Chronic circadian misalignment accelerates immune senescence and abbreviates lifespan in mice

Modern society characterized by a 24/7 lifestyle leads to misalignment between environmental cycles and endogenous circadian rhythms. Persisting circadian misalignment leads to deleterious effects on health and healthspan. However, the underlying mechanism remains not fully understood. Here, we subjected adult, wild-type mice to distinct chronic jet-lag paradigms, which showed that long-term circadian misalignment induced significant early mortality. Non-biased RNA sequencing analysis using liver and kidney showed marked activation of gene regulatory pathways associated with the immune system and immune disease in both organs. In accordance, we observed enhanced steatohepatitis with infiltration of inflammatory cells. The investigation of senescence-associated immune cell subsets from the spleens and mesenteric lymph nodes revealed an increase in PD-1+CD44high CD4 T cells as well as CD95+GL7+ germinal center B cells, indicating that the long-term circadian misalignment exacerbates immune senescence and consequent chronic inflammation. Our results underscore immune homeostasis as a pivotal interventional target against clock-related disorders.

including the SCN obtained from Per2 Luc mice of long-term exposure to ADV and LD conditions, and analyzed the bioluminescence rhythms in both whole SCN and single-cell levels. Since it has been reported that the age of mice affects the synchronicity of intra-SCN neuronal rhythms 23,24 , we prepared similar aged mice to evaluate the SCN rhythm against the long-term circadian misalignment. For example, SCN of ADV-conditioned mice (110-, 90-, 65-week-old (wo)) were compared with SCN of LD mice (112-, 108-, 103-wo). Whole SCN-level bioluminescence imaging revealed severe impairment of the circadian oscillation in the SCN from all ADV-conditioned mice, whereas robust circadian bioluminescence oscillation persisted in the SCN of similar aged LD mice. ( Fig. 2A-C). In contrast to the study involving single phase shift in rats 25 , we did not observe the desynchrony in the phase relationship between ventrolateral and dorsomedial regions of SCN in the ADV-conditioned mice ( Fig. 2A), suggesting that the intra-SCN desynchronization here is due to the long-term shifting. Single-cell-level analysis revealed that each neuron exhibited apparent circadian oscillation-even in the ADV condition-but the phases of analyzed ADV cells were widely dispersed and the overall amplitudes were lower even in relatively younger (65-wo) mouse ( Fig. 2D-G), suggesting that intra-SCN desynchronization in mice with long-term exposure to the non-adjustive ADV perturbation of light-dark condition. These findings showed that the long-term ADV-conditioned mice displayed circadian desynchronization within the SCN, strongly suggesting a causal relationship between circadian disruption and early mortality.
Comprehensive analysis of the long-term circadian misalignment-conditioned peripheral organs. The failure of tuning the peripheral clocks via the autonomic nervous system and/or endocrine system, as a result of SCN desynchronization, has been shown to initiate various systemic physiological dysfunctions 2,26,27 . The liver is a central peripheral organ whose functions are regulated by the circadian system 26 . We therefore performed non-biased evaluation using polyA-selected RNA sequencing (RNA-seq) on liver tissues obtained from 85-week ADV-conditioned and LD-conditioned mice (aged 97 weeks) without apparent weakness or injury. Expression levels of the core clock genes are divergent, but comparable between LD-and ADV-conditioned mice (Fig. S2).
To interrogate cellular mechanisms, we performed KEGG pathway enrichment analysis between ADV and LD groups. Our analysis revealed a wide spectrum of alteration in functional pathways in the 85-week ADV-conditioned liver when compared with those of the LD-conditioned livers. Interestingly, however, in the ADV condition, immune diseases and inflammation-related pathways, such as rheumatoid arthritis and inflammatory bowel disease associated pathways, were markedly activated, while metabolic pathways of lipids and amino acids were extensively repressed (Fig. 3A).
We also performed KEGG pathway enrichment analysis of kidney gene expression between the groups (Fig. 3B). Interestingly, as in the liver, pathways associated with immune system and immune diseases in the ADV-conditioned kidney-unlike those in LD-conditioned kidneys-were also activated. These results provide evidence for a systemic enhancement of inflammatory response in 85-week ADV-conditioned mice.
