sleep. Sleep loss, defined as insufficient or disrupted sleep, can arise from real-world circumstances (e.g., behavioral, environmental, clinical), and it is studied by experimental manipulations, including sleep restriction, sleep deprivation, or sleep fragmentation (see Table 1). Notably, even one night of sleep loss triggers a cascade of molecular, cellular, and systemic responses aimed at maintaining homeostasis in the face of inadequate restoration (He et al., 2014). In this way, the brain engages several mechanisms, including changes in neuronal activity within regions critical for maintaining cognitive and behavioral performance (Elmenhorst et al., 2017; Boonstra et al., 2007; Weiss et al., 2021), as well as broader cellular responses involved in homeostatic regulation (Williams et al., 2020; Zhang et al., 2025). As the blood-brain barrier (BBB) regulates nutrient transport, protection, and the clearance of metabolic waste products from the central nervous system (CNS) to the circulation (Pardridge et al., 1994; Abbott et al., 2010), understanding the BBB dynamics during sleep and sleep loss is critical for understanding the complex molecular signaling during prolonged wakefulness that can modulate the brain microenvironment. Under pathophysiological challenges, the BBB transiently responds to local and circulating signals (Segarra et al., 2021; Friedman et al., 2025) to support repair processes that promote allostasis. This is modeled by the increase in BBB permeability to immune cells during neuroinflammatory events (Galea, 2021), or the response of capillary vasodilatation and the enhanced function of glucose transport during heightened neuronal activity (Koepsell, 2020).

Table 1.

Definitions of key sleep and sleep-loss terminology used in this review.

Term	Definition
Sleep-wake cycle	The temporal organization and alternation of sleep and wakefulness across a 24-h period, regulated by the interaction of the circadian and homeostatic systems.
Circadian system	An endogenous timekeeping network organized by the suprachiasmatic nucleus that generates and synchronizes approximately 24-h rhythms in physiology and behavior. It coordinates peripheral biological clocks throughout the body and aligns internal biological rhythms with environmental time cues (zeitgebers), particularly the light–dark cycle.
Homeostatic system	An endogenous regulatory mechanism that tracks prior wakefulness and sleep, governing the accumulation and dissipation of sleep pressure to regulate sleep propensity and sleep need.
Sleep loss	An umbrella term referring to any reduction in the amount or quality of sleep obtained relative to an individual's physiological sleep need. In research, sleep loss results from experimental sleep restriction, sleep deprivation, sleep fragmentation, or other manipulations that reduce total sleep obtained.
Sleep restriction	Sleep opportunity or duration is deliberately or externally limited, resulting in less sleep than is physiologically needed. Experimental sleep restriction protocols typically involve limiting sleep opportunities (e.g., 4–6 h of sleep) while still allowing some sleep to occur, thereby producing cumulative sleep debt.
Sleep deprivation	The state resulting from insufficient sleep due to sleep restriction, prolonged wakefulness, sleep fragmentation, or other disruptions to normal sleep. It serves as an umbrella term that includes total sleep deprivation, partial sleep deprivation, and sleep restriction.
Total sleep deprivation	A specific form of acute sleep deprivation in which a human or model organism remains awake throughout an entire sleep period, obtaining no sleep during a defined deprivation interval, resulting in sustained wakefulness across one or more habitual sleep periods. In human experimental paradigms, total sleep deprivation typically involves 24-48 h of sustained wakefulness, although longer durations have also been studied.
Human models of sleep loss:	Acute: Refers to a short-term period of sleep loss occurring over hours to days. In experimental settings, acute sleep loss may range from a single night of restricted sleep (e.g., <5 h of sleep during one sleep opportunity) to continuous total sleep deprivation lasting multiple nights.
	Chronic: Refers to a prolonged period of insufficient sleep occurring over weeks to months. Chronic sleep loss typically involves habitual sleep durations less than the recommended 7–9 h per night on most nights (e.g., at least 3 nights per week) for a period of 3 months or longer.
Rodent models of sleep loss:	Acute: Refers to a continuous period of enforced wakefulness or restricted sleep lasting anywhere from 3 to 24 h.
	Chronic: Refers to a period of consistent sleep restriction that generally spans 3 to 21 days, with short sleep opportunities of 3-6 h.

Growing evidence suggests that a bidirectional relationship between the BBB and the sleep-wake cycle exists (He et al., 2014; Elmenhorst et al., 2017; Zhang et al., 2025). Based on the increase in BBB permeability to exogenous tracers during sleep loss (Hurtado-Alvarado et al., 2017; Avilez-Avilez et al., 2024, 2025), it appears that the balance between production and metabolism of molecules that accumulate during prolonged wakefulness (e.g., adenosine and proinflammatory cytokines) (Pardridge et al., 1994; Porkka-Heiskanen et al., 1997) may have a temporal effect on BBB dynamics concomitant to sleep pressure. Herein, we first introduce key properties of the BBB and the cellular mechanisms of BBB permeability associated with the sleep-wake state. Next, we highlight specific physiological mechanisms that constitute the bidirectional relationship between the BBB and sleep. Finally, we identify specific indicators of BBB dysfunction linked to sleep loss, emphasizing changes in a primary BBB efflux system, and discuss the adaptive nature of these changes.

2. Beyond the restrictive blood-brain barrier: A complex and heterogenous system

Beyond its traditional role as a selective barrier, the mammalian BBB exhibits complex and dynamic functions that support adaptive responses to physiological variations, including aging, hormonal changes, circadian input (Galea, 2021), etc. The BBB is formed by specialized endothelial cells (ECs) lining the brain microvasculature and is supported by pericytes, astrocytic endfeet, the basal lamina, and the endothelial glycocalyx (i.e., an extracellular matrix layer coating the luminal surface of vascular ECs) (Abbott et al., 2010). The BBB, combined with neurons, microglia, astrocytes, and pericytes, comprises the dynamic neurovascular unit (NVU), which regulates cerebral blood flow, ion homeostasis, nutrient transport, and waste clearance (Zhao et al., 2022). BBB ECs establish a highly specialized physical barrier through complex tight junctions (TJs) that regulate molecular exchange between the blood and brain (Xiang et al., 2025; Liu et al., 2012). In addition to limiting paracellular transport (passive transfer of substances through intercellular spaces between cells), brain ECs exhibit low rates of transcytosis (movement of macromolecules across the interior of cells), express a diverse array of selective influx and efflux transporters, and maintain low levels of leukocyte adhesion molecules, thereby restricting immune cell entry into the CNS (Muldoon et al., 2013; Chaves et al., 2024; Ohtsuki, 2004). BBB function is further supported by multiple coordinated transport systems: (1) blood-to-brain influx mechanisms that supply essential nutrients and signaling molecules, such as glucose, amino acids, hormones, and nucleotides, to the brain and facilitate metabolic communication (Koepsell, 2020; Campos-Bedolla et al., 2014a); (2) brain-to-blood efflux pumps that limit the accumulation of xenobiotics and other potentially harmful compounds within the brain by transporting them into circulation (Löscher et al., 2005), and; (3) brain-to-blood receptor-mediated clearance systems, such as low-density lipoprotein receptor-related protein 1 (LRP1), that facilitate the removal of metabolites and neurotoxic proteins, including amyloid-β (Aβ), from the brain interstitial fluid (Zhao et al., 2016; Ding et al., 2025). Through its tightly regulated junctional architecture, low vesicular trafficking rates, and specialized transport systems, the BBB plays a critical role in maintaining the biochemical stability of the brain microenvironment while serving as a dynamic interface between the peripheral circulation and the CNS that may influence the sleep-wake cycle.

