You are on page 1of 9

Oral Treatment with Live Lactobacillus reuteri Inhibits

the Allergic Airway Response in Mice


Paul Forsythe1, Mark D. Inman2, and John Bienenstock1
1
The Brain–Body Institute and Department of Pathology and Molecular Medicine and 2Firestone Institute for Respiratory Health,
McMaster University, and St. Joseph’s Healthcare, Hamilton, Ontario, Canada

Rationale: Clinical trials have demonstrated that probiotics may be


effective in the treatment and prevention of atopic disease in chil-
AT A GLANCE COMMENTARY
dren but there have been few reports of therapeutic effects of oral
probiotics outside the gastrointestinal tract. Scientific Knowledge on the Subject
Objectives: We investigated the effect of two probiotic organisms
Despite indications that probiotics can modulate immune
on the response to antigen challenge in a mouse model of allergic
airway inflammation.
responses in the lung, there have been no reports of the
Methods: We used an ovalbumin-sensitized asthma model in BALB/c effects of oral treatment with a probiotic on major charac-
and Toll-like receptor 9–deficient mice. Animals were treated with teristics of asthma, including airway inflammation and hyp-
probiotic organisms via gavaging needle before antigen challenge. erresponsiveness.
After antigen challenge, airway responsiveness to methacholine,
influx of inflammatory cells to the lung, and cytokine levels in bron-
What This Study Adds to the Field
choalveolar lavage fluid were assessed.
Results: Oral treatment with live Lactobacillus reuteri but not Lactoba- Oral treatment with a probiotic organism can attenuate
cillus salivarius significantly attenuated the influx of eosinophils to major characteristics of an asthmatic response, including
the airway lumen and parenchyma and reduced the levels of tumor airway eosinophilia, local cytokine responses, and hyperres-
necrosis factor, monocyte chemoattractant protein-1, IL-5, and ponsiveness.
IL-13 in bronchoalveolar lavage fluid of antigen-challenged animals,
but there was no change in eotaxin or IL-10. L. reuteri but not
L. salivarius also decreased allergen-induced airway hyperrespon-
siveness. These responses were dependent on Toll-like receptor 9
and were associated with increased activity of indoleamine 2,3- of probiotics in the treatment of experimental colitis and food
dioxygenase. Killed organisms did not mimic the ability of the live allergy (2, 3, 5) and several studies have characterized the ability
L. reuteri to attenuate inflammation or airway hyperresponsiveness. of various strains of probiotics to alter the activity and cytokine
Conclusion: Oral treatment with live L. reuteri can attenuate major production of the gut and associated lymphoid tissue, clarifica-
characteristics of an asthmatic response in a mouse model of allergic tion of the underlying mechanisms of these antiinflammatory
airway inflammation. These results suggest that oral treatment with effects is still obscure.
specific live probiotic strains may have therapeutic potential in the In addition to effects in the gut, there is evidence that pro-
treatment of allergic airway disease. biotics, given perinatally, may be protective against manifesta-
tions of atopic disease. Clinical trials have demonstrated that
Keywords: airway inflammation; bronchial hyperresponsiveness;
probiotics; Toll-like receptor 9; mouse model
Lactobacillus rhamnosus GG may be effective in treatment and
prevention of early atopic disease in children (6, 7). Lactobacillus
Probiotics are defined as live microorganisms, which, when con- fermentum was shown to be beneficial in improving the extent
sumed in adequate numbers, confer a health benefit on the host and severity of atopic dermatitis in young children (8). Such
(1). Organisms used as probiotics are most frequently of the findings have lead to an increased interest in the role of commen-
Lactobacillus or Bifidobacterium species and are generally com- sal organisms in the regulation of systemic immune responses.
mensals, occurring naturally as part of the gut microbiota. There Although long-term follow-up studies of children at high risk
is increasing evidence to support a therapeutic role for probiotics for developing allergy have not shown significant beneficial ef-
in the treatment of various inflammatory conditions. Random- fects of probiotics on the incidence of asthma (7), there is evi-
ized controlled trials of probiotics have proved successful in dence indicating that oral administration of certain microbial
patients with chronic pouchitis and irritable bowel syndrome (2, organisms can modulate immune responses in the lung. Rats
3). A multispecies probiotic (VSL#3; VSL, Gaithersburg, MD) orally immunized with killed Pseudomonas aeruginosa exhibited
given to IL-10–deficient mice with established colitis normalized enhanced bacterial clearance from the airways compared with
gut barrier function, reduced proinflammatory cytokines IL-1 nonimmunized donors after intratracheal challenge with live
and tumor necrosis factor (TNF)-␣, and diminished histologic P. aeruginosa (9). This response was associated with increases in
evidence of disease (4). Although there is support for the efficacy bronchoalveolar neutrophils and in recruitment and phagocytic
activity of alveolar macrophages. Furthermore, infection with
the gut-restricted bacterium Citrobacter rodentium attenuated
the airway eosinophilia that results from pulmonary Cryptococ-
cus neoformans infection (10). A double-blind randomized trial
(Received in original form June 19, 2006; accepted in final form December 20, 2006 )
found that oral treatment with the probiotic L. rhamnosus GG
Correspondence and requests for reprints should be addressed to John Bienens- reduced the rate and severity of respiratory virus infection in
tock, M.D., Departments of Pathology and Molecular Medicine, McMaster Univer- children (11). Oral treatment of mice with Lactobacillus casei
sity, Brain–Body Institute, St. Joseph’s Healthcare, 50 Charlton Avenue East, T3304
Hamilton, ON, L8N 4A6 Canada. E-mail: bienens@mcmaster.ca
increased pulmonary natural killer cell activity and enhanced
IFN-␥ and TNF production by nasal lymphocytes. The probiotic-
Am J Respir Crit Care Med Vol 175. pp 561–569, 2007
Originally Published in Press as DOI: 10.1164/rccm.200606-821OC on January 4, 2007 treated animals also had reduced viral titers in nasal washings
Internet address: www.atsjournals.org after influenza infection (12).
562 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 175 2007

