Molecular Psychiatry (2017) 00, 17

2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved 1359-4184/17
www.nature.com/mp

ORIGINAL ARTICLE
Selective increase of cerebrospinal uid IL-6 during
experimental systemic inammation in humans: association
with depressive symptoms
H Engler1, P Brendt2, J Wischermann2, A Wegner3, R Rhling1, T Schoemberg4, U Meyer5, R Gold6, J Peters2, S Benson1 and
M Schedlowski1

Systemic inammation is accompanied by profound behavioral and mood changes that resemble symptoms of depression.
Findings in animals suggest that pro-inammatory cytokines released by activated immune cells in the periphery evoke these
behavioral symptoms by driving inammatory changes in the brain. However, experimental data in humans are lacking. Here we
demonstrate in healthy male volunteers (10 endotoxin treated, 8 placebo treated) that intravenous administration of low-dose
endotoxin (0.8 ng/kg body weight), a prototypical pathogen-associated molecular pattern that activates the innate immune system,
not only induces a signicant increase in peripheral blood cytokine concentrations (that is, tumor necrosis factor-, interleukin
(IL)-6, IL-10) but also results, with some latency, in a robust and selective increase of IL-6 in the cerebrospinal uid (CSF). Moreover,
we found a strong association between the endotoxin-induced increase of IL-6 in the CSF and the severity of mood impairment,
with larger increases in CSF IL-6 concentration followed by a greater deterioration in mood. Taken together, these ndings suggest
that the appearance of depressive symptoms in inammatory conditions might be primarily linked to an increase in central IL-6
concentration, identifying IL-6 as a potential therapeutic target in mood disorders.

Molecular Psychiatry advance online publication, 31 January 2017; doi:10.1038/mp.2016.264

INTRODUCTION behavioral changes (for example, anhedonia, social withdrawal,

Inammation and immune dysregulation are increasingly recog-
nized as important factors in the etiology and pathophysiology of
neuropsychiatric disorders.1,2 An extensive body of evidence
particularly supports a bidirectional link between inammation
and mood disorders.3,4 On the one hand, a substantial proportion
of patients with major depression or bipolar disorder exhibit signs
of systemic inammation, including increased concentrations of
inammatory mediators (for example, cytokines and acute phase
proteins) in the blood or cerebrospinal uid (CSF).59 On the
other hand, major depression is highly prevalent in chronic
inammatory conditions such as rheumatoid arthritis, psoriasis,
inammatory bowel diseases and multiple sclerosis,1013 and anti-
inammatory therapies effectively improve mood in these
patients.14,15 Moreover, a large fraction of patients undergoing
cytokine therapy for the treatment of cancer or hepatitis C virus
infection develop depressive symptoms.1618 Finally, acute
systemic inammation is accompanied by profound behavioral
changes (for example, deterioration in mood, anhedonia, cogni-
tive decits, fatigue), commonly referred to as sickness behavior,
that show striking similarities to symptoms of depression.19,20
Experimental ndings in animals suggest that cytokines have a
key role in mediating the behavioral effects of inammation.
Administration of pro-inammatory cytokines such as interleukin
(IL)-1, IL-6, and tumor necrosis factor (TNF)- to rodents induces
decreased activity) reminiscent of human depression, and several
afferent pathways have been identied through which peripheral
cytokine signals can reach the brain.20,21 These routes include
activation of vagal afferent bers projecting to the nucleus of the
solitary tract,22 active transport across the bloodbrain barrier
(BBB) via cytokine-specic saturable transporters23 and diffusion in
brain areas lacking a BBB such as the circumventricular organs
and the choroid plexus (CP).24 Engagement of immune-to-brain
pathways triggers inammatory changes within the brain,
ultimately affecting neuronal activity and neurotransmitter release
in brain areas critically involved in mood regulation such as limbic
and cortical regions.4,25,26 Together, these data support the idea
that peripheral inammatory responses drive inammation in the
brain leading to depressive symptoms. However, while rodent
models have provided important insights into the cellular and
molecular mechanisms underlying inammation-induced beha-
vioral changes, human emotional states cannot be sufciently
modeled in animals. This calls for translational approaches linking
peripheral inammation and affective symptoms in humans.
In healthy humans, systemic inammation can be experimen-
tally elicited by administration of puried bacterial endotoxin
(lipopolysaccharide (LPS)).27 LPS is a prototypical pathogen-
associated molecular pattern that activates the innate immune
system through a Toll-like receptor 4-dependent pathway.

1
Institute of Medical Psychology and Behavioral Immunobiology, University Hospital Essen, University of Duisburg-Essen, Essen, Germany; 2Clinic for Anesthesiology and
Intensive Care Medicine, University Hospital Essen, University of Duisburg-Essen, Essen, Germany; 3Department of Orthopedic and Trauma Surgery, University Hospital Essen,
University of Duisburg-Essen, Essen, Germany; 4Department of Neurosurgery, University Hospital Essen, University of Duisburg-Essen, Essen, Germany; 5Institute of Pharmacology
and Toxicology, University of Zurich-Vetsuisse, Zurich, Switzerland and 6Department of Neurology, St Josef-Hospital, Ruhr University, Bochum, Germany. Correspondence:
Professor H Engler, Institute of Medical Psychology and Behavioral Immunobiology, University Hospital Essen, University of Duisburg-Essen, Hufelandstrasse 55, Essen D-45122,
Germany.
