Spatiotemporal maps reveal regional differences in the

effects on gut motility for Lactobacillus reuteri and

rhamnosus strains

R. Y. WU,* M. PASYK,* B. WANG,** P. FORSYTHE,*,�,§ J. BIENENSTOCK,*,� Y.-K. MAO,* P. SHARMA,*A. M. STANISZ* & W. A. KUNZE*,�,–

*McMaster Brain-Body Institute, St. Joseph’s Healthcare, Hamilton, ON, Canada

�Department of Medicine, McMaster University, Hamilton, ON, Canada

�Farncombe Family Digestive Health Research Institute, McMaster University, Hamilton, ON, Canada

§Firestone Institute for Respiratory Health, Hamilton, ON, Canada

–Department of Psychiatry and Behavioural Neurosciences, McMaster University, Hamilton, ON, Canada

**Centre for Simulation-Based Learning, Health Sciences Centre, McMaster University, Hamilton, ON, Canada

Abstract

Background Commensal bacteria such as probiotics

that are neuroactive acutely affect the amplitudes of

intestinal migrating motor complexes (MMCs). What

is lacking for an improved understanding of these

motility effects are region specific measurements of

velocity and frequency. We have combined intralu-

minal pressure recordings with spatiotemporal dia-

meter maps to analyze more completely effects of

different strains of beneficial bacteria on motility.

Methods Intraluminal peak pressure (PPr) was mea-

sured and video recordings made of mouse ex vivo

jejunum and colon segments before and after intra-

luminal applications of Lactobacillus rhamnosus

(JB-1) or Lactobacillus reuteri (DSM 17938). Migrating

motor complex frequency and velocity were calcu-

lated. Key Results JB-1 decreased jejunal frequencies

by 56% and 34% in colon. Jejunal velocities increased

171%, but decreased 31% in colon. Jejunal PPr

decreased by 55% and in colon by 21%. DSM 17938

increased jejunal frequencies 63% and in colon 75%;

jejunal velocity decreased 57%, but increased in colon

146%; jejunal PPr was reduced 26% and 12% in colon.

TRAM-34 decreased frequency by 71% and increased

velocity 200% for jejunum, but increased frequency

46% and velocity 50% for colon; PPr was decreased

59% for jejunum and 39% for colon. Conclusions &

Inferences The results show that probiotics and other

beneficial bacteria have strain and region-specific

actions on gut motility that can be successfully dis-

criminated using spatiotemporal mapping of diameter

changes. Effects are not necessarily the same in colon

and jejunum. Further research is needed on the

detailed effects of the strains on enteric neuron

currents for each gut region.

Keywords Lactobacilli, migrating motor complex,

motility, probiotics, spatiotemporal map.

INTRODUCTION

Ingestion of commensal and probiotic bacteria may

modulate gut motility to benefit the host. In general,

beneficial effects of ingestion may also include alter-

ations in the composition of the microbiome,1 changes

in rat enteric neuron function,2 alterations in mouse

brain neurochemistry and behavior,3,4 reduction in

blood pressure,5,6 increased anti-inflammatory and

immunoregulatory activity,7 and modulation of motil-

ity with reductions in either diarrhea or constipation.8

The cellular mechanisms through which these effects

are mediated are not clear. The influence of gastroin-

testinal commensals on motility is demonstrated from

studies of germfree mice 9,10 and those whose normal

bacterial community has been disrupted using

antibiotics.11–13

Address for Correspondence

Wolfgang A. Kunze, The McMaster Brain-Body Institute,T3306 Juravinski Innovation Tower, St Joseph’s Healthcare,50 Charlton Ave. East, Hamilton, ON L8N 4A6, Canada.Tel: +1 9055221155 ext. 35994; fax: +1 9055406593;e-mail: kunzew@mcmaster.caReceived: 21 September 2012Accepted for publication: 9 December 2012