Despite indications that probiotic treatment can modulate ODN]), and an inactive control oligonucleotide (mutated oligodeoxyri-
some immune responses in the lung, there have been no reports bonucleotides [M-ODN]) were supplied to us by Dr. E. Raz, University
of the effects of oral treatment with a probiotic organism on of California, San Diego.
major characteristics of asthma, including allergic airway in-
Treatment Protocols
flammation and bronchial hyperresponsiveness. Here, we de-
scribe the ability of one probiotic organism, Lactobacillus reuteri, Commencing on Day 6 (i.e., at time of second OVA sensitization),
but not another, Lactobacillus salivarius, to attenuate antigen- animals received 1 ⫻ 109 live, heat-killed, or irradiated L. reuteri,
L. salivarius, or equivalent isolated L. reuteri DNA (50 ␮g) in 200 ␮l
induced eosinophil influx to the airway as well as local cytokine
of MRS broth via a gavaging needle for 9 consecutive days. Animals
responses and hyperresponsiveness to methacholine in an oval- gavaged with broth alone served as control animals. In other experi-
bumin (OVA)-sensitized mouse model of allergic airway disease. ments, isolated L. reuteri DNA, calf thymus DNA, or ISS-ODN
Portions of the data presented in this article have previously (50 ␮g) was given intraperitoneally on Days 6, 8, 10, and 12.
been published in abstract form (13).
Airway Responsiveness
METHODS Airway responsiveness was assessed based on the response of total
respiratory system resistance (RRS) to saline and increasing intravenous
Animals doses of methacholine. RRS was measured as described previously (17)
Adult male BALB/c or Toll-like receptor 9–deficient (TLR9⫺/⫺) mice using the flow interrupter technique, modified for use in mice. The
(20–25 g) were maintained in an automatic light/dark cycle (light periods slope of the dose response was calculated by linear regression between
of 12 h) and provided water and chow ad libitum. Mice were acclima- the measured RRS and the log10-transformed methacholine dose, using
tized to the animal facility for 1 week before experimentation. Trans- data from the 10-, 33-, and 100-␮g doses.
genic TLR9⫺/⫺ mice on a balb/c background were obtained from Dr.
K. Rosenthal (McMaster University) with permission from Dr. S. Akira BAL
(Department of Host Defense, Osaka University, Osaka, Japan). Age- Two aliquots of 250 ␮l PBS were injected and withdrawn through a
matched (8–9 wk old) animals were used in all experiments. These tracheal cannula. Airway inflammation was assessed by inflammatory
experiments were performed in accordance with guidelines of the Cana- cell counts in BAL fluid. Cells were removed from BAL fluid by centri-
dian Council for Animal Care. fugation at 200 ⫻ g for 15 minutes, and supernatants stored at ⫺80⬚C
until evaluation of cytokine content. Cells were resuspended in PBS
Experimental Asthma (1 ml). BAL cells were stained with Trypan blue, and viable cells
This study used an OVA-sensitized mouse model of allergic airway counted using a hemocytometer. Smears of BAL cells were prepared
inflammation. Briefly, mice were sensitized by intraperitoneal injection with a Cytospin (Thermo Shandon, Pittsburgh, PA) and stained with
of 20 ␮g OVA adsorbed with 500 ␮g alum in saline on Day 0 and Day HEMA 3 reagent (Biochemical Sciences, Swedesboro, NJ) for differen-
6. On Days 12 and 14, mice were challenged intranasally with 5 ␮g tial cell counts. A total of 200 cells were counted for each lavage, and
OVA per mouse (14). Twenty-four hours after the last challenge (Day cells were classified, based on morphologic criteria, as macrophages,
15), mice were subjected to measurements of airway responsiveness neutrophils, lymphocytes, or eosinophils. An independent observer
followed by bronchoalveolar lavage (BAL). OVA/alum-sensitized, sa- blinded to the experimental conditions performed all cell counts.
line-challenged mice served as control animals.
Lung Histology
Bacterial Preparations
Lungs were inflated with 10% formalin (to a pressure of 20 cm H2O),
L. reuteri were purchased originally from the American Type Culture fixed for 24 hours, and embedded in paraffin. Fixed and embedded
Collection (ATCC No. 23272; Manassas, VA). L. salivarius were a gift tissue was stained with hematoxylin and eosin for histologic assessment
from Dr. B. Kiely (Alimentary Health, Cork, Ireland). Both strains are using light microscopy.
of human intestinal origin and have demonstrated antiinflammatory
effects in animal models of colitis (15, 16). From frozen stocks (⫺80⬚C), Cytokine Measurement
bacteria were suspended in Man-Rogosa-Sharpe liquid medium (MRS
Cytokines (IL-6, IL-10, IL-12, monocyte chemoattractant protein
broth; Difco Laboratories, Detroit, MI) plated in MRS agar, cultured
[MCP]-1, TNF, IFN-␥) were assessed using the BD Cytometric Bead
anaerobically at 37⬚C for 24 hours, then inoculated in fresh MRS broth
Array System (BD Biosciences, San Jose, CA) or commercial ELISA
and grown at 37⬚C under anaerobic conditions for 48 hours in 50-ml
for individual cytokines (IL-5, IL-13, and eotaxin; R&D Systems, Min-
tubes. After 2 days, tubes were centrifuged at 2,000 rpm for 15 minutes
neapolis, MN).
at 20⬚C and washed twice with sterile phosphate-buffered saline (PBS)
to a concentration of 6 ⫻ 108 bacteria/ml as determined by a Vitek Indoleamine 2,3-Dioxygenase Activity
colorimeter (bioMérieux, Hazelwood, MO). Bacterial suspensions were
centrifuged in 15-ml tubes at 2,000 rpm for 15 minutes at 20⬚C, superna- Maximal indoleamine 2,3-dioxygenase (IDO) activity was indirectly
tants discarded, and bacteria resuspended in MRS broth to give a determined as described previously (18). Briefly, lung tissue homoge-
concentration of 5 ⫻ 109/ml. Heat-killed L. reuteri were prepared by nates were incubated with an excess of exogenous tryptophan (Tryp;
heating aliquots of viable bacterial suspensions for 20 minutes at 80⬚C. 780 ␮M) as a substrate in the reaction buffer for 30 minutes in vitro.
Suspensions of L. reuteri were killed by ␥-irradiation with Cobalt 60 The resulting kynurenine (Kyn) levels in the reaction homogenates
for 20 hours at 8.05 Gy/minute. The resulting viability was determined were measured by HPLC. IDO activity was expressed as nanograms
by plating on MRS agar plates under anaerobic conditions for 72 hours of Kyn per milligram of protein per 30-minute interval (14). Protein
at 37⬚C. No bacterial growth was detected in either the heat-killed or concentrations were measured using a BioRad protein assay (BioRad,
irradiated L. reuteri preparations after 72 hours of culture. Hercules, CA). To determine systemic IDO activity, plasma Kyn and
Tryp were simultaneously determined in HPLC and expressed as a
DNA Isolation ratio (19).
Genomic DNA was isolated from L. reuteri using the EndoFree DNA
isolation kit (Qiagen, Valencia, CA) according to the manufacturer’s
Statistics
instructions. The purity of DNA was confirmed by measuring the ultra- Experimental results are expressed as means ⫾ SEM. Statistical analy-
violet 260/280 absorbance ratio (⬎ 1.8). Calf thymus DNA (Sigma ses were performed with unpaired two-tailed Student’s t tests or one-
Chemical Co., St Louis, MO) was used as a control in all experiments way analysis of variance, followed by Newman-Keuls test for comparing
using bacterial DNA. A synthetic immunostimulatory CpG-containing all pairs of groups (GraphPad Prism, version 4.0; GraphPad, San Diego,
oligodeoxyribonucleotide sequence, which acts as a ligand for TLR-9 CA). A p value of less than 0.05 was considered statistically significant,
(immunostimulatory CpG-containing oligodeoxyribonucleotides [ISS- and n represents the number of experiments performed.
Forsythe, Inman, and Bien: Probiotics and Airway Inflammation 563

