of 31
A gut microbial factor modulates locomotor behavior in
Drosophila
Catherine E. Schretter
1
,
Jost Vielmetter
2
,
Imre Bartos
3
,
Zsuzsa Marka
3
,
Szabolcs Marka
3
,
Sulabha Argade
4
, and
Sarkis K. Mazmanian
1
1
Division of Biology and Biological Engineering, California Institute of Technology, Pasadena,
California
2
Protein Expression Center, Beckman Institute, California Institute of Technology, Pasadena,
California
3
Department of Physics, Columbia University, New York, New York
4
GlycoAnalytics Core, University of California, San Diego, California
While research into the biology of animal behavior has primarily focused on the central
nervous system, cues from peripheral tissues and the environment have been implicated in
brain development and function
1
. Emerging data suggest bidirectional communication
between the gut and the brain affects behaviors including anxiety, cognition, nociception,
and social interaction, among others
1
9
. Coordinated locomotor behavior is critical for the
survival and propagation of animals, and is regulated by internal and external sensory
inputs
10
,
11
. However, little is known regarding influences by the gut microbiome on host
locomotion, or the molecular and cellular mechanisms involved. Here we report that germ-
free status or antibiotic treatment result in hyperactive locomotor behavior in
Drosophila
melanogaster
. Increased walking speed and daily activity found in the absence of a gut
microbiome are rescued by mono-colonization with specific bacteria, including the fly
commensal
Lactobacillus brevis
. The bacterial enzyme xylose isomerase (Xi) from
L. brevis
is sufficient to recapitulate the locomotor effects of microbial colonization via modulation of
sugar metabolism in flies. Notably, we discover that thermogenetic activation of
octopaminergic neurons or exogenous administration of octopamine, the invertebrate
counterpart of noradrenaline, abrogates Xi-induced effects on
Drosophila
locomotion. These
findings reveal a previously unappreciated role for the gut microbiome in modulating
<license-p>Users may view, print, copy, and download text and data-mine the content in such documents, for the purposes of
academic research, subject always to the full Conditions of use:<uri xlink:href="http://www.nature.com/authors/editorial_policies/
license.html#terms">http://www.nature.com/authors/editorial_policies/license.html#terms</uri></license-p>
Correspondence and requests for materials should be addressed to C.E.S. (cschrett@caltech.edu) or S.K.M (sarkis@caltech.edu).
Author Contributions
C.E.S. designed, performed, and analyzed most of the experiments. J.V. assisted with experimental design for
biochemical analysis. I.B., Z.M., and S.M. assisted with gait analysis experiments. S.A. performed carbohydrate quantification. C.E.S.
and S.K.M supervised the project. C.E.S. and S.K.M. wrote the manuscript with assistance from all of the authors.
Code availability.
Custom code for bout analysis is available from the corresponding authors upon request.
Data availability.
All data sets generated are available from the corresponding authors upon reasonable request.
Author Information
The authors declare no competing financial interests.
HHS Public Access
Author manuscript
Nature
. Author manuscript; available in PMC 2019 April 24.
Published in final edited form as:
Nature
. 2018 November ; 563(7731): 402–406. doi:10.1038/s41586-018-0634-9.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
locomotion, and identify octopaminergic neurons as mediators of peripheral microbial cues
that regulate motor behavior in animals.
Coordinated locomotion is required for fundamental activities of life such as foraging, social
interaction, and mating, and involves the integration of multiple contextual factors including
the internal state of the animal and external sensory stimuli
10
,
11
. The intestine represents a
major conduit for exposure to environmental signals that influence host physiology, and is
connected to the brain through both neuronal and humoral pathways. Recently, seminal
studies have uncovered that the intestinal microbiome regulates developmental and
functional features of the nervous system
1
,
2
, though gut bacterial effects on the
neuromodulators and neuronal circuits involved in locomotion remain poorly understood.
Since central mechanisms of locomotion, including sensory feedback and neuronal circuits
integrating these modalities, are shared in lineages spanning arthropods and vertebrates
11
13
,
we employed the fruit fly
Drosophila melanogaster
to explore host-microbiome interactions
that contribute to locomotor behavior. Locomotion was examined in the presence
(conventional; Conv) and absence (axenic; Ax) of commensal bacteria. In comparison to
conventionally-reared animals, axenic female adult flies exhibit increased walking speed and
daily activity (Fig. 1a – b, and 1g).
Drosophila
locomotion is characterized by a pattern of
intermittent periods of pauses and activity bouts
11
,
14
, during the latter of which the average
speed of the fly is above a set threshold of 0.25 mm/second. An increased average speed
may be related to changes in temporal patterns, including the number and/or duration of
walking bouts
14
. We discovered that axenic flies display an increased average walking bout
length in addition to a decreased average pause length, while remaining indistinguishable in
the number of bouts compared to animals harboring a microbial community (Fig. 1c – f).
These data reveal that the microbiota modulates walking speed and temporal patterns of
locomotion in
Drosophila
.
The microbial community of
Drosophila melanogaster
contains 5 – 20 bacterial species
15
,
16
.
In laboratory-raised flies, two of the dominant species are
Lactobacillus brevis
and
Lactobacillus plantarum
15
. Specific bacteria in this community affect distinct features of
Drosophila
physiology, and even closely related microbial taxa can exhibit unique biological
influences on the host
15
,
17
,
18
. Accordingly, we examined whether locomotor performance
was impacted differentially by individual bacterial species. Despite similar levels of
colonization (Extended Data Fig. 1a), mono-association with
L. brevis
, but not
L. plantarum,
starting at eclosion is sufficient to correct speed and daily activity deficits in axenic flies
(Fig. 1a– b, 1g, and Extended Data Fig. 1b – e). Varying the strain of
L. brevis
or host diet
did not alter bacterial influences on host speed (Extended Data Fig. 1c – e), and
L. brevis
is
able to largely restore temporal patterns of locomotion (Fig. 1c – f and Extended Data Fig.
1f). Detailed gait analysis reveals that
L. brevis
-associated flies display comparable
locomotor coordination to that of conventionally-reared flies (Fig. 1h and Extended Data
Fig. 1g). Further, axenic flies co-colonized with a 1:1 mixture of
L. brevis
and
L. plantarum
display similar changes in speed to flies mono-associated with
L. brevis
(Extended Data Fig.
1h).
To investigate whether the effects of microbial exposure are dependent on host
developmental stage, we mono-colonized flies at 3 – 5 days post-eclosion (Extended Data
Schretter et al.