The BBB is a critical support system for the intricate and dynamic functions of the healthy brain, spanning metabolic, physiological, and behavioral domains. Neural activity continuously fluctuates in response to an organism's real-time demands, and these changes are tightly coupled with adaptive modifications in BBB function (Zhao et al., 2022). For example, neurons that lack internal energy stores will require rapid adjustments in glucose delivery via BBB transporters, such as GLUT1, during periods of heightened firing (Koepsell, 2020). Similarly, transitions between brain states, such as the initiation of non-rapid eye movement (NREM) sleep, are marked by robust reductions in interneuron activity in regions including the cortex and hippocampus (Miyawaki et al., 2016). Through neurovascular coupling, cellular components of the NVU dynamically regulate EC permeability and cerebral blood flow to meet local metabolic needs (Pulido et al., 2020; Kaplan et al., 2020). Astrocytes, via their end-feet, secrete growth factors that promote the expression and proper assembly of endothelial TJ proteins (Emsley et al., 2006; Tan et al., 2020; Bohannon et al., 2021; Chhatbar et al., 2022), while microglia respond to metabolic signals and help guide neurotransmitters and neuromodulators toward the vasculature to maintain BBB structure (Ronaldson et al., 2020). Pericytes further contribute by dilating the microvasculature in response to neuronal activity, driving cerebral blood flow and modulating barrier permeability (Hall et al., 2014).

Importantly, these interactions are not uniform throughout the brain but instead display region-specific BBB properties. Variations in vascular density, transporter expression, and perivascular cell coverage create topographically distinct permeability profiles (Emsley et al., 2006; Tan et al., 2020; Bohannon et al., 2021; Chhatbar et al., 2022). For example, the hippocampus exhibits lower pericyte coverage than the cortex, a fact that partially explains its particular susceptibility to BBB dysfunction under pathophysiological challenges (Procter et al., 2021).

Many of the regional differences of the BBB have been most clearly identified in disease states, where gliosis and chronic neuroinflammation coexist with a persistent leakage of circulating molecules into the brain parenchyma (Erickson et al., 2013). A clear example of the importance of BBB heterogeneity is that, although disorders such as Alzheimer's disease (AD) and Parkinson's disease (PD) are associated with widespread BBB disruption (higher permeability to exogenous tracers), their behavioral manifestations arise from region-specific neuronal dysfunction that likely shapes local BBB responses (Denkinger et al., 2024; Al-Bachari et al., 2020). Accordingly, patients with AD exhibit increased gadolinium permeability in the thalamus, basal ganglia, amygdala, hippocampus, and occipital cortex (Denkinger et al., 2024), whereas patients with PD show greater BBB permeability in the substantia nigra and posterior cortex (Al-Bachari et al., 2020). Similarly, chronic sleep loss is consistently associated with increased BBB permeability (He et al., 2014), although the magnitude and regional distribution of these effects vary depending on the permeability tracer, brain region, sleep-loss paradigm, animal model/strain, and age (see Table 2 for a summary of findings). For instance, hippocampal BBB vulnerability is commonly reported in studies using sleep restriction paradigms, which show that 2 h of recovery sleep is sufficient to reverse BBB changes to normal function within the cortex, while the hippocampus exhibits sustained increases in permeability and altered TJ gene expression (Hurtado-Alvarado et al., 2017). Consistent with this, experimental pericyte depletion produces marked regional differences in BBB permeability that cannot be explained solely by equivalent pericyte loss or TJ disruption (Villaseñor et al., 2017).

Table 2.

Summary of experimental studies examining the effects of sleep loss on blood–brain barrier (BBB) integrity and permeability in animal and human models.

Sleep manipulation	Model and sex	Additional manipulation/measurement	BBB permeability tracer changes	Effect of recovery sleep (compared to baseline)	Brain region	Study
72h of sleep deprivation by inverted flowerpot	Mice- C57BL6 Male	N/A	↑ Evans Blue	N/A	Hippocampus	Sun et al. (2020), PMID: 33329306.
			↓ ZO-1
			↓ Ocln
48h of sleep deprivation by moving pedal within a water box with either short nap (every 20 min for 3 min) or long nap (20 min every 3 h) opportunities	Rat -Wistar Male	Implemented a reversed circadian rhythm model: 12h of sleep deprivation during opposite active/sleep timing (7am-7pm)	↓ GLUT-1	Short nap: ↓ Long nap: → baseline	Amygdala left hemisphere	Moghadam et al. (2019), PMID: 34466456.
			= GLUT-1	Short nap:  =  Long nap:  =	Amygdala right hemisphere
20h of sleep restriction by MMPM with 4h of recovery for 10 days, followed by 40 or 120 min of recovery on final day	Rat - Wistar Male	N/A	↑ Na Fluorescein	40 min: ↑ 120 min: → baseline	Hippocampus	Hurtado-Alvarado et al. (2017), PMID: 28543440.
		N/A	↓ TJ proteins	40 min: ↑ 120 min: → baseline	Hippocampus
20 h of sleep restriction by MMPM with 4h of recovery for 10 days	Mice- C57BL6 & BALB/c Male	↑ pro-inflammatory cytokines in C57BL6	↑ Evans Blue in C57BL6	N/A	Cortex, Hippocampus, Vermis	Hurtado-Alvarado et al. (2018), PMID: 29154957.
			↑ 10K Da Dextran in C57BL6	N/A	Cortex, Hippocampus, Vermis
			↑ Na Fluorescein in C57BL6	N/A	Cortex, Hippocampus, Vermis
20h of sleep restriction by MMPM with 4h of recovery for: a) 3 days b) 5 days c) 10 days	Mice - C57BL6 Male	N/A	Evans Blue: 3 days (=) 5 days (↑) 10 days (↑)	N/A	Cortex, Hippocampus	Avilez-Avilez et al. (2025), PMID: 40824389.
		N/A	Na Flourecein: 3 days (=) 5 days (=) 10 days (↑)	N/A	Cortex
		N/A	Na Flourecein: 3 days (=) 5 days (↑) 10 days (↑)	N/A	Hippocampus
		N/A	TJ – Cldn 5: 3 days (↓) 5 days (↓) 10 days (↓)	N/A	Cortex, Hippocampus
		N/A	TJ – ZO 1: 3 days (=) 5 days (↓) 10 days (↓)	N/A	Cortex, Hippocampus
20h of sleep restriction by MMPM with 4h of recovery for 10 days	Rat -Wistar Male	N/A	↓ PDGFR-β	N/A	Cortex, Hippocampus Microvessles	Medina-Flores et al. (2020), PMID: 32485292.
		N/A	↓ Connexin 43	N/A	Cortex, Microvessles
		N/A	= Connexin 43	N/A	Hippocampus Microvessles
		N/A	↑ Na Fluorescein	N/A	Cortex, Hippocampus
		N/A	↑ Rhodamine	N/A	Cortex, Hippocampus
		N/A	↑ Evans Blue	N/A	Cortex, Hippocampus
		N/A	↓ Cldn 5	N/A	Cortex, Hippocampus Microvessles
		N/A	↓ Ocln	N/A	Cortex, Hippocampus Microvessles
20h of sleep restriction by MMPM with 4h of recovery for 10 days	Rat - Wistar Male	A2A receptor antagonist SCH-58261 (in vivo – i.c.v)	↑ 70K Da Dextran-FITC, reverted by SCH-58261	→ baseline	Cortex, Hippocampus, Basal Ganglia, Vermis	Hurtado-Alvarado G et al. (2016), PMID: 27893847.
			↑ Evans Blue, reverted by SCH-58261	→ baseline	Cortex, Hippocampus, Basal Ganglia, Vermis
20h of REM sleep restriction by MMPM with 4h of recovery for 10 days followed by 40 or 120 min of recovery on final day	Rat -Wistar Male	N/A	↑ Evans Blue in whole brain	↑ 40 min	Cingulate gyrus, Insular cortex, Amygdala, Basal nuclei, Septal area, Preoptic arear, Motor cortex, Somatosensory cortex, Rhinal cortex, Hippocampus, Midbrain, Vermis	Gómez-González et al. (2013), PMID: 23713739.
				↑ 120 min	Cingulate gyrus, Insular cortex, Amygdala, Basal nuclei, Septal area, Preoptic arear, Motor cortex, Somatosensory cortex, Rhinal cortex, Hippocampus, Midbrain, Vermis
18h of sleep restriction by rotating bar with 4h of recovery for 6 days	Mice - C57BL6 Male	N/A	↓ GLUT-1	24 h: → baseline	Cerebral microvessels, Whole brain	He et al. (2014), PMID: 25355222.
			↓ H-2DG	24 h: → baseline	Cortex, Subcortical region, Hypothalamus
			↑ Na Fluorescein	24 h: → baseline	Cortex, Subcortical region, Brainstem, Cerebellum
12h of sleep deprivation by wheel	Mice - C57BL6 Male	Adenosine (in vivo – i.c.v)	= Na Fluorescein	N/A	Hippocampus	Harazin et al. (2026), PMID: 42112566.
8h total sleep deprivation	Human - Male	N/A	↑ NSE	N/A	Indirect measurements in serum	Benedict et al. (2014), PMID: 24470708.
		N/A	↑ S-100β	N/A
3 repeated cycles of 7-day sleep restriction by MMPM with 4h of recovery and 7-day intervals in between	Mice - C57BL6 Male	LPS or saline by i.p administration 14 days after the last cycle of sleep restriction	↓ Cldn 5 of LPS mice	N/A	Hippocampus	Xu et al. (2020), PMID: 32741775.
			↓ ZO-1of LPS mice	N/A	Hippocampus
			↓ Ocln of LPS mice	N/A	Hippocampus
			↑ Evans Blue of LPS mice	N/A	Whole Brain
2h of sleep delay by slow rotating wheel	Rat -Wistar Male	N/A	↓ 2-NBDG	N/A	Arcuate Nucleus	Hurtado-Alvarado et al. (2023), PMID: 37725978.
2min intervals of sleep fragmentation by rotating bar during the morning for 4 weeks or 9 weeks	Mice - C57BL6 Male	N/A	↑ 3-5 kDa FITC-dextran	2-week recovery: ↑	Whole brain	Puech et al. (2023), PMID: 37373028.
				6-week recovery: → baseline
Sleep restriction by gentle stimulation using a soft brush for the first 4 h daily for 7 days. Samples taken at: a) 24h after b) 72h after 7 days after sleep restriction	Mice -CD-1 Male Female	N/A	Radiolabeled 14C-sucrose: Males and Females- 1.24h (=) 2.72h (=) 3.7 days (=)	N/A	Whole brain	Hinterberger et al. (2025), PMID: 41413780.
			Radiolabeled 14C-sucrose: Male – 1.24h (=) 2.72h (=) 3.7 days (=)	N/A	Prefrontal cortex
			Radiolabeled 14C-sucrose: Female – 1.24h (=) 2.72h (↑) 3.7 days (=)	N/A	Prefrontal cortex
			Radiolabeled 14C-sucrose: Male – 1.24h (=) 2.72h (↓) 3.7 days (↑)	N/A	Hypothalamus
			Radiolabeled 14C-sucrose: Female – 1.24h (=) 2.72h (=) 3.7 days (↑)	N/A	Hypothalamus
			Radiolabeled 14C-sucrose: Male – 1.24h (=) 2.72h (↓) 3.7 days (↓)	N/A	Hippocampus
			Radiolabeled 14C-sucrose: Female – 1.24h (=) 2.72h (↑) 3.7 days (↑)	N/A	Hippocampus

