Psychoneuroendocrinology (2014) 44, 71—82

Available online at www.sciencedirect.com

ScienceDirect j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / p s y n e u e n

Circadian rhythmicity, variability and correlation of interleukin-6 levels in plasma and cerebrospinal fluid of healthy men Agorastos Agorastos a,b, Richard L. Hauger a,c,d, Donald A. Barkauskas e, Tobias Moeller-Bertram a,c,d, Paul L. Clopton c, Uzair Haji c, James B. Lohr a,c,d, Thomas D. Geracioti Jr.f, Piyush M. Patel c,d, George P. Chrousos g, Dewleen G. Baker a,c,d,* a

VA Center of Excellence for Stress and Mental Health, San Diego, CA, USA Department of Psychiatry and Psychotherapy, University Medical Center Hamburg-Eppendorf, Hamburg, Germany c VA San Diego Healthcare System, San Diego, CA, USA d Department of Psychiatry, University of California, San Diego, CA, USA e Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA f Department of Psychiatry, University of Cincinnati, Cincinnati, OH, USA g First Department of Pediatrics, Athens University Medical School, Aghia Sophia Children’s Hospital, Athens, Greece b

Received 22 November 2013; received in revised form 26 February 2014; accepted 27 February 2014

KEYWORDS Interleukin-6 (IL-6); Cytokines; Circadian; Immune; Inflammation; Blood—brain-barrier (BBB); Cerebrospinal fluid (CSF); Sleep; Stress; Neuroimmunology

Summary Background: Interleukin-6 (IL-6) is a cytokine with pleiotropic actions in both the periphery of the body and the central nervous system (CNS). Altered IL-6 secretion has been associated with inflammatory dysregulation and several adverse health consequences. However, little is known about the physiological circadian characteristics and dynamic inter-correlation between circulating and CNS IL-6 levels in humans, or their significance. Methods: Simultaneous assessment of plasma and cerebrospinal fluid (CSF) IL-6 levels was performed hourly in 11 healthy male volunteers over 24 h, to characterize physiological IL-6 secretion levels in both compartments. Results: IL-6 levels showed considerable within- and between-subject variability in both plasma and CSF, with plasma/CSF ratios revealing consistently higher levels in the CSF. Both CSF and plasma IL-6 levels showed a distinctive circadian variation, with CSF IL-6 levels exhibiting a main

* Corresponding author at: Tel.: +1 858-552-8585. E-mail address: [email protected] (D.G. Baker). http://dx.doi.org/10.1016/j.psyneuen.2014.02.020 0306-4530/Published by Elsevier Ltd.

72

A. Agorastos et al. 24 h, and plasma a biphasic 12 h, circadian component. Plasma peaks were roughly at 4 p.m. and 4 a.m., while the CSF peak was at around 7 p.m. There was no correlation between coincident CSF and plasma IL-6 values, but evidence for significant correlations at a negative 7—8 h time lag. Conclusions: This study provides evidence in humans for a circadian IL-6 rhythm in CSF and confirms prior observations reporting a plasma biphasic circadian pattern. Our results indicate differential IL-6 regulation across the two compartments and are consistent with local production of IL-6 in the CNS. Possible physiological significance is discussed and implications for further research are highlighted. Published by Elsevier Ltd.

1. Introduction Interleukin-6 (IL-6) is a 184 amino acid cytokine discovered three decades ago as a protein involved in B-cell differentiation (Hirano et al., 1986). However, IL-6 is a far more versatile peptide, influencing numerous cell types and being responsible for a plethora of pro- and anti-inflammatory immune functions, as well as metabolic, endocrine, autocrine and paracrine effects in humans (Mihara et al., 2012; Papanicolaou and Vgontzas, 2000; Scheller et al., 2011; Spooren et al., 2011). The pleiotropic biology of IL-6 is mediated through a number of different signal transduction pathways (Heinrich et al., 2003; Rose-John and Heinrich, 1994). In the periphery, IL-6 is produced by immune or immuneaccessory cells, non-immune endocrine and endothelial cells, as well as white adipose tissue and skeletal muscle (MohamedAli et al., 1997; Vgontzas et al., 1999). Its expression is stimulated by sympathetic nervous system activity, catecholamine production, and other cytokines, while suppressed by glucocorticoids and estrogens (Mohamed-Ali et al., 2001; Papanicolaou et al., 1996a; Sanceau et al., 1995; Vicennati et al., 2002; Zhou et al., 1993). IL-6 plays a central role in immune responses by regulating acute-phase reactions, host defense mechanisms and haematopoiesis, angiogenesis, thyroid function, and the hepatic synthesis and secretion of Creactive protein (CRP) (Akira et al., 1993; Castell et al., 1989; Heinrich et al., 1990; Mihara et al., 2012). Furthermore, IL-6 stimulates osteoclastogenesis and intermediary metabolism and regulates lipid metabolism and endothelial function (Akira et al., 1993; Keller et al., 1996; Manolagas and Jilka, 1995; Stouthard et al., 1995). In the CNS, IL-6 is produced by microglial cells, astrocytes and neurons. Stimuli of IL-6 production are very diverse, including a broad spectrum of neurotransmitters, neuropeptides, cytokines, pathogens or neuronal depolarization (Gruol and Nelson, 1997; Spooren et al., 2011; Van Wagoner et al., 1999). CNS IL-6 is considered a neurotrophic factor targeting overall homeostasis and development of the nervous system (Spooren et al., 2011). CNS IL-6 plays a crucial role in the maintenance of blood—brain-barrier integrity and is a key player in mediating pleitropic effects on astrocytes and microglia (Fee et al., 2000; Gruol and Nelson, 1997; Marz et al., 1999; Spooren et al., 2011; Streit et al., 2000; Swartz et al., 2001; Tilgner et al., 2001). IL-6 may, thus, play an important role in the complex autonomic, psychoneuroendocrine and metabolic interplay of inflammation in the CNS (Helwig et al., 2008; Spooren et al., 2011). A growing body of research suggests that IL-6 functions as a stress biomarker. Elevated plasma IL-6 levels have

repeatedly been reported in response to physical and psychological stress and altered IL-6 secretion has been frequently associated with a variety of health consequences through stress-immune system dysregulation (Chourbaji et al., 2006; Chrousos, 1995; Elenkov et al., 2005; Steptoe et al., 2007; Sternberg, 2001). While acute stress-induced IL6 increases might be useful in maintaining homeostasis, a long-term IL-6 increase is indicative of chronic stress and unfavorable health outcomes (Hansel et al., 2010). IL-6, thus, appears to have a key role mediating a rapid interplay between immune system and CNS function. However, the relation between peripheral versus central IL-6 has not been thoroughly investigated. Prior study results suggested circadian and even ultradian variations of circulating IL-6 concentrations, with generally lower levels during daytime and higher during the night (Bauer et al., 1994; Crofford et al., 1997; Gudewill et al., 1992; Izawa et al., 2013b; Kanabrocki et al., 1999; Lissoni et al., 1998; Perry et al., 2009; Redwine et al., 2000; Sothern et al., 1995a; Vgontzas et al., 1999, 2002, 2005), while disturbed day—night cycles and sleep deprivation affect these rhythmic oscillations (Redwine et al., 2000; Vgontzas et al., 1999, 2002, 2003). However, less is known about the regulation of IL-6 release in the brain under physiological conditions, or about the dynamic interplay between plasma and CSF IL-6 levels in humans across a 24 h period. Some prior studies that have shown no significant correlation between CSF and plasma IL-6 levels have proposed differential regulation mechanisms of IL-6 in CNS and the periphery (Lindqvist et al., 2009; Stenlof et al., 2003), but this has been studied little. The main objective of the present study was the precise and simultaneous examination of sequential 24 h plasma and CSF IL-6 measurements, with a primary goal of defining IL-6 circadian secretion patterns and to elucidate relations, if any, between peripheral and CNS IL-6 concentrations in healthy male volunteers.

2. Methods 2.1. Subjects We collected data from 11 healthy male, U.S. civilian study volunteers, who participated in a serial CSF and plasma sampling study approved by the Institutional Review Board of the University of California, San Diego Medical Center and the Research Committee of the San Diego Veterans Affairs Medical Center. One of the volunteers had only a single CSF sample and was excluded from analysis. The 10 remaining participants were mentally healthy, having met study

Circadian rhythmicity, variability and correlation of interleukin-6 levels in plasma and cerebrospinal fluid inclusion/exclusion criteria, which prohibited presence or history of any DSM-IV Axis I mental health disorder or abuse/ dependency of alcohol, tobacco or other illicit substances. Likewise, based on a thorough physical examination of the volunteers, including blood laboratory tests, chest X-ray and electrocardiogram, all volunteers were confirmed to be physically healthy. No subject reported a history of any chronic or acute inflammatory or immune system-associated medical conditions. No subject had a positive urine toxicology screen, was using any prescribed medications, or had used either prescribed or over the counter medications for at least five days prior to admission to the clinical research center (CRC). As our protocol called for exclusion of subjects with prescribed or over the counter drug use, no volunteer endorsed glucocorticoid use during the last weeks, or intake of NSAID medication in the last five days prior admission to the CRC. Physical examination in the afternoon of CRC admission also provided assurance that no subject had any acute clinical manifestations or febrile body temperature.