E-mail: harald.engler@uk-essen.de
Received 12 August 2016; revised 3 November 2016; accepted 14 December 2016
 Endotoxin-induced CSF IL-6 and depressive symptoms
H Engler et al
2
Intravenous injection of endotoxin rapidly leads to increased
cytokine concentrations in the circulation, and this is accompanied
by affective symptoms, such as negative mood, anhedonia and
social disconnection resembling human depression.2832 A recent
positron emission tomographic study has provided initial evi-
dence that LPS administration might also be associated with
inammatory changes in the human brain,33 but this has not been
conrmed by biochemical ndings. Moreover, the temporal
dynamics and molecular signature of the central inammatory
response and its potential association with behavioral symptoms
are unknown.
One feasible approach of measuring inammatory responses in
the human central nervous system is the analysis of cytokines in
the CSF, which is produced in the CP and exchanges with brain
interstitial uid. CSF can be safely obtained by lumbar puncture
and CSF cytokine concentrations are clinically used as markers of
central inammation. In the present study, we simultaneously
assessed plasma and CSF cytokine responses to low-dose
endotoxin administration in healthy volunteers over a period of
24 h and explored the potential associations between peripheral
and central cytokine levels as well as mood changes. We
hypothesized that experimental endotoxemia would result in
increased cytokine concentrations both in the circulation and CSF.
Furthermore, we expected to nd positive associations between
CSF levels of pro-inammatory cytokines and the severity of
depressive symptoms.
MATERIALS AND METHODS
Participants
Eighteen healthy male volunteers with a mean age of 27.8 1.2 years
(range: 2042 years) and a mean body mass index of 24.6 0.6 kg/m2
(range: 20.428.4 kg/m2), all of Caucasian ethnicity, were recruited by
public advertisement and underwent an extensive screening and safety
procedure consisting of a physical examination, a structured personal
interview and the measurement of blood and clinical chemistry param-
eters (that is, complete blood cell count, C-reactive protein (CRP), coagula-
tion factors, liver enzymes and renal parameters). General exclusion criteria
were any preexisting or current physical or psychiatric illness, body mass
index o18 or 29 kg/m2, current medications, smoking, regular alcohol
use (44 drinks per week) and anxiety and/or depression scores exceeding
published cutoffs of the Hospital Anxiety and Depression Scale.34
Participants were instructed to refrain from excessive exercise for 48 h
before testing. The study was conducted in accordance with the Declaration
of Helsinki and was approved by the Institutional Ethics Review Board
of the Medical Faculty of the University of Duisburg-Essen (Approval
No. 13-5408-BO). Signed informed consent was obtained and subjects
received nancial compensation for their participation in the study.
Study design
The study was conducted in the Clinic of Anesthesiology and Intensive
Care Medicine, University Hospital Essen and was supervised by a senior
anesthesiologist, with a neurosurgeon as an independent monitor.
Subjects were randomly assigned to receive either endotoxin (N = 10) or
placebo (N = 8). Randomization was conducted using a web-based tool
(www.randomizer.org). Upon arrival at the study site, participants rst
underwent a neurological check-up and were then prepared for the study.
On each arm, an intravenous catheter was inserted into a forearm vein for
repeated blood collection and LPS/placebo injection, respectively. Subse-
quently, a local anesthetic (Scandicain 1%, AstraZeneca, Wedel, Germany)
was injected subcutaneously at the lumbar L3/4 interspace and a catheter
(IntraLong Set, 27 G catheter, Pajunk Geisingen, Germany) was inserted
into the intrathecal space for serial CSF collection. After a resting period of
30 min, a rst blood and CSF sample were obtained (baseline). Thirty
minutes later, subjects received in a randomized and double-blinded
manner an intravenous injection of either endotoxin (0.8 ng kg 1 of body
weight; reference standard endotoxin from Escherichia coli, lot H0K354;
United States Pharmacopeia, Rockville, MD, USA) or placebo (sterile,
pyrogen-free isotonic NaCl solution, B Braun Melsungen, Melsungen,
Germany). LPS had been subjected to a microbial safety testing routine by
the German Federal Agency for Sera and Vaccines (Paul-Ehrlich Institute,
Molecular Psychiatry (2017), 1 7
Langen, Germany) and was stored in endotoxin-free borosilicate tubes
(Pyroquant Diagnostik, Mrfelden-Waldorf, Germany) at 20 C until use.
Additional blood and CSF samples were collected 1, 2, 3, 4, 6 and 24 h after
the injection. Heart rate and arterial blood pressure were continuously
recorded (Siemens SC 9000 XL, Drger Medical, Lbeck, Germany), and
body temperature was measured every 30 min with a tympanic thermo-
meter (Genius 2, Covidien, Tullamore, Ireland). After collection of the 6-h
sample, participants were transferred to a hospital ward room where they
stayed overnight under medical surveillance. On the next day (that is, 24 h
after LPS/placebo injection), a nal blood and CSF sample were collected,
and catheters were removed. After a neurological examination, partici-
pants were discharged and instructed to refrain from physical activity for
7 days. A medical follow-up was conducted 1 week after the study.
