Psychoactive bacteria Lactobacillus rhamnosus (JB-1) elicits rapid frequency facilitation in vagal afferents

Psychoactive bacteria Lactobacillus rhamnosus (JB-1) elicits rapid frequencyfacilitation in vagal afferents

Azucena Perez-Burgos,1 Bingxian Wang,1 Yu-Kang Mao,1 Bhavik Mistry,1 Karen-Anne McVey Neufeld,1

John Bienenstock,1,3 and Wolfgang Kunze1,2

1McMaster Brain-Body Institute, St. Joseph’s Healthcare, Hamilton, Ontario, Canada; 2Department of Psychiatry andBehavioural Neurosciences, McMaster University, Hamilton, Ontario, Canada; and, 3Department of Pathology and MolecularMedicine, McMaster University, Hamilton, Ontario, Canada

Submitted 27 March 2012; accepted in final form 31 October 2012

Perez-Burgos A, Wang B, Mao Y, Mistry B, Neufeld KM,Bienenstock J, Kunze W. Psychoactive bacteria Lactobacillus rham-nosus (JB-1) elicits rapid frequency facilitation in vagal afferents. AmJ Physiol Gastrointest Liver Physiol 304: G211–G220, 2013. Firstpublished November 8, 2012; doi:10.1152/ajpgi.00128.2012.—Mounting evidence supports the influence of the gut microbiome onthe local enteric nervous system and its effects on brain chemistry andrelevant behavior. Vagal afferents are involved in some of theseeffects. We previously showed that ingestion of the probiotic bacte-rium Lactobacillus rhamnosus (JB-1) caused extensive neurochemicalchanges in the brain and behavior that were abrogated by priorvagotomy. Because information can be transmitted to the brain viaprimary afferents encoded as neuronal spike trains, our goal was torecord those induced by JB-1 in vagal afferents in the mesentericnerve bundle and thus determine the nature of the signals sent to thebrain. Male Swiss Webster mice jejunal segments were cannulated exvivo, and serosal and luminal compartments were perfused separately.Bacteria were added intraluminally. We found no evidence for trans-location of labeled bacteria across the epithelium during the experi-ment. We recorded extracellular multi- and single-unit neuronal ac-tivity with glass suction pipettes. Within minutes of application, JB-1increased the constitutive single- and multiunit firing rate of themesenteric nerve bundle, but Lactobacillus salivarius (a negativecontrol) or media alone were ineffective. JB-1 significantly aug-mented multiunit discharge responses to an intraluminal distensionpressure of 31 hPa. Prior subdiaphragmatic vagotomy abolished all ofthe JB-1-evoked effects. This detailed exploration of the neuronalspike firing that encodes behavioral signaling to the brain may beuseful to identify effective psychoactive bacteria and thereby offer analternative new perspective in the field of psychiatry and comorbidconditions.

microbiome; probiotics; vagus; mesenteric nerve; gut-brain axis

IT IS KNOWN THAT THE ACQUISITION of a normal microbiome isnecessary for the development of a complete immunologicalrepertoire (8, 19), including that of the immune regulatorysystem (29). Even monoassociation of germ-free mice with, forexample, the segmented filamentous bacterium can restore theimmune system to normality, indicating the strain dependencyof these events (35, 52). Also, the presence of a normalmicrobiome may be constitutively anti-inflammatory and anti-nociceptive (31, 51, 54). However, very little is known aboutthe role of the gut microbiome in the development or functionof the nervous system.

It has been recently reported that nonpathogenic intestinalmicroorganisms can signal to the brain as part of the so-calledmicrobiome-gut-brain axis (9, 19, 22, 38). Changes in themicrobiome have been noted in the behavior of healthy non-infected animals, the most striking of which have been dem-onstrated in recent experiments showing that certain aspects ofbehavior in different strains of adult mice were dependent onthe content of their respective fecal microbiomes (1). Most ofthe information reaching the brain via primary sensory affer-ents, including the vagus fibers, has to be encoded in thelanguage of neuronal spike trains that represent the nature andintensity of the stimulus (20). Therefore, knowing how thesensory spike trains are affected by commensals or probioticsmight enable us to screen for and identify new potentiallybeneficial gut microorganisms, and eventually their activemolecules by their effects on primary afferent firing.

In terms of nociception, treatment of mice with nonabsorb-able antibiotics increased pseudoaffective responses to colo-rectal distension, and this could be prevented by ingestion ofLactobacillus paracasei (54). Similar responses in rats reflect-ing perception of visceral pain evoked by colorectal distensionwere almost completely inhibited by 9 days of gavage with L.rhamnosus (JB-1) (31). Spontaneous and distension-evokedaction potential firing recorded from single dorsal root fiberswas also decreased by prior ingestion of the same bacteria (31,37). In addition, JB-1 ingestion quenched long-term dorsal rootganglion neuron hyperexcitability elicited by repeated colorec-tal distension in healthy animals. Ingestion of JB-1 withinminutes of application to the epithelium (34, 55) facilitatesfiring of enteric intrinsic primary afferent neurons (34) andmoderates peristaltic reflexes (55). These observations havetaught us that changes in the gut microbiome can significantlyinfluence the activity of the enteric nervous system and spinalpathways.

Moreover, we recently reported that feeding JB-1 for 28days to conventionally housed mice decreased anxiety-relatedbehavioral scores and induced widespread changes in theexpression of brain �-aminobutyric acid (GABA) receptors (7).An important finding in this study was that the intact vagus wasessential since prior vagotomy abrogated these behavioral andcentral neurochemical effects.

It is not clear from current knowledge whether psychoactivebacteria are characterizable in terms of their potential vagalactivation or even other evidence of neuroactivity. For exam-ple, JB-1 decreases constitutive and distension-evoked firing indorsal root ganglia neurons (31), which is consistent with itsantinociceptive properties, yet its effects on myenteric intrinsic

Address for reprint requests and other correspondence: A. Perez-Burgos,The McMaster Brain-Body Institute, Juravinsky Innovation Tower, St. Jo-seph’s Healthcare, 50 Charlton Ave. East, Hamilton, ON, Canada L8N 4A6(e-mail: [email protected]).

Am J Physiol Gastrointest Liver Physiol 304: G211–G220, 2013.First published November 8, 2012; doi:10.1152/ajpgi.00128.2012.

0193-1857/13 Copyright © 2013 the American Physiological Societyhttp://www.ajpgi.org G211

mailto:[email protected]

primary afferent neurons is one of enhanced intrinsic excitabil-ity (34).

