Content uploaded by Michael Maes
Author content
All content in this area was uploaded by Michael Maes on Sep 19, 2018
Content may be subject to copyright.
To cite this article:
Neuroendocrinol Lett 2008; 29(1): 117–124
ORIGINAL ARTICLE
Neuroendocrinology Letters Volume 29 No. 1 2008
The gut-brain barrier in major depression:
Intestinal mucosal dysfunction with an increased
translocation of LPS from gram negative
enterobacteria (leaky gut) plays a role in the
inflammatory pathophysiology of depression
Michael Maes 1, Marta Kubera 2 and Jean-Claude Leunis 3
MCare4U Outpatient Clinics, Belgium;
Department of Experimental Neuroendocrinology, Institute of Pharmacology, Polish Academy of
Sciences, Krakow, Poland;
Laboratory Ategis, Waver, Belgium.
Correspondence to: Prof. Dr. M.Maes, M.D., Ph.D. Director: M-Care4U Outpatient Clinics,
Olmenlaan 9, 2610 Antwerp, Belgium. tel: +32-3-4809282 fax: +32-3-2889185
www.michaelmaes.com e-mail : crc.mh@telenet.be
Submitted: 2008-01-06 Accepted: 2008-01-27 Published online: 2008-02-22
Key words: major depression; chronic fatigue syndrome; inflammation; enterobacteria;
leaky gut; gut permeability; cytokines; LPS; oxidative stress
Neuroendocrinol Let t 2008; 29(1):117–124 PMID: 18283240 NEL290108A12 © 2008 Neuroendocrinology Letters • www.nel.edu
Abstract
There is now evidence that major depression (MDD) is accompanied by an ac-
tivation of the inflammatory response system (IRS) and that pro-inflammatory
cytokines and lipopolysacharide (LPS) may induce depressive symptoms.
The aim of the present study was to examine whether an increased gastrointestinal
permeability with an increased translocation of LPS from gram negative bacteria
may play a role in the pathophysiology of MDD. Toward this end, the present study
examines the serum concentrations of IgM and IgA against LPS of the gram-nega-
tive enterobacteria, Hafnia Alvei, Pseudomonas Aeruginosa, Morganella Morganii,
Pseudomonas Putida, Citrobacter Koseri, and Klebsielle Pneumoniae in MDD
patients and normal controls.
We found that the prevalences and median values for serum IgM and IgA against
LPS of enterobacteria are significantly greater in patients with MDD than in nor-
mal volunteers. These differences are significant to the extent that a significant
diagnostic performance is obtained, i.e. the area under the ROC curve is 90.1%.
The symptom profiles of increased IgM and IgA levels are fatigue, autonomic and
gastro-intestinal symptoms and a subjective feeling of infection. The results show
that intestinal mucosal dysfunction characterized by an increased translocation of
gram-negative bacteria (leaky gut) plays a role in the inflammatory pathophysiol-
ogy of depression. It is suggested that the increased LPS translocation may mount
an immune response and thus IRS activation in some patients with MDD and may
induce specific “sickness behaviour” symptoms. It is suggested that patients with
MDD should be checked for leaky gut by means of the IgM and IgA panel used in
the present study and accordingly should be treated for leaky gut.
1.
2.
3.
118
Copyright © 2008 Neuroendocrinology Letters ISSN 0172–780X • www.nel.edu
Michael Maes, Marta Kubera and Jean-Claude Leunis
INTRODUCTION
There is now evidence that activation of the inflamma-
tory response system (IRS) plays a role in the patho-
physiology of major depression (MDD). This theory
has been described as the “cytokine” or “monocyte-T
lymphocyte hypothesis” of depression, although recent-
ly we described this newly discovered pathway as the
“IRS activation” theory [1–4] and now as the “inflam-
matory and neurodegenerative (I&ND) hypothesis” of
depression [5].
The IRS findings in MDD show: a) an increased pro-
duction of pro-inflammatory cytokines, such as inter-
leukin-6 (IL-6), IL-1β and tumor necrosis factor alpha
(TNFα); b) an increased expression of T lymphocyte
activation markers, such as HLA-DR+ and CD25+; c)
the presence of an acute phase response, with increased
serum levels of haptoglobin, decreased serum zinc levels
and increased serum concentrations of the alpha2 glob-
ulin fraction obtained by electrophoresis; d) and signs
of poor cellular immunity, such as lowered natural kill-
er cell cytotoxity and decreased mitogen–induced lym-
phocyte responses; and e) an increased induction of in-
doleamine oxidase (IDO) with consequent tryptophan
depletion and formation of TRYCATs (tryptophan cat-
abolites along the IDO pathway) [6].
Activation of the IRS is found in animal models of
MDD, such as the chronic mild stress and the olfac-
tory bulbectomized rat models of depression [7–10].
Newly generated rat models of depression are based
on induced inflammation, e.g. the LPS-induced model
[11] and sustained administration of IL-6 by infecting
healthy MRL +/+, C3H.SW and Balb/C mice with ad-
enovirus vector carrying cDNA for murine IL-6 or mice
infected with Ad5mIL6 adenovirus [12,13].
It is known that systemic LPS and administration of
pro-inflammatory cytokines cause chronic central neu-
roinflammation. For example, systemic LPS results in
rapid increases in brain TNFα levels, which may remain
elevated for 10 months [14]. Moreover, brain microglia
are activated to produce chronically elevated pro-in-
flammatory factors in the brain [14]. Central neuroin-
flammation and an increased production of pro-inflam-
matory cytokines, such as TNFα, IL-1β and IL-6, may
induce a behaviour complex, i.e. sickness behaviour,
characterized by the appearance of symptoms such as
anorexia, psychomotor retardation, malaise, loss of in-
terest, etc. [14]. This symptom complex is quite similar
to the symptoms of MDD [15].
External (psychological) as well as internal (organic
stressors) stressors, which induce IRS activation, are re-
lated to the appearance of depressive episodes. Exam-
ples are severe negative life events, some inflammato-
ry and autoimmune illnesses, the postnatal period, etc.
