You are on page 1of 9

Basic Sciences

Cholinergic Pathway Suppresses Pulmonary


Innate Immunity Facilitating Pneumonia After Stroke
Odilo Engel, DVM*; Levent Akyüz, MSc*; Andrey C. da Costa Goncalves, PhD;
Katarzyna Winek, MD; Claudia Dames, MSc; Mareike Thielke, MLS; Susanne Herold, MD;
Chotima Böttcher, PhD; Josef Priller, MD; Hans Dieter Volk, MD; Ulrich Dirnagl, MD;
Christian Meisel, MD*; Andreas Meisel, MD*

Background and Purpose—Temporary immunosuppression has been identified as a major risk factor for the development
of pneumonia after acute central nervous system injury. Although overactivation of the sympathetic nervous system
was previously shown to mediate suppression of systemic cellular immune responses after stroke, the role of the
parasympathetic cholinergic anti-inflammatory pathway in the antibacterial defense in lung remains largely elusive.
Methods—The middle cerebral artery occlusion model in mice was used to examine the influence of the parasympathetic
nervous system on poststroke immunosuppression. We used heart rate variability measurement by telemetry, vagotomy,
α7 nicotinic acetylcholine receptor–deficient mice, and parasympathomimetics (nicotine, PNU282987) to measure and
modulate parasympathetic activity.
Results—Here, we demonstrate a rapidly increased parasympathetic activity in mice after experimental stroke. Inhibition
of cholinergic signaling by either vagotomy or by using α7 nicotinic acetylcholine receptor–deficient mice reversed
pulmonary immune hyporesponsiveness and prevented pneumonia after stroke. In vivo and ex vivo studies on the role
of α7 nicotinic acetylcholine receptor on different lung cells using bone marrow chimeric mice and isolated primary
cells indicated that not only macrophages but also alveolar epithelial cells are a major cellular target of cholinergic anti-
inflammatory signaling in the lung.
Conclusions—Thus, cholinergic pathways play a pivotal role in the development of pulmonary infections after acute central
nervous system injury.   (Stroke. 2015;46:3232-3240. DOI: 10.1161/STROKEAHA.115.008989.)
Key Words: bone marrow chimeric mice ◼ cholinergic anti-inflammatory pathway ◼ parasympathetic nervous
system ◼ pneumonia ◼ stroke-induced immunodepression

T he management of medical complications is one of the


most critical factors in the therapy of stroke because 95%
of stroke patients experience at least one relevant medical
central nervous system injuries like stroke, spinal cord injury,
or traumatic brain injury disturb the normally well-balanced
interplay between these 2 supersystems, resulting in a profound
complication in the first 3 months after stroke.1 In both experi- and long-lasting immunodepression.6,7 Overactivation of the
mental and clinical studies, infections are the most common sympathetic nervous system after stroke was shown to mediate
complication after stroke.2 Among these, pneumonia is the suppression of peripheral cellular immune responses and to con-
most relevant poststroke infection,3 accounting for the highest tribute to development of poststroke bacterial lung infections.8–10
attributable mortality after stroke.4 However, changes in the pulmonary immune compartment
To maintain body homeostasis in response to challenges and the underlying mechanisms of impaired antibacterial lung
from the environment, the nervous and immune systems are immune responses after stroke are currently not understood.
closely interconnected in an intense bidirectional communica- Besides resident lung immune cells, such as alveolar
tion.5 Recent experimental and clinical evidence indicates that macrophages (MΦ), alveolar epithelial cells (AEC) express

Received March 11, 2015; final revision received August 19, 2015; accepted August 24, 2015.
From the Department of Experimental Neurology (O.E., K.W., M.T., U.D., A.M.), Department of Neurology (U.D., A.M.), NeuroCure Clinical Research
(U.D., A.M.), Institute for Medical Immunology (L.A., C.D., H.D.V., C.M.), BCRT Berlin Brandenburg Centre for Regenerative Medicine (L.A., H.D.V.),
Department of Neuropsychiatry and Laboratory of Molecular Psychiatry (C.B., J.P.), and Center for Stroke Research Berlin (O.E., K.W., M.T., U.D., A.M.),
Charité University Medicine Berlin, Berlin, Germany; German Center for Neurodegeneration Research (DZNE), partner site Berlin, Germany (J.P., U.D.);
Department of Internal Medicine II, Justus-Liebig-University, Universities Giessen and Marburg Lung Center, Member of the German Center for Lung
Research (DZL) (S.H.); and Max Delbrück Center for Molecular Medicine, Berlin, Germany (A.C.d.C.G.).
Guest Editor for this article was Malcolm R. Macleod, PhD.
*O. Engel, L. Akyüz, Drs C. Meisel, and A. Meisel contributed equally.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.
115.008989/-/DC1.
Correspondence to Andreas Meisel, MD, NeuroCure Clinical Research Center, Charité–Universitätsmedizin Berlin, Charitéplatz 1, 10117 Berlin,
Germany. E-mail andreas.meisel@charite.de
© 2015 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.115.008989