The present experiments were therefore designed to studyhow probiotic bacteria might affect the electrical discharges inthe mesenteric nerve bundle, which contains vagus afferentfibers. We have recorded from afferent fibers, since theyemerge from the intestine in an ex vivo intestinal jejunalsegment mesenteric nerve recording preparation (32). This hasallowed us to control bacterial concentrations in the lumen andmeasure spike firing onset latencies from the moment ofbacterial application to the epithelium (55). Our hypothesiswas that JB-1 might reduce the firing rate of vagal fibers, sincethis bacteria had a vagally dependent anxiolytic effect (7). Wehave shown that introduction of JB-1, and not another Lacto-bacillus strain, in the lumen of such a segment within minutescauses an increase in the constitutive firing rate, increases theneuronal response to gut distension, and that both of theseeffects are prevented by prior subdiaphragmatic vagotomy.

MATERIALS AND METHODS

All procedures adhered to Canadian Council on Animal Careguidelines and were approved by the Animal Research Ethics Boardof McMaster University (permit no. 08-08-35).

Extracellular Recordings

Adult male Swiss Webster mice (20–30 g) were procured fromCharles River Laboratories (Wilmington, MA). The mice were killedby cervical dislocation. All ensuing procedures were ex vivo.

Segments of excised distal jejunum (�2.5 cm) with attachedmesenteric tissue were removed from dead animals and immediatelyplaced in a Sylgard-coated recording petri dish filled with Krebsbuffer of the following composition (in mM): 118 NaCl, 4.8 KCl, 25NaHCO3, 1.0 NaH2PO4, 1.2 MgSO4, 11.1 glucose, and 2.5 CaCl2bubbled with carbogen (95% O2-5% CO2). The oral and anal portionsof gut were cannulated with plastic tubing and gently emptied usingan attached syringe filled with oxygenated Krebs buffer. The gut andmesenteric tissue were pinned out, the mesenteric nerve bundle wasisolated by careful dissection under a stereomicroscope, and thepreparation was transferred to an inverted microscope and then lumengravity perfused at 0.5–1 ml/min with room temperature (�22°C)oxygenated Krebs and/or additives using several Mariotte bottles (56).The serosal compartment was separately perfused with prewarmed(34°C) Krebs solution at 3–5 ml/min. The nerve bundle was gentlysucked into a glass pipette attached to a patch-clamp electrode holder(CV-7B; Molecular Devices, Sunnyvale, CA), and extracellular nerverecordings were made using a Multi-Clamp 700B amplifier andDigidata 1440A signal converter (Molecular Devices). Electrical sig-nals were bandpass-filtered at 0.1–2 kHz, sampled at 20 kHz, andstored on a personal computer running pClamp 10 software (Molec-ular Devices) for post hoc analysis. Constitutive multiunit electricalactivity was always recorded from the mesenteric nerve bundle, evenin the absence of active distension. Repeated distensions of thesegment were made by infusing Krebs in the lumen of jejunalsegments at a constant gravity pressure head of 14 or 31 hPa andclosing the outflow tube for 1 min each time (up to a maximum of 5times consecutively). Segments were allowed to rest with no imposeddistension for 9 min between repeated distensions. We chose severalsingle units in selected multiunit recordings and made templates withthem using the spike identification module in Clampfit. The templateswere then used to identify the individual single units by templatematching successive spikes for shape, duration, and size.

Vagotomy

For some experiments, a subdiaphragmatic vagotomy was carriedout as previously described (53). Animals were allowed to recover for10–14 days before harvesting the jejunum and mesenteric tissue forelectrophysiological experiments. Sham vagotomy was performed inthree animals. Postoperatively, the body weight and general health ofthe mice were measured daily (5, 26). We found no evidence ofsignificant differences in weight gain 1 wk postsurgery in eithervagotomized or sham-treated animals [data not shown (43)]. Allvagotomized mice were tested for completeness of the procedure byrecording after each experiment the responses to serosal application ofcholecystokinin (CCK). Vagotomy was deemed to have been effectivewhen CCK did not increase mesenteric nerve firing rate (25).

Integrity of Jejunal Segments

To test the integrity of the jejunal segments with respect totranslocation of bacteria across the epithelium during an experiment,we labeled JB-1 with 5-(6)-carboxyfluorescein succinimidyl ester(CFSE), placed these at a concentration of 109/ml Krebs in the lumen,and, after incubation for 75 min, fixed the tissue in paraformaldehydeand examined sections for their presence below the epithelium inconfocal microscopy (dual-laser microscopy, LSM 510; Carl Zeiss).

JB-1 were washed two times in PBS and suspended in a finalconcentration of CFSE of 5 �M in PBS supplemented with 5% fetalcalf serum and incubated for 25 min at 37°C. The tissues (n � 3) werefixed in 4% paraformaldehyde at 4°C overnight and then washed threetimes in PBS for 10 min each and sectioned at 10 and 30 �m;transverse sections were transferred to microscope slides andmounted. These were then reviewed in optical slices by the Z-stackingmethodology.

Probiotics and Drugs

L. rhamnosus (JB-1), previously referred to as L. reuteri andsubsequently identified by genomic analysis as an L. rhamnosus, weretaken from in-house stock (34, 37, 54); for more information, pleasesee the supplementary information in Bravo et al. (7). L. salivariuswas a gift from the Alimentary Pharmabiotic Centre (UniversityCollege Cork, Cork, Ireland) (44). Bacterial cell numbers were deter-mined optically, and viability was checked by the ability to grow afterplating on growth medium agar plates. All other methods were asreported previously (34, 37, 55). Briefly, live L. rhamnosus weregrown from frozen (�80°C) 1-ml aliquots that consisted of 5 � 109

cells in Man-Rogosa Sharpe broth (Difco Laboratories, Sparks, MD).Bacteria from frozen stocks were thawed and centrifuged at 2,000 rpmfor 15 min, and the pellet was suspended in an equal volume of Krebsbuffer. The suspension was again centrifuged, and the bacteria wasremoved and resuspended in Krebs at the original concentration. Justbefore use, the bacteria were diluted to working concentrations withKrebs buffer. For some experiments the bacteria were diluted directlyto working concentrations after thawing (in broth); the bacteria werealways applied in the lumen of the jejunal segment. Cholecystokinin(25–33) sulfated (CCK; AnaSpec, Fremont, CA) was dissolved indimethyl sulfoxide (DMSO) to make a 1 mM stock solution. Aliquotswere diluted on the day of the experiment to a working concentrationin Krebs buffer, with a final DMSO concentration �0.0001%.