Page 2
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Fig. 2a), a time point in which the development of the GI tract and remodeling of the
nervous system are complete
19
21
. Colonization with
L. brevis
in fully developed animals
decreases locomotor speed and average walking bout length to levels similar in flies treated
immediately following eclosion (Extended Data Fig. 2b – e). Changes in locomotion are
likely independent of bacterial effects on host development, as conventionally-reared flies
treated after eclosion with broad spectrum antibiotics exhibit similar walking speeds to
animals born under axenic conditions (Extended Data Fig. 2f). Administration of antibiotics
increases fly locomotion in two different wild-type lines (Extended Data Fig. 2g).
Furthermore, colonization with
L. brevis
, but not
L. plantarum
, after the removal of
antibiotics reduces locomotor behavior to levels similar to conventional flies (Fig. 1i and
Extended Data Fig. 2h – l). From these data, we conclude that locomotion is modulated by
select bacterial species of the
Drosophila
microbiome, and is mediated by active signaling,
rather than developmental influences.
Gut bacteria secrete molecular products that regulate aspects of host physiology, including
immunity and feeding behavior
22
,
23
. To explore how microbes influence locomotion, we
administered either cell-free supernatant (CFS) harvested from bacterial cultures or heat-
killed bacteria to axenic flies. CFS alone from
L. brevis
(
L.b
CFS) reduces hyperactivity in
axenic flies, while heat-killing bacteria ablates modulation of locomotion (Fig. 2a and
Extended Data Fig. 3a – e), demonstrating a requirement for metabolically active
L. brevis
.
Previous studies have revealed that
L. brevis
produces uracil
18
, a molecule that affects the
host immune response and may impact locomotion
22
. However, administration of
physiologic levels of uracil to axenic flies did not alter walking speed (Extended Data Fig.
3f). We next explored whether immunity or feeding behavior impact microbial-mediated
locomotion. Depletion of the microbiome in Immune Deficiency (IMD) and Toll knockout
flies using antibiotics results in similar increases in walking speed compared to wild-type
flies (Extended Data Fig. 4a – b). There are no differences in the expression of anti-
microbial peptides or the dual oxidase gene,
Duox,
in
L.b
CFS-treated axenic flies
(Extended Data Fig. 4c). Moreover, while food intake may be influenced by bacterial species
and can inhibit locomotor behavior
23
25
, there is no significant change in the amount of food
ingested by
L.b
CFS-treated flies compared to controls (Extended Data Fig. 4d – e).
Bacterial metabolism of amino acids and carbohydrates is associated with changes in host
behavior
6
,
8
; however, it is not known whether bacterial metabolic enzymes influence host
locomotion. Employing biochemical analysis of
L.b
CFS and comparative functional
analysis of bacterial strains
26
28
, we determined that bacterial locomotor effects are
mediated via proteinaceous molecule(s) present in select bacteria, including
L. brevis
and
E.
coli
(Extended Data Fig. 5a – e). Subsequently, a screen of
E. coli
strains containing single
gene mutations related to amino acid and carbohydrate metabolism identified xylose
isomerase (Xi) as a candidate factor modulating locomotor behavior (Extended Data Fig.
5f). Xi is an enzyme that catalyzes the reversible isomerization of certain sugars, including
the conversion of D-glucose to D-fructose
29
, and is present only in
L. brevis
and
E. coli
of
the bacterial strains tested (Extended Data Fig. 5e). Administration of His-tagged Xi from
L.
brevis
(Xi*) reduces locomotor behavior in axenic flies to levels similar to
L.b
CFS and
conventional flies (Fig. 2b – c and Extended Data Fig. 5g – h). The addition of His-tagged
L-arabinose isomerase (Ai*), an enzyme that is not differentially expressed among the
Schretter et al.
Page 3
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
bacteria tested, is not sufficient to influence host speed in axenic flies (Fig. 2c). Furthermore,
we generated a chromosomal deletion of the xylose isomerase gene
xylA
in
L. brevis
, and
demonstrate the mutant strain lacks the ability to modulate host speed and daily activity
(Fig. 2d and Extended Data Fig. 5g). No changes in survival or intestinal cellular apoptosis
occur at the time of motor testing (Extended Data Fig. 5i – j). Additionally, Xi*-treatment
did not significantly alter sleep in axenic flies (Extended Data Fig. 6). Neither the addition
of predicted products of Xi (D-fructose, D-glucose, D-xylose, and D-xylulose) alone, nor Xi
inactivated by EDTA treatment
29
, reduces walking speed in axenic flies (Extended Data Fig.
7a – c). We next sought to explore Xi activity through carbohydrate analysis of whole flies,
which revealed flies given Xi* exhibit increased ribose and reduced trehalose levels
compared to axenic controls (Fig. 2e), with no differences in these sugars in the fly media
(Extended Data Fig. 7d). While EDTA-treated Xi* did not significantly alter trehalose
levels, these flies still display heightened levels of ribose compared to axenic controls
(Extended Data Fig. 7e). Additionally, similar to previous findings
30
, conventional and
L.
brevis
-colonized flies show reduced levels of trehalose compared to axenic groups
(Extended Data Figures 7f – g). Administration of trehalose alone reverses microbial effects
on host speed, while supplementation with arabinose or ribose does not (Fig. 2f and
Extended Data Fig. 7h – k). Collectively, these results demonstrate that xylose isomerase
from
L. brevis
is sufficient to control locomotion in
Drosophila
, likely via modulation of key
carbohydrates, such as trehalose.
Specific neuronal pathways regulate complex behaviors in animals
31
33
, and can be
modulated by peripheral inputs, including intestinal and circulating carbohydrate levels
34
.
To explore the involvement of various neuronal subsets in bacterial-induced motor behavior,
we used the thermosensitive cation channel
Drosophila
TRPA1 (dTRPA1) to activate
neuronal populations previously implicated in locomotion
35
, via a repertoire of GAL4-driver
lines. In combination with
UAS-dTrpA1
at the activity-inducing temperature (27 ̊C), we
observe that activation of only two GAL4 lines that both label octopaminergic neurons,
tyrosine decarboxylase (Tdc2) and tyramine beta-hydroxylase (T
β
h), override
L. brevis
modulation of locomotion (Fig. 3a and Extended Data Fig. 8). Accordingly, activation of
Tdc-expressing cells abrogated the effects of Xi*-treatment and differences between
conventional and antibiotic-treated groups (Fig. 3b – c and Extended Data Fig. 9). The
ability of
L. brevis
to decrease locomotion, however, is not changed by the activation of
dopaminergic, serotoninergic, GABAergic, or cholinergic neurons (Fig 3a and Extended
Data Fig. 8e – h). The administration of octopamine to conventional, Xi*-, or
L.b
CFS-
treated flies increases host walking speed to levels similar to that of axenic flies (Fig. 3d – e
and Extended Data Fig. 10a). Further,
Tdc2
and
T
β
h
transcript levels are reduced in RNA
extracted from the heads of Xi*- and
L.b
CFS-treated flies (Extended Data Figure 10b – c).