Cytokine and CRP concentrations
Blood was drawn into tubes containing either EDTA or clot activator
(S-Monovette, Sarstedt, Nmbrecht, Germany) and CSF was collected in
pyrogen-free polypropylene tubes (Falcon, BD Biosciences, Heidelberg,
Germany). Plasma/serum and cell-free CSF were obtained by centrifugation
(2000 g, 10 min, 4 C) and were stored at 80 C until analysis. Concentra-
tions of TNF-, IL-6, IL-10 and IL-1 in plasma and CSF were measured by
enzyme-linked immunosorbent assay (ELISA) (Human Quantikine ELISA,
R&D Systems, Minneapolis, MN, USA) according to the manufacturers
protocols. The sensitivity of the assays was 0.11 pg ml 1 for TNF-,
0.70 pg ml 1 for IL-6, 0.09 pg ml 1 for IL-10 and 0.14 pg ml 1 for IL-1.
Mean inter- and intra-assay coefcients of variation were 10%. As IL-1
was below the limit of detection in the majority of plasma and CSF
samples, IL-1 data are not presented. Serum concentration of CRP was
measured by ELISA (CRP high-sensitive ELISA, IBL International, Hamburg,
Germany) according to the manufacturers instructions. The sensitivity of
the assay was 0.02 mg l 1 and mean interassay and intra-assay coefcients
of variation were o 7%.
Cell counts
Complete blood counts, including white blood cell differential, were
determined with a Sysmex KX-N21 hematology analyzer (Sysmex Europe,
Norderstedt, Germany). CSF cell counts were obtained using a Sysmex
XE-500 Automated Hematology System (Sysmex Europe).
Albumin and IgG concentrations
Serum and CSF albumin and immunoglobulin (IgG) concentrations were
determined by nephelometry on a Beckman Coulter IMMAGE Immuno-
chemistry System (Beckman Coulter, Krefeld, Germany). The CSF (mg l 1)
to serum (g l 1) ratios of albumin and IgG served as indices to evaluate the
integrity of the BBB, with increases in these ratios indicating increased
permeability.
Mood assessment
Normal mood (euthymia) and depressed mood (dysthymia) were assessed
before (baseline) as well as 1, 2, 3, 4, 6 and 24 h after endotoxin or placebo
injection with two ve-item subscales of the state version of the
standardized and validated State-Trait Anxiety Depression Inventory.35
Statistical analysis
Data analysis was performed using SPSS 22.0 (SPSS, Chicago, IL, USA) and
the level of signicance was set at Po0.05. Normality of residuals was
examined using the ShapiroWilk test and data were log-transformed
when necessary. Group differences in behavioral and physiological
measures were analyzed by repeated-measures analysis of variance
(ANOVA) with LPS as between-subject factor and time as within-
subject factor. GreenhouseGeisser correction was performed when the
assumption of sphericity was violated. If ANOVA revealed a signicant
LPS time interaction, independent samples t-tests (two-tailed) were
computed to compare LPS and placebo groups at the different sampling
points. Only interaction effects are reported. A priori power calculations
using G*Power 3.1 (Faul et al.36) revealed sufcient statistical power
(483%) to detect medium effect sizes (f = 0.25) in repeated-measures
ANOVAs for plasma/CSF cytokines and mood parameters, which were the
primary readouts. The Spearmens rank correlation coefcient (Spearmans
; one tailed) was used to explore potential associations between cytokine
concentrations and behavioral measures within the endotoxin group. The
2017 Macmillan Publishers Limited, part of Springer Nature.
 Endotoxin-induced CSF IL-6 and depressive symptoms
H Engler et al
3
BenjaminiHochberg procedure was used to control for the false discovery QIgG: F(6,96) = 0.73, P = 0.63; Figure 2a). Furthermore, leukocyte
rate (q = 0.05) in multiple comparisons. counts in the CSF did not signicantly differ between treatment
groups at 2 and 6 h after injection, indicating that administration
of low-dose LPS did not result in leukocyte inltration into the CSF
RESULTS (F(1,16) = 0.33, P = 0.86; Figure 2b).
Sample characteristics
Ten subjects were injected with endotoxin (0.8 ng kg 1) and eight Mood changes
subjects served as controls, receiving placebo. Demographic and Participants who had received endotoxin exhibited pronounced
clinical characteristics of the two treatment groups are shown in
mood changes, whereas placebo-injected subjects did not show
Table 1. Groups did not signicantly differ in age and body mass
these behavioral symptoms (Figure 3). Specically, administration
index. Hematological parameters, inammatory markers, coagula-
of LPS resulted in worsened mood as evident from a signicant
tion factors and liver and renal parameters as well as Hospital
increase in dysthymia (F(6,96) = 2.50, P = 0.027), with the largest
Anxiety and Depression Scale anxiety and depression scores were
effects observed at 3 h after LPS injection. Although euthymia
within normal range, without evidence for group differences at
scores showed a slight decrease in LPS compared with placebo-
baseline.
injected subjects over time, no signicant time group interaction
was found (F(6,96) = 0.44, P = 0.85).