Maes et al. [16–18] were the first to show that – in hu-
mans – psychological stress induces IRS activation with
increased production of pro-inflammatory cytokines,
such as IFNγ and TNFα. In experimental animals in-
creased IL-1β and IL-6 levels have been detected in the
blood and various brain regions [19]. The IRS activa-
tion theory of MDD is fueled by the high comorbidity
of MDD with inflammatory disorders such as multiple
sclerosis (MS), coronary-heart disorder, HIV-infection,
inflammatory bowel disease and rheumatoid arthritis
[1,20]. For example, in MS, the depressive episodes are
preceded by increased IFNγ production [21], suggesting
that IRS activation may underpin depression in MS.
The aim of the present study is to examine anoth-
er immune pathway which may underpin MDD, i.e. an
immune response mounted against an increased trans-
location of LPS from gram-negative enterobacteria. Re-
cently, we published that the median values for serum
IgA and IgM against LPS of enterobacteria are signifi-
cantly greater in patients with chronic fatigue syndrome
(CFS) than in normal volunteers, suggesting that an in-
creased translocation of LPS from enterobacteria is a
new pathway underpinning CFS [22,23]. This condi-
tion can also be described as increased gut permeability
or leaky gut and indicates intestinal mucosal dysfunc-
tion (IMD). Since there is a strong degree of comorbid-
ity between MDD and CFS, and since fatigue is one of
the key symptoms of major depression and gastrointes-
tinal symptoms frequently occur in MDD [24,25], we
hypothesized that also MDD might be accompanied by
an increased translocation of LPS.
The specific aims of the present study are to examine
whether MDD is accompanied by increased serum lev-
els of IgM and IgA against the LPS of 6 enterobacteria,
i.e. Hafnei Alvei, Pseudomanes Aeruginosa, Morganel-
la Morganii, Pseudomanus Putida, Citrobacter Koseri,
and Klebsiella Pneumoniae indicating an immune re-
sponse directed to endotoxins secreted by gram-nega-
tive enterobacteria and which cannot be detected when
the gut-intestinal lining is intact.
SUBJECTS AND METHODS SUBJECTS
Fifty-one subjects participated in the present study, 23
controls (staff or their family members), and 28 MDD
patients admitted to the M-Care4U Outpatient Clinics,
Belgium. The patients were classified according to the
Diagnostic and Statistical Manual of Mental Disorders,
4th edition [26]. We have excluded:
a) subjects with life-time diagnosis of other psychiat-
ric DSM-IV-TR disorders, such as anxiety disor-
ders, schizophrenia, substance use disorders and
organic mental disorders;
b) subjects with CFS as diagnosed by the CDC criteria
[27]; c) subjects with other medical illness, such as
other inflammatory or autoimmune disorders;
d) subjects who ever had been treated with anti-psy-
chotic drugs or anticonvulsants and subjects who
had been taking psychotropic drugs during the last
year prior to the studies; e) subjects with abnor-
mal values for routine blood tests, such as alanine
119
Neuroendocrinology Letters Vol. 29 No. 1 2008 • Article available online: http://node.nel.edu
The gut-brain barrier in major depression: Intestinal mucosal dysfunction ...
aminotransferase (ALT), alkaline phosphatase
(ALP), blood urea nitrogen (BUN), calcium, cre-
atinine, electrolytes, thyroid stimulating hormone
(TSH), total protein, and iron or transferrin satu-
ration; and
f) subjects with acute inflammatory and allergic reac-
tions for at least 1 month prior to the study.
The total sum of the FibroFatigue (FF) scale, i.e. the
Fibromyalgia and Chronic Fatigue Syndrome Rating
Scale [28,29], was used in the MDD patients to com-
pute the severity of specific symptoms, some of which
have been shown to be correlated to the IgM and IgA
levels against LPS in CFS [22]. This scale measures 12
items reminiscent for CFS and fibromyalgia: pain, mus-
cular tension, fatigue, concentration difficulties, failing
memory, irritability, sadness, sleep disturbances, auto-
nomic disturbances, irritable bowel, headache, and sub-
jective experience of infection. Patients and controls
gave written informed consent after the study protocol
was fully explained. The study has been approved by
the local ethical committee.
METHODS
Blood was taken during the morning hours for the de-
termination of IgM and IgA against the LPS of 6 differ-
ent enterobacteria, i.e. Hafnei Alvei, Pseudomanes Aer-
uginosa, Morganella Morganii, Pseud omanus Putida,
Citrobacter Koseri and Klebsiella Pneumoniae. Serum
IgM and IgA levels were measured by means of an in-
direct ELISA method according to the methods out-
lined by the manufacturer (Gemacbio, The Ultimate
Biopharmaceuticals, France). Each serum sample was
measured in duplicate and tested simultaneously with
three standard solutions. The optical densities (OD)
of the three standards are expressed as Z values and
from this the reference linear curve is calculated as Z
= f(OD) with Z = a OD + b. Thus, the Z value of the
lowest standard can be negative. This curve allows to
deduce the mean values of the duplicate measurements
of the OD values. The biological interassay CV values
were < 10%.
Statistics
Group mean differences were assessed by means of
analysis of variance (ANOVA) or analysis of covariance
(ANCOVA). The diagnostic performance was checked
by means of ROC (receiver operating characteristics)
analysis with computation of the area under the ROC
curve, sensitivity, specificity and predictive value of a
positive test result (PV+) and with kappa statistics. Data
reduction was obtained by means of principal compo-
nent (PC) analyses. In order to assess the “total LPS
translocation load” we have employed three different
indices: a) the first PC of the 12 Ig (IgM and IgA) val-
ues; b) the total sum of the 12 Ig (IgM and IgA) levels;
and c) the peak Ig (IgM or IgA) levels, i.e. the highest of
the 12 Ig values. The same indices were also employed
to assess the IgM- versus the IgA-related translocation
loads. Toward this end, we computed the first PCs of
the 6 IgM and 6 IgA data; the peak IgM and peak IgA
data; and the total sum of the 6 IgM and 6 IgA data. We
employed Pearson’s product moment correlation coef-
ficients, Spearman’s rank order correlations, and multi-
ple regression analyses in order to examine the relation-
ships between variables. The relationships between the
IgM / IgA values and the symptoms of the FF scale were
assessed by means of canonical correlation analyses and
by means of stepwise (F-to-enter p=0.05) linear discri-
minant analysis (LDA) with the 12 FF scale items as the
discriminatory variables and the dichotomized peak Ig
(IgM and IgA) data as grouping variable. The signifi-
cance was set at α=0.05 (two tailed).