Off-Line Data Analysis

Multi- and single-unit spontaneous firing frequencies of mesentericafferent nerve bundles were measured using Clampfit 10.2 (MolecularDevices) and Origin 8.5 (Northampton, MA) software. Both methods(multiunit spike discharge and waveform analysis) of measurementare routinely used to determine changes in the excitability of themesenteric nerve fibers induced by different treatments (6, 25, 27, 32,45, 49). The timing of the multiunit spikes was determined using thepeak detection module of Clampfit, and average frequency was

G212 LACTOBACILLUS RAPIDLY EXCITES VAGAL AFFERENT FIBERS

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00128.2012 • www.ajpgi.org

calculated from spike intervals. Single-unit activity was isolated fromthe multiunit activity using the spike shape automatic template detec-tion tool of Clampfit (computerized waveform analysis). After tem-plate detection, discrimination was always checked by visual inspec-tion, and nonmatching spike events were discarded (�0.2%).

Statistics

Data are expressed as means � SE with n referring to the totalnumber of the jejunal segments recorded; the maximum number of

segments recorded from the same animal was two. The Wilcoxon testwas used for paired data comparisons, and Friedman test with Dunn’spost hoc test for repeated measures ANOVA were performed usingPrism software 5.0 (GraphPad Software, San Diego, CA). Becauselarge variations in spontaneous activity may occur between onepreparation and another in multiunit neural activity (49), all compar-isons were paired with before and after treatment recordings madewhere each nerve bundle served as its own control. Differences wereconsidered significant if P � 0.05.

Fig. 1. Extracellular multiunit activity from mesenteric nerve bundle is stable during experiments. A: firing frequency during 60 min does not vary significantlyin the absence of stimuli. Distension stimulus increases firing frequency only transiently. Each point illustrates the average of firing frequency � SE of each 0.6min of one nerve bundle recording. B: firing frequency of first and last 5-min recording not exposed to drugs or bacteriological treatment do not vary significantly(n � 7, P � 0.46, Wilcoxon test). C: average rundown in the same group as B is 3.3 � 4%. D: gut distension with 32 cmH2O (31 hPa) applied with a gravitysystem during �1 min (each 10 min) induces a remarkable increase in basal firing frequency. After the first distension, subsequent distensions induce a smallerincrease in firing frequency that remains constant for up to three further distensions (n � 4, P � 0.67, Friedman test with Dunn’s post hoc). The second distensionwas used as a control in further experiments. E: labeled Lactobacillus rhamnosus (JB-1) with 5-(6)-carboxyfluorescein succinimidyl ester (CFSE) placed in thelumen at 109/ml Krebs during 75 min. The section reveals fluorescent bacteria in the lumen but not in the epithelial layer or below in the mucosa or deeper layers(n � 3).

G213LACTOBACILLUS RAPIDLY EXCITES VAGAL AFFERENT FIBERS


RESULTS

Stability of the Recordings and Integrity of the Tissue

We always recorded spontaneous neural activity for at least5–10 min before any stimuli were applied (gut distension,bacteria, etc.). Application of 1 �M of the sodium channelblocker tetrodotoxin always eliminated this neural activity(data not shown). Complete experiments lasted �1 h, andduring this time the preparation remained stable and healthy(see Fig. 1, A–C). The mucosa of distal jejunum can remainhealthy up to �4 h in similar experiments involving Ussingchambers (28). Effects of bacteria began to be evident after10–15 min of luminal application.

Distensions consistently increased the average multi- andsingle-unit firing frequency (Fig. 1A); the response to the firstdistension was the largest with subsequent (up to 4 tried)responses being of equal intensity (Fig. 1D). Because of this, infurther experiments bacteria were applied after the seconddistension, and we made up to three further measures of the gutresponse to distension.

We failed to find evidence of translocation of bacteria acrossthe epithelium. Examination of the slides in confocal micros-copy from several sections in each of three jejunal segmentstaken over the time course of the experiments revealed num-bers of labeled bacteria in the lumen, but no fluorescentbacteria were seen in the epithelial layer or below it in themucosa, submucosa, or myenteric plexus (Fig. 1E).

Effects of JB-1 on the Spontaneous Firing Frequency ofMesenteric Afferent Nerve Bundles

Application of 5 � 107 colony-forming units (cfu)/ml ofJB-1 (in broth) had no effect on basal multiunit firing fre-quency (before and after JB-1: 20.7 � 2.4 vs. 19.7 � 2.1 Hz)(n � 10, P � 0.15, Fig. 2A). JB-1 at 1 � 108 cfu/ml also hadno apparent effect on the nerve firing frequency (data notshown). When the JB-1 concentration was increased to 1 � 109

cfu/ml, we observed a transient increase in the basal firingfrequency in both conditions (with Krebs and broth; no signif-icant differences were found between these two groups) [from14.2 � 2.5 to 19.6 � 3.3 Hz, n � 17/24 (70.8% of cases), P �

Fig. 2. L. rhamnosus (JB-1) increases the spontaneous firing frequency in the gross fiber population of the mesenteric afferent nerve bundle. A: graph showsspontaneous firing frequency before and after 5 � 107 JB-1/ml (in broth; P � 0.15, n � 10). B: JB-1 [1 � 109 colony-forming units (cfu)/ml; ***P � 0.0002,n � 17/24 (70.8%), Wilcoxon test]. C: effects of broth alone (P � 0.21, n � 12). D: 1 � 109 L. salivarius/ml (in Krebs, P � 0.36, n � 11). E: 1 � 109 JB-1/ml(in Krebs) after treatment with 3 �M of nicardipine [L-type Ca2 channels blocker; **P � 0.002, n � 9/13 (69.2%), Wilcoxon test]. F: representative tracesof multiunit extracellular recordings in control conditions (Krebs in lumen; left). Right, after addition of 1 � 109 JB-1/ml. Dotted rectangles show details withfaster time base. Scales are the same for both traces.



0.0002; pooled results, Fig. 2, B and F]. These effects wereevoked within 10–15 min of the onset of intraluminal appli-cation, even when the bacteria were previously washed anddiluted in Krebs solution, and we used this effective concentra-tion (1 � 109 cfu/ml) in all subsequent experiments. Table 1summarizes the degree of JB-1 actions (% of increase) onconstitutive firing frequency and on the response to distensionin all experimental groups. We saw no time-dependent run-down in mesenteric nerve discharge, and no differences whenKrebs was substituted for diluted broth (25.1 � 3.8 vs. 23.9 �3.4 Hz) (n � 12, P � 0.21, Fig. 2C). Because we previouslyshowed that L. salivarius had no effect on gut contractility(55), we tested it in our system. This strain of Lactobacillushad no effect at 1 � 109 cells/ml on the basal firing frequency(11.1 � 2.5 vs. 9.9 � 2.0 Hz; n � 11, P � 0.36, Fig. 2D) sowe used it as a further negative control.