Abbreviation Legend.

Abbreviation:	Definition:
MMPM	Modified multiple platform method
i.c.v	Intracerebroventricular injection
i.p	Intraperitoneal injection
TJ	Tight junction proteins
Cldn-5	Claudin-5
ZO-1	Zonula occludens-1
Ocln	Occludins
Glut-1	Glutamate Transporter 1
H-2DG	Tritiated 2-deoxy-D-glucose
2-NBDG	2-deoxy-2-D-glucose
3-5 kDa FITC-dextran	Fluorescently labeled polysaccharide
LPS	Lipopolysaccharides

Symbol Legend.

Symbol:	Definition:
↑	Increase
=	No change
↓	Decrease
→	Reversal

Because sleep profoundly alters neuronal (Dworak et al., 2010) and glial (Flores-Valle et al., 2025) activity throughout the brain, sleep and sleep loss are likely to produce region-specific changes in BBB function. In the following sections, we critically examine the physiological evidence supporting a bidirectional relationship between sleep and the BBB through a framework that emphasizes (1) regional heterogeneity of the BBB, including differences in vascular and glial composition; (2) diversity in selective transport mechanisms; and (3) region-specific neuronal activity and metabolic demands. Nonetheless, we must acknowledge that important biological variables remain underrepresented in studies of sleep and BBB function. In particular, the influence of biological sex has received limited attention despite growing recognition that sex-dependent differences may shape neurovascular physiology (for more information, see (Moon et al., 2021; Salvi de Souza et al., 2025; Shao et al., 2024; Swift et al., 2020)).

Under transient conditions of stress or pathophysiology, numerous factors can alter BBB permeability (e.g., inflammation, age, sex, cellular organization, gene expression), demonstrating that the BBB is a dynamic interface capable of active remodeling. We therefore propose that the BBB exhibits plasticity in response to changes in sleep–wake history, such that acute sleep loss and subsequent recovery induce adaptive, reversible remodeling, whereas chronic sleep loss drives maladaptive barrier dysfunction - typically characterized by TJ disruption, increased vesicular transport (transcytosis), upregulated adhesion molecule expression, enhanced leukocyte infiltration, and impaired communication between ECs, pericytes, and astrocytes, collectively increasing barrier permeability. Accordingly, this review is guided by the framework that adaptive BBB remodeling is reversible and preserves physiological and behavioral function, whereas maladaptive remodeling persists beyond recovery and compromises BBB integrity. These persistent alterations may impair waste clearance, promote neuroinflammation, contribute to cognitive and metabolic dysfunction, and increase vulnerability to neurodegenerative disease (Friedman et al., 2025; Banks et al., 2021), thereby facilitating long-term neuropathology.

3. The bidirectional nature of sleep and the BBB

The idea that sleep and BBB function are bidirectionally linked emerged from early observations that BBB activity fluctuates across the sleep-wake cycle, with barrier integrity and regulatory functions enhanced during sleep and reduced during wakefulness. Building on these observations, Korth (1995) proposed one unifying hypothesis, stating that sleep functions (in part) to preserve BBB integrity and limit brain exposure to peripheral inflammatory signals that accumulate during wakefulness. The inflammatory signals of this hypothesis primarily involve microbiome-derived bacterial cell wall components, whose concentrations coincide with rising levels of sleep-promoting proinflammatory cytokines, including tumor necrosis factor alpha (TNF-α) and interleukin-1 beta (IL-1β) (Korth, 1995). Two major lines of evidence have since strengthened this hypothesis by demonstrating reciprocal communication between sleep and BBB physiology. First, bacterial products (such as muramyl dipeptide, lipopolysaccharide, and sleep-promoting peptides) can alter sleep behaviors, such as significant increases in sleep duration following administration (Szentirmai et al., 2024a). These changes in sleep demonstrate the ability of peripheral immune signals to influence the brain that can then have downstream effects on behavior (Krueger et al., 1982, 2016; Szentirmai et al., 2024b). Second, small peptides in circulation (e.g., insulin, growth hormone-releasing hormone, galanin, and delta sleep-promoting peptide) can cross the BBB (for review, see (Banks, 2025)) in sufficient amounts to modulate neural function. The ease with which they cross the BBB appears to increase with sleep loss, indicating that sleep and the BBB may communicate bidirectionally to maintain healthy brain exposure to peripheral signals. In this perspective, the central hypothesis of this review is that the BBB is not merely a passive barrier shaped by sleep (or sleep loss), but a dynamic and responsive interface that actively regulates body-brain communication by integrating local and peripheral signals across sleep and wake states. Sleep preserves BBB integrity and selective transport, whereas sleep loss progressively disrupts these processes (see Fig. 1), thereby increasing brain exposure to peripheral inflammatory and metabolic signals. We propose that this reciprocal relationship underlies many of the vascular, immune, and neural consequences of sleep loss discussed throughout this review.