We further investigated the alteration of gene regulatory networks in the liver induced by the ADV condition with transcription factor (TF) enrichment analysis (Fig. 4). Gene expression modules regulated under core circadian transcriptional factors-including BMAL1, CLOCK, PER1, PER2, CRY1, CRY2, and NR1D1-were  www.nature.com/scientificreports www.nature.com/scientificreports/ We next examined the histology of these liver tissues (Fig. 5). Consistent with gene expression analysis above, clusters of infiltrating inflammatory cells were enhanced in the ADV mice, indicating low-grade chronic inflammation (Fig. 5A,B). Furthermore, a significant reduction in the number of liver parenchymal cells surrounding the central veins in the ADV-conditioned mice when compared with those of the LD-conditioned mice suggested that the degeneration of liver parenchymal cells was enhanced by long-term ADV exposure (Fig. 5C,D). In ADV-conditioned mice, fat deposition in the liver was also increased, implying hepatic steatosis (Fig. 5E). www.nature.com/scientificreports www.nature.com/scientificreports/ These results show that low-grade chronic inflammation in liver parenchymal cells is caused by exposure to the long-term non-adjustive CJL condition, supporting the results from comprehensive gene regulatory network analysis.
At the molecular level, the expression of fibrosis-associated genes-alpha-actin-2 (Acta2), also known as alpha-smooth muscle actin (α-SMA), collagens (Col1a1, Col1a2, and Col3a1) and lumican (Lum) was upregulated in the ADV-conditioned livers, as detected by the comprehensive gene expression analysis using RNA-seq (Fig. 5F). Consistent with the histological analysis, these results suggest that long-term non-adjustive CJL-conditioning promotes low-grade chronic inflammation with fibrosis similar to chronic steatohepatitis. www.nature.com/scientificreports www.nature.com/scientificreports/ Long-term Circadian Misalignment conditioning accelerates immune senescence. In addition to the activated pathways associating with immune system and diseases, our recent pilot study has also indicated that ADV mice with early death often exhibited severe inflammation at their humane endpoints 22 . These findings raised the possibility that the pathophysiology of chronic circadian misalignment induced by the long-term ADV condition may be associated with immune dysfunction. To determine whether the chronic inflammation we observed correlates with immune senescence 28 , we next analyzed senescence-associated (SA) immune cell subsets in the spleen and mesenteric lymph nodes (mLNs) from middle-aged mice (aged 77 weeks) exposed to CJL conditions for 65 weeks. As mice age, frequencies of PD-1 + CD44 high CD4 T cells (SA-T cells) 29 , CD153 + SA-T cells 30 , follicular helper T cells (Tfh cells) 30 , and regulatory T cells (Treg cells) 31 increase. Although lymphocyte development was unchanged (Fig. S3), SA-T, CD153 + SA-T, Tfh, and Treg cells increased in the spleen (Fig. 6A,B), and SA-T and Tfh cells increased in mLNs in ADV-conditioned mice (Fig. 6C,D), compared with those of LD-conditioned mice. Next, we cultured mLN cells to differentiate into type 1 helper T (Th1), Th2, and Th17 cells. Differentiation of IL-17A + Th17 cells in ADV-conditioned mice was elevated, and that of IFNγ + Th1 cells tended towards elevation (but this trend was not significant), whereas that of IL-4 + Th2 cells was unchanged (Fig. 6E). These results indicate that chronic circadian misalignment accelerates T cell senescence and enrichment of Th1 and Th17 cells.
SA-T and Tfh cells in aged mice stimulate B cells and induce germinal center (GC) formation and class switch 30,32 ; we therefore examined GC B and class-switched B cells. In ADV-conditioned mice, the frequency and number of CD95 + GL7 + GC B cells increased, and those of class-switched IgG1 + B cells showed a trend to increase (p = 0.19 or 0.18) when compared with those of LD-conditioned mice (Fig. 6F,G). These results provide evidence for the activation of Tfh cells and the acceleration of GC formation and class switch as a result of chronic circadian misalignment. SA-T cells produce high amounts of osteopontin, causing inflammation and autoantibody production 29,30,32 , and Th1 and Th17 cells exacerbate autoimmunity 33 . Our results suggest that by enhancing generation of Th1 and Th17 cells, long-term exposure to non-adjustive CJL condition aggravates inflammation and autoimmunity. Consistently, KEGG pathway enrichment analysis in the livers and kidneys of 85-week ADV-conditioned mice indicated the activation of gene regulatory pathways associated with autoimmune diseases such as rheumatoid arthritis (Fig. 3).
Since circadian rhythms regulate innate immunity 34 and circadian disruption increases production of pro-inflammatory cytokines in myeloid cells after LPS stimulation 35 , we next analyzed myeloid cell subsets. Long-term non-adjustive ADV-conditioned mice showed increases in monocytes in the peripheral blood and of plasmacytoid dendritic cells (pDCs) and conventional dendritic cells (cDCs) in the spleen, while other myeloid subsets remained unchanged (Fig. S4A-C). In addition, the expression of inflammatory cytokines in cDCs and marginal zone (MZ) macrophages of the spleen remained unchanged (Fig. S4D), suggesting that the overall cytokine expression is increased in the spleen. These results demonstrate that long-term circadian misalignment causes a mild increase in myeloid cells in the spleen and blood. In conjunction with immune senescence, such as an increase in monocytes 36 , in the innate immunity 37 , our results of age-associated decline in adaptive immunity strongly support the acceleration of immune senescence and the induction of chronic inflammation by long-term circadian misalignment.