To test if the JB-1 effect on the multiunit activity wasmediated via muscle contraction, the smooth muscle wasparalyzed using the L-type calcium channel blocker nicardi-pine (33). When 3 �M of nicardipine was added to the Krebsin the perfusion system, before applying intraluminal JB-1, thebacteria still increased the multiunit firing frequency [15.1 � 7vs. 19.3 � 9 Hz, n � 9/13 (69.2%), P � 0.002, Fig. 2E].

The Firing Frequency Increase Induced by JB-1 Could beDetected Using Single-Unit Recording

The extracellular multiunit recording data were derived fromthe average discharge of multiple single fibers, and suchaveraging may have obscured specific treatment effects. Be-cause the mesenteric nerve bundle contains both spinal andvagal fibers (Fig. 3B) (4), it is possible that subpopulations offibers could respond differently to different treatments (25). Tocontrol for this we measured the effects of applying intralumi-nal bacteria on single-unit firing frequency (49). Therefore, wetested the effects of JB-1 on shape-identified (25, 49) single-unit activity on 1-min time segments taken before and at 10–15min after adding bacteria (see MATERIALS AND METHODS and Fig.

3A) (49). We compared the effects on single-unit firing fre-quency with results from several responsive multiunit data.Seventy percent of single units showed a clear increase in thespontaneous firing frequency when JB-1 (1 � 109 cfu/ml) wasapplied. When we averaged the results for all the analyzedsingle units, the increase was from 2 � 0.4 to 2.8 � 0.5 Hz(n � 17, P � 0.001), which is similar to the percentage ofincrease found in the multiunit analysis with this JB-1 concen-tration. Next, we included for analysis only fibers that had anincrease in the firing frequency [increase from 2.3 � 0.6 to3.3 � 0.7 Hz, n � 12/17 (70%), P � 0.001, Fig. 4, A and D]to show that the increase in multiunit firing frequency was notprimarily due to an increase in the total number of active fibersbut to an increase in the firing rate of individual previouslyactive primary afferent fibers. Broth alone did not have aneffect on firing frequency of extracted single fiber activity(1.9 � 0.2 vs. 1.8 � 0.2 Hz, n � 16, P � 0.07, Fig. 4B). With3 �M of nicardipine previously added to the Krebs, JB-1 stillincreased the single-unit firing frequency (1.1 � 0.1 vs. 1.4 �0.1 Hz, n � 18, P � 0.006); we again wanted to include in theanalysis only the fibers that increased the firing frequency withJB-1; in this case the effect was from 1 � 0.2 to 1.5 � 0.2 Hz[n � 12/18 (66.6%), P � 0.001, Fig. 4C].

Effect of JB-1 on Distension-Evoked Increases in Firing Frequency

Distensions consistently increased average multiunit andsingle-unit firing frequencies (see MATERIALS AND METHODS,

Fig. 3. Single-unit activity extracted from multiunit recordings. A: top, exam-ple of extracellular multiunit recording segment that contains the activity of 3distinguishable single fibers comparable in shape, size, and duration. Individ-ual action potentials of each single fiber are marked on top of the voltage tracewith different grayscale tone and detected according to fit to a template fromthe same recording (Clampfit program analysis). Bottom, superimposed re-cordings from multiple events from examples of each single fiber. B: illustra-tion of how the mesenteric nerve bundle is constituted by a mixed populationof single vagal and spinal fibers that innervates the gut wall.

Table 1. Degree of JB-1 actions on constitutive firingfrequency and on the response to distension in allexperimental groups

Multiunit constitutive firing frequency

5 � 107 2 � 4% (n � 10 of 10)1 � 109 50 � 12% (n � 17 of 24†)1 � 109 3 �M nicardipine 26 � 5% (n � 9 of 13†)1 � 109 (vagotomized) 2 � 6% (n � 11 of 11)1 � 109 (sham-treated) 51 � 10% (n � 5 of 5†)

Single-unit constitutive firing frequency

1 � 109 59 � 13% (n � 12 of 17†)1 � 109 3 �M nicardipine 64 � 13% (n � 12 of 18†)

Multiunit response to distension*

1 � 109 108 � 51% (n � 10 of 14†)1 � 109 (vagotomized) 11 � 13% (n � 5 of 5)

Single-unit response to distension*

1 � 109 118 � 29% (n � 11 of 17†)

Values are means � SE; n, no. of jejunal segments recorded. JB-1,Lactobacillus rhamnosus; cfu, colony-forming units. *Degree to which JB-1increased distension-evoked firing (%). †Only responsive fibers were takeninto account.



Figs. 1A and 5A). The percentage firing frequency increaseabove resting spontaneous discharge during the second disten-sion was 352 � 115 vs. 305 � 82% on the third distension and345 � 112% on the fourth distension (the lumen containedonly Krebs in these experiments) (Fig. 1D).

In another set of experiments, we examined if JB-1 couldmodify the frequency increase evoked by distension. Disten-sion in the presence of JB-1 at 1 � 109 cfu/ml (added after the2nd distension) induced a robust increment compared with theresponse evoked previously by gut distension when only Krebswas present [244.4 � 34% on the 2nd distension vs. 406.8 �62% on the 4th distension; n � 10/14 (71.4%), P � 0.001, Fig.5B]. Also, JB-1 at 1 � 108 cfu/ml or broth by itself (data notshown) did not have an effect on the nerve response to gutdistension. To test if this bacteria increases the response todistension in all the population of responsive mesenteric nervefibers or in a subset of them, we carried out single-unitextractions before and during distension from the experimentswith and without (Krebs in the lumen) the effective dose ofJB-1; we have found that 64.7% of the single fibers showed anincrease in the response to distension; this percentage is similarto the percentage of single fibers that increased their constitu-tive firing frequency (see Fig. 2A) [314.9 � 61% on the 2nddistension vs. 577.9 � 77% on the 4th distension; n � 11/17(64.7%), P � 0.001, Fig. 5, C and D].

Vagotomy Eliminated the JB-1 Effects on Spontaneous NerveActivity and on the Response to Gut Distension

Mesenteric nerve response to CCK. The mesenteric nervecontains vagal and spinal afferent fibers. Because CCK strongly

stimulates vagal but not spinal afferents, it can be used as a control totest that vagal afferent fibers are functionally intact and to establishthat surgical vagotomy quenched all afferent vagal traffic (25, 48).Hence, we tested the effects of adding CCK to the serosal perfusate.