RESULTS
There were no significant differences (F=1.6, df=4/49,
p=0.2) in age between normal controls (mean ±SD =
40.0 ± 12.9 years) and MDD patients (44.3 ±11.2 years).
There were no significant differences (χ2 Yates = 0.1,
df=1, p=0.7) in the female / male ratio between normal
controls (16 / 7) and MDD patients (17 / 11). There
were no significant correlations between age and any
of the serum IgM or IgA levels directed against the LPS
of the different enterobacteria and no significant differ-
ences in the serum IgM or IgA values between males
and females. The FF score was not significantly differ-
ent between men and women (F=0.06, df=1/26, p=0.8)
and there was no significant correlation between age
and the FF score (r=–0.33, p=0.08).
Table 1 shows the serum IgM values in the MDD
patients and controls. ANOVAs showed that the IgM
levels directed against the LPS from Pseudomonas Aer-
uginosa and Putida were significantly greater in MDD
that in the normal controls. Table 1 shows that there
were significant differences between the study groups
in the 6 IgA levels against LPS except against Pseu-
domonas Aeruginosa. Covarying for age and sex did
not change any of the above results. By means of Fish-
er’s exact probability test we found a significantly (ψ =
0.50, p=0.0002) greater number of MDD patients (12 /
28) with abnormally increased IgM levels (i.e. anyone
of the 6 IgM values > 2 Z values) than in controls (0 /
23). Also, the prevalences of MDD patients (11 / 28)
with abnormally increased IgA levels (i.e. anyone of the
6 IgA > 2 Z values) were significantly (ψ = 0.35, p=0.01)
higher than in normal controls (2 / 23).
Table 2 gives the peak serum Ig values (IgM or IgA,
alone and together), the total sum of the Ig data (IgM or
IgA, alone and together) and the first PC of the IgM and
the IgA data. However, we were unable to find that one
PC could reflect the 12 Ig (IgM and IgA) data: indeed,
the first PC of the 12 Ig data explained only 38.2% of
the variance, while the 6 IgM variables loaded highly on
the first PC and the 6 IgA data on the second PC even
120
Copyright © 2008 Neuroendocrinology Letters ISSN 0172–780X • w ww.nel.edu
Michael Maes, Marta Kubera and Jean-Claude Leunis
without performing a rotation. Thus, there is no com-
mon PC which can reflect the 12 Ig variables. The peak
IgM and IgA values and the total sum of all 6 IgM and
IgA values as well as the first PC subtracted from the 6
IgM and IgA values were significantly higher in MDD
patients than in normal controls. The peak Ig (either
IgM or IgA) values and the sum of the 12 Ig values were
both significantly higher in MDD patients than in nor-
mal controls. Covarying for age and sex in ANCOVAs
did not change any of these results. ROC analysis per-
formed on the peak IgM or IgA values showed that the
area under the ROC curve (AUC) was highly significant
(AUC=90.8%). Accordingly, the diagnostic perform-
ance computed was highly significant; at a cut off value
for the peak IgM or IgA values > 2.2 Z, the diagnostic
performance was: sensitivity=67.9%, specificity=95.6%,
and PV+=95.0% (κ=0.62, t=5.66, p=0.00002).
By means of Spearman’s rank order correlation coef-
ficients we found a significant positive correlation be-
tween the peak IgM-IgA data and the total sum on the
FF scale (r=0.49, p=0.007). Two items of the FF scale
were significantly related to the peak IgM-IgA data, i.e
fatigue (r=0.54, p=0.003) and irritable bowel (r=0.56,
p=0.002). Table 3 shows the results of a canonical cor-
relation analysis with peak IgM and peak IgA values as
explanatory variables and the 12 items of the FF scale as
dependent variables. We found that the symptom pro-
files of increased IgM and IgA were fatigue, autonomic
disturbances, irritable bowel and a subjective feeling of
infection. In order to further examine the relationships
between serum IgM and IgA levels and the severity of
the FF scale, we carried out stepwise LDA with the 12
FF scale symptoms as the explanatory variables and the
dichotomized peak IgM (higher versus lower than 2 Z
values) and peak IgA (higher versus lower 2 Z values)
as groups. We found that sadness (loading=0.71), irri-
table bowel (loading=0.58) and muscular tension (load-
ing=0.40) were significant discriminatory variables for
the dichotomized peak IgM values (F=18.5, df=1/26,
p=0.0004). The dichotomized peak IgA groups were
best discriminated (F=5.7, df=1/26, p=0.02) using fa-
tigue (loading = 0.86) and irritable bowel (loading =
0.77) as discriminatory variables. Finally, the dichot-
omized peak IgM or IgA values were best predicted by
fatigue (loading = 0.87), irritable bowel (loading = 0.75)
and a subjective feeling of infection (loading = 0.65)
(F=18.2, df=1/26, p=0.0004).
DISCUSSION
The findings of the present study show that MDD is ac-
companied by increased serum levels of IgM and IgA
directed against LPS of gram-negative enterobacteria
and that the IgM-IgA values are related to symptoms
reminiscent of MDD and CFS, e.g. fatigue, autonomic
and gastro-intestinal symptoms, and a subjective feel-
ing of infection.