Inammatory response
Low-dose endotoxin administration induced an acute systemic Relationship between cytokine and mood changes
inammatory response as evident from a signicant elevation in
To explore the potential associations between the endotoxin-
body temperature (F(6,96) = 11.2, P o 0.001; Figure 1a), a pro-
induced increase in CSF IL-6 concentration and the deterioration
nounced leukocytosis (F(6,96) = 8.9, P o0.001; Figure 1b) and a
in mood, we additionally performed correlation analyses between
signicant rise in serum CRP levels (F(3,48) = 159.93, P o 0.001;
CSF IL-6 levels and dysthymia scores from 2 to 6 h after LPS
Figure 1c) as well as signicant increases in plasma concentrations
injection. Dysthymia scores showed strong positive correlations
of TNF- (F(6,96) = 65.76, P o 0.001), IL-6 (F(6,96) = 26.18, P o 0.001)
with CSF IL-6 concentrations across all time points investigated
and IL-10 (F(6,96) = 21.91, P o 0.001; Figure 1d). The increase in
(Supplementary Table S1). Correlations between 2 h CSF IL-6 and
plasma cytokines was transient, peaking at 2 h after LPS injection,
3 h dysthymia score (Spearmans = 0.759) as well as between 3 h
and rapidly declining thereafter. The endotoxin-induced cytokine
CSF IL-6 and 6 h dysthymia score (Spearmans = 0.817) remained
response in the CSF markedly differed from that in the peripheral
signicant after false discovery rate correction for multiple
blood (Figure 1e). First, a signicant increase in CSF cytokines
comparisons (Po 0.05; one tailed). No signicant correlations
was found only for IL-6 (F(6,96) = 4.62, P = 0.005), whereas
between plasma cytokine concentrations and dysthymia scores
CSF concentrations of TNF- (F(6,96) = 1.54, P = 0.17) and IL-10
were found.
(F(6,96) = 0.36, P = 0.91) remained unaffected. Second, the
endotoxin-induced increase of IL-6 in the CSF was time delayed
relative to the increase of IL-6 in the circulation, with signicant DISCUSSION
group differences between 3 and 6 h after injection, and IL-6 levels
We believe our study demonstrates for the rst time that
in the CSF continued to increase even after the peripheral IL-6
intravenous administration of low-dose endotoxin in humans
response had already vanished.
not only evokes peripheral pro- and anti-inammatory cytokine
responses but also results in a robust and selective increase of IL-6
BBB integrity and CSF cell counts in the CSF, suggesting that the molecular signature of the
Both the CSF/serum albumin (QAlb) and the CSF/serum IgG (QIgG) inammatory response substantially differs between the periph-
ratios were within normal range in all subjects and were not eral and central compartments. Moreover, we found a strong
affected by endotoxin administration (QAlb: F(6,96) = 0.50, P = 0.81; association between the endotoxin-induced increase of IL-6 in the
CSF and the severity of mood changes, showing that higher CSF
Table 1. Demographic and clinical characteristics IL-6 concentrations were followed by a greater deterioration
in mood.
Variable Placebo Endotoxin P-value One intriguing nding of the present study was the selective
(N = 8) (N = 10) increase of IL-6 in the CSF, while CSF concentrations of other key
cytokines such as TNF-, IL-1 and IL-10 remained unaffected.
Age (years) 30.2 1.9 25.8 1.1 0.07 Similar observations have been made in hepatitis C virus patients
BMI (kg/m2) 24.9 0.8 24.7 1.0 0.92
under cytokine therapy, exhibiting increased CSF concentrations
WBC count (103 l 1) 5.7 0.5 5.6 0.6 0.91
Hematocrit (%) 42.3 1.2 42.2 0.8 0.96 of IL-6 but not of TNF- and IL-1 after chronic IFN-/ribavirin
CRP (mg l 1) 0.51 0.16 0.59 0.24 0.34 treatment.37 Our data in healthy subjects extend and corroborate
TNF- (pg ml 1) 1.30 0.42 1.28 0.29 0.96 these ndings and strengthen the notion that in humans there is
IL-6 (pg ml 1) 2.50 0.38 3.22 1.06 0.57 no one-to-one translation between cytokine changes in the
IL-10 (pg ml 1) 0.29 0.07 0.47 0.11 0.20 systemic circulation and the CSF. Hence, our ndings differ from
Fibrinogen (mg dl 1) 210 9 209 8 0.94 those in rodents where peripheral LPS administration (intrave-
INR 1.01 0.01 1.02 0.01 0.58 nously or intraperitoneally) increased concentrations of various
GGT (U l 1) 25 7 20 3 0.47
Serum creatinine 1.08 0.03 1.01 0.04 0.15
cytokines, including TNF-, IL-1 and IL-6, in both circulation and
(mg dl 1) brain.3840 However, the latter effects emerged in response to
HADS Anxiety Score 3.9 1.2 3.0 1.0 0.51 much higher (104- to 106-fold) endotoxin doses and may have
HADS Depression Score 2.9 0.6 1.6 0.4 0.12 been accompanied by disruption of the BBB. In addition, cytokine
analyses in animals were carried out in brain tissue and not in CSF,
Abbreviations: BMI, body mass index; CRP, C-reactive protein; GGT, gamma
leaving the possibility that these differences in human and animal
glutamyl transpeptidase; HADS, Hospital Anxiety and Depression Scale; IL,
interleukin; INR, International Normalized Ratio; TNF, tumor necrosis factor; ndings might be related to differences in sampling sites. This
WBC, white blood cell. Data are presented as mean s.e.m. issue needs further investigation and calls for comparative time-
course studies assessing in parallel cytokine proles in peripheral

2017 Macmillan Publishers Limited, part of Springer Nature. Molecular Psychiatry (2017), 1 7
 Endotoxin-induced CSF IL-6 and depressive symptoms
H Engler et al
4

Figure 1. Low-dose endotoxin administration induces a systemic and central inammatory response. (a) Body temperature, (b) white blood
cell (WBC) counts, (c) serum C-reactive protein (CRP) concentration, (d) plasma and (e) cerebrospinal uid (CSF) concentrations of tumor
necrosis factor (TNF)-, interleukin (IL)-6 and IL-10 before and after intravenous injection of endotoxin (0.8 ng kg 1 lipopolysaccharide; N = 10)
or placebo (N = 8). Means and s.e.m. are shown. Repeated-measure analysis of variance with post hoc independent t-tests, *P 0.05; **P 0.01;
***P 0.001.

blood, CSF and brain tissue of animals challenged with more
moderate doses of LPS.