RESULTS white blood cells, respectively; n ⫽ 6, p ⬍ 0.001) (data not


shown).
Inflammatory Cell Influx to the Airway
Total cell numbers in BAL fluids were significantly increased Cytokine Levels in BAL Fluid
24 hours after the final antigen challenge in OVA-sensitized mice After OVA challenge, levels of MCP-1, TNF, IL-5, IL-13, and
compared with OVA-sensitized, saline-challenged mice (155.9 ⫾ eotaxin were significantly increased in BAL fluid compared with
30.1 ⫻ 104 vs. 18.08 ⫾ 0.8 ⫻ 104, respectively; n ⫽ 12, p ⬍ saline-challenged animals (Figure 3). Levels of IFN-␥, IL-12,
0.001) (Figure 1A), thus confirming that challenge with OVA and IL-10 were not significantly changed compared with saline
was effective. Although the cell population in BAL fluid from controls. Treatment with L. reuteri significantly attenuated the
saline-challenged mice consisted almost exclusively of alveolar increase in MCP-1, TNF, IL-5, and IL-13 but did not alter eotaxin
macrophages, OVA challenge caused a dramatic increase in the levels. There was no effect of L. reuteri on IL-10, whereas IFN-␥
proportion of eosinophils (Figure 1B). OVA-challenged mice and IL-12 were below the limit of detection in all groups assessed.
gavaged with L. reuteri had a significant reduction in cells recov- Treatment with either heat-killed or irradiated L. reuteri also
ered in BAL fluid compared with MRS broth–fed animals (59.2 significantly attenuated the increase in MCP-1 TNF and IL-5
⫾ 11.8 ⫻ 104 vs. 155.9 ⫾ 30.1 ⫻ 104, respectively; n ⫽ 12, p ⫽ after antigen challenge; however, in contrast to treatment with
0.004) (Figure 1A). This corresponded to a significant decrease live bacteria, killed organisms did not significantly alter levels
in both eosinophil (38.2 ⫾ 2.7 ⫻ 104 vs. 114.6 ⫾ 19.0 ⫻ 104) and of IL-13 (Figure 3). L. salivarius had no effect on cytokine levels
macrophage numbers (19.1 ⫾ 3.6 ⫻ 104 vs. 38.6 ⫾ 4.5 ⫻ 104) in BAL fluid (data not shown).
(Figure 1C). Histologic analysis demonstrated that the increase
in eosinophils in the lung parenchyma of OVA-sensitized and Airways Response to Methacholine
-challenged mice was also reduced by treatment with live Challenge with OVA in sensitized animals produced an increase
L. reuteri (Figure 2). Treatment of animals with either heat- in airway responsiveness to methacholine, as determined by an
killed or ␥-irradiated organisms did not attenuate eosinophil increase in the maximum resistance from 3.9 ⫾ 0.9 to 7.0 ⫾
influx to the airway (Figure 1). In contrast to L. reuteri, 0.8 cm H2O/ml/second (n ⫽ 15, p ⬍ 0.01) and the slope of
live L. salivarius did not modulate cellular influx to the airway dose–response curve from 1.9 ⫾ 0.7 to 4.8 ⫾ 0.7 (p ⬍ 0.01). This
(Figure 1D). hyperresponsiveness was significantly reduced by treatment with
live L. reuteri (maximum resistance, 4.5 ⫾ 0.7 cm H2O/ml/s;
Blood Eosinophils slope, 2.9 ⫾ 0.6) but not with heat-killed or irradiated organisms
To determine if the reduction in eosinophil influx to the airway (Figure 4).
was entirely due to a decrease in migration from the blood, or L. salivarius did not modulate airway responsiveness to meth-
also a result of decreased production/release from bone marrow, acholine (data not shown).
we measured circulating eosinophils. In OVA-challenged ani-
mals, treatment with live L. reuteri resulted in a significant reduc- IDO Activity
tion in the population of circulating eosinophils compared with The plasma ratio of Kyn to Tryp was significantly increased after
MRS broth–fed animals (4.7 ⫾ 0.54 vs. 1.6 ⫾ 0.27% of total treatment with live L. reuteri versus MRS broth (0.24 ⫾ 0.05 vs.