As Tdc and T
β
h are important for octopamine synthesis, these results further link
octopamine to Xi-induced locomotor effects. Octopamine and tyramine are involved in
multiple aspects of host physiology, including metabolism and behavior, and display
opposite roles in regulating certain motor behaviors
36
44
. While administration of tyramine
did not influence walking speed in Xi* and
L.b
CFS conditions (Fig. 3e and Extended Data
Fig. 10d), antibiotic-treated flies carrying a null allele for Tdc (
Tdc2
RO
54
) no longer display
differences in locomotion upon supplementation with Xi* (Fig. 3f), suggesting an indirect
Schretter et al.
Page 4
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
role for tyramine. Limiting the expression of a transgene for diphtheria toxin (
DTI
) to
octopaminergic and tyraminergic neurons outside of the ventral nerve cord
39
,
45
results in
equivalent average speeds between antibiotic and Xi*-treated flies (Extended Data Fig. 10e),
implicating the involvement of neurons in the supraesophageal and the subesophageal zones
in microbial effects on motor behavior. Octopamine signaling is necessary for locomotor
changes, as axenic flies administered with mianserin, an octopamine receptor antagonist, and
antibiotic-treated flies carrying a null allele for T
β
h (
T
β
H
M18
) or expressing T
β
h RNAi no
longer respond to Xi* or
L.b
CFS treatment (Fig. 3g and Extended Data Fig. 10f – h).
Similar results are also found under conventional conditions compared to antibiotic-treated
groups (Extended Data Fig. 10i – k). Collectively, we conclude that defined products of the
microbiome, and specifically Xi, negatively regulate octopaminergic pathways to control
Drosophila
locomotion (Extended Data Fig. 10l).
The microbiome influences neurodevelopment, regulates behavior, and contributes to
various neurologic and neuropsychiatric disorders. Herein, we demonstrate that gut bacteria
modulate locomotion in female
Drosophila
. The depletion of the gut microbiota increases
host exploratory behavior, and the commensal bacterium
L. brevis
is sufficient to regulate
locomotion. In addition, we establish that xylose isomerase from
L. brevis
corrects the
locomotor phenotypes of axenic flies, a process that is mediated by trehalose and
octopamine signaling in the host. However, further work is needed to identify the exact
neurons and neuronal mechanisms involved, including potential changes in firing patterns. It
would also be important to clarify the sex-specific aspects of these microbial effects on
locomotion
30
,
46
. It is intriguing that germ-free mice display hyperactivity similar to axenic
Drosophila
, and specific bacteria have been shown to decrease locomotor activity in mice
1
,
47
,
48
, although the neuronal pathways implicated in mammalian systems have yet to be
identified. The mammalian counterpart of octopamine, noradrenaline, modulates
locomotion
31
,
49
, potentially implicating adrenergic circuitry as a conserved pathway that is
co-opted by the microbiome in flies and mammals. In addition to motor behavior,
octopamine signaling is linked to sugar metabolism, and trehalose serves as a major energy
source for
Drosophila
36
. Xylose isomerase may therefore facilitate adrenergic regulation of
host physiology through orchestrating metabolic homeostasis, such as via altering internal
energy storage, although additional work is needed to define how the microbiome mediates
interactions between sugar metabolism and octopamine signaling. The inextricable link
between metabolic state and locomotion suggests that peripheral influences on metabolism
may signal via neuronal pathways to modulate physical activity. As animals have become
metabolically intertwined with their microbiomes, perhaps it is not surprising that a
fundamental trait such as locomotion is influenced by host-microbial symbiosis.
METHODS
Fly Stocks and Rearing.
We obtained Canton-S (#64349),
Imd
−/−
(#55711),
Ti
−/−
(#30652),
UAS-dTrpA1
(#26264),
Tdc2-GAL4
(#52243),
T
β
h-GAL4
(#48332),
Th-GAL4
(#8488),
Ddc-GAL4
(#7009),
Gad1-GAL4
(#51630),
ChAT-GAL4
(#60317),
Elav-GAL4
(#46655),
UAS-T
β
h
RNAi
(#27667),
UAS-DTI
(#25039), and
pBDPG4U-GAL4
(#68384) lines from Bloomington
Schretter et al.
Page 5
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Drosophila
Stock Center at Indiana University. Other fly stocks used were Oregon
R
(kindly
provided by A. A. Aravin and K. Fejes T
ό
th),
T
β
H
M18
(kindly provided by M. H.
Dickinson)
50
,
Tdc2
R054
, and
tsh-GAL80
(both kindly provided by D. J. Anderson)
51
,
52
. To
minimize the effect of genetic background on behaviors, mutant fly lines were outcrossed
for at least three generations onto a wild-type background.
Flies were cultured at 25 ̊C and 60% humidity on a 12-hr. light:12-hr. dark cycle and kept in
vials containing fresh fly media made at California Institute of Technology consisting of
cornmeal, yeast, molasses, agar, p-hydroxy-benzoic acid methyl ester. Other dietary
compositions used were created through altering this standard diet or the Nutri-Fly “German
Food” Formula (Genesee Scientific) and were calculated using previously published
nutritional data
53
. Axenic flies were generated using standard methods
18
,
54
58
. Briefly,
embryos from conventional flies were washed in bleach, ethanol, and sterile PBS before
being cultivated on fresh irradiated media
54
. Axenic stocks were maintained through the
application of an irradiated diet supplemented with antibiotics (500 μg/ml ampicillin,
Putney; 50 μg/ml tetracycline, Sigma; 200 μg/ml rifamycin, Sigma) for at least one
generation. For experiments, virgin female flies were collected shortly after eclosion and
placed at random into vials (10 – 15 flies per vial) containing irradiated media without
antibiotics. Vials were changed every 3 – 4 days using sterile methods. The antibiotic
supplemented diet was applied to conventional flies shortly after eclosion to generate
antibiotic-treated (ABX) flies. Both antibiotic-treated and axenic flies were tested for
contaminants through plating animal lysates on Man, Rogosa, and Sharpe (MRS, BD
Biosciences), Mannitol (25 g/L Mannitol, Sigma; 5 g/L Yeast extract, BD Biosciences; 3 g/L
Peptone, BD Biosciences), and Luria-Bertani (LB, BD Biosciences) nutrient agar plates.
