A circadian clock regulates efflux by the blood-brain barrier in mice and human cells

Zhang, Shirley L.; Lahens, Nicholas F.; Yue, Zhifeng; Arnold, Denice M.; Pakstis, Peter P.; Schwarz, Jessica E.; Sehgal, Amita

doi:10.1038/s41467-020-20795-9

Download PDF

Article
Open access
Published: 27 January 2021

A circadian clock regulates efflux by the blood-brain barrier in mice and human cells

Nature Communications volume 12, Article number: 617 (2021) Cite this article

12k Accesses
59 Citations
105 Altmetric
Metrics details

Subjects

Abstract

The blood-brain barrier (BBB) is critical for neural function. We report here circadian regulation of the BBB in mammals. Efflux of xenobiotics by the BBB oscillates in mice, with highest levels during the active phase and lowest during the resting phase. This oscillation is abrogated in circadian clock mutants. To elucidate mechanisms of circadian regulation, we profiled the transcriptome of brain endothelial cells; interestingly, we detected limited circadian regulation of transcription, with no evident oscillations in efflux transporters. We recapitulated the cycling of xenobiotic efflux using a human microvascular endothelial cell line to find that the molecular clock drives cycling of intracellular magnesium through transcriptional regulation of TRPM7, which appears to contribute to the rhythm in efflux. Our findings suggest that considering circadian regulation may be important when therapeutically targeting efflux transporter substrates to the CNS.

Blood-brain barrier associated tight junction disruption is a hallmark feature of major psychiatric disorders

Article Open access 02 November 2020

Physiological blood–brain transport is impaired with age by a shift in transcytosis

Article 01 July 2020

Defining diurnal fluctuations in mouse choroid plexus and CSF at high molecular, spatial, and temporal resolution

Article Open access 22 June 2023

Introduction

The blood–brain barrier (BBB) is a structure in the CNS vasculature that insulates the brain from the periphery. A key function of the BBB is to protect the brain from xenobiotics and other potentially harmful environmental insults; however, it also prevents diagnostic and therapeutic drugs from entering the brain^1,2. Understanding regulation of the BBB is important for addressing effects of the periphery on the brain and also for improving methods of drug delivery. The BBB is comprised of capillary endothelial cells, surrounded by astrocytic endfeet and pericytes^3,4. BBB endothelial cells constitute a polarized cell layer held together with tight junctions and characterized by low transcytosis, which together restrict paracellular and transcellular passage of molecules into the brain^5,6. The luminal membrane of the BBB has high concentrations of efflux transporters, notably ABCB1 (ATPase binding cassette also known as permeability glycoprotein [Pgp] and multi-drug resistant protein 1 [MDR1]), which is an ATP-dependent pump with broad substrate specificity^7,8.

Circadian clocks are endogenous timekeeping processes that cycle with a period of ~24 h and drive rhythms in most physiological processes. A synchronizing central clock resides in the suprachiasmatic nucleus in the hypothalamus; however, the core molecular clock is present in nearly every cell in mammals^9,10. These peripheral clocks share the same clock machinery as the central clock, but regulate tissue-specific genes and functions^11,12,13,14. The basic mechanism of the clock is a transcription-translation feedback loop in which the Period (Per) and Cryptochrome (Cry) genes are rhythmically transcribed and the two proteins negatively regulate their transcriptional activators, BMAL1 (brain and muscle aryl hydrocarbon receptor nuclear translocator-like protein 1) and either CLOCK (circadian locomotor output cycles kaput) or NPAS2 (neuronal PAS domain protein)¹⁵. Because of the pervasiveness of circadian rhythms in cellular processes^16,17, they are likely involved in the tissue-response to circulating factors, including drugs; however, the extent to which the circadian clock regulates access of molecules/drugs to different mammalian tissues, in particular to the brain, is largely unknown.

To address time-of-day influences on the BBB, we eliminated the molecular clock specifically in the mouse endothelium and used a xenobiotic to measure ABCB1-mediated efflux from the brain. We find that the BBB endothelium has a cell autonomous circadian clock that is required for the generation of rhythms of efflux through ABCB1. We then examined molecular mechanisms of clock-mediated control by conducting transcriptomic analysis using RNA-seq. Surprisingly, unlike the robust rhythms of RNA expression noted in other tissues, very few transcripts cycle in the endothelium and these do not include efflux transporters. We recapitulated efflux cycling in human microvascular endothelial cells and demonstrate oscillations in the level of intracellular free magnesium, a required cofactor for robust efflux. We find binding of BMAL1 to the promoter of the highly expressed magnesium transporter TRPM7 as well as cycling of TRPM7 transcript and protein. Together these data suggest a model in which BMAL1 regulates TRPM7 transcription, which affects intracellular magnesium homeostasis in a rhythmic fashion and produces time-of-day changes in efflux. Our data show the interaction of the molecular clock with BBB function in mammalian cells.

Results

Efflux of xenobiotics at the mouse BBB is gated by the circadian clock

Activity of efflux transporters at the BBB is an important determinant of BBB permeability¹⁸. To measure brain efflux, we intravenously injected Rhodamine B (RHB), a small, lipophilic, fluorescent compound that is a known substrate of ABCB1 (p-glycoprotein) into the jugular vein of the mouse. After 90 min, we harvested the blood, sacrificed and perfused the mouse, and harvested the brain. The amount of RHB fluorescence was measured in the serum and brain homogenate (Fig. 1a). To verify that the level of RHB measured in the brain was a result of biological processes rather than experimental error (such as contamination by blood), we used the ABCB1-specific noncompetitive inhibitor tariquidar (K_d = 5.1 ± 0.9 nM) to block efflux of RHB. We also controlled for possible differences in blood concentrations of RHB, by normalizing RHB levels in the brain to serum levels. We found that administering tariquidar increased the level of RHB in the brain relative to the serum (Fig. 1b), verifying that RHB is effluxed by ABCB1 at the mouse BBB.

**Fig. 1: Efflux of ABCB1 substrate from the brain oscillates over the course of the day.**

To determine whether efflux cycles through the circadian day, we intravenously injected RHB into WT mice at 6 time points throughout the circadian day (zeitgeber [ZT] 2, 6, 10, 14, 18, 22 with lights on and off corresponding to ZT0 and ZT12, respectively) and measured the brain–serum ratio 90 min later. We find that the level of RHB in the brain oscillates throughout the day with the peak early after lights on and trough after lights off, indicating that p-glycoprotein-mediated efflux is highest at night, when the animals are active (Fig. 1c). Levels of RHB in the serum did not show differences over time (Supplementary Fig. 1a).

Higher blood flow due to increased cardiac output and blood pressure during the active period of the mice may increase RHB circulation to the brain¹⁹; however, in our experiments we find decreased RHB levels in the brain at night, suggesting that increased blood circulation does not explain these results. Nevertheless, to ensure that our measurements reflect efflux over time, we sought to determine the amount of RHB lost from the brain at different times of day. We first measured the RHB brain–serum ratio over the circadian day at 60 min post-injection and found, as expected, that RHB accumulated in the brain to higher levels than at 90 min and the RHB brain–serum ratio trended toward a 24-h cycle (JTK pCycle = 0.1) (Supplementary Fig. 1b). By comparing the RHB blood-serum ratio at 60 and 90 min post-injection, we extrapolated the rate of efflux per minute for each ZT time and found that the highest rate was indeed at night (Supplementary Fig. 1c).