2.2. Measures After the documentation of basic demographic information, presence and history of psychiatric co-morbidities were assessed by a supervised trained psychologist by the Structured Clinician Interview for the DSM-IV-TR Axis I Disorders (SCID-I) (First et al., 2002) and Hamilton-depression scale (HDRS) (Hamilton, 1960), while history of physical illness and family history psychiatric and physical illness was assessed by unstructured exploratory clinical interviews.

73

bedside. All material was subsequently stored in a highperformance laboratory freezer at 80 8C, until immediately before assay. Sampling lasted for 24 h, at which time the subarachnoid catheter was removed. Subjects were encouraged to maintain a regular sleep time at around 10 p.m. (lights off) during each night of study participation, while relative silence (no electronic media or disturbing conversation) was maintained in the study room. A 24 h time period starting at 11:30 h was used for this study.

2.4. Assays 2.4.1. IL-6 measurements in CSF samples IL-6 concentrations were measured in CSF samples using the Quantikine ELISA (regular sensitivity) kit (R&D Systems, Minneapolis, MN). Lower detection limit and linear range were 0.7 pg/ml and 3.1—300 pg/ml respectively (intra-assay coefficient of variation: 2%; inter-assay coefficient of variation: 4%). IL-6 levels greater than the highest standard were diluted with calibrator media to be within the linear working range and re-assayed. 2.4.2. IL-6 measurements in plasma samples IL-6 concentrations were measured in plasma samples using the Quantikine HS ELISA (high sensitivity) kit (R&D Systems, Minneapolis, MN). Lower detection limit and linear range were 0.04 pg/ml and 0.16—10 pg/ml respectively (intraassay coefficient of variation: 7%; inter-assay coefficient of variation: 9%). IL-6 levels greater than the highest standard were diluted with calibrator media to be within the linear working range and re-assayed.

2.3. Procedures 2.5. Statistical analyses Study volunteers recruited by verbal or printed advertisement who met phone-screening enrollment criteria were invited to the laboratory for an introduction to the study. Following signing of the consent form, a series of assessments (full physical and neurological examinations, chest X-ray, electrocardiogram, blood draw, urine toxicology screen and mental health assessment) was completed in an office setting. Volunteers who met study inclusion and exclusion criteria were scheduled for admission to the Clinical Research Center (CRC). From the time of admission, study participants remained on a controlled low monoamine diet receiving three standard meals (each a 666 calories: 20% protein, 24% fat, and 56% carbohydrates), in addition to an evening snack of 300 calories. At 8 p.m. local time, the evening of CRC admission, an indwelling venous catheter was placed for blood withdrawal and a standard meal was provided. Participants fasted for approximately 12 h until 8 a.m. the following day, when a 20-gauge catheter was placed in the lumbar subarachnoid space at the L3/4 or L4/5 level. Normal saline solution was infused (100 ml/h) throughout the experiment through a second IV line. From that point, all subjects stayed in bed for 24 h. At 11:00 a.m. continuous CSF withdrawal into iced test tubes was begun at a rate of 0.03 ml/min. Every 30 min the 0.9 ml of CSF collected was separated into four 0.225 ml aliquots. 7.5 ml blood was withdrawn into iced glass test EDTA-coated tubes every 30 min and centrifuged in a refrigerated centrifuge to separate plasma. All fluids were harvested, processed and directly placed on dry ice near the

Statistical analyses were conducted using R, version 2.14.2. An error probability of p  .05 was accepted as statistically significant. A total of 187 CSF and 163 plasma observations were included in our analyses. Preliminary analyses were performed to assess normality, linearity and homoscedasticity. Values with normal distribution are reported as mean (standard error) and skewed values as median values (min— max). Ratios were calculated as plasma/CSF IL-6. 24 h serial CSF and plasma IL-6 levels, as well as plasma/CSF ratios, showed a skewed distribution and were therefore log2 transformed for further analyses. Relations were investigated using the Pearson product-moment correlation coefficient r for mean log2-transformed values. To assess the dispersion in the distribution of our results, the coefficient of variation (CV) has been calculated for individuals with more than 50% of valid observations. The coefficient of variation reflects the absolute value of the relative standard deviation (RSD), expressed as a percentage. The CV is essentially the same as the standard deviation of the logarithm of the data and given by the form: CV(data)= SD[log(data)]  100. CV below 100% is considered low-variance, while CV over 100% indicates high-variance. The circadian rhythmicity of IL-6 was assessed by utilizing linear mixed-effects models. The fixed effects considered were age, BMI, linear trend in time, and 24-, 12-, and 8 h circadian rhythms. Age was centered by subtracting 30, BMI was centered by subtracting 26, and time 0 represented

(6.4—54.6) 100 50 — 130 30 130 50 —

(1.2—7.7) (0.9—8.3) (0.9—8.4) (1.6—19.4) (0.9—30.6)

(0.9—30.6) (1.4—5.8)

29 26 23 28 37 21 26

Descriptive statistics are given as median values (Md) and Min—Max values for skewed untransformed data. CSF: cerebrospinal fluid; CV: coefficient of variation; Age is reported in years; BMI: body mass index calculated as kg/m2; C: Caucasian; H: Hispanic; For descriptive reasons only participants with more than 50% of data available are used to calculate median personal plasma and CSF levels as well as ratios and CV. * p < 0.05; # trend (0.1 > p > 0.05); n.s., not significant. a Less than 50% of plasma IL-6 data available. b Less than 50% of CSF IL-6 data available.

0.01—0.68 — — 0.01—1.88 0.00—0.07 — 0.00—0.17 0.12—1.80 0.02—1.31 0.00—0.04 0.00—1.88 0.04 — 70 0.15 0.02 — 0.02 0.30 0.07 0.02 0.04 100 120

(17.3—278.6) (23.7—829.5) 31.0 (13.2—116.5) (19.3—515.7) — (9.2—340.3) (4.6—10.0) (3.4—121.6) (269.5—711.53) (3.4—829.5) 103.1 387.3 — 36.0 69.7 — 134.0 6.6 51.8 341.3 72.8 110 — — 130 60 — 50 100 130 90 (1.2—19.1) — —

3.7 — C 6.0 2.7 — 2.1 1.8 2.6 7.7 3.2 C C 3 C C C/H C C C C 2 0 20.54 0 2 0 0 0 0 0 27.11 27.80 31 21.73 24.80 28.27 25.53 23.74 21.76 18.40

Min—Max Md CV (%) Min—Max Md

CSF IL-6

CV (%) Min—Max Md Ethnicity HDRS

42 38

Individual plasma/CSF IL-6 concentration ratios ranged from .003 to 1.88 across all subjects (median score: .055). Plasma IL-6 levels were consistently lower that the CSF levels, but did not illustrate any constant pattern, as they showed high variance. In participants with higher plasma IL-6 levels, correspondingly higher CSF IL-6 concentrations were not observed. No significant correlation between plasma/CSF

Subject 1 Subject 2 a Subject 3 a Subject 4 Subject 5 Subject 6a,b Subject 7 Subject 8 Subject 9 Subject 10 Whole Sample Median

3.3. Plasma/CSF IL-6 ratios

BMI

The group median levels of CSF and plasma IL-6 were 59.8 (3.4—829.5) and 3.1 (0.9—30.6) pg/mL respectively (cf. Table 1). The IL-6 geometric means for CSF and plasma are illustrated in Fig. 1. Between-subject IL-6 concentration variability ranged from 10 to 50% in plasma (mean: 35.1%, highest variability around 1:30 p.m. and 6:30 a.m.) and 40—80% in CSF (mean: 61.9%, highest variability around 11:30 p.m.). Our results showed no significant correlations between either age or BMI and CSF (age: r = 07, p = .85; BMI: r = .25, p = .50) or plasma IL-6 (age: r = .11, p = .80; BMI: r = .24, p = .58) levels, respectively. There were no significant correlations between plasma or CSF IL-6 concentrations and hemodynamic measures (resting heart rate, systolic and diastolic blood pressure) (data not shown).