The results of our study show that there is an IgM
and IgA-related immune response raised to the LPS
of enterobacteria in MDD. Increments in serum IgM
levels can be seen in immune activation and mucosal
immunity. B1 lymphocytes are a significant source of
natural serum IgM and they constitute a first line of
defence against systemic viral and bacterial infections
[30]. Moreover, B1 cells migrate to the intestinal lamina
propria to differentiate into IgA-producing serum cells,
which in turn play a role in mucosal immunity [30].
The increased serum IgM and IgA levels against LPS
in MDD indicate that MDD is accompanied by an in-
creased gut permeability and that there is an immune
response directed against LPS of enterobacteria. Indeed,
the intestinal epithelial barrier has critical functions,
such as a) the formation of a barrier, which separates
the luminal contents from the interstitium, and which
protects against micro-organisms including gram neg-
ative bacteria, larger toxic and antigenic molecules; b)
the transportation of fluids, electrolytes and nutrients
across the intestinal wall; and c) the secretion of IgA
to bind to bacteria thus preventing their attachment to
epithelial cells.
The function of the intestinal barrier may be com-
promised by IRS activation. The latter may cause a
loss of the protective barrier function [31–33], which
in turn causes enlarged spaces between the cells of the
gut wall [31–33]. In this respect, the important inflam-
matory mediators which induce “leaky gut”, IFNγ and
IL-6, are both significantly increased in MDD [34,35].
These disruptions of the intestinal epithelium allow
normally poorly invasive enterobacteria to exploit lipid
raft-mediated transcytotic pathways or the enlarged
spaces to cross the gut wall [31–33]. Thus, IRS activa-
tion – through an increased production of IFNγ and
IL-6 – is an essential factor in the loss of the epithelial
barrier function [31,33]. The former may induce an in-
creased translocation of LPS and thus cause increased
serum concentrations of LPS which, in turn, may trig-
ger an IgM or IgA-mediated immune response to LPS
[22,23].
Systemic increases in LPS not only cause a system-
ic inflammation, but also a central neuroinflammation;
increased brain tumor necrosis factor-α (TNFα) activi-
ties, which may remain elevated for 10 months; and ac-
tivation of brain microglia with a chronically elevated
production of pro-inflammatory mediators [14]. It is
well-known that an increased production of pro-in-
flammatory cytokines, such as IL-1, IL-6 and TNFα, ei-
ther peripheral or central, and brain neuroinflammation
may induce the sickness behaviour complex [14]. Also
systemic LPS may provoke sickness behaviour [36,37].
As explained previously, symptoms of sickness behav-
iour, such as anorexia, soporific effects, disturbances of
locomotor activity and exploration, and anhedonia bear
a strong similarity with those of MDD [15].
Increased LPS translocation may play a role in the
O&NS, which occurs in MDD. Indeed, MDD is ac-
121
Neuroendocrinology Letters Vol. 29 No. 1 2008 • Article available online: http://node.nel.edu
The gut-brain barrier in major depression: Intestinal mucosal dysfunction ...
Table 1. Measurements of serum IgM levels against the LPS of Hafnia Alvei, Pseudomonas Aeruginosa, Morganella
Morganii, Pseudomonas Putida, Citrobacter Koseri and Klebsiella Pneumoniae in normal controls and MDD patients.
Variables normal controls MDD p
Hafnei Alvei IgM
IgA
–0.41 (0.98)
–0.85 (0.71)
0.33 (2.03)
0.32 (2.13)
0.1
0.01
Pseudomonas Aeruginosa IgM
IgA
– 0.28 (1.00)
–0.40 (1.14)
0.68 (1.83)
0.27 (1.83)
0.02
0.13
Morganella Morganii IgM
IgA
–0.10 (0.91)
–0.88 (0.58)
0.98 (2.51)
1.00 (2.73)
0.052
0.002
Pseudomonas Putida IgM
IgA
–0.15 (0.85)
–0.56 (0.94)
1.19 (2.40)
0.26 (1.44)
0.01
0.02
Citrobacter Koseri IgM
IgA
–0.18 (1.05)
–0.60 (0.62)
0.61 (1.81)
1.09 (3.16)
0.07
0.01
Klebsiella Pneumoniae IgM
IgA
–0.46 (0.76)
–0.93 (1.39)
0.07 (1.67)
0.99 (1.90)
0.17
0.0004
All results are shown as mean (±SD). All results of ANOVAs (df=1/49)
Table 2. Measurements of peak IgM or IgA values, total sum of the 6 IgM or 6 IgA data, the first principal component
(PC) of the 6 IgM and 6 IgA data, as well as the peak, total sum and the first PC of the 12 Ig (IgM and IgA) data.
Variables normal controls MDD F value p
Peak IgM
Sum of the 6 IgM
First PC of the 6 IgM data
–0.48 (0.83)
–1.57 (4.48)
2.39 (1.09)
2.09 (2.67)
3.87 (10.63)
3.68 (2.56)
7.7
5.2
5.1
0.007
0.02
0.03
Peak IgA
Sum of the 6 IgA
First PC of the 6 IgA data
0.13 (1.12)
–4.22 (3.81)
–0.95 (085)
2.59 (3.09)
3.90 (10.79)
0.78 (2.32)
13.2
11.8
11.5
0.0009
0.002
0.002
Peak Ig (IgM and IgA) data
Sum of the 12 Ig (IgM and IgA) data
0.13 (1.12)
–5.79 (6.41)
2.59 (3.09)
7.77 (13.89)
27.5
18.6
0.00003
0.0002
All results are shown as mean (±SD). All results of ANOVAs (df=1/49)
Table 3. Results of canonical correlation analysis with the regression of the first principal
component (PC) of the IgM and the first PC of the IgA data directed against the LPS of Hafnia Alvei,
Pseudomonas Aeruginosa, Morganella Morganii, Pseudomonas Putida, Citrobacter Koseri, and
Klebsiella Pneumoniae on the one hand and the symptoms of the FibroFatigue scale, on the other.