Bacterial Strains.
Lactobacillus brevis
EW
,
Lactobacillus plantarum
WJL
, and
Acetobacter pomorum
were
obtained from laboratory-reared flies in the laboratory of Won-Jae Lee (Seoul National
University)
18
,
56
,
58
.
Lactobacillus brevis
Bb14
(ATCC, #14869) and
Lactobacillus brevis
P−2
(ATCC, #27305) were isolated from human feces and fermented beverages, respectively.
Escherichia coli
K12
(CGSC, #7636) was grown in LB broth and
Escherichia coli
∆tyrA
(CGSC, #9131),
Escherichia coli
∆trpC
(CGSC, #10049),
Escherichia coli
∆manX
(CGSC,
#9511),
Escherichia coli
∆treA
(CGSC, #9090), and
Escherichia coli
∆xylA
(CGSC,
#10610)
28
were grown in LB broth supplemented with kanamycin (50 μg/mL).
Lactobacillus brevis
and
Lactobacillus plantarum
cultures were grown overnight in a
standing 37 ̊C incubator in MRS broth (BD Biosciences). For mono-associations, fresh
stationary phase bacterial cultures (OD
600
= 1.0, 40 μL) were added directly to fly vials.
Associations with two bacteria were performed in a 1:1 mixture. For heat-killed
experiments, fresh cultures of
Lactobacillus brevis
(OD
600
= 1.0) were washed 3 times in
sterile PBS, incubated at 100 ̊C for 30 min., and cooled to room temperature before
administering to flies. All treatments were supplied daily through application to the fly
media (40 μL) for 6 days following eclosion.
Schretter et al.
Page 6
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Bacterial Supernatant Preparations.
Cell-free supernatants (CFS) of specified bacterial strains were harvested from bacterial
cultures (OD
600
= 1.0) by centrifuging at 13,000 ×
g
for 10 min. and subsequent filtration
through a 0.22-μm sterile filter (Millipore). CFS was dialyzed in MilliQ water with a 3.5
kDa membrane (Thermo Scientific) overnight at 4 ̊C to generate
L.b
CFS and
L.p
CFS
samples. Each of these treatments were supplied daily through application to the fly media
(40 μL) for 6 days following eclosion.
Heat and Enzymatic Treatment of
L.b
CFS.
For heat-inactivation experiments, freshly prepared
L.b
CFS samples were incubated at
100 ̊C for 30 min. and cooled to room temperature before administering to flies. For
proteinase K (PK) and trypsin (Typ) treatment, overnight dialysis of CFS was performed in
Tris-HCl (pH 8 for PK and pH 8.5 for Typ) after which samples were treated with either PK
(100 μg/mL, Invitrogen) or Typ (0.05 μg/mL, Sigma) at 37 ̊C for 24 or 7 hrs., respectively. A
proteinase inhibitor cocktail (Sigma) was added to stop the reaction and subsequently
removed through overnight dialysis (Thermo Scientific) at 4 ̊C in MilliQ water. Aliquots of
the samples were run on a 4–20% Tris-glycine gel (Invitrogen) to confirm protein cleavage.
Controls followed the same protocol except for the addition of proteinase K or trypsin. For
amylase digests, 20 μL of 100 mU/mL amylase (Sigma) was added to either freshly prepared
L.b
CFS or a PBS control for 30 min. and inhibited through lowering the pH to 4.5. Each of
these treatments were supplied daily through application to the fly media (40 μL) for 6 days
following eclosion.
Production of His-tagged proteins (Xi* and Ai*)
An expression plasmid for the production of His-tagged xylose isomerase from
L.b
, here
termed as Xi*, was constructed by amplification of its gene and cloning the resulting PCR
product in the pQE30 cloning vector (Qiagen) using SLIC ligation. The following primer
sequences were used for the construct: 5’-
CGCATCACCATCACCATCACGGATCTTACTTGCTCAACGTATCGATGATGTAA-3’
and 5’-
GGGGTACCGAGCTCGCATGCGGATCATGACTGAAGAATACTGGAAAGGC-3’.
Conformation of the resulting plasmid was verified and transformed into
E. coli
(Turbo,
NEB). This strain was then grown in LB containing ampicillin (100 μg/mL) and
chloramphenicol (25 μg/mL) with shaking at 220 rpm at 37 ̊C for 1 hr. before the addition of
0.1 mM IPTG. After 4 hrs. of shaking at 220 rpm at 37 ̊C, cells were pelleted and lysed
using lysozyme (Sigma) and bead beating with matrix B beads (MP Biomedicals) for 45 sec.
Supernatant was collected after centrifugation and the Xi* protein purified through metal
affinity purification under native conditions using HisPur™ Ni-NTA Spin Columns (Thermo
Scientific). Protein purification was verified through western blot using an Anti-6X His tag®
antibody (Abcam) and quantified using a Pierce BCA Protein Assay kit (Thermo Scientific)
after which protein was stored at −20 ̊C. Expression and purification of His-tagged L-
arabinose isomerase from
L.b
, here termed as Ai*, was performed under the exact same
conditions and the following primer sequences were used for the construct: 5’-
GGGGTACCGAGCTCGCATGCGGATCATGTTATCAGTTCCAGATTATGAATTTTGG-3
Schretter et al.
Page 7
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
’ and 5’-
CGCATCACCATCACCATCACGGATCCTTACTTGATGAACGCCTTTGTCAT-3’. For
EDTA treatment
29
, purified Xi* was combined with 5 mM EDTA for 44 hrs. at 4°C and
subsequently dialyzed prior to administering to flies through application to the fly media (40
μL) for 6 days following eclosion.