Age

3.2. IL-6 levels in CSF and plasma

Subjects

The mean (SE) age of the eleven healthy volunteers was 30 (1.97) years; range 21—42 years. Mean body mass index (BMI) was 23.9 (1.60); range 16.5—31.0. None of the participants had a HDRS score higher than 3 points.

Plasma IL-6

3.1. Sample characteristics

Levels, variance, ratios and correlations of plasma and cerebrospinal fluid IL-6 in each sample across 24 h.

3. Results

Table 1

11:00 AM on the first day. The random effects tested were random intercepts and slopes for each subject, and the residuals in the model were tested for the presence of an autoregressive correlation structure of order 1 (AR(1) correlation). The necessity of the AR(1) correlation structure plus random intercepts and/or slopes was tested on a model with all potential fixed effects present. Once the correct random effects were determined, the best model for the fixed effects was determined by a stepwise model selection procedure using the Akaike information criterion (AIC). The 24 h circadian component was handled by either entering in or deleting from the model both cos(2pt/24) and sin(2pt/24) simultaneously, and similarly for the 12- and 8 h circadian components. To control for bias by analyzing patients with only a few observations, the final model was run with all patients who had at least 20 observations, and the models were compared to make sure the coefficients and p-values were qualitatively and quantitatively similar. Time-ordered relations between plasma and CSF IL-6 were investigated by time-cross-correlation analyses similar to prior studies (Alesci et al., 2005; Crofford et al., 1997). Cross-correlation was computed at various time lags covering the 24 h period, by leading or lagging the concentration-time series of plasma IL-6 relative to the concentration-time series of CSF IL-6.

A. Agorastos et al.

Plasma/CSF ratio

74

Circadian rhythmicity, variability and correlation of interleukin-6 levels in plasma and cerebrospinal fluid

75

Figure 1 Geometric means of IL-6 concentrations in cerebrospinal fluid and plasma across 24 h. This figure presents the cerebrospinal fluid and plasma IL-6 values across the investigated 24 h timeframe in the whole sample. IL-6 concentrations showed a skewed distribution and were therefore log2-transformed. The arithmetic means and SEs calculated from log-transformed data were backtransformed to their original scale through exponentiation, so the sample means represent geometric means on the raw scale. Geometric means and SEs were plotted over time.

IL-6 concentration ratio and any personal or physiological parameter was observed (data not shown).

3.4. Linear trend over the 24 h sampling timeframe Our results indicated a significant linear trend for both CSF ( p = .001) and plasma ( p = .003) over the 24 h timeframe. The estimated IL-6 increases were 0.098 log2(pg/h) for CSF IL-6 and 0.055 log2(pg/hr) for plasma IL-6, respectively.

3.5. Circadian rhythmicity The results of the harmonic analysis indicated that CSF levels showed a strong 24 h component [p = .00003; amplitude: 1.10 (SE = 0.24)] and a relatively weak 12 h and 8 h component [p = .053; amplitude: 0.31 (SE = 0.13) and p = .055; amplitude: 0.20 (SE = 0.09) respectively]. CSF peak was at around 7:20 p.m. (SE = 49 min) (cf. Fig. 2). Plasma measures showed only a 12 h component [p = .0016; amplitude: 0.53 (SE = 0.15)], while the 8 h and 24 h circadian components were not significant ( p = .46 and p = .21, respectively). Plasma peaks were at roughly 4:12 p.m. and 4:12 a.m. (SE = 34 min) (cf. Fig. 2). Harmonic analysis revealed neither a significant linear trend nor any circadian components of plasma/CSF ratios (data not shown).

3.6. Correlations between CSF and plasma IL-6 No correlation pattern between mean CSF and plasma IL-6 values were observed within subjects (data not shown).

Cross-lagged correlation analyses between plasma and CSF IL-6 concentrations revealed a pattern of correlations with negative lag time and specifically significant positive correlations at 7 and 8 h time lags, meaning that earlier plasma IL-6 concentrations correlated with later CSF concentrations within the 24 h time frame investigated. Positive lag time analysis (later plasma concentrations, earlier CSF concentrations) showed no correlations between plasma and CSF concentrations. This pattern was also confirmed by investigating time lagged cross-correlations of plasma and CSF IL-6 in each individual separately (data not shown). The mean of the coefficients of correlation (Rx)  SE of the mean (SEM) between plasma and CSF IL-6 concentration-time series at lag time x is presented in Fig. 3.

4. Discussion Immunological research has repeatedly used IL-6 as an important inflammatory marker (Browning et al., 2004). In this study, simultaneous hourly assessment of plasma and CSF IL-6 levels was performed over 24 h, to further characterize physiological characteristics and circadian variation of IL-6 secretion patterns in the peripheral and central compartments in healthy males. The main findings of this study are: i) evidence for a 24 h CSF IL-6 circadian rhythm, ii) replication of prior results showing a biphasic circadian plasma IL-6 pattern, iii) throughout higher CSF than plasma IL-6 concentrations, iv) large within- and between-subject variability in IL-6 concentrations gathered over a 24 h timeframe and v) no correlation between coincident CSF and plasma IL-6 values, but evidence for significant correlations at a negative 7—8 h time lag.

76

A. Agorastos et al.

Figure 2 Circadian rhythmicity of IL-6 in cerebrospinal fluid and plasma across 24 h. This figure presents point estimates and SEs of harmonic analysis illustrating the circadian rhythmicity of IL-6 in cerebrospinal fluid and plasma across the investigated 24 h timeframe in the whole sample. The circadian rhythmicity of IL-6 was assessed by utilizing linear mixed-effects models (cf. statistics section). The circadian rhythms in the final model were calculated as coefficients of the cosine and sine components, and translated into amplitude and peak location. The standard errors of the amplitude and peak location were calculated from those of the sine and cosine terms in the final model by the multivariate delta-method.

4.1. IL-6 levels and between-subject variability Our results support prior findings reporting higher IL-6 concentrations in CSF than in the periphery in healthy humans (Maier et al., 2005; Steensberg et al., 2006; Stenlof et al.,

2003). On the other hand, we could not confirm prior research reporting positive correlations between age, BMI and plasma IL-6 levels (Carey et al., 2004; Sothern et al., 1995b; Vgontzas et al., 1999). We interpret our inability to find these correlations as due to our small study sample and

Figure 3 Cross-correlation analysis between plasma and CSF IL-6 over the 24 h study period Cross-correlation analysis presenting the mean of the individual values of the coefficient of correlation Rx for each subject at lag time x (solid line). The dark gray area represents the standard error of the mean (SEM) above and below zero and indicates the limits of statistical significance for cross-correlation at p = .05 level. Significant correlation at any lagtime is achieved when the solid line falls outside the dark gray area. The light gray area indicates the limits of statistical significance for cross-correlations after Bonferroni correction. Positive lags mean plasma IL-6 is at later point then CSF IL-6.

Circadian rhythmicity, variability and correlation of interleukin-6 levels in plasma and cerebrospinal fluid the narrow age and BMI ranges of our healthy young male volunteers. Our study contributes to the sparse literature on betweensubject variability in IL-6 concentrations in healthy humans. Our findings provide robust evidence for a considerable between-subject variability of IL-6 levels in healthy humans (plasma: 10—50%; CSF: 40—80%). In plasma, prior findings have suggested a between-subject IL-6 concentration variability above 30% (Dugue and Leppanen, 1998), however the study cited used only 3 repeated plasma measures within 3.5 h (8:00 a.m.—11:30 a.m.), thus, did not cover an entire 24 h timeframe. To our knowledge, no study has yet reported on data showing between-subject variability in CSF IL-6 concentrations.

4.2. IL-6 circadian pattern Our study shows a distinctive 12 h biphasic circadian component with zeniths at approximately 4 p.m. and 4 a.m. and nadirs at approximately 10 a.m. and 10 p.m. These findings are partly in accordance with prior studies showing circadian, diurnal and overnight variations of IL-6 in plasma (Bauer et al., 1994; Crofford et al., 1997; Gudewill et al., 1992; Izawa et al., 2013a; Kanabrocki et al., 1999; Lissoni et al., 1998; Perry et al., 2009; Redwine et al., 2000; Sothern et al., 1995a,b; Vgontzas et al., 2002) and are fairly concordant with the findings of Vgontzas et al. showing a bi-phasic plasma IL-6 circadian pattern, with, however, slightly differing nadir and zenith time points (Vgontzas et al.: nadirs at 8 a.m. and 9 p.m.; zeniths at 7 p.m. and 5 a.m.) (Vgontzas et al., 1999, 2005). Perhaps the most notable finding of this study is the evidence for a main 24 h circadian IL-6 pattern in the CNS. To the best of our knowledge, there have been no comparable prior research findings in humans. Our results, thus, suggest circadian IL-6 patterns in the CNS and the periphery that differ in periodicity. Accordingly, we did not observe any correlation between coincident CSF and plasma IL-6 values. On the other hand, on cross-lagged correlation analyses, our data also reveals significant correlations between plasma and CSF IL-6 levels at a 7—8 h lag, plasma preceding CSF concentrations. The physiological significance of this finding, if replicated, merits further investigation.