3232
Engel et al   Cholinergic Pathway in Stroke-Associated Pneumonia    3233

a variety of receptors recognizing pathogens or their prod- Drug Preparation and Administration
ucts.11 Rapid activation of MΦ and airway epithelium are of Nicotine (Sigma Aldrich) was given via drinking water in a concen-
crucial importance for effective host defense during bacterial tration of 100 μg/mL. The effects on heart rate (HR) were monitored
lung infections.12 Both MΦ and lung epithelial cells have been by pulse oximetry (MouseOx; STARR life sciences Corp.).
shown to express the α7 nicotinergic acetylcholine receptor
(α7nAChR),13 which has been identified as a crucial down- Creation of Bone Marrow Chimeric Mice
stream element of the so-called cholinergic anti-inflammatory For generation of bone marrow (BM) chimeric mice, BM cells
pathway.14 This neuronal pathway, involving the parasym- were isolated under sterile conditions from the tibias and femurs
of donor mice. 5×106 BM cells were IV injected into 6- to 8-week-
pathetic vagus nerve and the neurotransmitter acetylcholine old lethally irradiated recipient mice 24 h after total body irra-
(ACh), was shown to act as a negative feedback loop to prevent diation (1100 cGy) using a 137Caesium Gammacell 40 Exactor
a potentially harmful overreaction of the immune system dur- (Theratronics).
ing inflammatory conditions, such as bacterial infections, by
suppressing the production of proinflammatory cytokines by Telemetry
activated macrophages on binding of ACh to the α7nAChR.15 Blood pressure (BP) and HR were measured by telemetry combined
Here we demonstrate that stroke directly activates cholinergic with fast Fourier transform analysis of BP and HR, as described
anti-inflammatory pathways, which are crucially involved in elsewhere.19 In mice aged 8 to 10 weeks, transmitters (TA11PA-C10
BP; Data Sciences International) were implanted in a subcutaneous
the development of poststroke lung infections. pocket along the right flank with the pressure-sensing catheter in the
abdominal aorta. After 7 days recovery, baseline was recorded us-
Materials and Methods ing DATAQUEST software (A.R.T. 2.1, Data Sciences International),
A detailed description of all materials and methods can be found in and sympathetic and parasympathetic activities were verified phar-
the online-only Data Supplement. macologically by applying metoprolol (4 mg/kg) and atropine (2 mg/
kg), respectively. Parasympathetic activity was assessed via spectral
analysis using PV-wave software (Visual Numerics). The power spec-
Animals and Housing tra of systolic BP, pulse interval time series, and the cross spectra
Male specific pathogen free C57Bl6/J mice (Charles River were calculated using fast Fourier transformation. Low-frequency
Laboratories, Sulzfeld, Germany), sex-mixed α7nAChR knockout components of pulse intervals spectrum (LF), root mean square of
(KO) mice,16 and wild-type (WT) littermates, respectively (B6.129S7- successive differences between adjacent normal pulse intervals, total
Chrna7tm1Bay/J; The Jackson Laboratory, Bar Harbor) were housed in power, and the baroreflex sensitivity (BRS-LF) were calculated. The
cages lined with chip bedding and environmental enrichment (mouse baroreceptor HR reflex was investigated using sequence technique
tunnel and igloo) on a 12 h light/dark cycle with ad libitum access to and spectral function analysis of spontaneous changes in systolic
food and water. Experiments were performed using 11- to 14-week- BP and HR, and BRS was defined as the mean magnitude value of
old mice in a­ ccordance with the European directive on the protec- transfer function between systolic BP and R-R interval in the low-
tion of animals used for scientific purposes and all other applicable frequency band (BRS-LF).
regulations and approved by the relevant authority, Landesamt für
Gesundheit und Soziales, Berlin, Germany.
Flow Cytometry
Cell phenotyping was performed on an LSRII flow cytometer us-
Experimental Stroke ing FACSDiva software (BD Biosciences) and FlowJo software
Under general isoflurane anaethesia, a small silicon-coated filament (Tree Star Inc.) with the following anti-mouse monoclonal antibod-
was temporarily introduced via the left common carotid artery into ies: CD45-PE-Cy7, CD3-APC, CD4-A700, CD8-PB, NK1.1-PE,
the circle of Willis blocking the origin of the middle cerebral artery CD11b-APC-Cy7, B220-FITC, SiglecF-PE, F4/80-A700, CD11c-
for 60 min (middle cerebral artery occlusion [MCAo])17 following the FITC, CD11b-PE-TR, Gr1-PB, and MHCII-APC-Cy7 (Biolegend).
standard operating procedures of our laboratory. Success of MCAo Fluorescence-activated cell sorter–based cell sorting was performed
was verified by Bederson score, and animals without neurological on a FACSAria III cell sorter (BD Biosciences).
deficit after operation were excluded from the study.
Isolation of Primary Alveolar Macrophages and
Vagotomy Alveolar Epithelial Cells
Five days before MCAo surgery and under isoflurane anesthesia, the MΦ were obtained by repeated gentle flushing of tracheotomized
left cervical vagus nerve was carefully and bluntly dissected from the lungs from naïve C57Bl6/J mice with NaCl, centrifugation (300g,
common carotid artery and, in the vagotomy but not sham-operated 10 min), and resuspension in Roswell Park Memorial Institute
group, transected. medium. For purification of AEC,20 lungs were flushed after per-
fusion and BAL with 1500 μL dispase followed by intratracheal
Broncho-Alveolar Lavage and Measurement of injection of 500 μL of 1% low melting temperature agarose. Lung
single-cell suspension (homogenization in DMEM) underwent
Bacterial Burden magnetic cell separation for depletion of nonepithelial cells using
Under anaethesia with midazolam (5 mg/kg body weight) and me- anti-CD45, anti-CD31, and anti-CD16/CD32 mAb (BioLegend).
detomidin (0.5 mg/kg body weight) IP (antagonization with flumaze- Purity of macrophages and epithelial cells was confirmed by flow
nil 0.5 mg/kg body weight and atipamezol 5 mg/kg body weight, if cytometry (>95% and >90%, respectively).
needed), mice were intubated with a 22G peripheral venous catheter.
Afterward, 0.4 mL of saline plus 0.2 mL air were applied over the
tubus and immediately withdrawn.18 For the determination of colo- Analysis of Ex Vivo Cytokine Production
ny-forming units, broncho-alveolar lavage (BAL) fluid was serially Lung single-cell suspensions (2×106/mL), isolated primary MΦ
diluted, plated on LB agar plates, and incubated at 37°C for 18 h, (1×105/mL), or primary AEC (1×10E5/mL) were stimulated with li-
and bacterial colonies were counted.8 The cut-off value (<104 colony- popolysaccharide (LPS) (1 μg/mL; Sigma) or Pam3CSK4 (200 ng/
forming units/mL) has been determined by investigating BAL fluid mL; Invivogen) in the absence or presence of different concentrations
of naïve mice. of nicotine or PNU282987 (both from Sigma) for 12 h at 37°C.
3234  Stroke  November 2015