Neurogastroenterol Motil (2013) 25, e205–e214 doi: 10.1111/nmo.12072

� 2013 Blackwell Publishing Ltd e205

Neurogastroenterology & Motility

Probiotics have been used therapeutically to help

treat motility disturbances in humans and experimen-

tal animals. Meta-analysis of the research literature

indicates that Lactobacillus species can be used to

treat diarrhea in adults or infants.14

Probiotic organisms decreased colon contraction

frequency and were used in double-blind controlled

studies to treat infectious (Lactobacillus reuteri DSM

1793815) or functional (Bifidobacterium lactis

HN01916) diarrhea in humans.15,16

DSM 17938 (DSM) gavage decreased rotavirus-in-

duced diarrhea in mice.17 It also reduced infantile colic

after 21 days oral administration18 in a randomized

double-blind controlled trial. Infantile colic is a well-

described common symptom in infants, but whether it

is due to spasm of smooth muscle in the colon or

jejunum is not known. It is plausible to suggest

beneficial effects in the reduction in pain might have

been due to decreased giant contractions of the small or

large bowel.19

Functional and slow transit constipation can occur

in infants or adulthood and is especially common in

the institutionalized elderly.20 Probiotics may be use-

ful in this condition. This would be particularly

important because few long-term effective treatments

for constipation exist, and those that are used often

have significant side effects.21 DSM supplementation

increased the frequency of bowel movements in

chronically constipated infants.22 Lactobacillus casei

Lcr35 was as effective as magnesium oxide in treating

children with chronic constipation,23 while Bifidobac-

terium breve decreased abdominal pain and fecal

incontinence while increasing defecation frequency

in chronically constipated children.24 In constipated

but otherwise healthy human adults, ingestion of

L. plantarum SN35N or SN13T strains increased

defecation frequency compared with control bacteria

(Lactococcus lactis A6 and Streptococcus thermo-

philes) in a randomized double-blind experimental

design.25 A mixture of Lactobacillus, Bifidobacterium,

and Streptococcus strains had a positive effect (reduced

diarrhea, constipation, and improved nutritional

status) on bowel movements among orthopedic reha-

bilitation in elderly (>80 years) patients.26

That microbiota can control or influence motility is

evident from the disruption of intestinal myoelectric

complexes of germfree rats.27 Conventionalization of

such animals with normal feces or even a single

bacterial strain restored adult type frequency of gut

motility complexes. It is not immediately obvious how

such studies add to the understanding of potential

therapeutic probiotic modulation of gut motility. Pro-

biotics do not need to colonize the gut to exert an effect

on the host and their action is transient requiring

regular repeated applications for the effect to be

maintained.28 Because of this, we have devised exper-

iments to study acute effects on motility, where the

bacteria are placed into the gut lumen,29 although

some investigators have added bacteria so that they

contact the serosa (intraperitoneal) rather than the

epithelial surface.30

To identify biomarkers for effective probiotics in a

mouse model for human diarrhea or constipation,31 it

would be desirable to differentiate between the effects

of different probiotic strains on gut contractions. The

motility effects of probiotics have been mainly mea-

sured using intraluminal pressure recordings from

ex vivo gut segments in modified Trendelenburg

preparations.2,29,32 In this way, the concentration of

bacteria or derived compounds can be controlled and

confounding influences of circulating hormones and

supra-intestinal autonomic reflexes are removed. Nine

day feeding of Lactobacillus rhamnosus (JB-1) to rats

decreased colon migrating motor complex (MMC)

amplitudes ex vivo,2 and this effect was reproduced

within 15 min of introducing the same bacterium into

the lumen of ex vivo segments of mouse jejunum.29

The dose responses obtained, suggest that, similar to

many therapeutic compounds, those produced by, or

emanating from JB-1, also have a dose dependent ‘drug-

like’ action on peristalsis.

The limitation of these approaches is that intralu-

minal pressure recordings sample a restricted informa-

tion set. They cannot measure parameters such as

MMC propagation velocity, or frequency when several

MMCs occur simultaneously. Neither do they readily

distinguish between MMCs and stationary contrac-

tions.33 Spatiotemporal maps constructed from video

movies of the contracting gut can overcome these

limitations because they do not record information

from a single locus. This allows spatial and temporal

patterns to be measured simultaneously.

For the present experiments, we have used ex vivo

segments of small and large mouse intestine to study

the effects of introducing intraluminal probiotic

bacteria on motility. We used DSM, a probiotic

strain with therapeutic effects in children and adults,

and JB-1 whose effects on rodent myenteric intestinal

primary afferent neurons have been previously stud-

ied.2 Lactobacillus salivarius and L. reuteri PTA

6475 were used as comparators and controls. Spatio-

temporal maps of motility were constructed from

video recordings before, during and after luminal

application of the bacteria, revealing motor patterns

which are not readily detected using other

methods.33

R. Y. Wu et al. Neurogastroenterology and Motility

� 2013 Blackwell Publishing Ltde206

METHODS

We used adult male Swiss Webster mice (20–30 g) from CharlesRiver Laboratories (Wilmington, MA, USA; http:// http://www.criver.com). The mice were killed by cervical dislocation,in line with McMaster guidelines for the use and care of animals.All ensuing procedures were ex vivo.

Organ bath motility recordings

Four or more centimeter long jejunal or distal colon segmentswere excised and the contents emptied by flushing the segmentwith Krebs saline under a 2 hPa gravity pressure head. Eachsegment was mounted in a 20-mL organ bath chamber andsubmerged in oxygenated Krebs (Fig. 1A). Oral and anal ends werecannulated, and the lumen was gravity perfused with carbogen-gassed Krebs using several Mariotte bottles.34 The intraluminalcompartment was perfused at 0.5 mL min)1 with room temper-ature buffer (19–22 �C) according to the precedent set by previousex vivo gut segment preparations29,32,35,36 for which MMC can berecorded without rundown for at least 2 h. Pilot experiments haveshown that raising the temperature to 34 �C decreases theduration for which steady-state MMCs could be recorded withoutrundown to 30 min; presumably because a higher mucosalmetabolic demand and the lack of a blood supply.

The organ chamber (serosal compartment) was perfused withprewarmed (34 �C), carbogen-gassed, Krebs solution at a rate of5 mL min)1). Oxygenated Krebs buffer was of the followingcomposition (mmol L)1): 118 NaCl, 4.8 KCl, 25 NaHCO3, 1.0NaH2PO4, 1.2 MgSO4, 11.1 glucose, and 2.5 CaCl2 bubbled withcarbogen gas (95% O2 and 5% CO2).At the beginning of theexperiment, intraluminal pressure was adjusted to 3 hPa byadjusting the heights of inflow and outflow tubes and therecordings were made at this filling pressure. Bacteria wereapplied by switching the oral luminal inflow from Krebs to Krebsplus 8.0 log cfu mL)129 bacteria by closing and opening theappropriate stopcocks, as illustrated in Fig. 1A.

Intraluminal pressure changes were measured at the midpointof the longitudinal axis of the gut segment using a Krebs-filled0.58-mm external diameter non-distensible polyethylene tube asdescribed in.32. The tube emerged from the anal and was attachedto a COBE pressure transducer (Sorin Biomedical Inc., Irvine, CA,USA). The pressure signal was amplified, digitized, stored on a PCcomputer, and analyzed off-line using PClamp 9 software (Molec-ular Devices; Sunnyvale, CA, USA).32 Peak phasic intraluminalpressure increases (PPr) were identified and measured as describedin.29

Images were recorded using a video camcorder (JVC EverioHard Disk Camcorder Model GZ-MG155U) which was placed10 cm above a gut segment (Fig. 1A). Recording was started insynchrony with the pressure recording using an 8–12 cm field of

CarbogenCarbogen

Stop cock

Pressure sensor connected to amplifier & recording system

Krebsinflow

Krebs outlfow

Krebs+ bacteria Krebs

Camera

Lumen

Perspex organ bath

Inflow tubeOutflow tube

Oral

Anal

Optical slice

2 hPa

30 s 0.4 cm

10 mm

30 s

A

B

C

D

Figure 1 Experimental setup for video and

pressure recordings. (A) Canulated intestinal

segment was maintained in preheated oxygen-

ated Krebs solution and additives are adminis-

tered through Mariotte inflow tubes. A pressure

sensor placed at the midpoint tracked the

intraluminal pressure, while a camera placed

atop captured videos of the intestinal move-

ment; (B) The spatiotemporal map was produced

via our StMap plugin which converts gut

diameter into black and white hues for repre-

sentation. Oral to anal black lines represent

contractions, lighter areas are relaxations. The

dotted line indicates the position of pressure

sensor; (C) Diameter changes at the dotted line;

(D) Pressure changes recorded at the dotted line.