4.3. Physiological significance Our results suggest higher IL-6 levels in the CSF than in plasma, as well as differing circadian secretion patterns in the CNS versus peripheral compartments. As active IL-6 transport from the periphery to the CNS occurs only at low levels, its contribution to cross-compartmental exchange is regarded as minimal (Banks et al., 1994; Hopkins, 2007; Steensberg et al., 2006). These findings support prior animal and human research findings that suggest differential regulation of IL-6 in the two compartments (De Simoni et al., 1995; Reyes and Coe, 1996, 1998; Steensberg et al., 2006), as well as local production of IL-6 in the CNS, which could possibly be the reason for the large differences in IL-6 levels between CNS and periphery. Our results, thus, are consistent with prior literature embracing an independent physiological role for CNS IL-6 (Gruol and Nelson, 1997; Juttler et al., 2002;

77

Otten et al., 1994). Under basal conditions, the circadian CNS secretion of IL-6 could potentially be a factor modulating neuronal activity and other brain functions, as recently suggested (Besedovsky and del Rey, 2011; Juttler et al., 2002). In addition, further indirect mechanisms might be involved in the interplay between plasma and CSF IL-6, possibly explaining their deviating concentration levels and timely-limited time-lagged correlation observed, such as positive and negative CNS-periphery cytokine feedback regulation, cross-communication of the immune system with the HPA axis and the autonomic nervous system or other mechanisms (Besedovsky and del Rey, 2011; Elenkov, 2008; Hopkins, 2007). In this context, the relation of plasma IL-6 to the hypothalamic—pituitary—adrenal (HPA) axis is considered of particular importance. IL-6 stimulates the HPA axis at the level of the hypothalamus, pituitary, and adrenal gland (Crofford et al., 1997; Mastorakos et al., 1993; Naitoh et al., 1988; Salas et al., 1990; Stouthard et al., 1995; Tominaga et al., 1991; Vankelecom et al., 1989; Wilder, 1995). Conversely, cortisol exerts an inhibitory effect on the peripheral production of IL-6 (Papanicolaou et al., 1996b) and is a major moderator of circadian IL-6 changes (Alesci et al., 2005; Fantidis et al., 2002). Norepinephrine, on the other hand, leads to an increase of plasma IL-6 (DeRijk et al., 1994; Sondergaard et al., 2000; van Gool et al., 1990), in part via adrenal mechanisms regulating splenic IL-6 production (Engler et al., 2010; Hanke et al., 2012; Judd, 1998). A recent report also suggested that basal IL-6 signaling in the hypothalamus is a potential determinant of plasticity in the HPA axis response, specifically during chronic stress exposure (Girotti et al., 2012). The functional significance of the circadian variation in IL-6 CSF and plasma levels is mainly unknown. Nevertheless, the circadian IL-6 secretion profile in both central and peripheral compartments, although divergent, underlines the potential major importance of the circadian system in the physiology of IL-6. The human central circadian system and the immune system are closely and bi-directionally connected (Bryant et al., 2004; Coogan and Wyse, 2008; Imeri and Opp, 2009; Irwin, 2002; Marshall and Born, 2002). Prior animal and human research provides evidence of a physiological link between the circadian system and IL-6 (Monje et al., 2011; Redwine et al., 2000; Vgontzas et al., 2002, 2005; Vgontzas and Chrousos, 2002). Melatonin, one of the most important natural substances modulating sleep and circadian rhythms (Macchi and Bruce, 2004), is thought to have a direct effect on IL-6 production though nuclear receptors bridging immunomodulatory, chronobiologic and inflammatory processes (Clapp-Lilly et al., 2001; GarciaMaurino et al., 2000; Giannoulia-Karantana et al., 2006; Guerrero et al., 2000; Lau et al., 2011; Lissoni et al., 1998; Srinivasan et al., 2008a,b). On the other hand, IL-6 appears to play a direct role in sleep regulation (Motzkus et al., 2002; Rohleder et al., 2012; Vgontzas et al., 1999), while several clinical findings confirm that disturbed day-night cycles and sleep deprivation affect the rhythmic intra-diem oscillations of plasma IL-6 (Monje et al., 2011; Vgontzas and Chrousos, 2002; Vgontzas et al., 1999, 2002, 2003). In addition, CNS IL-6 has been implicated in cross-regulation of several central processes that also follow circadian patterns, such as thermoregulation, sleep, appetite, fati-

78 gue, HPA axis and SNS modulation, and corticotrophin releasing factor (CRF) and adrenocorticotropic hormone (ACTH) stimulation (Bethin et al., 2000; Biddle, 2006; Bluthe et al., 2000; Conti et al., 2004; Das, 2001; Juttler et al., 2002; Papanicolaou and Vgontzas, 2000; Vgontzas et al., 1999). Differential regulation of rhythmic immune functions between compartments, coordinated by different oscillating clocks on multiple levels could be a possible explanation of our findings and might contribute to a better understanding of the time-related interplay between immune-modulators and the circadian system (Keller et al., 2009). Loss of cellular and systemic rhythmicity, also referred to as chronodisruption (Erren and Reiter, 2009), may, thus, alter IL-6 secretion and potentially represent a link between this cytokine and an increased risk for inflammatory, cardiovascular and other diseases (Burgos et al., 2006; Castanon-Cervantes et al., 2010; Elenkov et al., 2005; Nader et al., 2010). However, the exact mechanisms that explain all the central and peripheral effects of IL-6 are still not fully determined. For example, our study reports on circadian secretion patterns of total free IL-6 concentrations, but does not provide any information on the concentration of soluble IL6 receptor (sIL-6R) or receptor-bound IL-6, a factor that may relate to differing IL-6 concentrations observed in the CSF and periphery (Rose-John et al., 2007). As the neural activity of IL-6 mainly depends on sIL-6R (Dimitrov et al., 2006; Marz et al., 1999), and plasma levels of soluble receptors of other cytokines (e.g., TNF-a and IL-2) also exhibit circadian patterns in healthy humans, further studies addressing the circadian patterns of sIL-6R levels and receptor-bound IL-6 in both the CNS and peripheral compartments may provide crucial new information on the physiology of IL-6 (Haack et al., 2004; Lemmer et al., 1992) and support the ongoing research on emerging anti-IL-6 treatment options (Nishimoto, 2010).

4.4. Clinical significance and research implications Our findings have implications for immunological research and clinical practice, as they point out that measurements of IL-6 in either CSF or plasma are variable across time and between individuals, thus indicating that caution should be used when interpreting a single IL-6 value as a valid immune status or treatment outcome marker. Our results suggest that future research should consider study designs with time-matched sample collection across participants, larger sample sizes and/or repeated measurements. Evidence of a distinctive IL-6 circadian rhythmicity in both CSF and plasma shown in our study provides further support for the presence of the biological correlate of diurnal variations seen in the inflammatory response, hormone levels and related clinical symptoms of various chronic diseases, such as rheumatoid arthritis and cancer (Berger et al., 2012; Cutolo et al., 2005, 2006; Perry et al., 2009; Spies et al., 2010). Since circadian rhythmicity of neuroendocrine pathways is closely coupled to immune/inflammatory reactions, our data are consistent with the repeatedly suggested time-adapted immunomodulatory treatment strategies in such diseases, in

A. Agorastos et al. order to increase efficacy, allow lower dosage and minimize side effects (Cutolo et al., 2006; Innominato et al., 2010; Iwata et al., 2011; Levi et al., 2007; Sewlall et al., 2010; Spies et al., 2010). Finally, the simultaneous assessment of peripheral and central IL-6 patterns with respect to other hormones, e.g. cortisol and cytokines would be of major future research importance, in order to better understand the bi-directional neuroendocrine-immune system interactions underlying a plethora of common human immune-related diseases (Elenkov, 2008). For example, in our study, plasma IL-6 showed its first peak in the late night hours, preceding the welldescribed morning cortisol rise in plasma, as described elsewhere (Crofford et al., 1997).