To determine whether gross vascular leakage changes over the day and night, which would contribute to RHB accumulation in the brain, we tracked the level of serum albumin in the brain using the high affinity Evans Blue dye. Wild-type mice were given intravenous injections of Evans Blue in the morning (ZT2-6), night (ZT14-18), or 24 h after treatment with vascular leak-inducing inflammatory stimulus, lipopolysaccharide (LPS), and the brain and liver were assessed 30 min after injection. Half of the mice given LPS exhibited Evans Blue staining in the brain, but uninflamed mice did not exhibit BBB vascular leakage in the morning or night (Fig. 1d), indicating that rhythms of RHB in the brain are not generated by vascular leakage.

To determine whether this oscillation was due to a circadian clock within the blood–brain barrier, we used an endothelial-specific (Tie2) cre recombinase to delete a floxed exon of Bmal1/Arntl (Bmal1^fl/fl), which is required for the generation of circadian rhythms. Control mice (Bmal1^fl/fl) have a robust rhythm in RHB efflux; however, in the absence of a functional clock in the BBB endothelium (Bmal1^fl/fl; Tie2-cre), the rhythms are lost and retention of RHB in the brain is high (Fig. 2a, Supplementary Fig. 2a). We then injected Evans Blue to assay for vascular leak and did not observe any differences between control and mutant mice (Supplementary Fig. 2b). To confirm circadian control, we obtained mice mutant for the Cryptochrome 1, Cryptochrome 2 proteins (Cry1^−/−Cry2^−/−), which repress Bmal1 function. We find that Cry1^−/−Cry2^−/− also lack RHB efflux rhythms compared to the heterozygous (Cry1^+/−Cry2^+/−) controls (Fig. 2b). Because BMAL1 is part of the positive arm of the circadian clock while CRY1 and CRY2 are part of the negative arm, we expected that deletion of Bmal1 or Cry1/2 would have opposing effects on the level of RHB in the brain and indeed Cry1^−/−Cry2^−/− mice trended toward lower levels of RHB in the brain (Supplementary Fig. 2c). The high variability of the Cry1^−/−Cry2^−/− mice may be due to some paravascular leakage from increased inflammation in the mice²⁰. Together, these data suggest that the circadian clock in the BBB regulates RHB efflux from the brain.

**Fig. 2: A circadian clock in the BBB regulates brain permeability of RHB.**

Few circadian transcripts in endothelial cells in the mouse brain

To determine the mechanism of circadian clock regulation, we first profiled the transcriptome of brain endothelial cells at different times of day in Bmal1^fl/fl control and Bmal1^fl/fl; Tie2-cre mutant mice. We homogenized and processed brains and used fluorescence-activated cell sorting (FACS) to isolate the brain endothelium marked with anti-CD31 (Supplementary Fig. 3a). Based upon previous single-cell sequencing data from brain endothelial cells, we surmise that ~80% of the cells we isolated are from capillaries that comprise the BBB^21,22. RNA was extracted from the isolated endothelial cells and used to prepare strand-specific libraries with the Clontech Pico V2 kit for 100 bp paired-end sequencing on an Illumina HiSeq 4000. The resulting reads were mapped to the mouse reference genome using STAR²³, and gene-level normalization and quantification were performed using the Pipeline Of RNA-seq Transformations (PORT). We compared the transcriptomes of the collected brain endothelial cells from control and mutant mice and found comparable levels of claudins and occludins (Supplementary Fig. 3b), indicating that the characteristic signature of BBB-forming endothelial cells in terms of tight junction expression level is unaffected by Bmal1 loss. Finally, the transcriptome of the brain endothelial cells was analyzed for 24-h rhythmic expression patterns using MetaCycle. Few genes (45 genes with Meta2d q < 0.6) were rhythmic in controls and even fewer (6 genes with Meta2d q < 0.6) in endothelial-specific Bmal1 knock-outs (Supplementary Datasets 1–3; Fig. 3a). These data are surprising as transcriptomic analysis of other tissues has revealed cycling of up to 20% of the genome²⁴, while here we find that <0.5% of transcripts of the brain endothelial cells are rhythmic. In a previous study of vascular endothelial cells that were isolated and synchronized by serum shock, 229 genes were found to be upregulated by Bmal1/CLOCK, suggesting heterogeneity among endothelia²⁵. Isolation of a single cell type might, however, be expected to yield a lower proportion of cycling genes, given that genes expressed rhythmically vary from cell type to cell type and tissue analysis likely represents the sum total of rhythmic genes across several cell types. This is supported by recent single-cell RNA-seq of SCN cells, which revealed cycling of an average of 5% of transcripts in any given cell type²⁶, although a higher proportion of the transcriptome cycles in whole tissue SCN¹³. Nonetheless, the number of cycling transcripts we observed is strikingly low given the robustness of the cycling in BBB function (see Figs. 1 and 2). Expression of core clock genes cycled in controls but not in mutants, revealing that the brain endothelium has an autonomous clock (Fig. 3b). Clock output genes also cycled (Supplementary Fig. 3c) in control brain endothelial cells although some were not represented in the 45 cycling genes suggesting that the algorithm used underestimates the number of clock-dependent genes. This could be due to the overall low number of sequence reads and/or heterogeneity of circadian expression within brain endothelial cells.

**Fig. 3: Cycling of circadian clock genes and a highly expressed magnesium transporter in brain endothelial cells.**

We did not observe circadian regulation of brain endothelium-specific genes. Several genes which exhibited robust cycling in the controls, such as the amino acid transporter Slc7a5a, were also rhythmic in the mutants suggesting that rhythms in other tissues can compensate or that the rhythms are dependent on oscillatory factors like sleep or light. Since ABC transporters are known to regulate xenobiotic efflux, we examined the transcripts of ABC-family transporters and found no consistent phases over time (Supplementary Fig. 4a). Abcb1a, the transporter relevant for RHB efflux and most homologous to human ABCB1 is highly expressed in the BBB, but was not rhythmic or dependent on Bmal1 (Supplementary Fig. 4b). One ABC efflux transporter, Abcg2, was rhythmic by Meta2d, but it was not dependent on Bmal1 (Supplementary Fig. 4b). Together these results suggest that the circadian clock does not regulate ABC transporters at the level of transcription.

Previous reports have suggested that circadian clocks regulate intracellular magnesium concentrations²⁷. Since magnesium is a necessary cofactor in the regulation of ABC transporters, we examined whether its regulation was circadian clock-dependent. We evaluated the RNA-seq results in control animals for all the known magnesium transporters and found Transient receptor potential cation channel, subfamily M, member 7 (Trpm7) to be highly expressed (Fig. 3c). Trpm7 transcripts in control animals trended toward having a rhythm, although this was not statistically significant (Supplementary Fig. 4c). We then sorted BBB from WT mice, extracted RNA, and used real-time PCR to quantify the level of Trpm7 transcripts, verifying that the mRNA levels are indeed rhythmic (Fig. 3d). This suggested the possibility that the clock regulates intracellular magnesium homeostasis to affect efflux.