Volume 25, Number 3, March 2013 Spatiotemporal maps: probiotic effects on motility

� 2013 Blackwell Publishing Ltd e207

view for the duration of the experiment. The camera output wasin raw video format (MOD) at 30 frames per second (fps). Ten-minute long video clips were excised from the MOD file usingvideo editing software (Avidemux version 2.5.0; http://www.avid-emux.org). The clips were then converted into the MOV formatusing a video converter (Zune converter version 1.1; http://ffmpeg.mplayerhq.hu). The final video clips were resampled to aresolution of 384 · 256 pixels and 25 fps.

Video recordings were analyzed using in-house image processingsoftware (StMap) developed as a plug-in for NIH ImageJ (version1.43c; NIH, Bethesda, MD, USA). The software converts the image(Fig. 1B) in each frame of the video into a black-and-whitesilhouette (Fig. 1C) and generates a spatiotemporal map using anedge detection routine. The routine first measures the diameter ateach position along the gut and then represents the physicaldiameter at each position as a hue value (ranges from 0 to 255,black–white). As gut diameter decreases during contractions, thehue value is reduced toward 0 and will be shown as darker values.As the software reads through each 10-min clip, it generates aspatiotemporal map – a pattern of alternating bands of light anddark hues that contains three sets of information: position along thegut, time, and gut diameter. Using these variables, the spatiotem-poral map becomes a motility ‘fingerprint’ whose sensitivity couldbe critically important in defining the detailed and perhaps nuancedeffects that specific bacterial strains could have on motility.

As the StMap measures the diameter changes at each position,StMap can be interpreted as a stacking of numerous 2D diametersvs time graphs. In fact, for a spatiotemporal map, if the location ofthe pressure transducer were identified (dotted line in Fig. 1B) andwere shown as a gray scale vs time graph (Fig. 1C), this graphwould be in register with the simultaneously recorded pressure vs

time recording at that locus (Fig. 1D). The diameter of the gut ateach point was plotted as a color as previously described byRoberts et al.36; and, for ease of visualization, 3D maps wereproduced by adding a z-axis that plotted the diameter (Fig 2A).

Motility parameters were measured from the spatiotemporalmap using the StMap plugin. Neurogenic, anally propagating,

migrating motor complexes MCs of the type described in Wanget al.29 generated thick dark bands that slant diagonally from leftto right. The propagation velocity (mm s)1) was measured fromthe slope of each band or trough in the each 3D map; the twomeasures are equivalent. For each treatment, slopes of 10–15successive MMCs were averaged to calculate the propagationvelocity. Migrating motor complex frequency (mHz) was calcu-lated by counting the number of MMCs during a 10-min segment.

Bacteria and drugs

JB-1 bacteria were taken from in-house stock (see Bravo et al.4)DSM18 and L. reuteri PTA 6475 (PTA 6475) were donated byBioGaia AB, Stockholm, Sweden. L. salivarius was a gift from theAlimentary Pharmabiotic Centre, University College Cork.37

Cell numbers were determined optically,38 and viability wasalways checked by ability to grow after plating on growth mediumagar plates. All other methods are as reported previously.2,38 Cellsfrom frozen stocks were thawed and centrifuged at 500 g for15 min, and the pellet was suspended in equal volume of Krebsbuffer. Then, the suspension was washed by centrifugation at thesame speed, and the cells were removed and resuspended in Krebsat the original concentration. Just prior to use, bacteria werediluted to a working concentration of 8 log cfu mL)1 in fresh Krebsbuffer.

Krebs containing bacteria were fed to the intraluminal com-partment while ion channel modulating drugs added to the Krebsbuffer perfusing the serosal compartment. The time required forthe drug solution to flow from the tap to the recording chamberwas 30 s. The intermediate conductance calcium-dependentpotassium (IKCa) ion channel blocker 1-[(2-chlorophenyl) diphe-nylmethyl]- 1H-pyrazole (TRAM-34) (Tocris Bioscience, Ellisville,MO, USA; http://www.tocris.com) distributed by Cedarlane Lab-oratories Ltd., Burlington, ON, Canada) was dissolved with puredimethyl sulphoxide (DMSO) to make 10-mmol L)1 stock solu-tions, and these were diluted in oxygenated Krebs to makeworking concentrations 30 min before use.

Statistics

Statistics were calculated using GraphPad Prism 5.0 (GraphPadSoftware, San Diego, CA, USA). Descriptive statistics are given asmeans ± SD, but in concentration-response plots, sampling errorsare displayed using SEM; the sample size is denoted by n. Thestatistically discernible difference for tests of significance was setat P = 0.05; all tests were paired t-test and two tailed.

RESULTS

Segments from 77 mice showed propagating MMC and

recordings were made from 43 jejunal and 34 distal

colonic segments taken from these animals. Migrating

motor complexes propagated in the oral to anal direc-

tion spanning from 50% to 100% of the ex vivo

segment. These contractions were in register with the

pressure pulses recorded by the intraluminal probe

when both were measured at the same locus (Fig. 1B).

Similar to Roberts et al.36 we also observed rare colonic

retrograde moving contractions, stationary or mixing

motility patterns or motor complexes that only prop-

agated for short distances (<50%); these were excluded

Krebs

Krebs + TTX

A

B

Figure 2 3D maps showing effect of TTX on jejunal motility. Hot areas

(peaks) are relaxations and cold areas (blue) valleys are contractions.

Slopes of troughs (mm s)1) give velocities of contractile wave. (A)

Control Krebs jejunal activity; (B) Administration of TTX caused the

abolition of migrating motor complexes (MMCs) and created a sta-

tionary mixing pattern as shown, this suggests that MMCs, and not the

stationary contractions, depend on neuronal activity (5/6 see Results).