Analysis of Cytokines in BAL and Cell Culture adaptation time for up slopes and down slopes of systemic
Supernatants BP were increased compared with baseline and sham controls
Cytokines in BAL and cell culture supernatant were measured by (Figure II in the online-only Data Supplement).
magnetic bead–based assay on a Bio-Plex 200 System (Luminex
xMAP Technology) with Bio-Plex Manager 6.0 software (Bio-Rad)
according to the manufacturer’s recommendations. Unilateral Vagotomy Reduces Bacterial Burden by
Restoring Pulmonary Immune Function
MCAo animals develop spontaneous bacterial pneumonia
Quantitative Reverse Transcriptase Polymerase
between day 1 and day 3 after experimental cerebral isch-
Chain Reaction
RNA was extracted from fluorescence-activated cell–sorted or
emia.8 To investigate the influence of parasympathetic acti-
Dynabead-purified lung cells using the NucleoSpin RNA II Kit vation on poststroke infections, we first performed unilateral
(Macherey-Nagel). cDNA synthesis was done using the QuantiTect cervical vagotomy. In mice that underwent vagotomy 5 days
kit (Qiagen) and measured by real-time quantitative reverse transcrip- before MCAo, we observed a significant reduction in bacte-
tase polymerase chain reaction (7500 Sequence Detection System, rial load on day 3 after MCAo compared with sham vagoto-
Applied Biosystems) with β-Actin as housekeeping gene.
mized stroke animals (Figure 2A). Importantly, administration
of nicotine in drinking water reversed the beneficial effect of
Statistics vagotomy on bacterial pneumonia after stroke (Figure 2A).
Treatment allocation was performed randomly and remained com-
pletely blinded for the examiners. Appropriate 2-sided statistical tests
Infarct volumes and HR did not significantly differ between
were used to compare groups for significant differences (P<0.05; SPSS the groups (Figure III in the online-only Data Supplement).
Statistics 18.0, IBM). Closed testing procedure was applied in telem- Assessment of ex vivo cytokine secretion by lung cells iso-
etry experiments, putting the comparisons at different time points in lated 1 day after MCAo revealed reduced release of proinflam-
a hierarchical order. A priori sample size calculation (G*Power3.021) matory cytokines interleukin (IL)-6 and tumor necrosis factor
assumed a change in bacterial burden by 100-fold as relevant or a dou-
bling of HR variability (HRV) parameters, respectively. Prespecified (TNF)-α after stimulation with the Toll-like receptor 4 (TLR4)
inclusion and exclusion criteria and detailed report on flow of ani- agonist LPS (Figure 2B and 2C) and the TLR1/2 agonist
mals through the experiments are reported in the online-only Data Pam3CSK4 (Figure IV in the online-only Data Supplement) in
Supplement (Tables I and II in the online-only Data Supplement). MCAo mice compared with sham controls. In contrast, lung
cells isolated from mice vagotomized 5 days before MCAo
Results released similar amounts of IL-6 and TNF-α after Pam3CSK4
Middle Cerebral Artery Occlusion Reduces HR and and LPS stimulation as lung cells from sham-operated controls.
Increases Blood Pressure Treatment of MCAo animals with nicotine in drinking water
decreased ex vivo proinflammatory cytokine production after
To obtain unbiased data on physiological responses after exper-
TLR stimulation of lung cells isolated from vagotomized mice
imental stroke in mice, we established a telemetric approach
to measure HR and BP in a quiet, undisturbed environment. In to levels observed in non- or sham-vagotomized animals (Figure
MCAo mice, HR was decreased during the first 3 days after 2B and 2C; Figure IV in the online-only Data Supplement).
surgery compared with baseline and sham-operated animals.
BP rose in both groups, but remained significantly elevated Role of α7nAChR in Diminished Antibacterial
for 5 days in MCAo animals compared with sham controls Defense in Lung After Experimental Stroke
(Figure IA and IB in the online-only Data Supplement). The α7nAChR receptor has been described as mediator of
Physical activity was slightly reduced in both groups com- cholinergic anti-inflammatory effects on macrophages.14
pared with baseline with no group differences in the first 2 Similar to the effects of vagotomy in WT animals, mice lack-
weeks (Figure IC in the online-only Data Supplement). ing the α7nAChR showed significantly less bacterial load in
lungs after MCAo compared with WT littermates (Figure 3A).
Increased HRV and Baroreflex Sensitivity Indicates Nicotine did not further increase lung bacterial burden in WT
Parasympathetic Activation After Experimental mice. In α7nAChR KO mice, bacterial loads in lung increased
Stroke only slightly by nicotine treatment but remained at lower lev-
In contrast to sham-operated animals, we observed a strong els compared with WT animals. Infarct volumes and HR did
increase in the LF power component of HRV reflecting not significantly differ between the groups (Figure 3B and
increased parasympathetic activity in MCAo-treated mice on 3C). The rather small number of male compared with female
day 1, 3, and 5 after surgery (Figure 1A). These findings were animals per group does not allow a definitive conclusion on
corroborated by changes in 2 time domain parameters describ- sex-related differences of α7nAChR KO mice in response to
ing parasympathetic activity in HRV, the standard deviation of stroke. However, subgroup analysis of infarct volumes and
RR intervals (Figure 1B), and the root mean square of succes- bacterial burden in lung does not suggest sex-specific differ-
sive RR differences (Figure 1C). ences (data not shown).
BRS is the ability to adapt the sinus frequency in response
to BP changes. Cross spectral analysis of the BRS-LF, calcu- Cholinergic Signaling Attenuates Antibacterial
lated as mean value of the transfer function between systemic Defense of Lung Myeloid and Epithelial Cells After
BP and pulse intervals in the LF band, revealed a significant Experimental Stroke
increase within the first 5 days after experimental stroke To differentiate between cholinergic effects on BM-derived
(Figure 1D). One, 3, and 5 days after experimental stroke, the immune cells and resident lung cells contributing to poststroke
Engel et al   Cholinergic Pathway in Stroke-Associated Pneumonia    3235