Oscillations of intracellular magnesium levels likely regulate efflux in a cultured human brain endothelial cell (BEC) line

We were unable to measure magnesium levels in the mouse BBB because tissue penetrance of magnesium indicators is poor; thus, we used an immortalized human cerebral microvascular endothelial (hCMEC/D3) cell line²⁸. First, we validated that the hCMEC/D3 cell line has a circadian rhythm by transducing the cell line with PER2-luciferase and measuring luminescence over several days. We find that the cells have a period of ~23 h (Fig. 4a). To determine whether circadian clocks in the cell lines produce oscillations in efflux, we treated hCMEC/D3 cells with dexamethasone to synchronize their endogenous circadian clocks²⁹. Twelve to forty-eight hours after synchronization, cells were incubated with the xenobiotic rhodamine 123 (RH123), a previously validated ABCB1 substrate of hCMEC/D3 cells³⁰. Fluorescence retained in the cells after 30 min at 37 ˚C compared to fluorescence retained in cells on ice was measured by flow cytometry. Increased retention of fluorescence by incubation with verapamil indicates that the fluorescence loss reflects ABCB1-mediated efflux of RH123 from the hCMEC/D3 (Fig. 4b). The amount of RH123 efflux by hCMEC/D3 cells oscillates over time and the rhythm is lost when efflux is inhibited with verapamil (Fig. 4c).

**Fig. 4: Cyclic efflux in human BECs is likely driven by intracellular magnesium oscillations.**

To measure the level of intracellular magnesium over time, we synchronized the hCMEC/D3 culture and used a ratiometric intracellular magnesium indicator Magfura2-AM. We verified that the indicator was sensitive to the level of intracellular magnesium in the hCMEC/D3 culture by using EDTA-AM to chelate magnesium (Fig. 4d), and then measured magnesium at different times following dexamethasone synchronization. The level of intracellular magnesium cycles roughly in phase with the cycle of efflux (Fig. 4e). Because the amplitude of the change in magnesium levels was low in this assay, we also stably expressed a genetically encoded magnesium FRET sensor, MARIO (magnesium ratiometric indicator for optical imaging) in hCMEC/D3. We measured the ratio of YFP/CFP in the cells for up to 48 h following synchronization with dexamethasone and found that the level of intracellular magnesium was rhythmic (Supplementary Fig. 5a). Use of a magnesium chelator (EDTA-am) verified that the FRET sensor was sensitive to intracellular magnesium, although the sensitivity was much lower than that of Magfura2-AM (Supplementary Fig. 5b). To determine whether magnesium levels affect efflux, we incubated cells with EDTA-am and/or verapamil and performed a RH123 efflux assay. We found similar reduction of efflux if either drug or both were applied to the cells, suggesting that changing concentrations of intracellular magnesium affect the activity of the ABCB1 transporter (Fig. 4f). Together these results suggest that oscillations of intracellular magnesium levels regulate the cycling of efflux activity.

Magnesium transporter TRPM7 is a direct target of the circadian clock

Our next objective was to understand how the circadian clock affects levels of intracellular magnesium. To determine whether the clock regulates magnesium transporters, we examined transcript levels in hCMEC/D3 cells following dexamethasone synchronization. We found both BMAL1 and TRPM7 transcripts cycle relative to tubulin (Fig. 5a, b). Because the acrophase of BMAL1 preceded the acrophase of TRPM7, we assessed whether BMAL1 directly binds the TRPM7 promoter by using an anti-BMAL1 antibody for chromatin immunoprecipitation (ChIP). Using the Eukaryotic Promoter Database to identify putative BMAL1-binding e-box sites in the TRPM7 promoter, we then used qPCR to determine the fold enrichment of BMAL1-binding relative to that of IgG controls. The increase in BMAL1 binding to the TRPM7 e-box is comparable to the level of BMAL1 binding to a known e-box in the PER2 promoter (Fig. 5c). ChIP-seq data sets have also identified the promoter region of TRPM7 to be a target of BMAL1^31,32. By contrast, ABCB1 transcript remained at comparable levels over a 24-hour period (Fig. 5d) and negligible enrichment of BMAL1 binding was found for two putative e-boxes on ABCB1 promoter (Supplementary Fig. 6). Together these results are consistent with the idea that TRPM7, but not ABCB1, is a direct target of the molecular clock.

**Fig. 5: *TRPM7* is a clock-controlled gene in human brain endothelial cell culture.**

We then determined whether the transcriptional oscillation of TRPM7 translates to oscillations in protein levels. Using an antibody against TRPM7 to assess protein expression by immunoblot following dexamethasone synchronization, we found that, indeed, protein levels of TRPM7 oscillate in phase with the mRNA levels (Fig. 5e). Immunoblot of ABCB1 protein shows that levels do not cycle (Fig. 5f), suggesting that the activity of ABCB1 is independent of the level of protein.

To directly manipulate the level of TRPM7 in the hCMEC/D3 line, we created stable cell lines using lentiviral vectors containing GFP and scrambled siRNA or siRNA targeting TRPM7. Interestingly, siRNA knockdown of TRPM7 resulted in >50% cell loss as well as lower copy numbers (reduced GFP expression), suggesting that TRPM7 is required for survival (Supplementary Fig. 7a). The resultant siTRPM7 cell line had 4-fold reduction in TRPM7 transcript levels (Supplementary Fig. 7b), which was sufficient to reduce intercellular free magnesium and abrogate the 24-h oscillation (Fig. 6a). Because of the presence of GFP in our stable cell lines, we could not use RH123 and thus used RHB for ABCB1-dependent efflux experiments. Consistent with the intracellular magnesium levels, the siTRPM7 cell line also showed lower levels and loss of oscillation of RHB efflux compared to the control cell line (Fig. 6b). These results demonstrate that TRPM7 can regulate the levels of intracellular magnesium and ABCB1-mediated efflux.

**Fig. 6: Knockdown of *TRPM7* reduces and dampens rhythms of intracellular Mg²⁺ and xenobiotic efflux.**

Discussion

We report circadian regulation of xenobiotic efflux in both mouse BBB and human cultured brain endothelial cells. The circadian regulation of BBB efflux explains previous findings in which levels of an ABCB1 substrate, quinidine, in the rat brain depended on the timing of administration³³. The molecular clock in the BBB may also serve to regulate brain levels of endogenous hormones and cytokines through the activity of transporters. Previous reports show concentrations of leptin, interleukin-6, and tumor necrosis factor α have different phases between blood and brain^34,35,36. Together, a growing body of work suggests that the BBB is not only physically restrictive, but also temporally restrictive.

A seemingly simple mechanism of clock-control of BBB efflux would be to directly regulate the cycling of transporter transcripts; however, we did not observe a consistent phase among the ABC transporter transcripts. In fact, very few transcripts are clock-controlled genes in the mouse brain endothelium, which is surprising given the extent of circadian regulation of the transcriptome in other tissues²⁴ as well as previous reports of cycling ABC-family transporters in the gut^37,38 and liver³⁹. Yet the impact of the clock on cell physiology in the BBB is dramatic. We find the clock regulates intracellular Mg²⁺ levels, which may affect cell physiology more broadly than transcriptional changes. For instance, changing Mg²⁺ or translational machinery may affect the activity or levels of many proteins; however, to achieve the same effect at the transcriptional level, clock proteins would need to bind and regulate each individual transcript.

Our results suggest that the efficiency of xenobiotic efflux of the BBB is regulated, at least in part, by oscillating free magnesium levels. This is consistent with previous reports of oscillating free magnesium concentrations and magnesium transporters in other cell types^27,40. Because Mg²⁺ is a cofactor for hundreds of MgATP-dependent enzymes, even modest changes in cytosolic Mg²⁺ may serve to regulate cellular energy expenditure over the course of a day⁴¹. Indeed, we find that inhibition of intracellular Mg²⁺ in the range of the daily variation can reduce efflux. In the BBB, such changes in intracellular free magnesium may also regulate the activity of other BBB transporters, such as those for amino acids and glucose transporters, if Mg²⁺ is rate-limiting for transport. In addition to magnesium transporter, changes in intracellular calcium may contribute to oscillations of Mg²⁺.