Fig. 1.

BBB transport and efflux systems during sleep loss. A. The majority of research on sleep loss and BBB dysfunction has focused on non-specific BBB transport systems, namely paracellular and adsorptive mediated transcytosis. Chronic sleep loss has been found to decrease TJ expression, thereby increasing paracellular transport of low- and high-dextran molecules, such as Evans blue and sodium fluorescein. Additionally, the increase in non-specific transcytosis during sleep loss allows for greater transport of caveolar vesicles, particularly within the hippocampus. These results are well-validated (depicted by green check mark) in rodent models. Changes in expression of other key proteins, such as tricellulin and Cldn-1, in response to sleep loss remains to be investigated (depicted by yellow question mark) B. Currently, the effect of acute to chronic sleep loss on specific transport and efflux activity remains unexplored (depicted by yellow question marks). Evidence regarding how receptor-mediated transcytosis responds to sleep loss is lacking. However, behavioral and metabolic cues of inadequate sleep suggest decreases in the entrance of metabolic signals (e.g., insulin) or inflammatory mediators (e.g., cytokines) to the brain. It is well-validated that glucose transporter (GLUT)-1 decreases during chronic sleep loss (depicted by green check mark), suggesting that carrier-mediated transport may be impaired, altering the delivery of key nutrients (such as amino acids) into the brain. Consistent results indicate increased permeability to Rhodamine 123, a substrate of Pgp during chronic sleep loss (depicted by green check mark), suggesting lower protein-mediated efflux activity, which could allow for the accumulation of wake-related molecules concomitantly to decreased chemical barrier protection. It is currently unknown if sleep loss alters Pgp expression, regional differences of Pgp activity, and the transport of endogenous Pgp substrates into the brain (such as hormones and glucocorticoids). Image created by Biorender®.

The following subsections discuss how changes in brain areas influence the BBB's adaptation or failure during sleep loss. Importantly, we emphasize critical gaps in knowledge that must be addressed for a complete understanding of how the BBB and sleep not only interact but support one another.

3.1. Cellular components

Components of the NVU differentially regulate both sleep and BBB function, providing multiple cellular pathways through which sleep loss may alter barrier permeability. Astrocytes appear particularly sensitive to sleep disruption. Across paradigms ranging from acute sleep deprivation to chronic sleep restriction, astrocytes undergo rapid structural and functional remodeling (Bellesi et al., 2017). In addition to their established roles in circadian regulation and sleep-wake behavior through calcium signaling (Bojarskaite et al., 2019) and immune-related gene expression (Que et al., 2023; Bellier et al., 2025), astrocytes directly support BBB integrity. Through their perivascular end-feet, astrocytes take up glucose, metabolize it to lactate as an energy source for neurons, and secrete growth factors that promote endothelial TJ assembly (Pociūtė et al., 2024).

Microglia likewise exhibit circadian fluctuations in morphology and molecular signaling while regulating somnogenic molecules within the brain (Lawrence et al., 2024). During acute sleep deprivation, microglial activation enhances neuron-glia communication (Kox et al., 2026) and modulates neurovascular signaling, thereby indirectly supporting BBB structure and contributing to dynamic changes in vascular permeability (Ronaldson et al., 2020).

Pericytes represent another important cellular link between neuronal activity and BBB function. By responding to neuronal activity with capillary dilation and functional hyperemia (i.e., increased cerebral blood flow), pericytes regulate local blood supply while contributing to transient changes in BBB permeability (Hall et al., 2014; Wu et al., 2023). Although their direct role in sleep regulation remains poorly understood, evidence suggests that pericyte integrity is influenced by sleep-wake history. A single episode of acute sleep deprivation does not elevate cerebrospinal fluid (CSF) markers of pericyte damage, whereas repeated chronic sleep deprivation does, suggesting that prolonged sleep loss compromises pericyte function and that recovery sleep may be required for restoration (Wu et al., 2023).

Importantly, astrocytes, microglia, and pericytes exhibit substantial regional heterogeneity. Single-cell transcriptomics and advanced imaging studies demonstrate that glial morphology, density, and gene expression differ across sleep-regulating regions, including the cortex, hippocampus, basal forebrain, thalamus, and hypothalamus (Vanrobaeys et al., 2023). This cellular diversity likely enables region-specific neurovascular responses to local neuronal activity, influencing synaptic plasticity, interstitial neurotransmitter homeostasis, and BBB function across sleep and wake states.

3.2. Proteins, permeability, and gene expression

In addition to regional differences in perivascular cell coverage, BBB heterogeneity is also reflected in variations in vascular density, giving rise to topographically distinct permeability profiles (Abbott et al., 2010; Zhao et al., 2022). Interestingly, the expression of major TJ proteins, including Zonula Occludens-1 (ZO-1), Claudin-5 (Cldn-5), and Occludin (Ocldn) appears relatively consistent across brain regions, including the hippocampus and cortex, suggesting that baseline paracellular barrier properties are largely preserved throughout the cerebral microvasculature (Blanchette et al., 2025). These observations indicate that regional differences in BBB function are likely driven by mechanisms beyond TJ abundance alone.

Consistent with this concept, transcriptomic studies demonstrate that sleep loss elicits highly region-specific molecular responses throughout the brain. Although no studies have systematically examined regional changes in BBB transporter expression following sleep loss, whole-brain transcriptomic analyses have revealed substantial spatial heterogeneity in gene expression. For example, 5 h of acute sleep deprivation in mice produced the largest transcriptional changes of differentially expressed genes (DEGs) in the hippocampus, cortex, and hypothalamus (Vanrobaeys et al., 2023). Spatial transcriptomic analyses further showed that more than 90% of the genes downregulated in the hippocampus were upregulated in the neocortex, with additional differences observed among subregions within these structures (Vanrobaeys et al., 2023). While these transcriptional changes were not specific to the BBB, region-specific alterations in gene expression could indirectly influence BBB integrity by affecting signaling pathways that regulate TJ organization and EC function. Indeed, genetic alterations that disrupt the expression, localization, or assembly of TJ proteins are known to compromise barrier integrity and increase paracellular permeability. Together, these findings suggest that sleep loss engages distinct molecular programs across brain regions that may differentially influence BBB structure and function, highlighting an important area for future investigation.

Despite regional complexity and the dynamic nature of BBB heterogeneity, most studies investigating BBB dysfunction during sleep loss have focused solely on alterations in TJ proteins, overlooking related physiological changes. Current evidence suggests that acute sleep deprivation has no effect on TJ protein expression or permeability to high-molecular-weight dextrans and sodium fluorescein (see Table 2), which is often reported as a holistic measurement of BBB functioning as a whole, rather than one possible mechanistic change. Furthermore, the expression of other critical proteins in TJ maintenance, such as tricellulin (Iwamoto et al., 2014) or Cldn-1, remains unexplored. Importantly, changes in barrier function may occur independently of altered TJ abundance. Therefore, it is critical to consider that dynamic reorganization of TJs and the cytoskeleton, as well as changes in junction-associated signaling pathways involving proteins such as β-catenin and ZO-1, can modify BBB permeability without requiring measurable changes in protein expression (Colpitts et al., 2026). These mechanisms remain largely unexplored in the context of sleep loss, but exploring region-specific alterations in gene expression in the BBB is an important area for future investigation.