5. Limitations Because of its invasive and intensive nature, only a relatively small number of volunteers were enrolled in this study. In addition, some subjects have only partial data available for analysis and we investigated only male subjects. Thus, our findings should be considered preliminary, may represent gender-specific effects, and merit replication in both genders. Study volunteers recruited by verbal or printed advertisements might differ regarding health background variables from the general population. However, all subjects were well matched in terms of age and ethnicity, and stringent exclusion criteria were met to minimize the probability of medical confounders. In addition, the harmonic analysis is functionally equivalent to that employed by Vgontzas et al. (1999, 2005), except that the linear mixed model can better handle missing data. Our results suggested a global raise of IL-6 levels over the 24 h timeframe in both compartments. The failure of IL-6 levels to return to baseline after 24 h is in accordance with the phenomenon reported in prior literature that show higher IL-6 levels over time from sampling through an indwelling catheter compared to single blood samples (Haack et al., 2000, 2002; Seiler et al., 1995). The possibility that our experimental paradigm including indwelling intravenous catheter could affect our results through local irritation and IL-6 production can therefore not be discounted. However, the robust biphasic plasma IL-6 circadian pattern and the similarity of our plasma data to prior studies (Vgontzas et al., 1999, 2005) imply only minor effects on our results on circadian rhythmicity. Although differences between normal ventricular and lumbar CSF have been suggested (Reiber et al., 2001), recent studies detect no rostro-caudal CSF gradient (Brandner et al., 2013), suggesting that IL-6 measurement in the CSF provides objective evidence of CSF IL-6 levels in the CNS and should be considered a valid method reflecting the dynamics in the CSF circulation over time. As sleep patterns also affect IL-6 secretion (Vgontzas and Chrousos, 2002; Vgontzas et al., 1999, 2002), we compared available data on physical activity and sleep patterns in our study subjects, monitored by a motion-sensitive, wrist-actigraphy monitor and calculated using validated algorithms (Sadeh et al., 1989). Among our subjects, there were no statistical significant inter-subject differences in sleep patterns (data not shown).

Circadian rhythmicity, variability and correlation of interleukin-6 levels in plasma and cerebrospinal fluid

6. Conclusions Psychoneuroimmunological research has offered mounting evidence of bi-directional communication between the immune and central nervous systems; IL-6 is thought to play a key endocrine mediator role in these communications. Our findings provide the first evidence in humans for a circadian IL-6 rhythm in the CSF and confirm prior results reporting a biphasic circadian plasma IL-6 pattern, herewith supporting the physiological need for a temporal organized, endogenous IL-6 pattern. The pathophysiological relevance of the central and peripheral IL-6 circadian secretion patterns in humans could be of major importance and remains to be further elucidated in further research.

Role of funding sources We gratefully acknowledge the fiscal support of the VA CSR&D Program, MERIT grant funded 2004—2008 (DGB) and support of the VA Center of Excellence for Stress and Mental Health San Diego, CA, USA (DGB, RHL, JBL).

Conflicts of interest The authors have no conflicts of interest.

Acknowledgements We also thank Dr. Robert Henry and the staff of the VA Clinical Research Center (CRC) for their assistance and acknowledge individuals who provided help with data collection and sample processing, Ekhator NN, Strawn JR (2004, Cincinnati), Fitzpatrick S, Valencerina A, Motazedi A, Turken A (2005— 2010, San Diego).

References Akira, S., Taga, T., Kishimoto, T., 1993. Interleukin-6 in biology and medicine. Adv. Immunol. 54, 1—78. Alesci, S., Martinez, P.E., Kelkar, S., Ilias, I., Ronsaville, D.S., Listwak, S.J., Ayala, A.R., Licinio, J., Gold, H.K., Kling, M.A., Chrousos, G.P., Gold, P.W., 2005. Major depression is associated with significant diurnal elevations in plasma interleukin-6 levels, a shift of its circadian rhythm, and loss of physiological complexity in its secretion: clinical implications. J. Clin. Endocrinol. Metab. 90, 2522—2530. Banks, W.A., Kastin, A.J., Gutierrez, E.G., 1994. Penetration of interleukin-6 across the murine blood—brain barrier. Neurosci. Lett. 179, 53—56. Bauer, J., Hohagen, F., Ebert, T., Timmer, J., Ganter, U., Krieger, S., Lis, S., Postler, E., Voderholzer, U., Berger, M., 1994. Interleukin6 serum levels in healthy persons correspond to the sleep—wake cycle. Clin. Investig. 72, 315. Berger, A.M., Hertzog, M., Geary, C.R., Fischer, P., Farr, L., 2012. Circadian rhythms, symptoms, physical functioning, and body mass index in breast cancer survivors. J. Cancer Surviv.. Besedovsky, H.O., del Rey, A., 2011. Central and peripheral cytokines mediate immune-brain connectivity. Neurochem. Res. 36, 1—6. Bethin, K.E., Vogt, S.K., Muglia, L.J., 2000. Interleukin-6 is an essential, corticotropin-releasing hormone-independent stimulator of the adrenal axis during immune system activation. Proc. Natl. Acad. Sci. U. S. A. 97, 9317—9322.

79

Biddle, C., 2006. The neurobiology of the human febrile response. AANA J. 74, 145—150. Bluthe, R.M., Michaud, B., Poli, V., Dantzer, R., 2000. Role of IL-6 in cytokine-induced sickness behavior: a study with IL-6 deficient mice. Physiol. Behav. 70, 367—373. Brandner, S., Thaler, C., Lewczuk, P., Lelental, N., Buchfelder, M., Kleindienst, A., 2013. Neuroprotein dynamics in the cerebrospinal fluid: intraindividual concomitant ventricular and lumbar measurements. Eur. Neurol. 70, 189—194. Browning, L.M., Krebs, J.D., Jebb, S.A., 2004. Discrimination ratio analysis of inflammatory markers: implications for the study of inflammation in chronic disease. Metabolism 53, 899—903. Bryant, P.A., Trinder, J., Curtis, N., 2004. Sick and tired: does sleep have a vital role in the immune system? Nat. Rev. Immunol. 4, 457—467. Burgos, I., Richter, L., Klein, T., Fiebich, B., Feige, B., Lieb, K., Voderholzer, U., Riemann, D., 2006. Increased nocturnal interleukin-6 excretion in patients with primary insomnia: a pilot study. Brain Behav. Immun. 20, 246—253. Carey, A.L., Bruce, C.R., Sacchetti, M., Anderson, M.J., Olsen, D.B., Saltin, B., Hawley, J.A., Febbraio, M.A., 2004. Interleukin-6 and tumor necrosis factor-alpha are not increased in patients with Type 2 diabetes: evidence that plasma interleukin-6 is related to fat mass and not insulin responsiveness. Diabetologia 47, 1029— 1037. Castanon-Cervantes, O., Wu, M., Ehlen, J.C., Paul, K., Gamble, K.L., Johnson, R.L., Besing, R.C., Menaker, M., Gewirtz, A.T., Davidson, A.J., 2010. Dysregulation of inflammatory responses by chronic circadian disruption. J. Immunol. 185, 5796—5805. Castell, J.V., Gomez-Lechon, M.J., David, M., Andus, T., Geiger, T., Trullenque, R., Fabra, R., Heinrich, P.C., 1989. Interleukin-6 is the major regulator of acute phase protein synthesis in adult human hepatocytes. FEBS Lett. 242, 237—239. Chourbaji, S., Urani, A., Inta, I., Sanchis-Segura, C., Brandwein, C., Zink, M., Schwaninger, M., Gass, P., 2006. IL-6 knockout mice exhibit resistance to stress-induced development of depressionlike behaviors. Neurobiol. Dis. 23, 587—594. Chrousos, G.P., 1995. The hypothalamic—pituitary—adrenal axis and immune-mediated inflammation. N. Engl. J. Med. 332, 1351— 1362. Clapp-Lilly, K.L., Smith, M.A., Perry, G., Duffy, L.K., 2001. Melatonin reduces interleukin secretion in amyloid-beta stressed mouse brain slices. Chem. Biol. Interact. 134, 101—107. Conti, B., Tabarean, I., Andrei, C., Bartfai, T., 2004. Cytokines and fever. Front. Biosci. 9, 1433—1449. Coogan, A.N., Wyse, C.A., 2008. Neuroimmunology of the circadian clock. Brain Res. 1232, 104—112. Crofford, L.J., Kalogeras, K.T., Mastorakos, G., Magiakou, M.A., Wells, J., Kanik, K.S., Gold, P.W., Chrousos, G.P., Wilder, R.L., 1997. Circadian relationships between interleukin (IL)-6 and hypothalamic—pituitary—adrenal axis hormones: failure of IL-6 to cause sustained hypercortisolism in patients with early untreated rheumatoid arthritis. J. Clin. Endocrinol. Metab. 82, 1279—1283. Cutolo, M., Sulli, A., Pizzorni, C., Secchi, M.E., Soldano, S., Seriolo, B., Straub, R.H., Otsa, K., Maestroni, G.J., 2006. Circadian rhythms: glucocorticoids and arthritis. Ann. N. Y. Acad. Sci. 1069, 289—299. Cutolo, M., Villaggio, B., Otsa, K., Aakre, O., Sulli, A., Seriolo, B., 2005. Altered circadian rhythms in rheumatoid arthritis patients play a role in the disease’s symptoms. Autoimmun. Rev. 4, 497— 502. Das, U.N., 2001. Is obesity an inflammatory condition? Nutrition 17, 953—966. De Simoni, M.G., Del Bo, R., De Luigi, A., Simard, S., Forloni, G., 1995. Central endotoxin induces different patterns of interleukin (IL)-1 beta and IL-6 messenger ribonucleic acid expression and