Figure 1. Effect of oral treatment with live


(LR), heat-killed (HK), and ␥-irradiated (IR)
L. reuteri (1 ⫻ 109 organisms daily for 9
consecutive days) on (A ) total and (B )
differential cell counts (macrophages, eo-
sinophils, neutrophils, and lymphocytes)
and (C ) absolute numbers of macro-
phages and eosinophils in bronchoal-
veolar lavage (BAL) fluid from ovalbumin
(OVA)-sensitized male mice 24 hours after
challenge with intranasal OVA or saline.
(D ) The effect of live L. salivarius treat-
ment (1 ⫻ 109 organisms daily for 9 con-
secutive days) on eosinophil numbers in
BAL fluid is also shown. Each column rep-
resents the mean ⫾ SEM (n ⫽ 10). *p ⬍
0.05; **p ⬍ 0.01 compared with Man-
Rogosa-Sharpe broth–treated control.
564 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 175 2007

Figure 2. Representative sections of lung tissue from L. reuteri–treated (A ) and untreated (B ) OVA-sensitized mice after antigen challenge. A section
from a saline-challenged control animal is shown for comparison (C ). Arrows indicate inflammatory cell influx to the parenchyma. This influx was
markedly reduced by treatment with live L. reuteri.

0.09 ⫾ 0.03, respectively; n ⫽ 10, p ⬍ 0.05). Heat-killed and and increased airway responsiveness compared with saline-
irradiated organisms did not modulate the Kyn:Tryp ratio (Fig- challenged animals. However, treatment with live L. reuteri did
ure 5A). Maximal IDO in lung tissue was not changed by treat- not result in attenuation of eosinophil influx (99.2 ⫾ 32.7 ⫻ 104),
ment with either live or killed organisms (Figure 5B). BAL cytokine levels (data not shown), or airway responsiveness
to methacholine (Figure 6).
The Role of TLR-9 in L. reuteri Treatment
As in wild-type animals, OVA sensitization and challenge Effect of Isolated L. reuteri DNA
of TLR9⫺/⫺ mice led to eosinophil influx to the airway (76.2 ⫾ Given the requirement for TLR-9 in the effects of L. reuteri,
12.4 ⫻ 104 vs. 0.62 ⫾ 0.1 ⫻ 104, n ⫽ 6, p ⬍ 0.001) (Figure 6) we assessed the ability of DNA isolated from the organism to

Figure 3. Effect of oral treatment with live


(LR), heat-killed (HK), and ␥-irradiated (IR)
L. reuteri (1 ⫻ 109 organisms daily for 9 con-
secutive days) on cytokine levels in BAL fluid
from OVA-sensitized male mice 24 hours
after challenge with intranasal OVA or saline.
Each column represents the mean ⫾ SEM (n
⫽ 10). *p ⬍ 0.05 compared with OVA chal-
lenged, Man-Rogosa-Sharpe broth–treated
control.
Forsythe, Inman, and Bien: Probiotics and Airway Inflammation 565

Figure 4. The effect of oral treatment with live


(LR) L. reuteri (A ) and heat-killed (HK) and ␥-
irradiated (IR) organisms (B ) on airway reactivity
(slope RRS) and maximum resistance (Max RRS) in
OVA-sensitized male mice 24 hours after chal-
lenge with intranasal OVA or saline. Airway re-
sponsiveness was measured in response to in-
creasing doses of intravenous methacholine.
Using the resulting dose–response curve, indices
of airway reactivity and maximal degree of bron-
choconstriction (Max RRS) were determined.
Each column represents the mean ⫾ SEM (n ⫽
10). *p ⬍ 0.05; **p ⬍ 0.01 compared OVA chal-
lenged, Man-Rogosa-Sharpe broth–treated con-
trol. MCh ⫽ methacholine.

reduce allergic airway inflammation. As previously described by viable L. reuteri, before intranasal antigen challenge, resulted
Hayashi and colleagues (20), intraperitoneal administration of in significant attenuation of inflammatory cell influx to the lung
the synthetic TLR-9 ligand ISS-ODN significantly reduced eosin- and airway responsiveness to methacholine. The decreases in
ophil influx to the airway (110.6 ⫾ 17.0 vs. 45.2 ⫾ 3.5 ⫻ 104, cellular influx and airway hyperresponsiveness (AHR) were as-
n ⫽ 10, p ⬍ 0.05); however, neither treatment with L. reuteri sociated with reduced levels of inflammatory cytokines (MCP-1,
nor calf thymus DNA had any effect on the allergic airway TNF, IL-5, IL-13) in the BAL fluid. However, oral consumption
response when administered either orally or through intraperito- of live L. reuteri did not result in an increase in the Th1-type
neal injection (data not shown). cytokines IL-12 and IFN-␥ or the antiinflammatory cytokine
IL-10.
DISCUSSION Strain-specific characteristics of this probiotic appear to be re-
sponsible for the antiinflammatory action we observed. Gavage
Studies have demonstrated strong immunomodulatory proper- with an alternative probiotic Lactobacillus strain, L. salivarius,
ties of lactobacilli, which include antiinflammatory and antitu- did not modulate the allergic airway response in our model.
mor activity, modulation of autoimmune diseases, and preven- Strain-specific effects of probiotics have been demonstrated in
tion of infections (4, 21–23). There is clinical evidence suggesting other systems, and the distinct immune responses between
that probiotic treatment is protective against atopic dermatitis strains may be due to differing inherent characteristics of the
and studies indicating that certain probiotic strains, when com- organisms, which include persistence in the gut, colonization,
bined with allergens, are candidates for mucosal vaccination and intrinsic immunogenicity. In this regard, the cytokine profile
against allergy. Kruisselbrink and colleagues constructed an elicited by a probiotic organism clearly plays an important role
L. plantarum strain that expressed an immunodominant T-cell in determining the immunologic outcome.
epitope of the Der p1 allergen of the house dust mite (24). In a detailed investigation, Maassen and colleagues (26) ana-
Intranasal administration of this organism suppressed antigen- lyzed eight common Lactobacillus strains with respect to induc-
induced Th1 and Th2 immune responses in Der p1–sensitized tion of cytokines by the murine gut mucosa in response to a
animals. Coapplication of Lactococcus lactis and L. plantarum parenterally administered antigen and found distinct cytokine
with the major birch pollen allergen (Bet v 1) caused suppression profiles elicited by different strains. L. reuteri induced several
of allergen-induced basophil degranulation (25). In a mouse proinflammatory and/or Th1 cytokines, such as IL-1␤, IL-2, and
model of food allergy, feeding mice with L. casei was shown to TNF, but not antiinflammatory or Th2 cytokines, such as IL-10
prevent responses to antigen challenge (5). and IL-4. L. casei tended to induce production of IL-10 and
Here, we have studied the effect of L. reuteri on an allergic IL-4. In this system, L. reuteri enhanced the systemic antibody
airway response in OVA-sensitized mice. Oral treatment with response to antigen. In contrast, in a study of the capacity of
566 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 175 2007