Together, our data suggest a model in which BMAL1 in the BBB endothelium drives transcript and protein expression of TRPM7 during active periods (Fig. 7). The increase in TRPM7 channel likely results in higher concentrations of intracellular free magnesium, which increases the activity of the ABCB1 transporter and thereby the level of xenobiotic efflux, reducing the amount of RHB in the brain. During resting periods, BMAL1 is low, reducing the level of TRPM7 and subsequently reducing the level of free magnesium in the BBB. Lower levels of intracellular magnesium decrease xenobiotic efflux, resulting in greater brain retention of RHB. In support of this model, loss of the CRY proteins, which are negative regulators of BMAL1, results in lower retention of RHB. However, we cannot exclude the possibility that other factors, such as the recently proposed regulation of the BBB by neuronal activity, contribute to rhythms in efflux⁴².

**Fig. 7: Model of clock regulation of xenobiotic efflux at the BBB.**

Our results show that accumulation of xenobiotics in the brain following intravenous injection of mice occurs with an inverted cycle relative to that of a previously described efflux cycle in Drosophila⁴³. This aligns with the rest:activity rhythms of the organism with highest efflux during the active period and lowest efflux during the resting period. High levels of efflux would be most advantageous during periods of activity because of the presumed increase in exposure to toxins from foraging or injury. Despite the evolutionary conservation of circadian regulation of xenobiotic efflux, we find differences in mechanism between flies and mammals. In flies, the circadian clock uses a non-cell autonomous mechanism to regulate efflux activity of the BBB; however, in mammals the endothelial BBB-clock drives cycling of efflux within the same cell type. Interestingly, though, both systems include an important role for magnesium. In flies, rhythmic expression of gap junctions restricts the passage of magnesium between two layers of the BBB to specific times of day; at these times, magnesium is depleted from the subperineurial layer, which contains efflux transporters and normally has very high magnesium, thereby resulting in a rhythm of magnesium⁴³. In mammals, on the other hand, our data suggest that magnesium rhythms are driven by cell-autonomous clock control of a magnesium transporter.

The cycling of ABCB1 activity may have implications in disease settings; for instance in Alzheimer’s disease, where ABCB1 contributes to the clearance of amyloid beta from extracellular spaces of the brain^44,45. Typically, amyloid beta is either broken down in the brain or cleared through the BBB endothelium or the cerebrospinal fluid. Interestingly, amyloid beta clearance is higher during sleep⁴⁶, whereas here we show the peak activity of ABCB1 during active hours. However, efficient clearance of amyloid beta through the endothelium requires both ABCB1 and LRP1 for phosphatidylinositol binding clathrin assembly protein (PICALM)-mediated transcytosis through BBB endothelium⁴⁷. Despite lower activity of ABCB1, it is possible that transcytosis is higher during sleep, as shown for endocytosis in the Drosophila BBB⁴⁸. Thus, increasing ABCB1 activity at this time could facilitate amyloid beta clearance.

A therapeutic application of this work is the possibility of improving drug delivery to the central nervous system (CNS). Given that many drug targets are clock-controlled genes, cells and tissues have different drug sensitivities throughout the circadian day⁴⁹. Utilizing circadian timing in medicine, also known as chronotherapy or chronomedicine, can improve outcomes and decrease side effects due to lower required doses to achieve the same efficacy⁵⁰. Here, we show that ABCB1 substrates are subject to circadian regulation at the BBB, indicating an additional level of circadian control for many drugs targeted to the CNS; however, other mechanisms of oscillatory regulation could also influence drug concentrations in the brain. Sleep has been shown to increase removal of metabolites from the brain⁴⁶, and robust circadian rhythms in the choroid plexus increase cerebrospinal fluid (CSF) production and turnover during the resting period^51,52. These mechanisms may allow brain retention of other classes of drugs in a time-of-day-dependent matter, which could inform decisions on optimal chronotherapy. Our results suggest that optimizing circadian timing of drug delivery in regards to BBB permissiveness can achieve increased CNS drug efficacy and should be taken into consideration in developing therapeutic regimens.

Methods

Animal care and husbandry

C57BL/6J (#000664), Bmal1^fl/fl (B6.129S4(Cg)-Arntl^tm1Weit/J #007668)⁵³, and Tie2cre (B6.Cg-Tg(Tek-cre)1Ywa/J, #008863)⁵⁴ mice were purchased from Jackson Laboratory animal facility. Cry1,Cry2 double-deficient mice were a gift from Katja Lamia originally from Aziz Sancar⁵⁵. Animals were fed ad libitum and entrained to 12 h:12 h light:dark cycles and given at least 1 day per hour of phase shift prior to experiments. All live animal experiments were performed according to protocols approved by the Institutional Animal Care and Use Committee of the University of Pennsylvania (Philadelphia, PA) in accordance with guidelines set by the NIH (Protocol #806387). Male and female mice were used for gene expression data (RNAseq and qPCR). Only male mice were used in functional experiments due to the decreased risk for surgical failures with their larger size. Mice from the same litters were randomly distributed to environmental cabinets for entrainment for different circadian time points.

RHB permeability assay in live mice

Mice between 8 and 12 weeks were used for permeability experiments. The surgeon was blinded to the genotype and circadian time of the mice. Mice were anesthesized with continuous isoflurane to maintain surgical plane and given meloxicam (5 mg/kg) and bupivacaine (0.1 mg/kg) subcutaneously. RHB (Sigma) was dissolved in 10% water and 90% PBS and was injected (5 mg/kg) into the jugular vein. Tariquidar (Selleck Chemicals) was suspended in 30% Propylene glycol and 5% Tween 80 and 65% of 5% dextrose in water and coinjected (30 mg/kg) into the jugular vein. Animals that did not receive a proper injection were sacrificed and excluded from analysis (~5–10% of the animals). The mice were allowed 90 min to recover prior to tissue harvest. Animals were anesthetized with isoflurane and blood was collected by retro-orbital bleed and serum was extracted after <10 min of coagulation. Mice were then sacrificed by cervical dislocation and remaining blood was removed through cardiac perfusion of 1x PBS before brains were harvested. Brains were sliced into halves and homogenized in equal volume by weight PBS using a steel ball and TissueLyser II (Qiagen) set to 25 Hz for 2 min. RHB fluorescence was measured with Victor3V Plate Reader (Perkin Elmer) at Excitation/Emission ex540/em590nm. Amount of RHB was determined by standard curve and ratios of brain to blood levels were calculated for each animal.

Vascular leakage assay

Evans Blue vascular leakage assay was adapted from previous studies⁵⁶. Briefly, mice were anethesized with isoflurance and administered 50 mg/kg Evans Blue via retro-orbital injection. After 30 min, mice were sacrificed and perfused with 10 units/mL heparin in PBS. Brain and liver were harvested and Evans Blue was extracted with formamide and the absorbance of the supernatant was measured at 620 nm using a spectrophotometer (Cytation 5, BioTek). Concentration of Evans Blue was determined using standard curve. For induction of vascular leak, 3 mg/kg LPS was administered to the mice by intraperitoneal injection 24-h prior to injection of Evans blue.

