were gavaged with 60,000 cpm of^14 C-GlcNAc-
labeled PG. Control mice were gavaged with
the equivalent dilution of nonlabeled PG. Col-
lection of blood and organs was performed
4 hours after gavage, as described below.
Radiolabeling of PG in live streptomycin-resistant
Escherichia colistrain FB8-LysA
TheE. coliFB8-LysA strain is unable to convert
meso-DAP to lysine. Thus, radioactivemeso-
DAP added to the growth media is incorpo-
rated specifically in the PG layer without the
need for purification ( 48 , 49 ). A streptomycin-
resistant strain ofE. coliFB8-LysA was gen-
erated by plating the bacteria on LB agar
containing a streptomycin sulfate gradient
(0 to 50 mg/ml; Sigma-Aldrich) followed by in-
cubation at 37°C. The colony exhibiting highest
resistance to streptomycin after 24 hours was
selected and amplified. Subculture of the re-
sultingE. coliFB8-LysA Strrstrain confirmed
that rapid growth was achieved in LB media
containing streptomycin sulfate (50 mg/ml)
and kanamycin (25 mg/ml; Sigma-Aldrich).
For radiolabeling, an exponential phase cul-
ture ofE. coliFB8-LysA Strrwas prepared in
LB medium containing streptomycin (50 mg/
ml) and kanamycin (25 mg/ml) and was used
to inoculate (1:100) 500 ml of prewarmed M9
minimal media supplemented with threonine,
methionine, and lysine (Sigma-Aldrich), each
at 100mg/ml. For radiolabeling, [^3 H]meso-DAP
(50mCi/liter; Moraveck Inc.) was added. A
nonlabeled control culture ofE. coliFB8-LysA
Strrwas prepared in parallel. Cultures were
incubated overnight at 37°C with aeration.
Final OD 600 was ~2.0. Bacteria were washed
five times by centrifugation at 4000g, 4°C
and resuspended with cold PBS. Removal of
free [^3 H]meso-DAP was confirmed by scintil-
lation counting of 1 ml of supernatant. To
favorE. coliFB8-LysA Strrintestinal coloni-
zation, streptomycin sulfate (5 g/liter) was
added to the drinking water of mice, begin-
ning 24 hours before gavage, and maintained
until the end of the experiment. Mice were
gavaged with ~10^10 colony-forming units (CFU)
of [^3 H]meso-DAP–labeledE. coliFB8-LysA Strr
or the unlabeled control strain. Blood and
tissues were collected 24 hours after gavage,
as described below. To assess colonization
efficacy,E. coliCFUs were quantified by serial
dilution of fecal homogenates spotted onto LB
agar containing streptomycin (50 mg/ml) and
kanamycin (25 mg/ml) after overnight incu-
bation at 37°C.
Processing and measurement of^14 C and^3 H
from blood and brain homogenates
Profound anesthesia was induced in mice by
intraperitoneal injection of ketamine (100 mg/
kg; Imalgene1000, Boehringer-Ingelheim) and
xylazine (8 mg/kg; Rompun 2%, Bayer). Blood
collection was performed upon section of the
inferior vena cava and was immediately fol-
lowed by transcardial perfusion with 20 ml of
PBS with a syringe connected to a 26G needle
to eliminate blood from the circulatory sys-
tem. The brain and small intestine were then
extracted. Approximately 8 cm of small intes-
tine proximal to the stomach (“duodenum”in
figs. S3F and S5F) and 8 cm of small intestine
proximal to the caecum (“ileum”in figs. S3F
andS5F)werecollected.Tissueweightswere
determined using a precision balance. Brain,
duodenum, ileum, and blood (200ml) were
transferred into glass vials and dissolved with
2 ml of Solvable (PerkinElmer) overnight at
60°C, followed by color-bleaching by addition
of 500ml(forblood)or200ml(fortissues)of
30% hydrogen peroxide (Millipore). The tubes
were incubated for 30 min at room temper-
ature and then 30 min at 60°C. The solutions
were then transferred to 20-ml HDPE scintil-
lation vials with urea cap containing a poly-
ethylene cone (Duran Wheaton Kimble) and
combined with 10 ml of UltimaGold LTT scintil-
lation cocktail (PerkinElmer). After a minimum
period of 4 hours for temperature equilibration
in the dark, the samples were analyzed using a
Tri-Carb 3110 TR Liquid Scintillation Analyzer
with QuantaSmart TriCarb LCS 3.00 software.(^14) C was measured in the range 0.0 to 156 keV
for 2 min and^3 H was measured in the range
2.0 to 18.6 keV for 5 min, both on the high-
sensitivity setting.