CCK increased spontaneous multiunit activity, had a shortonset latency (�30 s), and was reversible and dose-dependent.The response to a range of CCK doses (1 pM-1 nM) wasexamined in individual jejunal segments (n � 15). The lowestdose that produced a just discernible effect was 1 pM, whereas100 pM increased the basal firing frequency from 8.8 � 3 to 15 �5 Hz (83.4 � 14% of increase; n � 5, P � 0.03, Fig. 6, A and B,top). The percentage of increase in firing frequency vs. CCKconcentration plot was fitted by a Hill equation of the form Y �bottom (top � bottom)/[1 10(LogEC50 � X)]. EC50 for CCKfiring frequency increase was 11 pM (Fig. 6A). We also tested ifJB-1 is able to induce an increase in the actions of a low dose ofCCK (10 pM) on intact mesenteric nerve bundles; there were nosignificant differences in the percentage of firing frequency in-crease induced by CCK before and after the administration ofJB-1 in the luminal jejunum (data not shown).

After vagotomy, 0/13 of individual jejunal segments testedresponded with an increase in multiunit firing when 100 pMCCK was applied; firing frequencies were 28.4 � 5 vs. 28.1 �5 Hz (n � 13, P � 0.1, Fig. 6, A and B, bottom). We concludethat the vagotomies were functionally effective.

JB-1 Effects on Neuronal Activity were Abolished After Vagotomy

We added 11 additional animals to our study to include theeffects of JB-1 on the mesenteric nerve bundle after vagotomy.Six animals were vagotomized, and five were sham-treated.

Fig. 4. Spontaneous firing frequency increase induced by JB-1 can be detected in single-unit activity. A: 1 � 109 JB-1 effects on basal single-fiber firing frequency[**P � 0.001, n � 12/17 (70.5%), Wilcoxon test]. B: effects of broth alone (P � 0.07, n � 16). C: 1 � 109 JB-1/ml (in Krebs) after treatment with 3 �M ofnicardipine [**P � 0.001, n � 12/18 (66.6%), Wilcoxon test]. D: representative traces of multiunit recordings: Krebs in lumen (left) and after addition of 1 �109 JB-1/ml (right); enlarged single-unit waveforms extracted from the traces shown above (bottom).



We tested the JB-1 effects on the spontaneous firing fre-quency and on the response induced by gut distension afterperforming vagotomies. No effect on the spontaneous firingfrequency of JB-1 was seen (28 � 7 vs. 29.9 � 7 Hz; n � 11,P � 0.06, Fig. 7, A and B). After vagotomy, JB-1 also failed toaugment the response to gut distension [percentage of firingfrequency increase: 308.2 � 93 and 321.6 � 75% before andafter the addition of JB-1, respectively (n � 5, P � 0.4, Fig.7C)]. The basal response to gut distension induced by 14 or 31hPa did not differ from that seen in intact animals. Weconclude that the excitatory actions of JB-1 on the mesentericafferents were carried by vagal fibers. To demonstrate thatspinal afferents were still intact after vagotomy we alwaysconfirmed that adding 1 �M capsaicin to the serosal compart-

ment evoked a strong excitatory multiunit response (data notshown) (39). JB-1 effects on the mesenteric nerve bundles withsham-treated animals were similar to those found in intactanimals; here, the effective dose of this bacteria increased thespontaneous firing frequency from 12.6 � 3.3 to 18.7 � 5 Hz(n � 5, P � 0.03, Fig. 7D).

DISCUSSION

The intestinal microbiome influences gut-brain communica-tion (1, 7, 9, 14, 19, 38, 47). This is further supported byobservations that the enteric nervous system and brain differbetween germ-free and conventional animals (1, 15, 25, 40,41). However, the underlying mechanisms are unknown, and

Fig. 5. JB-1 augmented the increase of nerve bundle firing frequency induced by gut distension. A: multiunit activity; increased firing frequency during gutdistension (for 1 min each 10 min) at 31 hPa or 32 cmH2O. Dotted rectangles show details with faster time base before (left) and after (right) gut distension.B: firing frequency increase induced by gut distension before and after addition of 1 � 109 JB-1/ml [**P � 0.001, n � 10/14 (71.4%), Wilcoxon test].C: single-unit waveforms exemplify the firing frequency increase (%) of single fibers before and after the addition of 1 � 109 JB-1/ml. D: graph showing theresults exemplified in C [**P � 0.001, n � 11/17 (64.7%), Wilcoxon test].

Fig. 6. Increase in basal firing frequency inducedby cholecystokinin (CCK) is lost after vagot-omy. A: dose-response plot from 15 experimentswas fitted with a logistic equation, EC50 � 11pM (intact animals). Œ, CCK effects on intactanimals; �, CCK effects after vagotomy. B: ef-fect of 100 pM CCK in intact (P � 0.03, n � 5;top) or vagotomized (P � 0.1, n � 13; bottom)animals.



most studies focus on the impact of altered signaling on thebrain (9, 13, 19, 38) and not on the intervening pathways fromgut luminal microbiome to the brain. Our data are incompatiblewith the hypothesis we initially offered in the Introduction.Indeed, we present the first evidence that in a normal mousejejunum a single probiotic (L. rhamnosus JB-1) provokes anincrease of the constitutive mesenteric nerve multi- and single-unit firing rate within minutes of luminal application. In addi-tion, JB-1 further augmented distension-evoked increases inmulti- and single-unit firing of mesenteric vagal afferents.These effects were due to single fibers that increase their firingfrequency rather than an increase in the number of active fibersrecruited.

The vagus nerve is an important afferent signaling pathwaybetween the gastrointestinal tract and the brain (24, 47). Thispathway is bidirectional, and clear evidence for a vagal efferentanti-inflammatory pathway exists (50). There are several linesof evidence that suggest that the vagus is involved in bothmood regulation and gut inflammation (13, 23). Because low-grade, even unrecognizable gut inflammation, induced bypathogenic bacteria appears to promote anxiety-like behavior(36), the introduction of bacteria such as probiotics, many ofwhich themselves induce anti-inflammatory circuits (18, 29),cannot be assumed to be acting on the brain independentlyfrom their effects on inflammation. There is also evidence thatchanges in the gut microbiota and/or ingestion of probioticscan experimentally influence behavior in animals that haveundergone some experimental manipulations (1, 3), but theseoccurred separately from influences of the autonomic nervoussystem. The anxiolytic effect of B. longum also inhibited the

generation of anxiety-like behavior in a chemical colitis model,and this effect, however, was inhibited by prior vagotomy (2).