Isolation of BBB endothelium

Brains were harvested from mice at time points around the circadian day (ZT2, 6, 10, 14, 18, 22). Brains were finely minced in RPMI media containing 10% CCS, incubated at 37 °C with 2 mg/mL collagenase IV (Gibco) and 200 μg/mL DNase I (Roche) for 30 min with shaking at 250 rpm, pipetting up and down once during incubation. Brain homogenate was demyelinated using myelin removal beads II (Miltenyi Biotec) as instructed by the manufacturer’s protocol. Cell suspensions were filtered using a 100 μm cell strainer. Cells were labeled with anti-CD90 (clone M5/49.4.1, BioXell) for 15 min on ice and incubated with BioMag Goat Anti-Rat IgG (Qiagen) for 30 min with rotation, and magnetically separated (Easy Sep). Supernatant was removed and spun down at 600 × g. Cells were resuspended in 250 μl, labeled with fluorescence-conjugated anti-CD31 (390, BioLegend), anti-CD90 (53-2.1, BioLegend) antibodies sorted using either a BD FACSAria or BD FACSMelody (BD Biosciences). Dead cells were excluded through 4′,6-diamidino-2-phenylindole uptake. Doublets were excluded through FSC-H by FSC-W and SSC-H by SSC-W parameters. Between 10,000 and 20,000 cells were sorted from a single mouse at >95% purity verified by post-sort analysis. Processing of tissue and BBB isolation takes ~4 h. Data from sort were analyzed using FlowJo V10.6 software (TreeStar).

RNA sequencing and analysis

RNA was extracted from sorted cells using RNeasy Plus Micro Kit (Qiagen) following the manufacturer’s protocol. RNA quality was assessed using 2100 Bioanalyzer Instrument (Agilent). RNA integrity numbers (RIN) of 6–8 were used to construct libraries. Strand-specific libraries were constructed using the SMARTer Stranded Total RNA-Seq Kit v2 - Pico Input Mammalian (Clontech) and sequenced across three lanes of a HiSeq 4000 (Illumina), generating 150 bp paired-end reads. All of the following analyses used the GRCm38 build of the mouse reference genome and gene models from Ensembl v91, both downloaded from the Ensembl portal⁵⁷. Reads were aligned to the reference genome using STAR v2.5.3a²³ with the following command line options: --outSAMtype BAM Unsorted; --outSAMunmapped Within KeepPairs; --outFilterMismatchNmax 33; --seedSearchStartLmax 33; --alignSJoverhangMin 8. In addition, STAR was provided with the gene models in GTF format during alignments and the “—outSAMattrRGline” was used to add a unique read group to each of the resulting BAM files (required to merge files below). Note that the three pairs of FASTQ files per library (two for each sequencing lane) were aligned separately. Following alignment, the three BAM files for each library were combined using the SAMtools v1.3.1⁵⁸ merge command with the “-n” argument, which maintains the BAM files in “unsorted” order (i.e., sorted by read ID). The Pipeline Of RNA-seq Transformations (PORT) v0.8.5-beta (https://github.com/itmat/Normalization) was used to perform gene-level normalization and quantification of the aligned data. Cycling transcripts with 24 h rhythmicity were identified using the meta2d function from the MetaCycle v1.1.0R package⁵⁹. The meta2d function was run with the following arguments: outIntegration = “both”, adjustPhase = “predictedPer”, combinePvalue = “fisher”, weightedPerPha = FALSE, cycMethod = c(“JTK”, “LS”), analysisStrategy = “selfUSE”, minper = 24, maxper = 24. For the MetaCycle analyses, we applied a minimum expression filter to remove unexpressed genes. For a gene to be included, it must have a normalized read count of five or greater in at least control six samples, or six Bmal1 mutant samples.

hCMEC/D3 cell cultures

Immortalized human brain capillary endothelial cells (Millipore SCC066) were cultured in EndoGRO-MV Complete media kit (SCME004) as recommended by the manufacturer. Cells were tested every 3–5 passages for mycoplasma contamination by DNA Hoechst stain. Cultureware was coated with rat tail collagen type I solution at a concentration of 0.1 mg/mL and was incubated for >1 h at 25 °C. Cells were cultured in an incubator at 37 °C with 5% CO₂. For synchronization of culture, cells were grown to ~60% confluence and incubated in 200 nM dexamethasone for 30 min. Circadian time points were obtained by staggering the synchronization of different cultures and assaying them at the same time.

Generation of stable cell lines

hCMEC/D3 expressing Per2-dLuc-GFP

Stable hCMEC/D3 cell lines expressing Per2-dLuc-eGFP were generated via lentiviral vector transduction as previously described⁶⁰. Briefly, LentiX 293T cells (Clontech) were grown to 70% confluence in 10 cm dishes, and co-transfected with pPer2-dLuc-eGFP, together with the packaging plasmids (pDVPR8.1 and pVSV-G, Addgene) in a 10:1:0.5 ratio, using Lipofectamine 3000 PLUS (Life Tech) according to the manufacturer’s directions. Twenty-four hours following transduction, the >50% transfection efficiency was confirmed via fluorescent microscopy to assess GFP positivity, and the media was exchanged. Forty-eight hours following transduction, the supernatant was collected from the transfected 293T cells, and the supernatant was centrifuged at 1000 × g for 5 min to pellet any 293T cells. Finally, stable hCMEC/D3 cell lines were generated by infecting 70% confluent cultures with the recombinant lentiviral vectors with 10 mg/mL polybrene (Sigma-Aldrich) twice over 2 consecutive days, and then individual GFP+ cells were sorted on the FACSMelody (BD Biosciences).

hCMEC/D3 expressing siRNA against TRPM7

Stable hCMEC/D3 cell lines expressing either scrambled siRNA (control) or TRPM7 siRNA with GFP were generated via lentiviral vector transduction using commercially available lentiviruses (LVP015-G and iV026186, abm). GFP+ cells were sorted in bulk on the FACSMelody (BD Biosciences). Due to the amount of cell death (50%) visually observed in the TRPM7 siRNA infected cells, we opted to group the cells to avoid mutations that may be present in individual clones.

Lumicycle assay

Cells containing Per2-dLuc-GFP were seeded in 35 mm dishes at 80% confluence. A 30 min 100 nM dexamethasone (Sigma) pulse was used to synchronize the cells. Real-time bioluminescence of the cells were monitored in LumiCycle recording media (hCMEC/D3 media with 10% FBS, 25 mM Hepes, 4.2 mM sodium bicarbonate pH 7.0) with 200 µM beetle luciferin potassium salt using LumiCycle luminometer (Actimetrics) as previously described⁶⁰. Luminescence data were analyzed using LumiCycle software (Actimetrics).

Cell culture efflux assay

For measuring efflux of hCMEC/D3 cells, an assay was adapted from Löscher Lab⁶¹. Cells were cultured and synchronized as described above. Cells were preloaded with 10 µM RH123 (Sigma-Aldrich) in Opti-MEM (Invitrogen) for 15 min on ice. Cells were washed twice to remove excess RH123 and resuspended in Opti-MEM without RH123. Half of the cells were incubated at 37 °C with 5% CO₂ and the decay of intracellular florescence was measured after 30 min. To control for the amount of RH123 preloaded, half of the culture was harvested and remained on ice. Fluorescence was measured using a BD FACSCanto II (BD Biosciences). For inhibition of efflux, 200 µM verapamil was added to the cells during the 15 min of RH123 preloading and the 30 min of incubation. For inhibition of magnesium, 10 µM of EDTA-am (Setareh Biotech) was added to the cells for the duration of 37 °C incubation.