R. Y. Wu et al. Neurogastroenterology and Motility

� 2013 Blackwell Publishing Ltde208

from analysis as they did not represent full MMCs nor

a repeatable and stable pattern.36 In no case (n = 8) did

application of TRAM-34 (serosally) or JB-1 (intralumi-

nally) convert these irregular patterns to conventional

MMCs as described by Roberts et al.36 Migrating

motor complex frequencies ranged for 8–14 mHz

which is comparable with Powell et al. 39 who reported

a MMC frequency of 8 mHz (1.5-min interevent inter-

val) in a similar ex vivo mouse gut preparation. Four

jejunal segments did not display anally propagating

MMCs but exhibited multiple stationary contractions

similar to those described by Gwynne et al.33 although

only when Krebs buffer was in the lumen. To test

whether stationary contractions are generated by the

enteric nervous system, we added 0.6 lmol L)1 TTX to

the Krebs buffer perfusing the serosal surface for six

jejunal segments. In 5/6 experiments, propulsive

MMCs were abolished and stationary, mixing-type

contractions developed (Fig. 2), suggesting that, in

mouse jejunum, MMCs, but not stationary contrac-

tions, are dependent on neural activity. Adding log

8 cfu mL)1 JB-1 to the lumen (n = 5) did not change the

stationary contractions nor convert them to propagat-

ing ones. The stationary contractions were not exam-

ined further.

To demonstrate the strain specificity of the motor

effects of specific bacteria, we tested additional bacte-

ria or candidate probiotics. Log 8 cfu mL)1 of the

bacteria were applied to the lumen for 30 min. The

anti-inflammatory probiotic bacteria L. salivarius,

which was without effect in a clinical trial in IBS,37

was also ineffective in previous jejunal motility exper-

iments.29 Another L. reuteri, PTA 6475 with potent

anti-inflammatory activity40 was also tested. In all, we

tested five jejunal and five colonic segments for each of

these bacterial strains. Neither L. salivarius nor PTA

6475 had any effect on MMC amplitudes, frequencies

or velocities (data not shown).

Intraluminal JB-1 and DSM modulated motility. The

onset latencies of the observed motor effects ranged

from 10 to 20 min and plateaued at 20–30 min, and the

effects were not reversible with 20 min Krebs

washout.29

Nine jejunal and seven colonic segments were

exposed to intraluminal JB-1 (Fig. 3): MMC frequencies

decreased (Krebs vs JB-1) from 41 ± 18 to 18 ± 8 mHz

(56%) (P = 0.004) for jejunum and from 14 ± 4.5 to

9.3 ± 5.0 mHz (34%) (P = 0.02) for colon. Migrating

motor complex velocities for jejunum increased from

1.7 ± 0.9 to 4.6 ± 1.8 mm s)1 (171%) (P = 0.004), but

decreased from 1.6 ± 0.5 to 1.1 ± 0.4 mm s)1 for colon

(31%) (P = 0.02). Jejunal PPr decreased from 17 ± 4.1 to

7.6 ± 4.4 hPa (55%) (P = 0.004), and also decreased

from 42 ± 4.1 to 33 ± 6.5 hPa (21%) (P = 0.03) for

colon.

Ten jejunal and seven colonic segments were

exposed to DSM (Fig. 4). Migrating motor complex

frequencies decreased from 51 ± 16 to 19 ± 16 mHz

(63%) (P = 0.008) for jejunum, but increased from

8.0 ± 4.7 to 14 ± 7.9 (75%) (P = 0.02) for colon. DSM

decreased jejunal MMC velocity from 2.0 ± 0.81 to

Krebs

JB-1

0

20

40

60

80

100Fr

eque

ncy

(mH

z)

Krebs

JB-1

0

2

4

6

8

10

Vel

ocity

(mm

s–1

)

PP

r (hP

a)

Krebs

JB-1

–10

0

10

20

30Fr

eque

ncy

(mH

z)

Krebs

JB1

0

5

10

15

20

25

Krebs

JB1

0.0

0.5

1.0

1.5

2.0

2.5

Vel

ocity

(mm

s–1

)

PP

r (hP

a)

Krebs

JB-1

0

20

40

60

Colon

** ** **

* * *

Krebs

Krebs + JB-1

JejunumA

B

C D E

F G H

Figure 3 Effects of JB1 on intestinal motility. (A) 3D map taken from

jejunal recording; (B) The administration of JB1 altered the frequency

and conduction velocity of the migrating motor complexes (MMCs);

(C–H) Summaries of before and after experiments. (C–E) For jejunum

JB1 reduced MMC frequency, increased MMC velocity, and reduced

MC peak pressure. (F–H) for colon JB-1 reduced MMC frequency (F),

reduced MMC velocity (G), and the MMC peak pressure (PPr) (H).

Volume 25, Number 3, March 2013 Spatiotemporal maps: probiotic effects on motility

� 2013 Blackwell Publishing Ltd e209

0.86 ± 1.5 mm s)1 (57%) (P = 0.04), but increased

velocity in colon from 1.3 ± 0.9 to 3.2 ± 1.6 mm s)1

(146%) (P = 0.02). PPr decreased from 12 ± 0.7 to

8.9 ± 3.0 (26%) (P = 0.08) hPa for jejunum and 42 ± 12

to 37 ± 6 hPa for colon (12%) (P = 0.1).

We have previously shown that heat-killed JB-1 had

no effect on mouse jejunum MMC PPr or frequency.29

However, we did not measure velocity in these exper-

iments and thus cannot exclude the possibility that

heat-killed bacteria somehow altered velocity without

affecting the other parameters. We thus used heat-

killed JB-1 (n = 5) and DSM (n = 5) to exclude this

possibility. Migrating motor complex frequencies

(mHz), velocities (mm s)1), and PPr (hPa) Krebs vs

heat-killed JB-1 were 41 ± 10 vs 45 ± 13 (P = 0.5),

1.5 ± 0.7 vs 1.3 ± 0.6 (P = 0.7), and 15 ± 4 vs 18 ± 4

(P = 0.3). The same parameters when DSM were

applied to the jejunum were as follows: 44 ± 5 vs

44 ± 8 (P = 0.9), 1.7 ± 0.5 vs 1.9 ± 0.6 (P = 0.7), and

15 ± 2 vs 16 ± 4 (P = 0.03).