Several mechanisms may lead to the endotoxin-induced
increase of IL-6 concentration in the CSF. Given that the integrity
of the bloodCSF barrier (BCSFB)/BBB was apparently not affected
in our LPS-treated subjects, blood-borne IL-6 may have entered
the CSF either by diffusion from brain areas with incomplete BBB
such as the circumventricular organs and the CP or by active
transport across the BBB/BCSFB via cytokine-specic transport
molecules. Such saturable transport systems have been described
not only for IL-6 but also for other cytokines.41 As both diffusion
and active transport require time, this might explain the delayed
increase of IL-6 in the CSF relative to its rise in the circulation.
However, IL-6 levels in the CSF continued to increase while the
plasma IL-6 response had already vanished, suggesting that other
mechanisms might be involved as well. Studies in animals have
shown that CP epithelial cells, forming the BCSFB, express both
Toll-like receptor 4 and cytokine receptors at the basolateral
(blood-facing) side, and stimulation of these receptors by
circulating LPS or cytokines can trigger IL-6 synthesis in these
cells, ultimately leading to its secretion into the CSF.4245 Thus
local production of IL-6 in the CP may represent an additional
mechanism contributing to the endotoxin-induced increase of IL-6
in the CSF. Finally, as the CSF exchanges with brain interstitial

Molecular Psychiatry (2017), 1 7 2017 Macmillan Publishers Limited, part of Springer Nature.

Figure 2. Low-dose endotoxin administration does not affect bloodcerebrospinal uid barrier/bloodbrain barrier integrity and
cerebrospinal uid (CSF) cell counts. (a) CSF/serum albumin quotient, (b) CSF/serum immunoglobulin G (IgG) quotient and (c) leukocyte
numbers in the CSF before and after intravenous injection of endotoxin (0.8 ng kg 1 lipopolysaccharide; N = 10) or placebo (N = 8). Means and
s.e.m. are shown. Repeated-measure analysis of variance.
Figure 3. Systemic inammation is accompanied by mood changes.
Scores for (a) euthymia and (b) dysthymia before and after
intravenous injection of endotoxin (0.8 ng kg 1 lipopolysaccharide;
N = 10) or placebo (N = 8). Means and s.e.m. are shown. Repeated-
measure analysis of variance with post hoc independent t-tests,
*P 0.05; **P 0.01.
uid, IL-6 produced by brain residential cells (for example,
microglia, astrocytes, neurons) may have reached the CSF through
the brainCSF interface.
Another interesting nding was the strong association between
the endotoxin-induced rise of IL-6 in the CSF and the severity of
mood impairment, showing that a larger increase in CSF IL-6
concentration was followed by a greater deterioration in mood.
This not only extends previous ndings from human endotoxemia
and typhoid vaccination studies reporting correlations between
circulating IL-6 levels and changes in negative affect4650 but is
also in line with clinical observations showing that patients with
major depressive disorder exhibit elevated levels of IL-6 in
peripheral blood and CSF.5,5153 Based on the present experi-
mental ndings and considering the clinical evidence of increased
peripheral and CSF IL-6 levels in depressed patients, IL-6 might be
a promising therapeutic target in the treatment of depression,
particularly in patients exhibiting signs of systemic inammation
2017 Macmillan Publishers Limited, part of Springer Nature.
Endotoxin-induced CSF IL-6 and depressive symptoms
H Engler et al
5
and/or showing resistance to conventional antidepressant med-
ication. In this regard, humanized monoclonal antibodies against
the IL-6 receptor (IL-6R) could be one adjunct treatment option54
as they inhibit the binding of IL-6 to the soluble and membrane-
bound forms of the IL-6R, thereby blocking IL-6 signal transduc-
tion. Such trials are currently actively being pursued in patients
with major depressive disorder (for example, NCT02473289 at
ClinicalTrials.gov), but at present, no clinical data are available as
to whether these biologicals effectively reduce depressive
symptoms. However, anti-IL-6R antibody therapy in rheumatoid
arthritis patients signicantly improved fatigue, indicating bene-
cial effects on behavioral outcomes.55
Of note, CSF IL-6 concentration also gradually increased in the
placebo group, although this increase was considerably less
pronounced compared with the LPS group and not associated
with any mood changes. The observed increase of IL-6 concentra-
tion in the CSF samples of placebo-treated subjects presumably
reects local IL-6 production at the catheter insertion site rather
than changes in overall CSF IL-6 concentration. Such effects on
IL-6 levels owing to local cytokine production have been
documented in blood samples collected via indwelling venous
catheters for an extended period of time5658 and may also occur
in CSF samples collected via indwelling spinal catheters. Our
ndings in the CSF of placebo-treated subjects thereby conrm
earlier ndings in untreated healthy volunteers showing a similar
increase in CSF IL-6 concentrations after spinal catheterization,
while CSF levels of TNF-, IL-1 and IL-10 remained unaffected.59
Contamination of the CSF samples with locally produced IL-6 at
the later time points might also explain the absence of mood
changes in the placebo group. One approach of investigating this
issue in future studies could be to discard the rst drops of CSF,
likely containing locally produced IL-6, before collecting the actual
sample for analyses.