Measuring magnesium in cell culture

Indicator (MagFura2)

Cells were plated in a 96-well plate 2–3 days prior to the assay. hCMEC/D3 media was removed and cells were incubated in 150 μl Opti-MEM for 10 min (Gibco). Magfura2-AM was suspended in DMSO to make a 5 mM stock solution. A dispersion solution was made by adding 4 μl stock solution and 5 μl of 20% (w/v) Pluronic F-127 (Life Tech) to 1 mL Opti-MEM. Fifty microliters of dispersion solution was added to the culture for 30 min at 37 °C. Cells were washed 2× with PBS and incubated in Opti-MEM for 30 min. Magfura2 was measured at ex330/em490nm and ex369/em490nm using a Cytation plate reader.

Genetic sensor (MARIO)

A plasmid encoding a magnesium FRET sensor, MARIO (magnesium ratiometric indicator for optical imaging) was obtained from Dr. Takeharu Nagai⁶². MARIO was cloned into pBABE retroviral vector using BamHI/EcoRI. pBABE-MARIO was transfected into ΦNIX cells using Lipofectamine 3000 PLUS (Life Tech). Twenty-four hours following transfection, the media was replaced. Forty-eight hours following transfection, the supernatant was collected and applied with 10 mg/mL polybrene to a culture of hCMEC/D3. Culture was sorted for YFP+ cells using a FACSMelody (BD Biosciences). The ratio of YFP (ex440/em528nm) and CFP (ex440/em480nm) were measured with a fluorescent plate reader.

Chromatin immunoprecipitation

Chromatin immunoprecipitation of BMAL1 protocol was adapted from previous report⁶³. 1 × 10⁶ cells were fixed with 1% formalin for 10 min. 125 mM glycine was added to stop the cross-linking reaction. Cells were washed with PBS and collected by scraping. Cell pellet was resuspended in swelling buffer (5 mM PIPES pH 8.0, 85 mM KCl, 1% NP40, and protease inhibitor cocktail), homogenized, and collected by centrifugation. Nuclei were resuspended in nuclear lysis buffer (50 mM Tris-HCl pH 8.0, 10 mM EDTA, 1% SDS, and protease inhibitor cocktail) and sonicated to an average length of about 300–500 base pairs, which was confirmed with agarose gel electrophoresis. Samples were diluted 10-fold with IP dilution buffer (16.7 mM Tris-HCl pH 8.0, 0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 167 mM NaCl, and protease inhibitor cocktail) and incubated with anti-BMAL1 antibody (5 μg, ab3350, Abcam) or IgG control antibody overnight at 4 °C. DNA complexes were collected on Dynabeads Protein A and serial washed with dialysis buffer (50 mM Tris-HCl pH 8.0, 2 mM EDTA, and 0.2% sarkosyl) and IP wash buffer (100 mM Tris-HCl pH 9.0, 500 mM LiCl, 1% NP40, 1% Deoxycholate, and protease inhibitor cocktail). Samples were removed from beads using elution buffer (50 mM NaHCO₃ and 1% SDS). Cross-linking was reversed by overnight incubation with 0.2 M NaCl at 67 °C and samples were then RNase A treated and purified. Real-time PCR was performed using PER2 and ABCB1 primers. Percentage of input was calculated and normalized to IgG.

Real-time PCR

RNA was extracted using RNeasy mini kit (Qiagen) and reverse transcribed to cDNA using random hexamers and Superscript II (Invitrogen). Real-time polymerase chain reaction (PCR) was performed using Sybr Green PCR Master Mix (Applied Biosystems) with the oligonucleotides described in Supplementary Table 4. Assays were run on ViiA7 Real-Time PCR system (Applied Biosystems). Relative gene expression was calculated using the ΔΔCt method normalizing to tubulin.

Immunoblot

hCMEC/D3 cells were harvested at the indicated time points following dexamethasone synchronization, and lysed with 1x Passive Lysis Buffer (Promega) on ice. Lysates were boiled for 10 min and run on 4–12% gradient Tris-Glycine gels (Life Technologies), transferred to PVDF membranes, and probed with the following antibodies: anti-PGP (1:1000, MDR-1 c219, ThermoFisher), anti-TRPM7 (1:500, EPR4582, Abcam), anti-BMAL1 (1:1000, A302-616A, Bethyl Laboratories), and anti-beta-Actin (1:20,000, mAbcam8224, Abcam). Blots were digitalized with a photo scanner (Canon) and bands were quantified using ImageJ software.

Statistical analysis

Circadian statistical analysis was performed in R using JTK_CYCLEv3.1⁶⁴. Sample size estimates for mouse experiments were based on previous circadian studies; no fewer than n = 12 at 6 time points per 24 h. Variance of data for each time point was similar. Student’s T-tests and ANOVAs were performed in Excel (Microsoft) and/or Prism (GraphPad). Determination of sample size and statistical power was performed with powerandsamplesize.com.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The data that support the findings of this study are available upon request. Source data are available for Figs. 1b–d, 2, 4c–f, 5, 6. The RNA-sequencing datasets have been deposited in GEO database under accession code GSE135874.