TRAM-34 was added to the Krebs buffer perfusing

the serosal surface of eight jejunal and ten colonic

segments (Fig. 5). TRAM-34 changed rat jejunal pat-

terns of intraluminal pressure waves so that they were

less evenly spaced than for Krebs controls. The

contractions appeared to occur more in clusters with

longer intervals between clusters.41 Consistent with

these former observations, we found that intervals

between clustered jejunal MMC lengthened (Fig. 5A,B)

so that the average frequency of MMCs decreased from

34 ± 24 to 10 ± 6 mHz (71%) (P = 0.008). For colon the

MMC pattern remained the same after adding TRAM-

34, but there was an increase in frequency from

5.9 ± 4.3 to 8.6 ± 3.1 (46%) (P = 0.03). MMC velocity

was increased from 1.5 ± 0.7 to 4.5 ± 3.6 (200%)

(P = 0.04) for jejunum, and from 1.0 ± 0.3

to 1.5 ± 0.04 mm s)1 (50%) (P = 0.03) for colon.

**

** *

* *

*

Jejunum

Colon

Krebs

TRAM-340

20

40

60

80

100

Krebs

TRAM-340

5

10

15

Krebs

TRAM-340

5

10

15

20

25

Freq

uenc

y (m

Hz)

Krebs

TRAM-340

5

10

15

20

Krebs

TRAM-340.0

0.5

1.0

1.5

2.0

2.5

Vel

ocity

(mm

s–1

)

PP

r (hP

a)

Freq

uenc

y (m

Hz)

Vel

ocity

(mm

s–1

)

PP

r (hP

a)

Krebs

TRAM-340

20

40

60

80

A B C

D E F

Figure 5 Effects of TRAM-34 on intestinal motility. (A–C) TRAM-34

decreased jejunal migrating motor complex (MMC) frequency, in-

creased MMC velocity, and decreased intraluminal peak pressure (PPr).

(D–F) Colonic MMC frequency and velocity were increased by TRAM-

34, but PPr decreased.

Krebs

DSM0

20

40

60

80

Freq

uenc

y (m

Hz)

**

Krebs

DSM0

1

2

3

4

Vel

ocity

(mm

s–1

)

**

PP

r (hP

a)

Freq

uenc

y (m

Hz)

Vel

ocity

(mm

s–1

)

PP

r (hP

a)

Krebs

DSM0

5

10

15 ns

Jejunum

Krebs

DSM0

10

20

30 *

Krebs

DSM

0

2

4

6

8 *

Krebs

DSM0

20

40

60

80 nsKrebs + L. reuteri DSM 17938

Krebs

Colon

A B C

F

D

E

G H

Figure 4 Effects of DSM on intestinal motility. (A–C) DSM reduced

jejunal migrating motor complex (MMC) frequency, velocity, but not

MMC peak pressure (PPr). (D–E) Representative 3D maps representing

colon motility before (D) and 20 min after (E) adding DSM to lumen.

(F–H) DSM increased colon MMC frequency and conduction velocity

with no effect on PPr.

R. Y. Wu et al. Neurogastroenterology and Motility

� 2013 Blackwell Publishing Ltde210

TRAM-34 reduced PPr for jejunum and colon, the

respective reductions being from 14 ± 5.5 to 5.8 ± 2.4

(59%) (P = 0.02) and from 54 ± 7.9 to 33 ± 8.3 hPa

(39%) (P = 0.002).

DISCUSSION

Evidence is accumulating that certain probiotic strains

have acute actions in vivo5 and ex vivo,29 on the host’s

autonomic reflexes. These effects can occur within

minutes suggesting that the inter-kingdom signaling

responsible for them would not rely on colonization,

alteration in the microbiome composition or other

longer term adjustments.

Our experiments utilized established ex vivo mouse

jejunal29 or colon 36,39 perfusion models for which

MMCs have been shown to be generated by the enteric

nervous system because they are abolished by silencing

the neurons using TTX.29,32,36 We applied the bacteria

intraluminally, a technique that has been used for

probiotics5 or 5-HT,42 and used the methods of intra-

luminal pressure recording and spatiotemporal map-

ping36,43–45 to measure MMC parameters as previously

described.46 The use of these substantiated techniques

to compare probiotics strains in colon vs jejunal

segments has not previously been attempted.

Our results (Table 1) clearly show that such dis-

crimination was indeed possible for the Lactobacillus

strains tested and TRAM-34 in terms of the effect on

motor behavior of both small and large intestine. In

this regard, we had the surprising result that DSM and

JB-1 acted differently on the small compared with the

large intestine. There is at present no theory of enteric

neurophysiology that might explain such region-

specific differences, highlighting the scantiness of

basic research data in this area. The finding that

L. salivarius which was ineffective in vivo in an IBS

clinical trial, was also without effect on MMCs in

jejunum29 and in the present experiments on frequen-

cies or velocity in the colon, support the specificity of

the results. In addition, heat-killed JB-1 had no effect on

jejunal MMC velocity, extending the previous findings

that killed JB-1 do not alter PPr or frequency.29 Our

findings may have important clinical ramifications as

the motor disturbances of small or large intestines could

give rise to different clinical signs and symptoms which

then may require region-specific treatments.

TRAM-34 is an IKCa channel blocker with a high

degree of specificity for this channel.47 The blocker is

thought to selectively act on intrinsic primary afferent

neurons-IPANs (AH cells), increasing their excitability

by reducing the postspike slow afterhyperpolarization

(sAHP), as only these neurons functionally express

IKCa channels when there is no inflammation.32,48,49

In healthy guinea pig distal colon 10 lmol L)1

TRAM-34 applied serosally reduced pellet velocity.50,51

Conversely, in guinea pig jejunum 1 lmol L)1 TRAM-

34 was reported to not alter the number or velocity of

long distance propagating contractions.52 In rat colon,

there was no change in MMC frequency for 3 lmol L)1

TRAM-34,32 and in mouse jejunum with 5 lmol L)1

TRAM-34 there was a non-significant trend for reduc-

tion in MMC frequency of 4.9–4.1 mHz.29 In vivo rat

jejunum TRAM-34 1 mg kg)1 clustered MMCs inter-

vals so that MMCs within each cluster may have

increased in frequency, but there appeared to be longer

intervals between the clusters.41Thus, there is at

present no current consensus on species-specific

effects of TRAM-34 on propagated propulsive contrac-

tions in small vs large intestine.