Figure 1. Markers of parasympathetic activity in heart rate (HR) variability are increased during the first days after experimental stroke.
Heart rate variability was assessed by frequency domain analysis (low-frequency HR variability oscillation [LF-HRV], A), time domain anal-
ysis as the standard deviation of RR intervals (SD-RRI, B), and the root mean square of successive RR differences (RMSSD, C). Barore-
flex sensitivity reflects the ability of the body to adapt the heart rate to changes of blood pressure, and less sensitive after middle cerebral
artery occlusion (MCAo) in low-frequency oscillations (BRS-LF, D). Data from 2 independent experiments with n=14 (baseline), n=11 (d1),
n=11 (d3), n=9 (d5) to n=7 (d14) animals; *indicates P<0.05 (Mann–Whitney U test in closed testing procedure).

immunodepression, we investigated the development of post- expression was not observed in endothelial cells, NK-, B-, and
stroke pneumonia in mice responding to α7nAChR-mediated T-lymphocytes, dendritic cells, or granulocytes (Figure 5).
cholinergic signaling either in BM-derived cells (WT→KO)
or resident lung cells (KO→WT). For control, we transplanted Diminished Cytokine Production in AEC and MΦ
BM cells from WT into WT animals (WT→WT) and from upon α7nAChR Stimulation
KO into KO animals (KO→KO). After experimental stroke, To investigate cell type–dependent α7nAChR-mediated
bacterial burden in lung was significantly reduced in KO→KO effects, we assessed TLR ligand–induced IL-6 release in pri-
mice compared with WT→WT mice (Figure 4A), corroborat- mary AEC and MΦ. Nicotine diminished LPS-induced IL-6
ing findings observed in α7nAChR KO and WT mice (com- production in AEC isolated from WT but not α7nAChR KO
pare Figure 3A). Both KO→WT and WT→KO chimeric mice mice in a dose–dependent manner (Figure 6A). Likewise,
showed an intermediate phenotype regarding susceptibility to α7nAChR-specific agonist PNU282987 suppressed IL-6
pneumonia compared with syngeneic controls (Figure 4A). release in LPS-stimulated AEC from WT but not α7nAChR
Lung cells isolated 3 days after MCAo from ischemic KO mice (Figure 6C). In contrast, nicotine and PNU282987
WT→WT mice produced significantly lower TNF-α levels after dose-dependently reduced LPS-induced IL-6 secretion in
LPS stimulation than lung cells from KO→KO mice, whereas MΦ from WT and to lesser extent also in α7nAChR KO mice
lung cells from WT→KO and KO→WT chimeric mice again (Figure 6B and 6D). These data suggest that cholinergic con-
exhibited an intermediate phenotype compared with KO→KO trol of TLR-induced proinflammatory cytokine release by MΦ
and WT→WT mice (Figure 4B). is partly independent from α7nAChR.

mRNA Expression Pattern of α7nAChR in Alveolar Discussion


Macrophages and Alveolar Epithelial Cells Besides neurological deficits leading to dysphagia and aspira-
To determine the cellular expression pattern of α7nAChR in tion, suppression of peripheral immune responses because of an
lung tissue, we sorted single cell suspensions from WT mice overactivation of stress pathways has been recognized as a criti-
by magnetic cell sorting. α7nAChR was ≈50-fold higher cal risk factor for the development of infectious complications
expressed in AEC compared with MΦ. In contrast, α7nAChR after stroke. Previous studies in experimental stroke models have
3236  Stroke  November 2015

Figure 2. A, Vagotomy prevents poststroke pneumonia. In contrast to sham vagotomy, vagotomy 5 days before middle cerebral artery
occlusion (MCAo) significantly reduced bacterial load in the lung at day 3 after experimental stroke. This effect is abrogated if animals
were treated before MCAo with the acetylcholine receptor agonist nicotine in drinking water. Dashed line: bacterial load in broncho-
alveolar lavage fluid from naïve mice (n=32; winsorized; Kruskal–Wallis test P=0.026; *indicating P<0.05 in Dunn’s post hoc test; pooled
data from 2 independent experiments). B and C, Vagotomy improves ex vivo cytokine release by lung cells after stroke. Vagotomy (+) or
sham vagotomy (Sham) was performed 5 days before MCAo, and nonvagotomzied animals (−) served as controls. Lung cells (2×106/mL)
were isolated on day 1 after MCAo or Sham surgery and stimulated ex vivo for 12 h with lipopolysaccharide (LPS). Vagotomy significantly
increased Toll-like receptor (TLR)-induced interleukin (IL)-6 (B) and tumor necrosis factor (TNF)-α (C) production after MCAo. Again, treat-
ment of animals with nicotine in drinking water reversed the beneficial effect of vagotomy on ex vivo TLR-induced cytokine release. Bar
graphs show mean±SD; n=39; 1-way ANOVA with Dunnett’s post hoc test using vagotomy MCAo without Nicotine as group of compari-
son (*indicates P<0.05, **P<0.01, and ***P<0.001; pooled data from 2 independent experiments).

provided evidence that an increased activity of the sympathetic as well as impaired ex vivo T cell function in stroke patients
nervous system after cerebral ischemia leads to rapid impair- before the onset of infectious complications.6,10 Thus, although
ment of peripheral cellular immune responses, in particular natu- pathophysiological responses in human diseases and respective
ral killer cell and T cell responses. These experimental findings mouse models may markedly differ as recently shown for the
were corroborated by clinical studies showing increased levels genomic response in acute inflammatory conditions,22 the MCAo
of catecholamines and reduced T cell counts in peripheral blood, model in mice has proven to be clinically relevant to elucidate