ated by killed bacteria (31). Furthermore, the dependency on


live organisms for efficacy appears to be strain specific. In vitro,
L. reuteri can reduce TNF-induced IL-8 production by human
intestinal epithelial cells (32). This effect is dependent on live
organisms that are in contact with the epithelial cells. However,
the same modulation of IL-8 release has been reported with
both live and heat-killed L. rhamnosus GG (33). In contrast to
the inability of killed L. reuteri to modulate the allergic response
in the airway, Hunt and colleagues demonstrated that intragas-
tric administration of heat-killed Mycobacterium vaccae signifi-
cantly reduced pulmonary inflammation after antigen challenge
in OVA-sensitized mice (34). However, treatment with M. vac-
cae resulted in a markedly different cytokine profile in BAL
fluid than that observed after L. reuteri feeding, with significantly
increased IL-10 levels and no change in IL-5. These differences
in cytokine profile and requirement for live organisms suggest
that killed M. vaccae and live L. reuteri exert their protective
effects via distinct mechanisms. Importantly, L. reuteri reduces
both inflammation and AHR, whereas the impact of intragastric
M. vaccae treatment on AHR has not been reported. It does
not necessarily follow that reduction in airway inflammation
leads to reduced airway responsiveness, and a causal relationship
between eosinophilic inflammation and AHR is far from
established.
Figure 5. Effect of oral treatment with live (LR), heat-killed (HK), and Despite clear evidence for strong immunomodulatory proper-
␥-irradiated (IR) L. reuteri (1 ⫻ 109 organisms daily for 9 consecutive ties of lactobacilli, the mechanisms underlying these effects are
days) on activity of indoleamine 2,3-dioxygenase (IDO) in plasma and poorly understood. Our results indicate that feeding with live
lung tissue as assessed by kynurenine (Kyn) to tryptophan (Tryp) ratio
L. reuteri increased systemic IDO activity as assessed by the
and maximum IDO activity per milligram of protein, respectively. IDO
ratio of Tryp to Kyn in plasma. There is evidence, in animal
activity was assessed in OVA-sensitized mice 24 hours after challenge
models of colitis, that a mixture of eight different probiotic
with intranasal OVA or saline. Each column represents the mean ⫾ SEM
(n ⫽ 10). *p ⬍ 0.05 compared with OVA-challenged, Man-Rogosa-
organisms (VSL#3) mediates their antiinflammatory effect
Sharpe broth–treated control. through activation of TLR-9 by unmethylated CpG motifs de-
rived from the organism’s own DNA (35). Furthermore, Hayashi
and coworkers recently demonstrated that parenterally adminis-
tered synthetic immunostimulatory DNA sequences (ISS-ODN)
Lactobacillus strains to induce cytokine production from bone that act as TLR-9 ligands significantly reduced allergen-induced
airway inflammation in mice (20). In this system, the antiin-
marrow–derived dendritic cells (DCs), L. reuteri proved to be
flammatory response was dependent on TLR-9–mediated activa-
a poor IL-12 inducer, whereas L. casei strongly induced IL-12,
tion of the tolerogenic enzyme IDO. However, unlike intraperi-
IL-6, and TNF production (27). Therefore, it is clear that changes
toneal treatment with ISS-ODN described by Hayashi and
in cytokine profile induced by probiotic strains may be site spe-
colleagues, we did not see an increase in maximum IDO activity
cific and dependent on the experimental system used.
in the lung after oral administration of L. reuteri (20). The ISS-
Furthermore, Pelto and colleagues (28) demonstrated that
ODN–induced increase in IDO activity in pulmonary epithelial
probiotic bacteria modulate phagocytosis differently in healthy
cells is dependent on IFN-␥, whereas no changes in IFN-␥ were
and allergic subjects. In healthy people, there was a stimulatory
detectable in the airway after L. reuteri treatment. Pulmonary
effect, whereas in subjects with milk hypersensitivity, there was IDO activity is essential to the antiinflammatory effects of ISS-
down-regulation of the immune response (28). Therefore, the ODN in the airway; therefore, any role for IDO in the attenua-
outcome of probiotic treatment may depend on the immunologic tion of the allergic response by L. reuteri will be distinct from
state of the host; in this regard, it will be of interest to determine that induced by the synthetic TLR-9 ligand.
the impact of specific probiotics on allergic inflammation in a IDO is expressed in various cell types, including fibroblasts
range of mouse strains with distinct immune responses. (36), trophoblasts (37), macrophages, epithelial cells, DCs (38),
Neither heat-killed nor ␥-irradiated organisms mimicked the and nonimmune cells of the lung (20). IDO has been implicated
ability of the live L. reuteri to protect against antigen-induced in T-cell tolerance to tumors and dysfunctional self-tolerance in
eosinophil influx or AHR. However, killed organisms did retain nonobese diabetic mice, as a protective negative regulator in
some immunomodulatory capacity in that they significantly at- autoimmune disorders (39), and has shown a protective role
tenuated antigen-induced increases in TNF, MCP-1, and IL-5. in a mouse model of colitis (40). Perhaps most notably, the
IL-13 is critical to the development of experimental asthma after maintenance of a clinically unresponsive state after aeroallergen
acute allergen exposure (29, 30). Delivery of IL-13 to the airway exposure in atopic individuals has been associated with increased
induces mucus production, AHR, and eosinophilia, whereas systemic IDO activity and IL-10 production (41). Thus, while
blockade of IL-13 in animal models of allergic asthma leads to we did not see an overt increase in IL-10 in BAL fluid, the
inhibition of all of these effects (29). Therefore, the observation enhancement of systemic IDO activity by an oral probiotic may
that live but not killed L. reuteri inhibit IL-13 may be central be significant to the therapeutic potential of these organisms.
to the distinct ability of live organisms to modulate cell influx The effect of L. reuteri on the allergic airway response was
and AHR. dependent on TLR-9. However, neither ␥-irradiated organisms
It is clear that certain physiologic effects of probiotics are that retain intact DNA nor isolated DNA, whether given system-
dependent on live organisms, whereas other effects can be medi- ically or orally, altered inflammatory cell influx or AHR. This
Forsythe, Inman, and Bien: Probiotics and Airway Inflammation 567