JB-1 has previously been studied in terms of action

on the gut’s neuromuscular machinery. We have

shown that when myenteric neurons have been

silenced by adding TTX to the superfusate, small and

large intestinal MMCs vanish and the bacterium does

not modulate remaining neurally independent contrac-

tile activity in either rat or mouse.29,32 Furthermore,

intraluminal application decreased the amplitude of

anally propagated MMCs. Repeated daily ingestion or

acute epithelial application of JB-1 reduced the magni-

tude and duration of the IPAN sAHP. The sAHP is

generated by IKCa channel opening and TRAM-34

similarly decreases IPAN sAHP. Arguing by anal-

ogy,29,32 we proposed that JB-1 exerted its effects on

IPAN excitability, and therefore motility, by reducing

IKCa channel opening and thus the sAHP.

The present results (Table 1) for JB-1 and DSM and

TRAM-34 demonstrate that a decrease in the IKCa

channel opening is insufficient to explain all the

temporal changes evoked in jejunal and colonic

Table 1 Summary table for the effects of JB1, DSM, and TRAM-34 on

jejunum and colon › indicates an increase, fl a decrease, and s no

change in MMC parameters measured

Jejunum Colon

JB-1

Frequency fl flVelocity › flPeak pressure fl fl

DSM

Frequency fl ›Velocity fl ›Peak pressure s s

TRAM-34

Frequency fl ›Velocity › ›Peak pressure fl fl

Volume 25, Number 3, March 2013 Spatiotemporal maps: probiotic effects on motility

� 2013 Blackwell Publishing Ltd e211

MMCs. Somatic ion channels other than IKCa, or

synaptic transmission between IPANs or from IPANs

to inter-or motor neurons might also be modulated by

the probiotics to achieve the overall motor effects. It is

also possible, but less likely, as they do not innervate

the mucosa, that inter-or motor neurons might have

their firing patterns altered by the bacteria.

As IPANs are thought to transmit to excitatory

motor neurons by a fast neurotransmission53 and

inhibitory in the neurons predominantly via slow

transmission,52,54 it has been postulated that the

timing of muscle contractile patterns depends on a

complex interaction between these neural subcircuits

and sensory feedback from contracting muscle fibers.52

In this schema, the degree of suppression of the IPAN

sAHP (TRAM-34 dose) may be critical in determining

final form and timing of the overall motor pattern.55

DSM has been shown to be effective in well-

conducted trials in human infantile colic18 and also

on functional gastric disturbances, thought to be due to

disordered gastric motility.56 Our results suggest that

in terms of infantile colic the beneficial effects of the

probiotic may be due to effects on jejunum as much as

the colon as jejunal MMC frequency and velocity were

decreased in our model (Table 1). In this respect, the

use of the mouse colon as a surrogate for human

pharmacologic effectiveness of bioactive molecules is

strongly supported by other data.31 Thus, the demon-

stration that another L. reuteri (PTA 6475), with strong

anti-inflammatory activity had no effects on either

small or large intestinal segments predicts that it

would have little to no effect in infantile colic.

Furthermore, DSM has been used clinically to help

treat pediatric functional constipation.22 In our system,

DSM increased both colonic MMC frequency and

velocity (Table 1), an outcome which is again in

register with the idea that the recording of mouse

colon MMCs can be used to screen for positive effects

of candidate bacteria or products. We tentatively

predict that based on the effects of DSM on adult

mouse colon, DSM might have therapeutic potential in

slow transit constipation in elderly adults.

In conclusion, we have presented a model system

that can effectively differentiate between Lactobacil-

lus strains in terms of several key parameters of mouse

gut MMCs, and between the actions they have on the

small vs large intestine. This approach may help to

screen and identify potential therapeutic effects of

currently used or newly identified probiotic strains,

and help in correlating the specific effects of such

bacteria on the enteric nervous system with their

actions on gut motility.

FUNDING

This study was supported by a grant from the Natural Sciencesand Engineering Council of Canada (371955-2009), the GugliettiFamily Foundation, and an unrestricted grant from BioGaia AB.

DISCLOSURE

No competing interests declared.

AUTHOR CONTRIBUTIONS

WAK and RYW designed the research and wrote the manuscriptand PF helped with the design; MP, BW, YM, PS, and AMSperformed the experiments; RYW and MP analyzed the data; JBcontributed to drafting and editing the manuscript. All authorsapproved the manuscript for publication. Experiments wereconducted in the McMaster Brain-Body Institute at St Joseph’sHealthcare Hamilton.

REFERENCES

1 Riboulet-Bisson E, Sturme MH, Jeff-ery IB et al. Effect of Lactobacillussalivarius bacteriocin Abp118 on themouse and pig intestinal microbiota.PLoS ONE 2012; 7: e31113.

2 Kunze WA, Mao YK, Wang B et al.Lactobacillus reuteri enhances excit-ability of colonic AH neurons byinhibiting calcium-dependent potas-sium channel opening. J Cell Mol

Med 2009; 13: 2261–70.3 Forsythe P, Kunze WA. Voices from

within: gut microbes and the CNS.Cell Mol Life Sci 2012. [Epub aheadof print]. doi: 10.1007/s00018-012-1028-z.

4 Bravo JA, Forsythe P, Chew MV et al.Ingestion of Lactobacillus strainregulates emotional behavior and cen-tral GABA receptor expression in amouse via the vagus nerve. Proc Natl

AcadSciUSA2011;108:16050–5.5 Tanida M, Yamano T, Maeda K,

Okumura N, Fukushima Y, Nagai K.Effects of intraduodenal injection ofLactobacillus johnsonii La1 on renalsympathetic nerve activity and bloodpressure in urethane-anesthetizedrats. Neurosci Lett 2005; 389: 109–14.