The study has several strengths but also some limitations that
warrant consideration. A notable strength is that plasma and CSF
cytokine concentrations were concurrently and repeatedly
assessed in the same individuals over a period of 24 h, taking
into account the temporal dynamics of the inammatory
response. Furthermore, unlike cytokine therapy studies that have
been carried out in antivirally treated hepatitis C virus patients, our
experiments were conducted in healthy volunteers under
standardized laboratory conditions, eliminating potentially con-
founding factors such as underlying disease or medication. One
limitation of the study is the relatively small sample size, which
Molecular Psychiatry (2017), 1 7
 Endotoxin-induced CSF IL-6 and depressive symptoms
H Engler et al
6
may have led in correlation analyses to false negative results
owing to insufcient statistical power. Therefore, the lack of
signicant correlations between plasma cytokine concentrations
on the one side and CSF IL-6 levels and mood changes on the
other side should be interpreted with caution. Another limitation
of the study is that only male participants were used. Given that
women mount stronger plasma pro-inammatory cytokine
responses to endotoxin29 and are more likely to develop mood
disorders compared with men,60 it would be important to
investigate in future studies whether sex differences in CSF
cytokine responses to peripheral endotoxin administration exist.
Additional limitations include the restricted number of cytokines
analyzed. Our selection was based on those cytokines that were
consistently found to be induced in the central nervous system of
endotoxin-treated animals. However, given the ndings of the
present study, it would have been valuable to include other
cytokines and soluble cytokine receptors as well. Finally, low-dose
endotoxemia induces only short-lived behavioral and immune
changes, and it remains to be determined whether chronic low-
grade inammation is accompanied by comparable cytokine
alterations in the CSF. Despite these limitations, we believe that
our experimental approach has a high translational value as a
putative model for affective disorders, providing important
insights into the mechanisms underlying immune-to-brain com-
munication. Furthermore, our ndings suggest that the appear-
ance of depressive symptoms in inammatory conditions might
be primarily linked to an increase in central IL-6 concentration.
Therefore, IL-6 and its signal transduction pathway may represent
a promising therapeutic target in the treatment of mood
disorders.
CONFLICT OF INTEREST
The authors declare no conict of interest.
ACKNOWLEDGMENTS
We thank Ursula Brecklinghaus and Alexandra Kornowski for excellent technical
assistance and Bettina Lschner and Dr Ingo Spreitzer (Paul-Ehrlich-Institute, Langen,
Germany) for endotoxin safety testing. This work was supported by an internal
research grant from the Medical Faculty of the University of Duisburg-Essen (IFORES)
awarded to HE.
REFERENCES
1 Khandaker GM, Cousins L, Deakin J, Lennox BR, Yolken R, Jones PB. Inammation
and immunity in schizophrenia: implications for pathophysiology and treatment.
Lancet Psychiatry 2015; 2: 258270.
2 Haroon E, Raison CL, Miller AH. Psychoneuroimmunology meets neuropsycho-
pharmacology: translational implications of the impact of inammation on
behavior. Neuropsychopharmacology 2012; 37: 137162.
3 Rosenblat JD, Cha DS, Mansur RB, McIntyre RS. Inamed moods: a review of the
interactions between inammation and mood disorders. Prog Neuropsycho-
pharmacol Biol Psychiatry 2014; 53: 2334.
4 Miller AH, Raison CL. The role of inammation in depression: from evolutionary
imperative to modern treatment target. Nat Rev Immunol 2016; 16: 2234.
5 Dowlati Y, Herrmann N, Swardfager W, Liu H, Sham L, Reim EK et al. A meta-
analysis of cytokines in major depression. Biol Psychiatry 2010; 67: 446457.
6 Howren MB, Lamkin DM, Suls J. Associations of depression with C-reactive
protein, IL-1, and IL-6: a meta-analysis. Psychosom Med 2009; 71: 171186.
7 Modabbernia A, Taslimi S, Brietzke E, Ashra M. Cytokine alterations in bipolar
disorder: a meta-analysis of 30 studies. Biol Psychiatry 2013; 74: 1525.
8 Miller AH, Maletic V, Raison CL. Inammation and its discontents: the role of
cytokines in the pathophysiology of major depression. Biol Psychiatry 2009; 65:
732741.
9 Lindqvist D, Janelidze S, Hagell P, Erhardt S, Samuelsson M, Minthon L et al.
Interleukin-6 is elevated in the cerebrospinal uid of suicide attempters and
related to symptom severity. Biol Psychiatry 2009; 66: 287292.
10 Matcham F, Rayner L, Steer S, Hotopf M. The prevalence of depression in
rheumatoid arthritis: a systematic review and meta-analysis. Rheumatology 2013;
52: 21362148.
Molecular Psychiatry (2017), 1 7
11 Fleming P, Roubille C, Richer V, Starnino T, McCourt C, McFarlane A et al. Effect of
biologics on depressive symptoms in patients with psoriasis: a systematic review.
J Eur Acad Dermatol Venereol 2015; 29: 10631070.
12 Graff LA, Walker JR, Bernstein CN. Depression and anxiety in inammatory bowel
disease: a review of comorbidity and management. Inamm Bowel Dis 2009; 15:
11051118.