Generation of
xylA
deletion mutant (∆xylA)
~1-kb DNA segments flanking the region to be deleted were PCR amplified using the
following primers: 5’-ATTCCAATACTACCACTAGCAACGACATCCGTAAAGT-3’; 5’-
AATTCGAGCTCGGTACCCGGGGATCCACAATCAGAATTGATCGCGGCAAC-3’; 5’-
TCGTTGCTAGTGGTAGTATTGGAATCCTAAACCAGATTTCTTATCTTGATG-3’; 5’-
GCCTGCAGGTCGACTCTAGAGGATCCCGCAAGTCTAGTGCGGCT-3’. The forward
primers were designed using to be partially complementary at their 5’ ends by 25 bp. The
fused PCR product was cloned into the BamHI site of the
Lactobacilli
vector pGID023 and
mobilized into
L.b
. Colonies selected for the erythromycin (Erm) resistance, indicating
integration of the vector into the host chromosome were re-plated onto MRS+Erm and
subsequently passaged over 5 days and plated onto MRS+Erm. Colonies selected for Erm
resistance were passaged again in MRS alone over 3 days and plated on MRS. Resulting
colonies were plated in replica on MRS and MRS+Erm. Erm sensitive colonies were
screened by PCR to distinguish wild-type revertants from strains with the desired mutation.
Drug treatments.
Axenic flies were either left untreated or administered with
L.b
CFS or Xi* for 3 days after
eclosion. After switching to new irradiated fly media, groups of axenic flies were treated
through application to the fly media (40 μL) with octopamine (OA, 10 mg/mL, Sigma),
tyramine (TA, 10 mg/mL, Sigma), L-dopa (1 mg/mL, Sigma), or mianserin (2 mg/mL) every
day for 3 days before testing, similar to previously published methods
33
,
37
.
Bacterial Load Quantification.
Intestines dissected from surface sterilized 7-day-old adult female flies were homogenized in
sterile PBS with ~100 μl matrix D beads using a bead beater. Lysate dilutions in PBS were
plated on MRS agar plates and enumerated after 24 hrs. at 37 ̊C.
Locomotion Assays.
Locomotor behavior was assayed through three previously established methods: the
Drosophila
Activity Monitoring System (DAMS, Trikinetics)
59
,
60
, video-assisted
tracking
61
63
, and gait analysis
64
.
Activity measurements.—
7-day-old individual female flies were cooled on ice for 1
min. and transferred into individual vials (25 × 95 mm) containing standard irradiated
media. Tubes were then inserted and secured into Drosophila activity monitors (DAMS,
Trikinetics) and kept in a fly incubator held at 25 ̊C. Flies were allowed to acclimate to the
new environment for 1 day before testing and midline crossing was sampled every min.
Average daily activity was calculated from the 2 days tested and actograms were generated
Schretter et al.
Page 8
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
using ActogramJ
60
. Sleep was defined as a 5 min. bout of inactivity as previously
described
65
.
Video-assisted tracking.—
Individual female flies were cooled on ice for 1 min. before
being introduced under sterile conditions into autoclaved arenas (3.5 cm diameter wells),
which allowed free movement but restricted flight. After a 1 hr. acclimation period, arenas
were placed onto a light box and recorded from above for a period of 10 min. at 30 frames
per sec. All testing took place between ZT 0 and ZT 3 (ZT, Zeitgeber time; lights are turned
on at ZT 0 and turned off at ZT 12) and both acclimation and testing occurred at 25 ̊C unless
otherwise stated. Videos were processed using Ethovision software or the Caltech
FlyTracker (
http://www.vision.caltech.edu/Tools/FlyTracker/
).
Bout analysis was performed using custom python scripts (available upon request). The
velocity curve was smoothed from the acquired video at 30 frames per sec. using a 15 sec.
moving average window. A minimum walking speed of 0.25 mm/s was given below which
flies were moving but not walking (‘pause bouts’) and above which they were designated as
walking (‘walking bouts’). Lengths were measured as time between bout onset and offset.
Gait analysis.—
Experiments used an internally illuminated glass surface with frustrated
total internal refraction (fTIR) to mark the flies’ contact with the glass
64
. The movement of
the flies and their contact was recorded with a high-framerate camera, and videos were
quantified using the FlyWalker software package. For further details of the parameters see
64
. All groups consisted of 7-day-old female flies and were tested at room temperature.
Feeding Assays.
Female flies were collected at the same time as described for
Locomotor Assays
. Flies were
transferred regularly onto fresh food until day 7, upon which the flies were starved for 2 hrs.
and subsequently transferred for 30 min. to an irradiated standard fly media dyed with
FD&C Blue no. 1 (Sigma) at a final concentration of 0.5 g dye per 100 g food. Flies were
allowed to feed on the food (3–4 biological replicates and 7 flies per replicate) at 25 ̊C after
which they were decapitated and their bodies collected. Each replicate was homogenized in
150 μL of PBS/0.05% Triton X-100 and centrifuged at 5,000 ×
g
for 1 min. to remove
debris. Absorbance for all groups was measured together at 630 nm and the amount of food
consumed was estimated from a standard curve of the same dye solution. The MAFE assay
was performed as described previously
66
67
. Briefly, individual flies were introduced into a
200 μL pipette tip, which was cut to expose the proboscis. Flies were first water satiated and
presented with 100 mM sucrose delivered in a fine graduated capillary (VWR). After flies
were unresponsive to 10 food stimuli, the assay was terminated and the total volume of food
was calculated.
Measurement of life span
Adult female flies were transferred under sterile conditions to irradiated fly media every 4 –
5 days. Survival in 3 or more independent cohorts containing 15 – 25 flies each was
monitored over time.
Schretter et al.
Page 9
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Apoptosis assay
Midguts from 7-day-old female flies were dissected in PBS containing 0.1% Triton X-100
and the apoptosis assay was performed as previously described
18
,
56
. The percentage of
apoptotic cells was determined by dividing the number of apoptotic cells by the total number
of cells in each section and multiplying by 100.
Measurement of carbohydrate levels
Fly (5 flies per sample) and fly media (0.1 g per sample) were homogenized in TE Buffer
(10 mM Tris, pH = 8, 1 mM EDTA) using a bead beater for 45 sec. followed by
centrifugation at 7,000 × g for 3 min. The supernatant was heat treated for 30 min. at 72°C
before being stored at −80°C before subsequent clean-up steps prior to running on HPAEC-
PAD.
100 μL of fly or fly media homogenate sample in TE buffer was diluted with 200 μL of
UltraPure distilled water (Invitrogen) and sonicated to get uniform solution. Samples were
centrifuged at 2,000 rpm for 15 sec. to precipitate insoluble material. 100 μL of the sample
was filtered through pre-washed Pall Nanosep® 3K Omega centrifugal device (MWCO
3KDa, Sigma-Aldrich) for 15 min. at 14,000 rpm, 7°C. The filtrate was dried on Speed Vac.