6 Jauhiainen T, Vapaatalo H, Poussa T,KyronpaloS, Rasmussen M, KorpelaR.Lactobacillus helveticus fermentedmilk lowers blood pressure in hyper-tensive subjects in 24-h ambulatory

bloodpressuremeasurement[ast].AmJHypertens2005;18:1600–5.

7 Quigley EM. Probiotics in the man-agement of colonic disorders. CurrGastroenterol Rep 2007; 9: 434–40.

8 MoayyediP,FordA,TalleyNet al.Theefficacyofprobioticsinthetreatmentofirritable bowel syndrome: a systematicreview.Gut2010;59:325–32.

9 Caenepeel P, Janssens J, Vantrappen G,Eyssen H, Coremans G. Interdigestivemyoelectric complex in germ-free rats.DigDisSci1989;34:1180–4.

10 Husebye E, Hellstrom PM, MidtvedtT. Intestinal microflora stimulatesmyoelectric activity of rat smallintestine by promoting cyclic initia-tion and aboral propagation of

R. Y. Wu et al. Neurogastroenterology and Motility

� 2013 Blackwell Publishing Ltde212

migrating myoelectric complex. DigDis Sci 1994; 39: 946–56.

11 Barbara G, Stanghellini V, Brandi Get al. Interactions between commen-sal bacteria and gut sensorimotorfunction in health and disease. Am J

Gastroenterol 2005; 100: 2560–8.12 Abraham BP, Sellin JH. Drug-induced

diarrhea. In: Guandalini S, Vaziri H,eds. Diarrhea. Chpt. 23 in ClinicalGastroenterology, New York: HumanaPress, 2011: 393–423.

13 Verdu E, Bercik P, Collins S. Effect ofprobiotics on gastrointestinal func-tion: evidence from animal models.Therap Adv Gastroenterol 2009; 2:31–5.

14 Ritchie ML, Romanuk TN. A meta-analysis of probiotic efficacy for gas-trointestinal diseases. PLoS ONE

2012; 7: e34938.15 Francavilla R, Lionetti E, Castellane-

ta S et al. Randomised clinical trial:Lactobacillus reuteri DSM 17938 vs.placebo in children with acute diar-rhoea – a double-blind study. Aliment

Pharmacol Ther 2012; 36: 363–9.16 Waller PA, Gopal PK, Leyer GJ et al.

Dose-response effect of Bifidobacte-

rium lactis HN019 on whole guttransit time and functional gastroin-testinal symptoms in adults. Scand J

Gastroenterol 2011; 46: 1057–64.17 Preidis GA, Saulnier DM, Blutt SE

et al. Host response to probioticsdetermined by nutritional status ofrotavirus-infected neonatal mice. J

Pediatr Gastroenterol Nutr 2012; 55:299–307.

18 Savino F, Cordisco L, Tarasco V et al.

Lactobacillus reuteri DSM 17938 ininfantile colic: a randomized, double-blind, placebo-controlled trial. Pedi-

atrics 2010; 126: e526–33.19 Sarna S. Colonic Motility: From

Bench Side to Bedside. San Rafael(CA): Morgan & Claypool LifeSciences, 2010.

20 Gallegos-Orozco JF, Foxx-OrensteinAE, Sterler SM, Stoa JM. Chronicconstipation in the elderly. Am

J Gastroenterol 2012; 107: 18–25.21 Tack J, Muller-Lissner S, Stanghellini

V et al. Diagnosis and treatment ofchronic constipation – a Europeanperspective. NeurogastroenterolMotil 2011; 23: 697–710.

22 Coccorullo P, Strisciuglio C, Marti-nelli M, Miele E, Greco L, Staiano A.Lactobacillus reuteri (DSM 17938) ininfants with functional chronicconstipation: a double-blind, ran-

domized, placebo-controlled study.J Pediatr 2010; 157: 598–602.

23 Bu LN, Chang MH, Ni YH, Chen HL,Cheng CC. Lactobacillus caseirhamnosus Lcr35 in children withchronic constipation. Pediatr Int:

Official J Japan Pediatr Soc 2007; 49:485–90.

24 Tabbers MM, de Milliano I, Rose-boom MG, Benninga MA. Is Bifido-

bacterium breve effective in thetreatment of childhood constipation?Results from a pilot study Nutr J

2011; 10: 19–24.25 Higashikawa F, Noda M, Awaya T,

Nomura K, Oku H, Sugiyama M.Improvement of constipation andliver function by plant-derived lacticacid bacteria: a double-blind, ran-domized trial. Nutrition 2010; 26:367–74.

26 Zaharoni H, Rimon E, Vardi H, FrigerM, Bolotin A, Shahar DR. Probioticsimprove bowel movements in hospi-talized elderly patients – thePROAGE study. J Nutr Health Aging

2011; 15: 215–20.27 Husebye E, Hellstrom PM, Sundler F,

Chen J, Midtvedt T. Influence ofmicrobial species on small intestinalmyoelectric activity and transit ingerm-free rats. Am J Physiol Gastro-

intest Liver Physiol 2001; 280: G368–80.

28 Bourlioux P, Koletzko B, Guarner F,Braesco V. The intestine and itsmicroflora are partners for the pro-tection of the host: report on theDanone Symposium ‘‘The IntelligentIntestine,’’ held in Paris, June 14,2002. Am J Clin Nutr 2003; 78: 675–83.

29 Wang B, Mao YK, Diorio C et al.

Luminal administration ex vivo of alive Lactobacillus species moderatesmouse jejunal motility within min-utes. FASEB J 2010; 24: 4078–88.

30 Massi M, Ioan P, Budriesi R et al.

Effects of probiotic bacteria on gas-trointestinal motility in guinea-pigisolated tissue. World J Gastroenterol

2006; 12: 5987–94.31 Keating C, Martinez V, Ewart L et al.

The validation of an in vitro colonicmotility assay as a biomarker forgastrointestinal adverse drug reac-tions. Toxicol Appl Pharmacol 2010;245: 299–309.