Accordingly, the following sections examine another major, yet understudied, determinant of BBB function—selective transport systems—and consider how regional differences in transporter expression may contribute to sleep-dependent regulation of the BBB (see Fig. 1).

3.3. Carrier-mediated transport

Carrier-mediated transport enables the selective movement of essential nutrients and metabolites across the BBB through specialized endothelial membrane proteins (Jones et al., 2007). Because molecules such as glucose, amino acids, hormones, and metabolic substrates cannot freely diffuse across the endothelial lipid bilayer, they rely on transporters including hexose (e.g., GLUT1), monocarboxylate (e.g., MCT1), amino acid (e.g., LAT1), and amine transport systems (Tsuji et al., 1999) within the BBB. Despite their fundamental role in brain metabolism, the regulation of these transporters by sleep remains largely unexplored (Tsuji et al., 1999; Pan et al., 2017).

Evidence from Drosophila models suggests that several carrier-mediated transporters, including LAT1, MCT1, carnitine transporters, and organic cation/anion/zwitterion transporters, contribute to sleep regulation or are altered by sleep deprivation (He et al., 2014; Avila et al., 2025; Cuddapah et al., 2019; Hu et al., 2020). In mammals, however, studies have focused almost exclusively on GLUT1, the principal glucose transporter at the BBB. Acute sleep deprivation does not alter GLUT1 expression in ECs from the hypothalamus, hippocampus, or cortex (Petit et al., 2013), whereas chronic sleep restriction reduces GLUT1 expression in cerebral microvessels and decreases brain glucose uptake (Hurtado-Alvarado et al., 2023).

GLUT1 expression is itself regionally heterogeneous. In rats, GLUT1 abundance and capillary density are greatest in the retrosplenial cortex, intermediate in the hippocampus, and lowest in the cerebellum, paralleling regional differences in glucose utilization (Zeller et al., 1997). Mutations in Slc2a1 (GLUT1) are also associated with abnormal sleep phenotypes (Furuse et al., 2019). However, despite marked changes in neuronal activity, glucose utilization, and cerebral blood flow across sleep stages (Boyle et al., 1994), it remains unknown whether GLUT1 is dynamically regulated during the sleep-wake cycle.

Little is known about the effects of sleep on other BBB transport systems. For example, MCT1, which transports lactate, pyruvate, and ketone bodies, is enriched in ECs of the cortex and hippocampus and supports astrocyte-neuron metabolic coupling (Pierre et al., 2005). Although direct evidence linking BBB MCT1 to sleep is lacking, pharmacological inhibition of monocarboxylate transport increases NREM and REM sleep while reducing wakefulness (Braga et al., 2024), suggesting that carrier-mediated transport represents an important but understudied mechanism linking sleep, brain metabolism, and BBB function.

3.4. Transcytosis-mediated transport

Transport via transcytosis represents vesicle-mediated movement through the EC cytoplasm rather than through the paracellular (interendothelial) spaces between cells. Current evidence indicates that chronic sleep restriction increases nonspecific caveolar (small, flask-shaped invaginations of the cell membrane) transcytosis in the hippocampus (Blanchette et al., 2025). In contrast, the effects of sleep loss on receptor-mediated transcytosis remain unexplored. This represents an important knowledge gap because receptor-mediated transcytosis governs the selective transport of numerous signaling molecules (Baghirov, 2025) involved in sleep and brain homeostasis, including insulin (Rennick-Zuefle et al., 2026), leptin (van Egmond et al., 2023), transferrin (Ridefelt et al., 2010), and cytokines (Krueger et al., 1995). Consequently, regional differences in endothelial receptor expression, which are rarely examined in studies of sleep and BBB function, are likely to influence the local transport and signaling of these molecules.

Emerging transcriptomic evidence supports the existence of regional specialization in receptor-mediated transcytosis. Functional single-cell transcriptomic analyses combining plasma uptake assays with EC sequencing have shown that protein uptake is particularly enriched in hippocampal ECs from young mice and is associated with elevated expression of receptor-mediated transcytosis genes, including Transferrin receptor 1 (TfRc), Insulin-like growth factor 1 receptor (IGF-1R), and Low-density lipoprotein receptor-related protein 8 (LRP8) (Yang et al., 2020). Interestingly, aging has been shown to shift BBB protein transport away from receptor-mediated pathways toward increased nonspecific caveolar transcytosis (Yang et al., 2020).

Whether sleep loss induces a similar shift to that of aging is currently unknown. However, because chronic sleep restriction increases endothelial senescence markers and the abundance of caveolae in the hippocampus (Avilez-Avilez et al., 2025), it is plausible that prolonged sleep loss promotes a transition from selective receptor-mediated transport toward less selective caveolar transcytosis. Determining whether sleep alters the balance between these transport pathways may provide important insight into how BBB selectivity changes during chronic sleep disruption.

3.5. Efflux systems

Brain clearance broadly refers to the removal of endogenous metabolites, proteins, xenobiotics, and other solutes from the central nervous system (CNS) through multiple pathways, including bulk fluid movement (glymphatic system), cellular uptake and degradation, metabolic transformation, and transport across CNS barriers (Liu et al., 2021). Decreased brain clearance has been widely implicated in sleep loss as a potential mechanistic link to AD. Particularly, the movement of interstitial fluid via the glymphatic system (for more information, see (Chong et al., 2022; Nycz et al., 2021)) along the perivascular spaces facilitates the removal of Aβ and is dependent on the aquaporin-4 (AQP4) water channel localized to astrocytic endfeet (Porter et al., 2026). Genetic variation in the AQP4 gene has been associated with AD-related phenotypes that are modified by sleep quality. For example, carriers of the minor allele of the AQP4 single nucleotide polymorphism (SNP) rs2339214 exhibit greater rates of gray matter atrophy in association with shorter sleep duration (Porter et al., 2026). Although the glymphatic system has received the greatest attention in the context of sleep loss and AD development, it represents only one mechanism in which the brain removes wake-related accumulation of waste. Unlike the broad concept of brain clearance, brain efflux refers specifically to transporter-mediated export of solutes from the brain to circulation and therefore differs from glymphatic function in both scope and mechanism (Löscher et al., 2005) (see Table 3 for summary of differences). While these systems are mechanistically distinct, they function in a coordinated manner. For example, glymphatic transport can facilitate the movement of Aβ toward CNS barriers, where efflux transporters mediate its export into the circulation (Yoon et al., 2012; Deane et al., 2009; Storck et al., 2018). Among BBB efflux transporters, P-glycoprotein (Pgp; multidrug resistance protein 1, MDR1) is a major ATP-dependent efflux transporter at the BBB and a determinant of the BBB's selective permeability (Löscher et al., 2005). By exporting a broad range of endogenous and exogenous substrates from ECs into the blood, Pgp serves as a key mechanism of neuroprotection and waste clearance at the BBB (Löscher et al., 2005).

Table 3.

Comparison of the glymphatic clearance system and P-glycoprotein (Pgp)-mediated efflux as complementary brain waste-removal pathways.

Current evidence indicates that Pgp function is dynamically regulated by both circadian (Savolainen et al., 2016) and neuronal activity–dependent mechanisms (Zhu et al., 2004). Most investigations have focused on circadian variation rather than the effects of sleep manipulation on Pgp activity. In rodents, evidence suggests that endothelial circadian clock genes drive diurnal fluctuations in Pgp expression and activity, with higher efflux observed during the sleep phase compared to the active phase (Savolainen et al., 2016) (Fig. 2A and B). These findings suggest that intrinsic endothelial clock mechanisms contribute to temporal regulation of BBB transport. However, an important future distinction will be to separate circadian phase effects from sleep loss-driven impairment.

Fig. 2.