80 IL-6 secretion in the brain and periphery. Endocrinology 136, 897— 902. DeRijk, R.H., Boelen, A., Tilders, F.J., Berkenbosch, F., 1994. Induction of plasma interleukin-6 by circulating adrenaline in the rat. Psychoneuroendocrinology 19, 155—163. Dimitrov, S., Lange, T., Benedict, C., Nowell, M.A., Jones, S.A., Scheller, J., Rose-John, S., Born, J., 2006. Sleep enhances IL-6 trans-signaling in humans. FASEB J. 20, 2174—2176. Dugue, B., Leppanen, E., 1998. Short-term variability in the concentration of serum interleukin-6 and its soluble receptor in subjectively healthy persons. Clin. Chem. Lab. Med. 36, 323—325. Elenkov, I.J., 2008. Neurohormonal-cytokine interactions: implications for inflammation, common human diseases and well-being. Neurochem. Int. 52, 40—51. Elenkov, I.J., Iezzoni, D.G., Daly, A., Harris, A.G., Chrousos, G.P., 2005. Cytokine dysregulation, inflammation and well-being. Neuroimmunomodulation 12, 255—269. Engler, H., Doenlen, R., Riether, C., Engler, A., Besedovsky, H.O., Del Rey, A., Pacheco-Lopez, G., Schedlowski, M., 2010. Chemical destruction of brain noradrenergic neurons affects splenic cytokine production. J. Neuroimmunol. 219, 75—80. Erren, T.C., Reiter, R.J., 2009. Defining chronodisruption. J. Pineal Res. 46, 245—247. Fantidis, P., Perez De Prada, T., Fernandez-Ortiz, A., Carcia-Touchard, A., Alfonso, F., Sabate, M., Hernandez, R., Escaned, J., Bauelos, C., Macaya, C., 2002. Morning cortisol production in coronary heart disease patients. Eur. J. Clin. Invest. 32, 304—308. Fee, D., Grzybicki, D., Dobbs, M., Ihyer, S., Clotfelter, J., Macvilay, S., Hart, M.N., Sandor, M., Fabry, Z., 2000. Interleukin 6 promotes vasculogenesis of murine brain microvessel endothelial cells. Cytokine 12, 655—665. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 2002. Nonpatient Edition (SCID-I/NP). Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Non-patient Edition (SCID-I/NP), Biometrics Research, New York State Psychiatric Institute, New York. Garcia-Maurino, S., Pozo, D., Calvo, J.R., Guerrero, J.M., 2000. Correlation between nuclear melatonin receptor expression and enhanced cytokine production in human lymphocytic and monocytic cell lines. J. Pineal Res. 29, 129—137. Giannoulia-Karantana, A., Vlachou, A., Polychronopoulou, S., Papassotiriou, I., Chrousos, G.P., 2006. Melatonin and immunomodulation: connections and potential clinical applications. Neuroimmunomodulation 13, 133—144. Girotti, M., Donegan, J.J., Morilak, D.A., 2012. Influence of hypothalamic IL-6/gp130 receptor signaling on the HPA axis response to chronic stress. Psychoneuroendocrinology. Gruol, D.L., Nelson, T.E., 1997. Physiological and pathological roles of interleukin-6 in the central nervous system. Mol. Neurobiol. 15, 307—339. Gudewill, S., Pollmacher, T., Vedder, H., Schreiber, W., Fassbender, K., Holsboer, F., 1992. Nocturnal plasma levels of cytokines in healthy men. Eur. Arch. Psychiatry Clin. Neurosci. 242, 53—56. Guerrero, J.M., Pozo, D., Garcia-Maurino, S., Carrillo, A., Osuna, C., Molinero, P., Calvo, J.R., 2000. Nuclear receptors are involved in the enhanced IL-6 production by melatonin in U937 cells. Biol. Signals Recept. 9, 197—202. Haack, M., Kraus, T., Schuld, A., Dalal, M., Koethe, D., Pollmacher, T., 2002. Diurnal variations of interleukin-6 plasma levels are confounded by blood drawing procedures. Psychoneuroendocrinology 27, 921—931. Haack, M., Pollmacher, T., Mullington, J.M., 2004. Diurnal and sleep— wake dependent variations of soluble TNF- and IL-2 receptors in healthy volunteers. Brain. Behav. Immun. 18, 361—367. Haack, M., Reichenberg, A., Kraus, T., Schuld, A., Yirmiya, R., Pollmacher, T., 2000. Effects of an intravenous catheter on the local production of cytokines and soluble cytokine receptors in healthy men. Cytokine 12, 694—698.

A. Agorastos et al. Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23, 56—62. Hanke, M.L., Powell, N.D., Stiner, L.M., Bailey, M.T., Sheridan, J.F., 2012. Beta adrenergic blockade decreases the immunomodulatory effects of social disruption stress. Brain. Behav. Immun. 26, 1150—1159. Hansel, A., Hong, S., Camara, R.J., von Kanel, R., 2010. Inflammation as a psychophysiological biomarker in chronic psychosocial stress. Neurosci. Biobehav. Rev. 35, 115—121. Heinrich, P.C., Behrmann, I., Haan, S., Hermanns, H.M., MullerNewen, G., Schaper, F., 2003. Principles of interleukin (IL)-6-type cytokine signalling and its regulation. Biochem. J. 374, 1—20. Heinrich, P.C., Castell, J.V., Andus, T., 1990. Interleukin-6 and the acute phase response. Biochem. J. 265, 621—636. Helwig, B.G., Craig, R.A., Fels, R.J., Blecha, F., Kenney, M.J., 2008. Central nervous system administration of interleukin-6 produces splenic sympathoexcitation. Auton. Neurosci. 141, 104—111. Hirano, T., Yasukawa, K., Harada, H., Taga, T., Watanabe, Y., Matsuda, T., Kashiwamura, S., Nakajima, K., Koyama, K., Iwamatsu, A., et al., 1986. Complementary DNA for a novel human interleukin (BSF-2) that induces B lymphocytes to produce immunoglobulin. Nature 324, 73—76. Hopkins, S.J., 2007. Central nervous system recognition of peripheral inflammation: a neural, hormonal collaboration. Acta Biomed. 78 (Suppl. 1) 231—247. Imeri, L., Opp, M.R., 2009. How (and why) the immune system makes us sleep. Nat. Rev. Neurosci. 10, 199—210. Innominato, P.F., Levi, F.A., Bjarnason, G.A., 2010. Chronotherapy and the molecular clock: clinical implications in oncology. Adv. Drug. Deliv. Rev. 62, 979—1001. Irwin, M., 2002. Effects of sleep and sleep loss on immunity and cytokines. Brain Behav. Immun. 16, 503—512. Iwata, K., Aizawa, K., Sakai, S., Jingami, S., Fukunaga, E., Yoshida, M., Hamada, A., Saito, H., 2011. The relationship between treatment time of gemcitabine and development of hematologic toxicity in cancer patients. Biol. Pharm. Bull. 34, 1765—1768. Izawa, S., Miki, K., Liu, X., Ogawa, N., 2013a. The diurnal patterns of salivary interleukin-6 and C-reactive protein in healthy young adults. Brain Behav. Immun. 27, 38—41. Izawa, S., Miki, K., Liu, X.X., Ogawa, N., 2013b. The diurnal patterns of salivary interleukin-6 and C-reactive protein in healthy young adults. Brain Behav. Immun. 27, 38—41. Judd, A.M., 1998. Cytokine expression in the rat adrenal cortex. Horm. Metab. Res. 30, 404—410. Juttler, E., Tarabin, V., Schwaninger, M., 2002. Interleukin-6 (IL-6): a possible neuromodulator induced by neuronal activity. Neuroscientist 8, 268—275. Kanabrocki, E.L., Sothern, R.B., Messmore, H.L., Roitman-Johnson, B., McCormick, J.B., Dawson, S., Bremner, F.W., Third, J.L., Nemchausky, B.A., Shirazi, P., Scheving, L.E., 1999. Circadian interrelationships among levels of plasma fibrinogen, blood platelets, and serum interleukin-6. Clin. Appl. Thromb. Hemost. 5, 37—42. Keller, E.T., Wanagat, J., Ershler, W.B., 1996. Molecular and cellular biology of interleukin-6 and its receptor. Front. Biosci. 1, d340— d357. Keller, M., Mazuch, J., Abraham, U., Eom, G.D., Herzog, E.D., Volk, H.D., Kramer, A., Maier, B., 2009. A circadian clock in macrophages controls inflammatory immune responses. Proc. Natl. Acad. Sci. U. S. A. 106, 21407—21412. Lau, W.W., Ng, J.K., Lee, M.M., Chan, A.S., Wong, Y.H., 2011. Interleukin-6 autocrine signaling mediates melatonin MT(1/2) receptor-induced STAT3 Tyr(705) phosphorylation. J. Pineal Res.. Lemmer, B., Schwulera, U., Thrun, A., Lissner, R., 1992. Circadian rhythm of soluble interleukin-2 receptor in healthy individuals. Eur. Cytokine Netw. 3, 335—336. Levi, F., Focan, C., Karaboue, A., de la Valette, V., Focan-Henrard, D., Baron, B., Kreutz, F., Giacchetti, S., 2007. Implications of