Figure 3. α7 nicotinic acetylcholine receptor (α7nAChR) mediates susceptibility to poststroke pneumonia. A, Wild-type (WT) and α7nAChR
knockout mice were subjected to middle cerebral artery occlusion (MCAo), and bacterial load in lungs was determined 3 days after experi-
mental stroke. α7nAChR showed significantly lower bacterial burden compared with WT animals. Treatment of animals with nicotine in drink-
ing water did not significantly increase bacterial load α7nAChR knockout mice. The dashed line indicates bacterial load in broncho-alveolar
lavage fluid from naïve mice. Neither heart rate (B) nor infarct volume (C) differed significantly between WT and α7 nAChR knockout mice.
(n=56; pooled data from 3 independent experiments; Kruskal–Wallis test P=0.001 with *P<0.05, **P<0.01, and ***P<0.001 in Dunn’s post
hoc test.)
Engel et al   Cholinergic Pathway in Stroke-Associated Pneumonia    3237

Figure 4. α7 nicotinic acetylcholine receptor (α7nAChR) on bone marrow–derived and resident lung epithelial cells is involved in cho-
linergic anti-inflammatory effects after stroke. To assess the relative contribution of α7nAChR expression on bone marrow–derived and
resident lung epithelial cells in antibacterial defense after middle cerebral artery occlusion (MCAo), we generated bone marrow chimeras
between α7nAChR knockout mice (α7 KO) and wild-type (WT) littermates, as described in Materials and Methods. A, Bacterial burden in
lung was significantly reduced only if α7nAChR was absent on both bone marrow–derived and epithelial cells (n=52; Kruskal–Wallis test
P=0.01 with * indicating P<0.05 in Dunn’s post hoc test, pooled data from 2 independent experiments). B, Lung cells (2×106/mL) were iso-
lated at day 3 after MCAo and stimulated ex vivo with lipopolysaccharide (1 μg/mL). Tumor necrosis factor (TNF)-α concentration in cell
culture supernatants was measured 12 h after stimulation. TNF-α release was significantly increased only if α7nAChR was absent on both
bone marrow–derived and epithelial cells (n=21; Kruskal–Wallis test P=0.01 with * indicating P<0.05 in Dunn’s post hoc test).

mechanisms of increased susceptibility to infection and to harmful overactivation of the immune system in response to
develop preventive treatment strategies.23 However, it remained infections or inflammation-induced tissue damage. Sensing
unclear whether cerebral ischemia affects resident lung cells, of peripheral inflammatory processes via vagal afferent fibers
which mediate first-line pulmonary immune responses during was initially proposed to trigger an anti-inflammatory response
respiratory infections. Our data suggest that lung innate immune via release of acetylcholine from vagal efferent fibers, which
responses are suppressed after stroke by increased cholinergic acts on α7nAChR on macrophages to suppress the release of
signaling, which is relayed by the parasympathetic vagus nerve proinflammatory cytokines.14 Recent evidence suggests, how-
and the α7nAChR expressed on AEC and MΦ. Hence, both an ever, that the neural reflex that is triggered by and dampens
impaired early T/NK-cell dysfunction mediated by sympathetic systemic inflammatory responses may be more complex and
nervous system and an impaired innate immune response in the involve non-neural cholinergic and noradrenergic sympathetic
lung mediated by the parasympathetic nervous system seem to mechanisms.24 Nevertheless, vagus nerve stimulation by cen-
be important additive mechanisms leading to decreased anti- trally acting muscarinic agonists, such as CNI-1493, has been
bacterial defense and poststroke bacterial infections. shown to suppress peripheral inflammatory responses that can
The cholinergic anti-inflammatory pathway has been be prevented by vagotomy, demonstrating that the vagus is
described as a protective mechanism that prevents potentially an essential efferent pathway of the central cholinergic anti-
inflammatory pathway.25
Assessment of HRV has been recognized to provide reli-
able, noninvasive information on the activity of the autonomic
nervous system, including its vagal and sympathetic compo-
nents. Using a telemetry system, we observed an immediate
and long-lasting activation of the parasympathetic nervous
system after cerebral ischemia, which is in line with previous
observations showing a similar increase in root mean square
of successive RR differences and LF-HRV after experimental
stroke in mice.26 Because HRV parameters have been shown
to differ between species with respect to the influence of para-
sympathetic and sympathetic outflow,27 we have verified the
specificity of selected HRV measures by pharmacological
blockade of sympathetic and parasympathetic input using the
selective antagonists metoprolol and atropine, respectively.19
Figure 5. α7 nicotinic acetylcholine receptor (α7nAChR) mRNA is
expressed only in lung epithelial cells and alveolar macrophages. According to the few clinical data, autonomic dysfunctions in
Single lung cell suspensions from naïve wild-type animals were stroke patients occur in the acute phase with a predominance
sorted by flow cytometry into alveolar epithelial cells (AEC), of parasympathetic rather than sympathetic dysfunction.28–30
endothelial cells, and different leukocyte subsets, and α7nAChR Although measurement of HRV reflects primarily auto-
mRNA levels was determined by reverse transcriptase poly-
merase chain reaction (n=3 independent fluorescence-activated nomic influences on the cardiovascular system, several
cell sorter (FACS) sort experiments, n.d. not detectable). human studies have linked changes in HRV with markers of
3238  Stroke  November 2015