The dry sample was reconstituted in 300 μL of UltraPure water and loaded onto pre-washed
Dionex OnGuard® IIH 1cc cartridge. The flow through and 2×1 mL elution with Ultrapure
water was collected in the same tube and lyophilized.
Monosaccharide analysis was done using Dionex CarboPac™ PA1 column (4X250mm) with
PA1 guard column (4×50mm). Flow rate 1ml/min. Pulsed amperometric detection with gold
electrode. The elution gradient was as follows: 0 min. – 20 min., 19 mM sodium hydroxide;
20 min. – 50 min., 0 mM - 212.5 mM sodium acetate gradient with 19 mM sodium
hydroxide; 50 min. – 65 min., 212.5 mM sodium acetate with 19 mM sodium hydroxide; 65
min. – 68 min., 212.5 mM – 0 mM sodium acetate with 19 mM sodium hydroxide; 68 min. –
85 min., 19 mM sodium hydroxide
Trehalose, arabinose, galactose, glucose, mannose, xylose, fructose, ribose, sucrose and
xylulose were used as standards. The monosaccharides were assigned based on the retention
time and quantified using Chromeleon™ 6.8 chromatography data system software. In
Extended Data Figures 7f – g, measurements of trehalose levels were performed following
the same isolation procedure and subsequently processed using a Trehalose Assay Kit
(Megazyme) according to the manufacturer’s instructions.
For experiments treating flies with trehalose, arabinose, or ribose, groups of axenic or axenic
flies previously treated with Xi* were administered with trehalose, arabinose, or ribose (10
mg/mL, Sigma) through application to the fly media (40 μL) for every day for 3 days before
testing.
RNA isolation and quantitative real-time PCR.
Heads (20 flies per sample) or decapitated bodies (5 flies per sample) were dissected on ice
and immediately processed using an Arcturus™ PicoPure™ RNA isolation kit (Applied
Biosystems) or a standard TRIzol™-Chloroform protocol (ThermoFisher). 1 μg of RNA was
Schretter et al.
Page 10
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
reverse transcribed using iScript cDNA Synthesis Kit, according to manufacturer’s protocol
(Bio-Rad) and diluted to 10 ng/μL based on the input concentration of total RNA.
Previously published primer pairs were used to target immune-related gene transcripts
18
,
68
.
Other primer sequences used include
Tdc
(F: GGTCTGCCGGACCACTTTC, R:
CACTCCGATGCGGAAGTCTG),
T
β
h
(F: GCTTATCCGACACAAAGCTGC, R:
GAAAGCATTCTGCAAGTGGAA),
Ddc
(F: TGGGATGAGCACACCATCTTG, R:
GTAGAAGGGAATCAAACCCTCG),
Tph
(F: TGTTTTCGCCCAAGGATTCGT, R:
CACCAGGTTTATGTCATGCTTCT). All primers were synthesized by Integrated DNA
Technologies. Real-time PCR for the house-keeping genes Rp49 and RpL32 were used to
ensure that input RNA was equal among all samples. Real-time PCR was performed on
cDNA using an ABI PRISM 7900 HT system (ThermoFisher) according to the
manufacturer’s instructions.
Data reporting and statistical analysis.
Sample size was based on previous literature in the field and experimenters were not blinded
as almost all data acquisition and analysis was automated. After eclosion, virgin female flies
with the same genotype were sorted into groups of 10–15 flies per vial at random. All flies
in each vial were administered with the same treatment regime. For each experiment, the
experimental and control flies were collected, treated, and tested at the same time. A Mann-
Whitney
U
or Kruskal-Wallis and Dunn’s post-hoc test was used for statistical analysis of
behavioral data and carbohydrate analysis. Comparisons with more than one variant were
first analyzed using Two-way ANOVA. An unpaired two-sided Student’s t-test or a One-way
ANOVA followed by a Bonferroni post-hoc test was used for statistical analysis of
quantitative RT–PCR results and CFU analysis. All statistical analysis was performed using
Prism Software (GraphPad, version 7).
P
values are indicated as follows: ****
P
< 0.0001;
***
P
< 0.001; **
P
< 0.01; and *
P
< 0.05. See Supplementary Material for more details on
statistical tests and exact
P
values for each figure. For boxplots, lower and upper whiskers
represent 1.5 interquartile range of the lower and upper quartiles, respectively; boxes
indicate lower quartile, median, and upper quartile, from bottom to top. When all points are
shown, whiskers represent range and boxes indicate lower quartile, median, and upper
quartile, from bottom to top. Bar graphs are presented as mean values +/− standard error of
the mean (S.E.M.).
Schretter et al.
Page 11
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Extended Data
Extended Data Figure 1 |. Effects of colonization level, bacterial strain, and host diet on L.
brevis-modulation of locomotion.
a
, Colony forming units (CFU) per individual fly for
L.p
or
L.b
mono-associated flies. Bars
represent mean +/− S.E.M.
L.p
, n = 15;
L.b
, n =18.
b
, Average speed of Conv, Ax, and
L.b
mono-associated female or male flies. Females: Conv, n = 90; Ax, n = 92;
L.b
, n = 89;
Males: Conv, n = 100; Ax, n = 100;
L.b
, n = 95.
c
d
, Average speed of Ax or flies mono-
associated with
L.b
strains EW, Bb14, or P-2. (
c
) Ax, n = 58;
L.b
EW, n = 57;
L.b
Bb14, n =
Schretter et al.
Page 12
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
57. (
d
) Ax, n = 45;
L.b
EW, n = 28;
L.b
P-2, n = 42.
e
, Average speed of Ax or
L.b
mono-
associated flies raised on different diet compositions from eclosion until day 7. Diet 1: Ax, n
= 20;
L.b
, n = 21; Diet 2: Ax, n = 18;
L.b
, n = 16; Diet 3: Ax, n = 6;
L.b
, n = 6.
f
, Average
speed during walking bouts for Conv, Ax,
L.p
, and
L.b
groups. Conv, n = 23; Ax, n = 35;
L.p
, n = 22;
L.b
, n = 22.
g
, Tripod index for Conv, Ax,
L.p
, and
L.b
groups. Conv, n = 6; Ax,
n = 7;
L.p
, n = 5;
L.b
, n = 5.
h
, Average speed of Ax flies or flies mono-associated with
L.p
or
L.b
alone or in combination (1:1). Ax, n = 18;
L.p
, n = 24;
L.b
, n = 24;
L.p
+
L.b
, n = 24.