We have recently published that chronic feeding of conven-tionally housed healthy mice with L. rhamnosus (JB-1) in-duced anxiolytic changes in behavior and in the level ofGABAA2, GABAA1, and GABAB1b receptor mRNA expres-sion in several brain structures (hippocampus, prefrontal cor-tex, cingulate cortex, amygdala, and locus coeruleus). Theseeffects were abrogated after vagotomy (7). These data establishthe importance in this model of the intact afferent vagus for thebrain GABA receptor mRNA changes and related anxiolyticeffects (12), but they provide no information about how affer-ent signals are altered by the probiotic. Our present data showthe nature of the change that one anxiolytic probiotic bacteriaproduces in afferent vagal firing. This contrasts with the effectsof another Lactobacillus, L. salivarius, which is not neuroac-tive in the gut (55) and also did not alter vagal firing. Theevidence suggests that neuroactive and psychoactive effects arebacterial strain specific (16). Also, we have found that if weapply JB-1 when the muscle is paralyzed by nicardipine, anL-type calcium channel blocker, the degree of effects on thespontaneous firing frequency is still present although lowerthan with JB-1 administered alone. Even so, we cannot discardthe possibility that JB-1 may modify the constitutive firingfrequency partially through changes in the peristaltic reflexesthat we have previously reported (55). However, because thedegree of changes in the constitutive firing frequency inducedby JB-1 on single fibers is very similar with and withoutnicardipine, we conclude that our results with multiunit record-

Fig. 7. JB-1 effects on firing frequency lost after vagotomy. A: representative traces of multiunit spontaneous firing frequency before (left) and after (right)addition of 1 � 109 JB-1. Dotted rectangles show details with faster time base. B: graph showing the results exemplified in A (P � 0.06, n � 11). C: firingfrequency increase induced by gut distension before and after the addition of 1 � 109 JB-1/ml (P � 0.4, n � 5). D: effects of 1 � 109 JB-1/ml on spontaneousfiring frequency of sham-treated mice (*P � 0.03, n � 5/5, Wilcoxon test).



ings were probably obscured by the number of nonresponsivesingle fibers.

Our experiments could not directly address the underlyingmechanisms of cellular transduction and signaling betweenJB-1 and the vagal primary afferents within the mucosa.However, commensals produce a large range of molecules,including fatty acids, cell wall oligosaccharides, and neu-rotransmitters (17, 30, 42). Some of these may interact withmucosal epithelial cells that release paracrine mediators to acton neuronal processes (38); certain molecules such as hydro-nium ions could also act directly on vagal afferents. A cogentapproach to the study of which bacterial molecule(s) mediateneural effects on the host awaits genomic and metabolomicanalysis of the probiotic (10). JB-1 did not facilitate the vagalresponse to CCK, suggesting that these bacteria probably didnot act via the CCK pathway. It is not clear which of theseeffects might explain the neuronal effects with onset latenciesof 10–15 min that we report here.

It remains to be elucidated exactly how the increase in theconstitutive vagal firing frequency and the augmentation of itsresponse to gut distension may change brain neurochemistryand behavior. However, electrical stimulation of vagal fibersalters concentrations of serotonin, norepinephrine, GABA, andglutamate within the brain (46), and, although controversial,vagal stimulation is a Food and Drug Administration (FDA)-approved treatment for patients with depression who do notrespond to classical drug therapy (11, 46) and is also FDAapproved for the adjunctive treatment of epilepsy (21).

In summary, we have shown that, within minutes of intro-duction of probiotic bacteria in the jejunal lumen of mice,vagal afferents were activated and that this activation wasrecorded as an increase in the spontaneous frequency of bothmulti- and single-unit firing frequency. This procedure furtheraugmented the distension-evoked increases of firing frequencythat in turn were all abrogated by prior subdiaphragmaticvagotomy. Thus, the mechanisms whereby some probioticbacteria influence brain and behavior may critically involve thefacilitation of vagal firing. It is therefore plausible that certainbacteria might be useful and safe adjuncts to therapy in someconditions. Our results support the fact that use of this pathwayis bacterial strain dependent as has been also shown forbacterial effects on behavior. Further detailed exploration ofthe neuronal spike trains that encode behavioral signaling tothe brain may be useful to identify those psychoactive bacteriathat are effective. It may also help uncover how they signal thebrain and thus offer novel approaches to the development ofnew forms of treatment for some psychiatric and other comor-bid disorders.

GRANTS

We authors gratefully acknowledge financial support for the conduct of thisresearch to the National Science and Engineering Research Council (no.RPGIN371955-09 to W. Kunze) and the Giovanni and Concetta GugliettiFamily Foundation. A. Perez-Burgos was supported by the Consejo Nacionalde Ciencia y Tecnología postdoctoral fellowship and the Brain-Body Institute.

DISCLOSURES

The funders had no role in study design, data collection and analysis,decision to publish, or preparation of the manuscript.

AUTHOR CONTRIBUTIONS

Author contributions: A.P.-B., Y.-K.M., B.M., and K.-A.M.N. performedexperiments; A.P.-B. and W.A.K. analyzed data; A.P.-B. and W.A.K. inter-

preted results of experiments; A.P.-B. prepared figures; A.P.-B. drafted man-uscript; A.P.-B., J.B., and W.A.K. edited and revised manuscript; A.P.-B.,B.W., Y.-K.M., B.M., K.-A.M.N., J.B., and W.A.K. approved final version ofmanuscript; B.W., J.B., and W.A.K. conception and design of research.

REFERENCES

1. Bercik P, Denou E, Collins J, Jackson W, Lu J, Jury J, Deng Y,Blennerhassett P, Macri J, McCoy KD, Verdu EF, Collins SM. Theintestinal microbiota affect central levels of brain-derived neurotropicfactor and behavior in mice. Gastroenterology 141: 599–609, 2011.

2. Bercik P, Park AJ, Sinclair D, Khoshdel A, Lu J, Huang X, Deng Y,Blennerhassett PA, Fahnestock M, Moine D, Berger B, Huizinga JD,Kunze W, McLean PG, Bergonzelli GE, Collins SM, Verdu EF. Theanxiolytic effect of Bifidobacterium longum NCC3001 involves vagalpathways for gut-brain communication. Neurogastroenterol Motil 23:1132–1139, 2011.