Circadian rhythmicity, variability and correlation of interleukin-6 levels in plasma and cerebrospinal fluid circadian clocks for the rhythmic delivery of cancer therapeutics. Adv. Drug Deliv. Rev. 59, 1015—1035. Lindqvist, D., Janelidze, S., Hagell, P., Erhardt, S., Samuelsson, M., Minthon, L., Hansson, O., Bjorkqvist, M., Traskman-Bendz, L., Brundin, L., 2009. Interleukin-6 is elevated in the cerebrospinal fluid of suicide attempters and related to symptom severity. Biol. Psychiatry 66, 287—292. Lissoni, P., Rovelli, F., Brivio, F., Brivio, O., Fumagalli, L., 1998. Circadian secretions of IL-2, IL-12 IL-6 and IL-10 in relation to the light/dark rhythm of the pineal hormone melatonin in healthy humans. Nat. Immun. 16, 1—5. Macchi, M.M., Bruce, J.N., 2004. Human pineal physiology and functional significance of melatonin. Front. Neuroendocrinol. 25, 177—195. Maier, B., Laurer, H.L., Rose, S., Buurman, W.A., Marzi, I., 2005. Physiological levels of pro- and anti-inflammatory mediators in cerebrospinal fluid and plasma: a normative study. J. Neurotrauma 22, 822—835. Manolagas, S.C., Jilka, R.L., 1995. Bone marrow, cytokines, and bone remodeling. Emerging insights into the pathophysiology of osteoporosis. N. Engl. J. Med. 332, 305—311. Marshall, L., Born, J., 2002. Brain-immune interactions in sleep. Int. Rev. Neurobiol. 52, 93—131. Marz, P., Otten, U., Rose-John, S., 1999. Neural activities of IL-6-type cytokines often depend on soluble cytokine receptors. Eur. J. Neurosci. 11, 2995—3004. Mastorakos, G., Chrousos, G.P., Weber, J.S., 1993. Recombinant interleukin-6 activates the hypothalamic—pituitary—adrenal axis in humans. J. Clin. Endocrinol. Metab. 77, 1690—1694. Mihara, M., Hashizume, M., Yoshida, H., Suzuki, M., Shiina, M., 2012. IL-6/IL-6 receptor system and its role in physiological and pathological conditions. Clin. Sci. (Lond.) 122, 143—159. Mohamed-Ali, V., Flower, L., Sethi, J., Hotamisligil, G., Gray, R., Humphries, S.E., York, D.A., Pinkney, J., 2001. beta-Adrenergic regulation of IL-6 release from adipose tissue: in vivo and in vitro studies. J. Clin. Endocrinol. Metab. 86, 5864—5869. Mohamed-Ali, V., Goodrick, S., Rawesh, A., Katz, D.R., Miles, J.M., Yudkin, J.S., Klein, S., Coppack, S.W., 1997. Subcutaneous adipose tissue releases interleukin-6, but not tumor necrosis factoralpha, in vivo. J. Clin. Endocrinol. Metab. 82, 4196—4200. Monje, F.J., Cabatic, M., Divisch, I., Kim, E.J., Herkner, K.R., Binder, B.R., Pollak, D.D., 2011. Constant darkness induces IL-6-dependent depression-like behavior through the NF-kappaB signaling pathway. J. Neurosci. 31, 9075—9083. Motzkus, D., Albrecht, U., Maronde, E., 2002. The human PER1 gene is inducible by interleukin-6. J. Mol. Neurosci. 18, 105—109. Nader, N., Chrousos, G.P., Kino, T., 2010. Interactions of the circadian CLOCK system and the HPA axis. Trends Endocrinol. Metab. 21, 277—286. Naitoh, Y., Fukata, J., Tominaga, T., Nakai, Y., Tamai, S., Mori, K., Imura, H., 1988. Interleukin-6 stimulates the secretion of adrenocorticotropic hormone in conscious, freely-moving rats. Biochem. Biophys. Res. Commun. 155, 1459—1463. Nishimoto, N., 2010. Interleukin-6 as a therapeutic target in candidate inflammatory diseases. Clin. Pharmacol. Ther. 87, 483—487. Otten, U., Scully, J.L., Ehrhard, P.B., Gadient, R.A., 1994. Neurotrophins: signals between the nervous and immune systems. Prog. Brain Res. 103, 293—305. Papanicolaou, D.A., Petrides, J.S., Tsigos, C., Bina, S., Kalogeras, K.T., Wilder, R., Gold, P.W., Deuster, P.A., Chrousos, G.P., 1996a. Exercise stimulates interleukin-6 secretion: inhibition by glucocorticoids and correlation with catecholamines. Am. J. Physiol. 271, E601—E605. Papanicolaou, D.A., Tsigos, C., Oldfield, E.H., Chrousos, G.P., 1996b. Acute glucocorticoid deficiency is associated with plasma elevations of interleukin-6: does the latter participate in the symptomatology of the steroid withdrawal syndrome and adrenal insufficiency? J. Clin. Endocrinol. Metab. 81, 2303—2306.