Figure 6. Lipopolysaccharide (LPS)-induced


release of interleukin (IL)-6 from alveolar epi-
thelium cells (AEC) as well as macrophages
(MF) is diminished by cholinergic stimulation.
Isolated primary AEC and macrophages
(MΦ) from wild-type (WT; white bars) or α7
nicotinic acetylcholine receptor (α7nAChR)
knockout (KO) mice (gray bars) were stimu-
lated ex vivo with LPS (1 μg/mL) for 12 h in
the absence or presence of different concen-
trations of the nonselective nAChR agonists
nicotine (A and B) or α7nAChR-selective
agonist PNU282987 (C and D). Both nico-
tine (A) and PNU282987 (C) suppressed
LPS-induced IL-6 production in AEC from WT
but not α7nAChR KO mice in a dose-depen-
dant manner. In contrast, nicotine (B) reduced
the LPS-induced IL-6 secretion in MF isolated
from both WT and α7nAChR mice, whereas
PNU282987 (D) at lower concentrations
selectively suppressed IL-6 secretion in MF
from WT, but not α7nAChR KO mice. Data
are presented as mean±SD with *P<0.05 and
**P<0.01 (2-way ANOVA and Bonferroni post
hoc test, n=4–12 per group, pooled data from
2 independent experiments).

inflammation and immunity.31,32 An increase in sympathetic -injury.37 Our study provides direct evidence for the crucial
and parasympathetic HRV parameters is associated with role of the cholinergic anti-inflammatory pathway in the
worse outcome after stroke.9,33,34 development of spontaneous pneumonia after MCAo in mice.
In the present study, we demonstrated that increased Both, MΦ and AEC play a crucial role in the first-line
parasympathetic activity after cerebral ischemia plays a cru- defense against pulmonary infections.11,12 Cholinergic signals
cial role in mediating increased susceptibility to bacterial transmitted via the α7nAChR have been previously shown to
pneumonia after stroke. Although MCAo animals developed limit inflammatory responses in these cells.38 Accordingly,
spontaneous bacterial infections, either disruption of para- we examined whether resident lung immune competent cells
sympathetic signaling in WT mice by vagotomy or inhibi- or BM-derived immune cell are predominantly targeted by
tion of cholinergic effects by using α7nAChR-deficient mice cholinergic anti-inflammatory signaling after stroke. Our
reduced bacterial burden in BAL fluid almost completely to results using a model of BM chimerism between WT and
levels observed in naïve mice (<104 colony-forming units / α7nAChR-deficient mice suggest that the inhibitory effects of
mL).35 Furthermore, vagotomy resulted in the restoration of α7nAChR-mediated cholinergic signaling are exerted on both
inflammatory responses triggered by TLR agonist in lung lung-resident cells and BM-derived immune cells because
cells in stroke animals. Moreover, treatment of MCAo animals chimeric mice showed an intermediate phenotype in terms
with the nonselective nAChR agonist nicotine prevented the of bacterial burden and ex vivo cytokine secretion capacity
beneficial effect of vagotomy on lung bacterial burden and on of whole lung cells compared with syngeneic WT→WT and
ex vivo cytokine secretion by lung cells in WT mice but only KO→KO controls.
slightly increased bacterial load in α7nAChR mice, indicating To further elucidate pulmonary cellular targets of cholin-
that suppression of pulmonary antibacterial responses after ergic anti-inflammatory signaling, we determined α7nAchR
stroke by cholinergic signals involves α7nAChR. Because we expression in sorted cells isolated from lung. We observed
have not assessed all physiological parameters under para- α7nAChR expression only in AEC and MΦ but not in endo-
sympathetic control in these experiments, other than immune thelial cells or other leukocyte subsets. Ex vivo functional
mechanisms might be involved in poststroke pneumonia. analysis revealed that the anti-inflammatory effects of cho-
However, vagotomy resulted in the restoration of inflamma- linergic signaling on AEC were largely dependent on the
tory responses triggered by TLR agonist in lung cells in stroke presence of the α7-subunit nAChR because suppression of
animals. Recently, Lafargue et al made the important obser- TLR-induced proinflammatory cytokine production by the
vation that mice deficient in α7nAChR showed reduced lung nonselective nAchR agonist nicotine and the selective α7
injury and mortality compared with WT mice in an aspiration- agonist PNU282987 was observed in primary AEC from
induced Pseudomonas aeruginosa pneumonia model after WT but not α7nAChR-deficient mice. In contrast, although
experimental stroke.36 However, the Pseudomonas aspiration PNU282987 showed α7nAChR-dependent inhibitory effects
pneumonia model does not reflect the pathophysiological on MΦ, nicotine dose-dependently suppressed IL-6 produc-
course of lung infections in stroke patients because it requires tion in these cells also in the absence of α7nAChR. Although
a high bacterial inoculum, which by itself may trigger (neu- α7nAChR was consistently detectable in MΦ, expression lev-
ral) mechanisms to perturb pulmonary immune defense other els were ≈100-fold lower compared with primary AEC. These
than those induced by central activation of cholinergic anti- data suggest that nicotinic receptors other than α7nAchR may
inflammatory mechanisms after acute central nervous system be involved in cholinergic anti-inflammatory signaling in MΦ.
Engel et al   Cholinergic Pathway in Stroke-Associated Pneumonia    3239

