Access to this full-text is provided by American Society for Microbiology.
Content available from mBio®
This content is subject to copyright. Terms and conditions apply.
A Solution to Antifolate Resistance in Group B Streptococcus:
Untargeted Metabolomics Identifies Human Milk
Oligosaccharide-Induced Perturbations That Result in
Potentiation of Trimethoprim
Schuyler A. Chambers,
a
Rebecca E. Moore,
a
Kelly M. Craft,
a
*Harrison C. Thomas,
a
Rishub Das,
a
Shannon D. Manning,
e
Simona G. Codreanu,
a,c
Stacy D. Sherrod,
a,c
David M. Aronoff,
b,f
John A. McLean,
a,c
Jennifer A. Gaddy,
b,d,f
Steven D. Townsend
a
a
Department of Chemistry, Vanderbilt University, Nashville, Tennessee, USA
b
Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee, USA
c
Center for Innovative Technology, Nashville, Tennessee, USA
d
Department of Veterans Affairs, Tennessee Valley Healthcare Systems, Nashville, Tennessee, USA
e
Department of Microbiology and Molecular Genetics, Michigan State University, East Lansing, Michigan, USA
f
Department of Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center, Nashville, Tennessee, USA
Schuyler A. Chambers and Rebecca E. Moore contributed equally to this work. Author order was determined alphabetically.
ABSTRACT Adjuvants can be used to potentiate the function of antibiotics whose
efficacy has been reduced by acquired or intrinsic resistance. In the present study,
we discovered that human milk oligosaccharides (HMOs) sensitize strains of group B
Streptococcus (GBS) to trimethoprim (TMP), an antibiotic to which GBS is intrinsically
resistant. Reductions in the MIC of TMP reached as high as 512-fold across a diverse
panel of isolates. To better understand HMOs’ mechanism of action, we character-
ized the metabolic response of GBS to HMO treatment using ultrahigh-performance
liquid chromatography–high-resolution tandem mass spectrometry (UPLC-HRMS/MS)
analysis. These data showed that when challenged by HMOs, GBS undergoes signifi-
cant perturbations in metabolic pathways related to the biosynthesis and incorpora-
tion of macromolecules involved in membrane construction. This study represents
reports the metabolic characterization of a cell that is perturbed by HMOs.
IMPORTANCE Group B Streptococcus is an important human pathogen that causes
serious infections during pregnancy which can lead to chorioamnionitis, funisitis,
premature rupture of gestational membranes, preterm birth, neonatal sepsis, and
death. GBS is evolving antimicrobial resistance mechanisms, and the work presented
in this paper provides evidence that prebiotics such as human milk oligosaccharides
can act as adjuvants to restore the utility of antibiotics.
KEYWORDS group B Streptococcus, human milk oligosaccharides, resistance,
adjuvants, antifolate drugs
The development of antibiotics is arguably one of the most important advances in
modern medicine. Antibiotics can be organized according to the cellular compo-
nent or system they engage and whether they inhibit cell growth (bacteriostatic) or
induce cell death (bactericidal). Although antibiotics that target cellular viability are
effective, these agents impose selective pressures that foster the evolution of resistant
phenotypes (1). Combination therapy has emerged as a powerful solution to resistance
issues that plague monotherapy (2). This approach, which involves codosing an anti-
Citation Chambers SA, Moore RE, Craft KM,
Thomas HC, Das R, Manning SD, Codreanu SG,
Sherrod SD, Aronoff DM, McLean JA, Gaddy JA,
Townsend SD. 2020. A solution to antifolate
resistance in group B Streptococcus: untargeted
metabolomics identifies human milk
oligosaccharide-induced perturbations that
result in potentiation of trimethoprim. mBio
11:e00076-20. https://doi.org/10.1128/mBio
.00076-20.
Editor Jimmy D. Ballard, University of
Oklahoma Health Sciences Center
Copyright © 2020 Chambers et al. This is an
open-access article distributed under the terms
of the Creative Commons Attribution 4.0
International license.
Address correspondence to Jennifer A. Gaddy,
jennifer.a.gaddy@vanderbilt.edu, or Steven D.
Townsend, steven.d.townsend@vanderbilt.edu.
*Present address: Kelly M. Craft, Department of
Chemistry & Chemical Biology, Harvard
University, Cambridge, Massachusetts, USA.
This article is a direct contribution from David
M. Aronoff, a Fellow of the American Academy
of Microbiology, who arranged for and secured
reviews by Christian Melander, University of
Notre Dame; Amit Basu, Brown University; and
Xin Zhang, The Pennsylvania State University.
Received 15 January 2020
Accepted 30 January 2020
Published
RESEARCH ARTICLE
Therapeutics and Prevention
crossm
March/April 2020 Volume 11 Issue 2 e00076-20 ®mbio.asm.org 1
17 March 2020
biotic with an adjuvant that potentiates its function or a second antibiotic with a
different target, can improve efficacy and suppress resistance evolution (2–7).
One bacterial pathogen group that showcases multidrug resistance is group B
Streptococcus (GBS) (8). GBS is a leading cause of neonatal sepsis, pneumonia, and
meningitis (9–14). Recent data also suggest that GBS is a frequent cause of chorioam-
nionitis, endometritis, pneumonia, and urosepsis in adults with underlying medical
conditions (i.e., diabetes mellitus or immunosuppression) (15–19). As these patterns of
pathogenesis suggest, GBS is considered a saprophytic organism, i.e., invasive GBS
disease is most commonly observed in weakened hosts.
Treatment of GBS disease relies primarily on penicillin and ampicillin, followed by
first-generation cephalosporins and vancomycin (20). Alternative antibiotics, such as
lincosamides, are used for patients with

-lactam allergies. Due to resistance evolution,
macrolides, aminoglycosides, and tetracyclines are no longer clinically efficacious (21–
24). While our group and others have observed that GBS is resistant to a wide range of
antibiotics, GBS resistance remains poorly characterized and is a frontier of concern in
the clinic.
In the early stages of this program, we hypothesized that human milk oligosaccha-
rides (HMOs) possess antimicrobial and antivirulence properties (25). Indeed, we dis-
covered that heterogeneous HMOs modulate growth and biofilm production for a
number of bacterial pathogens (26, 27). We also determined the identities of several
single-entity HMOs with potent antimicrobial activity against GBS (28–31). In addition
to structure-activity relationship (SAR) studies, we found that HMOs potentiate the
activity of select intracellular targeting antibiotics (32). This included three antibiotics to
which GBS has evolved resistance, aminoglycosides, macrolides, and tetracyclines
(33–42). At their 50% inhibitory concentration (IC
50
), HMO extracts reduced the MICs of
certain intracellular targeting antibiotics up to 32-fold (Table 1). Interestingly, HMO
treatment did not affect

-lactam or glycopeptide activity; antibiotics that interfere
with cell wall synthesis. Based on this activity pattern, we hypothesized that HMOs
function by increasing membrane permeability, which would be an unprecedented
mode of action in GBS. This hypothesis was validated when HMOs were found to
increase membrane permeability by ca. 30% using a LIVE/DEAD BacLight assay (32).
Based on their ability to increase cellular permeability, a second-generation combi-
nation study was initiated to further characterize HMO enhancement of intracellular
targeting antibiotics in the GBS model. We took particular interest in trimethoprim
(TMP), an antifolate used in the treatment of enteric, respiratory, skin, and urinary tract
infections (43). Mechanistically, TMP is a bacteriostatic agent that inhibits dihydrofolate
reductase (DHFR), an enzyme within the folate biosynthesis pathway (44). Importantly,
interference with this pathway inhibits pyrimidine and purine biosynthesis, with down-
stream effects on bacterial DNA synthesis. Furthermore, a wide range of streptococcal
strains, including GBS, are intrinsically resistant to TMP (45–51). Resistance is typically
mediated by one of the following five mechanisms: (i) poor membrane permeability, (ii)
an impervious DHFR, (iii) mutations in the inherent DHFR, (iv) upregulation of gene
expression or gene duplication to increase DHFR production, and (v) horizontal transfer
of dfr genes that encode resistant DHFRs (45). We hypothesized that if TMP has
difficulty gaining penetrance into the GBS cell, HMOs could be used to sensitize GBS to
TABLE 1 Established patterns of HMO potentiation of antibiotic activity
Antibiotic in THB medium (strain)
MIC (
g/ml)
Fold reductionOverall With 5.0 mg/ml HMOs
Penicillin (CNCTC 10/84) 0.03 0.015 2
Vancomycin (CNCTC 10/84) 2 1 2
Clindamycin (GB2) 0.0312 0.0078 4
Gentamicin (GB590) 16 1 16
Erythromycin (GB590) 0.0312 0.001 32
Minocycline (CNCTC 10/84) 0.0625 0.0019 32
Chambers et al. ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 2
TMP. Described herein are the results of testing this hypothesis using heterogeneous
HMO extracts. To further evaluate the mechanism of HMO sensitization, ultrahigh-
performance liquid chromatography–high-resolution tandem mass spectrometry anal-
ysis (UPLC-HRMS/MS) was used to characterize the immediate metabolic response of
GBS to HMO-induced perturbations.