81

Papanicolaou, D.A., Vgontzas, A.N., 2000. Interleukin-6: the endocrine cytokine. J. Clin. Endocrinol. Metab. 85, 1331—1333. Perry, M.G., Kirwan, J.R., Jessop, D.S., Hunt, L.P., 2009. Overnight variations in cortisol, interleukin 6, tumour necrosis factor alpha and other cytokines in people with rheumatoid arthritis. Ann. Rheum. Dis. 68, 63—68. Redwine, L., Hauger, R.L., Gillin, J.C., Irwin, M., 2000. Effects of sleep and sleep deprivation on interleukin-6, growth hormone, cortisol, and melatonin levels in humans. J. Clin. Endocrinol. Metab. 85, 3597—3603. Reiber, H., Otto, M., Trendelenburg, C., Wormek, A., 2001. Reporting cerebrospinal fluid data: knowledge base and interpretation software. Clin. Chem. Lab. Med. 39, 324—332. Reyes, T.M., Coe, C.L., 1996. Interleukin-1 beta differentially affects interleukin-6 and soluble interleukin-6 receptor in the blood and central nervous system of the monkey. J. Neuroimmunol. 66, 135—141. Reyes, T.M., Coe, C.L., 1998. Resistance of central nervous system interleukin-6 to glucocorticoid inhibition in monkeys. Am. J. Physiol. 275, R612—R618. Rohleder, N., Aringer, M., Boentert, M., 2012. Role of interleukin-6 in stress, sleep, and fatigue. Ann. N. Y. Acad. Sci. 1261, 88—96. Rose-John, S., Heinrich, P.C., 1994. Soluble receptors for cytokines and growth factors: generation and biological function. Biochem J. 300 (Pt 2) 281—290. Rose-John, S., Waetzig, G.H., Scheller, J., Grotzinger, J., Seegert, D., 2007. The IL-6/sIL-6R complex as a novel target for therapeutic approaches. Expert Opin. Ther. Targets. 11, 613—624. Sadeh, A., Alster, J., Urbach, D., Lavie, P.A., 1989. Actigraphically based automatic bedtime sleep-wake scoring: validity and clinical applications. Ambul. Monit. 2, 209—216. Salas, M.A., Evans, S.W., Levell, M.J., Whicher, J.T., 1990. Interleukin-6 and ACTH act synergistically to stimulate the release of corticosterone from adrenal gland cells. Clin. Exp. Immunol. 79, 470—473. Sanceau, J., Kaisho, T., Hirano, T., Wietzerbin, J., 1995. Triggering of the human interleukin-6 gene by interferon-gamma and tumor necrosis factor-alpha in monocytic cells involves cooperation between interferon regulatory factor-1, NF kappa B, and Sp1 transcription factors. J. Biol. Chem. 270, 27920—27931. Scheller, J., Chalaris, A., Schmidt-Arras, D., Rose-John, S., 2011. The pro- and anti-inflammatory properties of the cytokine interleukin-6. Biochim. Biophys. Acta 1813, 878—888. Seiler, W., Muller, H., Hiemke, C., 1995. Diurnal variations of plasma interleukin-6 in man: methodological implications of continuous use of indwelling cannulae. Ann. N. Y. Acad. Sci. 762, 468—470. Sewlall, S., Pillay, V., Danckwerts, M.P., Choonara, Y.E., Ndesendo, V.M., du Toit, L.C., 2010. A timely review of state-of-the-art chronopharmaceuticals synchronized with biological rhythms. Curr. Drug. Deliv. 7, 370—388. Sondergaard, S.R., Ostrowski, K., Ullum, H., Pedersen, B.K., 2000. Changes in plasma concentrations of interleukin-6 and interleukin-1 receptor antagonists in response to adrenaline infusion in humans. Eur. J. Appl. Physiol. 83, 95—98. Sothern, R.B., Roitman-Johnson, B., Kanabrocki, E.L., Yager, J.G., Fuerstenberg, R.K., Weatherbee, J.A., Young, M.R., Nemchausky, B.M., Scheving, L.E., 1995a. Circadian characteristics of interleukin-6 in blood and urine of clinically healthy men. In Vivo 9, 331—339. Sothern, R.B., Roitman-Johnson, B., Kanabrocki, E.L., Yager, J.G., Roodell, M.M., Weatherbee, J.A., Young, M.R., Nenchausky, B.M., Scheving, L.E., 1995b. Circadian characteristics of circulating interleukin-6 in men. J. Allergy Clin. Immunol. 95, 1029— 1035. Spies, C.M., Cutolo, M., Straub, R.H., Burmester, G.R., Buttgereit, F., 2010. More night than day—circadian rhythms in polymyalgia rheumatica and ankylosing spondylitis. J. Rheumatol. 37, 894—899.

82 Spooren, A., Kolmus, K., Laureys, G., Clinckers, R., De Keyser, J., Haegeman, G., Gerlo, S., 2011. Interleukin-6, a mental cytokine. Brain Res. Rev. 67, 157—183. Srinivasan, V., Spence, D.W., Pandi-Perumal, S.R., Trakht, I., Cardinali, D.P., 2008a. Therapeutic actions of melatonin in cancer: possible mechanisms. Integr. Cancer Ther. 7, 189—203. Srinivasan, V., Spence, D.W., Trakht, I., Pandi-Perumal, S.R., Cardinali, D.P., Maestroni, G.J., 2008b. Immunomodulation by melatonin: its significance for seasonally occurring diseases. Neuroimmunomodulation 15, 93—101. Steensberg, A., Dalsgaard, M.K., Secher, N.H., Pedersen, B.K., 2006. Cerebrospinal fluid IL-6 HSP72, and TNF-alpha in exercising humans. Brain Behav. Immun. 20, 585—589. Stenlof, K., Wernstedt, I., Fjallman, T., Wallenius, V., Wallenius, K., Jansson, J.O., 2003. Interleukin-6 levels in the central nervous system are negatively correlated with fat mass in overweight/ obese subjects. J. Clin. Endocrinol. Metab. 88, 4379—4383. Steptoe, A., Hamer, M., Chida, Y., 2007. The effects of acute psychological stress on circulating inflammatory factors in humans: a review and meta-analysis. Brain Behav. Immun. 21, 901—912. Sternberg, E.M., 2001. Neuroendocrine regulation of autoimmune/ inflammatory disease. J. Endocrinol. 169, 429—435. Stouthard, J.M., Romijn, J.A., Van der Poll, T., Endert, E., Klein, S., Bakker, P.J., Veenhof, C.H., Sauerwein, H.P., 1995. Endocrinologic and metabolic effects of interleukin-6 in humans. Am. J. Physiol. 268, E813—E819. Streit, W.J., Hurley, S.D., McGraw, T.S., Semple-Rowland, S.L., 2000. Comparative evaluation of cytokine profiles and reactive gliosis supports a critical role for interleukin-6 in neuron-glia signaling during regeneration. J. Neurosci. Res. 61, 10—20. Swartz, K.R., Liu, F., Sewell, D., Schochet, T., Campbell, I., Sandor, M., Fabry, Z., 2001. Interleukin-6 promotes post-traumatic healing in the central nervous system. Brain Res. 896, 86—95. Tilgner, J., Volk, B., Kaltschmidt, C., 2001. Continuous interleukin-6 application in vivo via macroencapsulation of interleukin-6expressing COS-7 cells induces massive gliosis. Glia 35, 234—245. Tominaga, T., Fukata, J., Naito, Y., Usui, T., Murakami, N., Fukushima, M., Nakai, Y., Hirai, Y., Imura, H., 1991. Prostaglandindependent in vitro stimulation of adrenocortical steroidogenesis by interleukins. Endocrinology 128, 526—531.

A. Agorastos et al. van Gool, J., van Vugt, H., Helle, M., Aarden, L.A., 1990. The relation among stress, adrenalin, interleukin 6 and acute phase proteins in the rat. Clin. Immunol. Immunopathol. 57, 200—210. Van Wagoner, N.J., Oh, J.W., Repovic, P., Benveniste, E.N., 1999. Interleukin-6 (IL-6) production by astrocytes: autocrine regulation by IL-6 and the soluble IL-6 receptor. J. Neurosci. 19, 5236—5244. Vankelecom, H., Carmeliet, P., Van Damme, J., Billiau, A., Denef, C., 1989. Production of interleukin-6 by folliculo-stellate cells of the anterior pituitary gland in a histiotypic cell aggregate culture system. Neuroendocrinology 49, 102—106. Vgontzas, A.N., Bixler, E.O., Lin, H.M., Prolo, P., Trakada, G., Chrousos, G.P., 2005. IL-6 and its circadian secretion in humans. Neuroimmunomodulation 12, 131—140. Vgontzas, A.N., Chrousos, G.P., 2002. Sleep, the hypothalamic— pituitary—adrenal axis, and cytokines: multiple interactions and disturbances in sleep disorders. Endocrinol. Metab. Clin. North Am. 31, 15—36. Vgontzas, A.N., Papanicolaou, D.A., Bixler, E.O., Lotsikas, A., Zachman, K., Kales, A., Prolo, P., Wong, M.L., Licinio, J., Gold, P.W., Hermida, R.C., Mastorakos, G., Chrousos, G.P., 1999. Circadian interleukin-6 secretion and quantity and depth of sleep. J. Clin. Endocrinol. Metab. 84, 2603—2607. Vgontzas, A.N., Zoumakis, M., Bixler, E.O., Lin, H.M., Prolo, P., VelaBueno, A., Kales, A., Chrousos, G.P., 2003. Impaired nighttime sleep in healthy old versus young adults is associated with elevated plasma interleukin-6 and cortisol levels: physiologic and therapeutic implications. J. Clin. Endocrinol. Metab. 88, 2087—2095. Vgontzas, A.N., Zoumakis, M., Papanicolaou, D.A., Bixler, E.O., Prolo, P., Lin, H.M., Vela-Bueno, A., Kales, A., Chrousos, G.P., 2002. Chronic insomnia is associated with a shift of interleukin-6 and tumor necrosis factor secretion from nighttime to daytime. Metabolism 51, 887—892. Vicennati, V., Vottero, A., Friedman, C., Papanicolaou, D.A., 2002. Hormonal regulation of interleukin-6 production in human adipocytes. Int. J. Obes. Relat. Metab. Disord. 26, 905—911. Wilder, R.L., 1995. Neuroendocrine-immune system interactions and autoimmunity. Annu. Rev. Immunol. 13, 307—338. Zhou, D., Kusnecov, A.W., Shurin, M.R., DePaoli, M., Rabin, B.S., 1993. Exposure to physical and psychological stressors elevates plasma interleukin 6: relationship to the activation of hypothalamic—pituitary—adrenal axis. Endocrinology 133, 2523—2530.