Several studies suggested that stimulation of the choliner- cell type 1-like immunostimulation. J Exp Med. 2003;198:725–736.
doi: 10.1084/jem.20021098.
gic anti-inflammatory pathway via vagus nerve and α7nAChR
9. Walter U, Kolbaske S, Patejdl R, Steinhagen V, Abu-Mugheisib M,
may also be a neuroprotective strategy.39–41 However, our data Grossmann A, et al. Insular stroke is associated with acute sympathetic
strongly suggest that inhibition of peripheral parasympathetic hyperactivation and immunodepression. Eur J Neurol. 2013;20:153–
actions is beneficial in terms of medical complications after 159. doi: 10.1111/j.1468-1331.2012.03818.x.
10. Chamorro Á, Meisel A, Planas AM, Urra X, van de Beek D, Veltkamp
stroke, whereas an additional stimulation of the cholinergic R. The immunology of acute stroke. Nat Rev Neurol. 2012;8:401–410.
pathway might exacerbate the risk for life-threatening bacte- doi: 10.1038/nrneurol.2012.98.
rial infections. Thus, immunomodulatory therapeutic strategies 11. Hippenstiel S, Opitz B, Schmeck B, Suttorp N. Lung epithelium as a
will have to consider and keep the balance between reversal of sentinel and effector system in pneumonia–molecular mechanisms of
pathogen recognition and signal transduction. Respir Res. 2006;7:97.
the peripheral stroke-induced immunodepression and concur- doi: 10.1186/1465-9921-7-97.
rent preservation of its putative neuroprotective effects. 12. Evans SE, Xu Y, Tuvim MJ, Dickey BF. Inducible innate resistance
In summary, although activation of cholinergic anti- of lung epithelium to infection. Annu Rev Physiol. 2010;72:413–435.
doi: 10.1146/annurev-physiol-021909-135909.
inflammatory mechanisms during local and systemic inflam- 13. Maouche K, Polette M, Jolly T, Medjber K, Cloëz-Tayarani I, Changeux
matory responses may be a protective feedback mechanism to JP, et al. {alpha}7 nicotinic acetylcholine receptor regulates airway
limit potentially harmful effects of overt inflammation, central epithelium differentiation by controlling basal cell proliferation. Am J
nervous system injury–induced activation of such pathway in Pathol. 2009;175:1868–1882. doi: 10.2353/ajpath.2009.090212.
14. Tracey KJ. The inflammatory reflex. Nature. 2002;420:853–859.
the absence of an inflammatory stimulus may inappropriately doi: 10.1038/nature01321.
inhibit first-line antimicrobial responses and increase suscep- 15. Rosas-Ballina M, Tracey KJ. Cholinergic control of inflammation. J Intern
tibility to infectious complications after acute central nervous Med. 2009;265:663–679. doi: 10.1111/j.1365-2796.2009.02098.x.
16. Orr-Urtreger A, Göldner FM, Saeki M, Lorenzo I, Goldberg L, De Biasi
system injury.
M, et al. Mice deficient in the alpha7 neuronal nicotinic acetylcholine
receptor lack alpha-bungarotoxin binding sites and hippocampal fast
Acknowledgments nicotinic currents. J Neurosci. 1997;17:9165–9171.
17. Engel O, Kolodziej S, Dirnagl U, Prinz V. Modeling stroke in mice—
We thank Sabine Kolodziej, Carena Teufelhart, Verena Wörtmann,
middle cerebral artery occlusion with the filament model. J Vis Exp.
Yvonne Amoneit, Claudia Muselmann-Genschow (Charité), and
2011:e2423
Ilona Kamer (MDC) for excellent technical assistance. 18. Walters DM, Wills-Karp M, Mitzner W. Assessment of cellular profile
and lung function with repeated bronchoalveolar lavage in individual
Sources of Funding mice. Physiol Genomics. 2000;2:29–36.
19. da Costa-Goncalves AC, Tank J, Plehm R, Diedrich A, Todiras M,
This study was supported by German Research Foundation (Exc257, Gollasch M, et al. Role of the multidomain protein spinophilin in blood
SFB-TRR84, SFB-TRR43, FOR1336, SFB1021 C05, Excellence pressure and cardiac function regulation. Hypertension. 2008;52:702–
Cluster Cardio-Pulmonary System), German Federal Ministry of 707. doi: 10.1161/HYPERTENSIONAHA.108.114355.
Education and Research (01EO0801, German Center for Lung 20. Unkel B, Hoegner K, Clausen BE, Lewe-Schlosser P, Bodner J,
Research), European Community FP7 (201024), and Hessen State Gattenloehner S, et al. Alveolar epithelial cells orchestrate DC func-
Ministry of Higher Education, Research and the Arts (Universities tion in murine viral pneumonia. J Clin Invest. 2012;122:3652–3664.
Giessen & Marburg Lung Center). doi: 10.1172/JCI62139.
21. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statisti-
cal power analysis program for the social, behavioral, and biomedical
Disclosures sciences. Behav Res Methods. 2007;39:175–191.
A patent on preventive antibacterial therapy after stroke has been 22. Seok J, Warren HS, Cuenca AG, Mindrinos MN, Baker HV, Xu W, et al;
filed to the European Patent Office (PCT/EP03/02246) by Drs Volk, Inflammation and Host Response to Injury, Large Scale Collaborative
Dirnagl, C. Meisel, and A. Meisel. The other authors report no Research Program. Genomic responses in mouse models poorly mimic
conflicts. human inflammatory diseases. Proc Natl Acad Sci U S A. 2013;110:3507–
3512. doi: 10.1073/pnas.1222878110.
23. Dirnagl U, Endres M. Found in translation: preclinical stroke research pre-
References dicts human pathophysiology, clinical phenotypes, and therapeutic outcomes.
1. Davenport RJ, Dennis MS, Wellwood I, Warlow CP. Complications after Stroke. 2014;45:1510–1518. doi: 10.1161/STROKEAHA.113.004075.
acute stroke. Stroke. 1996;27:415–420. doi: 10.1161/01.STR.27.3.415. 24. Martelli D, Yao ST, McKinley MJ, McAllen RM. Reflex control
2. Emsley HC, Hopkins SJ. Acute ischaemic stroke and infection: recent of inflammation by sympathetic nerves, not the vagus. J Physiol.
and emerging concepts. Lancet Neurol. 2008;7:341–353. doi: 10.1016/ 2014;592(Pt 7):1677–1686. doi: 10.1113/jphysiol.2013.268573.
S1474-4422(08)70061-9. 25. Olofsson PS, Rosas-Ballina M, Levine YA, Tracey KJ. Rethinking
3. Kalra L, Yu G, Wilson K, Roots P. Medical complications during stroke inflammation: neural circuits in the regulation of immunity. Immunol
rehabilitation. Stroke. 1995;26:990–994. doi: 10.1161/01.STR.26.6.990. Rev. 2012;248:188–204. doi: 10.1111/j.1600-065X.2012.01138.x.
4. Koennecke HC, Belz W, Berfelde D, Endres M, Fitzek S, Hamilton F, et 26. Lubjuhn J, Gastens A, von Wilpert G, Bargiotas P, Herrmann O,
al; Berlin Stroke Register Investigators. Factors influencing in-hospital Murikinati S, et al. Functional testing in a mouse stroke model induced
mortality and morbidity in patients treated on a stroke unit. Neurology. by occlusion of the distal middle cerebral artery. J Neurosci Methods.
2011;77:965–972. doi: 10.1212/WNL.0b013e31822dc795. 2009;184:95–103. doi: 10.1016/j.jneumeth.2009.07.029.
5. Steinman L. Elaborate interactions between the immune and nervous 27. Gehrmann J, Hammer PE, Maguire CT, Wakimoto H, Triedman JK,
systems. Nat Immunol. 2004;5:575–581. doi: 10.1038/ni1078. Berul CI. Phenotypic screening for heart rate variability in the mouse.
6. Meisel C, Schwab JM, Prass K, Meisel A, Dirnagl U. Central nervous Am J Physiol Heart Circ Physiol. 2000;279:H733–H740.
system injury-induced immune deficiency syndrome. Nat Rev Neurosci. 28. Tokgözoglu SL, Batur MK, Top uoglu MA, Saribas O, Kes S, Oto A.
2005;6:775–786. doi: 10.1038/nrn1765. Effects of stroke localization on cardiac autonomic balance and sudden
7. Guilliams M, De Kleer I, Henri S, Post S, Vanhoutte L, De Prijck S, death. Stroke. 1999;30:1307–1311. doi: 10.1161/01.STR.30.7.1307.
et al. Alveolar macrophages develop from fetal monocytes that differ- 29. Korpelainen JT, Sotaniemi KA, Huikuri HV, Myllyä VV. Abnormal heart
entiate into long-lived cells in the first week of life via GM-CSF. J Exp rate variability as a manifestation of autonomic dysfunction in hemi-
Med. 2013;210:1977–1992. doi: 10.1084/jem.20131199. spheric brain infarction. Stroke. 1996;27:2059–2063. doi: 10.1161/01.
8. Prass K, Meisel C, Höflich C, Braun J, Halle E, Wolf T, et al. Stroke- STR.27.11.2059.
induced immunodeficiency promotes spontaneous bacterial infections 30. Xiong L, Leung HH, Chen XY, Han JH, Leung TW, Soo YO,
and is mediated by sympathetic activation reversal by poststroke T helper et al. Comprehensive assessment for autonomic dysfunction in
3240  Stroke  November 2015