References

Abbott, N. J., Patabendige, A. A. K., Dolman, D. E. M., Yusof, S. R. & Begley, D. J. Structure and function of the blood–brain barrier. Neurobiol. Dis. 37, 13–25 (2010).
Article CAS PubMed Google Scholar
Pardridge, W. M. The blood-brain barrier: bottleneck in brain drug development. NeuroRx 2, 3–14 (2005).
Article PubMed PubMed Central Google Scholar
Abbott, N. J., Rönnbäck, L. & Hansson, E. Astrocyte-endothelial interactions at the blood-brain barrier. Nat. Rev. Neurosci. 7, 41–53 (2006).
Article CAS PubMed Google Scholar
Armulik, A. et al. Pericytes regulate the blood-brain barrier. Nature 468, 557–561 (2010).
Article ADS CAS PubMed Google Scholar
Cereijido, M., Valdés, J., Shoshani, L. & Contreras, R. G. Role of tight junctions in establishing and maintaining cell polarity. Annu. Rev. Physiol. 60, 161–177 (1998).
Article CAS PubMed Google Scholar
Schneeberger, E. E. & Lynch, R. D. The tight junction: a multifunctional complex. Am. J. Physiol. Cell Physiol. 286, C1213–C1228 (2004).
Cordon-Cardo, C. et al. Multidrug-resistance gene (P-glycoprotein) is expressed by endothelial cells at blood-brain barrier sites. Proc. Natl Acad. Sci. USA 86, 695–698 (1989).
Article ADS CAS PubMed PubMed Central Google Scholar
Schinkel, A. H. et al. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood-brain barrier and to increased sensitivity to drugs. Cell 77, 491–502 (1994).
Article CAS PubMed Google Scholar
Mohawk, J. A., Green, C. B. & Takahashi, J. S. Central and peripheral circadian clocks in mammals. Annu Rev. Neurosci. 35, 445–462 (2012).
Article CAS PubMed PubMed Central Google Scholar
Hastings, M. H., Reddy, A. B. & Maywood, E. S. A clockwork web: circadian timing in brain and periphery, in health and disease. Nat. Rev. Neurosci. 4, 649–661 (2003).
Article CAS PubMed Google Scholar
Yoo, S.-H. et al. PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc. Natl Acad. Sci. USA 101, 5339–5346 (2004).
Article ADS CAS PubMed PubMed Central Google Scholar
Yan, J., Wang, H., Liu, Y. & Shao, C. Analysis of gene regulatory networks in the mammalian circadian rhythm. PLoS Comput. Biol. 4, e1000193 (2008).
Article ADS MathSciNet PubMed PubMed Central CAS Google Scholar
Panda, S. et al. Coordinated transcription of key pathways in the mouse by the circadian clock. Cell 109, 307–320 (2002).
Article CAS PubMed Google Scholar
Lamia, K. A., Storch, K.-F. & Weitz, C. J. Physiological significance of a peripheral tissue circadian clock. Proc. Natl Acad. Sci. USA 105, 15172–15177 (2008).
Article ADS CAS PubMed PubMed Central Google Scholar
Dunlap, J. C. Molecular bases for circadian clocks. Cell 96, 271–290 (1999).
Article CAS PubMed Google Scholar
McCarthy, J. J. et al. Identification of the circadian transcriptome in adult mouse skeletal muscle. Physiol. Genomics 31, 86–95 (2007).
Article CAS PubMed Google Scholar
Storch, K. F. et al. Extensive and divergent circadian gene expression in liver and heart. Nature 417, 78–83 (2002).
Article ADS CAS PubMed Google Scholar
Löscher, W. & Potschka, H. Blood-brain barrier active efflux transporters: ATP-binding cassette gene family. NeuroRx 2, 86–98 (2005).
Article PubMed PubMed Central Google Scholar
Curtis, A. M. et al. Circadian variation of blood pressure and the vascular response to asynchronous stress. Proc. Natl Acad. Sci. USA 104, 3450–3455 (2007).
Article ADS CAS PubMed PubMed Central Google Scholar
Narasimamurthy, R. et al. Circadian clock protein cryptochrome regulates the expression of proinflammatory cytokines. Proc. Natl Acad. Sci. USA 109, 12662–12667 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Kalucka, J. et al. Single-cell transcriptome atlas of murine endothelial cells. Cell 180, 764–779 (2020).
Article CAS PubMed Google Scholar
Chen, M. B., Yang, A. C., Lehallier, B., Quake, S. R. & Wyss-Coray, T. Brain endothelial cells are exquisite sensors of age-related circulatory cues. https://doi.org/10.1016/j.celrep.2020.03.012 (2020).
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
Zhang, R., Lahens, N. F., Ballance, H. I., Hughes, M. E. & Hogenesch, J. B. A circadian gene expression atlas in mammals: Implications for biology and medicine. Proc. Natl Acad. Sci. USA 111, 16219–16224 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Takeda, N. et al. Thrombomodulin is a clock-controlled gene in vascular endothelial cells. J. Biol. Chem. 282, 32561–32567 (2007).
Article CAS PubMed Google Scholar
Wen, S. et al. Spatiotemporal single-cell analysis of gene expression in the mouse suprachiasmatic nucleus. Nat. Neurosci. 23, 456–467 (2020).
Article CAS PubMed Google Scholar
Feeney, K. A. et al. Daily magnesium fluxes regulate cellular timekeeping and energy balance. Nature 532, 375–379 (2016).
Article ADS CAS PubMed PubMed Central Google Scholar
Weksler, B. B. et al. Blood-brain barrier-specific properties of a human adult brain endothelial cell line. FASEB J. 19, 1872–1874 (2005).
Article CAS PubMed Google Scholar
Balsalobre, A. et al. Resetting of circadian time in peripheral tissues by glucocorticoid signaling. Science 289, 2344–2347 (2000).
Article ADS CAS PubMed Google Scholar
Tai, L. M. et al. Polarized P-glycoprotein expression by the immortalised human brain endothelial cell line, hCMEC/D3, restricts apical-to-basolateral permeability to rhodamine 123. Brain Res. 1292, 14–24 (2009).
Article ADS CAS PubMed Google Scholar
Koike, N. et al. Transcriptional architecture and chromatin landscape of the core circadian clock in mammals. Science 338, 349–354 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Beytebiere, J. R. et al. Tissue-specific BMAL1 cistromes reveal that rhythmic transcription is associated with rhythmic enhancer-enhancer interactions. Genes Dev. 33, 294–309 (2019).
Article CAS PubMed PubMed Central Google Scholar
Kervezee, L. et al. Diurnal variation in P-glycoprotein-mediated transport and cerebrospinal fluid turnover in the brain. AAPS J. 16, 1029–1037 (2014).
Article CAS PubMed PubMed Central Google Scholar
Pan, W. & Kastin, A. J. Diurnal variation of leptin entry from blood to brain involving partial saturation of the transport system. Life Sci. 68, 2705–2714 (2001).
Article CAS PubMed Google Scholar
Agorastos, A. et al. Circadian rhythmicity, variability and correlation of interleukin-6 levels in plasma and cerebrospinal fluid of healthy men. Psychoneuroendocrinology 44, 71–82 (2014).
Article CAS PubMed Google Scholar
Pan, W., Cornélissen, G., Halberg, F. & Kastin, A. J. Selected Contribution: Circadian rhythm of tumor necrosis factor-α uptake into mouse spinal cord. J. Appl. Physiol. 92, 1357–1362 (2002).
Article CAS PubMed Google Scholar
Stearns, A. T., Balakrishnan, A., Rhoads, D. B., Ashley, S. W. & Tavakkolizadeh, A. Diurnal rhythmicity in the transcription of jejunal drug transporters. J. Pharmacol. Sci. 108, 144–148 (2008).
Article CAS PubMed PubMed Central Google Scholar
Okyar, A. et al. Sex-, feeding-, and circadian time-dependency of P-glycoprotein expression and activity - implications for mechanistic pharmacokinetics modeling. Sci. Rep. 9, 10505 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Hughes, M. E. et al. Harmonics of circadian gene transcription in mammals. PLoS Genet. 5, e1000442 (2009).
Pizarro, A., Hayer, K., Lahens, N. F. & Hogenesch, J. B. CircaDB: a database of mammalian circadian gene expression profiles. Nucleic Acids Res. 41, D1009–D1013 (2012).
Article PubMed PubMed Central CAS Google Scholar
de Baaij, J. H. F., Hoenderop, J. G. J. & Bindels, R. J. M. Magnesium in man: implications for health and disease. Physiol. Rev. 95, 1–46 (2015).
Article PubMed CAS Google Scholar
Pulido, R. S. et al. Neuronal activity regulates blood-brain barrier efflux transport through endothelial circadian genes. Neuron https://doi.org/10.1016/j.neuron.2020.09.002 (2020).
Zhang, S. L., Yue, Z., Arnold, D. M., Artiushin, G. & Sehgal, A. A circadian clock in the blood-brain barrier regulates xenobiotic efflux. Cell 173, 130–139 (2018).
Article CAS PubMed PubMed Central Google Scholar
Lam, F. C. et al. beta-Amyloid efflux mediated by p-glycoprotein. J. Neurochem. 76, 1121–1128 (2001).
Article CAS PubMed Google Scholar
Cirrito, J. R. et al. P-glycoprotein deficiency at the blood-brain barrier increases amyloid- deposition in an Alzheimer disease mouse model. J. Clin. Invest. 115, 3285–3290 (2005).
Article CAS PubMed PubMed Central Google Scholar
Xie, L. et al. Sleep drives metabolite clearance from the adult brain. Science 342, 373–377 (2013).
Article ADS CAS PubMed Google Scholar
Storck, S. E. et al. The concerted amyloid-beta clearance of LRP1 and ABCB1/P-gp across the blood-brain barrier is linked by PICALM. Brain. Behav. Immun. 73, 21–33 (2018).
Article CAS PubMed PubMed Central Google Scholar
Artiushin, G., Zhang, S. L., Tricoire, H. & Sehgal, A. Endocytosis at the Drosophila blood–brain barrier as a function for sleep. Elife 7, e43326 (2018).
Levi, F. & Schibler, U. Circadian rhythms: mechanisms and therapeutic implications. Annu. Rev. Pharmacol. Toxicol. 47, 593–628 (2007).
Article CAS PubMed Google Scholar
Lévi, F., Zidani, R. & Misset, J. L. Randomised multicentre trial of chronotherapy with oxaliplatin, fluorouracil, and folinic acid in metastatic colorectal cancer. Lancet 350, 681–686 (1997).
Article PubMed Google Scholar
Nilsson, C. et al. Circadian variation in human cerebrospinal fluid production measured by magnetic resonance imaging. Am. J. Physiol. - Regul. Integr. Comp. Physiol. 262, R20–R24 (1992).
Myung, J. et al. The choroid plexus is an important circadian clock component. Nat. Commun. 9, 1–13 (2018).
Article ADS CAS Google Scholar
Storch, K.-F. & Weitz, C. J. Daily rhythms of food-anticipatory behavioral activity do not require the known circadian clock. Proc. Natl Acad. Sci. USA 106, 6808–6813 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Kisanuki, Y. Y. et al. Tie2-Cre transgenic mice: a new model for endothelial cell-lineage analysis in vivo. Dev. Biol. 230, 230–242 (2001).
Article CAS PubMed Google Scholar
Thresher, R. J. et al. Role of mouse cryptochrome blue-light photoreceptor in circadian photoresponses. Science 282, 1490–1494 (1998).
Article CAS PubMed Google Scholar
Radu, M. & Chernoff, J. An in vivo assay to test blood vessel permeability. J. Vis. Exp. https://doi.org/10.3791/50062 (2013).
Zerbino, D. R. et al. Ensembl 2018. Nucleic Acids Res. 46, D754–D761 (2018).
Article CAS PubMed Google Scholar
Li, H. et al. The sequence alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).
Article PubMed PubMed Central CAS Google Scholar
Wu, G., Anafi, R. C., Hughes, M. E., Kornacker, K. & Hogenesch, J. B. MetaCycle: an integrated R package to evaluate periodicity in large scale data. Bioinformatics 32, 3351–3353 (2016).
Article CAS PubMed PubMed Central Google Scholar
Ramanathan, C., Khan, S. K., Kathale, N. D., Xu, H. & Liu, A. C. Monitoring cell-autonomous circadian clock rhythms of gene expression using luciferase bioluminescence reporters. J. Vis. Exp. https://doi.org/10.3791/4234 (2012).
Noack, A. et al. Drug-induced trafficking of p-glycoprotein in human brain capillary endothelial cells as demonstrated by exposure to mitomycin C. PLoS ONE 9, e88154 (2014).
Article ADS PubMed PubMed Central CAS Google Scholar
Maeshima, K. et al. A transient rise in free Mg²⁺ ions released from ATP-Mg hydrolysis contributes to mitotic chromosome condensation. Curr. Biol. 28, 444–451.e6 (2018).
Article CAS PubMed Google Scholar
Miki, T., Matsumoto, T., Zhao, Z. & Lee, C. C. p53 regulates Period2 expression and the circadian clock. Nat. Commun. 4, 2444 (2013).
Article ADS PubMed CAS Google Scholar
Hughes, M. E., Hogenesch, J. B. & Kornacker, K. JTK_CYCLE: an efficient nonparametric algorithm for detecting rhythmic components in genome-scale data sets. J. Biol. Rhythm. 25, 372–380 (2010).
Article Google Scholar

Download references

Acknowledgements

This work was supported by grants from the NIH (1K99HL147212) (to S.L.Z.), R37NS048471 (to A.S.), and 5UL1TR000003 (to Institute for Translational Medicine and Therapeutics at University of Pennsylvania) and from the Howard Hughes Medical Institute (to A.S.). We thank the Next-Generation Sequencing Core at University of Pennsylvania for construction of libraries for RNA-sequencing, High Performance Computing at Penn Medicine (PMACS HPC; funded by 1S10OD012312-NIH) for cluster computing support.

Author information

Authors and Affiliations

Chronobiology and Sleep Institute, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA
Shirley L. Zhang, Zhifeng Yue, Denice M. Arnold, Peter P. Pakstis, Jessica E. Schwarz & Amita Sehgal
Howard Hughes Medical Institute, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA
Shirley L. Zhang, Zhifeng Yue & Amita Sehgal
Institute for Translational Medicine and Therapeutics (ITMAT), Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
Nicholas F. Lahens

Authors

Shirley L. Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Nicholas F. Lahens
View author publications
You can also search for this author in PubMed Google Scholar
Zhifeng Yue
View author publications
You can also search for this author in PubMed Google Scholar
Denice M. Arnold
View author publications
You can also search for this author in PubMed Google Scholar
Peter P. Pakstis
View author publications
You can also search for this author in PubMed Google Scholar
Jessica E. Schwarz
View author publications
You can also search for this author in PubMed Google Scholar
Amita Sehgal
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

S.L.Z. and A.S. planned and designed the project. S.L.Z. performed and directed experiments, processed and analyzed data, and wrote the manuscript. N.F.L. processed and analyzed RNA-seq dataset. Z.Y. performed mouse experiments and molecular biology experiments. D.M.A. performed and analyzed molecular biology experiments. P.P.P. and J.E.S. performed cell culture experiments. A.S. managed the project and edited the manuscript with contributions from all co-authors.

Corresponding authors

Correspondence to Shirley L. Zhang or Amita Sehgal.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks Karyn Esser and the other anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Description of Additional Supplementary Files

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Zhang, S.L., Lahens, N.F., Yue, Z. et al. A circadian clock regulates efflux by the blood-brain barrier in mice and human cells. Nat Commun 12, 617 (2021). https://doi.org/10.1038/s41467-020-20795-9

Download citation

Received: 23 August 2019
Accepted: 18 December 2020
Published: 27 January 2021
DOI: https://doi.org/10.1038/s41467-020-20795-9

This article is cited by

Zika Virus Infection Alters the Circadian Clock Expression in Human Neuronal Monolayer and Neurosphere Cultures
- Thaíse Yasmine Vasconcelos de Lima Cavalcanti
- Morganna Costa Lima
- Rodrigo Antonio Peliciari-Garcia
Cellular and Molecular Neurobiology (2024)
Microfluidic models of the neurovascular unit: a translational view
- Nienke R. Wevers
- Helga E. De Vries
Fluids and Barriers of the CNS (2023)
Circadian rhythms in the blood–brain barrier: impact on neurological disorders and stress responses
- Nicolette Schurhoff
- Michal Toborek
Molecular Brain (2023)
Ocular steroidome in human eyes and in eyes with complex central serous chorioretinopathy (CSCR)
- Marta Zola
- Elodie Bousquet
- Francine Behar-Cohen
Scientific Reports (2023)
Activation of aryl hydrocarbon receptor (AhR) in Alzheimer’s disease: role of tryptophan metabolites generated by gut host-microbiota
- Antero Salminen
Journal of Molecular Medicine (2023)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.