IgM 0.78
IgA 0.54
Aches & pain 0.24
Muscular tension 0.10
Fatigue 0.36
Concentration difficulties 0.21
Failing memory 0.28
Irritability 0.28
Sadness 0.24
Sleep disturbances –0.13
Autonomic disturbances 0.59
Irritable bowel 0.86
Headache 0.18
Subjective experience of infection 0.47
Canonical correlation coefficient r=0.67
Shown are the significant loadings (> 0.35) of the canonical regressions of the
IgM / IgA data on the different symptoms of the FibroFatigue scale.
122
Copyright © 2008 Neuroendocrinology Letters ISSN 0172–780X • w ww.nel.edu
Michael Maes, Marta Kubera and Jean-Claude Leunis
companied by O&NS as indicated by increased levels of
malondialdehyde (MDA), a byproduct of polyunsatu-
rated fatty acid peroxidation and arachidonic acid, 8-
hydroxy-2-deoxyguanosine, indicating damage to DNA
by oxygen radicals [38–40]; increased catalase and
MDA levels [41,42], increased peroxidase and catalase
activities in blood and saliva [43]; and increased IgM
levels directed against nitro-BSA (bovine serum albu-
min), indicating increased nitrosative stress [44]. It is
well known that LPS causes induction of nuclear factor
kappa beta (NFκβ), the major upstream, intracellular
mechanism which regulates inflammatory and O&NS
mediators, such as cyclo-oxygenase (COX-2) and in-
ducible NO synthase (iNOS) [45,46].
Based on the above, we may conclude that an in-
creased translocation of LPS is another pathway which
may explain the inflammatory pathophysiology of
MDD and which underpins the link between IMD and
MDD. For example, external stressors (psychological
stress) may compromise the intestinal barrier [47,48]
as well as inducing the cytokine network causing in-
creased IFNγ and IL-6 production [16], which in turn
may further endanger the mucosal gut barrier.
Increased LPS translocation offers also an explana-
tion for the increased incidence of MDD during alco-
hol dependence. Thus, alcoholism is known to induce
an increased translocation of LPS [49] and is accom-
panied by activation of the IRS with an increased ca-
tabolism of tryptophan [50,51]. Internal (organic) stres-
sors, such as infections (AIDS/HIV) [52], autoimmune
disorders [53] and inflammatory bowel disease [54]
– which all show some degree of comorbidity with de-
pression – may endanger gut mucosal function and may
increase gut barrier permeability. Thus, the LPS path-
way may either primarily (an increased translocation of
LPS inducing inflammation) or secondarily (a primary
inflammation may induce LPS translocation through
increased IL-6 and IFNγ production) be involved in
the inflammatory pathophysiology of MDD. Conse-
quently, the intertwined systemic inflammation and the
IgM and IgA-mediated immune response following LPS
translocation may further aggravate the depressogenic
effects of IRS activation.
The symptom profiles of increased IgM and IgA lev-
els to LPS of enterobacteria in MDD are fatigue, sad-
ness, autonomic and gastro-intestinal symptoms, mus-
cular tension, and a subjective feeling of infection.
These findings corroborate our previous report that, in
CFS, there are significant positive correlations between
the IgA responses to LPS of enterobacteria and symp-
toms of the FibroFatigue scale, such as muscular ten-
sion, fatigue, concentration difficulties, failing memory,
autonomic disturbances, irritable bowel and the subjec-
tive experience of infection [22]. These correlations
probably reflect causal relationships between IRS acti-
vation and O&NS inducing the abovementioned symp-
toms occurring in CFS and MDD. As discussed above,
LPS and pro-inflammatory cytokines are depressogen-
ic and induce sickness behaviour, e.g. fatigue, anorex-
ia, weight loss, sleep disorders, psychomoror retarda-
tion, etc. Also, O&NS is implicated in the production
of musceloskeletal pain, muscle fatigue resistance, re-
duced responses to aerobic exercise and reduced maxi-
mal exercise time [55–59]. The significant correlation
between gastro-intestinal symptoms and the Ig-me-
diated immune response against LPS in MDD and in
CFS indicates that the gastro-intestinal symptoms in
some patients with MDD (and in CFS) reflect – in part
– disorders in gut-intestinal permeability and cannot be
considered as a symptom of mental stress as most psy-
chiatrists tend to confirm. The significant positive cor-
relation between the IgM and IgA-mediated immune
response against LPS and the subjective experience of
infection in MDD indicates that the latter is an index of
the inflammation from which the patients are suffering.
The canonical correlation analysis which was carried
out showed also a significant correlation between the
Ig-mediated immune response against LPS and auto-
nomic disturbances, such as gastroparesis, neuropathy,
vascular neuropathy, and dysautonomia. This may be
explained since LPS as well as IRS activation and O&NS
have multivarious and profound effects on the autono-
mous nervous system [60–66].
Increased gut permeability may be another factor
explaining the occurrence of autoimmunity in MDD.
There are many reports that MDD is accompanied
by autoimmune responses, such as against phopholi-
pids [67]. Indeed, enterobacteria may act as superan-
tigens for T lymphocytes or may induce autoimmu-
nity through a mechanism called molecular mimicry
[68,69]. This may be explained since enterobacteria
have antigenic sites which are very similar to those of
the lipid structures of neuronal tissue. These antigens
will go into various tissues and trigger inflammation
and once autoantibodies are formed the inflammation
may become more chronic. Thus, systemic LPS caused
by an increased translocation not only induces periph-
eral inflammation and O&NS, but also induces a cen-
tral neuroinflammation and eventually an autoimmune
responses directed against neuronal tissues.
The results of the present study show that patients
with MDD should be checked for the presence of leaky
gut by the measurements of IgM and IgA against the
LPS of gram-negative bacteria. The results of the present
study suggest that MDD patients who have a leaky gut
should be treated with specific antioxidants with an ef-
ficacity for leaky-gut and a leaky gut diet [23].