Figure 6. Effect of oral treatment with live


(LR) L. reuteri (1 ⫻ 109 organisms daily
for 9 consecutive days) on (A ) eosinophil
numbers in BAL fluid and (B ) airway re-
sponsiveness to methacholine in OVA-
sensitized TLR-9–deficient mice. Cell
counts and airway responsiveness were
assessed 24 hours after challenge with intra-
nasal OVA or saline. Airway responsiveness
was measured in response to increasing
doses of intravenous methacholine. Using
the resulting dose–response curve, indices
of airway reactivity (slope RRS) and maxi-
mal degree of bronchoconstriction (Max
RRS) were determined. Each column repre-
sents the mean ⫾ SEM (n ⫽ 6). WT ⫽
wild type. *p ⬍ 0.05 compared with OVA-
challenged, Man-Rogosa-Sharpe broth–
treated control.

may indicate that the antiinflammatory effects of oral L. reuteri CD8⫹Treg cells) (46). Indeed, DCs expressing IDO contribute
treatment in our model are not solely mediated by bacterial to the generation and maintenance of peripheral tolerance by
DNA acting on TLR-9, and it is possible that metabolic products depleting autoreactive T cells and by inducing Treg responses
of L. reuteri act in concert with TLR-9 to mediate the antiin- (47). In other studies, L. reuteri and L. casei, but not L. plan-
flammatory response. Another potential explanation for our re- tarum, have been demonstrated to prime monocyte-derived DCs
sults is that intact, live organisms are required for effective pre- to drive the development of Treg cells (48) In vivo, oral treatment
sentation of bacterial DNA to TLR-9. TLR-9 is an intracellular with L. rhamnosus induced T-cell hyporesponsiveness in healthy
receptor, localized to late endosomes or lysosomes (42). Results human volunteers and patients with Crohn’s disease with corre-
from recent studies suggest that the intracellular localization of sponding in vitro studies, suggesting this occurred via modulation
TLR-9 controls access of the receptor to different sources of of DC function (49). The capacity of lactobacilli to variably
DNA (43). In mammals, the extracellular environment contains induce maturation and the cytokine profile expressed by DCs
DNase I, which promotes degradation or extracellular DNA. indicate that different species of probiotics may differentially
Therefore, the isolated DNA from L. reuteri, whether adminis-
determine whether a DC drives Th1, Th2, or a Treg response,
tered orally or intraperitoneally, was presumably degraded ex-
and subsequently whether DCs have therapeutic potential in the
tracellularly before it could interact with TLR-9. This is consis-
treatment of allergic disorders.
tent with the finding that purified viral DNA is a poor inducer
In conclusion, we report that oral administration of live L.
of TLR-9 when compared with the intact virus (44). Similarly,
reuteri but not L. salivarius resulted in significant attenuation of
Rachmilewitz and colleagues found that, although oral treatment
with intact probiotic organisms could attenuate disease severity the allergic airway response. This effect depends on both viable
in a mouse model of colitis via TLR-9, DNA isolated from this organisms and TLR-9 and is associated with systemic evidence
same mixture of probiotic organisms was ineffective via the oral reflecting IDO activation. Our results indicate that the immuno-
route (35). modulatory effects of oral treatment with L. reuteri are not
Although, to date, little is known about the fate of probiotic limited to the gastrointestinal tract and we present the first report
organisms in the gastrointestinal tract, Macpherson and Uhr (45) that oral treatment with a probiotic organism can attenuate
demonstrated that intestinal DCs can ingest and retain small major characteristics of an asthmatic response, including airway
numbers of live commensals for several days. In this state, they eosinophilia, local cytokine responses, and hyperresponsiveness
drive DCs to a tolerogenic phenotype. Furthermore, these DCs to methacholine. These beneficial effects appear to be strain
did not move beyond the draining mesenteric lymph node. specific, and it is clear that realizing the true potential of pro-
Therefore, if this mechanism is in play in our model system, biotics as regulators of allergic airway disease will require a
local effects at or before the mesenteric lymph node (MLN) must better understanding of the mechanisms behind the quantitative
be assumed to be responsible for the systemic down-regulation of and qualitative differences in immune regulation that exist
the allergic response observed in the lung. Given that DCs are among candidate organisms.
pivotal in early bacterial recognition and are essential in pre-
venting asthmatic reactions to inhaled antigens, these cells may Conflict of Interest Statement : None of the authors has a financial relationship
with a commercial entity that has an interest in the subject of this manuscript.
be central in mediating the beneficial effects of probiotics on the
allergic airway response. DCs can induce a range of regulatory Acknowledgment : The authors thank Ursula Kadela, Gudrun Goettsche, and
T-cell subtypes (CD4⫹ CD25⫹, IL-10–producing Treg, and Jennifer Wattie for their invaluable technical support.
568 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 175 2007