different phases after ischemic stroke. Int J Stroke. 2013;8:645–651. aeruginosa lung injury. FASEB J. 2012;26:2919–2929. doi: 10.1096/
doi: 10.1111/j.1747-4949.2012.00829.x. fj.11-197384.
31. Thayer JF. Vagal tone and the inflammatory reflex. Cleve Clin J Med. 37. Ben Mohamed F, Mohamed FB, Garcia-Verdugo I, Medina M, Balloy
2009;76(suppl 2):S23–S26. doi: 10.3949/ccjm.76.s2.05. V, Chignard M, et al. A crucial role of Flagellin in the induction of
32. Marsland AL, Gianaros PJ, Prather AA, Jennings JR, Neumann SA, airway mucus production by Pseudomonas aeruginosa. PLoS One.
Manuck SB. Stimulated production of proinflammatory cytokines cova- 2012;7:e39888. doi: 10.1371/journal.pone.0039888.
ries inversely with heart rate variability. Psychosom Med. 2007;69:709– 38. Greene CM, Ramsay H, Wells RJ, O’Neill SJ, McElvaney NG. Inhibition
716. doi: 10.1097/PSY.0b013e3181576118. of Toll-like receptor 2-mediated interleukin-8 production in Cystic Fibrosis
33. Sykora M, Diedler J, Poli S, Rizos T, Turcani P, Veltkamp R, et al. airway epithelial cells via the alpha7-nicotinic acetylcholine receptor.
Autonomic shift and increased susceptibility to infections after acute Mediators Inflamm. 2010;2010:423241. doi: 10.1155/2010/423241.
intracerebral hemorrhage. Stroke. 2011;42:1218–1223. doi: 10.1161/ 39. Miyamoto O, Pang J, Sumitani K, Negi T, Hayashida Y, Itano T.
STROKEAHA.110.604637. Mechanisms of the anti-ischemic effect of vagus nerve stimulation in the
34. Günther A, Salzmann I, Nowack S, Schwab M, Surber R, Hoyer H, gerbil hippocampus. Neuroreport. 2003;14:1971–1974. doi: 10.1097/01.
et al. Heart rate variability: a potential early marker of sub-acute wnr.0000091686.94870.61.
post-stroke infections. Acta Neurol Scand. 2012;126:189–196. 40. Ottani A, Giuliani D, Mioni C, Galantucci M, Minutoli L, Bitto A, et al.
doi: 10.1111/j.1600-0404.2011.01626.x. Vagus nerve mediates the protective effects of melanocortins against
35. Sze MA, Tsuruta M, Yang SW, Oh Y, Man SF, Hogg JC, et al. Changes cerebral and systemic damage after ischemic stroke. J Cereb Blood Flow
in the bacterial microbiota in gut, blood, and lungs following acute LPS Metab. 2009;29:512–523. doi: 10.1038/jcbfm.2008.140.
instillation into mice lungs. PLoS One. 2014;9:e111228. doi: 10.1371/ 41. Kalappa BI, Sun F, Johnson SR, Jin K, Uteshev VV. A positive allo-
journal.pone.0111228. steric modulator of α7 nAChRs augments neuroprotective effects of
36. Lafargue M, Xu L, Carlès M, Serve E, Anjum N, Iles KE, et al. Stroke- endogenous nicotinic agonists in cerebral ischaemia. Br J Pharmacol.
induced activation of the α7 nicotinic receptor increases Pseudomonas 2013;169:1862–1878. doi: 10.1111/bph.12247.

You might also like