Acknowledgments
The research reported was supported by a NARSAD
Distinguished researcher award to M.Maes and by M-
CARE4U and CRC-MH, Antwerp, Belgium. The secre-
tarial assistance of Indra Corten is greatly appreciated.
123
Neuroendocrinology Letters Vol. 29 No. 1 2008 • Article available online: http://node.nel.edu
The gut-brain barrier in major depression: Intestinal mucosal dysfunction ...
REFERENCES
Maes M: A review on the acute phase response in major depres-
sion. Rev Neurosci 1993; 4(4): 407–416. Review.
Maes M: Evidence for an immune response in major depression:
a review and hypothesis. Prog Neuropsychopharmacol Biol Psy-
chiatry 1995; 19(1): 11–38. Review.
Maes M: Major depression and activation of the inflammatory
response system. Adv Exp Med Biol 1999; 461: 25–46. Review.
Schiepers OJ, Wichers MC, Maes M: Cytokines and major depression.
Prog Neuropsychopharmacol Biol Psychiatry 2005; 29(2): 201–
217. Review. Erratum in: Prog Neuropsychopharmacol Biol Psy-
chiatry 2005; 29(4): 637–638.
Maes M, Brene S, Hibbeln J, etc.: An Experimental Medicine
Approach to examine the Inflammatory & Neurodegenerative
hypothesis in Depression: leads for future research and drug
development in depression. Submitted.
Maes M, Mihaylova I, Ruyter MD, Kubera M, Bosmans E: The
immune effects of TRYCATs (tryptophan catabolites along the
ido pathway): Relevance for depression – and other conditions
characterized by tryptophan depletion induced by inflamma-
tion. Neuro Endocrinol Lett 2007; 28(6): 826–831.
Kubera M, Basta-Kaim A, Holan V, Simbirtsev A, Roman A, Piga-
reva N, Prokopieva E, Sham J: Effect of mild chronic stress, as a
model of depression, on the immunoreactivity of C57BL/6 mice.
Int J Immunopharmacol 1998; 20(12): 781–789.
Kubera M, Symbirtsev A, Basta-Kaim A, Borycz J, Roman A, Papp
M, Claesson M: Effect of chronic treatment with imipramine
on interleukin 1 and interleukin 2 production by splenocytes
obtained from rats subjected to a chronic mild stress model of
depression. Pol J Pharmacol 1996; 48(5): 503–506.
Goshen I, Kreisel T, Ben-Menachem-Zidon O, Licht T, Weiden-
feld J, Ben-Hur T, Yirmiya R: Brain interleukin-1 mediates chronic
stress-induced depression in mice via adrenocortical activation
and hippocampal neurogenesis suppression. Mol Psychiatry
2007 [Epub ahead of print]
Song C, Leonard BE: Interleukin-2-induced changes in behav-
ioural, neurotransmitter, and immunological parameters in the
olfactory bulbectomized rat. Neuroimmunomodulation 1995;
2(5): 263–273.
Yirmiya R: Endotoxin produces a depressive-like episode in rats.
Brain Res 1996; 711(1–2): 163–174.
Sakiƒc B, Szechtman H, Braciak T, Richards C, Gauldie J, Denburg
JA: Reduced preference for sucrose in autoimmune mice: a pos-
sible role of interleukin-6. Brain Res Bull 1997; 44(2): 155–165.
Sakiƒc B, Gauldie J, Denburg JA, Szechtman H: Behavioral effects
of infection with IL-6 adenovector. Brain Behav Immun 2001;
15(1): 25–42.
Qin L, Wu X, Block ML, Liu Y, Breese GR, Hong JS, Knapp DJ, Crews
FT: Systemic LPS causes chronic neuroinflammation and pro-
gressive neurodegeneration. Glia 2007; 55(5): 453–462.
Maes M, Meltzer HY, Scharpé S, Cooreman W, Uyttenbroeck W, Suy
E, Vandervorst C, Calabrese J, Raus J, Cosyns P: Psychomotor retar-
dation, anorexia, weight loss, sleep disturbances, and loss of
energy: psychopathological correlates of hyperhaptoglobinemia
during major depression. Psychiatry Res 1993; 47(3): 229–241.
Maes M, Song C, Lin A, De Jongh R, Van Gastel A, Kenis G, Bos-
mans E, De Meester I, Benoy I, Neels H, Demedts P, Janca A, Schar-
pé S, Smith RS: The effects of psychological stress on humans:
increased production of pro-inflammatory cytokines and a Th1-
like response in stress-induced anxiety. Cytokine 1998; 10(4):
313–318.
Maes M, Lin AH, Delmeire L, Van Gastel A, Kenis G, De Jongh R,
Bosmans E: Elevated serum interleukin-6 (IL-6) and IL-6 receptor
concentrations in posttraumatic stress disorder following acci-
dental man-made traumatic events. Biol Psychiatry 1999; 45(7):
833–839.
Song C, Kenis G, van Gastel A, Bosmans E, Lin A, de Jong R, Neels
H, Scharpé S, Janca A, Yasukawa K, Maes M: Influence of psycholog-
ical stress on immune-inflammatory variables in normal humans.
Part II. Altered serum concentrations of natural anti-inflamma-
tory agents and soluble membrane antigens of monocytes and T
lymphocytes. Psychiatry Res 1999; 85(3): 293–303.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Ishikawa I, Kitamura H, Kimura K, Saito M: Brain interleukin-1
is involved in blood interleukin-6 response to immobilization
stress in rats. Jpn J Vet Res 2001; 49(1): 19–25.
Yirmiya R, Weidenfeld J, Pollak Y, Morag M, Morag A, Avitsur R,
Barak O, Reichenberg A, Cohen E, Shavit Y, Ovadia H: Cytokines,
“depression due to a general medical condition,” and antidepres-
sant drugs. Adv Exp Med Biol 1999; 461: 283–316. Review.