13 Patten SB, Beck CA, Williams JV, Barbui C, Metz LM. Major depression in multiple
sclerosis: a population-based perspective. Neurology 2003; 61: 15241527.
14 Tyring S, Gottlieb A, Papp K, Gordon K, Leonardi C, Wang A et al. Etanercept and
clinical outcomes, fatigue, and depression in psoriasis: double-blind placebo-
controlled randomised phase III trial. Lancet 2006; 367: 2935.
15 Abbott R, Whear R, Nikolaou V, Bethel A, Coon JT, Stein K et al. Tumour necrosis
factor- inhibitor therapy in chronic physical illness: a systematic review and
meta-analysis of the effect on depression and anxiety. J Psychosom Res 2015; 79:
175184.
16 Capuron L, Hauser P, Hinze-Selch D, Miller AH, Neveu PJ. Treatment of cytokine-
induced depression. Brain Behav Immun 2002; 16: 575580.
17 Raison CL, Capuron L, Miller AH. Cytokines sing the blues: inammation and the
pathogenesis of depression. Trends Immunol 2006; 27: 2431.
18 Dowell NG, Cooper EA, Tibble J, Voon V, Critchley HD, Cercignani M et al. Acute
changes in striatal microstructure predict the development of interferon-alpha
induced fatigue. Biol Psychiatry 2016; 79: 320328.
19 Maes M, Berk M, Goehler L, Song C, Anderson G, Galecki P et al. Depression and
sickness behavior are Janus-faced responses to shared inammatory pathways.
BMC Med 2012; 10: 66.
20 Dantzer R, O'Connor JC, Freund GG, Johnson RW, Kelley KW. From inammation
to sickness and depression: when the immune system subjugates the brain. Nat
Rev Neurosci 2008; 9: 4656.
21 Quan N, Banks WA. Brain-immune communication pathways. Brain Behav Immun
2007; 21: 727735.
22 Goehler LE, Gaykema RP, Hansen MK, Anderson K, Maier SF, Watkins LR. Vagal
immune-to-brain communication: a visceral chemosensory pathway. Auton
Neurosci 2000; 85: 4959.
23 Banks WA. The blood-brain barrier in psychoneuroimmunology. Neurol Clin 2006;
24: 413419.
24 Vitkovic L, Konsman JP, Bockaert J, Dantzer R, Homburger V, Jacque C. Cytokine
signals propagate through the brain. Mol Psychiatry 2000; 5: 604615.
25 Engler H, Doenlen R, Engler A, Riether C, Prager G, Niemi MB et al. Acute
amygdaloid response to systemic inammation. Brain Behav Immun 2011; 25:
13841392.
26 Poon DC, Ho YS, Chiu K, Wong HL, Chang RC. Sickness: from the focus on cyto-
kines, prostaglandins, and complement factors to the perspectives of neurons.
Neurosci Biobehav Rev 2015; 57: 3045.
27 Schedlowski M, Engler H, Grigoleit JS. Endotoxin-induced experimental systemic
inammation in humans: a model to disentangle immune-to-brain communica-
tion. Brain Behav Immun 2014; 35: 18.
28 Reichenberg A, Yirmiya R, Schuld A, Kraus T, Haack M, Morag A et al. Cytokine-
associated emotional and cognitive disturbances in humans. Arch Gen Psychiatry
2001; 58: 445452.
29 Engler H, Benson S, Wegner A, Spreitzer I, Schedlowski M, Elsenbruch S. Men and
women differ in inammatory and neuroendocrine responses to endotoxin but
not in the severity of sickness symptoms. Brain Behav Immun 2016; 52: 1826.
30 Eisenberger NI, Berkman ET, Inagaki TK, Rameson LT, Mashal NM, Irwin MR.
Inammation-induced anhedonia: endotoxin reduces ventral striatum responses
to reward. Biol Psychiatry 2010; 68: 748754.
31 Eisenberger NI, Inagaki TK, Mashal NM, Irwin MR. Inammation and social
experience: an inammatory challenge induces feelings of social disconnection in
addition to depressed mood. Brain Behav Immun 2010; 24: 558563.
32 DellaGioia N, Hannestad J. A critical review of human endotoxin administration as
an experimental paradigm of depression. Neurosci Biobehav Rev 2010; 34:
130143.
33 Sandiego CM, Gallezot JD, Pittman B, Nabulsi N, Lim K, Lin SF et al. Imaging robust
microglial activation after lipopolysaccharide administration in humans with PET.
Proc Natl Acad Sci USA 2015; 112: 1246812473.
34 Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr
Scand 1983; 67: 361370.
35 Laux L, Hock M, Bergner-Kther R, Hodapp V, Renner K-H, Merzbacher G. The
State-Trait Anxiety Depression Inventory (STADI). Hogrefe: Gttingen, Germany,
2013.
36 Faul F, Erdfelder E, Lang AG, Buchner AG. *Power 3: a exible statistical power
analysis program for the social, behavioral, and biomedical sciences. Behav Res
Methods 2007; 39: 175191.
37 Raison CL, Borisov AS, Majer M, Drake DF, Pagnoni G, Woolwine BJ et al. Activation
of central nervous system inammatory pathways by interferon-alpha: relation-
ship to monoamines and depression. Biol Psychiatry 2009; 65: 296303.
2017 Macmillan Publishers Limited, part of Springer Nature.

38 Kakizaki Y, Watanobe H, Kohsaka A, Suda T. Temporal proles of interleukin-1beta,
interleukin-6, and tumor necrosis factor-alpha in the plasma and
hypothalamic paraventricular nucleus after intravenous or intraperitoneal
administration of lipopolysaccharide in the rat: estimation by push-pull perfusion.