Sleep loss influences BBB function from adaptability to failure - summarizes a tractable, mechanistic framework of this review. A: Independent studies have shown that neuroglia production of adenosine and TNF-α rise during wakefulness to modulate neuronal activity via their receptors, inducing somnolence and promoting sleep. Glutamatergic neuronal activity has been associated with reduced Pgp function by decreasing Pgp mRNA transcription, allowing for the accumulation of signaling molecules and Pgp substrates in the brain throughout wake. B: Following sleep onset, production of adenosine and TNF-α decreases, while the clearance of metabolites from the brain to the blood increases in association with altered neuronal activity (for example, the slowing of neuronal activity during NREM sleep). This can contribute to enhanced Pgp function by removing the inhibitory signal of adenosine and TNF-α on Pgp gene expression; thus, limiting the influx of lipophilic molecules into the brain and promoting Pgp-mediated clearance of molecules such as Aβ. C: The primary hypothesis of this review is that BBB changes in response to sleep loss may initially occur for the adaptability of the organism by modifying brain permeability in response to neuronal requirements. This adaptability can be expressed as reversible changes in transport and clearance activity with recovery sleep. If acute sleep loss becomes chronic, the initial protective changes in BBB activity can become maladaptive, promoting sustained hyperpermeability that allows potentially toxic blood-borne molecules to enter the brain. Consistent alterations in transport and efflux activity may promote health consequences that increase the propensity for neurodegenerative diseases. Confocal pictures following immunofluorescence (provided by GHA) in the bottom panel illustrate a possible increase in Pgp substrates (indicated by the question mark for requiring further investigation) within the brain during acute sleep loss, indicating that Pgp efflux activity is reduced. Sustained downregulation of Pgp may allow for the accumulation of wake-related signaling molecules that drive neuroinflammation and alter the neural environment, contributing to downstream behavioral and physiological consequences.

Image created by Biorender®.

In addition to circadian control, Pgp is likely sensitive to neuronal activity. Both optogenetic activation of glutamatergic neurons and sensory stimulation reduce cortical Pgp expression and function (Pulido et al., 2020; Hall et al., 2014). Together, these findings suggest that fluctuations in neuronal activity across sleep-wake states may dynamically regulate BBB efflux capacity (Fig. 2A and B). One possibility is that reduced neuronal activity during NREM sleep enhances Pgp-mediated efflux, whereas sustained wakefulness suppresses efflux capacity; however, this remains to be directly tested.

Importantly, Pgp also transports a range of endogenous signaling molecules, including steroid hormones such as cortisol, aldosterone, and estrogens, in addition to xenobiotics (Savolainen et al., 2016). Thus, activity-dependent changes in Pgp may influence not only neuroprotection (Mahringer et al., 2016) but also brain–body signaling. Whether Pgp regulation differs across brain regions in response to sleep-wake state transitions remains unknown. However, one hypothesis that may mechanistically explain the accumulation of wake-associated molecules in the brain during sleep is that Pgp activity appears to peak during the sleep phase (Savolainen et al., 2016), suggesting that effective Pgp-mediated clearance depends on sufficient sleep. In contrast, extended wakefulness likely impairs Pgp function in concert with sustained neuronal activation and altered NVU signaling (Pulido et al., 2020; Zhu et al., 2004), permitting the accumulation of molecules that rise with sleep pressure (somnogenic) that are normally cleared during sleep. Fig. 2 illustrates this hypothesis; Fig. 2A depicts the decreased efflux activity, heightened glutamatergic activity, and increased expression of somnogenic signaling molecules, whereas Fig. 2B portrays the increase in efflux activity alongside altered neuronal activity during sleep that aids in the clearance of circulating molecules (such as pro-inflammatory cytokines and peptides) from the brain. Fig. 2C details the primary hypothesis of this review, describing how changes in BBB transport and efflux systems may initially occur for the adaptation of the organism in response to acute sleep loss. Should acute sleep loss become chronic, these adaptations can become maladaptive. The sustained decrease in transport and efflux activity may impair the delivery of nutrients to the brain and allow for the accumulation of harmful substrates in the brain, leading to health consequences that may increase susceptibility to neurodegeneration. Overall, the interaction between neuronal activity, NVU function, BBB permeability, and Pgp regulation provides a viable model through which sleep loss compromises BBB integrity, reduces barrier selectivity, and limits the clearance of potentially harmful brain metabolites.

4. Unmasking the mechanisms behind BBB changes during acute sleep loss

Sleep regulation depends on dynamic fluctuations in systemic and central sleep-regulatory molecules that arise from wake-related processes such as energy expenditure. Sleep-wake activity is governed by the interaction of the circadian system, which governs sleep timing, and the homeostatic system, which reflects sleep need (see Table 1) (Borbély, 1982; Borbély et al., 1999; Czeisler et al., 2007). In addition to these global regulatory processes, sleep pressure may accumulate locally within specific neural circuits (Krueger et al., 2011a). The theory of local sleep proposes that, during prolonged wakefulness, discrete populations of neurons or brain regions can exhibit sleep-like activity patterns while the organism remains behaviorally awake (Krueger et al., 2019). These localized neuronal off-states are thought to arise from use-dependent increases in sleep pressure, whereby heavily utilized neural circuits require recovery before the rest of the brain (Krueger et al., 2011a). As wakefulness is extended, local sleep events become more frequent and widespread, potentially contributing to cognitive lapses and impaired performance (Rector et al., 2009). Consequently, the sleep-wake cycle may be regulated not only at the whole-brain level through circadian and homeostatic mechanisms, but also at the regional level through activity-dependent recovery processes within individual neural networks. Because the BBB exhibits substantial regional heterogeneity in its cellular composition (Emsley et al., 2006; Bohannon et al., 2021), transporter expression (Blanchette et al., 2025), and functional properties, and because sleep appears to regulate multiple aspects of BBB function, the local sleep framework raises the possibility that sleep-loss-induced BBB dysfunction may also occur at a localized level and/or a region-specific level. Under this model, brain regions that have more baseline BBB vulnerability may express even higher alterations to BBB permeability, transport, and efflux function if neuronal activity increases locally during sleep loss.

Following the local sleep framework, BBB-mediated transport of sleep-regulatory molecules may provide a mechanism through which regional physiological demands are translated into localized changes in neural activity and sleep propensity. Scientific consensus holds that sleep serves multiple vital functions rather than a single purpose, with various theories focusing on benefits such as synaptic downscaling, plasticity, cellular repair, and brain waste clearance (Berndt et al., 2019). Notably, the BBB may contribute directly or indirectly to each of these processes through its regulation of molecular exchange between the brain and peripheral circulation. Via the regulation of molecular trafficking and selective permeability, the BBB likely acts as a critical interface through which local and peripheral signals influence sleep-wake regulation. During sleep loss (stemming anywhere from restriction to total deprivation), circulating levels of several proinflammatory mediators increase, including TNF-α (Krueger et al., 1995), interleukin-6 (IL-6) (Vgontzas et al., 1999; Sang et al., 2023), interleukin-17 (IL-17) (Sang et al., 2023), prostaglandin D2 (Sang et al., 2023; Chua et al., 2015), and C-reactive protein (CRP) (Meier-Ewert et al., 2004), many of which have been shown to modulate sleep and alter BBB function. For example, the proinflammatory cytokine IL-6 is linked to sleep promotion, as central administration increases NREM sleep (Hogan et al., 2003), and it has also been found to upregulate during sleep loss in both peripheral circulation and the brain. Elevated IL-6 levels have been associated with altered endothelial function and vascular tone (Blecharz-Lang et al., 2018; Woods et al., 1999), which may contribute to impaired BBB integrity under conditions of sustained inflammation. Similar increases in multiple proinflammatory cytokines occur during prolonged wakefulness. Through interactions with the BBB, cytokines can act on NVU components to disrupt TJ organization and integrity, increase endothelial stress, and enhance BBB permeability (Gryka-Marton et al., 2024), thereby facilitating greater brain exposure to peripheral signals. Consistent with the local sleep framework, sleep-loss-induced changes in BBB permeability may occur preferentially in specific brain regions, resulting in regional differences in exposure to circulating signaling molecules. One hypothesis arising from this framework is that increased exposure of brain regions involved in attention, executive function, and vigilance to circulating signaling molecules may contribute to the cognitive impairments observed during sleep loss. However, not all cytokines are transported into the brain by the BBB, and transport rates vary among cytokine type, brain region, age, and disease state (Banks, 2005; Banks et al., 1989, 1995). This complexity has fueled debate regarding the specific contributions of individual cytokines to BBB dysfunction during sleep loss, highlighting both the bidirectional relationship between systemic sleep-regulatory mechanisms and BBB function, as well as the need for continued investigation.