RESULTS AND DISCUSSION
HMOs demonstrate synergy with TMP against group B Streptococcus.HMOs
were isolated from donor breast milk and pooled to create two HMO cocktails; the first
(HMO-1) used milk from 10 donors, while the second (HMO-2) used milk from 7 donors.
Prior to potentiation studies, the MIC of the HMOs and TMP were determined sepa-
rately in each strain of GBS grown in Todd-Hewitt broth (THB) using a broth microdi-
lution assay (Table 2). HMOs were assayed against five strains of GBS of various
serotypes to determine the strain specificity of antibiotic potentiation. The strains
selected are all clinical isolates. CNCTC 10/84 is commercially available (52). Isolates
GB00590, GB00002, GB00651, and GB00083 were recovered from colonized pregnant
women (53, 54). GBS strains are divided into 10 serotypes (1a, 1b, and II to IX) based on
a serological reaction against their capsular polysaccharides (55). GB2, GB590, and
CNCTC 10/84 are serotypes Ia, III, and V, respectively. These three serotypes are the
most common isolates associated with early-onset disease in the United States and
together account for over 80% of all isolates (56). GB651 and GB83 are serotypes Ib and
IV, respectively. Globally, the five strains represent 85% of all isolate serotypes (57).
HMOs were dosed at their 25% inhibitory concentrations (IC
25
s) in CNCTC 10/84 and
GB2. The growth of the remaining GBS strains were so rapidly affected by HMO
treatment that subsequent IC
50
curve fitting yielded immeasurable confidence limits.
For these strains, the IC
25
from a similar strain having a superior fit dose-response curve
was used (see Fig. S1A to E in the supplemental material). In each strain, the MIC of TMP
was 512
g/ml or higher (Table 2). In GB2, a 512-fold reduction in MIC was observed.
A 256-fold reduction in MIC was observed in CNCTC 10/84. For the three remaining
strains (GB651, GB83, and GB590), the fold reductions in MIC were 16, 16, and 64,
respectively. The potentiation patterns described above are remarkable for several
reasons. First, they represent the greatest magnitude of antibiotic enhancement that
we have observed. Second, GBS is not susceptible to antifolate antibiotics, so the
chemotherapeutic regime is effective at sensitizing GBS to TMP.
Next, checkerboard assays were conducted with GB2 and GB590, stains for which
strong and weak potentiation of TMP were observed, respectively, to determine if the
HMO-TMP combination was synergistic or additive in nature (Fig. S2A and B). Synergy
is measured using the fractional inhibitory concentration (FIC) index value and is
defined when the FIC is ⱕ0.5 for each combination of compounds. It was demonstrated
that in GB590, synergy was achieved when dosing HMOs from 1.28 to 2.56 mg/ml in
combination with TMP dosed at 8 to 128
g/ml (⌺FIC values, 0.281 to 0.508). In GB2,
the combination was synergistic with treatment of HMOs between 0.64 and 1.28 mg/ml
in conjunction with TMP at 4 to 32
g/ml (⌺FIC values, 0.281 to 0.508). These assays
firmly demonstrate the HMO-TMP combination to be synergistic, and they characterize
the dosing windows required to achieve this effect.
TABLE 2 HMO potentiation of TMP
Strain in THB medium
MIC (
g/ml) for:
Fold reductionHMOs TMP
TMP with
1.42 mg/ml HMO
CNCTC 10/84 5.12
a
⬎1,024 8
a
ⱖ256
GB2 2.56
a
1,024 2
a
512
GB590 5.12
a
⬎1,024 32
a
ⱖ64
GB651 5.12
b
512 32
b
16
GB83 5.12
b
⬎1,024 128
b
ⱖ16
a
HMO-1.
b
HMO-2.
HMO Changes Enhance Antibiotics against GBS ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 3
After evaluating the level of synergy in the cocktail, we conducted an experiment to
validate whether growth inhibition was due to HMO enablement of cognate engage-
ment of TMP with the folate pathway. By inhibiting folate biosynthesis, antifolates
inhibit the de novo biosynthesis of thymidylate and purine nucleotides. However, in
addition to de novo synthesis, cells can produce these nucleotides via salvage pathways
that use free thymidine or purine bases as precursors for the corresponding nucleo-
tides. Thus, we hypothesized that if HMOs facilitate TMP inhibition of the de novo
synthesis pathway, the addition of the preformed nucleotide precursors thymidine or
hypoxanthine would dampen the growth inhibitory capabilities of the HMO-TMP
combination, as thymidine and hypoxanthine serve as precursors in the pyrimidine and
purine salvage pathways, respectively (58).
In the experiment, we evaluated the MIC of the HMO-TMP cocktail in the presence
of thymidine (Table 3). The experiment was conducted in strains GB2 and GB590.
Against GB2, HMO supplementation decreased the MIC of TMP from 1,024
g/ml to
2
g/ml (512-fold reduction). Against GB590, the MIC of TMP was reduced from
⬎1,024
g/ml to 32
g/ml (at least a 64-fold reduction) (Table 2). In the presence of
added thymidine, the MICs of TMP in the HMO-TMP combination increased 8-fold to
16
g/ml in GB2 and 4-fold to 128
g/ml in GB590. These results support the hypoth-
esis that supplemental thymidine mitigates the effects of the HMO-TMP combination
and is able to partially salvage the folate biosynthetic pathway. Importantly, the MIC of
the HMO cocktail individually did not change in the presence of added thymidine. This
indicates that the folate pathway is not a target for HMOs. We therefore conclude that,
in the presence of HMOs, TMP gains penetrance into the group B streptococcal cell and
exhibits on-target inhibition of the folate cycle.
The final assay in this study was a comparison of the HMO-TMP combination with
the clinically useful TMP-sulfadiazine (SDZ) combination (Table S2). In both GB590 and
GB2, the TMP-sulfadiazine combination was largely ineffective, with an MIC of ⱖ512
g/ml, while the HMO-TMP combination sees the potentiation profile described above
(Table 2). This result demonstrates that while antifolate-based antibiotic combination
treatments remain largely ineffective against GBS, the HMO-TMP combination is oper-
ative. This insight offers new consideration for the use of existing combination thera-
pies in patient care.
Characterizing the HMO mode of action using untargeted metabolomics. The
mode of action of an antimicrobial agent cannot accurately be described in terms of a
single static target; rather, the complete induced response must be evaluated. In
theory, a single chemotherapeutic could have a wide range of direct and indirect
targets, simultaneously interfering with multiple enzymes or pathways. Accordingly,
the final stage of the study focused on utilizing global, untargeted metabolomic
analysis to characterize the early response of GBS to HMO-mediated perturbations. The
analysis described provides HMO-mediated perturbations. For this study, strain GB2
was used, as it was most susceptible to treatment with HMOs (MIC, 2.56
g/ml). Two
groups were analyzed and compared, with the first being an untreated GB2 control and
the second being GB2 treated with HMOs dosed at 1 mg/ml. This concentration
promoted cellular death (ca. 20 to 40%) compared to the untreated controls but also
provided enough remaining cellular mass for analysis (minimum, 200
g).
Our experimental design from sample collection through data analysis is depicted in
Fig. 1A. The annotated and statistically significant metabolites observed in the exper-
iment (Fig. 1B) were subjected to traditional pathway analysis (Fig. S3). The results
TABLE 3 HMO potentiation of TMP in the presence of thymidine
Strain in THB medium plus
20
g/ml thymidine
MIC (
g/ml) for:
Fold reductionTMP
TMP with HMO-1
(dose [mg/ml])
GB2 1,024 16 (1.43) 64
GB590 ⬎1,024 128 (1.42) 8
Chambers et al. ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 4
showed the most statistically perturbed metabolic pathways to be linoleic acid metab-
olism, sphingolipid metabolism, glycerophospholipid metabolism, and pyrimidine me-
tabolism (Fig. 1C and D). Characterized below are perturbations to both linoleic acid
and glycerophospholipid metabolism (Fig. 2 and 3). We focus on these pathways, as
each is critical to membrane formation and structural integrity, i.e., each pathway
contributes to the synthesis of membrane-bound macromolecules and their corre-
sponding precursors (59).
Based on statistical significance, linoleic acid metabolism is the metabolic pathway
most impacted when GBS is exposed to HMOs (Fig. 2A and S4). Linoleic acid meta-
bolites play a critical role in both cellular signaling and the stress response. Each is also
critical to proper membrane construction (60, 61). All identified linoleic acid metabolites
were accumulated in the HMO-treated population, with several metabolites having a
⬎100-fold increase from the untreated controls (Fig. 2B). Two epoxyoctadecanoic acid
metabolites were of particular interest, epoxyoctadecanoic acids (EpOMEs) and dihy-
droxyoctadecanoic acids (DiHOMEs). Accumulation of these metabolites is linked to
changes in Na
⫹
and K
⫹
ion channels and, subsequently, cell membrane fluidity (62). In
addition to the roles of EpOMEs and DiHOMEs in cell membrane construction, linoleic
acid metabolites have been shown to have a critical role in cellular signaling and the
stress response.
Glycerophospholipid metabolism was also significantly impacted, and in general, we
observed an accumulation of these metabolites compared to the control (Fig. 3A and
S5). Glycerophospholipid metabolites were observed with significant fold changes
compared to the control. For example, PE(17:0/0:0), PE(P-16:0/0:0), and PE(19:1/12:0),
known degradation products of phosphatidylethanolamine (PE), one of three major
components of the cellular membrane, were observed to have up to a 50-fold change
FIG 1 Workflow and pathway analysis using global, untargeted metabolomics data analysis. (A) Overview of global,
untargeted metabolomic workflow. (B) Global output of identified metabolites from RPLC and HILIC methods and
subsequent filtering for significance according to a Pvalue of ⱕ0.05 and fold change of ⱖ|2|. (C) Table output of metabolic
pathway enrichment analysis. The number of total metabolites in the pathway, the number of hits, and the Pvalue were
calculated using MetaboAnalyst 4.0. CoA, coenzyme A. (D) Metabolomic pathway analysis visualization. Shown is a
graphical representation analysis using the statistically significant metabolite compounds (Pⱕ0.05; fold change, ⱖ|2|)
annotated from RPLC and HILIC analyses. Matched pathways were arranged by Pvalues (from pathway enrichment
analysis) on the yaxis, and pathway impact values (from pathway topology analysis) are shown on the xaxis; node color
is based on pathway Pvalue, and node radius is determined based on pathway impact values; individual nodes represent
individual pathways.
HMO Changes Enhance Antibiotics against GBS ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 5
increase compared to the untreated control (63). This observed accumulation of lipid
metabolites indicates an increased rate of breakdown of critical cell membrane com-
ponents when bacteria are dosed with HMOs. In fact, this type of metabolite accumu-
lation has been previously observed in other model organisms when exposed to
antibiotics (64–67).
In addition to dysregulating pathways directly related to the production of mem-
brane components, HMOs perturbed additional pathways related to essential cellular
function. These include, for example, increased accumulation of purine and pyrimidine
nucleotide precursors. An inability to synthesize nucleotides would lead to perturba-
tions in DNA and RNA synthesis. HMO treatment also led to an accumulation of
metabolites in the cysteine and methionine metabolic pathway. Cysteine biosynthesis
is the primary pathway for incorporating sulfur into cellular components. In addition to
serving as a precursor of methionine, cysteine is also the direct precursor to biotin,
thiamine, and lipoic acid. Methionine is an essential amino acid in all organisms, as it
is both proteinogenic and a component of the cofactor S-adenosyl methionine. Inter-
estingly, sphingolipid metabolism was observed to be significantly perturbed upon
HMO treatment. While GBS does not synthesize sphingolipids directly, it does rely on
the host for access to these compounds for the biosynthesis of cell membrane
components. This metabolic change could suggest that HMO treatment has an impact
on host-microbe interactions (68, 69). Finally, several cell wall synthesis-associated
metabolites were also identified as being accumulated in the treated sample, but a
higher experimental mass range would be needed to identify a more significant
FIG 2 Linoleic acid-associated metabolite identification and statistical representation. (A) Heat map visualization of the
significantly differently regulated linoleic acid metabolic pathway upon HMO treatment. Linoleic acid metabolism
members shown here were detected by RPLC-positive LC-MS/MS analysis. Samples (columns) and metabolite compounds
(rows) were processed using Euclidean average clustering via MetaboAnalyst 4.0. The heat map was generated for
Pareto-scaled, log-transformed data, and colors are displayed by relative abundance, ranging from low (blue) to high (red),
as shown in the legend. (B) Corresponding data table of linoleic acid metabolites, where the asterisk (*) denotes
significance with a Pvalue of ⱕ0.05 and fold change of ⱖ|2|. ODE, octadecadienoic acid; 13-HOTE, 13-hydroxyoctadeca-
9,11,15-trienoic acid; 9,10-DiHOME, 9,10-dihydroxyoctadec-12-enoic acid; 9,12,13-TriHOME, 9,12,13-trihydroxyoctadecanoic
acid; 13S-HODE, 13S-octadecadienoic acid; 9(10)-EpHOME, 9(10)-epoxyhydroxyoctadecanoic acid; ID, identifier.
Chambers et al. ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 6
amount of these precursors and better identify a global trend. The variety of metabolic
pathways perturbed not only illuminates the synergistic nature of the TMP and HMO
combination treatment but demonstrates that antimicrobial agents have broad effects
on cellular biology (Fig. S6).
In summary, this study demonstrates that HMOs potentiate trimethoprim in group
BStreptococcus bacteria, with a synergistic profile that spans the most prevalent
serotypes worldwide. This potentiation profile makes antifolate antibiotics of potential
use in an organism where they have been long considered resistant. Moreover, this
combination could represent an alternative treatment for GBS-positive mothers with
penicillin allergies given the high rates of resistance observed with alternative agents.
To characterize the mechanism of HMO-mediated antimicrobial activity, we have
presented the first global, untargeted metabolomic analysis of HMO-mediated pertur-
bations within any cell type and have shown significant impacts on cell membrane-
affiliated macromolecules. While a number of high-throughput methods have been
performed to elucidate an antibiotic’s mode of action (e.g., cytological profiling, genetic
screens, or gene expression and proteomic profiling), direct experimental evidence that
FIG 3 Glycerophospholipid-associated metabolite identification and statistical representation. (A) Heat map visualization
of the significantly differently regulated glycerophospholipid metabolism pathway upon HMO treatment. Glycerophos-
pholipid members shown here were detected by HILIC-positive LC-MS/MS analysis. Samples (columns) and metabolite
compounds (rows) were processed using Euclidean average clustering via MetaboAnalyst 4.0. The heat map was generated
for Pareto-scaled, log-transformed data, and colors are displayed by relative abundance, ranging from low (blue) to high
(red), as shown in the legend. SM, sphingomyelin; PC, phosphocholine; PI, phosphoinositol; DG, diglyceride; LysoPC,
lysophosphatidylcholine. (B) Corresponding data table of glycerophospholipid metabolites, where the asterisk (*) denotes
significance with a Pvalue of ⱕ0.05 and fold change of ⱖ|2|.
HMO Changes Enhance Antibiotics against GBS ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 7
rapid metabolic changes are causal in facilitating the microbial response to antibiotics
is lacking. In fact, little is described about the downstream phenotypic changes induced
by antimicrobial agents. In the future, metabolomic experiments will be employed to
better describe the phenotypic response of GBS to HMO-induced effects. We hypoth-
esize that metabolomics will enable the characterization of the indirect connections
critical to HMOs’ mechanism of action. Since metabolites present the final phenotypic
manifestation of an organism and the final endpoint of biochemical reactions reflects
the interplay between gene expression, protein function, and the environment, we
argue that further metabolomics analyses are necessary to understand the HMO mode
of action (70, 71).
MATERIALS AND METHODS
Antibiotics and additional chemicals. Trimethoprim lactate 98% was purchased from Alfa Aesar.

-Galactosidase from Kluyveromyces lactis, at 2,600 units/g, was purchased from Sigma-Aldrich. Aceto-
nitrile (ACN; catalog no. A955-1), methanol (MeOH; catalog no. A456-1), and water (catalog no. W6-1,
liquid chromatography-mass spectrometry [LC-MS] grade; Optima) for the mass spectrometry analysis
were obtained from Thermo Fisher Scientific.
HMO isolation. Human milk was obtained from 17 healthy, lactating women between 3 days and
3 months postpartum and stored between ⫺80 and –20°C. Deidentified milk was provided by Jörn-
Hendrik Weitkamp from the Vanderbilt Department of Pediatrics, under a collection protocol approved
by the Vanderbilt University institutional review board (IRB no. 100897), or from Medolac. Milk samples
were thawed and then centrifuged for 45 min. Following centrifugation, the resultant top lipid layer was
removed. The proteins were then removed by diluting the remaining sample with roughly 1:1 (vol/vol)
180 or 200 proof ethanol, chilling the sample briefly, and centrifuging for 45 min, followed by removal
of the resulting HMO-containing supernatant. Following concentration of the supernatant in vacuo, the
HMO-containing extract was dissolved in 0.2 M phosphate buffer (pH 6.5) and heated to 37°C.

-Galactosidase from Kluyveromyces lactis was added, and the reaction mixture was stirred until lactose
hydrolysis was complete. The reaction mixture was diluted with roughly 1:0.5 (vol/vol) 180 or 200 proof
ethanol, chilled briefly, and then centrifuged for 30 min. The supernatant was removed and concentrated
in vacuo, and the remaining salts, glucose, and galactose were separated from the oligosaccharides using
size exclusion chromatography with P-2 gel (H
2
O eluent). The oligosaccharides were then dried by
lyophilization. Correspondingly, HMO isolates from donors were combined and solubilized in water to
reach a final concentration of 102.6 mg/ml.
Bacterial strains and culture conditions. The bacterial strains are shown in Table S1. All strains were
grown on tryptic soy agar plates supplemented with 5% sheep blood (blood agar plates) at 37°C in
ambient air overnight. All strains were subcultured from blood agar plates into 5 ml of Todd-Hewitt broth
(THB) and incubated under shaking conditions at 180 rpm at 37°C overnight. Following overnight
incubation, bacterial density was quantified through absorbance readings at 600 nm (OD
600
) using a
Promega GloMax-Multi detection system plate reader. Bacterial numbers were determined using the
predetermined coefficient of an OD
600
of 1, equal to 10
9
CFU/ml.
Broth microdilution method for determination of MICs. All strains were grown overnight as
described above and used to inoculate fresh THB or THB plus 20
g/ml thymidine to achieve 5 ⫻10
5
CFU/ml. To 96-well tissue culture-treated, sterile polystyrene plates was added the inoculated medium
in the presence of increasing concentrations of antibiotic or HMO cocktail to achieve a final volume of
100
l per well. Bacteria grown in medium in the absence of any compounds served as the controls. The
plates were incubated under static conditions at 37°C in ambient air for 24 h. Bacterial growth was
quantified through absorbance readings (OD
600
). The MICs were assigned at the lowest concentration of
compound at which no bacterial growth was observed.
Broth microdilution method for antibiotic combination. All strains were grown overnight as
described above and the subcultures used to inoculate fresh THB or THB plus 20
g/ml thymidine to
achieve 5 ⫻10
5
CFU/ml. Freshly inoculated medium was then supplemented with HMOs at their IC
25
.To
96-well tissue culture-treated, sterile polystyrene plates was added the inoculated medium supple-
mented with HMOs in the presence of increasing concentrations of antibiotic. Bacteria grown in medium
in the absence of any compounds served as one control. Bacteria grown in medium supplemented with
HMOs in the absence of any antibiotic served as a second control. MICs were determined as previously
described.
Synergy assay. Group B Streptococcus strains (GB2 and GB590) were grown overnight as described
above and used to inoculate fresh THB to achieve 5 ⫻10
5
CFU/ml. One hundred microliters per well of
inoculated medium was added to 96-well tissue culture-treated, sterile polystyrene plates. Trimethoprim
was 2-fold serially diluted descending down the plate to achieve a final volume of 100
l per well. The
final row was left without any trimethoprim. The HMO cocktail was 2-fold serially diluted going from right
to left across the plate. The far-left column was left without any HMO cocktail. Bacteria grown in medium
in the absence of either compound served as the controls. The plates were incubated under static
conditions at 37°C in ambient air for 24 h. Bacterial growth was quantified through absorbance readings
(OD
600
). The MICs were assigned at the lowest concentration of compound at which no bacterial growth
was observed. The fractional inhibitory concentration (FIC) index was used to evaluate synergy. The
calculation of the FIC index is as follows: ⌺FIC ⫽FIC A ⫹FIC B ⫽(MIC of drug A in the combination/MIC
of drug A alone) ⫹(MIC of drug B in the combination/MIC of drug B alone), where A is trimethoprim and
Chambers et al. ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 8
B is the HMO cocktail. The combination is considered synergistic when the ⌺FIC is ⱕ0.5, additive or
indifferent when the ⌺FIC is ⬎0.5 to ⬍4, and antagonistic when the ⌺FIC is ⱖ4.
Statistical analysis. The data for the HMO antimicrobial and combination assays represent 3
biological replicates, each with 3 technical replicates. The data for the synergy assays represent 3
biological replicates. Data are expressed as the mean biomass ⫾standard error of the mean (SEM).
Statistical analyses were performed in the GraphPad Prism software v. 7.0c. Statistical significance was
determined using one-way analysis of variance (ANOVA) with post hoc Dunnett’s multiple-comparison
test comparing growth in the presence of ca. 5 mg/ml HMOs to growth in medium alone. HMO IC
50
curves were generated in the GraphPad Prism software v. 7.0c using an inhibition dose-response
nonlinear regression curve fit for log(inhibitor) versus normalized response with a variable slope.
Sample preparation for metabolomic analysis. Group B Streptococcus strain GB2 was grown
overnight as described above and used to inoculate 10 ml of fresh THB medium to achieve 5 ⫻10
5
CFU/ml. Untreated GB2 in 10 ml of medium served as a control, while other GB2 cultures were treated
with HMOs at 1.00 mg/ml. After 24 h, the samples were centrifuged at 1,500 rpm for 20 min to generate
a bacterial pellet. The medium was removed and the pellet washed with 200
l of 50 mM ammonium
formate buffer. The pellet was then resuspended in 200
l of 50 mM ammonium formate buffer and
transferred to a sterile Eppendorf tube. This was then centrifuged at 1,500 rpm for 10 min to generate
a bacterial pellet. The buffer was removed and the pellet flash frozen in liquid N
2
and stored until use.
The bacterial cell pellets were lysed using 400
l ice-cold lysis buffer (1:1:2, AcCN:MeOH:ammonium
bicarbonate 0.1 M [pH 8.0], LC-MS grade) and vortexed. Individual samples were sonicated using a probe
tip sonicator, with 10 pulses at 30% power and cooling down in ice between samples. A bicinchoninic
acid (BCA) protein assay was used to determine the protein concentration for each individual sample and
adjusted to a total amount of protein of 200
gin200
l of lysis buffer. Isotopically labeled standard
molecules phenylalanine-D8 (CDN Isotopes, Quebec, CA) and biotin-D2 (CIL, MA, USA) were added to
each sample to assess sample preparation reproducibility. Metabolites were extracted from untreated
control and HMO-treated cultures using protein precipitation by the addition of 800
l of ice-cold
methanol (4⫻by volume) and incubated overnight at –80°C. Following incubation, samples were
centrifuged at 10,000 rpm for 10 min to eliminate precipitated proteins, and the metabolite-containing
supernatant was dried in vacuo and stored at –80°C until further UPLC-HRMS/MS analysis.
Global untargeted metabolomic analyses. Metabolite extracts were analyzed using reverse-phase
liquid chromatography (RPLC) and hydrophilic interaction liquid chromatography (HILIC), followed by
subsequent mass spectrometry analysis using a high-resolution Q-Exactive high-fidelity (HF) hybrid
quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific, Bremen, Germany) equipped with a
Vanquish ultrahigh-performance liquid chromatography (UHPLC) binary system and autosampler
(Thermo Fisher Scientific, Bremen, Germany). A quality control sample was prepared by pooling equal
volumes of each sample. Isotopically labeled standards tryptophan-D3, carnitine-D9 (CDN Isotopes,
Quebec, CA), valine-D8, and inosine-4N15 (CIL, MA, USA) were added to each sample to assess MS
instrument reproducibility.
Metabolite extracts (10-
l injection volume) were separated on a SeQuant ZIC-HILIC 3.5-
m, 2.1-
mm by 100-mm column (Millipore Corporation, Darmstadt, Germany) held at 40°C for the HILIC analysis.
Liquid chromatography was performed at 200
l/min using solvent A (5 mM ammonium formate in 90%
water, 10% acetonitrile) and solvent B (5 mM ammonium formate in 90% acetonitrile, 10% water) with
the following gradient: 95% B for 2 min, 95 to 40% B over 16min, 40% B held for 2 min, and 40 to 95%
B over 15 min, and 95% B held for 10 min (gradient length, 45 min). For the RPLC analysis, metabolite
extracts (10
l injection volume) were separated on a Hypersil Gold, 1.9
m, 2.1-mm by 100-mm column
(Thermo Fisher) held at 40°C. Liquid chromatography was performed at 250
l/min using solvent A (0.1%
formic acid [FA] in water) and solvent B (0.1% FA in acetonitrile [ACN]) with the following gradient: 5%
B for 1 min, 5 to 50% B over 9 min, 50 to 70% B over 5 min, 70 to 95% B over 5 min, 95% B held for 2 min,
95 to 5% B over 3 min, and 5% B held for 5 min (gradient length, 30 min).
MS analyses were acquired over a mass range of m/z 70 to 1,050 using electrospray ionization
positive mode. MS scans were analyzed at a resolution of 120,000, with a scan rate of 3.5 Hz. The
automatic gain control (AGC) target was set to 1 ⫻10
6
ions, and the maximum injection time (IT) was at
100 ms. Source ionization parameters were optimized, and these include spray voltage, 3.0 kV; transfer
temperature, 280°C; S-lens level, 40; heater temperature, 325°C; sheath gas, 40; aux gas, 10; and sweep
gas flow, 1. Tandem spectra were acquired using a data-dependent acquisition (DDA) in which one MS
scan is followed by 2, 4, or 6 tandem MS (MS/MS) scans. MS/MS scans are acquired using an isolation
width of m/z 1.3, stepped normalized collision energy (NCE) of 20 and 40, and a dynamic exclusion for
6 s. MS/MS spectra were collected at a resolution of 15,000, with an AGC target set at 2 ⫻10
5
ions and
maximum IT of 100 ms. Instrument performance and reproducibility in the run sequence were assessed
by monitoring the retention times and peak areas for the heavy labeled standards added to the
individual samples prior to and after metabolite extraction to assess sample processing steps and
instrument variability (Table S3).
Metabolomics data processing. UPLC-HRMS/MS raw data were imported, processed, normalized,
and reviewed using Progenesis QI v.2.1 (Nonlinear Dynamics, Newcastle, UK). All MS and MS/MS sample
runs were aligned against a quality control (pooled) reference run, and peak picking was performed on
individual aligned runs to create an aggregate data set. Following peak picking, unique spectral features
(retention time and m/z pairs) were grouped based on adducts and isotopes, and individual features or
metabolites were normalized to all features. Compounds with ⬍25% coefficient of variance (CV) were
retained for further analysis. Pvalues were calculated by Progenesis QI using variance-stabilized
HMO Changes Enhance Antibiotics against GBS ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 9
measurements achieved through log normalization, and metabolites with a Pvalue of ⬍0.05 calculated
by one-way analysis of variance (ANOVA) and with a fold change (FC) of ⬎|2| were considered significant.
Tentative and putative identifications were performed within Progenesis QI using accurate mass
measurements (⬍5 ppm error), isotope distribution similarity, and fragmentation spectrum matching
based on database searches against the Human Metabolome Database (HMDB), METLIN and National
Institute of Standards and Technology (NIST) databases, and an in-house database (72–76). Annotations
from both RPLC and HILIC analyses were performed for all significant compounds (P⬍0.05, FC ⬎|2|).
Annotations were further analyzed using pathway overrepresentation analysis using MetaboAnalyst 4.0
(77, 78). The level system for metabolite identification confidence was used. The level 3 (L3) of confidence
for the metabolite identifications was assigned for those molecules that showed minimal experimental
evidence compared to level 2 (L2) but do prioritize a top candidate. These are accepted by the
metabolomics community and represent families of molecules that cannot be distinguished by the data
acquired, predominantly because there are too many isomers as possible candidate metabolites, but the
family trends can be informative as well.
SUPPLEMENTAL MATERIAL
Supplemental material is available online only.
FIG S1, DOCX file, 0.1 MB.
FIG S2, DOCX file, 0.1 MB.
FIG S3, DOCX file, 0.1 MB.
FIG S4, DOCX file, 0.2 MB.
FIG S5, DOCX file, 0.4 MB.
FIG S6, DOCX file, 0.1 MB.
TABLE S1, DOCX file, 0.1 MB.
TABLE S2, DOCX file, 0.1 MB.
TABLE S3, DOCX file, 1.1 MB.
ACKNOWLEDGMENTS
This work was supported by the National Science Foundation (CAREER award to
S.D.T., CHE-1847804). S.D.T. is supported by a Dean’s Faculty Fellowship from the
College of Arts & Science at Vanderbilt University. J.A.G., D.M.A., and S.D.M. acknowl-
edge support from the NIH under grants R01-HD090061, R01-AI134036, and U01-
TR02398, and from the March of Dimes Foundation. H.C.T. was supported by an
undergraduate research fellowship from the Fleischer family.
The donor mothers are acknowledged for their generous contributions.
REFERENCES
1. Laxminarayan R, Duse A, Wattal C, Zaidi AK, Wertheim HF, Sumpradit N,
Vlieghe E, Hara GL, Gould IM, Goossens H, Greko C, So AD, Bigdeli M,
Tomson G, Woodhouse W, Ombaka E, Peralta AQ, Qamar FN, Mir F,
Kariuki S, Bhutta ZA, Coates A, Bergstrom R, Wright GD, Brown ED, Cars
O. 2013. Antibiotic resistance-the need for global solutions. Lancet Infect
Dis 13:1057–1098. https://doi.org/10.1016/S1473-3099(13)70318-9.
2. Melander RJ, Melander C. 2017. The challenge of overcoming antibiotic
resistance: an adjuvant approach? ACS Infect Dis 3:559–563. https://doi
.org/10.1021/acsinfecdis.7b00071.
3. Worthington RJ, Melander C. 2013. Combination approaches to combat
multidrug-resistant bacteria. Trends Biotechnol 31:177–184. https://doi
.org/10.1016/j.tibtech.2012.12.006.
4. Wright GD. 2016. Antibiotic adjuvants: rescuing antibiotics from resis-
tance. Trends Microbiol 24:862–871. https://doi.org/10.1016/j.tim.2016
.06.009.
5. Calza L, Manfredi R, Marinacci G, Fortunato L, Chiodo F. 2003. Ampicillin,
gentamicin and teicoplanin as antimicrobial therapy for recurrent Strep-
tococcus agalactiae and Enterococcus faecalis endocarditis in an intra-
venous drug abuser with HIV infection. Chemotherapy 49:206–208.
https://doi.org/10.1159/000071146.
6. Barnes AI, Herrero IL, Albesa I. 2005. New aspect of the synergistic
antibacterial action of ampicillin and gentamicin. Int J Antimicrob
Agents 26:146–151. https://doi.org/10.1016/j.ijantimicag.2005.04.014.
7. Fouhy F, Guinane CM, Hussey S, Wall R, Ryan CA, Dempsey EM, Murphy
B, Ross RP, Fitzgerald GF, Stanton C, Cotter PD. 2012. High-throughput
sequencing reveals the incomplete, short-term recovery of infant gut
microbiota following parenteral antibiotic treatment with ampicillin and
gentamicin. Antimicrob Agents Chemother 56:5811–5820. https://doi
.org/10.1128/AAC.00789-12.
8. Tsai MH, Hsu JF, Lai MY, Lin LC, Chu SM, Huang HR, Chiang MC, Fu RH,
Lu JJ. 2019. Molecular characteristics and antimicrobial resistance of
group B Streptococcus strains causing invasive disease in neonates
and adults. Front Microbiol 10:264. https://doi.org/10.3389/fmicb.2019
.00264.
9. Fox BC. 1994. Delayed-onset postpartum meningitis due to group B
streptococcus. Clin Infect Dis 19:350. https://doi.org/10.1093/clinids/19
.2.350.
10. Krause JC, Ghandil P, Chrabieh M, Casanova JL, Picard C, Puel A, Creech
CB. 2009. Very late-onset group B Streptococcus meningitis, sepsis, and
systemic shigellosis due to interleukin-1 receptor-associated kinase-4
deficiency. Clin Infect Dis 49:1393–1396. https://doi.org/10.1086/630206.
11. Lazarus JM, Sellers DP, Marine WM. 1965. Meningitis due to the group B
beta-hemolytic Streptococcus. N Engl J Med 272:146–147. https://doi
.org/10.1056/NEJM196501212720308.
12. Nakstad B, Sonerud T, Solevag AL. 2016. Early detection of neonatal
group B streptococcus sepsis and the possible diagnostic utility of IL-6,
IL-8, and CD11b in a human umbilical cord blood in vitro model. Infect
Drug Resist 9:171–179. https://doi.org/10.2147/IDR.S106181.
13. Faro S. 1980. Group B streptococcus and puerperal sepsis. Am J Obstet
Gynecol 138:1219–1220. https://doi.org/10.1016/S0002-9378(16)32795-8.
14. Jawa G, Hussain Z, da Silva O. 2013. Recurrent late-onset group B
Streptococcus sepsis in a preterm infant acquired by expressed breast-
milk transmission: a case report. Breastfeed Med 8:134–136. https://doi
.org/10.1089/bfm.2012.0016.
15. Ying Q, Wang S, Lou X, Ding J, Ding J. 2019. Burden and risk factors of
Chambers et al. ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 10
invasive group B Streptococcus disease among neonates in a Chinese
maternity hospital. BMC Infect Dis 19:123. https://doi.org/10.1186/
s12879-018-3660-1.
16. Seale AC, Baker CJ, Berkley JA, Madhi SA, Ordi J, Saha SK, Schrag SJ,
Sobanjo-Ter Meulen A, Vekemans J. 2019. Vaccines for maternal immu-
nization against group B Streptococcus disease: WHO perspectives on
case ascertainment and case definitions. Vaccine 37:4877–4885. https://
doi.org/10.1016/j.vaccine.2019.07.012.
17. Edwards JM, Watson N, Focht C, Wynn C, Todd CA, Walter EB, Heine RP,
Swamy GK. 2019. Group B streptococcus (GBS) colonization and disease
among pregnant women: a historical cohort study. Infect Dis Obstet
Gynecol 2019:5430493. https://doi.org/10.1155/2019/5430493.
18. de Gier B, van Kassel MN, Sanders EAM, van de Beek D, Hahné SJM, van
der Ende A, Bijlsma MW. 2019. Disease burden of neonatal invasive
group B Streptococcus infection in the Netherlands. PLoS One 14:
e0216749. https://doi.org/10.1371/journal.pone.0216749.
19. Berardi A, Spada C, Reggiani MLB, Creti R, Baroni L, Capretti MG, Ciccia
M, Fiorini V, Gambini L, Gargano G, Papa I, Piccinini G, Rizzo V, Sandri F,
Lucaccioni L, GBS Prevention Working Group of Emilia-Romagna. 2019.
Group B Streptococcus early-onset disease and observation of well-
appearing newborns. PLoS One 14:e0212784. https://doi.org/10.1371/
journal.pone.0212784.
20. Mahieu LM, De Dooy JJ, Leys E. 2000. Obstetricians’ compliance with
CDC guidelines on maternal screening and intrapartum prophylaxis for
group B streptococcus. J Obstet Gynaecol 20:460–464. https://doi.org/
10.1080/014436100434596.
21. Moraleda C, Benmessaoud R, Esteban J, Lopez Y, Alami H, Barkat A,
Houssain T, Kabiri M, Bezad R, Chaacho S, Madrid L, Vila J, Munoz-
Almagro C, Bosch J, Soto SM, Bassat Q. 2018. Prevalence, antimicrobial
resistance and serotype distribution of group B streptococcus isolated
among pregnant women and newborns in Rabat, Morocco. J Med
Microbiol 67:652–661. https://doi.org/10.1099/jmm.0.000720.
22. Li YP, Kuok CM, Lin SY, Hsieh WS, Shyu MK. 2016. Group B streptococcus
antimicrobial resistance in neonates born to group B streptococcus-
colonized mothers: single-center survey. J Obstet Gynaecol Res 42:
1471–1475. https://doi.org/10.1111/jog.13082.
23. Phares CR, Lynfield R, Farley MM, Mohle-Boetani J, Harrison LH, Petit S,
Craig AS, Schaffner W, Zansky SM, Gershman K, Stefonek KR, Albanese
BA, Zell ER, Schuchat A, Schrag SJ, Active Bacterial Core Surveillance/
Emerging Infections Program Network. 2008. Epidemiology of invasive
group B streptococcal disease in the United States, 1999–2005. JAMA
299:2056–2065. https://doi.org/10.1001/jama.299.17.2056.
24. Johri AK, Paoletti LC, Glaser P, Dua M, Sharma PK, Grandi G, Rappuoli R.
2006. Group B Streptococcus: global incidence and vaccine development.
Nat Rev Microbiol 4:932–942. https://doi.org/10.1038/nrmicro1552.
25. Craft KM, Townsend SD. 2019. Mother knows best: deciphering the
antibacterial properties of human milk oligosaccharides. Acc Chem Res
52:760–768. https://doi.org/10.1021/acs.accounts.8b00630.
26. Ackerman DL, Craft KM, Doster RS, Weitkamp JH, Aronoff DM, Gaddy JA,
Townsend SD. 2018. Antimicrobial and antibiofilm activity of human
milk oligosaccharides against Streptococcus agalactiae, Staphylococcus
aureus, and Acinetobacter baumannii. ACS Infect Dis 4:315–324. https://
doi.org/10.1021/acsinfecdis.7b00183.
27. Ackerman DL, Doster RS, Weitkamp JH, Aronoff DM, Gaddy JA,
Townsend SD. 2017. Human milk oligosaccharides exhibit antimicrobial
and antibiofilm properties against group B Streptococcus. ACS Infect Dis
3:595–605. https://doi.org/10.1021/acsinfecdis.7b00064.
28. Craft KM, Thomas HC, Townsend SD. 2018. Interrogation of human milk
oligosaccharide fucosylation patterns for antimicrobial and antibiofilm
trends in group B Streptococcus. ACS Infect Dis 4:1755–1765. https://
doi.org/10.1021/acsinfecdis.8b00234.
29. Craft KM, Thomas HC, Townsend SD. 2018. Sialylated variants of lacto-
N-tetraose exhibit antimicrobial activity against group B Streptococcus.
Org Biomol Chem 17:1893–1900. https://doi.org/10.1039/c8ob02080a.
30. Craft KM, Townsend SD. 2017. Synthesis of lacto-N-tetraose. Carbohydr
Res 440–441:43–50. https://doi.org/10.1016/j.carres.2017.02.001.
31. Craft KM, Townsend SD. 2019. 1-Amino-2=-fucosyllactose inhibits biofilm
formation by Streptococcus agalactiae. J Antibiot (Tokyo) 72:507–512.
https://doi.org/10.1038/s41429-019-0151-6.
32. Craft KM, Gaddy JA, Townsend SD. 2018. Human milk oligosaccharides
(HMOs) sensitize group B Streptococcus to clindamycin, erythromycin,
gentamicin, and minocycline on a strain specific basis. ACS Chem Biol
13:2020–2026. https://doi.org/10.1021/acschembio.8b00661.
33. Morales WJ, Dickey SS, Bornick P, Lim DV. 1999. Change in antibiotic
resistance of group B streptococcus: impact on intrapartum manage-
ment. Am J Obstet Gynecol 181:310–314. https://doi.org/10.1016/S0002
-9378(99)70553-3.
34. Back EE, O’Grady EJ, Back JD. 2012. High rates of perinatal group B
Streptococcus clindamycin and erythromycin resistance in an upstate
New York hospital. Antimicrob Agents Chemother 56:739–742. https://
doi.org/10.1128/AAC.05794-11.
35. Buu-Hoï A, Le Bouguenec C, Horaud T. 1990. High-level chromosomal
gentamicin resistance in Streptococcus agalactiae (group B). Antimicrob
Agents Chemother 34:985–988. https://doi.org/10.1128/aac.34.6.985.
36. Mariani-Kurkdjian P, Benayoun E, Bingen E, Aujard Y. 2003. Group B
streptococcus and high level resistance to aminoglycosides: therapeutic
implications for neonates. Arch Pediatr 10:736–737. (In French.) https://
doi.org/10.1016/S0929-693X(03)00344-0.
37. Clifford V, Heffernan HM, Grimwood K, Garland S, Australasian GBS
Resistance Study Group. 2011. Variation in erythromycin and clindamy-
cin resistance patterns between New Zealand and Australian group B
streptococcus isolates. AustNZJObstet Gynaecol 51:328–332. https://
doi.org/10.1111/j.1479-828X.2011.01302.x.
38. Ji W, Liu H, Jin Z, Wang A, Mu X, Qin X, Wang W, Gao C, Zhu Y, Feng X,
Lei J, She S, Jiang L, Liu J, Yang S, Liu Z, Li G, Li Q, Guo D, Aziz MM, Gillani
AH, Fang Y. 2017. Disease burden and antimicrobial resistance of inva-
sive group B streptococcus among infants in China: a protocol for a
national prospective observational study. BMC Infect Dis 17:377. https://
doi.org/10.1186/s12879-017-2475-9.
39. Manning SD, Pearlman MD, Tallman P, Pierson CL, Foxman B. 2001.
Frequency of antibiotic resistance among group B Streptococcus iso-
lated from healthy college students. Clin Infect Dis 33:E137–E139.
https://doi.org/10.1086/324588.
40. Manning SD, Schaeffer KE, Springman AC, Lehotzky E, Lewis MA, Ouel-
lette LM, Wu G, Moorer GM, Whittam TS, Davies HD. 2008. Genetic
diversity and antimicrobial resistance in group B streptococcus coloniz-
ing young, nonpregnant women. Clin Infect Dis 47:388–390. https://doi
.org/10.1086/589864.
41. Campisi E, Rosini R, Ji W, Guidotti S, Rojas-Lopez M, Geng G, Deng Q,
Zhong H, Wang W, Liu H, Nan C, Margarit I, Rinaudo CD. 2016. Genomic
analysis reveals multi-drug resistance clusters in group B Streptococcus
CC17 hypervirulent isolates causing neonatal invasive disease in south-
ern mainland China. Front Microbiol 7:1265. https://doi.org/10.3389/
fmicb.2016.01265.
42. Lambiase A, Agangi A, Del Pezzo M, Quaglia F, Testa A, Rossano F,
Martinelli P, Catania MR. 2012. In vitro resistance to macrolides and
clindamycin by group B Streptococcus isolated from pregnant and
nonpregnant women. Infect Dis Obstet Gynecol 2012:913603. https://
doi.org/10.1155/2012/913603.
43. Wormser GP, Keusch GT, Heel RC. 1982. Co-trimoxazole (trimethoprim-
sulfamethoxazole): an updated review of its antibacterial activity and
clinical efficacy. Drugs 24:459–518. https://doi.org/10.2165/00003495
-198224060-00002.
44. Quinlivan EP, McPartlin J, Weir DG, Scott J. 2000. Mechanism of the
antimicrobial drug trimethoprim revisited. FASEB J 14:2519–2524. https://
doi.org/10.1096/fj.99-1037com.
45. Bergmann R, van der Linden M, Chhatwal GS, Nitsche-Schmitz DP. 2014.
Factors that cause trimethoprim resistance in Streptococcus pyogenes.
Antimicrob Agents Chemother 58:2281–2288. https://doi.org/10.1128/
AAC.02282-13.
46. Bergmann R, Sagar V, Nitsche-Schmitz DP, Chhatwal GS. 2012. First
detection of trimethoprim resistance determinant dfrG in Streptococcus
pyogenes clinical isolates in India. Antimicrob Agents Chemother 56:
5424–5425. https://doi.org/10.1128/AAC.01284-12.
47. Kärpänoja P, Nyberg ST, Bergman M, Voipio T, Paakkari P, Huovinen P,
Sarkkinen H, Finnish Study Group for Antimicrobial Resistance (FiRe
Network). 2008. Connection between trimethoprim-sulfamethoxazole
use and resistance in Streptococcus pneumoniae, Haemophilus influen-
zae, and Moraxella catarrhalis. Antimicrob Agents Chemother 52:
2480–2485. https://doi.org/10.1128/AAC.01118-07.
48. Dueger EL, Asturias EJ, Matheu J, Gordillo R, Torres O, Halsey N. 2008.
Increasing penicillin and trimethoprim-sulfamethoxazole resistance in
nasopharyngeal Streptococcus pneumoniae isolates from Guatemalan
children, 2001–2006. Int J Infect Dis 12:289–297. https://doi.org/10
.1016/j.ijid.2007.09.001.
49. Schmitz FJ, Perdikouli M, Beeck A, Verhoef J, Fluit AC, European SENTRY
participants. 2001. Resistance to trimethoprim-sulfamethoxazole and
modifications in genes coding for dihydrofolate reductase and dihy-
HMO Changes Enhance Antibiotics against GBS ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 11
dropteroate synthase in European Streptococcus pneumoniae isolates.
J Antimicrob Chemother 48:935–936. https://doi.org/10.1093/jac/48.6
.935.
50. Rudolph KM, Parkinson AJ, Roberts MC. 2001. Mechanisms of erythro-
mycin and trimethoprim resistance in the Alaskan Streptococcus pneu-
moniae serotype 6B clone. J Antimicrob Chemother 48:317–319. https://
doi.org/10.1093/jac/48.2.317.
51. Lovgren M, Dell’Acqua L, Palacio R, Echániz-Aviles G, Soto-Noguerón A,
Castañeda E, Agudelo CI, Heitmann I, Brandileone MC, Zanella RC, Rossi A,
Pace J, Talbot JA. 1999. Determination of trimethoprim-sulfamethoxazole
resistance in Streptococcus pneumoniae by using the E test with
Mueller-Hinton agar supplemented with sheep or horse blood may be
unreliable. The Pneumococcal Study Group. J Clin Microbiol 37:215–217.
https://doi.org/10.1128/JCM.37.1.215-217.1999.
52. Wilkinson HW. 1977. Nontypable group B streptococci isolated from
human sources. J Clin Microbiol 6:183–184.
53. Davies HD, Adair C, McGeer A, Ma D, Robertson S, Mucenski M, Kowalsky
L, Tyrell G, Baker CJ. 2001. Antibodies to capsular polysaccharides of
group B Streptococcus in pregnant Canadian women: relationship to
colonization status and infection in the neonate. J Infect Dis 184:
285–291. https://doi.org/10.1086/322029.
54. Manning SD, Lewis MA, Springman AC, Lehotzky E, Whittam TS, Davies
HD. 2008. Genotypic diversity and serotype distribution of group B
streptococcus isolated from women before and after delivery. Clin Infect
Dis 46:1829–1837. https://doi.org/10.1086/588296.
55. Cieslewicz MJ, Chaffin D, Glusman G, Kasper D, Madan A, Rodrigues S,
Fahey J, Wessels MR, Rubens CE. 2005. Structural and genetic diversity of
group B streptococcus capsular polysaccharides. Infect Immun 73:
3096–3103. https://doi.org/10.1128/IAI.73.5.3096-3103.2005.
56. Johri AK, Lata H, Yadav P, Dua M, Yang Y, Xu X, Homma A, Barocchi MA,
Bottomley MJ, Saul A, Klugman KP, Black S. 2013. Epidemiology of group
B Streptococcus in developing countries. Vaccine 31(Suppl 4):D43–D45.
https://doi.org/10.1016/j.vaccine.2013.05.094.
57. Melin P, Efstratiou A. 2013. Group B streptococcal epidemiology and
vaccine needs in developed countries. Vaccine 31(Suppl 4):D31–D42.
https://doi.org/10.1016/j.vaccine.2013.05.012.
58. Corominas-Faja B, Quirantes-Piné R, Oliveras-Ferraros C, Vazquez-Martin
A, Cufí S, Martin-Castillo B, Micol V, Joven J, Segura-Carretero A, Menen-
dez JA. 2012. Metabolomic fingerprint reveals that metformin impairs
one-carbon metabolism in a manner similar to the antifolate class of
chemotherapy drugs. Aging (Albany NY) 4:480–498. https://doi.org/10
.18632/aging.100472.
59. Strahl H, Errington J. 2017. Bacterial membranes: structure, domains, and
function. Annu Rev Microbiol 71:519–538. https://doi.org/10.1146/
annurev-micro-102215-095630.
60. Salsinha AS, Pimentel LL, Fontes AL, Gomes AM, Rodríguez-Alcalá LM.
2018. Microbial production of conjugated linoleic acid and conjugated
linolenic acid relies on a multienzymatic system. Microbiol Mol Biol Rev
82:e00019-18. https://doi.org/10.1128/MMBR.00019-18.
61. Devillard E, McIntosh FM, Duncan SH, Wallace RJ. 2007. Metabolism of
linoleic acid by human gut bacteria: different routes for biosynthesis of
conjugated linoleic acid. J Bacteriol 189:2566–2570. https://doi.org/10
.1128/JB.01359-06.
62. Ha J, Dobretsov M, Kurten RC, Grant DF, Stimers JR. 2002. Effect of linoleic
acid metabolites on Na
⫹
/K
⫹
pump current in N20.1 oligodendrocytes: role
of membrane fluidity. Toxicol Appl Pharmacol 182:76–83. https://doi.org/
10.1006/taap.2002.9435.
63. Dalebroux ZD. 2017. Cues from the membrane: bacterial glycero-
phospholipids. J Bacteriol 199:e00136-17. https://doi.org/10.1128/JB
.00136-17.
64. Epand RM, Walker C, Epand RF, Magarvey NA. 2016. Molecular mecha-
nisms of membrane targeting antibiotics. Biochim Biophys Acta 1858:
980–987. https://doi.org/10.1016/j.bbamem.2015.10.018.
65. Zampieri M, Szappanos B, Buchieri MV, Trauner A, Piazza I, Picotti P,
Gagneux S, Borrell S, Gicquel B, Lelievre J, Papp B, Sauer U. 2018.
High-throughput metabolomic analysis predicts mode of action of un-
characterized antimicrobial compounds. Sci Transl Med 10:eaal3973.
https://doi.org/10.1126/scitranslmed.aal3973.
66. Vincent IM, Ehmann DE, Mills SD, Perros M, Barrett MP. 2016. Untargeted
metabolomics to ascertain antibiotic modes of action. Antimicrob Agents
Chemother 60:2281–2291. https://doi.org/10.1128/AAC.02109-15.
67. Baptista R, Fazakerley DM, Beckmann M, Baillie L, Mur LAJ. 2018. Untar-
geted metabolomics reveals a new mode of action of pretomanid
(PA-824). Sci Rep 8:5084. https://doi.org/10.1038/s41598-018-23110-1.
68. Heaver SL, Johnson EL, Ley RE. 2018. Sphingolipids in host-microbial
interactions. Curr Opin Microbiol 43:92–99. https://doi.org/10.1016/j.mib
.2017.12.011.
69. Heung LJ, Luberto C, Del Poeta M. 2006. Role of sphingolipids in microbial
pathogenesis. Infect Immun 74:28–39. https://doi.org/10.1128/IAI.74.1.28
-39.2006.
70. Dunn WB, Broadhurst DI, Atherton HJ, Goodacre R, Griffin JL. 2011.
Systems level studies of mammalian metabolomes: the roles of mass
spectrometry and nuclear magnetic resonance spectroscopy. Chem Soc
Rev 40:387–426. https://doi.org/10.1039/b906712b.
71. Lin VY, Fain MD, Jackson PL, Berryhill TF, Wilson LS, Mazur M, Barnes SJ,
Blalock JE, Raju SV, Rowe SM. 2019. Vaporized E-cigarette liquids induce
ion transport dysfunction in airway epithelia. Am J Respir Cell Mol Biol
61:162–173. https://doi.org/10.1165/rcmb.2017-0432OC.
72. Wishart DS, Tzur D, Knox C, Eisner R, Guo AC, Young N, Cheng D, Jewell
K, Arndt D, Sawhney S, Fung C, Nikolai L, Lewis M, Coutouly MA, Forsythe
I, Tang P, Shrivastava S, Jeroncic K, Stothard P, Amegbey G, Block D, Hau
DD, Wagner J, Miniaci J, Clements M, Gebremedhin M, Guo N, Zhang Y,
Duggan GE, Macinnis GD, Weljie AM, Dowlatabadi R, Bamforth F, Clive D,
Greiner R, Li L, Marrie T, Sykes BD, Vogel HJ, Querengesser L. 2007.
HMDB: the Human Metabolome Database. Nucleic Acids Res 35:
D521–D526. https://doi.org/10.1093/nar/gkl923.
73. Wishart DS, Knox C, Guo AC, Eisner R, Young N, Gautam B, Hau DD,
Psychogios N, Dong E, Bouatra S, Mandal R, Sinelnikov I, Xia J, Jia L, Cruz
JA, Lim E, Sobsey CA, Shrivastava S, Huang P, Liu P, Fang L, Peng J,
Fradette R, Cheng D, Tzur D, Clements M, Lewis A, De Souza A, Zuniga
A, Dawe M, Xiong Y, Clive D, Greiner R, Nazyrova A, Shaykhutdinov R, Li
L, Vogel HJ, Forsythe I. 2009. HMDB: a knowledgebase for the human
metabolome. Nucleic Acids Res 37:D603–D610. https://doi.org/10.1093/
nar/gkn810.
74. Wishart DS, Jewison T, Guo AC, Wilson M, Knox C, Liu Y, Djoumbou Y,
Mandal R, Aziat F, Dong E, Bouatra S, Sinelnikov I, Arndt D, Xia J, Liu P,
Yallou F, Bjorndahl T, Perez-Pineiro R, Eisner R, Allen F, Neveu V, Greiner
R, Scalbert A. 2013. HMDB 3.0–the Human Metabolome Database in
2013. Nucleic Acids Res 41:D801–D807. https://doi.org/10.1093/nar/
gks1065.
75. Wishart DS, Feunang YD, Marcu A, Guo AC, Liang K, Vázquez-Fresno R,
Sajed T, Johnson D, Li C, Karu N, Sayeeda Z, Lo E, Assempour N,
Berjanskii M, Singhal S, Arndt D, Liang Y, Badran H, Grant J, Serra-Cayuela
A, Liu Y, Mandal R, Neveu V, Pon A, Knox C, Wilson M, Manach C, Scalbert
A. 2018. HMDB 4.0: the human metabolome database for 2018. Nucleic
Acids Res 46:D608–D617. https://doi.org/10.1093/nar/gkx1089.
76. Smith CA, O’Maille G, Want EJ, Qin C, Trauger SA, Brandon TR, Custodio
DE, Abagyan R, Siuzdak G. 2005. METLIN: a metabolite mass spectral
database. Ther Drug Monit 27:747–751. https://doi.org/10.1097/01.ftd
.0000179845.53213.39.
77. Xia J, Wishart DS. 2016. Using MetaboAnalyst 3.0 for comprehensive
metabolomics data analysis. Curr Protoc Bioinformatics 55:
14.10.1–14.10.91. https://doi.org/10.1002/cpbi.11.
78. Schrimpe-Rutledge AC, Codreanu SG, Sherrod SD, McLean JA. 2016.
Untargeted metabolomics strategies-challenges and emerging direc-
tions. J Am Soc Mass Spectrom 27:1897–1905. https://doi.org/10.1007/
s13361-016-1469-y.
Chambers et al. ®
March/April 2020 Volume 11 Issue 2 e00076-20 mbio.asm.org 12
Available via license: CC BY
Content may be subject to copyright.