Mohr DC, Goodkin DE, Islar J, Hauser SL, Genain CP: Treatment
of depression is associated with suppression of nonspecific
and antigen-specific T(H)1 responses in multiple sclerosis. Arch
Neurol 2001; 58(7): 1081–1086.
Maes M, Mihaylova I, Leunis JC: Increased serum IgA and IgM
against LPS of enterobacteria in chronic fatigue syndrome (CFS):
indication for the involvement of gram-negative enterobacteria
in the etiology of CFS and for the presence of an increased gut-
intestinal permeability. J Affect Disord. 2007; 99(1–3): 237–240.
Maes M, Coucke F, Leunis JC: Normalization of the increased
translocation of endotoxin from gram negative enterobacteria
(leaky gut) is accompanied by a remission of chronic fatigue syn-
drome. Neuro Endocrinol Lett 2007; 28(6): 739–744.
Maes M: From Freud to Omega-3. 2005, Brussels, Standaard Uit-
geverij
Maes M, Schotte C, Maes L, Cosyns P: Clinical subtypes of uni-
polar depression: Part II. Quantitative and qualitative clinical
differences between the vital and nonvital depression groups.
Psychiatr Res 1990; 34(1): 43–57.
American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, Fourth Edition. Text Revision (DSM-
IV-TR), 2000; Washington DC.
Fukuda K, Straus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff
A: The chronic fatigue syndrome: a comprehensive approach to
its definition and study. International Chronic Fatigue Syndrome
Study Group. Ann Int Med 1994; 121(12): 953–959.
Zachrisson O, Regland B, Jahreskog M, Kron M, Gottfries CG: A rating
scale for fibromyalgia and chronic fatigue syndrome (the FibroFa-
tigue scale). J Psychosom Res 2002; 52(6): 501–509.
Shapiro CM, Moller HJ: Chronic fatigue: listen and measure. J
Psychosom Res 2002; 52(6): 427–436.
Thurnheer MC, Zuercher AW, Cebra JJ, Bos NA: B1 cells contrib-
ute to serum IgM, but not to intestinal IgA, production in gno-
tobiotic Ig allotype chimeric mice. J Immunol 2003; 170(9):
4564–4571.
Clark E, Hoare C, Tanianis-Hughes J, Carlson GL, Warhurst G:
Interferon gamma induces translocation of commensal Esche-
richia coli across gut epithelial cells via a lipid raft-mediated pro-
cess. Gastroenterology 2005; 128(5): 1258–1267.
Chavez AM, Menconi MJ, Hodin RA, Fink MP: Cytokine-induced
intestinal epithelial hyperpermeability: role of nitric oxide. Crit
Care Med 1999; 27(10): 2246–2251.
Yang R, Han X, Uchiyama T, Watkins SK, Yaguchi A, Delude RL,
Fink MP: IL-6 is essential for development of gut barrier dysfunc-
tion after hemorrhagic shock and resuscitation in mice. Am J
Physiol Gastrointest Liver Physiol 2003; 285(3): G621–629.
Maes M, Scharpé S, Meltzer HY, Okayli G, Bosmans E, D’Hondt
P, Vanden Bossche BV, Cosyns P: Increased neopterin and inter-
feron-gamma secretion and lower availability of L-tryptophan
in major depression: further evidence for an immune response.
Psychiatry Res 1994; 54(2): 143–160.
Maes M, Scharpé S, Meltzer HY, Bosmans E, Suy E, Calabrese J,
Cosyns P: Relationships between interleukin-6 activity, acute
phase proteins, and function of the hypothalamic-pituitary-adre-
nal axis in severe depression. Psychiatry Res 1993; 49(1): 11–27.
Borowski T, Kokkinidis L, Merali Z, Anisman H: Lipopolysaccha-
ride, central in vivo biogenic amine variations, and anhedonia.
Neuroreport 1998; 9(17): 3797–3802.
Lacosta S, Merali Z, Anisman H: Behavioral and neurochemical
consequences of lipopolysaccharide in mice: anxiogenic–like
effects. Brain Res 1999; 818(2): 291–303.
Sarandol A, Sarandol E, Eker SS, Erdinc S, Vatansever E, Kirli S:
Major depressive disorder is accompanied with oxidative stress:
short-term antidepressant treatment does not alter oxidative-
antioxidative systems. Hum Psychopharmacol 2007; 22(2): 67–
73.
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
124
Copyright © 2008 Neuroendocrinology Letters ISSN 0172–780X • w ww.nel.edu
Michael Maes, Marta Kubera and Jean-Claude Leunis
Forlenza MJ, Miller GE: Increased serum levels of 8-hydroxy-2’-
deoxyguanosine in clinical depression. Psychosom Med 2006;
68(1): 1–7.
Irie M, Miyata M, Kasai H: Depression and possible cancer risk due
to oxidative DNA damage. J Psychiatr Res 2005; 39(6): 553–560.
Khanzode SD, Dakhale GN, Khanzode SS, Saoji A, Palasodkar R:
Oxidative damage and major depression: the potential antioxi-
dant action of selective serotonin re-uptake inhibitors. Redox
Report 2003; 8(6): 365–370.
Ozcan ME, Gulec M, Ozerol E, Polat R, Akyol O: Antioxidant
enzyme activities and oxidative stress in affective disorders. Int
Clin Psychopharmacol 2004; 19(2): 89–95.
Lukash AI, Zaika VG, Kucherenko AO, Miliutina NP: Free radical
processes and antioxidant system in depression and treatment
efficiency. Zh Nevrol Psikhiatr Im S S Korsakova 2002; 102(9): 41–
44.
Maes M, Mihaylova I, Kubera M, Leunis J-C: An IgM-mediated
immune response directed against nitro-bovine serum albumin
(nitro-BSA) in chronic fatigue syndrome (CFS) and major depres-
sion (MDD): evidence that nitrosative stress is another factor
underpinning the comorbidity between CFS and MDD. Neuro
Endocrinol Lett 2008 (in press).