During prolonged wakefulness, increased production of adenosine, along with sleep-regulatory cytokines such as TNF-α, contributes to the accumulation of sleep pressure (Porkka-Heiskanen et al., 1997, 2000; Floyd et al., 1997). Experimental evidence further suggests that both molecules are capable of modulating BBB function after entering the brain and acting on neurovascular components (Zipp et al., 2023), ECs (Voirin et al., 2020), and interactions with Pgp activity. Consequently, adenosine and TNF-α have emerged as candidate mediators linking sleep pressure to BBB regulation during sleep loss, although the extent to which endogenous concentrations achieved during sleep deprivation produce these effects in vivo remains unclear.

4.1. Adenosine & TNF-α: Dual roles in sleep and BBB regulation

Prolonged wakefulness initiates a biochemical cascade that elevates sleep-promoting molecules such as adenosine and TNF-α, driven by the homeostatic system, thereby heightening sleep need. Physiological levels of adenosine and TNF-α in the brain exhibit a 10-times difference between sleep and wakefulness, with highest concentrations at sleep onset (Floyd et al., 1997; Porkka-Heiskanen et al., 2002, 2011). Both adenosine and TNF-α are inflammatory mediators that can impair BBB integrity by disrupting TJ structure and organization, a mechanism well documented in neuroinflammatory disorders (Voirin et al., 2020; Carman et al., 2011). However, whether these molecules similarly influence BBB dynamics under differing physiological conditions remains unclear. Most evidence for BBB modulation by adenosine and TNF-α comes from pathological conditions, where circulating and brain concentrations can increase by up to 1000-fold. In contrast, sleep loss induces much smaller increases in these molecules, making it difficult to determine whether they have comparable effects on BBB integrity. For example, adenosine concentrations typically range from approximately 10–150 nM across the sleep–wake cycle, whereas pathological conditions such as ischemia or hypoxia can elevate adenosine into the micromolar (μM) range (Salmaso et al., 2025). Consequently, the effects of acute sleep loss on BBB function should not be inferred directly from findings in clinical or experimental models of severe inflammation or ischemia.

Adenosine is produced through the sequential dephosphorylation of ATP as cellular energy demands increase (Agteresch et al., 1999). Increased neuronal activity during extended wakefulness accelerates ATP breakdown, thereby increasing extracellular adenosine (Porkka-Heiskanen et al., 2002, 2011). The progressive accumulation of adenosine during wakefulness and the decrease after sleep reflect the fundamental principle that greater neural activity and energy expenditure during periods of wake increase the homeostatic sleep drive (Porkka-Heiskanen et al., 2002, 2011). The activation of adenosine receptors depends on the extracellular concentration of adenosine and the distribution of receptors in several neural cells. Notably, adenosine exerts a polarized effect on the BBB – it tightens the paracellular barrier (i.e., decreasing permeability) on the luminal side while increasing permeability on the brain side (Harazin et al., 2026). Adenosine receptor A_2A (ADORA2A) signaling is both critical in sleep regulation and has been implicated in the regulation of BBB permeability, particularly to multiple immune regulatory molecules, including dextrans and immunoglobulins (Carman et al., 2011). Pharmacological activation of ADORA2A increases drug entry into the brain, which may be related to interactions with Pgp, although the precise mechanisms underlying this effect remain incompletely understood (Kim et al., 2016). In cultured human brain ECs, acute ADORA2A activation rapidly decreases Pgp activity, and in mice, ADORA2A agonists promote Pgp degradation through matrix metalloproteinase-dependent pathways (Kim et al., 2016). These findings demonstrate that adenosine signaling is capable of modulating BBB transport function. However, whether endogenous adenosine concentrations achieved during physiological sleep deprivation are sufficient to produce comparable effects on BBB permeability or Pgp activity in vivo remains unknown.

In the CNS, TNF-α is produced by microglia, astrocytes, and neurons and contributes to synaptic plasticity, neuronal regulation, and tissue remodeling, while at higher concentrations it promotes inflammatory signaling (Pan et al., 1997). Both systemic and central administration of TNF-α promotes spontaneous sleep and increases NREM sleep (Krueger et al., 1995, 2011b; Krueger, 2008). Endogenous TNF-α levels also exhibit a diurnal rhythm within sleep-regulatory regions of the hypothalamus, with peak concentrations occurring during wakefulness and lower concentrations during sleep (Floyd et al., 1997). Evidence indicates that lower concentrations rapidly and reversibly decrease Pgp expression in ECs in vitro without altering TJ permeability (Hartz et al., 2006). TNF-α-driven increases in glutamatergic neuronal activity may also engage downstream signaling pathways that suppress Pgp expression, linking inflammatory signaling, neuronal activity, and BBB efflux regulation (Pickering et al., 2005). However, the magnitude and direction of TNF-α-mediated BBB effects can also be influenced by several factors including the duration of exposure, receptor signaling, cell type, and physiological context. Consequently, it remains unclear whether endogenous TNF-α elevations associated with sleep loss produce BBB responses observed in experimental inflammatory and disease models.

Together, adenosine and TNF-α are well-established regulators of sleep and contributors to the accumulation of sleep pressure during wakefulness (Porkka-Heiskanen et al., 1997, 2000; Krueger, 2008), but their roles in BBB regulation during sleep loss remain less well defined. Available evidence suggests that both molecules are capable of modulating BBB permeability and efflux function under certain experimental conditions and may therefore represent mechanistic links between homeostatic sleep need and BBB regulation. One proposed mechanism involves the downregulation of Pgp, which shows a reduced BBB efflux capacity during wakefulness associated with neuronal activity. Via this change in Pgp activity during wakefulness, we can hypothesize that similar effects occur during prolonged wakefulness when neuronal activity can remain aberrantly elevated. In addition, elevated adenosine and TNF-α signaling can downregulate Pgp (Kim et al., 2016; Hartz et al., 2006) which may contribute to the accumulation of metabolic waste and other substrates within the brain by impairing efflux capacity. Moreover, TNF-α-driven increases in glutamatergic neuronal activity (Pickering et al., 2005) can indirectly engage downstream signaling pathways that suppress Pgp expression, linking inflammatory signaling, neuronal activity, and BBB efflux regulation. If chronic, impairment in Pgp-mediated efflux systems may contribute to neuroinflammation, disruption of neural signaling, and impairment to neurobehavioral performance due to excessive accumulation of waste and other harmful substrates. Although this framework represents a working hypothesis based on limited evidence, it provides a testable model through which sleep-regulatory signaling molecules may influence BBB function during sleep loss.