3. Bercik P, Verdu EF, Foster JA, Macri J, Potter M, Huang X,Malinowski P, Jackson W, Blennerhassett P, Neufeld KA, Lu J, KhanWI, Corthesy-Theulaz I, Cherbut C, Bergonzelli GE, Collins SM.Chronic gastrointestinal inflammation induces anxiety-like behavior andalters central nervous system biochemistry in mice. Gastroenterology 139:2102–2112 e2101, 2010.

4. Berthoud HR, Blackshaw LA, Brookes SJ, Grundy D. Neuroanatomyof extrinsic afferents supplying the gastrointestinal tract. Neurogastroen-terol Motil 16, Suppl 1: 28–33, 2004.

5. Bluthe RM, Michaud B, Kelley KW, Dantzer R. Vagotomy attenuatesbehavioural effects of interleukin-1 injected peripherally but not centrally.Neuroreport 7: 1485–1488, 1996.

6. Booth CE, Kirkup AJ, Hicks GA, Humphrey PP, Grundy D. Soma-tostatin sst(2) receptor-mediated inhibition of mesenteric afferent nervesof the jejunum in the anesthetized rat. Gastroenterology 121: 358–369,2001.

7. Bravo JA, Forsythe P, Chew MV, Escaravage E, Savignac HM, DinanTG, Bienenstock J, Cryan JF. Ingestion of Lactobacillus strain regulatesemotional behavior and central GABA receptor expression in a mouse viathe vagus nerve. Proc Natl Acad Sci USA 108: 16050–16055, 2011.

8. Cebra JJ. Influences of microbiota on intestinal immune system devel-opment. Am J Clin Nutr 69: 1046S–1051S, 1999.

9. Collins SM, Bercik P. The relationship between intestinal microbiota andthe central nervous system in normal gastrointestinal function and disease.Gastroenterology 136: 2003–2014, 2009.

10. Collins SM, Surette M, Bercik P. The interplay between the intestinalmicrobiota and the brain. Nat Rev Microbiol 10: 735–742, 2012.

11. Cristancho P, Cristancho MA, Baltuch GH, Thase ME, O’ReardonJP. Effectiveness and safety of vagus nerve stimulation for severe treat-ment-resistant major depression in clinical practice after FDA approval:outcomes at 1 year. J Clin Psychiatry 72: 1376–1382, 2011.

12. Cryan JF, Kaupmann K. Don’t worry ‘B’ happy!: a role for GABA(B)receptors in anxiety and depression. Trends Pharmacol Sci 26: 36–43,2005.

13. Cryan JF, O’Mahony SM. The microbiome-gut-brain axis: from bowelto behavior. Neurogastroenterol Motil 23: 187–192, 2011.

14. Desbonnet L, Garrett L, Clarke G, Kiely B, Cryan JF, Dinan TG.Effects of the probiotic Bifidobacterium infantis in the maternal separationmodel of depression. Neuroscience 170: 1179–1188, 2010.

15. Diaz Heijtz R, Wang S, Anuar F, Qian Y, Bjorkholm B, SamuelssonA, Hibberd ML, Forssberg H, Pettersson S. Normal gut microbiotamodulates brain development and behavior. Proc Natl Acad Sci USA 108:3047–3052, 2011.

16. Duncker SC, Kamiya T, Wang L, Yang P, Bienenstock J. ProbioticLactobacillus reuteri alleviates the response to gastric distension in rats. JNutr 141: 1813–1818, 2011.

17. Fanning S, Hall LJ, Cronin M, Zomer A, MacSharry J, Goulding D,Motherway MO, Shanahan F, Nally K, Dougan G, van Sinderen D.Bifidobacterial surface-exopolysaccharide facilitates commensal-host in-teraction through immune modulation and pathogen protection. Proc NatlAcad Sci USA 109: 2108–2113, 2012.

18. Forsythe P, Bienenstock J. Immunomodulation by commensal and pro-biotic bacteria. Immunol Invest 39: 429–448, 2010.

19. Forsythe P, Sudo N, Dinan T, Taylor VH, Bienenstock J. Mood and gutfeelings. Brain Behav Immun 24: 9–16, 2010.

20. Furness JB, Kunze WA, Clerc N. Nutrient tasting and signaling mech-anisms in the gut. II The intestine as a sensory organ: neural, endocrine,



and immune responses. Am J Physiol Gastrointest Liver Physiol 277:G922–G928, 1999.

21. George MS, Nahas Z, Borckardt JJ, Anderson B, Burns C, Kose S,Short EB. Vagus nerve stimulation for the treatment of depression andother neuropsychiatric disorders. Exp Rev Neurotherapeut 7: 63–74, 2007.

22. Goehler LE, Gaykema RP, Opitz N, Reddaway R, Badr N, Lyte M.Activation in vagal afferents and central autonomic pathways: earlyresponses to intestinal infection with Campylobacter jejuni. Brain BehavImmun 19: 334–344, 2005.

23. Goehler LE, Lyte M, Gaykema RP. Infection-induced viscerosensorysignals from the gut enhance anxiety: implications for psychoneuroimmu-nology. Brain Behav Immun 21: 721–726, 2007.

24. Grundy D. What activates visceral afferents? Gut 53, Suppl 2: ii5–ii8,2004.

25. Hillsley K, Grundy D. Serotonin and cholecystokinin activate differentpopulations of rat mesenteric vagal afferents. Neurosci Lett 255: 63–66,1998.

26. Hosoi T, Okuma Y, Ono A, Nomura Y. Subdiaphragmatic vagotomyfails to inhibit intravenous leptin-induced IL-1beta expression in thehypothalamus. Am J Physiol Regul Integr Comp Physiol 282: R627–R631,2002.

27. Ibeakanma C, Ochoa-Cortes F, Miranda-Morales M, McDonald T,Spreadbury I, Cenac N, Cattaruzza F, Hurlbut D, Vanner S, BunnettN, Vergnolle N, Vanner S. Brain-gut interactions increase peripheralnociceptive signaling in mice with postinfectious irritable bowel syn-drome. Gastroenterology 141: 2098–2108, 2011.

28. Inagaki E, Natori Y, Ohgishi Y, Hayashi H, Suzuki Y. Segmentaldifference of mucosal damage along the length of a mouse small intestinein an Ussing chamber. J Nutr Sci Vitaminol 51: 406–412, 2005.

29. Ivanov II, Littman DR. Modulation of immune homeostasis by commen-sal bacteria. Curr Opin Microbiol 14: 106–114, 2011.

30. Iyer LM, Aravind L, Coon SL, Klein DC, Koonin EV. Evolution ofcell-cell signaling in animals: did late horizontal gene transfer frombacteria have a role? Trends Genetics 20: 292–299, 2004.