32 Wang B, Mao YK, Diorio C et al.

Lactobacillus reuteri ingestion andIK(Ca) channel blockade have similareffects on rat colon motility and

myenteric neurones. Neurogastroen-terol Motil 2010; 22: 98–107.

33 Gwynne RM, Thomas EA, Goh SM,Sjovall H, Bornstein JC. Segmentationinduced by intraluminal fatty acid inisolated guinea-pig duodenum andjejunum. J Physiol 2004; 556: 557–69.

34 McCarthy EL. Mariotte’s bottle.Science 1934; 80: 100.

35 Huizinga JD, Ambrous K, Der-Sila-phet T. Co-operation between neuraland myogenic mechanisms in thecontrol of distension-induced peri-stalsis in the mouse small intestine.J Physiol 1998; 506: 843–56.

36 Roberts RR, Murphy JF, Young HM,Bornstein JC. Development of colonicmotility in the neonatal mouse-stud-ies using spatiotemporal maps. Am JPhysiol Gastrointest Liver Physiol

2007; 292: G930–8.37 O’Mahony L, McCarthy J, Kelly P

et al. Lactobacillus and Bifidobacte-rium in irritable bowel syndrome:symptom responses and relationshipto cytokine profiles. Gastroenterol-

ogy 2005; 128: 541–51.38 Kamiya T, Wang L, Forsythe P et al.

Inhibitory effects of Lactobacillus

reuteri on visceral pain induced bycolorectal distension in Sprague-Dawley rats. Gut 2006; 55: 191–6.

39 Powell AK, Fida R, Bywater RA.Motility in the isolated mouse colon:migrating motor complexes, myo-electric complexes and pressurewaves. Neurogastroenterol Motil

2003; 15: 257–66.40 Liu Y, Fatheree NY, Mangalat N,

Rhoads JM. Human-derived probioticLactobacillus reuteri strains differ-entially reduce intestinal inflamma-tion. Am J Physiol Gastrointest Liver

Physiol 2010; 299: G1087–96.41 Ferens D, Baell J, Lessene G, Smith

JE, Furness JB. Effects of modulatorsof Ca(2 + )-activated, intermediate-conductance potassium channels onmotility of the rat small intestine, invivo. Neurogastroenterol Motil 2007;19: 383–9.

42 Bulbring E, Lin RC. The effect ofintraluminal application of5-hydroxytryptamine and 5-hydroxy-tryptophan on peristalsis; the localproduction of 5-HT and its release inrelation to intraluminal pressure andpropulsive activity. J Physiol 1958;140: 381–407.

43 Hennig GW, Costa M, Chen BN,Brookes SJ. Quantitative analysis ofperistalsis in the guinea-pig small

Volume 25, Number 3, March 2013 Spatiotemporal maps: probiotic effects on motility

� 2013 Blackwell Publishing Ltd e213

intestine using spatio-temporalmaps. J Physiol 1999; 517(Pt 2): 575–90.

44 Bercik P, Bouley L, Dutoit P, BlumA, Kucera P. Quantitative analysisof intestinal motor patterns: spatio-temporal organization of nonneuralpacemaker sites in the rat ileum.Gastroenterology 2000; 119: 386–94.

45 Seerden TC, Lammers WJ, De WinterBY, De Man JG, Pelckmans PA. Spa-tiotemporal electrical and motilitymapping of distension-induced prop-agating oscillations in the murinesmall intestine. Am J Physiol Gas-

trointest Liver Physiol 2005; 289:G1043–51.

46 Lammers WJ, Cheng LK. Simulationand analysis of spatio-temporal mapsof gastrointestinal motility. Biomed

Eng Online 2008; 7: 2.47 Wulff H, Miller MJ, Hansel W,

Grissmer S, Cahalan MD, ChandyKG. Design of a potent and selectiveinhibitor of the intermediate-conductance Ca2 + -activated K+channel, IKCa1: a potential immu-nosuppressant. Proc Natl Acad Sci

USA 2000; 97: 8151–6.

48 Nurgali K, Nguyen TV, MatsuyamaH, Thacker M, Robbins HL, FurnessJB. Phenotypic changes of morpho-logically identified guinea-pig myen-teric neurons following intestinalinflammation. J Physiol 2007; 583:593–609.

49 Neylon CB, Nurgali K, Hunne B et al.Intermediate-conductance calcium-activated potassium channels in en-teric neurones of the mouse: phar-macological, molecular andimmunochemical evidence for theirrole in mediating the slow afterhy-perpolarization. J Neurochem 2004;90: 1414–22.

50 Hoffman J, McKnight N, Sharkey K,Mawe GM. The relationship betweeninflammation-induced neuronalexcitability and disrupted motoractivity in the guinea pig distal colon.Neurogastroenterol Motil 2011; 23:673–e279.

51 Strong DS, Sharkey KA, Mawe GM.Relationship between AH neuronexcitability and peristalsis in normalversus inflamed guinea pig distal co-lon. Neurogastroenterol Motil 2006;18: 695: A95.

52 Chambers JD, Bornstein JC, ThomasEA. Multiple neural oscillators andmuscle feedback are required for theintestinal fed state motor program.PLoS ONE 2011; 6: e19597.

53 Stebbing MJ, Bornstein JC. Electro-physiological mapping of fast excit-atory synaptic inputs tomorphologically and chemicallycharacterized myenteric neurons ofguinea-pig small intestine. Neurosci-

ence 1996; 73: 1017–28.54 Kunze WA, Furness JB, Bornstein JC.

Simultaneous intracellular recordingsfrom enteric neurons reveal thatmyenteric AH neurons transmit viaslow excitatory postsynaptic poten-tials. Neuroscience 1993; 55: 685–94.

55 Gwynne RM, Bornstein JC. Mecha-nisms underlying nutrient-inducedsegmentation in isolated guinea pigsmall intestine. Am J Physiol Gas-

trointest Liver Physiol 2007; 292:G1162–72.

56 Indrio F, Riezzo G, Raimondi F et al.

Lactobacillus reuteri accelerates gas-tric emptying and improves regurgi-tation in infants. Eur J Clin Invest

2011; 41: 417–22.

R. Y. Wu et al. Neurogastroenterology and Motility

� 2013 Blackwell Publishing Ltde214