Box-and-whisker plots show median and IQR. *
P
< 0.05, **
P
< 0.01, ***
P
< 0.001, ****
P
< 0.0001. Specific
P
values are in the Supplementary Material. Unpaired Student’s t-test
(
a
), Kruskal-Wallis and Dunn’s (
b
d
and
f
h
), or Mann-Whitney
U
(
e
) post-hoc tests
were used for statistical analysis. Data are representative of at least 3 independent trials for
each experiment.
Schretter et al.
Page 13
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Extended Data Figure 2 |. Post-eclosion microbial signals decrease host locomotion.
a
, Experimental design (
b
e
) in which Ax flies were associated with
L.b
either directly
after (day 0, dark green arrows) or 3 – 5 days (light green arrows) following eclosion.
b
d,
Average speed (
b
), average bout length (
c
), and average speed during walking bouts (
d
) of
Ax and flies mono-associated with
L.b
at either day 0 or day 3 – 5. (
b
) Ax, n = 46;
L.b
0d, n
= 47;
L.b
3–5d, n = 43. (
c
) Ax, n = 18;
L.b
0d, n = 18;
L.b
3–5d, n = 6. (
d
) Ax, n = 36;
L.b
0d, n = 36;
L.b
3–5d, n = 12.
e
, Average speed of Conv, Ax, and flies mono-associated with
L.b
at either day 0 or day 3 – 5. Conv, n = 11; Ax, n = 53;
L.b
0d, n = 53;
L.b
3–5d, n = 52.
Schretter et al.
Page 14
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
f
, Average speed of Conv, Ax, and Conv flies treated with antibiotics for 3 days after
eclosion (ABX). Conv, n = 32; Ax, n =36; ABX, n = 36.
g
, Average speed of Oregon
R
(OR)
and Canton S (CS) Conv flies and Conv flies treated with antibiotics for 3 days after
eclosion (ABX). OR: Conv, n = 20; ABX, n = 22; CS: Conv, n = 12; ABX, n = 17.
h
,
Experimental design (
i
l
) in which conventionally-reared flies were treated with antibiotics
(ABX, black arrow) for 3 days following eclosion. All flies were subsequently placed on
irradiated media either without supplementation (ABX) or associated with
L.p
(blue arrows)
or
L.b
(green arrows) for the 3 days prior to testing.
i
k
, Average speed (
i
), average bout
length (
j
), and average speed during walking bouts (
k
) calculated for ABX,
L.p
-, and
L.b
-
associated flies. (
i
) ABX, n = 29;
L.p
, n = 24;
L.b
, n = 35. (
j
) ABX, n = 36;
L.p
, n = 30;
L.b
,
n = 35. (
k
) ABX, n = 42;
L.p
, n = 30;
L.b
, n = 35.
l
, Daily activity of ABX,
L.p
and,
L.b
groups (virgin female Oregon
R
flies) over a 2-day light-dark cycle period each lasting 12
hours, starting at time 0. White boxes represent lights on and gray boxes represent lights off.
n = 6/condition. Box-and-whisker plots show median and IQR. *
P
< 0.05, **
P
< 0.01.
Specific
P
values are in the Supplementary Material. Kruskal-Wallis and Dunn’s (
b
f
and
i
– l
) or Mann-Whitney
U
(
g
) post-hoc tests were used for statistical analysis. Data are
representative of at least 2 independent trials for each experiment.
Extended Data Figure 3 |. Bacterial-derived products from L. brevis alter locomotion.
a
, Average speed of Ax,
L.p
or
L.b
mono-associated, and Ax flies treated with cell-free
supernatant (CFS) from
L.p
or
L.b
. Ax, n = 45;
L.p
, n = 17;
L.b
, n = 42;
L.p
CFS, n = 17;
L.b
CFS, n = 16.
b – e
, Average speed (
b
), average bout length (
c
), average speed during
Schretter et al.
Page 15
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
walking bouts (
d
), and daily activity (
e
) of Ax and Ax virgin female Oregon
R
flies treated
with CFS from either
L.p
or
L.b
. White boxes represent lights on and gray boxes represent
lights off. (
b
) Ax, n = 23;
L.p
CFS, n = 20;
L.b
CFS, n = 20. (
c
) Ax, n = 23;
L.p
CFS, n =
20;
L.b
CFS, n = 17. (
d
) Ax, n = 22;
L.p
CFS, n = 21;
L.b
CFS, n = 17. (
e
) Ax, n = 8;
L.p
CFS, n = 8;
L.b
CFS, n = 4.
f
, Average speed of Ax,
L.b
mono-associated, and Ax uracil-
treated flies. Ax, n = 96;
L.b
, n = 88; 0.1 Uracil, n = 41; 10 Uracil, n = 18. Box-and-whisker
plots show median and IQR. *
P
< 0.05, **
P
< 0.01, ***
P
< 0.001, ****
P
< 0.0001.
Specific
P
values are in the Supplementary Material. Kruskal-Wallis and Dunn’s post-hoc
tests were used for statistical analysis. Data are representative of at least 2 independent trials
for each experiment.
Extended Data Figure 4 |. Locomotor phenotypes are independent of food intake, anti-microbial
peptides, as well as the Immune Deficiency (IMD) and Toll pathways.
a
, Average speed of wild-type background (Oregon
R
, Wt) and
Imd
−/−
flies placed on either
media alone or media supplemented with antibiotics (ABX) following eclosion. Wt: Conv, n
= 16; ABX, n = 17;
IMD
−/
: Conv, n = 24; ABX, n = 25.
b
, Average speed of wild-type
background (Canton S, Wt) and
Ti
−/−
flies placed on either media alone or media
supplemented with antibiotics (ABX) following eclosion. Wt: Conv, n = 15; ABX, n = 17;
Ti
−/
: Conv, n = 10; ABX, n = 11.
c
, qRT-PCR of immune-related transcripts in Ax and Ax
L.p
- or
L.b
- CFS treated flies. Bars represent mean +/− S.E.M.