Maes M, Mihaylova I, Bosmans E: Not in the mind of neurasthenic
lazybones but in the cell nucleus: patients with chronic fatigue
syndrome have increased production of nuclear factor kappa
beta. Neuro Endocrinol Lett 2007; 28(4): 456–462.
Maes M, Mihaylova I, Kubera M, Bosmans E: Not in the mind but in
the cell: increased production of cyclo-oxygenase-2 and inducible
NO synthase in chronic fatigue syndrome. Neuro Endocrinol Lett
2007; 28(4): 463–469.
Meddings JB, Swain MG: Environmental stress-induced gastroin-
testinal permeability is mediated by endogenous glucocorticoids
in the rat. Gastroenterology 2000; 119(4): 1019–1028.
Cameron HL, Perdue MH: Stress impairs murine intestinal barrier
function: improvement by glucagon-like peptide-2. J Pharmacol
Exp Ther 2005; 314(1): 214–220.
Napolitano LM, Koruda MJ, Zimmerman K, McCowan K, Chang
J, Meyer AA: Chronic ethanol intake and burn injury: evidence
for synergistic alteration in gut and immune integrity. J Trauma
1995; 38(2): 198–207.
Maes M, Lin A, Bosmans E, Vandoolaeghe E, Bonaccorso S, Kenis
G, De Jongh R, Verkerk R, Song C, Scharpé S, Neels H: Serotonin-
immune interactions in detoxified chronic alcoholic patients
without apparent liver disease: activation of the inflammatory
response system and lower plasma total tryptophan. Psychiatry
Res 1998; 78(3): 151–161.
Song C, Lin A, De Jong R, Vandoolaeghe E, Kenis G, Bosmans E,
Whelan A, Scharpe S, Maes M: Cytokines in detoxified patients with
chronic alcoholism without liver disease: increased monocytic
cytokine production. Biol Psychiatry 1999; 45(9): 1212–1216.
Lim SG, Menzies IS, Lee CA, et al: Intestinal permeability and
function in patients infected with human immunodeficiency
virus. A comparison with coeliac disease. Scand J Gastroenterol
1993; 28: 573–580.
Sundqvist T, Lindstrom F, Magnusson KE, et al: Influence of fasting on
intestinal permeability and disease activity in patients with rheu-
matoid arthritis. Scand J Rheumatol 1982; 11: 33–38.
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Shanahan F: Current concepts of the pathogenesis of inflamma-
tory bowel disease. Ir J Med Sci 1994; 163: 544–549.
Vecchiet J, Cipollone F, Falasca K, Mezzetti A, Pizzigallo E, Buc-
ciarelli T, De Laurentis S, Affaitati G, De Cesare D, Giamberardino
MA: Relationship between musculoskeletal symptoms and blood
markers of oxidative stress in patients with chronic fatigue syn-
drome. Neurosci Lett 2003; 335(3): 151–154.
Matuszczak Y, Farid M, Jones J, Lansdowne S, Smith MA, Taylor
AA, Reid MB: Effects of N-acetylcysteine on glutathione oxida-
tion and fatigue during handgrip exercise. Muscle Nerve 2005;
32(5): 633–638.
Luan Y, Xu W: The function of the selective inhibitors of cyclooxy-
genase 2. Mini Rev Med Chem 2006; 6(12): 1375–1381.
Mets T, Bautmans I, Njemini R, Lambert M, Demanet C: The influ-
ence of celecoxib on muscle fatigue resistance and mobility in
elderly patients with inflammation. Am J Geriatr Pharmacother
2004; 2(4): 230–238.
McIver KL, Evans C, Kraus RM, Ispas L, Sciotti VM, Hickner RC: NO-
mediated alterations in skeletal muscle nutritive blood flow and
lactate metabolism in fibromyalgia. Pain 2005; 120(1–2): 161–
169.
Hartrick CT: Increased production of nitric oxide stimulated by
interferon-gamma from peripheral blood monocytes in patients
with complex regional pain syndrome. Neurosci Lett 2002;
323(1): 75–77.
Stoyneva Z, Lyapina M, Tzvetkov D, Vodenicharov E: Current
pathophysiological views on vibration-induced Raynaud’s phe-
nomenon. Cardiovasc Res 2003; 57(3): 615–624.
Alexander GM, Perreault MJ, Reichenberger ER, Schwartzman RJ:
Changes in immune and glial markers in the CSF of patients with
Complex Regional Pain Syndrome. Brain Behav Immun 2007;
21(5): 668–676.
Zhao J, Wecht JM, Zhang Y, Wen X, Zeman R, Bauman WA, Car-
dozo C: iNOS expression in rat aorta is increased after spinal cord
transection: a possible cause of orthostatic hypotension in man.
Neurosci Lett 2007; 415(3): 210–214.
Delrue-Perollet C, Li KS, Vitiello S, Neveu PJ: Peripheral catechola-
mines are involved in the neuroendocrine and immune effects of
LPS. Brain Behav Immun 1995; 9(2): 149–162.
Mailman D: A role for abdominal vagal afferents in lipopolysac-
charide-induced hypotension. Shock 2002; 18(2): 177–181.
Rogausch H, Vo NT, Del Rey A, Besedovsky HO: Increased sensi-
tivity of the baroreceptor reflex after bacterial endotoxin. Ann N
Y Acad Sci 2000; 917: 165–168.
Maes M, Meltzer H, Jacobs J, Suy E, Calabrese J, Minner B, Raus J:
Autoimmunity in depression: increased antiphospholipid auto-
antibodies. Acta Psychiatr Scand 1993; 87(3): 160–166.
Levin MC, Lee SM, Kalume F, Morcos Y, Dohan FC Jr, Hasty KA,
Callaway JC, Zunt J, Desiderio D, Stuart JM: Autoimmunity due to
molecular mimicry as a cause of neurological disease. Nat Med
2002; 8(5):509–513.
Wucherpfennig KW: Structural basis of molecular mimicry. J
Autoimmun 2001;16(3): 293–302.
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69