Endocr J 1999; 46: 487496.
39 Datta SC, Opp MR. Lipopolysaccharide-induced increases in cytokines in discrete
mouse brain regions are detectable using Luminex xMAP technology. J Neurosci
Meth 2008; 175: 119124.
40 Erickson MA, Banks WA. Cytokine and chemokine responses in serum and brain
after single and repeated injections of lipopolysaccharide: multiplex quantica-
tion with path analysis. Brain Behav Immun 2011; 25: 16371648.
41 Banks WA. The blood-brain barrier in neuroimmunology: Tales of separation and
assimilation. Brain Brain Behav Immun 2015; 44: 18.
42 Vallieres L, Rivest S. Regulation of the genes encoding interleukin-6, its receptor,
and gp130 in the rat brain in response to the immune activator lipopoly-
saccharide and the proinammatory cytokine interleukin-1beta. J Neurochem
1997; 69: 16681683.
43 LeMay LG, Vander AJ, Kluger MJ. Role of interleukin-6 in fever in rats. Am J Physiol
1990; 258: R798R803.
44 Reyes TM, Fabry Z, Coe CL. Brain endothelial cell production of a neuroprotective
cytokine, interleukin-6, in response to noxious stimuli. Brain Res 1999; 851:
215220.
45 Marques F, Sousa JC, Coppola G, Falcao AM, Rodrigues AJ, Geschwind DH et al.
Kinetic prole of the transcriptome changes induced in the choroid plexus by
peripheral inammation. J Cereb Blood Flow Metab 2009; 29: 921932.
46 Grigoleit JS, Kullmann JS, Wolf OT, Hammes F, Wegner A, Jablonowski S et al.
Dose-dependent effects of endotoxin on neurobehavioral functions in humans.
PloS One 2011; 6: e28330.
47 Wegner A, Elsenbruch S, Maluck J, Grigoleit JS, Engler H, Jager M et al.
Inammation-induced hyperalgesia: effects of timing, dosage, and negative affect
on somatic pain sensitivity in human experimental endotoxemia. Brain Behav
Immun 2014; 41: 4654.
48 Lasselin J, Elsenbruch S, Lekander M, Axelsson J, Karshikoff B, Grigoleit JS et al.
Mood disturbance during experimental endotoxemia: predictors of state anxiety
Supplementary Information accompanies the paper on the Molecular Psychiatry website (http://www.nature.com/mp)
2017 Macmillan Publishers Limited, part of Springer Nature.
Endotoxin-induced CSF IL-6 and depressive symptoms
H Engler et al
7
as a psychological component of sickness behavior. Brain Behav Immun 2016;
57: 3037, in press).
49 Wright CE, Strike PC, Brydon L, Steptoe A. Acute inammation and negative
mood: mediation by cytokine activation. Brain Behav Immun 2005; 19: 345350.
50 Harrison NA, Brydon L, Walker C, Gray MA, Steptoe A, Critchley HD. Inammation
causes mood changes through alterations in subgenual cingulate activity and
mesolimbic connectivity. Biol Psychiatry 2009; 66: 407414.
51 Sasayama D, Hattori K, Wakabayashi C, Teraishi T, Hori H, Ota M et al. Increased
cerebrospinal uid interleukin-6 levels in patients with schizophrenia and those
with major depressive disorder. J Psychiatr Res 2013; 47: 401406.
52 Goldsmith DR, Rapaport MH, Miller BJ. A meta-analysis of blood cytokine network
alterations in psychiatric patients: comparisons between schizophrenia, bipolar
disorder and depression. Mol Psychiatry 2016; 21: 16961709.
53 Valkanova V, Ebmeier KP, Allan CL. CRP, IL-6 and depression: a systematic review
and meta-analysis of longitudinal studies. J Affect Disord 2013; 150: 736744.
54 Venkiteshwaran A. Tocilizumab. mAbs 2009; 1: 432438.
55 Strand V, Singh JA. Newer biological agents in rheumatoid arthritis: impact on
health-related quality of life and productivity. Drugs 2010; 70: 121145.
56 Seiler W, Mller H, Hiemke C. Interleukin-6 in plasma collected with an indwelling
cannula reects local, not systemic, concentrations. Clin Chem 1994; 40:
17781779.
57 Gudmundsson A, Ershler WB, Goodman B, Lent SJ, Barczi S, Carnes M. Serum
concentrations of interleukin-6 are increased when sampled through an
indwelling venous catheter. Clin Chem 1997; 43: 21992201.
58 Haack M, Reichenberg A, Kraus T, Schuld A, Yirmiya R, Pollmcher T. Effects of an
intravenous catheter on the local production of cytokines and soluble cytokine
receptors in healthy men. Cytokine 2000; 12: 694698.
59 Llano DA, Li J, Warring JF, Ellis T, Devanarayan V, Witte DG, Lenz RA. Cerebrospinal
uid cytokine dynamics differ between Alzheimer disease patients and elderly
controls. Alzheimer Dis Assoc Disord 2012; 26: 322328.
60 Kessler R, McGonagle K, Zhao S, Nelson C, Hughes M, Eshleman S et al.
Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the
United States. Results from the National Comorbidity Survey. Arch Gen Psychiatry
1994; 51: 819.
Molecular Psychiatry (2017), 1 7