Dpt
: Ax, n = 8;
L.p
CFS, n =
10;
L.b
CFS, n = 10;
Drs
: Ax, n = 10;
L.p
CFS, n = 10;
L.b
CFS, n = 10;
Cec
: Ax, n = 8;
L.p
CFS, n = 10;
L.b
CFS, n = 10;
AttA
: Ax, n = 5;
L.p
CFS, n = 5;
L.b
CFS, n = 5;
Duox
: Ax, n
= 3;
L.p
CFS, n = 5;
L.b
CFS, n = 5.
d
, Amount ingested by Ax and Ax
L.p
- or
L.b
- CFS
Schretter et al.
Page 16
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
treated flies over 10 trials during MAFE assay. Ax, n = 6;
L.p
CFS, n =5;
L.b
CFS, n = 6. e,
Intestinal content measured through supplementing the diet of Conv, Ax, and
L.p
- or
L.b
-
CFS treated Ax flies with blue food dye. Conv, n = 7; Ax, n = 13;
L.p
CFS, n = 7;
L.b
CFS,
n = 10. Box-and-whisker plots show median and IQR. *
P
< 0.05, **
P
< 0.01, ****
P
<
0.0001. Specific
P
values are in the Supplementary Material. Mann-Whitney
U
(
a
b
), One-
way ANOVA and Bonferroni (c), or Kruskal-Wallis and Dunn’s (
d
e
) post-hoc tests were
used for statistical analysis. Data are representative of at least 2 independent trials for each
experiment.
Dpt
, Diptericin;
Drs
, Drosomycin;
Cec
, Cecropin;
AttA
, Attacin-A;
Duox
, Dual
Oxidase.
Schretter et al.
Page 17
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Extended Data Figure 5 |. Modulation of locomotion by the bacterial enzyme, xylose isomerase.
a
c
, Average speed of Ax or Ax flies treated with unaltered, protease- (Typ, Trypsin; PK,
Proteinase-K), or heat-treated (100°C)
L.b
CFS. (
a
) Ax, n = 18;
L.b
CFS, n = 18; +Typ, n =
17; -Typ, n = 17. (
b
) Ax, n = 23;
L.b
CFS, n = 18; +PK, n = 23; -PK, n = 23. (
c
) n = 18.
d
,
Average speed of Ax flies administered with amylase-treated PBS (Ax), amylase-treated
L.b
CFS (+amyl
L.b
CFS), or unaltered
L.b
CFS (-amyl
L.b
CFS). Ax, n = 30; +amyl, n = 17; -
amyl, n = 30.
e
, Average speed of Ax flies or flies mono-associated with
L.b
,
L.p
,
A.
pomorum
(
A.p
), or
E. coli
(
E.c
). Below shows the presence (+) or absence (−) of Xi based
Schretter et al.
Page 18
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
on NCBI Blastn (
xylA
locus) and Blastp (Xi) results. Ax, n = 30;
L.b
, n = 30;
L.p
, n = 29;
A.p
, n = 30;
E.c
, n = 18.
f
, Average speed of Ax and flies mono-associated with either WT
E.c
or single gene knockout strains of
E.c
(∆
tyrA
, ∆
trpC
, ∆
manX
, ∆
treA
, ∆
xylA
). Ax, n =
65;
E.c
; n = 52;
E.c
tyrA
, n = 18;
E.c
trpC
, n = 17;
E.c
manX
, n = 45;
E.c
treA
, n = 46;
E.c
xylA
, n = 20.
g
, Daily activity of Conv, Ax, and Ax virgin female Oregon
R
flies treated
with
L.b
CFS,
L.b
∆xylA
CFS, or Xi* over a 2-day light-dark cycle period each lasting 12
hours, starting at time 0. White boxes represent lights on and gray boxes represent lights off.
Conv, n = 16; Ax, n = 24;
L.b
CFS, n = 19;
L.b
∆xylA
CFS, n = 20; Xi*, n = 8.
h
, Average
speed of Ax and Ax flies treated with
L.b
CFS or Xi*. Ax, n = 16;
L.b
CFS, n = 11; 10
μg/mL Xi*, n = 12; 100 μg/mL Xi*, n = 14.
i
, Lifespan measurements for Ax and Ax treated
with
L.p
CFS,
L.b
CFS, or Xi*. Asterisks above represent significance at the time point
measured by Kruskal-Wallis and Dunn’s post-hoc test. Inset image shows survival at day 7,
error bars represent mean +/− S.E.M. Ax, n = 4 groups;
L.p
CFS, n = 5 groups;
L.b
CFS, n =
5 groups; Xi*, n = 4 groups.
j
, Percentage of apoptotic cells in the intestine of Conv, Ax, and
Ax treated with
L.p
CFS,
L.b
CFS, or Xi*. Bars represent mean +/− S.E.M. Conv, n = 7; Ax,
n = 5;
L.p
CFS, n = 4;
L.b
CFS, n = 6; Xi*, n = 6. Box-and-whisker plots show median and
IQR. *
P
< 0.05, **
P
< 0.01, ***
P
< 0.001, ****
P
< 0.0001. Specific
P
values are in the
Supplementary Material. Kruskal-Wallis and Dunn’s (
a – i
) or Log-rank (
i
) post-hoc tests
were used for statistical analysis. Data are representative of at least 2 independent trials for
each experiment.
Extended Data Figure 6 |. Sleep analysis for mono-colonized flies and flies administered with
bacteria factors.
Schretter et al.
Page 19
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
a
, 24-hour sleep profiles of Conv, Ax,
L.p-
, and
L.b
-colonized virgin female Oregon
R
flies
with the number of sleep bouts in 60 min time window and total sleep in the light or dark
phase. Error bars represent mean +/− S.E.M. n = 8/condition.
b
, 24-hour sleep profiles of
Conv, Ax,
L.b
CFS-,
L.b
∆xylA
CFS-, and Xi* treated Ax virgin female Oregon
R
flies with
the number of sleep bouts in 60 min time window and total sleep in the light or dark phase.
Error bars represent mean +/− S.E.M. Conv, n = 17; Ax, n = 25;
L.b
CFS-, n = 19;
L.b
∆xylA
CFS-, n = 21; Xi*, n = 8. Box-and-whisker plots show median and IQR. *
P
< 0.05. Specific
P
values are in the Supplementary Material. Kruskal-Wallis and Dunn’s post-hoc tests were
used for statistical analysis. Data are representative of at least 2 independent trials for each
experiment.
Extended Data Figure 7 |. Xylose isomerase activity and key carbohydrates are involved in Xi-
mediated changes in locomotion.
Schretter et al.
Page 20
Nature
. Author manuscript; available in PMC 2019 April 24.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript