Methods
Descriptions herein refer to a shift to acetate, but the same methods
also apply to other shifts. No statistical methods were used to prede-
termine sample size. The experiments were not randomized, and the
investigators were not blinded to allocation during experiments and
outcome assessment.
Strain construction
All E. coli strains used here were derived from E. coli strain K-12 NCM3722
(ref.^28 ). B. thetaiotaomicron was obtained from the American Type
Culture Collection (ATCC catalogue number 29148).
Ptet–aceB (NQ1350) and Ptet–ppsA (NQ1357) strains. The DNA
region containing the kmr gene, rrnBT and the promoter Ptet of
the pKDT Ptet plasmid^29 was amplified by polymerase chain reac-
tion (PCR) with upstream and downstream primers ptet–aceB–in-
sert (forward)/ptet–aceB–insert (reverse) and Ptet–ppsA–insert
(forward)/Ptet–ppsA-insert (reverse), respectively, and then in-
tegrated into the chromosome of E. coli strain NQ309 to replace
the chromosomal promoters of aceB and ppsA (each from −150
base pairs (bp) to −1 bp relative to the transcriptional start site)^30.
Each of the Ptet–promoter substitutions was then transferred to
strain NQ1358 (NCM3722 ycaD: Ptet–tetR Δkmr)^29 backgrounds by
phage P1 vir-mediated transduction, resulting in the strains NQ1350
and NQ1357.
cra deletion strain (NQ1077). The Δcra deletion allele in strain LJ2801
(E. coli Genetic Stock Center), in which a kmr gene is substituted for cra,
was transferred to wild-type strain NCM3722, resulting in strain NQ1077.
PykF (NQ1543), PfkA (NQ1544) and ArgA (NQ1545) overexpres-
sion strains. Overexpression pNT3 plasmids (from the library in
ref.^31 ) expressing the genes pykF, pfkA or argA from plasmid Ptac were
purified and transformed into wild-type strain NCM3722, resulting in
the strains NQ1543, NQ1544 and NQ1545, respectively.
glpK22 mutant strain (NQ898). To create a strain that grows faster on
glycerol, we replaced the glpK gene in strain NCM3722 with the glpK22
variant^17 through two P1 transduction steps. First, the pfkA::km marker
was transferred into strain NCM3722 using phage P1 vir, prepared from
the Keio collection^32. The resulting strain (NQ632) from the transduc-
tion cannot use mannitol as its sole carbon source. Second, phage P1
vir prepared from strain CGSC5511 (Lin-43)^16 containing glpK22 was
transfected into NQ632. Selecting a colony that grew on mannitol mini-
mum medium yielded a strain, NQ898, containing the glpK22 mutation
in an NCM3722 background.
YCE44 strain. The recipient strain NCM3722 was used for P1 transduc-
tion^33 with P1 lysate prepared from the Keio collection^32 to create the
fliC::Kan mutant. This mutant was then transformed with the Pcp20
plasmid^30 to flip out the kanamycin marker. The resulting strain was
then used as a recipient strain for P1 transduction with BO37 (ref.^34 ) P1
lysate to create the final target strain, YCE44 (NCM3722, fliC::FRT–FRT,
glmS::PRNAI–mCherry1–11–mKate–T1 terminator–FRT Kan FRT::pstS).
The donor strain BO37 was provided by the Paulsson Laboratory^34.
Strains used herein. Except for wild-type strain BW25113, used as a
control, all strains herein were derived from E. coli K-12 strain NCM3722
(refs.^28 ,^35 ,^36 ), provided by the S. Kustu laboratory. See Supplementary
Table S1 for a summary of strains.
Growth of bacterial cultures
Growth media. Unless otherwise indicated, we used N+C+ minimal me-
dium^37 , which contains K 2 SO 4 (1 g), K 2 HPO 4 .3H 2 O (17.7 g), KH 2 PO 4 (4.7 g),
MgSO 4 .7H 2 O (0.1 g) and NaCl (2.5 g) in one litre, and is supplemented
with 20 mM NH 4 Cl and specified carbon sources. Carbon-source con-
centrations were based on the number of carbon atoms in the molecule:
20 mM for C 6 carbons, 30 mM for C 4 carbons and 40 mM for C 3 carbons.
The base minimal medium used for the anaerobic growth of E. coli
strain NCM3722 consisted of KH 2 PO 4 (2 g), K 2 HPO 4 (14.8 g), NaCl (0.58 g),
NH 4 Cl (0.54 g) and Na 2 SO 4 (0.07 g), and 1,000× mineral solution (1 ml)
per litre. One litre of the ×1,000 mineral solution contained MgCl 2
(60 g), CaCl 2 (5.5 g), FeSO 4 .7H 2 O (5.5 g), MnCl 2 (19.7 mg), CoCl 2 (23.8 mg),
Ni 2 SO 4 (26.2 mg), CuCl 2 (15.9 mg), (NH 4 ) 2 MoO 4 (23.5 mg), SeO 2 (11 mg),
ZnSO 4 (28.7 mg) and H 3 BO 4 (6.2 mg) dissolved in 100 mM HCl. For con-
sistency of comparisons, the same medium was used for the aerobic
growth of strain NCM3722 (Extended Data Fig. 9). Carbon sources were
added as indicated.
The medium used for the anaerobic growth of B. thetaiotaomicron
was the same as that used for the anaerobic growth of E. coli, but also
included 2 mg cyanocobalamin, 2 mg haemin and 0.6 g cysteine per
litre. For anoxic media, Hungate tubes (16 mm × 125 mm) filled with 7 ml
medium were shaken at 270 rpm under a 7% CO 2 , 93% N 2 atmosphere
pressurized to 1.5 atm for 75 min. Cultures were transferred anoxically
into Hungate tubes with disposable syringes.Growth measurements. Batch culture aerobic growth was per-
formed in a 37 °C water-bath shaker or air incubator shaking at
250 rpm. The culture volume was at most 10 ml in 25 mm × 150 mm
test tubes. For seed culture, one colony from fresh LB agar plates
was inoculated into liquid LB and cultured at 37 °C with shaking.
Cells were then diluted into minimal medium and cultured in a
37 °C water-bath shaker overnight (preculture). The overnight pre-
culture was allowed to grow for at least three doublings. Cells from
the overnight preculture were then diluted to OD 600 = 0.005–0.025
in identical prewarmed minimal medium, and cultured in 37 °C wa-
ter bath shaker (experimental culture). At every half-doubling, we
collected 200 μl of cell culture in a Sterna submicrometre cuvette for
OD 600 measurement using a thermal spectrophotometer, after allow-
ing at least four generations of growth. The time taken for each sample
collection was less than 30 s and had no measureable effect on cell
growth.
Anaerobic growth was performed similarly with a few exceptions. All
growth for B. thetaiotaomicron was carried out in Hungate tubes. For
seed culture, a single colony from Wilkens–Chalgren agar plates was
inoculated into anoxic Hungate tubes filled with 7 ml Wilkens–Chal-
gren broth and incubated at 37 °C with shaking. Cells were then diluted
roughly 300-fold into preculture medium to grow overnight. The next
day, cells were diluted to OD 600 = 0.01–0.025 for experimental cultures
in the same medium as the preculture. OD 600 measurements of cells in
Hungate tubes were made with a Thermo Genesys 20 modified to hold
Hungate tubes in place of a cuvette. To maintain the temperature of the
culture tube, we removed tubes from the water-bath shakers to measure
OD 600 and returned them within 30 s. The OD 600 measured through the
Hungate tubes was equivalent to the OD 600 measured through a cuvette
in the range of at least 0.04–0.5.
Anaerobic growth of E. coli strain NCM3722 was measured similarly
to that of B. thetaiotaomicron, except that seed cultures were
performed aerobically in LB broth before being diluted roughly
300-fold into anoxic Hungate tubes for overnight preculture with
the same media as the experimental culture. Cells were again diluted
into fresh Hungate tubes with OD 600 = 0.01–0.025 for experimental
culture, and growth was measured with the modified Thermo
Genesys 20.pH changes. Because anaerobic growth of E. coli and B. thetaiotaomi-
cron involves copious acid production, the pH of cultures was moni-
tored. Typical pH changes for the anaerobic growth of NCM3722 were
from 7.2 (fresh anoxic medium) to 6.7 (at an OD 600 of roughly 0.4). For