Food consumption
For overnight food consumption, mice were
single-housed around 7 p.m. and a pre-weighed
amount of food pellets was added into the cage.
The next day (or 2 days later for the 40-hour
measurements) around 10 a.m., food pellets
were weighed and the weight difference was
noted as the food was consumed. Automated
food consumption analysis was performed using
an automated food dispenser (FED system)
( 50 ) that was adapted with minor modifica-
tions. Mice were habituated to the FED system
a few days prior to the analysis. During the test
period, mice were individually housed with
the FED device for 3 days. Mouse food con-
sumption was evaluated on day 2. A meal bout
was defined as a continuous period of food
intake (longest nonfeeding interval accepted
during a meal bout <2 min). To evaluate food
consumption after MDP gavage, mice were
individually housed with a FED device in their
home cage for 5 days. Day 0: FED in; day 1:
MDP control gavage; day 4: MDP gavage;
day 5: FED was removed, and mice were
placed together in their original cage. Mice
were gavaged (1 mg in 0.1 ml) at 5 p.m. and
food consumption was analyzed during the
four subsequent hours. Outliers were iden-
tified using Grubbs’method (a= 0.05) and
then removed. Mice that ate less than 4 g in
40 hours were also excluded from the analysis.
Whenever possible, behavioral measures were
performed blinded to mouse genotype.
Nest building
Mice were individually housed around 7 p.m.
The next day at around 11 a.m., mice received
a new pre-weighed cotton nestlet. Four hours
later, any cotton that was not unrolled was
removed and weighed. Mice were placed to-
gether in their original cages after the test.
Whenever possible, behavioral measures were
performed blinded to mouse genotype.
Light-dark box test
A two-compartment box containing a dark
chamber (black walls with upper lid) and a
light chamber (plexiglass walls, no upper lid)
wasused.Thechamberswereconnectedbya
10-cm^2 door in the middle of the wall. Animals
were placed in one corner of the light chamber
facing the wall and were allowed to freely
explore for 6 min. The time spent in the dark
chamber and the number of transitions from
one chamber to the other were annotated
manually afterward by looking at recorded
sessions.
Temperature analysis
Telemetric temperature probes (Anipill pills,
BodyCap medical) were implanted subcuta-
neously. After general gas anesthesia (isoflur-
ane 3% for induction, followed by 1.5% for
maintenance; diluted in 98.5% oxygen at a
flow rate of 1 liter/min) and analgesia (bupre-
norphine, Buprecare; 0.1 mg/kg, subcutaneous),
the skin in the scapular region was shaved,
sterilized, locally anesthetized (lidocaine, Xylovet,
2%), and incised for probe insertion in a sub-
cutaneous chamber. After probe insertion and
skin suturing, mice were given buprenorphine
every 24 hours for the next 48 hours. Mice
were allowed to recover for a week before any
data were collected. Temperature was mea-
sured every 15 min.
Drugs used
b 3 -Adrenergic agonist (CL 316,243; Sigma-
Aldrich) was diluted in sterile water and in-
jected intraperitoneally (500 ng per animal).
MDP (L-D isomer, active; tlrl-mdp) and MDP
control (MDPctr; L-L isomer, inactive; tlrl-
mdpcl) were purchased from InvivoGen. MDP
and MDPctr were diluted in LAL-water
(10 mg/ml), and 1 mg of either MDP of MDPctr
wasadministeredbyoralgavagewithasterile
gavage needle. Mice were systematically hab-
ituated to oral gavage prior to the test.
Glucose and insulin measurements
Mice were fasted for 6 hours. Blood samples
from the tail vein were collected using micro-
vette tubes (Sarstedt 16444). For blood glucose
measurements, samples were immediately
analyzed with a glucose meter (Accu Check,
Gabanyiet al.,Science 376 , eabj3986 (2022) 15 April 2022 9 of 12
RESEARCH | RESEARCH ARTICLE