References 23. Matsuzaki T, Hashimoto S, Yokokura T. Effects on antitumor activity


and cytokine production in the thoracic cavity by intrapleural adminis-
1. Guarner F, Schaafsma GJ. Probiotics. Int J Food Microbiol 1998;39:237–
tration of Lactobacillus casei in tumor-bearing mice. Med Microbiol
238. Immunol (Berl) 1996;185:157–161.
2. Gionchetti P, Rizzello F, Helwig U, Venturi A, Lammers KM, Brigidi 24. Kruisselbrink A, Heijne Den Bak-Glashouwer MJ, Havenith CE, Thole
P, Vitali B, Poggioli G, Miglioli M, Campieri M. Prophylaxis of pou- JE, Janssen R. Recombinant Lactobacillus plantarum inhibits house
chitis onset with probiotic therapy: a double-blind, placebo-controlled dust mite-specific T-cell responses. Clin Exp Immunol 2001;126:2–8.
trial. Gastroenterology 2003;124:1202–1209. 25. Repa A, Grangette C, Daniel C, Hochreiter R, Hoffmann-Sommergruber
3. O’Mahony L, McCarthy J, Kelly P, Hurley G, Luo F, Chen K, O’Sullivan K, Thalhamer J, Kraft D, Breiteneder H, Mercenier A, Wiedermann
GC, Kiely B, Collins JK, Shanahan F, et al. Lactobacillus and bifido- U. Mucosal co-application of lactic acid bacteria and allergen induces
bacterium in irritable bowel syndrome: symptom responses and rela- counter-regulatory immune responses in a murine model of birch
tionship to cytokine profiles. Gastroenterology 2005;128:541–551. pollen allergy. Vaccine 2003;22:87–95.
4. Madsen K, Cornish A, Soper P, McKaigney C, Jijon H, Yachimec C, 26. Maassen CB, van Holten-Neelen C, Balk F, den Bak-Glashouwer MJ,
Doyle J, Jewell L, De Simone C. Probiotic bacteria enhance murine Leer RJ, Laman JD, Boersma WJ, Claassen E. Strain-dependent in-
and human intestinal epithelial barrier function. Gastroenterology duction of cytokine profiles in the gut by orally administered Lactoba-
2001;121:580–591. cillus strains. Vaccine 2000;18:2613–2623.
5. Kim H, Kwack K, Kim DY, Ji GE. Oral probiotic bacterial administration 27. Christensen HR, Frokiaer H, Pestka JJ. Lactobacilli differentially modu-
suppressed allergic responses in an ovalbumin-induced allergy mouse late expression of cytokines and maturation surface markers in murine
model. FEMS Immunol Med Microbiol 2005;45:259–267. dendritic cells. J Immunol 2002;168:171–178.
6. Kalliomaki M, Salminen S, Poussa T, Arvilommi H, Isolauri E. Probiotics 28. Pelto L, Isolauri E, Lilius EM, Nuutila J, Salminen S. Probiotic bacteria
and prevention of atopic disease: 4-year follow-up of a randomised down-regulate the milk-induced inflammatory response in milk-hyper-
placebo-controlled trial. Lancet 2003;361:1869–1871. sensitive subjects but have an immunostimulatory effect in healthy
7. Kalliomaki M, Salminen S, Arvilommi H, Kero P, Koskinen P, Isolauri subjects. Clin Exp Allergy 1998;28:1474–1479.
E. Probiotics in primary prevention of atopic disease: a randomised 29. Kuperman DA, Huang X, Koth LL, Chang GH, Dolganov GM, Zhu Z,
placebo-controlled trial. Lancet 2001;357:1076–1079. Elias JA, Sheppard D, Erle DJ. Direct effects of interleukin-13 on
8. Weston S, Halbert A, Richmond P, Prescott SL. Effects of probiotics epithelial cells cause airway hyperreactivity and mucus overproduction
on atopic dermatitis: a randomised controlled trial. Arch Dis Child in asthma. Nat Med 2002;8:885–889.
2005;90:892–897. 30. Kamiya T, Wang L, Forsythe P, Goettsche G, Mao Y, Wang Y, Tougas
9. Cripps AW, Dunkley ML, Clancy RL. Mucosal and systemic immuniza- G, Bienenstock J. Inhibitory effects of Lactobacillus reuteri on visceral
tions with killed Pseudomonas aeruginosa protect against acute respi- pain induced by colorectal distension in Sprague-Dawley rats. Gut
ratory infection in rats. Infect Immun 1994;62:1427–1436. 2006;55:191–196.
10. Williams AE, Edwards L, Hussell T. Colonic bacterial infection abrogates 31. Leigh R, Ellis R, Wattie J, Donaldson DD, Inman MD. Is interleukin-
eosinophilic pulmonary disease. J Infect Dis 2006;193:223–230. 13 critical in maintaining airway hyperresponsiveness in allergen-chal-
11. Hatakka K, Savilahti E, Ponka A, Meurman JH, Poussa T, Nase L, lenged mice? Am J Respir Crit Care Med 2004;170:851–856.
Saxelin M, Korpela R. Effect of long term consumption of probiotic 32. Ma D, Forsythe P, Bienenstock J. Live Lactobacillus reuteri is essential
milk on infections in children attending day care centres: double blind, for the inhibitory effect on tumor necrosis factor alpha-induced
randomized trial. BMJ 2001;322:1–5. interleukin-8 expression. Infect Immun 2004;72:5308–5314.
12. Hori T, Kiyoshima J, Shida K, Yasui H. Augmentation of cellular immu- 33. Zhang L, Li N, Caicedo R, Neu J. Alive and dead Lactobacillus rhamno-
nity and reduction of influenza virus titer in aged mice fed Lactobacillus sus GG decrease tumor necrosis factor-␣-induced interleukin-8 pro-
casei strain Shirota. Clin Diagn Lab Immunol 2002;9:105–108. duction in Caco-2 cells. J Nutr 2005;135:1752–1756.
13. Forsythe P, Wattie J, Inman M, Bienenstock J. Oral treatment with live 34. Hunt JR, Martinelli R, Adams VC, Rook GA, Brunet LR. Intragastric
lactobacillus reuteri attenuates the allergic airway response in mice. administration of mycobacterium vaccae inhibits severe pulmonary
J Allergy Clin Immunol 2006;117:S315. allergic inflammation in a mouse model. Clin Exp Allergy 2005;35:685–
14. Madsen KL, Doyle JS, Jewell LD, Tavernini MM, Fedorak RN. Lactoba- 690.
35. Rachmilewitz D, Katakura K, Karmeli F, Hayashi T, Reinus C, Rudensky
cillus species prevents colitis in interleukin 10 gene-deficient mice.
B, Akira S, Takeda K, Lee J, Takabayashi K, et al. Toll-like receptor
Gastroenterology 1999;116:1107–1114.
9 signaling mediates the anti-inflammatory effects of probiotics in
15. Sheil B, McCarthy J, O’Mahony L, Bennett MW, Ryan P, Fitzgibbon JJ,
murine experimental colitis. Gastroenterology 2004;126:520–528.
Kiely B, Collins JK, Shanahan F. Is the mucosal route of administration
36. Yuan W, Collado-Hidalgo A, Yufit T, Taylor M, Varga J. Modulation of
essential for probiotic function? Subcutaneous administration is associ-
cellular tryptophan metabolism in human fibroblasts by transforming
ated with attenuation of murine colitis and arthritis. Gut 2004;53:694–
growth factor-beta: selective inhibition of indoleamine 2,3-dioxygenase
700.
and tryptophanyl-tRNA synthetase gene expression. J Cell Physiol
16. Ebeling C, Forsythe P, Ng J, Gordon JR, Hollenberg M, Vliagoftis H. 1998;177:174–186.
Proteinase-activated receptor 2 activation in the airways enhances 37. Kudo Y, Boyd CA, Spyropoulou I, Redman CW, Takikawa O, Katsuki
antigen-mediated airway inflammation and airway hyperresponsive- T, Hara T, Ohama K, Sargent IL. Indoleamine 2,3-dioxygenase: distri-
ness through different pathways. J Allergy Clin Immunol 2005;115: bution and function in the developing human placenta. J Reprod Im-
623–630. munol 2004;61:87–98.
17. Leigh R, Southam DS, Ellis R, Wattie JN, Sehmi R, Wan Y, Inman MD. 38. Mellor AL, Munn DH. IDO expression by dendritic cells: tolerance and
T-cell-mediated inflammation does not contribute to the maintenance tryptophan catabolism. Nat Rev Immunol 2004;4:762–774.
of airway dysfunction in mice. J Appl Physiol 2004;97:2258–2265. 39. Grohmann U, Fallarino F, Puccetti P. Tolerance, DCs and tryptophan:
18. Takikawa O, Kuroiwa T, Yamazaki F, Kido R. Mechanism of interferon- much ado about IDO. Trends Immunol 2003;24:242–248.
gamma action: characterization of indoleamine 2,3-dioxygenase in cul- 40. Gurtner GJ, Newberry RD, Schloemann SR, McDonald KG, Stenson
tured human cells induced by interferon-gamma and evaluation of the WF. Inhibition of indoleamine 2,3-dioxygenase augments trinitroben-
enzyme-mediated tryptophan degradation in its anticellular activity. zene sulfonic acid colitis in mice. Gastroenterology 2003;125:1762–
J Biol Chem 1988;263:2041–2048. 1773.
19. Laich A, Neurauter G, Widner B, Fuchs D. More rapid method for 41. von Bubnoff D, Fimmers R, Bogdanow M, Matz H, Koch S, Bieber T.
simultaneous measurement of tryptophan and kynurenine by HPLC. Asymptomatic atopy is associated with increased indoleamine 2,3-
Clin Chem 2002;48:579–581. dioxygenase activity and interleukin-10 production during seasonal
20. Hayashi T, Beck L, Rossetto C, Gong X, Takikawa O, Takabayashi K, allergen exposure. Clin Exp Allergy 2004;34:1056–1063.
Broide DH, Carson DA, Raz E. Inhibition of experimental asthma 42. Wagner H. The immunobiology of the TLR9 subfamily. Trends Immunol
by indoleamine 2,3-dioxygenase. J Clin Invest 2004;114:270–279. 2004;25:381–386.
21. Reid G, Jass J, Sebulsky MT, McCormick JK. Potential uses of probiotics 43. Barton GM, Kagan JC, Medzhitov R. Intracellular localization of Toll-
in clinical practice. Clin Microbiol Rev 2003;16:658–672. like receptor 9 prevents recognition of self DNA but facilitates access
22. Moudgil KD, Kim E, Yun OJ, Chi HH, Brahn E, Sercarz EE. Environ- to viral DNA. Nat Immunol 2006;7:49–56.
mental modulation of autoimmune arthritis involves the spontaneous 44. Lund J, Sato A, Akira S, Medzhitov R, Iwasaki A. Toll-like receptor 9-
microbial induction of T cell responses to regulatory determinants mediated recognition of herpes simplex virus-2 by plasmacytoid den-
within heat shock protein 65. J Immunol 2001;166:4237–4243. dritic cells. J Exp Med 2003;198:513–520.
Forsythe, Inman, and Bien: Probiotics and Airway Inflammation 569

45. Macpherson AJ, Uhr T. Induction of protective IgA by intestinal den- 48. Hart AL, Lammers K,. Brigidi P, Vitali B, Rizzello F, Gionchetti P,
dritic cells carrying commensal bacteria. Science 2004;303:1662–1665. Campieri M, Kamm MA, Knight SC, Stagg AJ. Modulation of human
46. Groux H, Fournier N, Cottrez F. Role of dendritic cells in the generation dendritic cell phenotype and function by probiotic bacteria. Gut
of regulatory T cells. Semin Immunol 2004;16:99–106. 2004;53:1602–1609.
47. Baban B, Hansen AM, Chandler PR, Manlapat A, Bingaman A, Kahler 49. Braat H, van den Brande J, van Tol E, Hommes D, Peppelenbosch M,
DJ, Munn DH, Mellor AL. Minor population of splenic dendritic cells van Deventer S. Lactobacillus rhamnosus induces peripheral hypore-
expressing CD19 mediates IDO-dependent T cell suppression via type sponsiveness in stimulated CD4⫹ T cells via modulation of dendritic
I IFN signaling following B7 ligation. Int Immunol 2005;17:909–919. cell function. Am J Clin Nutr 2004;80:1618–1625.