Discussion
In this reverse-translational study to understand the pathophysiological mechanisms for adverse consequences of chronic circadian impairment, we examined the effects of long-term (up to 85 weeks from 12 weeks of age) CJL using adult wild-type mice subjected to non-adjustive (ADV) and adjustive (DEL) conditions. The mouse age span in this study may be analogous to adolescence until retirement in humans 38 .
Importantly, our study revealed that the lifespan of mice exposed to ADV conditioning was significantly shorter than that of mice exposed to control LD or DEL conditions (Fig. 1). In ADV-conditioned mice, in accordance with recent reports, we observed severe desynchronization of circadian oscillation within the SCN 20,21 , whereas we detected no apparent desynchronization of SCN neuronal rhythms in the SCN slices of LD-conditioned mice of similar age (103~112-week-old) (Fig. 2). These observations suggest that intra-SCN desynchronization may contribute to the initiation of circadian misalignment-associated pathophysiological mechanisms. Recently, light-dark cycle entrains peripheral tissues including liver without functional SCN clock 39,40 , therefore other pathways than SCN-dependent axis may also contribute to the deleterious consequences associated with the long-term perturbation of light-dark cycle. In any case, RNA-seq data of liver and kidney tissues obtained from 85-week ADV-conditioned mice indicated that the immune function and inflammation-related pathways were markedly activated (Fig. 3). Moreover, transcriptional factor (TF) enrichment analysis using the RNA-seq data revealed that the gene modules regulated by core clock components such as CLOCK, BMAL1, PER1, PER2, CRY1 and CRY2 were widely suppressed in ADV-conditioned liver (Fig. 4). These findings strongly suggest that the environmental perturbation by non-adjustive shifted condition of light-dark cycle resulted in the "reprogramming" of the circadian gene regulatory networks in the peripheral tissues. Moreover, some immune system-related transcriptional factors were revealed in the TF enrichment analysis (Fig. 4). Decreased expression of Nfil3 may be related to the enrichment of Th17 in the intestine and the spleen 41 , consistent with the negative-regulation of NFIL3 activity in the ADV-conditioned mice compared with the LD-conditioned mice (Fig. 4). Furthermore, IRF8 activity appeared to be positively regulated in ADV-conditioned mice (Fig. 4). It has been reported that the increase in IRF8 promotes the increases in the myeloid cells including monocyte, DC, and pDC 42 . This is consistent with the elevated abundance of monocytes, pDCs, and cDCs in the peripheral blood or spleen of the ADV mice (Fig. S4A,C). These results demonstrate that immune network is markedly affected by the long-term non-adjustive CJL.

Scientific RepoRtS |
(2020) 10:2569 | https://doi.org/10.1038/s41598-020-59541-y www.nature.com/scientificreports www.nature.com/scientificreports/ in mice. A non-biased evaluation using comprehensive RNA-seq analysis to survey the physiological functions impaired by the circadian misalignment in the mouse liver revealed the activation of pathways associated with the immune system, in addition to suppression of metabolic pathways, as apparent pathophysiological consequences of exposure to ADV conditions for 85 weeks. Furthermore, we also observed the activation of immune system and immune disease-related pathways in 85-week ADV-conditioned kidneys (Fig. 3). Intriguingly, the subset analysis of immune cells in these ADV conditioned mice has shown the significant increases in SA-T, Tfh, Treg cells, and GC B cells (Fig. 6), which suggests that the acceleration of immune senescence and ageing-related changes in immune cell functions may play a critical role for establishing the pathophysiological consequence such as hepatic steatohepatitis.
Recent reports have indicated that immune system is broadly altered by the circadian dysregulation 45,46 . For example, circadian rhythms are closely linked with the sympathetic activity rhythm, and its dysregulation modulates immune functions [47][48][49][50] . In addition, endocrine dysregulation, often induced by jet-lag 2,26,27 , has been shown to be associated with immune senescence 51 . Moreover, since jet-lag induces dysbiosis in both mice and humans 52 , modification of gut microbiota may also contribute to the acceleration of immune senescence 53 . Together with these findings, our results suggest that the disruption of immune homeostasis induced by chronic circadian misalignment is a key mechanism underlying the systemic pathophysiological changes.
In addition to the enhancement of liver steatosis, we observed low-grade chronic inflammation with elevated fibrosis markers in the livers of ADV mice (Fig. 5). Recently, Kettner et al. reported that long-term CJL induced hepatocellular carcinogenesis in mice 48 . Although we failed to detect any macroscopically apparent hepatic tumors in our study, we observed common pathophysiological features -including chronic steatohepatitis -after long-term circadian misalignment. Since both immune senescence and low-grade chronic inflammation with elevated fibrosis markers are characteristic features of ageing 54 , these findings suggest that long-term circadian misalignment accelerates ageing via an accelerated immune senescence, ultimately aggravating chronic inflammation in mice (Fig. S5).
Chronic inflammation with immune senescence significantly increases risks of various chronic diseases such as metabolic syndrome, cardiovascular disease, inflammatory diseases, and cancers 55 . Our mouse model system of exposure to long-term non-adjustive shifted light conditions may mirror the pathophysiology of chronic circadian rhythm disruption in humans. In particular, our findings suggest that immune senescence and chronic low-grade inflammation are accelerated by long-term circadian perturbation via non-adjustive light-dark shift, leading to premature death. Therefore, immune homeostasis may be a crucial target for intervention against clock-associated diseases.

Methods
Animals and experimental conditions. Male C57BL/6J mice (10 weeks old) were purchased from SLC (Hamamatsu, Japan). The Per2 Luc knock-in mice were originally developed by Dr. Joseph Takahashi's group 14 and maintained in our facility. Mice were housed in plastic cages in light-shield mouse housing boxes (1800 × 360 × 520 mm) at a room temperature of 25.3 ± 0.3 °C with 55% humidity. The mice were acclimatized to the environment in the Animal Experimentation Center of the Kyoto Prefectural University of Medicine for 2 weeks. Mice were housed in groups (4-5 mice per cage) and light intensity was set at 200 lx within the cages. The mice were allowed ad libitum access to food and water, and were regularly subjected to observations to identify apparent abnormalities during captivity.
Chronic jet lag (CJL) conditions were designed based on our previous pilot study 22 . Mice acclimatized to the environment were divided into three groups of light-schedule conditions: LD-condition (light-dark condition with an 8:00-20:00 light period, n = 24); DEL-condition (8-h phase delay once every 7 days, n = 20); and ADV-condition (8-h phase advance once every 4 days, n = 20). Body weight of each mouse was measured once every two weeks during these CJL experimental periods.
For tissue sampling, blood was collected by cardiac puncture under deep terminal anesthesia by isoflurane and added to an equal volume of 0.5 mM EDTA/PBS. Then mice were decapitated and the livers or kidneys were collected and snap-frozen in liquid nitrogen for RNA-seq analysis or fixed in 10% neutral buffered formalin for histological analysis immediately. Mesenteric lymph nodes (mLNs), bone marrow from femora, spleens, and peripheral bloods were collected and used for flow cytometry analysis.
All experiments were approved by the Experimental Animals Committee, Kyoto Prefectural University of Medicine, and were performed in accordance with the institutional guidelines and Guidelines for Proper Conduct of Animal Experiments by the Science Council of Japan.
Behavioral analysis. The behavioral analysis of the mice exposed to the CJL conditions was performed as our previous report 22 . Each mouse was housed separately in a cage (170 × 350 × 145 mm) with a 120-mm diameter running wheel (SANKO, Osaka, Japan). The wheel-running frequency was measured by counting the number of signals from a magnet sensor (59070-010, Littelfuse Inc., Chicago, IL, USA). ClockLab software (Actimetrics, Wilmette, IL, USA) was used to analyze and display the behavioral activity in wheel revolutions per 5-minute bin.
SCN slice culture and bioluminescence recording. The SCN slice culture was carried out as previously reported 56 . Briefly, the 200-µm thickness of the SCN from male and female Per2 Luc knock-in mice was cultured in a recording medium [DMEM supplemented with 15 mM HEPES, 1.2 g/L NaHCO 3 , 20 mg/L kanamycin, 5 µg/ mL insulin, 100 µg/mL human transferrin, 100 µM putrescine, 20 nM progesterone, and 30 nM sodium selenite (Sigma)] on a culture membrane (Millicell-CM; Millipore). The bioluminescence images were acquired every hour with a 59-min exposure duration by using an integrated incubator-microscope system (LV100 and LV200; Olympus). The analysis was performed using AquaCosmos 2.6.3.6 software (Hamamatsu Photonics) and ImageJ 1.47 v software.

Statistical analysis and data analysis.
For the survival analysis, the Kaplan-Meier survival curves were constructed by R 3.4.3 software. The stratified Cox proportional hazard regression model analysis was performed by using SAS 9.4 software (SAS institute, Inc., Cary, NC, USA).
For fast Fourier transform (FFT) analysis, the raw data was detrended by subtracting a 24-h moving average and the relative spectral power density (relative power) at the peak within the range of 21-26 h was obtained by using Microsoft Excel 2010.
For calculating amplitudes of single-cell bioluminescence traces, the detrended data were used for the analysis. Cosine wave fitting was carried out with the following equation, using GraphPad Prism 6.0 software: where A: amplitude, T: period, and ϕ: acrophase.