31. Kamiya T, Wang L, Forsythe P, Goettsche G, Mao Y, Wang Y,Tougas G, Bienenstock J. Inhibitory effects of Lactobacillus reuteri onvisceral pain induced by colorectal distension in Sprague-Dawley rats. Gut55: 191–196, 2006.

32. Keating C, Beyak M, Foley S, Singh G, Marsden C, Spiller R, GrundyD. Afferent hypersensitivity in a mouse model of post-inflammatory gutdysfunction: role of altered serotonin metabolism. J Physiol 586: 4517–4530, 2008.

33. Kunze WA, Furness JB, Bertrand PP, Bornstein JC. Intracellularrecording from myenteric neurons of the guinea-pig ileum that respond tostretch. J Physiol 506: 827–842, 1998.

34. Kunze WA, Mao YK, Wang B, Huizinga JD, Ma X, Forsythe P,Bienenstock J. Lactobacillus reuteri enhances excitability of colonic AHneurons by inhibiting calcium-dependent potassium channel opening. JCell Mol Med 13: 2261–2270, 2009.

35. Lyons A, O’Mahony D, O’Brien F, MacSharry J, Sheil B, Ceddia M,Russell WM, Forsythe P, Bienenstock J, Kiely B, Shanahan F,O’Mahony L. Bacterial strain-specific induction of Foxp3 T regulatorycells is protective in murine allergy models. Clin Exp Allergy 40: 811–819,2010.

36. Lyte M, Li W, Opitz N, Gaykema RP, Goehler LE. Induction ofanxiety-like behavior in mice during the initial stages of infection with theagent of murine colonic hyperplasia Citrobacter rodentium. Physiol Behav89: 350–357, 2006.

37. Ma X, Mao YK, Wang B, Huizinga JD, Bienenstock J, Kunze W.Lactobacillus reuteri ingestion prevents hyperexcitability of colonic DRGneurons induced by noxious stimuli. Am J Physiol Gastrointest LiverPhysiol 296: G868–G875, 2009.

38. Mayer EA. Gut feelings: the emerging biology of gut-brain communica-tion. Nat Rev Neurosci 12: 453–466, 2011.

39. Miranda-Morales M, Ochoa-Cortes F, Stern E, Lomax AE, Vanner S.Axon reflexes evoked by transient receptor potential vanilloid 1 activationare mediated by tetrodotoxin-resistant voltage-gated Na channels inintestinal afferent nerves. J Pharmacol Exp Ther 334: 566–575, 2010.

40. Neufeld KA, Foster JA. Effects of gut microbiota on the brain: implica-tions for psychiatry. J Psychiatry Neurosci 34: 230–231, 2009.

41. Neufeld KM, Kang N, Bienenstock J, Foster JA. Reduced anxiety-likebehavior and central neurochemical change in germ-free mice. Neurogas-troenterol Motil 23: 255–264, e119, 2011.

42. O’Hara AM, Shanahan F. Mechanisms of action of probiotics in intes-tinal diseases. Sci World J 7: 31–46, 2007.

43. O’Mahony C, van der Kleij H, Bienenstock J, Shanahan F, O’MahonyL. Loss of vagal anti-inflammatory effect: in vivo visualization andadoptive transfer. Am J Physiol Regul Integr Comp Physiol 297: R1118–R1126, 2009.

44. O’Mahony L, McCarthy J, Kelly P, Hurley G, Luo F, Chen K,O’Sullivan GC, Kiely B, Collins JK, Shanahan F, Quigley EM.Lactobacillus and bifidobacterium in irritable bowel syndrome: symptomresponses and relationship to cytokine profiles. Gastroenterology 128:541–551, 2005.

45. Peiris M, Bulmer DC, Baker MD, Boundouki G, Sinha S, Hobson A,Lee K, Aziz Q, Knowles CH. Human visceral afferent recordings:preliminary report. Gut 60: 204–208, 2011.

46. Ressler KJ, Mayberg HS. Targeting abnormal neural circuits in moodand anxiety disorders: from the laboratory to the clinic. Nat Neurosci 10:1116–1124, 2007.

47. Rhee SH, Pothoulakis C, Mayer EA. Principles and clinical implicationsof the brain-gut-enteric microbiota axis. Nat Rev Gastroenterol Hepatol 6:306–314, 2009.

48. Richards W, Hillsley K, Eastwood C, Grundy D. Sensitivity of vagalmucosal afferents to cholecystokinin and its role in afferent signal trans-duction in the rat. J Physiol 497: 473–481, 1996.

49. Rong W, Hillsley K, Davis JB, Hicks G, Winchester WJ, Grundy D.Jejunal afferent nerve sensitivity in wild-type and TRPV1 knockout mice.J Physiol 560: 867–881, 2004.

50. Rosas-Ballina M, Tracey KJ. Cholinergic control of inflammation. J IntMed 265: 663–679, 2009.

51. Steed H, Macfarlane GT, Macfarlane S. Prebiotics, synbiotics andinflammatory bowel disease. Mol Nutr Food Res 52: 898–905, 2008.

52. Talham GL, Jiang HQ, Bos NA, Cebra JJ. Segmented filamentousbacteria are potent stimuli of a physiologically normal state of the murinegut mucosal immune system. Infec Immun 67: 1992–2000, 1999.

53. van der Kleij H, O’Mahony C, Shanahan F, O’Mahony L, Bienen-stock J. Protective effects of Lactobacillus rhamnosus and Bifidobacte-rium infantis in murine models for colitis do not involve the vagus nerve.Am J Physiol Regul Integr Comp Physiol 295: R1131–R1137, 2008.

54. Verdu EF, Bercik P, Verma-Gandhu M, Huang XX, Blennerhassett P,Jackson W, Mao Y, Wang L, Rochat F, Collins SM. Specific probiotictherapy attenuates antibiotic induced visceral hypersensitivity in mice. Gut55: 182–190, 2006.

55. Wang B, Mao YK, Diorio C, Pasyk M, Wu RY, Bienenstock J, KunzeWA. Luminal administration ex vivo of a live Lactobacillus speciesmoderates mouse jejunal motility within minutes. FASEB journal 24:4078–4088, 2010.

56. Wang B, Mao YK, Diorio C, Wang L, Huizinga JD, Bienenstock J,Kunze W. Lactobacillus reuteri ingestion and IK(Ca) channel blockadehave similar effects on rat colon motility and myenteric neurones. Neu-rogastroenterol Motil 22: 98–107, 2010.



Documents

Psychoactive bacteria Lactobacillus rhamnosus (JB-1) elicits rapid frequency facilitation in vagal afferents