5. Blood-brain barrier adaptation during sleep loss: A double-edged sword

Many physiological responses to stress are initially adaptive, preserving homeostasis but becoming maladaptive when sustained. Although this review proposes that BBB responses to sleep loss are initially adaptive and reversible, direct evidence demonstrating that these changes confer functional benefits remain limited. The available evidence shows that sleep can restore BBB permeability to exogenous tracers, suggesting transient BBB remodeling after chronic sleep restriction (Hurtado-Alvarado et al., 2017) (see Table 2); however, short-term sleep recovery does not fully restore synaptic function (García-Aviles et al., 2025). Whether acute sleep loss modulates BBB dynamics to support homeostasis by transient changes in BBB transporters, junctional proteins, and efflux pathways remains unknown.

Evidence from human sleep deprivation studies further illustrates the transition from adaptive to maladaptive responses, suggesting that the consequences of chronic sleep debt cannot always be fully reversed by equivalent recovery sleep. Following total sleep deprivation, subjective sleepiness in humans can normalize after a short period of recovery sleep, whereas objective deficits in attention (Hudson et al., 2020), elevated inflammatory mediators (e.g., TNF-α, IL-6, CRP) (Faraut et al., 2012), and altered brain adenosine levels (Kim et al., 2015) persist for hours or days. These findings suggest that subjective recovery may not accurately reflect physiological recovery, as objective markers of inflammation, cognition, and metabolism remain disrupted despite improvements in perceived sleepiness. Nonetheless, during acute sleep loss, increases in adenosine and TNF-α, together with BBB transport adaptations, may support energy homeostasis, regulate cerebral blood flow, signal mounting sleep need, and facilitate compensatory clearance while sleep-dependent efflux is reduced (Almeida et al., 2025; Liu, 2005). Further, these signaling pathways may help maintain neural function during prolonged wakefulness. For example, acute TNF-α signaling can promote synaptic downscaling to stabilize neural networks under strain (Stellwagen et al., 2006). During recovery, elevated adenosine and TNF-α promote NREM sleep (Van Cauter et al., 2008), during which restorative processes, including glymphatic clearance (Chong et al., 2022; Nycz et al., 2021), are enhanced before levels of these molecules subsequently decline. Collectively, these molecular and BBB-related responses may transiently preserve homeostasis and facilitate recovery sleep. However, when sleep disruption becomes chronic and these responses remain engaged, persistent astrocytic activation, excitotoxic stress, and metabolic strain may shift these initially adaptive mechanisms toward maladaptation, contributing to sustained neurocognitive (Van Dongen et al., 2003), metabolic (Van Cauter et al., 2008), and immune dysfunction (Garbarino et al., 2021).

The magnitude of physiological strain imposed by acute sleep loss, and whether transient alterations in BBB transporters, junctional proteins, and efflux pathways are exclusively protective, are not yet fully understood. Nevertheless, the available evidence is consistent with this framework, suggesting that transient BBB remodeling during acute sleep loss supports homeostasis, whereas persistent alterations induced by recurrent sleep loss impair recovery and prevent normalization of the neural microenvironment.

6. Conclusions and future directions

The dynamic interplay between sleep and the BBB establishes this interface as an active regulator of brain homeostasis rather than a static physical barrier (see Box 1 for a summary). The evidence reviewed here shows that BBB function is highly plastic, adapting across the sleep–wake cycle through coordinated changes in NVU signaling, transporter activity, and efflux systems. These processes likely respond rapidly to building sleep need, indicated by mediators such as adenosine and TNF-α, yet their effects are unlikely to be uniform across the brain. Instead, regional differences in neuronal activity, vascular composition, and transporter expression likely produce localized BBB responses that vary with vigilance state. Current evidence of Pgp regulation provides a plausible prospective mechanistic framework (Fig. 2A and B) that presents both opportunity and challenge: altered BBB gating may be leveraged to enhance central drug delivery, yet insufficient sleep could also render BBB-excluded compounds unpredictably bioactive in the brain.

Box 1. Practice Points.

• 1The BBB is critical for maintaining ionic homeostasis within the brain microenvironment and serves as a dynamic interface through which circulating molecules can influence the sleep–wake cycle.
• 2Cellular components of the NVU can dynamically regulate BBB permeability in response to the physiological demands of the organism.
• 3Chronic sleep restriction impairs BBB integrity and alters its selective permeability, thereby decreasing glucose transport while increasing permeability to high-molecular weight dextrans.
• 4Pgp-mediated efflux activity fluctuates across the sleep-wake cycle, with reduced glutamatergic activity during sleep associated with enhanced efflux function relative to wakefulness.
• 5Sleep deprivation promotes the accumulation of TNF-α and adenosine, which might modulate BBB efflux capacity by downregulating Pgp expression in endothelial cells.
• 6Through their regulatory effects on Pgp, the primary efflux transporter of the BBB, increased adenosine and TNF-α signaling during sleep loss may contribute to the accumulation of metabolic waste and other substrates within specific brain regions.

Alt-text: Box 1

This regional and functional plasticity has important implications for both neuroscience and therapeutic strategies. Sleep-dependent changes in BBB function may influence not only neuroprotection and waste removal but also nutrient delivery, brain–body communication, and the efficacy of CNS-targeted therapeutics. Even in the context of increased permeability and TJ disorganization, sleep loss may paradoxically impair access to essential nutrients required for neural signaling and effective brain–body communication (Campos-Bedolla et al., 2014b). Understanding how these processes are coordinated across the sleep–wake cycle will be essential for elucidating the physiological consequences of insufficient sleep and identifying strategies to preserve brain function.

In closing, we argue that advancing the field will require moving beyond an overreliance on exogenous permeability tracers, which are well suited for detecting structural BBB disruption but fail to capture the dynamic transporter- and signaling-mediated processes that define BBB function in the healthy brain. Rather than asking whether the BBB is simply “open” or “closed,” future studies should determine which transport pathways are engaged, how they differ by brain region, and how they change across sleep-wake states. Viewing the BBB as both a sensor and an effector of sleep-dependent physiology provides a new framework for understanding brain homeostasis and disease vulnerability, while offering new opportunities for therapeutic intervention in an increasingly sleep-deprived society (see Box 2 for pivotal research routes devised from current limitations).

Box 2. Research Agenda.

• 1Further research is needed to determine how sleep loss alters the entry of molecules and nutrients into specific brain regions and to characterize the resulting effects on behavior and cognition.
• 2Future studies should evaluate whether the timing of drug administration relative to the sleep–wake cycle influences drug efficacy by comparing administration during sleep, when Pgp activity is elevated, with administration during wakefulness, when Pgp activity is reduced. Such studies could provide insight into whether sleep–wake-dependent changes in Pgp function affect drug delivery to the brain.
• 3Future investigations should examine genetic variants affecting BBB components, including tight junction proteins such as occludin and claudins, and determine how these variants influence sleep through altered molecular transport and changes in the ionic microenvironment of specific brain regions.
• 4Further evidence is needed to determine whether BBB integrity within specific brain regions contributes to regulation of the sleep-wake cycle. Experimental studies examining the effects of localized BBB disruption on sleep behavior could help establish a mechanistic link between regional BBB dysfunction and sleep disturbances.

Alt-text: Box 2

CRediT authorship contribution statement

Sofia K. Fluke: Investigation, Visualization, Writing – original draft, Writing – review & editing. Gabriela Hurtado-Alvarado: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. Christopher J. Davis: Supervision, Writing – review & editing. Carolina Escobar: Supervision, Writing – review & editing. Natalí N. Guerrero-Vargas: Supervision, Writing – review & editing. Luis Ernesto García-Páez: Investigation, Visualization, Writing – review & editing. Rocío Ruiz-Manzano: Investigation, Visualization, Writing – review & editing. Arturo Contis-Montes de Oca: Investigation, Visualization, Writing – review & editing. Brieann C. Satterfield: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.

Funding

IBRO Rising Star award and DGAPA-PAPIIT award IA202725 to GHA for support with the tools required for writing and image design. US Army Research Office award W911NF2210223 (subward SUBK00016417) to BCS for support.

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

No data was used for the research described in the article.